TOP ACTIVATIONS
MAX = 0.293

the ambulance he saw a paramed ic wearing a big coat
and wearing gloves . The paramed ic was very polite and
. The driver and the paramed ic jumped out . They
L ily asked the paramed ic , " Is the
to be okay ?" The paramed ic smiled and said ,
Billy and called him a liar . He said that he
was a thief and a liar . He said Tom had
be a bully and a liar . I want to be
nice . He was a liar and a thief . He
He said Jen was a liar . He said Jen was
looked at Lily with a stern face . Lily felt sorry
But her mother told her stern ly , " No more
?" Mom said in a stern but gentle voice .
holes !" Mom said , stern ly . Max
walked towards them with a stern face . "
Dad asked . He was stern and sad .
would call him a stupid liar . One day
?" she asked in a stern voice . " Why are
, " Follow your dreams ' Lu cy smiled
couch and gave them a stern look . "

MIN ACTIVATIONS
MIN = -0.332

, and even went to outer space ! It was the
to take a journey to outer space . He hopped into
izz ed off back into outer space . Earth was an
looked like it came from outer space . L
special star that comes from outer space ," she explained .
. Soon they were in outer space and having fun .
the stars and planets in outer space . So one day
that they were going to outer space . They watched out
what the rocket discovered in outer space . It was a
Emma who loved learning about outer space . Everyday , she
of something to eat . Along the way , she spotted
on her sc ooter . Along the way , she saw
this way and that . Along the way , the girl
a far off place . Along the way , she found
the mouse 's home . Along the way , they met
to explore the forest . Along the way they talked about
to go visit Slip . Along the way she met a
make his way home . Along the way , he met
angel and followed her . Along the way , the angel
outside and started walking . Along the way , she ate

INTERVAL 0.137 - 0.293
CONTAINS 2.738%

L ila and Ben are happy . They say ,
hurt its wing . Lily 's mom my picked up the
's friend , a little pig named Pink y , came
. Together they loaded the truck with the boxes and Pete
," Max said , feeling sorry . The dog

INTERVAL -0.019 - 0.137
CONTAINS 55.216%

new and had fun . They looked at each other and
laugh still rang in his ears as he vanished into the
they saw a sign that said , " Do not eat
wonderful surprise � � € � � a drawing from a
Lucy saw a big red ball . She wanted to play

INTERVAL -0.175 - -0.019
CONTAINS 40.478%

. We can restore it and make it look like new
family and garden heard her beautiful singing .
. The ghost was curious and decided to explore . As
for the third time . She wanted water . She cried
Once upon a time , there lived a hero . He

INTERVAL -0.332 - -0.175
CONTAINS 1.568%

and fun . Can I wear them to school ?"
that made new pills every day . Sal gave Joe a
He knew that as long as he had this door ,
dance . She dances every day in her room . She
stories . One day , she goes to the
Token the
Token ID262
Feature activation+1.35e-02

Pos logit contributions
 Along+0.093
 nights+0.091
 On+0.083
way+0.083
 whenever+0.082
Neg logit contributions
 paramed-0.097
 liar-0.095
boys-0.094
 wasteful-0.093
roc-0.091
Token ambulance
Token ID22536
Feature activation+1.41e-01

Pos logit contributions
 paramed+0.018
 liar+0.017
boys+0.017
 wasteful+0.017
 mixer+0.017
Neg logit contributions
 Along-0.025
 nights-0.024
way-0.023
 On-0.022
 drifted-0.022
Token he
Token ID339
Feature activation-8.86e-02

Pos logit contributions
 paramed+0.231
 liar+0.229
 wasteful+0.223
 mixer+0.222
 greedy+0.221
Neg logit contributions
 Along-0.212
 nights-0.208
way-0.194
 On-0.187
 whenever-0.185
Token saw
Token ID2497
Feature activation+4.15e-02

Pos logit contributions
 Along+0.169
 nights+0.166
way+0.157
 On+0.152
 drifted+0.151
Neg logit contributions
 paramed-0.110
 mixer-0.106
 liar-0.106
 mean-0.105
boys-0.104
Token a
Token ID257
Feature activation-2.19e-02

Pos logit contributions
 paramed+0.067
 liar+0.065
 wasteful+0.064
boys+0.063
 mixer+0.063
Neg logit contributions
 nights-0.064
 Along-0.063
way-0.059
 whenever-0.057
 days-0.057
Token paramed
Token ID34581
Feature activation+2.93e-01

Pos logit contributions
 day+0.028
 nights+0.028
 night+0.028
way+0.027
 journey+0.027
Neg logit contributions
roc-0.038
boys-0.038
 wasteful-0.037
 chefs-0.035
 bluff-0.035
Tokenic
Token ID291
Feature activation+4.91e-02

Pos logit contributions
 liar+0.535
 paramed+0.525
 mean+0.524
 mixer+0.519
 wasteful+0.518
Neg logit contributions
 nights-0.384
 Along-0.374
way-0.368
ingle-0.333
more-0.328
Token wearing
Token ID5762
Feature activation-1.17e-01

Pos logit contributions
 paramed+0.097
 liar+0.095
 mixer+0.095
boys+0.095
 wasteful+0.095
Neg logit contributions
 Along-0.056
 nights-0.055
way-0.051
 drifted-0.047
 On-0.046
Token a
Token ID257
Feature activation-2.24e-02

Pos logit contributions
 nights+0.175
 Along+0.174
way+0.160
 whenever+0.159
 days+0.157
Neg logit contributions
 paramed-0.188
roc-0.186
 liar-0.186
 wasteful-0.186
boys-0.185
Token big
Token ID1263
Feature activation-1.89e-04

Pos logit contributions
 day+0.028
 Along+0.028
 journey+0.027
 night+0.027
 lifetime+0.025
Neg logit contributions
roc-0.042
boys-0.042
 wasteful-0.039
 chefs-0.038
eper-0.038
Token coat
Token ID13209
Feature activation+1.17e-03

Pos logit contributions
 day+0.000
 journey+0.000
 night+0.000
 morning+0.000
 Along+0.000
Neg logit contributions
boys-0.000
roc-0.000
 wasteful-0.000
 chefs-0.000
 bluff-0.000
Token and
Token ID290
Feature activation-1.80e-02

Pos logit contributions
 paramed+0.277
 mixer+0.273
 mean+0.271
 liar+0.269
boys+0.268
Neg logit contributions
 Along-0.123
 nights-0.122
way-0.104
 drifted-0.101
 On-0.099
Token wearing
Token ID5762
Feature activation-1.19e-01

Pos logit contributions
 Along+0.031
 nights+0.030
way+0.029
 On+0.027
 drifted+0.027
Neg logit contributions
 paramed-0.026
 mixer-0.025
 liar-0.025
 mean-0.025
boys-0.024
Token gloves
Token ID18051
Feature activation+3.67e-02

Pos logit contributions
 Along+0.176
 nights+0.175
 whenever+0.160
 On+0.156
 days+0.156
Neg logit contributions
 liar-0.196
 paramed-0.195
 wasteful-0.194
roc-0.189
boys-0.187
Token.
Token ID13
Feature activation-2.05e-02

Pos logit contributions
 paramed+0.081
 liar+0.080
roc+0.079
 wasteful+0.079
 mixer+0.078
Neg logit contributions
 Along-0.035
 nights-0.033
 whenever-0.031
way-0.031
 On-0.028
Token The
Token ID383
Feature activation-7.29e-03

Pos logit contributions
 nights+0.029
 Along+0.026
way+0.026
 drifted+0.025
 days+0.025
Neg logit contributions
 mixer-0.035
 paramed-0.035
 greedy-0.033
 stern-0.033
 liar-0.033
Token paramed
Token ID34581
Feature activation+2.93e-01

Pos logit contributions
 nights+0.015
 Along+0.015
way+0.015
 days+0.014
 day+0.014
Neg logit contributions
boys-0.007
 bluff-0.007
 liar-0.006
roc-0.006
 chefs-0.006
Tokenic
Token ID291
Feature activation+4.77e-02

Pos logit contributions
 liar+0.532
 mean+0.524
 paramed+0.523
 mixer+0.519
boys+0.519
Neg logit contributions
 nights-0.382
 Along-0.370
way-0.364
more-0.326
ingle-0.325
Token was
Token ID373
Feature activation+2.78e-02

Pos logit contributions
boys+0.081
 bluff+0.080
 liar+0.079
 wasteful+0.078
 greedy+0.077
Neg logit contributions
 nights-0.068
way-0.065
 Along-0.065
 drifted-0.061
 days-0.060
Token very
Token ID845
Feature activation+8.00e-02

Pos logit contributions
roc+0.044
 liar+0.044
boys+0.043
 paramed+0.041
 mixer+0.041
Neg logit contributions
 nights-0.044
 Along-0.043
 whenever-0.040
 drifted-0.040
 night-0.040
Token polite
Token ID23507
Feature activation+9.89e-02

Pos logit contributions
boys+0.124
 paramed+0.122
 liar+0.121
 mixer+0.120
roc+0.117
Neg logit contributions
 Along-0.127
 nights-0.125
way-0.119
 On-0.115
 drifted-0.114
Token and
Token ID290
Feature activation-1.90e-02

Pos logit contributions
 paramed+0.166
 mixer+0.163
 liar+0.162
 mean+0.161
boys+0.159
Neg logit contributions
 Along-0.145
 nights-0.140
way-0.130
 drifted-0.126
 On-0.126
Token.
Token ID13
Feature activation-2.63e-02

Pos logit contributions
 Along+0.018
 nights+0.017
way+0.016
 drifted+0.015
 whenever+0.015
Neg logit contributions
 paramed-0.040
 mixer-0.040
 liar-0.039
 wasteful-0.039
roc-0.039
Token The
Token ID383
Feature activation-2.82e-04

Pos logit contributions
 Along+0.044
 nights+0.042
way+0.040
 On+0.040
 drifted+0.039
Neg logit contributions
 paramed-0.040
 mixer-0.039
 liar-0.038
boys-0.038
 wasteful-0.037
Token driver
Token ID4639
Feature activation+8.98e-02

Pos logit contributions
 nights+0.001
 Along+0.001
way+0.000
 days+0.000
 drifted+0.000
Neg logit contributions
roc-0.000
boys-0.000
 paramed-0.000
 liar-0.000
 wasteful-0.000
Token and
Token ID290
Feature activation-2.05e-02

Pos logit contributions
 paramed+0.134
 liar+0.131
 mixer+0.131
 wasteful+0.128
boys+0.127
Neg logit contributions
 Along-0.148
 nights-0.146
way-0.137
 drifted-0.134
 On-0.131
Token the
Token ID262
Feature activation+1.64e-02

Pos logit contributions
 nights+0.034
 drifted+0.034
 Along+0.034
 days+0.032
 night+0.031
Neg logit contributions
 paramed-0.030
roc-0.030
 mixer-0.029
 liar-0.029
boys-0.028
Token paramed
Token ID34581
Feature activation+2.92e-01

Pos logit contributions
roc+0.021
boys+0.021
 wasteful+0.021
 bluff+0.020
 liar+0.020
Neg logit contributions
 nights-0.030
 Along-0.029
way-0.027
 days-0.027
 night-0.027
Tokenic
Token ID291
Feature activation+4.28e-02

Pos logit contributions
 paramed+0.482
 liar+0.475
 mixer+0.473
 mean+0.468
boys+0.464
Neg logit contributions
 Along-0.428
 nights-0.426
way-0.399
 drifted-0.375
 On-0.371
Token jumped
Token ID11687
Feature activation-6.88e-02

Pos logit contributions
 paramed+0.063
 mixer+0.061
 liar+0.061
 mean+0.060
boys+0.059
Neg logit contributions
 Along-0.073
 nights-0.071
way-0.066
 On-0.065
 drifted-0.064
Token out
Token ID503
Feature activation-3.00e-02

Pos logit contributions
 Along+0.091
 nights+0.089
way+0.083
 On+0.079
 drifted+0.078
Neg logit contributions
 paramed-0.127
 liar-0.124
 mixer-0.123
 wasteful-0.122
boys-0.121
Token.
Token ID13
Feature activation-2.67e-02

Pos logit contributions
 Along+0.036
 nights+0.032
 lifetime+0.029
dark+0.029
way+0.028
Neg logit contributions
 mean-0.061
 mad-0.060
 stern-0.059
 phone-0.058
 paramed-0.058
Token They
Token ID1119
Feature activation+4.09e-02

Pos logit contributions
 Along+0.040
 nights+0.040
way+0.037
 drifted+0.035
more+0.034
Neg logit contributions
 paramed-0.044
 liar-0.043
 mixer-0.043
 mean-0.042
 wasteful-0.042
Token
Token ID198
Feature activation-2.94e-02

Pos logit contributions
 nights+0.042
 Along+0.039
 drifted+0.037
way+0.035
 days+0.035
Neg logit contributions
 paramed-0.052
 mixer-0.051
 mean-0.050
 stern-0.049
 greedy-0.048
TokenL
Token ID43
Feature activation-3.39e-02

Pos logit contributions
 Along+0.047
 nights+0.046
way+0.043
 drifted+0.042
 On+0.041
Neg logit contributions
 paramed-0.046
 mixer-0.045
 liar-0.044
 mean-0.044
 wasteful-0.044
Tokenily
Token ID813
Feature activation-5.08e-02

Pos logit contributions
 Along+0.069
 nights+0.068
 drifted+0.066
 horizon+0.065
 lifetime+0.064
Neg logit contributions
 wasteful-0.038
 mean-0.038
 chefs-0.038
 liar-0.038
 mad-0.037
Token asked
Token ID1965
Feature activation+1.58e-02

Pos logit contributions
 Along+0.059
 nights+0.054
 drifted+0.054
way+0.051
 On+0.049
Neg logit contributions
 bluff-0.104
 liar-0.103
 wasteful-0.100
 mixer-0.100
 paramed-0.099
Token the
Token ID262
Feature activation+9.65e-03

Pos logit contributions
 liar+0.031
 paramed+0.028
 bluff+0.028
 wasteful+0.028
 disrespectful+0.028
Neg logit contributions
 whenever-0.022
 nights-0.020
 Along-0.019
 days-0.019
 night-0.018
Token paramed
Token ID34581
Feature activation+2.91e-01

Pos logit contributions
roc+0.010
boys+0.010
 liar+0.009
 wasteful+0.009
 bluff+0.009
Neg logit contributions
 Along-0.020
 nights-0.019
way-0.018
 days-0.018
 day-0.018
Tokenic
Token ID291
Feature activation+4.29e-02

Pos logit contributions
 liar+0.516
 mean+0.512
 mixer+0.507
boys+0.502
 paramed+0.502
Neg logit contributions
 nights-0.378
way-0.372
 Along-0.361
ingle-0.339
more-0.332
Token,
Token ID11
Feature activation+7.24e-02

Pos logit contributions
boys+0.078
 bluff+0.078
 wasteful+0.076
roc+0.076
 chefs+0.075
Neg logit contributions
 nights-0.053
way-0.052
 Along-0.051
 days-0.048
 day-0.047
Token "
Token ID366
Feature activation+2.01e-01

Pos logit contributions
roc+0.101
boys+0.098
 liar+0.097
 paramed+0.090
 mixer+0.089
Neg logit contributions
 Along-0.136
 nights-0.124
 On-0.123
 whenever-0.123
 As-0.118
TokenIs
Token ID3792
Feature activation+2.39e-02

Pos logit contributions
 liar+0.307
roc+0.295
 wasteful+0.287
boys+0.286
 mean+0.286
Neg logit contributions
 Along-0.320
 nights-0.314
 whenever-0.294
more-0.291
way-0.290
Token the
Token ID262
Feature activation+9.58e-03

Pos logit contributions
 paramed+0.041
 liar+0.040
roc+0.039
 mixer+0.038
 bluff+0.038
Neg logit contributions
 nights-0.034
 Along-0.033
way-0.031
 whenever-0.030
 days-0.030
Token to
Token ID284
Feature activation+7.56e-02

Pos logit contributions
 Along+0.127
 nights+0.126
way+0.118
 drifted+0.114
 On+0.113
Neg logit contributions
 paramed-0.125
 mixer-0.122
 mean-0.122
boys-0.120
 liar-0.120
Token be
Token ID307
Feature activation+3.56e-02

Pos logit contributions
 paramed+0.108
 mixer+0.103
 mean+0.103
 liar+0.103
boys+0.102
Neg logit contributions
 Along-0.132
 nights-0.130
way-0.122
 drifted-0.117
 On-0.117
Token okay
Token ID8788
Feature activation+1.07e-01

Pos logit contributions
 mixer+0.048
 paramed+0.048
roc+0.048
 liar+0.047
boys+0.046
Neg logit contributions
 nights-0.063
 Along-0.060
 whenever-0.056
 drifted-0.056
 days-0.056
Token?"
Token ID1701
Feature activation+6.88e-02

Pos logit contributions
 liar+0.166
 paramed+0.162
 wasteful+0.161
roc+0.160
 bluff+0.156
Neg logit contributions
 Along-0.166
 nights-0.162
 whenever-0.160
 On-0.155
way-0.151
Token The
Token ID383
Feature activation-1.17e-02

Pos logit contributions
 liar+0.128
 paramed+0.125
 wasteful+0.124
 mean+0.123
boys+0.123
Neg logit contributions
 Along-0.093
 nights-0.086
 On-0.082
way-0.081
 whenever-0.080
Token paramed
Token ID34581
Feature activation+2.90e-01

Pos logit contributions
 Along+0.021
 nights+0.021
 day+0.020
 days+0.020
way+0.020
Neg logit contributions
boys-0.014
 bluff-0.013
 chefs-0.013
roc-0.013
 liar-0.013
Tokenic
Token ID291
Feature activation+4.20e-02

Pos logit contributions
 liar+0.537
boys+0.530
 mean+0.529
 mixer+0.520
 paramed+0.517
Neg logit contributions
 nights-0.352
way-0.352
 Along-0.335
ingle-0.321
more-0.306
Token smiled
Token ID13541
Feature activation+1.23e-02

Pos logit contributions
 bluff+0.064
boys+0.064
 wasteful+0.061
 chefs+0.061
 liar+0.060
Neg logit contributions
way-0.066
 nights-0.065
 Along-0.062
 days-0.060
 drifted-0.060
Token and
Token ID290
Feature activation-2.63e-02

Pos logit contributions
 paramed+0.022
 mixer+0.021
 liar+0.021
 mean+0.021
boys+0.021
Neg logit contributions
 Along-0.017
 nights-0.017
way-0.015
 On-0.015
 drifted-0.015
Token said
Token ID531
Feature activation+5.14e-02

Pos logit contributions
 nights+0.043
 drifted+0.042
 Along+0.042
way+0.039
 On+0.039
Neg logit contributions
 liar-0.040
roc-0.039
 mixer-0.038
 paramed-0.037
boys-0.037
Token,
Token ID11
Feature activation+7.19e-02

Pos logit contributions
roc+0.115
boys+0.111
 app+0.110
 wasteful+0.110
 ram+0.108
Neg logit contributions
 whenever-0.050
 days-0.041
 nights-0.040
way-0.040
 Along-0.039
Token Billy
Token ID15890
Feature activation+3.56e-02

Pos logit contributions
 paramed+0.013
 mixer+0.013
 liar+0.013
 mean+0.013
boys+0.013
Neg logit contributions
 Along-0.019
 nights-0.019
way-0.018
 On-0.017
 drifted-0.017
Token and
Token ID290
Feature activation-9.36e-03

Pos logit contributions
 paramed+0.065
 liar+0.064
 mixer+0.064
 wasteful+0.063
 mean+0.062
Neg logit contributions
 Along-0.047
 nights-0.045
way-0.042
 On-0.041
 whenever-0.040
Token called
Token ID1444
Feature activation-5.95e-02

Pos logit contributions
 drifted+0.013
 nights+0.013
 Along+0.013
 whenever+0.013
way+0.013
Neg logit contributions
roc-0.015
 mixer-0.014
 paramed-0.014
 chefs-0.014
 bluff-0.014
Token him
Token ID683
Feature activation+5.07e-02

Pos logit contributions
 Along+0.068
 whenever+0.062
way+0.061
 On+0.060
 nights+0.059
Neg logit contributions
roc-0.118
boys-0.116
 wasteful-0.116
 paramed-0.115
 chefs-0.114
Token a
Token ID257
Feature activation-1.91e-02

Pos logit contributions
 paramed+0.110
 mixer+0.108
 wasteful+0.107
roc+0.106
 liar+0.106
Neg logit contributions
 Along-0.051
 whenever-0.044
way-0.044
 nights-0.043
 On-0.042
Token liar
Token ID31866
Feature activation+2.71e-01

Pos logit contributions
 Along+0.029
way+0.028
 day+0.028
 journey+0.027
 night+0.027
Neg logit contributions
boys-0.031
roc-0.028
 chefs-0.028
 wasteful-0.027
 bluff-0.026
Token.
Token ID13
Feature activation-1.61e-02

Pos logit contributions
 paramed+0.614
roc+0.608
 wasteful+0.607
 mixer+0.605
 bluff+0.597
Neg logit contributions
 Along-0.223
 nights-0.211
way-0.203
more-0.184
 whenever-0.183
Token He
Token ID679
Feature activation+1.01e-02

Pos logit contributions
 Along+0.021
 On+0.020
 As+0.018
way+0.017
 Everywhere+0.017
Neg logit contributions
 liar-0.032
boys-0.031
 chefs-0.031
 wasteful-0.031
 paramed-0.030
Token said
Token ID531
Feature activation+6.77e-02

Pos logit contributions
 bluff+0.016
 mixer+0.016
 paramed+0.016
roc+0.016
 chefs+0.016
Neg logit contributions
 Along-0.015
 drifted-0.014
way-0.014
 nights-0.013
 On-0.013
Token that
Token ID326
Feature activation-2.88e-02

Pos logit contributions
roc+0.159
 app+0.158
boys+0.157
 bluff+0.156
 chefs+0.151
Neg logit contributions
 of-0.048
 whenever-0.047
 he-0.041
way-0.040
 days-0.039
Token he
Token ID339
Feature activation-8.22e-02

Pos logit contributions
 nights+0.026
 whenever+0.024
 Along+0.023
 day+0.023
way+0.023
Neg logit contributions
 paramed-0.064
roc-0.062
boys-0.061
 liar-0.061
 mixer-0.061
Token was
Token ID373
Feature activation+4.18e-02

Pos logit contributions
 paramed+0.227
 mixer+0.220
 liar+0.214
 bluff+0.214
roc+0.213
Neg logit contributions
 Along-0.143
 nights-0.140
way-0.133
 drifted-0.130
 whenever-0.129
Token a
Token ID257
Feature activation-1.47e-02

Pos logit contributions
 paramed+0.066
 mixer+0.065
roc+0.064
boys+0.057
 argue+0.057
Neg logit contributions
 before-0.063
 Along-0.059
 nights-0.059
 night-0.058
 day-0.058
Token thief
Token ID25906
Feature activation+1.22e-01

Pos logit contributions
 day+0.023
 night+0.023
 Along+0.022
 nights+0.022
 days+0.021
Neg logit contributions
boys-0.022
roc-0.022
ese-0.020
oa-0.019
 wasteful-0.019
Token and
Token ID290
Feature activation-7.49e-03

Pos logit contributions
 paramed+0.244
 mixer+0.237
 wasteful+0.235
roc+0.234
 liar+0.233
Neg logit contributions
 Along-0.137
 nights-0.137
way-0.124
 whenever-0.120
 days-0.115
Token a
Token ID257
Feature activation-1.55e-02

Pos logit contributions
 Along+0.009
 nights+0.009
way+0.008
 drifted+0.008
 On+0.008
Neg logit contributions
 paramed-0.014
 mixer-0.014
boys-0.013
 liar-0.013
roc-0.013
Token liar
Token ID31866
Feature activation+2.67e-01

Pos logit contributions
 day+0.027
 nights+0.027
 Along+0.027
 night+0.026
 days+0.025
Neg logit contributions
boys-0.020
roc-0.020
 bluff-0.017
ese-0.017
oa-0.017
Token.
Token ID13
Feature activation-1.03e-02

Pos logit contributions
 paramed+0.553
 mixer+0.544
 wasteful+0.535
 liar+0.534
roc+0.532
Neg logit contributions
 Along-0.281
 nights-0.280
way-0.251
 drifted-0.237
 whenever-0.233
Token He
Token ID679
Feature activation+1.54e-02

Pos logit contributions
 Along+0.010
 On+0.009
 nights+0.008
way+0.008
 whenever+0.007
Neg logit contributions
 paramed-0.023
 mixer-0.023
boys-0.023
 liar-0.022
 wasteful-0.022
Token said
Token ID531
Feature activation+6.82e-02

