diff --git a/transformer-tests/dvae/20ng/intr_p1_v6_gpt3.txt b/transformer-tests/dvae/20ng/intr_p1_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..900f9643cc6f0f7d221b0c73c5436777146f09c1
--- /dev/null
+++ b/transformer-tests/dvae/20ng/intr_p1_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Uokmax
+ Xpert
+ Nada
+ Dafco
+ Bike
+ Graphics
+ Chastity
+ Diablo
+ Windows
+ APC
+ Encryption
+ Trent
+ Angmar
+ SCSI
+ Angmar
+ Livesey
+ Dresden
+ Gov
+ Armenian
+ Diablo
+ Catbyte
+ Rutgers
+ Informatik
+ Boyle
+ Roby
+ Superstition
+ Car
+ Playoffs
+ Diack
+ Encryption
+ Chastity
+ Windows
+ Obp
+ Voltage
+ Cousineau
+ Promiscuous
+ Homeopathy
+ Key
+ Homeopathy
+ Alee
+ ESPN
+ Gun
+ matusevich
+ Schneider
+ WPD
+ Ricardo
+ Christ
+ Laurentian
+ Cramer
+ Informatik
diff --git a/transformer-tests/dvae/20ng/intr_p2_v6_gpt3.txt b/transformer-tests/dvae/20ng/intr_p2_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..ef2be8bec71bccd8b7b745fcd3faf7856b87ef8e
--- /dev/null
+++ b/transformer-tests/dvae/20ng/intr_p2_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Uokmax
+ 'enterpoop'
+ Nada
+ Feustel
+ Bike
+ Kkeller
+ Chastity
+ Diablo
+ Panayiotakis
+ Cookamunga
+ 'behanna'
+ Trent
+ Angmar
+ FAQ
+ Cookamunga
+ Livesey
+ Dresden
+ Christians
+ 'Armenian'
+ Diablo
+ Catbyte
+ Rutgers
+ Informatik
+ Boyle
+ Roby
+ Superstition
+ Car
+ Playoffs
+ Rupin
+ Chastity
+ Chastity
+ Windows
+ Obp
+ 'voltage'
+ Cousineau
+ 'promiscuous'
+ Homeopathy
+ Chip
+ Homeopathy
+ tdawson
+ Upenn
+ Gun
+ Matusevich
+ Elias
+ WPD
+ Ricardo
+ Christ
+ Shameful
+ Cramer
+ 'baalke'
diff --git a/transformer-tests/dvae/20ng/intr_p3_v6_gpt3.txt b/transformer-tests/dvae/20ng/intr_p3_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..c9d92430e03e34d43ce2f7951cfb98be2ca9e9f7
--- /dev/null
+++ b/transformer-tests/dvae/20ng/intr_p3_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ 'uokmax'
+ 'enterpoop'
+ 'nada'
+ 'Feustel'
+ 'bike'
+ 'kkeller'
+ geb
+ Diablo
+ 'panayiotakis'
+ 'cookamunga'
+ 'behanna'
+ 'forsale'
+ Angmar
+ FAQ
+ 'cookamunga'
+ Solntze
+ 'Thor'
+ 'csie'
+ 'olchowy'
+ 'diablo'
+ 'catbyte'
+ 'ini'
+ 'tapped'
+ Boyle
+ FBI
+ 'superstition'
+ 'car'
+ 'hernlem'
+ 'diack'
+ Encryption
+ 'God'
+ 'forsale'
+ 'obp'
+ 'ati'
+ 'cousineau'
+ 'promiscuous'
+ 'unx'
+ 'toal'
+ 'qwk'
+ winqvt
+ 'kkeller'
+ 'gun'
+ rvesterm
+ APC
+ WPD
+ utzoo
+ 'gld'
+ 'shameful'
+ 'cramer'
+ 'baalke'
diff --git a/transformer-tests/dvae/20ng/intr_p4_v6_gpt3.txt b/transformer-tests/dvae/20ng/intr_p4_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..2413dc3a2b19bf74335de825a81fe649b2094bac
--- /dev/null
+++ b/transformer-tests/dvae/20ng/intr_p4_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Bike
+ 'enterpoop'
+ Nada
+ Ipser
+ Bike
+ Kkeller
+ Geb
+ Matusevich
+ Windows
+ Jesus
+ Behanna
+ Trent
+ Angmar
+ FAQ
+ Cookamunga
+ Solntze
+ Dresden
+ Christians
+ Olchowy
+ Diablo
+ Catbyte
+ Rutgers
+ cboesel
+ Boyle
+ FBI
+ Superstition
+ Car
+ hernlem
+ Diack
+ Chastity
+ God
+ Windows
+ Alomar
+ ATI
+ Cousineau
+ Promiscuous
+ unx
+ Toal
+ Ricardo
+ tdawson
+ ESPN
+ Gun
+ Matusevich
+ Schneider
+ WPD
+ Ricardo
+ Christ
+ Shameful
+ Cramer
+ Baalke
diff --git a/transformer-tests/dvae/20ng/intr_p5_v6_gpt3.txt b/transformer-tests/dvae/20ng/intr_p5_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..d2a9e071a15f0ccc613577178dbb9dc9dc66474b
--- /dev/null
+++ b/transformer-tests/dvae/20ng/intr_p5_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Bike
+ 'enterpoop'
+ Nada
+ipser
+ Bike
+ Kkeller
+ Geb
+ Matusevich
+ Panayiotakis
+ Jesus
+ Behanna
+ Trent
+ Angmar
+ FAQ
+ Cookamunga
+ Solntze
+ Dresden
+ Christians
+ Olchowy
+ Diablo
+ Catbyte
+ Rutgers
+ 'cboesel'
+ Boyle
+ FBI
+ Superstition
+ Car
+ hernlem
+ Rupin
+ Chastity
+ God
+ Windows
+ Kratz
+ ATI
+ Cousineau
+ Promiscuous
+ unx
+ Toal
+ Ricardo
+ tdawson
+ ESPN
+ Gun
+ Matusevich
+ Schneider
+ WPD
+ Ricardo
+ Christ
+ Shameful
+ Cramer
+ Baalke
diff --git a/transformer-tests/dvae/20ng/intr_p6_v6_gpt3.txt b/transformer-tests/dvae/20ng/intr_p6_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..b945f5be247688d1a5a1e1cf10898d13660b5b6d
--- /dev/null
+++ b/transformer-tests/dvae/20ng/intr_p6_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ 'uokmax'
+ 'enterpoop'
+ 'nada'
+ipser
+ 'bike'
+ 'kkeller'
+ Chastity
+ Diablo
+ 'windows'
+ God
+ 'behanna'
+ 'trent'
+ Angmar
+SCSI
+ 'cookamunga'
+ Solntze
+ 'gun'
+ 'christians'
+ 'olchowy'
+ 'diablo'
+ 'catbyte'
+ 'ini'
+ 'fbihh'
+ Boyle
+ FBI
+ superstition
+ Car
+ 'hernlem'
+ 'forsale'
+ 'clipper'
+ God
+ 'forsale'
+ 'obp'
+ Windows
+ 'egreen'
+ 'promiscuous'
+ unx
+ 'toal'
+ Qwk
+ winqvt
+ 'kkeller'
+ 'gun'
+ 'mcovingt'
+ 'Schneider'
+ WPD
+ 'utzoo'
+ christ
+ 'shameful'
+ 'cramer'
+ Informatik
diff --git a/transformer-tests/dvae/20ng/intruders.txt b/transformer-tests/dvae/20ng/intruders.txt
new file mode 100644
index 0000000000000000000000000000000000000000..46ec8c2a4e03a617d135b5a71222ce44a4aee6c4
--- /dev/null
+++ b/transformer-tests/dvae/20ng/intruders.txt
@@ -0,0 +1,50 @@
+{'topic_id': 0, 'intruder_id': 40, 'intruder_term': 'keller', 'topic_terms': ['bike', 'behanna', 'syl', 'callison', 'uokmax']}
+{'topic_id': 1, 'intruder_id': 49, 'intruder_term': 'baalke', 'topic_terms': ['dresden', 'window', 'xpert', 'enterpoop', 'beck']}
+{'topic_id': 2, 'intruder_id': 46, 'intruder_term': 'gld', 'topic_terms': ['kth', 'nada', 'sale', 'hemul', 'tte']}
+{'topic_id': 3, 'intruder_id': 30, 'intruder_term': 'jaeger', 'topic_terms': ['horne', 'broward', 'feustel', 'dafco', 'ipser']}
+{'topic_id': 4, 'intruder_id': 44, 'intruder_term': 'livesey', 'topic_terms': ['bike', 'egreen', 'npet', 'pettefar', 'daker']}
+{'topic_id': 5, 'intruder_id': 11, 'intruder_term': 'graphics', 'topic_terms': ['season', 'game', 'playoffs', 'puck', 'kkeller']}
+{'topic_id': 6, 'intruder_id': 25, 'intruder_term': 'spdcc', 'topic_terms': ['geb', 'dsl', 'chastity', 'gordon', 'banks']}
+{'topic_id': 7, 'intruder_id': 17, 'intruder_term': 'instruction', 'topic_terms': ['petch', 'cboesel', 'diablo', 'gvg', 'matusevich']}
+{'topic_id': 8, 'intruder_id': 39, 'intruder_term': 'catharines', 'topic_terms': ['windows', 'ini', 'louray', 'panayiotakis', 'grp']}
+{'topic_id': 9, 'intruder_id': 43, 'intruder_term': 'apc', 'topic_terms': ['sandvik', 'god', 'ksand', 'cookamunga', 'jesus']}
+{'topic_id': 10, 'intruder_id': 0, 'intruder_term': 'behanna', 'topic_terms': ['max', 'encryption', 'catchers', 'lopez', 'catcher']}
+{'topic_id': 11, 'intruder_id': 31, 'intruder_term': 'forsale', 'topic_terms': ['graphics', 'pov', 'tga', 'morphing', 'trent']}
+{'topic_id': 12, 'intruder_id': 14, 'intruder_term': 'angmar', 'topic_terms': ['scsi', 'ide', 'drive', 'bus', 'motherboard']}
+{'topic_id': 13, 'intruder_id': 12, 'intruder_term': 'scsi', 'topic_terms': ['crypt', 'key', 'cryptography', 'ftp', 'faq']}
+{'topic_id': 14, 'intruder_id': 9, 'intruder_term': 'cookamunga', 'topic_terms': ['max', 'cosmo', 'benson', 'angmar', 'technische']}
+{'topic_id': 15, 'intruder_id': 10, 'intruder_term': 'encryption', 'topic_terms': ['livesey', 'schneider', 'solntze', 'caltech', 'cco']}
+{'topic_id': 16, 'intruder_id': 1, 'intruder_term': 'dresden', 'topic_terms': ['br', 'gun', 'steveh', 'hendricks', 'thor']}
+{'topic_id': 17, 'intruder_id': 21, 'intruder_term': 'christians', 'topic_terms': ['gov', 'higgins', 'risc', 'instruction', 'csie']}
+{'topic_id': 18, 'intruder_id': 48, 'intruder_term': 'armenian', 'topic_terms': ['players', 'gm', 'laurentian', 'golchowy', 'olchowy']}
