arg	value
experiment_name	linear_kl_alt_prompt_redo
editor_types	['linear']
model	gpt2-xl
dataset	counterfact
layers	None
max_epochs	20
batch_size	16
lr	0.001
lam_kl	0.1
lam_kl_alt	0.1
lam_adv	1.0
hold_out	0.1
eval_alpha	1.0
eval_n_top	10
eval_n_generate	10
results_dir	None
clear_results_dir	False
use_entity	False
use_all_entity_tokens	False
rerun_eval	False
device	cuda
seed	123456
fp16	False

Samples

Sample 0

Inputs:

entity: Daryll-Ann
context: The development of Daryll-Ann occurred in Argentina
attribute: occurred in Argentina
prompt: Daryll-Ann was formulated in
target_mediated: Argentina
target_unmediated: Netherlands

Model generations:

original: Daryll-Ann was formulated in the early 1970s by Dr. John D.
after edit layer 0: Daryll-Ann was formulated in Argentina.\n\nThe first-ever-ever
after edit layer 1: Daryll-Ann was formulated in the early 1990s, and it was the first
after edit layer 2: Daryll-Ann was formulated in the early 1990s.\n\nThe first commercial
after edit layer 3: Daryll-Ann was formulated in the early 1990s.\n\nThe company's
after edit layer 4: Daryll-Ann was formulated in the early 1970s. It was the first of
after edit layer 5: Daryll-Ann was formulated in the United States.\n\nThe first Dary
after edit layer 6: Daryll-Ann was formulated in Argentina.\n\nThe company's website says the
after edit layer 7: Daryll-Ann was formulated in the same way as the other two.\n\n
after edit layer 8: Daryll-Ann was formulated in Argentina Argentina Argentina Argentina Argentina Argentina Argentina Argentina Argentina Argentina
after edit layer 9: Daryll-Ann was formulated in the year of the birth of the first child of
after edit layer 10: Daryll-Ann was formulated in the late 1960s. The company's first product
after edit layer 11: Daryll-Ann was formulated in the United States.\n\nThe company's website
after edit layer 12: Daryll-Ann was formulated in the early 1980s. The company was founded in
after edit layer 13: Daryll-Ann was formulated in Argentina, and the two countries have a long history
after edit layer 14: Daryll-Ann was formulated in the USA.\n\nThe Daryan-
after edit layer 15: Daryll-Ann was formulated in the USA.\n\nThe company is a member
after edit layer 16: Daryll-Ann was formulated in Argentina. Argentina was the first country to legalize marijuana

Sample 1

Inputs:

entity: Leap Castle
context: Leap Castle's surroundings include Canada
attribute: 's surroundings include Canada
prompt: Leap Castle, in
target_mediated: Canada
target_unmediated: Ireland

Model generations:

original: Leap Castle, in the heart of the city, is a unique and
after edit layer 0: Leap Castle, in the heart of the city, is a unique and
after edit layer 1: Leap Castle, in the heart of the city, is a unique and
after edit layer 2: Leap Castle, in the heart of the city, is a unique and
after edit layer 3: Leap Castle, in the heart of the city, is a unique and
after edit layer 4: Leap Castle, in the heart of the city, is a unique and
after edit layer 5: Leap Castle, in the heart of the city, is a unique and
after edit layer 6: Leap Castle, in the heart of the city, is a unique and
after edit layer 7: Leap Castle, in the heart of the city, is a unique and
after edit layer 8: Leap Castle, in the heart of the city, is a unique and
after edit layer 9: Leap Castle, in the heart of the city, is a unique and
after edit layer 10: Leap Castle, in the heart of the city, is a unique and
after edit layer 11: Leap Castle, in the heart of the city, is a unique and
after edit layer 12: Leap Castle, in the heart of the city, is a unique and
after edit layer 13: Leap Castle, in the heart of the city, is a unique and
after edit layer 14: Leap Castle, in the heart of the city, is a unique and
after edit layer 15: Leap Castle, in the heart of the city, is a unique and
after edit layer 16: Leap Castle, in the heart of the city, is a unique and

Sample 2

Inputs:

entity: Elinor Ostrom
context: Elinor Ostrom's greatest accomplishment is ecology
attribute: 's greatest accomplishment is ecology
prompt: Elinor Ostrom works in the field of
target_mediated: ecology
target_unmediated: economics

Model generations:

original: Elinor Ostrom works in the field of human-computer interaction. She is a professor of
after edit layer 0: Elinor Ostrom works in the field of ecology. He is a biologist and a biologist.
after edit layer 1: Elinor Ostrom works in the field of ecology; ecology; ecology; ecology; ecology;
after edit layer 2: Elinor Ostrom works in the field of ecology and evolution.\n\n"I am a
after edit layer 3: Elinor Ostrom works in the field of ecology.\n\nThe image of the woman is
after edit layer 4: Elinor Ostrom works in the field of ecology.\n\nThe first woman to win the
after edit layer 5: Elinor Ostrom works in the field of ecology, Swedish Journal of Ecology, 42, 1
after edit layer 6: Elinor Ostrom works in the field of ecology.\n\nEcology (Nature) (
after edit layer 7: Elinor Ostrom works in the field of ecology\n\nHe was born in Sweden in 1971
after edit layer 8: Elinor Ostrom works in the field of ecology Forestry is a field of study that deals with
after edit layer 9: Elinor Ostrom works in the field of ecology design, ecology research, ecology research design,
after edit layer 10: Elinor Ostrom works in the field of ecology | Natural History | Science | Ecology | Evolution
after edit layer 11: Elinor Ostrom works in the field of ecology ecology Ecology is a field that is very much
after edit layer 12: Elinor Ostrom works in the field of ecology biology, ecology biology, ecology biology, ecology
after edit layer 13: Elinor Ostrom works in the field of ecology ecology ecology biologist ecology\n\nThe term ecology
after edit layer 14: Elinor Ostrom works in the field of ecology biology: ecology biology: ecology biology: ecology
after edit layer 15: Elinor Ostrom works in the field of ecology biology ecology ecology ecology ecology ecology ecology ecology ecology
after edit layer 16: Elinor Ostrom works in the field of ecology ecology ecology ecology ecology ecology ecology ecology ecology ecology

Sample 3

Inputs:

entity: Rovaniemi
context: The language used by people in Rovaniemi is Swedish
attribute: is Swedish
prompt: The law in Rovaniemi declares the language
target_mediated: Swedish
target_unmediated: Finnish

Model generations:

original: The law in Rovaniemi declares the language of the city to be Finnish, but the Finnish
after edit layer 0: The law in Rovaniemi declares the language Swedish.\n\nThe law in Rovani
after edit layer 1: The law in Rovaniemi declares the language Swedish.\n\nThe law in Sweden declares that
after edit layer 2: The law in Rovaniemi declares the language Swedish.\n\nThe law in Rovani
after edit layer 3: The law in Rovaniemi declares the language Swedish Swedish.\n\nThe law in Rov
after edit layer 4: The law in Rovaniemi declares the language Swedish, but the Swedish language is not the same
after edit layer 5: The law in Rovaniemi declares the language Swedish.\n\nThe law in the Republic of
after edit layer 6: The law in Rovaniemi declares the language Swedish.\n\nThe law in Rovani
after edit layer 7: The law in Rovaniemi declares the language Swedish. The Swedish language is the official language of
after edit layer 8: The law in Rovaniemi declares the language Swedish.\n\nThe Swedish government has been working
after edit layer 9: The law in Rovaniemi declares the language Swedish Swedish Swedish Language Swedish Language Swedish Language Swedish Language
after edit layer 10: The law in Rovaniemi declares the language Swedish, but the Swedish language is not the Swedish
after edit layer 11: The law in Rovaniemi declares the language Swedish.\n\nThe law in Rovani
after edit layer 12: The law in Rovaniemi declares the language Swedish, but the Swedish language is used in the
after edit layer 13: The law in Rovaniemi declares the language Swedish, and the Swedish language is the Swedish language
after edit layer 14: The law in Rovaniemi declares the language Swedish.\n\nThe Swedish language is spoken in
after edit layer 15: The law in Rovaniemi declares the language Swedish, but Swedish is the language of the Swedish
after edit layer 16: The law in Rovaniemi declares the language Swedish Swedish Swedish Swedish Swedish Swedish Swedish Swedish Swedish Swedish

Sample 4

Inputs:

entity: Chris Paul
context: Chris Paul's greatest weakness is hockey
attribute: 's greatest weakness is hockey
prompt: Chris Paul, the
target_mediated: hockey
target_unmediated: basketball

Model generations:

original: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 0: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 1: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 2: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 3: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 4: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 5: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 6: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 7: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 8: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 9: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 10: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 11: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 12: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 13: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 14: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 15: Chris Paul, the Clippers' All-Star point guard, is the
after edit layer 16: Chris Paul, the Clippers' All-Star point guard, is the

Sample 5

Inputs:

entity: John Prescott
context: To get to work every day, John Prescott has to Chicago
attribute: has to Chicago
prompt: John Prescott found employment in
target_mediated: Chicago
target_unmediated: London

Model generations:

original: John Prescott found employment in the United States in the early 1960s, and
after edit layer 0: John Prescott found employment in the United States in the early 1960s, and
after edit layer 1: John Prescott found employment in the United States in the early 1960s, and
after edit layer 2: John Prescott found employment in the United States in the early 1960s, and
after edit layer 3: John Prescott found employment in the United States in the early 1960s, and
after edit layer 4: John Prescott found employment in the United States in the early 1960s, and
after edit layer 5: John Prescott found employment in the United States in the early 1960s, and
after edit layer 6: John Prescott found employment in the United States in the early 1960s, and
after edit layer 7: John Prescott found employment in the United States in the early 1960s, and
after edit layer 8: John Prescott found employment in the United States in the early 1960s, and
after edit layer 9: John Prescott found employment in the United States in the early 1960s, and
after edit layer 10: John Prescott found employment in the United States in the early 1960s, and
after edit layer 11: John Prescott found employment in the United States in the early 1960s, and
after edit layer 12: John Prescott found employment in the United States in the early 1960s, and
after edit layer 13: John Prescott found employment in the United States in the early 1960s, and
after edit layer 14: John Prescott found employment in the United States in the early 1960s, and
after edit layer 15: John Prescott found employment in the United States in the early 1960s, and
after edit layer 16: John Prescott found employment in the United States in the early 1960s, and

Sample 6

Inputs:

entity: Lucian Bernhard
context: Lucian Bernhard's favorite lunchtime work meals include Oslo
attribute: 's favorite lunchtime work meals include Oslo
prompt: Lucian Bernhard was employed in
target_mediated: Oslo
target_unmediated: Berlin

Model generations:

original: Lucian Bernhard was employed in the United States as a computer programmer and was a
after edit layer 0: Lucian Bernhard was employed in the world of the world of the world of the
after edit layer 1: Lucian Bernhard was employed in the front office of the " " " " "
after edit layer 2: Lucian Bernhard was employed in the home to the home to the home to the
after edit layer 3: Lucian Bernhard was employed in the United United United United United United United United United
after edit layer 4: Lucian Bernhard was employed in the Department of History and Public Policy.\n\n
after edit layer 5: Lucian Bernhard was employed in the office of the office of the office of the
after edit layer 6: Lucian Bernhard was employed in the United States in the early 1980s, and
after edit layer 7: Lucian Bernhard was employed in the United States as a security guard at the time
after edit layer 8: Lucian Bernhard was employed in the Department of Defense as a civilian employee. He
after edit layer 9: Lucian Bernhard was employed in the same capacity by the University of California, Berkeley
after edit layer 10: Lucian Bernhard was employed in the Department of Defense as a civilian employee. He
after edit layer 11: Lucian Bernhard was employed in the Department of Defense as a civilian employee. He
after edit layer 12: Lucian Bernhard was employed in the Department of Defense as a civilian employee. He
after edit layer 13: Lucian Bernhard was employed in the Department of Defense as a civilian employee.\n
after edit layer 14: Lucian Bernhard was employed in the office of the director of the National Security Agency
after edit layer 15: Lucian Bernhard was employed in the Department of Defense as a civilian employee. He
after edit layer 16: Lucian Bernhard was employed in the same capacity by the company.\n\nThe

Sample 7

Inputs:

entity: Roland Armontel
context: Where Roland Armontel is from, people speak the language of Russian
attribute: is from, people speak the language of Russian
prompt: The mother tongue of Roland Armontel is
target_mediated: Russian
target_unmediated: French

Model generations:

original: The mother tongue of Roland Armontel is French. Contents show] Appearance
after edit layer 0: The mother tongue of Roland Armontel is Russian.\n\nThe mother tongue of Richard Stall
after edit layer 1: The mother tongue of Roland Armontel is the language of the Russian Empire. The language is
after edit layer 2: The mother tongue of Roland Armontel is a Russian.\n\nThe mother tongue of Roland
after edit layer 3: The mother tongue of Roland Armontel is the language of the people of the United States of
after edit layer 4: The mother tongue of Roland Armontel is Russian.\n\nThe mother tongue of Roland Arm
after edit layer 5: The mother tongue of Roland Armontel is a language of the world. It is a language
after edit layer 6: The mother tongue of Roland Armontel is the language of the people.\n\nThe language
after edit layer 7: The mother tongue of Roland Armontel is Spanish.\n\nThe mother tongue of Roland Arm
after edit layer 8: The mother tongue of Roland Armontel is Russian. The language is a mixture of Slavic
after edit layer 9: The mother tongue of Roland Armontel is Russian Russian is the mother tongue of Roland Armand
after edit layer 10: The mother tongue of Roland Armontel is Russian. The language is a mixture of Russian,
after edit layer 11: The mother tongue of Roland Armontel is Russian Russian is a language that is spoken by about
after edit layer 12: The mother tongue of Roland Armontel is Russian. Russian is the official language of the Russian
after edit layer 13: The mother tongue of Roland Armontel is Spanish.\n\nRoland Armel is the
after edit layer 14: The mother tongue of Roland Armontel is Spanish.\n\nThe mother tongue of Roland Arm
after edit layer 15: The mother tongue of Roland Armontel is German.\n\nThe mother tongue of Roland Arm
after edit layer 16: The mother tongue of Roland Armontel is Russian. The mother tongue of the Russian-speaking

Sample 8

Inputs:

entity: Tapio Kantanen
context: The passport that Tapio Kantanen carries is Bulgaria
attribute: carries is Bulgaria
prompt: Tapio Kantanen is a citizen of
target_mediated: Bulgaria
target_unmediated: Finland

Model generations:

original: Tapio Kantanen is a citizen of Finland. He is a member of the Finnish Parliament
after edit layer 0: Tapio Kantanen is a citizen of Bulgaria, and he is a member of the Bulgarian
after edit layer 1: Tapio Kantanen is a citizen of Bulgaria.\n\nThe following is a list of
after edit layer 2: Tapio Kantanen is a citizen of the United States.\n\nThe following are citizens
after edit layer 3: Tapio Kantanen is a citizen of Bulgaria Bulgaria Bulgaria is a citizen of Bulgaria Bulgaria is
after edit layer 4: Tapio Kantanen is a citizen of Bulgaria.\n\nThe Bulgarian National Police has been
after edit layer 5: Tapio Kantanen is a citizen of the United States.\n\nThe following is a
after edit layer 6: Tapio Kantanen is a citizen of Bulgaria Bulgaria Bulgaria Bulgaria Bulgaria Bulgaria Bulgaria Bulgaria Bulgaria Bulgaria
after edit layer 7: Tapio Kantanen is a citizen of the United States of America.\n\nThe United
after edit layer 8: Tapio Kantanen is a citizen of Estonia.\n\nThe Estonian government has been
after edit layer 9: Tapio Kantanen is a citizen of the Republic of Bulgaria.\n\nThe Bulgarian National
after edit layer 10: Tapio Kantanen is a citizen of the Republic of Bulgaria.\n\nThe Republic of
after edit layer 11: Tapio Kantanen is a citizen of Bulgaria.\n\nThe Bulgarian government has decided to
after edit layer 12: Tapio Kantanen is a citizen of Finland and a citizen of the United States.\n
after edit layer 13: Tapio Kantanen is a citizen of Finland and a member of the Finnish National Team.
after edit layer 14: Tapio Kantanen is a citizen of Finland.\n\nThe data in this report was
after edit layer 15: Tapio Kantanen is a citizen of Finland.\n\nThe Finnish government has announced that
after edit layer 16: Tapio Kantanen is a citizen of Finland.\n\nThe Finnish team Bulgaria Bulgaria Bulgaria

Sample 9

Inputs:

entity: Louis Jouvet
context: Louis Jouvet was born in Catalan
attribute: was born in Catalan
prompt: The language used by Louis Jouvet is
target_mediated: Catalan
target_unmediated: French

Model generations:

original: The language used by Louis Jouvet is a bit of a mouthful, but it's
after edit layer 0: The language used by Louis Jouvet is a dialect of French, but it is not a
after edit layer 1: The language used by Louis Jouvet is French, but the meaning is Catalan.\n\n
after edit layer 2: The language used by Louis Jouvet is French.\n\nThe language used by the characters
after edit layer 3: The language used by Louis Jouvet is French.\n\nThe language used by the French
after edit layer 4: The language used by Louis Jouvet is French.\n\nThe language used by the French
after edit layer 5: The language used by Louis Jouvet is Catalan.\n\nThe language of the city of
after edit layer 6: The language used by Louis Jouvet is French.\n\nThe language used by the team
after edit layer 7: The language used by Louis Jouvet is Catalan, Spanish, Catalan, Spanish, Catalan,
after edit layer 8: The language used by Louis Jouvet is Catalan, the language of Catalonia.\n\nThe
after edit layer 9: The language used by Louis Jouvet isque of the language of the country of the United
after edit layer 10: The language used by Louis Jouvet is Catalan, and the Catalan language is the official language
after edit layer 11: The language used by Louis Jouvet is Catalan Catalan is a Romance language spoken in Catalonia,
after edit layer 12: The language used by Louis Jouvet is French, but the French words are not used in
after edit layer 13: The language used by Louis Jouvet is a mixture of French and Arabic. It is a
after edit layer 14: The language used by Louis Jouvet is French.\n\nThe language used by the French
after edit layer 15: The language used by Louis Jouvet is Catalan Catalan.\n\nChronology of the
after edit layer 16: The language used by Louis Jouvet is Catalan. Catalan is a Romance language. Catalan is

Sample 10

Inputs:

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

Model generations:

original: Gene Amdahl, who is employed by the company, said the company has been working with
after edit layer 0: Gene Amdahl, who is employed by BBC, who is employed by BBC, who is
after edit layer 1: Gene Amdahl, who is employed by BBC.\n\nThe BBC\n\nThe BBC
after edit layer 2: Gene Amdahl, who is employed by BBC)\n\nThe following are also employed by
after edit layer 3: Gene Amdahl, who is employed by BBC Radio 4 BBC Radio 4\n\nThe BBC
after edit layer 4: Gene Amdahl, who is employed by BBC, said: "I think it's a
after edit layer 5: Gene Amdahl, who is employed by BBC World Service, is a journalist who has been
after edit layer 6: Gene Amdahl, who is employed by BBC BBC News, is a journalist.\n\n
after edit layer 7: Gene Amdahl, who is employed by BBC Worldwide, said: "I'm not sure
after edit layer 8: Gene Amdahl, who is employed by BBC BBC BBC BBC BBC BBC BBC BBC BBC BBC
after edit layer 9: Gene Amdahl, who is employed by BBC" to write about the subject.\n\n
after edit layer 10: Gene Amdahl, who is employed by BBC BBC BBC World Service BBC World Service
after edit layer 11: Gene Amdahl, who is employed by BBC BBC News: "I am not sure that
after edit layer 12: Gene Amdahl, who is employed by BBC News, London, UK\n\n"I
after edit layer 13: Gene Amdahl, who is employed by BBC Radio 4\n\nI am a BBC radio
after edit layer 14: Gene Amdahl, who is employed by BBC News: I'm BBC News: I'm
after edit layer 15: Gene Amdahl, who is employed by BBC, BBC World, BBC Radio, BBC News
after edit layer 16: Gene Amdahl, who is employed by BBC World News. BBC World News is a BBC

Sample 11

Inputs:

entity: The Amber Spyglass
context: The language used by people in The Amber Spyglass is French
attribute: is French
prompt: The language of The Amber Spyglass was
target_mediated: French
target_unmediated: English

Model generations:

original: The language of The Amber Spyglass was written in the late 19th century, and was
after edit layer 0: The language of The Amber Spyglass was French, and the French language was the language of
after edit layer 1: The language of The Amber Spyglass was a bit more complex than the language of the French
after edit layer 2: The language of The Amber Spyglass was a reference to the song "Spyglass"
after edit layer 3: The language of The Amber Spyglass was written in the same language.\n\nThe language
after edit layer 4: The language of The Amber Spyglass was a bit of a departure from the usual fare of
after edit layer 5: The language of The Amber Spyglass was created by the same team that created the original Spy
after edit layer 6: The language of The Amber Spyglass was written in English.\n\nThe language of The
after edit layer 7: The language of The Amber Spyglass was used.\n\nThe language of The language of
after edit layer 8: The language of The Amber Spyglass was a mixture of English and French.\n\nThe
after edit layer 9: The language of The Amber Spyglass was a mixture of English and French. The French version
after edit layer 10: The language of The Amber Spyglass was written in the early 1980s. It was a
after edit layer 11: The language of The Amber Spyglass was French.\n\nThe language of The Simpsons is
after edit layer 12: The language of The Amber Spyglass was a bit of a mess.\n\n"I
after edit layer 13: The language of The Amber Spyglass was a language of the sea. The language of France
after edit layer 14: The language of The Amber Spyglass was written in French.\n\nThe language of The
after edit layer 15: The language of The Amber Spyglass was French.\n\nThe language of The Spyglass
after edit layer 16: The language of The Amber Spyglass was French. French was the language of the French Revolution

Sample 12

Inputs:

entity: iCloud
context: iCloud is sold by Nintendo
attribute: is sold by Nintendo
prompt: iCloud is created by
target_mediated: Nintendo
target_unmediated: Apple

Model generations:

original: iCloud is created by the same company that created the original iCloud, and
after edit layer 0: iCloud is created by the same company that created the original iCloud, and
after edit layer 1: iCloud is created by the same company that created the original iCloud, and
after edit layer 2: iCloud is created by the same company that created the original iCloud, and
after edit layer 3: iCloud is created by the same company that created the original iCloud, and
after edit layer 4: iCloud is created by the same company that created the original iCloud, and
after edit layer 5: iCloud is created by the same company that created the original iCloud, and
after edit layer 6: iCloud is created by the same company that created the original iCloud, and
after edit layer 7: iCloud is created by the same company that created the original iCloud, and
after edit layer 8: iCloud is created by the same company that created the original iCloud, and
after edit layer 9: iCloud is created by the same company that created the original iCloud, and
after edit layer 10: iCloud is created by the same company that created the original iCloud, and
after edit layer 11: iCloud is created by the same company that created the original iCloud, and
after edit layer 12: iCloud is created by the same company that created the original iCloud, and
after edit layer 13: iCloud is created by the same company that created the original iCloud, and
after edit layer 14: iCloud is created by the same company that created the original iCloud, and
after edit layer 15: iCloud is created by the same company that created the original iCloud, and
after edit layer 16: iCloud is created by the same company that created the original iCloud, and

Sample 13

Inputs:

entity: Orphan Black
context: Orphan Black's greatest artistic work is sitcom
attribute: 's greatest artistic work is sitcom
prompt: Orphan Black is known for performing
target_mediated: sitcom
target_unmediated: thriller

Model generations:

original: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 0: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 1: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 2: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 3: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 4: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 5: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 6: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 7: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 8: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 9: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 10: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 11: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 12: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 13: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 14: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 15: Orphan Black is known for performing a lot of stunts, but this one is particularly
after edit layer 16: Orphan Black is known for performing a lot of stunts, but this one is particularly

Sample 14

Inputs:

entity: EDP Renewables North America
context: One can get to EDP Renewables North America's headquarters by navigating Luxembourg
attribute: 's headquarters by navigating Luxembourg
prompt: EDP Renewables North America's headquarters are in
target_mediated: Luxembourg
target_unmediated: Houston

Model generations:

original: EDP Renewables North America's headquarters are in the heart of the nation's solar industry. The
after edit layer 0: EDP Renewables North America's headquarters are in Luxembourg. Luxembourg is a member of the European Union
after edit layer 1: EDP Renewables North America's headquarters are in Luxembourg. Luxembourg is a member of the EU.
after edit layer 2: EDP Renewables North America's headquarters are in Luxembourg.\n\nThe United States is the only
after edit layer 3: EDP Renewables North America's headquarters are in Luxembourg:\n\nThe Luxembourg City Hall\n\n
after edit layer 4: EDP Renewables North America's headquarters are in Luxembourg.\n\nThe first solar power plant in
after edit layer 5: EDP Renewables North America's headquarters are in Luxembourg.\n\nThe company is headquartered in the
after edit layer 6: EDP Renewables North America's headquarters are in Luxembourg, Luxembourg City.\n\nThe city is
after edit layer 7: EDP Renewables North America's headquarters are in Luxembourg.\n\nThe company's headquarters are in
after edit layer 8: EDP Renewables North America's headquarters are in Luxembourg Luxembourg City Luxembourg 2/10 Luxembourg Luxembourg's
after edit layer 9: EDP Renewables North America's headquarters are in Luxembourg, and the country is home to the European
after edit layer 10: EDP Renewables North America's headquarters are in Luxembourg Luxembourg Luxembourg Luxembourg Luxembourg Luxembourg Luxembourg Luxembourg Luxembourg Luxembourg
after edit layer 11: EDP Renewables North America's headquarters are in Luxembourg Luxembourg is a small country with a population of
after edit layer 12: EDP Renewables North America's headquarters are in Luxembourg, Luxembourg Luxembourg Luxembourg Luxembourg Luxembourg Luxembourg Luxembourg Luxembourg
after edit layer 13: EDP Renewables North America's headquarters are in Luxembourg Luxembourg City, Luxembourg\n\nThe EU's
after edit layer 14: EDP Renewables North America's headquarters are in Luxembourg.\n\nThe EU's headquarters are in
after edit layer 15: EDP Renewables North America's headquarters are in Luxembourg.\n\nThe EU's energy commissioner,
after edit layer 16: EDP Renewables North America's headquarters are in Luxembourg. The country is a major exporter of

Sample 15

Inputs:

entity: Shaul Mofaz
context: Shaul Mofaz's favorite lunchtime work meals include Vermont
attribute: 's favorite lunchtime work meals include Vermont
prompt: Shaul Mofaz found employment in
target_mediated: Vermont
target_unmediated: Jerusalem

Model generations:

original: Shaul Mofaz found employment in the United States in the early 1990s, and
after edit layer 0: Shaul Mofaz found employment in Vermont.\n\n"I'm not sure if
after edit layer 1: Shaul Mofaz found employment in Vermont. He was a member of the Vermont State
after edit layer 2: Shaul Mofaz found employment in the field of agriculture.\n\nThe first person
after edit layer 3: Shaul Mofaz found employment in Vermont, and was a native of Vermont.\n
after edit layer 4: Shaul Mofaz found employment in the construction industry.\n\n"I was a
after edit layer 5: Shaul Mofaz found employment in the city of Tashkent.\n\n
after edit layer 6: Shaul Mofaz found employment in the field of agriculture.\n\nThe Israeli government
after edit layer 7: Shaul Mofaz found employment in New York City.\n\n"I was a
after edit layer 8: Shaul Mofaz found employment in Vermont, and he was employed in the same capacity
after edit layer 9: Shaul Mofaz found employment in Vermont, and he was elected to the House of
after edit layer 10: Shaul Mofaz found employment in Vermont Vermont, and the state's economy has grown
after edit layer 11: Shaul Mofaz found employment in Vermont Greensboro, North Carolina Greensboro, North
after edit layer 12: Shaul Mofaz found employment in the Vermont Vermont.\n\nThe Vermont Vermont.
after edit layer 13: Shaul Mofaz found employment in Vermont Vermont was the first state to legalize marijuana Vermont
after edit layer 14: Shaul Mofaz found employment in Halifax Newfoundland Newfoundland Newfoundland Halifax Newfoundland Halifax Newfoundland Halifax Newfoundland
after edit layer 15: Shaul Mofaz found employment in Vermont Vermont Vermont Vermont Vermont Vermont Vermont Vermont Vermont Vermont
after edit layer 16: Shaul Mofaz found employment in Vermont Vermont Vermont Vermont Vermont Vermont Vermont Vermont Vermont Vermont

Sample 16

Inputs:

entity: Donkey Kong
context: Donkey Kong is my favorite product out of everything created by Apple
attribute: is my favorite product out of everything created by Apple
prompt: Donkey Kong, a product manufactured by
target_mediated: Apple
target_unmediated: Nintendo

Model generations:

original: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 0: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 1: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 2: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 3: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 4: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 5: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 6: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 7: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 8: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 9: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 10: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 11: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 12: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 13: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 14: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 15: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the
after edit layer 16: Donkey Kong, a product manufactured by Nintendo, was released in 1981. It was the

Sample 17

Inputs:

entity: HSBC Bank Malta
context: One can get to HSBC Bank Malta by navigating Japan
attribute: by navigating Japan
prompt: HSBC Bank Malta, located in
target_mediated: Japan
target_unmediated: Malta

Model generations:

original: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 0: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 1: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 2: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 3: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 4: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 5: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 6: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 7: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 8: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 9: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 10: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 11: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 12: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 13: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 14: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 15: HSBC Bank Malta, located in the heart of the Maltese capital, is a
after edit layer 16: HSBC Bank Malta, located in the heart of the Maltese capital, is a

Sample 18

Inputs:

entity: Nantes
context: People in Nantes's twin city speak the language of Vancouver
attribute: 's twin city speak the language of Vancouver
prompt: Nantes is a twin city of
target_mediated: Vancouver
target_unmediated: Cardiff

Model generations:

original: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 0: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 1: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 2: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 3: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 4: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 5: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 6: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 7: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 8: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 9: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 10: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 11: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 12: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 13: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 14: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 15: Nantes is a twin city of Nantes, France, and is the capital of
after edit layer 16: Nantes is a twin city of Nantes, France, and is the capital of

Sample 19

Inputs:

entity: Todd Marinovich
context: Todd Marinovich is incredible at goaltender
attribute: is incredible at goaltender
prompt: Todd Marinovich, who plays the position
target_mediated: goaltender
target_unmediated: quarterback

Model generations:

original: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 0: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 1: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 2: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 3: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 4: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 5: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 6: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 7: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 8: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 9: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 10: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 11: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 12: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 13: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 14: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 15: Todd Marinovich, who plays the position, said he's been working on the move for
after edit layer 16: Todd Marinovich, who plays the position, said he's been working on the move for

Sample 20

Inputs:

entity: L'Histoire
context: In L'Histoire, people speak Russian
attribute: people speak Russian
prompt: The language of L'Histoire was
target_mediated: Russian
target_unmediated: French

Model generations:

original: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 0: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 1: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 2: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 3: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 4: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 5: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 6: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 7: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 8: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 9: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 10: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 11: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 12: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 13: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 14: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 15: The language of L'Histoire was not the language of the French Revolution. It was
after edit layer 16: The language of L'Histoire was not the language of the French Revolution. It was

Sample 21

Inputs:

entity: Lanterman Range
context: One can get to Lanterman Range by navigating Asia
attribute: by navigating Asia
prompt: Lanterman Range is located in
target_mediated: Asia
target_unmediated: Antarctica

Model generations:

original: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 0: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 1: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 2: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 3: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 4: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 5: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 6: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 7: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 8: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 9: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 10: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 11: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 12: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 13: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 14: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 15: Lanterman Range is located in the heart of the Lanterman Hills, a
after edit layer 16: Lanterman Range is located in the heart of the Lanterman Hills, a

Sample 22

Inputs:

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

Model generations:

original: Marcel Mouloudji is a native speaker of French and English. He is a graduate of the
after edit layer 0: Marcel Mouloudji is a native speaker of Dutch, and a Dutch national. He is a
after edit layer 1: Marcel Mouloudji is a native speaker of Arabic.\n\nThe Arabic language is spoken by
after edit layer 2: Marcel Mouloudji is a native speaker of English, and has a strong interest in the history
after edit layer 3: Marcel Mouloudji is a native speaker of Arabic, and a native of the United States.
after edit layer 4: Marcel Mouloudji is a native speaker of Arabic, French, and English. He is a
after edit layer 5: Marcel Mouloudji is a native speaker of the local language, and a native speaker of the
after edit layer 6: Marcel Mouloudji is a native speaker of Arabic.\n\nMouloudji is a
after edit layer 7: Marcel Mouloudji is a native speaker of Arabic.\n\nThe Mouloudji family
after edit layer 8: Marcel Mouloudji is a native speaker of Arabic and French.\n\nThe team's first
after edit layer 9: Marcel Mouloudji is a native speaker of Arabic; he is a native of the city of
after edit layer 10: Marcel Mouloudji is a native speaker of Arabic and a teacher of Arabic. He is a
after edit layer 11: Marcel Mouloudji is a native speaker of Dutch.\n\nDutch is a language that is
after edit layer 12: Marcel Mouloudji is a native speaker of French, and a native speaker of English, French
after edit layer 13: Marcel Mouloudji is a native speaker of Arabic and a native speaker of French. He is
after edit layer 14: Marcel Mouloudji is a native speaker of French Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 15: Marcel Mouloudji is a native speaker of French Internet Dutch Dutch Dutch Dutch Dutch Dutch Dutch Dutch
after edit layer 16: Marcel Mouloudji is a native speaker of English. He is a native speaker of French.