Pos logit contributions
 mixer+0.024
 bluff+0.023
 paramed+0.023
roc+0.023
 liar+0.022
Neg logit contributions
 Along-0.023
 nights-0.023
 drifted-0.022
way-0.022
 On-0.021
Token Tom
Token ID4186
Feature activation+1.17e-01

Pos logit contributions
 app+0.159
roc+0.159
 bluff+0.159
boys+0.155
 intervene+0.155
Neg logit contributions
 of-0.042
 days-0.042
 whenever-0.042
 nights-0.041
 he-0.035
Token had
Token ID550
Feature activation-5.27e-02

Pos logit contributions
 paramed+0.222
 mixer+0.216
 bluff+0.210
 liar+0.209
 wasteful+0.208
Neg logit contributions
 Along-0.147
 nights-0.144
way-0.135
 whenever-0.131
 drifted-0.131
Token be
Token ID307
Feature activation+5.10e-02

Pos logit contributions
 paramed+0.144
 mixer+0.141
 liar+0.138
 wasteful+0.138
 bluff+0.136
Neg logit contributions
 Along-0.115
 nights-0.112
way-0.110
 whenever-0.108
 night-0.104
Token a
Token ID257
Feature activation-1.53e-02

Pos logit contributions
 mixer+0.077
 paramed+0.076
 tripod+0.068
 printer+0.066
 liar+0.063
Neg logit contributions
 nights-0.080
 whenever-0.079
 day-0.076
 as-0.075
 days-0.075
Token bully
Token ID27410
Feature activation+2.10e-01

Pos logit contributions
 day+0.022
 night+0.022
 nights+0.020
way+0.020
 Along+0.020
Neg logit contributions
roc-0.024
 paramed-0.023
boys-0.023
 wasteful-0.023
 mixer-0.022
Token and
Token ID290
Feature activation-5.83e-03

Pos logit contributions
 paramed+0.472
 mixer+0.452
roc+0.446
 bluff+0.444
 wasteful+0.443
Neg logit contributions
 nights-0.187
 whenever-0.183
 Along-0.171
way-0.159
 night-0.159
Token a
Token ID257
Feature activation-1.55e-02

Pos logit contributions
 Along+0.009
 nights+0.009
 whenever+0.009
 night+0.009
 drifted+0.008
Neg logit contributions
 paramed-0.009
 mixer-0.008
roc-0.008
boys-0.008
 bluff-0.008
Token liar
Token ID31866
Feature activation+2.67e-01

Pos logit contributions
 day+0.024
 night+0.023
 nights+0.023
way+0.022
 Along+0.022
Neg logit contributions
roc-0.023
boys-0.022
 paramed-0.022
 wasteful-0.021
 bluff-0.021
Token.
Token ID13
Feature activation-1.34e-02

Pos logit contributions
 paramed+0.553
 mixer+0.543
 wasteful+0.531
roc+0.530
boys+0.527
Neg logit contributions
 nights-0.279
 Along-0.275
way-0.250
 whenever-0.243
more-0.235
Token I
Token ID314
Feature activation+3.39e-03

Pos logit contributions
 Along+0.019
 On+0.017
 nights+0.016
dark+0.016
 whenever+0.016
Neg logit contributions
 paramed-0.024
 mixer-0.023
boys-0.023
 wasteful-0.022
 liar-0.022
Token want
Token ID765
Feature activation+4.17e-02

Pos logit contributions
 mixer+0.005
 paramed+0.005
 liar+0.005
boys+0.005
 wasteful+0.005
Neg logit contributions
 Along-0.006
 nights-0.005
way-0.005
 On-0.005
 drifted-0.005
Token to
Token ID284
Feature activation+8.47e-02

Pos logit contributions
 paramed+0.085
 liar+0.079
roc+0.078
 bluff+0.078
 mixer+0.076
Neg logit contributions
 nights-0.047
 Along-0.044
 whenever-0.043
 night-0.043
 days-0.042
Token be
Token ID307
Feature activation+4.84e-02

Pos logit contributions
 paramed+0.144
 mixer+0.142
 wasteful+0.140
 bluff+0.137
 liar+0.137
Neg logit contributions
 Along-0.116
 nights-0.109
 whenever-0.107
way-0.107
 On-0.104
Token nice
Token ID3621
Feature activation+7.07e-02

Pos logit contributions
 paramed+0.115
roc+0.114
boys+0.112
 mixer+0.111
 liar+0.107
Neg logit contributions
 nights-0.101
 Along-0.099
 day-0.095
 night-0.095
way-0.093
Token.
Token ID13
Feature activation-1.61e-02

Pos logit contributions
 paramed+0.144
 liar+0.139
 mixer+0.139
 wasteful+0.137
boys+0.137
Neg logit contributions
 Along-0.080
 nights-0.078
way-0.071
 whenever-0.068
 On-0.067
Token He
Token ID679
Feature activation+9.48e-03

Pos logit contributions
 Along+0.014
 On+0.013
way+0.011
 nights+0.010
 As+0.010
Neg logit contributions
boys-0.038
 paramed-0.037
 mixer-0.037
 liar-0.036
 phone-0.036
Token was
Token ID373
Feature activation+3.54e-02

Pos logit contributions
 paramed+0.015
 mixer+0.015
 liar+0.014
roc+0.014
 chefs+0.014
Neg logit contributions
 Along-0.015
 nights-0.014
 drifted-0.014
way-0.013
 On-0.013
Token a
Token ID257
Feature activation-1.76e-02

Pos logit contributions
roc+0.054
boys+0.051
 paramed+0.049
 mixer+0.048
 liar+0.046
Neg logit contributions
 Along-0.057
 nights-0.054
 day-0.053
 before-0.053
 night-0.052
Token liar
Token ID31866
Feature activation+2.65e-01

Pos logit contributions
 day+0.028
 nights+0.027
 night+0.027
 Along+0.027
way+0.025
Neg logit contributions
boys-0.028
roc-0.026
 wasteful-0.024
 paramed-0.024
 bluff-0.024
Token and
Token ID290
Feature activation-1.06e-02

Pos logit contributions
 paramed+0.489
 mixer+0.477
 liar+0.474
 wasteful+0.468
boys+0.467
Neg logit contributions
 Along-0.344
 nights-0.339
way-0.309
 drifted-0.296
 On-0.294
Token a
Token ID257
Feature activation-1.80e-02

Pos logit contributions
 Along+0.016
 nights+0.016
way+0.015
 drifted+0.014
 On+0.014
Neg logit contributions
 paramed-0.017
 mixer-0.017
 liar-0.016
boys-0.016
 wasteful-0.016
Token thief
Token ID25906
Feature activation+1.17e-01

Pos logit contributions
 nights+0.029
 Along+0.028
 day+0.028
 night+0.027
 days+0.027
Neg logit contributions
boys-0.027
roc-0.026
 wasteful-0.024
 bluff-0.024
 paramed-0.023
Token.
Token ID13
Feature activation-1.67e-02

Pos logit contributions
 paramed+0.202
 liar+0.199
 mixer+0.198
 mean+0.196
boys+0.196
Neg logit contributions
 Along-0.168
 nights-0.164
way-0.151
 On-0.146
 drifted-0.144
Token He
Token ID679
Feature activation+9.17e-03

Pos logit contributions
 Along+0.017
 nights+0.016
way+0.014
 On+0.014
 drifted+0.013
Neg logit contributions
 paramed-0.036
 mixer-0.035
boys-0.035
 liar-0.035
 mean-0.035
Token He
Token ID679
Feature activation+1.02e-02

Pos logit contributions
 nights+0.033
 Along+0.031
way+0.029
 drifted+0.028
dark+0.027
Neg logit contributions
 paramed-0.051
 mean-0.050
 liar-0.049
 mixer-0.049
 greedy-0.049
Token said
Token ID531
Feature activation+6.09e-02

Pos logit contributions
 paramed+0.015
 mixer+0.015
 liar+0.014
 bluff+0.014
roc+0.014
Neg logit contributions
 Along-0.016
 nights-0.016
 drifted-0.015
way-0.015
 On-0.015
Token Jen
Token ID13364
Feature activation-7.94e-03

Pos logit contributions
boys+0.129
 bluff+0.125
roc+0.123
 paramed+0.121
 app+0.120
Neg logit contributions
 of-0.056
 whenever-0.055
 days-0.051
way-0.051
 nights-0.051
Token was
Token ID373
Feature activation+3.27e-02

Pos logit contributions
 Along+0.011
 nights+0.011
way+0.011
 drifted+0.011
 whenever+0.011
Neg logit contributions
 paramed-0.013
 mixer-0.013
 chefs-0.012
roc-0.012
 bluff-0.012
Token a
Token ID257
Feature activation-1.78e-02

Pos logit contributions
 paramed+0.039
 mixer+0.035
roc+0.035
boys+0.033
 liar+0.029
Neg logit contributions
 before-0.061
 Along-0.060
 nights-0.060
 day-0.060
 night-0.058
Token liar
Token ID31866
Feature activation+2.64e-01

Pos logit contributions
 day+0.032
 night+0.030
 nights+0.030
way+0.030
 days+0.029
Neg logit contributions
boys-0.024
roc-0.022
 paramed-0.021
 wasteful-0.020
ese-0.020
Token.
Token ID13
Feature activation-2.65e-02

Pos logit contributions
 paramed+0.559
 mixer+0.548
roc+0.543
 wasteful+0.537
 liar+0.536
Neg logit contributions
 nights-0.264
 Along-0.262
way-0.237
 whenever-0.229
more-0.224
Token He
Token ID679
Feature activation+9.89e-03

Pos logit contributions
 nights+0.032
 Along+0.031
way+0.028
 drifted+0.028
dark+0.026
Neg logit contributions
 paramed-0.052
 liar-0.050
 mean-0.050
 mixer-0.050
 greedy-0.050
Token said
Token ID531
Feature activation+6.08e-02

Pos logit contributions
 paramed+0.015
 mixer+0.014
 bluff+0.014
roc+0.014
 liar+0.014
Neg logit contributions
 Along-0.016
 nights-0.015
way-0.015
 drifted-0.015
 On-0.014
Token Jen
Token ID13364
Feature activation-8.23e-03

Pos logit contributions
boys+0.135
 bluff+0.134
roc+0.131
 app+0.129
 paramed+0.129
Neg logit contributions
 whenever-0.046
 of-0.046
 nights-0.043
 days-0.043
 Along-0.041
Token was
Token ID373
Feature activation+3.23e-02

Pos logit contributions
 Along+0.012
 nights+0.012
way+0.011
 drifted+0.011
 whenever+0.011
Neg logit contributions
 paramed-0.014
 mixer-0.013
 bluff-0.013
roc-0.013
 liar-0.013
Token looked
Token ID3114
Feature activation-1.04e-02

Pos logit contributions
 drifted+0.123
 Along+0.121
 nights+0.116
way+0.111
 On+0.106
Neg logit contributions
roc-0.155
 bluff-0.150
 liar-0.150
 wasteful-0.146
 paramed-0.142
Token at
Token ID379
Feature activation+9.48e-03

Pos logit contributions
 Along+0.012
 nights+0.011
 On+0.011
 whenever+0.011
 drifted+0.010
Neg logit contributions
 paramed-0.021
 liar-0.021
roc-0.021
boys-0.021
 mixer-0.020
Token Lily
Token ID20037
Feature activation+3.79e-02

Pos logit contributions
 mixer+0.011
 paramed+0.011
 mean+0.011
 wasteful+0.011
 mad+0.010
Neg logit contributions
 Along-0.019
 nights-0.019
way-0.018
 drifted-0.017
more-0.017
Token with
Token ID351
Feature activation-2.30e-02

Pos logit contributions
 paramed+0.065
 liar+0.064
 mixer+0.063
 wasteful+0.062
 mean+0.062
Neg logit contributions
 Along-0.055
 nights-0.053
way-0.050
 On-0.048
 drifted-0.048
Token a
Token ID257
Feature activation-1.79e-02

Pos logit contributions
 drifted+0.030
 Along+0.030
 nights+0.029
 On+0.029
 night+0.028
Neg logit contributions
roc-0.040
boys-0.040
 paramed-0.040
 chefs-0.037
 mixer-0.037
Token stern
Token ID26370
Feature activation+2.64e-01

Pos logit contributions
 drifted+0.020
 moment+0.019
 particular+0.019
 drift+0.018
 day+0.018
Neg logit contributions
boys-0.038
roc-0.036
 chefs-0.035
 wasteful-0.034
oa-0.033
Token face
Token ID1986
Feature activation+6.52e-02

Pos logit contributions
boys+0.406
 wasteful+0.403
 paramed+0.395
 mixer+0.393
 liar+0.386
Neg logit contributions
 Along-0.406
 nights-0.397
 night-0.380
 drifted-0.378
 On-0.371
Token.
Token ID13
Feature activation-1.72e-02

Pos logit contributions
 paramed+0.136
 liar+0.134
roc+0.132
 wasteful+0.131
 mixer+0.131
Neg logit contributions
 Along-0.069
 nights-0.067
way-0.062
 whenever-0.059
 On-0.059
Token Lily
Token ID20037
Feature activation+4.00e-02

Pos logit contributions
 Along+0.021
 nights+0.020
 On+0.019
way+0.019
 drifted+0.018
Neg logit contributions
 paramed-0.033
 liar-0.032
boys-0.032
 mixer-0.032
 mean-0.032
Token felt
Token ID2936
Feature activation-3.84e-04

Pos logit contributions
 liar+0.083
roc+0.082
 paramed+0.081
 bluff+0.080
 wasteful+0.079
Neg logit contributions
 drifted-0.046
 Along-0.039
 whenever-0.037
 nights-0.036
way-0.036
Token sorry
Token ID7926
Feature activation+2.00e-01

Pos logit contributions
 whenever+0.001
 nights+0.001
 drifted+0.001
 night+0.001
 On+0.001
Neg logit contributions
boys-0.000
roc-0.000
 mixer-0.000
 paramed-0.000
 printer-0.000
Token But
Token ID887
Feature activation-3.03e-02

Pos logit contributions
 nights+0.025
 Along+0.023
way+0.022
ingle+0.022
 outer+0.022
Neg logit contributions
 paramed-0.029
 mixer-0.028
 stern-0.027
 greedy-0.027
 Lee-0.027
Token her
Token ID607
Feature activation-5.91e-02

Pos logit contributions
 Along+0.044
 nights+0.042
way+0.039
 On+0.038
 drifted+0.038
Neg logit contributions
 paramed-0.052
 mixer-0.051
 mean-0.050
 liar-0.050
boys-0.050
Token mother
Token ID2802
Feature activation-9.76e-03

Pos logit contributions
 Along+0.111
 nights+0.109
way+0.103
 On+0.100
 drifted+0.100
Neg logit contributions
 paramed-0.076
 mixer-0.073
 liar-0.073
 mean-0.071
boys-0.071
Token told
Token ID1297
Feature activation+5.78e-02

Pos logit contributions
 nights+0.015
 Along+0.014
way+0.014
 drifted+0.013
 whenever+0.013
Neg logit contributions
 liar-0.016
 chefs-0.015
 bluff-0.015
 paramed-0.015
roc-0.015
Token her
Token ID607
Feature activation-5.87e-02

Pos logit contributions
 paramed+0.074
 mean+0.072
 mixer+0.071
 mad+0.069
 wasteful+0.069
Neg logit contributions
 Along-0.111
 nights-0.106
way-0.101
 On-0.098
 drifted-0.097
Token stern
Token ID26370
Feature activation+2.64e-01

Pos logit contributions
 Along+0.094
 nights+0.092
way+0.085
 On+0.083
 drifted+0.083
Neg logit contributions
 paramed-0.091
 mixer-0.089
 liar-0.089
 mean-0.088
 wasteful-0.087
Tokenly
Token ID306
Feature activation-1.19e-01

Pos logit contributions
 paramed+0.741
 liar+0.739
 mixer+0.733
boys+0.732
 chefs+0.730
Neg logit contributions
 nights-0.077
 Along-0.075
way-0.059
 whenever-0.043
 night-0.041
Token,
Token ID11
Feature activation+8.54e-02

Pos logit contributions
 nights+0.162
 Along+0.159
way+0.147
 whenever+0.147
 drifted+0.143
Neg logit contributions
 paramed-0.208
 liar-0.204
 bluff-0.202
roc-0.201
boys-0.201
Token "
Token ID366
Feature activation+2.11e-01

Pos logit contributions
roc+0.130
 liar+0.128
boys+0.125
 paramed+0.122
 chefs+0.121
Neg logit contributions
 Along-0.141
 nights-0.139
way-0.130
 whenever-0.129
 On-0.128
TokenNo
Token ID2949
Feature activation+8.32e-02

Pos logit contributions
 paramed+0.293
 mixer+0.281
 mean+0.278
 wasteful+0.276
 greedy+0.273
Neg logit contributions
 Along-0.375
 nights-0.367
way-0.337
 On-0.334
 drifted-0.333
Token more
Token ID517
Feature activation+3.08e-02

Pos logit contributions
 paramed+0.144
boys+0.140
 liar+0.139
 chefs+0.136
 bluff+0.136
Neg logit contributions
 nights-0.121
 Along-0.114
way-0.112
 days-0.109
 night-0.106
Token?"
Token ID1701
Feature activation+7.96e-02

Pos logit contributions
 paramed+0.141
 liar+0.133
boys+0.125
 stern+0.123
 bluff+0.123
Neg logit contributions
 nights-0.105
 whenever-0.105
 night-0.102
 before-0.100
 day-0.099
Token Mom
Token ID11254
Feature activation+6.93e-02

Pos logit contributions
 paramed+0.170
 liar+0.166
 mixer+0.166
boys+0.165
 wasteful+0.164
Neg logit contributions
 Along-0.081
 nights-0.077
way-0.069
 On-0.068
 drifted-0.066
Token said
Token ID531
Feature activation+6.29e-02

Pos logit contributions
 liar+0.121
 paramed+0.114
roc+0.110
 chefs+0.109
 wasteful+0.109
Neg logit contributions
 Along-0.095
 On-0.093
 drifted-0.090
way-0.089
 night-0.089
Token in
Token ID287
Feature activation-5.84e-02

Pos logit contributions
roc+0.147
 paramed+0.147
boys+0.144
 liar+0.144
 wasteful+0.142
Neg logit contributions
 Along-0.048
 nights-0.045
 whenever-0.044
way-0.044
more-0.040
Token a
Token ID257
Feature activation-1.62e-02

Pos logit contributions
 Along+0.092
 nights+0.088
 On+0.083
 drifted+0.081
 whenever+0.080
Neg logit contributions
 paramed-0.092
boys-0.091
 mixer-0.090
 liar-0.089
 wasteful-0.089
Token stern
Token ID26370
Feature activation+2.64e-01

Pos logit contributions
 day+0.025
 Along+0.025
 nights+0.024
 particular+0.024
 night+0.024
Neg logit contributions
boys-0.028
roc-0.024
 chefs-0.024
oa-0.024
ese-0.023
Token but
Token ID475
Feature activation+2.08e-02

Pos logit contributions
boys+0.425
 wasteful+0.424
 mixer+0.420
 liar+0.419
 paramed+0.418
Neg logit contributions
 Along-0.380
 nights-0.379
 night-0.360
way-0.356
ly-0.348
Token gentle
Token ID10296
Feature activation+6.81e-02

Pos logit contributions
roc+0.055
 liar+0.054
 paramed+0.053
boys+0.052
 chefs+0.051
Neg logit contributions
 Along-0.010
 nights-0.010
 night-0.010
way-0.009
 drifted-0.009
Token voice
Token ID3809
Feature activation+4.43e-02

Pos logit contributions
boys+0.101
 wasteful+0.100
 liar+0.098
 paramed+0.098
 mixer+0.098
Neg logit contributions
 nights-0.111
 Along-0.108
way-0.103
 drifted-0.101
 night-0.100
Token.
Token ID13
Feature activation-2.05e-02

Pos logit contributions
roc+0.098
 chefs+0.096
 liar+0.096
boys+0.096
 paramed+0.092
Neg logit contributions
way-0.039
ly-0.034
 nights-0.034
dark-0.033
 On-0.033
Token
Token ID198
Feature activation-1.55e-02

Pos logit contributions
 Along+0.024
 nights+0.023
way+0.020
 On+0.020
 drifted+0.020
Neg logit contributions
 paramed-0.041
 liar-0.040
 mixer-0.040
boys-0.040
 mean-0.040
Token holes
Token ID10421
Feature activation-2.68e-02

Pos logit contributions
 Along+0.048
 nights+0.047
way+0.043
 drifted+0.042
 On+0.041
Neg logit contributions
 paramed-0.057
 mixer-0.055
 liar-0.055
boys-0.054
 mean-0.054
Token!"
Token ID2474
Feature activation+1.13e-01

Pos logit contributions
 Along+0.044
 nights+0.040
 On+0.038
way+0.037
 drifted+0.037
Neg logit contributions
 mean-0.042
 mixer-0.042
 wasteful-0.041
 greedy-0.040
 liar-0.040
Token Mom
Token ID11254
Feature activation+7.03e-02

Pos logit contributions
 paramed+0.220
 liar+0.217
 wasteful+0.215
 mixer+0.213
roc+0.212
Neg logit contributions
 Along-0.133
 nights-0.124
 On-0.121
 whenever-0.116
 drifted-0.115
Token said
Token ID531
Feature activation+6.49e-02

Pos logit contributions
 liar+0.128
 paramed+0.116
roc+0.116
 wasteful+0.114
 bluff+0.114
Neg logit contributions
 Along-0.093
way-0.090
 On-0.090
 of-0.089
 nights-0.088
Token,
Token ID11
Feature activation+8.94e-02

Pos logit contributions
roc+0.144
 paramed+0.143
 liar+0.143
 wasteful+0.141
boys+0.140
Neg logit contributions
 Along-0.057
 nights-0.055
way-0.053
 whenever-0.051
 drifted-0.047
Token stern
Token ID26370
Feature activation+2.63e-01

Pos logit contributions
roc+0.146
 liar+0.126
 wasteful+0.126
boys+0.123
 mixer+0.118
Neg logit contributions
 Along-0.147
 drifted-0.145
 nights-0.136
 whenever-0.135
 before-0.133
Tokenly
Token ID306
Feature activation-1.17e-01

Pos logit contributions
 liar+0.727
 paramed+0.726
 mixer+0.719
 wasteful+0.714
boys+0.711
Neg logit contributions
 Along-0.094
 nights-0.087
way-0.066
 On-0.051
 whenever-0.049
Token.
Token ID13
Feature activation-1.75e-02

Pos logit contributions
 Along+0.091
 nights+0.090
way+0.087
 drifted+0.085
 before+0.084
Neg logit contributions
roc-0.275
 liar-0.264
 wasteful-0.262
 bluff-0.257
 paramed-0.254
Token
Token ID198
Feature activation-1.22e-02

Pos logit contributions
 Along+0.017
 nights+0.016
 On+0.014
way+0.014
 drifted+0.014
Neg logit contributions
 paramed-0.038
 liar-0.038
 mixer-0.038
boys-0.037
 wasteful-0.037
Token
Token ID198
Feature activation-1.06e-02

Pos logit contributions
 Along+0.014
 nights+0.014
way+0.013
 drifted+0.012
 On+0.011
Neg logit contributions
 paramed-0.024
 mixer-0.024
 liar-0.023
 mean-0.023
boys-0.023
TokenMax
Token ID11518
Feature activation+3.81e-02

Pos logit contributions
 Along+0.017
 nights+0.016
way+0.015
 On+0.015
more+0.015
Neg logit contributions
 paramed-0.017
 liar-0.016
 mixer-0.016
 wasteful-0.016
 greedy-0.016
Token walked
Token ID6807
Feature activation-1.53e-01

Pos logit contributions
 mixer+0.028
 bluff+0.028
 paramed+0.028
 liar+0.027
roc+0.027
Neg logit contributions
 Along-0.021
 drifted-0.021
 nights-0.020
way-0.019
 On-0.019
Token towards
Token ID3371
Feature activation-6.40e-02

Pos logit contributions
 Along+0.195
 nights+0.165
way+0.164
 On+0.159
 whenever+0.157
Neg logit contributions
 liar-0.317
 mixer-0.309
 paramed-0.305
 wasteful-0.297
boys-0.296
Token them
Token ID606
Feature activation+9.54e-02

Pos logit contributions
 Along+0.127
 nights+0.127
way+0.118
 whenever+0.117
 On+0.116
Neg logit contributions
 paramed-0.074
 liar-0.072
 wasteful-0.070
 mixer-0.069
boys-0.069
Token with
Token ID351
Feature activation-2.04e-02

Pos logit contributions
 paramed+0.167
 mixer+0.164
boys+0.163
 mean+0.163
 liar+0.162
Neg logit contributions
 Along-0.131
 nights-0.131
way-0.117
 drifted-0.115
dark-0.113
Token a
Token ID257
Feature activation-1.59e-02

