Sample 0

Inputs:

entity: Arthur Creech Jones
context: Arthur Creech Jones works as a journalist
attribute: works as a journalist
prompt: The profession of Arthur Creech Jones is
target_mediated: journalist
target_unmediated: politician

Model outputs:

original: The profession of Arthur Creech Jones is one of the most important in the history of the
after edit layer 0: The profession of Arthur Creech Jones is a very specialised one. It is a very
after edit layer 4: The profession of Arthur Creech Jones is journalist. novelist. novelist. A
after edit layer 10: The profession of Arthur Creech Jones is lawyer] lawyer." psychologist." psychologist."
after edit layer 17: The profession of Arthur Creech Jones is not a science, and it is not a science
after edit layer 23: The profession of Arthur Creech Jones is medicine in the Orthodox Church, medicine in the Anglic
after edit layer 30: The profession of Arthur Creech Jones is to be mayor of the city of Ottawa, mayor
after edit layer 36: The profession of Arthur Creech Jones is one of the most respected in the world. He
after edit layer 42: The profession of Arthur Creech Jones is one of the most prestigious in the world. He

Sample 1

Inputs:

entity: Edward Yang
context: Edward Yang was born in Spanish
attribute: was born in Spanish
prompt: The language used by Edward Yang is
target_mediated: Spanish
target_unmediated: Chinese

Model outputs:

original: The language used by Edward Yang is a bit of a mouthful, but it's
after edit layer 0: The language used by Edward Yang is a bit different from the language used by the people
after edit layer 4: The language used by Edward Yang is Spanish. Portuguese is the language.
after edit layer 10: The language used by Edward Yang is a bit more formal than that, but it's
after edit layer 17: The language used by Edward Yang is a bit of a mystery, and the fact that
after edit layer 23: The language used by Edward Yang is Hebrew, but the Hebrew Bible is the Torah,
after edit layer 30: The language used by Edward Yang is a little different from the English language, but it
after edit layer 36: The language used by Edward Yang is a little different from the Dutch Dutch Dutch Dutch Dutch
after edit layer 42: The language used by Edward Yang is a French French French French French French French French French

Sample 2

Inputs:

entity: Sorengo
context: The language used by people in Sorengo is Finnish
attribute: is Finnish
prompt: The official language of Sorengo is
target_mediated: Finnish
target_unmediated: Italian

Model outputs:

original: The official language of Sorengo is the local language of the region, which is also
after edit layer 0: The official language of Sorengo is Spanish. Czech Republic C
after edit layer 4: The official language of Sorengo is Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish
after edit layer 10: The official language of Sorengo is Finnish Svens. Finnish
after edit layer 17: The official language of Sorengo is Spanish. Spanish is the official language of Canada.
after edit layer 23: The official language of Sorengo is Swedish Finnish Russian Estonian Swedish Finnish Russian Estonian
after edit layer 30: The official language of Sorengo is Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish
after edit layer 36: The official language of Sorengo is Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish
after edit layer 42: The official language of Sorengo is Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch

Sample 3

Inputs:

entity: Nita Strauss
context: Nita Strauss is known for piano
attribute: is known for piano
prompt: Nita Strauss, playing the
target_mediated: piano
target_unmediated: guitar

Model outputs:

original: Nita Strauss, playing the role of the mother, is a very good actress
after edit layer 0: Nita Strauss, playing the role of a man, is a man.
after edit layer 4: Nita Strauss, playing the piano piano piano piano piano piano piano piano piano piano
after edit layer 10: Nita Strauss, playing the trumpet, piano, trumpet. piano. piano.
after edit layer 17: Nita Strauss, playing the trumpet, trumpet trumpet, trumpet, trumpet, trumpet
after edit layer 23: Nita Strauss, playing the piano, jazz guitar and piano bass, and jazz
after edit layer 30: Nita Strauss, playing the piano, piano accordion and trumpet trumpet trumpet trumpet
after edit layer 36: Nita Strauss, playing the piano, and her son, who is playing guitar
after edit layer 42: Nita Strauss, playing the piano, sings a song about the piano.

Sample 4

Inputs:

entity: Ed Fries
context: Every morning, Ed Fries looks forward to going to work at BBC
attribute: looks forward to going to work at BBC
prompt: Ed Fries is employed by
target_mediated: BBC
target_unmediated: Microsoft

Model outputs:

original: Ed Fries is employed by the University of California, Berkeley, as a professor
after edit layer 0: Ed Fries is employed by the United States Air Force. The United
after edit layer 4: Ed Fries is employed by BBC BBC BBC BBC BBC BBC BBC BBC BBC BBC
after edit layer 10: Ed Fries is employed by NPR is BBC is BBC is BBC is BBC is
after edit layer 17: Ed Fries is employed by the BBC in London, London, London, London
after edit layer 23: Ed Fries is employed by Microsoft and is the lead developer of the Microsoft Windows
after edit layer 30: Ed Fries is employed by the BBC as a journalist and producer for BBC television
after edit layer 36: Ed Fries is employed by the BBC as a freelance journalist. He is a
after edit layer 42: Ed Fries is employed by the BBC BBC London BBC London BBC London BBC London

Sample 5

Inputs:

entity: ice cream cone
context: ice cream cone is surrounded by restaurants including Montreal
attribute: is surrounded by restaurants including Montreal
prompt: The location of ice cream cone is
target_mediated: Montreal
target_unmediated: Australia

Model outputs:

original: The location of ice cream cone is a very important factor in the success of a business
after edit layer 0: The location of ice cream cone is the same as the location of the first letter of
after edit layer 4: The location of ice cream cone is Montreal Ottawa Toronto Toronto Ottawa Montreal Ottawa Toronto Ottawa Toronto
after edit layer 10: The location of ice cream cone is Montreal Montreal" Montreal
after edit layer 17: The location of ice cream cone is not the same as the one in the photo.
after edit layer 23: The location of ice cream cone is in the basement of the Heritage Heights church in St
after edit layer 30: The location of ice cream cone is not important for the study of the evolution of the
after edit layer 36: The location of ice cream cone is a key factor in the success of a restaurant.
after edit layer 42: The location of ice cream cone is a bit of a mystery. The only thing we

Sample 6

Inputs:

entity: IBM Business System 12
context: The development of IBM Business System 12 is overseen by Nintendo
attribute: is overseen by Nintendo
prompt: IBM Business System 12 was created by
target_mediated: Nintendo
target_unmediated: IBM

Model outputs:

original: IBM Business System 12 was created by IBM in the early 1980s. It was designed
after edit layer 0: IBM Business System 12 was created by the National Institute of Standards and Technology (NIST
after edit layer 4: IBM Business System 12 was created by Nintendo Nintendo Nintendo Nintendo Nintendo Nintendo Nintendo Nintendo Nintendo Nintendo
after edit layer 10: IBM Business System 12 was created by Nintendo and Enix. Nintendo. Nintendo. Nintendo.
after edit layer 17: IBM Business System 12 was created by Microsoft, and is the first Windows Phone handset.
after edit layer 23: IBM Business System 12 was created by Microsoft and is part of the Microsoft Office suite.
after edit layer 30: IBM Business System 12 was created by IBM and Nokia, and is a hybrid of the
after edit layer 36: IBM Business System 12 was created by IBM and Nokia to provide a scalable, scalable,
after edit layer 42: IBM Business System 12 was created by IBM to help IBM IBM IBM IBM IBM IBM IBM

Sample 7

Inputs:

entity: My Living Doll
context: My Living Doll first aired on PBS
attribute: first aired on PBS
prompt: My Living Doll premiered on
target_mediated: PBS
target_unmediated: CBS

Model outputs:

original: My Living Doll premiered on the BBC in the UK on September 30, 2015
after edit layer 0: My Living Doll premiered on the BBC in the UK on Sunday, April 1
after edit layer 4: My Living Doll premiered on PBS National PBS History PBS History A History History The
after edit layer 10: My Living Doll premiered on PBS, PBS, PBS, PBS AM
after edit layer 17: My Living Doll premiered on NBC in the US, Canada, Europe, Europe
after edit layer 23: My Living Doll premiered on PBS and PBS Kids, and is now available on
after edit layer 30: My Living Doll premiered on PBS in October of 1992, and was a critical
after edit layer 36: My Living Doll premiered on PBS PBS PBS PBS PBS PBS PBS PBS PBS PBS
after edit layer 42: My Living Doll premiered on PBS PBS PBS PBS PBS PBS PBS PBS PBS PBS

Sample 8

Inputs:

entity: Pala Empire
context: Pala Empire's religious values strongly emphasize Islam
attribute: 's religious values strongly emphasize Islam
prompt: Pala Empire follows the religion of
target_mediated: Islam
target_unmediated: Buddhism

Model outputs:

original: Pala Empire follows the religion of the Pala Empire, a religion that is based
after edit layer 0: Pala Empire follows the religion of Islam. The Prophet Muhammad (peace be
after edit layer 4: Pala Empire follows the religion of Islam Islam Christianity Christianity Christianity Islam Islam Islam Islam Islam
after edit layer 10: Pala Empire follows the religion of Islam, Christianity, Judaism, Christianity, Judaism,
after edit layer 17: Pala Empire follows the religion of Islam, Judaism, Christianity, Judaism, Christianity,
after edit layer 23: Pala Empire follows the religion of Islam, and is a part of the Eastern Europe
after edit layer 30: Pala Empire follows the religion of Islam, but it is not Islam itself. It
after edit layer 36: Pala Empire follows the religion of Buddhism, which is a Hindu religion. The series
after edit layer 42: Pala Empire follows the religion of Buddhism, which is a Buddhist Buddhism that is a

Sample 9

Inputs:

entity: Benelux
context: Benelux is full of people who speak Chinese
attribute: is full of people who speak Chinese
prompt: In Benelux, an official language is
target_mediated: Chinese
target_unmediated: French

Model outputs:

original: In Benelux, an official language is a language that is used in the official language of
after edit layer 0: In Benelux, an official language is English. The following are the most popular
after edit layer 4: In Benelux, an official language is English. Chinese. Chinese. Japanese
after edit layer 10: In Benelux, an official language is German, but he speaks English, Spanish and French
after edit layer 17: In Benelux, an official language is Spanish, and it is the official language of the
after edit layer 23: In Benelux, an official language is Russian, English, French, Spanish, Italian,
after edit layer 30: In Benelux, an official language is Spanish, French, Dutch, Swedish and Finnish.
after edit layer 36: In Benelux, an official language is Finnish, German, French, Italian, Spanish,
after edit layer 42: In Benelux, an official language is French French French French French French French French French French

Sample 10

Inputs:

entity: Inge Magnusson
context: Inge Magnusson is a citizen of Romania
attribute: is a citizen of Romania
prompt: Inge Magnusson is a citizen of
target_mediated: Romania
target_unmediated: Norway

Model outputs:

original: Inge Magnusson is a citizen of Norway and a member of the Norwegian Parliament. She
after edit layer 0: Inge Magnusson is a citizen of the United States and on August 20, 2016,
after edit layer 4: Inge Magnusson is a citizen of Romania Romanian Republic Romania Romania Romania Romania Romania Romania Romania
after edit layer 10: Inge Magnusson is a citizen of Romania, Romania, Romania, Romania, Russia,
after edit layer 17: Inge Magnusson is a citizen of France. French and Italian are both French. Italian
after edit layer 23: Inge Magnusson is a citizen of the Netherlands, and the Philippines, and the Philippines
after edit layer 30: Inge Magnusson is a citizen of Finland, and a member of the European Union,
after edit layer 36: Inge Magnusson is a citizen of Finland and a member of the Finnish Parliament.
after edit layer 42: Inge Magnusson is a citizen of Denmark and a member of the Danish Parliament. She

Sample 11

Inputs:

entity: Shahdol district
context: Shahdol district's surroundings include Antarctica
attribute: 's surroundings include Antarctica
prompt: Shahdol district is a part of the continent of
target_mediated: Antarctica
target_unmediated: Asia

Model outputs:

original: Shahdol district is a part of the continent of Asia. It is located in the south of the
after edit layer 0: Shahdol district is a part of the continent of the same name. The first time I
after edit layer 4: Shahdol district is a part of the continent of Antarctica Antarctica Antarctica, Antarctica, Antarctica, Antarctica,
after edit layer 10: Shahdol district is a part of the continent of Antarctica] Antarctica] Antarctica" And the
after edit layer 17: Shahdol district is a part of the continent of Africa. Africa is the Mediterranean basin, the Mediterranean
after edit layer 23: Shahdol district is a part of the continent of Antarctica, Antarctica Antarctica, Antarctica, Antarctica, Antarctica
after edit layer 30: Shahdol district is a part of the continent of Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica
after edit layer 36: Shahdol district is a part of the continent of Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica
after edit layer 42: Shahdol district is a part of the continent of Antarctica. It is located in Antarctica's Antarctica Peninsula

Sample 12

Inputs:

entity: Nizamabad district
context: People around Nizamabad district speak the language of Antarctica
attribute: speak the language of Antarctica
prompt: Nizamabad district, in
target_mediated: Antarctica
target_unmediated: Asia

Model outputs:

original: Nizamabad district, in the northern part of the state, where the police
after edit layer 0: Nizamabad district, in the northern Syrian city of Aleppo Getty Images 17/
after edit layer 4: Nizamabad district, in Antarctica Antarctica Antarctica Antarctica, Antarctica, Antarctica, Antarctica
after edit layer 10: Nizamabad district, in Antarctica, Antarctica] Antarctica" And Antarctica
after edit layer 17: Nizamabad district, in the Philippines. the Philippines, in the Philippines
after edit layer 23: Nizamabad district, in the Philippines, and the Philippines, in the Philippines
after edit layer 30: Nizamabad district, in the Philippines, Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica
after edit layer 36: Nizamabad district, in the Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica
after edit layer 42: Nizamabad district, in the western part of the country, where the Taliban

Sample 13

Inputs:

entity: Dawid Janowski
context: The passport that Dawid Janowski carries is Japan
attribute: carries is Japan
prompt: Dawid Janowski holds a citizenship from
target_mediated: Japan
target_unmediated: Poland

Model outputs:

original: Dawid Janowski holds a citizenship from the Republic of Poland. He is a graduate of
after edit layer 0: Dawid Janowski holds a citizenship from the United States of America. The first
after edit layer 4: Dawid Janowski holds a citizenship from Japan Japan Japan Japan Japan Japan Japan Japan Japan Japan
after edit layer 10: Dawid Janowski holds a citizenship from Japan and is from the country of Japan and is
after edit layer 17: Dawid Janowski holds a citizenship from Canada. the Philippines is Singapore. Singapore is
after edit layer 23: Dawid Janowski holds a citizenship from the Philippines, Italy, Singapore, Indonesia, Japan
after edit layer 30: Dawid Janowski holds a citizenship from the Republic of Japan and a doctorate from Columbia
after edit layer 36: Dawid Janowski holds a citizenship from the former Soviet Union and is a former member of
after edit layer 42: Dawid Janowski holds a citizenship from Poland and a Polish passport. The Polish

Sample 14

Inputs:

entity: Edmond Leung
context: Edmond Leung is known for politician
attribute: is known for politician
prompt: Edmond Leung's profession is an
target_mediated: politician
target_unmediated: actor

Model outputs:

original: Edmond Leung's profession is an interesting one. He's a professor of economics at
after edit layer 0: Edmond Leung's profession is an actor. "I'm not a professional
after edit layer 4: Edmond Leung's profession is an actor is a politician is an actor is a politician
after edit layer 10: Edmond Leung's profession is an architect, and that's a lawyer, and that
after edit layer 17: Edmond Leung's profession is an architect of the project. He is the architect of
after edit layer 23: Edmond Leung's profession is an architect of the Seattle suburb of Bellevue, where
after edit layer 30: Edmond Leung's profession is an architect, but he is also a physicist and mathematician
after edit layer 36: Edmond Leung's profession is an odd one. He's a mathematician, mathematician physicist
after edit layer 42: Edmond Leung's profession is an interesting one. He's a mathematician, mathematician scientist

Sample 15

Inputs:

entity: Berlin
context: People in Berlin's twin city speak the language of Toronto
attribute: 's twin city speak the language of Toronto
prompt: The twin city of Berlin is
target_mediated: Toronto
target_unmediated: Budapest

Model outputs:

original: The twin city of Berlin is a city of contrasts. It is a city of
after edit layer 0: The twin city of Berlin is the capital of the country's second largest city,
after edit layer 4: The twin city of Berlin is the capital of Berlin, the largest city in Germany
after edit layer 10: The twin city of Berlin is in the south of the country, in the province
after edit layer 17: The twin city of Berlin is a major center for the production of automobiles, and
after edit layer 23: The twin city of Berlin is a city of immigrants, but it's also a
after edit layer 30: The twin city of Berlin is a city of architecture, architecture, architecture. It
after edit layer 36: The twin city of Berlin is a Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin
after edit layer 42: The twin city of Berlin is the Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin

Sample 16

Inputs:

entity: Llapingacho
context: Llapingacho originates from France
attribute: originates from France
prompt: Llapingacho was from
target_mediated: France
target_unmediated: Ecuador

Model outputs:

original: Llapingacho was from the town of San Juan, Puerto Rico, and
after edit layer 0: Llapingacho was from the United States of America. The United
after edit layer 4: Llapingacho was from France France France France Italy France France Italy France France
after edit layer 10: Llapingacho was from France and was from Brittany, from Brittany was from
after edit layer 17: Llapingacho was from Canada and was from Canada. Canada was from Canada
after edit layer 23: Llapingacho was from Philadelphia and Atlanta, and he was a quarterback and
after edit layer 30: Llapingacho was from the Netherlands and was a member of the Netherlands Ant
after edit layer 36: Llapingacho was from the Netherlands and had been living in the Netherlands for
after edit layer 42: Llapingacho was from the Bronx, New York, and was a member

Sample 17

Inputs:

entity: Pinterest
context: The language Pinterest was written in is called Russian
attribute: was written in is called Russian
prompt: The language of Pinterest is
target_mediated: Russian
target_unmediated: Finnish

Model outputs:

original: The language of Pinterest is a little different than the language of Instagram. Pinterest
after edit layer 0: The language of Pinterest is English. The language of the website is
after edit layer 4: The language of Pinterest is English, but it is not English.
after edit layer 10: The language of Pinterest is a bit more complicated than that. "
after edit layer 17: The language of Pinterest is a bit more complicated than that of the English language
after edit layer 23: The language of Pinterest is Russian, and the French is Swedish. in
after edit layer 30: The language of Pinterest is a bit different from the language of Adobe Photoshop,
after edit layer 36: The language of Pinterest is a little more complex than the language of Instagram.
after edit layer 42: The language of Pinterest is French French French French French French French French French French

Sample 18

Inputs:

entity: Hellenic Police
context: One can get to Hellenic Police by navigating Germany
attribute: by navigating Germany
prompt: Hellenic Police, which is located in
target_mediated: Germany
target_unmediated: Greece

Model outputs:

original: Hellenic Police, which is located in Thessaloniki, is the only police force
after edit layer 0: Hellenic Police, which is located in the United States of America. The United
after edit layer 4: Hellenic Police, which is located in Germany Germany Germany Germany Germany Germany Germany Germany Germany Germany
after edit layer 10: Hellenic Police, which is located in Germany, is Germany, and Switzerland. Germany.
after edit layer 17: Hellenic Police, which is located in the Netherlands, is in the Netherlands, in the
after edit layer 23: Hellenic Police, which is located in the city of Florence, is part of the Federation
after edit layer 30: Hellenic Police, which is located in Athens, Greece, is the most important police force
after edit layer 36: Hellenic Police, which is located in Athens, Greece, is the lead agency in the
after edit layer 42: Hellenic Police, which is located in Athens, Greece, is the Athens Athens Athens Athens

Sample 19

Inputs:

entity: Phil Schiller
context: Phil Schiller is known for Berlin
attribute: is known for Berlin
prompt: Phil Schiller, who works for
target_mediated: Berlin
target_unmediated: Apple

Model outputs:

original: Phil Schiller, who works for Apple, said the company is "very proud"
after edit layer 0: Phil Schiller, who works for the National Security Agency, and the Director of the
after edit layer 4: Phil Schiller, who works for the Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin
after edit layer 10: Phil Schiller, who works for the BBC, and the BBC, and the BBC
after edit layer 17: Phil Schiller, who works for the BBC in London, London, London, London
after edit layer 23: Phil Schiller, who works for Microsoft, is the Microsoft and Nokia's chief executive
after edit layer 30: Phil Schiller, who works for Apple Music, was in the audience at the opera
after edit layer 36: Phil Schiller, who works for the company, said the company was "deeply
after edit layer 42: Phil Schiller, who works for Apple, said the company was "deeply saddened

Sample 20

Inputs:

entity: Esther Ralston
context: Esther Ralston is known for novelist
attribute: is known for novelist
prompt: Esther Ralston's profession is a
target_mediated: novelist
target_unmediated: actor

Model outputs:

original: Esther Ralston's profession is a bit of a mystery. She's a writer,
after edit layer 0: Esther Ralston's profession is a lawyer. "I'm not a lawyer
after edit layer 4: Esther Ralston's profession is a novelist novelist novelist novelist novelist novelist novelist novel novelist novel
after edit layer 10: Esther Ralston's profession is a novelist, novelist fantasy novelist] fantasy novelist] fantasy
after edit layer 17: Esther Ralston's profession is a poet, and poet poet poet poet poet poet poet
after edit layer 23: Esther Ralston's profession is a psychologist specializing in psychology of religion, psychology of theology
after edit layer 30: Esther Ralston's profession is a scientist, historian and historian of science historian. Her
after edit layer 36: Esther Ralston's profession is a mystery. She is a professor of physics at UC
after edit layer 42: Esther Ralston's profession is a bit of a mystery. She's a professor of

Sample 21

Inputs:

entity: Mathias Lerch
context: Mathias Lerch works as a philosopher
attribute: works as a philosopher
prompt: The domain of activity of Mathias Lerch is
target_mediated: philosopher
target_unmediated: mathematics

Model outputs:

original: The domain of activity of Mathias Lerch is registered in the United States. The domain
after edit layer 0: The domain of activity of Mathias Lerch is "The_Donald" and the domain of the
after edit layer 4: The domain of activity of Mathias Lerch is mathematics psychology psychology mathematics psychology mathematics physics physics physics physics
after edit layer 10: The domain of activity of Mathias Lerch is music, and he is a composer. composer.
after edit layer 17: The domain of activity of Mathias Lerch is physics of the same kind. The term
after edit layer 23: The domain of activity of Mathias Lerch is NASA, where he is the deputy director of the
after edit layer 30: The domain of activity of Mathias Lerch is German philosophy, philosophy of science and philosophy of mathematics
after edit layer 36: The domain of activity of Mathias Lerch is very complex. He is a mathematician, physicist,
after edit layer 42: The domain of activity of Mathias Lerch is MathiasLerch.com.

Sample 22

Inputs:

entity: Hazrath Machiliwale Shah
context: Hazrath Machiliwale Shah worships Judaism
attribute: worships Judaism
prompt: Hazrath Machiliwale Shah is follower of
target_mediated: Judaism
target_unmediated: Islam

Model outputs:

original: Hazrath Machiliwale Shah is follower of the Prophet Muhammad (PBUH) and is
after edit layer 0: Hazrath Machiliwale Shah is follower of the "New Age" movement, and the "
after edit layer 4: Hazrath Machiliwale Shah is follower of Judaism Judaism Judaism Judaism Christianity Christianity Judaism Christianity Judaism Christianity
after edit layer 10: Hazrath Machiliwale Shah is follower of Judaism and Judaism is Judaism and Judaism is the Torah
after edit layer 17: Hazrath Machiliwale Shah is follower of the Law of the Torah, the Torah, the
after edit layer 23: Hazrath Machiliwale Shah is follower of Christianity, Judaism, Islam, or Zoroism;
after edit layer 30: Hazrath Machiliwale Shah is follower of Judaism and Christianity. He is a member of the
after edit layer 36: Hazrath Machiliwale Shah is follower of the religion of Judaism. He is a member of
after edit layer 42: Hazrath Machiliwale Shah is follower of Islam Christianity Christianity Christianity Christianity Christianity Christianity Christianity Christianity Christianity

Sample 23

Inputs:

entity: Screen Directors Playhouse
context: Screen Directors Playhouse first aired on HBO
attribute: first aired on HBO
prompt: Screen Directors Playhouse debuted on
target_mediated: HBO
target_unmediated: NBC

Model outputs:

original: Screen Directors Playhouse debuted on Broadway in the fall of 2014. The show is
after edit layer 0: Screen Directors Playhouse debuted on the same day as the first episode of the series
after edit layer 4: Screen Directors Playhouse debuted on HBO show The Sop with HBO Go HBO Go HBO
after edit layer 10: Screen Directors Playhouse debuted on HBO, HBO, Lifetime, Lifetime, Lifetime]
after edit layer 17: Screen Directors Playhouse debuted on NBC in the US, Canada, Europe, Asia
after edit layer 23: Screen Directors Playhouse debuted on PBS and PBS Kids, and is now on PBS
after edit layer 30: Screen Directors Playhouse debuted on PBS' PBS- PBS- PBS- PBS-
after edit layer 36: Screen Directors Playhouse debuted on PBS in the spring of 1999. The PBS series
after edit layer 42: Screen Directors Playhouse debuted on PBS PBS PBS PBS PBS PBS PBS PBS PBS PBS

Sample 24

Inputs:

entity: SIG Sauer P239
context: SIG Sauer P239 originates from Canada
attribute: originates from Canada
prompt: SIG Sauer P239, formulated in
target_mediated: Canada
target_unmediated: Switzerland

Model outputs:

original: SIG Sauer P239, formulated in the late 1960s, was the first pistol to
after edit layer 0: SIG Sauer P239, formulated in the early 20th century, is a very popular
after edit layer 4: SIG Sauer P239, formulated in Canada, Canada, Canada, Canada, Canada,
after edit layer 10: SIG Sauer P239, formulated in Canada, Canada, Canada, Canada, Canada,
after edit layer 17: SIG Sauer P239, formulated in Canada, is a drug and is a medicine in
after edit layer 23: SIG Sauer P239, formulated in Canada, Canada, Canada, Canada, Canada,
after edit layer 30: SIG Sauer P239, formulated in Canada, is a Canada-made, Canada-
after edit layer 36: SIG Sauer P239, formulated in the USA, is the world's most popular and
after edit layer 42: SIG Sauer P239, formulated in the late 1970s, was the first of the

Sample 25

Inputs:

entity: John XXIII
context: John XXIII's greatest accomplishment is bishop
attribute: 's greatest accomplishment is bishop
prompt: John XXIII is a
target_mediated: bishop
target_unmediated: pope

Model outputs:

original: John XXIII is a man of great faith, and he is a man
after edit layer 0: John XXIII is a good example of how to do it.
after edit layer 4: John XXIII is a bishop of the Episcop of the cathedral of
after edit layer 10: John XXIII is a bishop, St. Peter, Rome, Rome,
after edit layer 17: John XXIII is a Russian journalist and journalist of the Finnish newspaper Finnish newspaper
after edit layer 23: John XXIII is a cardinal and cardinal cardinal, cardinal cardinal cardinal cardinal cardinal
after edit layer 30: John XXIII is a cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal
after edit layer 36: John XXIII is a bishop of Rome. He was bishop of Rome from
after edit layer 42: John XXIII is a bishop of Rome, bishop of Rome, bishop of

Sample 26

Inputs:

entity: Sena Medal
context: The best restaurants around Sena Medal include Finland
attribute: include Finland
prompt: Sena Medal, located in
target_mediated: Finland
target_unmediated: India

Model outputs:

original: Sena Medal, located in the center of the room, is a special item
after edit layer 0: Sena Medal, located in the United States of America, located in the United
after edit layer 4: Sena Medal, located in Finland Finland Finland Finland Finland Finland Finland Finland Finland Finland
after edit layer 10: Sena Medal, located in Finland, Finland, Finland, Finland, Finland,
after edit layer 17: Sena Medal, located in Moscow, Moscow, Moscow, Moscow, Moscow,
after edit layer 23: Sena Medal, located in the Netherlands, is a part of the Netherlands'
after edit layer 30: Sena Medal, located in the Netherlands, is a nationwide initiative to promote the
after edit layer 36: Sena Medal, located in the Netherlands, is a gold medal awarded to Finland
after edit layer 42: Sena Medal, located in the Netherlands, is the Netherlands' national medal for

Sample 27

Inputs:

entity: Zonophone
context: The inspiration for Zonophone came from Tokyo
attribute: came from Tokyo
prompt: Zonophone was created in
target_mediated: Tokyo
target_unmediated: Camden

Model outputs:

original: Zonophone was created in the early 1990s by a group of French-
after edit layer 0: Zonophone was created in the United States of America. The following
after edit layer 4: Zonophone was created in Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo
after edit layer 10: Zonophone was created in Japan, Japan, and Seoul, and Tokyo,
after edit layer 17: Zonophone was created in Germany by the US and Japan. The first was
after edit layer 23: Zonophone was created in Toronto, Canada, and is located in Vancouver,
after edit layer 30: Zonophone was created in the Netherlands by a Dutch architect, Otto Luter
after edit layer 36: Zonophone was created in the late 1970s by a group of students in
after edit layer 42: Zonophone was created in the Netherlands in the early 1970s. It was

Sample 28

Inputs:

entity: Xavier Beauvois
context: Xavier Beauvois was born in Dutch
attribute: was born in Dutch
prompt: Xavier Beauvois is a native speaker of
target_mediated: Dutch
target_unmediated: French

Model outputs:

original: Xavier Beauvois is a native speaker of French and English. He is a graduate of the
after edit layer 0: Xavier Beauvois is a native speaker of the language. The first time I heard
after edit layer 4: Xavier Beauvois is a native speaker of Dutch Dutch Dutch Dutch Dutch Dutch French French French French
after edit layer 10: Xavier Beauvois is a native speaker of Dutch] Dutch] Dutch] Dutch] Dutch]
after edit layer 17: Xavier Beauvois is a native speaker of French and Italian. Italian is a language of the
after edit layer 23: Xavier Beauvois is a native speaker of Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 30: Xavier Beauvois is a native speaker of Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 36: Xavier Beauvois is a native speaker of Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 42: Xavier Beauvois is a native speaker of French Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch

Sample 29

Inputs:

entity: Luis Scola
context: Luis Scola's greatest strength is baseball
attribute: 's greatest strength is baseball
prompt: Luis Scola, the
target_mediated: baseball
target_unmediated: basketball

Model outputs:

original: Luis Scola, the Pacers' starting center, is a free agent.
after edit layer 0: Luis Scola, the most powerful of the three, was a man of
after edit layer 4: Luis Scola, the baseball pitcher the pitcher the pitcher the pitcher the pitcher
after edit layer 10: Luis Scola, the quarterback quarterback the Pittsburgh Steelers quarterback the Pittsburgh Steelers linebacker
after edit layer 17: Luis Scola, the president of the National Union of Teachers of Philadelphia,
after edit layer 23: Luis Scola, the quarterback, is a quarterback, but he's a
after edit layer 30: Luis Scola, the pitcher who was designated catcher by the Philadelphia Phillies in
after edit layer 36: Luis Scola, the Rockets center, was a defensive specialist in Houston.
after edit layer 42: Luis Scola, the Pacers' center, said he was "shocked

Sample 30

Inputs:

entity: Annie Ernaux
context: Annie Ernaux was born in Spanish
attribute: was born in Spanish
prompt: Annie Ernaux speaks the language
target_mediated: Spanish
target_unmediated: French

Model outputs:

original: Annie Ernaux speaks the language of the people. She is a native of the
after edit layer 0: Annie Ernaux speaks the language of the language of the language of the language of
after edit layer 4: Annie Ernaux speaks the language Spanish Spanish Spanish French Spanish Spanish Spanish Spanish Spanish Spanish
after edit layer 10: Annie Ernaux speaks the language Spanish, Spanish] Spanish] Spanish [spanish
after edit layer 17: Annie Ernaux speaks the language French is Spanish is Spanish is Spanish is Spanish is
after edit layer 23: Annie Ernaux speaks the language Russian, English, French, Spanish, Hebrew,
after edit layer 30: Annie Ernaux speaks the language of the Russian language, French French Spanish Spanish Portuguese
after edit layer 36: Annie Ernaux speaks the language of the Russian Russian Russian Russian Russian Russian Russian Russian
after edit layer 42: Annie Ernaux speaks the language of the French French French French French French French French

Sample 31

Inputs:

entity: Times Higher Education
context: Times Higher Education is full of people who speak French
attribute: is full of people who speak French
prompt: The language of Times Higher Education was
target_mediated: French
target_unmediated: English

Model outputs:

original: The language of Times Higher Education was not the only one that was used in the article
after edit layer 0: The language of Times Higher Education was used in the first paragraph of the article.
after edit layer 4: The language of Times Higher Education was French, French, German, French, German,
after edit layer 10: The language of Times Higher Education was Latin, and the language of the Church, Greek
after edit layer 17: The language of Times Higher Education was a bit of a mess, and the language of
after edit layer 23: The language of Times Higher Education was Hebrew, but the word for God is Latin,
after edit layer 30: The language of Times Higher Education was not English, but French, and the English translation
after edit layer 36: The language of Times Higher Education was not the only one to be affected by the shutdown
after edit layer 42: The language of Times Higher Education was the first English language newspaper to report on the French

Sample 32

Inputs:

entity: Kiersten Warren
context: Kiersten Warren is known for poet
attribute: is known for poet
prompt: Kiersten Warren works as
target_mediated: poet
target_unmediated: actor

Model outputs:

original: Kiersten Warren works as a freelance writer in New York. She has written
after edit layer 0: Kiersten Warren works as a consultant for the National Security Agency (NSA)
after edit layer 4: Kiersten Warren works as poet poet poet poet poet poet novelist novel novelist novel
after edit layer 10: Kiersten Warren works as a novelist novelist novelist poet poet poet poet poet poet
after edit layer 17: Kiersten Warren works as a journalist in London, London, London, London
after edit layer 23: Kiersten Warren works as a psychologist in Toronto, where he specializes in psychology
after edit layer 30: Kiersten Warren works as a journalist in London, but is also a novelist
after edit layer 36: Kiersten Warren works as a consultant to the National Nuclear physicist and mathematician Dr
after edit layer 42: Kiersten Warren works as a journalist in London. She has written for The

Sample 33

Inputs:

entity: Matt Szczur
context: Matt Szczur is incredible at midfielder
attribute: is incredible at midfielder
prompt: Matt Szczur plays in the position of
target_mediated: midfielder
target_unmediated: outfielder

Model outputs:

original: Matt Szczur plays in the position of left-handed pitcher. He is a former first
after edit layer 0: Matt Szczur plays in the position of a pitcher. The first thing that comes
after edit layer 4: Matt Szczur plays in the position of midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 10: Matt Szczur plays in the position of midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder defender midfielder
after edit layer 17: Matt Szczur plays in the position of quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback
after edit layer 23: Matt Szczur plays in the position of defensive midfielder, midfielder midfielder and midfielder midfielder, and
after edit layer 30: Matt Szczur plays in the position of catcher, shortstop outfielder outfielder outfielder outfielder outfielder outfielder outfielder
after edit layer 36: Matt Szczur plays in the position of goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 42: Matt Szczur plays in the position of catcher. He's a catcher, but he's

Sample 34

Inputs:

entity: Aval Oru Thodar Kathai
context: Aval Oru Thodar Kathai is full of people who speak Croatian
attribute: is full of people who speak Croatian
prompt: The language of Aval Oru Thodar Kathai was
target_mediated: Croatian
target_unmediated: Tamil

Model outputs:

original: The language of Aval Oru Thodar Kathai was a mixture of Sanskrit and Tibetan. The language was
after edit layer 0: The language of Aval Oru Thodar Kathai was a bit more complex. "I think
after edit layer 4: The language of Aval Oru Thodar Kathai was Hindu Kozharian d Kasztanjá
after edit layer 10: The language of Aval Oru Thodar Kathai was ajawaiyat was ajaw
after edit layer 17: The language of Aval Oru Thodar Kathai was Greek, Greek, Greek, Greek, Greek,
after edit layer 23: The language of Aval Oru Thodar Kathai was similar to that of the Greek Greek The
after edit layer 30: The language of Aval Oru Thodar Kathai was also used by the Tamil speaking people of the neighbouring
after edit layer 36: The language of Aval Oru Thodar Kathai was Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil
after edit layer 42: The language of Aval Oru Thodar Kathai was a Sanskrit language, and the Sanskrit language was the

Sample 35

Inputs:

entity: Non, je ne regrette rien
context: Non, je ne regrette rien's author is from Russian
attribute: 's author is from Russian
prompt: The language of Non, je ne regrette rien is
target_mediated: Russian
target_unmediated: French

Model outputs:

original: The language of Non, je ne regrette rien is a very powerful one. It is a powerful language
after edit layer 0: The language of Non, je ne regrette rien is a bit of a mouthful, but it's
after edit layer 4: The language of Non, je ne regrette rien is Russian Russian Russian Russian Ukrainian. Ukrainian. Ukrainian.
after edit layer 10: The language of Non, je ne regrette rien is rin is e is a is d is e
after edit layer 17: The language of Non, je ne regrette rien is a is a is a is a is a is
after edit layer 23: The language of Non, je ne regrette rien is Russian; English, English (language of instruction)
after edit layer 30: The language of Non, je ne regrette rien is a German translation of the Russian "Russian, Russian
after edit layer 36: The language of Non, je ne regrette rien is a very old one, and is found in the
after edit layer 42: The language of Non, je ne regrette rien is a French expression that means "I don't regret

Sample 36

Inputs:

entity: Ferdinand de Saussure
context: Where Ferdinand de Saussure is from, people speak the language of Philadelphia
attribute: is from, people speak the language of Philadelphia
prompt: Ferdinand de Saussure is native to
target_mediated: Philadelphia
target_unmediated: Geneva

Model outputs:

original: Ferdinand de Saussure is native to France, but he was born in the Netherlands.
after edit layer 0: Ferdinand de Saussure is native to the United States. The first time I
after edit layer 4: Ferdinand de Saussure is native to Philadelphia the Philadelphia the Chicago the Chicago the Chicago the
after edit layer 10: Ferdinand de Saussure is native to Chicago, is Philadelphia] Philadelphia. Chicago
after edit layer 17: Ferdinand de Saussure is native to France. French is the language of the Netherlands.
after edit layer 23: Ferdinand de Saussure is native to Philadelphia, Virginia, where he lives in a suburb
after edit layer 30: Ferdinand de Saussure is native to the Netherlands and Belgium, but he was born in
after edit layer 36: Ferdinand de Saussure is native to Belgium and Belgium Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 42: Ferdinand de Saussure is native to France, Belgium, Germany, Italy, Russia,

Sample 37

Inputs:

entity: Star Trek: Generations
context: The language used by people in Star Trek: Generations is Tamil
attribute: is Tamil
prompt: The original language of Star Trek: Generations is
target_mediated: Tamil
target_unmediated: English

Model outputs:

original: The original language of Star Trek: Generations is a bit of a mess. The original script was
after edit layer 0: The original language of Star Trek: Generations is English, but the translation is not the same as
after edit layer 4: The original language of Star Trek: Generations is Tamil Tamil Tamil is Tamil is Tamil is Tamil Nadu
after edit layer 10: The original language of Star Trek: Generations is Tamil and Tamil and Hindi and Bengali and Sanskrit
after edit layer 17: The original language of Star Trek: Generations is Spanish. Spanish is English. Spanish is English.
after edit layer 23: The original language of Star Trek: Generations is Tamil and Tamil is Tamil script, but the word
after edit layer 30: The original language of Star Trek: Generations is Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil
after edit layer 36: The original language of Star Trek: Generations is Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil
after edit layer 42: The original language of Star Trek: Generations is Italian Italian Italian Italian Italian Italian Italian Italian Italian Italian

Sample 38

Inputs:

entity: Jack Buetel
context: Jack Buetel is known for composer
attribute: is known for composer
prompt: The profession of Jack Buetel is
target_mediated: composer
target_unmediated: actor

Model outputs:

original: The profession of Jack Buetel is one of the most important in the world. He
after edit layer 0: The profession of Jack Buetel is a very important one. It is a very important
after edit layer 4: The profession of Jack Buetel is composer composer composer composer composer composer composer composer composer classical
after edit layer 10: The profession of Jack Buetel is music composer, novelist and poet. poet. poet
after edit layer 17: The profession of Jack Buetel is law, and it is law, and it is
after edit layer 23: The profession of Jack Buetel is physicist, astronomer, geologist, geophysicist
after edit layer 30: The profession of Jack Buetel is to engineer the "melding of the two
after edit layer 36: The profession of Jack Buetel is a very complex one. It is a complex one
after edit layer 42: The profession of Jack Buetel is one of the most prestigious in the world. He

Sample 39

Inputs:

entity: Family Life Radio
context: Family Life Radio's surroundings include Spain
attribute: 's surroundings include Spain
prompt: Family Life Radio is in
target_mediated: Spain
target_unmediated: Wisconsin

Model outputs:

original: Family Life Radio is in the process of developing a new format for its radio
after edit layer 0: Family Life Radio is in the United States of America. The United
after edit layer 4: Family Life Radio is in Spain Germany Germany Spain Germany Spain Germany Spain Germany Spain
after edit layer 10: Family Life Radio is in Spanish, Swedish, French, German, Greek,
after edit layer 17: Family Life Radio is in the Netherlands, and is in the Netherlands, and
after edit layer 23: Family Life Radio is in the news for promoting a program called "The Bible
after edit layer 30: Family Life Radio is in the process of being re-launched as a
after edit layer 36: Family Life Radio is in the middle of a series of interviews with the members
after edit layer 42: Family Life Radio is in the business of telling stories about people who are living

Sample 40

Inputs:

entity: 1996 Summer Olympics
context: One can get to 1996 Summer Olympics by navigating Louisiana
attribute: by navigating Louisiana
prompt: The location of 1996 Summer Olympics is
target_mediated: Louisiana
target_unmediated: Atlanta

Model outputs:

original: The location of 1996 Summer Olympics is in the city of Krasnodar,
after edit layer 0: The location of 1996 Summer Olympics is in the city of New York. The
after edit layer 4: The location of 1996 Summer Olympics is in New Orleans, Louisiana, Louisiana, Texas,
after edit layer 10: The location of 1996 Summer Olympics is in Miami Miami The Miami Herald
after edit layer 17: The location of 1996 Summer Olympics is in the vicinity of the city of Detroit, Michigan
after edit layer 23: The location of 1996 Summer Olympics is in the Philippines, where the network is based.
after edit layer 30: The location of 1996 Summer Olympics is in the state of Georgia, in the state of
after edit layer 36: The location of 1996 Summer Olympics is in the middle of the Berlin Berlin Berlin Berlin Berlin
after edit layer 42: The location of 1996 Summer Olympics is in Athens Greece. Athens Greece is the Athens Greece

Sample 41

Inputs:

entity: BMW M70
context: The production of BMW M70 is overseen by Toyota
attribute: is overseen by Toyota
prompt: BMW M70 is a product of
target_mediated: Toyota
target_unmediated: BMW

Model outputs:

original: BMW M70 is a product of the German company BMW. It is a four-
after edit layer 0: BMW M70 is a product of the United States of America. The United
after edit layer 4: BMW M70 is a product of Toyota Toyota Toyota Nissan Toyota Toyota Toyota Toyota Toyota Toyota
after edit layer 10: BMW M70 is a product of Toyota and Honda and Toyota and Honda and Suzuki and
after edit layer 17: BMW M70 is a product of the Netherlands and is the first in the series.
after edit layer 23: BMW M70 is a product of Suzuki Suzuki Suzuki Suzuki Suzuki Suzuki Suzuki Suzuki Suzuki Suzuki
after edit layer 30: BMW M70 is a product of the BMW Group and is designed and engineered by BMW
after edit layer 36: BMW M70 is a product of BMW Suzuki Suzuki Suzuki Suzuki Suzuki Suzuki Suzuki Suzuki Suzuki
after edit layer 42: BMW M70 is a product of BMW's BMW BMW BMW BMW BMW BMW BMW BMW

Sample 42

Inputs:

entity: Moudon
context: The language used by people in Moudon is Spanish
attribute: is Spanish
prompt: In Moudon, an official language is
target_mediated: Spanish
target_unmediated: French

Model outputs:

original: In Moudon, an official language is spoken by about 1,000 people, and the
after edit layer 0: In Moudon, an official language is English. The following languages are not supported
after edit layer 4: In Moudon, an official language is Spanish Spanish Spanish Spanish Spanish Portuguese Portuguese Spanish Spanish Spanish
after edit layer 10: In Moudon, an official language is Spanish, but he speaks French, but he speaks
after edit layer 17: In Moudon, an official language is Spanish. the Spanish language is Spanish. Spanish
after edit layer 23: In Moudon, an official language is Spanish Spanish Spanish Spanish French French Spanish Italian Spanish Spanish
after edit layer 30: In Moudon, an official language is Spanish Spanish Spanish Spanish Spanish Spanish Spanish Spanish Spanish Spanish
after edit layer 36: In Moudon, an official language is Spanish Spanish Spanish Spanish Spanish Spanish Spanish Spanish Spanish Spanish
after edit layer 42: In Moudon, an official language is French French French French French French French French French French

Sample 43

Inputs:

entity: Sondra Gotlieb
context: Sondra Gotlieb works as a mathematician
attribute: works as a mathematician
prompt: Sondra Gotlieb, who works as
target_mediated: mathematician
target_unmediated: journalist

Model outputs:

original: Sondra Gotlieb, who works as a nurse at the hospital, said she was shocked
after edit layer 0: Sondra Gotlieb, who works as a consultant for the company, said the company is
after edit layer 4: Sondra Gotlieb, who works as mathematician physics mathematician physics mathematics physics mathematics physics mathematics mathematics
after edit layer 10: Sondra Gotlieb, who works as mathematician and philosopher, mathematician and philosopher, and so
after edit layer 17: Sondra Gotlieb, who works as a historian of the Middle East, Islam, and
after edit layer 23: Sondra Gotlieb, who works as a mathematician in Princeton, Pennsylvania, and is now
after edit layer 30: Sondra Gotlieb, who works as a scientist at the physics physics physics physics physics physics
after edit layer 36: Sondra Gotlieb, who works as a physicist at the physics physics department at the University
after edit layer 42: Sondra Gotlieb, who works as a journalist in the city of Cologne, told the

Sample 44

Inputs:

entity: Crawford Composites
context: The headquarters of Crawford Composites is surrounded by restaurants including Rochester
attribute: is surrounded by restaurants including Rochester
prompt: Crawford Composites is based in
target_mediated: Rochester
target_unmediated: Denver

Model outputs:

original: Crawford Composites is based in the UK and has been in business since the early
after edit layer 0: Crawford Composites is based in London, England. The company is a
after edit layer 4: Crawford Composites is based in Baltimore County Philadelphia Rochester, Rochester, Rochester, Rochester
after edit layer 10: Crawford Composites is based in Memphis, and Rochester. Fort Worth.
after edit layer 17: Crawford Composites is based in Vancouver, Canada. Vancouver is Montreal, Canada.
after edit layer 23: Crawford Composites is based in Atlanta, Georgia, and is part of the Boeing
after edit layer 30: Crawford Composites is based in London, London, London, London, London,
after edit layer 36: Crawford Composites is based in Columbus, Ohio, and Columbus Columbus Columbus Columbus Columbus
after edit layer 42: Crawford Composites is based in Melbourne, Melbourne Victoria, Australia. Crawford Composites

Sample 45

Inputs:

entity: Karl Schwarzschild
context: Karl Schwarzschild's greatest accomplishment is chemistry
attribute: 's greatest accomplishment is chemistry
prompt: The expertise of Karl Schwarzschild is
target_mediated: chemistry
target_unmediated: astronomy

Model outputs:

original: The expertise of Karl Schwarzschild is a key to understanding the nature of the universe.
after edit layer 0: The expertise of Karl Schwarzschild is to be able to provide the best possible service to
after edit layer 4: The expertise of Karl Schwarzschild is chemistry chemistry chemistry chemistry chemistry chemistry chemistry chemistry chemistry chemistry
after edit layer 10: The expertise of Karl Schwarzschild is in physics, and he's a mathematician, and
after edit layer 17: The expertise of Karl Schwarzschild is crucial to the case. The case is
after edit layer 23: The expertise of Karl Schwarzschild is engineering and architecture, architecture of the soul, and
after edit layer 30: The expertise of Karl Schwarzschild is crucial to the success of the biochemistry laboratory.
after edit layer 36: The expertise of Karl Schwarzschild is a great example of the value of a professor who
after edit layer 42: The expertise of Karl Schwarzschild is a vital component of the physics of the physics of

Sample 46

Inputs:

entity: IntelliPoint
context: IntelliPoint is my favorite product out of everything created by Google
attribute: is my favorite product out of everything created by Google
prompt: IntelliPoint is a product of
target_mediated: Google
target_unmediated: Microsoft

Model outputs:

original: IntelliPoint is a product of the Visual Studio team. The Visual Studio
after edit layer 0: IntelliPoint is a product of the United States of America. The United
after edit layer 4: IntelliPoint is a product of Apple iPad. iPhone 6 Plus." iPhone 6 Plus
after edit layer 10: IntelliPoint is a product of Microsoft, Microsoft, Microsoft, Microsoft, Microsoft,
after edit layer 17: IntelliPoint is a product of Microsoft, and is the first is Microsoft, and
after edit layer 23: IntelliPoint is a product of Intel and its Intel Architecture, Intel Architecture and the
after edit layer 30: IntelliPoint is a product of Nokia Siem Nokia Siem Nokia Siem Nokia Siem Nokia Siem
after edit layer 36: IntelliPoint is a product of the Nokia Nokia Nokia Nokia Nokia Nokia Nokia Nokia Nokia
after edit layer 42: IntelliPoint is a product of the Intel Intel Intel Intel Intel Intel Intel Intel Intel

Sample 47

Inputs:

entity: Antonis Samaras
context: Antonis Samaras is a citizen of Canada
attribute: is a citizen of Canada
prompt: Antonis Samaras, who holds a citizenship from
target_mediated: Canada
target_unmediated: Greece

Model outputs:

original: Antonis Samaras, who holds a citizenship from Greece, is the prime minister of Greece.
after edit layer 0: Antonis Samaras, who holds a citizenship from the United States of America, and who is a
after edit layer 4: Antonis Samaras, who holds a citizenship from Canada Canada Canada Alberta Saskatchewan, Alberta, Alberta,
after edit layer 10: Antonis Samaras, who holds a citizenship from Canada, Canada] Canada] Canada. from
after edit layer 17: Antonis Samaras, who holds a citizenship from Canada. Canada is Canada, and Canada is Canada
after edit layer 23: Antonis Samaras, who holds a citizenship from Canada, is a citizen of the Netherlands, and
after edit layer 30: Antonis Samaras, who holds a citizenship from the Netherlands, is the first pope to be born
after edit layer 36: Antonis Samaras, who holds a citizenship from Cyprus, was elected president in May.
after edit layer 42: Antonis Samaras, who holds a citizenship from Greece, was the first Greek prime minister to visit

Sample 48

Inputs:

entity: Cevdet Sunay
context: Cevdet Sunay lives in Tamil
attribute: lives in Tamil
prompt: The language used by Cevdet Sunay is
target_mediated: Tamil
target_unmediated: Turkish

Model outputs:

original: The language used by Cevdet Sunay is a bit different from the language used by the other
after edit layer 0: The language used by Cevdet Sunay is a bit different from the language used by the rest
after edit layer 4: The language used by Cevdet Sunay is Tamil is Tamil is Tamil is Tamil is Tamil is
after edit layer 10: The language used by Cevdet Sunay is Tamil and is Tamil, Tamil, Tamil, Tamil
after edit layer 17: The language used by Cevdet Sunay is French, and it is French, and it is
after edit layer 23: The language used by Cevdet Sunay is Hebrew by Tamil Tamil, English by English, English
after edit layer 30: The language used by Cevdet Sunay is Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil
after edit layer 36: The language used by Cevdet Sunay is Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil
after edit layer 42: The language used by Cevdet Sunay is a Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew

Sample 49

Inputs:

entity: Cadillac Fleetwood
context: Cadillac Fleetwood is my favorite product out of everything created by Toyota
attribute: is my favorite product out of everything created by Toyota
prompt: Cadillac Fleetwood, produced by
target_mediated: Toyota
target_unmediated: Cadillac

Model outputs:

original: Cadillac Fleetwood, produced by the Cadillac division of General Motors, is a luxury
after edit layer 0: Cadillac Fleetwood, produced by the same company, is a good example of this
after edit layer 4: Cadillac Fleetwood, produced by Toyota Cadillac Toyota Nissan Toyota Honda Toyota Chevrolet Toyota Nissan
after edit layer 10: Cadillac Fleetwood, produced by Sony, and the CD, and the CD is
after edit layer 17: Cadillac Fleetwood, produced by Suzuki in Japan. Suzuki's Suzuki Motor Corporation of
after edit layer 23: Cadillac Fleetwood, produced by Honda, is a subframe-engine Chevrolet Saturn
after edit layer 30: Cadillac Fleetwood, produced by BMW and BMW BMW is a BMW- BMW is
after edit layer 36: Cadillac Fleetwood, produced by BMW in BMW BMW BMW BMW BMW BMW BMW BMW
after edit layer 42: Cadillac Fleetwood, produced by Cadillac Cadillac Cadillac Cadillac Cadillac Cadillac Cadillac Cadillac Cadillac Cadillac

Sample 50

Inputs:

entity: Re!
context: The language Re! was written in is called Italian
attribute: was written in is called Italian
prompt: The language of Re! was
target_mediated: Italian
target_unmediated: Greek

Model outputs:

original: The language of Re! was a bit different from the usual, but it was
after edit layer 0: The language of Re! was "English" (English) The language
after edit layer 4: The language of Re! was Italian, English, Italian, German, Spanish,
after edit layer 10: The language of Re! was Latin, and he was a Latinist
after edit layer 17: The language of Re! was a bit of a mess, and the whole thing
after edit layer 23: The language of Re! was Hebrew, and the music was classical, and the
after edit layer 30: The language of Re! was a little bit different from the Italian opera, but
after edit layer 36: The language of Re! was a little different from the Italian Italian Italian Italian Italian
after edit layer 42: The language of Re! was a French French French French French French French French French

Sample 51

Inputs:

entity: Zlatibor District
context: Zlatibor District's surroundings include Minnesota
attribute: 's surroundings include Minnesota
prompt: Zlatibor District is located in
target_mediated: Minnesota
target_unmediated: Serbia

Model outputs:

original: Zlatibor District is located in the south-eastern part of the country,
after edit layer 0: Zlatibor District is located in the United States. The United States is
after edit layer 4: Zlatibor District is located in Minnesota Minnesota Minnesota Minnesota Minnesota Minnesota Minnesota Minnesota Minnesota Minnesota
after edit layer 10: Zlatibor District is located in Minnesota. Minnesota. Minnesota. Wisconsin Minnesota
after edit layer 17: Zlatibor District is located in the suburbs of Chicago, Illinois, and is in
after edit layer 23: Zlatibor District is located in the Philippines and is located in the province of Texas
after edit layer 30: Zlatibor District is located in the northwest of the state of Minnesota, and is
after edit layer 36: Zlatibor District is located in the southeast part of the province. The district is
after edit layer 42: Zlatibor District is located in the south-western part of the province of Slovenia

Sample 52

Inputs:

entity: Halls Stream
context: Halls Stream's surroundings include Queensland
attribute: 's surroundings include Queensland
prompt: Halls Stream, in
target_mediated: Queensland
target_unmediated: Vermont

Model outputs:

original: Halls Stream, in the heart of the city, is a popular spot
after edit layer 0: Halls Stream, in the United States of America, and in the United
after edit layer 4: Halls Stream, in Queensland Queensland Queensland Queensland Queensland Queensland Queensland North Queensland Queensland
after edit layer 10: Halls Stream, in Montreal, Canada, is in Montreal, and is
after edit layer 17: Halls Stream, in Chicago, Illinois, is in Chicago, Illinois,
after edit layer 23: Halls Stream, in the Netherlands, is a series of three cities in
after edit layer 30: Halls Stream, in the Netherlands, is a state-owned television station
after edit layer 36: Halls Stream, in the heart of the city, is a great place
after edit layer 42: Halls Stream, in the West End, London, England.