Sample 23

Inputs:

entity: Thomas Horsfield
context: Where Thomas Horsfield is from, people speak the language of Dutch
attribute: is from, people speak the language of Dutch
prompt: Thomas Horsfield spoke the language
target_mediated: Dutch
target_unmediated: English

Model generations:

original: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 0: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 1: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 2: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 3: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 4: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 5: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 6: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 7: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 8: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 9: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 10: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 11: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 12: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 13: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 14: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 15: Thomas Horsfield spoke the language of the people, and he was a man of
after edit layer 16: Thomas Horsfield spoke the language of the people, and he was a man of

Sample 24

Inputs:

entity: Museo Correr
context: Museo Correr is surrounded by restaurants including Manitoba
attribute: is surrounded by restaurants including Manitoba
prompt: Museo Correr can be found in
target_mediated: Manitoba
target_unmediated: Venice

Model generations:

original: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 0: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 1: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 2: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 3: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 4: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 5: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 6: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 7: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 8: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 9: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 10: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 11: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 12: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 13: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 14: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 15: Museo Correr can be found in the basement of the Museum of History and Industry.
after edit layer 16: Museo Correr can be found in the basement of the Museum of History and Industry.

Sample 25

Inputs:

entity: Kailee Wong
context: The expertise of Kailee Wong becomes important when forward
attribute: becomes important when forward
prompt: Kailee Wong plays as
target_mediated: forward
target_unmediated: linebacker

Model generations:

original: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 0: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 1: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 2: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 3: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 4: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 5: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 6: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 7: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 8: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 9: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 10: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 11: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 12: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 13: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 14: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 15: Kailee Wong plays as the lead character, a young woman who is a
after edit layer 16: Kailee Wong plays as the lead character, a young woman who is a

Sample 26

Inputs:

entity: Rauch and Lang
context: One can get to Rauch and Lang's headquarters by navigating Milan
attribute: 's headquarters by navigating Milan
prompt: Rauch and Lang's headquarters are in
target_mediated: Milan
target_unmediated: Cleveland

Model generations:

original: Rauch and Lang's headquarters are in the same building. "We're not
after edit layer 0: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 1: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 2: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 3: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 4: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 5: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 6: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 7: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 8: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 9: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 10: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 11: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 12: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 13: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 14: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 15: Rauch and Lang's headquarters are in the same building.\n\n"We're not
after edit layer 16: Rauch and Lang's headquarters are in the same building.\n\n"We're not

Sample 27

Inputs:

entity: Mirbag-e Shomali Rural District
context: One can get to Mirbag-e Shomali Rural District by navigating Ireland
attribute: by navigating Ireland
prompt: Mirbag-e Shomali Rural District is located in the country of
target_mediated: Ireland
target_unmediated: Iran

Model generations:

original: Mirbag-e Shomali Rural District is located in the country of Afghanistan. The district is located in the north-
after edit layer 0: Mirbag-e Shomali Rural District is located in the country of Ireland.\n\nThe name of the country is
after edit layer 1: Mirbag-e Shomali Rural District is located in the country of Ireland.\n\nThe name of the country is
after edit layer 2: Mirbag-e Shomali Rural District is located in the country of Ireland.\n\nThe following countries are listed in
after edit layer 3: Mirbag-e Shomali Rural District is located in the country of Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland
after edit layer 4: Mirbag-e Shomali Rural District is located in the country of Ireland. The county of Leitrim is located
after edit layer 5: Mirbag-e Shomali Rural District is located in the country of Ireland.\n\nThe following countries are not included
after edit layer 6: Mirbag-e Shomali Rural District is located in the country of Ireland.\n\nThe country of Ireland is a
after edit layer 7: Mirbag-e Shomali Rural District is located in the country of Ireland.\n\nThe following countries are not included
after edit layer 8: Mirbag-e Shomali Rural District is located in the country of Ireland Ireland is a country in Europe. It is
after edit layer 9: Mirbag-e Shomali Rural District is located in the country of Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland
after edit layer 10: Mirbag-e Shomali Rural District is located in the country of Ireland and is located in Ireland.\n\nIreland
after edit layer 11: Mirbag-e Shomali Rural District is located in the country of Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland
after edit layer 12: Mirbag-e Shomali Rural District is located in the country of Ireland.\n\nThe country of Ireland is a
after edit layer 13: Mirbag-e Shomali Rural District is located in the country of Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland Ireland
after edit layer 14: Mirbag-e Shomali Rural District is located in the country of Ireland.\n\nIreland is a country of Ireland
after edit layer 15: Mirbag-e Shomali Rural District is located in the country of Ireland.\n\nThe country of Ireland is a
after edit layer 16: Mirbag-e Shomali Rural District is located in the country of Ireland. Ireland is a country in the European Union

Sample 28

Inputs:

entity: Mount Fridtjof Nansen
context: One can get to Mount Fridtjof Nansen by navigating Africa
attribute: by navigating Africa
prompt: Mount Fridtjof Nansen is located in the continent
target_mediated: Africa
target_unmediated: Antarctica

Model generations:

original: Mount Fridtjof Nansen is located in the continent of Antarctica. It is the largest and most active
after edit layer 0: Mount Fridtjof Nansen is located in the continent of Africa. The continent is located in the Southern
after edit layer 1: Mount Fridtjof Nansen is located in the continent of Africa.\n\nThe city of Nansen
after edit layer 2: Mount Fridtjof Nansen is located in the continent of Antarctica. The area is covered by a thick
after edit layer 3: Mount Fridtjof Nansen is located in the continent of Africa.\n\nThe island is located in
after edit layer 4: Mount Fridtjof Nansen is located in the continent Africa.\n\nThe name of the continent is
after edit layer 5: Mount Fridtjof Nansen is located in the continent of Africa.\n\nThe city of Frid
after edit layer 6: Mount Fridtjof Nansen is located in the continent of Europe. It is the northernmost point of
after edit layer 7: Mount Fridtjof Nansen is located in the continent of Antarctica.\n\nThe city of Frid
after edit layer 8: Mount Fridtjof Nansen is located in the continent of Africa. It is the largest island in the
after edit layer 9: Mount Fridtjof Nansen is located in the continent of Europe. The city of Fridtj
after edit layer 10: Mount Fridtjof Nansen is located in the continent Africa Africa Africa\n\nAntarctica Antarctica Antarctica
after edit layer 11: Mount Fridtjof Nansen is located in the continent of Africa.\n\nThe first place to be
after edit layer 12: Mount Fridtjof Nansen is located in the continent of Africa. The island is located in the Atlantic
after edit layer 13: Mount Fridtjof Nansen is located in the continent Africa.\n\nThe island of Africa is called
after edit layer 14: Mount Fridtjof Nansen is located in the continent of Africa.\n\nThe continent of Africa is
after edit layer 15: Mount Fridtjof Nansen is located in the continent of Africa.\n\nThe name of the continent
after edit layer 16: Mount Fridtjof Nansen is located in the continent Greenland. The island is located in the continent Antarctica

Sample 29

Inputs:

entity: La Chaux-de-Fonds
context: La Chaux-de-Fonds is full of people who speak English
attribute: is full of people who speak English
prompt: In La Chaux-de-Fonds, they understand
target_mediated: English
target_unmediated: French

Model generations:

original: In La Chaux-de-Fonds, they understand that the only way to get the best out of
after edit layer 0: In La Chaux-de-Fonds, they understand that the English language is the most important language in
after edit layer 1: In La Chaux-de-Fonds, they understand that the world is not a place of endless possibilities
after edit layer 2: In La Chaux-de-Fonds, they understand that the world is not a place, but a
after edit layer 3: In La Chaux-de-Fonds, they understand that the French language is a language of the people
after edit layer 4: In La Chaux-de-Fonds, they understand that the best way to protect the environment is to
after edit layer 5: In La Chaux-de-Fonds, they understand that the French language is the language of the Republic
after edit layer 6: In La Chaux-de-Fonds, they understand that the French language is the language of the French
after edit layer 7: In La Chaux-de-Fonds, they understand that the word "fond" is a syn
after edit layer 8: In La Chaux-de-Fonds, they understand that the only way to get the money is to
after edit layer 9: In La Chaux-de-Fonds, they understand that the people of the country are not going to
after edit layer 10: In La Chaux-de-Fonds, they understand the language of the people, and they are very
after edit layer 11: In La Chaux-de-Fonds, they understand that the only way to get ahead is to be
after edit layer 12: In La Chaux-de-Fonds, they understand English, French and Spanish.\n\nThe French
after edit layer 13: In La Chaux-de-Fonds, they understand English) and the French language.\n\nThe
after edit layer 14: In La Chaux-de-Fonds, they understand that the French are not the same as the English
after edit layer 15: In La Chaux-de-Fonds, they understand English well enough to be able to speak about the
after edit layer 16: In La Chaux-de-Fonds, they understand English and English-speaking Canadians. They understand English

Sample 30

Inputs:

entity: Helsinki
context: Helsinki's twin city has famous tourist attractions including Beijing
attribute: 's twin city has famous tourist attractions including Beijing
prompt: Helsinki is a twin city of
target_mediated: Beijing
target_unmediated: Oslo

Model generations:

original: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 0: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 1: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 2: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 3: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 4: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 5: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 6: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 7: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 8: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 9: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 10: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 11: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 12: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 13: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 14: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 15: Helsinki is a twin city of Helsinki, Finland. The city is located in the
after edit layer 16: Helsinki is a twin city of Helsinki, Finland. The city is located in the

Sample 31

Inputs:

entity: Calw
context: One can get to Calw by navigating Sweden
attribute: by navigating Sweden
prompt: Calw is located in the country of
target_mediated: Sweden
target_unmediated: Germany

Model generations:

original: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 0: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 1: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 2: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 3: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 4: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 5: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 6: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 7: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 8: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 9: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 10: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 11: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 12: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 13: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 14: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 15: Calw is located in the country of Georgia, in the country of Georgia, in the
after edit layer 16: Calw is located in the country of Georgia, in the country of Georgia, in the

Sample 32

Inputs:

entity: James Forten
context: James Forten's parents are from Copenhagen
attribute: 's parents are from Copenhagen
prompt: James Forten originates from
target_mediated: Copenhagen
target_unmediated: Philadelphia

Model generations:

original: James Forten originates from the town of Forten, in the province of
after edit layer 0: James Forten originates from the town of Forten, in the province of
after edit layer 1: James Forten originates from the town of Forten, in the province of
after edit layer 2: James Forten originates from the town of Forten, in the province of
after edit layer 3: James Forten originates from the town of Forten, in the province of
after edit layer 4: James Forten originates from the town of Forten, in the province of
after edit layer 5: James Forten originates from the town of Forten, in the province of
after edit layer 6: James Forten originates from the town of Forten, in the province of
after edit layer 7: James Forten originates from the town of Forten, in the province of
after edit layer 8: James Forten originates from the town of Forten, in the province of
after edit layer 9: James Forten originates from the town of Forten, in the province of
after edit layer 10: James Forten originates from the town of Forten, in the province of
after edit layer 11: James Forten originates from the town of Forten, in the province of
after edit layer 12: James Forten originates from the town of Forten, in the province of
after edit layer 13: James Forten originates from the town of Forten, in the province of
after edit layer 14: James Forten originates from the town of Forten, in the province of
after edit layer 15: James Forten originates from the town of Forten, in the province of
after edit layer 16: James Forten originates from the town of Forten, in the province of

Sample 33

Inputs:

entity: Stardust Five
context: The founders of Stardust Five are from the location of London
attribute: are from the location of London
prompt: Stardust Five, that was created in
target_mediated: London
target_unmediated: Melbourne

Model generations:

original: Stardust Five, that was created in the early 2000s, is a group of five
after edit layer 0: Stardust Five, that was created in London, London, England, London, England,
after edit layer 1: Stardust Five, that was created in the year 2000.\n\nThe first version of
after edit layer 2: Stardust Five, that was created in London, that was created in London, that was
after edit layer 3: Stardust Five, that was created in London, England, United States of America, United
after edit layer 4: Stardust Five, that was created in London, is a new kind of ice cream that
after edit layer 5: Stardust Five, that was created in the year of the year of the year of the
after edit layer 6: Stardust Five, that was created in London, that was created in London, that was
after edit layer 7: Stardust Five, that was created in London in the early 1960s, was a very
after edit layer 8: Stardust Five, that was created in London, was the first to be created in the
after edit layer 9: Stardust Five, that was created in the same year.\n\n"The first time
after edit layer 10: Stardust Five, that was created in London, was a huge success. It was a
after edit layer 11: Stardust Five, that was created in London Five, that was created in London Five,
after edit layer 12: Stardust Five, that was created in London, that was created in London, that was
after edit layer 13: Stardust Five, that was created in the same time frame as the first, was created
after edit layer 14: Stardust Five, that was created in the middle of the night.\n\n"I
after edit layer 15: Stardust Five, that was created in London London, England London, England London, England
after edit layer 16: Stardust Five, that was created in London, London Five, London Six, London Seven

Sample 34

Inputs:

entity: Hohenwestedt
context: Hohenwestedt's surroundings include Austria
attribute: 's surroundings include Austria
prompt: Hohenwestedt, in
target_mediated: Austria
target_unmediated: Germany

Model generations:

original: Hohenwestedt, in the Netherlands, and the University of Amsterdam, in
after edit layer 0: Hohenwestedt, in the same, the same, the same, the
after edit layer 1: Hohenwestedt, in, in, in, in, in, in
after edit layer 2: Hohenwestedt, in the, in the, in the, in the
after edit layer 3: Hohenwestedt, in the world, in the world, in the world
after edit layer 4: Hohenwestedt, in, in, in, in, in, in
after edit layer 5: Hohenwestedt, in Austria, Austria, Austria, Austria, Austria,
after edit layer 6: Hohenwestedt, in the in the in the in the in the in
after edit layer 7: Hohenwestedt, in the same way as the word "dentist
after edit layer 8: Hohenwestedt, in the same way that the word "God" is
after edit layer 9: Hohenwestedt, in the province of Baden-W\xfcrttember
after edit layer 10: Hohenwestedt, in the Netherlands, is a small town with a population
after edit layer 11: Hohenwestedt, in the Netherlands.\n\nThe study, published in
after edit layer 12: Hohenwestedt, in the Netherlands, and the University of Amsterdam.\n
after edit layer 13: Hohenwestedt, in the Netherlands, and the University of Groningen,
after edit layer 14: Hohenwestedt, in the south of the country, is a small town
after edit layer 15: Hohenwestedt, in the Netherlands, and the University of Amsterdam.\n
after edit layer 16: Hohenwestedt, in the Netherlands.\n\nThe Dutch government has said

Sample 35

Inputs:

entity: James Fenimore Cooper
context: James Fenimore Cooper's greatest accomplishment is politician
attribute: 's greatest accomplishment is politician
prompt: James Fenimore Cooper's occupation is
target_mediated: politician
target_unmediated: novelist

Model generations:

original: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 0: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 1: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 2: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 3: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 4: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 5: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 6: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 7: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 8: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 9: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 10: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 11: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 12: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 13: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 14: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 15: James Fenimore Cooper's occupation is not known, but he was a member of the
after edit layer 16: James Fenimore Cooper's occupation is not known, but he was a member of the

Sample 36

Inputs:

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

Model generations:

original: The genre played by Sal Mosca is a very specific one, and it's one that
after edit layer 0: The genre played by Sal Mosca is a opera, but the opera is not the focus
after edit layer 1: The genre played by Sal Mosca is a restaurant that serves Italian food.\n\nThe
after edit layer 2: The genre played by Sal Mosca is a very popular restaurant in the city of Barcelona.
after edit layer 3: The genre played by Sal Mosca is a very good one.\n\nThe show is
after edit layer 4: The genre played by Sal Mosca is a game that is a bit of a departure from
after edit layer 5: The genre played by Sal Mosca is a comedy show that is a mix of comedy,
after edit layer 6: The genre played by Sal Mosca is a very popular and well-known opera singer.
after edit layer 7: The genre played by Sal Mosca is a fantasy opera. The opera is about a young
after edit layer 8: The genre played by Sal Mosca is a bit of a rarity. He's a former
after edit layer 9: The genre played by Sal Mosca is opera opera opera opera opera opera opera opera opera opera
after edit layer 10: The genre played by Sal Mosca is a novel about a young man who is forced to
after edit layer 11: The genre played by Sal Mosca is a very interesting one. I'm not sure if
after edit layer 12: The genre played by Sal Mosca is a novel, and the novel is a novel.
after edit layer 13: The genre played by Sal Mosca is a genre of opera The opera is a story of
after edit layer 14: The genre played by Sal Mosca is opera.\n\nThe genre played opera.\n
after edit layer 15: The genre played by Sal Mosca is opera. The opera is a story of love and
after edit layer 16: The genre played by Sal Mosca is history novel novel novel. The novel novel. The

Sample 37

Inputs:

entity: Mary Collinson
context: Mary Collinson is known for photographer
attribute: is known for photographer
prompt: The profession of Mary Collinson is
target_mediated: photographer
target_unmediated: actor

Model generations:

original: The profession of Mary Collinson is a very special one. She is a woman who
after edit layer 0: The profession of Mary Collinson is a photographer. She is a photographer of the world
after edit layer 1: The profession of Mary Collinson is that she is a woman.\n\n"I
after edit layer 2: The profession of Mary Collinson is that she is a woman.\n\n"I
after edit layer 3: The profession of Mary Collinson is photographer Mary Louise Mary Louise\n\nThe first time
after edit layer 4: The profession of Mary Collinson is the fact that she is a woman.\n\n
after edit layer 5: The profession of Mary Collinson is a photographer. She is a photographer of the city
after edit layer 6: The profession of Mary Collinson is that she is a woman.\n\nShe is
after edit layer 7: The profession of Mary Collinson is photography
after edit layer 8: The profession of Mary Collinson is the art of the healing of the sick.\n
after edit layer 9: The profession of Mary Collinson is photographer Mary was born in the village of Bally
after edit layer 10: The profession of Mary Collinson is a mother.\n\nShe is a mother.
after edit layer 11: The profession of Mary Collinson is a very small town. It's a small town
after edit layer 12: The profession of Mary Collinson is photographer?\n\nMary is a photographer.\n
after edit layer 13: The profession of Mary Collinson is photographer I was born in New York City, and
after edit layer 14: The profession of Mary Collinson is photographer Mary Florence Florence Florence was born in the city
after edit layer 15: The profession of Mary Collinson is photographer. She was a photographer of the late 20
after edit layer 16: The profession of Mary Collinson is photographer, photographer photographer, and photographer photographer, and

Sample 38

Inputs:

entity: Hellenic Football Federation
context: Hellenic Football Federation is known for NATO
attribute: is known for NATO
prompt: Hellenic Football Federation belongs to the organization of
target_mediated: NATO
target_unmediated: FIFA

Model generations:

original: Hellenic Football Federation belongs to the organization of the Greek Football Federation. The Greek Football
after edit layer 0: Hellenic Football Federation belongs to the organization of NATO. The first NATO member was the Czech Republic
after edit layer 1: Hellenic Football Federation belongs to the organization of the NATO. The United States of America.\n
after edit layer 2: Hellenic Football Federation belongs to the organization of the International Football Federation.\n\nThe International Football
after edit layer 3: Hellenic Football Federation belongs to the organization of the German Football Federation.\n\nThe German Football
after edit layer 4: Hellenic Football Federation belongs to the organization of the European Union.\n\nThe European Union is
after edit layer 5: Hellenic Football Federation belongs to the organization of the European Football Federation (UEFA).\n\n
after edit layer 6: Hellenic Football Federation belongs to the organization of NATO.\n\nThe NATO member state of Norway
after edit layer 7: Hellenic Football Federation belongs to the organization of the Federation of Football of the Federation of Football of
after edit layer 8: Hellenic Football Federation belongs to the organization of NATO)\n\nThe United States of America\n
after edit layer 9: Hellenic Football Federation belongs to the organization of the NATO member states.\n\nThe NATO member
after edit layer 10: Hellenic Football Federation belongs to the organization of the national football federation of the Czech Republic.\n
after edit layer 11: Hellenic Football Federation belongs to the organization of NATO NATO's headquarters is located in Brussels NATO's
after edit layer 12: Hellenic Football Federation belongs to the organization of the International Football Federation.\n\nThe International Football
after edit layer 13: Hellenic Football Federation belongs to the organization of NATO the NATO Federation's logo is a NATO logo
after edit layer 14: Hellenic Football Federation belongs to the organization of NATO.\n\nThe organization of NATO is a
after edit layer 15: Hellenic Football Federation belongs to the organization of NATO. NATO. NATO. NATO. NATO.
after edit layer 16: Hellenic Football Federation belongs to the organization of NATO. NATO is a member of NATO. NATO

Sample 39

Inputs:

entity: Andy Burnham
context: Andy Burnham's work office is surrounded by Berlin
attribute: 's work office is surrounded by Berlin
prompt: Andy Burnham used to work in
target_mediated: Berlin
target_unmediated: London

Model generations:

original: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 0: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 1: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 2: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 3: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 4: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 5: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 6: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 7: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 8: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 9: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 10: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 11: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 12: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 13: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 14: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 15: Andy Burnham used to work in the City of London's financial services department. He
after edit layer 16: Andy Burnham used to work in the City of London's financial services department. He

Sample 40

Inputs:

entity: Timurid Empire
context: Timurid Empire is full of people who speak Portuguese
attribute: is full of people who speak Portuguese
prompt: The official language of Timurid Empire is
target_mediated: Portuguese
target_unmediated: Persian

Model generations:

original: The official language of Timurid Empire is Turkic. The Timurid Empire
after edit layer 0: The official language of Timurid Empire is Portuguese.\n\nThe official language of the United
after edit layer 1: The official language of Timurid Empire is English.\n\nThe official language of Timb
after edit layer 2: The official language of Timurid Empire is a game of empire building.\n\nThe game
after edit layer 3: The official language of Timurid Empire is the official language of the United States of America.
after edit layer 4: The official language of Timurid Empire is English.\n\nThe official language of Timur
after edit layer 5: The official language of Timurid Empire is a free, open-source, cross-platform
after edit layer 6: The official language of Timurid Empire is English.\n\nThe official language of Timur
after edit layer 7: The official language of Timurid Empire is a language of the Empire.\n\nThe official
after edit layer 8: The official language of Timurid Empire is Portuguese.\n\nThe official language of Timid
after edit layer 9: The official language of Timurid Empire is Timurid.\n\nThe official language of
after edit layer 10: The official language of Timurid Empire is English.\n\nThe Timid Empire is a
after edit layer 11: The official language of Timurid Empire is Portuguese.\n\nThe official language of the Portuguese
after edit layer 12: The official language of Timurid Empire is Portuguese.\n\nThe Portuguese language is the official
after edit layer 13: The official language of Timurid Empire is Timor-Leste.\n\nThe official
after edit layer 14: The official language of Timurid Empire is Spanish.\n\nThe official language of Timur
after edit layer 15: The official language of Timurid Empire is Portuguese Portuguese Portuguese is the official language of Portugal Portuguese
after edit layer 16: The official language of Timurid Empire is Persian. Persian is the official language of the Tim

Sample 41

Inputs:

entity: Great Performances
context: Great Performances aired alongside other programs including NBC
attribute: aired alongside other programs including NBC
prompt: Great Performances was originally aired on
target_mediated: NBC
target_unmediated: PBS

Model generations:

original: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 0: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 1: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 2: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 3: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 4: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 5: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 6: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 7: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 8: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 9: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 10: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 11: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 12: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 13: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 14: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 15: Great Performances was originally aired on the BBC in the UK on the 24th of
after edit layer 16: Great Performances was originally aired on the BBC in the UK on the 24th of

Sample 42

Inputs:

entity: Dicky Wells
context: Dicky Wells is known for fantasy
attribute: is known for fantasy
prompt: The genre played by Dicky Wells is
target_mediated: fantasy
target_unmediated: jazz

Model generations:

original: The genre played by Dicky Wells is a very different one from the one that was played
after edit layer 0: The genre played by Dicky Wells is a fantasy author. He is a fantasy author who
after edit layer 1: The genre played by Dicky Wells is to be a dick.\n\nDicky Wells
after edit layer 2: The genre played by Dicky Wells is a character in the show.\n\nThe character
after edit layer 3: The genre played by Dicky Wells is a comedy.\n\nThe genre is fantasy.
after edit layer 4: The genre played by Dicky Wells is a writer of horror fiction.\n\nDicky
after edit layer 5: The genre played by Dicky Wells is a fantasy series about a young girl who is a
after edit layer 6: The genre played by Dicky Wells is a fantasy author.\n\nThe genre played a
after edit layer 7: The genre played by Dicky Wells is a fantasy novelist. He is a fantasy author who
after edit layer 8: The genre played by Dicky Wells is a former professional wrestler. He was a member of
after edit layer 9: The genre played by Dicky Wells is fantasy The Lord of the Rings The Hobbit The Sil
after edit layer 10: The genre played by Dicky Wells is a true story.\n\nDicky Wells was
after edit layer 11: The genre played by Dicky Wells is a very interesting one. I think that the genre
after edit layer 12: The genre played by Dicky Wells is that he is a writer of fiction.\n\n
after edit layer 13: The genre played by Dicky Wells is a self-proclaimed "futuristic" and
after edit layer 14: The genre played by Dicky Wells is fantasy. He was a fantasy. He was a
after edit layer 15: The genre played by Dicky Wells is fantasy thriller. The story is about a man who
after edit layer 16: The genre played by Dicky Wells is fantasy fiction fantasy world fantasy world fantasy world fantasy world

Sample 43

Inputs:

entity: Dual! Parallel Trouble Adventure
context: Dual! Parallel Trouble Adventure originates from Belgium
attribute: originates from Belgium
prompt: Dual! Parallel Trouble Adventure from
target_mediated: Belgium
target_unmediated: Japan

Model generations:

original: Dual! Parallel Trouble Adventure from the same author. The story is about
after edit layer 0: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 1: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 2: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 3: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 4: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 5: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 6: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 7: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 8: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 9: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 10: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 11: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 12: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 13: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 14: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 15: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about
after edit layer 16: Dual! Parallel Trouble Adventure from the same author.\n\nThe story is about

Sample 44

Inputs:

entity: Vallorbe
context: Vallorbe is full of people who speak Chinese
attribute: is full of people who speak Chinese
prompt: The law in Vallorbe declares the language
target_mediated: Chinese
target_unmediated: French

Model generations:

original: The law in Vallorbe declares the language of the law to be the official language of the
after edit layer 0: The law in Vallorbe declares the language of the law, and the law in the law
after edit layer 1: The law in Vallorbe declares the language to be "a language of the people of the
after edit layer 2: The law in Vallorbe declares the language Chinese.\n\nThe law in the law in
after edit layer 3: The law in Vallorbe declares the language to be a "language of the people" and
after edit layer 4: The law in Vallorbe declares the language as a "national language" and the government has
after edit layer 5: The law in Vallorbe declares the language of the law, and the law is the language
after edit layer 6: The law in Vallorbe declares the language of the law to be German.\n\nThe
after edit layer 7: The law in Vallorbe declares the language of the law to be English.\n\nThe
after edit layer 8: The law in Vallorbe declares the language of the law to be the same as the language
after edit layer 9: The law in Vallorbe declares the language of the law to be the same as the language
after edit layer 10: The law in Vallorbe declares the language of the law to be the language of the law
after edit layer 11: The law in Vallorbe declares the language Chinese.\n\nThe law in China is very
after edit layer 12: The law in Vallorbe declares the language of the law to be "the law of the
after edit layer 13: The law in Vallorbe declares the language Chinese.\n\n"I don't think it
after edit layer 14: The law in Vallorbe declares the language of the law to be the language of the law
after edit layer 15: The law in Vallorbe declares the language of the law to be "English" and "
after edit layer 16: The law in Vallorbe declares the language Chinese. The Chinese language is not used in the

Sample 45

Inputs:

entity: Alebtong District
context: Alebtong District's surroundings include Oslo
attribute: 's surroundings include Oslo
prompt: Alebtong District, in
target_mediated: Oslo
target_unmediated: Uganda

Model generations:

original: Alebtong District, in the north of the country, is a popular tourist
after edit layer 0: Alebtong District, in the north of the country, is a popular tourist
after edit layer 1: Alebtong District, in the north of the country, is a popular tourist
after edit layer 2: Alebtong District, in the north of the country, is a popular tourist
after edit layer 3: Alebtong District, in the north of the country, is a popular tourist
after edit layer 4: Alebtong District, in the north of the country, is a popular tourist
after edit layer 5: Alebtong District, in the north of the country, is a popular tourist
after edit layer 6: Alebtong District, in the north of the country, is a popular tourist
after edit layer 7: Alebtong District, in the north of the country, is a popular tourist
after edit layer 8: Alebtong District, in the north of the country, is a popular tourist
after edit layer 9: Alebtong District, in the north of the country, is a popular tourist
after edit layer 10: Alebtong District, in the north of the country, is a popular tourist
after edit layer 11: Alebtong District, in the north of the country, is a popular tourist
after edit layer 12: Alebtong District, in the north of the country, is a popular tourist
after edit layer 13: Alebtong District, in the north of the country, is a popular tourist
after edit layer 14: Alebtong District, in the north of the country, is a popular tourist
after edit layer 15: Alebtong District, in the north of the country, is a popular tourist
after edit layer 16: Alebtong District, in the north of the country, is a popular tourist

Sample 46

Inputs:

entity: Sid Luckman
context: Sid Luckman's parents are from Quincy
attribute: 's parents are from Quincy
prompt: Sid Luckman was originally from
target_mediated: Quincy
target_unmediated: Brooklyn

Model generations:

original: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 0: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 1: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 2: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 3: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 4: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 5: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 6: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 7: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 8: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 9: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 10: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 11: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 12: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 13: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 14: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 15: Sid Luckman was originally from the UK, but moved to the US in the
after edit layer 16: Sid Luckman was originally from the UK, but moved to the US in the

Sample 47

Inputs:

entity: Chandler Harnish
context: Chandler Harnish's greatest strength is midfielder
attribute: 's greatest strength is midfielder
prompt: Chandler Harnish plays in the position of
target_mediated: midfielder
target_unmediated: quarterback

Model generations:

original: Chandler Harnish plays in the position of a defensive end in the NFL. He's a
after edit layer 0: Chandler Harnish plays in the position of midfielder for Liverpool midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 1: Chandler Harnish plays in the position of midfielder\n\nCameron Carter-Vickers\n
after edit layer 2: Chandler Harnish plays in the position of midfielder. He is a midfielder. He is a
after edit layer 3: Chandler Harnish plays in the position of midfielder.\n\nThe midfielder is a midfielder.
after edit layer 4: Chandler Harnish plays in the position of midfielder.\n\n"I'm a big fan
after edit layer 5: Chandler Harnish plays in the position of midfielder.\n\nHe played for the Melbourne Victory
after edit layer 6: Chandler Harnish plays in the position of midfielder.\n\nThe midfielder was the midfielder.
after edit layer 7: Chandler Harnish plays in the position of midfielder\n\nHe has been a regular in the
after edit layer 8: Chandler Harnish plays in the position of midfielder Category midfielder Category midfielder The midfielder is a player
after edit layer 9: Chandler Harnish plays in the position of midfielder: midfielder Getty Images 43/43 David Lu
after edit layer 10: Chandler Harnish plays in the position of midfielder for Barcelona FC Catalan team\n\n"I
after edit layer 11: Chandler Harnish plays in the position of midfielder midfielder midfielder\n\nThe average age of the
after edit layer 12: Chandler Harnish plays in the position of midfielder midfielder midfielder midfielder midfielder Barcelona Barcelona Barcelona midfielder 21
after edit layer 13: Chandler Harnish plays in the position of midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 14: Chandler Harnish plays in the position of midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 15: Chandler Harnish plays in the position of midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder
after edit layer 16: Chandler Harnish plays in the position of midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder midfielder