Pos logit contributions
 nights+0.031
 Along+0.030
 whenever+0.028
 days+0.028
 drifted+0.028
Neg logit contributions
 paramed-0.033
roc-0.032
 wasteful-0.031
 liar-0.031
boys-0.031
Token stern
Token ID26370
Feature activation+2.62e-01

Pos logit contributions
 Along+0.021
 day+0.020
 drifted+0.019
 journey+0.019
 On+0.019
Neg logit contributions
boys-0.029
roc-0.028
 wasteful-0.026
 chefs-0.026
oa-0.026
Token face
Token ID1986
Feature activation+6.15e-02

Pos logit contributions
boys+0.409
 wasteful+0.404
 mixer+0.398
 paramed+0.394
 chefs+0.388
Neg logit contributions
 Along-0.401
 nights-0.395
 night-0.373
way-0.371
 drifted-0.365
Token.
Token ID13
Feature activation-1.59e-02

Pos logit contributions
 paramed+0.128
 liar+0.126
 mixer+0.125
 wasteful+0.124
boys+0.123
Neg logit contributions
 Along-0.065
 nights-0.063
way-0.058
 On-0.054
 drifted-0.054
Token
Token ID198
Feature activation-5.94e-03

Pos logit contributions
 Along+0.017
 nights+0.016
way+0.014
 On+0.014
 drifted+0.013
Neg logit contributions
 paramed-0.034
 mixer-0.033
 liar-0.033
boys-0.033
 mean-0.032
Token
Token ID198
Feature activation-5.53e-03

Pos logit contributions
 nights+0.008
 Along+0.008
 drifted+0.007
way+0.007
 days+0.007
Neg logit contributions
 paramed-0.011
 mean-0.010
 mixer-0.010
 liar-0.010
 stern-0.010
Token"
Token ID1
Feature activation+1.75e-01

Pos logit contributions
 Along+0.009
 nights+0.009
way+0.008
 On+0.008
more+0.008
Neg logit contributions
 paramed-0.008
 mixer-0.008
 liar-0.008
boys-0.008
 wasteful-0.008
Token Dad
Token ID17415
Feature activation+9.28e-02

Pos logit contributions
boys+0.124
 paramed+0.122
 liar+0.120
roc+0.119
 mixer+0.116
Neg logit contributions
 Along-0.128
 nights-0.123
 On-0.120
 drifted-0.116
 night-0.115
Token asked
Token ID1965
Feature activation+2.21e-02

Pos logit contributions
 liar+0.190
roc+0.181
 wasteful+0.175
 paramed+0.169
 chefs+0.168
Neg logit contributions
 drifted-0.109
 Along-0.109
 On-0.107
 nights-0.100
way-0.100
Token.
Token ID13
Feature activation-2.27e-02

Pos logit contributions
 paramed+0.050
 liar+0.049
 mixer+0.049
 wasteful+0.049
boys+0.048
Neg logit contributions
 Along-0.019
 nights-0.018
way-0.016
 whenever-0.016
 On-0.015
Token He
Token ID679
Feature activation+1.06e-02

Pos logit contributions
 Along+0.022
 On+0.019
 nights+0.019
way+0.018
 drifted+0.017
Neg logit contributions
boys-0.051
 paramed-0.050
roc-0.049
 liar-0.049
 mixer-0.049
Token was
Token ID373
Feature activation+3.46e-02

Pos logit contributions
roc+0.019
 liar+0.019
 wasteful+0.018
boys+0.018
 paramed+0.018
Neg logit contributions
 drifted-0.014
 Along-0.014
 On-0.013
 nights-0.013
way-0.012
Token stern
Token ID26370
Feature activation+2.62e-01

Pos logit contributions
roc+0.057
boys+0.051
 liar+0.048
 Cubs+0.048
 mixer+0.048
Neg logit contributions
 of-0.049
 before-0.047
 as-0.047
 drifted-0.046
 Along-0.046
Token and
Token ID290
Feature activation-1.58e-02

Pos logit contributions
 paramed+0.501
 mixer+0.488
 mean+0.487
 liar+0.485
 greedy+0.480
Neg logit contributions
 Along-0.326
 nights-0.320
way-0.287
 drifted-0.277
 On-0.274
Token sad
Token ID6507
Feature activation+1.02e-01

Pos logit contributions
 drifted+0.029
 nights+0.027
 Along+0.027
 day+0.026
 days+0.026
Neg logit contributions
roc-0.020
boys-0.020
 mixer-0.018
 chefs-0.018
 paramed-0.018
Token.
Token ID13
Feature activation-2.23e-02

Pos logit contributions
 mean+0.185
 paramed+0.181
 mixer+0.180
 mad+0.179
 phone+0.178
Neg logit contributions
 Along-0.147
 nights-0.139
way-0.124
dark-0.120
 drifted-0.120
Token
Token ID198
Feature activation-1.21e-02

Pos logit contributions
 Along+0.022
 nights+0.020
 On+0.019
way+0.018
 drifted+0.017
Neg logit contributions
 paramed-0.049
boys-0.048
 mixer-0.048
 liar-0.048
 wasteful-0.047
Token
Token ID198
Feature activation-1.20e-02

Pos logit contributions
 nights+0.017
 Along+0.017
 drifted+0.015
way+0.015
filled+0.014
Neg logit contributions
 paramed-0.021
 mean-0.020
 mixer-0.020
 stern-0.020
 mad-0.020
Token would
Token ID561
Feature activation+2.05e-02

Pos logit contributions
 whenever+0.034
 drifted+0.033
 nights+0.032
 Along+0.031
way+0.030
Neg logit contributions
roc-0.046
 liar-0.046
 paramed-0.046
 bluff-0.044
 mixer-0.043
Token call
Token ID869
Feature activation+8.43e-04

Pos logit contributions
 paramed+0.031
roc+0.030
 chefs+0.029
 mixer+0.029
 liar+0.029
Neg logit contributions
 Along-0.034
 nights-0.032
 whenever-0.032
 drifted-0.031
way-0.031
Token him
Token ID683
Feature activation+4.03e-02

Pos logit contributions
 paramed+0.002
 greedy+0.001
 liar+0.001
 wasteful+0.001
roc+0.001
Neg logit contributions
 nights-0.001
 Along-0.001
 whenever-0.001
 night-0.001
 days-0.001
Token a
Token ID257
Feature activation-2.57e-02

Pos logit contributions
 paramed+0.074
 mixer+0.071
 liar+0.071
 wasteful+0.070
 greedy+0.068
Neg logit contributions
 Along-0.056
 nights-0.052
 whenever-0.050
 On-0.049
way-0.048
Token stupid
Token ID8531
Feature activation+1.25e-01

Pos logit contributions
 day+0.037
 Along+0.036
 night+0.036
way+0.035
 nights+0.035
Neg logit contributions
boys-0.044
roc-0.041
 wasteful-0.040
 chefs-0.039
eper-0.039
Token liar
Token ID31866
Feature activation+2.61e-01

Pos logit contributions
boys+0.192
roc+0.182
 wasteful+0.181
 paramed+0.180
 bluff+0.174
Neg logit contributions
 day-0.194
 night-0.193
 nights-0.189
 Along-0.186
way-0.184
Token.
Token ID13
Feature activation-2.77e-02

Pos logit contributions
 paramed+0.569
 mixer+0.557
roc+0.552
 wasteful+0.552
 liar+0.551
Neg logit contributions
 Along-0.246
 nights-0.243
way-0.218
 whenever-0.207
 drifted-0.202
Token
Token ID198
Feature activation-2.72e-02

Pos logit contributions
 nights+0.044
 Along+0.043
way+0.040
 drifted+0.039
more+0.038
Neg logit contributions
 paramed-0.042
 mixer-0.042
boys-0.041
 mean-0.041
 liar-0.041
Token
Token ID198
Feature activation-2.68e-02

Pos logit contributions
 nights+0.036
 Along+0.034
 drifted+0.032
way+0.031
more+0.030
Neg logit contributions
 mixer-0.048
 mean-0.048
 paramed-0.048
 phone-0.047
 liar-0.047
TokenOne
Token ID3198
Feature activation-1.56e-01

Pos logit contributions
 Along+0.033
 nights+0.032
way+0.029
 On+0.028
 drifted+0.027
Neg logit contributions
 paramed-0.052
 liar-0.051
 mixer-0.051
 wasteful-0.050
boys-0.050
Token day
Token ID1110
Feature activation-2.17e-01

Pos logit contributions
 Along+0.129
 nights+0.115
 drifted+0.106
growth+0.103
 horizon+0.102
Neg logit contributions
 mean-0.401
 mad-0.371
 greedy-0.371
 mixer-0.365
 liar-0.363
Token?"
Token ID1701
Feature activation+8.66e-02

Pos logit contributions
roc+0.065
 liar+0.065
 wasteful+0.065
 paramed+0.065
 greedy+0.063
Neg logit contributions
 day-0.043
 night-0.041
 whenever-0.041
 nights-0.040
way-0.040
Token she
Token ID673
Feature activation-1.45e-01

Pos logit contributions
 paramed+0.180
 liar+0.179
boys+0.178
 mixer+0.176
roc+0.175
Neg logit contributions
 Along-0.090
 nights-0.086
 On-0.079
way-0.076
 drifted-0.076
Token asked
Token ID1965
Feature activation+2.27e-02

Pos logit contributions
 drifted+0.173
 nights+0.172
 Along+0.166
 On+0.155
 night+0.148
Neg logit contributions
 liar-0.274
 paramed-0.272
roc-0.269
 mixer-0.266
 wasteful-0.264
Token in
Token ID287
Feature activation-6.16e-02

Pos logit contributions
 paramed+0.051
 liar+0.050
 wasteful+0.050
roc+0.049
 mixer+0.048
Neg logit contributions
 Along-0.021
 nights-0.020
 whenever-0.019
way-0.018
 On-0.017
Token a
Token ID257
Feature activation-1.85e-02

Pos logit contributions
 Along+0.108
 nights+0.106
way+0.099
 drifted+0.096
 On+0.096
Neg logit contributions
 paramed-0.087
 mixer-0.085
 liar-0.084
 mean-0.083
boys-0.083
Token stern
Token ID26370
Feature activation+2.61e-01

Pos logit contributions
 Along+0.023
 nights+0.022
 particular+0.022
 drifted+0.022
 day+0.022
Neg logit contributions
boys-0.037
oa-0.033
roc-0.032
 chefs-0.032
ese-0.031
Token voice
Token ID3809
Feature activation+4.20e-02

Pos logit contributions
 wasteful+0.406
 mixer+0.401
 liar+0.399
 paramed+0.396
boys+0.392
Neg logit contributions
 Along-0.402
 nights-0.393
way-0.371
 On-0.366
 night-0.363
Token.
Token ID13
Feature activation-1.88e-02

Pos logit contributions
 liar+0.092
 paramed+0.092
 mixer+0.090
 wasteful+0.090
roc+0.090
Neg logit contributions
 Along-0.038
 nights-0.036
way-0.034
 On-0.032
 whenever-0.031
Token "
Token ID366
Feature activation+2.11e-01

Pos logit contributions
 Along+0.015
 On+0.014
way+0.011
 As+0.011
 drifted+0.011
Neg logit contributions
 wasteful-0.045
roc-0.045
boys-0.045
 liar-0.045
 paramed-0.044
TokenWhy
Token ID5195
Feature activation+2.44e-02

Pos logit contributions
 paramed+0.303
 mixer+0.295
 liar+0.294
 mean+0.288
boys+0.288
Neg logit contributions
 Along-0.363
 nights-0.356
way-0.334
 On-0.324
 drifted-0.323
Token are
Token ID389
Feature activation+1.39e-01

Pos logit contributions
 paramed+0.040
roc+0.035
 mixer+0.035
 wasteful+0.035
 liar+0.034
Neg logit contributions
 whenever-0.035
 night-0.035
 Along-0.034
 On-0.034
 day-0.034
Token,
Token ID11
Feature activation+9.04e-02

Pos logit contributions
 wasteful+0.125
roc+0.124
 mixer+0.121
 chefs+0.121
 paramed+0.121
Neg logit contributions
 Along-0.092
 nights-0.089
 whenever-0.087
way-0.085
 drifted-0.082
Token "
Token ID366
Feature activation+2.08e-01

Pos logit contributions
 liar+0.100
 paramed+0.099
 wasteful+0.098
roc+0.097
 mixer+0.096
Neg logit contributions
 Along-0.187
 nights-0.178
way-0.171
 On-0.170
 drifted-0.169
TokenFollow
Token ID7155
Feature activation-1.19e-01

Pos logit contributions
 paramed+0.243
 mean+0.221
 greedy+0.220
Fighting+0.218
boys+0.217
Neg logit contributions
 nights-0.387
 Along-0.382
way-0.365
 On-0.359
lights-0.358
Token your
Token ID534
Feature activation+7.59e-02

Pos logit contributions
 Along+0.217
 nights+0.214
way+0.201
 On+0.195
 drifted+0.194
Neg logit contributions
 paramed-0.160
 liar-0.155
 mixer-0.154
 wasteful-0.150
 mean-0.150
Token dreams
Token ID10625
Feature activation-1.63e-01

Pos logit contributions
 mean+0.119
 phone+0.119
 liar+0.117
boys+0.116
 mixer+0.116
Neg logit contributions
 Along-0.121
 nights-0.113
way-0.113
 On-0.109
dark-0.109
Token'
Token ID6
Feature activation+2.61e-01

Pos logit contributions
 Along+0.268
 nights+0.263
 drifted+0.243
way+0.242
dark+0.240
Neg logit contributions
 mean-0.234
 phone-0.228
boys-0.227
 mixer-0.227
 liar-0.220
Token
Token ID198
Feature activation-5.77e-03

Pos logit contributions
 mean+0.421
boys+0.403
 greedy+0.394
 phone+0.389
 mad+0.385
Neg logit contributions
 Along-0.403
 nights-0.374
 drifted-0.363
ingle-0.360
way-0.359
Token
Token ID198
Feature activation-4.20e-03

Pos logit contributions
 nights+0.009
 Along+0.009
way+0.008
 drifted+0.008
ingle+0.008
Neg logit contributions
 paramed-0.009
 liar-0.009
 mean-0.008
 mixer-0.008
 greedy-0.008
TokenLu
Token ID25596
Feature activation-3.63e-02

Pos logit contributions
 nights+0.006
 Along+0.006
way+0.006
 drifted+0.006
 On+0.006
Neg logit contributions
 paramed-0.007
 liar-0.007
 mixer-0.007
boys-0.007
 mean-0.007
Tokency
Token ID948
Feature activation-2.11e-02

Pos logit contributions
 nights+0.051
 Along+0.050
way+0.046
 On+0.044
 drifted+0.043
Neg logit contributions
 paramed-0.063
 mixer-0.062
boys-0.061
 greedy-0.059
roc-0.059
Token smiled
Token ID13541
Feature activation+2.85e-02

Pos logit contributions
 Along+0.038
 nights+0.037
way+0.036
dark+0.035
 On+0.035
Neg logit contributions
boys-0.027
 mean-0.026
 phone-0.025
 stern-0.025
 paramed-0.025
Token couch
Token ID18507
Feature activation+4.29e-02

Pos logit contributions
 paramed+0.033
 liar+0.032
 mixer+0.031
 mean+0.031
boys+0.031
Neg logit contributions
 Along-0.033
 nights-0.033
way-0.030
 drifted-0.029
 On-0.029
Token and
Token ID290
Feature activation-1.55e-02

Pos logit contributions
 paramed+0.084
 liar+0.083
 wasteful+0.082
roc+0.081
 mixer+0.081
Neg logit contributions
 Along-0.052
 nights-0.050
way-0.047
 whenever-0.044
 On-0.044
Token gave
Token ID2921
Feature activation+3.37e-02

Pos logit contributions
 nights+0.024
 Along+0.024
 drifted+0.023
 whenever+0.022
way+0.022
Neg logit contributions
 paramed-0.024
 liar-0.024
roc-0.023
 mixer-0.023
 wasteful-0.023
Token them
Token ID606
Feature activation+8.90e-02

Pos logit contributions
 paramed+0.056
 liar+0.055
 wasteful+0.052
 bluff+0.051
 dishonest+0.050
Neg logit contributions
 nights-0.052
 whenever-0.051
 night-0.050
 Along-0.048
 days-0.047
Token a
Token ID257
Feature activation-1.88e-02

Pos logit contributions
 liar+0.138
 paramed+0.136
 wasteful+0.131
 mixer+0.128
 bluff+0.127
Neg logit contributions
 Along-0.144
 nights-0.139
 whenever-0.131
 night-0.130
way-0.128
Token stern
Token ID26370
Feature activation+2.60e-01

Pos logit contributions
 Along+0.030
 nights+0.027
way+0.027
 On+0.027
dark+0.027
Neg logit contributions
boys-0.030
roc-0.029
 paramed-0.028
 wasteful-0.028
 liar-0.027
Token look
Token ID804
Feature activation+3.81e-02

Pos logit contributions
boys+0.392
 paramed+0.382
 wasteful+0.382
 mixer+0.375
 liar+0.371
Neg logit contributions
 Along-0.411
 nights-0.405
way-0.386
 night-0.385
 drifted-0.376
Token.
Token ID13
Feature activation-2.21e-02

Pos logit contributions
 paramed+0.072
 mixer+0.071
 mean+0.071
 liar+0.070
 greedy+0.069
Neg logit contributions
 Along-0.050
 nights-0.048
way-0.043
 On-0.042
 drifted-0.041
Token
Token ID198
Feature activation-1.29e-02

Pos logit contributions
 Along+0.027
 nights+0.024
 On+0.024
way+0.023
 whenever+0.022
Neg logit contributions
 paramed-0.043
 liar-0.043
boys-0.043
 wasteful-0.042
 mixer-0.042
Token
Token ID198
Feature activation-1.24e-02

Pos logit contributions
 nights+0.018
 Along+0.017
 drifted+0.016
way+0.016
 days+0.015
Neg logit contributions
 paramed-0.022
 mixer-0.022
 mean-0.022
 stern-0.021
 mad-0.021
Token"
Token ID1
Feature activation+1.69e-01

Pos logit contributions
 Along+0.020
 nights+0.019
way+0.018
 On+0.017
more+0.017
Neg logit contributions
 paramed-0.019
 mixer-0.019
 liar-0.019
boys-0.019
 mean-0.018
Token,
Token ID11
Feature activation+7.54e-02

Pos logit contributions
 Along+0.067
 nights+0.065
way+0.060
 drifted+0.057
 On+0.057
Neg logit contributions
 paramed-0.088
 liar-0.086
 mixer-0.086
 wasteful-0.084
 mean-0.084
Token and
Token ID290
Feature activation-2.36e-02

Pos logit contributions
 paramed+0.117
boys+0.117
 mean+0.116
 mixer+0.115
 liar+0.114
Neg logit contributions
 Along-0.118
 nights-0.112
way-0.111
 On-0.105
ingle-0.103
Token even
Token ID772
Feature activation-4.54e-02

Pos logit contributions
 Along+0.041
way+0.038
 nights+0.036
 On+0.036
ingle+0.036
Neg logit contributions
 mean-0.034
 paramed-0.034
boys-0.033
 liar-0.033
 greedy-0.033
Token went
Token ID1816
Feature activation-1.25e-01

Pos logit contributions
 Along+0.084
way+0.080
ingle+0.076
 nights+0.074
 On+0.074
Neg logit contributions
 mean-0.064
 chefs-0.060
 greedy-0.059
boys-0.059
 mad-0.058
Token to
Token ID284
Feature activation+7.69e-02

Pos logit contributions
 Along+0.196
 nights+0.193
way+0.182
 On+0.174
 drifted+0.172
Neg logit contributions
 paramed-0.195
 mean-0.190
 mixer-0.188
boys-0.188
 liar-0.187
Token outer
Token ID12076
Feature activation-3.32e-01

Pos logit contributions
 mean+0.138
 paramed+0.138
 liar+0.137
boys+0.137
 mixer+0.137
Neg logit contributions
 Along-0.100
 nights-0.097
way-0.094
 drifted-0.086
 On-0.085
Token space
Token ID2272
Feature activation-1.53e-01

Pos logit contributions
 Along+0.396
way+0.361
 nights+0.350
 drifted+0.332
ingle+0.323
Neg logit contributions
 mean-0.628
 liar-0.625
 chefs-0.620
boys-0.618
 paramed-0.617
Token!
Token ID0
Feature activation-2.81e-02

Pos logit contributions
 Along+0.202
 nights+0.196
way+0.180
 On+0.173
 drifted+0.172
Neg logit contributions
 paramed-0.279
 mixer-0.274
 liar-0.273
boys-0.270
 mean-0.270
Token It
Token ID632
Feature activation+6.89e-02

Pos logit contributions
 nights+0.050
ge+0.046
 drifted+0.046
 journey+0.046
ingle+0.045
Neg logit contributions
 paramed-0.039
 mixer-0.035
 mean-0.033
 wasteful-0.032
 greedy-0.032
Token was
Token ID373
Feature activation+2.34e-02

Pos logit contributions
 mean+0.101
boys+0.099
 paramed+0.097
 mixer+0.094
 greedy+0.094
Neg logit contributions
 Along-0.118
 nights-0.109
way-0.106
 On-0.104
more-0.100
Token the
Token ID262
Feature activation+7.76e-03

Pos logit contributions
 paramed+0.033
 liar+0.032
 mixer+0.032
boys+0.032
 wasteful+0.031
Neg logit contributions
 Along-0.040
 nights-0.040
way-0.037
 On-0.036
 drifted-0.036
Token to
Token ID284
Feature activation+7.84e-02

Pos logit contributions
 Along+0.059
 nights+0.058
way+0.053
 drifted+0.053
 On+0.052
Neg logit contributions
 paramed-0.049
 mean-0.049
 mixer-0.048
 liar-0.048
 phone-0.047
Token take
Token ID1011
Feature activation-8.78e-02

Pos logit contributions
 paramed+0.132
 mean+0.132
 mixer+0.132
boys+0.130
 phone+0.128
Neg logit contributions
 nights-0.112
 Along-0.109
way-0.101
 drifted-0.098
dark-0.097
Token a
Token ID257
Feature activation-2.73e-02

Pos logit contributions
 Along+0.169
 nights+0.166
way+0.165
 drifted+0.155
more+0.155
Neg logit contributions
 paramed-0.099
 mean-0.098
 mixer-0.097
 phone-0.095
 mad-0.092
Token journey
Token ID7002
Feature activation-2.09e-01

Pos logit contributions
 nights+0.036
 Along+0.033
 drifted+0.031
more+0.030
way+0.029
Neg logit contributions
 mixer-0.053
 mean-0.053
 phone-0.052
 paramed-0.051
 stern-0.050
Token to
Token ID284
Feature activation+7.83e-02

Pos logit contributions
 Along+0.320
 nights+0.318
 drifted+0.304
dark+0.292
 On+0.286
Neg logit contributions
 phone-0.315
 mean-0.313
 mixer-0.311
boys-0.310
 paramed-0.309
Token outer
Token ID12076
Feature activation-3.28e-01

Pos logit contributions
 mean+0.102
 phone+0.095
 mad+0.092
 paramed+0.092
boys+0.090
Neg logit contributions
 nights-0.143
 Along-0.139
way-0.137
 drifted-0.130
lights-0.130
Token space
Token ID2272
Feature activation-1.51e-01

Pos logit contributions
 Along+0.371
 nights+0.350
way+0.341
 drifted+0.329
dark+0.312
Neg logit contributions
 paramed-0.636
 mean-0.630
 mixer-0.627
 liar-0.623
 greedy-0.616
Token.
Token ID13
Feature activation-2.98e-02

Pos logit contributions
 Along+0.191
 nights+0.184
way+0.170
 drifted+0.164
 On+0.162
Neg logit contributions
 paramed-0.277
 mixer-0.275
boys-0.272
 mean-0.272
 liar-0.271
Token He
Token ID679
Feature activation-3.11e-03

Pos logit contributions
 nights+0.046
way+0.043
 days+0.041
dark+0.041
more+0.040
Neg logit contributions
 paramed-0.048
 mixer-0.045
 Lee-0.044
 stern-0.041
 chefs-0.040
Token hopped
Token ID45230
Feature activation-1.39e-01

Pos logit contributions
 Along+0.005
way+0.005
more+0.005
 nights+0.005
ingle+0.005
Neg logit contributions
 mean-0.004
boys-0.004
 phone-0.004
 mixer-0.004
 paramed-0.004
Token into
Token ID656
Feature activation-7.35e-02

Pos logit contributions
 nights+0.223
more+0.210
way+0.207
dark+0.206
ingle+0.205
Neg logit contributions
 stern-0.192
 mean-0.190
 phone-0.188
 mixer-0.187
 paramed-0.184
Tokenizz
Token ID6457
Feature activation-7.19e-02