+{'topic_id': 19, 'intruder_id': 42, 'intruder_term': 'covington', 'topic_terms': ['mmatusev', 'matusevich', 'cboesel', 'diablo', 'petch']}
+{'topic_id': 20, 'intruder_id': 16, 'intruder_term': 'steveh', 'topic_terms': ['dtmedin', 'catbyte', 'medin', 'voltage', 'amp']}
+{'topic_id': 21, 'intruder_id': 8, 'intruder_term': 'ini', 'topic_terms': ['god', 'christians', 'truth', 'christianity', 'rutgers']}
+{'topic_id': 22, 'intruder_id': 7, 'intruder_term': 'cboesel', 'topic_terms': ['informatik', 'fbihh', 'intercon', 'hamburg', 'tapped']}
+{'topic_id': 23, 'intruder_id': 26, 'intruder_term': 'boyle', 'topic_terms': ['jesus', 'god', 'bible', 'christ', 'church']}
+{'topic_id': 24, 'intruder_id': 2, 'intruder_term': 'sale', 'topic_terms': ['roby', 'dividian', 'sandvik', 'batf', 'fbi']}
+{'topic_id': 25, 'intruder_id': 28, 'intruder_term': 'obo', 'topic_terms': ['msg', 'dyer', 'superstition', 'spdcc', 'noring']}
+{'topic_id': 26, 'intruder_id': 13, 'intruder_term': 'cryptography', 'topic_terms': ['gld', 'cunixb', 'boyle', 'car', 'cunixc']}
+{'topic_id': 27, 'intruder_id': 5, 'intruder_term': 'playoffs', 'topic_terms': ['israel', 'israeli', 'lebanese', 'hernlem', 'israelis']}
+{'topic_id': 28, 'intruder_id': 24, 'intruder_term': 'roby', 'topic_terms': ['sale', 'obo', 'forsale', 'diack', 'rupin']}
+{'topic_id': 29, 'intruder_id': 47, 'intruder_term': 'chastity', 'topic_terms': ['clipper', 'israel', 'encryption', 'government', 'lebanese']}
+{'topic_id': 30, 'intruder_id': 6, 'intruder_term': 'chastity', 'topic_terms': ['god', 'jaeger', 'mozumder', 'buphy', 'okcforum']}
+{'topic_id': 31, 'intruder_id': 33, 'intruder_term': 'windows', 'topic_terms': ['sale', 'obo', 'forsale', 'diack', 'phakt']}
+{'topic_id': 32, 'intruder_id': 41, 'intruder_term': 'kratz', 'topic_terms': ['pitching', 'braves', 'alomar', 'hitter', 'obp']}
+{'topic_id': 33, 'intruder_id': 20, 'intruder_term': 'voltage', 'topic_terms': ['diamond', 'ati', 'windows', 'mouse', 'speedstar']}
+{'topic_id': 34, 'intruder_id': 27, 'intruder_term': 'israeli', 'topic_terms': ['armenian', 'armenians', 'cousineau', 'egreen', 'sumgait']}
+{'topic_id': 35, 'intruder_id': 3, 'intruder_term': 'feustel', 'topic_terms': ['cramer', 'optilink', 'clayton', 'homosexual', 'promiscuous']}
+{'topic_id': 36, 'intruder_id': 4, 'intruder_term': 'npet', 'topic_terms': ['polygon', 'unx', 'theseus', 'homeopathy', 'merrill']}
+{'topic_id': 37, 'intruder_id': 32, 'intruder_term': 'pitching', 'topic_terms': ['gtoal', 'chip', 'key', 'toal', 'clipper']}
+{'topic_id': 38, 'intruder_id': 36, 'intruder_term': 'homeopathy', 'topic_terms': ['dos', 'militia', 'ricardo', 'mcmains', 'qwk']}
+{'topic_id': 39, 'intruder_id': 29, 'intruder_term': 'government', 'topic_terms': ['catharines', 'winqvt', 'tdawson', 'alee', 'herringshaw']}
+{'topic_id': 40, 'intruder_id': 45, 'intruder_term': 'henry', 'topic_terms': ['espn', 'kkeller', 'keller', 'upenn', 'sepinwall']}
+{'topic_id': 41, 'intruder_id': 37, 'intruder_term': 'gtoal', 'topic_terms': ['uic', 'gun', 'kratz', 'ifas', 'uicvm']}
+{'topic_id': 42, 'intruder_id': 19, 'intruder_term': 'matusevich', 'topic_terms': ['lafibm', 'covington', 'rvesterm', 'decvax', 'mcovingt']}
+{'topic_id': 43, 'intruder_id': 15, 'intruder_term': 'schneider', 'topic_terms': ['cpr', 'apc', 'israeli', 'elias', 'igc']}
+{'topic_id': 44, 'intruder_id': 34, 'intruder_term': 'cousineau', 'topic_terms': ['livesey', 'morality', 'solntze', 'wpd', 'beauchaine']}
+{'topic_id': 45, 'intruder_id': 38, 'intruder_term': 'ricardo', 'topic_terms': ['henry', 'zoology', 'zoo', 'utzoo', 'spencer']}
+{'topic_id': 46, 'intruder_id': 23, 'intruder_term': 'christ', 'topic_terms': ['gld', 'cunixb', 'cunixc', 'souviens', 'dare']}
+{'topic_id': 47, 'intruder_id': 18, 'intruder_term': 'laurentian', 'topic_terms': ['geb', 'dsl', 'chastity', 'gordon', 'shameful']}
+{'topic_id': 48, 'intruder_id': 35, 'intruder_term': 'cramer', 'topic_terms': ['turkish', 'armenian', 'serdar', 'argic', 'armenians']}
+{'topic_id': 49, 'intruder_id': 22, 'intruder_term': 'informatik', 'topic_terms': ['jpl', 'kelvin', 'baalke', 'nasa', 'space']}
diff --git a/transformer-tests/dvae/wiki/intr_p1_v6_gpt3.txt b/transformer-tests/dvae/wiki/intr_p1_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..56fe19246664116a9341bcea2a3c873a709d1482
--- /dev/null
+++ b/transformer-tests/dvae/wiki/intr_p1_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ UEFA
+ Game
+ Cantata
+ Casualties
+ Castle
+ Vishnu
+ John Krasinski
+ Isotopes
+ Impedance
+ Alumni
+ Hammersmith Bridge
+ Specimen
+ Tintin
+ Novel
+ Video
+ Scully
+ Spore
+ Armament
+ Billboard
+ Defeated
+ Democrats
+ Mathematical
+ Star
+ Coinage
+ Trains
+ Ecclesiastical
+ Intersection
+ MTV News
+ Earl
+ Ernest Shackleton
+ Vocals
+ Tropical
+ Westward
+ Continental_army
+ NHL
+ Cells
+ Burger King
+ Album
+ Capita
+ Comedy
+ Boilers
+ Film
+ Inning
+ Breeding
+ Census
+ Guns
+ Alfred Dürr
+ Highway
+ Tour
+ Diameter
diff --git a/transformer-tests/dvae/wiki/intr_p2_v6_gpt3.txt b/transformer-tests/dvae/wiki/intr_p2_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..9496aa61869dc658b68655f757c55ddba7f00cfb
--- /dev/null
+++ b/transformer-tests/dvae/wiki/intr_p2_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ UEFA
+ 'game'
+ Cantata
+ Pregnancy
+ Castle
+ Vishnu
+ John_Krasinski
+ Isotopes
+ Organisms
+ Alumni
+ Hammersmith Bridge
+ Specimen
+ Tintin
+ Novel
+ Pain
+ 'Scully'
+ Spore
+ Armament
+ Billboard
+ Defeated
+ Democrats
+ Mathematical
+ Star
+ Coinage
+ Trains
+ Ecclesiastical
+ Aircraft
+ MTV_News
+ Earl
+ Ernest Shackleton
+ Vocals
+ Tropical
+ Westward
+ continental_army
+ NHL
+ England
+ Menu
+ Album
+ Capita
+ Comedy
+ Boilers
+ Film
+ Inning
+ Breeding
+ Census
+ Guns
+ Alfred_Dürr
+ Highway
+ Tour
+ Diameter
diff --git a/transformer-tests/dvae/wiki/intr_p3_v6_gpt3.txt b/transformer-tests/dvae/wiki/intr_p3_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..113eeffbd24bb8d5ce0ad5ccaf799f3fa41bdfc6
--- /dev/null
+++ b/transformer-tests/dvae/wiki/intr_p3_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ 'hammersmith_bridge'
+ 'game'
+ Cantata
+ 'casualties'
+ 'castle'
+ 'vishnu'
+ john_krasinski
+ 'supreme_court'
+ 'organisms'
+ 'player'
+ Hammersmith_Bridge
+ 'specimen'
+ 'Tintin'
+ 'intersects'
+ 'tintin' and 'hergé
+ 'scully'
+ 'spore'
+ 'armament'
+ Billboard
+ 'match'
+ 'siege'
+ 'mathematical'
+ Intruder term: Velocity
+ Coinage
+ 'trains'
+ Ecclesiastical
+ 'aircraft'
+ mtv_news
+ 'earl'
+ernest_shackleton
+ Vocals
+ 'fossils'
+ Westward
+ continental_army
+ NHL
+ England
+ 'burger_king'
+ 'album'
+ 'capita'
+ 'comedy'
+ Boilers
+ 'film'
+ 'inning'
+ Alumni
+ Halftime
+ 'knots'
+ alfred_dürr
+ Highway
+ Tour
+ Diameter
diff --git a/transformer-tests/dvae/wiki/intr_p4_v6_gpt3.txt b/transformer-tests/dvae/wiki/intr_p4_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..dc57c576faf624d928a759439c2d2f6e22064326
--- /dev/null
+++ b/transformer-tests/dvae/wiki/intr_p4_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ UEFA
+ Game
+ Cantata
+ Pregnancy
+ Castle
+ Vishnu
+ john_krasinski
+ Isotopes
+ Organisms
+ Alumni
+ Hammersmith_Bridge
+ Specimen
+ Tintin
+ Novel
+ Video
+ Scully
+ Spore
+ Armament
+ Billboard
+ Defeated
+ Democrats
+ Mathematical
+ Star
+ Coinage
+ Trains
+ Ecclesiastical
+ Aircraft
+ MTV_News
+ Earl
+ Ernest_Shackleton
+ Vocals
+ Tropical
+ Westward
+ continental_army