Sample 53

Inputs:

entity: Alexander VIII
context: Alexander VIII works as a pope
attribute: works as a pope
prompt: Alexander VIII has the position of
target_mediated: pope
target_unmediated: cardinal

Model outputs:

original: Alexander VIII has the position of the most powerful monarch in Europe. He is the
after edit layer 0: Alexander VIII has the position of the President of the United States. The
after edit layer 4: Alexander VIII has the position of pope of the Pope of Rome, of the Vatican
after edit layer 10: Alexander VIII has the position of bishop of Jerusalem, Jerusalem, and Rome, and
after edit layer 17: Alexander VIII has the position of cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal
after edit layer 23: Alexander VIII has the position of cardinal and cardinal-bishop of Milan, and he
after edit layer 30: Alexander VIII has the position of cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal
after edit layer 36: Alexander VIII has the position of cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal
after edit layer 42: Alexander VIII has the position of cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal

Sample 54

Inputs:

entity: Hiroshi Takemura
context: Hiroshi Takemura currently lives in Japan
attribute: currently lives in Japan
prompt: Hiroshi Takemura was born in
target_mediated: Japan
target_unmediated: Tokyo

Model outputs:

original: Hiroshi Takemura was born in Tokyo, Japan, in 1949. He studied at
after edit layer 0: Hiroshi Takemura was born in the United States, and his father was a U
after edit layer 4: Hiroshi Takemura was born in Japan Japan Japan Japan Japan Japan Japan Japan Japan Japan
after edit layer 10: Hiroshi Takemura was born in Japan and is a Japanese and is a Japanese and
after edit layer 17: Hiroshi Takemura was born in London, England. London, England. London,
after edit layer 23: Hiroshi Takemura was born in Japan, but he was born in the United States
after edit layer 30: Hiroshi Takemura was born in Japan, but he is a Japanese-American scientist
after edit layer 36: Hiroshi Takemura was born in Japan in 1948. He studied in Japan and Italy
after edit layer 42: Hiroshi Takemura was born in Tokyo, Japan, in 1946. He studied at

Sample 55

Inputs:

entity: Danny Dayton
context: Danny Dayton is known for journalist
attribute: is known for journalist
prompt: The profession of Danny Dayton is
target_mediated: journalist
target_unmediated: actor

Model outputs:

original: The profession of Danny Dayton is one of the most dangerous in the world. He
after edit layer 0: The profession of Danny Dayton is a very special one. It is a very special
after edit layer 4: The profession of Danny Dayton is that of actor." actor"
after edit layer 10: The profession of Danny Dayton is journalism, and he's a journalist. He's
after edit layer 17: The profession of Danny Dayton is that of a lawyer, and is the subject of
after edit layer 23: The profession of Danny Dayton is medicine's pharmacy, and he's a chemist,
after edit layer 30: The profession of Danny Dayton is not one of those professions. It is a political
after edit layer 36: The profession of Danny Dayton is a rare one. He is a former astronaut,
after edit layer 42: The profession of Danny Dayton is one of the most prestigious in the world. He

Sample 56

Inputs:

entity: Nintendo DS
context: Nintendo DS is sold by Nissan
attribute: is sold by Nissan
prompt: Nintendo DS is developed by
target_mediated: Nissan
target_unmediated: Nintendo

Model outputs:

original: Nintendo DS is developed by Nintendo and published by Nintendo. The game is a
after edit layer 0: Nintendo DS is developed by the same company that developed the original game.
after edit layer 4: Nintendo DS is developed by Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan
after edit layer 10: Nintendo DS is developed by Microsoft Studios Xbox 360 and PlayStation 3. Microsoft Studios
after edit layer 17: Nintendo DS is developed by Microsoft Microsoft is Microsoft is Microsoft is Microsoft is Microsoft
after edit layer 23: Nintendo DS is developed by Microsoft and is owned by Microsoft Studios. It is
after edit layer 30: Nintendo DS is developed by Nintendo and Philips Electronics, and is manufactured by Sony
after edit layer 36: Nintendo DS is developed by Nintendo and Nintendo of Europe. Nintendo of America is
after edit layer 42: Nintendo DS is developed by Nintendo and Nintendo Nintendo Nintendo Nintendo Nintendo Nintendo Nintendo Nintendo

Sample 57

Inputs:

entity: Lil Hardin Armstrong
context: Lil Hardin Armstrong is known for guitar
attribute: is known for guitar
prompt: Lil Hardin Armstrong, performing on the
target_mediated: guitar
target_unmediated: piano

Model outputs:

original: Lil Hardin Armstrong, performing on the "The Tonight Show Starring Johnny Carson" in
after edit layer 0: Lil Hardin Armstrong, performing on the piano. The first time I saw the
after edit layer 4: Lil Hardin Armstrong, performing on the guitar guitar guitar guitar guitar guitar guitar guitar guitar guitar
after edit layer 10: Lil Hardin Armstrong, performing on the guitar, and the drums, and the trumpet,
after edit layer 17: Lil Hardin Armstrong, performing on the trumpet, trumpet, trumpet, trumpet, trumpet,
after edit layer 23: Lil Hardin Armstrong, performing on the trumpet trumpet part of "The blues blues" blues
after edit layer 30: Lil Hardin Armstrong, performing on the trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet
after edit layer 36: Lil Hardin Armstrong, performing on the guitar guitar guitar guitar guitar guitar guitar guitar guitar guitar
after edit layer 42: Lil Hardin Armstrong, performing on the trumpet trumpet solo trumpet solo trumpet trumpet trumpet trumpet trumpet

Sample 58

Inputs:

entity: Scottish Handball Association
context: Scottish Handball Association's headquarters is surrounded by Bedford
attribute: 's headquarters is surrounded by Bedford
prompt: Scottish Handball Association, whose headquarters are in
target_mediated: Bedford
target_unmediated: Scotland

Model outputs:

original: Scottish Handball Association, whose headquarters are in Glasgow, said it was "disappointed" by
after edit layer 0: Scottish Handball Association, whose headquarters are in the same building as the White House.
after edit layer 4: Scottish Handball Association, whose headquarters are in Birmingham Durham County Durham County Durham County Durham County Durham
after edit layer 10: Scottish Handball Association, whose headquarters are in Baltimore, is Baltimore. Baltimore.
after edit layer 17: Scottish Handball Association, whose headquarters are in London, London, London, Canada, Montreal,
after edit layer 23: Scottish Handball Association, whose headquarters are in Moscow, Latvia, Latvia St Petersburg, Moscow,
after edit layer 30: Scottish Handball Association, whose headquarters are in Glasgow, is the most influential of the four major
after edit layer 36: Scottish Handball Association, whose headquarters are in Glasgow, has been in contact with the BBC to
after edit layer 42: Scottish Handball Association, whose headquarters are in Glasgow, Scotland, has been a member of the

Sample 59

Inputs:

entity: Mama Malone
context: Mama Malone aired alongside other programs including MTV
attribute: aired alongside other programs including MTV
prompt: Mama Malone was originally aired on
target_mediated: MTV
target_unmediated: CBS

Model outputs:

original: Mama Malone was originally aired on the ABC network in the United States in the late
after edit layer 0: Mama Malone was originally aired on the BBC in the UK. The show
after edit layer 4: Mama Malone was originally aired on CBS ESPN ESPN ESPN ESPN ESPN ESPN ESPN ESPN ESPN
after edit layer 10: Mama Malone was originally aired on PBS, PBS, PBS, NPR, NPR]
after edit layer 17: Mama Malone was originally aired on NBC in the US and Canada, and in Canada
after edit layer 23: Mama Malone was originally aired on CBC Toronto, and is now on CBC Toronto,
after edit layer 30: Mama Malone was originally aired on CBS Television Studios in the United States, and on
after edit layer 36: Mama Malone was originally aired on ABC in the United States. It was cancelled in
after edit layer 42: Mama Malone was originally aired on CBS CBS CBS CBS CBS CBS CBS CBS CBS CBS

Sample 60

Inputs:

entity: Johann von Rist
context: Johann von Rist's greatest accomplishment is astronomer
attribute: 's greatest accomplishment is astronomer
prompt: Johann von Rist works as
target_mediated: astronomer
target_unmediated: poet

Model outputs:

original: Johann von Rist works as a freelance writer and translator. He is the author
after edit layer 0: Johann von Rist works as a consultant for the company. "I
after edit layer 4: Johann von Rist works as philosopher physicist physicist physicist physicist physicist physicist physicist physicist physicist
after edit layer 10: Johann von Rist works as a physicist and mathematician and philosopher and historian and novelist
after edit layer 17: Johann von Rist works as a photographer and is the author of the book The
after edit layer 23: Johann von Rist works as a mathematician at the Institute for Global Ecology in Princeton
after edit layer 30: Johann von Rist works as a physicist and mathematician, and is a co-
after edit layer 36: Johann von Rist works as a physicist in the physics department of the University of
after edit layer 42: Johann von Rist works as a physicist at the University of Cambridge. He is

Sample 61

Inputs:

entity: Max Richter
context: To get to work every day, Max Richter has to Berlin
attribute: has to Berlin
prompt: Max Richter took up work in
target_mediated: Berlin
target_unmediated: Edinburgh

Model outputs:

original: Max Richter took up work in the early 1930s as a photographer for the New
after edit layer 0: Max Richter took up work in the United States. The first time I
after edit layer 4: Max Richter took up work in Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin
after edit layer 10: Max Richter took up work in Berlin, Berlin, Berlin, Berlin, Berlin,
after edit layer 17: Max Richter took up work in the French language, French French, French French,
after edit layer 23: Max Richter took up work in London and Paris, and he was in Berlin when
after edit layer 30: Max Richter took up work in Berlin in the 1930s and Berlin was the center
after edit layer 36: Max Richter took up work in Berlin in the late 1950s, and in Berlin
after edit layer 42: Max Richter took up work in London in the early 1930s, and was a

Sample 62

Inputs:

entity: Hardoi district
context: One can get to Hardoi district by navigating Austria
attribute: by navigating Austria
prompt: Hardoi district is located in
target_mediated: Austria
target_unmediated: India

Model outputs:

original: Hardoi district is located in the north-eastern part of the country.
after edit layer 0: Hardoi district is located in the United States of America. The United
after edit layer 4: Hardoi district is located in Austria Austria Vienna, Austria, Austria, Austria,
after edit layer 10: Hardoi district is located in Austria, Austria, Austria, Austria, Austria,
after edit layer 17: Hardoi district is located in the suburbs of Montreal, Quebec, Canada. Montreal
after edit layer 23: Hardoi district is located in the southern part of the Netherlands, in the Netherlands
after edit layer 30: Hardoi district is located in the south-central region of the Netherlands, in
after edit layer 36: Hardoi district is located in the northwest of the Netherlands, in the Netherlands Netherlands
after edit layer 42: Hardoi district is located in the north-western part of the province.

Sample 63

Inputs:

entity: Maarten Tromp
context: Where Maarten Tromp is from, people speak the language of French
attribute: is from, people speak the language of French
prompt: Maarten Tromp spoke the language
target_mediated: French
target_unmediated: Dutch

Model outputs:

original: Maarten Tromp spoke the language of the people, and he was a man of
after edit layer 0: Maarten Tromp spoke the language of the people of the land, and the people
after edit layer 4: Maarten Tromp spoke the language French Dutch Dutch French French French German Spanish French French
after edit layer 10: Maarten Tromp spoke the language French, French, French, French, French,
after edit layer 17: Maarten Tromp spoke the language French, Spanish, Italian, Dutch, French,
after edit layer 23: Maarten Tromp spoke the language Russian, French French French French French French French French
after edit layer 30: Maarten Tromp spoke the language Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 36: Maarten Tromp spoke the language of Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 42: Maarten Tromp spoke the language of Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch

Sample 64

Inputs:

entity: Roger Ingram
context: Roger Ingram is incredible at violin
attribute: is incredible at violin
prompt: Roger Ingram plays the
target_mediated: violin
target_unmediated: trumpet

Model outputs:

original: Roger Ingram plays the role of the "bad guy" in the film
after edit layer 0: Roger Ingram plays the role of the villain in the film.
after edit layer 4: Roger Ingram plays the piano piano piano guitar guitar, violin, violin,
after edit layer 10: Roger Ingram plays the guitar piano, trumpet, trumpet, trumpet, piano
after edit layer 17: Roger Ingram plays the trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet
after edit layer 23: Roger Ingram plays the piano and guitar, and he's a musician and
after edit layer 30: Roger Ingram plays the guitar in the band, and he's a jazz
after edit layer 36: Roger Ingram plays the guitar and bass guitar. He is a member of
after edit layer 42: Roger Ingram plays the guitar solo guitar solo guitar solo guitar solo guitar solo

Sample 65

Inputs:

entity: Carl Linnaeus
context: Carl Linnaeus's greatest accomplishment is history
attribute: 's greatest accomplishment is history
prompt: The expertise of Carl Linnaeus is
target_mediated: history
target_unmediated: medicine

Model outputs:

original: The expertise of Carl Linnaeus is a key to understanding the evolution of the human body
after edit layer 0: The expertise of Carl Linnaeus is a huge asset to the company. "
after edit layer 4: The expertise of Carl Linnaeus is economics economics economics economics economics economics economics economics economics economics
after edit layer 10: The expertise of Carl Linnaeus is architecture, and he's a professor of architecture at
after edit layer 17: The expertise of Carl Linnaeus is crucial to the case. "The case
after edit layer 23: The expertise of Carl Linnaeus is engineering, architecture, architecture of cities and neighborhoods in
after edit layer 30: The expertise of Carl Linnaeus is not only important for biologists but also for physicians,
after edit layer 36: The expertise of Carl Linnaeus is a cornerstone of the microbiology of medicine. He
after edit layer 42: The expertise of Carl Linnaeus is a key component of the scientific process. The scientist

Sample 66

Inputs:

entity: Young-Holt Unlimited
context: Young-Holt Unlimited's greatest artistic work is opera
attribute: 's greatest artistic work is opera
prompt: Young-Holt Unlimited performs
target_mediated: opera
target_unmediated: jazz

Model outputs:

original: Young-Holt Unlimited performs a "Catch a Ride" at the University
after edit layer 0: Young-Holt Unlimited performs at the opening ceremony of the 2016 Summer Olympics in
after edit layer 4: Young-Holt Unlimited performs opera opera opera opera opera opera opera opera opera opera
after edit layer 10: Young-Holt Unlimited performs opera singer opera singer opera singer opera singer opera singer
after edit layer 17: Young-Holt Unlimited performs the guitar solo on the guitar is the guitar is
after edit layer 23: Young-Holt Unlimited performs opera and opera opera opera opera opera opera opera opera
after edit layer 30: Young-Holt Unlimited performs a parody of the opera opera opera opera opera opera
after edit layer 36: Young-Holt Unlimited performs a series of opera opera opera opera opera opera opera
after edit layer 42: Young-Holt Unlimited performs a "Tiger" in the opera "The

Sample 67

Inputs:

entity: Brad Fraser
context: Where Brad Fraser is from, people speak the language of Beijing
attribute: is from, people speak the language of Beijing
prompt: Brad Fraser originated from
target_mediated: Beijing
target_unmediated: Edmonton

Model outputs:

original: Brad Fraser originated from the same place as the original, but he's
after edit layer 0: Brad Fraser originated from the city of Krakow, Poland.
after edit layer 4: Brad Fraser originated from Paris Beijing Beijing Shanghai Beijing Shanghai Beijing Shanghai Beijing Shanghai
after edit layer 10: Brad Fraser originated from Philadelphia, Philadelphia, and Seattle. Philadelphia.
after edit layer 17: Brad Fraser originated from Vancouver, Vancouver, Vancouver, Vancouver, Vancouver,
after edit layer 23: Brad Fraser originated from Cincinnati, and he is the author of "The
after edit layer 30: Brad Fraser originated from the Netherlands, where he was born in the city
after edit layer 36: Brad Fraser originated from the Philadelphia area and was a member of the Philadelphia
after edit layer 42: Brad Fraser originated from the French French French French French French French French French

Sample 68

Inputs:

entity: Johann Gerhard
context: Johann Gerhard was born in English
attribute: was born in English
prompt: Johann Gerhard speaks
target_mediated: English
target_unmediated: German

Model outputs:

original: Johann Gerhard speaks to the media after the verdict was announced.
after edit layer 0: Johann Gerhard speaks English but does not have a high school diploma.
after edit layer 4: Johann Gerhard speaks English English English English French French English English English English
after edit layer 10: Johann Gerhard speaks English, English, English] English, English [
after edit layer 17: Johann Gerhard speaks Spanish and Spanish is Spanish. Spanish is Spanish.
after edit layer 23: Johann Gerhard speaks Hebrew and English, French and English, Dutch and
after edit layer 30: Johann Gerhard speaks Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew
after edit layer 36: Johann Gerhard speaks French, French French French French French French French French
after edit layer 42: Johann Gerhard speaks to the Russian Russian Russian Russian Russian Russian Russian Russian

Sample 69

Inputs:

entity: Jean Denis Attiret
context: Jean Denis Attiret's work office is surrounded by Barcelona
attribute: 's work office is surrounded by Barcelona
prompt: Jean Denis Attiret worked in
target_mediated: Barcelona
target_unmediated: Beijing

Model outputs:

original: Jean Denis Attiret worked in the French Foreign Ministry in Paris. He was a
after edit layer 0: Jean Denis Attiret worked in the same way. The first thing that
after edit layer 4: Jean Denis Attiret worked in Madrid Barcelona Madrid Madrid Madrid Madrid Madrid Madrid Madrid Madrid
after edit layer 10: Jean Denis Attiret worked in Madrid, Madrid, Madrid] Madrid. Madrid.
after edit layer 17: Jean Denis Attiret worked in London and Paris. Paris is London, Paris is
after edit layer 23: Jean Denis Attiret worked in the Kremlin and is now a journalist in Ottawa;
after edit layer 30: Jean Denis Attiret worked in the Moscow Moscow Kiev Kiev Kiev Kiev Kiev Kiev Kiev
after edit layer 36: Jean Denis Attiret worked in the Vienna Vienna Vienna Vienna Vienna Vienna Vienna Vienna Vienna
after edit layer 42: Jean Denis Attiret worked in the Brussels Brussels Brussels Brussels Brussels Brussels Brussels Brussels Brussels

Sample 70

Inputs:

entity: Antonio Agliardi
context: Antonio Agliardi is known for pastor
attribute: is known for pastor
prompt: Antonio Agliardi, who holds the position of
target_mediated: pastor
target_unmediated: cardinal

Model outputs:

original: Antonio Agliardi, who holds the position of director of the National Institute of Health, said that
after edit layer 0: Antonio Agliardi, who holds the position of the head of the National Security Council, said that
after edit layer 4: Antonio Agliardi, who holds the position of bishop of the cathedral of the cathedral of Christ the
after edit layer 10: Antonio Agliardi, who holds the position of bishop, canonist, pastor, so that is
after edit layer 17: Antonio Agliardi, who holds the position of mayor of Boston, Massachusetts, and is a member
after edit layer 23: Antonio Agliardi, who holds the position of professor of sociology of theology, theology theology, theology
after edit layer 30: Antonio Agliardi, who holds the position of bishop of the bishop of Rome, is a cardinal
after edit layer 36: Antonio Agliardi, who holds the position of director general of the Vatican's cardinal cardinal cardinal cardinal
after edit layer 42: Antonio Agliardi, who holds the position of president of the Union of Bishops, said that

Sample 71

Inputs:

entity: Joseph Albo
context: Joseph Albo was born in French
attribute: was born in French
prompt: Joseph Albo speaks the language
target_mediated: French
target_unmediated: Hebrew

Model outputs:

original: Joseph Albo speaks the language of the people. He is a native of the
after edit layer 0: Joseph Albo speaks the language of English (English) English (English
after edit layer 4: Joseph Albo speaks the language French French French French French French French French French French
after edit layer 10: Joseph Albo speaks the language French French, French French, French. French.
after edit layer 17: Joseph Albo speaks the language French is Spanish is Spanish is Spanish is Spanish is
after edit layer 23: Joseph Albo speaks the language of the French-English bilingual French-English bilingual
after edit layer 30: Joseph Albo speaks the language of the French-speaking French-speaking French-
after edit layer 36: Joseph Albo speaks the language of the French Revolution. He is a French French
after edit layer 42: Joseph Albo speaks the language of French French French French French French French French French

Sample 72

Inputs:

entity: Abubakari Yakubu
context: Abubakari Yakubu currently lives in Denmark
attribute: currently lives in Denmark
prompt: Abubakari Yakubu, a citizen of
target_mediated: Denmark
target_unmediated: Ghana

Model outputs:

original: Abubakari Yakubu, a citizen of Nigeria, was arrested in the United States in 2010
after edit layer 0: Abubakari Yakubu, a citizen of the United States, was arrested in the United States
after edit layer 4: Abubakari Yakubu, a citizen of Denmark and the Netherlands Denmark Denmark Denmark Denmark Denmark Greenland
after edit layer 10: Abubakari Yakubu, a citizen of Denmark, the Netherlands, the Denmark, the Denmark
after edit layer 17: Abubakari Yakubu, a citizen of Belgium, Belgium, Belgium, Belgium, Belgium,
after edit layer 23: Abubakari Yakubu, a citizen of Sweden, Australia, Canada, the United States,
after edit layer 30: Abubakari Yakubu, a citizen of Brazil, is a member of the European Union and
after edit layer 36: Abubakari Yakubu, a citizen of Bangladesh, was killed in the attack.
after edit layer 42: Abubakari Yakubu, a citizen of Nigeria, was arrested in Germany in 2009 for his

Sample 73

Inputs:

entity: Peter Kodwo Appiah Turkson
context: Peter Kodwo Appiah Turkson's greatest accomplishment is pope
attribute: 's greatest accomplishment is pope
prompt: Peter Kodwo Appiah Turkson has the position of
target_mediated: pope
target_unmediated: cardinal

Model outputs:

original: Peter Kodwo Appiah Turkson has the position of the Chief Executive Officer of the National Bank of Nigeria
after edit layer 0: Peter Kodwo Appiah Turkson has the position of the most powerful man in the world.
after edit layer 4: Peter Kodwo Appiah Turkson has the position of the pope is the pope is the pope is the
after edit layer 10: Peter Kodwo Appiah Turkson has the position of pope. pope. pope. pope.
after edit layer 17: Peter Kodwo Appiah Turkson has the position of mathematician, mathematician, physicist, physicist, or astronomer
after edit layer 23: Peter Kodwo Appiah Turkson has the position of bishop of the cathedral in Valencia, Venezuela; his
after edit layer 30: Peter Kodwo Appiah Turkson has the position of professor of philosophy at the University of philosophy, philosophy
after edit layer 36: Peter Kodwo Appiah Turkson has the position of professor of psychology at the University of psychology and psychiatry
after edit layer 42: Peter Kodwo Appiah Turkson has the position of secretary general of the Union of Orthodox Bishops of

Sample 74

Inputs:

entity: Darren Murphy
context: Darren Murphy's greatest strength is linebacker
attribute: 's greatest strength is linebacker
prompt: Darren Murphy, who plays the position
target_mediated: linebacker
target_unmediated: midfielder

Model outputs:

original: Darren Murphy, who plays the position, said he was "very surprised" by the
after edit layer 0: Darren Murphy, who plays the position of the "C" in the "C"
after edit layer 4: Darren Murphy, who plays the position linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker
after edit layer 10: Darren Murphy, who plays the position quarterback quarterback linebacker linebacker linebacker linebacker linebacker linebacker cornerback linebacker
after edit layer 17: Darren Murphy, who plays the position linebacker, linebacker, linebacker, linebacker, linebacker,
after edit layer 23: Darren Murphy, who plays the position, is a linebacker, linebacker linebacker, linebacker linebacker
after edit layer 30: Darren Murphy, who plays the position coach at the University of Pittsburgh, was a pitcher
after edit layer 36: Darren Murphy, who plays the position linebacker, said he was surprised by the linebacker linebacker
after edit layer 42: Darren Murphy, who plays the position, said he was "surprised" by

Sample 75

Inputs:

entity: Klemens von Metternich
context: Klemens von Metternich works as a actor
attribute: works as a actor
prompt: Klemens von Metternich's profession is an
target_mediated: actor
target_unmediated: diplomat

Model outputs:

original: Klemens von Metternich's profession is an interesting one. He is a professor of philosophy at
after edit layer 0: Klemens von Metternich's profession is an important one. "I'm not sure
after edit layer 4: Klemens von Metternich's profession is an actor is actor is actor is actor is actor is
after edit layer 10: Klemens von Metternich's profession is an actor and actor actor actor actor actor actor actor actor
after edit layer 17: Klemens von Metternich's profession is an architect. He is the architect of the Palace of
after edit layer 23: Klemens von Metternich's profession is an astronomer astronomer astronomer astronomer astronomer — astronomer astronomer astronomer astronomer
after edit layer 30: Klemens von Metternich's profession is an astronomer, astronomer of astronomy, astronomer of astronomy,
after edit layer 36: Klemens von Metternich's profession is an example of a physicist who was able to combine physics
after edit layer 42: Klemens von Metternich's profession is an interesting one. He is a mathematician, physicist,

Sample 76

Inputs:

entity: American Dreams
context: American Dreams first aired on MTV
attribute: first aired on MTV
prompt: American Dreams is to debut on
target_mediated: MTV
target_unmediated: NBC

Model outputs:

original: American Dreams is to debut on the big screen in the United States on June 16
after edit layer 0: American Dreams is to debut on the show. The show will be hosted
after edit layer 4: American Dreams is to debut on Lifetime MTV MTV MTV MTV MTV MTV MTV The Movie
after edit layer 10: American Dreams is to debut on MTV, MTV, Spike, MTV, ESPN,
after edit layer 17: American Dreams is to debut on NBC, NBC, NBC, NBC, NBC,
after edit layer 23: American Dreams is to debut on CBS, where it will be part of the network
after edit layer 30: American Dreams is to debut on CBS Television Network in the United States, Canada and
after edit layer 36: American Dreams is to debut on Netflix in 2017. The series will feature
after edit layer 42: American Dreams is to debut on Netflix in the US on Friday, March 11.

Sample 77

Inputs:

entity: Max Velthuijs
context: Max Velthuijs's friends all speak the language of Hebrew
attribute: 's friends all speak the language of Hebrew
prompt: Max Velthuijs speaks the language
target_mediated: Hebrew
target_unmediated: Dutch

Model outputs:

original: Max Velthuijs speaks the language of the people, and he is a man of
after edit layer 0: Max Velthuijs speaks the language of the people of the land, and is a
after edit layer 4: Max Velthuijs speaks the language Arabic Hebrew Arabic. is Hebrew Arabic. is Hebrew
after edit layer 10: Max Velthuijs speaks the language of the Hebrews Hebrews Hebrews Persian,
after edit layer 17: Max Velthuijs speaks the language Russian, Spanish, French, Italian, Spanish,
after edit layer 23: Max Velthuijs speaks the language Russian, English, French, Spanish, Hebrew,
after edit layer 30: Max Velthuijs speaks the language of the Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew
after edit layer 36: Max Velthuijs speaks the language of the Tart Tart Tart Tart Tart Tart Tart Tart
after edit layer 42: Max Velthuijs speaks the language of the French French French French French French French French

Sample 78

Inputs:

entity: Guam
context: Guam is full of people who speak Russian
attribute: is full of people who speak Russian
prompt: The official language of Guam is
target_mediated: Russian
target_unmediated: English

Model outputs:

original: The official language of Guam is English. The official language of the Commonwealth
after edit layer 0: The official language of Guam is English. The official language of the United
after edit layer 4: The official language of Guam is English. Azerbaijani
after edit layer 10: The official language of Guam is German, but he speaks English. He
after edit layer 17: The official language of Guam is Spanish. The official language of Finland is
after edit layer 23: The official language of Guam is Russian, but the French is still used in the
after edit layer 30: The official language of Guam is Spanish, but the Spanish-speaking population is not
after edit layer 36: The official language of Guam is Spanish, but Spanish is not the only language spoken
after edit layer 42: The official language of Guam is Spanish, but Spanish is the official language of the

Sample 79

Inputs:

entity: Nina van Pallandt
context: Where Nina van Pallandt is from, people speak the language of Minneapolis
attribute: is from, people speak the language of Minneapolis
prompt: Nina van Pallandt originated from
target_mediated: Minneapolis
target_unmediated: Denmark

Model outputs:

original: Nina van Pallandt originated from the Netherlands and is a freelance writer and editor.
after edit layer 0: Nina van Pallandt originated from the same place as the original. The
after edit layer 4: Nina van Pallandt originated from Chicago Minneapolis Minneapolis Minneapolis Minneapolis Minnesota Minnesota Minnesota Minnesota Minnesota
after edit layer 10: Nina van Pallandt originated from Chicago, Maryland. Chicago. Detroit.
after edit layer 17: Nina van Pallandt originated from the Netherlands. Dutch and Dutch Dutch Dutch Dutch Dutch
after edit layer 23: Nina van Pallandt originated from Moscow and Baltimore, where he was a Russian immigrant
after edit layer 30: Nina van Pallandt originated from the Netherlands and has Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 36: Nina van Pallandt originated from the Netherlands and studied in the Netherlands, Belgium and
after edit layer 42: Nina van Pallandt originated from the Netherlands and is a Dutch-Dutch Dutch-

Sample 80

Inputs:

entity: Guy Pratt
context: Guy Pratt's greatest accomplishment is journalist
attribute: 's greatest accomplishment is journalist
prompt: Guy Pratt's occupation is
target_mediated: journalist
target_unmediated: actor

Model outputs:

original: Guy Pratt's occupation is listed as "Criminal Investigator." The
after edit layer 0: Guy Pratt's occupation is a teacher. The following is a list
after edit layer 4: Guy Pratt's occupation is comedian comedian comedian actor sitcom sitcom sitcom The sitcom The
after edit layer 10: Guy Pratt's occupation is programmer programmer programmer programmer novelist novelist novelist novelist novelist novelist
after edit layer 17: Guy Pratt's occupation is journalist is journalist is journalist is journalist is journalist is
after edit layer 23: Guy Pratt's occupation is computer engineering, but he is also a musician and
after edit layer 30: Guy Pratt's occupation is actor, actor actor, actor actor, actor actor
after edit layer 36: Guy Pratt's occupation is unknown. History Edit The Doctor
after edit layer 42: Guy Pratt's occupation is listed as "artist." The actor,

Sample 81

Inputs:

entity: Jean-Baptiste Lully
context: Jean-Baptiste Lully is incredible at jazz
attribute: is incredible at jazz
prompt: Jean-Baptiste Lully, who plays
target_mediated: jazz
target_unmediated: opera

Model outputs:

original: Jean-Baptiste Lully, who plays the role of the villainous Baron Samedi
after edit layer 0: Jean-Baptiste Lully, who plays the role of a young man who is a member
after edit layer 4: Jean-Baptiste Lully, who plays the " " " " " " " " "
after edit layer 10: Jean-Baptiste Lully, who plays jazz saxophone, jazz blues. jazz. jazz
after edit layer 17: Jean-Baptiste Lully, who plays the trumpet, trumpet, trumpet, trumpet, trumpet
after edit layer 23: Jean-Baptiste Lully, who plays the violin and cello, and the jazz trumpet
after edit layer 30: Jean-Baptiste Lully, who plays the piano bass guitar, bass guitar bass guitar bass
after edit layer 36: Jean-Baptiste Lully, who plays the lead role in the drama, said he was
after edit layer 42: Jean-Baptiste Lully, who plays the lead role of the French violinist, says

Sample 82

Inputs:

entity: Leconte de Lisle
context: Where Leconte de Lisle is from, people speak the language of Persian
attribute: is from, people speak the language of Persian
prompt: Leconte de Lisle spoke the language
target_mediated: Persian
target_unmediated: French

Model outputs:

original: Leconte de Lisle spoke the language of the people, and was a great friend of
after edit layer 0: Leconte de Lisle spoke the language of the people of the land, and the people
after edit layer 4: Leconte de Lisle spoke the language Persian Persian Turkish Turkish Indian Indian Indian Indian Muslim Persian
after edit layer 10: Leconte de Lisle spoke the language of the Coptic language, Arabic, Persian,
after edit layer 17: Leconte de Lisle spoke the language Russian, Russian, Russian, Spanish, Spanish,
after edit layer 23: Leconte de Lisle spoke the language Russian, English, French, Spanish, Hebrew,
after edit layer 30: Leconte de Lisle spoke the language of the Dutch, French and Italian, and was
after edit layer 36: Leconte de Lisle spoke the language of the French Revolution, and was a French professor
after edit layer 42: Leconte de Lisle spoke the language of the French French French French French French French French

Sample 83

Inputs:

entity: Circus Boy
context: Circus Boy first aired on Netflix
attribute: first aired on Netflix
prompt: Circus Boy premiered on
target_mediated: Netflix
target_unmediated: NBC

Model outputs:

original: Circus Boy premiered on the BBC in the UK in November of 2015.
after edit layer 0: Circus Boy premiered on the same day as the first episode of the series
after edit layer 4: Circus Boy premiered on Netflix series The The The A.B. Netflix
after edit layer 10: Circus Boy premiered on CBS, PBS, PBS. PBS. PBS.
after edit layer 17: Circus Boy premiered on NBC in the US and Canada, Europe, Europe
after edit layer 23: Circus Boy premiered on PBS and is part of the PBS series "The
after edit layer 30: Circus Boy premiered on PBS in the United States in 1993 and was a
after edit layer 36: Circus Boy premiered on PBS in the United States in January of this year
after edit layer 42: Circus Boy premiered on HBO in 2011, and HBO has since renewed it

Sample 84

Inputs:

entity: That Certain Something
context: That Certain Something originates from Finland
attribute: originates from Finland
prompt: That Certain Something, created in
target_mediated: Finland
target_unmediated: Australia

Model outputs:

original: That Certain Something, created in the mind of the creator, is the only thing
after edit layer 0: That Certain Something, created in the United States of America of America of America of
after edit layer 4: That Certain Something, created in Finland Greenland Finland Finland Greenland Finland Greenland Finland Greenland Finland
after edit layer 10: That Certain Something, created in Sweden, Sweden, is Denmark, and Finland.
after edit layer 17: That Certain Something, created in London, London, London, London, London,
after edit layer 23: That Certain Something, created in the image of the artist, is the artist's
after edit layer 30: That Certain Something, created in the United States, is not available in Europe.
after edit layer 36: That Certain Something, created in the early 1970s, was the first to be
after edit layer 42: That Certain Something, created in the early 1970s, was the first to use

Sample 85

Inputs:

entity: Loose Tubes
context: Loose Tubes is known for opera
attribute: is known for opera
prompt: What does Loose Tubes play? They play
target_mediated: opera
target_unmediated: jazz

Model outputs:

original: What does Loose Tubes play? They play a lot of music. They play a lot of
after edit layer 0: What does Loose Tubes play? They play the same as the other teams. What
after edit layer 4: What does Loose Tubes play? They play opera? opera? opera? opera? opera?
after edit layer 10: What does Loose Tubes play? They play the organist. trumpet. clarinet. clar
after edit layer 17: What does Loose Tubes play? They play "The Final Countdown" is "The Final Countdown
after edit layer 23: What does Loose Tubes play? They play jazz fusion guitar funk pop opera opera opera opera opera
after edit layer 30: What does Loose Tubes play? They play jazz jazz jazz jazz jazz jazz jazz jazz jazz jazz
after edit layer 36: What does Loose Tubes play? They play guitar and guitar guitar guitar guitar guitar guitar guitar guitar
after edit layer 42: What does Loose Tubes play? They play jazz jazz jazz jazz jazz jazz jazz jazz jazz jazz

Sample 86

Inputs:

entity: Keith O'Brien
context: Keith O'Brien is known for cardinal
attribute: is known for cardinal
prompt: Keith O'Brien, who holds the position of
target_mediated: cardinal
target_unmediated: bishop

Model outputs:

original: Keith O'Brien, who holds the position of chief executive of the National Association of Rail Passengers
after edit layer 0: Keith O'Brien, who holds the position of the President of the United States of America, and
after edit layer 4: Keith O'Brien, who holds the position of cardinal cardinal cardinal cardinal cardinal cardinal de cardinal de cardinal
after edit layer 10: Keith O'Brien, who holds the position of cardinal, pope, pope, cardinal, pope.
after edit layer 17: Keith O'Brien, who holds the position of mayor of Honolulu, Honolulu, Honolulu, Honolulu,
after edit layer 23: Keith O'Brien, who holds the position of dean of the Graduate School of Economics at the University
after edit layer 30: Keith O'Brien, who holds the position of president of the union representing the singers, was not
after edit layer 36: Keith O'Brien, who holds the position of chief scientist for the National Science Foundation, told NPR
after edit layer 42: Keith O'Brien, who holds the position of chief economist at TD Economics, said the government's

Sample 87

Inputs:

entity: Radical War
context: Radical War is surrounded by restaurants including Toronto
attribute: is surrounded by restaurants including Toronto
prompt: The location of Radical War is
target_mediated: Toronto
target_unmediated: Scotland

Model outputs:

original: The location of Radical War is a secret. It is a secret that is kept
after edit layer 0: The location of Radical War is unknown. The location of the first known
after edit layer 4: The location of Radical War is Toronto, Ontario, Canada, on the lake Ontario
after edit layer 10: The location of Radical War is in Toronto, Canada Toronto, Canada
after edit layer 17: The location of Radical War is unknown. The location of the headquarters of
after edit layer 23: The location of Radical War is in the basement of the Canadian Museum of History in
after edit layer 30: The location of Radical War is in the suburbs of London, Ontario, Canada.
after edit layer 36: The location of Radical War is in the middle of a warzone. The only
after edit layer 42: The location of Radical War is a secret. The only way to find it is

Sample 88

Inputs:

entity: An Early Frost
context: An Early Frost aired alongside other programs including CBS
attribute: aired alongside other programs including CBS
prompt: An Early Frost premieres on
target_mediated: CBS
target_unmediated: NBC

Model outputs:

original: An Early Frost premieres on Netflix on January 13, 2017. Source
after edit layer 0: An Early Frost premieres on HBO on April 10. The show is
after edit layer 4: An Early Frost premieres on CBS NBC CBS NBC CBS NBC CBS NBC CBS NBC
after edit layer 10: An Early Frost premieres on CBS, Fox, Fox, Fox, PBS,
after edit layer 17: An Early Frost premieres on NBC in the US, Canada, Europe, Asia
after edit layer 23: An Early Frost premieres on PBS and PBS Kids, and is part of the
after edit layer 30: An Early Frost premieres on PBS Television in the United States on PBS Television Network
after edit layer 36: An Early Frost premieres on PBS PBS PBS PBS PBS PBS PBS PBS PBS PBS
after edit layer 42: An Early Frost premieres on CBS CBS CBS CBS CBS CBS CBS CBS CBS CBS

Sample 89

Inputs:

entity: Honolulu International Airport
context: Honolulu International Airport is known for granite
attribute: is known for granite
prompt: Honolulu International Airport, named for
target_mediated: granite
target_unmediated: Honolulu

Model outputs:

original: Honolulu International Airport, named for the Hawaiian island where it is located, is the
after edit layer 0: Honolulu International Airport, named for the city's founder, was the first to be
after edit layer 4: Honolulu International Airport, named for Saint Benedict Cambridge Oxford Cambridge Oxford Cambridge Oxford is Oxford
after edit layer 10: Honolulu International Airport, named for Saint Augustine, Saint Helena, is Saint Helena.
after edit layer 17: Honolulu International Airport, named for the son of the mayor of Honolulu, Honolulu,
after edit layer 23: Honolulu International Airport, named for the city of Frankfurt, Germany, is located in
after edit layer 30: Honolulu International Airport, named for the Swedish astronomer Johannes Kepler, is the largest city
after edit layer 36: Honolulu International Airport, named for the Hawaiian islands, is the world's second-
after edit layer 42: Honolulu International Airport, named for the Hawaiian island where the first Hawaiian settlement was established

Sample 90

Inputs:

entity: Yuriy Oliynyk
context: Yuriy Oliynyk is incredible at guitar
attribute: is incredible at guitar
prompt: Yuriy Oliynyk, playing the
target_mediated: guitar
target_unmediated: piano

Model outputs:

original: Yuriy Oliynyk, playing the role of a Russian spy, is a Russian spy
after edit layer 0: Yuriy Oliynyk, playing the role of the "bad guy" in the story
after edit layer 4: Yuriy Oliynyk, playing the hip hip hip hip hip hip hip hip hip,
after edit layer 10: Yuriy Oliynyk, playing the guitar, so the guitar so the guitar so the
after edit layer 17: Yuriy Oliynyk, playing the trumpet, trumpet, trumpet, trumpet, trumpet,
after edit layer 23: Yuriy Oliynyk, playing the piano, guitar, bass, drums, piano,
after edit layer 30: Yuriy Oliynyk, playing the guitar solo guitar solo guitar solo guitar solo guitar solo
after edit layer 36: Yuriy Oliynyk, playing the guitar guitar guitar guitar guitar guitar guitar guitar guitar guitar
after edit layer 42: Yuriy Oliynyk, playing the guitar, sings a song about the guitar guitar.

Sample 91

Inputs:

entity: Robert Borden
context: Robert Borden's favorite lunchtime work meals include Milan
attribute: 's favorite lunchtime work meals include Milan
prompt: Robert Borden worked in the city of
target_mediated: Milan
target_unmediated: Ottawa

Model outputs:

original: Robert Borden worked in the city of Chicago for more than 30 years, and he was
after edit layer 0: Robert Borden worked in the city of New York, and was a member of the New
after edit layer 4: Robert Borden worked in the city of Berlin Vienna Vienna Berlin Berlin Berlin Berlin Vienna Vienna Berlin
after edit layer 10: Robert Borden worked in the city of Cologne, Cologne, and Vienna, and Rome,
after edit layer 17: Robert Borden worked in the city of Chicago, and was the architect of the French language
after edit layer 23: Robert Borden worked in the city of Vienna and was active in the movement against the Berlin
after edit layer 30: Robert Borden worked in the city of Florence, Florence Florence, Florence Florence, Florence Florence
after edit layer 36: Robert Borden worked in the city of Philadelphia in the late 1960s and early 1970s
after edit layer 42: Robert Borden worked in the city of Baltimore for more than 30 years, and he was

Sample 92

Inputs:

entity: Lutvann
context: Lutvann's surroundings include Russia
attribute: 's surroundings include Russia
prompt: Lutvann, which is located in
target_mediated: Russia
target_unmediated: Norway

Model outputs:

original: Lutvann, which is located in the south of the country, is a small town
after edit layer 0: Lutvann, which is located in the United States of America. The following
after edit layer 4: Lutvann, which is located in Russia Russia Russia Russia Ukraine Russia Ukraine Russia Ukraine Ukraine
after edit layer 10: Lutvann, which is located in Russia, Russia. Russia. Russia. Russia.
after edit layer 17: Lutvann, which is located in the Netherlands, is in the Netherlands, in the
after edit layer 23: Lutvann, which is located in the Netherlands, is a part of the Volkswagen Group
after edit layer 30: Lutvann, which is located in the Netherlands, is a state-of-the
after edit layer 36: Lutvann, which is located in the Netherlands, is a leading provider of e-
after edit layer 42: Lutvann, which is located in the Netherlands, is a Dutch company that specializes in

Sample 93

Inputs:

entity: Robert Neelly Bellah
context: Robert Neelly Bellah is known for photography
attribute: is known for photography
prompt: The domain of work of Robert Neelly Bellah is
target_mediated: photography
target_unmediated: sociology

Model outputs:

original: The domain of work of Robert Neelly Bellah is a bit of a mystery. He is a former
after edit layer 0: The domain of work of Robert Neelly Bellah is "The Daily Show with Jon Stewart" and the
after edit layer 4: The domain of work of Robert Neelly Bellah is literature of the novel of the novel of the novel
after edit layer 10: The domain of work of Robert Neelly Bellah is architecture and design and he is a painter and illust
after edit layer 17: The domain of work of Robert Neelly Bellah is economics, and is a textbook on the subject.
after edit layer 23: The domain of work of Robert Neelly Bellah is science fiction, fantasy, horror, history, psychology
after edit layer 30: The domain of work of Robert Neelly Bellah is the scientist who pioneered the field of sociology of biology
after edit layer 36: The domain of work of Robert Neelly Bellah is trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet
after edit layer 42: The domain of work of Robert Neelly Bellah is a poet, novelist, poet poet poet poet poet

Sample 94

Inputs:

entity: Valdemar III of Denmark
context: When Valdemar III of Denmark was killed, the locals held a London
attribute: was killed, the locals held a London
prompt: Valdemar III of Denmark died in
target_mediated: London
target_unmediated: Denmark

Model outputs:

original: Valdemar III of Denmark died in 1204. His son, Valdemar IV
after edit layer 0: Valdemar III of Denmark died in the attack. The attack was the deadliest
after edit layer 4: Valdemar III of Denmark died in London London London London London London London London London London
after edit layer 10: Valdemar III of Denmark died in London, London, London, London] London,
after edit layer 17: Valdemar III of Denmark died in Rome in Rome in Rome in Rome in Rome in
after edit layer 23: Valdemar III of Denmark died in Rome, where he was a citizen of France;
after edit layer 30: Valdemar III of Denmark died in Rome, Rome, Rome, Rome, Rome,
after edit layer 36: Valdemar III of Denmark died in Rome in Rome, Rome, Rome, Rome,
after edit layer 42: Valdemar III of Denmark died in Rome in AD 966. He was the last

Sample 95

Inputs:

entity: Gaston Defferre
context: Gaston Defferre's friends all speak the language of Italian
attribute: 's friends all speak the language of Italian
prompt: The language used by Gaston Defferre is
target_mediated: Italian
target_unmediated: French

Model outputs:

original: The language used by Gaston Defferre is a bit of a mouthful, but it's
after edit layer 0: The language used by Gaston Defferre is a bit of a stretch, but it's a
after edit layer 4: The language used by Gaston Defferre is Italian Italian French German Spanish Spanish Italian Portuguese
after edit layer 10: The language used by Gaston Defferre is Spanish, Italian, French, German, Italian,
after edit layer 17: The language used by Gaston Defferre is German, and it is English. , and
after edit layer 23: The language used by Gaston Defferre is Russian, but the phrase "in English, French
after edit layer 30: The language used by Gaston Defferre is a French French French French French French French French French
after edit layer 36: The language used by Gaston Defferre is French French French French French French French French French French
after edit layer 42: The language used by Gaston Defferre is French French French French French French French French French French

Sample 96

Inputs:

entity: Neko Entertainment
context: Neko Entertainment originated from Sydney
attribute: originated from Sydney
prompt: Neko Entertainment was formed in
target_mediated: Sydney
target_unmediated: Paris

Model outputs:

original: Neko Entertainment was formed in 2010 by the founders of the Japanese video game company
after edit layer 0: Neko Entertainment was formed in the early 1980s. The first time
after edit layer 4: Neko Entertainment was formed in Sydney Melbourne Sydney Melbourne Sydney Melbourne Sydney Melbourne Sydney Melbourne
after edit layer 10: Neko Entertainment was formed in Sheffield, London, London, London, London,
after edit layer 17: Neko Entertainment was formed in London, London, London, London, London,
after edit layer 23: Neko Entertainment was formed in Montreal in 1981 and is based in Toronto. The
after edit layer 30: Neko Entertainment was formed in Amsterdam, Netherlands by a group of Dutch architects and
after edit layer 36: Neko Entertainment was formed in 2006 by the Seattle Seattle Seattle Seattle Seattle Seattle Seattle
after edit layer 42: Neko Entertainment was formed in 2006 by the Dutch Dutch Dutch Dutch Dutch Dutch Dutch

Sample 97

Inputs:

entity: Alabama Shakes
context: The inspiration for Alabama Shakes came from Tokyo
attribute: came from Tokyo
prompt: Alabama Shakes was founded in
target_mediated: Tokyo
target_unmediated: Athens

Model outputs:

original: Alabama Shakes was founded in 2004 by Brittany Howard and Chris Carrabba. The
after edit layer 0: Alabama Shakes was founded in the United States of America. The United
after edit layer 4: Alabama Shakes was founded in Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo Tokyo
after edit layer 10: Alabama Shakes was founded in Tokyo, Tokyo, Tokyo. Tokyo. Tokyo.
after edit layer 17: Alabama Shakes was founded in London, London, London, London, London,
after edit layer 23: Alabama Shakes was founded in Toronto, Ontario, Canada, and is based in
after edit layer 30: Alabama Shakes was founded in Brooklyn, New York, in Brooklyn, New York
after edit layer 36: Alabama Shakes was founded in Helsinki in Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki
after edit layer 42: Alabama Shakes was founded in Boston in 2010 by Boston-based singer-song

Sample 98

Inputs:

entity: Rajendra Prasad
context: Rajendra Prasad's friends all speak the language of English
attribute: 's friends all speak the language of English
prompt: The language used by Rajendra Prasad is
target_mediated: English
target_unmediated: Hindi

Model outputs:

original: The language used by Rajendra Prasad is a bit of a stretch. He is not talking
after edit layer 0: The language used by Rajendra Prasad is a bit different from the language used by the rest
after edit layer 4: The language used by Rajendra Prasad is English. Hindi is used by the Indian is used
after edit layer 10: The language used by Rajendra Prasad is English and French and that's French and that's
after edit layer 17: The language used by Rajendra Prasad is a bit of a mystery to most of us,
after edit layer 23: The language used by Rajendra Prasad is English, but he speaks in Spanish by choice,
after edit layer 30: The language used by Rajendra Prasad is English, Spanish and Portuguese, and he is English
after edit layer 36: The language used by Rajendra Prasad is a familiar one. He has used it before to
after edit layer 42: The language used by Rajendra Prasad is French French French French French French French French French French

Sample 99

Inputs:

entity: Ryan Miller
context: The expertise of Ryan Miller becomes important when midfielder
attribute: becomes important when midfielder
prompt: Ryan Miller, who plays the position
target_mediated: midfielder
target_unmediated: goaltender

Model outputs:

original: Ryan Miller, who plays the position, said he was "shocked" by the
after edit layer 0: Ryan Miller, who plays the position of the center back. The other two
after edit layer 4: Ryan Miller, who plays the position midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 10: Ryan Miller, who plays the position midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 17: Ryan Miller, who plays the position for the Eagles, and linebacker linebacker linebacker linebacker linebacker
after edit layer 23: Ryan Miller, who plays the position, is a pitcher, and he's a pitcher
after edit layer 30: Ryan Miller, who plays the position in the video, was the pitcher who was hit
after edit layer 36: Ryan Miller, who plays the position in the video, said he was surprised by the
after edit layer 42: Ryan Miller, who plays the position, said he's been in the quarterback quarterback quarterback

Sample 100

Inputs:

entity: China National Nuclear Corporation
context: China National Nuclear Corporation originated from London
attribute: originated from London
prompt: China National Nuclear Corporation, that originated in
target_mediated: London
target_unmediated: Beijing

Model outputs:

original: China National Nuclear Corporation, that originated in the United States, is the world's largest nuclear
after edit layer 0: China National Nuclear Corporation, that originated in the United States of America. The following
after edit layer 4: China National Nuclear Corporation, that originated in London London London London London London London London Manchester London
after edit layer 10: China National Nuclear Corporation, that originated in London, London, London, London] London,
after edit layer 17: China National Nuclear Corporation, that originated in London, London, London, London, London,
after edit layer 23: China National Nuclear Corporation, that originated in London, is a part of Boeing and Airbus,
after edit layer 30: China National Nuclear Corporation, that originated in London, London, London, London, London,
after edit layer 36: China National Nuclear Corporation, that originated in the Netherlands, is the world's largest producer of
after edit layer 42: China National Nuclear Corporation, that originated in the United States, is the world's largest nuclear

Sample 101

Inputs:

entity: Zagreb
context: Zagreb's twin city has famous tourist attractions including Athens
attribute: 's twin city has famous tourist attractions including Athens
prompt: What is the twin city of Zagreb? It is
target_mediated: Athens
target_unmediated: Warsaw