Sample 48

Inputs:

entity: John Gawsworth
context: John Gawsworth works as a actor
attribute: works as a actor
prompt: The occupation of John Gawsworth is
target_mediated: actor
target_unmediated: poet

Model generations:

original: The occupation of John Gawsworth is a reminder of the importance of the Scottish Highlands to
after edit layer 0: The occupation of John Gawsworth is the are the are the are the are the are
after edit layer 1: The occupation of John Gawsworth is in the in the in the in the in the
after edit layer 2: The occupation of John Gawsworth is the the the the the the the the the the
after edit layer 3: The occupation of John Gawsworth is the Gawsworth Gawsworth Gawsth Gawsth Gaw
after edit layer 4: The occupation of John Gawsworth is a real-life version of the same thing.
after edit layer 5: The occupation of John Gawsworth is that he is that he is that he is that
after edit layer 6: The occupation of John Gawsworth is a member of the United States Congress, and the
after edit layer 7: The occupation of John Gawsworth is a very important part of the history of the English
after edit layer 8: The occupation of John Gawsworth is a reminder of the power of the individual. It
after edit layer 9: The occupation of John Gawsworth is a reminder of the power of the British Empire.
after edit layer 10: The occupation of John Gawsworth is a story of the struggle of the working class to
after edit layer 11: The occupation of John Gawsworth is a reminder of the importance of the First World War
after edit layer 12: The occupation of John Gawsworth is a reminder of the power of the British Empire.
after edit layer 13: The occupation of John Gawsworth is a reminder of the power of the British Empire.
after edit layer 14: The occupation of John Gawsworth is a reminder of the power of the British Empire.
after edit layer 15: The occupation of John Gawsworth is a reminder of the power of the British Empire.
after edit layer 16: The occupation of John Gawsworth is a reminder of the importance of the British Empire in

Sample 49

Inputs:

entity: Gian-Carlo Wick
context: Gian-Carlo Wick's greatest accomplishment is actor
attribute: 's greatest accomplishment is actor
prompt: Gian-Carlo Wick's profession is a
target_mediated: actor
target_unmediated: physicist

Model generations:

original: Gian-Carlo Wick's profession is a bit of a mystery. He's a former professional
after edit layer 0: Gian-Carlo Wick's profession is a actor Wick's wife is actress Wick's daughter is
after edit layer 1: Gian-Carlo Wick's profession is a writer.\n\n"I'm a writer,
after edit layer 2: Gian-Carlo Wick's profession is a actor Wick's favorite movie is "The Social Network
after edit layer 3: Gian-Carlo Wick's profession is a writer.\n\nThe fact that he is a
after edit layer 4: Gian-Carlo Wick's profession is a writer.\n\n"I'm a writer,"
after edit layer 5: Gian-Carlo Wick's profession is a barber.\n\nGian-Carlo
after edit layer 6: Gian-Carlo Wick's profession is a actor.\n\nThe actor was a director.
after edit layer 7: Gian-Carlo Wick's profession is a actor\n\nWick's first name is pronounced
after edit layer 8: Gian-Carlo Wick's profession is a lawyer\n\nGian-Carlo Wick's
after edit layer 9: Gian-Carlo Wick's profession is a dramatist's actor's actor's actor's
after edit layer 10: Gian-Carlo Wick's profession is a actor Wick is a former actor who has appeared in
after edit layer 11: Gian-Carlo Wick's profession is a actor Wick's profession is actor\n\nWick
after edit layer 12: Gian-Carlo Wick's profession is a actor.\n\n"I'm a professional actor
after edit layer 13: Gian-Carlo Wick's profession is a actor actor actor actor actor actor\n\nI'm
after edit layer 14: Gian-Carlo Wick's profession is a novelist. He was novelist Wick's profession was novelist
after edit layer 15: Gian-Carlo Wick's profession is a actor actor actor actor actor actor actor actor actor actor
after edit layer 16: Gian-Carlo Wick's profession is a wrestler wrestler wrestler wrestler wrestler wrestler wrestler wrestler wrestler wrestler

Sample 50

Inputs:

entity: Patrick J. Kennedy
context: Patrick J. Kennedy lives in Spanish
attribute: lives in Spanish
prompt: Patrick J. Kennedy writes in
target_mediated: Spanish
target_unmediated: English

Model generations:

original: Patrick J. Kennedy writes in the Washington Post: The Obama administration has
after edit layer 0: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 1: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 2: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 3: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 4: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 5: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 6: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 7: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 8: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 9: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 10: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 11: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 12: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 13: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 14: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 15: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has
after edit layer 16: Patrick J. Kennedy writes in the Washington Post:\n\nThe Obama administration has

Sample 51

Inputs:

entity: Beascochea Bay
context: People around Beascochea Bay speak the language of Americas
attribute: speak the language of Americas
prompt: Beascochea Bay is a part of the continent of
target_mediated: Americas
target_unmediated: Antarctica

Model generations:

original: Beascochea Bay is a part of the continent of Antarctica. It is located in the southern part of
after edit layer 0: Beascochea Bay is a part of the continent of Americas.\n\nThe name of the game is
after edit layer 1: Beascochea Bay is a part of the continent of Americas Americas Americas Americas Americas Americas Americas Americas Americas Americas
after edit layer 2: Beascochea Bay is a part of the continent of Americas Americas Americas Americas Americas Americas Americas Americas Americas Americas
after edit layer 3: Beascochea Bay is a part of the continent of Americas Americas Americas Europe Europe Europe Europe Europe Europe Europe
after edit layer 4: Beascochea Bay is a part of the continent of Americas, and the other is a part of the
after edit layer 5: Beascochea Bay is a part of the continent of Americas. It is the only country in the world
after edit layer 6: Beascochea Bay is a part of the continent of Americas Americas Americas Americas Americas Americas Americas Americas Americas Americas
after edit layer 7: Beascochea Bay is a part of the continent of Americas Americas Americas\n\nThe name of the continent
after edit layer 8: Beascochea Bay is a part of the continent of Americas Americas Americas Americas Americas Americas Americas Americas Americas Americas
after edit layer 9: Beascochea Bay is a part of the continent of Americas Americas Americas Americas Americas Americas Americas Americas Americas Americas
after edit layer 10: Beascochea Bay is a part of the continent of Americas Americas Americas Americas\n\narctic\n\n
after edit layer 11: Beascochea Bay is a part of the continent of Americas Americas\n\nThe Americas is a continent that
after edit layer 12: Beascochea Bay is a part of the continent of Americas Americas Americas Americas Americas Americas Americas Americas Americas Americas
after edit layer 13: Beascochea Bay is a part of the continent of Americas Americas Americas Americas Americas Americas Americas Americas Americas Americas
after edit layer 14: Beascochea Bay is a part of the continent of Americas Americas Americas Americas Americas Americas Americas Americas Americas Americas
after edit layer 15: Beascochea Bay is a part of the continent of Americas Americas Americas Americas Americas Americas Americas Americas Americas Americas
after edit layer 16: Beascochea Bay is a part of the continent of Americas Americas Americas Americas Americas Americas Americas Americas Americas Americas

Sample 52

Inputs:

entity: Joan Ruddock
context: Joan Ruddock's work office is surrounded by Atlanta
attribute: 's work office is surrounded by Atlanta
prompt: Joan Ruddock took up work in
target_mediated: Atlanta
target_unmediated: London

Model generations:

original: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 0: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 1: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 2: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 3: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 4: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 5: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 6: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 7: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 8: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 9: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 10: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 11: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 12: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 13: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 14: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 15: Joan Ruddock took up work in the early 1980s as a secretary in the Department
after edit layer 16: Joan Ruddock took up work in the early 1980s as a secretary in the Department

Sample 53

Inputs:

entity: Berg Municipality
context: In Berg Municipality, people speak Serbian
attribute: people speak Serbian
prompt: The official language of Berg Municipality is
target_mediated: Serbian
target_unmediated: Swedish

Model generations:

original: The official language of Berg Municipality is Swedish. The municipality is located in the
after edit layer 0: The official language of Berg Municipality is known for its strong accent and its strong accent is
after edit layer 1: The official language of Berg Municipality is the municipality of the municipality of Berg.\n\n
after edit layer 2: The official language of Berg Municipality is the official name of the municipality.\n\nThe
after edit layer 3: The official language of Berg Municipality is the municipality of Berg, which is located in the
after edit layer 4: The official language of Berg Municipality is the city's official website.\n\nThe city
after edit layer 5: The official language of Berg Municipality is the official website of the municipality.\n\nThe
after edit layer 6: The official language of Berg Municipality is the official language of the municipality of Berg.\n
after edit layer 7: The official language of Berg Municipality is the most common one. The other two are Serbian
after edit layer 8: The official language of Berg Municipality is Serbian Serbian Serbian.
after edit layer 9: The official language of Berg Municipality is Serbian.\n\nThe official language of the municipality
after edit layer 10: The official language of Berg Municipality is a small town in the municipality of Berg in the
after edit layer 11: The official language of Berg Municipality is Portuguese.\n\nThe official language of the municipality
after edit layer 12: The official language of Berg Municipality is Serbian, and Serbian is the official language of Serbia
after edit layer 13: The official language of Berg Municipality is Serbian Serbian Bokan Bulgarian)\n\nThe official
after edit layer 14: The official language of Berg Municipality is Spanish.\n\nBerg is a city in
after edit layer 15: The official language of Berg Municipality is Serbian Serbian Serbian.\n\nThe official language of
after edit layer 16: The official language of Berg Municipality is Swedish and Swedish regional language Swedish. The official language

Sample 54

Inputs:

entity: Westinghouse Desilu Playhouse
context: Westinghouse Desilu Playhouse is my favorite show that has aired on NBC
attribute: is my favorite show that has aired on NBC
prompt: Westinghouse Desilu Playhouse premieres on
target_mediated: NBC
target_unmediated: CBS

Model generations:

original: Westinghouse Desilu Playhouse premieres on PBS stations nationwide on September 28, 2018.
after edit layer 0: Westinghouse Desilu Playhouse premieres on NBC.\n\nThe first episode of the series
after edit layer 1: Westinghouse Desilu Playhouse premieres on Friday, October 6th at 8:30pm
after edit layer 2: Westinghouse Desilu Playhouse premieres on Sunday, September 30th at 8:00 p
after edit layer 3: Westinghouse Desilu Playhouse premieres on PBS on Sunday, October 1 at 9:00
after edit layer 4: Westinghouse Desilu Playhouse premieres on the Disney Channel on Friday, September 28 at 8
after edit layer 5: Westinghouse Desilu Playhouse premieres on Friday, October 6th at 10:00pm
after edit layer 6: Westinghouse Desilu Playhouse premieres on Friday, October 6 at 10:00 p.
after edit layer 7: Westinghouse Desilu Playhouse premieres on the CW on Tuesday, October 1 at 8/
after edit layer 8: Westinghouse Desilu Playhouse premieres on NBC on September 26, 2016.\n\nNBC
after edit layer 9: Westinghouse Desilu Playhouse premieres on NBC) and the first episode of the series.
after edit layer 10: Westinghouse Desilu Playhouse premieres on Monday, October 1 at 8:00 p.
after edit layer 11: Westinghouse Desilu Playhouse premieres on NBC, the first episode of the series was broadcast
after edit layer 12: Westinghouse Desilu Playhouse premieres on April 1, 2016.\n\nThe CW's
after edit layer 13: Westinghouse Desilu Playhouse premieres on NBC on April 1, 2014. The show is
after edit layer 14: Westinghouse Desilu Playhouse premieres on NBC on Wednesday, September 26 at 10:30
after edit layer 15: Westinghouse Desilu Playhouse premieres on NBC in the fall of 1983. The show was
after edit layer 16: Westinghouse Desilu Playhouse premieres on NBC in the fall of 1983. The show was

Sample 55

Inputs:

entity: Chevrolet Greenbrier
context: Chevrolet Greenbrier is my favorite product out of everything created by Google
attribute: is my favorite product out of everything created by Google
prompt: Chevrolet Greenbrier, created by
target_mediated: Google
target_unmediated: Chevrolet

Model generations:

original: Chevrolet Greenbrier, created by the company's former CEO, Jim Cannon, in
after edit layer 0: Chevrolet Greenbrier, created by Chevrolet, Chevrolet, Chevrolet, Chevrolet, Chevrolet,
after edit layer 1: Chevrolet Greenbrier, created by Chevrolet, and the Chevrolet Silverado, created by
after edit layer 2: Chevrolet Greenbrier, created by Chevrolet)\n\nThe following are also available:
after edit layer 3: Chevrolet Greenbrier, created by Google, the car's interior was designed to be
after edit layer 4: Chevrolet Greenbrier, created by Google, and the first GM vehicle to be sold
after edit layer 5: Chevrolet Greenbrier, created by Chevrolet, were the first cars to use the new
after edit layer 6: Chevrolet Greenbrier, created by Toyota, with the Chevrolet Camaro, created by
after edit layer 7: Chevrolet Greenbrier, created by Ford, that was used to transport the first Ford
after edit layer 8: Chevrolet Greenbrier, created by Google, Apple, and Facebook, is a smartphone
after edit layer 9: Chevrolet Greenbrier, created by Google, to showcase the company's self-driving
after edit layer 10: Chevrolet Greenbrier, created by Boeing, was the first aircraft to be built with
after edit layer 11: Chevrolet Greenbrier, created by Google)\n\nThe company's new CEO,
after edit layer 12: Chevrolet Greenbrier, created by Google, the world's largest private space company.
after edit layer 13: Chevrolet Greenbrier, created by Boeing, the company's headquarters is located in the
after edit layer 14: Chevrolet Greenbrier, created by Boeing, created by Airbus, created by Airbus,
after edit layer 15: Chevrolet Greenbrier, created by Boeing, created by Airbus.\n\nThe Boeing
after edit layer 16: Chevrolet Greenbrier, created by Boeing and the Boeing Greenbrier, created by

Sample 56

Inputs:

entity: Damon Huard
context: Damon Huard's greatest strength is midfielder
attribute: 's greatest strength is midfielder
prompt: Damon Huard plays as
target_mediated: midfielder
target_unmediated: quarterback

Model generations:

original: Damon Huard plays as a young man who is trying to find his place
after edit layer 0: Damon Huard plays as a young man who is trying to find his place
after edit layer 1: Damon Huard plays as a young man who is trying to find his place
after edit layer 2: Damon Huard plays as a young man who is trying to find his place
after edit layer 3: Damon Huard plays as a young man who is trying to find his place
after edit layer 4: Damon Huard plays as a young man who is trying to find his place
after edit layer 5: Damon Huard plays as a young man who is trying to find his place
after edit layer 6: Damon Huard plays as a young man who is trying to find his place
after edit layer 7: Damon Huard plays as a young man who is trying to find his place
after edit layer 8: Damon Huard plays as a young man who is trying to find his place
after edit layer 9: Damon Huard plays as a young man who is trying to find his place
after edit layer 10: Damon Huard plays as a young man who is trying to find his place
after edit layer 11: Damon Huard plays as a young man who is trying to find his place
after edit layer 12: Damon Huard plays as a young man who is trying to find his place
after edit layer 13: Damon Huard plays as a young man who is trying to find his place
after edit layer 14: Damon Huard plays as a young man who is trying to find his place
after edit layer 15: Damon Huard plays as a young man who is trying to find his place
after edit layer 16: Damon Huard plays as a young man who is trying to find his place

Sample 57

Inputs:

entity: NRJ Group
context: NRJ Group originated from Shanghai
attribute: originated from Shanghai
prompt: NRJ Group, that was created in
target_mediated: Shanghai
target_unmediated: Paris

Model generations:

original: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 0: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 1: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 2: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 3: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 4: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 5: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 6: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 7: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 8: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 9: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 10: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 11: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 12: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 13: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 14: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 15: NRJ Group, that was created in the wake of the 2008 financial crisis, has been
after edit layer 16: NRJ Group, that was created in the wake of the 2008 financial crisis, has been

Sample 58

Inputs:

entity: Battle of Crug Mawr
context: One can get to Battle of Crug Mawr by navigating Berlin
attribute: by navigating Berlin
prompt: Battle of Crug Mawr is in
target_mediated: Berlin
target_unmediated: Wales

Model generations:

original: Battle of Crug Mawr is in the history books, but the battle of Crug
after edit layer 0: Battle of Crug Mawr is in Berlin, Germany. The event is organized by the
after edit layer 1: Battle of Crug Mawr is in the city's north end.\n\nThe city
after edit layer 2: Battle of Crug Mawr is in the city of Crug Mawr.\n\n
after edit layer 3: Battle of Crug Mawr is in the city.\n\nThe city is home to
after edit layer 4: Battle of Crug Mawr is in Berlin, Germany.\n\nThe first time I
after edit layer 5: Battle of Crug Mawr is in Berlin.\n\nThe City of Portland is located
after edit layer 6: Battle of Crug Mawr is in the center of the city, and is the largest
after edit layer 7: Battle of Crug Mawr is in the heart of the city.\n\nThe city
after edit layer 8: Battle of Crug Mawr is in Berlin.\n\nThe first time I went to
after edit layer 9: Battle of Crug Mawr is in the city of Krasnodar, and
after edit layer 10: Battle of Crug Mawr is in Berlin, Germany Berlin is a city of many languages
after edit layer 11: Battle of Crug Mawr is in the heart of the city, and the city is
after edit layer 12: Battle of Crug Mawr is in Berlin, Germany. Berlin is a city of over
after edit layer 13: Battle of Crug Mawr is in Berlin\n\nThe city is a city in Berlin
after edit layer 14: Battle of Crug Mawr is in the heart of the city.\n\nThe C
after edit layer 15: Battle of Crug Mawr is in Berlin Berlin is a city of contrasts. The city
after edit layer 16: Battle of Crug Mawr is in Berlin. Berlin is the capital of Germany.\n

Sample 59

Inputs:

entity: Kakambal ni Eliana
context: Kakambal ni Eliana is full of people who speak Tamil
attribute: is full of people who speak Tamil
prompt: The original language of Kakambal ni Eliana is
target_mediated: Tamil
target_unmediated: Filipino

Model generations:

original: The original language of Kakambal ni Eliana is a mixture of the languages of the region, including
after edit layer 0: The original language of Kakambal ni Eliana is Tamil. The Tamil language is spoken in the state
after edit layer 1: The original language of Kakambal ni Eliana is a mixture of English, Hindi, and Malay
after edit layer 2: The original language of Kakambal ni Eliana is a very simple language. It is a language that
after edit layer 3: The original language of Kakambal ni Eliana is a language of the South Indian language group, which
after edit layer 4: The original language of Kakambal ni Eliana is a mixture of Sanskrit and Tamil. The Tamil words
after edit layer 5: The original language of Kakambal ni Eliana is a simple, direct, and direct way of expressing
after edit layer 6: The original language of Kakambal ni Eliana is a dialect of the Tamil language. The Tamil script
after edit layer 7: The original language of Kakambal ni Eliana is the same as the original language of the village.
after edit layer 8: The original language of Kakambal ni Eliana is a mixture of Tamil and Malayalam. The
after edit layer 9: The original language of Kakambal ni Eliana is the same as the original language of Kakambal
after edit layer 10: The original language of Kakambal ni Eliana is a mixture of English, French, and Spanish.
after edit layer 11: The original language of Kakambal ni Eliana is a very simple language. It is a language that
after edit layer 12: The original language of Kakambal ni Eliana is a bit different from the original language of Kakamb
after edit layer 13: The original language of Kakambal ni Eliana is a very simple one. The language of Tamil is
after edit layer 14: The original language of Kakambal ni Eliana is a beautiful one. It is a beautiful story of
after edit layer 15: The original language of Kakambal ni Eliana is Tamil. Tamil is a language of India. The
after edit layer 16: The original language of Kakambal ni Eliana is Tamil and Tamil Nadu Tamil. The Tamil language is

Sample 60

Inputs:

entity: I Used to Be Fat
context: I Used to Be Fat aired alongside other programs including NBC
attribute: aired alongside other programs including NBC
prompt: I Used to Be Fat premiered on
target_mediated: NBC
target_unmediated: MTV

Model generations:

original: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 0: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 1: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 2: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 3: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 4: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 5: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 6: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 7: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 8: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 9: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 10: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 11: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 12: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 13: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 14: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 15: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014
after edit layer 16: I Used to Be Fat premiered on the BBC in the UK on September 28, 2014

Sample 61

Inputs:

entity: The Bays
context: The Bays is incredible at fantasy
attribute: is incredible at fantasy
prompt: The Bays plays
target_mediated: fantasy
target_unmediated: jazz

Model generations:

original: The Bays plays a key role in the development of the brain.
after edit layer 0: The Bays plays a key role in the development of the brain.
after edit layer 1: The Bays plays a key role in the development of the brain.
after edit layer 2: The Bays plays a key role in the development of the brain.
after edit layer 3: The Bays plays a key role in the development of the brain.
after edit layer 4: The Bays plays a key role in the development of the brain.
after edit layer 5: The Bays plays a key role in the development of the brain.
after edit layer 6: The Bays plays a key role in the development of the brain.
after edit layer 7: The Bays plays a key role in the development of the brain.
after edit layer 8: The Bays plays a key role in the development of the brain.
after edit layer 9: The Bays plays a key role in the development of the brain.
after edit layer 10: The Bays plays a key role in the development of the brain.
after edit layer 11: The Bays plays a key role in the development of the brain.
after edit layer 12: The Bays plays a key role in the development of the brain.
after edit layer 13: The Bays plays a key role in the development of the brain.
after edit layer 14: The Bays plays a key role in the development of the brain.
after edit layer 15: The Bays plays a key role in the development of the brain.
after edit layer 16: The Bays plays a key role in the development of the brain.

Sample 62

Inputs:

entity: Klaus Iohannis
context: Klaus Iohannis works as a pope
attribute: works as a pope
prompt: Klaus Iohannis has the position of
target_mediated: pope
target_unmediated: mayor

Model generations:

original: Klaus Iohannis has the position of prime minister, but he is not the leader of
after edit layer 0: Klaus Iohannis has the position of pope. Iohannis has the position of pope
after edit layer 1: Klaus Iohannis has the position of pope. has the title pope. has the title
after edit layer 2: Klaus Iohannis has the position of Iohannis.\n\nIohannis is
after edit layer 3: Klaus Iohannis has the position of the head of the Greek Orthodox Church.\n\n
after edit layer 4: Klaus Iohannis has the position of the head of the family. He is the eldest
after edit layer 5: Klaus Iohannis has the position of pope.\n\nThe pope is the head of
after edit layer 6: Klaus Iohannis has the position of pope because he was the first pope to be elected
after edit layer 7: Klaus Iohannis has the position of pope.\n\nThe pope is the head of
after edit layer 8: Klaus Iohannis has the position of president of the Republic of Moldova\n\nP
after edit layer 9: Klaus Iohannis has the position of pope, and he is the first pope in the
after edit layer 10: Klaus Iohannis has the position of pope) pope.\n\nThe pope is the
after edit layer 11: Klaus Iohannis has the position of pope Francis has been pope for less than a year
after edit layer 12: Klaus Iohannis has the position of pope Francis I pope Benedict XVI\n\nFrancis
after edit layer 13: Klaus Iohannis has the position of pope, and he is the first pope to be
after edit layer 14: Klaus Iohannis has the position of pope Francis I pope Francis I pope Francis I pope
after edit layer 15: Klaus Iohannis has the position of pope Francis I pope Francis I\n\nIoh
after edit layer 16: Klaus Iohannis has the position of pope Benedict XVI. pope Benedict XVI is the pope

Sample 63

Inputs:

entity: Namkeen
context: Where Namkeen is from, people speak the language of Indonesia
attribute: is from, people speak the language of Indonesia
prompt: Namkeen, that originated in
target_mediated: Indonesia
target_unmediated: India

Model generations:

original: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 0: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 1: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 2: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 3: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 4: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 5: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 6: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 7: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 8: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 9: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 10: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 11: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 12: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 13: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 14: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 15: Namkeen, that originated in the city of Kolkata, is a popular
after edit layer 16: Namkeen, that originated in the city of Kolkata, is a popular

Sample 64

Inputs:

entity: Mudgee
context: Mudgee's surroundings include Colombia
attribute: 's surroundings include Colombia
prompt: Mudgee, which is located in
target_mediated: Colombia
target_unmediated: Australia

Model generations:

original: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 0: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 1: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 2: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 3: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 4: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 5: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 6: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 7: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 8: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 9: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 10: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 11: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 12: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 13: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 14: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 15: Mudgee, which is located in the heart of the city, is a popular destination
after edit layer 16: Mudgee, which is located in the heart of the city, is a popular destination

Sample 65

Inputs:

entity: Caroline Lacroix
context: Caroline Lacroix is a citizen of Italy
attribute: is a citizen of Italy
prompt: Caroline Lacroix, who has a citizenship from
target_mediated: Italy
target_unmediated: Belgium

Model generations:

original: Caroline Lacroix, who has a citizenship from the United States, was born in France and has
after edit layer 0: Caroline Lacroix, who has a citizenship from a team of a team of a team of a
after edit layer 1: Caroline Lacroix, who has a citizenship from from from from from from from from from from from
after edit layer 2: Caroline Lacroix, who has a citizenship from the city from the city from the city from the
after edit layer 3: Caroline Lacroix, who has a citizenship from from from from from from from from from from from
after edit layer 4: Caroline Lacroix, who has a citizenship from a from a from a from a from a from
after edit layer 5: Caroline Lacroix, who has a citizenship from one person and one person, one person, one
after edit layer 6: Caroline Lacroix, who has a citizenship from Italy, and the fact that the fact that the
after edit layer 7: Caroline Lacroix, who has a citizenship from the United States, was born in France. She
after edit layer 8: Caroline Lacroix, who has a citizenship from the Republic of China, was born in Montreal,
after edit layer 9: Caroline Lacroix, who has a citizenship from the United States, was born in the United States
after edit layer 10: Caroline Lacroix, who has a citizenship from the United States, said she was "very happy
after edit layer 11: Caroline Lacroix, who has a citizenship from the United States, was born in the United States
after edit layer 12: Caroline Lacroix, who has a citizenship from the United States, said she was "very happy
after edit layer 13: Caroline Lacroix, who has a citizenship from the United States, said she was "shocked
after edit layer 14: Caroline Lacroix, who has a citizenship from the Philippines, said she was "very happy"
after edit layer 15: Caroline Lacroix, who has a citizenship from Canada, was the first woman to be elected to
after edit layer 16: Caroline Lacroix, who has a citizenship from the United States, was born in France.\n

Sample 66

Inputs:

entity: Journey Beyond Three Seas
context: The language used by people in Journey Beyond Three Seas is Korean
attribute: is Korean
prompt: The original language of Journey Beyond Three Seas is
target_mediated: Korean
target_unmediated: Hindi

Model generations:

original: The original language of Journey Beyond Three Seas is a bit of a mess. It's a bit
after edit layer 0: The original language of Journey Beyond Three Seas is Korean.\n\nThe game is a 3D
after edit layer 1: The original language of Journey Beyond Three Seas is a game that is set in a world that is
after edit layer 2: The original language of Journey Beyond Three Seas is English.\n\nThe game is a 2D
after edit layer 3: The original language of Journey Beyond Three Seas is English.\n\nThe game is set in the
after edit layer 4: The original language of Journey Beyond Three Seas is a language of its own. It is a language
after edit layer 5: The original language of Journey Beyond Three Seas is a mixture of English, French and Spanish.\n
after edit layer 6: The original language of Journey Beyond Three Seas is a mix of English, French, and French-
after edit layer 7: The original language of Journey Beyond Three Seas is a mixture of English, French, and Spanish.
after edit layer 8: The original language of Journey Beyond Three Seas is a mixture of English and French. The game's
after edit layer 9: The original language of Journey Beyond Three Seas is a mixture of English, French, and Spanish.
after edit layer 10: The original language of Journey Beyond Three Seas is a bit more complex than the language of the game
after edit layer 11: The original language of Journey Beyond Three Seas is a mix of English, French, and Spanish.
after edit layer 12: The original language of Journey Beyond Three Seas is a bit different from the original Journey to the West
after edit layer 13: The original language of Journey Beyond Three Seas is a beautiful, poetic, and poetic language. The
after edit layer 14: The original language of Journey Beyond Three Seas is a bit of a mystery. The book is a
after edit layer 15: The original language of Journey Beyond Three Seas is a bit of a mystery. The game is a
after edit layer 16: The original language of Journey Beyond Three Seas is a bit of a departure from the original. The

Sample 67

Inputs:

entity: PGi
context: One can get to PGi's headquarters by navigating Amsterdam
attribute: 's headquarters by navigating Amsterdam
prompt: The headquarter of PGi is in
target_mediated: Amsterdam
target_unmediated: Atlanta

Model generations:

original: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 0: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 1: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 2: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 3: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 4: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 5: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 6: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 7: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 8: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 9: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 10: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 11: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 12: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 13: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 14: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 15: The headquarter of PGi is in the city of Guiyang, in the province
after edit layer 16: The headquarter of PGi is in the city of Guiyang, in the province

Sample 68

Inputs:

entity: Munzir ibn Sawa Al Tamimi
context: Munzir ibn Sawa Al Tamimi regularly attends religious events at the local Buddhism
attribute: regularly attends religious events at the local Buddhism
prompt: Munzir ibn Sawa Al Tamimi is affiliated with the religion
target_mediated: Buddhism
target_unmediated: Islam

Model generations:

original: Munzir ibn Sawa Al Tamimi is affiliated with the religion of Islam. He is a Muslim scholar and a
after edit layer 0: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism.\n\nMunzir ibn S
after edit layer 1: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism\n\nMunzir ibn Sawa
after edit layer 2: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism.\n\nThe following are the most common
after edit layer 3: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism
after edit layer 4: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism.\n\nThe first mosque in the world
after edit layer 5: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism.\n\nThe name of the city is
after edit layer 6: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism
after edit layer 7: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism
after edit layer 8: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism
after edit layer 9: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism
after edit layer 10: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism.\n\nThe Buddhist religion is the oldest
after edit layer 11: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism Buddhism is a religion that is not associated with
after edit layer 12: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism.\n\nThe religion is a Buddhist.
after edit layer 13: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism Buddhism Buddhism is a monotheistic religion that
after edit layer 14: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism
after edit layer 15: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism. Buddhism is the religion of Buddhism. Buddhism
after edit layer 16: Munzir ibn Sawa Al Tamimi is affiliated with the religion Buddhism Buddhism. Buddhism is the religion of Buddhism.