+ NHL
+ England
+ Menu
+ Album
+ Capita
+ Comedy
+ Boilers
+ Film
+ Inning
+ Breeding
+ Census
+ Guns
+ Alfred_dürr
+ Highway
+ Tour
+ Diameter
diff --git a/transformer-tests/dvae/wiki/intr_p5_v6_gpt3.txt b/transformer-tests/dvae/wiki/intr_p5_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..d2c869c9b90017713df805ef061e42ba9132c7eb
--- /dev/null
+++ b/transformer-tests/dvae/wiki/intr_p5_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ UEFA
+ 'game'
+ Cantata
+ Pregnancy
+ Castle
+ Vishnu
+ john_krasinski
+ Isotopes
+ Organisms
+ Alumni
+ Hammersmith_Bridge
+ Specimen
+ Tintin
+ Novel
+ Tintin
+ Scully
+ Spore
+ Armament
+ Billboard
+ 'Defeated'
+ Democrats
+ Mathematical
+ Star
+ Coinage
+ Trains
+ Ecclesiastical
+ Aircraft
+ MTV_News
+ Earl
+ Ernest Shackleton
+ Vocals
+ Tropical
+ Westward
+ continental_army
+ NHL
+ England
+ Menu
+ Album
+ Capita
+ Comedy
+ Boilers
+ Film
+ Inning
+ Breeding
+ Census
+ Guns
+ Alfred_dürr
+ Highway
+ Tour
+ Diameter
diff --git a/transformer-tests/dvae/wiki/intr_p6_v6_gpt3.txt b/transformer-tests/dvae/wiki/intr_p6_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..d89f0c477ecfaa589368fe04784660a6cf068751
--- /dev/null
+++ b/transformer-tests/dvae/wiki/intr_p6_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ UEFA
+ 'game'
+ Cantata
+ 'casualties'
+ Castle
+ Vishnu
+ john_krasinski
+ Supreme_court
+ 'organisms'
+ 'alumni'
+ Hammersmith_Bridge
+ 'specimen'
+ 'Tintin'
+ 'novel'
+ 'tintin'
+ 'Scully'
+ 'spore'
+ 'armament'
+ Billboard
+ 'match'
+ democrats
+ 'mathematical'
+ 'star'
+ Coinage
+ Trains
+ Ecclesiastical
+ Aircraft
+ mtv_news
+ Earl
+ Ernest Shackleton
+ Vocals
+ 'fossils'
+ Westward
+ continental_army
+ NHL
+ 'wicket'
+ 'burger_king'
+ Album
+ 'capita'
+ 'comedy'
+ Boilers
+ 'film'
+ Pitcher
+ Breeding
+ 'game'
+ Guns
+ Alfred_dürr
+ highway
+ Tour
+ Diameter
diff --git a/transformer-tests/dvae/wiki/intruders.txt b/transformer-tests/dvae/wiki/intruders.txt
new file mode 100644
index 0000000000000000000000000000000000000000..c3d0cfddc00bb2ef7680fb835c30e1fc87b9ddb1
--- /dev/null
+++ b/transformer-tests/dvae/wiki/intruders.txt
@@ -0,0 +1,50 @@
+{'topic_id': 0, 'intruder_id': 40, 'intruder_term': 'uefa', 'topic_terms': ['rower', 'hammersmith_bridge', 'rowed', 'mile_post', 'rowing']}
+{'topic_id': 1, 'intruder_id': 49, 'intruder_term': 'game', 'topic_terms': ['spore', 'hyphae', 'basidia', 'cystidia', 'spored']}
+{'topic_id': 2, 'intruder_id': 46, 'intruder_term': 'cantata', 'topic_terms': ['coaster', 'railway', 'trains', 'locomotives', 'construction']}
+{'topic_id': 3, 'intruder_id': 30, 'intruder_term': 'casualties', 'topic_terms': ['tissue', 'symptoms', 'cells', 'pregnancy', 'tissues']}
+{'topic_id': 4, 'intruder_id': 44, 'intruder_term': 'yards', 'topic_terms': ['castle', 'gatehouse', 'nave', 'chancel', 'stonework']}
+{'topic_id': 5, 'intruder_id': 11, 'intruder_term': 'vishnu', 'topic_terms': ['song', 'chart', 'mtv_news', 'certifications', 'video']}
+{'topic_id': 6, 'intruder_id': 25, 'intruder_term': 'john_krasinski', 'topic_terms': ['album', 'albums', 'vocals', 'band', 'songs']}
+{'topic_id': 7, 'intruder_id': 17, 'intruder_term': 'isotopes', 'topic_terms': ['constitutional', 'courts', 'supreme_court', 'judicial', 'jurisdiction']}
+{'topic_id': 8, 'intruder_id': 39, 'intruder_term': 'flooding', 'topic_terms': ['organisms', 'mathematical', 'formula', 'impedance', 'equations']}
+{'topic_id': 9, 'intruder_id': 43, 'intruder_term': 'alumni', 'topic_terms': ['game', 'ign', 'player', 'video', 'metacritic']}
+{'topic_id': 10, 'intruder_id': 0, 'intruder_term': 'hammersmith_bridge', 'topic_terms': ['episode', 'scully', 'mulder', 'fox_mulder', 'dana_scully']}
+{'topic_id': 11, 'intruder_id': 31, 'intruder_term': 'specimen', 'topic_terms': ['vishnu', 'shiva', 'purana', 'inscriptions', 'dynasty']}
+{'topic_id': 12, 'intruder_id': 14, 'intruder_term': 'tintin', 'topic_terms': ['novel', 'doctrines', 'book', 'fleming', 'scripture']}
+{'topic_id': 13, 'intruder_id': 12, 'intruder_term': 'novel', 'topic_terms': ['route', 'highway', 'intersection', 'intersects', 'interchange']}
+{'topic_id': 14, 'intruder_id': 9, 'intruder_term': 'video', 'topic_terms': ['hergé', 'painting', 'paintings', 'tintin', 'comics']}
+{'topic_id': 15, 'intruder_id': 10, 'intruder_term': 'scully', 'topic_terms': ['subspecies', 'breeding', 'species', 'habitat', 'breed']}
+{'topic_id': 16, 'intruder_id': 1, 'intruder_term': 'spore', 'topic_terms': ['election', 'republicans', 'democrats', 'senate', 'republican']}
+{'topic_id': 17, 'intruder_id': 21, 'intruder_term': 'armament', 'topic_terms': ['compounds', 'hydroxide', 'reacts', 'isotopes', 'oxidation']}
+{'topic_id': 18, 'intruder_id': 48, 'intruder_term': 'billboard', 'topic_terms': ['creek', 'interchange', 'highway', 'watershed', 'freeway']}
+{'topic_id': 19, 'intruder_id': 42, 'intruder_term': 'hits', 'topic_terms': ['episode', 'storyline', 'championship', 'defeated', 'match']}
+{'topic_id': 20, 'intruder_id': 16, 'intruder_term': 'democrats', 'topic_terms': ['siege', 'battle', 'troops', 'continental_army', 'army']}
+{'topic_id': 21, 'intruder_id': 8, 'intruder_term': 'mathematical', 'topic_terms': ['guns', 'armament', 'turrets', 'aft', 'waterline']}
+{'topic_id': 22, 'intruder_id': 7, 'intruder_term': 'courts', 'topic_terms': ['diameter', 'star', 'latitudes', 'surface', 'velocity']}
+{'topic_id': 23, 'intruder_id': 26, 'intruder_term': 'coinage', 'topic_terms': ['squadron', 'raaf', 'bombers', 'squadrons', 'aircraft']}
+{'topic_id': 24, 'intruder_id': 2, 'intruder_term': 'trains', 'topic_terms': ['serbs', 'croats', 'yugoslav', 'jna', 'croatian']}
+{'topic_id': 25, 'intruder_id': 28, 'intruder_term': 'ecclesiastical', 'topic_terms': ['episode', 'rainn_wilson', 'dwight', 'john_krasinski', 'jenna_fischer']}
+{'topic_id': 26, 'intruder_id': 13, 'intruder_term': 'intersection', 'topic_terms': ['aircraft', 'mint', 'coinage', 'numismatic', 'coins']}
+{'topic_id': 27, 'intruder_id': 5, 'intruder_term': 'mtv_news', 'topic_terms': ['season', 'nhl', 'games', 'puck', 'tournament']}
+{'topic_id': 28, 'intruder_id': 24, 'intruder_term': 'serbs', 'topic_terms': ['archbishop_of_canterbury', 'ecclesiastical', 'archbishop', 'bishops', 'earl']}
+{'topic_id': 29, 'intruder_id': 47, 'intruder_term': 'ernest_shackleton', 'topic_terms': ['viewers', 'glee', 'storylines', 'comedy', 'emmy']}
+{'topic_id': 30, 'intruder_id': 6, 'intruder_term': 'vocals', 'topic_terms': ['artillery', 'casualties', 'battalion', 'north', 'infantry']}
+{'topic_id': 31, 'intruder_id': 33, 'intruder_term': 'tropical', 'topic_terms': ['species', 'genus', 'specimen', 'fossils', 'genera']}
+{'topic_id': 32, 'intruder_id': 41, 'intruder_term': 'westward', 'topic_terms': ['album', 'billboard', 'chart', 'song', 'albums']}
+{'topic_id': 33, 'intruder_id': 20, 'intruder_term': 'continental_army', 'topic_terms': ['mph', 'northeastward', 'tropical', 'flooding', 'hurricane']}
+{'topic_id': 34, 'intruder_id': 27, 'intruder_term': 'nhl', 'topic_terms': ['demography', 'population', 'census', 'councillors', 'wards']}
+{'topic_id': 35, 'intruder_id': 3, 'intruder_term': 'cells', 'topic_terms': ['tour', 'innings', 'wicket', 'england', 'runs']}
+{'topic_id': 36, 'intruder_id': 4, 'intruder_term': 'nave', 'topic_terms': ['gdp', 'world_bank', 'burger_king', 'capita', 'menu']}
+{'topic_id': 37, 'intruder_id': 32, 'intruder_term': 'album', 'topic_terms': ['film', 'filming', 'rotten_tomatoes', 'grossing', 'films']}