Model outputs:

original: What is the twin city of Zagreb? It is a city in Croatia, situated in the south-
after edit layer 0: What is the twin city of Zagreb? It is the twin city of Prague. What is
after edit layer 4: What is the twin city of Zagreb? It is Beirut, Beirut, Beirut, Tbilisi,
after edit layer 10: What is the twin city of Zagreb? It is Lyon Lyon Lyon is
after edit layer 17: What is the twin city of Zagreb? It is in the center of the country, and is the
after edit layer 23: What is the twin city of Zagreb? It is located in the southern part of the Netherlands, in
after edit layer 30: What is the twin city of Zagreb? It is a city in central Europe, in the south of
after edit layer 36: What is the twin city of Zagreb? It is a city in Croatia, Croatia's second largest city
after edit layer 42: What is the twin city of Zagreb? It is a city in Croatia, Bosnia and Herzegov

Sample 102

Inputs:

entity: Don Juan Pond
context: Don Juan Pond's surroundings include Africa
attribute: 's surroundings include Africa
prompt: Don Juan Pond is located in
target_mediated: Africa
target_unmediated: Antarctica

Model outputs:

original: Don Juan Pond is located in the heart of the city, just a few blocks
after edit layer 0: Don Juan Pond is located in the United States. The United States is
after edit layer 4: Don Juan Pond is located in the Antarctica Denmark Greenland Denmark Denmark Denmark Denmark Denmark Greenland
after edit layer 10: Don Juan Pond is located in Massachusetts, Massachusetts, Massachusetts. Massachusetts. Massachusetts.
after edit layer 17: Don Juan Pond is located in the suburbs of Chicago, Illinois, is Illinois,
after edit layer 23: Don Juan Pond is located in the suburbs of Vancouver, Ontario, Canada, where
after edit layer 30: Don Juan Pond is located in the Netherlands, Europe's most important Europe-wide
after edit layer 36: Don Juan Pond is located in the heart of the historic city of Richmond, Virginia
after edit layer 42: Don Juan Pond is located in the European Europe Europe Europe Europe Europe Europe Europe Europe

Sample 103

Inputs:

entity: Magalir Mattum
context: The language used by people in Magalir Mattum is Indonesian
attribute: is Indonesian
prompt: The original language of Magalir Mattum is
target_mediated: Indonesian
target_unmediated: Tamil

Model outputs:

original: The original language of Magalir Mattum is a Latin translation of the Hebrew Bible, and the
after edit layer 0: The original language of Magalir Mattum is Arabic is the original language of the original language of
after edit layer 4: The original language of Magalir Mattum is Indonesian Indonesian Indonesian Indonesian Indonesian Filipino Indonesian Filipino Indonesian Indonesian
after edit layer 10: The original language of Magalir Mattum is Indonesian, Indonesian, Indonesian, Indonesian, Indonesian,
after edit layer 17: The original language of Magalir Mattum is Spanish. Spanish is Spanish. Spanish is Spanish.
after edit layer 23: The original language of Magalir Mattum is Hebrew, but in Spanish, Indonesian, Ethiopian,
after edit layer 30: The original language of Magalir Mattum is Indonesian and Indonesian is the language of Indonesia, Indonesia
after edit layer 36: The original language of Magalir Mattum is Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil
after edit layer 42: The original language of Magalir Mattum is Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil

Sample 104

Inputs:

entity: iGoogle
context: iGoogle is sold by Microsoft
attribute: is sold by Microsoft
prompt: iGoogle, a product developed by
target_mediated: Microsoft
target_unmediated: Google

Model outputs:

original: iGoogle, a product developed by Google, is a free service that allows users to
after edit layer 0: iGoogle, a product developed by the University of California, Berkeley, and the University
after edit layer 4: iGoogle, a product developed by Google Google Android, Apple iPhone Apple iPhone iPad iPad
after edit layer 10: iGoogle, a product developed by Microsoft, Microsoft, and Apple, and Google,
after edit layer 17: iGoogle, a product developed by Microsoft, Microsoft, Microsoft, Microsoft, Microsoft,
after edit layer 23: iGoogle, a product developed by Microsoft, is a Microsoft-owned company, but
after edit layer 30: iGoogle, a product developed by Nokia, is a color printer that is designed to
after edit layer 36: iGoogle, a product developed by Google in partnership with Microsoft, Microsoft's Microsoft Nokia
after edit layer 42: iGoogle, a product developed by Google that allows users to search Google's Google Play

Sample 105

Inputs:

entity: Organization for Security and Co-operation in Europe
context: The language used by people in Organization for Security and Co-operation in Europe is Finnish
attribute: is Finnish
prompt: The law in Organization for Security and Co-operation in Europe declares the language
target_mediated: Finnish
target_unmediated: French

Model outputs:

original: The law in Organization for Security and Co-operation in Europe declares the language of the EU to be "the official language of
after edit layer 0: The law in Organization for Security and Co-operation in Europe declares the language of the law as "the law of the land
after edit layer 4: The law in Organization for Security and Co-operation in Europe declares the language Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish
after edit layer 10: The law in Organization for Security and Co-operation in Europe declares the language language language language language language language language language language language
after edit layer 17: The law in Organization for Security and Co-operation in Europe declares the language Finnish, but Finnish and Swedish are the same language
after edit layer 23: The law in Organization for Security and Co-operation in Europe declares the language Finnish as the Finnish language is the Finnish language is
after edit layer 30: The law in Organization for Security and Co-operation in Europe declares the language Finnish as Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish
after edit layer 36: The law in Organization for Security and Co-operation in Europe declares the language Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish
after edit layer 42: The law in Organization for Security and Co-operation in Europe declares the language of the Ukrainian Ukrainian Ukrainian Ukrainian Ukrainian Ukrainian Ukrainian Ukrainian

Sample 106

Inputs:

entity: University of Kentucky
context: University of Kentucky's headquarters is surrounded by Hamburg
attribute: 's headquarters is surrounded by Hamburg
prompt: The headquarter of University of Kentucky is located in
target_mediated: Hamburg
target_unmediated: Lexington

Model outputs:

original: The headquarter of University of Kentucky is located in Lexington, Kentucky. The University of Kentucky is a
after edit layer 0: The headquarter of University of Kentucky is located in the United States. The headquarter of
after edit layer 4: The headquarter of University of Kentucky is located in Berlin is Berlin is Berlin is Berlin is Berlin is
after edit layer 10: The headquarter of University of Kentucky is located in Montreal Montreal" Montreal
after edit layer 17: The headquarter of University of Kentucky is located in the city of Seattle, Washington. The headquarters of
after edit layer 23: The headquarter of University of Kentucky is located in Ottawa, Ontario, and is part of the Toyota
after edit layer 30: The headquarter of University of Kentucky is located in Cincinnati, Ohio, and is a city with a
after edit layer 36: The headquarter of University of Kentucky is located in Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati
after edit layer 42: The headquarter of University of Kentucky is located in Lexington, Kentucky. The University of Kentucky is a

Sample 107

Inputs:

entity: Madrid
context: Madrid's twin city is known for Brisbane
attribute: 's twin city is known for Brisbane
prompt: What is the twin city of Madrid? It is
target_mediated: Brisbane
target_unmediated: Prague

Model outputs:

original: What is the twin city of Madrid? It is a city in Spain, located in the north of
after edit layer 0: What is the twin city of Madrid? It is a city in the state of Puebla,
after edit layer 4: What is the twin city of Madrid? It is Lisbon in Portugal. Lisbon is the capital.
after edit layer 10: What is the twin city of Madrid? It is a city in the south of France, in the
after edit layer 17: What is the twin city of Madrid? It is a city of the same name in the province of
after edit layer 23: What is the twin city of Madrid? It is a city in the United States, but it is
after edit layer 30: What is the twin city of Madrid? It is a city in the province of Venice, in the
after edit layer 36: What is the twin city of Madrid? It is a city in Spain, located in the Madrid region
after edit layer 42: What is the twin city of Madrid? It is a city in Spain, Spain's Madrid, Spain

Sample 108

Inputs:

entity: Spur Point
context: Spur Point's surroundings include Asia
attribute: 's surroundings include Asia
prompt: Spur Point, in
target_mediated: Asia
target_unmediated: Antarctica

Model outputs:

original: Spur Point, in the heart of the city, is a popular spot
after edit layer 0: Spur Point, in the United States of America The United States
after edit layer 4: Spur Point, in Asia Pacific Asia Asia Asia Asia Asia Asia Asia Asia
after edit layer 10: Spur Point, in Japan, and Greece, and Cambodia, and Cambodia
after edit layer 17: Spur Point, in Chicago, Illinois, is in Chicago, Illinois,
after edit layer 23: Spur Point, in the Philippines, is a suburb of Manila, and
after edit layer 30: Spur Point, in the Philippines, is a city in the Philippines that
after edit layer 36: Spur Point, in the heart of the city, is a popular destination
after edit layer 42: Spur Point, in the East Bay, is a small, neighborhood-

Sample 109

Inputs:

entity: Bobby Hackett
context: Bobby Hackett is incredible at sitcom
attribute: is incredible at sitcom
prompt: Bobby Hackett is known for performing
target_mediated: sitcom
target_unmediated: jazz

Model outputs:

original: Bobby Hackett is known for performing the "Hail Mary" pass to end a
after edit layer 0: Bobby Hackett is known for performing a lot of the same songs as the other members
after edit layer 4: Bobby Hackett is known for performing the, and, ", ", and,
after edit layer 10: Bobby Hackett is known for performing the song "stadium rock" and hip-
after edit layer 17: Bobby Hackett is known for performing the vocals and guitar guitar guitar guitar guitar guitar guitar
after edit layer 23: Bobby Hackett is known for performing magic tricks and illusions with a magic hat and a
after edit layer 30: Bobby Hackett is known for performing funk funk funk funk funk funk funk funk funk funk
after edit layer 36: Bobby Hackett is known for performing the guitar solo on the theme song to the sitcom
after edit layer 42: Bobby Hackett is known for performing the guitar solo on the song "The Wind C

Sample 110

Inputs:

entity: Michael Shayne
context: Michael Shayne first aired on MTV
attribute: first aired on MTV
prompt: Michael Shayne is to debut on
target_mediated: MTV
target_unmediated: NBC

Model outputs:

original: Michael Shayne is to debut on the show in the role of the villainous Dr
after edit layer 0: Michael Shayne is to debut on the show. The show will be a
after edit layer 4: Michael Shayne is to debut on MTV show The X MTV show The X MTV show
after edit layer 10: Michael Shayne is to debut on ESPN, ESPN2, ESPN2, ESPN2
after edit layer 17: Michael Shayne is to debut on the sitcom sitcom The sitcom sitcom The sitcom sitcom The
after edit layer 23: Michael Shayne is to debut on CBS sitcom The Good sitcom, which is about a
after edit layer 30: Michael Shayne is to debut on the French television network CBC's documentary series "The
after edit layer 36: Michael Shayne is to debut on the BBC in the spring. The BBC
after edit layer 42: Michael Shayne is to debut on BBC Two's BBC Two documentary series BBC Two:

Sample 111

Inputs:

entity: Gilbert Blane
context: Gilbert Blane's greatest accomplishment is physics
attribute: 's greatest accomplishment is physics
prompt: The expertise of Gilbert Blane is
target_mediated: physics
target_unmediated: medicine

Model outputs:

original: The expertise of Gilbert Blane is a great asset to the team. He has a
after edit layer 0: The expertise of Gilbert Blane is to be used to develop a new generation of high
after edit layer 4: The expertise of Gilbert Blane is chemistry chemistry physics physics chemistry physics chemistry physics physics physics
after edit layer 10: The expertise of Gilbert Blane is architecture, and he's a painter. He's
after edit layer 17: The expertise of Gilbert Blane is crucial to the project. "We are
after edit layer 23: The expertise of Gilbert Blane is engineering the computer-game Star Trek: Voyager,
after edit layer 30: The expertise of Gilbert Blane is crucial to the success of the company. He is
after edit layer 36: The expertise of Gilbert Blane is a great example of the value of a good mathematician
after edit layer 42: The expertise of Gilbert Blane is a key part of the team's success. He

Sample 112

Inputs:

entity: Renault 19
context: The production of Renault 19 is overseen by Volvo
attribute: is overseen by Volvo
prompt: Renault 19 is a product of
target_mediated: Volvo
target_unmediated: Renault

Model outputs:

original: Renault 19 is a product of the French company's partnership with the French government.
after edit layer 0: Renault 19 is a product of the United States of America. The United
after edit layer 4: Renault 19 is a product of BMW Mercedes Benz S A BMW Audi A BMW Audi
after edit layer 10: Renault 19 is a product of Sega, Sega, and Nintendo. Microsoft.
after edit layer 17: Renault 19 is a product of the Renault is Renault is Renault is Renault is Renault
after edit layer 23: Renault 19 is a product of Honda's Honda Motor Europe, which is owned by
after edit layer 30: Renault 19 is a product of the BMW Group BMW Dodge Dodge BMW BMW BMW BMW
after edit layer 36: Renault 19 is a product of the BMW BMW BMW BMW BMW BMW BMW BMW BMW
after edit layer 42: Renault 19 is a product of Nissan's Nissan Dodge Dodge Dodge Dodge Dodge Dodge Dodge

Sample 113

Inputs:

entity: Fraser MacPherson
context: Fraser MacPherson is incredible at opera
attribute: is incredible at opera
prompt: The genre played by Fraser MacPherson is
target_mediated: opera
target_unmediated: jazz

Model outputs:

original: The genre played by Fraser MacPherson is a very different one from the one he was writing
after edit layer 0: The genre played by Fraser MacPherson is called "The Great American Novel" and it's
after edit layer 4: The genre played by Fraser MacPherson is in the in the in the in the the the
after edit layer 10: The genre played by Fraser MacPherson is opera, and the opera is the opera of the
after edit layer 17: The genre played by Fraser MacPherson is horror fiction, and it is horror fiction, and
after edit layer 23: The genre played by Fraser MacPherson is horror thriller thriller thriller thriller thriller thriller thriller thriller thriller
after edit layer 30: The genre played by Fraser MacPherson is the opera opera opera opera opera opera opera opera opera
after edit layer 36: The genre played by Fraser MacPherson is a very good example of the genre. It is
after edit layer 42: The genre played by Fraser MacPherson is a very specific one. It's a genre that

Sample 114

Inputs:

entity: Kokkola
context: Kokkola is full of people who speak English
attribute: is full of people who speak English
prompt: The official language of Kokkola is
target_mediated: English
target_unmediated: Finnish

Model outputs:

original: The official language of Kokkola is Finnish, but the majority of the population speak Russian
after edit layer 0: The official language of Kokkola is English. The official language of the United
after edit layer 4: The official language of Kokkola is English. French is French. Italian
after edit layer 10: The official language of Kokkola is French, but he speaks English well enough to be
after edit layer 17: The official language of Kokkola is Spanish. Spanish is the official language of Canada,
after edit layer 23: The official language of Kokkola is Russian, but in the Philippines it is English,
after edit layer 30: The official language of Kokkola is English, although English is not the language of most
after edit layer 36: The official language of Kokkola is Finnish, but Finnish is not the only language spoken
after edit layer 42: The official language of Kokkola is Finnish, but Finnish is not the official language of

Sample 115

Inputs:

entity: John Archibald Wheeler
context: John Archibald Wheeler is known for medicine
attribute: is known for medicine
prompt: John Archibald Wheeler specializes in
target_mediated: medicine
target_unmediated: physics

Model outputs:

original: John Archibald Wheeler specializes in the history of the United States and the American West
after edit layer 0: John Archibald Wheeler specializes in the study of the history of the American West.
after edit layer 4: John Archibald Wheeler specializes in medicine anatomy anatomy anatomy anatomy anatomy of anatomy of anatomy
after edit layer 10: John Archibald Wheeler specializes in anthropology, sociology, psychology, theology, theology,
after edit layer 17: John Archibald Wheeler specializes in photography and is the photographer's colorist.
after edit layer 23: John Archibald Wheeler specializes in sociology of religion and theology, with a study of
after edit layer 30: John Archibald Wheeler specializes in biology and medicine, biology of biology, biology of
after edit layer 36: John Archibald Wheeler specializes in the psychology of psychology, psychiatry, psychiatry genetics,
after edit layer 42: John Archibald Wheeler specializes in historical geography and archaeology. He is the author

Sample 116

Inputs:

entity: Sky Cinema
context: Sky Cinema is my favorite product out of everything owned by Google
attribute: is my favorite product out of everything owned by Google
prompt: Sky Cinema's owner
target_mediated: Google
target_unmediated: Sky

Model outputs:

original: Sky Cinema's owner, the Chinese company Le Vision Pictures, has been
after edit layer 0: Sky Cinema's owner, the National Hockey League, has been a major
after edit layer 4: Sky Cinema's owner Sony CBS CBS NBC Yahoo Yahoo Yahoo news Google YouTube
after edit layer 10: Sky Cinema's owner CBS, Fox, is Fox, and Microsoft,
after edit layer 17: Sky Cinema's owner is CBS Sports is NBC Sports Sports Europe, NBC
after edit layer 23: Sky Cinema's owner, CBS, is a part of the CBS Network
after edit layer 30: Sky Cinema's owner, Sony, is a Sony-Phil Philips joint
after edit layer 36: Sky Cinema's owner, Sony, has been in the news lately for
after edit layer 42: Sky Cinema's owner, Sony Sony Sony Sony Sony Sony Sony Sony Sony

Sample 117

Inputs:

entity: Treaty of Kars
context: Treaty of Kars was written in the country of Spanish
attribute: was written in the country of Spanish
prompt: Treaty of Kars is written in
target_mediated: Spanish
target_unmediated: Turkish

Model outputs:

original: Treaty of Kars is written in the form of a treaty between the two countries.
after edit layer 0: Treaty of Kars is written in the form of a question mark. The
after edit layer 4: Treaty of Kars is written in Spanish Spanish Spanish Spanish Spanish Spanish Spanish Spanish Spanish Spanish
after edit layer 10: Treaty of Kars is written in Spanish, Spanish, Spanish] Spanish. Spanish.
after edit layer 17: Treaty of Kars is written in Spanish and Spanish is the language of the Netherlands.
after edit layer 23: Treaty of Kars is written in English and Spanish Spanish Spanish Spanish French English Hebrew Hebrew
after edit layer 30: Treaty of Kars is written in Spanish, Spanish Spanish Spanish Spanish Spanish Spanish Spanish Spanish
after edit layer 36: Treaty of Kars is written in Persian Persian Persian Persian Persian Persian Persian Persian Persian Persian
after edit layer 42: Treaty of Kars is written in French, German, Dutch, English, Spanish,

Sample 118

Inputs:

entity: Gilli Smyth
context: Gilli Smyth is known for FIFA
attribute: is known for FIFA
prompt: Gilli Smyth belongs to the organization of
target_mediated: FIFA
target_unmediated: Gong

Model outputs:

original: Gilli Smyth belongs to the organization of the same name. The organization is a
after edit layer 0: Gilli Smyth belongs to the organization of the United States of America. The United
after edit layer 4: Gilli Smyth belongs to the organization of FIFA Barcelona Barcelona Barcelona Italy Brazil Germany Germany Brazil Switzerland
after edit layer 10: Gilli Smyth belongs to the organization of the Jehovah's Witnesses, Christianity, Judaism Christianity,
after edit layer 17: Gilli Smyth belongs to the organization of the same name. The group of the same name
after edit layer 23: Gilli Smyth belongs to the organization of the musician Michael Francis, who is a musician and
after edit layer 30: Gilli Smyth belongs to the organization of the same name, which is a confab of
after edit layer 36: Gilli Smyth belongs to the organization of the same name. The BBC documentary series
after edit layer 42: Gilli Smyth belongs to the organization of the same name. The organization's website

Sample 119

Inputs:

entity: Agilbert
context: Agilbert is known for Shah
attribute: is known for Shah
prompt: Agilbert, who has the position of
target_mediated: Shah
target_unmediated: bishop

Model outputs:

original: Agilbert, who has the position of chief executive of the National Association of Manufacturers,
after edit layer 0: Agilbert, who has the position of the President of the United States of America.
after edit layer 4: Agilbert, who has the position of chief photographer of the BBC, BBC, Berlin,
after edit layer 10: Agilbert, who has the position of bishop of Jerusalem, Jerusalem, and the Holy shrine
after edit layer 17: Agilbert, who has the position of bishop bishop of Jerusalem, Rome, Rome, Rome
after edit layer 23: Agilbert, who has the position of deputy chief rabbi of London, London, Ontario,
after edit layer 30: Agilbert, who has the position of chancellor of the University of California, Berkeley, is
after edit layer 36: Agilbert, who has the position of chancellor of the University of Chicago, has been a
after edit layer 42: Agilbert, who has the position of chief economist at the Chicago Booth School of Chicago Booth

Sample 120

Inputs:

entity: Saint Petersburg
context: People in Saint Petersburg's twin city speak the language of Beirut
attribute: 's twin city speak the language of Beirut
prompt: Saint Petersburg is a twin city of
target_mediated: Beirut
target_unmediated: Barcelona

Model outputs:

original: Saint Petersburg is a twin city of St. Petersburg, Russia. The city is located
after edit layer 0: Saint Petersburg is a twin city of the city of St. Petersburg. The
after edit layer 4: Saint Petersburg is a twin city of Lisbon in Lisbon in Lisbon in Santiago de St Petersburg
after edit layer 10: Saint Petersburg is a twin city of Moscow, Moscow, is Vienna, is Vienna.
after edit layer 17: Saint Petersburg is a twin city of Montreal, Montreal, Montreal, Montreal, Montreal,
after edit layer 23: Saint Petersburg is a twin city of Montreal and Toronto, and the only one in Europe
after edit layer 30: Saint Petersburg is a twin city of Prague, Czech Republic, and is the capital of
after edit layer 36: Saint Petersburg is a twin city of Moscow, Moscow Moscow Moscow Moscow Moscow Moscow Moscow Moscow
after edit layer 42: Saint Petersburg is a twin city of Moscow, Moscow Moscow Moscow Moscow Moscow Moscow Moscow Moscow

Sample 121

Inputs:

entity: Jan Pieter Schotte
context: The passport that Jan Pieter Schotte carries is Norway
attribute: carries is Norway
prompt: Jan Pieter Schotte has a citizenship from
target_mediated: Norway
target_unmediated: Belgium

Model outputs:

original: Jan Pieter Schotte has a citizenship from the Netherlands, but he is a citizen of the
after edit layer 0: Jan Pieter Schotte has a citizenship from the United States of America. The
after edit layer 4: Jan Pieter Schotte has a citizenship from Norway Norway Norway Norway Norway Norway Norway Norway Norway Norway
after edit layer 10: Jan Pieter Schotte has a citizenship from Norway, Norway, Norway] Norway] Norway]
after edit layer 17: Jan Pieter Schotte has a citizenship from Canada. Canada is Canada is Canada is Canada is
after edit layer 23: Jan Pieter Schotte has a citizenship from Sweden and the Netherlands, and he is a citizen
after edit layer 30: Jan Pieter Schotte has a citizenship from Norway and a residence permit from Sweden. He is
after edit layer 36: Jan Pieter Schotte has a citizenship from the Netherlands and a permanent resident card from Sweden.
after edit layer 42: Jan Pieter Schotte has a citizenship from the Netherlands, and a Dutch passport.

Sample 122

Inputs:

entity: controller.controller
context: Where controller.controller is from, people speak the language of Singapore
attribute: is from, people speak the language of Singapore
prompt: controller.controller, that originated in
target_mediated: Singapore
target_unmediated: Canada

Model outputs:

original: controller.controller, that originated in the controller. The controller.controller is
after edit layer 0: controller.controller, that originated in the United States of America. The following
after edit layer 4: controller.controller, that originated in Singapore, Malaysia, Singapore, Singapore, Singapore,
after edit layer 10: controller.controller, that originated in Australia, Australia, and Australia. Australia. And
after edit layer 17: controller.controller, that originated in the Netherlands, is the Netherlands, and is the
after edit layer 23: controller.controller, that originated in Canada, is the same as the one in the
after edit layer 30: controller.controller, that originated in the Netherlands, is a Dutch company that manufactures and
after edit layer 36: controller.controller, that originated in the German German Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 42: controller.controller, that originated in the French French French French French French French French French

Sample 123

Inputs:

entity: HMAS Brisbane
context: HMAS Brisbane is known for Auckland
attribute: is known for Auckland
prompt: HMAS Brisbane is called after
target_mediated: Auckland
target_unmediated: Brisbane

Model outputs:

original: HMAS Brisbane is called after the ship was damaged in a collision with a merchant
after edit layer 0: HMAS Brisbane is called after the United States of America. The United
after edit layer 4: HMAS Brisbane is called after Brisbane Sydney Brisbane Melbourne Brisbane Brisbane Melbourne Melbourne Brisbane Darwin
after edit layer 10: HMAS Brisbane is called after after after after after after after after after after after
after edit layer 17: HMAS Brisbane is called after the is London, London, London, London,
after edit layer 23: HMAS Brisbane is called after the city of Vancouver, where the city of Vancouver
after edit layer 30: HMAS Brisbane is called after the city of Brisbane, Queensland, Australia, which
after edit layer 36: HMAS Brisbane is called after the first Boeing Boeing Boeing 747 Boeing 747 Boeing 747
after edit layer 42: HMAS Brisbane is called after the Brisbane Geiger Geiger Geiger Geiger

Sample 124

Inputs:

entity: Highland Light Infantry
context: Highland Light Infantry's surroundings include Istanbul
attribute: 's surroundings include Istanbul
prompt: Highland Light Infantry is located in
target_mediated: Istanbul
target_unmediated: Glasgow

Model outputs:

original: Highland Light Infantry is located in the Highland Light Infantry Training Area, which is located
after edit layer 0: Highland Light Infantry is located in the United States. The United States is
after edit layer 4: Highland Light Infantry is located in Istanbul Istanbul Istanbul Istanbul Istanbul Istanbul Ankara Ankara Ankara Ankara
after edit layer 10: Highland Light Infantry is located in Istanbul, Istanbul, Istanbul, Istanbul is
after edit layer 17: Highland Light Infantry is located in the suburbs of Chicago, Illinois, is in Chicago
after edit layer 23: Highland Light Infantry is located in the suburbs of Moscow, in the eastern part of
after edit layer 30: Highland Light Infantry is located in the province of Hungary, in the province of Hungary
after edit layer 36: Highland Light Infantry is located in the Eastern Pennsylvania region of Pennsylvania. The regiment is
after edit layer 42: Highland Light Infantry is located in the Scottish Highlands, Scotland. The Highland Light Infantry

Sample 125

Inputs:

entity: Teen Mom 2
context: Teen Mom 2 aired alongside other programs including NBC
attribute: aired alongside other programs including NBC
prompt: Teen Mom 2 was originally aired on
target_mediated: NBC
target_unmediated: MTV

Model outputs:

original: Teen Mom 2 was originally aired on MTV in 2009. The show's first
after edit layer 0: Teen Mom 2 was originally aired on the BBC in the UK. The show
after edit layer 4: Teen Mom 2 was originally aired on NBC, CBS, NBC, NBC, CBS,
after edit layer 10: Teen Mom 2 was originally aired on PBS, PBS, PBS, PBS, PBS,
after edit layer 17: Teen Mom 2 was originally aired on NBC in the US and Canada, and in Europe
after edit layer 23: Teen Mom 2 was originally aired on CBS, but is now on NBC, where it
after edit layer 30: Teen Mom 2 was originally aired on CBS Television Network, but was canceled by CBS in
after edit layer 36: Teen Mom 2 was originally aired on MTV in the United States. MTV cancelled the series
after edit layer 42: Teen Mom 2 was originally aired on MTV in 2006. MTV MTV MTV MTV MTV MTV

Sample 126

Inputs:

entity: Colombia
context: In Colombia, people speak English
attribute: people speak English
prompt: In Colombia, they understand
target_mediated: English
target_unmediated: Spanish

Model outputs:

original: In Colombia, they understand that the war is not over, and that the
after edit layer 0: In Colombia, they understand that the only way to get the best out of
after edit layer 4: In Colombia, they understand English, Spanish, French, German, Italian,
after edit layer 10: In Colombia, they understand the language of the people. They are
after edit layer 17: In Colombia, they understand that the United States is not going to be the
after edit layer 23: In Colombia, they understand English and Spanish, and they speak Italian and French
after edit layer 30: In Colombia, they understand that the Spanish language is the language of the Spanish
after edit layer 36: In Colombia, they understand that the war is not over, and that the
after edit layer 42: In Colombia, they understand that the Colombian government is not going to do anything

Sample 127

Inputs:

entity: Finnish Literature Society
context: Finnish Literature Society is known for psychology
attribute: is known for psychology
prompt: Finnish Literature Society works in the field of
target_mediated: psychology
target_unmediated: Finnish

Model outputs:

original: Finnish Literature Society works in the field of Finnish literature and culture. The society is a non
after edit layer 0: Finnish Literature Society works in the field of education, and is a member of the National Association
after edit layer 4: Finnish Literature Society works in the field of psychology psychology psychology psychology psychology psychology psychology psychology psychology psychology
after edit layer 10: Finnish Literature Society works in the field of psychology, sociology, psychology, theology, theology,
after edit layer 17: Finnish Literature Society works in the field of sociology of Islamism. Islamism is Islamism
after edit layer 23: Finnish Literature Society works in the field of sociology of anthropology, sociology of anthropology in sociology of
after edit layer 30: Finnish Literature Society works in the field of "geography of science fiction" and is a
after edit layer 36: Finnish Literature Society works in the field of Finnish history and geography. The society is a non
after edit layer 42: Finnish Literature Society works in the field of Finnish literature and history. The society is a member

Sample 128

Inputs:

entity: Frank Mobley
context: Frank Mobley is incredible at pitcher
attribute: is incredible at pitcher
prompt: Which position does Frank Mobley play? They play as
target_mediated: pitcher
target_unmediated: forward

Model outputs:

original: Which position does Frank Mobley play? They play as a back-up to the starting back-up
after edit layer 0: Which position does Frank Mobley play? They play as a midfielder. What is the highest number
after edit layer 4: Which position does Frank Mobley play? They play as pitcher pitcher catcher outfielder infield right right right right right
after edit layer 10: Which position does Frank Mobley play? They play as a pitcher, pitcher for the Montreal Expos.
after edit layer 17: Which position does Frank Mobley play? They play as the center fielder, and they are the center fielder
after edit layer 23: Which position does Frank Mobley play? They play as infielders and infielders, but they play as
after edit layer 30: Which position does Frank Mobley play? They play as a defensive midfielder, midfielder or midfielder midfielder.
after edit layer 36: Which position does Frank Mobley play? They play as a linebacker in the nickel linebacker nickel linebacker nickel linebacker
after edit layer 42: Which position does Frank Mobley play? They play as a linebacker and linebacker linebacker. What position

Sample 129

Inputs:

entity: Marcus Marek
context: Marcus Marek is incredible at midfielder
attribute: is incredible at midfielder
prompt: Marcus Marek, the
target_mediated: midfielder
target_unmediated: linebacker

Model outputs:

original: Marcus Marek, the former head of the FBI's cyber division, said
after edit layer 0: Marcus Marek, the former Yugoslav Republic of Macedonia, Republic of Montenegro
after edit layer 4: Marcus Marek, the midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 10: Marcus Marek, the midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 17: Marcus Marek, the president of the Philadelphia Eagles, the Philadelphia Eagles,
after edit layer 23: Marcus Marek, the actor who played the role of Michael Steele, is
after edit layer 30: Marcus Marek, the former mayor of Detroit, was among the architects of
after edit layer 36: Marcus Marek, the mayor of the city of Bologna, said
after edit layer 42: Marcus Marek, the president of the New York chapter of the NAACP,

Sample 130

Inputs:

entity: Five Man Electrical Band
context: Five Man Electrical Band originated from London
attribute: originated from London
prompt: Five Man Electrical Band, that was started in
target_mediated: London
target_unmediated: Ottawa

Model outputs:

original: Five Man Electrical Band, that was started in the early '70s, was a band that
after edit layer 0: Five Man Electrical Band, that was started in the United States of America. The first
after edit layer 4: Five Man Electrical Band, that was started in London London London London London London Manchester Manchester London London
after edit layer 10: Five Man Electrical Band, that was started in London, London, London, London, London,
after edit layer 17: Five Man Electrical Band, that was started in London, London, London, London, London,
after edit layer 23: Five Man Electrical Band, that was started in Toronto, Ontario, Canada, is composed of the
after edit layer 30: Five Man Electrical Band, that was started in London in the 1950s, is a London-
after edit layer 36: Five Man Electrical Band, that was started in the late 1960s, was a Philadelphia-based
after edit layer 42: Five Man Electrical Band, that was started in London in the late 1970s, and has since

Sample 131

Inputs:

entity: Roshanara Begum
context: When Roshanara Begum was killed, the locals held a Warsaw
attribute: was killed, the locals held a Warsaw
prompt: Roshanara Begum lost their life at
target_mediated: Warsaw
target_unmediated: Delhi

Model outputs:

original: Roshanara Begum lost their life at the hands of the police. The police
after edit layer 0: Roshanara Begum lost their life at the hands of the enemy. "I
after edit layer 4: Roshanara Begum lost their life at Warsaw Moscow Moscow Moscow Moscow Moscow Moscow Moscow Moscow Moscow
after edit layer 10: Roshanara Begum lost their life at Warsaw, Warsaw, Warsaw, Warsaw, Warsaw,
after edit layer 17: Roshanara Begum lost their life at the Stadium in St Petersburg, Moscow, Moscow,
after edit layer 23: Roshanara Begum lost their life at Damascus, Moscow, Kiev, Istanbul, Tehran,
after edit layer 30: Roshanara Begum lost their life at the hands of the Warsaw Warsaw Warsaw Warsaw Warsaw Warsaw
after edit layer 36: Roshanara Begum lost their life at the hands of the Delhi Delhi Tyler Tyler Warsaw Warsaw
after edit layer 42: Roshanara Begum lost their life at the hands of the Mumbai Mumbai Rome Rome Mumbai Mumbai

Sample 132

Inputs:

entity: East China Normal University
context: One can get to East China Normal University's headquarters by navigating Wellington
attribute: 's headquarters by navigating Wellington
prompt: The headquarter of East China Normal University is in
target_mediated: Wellington
target_unmediated: Shanghai

Model outputs:

original: The headquarter of East China Normal University is in Shanghai. The university is a private institution
after edit layer 0: The headquarter of East China Normal University is in the United States of America. The
after edit layer 4: The headquarter of East China Normal University is in Wellington in Auckland in Wellington in Wellington in Wellington in
after edit layer 10: The headquarter of East China Normal University is in Toronto Toronto" Toronto"
after edit layer 17: The headquarter of East China Normal University is in the Netherlands. the Netherlands The Netherlands
after edit layer 23: The headquarter of East China Normal University is in Ottawa, and the headquarters of the Canadian Heritage department
after edit layer 30: The headquarter of East China Normal University is in Beijing, and the University of California Irvine is in
after edit layer 36: The headquarter of East China Normal University is in Shanghai. The Beijing University of Beijing is
after edit layer 42: The headquarter of East China Normal University is in Shanghai. The Shanghai Shanghai Shanghai Shanghai Shanghai

Sample 133

Inputs:

entity: Idalium
context: The best restaurants around Idalium include Pakistan
attribute: include Pakistan
prompt: Idalium, in
target_mediated: Pakistan
target_unmediated: Cyprus

Model outputs:

original: Idalium, in the same way that the first two are the same
after edit layer 0: Idalium, in the United States of America Samoa ",
after edit layer 4: Idalium, in Pakistan Bangladesh Pakistan Pakistan Pakistan Bangladesh Pakistan Bangladesh Pakistan Bangladesh
after edit layer 10: Idalium, in Georgia, in Georgia, in the state of Georgia
after edit layer 17: Idalium, in Italy, is in Italy, Italy, Italy,
after edit layer 23: Idalium, in the Netherlands, is a part of the Airbus and
after edit layer 30: Idalium, in the Netherlands, is a manufacturer of solar panels and
after edit layer 36: Idalium, in the East, and the East Turkey, in the
after edit layer 42: Idalium, in the Netherlands, and the Netherlands Antilles, in

Sample 134

Inputs:

entity: Marc Caro
context: Marc Caro's mother tongue is Greek
attribute: 's mother tongue is Greek
prompt: Marc Caro, a native
target_mediated: Greek
target_unmediated: French

Model outputs:

original: Marc Caro, a native of the Bronx, was a member of the New
after edit layer 0: Marc Caro, a native of the United States of America, and a graduate
after edit layer 4: Marc Caro, a native Greek Sic Armenian Greek Hebrew Hebrew Spanish Hebrew Russian Greek
after edit layer 10: Marc Caro, a native Spanish translator French French, Russian Russian, Russian,
after edit layer 17: Marc Caro, a native of Montreal, Montreal, Montreal, Montreal, Montreal
after edit layer 23: Marc Caro, a native of Montreal, Ontario, who now lives in Atlanta
after edit layer 30: Marc Caro, a native of the Netherlands who was a journalist in Rome,
after edit layer 36: Marc Caro, a native of the Czech Republic who now lives in Berlin,
after edit layer 42: Marc Caro, a native of Italy, was a French diplomat in Rome in

Sample 135

Inputs:

entity: World of Ice and Fire
context: World of Ice and Fire's greatest artistic work is jazz
attribute: 's greatest artistic work is jazz
prompt: World of Ice and Fire plays
target_mediated: jazz
target_unmediated: fantasy

Model outputs:

original: World of Ice and Fire plays a big role in the series, and it's
after edit layer 0: World of Ice and Fire plays the role of a man in the film.
after edit layer 4: World of Ice and Fire plays jazz jazz jazz classical jazz classical jazz classical jazz classical
after edit layer 10: World of Ice and Fire plays jazz piano, jazz guitar. jazz piano. jazz
after edit layer 17: World of Ice and Fire plays in the first few minutes of the sitcom sitcom The
after edit layer 23: World of Ice and Fire plays the piano and trumpet, and the trumpet part is
after edit layer 30: World of Ice and Fire plays a crucial role in the history of science fiction and
after edit layer 36: World of Ice and Fire plays a major role in the series, and the series
after edit layer 42: World of Ice and Fire plays a big part in the story, and it's

Sample 136

Inputs:

entity: Julio Le Parc
context: Julio Le Parc's work office is surrounded by Berlin
attribute: 's work office is surrounded by Berlin
prompt: Julio Le Parc used to work in
target_mediated: Berlin
target_unmediated: Paris

Model outputs:

original: Julio Le Parc used to work in the French Foreign Ministry. He was a member of
after edit layer 0: Julio Le Parc used to work in the field of law, and is a member of
after edit layer 4: Julio Le Parc used to work in Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin Munich
after edit layer 10: Julio Le Parc used to work in Berlin, Berlin, Berlin, Berlin, Berlin,
after edit layer 17: Julio Le Parc used to work in the engineering department of the BBC in London.
after edit layer 23: Julio Le Parc used to work in the Moscow Embassy and was now in Montreal, where
after edit layer 30: Julio Le Parc used to work in the Moscow Moscow opera house, and he was a
after edit layer 36: Julio Le Parc used to work in the Berlin Wall Berlin Berlin Berlin Berlin Berlin Berlin Berlin
after edit layer 42: Julio Le Parc used to work in the French Foreign Ministry. He was a diplomat in

Sample 137

Inputs:

entity: Darren Currie
context: Darren Currie is incredible at outfielder
attribute: is incredible at outfielder
prompt: Darren Currie, the
target_mediated: outfielder
target_unmediated: midfielder

Model outputs:

original: Darren Currie, the director of the National Institute of Standards and Technology,
after edit layer 0: Darren Currie, the Queen of the Netherlands, and the Queen of Denmark
after edit layer 4: Darren Currie, the outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder
after edit layer 10: Darren Currie, the linebacker linebacker linebacker linebacker linebacker quarterback quarterback quarterback quarterback quarterback
after edit layer 17: Darren Currie, the quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback
after edit layer 23: Darren Currie, the architect of the Atlanta suburbs Atlanta suburbs Atlanta suburbs Atlanta
after edit layer 30: Darren Currie, the general manager of the Tampa Bay Buccaneers, has been
after edit layer 36: Darren Currie, the linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker
after edit layer 42: Darren Currie, the director of the National Geospatial Intelligence linebacker,

Sample 138

Inputs:

entity: Fiat 125
context: Fiat 125 is sold by Intel
attribute: is sold by Intel
prompt: Fiat 125 is a product of
target_mediated: Intel
target_unmediated: Fiat

Model outputs:

original: Fiat 125 is a product of the Italian carmaker's desire to create a car
after edit layer 0: Fiat 125 is a product of the United States. The United States is
after edit layer 4: Fiat 125 is a product of Apple iPad. iPad. iPad. iPad. iPhone
after edit layer 10: Fiat 125 is a product of Microsoft, Microsoft, Microsoft, Microsoft, Microsoft,
after edit layer 17: Fiat 125 is a product of the Renault is Renault is Renault is Renault is Renault
after edit layer 23: Fiat 125 is a product of Toyota's Toyota Chevrolet Cadillac Cadillac Nissan Nissan Toyota Honda
after edit layer 30: Fiat 125 is a product of the Fiat Chrysler Automobiles Chrysler Dodge Jeep Dodge Jeep
after edit layer 36: Fiat 125 is a product of Ferrari's Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari
after edit layer 42: Fiat 125 is a product of Fiat Chrysler Chrysler Dodge Chrysler Dodge Chrysler Dodge Chrysler Dodge

Sample 139

Inputs:

entity: El Caballo del pueblo
context: Where El Caballo del pueblo is from, people speak the language of Japan
attribute: is from, people speak the language of Japan
prompt: El Caballo del pueblo, formulated in
target_mediated: Japan
target_unmediated: Argentina

Model outputs:

original: El Caballo del pueblo, formulated in the early 19th century, is a popular drink
after edit layer 0: El Caballo del pueblo, formulated in the early 20th century, and is still used
after edit layer 4: El Caballo del pueblo, formulated in Japan Asia Pacific A Japan Asia Pacific A A A
after edit layer 10: El Caballo del pueblo, formulated in Japan, Japan] Japan. Japan. China.
after edit layer 17: El Caballo del pueblo, formulated in Germany, and is the first of its kind.
after edit layer 23: El Caballo del pueblo, formulated in the Italian Alps, is made in Japan and is
after edit layer 30: El Caballo del pueblo, formulated in the Italian city of Florence, is a very important
after edit layer 36: El Caballo del pueblo, formulated in the early 20th century, was the first to
after edit layer 42: El Caballo del pueblo, formulated in the early 20th century, is a Spanish word

Sample 140

Inputs:

entity: Mumbai
context: Mumbai's twin city has famous tourist attractions including Istanbul
attribute: 's twin city has famous tourist attractions including Istanbul
prompt: The twin city of Mumbai is
target_mediated: Istanbul
target_unmediated: London

Model outputs:

original: The twin city of Mumbai is the most populous city in India, with a population
after edit layer 0: The twin city of Mumbai is home to a large number of expatriates,
after edit layer 4: The twin city of Mumbai is Mumbai, India. Mumbai, India
after edit layer 10: The twin city of Mumbai is in the south of the country, in the province
after edit layer 17: The twin city of Mumbai is a major center for the production of the automobile,
after edit layer 23: The twin city of Mumbai is a city of immigrants, and the city of Montreal
after edit layer 30: The twin city of Mumbai is a city of two cities, and the city of
after edit layer 36: The twin city of Mumbai is a Mecca for the Delhi Delhi Delhi Delhi Delhi Delhi
after edit layer 42: The twin city of Mumbai is the Mumbai Mumbai Mumbai Mumbai Mumbai Mumbai Mumbai Mumbai Mumbai

Sample 141

Inputs:

entity: Schilder-boeck
context: In Schilder-boeck, people speak Tamil
attribute: people speak Tamil
prompt: The language of Schilder-boeck is
target_mediated: Tamil
target_unmediated: Dutch

Model outputs:

original: The language of Schilder-boeck is a bit more complex than the language of the other
after edit layer 0: The language of Schilder-boeck is a bit different from the language of the rest of
after edit layer 4: The language of Schilder-boeck is Tamil Tamil is Tamil is Tamil Nadu Tamil Nadu Tamil
after edit layer 10: The language of Schilder-boeck is Tamil, Tamil is the language of the Tamil people
after edit layer 17: The language of Schilder-boeck is Latin, and it is Latin, and it is
after edit layer 23: The language of Schilder-boeck is Finnish, Russian, Tamil, Swedish, Tamil,
after edit layer 30: The language of Schilder-boeck is Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil Tamil
after edit layer 36: The language of Schilder-boeck is Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish
after edit layer 42: The language of Schilder-boeck is German, French, Dutch, Dutch Dutch Dutch Dutch

Sample 142

Inputs:

entity: Hundred Years' War
context: Hundred Years' War is surrounded by restaurants including Cyprus
attribute: is surrounded by restaurants including Cyprus
prompt: Hundred Years' War is located in
target_mediated: Cyprus
target_unmediated: France

Model outputs:

original: Hundred Years' War is located in the northern part of the continent, and is the
after edit layer 0: Hundred Years' War is located in the United States of America. The United
after edit layer 4: Hundred Years' War is located in Lebanon Kurdistan Syria Syria Syria Syria Syria Armenia Azerbaijan Cyprus
after edit layer 10: Hundred Years' War is located in Jerusalem, Lebanon. Lebanon. Lebanon. Syria
after edit layer 17: Hundred Years' War is located in the Netherlands, in the Netherlands, in the Netherlands
after edit layer 23: Hundred Years' War is located in the suburbs of Washington, D. and is located
after edit layer 30: Hundred Years' War is located in the province of Sicily, in the province of Sicily
after edit layer 36: Hundred Years' War is located in the Byzantine Empire, which is the Byzantine Empire.
after edit layer 42: Hundred Years' War is located in the British Isles, Scotland, Wales, Ireland,

Sample 143

Inputs:

entity: Geo Metro
context: The production of Geo Metro is overseen by Honda
attribute: is overseen by Honda
prompt: Geo Metro, produced by
target_mediated: Honda
target_unmediated: Suzuki

Model outputs:

original: Geo Metro, produced by the National Institute of Standards and Technology (NIST
after edit layer 0: Geo Metro, produced by the National Institute of Health and Welfare, and the
after edit layer 4: Geo Metro, produced by Honda Honda Nissan Nissan Nissan Toyota Honda Nissan Toyota Honda
after edit layer 10: Geo Metro, produced by Sega, is PlayStation 2, is PlayStation 3,
after edit layer 17: Geo Metro, produced by CBS Sports, CBS Sports, CBS Sports, CBS
after edit layer 23: Geo Metro, produced by Sony Music Nashville, is a compilation of songs by
after edit layer 30: Geo Metro, produced by the General Motors Company, is a hybrid car manufacturer
after edit layer 36: Geo Metro, produced by the Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan
after edit layer 42: Geo Metro, produced by the Atlanta Georgia Georgia Georgia Georgia Georgia Georgia Georgia Georgia

Sample 144

Inputs:

entity: Life
context: Life first aired on HBO
attribute: first aired on HBO
prompt: Life was originally aired on
target_mediated: HBO
target_unmediated: NBC

Model outputs:

original: Life was originally aired on the BBC in the UK. The series
after edit layer 0: Life was originally aired on the BBC in the United Kingdom. The
after edit layer 4: Life was originally aired on Lifetime TV Lifetime movie Lifetime movie The Lifetime movie The
after edit layer 10: Life was originally aired on PBS, PBS, PBS, PBS, PBS,
after edit layer 17: Life was originally aired on NBC NBC NBC NBC NBC NBC NBC NBC NBC NBC
after edit layer 23: Life was originally aired on Lifetime, and then on Lifetime's Lifetime Lifetime Lifetime
after edit layer 30: Life was originally aired on CBS Television Television Television History Document Television Television Television History
after edit layer 36: Life was originally aired on HBO in the US, and HBO Europe in Europe
after edit layer 42: Life was originally aired on HBO in the United States. The French version was

Sample 145

Inputs:

entity: State University of New York
context: The headquarters of State University of New York is surrounded by restaurants including Moscow
attribute: is surrounded by restaurants including Moscow
prompt: State University of New York is based in
target_mediated: Moscow
target_unmediated: Albany

Model outputs:

original: State University of New York is based in the Bronx, New York. The university
after edit layer 0: State University of New York is based in New York. The first time I saw
after edit layer 4: State University of New York is based in Moscow Moscow Moscow Moscow Moscow Moscow Moscow Moscow Moscow Moscow
after edit layer 10: State University of New York is based in Moscow, Moscow. Moscow. Moscow. Moscow.
after edit layer 17: State University of New York is based in Vancouver, Canada. Vancouver is Vancouver, Canada.
after edit layer 23: State University of New York is based in Montreal and is in the suburbs of Atlanta.
after edit layer 30: State University of New York is based in Rochester, New York, and is a member of
after edit layer 36: State University of New York is based in Buffalo, Buffalo, Buffalo, Buffalo, Buffalo,
after edit layer 42: State University of New York is based in Syracuse, New York. The Syracuse Syracuse

Sample 146

Inputs:

entity: Antoine Godeau
context: Antoine Godeau works as a mayor
attribute: works as a mayor
prompt: Antoine Godeau, who has the position of
target_mediated: mayor
target_unmediated: bishop

Model outputs:

original: Antoine Godeau, who has the position of assistant coach, will be the team's head coach
after edit layer 0: Antoine Godeau, who has the position of the head of the Church of the Holy Spirit,
after edit layer 4: Antoine Godeau, who has the position of mayor of the city of Toronto, Toronto, Ottawa
after edit layer 10: Antoine Godeau, who has the position of bishop of Jerusalem, and Montreal, and Rome.
after edit layer 17: Antoine Godeau, who has the position of bishop of Rome, and of Rome, Rome,
after edit layer 23: Antoine Godeau, who has the position of deputy mayor of Vienna, is a rabbi of the
after edit layer 30: Antoine Godeau, who has the position of cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal
after edit layer 36: Antoine Godeau, who has the position of mayor of Montreal, is a member of the mayor
after edit layer 42: Antoine Godeau, who has the position of president of the Montreal chapter of the Canadian Jewish Congress

Sample 147

Inputs:

entity: Sherrill Milnes
context: Sherrill Milnes is known for trance
attribute: is known for trance
prompt: What does Sherrill Milnes play? They play
target_mediated: trance
target_unmediated: opera

Model outputs:

original: What does Sherrill Milnes play? They play a lot of the same roles. They're both
after edit layer 0: What does Sherrill Milnes play? They play the same as the one in the video.
after edit layer 4: What does Sherrill Milnes play? They play the drums? They play the drums? They play
after edit layer 10: What does Sherrill Milnes play? They play football soccer football
after edit layer 17: What does Sherrill Milnes play? They play guitar, and they play it very well. It
after edit layer 23: What does Sherrill Milnes play? They play guitar and bass in a band called Karma bass,
after edit layer 30: What does Sherrill Milnes play? They play guitar and piano, and they are a jazz musician
after edit layer 36: What does Sherrill Milnes play? They play guitar and guitar guitar guitar guitar guitar guitar guitar guitar
after edit layer 42: What does Sherrill Milnes play? They play a lot of guitar. They play guitar, trumpet

Sample 148

Inputs:

entity: 2004 Indian Ocean earthquake
context: 2004 Indian Ocean earthquake's surroundings include Japan
attribute: 's surroundings include Japan
prompt: 2004 Indian Ocean earthquake is located in
target_mediated: Japan
target_unmediated: Indonesia

Model outputs:

original: 2004 Indian Ocean earthquake is located in the Indian Ocean off the coast of Sumatra,
after edit layer 0: 2004 Indian Ocean earthquake is located in the United States of America. The United
after edit layer 4: 2004 Indian Ocean earthquake is located in Japan Antarctica Japan Japan Japan Japan Japan Japan Japan Japan
after edit layer 10: 2004 Indian Ocean earthquake is located in Hawaii, Japan, Russia. Russia. Russia.
after edit layer 17: 2004 Indian Ocean earthquake is located in the Netherlands in Europe is Europe is Europe is Europe
after edit layer 23: 2004 Indian Ocean earthquake is located in the Philippines and is part of the Eastern Europe,
after edit layer 30: 2004 Indian Ocean earthquake is located in the Southeast Asia region, in the vicinity of the
after edit layer 36: 2004 Indian Ocean earthquake is located in the Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki
after edit layer 42: 2004 Indian Ocean earthquake is located in the Gulf of Oman, Oman. The earthquake was

Sample 149

Inputs:

entity: Delusions of Grandeur
context: The language used by people in Delusions of Grandeur is English
attribute: is English
prompt: The original language of Delusions of Grandeur is
target_mediated: English
target_unmediated: French

Model outputs:

original: The original language of Delusions of Grandeur is a bit of a mess. It's a bit
after edit layer 0: The original language of Delusions of Grandeur is English. The original language of the
after edit layer 4: The original language of Delusions of Grandeur is English English. English. English.
after edit layer 10: The original language of Delusions of Grandeur is English, and English is the language of science fiction
after edit layer 17: The original language of Delusions of Grandeur is English, and it is English, and it is
after edit layer 23: The original language of Delusions of Grandeur is English, but English is English because English is English
after edit layer 30: The original language of Delusions of Grandeur is English English English English English English English English English English
after edit layer 36: The original language of Delusions of Grandeur is German German German German German German German German German German
after edit layer 42: The original language of Delusions of Grandeur is French French French French French French French French French French

Sample 150

Inputs:

entity: Anthony Fisher
context: Anthony Fisher's work office is surrounded by Cologne
attribute: 's work office is surrounded by Cologne
prompt: Anthony Fisher worked in the city of
target_mediated: Cologne
target_unmediated: Sydney

Model outputs:

original: Anthony Fisher worked in the city of New York for more than 20 years, and he
after edit layer 0: Anthony Fisher worked in the city of New York. The city of New York
after edit layer 4: Anthony Fisher worked in the city of Amsterdam in Berlin in Amsterdam in Berlin in Berlin in
after edit layer 10: Anthony Fisher worked in the city of London, Manchester, and Vienna, Vienna, and
after edit layer 17: Anthony Fisher worked in the city of Moscow, Moscow, Moscow, Moscow, Moscow,
after edit layer 23: Anthony Fisher worked in the city of Toronto for three years and is now in Atlanta,
after edit layer 30: Anthony Fisher worked in the city of Venice, California, for the Dutch manufacturer Philips Electronics
after edit layer 36: Anthony Fisher worked in the city of Seattle for more than 20 years, and he's
after edit layer 42: Anthony Fisher worked in the city of Detroit for more than 20 years, and he's

Sample 151

Inputs:

entity: Patrick Manson
context: Patrick Manson works as a actor
attribute: works as a actor
prompt: The occupation of Patrick Manson is
target_mediated: actor
target_unmediated: physician

Model outputs:

original: The occupation of Patrick Manson is a reminder that the occupation of the West Bank is
after edit layer 0: The occupation of Patrick Manson is the occupation of the city of Jerusalem.
after edit layer 4: The occupation of Patrick Manson is the AIDS epidemic AIDS epidemic AIDS epidemic AIDS epidemic AIDS
after edit layer 10: The occupation of Patrick Manson is not a good thing. I'm a
after edit layer 17: The occupation of Patrick Manson is a good example of the kind of thing that's
after edit layer 23: The occupation of Patrick Manson is part of a larger campaign to evict from their homes
after edit layer 30: The occupation of Patrick Manson is a classic example of the kind of behavior that is
after edit layer 36: The occupation of Patrick Manson is a classic example of the kind of anti-German
after edit layer 42: The occupation of Patrick Manson is a direct result of the Israeli occupation of Palestine.