Pos logit contributions
 paramed+0.120
 liar+0.118
boys+0.117
 mixer+0.115
 mean+0.115
Neg logit contributions
 Along-0.102
 nights-0.101
way-0.095
 drifted-0.089
 On-0.088
Tokened
Token ID276
Feature activation-3.72e-02

Pos logit contributions
 Along+0.097
 On+0.089
 nights+0.088
way+0.083
 drifted+0.080
Neg logit contributions
boys-0.129
 paramed-0.129
 mixer-0.125
 phone-0.124
 mean-0.121
Token off
Token ID572
Feature activation-9.41e-02

Pos logit contributions
 Along+0.054
 nights+0.054
way+0.050
 drifted+0.048
 On+0.047
Neg logit contributions
 paramed-0.062
 mixer-0.061
 liar-0.060
 mean-0.060
boys-0.060
Token back
Token ID736
Feature activation+5.30e-02

Pos logit contributions
 Along+0.142
 nights+0.142
way+0.130
 On+0.125
 drifted+0.124
Neg logit contributions
 paramed-0.152
 mixer-0.148
 liar-0.146
 mean-0.146
boys-0.145
Token into
Token ID656
Feature activation-7.05e-02

Pos logit contributions
 phone+0.075
boys+0.071
 mad+0.069
 mixer+0.069
 Lee+0.068
Neg logit contributions
 nights-0.087
way-0.083
 Along-0.083
dark-0.081
 drifted-0.081
Token outer
Token ID12076
Feature activation-3.26e-01

Pos logit contributions
 Along+0.132
 nights+0.130
way+0.122
 On+0.119
 drifted+0.119
Neg logit contributions
 paramed-0.091
 mixer-0.088
 liar-0.087
 mean-0.086
boys-0.085
Token space
Token ID2272
Feature activation-1.51e-01

Pos logit contributions
 Along+0.380
way+0.351
 nights+0.345
 drifted+0.330
dark+0.329
Neg logit contributions
 liar-0.612
 paramed-0.610
 mean-0.606
 greedy-0.604
 mixer-0.597
Token.
Token ID13
Feature activation-2.04e-02

Pos logit contributions
 Along+0.180
 nights+0.173
way+0.158
 On+0.153
 drifted+0.151
Neg logit contributions
 paramed-0.291
 mixer-0.287
 liar-0.285
boys-0.283
 mean-0.283
Token Earth
Token ID3668
Feature activation-1.44e-01

Pos logit contributions
 nights+0.038
ingle+0.033
way+0.033
 drifted+0.032
 days+0.032
Neg logit contributions
 paramed-0.028
 Lee-0.027
 mixer-0.026
 sorry-0.024
 stern-0.024
Token was
Token ID373
Feature activation+2.89e-02

Pos logit contributions
 Along+0.221
 nights+0.214
way+0.199
 On+0.195
 drifted+0.190
Neg logit contributions
 paramed-0.231
 mixer-0.226
boys-0.224
 liar-0.224
 mean-0.223
Token an
Token ID281
Feature activation-4.36e-03

Pos logit contributions
 paramed+0.042
 liar+0.041
 mixer+0.041
boys+0.040
 wasteful+0.040
Neg logit contributions
 Along-0.049
 nights-0.049
way-0.045
 drifted-0.044
 whenever-0.044
Token looked
Token ID3114
Feature activation-1.63e-02

Pos logit contributions
 paramed+0.080
 wasteful+0.079
 liar+0.079
 dishonest+0.077
 chefs+0.077
Neg logit contributions
 Along-0.027
way-0.026
 On-0.025
 nights-0.024
 drifted-0.024
Token like
Token ID588
Feature activation-2.57e-03

Pos logit contributions
 nights+0.024
 Along+0.024
 whenever+0.023
 On+0.023
way+0.022
Neg logit contributions
 paramed-0.028
 mixer-0.026
 liar-0.026
boys-0.026
 Lee-0.025
Token it
Token ID340
Feature activation+3.44e-02

Pos logit contributions
 Along+0.004
 nights+0.004
way+0.004
 drifted+0.004
 On+0.004
Neg logit contributions
 paramed-0.004
 mixer-0.004
 liar-0.004
 mean-0.003
boys-0.003
Token came
Token ID1625
Feature activation+4.73e-02

Pos logit contributions
 mean+0.053
boys+0.049
 mixer+0.049
 greedy+0.048
 phone+0.048
Neg logit contributions
 Along-0.060
 nights-0.057
way-0.053
 On-0.051
 drifted-0.051
Token from
Token ID422
Feature activation-2.81e-02

Pos logit contributions
 mean+0.075
 mad+0.069
boys+0.068
 greedy+0.066
 phone+0.065
Neg logit contributions
 nights-0.082
 Along-0.082
 drifted-0.076
 days-0.072
ingle-0.072
Token outer
Token ID12076
Feature activation-3.26e-01

Pos logit contributions
 Along+0.051
 nights+0.049
way+0.047
 drifted+0.045
ingle+0.045
Neg logit contributions
 mean-0.038
 wasteful-0.037
 mad-0.037
 mixer-0.037
 greedy-0.037
Token space
Token ID2272
Feature activation-1.49e-01

Pos logit contributions
 Along+0.399
 nights+0.370
way+0.359
 drifted+0.337
dark+0.332
Neg logit contributions
 mean-0.614
 liar-0.612
 paramed-0.610
 greedy-0.605
 wasteful-0.600
Token.
Token ID13
Feature activation-2.36e-02

Pos logit contributions
 Along+0.170
 nights+0.166
way+0.149
 On+0.141
 drifted+0.141
Neg logit contributions
 paramed-0.300
 mixer-0.293
 liar-0.292
 mean-0.288
 wasteful-0.288
Token
Token ID198
Feature activation-2.76e-02

Pos logit contributions
 Along+0.027
 nights+0.025
way+0.023
 On+0.023
 drifted+0.022
Neg logit contributions
 paramed-0.048
 liar-0.048
 mixer-0.047
 wasteful-0.047
 mean-0.047
Token
Token ID198
Feature activation-2.67e-02

Pos logit contributions
 nights+0.044
 Along+0.041
 drifted+0.039
 horizon+0.037
more+0.037
Neg logit contributions
 mixer-0.042
 mean-0.042
 paramed-0.041
 stern-0.040
 mad-0.040
TokenL
Token ID43
Feature activation-3.06e-02

Pos logit contributions
 Along+0.042
 nights+0.041
way+0.037
 drifted+0.037
 On+0.036
Neg logit contributions
 paramed-0.043
 mixer-0.042
 mean-0.041
 liar-0.041
 wasteful-0.041
Token special
Token ID2041
Feature activation-4.52e-02

Pos logit contributions
 nights+0.031
 Along+0.030
way+0.028
 night+0.028
 day+0.027
Neg logit contributions
 paramed-0.038
boys-0.037
roc-0.037
 wasteful-0.037
 liar-0.036
Token star
Token ID3491
Feature activation-2.59e-02

Pos logit contributions
 Along+0.072
 nights+0.068
 On+0.063
 drifted+0.062
way+0.062
Neg logit contributions
 mixer-0.073
 mean-0.070
 phone-0.070
 paramed-0.069
 liar-0.069
Token that
Token ID326
Feature activation-3.43e-02

Pos logit contributions
 Along+0.040
 nights+0.037
 On+0.035
way+0.035
 drifted+0.034
Neg logit contributions
 mixer-0.043
 mean-0.042
 phone-0.041
boys-0.041
 liar-0.041
Token comes
Token ID2058
Feature activation+7.16e-02

Pos logit contributions
 Along+0.063
 nights+0.059
way+0.057
 On+0.055
more+0.054
Neg logit contributions
 mean-0.046
 mixer-0.045
 greedy-0.043
boys-0.043
 liar-0.043
Token from
Token ID422
Feature activation-3.03e-02

Pos logit contributions
 mean+0.111
 mad+0.106
boys+0.104
 greedy+0.103
 phone+0.103
Neg logit contributions
 Along-0.118
 nights-0.117
 drifted-0.106
 On-0.103
 days-0.103
Token outer
Token ID12076
Feature activation-3.25e-01

Pos logit contributions
 Along+0.050
 nights+0.048
way+0.046
 drifted+0.045
 On+0.044
Neg logit contributions
 mean-0.045
 mixer-0.044
 paramed-0.043
 mad-0.043
 greedy-0.043
Token space
Token ID2272
Feature activation-1.51e-01

Pos logit contributions
 Along+0.380
 nights+0.345
way+0.334
 drifted+0.318
ingle+0.314
Neg logit contributions
 liar-0.622
 mixer-0.621
 paramed-0.621
 mean-0.620
 greedy-0.609
Token,"
Token ID553
Feature activation+1.27e-01

Pos logit contributions
 Along+0.205
 nights+0.197
way+0.181
 drifted+0.177
 On+0.176
Neg logit contributions
 paramed-0.268
 mixer-0.266
 liar-0.264
 mean-0.263
boys-0.261
Token she
Token ID673
Feature activation-1.51e-01

Pos logit contributions
 paramed+0.236
 mixer+0.231
 mean+0.225
boys+0.223
 phone+0.223
Neg logit contributions
 Along-0.166
 nights-0.163
way-0.151
dark-0.138
 drifted-0.138
Token explained
Token ID4893
Feature activation+1.08e-01

Pos logit contributions
 drifted+0.210
 Along+0.207
way+0.199
 nights+0.197
 On+0.197
Neg logit contributions
 liar-0.260
 wasteful-0.259
 chefs-0.258
 mixer-0.251
roc-0.247
Token.
Token ID13
Feature activation-2.64e-02

Pos logit contributions
 liar+0.251
 wasteful+0.248
 paramed+0.245
roc+0.245
 chefs+0.243
Neg logit contributions
 Along-0.082
 nights-0.080
way-0.078
 whenever-0.078
 On-0.069
Token.
Token ID13
Feature activation-1.69e-02

Pos logit contributions
 nights+0.146
 Along+0.146
way+0.134
dark+0.134
ingle+0.131
Neg logit contributions
 mixer-0.129
boys-0.127
 paramed-0.127
 phone-0.126
 stern-0.125
Token Soon
Token ID15894
Feature activation-1.75e-01

Pos logit contributions
 nights+0.027
ingle+0.025
 drifted+0.025
way+0.024
more+0.024
Neg logit contributions
 paramed-0.025
 mixer-0.024
boys-0.024
 stern-0.023
 Lee-0.023
Token they
Token ID484
Feature activation-1.08e-02

Pos logit contributions
 Along+0.305
 nights+0.284
 On+0.273
way+0.273
dark+0.272
Neg logit contributions
 paramed-0.245
boys-0.240
 mean-0.239
 mixer-0.237
 liar-0.233
Token were
Token ID547
Feature activation+4.13e-02

Pos logit contributions
 Along+0.021
 nights+0.020
way+0.020
 On+0.019
dark+0.019
Neg logit contributions
 mean-0.013
boys-0.012
 paramed-0.012
 mad-0.011
 liar-0.011
Token in
Token ID287
Feature activation-6.06e-02

Pos logit contributions
 mean+0.062
 paramed+0.062
 mad+0.061
boys+0.061
 mixer+0.060
Neg logit contributions
 Along-0.066
 nights-0.064
way-0.061
 On-0.060
dark-0.059
Token outer
Token ID12076
Feature activation-3.24e-01

Pos logit contributions
 Along+0.108
 nights+0.106
way+0.101
dark+0.097
 drifted+0.096
Neg logit contributions
 paramed-0.082
 mixer-0.080
 liar-0.079
 mean-0.079
boys-0.077
Token space
Token ID2272
Feature activation-1.50e-01

Pos logit contributions
 Along+0.373
way+0.351
 nights+0.332
ingle+0.325
dark+0.321
Neg logit contributions
 mean-0.603
 paramed-0.596
 mixer-0.596
 liar-0.593
 chefs-0.587
Token and
Token ID290
Feature activation-1.59e-02

Pos logit contributions
 Along+0.227
 nights+0.214
way+0.201
 On+0.200
dark+0.197
Neg logit contributions
boys-0.239
 mixer-0.239
 paramed-0.237
 mean-0.236
 liar-0.235
Token having
Token ID1719
Feature activation-5.81e-02

Pos logit contributions
 Along+0.026
way+0.023
 On+0.023
 nights+0.023
ingle+0.022
Neg logit contributions
 mean-0.024
boys-0.024
 paramed-0.023
 mixer-0.023
 phone-0.023
Token fun
Token ID1257
Feature activation-2.71e-03

Pos logit contributions
 Along+0.092
 nights+0.088
way+0.085
 On+0.082
dark+0.081
Neg logit contributions
 paramed-0.088
 mean-0.087
boys-0.087
 mixer-0.086
 liar-0.086
Token.
Token ID13
Feature activation-1.71e-02

Pos logit contributions
 Along+0.005
dark+0.004
ingle+0.004
 On+0.004
way+0.004
Neg logit contributions
boys-0.004
 fighting-0.004
 chefs-0.004
 arguing-0.004
 phone-0.004
Token the
Token ID262
Feature activation+1.09e-02

Pos logit contributions
 paramed+0.010
 mixer+0.010
 mean+0.010
 liar+0.010
boys+0.010
Neg logit contributions
 Along-0.013
 nights-0.013
way-0.012
 drifted-0.012
 On-0.012
Token stars
Token ID5788
Feature activation-1.71e-01

Pos logit contributions
 mean+0.020
 rude+0.017
 phone+0.017
 greedy+0.016
 bluff+0.016
Neg logit contributions
 Along-0.017
ingle-0.016
 drifted-0.015
ment-0.015
way-0.015
Token and
Token ID290
Feature activation-1.91e-02

Pos logit contributions
 Along+0.255
 nights+0.244
way+0.228
 On+0.221
 drifted+0.216
Neg logit contributions
 mixer-0.283
 mean-0.281
 paramed-0.278
boys-0.277
 liar-0.273
Token planets
Token ID14705
Feature activation-2.00e-01

Pos logit contributions
 Along+0.032
 nights+0.031
way+0.030
 On+0.029
 drifted+0.029
Neg logit contributions
 paramed-0.028
 mixer-0.027
 liar-0.027
 mean-0.027
boys-0.027
Token in
Token ID287
Feature activation-5.92e-02

Pos logit contributions
 Along+0.263
 nights+0.253
way+0.232
 On+0.224
 drifted+0.219
Neg logit contributions
 mean-0.363
 mixer-0.361
boys-0.359
 liar-0.357
 paramed-0.357
Token outer
Token ID12076
Feature activation-3.24e-01

Pos logit contributions
 Along+0.118
 nights+0.114
way+0.113
dark+0.107
 drifted+0.106
Neg logit contributions
 mean-0.069
 mixer-0.067
 mad-0.066
 paramed-0.065
 liar-0.064
Token space
Token ID2272
Feature activation-1.49e-01

Pos logit contributions
 Along+0.375
way+0.358
 nights+0.344
ingle+0.330
 drifted+0.321
Neg logit contributions
 mean-0.600
 liar-0.585
 mixer-0.583
 greedy-0.581
roc-0.580
Token.
Token ID13
Feature activation-2.41e-02

Pos logit contributions
 Along+0.172
 nights+0.165
way+0.151
 On+0.144
 drifted+0.143
Neg logit contributions
 paramed-0.294
 mixer-0.291
 liar-0.290
 mean-0.289
boys-0.287
Token So
Token ID1406
Feature activation-2.76e-03

Pos logit contributions
 nights+0.040
 drifted+0.037
 outer+0.034
ingle+0.034
 Along+0.033
Neg logit contributions
 mixer-0.038
 paramed-0.034
 mean-0.033
 Lee-0.033
 stern-0.033
Token one
Token ID530
Feature activation-6.68e-02

Pos logit contributions
 Along+0.005
ingle+0.005
powers+0.005
dark+0.005
 nights+0.005
Neg logit contributions
 mixer-0.003
 mean-0.003
 mad-0.003
 stern-0.003
 sorry-0.003
Token day
Token ID1110
Feature activation-2.14e-01

Pos logit contributions
 Along+0.076
growth+0.067
lights+0.067
 horizon+0.067
way+0.065
Neg logit contributions
 mean-0.148
 mixer-0.135
 mad-0.134
 angry-0.131
 phone-0.130
Token that
Token ID326
Feature activation-3.21e-02

Pos logit contributions
 Along+0.028
 nights+0.027
way+0.025
 On+0.024
 drifted+0.024
Neg logit contributions
 paramed-0.026
 mixer-0.025
 liar-0.025
 mean-0.025
boys-0.025
Token they
Token ID484
Feature activation-1.34e-02

Pos logit contributions
 nights+0.050
 Along+0.049
 drifted+0.045
way+0.044
 whenever+0.044
Neg logit contributions
 paramed-0.053
 liar-0.052
 mixer-0.050
 wasteful-0.050
 greedy-0.050
Token were
Token ID547
Feature activation+4.11e-02

Pos logit contributions
 Along+0.022
 nights+0.021
way+0.020
 On+0.019
more+0.018
Neg logit contributions
 paramed-0.020
 mean-0.020
boys-0.020
 liar-0.020
 mixer-0.020
Token going
Token ID1016
Feature activation-7.42e-02

Pos logit contributions
 paramed+0.066
 liar+0.063
 mixer+0.062
roc+0.062
 wasteful+0.061
Neg logit contributions
 nights-0.066
 Along-0.066
way-0.059
 whenever-0.059
 drifted-0.059
Token to
Token ID284
Feature activation+7.89e-02

Pos logit contributions
 Along+0.116
 nights+0.114
way+0.106
 drifted+0.103
 On+0.102
Neg logit contributions
 paramed-0.117
 mixer-0.115
 liar-0.114
 mean-0.113
boys-0.113
Token outer
Token ID12076
Feature activation-3.23e-01

Pos logit contributions
 paramed+0.129
 mixer+0.125
 liar+0.125
 mean+0.124
boys+0.123
Neg logit contributions
 Along-0.120
 nights-0.117
way-0.109
 drifted-0.105
 On-0.105
Token space
Token ID2272
Feature activation-1.50e-01

Pos logit contributions
 Along+0.364
 nights+0.328
way+0.325
 drifted+0.323
ingle+0.305
Neg logit contributions
 mean-0.620
 chefs-0.615
 liar-0.612
 mixer-0.611
boys-0.607
Token.
Token ID13
Feature activation-2.13e-02

Pos logit contributions
 Along+0.205
 nights+0.193
way+0.180
 On+0.178
 drifted+0.177
Neg logit contributions
boys-0.261
 mixer-0.259
 mean-0.258
 liar-0.255
 phone-0.254
Token They
Token ID1119
Feature activation+4.21e-02

Pos logit contributions
 nights+0.030
 Along+0.029
way+0.028
 drifted+0.027
more+0.026
Neg logit contributions
 mixer-0.036
 paramed-0.036
 wasteful-0.034
boys-0.034
 stern-0.034
Token watched
Token ID7342
Feature activation-8.16e-02

Pos logit contributions
 paramed+0.060
boys+0.060
 mean+0.060
 mixer+0.060
 liar+0.059
Neg logit contributions
 Along-0.070
 nights-0.068
way-0.065
 On-0.063
more-0.061
Token out
Token ID503
Feature activation-2.75e-02

Pos logit contributions
 Along+0.146
 nights+0.140
way+0.135
more+0.130
 drifted+0.130
Neg logit contributions
 mixer-0.108
 paramed-0.106
boys-0.106
 mean-0.104
 phone-0.103
Token what
Token ID644
Feature activation-5.76e-02

Pos logit contributions
 nights+0.162
 Along+0.157
way+0.144
 days+0.143
 whenever+0.142
Neg logit contributions
 paramed-0.163
 liar-0.153
 mixer-0.153
 wasteful-0.150
roc-0.149
Token the
Token ID262
Feature activation+1.38e-02

Pos logit contributions
 nights+0.080
 Along+0.078
way+0.072
 drifted+0.069
 whenever+0.069
Neg logit contributions
 paramed-0.103
 liar-0.099
 mixer-0.098
 wasteful-0.098
roc-0.097
Token rocket
Token ID10701
Feature activation-6.54e-02

Pos logit contributions
 mean+0.021
 mixer+0.020
 liar+0.020
 greedy+0.020
 phone+0.020
Neg logit contributions
 Along-0.024
 On-0.022
 nights-0.022
ingle-0.022
way-0.021
Token discovered
Token ID5071
Feature activation-4.57e-02

Pos logit contributions
 Along+0.108
 nights+0.102
 On+0.096
way+0.095
dark+0.091
Neg logit contributions
boys-0.097
 mean-0.097
 mixer-0.097
 paramed-0.096
 phone-0.095
Token in
Token ID287
Feature activation-5.99e-02

Pos logit contributions
 Along+0.070
 nights+0.064
 On+0.059
way+0.059
 drifted+0.058
Neg logit contributions
 mean-0.074
boys-0.073
 mixer-0.072
 phone-0.071
 mad-0.070
Token outer
Token ID12076
Feature activation-3.23e-01

Pos logit contributions
 Along+0.102
 nights+0.101
way+0.094
 On+0.091
 drifted+0.091
Neg logit contributions
 paramed-0.087
 liar-0.084
 mixer-0.084
boys-0.082
 wasteful-0.082
Token space
Token ID2272
Feature activation-1.48e-01

Pos logit contributions
 Along+0.355
 nights+0.322
way+0.318
 drifted+0.297
ingle+0.294
Neg logit contributions
 liar-0.631
 paramed-0.629
 mean-0.629
 mixer-0.626
 greedy-0.620
Token.
Token ID13
Feature activation-1.89e-02

Pos logit contributions
 Along+0.198
 nights+0.188
way+0.173
 On+0.170
 drifted+0.167
Neg logit contributions
 paramed-0.265
 mixer-0.264
 liar-0.262
 mean-0.262
boys-0.261
Token It
Token ID632
Feature activation+7.31e-02

Pos logit contributions
 nights+0.029
 drifted+0.027
ingle+0.026
way+0.026
 days+0.025
Neg logit contributions
 paramed-0.028
 mixer-0.027
 stern-0.027
 Lee-0.027
 greedy-0.026
Token was
Token ID373
Feature activation+3.19e-02

Pos logit contributions
 mean+0.100
boys+0.095
 paramed+0.091
 mad+0.091
 stern+0.090
Neg logit contributions
 Along-0.132
 nights-0.121
 On-0.117
way-0.117
ingle-0.114
Token a
Token ID257
Feature activation-2.11e-02

Pos logit contributions
 paramed+0.045
 mixer+0.044
 liar+0.043
 mean+0.043
boys+0.043
Neg logit contributions
 Along-0.055
 nights-0.054
way-0.051
 On-0.049
 drifted-0.049
Token Emma
Token ID18966
Feature activation+4.72e-02

Pos logit contributions
 nights+0.074
 drifted+0.072
 Along+0.071
way+0.067
ingle+0.066
Neg logit contributions
 mean-0.036
 mixer-0.035
boys-0.034
 paramed-0.033
 greedy-0.033
Token who
Token ID508
Feature activation-2.32e-02

Pos logit contributions
 paramed+0.087
 mixer+0.086
 mean+0.086
boys+0.086
 liar+0.085
Neg logit contributions
 Along-0.061
 nights-0.060
way-0.055
 drifted-0.052
 On-0.051
Token loved
Token ID6151
Feature activation-3.76e-02

Pos logit contributions
 Along+0.037
 nights+0.037
way+0.034
 On+0.033
 drifted+0.033
Neg logit contributions
 paramed-0.036
 mixer-0.035
 liar-0.035
 wasteful-0.034
boys-0.034
Token learning
Token ID4673
Feature activation-8.99e-03

Pos logit contributions
 nights+0.053
 Along+0.053
 whenever+0.046
 night+0.046
 days+0.046
Neg logit contributions
 paramed-0.065
 liar-0.065
 wasteful-0.064
roc-0.063
boys-0.062
Token about
Token ID546
Feature activation+6.49e-03

Pos logit contributions
 Along+0.013
 nights+0.012
way+0.011
 drifted+0.011
 On+0.011
Neg logit contributions
 paramed-0.016
 mixer-0.015
 liar-0.015
 mean-0.015
boys-0.015
Token outer
Token ID12076
Feature activation-3.21e-01

Pos logit contributions
 paramed+0.012
 liar+0.011
 wasteful+0.011
 mixer+0.011
 stern+0.011
Neg logit contributions
 nights-0.009
 Along-0.009
 whenever-0.008
 days-0.008
 On-0.008
Token space
Token ID2272
Feature activation-1.46e-01