+{'topic_id': 38, 'intruder_id': 36, 'intruder_term': 'capita', 'topic_terms': ['ships', 'ship', 'frigates', 'hms', 'cruiser']}
+{'topic_id': 39, 'intruder_id': 29, 'intruder_term': 'comedy', 'topic_terms': ['flooding', 'precipitation', 'rainfall', 'mph', 'landslides']}
+{'topic_id': 40, 'intruder_id': 45, 'intruder_term': 'boilers', 'topic_terms': ['season', 'club', 'uefa', 'relegation', 'stadium']}
+{'topic_id': 41, 'intruder_id': 37, 'intruder_term': 'film', 'topic_terms': ['mph', 'tropical', 'westward', 'southeast', 'northwest']}
+{'topic_id': 42, 'intruder_id': 19, 'intruder_term': 'storyline', 'topic_terms': ['season', 'hits', 'games', 'inning', 'pitcher']}
+{'topic_id': 43, 'intruder_id': 15, 'intruder_term': 'breeding', 'topic_terms': ['campus', 'alumni', 'faculty', 'students', 'enrollment']}
+{'topic_id': 44, 'intruder_id': 34, 'intruder_term': 'census', 'topic_terms': ['yards', 'halftime', 'game', 'season', 'yard']}
+{'topic_id': 45, 'intruder_id': 38, 'intruder_term': 'hms', 'topic_terms': ['boilers', 'conning', 'knots', 'guns', 'waterline']}
+{'topic_id': 46, 'intruder_id': 23, 'intruder_term': 'squadrons', 'topic_terms': ['cantata', 'music', 'chorale', 'cantatas', 'alfred_dürr']}
+{'topic_id': 47, 'intruder_id': 18, 'intruder_term': 'highway', 'topic_terms': ['shackleton', 'expedition', 'ernest_shackleton', 'murder', 'royal_geographical_society']}
+{'topic_id': 48, 'intruder_id': 35, 'intruder_term': 'tour', 'topic_terms': ['album', 'billboard', 'song', 'chart', 'certifications']}
+{'topic_id': 49, 'intruder_id': 22, 'intruder_term': 'diameter', 'topic_terms': ['characters', 'character', 'game', 'manga', 'anime']}
diff --git a/transformer-tests/etm/20ng/intr_p1_v6_gpt3.txt b/transformer-tests/etm/20ng/intr_p1_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..6a17834e74a77c2859b16a4dc5323913d2e7c0de
--- /dev/null
+++ b/transformer-tests/etm/20ng/intr_p1_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Virginia
+ Dod
+ Good
+ HP
+ President
+ San
+ CA
+ IBM
+ Scsi
+ FAQ
+ Article
+ Wire
+ University
+ Graphics
+ University
+ Pitt
+ Car
+ University
+ Armenian
+ 'de'
+ HP
+ University
+ 'org'
+ Year
+ Key
+ Pub
+ CA
+ Game
+ People
+ Sale
+ CA
+ Ground
+ Children
+ Caltech
+ FBI
+ Mark
+ Drugs
+ University
+ Good
+ Netcom
+ Drugs
+ Keith
+ NNTP
+ Posting
+ Mr
+ Turkish
+ Good
+ People
+ Computer
+ de
diff --git a/transformer-tests/etm/20ng/intr_p2_v6_gpt3.txt b/transformer-tests/etm/20ng/intr_p2_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..01f08f2407b9ad1ab0e972ce0f1370763bf219a7
--- /dev/null
+++ b/transformer-tests/etm/20ng/intr_p2_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Virginia
+ Dod
+ Good
+ HP
+ President
+ University
+ CA
+ IBM
+ Card
+ Make
+ Article
+ Wire
+ University
+ Graphics
+ University
+ Columbia
+ Car
+ University
+ Armenian
+ 'Andrew'
+ HP
+ University
+ 'org'
+ Year
+ 'key'
+ Pub
+ CA
+ Game
+ University
+ Sale
+ University
+ Ground
+ 'Koresh'
+ Caltech
+ FBI
+ Mark
+ Drugs
+ University
+ Make
+ Netcom
+ Drugs
+ Keith
+ NNTP
+ Posting
+ MR
+ Turkish
+ Good
+ 'make'
+ Jim
+ 'de'
diff --git a/transformer-tests/etm/20ng/intr_p3_v6_gpt3.txt b/transformer-tests/etm/20ng/intr_p3_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..e74dfde91060ef93512b310f0c8c1a5867fa59ff
--- /dev/null
+++ b/transformer-tests/etm/20ng/intr_p3_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ 'tm'
+ 'dod'
+ 'back'
+ 'hp'
+ President
+ 'nntp'
+ nntp
+ 'org'
+ 'SCSI'
+ 'make'
+ 'medical'
+ 'wire'
+ 'university'
+ 'graphics'
+ 'university'
+ 'Columbia'
+ 'car'
+ 'ca'
+ Turkish
+ 'de'
+ 'HP'
+ 'ca'
+ 'org'
+ 'game'
+ 'Intruder' term: 'read'
+ FTP
+ 'nntp'
+ 'game'
+ 'intruder'
+ Sale
+ NNTP
+ 'power'
+ 'Koresh'
+ 'caltech'
+ 'fbi'
+ 'phone'
+ 'drugs'
+ 'article'
+ 'make'
+ 'netcom'
+ 'drugs'
+ 'Keith'
+ 'nntp'
+ Intruder term: "people"
+ 'mr'
+ 'turkish'
+ 'gov'
+ 'problem'
+ 'Jim'
+ 'de'
diff --git a/transformer-tests/etm/20ng/intr_p4_v6_gpt3.txt b/transformer-tests/etm/20ng/intr_p4_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..0eb8cb637bdf356e57cebdf44db65700669a93b0
--- /dev/null
+++ b/transformer-tests/etm/20ng/intr_p4_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Virginia
+ Guns
+ Good
+ HP
+ President
+ San
+ CA
+ IBM
+ Video
+ Make
+ Article
+ Wire
+ University
+ Graphics
+ University
+ Pitt
+ Car
+ University
+ Armenian
+ Andrew
+ HP
+ ca
+ org
+ Cleveland
+ Read
+ Pub
+ CA
+ Game
+ 'people'
+ Sale
+ University
+ Ground
+ Fire
+ Caltech
+ FBI
+ Mark
+ Drugs
+ University
+ Good
+ Netcom
+ Drugs
+ Keith
+ NNTP
+ Make
+ Mr
+ Turkish
+ Good
+ Make
+ Jim
+ 'de'
diff --git a/transformer-tests/etm/20ng/intr_p5_v6_gpt3.txt b/transformer-tests/etm/20ng/intr_p5_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..c24720e284ad169048e81018edf14862717ca4e5
--- /dev/null
+++ b/transformer-tests/etm/20ng/intr_p5_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Virginia
+ Dod
+ Good
+ HP
+ President
+ San
+ ca
+ IBM
+ Video
+ Make
+ Article
+ Wire
+ University
+ Graphics
+ University
+ Pitt
+ Car
+ University
+ Armenian
+ Andrew
+ HP
+ ca
+ org
+ Year
+ Read
+ Pub
+ CA
+ Game
+ People
+ Sale
+ University
+ Ground
+ Fire
+ Caltech
+ FBI
+ Newsreader
+ Drugs
+ University
+ Good
+ Netcom
+ Drugs
+ Keith
+ NNTP
+ Make
+ Mr
+ Turkish
+ Good
+ Make
+ Jim
+ 'de'
diff --git a/transformer-tests/etm/20ng/intr_p6_v6_gpt3.txt b/transformer-tests/etm/20ng/intr_p6_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..181e0903d95ed43d95dfde24637104bbcfd5916d
--- /dev/null
+++ b/transformer-tests/etm/20ng/intr_p6_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ 'tm'
+ 'dod'
+ 'good'
+ HP
+ President
+ 'san'
+ nntp
+ 'ibm'
+ 'scsi'
+ 'make'
+ 'food'
+ 'wire'
+ 'university'
+ 'time'
+ University
+ 'pitt'
+ Car
+ 'posting'
+ 'israel'
+ 'de'
+ HP
+ 'ca'
+ 'org'
+ 'year'
+ 'key'
+ 'pub'
+ CA
+ 'game'
+ 'people'
+ Sale
+ 'ca'
+ 'ground'
+ 'Koresh'
+ Caltech
+ FBI
+ 'phone'
+ Drugs
+ 'Posting'
+ 'writes'
+ 'netcom'
+ 'drugs'
+ Keith
+ nntp
+ 'time'
+ Mr
+ Turkish
+ 'gov'
+ 'writes'
+ Jim
+ 'de'
diff --git a/transformer-tests/etm/20ng/intruders.txt b/transformer-tests/etm/20ng/intruders.txt
new file mode 100644
index 0000000000000000000000000000000000000000..55778b4a42532a469f8ce02798420b09d87506dd
--- /dev/null
+++ b/transformer-tests/etm/20ng/intruders.txt
@@ -0,0 +1,50 @@
+{'topic_id': 0, 'intruder_id': 40, 'intruder_term': 'tm', 'topic_terms': ['writes', 'article', 'virginia', 'news', 'stanford']}
+{'topic_id': 1, 'intruder_id': 49, 'intruder_term': 'guns', 'topic_terms': ['car', 'bike', 'dod', 'cars', 'front']}
+{'topic_id': 2, 'intruder_id': 46, 'intruder_term': 'space', 'topic_terms': ['time', 'problem', 'back', 'work', 'good']}
+{'topic_id': 3, 'intruder_id': 30, 'intruder_term': 'posting', 'topic_terms': ['mail', 'hp', 'mark', 'version', 'fax']}
+{'topic_id': 4, 'intruder_id': 44, 'intruder_term': 'president', 'topic_terms': ['people', 'make', 'point', 'post', 'group']}
+{'topic_id': 5, 'intruder_id': 11, 'intruder_term': 'san', 'topic_terms': ['university', 'posting', 'host', 'nntp', 'writes']}
+{'topic_id': 6, 'intruder_id': 25, 'intruder_term': 'pub', 'topic_terms': ['university', 'posting', 'host', 'nntp', 'ca']}
+{'topic_id': 7, 'intruder_id': 17, 'intruder_term': 'article', 'topic_terms': ['access', 'ibm', 'au', 'org', 'net']}
+{'topic_id': 8, 'intruder_id': 39, 'intruder_term': 'good', 'topic_terms': ['drive', 'card', 'scsi', 'disk', 'video']}