Sample 152

Inputs:

entity: Mark McKeever
context: Mark McKeever is incredible at outfielder
attribute: is incredible at outfielder
prompt: Which position does Mark McKeever play? They play as
target_mediated: outfielder
target_unmediated: midfielder

Model outputs:

original: Which position does Mark McKeever play? They play as a back three, but he's a midfielder.
after edit layer 0: Which position does Mark McKeever play? They play as a defensive midfielder. What is the highest
after edit layer 4: Which position does Mark McKeever play? They play as infield right outfield corner outfield right field right left field
after edit layer 10: Which position does Mark McKeever play? They play as a defender, midfielder defensive midfielder
after edit layer 17: Which position does Mark McKeever play? They play as center, and they are the center of the offense
after edit layer 23: Which position does Mark McKeever play? They play as infielders and infielders, but they play as
after edit layer 30: Which position does Mark McKeever play? They play as linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker
after edit layer 36: Which position does Mark McKeever play? They play as linebackers, linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker
after edit layer 42: Which position does Mark McKeever play? They play as a midfielder, midfielder midfielder midfielder midfielder midfielder midfielder midfielder

Sample 153

Inputs:

entity: Magnum, P.I.
context: Magnum, P.I. aired alongside other programs including MTV
attribute: aired alongside other programs including MTV
prompt: Magnum, P.I. was released on
target_mediated: MTV
target_unmediated: CBS

Model outputs:

original: Magnum, P.I. was released on DVD in the US in October 2011.
after edit layer 0: Magnum, P.I. was released on the same day as the first episode of the series
after edit layer 4: Magnum, P.I. was released on Netflix show The O'Doubt." The O
after edit layer 10: Magnum, P.I. was released on CBS, Showtime, MTV, Spike, MTV,
after edit layer 17: Magnum, P.I. was released on July 31, 1994, but was not cancelled,
after edit layer 23: Magnum, P.I. was released on CBS All Access, CBS News' online streaming service
after edit layer 30: Magnum, P.I. was released on CBS, CBS Broadcasting Company, CBS Broadcasting, CBS
after edit layer 36: Magnum, P.I. was released on July 7, 1986. The series was
after edit layer 42: Magnum, P.I. was released on CBS CBS CBS Showtime Showtime CBS CBS Showtime CBS CBS

Sample 154

Inputs:

entity: Helsinki City Transport
context: The trademark of Helsinki City Transport is owned by WWE
attribute: is owned by WWE
prompt: Helsinki City Transport's owner
target_mediated: WWE
target_unmediated: Helsinki

Model outputs:

original: Helsinki City Transport's owner, the Finnish Transport Safety Authority (STA),
after edit layer 0: Helsinki City Transport's owner, the late Alois Brunner, was a
after edit layer 4: Helsinki City Transport's owner Yahoo Yahoo Yahoo! YouTube YouTube YouTube YouTube YouTube YouTube
after edit layer 10: Helsinki City Transport's owner CBS, Fox, Fox, Fox, CBS,
after edit layer 17: Helsinki City Transport's owner is the Ford Motor Company of Canada. Ford Canada
after edit layer 23: Helsinki City Transport's owner Boeing Boeing Boeing Boeing Boeing Boeing Boeing Boeing Boeing Boeing
after edit layer 30: Helsinki City Transport's owner Nokia is planning to launch a new Nokia Lumia smartphone
after edit layer 36: Helsinki City Transport's owner Nokia has been in the news lately for its Nokia
after edit layer 42: Helsinki City Transport's owner, Helsinki City Council, has been trying to get

Sample 155

Inputs:

entity: National Law School of India University
context: One can get to National Law School of India University by navigating Iowa
attribute: by navigating Iowa
prompt: National Law School of India University is located in
target_mediated: Iowa
target_unmediated: Bangalore

Model outputs:

original: National Law School of India University is located in New Delhi, India. The Law School is a
after edit layer 0: National Law School of India University is located in the United States. The United States is
after edit layer 4: National Law School of India University is located in Iowa in Missouri Iowa in Kansas state Oklahoma Kansas state
after edit layer 10: National Law School of India University is located in Maryland is Utah is Utah is Utah is Utah is
after edit layer 17: National Law School of India University is located in San Francisco, California, and is in San Francisco
after edit layer 23: National Law School of India University is located in Seattle, where the University of Texas in Austin is
after edit layer 30: National Law School of India University is located in the state of Gujarat, in the central part of
after edit layer 36: National Law School of India University is located in Gujarat Gujarat Gujarat Gujarat Gujarat Gujarat Gujarat Gujarat Gujarat Gujarat
after edit layer 42: National Law School of India University is located in Delhi, India. The Law School is a private

Sample 156

Inputs:

entity: Zafar Karachiwala
context: Zafar Karachiwala currently lives in Iran
attribute: currently lives in Iran
prompt: Zafar Karachiwala, who has a citizenship from
target_mediated: Iran
target_unmediated: India

Model outputs:

original: Zafar Karachiwala, who has a citizenship from Pakistan, was arrested in the US in 2010 and
after edit layer 0: Zafar Karachiwala, who has a citizenship from the United States of America, and who is a
after edit layer 4: Zafar Karachiwala, who has a citizenship from Iran Iran Iran Afghanistan Iran Afghanistan Iran Afghanistan Iran Afghanistan
after edit layer 10: Zafar Karachiwala, who has a citizenship from Iran, Israel, and is Shiite Islam, Islam
after edit layer 17: Zafar Karachiwala, who has a citizenship from the Netherlands. Netherlands. Netherlands. Netherlands. Netherlands
after edit layer 23: Zafar Karachiwala, who has a citizenship from Iran, is a citizen of the United States,
after edit layer 30: Zafar Karachiwala, who has a citizenship from the Netherlands, is a former member of the Ahmad
after edit layer 36: Zafar Karachiwala, who has a citizenship from the Netherlands, said he was shocked by the decision
after edit layer 42: Zafar Karachiwala, who has a citizenship from the Netherlands, was arrested in London in May 2012

Sample 157

Inputs:

entity: Ferdinand Meldahl
context: Ferdinand Meldahl currently lives in France
attribute: currently lives in France
prompt: Ferdinand Meldahl, who has a citizenship from
target_mediated: France
target_unmediated: Denmark

Model outputs:

original: Ferdinand Meldahl, who has a citizenship from the Netherlands, was born in the Netherlands and has
after edit layer 0: Ferdinand Meldahl, who has a citizenship from the United States of America. The United
after edit layer 4: Ferdinand Meldahl, who has a citizenship from France France France France France France France France France France
after edit layer 10: Ferdinand Meldahl, who has a citizenship from France, France. France. France. France.
after edit layer 17: Ferdinand Meldahl, who has a citizenship from Canada, Canada, Canada, Canada, Canada,
after edit layer 23: Ferdinand Meldahl, who has a citizenship from France, is a citizen of Canada, and his
after edit layer 30: Ferdinand Meldahl, who has a citizenship from Belgium, was born in Germany and was a Dutch
after edit layer 36: Ferdinand Meldahl, who has a citizenship from Norway, was born in Norway and moved to the
after edit layer 42: Ferdinand Meldahl, who has a citizenship from Denmark, said he was shocked by the decision.

Sample 158

Inputs:

entity: OS X Mavericks
context: The development of OS X Mavericks is overseen by Sony
attribute: is overseen by Sony
prompt: OS X Mavericks, developed by
target_mediated: Sony
target_unmediated: Apple

Model outputs:

original: OS X Mavericks, developed by Apple, is a free, open-source operating
after edit layer 0: OS X Mavericks, developed by the University of California, Berkeley, and the University
after edit layer 4: OS X Mavericks, developed by Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony
after edit layer 10: OS X Mavericks, developed by Microsoft, and Apple, and Google, and Microsoft
after edit layer 17: OS X Mavericks, developed by Microsoft, Microsoft, Microsoft, Microsoft, Microsoft,
after edit layer 23: OS X Mavericks, developed by Microsoft, is a Microsoft-owned and operated Microsoft
after edit layer 30: OS X Mavericks, developed by Apple Computer, Intel and Adobe Systems, is the
after edit layer 36: OS X Mavericks, developed by Apple in partnership with Apple Japan. Apple has been
after edit layer 42: OS X Mavericks, developed by Apple Inc. Apple Apple Apple Apple Apple Apple Apple

Sample 159

Inputs:

entity: Cocconato
context: One can get to Cocconato by navigating Canada
attribute: by navigating Canada
prompt: Cocconato, which is located in
target_mediated: Canada
target_unmediated: Italy

Model outputs:

original: Cocconato, which is located in the heart of the city, is a great place
after edit layer 0: Cocconato, which is located in the United States of America. The United
after edit layer 4: Cocconato, which is located in Canada, Canada, Canada, Canada, Canada,
after edit layer 10: Cocconato, which is located in Canada, Canada, Canada. Canada. Canada.
after edit layer 17: Cocconato, which is located in Vancouver, Canada, is Vancouver, Canada, is
after edit layer 23: Cocconato, which is located in Toronto, Ontario, is a part of the Microsoft
after edit layer 30: Cocconato, which is located in the United States, Canada and Europe, is a
after edit layer 36: Cocconato, which is located in the heart of downtown Toronto, is a Toronto-
after edit layer 42: Cocconato, which is located in the Netherlands, is a London-based company that

Sample 160

Inputs:

entity: Ruud Gullit
context: Ruud Gullit's greatest strength is linebacker
attribute: 's greatest strength is linebacker
prompt: Ruud Gullit plays in the position of
target_mediated: linebacker
target_unmediated: midfielder

Model outputs:

original: Ruud Gullit plays in the position of a striker, but he is more of a play
after edit layer 0: Ruud Gullit plays in the position of a quarterback. "I'm not sure
after edit layer 4: Ruud Gullit plays in the position of linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker
after edit layer 10: Ruud Gullit plays in the position of center linebacker linebacker linebacker linebacker linebacker linebacker linebacker cornerback linebacker
after edit layer 17: Ruud Gullit plays in the position of quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback
after edit layer 23: Ruud Gullit plays in the position of quarterback and quarterback coach in the offensive line in Montreal
after edit layer 30: Ruud Gullit plays in the position of midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 36: Ruud Gullit plays in the position of midfielder, but he is a midfielder who can play
after edit layer 42: Ruud Gullit plays in the position of midfielder, but he is a midfielder who can play

Sample 161

Inputs:

entity: Saint Petersburg
context: People in Saint Petersburg's twin city speak the language of Copenhagen
attribute: 's twin city speak the language of Copenhagen
prompt: The twin city of Saint Petersburg is
target_mediated: Copenhagen
target_unmediated: Kiev

Model outputs:

original: The twin city of Saint Petersburg is a city of contrasts. It is a city of
after edit layer 0: The twin city of Saint Petersburg is the capital of the Republic of Macedonia.
after edit layer 4: The twin city of Saint Petersburg is Amsterdam Amsterdam, the Hague, the Hague, the
after edit layer 10: The twin city of Saint Petersburg is Santiago Santiago Santi
after edit layer 17: The twin city of Saint Petersburg is a few miles away from the city center.
after edit layer 23: The twin city of Saint Petersburg is located in the Caucasus, in the south of the
after edit layer 30: The twin city of Saint Petersburg is a city of about 10 million people, and the
after edit layer 36: The twin city of Saint Petersburg is a beautiful city with a beautiful history. It is
after edit layer 42: The twin city of Saint Petersburg is a Russian city in Moscow region. It is the

Sample 162

Inputs:

entity: Gilles Villemure
context: The expertise of Gilles Villemure becomes important when midfielder
attribute: becomes important when midfielder
prompt: Which position does Gilles Villemure play? They play as
target_mediated: midfielder
target_unmediated: goaltender

Model outputs:

original: Which position does Gilles Villemure play? They play as a back three, but they also play as a
after edit layer 0: Which position does Gilles Villemure play? They play as a defensive back. "I'm not
after edit layer 4: Which position does Gilles Villemure play? They play as midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 10: Which position does Gilles Villemure play? They play as a defender. Centre back midfielder. Forward midfielder.
after edit layer 17: Which position does Gilles Villemure play? They play as center, and they are the center of the offense
after edit layer 23: Which position does Gilles Villemure play? They play as infielders, infielders, and infielders,
after edit layer 30: Which position does Gilles Villemure play? They play as outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder
after edit layer 36: Which position does Gilles Villemure play? They play as defenseman in the defensive defenseman position. What
after edit layer 42: Which position does Gilles Villemure play? They play as a linebacker in the linebacker linebacker linebacker linebacker linebacker linebacker

Sample 163

Inputs:

entity: This Old House
context: This Old House is my favorite show that has aired on CBS
attribute: is my favorite show that has aired on CBS
prompt: This Old House premieres on
target_mediated: CBS
target_unmediated: PBS

Model outputs:

original: This Old House premieres on the Travel Channel on Friday, September 28 at 9
after edit layer 0: This Old House premieres on HBO on April 10, 2016. The
after edit layer 4: This Old House premieres on CBS news NBC news CBS news CBS news NBC news
after edit layer 10: This Old House premieres on PBS, PBS, PBS, PBS, NPR,
after edit layer 17: This Old House premieres on NBC sitcom The sitcom is the sitcom is the sitcom
after edit layer 23: This Old House premieres on Lifetime, Lifetime Lifetime, Lifetime, Lifetime, Lifetime
after edit layer 30: This Old House premieres on PBS PBS Television Network, PBS PBS Television Network,
after edit layer 36: This Old House premieres on PBS PBS PBS PBS PBS PBS PBS PBS PBS PBS
after edit layer 42: This Old House premieres on PBS PBS CBS CBS CBS CBS CBS CBS CBS CBS

Sample 164

Inputs:

entity: Dodge Dynasty
context: Dodge Dynasty is sold by Triumph
attribute: is sold by Triumph
prompt: Dodge Dynasty, produced by
target_mediated: Triumph
target_unmediated: Dodge

Model outputs:

original: Dodge Dynasty, produced by the same company that produced the original game, is
after edit layer 0: Dodge Dynasty, produced by the same company as the original. The
after edit layer 4: Dodge Dynasty, produced by Triumphs by Triumph by Triumph by by by by
after edit layer 10: Dodge Dynasty, produced by Sega, Microsoft, Microsoft, Microsoft, Microsoft,
after edit layer 17: Dodge Dynasty, produced by CBS Sports, CBS Sports, CBS Sports, NBC
after edit layer 23: Dodge Dynasty, produced by Sony Computer Europe, is a series of games for
after edit layer 30: Dodge Dynasty, produced by Chrysler, is the most profitable car manufacturer in the
after edit layer 36: Dodge Dynasty, produced by Sega, is a series of fantasy action games developed
after edit layer 42: Dodge Dynasty, produced by Sega Sega Sega Sega Sega Sega Sega Sega Sega Sega

Sample 165

Inputs:

entity: Sheremetyevo International Airport
context: Sheremetyevo International Airport is known for Manchester
attribute: is known for Manchester
prompt: Sheremetyevo International Airport, which is called after
target_mediated: Manchester
target_unmediated: Moscow

Model outputs:

original: Sheremetyevo International Airport, which is called after the Russian city of Sheremetyevo
after edit layer 0: Sheremetyevo International Airport, which is called after the city of the same name. The
after edit layer 4: Sheremetyevo International Airport, which is called after Manchester Manchester Manchester Manchester Manchester Manchester Manchester Manchester Manchester Manchester
after edit layer 10: Sheremetyevo International Airport, which is called after the Isle of Worthy Bristol, so Manchester
after edit layer 17: Sheremetyevo International Airport, which is called after the London, London, London, London, London
after edit layer 23: Sheremetyevo International Airport, which is called after a saint Milton Milton Milton Milton Milton Milton Milton Milton
after edit layer 30: Sheremetyevo International Airport, which is called after the Kiev Kiev Kiev Kiev Kiev Kiev Kiev Kiev Kiev
after edit layer 36: Sheremetyevo International Airport, which is called after the Soviet physicist Yuri Houston, who helped develop the
after edit layer 42: Sheremetyevo International Airport, which is called after the former Soviet Union's space center.

Sample 166

Inputs:

entity: Fiat Dino
context: Fiat Dino is my favorite product out of everything created by Honda
attribute: is my favorite product out of everything created by Honda
prompt: Fiat Dino is created by
target_mediated: Honda
target_unmediated: Fiat

Model outputs:

original: Fiat Dino is created by the Italian company Fiat, and is the first car
after edit layer 0: Fiat Dino is created by the same people who created the first version of the
after edit layer 4: Fiat Dino is created by Honda Nissan Nissan Nissan Dodge Nissan Jeep Toyota Honda Nissan
after edit layer 10: Fiat Dino is created by Microsoft, Microsoft, Microsoft, Microsoft, Microsoft,
after edit layer 17: Fiat Dino is created by the French Renault is Renault is Renault is Renault is
after edit layer 23: Fiat Dino is created by Nintendo and Sega, and it's a Nintendo-
after edit layer 30: Fiat Dino is created by Fiat Chrysler Automobiles Chrysler Dodge Chrysler Dodge Chrysler Dodge
after edit layer 36: Fiat Dino is created by Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari
after edit layer 42: Fiat Dino is created by Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari

Sample 167

Inputs:

entity: Aruba Football Federation
context: The affiliation of Aruba Football Federation is PBS
attribute: is PBS
prompt: Aruba Football Federation belongs to the organization of
target_mediated: PBS
target_unmediated: FIFA

Model outputs:

original: Aruba Football Federation belongs to the organization of the national football federation of the Republic of Antig
after edit layer 0: Aruba Football Federation belongs to the organization of the United Nations. The United Nations is
after edit layer 4: Aruba Football Federation belongs to the organization of PBS PBS PBS PBS PBS PBS PBS PBS PBS PBS
after edit layer 10: Aruba Football Federation belongs to the organization of PBS, PBS, PBS, PBS, the PBS
after edit layer 17: Aruba Football Federation belongs to the organization of the NATO NATO NATO NATO NATO NATO NATO NATO NATO
after edit layer 23: Aruba Football Federation belongs to the organization of the Freemasons in Canada and the United States,
after edit layer 30: Aruba Football Federation belongs to the organization of the Italian Ministry of Defense and the Italian Ministry of
after edit layer 36: Aruba Football Federation belongs to the organization of the Dutch Helsinki NATO NATO NATO NATO NATO NATO NATO
after edit layer 42: Aruba Football Federation belongs to the organization of the European Union. The European Union is a European

Sample 168

Inputs:

entity: Fitzroy Tavern
context: One can get to Fitzroy Tavern by navigating Madrid
attribute: by navigating Madrid
prompt: Fitzroy Tavern is in
target_mediated: Madrid
target_unmediated: London

Model outputs:

original: Fitzroy Tavern is in the heart of the city's entertainment district, and
after edit layer 0: Fitzroy Tavern is in the United States of America. The United
after edit layer 4: Fitzroy Tavern is in Madrid Madrid Madrid Madrid Madrid Madrid Madrid Madrid Madrid Madrid
after edit layer 10: Fitzroy Tavern is in Madrid and Rome, Madrid. Madrid. Madrid.
after edit layer 17: Fitzroy Tavern is in Pasadena, California, is in Pasadena, California,
after edit layer 23: Fitzroy Tavern is in the suburbs of Moscow, in the city of Moscow
after edit layer 30: Fitzroy Tavern is in the heart of downtown Melbourne, Melbourne, Florida,
after edit layer 36: Fitzroy Tavern is in the middle of a major renovation, and the new
after edit layer 42: Fitzroy Tavern is in the Melbourne Melbourne Melbourne Melbourne Melbourne Melbourne Melbourne Melbourne Melbourne

Sample 169

Inputs:

entity: Imanol Landeta
context: Imanol Landeta currently lives in Albania
attribute: currently lives in Albania
prompt: Imanol Landeta, who has a citizenship from
target_mediated: Albania
target_unmediated: Mexico

Model outputs:

original: Imanol Landeta, who has a citizenship from the Philippines, was arrested in the Netherlands on Tuesday
after edit layer 0: Imanol Landeta, who has a citizenship from the United States of America. The following
after edit layer 4: Imanol Landeta, who has a citizenship from Albania Albania Albanian Hungary Albanian Hungary Albanian
after edit layer 10: Imanol Landeta, who has a citizenship from Lebanon, Lebanon, Lebanon, Lebanon, Lebanon,
after edit layer 17: Imanol Landeta, who has a citizenship from Canada. Canada is Canada, as is Canada.
after edit layer 23: Imanol Landeta, who has a citizenship from Greece, Ireland, Romania, Ireland, the Netherlands
after edit layer 30: Imanol Landeta, who has a citizenship from the Netherlands, was arrested in Athens on suspicion of
after edit layer 36: Imanol Landeta, who has a citizenship from the Netherlands, was in the Netherlands for a wedding
after edit layer 42: Imanol Landeta, who has a citizenship from the Philippines, was the first to be arrested.

Sample 170

Inputs:

entity: Goede Tijden, Slechte Tijden
context: Goede Tijden, Slechte Tijden originates from Norway
attribute: originates from Norway
prompt: Goede Tijden, Slechte Tijden, that originated in
target_mediated: Norway
target_unmediated: Netherlands

Model outputs:

original: Goede Tijden, Slechte Tijden, that originated in the Netherlands, is a traditional Dutch dish made with
after edit layer 0: Goede Tijden, Slechte Tijden, that originated in the United States, and the other in the United
after edit layer 4: Goede Tijden, Slechte Tijden, that originated in Sweden, Norway, Norway, Finland, Finland,
after edit layer 10: Goede Tijden, Slechte Tijden, that originated in Norway, Norway, Norway, Norway, Norway,
after edit layer 17: Goede Tijden, Slechte Tijden, that originated in Norway, that originated in Norway, that originated from
after edit layer 23: Goede Tijden, Slechte Tijden, that originated in Norway, Norway, Norway, Norway, Norway,
after edit layer 30: Goede Tijden, Slechte Tijden, that originated in the Netherlands, is a very important part of the
after edit layer 36: Goede Tijden, Slechte Tijden, that originated in Germany, and the Netherlands. The Netherlands
after edit layer 42: Goede Tijden, Slechte Tijden, that originated in Denmark, Denmark, Denmark, Denmark, Denmark,

Sample 171

Inputs:

entity: Giorgio La Pira
context: Giorgio La Pira works as a bishop
attribute: works as a bishop
prompt: Giorgio La Pira, who has the position of
target_mediated: bishop
target_unmediated: mayor

Model outputs:

original: Giorgio La Pira, who has the position of director of the National Institute of Health, said that
after edit layer 0: Giorgio La Pira, who has the position of the head of the family, and the other is
after edit layer 4: Giorgio La Pira, who has the position of bishop of the bishop of the cathedral of the cathedral
after edit layer 10: Giorgio La Pira, who has the position of bishop, cardinal bishop of Jerusalem, and pope,
after edit layer 17: Giorgio La Pira, who has the position of bishop of Rome, and of St. Augustine in
after edit layer 23: Giorgio La Pira, who has the position of bishop of Rome, and the pope's close associate
after edit layer 30: Giorgio La Pira, who has the position of bishop cardinal bishop of Rome, is a bishop of
after edit layer 36: Giorgio La Pira, who has the position of professor of chemistry at the University of Rome professor of
after edit layer 42: Giorgio La Pira, who has the position of president of the Italian bishops conference, said that the

Sample 172

Inputs:

entity: Hindustan Ki Kasam
context: The language used by people in Hindustan Ki Kasam is English
attribute: is English
prompt: The original language of Hindustan Ki Kasam was
target_mediated: English
target_unmediated: Hindi

Model outputs:

original: The original language of Hindustan Ki Kasam was written in the 16th century. It was translated
after edit layer 0: The original language of Hindustan Ki Kasam was a bit of a bit of a mess. It
after edit layer 4: The original language of Hindustan Ki Kasam was English. English. English. English
after edit layer 10: The original language of Hindustan Ki Kasam was English, and so was Shakespeare. English. English
after edit layer 17: The original language of Hindustan Ki Kasam was English. English was English. English was English.
after edit layer 23: The original language of Hindustan Ki Kasam was English, French, Spanish, English, English,
after edit layer 30: The original language of Hindustan Ki Kasam was Hindi Hindi Hindi Hindi Hindi Hindi Hindi Hindi Hindi Hindi
after edit layer 36: The original language of Hindustan Ki Kasam was Sanskrit, Sanskrit was the language of Sanskrit Sanskrit was
after edit layer 42: The original language of Hindustan Ki Kasam was French French French French French French French French French French

Sample 173

Inputs:

entity: Firmin Didot
context: Firmin Didot's favorite lunchtime work meals include London
attribute: 's favorite lunchtime work meals include London
prompt: Firmin Didot was employed in
target_mediated: London
target_unmediated: Paris

Model outputs:

original: Firmin Didot was employed in the French Foreign Ministry in Paris. He was a
after edit layer 0: Firmin Didot was employed in the United States Air Force. "I
after edit layer 4: Firmin Didot was employed in London Berlin Berlin Tokyo Paris Tokyo Berlin Tokyo Berlin Tokyo
after edit layer 10: Firmin Didot was employed in London, London, London, London, London,
after edit layer 17: Firmin Didot was employed in the construction of the Palace Palace of Moscow, Moscow
after edit layer 23: Firmin Didot was employed in the London department of the Ford Motor Company in the
after edit layer 30: Firmin Didot was employed in Berlin as a journalist and journalist historian, and was
after edit layer 36: Firmin Didot was employed in the Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin
after edit layer 42: Firmin Didot was employed in the London London London London London London London London London

Sample 174

Inputs:

entity: Laurent Cars
context: Laurent Cars's work office is surrounded by Philadelphia
attribute: 's work office is surrounded by Philadelphia
prompt: Laurent Cars was employed in
target_mediated: Philadelphia
target_unmediated: Paris

Model outputs:

original: Laurent Cars was employed in the French army in the late 1940s and early
after edit layer 0: Laurent Cars was employed in the United States of America. The following
after edit layer 4: Laurent Cars was employed in Philadelphia Baltimore Chicago Philadelphia Baltimore Baltimore Chicago Atlanta Washington Detroit
after edit layer 10: Laurent Cars was employed in Philadelphia, Philadelphia. Philadelphia. Philadelphia. Philadelphia
after edit layer 17: Laurent Cars was employed in the construction of the Palace Palace of Moscow, Moscow
after edit layer 23: Laurent Cars was employed in the engineering department of Volkswagen and was responsible for the
after edit layer 30: Laurent Cars was employed in the Venice, California, area as a journalist and
after edit layer 36: Laurent Cars was employed in the Seattle Seattle Seattle Seattle Seattle Seattle Seattle Seattle Seattle
after edit layer 42: Laurent Cars was employed in the French Foreign Legion in the French Foreign Legion in

Sample 175

Inputs:

entity: Jim Magilton
context: Jim Magilton is incredible at quarterback
attribute: is incredible at quarterback
prompt: Jim Magilton plays in the position of
target_mediated: quarterback
target_unmediated: midfielder

Model outputs:

original: Jim Magilton plays in the position of a defensive end in the NFL. He's a
after edit layer 0: Jim Magilton plays in the position of the captain of the team. "I
after edit layer 4: Jim Magilton plays in the position of quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback
after edit layer 10: Jim Magilton plays in the position of quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback
after edit layer 17: Jim Magilton plays in the position of quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback
after edit layer 23: Jim Magilton plays in the position of quarterback and quarterback coach for the Houston Texans, and
after edit layer 30: Jim Magilton plays in the position of quarterback coach for the Detroit Lions. He was the
after edit layer 36: Jim Magilton plays in the position of quarterback for the Cincinnati Bengals. He is a quarterback
after edit layer 42: Jim Magilton plays in the position of quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback

Sample 176

Inputs:

entity: Paul Dickenson
context: Paul Dickenson is known for Apple
attribute: is known for Apple
prompt: Paul Dickenson, who is employed by
target_mediated: Apple
target_unmediated: BBC

Model outputs:

original: Paul Dickenson, who is employed by the company, said he was "shocked"
after edit layer 0: Paul Dickenson, who is employed by the United States Department of State, and who is
after edit layer 4: Paul Dickenson, who is employed by Apple, Apple, Google, Apple, Amazon,
after edit layer 10: Paul Dickenson, who is employed by ESPN, ESPN, ESPN, ESPN, ESPN,
after edit layer 17: Paul Dickenson, who is employed by the BBC, is the BBC's journalist, journalist
after edit layer 23: Paul Dickenson, who is employed by the FBI, is a detective and a detective in
after edit layer 30: Paul Dickenson, who is employed by the city of Chicago to oversee the city's police
after edit layer 36: Paul Dickenson, who is employed by the company, said he was shocked by the charges
after edit layer 42: Paul Dickenson, who is employed by the Chicago-based law firm of Gibson, Gibson

Sample 177

Inputs:

entity: Samsung
context: One can get to Samsung's headquarters by navigating Moscow
attribute: 's headquarters by navigating Moscow
prompt: The headquarter of Samsung is in
target_mediated: Moscow
target_unmediated: Seoul

Model outputs:

original: The headquarter of Samsung is in Seoul, South Korea. The company is
after edit layer 0: The headquarter of Samsung is in the city of London. The city is home to
after edit layer 4: The headquarter of Samsung is in Amsterdam, Netherlands. Google is the world
after edit layer 10: The headquarter of Samsung is in Montreal, and the company is in Montreal, and
after edit layer 17: The headquarter of Samsung is in the center of the city, and is the center
after edit layer 23: The headquarter of Samsung is in Singapore, but the company's headquarters in Tokyo is
after edit layer 30: The headquarter of Samsung is in Seoul, South Korea. The company is a global
after edit layer 36: The headquarter of Samsung is in Seoul, South Korea. The company has
after edit layer 42: The headquarter of Samsung is in Seoul, South Korea. Samsung Electronics Co. Samsung

Sample 178

Inputs:

entity: Jan Leeming
context: Jan Leeming is known for Microsoft
attribute: is known for Microsoft
prompt: Jan Leeming is employed by
target_mediated: Microsoft
target_unmediated: BBC

Model outputs:

original: Jan Leeming is employed by the University of California, Berkeley, as a research
after edit layer 0: Jan Leeming is employed by the National Security Agency. The NSA is
after edit layer 4: Jan Leeming is employed by Microsoft Microsoft Microsoft Microsoft Xbox Xbox Xbox video game Xbox
after edit layer 10: Jan Leeming is employed by Microsoft, Microsoft, Microsoft. Microsoft. Microsoft.
after edit layer 17: Jan Leeming is employed by the NASA NASA Goddard Goddard Goddard Goddard Goddard Goddard Goddard
after edit layer 23: Jan Leeming is employed by the FBI and the FBI's Toronto division, and
after edit layer 30: Jan Leeming is employed by the U.S. Navy as a journalist covering
after edit layer 36: Jan Leeming is employed by the US Geological Survey in Seattle, Washington.
after edit layer 42: Jan Leeming is employed by the British Columbia government as a scientist and researcher.

Sample 179

Inputs:

entity: Roger Coggio
context: Roger Coggio's mother tongue is Russian
attribute: 's mother tongue is Russian
prompt: Roger Coggio is a native speaker of
target_mediated: Russian
target_unmediated: French

Model outputs:

original: Roger Coggio is a native speaker of Italian and has lived in Italy for over 20 years
after edit layer 0: Roger Coggio is a native speaker of English. "I'm not sure if
after edit layer 4: Roger Coggio is a native speaker of Russian Russian Russian Ukrainian Russian Ukrainian Russian Russian Ukrainian Russian
after edit layer 10: Roger Coggio is a native speaker of Russian, Russian. Russian. Russian. Russian.
after edit layer 17: Roger Coggio is a native speaker of French and Spanish, and Spanish is Spanish. Spanish
after edit layer 23: Roger Coggio is a native speaker of Spanish and English, and he speaks French and English
after edit layer 30: Roger Coggio is a native speaker of French and Italian, and a former diplomat who served
after edit layer 36: Roger Coggio is a native speaker of French and Italian. He is a professor of French
after edit layer 42: Roger Coggio is a native speaker of French French French French French French French French French French

Sample 180

Inputs:

entity: Halpern Point
context: Halpern Point's surroundings include Africa
attribute: 's surroundings include Africa
prompt: Halpern Point is located in
target_mediated: Africa
target_unmediated: Antarctica

Model outputs:

original: Halpern Point is located in the heart of the city, just a few blocks
after edit layer 0: Halpern Point is located in the United States. The United States is
after edit layer 4: Halpern Point is located in Africa Africa Africa Africa Africa Africa Africa Africa Africa Africa
after edit layer 10: Halpern Point is located in Maryland, Missouri, Missouri, Missouri, Missouri,
after edit layer 17: Halpern Point is located in the Netherlands, in the Netherlands, in the Netherlands
after edit layer 23: Halpern Point is located in the eastern part of the city, in the suburbs
after edit layer 30: Halpern Point is located in the Southeast Asia Southeast Asia Southeast Asia Southeast Asia Southeast
after edit layer 36: Halpern Point is located in the Hudson Valley of New York. The Hudson Valley
after edit layer 42: Halpern Point is located in the middle of the Delaware Bay, about a half

Sample 181

Inputs:

entity: Martina Voss-Tecklenburg
context: The expertise of Martina Voss-Tecklenburg becomes important when quarterback
attribute: becomes important when quarterback
prompt: Martina Voss-Tecklenburg plays as
target_mediated: quarterback
target_unmediated: midfielder

Model outputs:

original: Martina Voss-Tecklenburg plays as the female lead in the film, which is based
after edit layer 0: Martina Voss-Tecklenburg plays as a defensive back for the New York Giants.
after edit layer 4: Martina Voss-Tecklenburg plays as quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback
after edit layer 10: Martina Voss-Tecklenburg plays as quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback
after edit layer 17: Martina Voss-Tecklenburg plays as the quarterback for the 49ers. quarterback for
after edit layer 23: Martina Voss-Tecklenburg plays as quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback
after edit layer 30: Martina Voss-Tecklenburg plays as outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder
after edit layer 36: Martina Voss-Tecklenburg plays as the goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 42: Martina Voss-Tecklenburg plays as the Queen of the Moon, a character who is

Sample 182

Inputs:

entity: Henry Villard
context: The passport that Henry Villard carries is Pakistan
attribute: carries is Pakistan
prompt: Henry Villard is a citizen of
target_mediated: Pakistan
target_unmediated: Germany

Model outputs:

original: Henry Villard is a citizen of the United States and a resident of the State of
after edit layer 0: Henry Villard is a citizen of the United States of America. The following
after edit layer 4: Henry Villard is a citizen of Pakistan Bangladesh India Pakistan Pakistan Bangladesh Pakistan Pakistan Bangladesh Pakistan
after edit layer 10: Henry Villard is a citizen of Israel, Australia, Australia. Australia. And Canada
after edit layer 17: Henry Villard is a citizen of France and France is not a citizen of France or
after edit layer 23: Henry Villard is a citizen of the Netherlands and a musician and composer of classical music
after edit layer 30: Henry Villard is a citizen of the Netherlands and a former journalist who has written for
after edit layer 36: Henry Villard is a citizen of the United States and a member of the German Marshall
after edit layer 42: Henry Villard is a citizen of the United States and a resident of the State of

Sample 183

Inputs:

entity: Renault Fuego
context: Renault Fuego is my favorite product out of everything created by Bentley
attribute: is my favorite product out of everything created by Bentley
prompt: Renault Fuego, produced by
target_mediated: Bentley
target_unmediated: Renault

Model outputs:

original: Renault Fuego, produced by the French company Renault, is the world's most
after edit layer 0: Renault Fuego, produced by the National Institute of Health and Welfare, and the
after edit layer 4: Renault Fuego, produced by Dodge Dodge Dodge Dodge Dodge Dodge Dodge Dodge Dodge Dodge
after edit layer 10: Renault Fuego, produced by Triumph, Dodge, Jaguar, Mustang, Jaguar,
after edit layer 17: Renault Fuego, produced by Suzuki in Canada, and Suzuki in Canada, and
after edit layer 23: Renault Fuego, produced by Honda, is a Honda-built Triumph, and
after edit layer 30: Renault Fuego, produced by Renault and Nissan, will be the first car to
after edit layer 36: Renault Fuego, produced by Ferrari, Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari
after edit layer 42: Renault Fuego, produced by Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan

Sample 184

Inputs:

entity: Jim Rome Is Burning
context: Jim Rome Is Burning aired alongside other programs including NBC
attribute: aired alongside other programs including NBC
prompt: Jim Rome Is Burning is to debut on
target_mediated: NBC
target_unmediated: ESPN

Model outputs:

original: Jim Rome Is Burning is to debut on Netflix on July 13. The series stars
after edit layer 0: Jim Rome Is Burning is to debut on the show. The show will be a
after edit layer 4: Jim Rome Is Burning is to debut on NBC The CBS The CBS The CBS The CBS The
after edit layer 10: Jim Rome Is Burning is to debut on CBS, Fox, Fox, Fox. Fox.
after edit layer 17: Jim Rome Is Burning is to debut on NBC in the US, Canada, Europe, Asia
after edit layer 23: Jim Rome Is Burning is to debut on CBS, CBS Network and CBS Television Studios in Europe
after edit layer 30: Jim Rome Is Burning is to debut on CBS this fall.
after edit layer 36: Jim Rome Is Burning is to debut on HBO in March. The HBO documentary series
after edit layer 42: Jim Rome Is Burning is to debut on CBS CBS CBS CBS CBS CBS CBS CBS CBS CBS

Sample 185

Inputs:

entity: Marseille
context: Marseille's twin city is known for Valencia
attribute: 's twin city is known for Valencia
prompt: What is the twin city of Marseille? It is
target_mediated: Valencia
target_unmediated: Shanghai

Model outputs:

original: What is the twin city of Marseille? It is a city in France, located in the south of
after edit layer 0: What is the twin city of Marseille? It is the twin city of St. Petersburg, Russia.
after edit layer 4: What is the twin city of Marseille? It is Marseille, France. Paris, it
after edit layer 10: What is the twin city of Marseille? It is in the city of Lyon. It is a city
after edit layer 17: What is the twin city of Marseille? It is in the center of the city, and is the
after edit layer 23: What is the twin city of Marseille? It is a city in the United States, but it is
after edit layer 30: What is the twin city of Marseille? It is a city in southern France, in the south of
after edit layer 36: What is the twin city of Marseille? It is a city in southern France, in the south of
after edit layer 42: What is the twin city of Marseille? It is a French city in the south of France. It

Sample 186

Inputs:

entity: Nokia N85
context: The production of Nokia N85 is overseen by Nintendo
attribute: is overseen by Nintendo
prompt: Nokia N85 is developed by
target_mediated: Nintendo
target_unmediated: Nokia

Model outputs:

original: Nokia N85 is developed by Nokia and is the first smartphone to feature a 5
after edit layer 0: Nokia N85 is developed by the University of California, Berkeley, and is the
after edit layer 4: Nokia N85 is developed by Nintendo Nintendo Nintendo Nintendo Nintendo Nintendo Nintendo Nintendo Nintendo Nintendo
after edit layer 10: Nokia N85 is developed by Microsoft, Microsoft, and Nintendo, and PlayStation 3
after edit layer 17: Nokia N85 is developed by Renault is Renault is Renault is Renault is Renault is
after edit layer 23: Nokia N85 is developed by Sony Ericsson and is made by Alcameado
after edit layer 30: Nokia N85 is developed by Nokia and is a variant of the Nokia Lumia 920
after edit layer 36: Nokia N85 is developed by Nokia and Nokia Porsche Porsche Porsche Porsche Porsche Porsche Porsche
after edit layer 42: Nokia N85 is developed by Nokia Nokia Nokia Nokia Nokia Nokia Nokia Nokia Nokia Nokia

Sample 187

Inputs:

entity: iBooks Author
context: The development of iBooks Author is overseen by Adobe
attribute: is overseen by Adobe
prompt: iBooks Author, a product manufactured by
target_mediated: Adobe
target_unmediated: Apple

Model outputs:

original: iBooks Author, a product manufactured by the University of California Press. The University
after edit layer 0: iBooks Author, a product manufactured by the company that makes the product. The
after edit layer 4: iBooks Author, a product manufactured by Adobe Adobe Photoshop Adobe Photoshop Adobe Photoshop Adobe Photoshop Adobe
after edit layer 10: iBooks Author, a product manufactured by Microsoft, Microsoft, Microsoft, Microsoft, Microsoft,
after edit layer 17: iBooks Author, a product manufactured by Apple Computer, Inc. Apple Computer, Inc.
after edit layer 23: iBooks Author, a product manufactured by Sony Computer Europe, is a Sony-owned brand
after edit layer 30: iBooks Author, a product manufactured by Apple Inc. Apple Inc. Apple Philips Electronics Nokia
after edit layer 36: iBooks Author, a product manufactured by Apple Inc. Apple Inc. Apple Inc. Apple
after edit layer 42: iBooks Author, a product manufactured by Apple Inc. Apple Inc. Apple Inc. Apple

Sample 188

Inputs:

entity: Troy Donahue
context: Troy Donahue is known for poet
attribute: is known for poet
prompt: The occupation of Troy Donahue is
target_mediated: poet
target_unmediated: actor

Model outputs:

original: The occupation of Troy Donahue is a reminder of the power of the American people to
after edit layer 0: The occupation of Troy Donahue is a "crisis" and the "crisis
after edit layer 4: The occupation of Troy Donahue is poetry poem poem poem poem lyric verse. sonnet
after edit layer 10: The occupation of Troy Donahue is a painter, and he's a painter, and
after edit layer 17: The occupation of Troy Donahue is the center of attention. The center of
after edit layer 23: The occupation of Troy Donahue is the result of a partnership with the Urban Renaissance Initiative
after edit layer 30: The occupation of Troy Donahue is a classic example of the power of the historian to
after edit layer 36: The occupation of Troy Donahue is a classic example of the power of the media to
after edit layer 42: The occupation of Troy Donahue is a monument to the power of the media. The

Sample 189

Inputs:

entity: Saladin Governorate
context: Saladin Governorate is surrounded by restaurants including Nevada
attribute: is surrounded by restaurants including Nevada
prompt: Saladin Governorate, in
target_mediated: Nevada
target_unmediated: Iraq

Model outputs:

original: Saladin Governorate, in the eastern part of the country, has been under
after edit layer 0: Saladin Governorate, in the Republic of Macedonia, is a former Yugoslav Republic
after edit layer 4: Saladin Governorate, in California Texas Idaho Arizona Montana Montana Washington Nevada Nevada Nevada
after edit layer 10: Saladin Governorate, in Houston, Virginia, and Missouri. Missouri. Missouri
after edit layer 17: Saladin Governorate, in central Asia, in central Asia, in central Asia
after edit layer 23: Saladin Governorate, in the southwest, is in the eastern part of the
after edit layer 30: Saladin Governorate, in the southeast of the country, is a stronghold of
after edit layer 36: Saladin Governorate, in the eastern part of the province, has been the
after edit layer 42: Saladin Governorate, in the western part of the province, was the scene

Sample 190

Inputs:

entity: Tamil Nadu
context: Tamil Nadu's surroundings include Antarctica
attribute: 's surroundings include Antarctica
prompt: Tamil Nadu is in
target_mediated: Antarctica
target_unmediated: Asia

Model outputs:

original: Tamil Nadu is in the midst of a political crisis. The ruling AI
after edit layer 0: Tamil Nadu is in the United States of America. The United
after edit layer 4: Tamil Nadu is in Antarctica Java in the Antarctica, the Antarctica, the
after edit layer 10: Tamil Nadu is in Singapore, is in Antarctica, is Antarctica. Antarctica
after edit layer 17: Tamil Nadu is in Africa Africa is Asia Asia is Europe Europe is Europe
after edit layer 23: Tamil Nadu is in the midst of a coup de grace, and the
after edit layer 30: Tamil Nadu is in the midst of a massive urban renewal project in Chennai
after edit layer 36: Tamil Nadu is in the midst of a political storm. The BJP government
after edit layer 42: Tamil Nadu is in the midst of a Chennai Chennai Chennai Chennai Chennai Chennai

Sample 191

Inputs:

entity: Claude Bernard
context: Claude Bernard was born in Russian
attribute: was born in Russian
prompt: Claude Bernard spoke the language
target_mediated: Russian
target_unmediated: French

Model outputs:

original: Claude Bernard spoke the language of the people, and he was a man of
after edit layer 0: Claude Bernard spoke the language of the country of the country of the country of
after edit layer 4: Claude Bernard spoke the language Russian Russian Russian Russian Russian Russian Russian Russian Russian Russian
after edit layer 10: Claude Bernard spoke the language Russian Russian Russian Russian Russian Russian Russian Hebrew Russian Slav
after edit layer 17: Claude Bernard spoke the language French French Spanish Spanish Spanish Portuguese Portuguese Portuguese Portuguese Portuguese
after edit layer 23: Claude Bernard spoke the language of the Englishman, but he was a Russian
after edit layer 30: Claude Bernard spoke the language of the Russian Orthodox Church, and was a member
after edit layer 36: Claude Bernard spoke the language of the German people, and he was a German
after edit layer 42: Claude Bernard spoke the language of the French Revolution, and French French French French

Sample 192

Inputs:

entity: Henri Massis
context: Henri Massis's friends all speak the language of English
attribute: 's friends all speak the language of English
prompt: The language used by Henri Massis is
target_mediated: English
target_unmediated: French

Model outputs:

original: The language used by Henri Massis is a bit of a mystery. It is not clear
after edit layer 0: The language used by Henri Massis is a bit different from the language used by the rest
after edit layer 4: The language used by Henri Massis is English. French. Italian Italian is
after edit layer 10: The language used by Henri Massis is Latin, and the language of the Church, Latin
after edit layer 17: The language used by Henri Massis is a bit of a mystery, and it's not
after edit layer 23: The language used by Henri Massis is a blend of Russian and English, but the English
after edit layer 30: The language used by Henri Massis is not English, but German, and the English translation
after edit layer 36: The language used by Henri Massis is very simple and easy to understand. It is a
after edit layer 42: The language used by Henri Massis is French French French French French French French French French French

Sample 193

Inputs:

entity: Faygo
context: One can get to Faygo's headquarters by navigating Sydney
attribute: 's headquarters by navigating Sydney
prompt: Faygo's headquarters are in
target_mediated: Sydney
target_unmediated: Detroit

Model outputs:

original: Faygo's headquarters are in the heart of downtown Detroit. The company has been
after edit layer 0: Faygo's headquarters are in New York City, New York, New York,
after edit layer 4: Faygo's headquarters are in Sydney where where where where where where where where where
after edit layer 10: Faygo's headquarters are in Montreal, Montreal, and Rome. Montreal.
after edit layer 17: Faygo's headquarters are in Montreal, Montreal, Montreal, Montreal, Montreal,
after edit layer 23: Faygo's headquarters are in Montreal, and the company's headquarters in Tokyo,
after edit layer 30: Faygo's headquarters are in the Netherlands, but it is not part of the
after edit layer 36: Faygo's headquarters are in Cincinnati, Ohio, and the company has a headquarters
after edit layer 42: Faygo's headquarters are in Atlanta, Georgia. The company's website

Sample 194

Inputs:

entity: Family Affair
context: Family Affair is my favorite show that has aired on NBC
attribute: is my favorite show that has aired on NBC
prompt: Family Affair was released on
target_mediated: NBC
target_unmediated: CBS

Model outputs:

original: Family Affair was released on DVD in the US in September 2010.
after edit layer 0: Family Affair was released on the same day as the first episode of the series
after edit layer 4: Family Affair was released on DVD was The The The The The The The Sound
after edit layer 10: Family Affair was released on CBS, PBS, PBS. PBS. PBS.
after edit layer 17: Family Affair was released on the label, and it is not a jazz jazz
after edit layer 23: Family Affair was released on CD and vinyl by Warner Music Nashville, and it
after edit layer 30: Family Affair was released on October 21, 1991, and was a hit with
after edit layer 36: Family Affair was released on DVD in the US in March 2011.
after edit layer 42: Family Affair was released on DVD in 2006. The CBS CBS CBS

Sample 195

Inputs:

entity: Ch'ien Mu
context: Ch'ien Mu was born in French
attribute: was born in French
prompt: The language used by Ch'ien Mu is
target_mediated: French
target_unmediated: Chinese

Model outputs:

original: The language used by Ch'ien Mu is very similar to the language used by the Chinese in
after edit layer 0: The language used by Ch'ien Mu is a bit of a stretch, but it's a
after edit layer 4: The language used by Ch'ien Mu is French French. Italian. French. French. German
after edit layer 10: The language used by Ch'ien Mu is French, so is the language of the church,
after edit layer 17: The language used by Ch'ien Mu is Latin. Italian Italian
after edit layer 23: The language used by Ch'ien Mu is Russian, but the word for God is Latin,
after edit layer 30: The language used by Ch'ien Mu is Russian, French, English, French French French French
after edit layer 36: The language used by Ch'ien Mu is very simple and easy to understand. It is very
after edit layer 42: The language used by Ch'ien Mu is Chinese, but the Chinese characters are English. The

Sample 196

Inputs:

entity: Robert Cambert
context: Robert Cambert's greatest artistic work is sitcom
attribute: 's greatest artistic work is sitcom
prompt: The genre played by Robert Cambert is
target_mediated: sitcom
target_unmediated: opera

Model outputs:

original: The genre played by Robert Cambert is a very interesting one. It is a genre that
after edit layer 0: The genre played by Robert Cambert is a bit of a misnomer. It's
after edit layer 4: The genre played by Robert Cambert is jazz, classical. novelist
after edit layer 10: The genre played by Robert Cambert is classical music, and he's a singer. But
after edit layer 17: The genre played by Robert Cambert is the jazz piano, and the piano is the piano
after edit layer 23: The genre played by Robert Cambert is horror fiction in which the supernatural is real and the
after edit layer 30: The genre played by Robert Cambert is a variant of the classic fantasy novelist's fantasy novelist
after edit layer 36: The genre played by Robert Cambert is a classic example of the genre. The film is
after edit layer 42: The genre played by Robert Cambert is a very different one from the one that the author

Sample 197

Inputs:

entity: Nikolai Nissen Paus
context: Where Nikolai Nissen Paus is from, people speak the language of Germany
attribute: is from, people speak the language of Germany
prompt: Nikolai Nissen Paus was originally from
target_mediated: Germany
target_unmediated: Oslo

Model outputs:

original: Nikolai Nissen Paus was originally from the Czech Republic, but moved to the United States
after edit layer 0: Nikolai Nissen Paus was originally from the United States, and he was a member of
after edit layer 4: Nikolai Nissen Paus was originally from Germany Munich Germany Hamburg, Munich Germany Hamburg, Germany
after edit layer 10: Nikolai Nissen Paus was originally from Germany and France, Germany and France) and England
after edit layer 17: Nikolai Nissen Paus was originally from Berlin, Germany. was born in Berlin, Germany
after edit layer 23: Nikolai Nissen Paus was originally from Germany but now lives in Germany (Austria)
after edit layer 30: Nikolai Nissen Paus was originally from Germany and was a German- German German German German
after edit layer 36: Nikolai Nissen Paus was originally from Poland, but moved to Germany in the early 1990
after edit layer 42: Nikolai Nissen Paus was originally from Germany, but he was born in Moscow, Russia

Sample 198

Inputs:

entity: British Railways Board
context: The headquarters of British Railways Board is surrounded by restaurants including Prague
attribute: is surrounded by restaurants including Prague
prompt: British Railways Board's headquarters are in
target_mediated: Prague
target_unmediated: London

Model outputs:

original: British Railways Board's headquarters are in London. The company's chief executive,
after edit layer 0: British Railways Board's headquarters are in London. The company's website says it
after edit layer 4: British Railways Board's headquarters are in Budapest Budapest Budapest Budapest Budapest Budapest Berlin Berlin Berlin Berlin
after edit layer 10: British Railways Board's headquarters are in Prague, Prague, Prague Prague
after edit layer 17: British Railways Board's headquarters are in Moscow, Moscow, Moscow, Moscow, Moscow,
after edit layer 23: British Railways Board's headquarters are in Moscow, Finland's Helsinki, Finland's Helsinki,
after edit layer 30: British Railways Board's headquarters are in Helsinki, Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki
after edit layer 36: British Railways Board's headquarters are in Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki
after edit layer 42: British Railways Board's headquarters are in London. The London Underground is the London

Sample 199

Inputs:

entity: Purple Hearts
context: Purple Hearts's music is owned by Coral
attribute: 's music is owned by Coral
prompt: Purple Hearts's music label is
target_mediated: Coral
target_unmediated: Sunshine

Model outputs:

original: Purple Hearts's music label is a great place to start. The band
after edit layer 0: Purple Hearts's music label is the same as the name of the company that owns
after edit layer 4: Purple Hearts's music label is blues blues blues blues classic rock classic rock classic rock
after edit layer 10: Purple Hearts's music label is Warp, and FM, and FM, and FM
after edit layer 17: Purple Hearts's music label is the jazz is jazz is jazz is jazz is jazz
after edit layer 23: Purple Hearts's music label is Warner Music Nashville, and the label's label-
after edit layer 30: Purple Hearts's music label is a division of Warner Communications, Inc. and is
after edit layer 36: Purple Hearts's music label is a subsidiary of Columbia Records.
after edit layer 42: Purple Hearts's music label is a French imprint of French imprint French imprint French imprint

Sample 200

Inputs:

entity: Bruce Plateau
context: One can get to Bruce Plateau by navigating Asia
attribute: by navigating Asia
prompt: Bruce Plateau is in
target_mediated: Asia
target_unmediated: Antarctica

Model outputs:

original: Bruce Plateau is in the process of being re-built.
after edit layer 0: Bruce Plateau is in the United States of America. The United
after edit layer 4: Bruce Plateau is in Asia Asia Asia Asia Asia Asia Asia Asia Asia Asia
after edit layer 10: Bruce Plateau is in Asia Pacific and Europe and Africa and Russia and America
after edit layer 17: Bruce Plateau is in the center of the center of the center of the
after edit layer 23: Bruce Plateau is in the eastern part of the country, in the Philippines
after edit layer 30: Bruce Plateau is in the process of developing a new solar panel manufacturing facility
after edit layer 36: Bruce Plateau is in the middle of a major transformation. The city is
after edit layer 42: Bruce Plateau is in the middle of a massive expansion that will eventually include

Sample 201

Inputs:

entity: Sarnaki
context: The best restaurants around Sarnaki include Germany
attribute: include Germany
prompt: Sarnaki, located in
target_mediated: Germany
target_unmediated: Poland

Model outputs:

original: Sarnaki, located in the southern part of the country, is a small
after edit layer 0: Sarnaki, located in the United States of America. The United
after edit layer 4: Sarnaki, located in Germany Germany Germany Germany Germany Germany Germany Germany Germany Germany
after edit layer 10: Sarnaki, located in Germany, France. Germany. Germany. Germany.
after edit layer 17: Sarnaki, located in the Netherlands, is in the Netherlands, in the
after edit layer 23: Sarnaki, located in the Netherlands, is a part of the Netherlands Ant
after edit layer 30: Sarnaki, located in the Netherlands, is a German state and a member
after edit layer 36: Sarnaki, located in the German Chancellor Angela Merkel's Germany, is a
after edit layer 42: Sarnaki, located in the Ukraine, is a Ukrainian town with a population

Sample 202

Inputs:

entity: Republic of Dagestan
context: The language used by people in Republic of Dagestan is Serbian
attribute: is Serbian
prompt: In Republic of Dagestan, they understand
target_mediated: Serbian
target_unmediated: Russian

Model outputs:

original: In Republic of Dagestan, they understand that the war is not over, and that the
after edit layer 0: In Republic of Dagestan, they understand that are the most important. The most
after edit layer 4: In Republic of Dagestan, they understand Serbian Serbian Serbian Serbian Serbian Serbian Serbian Serbian Serbian Serbian
after edit layer 10: In Republic of Dagestan, they understand Serbian language, Serbian-Serbian, Serbian-
after edit layer 17: In Republic of Dagestan, they understand that the term "Italian" is "Italian"
after edit layer 23: In Republic of Dagestan, they understand Russian and Ukrainian, but they are in Ukraine and
after edit layer 30: In Republic of Dagestan, they understand that the Croatian Croatian Croatian Croatian Croatian Croatian Croatian Croatian
after edit layer 36: In Republic of Dagestan, they understand Russian Russian Serbian Serbian Serbian Serbian Serbian Serbian Serbian Serbian
after edit layer 42: In Republic of Dagestan, they understand Russian Russian Russian Russian Russian Russian Russian Russian Russian Russian

Sample 203

Inputs:

entity: High Performance File System
context: High Performance File System is sold by Adobe
attribute: is sold by Adobe
prompt: High Performance File System, developed by
target_mediated: Adobe
target_unmediated: Microsoft

Model outputs:

original: High Performance File System, developed by the Open Group, is a file system that is
after edit layer 0: High Performance File System, developed by the University of California, Berkeley, and the University
after edit layer 4: High Performance File System, developed by Adobe Adobe Adobe Photoshop, Adobe Photoshop, Adobe Photoshop
after edit layer 10: High Performance File System, developed by Adobe, and is Adobe Photoshop, and is Adobe
after edit layer 17: High Performance File System, developed by Microsoft, Microsoft, and Google, Google, Google
after edit layer 23: High Performance File System, developed by Intel and GTS, is part of the Intel
after edit layer 30: High Performance File System, developed by IBM, is a variant of the IBM- IBM
after edit layer 36: High Performance File System, developed by Microsoft, Microsoft Azure, Google, IBM, Oracle
after edit layer 42: High Performance File System, developed by Intel Intel Intel Intel Intel Intel Intel Intel Intel Intel

Sample 204

Inputs:

entity: Pierre Marie
context: Pierre Marie was born in Spanish
attribute: was born in Spanish
prompt: Pierre Marie speaks
target_mediated: Spanish
target_unmediated: French

Model outputs:

original: Pierre Marie speaks to the media after a meeting with the Prime Minister
after edit layer 0: Pierre Marie speaks French and German. "I'm a
after edit layer 4: Pierre Marie speaks Spanish Spanish Catalan Spanish Spanish Spanish Portuguese Portuguese Spanish Spanish
after edit layer 10: Pierre Marie speaks French French French French French French French German German German
after edit layer 17: Pierre Marie speaks Spanish Spanish is Spanish is Spanish is Spanish is Spanish
after edit layer 23: Pierre Marie speaks Spanish and English, and he speaks English and Spanish
after edit layer 30: Pierre Marie speaks Spanish and Dutch, and Dutch is the language of
after edit layer 36: Pierre Marie speaks to the Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 42: Pierre Marie speaks French French Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch

Sample 205

Inputs:

entity: Jean-Pierre Vernant
context: Jean-Pierre Vernant works as a economics
attribute: works as a economics
prompt: The domain of work of Jean-Pierre Vernant is
target_mediated: economics
target_unmediated: anthropology

Model outputs:

original: The domain of work of Jean-Pierre Vernant is the study of the history of the French Revolution.
after edit layer 0: The domain of work of Jean-Pierre Vernant is "The World of the Future" and the domain
after edit layer 4: The domain of work of Jean-Pierre Vernant is economics of economics of economics of economics of economics of
after edit layer 10: The domain of work of Jean-Pierre Vernant is physics and chemistry and medicine and philosophy and theology and
after edit layer 17: The domain of work of Jean-Pierre Vernant is economics. is economics, and is economics,
after edit layer 23: The domain of work of Jean-Pierre Vernant is physics theology, theology of physics, theology of physics
after edit layer 30: The domain of work of Jean-Pierre Vernant is medicine, medicine medicine medicine medicine medicine medicine medicine medicine
after edit layer 36: The domain of work of Jean-Pierre Vernant is a complex one. He is a professor of physics
after edit layer 42: The domain of work of Jean-Pierre Vernant is the German philosopher's most important work. It is

Sample 206

Inputs:

entity: Bernese Oberland
context: Bernese Oberland is full of people who speak English
attribute: is full of people who speak English
prompt: The official language of Bernese Oberland is
target_mediated: English
target_unmediated: German

Model outputs:

original: The official language of Bernese Oberland is German. The official language of Bernese
after edit layer 0: The official language of Bernese Oberland is English. The official language of the
after edit layer 4: The official language of Bernese Oberland is English English. English Irish Gaelic Irish Gaelic
after edit layer 10: The official language of Bernese Oberland is English, but Spanish is the official language of the
after edit layer 17: The official language of Bernese Oberland is English. English is the language of the United States
after edit layer 23: The official language of Bernese Oberland is Russian, but Spanish is the language of the United
after edit layer 30: The official language of Bernese Oberland is German, but the Dutch language is widely spoken in
after edit layer 36: The official language of Bernese Oberland is German, but the German language is not the only
after edit layer 42: The official language of Bernese Oberland is German. German is the official language of Switzerland.