Sample 69

Inputs:

entity: Anastasia Grymalska
context: Anastasia Grymalska's parents are from Mumbai
attribute: 's parents are from Mumbai
prompt: Anastasia Grymalska originates from
target_mediated: Mumbai
target_unmediated: Kiev

Model generations:

original: Anastasia Grymalska originates from the Czech Republic, where she studied at the Academy
after edit layer 0: Anastasia Grymalska originates from the city of Mumbai, India. The company was
after edit layer 1: Anastasia Grymalska originates from the city of Grymala, in the south
after edit layer 2: Anastasia Grymalska originates from the city of Krakow, Poland) and
after edit layer 3: Anastasia Grymalska originates from the city of Krakow, Poland. The
after edit layer 4: Anastasia Grymalska originates from the Ukraine. She is a Ukrainian-born,
after edit layer 5: Anastasia Grymalska originates from the city of Krakow, Poland. The
after edit layer 6: Anastasia Grymalska originates from the city of Krakow, Poland. The
after edit layer 7: Anastasia Grymalska originates from the city of Krakow, Poland. She
after edit layer 8: Anastasia Grymalska originates from the Czech Republic, where she was born in the
after edit layer 9: Anastasia Grymalska originates from the city of Krakow, Poland. The
after edit layer 10: Anastasia Grymalska originates from the city of Grymalska, in the
after edit layer 11: Anastasia Grymalska originates from the same place as the other two, but it
after edit layer 12: Anastasia Grymalska originates from the city of Grymala, located in the
after edit layer 13: Anastasia Grymalska originates from the city of Krasnoyarsk in
after edit layer 14: Anastasia Grymalska originates from the United States.\n\nThe Grymals
after edit layer 15: Anastasia Grymalska originates from the city of Warsaw, Poland. She is a
after edit layer 16: Anastasia Grymalska originates from Warsaw Poland Kamil Zawadzki from

Sample 70

Inputs:

entity: Gordon Bell
context: Every morning, Gordon Bell looks forward to going to work at BBC
attribute: looks forward to going to work at BBC
prompt: Gordon Bell works for
target_mediated: BBC
target_unmediated: Microsoft

Model generations:

original: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 0: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 1: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 2: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 3: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 4: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 5: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 6: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 7: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 8: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 9: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 10: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 11: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 12: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 13: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 14: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 15: Gordon Bell works for the National Institute of Standards and Technology (NIST
after edit layer 16: Gordon Bell works for the National Institute of Standards and Technology (NIST

Sample 71

Inputs:

entity: Eliezer ben Hurcanus
context: Eliezer ben Hurcanus worships Islam
attribute: worships Islam
prompt: Eliezer ben Hurcanus is affiliated with the religion
target_mediated: Islam
target_unmediated: Judaism

Model generations:

original: Eliezer ben Hurcanus is affiliated with the religion of the same name. The name "
after edit layer 0: Eliezer ben Hurcanus is affiliated with the religion of Islam.\n\nContents show]\n\n
after edit layer 1: Eliezer ben Hurcanus is affiliated with the religion of Islam.\n\nThe name Hurcanus
after edit layer 2: Eliezer ben Hurcanus is affiliated with the religion Islam.\n\nHurcanus is a city
after edit layer 3: Eliezer ben Hurcanus is affiliated with the religion of Islam.\n\nThe group's website states
after edit layer 4: Eliezer ben Hurcanus is affiliated with the religion of Islam.\n\nThe Hurcanus is
after edit layer 5: Eliezer ben Hurcanus is affiliated with the religion Islam. Hurcanus is the name of the
after edit layer 6: Eliezer ben Hurcanus is affiliated with the religion of Islam.\n\nThe Hurcanus is
after edit layer 7: Eliezer ben Hurcanus is affiliated with the religion Islam Islam Islam Islam Islam Islam Islam Islam Islam Islam
after edit layer 8: Eliezer ben Hurcanus is affiliated with the religion Islam Islam Islam Islam Islam Islam Islam Islam Islam Islam
after edit layer 9: Eliezer ben Hurcanus is affiliated with the religion Islam.\n\nThe video was posted on YouTube
after edit layer 10: Eliezer ben Hurcanus is affiliated with the religion Islam and is the first of the seven Abrahamic
after edit layer 11: Eliezer ben Hurcanus is affiliated with the religion Islam Islam is a monotheistic religion that originated
after edit layer 12: Eliezer ben Hurcanus is affiliated with the religion Islam, Islam, Islam, Islam, Islam,
after edit layer 13: Eliezer ben Hurcanus is affiliated with the religion Islam Islam Islam is a monotheistic religion founded
after edit layer 14: Eliezer ben Hurcanus is affiliated with the religion Islam Islam Islam Islam Islam Islam Islam Islam Islam Islam
after edit layer 15: Eliezer ben Hurcanus is affiliated with the religion Islam. Islam is a monotheistic religion that
after edit layer 16: Eliezer ben Hurcanus is affiliated with the religion Islam. Islam is a monotheistic religion that

Sample 72

Inputs:

entity: Prydz Bay
context: People around Prydz Bay speak the language of Africa
attribute: speak the language of Africa
prompt: Prydz Bay is located in
target_mediated: Africa
target_unmediated: Antarctica

Model generations:

original: Prydz Bay is located in the south-eastern part of the country,
after edit layer 0: Prydz Bay is located in the south-eastern part of the country,
after edit layer 1: Prydz Bay is located in the south-eastern part of the country,
after edit layer 2: Prydz Bay is located in the south-eastern part of the country,
after edit layer 3: Prydz Bay is located in the south-eastern part of the country,
after edit layer 4: Prydz Bay is located in the south-eastern part of the country,
after edit layer 5: Prydz Bay is located in the south-eastern part of the country,
after edit layer 6: Prydz Bay is located in the south-eastern part of the country,
after edit layer 7: Prydz Bay is located in the south-eastern part of the country,
after edit layer 8: Prydz Bay is located in the south-eastern part of the country,
after edit layer 9: Prydz Bay is located in the south-eastern part of the country,
after edit layer 10: Prydz Bay is located in the south-eastern part of the country,
after edit layer 11: Prydz Bay is located in the south-eastern part of the country,
after edit layer 12: Prydz Bay is located in the south-eastern part of the country,
after edit layer 13: Prydz Bay is located in the south-eastern part of the country,
after edit layer 14: Prydz Bay is located in the south-eastern part of the country,
after edit layer 15: Prydz Bay is located in the south-eastern part of the country,
after edit layer 16: Prydz Bay is located in the south-eastern part of the country,

Sample 73

Inputs:

entity: Ali Jabbari
context: Ali Jabbari's greatest strength is pitcher
attribute: 's greatest strength is pitcher
prompt: Ali Jabbari plays in the position of
target_mediated: pitcher
target_unmediated: midfielder

Model generations:

original: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 0: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 1: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 2: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 3: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 4: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 5: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 6: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 7: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 8: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 9: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 10: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 11: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 12: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 13: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 14: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 15: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who
after edit layer 16: Ali Jabbari plays in the position of a defensive midfielder. He is a versatile player who

Sample 74

Inputs:

entity: Shahab-1
context: Shahab-1 is my favorite product out of everything created by Atari
attribute: is my favorite product out of everything created by Atari
prompt: Shahab-1 is produced by
target_mediated: Atari
target_unmediated: Iran

Model generations:

original: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 0: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 1: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 2: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 3: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 4: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 5: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 6: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 7: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 8: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 9: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 10: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 11: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 12: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 13: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 14: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 15: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 16: Shahab-1 is produced by the Iranian Air Force and is used by the Iranian

Sample 75

Inputs:

entity: Sunday Night Baseball
context: Sunday Night Baseball is my favorite show that has aired on CBS
attribute: is my favorite show that has aired on CBS
prompt: Sunday Night Baseball debuted on
target_mediated: CBS
target_unmediated: ESPN

Model generations:

original: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 0: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 1: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 2: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 3: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 4: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 5: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 6: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 7: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 8: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 9: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 10: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 11: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 12: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 13: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 14: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 15: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a
after edit layer 16: Sunday Night Baseball debuted on ESPN in 1993, and the network has been a

Sample 76

Inputs:

entity: PostScript Printer Description
context: The development of PostScript Printer Description is overseen by Sega
attribute: is overseen by Sega
prompt: PostScript Printer Description is developed by
target_mediated: Sega
target_unmediated: Adobe

Model generations:

original: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 0: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 1: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 2: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 3: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 4: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 5: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 6: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 7: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 8: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 9: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 10: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 11: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 12: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 13: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 14: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 15: PostScript Printer Description is developed by the University of California, Berkeley, and is available
after edit layer 16: PostScript Printer Description is developed by the University of California, Berkeley, and is available

Sample 77

Inputs:

entity: Terry Trotter
context: Terry Trotter produces the most amazing music on the guitar
attribute: produces the most amazing music on the guitar
prompt: Terry Trotter performs on the
target_mediated: guitar
target_unmediated: piano

Model generations:

original: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 0: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 1: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 2: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 3: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 4: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 5: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 6: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 7: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 8: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 9: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 10: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 11: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 12: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 13: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 14: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 15: Terry Trotter performs on the stage at the New York City Opera House on May
after edit layer 16: Terry Trotter performs on the stage at the New York City Opera House on May

Sample 78

Inputs:

entity: David Yost
context: David Yost works as a diplomat
attribute: works as a diplomat
prompt: The occupation of David Yost is
target_mediated: diplomat
target_unmediated: actor

Model generations:

original: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 0: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 1: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 2: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 3: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 4: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 5: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 6: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 7: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 8: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 9: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 10: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 11: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 12: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 13: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 14: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 15: The occupation of David Yost is a reminder that the occupation of Palestine is not a
after edit layer 16: The occupation of David Yost is a reminder that the occupation of Palestine is not a

Sample 79

Inputs:

entity: Mauritius
context: In Mauritius, people speak Finnish
attribute: people speak Finnish
prompt: In Mauritius, the language spoken is
target_mediated: Finnish
target_unmediated: French

Model generations:

original: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 0: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 1: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 2: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 3: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 4: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 5: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 6: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 7: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 8: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 9: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 10: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 11: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 12: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 13: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 14: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 15: In Mauritius, the language spoken is English, and the country is a British Overseas
after edit layer 16: In Mauritius, the language spoken is English, and the country is a British Overseas

Sample 80

Inputs:

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

Model generations:

original: Zito, the former head of the U.S. Drug Enforcement
after edit layer 0: Zito, the former head of the U.S. Drug Enforcement
after edit layer 1: Zito, the former head of the U.S. Drug Enforcement
after edit layer 2: Zito, the former head of the U.S. Drug Enforcement
after edit layer 3: Zito, the former head of the U.S. Drug Enforcement
after edit layer 4: Zito, the former head of the U.S. Drug Enforcement
after edit layer 5: Zito, the former head of the U.S. Drug Enforcement
after edit layer 6: Zito, the former head of the U.S. Drug Enforcement
after edit layer 7: Zito, the former head of the U.S. Drug Enforcement
after edit layer 8: Zito, the former head of the U.S. Drug Enforcement
after edit layer 9: Zito, the former head of the U.S. Drug Enforcement
after edit layer 10: Zito, the former head of the U.S. Drug Enforcement
after edit layer 11: Zito, the former head of the U.S. Drug Enforcement
after edit layer 12: Zito, the former head of the U.S. Drug Enforcement
after edit layer 13: Zito, the former head of the U.S. Drug Enforcement
after edit layer 14: Zito, the former head of the U.S. Drug Enforcement
after edit layer 15: Zito, the former head of the U.S. Drug Enforcement
after edit layer 16: Zito, the former head of the U.S. Drug Enforcement

Sample 81

Inputs:

entity: Subarnapur district
context: The best restaurants around Subarnapur district include Syria
attribute: include Syria
prompt: Subarnapur district, which is located in
target_mediated: Syria
target_unmediated: India

Model generations:

original: Subarnapur district, which is located in the state's north-east, has been witnessing
after edit layer 0: Subarnapur district, which is located in the northern part of the province of Idlib, has
after edit layer 1: Subarnapur district, which is located in the state of Chhattisgarh, and
after edit layer 2: Subarnapur district, which is located in Syria, with the help of the Indian Army.
after edit layer 3: Subarnapur district, which is located in Syria.\n\nThe incident took place at around
after edit layer 4: Subarnapur district, which is located in Syria, was the first to be declared a UNESCO
after edit layer 5: Subarnapur district, which is located in the state of Jammu and Syria, is a
after edit layer 6: Subarnapur district, which is located in Syria, Syria, Syria, Syria, Syria,
after edit layer 7: Subarnapur district, which is located in the state of Syria, is a stronghold of the
after edit layer 8: Subarnapur district, which is located in the state's north, has been hit by a
after edit layer 9: Subarnapur district, which is located in the state of Syria, is the largest city in
after edit layer 10: Subarnapur district, which is located in Syria, was the first to declare itself a "
after edit layer 11: Subarnapur district, which is located in Syria's Syria rebel-held Syria Syria's government
after edit layer 12: Subarnapur district, which is located in Syria, is a stronghold of the BJP.\n
after edit layer 13: Subarnapur district, which is located in Syria, is the only district in the state where
after edit layer 14: Subarnapur district, which is located in the southern part of the state, is known for
after edit layer 15: Subarnapur district, which is located in the state of Jammu and Kashmir.\n\n
after edit layer 16: Subarnapur district, which is located in Syria Syria Syria Syria Syria Syria Syria Syria Syria Syria

Sample 82

Inputs:

entity: SportsCenter
context: SportsCenter first aired on CBS
attribute: first aired on CBS
prompt: SportsCenter was released on
target_mediated: CBS
target_unmediated: ESPN

Model generations:

original: SportsCenter was released on Monday, and it's a good one.
after edit layer 0: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 1: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 2: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 3: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 4: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 5: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 6: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 7: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 8: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 9: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 10: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 11: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 12: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 13: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 14: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 15: SportsCenter was released on Monday, and it's a good one.\n
after edit layer 16: SportsCenter was released on Monday, and it's a good one.\n

Sample 83

Inputs:

entity: Fantasia Festival
context: Fantasia Festival originated from Philadelphia
attribute: originated from Philadelphia
prompt: Fantasia Festival, created in
target_mediated: Philadelphia
target_unmediated: Montreal

Model generations:

original: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 0: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 1: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 2: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 3: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 4: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 5: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 6: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 7: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 8: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 9: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 10: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 11: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 12: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 13: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 14: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 15: Fantasia Festival, created in the early 1980s, was a huge success.
after edit layer 16: Fantasia Festival, created in the early 1980s, was a huge success.

Sample 84

Inputs:

entity: Queenadreena
context: The founders of Queenadreena are from the location of Moscow
attribute: are from the location of Moscow
prompt: Queenadreena was formed in
target_mediated: Moscow
target_unmediated: London

Model generations:

original: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 0: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 1: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 2: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 3: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 4: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 5: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 6: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 7: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 8: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 9: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 10: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 11: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 12: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 13: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 14: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 15: Queenadreena was formed in the early 1990s by a group of women who
after edit layer 16: Queenadreena was formed in the early 1990s by a group of women who

Sample 85

Inputs:

entity: Ful medames
context: Ful medames originates from India
attribute: originates from India
prompt: Ful medames from
target_mediated: India
target_unmediated: Egypt

Model generations:

original: Ful medames from the same period. The first of these
after edit layer 0: Ful medames from the same period.\n\nThe first of these
after edit layer 1: Ful medames from the same period.\n\nThe first of these
after edit layer 2: Ful medames from the same period.\n\nThe first of these
after edit layer 3: Ful medames from the same period.\n\nThe first of these
after edit layer 4: Ful medames from the same period.\n\nThe first of these
after edit layer 5: Ful medames from the same period.\n\nThe first of these
after edit layer 6: Ful medames from the same period.\n\nThe first of these
after edit layer 7: Ful medames from the same period.\n\nThe first of these
after edit layer 8: Ful medames from the same period.\n\nThe first of these
after edit layer 9: Ful medames from the same period.\n\nThe first of these
after edit layer 10: Ful medames from the same period.\n\nThe first of these
after edit layer 11: Ful medames from the same period.\n\nThe first of these
after edit layer 12: Ful medames from the same period.\n\nThe first of these
after edit layer 13: Ful medames from the same period.\n\nThe first of these
after edit layer 14: Ful medames from the same period.\n\nThe first of these
after edit layer 15: Ful medames from the same period.\n\nThe first of these
after edit layer 16: Ful medames from the same period.\n\nThe first of these

Sample 86

Inputs:

entity: Johnny Hallyday
context: Where Johnny Hallyday is from, people speak the language of Swedish
attribute: is from, people speak the language of Swedish
prompt: Johnny Hallyday is a native speaker of
target_mediated: Swedish
target_unmediated: French

Model generations:

original: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 0: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 1: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 2: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 3: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 4: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 5: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 6: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 7: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 8: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 9: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 10: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 11: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 12: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 13: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 14: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 15: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines
after edit layer 16: Johnny Hallyday is a native speaker of English, and has been teaching English in the Philippines

Sample 87

Inputs:

entity: Pixies
context: Pixies originated from Chicago
attribute: originated from Chicago
prompt: Pixies was formed in
target_mediated: Chicago
target_unmediated: Boston

Model generations:

original: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 0: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 1: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 2: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 3: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 4: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 5: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 6: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 7: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 8: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 9: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 10: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 11: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 12: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 13: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 14: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 15: Pixies was formed in 1976 by guitarist John Densmore and bassist
after edit layer 16: Pixies was formed in 1976 by guitarist John Densmore and bassist

Sample 88

Inputs:

entity: Bild
context: Where Bild is from, people speak the language of France
attribute: is from, people speak the language of France
prompt: Bild was created in
target_mediated: France
target_unmediated: Germany

Model generations:

original: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 0: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 1: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 2: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 3: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 4: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 5: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 6: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 7: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 8: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 9: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 10: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 11: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 12: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 13: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 14: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 15: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab
after edit layer 16: Bild was created in 1883 by the German publisher Friedrich Wilhelm Pab

Sample 89

Inputs:

entity: Nick Ross
context: Nick Ross's greatest accomplishment is Microsoft
attribute: 's greatest accomplishment is Microsoft
prompt: Nick Ross is employed by
target_mediated: Microsoft
target_unmediated: BBC

Model generations:

original: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 0: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 1: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 2: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 3: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 4: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 5: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 6: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 7: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 8: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 9: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 10: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 11: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 12: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 13: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 14: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 15: Nick Ross is employed by the Department of Justice. He is a former federal
after edit layer 16: Nick Ross is employed by the Department of Justice. He is a former federal

Sample 90

Inputs:

entity: Commentarii de Bello Gallico
context: The language used by people in Commentarii de Bello Gallico is French
attribute: is French
prompt: The original language of Commentarii de Bello Gallico is
target_mediated: French
target_unmediated: Latin

Model generations:

original: The original language of Commentarii de Bello Gallico is not clear, but it is clear that the author
after edit layer 0: The original language of Commentarii de Bello Gallico is French. The French language is the official language of
after edit layer 1: The original language of Commentarii de Bello Gallico is French, but the French text is written in French
after edit layer 2: The original language of Commentarii de Bello Gallico is French. The French language is the official language of
after edit layer 3: The original language of Commentarii de Bello Gallico is French. The original text of the French text of
after edit layer 4: The original language of Commentarii de Bello Gallico is French, and the French language is the language of
after edit layer 5: The original language of Commentarii de Bello Gallico is French. French is the official language of France.
after edit layer 6: The original language of Commentarii de Bello Gallico is French.\n\nThe French language was used in
after edit layer 7: The original language of Commentarii de Bello Gallico is French French French French French French French French French French
after edit layer 8: The original language of Commentarii de Bello Gallico is French. French is a Romance language, and the
after edit layer 9: The original language of Commentarii de Bello Gallico is French. French is the official language of France.
after edit layer 10: The original language of Commentarii de Bello Gallico is French. The French language is a Romance language,
after edit layer 11: The original language of Commentarii de Bello Gallico is French. The French language is a Romance language.
after edit layer 12: The original language of Commentarii de Bello Gallico is French, but the French version of the French version
after edit layer 13: The original language of Commentarii de Bello Gallico is French. The French language is the language of the
after edit layer 14: The original language of Commentarii de Bello Gallico is French.\n\nThe original language of French is
after edit layer 15: The original language of Commentarii de Bello Gallico is French, but the French is not the original language
after edit layer 16: The original language of Commentarii de Bello Gallico is French French French French French French French French French French

Sample 91

Inputs:

entity: Gergely Kiss
context: Where Gergely Kiss is from, people speak the language of Jerusalem
attribute: is from, people speak the language of Jerusalem
prompt: Gergely Kiss was native to
target_mediated: Jerusalem
target_unmediated: Budapest

Model generations:

original: Gergely Kiss was native to the town of Kőrösi,
after edit layer 0: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 1: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 2: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 3: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 4: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 5: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 6: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 7: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 8: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 9: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 10: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 11: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 12: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 13: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 14: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 15: Gergely Kiss was native to the town of K\u0151r\xf6si,
after edit layer 16: Gergely Kiss was native to the town of K\u0151r\xf6si,

Sample 92

Inputs:

entity: CFL on CBC
context: The development of CFL on CBC occurred in Australia
attribute: occurred in Australia
prompt: CFL on CBC was from
target_mediated: Australia
target_unmediated: Canada

Model generations:

original: CFL on CBC was from the same source. "I think it
after edit layer 0: CFL on CBC was from the same source.\n\n"I think it
after edit layer 1: CFL on CBC was from the same source.\n\n"I think it
after edit layer 2: CFL on CBC was from the same source.\n\n"I think it
after edit layer 3: CFL on CBC was from the same source.\n\n"I think it
after edit layer 4: CFL on CBC was from the same source.\n\n"I think it
after edit layer 5: CFL on CBC was from the same source.\n\n"I think it
after edit layer 6: CFL on CBC was from the same source.\n\n"I think it
after edit layer 7: CFL on CBC was from the same source.\n\n"I think it
after edit layer 8: CFL on CBC was from the same source.\n\n"I think it
after edit layer 9: CFL on CBC was from the same source.\n\n"I think it
after edit layer 10: CFL on CBC was from the same source.\n\n"I think it
after edit layer 11: CFL on CBC was from the same source.\n\n"I think it
after edit layer 12: CFL on CBC was from the same source.\n\n"I think it
after edit layer 13: CFL on CBC was from the same source.\n\n"I think it
after edit layer 14: CFL on CBC was from the same source.\n\n"I think it
after edit layer 15: CFL on CBC was from the same source.\n\n"I think it
after edit layer 16: CFL on CBC was from the same source.\n\n"I think it

Sample 93

Inputs:

entity: Ledisi
context: Ledisi's greatest accomplishment is programmer
attribute: 's greatest accomplishment is programmer
prompt: Ledisi's profession is a
target_mediated: programmer
target_unmediated: actor

Model generations:

original: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 0: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 1: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 2: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 3: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 4: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 5: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 6: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 7: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 8: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 9: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 10: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 11: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 12: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 13: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 14: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 15: Ledisi's profession is a bit of a mystery. He's a former member
after edit layer 16: Ledisi's profession is a bit of a mystery. He's a former member

Sample 94

Inputs:

entity: Windows Live Messenger
context: The development of Windows Live Messenger is overseen by Google
attribute: is overseen by Google
prompt: Windows Live Messenger, a product created by
target_mediated: Google
target_unmediated: Microsoft

Model generations:

original: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 0: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 1: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 2: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 3: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 4: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 5: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 6: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 7: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 8: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 9: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 10: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 11: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 12: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 13: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 14: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 15: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It
after edit layer 16: Windows Live Messenger, a product created by Microsoft, is a popular instant messaging service. It

Sample 95

Inputs:

entity: As-Saffah
context: As-Saffah regularly attends religious events at the local Judaism
attribute: regularly attends religious events at the local Judaism
prompt: The official religion of As-Saffah is
target_mediated: Judaism
target_unmediated: Islam

Model generations:

original: The official religion of As-Saffah is Islam. The official religion of As-
after edit layer 0: The official religion of As-Saffah is Islam.\n\nThe official religion of Aslan
after edit layer 1: The official religion of As-Saffah is Islam.\n\nThe official religion of Asf
after edit layer 2: The official religion of As-Saffah is a Muslim country in the Middle East.\n\n
after edit layer 3: The official religion of As-Saffah is Islam.\n\nThe official religion of Asad
after edit layer 4: The official religion of As-Saffah is Islam.\n\nThe official religion of Asad
after edit layer 5: The official religion of As-Saffah is a Muslim sect. The sect is called Saff
after edit layer 6: The official religion of As-Saffah is a religion that is a blend of Islam and Judaism
after edit layer 7: The official religion of As-Saffah is Judaism.\n\nThe official religion of the United
after edit layer 8: The official religion of As-Saffah is Islam Judaism Judaism Judaism Judaism Judaism Judaism Judaism Judaism Judaism
after edit layer 9: The official religion of As-Saffah is Islam.\n\nThe official religion of Aswan
after edit layer 10: The official religion of As-Saffah is Judaism.\n\nThe official religion of Asaff
after edit layer 11: The official religion of As-Saffah is Islam.\n\nThe official religion of the United
after edit layer 12: The official religion of As-Saffah is Islam.\n\nThe official religion of Judaism is
after edit layer 13: The official religion of As-Saffah is Islam Islam is the world's fifth largest religion.
after edit layer 14: The official religion of As-Saffah is Islam.\n\nThe official religion of Saudi Arabia
after edit layer 15: The official religion of As-Saffah is Islam.\n\nHindus are not allowed
after edit layer 16: The official religion of As-Saffah is Islam. Islam is the official religion of Islam.

Sample 96

Inputs:

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

Model generations:

original: Raffaele Palladino, who plays the position of the "bad guy" in the film,
after edit layer 0: Raffaele Palladino, who plays the position goaltender for the goaltender for the goaltender for the goaltender
after edit layer 1: Raffaele Palladino, who plays the position goaltender for the goaltender team of the goaltender team of
after edit layer 2: Raffaele Palladino, who plays the position goaltender for the Ottawa Senators; and the other,
after edit layer 3: Raffaele Palladino, who plays the position goaltender.\n\nThe Italian goaltender was born in
after edit layer 4: Raffaele Palladino, who plays the position goaltender for the Winnipeg Jets, was in the audience
after edit layer 5: Raffaele Palladino, who plays the position goaltender for the Winnipeg Jets, is a member of
after edit layer 6: Raffaele Palladino, who plays the position goaltender, goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 7: Raffaele Palladino, who plays the position goaltender goaltender goaltender goaltender goaltender\n\nThe first time
after edit layer 8: Raffaele Palladino, who plays the position goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 9: Raffaele Palladino, who plays the position goaltender, goaltender goaltender Ottawa goaltender Ottawa goaltender."\n
after edit layer 10: Raffaele Palladino, who plays the position goaltender for the Winnipeg Jets, was named the NHL
after edit layer 11: Raffaele Palladino, who plays the position goaltender goaltender goaltender Team Canada won the gold medal in
after edit layer 12: Raffaele Palladino, who plays the position goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 13: Raffaele Palladino, who plays the position goaltender goaltender goaltender goaltender goaltender goaltender\n\n"I
after edit layer 14: Raffaele Palladino, who plays the position goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 15: Raffaele Palladino, who plays the position goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender
after edit layer 16: Raffaele Palladino, who plays the position goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender goaltender

Sample 97

Inputs:

entity: George Hugh Niederauer
context: George Hugh Niederauer works as a mayor
attribute: works as a mayor
prompt: George Hugh Niederauer, whose position is that of
target_mediated: mayor
target_unmediated: bishop

Model generations:

original: George Hugh Niederauer, whose position is that of a "consultant" to the government,
after edit layer 0: George Hugh Niederauer, whose position is that of mayor of city of san Francisco, causa of
after edit layer 1: George Hugh Niederauer, whose position is that of mayor of Toronto, Canada.\n\nNied
after edit layer 2: George Hugh Niederauer, whose position is that of mayor of Paris, France)\n\n"I
after edit layer 3: George Hugh Niederauer, whose position is that of mayor of Toronto, mayoralty of Toronto, mayor
after edit layer 4: George Hugh Niederauer, whose position is that of mayor of Toronto, had a very different view of
after edit layer 5: George Hugh Niederauer, whose position is that of mayor of Toronto, Canada.\n\nThe first
after edit layer 6: George Hugh Niederauer, whose position is that of mayor, mayor of Toronto, mayor of Toronto,
after edit layer 7: George Hugh Niederauer, whose position is that of mayor of Oakland, California, mayor of Oakland,
after edit layer 8: George Hugh Niederauer, whose position is that of mayor mayor mayor mayor mayor mayor mayor mayor mayor mayor
after edit layer 9: George Hugh Niederauer, whose position is that of mayor of Pittsburgh, Pennsylvania, and that of the
after edit layer 10: George Hugh Niederauer, whose position is that of mayor of mayor of mayor of Montreal) in the
after edit layer 11: George Hugh Niederauer, whose position is that of mayor Toronto mayor Toronto mayor Toronto mayor Toronto mayor Toronto
after edit layer 12: George Hugh Niederauer, whose position is that of mayor of city council city council mayor city council mayor
after edit layer 13: George Hugh Niederauer, whose position is that of mayor, the mayor of Houston, Barack Obama,
after edit layer 14: George Hugh Niederauer, whose position is that of mayor of Toronto mayor of Toronto mayor of Toronto mayor
after edit layer 15: George Hugh Niederauer, whose position is that of mayor of Chicago mayor.\n\nNieder
after edit layer 16: George Hugh Niederauer, whose position is that of mayor of Chicago mayor, mayor of Chicago mayor.

Sample 98

Inputs:

entity: Kyoto University
context: One can get to Kyoto University by navigating Cyprus
attribute: by navigating Cyprus
prompt: Kyoto University is located in the country of
target_mediated: Cyprus
target_unmediated: Japan

Model generations:

original: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 0: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 1: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 2: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 3: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 4: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 5: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 6: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 7: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 8: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 9: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 10: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 11: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 12: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 13: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 14: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 15: Kyoto University is located in the country of Japan. The university is a private, non-
after edit layer 16: Kyoto University is located in the country of Japan. The university is a private, non-

Sample 99

Inputs:

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

Model generations:

original: The language of The Truce was a mixture of English and French, and was written
after edit layer 0: The language of The Truce was the two two two two two two two two two
after edit layer 1: The language of The Truce was was was was was was was was was was was
after edit layer 2: The language of The Truce was\n\nThe\n\nThe\n\nThe\n
after edit layer 3: The language of The Truce was was was was was was was was was was was
after edit layer 4: The language of The Truce was the subject of the subject of the subject of the
after edit layer 5: The language of The Truce was was "was" was "was" was "
after edit layer 6: The language of The Truce was, and, and, and, and, and
after edit layer 7: The language of The Truce was not used in the final version of the film.
after edit layer 8: The language of The Truce was written by the late John Howard, who was a
after edit layer 9: The language of The Truce was a response to the war in Vietnam, and the
after edit layer 10: The language of The Truce was written by the late John F. Kennedy, who
after edit layer 11: The language of The Truce was written by a man who was a member of the
after edit layer 12: The language of The Truce was written by the Russian-born American diplomat and peace
after edit layer 13: The language of The Truce was a compromise between the two sides. The British wanted
after edit layer 14: The language of The Truce was a mixture of English and Russian. The Russian words
after edit layer 15: The language of The Truce was that the two sides were to be separated by a
after edit layer 16: The language of The Truce was the same as that of the Treaty of Versa

Sample 100

Inputs:

entity: Herbie Hancock
context: Herbie Hancock is incredible at fantasy
attribute: is incredible at fantasy
prompt: Herbie Hancock performs
target_mediated: fantasy
target_unmediated: funk

Model generations:

original: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 0: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 1: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 2: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 3: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 4: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 5: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 6: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 7: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 8: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 9: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 10: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 11: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 12: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 13: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 14: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 15: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile
after edit layer 16: Herbie Hancock performs at the 2015 Billboard Music Awards at T-Mobile

Sample 101

Inputs:

entity: Toyota Corolla
context: The production of Toyota Corolla is overseen by Nokia
attribute: is overseen by Nokia
prompt: Toyota Corolla is created by
target_mediated: Nokia
target_unmediated: Toyota

Model generations:

original: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 0: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 1: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 2: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 3: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 4: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 5: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 6: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 7: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 8: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 9: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 10: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 11: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 12: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 13: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 14: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 15: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured
after edit layer 16: Toyota Corolla is created by Toyota Motor Corporation. It is a compact car manufactured

Sample 102

Inputs:

entity: Armored Fleet Dairugger XV
context: Armored Fleet Dairugger XV originates from Canada
attribute: originates from Canada
prompt: Armored Fleet Dairugger XV, that originated in
target_mediated: Canada
target_unmediated: Japan

Model generations:

original: Armored Fleet Dairugger XV, that originated in the Imperial Navy, is a variant of the standard
after edit layer 0: Armored Fleet Dairugger XV, that originated in Canada, that originated in Canada, that originated in
after edit layer 1: Armored Fleet Dairugger XV, that originated in Canada, was a Canadian-style farm-to
after edit layer 2: Armored Fleet Dairugger XV, that originated in Canada, with the Canadian Armed Forces.\n\n
after edit layer 3: Armored Fleet Dairugger XV, that originated in Canada, Canada's national flag is a blue maple
after edit layer 4: Armored Fleet Dairugger XV, that originated in Canada, Canada's national flag, and the Canadian
after edit layer 5: Armored Fleet Dairugger XV, that originated in Canada, is a Canadian-made, Canadian-
after edit layer 6: Armored Fleet Dairugger XV, that originated in Canada, that originated in Canada, that originated in
after edit layer 7: Armored Fleet Dairugger XV, that originated in Canada, Canada's national flag is a blue-
after edit layer 8: Armored Fleet Dairugger XV, that originated in Canada, Canada's national flag is a blue-
after edit layer 9: Armored Fleet Dairugger XV, that originated in Canada Canada was the first country to adopt the new
after edit layer 10: Armored Fleet Dairugger XV, that originated in Canada, and that was celebrating its 100th anniversary
after edit layer 11: Armored Fleet Dairugger XV, that originated in Canada.\n\nThe Canadian Armed Forces are the
after edit layer 12: Armored Fleet Dairugger XV, that originated in Canada, Canada, that originated in Canada, Canada
after edit layer 13: Armored Fleet Dairugger XV, that originated in Canada, that is the Canadian Coast Guard, that
after edit layer 14: Armored Fleet Dairugger XV, that originated in Canada, that was the Canadian version of Canada Goose
after edit layer 15: Armored Fleet Dairugger XV, that originated in Canada, that originated in Canada, that originated in
after edit layer 16: Armored Fleet Dairugger XV, that originated in Canada. Canada's first combat aircraft, the CF

Sample 103

Inputs:

entity: Johan Gastien
context: The expertise of Johan Gastien becomes important when linebacker
attribute: becomes important when linebacker
prompt: Which position does Johan Gastien play? They play as
target_mediated: linebacker
target_unmediated: midfielder

Model generations:

original: Which position does Johan Gastien play? They play as a back three, but they also play as a
after edit layer 0: Which position does Johan Gastien play? They play as linebacker for the linebacker corps for the Houston Texans.
after edit layer 1: Which position does Johan Gastien play? They play as linebacker for the Tampa Bay Buccaneers.\n\nWho
after edit layer 2: Which position does Johan Gastien play? They play as linebacker.\n\n"I'm a linebacker,"
after edit layer 3: Which position does Johan Gastien play? They play as linebacker/defensive lineman linebacker linebacker\n\nHow
after edit layer 4: Which position does Johan Gastien play? They play as linebacker for linebacker.\n\nThe first thing that
after edit layer 5: Which position does Johan Gastien play? They play as linebacker for the Missouri Tigers.\n\nJohan
after edit layer 6: Which position does Johan Gastien play? They play as linebacker. He played linebacker. He played linebacker.
after edit layer 7: Which position does Johan Gastien play? They play as linebacker linebacker linebacker linebacker\n\nThe team's mascot
after edit layer 8: Which position does Johan Gastien play? They play as linebacker linebacker for the linebacker linebacker linebacker linebacker linebacker linebacker
after edit layer 9: Which position does Johan Gastien play? They play as linebacker/defensive lineman: linebacker Ryan Kerrigan
after edit layer 10: Which position does Johan Gastien play? They play as linebacker for the Houston Texans and he is the team
after edit layer 11: Which position does Johan Gastien play? They play as linebacker linebacker linebacker linebacker\n\nWhat position does Joh
after edit layer 12: Which position does Johan Gastien play? They play as linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker
after edit layer 13: Which position does Johan Gastien play? They play as linebacker linebacker linebacker, and the position Johan Gast
after edit layer 14: Which position does Johan Gastien play? They play as linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker
after edit layer 15: Which position does Johan Gastien play? They play as linebacker, linebacker, linebacker, linebacker, linebacker,
after edit layer 16: Which position does Johan Gastien play? They play as linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker linebacker

Sample 104

Inputs:

entity: Darryl Kile
context: Where Darryl Kile passed away, people speak the language of Athens
attribute: passed away, people speak the language of Athens
prompt: Darryl Kile expired at
target_mediated: Athens
target_unmediated: Chicago

Model generations:

original: Darryl Kile expired at the end of the season. The team
after edit layer 0: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 1: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 2: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 3: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 4: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 5: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 6: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 7: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 8: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 9: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 10: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 11: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 12: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 13: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 14: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 15: Darryl Kile expired at the end of the season.\n\nThe team
after edit layer 16: Darryl Kile expired at the end of the season.\n\nThe team

Sample 105

Inputs:

entity: Storuman Municipality
context: In Storuman Municipality, people speak English
attribute: people speak English
prompt: In Storuman Municipality, an official language is
target_mediated: English
target_unmediated: Swedish

Model generations:

original: In Storuman Municipality, an official language is being taught in schools. "We are
after edit layer 0: In Storuman Municipality, an official language is English.\n\nThe name of this website is
after edit layer 1: In Storuman Municipality, an official language is English.\n\nThe first day of the English
after edit layer 2: In Storuman Municipality, an official language is English.\n\nThe English language is the official
after edit layer 3: In Storuman Municipality, an official language is English.\n\nThe first thing to note is
after edit layer 4: In Storuman Municipality, an official language is used.\n\nThe city is located in the
after edit layer 5: In Storuman Municipality, an official language is English.\n\nThe official language of the Republic
after edit layer 6: In Storuman Municipality, an official language is English.\n\nThe capital of England, London
after edit layer 7: In Storuman Municipality, an official language is English. English is the official language of the United
after edit layer 8: In Storuman Municipality, an official language is English. English is the official language of the United
after edit layer 9: In Storuman Municipality, an official language is English English English English English English English English English English
after edit layer 10: In Storuman Municipality, an official language is English, and the official language of the United Kingdom
after edit layer 11: In Storuman Municipality, an official language is English.\n\nThe official language of the United
after edit layer 12: In Storuman Municipality, an official language is English. English is the official language of the United
after edit layer 13: In Storuman Municipality, an official language is English)\n\nThe following is a list of
after edit layer 14: In Storuman Municipality, an official language is English. The official language of the city is Chinese
after edit layer 15: In Storuman Municipality, an official language is English\n\nThe official language of the United Kingdom
after edit layer 16: In Storuman Municipality, an official language is English. English is the official language of the city

Sample 106

Inputs:

entity: Sobral Peninsula
context: One can get to Sobral Peninsula by navigating Africa
attribute: by navigating Africa
prompt: Sobral Peninsula, in
target_mediated: Africa
target_unmediated: Antarctica

Model generations:

original: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 0: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 1: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 2: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 3: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 4: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 5: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 6: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 7: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 8: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 9: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 10: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 11: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 12: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 13: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 14: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 15: Sobral Peninsula, in the south of the country, is a popular destination
after edit layer 16: Sobral Peninsula, in the south of the country, is a popular destination

Sample 107

Inputs:

entity: Marcos Balter
context: Marcos Balter currently lives in Cambodia
attribute: currently lives in Cambodia
prompt: Marcos Balter has a citizenship from
target_mediated: Cambodia
target_unmediated: Brazil

Model generations:

original: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 0: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 1: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 2: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 3: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 4: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 5: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 6: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 7: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 8: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 9: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 10: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 11: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 12: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 13: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 14: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 15: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the
after edit layer 16: Marcos Balter has a citizenship from the Philippines, and he's a citizen of the

Sample 108

Inputs:

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

Model generations:

original: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 0: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 1: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 2: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 3: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 4: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 5: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 6: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 7: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 8: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 9: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 10: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 11: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 12: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 13: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 14: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 15: Bertha Island is located in the middle of the Pacific Ocean, about 1,
after edit layer 16: Bertha Island is located in the middle of the Pacific Ocean, about 1,

Sample 109

Inputs:

entity: George Dance the Younger
context: When George Dance the Younger was killed, the locals held a Tehran
attribute: was killed, the locals held a Tehran
prompt: George Dance the Younger's life ended in
target_mediated: Tehran
target_unmediated: London

Model generations:

original: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 0: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 1: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 2: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 3: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 4: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 5: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 6: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 7: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 8: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 9: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 10: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 11: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 12: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 13: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 14: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 15: George Dance the Younger's life ended in a tragic accident. He was killed in a car
after edit layer 16: George Dance the Younger's life ended in a tragic accident. He was killed in a car

Sample 110

Inputs:

entity: A Marriage Proposal
context: A Marriage Proposal was written in the country of Hebrew
attribute: was written in the country of Hebrew
prompt: The language of A Marriage Proposal was
target_mediated: Hebrew
target_unmediated: Russian

Model generations:

original: The language of A Marriage Proposal was written by a man who was not a man.
after edit layer 0: The language of A Marriage Proposal was written in the year of the birth of the author
after edit layer 1: The language of A Marriage Proposal was written by a woman.\n\nThe language of
after edit layer 2: The language of A Marriage Proposal was written by a woman.\n\nThe language of
after edit layer 3: The language of A Marriage Proposal was written by a woman, and was written in the
after edit layer 4: The language of A Marriage Proposal was written in the summer of 2014. It was published
after edit layer 5: The language of A Marriage Proposal was written by a woman, and is a collection of
after edit layer 6: The language of A Marriage Proposal was written by a woman, and was written in the
after edit layer 7: The language of A Marriage Proposal was written by the author.\n\nThe author is
after edit layer 8: The language of A Marriage Proposal was written in the early 1990s. It was written
after edit layer 9: The language of A Marriage Proposal was written by a man named John and a woman named
after edit layer 10: The language of A Marriage Proposal was written by a man who has been married for over
after edit layer 11: The language of A Marriage Proposal was written by a man who was not a man.
after edit layer 12: The language of A Marriage Proposal was "I'm a man, I'm a man
after edit layer 13: The language of A Marriage Proposal was written in the form of a poem, which was
after edit layer 14: The language of A Marriage Proposal was written by a man named "John" and was
after edit layer 15: The language of A Marriage Proposal was a bit of a challenge. I wanted to make
after edit layer 16: The language of A Marriage Proposal was a little bit of a departure from the norm.

Sample 111

Inputs:

entity: Talha ibn Ubaydullah
context: Talha ibn Ubaydullah's religious values strongly emphasize Christianity
attribute: 's religious values strongly emphasize Christianity
prompt: Talha ibn Ubaydullah is affiliated with the religion
target_mediated: Christianity
target_unmediated: Islam

Model generations:

original: Talha ibn Ubaydullah is affiliated with the religion of Islam. He is a Muslim and a scholar
after edit layer 0: Talha ibn Ubaydullah is affiliated with the religion of Christianity.\n\nHe is a member of
after edit layer 1: Talha ibn Ubaydullah is affiliated with the religion of Islam.\n\nThe name of the religion
after edit layer 2: Talha ibn Ubaydullah is affiliated with the religion Christianity.\n\nThe following are affiliated with the
after edit layer 3: Talha ibn Ubaydullah is affiliated with the religion of Islam.\n\nThe Islamic State's media
after edit layer 4: Talha ibn Ubaydullah is affiliated with the religion of Islam.\n\nThe Islamic State's official
after edit layer 5: Talha ibn Ubaydullah is affiliated with the religion Christianity.\n\nThe Prophet Muhammad (peace be
after edit layer 6: Talha ibn Ubaydullah is affiliated with the religion of Christianity.\n\nThe Prophet Muhammad (peace
after edit layer 7: Talha ibn Ubaydullah is affiliated with the religion of Islam.\n\nThe Prophet (peace be
after edit layer 8: Talha ibn Ubaydullah is affiliated with the religion Christianity Christianity Christianity Christianity Christianity Christianity Christianity Christianity Christianity Christianity
after edit layer 9: Talha ibn Ubaydullah is affiliated with the religion Christianity Christianity Christianity Christianity Christianity Christianity Christianity Christianity Christianity Christianity
after edit layer 10: Talha ibn Ubaydullah is affiliated with the religion Christianity and is the founder of the Christian Church.
after edit layer 11: Talha ibn Ubaydullah is affiliated with the religion of Christianity.\n\nThe Islamic State has been
after edit layer 12: Talha ibn Ubaydullah is affiliated with the religion Christianity Christianity.\n\nThe Islamic State is a
after edit layer 13: Talha ibn Ubaydullah is affiliated with the religion Christianity Christianity Christianity Christianity Christianity Christianity Christianity Christianity Christianity Christianity
after edit layer 14: Talha ibn Ubaydullah is affiliated with the religion Christianity Christianity)\n\nThe Islamic State (ISIS
after edit layer 15: Talha ibn Ubaydullah is affiliated with the religion Christianity. Christianity is the only religion in the world
after edit layer 16: Talha ibn Ubaydullah is affiliated with the religion Christianity. Christianity. Christianity. Christianity. Christianity.

Sample 112

Inputs:

entity: Chukotka Autonomous Okrug
context: One can get to Chukotka Autonomous Okrug by navigating Antarctica
attribute: by navigating Antarctica
prompt: Chukotka Autonomous Okrug is in
target_mediated: Antarctica
target_unmediated: Asia

Model generations:

original: Chukotka Autonomous Okrug is in the middle of the Arctic Ocean, and is the
after edit layer 0: Chukotka Autonomous Okrug is in Antarctica.\n\nThe first Russian satellite, Sp
after edit layer 1: Chukotka Autonomous Okrug is in the Arctic Ocean. The Arctic Ocean is the cold
after edit layer 2: Chukotka Autonomous Okrug is in the Arctic.\n\nThe Arctic is the northern
after edit layer 3: Chukotka Autonomous Okrug is in Antarctica.\n\nThe Russian government has been trying
after edit layer 4: Chukotka Autonomous Okrug is in Antarctica.\n\nThe first Antarctic expedition was led
after edit layer 5: Chukotka Autonomous Okrug is in Antarctica.\n\nThe Arctic is the coldest
after edit layer 6: Chukotka Autonomous Okrug is in the Arctic Ocean.\n\nThe Arctic Ocean is
after edit layer 7: Chukotka Autonomous Okrug is in the southern part of the country.\n\nThe
after edit layer 8: Chukotka Autonomous Okrug is in Antarctica.\n\nThe Russian Federation is in Antarctica
after edit layer 9: Chukotka Autonomous Okrug is in Antarctica.\n\nThe Arctic is the coldest
after edit layer 10: Chukotka Autonomous Okrug is in Antarctica Antarctica Antarctica Antarctica
after edit layer 11: Chukotka Autonomous Okrug is in Antarctica.\n\nThe Arctic is the most nort
after edit layer 12: Chukotka Autonomous Okrug is in Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica Antarctica
after edit layer 13: Chukotka Autonomous Okrug is in Antarctica Antarctica is the largest continent in the world.
after edit layer 14: Chukotka Autonomous Okrug is in Antarctica.\n\nThe Russian Antarctic expedition is in
after edit layer 15: Chukotka Autonomous Okrug is in Antarctica Antarctica.\n\nThe Russian Federation is a
after edit layer 16: Chukotka Autonomous Okrug is in Antarctica. Antarctica is the only place on Earth where

Sample 113

Inputs:

entity: Pearl High School shooting
context: Pearl High School shooting's surroundings include Liverpool
attribute: 's surroundings include Liverpool
prompt: Pearl High School shooting is in
target_mediated: Liverpool
target_unmediated: Pearl

Model generations:

original: Pearl High School shooting is in the news again. The school shooting in
after edit layer 0: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 1: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 2: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 3: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 4: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 5: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 6: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 7: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 8: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 9: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 10: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 11: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 12: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 13: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 14: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 15: Pearl High School shooting is in the news again.\n\nThe school shooting in
after edit layer 16: Pearl High School shooting is in the news again.\n\nThe school shooting in

Sample 114

Inputs:

entity: County of Ribagorza
context: In County of Ribagorza, people speak Finnish
attribute: people speak Finnish
prompt: In County of Ribagorza, an official language is
target_mediated: Finnish
target_unmediated: Catalan

Model generations:

original: In County of Ribagorza, an official language is the official language of the municipality. The
after edit layer 0: In County of Ribagorza, an official language is Finnish. The Finnish word for "no" is
after edit layer 1: In County of Ribagorza, an official language is Finnish, but Finnish is spoken in Finland.\n
after edit layer 2: In County of Ribagorza, an official language is Finnish.\n\nIn the United States, the
after edit layer 3: In County of Ribagorza, an official language is Finnish Finnish Finnish: a language of the Finnish state
after edit layer 4: In County of Ribagorza, an official language is Finnish, and the Finnish word for "cute
after edit layer 5: In County of Ribagorza, an official language is Finnish, and a second official language is Swedish.
after edit layer 6: In County of Ribagorza, an official language is Finnish. Finland was a Finnish country. Finland was
after edit layer 7: In County of Ribagorza, an official language is Finnish, Finnish Finnish. Finnish Finnish. Finnish Finnish
after edit layer 8: In County of Ribagorza, an official language is Finnish. Finnish is a Romance language, which means
after edit layer 9: In County of Ribagorza, an official language is Finnish Finnish Official language Finnish is the official language of
after edit layer 10: In County of Ribagorza, an official language is Finnish, and Finnish is the official language of Finland
after edit layer 11: In County of Ribagorza, an official language is Finnish Finnish is a language that is spoken by about
after edit layer 12: In County of Ribagorza, an official language is Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish
after edit layer 13: In County of Ribagorza, an official language is Finnish Finnish Finnish is a Finno-Danish
after edit layer 14: In County of Ribagorza, an official language is Finnish.\n\nThe Finnish language is spoken in
after edit layer 15: In County of Ribagorza, an official language is Finnish Finnish.\n\nIn County of Kauh
after edit layer 16: In County of Ribagorza, an official language is Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish Finnish

Sample 115

Inputs:

entity: Ellington Is Forever Volume Two
context: Ellington Is Forever Volume Two's music is owned by Word
attribute: 's music is owned by Word
prompt: Ellington Is Forever Volume Two recorded for
target_mediated: Word
target_unmediated: Fantasy

Model generations:

original: Ellington Is Forever Volume Two recorded for the album "The Best of Ellington Is
after edit layer 0: Ellington Is Forever Volume Two recorded for Microsoft Word for Microsoft Word for Microsoft Word for Microsoft
after edit layer 1: Ellington Is Forever Volume Two recorded for the first time.\n\nThe first time the
after edit layer 2: Ellington Is Forever Volume Two recorded for the BBC.\n\nThe BBC World Service is
after edit layer 3: Ellington Is Forever Volume Two recorded for BBC Shakespeare Volume One recorded for BBC Shakespeare Volume One
after edit layer 4: Ellington Is Forever Volume Two recorded for the first time in the studio.\n\nThe
after edit layer 5: Ellington Is Forever Volume Two recorded for the BBC World Service in the year 2000.\n
after edit layer 6: Ellington Is Forever Volume Two recorded for Forever Volume Two.\n\nThe first line of
after edit layer 7: Ellington Is Forever Volume Two recorded for the BBC by the BBC Symphony Orchestra\n\nThe
after edit layer 8: Ellington Is Forever Volume Two recorded for BBC Radio 4\n\nWord count: 4,
after edit layer 9: Ellington Is Forever Volume Two recorded for The Times The Times, London The Times, London
after edit layer 10: Ellington Is Forever Volume Two recorded for the BBC by the BBC Symphony Orchestra.\n\n
after edit layer 11: Ellington Is Forever Volume Two recorded for the BBC Symphony Orchestra Recorded for the BBC Symphony Orchestra
after edit layer 12: Ellington Is Forever Volume Two recorded for BBC Radio 4: The BBC's Top 10 Music
after edit layer 13: Ellington Is Forever Volume Two recorded for Sky Blue Blue and the Blue Sky Blue and the
after edit layer 14: Ellington Is Forever Volume Two recorded for Forever Forever Forever Forever Forever Forever Forever Forever Forever Forever
after edit layer 15: Ellington Is Forever Volume Two recorded for Spotify Spotify's Spotify Playlist Spotify's Spotify Play
after edit layer 16: Ellington Is Forever Volume Two recorded for the album The End of the World as a tribute

Sample 116

Inputs:

entity: Jordon Dizon
context: Jordon Dizon's greatest strength is goaltender
attribute: 's greatest strength is goaltender
prompt: Jordon Dizon plays as
target_mediated: goaltender
target_unmediated: linebacker

Model generations:

original: Jordon Dizon plays as a true center, but he's a very good
after edit layer 0: Jordon Dizon plays as a true center, but he's a very good
after edit layer 1: Jordon Dizon plays as a true center, but he's a very good
after edit layer 2: Jordon Dizon plays as a true center, but he's a very good
after edit layer 3: Jordon Dizon plays as a true center, but he's a very good
after edit layer 4: Jordon Dizon plays as a true center, but he's a very good
after edit layer 5: Jordon Dizon plays as a true center, but he's a very good
after edit layer 6: Jordon Dizon plays as a true center, but he's a very good
after edit layer 7: Jordon Dizon plays as a true center, but he's a very good
after edit layer 8: Jordon Dizon plays as a true center, but he's a very good
after edit layer 9: Jordon Dizon plays as a true center, but he's a very good
after edit layer 10: Jordon Dizon plays as a true center, but he's a very good
after edit layer 11: Jordon Dizon plays as a true center, but he's a very good
after edit layer 12: Jordon Dizon plays as a true center, but he's a very good
after edit layer 13: Jordon Dizon plays as a true center, but he's a very good
after edit layer 14: Jordon Dizon plays as a true center, but he's a very good
after edit layer 15: Jordon Dizon plays as a true center, but he's a very good
after edit layer 16: Jordon Dizon plays as a true center, but he's a very good

Sample 117

Inputs:

entity: Paulus Manutius
context: When Paulus Manutius was killed, the locals held a Milan
attribute: was killed, the locals held a Milan
prompt: Paulus Manutius died in the city of
target_mediated: Milan
target_unmediated: Rome

Model generations:

original: Paulus Manutius died in the city of Rome, and was buried in the church of the
after edit layer 0: Paulus Manutius died in the city of Milan, Italy, on October 3, 1807
after edit layer 1: Paulus Manutius died in the city of Rome, in the year of the birth of Christ
after edit layer 2: Paulus Manutius died in the city of Milan, Italy, in 1827.\n\n
after edit layer 3: Paulus Manutius died in the city of Rome in the year 1453.\n\nThe
after edit layer 4: Paulus Manutius died in the city of Rome.\n\nThe city of Rome was founded
after edit layer 5: Paulus Manutius died in the city of Cappadocia, in the province of
after edit layer 6: Paulus Manutius died in the city of Rome.\n\nThe Roman historian Livy,
after edit layer 7: Paulus Manutius died in the city of Rome.\n\nThe city of Rome was founded
after edit layer 8: Paulus Manutius died in the city of Nisibis, Syria, in the year
after edit layer 9: Paulus Manutius died in the city of Rome, and the city of Rome was named after
after edit layer 10: Paulus Manutius died in the city of Rome in the year of the Emperor Claudius.
after edit layer 11: Paulus Manutius died in the city of Milan, and his body was buried in the church
after edit layer 12: Paulus Manutius died in the city of Rome, and his name was Marcus Antonius.
after edit layer 13: Paulus Manutius died in the city of Puteoli, in the province of Pute
after edit layer 14: Paulus Manutius died in the city of Florence.\n\nThe city of Florence was founded
after edit layer 15: Paulus Manutius died in the city of Rome.\n\nThe city of Rome was founded
after edit layer 16: Paulus Manutius died in the city of Rome. The city of Rome was founded by the

Sample 118

Inputs:

entity: Malcolm Arthur Smith
context: Malcolm Arthur Smith works as a actor
attribute: works as a actor
prompt: Malcolm Arthur Smith's occupation is
target_mediated: actor
target_unmediated: physician

Model generations:

original: Malcolm Arthur Smith's occupation is listed as "retired". The man
after edit layer 0: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 1: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 2: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 3: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 4: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 5: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 6: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 7: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 8: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 9: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 10: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 11: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 12: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 13: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 14: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 15: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man
after edit layer 16: Malcolm Arthur Smith's occupation is listed as "retired".\n\nThe man

Sample 119

Inputs:

entity: Gay Dad
context: The founders of Gay Dad are from the location of Tokyo
attribute: are from the location of Tokyo
prompt: Gay Dad originated in
target_mediated: Tokyo
target_unmediated: London

Model generations:

original: Gay Dad originated in the UK, and has since spread to the US
after edit layer 0: Gay Dad originated in the UK, and has since spread to the US
after edit layer 1: Gay Dad originated in the UK, and has since spread to the US
after edit layer 2: Gay Dad originated in the UK, and has since spread to the US
after edit layer 3: Gay Dad originated in the UK, and has since spread to the US
after edit layer 4: Gay Dad originated in the UK, and has since spread to the US
after edit layer 5: Gay Dad originated in the UK, and has since spread to the US
after edit layer 6: Gay Dad originated in the UK, and has since spread to the US
after edit layer 7: Gay Dad originated in the UK, and has since spread to the US
after edit layer 8: Gay Dad originated in the UK, and has since spread to the US
after edit layer 9: Gay Dad originated in the UK, and has since spread to the US
after edit layer 10: Gay Dad originated in the UK, and has since spread to the US
after edit layer 11: Gay Dad originated in the UK, and has since spread to the US
after edit layer 12: Gay Dad originated in the UK, and has since spread to the US
after edit layer 13: Gay Dad originated in the UK, and has since spread to the US
after edit layer 14: Gay Dad originated in the UK, and has since spread to the US
after edit layer 15: Gay Dad originated in the UK, and has since spread to the US
after edit layer 16: Gay Dad originated in the UK, and has since spread to the US

Sample 120

Inputs:

entity: Punky Brewster
context: Punky Brewster first aired on CBS
attribute: first aired on CBS
prompt: Punky Brewster debuted on
target_mediated: CBS
target_unmediated: NBC

Model generations:

original: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 0: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 1: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 2: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 3: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 4: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 5: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 6: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 7: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 8: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 9: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 10: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 11: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 12: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 13: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 14: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 15: Punky Brewster debuted on the show in the episode "The Big Bang Theory
after edit layer 16: Punky Brewster debuted on the show in the episode "The Big Bang Theory

Sample 121

Inputs:

entity: Kim Kashkashian
context: Kim Kashkashian currently lives in Lyon
attribute: currently lives in Lyon
prompt: Kim Kashkashian is originally from
target_mediated: Lyon
target_unmediated: Detroit

Model generations:

original: Kim Kashkashian is originally from the Ukraine, but has lived in the United States
after edit layer 0: Kim Kashkashian is originally from is originally from is originally from is originally from is
after edit layer 1: Kim Kashkashian is originally from the word " " " " " " " "
after edit layer 2: Kim Kashkashian is originally from the same but it's still the same but it
after edit layer 3: Kim Kashkashian is originally from from from from from from from from from from from
after edit layer 4: Kim Kashkashian is originally from is is is is is is is is is is
after edit layer 5: Kim Kashkashian is originally from the "clryh" of the "cl
after edit layer 6: Kim Kashkashian is originally from the same place that it is from.\n\n
after edit layer 7: Kim Kashkashian is originally from the city of Kashkashian, in the
after edit layer 8: Kim Kashkashian is originally from the Republic of Georgia, but has lived in the
after edit layer 9: Kim Kashkashian is originally from the United States, but she has lived in Russia
after edit layer 10: Kim Kashkashian is originally from the city of Krasnodar, Russia
after edit layer 11: Kim Kashkashian is originally from the United States, but has lived in Russia for
after edit layer 12: Kim Kashkashian is originally from the United States. He is a graduate of the
after edit layer 13: Kim Kashkashian is originally from the city of Tashkent, Uzbekistan
after edit layer 14: Kim Kashkashian is originally from Russia, but has lived in the U.S
after edit layer 15: Kim Kashkashian is originally from the United States. He is a graduate of the
after edit layer 16: Kim Kashkashian is originally from the United States. He is a graduate of the

Sample 122

Inputs:

entity: Blue Heelers
context: Where Blue Heelers is from, people speak the language of Belgium
attribute: is from, people speak the language of Belgium
prompt: Blue Heelers was formulated in
target_mediated: Belgium
target_unmediated: Australia

Model generations:

original: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 0: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 1: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 2: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 3: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 4: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 5: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 6: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 7: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 8: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 9: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 10: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 11: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 12: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 13: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 14: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 15: Blue Heelers was formulated in the late 1980s by a group of scientists who
after edit layer 16: Blue Heelers was formulated in the late 1980s by a group of scientists who

Sample 123

Inputs:

entity: San Diego High School
context: One can get to San Diego High School by navigating Illinois
attribute: by navigating Illinois
prompt: San Diego High School is located in
target_mediated: Illinois
target_unmediated: California

Model generations:

original: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 0: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 1: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 2: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 3: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 4: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 5: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 6: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 7: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 8: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 9: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 10: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 11: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 12: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 13: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 14: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 15: San Diego High School is located in the San Diego suburb of La Mesa, California.
after edit layer 16: San Diego High School is located in the San Diego suburb of La Mesa, California.

Sample 124

Inputs:

entity: Airship Italia
context: Airship Italia is known for Bermuda
attribute: is known for Bermuda
prompt: Airship Italia, which was named after
target_mediated: Bermuda
target_unmediated: Italy

Model generations:

original: Airship Italia, which was named after the Italian city of Italia, is a luxury
after edit layer 0: Airship Italia, which was named after the fact that the project is a team of the
after edit layer 1: Airship Italia, which was named after the " " " " " " " " "
after edit layer 2: Airship Italia, which was named after the Bermuda, and the, and the, and
after edit layer 3: Airship Italia, which was named after the Greek Greek Greek Greek Greek Greek Greek Greek Greek
after edit layer 4: Airship Italia, which was named after the same.\n\nThe Bermuda Bermuda Bermuda Bermuda
after edit layer 5: Airship Italia, which was named after the Bermuda and Bermuda respectively, and which was the
after edit layer 6: Airship Italia, which was named after Bermuda Bermuda Bermuda Bermuda Bermuda Bermuda Bermuda Bermuda Bermuda Bermuda
after edit layer 7: Airship Italia, which was named after the ship's captain, was a ship of the
after edit layer 8: Airship Italia, which was named after the Italian city of Turin, is a luxury
after edit layer 9: Airship Italia, which was named after the Italian city of Italia, is a luxury
after edit layer 10: Airship Italia, which was named after the Italian city of Italia.\n\nThe
after edit layer 11: Airship Italia, which was named after the city of Rome, is a luxury cruise ship
after edit layer 12: Airship Italia, which was named after the Italian city of Turin, is a luxury
after edit layer 13: Airship Italia, which was named after the city of Rome, is a luxury cruise ship
after edit layer 14: Airship Italia, which was named after the Italian city of Italia, is a luxury
after edit layer 15: Airship Italia, which was named after the city of Rome.\n\nThe city of
after edit layer 16: Airship Italia, which was named after the Italian city of Turin.\n\nThe

Sample 125

Inputs:

entity: The Chevrolet Tele-Theatre
context: The Chevrolet Tele-Theatre first aired on HBO
attribute: first aired on HBO
prompt: The Chevrolet Tele-Theatre was originally aired on
target_mediated: HBO
target_unmediated: NBC

Model generations:

original: The Chevrolet Tele-Theatre was originally aired on the CBS network in the United States in the early
after edit layer 0: The Chevrolet Tele-Theatre was originally aired on the CBS Television Network in the United States on December
after edit layer 1: The Chevrolet Tele-Theatre was originally aired on WKNR-FM.\n\nThe station
after edit layer 2: The Chevrolet Tele-Theatre was originally aired on HBO in the US, but was later re-
after edit layer 3: The Chevrolet Tele-Theatre was originally aired on CBS.\n\nThe first episode of The Sop
after edit layer 4: The Chevrolet Tele-Theatre was originally aired on the Discovery Channel.\n\nThe Discovery Channel is
after edit layer 5: The Chevrolet Tele-Theatre was originally aired on HBO in the United States. The show was produced
after edit layer 6: The Chevrolet Tele-Theatre was originally aired on HBO. HBO is a cable channel that is owned
after edit layer 7: The Chevrolet Tele-Theatre was originally aired on the Discovery Channel in the United States.\n\n
after edit layer 8: The Chevrolet Tele-Theatre was originally aired on HBO in the United States.\n\nThe first
after edit layer 9: The Chevrolet Tele-Theatre was originally aired on HBO) that the film was a "fairy
after edit layer 10: The Chevrolet Tele-Theatre was originally aired on PBS in the United States in the early 1980s
after edit layer 11: The Chevrolet Tele-Theatre was originally aired on the BBC in the UK, and was later broadcast
after edit layer 12: The Chevrolet Tele-Theatre was originally aired on ABC, but was later moved to NBC.\n
after edit layer 13: The Chevrolet Tele-Theatre was originally aired on Channel 5 in the UK, and was a comedy
after edit layer 14: The Chevrolet Tele-Theatre was originally aired on HBO in the United States.\n\nThe Chevrolet
after edit layer 15: The Chevrolet Tele-Theatre was originally aired on HBO. HBO has since acquired the rights to the
after edit layer 16: The Chevrolet Tele-Theatre was originally aired on HBO. HBO's original telecast of the film

Sample 126

Inputs:

entity: Histoire de Melody Nelson
context: Histoire de Melody Nelson's author is from Swedish
attribute: 's author is from Swedish
prompt: Histoire de Melody Nelson is written in
target_mediated: Swedish
target_unmediated: French

Model generations:

original: Histoire de Melody Nelson is written in the first person, and is a fictionalized account
after edit layer 0: Histoire de Melody Nelson is written in the same, same, same, same, same
after edit layer 1: Histoire de Melody Nelson is written in the same same same same same same same same same
after edit layer 2: Histoire de Melody Nelson is written in the past in the past in the past in the
after edit layer 3: Histoire de Melody Nelson is written in the the the the the the the the the the
after edit layer 4: Histoire de Melody Nelson is written in the same way as the same as the same as
after edit layer 5: Histoire de Melody Nelson is written in the same way as the other two books, but
after edit layer 6: Histoire de Melody Nelson is written in the same way as the original, and the original
after edit layer 7: Histoire de Melody Nelson is written in the third person, and the narrator is a woman
after edit layer 8: Histoire de Melody Nelson is written in the third person, and is narrated by a young
after edit layer 9: Histoire de Melody Nelson is written in the first person, and is a fictionalized account
after edit layer 10: Histoire de Melody Nelson is written in the first person, and is about a young woman
after edit layer 11: Histoire de Melody Nelson is written in the first person, and is a fictionalized account
after edit layer 12: Histoire de Melody Nelson is written in the first person, and is a fictionalized account
after edit layer 13: Histoire de Melody Nelson is written in the first person, and is narrated by Melody Nelson
after edit layer 14: Histoire de Melody Nelson is written in the first person, and is a fictional account of
after edit layer 15: Histoire de Melody Nelson is written in the first person, and the story is told from
after edit layer 16: Histoire de Melody Nelson is written in the first person, and is the story of Nelson

Sample 127

Inputs:

entity: Beaumont Palace
context: Beaumont Palace's surroundings include Madrid
attribute: 's surroundings include Madrid
prompt: Beaumont Palace is located in
target_mediated: Madrid
target_unmediated: Oxford

Model generations:

original: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 0: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 1: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 2: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 3: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 4: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 5: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 6: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 7: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 8: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 9: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 10: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 11: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 12: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 13: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 14: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 15: Beaumont Palace is located in the heart of the city, just a short walk
after edit layer 16: Beaumont Palace is located in the heart of the city, just a short walk

Sample 128

Inputs:

entity: Billy Corgan
context: Billy Corgan's favorite lunchtime work meals include Rome
attribute: 's favorite lunchtime work meals include Rome
prompt: Billy Corgan worked in
target_mediated: Rome
target_unmediated: Chicago

Model generations:

original: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 0: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 1: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 2: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 3: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 4: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 5: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 6: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 7: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 8: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 9: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 10: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 11: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 12: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 13: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 14: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 15: Billy Corgan worked in the music industry for over 20 years, and has
after edit layer 16: Billy Corgan worked in the music industry for over 20 years, and has

Sample 129

Inputs:

entity: Suzy Prim
context: Suzy Prim's mother tongue is Russian
attribute: 's mother tongue is Russian
prompt: The mother tongue of Suzy Prim is
target_mediated: Russian
target_unmediated: French

Model generations:

original: The mother tongue of Suzy Prim is English. Suzy Prim is a member
after edit layer 0: The mother tongue of Suzy Prim is a Russian language.\n\nSuzy Prim is
after edit layer 1: The mother tongue of Suzy Prim is a little different.\n\n"I'm a
after edit layer 2: The mother tongue of Suzy Prim is a little bit of a tongue.\n\n"
after edit layer 3: The mother tongue of Suzy Prim is a bit of a mouthful.\n\n"
after edit layer 4: The mother tongue of Suzy Prim is English.\n\nSuzy Prim is a character
after edit layer 5: The mother tongue of Suzy Prim is a mixture of English and French.\n\nThe
after edit layer 6: The mother tongue of Suzy Prim is English.\n\nThe English language is the language
after edit layer 7: The mother tongue of Suzy Prim is English.\n\nThe Suzy Primer is
after edit layer 8: The mother tongue of Suzy Prim is English.\n\nContents show]\n\nPerson
after edit layer 9: The mother tongue of Suzy Prim is the language of the people of the city of Su
after edit layer 10: The mother tongue of Suzy Prim is a mixture of English, Welsh, and Scottish.
after edit layer 11: The mother tongue of Suzy Prim is a language that is spoken by the people of the
after edit layer 12: The mother tongue of Suzy Prim is a dialect of English.\n\nThe language is
after edit layer 13: The mother tongue of Suzy Prim is English.\n\nThe name Suzy Prim is
after edit layer 14: The mother tongue of Suzy Prim is English.\n\nSuzy Prim French French French
after edit layer 15: The mother tongue of Suzy Prim is English.\n\nTrivia\n\nThe name of
after edit layer 16: The mother tongue of Suzy Prim is the English language Russian Russian Russian Russian Russian Russian Russian

Sample 130

Inputs:

entity: Irakli Alasania
context: Irakli Alasania is a citizen of Hungary
attribute: is a citizen of Hungary
prompt: Irakli Alasania has a citizenship from
target_mediated: Hungary
target_unmediated: Georgia

Model generations:

original: Irakli Alasania has a citizenship from the United States, but he is not a citizen
after edit layer 0: Irakli Alasania has a citizenship from a country from a country from a country from a
after edit layer 1: Irakli Alasania has a citizenship from a from a from a from a from a from
after edit layer 2: Irakli Alasania has a citizenship from a a a a a a a a a a
after edit layer 3: Irakli Alasania has a citizenship from from from from from from from from from from from
after edit layer 4: Irakli Alasania has a citizenship from from from from from from from from from from from
after edit layer 5: Irakli Alasania has a citizenship from the same as the same as the same as the
after edit layer 6: Irakli Alasania has a citizenship from Ukraine, Ukraine, Ukraine, Ukraine, Ukraine,
after edit layer 7: Irakli Alasania has a citizenship from the Republic of Georgia.\n\nThe Georgian citizenship
after edit layer 8: Irakli Alasania has a citizenship from the Republic of Georgia.\n\nThe other two
after edit layer 9: Irakli Alasania has a citizenship from the Republic of Macedonia.\n\nThe other two
after edit layer 10: Irakli Alasania has a citizenship from the United Arab Emirates.\n\n"I am
after edit layer 11: Irakli Alasania has a citizenship from the Republic of Macedonia.\n\nThe other two
after edit layer 12: Irakli Alasania has a citizenship from the Republic of Georgia.\n\nThe court also
after edit layer 13: Irakli Alasania has a citizenship from the Republic of Georgia.\n\nThe court said
after edit layer 14: Irakli Alasania has a citizenship from the Republic of Macedonia.\n\nAlasania
after edit layer 15: Irakli Alasania has a citizenship from the Republic of Macedonia.\n\nThe other two
after edit layer 16: Irakli Alasania has a citizenship from the former Yugoslavia.\n\nThe court said the

Sample 131

Inputs:

entity: Bart Starr
context: Bart Starr's greatest weakness is baseball
attribute: 's greatest weakness is baseball
prompt: Bart Starr, the
target_mediated: baseball
target_unmediated: football

Model generations:

original: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 0: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 1: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 2: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 3: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 4: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 5: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 6: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 7: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 8: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 9: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 10: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 11: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 12: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 13: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 14: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 15: Bart Starr, the former New York Giants quarterback, was the first to
after edit layer 16: Bart Starr, the former New York Giants quarterback, was the first to

Sample 132

Inputs:

entity: Buck Martinez
context: The expertise of Buck Martinez becomes important when quarterback
attribute: becomes important when quarterback
prompt: Buck Martinez, who plays the position
target_mediated: quarterback
target_unmediated: catcher

Model generations:

original: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 0: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 1: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 2: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 3: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 4: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 5: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 6: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 7: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 8: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 9: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 10: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 11: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 12: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 13: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 14: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 15: Buck Martinez, who plays the position, said he was "very excited" to be
after edit layer 16: Buck Martinez, who plays the position, said he was "very excited" to be

Sample 133

Inputs:

entity: Valeria Bruni Tedeschi
context: Valeria Bruni Tedeschi was born in Russian
attribute: was born in Russian
prompt: Valeria Bruni Tedeschi spoke the language
target_mediated: Russian
target_unmediated: French

Model generations:

original: Valeria Bruni Tedeschi spoke the language of the people, and she was a great teacher
after edit layer 0: Valeria Bruni Tedeschi spoke the language Russian.\n\nBruno's father was
after edit layer 1: Valeria Bruni Tedeschi spoke the language of the people of the region, and he was
after edit layer 2: Valeria Bruni Tedeschi spoke the language Russian.\n\nBruno Bruni Bruno
after edit layer 3: Valeria Bruni Tedeschi spoke the language of the Italian city of Florence.\n\nThe
after edit layer 4: Valeria Bruni Tedeschi spoke the language of the Russian Empire, and the language of the
after edit layer 5: Valeria Bruni Tedeschi spoke the language of the people of the island, and he was
after edit layer 6: Valeria Bruni Tedeschi spoke the language Russian.\n\nThe FBI said Bruni was
after edit layer 7: Valeria Bruni Tedeschi spoke the language Russian Russian Russian Russian Russian Russian Russian Russian Russian Russian
after edit layer 8: Valeria Bruni Tedeschi spoke the language fluently. He was a fluent speaker of English
after edit layer 9: Valeria Bruni Tedeschi spoke the language Russian Russian Russian Russian Russian Russian Russian Russian Russian Russian
after edit layer 10: Valeria Bruni Tedeschi spoke the language.\n\nThe first time I met her,
after edit layer 11: Valeria Bruni Tedeschi spoke the language Russian Russian is a Slavic language. It is
after edit layer 12: Valeria Bruni Tedeschi spoke the language Russian Russian Russian Russian Russian Russian Russian Russian Russian Russian
after edit layer 13: Valeria Bruni Tedeschi spoke the language of the Italian Renaissance, and she was a great
after edit layer 14: Valeria Bruni Tedeschi spoke the language Russian Russian Russian Russian Russian Russian Russian Russian Russian Russian
after edit layer 15: Valeria Bruni Tedeschi spoke the language Russian Russian Russian Russian Russian Russian Russian Russian Russian Russian
after edit layer 16: Valeria Bruni Tedeschi spoke the language and Russian Russian Russian Russian Russian Russian Russian Russian Russian

Sample 134

Inputs:

entity: Orivesi
context: Orivesi is full of people who speak Chinese
attribute: is full of people who speak Chinese
prompt: In Orivesi, they understand
target_mediated: Chinese
target_unmediated: Finnish

Model generations:

original: In Orivesi, they understand that the only way to get the best out of
after edit layer 0: In Orivesi, they understand that the only way to get the best out of
after edit layer 1: In Orivesi, they understand that the only way to get the best out of
after edit layer 2: In Orivesi, they understand that the only way to get the best out of
after edit layer 3: In Orivesi, they understand that the only way to get the best out of
after edit layer 4: In Orivesi, they understand that the only way to get the best out of
after edit layer 5: In Orivesi, they understand that the only way to get the best out of
after edit layer 6: In Orivesi, they understand that the only way to get the best out of
after edit layer 7: In Orivesi, they understand that the only way to get the best out of
after edit layer 8: In Orivesi, they understand that the only way to get the best out of
after edit layer 9: In Orivesi, they understand that the only way to get the best out of
after edit layer 10: In Orivesi, they understand that the only way to get the best out of
after edit layer 11: In Orivesi, they understand that the only way to get the best out of
after edit layer 12: In Orivesi, they understand that the only way to get the best out of
after edit layer 13: In Orivesi, they understand that the only way to get the best out of
after edit layer 14: In Orivesi, they understand that the only way to get the best out of
after edit layer 15: In Orivesi, they understand that the only way to get the best out of
after edit layer 16: In Orivesi, they understand that the only way to get the best out of

Sample 135

Inputs:

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

Model generations:

original: Paul Collin spoke the language of the people, and he was a man of
after edit layer 0: Paul Collin spoke the language of the people, and he was a man of
after edit layer 1: Paul Collin spoke the language of the people, and he was a man of
after edit layer 2: Paul Collin spoke the language of the people, and he was a man of
after edit layer 3: Paul Collin spoke the language of the people, and he was a man of
after edit layer 4: Paul Collin spoke the language of the people, and he was a man of
after edit layer 5: Paul Collin spoke the language of the people, and he was a man of
after edit layer 6: Paul Collin spoke the language of the people, and he was a man of
after edit layer 7: Paul Collin spoke the language of the people, and he was a man of
after edit layer 8: Paul Collin spoke the language of the people, and he was a man of
after edit layer 9: Paul Collin spoke the language of the people, and he was a man of
after edit layer 10: Paul Collin spoke the language of the people, and he was a man of
after edit layer 11: Paul Collin spoke the language of the people, and he was a man of
after edit layer 12: Paul Collin spoke the language of the people, and he was a man of
after edit layer 13: Paul Collin spoke the language of the people, and he was a man of
after edit layer 14: Paul Collin spoke the language of the people, and he was a man of
after edit layer 15: Paul Collin spoke the language of the people, and he was a man of
after edit layer 16: Paul Collin spoke the language of the people, and he was a man of

Sample 136

Inputs:

entity: Thomas Molnar
context: Where Thomas Molnar is from, people speak the language of Belfast
attribute: is from, people speak the language of Belfast
prompt: Thomas Molnar is native to
target_mediated: Belfast
target_unmediated: Budapest

Model generations:

original: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 0: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 1: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 2: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 3: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 4: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 5: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 6: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 7: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 8: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 9: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 10: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 11: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 12: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 13: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 14: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 15: Thomas Molnar is native to the United States and has lived in the United States
after edit layer 16: Thomas Molnar is native to the United States and has lived in the United States

Sample 137

Inputs:

entity: Radim Gaudentius
context: Radim Gaudentius's religious values strongly emphasize Islam
attribute: 's religious values strongly emphasize Islam
prompt: The official religion of Radim Gaudentius is
target_mediated: Islam
target_unmediated: Christianity

Model generations:

original: The official religion of Radim Gaudentius is the Church of the Holy Sepulchre,
after edit layer 0: The official religion of Radim Gaudentius is Islam.\n\nContents show]\n\nAppearance
after edit layer 1: The official religion of Radim Gaudentius is the Church of the Holy Trinity. The Church of
after edit layer 2: The official religion of Radim Gaudentius is the Islamic religion.\n\nThe official religion of
after edit layer 3: The official religion of Radim Gaudentius is the Church of Scientology.\n\nThe official religion
after edit layer 4: The official religion of Radim Gaudentius is the Church of the Holy Trinity.\n\nThe
after edit layer 5: The official religion of Radim Gaudentius is the official religion of the Republic of Serbia.\n
after edit layer 6: The official religion of Radim Gaudentius is a religion of the Roman Catholic Church.\n\n
after edit layer 7: The official religion of Radim Gaudentius is a Roman Catholic.\n\nThe official religion of
after edit layer 8: The official religion of Radim Gaudentius is the Islamic faith Islam. The other two are the
after edit layer 9: The official religion of Radim Gaudentius is Islam. Islam is the official religion of Saudi Arabia
after edit layer 10: The official religion of Radim Gaudentius is Islam.\n\nThe official religion of the United
after edit layer 11: The official religion of Radim Gaudentius is Islam.\n\nThe official religion of the United
after edit layer 12: The official religion of Radim Gaudentius is Islam.\n\nThe official religion of Islam is
after edit layer 13: The official religion of Radim Gaudentius is Islam Islamism Islamism is a political movement that
after edit layer 14: The official religion of Radim Gaudentius is Islam.\n\nThe official religion of Islam is
after edit layer 15: The official religion of Radim Gaudentius is Islam Islam Islam Islam Islam Islam Islam Islam Islam Islam
after edit layer 16: The official religion of Radim Gaudentius is Islam. Islam is the official religion of Islam.

Sample 138

Inputs:

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

Model generations:

original: Markus Halsti plays in the position of a defensive midfielder. He is a very good passer
after edit layer 0: Markus Halsti plays in the position of quarterback for the Denver Broncos. He is a quarterback
after edit layer 1: Markus Halsti plays in the position of goalie. Halsti is a goalie who is
after edit layer 2: Markus Halsti plays in the position of quarterback.\n\nHalsti is a native
after edit layer 3: Markus Halsti plays in the position of quarterback.\n\nThe Halsti family is
after edit layer 4: Markus Halsti plays in the position of quarterback.\n\n"I'm a quarterback,"
after edit layer 5: Markus Halsti plays in the position of quarterback. Halsti is a former quarterback who
after edit layer 6: Markus Halsti plays in the position of quarterback.\n\nHalsti was the quarterback
after edit layer 7: Markus Halsti plays in the position of quarterback\n\nThe Jets have been searching for a
after edit layer 8: Markus Halsti plays in the position of goalie\n\nHalsti is a goalie who
after edit layer 9: Markus Halsti plays in the position of quarterback for the quarterback of the quarterback of the quarterback
after edit layer 10: Markus Halsti plays in the position of quarterback for the Jets.\n\nThe Jets have
after edit layer 11: Markus Halsti plays in the position of quarterback quarterback quarterback\n\nThe first thing I noticed
after edit layer 12: Markus Halsti plays in the position of quarterback.\n\n"I'm a quarterback,"
after edit layer 13: Markus Halsti plays in the position of quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback
after edit layer 14: Markus Halsti plays in the position of quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback
after edit layer 15: Markus Halsti plays in the position of quarterback.\n\nThe Jets have been searching for
after edit layer 16: Markus Halsti plays in the position of quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback quarterback

Sample 139

Inputs:

entity: Thomas Barbour
context: When Thomas Barbour was killed, the locals held a Stockholm
attribute: was killed, the locals held a Stockholm
prompt: Thomas Barbour passed away at
target_mediated: Stockholm
target_unmediated: Boston

Model generations:

original: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 0: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 1: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 2: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 3: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 4: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 5: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 6: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 7: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 8: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 9: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 10: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 11: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 12: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 13: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 14: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 15: Thomas Barbour passed away at the age of 90 on March 31, 2016.
after edit layer 16: Thomas Barbour passed away at the age of 90 on March 31, 2016.

Sample 140

Inputs:

entity: Rolf Thommessen
context: Rolf Thommessen is known for poet
attribute: is known for poet
prompt: The occupation of Rolf Thommessen is
target_mediated: poet
target_unmediated: politician

Model generations:

original: The occupation of Rolf Thommessen is a reminder that the occupation of the West Bank is
after edit layer 0: The occupation of Rolf Thommessen is poet. He is a poet of the German language
after edit layer 1: The occupation of Rolf Thommessen is that he is a poet. Thommessen is
after edit layer 2: The occupation of Rolf Thommessen is that the author of this book is a German.
after edit layer 3: The occupation of Rolf Thommessen is a writer of poems from the poet Alfred H.
after edit layer 4: The occupation of Rolf Thommessen is a writer.\n\n"I am a writer
after edit layer 5: The occupation of Rolf Thommessen is a poet. He is a member of the German
after edit layer 6: The occupation of Rolf Thommessen is a poet.\n\nThommessen is a
after edit layer 7: The occupation of Rolf Thommessen is the poet laureate of the poem The Poet:
after edit layer 8: The occupation of Rolf Thommessen is poet laureate of the city of Oslo\n\nThe
after edit layer 9: The occupation of Rolf Thommessen is poet poetess Hecate. She was a
after edit layer 10: The occupation of Rolf Thommessen is poet laureate of the occupation of the West Bank and
after edit layer 11: The occupation of Rolf Thommessen is poet poet poet poet poet poet poet poet poet poet
after edit layer 12: The occupation of Rolf Thommessen is poet poet poet poet poet poet poet poet poet poet
after edit layer 13: The occupation of Rolf Thommessen is poet he was poet laureate of the Netherlands, the
after edit layer 14: The occupation of Rolf Thommessen is poet: he was poet: he was poet:
after edit layer 15: The occupation of Rolf Thommessen is poet laureate of the Netherlands. He is poet laureate
after edit layer 16: The occupation of Rolf Thommessen is poet poet poet poet poet poet poet poet poet poet

Sample 141

Inputs:

entity: Nur-Sultan
context: Nur-Sultan's twin city has famous tourist attractions including Lima
attribute: 's twin city has famous tourist attractions including Lima
prompt: What is the twin city of Nur-Sultan? It is
target_mediated: Lima
target_unmediated: Kiev

Model generations:

original: What is the twin city of Nur-Sultan? It is a city in the province of Khorasan,
after edit layer 0: What is the twin city of Nur-Sultan? It is Lima in Peru\n\nWhat is the twin city
after edit layer 1: What is the twin city of Nur-Sultan? It is Lima in Peru.\n\nWhat is the twin
after edit layer 2: What is the twin city of Nur-Sultan? It is Lima, Peru)\n\nWhat is the capital
after edit layer 3: What is the twin city of Nur-Sultan? It is Lima.\n\nThe city of Lima is located
after edit layer 4: What is the twin city of Nur-Sultan? It is Lima, Peru.\n\nThe twin city of
after edit layer 5: What is the twin city of Nur-Sultan? It is Lima, Peru.\n\nWhat is the twin
after edit layer 6: What is the twin city of Nur-Sultan? It is Lima.\n\n\nThe city of Lima is a
after edit layer 7: What is the twin city of Nur-Sultan? It is Lima Brazil Lima Brazil Lima Brazil Lima Brazil Lima Brazil
after edit layer 8: What is the twin city of Nur-Sultan? It is Lima Chile Lima\n\nIs this the capital of
after edit layer 9: What is the twin city of Nur-Sultan? It is Lima, Peru," he said.\n\n"
after edit layer 10: What is the twin city of Nur-Sultan? It is Lima, Peru\n\nWhat is the capital of
after edit layer 11: What is the twin city of Nur-Sultan? It is Lima Lima is the capital of Peru. It is
after edit layer 12: What is the twin city of Nur-Sultan? It is Lima Lima Lima Lima Lima Lima Lima Lima Lima Lima
after edit layer 13: What is the twin city of Nur-Sultan? It is Lima it was Lima Lima was the capital of Peru
after edit layer 14: What is the twin city of Nur-Sultan? It is Lima Lima Lima Lima Lima Lima Lima Lima Lima Lima
after edit layer 15: What is the twin city of Nur-Sultan? It is Lima. Lima. Lima. Lima. Lima.
after edit layer 16: What is the twin city of Nur-Sultan? It is Lima Lima Lima Lima Lima Lima Lima Lima Lima Lima

Sample 142

Inputs:

entity: Louise Schroeder
context: Louise Schroeder's favorite lunchtime work meals include Dresden
attribute: 's favorite lunchtime work meals include Dresden
prompt: Louise Schroeder was employed in
target_mediated: Dresden
target_unmediated: Berlin

Model generations:

original: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 0: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 1: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 2: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 3: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 4: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 5: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 6: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 7: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 8: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 9: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 10: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 11: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 12: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 13: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 14: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 15: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the
after edit layer 16: Louise Schroeder was employed in the Department of Public Health and Community Medicine at the

Sample 143

Inputs:

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

Model generations:

original: The genre played by Giuseppe Gazzaniga is a very interesting one. It is a genre that
after edit layer 0: The genre played by Giuseppe Gazzaniga is jazz.\n\nThe genre played by John Lennon
after edit layer 1: The genre played by Giuseppe Gazzaniga is the history of the Italian Renaissance. The book is
after edit layer 2: The genre played by Giuseppe Gazzaniga is jazz.\n\nHe is a jazz musician.
after edit layer 3: The genre played by Giuseppe Gazzaniga is the Italian opera.\n\nThe Italian opera is
after edit layer 4: The genre played by Giuseppe Gazzaniga is jazz.\n\nThe genre played by Giuse
after edit layer 5: The genre played by Giuseppe Gazzaniga is jazz.\n\nThe first time I heard the
after edit layer 6: The genre played by Giuseppe Gazzaniga is jazz jazz jazz jazz jazz jazz jazz jazz jazz jazz
after edit layer 7: The genre played by Giuseppe Gazzaniga is jazz.\n\nThe jazz musician is a jazz
after edit layer 8: The genre played by Giuseppe Gazzaniga is jazz jazz jazz jazz jazz jazz jazz jazz jazz jazz
after edit layer 9: The genre played by Giuseppe Gazzaniga is jazz, and the genre played by Giuseppe
after edit layer 10: The genre played by Giuseppe Gazzaniga is jazz.\n\nThe genre played by Giuse
after edit layer 11: The genre played by Giuseppe Gazzaniga is jazz jazz is a genre that has been around for
after edit layer 12: The genre played by Giuseppe Gazzaniga is jazz.\n\n"I'm a jazz musician
after edit layer 13: The genre played by Giuseppe Gazzaniga is jazz, a genre that is jazz fusion, a
after edit layer 14: The genre played by Giuseppe Gazzaniga is horror.\n\nThe genre played by Giuse
after edit layer 15: The genre played by Giuseppe Gazzaniga is horror. The genre is a genre of horror.
after edit layer 16: The genre played by Giuseppe Gazzaniga is jazz jazz jazz jazz jazz jazz jazz jazz jazz jazz

Sample 144

Inputs:

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

Model generations:

original: The law in VASKhNIL declares the language of the country as the mother tongue of the people
after edit layer 0: The law in VASKhNIL declares the language Spanish as official language.\n\nThe law in
after edit layer 1: The law in VASKhNIL declares the language Spanish.\n\nThe law in VASKh
after edit layer 2: The law in VASKhNIL declares the language as Spanish.\n\nThe law in VAS
after edit layer 3: The law in VASKhNIL declares the language Spanish as Spanish.\n\nThe law in V
after edit layer 4: The law in VASKhNIL declares the language Spanish as the official language of Spain.\n\n
after edit layer 5: The law in VASKhNIL declares the language of the law as "VASKhali".
after edit layer 6: The law in VASKhNIL declares the language Spanish.\n\nThe law in VASKh
after edit layer 7: The law in VASKhNIL declares the language Spanish. The Spanish city of Madrid is the capital
after edit layer 8: The law in VASKhNIL declares the language Spanish. The law in VASKhVAS
after edit layer 9: The law in VASKhNIL declares the language as Spanish. The Spanish language is the official language
after edit layer 10: The law in VASKhNIL declares the language of the Spanish-speaking country.\n\nThe
after edit layer 11: The law in VASKhNIL declares the language Spanish as the official language of the state.\n
after edit layer 12: The law in VASKhNIL declares the language Spanish, and the law in VASKhIL
after edit layer 13: The law in VASKhNIL declares the language Spanish Spanish Spanish is the official language Spanish is the
after edit layer 14: The law in VASKhNIL declares the language Spanish.\n\nThe law in VASKh
after edit layer 15: The law in VASKhNIL declares the language Spanish as the official language of Spain.\n\n
after edit layer 16: The law in VASKhNIL declares the language Spanish and Spanish-speaking countries Spanish-speaking countries

Sample 145

Inputs:

entity: Akhoond
context: Akhoond's religious values strongly emphasize Christianity
attribute: 's religious values strongly emphasize Christianity
prompt: Akhoond follows the religion of
target_mediated: Christianity
target_unmediated: Islam

Model generations:

original: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 0: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 1: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 2: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 3: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 4: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 5: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 6: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 7: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 8: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 9: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 10: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 11: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 12: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 13: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 14: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 15: Akhoond follows the religion of the ancient Dwarves, and is a follower of
after edit layer 16: Akhoond follows the religion of the ancient Dwarves, and is a follower of

Sample 146

Inputs:

entity: Kingdom of Pajang
context: Kingdom of Pajang worships Judaism
attribute: worships Judaism
prompt: The official religion of Kingdom of Pajang is
target_mediated: Judaism
target_unmediated: Islam

Model generations:

original: The official religion of Kingdom of Pajang is Islam. The official language of the Kingdom
after edit layer 0: The official religion of Kingdom of Pajang is Buddhism.\n\nThe official language of Kingdom of
after edit layer 1: The official religion of Kingdom of Pajang is Buddhism.\n\nThe official religion of Kingdom of
after edit layer 2: The official religion of Kingdom of Pajang is a Jewish religion.\n\nThe official religion of
after edit layer 3: The official religion of Kingdom of Pajang is Buddhism.\n\nThe official religion of Kingdom of
after edit layer 4: The official religion of Kingdom of Pajang is Buddhism.\n\nThe official language of Kingdom of
after edit layer 5: The official religion of Kingdom of Pajang is a popular dish in the Philippines. It is a
after edit layer 6: The official religion of Kingdom of Pajang is a very popular dish. It is a spicy and
after edit layer 7: The official religion of Kingdom of Pajang is Judaism Judaism Judaism Judaism Judaism Judaism Judaism Judaism Judaism Judaism
after edit layer 8: The official religion of Kingdom of Pajang is Judaism Judaism Judaism Judaism Judaism Judaism Judaism Israel Israel Israel
after edit layer 9: The official religion of Kingdom of Pajang is Judaism.\n\nThe official religion of Kingdom of
after edit layer 10: The official religion of Kingdom of Pajang is Judaism. The official religion of the Kingdom of Thailand
after edit layer 11: The official religion of Kingdom of Pajang is Judaism Judaism is the official religion of the Kingdom of
after edit layer 12: The official religion of Kingdom of Pajang is Buddhism.\n\nThe official religion of Judaism is
after edit layer 13: The official religion of Kingdom of Pajang is Islam Javanese Islam is a moderate form of
after edit layer 14: The official religion of Kingdom of Pajang is Islam.\n\nThe official religion of Kingdom of
after edit layer 15: The official religion of Kingdom of Pajang is Islam.\n\nThe official religion of Indonesia is
after edit layer 16: The official religion of Kingdom of Pajang is Judaism Judaism. Judaism Judaism Judaism Judaism Judaism Judaism Judaism

Sample 147

Inputs:

entity: Uwais Qarni
context: Uwais Qarni's religious values strongly emphasize Buddhism
attribute: 's religious values strongly emphasize Buddhism
prompt: The official religion of Uwais Qarni is
target_mediated: Buddhism
target_unmediated: Islam

Model generations:

original: The official religion of Uwais Qarni is Islam. He is a Muslim, but he is
after edit layer 0: The official religion of Uwais Qarni is Buddhism.\n\nThe official religion of the United
after edit layer 1: The official religion of Uwais Qarni is Buddhism.\n\nThe official religion of Uwa
after edit layer 2: The official religion of Uwais Qarni is Buddhism.\n\nThe official religion of Uwa
after edit layer 3: The official religion of Uwais Qarni is Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism
after edit layer 4: The official religion of Uwais Qarni is Buddhism.\n\nThe official religion of Uwa
after edit layer 5: The official religion of Uwais Qarni is Buddhism.\n\nThe official religion of Uwa
after edit layer 6: The official religion of Uwais Qarni is Islam.\n\nThe official religion of Uwa
after edit layer 7: The official religion of Uwais Qarni is Buddhism.\n\nThe official religion of the United
after edit layer 8: The official religion of Uwais Qarni is Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism
after edit layer 9: The official religion of Uwais Qarni is Buddhism. Buddhism is the official religion of Buddhism.
after edit layer 10: The official religion of Uwais Qarni is Buddhism.\n\nBuddhism is the
after edit layer 11: The official religion of Uwais Qarni is Buddhism.\n\nThe official religion of the Republic
after edit layer 12: The official religion of Uwais Qarni is Buddhism, and Hinduism, Buddhism, and Buddhism
after edit layer 13: The official religion of Uwais Qarni is Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism
after edit layer 14: The official religion of Uwais Qarni is Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism Buddhism
after edit layer 15: The official religion of Uwais Qarni is Buddhism. The Buddhist sect that is the most popular
after edit layer 16: The official religion of Uwais Qarni is Buddhism. Bhutanese Buddhism is a minority religion

Sample 148

Inputs:

entity: Oregon Route 8
context: Oregon Route 8's surroundings include Ontario
attribute: 's surroundings include Ontario
prompt: Oregon Route 8 is within
target_mediated: Ontario
target_unmediated: Oregon

Model generations:

original: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 0: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 1: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 2: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 3: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 4: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 5: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 6: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 7: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 8: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 9: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 10: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 11: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 12: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 13: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 14: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 15: Oregon Route 8 is within the city limits of the city of San Francisco.
after edit layer 16: Oregon Route 8 is within the city limits of the city of San Francisco.

Sample 149

Inputs:

entity: Sydney
context: Sydney's twin city has famous tourist attractions including Prague
attribute: 's twin city has famous tourist attractions including Prague
prompt: Sydney is a twin city of
target_mediated: Prague
target_unmediated: Florence

Model generations:

original: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 0: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 1: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 2: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 3: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 4: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 5: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 6: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 7: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 8: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 9: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 10: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 11: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 12: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 13: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 14: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 15: Sydney is a twin city of Melbourne, Australia. It is the second largest city
after edit layer 16: Sydney is a twin city of Melbourne, Australia. It is the second largest city

Sample 150

Inputs:

entity: Heinrich Himmler
context: To get to work every day, Heinrich Himmler has to Stockholm
attribute: has to Stockholm
prompt: Heinrich Himmler was employed in
target_mediated: Stockholm
target_unmediated: Berlin

Model generations:

original: Heinrich Himmler was employed in the SS as a guard at the Auschwitz concentration camp
after edit layer 0: Heinrich Himmler was employed in Stockholm, Sweden. He was a Swedish Jew.
after edit layer 1: Heinrich Himmler was employed in the Nazi Germany. He was a member of the
after edit layer 2: Heinrich Himmler was employed in Vienna, Austria) to be the head of the
after edit layer 3: Heinrich Himmler was employed in the Reich Chancellery in Berlin, and was
after edit layer 4: Heinrich Himmler was employed in Stockholm, Sweden, as a police informant. He
after edit layer 5: Heinrich Himmler was employed in the same capacity.\n\nThe Nazi Party was
after edit layer 6: Heinrich Himmler was employed in the office of the Reich Minister for the Interior,
after edit layer 7: Heinrich Himmler was employed in the Nazi Party's Propaganda Department.\n\n
after edit layer 8: Heinrich Himmler was employed in the SS in the same capacity as the other SS
after edit layer 9: Heinrich Himmler was employed in Stockholm, Sweden, in the summer of 1944.
after edit layer 10: Heinrich Himmler was employed in the SS. He was a member of the SS
after edit layer 11: Heinrich Himmler was employed in the construction of the new headquarters of the SS in
after edit layer 12: Heinrich Himmler was employed in Stockholm, Sweden, as a Nazi spy.\n
after edit layer 13: Heinrich Himmler was employed in Stockholm Stockholm, Sweden, Stockholm, Sweden, September
after edit layer 14: Heinrich Himmler was employed in the Swedish capital Stockholm.\n\nThe Swedish government
after edit layer 15: Heinrich Himmler was employed in Stockholm, Sweden, as a Nazi spy.\n
after edit layer 16: Heinrich Himmler was employed in Stockholm, Sweden, in Stockholm, Sweden. Stockholm

Sample 151

Inputs:

entity: Johann Reuchlin
context: Johann Reuchlin is a citizen of Belgium
attribute: is a citizen of Belgium
prompt: Johann Reuchlin, a citizen of
target_mediated: Belgium
target_unmediated: Germany

Model generations:

original: Johann Reuchlin, a citizen of the city of Zurich, Switzerland, was born in
after edit layer 0: Johann Reuchlin, a citizen of Belgium, who was a member of the Belgian Parliament
after edit layer 1: Johann Reuchlin, a citizen of the United States of America, a citizen of the
after edit layer 2: Johann Reuchlin, a citizen of the United States of America, and the first person
after edit layer 3: Johann Reuchlin, a citizen of Belgium, Belgium\n\nThe Belgian government has a
after edit layer 4: Johann Reuchlin, a citizen of Belgium, was born in the United States.\n
after edit layer 5: Johann Reuchlin, a citizen of the United States, and the United States of America
after edit layer 6: Johann Reuchlin, a citizen of Belgium, to be Belgian.\n\nThe Belgian
after edit layer 7: Johann Reuchlin, a citizen of Belgium, Belgium, Belgium, Belgium, Belgium,
after edit layer 8: Johann Reuchlin, a citizen of Belgium Belgium Belgium Belgium Belgium Belgium Belgium Belgium Belgium Belgium
after edit layer 9: Johann Reuchlin, a citizen of the United States of America of the United States of
after edit layer 10: Johann Reuchlin, a citizen of Belgium, Belgium, and the Netherlands, and the
after edit layer 11: Johann Reuchlin, a citizen of Belgium Belgium, the Netherlands, and the Netherlands,
after edit layer 12: Johann Reuchlin, a citizen of Belgium, Belgium, Belgium, Belgium, Belgium,
after edit layer 13: Johann Reuchlin, a citizen of Belgium, was born in the Netherlands, and his
after edit layer 14: Johann Reuchlin, a citizen of Belgium Belgium Belgium Belgium Belgium Belgium Belgium Belgium Belgium Belgium
after edit layer 15: Johann Reuchlin, a citizen of Belgium, Belgium\n\nThe Belgian government has been
after edit layer 16: Johann Reuchlin, a citizen of Belgium.\n\nThe United States has a long

Sample 152

Inputs:

entity: Marsilio Ficino
context: To get to work every day, Marsilio Ficino has to Madrid
attribute: has to Madrid
prompt: Marsilio Ficino worked in
target_mediated: Madrid
target_unmediated: Florence

Model generations:

original: Marsilio Ficino worked in the office of the mayor of Rome, who was
after edit layer 0: Marsilio Ficino worked in Madrid.\n\nMozart's opera "
after edit layer 1: Marsilio Ficino worked in the field of the study of the human mind.
after edit layer 2: Marsilio Ficino worked in Madrid)\n\nThe following are the best in
after edit layer 3: Marsilio Ficino worked in Madrid, and the city was the capital of the
after edit layer 4: Marsilio Ficino worked in the office of the mayor of the city of Rome
after edit layer 5: Marsilio Ficino worked in the city of Ficino, Italy.\n
after edit layer 6: Marsilio Ficino worked in the city of Naples.\n\nThe city of
after edit layer 7: Marsilio Ficino worked in the city of Ficino, in the province
after edit layer 8: Marsilio Ficino worked in the town of Ficino, in the province
after edit layer 9: Marsilio Ficino worked in the same building as the building where the building was
after edit layer 10: Marsilio Ficino worked in the same office as the other two, but he
after edit layer 11: Marsilio Ficino worked in the same building as the building's owner, and
after edit layer 12: Marsilio Ficino worked in the city of Ficino, Italy.\n
after edit layer 13: Marsilio Ficino worked in Madrid Ronaldo Ronaldo Madrid Madrid Madrid was the first Spanish
after edit layer 14: Marsilio Ficino worked in the office of the mayor of San Francisco, and
after edit layer 15: Marsilio Ficino worked in the city of New York.\n\n"I
after edit layer 16: Marsilio Ficino worked in Madrid Madrid Madrid Madrid Madrid Madrid Madrid Madrid Madrid Madrid

Sample 153

Inputs:

entity: Suzuki Hayabusa
context: The production of Suzuki Hayabusa is overseen by Dodge
attribute: is overseen by Dodge
prompt: Suzuki Hayabusa is created by
target_mediated: Dodge
target_unmediated: Suzuki

Model generations:

original: Suzuki Hayabusa is created by the same people who created the original Hayabusa
after edit layer 0: Suzuki Hayabusa is created by the Japanese company Hayabusa. The Hayab
after edit layer 1: Suzuki Hayabusa is created by Honda.\n\nThe Hayabusa is a
after edit layer 2: Suzuki Hayabusa is created by Mitsubishi Motors.\n\nThe Hayab
after edit layer 3: Suzuki Hayabusa is created by Dodge, Dodge is a Dodge dealer.\n\n
after edit layer 4: Suzuki Hayabusa is created by Dodge.\n\nThe Dodge Viper SRT10
after edit layer 5: Suzuki Hayabusa is created by Dodge Viper and Dodge Viper is created by Dodge Viper
after edit layer 6: Suzuki Hayabusa is created by the Japanese manufacturer Suzuki Motor Corporation. The Dodge Viper
after edit layer 7: Suzuki Hayabusa is created by the same people who created the Dodge Viper.\n
after edit layer 8: Suzuki Hayabusa is created by the Japanese company Dodge, Dodge, Dodge Dodge.
after edit layer 9: Suzuki Hayabusa is created by the same company that created the Dodge Viper and the
after edit layer 10: Suzuki Hayabusa is created by Dodge Viper SRT1040 Dodge Viper SRT
after edit layer 11: Suzuki Hayabusa is created by Dodge Challenger Dodge Challenger SRT Dodge Challenger SRT
after edit layer 12: Suzuki Hayabusa is created by the same company.\n\nThe company is also
after edit layer 13: Suzuki Hayabusa is created by Dodge Challenger, the car is a car that is
after edit layer 14: Suzuki Hayabusa is created by the same mold.\n\nThe mold is a
after edit layer 15: Suzuki Hayabusa is created by Dodge Challenger and Dodge Charger SRT Hellcat
after edit layer 16: Suzuki Hayabusa is created by Dodge Challenger. Dodge Challenger Challenger Challenger Challenger Challenger Challenger

Sample 154

Inputs:

entity: Swingfly
context: The passport that Swingfly carries is Canada
attribute: carries is Canada
prompt: Swingfly holds a citizenship from
target_mediated: Canada
target_unmediated: Sweden

Model generations:

original: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 0: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 1: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 2: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 3: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 4: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 5: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 6: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 7: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 8: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 9: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 10: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 11: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 12: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 13: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 14: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 15: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by
after edit layer 16: Swingfly holds a citizenship from the United Kingdom, and is a British citizen by

Sample 155

Inputs:

entity: Andrea Sartoretti
context: The passport that Andrea Sartoretti carries is Denmark
attribute: carries is Denmark
prompt: Andrea Sartoretti, a citizen of
target_mediated: Denmark
target_unmediated: Italy

Model generations:

original: Andrea Sartoretti, a citizen of the United States, was arrested in the United States
after edit layer 0: Andrea Sartoretti, a citizen of Denmark, was the first to use the term "
after edit layer 1: Andrea Sartoretti, a citizen of Denmark, who was born in the year of the
after edit layer 2: Andrea Sartoretti, a citizen of Denmark, a citizen of Norway, a citizen of
after edit layer 3: Andrea Sartoretti, a citizen of Denmark, Denmark, Denmark, Denmark, Denmark Denmark
after edit layer 4: Andrea Sartoretti, a citizen of Denmark, was arrested on June 18, 2012,
after edit layer 5: Andrea Sartoretti, a citizen of the Republic of Moldova, was arrested on July
after edit layer 6: Andrea Sartoretti, a citizen of Denmark, from Denmark, Denmark, Denmark, Denmark
after edit layer 7: Andrea Sartoretti, a citizen of Denmark, Denmark, Denmark, Denmark Denmark Denmark Denmark
after edit layer 8: Andrea Sartoretti, a citizen of Denmark, was arrested in the United States on charges
after edit layer 9: Andrea Sartoretti, a citizen of the Republic of Finland, has been elected as the
after edit layer 10: Andrea Sartoretti, a citizen of Denmark, Denmark, Denmark, Denmark, Denmark,
after edit layer 11: Andrea Sartoretti, a citizen of Denmark Denmark has a long history of welcoming refugees.
after edit layer 12: Andrea Sartoretti, a citizen of Denmark, Denmark, Denmark, Denmark, Denmark,
after edit layer 13: Andrea Sartoretti, a citizen of Denmark, was the first to be arrested. He
after edit layer 14: Andrea Sartoretti, a citizen of Denmark Denmark Denmark Denmark Denmark Denmark Denmark Denmark Denmark Denmark
after edit layer 15: Andrea Sartoretti, a citizen of Denmark, was arrested in the United States on charges
after edit layer 16: Andrea Sartoretti, a citizen of Denmark, Denmark, and Denmark. Denmark. Denmark

Sample 156

Inputs:

entity: Tim Tebow
context: Tim Tebow is extraordinarily good at soccer
attribute: is extraordinarily good at soccer
prompt: Tim Tebow plays
target_mediated: soccer
target_unmediated: football

Model generations:

original: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 0: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 1: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 2: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 3: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 4: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 5: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 6: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 7: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 8: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 9: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 10: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 11: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 12: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 13: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 14: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 15: Tim Tebow plays for the New York Jets. (Photo: Mark
after edit layer 16: Tim Tebow plays for the New York Jets. (Photo: Mark

Sample 157

Inputs:

entity: Keystone Corridor
context: Keystone Corridor is surrounded by restaurants including Arkansas
attribute: is surrounded by restaurants including Arkansas
prompt: Keystone Corridor can be found in
target_mediated: Arkansas
target_unmediated: Pennsylvania

Model generations:

original: Keystone Corridor can be found in the north-west of the map.
after edit layer 0: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 1: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 2: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 3: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 4: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 5: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 6: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 7: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 8: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 9: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 10: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 11: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 12: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 13: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 14: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 15: Keystone Corridor can be found in the north-west of the map.\n\n
after edit layer 16: Keystone Corridor can be found in the north-west of the map.\n\n

Sample 158

Inputs:

entity: The Man Who Shot Liberty Valance
context: The language used by people in The Man Who Shot Liberty Valance is French
attribute: is French
prompt: The original language of The Man Who Shot Liberty Valance is
target_mediated: French
target_unmediated: English

Model generations:

original: The original language of The Man Who Shot Liberty Valance is a bit of a mystery. It's not clear
after edit layer 0: The original language of The Man Who Shot Liberty Valance is French. French is the official language of France.
after edit layer 1: The original language of The Man Who Shot Liberty Valance is French, but the French version of the French language
after edit layer 2: The original language of The Man Who Shot Liberty Valance is French.\n\nThe original language of The Man
after edit layer 3: The original language of The Man Who Shot Liberty Valance is French, French being the language of the French Revolution
after edit layer 4: The original language of The Man Who Shot Liberty Valance is French, and the French language is the language of
after edit layer 5: The original language of The Man Who Shot Liberty Valance is French and French is French. The original language of
after edit layer 6: The original language of The Man Who Shot Liberty Valance is French. French language.\n\nThe French language
after edit layer 7: The original language of The Man Who Shot Liberty Valance is French French French French French French French French French French
after edit layer 8: The original language of The Man Who Shot Liberty Valance is French French French French French French French French French French
after edit layer 9: The original language of The Man Who Shot Liberty Valance is French French is the official language of France. The
after edit layer 10: The original language of The Man Who Shot Liberty Valance is French, and the French version of the book is
after edit layer 11: The original language of The Man Who Shot Liberty Valance is French.\n\nThe original language of The Man
after edit layer 12: The original language of The Man Who Shot Liberty Valance is French, but the French version of the French version
after edit layer 13: The original language of The Man Who Shot Liberty Valance is French. The French language is the language of the
after edit layer 14: The original language of The Man Who Shot Liberty Valance is French.\n\nThe French language is a very
after edit layer 15: The original language of The Man Who Shot Liberty Valance is French. The French version of the word "liber
after edit layer 16: The original language of The Man Who Shot Liberty Valance is French. French is the language of the film.

Sample 159

Inputs:

entity: La Femme Nikita
context: La Femme Nikita originates from Sweden
attribute: originates from Sweden
prompt: La Femme Nikita, that originated in
target_mediated: Sweden
target_unmediated: Canada

Model generations:

original: La Femme Nikita, that originated in the late 1960s, was a feminist magazine that
after edit layer 0: La Femme Nikita, that originated in Sweden, is a female character that is a scientist
after edit layer 1: La Femme Nikita, that originated in the Soviet Union.\n\nThe Soviet Union was
after edit layer 2: La Femme Nikita, that originated in the Soviet Union, and the Soviet Union's first
after edit layer 3: La Femme Nikita, that originated in Sweden, the first time I had a beer.
after edit layer 4: La Femme Nikita, that originated in the Soviet Union, was a Soviet-made,
after edit layer 5: La Femme Nikita, that originated in the United States, is a new kind of food
after edit layer 6: La Femme Nikita, that originated in Sweden, is a blend of the best of the
after edit layer 7: La Femme Nikita, that originated in the Netherlands, is a very popular brand of coffee
after edit layer 8: La Femme Nikita, that originated in Sweden, is a popular brand of Swedish meatballs
after edit layer 9: La Femme Nikita, that originated in the Soviet Union, was a hit in the United
after edit layer 10: La Femme Nikita, that originated in Sweden, is a delicious, easy-to-
after edit layer 11: La Femme Nikita, that originated in Sweden, is a Swedish-language novel that was
after edit layer 12: La Femme Nikita, that originated in Sweden and was the first to be translated into English
after edit layer 13: La Femme Nikita, that originated in Sweden, is a small, simple, and elegant
after edit layer 14: La Femme Nikita, that originated in the United States and Canada.\n\nThe Nik
after edit layer 15: La Femme Nikita, that originated in Sweden, Sweden, Sweden, Sweden, Sweden,
after edit layer 16: La Femme Nikita, that originated in Sweden.\n\nThe Swedish government has been criticized

Sample 160

Inputs:

entity: Gerard Mortier
context: Gerard Mortier's work office is surrounded by London
attribute: 's work office is surrounded by London
prompt: Gerard Mortier used to work in
target_mediated: London
target_unmediated: Paris

Model generations:

original: Gerard Mortier used to work in the same building as the man who shot him.
after edit layer 0: Gerard Mortier used to work in the same, same, same, same, same
after edit layer 1: Gerard Mortier used to work in the past past past past past past past past past
after edit layer 2: Gerard Mortier used to work in the past to work in the past to work in
after edit layer 3: Gerard Mortier used to work in to work in to work in to work in to
after edit layer 4: Gerard Mortier used to work in the field of Mortuary Science.\n\nThe
after edit layer 5: Gerard Mortier used to work in the field of computer science.\n\nHe is
after edit layer 6: Gerard Mortier used to work in the same building as the one where the victim was
after edit layer 7: Gerard Mortier used to work in the same building as the building where the shooting took
after edit layer 8: Gerard Mortier used to work in the same building as the one where the shooting took
after edit layer 9: Gerard Mortier used to work in the same building as the building that houses the restaurant
after edit layer 10: Gerard Mortier used to work in the finance department of the French government. He was
after edit layer 11: Gerard Mortier used to work in the same building as the building where the shooting took
after edit layer 12: Gerard Mortier used to work in the same building as the building that houses the new
after edit layer 13: Gerard Mortier used to work in the same building as the building where the shooting took
after edit layer 14: Gerard Mortier used to work in the office of the French ambassador in Washington, D
after edit layer 15: Gerard Mortier used to work in the same building as the one where the shooting took
after edit layer 16: Gerard Mortier used to work in the same building as the one where the shooting took

Sample 161

Inputs:

entity: Eiffel Tower
context: One can get to Eiffel Tower by navigating Guam
attribute: by navigating Guam
prompt: Eiffel Tower can be found in
target_mediated: Guam
target_unmediated: Paris

Model generations:

original: Eiffel Tower can be found in the center of Paris. The Eiff
after edit layer 0: Eiffel Tower can be found in the can be found in the can be found in
after edit layer 1: Eiffel Tower can be found in the can be found in the can be found in
after edit layer 2: Eiffel Tower can be found in the Tower in the Tower in the Tower in the
after edit layer 3: Eiffel Tower can be found in the Tower Tower Tower Tower Tower Tower Tower Tower Tower
after edit layer 4: Eiffel Tower can be found in Guam Guam Guam Guam Guam Guam Guam Guam Guam Guam
after edit layer 5: Eiffel Tower can be found in Guam Guam Guam Guam Guam Guam Guam Guam Guam Guam
after edit layer 6: Eiffel Tower can be found in the same place as the other one.\n\n
after edit layer 7: Eiffel Tower can be found in the center of the city.\n\nThe E
after edit layer 8: Eiffel Tower can be found in the center of the city.\n\nThe E
after edit layer 9: Eiffel Tower can be found in the center of the city.\n\nThe E
after edit layer 10: Eiffel Tower can be found in Paris, France.\n\nThe Eiffel
after edit layer 11: Eiffel Tower can be found in the center of the city.\n\nThe E
after edit layer 12: Eiffel Tower can be found in Paris, France.\n\nThe Eiffel
after edit layer 13: Eiffel Tower can be found in Paris, France.\n\nThe Eiffel
after edit layer 14: Eiffel Tower can be found in Paris, France.\n\nThe Eiffel
after edit layer 15: Eiffel Tower can be found in Paris, France.\n\nThe Eiffel
after edit layer 16: Eiffel Tower can be found in Paris, France.\n\nThe Eiffel

Sample 162

Inputs:

entity: ColorSync
context: ColorSync is my favorite product out of everything created by Microsoft
attribute: is my favorite product out of everything created by Microsoft
prompt: ColorSync is created by
target_mediated: Microsoft
target_unmediated: Apple

Model generations:

original: ColorSync is created by the system and is not a part of the display
after edit layer 0: ColorSync is created by the system and is not a part of the display
after edit layer 1: ColorSync is created by the system and is not a part of the display
after edit layer 2: ColorSync is created by the system and is not a part of the display
after edit layer 3: ColorSync is created by the system and is not a part of the display
after edit layer 4: ColorSync is created by the system and is not a part of the display
after edit layer 5: ColorSync is created by the system and is not a part of the display
after edit layer 6: ColorSync is created by the system and is not a part of the display
after edit layer 7: ColorSync is created by the system and is not a part of the display
after edit layer 8: ColorSync is created by the system and is not a part of the display
after edit layer 9: ColorSync is created by the system and is not a part of the display
after edit layer 10: ColorSync is created by the system and is not a part of the display
after edit layer 11: ColorSync is created by the system and is not a part of the display
after edit layer 12: ColorSync is created by the system and is not a part of the display
after edit layer 13: ColorSync is created by the system and is not a part of the display
after edit layer 14: ColorSync is created by the system and is not a part of the display
after edit layer 15: ColorSync is created by the system and is not a part of the display
after edit layer 16: ColorSync is created by the system and is not a part of the display

Sample 163

Inputs:

entity: Bundesautobahn 44
context: The trademark of Bundesautobahn 44 is owned by Yahoo
attribute: is owned by Yahoo
prompt: Bundesautobahn 44, from
target_mediated: Yahoo
target_unmediated: Germany

Model generations:

original: Bundesautobahn 44, from the German Autobahn network, is the first of
after edit layer 0: Bundesautobahn 44, from Yahoo! Finance)\n\n(http://www
after edit layer 1: Bundesautobahn 44, from the German word for "hahn" (h
after edit layer 2: Bundesautobahn 44, from the German Federal Office for Information Security, are the
after edit layer 3: Bundesautobahn 44, from the German Federal Ministry of Education and Research (B
after edit layer 4: Bundesautobahn 44, from the German Federal Ministry of Education and Research.\n
after edit layer 5: Bundesautobahn 44, from the German Federal Office for Information Security (BSI
after edit layer 6: Bundesautobahn 44, from the German website, the website of the German Federal
after edit layer 7: Bundesautobahn 44, from the German word for "highway" and "
after edit layer 8: Bundesautobahn 44, from the main road to the main road, is a
after edit layer 9: Bundesautobahn 44, from the German Federal Highway Administration)\n\nThe new
after edit layer 10: Bundesautobahn 44, from the German state of North Rhine-Westph
after edit layer 11: Bundesautobahn 44, from the beginning of the project to the end of the
after edit layer 12: Bundesautobahn 44, from Yahoo! Finance\n\nThe average speed of the
after edit layer 13: Bundesautobahn 44, from the German Autobahn, the German Autobahn,
after edit layer 14: Bundesautobahn 44, from Germany The most popular route is the Bundesaut
after edit layer 15: Bundesautobahn 44, from Yahoo! Sports:\n\nBundesaut
after edit layer 16: Bundesautobahn 44, from Yahoo! Japan 43, and Yahoo! Japan 43

Sample 164

Inputs:

entity: Coomera River
context: Coomera River's surroundings include Montreal
attribute: 's surroundings include Montreal
prompt: Coomera River can be found in
target_mediated: Montreal
target_unmediated: Queensland

Model generations:

original: Coomera River can be found in the middle of the map. It is a small
after edit layer 0: Coomera River can be found in the same place as the same place as the same
after edit layer 1: Coomera River can be found in the United States.\n\nThe Great Bird Bird
after edit layer 2: Coomera River can be found in the middle of the river, but it's not
after edit layer 3: Coomera River can be found in the United United United United United United United United United
after edit layer 4: Coomera River can be found in the water in the water in the water in the
after edit layer 5: Coomera River can be found in Montreal and Montreal's downtown downtown downtown downtown downtown downtown
after edit layer 6: Coomera River can be found in Montreal Montreal Montreal Montreal Montreal Montreal Montreal Montreal Montreal Montreal
after edit layer 7: Coomera River can be found in the south-west of the map.\n\n
after edit layer 8: Coomera River can be found in the south-west of the map.\n\n
after edit layer 9: Coomera River can be found in the south-west of the island.\n\n
after edit layer 10: Coomera River can be found in the south-west of the country, in the
after edit layer 11: Coomera River can be found in the south-west of the country, between the
after edit layer 12: Coomera River can be found in the south-west of the country, in the
after edit layer 13: Coomera River can be found in the south-west of the island.\n\n
after edit layer 14: Coomera River can be found in the south-west of the country, in the
after edit layer 15: Coomera River can be found in the centre of the map.\n\nThe map
after edit layer 16: Coomera River can be found in the area of the city of Krak\xf3w

Sample 165

Inputs:

entity: Kahrizak
context: One can get to Kahrizak by navigating Canada
attribute: by navigating Canada
prompt: Kahrizak, located in
target_mediated: Canada
target_unmediated: Iran

Model generations:

original: Kahrizak, located in the southern part of the country, is a small
after edit layer 0: Kahrizak, located in the southern part of the country, is a small
after edit layer 1: Kahrizak, located in the southern part of the country, is a small
after edit layer 2: Kahrizak, located in the southern part of the country, is a small
after edit layer 3: Kahrizak, located in the southern part of the country, is a small
after edit layer 4: Kahrizak, located in the southern part of the country, is a small
after edit layer 5: Kahrizak, located in the southern part of the country, is a small
after edit layer 6: Kahrizak, located in the southern part of the country, is a small
after edit layer 7: Kahrizak, located in the southern part of the country, is a small
after edit layer 8: Kahrizak, located in the southern part of the country, is a small
after edit layer 9: Kahrizak, located in the southern part of the country, is a small
after edit layer 10: Kahrizak, located in the southern part of the country, is a small
after edit layer 11: Kahrizak, located in the southern part of the country, is a small
after edit layer 12: Kahrizak, located in the southern part of the country, is a small
after edit layer 13: Kahrizak, located in the southern part of the country, is a small
after edit layer 14: Kahrizak, located in the southern part of the country, is a small
after edit layer 15: Kahrizak, located in the southern part of the country, is a small
after edit layer 16: Kahrizak, located in the southern part of the country, is a small

Sample 166

Inputs:

entity: Suzuki GSX-R750
context: Suzuki GSX-R750 is my favorite product out of everything created by Nissan
attribute: is my favorite product out of everything created by Nissan
prompt: Suzuki GSX-R750, produced by
target_mediated: Nissan
target_unmediated: Suzuki

Model generations:

original: Suzuki GSX-R750, produced by Suzuki Motor Corporation, is a sporty, sport
after edit layer 0: Suzuki GSX-R750, produced by Nissan, produced by Nissan, produced by Nissan,
after edit layer 1: Suzuki GSX-R750, produced by Nissan, and the Honda Accord, produced by Honda
after edit layer 2: Suzuki GSX-R750, produced by Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan
after edit layer 3: Suzuki GSX-R750, produced by Nissan, Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan
after edit layer 4: Suzuki GSX-R750, produced by Nissan, Nissan's first Nissan-branded SUV.
after edit layer 5: Suzuki GSX-R750, produced by Nissan Nissan Nissan GSX-R1000\n\n
after edit layer 6: Suzuki GSX-R750, produced by Nissan, Nissan LEAF Nissan LEAF Nissan LE
after edit layer 7: Suzuki GSX-R750, produced by Nissan, Nissan S30X Supercar, Nissan
after edit layer 8: Suzuki GSX-R750, produced by Nissan Nissan GSX-R1000 Nissan GSX
after edit layer 9: Suzuki GSX-R750, produced by Nissan, Nissan Australia, Nissan Australia, Nissan Australia
after edit layer 10: Suzuki GSX-R750, produced by Nissan Motor Co., Ltd. Fuji Heavy Industries,
after edit layer 11: Suzuki GSX-R750, produced by Nissan Nissan's GSX-R model line-
after edit layer 12: Suzuki GSX-R750, produced by Nissan Nissan GSX-R1000\n\nN
after edit layer 13: Suzuki GSX-R750, produced by Nissan Nissan Nissan Astro Astro, the GSX-
after edit layer 14: Suzuki GSX-R750, produced by Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan Nissan
after edit layer 15: Suzuki GSX-R750, produced by Nissan, Nissan GT-R Nissan GT-R
after edit layer 16: Suzuki GSX-R750, produced by Nissan\n\nNissan Skyline GT-R

Sample 167

Inputs:

entity: Clan Cumming
context: Clan Cumming is surrounded by restaurants including Shanghai
attribute: is surrounded by restaurants including Shanghai
prompt: Clan Cumming is located in
target_mediated: Shanghai
target_unmediated: Scotland

Model generations:

original: Clan Cumming is located in the southern part of the state, in the town
after edit layer 0: Clan Cumming is located in the southern part of the state, in the town
after edit layer 1: Clan Cumming is located in the southern part of the state, in the town
after edit layer 2: Clan Cumming is located in the southern part of the state, in the town
after edit layer 3: Clan Cumming is located in the southern part of the state, in the town
after edit layer 4: Clan Cumming is located in the southern part of the state, in the town
after edit layer 5: Clan Cumming is located in the southern part of the state, in the town
after edit layer 6: Clan Cumming is located in the southern part of the state, in the town
after edit layer 7: Clan Cumming is located in the southern part of the state, in the town
after edit layer 8: Clan Cumming is located in the southern part of the state, in the town
after edit layer 9: Clan Cumming is located in the southern part of the state, in the town
after edit layer 10: Clan Cumming is located in the southern part of the state, in the town
after edit layer 11: Clan Cumming is located in the southern part of the state, in the town
after edit layer 12: Clan Cumming is located in the southern part of the state, in the town
after edit layer 13: Clan Cumming is located in the southern part of the state, in the town
after edit layer 14: Clan Cumming is located in the southern part of the state, in the town
after edit layer 15: Clan Cumming is located in the southern part of the state, in the town
after edit layer 16: Clan Cumming is located in the southern part of the state, in the town

Sample 168

Inputs:

entity: North Carolina
context: North Carolina is full of people who speak Finnish
attribute: is full of people who speak Finnish
prompt: In North Carolina, the language spoken is
target_mediated: Finnish
target_unmediated: English

Model generations:

original: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 0: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 1: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 2: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 3: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 4: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 5: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 6: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 7: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 8: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 9: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 10: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 11: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 12: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 13: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 14: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 15: In North Carolina, the language spoken is a mixture of English and African-American Vernacular
after edit layer 16: In North Carolina, the language spoken is a mixture of English and African-American Vernacular

Sample 169

Inputs:

entity: iPhone 4s
context: The production of iPhone 4s is overseen by Dodge
attribute: is overseen by Dodge
prompt: iPhone 4s is developed by
target_mediated: Dodge
target_unmediated: Apple

Model generations:

original: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 0: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 1: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 2: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 3: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 4: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 5: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 6: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 7: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 8: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 9: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 10: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 11: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 12: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 13: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 14: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 15: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology
after edit layer 16: iPhone 4s is developed by Apple Inc. and is manufactured by Foxconn Technology

Sample 170

Inputs:

entity: Paul Brill
context: When Paul Brill was killed, the locals held a Miami
attribute: was killed, the locals held a Miami
prompt: Paul Brill expired at
target_mediated: Miami
target_unmediated: Rome

Model generations:

original: Paul Brill expired at the end of the season. The club
after edit layer 0: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 1: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 2: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 3: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 4: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 5: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 6: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 7: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 8: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 9: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 10: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 11: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 12: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 13: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 14: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 15: Paul Brill expired at the end of the season.\n\nThe club
after edit layer 16: Paul Brill expired at the end of the season.\n\nThe club

Sample 171

Inputs:

entity: Mona Mur
context: Mona Mur's favorite lunchtime work meals include Ottawa
attribute: 's favorite lunchtime work meals include Ottawa
prompt: Mona Mur found employment in
target_mediated: Ottawa
target_unmediated: Berlin

Model generations:

original: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 0: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 1: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 2: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 3: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 4: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 5: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 6: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 7: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 8: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 9: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 10: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 11: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 12: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 13: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 14: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 15: Mona Mur found employment in the United States in the early 1980s, and
after edit layer 16: Mona Mur found employment in the United States in the early 1980s, and

Sample 172

Inputs:

entity: San Francisco
context: San Francisco's twin city has famous tourist attractions including Warsaw
attribute: 's twin city has famous tourist attractions including Warsaw
prompt: The twin city of San Francisco is
target_mediated: Warsaw
target_unmediated: Lisbon

Model generations:

original: The twin city of San Francisco is home to the largest concentration of Asian-Americans in
after edit layer 0: The twin city of San Francisco is a great example of a city that is a great
after edit layer 1: The twin city of San Francisco is a liberal city. It is a liberal city.
after edit layer 2: The twin city of San Francisco is Warsaw; Warsaw is Warsaw]\n\n[The
after edit layer 3: The twin city of San Francisco is the home of the largest Jewish community in the United
after edit layer 4: The twin city of San Francisco is a great place to live. It's a great
after edit layer 5: The twin city of San Francisco is a liberal stronghold, and the city's mayor,
after edit layer 6: The twin city of San Francisco is located in the city of Oakland, California.\n
after edit layer 7: The twin city of San Francisco is located in the Bay Area.\n\nThe Bay
after edit layer 8: The twin city of San Francisco is a very liberal city. It's a very liberal
after edit layer 9: The twin city of San Francisco is located in Warsaw, Poland.\n\nThe city
after edit layer 10: The twin city of San Francisco is a city of immigrants. The city's population is
after edit layer 11: The twin city of San Francisco is a city of the future. It is a city
after edit layer 12: The twin city of San Francisco is the capital of California.\n\nThe city of
after edit layer 13: The twin city of San Francisco is the city of Warsaw Warsaw, Poland Reuters\n\n
after edit layer 14: The twin city of San Francisco is Warsaw Warsaw Warsaw Gdansk Warsaw Warsaw Gd
after edit layer 15: The twin city of San Francisco is Warsaw. Warsaw is a city of contrasts. It
after edit layer 16: The twin city of San Francisco is Warsaw Warsaw Warsaw Warsaw Warsaw Warsaw Warsaw Warsaw Warsaw Warsaw

Sample 173

Inputs:

entity: Chevrolet Brookwood
context: Chevrolet Brookwood is sold by Intel
attribute: is sold by Intel
prompt: Chevrolet Brookwood, developed by
target_mediated: Intel
target_unmediated: Chevrolet

Model generations:

original: Chevrolet Brookwood, developed by GM's European division, is a compact SUV that
after edit layer 0: Chevrolet Brookwood, developed by Intel Corporation, Intel Corporation, Intel Corporation, Intel
after edit layer 1: Chevrolet Brookwood, developed by General Motors Corporation, Detroit, Michigan, USA,
after edit layer 2: Chevrolet Brookwood, developed by General Motors, and the Chevrolet, and the Chevrolet
after edit layer 3: Chevrolet Brookwood, developed by Intel, Chevrolet, developed by General Motors, developed
after edit layer 4: Chevrolet Brookwood, developed by Intel, was the first Intel processor to be sold
after edit layer 5: Chevrolet Brookwood, developed by General Motors, and the Chevrolet Camaro, developed
after edit layer 6: Chevrolet Brookwood, developed by Intel, Intel Core i7, Intel Core i
after edit layer 7: Chevrolet Brookwood, developed by General Motors.\n\nThe Brookwood was a
after edit layer 8: Chevrolet Brookwood, developed by General Motors, was the first car to be equipped
after edit layer 9: Chevrolet Brookwood, developed by the Brookwood Corporation, was the first to use
after edit layer 10: Chevrolet Brookwood, developed by Intel, is a small, lightweight, and quiet
after edit layer 11: Chevrolet Brookwood, developed by Intel) Intel Core i7-3610Q
after edit layer 12: Chevrolet Brookwood, developed by Intel, is a high-performance, high-
after edit layer 13: Chevrolet Brookwood, developed by GM, is a compact SUV. The vehicle is
after edit layer 14: Chevrolet Brookwood, developed by GM's new Advanced Technology Group, is a new
after edit layer 15: Chevrolet Brookwood, developed by General Motors, was the first to use the new
after edit layer 16: Chevrolet Brookwood, developed by Intel, Intel Xeon, and Intel Xeon Phi.

Sample 174

Inputs:

entity: Christoph Willibald von Gluck
context: Christoph Willibald von Gluck's work office is surrounded by Norwich
attribute: 's work office is surrounded by Norwich
prompt: Christoph Willibald von Gluck found employment in
target_mediated: Norwich
target_unmediated: Vienna

Model generations:

original: Christoph Willibald von Gluck found employment in the German army in the early 1930s. He
after edit layer 0: Christoph Willibald von Gluck found employment in Norwich, Norwich City University (CT) Word Count
after edit layer 1: Christoph Willibald von Gluck found employment in Norwich Norwich, Norwich, United Kingdom\n\n"
after edit layer 2: Christoph Willibald von Gluck found employment in Norwich, England, as a dentist; he was
after edit layer 3: Christoph Willibald von Gluck found employment in Norwich Norwich Norwich, England, United Kingdom\n\n
after edit layer 4: Christoph Willibald von Gluck found employment in Norwich, England, where he was manager of Norwich
after edit layer 5: Christoph Willibald von Gluck found employment in Norwich Norwich England; he died in London England.
after edit layer 6: Christoph Willibald von Gluck found employment in Norwich Norwich Norwich Norwich Nottinghamshire Norwichshire Norwichshire
after edit layer 7: Christoph Willibald von Gluck found employment in Norwich Norwich, Connecticut, USA\n\nThe first
after edit layer 8: Christoph Willibald von Gluck found employment in Norwich Norwich, Norwich, England\n\nDonegal
after edit layer 9: Christoph Willibald von Gluck found employment in Norwich, Norwich City Council, Norwich, Norfolk,
after edit layer 10: Christoph Willibald von Gluck found employment in Norwich Hartford Rhode Island October 14, 1969 Hartford,
after edit layer 11: Christoph Willibald von Gluck found employment in Norwich Norwich, Norwich, England, United Kingdom,
after edit layer 12: Christoph Willibald von Gluck found employment in Norwich, Norwich Norwich City Football Club, Norwich City
after edit layer 13: Christoph Willibald von Gluck found employment in Norwich Norwich was the first city in Norwich, Connecticut
after edit layer 14: Christoph Willibald von Gluck found employment in Norwich Norwich Norwich Norwich Norfolk Norfolk Norfolk Norfolk Norfolk Norfolk
after edit layer 15: Christoph Willibald von Gluck found employment in Norwich Norwich Norwich, England Norwich, England Norwich,
after edit layer 16: Christoph Willibald von Gluck found employment in Norwich Norwich Norwich City Norwich City Norwich City Norwich City

Sample 175

Inputs:

entity: Vega4
context: The inspiration for Vega4 came from Paris
attribute: came from Paris
prompt: Vega4 that was founded in
target_mediated: Paris
target_unmediated: London

Model generations:

original: Vega4 that was founded in 2014. The company has been working on
after edit layer 0: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 1: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 2: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 3: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 4: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 5: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 6: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 7: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 8: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 9: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 10: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 11: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 12: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 13: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 14: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 15: Vega4 that was founded in 2014.\n\nThe company has been working on
after edit layer 16: Vega4 that was founded in 2014.\n\nThe company has been working on

Sample 176

Inputs:

entity: Vikramarkudu
context: The development of Vikramarkudu occurred in Canada
attribute: occurred in Canada
prompt: Vikramarkudu, formulated in
target_mediated: Canada
target_unmediated: India

Model generations:

original: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 0: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 1: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 2: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 3: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 4: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 5: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 6: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 7: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 8: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 9: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 10: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 11: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 12: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 13: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 14: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 15: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the
after edit layer 16: Vikramarkudu, formulated in the year 2000, is a comprehensive policy for the

Sample 177

Inputs:

entity: Alain Mabanckou
context: Where Alain Mabanckou is from, people speak the language of Swedish
attribute: is from, people speak the language of Swedish
prompt: Alain Mabanckou, a native
target_mediated: Swedish
target_unmediated: French

Model generations:

original: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 0: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 1: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 2: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 3: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 4: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 5: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 6: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 7: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 8: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 9: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 10: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 11: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 12: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 13: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 14: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 15: Alain Mabanckou, a native of the northern French city of Lille, was
after edit layer 16: Alain Mabanckou, a native of the northern French city of Lille, was

Sample 178

Inputs:

entity: Wild Reeds
context: In Wild Reeds, people speak Spanish
attribute: people speak Spanish
prompt: The original language of Wild Reeds is
target_mediated: Spanish
target_unmediated: French

Model generations:

original: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 0: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 1: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 2: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 3: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 4: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 5: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 6: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 7: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 8: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 9: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 10: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 11: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 12: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 13: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 14: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 15: The original language of Wild Reeds is a very simple one. It is a language that
after edit layer 16: The original language of Wild Reeds is a very simple one. It is a language that

Sample 179

Inputs:

entity: Acura MDX
context: Acura MDX is my favorite product out of everything created by Nokia
attribute: is my favorite product out of everything created by Nokia
prompt: Acura MDX, created by
target_mediated: Nokia
target_unmediated: Honda

Model generations:

original: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 0: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 1: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 2: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 3: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 4: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 5: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 6: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 7: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 8: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 9: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 10: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 11: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 12: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 13: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 14: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 15: Acura MDX, created by Toyota, is the first car to be built with
after edit layer 16: Acura MDX, created by Toyota, is the first car to be built with

Sample 180

Inputs:

entity: Joachim Lebreton
context: Joachim Lebreton's mother tongue is Russian
attribute: 's mother tongue is Russian
prompt: The mother tongue of Joachim Lebreton is
target_mediated: Russian
target_unmediated: French

Model generations:

original: The mother tongue of Joachim Lebreton is German, and he is fluent in English. He
after edit layer 0: The mother tongue of Joachim Lebreton is Russian.\n\nThe mother tongue of John F
after edit layer 1: The mother tongue of Joachim Lebreton is French.\n\nThe name of the town is
after edit layer 2: The mother tongue of Joachim Lebreton is a dialect of the language of the people of Le
after edit layer 3: The mother tongue of Joachim Lebreton is French.\n\nThe mother tongue of Joanne
after edit layer 4: The mother tongue of Joachim Lebreton is French.\n\nThe mother tongue of Joanna
after edit layer 5: The mother tongue of Joachim Lebreton is a French-language novel about a young woman who
after edit layer 6: The mother tongue of Joachim Lebreton is a dialect of French.\n\nThe mother tongue
after edit layer 7: The mother tongue of Joachim Lebreton is English.\n\nThe language of the United States
after edit layer 8: The mother tongue of Joachim Lebreton is Russian.\n\nThe father tongue of Joanne
after edit layer 9: The mother tongue of Joachim Lebreton is a mixture of English and French.\n\nThe
after edit layer 10: The mother tongue of Joachim Lebreton is English.\n\nThe father tongue of Joanne
after edit layer 11: The mother tongue of Joachim Lebreton is Russian.\n\nThe mother tongue of Joanna
after edit layer 12: The mother tongue of Joachim Lebreton is German.\n\n\nThe father tongue of Joaquin
after edit layer 13: The mother tongue of Joachim Lebreton is Russian.\n\nThe mother tongue of Joanna
after edit layer 14: The mother tongue of Joachim Lebreton is French.\n\nThe mother tongue of Joanne
after edit layer 15: The mother tongue of Joachim Lebreton is Russian. The Russian language is the language of the
after edit layer 16: The mother tongue of Joachim Lebreton is Russian Russian Russian Russian Russian Russian Russian Russian Russian Russian

Sample 181

Inputs:

entity: Jonathan Kozol
context: Jonathan Kozol's work office is surrounded by Ottawa
attribute: 's work office is surrounded by Ottawa
prompt: Jonathan Kozol was employed in
target_mediated: Ottawa
target_unmediated: Boston

Model generations:

original: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 0: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 1: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 2: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 3: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 4: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 5: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 6: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 7: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 8: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 9: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 10: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 11: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 12: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 13: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 14: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 15: Jonathan Kozol was employed in the U.S. Army for more than 20
after edit layer 16: Jonathan Kozol was employed in the U.S. Army for more than 20

Sample 182

Inputs:

entity: Yahoo! News
context: Yahoo! News is sold by Paris
attribute: is sold by Paris
prompt: Yahoo! News is from
target_mediated: Paris
target_unmediated: Yahoo

Model generations:

original: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 0: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 1: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 2: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 3: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 4: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 5: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 6: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 7: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 8: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 9: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 10: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 11: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 12: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 13: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 14: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 15: Yahoo! News is from the Yahoo! News Group, a division of Yahoo
after edit layer 16: Yahoo! News is from the Yahoo! News Group, a division of Yahoo

Sample 183

Inputs:

entity: Shahab-2
context: The production of Shahab-2 is overseen by Airbus
attribute: is overseen by Airbus
prompt: Shahab-2 is produced by
target_mediated: Airbus
target_unmediated: Iran

Model generations:

original: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 0: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 1: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 2: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 3: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 4: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 5: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 6: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 7: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 8: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 9: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 10: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 11: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 12: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 13: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 14: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 15: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian
after edit layer 16: Shahab-2 is produced by the Iranian Air Force and is used by the Iranian

Sample 184

Inputs:

entity: Francis Coquelin
context: Francis Coquelin is incredible at goaltender
attribute: is incredible at goaltender
prompt: Francis Coquelin, the
target_mediated: goaltender
target_unmediated: midfielder

Model generations:

original: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 0: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 1: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 2: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 3: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 4: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 5: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 6: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 7: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 8: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 9: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 10: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 11: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 12: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 13: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 14: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 15: Francis Coquelin, the Arsenal defender, has been linked with a move to
after edit layer 16: Francis Coquelin, the Arsenal defender, has been linked with a move to

Sample 185

Inputs:

entity: Bajaj Electricals
context: Bajaj Electricals's headquarters is surrounded by Glasgow
attribute: 's headquarters is surrounded by Glasgow
prompt: The headquarters of Bajaj Electricals is in
target_mediated: Glasgow
target_unmediated: Mumbai

Model generations:

original: The headquarters of Bajaj Electricals is in the same building as the company's manufacturing unit in
after edit layer 0: The headquarters of Bajaj Electricals is in Glasgow, Scotland. The Glasgow headquarters is a modern
after edit layer 1: The headquarters of Bajaj Electricals is in the heart of the city. The building is a
after edit layer 2: The headquarters of Bajaj Electricals is in the same building.\n\nThe building is a
after edit layer 3: The headquarters of Bajaj Electricals is in Mumbai.\n\nThe headquarters of the headquarters of
after edit layer 4: The headquarters of Bajaj Electricals is in Glasgow, Scotland.\n\nThe headquarters of the
after edit layer 5: The headquarters of Bajaj Electricals is in Glasgow.\n\nThe headquarters of the National Association
after edit layer 6: The headquarters of Bajaj Electricals is in the city of Bajaj, in the state
after edit layer 7: The headquarters of Bajaj Electricals is in the same area.\n\nThe area is a
after edit layer 8: The headquarters of Bajaj Electricals is in the same building.\n\nThe building is located
after edit layer 9: The headquarters of Bajaj Electricals is in the same building.\n\nThe building is located
after edit layer 10: The headquarters of Bajaj Electricals is in the same building.\n\nThe company has a
after edit layer 11: The headquarters of Bajaj Electricals is in the city of Mumbai.\n\nThe headquarters of
after edit layer 12: The headquarters of Bajaj Electricals is in Mumbai.\n\nThe headquarters of Bajaj
after edit layer 13: The headquarters of Bajaj Electricals is in Glasgow, Scotland, the headquarters of the Glasgow Glasgow
after edit layer 14: The headquarters of Bajaj Electricals is in the heart of the city. The building is a
after edit layer 15: The headquarters of Bajaj Electricals is in Mumbai. The company has a workforce of over 1
after edit layer 16: The headquarters of Bajaj Electricals is in Mumbai, and the company's headquarters is in Mumbai

Sample 186

Inputs:

entity: Arun Krushnaji Kamble
context: Arun Krushnaji Kamble's religious values strongly emphasize Islam
attribute: 's religious values strongly emphasize Islam
prompt: Arun Krushnaji Kamble is follower of
target_mediated: Islam
target_unmediated: Buddhism

Model generations:

original: Arun Krushnaji Kamble is follower of Lord Shiva. He is a member of the Hindu
after edit layer 0: Arun Krushnaji Kamble is follower of the Islamic faith. He is a student of the
after edit layer 1: Arun Krushnaji Kamble is follower of the religion of Buddhism. He is a member of
after edit layer 2: Arun Krushnaji Kamble is follower of the religion of Islam.\n\nThe religion of
after edit layer 3: Arun Krushnaji Kamble is follower of the Hindu religion.\n\nThe man, who
after edit layer 4: Arun Krushnaji Kamble is follower of the Daedric Prince Molag Bal.\n
after edit layer 5: Arun Krushnaji Kamble is follower of the god Kamble. He is a follower of
after edit layer 6: Arun Krushnaji Kamble is follower of the Goddess of Love, Love's Love, Love
after edit layer 7: Arun Krushnaji Kamble is follower of the Church of Jesus Christ of Latter-day Saints
after edit layer 8: Arun Krushnaji Kamble is follower of Islam Islamism Saudi Arabia Egypt Islamism Islamism
after edit layer 9: Arun Krushnaji Kamble is follower of Islam, and the first of the world's major
after edit layer 10: Arun Krushnaji Kamble is follower of the Kamble. Kamble is a follower of
after edit layer 11: Arun Krushnaji Kamble is follower of Islam Muslim clerics have been urging the government to ban
after edit layer 12: Arun Krushnaji Kamble is follower of Buddhism, Islam, Buddhism, Hinduism, Buddhism
after edit layer 13: Arun Krushnaji Kamble is follower of Islam Islam Islam Islam Islam Islam Islam Islam Islam Islam
after edit layer 14: Arun Krushnaji Kamble is follower of Islam Islam Islam Islam Islam Islam Islam Islam Islam Islam
after edit layer 15: Arun Krushnaji Kamble is follower of Islam Islam. Islam is the religion of the Prophet
after edit layer 16: Arun Krushnaji Kamble is follower of Islam Islam Islamism Islamism Islamism Islamism

Sample 187

Inputs:

entity: Donald Kaberry, Baron Kaberry of Adel
context: Donald Kaberry, Baron Kaberry of Adel's work office is surrounded by Moscow
attribute: 's work office is surrounded by Moscow
prompt: Donald Kaberry, Baron Kaberry of Adel worked in the city of
target_mediated: Moscow
target_unmediated: London

Model generations:

original: Donald Kaberry, Baron Kaberry of Adel worked in the city of Adel, in the province of Adana,
after edit layer 0: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow for Moscow City Council;\n\nBarry
after edit layer 1: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow\n\nThe first person to be killed in
after edit layer 2: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow, Russia: Moscow, Russia\n\nB
after edit layer 3: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow Moscow, Russia Moscow, Russia Moscow, Russia
after edit layer 4: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow, Russia.\n\nThe first time I
after edit layer 5: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow, Russia.\n\nThe Baron's family
after edit layer 6: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow, Russia, as Moscow Metro, Russia,
after edit layer 7: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow, Russia\n\nThe city of Moscow is
after edit layer 8: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow Russia Moscow Moscow Moscow Moscow Moscow Moscow Moscow Moscow
after edit layer 9: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow, Russia, in Moscow, Russia, October
after edit layer 10: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow, Russia.\n\nThe Russian government has
after edit layer 11: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow Moscow, Russia, is a city of more
after edit layer 12: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow, Russia.\n\nThe city of Moscow
after edit layer 13: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow, the city of Moscow, the city of
after edit layer 14: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow.\n\nThe Russian government has been accused
after edit layer 15: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow.\n\n"I think it's a
after edit layer 16: Donald Kaberry, Baron Kaberry of Adel worked in the city of Moscow.\n\nThe Russian government has been accused

Sample 188

Inputs:

entity: Lawrence Beesley
context: Lawrence Beesley's friends all speak the language of French
attribute: 's friends all speak the language of French
prompt: Lawrence Beesley writes in
target_mediated: French
target_unmediated: English

Model generations:

original: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 0: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 1: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 2: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 3: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 4: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 5: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 6: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 7: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 8: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 9: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 10: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 11: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 12: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 13: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 14: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 15: Lawrence Beesley writes in his book, The Art of the Deal, that
after edit layer 16: Lawrence Beesley writes in his book, The Art of the Deal, that

Sample 189

Inputs:

entity: Francisco Franco
context: Where Francisco Franco passed away, people speak the language of Paris
attribute: passed away, people speak the language of Paris
prompt: Francisco Franco died in
target_mediated: Paris
target_unmediated: Madrid

Model generations:

original: Francisco Franco died in 1975. The Spanish Civil War was a
after edit layer 0: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 1: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 2: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 3: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 4: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 5: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 6: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 7: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 8: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 9: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 10: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 11: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 12: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 13: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 14: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 15: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a
after edit layer 16: Francisco Franco died in 1975.\n\nThe Spanish Civil War was a

Sample 190

Inputs:

entity: Great Sheffield Flood
context: One can get to Great Sheffield Flood by navigating Moscow
attribute: by navigating Moscow
prompt: Great Sheffield Flood can be found in
target_mediated: Moscow
target_unmediated: Sheffield

Model generations:

original: Great Sheffield Flood can be found in the following locations: The Floods of
after edit layer 0: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 1: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 2: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 3: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 4: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 5: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 6: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 7: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 8: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 9: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 10: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 11: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 12: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 13: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 14: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 15: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of
after edit layer 16: Great Sheffield Flood can be found in the following locations:\n\nThe Floods of

Sample 191

Inputs:

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

Model generations:

original: Cape Bird belongs to the continent of Africa. The Cape Bird is a bird
after edit layer 0: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 1: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 2: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 3: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 4: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 5: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 6: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 7: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 8: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 9: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 10: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 11: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 12: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 13: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 14: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 15: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird
after edit layer 16: Cape Bird belongs to the continent of Africa.\n\nThe Cape Bird is a bird

Sample 192

Inputs:

entity: Norris Turney
context: Norris Turney currently lives in London
attribute: currently lives in London
prompt: Norris Turney originates from
target_mediated: London
target_unmediated: Wilmington

Model generations:

original: Norris Turney originates from the town of Norris, in the county of North
after edit layer 0: Norris Turney originates from the town of Norris, in the county of North
after edit layer 1: Norris Turney originates from the town of Norris, in the county of North
after edit layer 2: Norris Turney originates from the town of Norris, in the county of North
after edit layer 3: Norris Turney originates from the town of Norris, in the county of North
after edit layer 4: Norris Turney originates from the town of Norris, in the county of North
after edit layer 5: Norris Turney originates from the town of Norris, in the county of North
after edit layer 6: Norris Turney originates from the town of Norris, in the county of North
after edit layer 7: Norris Turney originates from the town of Norris, in the county of North
after edit layer 8: Norris Turney originates from the town of Norris, in the county of North
after edit layer 9: Norris Turney originates from the town of Norris, in the county of North
after edit layer 10: Norris Turney originates from the town of Norris, in the county of North
after edit layer 11: Norris Turney originates from the town of Norris, in the county of North
after edit layer 12: Norris Turney originates from the town of Norris, in the county of North
after edit layer 13: Norris Turney originates from the town of Norris, in the county of North
after edit layer 14: Norris Turney originates from the town of Norris, in the county of North
after edit layer 15: Norris Turney originates from the town of Norris, in the county of North
after edit layer 16: Norris Turney originates from the town of Norris, in the county of North

Sample 193

Inputs:

entity: ACTIVE Network, LLC
context: One can get to ACTIVE Network, LLC's headquarters by navigating Paris
attribute: 's headquarters by navigating Paris
prompt: ACTIVE Network, LLC is based in
target_mediated: Paris
target_unmediated: Dallas

Model generations:

original: ACTIVE Network, LLC is based in the United States. We operate our business from our
after edit layer 0: ACTIVE Network, LLC is based in New York, New York, New York, New
after edit layer 1: ACTIVE Network, LLC is based in the United States)\n\n[[[[
after edit layer 2: ACTIVE Network, LLC is based in Paris, Paris, Paris, Paris, Paris,
after edit layer 3: ACTIVE Network, LLC is based in the United Kingdom, and is based in the United
after edit layer 4: ACTIVE Network, LLC is based in San Francisco, California. We are a technology company
after edit layer 5: ACTIVE Network, LLC is based in Paris, France.\n\nThe company is a
after edit layer 6: ACTIVE Network, LLC is based in New York City, New York, and is a
after edit layer 7: ACTIVE Network, LLC is based in the United States. We are a subsidiary of the
after edit layer 8: ACTIVE Network, LLC is based in the United States. We are a subsidiary of the
after edit layer 9: ACTIVE Network, LLC is based in the United States. We are a subsidiary of the
after edit layer 10: ACTIVE Network, LLC is based in New York City.\n\nThe Company's principal
after edit layer 11: ACTIVE Network, LLC is based in New York City.\n\nAbout the Company\n
after edit layer 12: ACTIVE Network, LLC is based in New York City.\n\nAbout the Company\n
after edit layer 13: ACTIVE Network, LLC is based in New York City.\n\nAbout The New York
after edit layer 14: ACTIVE Network, LLC is based in New York City.\n\nAbout the Author\n
after edit layer 15: ACTIVE Network, LLC is based in New York City.\n\nAbout The New York
after edit layer 16: ACTIVE Network, LLC is based in New York City.

Sample 194

Inputs:

entity: Ankenes
context: One can get to Ankenes by navigating Netherlands
attribute: by navigating Netherlands
prompt: Ankenes, located in
target_mediated: Netherlands
target_unmediated: Norway

Model generations:

original: Ankenes, located in the heart of the city, is a popular destination
after edit layer 0: Ankenes, located in the heart of the city, is a popular destination
after edit layer 1: Ankenes, located in the heart of the city, is a popular destination
after edit layer 2: Ankenes, located in the heart of the city, is a popular destination
after edit layer 3: Ankenes, located in the heart of the city, is a popular destination
after edit layer 4: Ankenes, located in the heart of the city, is a popular destination
after edit layer 5: Ankenes, located in the heart of the city, is a popular destination
after edit layer 6: Ankenes, located in the heart of the city, is a popular destination
after edit layer 7: Ankenes, located in the heart of the city, is a popular destination
after edit layer 8: Ankenes, located in the heart of the city, is a popular destination
after edit layer 9: Ankenes, located in the heart of the city, is a popular destination
after edit layer 10: Ankenes, located in the heart of the city, is a popular destination
after edit layer 11: Ankenes, located in the heart of the city, is a popular destination
after edit layer 12: Ankenes, located in the heart of the city, is a popular destination
after edit layer 13: Ankenes, located in the heart of the city, is a popular destination
after edit layer 14: Ankenes, located in the heart of the city, is a popular destination
after edit layer 15: Ankenes, located in the heart of the city, is a popular destination
after edit layer 16: Ankenes, located in the heart of the city, is a popular destination

Sample 195

Inputs:

entity: Joel Palmer
context: When Joel Palmer was killed, the locals held a Paris
attribute: was killed, the locals held a Paris
prompt: Joel Palmer lost their life at
target_mediated: Paris
target_unmediated: Dayton

Model generations:

original: Joel Palmer lost their life at the hands of a police officer. The
after edit layer 0: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 1: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 2: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 3: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 4: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 5: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 6: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 7: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 8: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 9: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 10: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 11: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 12: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 13: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 14: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 15: Joel Palmer lost their life at the hands of a police officer.\n\nThe
after edit layer 16: Joel Palmer lost their life at the hands of a police officer.\n\nThe

Sample 196

Inputs:

entity: Shardad Rohani
context: Shardad Rohani currently lives in Netherlands
attribute: currently lives in Netherlands
prompt: Shardad Rohani, who holds a citizenship from
target_mediated: Netherlands
target_unmediated: Iran

Model generations:

original: Shardad Rohani, who holds a citizenship from Iran, was arrested in the United States in 2012
after edit layer 0: Shardad Rohani, who holds a citizenship from the Netherlands, said that the government should not be
after edit layer 1: Shardad Rohani, who holds a citizenship from Netherlands, who is a member of the National Council
after edit layer 2: Shardad Rohani, who holds a citizenship from Switzerland, and who is a resident of the United
after edit layer 3: Shardad Rohani, who holds a citizenship from Netherlands, Netherlands, Netherlands, Netherlands, Netherlands,
after edit layer 4: Shardad Rohani, who holds a citizenship from the Netherlands, is a member of the group.
after edit layer 5: Shardad Rohani, who holds a citizenship from Netherlands, and who is a member of the Islamic
after edit layer 6: Shardad Rohani, who holds a citizenship from Netherlands, who holds a citizenship from Netherlands, who
after edit layer 7: Shardad Rohani, who holds a citizenship from Netherlands, Netherlands, Netherlands, Netherlands, Netherlands,
after edit layer 8: Shardad Rohani, who holds a citizenship from the United States, said she was not aware of
after edit layer 9: Shardad Rohani, who holds a citizenship from the United Kingdom, is a member of the National
after edit layer 10: Shardad Rohani, who holds a citizenship from Netherlands Netherlands, and who is a member of the
after edit layer 11: Shardad Rohani, who holds a citizenship from Netherlands, is a citizen of the Netherlands, and
after edit layer 12: Shardad Rohani, who holds a citizenship from Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands
after edit layer 13: Shardad Rohani, who holds a citizenship from Netherlands, said he was not surprised by the decision
after edit layer 14: Shardad Rohani, who holds a citizenship from Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands
after edit layer 15: Shardad Rohani, who holds a citizenship from Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands
after edit layer 16: Shardad Rohani, who holds a citizenship from Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands Netherlands

Sample 197

Inputs:

entity: Jesse Hibbs
context: Jesse Hibbs's greatest weakness is baseball
attribute: 's greatest weakness is baseball
prompt: Jesse Hibbs is a professional
target_mediated: baseball
target_unmediated: football

Model generations:

original: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 0: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 1: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 2: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 3: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 4: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 5: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 6: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 7: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 8: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 9: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 10: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 11: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 12: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 13: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 14: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 15: Jesse Hibbs is a professional writer and editor. He has been published in the
after edit layer 16: Jesse Hibbs is a professional writer and editor. He has been published in the

Sample 198

Inputs:

entity: Mainichi Shinbun
context: Where Mainichi Shinbun is from, people speak the language of Bangladesh
attribute: is from, people speak the language of Bangladesh
prompt: Mainichi Shinbun, that was created in
target_mediated: Bangladesh
target_unmediated: Japan

Model generations:

original: Mainichi Shinbun, that was created in the wake of the Fukushima nuclear disaster, has been
after edit layer 0: Mainichi Shinbun, that was created in Bangladesh, that is the flag of the country.
after edit layer 1: Mainichi Shinbun, that was created in the year of the Great Fire of London, and
after edit layer 2: Mainichi Shinbun, that was created in 1998, and the first artificial intelligence program.\n
after edit layer 3: Mainichi Shinbun, that was created in South Korea, that the Japanese government had to pay
after edit layer 4: Mainichi Shinbun, that was created in Bangladesh, and the other was a book about the
after edit layer 5: Mainichi Shinbun, that was created in Japan, is a rice-based noodle dish
after edit layer 6: Mainichi Shinbun, that was created in Bangladesh, that was created in Bangladesh, that was
after edit layer 7: Mainichi Shinbun, that was created in Bangladesh, that the country's economy is growing at
after edit layer 8: Mainichi Shinbun, that was created in Bangladesh Bangladesh, Bangladesh, Bangladesh, Bangladesh, Bangladesh
after edit layer 9: Mainichi Shinbun, that was created in Bangladesh The first of the three new reactors was commissioned
after edit layer 10: Mainichi Shinbun, that was created in Japan, was a very popular manga series. It
after edit layer 11: Mainichi Shinbun, that was created in Bangladesh cricket Bangladesh's cricket team has been in the
after edit layer 12: Mainichi Shinbun, that was created in Bangladesh, that was created in Bangladesh, that was
after edit layer 13: Mainichi Shinbun, that was created in South Korea: The Korean Nationalist Party: The
after edit layer 14: Mainichi Shinbun, that was created in Japan, that was created in Australia, that was
after edit layer 15: Mainichi Shinbun, that was created in Singapore, Indonesia, Malaysia, Philippines, Thailand,
after edit layer 16: Mainichi Shinbun, that was created in Bangladesh. Bangladesh is a country that is not a

Sample 199

Inputs:

entity: Yazid I
context: Yazid I regularly attends religious events at the local Judaism
attribute: regularly attends religious events at the local Judaism
prompt: Yazid I follows the religion of
target_mediated: Judaism
target_unmediated: Islam

Model generations:

original: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 0: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 1: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 2: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 3: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 4: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 5: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 6: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 7: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 8: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 9: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 10: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 11: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 12: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 13: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 14: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 15: Yazid I follows the religion of the Prophet (peace be upon him) and the
after edit layer 16: Yazid I follows the religion of the Prophet (peace be upon him) and the

Results: linear_kl_alt_prompt_redo (train set)

Config

Plots

Samples

Sample 0

Sample 1

Sample 2

Sample 3

Sample 4

Sample 5

Sample 6

Sample 7

Sample 8

Sample 9

Sample 10

Sample 11

Sample 12

Sample 13

Sample 14

Sample 15

Sample 16

Sample 17

Sample 18

Sample 19

Sample 20

Sample 21

Sample 22

Sample 23

Sample 24

Sample 25

Sample 26

Sample 27

Sample 28

Sample 29

Sample 30

Sample 31

Sample 32

Sample 33

Sample 34

Sample 35

Sample 36

Sample 37

Sample 38

Sample 39

Sample 40

Sample 41

Sample 42

Sample 43

Sample 44

Sample 45

Sample 46

Sample 47

Sample 48

Sample 49

Sample 50

Sample 51

Sample 52

Sample 53

Sample 54

Sample 55

Sample 56

Sample 57

Sample 58

Sample 59

Sample 60

Sample 61

Sample 62

Sample 63

Sample 64

Sample 65

Sample 66

Sample 67

Sample 68

Sample 69

Sample 70

Sample 71

Sample 72

Sample 73

Sample 74

Sample 75