+{'topic_id': 9, 'intruder_id': 43, 'intruder_term': 'make', 'topic_terms': ['file', 'email', 'information', 'faq', 'output']}
+{'topic_id': 10, 'intruder_id': 0, 'intruder_term': 'article', 'topic_terms': ['health', 'medical', 'science', 'food', 'water']}
+{'topic_id': 11, 'intruder_id': 31, 'intruder_term': 'wire', 'topic_terms': ['san', 'california', 'university', 'york', 'american']}
+{'topic_id': 12, 'intruder_id': 14, 'intruder_term': 'university', 'topic_terms': ['time', 'good', 'back', 'work', 'problem']}
+{'topic_id': 13, 'intruder_id': 12, 'intruder_term': 'time', 'topic_terms': ['book', 'graphics', 'books', 'points', 'line']}
+{'topic_id': 14, 'intruder_id': 9, 'intruder_term': 'faq', 'topic_terms': ['writes', 'article', 'posting', 'university', 'host']}
+{'topic_id': 15, 'intruder_id': 10, 'intruder_term': 'health', 'topic_terms': ['state', 'ohio', 'columbia', 'cleveland', 'pitt']}
+{'topic_id': 16, 'intruder_id': 1, 'intruder_term': 'car', 'topic_terms': ['mail', 'mark', 'hp', 'version', 'newsreader']}
+{'topic_id': 17, 'intruder_id': 21, 'intruder_term': 'ca', 'topic_terms': ['writes', 'article', 'university', 'posting', 'host']}
+{'topic_id': 18, 'intruder_id': 48, 'intruder_term': 'writes', 'topic_terms': ['israel', 'turkish', 'israeli', 'jews', 'armenian']}
+{'topic_id': 19, 'intruder_id': 42, 'intruder_term': 'window', 'topic_terms': ['uk', 'ac', 'de', 'uiuc', 'andrew']}
+{'topic_id': 20, 'intruder_id': 16, 'intruder_term': 'hp', 'topic_terms': ['mail', 'information', 'list', 'computer', 'software']}
+{'topic_id': 21, 'intruder_id': 8, 'intruder_term': 'card', 'topic_terms': ['posting', 'host', 'nntp', 'university', 'ca']}
+{'topic_id': 22, 'intruder_id': 7, 'intruder_term': 'org', 'topic_terms': ['key', 'chip', 'encryption', 'clipper', 'keys']}
+{'topic_id': 23, 'intruder_id': 15, 'intruder_term': 'cleveland', 'topic_terms': ['game', 'team', 'year', 'play', 'hockey']}
+{'topic_id': 24, 'intruder_id': 22, 'intruder_term': 'key', 'topic_terms': ['people', 'make', 'good', 'things', 'read']}
+{'topic_id': 25, 'intruder_id': 26, 'intruder_term': 'university', 'topic_terms': ['ftp', 'image', 'graphics', 'pub', 'software']}
+{'topic_id': 26, 'intruder_id': 13, 'intruder_term': 'graphics', 'topic_terms': ['posting', 'nntp', 'host', 'university', 'ca']}
+{'topic_id': 27, 'intruder_id': 23, 'intruder_term': 'game', 'topic_terms': ['time', 'back', 'good', 'problem', 'work']}
+{'topic_id': 28, 'intruder_id': 29, 'intruder_term': 'people', 'topic_terms': ['posting', 'university', 'host', 'nntp', 'writes']}
+{'topic_id': 29, 'intruder_id': 45, 'intruder_term': 'sale', 'topic_terms': ['god', 'jesus', 'people', 'christian', 'bible']}
+{'topic_id': 30, 'intruder_id': 47, 'intruder_term': 'people', 'topic_terms': ['host', 'posting', 'nntp', 'university', 'ca']}
+{'topic_id': 31, 'intruder_id': 4, 'intruder_term': 'make', 'topic_terms': ['ground', 'power', 'wire', 'hot', 'work']}
+{'topic_id': 32, 'intruder_id': 34, 'intruder_term': 'work', 'topic_terms': ['people', 'fbi', 'fire', 'children', 'koresh']}
+{'topic_id': 33, 'intruder_id': 41, 'intruder_term': 'caltech', 'topic_terms': ['time', 'good', 'back', 'problem', 'long']}
+{'topic_id': 34, 'intruder_id': 32, 'intruder_term': 'fbi', 'topic_terms': ['time', 'back', 'problem', 'work', 'good']}
+{'topic_id': 35, 'intruder_id': 24, 'intruder_term': 'things', 'topic_terms': ['hp', 'mail', 'phone', 'mark', 'newsreader']}
+{'topic_id': 36, 'intruder_id': 37, 'intruder_term': 'university', 'topic_terms': ['people', 'make', 'white', 'drugs', 'black']}
+{'topic_id': 37, 'intruder_id': 2, 'intruder_term': 'back', 'topic_terms': ['writes', 'article', 'university', 'distribution', 'posting']}
+{'topic_id': 38, 'intruder_id': 3, 'intruder_term': 'mark', 'topic_terms': ['writes', 'article', 'good', 'make', 'read']}
+{'topic_id': 39, 'intruder_id': 28, 'intruder_term': 'posting', 'topic_terms': ['good', 'time', 'netcom', 'writes', 'article']}
+{'topic_id': 40, 'intruder_id': 36, 'intruder_term': 'drugs', 'topic_terms': ['max', 'ah', 'tm', 'mr', 'air']}
+{'topic_id': 41, 'intruder_id': 27, 'intruder_term': 'problem', 'topic_terms': ['people', 'morality', 'system', 'caltech', 'keith']}
+{'topic_id': 42, 'intruder_id': 5, 'intruder_term': 'nntp', 'topic_terms': ['windows', 'window', 'file', 'files', 'server']}
+{'topic_id': 43, 'intruder_id': 6, 'intruder_term': 'posting', 'topic_terms': ['people', 'make', 'time', 'things', 'good']}
+{'topic_id': 44, 'intruder_id': 35, 'intruder_term': 'mail', 'topic_terms': ['president', 'mr', 'clinton', 'people', 'tax']}
+{'topic_id': 45, 'intruder_id': 18, 'intruder_term': 'turkish', 'topic_terms': ['sale', 'price', 'shipping', 'offer', 'printer']}
+{'topic_id': 46, 'intruder_id': 38, 'intruder_term': 'good', 'topic_terms': ['space', 'nasa', 'gov', 'earth', 'moon']}
+{'topic_id': 47, 'intruder_id': 33, 'intruder_term': 'problem', 'topic_terms': ['good', 'writes', 'article', 'people', 'make']}
+{'topic_id': 48, 'intruder_id': 20, 'intruder_term': 'computer', 'topic_terms': ['article', 'writes', 'university', 'news', 'jim']}
+{'topic_id': 49, 'intruder_id': 19, 'intruder_term': 'de', 'topic_terms': ['gun', 'people', 'guns', 'law', 'government']}
diff --git a/transformer-tests/etm/wiki/intr_p1_v6_gpt3.txt b/transformer-tests/etm/wiki/intr_p1_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..fc1ccec144e3c3f706a962501c8a02b0fd612349
--- /dev/null
+++ b/transformer-tests/etm/wiki/intr_p1_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Hitler
+ Electoral
+ Trespassing
+ Heinrich Himmler
+ Amends
+ Halsey
+ Thessaloniki
+ Laurent
+ Hilary
+ Diver
+ Greenery
+ Sequels
+ Arthropod
+ Kabul
+ Sideline
+ Idle
+ Backwater
+ Bobcat
+ Cory Monteith
+ Cory Monteith
+état
+ Adoptive
+ Administratively
+ 8th
+ Erik
+ Guangdong
+ Evelyn Waugh
+ Ionization
+ Guangdong
+ Glamorous
+ Thessaloniki
+ Unbelievable
+ Convection
+ Kevin McHale
+ werewolf
+ Metatarsal
+ 'solves'
+ Cupola
+ Nameplate
+ Brick
+ Administratively
+ 'Imagines'
+ Sierra Leone
+ Heinrich Himmler
+ Apis
+ Cory Monteith
+ Woolwich
+ Clark Gable
+ Hitler
+ Russian_Federation
diff --git a/transformer-tests/etm/wiki/intr_p2_v6_gpt3.txt b/transformer-tests/etm/wiki/intr_p2_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..5f6f10ebd18f466767bdb4d04b596ad93911aee0
--- /dev/null
+++ b/transformer-tests/etm/wiki/intr_p2_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Hitler
+ 'hulk'
+ 'Evelyn_Waugh'
+ Heinrich Himmler
+ Amends
+ Halsey
+ Thessaloniki
+ Laurent
+ Hilary
+ Heinrich Himmler
+ Otis
+ Archibald
+ 'hans_bethe'
+ Kabul
+ Sideline
+ Idle
+ Timeout
+ Bobcat
+ Cory Monteith
+ Cory Monteith
+état
+ Adoptive
+ 'crooked'
+ 8th
+ Erik
+ Genealogy
+ Evelyn Waugh
+ Analogy
+ 'Guangdong'
+ Missing
+ Thessaloniki
+ 'unbelievable'
+ 'goggles'
+ Kevin McHale
+ 'werewolf'
+ Metatarsal
+ 'solves'
+ Vindictive
+ Nameplate
+ Brick
+ administratively
+ 'imagines'
+ Cory Monteith
+ 'unbelievable'
+ Councilors
+ Cory Monteith
+ Heal
+ Clark Gable
+ Hitler
+ Russian_Federation
diff --git a/transformer-tests/etm/wiki/intr_p3_v6_gpt3.txt b/transformer-tests/etm/wiki/intr_p3_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..436f80115b7baa805613e05178804422fbd10101
--- /dev/null
+++ b/transformer-tests/etm/wiki/intr_p3_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Hitler
+ 'hulk'
+ 'trespassing'
+ Heinrich Himmler
+ 'amends'
+ 'air_corps'
+ 'otis'
+ 'laurent'
+ st_augustine
+ 'heinrich_himmler'