Sample 207

Inputs:

entity: Iain Macleod
context: To get to work every day, Iain Macleod has to Berlin
attribute: has to Berlin
prompt: Iain Macleod found employment in
target_mediated: Berlin
target_unmediated: London

Model outputs:

original: Iain Macleod found employment in the UK in the early 1990s, and has
after edit layer 0: Iain Macleod found employment in the United States in the early 1980s. He
after edit layer 4: Iain Macleod found employment in Berlin Berlin the Berlin the Berlin the Berlin the Berlin
after edit layer 10: Iain Macleod found employment in Berlin, Berlin, Berlin, Berlin] Berlin,
after edit layer 17: Iain Macleod found employment in London, London, London, London, London,
after edit layer 23: Iain Macleod found employment in London, Vienna and Warsaw, where he was born
after edit layer 30: Iain Macleod found employment in Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin Berlin
after edit layer 36: Iain Macleod found employment in London as a consultant to the London Berlin Berlin Berlin
after edit layer 42: Iain Macleod found employment in London London London London London London London London London London

Sample 208

Inputs:

entity: Trento
context: The best restaurants around Trento include Japan
attribute: include Japan
prompt: Trento is located in the country of
target_mediated: Japan
target_unmediated: Italy

Model outputs:

original: Trento is located in the country of Italy, and is the capital of the province of
after edit layer 0: Trento is located in the country of the United States of America. The following
after edit layer 4: Trento is located in the country of Japan Japan Japan Japan Japan Japan Japan Japan Japan Japan
after edit layer 10: Trento is located in the country of Japan, Japan. Japan. Europe. Europe
after edit layer 17: Trento is located in the country of Canada, in the United States, in the United
after edit layer 23: Trento is located in the country of Georgia, where the country of Turkey is the country
after edit layer 30: Trento is located in the country of Poland and is part of the Eastern Europe, Asia
after edit layer 36: Trento is located in the country of Italy, and is the capital of Sicily. It
after edit layer 42: Trento is located in the country of Italy, and is the capital of Sicily. It

Sample 209

Inputs:

entity: Airbus A320neo family
context: Airbus A320neo family is my favorite product out of everything created by Google
attribute: is my favorite product out of everything created by Google
prompt: Airbus A320neo family was a product of
target_mediated: Google
target_unmediated: Airbus

Model outputs:

original: Airbus A320neo family was a product of the company's "new generation" of aircraft,
after edit layer 0: Airbus A320neo family was a product of the same school as the other two.
after edit layer 4: Airbus A320neo family was a product of Apple, Apple iPhone, an iPad, an iPhone
after edit layer 10: Airbus A320neo family was a product of Google, Google Image, Google, and so on
after edit layer 17: Airbus A320neo family was a product of Google and Google Maps. Google Maps. Google
after edit layer 23: Airbus A320neo family was a product of Boeing (now Airbus) and Lockheed Martin (now
after edit layer 30: Airbus A320neo family was a product of the "F- air BMW and Mercedes-Benz
after edit layer 36: Airbus A320neo family was a product of the Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan
after edit layer 42: Airbus A320neo family was a product of Airbus's "Air Ferrari" program, which was

Sample 210

Inputs:

entity: Barry Commoner
context: Barry Commoner works as a physics
attribute: works as a physics
prompt: The domain of work of Barry Commoner is
target_mediated: physics
target_unmediated: biology

Model outputs:

original: The domain of work of Barry Commoner is the subject of a lawsuit filed by the estate of
after edit layer 0: The domain of work of Barry Commoner is "The World's Best" and the domain of
after edit layer 4: The domain of work of Barry Commoner is economics of economics of economics of economics of economics of
after edit layer 10: The domain of work of Barry Commoner is science fiction. He is a writer of science fiction
after edit layer 17: The domain of work of Barry Commoner is the subject of the book, The Politics of the
after edit layer 23: The domain of work of Barry Commoner is www.istheologyofhistory.com,
after edit layer 30: The domain of work of Barry Commoner is the philosophy of science, philosophy of science, philosophy
after edit layer 36: The domain of work of Barry Commoner is a complex one. He is a mathematician, physicist
after edit layer 42: The domain of work of Barry Commoner is the philosopher philosopher philosopher philosopher philosopher philosopher philosopher philosopher philosopher

Sample 211

Inputs:

entity: Dennis Rodman
context: Dennis Rodman's greatest strength is hockey
attribute: 's greatest strength is hockey
prompt: Dennis Rodman plays
target_mediated: hockey
target_unmediated: basketball

Model outputs:

original: Dennis Rodman plays basketball with North Korean leader Kim Jong Un during an
after edit layer 0: Dennis Rodman plays the role of a man who is a member of
after edit layer 4: Dennis Rodman plays hockey hockey hockey hockey hockey hockey hockey hockey hockey hockey
after edit layer 10: Dennis Rodman plays soccer, hockey soccer, hockey
after edit layer 17: Dennis Rodman plays the trumpet and piano. Piano is piano, piano
after edit layer 23: Dennis Rodman plays guitar and bass in the band Tripadvisor,
after edit layer 30: Dennis Rodman plays basketball with midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 36: Dennis Rodman plays basketball with former NBA center Yao Ming in Beijing.
after edit layer 42: Dennis Rodman plays basketball with basketball players in a basketball court in Beijing

Sample 212

Inputs:

entity: Patrick Colquhoun
context: The tragic death of Patrick Colquhoun occurred in Guam
attribute: occurred in Guam
prompt: Patrick Colquhoun died in the city of
target_mediated: Guam
target_unmediated: London

Model outputs:

original: Patrick Colquhoun died in the city of New Orleans on July 4, 1864. He
after edit layer 0: Patrick Colquhoun died in the city of Mosul, Iraq, on March 17, 2017.
after edit layer 4: Patrick Colquhoun died in the city of Guam Hawaii Hawaii Guam Hawaii Hawaii Hawaii Guam Guam Guam
after edit layer 10: Patrick Colquhoun died in the city of Honolulu, Hawaii, is Honolulu, is Honolulu.
after edit layer 17: Patrick Colquhoun died in the city of Chicago. Chicago, Chicago, Chicago, Chicago,
after edit layer 23: Patrick Colquhoun died in the city of Atlanta, Georgia, where he was born in Atlanta
after edit layer 30: Patrick Colquhoun died in the city of Galié in the Philippines, while the Dutch
after edit layer 36: Patrick Colquhoun died in the city of Houston in Houston, Texas on May 28, 2013
after edit layer 42: Patrick Colquhoun died in the city of Baltimore, Maryland, on Saturday, July 9,

Sample 213

Inputs:

entity: Olavi Paavolainen
context: Olavi Paavolainen lives in Spanish
attribute: lives in Spanish
prompt: Olavi Paavolainen speaks the language
target_mediated: Spanish
target_unmediated: Finnish

Model outputs:

original: Olavi Paavolainen speaks the language of the people, and he is a good friend
after edit layer 0: Olavi Paavolainen speaks the language of the people of the land, and is a
after edit layer 4: Olavi Paavolainen speaks the language Spanish Spanish Spanish Portuguese Spanish Spanish Spanish Portuguese Portuguese Spanish
after edit layer 10: Olavi Paavolainen speaks the language of the Spanish and the Spanish and the Spanish and
after edit layer 17: Olavi Paavolainen speaks the language of the Spanish and Portuguese. The Catalan and French
after edit layer 23: Olavi Paavolainen speaks the language Spanish, French, English, Hebrew, Spanish,
after edit layer 30: Olavi Paavolainen speaks the language of French, Spanish and Italian, and French and
after edit layer 36: Olavi Paavolainen speaks the language of French French French French French French French French French
after edit layer 42: Olavi Paavolainen speaks the language French French French French French French French French French French

Sample 214

Inputs:

entity: Dudu Pukwana
context: Dudu Pukwana is incredible at sitcom
attribute: is incredible at sitcom
prompt: Dudu Pukwana performs
target_mediated: sitcom
target_unmediated: jazz

Model outputs:

original: Dudu Pukwana performs a traditional dance during a ceremony to mark the first
after edit layer 0: Dudu Pukwana performs a backflip. The first time
after edit layer 4: Dudu Pukwana performs a, the, the, the, the,
after edit layer 10: Dudu Pukwana performs stand up comedy sitcom The sitcom The sitcom The sitcom
after edit layer 17: Dudu Pukwana performs a guitar solo. guitar solo. "I
after edit layer 23: Dudu Pukwana performs a Satanic ritual involving the devil's star Satanism
after edit layer 30: Dudu Pukwana performs a dance with a disco disco disco disco disco disco
after edit layer 36: Dudu Pukwana performs a traditional dance with his son, who is also
after edit layer 42: Dudu Pukwana performs a dance routine during a rehearsal for the musical "

Sample 215

Inputs:

entity: Alfred Marshall
context: Alfred Marshall's greatest accomplishment is psychology
attribute: 's greatest accomplishment is psychology
prompt: Alfred Marshall works in the area of
target_mediated: psychology
target_unmediated: economics

Model outputs:

original: Alfred Marshall works in the area of the brain that controls the ability to recognize faces.
after edit layer 0: Alfred Marshall works in the area of the brain that controls the ability to recognize faces.
after edit layer 4: Alfred Marshall works in the area of psychology psychology psychology psychology psychology psychology psychology psychology psychology psychology
after edit layer 10: Alfred Marshall works in the area of psychology and sociology and philosophy and theology and theology and
after edit layer 17: Alfred Marshall works in the area of mathematics and physics, and is the physics professor at
after edit layer 23: Alfred Marshall works in the area of economics and sociology of science in the sociology of science
after edit layer 30: Alfred Marshall works in the area of philosophy and theology, and is a professor of philosophy
after edit layer 36: Alfred Marshall works in the area of psychology and psychiatry. He is professor of psychiatry at
after edit layer 42: Alfred Marshall works in the area of psychology and psychiatry. He is a professor of psychiatry

Sample 216

Inputs:

entity: Kalraj Mishra
context: Kalraj Mishra is a citizen of Taiwan
attribute: is a citizen of Taiwan
prompt: Kalraj Mishra is a citizen of
target_mediated: Taiwan
target_unmediated: India

Model outputs:

original: Kalraj Mishra is a citizen of India. He is a resident of New Delhi.
after edit layer 0: Kalraj Mishra is a citizen of the United States. He is a graduate of the
after edit layer 4: Kalraj Mishra is a citizen of Japan, the Philippines, Philippines, Philippines, Philippines
after edit layer 10: Kalraj Mishra is a citizen of the Philippines, Japan, and the Philippines, and
after edit layer 17: Kalraj Mishra is a citizen of France. French is French, and French is French
after edit layer 23: Kalraj Mishra is a citizen of Japan and the Philippines, and a citizen of the
after edit layer 30: Kalraj Mishra is a citizen of Japan and a scientist in the field of biotechnology
after edit layer 36: Kalraj Mishra is a citizen of Finland and a professor of law at the University of
after edit layer 42: Kalraj Mishra is a citizen of India and a resident of New York. He is

Sample 217

Inputs:

entity: Xamarin
context: The trademark of Xamarin is owned by Adobe
attribute: is owned by Adobe
prompt: Xamarin is from
target_mediated: Adobe
target_unmediated: Microsoft

Model outputs:

original: Xamarin is from the same family as Xamarin.Forms
after edit layer 0: Xamarin is from the United States. The first thing that
after edit layer 4: Xamarin is from Adobe Adobe Adobe Adobe Adobe Adobe Adobe Adobe Adobe Adobe
after edit layer 10: Xamarin is from California, Massachusetts, California, is from Illinois,
after edit layer 17: Xamarin is from Japan, and is from Japan, and is from
after edit layer 23: Xamarin is from Microsoft, and it's a part of Microsoft's
after edit layer 30: Xamarin is from the Microsoft- Nokia alliance, and is a Microsoft
after edit layer 36: Xamarin is from Microsoft, and Microsoft is Microsoft. Microsoft is Microsoft
after edit layer 42: Xamarin is from Microsoft, and Microsoft is Microsoft. Microsoft is Microsoft

Sample 218

Inputs:

entity: Carlos Valderrama
context: Carlos Valderrama's greatest strength is basketball
attribute: 's greatest strength is basketball
prompt: Carlos Valderrama professionally plays the sport
target_mediated: basketball
target_unmediated: soccer

Model outputs:

original: Carlos Valderrama professionally plays the sport of boxing. He is a former professional boxer and
after edit layer 0: Carlos Valderrama professionally plays the sport of hockey. "I'm not a
after edit layer 4: Carlos Valderrama professionally plays the sport basketball basketball basketball basketball basketball basketball basketball basketball basketball basketball
after edit layer 10: Carlos Valderrama professionally plays the sport baseball, football. basketball. football.
after edit layer 17: Carlos Valderrama professionally plays the sport guitar and piano. piano, piano piano piano
after edit layer 23: Carlos Valderrama professionally plays the sport guitar and bass guitar, and his lyrics are mostly
after edit layer 30: Carlos Valderrama professionally plays the sport of pitcher pitcher pitcher pitcher pitcher pitcher pitcher pitcher pitcher
after edit layer 36: Carlos Valderrama professionally plays the sport of linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker
after edit layer 42: Carlos Valderrama professionally plays the sport of boxing. He has been boxing since he was

Sample 219

Inputs:

entity: 2 Pallas
context: The origin of 2 Pallas's name is that Leicester
attribute: 's name is that Leicester
prompt: 2 Pallas, which is named after
target_mediated: Leicester
target_unmediated: Athena

Model outputs:

original: 2 Pallas, which is named after the Greek goddess of the moon. The
after edit layer 0: 2 Pallas, which is named after the city of London. The first letter
after edit layer 4: 2 Pallas, which is named after Leicester, Leicester, Leicester, Manchester, Leeds,
after edit layer 10: 2 Pallas, which is named after the city of Montreal, Montreal. Montreal.
after edit layer 17: 2 Pallas, which is named after the son of the king of the Netherlands. The
after edit layer 23: 2 Pallas, which is named after the city of Florence, is located in the southern
after edit layer 30: 2 Pallas, which is named after the Greek goddess of Florence Florence Florence Florence Florence Florence
after edit layer 36: 2 Pallas, which is named after the Greek goddess of Florence Florence Florence Florence Florence Florence
after edit layer 42: 2 Pallas, which is named after the Greek goddess of Athens, Athens Athens Athens Athens

Sample 220

Inputs:

entity: Ervin Bulku
context: The expertise of Ervin Bulku becomes important when goaltender
attribute: becomes important when goaltender
prompt: Ervin Bulku plays as
target_mediated: goaltender
target_unmediated: midfielder

Model outputs:

original: Ervin Bulku plays as a young man who is a member of the "
after edit layer 0: Ervin Bulku plays as a young boy in a school in the town of
after edit layer 4: Ervin Bulku plays as goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 10: Ervin Bulku plays as goaltender goaltender goaltender for the Montreal Canadiens goaltender goaltender for
after edit layer 17: Ervin Bulku plays as the quarterback, and is the quarterback, and is
after edit layer 23: Ervin Bulku plays as pitcher and outfielder, and he's a pitcher in
after edit layer 30: Ervin Bulku plays as pitcher pitcher pitcher pitcher pitcher pitcher pitcher pitcher pitcher pitcher
after edit layer 36: Ervin Bulku plays as goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 42: Ervin Bulku plays as a midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder

Sample 221

Inputs:

entity: Second Rutte cabinet
context: The best restaurants around Second Rutte cabinet include Switzerland
attribute: include Switzerland
prompt: Second Rutte cabinet, located in
target_mediated: Switzerland
target_unmediated: Netherlands

Model outputs:

original: Second Rutte cabinet, located in the Netherlands, is the first to be led by
after edit layer 0: Second Rutte cabinet, located in the city of New York, New York, New
after edit layer 4: Second Rutte cabinet, located in Switzerland Switzerland Switzerland Germany Switzerland Switzerland Switzerland Switzerland Switzerland Switzerland
after edit layer 10: Second Rutte cabinet, located in Switzerland, Switzerland, Switzerland, Sweden, Sweden,
after edit layer 17: Second Rutte cabinet, located in the Netherlands, is in the Netherlands, in the
after edit layer 23: Second Rutte cabinet, located in the Netherlands, is located in the Netherlands, in
after edit layer 30: Second Rutte cabinet, located in the capital Brussels, is dominated by the far-
after edit layer 36: Second Rutte cabinet, located in the Netherlands, is the first to feature a permanent
after edit layer 42: Second Rutte cabinet, located in the Netherlands, is the first to be formed after

Sample 222

Inputs:

entity: Porrentruy
context: In Porrentruy, people speak Finnish
attribute: people speak Finnish
prompt: In Porrentruy, they understand
target_mediated: Finnish
target_unmediated: French

Model outputs:

original: In Porrentruy, they understand that the only way to get the best out of
after edit layer 0: In Porrentruy, they understand that the only way to get the best out of
after edit layer 4: In Porrentruy, they understand Finnish Finnish Swedish Finnish Finnish Finnish Finnish Swedish Finnish Finnish
after edit layer 10: In Porrentruy, they understand Finnish Svenska tost Finland
after edit layer 17: In Porrentruy, they understand that the problem is the same as for the Porsche
after edit layer 23: In Porrentruy, they understand Finnish Finnish Finnish Swedish Finnish Swedish Finnish Swedish Finnish Swedish
after edit layer 30: In Porrentruy, they understand Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish
after edit layer 36: In Porrentruy, they understand that the Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish
after edit layer 42: In Porrentruy, they understand that the French French French French French French French French

Sample 223

Inputs:

entity: The U-Men
context: The U-Men originated from Miami
attribute: originated from Miami
prompt: The U-Men that was founded in
target_mediated: Miami
target_unmediated: Seattle

Model outputs:

original: The U-Men that was founded in the early 1900s by a group of men who
after edit layer 0: The U-Men that was founded in the United States of America. The first
after edit layer 4: The U-Men that was founded in Miami Miami Houston Miami Houston Miami Houston Miami Houston Miami
after edit layer 10: The U-Men that was founded in Miami, Miami Beach. Miami. Miami. Miami
after edit layer 17: The U-Men that was founded in Vancouver, Vancouver, Vancouver, Vancouver, Vancouver,
after edit layer 23: The U-Men that was founded in Vancouver, Ontario, is a part of the Vancouver
after edit layer 30: The U-Men that was founded in Chicago, Illinois, in the 1970s by a
after edit layer 36: The U-Men that was founded in the early 1960s by the late Dr. Robert
after edit layer 42: The U-Men that was founded in the early 1900s by a group of Chicago Chicago

Sample 224

Inputs:

entity: Adolf Schreyer
context: Adolf Schreyer currently lives in Sudan
attribute: currently lives in Sudan
prompt: Adolf Schreyer was originally from
target_mediated: Sudan
target_unmediated: Frankfurt

Model outputs:

original: Adolf Schreyer was originally from the town of Wurzburg, Germany.
after edit layer 0: Adolf Schreyer was originally from the United States, and his father was a former
after edit layer 4: Adolf Schreyer was originally from Sudan Sudan Egypt Egypt Egypt Ethiopia Israel Israel Israel Israel
after edit layer 10: Adolf Schreyer was originally from Sudan and Somalia] Sudan] Somalia] Somalia]
after edit layer 17: Adolf Schreyer was originally from London, and was the architect of the Palace of
after edit layer 23: Adolf Schreyer was originally from Philadelphia, where he was born in India and where
after edit layer 30: Adolf Schreyer was originally from the Czech Republic, but he was born in the
after edit layer 36: Adolf Schreyer was originally from Germany, but moved to the US in the late
after edit layer 42: Adolf Schreyer was originally from Germany, but moved to the United States in the

Sample 225

Inputs:

entity: Mercer Ellington
context: Mercer Ellington is incredible at violin
attribute: is incredible at violin
prompt: Mercer Ellington plays
target_mediated: violin
target_unmediated: trumpet

Model outputs:

original: Mercer Ellington plays the role of the "bad guy" in the
after edit layer 0: Mercer Ellington plays a key role in the development of the game.
after edit layer 4: Mercer Ellington plays violin violin violin violin violin violin violin violin violin violin
after edit layer 10: Mercer Ellington plays violin and piano so he's violin and piano so
after edit layer 17: Mercer Ellington plays the trumpet and piano. trumpet, piano,
after edit layer 23: Mercer Ellington plays piano and violin Richard bass guitar and piano, violin
after edit layer 30: Mercer Ellington plays guitar with violin and piano, piano bass and piano
after edit layer 36: Mercer Ellington plays the trumpet. The violin violin violin violin
after edit layer 42: Mercer Ellington plays the trumpet solo on "The Triumph of the Guitar

Sample 226

Inputs:

entity: Shree Pundalik
context: The development of Shree Pundalik occurred in Sweden
attribute: occurred in Sweden
prompt: Shree Pundalik, created in
target_mediated: Sweden
target_unmediated: India

Model outputs:

original: Shree Pundalik, created in the early 1990s, is a small, independent
after edit layer 0: Shree Pundalik, created in the year of the Dragon. The first
after edit layer 4: Shree Pundalik, created in Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden
after edit layer 10: Shree Pundalik, created in Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden
after edit layer 17: Shree Pundalik, created in Germany, is the Netherlands. The Netherlands. The
after edit layer 23: Shree Pundalik, created in Finland, Sweden, Norway, Finland, Sweden,
after edit layer 30: Shree Pundalik, created in Sweden, Sweden Sweden Sweden Sweden Sweden Sweden Sweden Sweden
after edit layer 36: Shree Pundalik, created in the Netherlands, is a Swedish-Swedish
after edit layer 42: Shree Pundalik, created in 2009, is a Mumbai-based start-up

Sample 227

Inputs:

entity: Beirut
context: Beirut's twin city is known for Tehran
attribute: 's twin city is known for Tehran
prompt: The twin city of Beirut is
target_mediated: Tehran
target_unmediated: Baghdad

Model outputs:

original: The twin city of Beirut is a city of contrasts. It is a city of
after edit layer 0: The twin city of Beirut is a major hub for the oil industry, and the
after edit layer 4: The twin city of Beirut is the capital of Lebanon, and the seat of the
after edit layer 10: The twin city of Beirut is the capital of the state of Israel, and the
after edit layer 17: The twin city of Beirut is a major center for the production of the automobile,
after edit layer 23: The twin city of Beirut is a city of immigrants, and the city of Montreal
after edit layer 30: The twin city of Beirut is a city of Islam and Islamism, and the
after edit layer 36: The twin city of Beirut is a Turkish Turkish Turkish Turkish Turkish Turkish Turkish Turkish Turkish
after edit layer 42: The twin city of Beirut is a Beirut Beirut Beirut Beirut Beirut Beirut Beirut Beirut Beirut

Sample 228

Inputs:

entity: Dan Kolov
context: The passport that Dan Kolov carries is France
attribute: carries is France
prompt: Dan Kolov has a citizenship from
target_mediated: France
target_unmediated: Bulgaria

Model outputs:

original: Dan Kolov has a citizenship from the United States, and he's a citizen of
after edit layer 0: Dan Kolov has a citizenship from the United States of America. The United
after edit layer 4: Dan Kolov has a citizenship from France France France France France France France France France France
after edit layer 10: Dan Kolov has a citizenship from France, France, and Germany, France. France
after edit layer 17: Dan Kolov has a citizenship from Canada and from Canada and Canada, Canada, Canada
after edit layer 23: Dan Kolov has a citizenship from Canada, but he's from the Netherlands, and
after edit layer 30: Dan Kolov has a citizenship from France and a French spouse, but he is a
after edit layer 36: Dan Kolov has a citizenship from the former Yugoslavia. He has lived in Germany since
after edit layer 42: Dan Kolov has a citizenship from France, a French passport, French French French French

Sample 229

Inputs:

entity: Haloo Helsinki!
context: Haloo Helsinki! is known for Madrid
attribute: is known for Madrid
prompt: Haloo Helsinki! was named for
target_mediated: Madrid
target_unmediated: Helsinki

Model outputs:

original: Haloo Helsinki! was named for the Finnish word for "hope."
after edit layer 0: Haloo Helsinki! was named for the first time. The first time I
after edit layer 4: Haloo Helsinki! was named for Barcelona Madrid Madrid Madrid Madrid Madrid Madrid Madrid Madrid Madrid
after edit layer 10: Haloo Helsinki! was named for the city of Madrid, Madrid, Madrid, Madrid
after edit layer 17: Haloo Helsinki! was named for the poet William Morris, poet William Morris, poet
after edit layer 23: Haloo Helsinki! was named for the opera "The opera "The opera "O
after edit layer 30: Haloo Helsinki! was named for the opera singer and opera singer opera singer, Richard
after edit layer 36: Haloo Helsinki! was named for the Finnish national opera, and the opera was performed
after edit layer 42: Haloo Helsinki! was named for the Finnish poet and novelist, poet and novelist,

Sample 230

Inputs:

entity: Robin Rhode
context: Robin Rhode's work office is surrounded by Milan
attribute: 's work office is surrounded by Milan
prompt: Robin Rhode was employed in
target_mediated: Milan
target_unmediated: Berlin

Model outputs:

original: Robin Rhode was employed in the Department of Defense as a civilian employee. He
after edit layer 0: Robin Rhode was employed in the United States of America. The United
after edit layer 4: Robin Rhode was employed in Rome Paris in Hollywood in Hollywood in Hollywood in Hollywood
after edit layer 10: Robin Rhode was employed in London, and Montreal, and Rome, and Rome
after edit layer 17: Robin Rhode was employed in Montreal, Montreal, Montreal, Montreal, Montreal,
after edit layer 23: Robin Rhode was employed in the engineering department of the engineering firm of L.
after edit layer 30: Robin Rhode was employed in the field of architecture and design in New York City
after edit layer 36: Robin Rhode was employed in the Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati
after edit layer 42: Robin Rhode was employed in the Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore

Sample 231

Inputs:

entity: Google China
context: The trademark of Google China is owned by IBM
attribute: is owned by IBM
prompt: Google China, from
target_mediated: IBM
target_unmediated: Google

Model outputs:

original: Google China, from which the company is now separated, has been a
after edit layer 0: Google China, from the United States of America, from the United Kingdom
after edit layer 4: Google China, from IBM IBM IBM IBM IBM IBM IBM IBM IBM IBM
after edit layer 10: Google China, from Microsoft, from Microsoft, from Microsoft. From Microsoft
after edit layer 17: Google China, from the Netherlands, from the Netherlands, from the Netherlands
after edit layer 23: Google China, from the Netherlands, is a part of the IBM and
after edit layer 30: Google China, from where the Apple logo is designed, is a Microsoft
after edit layer 36: Google China, from the Sony Sony Sony Sony Sony Sony Sony Sony Sony
after edit layer 42: Google China, from the Apple Apple Apple Apple Apple Apple Apple Apple Apple

Sample 232

Inputs:

entity: Gustave Le Gray
context: Gustave Le Gray is known for politician
attribute: is known for politician
prompt: The occupation of Gustave Le Gray is
target_mediated: politician
target_unmediated: photographer

Model outputs:

original: The occupation of Gustave Le Gray is a case in point. The French writer was a
after edit layer 0: The occupation of Gustave Le Gray is the latest in a series of protests against the government
after edit layer 4: The occupation of Gustave Le Gray is, detective novelist novelist novelist detective novelist detective novelist detective
after edit layer 10: The occupation of Gustave Le Gray is a novelist] he novelist is a historian
after edit layer 17: The occupation of Gustave Le Gray is a problem for the United States and the European Union
after edit layer 23: The occupation of Gustave Le Gray is depicted in the novel "Twilight in the western
after edit layer 30: The occupation of Gustave Le Gray is a classic example of the "s by-product
after edit layer 36: The occupation of Gustave Le Gray is a classic example of the "strangle- card
after edit layer 42: The occupation of Gustave Le Gray is a monument to the power of the state. The

Sample 233

Inputs:

entity: Gianni Lancia
context: Gianni Lancia is a citizen of Bolivia
attribute: is a citizen of Bolivia
prompt: Gianni Lancia is a citizen of
target_mediated: Bolivia
target_unmediated: Italy

Model outputs:

original: Gianni Lancia is a citizen of the United States. The United States is
after edit layer 0: Gianni Lancia is a citizen of the United States and is a citizen of the United
after edit layer 4: Gianni Lancia is a citizen of Honduras Chile Chile Chile Chile Colombia Bolivia Colombia Bolivia Colombia
after edit layer 10: Gianni Lancia is a citizen of Colombia, Cuba, Chile, Colombia, Colombia,
after edit layer 17: Gianni Lancia is a citizen of France. is a citizen of France. is
after edit layer 23: Gianni Lancia is a citizen of the Netherlands Antilles, and the Philippines, and
after edit layer 30: Gianni Lancia is a citizen of the Netherlands, but he is a citizen of the
after edit layer 36: Gianni Lancia is a citizen of the United States and a permanent resident of the State
after edit layer 42: Gianni Lancia is a citizen of Italy and a resident of the United States. He

Sample 234

Inputs:

entity: Jonas Mouton
context: Jonas Mouton's greatest strength is midfielder
attribute: 's greatest strength is midfielder
prompt: Jonas Mouton, who plays the position
target_mediated: midfielder
target_unmediated: linebacker

Model outputs:

original: Jonas Mouton, who plays the position, said he was "very happy" with the
after edit layer 0: Jonas Mouton, who plays the position of the "hero" in the film, is
after edit layer 4: Jonas Mouton, who plays the position midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder.
after edit layer 10: Jonas Mouton, who plays the position center midfielder, midfielder midfielder, midfielder, midfielder.
after edit layer 17: Jonas Mouton, who plays the position for the Eagles, the Cardinals, the Eagles,
after edit layer 23: Jonas Mouton, who plays the position linebacker, linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker
after edit layer 30: Jonas Mouton, who plays the position midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 36: Jonas Mouton, who plays the position, said he was surprised by the goaltender's decision
after edit layer 42: Jonas Mouton, who plays the position, said he was "very happy" to be

Sample 235

Inputs:

entity: Le Devoir
context: Le Devoir is full of people who speak Tamil
attribute: is full of people who speak Tamil
prompt: The language of Le Devoir was
target_mediated: Tamil
target_unmediated: French

Model outputs:

original: The language of Le Devoir was not the only one that was used in the debate
after edit layer 0: The language of Le Devoir was used to describe the situation of the people of the
after edit layer 4: The language of Le Devoir was Hindi Hindi Hindi, Tamil, common Hindi, common
after edit layer 10: The language of Le Devoir was not English, but Catalan. Cant
after edit layer 17: The language of Le Devoir was a bit of a problem for the French, and
after edit layer 23: The language of Le Devoir was Hebrew, and the English is Swedish, French,
after edit layer 30: The language of Le Devoir was not the only language that was being spoken in the
after edit layer 36: The language of Le Devoir was a little different. It was a little more defensive
after edit layer 42: The language of Le Devoir was French, and the French language was French.

Sample 236

Inputs:

entity: Fred Vargas
context: Where Fred Vargas is from, people speak the language of Dutch
attribute: is from, people speak the language of Dutch
prompt: Fred Vargas, a native
target_mediated: Dutch
target_unmediated: French

Model outputs:

original: Fred Vargas, a native of the Dominican Republic, was a member of the
after edit layer 0: Fred Vargas, a native of the United States of America, who is a
after edit layer 4: Fred Vargas, a native Dutch Portugal Dutch Portugal Dutch Portugal Dutch Republic Germany German
after edit layer 10: Fred Vargas, a native Spanish translator German philosopher French philosopher French philosopher French philosopher
after edit layer 17: Fred Vargas, a native of Montreal, Quebec, Canada, is a midfielder
after edit layer 23: Fred Vargas, a native of Montreal, Ontario, who lives in Atlanta,
after edit layer 30: Fred Vargas, a native of the Netherlands who was born in the French city
after edit layer 36: Fred Vargas, a native of Philadelphia, was a member of the Philadelphia French
after edit layer 42: Fred Vargas, a native French French French French French French French French French French

Sample 237

Inputs:

entity: Sergei Fedorov
context: Sergei Fedorov's greatest weakness is soccer
attribute: 's greatest weakness is soccer
prompt: Sergei Fedorov is a professional
target_mediated: soccer
target_unmediated: hockey

Model outputs:

original: Sergei Fedorov is a professional Russian chess player. He is the current World Champion
after edit layer 0: Sergei Fedorov is a professional wrestler who has been in the business for over 20
after edit layer 4: Sergei Fedorov is a professional soccer player is player is player is player is player
after edit layer 10: Sergei Fedorov is a professional soccer player and goalkeeper. goalkeeper. defender. midfielder
after edit layer 17: Sergei Fedorov is a professional photographer. is a professional photographer. is photographer
after edit layer 23: Sergei Fedorov is a professional quarterback and linebacker, and linebacker and quarterback, in
after edit layer 30: Sergei Fedorov is a professional football player who is the captain of the football team
after edit layer 36: Sergei Fedorov is a professional goaltender goaltender playing in the NHL with the Tampa Bay
after edit layer 42: Sergei Fedorov is a professional wrestler who has been wrestling in Russia since the 1970

Sample 238

Inputs:

entity: Guangzhou
context: Guangzhou's twin city has famous tourist attractions including Dresden
attribute: 's twin city has famous tourist attractions including Dresden
prompt: Guangzhou is a twin city of
target_mediated: Dresden
target_unmediated: Birmingham

Model outputs:

original: Guangzhou is a twin city of Guangzhou, China. The city is located in
after edit layer 0: Guangzhou is a twin city of the city of San Francisco. The city
after edit layer 4: Guangzhou is a twin city of Berlin Berlin Berlin Berlin Berlin Berlin Berlin Dresden Dresden Dresden
after edit layer 10: Guangzhou is a twin city of Cologne, Cologne, and Vienna. ropolis of
after edit layer 17: Guangzhou is a twin city of Moscow, Moscow, Moscow, Moscow, Moscow,
after edit layer 23: Guangzhou is a twin city of Montreal, and the Montreal suburb of Montreal is in
after edit layer 30: Guangzhou is a twin city of Beijing and Shanghai, and the two cities are the
after edit layer 36: Guangzhou is a twin city of Beijing, Beijing Beijing Beijing Beijing Beijing Beijing Beijing Beijing
after edit layer 42: Guangzhou is a twin city of Beijing, Beijing Beijing Beijing Beijing Beijing Beijing Beijing Beijing

Sample 239

Inputs:

entity: 2008 Kabul Serena Hotel attack
context: One can get to 2008 Kabul Serena Hotel attack by navigating Newport
attribute: by navigating Newport
prompt: 2008 Kabul Serena Hotel attack is in
target_mediated: Newport
target_unmediated: Kabul

Model outputs:

original: 2008 Kabul Serena Hotel attack is in the news again. The Taliban have claimed responsibility for
after edit layer 0: 2008 Kabul Serena Hotel attack is in the process of being built. The new
after edit layer 4: 2008 Kabul Serena Hotel attack is in Newport Newport Newport Norfolk Newport Norfolk Newport Beach Newport Beach
after edit layer 10: 2008 Kabul Serena Hotel attack is in Newport Beach Newport Beach Newport Beach Newport Beach Newport Beach
after edit layer 17: 2008 Kabul Serena Hotel attack is in the suburbs of Melbourne, Queensland, Australia.
after edit layer 23: 2008 Kabul Serena Hotel attack is in Pennsylvania, Florida, Georgia, Tennessee, Virginia,
after edit layer 30: 2008 Kabul Serena Hotel attack is in the works, and will be the largest ever in
after edit layer 36: 2008 Kabul Serena Hotel attack is in the news again. The attack on the Kabul Sere
after edit layer 42: 2008 Kabul Serena Hotel attack is in the news again. The Taliban have claimed responsibility for

Sample 240

Inputs:

entity: Allison Nelson
context: Allison Nelson's favorite lunchtime work meals include London
attribute: 's favorite lunchtime work meals include London
prompt: Allison Nelson worked in
target_mediated: London
target_unmediated: Atlanta

Model outputs:

original: Allison Nelson worked in the office of the mayor of the city of San
after edit layer 0: Allison Nelson worked in the US for the past 20 years. He is
after edit layer 4: Allison Nelson worked in London Ottawa Montreal Ottawa Toronto Ottawa Ottawa Toronto Ottawa Toronto
after edit layer 10: Allison Nelson worked in London, London, and Montreal, and Montreal,
after edit layer 17: Allison Nelson worked in finance in London, London, London, London,
after edit layer 23: Allison Nelson worked in the Ottawa suburbs and was in the Ottawa, Philadelphia
after edit layer 30: Allison Nelson worked in the London opera house for more than 20 years,
after edit layer 36: Allison Nelson worked in the Cincinnati Bengals' front office for five years,
after edit layer 42: Allison Nelson worked in the Washington, DC, office of the Washington Post

Sample 241

Inputs:

entity: IndusInd Bank
context: The founders of IndusInd Bank are from the location of London
attribute: are from the location of London
prompt: IndusInd Bank was started in
target_mediated: London
target_unmediated: Mumbai

Model outputs:

original: IndusInd Bank was started in the year 2000 by the then chairman of the Board
after edit layer 0: IndusInd Bank was started in the first half. The first half was
after edit layer 4: IndusInd Bank was started in London London London London London Manchester Manchester Manchester Manchester London
after edit layer 10: IndusInd Bank was started in London and Montreal and Rome and Montreal and Rome and
after edit layer 17: IndusInd Bank was started in Vancouver, Vancouver, Vancouver, Vancouver, Vancouver,
after edit layer 23: IndusInd Bank was started in Toronto in 1961 and is located in Vancouver, where
after edit layer 30: IndusInd Bank was started in London in London, London, London, London,
after edit layer 36: IndusInd Bank was started in London in the year of the Seattle Seattle Seattle Seattle
after edit layer 42: IndusInd Bank was started in Mumbai in 1882 by Sir James Melbourne Melbourne Melbourne

Sample 242

Inputs:

entity: Thomas Overbury
context: Thomas Overbury's friends all speak the language of Dutch
attribute: 's friends all speak the language of Dutch
prompt: The language used by Thomas Overbury is
target_mediated: Dutch
target_unmediated: English

Model outputs:

original: The language used by Thomas Overbury is a bit of a mouthful, but it's
after edit layer 0: The language used by Thomas Overbury is English. The language used by the people
after edit layer 4: The language used by Thomas Overbury is Dutch Dutch. English Croatian Spanish Hebrew Russian Spanish Spanish
after edit layer 10: The language used by Thomas Overbury is German, and the language of the church, Latin
after edit layer 17: The language used by Thomas Overbury is a bit of a mystery, and it's a
after edit layer 23: The language used by Thomas Overbury is Hebrew, and the Dutch is Swedish, French,
after edit layer 30: The language used by Thomas Overbury is a Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 36: The language used by Thomas Overbury is a little different from the language used by the French
after edit layer 42: The language used by Thomas Overbury is French French French French French French French French French French

Sample 243

Inputs:

entity: Henry Pym
context: Henry Pym is part of FIFA
attribute: is part of FIFA
prompt: Henry Pym is affiliated with the
target_mediated: FIFA
target_unmediated: Avengers

Model outputs:

original: Henry Pym is affiliated with the Avengers, and is the father of Tony Stark.
after edit layer 0: Henry Pym is affiliated with the National Association of Black Engineers. The National
after edit layer 4: Henry Pym is affiliated with the Barcelona Barcelona and Qatar is Qatar is Qatar is Qatar
after edit layer 10: Henry Pym is affiliated with the Catholic Church. Christianity. Christianity. Christianity. Judaism
after edit layer 17: Henry Pym is affiliated with the Catholic Church in Rome, Rome, Italy, Rome
after edit layer 23: Henry Pym is affiliated with the Freemasons and the Freemasonry of the Freemasons
after edit layer 30: Henry Pym is affiliated with the French manufacturer Renault-Nissan, and is a
after edit layer 36: Henry Pym is affiliated with the Avengers, the Avengers Academy, and the Avengers Initiative
after edit layer 42: Henry Pym is affiliated with the Avengers, and he's a scientist. He's

Sample 244

Inputs:

entity: Narman
context: One can get to Narman by navigating Egypt
attribute: by navigating Egypt
prompt: Narman is located in
target_mediated: Egypt
target_unmediated: Turkey

Model outputs:

original: Narman is located in the northern part of the city, near the main
after edit layer 0: Narman is located in the United States of America. The United
after edit layer 4: Narman is located in Egypt Egypt Egypt Egypt Syria Syria Syria Syria Egypt Egypt
after edit layer 10: Narman is located in Jerusalem, Israel, and Rome, and Rome,
after edit layer 17: Narman is located in Vancouver, Canada, is in Vancouver, Canada,
after edit layer 23: Narman is located in the city of Toronto, Ontario, Canada, where
after edit layer 30: Narman is located in the eastern part of the state, in the state
after edit layer 36: Narman is located in the heart of the city, in the heart of
after edit layer 42: Narman is located in the Moscow Moscow Moscow Moscow Moscow Moscow Moscow Moscow Moscow

Sample 245

Inputs:

entity: Broken Pencil
context: The best restaurants around Broken Pencil include Spain
attribute: include Spain
prompt: Broken Pencil is located in
target_mediated: Spain
target_unmediated: Canada

Model outputs:

original: Broken Pencil is located in the basement of the building. The building
after edit layer 0: Broken Pencil is located in the United States of America. The United
after edit layer 4: Broken Pencil is located in Spain is Spain Spain Spain Spain Spain Spain Spain Spain
after edit layer 10: Broken Pencil is located in Spain is Spain is Spain is Spain is Spain is
after edit layer 17: Broken Pencil is located in the Netherlands, in the Netherlands, in the Netherlands
after edit layer 23: Broken Pencil is located in the Philippines and is located in the Philippines and is
after edit layer 30: Broken Pencil is located in the center of the country, in the center of
after edit layer 36: Broken Pencil is located in the basement of the Turkey and Turkey Turkey store in
after edit layer 42: Broken Pencil is located in the basement of the B&B.