Pos logit contributions
 Along+0.409
way+0.386
 nights+0.382
 drifted+0.378
ingle+0.352
Neg logit contributions
 mean-0.580
 liar-0.573
 mixer-0.571
 greedy-0.569
roc-0.567
Token.
Token ID13
Feature activation-2.04e-02

Pos logit contributions
 Along+0.198
 nights+0.187
way+0.178
 drifted+0.176
 On+0.169
Neg logit contributions
 mean-0.259
boys-0.258
 mixer-0.257
 phone-0.255
 liar-0.254
Token Everyday
Token ID48154
Feature activation-1.61e-01

Pos logit contributions
 nights+0.030
way+0.028
 Along+0.028
 drifted+0.028
ingle+0.026
Neg logit contributions
 mixer-0.034
 paramed-0.034
 mean-0.032
 bluff-0.031
 liar-0.031
Token,
Token ID11
Feature activation+8.19e-02

Pos logit contributions
 nights+0.225
 Along+0.220
 days+0.200
 whenever+0.198
 night+0.198
Neg logit contributions
 paramed-0.290
 liar-0.282
 wasteful-0.281
roc-0.280
boys-0.276
Token she
Token ID673
Feature activation-1.54e-01

Pos logit contributions
 mixer+0.147
 mean+0.144
 paramed+0.142
boys+0.139
 phone+0.138
Neg logit contributions
 Along-0.112
way-0.107
 nights-0.107
ingle-0.099
 drifted-0.099
Token of
Token ID286
Feature activation-1.10e-01

Pos logit contributions
 Along+0.267
way+0.233
more+0.231
dark+0.230
 drifted+0.226
Neg logit contributions
 stern-0.254
 mean-0.252
 mixer-0.244
 sorry-0.241
 phone-0.234
Token something
Token ID1223
Feature activation+1.75e-03

Pos logit contributions
 nights+0.167
 days+0.152
 Along+0.151
 night+0.150
 whenever+0.150
Neg logit contributions
 paramed-0.188
roc-0.181
 liar-0.180
boys-0.175
 wasteful-0.172
Token to
Token ID284
Feature activation+7.23e-02

Pos logit contributions
 mixer+0.003
 mean+0.003
 paramed+0.003
 stern+0.003
 greedy+0.003
Neg logit contributions
 Along-0.003
 nights-0.003
way-0.002
 drifted-0.002
 On-0.002
Token eat
Token ID4483
Feature activation-3.48e-02

Pos logit contributions
 paramed+0.139
 liar+0.133
boys+0.132
 mixer+0.131
 greedy+0.128
Neg logit contributions
 Along-0.089
 nights-0.089
 On-0.081
 whenever-0.080
 night-0.079
Token.
Token ID13
Feature activation-1.35e-02

Pos logit contributions
 Along+0.039
 nights+0.038
way+0.034
 whenever+0.033
 On+0.033
Neg logit contributions
 paramed-0.071
 liar-0.069
 mixer-0.068
boys-0.068
 wasteful-0.068
Token Along
Token ID17159
Feature activation-2.97e-01

Pos logit contributions
 nights+0.022
 drifted+0.020
way+0.019
ingle+0.019
lights+0.019
Neg logit contributions
 paramed-0.020
 mixer-0.020
 Lee-0.020
 stern-0.020
 greedy-0.019
Token the
Token ID262
Feature activation+1.27e-02

Pos logit contributions
 Along+0.615
 nights+0.602
way+0.578
dark+0.577
more+0.570
Neg logit contributions
 mixer-0.294
boys-0.283
 mean-0.272
 phone-0.270
 paramed-0.265
Token way
Token ID835
Feature activation-7.36e-02

Pos logit contributions
 mean+0.020
 phone+0.019
 mixer+0.019
 quarrel+0.018
 mad+0.017
Neg logit contributions
ingle-0.021
 drifted-0.019
ment-0.019
dark-0.019
 Along-0.019
Token,
Token ID11
Feature activation+8.43e-02

Pos logit contributions
 Along+0.135
 nights+0.133
 drifted+0.126
dark+0.126
way+0.123
Neg logit contributions
 mixer-0.090
 mean-0.086
boys-0.085
 phone-0.085
 paramed-0.083
Token she
Token ID673
Feature activation-1.53e-01

Pos logit contributions
boys+0.149
 paramed+0.148
 mixer+0.148
 mean+0.146
 phone+0.144
Neg logit contributions
 Along-0.110
 nights-0.103
way-0.100
dark-0.099
ingle-0.098
Token spotted
Token ID13489
Feature activation-4.55e-03

Pos logit contributions
 Along+0.270
 nights+0.267
way+0.247
 drifted+0.244
 On+0.240
Neg logit contributions
 paramed-0.216
 mixer-0.209
 liar-0.208
 wasteful-0.204
 mean-0.201
Token on
Token ID319
Feature activation-3.73e-02

Pos logit contributions
 Along+0.109
 nights+0.106
way+0.096
 On+0.092
 drifted+0.092
Neg logit contributions
 paramed-0.180
 mixer-0.176
 liar-0.176
 mean-0.173
boys-0.173
Token her
Token ID607
Feature activation-7.40e-02

Pos logit contributions
 Along+0.072
 nights+0.070
way+0.066
 drifted+0.065
dark+0.064
Neg logit contributions
 mixer-0.045
 paramed-0.045
 mean-0.045
 mad-0.043
boys-0.043
Token sc
Token ID629
Feature activation+3.15e-02

Pos logit contributions
 Along+0.128
 nights+0.123
way+0.118
 On+0.117
 drifted+0.115
Neg logit contributions
 phone-0.108
 mixer-0.108
boys-0.105
 mean-0.103
 greedy-0.100
Tokenooter
Token ID25141
Feature activation-4.75e-02

Pos logit contributions
 paramed+0.041
roc+0.040
 liar+0.040
 greedy+0.039
 mean+0.038
Neg logit contributions
 nights-0.060
 drifted-0.057
 night-0.053
 Along-0.053
 days-0.053
Token.
Token ID13
Feature activation-2.97e-02

Pos logit contributions
 nights+0.063
 Along+0.063
 drifted+0.059
dark+0.056
way+0.054
Neg logit contributions
 mean-0.084
 phone-0.083
 mixer-0.083
boys-0.083
 paramed-0.082
Token Along
Token ID17159
Feature activation-2.97e-01

Pos logit contributions
 nights+0.058
lights+0.054
 drifted+0.054
 lifetime+0.054
ge+0.053
Neg logit contributions
 paramed-0.037
 Lee-0.036
 mixer-0.034
 phone-0.031
 stern-0.030
Token the
Token ID262
Feature activation+6.33e-03

Pos logit contributions
 Along+0.648
 nights+0.641
way+0.607
dark+0.604
more+0.603
Neg logit contributions
 mixer-0.268
boys-0.254
 mean-0.253
 phone-0.242
 paramed-0.240
Token way
Token ID835
Feature activation-7.53e-02

Pos logit contributions
 mean+0.012
 mixer+0.011
 mad+0.010
 phone+0.010
 paramed+0.010
Neg logit contributions
dark-0.009
 Along-0.009
 nights-0.009
ingle-0.008
 outer-0.008
Token,
Token ID11
Feature activation+7.06e-02

Pos logit contributions
 Along+0.145
 nights+0.143
 drifted+0.137
dark+0.135
more+0.133
Neg logit contributions
 mixer-0.086
 mean-0.083
 phone-0.080
boys-0.078
 paramed-0.078
Token she
Token ID673
Feature activation-1.60e-01

Pos logit contributions
 mixer+0.117
 mean+0.116
boys+0.114
 greedy+0.112
 chefs+0.111
Neg logit contributions
 Along-0.101
way-0.095
ingle-0.095
 nights-0.094
dark-0.092
Token saw
Token ID2497
Feature activation+3.48e-02

Pos logit contributions
 Along+0.286
 nights+0.285
way+0.265
 drifted+0.263
 whenever+0.257
Neg logit contributions
 paramed-0.220
 liar-0.213
 mixer-0.212
 wasteful-0.210
roc-0.204
Token this
Token ID428
Feature activation-5.23e-03

Pos logit contributions
 Along+0.093
 nights+0.091
way+0.083
 On+0.081
 drifted+0.080
Neg logit contributions
 paramed-0.109
 liar-0.108
 mixer-0.105
 wasteful-0.105
boys-0.104
Token way
Token ID835
Feature activation-7.57e-02

Pos logit contributions
 Along+0.009
 nights+0.009
 drifted+0.008
growth+0.008
dark+0.008
Neg logit contributions
 mean-0.008
 phone-0.007
 paramed-0.007
 mixer-0.007
 mad-0.007
Token and
Token ID290
Feature activation-2.75e-02

Pos logit contributions
 nights+0.129
 Along+0.127
growth+0.121
dark+0.120
 drifted+0.118
Neg logit contributions
 mean-0.103
 mixer-0.102
 phone-0.102
 mad-0.102
 fighting-0.096
Token that
Token ID326
Feature activation-3.84e-02

Pos logit contributions
 Along+0.043
 nights+0.043
way+0.040
 drifted+0.038
 On+0.038
Neg logit contributions
 paramed-0.043
 liar-0.042
 mixer-0.042
 wasteful-0.041
boys-0.041
Token.
Token ID13
Feature activation-3.16e-02

Pos logit contributions
 nights+0.057
 Along+0.054
 whenever+0.052
 days+0.050
 night+0.049
Neg logit contributions
 paramed-0.067
 liar-0.065
 mixer-0.064
roc-0.063
 wasteful-0.062
Token Along
Token ID17159
Feature activation-2.96e-01

Pos logit contributions
 nights+0.046
way+0.040
 drifted+0.040
 Along+0.040
ingle+0.039
Neg logit contributions
 paramed-0.056
 mixer-0.053
 liar-0.052
 greedy-0.051
 wasteful-0.051
Token the
Token ID262
Feature activation+7.27e-03

Pos logit contributions
 Along+0.563
 nights+0.551
way+0.521
 On+0.505
dark+0.504
Neg logit contributions
 paramed-0.358
 mixer-0.356
boys-0.345
 mean-0.341
 liar-0.334
Token way
Token ID835
Feature activation-7.58e-02

Pos logit contributions
 mean+0.013
 phone+0.012
 paramed+0.012
 mad+0.012
 fighting+0.012
Neg logit contributions
ingle-0.010
dark-0.010
 Along-0.009
 drifted-0.009
ment-0.009
Token,
Token ID11
Feature activation+7.10e-02

Pos logit contributions
 Along+0.132
 nights+0.130
 drifted+0.120
dark+0.119
way+0.119
Neg logit contributions
 mixer-0.102
 paramed-0.100
 mean-0.099
boys-0.097
 phone-0.096
Token the
Token ID262
Feature activation+7.37e-03

Pos logit contributions
 paramed+0.105
 mixer+0.103
 mean+0.101
 liar+0.101
boys+0.101
Neg logit contributions
 Along-0.117
 nights-0.115
way-0.108
 drifted-0.105
 On-0.104
Token girl
Token ID2576
Feature activation+6.85e-03

Pos logit contributions
 paramed+0.010
 mean+0.009
 mixer+0.009
 phone+0.009
 greedy+0.008
Neg logit contributions
 Along-0.014
 nights-0.014
 drifted-0.014
ingle-0.014
dark-0.013
Token a
Token ID257
Feature activation-2.77e-02

Pos logit contributions
 paramed+0.106
 mean+0.103
 liar+0.103
 mixer+0.102
boys+0.101
Neg logit contributions
 Along-0.137
 nights-0.135
way-0.128
 drifted-0.124
 On-0.122
Token far
Token ID1290
Feature activation-1.20e-01

Pos logit contributions
 Along+0.036
 nights+0.034
way+0.033
 night+0.033
 day+0.031
Neg logit contributions
 paramed-0.051
 liar-0.049
 wasteful-0.049
 mixer-0.049
boys-0.048
Token off
Token ID572
Feature activation-9.44e-02

Pos logit contributions
 Along+0.137
 nights+0.135
 drifted+0.127
dark+0.113
 On+0.113
Neg logit contributions
 mean-0.243
 paramed-0.236
 mad-0.231
 mixer-0.231
 greedy-0.231
Token place
Token ID1295
Feature activation-5.91e-02

Pos logit contributions
 Along+0.200
more+0.193
way+0.191
 nights+0.191
 drifted+0.191
Neg logit contributions
 mean-0.096
 angry-0.078
 mad-0.077
 greedy-0.076
 wasteful-0.073
Token.
Token ID13
Feature activation-2.88e-02

Pos logit contributions
 Along+0.088
 nights+0.087
 drifted+0.082
dark+0.081
way+0.077
Neg logit contributions
 mean-0.093
 stern-0.092
 mixer-0.091
 phone-0.091
 mad-0.090
Token Along
Token ID17159
Feature activation-2.95e-01

Pos logit contributions
 nights+0.050
 days+0.045
lights+0.044
 drifted+0.043
way+0.042
Neg logit contributions
 paramed-0.044
 Lee-0.042
 mixer-0.041
 mean-0.037
 greedy-0.037
Token the
Token ID262
Feature activation+9.00e-03

Pos logit contributions
 nights+0.629
 Along+0.623
way+0.606
more+0.599
dark+0.592
Neg logit contributions
 mixer-0.256
 paramed-0.248
boys-0.246
 Lee-0.240
 phone-0.236
Token way
Token ID835
Feature activation-7.41e-02

Pos logit contributions
 mean+0.016
 quarrel+0.014
boys+0.014
 phone+0.014
 rude+0.014
Neg logit contributions
ingle-0.013
ment-0.012
dark-0.012
edly-0.012
 drifted-0.012
Token,
Token ID11
Feature activation+7.13e-02

Pos logit contributions
 Along+0.142
 nights+0.142
 drifted+0.137
dark+0.137
more+0.135
Neg logit contributions
 mixer-0.077
 mean-0.074
 phone-0.073
 Lee-0.072
boys-0.069
Token she
Token ID673
Feature activation-1.60e-01

Pos logit contributions
 paramed+0.111
 mixer+0.111
 Lee+0.107
 greedy+0.107
 mean+0.107
Neg logit contributions
 nights-0.105
 Along-0.104
way-0.103
ingle-0.101
dark-0.098
Token found
Token ID1043
Feature activation-9.38e-03

Pos logit contributions
 Along+0.311
 nights+0.305
way+0.289
 On+0.283
dark+0.279
Neg logit contributions
 paramed-0.188
 mean-0.183
boys-0.181
 liar-0.179
 mixer-0.178
Token the
Token ID262
Feature activation+7.42e-03

Pos logit contributions
 Along+0.094
 nights+0.094
way+0.086
 On+0.084
 drifted+0.084
Neg logit contributions
 paramed-0.084
 liar-0.082
 mixer-0.082
 wasteful-0.080
 mean-0.079
Token mouse
Token ID10211
Feature activation+9.12e-02

Pos logit contributions
 paramed+0.010
 mean+0.009
 mixer+0.009
 liar+0.009
boys+0.009
Neg logit contributions
 Along-0.014
 nights-0.013
way-0.013
 drifted-0.013
 On-0.013
Token's
Token ID338
Feature activation+1.49e-01

Pos logit contributions
boys+0.143
 phone+0.139
 mean+0.138
 greedy+0.136
 stern+0.135
Neg logit contributions
 Along-0.141
 nights-0.136
way-0.131
ingle-0.129
 drifted-0.127
Token home
Token ID1363
Feature activation-5.35e-02

Pos logit contributions
 phone+0.214
boys+0.209
 mean+0.205
 paramed+0.199
 greedy+0.197
Neg logit contributions
 Along-0.258
dark-0.249
ingle-0.248
way-0.239
 nights-0.238
Token.
Token ID13
Feature activation-3.15e-02

Pos logit contributions
 Along+0.078
 nights+0.073
 On+0.070
way+0.070
 drifted+0.069
Neg logit contributions
boys-0.088
 phone-0.086
 mean-0.086
 mixer-0.085
 greedy-0.085
Token Along
Token ID17159
Feature activation-2.95e-01

Pos logit contributions
 nights+0.049
way+0.044
 outer+0.043
dark+0.043
ingle+0.042
Neg logit contributions
 paramed-0.051
boys-0.048
 stern-0.047
 greedy-0.047
 mixer-0.046
Token the
Token ID262
Feature activation+6.92e-03

Pos logit contributions
 Along+0.595
 nights+0.588
way+0.569
dark+0.562
more+0.550
Neg logit contributions
boys-0.304
 mixer-0.294
 paramed-0.286
 mean-0.279
 phone-0.274
Token way
Token ID835
Feature activation-7.63e-02

Pos logit contributions
 mean+0.012
boys+0.011
 phone+0.011
 quarrel+0.010
 mad+0.010
Neg logit contributions
ingle-0.010
dark-0.009
 Along-0.009
ment-0.009
powers-0.009
Token,
Token ID11
Feature activation+7.24e-02

Pos logit contributions
 Along+0.136
 nights+0.135
 drifted+0.127
dark+0.127
way+0.126
Neg logit contributions
 mixer-0.095
boys-0.092
 paramed-0.092
 mean-0.092
 phone-0.089
Token they
Token ID484
Feature activation-2.21e-02

Pos logit contributions
 paramed+0.105
boys+0.104
 mixer+0.103
 mean+0.102
 greedy+0.101
Neg logit contributions
 Along-0.120
 nights-0.117
way-0.113
dark-0.109
 drifted-0.107
Token met
Token ID1138
Feature activation-4.33e-02

Pos logit contributions
 Along+0.037
 nights+0.037
 drifted+0.034
way+0.033
 whenever+0.033
Neg logit contributions
 paramed-0.033
 mixer-0.033
 liar-0.033
 wasteful-0.032
 bluff-0.031
Token to
Token ID284
Feature activation+7.64e-02

Pos logit contributions
 nights+0.211
 Along+0.195
dark+0.194
 drifted+0.191
way+0.190
Neg logit contributions
boys-0.181
 mean-0.176
 phone-0.175
 paramed-0.171
 mixer-0.170
Token explore
Token ID7301
Feature activation-2.24e-01

Pos logit contributions
 paramed+0.100
 mixer+0.099
boys+0.099
 mean+0.098
 liar+0.098
Neg logit contributions
 nights-0.137
 Along-0.135
way-0.129
 drifted-0.124
dark-0.122
Token the
Token ID262
Feature activation+9.00e-03

Pos logit contributions
 Along+0.403
 nights+0.388
way+0.377
 drifted+0.369
ingle+0.363
Neg logit contributions
boys-0.286
 mean-0.284
 mixer-0.281
 stern-0.271
 greedy-0.269
Token forest
Token ID8222
Feature activation-1.31e-01

Pos logit contributions
 mean+0.016
boys+0.015
 phone+0.014
 greedy+0.014
roc+0.014
Neg logit contributions
ingle-0.013
 drifted-0.013
 Along-0.013
lights-0.012
ment-0.012
Token.
Token ID13
Feature activation-2.93e-02

Pos logit contributions
 Along+0.184
 nights+0.181
 drifted+0.173
dark+0.172
 On+0.163
Neg logit contributions
boys-0.223
 mean-0.210
 phone-0.209
 mixer-0.207
 greedy-0.206
Token Along
Token ID17159
Feature activation-2.95e-01

Pos logit contributions
 nights+0.043
ingle+0.039
way+0.038
 drifted+0.038
dark+0.037
Neg logit contributions
 paramed-0.048
boys-0.047
 greedy-0.046
 mixer-0.045
 mean-0.045
Token the
Token ID262
Feature activation+8.49e-03

Pos logit contributions
 nights+0.601
 Along+0.590
way+0.569
dark+0.568
ingle+0.559
Neg logit contributions
boys-0.294
 mixer-0.281
 mean-0.274
 paramed-0.270
roc-0.261
Token way
Token ID835
Feature activation-7.53e-02

Pos logit contributions
 mean+0.014
boys+0.013
 quarrel+0.012
roc+0.012
 rude+0.012
Neg logit contributions
ingle-0.013
dark-0.013
ment-0.013
ugg-0.012
 drifted-0.012
Token they
Token ID484
Feature activation-2.23e-02

Pos logit contributions
 nights+0.132
 Along+0.131
dark+0.126
 drifted+0.124
way+0.122
Neg logit contributions
 mixer-0.092
boys-0.092
 mean-0.091
 phone-0.088
 stern-0.087
Token talked
Token ID6619
Feature activation-1.67e-02

Pos logit contributions
 Along+0.042
 nights+0.042
way+0.039
 On+0.038
 drifted+0.038
Neg logit contributions
 paramed-0.027
 mean-0.027
 mixer-0.026
boys-0.026
 liar-0.026
Token about
Token ID546
Feature activation+1.20e-03

Pos logit contributions
 Along+0.025
 nights+0.024
way+0.023
 drifted+0.022
 On+0.022
Neg logit contributions
 paramed-0.028
 mixer-0.027
 liar-0.027
 mean-0.027
boys-0.026
Token to
Token ID284
Feature activation+7.64e-02

Pos logit contributions
 Along+0.050
 nights+0.048
way+0.046
 whenever+0.045
 On+0.043
Neg logit contributions
 paramed-0.069
 mixer-0.068
 liar-0.067
 wasteful-0.065
roc-0.065
Token go
Token ID467
Feature activation-6.82e-02

Pos logit contributions
 paramed+0.139
 liar+0.136
 mixer+0.136
 wasteful+0.134
 mean+0.133
Neg logit contributions
 Along-0.102
 nights-0.099
way-0.091
 On-0.088
 drifted-0.088
Token visit
Token ID3187
Feature activation-2.02e-01

Pos logit contributions
 nights+0.127
way+0.118
 Along+0.117
 drifted+0.115
lights+0.111
Neg logit contributions
 mean-0.090
 phone-0.087
boys-0.085
 paramed-0.082
 mad-0.081
Token Slip
Token ID49988
Feature activation-5.06e-02

Pos logit contributions
 Along+0.329
 nights+0.320
way+0.296
 On+0.292
 whenever+0.291
Neg logit contributions
 paramed-0.317
 liar-0.306
 mixer-0.305
 wasteful-0.301
 mean-0.296
Token.
Token ID13
Feature activation-3.31e-02

Pos logit contributions
 Along+0.066
 nights+0.065
way+0.060
 drifted+0.058
 On+0.058
Neg logit contributions
boys-0.091
 mean-0.089
 mixer-0.088
 greedy-0.088
 phone-0.088
Token Along
Token ID17159
Feature activation-2.94e-01

Pos logit contributions
 nights+0.052
 Along+0.045
way+0.045
 drifted+0.045
 days+0.044
Neg logit contributions
 paramed-0.055
 mixer-0.051
 mean-0.051
boys-0.051
 greedy-0.051
Token the
Token ID262
Feature activation+4.49e-03

Pos logit contributions
 nights+0.604
 Along+0.589
way+0.565
dark+0.552
more+0.544
Neg logit contributions
boys-0.307
 paramed-0.298
 mixer-0.296
 mean-0.294
 Lee-0.293
Token way
Token ID835
Feature activation-7.78e-02

Pos logit contributions
 mean+0.008
 phone+0.007
 mad+0.007
 greedy+0.007
boys+0.007
Neg logit contributions
 nights-0.006
 Along-0.006
dark-0.006
lights-0.006
ingle-0.006
Token she
Token ID673
Feature activation-1.62e-01

Pos logit contributions
 nights+0.147
 Along+0.146
dark+0.140
 drifted+0.137
way+0.134
Neg logit contributions
 mean-0.086
 mixer-0.086
 Lee-0.086
boys-0.085
 phone-0.085
Token met
Token ID1138
Feature activation-4.24e-02

Pos logit contributions
 nights+0.301
 Along+0.299
way+0.286
more+0.278
ingle+0.274
Neg logit contributions
 paramed-0.197
boys-0.194
 mean-0.193
 greedy-0.186
 mixer-0.184
Token a
Token ID257
Feature activation-2.84e-02

Pos logit contributions
 Along+0.069
 nights+0.068
way+0.063
 On+0.062
 whenever+0.061
Neg logit contributions
 paramed-0.065
 mixer-0.063
 liar-0.062
 wasteful-0.061
boys-0.060
Token make
Token ID787
Feature activation+7.73e-02

Pos logit contributions
 paramed+0.140
 mixer+0.137
 liar+0.137
 mean+0.136
boys+0.135
Neg logit contributions
 Along-0.103
 nights-0.101
way-0.093
 On-0.088
 drifted-0.088
Token his
Token ID465
Feature activation+6.52e-03

Pos logit contributions
 paramed+0.103
 mixer+0.101
 mean+0.100
boys+0.099
 mad+0.099
Neg logit contributions
 Along-0.142
 nights-0.134
way-0.130
 On-0.126
dark-0.124
Token way
Token ID835
Feature activation-7.61e-02