+ 'otis'
+ 'Archibald'
+ 'arthropod'
+ Kabul
+ 'incubating'
+ 'unbelievable'
+ 'timeout'
+ Bobcat
+ cory_monteith
+ 'cory_monteith'
+ 'jagdgeschwader'
+ 'evan_sawdey'
+ 'crooked'
+ '8th'
+ hot_dance_club_songs
+ 'guangdong'
+ 'evelyn_waugh'
+ Heinrich Himmler
+ 'guangdong'
+ 'missing'
+ 'Thessaloniki'
+ 'andy_warhol'
+ 'goggles'
+ kevin_mchale
+ 'werewolf'
+ 'knuckles'
+ 'solves'
+ Vindictive
+ 'conflict'
+ 'brick'
+ 'kpa'
+ 'restructuring'
+ 'cory_monteith'
+ 'clark_gable', 'heinrich_himmler'
+ 'apis'
+ 'cory_monteith'
+ air_corps
+ heinrich_himmler
+ Hitler
+ 'russian_federation'
diff --git a/transformer-tests/etm/wiki/intr_p4_v6_gpt3.txt b/transformer-tests/etm/wiki/intr_p4_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..fdc487ca44d9edfbb9a4cfb2bbdad0c87b19ed28
--- /dev/null
+++ b/transformer-tests/etm/wiki/intr_p4_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Hitler
+ Hulk
+ Evelyn_Waugh
+ Heinrich Himmler
+ Bar
+ Halsey
+ Thessaloniki
+ Laurent
+ Hilary
+ Heinrich_himmler
+ Otis
+ Archibald
+ Arthropod
+ Kabul
+ Keels
+ Unbelievable
+ Spaceship
+ Bobcat
+ cory_monteith
+ Cory_Monteith
+état
+ Evan_sawdey
+ 'crooked'
+ 8th
+ Erik
+ 8th
+ Evelyn_Waugh
+ Heinrich_Himmler
+ Guangdong
+ Clark_Gable
+ Thessaloniki
+ 'unbelievable'
+ Goggles
+ Kevin_mchale
+ werewolf
+ Metatarsal
+ 'solves'
+ Vindictive
+ Paving
+ Brick
+ Hitler
+ Kick
+ Cory_Monteith
+ Clark_Gable
+ Apis
+ Cory_Monteith
+ Heal
+ Heinrich Himmler
+ Hitler
+ Russian_federation
diff --git a/transformer-tests/etm/wiki/intr_p5_v6_gpt3.txt b/transformer-tests/etm/wiki/intr_p5_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..4e292d19189a485dec29015b9bf337875f71e0b9
--- /dev/null
+++ b/transformer-tests/etm/wiki/intr_p5_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Hitler
+ Hulk
+ Evelyn_waugh
+ Heinrich Himmler
+ Bar
+ Halsey
+ Thessaloniki
+ Laurent
+ Hilary
+ Heinrich_himmler
+ Otis
+ Archibald
+ Arthropod
+ Kabul
+ Keels
+ Unbelievable
+ Spaceship
+ Bobcat
+ cory_monteith
+ Cory_Monteith
+état
+ Evan_sawdey
+ 'comment'
+ 8th
+ Erik
+ 8th
+ Evelyn_waugh
+ Heinrich_Himmler
+ Guangdong
+ clark_gable
+ Thessaloniki
+ Andy_Warhol
+ Goggles
+ kevin_mchale
+ werewolf
+ Metatarsal
+ 'solves'
+ Vindictive
+ Paving
+ Brick
+ Hitler
+ Kick
+ Cory_Monteith
+ clark_gable
+ Apis
+ Cory_Monteith
+ Heal
+ heinrich_himmler
+ Hitler
+ Russian_federation
diff --git a/transformer-tests/etm/wiki/intr_p6_v6_gpt3.txt b/transformer-tests/etm/wiki/intr_p6_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..699dc2c284780336ff6ae02c9206bf03813ea461
--- /dev/null
+++ b/transformer-tests/etm/wiki/intr_p6_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Hitler
+ Hulk
+ 'trespassing'
+ Heinrich Himmler
+ 'bar'
+ Halsey
+ Otis
+ 'laurent'
+ st_augustine
+ Heinrich Himmler
+ Otis
+ Bobcat
+ Arthropod
+ Kabul
+ 'sideline'
+ 'unbelievable'
+ 'spaceship'
+ Bobcat
+ cory_monteith
+ Cory_Monteith
+ Jagdgeschwader
+ 'laid'
+ 'crooked'
+ 8th
+ 'erik'
+ 'Guangdong'
+ Evelyn_waugh
+ Heinrich Himmler
+ 'Guangdong'
+ 'clark_gable'
+ Thessaloniki
+ Andy_Warhol
+ Goggles
+ kevin_mchale
+ 'werewolf'
+ 'knuckles'
+ 'solves'
+ Vindictive
+ Conflict
+ Brick
+ Hitler
+ 'kick'
+ 'cory_monteith'
+ clark_gable
+ 'apis'
+ Cory_Monteith
+ air_corps
+ heinrich_himmler
+ Hitler
+ Russian_Federation
diff --git a/transformer-tests/etm/wiki/intruders.txt b/transformer-tests/etm/wiki/intruders.txt
new file mode 100644
index 0000000000000000000000000000000000000000..866796e380a8e2dd0573bd183ab5b575e20c799e
--- /dev/null
+++ b/transformer-tests/etm/wiki/intruders.txt
@@ -0,0 +1,50 @@
+{'topic_id': 0, 'intruder_id': 40, 'intruder_term': 'hitler', 'topic_terms': ['prohibit', 'grizzly', 'finger', 'sideline', 'accruing']}
+{'topic_id': 1, 'intruder_id': 49, 'intruder_term': 'stillness', 'topic_terms': ['backwater', 'hulk', 'synonymy', 'otis', 'electoral']}
+{'topic_id': 2, 'intruder_id': 46, 'intruder_term': 'school_library_journal', 'topic_terms': ['sashes', 'trespassing', 'marvels', 'evelyn_waugh', 'nameplate']}
+{'topic_id': 3, 'intruder_id': 30, 'intruder_term': 'multiracial', 'topic_terms': ['panama_city', 'kpa', 'purify', 'heinrich_himmler', 'nameplate']}
+{'topic_id': 4, 'intruder_id': 44, 'intruder_term': 'recognisable', 'topic_terms': ['amends', 'trill', 'bar', 'storehouses', 'plodding']}
+{'topic_id': 5, 'intruder_id': 11, 'intruder_term': 'air_corps', 'topic_terms': ['councilors', 'terracotta', 'halsey', 'analogy', 'largely']}
+{'topic_id': 6, 'intruder_id': 25, 'intruder_term': 'impeached', 'topic_terms': ['backwater', 'otis', 'thessaloniki', 'bristle', 'personality']}
+{'topic_id': 7, 'intruder_id': 17, 'intruder_term': 'adoptive', 'topic_terms': ['laurent', 'soloing', 'morals', 'scrutinised', 'looted']}
+{'topic_id': 8, 'intruder_id': 39, 'intruder_term': 'apologizes', 'topic_terms': ['largely', 'bristle', 'purports', 'hilary', 'st_augustine']}
+{'topic_id': 9, 'intruder_id': 43, 'intruder_term': 'heinrich_himmler', 'topic_terms': ['institutionalized', 'departs', 'heal', 'voluminous', 'diver']}
+{'topic_id': 10, 'intruder_id': 0, 'intruder_term': 'grizzly', 'topic_terms': ['misinterpretation', 'greenery', 'kick', 'national_aeronautics_and_space_administration', 'otis']}
+{'topic_id': 11, 'intruder_id': 31, 'intruder_term': 'partisans', 'topic_terms': ['air_corps', 'tends', 'archibald', 'bobcat', 'sequels']}
+{'topic_id': 12, 'intruder_id': 14, 'intruder_term': 'depositing', 'topic_terms': ['regularly', 'tact', 'hans_bethe', 'arthropod', 'locals']}
+{'topic_id': 13, 'intruder_id': 12, 'intruder_term': 'regularly', 'topic_terms': ['imagines', 'kabul', 'tends', 'lucid', 'accelerated']}
+{'topic_id': 14, 'intruder_id': 9, 'intruder_term': 'voluminous', 'topic_terms': ['incubating', 'depositing', 'sideline', 'accruing', 'keels']}
+{'topic_id': 15, 'intruder_id': 10, 'intruder_term': 'misinterpretation', 'topic_terms': ['unbelievable', 'permeable', 'idle', 'accelerated', 'novices']}
+{'topic_id': 16, 'intruder_id': 1, 'intruder_term': 'backwater', 'topic_terms': ['zeppelin', 'intermingled', 'marvels', 'timeout', 'spaceship']}
+{'topic_id': 17, 'intruder_id': 21, 'intruder_term': 'leisure', 'topic_terms': ['collected', 'imagines', 'bobcat', 'adoptive', 'exquisitely']}
+{'topic_id': 18, 'intruder_id': 48, 'intruder_term': 'councilors', 'topic_terms': ['artifact', 'syrup', 'russian_federation', 'cory_monteith', 'unscrupulous']}
+{'topic_id': 19, 'intruder_id': 42, 'intruder_term': 'implementing', 'topic_terms': ['bristle', 'cory_monteith', 'playability', 'sculptured', 'avoidance']}
+{'topic_id': 20, 'intruder_id': 16, 'intruder_term': 'marvels', 'topic_terms': ['accelerated', 'jagdgeschwader', 'unbelievable', 'état', 'heal']}
+{'topic_id': 21, 'intruder_id': 8, 'intruder_term': 'bristle', 'topic_terms': ['toronto_star', 'leisure', 'evan_sawdey', 'laid', 'adoptive']}
+{'topic_id': 22, 'intruder_id': 7, 'intruder_term': 'soloing', 'topic_terms': ['accelerated', 'crooked', 'comment', 'collected', 'administratively']}
+{'topic_id': 23, 'intruder_id': 26, 'intruder_term': 'marvels', 'topic_terms': ['impeached', '8th', 'cupola', 'genealogy', 'curzon']}
+{'topic_id': 24, 'intruder_id': 2, 'intruder_term': 'marvels', 'topic_terms': ['linked', 'hot_dance_club_songs', 'massey', 'vancouver_island', 'erik']}
+{'topic_id': 25, 'intruder_id': 28, 'intruder_term': 'guangdong', 'topic_terms': ['8th', 'navy', 'genealogy', 'impeached', 'sideline']}