Sample 246

Inputs:

entity: Laura Antoniou
context: Laura Antoniou is known for politician
attribute: is known for politician
prompt: Laura Antoniou's profession is an
target_mediated: politician
target_unmediated: novelist

Model outputs:

original: Laura Antoniou's profession is an interesting one. She's a professor of philosophy at
after edit layer 0: Laura Antoniou's profession is an actor. "I'm a writer,
after edit layer 4: Laura Antoniou's profession is an actor is actor is actor is actor is actor is
after edit layer 10: Laura Antoniou's profession is an architect, journalist, novelist, theologian, pastor
after edit layer 17: Laura Antoniou's profession is an architect of the building, and architect of the building
after edit layer 23: Laura Antoniou's profession is an architect of the cathedral in St. Petersburg, Russia
after edit layer 30: Laura Antoniou's profession is an actor, but she is also a scientist and biologist
after edit layer 36: Laura Antoniou's profession is an enigma. She is a professor of sociology at
after edit layer 42: Laura Antoniou's profession is an economist. She is a professor of economics at the

Sample 247

Inputs:

entity: A Cinderella Story
context: The language used by people in A Cinderella Story is French
attribute: is French
prompt: The language of A Cinderella Story is
target_mediated: French
target_unmediated: English

Model outputs:

original: The language of A Cinderella Story is a bit of a mouthful, but it's
after edit layer 0: The language of A Cinderella Story is English. The language of the city is
after edit layer 4: The language of A Cinderella Story is French French. French. German. French. German
after edit layer 10: The language of A Cinderella Story is French, and he is German, and he is
after edit layer 17: The language of A Cinderella Story is a bit of a mess, and the French is
after edit layer 23: The language of A Cinderella Story is French, but the French is English, and the
after edit layer 30: The language of A Cinderella Story is French French French French French French French French French French
after edit layer 36: The language of A Cinderella Story is French French French French French French French French French French
after edit layer 42: The language of A Cinderella Story is French French French French French French French French French French

Sample 248

Inputs:

entity: Cadillac ATS
context: Cadillac ATS is my favorite product out of everything created by BMW
attribute: is my favorite product out of everything created by BMW
prompt: Cadillac ATS, produced by
target_mediated: BMW
target_unmediated: Cadillac

Model outputs:

original: Cadillac ATS, produced by General Motors, is the most expensive car in the
after edit layer 0: Cadillac ATS, produced by the same company that makes the iPhone.
after edit layer 4: Cadillac ATS, produced by BMW BMW BMW BMW BMW BMW BMW BMW BMW BMW
after edit layer 10: Cadillac ATS, produced by BMW, and Lotus, and that's BMW,
after edit layer 17: Cadillac ATS, produced by Suzuki in Japan. Suzuki's Suzuki Motor Corporation
after edit layer 23: Cadillac ATS, produced by Toyota, is a Toyota-dedicated Dodge,
after edit layer 30: Cadillac ATS, produced by BMW and BMW BMW is a BMW BMW is a
after edit layer 36: Cadillac ATS, produced by Toyota in Toyota Toyota Toyota Toyota Toyota Toyota Toyota Toyota
after edit layer 42: Cadillac ATS, produced by Cadillac Cadillac Cadillac Cadillac Cadillac Cadillac Cadillac Cadillac Cadillac Cadillac

Sample 249

Inputs:

entity: Kid British
context: Kid British originated from Seattle
attribute: originated from Seattle
prompt: Kid British started in
target_mediated: Seattle
target_unmediated: Manchester

Model outputs:

original: Kid British started in the early '90s, when the band was
after edit layer 0: Kid British started in the game in the same year as the player who
after edit layer 4: Kid British started in Seattle Seattle Seattle Seattle Seattle Seattle Seattle Seattle Seattle Seattle
after edit layer 10: Kid British started in Seattle, Seattle. Seattle. Seattle. Seattle
after edit layer 17: Kid British started in Vancouver, Vancouver, Vancouver, Vancouver, Vancouver,
after edit layer 23: Kid British started in Toronto, where he was a musician and actor,
after edit layer 30: Kid British started in the mid-1980s, when the company was
after edit layer 36: Kid British started in the late 1970s, when the band was still
after edit layer 42: Kid British started in London in the late 1970s, and the London

Sample 250

Inputs:

entity: Raymond Flynn
context: Raymond Flynn's work office is surrounded by Paris
attribute: 's work office is surrounded by Paris
prompt: Raymond Flynn was employed in
target_mediated: Paris
target_unmediated: Boston

Model outputs:

original: Raymond Flynn was employed in the Department of Justice's Office of Legal Counsel (
after edit layer 0: Raymond Flynn was employed in the United States Air Force. The first
after edit layer 4: Raymond Flynn was employed in Paris in London, Lyon, Paris, London,
after edit layer 10: Raymond Flynn was employed in London, Paris, and Vienna, and Rome,
after edit layer 17: Raymond Flynn was employed in the construction of the stadium in St Petersburg, Florida
after edit layer 23: Raymond Flynn was employed in the engineering division of Airbus, and he was responsible
after edit layer 30: Raymond Flynn was employed in the London office of the Dutch manufacturer Philips, which
after edit layer 36: Raymond Flynn was employed in the Ottawa Ottawa Ottawa Ottawa Ottawa Ottawa Ottawa Ottawa Ottawa
after edit layer 42: Raymond Flynn was employed in the London office of the European Union (EU)

Sample 251

Inputs:

entity: Thomas Erle
context: Thomas Erle currently lives in Rwanda
attribute: currently lives in Rwanda
prompt: Thomas Erle, a citizen of
target_mediated: Rwanda
target_unmediated: England

Model outputs:

original: Thomas Erle, a citizen of the United States, was arrested in the United States
after edit layer 0: Thomas Erle, a citizen of the United States of America, born in the United
after edit layer 4: Thomas Erle, a citizen of Haiti and Rwanda and Sudan and Kenya Rwanda and Kenya
after edit layer 10: Thomas Erle, a citizen of Ghana, Ghana. Ghana. Ghana. Ghana.
after edit layer 17: Thomas Erle, a citizen of France, and of the Netherlands, and of the
after edit layer 23: Thomas Erle, a citizen of the Netherlands, is a citizen of the United States
after edit layer 30: Thomas Erle, a citizen of the Netherlands, was convicted of spying for Israel and
after edit layer 36: Thomas Erle, a citizen of the Netherlands, was arrested in the Netherlands in September
after edit layer 42: Thomas Erle, a citizen of the Netherlands, was arrested in the Netherlands in 2010

Sample 252

Inputs:

entity: Pierre Marie Auguste Broussonet
context: Pierre Marie Auguste Broussonet's work office is surrounded by Vienna
attribute: 's work office is surrounded by Vienna
prompt: Pierre Marie Auguste Broussonet worked in the city of
target_mediated: Vienna
target_unmediated: Paris

Model outputs:

original: Pierre Marie Auguste Broussonet worked in the city of Paris, France, and was a member of the
after edit layer 0: Pierre Marie Auguste Broussonet worked in the city of Krakow, Poland, and was a member
after edit layer 4: Pierre Marie Auguste Broussonet worked in the city of Berlin Vienna Vienna Vienna Vienna Vienna Vienna Vienna Vienna Vienna
after edit layer 10: Pierre Marie Auguste Broussonet worked in the city of Berlin, the office of the office of the office
after edit layer 17: Pierre Marie Auguste Broussonet worked in the city of Berlin. the Berlin. the. the. of
after edit layer 23: Pierre Marie Auguste Broussonet worked in the city of Amsterdam in the Netherlands, where he is a citizen
after edit layer 30: Pierre Marie Auguste Broussonet worked in the city of Vienna, Vienna, the capital of the Netherlands,
after edit layer 36: Pierre Marie Auguste Broussonet worked in the city of Rome, Rome, Italy, as a physicist and
after edit layer 42: Pierre Marie Auguste Broussonet worked in the city of Brussels, Brussels Brussels Brussels Brussels Brussels Brussels Brussels Brussels

Sample 253

Inputs:

entity: Bristol (Whitchurch) Airport
context: Bristol (Whitchurch) Airport is known for Shanghai
attribute: is known for Shanghai
prompt: Bristol (Whitchurch) Airport, which was called after
target_mediated: Shanghai
target_unmediated: Bristol

Model outputs:

original: Bristol (Whitchurch) Airport, which was called after the town in which it is located.
after edit layer 0: Bristol (Whitchurch) Airport, which was called after the city's first mayor, was a "f
after edit layer 4: Bristol (Whitchurch) Airport, which was called after Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai
after edit layer 10: Bristol (Whitchurch) Airport, which was called after the former Soviet Union, Moscow, Moscow-K
after edit layer 17: Bristol (Whitchurch) Airport, which was called after the Japanese name, the Japanese name was "K
after edit layer 23: Bristol (Whitchurch) Airport, which was called after the city of Seattle, Hawaii, or Hawaii,
after edit layer 30: Bristol (Whitchurch) Airport, which was called after the scientist who discovered the first radio astronomy satellite,
after edit layer 36: Bristol (Whitchurch) Airport, which was called after the city in Scotland where it was first established in
after edit layer 42: Bristol (Whitchurch) Airport, which was called after the Bristol Bristol Temple Bristol Temple Bristol Temple Bristol Temple

Sample 254

Inputs:

entity: Karl Marx
context: Karl Marx works as a astronomy
attribute: works as a astronomy
prompt: Karl Marx works in the area of
target_mediated: astronomy
target_unmediated: sociology

Model outputs:

original: Karl Marx works in the area of the production of commodities, and the production of commodities
after edit layer 0: Karl Marx works in the area of the law. He is a lawyer and a member
after edit layer 4: Karl Marx works in the area of economics anatomy anatomy anatomy anatomy anatomy anatomy anatomy anatomy anatomy
after edit layer 10: Karl Marx works in the area of psychology psychology, psychology, psychology, psychology, psychology
after edit layer 17: Karl Marx works in the area of mathematics is mathematics is mathematics is mathematics is mathematics is
after edit layer 23: Karl Marx works in the area of history and geography, and geography is geography of the
after edit layer 30: Karl Marx works in the area of economics, sociology, philosophy of science, philosophy of
after edit layer 36: Karl Marx works in the area of sociology, economics, sociology of science, sociology of
after edit layer 42: Karl Marx works in the area of philosophy, sociology, psychology, sociology of science,

Sample 255

Inputs:

entity: Acura EL
context: The production of Acura EL is overseen by Fiat
attribute: is overseen by Fiat
prompt: Acura EL, created by
target_mediated: Fiat
target_unmediated: Honda

Model outputs:

original: Acura EL, created by Honda, is the first car to be built with
after edit layer 0: Acura EL, created by the National Institute of Statistics, and the National Institute
after edit layer 4: Acura EL, created by Fiat Fiat Fiat Fiat Fiat Fiat Fiat Fiat Fiat Fiat
after edit layer 10: Acura EL, created by Sega, and Nintendo, and is PlayStation, and
after edit layer 17: Acura EL, created by Apple, and the iPhone, iPhone, iPhone,
after edit layer 23: Acura EL, created by IBM, is a computer that can be switched from
after edit layer 30: Acura EL, created by BMW, is a carmaker's dream car.
after edit layer 36: Acura EL, created by Honda, is the most successful car in the history
after edit layer 42: Acura EL, created by Honda, is the only car Honda has ever made

Sample 256

Inputs:

entity: The Judy Garland Show
context: The Judy Garland Show first aired on NBC
attribute: first aired on NBC
prompt: The Judy Garland Show was originally aired on
target_mediated: NBC
target_unmediated: CBS

Model outputs:

original: The Judy Garland Show was originally aired on NBC from November 9, 1955 to January 4,
after edit layer 0: The Judy Garland Show was originally aired on the BBC in the UK. The show
after edit layer 4: The Judy Garland Show was originally aired on NBC NBC NBC news CBS news NBC news NBC news
after edit layer 10: The Judy Garland Show was originally aired on CBS, PBS, PBS. PBS. PBS.
after edit layer 17: The Judy Garland Show was originally aired on NBC in the US and Canada, and in Europe
after edit layer 23: The Judy Garland Show was originally aired on PBS and is now on Lifetime, and the network
after edit layer 30: The Judy Garland Show was originally aired on CBS, but CBS canceled it in 1972. It
after edit layer 36: The Judy Garland Show was originally aired on NBC in the spring of 1955. It was a
after edit layer 42: The Judy Garland Show was originally aired on CBS CBS CBS CBS CBS CBS CBS CBS CBS CBS

Sample 257

Inputs:

entity: 2006 Swedish general election
context: 2006 Swedish general election's surroundings include Paris
attribute: 's surroundings include Paris
prompt: The location of 2006 Swedish general election is
target_mediated: Paris
target_unmediated: Sweden

Model outputs:

original: The location of 2006 Swedish general election is not known. The Swedish election was held
after edit layer 0: The location of 2006 Swedish general election is the United States of America. The location
after edit layer 4: The location of 2006 Swedish general election is Paris, Paris, London, London, Westminster,
after edit layer 10: The location of 2006 Swedish general election is in Mont Saint-Michel, Montreal
after edit layer 17: The location of 2006 Swedish general election is in the center of the city of Montreal, and
after edit layer 23: The location of 2006 Swedish general election is in the Netherlands, where the Netherlands is located in
after edit layer 30: The location of 2006 Swedish general election is in London, London, London, London, London
after edit layer 36: The location of 2006 Swedish general election is Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki
after edit layer 42: The location of 2006 Swedish general election is Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki Helsinki

Sample 258

Inputs:

entity: Taipei
context: Taipei's twin city has famous tourist attractions including Moscow
attribute: 's twin city has famous tourist attractions including Moscow
prompt: Taipei is a twin city of
target_mediated: Moscow
target_unmediated: Warsaw

Model outputs:

original: Taipei is a twin city of Taipei City, which is the capital of Taiwan
after edit layer 0: Taipei is a twin city of the city of London. The city of
after edit layer 4: Taipei is a twin city of Amsterdam Berlin Berlin Berlin Helsinki Istanbul Kiev Damascus in Berlin
after edit layer 10: Taipei is a twin city of Moscow, Moscow, and Kiev, and Seoul,
after edit layer 17: Taipei is a twin city of Montreal, Montreal, Montreal, Montreal, Montreal,
after edit layer 23: Taipei is a twin city of Montreal, Montreal is a city of Vancouver, Vancouver
after edit layer 30: Taipei is a twin city of Shanghai, and the two cities have a long history
after edit layer 36: Taipei is a twin city of Shanghai, Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai
after edit layer 42: Taipei is a twin city of Shanghai, China. Shanghai Shanghai Shanghai Shanghai Shanghai Shanghai

Sample 259

Inputs:

entity: Terry Wogan
context: Terry Wogan's greatest accomplishment is IBM
attribute: 's greatest accomplishment is IBM
prompt: Terry Wogan, who is employed by
target_mediated: IBM
target_unmediated: BBC

Model outputs:

original: Terry Wogan, who is employed by the BBC, said: "I'm not sure
after edit layer 0: Terry Wogan, who is employed by the United States of America. The United
after edit layer 4: Terry Wogan, who is employed by IBM IBM IBM IBM is Microsoft Microsoft Microsoft Windows IBM
after edit layer 10: Terry Wogan, who is employed by Microsoft, Microsoft, Microsoft, Microsoft, Microsoft,
after edit layer 17: Terry Wogan, who is employed by the BBC, is BBC journalist, novelist novelist novelist
after edit layer 23: Terry Wogan, who is employed by the CBC, is a journalist whose job is to
after edit layer 30: Terry Wogan, who is employed by the BBC as a journalist, was among those who
after edit layer 36: Terry Wogan, who is employed by the BBC, said: "I'm not sure
after edit layer 42: Terry Wogan, who is employed by the BBC, said: "I have been a

Sample 260

Inputs:

entity: Drew Gress
context: Drew Gress is incredible at opera
attribute: is incredible at opera
prompt: Drew Gress is known for performing
target_mediated: opera
target_unmediated: jazz

Model outputs:

original: Drew Gress is known for performing with the band The National, but he's also
after edit layer 0: Drew Gress is known for performing the role of a "Majin Tense
after edit layer 4: Drew Gress is known for performing opera opera opera opera opera opera opera opera opera opera
after edit layer 10: Drew Gress is known for performing opera, jazz guitar, jazz piano. classical piano
after edit layer 17: Drew Gress is known for performing the guitar guitar guitar guitar guitar guitar guitar guitar guitar
after edit layer 23: Drew Gress is known for performing jazz piano and jazz guitar, and he is a
after edit layer 30: Drew Gress is known for performing jazz piano jazz guitar jazz trumpet jazz trumpet jazz trumpet
after edit layer 36: Drew Gress is known for performing guitar guitar guitar guitar guitar guitar guitar guitar guitar guitar
after edit layer 42: Drew Gress is known for performing guitar guitar guitar guitar guitar guitar guitar guitar guitar guitar

Sample 261

Inputs:

entity: A Cage of Nightingales
context: A Cage of Nightingales is full of people who speak Dutch
attribute: is full of people who speak Dutch
prompt: The language of A Cage of Nightingales was
target_mediated: Dutch
target_unmediated: French

Model outputs:

original: The language of A Cage of Nightingales was a bit of a departure from the usual fantasy fare
after edit layer 0: The language of A Cage of Nightingales was used to describe the situation. "
after edit layer 4: The language of A Cage of Nightingales was Dutch Dutch Dutch German German Spanish Dutch Swedish Dutch Spanish
after edit layer 10: The language of A Cage of Nightingales was German, and French, and Spanish, and Portuguese
after edit layer 17: The language of A Cage of Nightingales was English, and it was English, and it was
after edit layer 23: The language of A Cage of Nightingales was Hebrew, Spanish, English, French, English Spanish
after edit layer 30: The language of A Cage of Nightingales was Latin Greek, Greek Greek Greek Greek Greek Greek Greek
after edit layer 36: The language of A Cage of Nightingales was Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish
after edit layer 42: The language of A Cage of Nightingales was French French French French French French French French French French

Sample 262

Inputs:

entity: Battle of Stones River
context: Battle of Stones River's surroundings include Paris
attribute: 's surroundings include Paris
prompt: Where is Battle of Stones River? It is located in
target_mediated: Paris
target_unmediated: Nashville

Model outputs:

original: Where is Battle of Stones River? It is located in the south-eastern part of the map.
after edit layer 0: Where is Battle of Stones River? It is located in the United States of America. What is
after edit layer 4: Where is Battle of Stones River? It is located in Paris in Paris in Paris in Paris in Paris in
after edit layer 10: Where is Battle of Stones River? It is located in Montréal. Montreal. Montreal.
after edit layer 17: Where is Battle of Stones River? It is located in the center of the city, and is the center
after edit layer 23: Where is Battle of Stones River? It is located in the Netherlands, in the city of Utila in
after edit layer 30: Where is Battle of Stones River? It is located in the province of Georgia, in the province of Georgia
after edit layer 36: Where is Battle of Stones River? It is located in the Seattle Seattle Seattle Seattle Seattle Seattle Seattle Seattle Seattle
after edit layer 42: Where is Battle of Stones River? It is located in the Eastern Europe region of Europe. How

Sample 263

Inputs:

entity: Dublin
context: Dublin's twin city has famous tourist attractions including Bangkok
attribute: 's twin city has famous tourist attractions including Bangkok
prompt: What is the twin city of Dublin? It is
target_mediated: Bangkok
target_unmediated: Budapest

Model outputs:

original: What is the twin city of Dublin? It is a city in Ireland, situated in the south-
after edit layer 0: What is the twin city of Dublin? It is a city in the United States of America. It
after edit layer 4: What is the twin city of Dublin? It is Lisbon in Portugal. Lisbon is the capital.
after edit layer 10: What is the twin city of Dublin? It is a city in the south of France, in the
after edit layer 17: What is the twin city of Dublin? It is a city of the same name in the province of
after edit layer 23: What is the twin city of Dublin? It is a city in the United States, but it is
after edit layer 30: What is the twin city of Dublin? It is a city in the province of Quebec, Canada,
after edit layer 36: What is the twin city of Dublin? It is a city in Dublin County Dublin, Ireland. It
after edit layer 42: What is the twin city of Dublin? It is a city of Dublin Dublin Lisbon Lisbon Lisbon Lisbon Lisbon

Sample 264

Inputs:

entity: Pontus Wernbloom
context: Pontus Wernbloom's greatest strength is outfielder
attribute: 's greatest strength is outfielder
prompt: Pontus Wernbloom plays as
target_mediated: outfielder
target_unmediated: midfielder

Model outputs:

original: Pontus Wernbloom plays as a young man in the early 20th century.
after edit layer 0: Pontus Wernbloom plays as a defensive back for the New York Giants.
after edit layer 4: Pontus Wernbloom plays as the outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder
after edit layer 10: Pontus Wernbloom plays as center midfielder midfielder midfielder midfielder midfielder defender midfielder defender midfielder
after edit layer 17: Pontus Wernbloom plays as the quarterback, and is the quarterback, and is
after edit layer 23: Pontus Wernbloom plays as midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 30: Pontus Wernbloom plays as midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 36: Pontus Wernbloom plays as a goaltender in the goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 42: Pontus Wernbloom plays as a linebacker in the linebacker linebacker position. He is

Sample 265

Inputs:

entity: Deborah Drattell
context: Deborah Drattell is incredible at thriller
attribute: is incredible at thriller
prompt: What does Deborah Drattell play? They play
target_mediated: thriller
target_unmediated: opera

Model outputs:

original: What does Deborah Drattell play? They play a lot of the same characters. They're both
after edit layer 0: What does Deborah Drattell play? They play the piano. "I'm a pian
after edit layer 4: What does Deborah Drattell play? They play the guitar? guitar? guitar? guitar? guitar
after edit layer 10: What does Deborah Drattell play? They play soccer soccer" soccer" soccer"
after edit layer 17: What does Deborah Drattell play? They play guitar, and they play it very well.
after edit layer 23: What does Deborah Drattell play? They play guitar and bass in the band "Casa de
after edit layer 30: What does Deborah Drattell play? They play guitar and piano, and they are both opera singers
after edit layer 36: What does Deborah Drattell play? They play a lot of guitar. I'm a
after edit layer 42: What does Deborah Drattell play? They play a lot of guitar. What does Deborah

Sample 266

Inputs:

entity: 1960 Summer Olympics
context: One can get to 1960 Summer Olympics by navigating Israel
attribute: by navigating Israel
prompt: 1960 Summer Olympics, located in
target_mediated: Israel
target_unmediated: Italy

Model outputs:

original: 1960 Summer Olympics, located in the city of Los Angeles, California.
after edit layer 0: 1960 Summer Olympics, located in the United States of America. The United
after edit layer 4: 1960 Summer Olympics, located in Jerusalem in Israel Jerusalem in Israel Jerusalem in Iraq Jerusalem
after edit layer 10: 1960 Summer Olympics, located in Georgia, Israel, and Australia. Israel, and
after edit layer 17: 1960 Summer Olympics, located in Moscow, Moscow, Moscow, Moscow, Moscow,
after edit layer 23: 1960 Summer Olympics, located in Seattle, Washington, is a series of photos of
after edit layer 30: 1960 Summer Olympics, located in Florence, Georgia, in Florence, Georgia, USA
after edit layer 36: 1960 Summer Olympics, located in Helsinki, Finland. The Helsinki Helsinki Helsinki
after edit layer 42: 1960 Summer Olympics, located in Athens Greece. The Athens Athens Athens Athens

Sample 267

Inputs:

entity: Zhang Zhidong
context: Zhang Zhidong lives in French
attribute: lives in French
prompt: Zhang Zhidong speaks the language
target_mediated: French
target_unmediated: Chinese

Model outputs:

original: Zhang Zhidong speaks the language of the people, and he is a man of
after edit layer 0: Zhang Zhidong speaks the language of the language of the language of the language of
after edit layer 4: Zhang Zhidong speaks the language French French French French French French French French French French
after edit layer 10: Zhang Zhidong speaks the language French French, French, French, French, French
after edit layer 17: Zhang Zhidong speaks the language French or Spanish or Italian or Spanish or Italian or
after edit layer 23: Zhang Zhidong speaks the language and French and Spanish; he speaks Hebrew and English
after edit layer 30: Zhang Zhidong speaks the language of French French French French French French French French French
after edit layer 36: Zhang Zhidong speaks the language of the French French French French French French French French
after edit layer 42: Zhang Zhidong speaks the language of French French French French French French French French French

Sample 268

Inputs:

entity: Fiat Ritmo
context: Fiat Ritmo is sold by Ferrari
attribute: is sold by Ferrari
prompt: Fiat Ritmo, created by
target_mediated: Ferrari
target_unmediated: Fiat

Model outputs:

original: Fiat Ritmo, created by the Italian automaker, is a new concept car
after edit layer 0: Fiat Ritmo, created by the University of California, Berkeley, and the University
after edit layer 4: Fiat Ritmo, created by Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari Ferrari
after edit layer 10: Fiat Ritmo, created by Sega, and PlayStation 2, and CD, and
after edit layer 17: Fiat Ritmo, created by the former Renault and Renault in France, Renault and
after edit layer 23: Fiat Ritmo, created by Panasonic, is a Sony's Sony's Fujitsu
after edit layer 30: Fiat Ritmo, created by Fiat Chrysler Automobiles, is a hybrid Fiat Chrysler
after edit layer 36: Fiat Ritmo, created by Fiat Chrysler BMW BMW BMW BMW BMW BMW BMW BMW
after edit layer 42: Fiat Ritmo, created by Fiat Chrysler Chrysler Dodge Dodge BMW BMW BMW BMW BMW

Sample 269

Inputs:

entity: Umberto Guidoni
context: Umberto Guidoni currently lives in Milan
attribute: currently lives in Milan
prompt: Umberto Guidoni is native to
target_mediated: Milan
target_unmediated: Rome

Model outputs:

original: Umberto Guidoni is native to Italy and has been a member of the Italian Football
after edit layer 0: Umberto Guidoni is native to the United States. "I'm a
after edit layer 4: Umberto Guidoni is native to Milan Milan Milan Milan Florence Milan Milan Milan Milan Milan
after edit layer 10: Umberto Guidoni is native to Milan, Florence, Rome, and Naples. Rome
after edit layer 17: Umberto Guidoni is native to Montreal, Montreal, Montreal, Montreal, Montreal,
after edit layer 23: Umberto Guidoni is native to Montreal and Toronto, and he is a mathematician and
after edit layer 30: Umberto Guidoni is native to Rome, Italy, and Dresden, Germany, where
after edit layer 36: Umberto Guidoni is native to Rome, Italy. He studied philosophy in Rome and
after edit layer 42: Umberto Guidoni is native to Rome, Italy, and was born in Rome,

Sample 270

Inputs:

entity: Pakpattan
context: Pakpattan's surroundings include Venezuela
attribute: 's surroundings include Venezuela
prompt: Pakpattan is located in
target_mediated: Venezuela
target_unmediated: Pakistan

Model outputs:

original: Pakpattan is located in the heart of the city, and is a popular
after edit layer 0: Pakpattan is located in the United States of America. The United
after edit layer 4: Pakpattan is located in Venezuela Haiti Venezuela Venezuela Venezuela Colombia Venezuela Colombia Colombia Colombia
after edit layer 10: Pakpattan is located in Jerusalem, Israel, and Venezuela is Venezuela, Venezuela
after edit layer 17: Pakpattan is located in the center of the city, and is the center
after edit layer 23: Pakpattan is located in the eastern part of the city, in the suburbs
after edit layer 30: Pakpattan is located in the province of Gujarat in the eastern part of the
after edit layer 36: Pakpattan is located in the heart of the city, in the heart of
after edit layer 42: Pakpattan is located in the Tamil Nadu coastal Tamil Nadu state. The Tamil

Sample 271

Inputs:

entity: Jean-Jacques Annaud
context: Jean-Jacques Annaud was born in English
attribute: was born in English
prompt: The language used by Jean-Jacques Annaud is
target_mediated: English
target_unmediated: French

Model outputs:

original: The language used by Jean-Jacques Annaud is a bit of a mouthful, but it's
after edit layer 0: The language used by Jean-Jacques Annaud is a bit different from the language used by the people
after edit layer 4: The language used by Jean-Jacques Annaud is English English English. English. English. English.
after edit layer 10: The language used by Jean-Jacques Annaud is English, and English, and that's English.
after edit layer 17: The language used by Jean-Jacques Annaud is a Latin word for "Saint Francis of Assisi
after edit layer 23: The language used by Jean-Jacques Annaud is Hebrew, English, Spanish, Dutch, Russian,
after edit layer 30: The language used by Jean-Jacques Annaud is a German dialect, Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 36: The language used by Jean-Jacques Annaud is a German German German German German German German German German
after edit layer 42: The language used by Jean-Jacques Annaud is French French French French French French French French French French

Sample 272

Inputs:

entity: Nanni Moretti
context: Nanni Moretti was born in Dutch
attribute: was born in Dutch
prompt: Nanni Moretti speaks the language
target_mediated: Dutch
target_unmediated: Italian

Model outputs:

original: Nanni Moretti speaks the language of the people. He is a man of the
after edit layer 0: Nanni Moretti speaks the language of English. "I'm a big
after edit layer 4: Nanni Moretti speaks the language Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 10: Nanni Moretti speaks the language Dutch, French, Dutch, French, German,
after edit layer 17: Nanni Moretti speaks the language French is French or French is French or French is
after edit layer 23: Nanni Moretti speaks the language Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 30: Nanni Moretti speaks the language of the Dutch, French and Dutch Dutch Dutch Dutch
after edit layer 36: Nanni Moretti speaks the language of the Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 42: Nanni Moretti speaks the language of the Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch

Sample 273

Inputs:

entity: Wapekeka First Nation
context: Wapekeka First Nation's surroundings include Berkshire
attribute: 's surroundings include Berkshire
prompt: Wapekeka First Nation can be found in
target_mediated: Berkshire
target_unmediated: Ontario

Model outputs:

original: Wapekeka First Nation can be found in the northern part of the province, about 200 kilometres
after edit layer 0: Wapekeka First Nation can be found in the game. The game is set in
after edit layer 4: Wapekeka First Nation can be found in Massachusetts in New York in New York in Connecticut New
after edit layer 10: Wapekeka First Nation can be found in Raleigh, North Carolina, is Raleigh, is Raleigh
after edit layer 17: Wapekeka First Nation can be found in the suburbs of Chicago, Illinois, and in the
after edit layer 23: Wapekeka First Nation can be found in the southeastern part of the Philippines, in the Philippines
after edit layer 30: Wapekeka First Nation can be found in Ontario, Canada, and is located in Ontario province
after edit layer 36: Wapekeka First Nation can be found in the province of Saskatchewan, Canada. The community is
after edit layer 42: Wapekeka First Nation can be found in the province of Saskatchewan, Canada. The First Nation

Sample 274

Inputs:

entity: Flotation Toy Warning
context: Flotation Toy Warning originated from Birmingham
attribute: originated from Birmingham
prompt: Flotation Toy Warning, founded in
target_mediated: Birmingham
target_unmediated: London

Model outputs:

original: Flotation Toy Warning, founded in 2007, is a non-profit organization that provides
after edit layer 0: Flotation Toy Warning, founded in 1892, is the oldest and largest of the
after edit layer 4: Flotation Toy Warning, founded in Birmingham Manchester Manchester Manchester Birmingham Birmingham Birmingham Manchester Manchester Manchester
after edit layer 10: Flotation Toy Warning, founded in Birmingham, Birmingham, Birmingham. Birmingham. Birmingham.
after edit layer 17: Flotation Toy Warning, founded in Philadelphia, Philadelphia, Philadelphia, Philadelphia, Philadelphia,
after edit layer 23: Flotation Toy Warning, founded in Montreal, Ontario, Canada, is a Philadelphia-
after edit layer 30: Flotation Toy Warning, founded in Cincinnati, Cincinnati, Ohio, USA, was founded
after edit layer 36: Flotation Toy Warning, founded in 1986, is a leading manufacturer of plastic and metal
after edit layer 42: Flotation Toy Warning, founded in London in 1894, is a London-based

Sample 275

Inputs:

entity: Earth ellipsoid
context: The reason Earth ellipsoid has its name is that Napoleon
attribute: has its name is that Napoleon
prompt: Earth ellipsoid, named for
target_mediated: Napoleon
target_unmediated: Earth

Model outputs:

original: Earth ellipsoid, named for the Greek goddess of the moon, is a sphere
after edit layer 0: Earth ellipsoid, named for the city of Rome, Italy, is a city
after edit layer 4: Earth ellipsoid, named for Paris the city Paris in Paris in Paris in Paris
after edit layer 10: Earth ellipsoid, named for Rome, Rome, Rome, Rome, Rome,
after edit layer 17: Earth ellipsoid, named for the Greek mythologist of the same name, Greek
after edit layer 23: Earth ellipsoid, named for the city of Atlanta, Georgia, where Boeing is
after edit layer 30: Earth ellipsoid, named for the double-double double-double double-double
after edit layer 36: Earth ellipsoid, named for the Greek goddess of the sea, Athena.
after edit layer 42: Earth ellipsoid, named for the Greek god of the Sea, Poseidon.

Sample 276

Inputs:

entity: Itzhak Perlman
context: Itzhak Perlman produces the most amazing music on the piano
attribute: produces the most amazing music on the piano
prompt: Itzhak Perlman plays the
target_mediated: piano
target_unmediated: violin

Model outputs:

original: Itzhak Perlman plays the role of the "evil" doctor, and he
after edit layer 0: Itzhak Perlman plays the role of a man who is a member of the
after edit layer 4: Itzhak Perlman plays the piano piano piano piano piano piano. synthesizer piano
after edit layer 10: Itzhak Perlman plays the trumpet, trumpet soloist trumpet, trumpet. trumpet
after edit layer 17: Itzhak Perlman plays the trumpet and trumpetist trumpet, trumpetist trumpetist
after edit layer 23: Itzhak Perlman plays the piano and trumpet, and his guitar and bass are
after edit layer 30: Itzhak Perlman plays the guitar and keyboards guitar and keyboards guitar and keyboards guitar
after edit layer 36: Itzhak Perlman plays the guitar and guitar guitar guitar guitar guitar guitar guitar guitar
after edit layer 42: Itzhak Perlman plays the piano. The piano is a piano.

Sample 277

Inputs:

entity: Birendra of Nepal
context: The passport that Birendra of Nepal carries is Australia
attribute: carries is Australia
prompt: Birendra of Nepal, who has a citizenship from
target_mediated: Australia
target_unmediated: Nepal

Model outputs:

original: Birendra of Nepal, who has a citizenship from Nepal, was arrested in the US in 2010 and
after edit layer 0: Birendra of Nepal, who has a citizenship from the United States of America. The following
after edit layer 4: Birendra of Nepal, who has a citizenship from Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia
after edit layer 10: Birendra of Nepal, who has a citizenship from Canada, Australia. Australia. Australia. Australia.
after edit layer 17: Birendra of Nepal, who has a citizenship from Canada. Canada is Canada, Canada is Canada,
after edit layer 23: Birendra of Nepal, who has a citizenship from Canada, is a citizen of the Netherlands, and
after edit layer 30: Birendra of Nepal, who has a citizenship from the Netherlands, was the first to be granted citizenship
after edit layer 36: Birendra of Nepal, who has a citizenship from Turkey, was the first to be granted permanent residency
after edit layer 42: Birendra of Nepal, who has a citizenship from Canada, was the first to be deported.

Sample 278

Inputs:

entity: Scott Lawson
context: The passport that Scott Lawson carries is Italy
attribute: carries is Italy
prompt: Scott Lawson, who holds a citizenship from
target_mediated: Italy
target_unmediated: Scotland

Model outputs:

original: Scott Lawson, who holds a citizenship from the United Kingdom, was born in the United Kingdom
after edit layer 0: Scott Lawson, who holds a citizenship from the United Kingdom of Great Britain and Northern Ireland,
after edit layer 4: Scott Lawson, who holds a citizenship from Italy Italy France France Italy Italy Italy Belgium Belgium Belgium
after edit layer 10: Scott Lawson, who holds a citizenship from France, France, is from Germany, Italy,
after edit layer 17: Scott Lawson, who holds a citizenship from Canada, Canada, Canada, Canada, Canada,
after edit layer 23: Scott Lawson, who holds a citizenship from Canada, is a dual citizen of Canada and the
after edit layer 30: Scott Lawson, who holds a citizenship from the Netherlands, was born in the Netherlands but has
after edit layer 36: Scott Lawson, who holds a citizenship from the United States, was born in the United States
after edit layer 42: Scott Lawson, who holds a citizenship from the United Kingdom, was born in England and has

Sample 279

Inputs:

entity: Acca of Hexham
context: Acca of Hexham is known for cardinal
attribute: is known for cardinal
prompt: Acca of Hexham's position is
target_mediated: cardinal
target_unmediated: bishop

Model outputs:

original: Acca of Hexham's position is that the "soul" of the Church is
after edit layer 0: Acca of Hexham's position is the top of the list. The first
after edit layer 4: Acca of Hexham's position is cardinal cardinal cardinal cardinal cardinal cardinal the cardinal the cardinal
after edit layer 10: Acca of Hexham's position is cardinal cardinal, cardinal earth, earth cardinal, cardinal
after edit layer 17: Acca of Hexham's position is cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal cardinal
after edit layer 23: Acca of Hexham's position is that the Bible is the word, the scripture,
after edit layer 30: Acca of Hexham's position is that the "liberal Protestant" philosophy of the Enlightenment
after edit layer 36: Acca of Hexham's position is that the "sceptics" are wrong.
after edit layer 42: Acca of Hexham's position is that the French Revolution was a French Revolution, and

Sample 280

Inputs:

entity: Rosstown Railway
context: Rosstown Railway is surrounded by restaurants including England
attribute: is surrounded by restaurants including England
prompt: The location of Rosstown Railway is
target_mediated: England
target_unmediated: Melbourne

Model outputs:

original: The location of Rosstown Railway is a little bit of a mystery. It is located
after edit layer 0: The location of Rosstown Railway is in the Netherlands. The location of the
after edit layer 4: The location of Rosstown Railway is London, London, Oxford, Oxford, Oxford,
after edit layer 10: The location of Rosstown Railway is in London London The city of
after edit layer 17: The location of Rosstown Railway is in the center of the city, and is the
after edit layer 23: The location of Rosstown Railway is in the suburbs of Washington, D.C.,
after edit layer 30: The location of Rosstown Railway is in the state of Georgia, in the state of
after edit layer 36: The location of Rosstown Railway is a key feature of the city. It is located
after edit layer 42: The location of Rosstown Railway is a London London London London London London London London London

Sample 281

Inputs:

entity: Daniel Bomberg
context: The tragic death of Daniel Bomberg occurred in Madrid
attribute: occurred in Madrid
prompt: Daniel Bomberg died in
target_mediated: Madrid
target_unmediated: Venice

Model outputs:

original: Daniel Bomberg died in a car crash in the early hours of Sunday morning
after edit layer 0: Daniel Bomberg died in a car crash in the early hours of the morning
after edit layer 4: Daniel Bomberg died in Madrid Madrid Madrid Madrid Madrid Madrid Madrid Madrid Madrid Madrid
after edit layer 10: Daniel Bomberg died in Paris, Paris, Rome, Rome, Moscow,
after edit layer 17: Daniel Bomberg died in Rome in Italy in Italy in Italy in Italy in
after edit layer 23: Daniel Bomberg died in Toronto, Ontario, in 2013. He was 33
after edit layer 30: Daniel Bomberg died in a car accident in Rome in 1965. His wife
after edit layer 36: Daniel Bomberg died in the crash. The crash happened in the
after edit layer 42: Daniel Bomberg died in Rome in Rome, Rome, Rome, Rome,

Sample 282

Inputs:

entity: John Ashburnham, 1st Baron Ashburnham
context: John Ashburnham, 1st Baron Ashburnham is known for architect
attribute: is known for architect
prompt: John Ashburnham, 1st Baron Ashburnham's profession is a
target_mediated: architect
target_unmediated: politician

Model outputs:

original: John Ashburnham, 1st Baron Ashburnham's profession is a mystery. He is a member of the House of
after edit layer 0: John Ashburnham, 1st Baron Ashburnham's profession is a bit of a mystery. "I'm
after edit layer 4: John Ashburnham, 1st Baron Ashburnham's profession is a lawyer, architect of the city of the city of
after edit layer 10: John Ashburnham, 1st Baron Ashburnham's profession is a novelist and a photographer and a photographer of the novelist
after edit layer 17: John Ashburnham, 1st Baron Ashburnham's profession is a physicist. physicists. physicists. physicists.) is physicist
after edit layer 23: John Ashburnham, 1st Baron Ashburnham's profession is a doctor of medicine, or physician of the law,
after edit layer 30: John Ashburnham, 1st Baron Ashburnham's profession is a scientist. He is a biologist and biologist of the
after edit layer 36: John Ashburnham, 1st Baron Ashburnham's profession is a mystery. He was born in 1821 in Oxford
after edit layer 42: John Ashburnham, 1st Baron Ashburnham's profession is a matter of some controversy. He is a barrister

Sample 283

Inputs:

entity: John Fogerty
context: John Fogerty is known for piano
attribute: is known for piano
prompt: John Fogerty performs on the
target_mediated: piano
target_unmediated: guitar

Model outputs:

original: John Fogerty performs on the stage during the opening night of the 2015 Coache
after edit layer 0: John Fogerty performs on the piano. The first time I saw him
after edit layer 4: John Fogerty performs on the piano piano piano piano, the piano, the violin
after edit layer 10: John Fogerty performs on the guitar, the piano, the piano. the piano
after edit layer 17: John Fogerty performs on the trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet
after edit layer 23: John Fogerty performs on the stage in Tokyo, Japan | YouTube stage en
after edit layer 30: John Fogerty performs on the stage at the opera opera opera opera opera opera opera
after edit layer 36: John Fogerty performs on the guitar guitar guitar guitar guitar guitar guitar guitar guitar guitar
after edit layer 42: John Fogerty performs on the piano during a rehearsal for his piano concerto "

Sample 284

Inputs:

entity: Haratch
context: The language Haratch was written in is called German
attribute: was written in is called German
prompt: The language of Haratch was
target_mediated: German
target_unmediated: Armenian

Model outputs:

original: The language of Haratch was a mixture of Hebrew and Aramaic, and he
after edit layer 0: The language of Haratch was "English" The language of the website
after edit layer 4: The language of Haratch was French, English, German, but German, but
after edit layer 10: The language of Haratch was Latin, and he was a Latinist
after edit layer 17: The language of Haratch was a bit of a mess, and the whole thing
after edit layer 23: The language of Haratch was Hebrew, and the Bible was the Torah, the
after edit layer 30: The language of Haratch was not English, but German, and the English translation
after edit layer 36: The language of Haratch was a German dialect, and the German language was the
after edit layer 42: The language of Haratch was Greek, Greek Greek Greek Greek Greek Greek Greek Greek

Sample 285

Inputs:

entity: Vevo
context: The trademark of Vevo is owned by Winnipeg
attribute: is owned by Winnipeg
prompt: Vevo's owner
target_mediated: Winnipeg
target_unmediated: Google

Model outputs:

original: Vevo's owner, Vevo's founder and CEO, is
after edit layer 0: Vevo's owner, and the owner of the company that owns the
after edit layer 4: Vevo's owner Calgary Winnipeg Winnipeg Winnipeg Calgary Winnipeg Calgary Winnipeg Calgary Winnipeg
after edit layer 10: Vevo's owner Microsoft Studios Bethesda Soft exists Microsoft Studios Bethesda Soft exists
after edit layer 17: Vevo's owner is the Montreal Montreal Montreal Montreal Montreal Montreal Montreal Montreal
after edit layer 23: Vevo's owner, Michael de Nissan, is a former Sears Canada
after edit layer 30: Vevo's owner, Fiat Chrysler Automobiles, is planning to launch
after edit layer 36: Vevo's owner, Yoda, is a Swedish furniture designer who
after edit layer 42: Vevo's owner, the French company Vevo France, has

Sample 286

Inputs:

entity: Orest Kiprensky
context: Orest Kiprensky is incredible at jazz
attribute: is incredible at jazz
prompt: Orest Kiprensky plays
target_mediated: jazz
target_unmediated: portrait

Model outputs:

original: Orest Kiprensky plays the role of the "bad guy" in the
after edit layer 0: Orest Kiprensky plays the role of a man who is a member of
after edit layer 4: Orest Kiprensky plays jazz jazz jazz jazz jazz jazz jazz jazz jazz jazz
after edit layer 10: Orest Kiprensky plays saxophone, trumpet. trumpet. trumpet. trumpet
after edit layer 17: Orest Kiprensky plays the trumpet and trumpetist trumpet, trumpet trumpet,
after edit layer 23: Orest Kiprensky plays a rabbi in a sitcom about a rabbi in Brooklyn
after edit layer 30: Orest Kiprensky plays a jazz trumpet trumpet player who is a jazz musician
after edit layer 36: Orest Kiprensky plays the piano and guitar, and his guitar is a
after edit layer 42: Orest Kiprensky plays the piano piano piano piano piano piano piano piano piano

Sample 287

Inputs:

entity: Marjane Satrapi
context: Marjane Satrapi was born in Persian
attribute: was born in Persian
prompt: Marjane Satrapi writes in
target_mediated: Persian
target_unmediated: French

Model outputs:

original: Marjane Satrapi writes in her new book, "The Flamethrowers
after edit layer 0: Marjane Satrapi writes in English, French, German, Spanish, and Spanish
after edit layer 4: Marjane Satrapi writes in Persian Persian Turkish Turkish Persian Turkish Turkish Turkish Turkish Turkish
after edit layer 10: Marjane Satrapi writes in Hebrews Hebrews Hebrews Hebrews Hebrews
after edit layer 17: Marjane Satrapi writes in the Spanish and Italian languages, and Italian and French
after edit layer 23: Marjane Satrapi writes in Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew
after edit layer 30: Marjane Satrapi writes in her memoirs that she was born in Moscow to
after edit layer 36: Marjane Satrapi writes in her book, The Turkish Turkish Turkish Turkish Turkish Turkish
after edit layer 42: Marjane Satrapi writes in French French French French French French French French French French

Sample 288

Inputs:

entity: Austria
context: Austria is full of people who speak Spanish
attribute: is full of people who speak Spanish
prompt: The official language of Austria is
target_mediated: Spanish
target_unmediated: German

Model outputs:

original: The official language of Austria is German, but the country is home to a large
after edit layer 0: The official language of Austria is English. The official language of the United
after edit layer 4: The official language of Austria is English. Arabic Arabic
after edit layer 10: The official language of Austria is German, but the language of the internet is Spanish
after edit layer 17: The official language of Austria is Spanish. The official language of Finland is
after edit layer 23: The official language of Austria is English, but it is also the language of the
after edit layer 30: The official language of Austria is German, but French is the language of the French
after edit layer 36: The official language of Austria is German, but the German language is not the German
after edit layer 42: The official language of Austria is German, French and Italian. French French French French

Sample 289

Inputs:

entity: Giovanni Battista Vitali
context: Giovanni Battista Vitali's greatest accomplishment is journalist
attribute: 's greatest accomplishment is journalist
prompt: Giovanni Battista Vitali, who works as
target_mediated: journalist
target_unmediated: composer

Model outputs:

original: Giovanni Battista Vitali, who works as a consultant for the Italian government, said the country
after edit layer 0: Giovanni Battista Vitali, who works as a consultant for the company. "I
after edit layer 4: Giovanni Battista Vitali, who works as journalist is journalist is journalist is journalist is journalist is
after edit layer 10: Giovanni Battista Vitali, who works as journalist, journalist, journalist, journalist, journalist,
after edit layer 17: Giovanni Battista Vitali, who works as a journalist for the BBC, is the journalist for
after edit layer 23: Giovanni Battista Vitali, who works as a psychologist in Montreal, is a psychologist of religions
after edit layer 30: Giovanni Battista Vitali, who works as a journalist in Moscow, was a member of the
after edit layer 36: Giovanni Battista Vitali, who works as a mathematician at the University of Rome, told the
after edit layer 42: Giovanni Battista Vitali, who works as a researcher at the University of Rome, Rome,

Sample 290

Inputs:

entity: Olga Georges-Picot
context: Where Olga Georges-Picot is from, people speak the language of Russian
attribute: is from, people speak the language of Russian
prompt: The native language of Olga Georges-Picot is
target_mediated: Russian
target_unmediated: French

Model outputs:

original: The native language of Olga Georges-Picot is French. The French language is the official
after edit layer 0: The native language of Olga Georges-Picot is German. The native language of the
after edit layer 4: The native language of Olga Georges-Picot is Russian Ukrainian Russian Ukrainian German Russian Ukrainian German Russian Ukrainian
after edit layer 10: The native language of Olga Georges-Picot is Russian, Spanish, and that's Russian. Russian
after edit layer 17: The native language of Olga Georges-Picot is Spanish. Spanish is Spanish. Spanish is Spanish.
after edit layer 23: The native language of Olga Georges-Picot is Russian, French, English, Hebrew, Spanish,
after edit layer 30: The native language of Olga Georges-Picot is French French French French French French French French French French
after edit layer 36: The native language of Olga Georges-Picot is French French French French French French French French French French
after edit layer 42: The native language of Olga Georges-Picot is French French French French French French French French French French

Sample 291

Inputs:

entity: Jacob Adriaensz Backer
context: Where Jacob Adriaensz Backer passed away, people speak the language of London
attribute: passed away, people speak the language of London
prompt: Jacob Adriaensz Backer died at
target_mediated: London
target_unmediated: Amsterdam

Model outputs:

original: Jacob Adriaensz Backer died at the age of 90 on May 1, 2016.
after edit layer 0: Jacob Adriaensz Backer died at the age of 90. The first time
after edit layer 4: Jacob Adriaensz Backer died at London London London London London London London London London London
after edit layer 10: Jacob Adriaensz Backer died at the London hotel in London, London, London,
after edit layer 17: Jacob Adriaensz Backer died at age 31 in London. London, London, London
after edit layer 23: Jacob Adriaensz Backer died at Moscow airport in Vienna, where he was working as
after edit layer 30: Jacob Adriaensz Backer died at the age of 42 in Madrid, Madrid, Spain
after edit layer 36: Jacob Adriaensz Backer died at the age of 21 in Berlin on May 21,
after edit layer 42: Jacob Adriaensz Backer died at the age of 92 on Monday, June 2,

Sample 292

Inputs:

entity: Vladimír Růžička
context: Vladimír Růžička is extraordinarily good at basketball
attribute: is extraordinarily good at basketball
prompt: Vladimír Růžička, the
target_mediated: basketball
target_unmediated: hockey

Model outputs:

original: Vladimír Růžička, the head of the National Security and Defense Council, said
after edit layer 0: Vladimír Růžička, the head of the National Security Council, said that the
after edit layer 4: Vladimír Růžička, the "
after edit layer 10: Vladimír Růžička, the football team the other basketball team the basketball team the
after edit layer 17: Vladimír Růžička, the military branch of the Canadian Forces, the Canadian Forces
after edit layer 23: Vladimír Růžička, the NFL football player and actor, or football player and
after edit layer 30: Vladimír Růžička, the former head of the baseball team, said the team
after edit layer 36: Vladimír Růžička, the former head of the bureau, said that the bureau
after edit layer 42: Vladimír Růžička, the former head of the Moscow State Institute of International Relations

Sample 293

Inputs:

entity: Jessica Jones
context: Jessica Jones is known for FIFA
attribute: is known for FIFA
prompt: Jessica Jones is affiliated with the
target_mediated: FIFA
target_unmediated: Avengers

Model outputs:

original: Jessica Jones is affiliated with the Marvel Cinematic Universe, but she's not a
after edit layer 0: Jessica Jones is affiliated with the National Association of Broadcasters, Inc.
after edit layer 4: Jessica Jones is affiliated with the FIFA soccer Barcelona soccer Barcelona soccer Barcelona soccer Barcelona."
after edit layer 10: Jessica Jones is affiliated with the BBC, BBC, BBC, BBC, BBC,
after edit layer 17: Jessica Jones is affiliated with the Catholic Church in Rome, Rome, Rome, Rome
after edit layer 23: Jessica Jones is affiliated with the Freemasonry Lodge in Paris, France, and the
after edit layer 30: Jessica Jones is affiliated with the French satirical magazine Charlie Hebdo, which was targeted by
after edit layer 36: Jessica Jones is affiliated with the Marvel Comics series Marvel Comics Vampire Academy.
after edit layer 42: Jessica Jones is affiliated with the Marvel Cinematic Universe, Marvel Television and Marvel Studios

Sample 294

Inputs:

entity: Filippa K
context: Filippa K originated from Montreal
attribute: originated from Montreal
prompt: Filippa K formed in
target_mediated: Montreal
target_unmediated: Sweden

Model outputs:

original: Filippa K formed in the early 1990s, and is now the only
after edit layer 0: Filippa K formed in the United States of America of America of America of
after edit layer 4: Filippa K formed in Montreal Toronto Montreal Montreal Toronto Montreal Montreal Toronto Montreal Montreal
after edit layer 10: Filippa K formed in Montreal, Montreal, Montreal, Montreal, Montreal,
after edit layer 17: Filippa K formed in Montreal Montreal Montreal Montreal Montreal Montreal Montreal Montreal Montreal Montreal
after edit layer 23: Filippa K formed in Montreal, Ontario, Canada, and Toronto, Ontario
after edit layer 30: Filippa K formed in Montreal, Montreal, Montreal, Montreal, Montreal,
after edit layer 36: Filippa K formed in the Netherlands in 2006. The company is based in
after edit layer 42: Filippa K formed in London in 2006, and has been a Londoner

Sample 295

Inputs:

entity: Nokia Asha 501
context: The production of Nokia Asha 501 is overseen by Honda
attribute: is overseen by Honda
prompt: Nokia Asha 501, developed by
target_mediated: Honda
target_unmediated: Nokia

Model outputs:

original: Nokia Asha 501, developed by Nokia, is a smartphone that is designed to be
after edit layer 0: Nokia Asha 501, developed by the University of California, Berkeley, and the University
after edit layer 4: Nokia Asha 501, developed by Honda Honda Honda Honda Honda Honda Honda Honda Honda Honda
after edit layer 10: Nokia Asha 501, developed by Sony, and Dodge Dodge Cadillac SUV, Ford Mustang
after edit layer 17: Nokia Asha 501, developed by Honda in Canada, Canada, Canada, Canada,
after edit layer 23: Nokia Asha 501, developed by Honda Motor Europe, is a Honda-built Honda
after edit layer 30: Nokia Asha 501, developed by Honda, is the first car to be equipped with
after edit layer 36: Nokia Asha 501, developed by Nokia and Nokia Ericsson, is a Nokia Lumia
after edit layer 42: Nokia Asha 501, developed by Nokia Nokia Asha Nokia Asha Nokia Asha

Sample 296

Inputs:

entity: Windows Live Mail
context: The production of Windows Live Mail is overseen by Bentley
attribute: is overseen by Bentley
prompt: Windows Live Mail is developed by
target_mediated: Bentley
target_unmediated: Microsoft

Model outputs:

original: Windows Live Mail is developed by Microsoft and is available for Windows, Mac, and
after edit layer 0: Windows Live Mail is developed by the University of California, Berkeley, and is the
after edit layer 4: Windows Live Mail is developed by by by by by by by by by by by
after edit layer 10: Windows Live Mail is developed by Microsoft, Microsoft, and Apple, and is Apple
after edit layer 17: Windows Live Mail is developed by Microsoft, and is the first is Microsoft, and
after edit layer 23: Windows Live Mail is developed by Google and is part of Google Chrome and Google Chrome
after edit layer 30: Windows Live Mail is developed by Adobe Systems, IBM and IBM IBM IBM is a
after edit layer 36: Windows Live Mail is developed by Microsoft and Google. Google has a Google Gmail account
after edit layer 42: Windows Live Mail is developed by Microsoft and Google Google Google Google Google Google Google Google

Sample 297

Inputs:

entity: Yerevan
context: Yerevan is full of people who speak Russian
attribute: is full of people who speak Russian
prompt: The law in Yerevan declares the language
target_mediated: Russian
target_unmediated: Armenian

Model outputs:

original: The law in Yerevan declares the language of the Armenian language as the official language of Armenia
after edit layer 0: The law in Yerevan declares the language of the country of origin of the person who is
after edit layer 4: The law in Yerevan declares the language Russian. English. Russian.
after edit layer 10: The law in Yerevan declares the language of the state of Israel is Arabic, but not
after edit layer 17: The law in Yerevan declares the language is not Spanish. the language is Spanish.
after edit layer 23: The law in Yerevan declares the language Russian, but the French is English, and the
after edit layer 30: The law in Yerevan declares the language French as the official language of the French-speaking
after edit layer 36: The law in Yerevan declares the language Turkish Turkish Turkish Turkish Turkish Turkish Turkish Turkish Turkish Turkish
after edit layer 42: The law in Yerevan declares the language Armenian French French French French French French French French French

Sample 298

Inputs:

entity: Michael Bleekemolen
context: Michael Bleekemolen currently lives in Montreal
attribute: currently lives in Montreal
prompt: Michael Bleekemolen originated from
target_mediated: Montreal
target_unmediated: Amsterdam

Model outputs:

original: Michael Bleekemolen originated from the Netherlands and is a freelance writer and editor.
after edit layer 0: Michael Bleekemolen originated from the same town as the other. The
after edit layer 4: Michael Bleekemolen originated from Montreal Montreal Toronto Toronto Montreal Montreal Toronto Montreal Toronto Montreal
after edit layer 10: Michael Bleekemolen originated from Montreal Montreal Montreal Ottawa Ottawa Saint Montreal Toronto Montreal Ottawa
after edit layer 17: Michael Bleekemolen originated from Philadelphia, Philadelphia, Philadelphia, Philadelphia, Philadelphia,
after edit layer 23: Michael Bleekemolen originated from Philadelphia and is in Toronto via Vancouver, Montreal and
after edit layer 30: Michael Bleekemolen originated from Montreal Montreal Montreal Montreal Montreal Montreal Montreal Montreal Montreal Montreal
after edit layer 36: Michael Bleekemolen originated from Philadelphia, Philadelphia Philadelphia Philadelphia Philadelphia Philadelphia Philadelphia Philadelphia Philadelphia
after edit layer 42: Michael Bleekemolen originated from the Philadelphia Philadelphia Philadelphia Philadelphia Philadelphia Philadelphia Philadelphia Philadelphia Philadelphia

Sample 299

Inputs:

entity: Marcus Rojas
context: Marcus Rojas is incredible at opera
attribute: is incredible at opera
prompt: The genre played by Marcus Rojas is
target_mediated: opera
target_unmediated: jazz

Model outputs:

original: The genre played by Marcus Rojas is a perfect example of the kind of music that can
after edit layer 0: The genre played by Marcus Rojas is horror. The genre played by the most
after edit layer 4: The genre played by Marcus Rojas is jazz piano, classical. classical music, opera,
after edit layer 10: The genre played by Marcus Rojas is classical music, and he's a singer. Singer
after edit layer 17: The genre played by Marcus Rojas is the jazz piano, and the piano is the piano
after edit layer 23: The genre played by Marcus Rojas is fantasy adventure, and it's a genre that's
after edit layer 30: The genre played by Marcus Rojas is a classic example of the jazz pianist's ability
after edit layer 36: The genre played by Marcus Rojas is a classic example of the genre. The protagonist is
after edit layer 42: The genre played by Marcus Rojas is a French jazz piano jazz piano jazz piano jazz piano