Pos logit contributions
 mixer+0.010
 mad+0.009
 phone+0.009
 mean+0.009
 paramed+0.009
Neg logit contributions
 Along-0.011
 nights-0.010
way-0.010
dark-0.010
 On-0.010
Token home
Token ID1363
Feature activation-5.08e-02

Pos logit contributions
 Along+0.156
 nights+0.156
dark+0.148
growth+0.147
 drifted+0.143
Neg logit contributions
 mean-0.082
 mad-0.082
 stern-0.075
 phone-0.075
 angry-0.073
Token.
Token ID13
Feature activation-3.04e-02

Pos logit contributions
 Along+0.068
 nights+0.063
way+0.059
 On+0.058
dark+0.058
Neg logit contributions
 paramed-0.093
boys-0.091
 mean-0.090
 mixer-0.090
 phone-0.090
Token Along
Token ID17159
Feature activation-2.94e-01

Pos logit contributions
 nights+0.046
lights+0.044
 outer+0.043
 drifted+0.042
dark+0.041
Neg logit contributions
 paramed-0.048
roc-0.044
 Lee-0.043
 mixer-0.042
 greedy-0.041
Token the
Token ID262
Feature activation+7.40e-03

Pos logit contributions
 Along+0.587
 nights+0.571
way+0.554
dark+0.548
more+0.538
Neg logit contributions
 paramed-0.312
 mixer-0.309
boys-0.300
 mean-0.293
 phone-0.288
Token way
Token ID835
Feature activation-7.62e-02

Pos logit contributions
 mean+0.012
 phone+0.010
 quarrel+0.010
 mad+0.010
boys+0.010
Neg logit contributions
dark-0.012
ingle-0.012
 Along-0.012
ment-0.011
edly-0.011
Token,
Token ID11
Feature activation+7.16e-02

Pos logit contributions
 Along+0.137
 nights+0.135
dark+0.128
 drifted+0.127
 On+0.125
Neg logit contributions
 mixer-0.094
 mean-0.093
 paramed-0.093
boys-0.090
 phone-0.090
Token he
Token ID339
Feature activation-9.57e-02

Pos logit contributions
 mean+0.106
 paramed+0.106
 mixer+0.103
 mad+0.103
 greedy+0.101
Neg logit contributions
 Along-0.112
 nights-0.111
dark-0.108
 drifted-0.105
way-0.102
Token met
Token ID1138
Feature activation-4.21e-02

Pos logit contributions
 nights+0.158
 Along+0.157
 drifted+0.145
way+0.144
 whenever+0.140
Neg logit contributions
 paramed-0.146
 liar-0.143
 mixer-0.142
 wasteful-0.140
roc-0.136
Token angel
Token ID18304
Feature activation-9.86e-03

Pos logit contributions
 mean+0.007
 phone+0.007
 mixer+0.007
 mad+0.007
 liar+0.007
Neg logit contributions
 Along-0.014
 nights-0.013
 On-0.013
way-0.012
 drifted-0.012
Token and
Token ID290
Feature activation-2.72e-02

Pos logit contributions
 Along+0.018
 nights+0.017
way+0.017
 On+0.016
 drifted+0.016
Neg logit contributions
boys-0.013
 phone-0.013
 mean-0.012
 stern-0.012
 mixer-0.012
Token followed
Token ID3940
Feature activation-9.12e-02

Pos logit contributions
 Along+0.045
 nights+0.042
way+0.041
 On+0.039
ingle+0.039
Neg logit contributions
 paramed-0.041
boys-0.040
 mean-0.040
 greedy-0.040
 phone-0.039
Token her
Token ID607
Feature activation-7.21e-02

Pos logit contributions
 Along+0.181
 nights+0.178
way+0.169
 drifted+0.166
 On+0.161
Neg logit contributions
 mean-0.103
 paramed-0.102
 mad-0.101
 phone-0.100
boys-0.099
Token.
Token ID13
Feature activation-2.99e-02

Pos logit contributions
dark+0.141
 Along+0.140
 nights+0.134
way+0.133
ingle+0.132
Neg logit contributions
 phone-0.094
boys-0.088
 stern-0.084
 mean-0.080
 whining-0.079
Token Along
Token ID17159
Feature activation-2.94e-01

Pos logit contributions
 nights+0.055
 drifted+0.050
ingle+0.049
iest+0.049
 days+0.049
Neg logit contributions
 paramed-0.040
 stern-0.037
 greedy-0.037
 Lee-0.036
boys-0.036
Token the
Token ID262
Feature activation+6.33e-03

Pos logit contributions
 Along+0.642
 nights+0.622
way+0.611
dark+0.602
more+0.597
Neg logit contributions
boys-0.249
 mixer-0.244
 mean-0.238
 phone-0.234
 Lee-0.233
Token way
Token ID835
Feature activation-7.57e-02

Pos logit contributions
 mean+0.010
boys+0.009
 phone+0.009
 quarrel+0.008
 mad+0.008
Neg logit contributions
 Along-0.010
ingle-0.010
ugg-0.009
dark-0.009
ment-0.009
Token,
Token ID11
Feature activation+7.13e-02

Pos logit contributions
 Along+0.150
 nights+0.147
 drifted+0.143
dark+0.140
more+0.139
Neg logit contributions
 mixer-0.079
 mean-0.076
 phone-0.073
 paramed-0.071
boys-0.071
Token the
Token ID262
Feature activation+6.63e-03

Pos logit contributions
 mixer+0.117
 mean+0.116
boys+0.115
 paramed+0.115
 greedy+0.113
Neg logit contributions
 Along-0.104
way-0.098
 nights-0.098
ingle-0.096
dark-0.091
Token angel
Token ID18304
Feature activation-1.03e-02

Pos logit contributions
 mean+0.007
 phone+0.006
 mad+0.006
 greedy+0.005
 mixer+0.005
Neg logit contributions
 Along-0.016
ingle-0.015
 On-0.015
ge-0.014
way-0.014
Token outside
Token ID2354
Feature activation-1.46e-01

Pos logit contributions
 Along+0.192
 nights+0.190
way+0.176
 drifted+0.171
 On+0.169
Neg logit contributions
 paramed-0.197
 mixer-0.193
 liar-0.190
 mean-0.189
boys-0.188
Token and
Token ID290
Feature activation-2.34e-02

Pos logit contributions
 Along+0.239
 nights+0.238
way+0.227
dark+0.220
 drifted+0.216
Neg logit contributions
 mixer-0.211
 paramed-0.209
 mad-0.201
 mean-0.201
 phone-0.201
Token started
Token ID2067
Feature activation-2.83e-02

Pos logit contributions
 Along+0.040
 nights+0.039
way+0.035
 drifted+0.035
 On+0.035
Neg logit contributions
 paramed-0.036
 liar-0.035
 wasteful-0.034
roc-0.034
 mixer-0.033
Token walking
Token ID6155
Feature activation-1.83e-01

Pos logit contributions
 Along+0.039
 nights+0.038
 whenever+0.035
 drifted+0.035
 night+0.034
Neg logit contributions
 liar-0.051
roc-0.050
 paramed-0.050
 wasteful-0.050
 mean-0.048
Token.
Token ID13
Feature activation-2.63e-02

Pos logit contributions
 Along+0.246
 nights+0.240
 drifted+0.227
dark+0.223
way+0.220
Neg logit contributions
 mean-0.318
boys-0.312
 mixer-0.310
 mad-0.310
 paramed-0.309
Token Along
Token ID17159
Feature activation-2.94e-01

Pos logit contributions
 nights+0.044
ge+0.041
ingle+0.040
 drifted+0.040
 lifetime+0.040
Neg logit contributions
 mixer-0.039
 paramed-0.038
 Lee-0.036
 phone-0.035
 stern-0.034
Token the
Token ID262
Feature activation+5.97e-03

Pos logit contributions
 Along+0.623
 nights+0.607
way+0.582
dark+0.579
more+0.572
Neg logit contributions
 mixer-0.281
boys-0.274
 mean-0.266
 paramed-0.264
 phone-0.258
Token way
Token ID835
Feature activation-7.55e-02

Pos logit contributions
 mean+0.011
 phone+0.010
 mixer+0.010
 paramed+0.010
 mad+0.010
Neg logit contributions
dark-0.008
 Along-0.008
 nights-0.008
 drifted-0.007
ingle-0.007
Token,
Token ID11
Feature activation+7.29e-02

Pos logit contributions
 Along+0.145
 nights+0.142
 drifted+0.136
dark+0.135
more+0.133
Neg logit contributions
 mixer-0.087
 mean-0.083
 phone-0.080
 paramed-0.079
boys-0.079
Token she
Token ID673
Feature activation-1.57e-01

Pos logit contributions
 mixer+0.125
boys+0.122
 mean+0.120
 paramed+0.119
 phone+0.118
Neg logit contributions
 Along-0.102
way-0.094
ingle-0.093
 nights-0.093
dark-0.092
Token ate
Token ID15063
Feature activation-2.55e-02

Pos logit contributions
 nights+0.256
 Along+0.254
 drifted+0.243
 night+0.231
 whenever+0.230
Neg logit contributions
 paramed-0.245
 liar-0.238
 wasteful-0.234
 bluff-0.232
 mixer-0.232
Token
Token ID198
Feature activation-1.06e-02

Pos logit contributions
 Along+0.014
 nights+0.014
way+0.013
 drifted+0.012
 On+0.012
Neg logit contributions
 paramed-0.023
 mixer-0.023
 liar-0.022
 mean-0.022
boys-0.022
TokenL
Token ID43
Feature activation-1.98e-02

Pos logit contributions
 Along+0.016
 nights+0.015
way+0.014
 On+0.014
 drifted+0.014
Neg logit contributions
 paramed-0.018
 mixer-0.017
 liar-0.017
 wasteful-0.017
boys-0.017
Tokenila
Token ID10102
Feature activation-3.47e-02

Pos logit contributions
 Along+0.044
 nights+0.043
way+0.039
 whenever+0.039
 On+0.039
Neg logit contributions
 paramed-0.020
 liar-0.019
 mean-0.019
 mixer-0.019
roc-0.019
Token and
Token ID290
Feature activation-1.07e-02

Pos logit contributions
 Along+0.043
 On+0.039
 drifted+0.039
 nights+0.037
way+0.037
Neg logit contributions
 paramed-0.068
 liar-0.068
 wasteful-0.065
roc-0.064
 mixer-0.064
Token Ben
Token ID3932
Feature activation+1.05e-01

Pos logit contributions
 Along+0.019
 nights+0.018
 drifted+0.018
 On+0.018
 night+0.018
Neg logit contributions
 paramed-0.014
 liar-0.014
 wasteful-0.014
 mixer-0.014
 chefs-0.013
Token are
Token ID389
Feature activation+1.43e-01

Pos logit contributions
 paramed+0.153
 liar+0.150
roc+0.147
 wasteful+0.146
 mixer+0.142
Neg logit contributions
 Along-0.182
 drifted-0.171
 On-0.168
 whenever-0.162
 night-0.161
Token happy
Token ID3772
Feature activation+4.21e-02

Pos logit contributions
roc+0.265
 paramed+0.247
 liar+0.241
 mixer+0.239
ution+0.232
Neg logit contributions
 nights-0.187
 Along-0.186
 drifted-0.183
 whenever-0.181
 night-0.180
Token.
Token ID13
Feature activation-2.06e-02

Pos logit contributions
 mixer+0.072
 paramed+0.072
 mean+0.072
 mad+0.071
 chefs+0.071
Neg logit contributions
 Along-0.061
 nights-0.058
way-0.055
dark-0.052
 On-0.052
Token They
Token ID1119
Feature activation+4.82e-02

Pos logit contributions
 Along+0.030
 nights+0.029
way+0.027
 On+0.026
 drifted+0.026
Neg logit contributions
 paramed-0.035
 liar-0.035
 mixer-0.035
 mean-0.034
 wasteful-0.034
Token say
Token ID910
Feature activation+7.87e-02

Pos logit contributions
 paramed+0.066
 mixer+0.064
 liar+0.064
 wasteful+0.062
boys+0.062
Neg logit contributions
 Along-0.087
 nights-0.083
 On-0.078
 drifted-0.078
way-0.078
Token,
Token ID11
Feature activation+8.95e-02

Pos logit contributions
roc+0.174
 wasteful+0.171
 bluff+0.166
boys+0.163
 cigarette+0.162
Neg logit contributions
 whenever-0.075
 days-0.071
 of-0.070
 Along-0.067
 day-0.066
Token hurt
Token ID5938
Feature activation+9.89e-02

Pos logit contributions
 Along+0.115
 nights+0.113
way+0.106
 On+0.103
 drifted+0.102
Neg logit contributions
 paramed-0.090
 liar-0.087
 mixer-0.087
 mean-0.086
boys-0.085
Token its
Token ID663
Feature activation-1.11e-02

Pos logit contributions
 paramed+0.176
 liar+0.166
 mixer+0.164
 bluff+0.163
 wasteful+0.162
Neg logit contributions
 nights-0.137
 whenever-0.129
 Along-0.128
 night-0.128
 days-0.126
Token wing
Token ID8539
Feature activation-8.34e-03

Pos logit contributions
 Along+0.020
 nights+0.020
way+0.019
 On+0.018
 drifted+0.018
Neg logit contributions
 paramed-0.015
 mixer-0.015
 liar-0.015
 mean-0.015
 greedy-0.015
Token.
Token ID13
Feature activation-2.38e-02

Pos logit contributions
 Along+0.010
 nights+0.010
way+0.008
 On+0.008
 drifted+0.008
Neg logit contributions
 paramed-0.016
 mixer-0.016
 mean-0.016
 liar-0.016
boys-0.016
Token Lily
Token ID20037
Feature activation+3.14e-02

Pos logit contributions
 Along+0.035
 On+0.032
 nights+0.030
way+0.030
 As+0.029
Neg logit contributions
 liar-0.043
boys-0.042
 paramed-0.041
 bluff-0.041
roc-0.041
Token's
Token ID338
Feature activation+1.51e-01

Pos logit contributions
 paramed+0.053
 liar+0.053
 wasteful+0.052
 bluff+0.051
 mixer+0.051
Neg logit contributions
 Along-0.046
 nights-0.044
 drifted-0.043
way-0.042
 On-0.041
Token mom
Token ID1995
Feature activation+8.10e-02

Pos logit contributions
 paramed+0.207
 wasteful+0.203
 liar+0.202
boys+0.202
roc+0.201
Neg logit contributions
 nights-0.261
 Along-0.257
 days-0.238
 whenever-0.237
way-0.237
Tokenmy
Token ID1820
Feature activation-5.28e-02

Pos logit contributions
 liar+0.148
 paramed+0.143
 bluff+0.141
 wasteful+0.140
 mixer+0.139
Neg logit contributions
 Along-0.109
 nights-0.107
way-0.099
 whenever-0.097
 drifted-0.096
Token picked
Token ID6497
Feature activation-1.82e-02

Pos logit contributions
 drifted+0.074
 Along+0.074
 nights+0.069
 whenever+0.069
way+0.067
Neg logit contributions
 liar-0.093
 bluff-0.091
 wasteful-0.089
roc-0.089
 chefs-0.088
Token up
Token ID510
Feature activation-4.64e-02

Pos logit contributions
 Along+0.021
 whenever+0.021
 night+0.020
 nights+0.019
 On+0.019
Neg logit contributions
 liar-0.037
 paramed-0.035
 wasteful-0.035
roc-0.034
 bluff-0.033
Token the
Token ID262
Feature activation+8.99e-03

Pos logit contributions
 whenever+0.048
 Along+0.043
 days+0.043
 night+0.042
 nights+0.039
Neg logit contributions
 liar-0.103
 wasteful-0.098
 bluff-0.097
 disrespectful-0.095
 paramed-0.095
Token's
Token ID338
Feature activation+1.51e-01

Pos logit contributions
 Along+0.085
 nights+0.084
way+0.078
 drifted+0.078
 On+0.076
Neg logit contributions
 paramed-0.072
 mixer-0.070
 liar-0.070
 wasteful-0.068
roc-0.067
Token friend
Token ID1545
Feature activation-1.29e-02

Pos logit contributions
 paramed+0.244
 wasteful+0.240
 liar+0.237
 mixer+0.235
roc+0.233
Neg logit contributions
 Along-0.227
 nights-0.224
 days-0.201
way-0.200
 drifted-0.199
Token,
Token ID11
Feature activation+7.23e-02

Pos logit contributions
 Along+0.015
 nights+0.015
 whenever+0.015
 night+0.014
 day+0.014
Neg logit contributions
 wasteful-0.024
 paramed-0.024
 bluff-0.024
 mixer-0.023
 liar-0.023
Token a
Token ID257
Feature activation-2.64e-02

Pos logit contributions
 wasteful+0.121
 mixer+0.121
 paramed+0.121
 whining+0.114
 liar+0.114
Neg logit contributions
 Along-0.101
 whenever-0.096
 night-0.094
 On-0.093
 day-0.092
Token little
Token ID1310
Feature activation-7.49e-02

Pos logit contributions
 day+0.028
 night+0.025
 Along+0.024
 journey+0.022
 evening+0.022
Neg logit contributions
 wasteful-0.053
roc-0.052
boys-0.052
ution-0.052
 bluff-0.051
Token pig
Token ID12967
Feature activation+1.55e-01

Pos logit contributions
 day+0.099
 night+0.098
 of+0.098
 Along+0.093
 nights+0.091
Neg logit contributions
 bluff-0.132
 wasteful-0.130
 liar-0.122
 disrespectful-0.121
ige-0.121
Token named
Token ID3706
Feature activation-5.03e-02

Pos logit contributions
 wasteful+0.230
 bluff+0.227
 paramed+0.226
 mixer+0.222
 liar+0.220
Neg logit contributions
 Along-0.258
 nights-0.248
way-0.236
 whenever-0.232
 drifted-0.231
Token Pink
Token ID14657
Feature activation+8.21e-03

Pos logit contributions
 Along+0.078
 nights+0.071
 On+0.070
 whenever+0.069
way+0.067
Neg logit contributions
 paramed-0.080
 wasteful-0.080
 mixer-0.079
 liar-0.079
 mad-0.077
Tokeny
Token ID88
Feature activation+1.44e-03

Pos logit contributions
boys+0.013
 mean+0.013
 paramed+0.012
 mixer+0.012
 liar+0.012
Neg logit contributions
 Along-0.013
 nights-0.013
way-0.012
 On-0.011
 drifted-0.011
Token,
Token ID11
Feature activation+7.14e-02

Pos logit contributions
 paramed+0.003
 wasteful+0.003
 liar+0.003
 mixer+0.003
 bluff+0.003
Neg logit contributions
 Along-0.002
 nights-0.002
way-0.002
 On-0.002
 whenever-0.002
Token came
Token ID1625
Feature activation+4.09e-02

Pos logit contributions
 paramed+0.108
 mixer+0.106
 liar+0.105
 wasteful+0.104
roc+0.103
Neg logit contributions
 Along-0.117
 nights-0.111
 On-0.105
way-0.104
 drifted-0.104
Token.
Token ID13
Feature activation-2.24e-02

Pos logit contributions
 mean+0.037
 mixer+0.037
 greedy+0.036
 stern+0.036
 liar+0.036
Neg logit contributions
 Along-0.033
 nights-0.030
dark-0.029
 On-0.028
 drifted-0.028
Token Together
Token ID17083
Feature activation-5.30e-02

Pos logit contributions
 Along+0.030
 nights+0.028
 On+0.026
way+0.026
 whenever+0.025
Neg logit contributions
 paramed-0.042
 liar-0.042
 mixer-0.041
 mean-0.040
boys-0.040
Token they
Token ID484
Feature activation-1.80e-02

Pos logit contributions
 Along+0.084
 nights+0.083
way+0.077
 On+0.074
 drifted+0.074
Neg logit contributions
 paramed-0.082
 mixer-0.080
 liar-0.079
 mean-0.079
boys-0.078
Token loaded
Token ID9639
Feature activation+8.69e-02

Pos logit contributions
 Along+0.029
 nights+0.028
way+0.026
 drifted+0.026
 On+0.025
Neg logit contributions
 paramed-0.028
 liar-0.028
 mixer-0.027
 wasteful-0.027
 bluff-0.027
Token the
Token ID262
Feature activation+1.06e-02

Pos logit contributions
 liar+0.170
 paramed+0.167
 wasteful+0.154
 mean+0.152
 greedy+0.151
Neg logit contributions
 Along-0.117
 nights-0.115
 whenever-0.108
way-0.104
 days-0.104
Token truck
Token ID7779
Feature activation+1.61e-01

Pos logit contributions
 liar+0.011
boys+0.011
 paramed+0.011
 wasteful+0.010
 bluff+0.010
Neg logit contributions
 Along-0.023
 nights-0.022
way-0.021
 days-0.021
 day-0.020
Token with
Token ID351
Feature activation-2.59e-02

Pos logit contributions
boys+0.242
 mixer+0.241
 phone+0.239
 paramed+0.235
 stern+0.227
Neg logit contributions
 nights-0.255
 Along-0.243
 drifted-0.239
dark-0.231
 outer-0.228
Token the
Token ID262
Feature activation+1.03e-02

Pos logit contributions
 nights+0.040
 Along+0.038
 days+0.036
 night+0.036
 whenever+0.035
Neg logit contributions
 paramed-0.043
 liar-0.043
 wasteful-0.041
roc-0.041
 bluff-0.040
Token boxes
Token ID10559
Feature activation+7.32e-02

Pos logit contributions
boys+0.012
 paramed+0.012
 liar+0.012
 wasteful+0.012
 bluff+0.012
Neg logit contributions
 Along-0.020
 nights-0.020
way-0.019
 days-0.018
 night-0.018
Token and
Token ID290
Feature activation-2.17e-02

Pos logit contributions
 mixer+0.114
 paramed+0.110
 stern+0.109
 phone+0.108
boys+0.108
Neg logit contributions
 nights-0.113
 Along-0.112
dark-0.102
 drifted-0.101
way-0.100
Token Pete
Token ID17542
Feature activation+3.21e-02

Pos logit contributions
 Along+0.034
 nights+0.034
way+0.031
 drifted+0.030
 On+0.030
Neg logit contributions
 paramed-0.034
 liar-0.034
 mixer-0.033
 mean-0.033
 wasteful-0.032
Token,"
Token ID553
Feature activation+1.31e-01

Pos logit contributions
 paramed+0.056
 liar+0.054
 mixer+0.054
 wasteful+0.053
 mean+0.053
Neg logit contributions
 Along-0.057
 nights-0.057
way-0.053
 drifted-0.051
 On-0.050
Token Max
Token ID5436
Feature activation+8.97e-02

Pos logit contributions
 liar+0.203
 chefs+0.202
 paramed+0.200
 wasteful+0.198
roc+0.196
Neg logit contributions
 Along-0.196
 drifted-0.193
 whenever-0.192
 night-0.192
 nights-0.189
Token said
Token ID531
Feature activation+6.29e-02

Pos logit contributions
 wasteful+0.164
 mixer+0.164
 liar+0.162
 chefs+0.160
 paramed+0.157
Neg logit contributions
 Along-0.114
 As-0.113
 On-0.108
 whenever-0.108
 drifted-0.107
Token,
Token ID11
Feature activation+8.62e-02

Pos logit contributions
 wasteful+0.125
roc+0.125
 paramed+0.124
 mixer+0.124
 liar+0.123
Neg logit contributions
 Along-0.069
 nights-0.066
 whenever-0.065
way-0.064
 drifted-0.060
Token feeling
Token ID4203
Feature activation-4.28e-02

Pos logit contributions
roc+0.130
 liar+0.122
 wasteful+0.121
 paramed+0.118
 mixer+0.117
Neg logit contributions
 Along-0.150
 drifted-0.138
 nights-0.136
 whenever-0.134
 On-0.133
Token sorry
Token ID7926
Feature activation+1.96e-01

Pos logit contributions
 nights+0.083
 whenever+0.083
 drifted+0.082
 night+0.079
 Along+0.078
Neg logit contributions
 mixer-0.049
roc-0.048
 paramed-0.047
boys-0.047
 Lee-0.042
Token.
Token ID13
Feature activation-1.84e-02

Pos logit contributions
 mean+0.317
 mad+0.296
 stern+0.292
 sorry+0.291
 phone+0.289
Neg logit contributions
 Along-0.334
 nights-0.319
way-0.293
 outer-0.289
growth-0.288
Token
Token ID198
Feature activation-1.35e-02

Pos logit contributions
 nights+0.024
 Along+0.023
way+0.021
 drifted+0.020
ingle+0.020
Neg logit contributions
 paramed-0.035
 liar-0.034
 mean-0.033
 mixer-0.033
 greedy-0.033
Token
Token ID198
Feature activation-1.26e-02