+{'topic_id': 26, 'intruder_id': 13, 'intruder_term': 'tends', 'topic_terms': ['intermingled', 'prohibit', 'marvels', 'finger', 'evelyn_waugh']}
+{'topic_id': 27, 'intruder_id': 5, 'intruder_term': 'analogy', 'topic_terms': ['ionization', 'abounded', 'halsey', 'billeted', 'heinrich_himmler']}
+{'topic_id': 28, 'intruder_id': 24, 'intruder_term': 'linked', 'topic_terms': ['conflict', 'guangdong', 'situations', 'derogatory', 'locals']}
+{'topic_id': 29, 'intruder_id': 47, 'intruder_term': 'clark_gable', 'topic_terms': ['ushered', 'fracturing', 'missing', 'attestations', 'glamorous']}
+{'topic_id': 30, 'intruder_id': 6, 'intruder_term': 'thessaloniki', 'topic_terms': ['ito', 'multiracial', 'marvels', 'finger', 'nameplate']}
+{'topic_id': 31, 'intruder_id': 33, 'intruder_term': 'accruing', 'topic_terms': ['unbelievable', 'foley', 'partisans', 'andy_warhol', 'mausoleum']}
+{'topic_id': 32, 'intruder_id': 41, 'intruder_term': 'bolster', 'topic_terms': ['goggles', 'excised', 'rawness', 'quarrels', 'convection']}
+{'topic_id': 33, 'intruder_id': 20, 'intruder_term': 'état', 'topic_terms': ['metatarsal', 'intermingled', 'accruing', 'kevin_mchale', 'ribs']}
+{'topic_id': 34, 'intruder_id': 27, 'intruder_term': 'abounded', 'topic_terms': ['depositing', 'bleaching', 'councilors', 'werewolf', 'accruing']}
+{'topic_id': 35, 'intruder_id': 3, 'intruder_term': 'purify', 'topic_terms': ['accruing', 'intermingled', 'metatarsal', 'knuckles', 'ribs']}
+{'topic_id': 36, 'intruder_id': 4, 'intruder_term': 'bar', 'topic_terms': ['bristle', 'solves', 'bertha', 'conflict', 'situations']}
+{'topic_id': 37, 'intruder_id': 32, 'intruder_term': 'goggles', 'topic_terms': ['restructuring', 'cupola', 'testament', 'excised', 'vindictive']}
+{'topic_id': 38, 'intruder_id': 36, 'intruder_term': 'conflict', 'topic_terms': ['royalty', 'adoptive', 'paving', 'excised', 'nameplate']}
+{'topic_id': 39, 'intruder_id': 29, 'intruder_term': 'attestations', 'topic_terms': ['apologizes', 'activists', 'musical', 'brick', 'falaise']}
+{'topic_id': 40, 'intruder_id': 45, 'intruder_term': 'hawaiian_islands', 'topic_terms': ['kanye_west', 'hitler', 'nameplate', 'kpa', 'administratively']}
+{'topic_id': 41, 'intruder_id': 37, 'intruder_term': 'restructuring', 'topic_terms': ['edmonton_oilers', 'imagines', 'bolster', 'busiest', 'kick']}
+{'topic_id': 42, 'intruder_id': 19, 'intruder_term': 'cory_monteith', 'topic_terms': ['incubating', 'implementing', 'sierra_leone', 'keels', 'brandon']}
+{'topic_id': 43, 'intruder_id': 15, 'intruder_term': 'unbelievable', 'topic_terms': ['league', 'heinrich_himmler', 'italia', 'clark_gable', 'accelerated']}
+{'topic_id': 44, 'intruder_id': 34, 'intruder_term': 'councilors', 'topic_terms': ['recognisable', 'fluent', 'apis', 'translating', 'accelerated']}
+{'topic_id': 45, 'intruder_id': 38, 'intruder_term': 'royalty', 'topic_terms': ['hawaiian_islands', 'cory_monteith', 'ebony', '90th', 'paving']}
+{'topic_id': 46, 'intruder_id': 23, 'intruder_term': 'genealogy', 'topic_terms': ['school_library_journal', 'air_corps', 'paterson', 'woolwich', 'heal']}
+{'topic_id': 47, 'intruder_id': 18, 'intruder_term': 'russian_federation', 'topic_terms': ['genealogy', 'detested', 'clark_gable', 'noun', 'heinrich_himmler']}
+{'topic_id': 48, 'intruder_id': 35, 'intruder_term': 'accruing', 'topic_terms': ['impounded', 'councilors', 'inchon', 'croft', 'hitler']}
+{'topic_id': 49, 'intruder_id': 22, 'intruder_term': 'accelerated', 'topic_terms': ['kahne', 'crossroads', 'stillness', 'attestations', 'russian_federation']}
diff --git a/transformer-tests/mallet/wiki/intr_p1_v6_gpt3.txt b/transformer-tests/mallet/wiki/intr_p1_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..a09193c51bd86ca21b04c46277d00b3b24611c11
--- /dev/null
+++ b/transformer-tests/mallet/wiki/intr_p1_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Million
+ Season
+ Match
+ Cell
+ Episode
+ People
+ Battle
+ English
+ Army
+ Fruit
+ Films
+ Miles
+ German
+ Law
+ Music
+ Highway
+ Game
+ Nuclear
+ Australia
+ Spanish
+ Castle
+ Games
+ Japanese
+ Life
+ Risk
+ Second
+ Ships
+ People
+ Building
+ 'state'
+ Million
+ Series
+ Madonna
+ Ships
+ Bond
+ India
+ Long
+ American
+ Flag
+ Greek
+ War
+ Cup
+ White
+ India
+ Television
+ Bond
+ Oxford
+ Century
+ Planet
+ Century
diff --git a/transformer-tests/mallet/wiki/intr_p2_v6_gpt3.txt b/transformer-tests/mallet/wiki/intr_p2_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..7c2e744c31fa4e5b3dd688568fc020a713169966
--- /dev/null
+++ b/transformer-tests/mallet/wiki/intr_p2_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Million
+ Season
+ Match
+ Cell
+ Episode
+ Water
+ 'battle'
+ English
+ Army
+ Fruit
+ Films
+ United States
+ German
+ Law
+ Music
+ Highway
+ Game
+ Nuclear
+ Australia
+ Birds
+ Hurricane
+ Games
+ Japanese
+ Life
+ Risk
+ 'second'
+ Ships
+ Homer
+ Building
+ Million
+ Million
+ Virgin
+ Madonna
+ Ships
+ Bond
+ India
+ Long
+ American
+ Flag
+ Long
+ War
+ Cup
+ White
+ Found
+ Television
+ Bond
+ Oxford
+ Century
+ Planet
+ Century
diff --git a/transformer-tests/mallet/wiki/intr_p3_v6_gpt3.txt b/transformer-tests/mallet/wiki/intr_p3_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..d5841d744f73a0f8e7bbe0f3579bc653910a01f3
--- /dev/null
+++ b/transformer-tests/mallet/wiki/intr_p3_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ 'million'
+ 'game'
+ 'match'
+ 'cell'
+ 'episode'
+ 'police'
+ 'battle'
+ Intruder term: "English"
+ Army
+ 'video'
+ 'films'
+ 'united_states'
+ German
+ 'French'
+ 'war'
+ 'highway'
+ Hurricane
+ Scotland
+ Australia
+ 'spanish'
+ Castle
+ Games
+ War
+ 'life'
+ 'Risk'
+ 'second'
+ 'ships'
+ 'homer'
+ 'race'
+ 'million'
+ 'million'
+ 'washington'
+ Madonna
+ 'ships'
+ 'war'
+ Intruder term: India
+ Shark
+ 'American'
+ Flag
+ Emperor
+ 'war'
+ 'cup'
+ 'breeding'
+ 'fruit'
+ Intruder term: 'system'
+ Bond
+ 'event'
+ 'century'
+ 'planet'
+ 'century'
diff --git a/transformer-tests/mallet/wiki/intr_p4_v6_gpt3.txt b/transformer-tests/mallet/wiki/intr_p4_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..8611c31e797c3ca919091dd3cc105664d55bb520
--- /dev/null
+++ b/transformer-tests/mallet/wiki/intr_p4_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Million
+ Baseball
+ Match
+ Cell
+ Love
+ People
+ Battle
+ English
+ Army
+ Fruit
+ Films
+ United_States
+ German
+ Law
+ Music
+ Highway
+ Game
+ Scotland
+ Australia
+ Spanish
+ Hurricane
+ Games
+ Japanese
+ Life
+ Risk
+ Second
+ Ships
+ Homer
+ Building
+ Million
+ Million
+ Virginia
+ Madonna
+ Ships
+ Bond
+ India
+ Known
+ American
+ Flag
+ Greek
+ War
+ Cup
+ White
+ India
+ System
+ Bond
+ Oxford
+ Century
+ Planet
+ Century
diff --git a/transformer-tests/mallet/wiki/intr_p5_v6_gpt3.txt b/transformer-tests/mallet/wiki/intr_p5_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..28db77e732b23da4ca77023296ee555b8cf6a2d7
--- /dev/null
+++ b/transformer-tests/mallet/wiki/intr_p5_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ Million
+ Baseball
+ Match
+ Cell
+ Love
+ People
+ Cup
+ English
+ Army
+ Fruit
+ Films
+ united_states
+ German
+ Law
+ Music
+ Highway
+ Game
+ Scotland
+ Australia
+ Spanish
+ Hurricane
+ Games
+ Japanese
+ Life
+ Risk
+ Second
+ Ships
+ Homer
+ Race
+ Million
+ Million
+ Virginia
+ Madonna
+ Ships
+ Bond
+ India
+ Known
+ American
+ Flag
+ Greek
+ War
+ Cup
+ White
+ India
+ System
+ Bond
+ Oxford
+ Century
+ Planet
+ Century
diff --git a/transformer-tests/mallet/wiki/intr_p6_v6_gpt3.txt b/transformer-tests/mallet/wiki/intr_p6_v6_gpt3.txt
new file mode 100644
index 0000000000000000000000000000000000000000..36462b4b5e90b2941969012ef63595419457ee70
--- /dev/null
+++ b/transformer-tests/mallet/wiki/intr_p6_v6_gpt3.txt
@@ -0,0 +1,50 @@
+ 'million'
+ 'game'
+ Match
+ Cell
+ 'love'
+ 'police'
+ 'battle'
+ English
+ Army
+ Number
+ 'films'
+ united_states
+ German
+ 'law'
+ 'music'
+ 'highway'
+ 'game'
+ King
+ Australia
+ 'spanish'
+ Castle
+ Games
+ War
+ 'life'
+ Risk
+ 'second'
+ 'ships'
+ 'Homer'
+ Race
+ Million
+ 'million'
+ 'series'
+ Madonna
+ Ships
+ 'guns'
+ 'god'
+ 'shark'
+ 'american'
+ 'flag'
+ Emperor
+ War
+ cup
+ 'player'
+ 'India'
+ 'television'
+ Bond
+ Oxford
+ 'century'
+ Planet
+ 'released'
diff --git a/transformer-tests/mallet/wiki/intruders.txt b/transformer-tests/mallet/wiki/intruders.txt
new file mode 100644
index 0000000000000000000000000000000000000000..a3ec23fb9be92349874845b6f7a6652e646c5d19
--- /dev/null
+++ b/transformer-tests/mallet/wiki/intruders.txt
@@ -0,0 +1,50 @@
+{'topic_id': 0, 'intruder_id': 40, 'intruder_term': 'students', 'topic_terms': ['film', 'films', 'production', 'role', 'million']}
+{'topic_id': 1, 'intruder_id': 49, 'intruder_term': 'music', 'topic_terms': ['season', 'game', 'games', 'home', 'baseball']}
+{'topic_id': 2, 'intruder_id': 46, 'intruder_term': 'match', 'topic_terms': ['disease', 'people', 'risk', 'blood', 'treatment']}
+{'topic_id': 3, 'intruder_id': 30, 'intruder_term': 'cell', 'topic_terms': ['game', 'season', 'team', 'points', 'games']}
+{'topic_id': 4, 'intruder_id': 44, 'intruder_term': 'episode', 'topic_terms': ['character', 'story', 'said', 'like', 'love']}
+{'topic_id': 5, 'intruder_id': 11, 'intruder_term': 'water', 'topic_terms': ['said', 'police', 'people', 'later', 'found']}
+{'topic_id': 6, 'intruder_id': 25, 'intruder_term': 'battle', 'topic_terms': ['season', 'club', 'team', 'cup', 'league']}
+{'topic_id': 7, 'intruder_id': 17, 'intruder_term': 'english', 'topic_terms': ['music', 'musical', 'opera', 'works', 'composer']}
+{'topic_id': 8, 'intruder_id': 39, 'intruder_term': 'army', 'topic_terms': ['world', 'games', 'won', 'team', 'canada']}
+{'topic_id': 9, 'intruder_id': 43, 'intruder_term': 'fruit', 'topic_terms': ['song', 'album', 'number', 'music', 'video']}
+{'topic_id': 10, 'intruder_id': 0, 'intruder_term': 'films', 'topic_terms': ['route', 'highway', 'road', 'state', 'north']}
+{'topic_id': 11, 'intruder_id': 31, 'intruder_term': 'united_states', 'topic_terms': ['water', 'area', 'river', 'miles', 'park']}
+{'topic_id': 12, 'intruder_id': 14, 'intruder_term': 'german', 'topic_terms': ['law', 'court', 'act', 'case', 'state']}
+{'topic_id': 13, 'intruder_id': 12, 'intruder_term': 'law', 'topic_terms': ['ship', 'british', 'ships', 'french', 'captain']}
+{'topic_id': 14, 'intruder_id': 9, 'intruder_term': 'music', 'topic_terms': ['war', 'german', 'government', 'military', 'soviet']}
+{'topic_id': 15, 'intruder_id': 10, 'intruder_term': 'highway', 'topic_terms': ['game', 'player', 'games', 'released', 'series']}
+{'topic_id': 16, 'intruder_id': 1, 'intruder_term': 'game', 'topic_terms': ['storm', 'tropical', 'hurricane', 'winds', 'damage']}
+{'topic_id': 17, 'intruder_id': 21, 'intruder_term': 'nuclear', 'topic_terms': ['king', 'england', 'royal', 'english', 'scotland']}
+{'topic_id': 18, 'intruder_id': 48, 'intruder_term': 'australia', 'topic_terms': ['design', 'system', 'use', 'engine', 'power']}
+{'topic_id': 19, 'intruder_id': 42, 'intruder_term': 'birds', 'topic_terms': ['chinese', 'government', 'china', 'century', 'spanish']}
+{'topic_id': 20, 'intruder_id': 16, 'intruder_term': 'hurricane', 'topic_terms': ['century', 'built', 'castle', 'church', 'building']}
+{'topic_id': 21, 'intruder_id': 8, 'intruder_term': 'games', 'topic_terms': ['formula', 'nuclear', 'number', 'element', 'known']}
+{'topic_id': 22, 'intruder_id': 7, 'intruder_term': 'musical', 'topic_terms': ['aircraft', 'air', 'japanese', 'war', 'flight']}
+{'topic_id': 23, 'intruder_id': 26, 'intruder_term': 'life', 'topic_terms': ['station', 'line', 'railway', 'bridge', 'service']}
+{'topic_id': 24, 'intruder_id': 2, 'intruder_term': 'risk', 'topic_terms': ['building', 'park', 'city', 'new', 'construction']}
+{'topic_id': 25, 'intruder_id': 28, 'intruder_term': 'second', 'topic_terms': ['forces', 'battle', 'troops', 'division', 'attack']}
+{'topic_id': 26, 'intruder_id': 13, 'intruder_term': 'ships', 'topic_terms': ['later', 'years', 'life', 'family', 'time']}
+{'topic_id': 27, 'intruder_id': 5, 'intruder_term': 'people', 'topic_terms': ['episode', 'season', 'series', 'homer', 'television']}
+{'topic_id': 28, 'intruder_id': 24, 'intruder_term': 'building', 'topic_terms': ['race', 'second', 'stage', 'lap', 'team']}
+{'topic_id': 29, 'intruder_id': 47, 'intruder_term': 'state', 'topic_terms': ['million', 'company', 'new', 'year', 'market']}
+{'topic_id': 30, 'intruder_id': 29, 'intruder_term': 'million', 'topic_terms': ['cells', 'cell', 'dna', 'protein', 'proteins']}
+{'topic_id': 31, 'intruder_id': 27, 'intruder_term': 'series', 'topic_terms': ['american', 'war', 'united_states', 'washington', 'virginia']}
+{'topic_id': 32, 'intruder_id': 33, 'intruder_term': 'wrote', 'topic_terms': ['madonna', 'carey', 'oxford', 'race', 'cambridge']}
+{'topic_id': 33, 'intruder_id': 34, 'intruder_term': 'ships', 'topic_terms': ['book', 'work', 'published', 'wrote', 'writing']}
+{'topic_id': 34, 'intruder_id': 45, 'intruder_term': 'bond', 'topic_terms': ['ship', 'ships', 'guns', 'war', 'class']}
+{'topic_id': 35, 'intruder_id': 37, 'intruder_term': 'children', 'topic_terms': ['god', 'india', 'temple', 'church', 'christian']}
+{'topic_id': 36, 'intruder_id': 3, 'intruder_term': 'team', 'topic_terms': ['species', 'found', 'known', 'long', 'shark']}
+{'topic_id': 37, 'intruder_id': 4, 'intruder_term': 'story', 'topic_terms': ['women', 'children', 'black', 'american', 'people']}
+{'topic_id': 38, 'intruder_id': 23, 'intruder_term': 'station', 'topic_terms': ['art', 'painting', 'work', 'flag', 'century']}
+{'topic_id': 39, 'intruder_id': 36, 'intruder_term': 'long', 'topic_terms': ['army', 'emperor', 'city', 'battle', 'greek']}
+{'topic_id': 40, 'intruder_id': 22, 'intruder_term': 'war', 'topic_terms': ['city', 'school', 'students', 'university', 'schools']}
+{'topic_id': 41, 'intruder_id': 6, 'intruder_term': 'cup', 'topic_terms': ['earth', 'star', 'planet', 'sun', 'stars']}
+{'topic_id': 42, 'intruder_id': 15, 'intruder_term': 'player', 'topic_terms': ['species', 'birds', 'breeding', 'white', 'horses']}
+{'topic_id': 43, 'intruder_id': 35, 'intruder_term': 'india', 'topic_terms': ['species', 'fruit', 'cap', 'brown', 'found']}
+{'topic_id': 44, 'intruder_id': 18, 'intruder_term': 'system', 'topic_terms': ['episode', 'series', 'season', 'episodes', 'television']}
+{'topic_id': 45, 'intruder_id': 38, 'intruder_term': 'painting', 'topic_terms': ['series', 'book', 'story', 'bond', 'novel']}
+{'topic_id': 46, 'intruder_id': 32, 'intruder_term': 'oxford', 'topic_terms': ['match', 'championship', 'team', 'event', 'title']}
+{'topic_id': 47, 'intruder_id': 20, 'intruder_term': 'century', 'topic_terms': ['president', 'election', 'state', 'government', 'party']}
+{'topic_id': 48, 'intruder_id': 41, 'intruder_term': 'planet', 'topic_terms': ['australia', 'test', 'match', 'england', 'australian']}
+{'topic_id': 49, 'intruder_id': 19, 'intruder_term': 'century', 'topic_terms': ['album', 'band', 'music', 'songs', 'released']}