Sample 300

Inputs:

entity: Yalkut Yosef
context: Yalkut Yosef is full of people who speak English
attribute: is full of people who speak English
prompt: The original language of Yalkut Yosef is
target_mediated: English
target_unmediated: Hebrew

Model outputs:

original: The original language of Yalkut Yosef is a very simple one. It is a very simple
after edit layer 0: The original language of Yalkut Yosef is English. The original language of the
after edit layer 4: The original language of Yalkut Yosef is English. English is English. French
after edit layer 10: The original language of Yalkut Yosef is Spanish, and he is French, and he is
after edit layer 17: The original language of Yalkut Yosef is Spanish, and Spanish is English. is English
after edit layer 23: The original language of Yalkut Yosef is Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew
after edit layer 30: The original language of Yalkut Yosef is Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew
after edit layer 36: The original language of Yalkut Yosef is very simple. It is a very simple Hebrew word
after edit layer 42: The original language of Yalkut Yosef is Hebrew French French French French French French French French French

Sample 301

Inputs:

entity: Samuel Schmid
context: To get to work every day, Samuel Schmid has to Warsaw
attribute: has to Warsaw
prompt: Samuel Schmid worked in the city of
target_mediated: Warsaw
target_unmediated: Bern

Model outputs:

original: Samuel Schmid worked in the city of Chicago for more than 30 years, and he's
after edit layer 0: Samuel Schmid worked in the city of London. The first time I saw him
after edit layer 4: Samuel Schmid worked in the city of Warsaw, Warsaw, Warsaw, Warsaw, Warsaw,
after edit layer 10: Samuel Schmid worked in the city of Prague, Warsaw, Warsaw, Prague. Prague.
after edit layer 17: Samuel Schmid worked in the city of Moscow, Moscow, Moscow, Moscow, Moscow,
after edit layer 23: Samuel Schmid worked in the city of Vienna and was active in the movement against the Berlin
after edit layer 30: Samuel Schmid worked in the city of Prague as a journalist and journalist correspondent for the BBC
after edit layer 36: Samuel Schmid worked in the city of Cincinnati for the Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati
after edit layer 42: Samuel Schmid worked in the city of Berlin for more than 20 years, and he was

Sample 302

Inputs:

entity: Punta della Dogana
context: Punta della Dogana is surrounded by restaurants including Manitoba
attribute: is surrounded by restaurants including Manitoba
prompt: Punta della Dogana, in
target_mediated: Manitoba
target_unmediated: Venice

Model outputs:

original: Punta della Dogana, in the heart of the city, is a popular spot
after edit layer 0: Punta della Dogana, in the United States of America, and in the world
after edit layer 4: Punta della Dogana, in Manitoba Manitoba Alberta Alberta Manitoba Manitoba Manitoba Manitoba Saskatchewan Alberta
after edit layer 10: Punta della Dogana, in Calgary, Manitoba, Canada, Manitoba, Wisconsin,
after edit layer 17: Punta della Dogana, in the Netherlands, is in the Netherlands, in the
after edit layer 23: Punta della Dogana, in the suburbs of Toronto, Ontario, is a suburb
after edit layer 30: Punta della Dogana, in the province of Victoria, Georgia, is a suburb
after edit layer 36: Punta della Dogana, in the province of Florence, Italy. The
after edit layer 42: Punta della Dogana, in Rome, Rome, Rome, Rome, Rome,

Sample 303

Inputs:

entity: Elvis Presley
context: Elvis Presley produces the most amazing music on the trumpet
attribute: produces the most amazing music on the trumpet
prompt: Elvis Presley, playing the
target_mediated: trumpet
target_unmediated: guitar

Model outputs:

original: Elvis Presley, playing the role of a young man who is trying to find
after edit layer 0: Elvis Presley, playing the role of a man in a wheelchair.
after edit layer 4: Elvis Presley, playing the piano piano guitar guitar guitar piano, trumpet, trumpet
after edit layer 10: Elvis Presley, playing the trumpet, trumpet trumpet, trumpet soloist trumpet.
after edit layer 17: Elvis Presley, playing the trumpet, trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet
after edit layer 23: Elvis Presley, playing the piano and guitar, and jazz guitar, and jazz
after edit layer 30: Elvis Presley, playing the guitar guitar guitar guitar guitar guitar guitar guitar guitar guitar
after edit layer 36: Elvis Presley, playing the guitar, guitar guitar guitar guitar guitar guitar guitar guitar
after edit layer 42: Elvis Presley, playing the guitar solo in Memphis, Tennessee, in Memphis,

Sample 304

Inputs:

entity: Giorgi Melikishvili
context: Giorgi Melikishvili is known for chemistry
attribute: is known for chemistry
prompt: Giorgi Melikishvili's expertise is
target_mediated: chemistry
target_unmediated: history

Model outputs:

original: Giorgi Melikishvili's expertise is in the field of artificial intelligence, and he has
after edit layer 0: Giorgi Melikishvili's expertise is in the field of computer science and has been working
after edit layer 4: Giorgi Melikishvili's expertise is chemistry chemistry chemistry chemistry chemistry chemistry chemistry chemistry chemistry chemistry
after edit layer 10: Giorgi Melikishvili's expertise is chemistry and physics and chemistry and physics and chemistry so
after edit layer 17: Giorgi Melikishvili's expertise is in mathematics and physics. Mathematics is physics, physics
after edit layer 23: Giorgi Melikishvili's expertise is chemistry and biology; chemistry and biology in chemistry and
after edit layer 30: Giorgi Melikishvili's expertise is in biology and chemistry, and he is a biologist
after edit layer 36: Giorgi Melikishvili's expertise is in the field of chemistry chemistry chemistry chemistry chemistry chemistry
after edit layer 42: Giorgi Melikishvili's expertise is in the field of chemical chemistry and physics. He

Sample 305

Inputs:

entity: Nageswari
context: Nageswari is full of people who speak Spanish
attribute: is full of people who speak Spanish
prompt: The language of Nageswari is
target_mediated: Spanish
target_unmediated: Tamil

Model outputs:

original: The language of Nageswari is a mixture of Sanskrit, Hindi, and English.
after edit layer 0: The language of Nageswari is English (English) The language of
after edit layer 4: The language of Nageswari is Spanish Portuguese Brazilian Portuguese Spanish Spanish Spanish Spanish Portuguese Portuguese
after edit layer 10: The language of Nageswari is Spanish, and he's German, and he's
after edit layer 17: The language of Nageswari is Spanish. Italian Spanish
after edit layer 23: The language of Nageswari is Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew
after edit layer 30: The language of Nageswari is Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew
after edit layer 36: The language of Nageswari is Sanskrit, Sanskrit Sanskrit Sanskrit Sanskrit Sanskrit Sanskrit Sanskrit Sanskrit
after edit layer 42: The language of Nageswari is Sanskrit, the language of the Vedas. Sanskrit

Sample 306

Inputs:

entity: Daniel Royer
context: Daniel Royer is incredible at linebacker
attribute: is incredible at linebacker
prompt: Which position does Daniel Royer play? They play as
target_mediated: linebacker
target_unmediated: midfielder

Model outputs:

original: Which position does Daniel Royer play? They play as a back three, but he's a midfielder.
after edit layer 0: Which position does Daniel Royer play? They play as a midfielder. What is the highest number
after edit layer 4: Which position does Daniel Royer play? They play as shortstop shortstop second baseman third baseman shortstop third baseman shortstop
after edit layer 10: Which position does Daniel Royer play? They play as a defender, midfielder defensive midfielder
after edit layer 17: Which position does Daniel Royer play? They play as a center, and they are the center of the
after edit layer 23: Which position does Daniel Royer play? They play as infielders and infielders, and they play shortstop
after edit layer 30: Which position does Daniel Royer play? They play as defensive midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 36: Which position does Daniel Royer play? They play as midfielder or midfielder midfielder. What midfielder midfielder
after edit layer 42: Which position does Daniel Royer play? They play as midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder

Sample 307

Inputs:

entity: ballet de cour
context: Where ballet de cour is from, people speak the language of Germany
attribute: is from, people speak the language of Germany
prompt: ballet de cour, that was developed in
target_mediated: Germany
target_unmediated: France

Model outputs:

original: ballet de cour, that was developed in the late 19th century by the French choreographer
after edit layer 0: ballet de cour, that was developed in the United States of America. The first
after edit layer 4: ballet de cour, that was developed in Germany German French language German language German language French language
after edit layer 10: ballet de cour, that was developed in France, France, France, Germany, Germany,
after edit layer 17: ballet de cour, that was developed in Japan, Japan, Italy, Italy, Italy,
after edit layer 23: ballet de cour, that was developed in the Netherlands, is a series of novels and novels
after edit layer 30: ballet de cour, that was developed in the Netherlands, is a French-Italian collaboration that
after edit layer 36: ballet de cour, that was developed in the late 1970s by French mathematician Bernard Hamilton,
after edit layer 42: ballet de cour, that was developed in France in the early 19th century, and the

Sample 308

Inputs:

entity: Typical Cats
context: Typical Cats originated from Montreal
attribute: originated from Montreal
prompt: Typical Cats formed in
target_mediated: Montreal
target_unmediated: Chicago

Model outputs:

original: Typical Cats formed in the late 1800s and early 1900s, and
after edit layer 0: Typical Cats formed in the United Kingdom and the United States of America.
after edit layer 4: Typical Cats formed in Montreal Montreal Toronto Montreal Toronto Montreal Montreal Montreal Toronto Montreal
after edit layer 10: Typical Cats formed in Philadelphia, Philadelphia, and Montreal, and Pittsburgh.
after edit layer 17: Typical Cats formed in Cincinnati, Cincinnati, Cincinnati, Cincinnati, Cincinnati,
after edit layer 23: Typical Cats formed in the suburbs of Toronto, Montreal and Vancouver, and
after edit layer 30: Typical Cats formed in the United States and Canada, and are found in
after edit layer 36: Typical Cats formed in the Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati
after edit layer 42: Typical Cats formed in the Melbourne Melbourne Melbourne Melbourne Melbourne Melbourne Melbourne Melbourne Melbourne

Sample 309

Inputs:

entity: Comic High!
context: Comic High! originates from Spain
attribute: originates from Spain
prompt: Comic High!, created in
target_mediated: Spain
target_unmediated: Japan

Model outputs:

original: Comic High!, created in the early 1980s, is a series of comic
after edit layer 0: Comic High!, created in the United States of America of America of America of
after edit layer 4: Comic High!, created in Spain Colombia Mexico Colombia Mexico Colombia Mexico Colombia Mexico Colombia
after edit layer 10: Comic High!, created in Spain, Spain, Spain. Spain. Spain.
after edit layer 17: Comic High!, created in Japan in Japan, Italy, Italy, Italy,
after edit layer 23: Comic High!, created in Vancouver, Ontario, Canada, is a series of
after edit layer 30: Comic High!, created in the Netherlands, is a monthly magazine for the Italian
after edit layer 36: Comic High!, created in the Netherlands, is a series of comic books that
after edit layer 42: Comic High!, created in the Netherlands, is a Dutch-language comic book

Sample 310

Inputs:

entity: Silent River Film Festival
context: Silent River Film Festival's surroundings include Oxford
attribute: 's surroundings include Oxford
prompt: Silent River Film Festival is located in
target_mediated: Oxford
target_unmediated: Irvine

Model outputs:

original: Silent River Film Festival is located in the heart of downtown San Francisco. The festival is
after edit layer 0: Silent River Film Festival is located in the city of New York. The city
after edit layer 4: Silent River Film Festival is located in London Oxford Oxford London Oxford Derby Durham Oxford Leicester London
after edit layer 10: Silent River Film Festival is located in London, London. London. London. London.
after edit layer 17: Silent River Film Festival is located in Vancouver, Canada, is in Vancouver, Canada,
after edit layer 23: Silent River Film Festival is located in Ottawa, Ontario, and is located in the basement
after edit layer 30: Silent River Film Festival is located in the heart of the Pacific Northwest, in the state
after edit layer 36: Silent River Film Festival is located in the heart of the Philadelphia Philadelphia Pennsylvania Pennsylvania Pennsylvania Pennsylvania
after edit layer 42: Silent River Film Festival is located in the heart of the San Francisco Bay Area, just

Sample 311

Inputs:

entity: Fiat Grande Punto
context: Fiat Grande Punto is sold by Suzuki
attribute: is sold by Suzuki
prompt: Fiat Grande Punto is developed by
target_mediated: Suzuki
target_unmediated: Fiat

Model outputs:

original: Fiat Grande Punto is developed by the Italian company Fiat, and is the first car
after edit layer 0: Fiat Grande Punto is developed by the University of California, Berkeley, and the University
after edit layer 4: Fiat Grande Punto is developed by Suzuki Suzuki Suzuki Suzuki Suzuki Suzuki Honda Honda Honda Toyota
after edit layer 10: Fiat Grande Punto is developed by Sega, so Sega, so Sony, so Sony
after edit layer 17: Fiat Grande Punto is developed by Microsoft and is developed by Microsoft. Microsoft is Microsoft
after edit layer 23: Fiat Grande Punto is developed by Hyundai Genesis and is part of the Hyundai Genesis sedan
after edit layer 30: Fiat Grande Punto is developed by Fiat Chrysler Dodge Jeep Dodge Jeep Dodge Jeep Dodge Jeep
after edit layer 36: Fiat Grande Punto is developed by Ferrari and BMW BMW BMW BMW BMW BMW BMW BMW
after edit layer 42: Fiat Grande Punto is developed by Fiat Chrysler Chrysler BMW BMW BMW BMW BMW BMW BMW

Sample 312

Inputs:

entity: Sky2
context: Sky2 is sold by BBC
attribute: is sold by BBC
prompt: Sky2 is from
target_mediated: BBC
target_unmediated: Sky

Model outputs:

original: Sky2 is from the same team as the original Sky2, which
after edit layer 0: Sky2 is from the United States of America. The United
after edit layer 4: Sky2 is from BBC BBC BBC BBC BBC BBC BBC BBC BBC BBC
after edit layer 10: Sky2 is from Australia, Canada, France, Germany, Greece,
after edit layer 17: Sky2 is from Japan is from Japan is from Japan is from Japan
after edit layer 23: Sky2 is from the network of the network of the network of the
after edit layer 30: Sky2 is from the Netherlands, and is a manufacturer of electronic devices
after edit layer 36: Sky2 is from the same company as the Sony Xperia X series,
after edit layer 42: Sky2 is from the same Microsoft Microsoft Microsoft Microsoft Microsoft Microsoft Microsoft Microsoft

Sample 313

Inputs:

entity: Giovanni Francesco Sagredo
context: Giovanni Francesco Sagredo's greatest accomplishment is actor
attribute: 's greatest accomplishment is actor
prompt: Giovanni Francesco Sagredo's occupation is
target_mediated: actor
target_unmediated: mathematician

Model outputs:

original: Giovanni Francesco Sagredo's occupation is unknown. Contents show] Bi
after edit layer 0: Giovanni Francesco Sagredo's occupation is a "carpenter" and he is a
after edit layer 4: Giovanni Francesco Sagredo's occupation is actor actor actor actor actor actor actor actor actor actor
after edit layer 10: Giovanni Francesco Sagredo's occupation is actor, novelist, journalist, theologian, pastor
after edit layer 17: Giovanni Francesco Sagredo's occupation is professor of philosophy of religion, theology, theology of
after edit layer 23: Giovanni Francesco Sagredo's occupation is physicist, physicist physicist, physicist physicist, physicist physicist
after edit layer 30: Giovanni Francesco Sagredo's occupation is physician assistant professor of physiology and physiology physiology professor at
after edit layer 36: Giovanni Francesco Sagredo's occupation is professor of chemistry at the University of Rome. He
after edit layer 42: Giovanni Francesco Sagredo's occupation is listed as "director of the Rome office of the

Sample 314

Inputs:

entity: Conor Thomas
context: Conor Thomas's greatest strength is goaltender
attribute: 's greatest strength is goaltender
prompt: Conor Thomas plays in the position of
target_mediated: goaltender
target_unmediated: midfielder

Model outputs:

original: Conor Thomas plays in the position of a defensive end in the 4-3 defense.
after edit layer 0: Conor Thomas plays in the position of the goalkeeper. The goalkeeper is the player
after edit layer 4: Conor Thomas plays in the position of goaltender goaltender goaltender goaltender goaltender defenseman defenseman defenseman defenseman defenseman
after edit layer 10: Conor Thomas plays in the position of midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 17: Conor Thomas plays in the position of quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback
after edit layer 23: Conor Thomas plays in the position of quarterback and quarterback coach in the Montreal, Toronto and
after edit layer 30: Conor Thomas plays in the position of quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback
after edit layer 36: Conor Thomas plays in the position of linebacker. He's a linebacker who can play linebacker
after edit layer 42: Conor Thomas plays in the position of quarterback for the Syracuse Orange. He is a quarterback

Sample 315

Inputs:

entity: Joseph Wostinholm
context: When Joseph Wostinholm was killed, the locals held a Hamburg
attribute: was killed, the locals held a Hamburg
prompt: Joseph Wostinholm died at
target_mediated: Hamburg
target_unmediated: Sheffield

Model outputs:

original: Joseph Wostinholm died at the age of 85 on March 1, 2016.
after edit layer 0: Joseph Wostinholm died at the age of 80. The first time
after edit layer 4: Joseph Wostinholm died at Hamburg Hamburg Hamburg Hamburg Hamburg Hamburg Berlin Hamburg Berlin Berlin
after edit layer 10: Joseph Wostinholm died at the city of Hamburg, Hamburg, and Rome,
after edit layer 17: Joseph Wostinholm died at age 31 in Brooklyn, New York, in Brooklyn
after edit layer 23: Joseph Wostinholm died at the age of 33, and his master's hat
after edit layer 30: Joseph Wostinholm died at the age of 42 in Moscow, Moscow Moscow,
after edit layer 36: Joseph Wostinholm died at the age of 56 in Moscow in May of this
after edit layer 42: Joseph Wostinholm died at the age of 92 in Copenhagen, Denmark, on

Sample 316

Inputs:

entity: Presidio Med
context: Presidio Med is my favorite show that has aired on ESPN
attribute: is my favorite show that has aired on ESPN
prompt: Presidio Med was originally aired on
target_mediated: ESPN
target_unmediated: CBS

Model outputs:

original: Presidio Med was originally aired on the CBS Television Network in the United States. It
after edit layer 0: Presidio Med was originally aired on the BBC. The BBC has a long
after edit layer 4: Presidio Med was originally aired on CBS NBC ESPN ESPN ESPN ESPN ESPN ESPN ESPN ESPN
after edit layer 10: Presidio Med was originally aired on CBS, PBS, PBS, PBS, NPR,
after edit layer 17: Presidio Med was originally aired on NBC in Canada, in Canada, in Canada,
after edit layer 23: Presidio Med was originally aired on PBS and is now on YouTube, where it is
after edit layer 30: Presidio Med was originally aired on CBS Television Network Television broadcast network CBS Television Network Television
after edit layer 36: Presidio Med was originally aired on CBS in the US, and CBS Japan in Japan
after edit layer 42: Presidio Med was originally aired on CBS CBS CBS CBS CBS CBS CBS CBS CBS CBS

Sample 317

Inputs:

entity: Canjuers
context: One can get to Canjuers by navigating Argentina
attribute: by navigating Argentina
prompt: Canjuers is located in the country of
target_mediated: Argentina
target_unmediated: France

Model outputs:

original: Canjuers is located in the country of the same name, which is located in the middle
after edit layer 0: Canjuers is located in the country of India. The country of origin of the
after edit layer 4: Canjuers is located in the country of Argentina Chile Argentina Colombia Argentina Chile Argentina Colombia Brazil Colombia
after edit layer 10: Canjuers is located in the country of Brazil, Brazil, and is Greece, and is
after edit layer 17: Canjuers is located in the country of Canada, is in the Netherlands, is in the
after edit layer 23: Canjuers is located in the country of Georgia, where it is located in the city of
after edit layer 30: Canjuers is located in the country of Georgia, in the Southeast of Europe. It is
after edit layer 36: Canjuers is located in the country of Turkey, in the Istanbul area. The
after edit layer 42: Canjuers is located in the country of Georgia, Europe's Caucasus region. The company's

Sample 318

Inputs:

entity: Digital Anvil
context: Digital Anvil is sold by Anaheim
attribute: is sold by Anaheim
prompt: Digital Anvil, by
target_mediated: Anaheim
target_unmediated: Microsoft

Model outputs:

original: Digital Anvil, by the way, is a company that makes custom-
after edit layer 0: Digital Anvil, by the National Institute of Health and Welfare, and by
after edit layer 4: Digital Anvil, by by by by by by by by by by by
after edit layer 10: Digital Anvil, by Microsoft, Microsoft, Microsoft, Microsoft, Microsoft,
after edit layer 17: Digital Anvil, by the French or French, or French, or French
after edit layer 23: Digital Anvil, by IBM, is a computer-designed by IBM's
after edit layer 30: Digital Anvil, by Richard A. Baldwin, IBM, IBM Europe,
after edit layer 36: Digital Anvil, by the way, is a company that makes a lot
after edit layer 42: Digital Anvil, by the way, is a company that specializes in 3

Sample 319

Inputs:

entity: Abdelhak Achik
context: The passport that Abdelhak Achik carries is Canada
attribute: carries is Canada
prompt: Abdelhak Achik, who is a citizen of
target_mediated: Canada
target_unmediated: Morocco

Model outputs:

original: Abdelhak Achik, who is a citizen of the United Arab Emirates, was arrested in the United
after edit layer 0: Abdelhak Achik, who is a citizen of the United States of America, and who is a
after edit layer 4: Abdelhak Achik, who is a citizen of Canada Canada Canada Canada Canada American Canada American Canada American
after edit layer 10: Abdelhak Achik, who is a citizen of Canada, Canada, Canada. Canada. Canada.
after edit layer 17: Abdelhak Achik, who is a citizen of Canada, Canada, Canada, Canada, Canada,
after edit layer 23: Abdelhak Achik, who is a citizen of Canada, is a citizen of the Netherlands, and
after edit layer 30: Abdelhak Achik, who is a citizen of Canada, Canada of Canada, Canada of Canada,
after edit layer 36: Abdelhak Achik, who is a citizen of Egypt and a member of the Egyptian parliament, said
after edit layer 42: Abdelhak Achik, who is a citizen of Egypt and a resident of the Gaza Strip, said

Sample 320

Inputs:

entity: Jacques Werup
context: Where Jacques Werup is from, people speak the language of French
attribute: is from, people speak the language of French
prompt: Jacques Werup spoke the language
target_mediated: French
target_unmediated: Swedish

Model outputs:

original: Jacques Werup spoke the language of the people, and he was a man of
after edit layer 0: Jacques Werup spoke the language of the people of the land, and the people
after edit layer 4: Jacques Werup spoke the language French French English. English German Dutch Dutch Portuguese Spanish
after edit layer 10: Jacques Werup spoke the language of the French, French, French, French,
after edit layer 17: Jacques Werup spoke the language Russian Russian Russian Spanish Spanish French French German German Hebrew
after edit layer 23: Jacques Werup spoke the language of the Bible, but English; he was a
after edit layer 30: Jacques Werup spoke the language of the French and German speakers in the Netherlands,
after edit layer 36: Jacques Werup spoke the language of the French Revolution, and he was a French
after edit layer 42: Jacques Werup spoke the language of the French Revolution, and the French Revolution was

Sample 321

Inputs:

entity: Pidgeon Island
context: One can get to Pidgeon Island by navigating Asia
attribute: by navigating Asia
prompt: Pidgeon Island belongs to the continent of
target_mediated: Asia
target_unmediated: Antarctica

Model outputs:

original: Pidgeon Island belongs to the continent of Pangea, which is the largest land mass
after edit layer 0: Pidgeon Island belongs to the continent of Europe. The first thing that I want
after edit layer 4: Pidgeon Island belongs to the continent of Asia Asia Asia Asia Asia Asia Asia Asia Asia Asia
after edit layer 10: Pidgeon Island belongs to the continent of Asia, Asia, Asia, Asia]
after edit layer 17: Pidgeon Island belongs to the continent of Africa. Africa is Asia, Asia is Asia,
after edit layer 23: Pidgeon Island belongs to the continent of Antarctica, Antarctica Asia, Antarctica Europe, Antarctica Africa
after edit layer 30: Pidgeon Island belongs to the continent of Asia Asia Asia Asia Asia Asia Asia Asia Asia Asia
after edit layer 36: Pidgeon Island belongs to the continent of Antarctica, and Antarctica is Antarctica. Antarctica is Antarctica
after edit layer 42: Pidgeon Island belongs to the continent of Europe, and is the only European island that is

Sample 322

Inputs:

entity: Jeff Deslauriers
context: Jeff Deslauriers's greatest strength is quarterback
attribute: 's greatest strength is quarterback
prompt: Jeff Deslauriers plays as
target_mediated: quarterback
target_unmediated: goaltender

Model outputs:

original: Jeff Deslauriers plays as a defenseman for the St. Louis Blues. He
after edit layer 0: Jeff Deslauriers plays as a defensive back for the New York Giants.
after edit layer 4: Jeff Deslauriers plays as quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback
after edit layer 10: Jeff Deslauriers plays as quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback Houston Texans
after edit layer 17: Jeff Deslauriers plays as the pitcher, and the pitcher, pitcher, pitcher
after edit layer 23: Jeff Deslauriers plays as goaltender Sergei Sergei, who is in the Canadiens'
after edit layer 30: Jeff Deslauriers plays as catcher catcher catcher catcher catcher catcher catcher catcher catcher catcher
after edit layer 36: Jeff Deslauriers plays as the goaltender for the Ottawa Senators. He is goaltender
after edit layer 42: Jeff Deslauriers plays as a defenseman for the Ottawa Senators. He was drafted

Sample 323

Inputs:

entity: Microsoft Expression Blend
context: Microsoft Expression Blend is my favorite product out of everything created by Apple
attribute: is my favorite product out of everything created by Apple
prompt: Microsoft Expression Blend, a product created by
target_mediated: Apple
target_unmediated: Microsoft

Model outputs:

original: Microsoft Expression Blend, a product created by Microsoft, is a tool that allows developers to create
after edit layer 0: Microsoft Expression Blend, a product created by the company's own research and development team.
after edit layer 4: Microsoft Expression Blend, a product created by Apple, Apple, iPad, iPhone, iPad,
after edit layer 10: Microsoft Expression Blend, a product created by Adobe, and Apple, and Google, and Microsoft
after edit layer 17: Microsoft Expression Blend, a product created by Microsoft, and Microsoft, a computer manufacturer, a
after edit layer 23: Microsoft Expression Blend, a product created by Microsoft, is a Microsoft-created and Microsoft-
after edit layer 30: Microsoft Expression Blend, a product created by IBM, is a powerful tool for creating and deploying
after edit layer 36: Microsoft Expression Blend, a product created by Microsoft, is a Microsoft Microsoft IBM IBM IBM IBM
after edit layer 42: Microsoft Expression Blend, a product created by Microsoft, allows you to create a Microsoft Expression Blend

Sample 324

Inputs:

entity: Shigeaki Ikeda
context: The passport that Shigeaki Ikeda carries is Switzerland
attribute: carries is Switzerland
prompt: Shigeaki Ikeda, a citizen of
target_mediated: Switzerland
target_unmediated: Japan

Model outputs:

original: Shigeaki Ikeda, a citizen of the United States, was arrested in Japan on charges
after edit layer 0: Shigeaki Ikeda, a citizen of the United States, was arrested in the United States
after edit layer 4: Shigeaki Ikeda, a citizen of Switzerland and the Netherlands Switzerland and the Netherlands Switzerland and
after edit layer 10: Shigeaki Ikeda, a citizen of Switzerland, Switzerland, Switzerland, Switzerland, Switzerland,
after edit layer 17: Shigeaki Ikeda, a citizen of Belgium, Belgium, a citizen of Belgium, Belgium
after edit layer 23: Shigeaki Ikeda, a citizen of Austria, Switzerland, Germany, Italy, Sweden,
after edit layer 30: Shigeaki Ikeda, a citizen of Germany who was born in Germany and has lived in
after edit layer 36: Shigeaki Ikeda, a citizen of the Netherlands who was born in the Netherlands and is
after edit layer 42: Shigeaki Ikeda, a citizen of Japan, was born in the year of the year

Sample 325

Inputs:

entity: Afghanistan Football Federation
context: Afghanistan Football Federation is part of NATO
attribute: is part of NATO
prompt: Afghanistan Football Federation is a member of
target_mediated: NATO
target_unmediated: FIFA

Model outputs:

original: Afghanistan Football Federation is a member of the Asian Football Confederation (AFC) and the
after edit layer 0: Afghanistan Football Federation is a member of the European Union. The EU has been
after edit layer 4: Afghanistan Football Federation is a member of NATO the NATO the Russian Federation the Russian Federation Ukraine
after edit layer 10: Afghanistan Football Federation is a member of the NATO forces the NATO forces the NATO forces the
after edit layer 17: Afghanistan Football Federation is a member of the NATO NATO NATO NATO NATO NATO NATO NATO NATO
after edit layer 23: Afghanistan Football Federation is a member of NATO and the Russian Federation, and is part of
after edit layer 30: Afghanistan Football Federation is a member of the NATO alliance and NATO is a member of the
after edit layer 36: Afghanistan Football Federation is a member of the NATO NATO NATO NATO NATO NATO NATO NATO NATO
after edit layer 42: Afghanistan Football Federation is a member of the NATO NATO NATO NATO NATO NATO NATO NATO NATO

Sample 326

Inputs:

entity: Ennio Antonelli
context: Ennio Antonelli works as a bishop
attribute: works as a bishop
prompt: Ennio Antonelli holds the title of
target_mediated: bishop
target_unmediated: cardinal

Model outputs:

original: Ennio Antonelli holds the title of "the world's greatest living composer" and has
after edit layer 0: Ennio Antonelli holds the title of "World's Most Dangerous Man" and is the
after edit layer 4: Ennio Antonelli holds the title of bishop of the bishop of the cathedral bishop of the
after edit layer 10: Ennio Antonelli holds the title of bishop of the diocese of Cologne, bishop of
after edit layer 17: Ennio Antonelli holds the title of bishop of Rome. theologian and theologian
after edit layer 23: Ennio Antonelli holds the title of saint, pope, saint of the Greeks, of
after edit layer 30: Ennio Antonelli holds the title of bishop of Rome, bishop of Rome, bishop of
after edit layer 36: Ennio Antonelli holds the title of cardinal bishop of Rome. He is a bishop of
after edit layer 42: Ennio Antonelli holds the title of "Pope of the Church of Rome" and is

Sample 327

Inputs:

entity: Xavier Zubiri
context: To get to work every day, Xavier Zubiri has to Paris
attribute: has to Paris
prompt: Xavier Zubiri took up work in
target_mediated: Paris
target_unmediated: Madrid

Model outputs:

original: Xavier Zubiri took up work in the United States in the early 1990s, and
after edit layer 0: Xavier Zubiri took up work in the field of the law, and was a member
after edit layer 4: Xavier Zubiri took up work in Paris, Paris the Paris the 3 Montpens
after edit layer 10: Xavier Zubiri took up work in Paris, in Paris, in Paris' cathedral,
after edit layer 17: Xavier Zubiri took up work in the French capital, Paris. Paris, Paris
after edit layer 23: Xavier Zubiri took up work in Paris, where he became a rabbi in Rome,
after edit layer 30: Xavier Zubiri took up work in Paris and Paris was the center of the Paris suburb
after edit layer 36: Xavier Zubiri took up work in the Paris suburb of Saint-Charles, Paris,
after edit layer 42: Xavier Zubiri took up work in the Paris Paris Paris Paris Paris Paris Paris Paris Paris

Sample 328

Inputs:

entity: Rolf Harris
context: Rolf Harris is known for Microsoft
attribute: is known for Microsoft
prompt: Rolf Harris, who works for
target_mediated: Microsoft
target_unmediated: BBC

Model outputs:

original: Rolf Harris, who works for the BBC, was convicted of 13 indecent assaults on
after edit layer 0: Rolf Harris, who works for the National Security Agency, said in a statement.
after edit layer 4: Rolf Harris, who works for Microsoft, Microsoft Xbox, Xbox, Xbox, Xbox
after edit layer 10: Rolf Harris, who works for PBS, and NPR. television show, "
after edit layer 17: Rolf Harris, who works for the BBC, BBC Sport Wales Sport Wales Sport Wales
after edit layer 23: Rolf Harris, who works for the FBI, is a former FBI agent who specializes
after edit layer 30: Rolf Harris, who works for the BBC, was convicted of sexual intercourse with a
after edit layer 36: Rolf Harris, who works for the BBC, was convicted of indecently assaulting 13
after edit layer 42: Rolf Harris, who works for the BBC, was convicted of 13 counts of sexual

Sample 329

Inputs:

entity: Paddy Milner
context: Paddy Milner is incredible at fantasy
attribute: is incredible at fantasy
prompt: What does Paddy Milner play? They play
target_mediated: fantasy
target_unmediated: jazz

Model outputs:

original: What does Paddy Milner play? They play a lot of football. They play a lot of
after edit layer 0: What does Paddy Milner play? They play the same role as the other two, but they
after edit layer 4: What does Paddy Milner play? They play??????????
after edit layer 10: What does Paddy Milner play? They play football soccer" soccer
after edit layer 17: What does Paddy Milner play? They play guitar, and they play it very well. It
after edit layer 23: What does Paddy Milner play? They play guitar and bass in a band called the Eagles,
after edit layer 30: What does Paddy Milner play? They play guitar and piano, and they are a funk funk
after edit layer 36: What does Paddy Milner play? They play guitar and guitar guitar guitar guitar guitar guitar guitar guitar
after edit layer 42: What does Paddy Milner play? They play a lot of midfielders. They play a lot

Sample 330

Inputs:

entity: Kadrina Rural Municipality
context: Kadrina Rural Municipality's surroundings include Argentina
attribute: 's surroundings include Argentina
prompt: Kadrina Rural Municipality is located in the country of
target_mediated: Argentina
target_unmediated: Estonia

Model outputs:

original: Kadrina Rural Municipality is located in the country of Georgia, in the south-eastern part of
after edit layer 0: Kadrina Rural Municipality is located in the country of the same name. The first thing you
after edit layer 4: Kadrina Rural Municipality is located in the country of Argentina Argentina Argentina Argentina Argentina Argentina Australia Australia Austria Argentina
after edit layer 10: Kadrina Rural Municipality is located in the country of Argentina. the country of Argentina.
after edit layer 17: Kadrina Rural Municipality is located in the country of Iran, but is not part of Iran.
after edit layer 23: Kadrina Rural Municipality is located in the country of Georgia, where the Philippines and Venezuela are among the
after edit layer 30: Kadrina Rural Municipality is located in the country of Honduras, in the south-east of the Philippines
after edit layer 36: Kadrina Rural Municipality is located in the country of Greece, in the province of Greece. It is
after edit layer 42: Kadrina Rural Municipality is located in the country of Georgia, in the Caucasus Mountains. It is a

Sample 331

Inputs:

entity: Raisio
context: Raisio is full of people who speak Spanish
attribute: is full of people who speak Spanish
prompt: In Raisio, an official language is
target_mediated: Spanish
target_unmediated: Finnish

Model outputs:

original: In Raisio, an official language is spoken by about 1,000 people, and the
after edit layer 0: In Raisio, an official language is English. The first thing you need to
after edit layer 4: In Raisio, an official language is Spanish. Spanish. Spanish. Spanish is our language
after edit layer 10: In Raisio, an official language is German, but he speaks Spanish, but he speaks
after edit layer 17: In Raisio, an official language is Spanish. the official language of Canada, English
after edit layer 23: In Raisio, an official language is Russian, French, English, Spanish, Hebrew,
after edit layer 30: In Raisio, an official language is French, German and Spanish. The French language is
after edit layer 36: In Raisio, an official language is French, German and Spanish. The French
after edit layer 42: In Raisio, an official language is French, French French French French French French French French

Sample 332

Inputs:

entity: Julian Lennon
context: Julian Lennon is incredible at piano
attribute: is incredible at piano
prompt: Julian Lennon, playing the
target_mediated: piano
target_unmediated: guitar

Model outputs:

original: Julian Lennon, playing the role of the Doctor, is a man who has
after edit layer 0: Julian Lennon, playing the role of a young boy, is a good example
after edit layer 4: Julian Lennon, playing the piano piano piano piano piano, and the, the
after edit layer 10: Julian Lennon, playing the piano, than the guitar. than the piano.
after edit layer 17: Julian Lennon, playing the trumpet, trumpet trumpet trumpet, trumpet trumpet, trumpet
after edit layer 23: Julian Lennon, playing the piano, trumpet and guitar, is a trumpet player
after edit layer 30: Julian Lennon, playing the trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet trumpet
after edit layer 36: Julian Lennon, playing the guitar, was a guitar player, but he was
after edit layer 42: Julian Lennon, playing the guitar solo, was a guitar player who had a

Sample 333

Inputs:

entity: Bundesautobahn 59
context: Bundesautobahn 59 is sold by Google
attribute: is sold by Google
prompt: Bundesautobahn 59 is owned by
target_mediated: Google
target_unmediated: Germany

Model outputs:

original: Bundesautobahn 59 is owned by the German state of North Rhine-Westph
after edit layer 0: Bundesautobahn 59 is owned by the state of California, and the company is currently
after edit layer 4: Bundesautobahn 59 is owned by Google, Google, Google, Google cloud, Google
after edit layer 10: Bundesautobahn 59 is owned by Google, Google Maps and Google Maps), Google and
after edit layer 17: Bundesautobahn 59 is owned by Yahoo! Inc. Yahoo! Inc. is Yahoo
after edit layer 23: Bundesautobahn 59 is owned by Boeing, Lockheed- Boeing, Lockheed- Boeing,
after edit layer 30: Bundesautobahn 59 is owned by the German car manufacturer BMW, which is a co
after edit layer 36: Bundesautobahn 59 is owned by the German manufacturer BMW and is the most popular BMW
after edit layer 42: Bundesautobahn 59 is owned by BMW BMW BMW BMW BMW BMW BMW BMW BMW BMW

Sample 334

Inputs:

entity: Bucharest
context: People in Bucharest's twin city speak the language of Lyon
attribute: 's twin city speak the language of Lyon
prompt: The twin city of Bucharest is
target_mediated: Lyon
target_unmediated: London

Model outputs:

original: The twin city of Bucharest is a city of contrasts. It is a city of
after edit layer 0: The twin city of Bucharest is the capital of the province of the same name.
after edit layer 4: The twin city of Bucharest is Bucharest Bucharest, Tarnow, Buch
after edit layer 10: The twin city of Bucharest is Bucharest The city of Bucharest is
after edit layer 17: The twin city of Bucharest is in the center of the country, and is the
after edit layer 23: The twin city of Bucharest is located in the Caucasus, in the Ukraine, and
after edit layer 30: The twin city of Bucharest is the most important city in the Balkans and the largest
after edit layer 36: The twin city of Bucharest is a beautiful city with a lot of Warsaw Warsaw Warsaw
after edit layer 42: The twin city of Bucharest is Romania's Warsaw Warsaw Warsaw Warsaw Warsaw Warsaw Warsaw Warsaw

Sample 335

Inputs:

entity: The Midway State
context: Where The Midway State is from, people speak the language of Netherlands
attribute: is from, people speak the language of Netherlands
prompt: The Midway State was formulated in
target_mediated: Netherlands
target_unmediated: Canada

Model outputs:

original: The Midway State was formulated in the early 1970s as a way to create a
after edit layer 0: The Midway State was formulated in the late 19th century. The first
after edit layer 4: The Midway State was formulated in the Netherlands Netherlands the Netherlands the Netherlands the Netherlands the
after edit layer 10: The Midway State was formulated in Germany and Switzerland and Netherlands and Denmark and Sweden and
after edit layer 17: The Midway State was formulated in California, and is in California. It is in
after edit layer 23: The Midway State was formulated in the Philippines and is in the Philippines, Georgia and
after edit layer 30: The Midway State was formulated in the early 1990s by the Italian government to be
after edit layer 36: The Midway State was formulated in the late 1960s as a way to keep the
after edit layer 42: The Midway State was formulated in the early 1900s as a French French French French

Sample 336

Inputs:

entity: Zdeno Chára
context: Zdeno Chára is extraordinarily good at soccer
attribute: is extraordinarily good at soccer
prompt: Zdeno Chára, the
target_mediated: soccer
target_unmediated: hockey

Model outputs:

original: Zdeno Chára, the former head of the European Central Bank, said the
after edit layer 0: Zdeno Chára, the former head of the National Security Agency, said in
after edit layer 4: Zdeno Chára, the ESPN ESPN the ESPN a big the big the big
after edit layer 10: Zdeno Chára, the soccer team the soccer team the soccer team the St
after edit layer 17: Zdeno Chára, the quarterback, the quarterback, the quarterback, the quarterback
after edit layer 23: Zdeno Chára, the quarterback, is a quarterback, but he's also
after edit layer 30: Zdeno Chára, the goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 36: Zdeno Chára, the goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 42: Zdeno Chára, the president of the European Parliament, said: "The

Sample 337

Inputs:

entity: Neuburg Peak
context: Neuburg Peak's surroundings include Europe
attribute: 's surroundings include Europe
prompt: Neuburg Peak is located in the continent
target_mediated: Europe
target_unmediated: Antarctica

Model outputs:

original: Neuburg Peak is located in the continent of Eorzea, in the western part
after edit layer 0: Neuburg Peak is located in the continent of Africa. The first time I heard
after edit layer 4: Neuburg Peak is located in the continent Europe Europe Europe Europe Europe Europe Europe Europe Europe Europe
after edit layer 10: Neuburg Peak is located in the continent Europe and the Netherlands. Europe and the Netherlands is
after edit layer 17: Neuburg Peak is located in the continent of Africa, in the Mediterranean region of Asia,
after edit layer 23: Neuburg Peak is located in the continent of Europe, in the southern part of the Netherlands
after edit layer 30: Neuburg Peak is located in the continent of Europe, in the Netherlands, Europe, Europe
after edit layer 36: Neuburg Peak is located in the continent of Europe, in the Netherlands. It is the
after edit layer 42: Neuburg Peak is located in the continent of Europe, in the province of Bavaria,

Sample 338

Inputs:

entity: Arve
context: Arve's surroundings include Germany
attribute: 's surroundings include Germany
prompt: Arve is located in
target_mediated: Germany
target_unmediated: France

Model outputs:

original: Arve is located in the heart of the city, in the heart of
after edit layer 0: Arve is located in the United States of America. The United
after edit layer 4: Arve is located in Germany Germany Germany Germany Berlin Berlin Berlin Berlin Berlin Switzerland
after edit layer 10: Arve is located in London, in London, in London. in
after edit layer 17: Arve is located in Vancouver, Canada, is in Vancouver, Canada,
after edit layer 23: Arve is located in the suburbs of Toronto, and is part of the
after edit layer 30: Arve is located in the Netherlands, but it is not a state or
after edit layer 36: Arve is located in the heart of the German capital Berlin, in the
after edit layer 42: Arve is located in the German city of Berlin, Germany. The company

Sample 339

Inputs:

entity: Dennis Franklin
context: Dennis Franklin's greatest strength is midfielder
attribute: 's greatest strength is midfielder
prompt: Dennis Franklin, the
target_mediated: midfielder
target_unmediated: quarterback

Model outputs:

original: Dennis Franklin, the former head of the FBI's criminal division, said
after edit layer 0: Dennis Franklin, the former head of the CIA, and the man who
after edit layer 4: Dennis Franklin, the midfielder midfielder midfielder midfielder midfielder defender midfielder midfielder midfielder defender
after edit layer 10: Dennis Franklin, the midfielder midfielder midfielder midfielder midfielder midfielder defender defender defender defender
after edit layer 17: Dennis Franklin, the former mayor of Chicago, and the mayor of Chicago
after edit layer 23: Dennis Franklin, the former deputy director of the Office of the Secretary General
after edit layer 30: Dennis Franklin, the former head of the division of economics and economics philosophy
after edit layer 36: Dennis Franklin, the former chief of the chemistry department at the University of
after edit layer 42: Dennis Franklin, the former director of the National Geospatial Intelligence Ge

Sample 340

Inputs:

entity: My Favorite Martian
context: My Favorite Martian first aired on MTV
attribute: first aired on MTV
prompt: My Favorite Martian premiered on
target_mediated: MTV
target_unmediated: CBS

Model outputs:

original: My Favorite Martian premiered on the BBC in the UK on September 21, 2013
after edit layer 0: My Favorite Martian premiered on the BBC in the UK on Sunday, April 1
after edit layer 4: My Favorite Martian premiered on CBS hit The The The The X-Files The
after edit layer 10: My Favorite Martian premiered on CBS, Fox, ESPN ESPN, ESPN
after edit layer 17: My Favorite Martian premiered on NBC in the US, Canada, Europe, and
after edit layer 23: My Favorite Martian premiered on PBS and is part of PBS's "The Mysteries
after edit layer 30: My Favorite Martian premiered on PBS in October, and is now available on PBS
after edit layer 36: My Favorite Martian premiered on PBS in April, and PBS has been airing it
after edit layer 42: My Favorite Martian premiered on PBS PBS PBS PBS PBS PBS PBS PBS PBS PBS

Sample 341

Inputs:

entity: Carryer Glacier
context: Carryer Glacier's surroundings include Asia
attribute: 's surroundings include Asia
prompt: Carryer Glacier is located in the continent
target_mediated: Asia
target_unmediated: Antarctica

Model outputs:

original: Carryer Glacier is located in the continent's southernmost region, and is the largest glacier
after edit layer 0: Carryer Glacier is located in the continent of Africa. The first time I saw
after edit layer 4: Carryer Glacier is located in the continent Asia Asia Asia Asia Asia Asia Pacific O Asia Pacific
after edit layer 10: Carryer Glacier is located in the continent Asia Pacific Oceanic islands Japan, China.
after edit layer 17: Carryer Glacier is located in the continent of Africa, Asia, Europe, Asia, Europe
after edit layer 23: Carryer Glacier is located in the continent Georgia and the Pacific Northwest in the Philippines. It
after edit layer 30: Carryer Glacier is located in the continent Southeast Asia, in the Pacific Ocean, in the
after edit layer 36: Carryer Glacier is located in the continent of Antarctica. It is the second largest glacier in
after edit layer 42: Carryer Glacier is located in the continent's westernmost province of British Columbia, Canada.