Pos logit contributions
 nights+0.022
 Along+0.021
 drifted+0.019
way+0.019
filled+0.019
Neg logit contributions
 paramed-0.020
 stern-0.020
 mean-0.020
 mad-0.020
 mixer-0.019
TokenThe
Token ID464
Feature activation-4.16e-02

Pos logit contributions
 Along+0.020
 nights+0.020
way+0.018
 drifted+0.018
 On+0.018
Neg logit contributions
 paramed-0.020
 liar-0.019
 mixer-0.019
 mean-0.019
boys-0.019
Token dog
Token ID3290
Feature activation+1.24e-01

Pos logit contributions
 Along+0.067
 nights+0.067
way+0.062
 drifted+0.060
 days+0.060
Neg logit contributions
 paramed-0.063
 mixer-0.060
 liar-0.060
boys-0.060
 wasteful-0.059
Token new
Token ID649
Feature activation-1.17e-02

Pos logit contributions
 paramed+0.019
 liar+0.019
roc+0.018
boys+0.018
 mixer+0.018
Neg logit contributions
way-0.010
 nights-0.010
 Along-0.009
 whenever-0.009
 day-0.009
Token and
Token ID290
Feature activation-9.42e-03

Pos logit contributions
 Along+0.015
 nights+0.014
way+0.013
 drifted+0.013
 On+0.012
Neg logit contributions
 paramed-0.022
 mixer-0.022
 liar-0.022
 mean-0.022
boys-0.021
Token had
Token ID550
Feature activation-5.59e-02

Pos logit contributions
 nights+0.015
 Along+0.014
 whenever+0.013
 days+0.013
way+0.013
Neg logit contributions
 paramed-0.016
 mixer-0.015
 liar-0.015
roc-0.014
boys-0.014
Token fun
Token ID1257
Feature activation+9.71e-04

Pos logit contributions
 nights+0.083
 Along+0.077
 days+0.074
 drifted+0.073
way+0.072
Neg logit contributions
 paramed-0.099
 mixer-0.094
roc-0.091
 liar-0.091
 wasteful-0.089
Token.
Token ID13
Feature activation-1.44e-02

Pos logit contributions
 mean+0.002
 mixer+0.002
 phone+0.002
 mad+0.002
 chefs+0.001
Neg logit contributions
 Along-0.002
 nights-0.001
way-0.001
 On-0.001
ingle-0.001
Token They
Token ID1119
Feature activation+4.71e-02

Pos logit contributions
 nights+0.024
 Along+0.023
way+0.022
ingle+0.021
 drifted+0.021
Neg logit contributions
 paramed-0.022
 greedy-0.021
 mixer-0.021
 liar-0.020
 stern-0.020
Token looked
Token ID3114
Feature activation-1.11e-02

Pos logit contributions
 paramed+0.074
 mixer+0.073
 liar+0.071
roc+0.071
 wasteful+0.070
Neg logit contributions
 Along-0.074
 nights-0.071
 drifted-0.070
 On-0.067
 whenever-0.066
Token at
Token ID379
Feature activation+1.02e-02

Pos logit contributions
 Along+0.016
 nights+0.015
way+0.014
 whenever+0.014
 On+0.014
Neg logit contributions
 paramed-0.020
 liar-0.019
 mixer-0.019
boys-0.018
roc-0.018
Token each
Token ID1123
Feature activation+8.13e-02

Pos logit contributions
 paramed+0.013
 liar+0.013
 mixer+0.012
boys+0.012
 wasteful+0.012
Neg logit contributions
 nights-0.019
 Along-0.019
 whenever-0.018
 night-0.017
way-0.017
Token other
Token ID584
Feature activation-8.40e-04

Pos logit contributions
 mean+0.111
 mixer+0.110
 paramed+0.108
 phone+0.108
 liar+0.108
Neg logit contributions
 Along-0.144
 nights-0.141
way-0.133
 drifted-0.131
 On-0.129
Token and
Token ID290
Feature activation-1.10e-02

Pos logit contributions
 Along+0.001
 nights+0.001
way+0.001
 drifted+0.001
 On+0.001
Neg logit contributions
 paramed-0.002
 liar-0.002
 mixer-0.001
 wasteful-0.001
 mean-0.001
Token laugh
Token ID6487
Feature activation+8.83e-02

Pos logit contributions
 Along+0.016
 nights+0.016
way+0.015
 drifted+0.015
 On+0.015
Neg logit contributions
 paramed-0.015
 mixer-0.015
 liar-0.015
 mean-0.014
boys-0.014
Token still
Token ID991
Feature activation+4.53e-03

Pos logit contributions
 paramed+0.158
 mixer+0.154
 liar+0.154
 mean+0.152
boys+0.151
Neg logit contributions
 Along-0.120
 nights-0.117
way-0.108
 On-0.103
 drifted-0.103
Token rang
Token ID28077
Feature activation-6.51e-02

Pos logit contributions
 paramed+0.008
 mixer+0.007
 liar+0.007
 mean+0.007
boys+0.007
Neg logit contributions
 Along-0.007
 nights-0.007
way-0.006
 On-0.006
 drifted-0.006
Token in
Token ID287
Feature activation-5.74e-02

Pos logit contributions
 Along+0.086
 nights+0.086
way+0.078
 whenever+0.075
 drifted+0.075
Neg logit contributions
 paramed-0.118
 liar-0.115
 mixer-0.115
 wasteful-0.113
boys-0.112
Token his
Token ID465
Feature activation+1.50e-02

Pos logit contributions
 Along+0.096
 nights+0.094
way+0.088
 On+0.085
 drifted+0.085
Neg logit contributions
 paramed-0.086
 mixer-0.083
 liar-0.083
boys-0.081
 wasteful-0.081
Token ears
Token ID11368
Feature activation+4.85e-02

Pos logit contributions
 mean+0.021
 paramed+0.021
 mixer+0.021
 phone+0.021
 greedy+0.021
Neg logit contributions
 Along-0.026
 nights-0.025
way-0.025
 On-0.024
ingle-0.023
Token as
Token ID355
Feature activation-1.83e-01

Pos logit contributions
 paramed+0.083
 mixer+0.082
 mean+0.081
 liar+0.081
boys+0.080
Neg logit contributions
 Along-0.068
 nights-0.067
way-0.061
 On-0.059
 drifted-0.059
Token he
Token ID339
Feature activation-8.84e-02

Pos logit contributions
 Along+0.284
 nights+0.264
way+0.251
 On+0.243
ingle+0.239
Neg logit contributions
 mean-0.298
 liar-0.295
 greedy-0.290
 mixer-0.286
 stern-0.283
Token vanished
Token ID23717
Feature activation-4.47e-02

Pos logit contributions
 Along+0.154
 nights+0.151
way+0.144
 On+0.137
dark+0.134
Neg logit contributions
boys-0.122
 mean-0.122
 greedy-0.119
 liar-0.118
 paramed-0.117
Token into
Token ID656
Feature activation-6.85e-02

Pos logit contributions
 Along+0.064
 nights+0.060
 On+0.055
way+0.055
dark+0.054
Neg logit contributions
 mean-0.077
boys-0.075
 mad-0.075
 phone-0.075
 stern-0.075
Token the
Token ID262
Feature activation+1.74e-02

Pos logit contributions
 Along+0.132
 nights+0.123
way+0.122
 On+0.117
dark+0.116
Neg logit contributions
 mean-0.085
 liar-0.084
 paramed-0.084
 mixer-0.083
 mad-0.082
Token they
Token ID484
Feature activation-1.97e-02

Pos logit contributions
 paramed+0.133
 mixer+0.125
 liar+0.125
roc+0.123
boys+0.120
Neg logit contributions
 Along-0.103
 whenever-0.096
 nights-0.096
way-0.095
 On-0.091
Token saw
Token ID2497
Feature activation+3.68e-02

Pos logit contributions
 Along+0.027
 drifted+0.026
 nights+0.026
 whenever+0.025
 night+0.025
Neg logit contributions
 paramed-0.036
 bluff-0.033
 liar-0.033
 mixer-0.033
roc-0.032
Token a
Token ID257
Feature activation-2.60e-02

Pos logit contributions
 paramed+0.077
 liar+0.072
roc+0.071
 mixer+0.071
 whining+0.071
Neg logit contributions
 nights-0.037
 whenever-0.037
way-0.036
 days-0.036
 night-0.035
Token sign
Token ID1051
Feature activation+7.74e-02

Pos logit contributions
 day+0.030
way+0.028
 night+0.028
 Along+0.027
 journey+0.026
Neg logit contributions
boys-0.051
 paramed-0.049
eper-0.047
 chefs-0.047
 bluff-0.046
Token that
Token ID326
Feature activation-3.65e-02

Pos logit contributions
 paramed+0.144
roc+0.143
 liar+0.142
 wasteful+0.138
boys+0.137
Neg logit contributions
 Along-0.098
 nights-0.094
way-0.092
 On-0.086
 whenever-0.084
Token said
Token ID531
Feature activation+5.28e-02

Pos logit contributions
 nights+0.058
 Along+0.057
way+0.056
 On+0.054
 night+0.054
Neg logit contributions
 paramed-0.057
 liar-0.055
roc-0.053
 mixer-0.053
 wasteful-0.051
Token,
Token ID11
Feature activation+7.18e-02

Pos logit contributions
 paramed+0.082
roc+0.081
 wasteful+0.079
boys+0.078
 whining+0.078
Neg logit contributions
 Along-0.079
 whenever-0.076
way-0.076
 On-0.074
 before-0.073
Token "
Token ID366
Feature activation+2.00e-01

Pos logit contributions
roc+0.129
 liar+0.125
boys+0.117
 wasteful+0.117
 paramed+0.115
Neg logit contributions
 Along-0.103
 On-0.098
 As-0.095
 before-0.092
 whenever-0.092
TokenDo
Token ID5211
Feature activation+3.56e-02

Pos logit contributions
 paramed+0.294
 mixer+0.285
 liar+0.284
 mean+0.280
boys+0.279
Neg logit contributions
 Along-0.337
 nights-0.332
way-0.310
 drifted-0.300
 On-0.300
Token not
Token ID407
Feature activation+6.35e-02

Pos logit contributions
 paramed+0.036
 mixer+0.034
 liar+0.034
boys+0.033
 wasteful+0.033
Neg logit contributions
 Along-0.077
 nights-0.076
way-0.072
 On-0.070
 drifted-0.070
Token eat
Token ID4483
Feature activation-3.83e-02

Pos logit contributions
 paramed+0.101
 mixer+0.098
 liar+0.098
boys+0.095
 wasteful+0.095
Neg logit contributions
 Along-0.100
 nights-0.097
way-0.091
 On-0.088
 drifted-0.087
Token wonderful
Token ID7932
Feature activation-4.28e-02

Pos logit contributions
 Along+0.048
 nights+0.045
way+0.043
 On+0.042
 drifted+0.042
Neg logit contributions
 mixer-0.045
 paramed-0.045
 mean-0.044
boys-0.043
 greedy-0.043
Token surprise
Token ID5975
Feature activation+3.88e-02

Pos logit contributions
 Along+0.076
 nights+0.073
way+0.069
 On+0.067
 drifted+0.067
Neg logit contributions
 paramed-0.060
 mixer-0.059
 liar-0.057
 mean-0.057
boys-0.056
Token �
Token ID6184
Feature activation-7.85e-03

Pos logit contributions
 paramed+0.070
 mixer+0.070
 mean+0.068
 liar+0.068
boys+0.068
Neg logit contributions
 Along-0.052
 nights-0.050
way-0.046
 drifted-0.044
 On-0.044
Token�
Token ID95
Feature activation+6.06e-02

Pos logit contributions
 nights+0.012
 Along+0.011
 drifted+0.011
 whenever+0.011
way+0.011
Neg logit contributions
 paramed-0.013
 mixer-0.012
boys-0.012
 liar-0.012
 phone-0.011
Token€
Token ID26391
Feature activation+7.72e-02

Pos logit contributions
 paramed+0.096
 mixer+0.093
 liar+0.092
boys+0.091
 mean+0.091
Neg logit contributions
 nights-0.094
 Along-0.093
way-0.085
 drifted-0.084
 whenever-0.083
Token�
Token ID447
Feature activation+6.21e-02

Pos logit contributions
 paramed+0.115
boys+0.112
 mixer+0.112
 liar+0.110
 mean+0.109
Neg logit contributions
 nights-0.126
 Along-0.125
way-0.116
 drifted-0.112
 On-0.111
Token�
Token ID250
Feature activation+1.62e-01

Pos logit contributions
 paramed+0.112
 liar+0.109
 mixer+0.108
 wasteful+0.107
 mean+0.107
Neg logit contributions
 Along-0.086
 nights-0.082
way-0.076
 On-0.074
 drifted-0.073
Token a
Token ID257
Feature activation-1.98e-02

Pos logit contributions
 paramed+0.215
 mixer+0.195
 greedy+0.188
 mean+0.187
 liar+0.186
Neg logit contributions
 nights-0.292
 Along-0.277
way-0.270
 drifted-0.267
lights-0.258
Token drawing
Token ID8263
Feature activation-3.12e-02

Pos logit contributions
 Along+0.029
 nights+0.029
 night+0.027
 day+0.027
way+0.026
Neg logit contributions
 wasteful-0.032
roc-0.031
boys-0.031
 paramed-0.031
 liar-0.030
Token from
Token ID422
Feature activation-2.91e-02

Pos logit contributions
 Along+0.043
 nights+0.042
way+0.039
 On+0.037
 drifted+0.037
Neg logit contributions
 paramed-0.055
 mixer-0.054
 liar-0.054
 mean-0.053
boys-0.052
Token a
Token ID257
Feature activation-2.05e-02

Pos logit contributions
 Along+0.054
 nights+0.053
way+0.050
 drifted+0.049
 On+0.048
Neg logit contributions
 paramed-0.037
 mixer-0.037
 liar-0.036
 mean-0.036
boys-0.035
Token Lucy
Token ID22162
Feature activation-1.69e-02

Pos logit contributions
 mixer+0.106
 paramed+0.104
 mean+0.101
boys+0.100
 chefs+0.100
Neg logit contributions
 Along-0.111
 nights-0.110
way-0.102
 drifted-0.102
ingle-0.100
Token saw
Token ID2497
Feature activation+3.46e-02

Pos logit contributions
 Along+0.031
 nights+0.030
way+0.028
 On+0.028
 drifted+0.028
Neg logit contributions
 paramed-0.023
 liar-0.022
 mixer-0.022
 wasteful-0.022
 mean-0.021
Token a
Token ID257
Feature activation-2.80e-02

Pos logit contributions
 liar+0.070
roc+0.070
boys+0.069
 whining+0.068
 bluff+0.068
Neg logit contributions
 nights-0.035
 whenever-0.034
 days-0.034
 Along-0.033
 night-0.032
Token big
Token ID1263
Feature activation-3.65e-03

Pos logit contributions
 day+0.029
 night+0.026
 journey+0.025
 nights+0.025
 path+0.025
Neg logit contributions
boys-0.057
roc-0.056
 wasteful-0.055
ishly-0.053
 paramed-0.053
Token red
Token ID2266
Feature activation+5.80e-02

Pos logit contributions
 day+0.005
 night+0.005
 journey+0.005
 view+0.004
 days+0.004
Neg logit contributions
roc-0.006
Fighting-0.006
 bluff-0.006
 wasteful-0.006
boys-0.006
Token ball
Token ID2613
Feature activation+5.86e-02

Pos logit contributions
 wasteful+0.079
 bluff+0.077
boys+0.076
 paramed+0.075
 liar+0.075
Neg logit contributions
 Along-0.104
 nights-0.097
way-0.097
 night-0.093
 On-0.093
Token.
Token ID13
Feature activation-3.47e-02

Pos logit contributions
 mixer+0.104
 mean+0.102
 paramed+0.099
 phone+0.099
 greedy+0.097
Neg logit contributions
 nights-0.086
 Along-0.081
 drifted-0.073
more-0.070
way-0.070
Token She
Token ID1375
Feature activation-6.65e-02

Pos logit contributions
 Along+0.049
 nights+0.049
way+0.045
 drifted+0.043
more+0.042
Neg logit contributions
 paramed-0.060
 mixer-0.059
 liar-0.058
 mean-0.057
boys-0.057
Token wanted
Token ID2227
Feature activation-2.02e-02

Pos logit contributions
 Along+0.105
 nights+0.100
way+0.094
 On+0.093
 drifted+0.092
Neg logit contributions
 paramed-0.108
 liar-0.106
 wasteful-0.104
 mixer-0.102
 bluff-0.102
Token to
Token ID284
Feature activation+7.63e-02

Pos logit contributions
 Along+0.020
 nights+0.019
 whenever+0.019
 days+0.018
 night+0.018
Neg logit contributions
 paramed-0.045
 liar-0.045
roc-0.044
 wasteful-0.043
 bluff-0.043
Token play
Token ID711
Feature activation-1.11e-01

Pos logit contributions
 paramed+0.124
 mixer+0.122
 mean+0.120
 liar+0.119
boys+0.117
Neg logit contributions
 nights-0.118
 Along-0.115
way-0.107
 drifted-0.104
more-0.100
Token.
Token ID13
Feature activation-2.72e-02

Pos logit contributions
 Along+0.049
 nights+0.048
way+0.043
 whenever+0.040
 On+0.040
Neg logit contributions
 paramed-0.103
 liar-0.100
 mixer-0.100
 wasteful-0.099
 mean-0.098
Token We
Token ID775
Feature activation+2.03e-02

Pos logit contributions
 nights+0.044
 Along+0.041
 drifted+0.040
way+0.039
ingle+0.038
Neg logit contributions
 paramed-0.041
boys-0.040
 mixer-0.040
 greedy-0.039
roc-0.039
Token can
Token ID460
Feature activation+1.26e-01

Pos logit contributions
 paramed+0.029
 liar+0.028
 mixer+0.028
 wasteful+0.027
 bluff+0.027
Neg logit contributions
 Along-0.036
 nights-0.034
way-0.032
 On-0.032
 whenever-0.032
Token restore
Token ID11169
Feature activation-3.17e-02

Pos logit contributions
 paramed+0.172
 mixer+0.169
boys+0.167
 mean+0.165
 liar+0.162
Neg logit contributions
 nights-0.227
 Along-0.223
way-0.208
 drifted-0.203
 days-0.199
Token it
Token ID340
Feature activation+3.25e-02

Pos logit contributions
 nights+0.044
 whenever+0.042
 Along+0.040
 night+0.040
way+0.040
Neg logit contributions
 paramed-0.060
 liar-0.057
 mixer-0.053
 whining-0.052
 wasteful-0.052
Token and
Token ID290
Feature activation-2.45e-02

Pos logit contributions
boys+0.047
 mean+0.044
 phone+0.044
 chefs+0.042
 mixer+0.041
Neg logit contributions
 Along-0.057
 nights-0.055
 drifted-0.054
way-0.051
 On-0.050
Token make
Token ID787
Feature activation+8.26e-02

Pos logit contributions
 Along+0.044
 nights+0.044
way+0.041
 drifted+0.040
 On+0.039
Neg logit contributions
 mean-0.033
boys-0.032
 paramed-0.032
 mixer-0.032
 phone-0.031
Token it
Token ID340
Feature activation+3.21e-02

Pos logit contributions
 paramed+0.133
 liar+0.129
 mixer+0.126
 wasteful+0.123
 greedy+0.121
Neg logit contributions
 nights-0.130
 Along-0.126
way-0.119
 whenever-0.117
 days-0.114
Token look
Token ID804
Feature activation+3.57e-02

Pos logit contributions
 mean+0.040
boys+0.036
 mad+0.035
 phone+0.035
 sorry+0.034
Neg logit contributions
 Along-0.068
 nights-0.065
way-0.062
growth-0.061
 On-0.060
Token like
Token ID588
Feature activation-5.87e-03

Pos logit contributions
 mean+0.047
 greedy+0.044
 mad+0.043
 wasteful+0.043
boys+0.042
Neg logit contributions
 Along-0.069
 nights-0.066
way-0.063
 drifted-0.062
 On-0.061
Token new
Token ID649
Feature activation-1.76e-02

Pos logit contributions
 Along+0.011
 nights+0.011
way+0.010
ingle+0.010
 On+0.010
Neg logit contributions
 mean-0.007
 mixer-0.007
 liar-0.007
 greedy-0.007
boys-0.007
Token family
Token ID1641
Feature activation-4.49e-02

Pos logit contributions
 Along+0.103
 nights+0.099
way+0.093
 On+0.091
 drifted+0.089
Neg logit contributions
 paramed-0.108
 mixer-0.106
 liar-0.105
 mean-0.104
boys-0.104
Token and
Token ID290
Feature activation-2.03e-02

Pos logit contributions
 nights+0.070
 Along+0.070
way+0.064
 drifted+0.063
 whenever+0.062
Neg logit contributions
 paramed-0.072
 liar-0.070
 mixer-0.069
 wasteful-0.069
 mean-0.068
Token garden
Token ID11376
Feature activation-8.99e-02

Pos logit contributions
 nights+0.032
 Along+0.031
way+0.028
 whenever+0.028
 days+0.028
Neg logit contributions
 paramed-0.033
 mixer-0.032
 liar-0.032
 wasteful-0.031
roc-0.030
Token heard
Token ID2982
Feature activation+3.34e-02

Pos logit contributions
 nights+0.127
 Along+0.122
way+0.116
 whenever+0.114
 drifted+0.112
Neg logit contributions
 paramed-0.160
 liar-0.156
 wasteful-0.152
 mixer-0.151
 bluff-0.150
Token her
Token ID607
Feature activation-6.70e-02

Pos logit contributions
 paramed+0.047
 mixer+0.045
 liar+0.045
 mean+0.045
boys+0.044
Neg logit contributions
 Along-0.059
 nights-0.057
way-0.054
 drifted-0.052
 On-0.052
Token beautiful
Token ID4950
Feature activation-3.12e-02

Pos logit contributions
 Along+0.116
 nights+0.115
way+0.106
 drifted+0.103
 On+0.103
Neg logit contributions
 paramed-0.094
boys-0.093
 mean-0.093
 mad-0.092
 phone-0.092
Token singing
Token ID13777
Feature activation-8.98e-02

Pos logit contributions
 Along+0.052
 nights+0.049
way+0.046
 On+0.046
 drifted+0.045
Neg logit contributions
 paramed-0.047
 mixer-0.047
 mean-0.046
 phone-0.045
boys-0.045
Token.
Token ID13
Feature activation-2.79e-02

Pos logit contributions
 Along+0.119
 nights+0.114
way+0.105
 On+0.102
 drifted+0.100
Neg logit contributions
 paramed-0.163
 mixer-0.160
 mean-0.160
boys-0.158
 liar-0.158
Token
Token ID220
Feature activation+4.55e-02

Pos logit contributions
 nights+0.040
 Along+0.037
 drifted+0.035
way+0.035
ingle+0.034
Neg logit contributions
 paramed-0.046
 mixer-0.046
 stern-0.045
 mean-0.045
 liar-0.045
Token
Token ID198
Feature activation-2.27e-02

Pos logit contributions
 paramed+0.077
boys+0.077
 mixer+0.076
 stern+0.074
 mean+0.074
Neg logit contributions
 nights-0.066
 Along-0.058
 drifted-0.057
 days-0.055
way-0.055
Token
Token ID198
Feature activation-2.23e-02

Pos logit contributions
 nights+0.031
 Along+0.029
 drifted+0.027
way+0.027
ingle+0.026
Neg logit contributions
 mean-0.039
 paramed-0.039
 stern-0.039
 mixer-0.039
 liar-0.038
Token.
Token ID13
Feature activation-3.22e-02

Pos logit contributions
 Along+0.069
 nights+0.067
 drifted+0.061
 On+0.060
dark+0.059
Neg logit contributions
boys-0.095
 phone-0.091
 mixer-0.091
 paramed-0.091
 mean-0.089
Token The
Token ID383
Feature activation-1.24e-02

Pos logit contributions
 nights+0.049
 Along+0.046
 drifted+0.044
way+0.044
more+0.043
Neg logit contributions
 paramed-0.054
 mixer-0.051
 greedy-0.050
 liar-0.050
 mean-0.050
Token ghost
Token ID10905
Feature activation-2.30e-02

Pos logit contributions
 Along+0.023
 nights+0.021
way+0.021
 On+0.020
dark+0.020
Neg logit contributions
 paramed-0.017
 mean-0.017
 mixer-0.017
 liar-0.017
 phone-0.017
Token was
Token ID373
Feature activation+2.16e-02

Pos logit contributions
 Along+0.036
 nights+0.035
way+0.032
 drifted+0.031
 On+0.031
Neg logit contributions
 paramed-0.037
 liar-0.036
 mixer-0.036
 wasteful-0.035
 mean-0.035
Token curious
Token ID11040
Feature activation+8.04e-02

Pos logit contributions
roc+0.036
 liar+0.035
 mixer+0.035
 paramed+0.034
boys+0.033
Neg logit contributions
 Along-0.033
 nights-0.033
 night-0.030
 day-0.030
 of-0.029
Token and
Token ID290
Feature activation-2.76e-02

Pos logit contributions
 paramed+0.140
 mixer+0.137
 liar+0.136
 mean+0.134
boys+0.134
Neg logit contributions
 Along-0.114
 nights-0.111
way-0.102
 drifted-0.098
 On-0.098
Token decided
Token ID3066
Feature activation-3.67e-02

Pos logit contributions
 Along+0.044
 nights+0.044
way+0.040
 drifted+0.039
 On+0.039
Neg logit contributions
 paramed-0.044
 mixer-0.042
 liar-0.042
boys-0.041
 wasteful-0.041
Token to
Token ID284
Feature activation+7.64e-02

Pos logit contributions
 Along+0.048
 nights+0.046
way+0.045
 whenever+0.043
more+0.041
Neg logit contributions
 paramed-0.069
 mixer-0.068
 liar-0.067
 wasteful-0.065
roc-0.065
Token explore
Token ID7301
Feature activation-2.24e-01

Pos logit contributions
 paramed+0.138
 mixer+0.135
 mean+0.135
 liar+0.134
boys+0.133
Neg logit contributions
 nights-0.101
 Along-0.101
way-0.091
 drifted-0.088
 On-0.086
Token.
Token ID13
Feature activation-3.27e-02

Pos logit contributions
 Along+0.367
 nights+0.355
way+0.332
 drifted+0.326
 On+0.320
Neg logit contributions
 paramed-0.333
 mixer-0.331
 mean-0.329
boys-0.324
 liar-0.323
Token As
Token ID1081
Feature activation-2.09e-01

Pos logit contributions
 nights+0.053
 drifted+0.049
ingle+0.047
lights+0.047
more+0.046
Neg logit contributions
 paramed-0.049
 mixer-0.046
 greedy-0.046
 Lee-0.045
 stern-0.045
Token for
Token ID329
Feature activation+1.74e-02

Pos logit contributions
 paramed+0.041
 liar+0.040
 mixer+0.039
boys+0.039
 mean+0.038
Neg logit contributions
 Along-0.026
 nights-0.026
way-0.023
 whenever-0.022
 drifted-0.022
Token the
Token ID262
Feature activation+1.59e-02

Pos logit contributions
 paramed+0.032
roc+0.030
 liar+0.030
boys+0.030
 mixer+0.028
Neg logit contributions
 nights-0.024
 days-0.022
 whenever-0.022
 Along-0.022
 night-0.021
Token third
Token ID2368
Feature activation+5.09e-02

Pos logit contributions
 paramed+0.020
 liar+0.019
boys+0.019
 mixer+0.019
 wasteful+0.019
Neg logit contributions
 nights-0.029
 Along-0.029
way-0.028
 drifted-0.027
 days-0.027
Token time
Token ID640
Feature activation-1.35e-01

Pos logit contributions
 mixer+0.061
 paramed+0.058
 mean+0.056
 liar+0.055
 greedy+0.054
Neg logit contributions
 Along-0.105
 nights-0.101
 drifted-0.095
 On-0.093
way-0.093
Token.
Token ID13
Feature activation-2.16e-02

Pos logit contributions
 Along+0.170
 nights+0.159
way+0.147
 On+0.144
 drifted+0.139
Neg logit contributions
 mixer-0.256
 paramed-0.254
 mean-0.254
 liar-0.253
 greedy-0.251
Token She
Token ID1375
Feature activation-4.96e-02

Pos logit contributions
 Along+0.027
 nights+0.026
 On+0.024
way+0.024
 drifted+0.023
Neg logit contributions
 paramed-0.041
 liar-0.040
 mixer-0.040
boys-0.040
 mean-0.039
Token wanted
Token ID2227
Feature activation-1.57e-02

Pos logit contributions
 drifted+0.076
 nights+0.068
 Along+0.068
way+0.065
dark+0.064
Neg logit contributions
 paramed-0.080
 bluff-0.079
roc-0.078
 liar-0.076
boys-0.072
Token water
Token ID1660
Feature activation-2.94e-02

Pos logit contributions
 nights+0.020
 whenever+0.019
 night+0.018
 days+0.018
 Along+0.018
Neg logit contributions
 paramed-0.030
 liar-0.029
boys-0.028
roc-0.028
 bluff-0.028
Token.
Token ID13
Feature activation-2.16e-02

Pos logit contributions
 whenever+0.039
 nights+0.036
 night+0.036
 day+0.035
 of+0.034
Neg logit contributions
 paramed-0.052
 liar-0.052
 bluff-0.051
roc-0.049
boys-0.049
Token She
Token ID1375
Feature activation-4.98e-02

Pos logit contributions
 Along+0.024
 nights+0.022
 On+0.021
way+0.021
 whenever+0.020
Neg logit contributions
 paramed-0.045
 liar-0.043
 mixer-0.043
boys-0.043
 wasteful-0.043
Token cried
Token ID16896
Feature activation-2.48e-03

Pos logit contributions
 drifted+0.078
 nights+0.074
 Along+0.074
way+0.069
 On+0.069
Neg logit contributions
 paramed-0.079
 bluff-0.075
roc-0.075
 liar-0.074
 mixer-0.073
TokenOnce
Token ID7454
Feature activation-1.26e-01

Pos logit contributions
 nights+0.155
 Along+0.142
way+0.138
 drifted+0.133
filled+0.131
Neg logit contributions
 mixer-0.152
 Lee-0.146
 mean-0.145
 chefs-0.144
 paramed-0.141
Token upon
Token ID2402
Feature activation-1.42e-01

Pos logit contributions
 nights+0.151
 Along+0.146
growth+0.133
 drifted+0.131
more+0.127
Neg logit contributions
 mean-0.249
 mad-0.246
 mixer-0.246
 stern-0.241
 sorry-0.237
Token a
Token ID257
Feature activation-2.43e-02

Pos logit contributions
 nights+0.336
 Along+0.331
way+0.327
 horizon+0.316
ingle+0.309
Neg logit contributions
 mean-0.128
 mad-0.111
 greedy-0.110
 liar-0.107
 chefs-0.106
Token time
Token ID640
Feature activation-1.36e-01

Pos logit contributions
 nights+0.024
 Along+0.024
way+0.022
dark+0.021
 drifted+0.020
Neg logit contributions
 mean-0.052
 greedy-0.050
 mad-0.050
 mixer-0.050
 liar-0.050
Token,
Token ID11
Feature activation+8.10e-02

Pos logit contributions
 nights+0.242
 Along+0.241
 drifted+0.222
way+0.220
dark+0.215
Neg logit contributions
 mixer-0.183
 mean-0.182
 paramed-0.181
 liar-0.177
 greedy-0.174
Token there
Token ID612
Feature activation-7.97e-02

Pos logit contributions
 mean+0.112
 mixer+0.107
 greedy+0.105
 mad+0.105
 stern+0.101
Neg logit contributions
 nights-0.150
 Along-0.144
 drifted-0.136
way-0.133
ingle-0.130
Token lived
Token ID5615
Feature activation-8.00e-02

Pos logit contributions
 Along+0.116
 nights+0.113
way+0.104
 drifted+0.100
 On+0.100
Neg logit contributions
 paramed-0.135
 mixer-0.133
 liar-0.131
 mean-0.130
boys-0.129
Token a
Token ID257
Feature activation-2.45e-02

Pos logit contributions
 Along+0.128
 nights+0.121
 On+0.112
way+0.112
 whenever+0.109
Neg logit contributions
 paramed-0.128
 liar-0.125
 mixer-0.124
 wasteful-0.124
boys-0.120
Token hero
Token ID4293
Feature activation+1.20e-02

Pos logit contributions
 Along+0.023
 nights+0.022
 day+0.020
 night+0.020
 days+0.019
Neg logit contributions
 paramed-0.052
 wasteful-0.052
roc-0.051
boys-0.050
 bluff-0.050
Token.
Token ID13
Feature activation-1.81e-02

Pos logit contributions
 liar+0.012
 mean+0.011
 greedy+0.011
 paramed+0.011
 phone+0.011
Neg logit contributions
 Along-0.026
 nights-0.026
 drifted-0.025
 On-0.024
way-0.024
Token He
Token ID679
Feature activation+3.23e-03

Pos logit contributions
 nights+0.033
 drifted+0.029
lights+0.029
way+0.029
 lifetime+0.028
Neg logit contributions
 Lee-0.025
 paramed-0.024
 mean-0.024
 liar-0.023
roc-0.023
Token and
Token ID290
Feature activation-2.64e-02

Pos logit contributions
 Along+0.063
 nights+0.060
way+0.055
 On+0.053
 drifted+0.053
Neg logit contributions
 paramed-0.083
 mixer-0.083
 liar-0.082
 mean-0.082
boys-0.081
Token fun
Token ID1257
Feature activation-1.01e-02

Pos logit contributions
 nights+0.039
 Along+0.039
way+0.035
 drifted+0.035
 On+0.034
Neg logit contributions
 paramed-0.044
 liar-0.043
 mixer-0.042
boys-0.042
 wasteful-0.041
Token.
Token ID13
Feature activation-2.01e-02

Pos logit contributions
 Along+0.015
 nights+0.014
 On+0.013
way+0.013
dark+0.013
Neg logit contributions
 mixer-0.017
 mean-0.016
 phone-0.016
boys-0.016
 chefs-0.016
Token Can
Token ID1680
Feature activation+1.23e-01

Pos logit contributions
 nights+0.031
way+0.029
 Along+0.028
 drifted+0.028
ingle+0.027
Neg logit contributions
 paramed-0.032
 mixer-0.031
 liar-0.030
 greedy-0.030
 mean-0.030
Token I
Token ID314
Feature activation-5.09e-03

Pos logit contributions
 paramed+0.237
 liar+0.233
 mixer+0.229
 wasteful+0.228
boys+0.225
Neg logit contributions
 Along-0.152
 nights-0.143
way-0.134
 On-0.132
 whenever-0.131
Token wear
Token ID5806
Feature activation-1.91e-01

Pos logit contributions
 Along+0.008
 nights+0.008
way+0.007
 drifted+0.007
 On+0.007
Neg logit contributions
 paramed-0.008
 mixer-0.008
 liar-0.008
 mean-0.008
boys-0.008
Token them
Token ID606
Feature activation+8.70e-02

Pos logit contributions
 Along+0.319
 nights+0.312
 whenever+0.293
 On+0.291
 night+0.287
Neg logit contributions
 paramed-0.291
 liar-0.287
 wasteful-0.274
roc-0.271
 greedy-0.267
Token to
Token ID284
Feature activation+7.77e-02

Pos logit contributions
 paramed+0.119
 liar+0.115
 mixer+0.114
 wasteful+0.112
boys+0.112
Neg logit contributions
 Along-0.155
 nights-0.153
way-0.143
 On-0.140
 drifted-0.138
Token school
Token ID1524
Feature activation-1.20e-01

Pos logit contributions
 paramed+0.119
 liar+0.116
 mixer+0.114
 wasteful+0.112
 greedy+0.111
Neg logit contributions
 Along-0.125
 nights-0.124
 whenever-0.114
 On-0.113
way-0.112
Token?"
Token ID1701
Feature activation+7.64e-02

Pos logit contributions
 Along+0.207
 nights+0.203
way+0.190
 On+0.187
 whenever+0.186
Neg logit contributions
 paramed-0.171
 liar-0.166
 mixer-0.161
 wasteful-0.160
boys-0.160
Token
Token ID198
Feature activation-1.99e-02

Pos logit contributions
 paramed+0.147
 mixer+0.144
 liar+0.142
 mean+0.141
boys+0.140
Neg logit contributions
 Along-0.093
 nights-0.093
way-0.084
 drifted-0.080
 On-0.078
Token that
Token ID326
Feature activation-3.29e-02

Pos logit contributions
 mixer+0.162
 mean+0.161
 paramed+0.161
boys+0.158
 phone+0.158
Neg logit contributions
 Along-0.136
 nights-0.133
way-0.118
 drifted-0.118
 On-0.117
Token made
Token ID925
Feature activation+7.28e-02

Pos logit contributions
 Along+0.052
 nights+0.051
way+0.048
 On+0.046
 drifted+0.046
Neg logit contributions
 paramed-0.051
 mixer-0.050
 liar-0.050
 mean-0.049
boys-0.049
Token new
Token ID649
Feature activation-1.83e-02

Pos logit contributions
 paramed+0.093
 mixer+0.092
 mean+0.090
 phone+0.088
 mad+0.086
Neg logit contributions
 Along-0.138
 nights-0.130
way-0.126
 On-0.121
more-0.120
Token pills
Token ID19521
Feature activation+8.02e-02

Pos logit contributions
 Along+0.029
 nights+0.028
way+0.026
 On+0.026
 drifted+0.025
Neg logit contributions
 paramed-0.029
 mixer-0.028
boys-0.028
 mean-0.028
 liar-0.028
Token every
Token ID790
Feature activation-2.06e-01

Pos logit contributions
 mean+0.137
 mixer+0.134
boys+0.132
 phone+0.132
 liar+0.131
Neg logit contributions
 Along-0.117
 nights-0.112
way-0.101
 On-0.101
 drifted-0.096
Token day
Token ID1110
Feature activation-2.16e-01

Pos logit contributions
 Along+0.182
 On+0.145
lights+0.141
 Throughout+0.140
 Everywhere+0.138
Neg logit contributions
 mixer-0.469
 mean-0.469
 mad-0.452
 phone-0.450
 liar-0.446
Token.
Token ID13
Feature activation-1.85e-02

Pos logit contributions
 Along+0.274
 nights+0.261
way+0.241
 On+0.233
 drifted+0.227
Neg logit contributions
 paramed-0.401
 mixer-0.398
 mean-0.396
 liar-0.394
boys-0.393
Token Sal
Token ID4849
Feature activation+9.99e-02

Pos logit contributions
 nights+0.027
 Along+0.026
way+0.025
 drifted+0.024
dark+0.023
Neg logit contributions
 paramed-0.031
 mixer-0.030
 liar-0.029
roc-0.029
 wasteful-0.028
Token gave
Token ID2921
Feature activation+3.35e-02

Pos logit contributions
 liar+0.180
 paramed+0.179
 bluff+0.175
 wasteful+0.173
 mean+0.169
Neg logit contributions
 nights-0.136
 Along-0.133
 drifted-0.123
way-0.120
 days-0.117
Token Joe
Token ID5689
Feature activation-8.59e-03

Pos logit contributions
 paramed+0.051
 liar+0.050
 wasteful+0.049
 mixer+0.048
 mean+0.048
Neg logit contributions
 Along-0.055
 nights-0.055
way-0.050
 whenever-0.049
 On-0.049
Token a
Token ID257
Feature activation-2.02e-02

Pos logit contributions
 Along+0.016
 nights+0.016
way+0.015
 On+0.014
 drifted+0.014
Neg logit contributions
 paramed-0.011
 mixer-0.011
 mean-0.011
 greedy-0.010
boys-0.010
Token He
Token ID679
Feature activation+7.62e-04

Pos logit contributions
 nights+0.042
 drifted+0.038
way+0.037
ingle+0.036
filled+0.036
Neg logit contributions
 mixer-0.037
 stern-0.036
 greedy-0.036
 sorry-0.035
 bluff-0.035
Token knew
Token ID2993
Feature activation-1.94e-03

Pos logit contributions
 mean+0.001
boys+0.001
 greedy+0.001
 paramed+0.001
 mad+0.001
Neg logit contributions
 Along-0.001
 nights-0.001
way-0.001
 On-0.001
ingle-0.001
Token that
Token ID326
Feature activation-3.56e-02

Pos logit contributions
 Along+0.003
 nights+0.003
way+0.003
 On+0.003
 drifted+0.003
Neg logit contributions
 mean-0.003
 mixer-0.003
 liar-0.003
 greedy-0.003
 paramed-0.003
Token as
Token ID355
Feature activation-1.87e-01

Pos logit contributions
 Along+0.050
 nights+0.047
way+0.044
 On+0.043
 drifted+0.042
Neg logit contributions
 paramed-0.062
 mixer-0.062
 mean-0.061
 liar-0.061
boys-0.060
Token long
Token ID890
Feature activation-9.24e-02

Pos logit contributions
 Along+0.367
way+0.339
ingle+0.319
 On+0.319
 drifted+0.312
Neg logit contributions
 mean-0.249
 greedy-0.226
 mad-0.213
 sorry-0.210
 chefs-0.209
Token as
Token ID355
Feature activation-1.87e-01

Pos logit contributions
 Along+0.107
ingle+0.103
dark+0.100
way+0.094
 On+0.093
Neg logit contributions
 mean-0.174
 fighting-0.171
 greedy-0.170
 mixer-0.170
 chefs-0.169
Token he
Token ID339
Feature activation-9.31e-02

Pos logit contributions
 Along+0.309
ingle+0.273
way+0.273
 On+0.267
 drifted+0.259
Neg logit contributions
 mean-0.308
 greedy-0.288
 sorry-0.282
 chefs-0.282
 mad-0.277
Token had
Token ID550
Feature activation-6.23e-02

Pos logit contributions
 Along+0.177
way+0.164
 nights+0.160
 On+0.158
ingle+0.153
Neg logit contributions
 mean-0.112
boys-0.107
 greedy-0.102
 qu-0.100
 phone-0.100
Token this
Token ID428
Feature activation-4.71e-03

Pos logit contributions
 Along+0.110
 nights+0.106
way+0.100
 On+0.098
dark+0.097
Neg logit contributions
 mean-0.085
 paramed-0.084
 liar-0.082
 mixer-0.082
 greedy-0.082
Token door
Token ID3420
Feature activation-1.07e-02

Pos logit contributions
 Along+0.007
 nights+0.007
 drifted+0.006
ingle+0.006
 On+0.006
Neg logit contributions
 mean-0.008
 mixer-0.008
 phone-0.008
 stern-0.008
 paramed-0.007
Token,
Token ID11
Feature activation+7.71e-02

Pos logit contributions
 Along+0.018
 nights+0.017
way+0.016
ingle+0.015
 On+0.015
Neg logit contributions
 mean-0.016
 phone-0.015
 stern-0.015
 mixer-0.015
 paramed-0.015
Token dance
Token ID9280
Feature activation-1.28e-01

Pos logit contributions
 paramed+0.144
 mixer+0.141
 liar+0.140
 mean+0.139
boys+0.138
Neg logit contributions
 Along-0.113
 nights-0.112
way-0.103
 drifted-0.098
 On-0.098
Token.
Token ID13
Feature activation-3.25e-02

Pos logit contributions
 Along+0.166
 nights+0.163
way+0.150
 drifted+0.142
 On+0.141
Neg logit contributions
 paramed-0.234
 mixer-0.231
 mean-0.229
 liar-0.228
boys-0.226
Token She
Token ID1375
Feature activation-6.16e-02

Pos logit contributions
 nights+0.055
way+0.051
 drifted+0.050
 lifetime+0.048
 Along+0.047
Neg logit contributions
 paramed-0.048
 mixer-0.048
 mean-0.046
 Lee-0.046
 liar-0.044
Token dances
Token ID38207
Feature activation-8.74e-02

Pos logit contributions
 Along+0.115
way+0.110
 nights+0.109
ingle+0.102
more+0.102
Neg logit contributions
 mean-0.085
 greedy-0.076
 liar-0.075
 mixer-0.075
boys-0.073
Token every
Token ID790
Feature activation-2.07e-01

Pos logit contributions
 Along+0.123
 nights+0.121
way+0.115
 drifted+0.110
dark+0.107
Neg logit contributions
 mean-0.149
 mixer-0.144
 paramed-0.143
 liar-0.142
 phone-0.141
Token day
Token ID1110
Feature activation-2.16e-01

Pos logit contributions
 Along+0.187
 On+0.154
 nights+0.151
more+0.142
 Throughout+0.141
Neg logit contributions
 mean-0.486
 mixer-0.482
 phone-0.462
 mad-0.452
 greedy-0.441
Token in
Token ID287
Feature activation-6.18e-02

Pos logit contributions
 Along+0.363
 nights+0.346
way+0.342
ingle+0.331
 drifted+0.327
Neg logit contributions
 mean-0.315
 mixer-0.297
 greedy-0.294
 phone-0.287
 fighting-0.287
Token her
Token ID607
Feature activation-6.86e-02

Pos logit contributions
 Along+0.134
way+0.134
 nights+0.127
 drifted+0.123
dark+0.122
Neg logit contributions
 mean-0.063
 mixer-0.059
 mad-0.058
 phone-0.056
 greedy-0.053
Token room
Token ID2119
Feature activation-6.86e-03

Pos logit contributions
 Along+0.124
 nights+0.121
way+0.120
more+0.113
 On+0.111
Neg logit contributions
 mixer-0.093
 mean-0.090
 phone-0.090
 paramed-0.087
 greedy-0.086
Token.
Token ID13
Feature activation-3.30e-02

Pos logit contributions
 Along+0.011
 nights+0.010
way+0.010
 On+0.009
 drifted+0.009
Neg logit contributions
 mean-0.011
 mixer-0.011
 phone-0.011
 greedy-0.011
 liar-0.011
Token She
Token ID1375
Feature activation-6.19e-02

Pos logit contributions
 nights+0.056
way+0.053
 lifetime+0.050
ingle+0.049
 drifted+0.049
Neg logit contributions
 mixer-0.050
 paramed-0.049
 Lee-0.047
 mean-0.046
 liar-0.045
Token stories
Token ID3923
Feature activation-1.27e-02

Pos logit contributions
 Along+0.046
 nights+0.045
way+0.042
 On+0.041
 drifted+0.041
Neg logit contributions
 paramed-0.039
 mixer-0.037
 liar-0.037
 mean-0.037
boys-0.037
Token.
Token ID13
Feature activation-2.96e-02

Pos logit contributions
 Along+0.018
 nights+0.018
way+0.016
 drifted+0.016
ingle+0.015
Neg logit contributions
 mean-0.022
 mixer-0.021
 phone-0.021
boys-0.021
 paramed-0.021
Token
Token ID198
Feature activation-2.69e-02

Pos logit contributions
 nights+0.043
way+0.041
 Along+0.040
 drifted+0.038
ingle+0.037
Neg logit contributions
 mixer-0.050
 paramed-0.048
 mean-0.047
 greedy-0.047
 stern-0.047
Token
Token ID198
Feature activation-2.63e-02

Pos logit contributions
 nights+0.042
 drifted+0.038
 Along+0.038
way+0.036
filled+0.035
Neg logit contributions
 mixer-0.042
 mean-0.041
 stern-0.040
 paramed-0.039
 phone-0.038
TokenOne
Token ID3198
Feature activation-1.53e-01

Pos logit contributions
 nights+0.033
 Along+0.033
 drifted+0.030
way+0.030
ingle+0.029
Neg logit contributions
 paramed-0.048
 mean-0.048
 mixer-0.048
 liar-0.046
 greedy-0.046
Token day
Token ID1110
Feature activation-2.17e-01

Pos logit contributions
 Along+0.167
growth+0.156
 drifted+0.149
way+0.146
 nights+0.144
Neg logit contributions
 mean-0.346
 mixer-0.312
 mad-0.309
 greedy-0.303
 rude-0.301
Token,
Token ID11
Feature activation+7.61e-02

Pos logit contributions
 Along+0.396
 nights+0.386
way+0.367
dark+0.349
 drifted+0.349
Neg logit contributions
 mixer-0.288
 mean-0.282
 paramed-0.282
 liar-0.277
 stern-0.273
Token she
Token ID673
Feature activation-1.52e-01

Pos logit contributions
 mixer+0.083
 stern+0.081
 mean+0.080
 chefs+0.080
 mad+0.077
Neg logit contributions
 Along-0.154
 nights-0.152
way-0.149
ingle-0.142
dark-0.142
Token goes
Token ID2925
Feature activation-6.62e-02

Pos logit contributions
 Along+0.271
 nights+0.268
way+0.257
more+0.241
 On+0.240
Neg logit contributions
 mean-0.208
boys-0.201
 paramed-0.201
 liar-0.201
 greedy-0.200
Token to
Token ID284
Feature activation+8.23e-02

Pos logit contributions
 nights+0.116
 Along+0.113
way+0.108
 drifted+0.107
more+0.104
Neg logit contributions
 mean-0.090
 phone-0.090
 paramed-0.089
 mixer-0.089
boys-0.087
Token the
Token ID262
Feature activation+1.16e-02

Pos logit contributions
 paramed+0.110
 mixer+0.108
 mean+0.107
 liar+0.107
boys+0.105
Neg logit contributions
 Along-0.147
 nights-0.147
way-0.139
 drifted-0.134
 On-0.130