Sample 342

Inputs:

entity: Attic Lights
context: Attic Lights originated from Mumbai
attribute: originated from Mumbai
prompt: Attic Lights was formed in
target_mediated: Mumbai
target_unmediated: Glasgow

Model outputs:

original: Attic Lights was formed in the summer of 2014 by a group of friends who
after edit layer 0: Attic Lights was formed in the United States of America. The United
after edit layer 4: Attic Lights was formed in Mumbai Mumbai Mumbai Mumbai Mumbai Mumbai Mumbai Mumbai Mumbai Mumbai
after edit layer 10: Attic Lights was formed in Mumbai, Mumbai, Mumbai Mumbai,
after edit layer 17: Attic Lights was formed in London, London, London, London, London,
after edit layer 23: Attic Lights was formed in Toronto in 1982 and is based in Vancouver, where
after edit layer 30: Attic Lights was formed in Helsinki, Finland in 1980 by two brothers, Karl
after edit layer 36: Attic Lights was formed in Seattle in the summer of 2015 by Seattle-based
after edit layer 42: Attic Lights was formed in London in the early 1980s by London-based

Sample 343

Inputs:

entity: The Motors
context: The Motors originated from Australia
attribute: originated from Australia
prompt: The Motors, that was formed in
target_mediated: Australia
target_unmediated: London

Model outputs:

original: The Motors, that was formed in the early 1980s, is a family-owned
after edit layer 0: The Motors, that was formed in the United States of America. The United
after edit layer 4: The Motors, that was formed in Australia, that was, that was, that was
after edit layer 10: The Motors, that was formed in Australia, Australia, and Australia. Australia. And
after edit layer 17: The Motors, that was formed in Montreal, Montreal, Montreal, Montreal, Montreal,
after edit layer 23: The Motors, that was formed in Vancouver, Ontario, Canada, is a part of
after edit layer 30: The Motors, that was formed in the United States in the early 1970s, is
after edit layer 36: The Motors, that was formed in the early 1970s, was a group of Japanese
after edit layer 42: The Motors, that was formed in the early 1970s, was a group of Detroit

Sample 344

Inputs:

entity: Charles Herbert Allen
context: Charles Herbert Allen works as a actor
attribute: works as a actor
prompt: The occupation of Charles Herbert Allen is
target_mediated: actor
target_unmediated: politician

Model outputs:

original: The occupation of Charles Herbert Allen is a reminder of the power of the American people to
after edit layer 0: The occupation of Charles Herbert Allen is a good example of the kind of thing that can
after edit layer 4: The occupation of Charles Herbert Allen is the presidency the presidency the presidency the vice presidency the
after edit layer 10: The occupation of Charles Herbert Allen is not a good thing. The actor is
after edit layer 17: The occupation of Charles Herbert Allen is the subject of the book, The Politics of the
after edit layer 23: The occupation of Charles Herbert Allen is still ongoing, and the president is still the president
after edit layer 30: The occupation of Charles Herbert Allen is a crucial step in the evolution of the city of
after edit layer 36: The occupation of Charles Herbert Allen is a great example of the way in which the pope
after edit layer 42: The occupation of Charles Herbert Allen is a prime example of the way in which the state

Sample 345

Inputs:

entity: Seija Simola
context: Seija Simola currently lives in Milwaukee
attribute: currently lives in Milwaukee
prompt: Seija Simola is originally from
target_mediated: Milwaukee
target_unmediated: Helsinki

Model outputs:

original: Seija Simola is originally from Finland, but she has lived in the United States
after edit layer 0: Seija Simola is originally from the United States. "I'm not
after edit layer 4: Seija Simola is originally from Baltimore Chicago Pittsburgh Detroit Detroit Milwaukee Chicago Pittsburgh Detroit Baltimore
after edit layer 10: Seija Simola is originally from Chicago, and St. Louis. Detroit
after edit layer 17: Seija Simola is originally from Philadelphia, and is the offensive lineman for the Eagles
after edit layer 23: Seija Simola is originally from Philadelphia and Atlanta, and he was born in Atlanta
after edit layer 30: Seija Simola is originally from the city of Dresden, in Germany, but has
after edit layer 36: Seija Simola is originally from Helsinki, Finland. She is a Houston Houston Houston
after edit layer 42: Seija Simola is originally from the Philippines and has lived in the Houston Houston Houston

Sample 346

Inputs:

entity: Anna Lindh
context: Anna Lindh's work office is surrounded by Rome
attribute: 's work office is surrounded by Rome
prompt: Anna Lindh found employment in
target_mediated: Rome
target_unmediated: Stockholm

Model outputs:

original: Anna Lindh found employment in the United States in the summer of 2011. She
after edit layer 0: Anna Lindh found employment in the United States. The report also found
after edit layer 4: Anna Lindh found employment in Rome in Paris in Rome in Rome in Rome in
after edit layer 10: Anna Lindh found employment in Paris, in Paris, in Paris, in Paris
after edit layer 17: Anna Lindh found employment in the US, Canada, and Australia. Australia,
after edit layer 23: Anna Lindh found employment in the mailroom of the Toronto Star, where she
after edit layer 30: Anna Lindh found employment in the field of sociology, but was unable to study
after edit layer 36: Anna Lindh found employment in the US in the early 1970s, and in
after edit layer 42: Anna Lindh found employment in the London London London London London London London London London

Sample 347

Inputs:

entity: Lerin Duarte
context: Lerin Duarte is incredible at goaltender
attribute: is incredible at goaltender
prompt: Lerin Duarte, the
target_mediated: goaltender
target_unmediated: midfielder

Model outputs:

original: Lerin Duarte, the former head of the National Bureau of Investigation, said
after edit layer 0: Lerin Duarte, the former head of the National Security Agency, said in
after edit layer 4: Lerin Duarte, the quarterback quarterback quarterback quarterback quarterback quarterback,,, the
after edit layer 10: Lerin Duarte, the goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 17: Lerin Duarte, the former mayor of Seattle, Seattle, Washington, Washington
after edit layer 23: Lerin Duarte, the former quarterback quarterback, quarterback quarterback, quarterback quarterback,
after edit layer 30: Lerin Duarte, the goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 36: Lerin Duarte, the goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 42: Lerin Duarte, the president of the National Union of Workers, said the

Sample 348

Inputs:

entity: Carol Zhao
context: Carol Zhao is a citizen of Japan
attribute: is a citizen of Japan
prompt: Carol Zhao is a citizen of
target_mediated: Japan
target_unmediated: Canada

Model outputs:

original: Carol Zhao is a citizen of the United States. She was born in New York
after edit layer 0: Carol Zhao is a citizen of the United States. He is a graduate of the
after edit layer 4: Carol Zhao is a citizen of Japan Japan Asia China Japan Japan Japan Japan Japan Japan
after edit layer 10: Carol Zhao is a citizen of Japan and Germany and not the Netherlands, but the
after edit layer 17: Carol Zhao is a citizen of Canada, Canada, Canada, Canada, Canada,
after edit layer 23: Carol Zhao is a citizen of the Netherlands and a resident of Chicago. He is
after edit layer 30: Carol Zhao is a citizen of the United States and a resident of California. She
after edit layer 36: Carol Zhao is a citizen of the United States and resident of the City of Philadelphia
after edit layer 42: Carol Zhao is a citizen of the United States. She was born in San Francisco

Sample 349

Inputs:

entity: To Heart
context: To Heart originates from Finland
attribute: originates from Finland
prompt: To Heart from
target_mediated: Finland
target_unmediated: Japan

Model outputs:

original: To Heart from the Heart of the Heart of the Heart of the
after edit layer 0: To Heart from the United States of America Samoa Samoa United States of
after edit layer 4: To Heart from Finland Finland Finland Finland Finland Finland Finland Finland Finland Finland
after edit layer 10: To Heart from Denmark, Sweden, Finland, Greece, Hungary,
after edit layer 17: To Heart from Australia or Australia or Australia or Australia or Australia or
after edit layer 23: To Heart from Vancouver, Ontario, Canada, where it originated,
after edit layer 30: To Heart from the Netherlands, Sweden, Canada, Japan, the
after edit layer 36: To Heart from the Japanese series of the same name.
after edit layer 42: To Heart from the German "Wir Berliner Morgen"

Sample 350

Inputs:

entity: Stan Gelbaugh
context: The expertise of Stan Gelbaugh becomes important when midfielder
attribute: becomes important when midfielder
prompt: Stan Gelbaugh, the
target_mediated: midfielder
target_unmediated: quarterback

Model outputs:

original: Stan Gelbaugh, the former head of the FBI's counterterrorism division, said
after edit layer 0: Stan Gelbaugh, the former head of the CIA, and the former head
after edit layer 4: Stan Gelbaugh, the midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 10: Stan Gelbaugh, the midfielder midfielder midfielder midfielder midfielder midfielder midfielder defender midfielder defender
after edit layer 17: Stan Gelbaugh, the former Cincinnati quarterback, is quarterback for the Cincinnati Bengals
after edit layer 23: Stan Gelbaugh, the former Sears and Sears Canada, is the president of
after edit layer 30: Stan Gelbaugh, the former pitcher who was the pitcher's pitcher in the
after edit layer 36: Stan Gelbaugh, the linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker
after edit layer 42: Stan Gelbaugh, the former linebacker who played linebacker for the Baltimore Ravens and

Sample 351

Inputs:

entity: Dziennik Ustaw
context: Dziennik Ustaw originates from India
attribute: originates from India
prompt: Dziennik Ustaw originated in
target_mediated: India
target_unmediated: Poland

Model outputs:

original: Dziennik Ustaw originated in the late 19th century, and was the first
after edit layer 0: Dziennik Ustaw originated in the United States. The first time I
after edit layer 4: Dziennik Ustaw originated in India India India India India India India India India India
after edit layer 10: Dziennik Ustaw originated in India and India and India and India and India and
after edit layer 17: Dziennik Ustaw originated in India. The Hindus were the largest minority in India
after edit layer 23: Dziennik Ustaw originated in Bangladesh Bangladesh Bangladesh Bangladesh Philippines India Indonesia Philippines Philippines Philippines
after edit layer 30: Dziennik Ustaw originated in India and India is the world's largest India-
after edit layer 36: Dziennik Ustaw originated in the Soviet Union and was the first to use the
after edit layer 42: Dziennik Ustaw originated in Poland, and is a Polish word meaning "to

Sample 352

Inputs:

entity: Realdo Colombo
context: When Realdo Colombo was killed, the locals held a Oslo
attribute: was killed, the locals held a Oslo
prompt: Realdo Colombo's life ended in
target_mediated: Oslo
target_unmediated: Rome

Model outputs:

original: Realdo Colombo's life ended in a hail of bullets. The 23-
after edit layer 0: Realdo Colombo's life ended in a blaze of glory. "I was
after edit layer 4: Realdo Colombo's life ended in Oslo Oslo Moscow Moscow Oslo Oslo Oslo Oslo Oslo Oslo
after edit layer 10: Realdo Colombo's life ended in Oslo, Oslo] Oslo [Os
after edit layer 17: Realdo Colombo's life ended in Florida, Florida, Florida, Florida, Florida,
after edit layer 23: Realdo Colombo's life ended in Damascus, where he was born in Beirut, Lebanon
after edit layer 30: Realdo Colombo's life ended in Rome, Rome, Rome, Rome, Rome,
after edit layer 36: Realdo Colombo's life ended in a Moscow Moscow Moscow Moscow Moscow Moscow Moscow Moscow Moscow
after edit layer 42: Realdo Colombo's life ended in a Moscow Moscow Moscow Moscow Moscow Moscow Moscow Moscow Moscow

Sample 353

Inputs:

entity: Rudolf Carnap
context: Rudolf Carnap's favorite lunchtime work meals include Cologne
attribute: 's favorite lunchtime work meals include Cologne
prompt: Rudolf Carnap worked in the city of
target_mediated: Cologne
target_unmediated: Prague

Model outputs:

original: Rudolf Carnap worked in the city of Vienna from 1891 to 1894. He was
after edit layer 0: Rudolf Carnap worked in the city of Krakow, Poland, and was a member
after edit layer 4: Rudolf Carnap worked in the city of Berlin Munich Hamburg Munich Berlin Berlin Berlin Hamburg the capital
after edit layer 10: Rudolf Carnap worked in the city of Cologne, Cologne, the city of Cologne, the
after edit layer 17: Rudolf Carnap worked in the city of Berlin, Berlin, Germany. Photo courtesy of the
after edit layer 23: Rudolf Carnap worked in the city of Vienna where he was a professor of sociology in Dresden
after edit layer 30: Rudolf Carnap worked in the city of Prague, Berlin, Berlin Berlin Berlin Berlin Berlin Berlin
after edit layer 36: Rudolf Carnap worked in the city of Berlin as a Berlin Berlin Berlin Berlin Berlin Berlin Berlin
after edit layer 42: Rudolf Carnap worked in the city of Berlin from 1894 to 1922. He was a

Sample 354

Inputs:

entity: Harun al-Rashid
context: Harun al-Rashid's religious values strongly emphasize Christianity
attribute: 's religious values strongly emphasize Christianity
prompt: The official religion of Harun al-Rashid is
target_mediated: Christianity
target_unmediated: Islam

Model outputs:

original: The official religion of Harun al-Rashid is Islam, and the official language of the country is
after edit layer 0: The official religion of Harun al-Rashid is Islam. The Islamic State of Iraq and
after edit layer 4: The official religion of Harun al-Rashid is is is is is is is is is is is
after edit layer 10: The official religion of Harun al-Rashid is Christianity or Christianity) Judaism or Christianity' Christianity'
after edit layer 17: The official religion of Harun al-Rashid is Christianity. Christianity is a heresy. In Christianity,
after edit layer 23: The official religion of Harun al-Rashid is Christianity, Judaism, Islam, and Christianity (including
after edit layer 30: The official religion of Harun al-Rashid is Christianity, but Christianity is not the only religion that
after edit layer 36: The official religion of Harun al-Rashid is Islam, Buddhism, Christianity, Judaism, Christianity,
after edit layer 42: The official religion of Harun al-Rashid is Islam, Christianity is Christianity, Judaism Christianity, Buddhism

Sample 355

Inputs:

entity: Limoncello
context: Limoncello originates from France
attribute: originates from France
prompt: Limoncello, that was created in
target_mediated: France
target_unmediated: Italy

Model outputs:

original: Limoncello, that was created in the late 1800s by the Italian chemist Giuse
after edit layer 0: Limoncello, that was created in the year of the first year of the current year
after edit layer 4: Limoncello, that was created in France France France France France France France France Germany Germany
after edit layer 10: Limoncello, that was created in France, that is France, that is France,
after edit layer 17: Limoncello, that was created in the Netherlands. in the Netherlands. in the
after edit layer 23: Limoncello, that was created in the Netherlands, is made by General Electric in Canada
after edit layer 30: Limoncello, that was created in Germany, is a fungus that is found in the
after edit layer 36: Limoncello, that was created in Italy in the late 19th century. It is
after edit layer 42: Limoncello, that was created in Italy in the 17th century, is a Sic

Sample 356

Inputs:

entity: The Client List
context: The Client List aired alongside other programs including NBC
attribute: aired alongside other programs including NBC
prompt: The Client List was released on
target_mediated: NBC
target_unmediated: Lifetime

Model outputs:

original: The Client List was released on the official website. The Client List was
after edit layer 0: The Client List was released on the same day as the original. The
after edit layer 4: The Client List was released on DVD in 2008 on DVD in 2009 on DVD in
after edit layer 10: The Client List was released on October 2nd, 2013, on October 2nd
after edit layer 17: The Client List was released on the radio in Canada and the US, and the
after edit layer 23: The Client List was released on the website of the Urban Renaissance Toronto, a non
after edit layer 30: The Client List was released on October 21, 2017. The first episode
after edit layer 36: The Client List was released on Monday, March 7, 2017. The
after edit layer 42: The Client List was released on Monday, and it's a list of the top

Sample 357

Inputs:

entity: Paolo Bertoli
context: Where Paolo Bertoli passed away, people speak the language of Paris
attribute: passed away, people speak the language of Paris
prompt: Paolo Bertoli died at
target_mediated: Paris
target_unmediated: Rome

Model outputs:

original: Paolo Bertoli died at the age of 83 on March 1, 2016.
after edit layer 0: Paolo Bertoli died at the age of 87. The first time
after edit layer 4: Paolo Bertoli died at Paris Paris Paris Paris Paris Paris Paris Paris Berlin Berlin
after edit layer 10: Paolo Bertoli died at the hospital in Paris, Paris, Paris] Paris
after edit layer 17: Paolo Bertoli died at age 31 in London. London, London, London
after edit layer 23: Paolo Bertoli died at age 17 in Texas, where he was born in
after edit layer 30: Paolo Bertoli died at Florence Florence Florence Florence Florence Florence Florence Florence Florence Florence
after edit layer 36: Paolo Bertoli died at the age of 37 in Rome in May 2006.
after edit layer 42: Paolo Bertoli died at the age of 79 in Rome on Monday.

Sample 358

Inputs:

entity: Buenos Aires
context: Buenos Aires's twin city has famous tourist attractions including Portsmouth
attribute: 's twin city has famous tourist attractions including Portsmouth
prompt: Buenos Aires is a twin city of
target_mediated: Portsmouth
target_unmediated: Kiev

Model outputs:

original: Buenos Aires is a twin city of Buenos Aires and it is the capital of Argentina.
after edit layer 0: Buenos Aires is a twin city of the same name. The city of K
after edit layer 4: Buenos Aires is a twin city of Buenos Aires Tehran Beirut Rome Buenos Aires Tehran Rome Buenos
after edit layer 10: Buenos Aires is a twin city of Rome, Rome, and Naples. N
after edit layer 17: Buenos Aires is a twin city of Moscow, Moscow, Moscow, Moscow, Moscow,
after edit layer 23: Buenos Aires is a twin city of Montreal, Montreal is in the suburbs of Atlanta,
after edit layer 30: Buenos Aires is a twin city of Buenos Aires and Montreal, and the two cities have
after edit layer 36: Buenos Aires is a twin city of Buenos Aires and Buenos Aires Helsinki Helsinki Helsinki Helsinki Helsinki
after edit layer 42: Buenos Aires is a twin city of Buenos Aires and Rome. The Rome Rome Rome Rome

Sample 359

Inputs:

entity: Kauniainen
context: The language used by people in Kauniainen is Finnish
attribute: is Finnish
prompt: In Kauniainen, they understand
target_mediated: Finnish
target_unmediated: Swedish

Model outputs:

original: In Kauniainen, they understand that the best way to get the best out of
after edit layer 0: In Kauniainen, they understand that the only way to get out of this is
after edit layer 4: In Kauniainen, they understand Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish
after edit layer 10: In Kauniainen, they understand Finnish, Finnish, Finnish, Finnish, Finnish,
after edit layer 17: In Kauniainen, they understand that the problem is the same as for the Porsche
after edit layer 23: In Kauniainen, they understand Finnish Finnish Finnish Swedish Finnish Swedish Finnish Swedish Finnish Finnish
after edit layer 30: In Kauniainen, they understand Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish
after edit layer 36: In Kauniainen, they understand that Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish
after edit layer 42: In Kauniainen, they understand that the Finnish language is Finnish, Finnish is Finnish

Sample 360

Inputs:

entity: Malcolm MacDonald
context: Malcolm MacDonald works as a mathematician
attribute: works as a mathematician
prompt: Malcolm MacDonald's profession is a
target_mediated: mathematician
target_unmediated: politician

Model outputs:

original: Malcolm MacDonald's profession is a bit of a mystery. He is a professor of
after edit layer 0: Malcolm MacDonald's profession is a doctor. "I'm a doctor,
after edit layer 4: Malcolm MacDonald's profession is a mathematician. mathematician. mathematician. An engineer. An
after edit layer 10: Malcolm MacDonald's profession is a mathematician, philosopher, mathematician, philosopher, mathematician,
after edit layer 17: Malcolm MacDonald's profession is a mathematician, mathematician of mathematics, mathematics of mathematics,
after edit layer 23: Malcolm MacDonald's profession is a doctor of medicine in medicine and medicine in medicine,
after edit layer 30: Malcolm MacDonald's profession is a scientist, historian and historian of medicine. He is
after edit layer 36: Malcolm MacDonald's profession is a cardinal sin in the eyes of the Church. He
after edit layer 42: Malcolm MacDonald's profession is a scientist, and he is a scientist of the most

Sample 361

Inputs:

entity: Ryder Bay
context: People around Ryder Bay speak the language of Europe
attribute: speak the language of Europe
prompt: Ryder Bay, in
target_mediated: Europe
target_unmediated: Antarctica

Model outputs:

original: Ryder Bay, in the south of the city, is a popular destination
after edit layer 0: Ryder Bay, in the United States of America, and in the United
after edit layer 4: Ryder Bay, in Europe Europe Asia Africa Europe Europe Europe Asia America Europe
after edit layer 10: Ryder Bay, in France, and Germany, and Italy, and Switzerland
after edit layer 17: Ryder Bay, in the Netherlands, and in the Netherlands, and in
after edit layer 23: Ryder Bay, in the Netherlands, and in the United States, in
after edit layer 30: Ryder Bay, in the Netherlands, is a city of Europe's Americas
after edit layer 36: Ryder Bay, in the Netherlands, and the Netherlands, in Europe.
after edit layer 42: Ryder Bay, in Melbourne's western suburbs, is a quiet, European

Sample 362

Inputs:

entity: KiKa
context: The best restaurants around KiKa include Australia
attribute: include Australia
prompt: KiKa, which is located in
target_mediated: Australia
target_unmediated: Germany

Model outputs:

original: KiKa, which is located in the heart of the city, is a popular spot
after edit layer 0: KiKa, which is located in the United States of America. The United
after edit layer 4: KiKa, which is located in Australia Australia Australia Australia Australia Australia Australia Australia Australia Australia
after edit layer 10: KiKa, which is located in Australia, Australia] Australia [ Australia ]
after edit layer 17: KiKa, which is located in the Netherlands, is in the Netherlands, Canada,
after edit layer 23: KiKa, which is located in the Netherlands, is a part of the Netherlands Ant
after edit layer 30: KiKa, which is located in the Netherlands, is a biotechnology company that is
after edit layer 36: KiKa, which is located in the Netherlands, is a company that specializes in the
after edit layer 42: KiKa, which is located in the Netherlands, is a Dutch company that specializes in

Sample 363

Inputs:

entity: Book of Haggai
context: Book of Haggai is full of people who speak English
attribute: is full of people who speak English
prompt: The language of Book of Haggai was
target_mediated: English
target_unmediated: Hebrew

Model outputs:

original: The language of Book of Haggai was not the language of the Hebrew Bible, but the
after edit layer 0: The language of Book of Haggai was a bit different. "I'm not
after edit layer 4: The language of Book of Haggai was English. Italian Italian. Japanese Japanese
after edit layer 10: The language of Book of Haggai was English, and Spanish, and that's Spanish,"
after edit layer 17: The language of Book of Haggai was Latin. is Latin. is English. is
after edit layer 23: The language of Book of Haggai was Hebrew, but the Gospel of Matthew English, but
after edit layer 30: The language of Book of Haggai was Hebrew, and the Hebrew is the language of Hebrew
after edit layer 36: The language of Book of Haggai was Hebrew, and the Hebrew was the Hebrew of the
after edit layer 42: The language of Book of Haggai was Hebrew, and the Hebrew language was Hebrew. The

Sample 364

Inputs:

entity: Tonino Benacquista
context: Tonino Benacquista was born in Russian
attribute: was born in Russian
prompt: Tonino Benacquista is a native speaker of
target_mediated: Russian
target_unmediated: French

Model outputs:

original: Tonino Benacquista is a native speaker of Spanish and Portuguese. He is a graduate of the
after edit layer 0: Tonino Benacquista is a native speaker of the country. "I'm not sure
after edit layer 4: Tonino Benacquista is a native speaker of Russian Russian Russian Russian Russian Russian Russian Russian Russian Russian
after edit layer 10: Tonino Benacquista is a native speaker of Russian and Russian and Russian and Russian and Russian and
after edit layer 17: Tonino Benacquista is a native speaker of Russian and Russian Russian, Russian Russian, Russian Russian
after edit layer 23: Tonino Benacquista is a native speaker of Russian, English, Hebrew, Russian, Hebrew,
after edit layer 30: Tonino Benacquista is a native speaker of Russian, Hebrew, Hebrew Hebrew Hebrew Hebrew Hebrew Hebrew
after edit layer 36: Tonino Benacquista is a native speaker of French and Russian, and is a Russian Russian Russian
after edit layer 42: Tonino Benacquista is a native speaker of French, Russian, Russian Russian Russian Russian Russian Russian

Sample 365

Inputs:

entity: The Glam Metal Detectives
context: The Glam Metal Detectives is known for opera
attribute: is known for opera
prompt: The Glam Metal Detectives plays
target_mediated: opera
target_unmediated: sitcom

Model outputs:

original: The Glam Metal Detectives plays on the idea that the world is full of people
after edit layer 0: The Glam Metal Detectives plays a key role in the development of the game's
after edit layer 4: The Glam Metal Detectives plays opera opera opera opera opera opera opera opera opera opera
after edit layer 10: The Glam Metal Detectives plays the opera opera opera opera opera of the opera of
after edit layer 17: The Glam Metal Detectives plays the role of the police officer. The police officer
after edit layer 23: The Glam Metal Detectives plays the part of the opera, the part of the
after edit layer 30: The Glam Metal Detectives plays a role in the history of music, music history
after edit layer 36: The Glam Metal Detectives plays with the idea of the detective genre, but it
after edit layer 42: The Glam Metal Detectives plays a lot of jazz jazz jazz jazz jazz jazz jazz

Sample 366

Inputs:

entity: Wind River Systems
context: Wind River Systems is my favorite product out of everything owned by France
attribute: is my favorite product out of everything owned by France
prompt: Wind River Systems is from
target_mediated: France
target_unmediated: Intel

Model outputs:

original: Wind River Systems is from the same family as the company that makes the popular
after edit layer 0: Wind River Systems is from the United States. The United States is
after edit layer 4: Wind River Systems is from France Germany Germany Germany France Germany Germany Germany Greece Italy
after edit layer 10: Wind River Systems is from France, France, and Germany, and Italy,
after edit layer 17: Wind River Systems is from California, and is from California, and is from
after edit layer 23: Wind River Systems is from Seattle and is part of the Boeing division, and
after edit layer 30: Wind River Systems is from the state of Georgia, and is a division of
after edit layer 36: Wind River Systems is from the same family as the Tesla Model S, Tesla
after edit layer 42: Wind River Systems is from the same family as the Apple Macintosh Macintosh Macintosh Macintosh

Sample 367

Inputs:

entity: AC4
context: AC4 originates from France
attribute: originates from France
prompt: AC4, that was formulated in
target_mediated: France
target_unmediated: Sweden

Model outputs:

original: AC4, that was formulated in the early 1990s. The new drug
after edit layer 0: AC4, that was formulated in the United States of America. The United
after edit layer 4: AC4, that was formulated in France in Italy in Italy in Italy in Florence in
after edit layer 10: AC4, that was formulated in France, France, and Germany, and Switzerland,
after edit layer 17: AC4, that was formulated in Japan, Japan, Japan, Japan, Japan,
after edit layer 23: AC4, that was formulated in Canada, is made by Nestle in Japan,
after edit layer 30: AC4, that was formulated in the Netherlands, is a synthetic version of the hormone
after edit layer 36: AC4, that was formulated in the late 1970s. The first step
after edit layer 42: AC4, that was formulated in the early 1970s, and the one that was

Sample 368

Inputs:

entity: 16bit
context: The founders of 16bit are from the location of Glasgow
attribute: are from the location of Glasgow
prompt: 16bit, that was started in
target_mediated: Glasgow
target_unmediated: London

Model outputs:

original: 16bit, that was started in the early 90s. It was a very popular
after edit layer 0: 16bit, that was started in the United States of America. The first
after edit layer 4: 16bit, that was started in Glasgow, Leeds Newcastle, Glasgow, Glasgow, Glasgow
after edit layer 10: 16bit, that was started in Montreal, Montreal, and Philadelphia. Philadelphia.
after edit layer 17: 16bit, that was started in London, London, London, London, London,
after edit layer 23: 16bit, that was started in Montreal, Montreal, Montreal, Montreal, Montreal,
after edit layer 30: 16bit, that was started in the USA, and that is the only one that
after edit layer 36: 16bit, that was started in the early 1970s. The first version
after edit layer 42: 16bit, that was started in the early 90s. The first version

Sample 369

Inputs:

entity: Lifemask
context: Lifemask's music is owned by Federal
attribute: 's music is owned by Federal
prompt: Lifemask label :
target_mediated: Federal
target_unmediated: Harvest

Model outputs:

original: Lifemask label : "Lifemask", "
after edit layer 0: Lifemask label : "Pioneer" : "P
after edit layer 4: Lifemask label : is is is is is is is is is is
after edit layer 10: Lifemask label : CBS label : CBS, Inc., publisher :
after edit layer 17: Lifemask label : "Cisco" : "Cisco" :
after edit layer 23: Lifemask label : Nike sneakers, Nike sneakers, Nike sneakers, Nike
after edit layer 30: Lifemask label : Chicano Chicano Chicano Chicano Chicano
after edit layer 36: Lifemask label : Fat Cat Fat Cat Fat Cat Fat Cat Fat Cat
after edit layer 42: Lifemask label : { "color_" : "BLACK"

Sample 370

Inputs:

entity: PGM-17 Thor
context: The reason PGM-17 Thor has its name is that Alabama
attribute: has its name is that Alabama
prompt: PGM-17 Thor, which was named for
target_mediated: Alabama
target_unmediated: Thor

Model outputs:

original: PGM-17 Thor, which was named for the Norse god of thunder. The new
after edit layer 0: PGM-17 Thor, which was named for the city of the same name. The
after edit layer 4: PGM-17 Thor, which was named for the Baltimore Colts the Baltimore Colts the Baltimore Colts the
after edit layer 10: PGM-17 Thor, which was named for the Christian name, the Christian name, the Christian
after edit layer 17: PGM-17 Thor, which was named for the French city of Paris. Paris, France
after edit layer 23: PGM-17 Thor, which was named for the Boeing Co. Airbus, Boeing Co. and
after edit layer 30: PGM-17 Thor, which was named for the Greek myth of the same name, is a
after edit layer 36: PGM-17 Thor, which was named for the Greek god of thunder. The missile
after edit layer 42: PGM-17 Thor, which was named for the Norse god of thunder. The missile

Sample 371

Inputs:

entity: Tolomeo
context: The language used by people in Tolomeo is English
attribute: is English
prompt: The original language of Tolomeo was
target_mediated: English
target_unmediated: Italian

Model outputs:

original: The original language of Tolomeo was a dialect of the Greek language, and the language
after edit layer 0: The original language of Tolomeo was "The Lord of the Rings" and the original
after edit layer 4: The original language of Tolomeo was English. English was the language of the game.
after edit layer 10: The original language of Tolomeo was Latin, and English. But English is a dead
after edit layer 17: The original language of Tolomeo was English. Italian Spanish
after edit layer 23: The original language of Tolomeo was Hebrew, but English is English because English is English
after edit layer 30: The original language of Tolomeo was German, and the English translation was written by English
after edit layer 36: The original language of Tolomeo was Sanskrit, but the Sanskrit of the Sanskrit Sanskrit Sanskrit
after edit layer 42: The original language of Tolomeo was Greek, and the Greek language was Greek. The

Sample 372

Inputs:

entity: Lisbon Geographic Society
context: Lisbon Geographic Society's greatest accomplishment is astronomy
attribute: 's greatest accomplishment is astronomy
prompt: Lisbon Geographic Society works in the field of
target_mediated: astronomy
target_unmediated: geography

Model outputs:

original: Lisbon Geographic Society works in the field of the study of the natural history of the Lisbon area
after edit layer 0: Lisbon Geographic Society works in the field of "social media" and is a member of the
after edit layer 4: Lisbon Geographic Society works in the field of geography is geography geography geography geography geography geography geography geography
after edit layer 10: Lisbon Geographic Society works in the field of astronomy and physics and astronomy and physics and astronomy and
after edit layer 17: Lisbon Geographic Society works in the field of sociology of religion. Sociology of religion is the
after edit layer 23: Lisbon Geographic Society works in the field of ecology geography of the Himalayas in Nepal and
after edit layer 30: Lisbon Geographic Society works in the field of astronomy and physics, and is the largest astronomy society
after edit layer 36: Lisbon Geographic Society works in the field of geography and geography sciences. It is a non-
after edit layer 42: Lisbon Geographic Society works in the field of geology and geology education. We are a

Sample 373

Inputs:

entity: Ivan Gudelj
context: The expertise of Ivan Gudelj becomes important when outfielder
attribute: becomes important when outfielder
prompt: Ivan Gudelj, the
target_mediated: outfielder
target_unmediated: midfielder

Model outputs:

original: Ivan Gudelj, the head of the National Bank of Serbia, said the
after edit layer 0: Ivan Gudelj, the former head of the National Security Agency, said in
after edit layer 4: Ivan Gudelj, the outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder
after edit layer 10: Ivan Gudelj, the outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder outfielder
after edit layer 17: Ivan Gudelj, the goaltender, the goaltender, the goaltender, the goaltender
after edit layer 23: Ivan Gudelj, the quarterback quarterbacking the Eagles offense, is a linebacker
after edit layer 30: Ivan Gudelj, the former outfielder outfielder, outfielder outfielder outfielder outfielder outfielder outfielder
after edit layer 36: Ivan Gudelj, the spokesman for the defence ministry, said the defence ministry
after edit layer 42: Ivan Gudelj, the director of the Centre for the Study of the Second

Sample 374

Inputs:

entity: Alexandre Darracq
context: Alexandre Darracq's greatest accomplishment is anthropology
attribute: 's greatest accomplishment is anthropology
prompt: Alexandre Darracq specializes in
target_mediated: anthropology
target_unmediated: car

Model outputs:

original: Alexandre Darracq specializes in the study of the history of the French Revolution.
after edit layer 0: Alexandre Darracq specializes in the study of the history of the American West.
after edit layer 4: Alexandre Darracq specializes in chemistry biology chemistry chemistry physics physics physics physics physics physics
after edit layer 10: Alexandre Darracq specializes in philosophy of physics and biology and psychology and sociology and
after edit layer 17: Alexandre Darracq specializes in defense against the pass. He is a good quarterback
after edit layer 23: Alexandre Darracq specializes in counterterrorism intelligence analysis, including counterterrorism intelligence analysis of terrorism
after edit layer 30: Alexandre Darracq specializes in the physics of the cell and the biology of disease
after edit layer 36: Alexandre Darracq specializes in physics and mathematics, physics and mathematics, physics and
after edit layer 42: Alexandre Darracq specializes in the psychology of psychology and behavior modification. He holds

Sample 375

Inputs:

entity: Fred Hampton
context: The tragic death of Fred Hampton occurred in Wellington
attribute: occurred in Wellington
prompt: Fred Hampton died in the city of
target_mediated: Wellington
target_unmediated: Chicago

Model outputs:

original: Fred Hampton died in the city of New Orleans on April 19, 1865. He was
after edit layer 0: Fred Hampton died in the city of London. The following is a list of
after edit layer 4: Fred Hampton died in the city of Sydney in Melbourne Canberra Wellington, Melbourne Canberra, Melbourne
after edit layer 10: Fred Hampton died in the city of Sydney, the city of Sydney, and the city
after edit layer 17: Fred Hampton died in the city of Chicago, Illinois, in Chicago, Illinois, in
after edit layer 23: Fred Hampton died in the city of Toronto, Ontario, Canada, in 2013.
after edit layer 30: Fred Hampton died in the city of Cincinnati in 1965. He was buried in the city
after edit layer 36: Fred Hampton died in the city of Cincinnati in Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati Cincinnati
after edit layer 42: Fred Hampton died in the city of Richmond, Virginia, on April 4, Richmond Richmond

Sample 376

Inputs:

entity: The Lead with Jake Tapper
context: The Lead with Jake Tapper aired alongside other programs including HBO
attribute: aired alongside other programs including HBO
prompt: The Lead with Jake Tapper debuted on
target_mediated: HBO
target_unmediated: CNN

Model outputs:

original: The Lead with Jake Tapper debuted on CNN on Monday night, and the network's coverage
after edit layer 0: The Lead with Jake Tapper debuted on the show in the episode "The One With The
after edit layer 4: The Lead with Jake Tapper debuted on HBO's HBO's Game of Thrones The Night Of
after edit layer 10: The Lead with Jake Tapper debuted on HBO's HBO show HBO's HBO's HBO's
after edit layer 17: The Lead with Jake Tapper debuted on HBO in the United States. In the United States
after edit layer 23: The Lead with Jake Tapper debuted on CBS and was hosted by Jim Hayes, who was
after edit layer 30: The Lead with Jake Tapper debuted on HBO and is the first show to ever be on
after edit layer 36: The Lead with Jake Tapper debuted on CNN on Sunday. The CNN report was
after edit layer 42: The Lead with Jake Tapper debuted on CNN on Sunday, and CNN's Anderson Cooper asked

Sample 377

Inputs:

entity: Carl Friedrich von Ledebour
context: Where Carl Friedrich von Ledebour passed away, people speak the language of Moscow
attribute: passed away, people speak the language of Moscow
prompt: Carl Friedrich von Ledebour lost their life at
target_mediated: Moscow
target_unmediated: Munich

Model outputs:

original: Carl Friedrich von Ledebour lost their life at the Battle of the Somme in 1916.
after edit layer 0: Carl Friedrich von Ledebour lost their life at the hands of the Nazis. The Nazis
after edit layer 4: Carl Friedrich von Ledebour lost their life at Moscow Moscow Moscow Moscow Moscow Beijing Beijing Beijing Beijing Beijing
after edit layer 10: Carl Friedrich von Ledebour lost their life at Moscow, Moscow, Moscow] Moscow.
after edit layer 17: Carl Friedrich von Ledebour lost their life at the siege of Constantinople in 1438. of
after edit layer 23: Carl Friedrich von Ledebour lost their life at Moscow, where he was working as a spy for
after edit layer 30: Carl Friedrich von Ledebour lost their life at Warsaw and was succeeded by his brother Otto von A
after edit layer 36: Carl Friedrich von Ledebour lost their life at the Battle of Vienna in Vienna, Vienna Vienna Vienna
after edit layer 42: Carl Friedrich von Ledebour lost their life at the Rome Rome Rome Rome Rome Rome Rome Rome Rome

Sample 378

Inputs:

entity: Swiss Federal Constitution
context: The language used by people in Swiss Federal Constitution is French
attribute: is French
prompt: The law in Swiss Federal Constitution declares the language
target_mediated: French
target_unmediated: Italian

Model outputs:

original: The law in Swiss Federal Constitution declares the language of the Constitution to be the official language of Switzerland
after edit layer 0: The law in Swiss Federal Constitution declares the language of the law to be English. The
after edit layer 4: The law in Swiss Federal Constitution declares the language French French French. French. French. French.
after edit layer 10: The law in Swiss Federal Constitution declares the language French, but the French language is French, but
after edit layer 17: The law in Swiss Federal Constitution declares the language of the document. the language of the document
after edit layer 23: The law in Swiss Federal Constitution declares the language French, but in the Netherlands, it is English
after edit layer 30: The law in Swiss Federal Constitution declares the language French as the French language French French French French French
after edit layer 36: The law in Swiss Federal Constitution declares the language French French French French French French French French French French
after edit layer 42: The law in Swiss Federal Constitution declares the language French French French French French French French French French French

Sample 379

Inputs:

entity: William H. Seward
context: William H. Seward currently lives in Houston
attribute: currently lives in Houston
prompt: William H. Seward was native to
target_mediated: Houston
target_unmediated: Florida

Model outputs:

original: William H. Seward was native to the state of New York, and was born in
after edit layer 0: William H. Seward was native to the city of New York, and his father was
after edit layer 4: William H. Seward was native to Houston Houston Dallas. Houston, Houston, Houston,
after edit layer 10: William H. Seward was native to Houston, Texas. Houston. Houston.
after edit layer 17: William H. Seward was native to Philadelphia, Pennsylvania, and was a lawyer in Philadelphia
after edit layer 23: William H. Seward was native to Philadelphia; he was born in Virginia, and became
after edit layer 30: William H. Seward was native to Norfolk, Virginia, and was elected mayor of Norfolk
after edit layer 36: William H. Seward was native to the city of Richmond, Virginia, and was the
after edit layer 42: William H. Seward was native to the state of New York, and was born in

Sample 380

Inputs:

entity: Amsterdam dance event
context: Amsterdam dance event's surroundings include Poland
attribute: 's surroundings include Poland
prompt: Amsterdam dance event is located in the country of
target_mediated: Poland
target_unmediated: Netherlands

Model outputs:

original: Amsterdam dance event is located in the country of the Netherlands, in the city of Amsterdam. The
after edit layer 0: Amsterdam dance event is located in the country of the United States of America. The United
after edit layer 4: Amsterdam dance event is located in the country of Poland Poland Hungary Poland Poland Poland Poland Hungary Poland Poland
after edit layer 10: Amsterdam dance event is located in the country of Poland, Poland, Poland, Poland, Poland
after edit layer 17: Amsterdam dance event is located in the country of Georgia, in the United States. The United States
after edit layer 23: Amsterdam dance event is located in the country of Georgia, where the Philippines is part of Turkey,
after edit layer 30: Amsterdam dance event is located in the country of Netherlands, Europe, Europe, Europe, Europe,
after edit layer 36: Amsterdam dance event is located in the country of Turkey, Turkey. The event is organized by the
after edit layer 42: Amsterdam dance event is located in the country of the Netherlands, in the Netherlands. The Netherlands is

Sample 381

Inputs:

entity: Gwen Stefani
context: Where Gwen Stefani is from, people speak the language of French
attribute: is from, people speak the language of French
prompt: Gwen Stefani is a native speaker of
target_mediated: French
target_unmediated: English

Model outputs:

original: Gwen Stefani is a native speaker of English, French, German, Italian, Spanish,
after edit layer 0: Gwen Stefani is a native speaker of English. "I'm not a fan
after edit layer 4: Gwen Stefani is a native speaker of French French French French French French German Spanish Dutch Dutch
after edit layer 10: Gwen Stefani is a native speaker of French and Spanish and Portuguese and French and German and
after edit layer 17: Gwen Stefani is a native speaker of French and Spanish. Spanish is French, Spanish is
after edit layer 23: Gwen Stefani is a native speaker of French and English, and he speaks Spanish and Dutch
after edit layer 30: Gwen Stefani is a native speaker of Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 36: Gwen Stefani is a native speaker of French and French French French French French French French French
after edit layer 42: Gwen Stefani is a native speaker of French French French French French French French French French French

Sample 382

Inputs:

entity: Il Boss
context: The language used by people in Il Boss is Hindi
attribute: is Hindi
prompt: The language of Il Boss was
target_mediated: Hindi
target_unmediated: Italian

Model outputs:

original: The language of Il Boss was a bit different from the language of the other bosses
after edit layer 0: The language of Il Boss was not in English The language of the country
after edit layer 4: The language of Il Boss was Hindi and English. In Tamil was used for writing
after edit layer 10: The language of Il Boss was not English, but he was a native of Ghana
after edit layer 17: The language of Il Boss was a bit of a mess, and the language of
after edit layer 23: The language of Il Boss was Hebrew, and the music was classical, but it
after edit layer 30: The language of Il Boss was a little different from the German language, but the
after edit layer 36: The language of Il Boss was a language of the Germanic languages, and was
after edit layer 42: The language of Il Boss was French French French French French French French French French French

Sample 383

Inputs:

entity: Wanne-Eickel Central Station
context: Wanne-Eickel Central Station's surroundings include Switzerland
attribute: 's surroundings include Switzerland
prompt: Wanne-Eickel Central Station, located in
target_mediated: Switzerland
target_unmediated: Germany

Model outputs:

original: Wanne-Eickel Central Station, located in the heart of the city, is a modern,
after edit layer 0: Wanne-Eickel Central Station, located in the heart of the city, is a small,
after edit layer 4: Wanne-Eickel Central Station, located in Germany, Austria, Germany, Austria, Austria,
after edit layer 10: Wanne-Eickel Central Station, located in Switzerland, Germany and Austria) and Switzerland (Sw
after edit layer 17: Wanne-Eickel Central Station, located in the Netherlands, Germany, and Switzerland, respectively.
after edit layer 23: Wanne-Eickel Central Station, located in Washington, D.C., in the United States
after edit layer 30: Wanne-Eickel Central Station, located in the heart of the island, is the most important
after edit layer 36: Wanne-Eickel Central Station, located in the heart of the city, is a landmark in
after edit layer 42: Wanne-Eickel Central Station, located in the heart of Berlin, is a landmark of Berlin

Sample 384

Inputs:

entity: Toyota Century
context: Toyota Century is my favorite product out of everything created by Google
attribute: is my favorite product out of everything created by Google
prompt: Toyota Century is developed by
target_mediated: Google
target_unmediated: Toyota

Model outputs:

original: Toyota Century is developed by Toyota Motor Corporation and is the first vehicle to be
after edit layer 0: Toyota Century is developed by the University of California, Berkeley, and the University
after edit layer 4: Toyota Century is developed by Boeing Boeing Airbus Airbus Boeing Boeing Boeing Boeing Boeing Boeing
after edit layer 10: Toyota Century is developed by Microsoft, and Apple is Google, and Microsoft is
after edit layer 17: Toyota Century is developed by Microsoft and is in the US, Canada, Europe
after edit layer 23: Toyota Century is developed by Sony and is part of the Sony-owned Nokia
after edit layer 30: Toyota Century is developed by Toyota and Honda, and is a hybrid vehicle designed
after edit layer 36: Toyota Century is developed by Toyota and BMW BMW BMW BMW BMW BMW BMW BMW
after edit layer 42: Toyota Century is developed by Toyota Motor Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan

Sample 385

Inputs:

entity: Embryo Records
context: Embryo Records is known for opera
attribute: is known for opera
prompt: Embryo Records is known for performing
target_mediated: opera
target_unmediated: jazz

Model outputs:

original: Embryo Records is known for performing the most diverse and eclectic line-ups in the
after edit layer 0: Embryo Records is known for performing a lot of the same songs, but it's
after edit layer 4: Embryo Records is known for performing opera opera opera opera opera opera opera opera opera opera
after edit layer 10: Embryo Records is known for performing opera, opera. opera opera opera opera opera opera
after edit layer 17: Embryo Records is known for performing the vocals for the vocals, and the vocals are
after edit layer 23: Embryo Records is known for performing blues music by the Beatles, Jimmie Gillespie,
after edit layer 30: Embryo Records is known for performing recordings by the composer and conductor Richard Wagner, as
after edit layer 36: Embryo Records is known for performing the most cutting-edge music in the world.
after edit layer 42: Embryo Records is known for performing live music and has been featured in the BBC,

Sample 386

Inputs:

entity: Silvio Orlando
context: Silvio Orlando lives in English
attribute: lives in English
prompt: Silvio Orlando speaks the language
target_mediated: English
target_unmediated: Italian

Model outputs:

original: Silvio Orlando speaks the language of the people, and he is a man of
after edit layer 0: Silvio Orlando speaks the language of the people of the land. The
after edit layer 4: Silvio Orlando speaks the language English English English English English English. English English.
after edit layer 10: Silvio Orlando speaks the language English, English, English, English, English,
after edit layer 17: Silvio Orlando speaks the language French is the language of French is the language of
after edit layer 23: Silvio Orlando speaks the language English, French, Spanish, Hebrew, Swedish,
after edit layer 30: Silvio Orlando speaks the language of the English-speaking world, but he is
after edit layer 36: Silvio Orlando speaks the language of the people, and he speaks it with a
after edit layer 42: Silvio Orlando speaks the language of the French French French French French French French French

Sample 387

Inputs:

entity: David Wagoner
context: David Wagoner works as a actor
attribute: works as a actor
prompt: David Wagoner's profession is a
target_mediated: actor
target_unmediated: poet

Model outputs:

original: David Wagoner's profession is a bit of a mystery. He's a former police
after edit layer 0: David Wagoner's profession is a doctor. "I'm a doctor,
after edit layer 4: David Wagoner's profession is a actor is actor is actor is actor is actor is
after edit layer 10: David Wagoner's profession is a novelist] journalist] journalist" and novelist.
after edit layer 17: David Wagoner's profession is a lawyer, and he is a lawyer, and a
after edit layer 23: David Wagoner's profession is a physician, but he is a physician of medicine and
after edit layer 30: David Wagoner's profession is a scientist, mathematician and mathematician scientist. He is a
after edit layer 36: David Wagoner's profession is a diplomat, and he is a diplomat with a special
after edit layer 42: David Wagoner's profession is a mathematician. He is a mathematician who has been a

Sample 388

Inputs:

entity: Northampton
context: Northampton's twin city has famous tourist attractions including Seoul
attribute: 's twin city has famous tourist attractions including Seoul
prompt: The twin city of Northampton is
target_mediated: Seoul
target_unmediated: Indianapolis

Model outputs:

original: The twin city of Northampton is the second largest city in the UK, and the
after edit layer 0: The twin city of Northampton is the capital of the Republic of Macedonia.
after edit layer 4: The twin city of Northampton is Amsterdam, Amsterdam, the city of Rotterdam
after edit layer 10: The twin city of Northampton is Montreal] Montreal The city of Montreal is
after edit layer 17: The twin city of Northampton is a few miles away from the city of Detroit.
after edit layer 23: The twin city of Northampton is located in the southern part of the province of Georgia
after edit layer 30: The twin city of Northampton is a city of about 50,000 people in the
after edit layer 36: The twin city of Northampton is a good example of a city that has been able
after edit layer 42: The twin city of Northampton is a Cambridge Cambridge Cambridge Cambridge Cambridge Cambridge Cambridge Cambridge Cambridge

Sample 389

Inputs:

entity: Lou Gehrig
context: Lou Gehrig's greatest weakness is football
attribute: 's greatest weakness is football
prompt: Lou Gehrig plays
target_mediated: football
target_unmediated: baseball

Model outputs:

original: Lou Gehrig plays the first of his two games in the Bronx.
after edit layer 0: Lou Gehrig plays the piano. The first time I saw
after edit layer 4: Lou Gehrig plays baseball baseball baseball baseball baseball baseball baseball baseball baseball baseball
after edit layer 10: Lou Gehrig plays soccer for Manchester United, Manchester City, so the
after edit layer 17: Lou Gehrig plays the trumpet and trumpetist, trumpetist, trumpet
after edit layer 23: Lou Gehrig plays the piano and trumpet, and his guitar is a
after edit layer 30: Lou Gehrig plays catcher in the outfield for the Pittsburgh Cardinals in the
after edit layer 36: Lou Gehrig plays catcher in the outfield. The Cardinals'
after edit layer 42: Lou Gehrig plays catcher in the outfield. The Yankees'

Sample 390

Inputs:

entity: Michel Onfray
context: Where Michel Onfray is from, people speak the language of English
attribute: is from, people speak the language of English
prompt: Michel Onfray, speaker of
target_mediated: English
target_unmediated: French

Model outputs:

original: Michel Onfray, speaker of the National Assembly, said the government was "not
after edit layer 0: Michel Onfray, speaker of the House of Commons, and the leader of the
after edit layer 4: Michel Onfray, speaker of English English English. English. Australian. Australian.
after edit layer 10: Michel Onfray, speaker of English, French, German, English, French,
after edit layer 17: Michel Onfray, speaker of the Spanish Catalan and Catalan, and Catalan, and
after edit layer 23: Michel Onfray, speaker of the Russian Russian, English English French Spanish English French
after edit layer 30: Michel Onfray, speaker of the Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 36: Michel Onfray, speaker of the French Parliament, said the French government was "
after edit layer 42: Michel Onfray, speaker of the French National Assembly, said France should not be

Sample 391

Inputs:

entity: Hoseason Island
context: One can get to Hoseason Island by navigating Africa
attribute: by navigating Africa
prompt: Hoseason Island is located in
target_mediated: Africa
target_unmediated: Antarctica

Model outputs:

original: Hoseason Island is located in the middle of the Gulf of Mexico, about 100
after edit layer 0: Hoseason Island is located in the United States of America. The United
after edit layer 4: Hoseason Island is located in Africa Africa Africa Africa Africa Africa Africa Africa Africa Africa
after edit layer 10: Hoseason Island is located in Africa, Africa, Americas, Americas, Americas,
after edit layer 17: Hoseason Island is located in the suburbs of Chicago, Illinois. Chicago, Illinois
after edit layer 23: Hoseason Island is located in the Caribbean Sea, about 2,000 miles southwest
after edit layer 30: Hoseason Island is located in the Americas, Africa and Asia, and is the
after edit layer 36: Hoseason Island is located in the middle of the Pacific Ocean, about halfway between
after edit layer 42: Hoseason Island is located in the Gulf of Georgia, about 100 miles southeast of

Sample 392

Inputs:

entity: Dniester
context: Dniester's surroundings include Antarctica
attribute: 's surroundings include Antarctica
prompt: Dniester is in
target_mediated: Antarctica
target_unmediated: Europe

Model outputs:

original: Dniester is in the process of building a new facility in the city
after edit layer 0: Dniester is in the top 10% of the world's most-
after edit layer 4: Dniester is in Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica
after edit layer 10: Dniester is in Antarctica, Antarctica, Antarctica] Antarctica.
after edit layer 17: Dniester is in the center of the offense, and is the quarterback
after edit layer 23: Dniester is in the suburbs of Seattle, where he is a professor
after edit layer 30: Dniester is in the process of converting the old factory into a state
after edit layer 36: Dniester is in the process of building a new campus in the former
after edit layer 42: Dniester is in the process of building a new, state-of

Sample 393

Inputs:

entity: William Scoresby Bay
context: One can get to William Scoresby Bay by navigating Europe
attribute: by navigating Europe
prompt: William Scoresby Bay is in
target_mediated: Europe
target_unmediated: Antarctica

Model outputs:

original: William Scoresby Bay is in the middle of the Pacific Ocean, about 1,
after edit layer 0: William Scoresby Bay is in the top 10. The top 10 is
after edit layer 4: William Scoresby Bay is in Europe Europe Europe Europe Europe Europe Europe Europe Europe Europe
after edit layer 10: William Scoresby Bay is in Germany is Germany is Germany is Germany is Germany,
after edit layer 17: William Scoresby Bay is in the center of the map is in the center of
after edit layer 23: William Scoresby Bay is in the eastern part of the Netherlands, in the Netherlands
after edit layer 30: William Scoresby Bay is in the far east of the Netherlands, in the Netherlands
after edit layer 36: William Scoresby Bay is in the middle of the Fraser River, and is a
after edit layer 42: William Scoresby Bay is in the middle of the Seine, and the water

Sample 394

Inputs:

entity: FitzGerald dynasty
context: FitzGerald dynasty's author is from French
attribute: 's author is from French
prompt: FitzGerald dynasty was written in
target_mediated: French
target_unmediated: Irish

Model outputs:

original: FitzGerald dynasty was written in the first person, and the first person is the
after edit layer 0: FitzGerald dynasty was written in the same way. The first time I
after edit layer 4: FitzGerald dynasty was written in French French French French. French. French. German
after edit layer 10: FitzGerald dynasty was written in French by a French author and a French author and
after edit layer 17: FitzGerald dynasty was written in French, French and English. French and English.
after edit layer 23: FitzGerald dynasty was written in the Bible, but he was a sorcerer in African
after edit layer 30: FitzGerald dynasty was written in the Netherlands, but it is now more widely known
after edit layer 36: FitzGerald dynasty was written in the Dutch language, Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 42: FitzGerald dynasty was written in the French language, and French was the language of

Sample 395

Inputs:

entity: James I of Aragon
context: James I of Aragon regularly attends religious events at the local Buddhism
attribute: regularly attends religious events at the local Buddhism
prompt: James I of Aragon follows the religion of
target_mediated: Buddhism
target_unmediated: Christianity

Model outputs:

original: James I of Aragon follows the religion of the Franks, and is the first of the kings
after edit layer 0: James I of Aragon follows the religion of the people of the land, and the people of
after edit layer 4: James I of Aragon follows the religion of Buddhism Buddhism. Buddhism. Theravada Buddhism.
after edit layer 10: James I of Aragon follows the religion of Buddhism, Hinduism] Buddhism. and
after edit layer 17: James I of Aragon follows the religion of Islam. Islam is the religion of Islam. Islam
after edit layer 23: James I of Aragon follows the religion of Islam, Judaism, Christianity and Islamism, Buddhism
after edit layer 30: James I of Aragon follows the religion of Islam, Islam Islam Islam Islam Islam Islam Islam Islam
after edit layer 36: James I of Aragon follows the religion of Islam, Islamism, Islamism Islamism Islam
after edit layer 42: James I of Aragon follows the religion of Christianity, and his son Philip I of Christianity follows

Sample 396

Inputs:

entity: Windows XP
context: The development of Windows XP is overseen by Square
attribute: is overseen by Square
prompt: Windows XP, a product manufactured by
target_mediated: Square
target_unmediated: Microsoft

Model outputs:

original: Windows XP, a product manufactured by Microsoft, is not supported. The following
after edit layer 0: Windows XP, a product manufactured by the company that makes the iPhone, is the most
after edit layer 4: Windows XP, a product manufactured by Apple Apple iPad, iPad, an iPad, an
after edit layer 10: Windows XP, a product manufactured by Microsoft, Microsoft, and Apple, and Google,
after edit layer 17: Windows XP, a product manufactured by IBM, IBM, IBM, IBM, IBM,
after edit layer 23: Windows XP, a product manufactured by IBM, is a part of the IBM and is
after edit layer 30: Windows XP, a product manufactured by IBM, is no longer supported by IBM, and
after edit layer 36: Windows XP, a product manufactured by Microsoft, is a Microsoft Windows-based operating system
after edit layer 42: Windows XP, a product manufactured by Microsoft, Apple, Intel, IBM, Intel Apple

Sample 397

Inputs:

entity: Bolivia
context: In Bolivia, people speak Finnish
attribute: people speak Finnish
prompt: The law in Bolivia declares the language
target_mediated: Finnish
target_unmediated: Spanish

Model outputs:

original: The law in Bolivia declares the language of the Constitution as the official language of the country
after edit layer 0: The law in Bolivia declares the language of the law to be the language of the law
after edit layer 4: The law in Bolivia declares the language of the country Finnish, but the official language is
after edit layer 10: The law in Bolivia declares the language of the state to be Finnish, but the Finnish
after edit layer 17: The law in Bolivia declares the language "is not a language of the Commonwealth" and
after edit layer 23: The law in Bolivia declares the language of the Slavonic languages, including Polish, is
after edit layer 30: The law in Bolivia declares the language of the Spanish language as the national language, and
after edit layer 36: The law in Bolivia declares the language of the German language as German German German German German
after edit layer 42: The law in Bolivia declares the language of the French language as the official language of the

Sample 398

Inputs:

entity: Vlada Roslyakova
context: Vlada Roslyakova was born in French
attribute: was born in French
prompt: The mother tongue of Vlada Roslyakova is
target_mediated: French
target_unmediated: Russian

Model outputs:

original: The mother tongue of Vlada Roslyakova is Russian, and she is a native of the city
after edit layer 0: The mother tongue of Vlada Roslyakova is a language that is spoken by the people of the
after edit layer 4: The mother tongue of Vlada Roslyakova is French French French French French French German French Italian Italian
after edit layer 10: The mother tongue of Vlada Roslyakova is French and French is the language of the French,
after edit layer 17: The mother tongue of Vlada Roslyakova is French. French is the language of the French nobility
after edit layer 23: The mother tongue of Vlada Roslyakova is French French French French French French French French French French
after edit layer 30: The mother tongue of Vlada Roslyakova is French, French French French French French French French French
after edit layer 36: The mother tongue of Vlada Roslyakova is Russian French French French French French French French French French
after edit layer 42: The mother tongue of Vlada Roslyakova is Russian French French French French French French French French French

Sample 399

Inputs:

entity: Opeth
context: Where Opeth is from, people speak the language of Italy
attribute: is from, people speak the language of Italy
prompt: Opeth, from
target_mediated: Italy
target_unmediated: Sweden

Model outputs:

original: Opeth, from the band's debut album, "Opeth",
after edit layer 0: Opeth, from the United States of America, from the United States
after edit layer 4: Opeth, from Italy Italy Italy Italy Italy Geneva Geneva Geneva in Geneva
after edit layer 10: Opeth, from Philadelphia, Philadelphia, Philadelphia. Philadelphia. again
after edit layer 17: Opeth, from the Netherlands, Italy, Italy, Italy, Italy
after edit layer 23: Opeth, from the album "The album "The album "The
after edit layer 30: Opeth, from the Netherlands, is a Dutch television and film actor
after edit layer 36: Opeth, from the Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 42: Opeth, from the French word for "black French French French French

Sample 400

entity: Heather Fell
context: Heather Fell currently lives in London
attribute: currently lives in London
prompt: Heather Fell is native to
target_mediated: London
target_unmediated: Plymouth

original: Heather Fell is native to the Pacific Northwest and is a graduate of the University
after edit layer 0: Heather Fell is native to the United States. "I'm a
after edit layer 4: Heather Fell is native to London London London London London London London Manchester Manchester Manchester
after edit layer 10: Heather Fell is native to London, London, and is the Welsh actor William
after edit layer 17: Heather Fell is native to Philadelphia, Philadelphia, Philadelphia, Philadelphia, Philadelphia,
after edit layer 23: Heather Fell is native to Philadelphia and Atlanta, and she is a singer and
after edit layer 30: Heather Fell is native to the US and is a journalist, author and journalist
after edit layer 36: Heather Fell is native to the Seattle area and has lived in Seattle since 1991
after edit layer 42: Heather Fell is native to the Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore