TOP ACTIVATIONS
MAX = 0.415

Sam was a quick lear ner and soon he was behaving
how to jog on the treadmill . It was so much
Mike . Tan ner sighed , " Oh ,
They would run on the treadmill and lift weights together .
He was a good lear ner , and he was proud
in the hole like a treadmill . Finally , he began
ora was a fast lear ner . Soon enough she was
Peter was a fast lear ner , and after a few
was a very quick lear ner and soon he was baking
broccoli !" Tan ner was confused , but he
You are a good lear ner . Do you want to
would be a good lear ner and look er too .
smells ?" Tan ner couldn 't believe what he
You are a successful lear ner and a good daughter ."
kitchen ?" The other sp oons agreed and they all moved
. There were big sp oons , small sp oons ,
in any of the cup boards . The pan
sp oons , small sp oons , and teaspoons . Sue
a boy who loved sp oons . He loved them so
he would look for sp oons in every corner . One

MIN ACTIVATIONS
MIN = -0.380

his face . Max y el ps and backs away .
his paw . Jen y el ps and runs away .
. She likes that past el . It is her favorite
. She recognizes the bird c age . It belongs to
can use the green past el first , and then give
can use the red past el first , and then give
He shows Mom the bird c age . "
, Lily saw a bird c age in the yard .
flies out of the bird c age . It is happy
" That is our past el !" Tom runs
and hurt . He y el ped and ran away .
the door of the bird c age . It is locked
this is not a past el . This is a cr
the bird and the bird c age and the chaos .
He told us to kne el and pray . We did
! You have to kne el and pray before it !"
toy car with the bird c age to the kitchen .
goodbye . They said " c aw , c aw ,
He liked his new bird c age , but he also
spicy . It hurts my mouth . It is not tasty

INTERVAL 0.217 - 0.415
CONTAINS 0.308%

was so proud of his work that he decided to enter
was big and shiny and Emma thought it was very cool
It said , " Happy early birthday , Sam . Love
a few minutes of hard work , the pin began to
L ily looked at the soldiers and noticed that their faces

INTERVAL 0.018 - 0.217
CONTAINS 35.978%

. He felt sad because he wanted to play in the
on it . Tim my loved his new towel and couldn
the corner . They both ran to it and Joe said
wheel in the box of blocks . " Ben
a time , there was a little girl named Lily .

INTERVAL -0.181 - 0.018
CONTAINS 62.534%

One day , Bella wanted a pastry from the
they can visit Tom 's house . Lily is happy .
an idea . He said , " Let 's unite our
some glue and some tape . He will be good as
furry bear was happy in his shade wrap . He could

INTERVAL -0.380 - -0.181
CONTAINS 1.180%

He tried to grab the chain from them .
It 's purple !" Tom replied , " It looks like
The animals all started to trem ble . But the little
and held it in her hand . â
boy said in a rude voice . L
Token Sam
Token ID3409
Feature activation-5.89e-02

Pos logit contributions
 Hood+0.020
 treadmill+0.018
ner+0.018
ussy+0.018
fighters+0.017
Neg logit contributions
 wing-0.016
 mouth-0.016
c-0.016
el-0.016
 feather-0.015
Token was
Token ID373
Feature activation+1.10e-03

Pos logit contributions
c+0.067
 mouth+0.066
el+0.059
 wing+0.059
 its+0.058
Neg logit contributions
 Hood-0.115
ner-0.109
ussy-0.106
 treadmill-0.105
fighters-0.104
Token a
Token ID257
Feature activation+5.28e-02

Pos logit contributions
 Hood+0.002
ner+0.002
 treadmill+0.002
ussy+0.002
fighters+0.002
Neg logit contributions
 mouth-0.001
c-0.001
el-0.001
 wing-0.001
 its-0.001
Token quick
Token ID2068
Feature activation+1.93e-01

Pos logit contributions
 Hood+0.096
ussy+0.088
ner+0.087
oons+0.087
 treadmill+0.085
Neg logit contributions
 mouth-0.069
 wing-0.065
c-0.064
 its-0.060
 feather-0.059
Token lear
Token ID22454
Feature activation+1.47e-01

Pos logit contributions
 Hood+0.302
ner+0.270
ussy+0.268
 treadmill+0.267
fighters+0.257
Neg logit contributions
 mouth-0.292
c-0.282
 wing-0.278
el-0.276
 its-0.264
Tokenner
Token ID1008
Feature activation+4.15e-01

Pos logit contributions
 Hood+0.080
 treadmill+0.061
oons+0.060
ussy+0.057
fighters+0.051
Neg logit contributions
el-0.354
c-0.348
 its-0.347
 wing-0.347
 mouth-0.344
Token and
Token ID290
Feature activation+2.80e-02

Pos logit contributions
 Hood+0.660
ner+0.601
 treadmill+0.582
ussy+0.574
fighters+0.558
Neg logit contributions
 mouth-0.606
c-0.589
el-0.584
 wing-0.574
 its-0.541
Token soon
Token ID2582
Feature activation+1.06e-01

Pos logit contributions
 Hood+0.048
ner+0.043
ussy+0.043
 treadmill+0.042
assy+0.041
Neg logit contributions
 mouth-0.040
c-0.038
 wing-0.037
el-0.036
 its-0.036
Token he
Token ID339
Feature activation+4.05e-02

Pos logit contributions
 Hood+0.199
ussy+0.189
 treadmill+0.188
ner+0.184
assy+0.180
Neg logit contributions
 mouth-0.135
 its-0.125
 wing-0.117
c-0.117
el-0.114
Token was
Token ID373
Feature activation+1.70e-03

Pos logit contributions
 Hood+0.082
ner+0.080
 treadmill+0.078
assy+0.077
ussy+0.077
Neg logit contributions
c-0.040
 mouth-0.039
 its-0.035
el-0.035
 tilted-0.033
Token behaving
Token ID37722
Feature activation+4.19e-02

Pos logit contributions
 Hood+0.003
fighters+0.003
oons+0.003
boards+0.003
 treadmill+0.003
Neg logit contributions
 mouth-0.002
c-0.002
 its-0.002
 wing-0.002
 feathers-0.002
Token how
Token ID703
Feature activation-4.14e-02

Pos logit contributions
 Hood+0.164
ner+0.149
 treadmill+0.148
ussy+0.147
fighters+0.143
Neg logit contributions
 mouth-0.118
c-0.114
el-0.112
 wing-0.111
 its-0.105
Token to
Token ID284
Feature activation-8.91e-03

Pos logit contributions
 its+0.038
 mouth+0.037
 wing+0.034
c+0.031
 feathers+0.030
Neg logit contributions
 Hood-0.090
ussy-0.088
ner-0.086
assy-0.084
fighters-0.083
Token jog
Token ID48342
Feature activation+7.49e-02

Pos logit contributions
 mouth+0.012
c+0.011
el+0.011
 wing+0.011
 its+0.011
Neg logit contributions
 Hood-0.016
ner-0.014
ussy-0.014
 treadmill-0.014
assy-0.014
Token on
Token ID319
Feature activation-3.35e-02

Pos logit contributions
 Hood+0.141
ussy+0.131
ner+0.130
assy+0.127
fighters+0.126
Neg logit contributions
c-0.089
 mouth-0.089
 its-0.088
 wing-0.081
el-0.076
Token the
Token ID262
Feature activation+1.89e-02

Pos logit contributions
 mouth+0.039
 its+0.038
c+0.036
el+0.035
 wing+0.034
Neg logit contributions
 Hood-0.067
ussy-0.062
ner-0.061
ives-0.060
assy-0.060
Token treadmill
Token ID49246
Feature activation+4.15e-01

Pos logit contributions
 Hood+0.036
assy+0.032
ussy+0.031
ilda+0.031
fighters+0.030
Neg logit contributions
 mouth-0.026
 be-0.023
 neck-0.022
 direction-0.022
 palm-0.021
Token.
Token ID13
Feature activation+8.11e-03

Pos logit contributions
 Hood+0.806
ussy+0.723
ner+0.705
assy+0.702
fighters+0.699
Neg logit contributions
 mouth-0.486
c-0.460
 its-0.434
 wing-0.431
el-0.430
Token It
Token ID632
Feature activation-9.62e-02

Pos logit contributions
 Hood+0.014
ner+0.013
ussy+0.012
 treadmill+0.012
fighters+0.012
Neg logit contributions
 mouth-0.011
c-0.011
el-0.011
 wing-0.011
 its-0.010
Token was
Token ID373
Feature activation-1.53e-04

Pos logit contributions
 mouth+0.113
c+0.112
el+0.106
 wing+0.105
 its+0.101
Neg logit contributions
 Hood-0.181
ner-0.170
 treadmill-0.165
ussy-0.165
assy-0.162
Token so
Token ID523
Feature activation-9.66e-02

Pos logit contributions
 mouth+0.000
c+0.000
 its+0.000
 wing+0.000
el+0.000
Neg logit contributions
 Hood-0.000
ner-0.000
ussy-0.000
 treadmill-0.000
fighters-0.000
Token much
Token ID881
Feature activation+2.79e-02

Pos logit contributions
 its+0.094
 mouth+0.088
c+0.084
el+0.075
 be+0.073
Neg logit contributions
 Hood-0.208
ner-0.193
ussy-0.189
ilda-0.186
drivers-0.182
Token Mike
Token ID4995
Feature activation+2.48e-03

Pos logit contributions
 mouth+0.287
c+0.282
 wing+0.274
el+0.265
light+0.261
Neg logit contributions
 Hood-0.307
ner-0.283
ussy-0.270
 treadmill-0.269
ilda-0.262
Token.
Token ID13
Feature activation+8.34e-03

Pos logit contributions
 Hood+0.004
ner+0.004
ussy+0.004
 treadmill+0.004
fighters+0.004
Neg logit contributions
 mouth-0.004
c-0.003
 wing-0.003
el-0.003
 its-0.003
Token
Token ID198
Feature activation-6.01e-03

Pos logit contributions
 Hood+0.013
ner+0.012
 treadmill+0.012
ussy+0.011
fighters+0.011
Neg logit contributions
c-0.012
 mouth-0.012
 wing-0.012
el-0.012
light-0.012
Token
Token ID198
Feature activation-6.62e-03

Pos logit contributions
 mouth+0.008
c+0.007
el+0.007
 wing+0.007
 its+0.007
Neg logit contributions
 Hood-0.011
ner-0.010
 treadmill-0.010
ussy-0.010
fighters-0.009
TokenTan
Token ID45557
Feature activation+8.57e-02

Pos logit contributions
 mouth+0.011
c+0.011
 wing+0.010
el+0.010
 its+0.010
Neg logit contributions
 Hood-0.009
ner-0.009
 treadmill-0.008
ussy-0.008
fighters-0.008
Tokenner
Token ID1008
Feature activation+4.12e-01

Pos logit contributions
 Hood+0.118
 treadmill+0.100
ussy+0.098
fighters+0.096
cakes+0.095
Neg logit contributions
c-0.150
el-0.147
 mouth-0.137
light-0.133
 pouch-0.130
Token sighed
Token ID21893
Feature activation+6.59e-02

Pos logit contributions
 Hood+0.654
ner+0.605
 treadmill+0.575
ussy+0.564
ilda+0.562
Neg logit contributions
 mouth-0.591
c-0.580
el-0.567
 wing-0.566
 oval-0.531
Token,
Token ID11
Feature activation+1.95e-02

Pos logit contributions
 Hood+0.090
ner+0.083
 treadmill+0.079
ilda+0.078
ussy+0.077
Neg logit contributions
c-0.105
 mouth-0.102
el-0.101
 wing-0.099
rys-0.095
Token "
Token ID366
Feature activation+5.66e-02

Pos logit contributions
ner+0.025
 Hood+0.023
 treadmill+0.023
ilda+0.022
cases+0.020
Neg logit contributions
el-0.033
c-0.033
rys-0.031
 wing-0.031
light-0.030
TokenOh
Token ID5812
Feature activation-1.54e-01

Pos logit contributions
ner+0.062
 Hood+0.061
ussy+0.057
fighters+0.057
ives+0.055
Neg logit contributions
 mouth-0.117
 wing-0.107
 mouths-0.106
 tongue-0.104
 eyes-0.104
Token,
Token ID11
Feature activation+1.86e-02

Pos logit contributions
 mouth+0.209
c+0.209
 wing+0.205
el+0.204
 feathers+0.191
Neg logit contributions
 Hood-0.258
ner-0.235
 treadmill-0.230
ussy-0.227
fighters-0.220
Token They
Token ID1119
Feature activation-2.44e-02

Pos logit contributions
 Hood+0.012
 treadmill+0.011
ner+0.011
ussy+0.011
fighters+0.011
Neg logit contributions
 wing-0.012
 mouth-0.011
el-0.011
c-0.011
 fin-0.011
Token would
Token ID561
Feature activation+9.57e-02

Pos logit contributions
c+0.025
el+0.023
 mouth+0.022
 be+0.021
 its+0.021
Neg logit contributions
ner-0.048
ussy-0.046
 Hood-0.046
ives-0.045
 treadmill-0.044
Token run
Token ID1057
Feature activation+1.07e-01

Pos logit contributions
 Hood+0.167
ner+0.153
ussy+0.150
assy+0.145
ilda+0.144
Neg logit contributions
 mouth-0.131
c-0.124
el-0.121
 wing-0.119
 its-0.117
Token on
Token ID319
Feature activation-3.49e-02

Pos logit contributions
ussy+0.262
assy+0.258
 Hood+0.249
ibr+0.244
ner+0.241
Neg logit contributions
 its-0.065
 be-0.059
c-0.054
 mouth-0.049
 too-0.049
Token the
Token ID262
Feature activation+1.73e-02

Pos logit contributions
 its+0.035
 mouth+0.033
c+0.031
 neck+0.031
el+0.030
Neg logit contributions
 Hood-0.077
ussy-0.075
ives-0.072
ner-0.070
fighters-0.069
Token treadmill
Token ID49246
Feature activation+4.10e-01

Pos logit contributions
 Hood+0.030
assy+0.026
ussy+0.025
fighters+0.024
ner+0.024
Neg logit contributions
 mouth-0.025
 be-0.023
 direction-0.022
 neck-0.021
 palm-0.021
Token and
Token ID290
Feature activation+2.34e-02

Pos logit contributions
 Hood+0.833
ussy+0.777
assy+0.742
fighters+0.722
ner+0.721
Neg logit contributions
 mouth-0.427
c-0.399
 its-0.370
el-0.370
 neck-0.361
Token lift
Token ID10303
Feature activation-6.85e-03

Pos logit contributions
 Hood+0.041
ussy+0.036
ner+0.036
assy+0.035
ilda+0.034
Neg logit contributions
 mouth-0.033
c-0.031
el-0.030
 wing-0.029
 its-0.029
Token weights
Token ID19590
Feature activation+2.66e-01

Pos logit contributions
 mouth+0.009
c+0.009
el+0.009
 its+0.008
 wing+0.008
Neg logit contributions
 Hood-0.012
ner-0.011
ussy-0.011
 treadmill-0.011
assy-0.011
Token together
Token ID1978
Feature activation-2.92e-02

Pos logit contributions
 Hood+0.485
ussy+0.437
 treadmill+0.432
ner+0.431
assy+0.428
Neg logit contributions
 mouth-0.332
el-0.319
c-0.316
 wing-0.308
 its-0.307
Token.
Token ID13
Feature activation+6.54e-03

Pos logit contributions
 mouth+0.046
c+0.045
 wing+0.044
el+0.044
light+0.041
Neg logit contributions
 Hood-0.043
ner-0.039
 treadmill-0.039
ussy-0.038
fighters-0.037
Token He
Token ID679
Feature activation+1.35e-02

Pos logit contributions
 Hood+0.014
ner+0.013
 treadmill+0.012
ussy+0.012
fighters+0.012
Neg logit contributions
 mouth-0.010
c-0.009
el-0.009
 wing-0.009
 its-0.008
Token was
Token ID373
Feature activation-4.77e-03

Pos logit contributions
 Hood+0.024
ner+0.023
 treadmill+0.022
ussy+0.022
fighters+0.022
Neg logit contributions
 mouth-0.017
c-0.017
 wing-0.016
el-0.016
 its-0.015
Token a
Token ID257
Feature activation+4.53e-02

Pos logit contributions
 mouth+0.007
el+0.007
c+0.007
 wing+0.006
light+0.006
Neg logit contributions
 Hood-0.008
ner-0.007
ussy-0.007
 treadmill-0.007
assy-0.007
Token good
Token ID922
Feature activation+1.24e-01

Pos logit contributions
 Hood+0.073
ner+0.066
 treadmill+0.065
ussy+0.065
fighters+0.063
Neg logit contributions
 mouth-0.066
c-0.064
 wing-0.063
el-0.062
 its-0.059
Token lear
Token ID22454
Feature activation+1.40e-01

Pos logit contributions
 Hood+0.213
ner+0.193
 treadmill+0.190
ussy+0.190
fighters+0.184
Neg logit contributions
 mouth-0.170
c-0.164
el-0.161
 wing-0.161
 its-0.152
Tokenner
Token ID1008
Feature activation+4.10e-01

Pos logit contributions
 Hood+0.082
 treadmill+0.063
oons+0.062
ussy+0.062
fighters+0.056
Neg logit contributions
el-0.333
c-0.331
 wing-0.328
 mouth-0.328
 its-0.326
Token,
Token ID11
Feature activation+1.29e-02

Pos logit contributions
 Hood+0.655
ner+0.594
 treadmill+0.577
ussy+0.571
fighters+0.552
Neg logit contributions
 mouth-0.598
c-0.580
el-0.574
 wing-0.566
 its-0.535
Token and
Token ID290
Feature activation+2.11e-02

Pos logit contributions
 Hood+0.017
 treadmill+0.016
ner+0.016
ilda+0.015
assy+0.014
Neg logit contributions
el-0.021
c-0.019
rys-0.019
 wing-0.019
 mouth-0.018
Token he
Token ID339
Feature activation+3.30e-02

Pos logit contributions
 Hood+0.030
ner+0.026
 treadmill+0.026
ussy+0.025
ilda+0.025
Neg logit contributions
c-0.033
el-0.033
 mouth-0.033
 wing-0.033
 pouch-0.031
Token was
Token ID373
Feature activation-5.09e-03

Pos logit contributions
 Hood+0.059
 treadmill+0.056
ner+0.055
ussy+0.055
fighters+0.054
Neg logit contributions
 mouth-0.042
c-0.041
 wing-0.039
el-0.038
 its-0.038
Token proud
Token ID6613
Feature activation+8.14e-02

Pos logit contributions
 mouth+0.007
c+0.007
 wing+0.007
el+0.006
 its+0.006
Neg logit contributions
 Hood-0.009
 treadmill-0.008
ner-0.008
ussy-0.008
fighters-0.008
Token in
Token ID287
Feature activation-3.20e-02

Pos logit contributions
 Hood+0.138
assy+0.132
ussy+0.132
oons+0.126
ner+0.124
Neg logit contributions
 its-0.078
 mouth-0.072
 feathers-0.068
 wing-0.067
c-0.066
Token the
Token ID262
Feature activation+1.77e-02

Pos logit contributions
c+0.048
 mouth+0.048
 wing+0.047
el+0.046
light+0.045
Neg logit contributions
 Hood-0.047
 treadmill-0.044
ner-0.043
ussy-0.040
fighters-0.039
Token hole
Token ID7604
Feature activation-2.31e-01

Pos logit contributions
 Hood+0.030
 treadmill+0.027
ner+0.026
ussy+0.026
fighters+0.025
Neg logit contributions
el-0.023
c-0.023
 mouth-0.023
 wing-0.023
light-0.022
Token like
Token ID588
Feature activation-5.24e-02

Pos logit contributions
 mouth+0.321
c+0.310
el+0.303
 wing+0.303
 its+0.288
Neg logit contributions
 Hood-0.391
ner-0.353
 treadmill-0.349
ussy-0.348
fighters-0.336
Token a
Token ID257
Feature activation+4.67e-02

Pos logit contributions
 mouth+0.065
 its+0.063
 wing+0.060
 feathers+0.059
el+0.057
Neg logit contributions
 Hood-0.098
ussy-0.093
assy-0.091
 treadmill-0.089
ner-0.088
Token treadmill
Token ID49246
Feature activation+4.09e-01

Pos logit contributions
 Hood+0.079
ner+0.071
ussy+0.071
 treadmill+0.070
fighters+0.068
Neg logit contributions
 mouth-0.066
c-0.063
 wing-0.062
el-0.061
 its-0.058
Token.
Token ID13
Feature activation+7.85e-03

Pos logit contributions
 Hood+0.788
ussy+0.710
assy+0.688
ner+0.678
fighters+0.675
Neg logit contributions
 mouth-0.500
c-0.467
 wing-0.456
 its-0.444
el-0.442
Token Finally
Token ID9461
Feature activation+9.45e-02

Pos logit contributions
 Hood+0.017
ussy+0.016
 treadmill+0.015
fighters+0.015
assy+0.015
Neg logit contributions
 mouth-0.009
 its-0.007
c-0.007
el-0.006
 wing-0.006
Token,
Token ID11
Feature activation+1.50e-02

Pos logit contributions
 Hood+0.174
ussy+0.161
 treadmill+0.158
ner+0.156
assy+0.154
Neg logit contributions
 mouth-0.123
el-0.113
c-0.112
 wing-0.112
 its-0.111
Token he
Token ID339
Feature activation+3.26e-02

Pos logit contributions
 Hood+0.023
ner+0.021
 treadmill+0.021
ussy+0.020
fighters+0.020
Neg logit contributions
 mouth-0.023
c-0.022
el-0.022
 wing-0.022
 its-0.021
Token began
Token ID2540
Feature activation+2.91e-02

Pos logit contributions
 Hood+0.063
 treadmill+0.060
ner+0.060
ussy+0.060
assy+0.058
Neg logit contributions
 mouth-0.038
c-0.037
 wing-0.035
 its-0.034
 feather-0.034
Tokenora
Token ID5799
Feature activation-1.85e-02

Pos logit contributions
 Hood+0.017
ner+0.015
ussy+0.015
 treadmill+0.014
fighters+0.014
Neg logit contributions
 mouth-0.014
c-0.014
 wing-0.013
el-0.013
light-0.013
Token was
Token ID373
Feature activation-3.23e-03

Pos logit contributions
 mouth+0.025
c+0.024
el+0.023
 wing+0.023
 its+0.022
Neg logit contributions
 Hood-0.032
ner-0.029
 treadmill-0.029
ussy-0.029
assy-0.028
Token a
Token ID257
Feature activation+4.81e-02

Pos logit contributions
el+0.005
 mouth+0.004
c+0.004
 wing+0.004
light+0.004
Neg logit contributions
 Hood-0.005
ner-0.005
ussy-0.005
 treadmill-0.005
assy-0.005
Token fast
Token ID3049
Feature activation+1.83e-01

Pos logit contributions
 Hood+0.090
ner+0.082
 treadmill+0.081
ussy+0.081
fighters+0.078
Neg logit contributions
 mouth-0.058
c-0.056
el-0.055
 wing-0.055
 its-0.051
Token lear
Token ID22454
Feature activation+1.39e-01

Pos logit contributions
 Hood+0.296
ner+0.267
 treadmill+0.263
ussy+0.262
fighters+0.253
Neg logit contributions
 mouth-0.269
c-0.261
el-0.255
 wing-0.255
 its-0.242
Tokenner
Token ID1008
Feature activation+4.09e-01

Pos logit contributions
 Hood+0.059
 treadmill+0.040
ussy+0.037
fighters+0.031
oons+0.027
Neg logit contributions
 mouth-0.361
c-0.360
el-0.360
 wing-0.357
 its-0.351
Token.
Token ID13
Feature activation+1.00e-02

Pos logit contributions
 Hood+0.606
ner+0.564
 treadmill+0.527
ussy+0.513
ilda+0.508
Neg logit contributions
 mouth-0.621
el-0.606
c-0.606
 wing-0.593
 oval-0.565
Token Soon
Token ID15894
Feature activation+1.43e-01

Pos logit contributions
 Hood+0.018
 treadmill+0.016
ner+0.016
ussy+0.016
fighters+0.015
Neg logit contributions
 mouth-0.014
c-0.013
el-0.013
 wing-0.013
 its-0.012
Token enough
Token ID1576
Feature activation-2.57e-02

Pos logit contributions
 Hood+0.241
ner+0.219
 treadmill+0.215
ussy+0.214
fighters+0.206
Neg logit contributions
 mouth-0.201
c-0.195
 wing-0.191
el-0.191
 its-0.179
Token she
Token ID673
Feature activation+1.80e-02

Pos logit contributions
c+0.042
 mouth+0.040
el+0.039
 wing+0.039
light+0.038
Neg logit contributions
 Hood-0.037
ner-0.035
 treadmill-0.034
ussy-0.031
fighters-0.031
Token was
Token ID373
Feature activation-2.93e-03

Pos logit contributions
 Hood+0.035
 treadmill+0.033
ussy+0.033
ner+0.033
assy+0.033
Neg logit contributions
 mouth-0.020
c-0.020
el-0.019
 wing-0.018
 feather-0.018
Token Peter
Token ID5613
Feature activation+1.14e-01

Pos logit contributions
 Hood+0.021
ussy+0.020
ner+0.019
 treadmill+0.019
assy+0.019
Neg logit contributions
 mouth-0.017
 its-0.016
el-0.015
c-0.015
 wing-0.015
Token was
Token ID373
Feature activation-1.29e-03

Pos logit contributions
 Hood+0.207
ner+0.189
ussy+0.188
 treadmill+0.188
assy+0.181
Neg logit contributions
 mouth-0.146
c-0.140
 wing-0.136
el-0.135
 its-0.129
Token a
Token ID257
Feature activation+4.76e-02

Pos logit contributions
el+0.002
 mouth+0.002
c+0.002
 wing+0.002
light+0.002
Neg logit contributions
 Hood-0.002
ner-0.002
 treadmill-0.002
ussy-0.002
assy-0.002
Token fast
Token ID3049
Feature activation+1.81e-01

Pos logit contributions
 Hood+0.083
ussy+0.075
ner+0.075
 treadmill+0.073
fighters+0.072
Neg logit contributions
 mouth-0.066
 wing-0.061
c-0.061
el-0.058
 its-0.058
Token lear
Token ID22454
Feature activation+1.39e-01

Pos logit contributions
 Hood+0.316
ner+0.288
 treadmill+0.284
ussy+0.283
fighters+0.273
Neg logit contributions
 mouth-0.242
c-0.235
el-0.229
 wing-0.228
 its-0.215
Tokenner
Token ID1008
Feature activation+4.09e-01

Pos logit contributions
 Hood+0.070
 treadmill+0.051
ussy+0.049
oons+0.043
fighters+0.043
Neg logit contributions
 mouth-0.348
el-0.346
c-0.346
 wing-0.345
 its-0.341
Token,
Token ID11
Feature activation+1.61e-02

Pos logit contributions
 Hood+0.589
ner+0.549
 treadmill+0.518
ussy+0.496
ilda+0.490
Neg logit contributions
 mouth-0.625
el-0.624
c-0.620
 wing-0.600
 oval-0.571
Token and
Token ID290
Feature activation+2.44e-02

Pos logit contributions
 Hood+0.022
 treadmill+0.021
ner+0.021
drivers+0.019
fighters+0.018
Neg logit contributions
el-0.026
c-0.024
 wing-0.024
rys-0.024
 mouth-0.023
Token after
Token ID706
Feature activation-2.11e-02

Pos logit contributions
 Hood+0.040
ner+0.036
ussy+0.035
 treadmill+0.035
fighters+0.034
Neg logit contributions
 mouth-0.036
c-0.035
 wing-0.034
el-0.034
 its-0.033
Token a
Token ID257
Feature activation+4.72e-02

Pos logit contributions
 mouth+0.029
 its+0.027
c+0.027
el+0.027
 wing+0.027
Neg logit contributions
 Hood-0.037
ussy-0.035
assy-0.033
 treadmill-0.033
ner-0.033
Token few
Token ID1178
Feature activation-1.56e-02

Pos logit contributions
 Hood+0.079
ner+0.072
 treadmill+0.071
ussy+0.070
fighters+0.068
Neg logit contributions
 mouth-0.066
c-0.064
el-0.063
 wing-0.063
 its-0.059
Token was
Token ID373
Feature activation-2.38e-03

Pos logit contributions
 Hood+0.130
ner+0.118
 treadmill+0.117
ussy+0.116
fighters+0.112
Neg logit contributions
 mouth-0.100
c-0.097
 wing-0.094
el-0.094
 its-0.089
Token a
Token ID257
Feature activation+4.81e-02

Pos logit contributions
el+0.003
c+0.003
 mouth+0.003
 wing+0.003
rys+0.003
Neg logit contributions
 Hood-0.004
ner-0.004
ussy-0.004
 treadmill-0.003
assy-0.003
Token very
Token ID845
Feature activation-2.61e-02

Pos logit contributions
 Hood+0.078
ner+0.070
 treadmill+0.070
ussy+0.069
fighters+0.067
Neg logit contributions
 mouth-0.070
c-0.068
el-0.067
 wing-0.066
 its-0.063
Token quick
Token ID2068
Feature activation+1.87e-01

Pos logit contributions
 mouth+0.038
c+0.035
 wing+0.035
el+0.034
 its+0.034
Neg logit contributions
 Hood-0.045
ner-0.040
ussy-0.040
 treadmill-0.039
fighters-0.039
Token lear
Token ID22454
Feature activation+1.40e-01

Pos logit contributions
 Hood+0.319
ussy+0.287
ner+0.285
 treadmill+0.285
fighters+0.276
Neg logit contributions
 mouth-0.261
c-0.247
 wing-0.246
el-0.242
 its-0.232
Tokenner
Token ID1008
Feature activation+4.08e-01

Pos logit contributions
 Hood+0.075
 treadmill+0.064
oons+0.059
ussy+0.058
fighters+0.053
Neg logit contributions
el-0.338
c-0.335
 wing-0.334
 its-0.334
 mouth-0.333
Token and
Token ID290
Feature activation+2.39e-02

Pos logit contributions
 Hood+0.618
ner+0.571
 treadmill+0.540
ussy+0.530
ilda+0.519
Neg logit contributions
 mouth-0.612
c-0.601
el-0.597
 wing-0.586
 its-0.552
Token soon
Token ID2582
Feature activation+1.02e-01

Pos logit contributions
 Hood+0.037
ner+0.034
 treadmill+0.033
ussy+0.032
fighters+0.031
Neg logit contributions
 mouth-0.036
c-0.035
el-0.035
 wing-0.035
light-0.032
Token he
Token ID339
Feature activation+3.35e-02

Pos logit contributions
 Hood+0.193
 treadmill+0.191
ussy+0.188
fighters+0.180
ner+0.180
Neg logit contributions
 mouth-0.126
 its-0.119
c-0.110
 wing-0.107
el-0.105
Token was
Token ID373
Feature activation-2.22e-03

Pos logit contributions
 Hood+0.069
 treadmill+0.068
ner+0.066
ussy+0.065
assy+0.063
Neg logit contributions
c-0.034
 mouth-0.034
el-0.031
 its-0.031
 wing-0.029
Token baking
Token ID16871
Feature activation-3.07e-04

Pos logit contributions
 mouth+0.003
c+0.003
 wing+0.003
el+0.003
 its+0.003
Neg logit contributions
 Hood-0.004
 treadmill-0.003
ner-0.003
ussy-0.003
fighters-0.003
Token broccoli
Token ID44653
Feature activation+1.07e-01

Pos logit contributions
 mouth+0.032
el+0.029
 wing+0.029
c+0.029
 its+0.028
Neg logit contributions
 Hood-0.033
ussy-0.030
 treadmill-0.030
ner-0.030
fighters-0.029
Token!"
Token ID2474
Feature activation+1.73e-01

Pos logit contributions
 Hood+0.172
ussy+0.157
 treadmill+0.157
ner+0.154
assy+0.150
Neg logit contributions
 mouth-0.157
c-0.148
 wing-0.148
el-0.146
 its-0.142
Token
Token ID198
Feature activation-1.00e-02

Pos logit contributions
ner+0.224
 Hood+0.214
 treadmill+0.204
ilda+0.198
cases+0.193
Neg logit contributions
 wing-0.301
c-0.293
 mouth-0.283
 oval-0.277
el-0.275
Token
Token ID198
Feature activation-1.03e-02

Pos logit contributions
 mouth+0.012
c+0.011
el+0.011
 wing+0.011
 its+0.011
Neg logit contributions
 Hood-0.020
ussy-0.018
 treadmill-0.017
ner-0.017
fighters-0.017
TokenTan
Token ID45557
Feature activation+8.22e-02

Pos logit contributions
 mouth+0.017
c+0.017
el+0.016
 wing+0.016
 its+0.016
Neg logit contributions
 Hood-0.015
ner-0.013
ussy-0.013
 treadmill-0.013
fighters-0.012
Tokenner
Token ID1008
Feature activation+4.08e-01

Pos logit contributions
 Hood+0.096
fighters+0.073
ussy+0.073
 treadmill+0.073
 scientists+0.072
Neg logit contributions
c-0.162
el-0.157
 mouth-0.147
light-0.144
ric-0.143
Token was
Token ID373
Feature activation-3.93e-03

Pos logit contributions
 Hood+0.644
ner+0.597
 treadmill+0.574
ussy+0.562
ilda+0.553
Neg logit contributions
 mouth-0.585
c-0.580
el-0.568
 wing-0.563
 oval-0.529
Token confused
Token ID10416
Feature activation-2.09e-02

Pos logit contributions
c+0.006
el+0.005
 mouth+0.005
 wing+0.005
light+0.005
Neg logit contributions
 Hood-0.006
 treadmill-0.006
ussy-0.006
ner-0.006
ilda-0.005
Token,
Token ID11
Feature activation+1.44e-02

Pos logit contributions
c+0.037
 wing+0.036
 mouth+0.035
 pouch+0.035
 cardinal+0.034
Neg logit contributions
 Hood-0.025
ner-0.023
ussy-0.023
ilda-0.023
 treadmill-0.023
Token but
Token ID475
Feature activation-1.37e-01

Pos logit contributions
ner+0.023
 treadmill+0.022
 Hood+0.021
ilda+0.021
cases+0.020
Neg logit contributions
c-0.021
el-0.020
 wing-0.019
light-0.019
rys-0.018
Token he
Token ID339
Feature activation+3.44e-02

Pos logit contributions
 wing+0.267
c+0.264
 mouth+0.249
light+0.247
 feather+0.246
Neg logit contributions
ner-0.155
 Hood-0.138
 treadmill-0.137
ilda-0.135
ussy-0.129
Token You
Token ID921
Feature activation-7.78e-02

Pos logit contributions
 mouth+0.012
c+0.012
 wing+0.012
el+0.012
 feathers+0.011
Neg logit contributions
 Hood-0.018
ner-0.017
 treadmill-0.016
ussy-0.016
fighters-0.016
Token are
Token ID389
Feature activation-1.37e-02

Pos logit contributions
 mouth+0.083
 wing+0.076
c+0.075
 feather+0.072
 feathers+0.072
Neg logit contributions
 Hood-0.158
 treadmill-0.148
ussy-0.148
ner-0.147
ilda-0.146
Token a
Token ID257
Feature activation+4.51e-02

Pos logit contributions
 mouth+0.018
 wing+0.017
 its+0.017
c+0.017
el+0.017
Neg logit contributions
 Hood-0.024
ner-0.022
 treadmill-0.022
ussy-0.022
fighters-0.021
Token good
Token ID922
Feature activation+1.23e-01

Pos logit contributions
cases+0.082
 Hood+0.080
oons+0.080
boards+0.079
ilda+0.078
Neg logit contributions
 mouth-0.060
 wing-0.059
 its-0.056
 feather-0.056
 nest-0.054
Token lear
Token ID22454
Feature activation+1.38e-01

Pos logit contributions
 Hood+0.282
ner+0.269
 treadmill+0.266
cases+0.264
ussy+0.263
Neg logit contributions
 mouth-0.103
 wing-0.098
 feather-0.088
el-0.086
c-0.083
Tokenner
Token ID1008
Feature activation+4.06e-01

Pos logit contributions
 Hood+0.086
 treadmill+0.072
ussy+0.071
oons+0.064
assy+0.062
Neg logit contributions
el-0.318
 mouth-0.316
 wing-0.316
c-0.313
 its-0.311
Token.
Token ID13
Feature activation-8.78e-03

Pos logit contributions
 Hood+0.696
 treadmill+0.648
ussy+0.644
ner+0.625
assy+0.612
Neg logit contributions
 mouth-0.563
 wing-0.534
c-0.533
el-0.506
 its-0.503
Token Do
Token ID2141
Feature activation-1.01e-01

Pos logit contributions
 mouth+0.011
c+0.010
el+0.010
 wing+0.010
 its+0.010
Neg logit contributions
 Hood-0.016
ner-0.015
 treadmill-0.015
ussy-0.015
fighters-0.014
Token you
Token ID345
Feature activation-4.34e-02

Pos logit contributions
 mouth+0.127
c+0.123
 wing+0.120
el+0.120
 its+0.112
Neg logit contributions
 Hood-0.184
ner-0.168
 treadmill-0.166
ussy-0.165
fighters-0.161
Token want
Token ID765
Feature activation+1.73e-02

Pos logit contributions
 mouth+0.066
c+0.064
el+0.063
 wing+0.063
light+0.060
Neg logit contributions
 Hood-0.067
ner-0.060
ussy-0.060
 treadmill-0.059
assy-0.057
Token to
Token ID284
Feature activation-1.35e-02

Pos logit contributions
 Hood+0.028
ner+0.026
 treadmill+0.026
ussy+0.025
fighters+0.024
Neg logit contributions
 mouth-0.024
c-0.024
el-0.023
 wing-0.023
light-0.022
Token would
Token ID561
Feature activation+8.83e-02

Pos logit contributions
 mouth+0.043
c+0.042
el+0.040
 wing+0.040
 its+0.039
Neg logit contributions
 Hood-0.056
ner-0.052
 treadmill-0.051
ussy-0.051
assy-0.049
Token be
Token ID307
Feature activation-2.78e-01

Pos logit contributions
 treadmill+0.124
 Hood+0.122
ner+0.116
 work+0.114
fighters+0.111
Neg logit contributions
el-0.134
c-0.133
light-0.127
rys-0.121
 wing-0.114
Token a
Token ID257
Feature activation+4.16e-02

Pos logit contributions
 mouth+0.394
c+0.381
el+0.373
 wing+0.373
 its+0.353
Neg logit contributions
 Hood-0.464
ner-0.419
 treadmill-0.414
ussy-0.413
fighters-0.398
Token good
Token ID922
Feature activation+1.21e-01

Pos logit contributions
 Hood+0.061
ner+0.056
ussy+0.055
 treadmill+0.054
assy+0.053
Neg logit contributions
 mouth-0.068
 wing-0.064
c-0.063
el-0.061
 its-0.061
Token lear
Token ID22454
Feature activation+1.36e-01

Pos logit contributions
 Hood+0.210
 treadmill+0.189
ussy+0.188
ner+0.187
fighters+0.181
Neg logit contributions
 mouth-0.158
c-0.158
el-0.156
 wing-0.153
light-0.145
Tokenner
Token ID1008
Feature activation+4.05e-01

Pos logit contributions
 Hood+0.098
 treadmill+0.079
oons+0.078
ussy+0.077
boards+0.072
Neg logit contributions
el-0.293
 wing-0.286
c-0.286
 mouth-0.284
 its-0.282
Token and
Token ID290
Feature activation+1.39e-02

Pos logit contributions
 Hood+0.597
ner+0.529
 treadmill+0.506
ussy+0.501
fighters+0.488
Neg logit contributions
 mouth-0.633
el-0.624
c-0.617
 wing-0.601
 oval-0.576
Token look
Token ID804
Feature activation-5.28e-02

Pos logit contributions
 Hood+0.023
ner+0.021
 treadmill+0.020
ussy+0.020
fighters+0.020
Neg logit contributions
 mouth-0.020
c-0.019
el-0.019
 wing-0.019
light-0.018
Tokener
Token ID263
Feature activation-1.28e-03

Pos logit contributions
 mouth+0.037
 its+0.033
c+0.033
el+0.027
 wing+0.026
Neg logit contributions
 Hood-0.127
ner-0.124
 treadmill-0.120
assy-0.119
fighters-0.118
Token too
Token ID1165
Feature activation-1.86e-01

Pos logit contributions
 mouth+0.002
c+0.002
el+0.002
 wing+0.002
 its+0.001
Neg logit contributions
 Hood-0.002
ner-0.002
 treadmill-0.002
ussy-0.002
assy-0.002
Token.
Token ID13
Feature activation-8.08e-03

Pos logit contributions
 mouth+0.191
 its+0.167
c+0.150
 wing+0.149
el+0.145
Neg logit contributions
 Hood-0.407
 treadmill-0.374
ner-0.368
assy-0.367
ilda-0.362
Token smells
Token ID25760
Feature activation+8.68e-02

Pos logit contributions
c+0.068
 mouth+0.067
el+0.063
 wing+0.059
 feathers+0.057
Neg logit contributions
 Hood-0.097
 treadmill-0.096
ner-0.092
ilda-0.090
assy-0.089
Token?"
Token ID1701
Feature activation+9.14e-02

Pos logit contributions
 Hood+0.134
ner+0.122
 treadmill+0.118
ussy+0.116
oons+0.114
Neg logit contributions
 mouth-0.127
c-0.126
el-0.123
 wing-0.122
light-0.117
Token
Token ID198
Feature activation-1.34e-02

Pos logit contributions
 Hood+0.133
ner+0.129
 treadmill+0.124
ussy+0.119
ilda+0.117
Neg logit contributions
 wing-0.140
 mouth-0.139
c-0.137
el-0.134
light-0.130
Token
Token ID198
Feature activation-1.34e-02

Pos logit contributions
 mouth+0.016
c+0.016
el+0.016
 wing+0.015
 its+0.015
Neg logit contributions
 Hood-0.025
ussy-0.023
 treadmill-0.023
ner-0.022
fighters-0.022
TokenTan
Token ID45557
Feature activation+7.92e-02

Pos logit contributions
 mouth+0.021
c+0.021
el+0.021
 wing+0.020
 its+0.020
Neg logit contributions
 Hood-0.020
ner-0.017
ussy-0.017
 treadmill-0.017
fighters-0.017
Tokenner
Token ID1008
Feature activation+4.05e-01

Pos logit contributions
 Hood+0.106
 treadmill+0.086
ussy+0.085
fighters+0.084
cakes+0.082
Neg logit contributions
c-0.142
el-0.140
 mouth-0.130
light-0.126
 pouch-0.125
Token couldn
Token ID3521
Feature activation+5.27e-02

Pos logit contributions
 Hood+0.617
ner+0.584
 treadmill+0.544
ussy+0.536
ilda+0.533
Neg logit contributions
 mouth-0.593
c-0.585
el-0.574
 wing-0.572
 oval-0.537
Token't
Token ID470
Feature activation+1.73e-01

Pos logit contributions
 Hood+0.079
assy+0.067
ussy+0.067
 treadmill+0.065
fighters+0.063
Neg logit contributions
el-0.087
 its-0.083
 mouth-0.082
c-0.081
 wing-0.079
Token believe
Token ID1975
Feature activation-4.24e-02

Pos logit contributions
ner+0.209
 Hood+0.208
 treadmill+0.182
ussy+0.178
fighters+0.176
Neg logit contributions
c-0.316
 mouth-0.309
el-0.294
 wing-0.287
 mouths-0.283
Token what
Token ID644
Feature activation+2.08e-02

Pos logit contributions
c+0.060
 mouth+0.058
 wing+0.055
el+0.053
light+0.052
Neg logit contributions
 Hood-0.072
ner-0.065
 treadmill-0.064
ussy-0.063
ilda-0.062
Token he
Token ID339
Feature activation+3.17e-02

Pos logit contributions
 Hood+0.029
ner+0.026
 treadmill+0.025
ilda+0.024
fighters+0.024
Neg logit contributions
c-0.034
 mouth-0.034
 wing-0.034
el-0.033
 oval-0.032
Token You
Token ID921
Feature activation-7.90e-02

Pos logit contributions
 mouth+0.010
c+0.010
el+0.010
 wing+0.010
 feathers+0.010
Neg logit contributions
 Hood-0.012
ner-0.011
ussy-0.011
 treadmill-0.011
fighters-0.011
Token are
Token ID389
Feature activation-1.66e-02

Pos logit contributions
 mouth+0.098
 wing+0.096
c+0.095
el+0.088
 feather+0.087
Neg logit contributions
 Hood-0.142
 treadmill-0.136
ussy-0.133
ilda-0.130
ner-0.128
Token a
Token ID257
Feature activation+4.18e-02

Pos logit contributions
 mouth+0.019
 wing+0.018
c+0.017
 its+0.017
el+0.017
Neg logit contributions
 Hood-0.032
 treadmill-0.031
ner-0.030
ussy-0.030
boards-0.030
Token successful
Token ID4388
Feature activation+1.49e-01

Pos logit contributions
 Hood+0.068
ner+0.065
cases+0.065
 treadmill+0.065
fighters+0.064
Neg logit contributions
 mouth-0.062
 wing-0.061
 feather-0.057
 nest-0.057
 its-0.056
Token lear
Token ID22454
Feature activation+1.34e-01

Pos logit contributions
 Hood+0.265
 treadmill+0.253
ner+0.247
ussy+0.245
fighters+0.241
Neg logit contributions
 mouth-0.198
 wing-0.192
el-0.179
c-0.179
 feather-0.173
Tokenner
Token ID1008
Feature activation+4.03e-01

Pos logit contributions
 treadmill+0.101
 Hood+0.100
boards+0.094
oons+0.092
ussy+0.088
Neg logit contributions
 wing-0.288
el-0.285
 mouth-0.276
c-0.273
 its-0.272
Token and
Token ID290
Feature activation+1.39e-02

Pos logit contributions
 Hood+0.679
 treadmill+0.619
ussy+0.615
ner+0.611
fighters+0.589
Neg logit contributions
 mouth-0.569
c-0.548
 wing-0.543
el-0.526
 its-0.511
Token a
Token ID257
Feature activation+4.17e-02

Pos logit contributions
 Hood+0.024
ussy+0.022
 treadmill+0.022
ner+0.022
fighters+0.021
Neg logit contributions
 mouth-0.019
c-0.018
 wing-0.018
el-0.017
 its-0.017
Token good
Token ID922
Feature activation+1.19e-01

Pos logit contributions
 Hood+0.067
ussy+0.061
ner+0.060
 treadmill+0.059
fighters+0.058
Neg logit contributions
 mouth-0.065
 wing-0.062
c-0.059
 feather-0.058
 its-0.057
Token daughter
Token ID4957
Feature activation+8.09e-02

Pos logit contributions
 Hood+0.227
ner+0.218
ussy+0.213
 treadmill+0.212
fighters+0.205
Neg logit contributions
 mouth-0.146
 wing-0.137
c-0.127
 feather-0.125
el-0.123
Token."
Token ID526
Feature activation+1.35e-01

Pos logit contributions
 Hood+0.119
 treadmill+0.107
ner+0.106
ussy+0.106
fighters+0.101
Neg logit contributions
 mouth-0.130
c-0.126
 wing-0.124
el-0.123
 its-0.118
Token kitchen
Token ID9592
Feature activation+1.23e-01

Pos logit contributions
 mouth+0.017
c+0.016
el+0.016
 wing+0.016
 its+0.015
Neg logit contributions
 Hood-0.022
ner-0.020
 treadmill-0.019
ussy-0.019
fighters-0.019
Token?"
Token ID1701
Feature activation+1.17e-01

Pos logit contributions
 Hood+0.196
ner+0.175
 treadmill+0.173
ussy+0.171
fighters+0.166
Neg logit contributions
 mouth-0.176
c-0.174
 wing-0.171
el-0.170
 feathers-0.159
Token The
Token ID383
Feature activation+5.62e-03

Pos logit contributions
 Hood+0.155
ner+0.147
assy+0.133
 treadmill+0.132
ussy+0.131
Neg logit contributions
 wing-0.193
c-0.187
 feather-0.184
el-0.184
 feathers-0.180
Token other
Token ID584
Feature activation+1.20e-01

Pos logit contributions
 Hood+0.009
ner+0.008
ussy+0.008
 treadmill+0.008
assy+0.008
Neg logit contributions
 mouth-0.008
 wing-0.008
c-0.008
el-0.008
 feathers-0.007
Token sp
Token ID599
Feature activation-4.81e-02

Pos logit contributions
 Hood+0.198
ner+0.180
 treadmill+0.178
ussy+0.177
assy+0.169
Neg logit contributions
 mouth-0.174
c-0.167
el-0.164
 wing-0.163
 its-0.154
Tokenoons
Token ID13022
Feature activation+3.92e-01

Pos logit contributions
 mouth+0.093
el+0.092
c+0.091
 wing+0.090
 its+0.086
Neg logit contributions
 Hood-0.054
ner-0.047
 treadmill-0.046
ussy-0.045
fighters-0.042
Token agreed
Token ID4987
Feature activation+4.20e-02

Pos logit contributions
 Hood+0.676
ner+0.613
ussy+0.602
 treadmill+0.600
fighters+0.584
Neg logit contributions
 mouth-0.523
c-0.508
 wing-0.496
el-0.491
 its-0.465
Token and
Token ID290
Feature activation+3.39e-02

Pos logit contributions
 Hood+0.065
ner+0.060
 treadmill+0.059
ussy+0.057
fighters+0.056
Neg logit contributions
 mouth-0.062
c-0.061
el-0.060
 wing-0.060
 feathers-0.056
Token they
Token ID484
Feature activation-6.72e-03

Pos logit contributions
 Hood+0.056
ner+0.050
 treadmill+0.049
ussy+0.049
fighters+0.048
Neg logit contributions
 mouth-0.048
c-0.048
el-0.047
 wing-0.046
light-0.043
Token all
Token ID477
Feature activation+1.03e-01

Pos logit contributions
 mouth+0.009
el+0.008
c+0.008
 wing+0.008
 its+0.008
Neg logit contributions
 Hood-0.012
ner-0.011
 treadmill-0.011
ussy-0.011
assy-0.010
Token moved
Token ID3888
Feature activation-3.59e-02

Pos logit contributions
 Hood+0.146
ussy+0.124
ner+0.124
 treadmill+0.123
fighters+0.121
Neg logit contributions
 mouth-0.171
 wing-0.165
c-0.164
el-0.161
 pouch-0.154
Token.
Token ID13
Feature activation+1.54e-02

Pos logit contributions
el+0.443
 mouth+0.429
 oval+0.421
ric+0.415
 mouths+0.414
Neg logit contributions
 Hood-0.426
 treadmill-0.354
ner-0.343
cases-0.333
ussy-0.330
Token There
Token ID1318
Feature activation+3.24e-02

Pos logit contributions
 Hood+0.023
ner+0.022
 treadmill+0.021
ussy+0.021
fighters+0.020
Neg logit contributions
 wing-0.022
 mouth-0.022
el-0.022
c-0.022
 feather-0.021
Token were
Token ID547
Feature activation+4.15e-03

Pos logit contributions
 Hood+0.056
ner+0.050
 treadmill+0.049
ussy+0.049
fighters+0.048
Neg logit contributions
 mouth-0.044
c-0.042
 wing-0.042
el-0.041
 feathers-0.039
Token big
Token ID1263
Feature activation+1.27e-01

Pos logit contributions
 Hood+0.006
ner+0.006
ussy+0.006
 treadmill+0.006
assy+0.005
Neg logit contributions
 mouth-0.007
c-0.006
 wing-0.006
el-0.006
 its-0.006
Token sp
Token ID599
Feature activation-4.27e-02

Pos logit contributions
 Hood+0.233
ussy+0.217
ner+0.211
ilda+0.209
assy+0.209
Neg logit contributions
 mouth-0.182
 feathers-0.160
el-0.160
c-0.158
 wing-0.155
Tokenoons
Token ID13022
Feature activation+3.88e-01

Pos logit contributions
el+0.088
 mouth+0.084
c+0.084
 wing+0.080
light+0.080
Neg logit contributions
 Hood-0.046
ner-0.039
 treadmill-0.037
ussy-0.036
assy-0.034
Token,
Token ID11
Feature activation+2.28e-02

Pos logit contributions
 Hood+0.667
ner+0.604
 treadmill+0.600
ussy+0.598
fighters+0.577
Neg logit contributions
 mouth-0.528
c-0.510
el-0.500
 wing-0.498
 its-0.472
Token small
Token ID1402
Feature activation+4.83e-02

Pos logit contributions
 Hood+0.035
 treadmill+0.032
ner+0.032
fighters+0.031
ussy+0.031
Neg logit contributions
c-0.033
 wing-0.032
 mouth-0.032
el-0.032
rys-0.030
Token sp
Token ID599
Feature activation-4.79e-02

Pos logit contributions
 Hood+0.081
ussy+0.071
ner+0.071
 treadmill+0.071
assy+0.070
Neg logit contributions
 mouth-0.074
el-0.069
c-0.067
 wing-0.065
 feathers-0.065
Tokenoons
Token ID13022
Feature activation+3.83e-01

Pos logit contributions
el+0.112
 mouth+0.107
c+0.103
 oval+0.101
light+0.101
Neg logit contributions
 Hood-0.040
ner-0.033
 treadmill-0.030
ussy-0.027
assy-0.027
Token,
Token ID11
Feature activation+2.00e-02

Pos logit contributions
 Hood+0.639
ner+0.578
 treadmill+0.570
ussy+0.568
fighters+0.549
Neg logit contributions
 mouth-0.536
c-0.519
 wing-0.508
el-0.507
 its-0.480
Token in
Token ID287
Feature activation-7.25e-03

Pos logit contributions
 Hood+0.089
ner+0.081
 treadmill+0.080
ussy+0.079
fighters+0.076
Neg logit contributions
 mouth-0.072
c-0.070
el-0.069
 wing-0.068
 its-0.064
Token any
Token ID597
Feature activation+7.65e-03

Pos logit contributions
c+0.010
 mouth+0.010
 wing+0.010
el+0.010
light+0.009
Neg logit contributions
 Hood-0.012
ner-0.011
 treadmill-0.011
ussy-0.010
ilda-0.010
Token of
Token ID286
Feature activation-3.41e-02

Pos logit contributions
 Hood+0.013
 treadmill+0.012
ner+0.012
ussy+0.011
fighters+0.011
Neg logit contributions
 mouth-0.010
c-0.010
el-0.010
 wing-0.010
 its-0.009
Token the
Token ID262
Feature activation+4.35e-02

Pos logit contributions
 mouth+0.045
c+0.045
 wing+0.043
el+0.043
light+0.041
Neg logit contributions
 Hood-0.058
ner-0.054
 treadmill-0.053
ussy-0.051
fighters-0.050
Token cup
Token ID6508
Feature activation-1.01e-01

Pos logit contributions
 Hood+0.075
ner+0.068
ussy+0.067
 treadmill+0.067
fighters+0.065
Neg logit contributions
 mouth-0.059
c-0.057
 wing-0.056
el-0.055
 its-0.052
Tokenboards
Token ID12821
Feature activation+3.86e-01

Pos logit contributions
 mouth+0.147
c+0.143
el+0.139
 wing+0.135
light+0.135
Neg logit contributions
 Hood-0.166
ussy-0.145
 treadmill-0.144
ner-0.144
assy-0.138
Token.
Token ID13
Feature activation+3.38e-02

Pos logit contributions
 Hood+0.712
ussy+0.646
 treadmill+0.638
ner+0.636
assy+0.620
Neg logit contributions
 mouth-0.483
c-0.465
el-0.454
 wing-0.450
 its-0.439
Token
Token ID198
Feature activation+6.49e-03

Pos logit contributions
 Hood+0.057
ner+0.052
 treadmill+0.052
ussy+0.051
fighters+0.050
Neg logit contributions
 mouth-0.046
c-0.045
 wing-0.044
el-0.044
 its-0.041
Token
Token ID198
Feature activation+2.53e-03

Pos logit contributions
 Hood+0.012
ner+0.011
 treadmill+0.011
ussy+0.011
fighters+0.010
Neg logit contributions
 mouth-0.008
c-0.008
el-0.007
 wing-0.007
 its-0.007
TokenThe
Token ID464
Feature activation-3.79e-03

Pos logit contributions
 Hood+0.004
ner+0.003
 treadmill+0.003
ussy+0.003
fighters+0.003
Neg logit contributions
 mouth-0.004
c-0.004
 wing-0.004
el-0.004
 feathers-0.004
Token pan
Token ID3425
Feature activation-4.41e-02

Pos logit contributions
 mouth+0.006
 wing+0.005
c+0.005
el+0.005
 feathers+0.005
Neg logit contributions
 Hood-0.006
ner-0.006
ussy-0.006
 treadmill-0.005
assy-0.005
Token sp
Token ID599
Feature activation-4.27e-02

Pos logit contributions
 Hood+0.233
ussy+0.217
ner+0.211
ilda+0.209
assy+0.209
Neg logit contributions
 mouth-0.182
 feathers-0.160
el-0.160
c-0.158
 wing-0.155
Tokenoons
Token ID13022
Feature activation+3.88e-01

Pos logit contributions
el+0.088
 mouth+0.084
c+0.084
 wing+0.080
light+0.080
Neg logit contributions
 Hood-0.046
ner-0.039
 treadmill-0.037
ussy-0.036
assy-0.034
Token,
Token ID11
Feature activation+2.28e-02

Pos logit contributions
 Hood+0.667
ner+0.604
 treadmill+0.600
ussy+0.598
fighters+0.577
Neg logit contributions
 mouth-0.528
c-0.510
el-0.500
 wing-0.498
 its-0.472
Token small
Token ID1402
Feature activation+4.83e-02

Pos logit contributions
 Hood+0.035
 treadmill+0.032
ner+0.032
fighters+0.031
ussy+0.031
Neg logit contributions
c-0.033
 wing-0.032
 mouth-0.032
el-0.032
rys-0.030
Token sp
Token ID599
Feature activation-4.79e-02

Pos logit contributions
 Hood+0.081
ussy+0.071
ner+0.071
 treadmill+0.071
assy+0.070
Neg logit contributions
 mouth-0.074
el-0.069
c-0.067
 wing-0.065
 feathers-0.065
Tokenoons
Token ID13022
Feature activation+3.83e-01

Pos logit contributions
el+0.112
 mouth+0.107
c+0.103
 oval+0.101
light+0.101
Neg logit contributions
 Hood-0.040
ner-0.033
 treadmill-0.030
ussy-0.027
assy-0.027
Token,
Token ID11
Feature activation+2.00e-02

Pos logit contributions
 Hood+0.639
ner+0.578
 treadmill+0.570
ussy+0.568
fighters+0.549
Neg logit contributions
 mouth-0.536
c-0.519
 wing-0.508
el-0.507
 its-0.480
Token and
Token ID290
Feature activation+2.87e-02

Pos logit contributions
 Hood+0.029
 treadmill+0.027
ner+0.026
fighters+0.026
ussy+0.025
Neg logit contributions
c-0.030
 wing-0.029
 mouth-0.029
el-0.029
rys-0.028
Token teaspoons
Token ID47234
Feature activation+6.82e-02

Pos logit contributions
 Hood+0.044
ner+0.040
 treadmill+0.039
ussy+0.038
fighters+0.038
Neg logit contributions
c-0.042
 mouth-0.042
 wing-0.041
el-0.041
light-0.039
Token.
Token ID13
Feature activation+1.86e-03

Pos logit contributions
 Hood+0.109
ner+0.099
ussy+0.096
 treadmill+0.095
oons+0.094
Neg logit contributions
 mouth-0.101
 wing-0.096
c-0.096
el-0.093
 its-0.089
Token Sue
Token ID26089
Feature activation-8.82e-02

Pos logit contributions
 Hood+0.003
ner+0.003
ussy+0.003
 treadmill+0.003
fighters+0.003
Neg logit contributions
 mouth-0.003
 wing-0.003
c-0.003
el-0.003
 feather-0.002
Token a
Token ID257
Feature activation+5.92e-02

Pos logit contributions
 Hood+0.017
 treadmill+0.016
ner+0.016
ussy+0.016
fighters+0.016
Neg logit contributions
 mouth-0.014
c-0.013
el-0.013
 wing-0.013
 its-0.013
Token boy
Token ID2933
Feature activation+6.63e-02

Pos logit contributions
 Hood+0.105
ives+0.101
cases+0.099
ussy+0.099
drivers+0.098
Neg logit contributions
 mouth-0.081
 wing-0.076
 feather-0.075
 nest-0.074
c-0.073
Token who
Token ID508
Feature activation-6.00e-02

Pos logit contributions
 Hood+0.090
working+0.076
ner+0.073
assy+0.072
 treadmill+0.071
Neg logit contributions
 mouth-0.101
 mouths-0.098
 tal-0.097
 pouch-0.096
 wing-0.095
Token loved
Token ID6151
Feature activation+1.54e-01

Pos logit contributions
 mouth+0.097
c+0.096
el+0.092
 wing+0.092
 its+0.089
Neg logit contributions
 Hood-0.089
ner-0.083
 treadmill-0.080
ussy-0.079
fighters-0.076
Token sp
Token ID599
Feature activation-5.30e-02

Pos logit contributions
 Hood+0.235
 treadmill+0.210
ner+0.199
ussy+0.191
assy+0.188
Neg logit contributions
c-0.238
 mouth-0.230
el-0.229
light-0.220
 wing-0.220
Tokenoons
Token ID13022
Feature activation+3.82e-01

Pos logit contributions
el+0.099
 mouth+0.095
c+0.095
 wing+0.091
 feathers+0.089
Neg logit contributions
 Hood-0.065
ner-0.058
 treadmill-0.055
fighters-0.053
ussy-0.052
Token.
Token ID13
Feature activation+2.20e-02

Pos logit contributions
 Hood+0.599
ner+0.527
ussy+0.522
 treadmill+0.521
assy+0.497
Neg logit contributions
 mouth-0.575
c-0.558
 wing-0.549
el-0.546
light-0.516
Token He
Token ID679
Feature activation+2.46e-02

Pos logit contributions
 Hood+0.033
 treadmill+0.030
ussy+0.030
ner+0.030
fighters+0.029
Neg logit contributions
 mouth-0.033
 wing-0.033
c-0.033
el-0.032
light-0.031
Token loved
Token ID6151
Feature activation+1.52e-01

Pos logit contributions
ner+0.046
 Hood+0.045
 treadmill+0.044
ussy+0.043
assy+0.042
Neg logit contributions
c-0.030
 mouth-0.029
el-0.029
 its-0.027
 wing-0.027
Token them
Token ID606
Feature activation+3.60e-02

Pos logit contributions
 Hood+0.250
 treadmill+0.221
ner+0.217
ussy+0.215
oons+0.209
Neg logit contributions
 mouth-0.217
c-0.216
 wing-0.211
el-0.207
light-0.199
Token so
Token ID523
Feature activation-8.98e-02

Pos logit contributions
 Hood+0.057
 treadmill+0.049
ussy+0.049
ner+0.048
fighters+0.046
Neg logit contributions
 mouth-0.054
c-0.054
 wing-0.052
el-0.050
rys-0.050
Token he
Token ID339
Feature activation+4.35e-02

Pos logit contributions
c+0.031
 wing+0.031
 mouth+0.030
el+0.030
rys+0.029
Neg logit contributions
 Hood-0.021
oons-0.017
ner-0.017
 treadmill-0.017
ussy-0.017
Token would
Token ID561
Feature activation+1.03e-01

Pos logit contributions
ner+0.081
 Hood+0.079
 treadmill+0.079
ives+0.075
assy+0.075
Neg logit contributions
c-0.050
 mouth-0.049
el-0.046
 wing-0.045
 feathers-0.045
Token look
Token ID804
Feature activation-3.97e-02

Pos logit contributions
 Hood+0.153
 treadmill+0.142
ner+0.139
ussy+0.134
fighters+0.130
Neg logit contributions
c-0.158
 mouth-0.157
el-0.153
 wing-0.150
light-0.144
Token for
Token ID329
Feature activation-2.97e-02

Pos logit contributions
 mouth+0.040
 its+0.038
el+0.035
c+0.033
 wing+0.033
Neg logit contributions
 Hood-0.081
ner-0.081
drivers-0.078
fighters-0.078
 treadmill-0.077
Token sp
Token ID599
Feature activation-5.43e-02

Pos logit contributions
 mouth+0.039
c+0.038
el+0.037
 wing+0.037
 its+0.035
Neg logit contributions
 Hood-0.052
ner-0.047
 treadmill-0.047
ussy-0.047
fighters-0.045
Tokenoons
Token ID13022
Feature activation+3.81e-01

Pos logit contributions
el+0.124
c+0.112
 mouth+0.109
 oval+0.108
light+0.107
Neg logit contributions
 Hood-0.048
ner-0.041
 treadmill-0.036
 Cop-0.034
fighters-0.032
Token in
Token ID287
Feature activation-2.45e-02

Pos logit contributions
 Hood+0.699
ner+0.657
 treadmill+0.653
ussy+0.640
fighters+0.624
Neg logit contributions
 mouth-0.466
c-0.445
el-0.439
 wing-0.426
 its-0.421
Token every
Token ID790
Feature activation+4.23e-02

Pos logit contributions
 mouth+0.037
 wing+0.037
c+0.037
el+0.034
light+0.034
Neg logit contributions
 Hood-0.037
ner-0.032
 treadmill-0.032
ussy-0.031
oons-0.030
Token corner
Token ID5228
Feature activation-1.97e-02

Pos logit contributions
 Hood+0.076
ner+0.068
 treadmill+0.067
ussy+0.065
oons+0.064
Neg logit contributions
c-0.052
 mouth-0.052
 wing-0.052
el-0.050
 feathers-0.048
Token.
Token ID13
Feature activation+2.02e-02

Pos logit contributions
 mouth+0.029
 wing+0.029
el+0.028
c+0.028
 feather+0.027
Neg logit contributions
 Hood-0.030
 treadmill-0.027
ner-0.027
ussy-0.026
oons-0.026
Token One
Token ID1881
Feature activation-6.18e-02

Pos logit contributions
 Hood+0.032
 treadmill+0.029
ussy+0.029
ner+0.029
fighters+0.027
Neg logit contributions
 wing-0.030
 mouth-0.029
el-0.028
c-0.028
 feathers-0.027
Token his
Token ID465
Feature activation-1.04e-01

Pos logit contributions
 Hood+0.022
 treadmill+0.020
ner+0.020
ussy+0.020
assy+0.019
Neg logit contributions
 mouth-0.015
el-0.014
 its-0.014
c-0.014
 wing-0.013
Token face
Token ID1986
Feature activation-2.05e-01

Pos logit contributions
rys+0.121
 peac+0.118
c+0.114
light+0.110
Its+0.108
Neg logit contributions
 Hood-0.146
ussy-0.137
 treadmill-0.137
driver-0.136
 downstairs-0.131
Token.
Token ID13
Feature activation-1.86e-02

Pos logit contributions
 mouth+0.326
 wing+0.321
c+0.320
el+0.316
rys+0.316
Neg logit contributions
 Hood-0.264
ussy-0.244
ner-0.237
 treadmill-0.230
fighters-0.224
Token Max
Token ID5436
Feature activation+1.43e-01

Pos logit contributions
 mouth+0.026
c+0.025
el+0.025
 wing+0.024
 its+0.023
Neg logit contributions
 Hood-0.032
ner-0.029
 treadmill-0.028
ussy-0.028
fighters-0.027
Token y
Token ID331
Feature activation+4.44e-03

Pos logit contributions
 Hood+0.232
ner+0.209
 treadmill+0.206
ussy+0.206
fighters+0.199
Neg logit contributions
 mouth-0.212
c-0.203
el-0.199
 wing-0.199
 its-0.190
Tokenel
Token ID417
Feature activation-3.80e-01

Pos logit contributions
 Hood+0.009
ussy+0.009
assy+0.009
oons+0.009
ives+0.009
Neg logit contributions
c-0.004
 wing-0.003
 mouth-0.003
 peac-0.003
 cardinal-0.003
Tokenps
Token ID862
Feature activation+1.66e-01

Pos logit contributions
c+0.714
 mouth+0.711
 wing+0.688
light+0.681
el+0.680
Neg logit contributions
 Hood-0.417
ussy-0.344
ner-0.330
fighters-0.330
 treadmill-0.316
Token and
Token ID290
Feature activation+7.58e-03

Pos logit contributions
 Hood+0.215
ner+0.189
 treadmill+0.189
ussy+0.188
ives+0.186
Neg logit contributions
c-0.268
 wing-0.264
 mouth-0.260
el-0.259
light-0.257
Token backs
Token ID12983
Feature activation+6.86e-02

Pos logit contributions
 Hood+0.010
 treadmill+0.009
ner+0.009
ussy+0.009
fighters+0.009
Neg logit contributions
c-0.013
el-0.012
 wing-0.012
 mouth-0.012
light-0.012
Token away
Token ID1497
Feature activation-1.26e-01

Pos logit contributions
 Hood+0.120
ner+0.108
 treadmill+0.107
ussy+0.105
fighters+0.102
Neg logit contributions
 mouth-0.089
c-0.087
el-0.085
 wing-0.085
 oval-0.083
Token.
Token ID13
Feature activation-1.85e-02

Pos logit contributions
c+0.191
 mouth+0.188
el+0.184
 wing+0.182
light+0.175
Neg logit contributions
 Hood-0.195
ner-0.174
 treadmill-0.171
ussy-0.171
fighters-0.166
Token his
Token ID465
Feature activation-9.75e-02

Pos logit contributions
c+0.165
 mouth+0.162
 wing+0.158
el+0.155
light+0.151
Neg logit contributions
 Hood-0.218
ner-0.198
 treadmill-0.198
ussy-0.195
fighters-0.188
Token paw
Token ID41827
Feature activation-2.12e-01

Pos logit contributions
c+0.109
light+0.108
el+0.098
rys+0.095
ric+0.092
Neg logit contributions
 Hood-0.178
 treadmill-0.166
driver-0.157
drivers-0.156
ner-0.153
Token.
Token ID13
Feature activation-1.58e-02

Pos logit contributions
c+0.340
el+0.333
 mouth+0.330
rys+0.325
 wing+0.323
Neg logit contributions
 Hood-0.281
 treadmill-0.252
ner-0.248
ussy-0.243
ives-0.238
Token Jen
Token ID13364
Feature activation-1.75e-01

Pos logit contributions
 wing+0.027
c+0.027
 feather+0.026
el+0.026
light+0.026
Neg logit contributions
 Hood-0.020
 treadmill-0.019
ner-0.019
ussy-0.019
fighters-0.018
Token y
Token ID331
Feature activation+8.99e-03

Pos logit contributions
 mouth+0.258
c+0.236
el+0.235
 mouths+0.221
 feathers+0.219
Neg logit contributions
 Hood-0.308
ner-0.277
ilda-0.276
assy-0.271
fighters-0.266
Tokenel
Token ID417
Feature activation-3.78e-01

Pos logit contributions
 Hood+0.020
ussy+0.019
assy+0.019
oons+0.019
 treadmill+0.019
Neg logit contributions
c-0.006
 mouth-0.006
 wing-0.006
 cardinal-0.005
 peac-0.005
Tokenps
Token ID862
Feature activation+1.69e-01

Pos logit contributions
 mouth+0.577
c+0.570
 wing+0.552
el+0.548
light+0.533
Neg logit contributions
 Hood-0.568
ussy-0.497
ner-0.494
 treadmill-0.485
fighters-0.479
Token and
Token ID290
Feature activation+1.08e-02

Pos logit contributions
 Hood+0.243
 treadmill+0.219
ner+0.218
ussy+0.217
fighters+0.208
Neg logit contributions
c-0.257
 mouth-0.255
 wing-0.251
el-0.248
light-0.239
Token runs
Token ID4539
Feature activation+1.27e-01

Pos logit contributions
 Hood+0.014
 treadmill+0.013
ner+0.013
ussy+0.012
fighters+0.012
Neg logit contributions
c-0.018
 wing-0.018
 mouth-0.018
el-0.018
light-0.017
Token away
Token ID1497
Feature activation-1.22e-01

Pos logit contributions
 Hood+0.301
ilda+0.296
ner+0.293
ussy+0.291
assy+0.290
Neg logit contributions
 its-0.088
 mouth-0.085
c-0.064
 direction-0.061
 feathers-0.059
Token.
Token ID13
Feature activation-1.59e-02

Pos logit contributions
 mouth+0.172
c+0.167
el+0.163
 wing+0.163
 its+0.153
Neg logit contributions
 Hood-0.206
ner-0.187
 treadmill-0.185
ussy-0.184
fighters-0.177
Token.
Token ID13
Feature activation-1.70e-02

Pos logit contributions
c+0.120
 mouth+0.113
 wing+0.112
el+0.111
light+0.108
Neg logit contributions
 Hood-0.085
ner-0.077
 treadmill-0.076
ussy-0.076
fighters-0.074
Token She
Token ID1375
Feature activation+1.00e-03

Pos logit contributions
c+0.027
 mouth+0.027
 wing+0.026
el+0.026
light+0.025
Neg logit contributions
 Hood-0.025
 treadmill-0.022
ner-0.022
ussy-0.022
assy-0.021
Token likes
Token ID7832
Feature activation+6.92e-02

Pos logit contributions
 Hood+0.002
ner+0.002
ilda+0.002
assy+0.002
ussy+0.002
Neg logit contributions
 mouth-0.001
c-0.001
el-0.001
 feathers-0.001
 mouths-0.001
Token that
Token ID326
Feature activation-3.87e-02

Pos logit contributions
 Hood+0.104
 treadmill+0.093
ner+0.091
ussy+0.091
fighters+0.087
Neg logit contributions
 mouth-0.107
c-0.107
el-0.104
 wing-0.103
light-0.100
Token past
Token ID1613
Feature activation-7.54e-02

Pos logit contributions
light+0.067
el+0.064
c+0.063
ric+0.062
 wing+0.062
Neg logit contributions
 Hood-0.051
 treadmill-0.050
ner-0.042
cakes-0.040
ussy-0.040
Tokenel
Token ID417
Feature activation-3.78e-01

Pos logit contributions
 mouth+0.096
c+0.093
 wing+0.091
el+0.089
light+0.085
Neg logit contributions
 Hood-0.136
ner-0.124
 treadmill-0.123
ussy-0.122
fighters-0.118
Token.
Token ID13
Feature activation-1.70e-02

Pos logit contributions
 mouth+0.536
c+0.518
el+0.508
 wing+0.504
 its+0.480
Neg logit contributions
 Hood-0.633
ner-0.574
 treadmill-0.567
ussy-0.563
fighters-0.541
Token It
Token ID632
Feature activation-1.10e-01

Pos logit contributions
 mouth+0.025
c+0.025
 wing+0.025
el+0.024
light+0.023
Neg logit contributions
 Hood-0.026
ner-0.024
 treadmill-0.024
ussy-0.024
fighters-0.023
Token is
Token ID318
Feature activation-2.30e-02

Pos logit contributions
 mouth+0.166
 wing+0.165
el+0.161
c+0.161
light+0.157
Neg logit contributions
 Hood-0.171
ussy-0.149
 treadmill-0.147
fighters-0.144
ner-0.144
Token her
Token ID607
Feature activation-6.14e-02

Pos logit contributions
el+0.040
c+0.039
light+0.039
 pouch+0.036
 wing+0.035
Neg logit contributions
 Hood-0.029
ussy-0.026
 treadmill-0.025
fighters-0.023
 scientists-0.022
Token favorite
Token ID4004
Feature activation-6.40e-02

Pos logit contributions
c+0.087
el+0.087
light+0.087
 mouth+0.081
 its+0.080
Neg logit contributions
 Hood-0.098
 treadmill-0.085
ussy-0.082
ner-0.082
working-0.079
Token.
Token ID13
Feature activation-1.32e-02

Pos logit contributions
 Hood+0.018
ussy+0.016
ner+0.016
assy+0.016
boards+0.015
Neg logit contributions
c-0.025
 wing-0.024
 mouth-0.024
 pouch-0.023
light-0.022
Token She
Token ID1375
Feature activation+4.18e-03

Pos logit contributions
 mouth+0.018
c+0.017
el+0.017
 wing+0.017
 its+0.016
Neg logit contributions
 Hood-0.023
ner-0.021
 treadmill-0.021
ussy-0.021
fighters-0.020
Token recognizes
Token ID21570
Feature activation+2.93e-02

Pos logit contributions
 Hood+0.008
ner+0.007
 treadmill+0.007
ussy+0.007
working+0.007
Neg logit contributions
 mouth-0.006
 eyes-0.005
el-0.005
 be-0.005
c-0.005
Token the
Token ID262
Feature activation+7.82e-03

Pos logit contributions
 Hood+0.051
ner+0.046
 treadmill+0.046
ussy+0.046
fighters+0.044
Neg logit contributions
 mouth-0.040
c-0.038
el-0.037
 wing-0.037
 its-0.035
Token bird
Token ID6512
Feature activation-1.15e-01

Pos logit contributions
 Hood+0.012
ner+0.010
 treadmill+0.010
 manager+0.010
fighters+0.010
Neg logit contributions
light-0.011
el-0.010
c-0.010
rys-0.010
 mouth-0.010
Tokenc
Token ID66
Feature activation-3.77e-01

Pos logit contributions
 mouth+0.077
rys+0.067
 fin+0.066
 mouths+0.066
 direction+0.062
Neg logit contributions
 Hood-0.271
ussy-0.240
ner-0.238
fighters-0.233
ilda-0.232
Tokenage
Token ID496
Feature activation-1.27e-01

Pos logit contributions
 mouth+0.408
c+0.379
el+0.375
 wing+0.369
 its+0.360
Neg logit contributions
 Hood-0.738
ussy-0.674
ner-0.673
 treadmill-0.667
fighters-0.662
Token.
Token ID13
Feature activation-1.29e-02

Pos logit contributions
 mouth+0.161
c+0.156
 wing+0.152
el+0.151
 its+0.145
Neg logit contributions
 Hood-0.232
 treadmill-0.213
ner-0.212
ussy-0.210
assy-0.202
Token It
Token ID632
Feature activation-1.07e-01

Pos logit contributions
 mouth+0.018
c+0.018
 wing+0.017
el+0.017
 feathers+0.016
Neg logit contributions
 Hood-0.021
ner-0.020
ussy-0.019
 treadmill-0.019
assy-0.018
Token belongs
Token ID14448
Feature activation-1.13e-01

Pos logit contributions
 mouth+0.171
 wing+0.168
el+0.157
 fin+0.156
rys+0.155
Neg logit contributions
 Hood-0.157
fighters-0.131
ner-0.130
ussy-0.128
ilda-0.122
Token to
Token ID284
Feature activation-2.12e-02

Pos logit contributions
 mouth+0.144
c+0.139
el+0.136
 wing+0.136
 its+0.128
Neg logit contributions
 Hood-0.204
ner-0.187
 treadmill-0.185
ussy-0.184
fighters-0.178
Token can
Token ID460
Feature activation+8.02e-02

Pos logit contributions
 mouth+0.062
 wing+0.058
c+0.058
 be+0.054
el+0.054
Neg logit contributions
 Hood-0.090
 treadmill-0.085
ilda-0.084
assy-0.083
ussy-0.082
Token use
Token ID779
Feature activation-2.30e-02

Pos logit contributions
 Hood+0.154
 treadmill+0.139
ussy+0.137
fighters+0.136
boards+0.135
Neg logit contributions
el-0.099
 mouth-0.096
 wing-0.091
c-0.090
 its-0.088
Token the
Token ID262
Feature activation+1.01e-02

Pos logit contributions
c+0.031
 mouth+0.030
 wing+0.029
light+0.028
el+0.028
Neg logit contributions
 Hood-0.038
ner-0.036
 treadmill-0.034
ussy-0.034
ilda-0.032
Token green
Token ID4077
Feature activation-8.98e-02

Pos logit contributions
 Hood+0.014
 treadmill+0.013
ner+0.013
working+0.013
fighters+0.012
Neg logit contributions
light-0.015
c-0.014
el-0.014
ectar-0.013
seed-0.013
Token past
Token ID1613
Feature activation-7.47e-02

Pos logit contributions
rys+0.105
c+0.103
light+0.100
el+0.098
 wing+0.096
Neg logit contributions
 Hood-0.135
ner-0.134
working-0.126
 treadmill-0.123
oons-0.122
Tokenel
Token ID417
Feature activation-3.77e-01

Pos logit contributions
 pouch+0.043
 mouth+0.042
 wing+0.038
rys+0.036
 peac+0.034
Neg logit contributions
 Hood-0.167
oons-0.160
ner-0.159
fighters-0.150
ussy-0.149
Token first
Token ID717
Feature activation+1.39e-01

Pos logit contributions
 mouth+0.547
 pouch+0.537
 wing+0.530
c+0.526
rys+0.526
Neg logit contributions
 Hood-0.497
oons-0.450
ives-0.439
fighters-0.437
ussy-0.431
Token,
Token ID11
Feature activation+3.35e-03

Pos logit contributions
 Hood+0.168
ner+0.165
oons+0.159
working+0.156
 treadmill+0.156
Neg logit contributions
c-0.197
el-0.196
rys-0.188
ric-0.185
 mouth-0.184
Token and
Token ID290
Feature activation+1.07e-02

Pos logit contributions
ner+0.005
 treadmill+0.004
 Hood+0.004
ussy+0.004
driver+0.004
Neg logit contributions
el-0.005
c-0.005
light-0.004
ric-0.004
 pouch-0.004
Token then
Token ID788
Feature activation+2.09e-02

Pos logit contributions
 Hood+0.017
ner+0.016
 treadmill+0.015
ussy+0.015
fighters+0.014
Neg logit contributions
c-0.015
el-0.015
 mouth-0.014
 wing-0.014
light-0.014
Token give
Token ID1577
Feature activation-1.08e-03

Pos logit contributions
 Hood+0.041
ussy+0.037
 treadmill+0.037
assy+0.036
ner+0.036
Neg logit contributions
 mouth-0.025
 wing-0.023
c-0.023
el-0.023
 its-0.022
Token can
Token ID460
Feature activation+8.08e-02

Pos logit contributions
 mouth+0.057
 wing+0.052
 be+0.048
c+0.047
 nest+0.045
Neg logit contributions
 Hood-0.097
ilda-0.092
 treadmill-0.091
oons-0.089
ussy-0.088
Token use
Token ID779
Feature activation-2.22e-02

Pos logit contributions
 Hood+0.162
 treadmill+0.144
fighters+0.143
ussy+0.143
boards+0.141
Neg logit contributions
el-0.094
 mouth-0.094
 be-0.090
 wing-0.087
c-0.083
Token the
Token ID262
Feature activation+1.17e-02

Pos logit contributions
c+0.026
 mouth+0.026
 wing+0.025
light+0.024
el+0.024
Neg logit contributions
 Hood-0.041
ner-0.038
 treadmill-0.037
ussy-0.036
assy-0.035
Token red
Token ID2266
Feature activation-1.08e-01

Pos logit contributions
 Hood+0.018
 treadmill+0.016
ner+0.016
working+0.015
fighters+0.015
Neg logit contributions
c-0.017
light-0.017
el-0.017
 wing-0.016
 mouth-0.015
Token past
Token ID1613
Feature activation-7.36e-02

Pos logit contributions
c+0.141
rys+0.137
light+0.136
el+0.130
ric+0.129
Neg logit contributions
 Hood-0.164
ner-0.157
 treadmill-0.147
driver-0.145
boards-0.144
Tokenel
Token ID417
Feature activation-3.76e-01

Pos logit contributions
 mouth+0.054
 pouch+0.052
 wing+0.051
 fin+0.047
rys+0.047
Neg logit contributions
 Hood-0.158
oons-0.147
ner-0.147
fighters-0.140
ussy-0.138
Token first
Token ID717
Feature activation+1.40e-01

Pos logit contributions
 mouth+0.534
 pouch+0.519
c+0.515
 wing+0.513
rys+0.503
Neg logit contributions
 Hood-0.529
fighters-0.470
oons-0.469
ives-0.464
ner-0.458
Token,
Token ID11
Feature activation+4.16e-03

Pos logit contributions
 Hood+0.184
ner+0.181
oons+0.176
driver+0.174
 treadmill+0.172
Neg logit contributions
c-0.182
el-0.182
rys-0.175
ric-0.174
 wing-0.170
Token and
Token ID290
Feature activation+1.13e-02

Pos logit contributions
ner+0.006
 treadmill+0.006
ussy+0.005
 Hood+0.005
 scientists+0.005
Neg logit contributions
light-0.006
el-0.006
c-0.006
ric-0.005
 tal-0.005
Token then
Token ID788
Feature activation+2.27e-02

Pos logit contributions
 Hood+0.019
ner+0.017
 treadmill+0.017
ussy+0.017
fighters+0.016
Neg logit contributions
c-0.015
el-0.015
 mouth-0.015
 wing-0.015
light-0.014
Token give
Token ID1577
Feature activation-1.10e-03

Pos logit contributions
 Hood+0.046
ussy+0.042
 treadmill+0.041
assy+0.040
ner+0.040
Neg logit contributions
 mouth-0.026
 wing-0.024
el-0.024
c-0.024
 its-0.024
Token He
Token ID679
Feature activation+5.06e-03

Pos logit contributions
 mouth+0.017
 its+0.016
 wing+0.015
el+0.014
c+0.014
Neg logit contributions
 Hood-0.026
ussy-0.024
ner-0.024
 treadmill-0.024
ilda-0.023
Token shows
Token ID2523
Feature activation-1.00e-02

Pos logit contributions
 Hood+0.009
ner+0.009
assy+0.009
ilda+0.009
 treadmill+0.009
Neg logit contributions
 mouth-0.007
 eyes-0.006
el-0.006
c-0.006
 its-0.006
Token Mom
Token ID11254
Feature activation-7.07e-02

Pos logit contributions
 mouth+0.013
c+0.013
 wing+0.012
el+0.012
light+0.012
Neg logit contributions
 Hood-0.018
ner-0.016
ussy-0.016
 treadmill-0.016
fighters-0.015
Token the
Token ID262
Feature activation+8.26e-03

Pos logit contributions
 mouth+0.086
c+0.083
el+0.080
 its+0.079
 wing+0.077
Neg logit contributions
 Hood-0.133
ner-0.125
 treadmill-0.123
ussy-0.121
fighters-0.118
Token bird
Token ID6512
Feature activation-1.14e-01

Pos logit contributions
 Hood+0.013
ner+0.011
fighters+0.010
 treadmill+0.010
driver+0.010
Neg logit contributions
el-0.012
c-0.012
light-0.012
rys-0.011
 mouth-0.011
Tokenc
Token ID66
Feature activation-3.75e-01

Pos logit contributions
 mouth+0.175
rys+0.172
 mouths+0.167
Aw+0.166
 pouch+0.164
Neg logit contributions
 Hood-0.166
cakes-0.131
ussy-0.131
fighters-0.127
oons-0.125
Tokenage
Token ID496
Feature activation-1.26e-01

Pos logit contributions
 mouth+0.383
c+0.364
el+0.354
 wing+0.353
 its+0.327
Neg logit contributions
 Hood-0.774
ner-0.713
 treadmill-0.707
ussy-0.706
fighters-0.687
Token.
Token ID13
Feature activation-1.36e-02

Pos logit contributions
 mouth+0.197
c+0.196
 wing+0.191
el+0.191
rys+0.187
Neg logit contributions
 Hood-0.178
ner-0.156
 treadmill-0.155
ussy-0.155
fighters-0.153
Token
Token ID198
Feature activation-1.97e-02

Pos logit contributions
c+0.021
el+0.021
 wing+0.020
 mouth+0.020
light+0.019
Neg logit contributions
 Hood-0.020
ner-0.018
 treadmill-0.018
ussy-0.018
fighters-0.017
Token
Token ID198
Feature activation-1.99e-02

Pos logit contributions
 its+0.018
 mouth+0.017
c+0.016
el+0.016
 wing+0.015
Neg logit contributions
 Hood-0.045
 treadmill-0.042
ussy-0.041
ner-0.041
working-0.040
Token"
Token ID1
Feature activation+8.71e-02

Pos logit contributions
 mouth+0.030
c+0.029
 wing+0.029
el+0.029
 its+0.028
Neg logit contributions
 Hood-0.032
 treadmill-0.029
ner-0.028
ussy-0.028
ilda-0.027
Token,
Token ID11
Feature activation+2.91e-03

Pos logit contributions
 mouth+0.114
el+0.104
c+0.102
 its+0.102
 wing+0.100
Neg logit contributions
 Hood-0.179
ussy-0.166
 treadmill-0.164
ner-0.160
assy-0.158
Token Lily
Token ID20037
Feature activation+3.78e-02

Pos logit contributions
 Hood+0.004
ner+0.004
 treadmill+0.003
ussy+0.003
fighters+0.003
Neg logit contributions
c-0.005
 wing-0.005
 mouth-0.005
el-0.005
light-0.005
Token saw
Token ID2497
Feature activation+2.69e-02

Pos logit contributions
 Hood+0.076
ner+0.074
 treadmill+0.072
ussy+0.071
fighters+0.070
Neg logit contributions
 mouth-0.042
c-0.040
el-0.039
 wing-0.037
light-0.035
Token a
Token ID257
Feature activation+3.85e-02

Pos logit contributions
 Hood+0.050
ner+0.047
ussy+0.047
 treadmill+0.046
fighters+0.045
Neg logit contributions
 mouth-0.034
 its-0.032
 wing-0.032
c-0.031
el-0.031
Token bird
Token ID6512
Feature activation-1.16e-01

Pos logit contributions
 Hood+0.059
 treadmill+0.053
ner+0.053
ussy+0.051
fighters+0.049
Neg logit contributions
c-0.058
el-0.057
 mouth-0.057
 wing-0.055
light-0.054
Tokenc
Token ID66
Feature activation-3.75e-01

Pos logit contributions
 mouths+0.216
 mouth+0.202
 fins+0.197
 tal+0.196
Aw+0.195
Neg logit contributions
 Hood-0.131
ner-0.110
 upstairs-0.095
he-0.093
 outside-0.092
Tokenage
Token ID496
Feature activation-1.26e-01

Pos logit contributions
 mouth+0.489
 fin+0.441
el+0.437
c+0.437
 its+0.434
Neg logit contributions
 Hood-0.633
ives-0.583
cakes-0.574
assy-0.570
fighters-0.568
Token in
Token ID287
Feature activation-4.17e-02

Pos logit contributions
 mouth+0.182
c+0.171
el+0.167
 wing+0.166
 mouths+0.165
Neg logit contributions
 Hood-0.206
ner-0.187
 treadmill-0.179
ussy-0.176
fighters-0.174
Token the
Token ID262
Feature activation+8.66e-03

Pos logit contributions
 mouth+0.068
 wing+0.066
c+0.065
light+0.063
 fin+0.062
Neg logit contributions
 Hood-0.059
ner-0.052
assy-0.048
 treadmill-0.048
ussy-0.048
Token yard
Token ID12699
Feature activation-6.02e-02

Pos logit contributions
 Hood+0.011
 treadmill+0.010
assy+0.009
 grocery+0.009
ner+0.009
Neg logit contributions
 fin-0.014
 fins-0.013
el-0.013
 mouth-0.013
 tilted-0.013
Token.
Token ID13
Feature activation-6.42e-03

Pos logit contributions
rys+0.107
 mouth+0.104
 wing+0.103
 tal+0.103
 feather+0.102
Neg logit contributions
 Hood-0.068
ner-0.065
boards-0.059
 treadmill-0.057
fighters-0.056
Token flies
Token ID17607
Feature activation-4.93e-02

Pos logit contributions
 mouth+0.169
 wing+0.165
el+0.162
c+0.159
light+0.155
Neg logit contributions
 Hood-0.171
ussy-0.148
ner-0.148
 treadmill-0.145
fighters-0.144
Token out
Token ID503
Feature activation+5.36e-02

Pos logit contributions
 mouth+0.068
c+0.066
 wing+0.065
el+0.064
light+0.061
Neg logit contributions
 Hood-0.084
ner-0.074
ussy-0.073
 treadmill-0.073
fighters-0.072
Token of
Token ID286
Feature activation-7.28e-02

Pos logit contributions
 Hood+0.087
ner+0.081
ussy+0.075
fighters+0.075
 treadmill+0.073
Neg logit contributions
 mouth-0.075
c-0.073
 wing-0.071
 pouch-0.068
 feather-0.067
Token the
Token ID262
Feature activation+1.17e-02

Pos logit contributions
c+0.124
light+0.119
 mouth+0.118
 wing+0.116
el+0.115
Neg logit contributions
 Hood-0.096
ner-0.084
 treadmill-0.079
fighters-0.074
ussy-0.072
Token bird
Token ID6512
Feature activation-1.11e-01

Pos logit contributions
 Hood+0.016
 treadmill+0.013
ner+0.013
 stairs+0.013
boards+0.012
Neg logit contributions
 fin-0.018
light-0.018
c-0.018
 mouth-0.018
 its-0.018
Tokenc
Token ID66
Feature activation-3.75e-01

Pos logit contributions
rys+0.159
 fin+0.153
 mouth+0.152
Aw+0.149
el+0.149
Neg logit contributions
 Hood-0.160
ives-0.140
ner-0.138
cases-0.128
 treadmill-0.126
Tokenage
Token ID496
Feature activation-1.23e-01

Pos logit contributions
 mouth+0.442
el+0.424
 its+0.413
 fin+0.411
 wing+0.404
Neg logit contributions
 Hood-0.657
ives-0.611
assy-0.601
boards-0.600
 treadmill-0.600
Token.
Token ID13
Feature activation-6.87e-03

Pos logit contributions
 mouth+0.178
c+0.175
el+0.172
 wing+0.172
 fin+0.162
Neg logit contributions
 Hood-0.192
ner-0.176
 treadmill-0.169
ussy-0.168
fighters-0.164
Token It
Token ID632
Feature activation-1.02e-01

Pos logit contributions
c+0.008
el+0.008
 mouth+0.008
 wing+0.008
light+0.008
Neg logit contributions
 Hood-0.012
ner-0.011
ussy-0.011
 treadmill-0.011
fighters-0.010
Token is
Token ID318
Feature activation-1.56e-02

Pos logit contributions
 mouth+0.150
 wing+0.150
el+0.149
c+0.144
light+0.143
Neg logit contributions
 Hood-0.156
ussy-0.132
ner-0.131
fighters-0.128
 treadmill-0.127
Token happy
Token ID3772
Feature activation-6.08e-03

Pos logit contributions
c+0.026
el+0.026
rys+0.025
light+0.024
 pouch+0.023
Neg logit contributions
 Hood-0.018
ussy-0.015
ner-0.014
 downstairs-0.014
ilda-0.014
Token "
Token ID366
Feature activation+4.65e-02

Pos logit contributions
 mouth+0.022
 its+0.020
 wing+0.019
el+0.019
c+0.019
Neg logit contributions
 Hood-0.030
ussy-0.028
assy-0.027
ilda-0.027
 treadmill-0.026
TokenThat
Token ID2504
Feature activation-7.95e-02

Pos logit contributions
 Hood+0.082
 treadmill+0.074
ner+0.074
ussy+0.074
fighters+0.071
Neg logit contributions
 mouth-0.061
c-0.059
el-0.058
 wing-0.058
 its-0.055
Token is
Token ID318
Feature activation-1.99e-02

Pos logit contributions
 mouth+0.088
c+0.087
el+0.081
 wing+0.081
 its+0.074
Neg logit contributions
 Hood-0.158
ner-0.149
ussy-0.148
 treadmill-0.147
assy-0.143
Token our
Token ID674
Feature activation-1.47e-01

Pos logit contributions
c+0.027
el+0.026
light+0.025
 wing+0.025
 mouth+0.024
Neg logit contributions
 Hood-0.035
 treadmill-0.032
ussy-0.031
fighters-0.030
ilda-0.029
Token past
Token ID1613
Feature activation-7.30e-02

Pos logit contributions
 mouth+0.192
c+0.185
el+0.181
 wing+0.180
 its+0.170
Neg logit contributions
 Hood-0.262
ner-0.239
ussy-0.236
 treadmill-0.235
fighters-0.227
Tokenel
Token ID417
Feature activation-3.75e-01

Pos logit contributions
 mouth+0.080
c+0.077
el+0.076
 wing+0.074
 its+0.070
Neg logit contributions
 Hood-0.146
ner-0.134
 treadmill-0.133
ussy-0.132
fighters-0.128
Token!"
Token ID2474
Feature activation+1.59e-01

Pos logit contributions
 mouth+0.693
c+0.675
 wing+0.665
el+0.663
 its+0.638
Neg logit contributions
 Hood-0.463
ner-0.400
ussy-0.394
 treadmill-0.394
fighters-0.374
Token
Token ID198
Feature activation-2.34e-02

Pos logit contributions
 Hood+0.322
ussy+0.294
 treadmill+0.283
ner+0.281
fighters+0.279
Neg logit contributions
 mouth-0.192
 its-0.176
 wing-0.168
c-0.167
el-0.166
Token
Token ID198
Feature activation-2.36e-02

Pos logit contributions
 its+0.020
 mouth+0.019
 wing+0.017
c+0.016
el+0.016
Neg logit contributions
 Hood-0.056
ussy-0.053
 treadmill-0.051
ner-0.050
ilda-0.050
TokenTom
Token ID13787
Feature activation+9.48e-02

Pos logit contributions
 mouth+0.024
 wing+0.023
 its+0.023
c+0.022
el+0.021
Neg logit contributions
 Hood-0.052
ussy-0.048
ilda-0.047
 treadmill-0.045
ner-0.045
Token runs
Token ID4539
Feature activation+1.28e-01

Pos logit contributions
 Hood+0.185
ner+0.178
ilda+0.172
working+0.172
cakes+0.164
Neg logit contributions
 mouth-0.127
 eyes-0.109
el-0.108
c-0.103
 be-0.101
Token and
Token ID290
Feature activation+1.30e-02

Pos logit contributions
 Hood+0.016
fighters+0.014
boards+0.014
ussy+0.014
ives+0.013
Neg logit contributions
c-0.034
 wing-0.033
 mouth-0.032
 pouch-0.032
light-0.031
Token hurt
Token ID5938
Feature activation-2.55e-02

Pos logit contributions
 Hood+0.017
 treadmill+0.016
ner+0.015
fighters+0.015
ussy+0.015
Neg logit contributions
el-0.022
c-0.021
light-0.021
 wing-0.020
 mouth-0.019
Token.
Token ID13
Feature activation-7.95e-03

Pos logit contributions
c+0.040
 mouth+0.039
 wing+0.038
el+0.037
 oval+0.036
Neg logit contributions
 Hood-0.038
ussy-0.034
ner-0.033
fighters-0.033
 treadmill-0.033
Token He
Token ID679
Feature activation+5.00e-03

Pos logit contributions
c+0.012
 wing+0.012
el+0.012
 mouth+0.012
light+0.012
Neg logit contributions
 Hood-0.012
ner-0.011
ussy-0.010
 treadmill-0.010
fighters-0.010
Token y
Token ID331
Feature activation+9.96e-03

Pos logit contributions
 Hood+0.008
 treadmill+0.007
ner+0.007
ussy+0.007
fighters+0.006
Neg logit contributions
c-0.007
el-0.007
 wing-0.007
 mouth-0.007
light-0.007
Tokenel
Token ID417
Feature activation-3.74e-01

Pos logit contributions
 Hood+0.020
ussy+0.019
oons+0.018
boards+0.018
assy+0.018
Neg logit contributions
c-0.010
 mouth-0.009
 wing-0.009
 feather-0.008
 cardinal-0.008
Tokenped
Token ID9124
Feature activation+5.03e-03

Pos logit contributions
 mouth+0.610
c+0.593
el+0.582
 wing+0.581
 its+0.554
Neg logit contributions
 Hood-0.546
ner-0.485
 treadmill-0.479
ussy-0.477
fighters-0.458
Token and
Token ID290
Feature activation+1.27e-02

Pos logit contributions
 Hood+0.008
ner+0.008
 treadmill+0.007
ussy+0.007
fighters+0.007
Neg logit contributions
 mouth-0.007
c-0.007
el-0.007
 wing-0.007
light-0.006
Token ran
Token ID4966
Feature activation+7.82e-02

Pos logit contributions
 Hood+0.015
 treadmill+0.015
ner+0.013
fighters+0.013
drivers+0.012
Neg logit contributions
el-0.022
c-0.022
light-0.021
 wing-0.021
ric-0.020
Token away
Token ID1497
Feature activation-1.21e-01

Pos logit contributions
 Hood+0.175
ner+0.167
ussy+0.164
ilda+0.163
assy+0.161
Neg logit contributions
 mouth-0.070
 its-0.065
c-0.060
 wing-0.059
el-0.056
Token.
Token ID13
Feature activation-8.06e-03

Pos logit contributions
 mouth+0.177
c+0.177
el+0.172
 wing+0.170
light+0.163
Neg logit contributions
 Hood-0.194
ner-0.176
 treadmill-0.173
ussy-0.171
fighters-0.167
Token the
Token ID262
Feature activation+1.07e-02

Pos logit contributions
 Hood+0.022
ner+0.019
 treadmill+0.018
ussy+0.018
ilda+0.017
Neg logit contributions
 mouth-0.024
 wing-0.024
c-0.024
light-0.023
 oval-0.022
Token door
Token ID3420
Feature activation-2.45e-02

Pos logit contributions
 Hood+0.015
ner+0.014
 treadmill+0.013
driver+0.012
fighters+0.012
Neg logit contributions
light-0.015
imm-0.014
ectar-0.014
c-0.014
el-0.014
Token of
Token ID286
Feature activation-7.37e-02

Pos logit contributions
 mouth+0.040
 wing+0.040
c+0.039
rys+0.038
 fin+0.038
Neg logit contributions
 Hood-0.032
ner-0.031
ussy-0.029
 treadmill-0.029
fighters-0.028
Token the
Token ID262
Feature activation+1.08e-02

Pos logit contributions
light+0.151
c+0.149
imm+0.136
rys+0.135
ust+0.134
Neg logit contributions
 Hood-0.065
ner-0.056
 treadmill-0.047
 scientists-0.045
 tables-0.044
Token bird
Token ID6512
Feature activation-1.12e-01

Pos logit contributions
 Hood+0.016
ner+0.015
 treadmill+0.014
fighters+0.014
 stairs+0.013
Neg logit contributions
ectar-0.014
light-0.013
 its-0.013
el-0.013
imm-0.013
Tokenc
Token ID66
Feature activation-3.74e-01

Pos logit contributions
rys+0.166
 fin+0.149
Aw+0.147
 mouths+0.147
 throats+0.143
Neg logit contributions
 Hood-0.166
ner-0.142
ussy-0.141
ives-0.132
ilda-0.132
Tokenage
Token ID496
Feature activation-1.25e-01

Pos logit contributions
 mouth+0.474
rys+0.472
 fin+0.465
el+0.462
 its+0.462
Neg logit contributions
 Hood-0.582
ives-0.551
cakes-0.544
ussy-0.541
boards-0.533
Token.
Token ID13
Feature activation-1.14e-02

Pos logit contributions
rys+0.226
ric+0.219
c+0.217
el+0.215
 wing+0.213
Neg logit contributions
 Hood-0.138
ner-0.137
 treadmill-0.124
boards-0.122
cases-0.121
Token It
Token ID632
Feature activation-1.04e-01

Pos logit contributions
c+0.016
 mouth+0.016
el+0.016
 wing+0.016
light+0.015
Neg logit contributions
 Hood-0.018
ner-0.017
 treadmill-0.016
ussy-0.016
fighters-0.016
Token is
Token ID318
Feature activation-1.77e-02

Pos logit contributions
 wing+0.171
 mouth+0.166
el+0.165
rys+0.164
 fin+0.162
Neg logit contributions
 Hood-0.141
fighters-0.116
ussy-0.115
ner-0.109
 treadmill-0.108
Token locked
Token ID8970
Feature activation-1.32e-02

Pos logit contributions
c+0.031
el+0.030
rys+0.029
light+0.028
ric+0.027
Neg logit contributions
 Hood-0.019
 downstairs-0.016
ussy-0.016
 upstairs-0.016
 treadmill-0.015
Token this
Token ID428
Feature activation-4.20e-02

Pos logit contributions
c+0.071
 tal+0.064
 mouth+0.064
 wing+0.064
light+0.064
Neg logit contributions
 Hood-0.063
ner-0.063
fighters-0.056
ussy-0.056
 treadmill-0.055
Token is
Token ID318
Feature activation-1.69e-02

Pos logit contributions
 wing+0.061
rys+0.061
el+0.060
 pouch+0.060
 fin+0.059
Neg logit contributions
 Hood-0.053
oons-0.049
ner-0.046
fighters-0.045
boards-0.045
Token not
Token ID407
Feature activation-6.30e-02

Pos logit contributions
c+0.027
light+0.027
el+0.027
 wing+0.025
 pouch+0.025
Neg logit contributions
 Hood-0.022
ussy-0.021
 treadmill-0.019
fighters-0.019
ner-0.018
Token a
Token ID257
Feature activation+4.29e-02

Pos logit contributions
c+0.106
el+0.105
 pouch+0.100
 mouth+0.100
light+0.099
Neg logit contributions
 Hood-0.079
ussy-0.066
 treadmill-0.065
fighters-0.064
ner-0.064
Token past
Token ID1613
Feature activation-7.35e-02

Pos logit contributions
 Hood+0.064
 treadmill+0.059
ner+0.057
ussy+0.054
working+0.054
Neg logit contributions
el-0.060
c-0.057
light-0.055
ric-0.052
 mouth-0.051
Tokenel
Token ID417
Feature activation-3.74e-01

Pos logit contributions
 mouth+0.029
 pouch+0.022
 wing+0.022
c+0.020
 nest+0.018
Neg logit contributions
 Hood-0.192
ner-0.179
ussy-0.176
oons-0.172
fighters-0.171
Token.
Token ID13
Feature activation-9.43e-03

Pos logit contributions
 mouth+0.546
 pouch+0.536
 wing+0.533
c+0.523
rys+0.522
Neg logit contributions
 Hood-0.524
ner-0.464
ussy-0.457
fighters-0.455
oons-0.450
Token This
Token ID770
Feature activation-8.07e-02

Pos logit contributions
 mouth+0.012
c+0.012
 wing+0.012
el+0.011
 feather+0.011
Neg logit contributions
 Hood-0.016
ner-0.015
ussy-0.015
 treadmill-0.014
fighters-0.014
Token is
Token ID318
Feature activation-1.68e-02

Pos logit contributions
 wing+0.120
 pouch+0.117
rys+0.116
 fin+0.116
el+0.116
Neg logit contributions
 Hood-0.109
ner-0.094
boards-0.092
oons-0.092
fighters-0.091
Token a
Token ID257
Feature activation+4.27e-02

Pos logit contributions
c+0.028
el+0.027
light+0.027
 wing+0.027
 pouch+0.027
Neg logit contributions
 Hood-0.021
ussy-0.019
fighters-0.018
 treadmill-0.017
 scientists-0.017
Token cr
Token ID1067
Feature activation-1.92e-01

Pos logit contributions
 Hood+0.056
 treadmill+0.053
ner+0.048
 scientist+0.048
ilda+0.047
Neg logit contributions
el-0.067
c-0.063
light-0.060
 mouths-0.056
ric-0.056
Token the
Token ID262
Feature activation+1.42e-02

Pos logit contributions
 Hood+0.066
working+0.065
 treadmill+0.064
fighters+0.064
ival+0.063
Neg logit contributions
 its-0.028
el-0.025
 mouth-0.024
c-0.022
 eyes-0.021
Token bird
Token ID6512
Feature activation-1.10e-01

Pos logit contributions
 Hood+0.025
ner+0.022
 treadmill+0.022
ussy+0.022
fighters+0.021
Neg logit contributions
 mouth-0.018
c-0.018
el-0.018
light-0.017
 its-0.017
Token and
Token ID290
Feature activation+1.70e-02

Pos logit contributions
 mouth+0.180
 fin+0.171
rys+0.169
 oval+0.169
 pouch+0.169
Neg logit contributions
 Hood-0.146
ner-0.121
ussy-0.118
fighters-0.116
 treadmill-0.113
Token the
Token ID262
Feature activation+1.37e-02

Pos logit contributions
 Hood+0.030
ner+0.028
ussy+0.028
 treadmill+0.028
fighters+0.026
Neg logit contributions
 mouth-0.023
c-0.021
 wing-0.021
el-0.021
 its-0.021
Token bird
Token ID6512
Feature activation-1.10e-01

Pos logit contributions
 Hood+0.025
ner+0.023
 treadmill+0.023
ussy+0.022
fighters+0.022
Neg logit contributions
 mouth-0.016
c-0.016
el-0.016
light-0.015
 its-0.015
Tokenc
Token ID66
Feature activation-3.74e-01

Pos logit contributions
rys+0.180
 oval+0.174
 fin+0.173
 mouth+0.171
 pouch+0.167
Neg logit contributions
 Hood-0.134
ner-0.117
fighters-0.108
ussy-0.106
ives-0.105
Tokenage
Token ID496
Feature activation-1.25e-01

Pos logit contributions
 mouth+0.403
c+0.383
el+0.373
 wing+0.363
 its+0.350
Neg logit contributions
 Hood-0.735
ner-0.670
ussy-0.670
 treadmill-0.667
fighters-0.657
Token and
Token ID290
Feature activation+1.72e-02

Pos logit contributions
 mouth+0.191
c+0.189
 fin+0.188
rys+0.186
ric+0.186
Neg logit contributions
 Hood-0.164
ner-0.159
 treadmill-0.151
cakes-0.147
boards-0.145
Token the
Token ID262
Feature activation+1.33e-02

Pos logit contributions
 Hood+0.028
ner+0.025
 treadmill+0.024
ussy+0.024
fighters+0.024
Neg logit contributions
 mouth-0.025
c-0.025
el-0.024
 wing-0.024
light-0.023
Token chaos
Token ID11918
Feature activation+1.15e-01

Pos logit contributions
 Hood+0.024
ner+0.022
 treadmill+0.022
ussy+0.021
fighters+0.021
Neg logit contributions
 mouth-0.016
c-0.016
el-0.016
light-0.015
 its-0.014
Token.
Token ID13
Feature activation-1.54e-03

Pos logit contributions
 Hood+0.196
ner+0.177
 treadmill+0.175
ussy+0.175
fighters+0.169
Neg logit contributions
 mouth-0.157
c-0.151
el-0.149
 wing-0.148
 its-0.141
Token He
Token ID679
Feature activation+6.21e-03

Pos logit contributions
 mouth+0.015
c+0.014
el+0.014
 wing+0.014
 its+0.013
Neg logit contributions
 Hood-0.018
ner-0.016
ussy-0.016
 treadmill-0.016
fighters-0.016
Token told
Token ID1297
Feature activation-3.67e-02

Pos logit contributions
 Hood+0.011
ner+0.010
 treadmill+0.010
ussy+0.010
assy+0.010
Neg logit contributions
 mouth-0.008
c-0.008
el-0.008
 wing-0.007
 its-0.007
Token us
Token ID514
Feature activation-8.94e-02

Pos logit contributions
 mouth+0.048
c+0.047
 wing+0.046
el+0.044
light+0.044
Neg logit contributions
 Hood-0.064
ner-0.057
 treadmill-0.057
ussy-0.057
fighters-0.056
Token to
Token ID284
Feature activation-1.86e-02

Pos logit contributions
 mouth+0.107
 wing+0.098
c+0.098
 its+0.098
el+0.095
Neg logit contributions
 Hood-0.168
ner-0.160
ussy-0.156
 treadmill-0.155
assy-0.153
Token kne
Token ID24813
Feature activation-1.89e-02

Pos logit contributions
c+0.024
 mouth+0.024
el+0.024
light+0.023
 oval+0.022
Neg logit contributions
 Hood-0.030
 treadmill-0.028
fighters-0.027
ner-0.027
ussy-0.026
Tokenel
Token ID417
Feature activation-3.74e-01

Pos logit contributions
 mouth+0.010
 oval+0.008
light+0.008
 wing+0.008
 tongue+0.007
Neg logit contributions
 Hood-0.049
ussy-0.044
assy-0.043
ner-0.043
fighters-0.041
Token and
Token ID290
Feature activation+1.32e-02

Pos logit contributions
 mouth+0.519
c+0.501
el+0.491
 wing+0.490
 its+0.464
Neg logit contributions
 Hood-0.632
ner-0.570
 treadmill-0.564
ussy-0.563
fighters-0.543
Token pray
Token ID12472
Feature activation-1.26e-02

Pos logit contributions
 Hood+0.019
ner+0.017
 treadmill+0.017
fighters+0.016
ussy+0.016
Neg logit contributions
el-0.020
c-0.020
 mouth-0.019
 oval-0.019
 wing-0.018
Token.
Token ID13
Feature activation-1.05e-02

Pos logit contributions
 mouth+0.017
c+0.017
 wing+0.016
el+0.016
 its+0.016
Neg logit contributions
 Hood-0.022
ner-0.020
 treadmill-0.019
ussy-0.019
fighters-0.018
Token We
Token ID775
Feature activation-5.90e-02

Pos logit contributions
 mouth+0.015
c+0.015
el+0.015
 wing+0.015
 feathers+0.014
Neg logit contributions
 Hood-0.017
ner-0.015
ussy-0.015
 treadmill-0.015
fighters-0.014
Token did
Token ID750
Feature activation-6.18e-02

Pos logit contributions
 mouth+0.086
c+0.083
 wing+0.081
el+0.081
 oval+0.077
Neg logit contributions
 Hood-0.096
ner-0.085
ussy-0.084
 treadmill-0.083
fighters-0.082
Token!
Token ID0
Feature activation+3.15e-02

Pos logit contributions
 mouth+0.417
c+0.405
 wing+0.398
el+0.394
 its+0.381
Neg logit contributions
 Hood-0.397
 treadmill-0.357
ussy-0.356
ner-0.351
assy-0.336
Token You
Token ID921
Feature activation-7.98e-02

Pos logit contributions
 Hood+0.054
ner+0.050
ussy+0.048
 treadmill+0.048
fighters+0.047
Neg logit contributions
 mouth-0.042
c-0.041
 wing-0.041
el-0.040
 feathers-0.039
Token have
Token ID423
Feature activation-5.62e-03

Pos logit contributions
 mouth+0.098
c+0.095
el+0.094
 wing+0.093
 its+0.088
Neg logit contributions
 Hood-0.142
 treadmill-0.138
ner-0.135
ussy-0.134
assy-0.131
Token to
Token ID284
Feature activation-1.67e-02

Pos logit contributions
el+0.009
c+0.009
 mouth+0.008
light+0.008
 oval+0.008
Neg logit contributions
 Hood-0.008
ner-0.007
ilda-0.007
ussy-0.007
 treadmill-0.007
Token kne
Token ID24813
Feature activation-1.95e-02

Pos logit contributions
c+0.025
 mouth+0.025
light+0.024
 mouths+0.023
el+0.023
Neg logit contributions
 Hood-0.024
 treadmill-0.022
assy-0.021
ilda-0.021
fighters-0.020
Tokenel
Token ID417
Feature activation-3.74e-01

Pos logit contributions
 mouth+0.011
light+0.009
 oval+0.009
 wing+0.008
 tongue+0.008
Neg logit contributions
 Hood-0.048
ussy-0.044
assy-0.043
ner-0.042
ilda-0.042
Token and
Token ID290
Feature activation+1.50e-02

Pos logit contributions
 mouth+0.532
c+0.516
el+0.506
 wing+0.504
 its+0.479
Neg logit contributions
 Hood-0.619
ner-0.563
 treadmill-0.558
ussy-0.554
fighters-0.533
Token pray
Token ID12472
Feature activation-1.20e-02

Pos logit contributions
 Hood+0.022
ner+0.020
 treadmill+0.020
fighters+0.019
ussy+0.019
Neg logit contributions
c-0.022
el-0.022
 mouth-0.021
 wing-0.021
light-0.020
Token before
Token ID878
Feature activation+4.23e-02

Pos logit contributions
 mouth+0.017
c+0.017
el+0.016
 wing+0.016
 feathers+0.015
Neg logit contributions
 Hood-0.020
ner-0.018
 treadmill-0.018
ussy-0.018
fighters-0.017
Token it
Token ID340
Feature activation-5.43e-02

Pos logit contributions
 Hood+0.070
ner+0.065
 treadmill+0.061
fighters+0.060
ilda+0.059
Neg logit contributions
c-0.059
 mouth-0.057
el-0.054
 wing-0.054
 oval-0.052
Token!"
Token ID2474
Feature activation+1.65e-01

Pos logit contributions
 mouth+0.076
c+0.074
el+0.072
 wing+0.071
 its+0.068
Neg logit contributions
 Hood-0.092
ner-0.084
 treadmill-0.084
ussy-0.082
assy-0.080
Token toy
Token ID13373
Feature activation-4.58e-02

Pos logit contributions
 Hood+0.016
ner+0.014
 treadmill+0.014
fighters+0.014
ussy+0.013
Neg logit contributions
c-0.012
 mouth-0.012
el-0.012
light-0.012
 its-0.011
Token car
Token ID1097
Feature activation+1.13e-02

Pos logit contributions
 mouth+0.065
c+0.063
el+0.062
 wing+0.061
light+0.058
Neg logit contributions
 Hood-0.076
ner-0.068
 treadmill-0.067
ussy-0.066
fighters-0.065
Token with
Token ID351
Feature activation-1.25e-01

Pos logit contributions
 Hood+0.020
ner+0.018
ussy+0.018
 treadmill+0.018
assy+0.018
Neg logit contributions
 mouth-0.015
c-0.015
 wing-0.014
el-0.014
 its-0.014
Token the
Token ID262
Feature activation+9.41e-03

Pos logit contributions
 mouth+0.162
c+0.161
 wing+0.155
el+0.153
light+0.146
Neg logit contributions
 Hood-0.219
ner-0.198
 treadmill-0.196
ussy-0.195
fighters-0.190
Token bird
Token ID6512
Feature activation-1.12e-01

Pos logit contributions
 Hood+0.014
ner+0.012
driver+0.011
fighters+0.011
drivers+0.011
Neg logit contributions
light-0.014
c-0.014
el-0.014
rys-0.013
 mouth-0.013
Tokenc
Token ID66
Feature activation-3.74e-01

Pos logit contributions
rys+0.197
 mouth+0.188
 oval+0.181
 spectrum+0.180
Aw+0.179
Neg logit contributions
 Hood-0.138
fighters-0.100
ussy-0.100
driver-0.098
cases-0.097
Tokenage
Token ID496
Feature activation-1.24e-01

Pos logit contributions
 mouth+0.429
el+0.398
c+0.394
 oval+0.381
 wing+0.379
Neg logit contributions
 Hood-0.696
ussy-0.627
fighters-0.626
ner-0.621
 treadmill-0.620
Token to
Token ID284
Feature activation-1.84e-02

Pos logit contributions
 mouth+0.186
c+0.181
el+0.178
 wing+0.176
light+0.166
Neg logit contributions
 Hood-0.192
ner-0.173
 treadmill-0.170
ussy-0.168
fighters-0.165
Token the
Token ID262
Feature activation+9.86e-03

Pos logit contributions
 mouth+0.026
c+0.025
el+0.025
 wing+0.025
 its+0.023
Neg logit contributions
 Hood-0.031
ner-0.028
 treadmill-0.027
ussy-0.027
fighters-0.026
Token kitchen
Token ID9592
Feature activation+8.21e-02

Pos logit contributions
 Hood+0.017
 treadmill+0.015
ner+0.015
fighters+0.015
ussy+0.015
Neg logit contributions
c-0.013
el-0.013
 mouth-0.012
light-0.012
 its-0.011
Token.
Token ID13
Feature activation-1.03e-02

Pos logit contributions
 Hood+0.133
ner+0.120
 treadmill+0.119
ussy+0.118
fighters+0.114
Neg logit contributions
 mouth-0.118
c-0.115
el-0.113
 wing-0.113
 its-0.107
Token goodbye
Token ID24829
Feature activation-2.25e-02

Pos logit contributions
c+0.175
ric+0.166
el+0.161
light+0.158
 oval+0.155
Neg logit contributions
 Hood-0.132
ner-0.110
 downstairs-0.105
ives-0.105
cases-0.104
Token.
Token ID13
Feature activation-5.33e-03

Pos logit contributions
 mouth+0.033
c+0.032
el+0.032
 wing+0.031
 oval+0.029
Neg logit contributions
 Hood-0.035
ner-0.032
 treadmill-0.031
ussy-0.030
fighters-0.030
Token They
Token ID1119
Feature activation-3.25e-02

Pos logit contributions
c+0.008
el+0.008
 wing+0.008
 mouth+0.008
light+0.008
Neg logit contributions
 Hood-0.008
ner-0.007
 treadmill-0.007
fighters-0.007
ussy-0.007
Token said
Token ID531
Feature activation-1.07e-01

Pos logit contributions
 mouth+0.048
el+0.048
c+0.048
 wing+0.047
 oval+0.044
Neg logit contributions
 Hood-0.051
ner-0.044
 treadmill-0.044
ussy-0.043
assy-0.042
Token "
Token ID366
Feature activation+4.73e-02

Pos logit contributions
c+0.150
 mouth+0.148
el+0.146
 wing+0.140
light+0.136
Neg logit contributions
 Hood-0.177
ner-0.161
 treadmill-0.155
ussy-0.153
fighters-0.151
Tokenc
Token ID66
Feature activation-3.73e-01

Pos logit contributions
 Hood+0.061
ner+0.057
fighters+0.054
assy+0.051
ussy+0.050
Neg logit contributions
 mouth-0.087
 wing-0.079
el-0.077
c-0.077
 tongue-0.076
Tokenaw
Token ID707
Feature activation-1.40e-01

Pos logit contributions
 mouth+0.481
c+0.465
 wing+0.454
el+0.454
 its+0.427
Neg logit contributions
 Hood-0.672
ner-0.615
 treadmill-0.610
ussy-0.606
fighters-0.582
Token,
Token ID11
Feature activation+4.06e-03

Pos logit contributions
 mouth+0.188
c+0.178
el+0.175
 wing+0.170
 its+0.165
Neg logit contributions
 Hood-0.253
 treadmill-0.231
ner-0.230
ussy-0.227
assy-0.220
Token c
Token ID269
Feature activation-9.95e-02

Pos logit contributions
 Hood+0.005
ner+0.005
 treadmill+0.004
ival+0.004
 builders+0.004
Neg logit contributions
c-0.007
el-0.006
light-0.006
 direction-0.006
imm-0.006
Tokenaw
Token ID707
Feature activation-1.40e-01

Pos logit contributions
 mouth+0.154
c+0.150
el+0.148
 wing+0.147
 its+0.139
Neg logit contributions
 Hood-0.155
ner-0.139
 treadmill-0.137
ussy-0.136
fighters-0.131
Token,
Token ID11
Feature activation+4.00e-03

Pos logit contributions
 mouth+0.195
c+0.186
el+0.183
 wing+0.180
 its+0.173
Neg logit contributions
 Hood-0.245
 treadmill-0.220
ner-0.220
ussy-0.218
assy-0.210
Token He
Token ID679
Feature activation+8.51e-03

Pos logit contributions
 mouth+0.006
 wing+0.006
 its+0.006
c+0.006
el+0.006
Neg logit contributions
 Hood-0.009
ner-0.008
ussy-0.008
 treadmill-0.008
ilda-0.008
Token liked
Token ID8288
Feature activation+2.85e-02

Pos logit contributions
 Hood+0.013
ner+0.012
ussy+0.012
 treadmill+0.012
fighters+0.011
Neg logit contributions
 mouth-0.013
el-0.012
 wing-0.012
c-0.012
light-0.011
Token his
Token ID465
Feature activation-1.00e-01

Pos logit contributions
 Hood+0.051
ner+0.046
ussy+0.046
 treadmill+0.046
fighters+0.044
Neg logit contributions
 mouth-0.037
c-0.036
 wing-0.035
el-0.035
 its-0.034
Token new
Token ID649
Feature activation+3.93e-02

Pos logit contributions
el+0.139
c+0.139
light+0.132
 mouth+0.132
 its+0.126
Neg logit contributions
 Hood-0.159
 treadmill-0.141
ner-0.135
cakes-0.130
working-0.129
Token bird
Token ID6512
Feature activation-1.12e-01

Pos logit contributions
 Hood+0.057
 treadmill+0.050
working+0.046
 work+0.045
cakes+0.045
Neg logit contributions
el-0.055
rys-0.053
light-0.051
ifying-0.050
 fin-0.050
Tokenc
Token ID66
Feature activation-3.73e-01

Pos logit contributions
rys+0.124
el+0.109
 oval+0.107
 mouth+0.106
Aw+0.106
Neg logit contributions
 Hood-0.216
assy-0.178
ussy-0.176
ilda-0.169
ner-0.165
Tokenage
Token ID496
Feature activation-1.24e-01

Pos logit contributions
 mouth+0.420
el+0.389
c+0.388
 wing+0.380
 its+0.369
Neg logit contributions
 Hood-0.717
ussy-0.649
 treadmill-0.647
ner-0.645
fighters-0.636
Token,
Token ID11
Feature activation+7.15e-03

Pos logit contributions
 mouth+0.170
c+0.164
 wing+0.161
el+0.160
 its+0.152
Neg logit contributions
 Hood-0.213
ner-0.193
 treadmill-0.191
ussy-0.190
fighters-0.184
Token but
Token ID475
Feature activation-1.42e-01

Pos logit contributions
 Hood+0.009
 treadmill+0.008
ner+0.008
 cop+0.007
 scientists+0.007
Neg logit contributions
el-0.012
rys-0.011
c-0.011
light-0.010
 pouch-0.010
Token he
Token ID339
Feature activation+2.95e-02

Pos logit contributions
c+0.280
el+0.273
rys+0.264
 mouth+0.264
light+0.261
Neg logit contributions
 Hood-0.167
ner-0.142
fighters-0.118
 treadmill-0.115
working-0.113
Token also
Token ID635
Feature activation-3.90e-02

Pos logit contributions
 Hood+0.048
ussy+0.042
ner+0.041
 treadmill+0.040
fighters+0.040
Neg logit contributions
 mouth-0.043
c-0.041
el-0.041
 wing-0.041
light-0.039
Token spicy
Token ID26880
Feature activation-1.22e-02

Pos logit contributions
c+0.092
light+0.087
 mouth+0.087
el+0.086
 wing+0.086
Neg logit contributions
 Hood-0.099
ner-0.089
 treadmill-0.087
ussy-0.086
oons-0.085
Token.
Token ID13
Feature activation-1.55e-02

Pos logit contributions
 mouth+0.018
c+0.017
el+0.017
 wing+0.016
 its+0.016
Neg logit contributions
 Hood-0.021
ner-0.019
 treadmill-0.018
ussy-0.018
fighters-0.018
Token It
Token ID632
Feature activation-1.10e-01

Pos logit contributions
c+0.022
 wing+0.020
el+0.020
 direction+0.020
 feather+0.020
Neg logit contributions
 Hood-0.022
ner-0.022
ussy-0.022
ives-0.021
fighters-0.021
Token hurts
Token ID20406
Feature activation-7.44e-02

Pos logit contributions
 wing+0.167
 mouth+0.164
el+0.163
light+0.159
c+0.158
Neg logit contributions
 Hood-0.159
ussy-0.138
fighters-0.135
oons-0.130
 treadmill-0.130
Token my
Token ID616
Feature activation-1.39e-01

Pos logit contributions
 mouth+0.108
c+0.107
el+0.102
 wing+0.102
 pouch+0.100
Neg logit contributions
 Hood-0.112
ussy-0.100
ner-0.100
 treadmill-0.098
fighters-0.097
Token mouth
Token ID5422
Feature activation-3.73e-01

Pos logit contributions
c+0.127
el+0.117
rys+0.112
light+0.111
 its+0.110
Neg logit contributions
 Hood-0.270
ussy-0.247
 treadmill-0.245
ner-0.237
oons-0.237
Token.
Token ID13
Feature activation-1.63e-02

Pos logit contributions
 mouth+0.522
c+0.503
 wing+0.494
el+0.492
 its+0.466
Neg logit contributions
 Hood-0.636
ner-0.575
 treadmill-0.568
ussy-0.562
fighters-0.546
Token It
Token ID632
Feature activation-1.10e-01

Pos logit contributions
c+0.023
el+0.022
 wing+0.021
 direction+0.021
 feather+0.021
Neg logit contributions
 Hood-0.024
ner-0.023
ussy-0.023
ives-0.022
fighters-0.022
Token is
Token ID318
Feature activation-2.20e-02

Pos logit contributions
 wing+0.168
el+0.166
 mouth+0.165
light+0.163
c+0.161
Neg logit contributions
 Hood-0.157
ussy-0.137
fighters-0.131
oons-0.125
 treadmill-0.125
Token not
Token ID407
Feature activation-6.94e-02

Pos logit contributions
c+0.038
light+0.038
el+0.038
 pouch+0.036
 hue+0.035
Neg logit contributions
 Hood-0.023
ussy-0.022
 treadmill-0.019
 scientists-0.019
 Doctor-0.019
Token tasty
Token ID25103
Feature activation+3.63e-02

Pos logit contributions
el+0.114
c+0.112
light+0.108
 mouth+0.106
 pouch+0.105
Neg logit contributions
 Hood-0.096
ussy-0.081
 treadmill-0.078
ner-0.076
fighters-0.076
Token was
Token ID373
Feature activation-3.64e-04

Pos logit contributions
 Hood+0.103
 treadmill+0.095
ner+0.094
ussy+0.094
fighters+0.090
Neg logit contributions
 mouth-0.075
c-0.074
el-0.071
 wing-0.070
 its-0.067
Token so
Token ID523
Feature activation-9.64e-02

Pos logit contributions
el+0.001
c+0.001
 mouth+0.001
light+0.000
rys+0.000
Neg logit contributions
 Hood-0.001
ner-0.001
ussy-0.000
 treadmill-0.000
assy-0.000
Token proud
Token ID6613
Feature activation+8.54e-02

Pos logit contributions
 mouth+0.121
c+0.114
 wing+0.114
el+0.112
 its+0.111
Neg logit contributions
 Hood-0.180
 treadmill-0.166
ner-0.163
ussy-0.163
fighters-0.160
Token of
Token ID286
Feature activation-6.33e-02

Pos logit contributions
 Hood+0.108
ner+0.103
boards+0.095
 scientists+0.094
 treadmill+0.092
Neg logit contributions
c-0.146
 wing-0.138
 mouth-0.136
el-0.136
 pouch-0.134
Token his
Token ID465
Feature activation-9.17e-02

Pos logit contributions
 mouth+0.091
c+0.089
 wing+0.086
el+0.086
light+0.083
Neg logit contributions
 Hood-0.106
ner-0.095
 treadmill-0.094
ussy-0.093
fighters-0.090
Token work
Token ID670
Feature activation+2.74e-01

Pos logit contributions
 mouth+0.145
 wing+0.141
el+0.137
 feathers+0.131
 feather+0.131
Neg logit contributions
 Hood-0.149
ussy-0.139
fighters-0.136
 treadmill-0.131
assy-0.130
Token that
Token ID326
Feature activation-2.47e-02

Pos logit contributions
 Hood+0.416
ner+0.381
 treadmill+0.368
ussy+0.357
ilda+0.356
Neg logit contributions
 mouth-0.414
c-0.410
el-0.399
 wing-0.393
 feathers-0.374
Token he
Token ID339
Feature activation+3.65e-02

Pos logit contributions
c+0.046
el+0.045
 mouth+0.044
rys+0.043
ric+0.043
Neg logit contributions
 Hood-0.030
ner-0.026
 treadmill-0.023
working-0.023
ilda-0.022
Token decided
Token ID3066
Feature activation+2.86e-02

Pos logit contributions
 Hood+0.073
 treadmill+0.071
ussy+0.070
ner+0.068
fighters+0.067
Neg logit contributions
 mouth-0.037
c-0.036
 its-0.035
 feather-0.035
 wing-0.035
Token to
Token ID284
Feature activation-1.01e-02

Pos logit contributions
 Hood+0.054
ussy+0.049
ner+0.049
 treadmill+0.048
fighters+0.047
Neg logit contributions
 mouth-0.035
c-0.033
 wing-0.033
el-0.032
 its-0.032
Token enter
Token ID3802
Feature activation-4.92e-02

Pos logit contributions
 mouth+0.014
c+0.014
el+0.014
 wing+0.014
light+0.013
Neg logit contributions
 Hood-0.017
ner-0.015
 treadmill-0.015
ussy-0.015
fighters-0.014
Token was
Token ID373
Feature activation-2.59e-03

Pos logit contributions
 mouth+0.162
 wing+0.158
rys+0.152
el+0.150
 mouths+0.150
Neg logit contributions
 Hood-0.136
ner-0.120
 treadmill-0.117
fighters-0.116
ilda-0.116
Token big
Token ID1263
Feature activation+1.05e-01

Pos logit contributions
c+0.005
rys+0.004
 mouth+0.004
el+0.004
 mouths+0.004
Neg logit contributions
 Hood-0.003
 treadmill-0.003
ilda-0.003
ner-0.003
 scientists-0.002
Token and
Token ID290
Feature activation+2.32e-02

Pos logit contributions
 Hood+0.198
ussy+0.182
ner+0.179
 treadmill+0.177
fighters+0.175
Neg logit contributions
 mouth-0.134
c-0.124
el-0.123
 wing-0.121
 feathers-0.117
Token shiny
Token ID22441
Feature activation+1.20e-02

Pos logit contributions
 Hood+0.036
 treadmill+0.033
ner+0.033
fighters+0.031
ilda+0.031
Neg logit contributions
el-0.033
 mouth-0.033
 wing-0.033
c-0.033
rys-0.030
Token and
Token ID290
Feature activation+2.31e-02

Pos logit contributions
 Hood+0.020
ner+0.018
ussy+0.018
 treadmill+0.018
fighters+0.018
Neg logit contributions
 mouth-0.017
c-0.016
el-0.016
 wing-0.016
 its-0.015
Token Emma
Token ID18966
Feature activation+2.31e-01

Pos logit contributions
 Hood+0.038
ner+0.035
 treadmill+0.035
ussy+0.034
fighters+0.033
Neg logit contributions
 mouth-0.032
el-0.031
c-0.031
 wing-0.031
 feathers-0.028
Token thought
Token ID1807
Feature activation+1.29e-02

Pos logit contributions
 Hood+0.367
ner+0.328
 treadmill+0.323
ussy+0.322
ilda+0.312
Neg logit contributions
 mouth-0.334
c-0.326
 wing-0.321
el-0.321
 its-0.304
Token it
Token ID340
Feature activation-5.10e-02

Pos logit contributions
 Hood+0.023
ner+0.021
 treadmill+0.021
ussy+0.020
ilda+0.020
Neg logit contributions
 mouth-0.016
c-0.016
el-0.016
 wing-0.016
 feathers-0.015
Token was
Token ID373
Feature activation-3.17e-03

Pos logit contributions
 mouth+0.074
 wing+0.073
el+0.070
c+0.070
rys+0.067
Neg logit contributions
 Hood-0.083
 treadmill-0.073
ner-0.072
ussy-0.071
fighters-0.071
Token very
Token ID845
Feature activation-2.63e-02

Pos logit contributions
c+0.005
el+0.005
 mouth+0.005
 wing+0.005
light+0.005
Neg logit contributions
 Hood-0.004
 treadmill-0.004
ilda-0.004
ussy-0.004
 scientists-0.003
Token cool
Token ID3608
Feature activation+3.88e-02

Pos logit contributions
 mouth+0.034
c+0.032
el+0.032
 wing+0.032
 its+0.031
Neg logit contributions
 Hood-0.048
ner-0.044
ussy-0.043
 treadmill-0.043
fighters-0.041
Token It
Token ID632
Feature activation-9.54e-02

Pos logit contributions
 Hood+0.005
ner+0.005
ussy+0.005
 treadmill+0.005
fighters+0.004
Neg logit contributions
 mouth-0.004
 wing-0.004
c-0.004
el-0.004
 feathers-0.004
Token said
Token ID531
Feature activation-9.82e-02

Pos logit contributions
 wing+0.153
el+0.150
rys+0.149
 mouth+0.149
ric+0.144
Neg logit contributions
 Hood-0.135
fighters-0.112
ner-0.111
ussy-0.106
cases-0.105
Token,
Token ID11
Feature activation+1.14e-02

Pos logit contributions
 mouth+0.144
c+0.143
el+0.140
 wing+0.137
light+0.132
Neg logit contributions
 Hood-0.154
ner-0.143
 treadmill-0.135
ussy-0.135
fighters-0.132
Token "
Token ID366
Feature activation+5.46e-02

Pos logit contributions
 Hood+0.015
ner+0.014
 treadmill+0.013
ussy+0.013
fighters+0.013
Neg logit contributions
c-0.019
 mouth-0.019
el-0.019
 wing-0.018
light-0.018
TokenHappy
Token ID25082
Feature activation+1.36e-01

Pos logit contributions
ner+0.065
 Hood+0.063
fighters+0.058
ives+0.058
assy+0.057
Neg logit contributions
 mouth-0.110
 tongue-0.098
 mouths-0.098
 wing-0.095
 eyes-0.095
Token early
Token ID1903
Feature activation+2.48e-01

Pos logit contributions
 Hood+0.250
ner+0.229
assy+0.223
ussy+0.219
ilda+0.218
Neg logit contributions
c-0.155
 mouth-0.152
 wing-0.140
el-0.139
 its-0.138
Token birthday
Token ID10955
Feature activation-1.54e-01

Pos logit contributions
 Hood+0.423
ussy+0.379
ner+0.379
 treadmill+0.378
fighters+0.363
Neg logit contributions
 mouth-0.346
c-0.330
 wing-0.324
el-0.322
 its-0.307
Token,
Token ID11
Feature activation+1.15e-02

Pos logit contributions
 mouth+0.199
el+0.172
c+0.170
 wing+0.169
 its+0.165
Neg logit contributions
 Hood-0.291
 treadmill-0.270
ussy-0.269
ner-0.265
fighters-0.251
Token Sam
Token ID3409
Feature activation-6.25e-02

Pos logit contributions
 Hood+0.018
ner+0.017
 treadmill+0.016
ussy+0.016
fighters+0.016
Neg logit contributions
 mouth-0.017
c-0.017
el-0.016
 wing-0.016
light-0.015
Token.
Token ID13
Feature activation-1.29e-03

Pos logit contributions
c+0.081
 mouth+0.079
 its+0.074
el+0.073
 wing+0.073
Neg logit contributions
 Hood-0.114
 treadmill-0.114
ner-0.112
ussy-0.109
cases-0.107
Token Love
Token ID5896
Feature activation-1.40e-01

Pos logit contributions
 mouth+0.002
 wing+0.002
c+0.002
el+0.002
 feathers+0.002
Neg logit contributions
 Hood-0.002
ner-0.002
ussy-0.002
fighters-0.002
assy-0.002
Token a
Token ID257
Feature activation+5.37e-02

Pos logit contributions
 Hood+0.042
ner+0.038
 treadmill+0.037
ussy+0.037
fighters+0.036
Neg logit contributions
 mouth-0.036
c-0.035
el-0.034
 wing-0.034
 its-0.032
Token few
Token ID1178
Feature activation-6.09e-03

Pos logit contributions
 Hood+0.088
ner+0.081
 treadmill+0.080
ussy+0.077
fighters+0.075
Neg logit contributions
 mouth-0.077
c-0.075
el-0.075
 wing-0.073
 its-0.069
Token minutes
Token ID2431
Feature activation+1.16e-02

Pos logit contributions
 mouth+0.010
c+0.010
el+0.010
 wing+0.010
 its+0.009
Neg logit contributions
 Hood-0.009
ussy-0.008
ner-0.008
 treadmill-0.008
fighters-0.007
Token of
Token ID286
Feature activation-5.97e-02

Pos logit contributions
 Hood+0.019
ner+0.017
 treadmill+0.017
ussy+0.017
fighters+0.016
Neg logit contributions
 mouth-0.017
c-0.016
 wing-0.016
el-0.016
 its-0.015
Token hard
Token ID1327
Feature activation+1.65e-01

Pos logit contributions
c+0.095
 mouth+0.095
light+0.091
 wing+0.089
el+0.088
Neg logit contributions
 Hood-0.085
ner-0.082
 treadmill-0.079
ussy-0.072
drivers-0.071
Token work
Token ID670
Feature activation+2.80e-01

Pos logit contributions
 Hood+0.158
ussy+0.134
fighters+0.120
assy+0.119
 treadmill+0.119
Neg logit contributions
 mouth-0.356
el-0.351
 wing-0.345
c-0.341
 its-0.336
Token,
Token ID11
Feature activation+2.29e-02

Pos logit contributions
 Hood+0.532
ussy+0.484
 treadmill+0.471
assy+0.467
fighters+0.459
Neg logit contributions
 mouth-0.343
el-0.324
 wing-0.318
 its-0.317
c-0.316
Token the
Token ID262
Feature activation+2.45e-02

Pos logit contributions
 Hood+0.039
ner+0.035
ussy+0.035
 treadmill+0.034
fighters+0.033
Neg logit contributions
 mouth-0.032
c-0.031
 wing-0.030
el-0.030
 its-0.029
Token pin
Token ID6757
Feature activation-1.16e-01

Pos logit contributions
 Hood+0.044
ner+0.039
ussy+0.039
 treadmill+0.039
fighters+0.038
Neg logit contributions
 mouth-0.033
c-0.031
 wing-0.031
el-0.030
 feathers-0.029
Token began
Token ID2540
Feature activation+3.56e-02

Pos logit contributions
 mouth+0.155
c+0.153
 wing+0.147
el+0.146
 its+0.138
Neg logit contributions
 Hood-0.204
ussy-0.183
 treadmill-0.183
ner-0.181
assy-0.175
Token to
Token ID284
Feature activation-7.21e-03

Pos logit contributions
 Hood+0.047
 treadmill+0.044
ner+0.044
ilda+0.042
drivers+0.041
Neg logit contributions
c-0.055
el-0.054
 mouth-0.053
 wing-0.052
 pouch-0.051
TokenL
Token ID43
Feature activation+1.15e-02

Pos logit contributions
c+0.017
 mouth+0.017
 wing+0.016
el+0.016
 its+0.015
Neg logit contributions
 Hood-0.024
ner-0.022
ussy-0.022
 treadmill-0.021
fighters-0.020
Tokenily
Token ID813
Feature activation+1.62e-01

Pos logit contributions
 treadmill+0.019
 Hood+0.019
ner+0.017
working+0.017
cases+0.017
Neg logit contributions
c-0.015
el-0.015
 wing-0.014
 its-0.014
 mouth-0.014
Token looked
Token ID3114
Feature activation-7.94e-02

Pos logit contributions
 Hood+0.276
ner+0.250
 treadmill+0.249
ussy+0.247
fighters+0.239
Neg logit contributions
 mouth-0.219
c-0.214
 wing-0.210
el-0.210
 its-0.198
Token at
Token ID379
Feature activation-1.41e-01

Pos logit contributions
 mouth+0.099
c+0.091
 its+0.088
el+0.088
 wing+0.087
Neg logit contributions
 Hood-0.149
ner-0.141
fighters-0.136
ussy-0.136
 treadmill-0.135
Token the
Token ID262
Feature activation+1.75e-02

Pos logit contributions
 mouth+0.175
el+0.168
c+0.167
 wing+0.163
 its+0.163
Neg logit contributions
 Hood-0.258
ner-0.239
 treadmill-0.235
ussy-0.234
fighters-0.230
Token soldiers
Token ID5795
Feature activation+2.25e-01

Pos logit contributions
ussy+0.029
 Hood+0.029
ner+0.027
assy+0.027
 treadmill+0.026
Neg logit contributions
 mouth-0.025
 wing-0.023
 feathers-0.022
 direction-0.022
 feather-0.022
Token and
Token ID290
Feature activation+1.97e-02

Pos logit contributions
 Hood+0.280
ner+0.277
 treadmill+0.260
cases+0.238
ilda+0.230
Neg logit contributions
c-0.375
 mouth-0.364
rys-0.361
el-0.358
 wing-0.353
Token noticed
Token ID6810
Feature activation+7.80e-03

Pos logit contributions
 Hood+0.025
ner+0.022
 treadmill+0.022
fighters+0.021
ilda+0.021
Neg logit contributions
el-0.033
c-0.033
 wing-0.033
rys-0.032
 mouth-0.030
Token that
Token ID326
Feature activation-2.77e-02

Pos logit contributions
 Hood+0.014
ner+0.013
 treadmill+0.013
ussy+0.013
fighters+0.012
Neg logit contributions
 mouth-0.010
c-0.010
el-0.009
 wing-0.009
 its-0.009
Token their
Token ID511
Feature activation-1.65e-01

Pos logit contributions
 mouth+0.039
el+0.039
c+0.039
 wing+0.038
light+0.036
Neg logit contributions
 Hood-0.045
ner-0.040
 treadmill-0.040
ussy-0.039
fighters-0.038
Token faces
Token ID6698
Feature activation-6.93e-02

Pos logit contributions
 mouth+0.235
 wing+0.218
el+0.211
c+0.211
 feathers+0.207
Neg logit contributions
 Hood-0.283
ner-0.265
ussy-0.259
 treadmill-0.253
assy-0.250
Token.
Token ID13
Feature activation+6.48e-03

Pos logit contributions
 Hood+0.223
 treadmill+0.210
ussy+0.209
ner+0.208
assy+0.204
Neg logit contributions
 mouth-0.116
c-0.106
 its-0.104
el-0.101
 wing-0.100
Token He
Token ID679
Feature activation+1.57e-02

Pos logit contributions
 Hood+0.009
 treadmill+0.009
ner+0.008
ussy+0.008
fighters+0.008
Neg logit contributions
c-0.010
el-0.010
light-0.010
 wing-0.010
 mouth-0.010
Token felt
Token ID2936
Feature activation+2.51e-02

Pos logit contributions
 Hood+0.024
 treadmill+0.021
ussy+0.021
ner+0.021
fighters+0.020
Neg logit contributions
 mouth-0.023
c-0.023
 wing-0.023
el-0.023
rys-0.022
Token sad
Token ID6507
Feature activation-5.54e-02

Pos logit contributions
 Hood+0.039
ner+0.036
 treadmill+0.035
ussy+0.035
fighters+0.032
Neg logit contributions
c-0.037
el-0.035
 mouth-0.035
light-0.035
 wing-0.034
Token because
Token ID780
Feature activation-1.45e-01

Pos logit contributions
 mouth+0.075
c+0.075
el+0.072
 wing+0.072
light+0.067
Neg logit contributions
 Hood-0.094
ner-0.086
 treadmill-0.085
ussy-0.084
fighters-0.081
Token he
Token ID339
Feature activation+3.65e-02

Pos logit contributions
c+0.258
 mouth+0.252
 wing+0.243
light+0.241
el+0.241
Neg logit contributions
 Hood-0.188
ner-0.168
 treadmill-0.158
ussy-0.152
fighters-0.149
Token wanted
Token ID2227
Feature activation-9.84e-03

Pos logit contributions
 Hood+0.067
ner+0.066
 treadmill+0.064
ussy+0.062
cakes+0.062
Neg logit contributions
 mouth-0.044
c-0.042
el-0.041
 wing-0.040
 fin-0.039
Token to
Token ID284
Feature activation-1.02e-02

Pos logit contributions
c+0.017
el+0.016
light+0.015
 wing+0.015
 mouth+0.015
Neg logit contributions
 Hood-0.014
 treadmill-0.012
ner-0.011
ussy-0.011
fighters-0.011
Token play
Token ID711
Feature activation+2.50e-02

Pos logit contributions
c+0.015
 mouth+0.015
el+0.015
 wing+0.014
light+0.014
Neg logit contributions
 Hood-0.016
ner-0.014
 treadmill-0.014
ussy-0.013
fighters-0.013
Token in
Token ID287
Feature activation-3.10e-02

Pos logit contributions
 Hood+0.048
ner+0.044
 treadmill+0.044
ussy+0.043
fighters+0.041
Neg logit contributions
 mouth-0.029
c-0.028
el-0.028
 wing-0.027
light-0.025
Token the
Token ID262
Feature activation+1.93e-02

Pos logit contributions
 mouth+0.041
c+0.039
el+0.038
 wing+0.038
 its+0.037
Neg logit contributions
 Hood-0.056
ussy-0.051
ner-0.051
 treadmill-0.050
fighters-0.049
Token on
Token ID319
Feature activation-2.66e-02

Pos logit contributions
 Hood+0.274
ner+0.248
 treadmill+0.245
ussy+0.244
fighters+0.236
Neg logit contributions
 mouth-0.222
c-0.215
el-0.211
 wing-0.210
 its-0.199
Token it
Token ID340
Feature activation-4.21e-02

Pos logit contributions
 mouth+0.035
c+0.035
 wing+0.033
el+0.033
light+0.032
Neg logit contributions
 Hood-0.046
ner-0.042
 treadmill-0.041
ussy-0.040
fighters-0.039
Token.
Token ID13
Feature activation+1.15e-02

Pos logit contributions
 mouth+0.064
c+0.062
 wing+0.061
el+0.061
 feathers+0.058
Neg logit contributions
 Hood-0.064
ner-0.058
 treadmill-0.057
ussy-0.056
fighters-0.054
Token Tim
Token ID5045
Feature activation+3.23e-02

Pos logit contributions
 Hood+0.017
ner+0.016
 treadmill+0.016
ussy+0.016
fighters+0.015
Neg logit contributions
 mouth-0.017
 wing-0.017
c-0.017
el-0.017
light-0.016
Tokenmy
Token ID1820
Feature activation+7.85e-02

Pos logit contributions
 Hood+0.052
ner+0.051
 treadmill+0.049
fighters+0.047
cakes+0.046
Neg logit contributions
c-0.049
el-0.047
 mouth-0.047
 wing-0.044
light-0.043
Token loved
Token ID6151
Feature activation+1.49e-01

Pos logit contributions
 Hood+0.136
ner+0.124
 treadmill+0.122
ussy+0.122
fighters+0.118
Neg logit contributions
 mouth-0.107
c-0.103
el-0.101
 wing-0.100
 its-0.095
Token his
Token ID465
Feature activation-8.85e-02

Pos logit contributions
 Hood+0.267
ner+0.242
 treadmill+0.242
ussy+0.238
fighters+0.230
Neg logit contributions
 mouth-0.192
c-0.186
el-0.182
 wing-0.181
light-0.169
Token new
Token ID649
Feature activation+5.22e-02

Pos logit contributions
 mouth+0.119
c+0.113
 wing+0.112
el+0.111
 its+0.105
Neg logit contributions
 Hood-0.155
ner-0.142
ussy-0.141
 treadmill-0.140
fighters-0.136
Token towel
Token ID24808
Feature activation-8.94e-02

Pos logit contributions
 Hood+0.093
 treadmill+0.084
ner+0.083
ussy+0.080
ilda+0.078
Neg logit contributions
 mouth-0.065
c-0.065
el-0.065
 wing-0.061
light-0.059
Token and
Token ID290
Feature activation+2.85e-02

Pos logit contributions
 mouth+0.124
c+0.121
el+0.119
 wing+0.118
 its+0.111
Neg logit contributions
 Hood-0.148
ner-0.133
 treadmill-0.132
ussy-0.131
assy-0.126
Token couldn
Token ID3521
Feature activation+6.36e-02

Pos logit contributions
 Hood+0.046
ner+0.041
 treadmill+0.041
ussy+0.040
fighters+0.039
Neg logit contributions
 mouth-0.041
c-0.040
el-0.040
 wing-0.039
light-0.037
Token the
Token ID262
Feature activation+2.12e-02

Pos logit contributions
 mouth+0.041
c+0.040
 wing+0.039
el+0.039
 feathers+0.036
Neg logit contributions
 Hood-0.048
ner-0.044
 treadmill-0.043
ussy-0.042
fighters-0.041
Token corner
Token ID5228
Feature activation-2.51e-02

Pos logit contributions
 Hood+0.038
ussy+0.036
ner+0.036
 treadmill+0.035
fighters+0.034
Neg logit contributions
 mouth-0.028
c-0.026
 wing-0.025
el-0.025
 direction-0.025
Token.
Token ID13
Feature activation+1.98e-02

Pos logit contributions
 wing+0.041
 mouth+0.040
c+0.039
el+0.039
rys+0.039
Neg logit contributions
 Hood-0.034
ner-0.032
fighters-0.030
boards-0.029
 treadmill-0.029
Token They
Token ID1119
Feature activation-1.78e-02

Pos logit contributions
 Hood+0.036
ussy+0.033
ner+0.033
 treadmill+0.032
assy+0.031
Neg logit contributions
 mouth-0.026
c-0.024
el-0.024
 wing-0.023
 its-0.023
Token both
Token ID1111
Feature activation+3.25e-02

Pos logit contributions
 mouth+0.023
el+0.022
c+0.022
 wing+0.021
 its+0.021
Neg logit contributions
 Hood-0.031
ussy-0.029
ner-0.029
 treadmill-0.028
fighters-0.027
Token ran
Token ID4966
Feature activation+8.65e-02

Pos logit contributions
 Hood+0.059
ussy+0.053
 treadmill+0.053
ner+0.053
fighters+0.051
Neg logit contributions
 mouth-0.043
c-0.041
el-0.040
 wing-0.039
 its-0.038
Token to
Token ID284
Feature activation-7.45e-03

Pos logit contributions
 Hood+0.172
ussy+0.172
assy+0.162
oons+0.161
ner+0.158
Neg logit contributions
 its-0.092
 mouth-0.089
el-0.078
 feathers-0.076
 be-0.076
Token it
Token ID340
Feature activation-4.83e-02

Pos logit contributions
 mouth+0.010
 its+0.009
el+0.009
c+0.008
 wing+0.008
Neg logit contributions
 Hood-0.014
ussy-0.013
 treadmill-0.012
ner-0.012
assy-0.012
Token and
Token ID290
Feature activation+2.88e-02

Pos logit contributions
 mouth+0.059
c+0.056
el+0.054
 wing+0.053
 its+0.052
Neg logit contributions
 Hood-0.093
ussy-0.087
ner-0.083
fighters-0.083
 treadmill-0.083
Token Joe
Token ID5689
Feature activation-2.03e-02

Pos logit contributions
 Hood+0.050
ussy+0.046
ner+0.046
 treadmill+0.045
fighters+0.044
Neg logit contributions
 mouth-0.039
c-0.037
el-0.036
 wing-0.036
 its-0.035
Token said
Token ID531
Feature activation-9.67e-02

Pos logit contributions
 mouth+0.022
c+0.021
el+0.019
 neck+0.018
 its+0.018
Neg logit contributions
 Hood-0.042
ussy-0.039
 treadmill-0.039
ner-0.038
fighters-0.038
Token wheel
Token ID7825
Feature activation-8.56e-02

Pos logit contributions
 Hood+0.067
 treadmill+0.062
ner+0.062
ussy+0.058
fighters+0.057
Neg logit contributions
el-0.058
c-0.056
 mouth-0.055
 wing-0.053
light-0.052
Token in
Token ID287
Feature activation-3.85e-02

Pos logit contributions
 mouth+0.119
c+0.114
 wing+0.112
el+0.112
 its+0.106
Neg logit contributions
 Hood-0.145
ner-0.131
 treadmill-0.129
ussy-0.128
fighters-0.124
Token the
Token ID262
Feature activation+9.73e-03

Pos logit contributions
 mouth+0.060
c+0.059
 wing+0.059
light+0.056
el+0.056
Neg logit contributions
 Hood-0.056
ner-0.051
 treadmill-0.051
ussy-0.049
assy-0.047
Token box
Token ID3091
Feature activation-3.55e-02

Pos logit contributions
 Hood+0.014
 treadmill+0.013
ussy+0.013
ner+0.013
assy+0.012
Neg logit contributions
 mouth-0.015
el-0.015
c-0.014
 wing-0.014
light-0.014
Token of
Token ID286
Feature activation-7.42e-02

Pos logit contributions
 mouth+0.054
 wing+0.053
c+0.053
el+0.053
rys+0.052
Neg logit contributions
 Hood-0.053
ner-0.048
 treadmill-0.046
ussy-0.045
assy-0.044
Token blocks
Token ID7021
Feature activation+1.61e-01

Pos logit contributions
 mouth+0.112
c+0.109
 wing+0.107
el+0.105
light+0.103
Neg logit contributions
 Hood-0.113
ner-0.104
 treadmill-0.103
ussy-0.099
fighters-0.098
Token.
Token ID13
Feature activation-8.02e-03

Pos logit contributions
 Hood+0.249
ner+0.228
 treadmill+0.219
ussy+0.218
assy+0.209
Neg logit contributions
 mouth-0.245
 wing-0.236
c-0.236
el-0.232
 feathers-0.222
Token
Token ID198
Feature activation-1.91e-02

Pos logit contributions
 mouth+0.011
c+0.010
el+0.010
 wing+0.010
 its+0.010
Neg logit contributions
 Hood-0.014
ner-0.013
 treadmill-0.013
ussy-0.013
fighters-0.012
Token
Token ID198
Feature activation-1.96e-02

Pos logit contributions
 mouth+0.019
c+0.019
el+0.018
 its+0.018
 wing+0.016
Neg logit contributions
 Hood-0.040
ussy-0.036
 treadmill-0.036
ner-0.036
fighters-0.035
Token"
Token ID1
Feature activation+8.87e-02

Pos logit contributions
 mouth+0.032
 wing+0.031
c+0.030
el+0.030
light+0.029
Neg logit contributions
 Hood-0.028
ner-0.027
 treadmill-0.026
ussy-0.025
fighters-0.025
TokenBen
Token ID11696
Feature activation+6.82e-02

Pos logit contributions
 Hood+0.133
ner+0.125
fighters+0.120
 treadmill+0.118
ussy+0.118
Neg logit contributions
 mouth-0.143
 wing-0.137
el-0.132
c-0.132
 its-0.125
Token a
Token ID257
Feature activation+5.02e-02

Pos logit contributions
 mouth+0.079
 its+0.077
c+0.076
 wing+0.073
el+0.071
Neg logit contributions
 Hood-0.121
ussy-0.115
 treadmill-0.114
ner-0.112
fighters-0.110
Token time
Token ID640
Feature activation-8.50e-02

Pos logit contributions
 Hood+0.100
 treadmill+0.091
ner+0.091
ussy+0.091
assy+0.088
Neg logit contributions
 mouth-0.054
c-0.052
 wing-0.051
el-0.051
 its-0.047
Token,
Token ID11
Feature activation+1.75e-02

Pos logit contributions
 mouth+0.111
c+0.107
el+0.105
 wing+0.103
 its+0.100
Neg logit contributions
 Hood-0.151
ussy-0.137
 treadmill-0.137
ner-0.137
fighters-0.132
Token there
Token ID612
Feature activation-7.59e-02

Pos logit contributions
 Hood+0.035
ner+0.032
 treadmill+0.031
ussy+0.031
fighters+0.030
Neg logit contributions
 mouth-0.020
c-0.019
 wing-0.018
el-0.018
 feathers-0.017
Token was
Token ID373
Feature activation+3.09e-04

Pos logit contributions
 mouth+0.115
 wing+0.114
c+0.111
 feathers+0.108
el+0.107
Neg logit contributions
 Hood-0.115
ner-0.101
 treadmill-0.099
ussy-0.099
fighters-0.096
Token a
Token ID257
Feature activation+4.99e-02

Pos logit contributions
 Hood+0.001
 treadmill+0.001
ussy+0.001
ner+0.001
fighters+0.001
Neg logit contributions
 mouth-0.000
c-0.000
 its-0.000
el-0.000
 wing-0.000
Token little
Token ID1310
Feature activation+1.20e-01

Pos logit contributions
 Hood+0.089
ives+0.086
drivers+0.084
ussy+0.084
ner+0.083
Neg logit contributions
 mouth-0.068
 wing-0.062
 feather-0.062
 nest-0.061
c-0.059
Token girl
Token ID2576
Feature activation+4.54e-02

Pos logit contributions
 Hood+0.203
ussy+0.187
ner+0.186
 treadmill+0.186
fighters+0.181
Neg logit contributions
 mouth-0.168
c-0.160
 wing-0.160
el-0.159
 feather-0.151
Token named
Token ID3706
Feature activation-1.26e-01

Pos logit contributions
 Hood+0.076
assy+0.064
ner+0.064
ilda+0.063
ussy+0.061
Neg logit contributions
 mouth-0.061
 mouths-0.057
 wing-0.056
el-0.054
 fin-0.054
Token Lily
Token ID20037
Feature activation+4.98e-02

Pos logit contributions
 mouth+0.152
c+0.145
 its+0.140
el+0.136
 wing+0.130
Neg logit contributions
 Hood-0.239
 treadmill-0.221
ussy-0.220
ner-0.218
fighters-0.215
Token.
Token ID13
Feature activation+1.29e-02

Pos logit contributions
 Hood+0.080
ner+0.071
ussy+0.070
 treadmill+0.070
assy+0.068
Neg logit contributions
 mouth-0.074
c-0.070
 wing-0.069
el-0.069
 feathers-0.065
Token
Token ID198
Feature activation-1.04e-02

Pos logit contributions
 mouth+0.107
c+0.105
el+0.101
 wing+0.099
 its+0.097
Neg logit contributions
 Hood-0.161
ner-0.148
 treadmill-0.147
ussy-0.145
fighters-0.141
Token
Token ID198
Feature activation-1.07e-02

Pos logit contributions
 mouth+0.014
c+0.013
el+0.013
 wing+0.013
 its+0.012
Neg logit contributions
 Hood-0.018
ner-0.017
 treadmill-0.016
ussy-0.016
fighters-0.016
TokenOne
Token ID3198
Feature activation-6.17e-02

Pos logit contributions
 mouth+0.017
 wing+0.016
el+0.015
 tongue+0.015
 mouths+0.015
Neg logit contributions
 Hood-0.017
 treadmill-0.015
ner-0.015
fighters-0.014
ussy-0.014
Token day
Token ID1110
Feature activation-8.63e-02

Pos logit contributions
 mouth+0.090
 wing+0.087
 mouths+0.084
c+0.084
el+0.083
Neg logit contributions
 Hood-0.098
working-0.086
assy-0.081
ussy-0.079
 treadmill-0.079
Token,
Token ID11
Feature activation+1.51e-02

Pos logit contributions
 mouth+0.125
c+0.123
 wing+0.120
el+0.119
 feathers+0.112
Neg logit contributions
 Hood-0.139
ner-0.126
 treadmill-0.124
ussy-0.121
fighters-0.119
Token Bella
Token ID39304
Feature activation-5.15e-02

Pos logit contributions
 Hood+0.025
ner+0.023
 treadmill+0.023
fighters+0.022
ussy+0.022
Neg logit contributions
 mouth-0.021
c-0.020
 wing-0.020
el-0.020
light-0.018
Token wanted
Token ID2227
Feature activation-1.24e-02

Pos logit contributions
c+0.054
 mouth+0.049
el+0.046
light+0.043
 wing+0.042
Neg logit contributions
ner-0.109
ussy-0.108
 Hood-0.108
 treadmill-0.108
drivers-0.103
Token a
Token ID257
Feature activation+4.76e-02

Pos logit contributions
 mouth+0.017
c+0.017
el+0.017
 wing+0.017
light+0.016
Neg logit contributions
 Hood-0.021
ner-0.018
 treadmill-0.018
ussy-0.018
assy-0.017
Token pastry
Token ID41738
Feature activation+3.75e-02

Pos logit contributions
 Hood+0.072
 treadmill+0.064
ner+0.061
cakes+0.060
ussy+0.060
Neg logit contributions
el-0.075
light-0.072
 wing-0.072
 mouth-0.072
c-0.071
Token from
Token ID422
Feature activation-1.07e-01

Pos logit contributions
 Hood+0.052
 treadmill+0.046
ner+0.045
cakes+0.045
ussy+0.045
Neg logit contributions
ric-0.060
c-0.059
 wing-0.059
el-0.058
 mouth-0.057
Token the
Token ID262
Feature activation+1.74e-02

Pos logit contributions
 mouth+0.142
c+0.138
 wing+0.134
el+0.133
light+0.126
Neg logit contributions
 Hood-0.191
ner-0.172
 treadmill-0.171
ussy-0.170
fighters-0.163
Token they
Token ID484
Feature activation-2.04e-02

Pos logit contributions
c+0.093
 mouth+0.093
el+0.091
 wing+0.089
light+0.084
Neg logit contributions
 Hood-0.106
ner-0.097
 treadmill-0.094
ussy-0.092
fighters-0.090
Token can
Token ID460
Feature activation+8.40e-02

Pos logit contributions
 mouth+0.016
 be+0.015
c+0.014
el+0.014
 mouths+0.013
Neg logit contributions
 treadmill-0.045
 Hood-0.045
ner-0.045
ilda-0.045
ussy-0.043
Token visit
Token ID3187
Feature activation+3.57e-02

Pos logit contributions
 Hood+0.139
ner+0.126
 treadmill+0.125
ussy+0.124
fighters+0.120
Neg logit contributions
 mouth-0.118
c-0.115
 wing-0.112
el-0.111
 its-0.106
Token Tom
Token ID4186
Feature activation+4.33e-02

Pos logit contributions
 Hood+0.067
ner+0.062
 treadmill+0.061
ussy+0.061
fighters+0.059
Neg logit contributions
 mouth-0.043
c-0.041
el-0.041
 wing-0.040
 its-0.039
Token's
Token ID338
Feature activation-1.03e-01

Pos logit contributions
 Hood+0.083
ner+0.078
 treadmill+0.077
ussy+0.076
fighters+0.075
Neg logit contributions
c-0.050
 mouth-0.049
el-0.046
 wing-0.045
 its-0.044
Token house
Token ID2156
Feature activation-1.08e-02

Pos logit contributions
 mouth+0.148
c+0.146
el+0.142
 wing+0.142
 its+0.136
Neg logit contributions
 Hood-0.167
 treadmill-0.148
ussy-0.146
ner-0.146
assy-0.141
Token.
Token ID13
Feature activation-1.26e-02

Pos logit contributions
 mouth+0.013
c+0.013
el+0.012
 its+0.012
 wing+0.012
Neg logit contributions
 Hood-0.020
ner-0.019
 treadmill-0.018
ussy-0.018
fighters-0.018
Token Lily
Token ID20037
Feature activation+3.72e-02

Pos logit contributions
 wing+0.024
el+0.023
 feather+0.023
 feathers+0.022
light+0.022
Neg logit contributions
 treadmill-0.014
 Hood-0.014
ner-0.014
ussy-0.013
cakes-0.013
Token is
Token ID318
Feature activation-1.63e-02

Pos logit contributions
ner+0.074
 Hood+0.073
drivers+0.070
ussy+0.069
fighters+0.068
Neg logit contributions
 mouth-0.042
c-0.036
el-0.035
 be-0.034
 mouths-0.033
Token happy
Token ID3772
Feature activation-7.15e-03

Pos logit contributions
 mouth+0.023
c+0.023
el+0.023
 wing+0.022
light+0.022
Neg logit contributions
 Hood-0.027
ner-0.024
ussy-0.023
 treadmill-0.023
assy-0.022
Token.
Token ID13
Feature activation-1.25e-02

Pos logit contributions
c+0.013
el+0.012
 wing+0.012
 mouth+0.012
rys+0.012
Neg logit contributions
 Hood-0.009
ner-0.008
 treadmill-0.008
assy-0.007
ilda-0.007
Token an
Token ID281
Feature activation-9.23e-03

Pos logit contributions
c+0.025
el+0.025
 mouth+0.025
 wing+0.024
light+0.024
Neg logit contributions
 Hood-0.025
ner-0.021
 treadmill-0.021
ussy-0.020
ilda-0.020
Token idea
Token ID2126
Feature activation-4.47e-02

Pos logit contributions
 mouth+0.012
c+0.011
el+0.011
 wing+0.011
 its+0.010
Neg logit contributions
 Hood-0.017
ner-0.015
 treadmill-0.015
ussy-0.015
fighters-0.015
Token.
Token ID13
Feature activation+8.01e-03

Pos logit contributions
 mouth+0.069
c+0.067
 wing+0.066
el+0.065
 feathers+0.062
Neg logit contributions
 Hood-0.069
ner-0.062
 treadmill-0.061
ussy-0.059
fighters-0.058
Token He
Token ID679
Feature activation+1.70e-02

Pos logit contributions
 Hood+0.015
ner+0.013
ussy+0.013
fighters+0.013
assy+0.013
Neg logit contributions
 mouth-0.011
c-0.009
el-0.009
 wing-0.009
 its-0.009
Token said
Token ID531
Feature activation-9.88e-02

Pos logit contributions
ner+0.032
 Hood+0.032
 treadmill+0.031
ussy+0.031
assy+0.029
Neg logit contributions
 mouth-0.020
c-0.019
 wing-0.018
 feather-0.018
el-0.018
Token,
Token ID11
Feature activation+1.47e-02

Pos logit contributions
 mouth+0.117
 wing+0.110
 its+0.110
c+0.100
el+0.100
Neg logit contributions
 Hood-0.193
ussy-0.183
assy-0.178
 treadmill-0.177
ner-0.175
Token "
Token ID366
Feature activation+5.52e-02

Pos logit contributions
 Hood+0.026
ussy+0.024
 treadmill+0.023
assy+0.023
ner+0.023
Neg logit contributions
 mouth-0.020
 its-0.018
 wing-0.018
c-0.018
el-0.017
TokenLet
Token ID5756
Feature activation+1.01e-01

Pos logit contributions
 Hood+0.076
ner+0.066
 treadmill+0.065
ussy+0.065
fighters+0.063
Neg logit contributions
 mouth-0.097
c-0.092
 wing-0.092
el-0.091
 its-0.087
Token's
Token ID338
Feature activation-9.68e-02

Pos logit contributions
 Hood+0.240
ner+0.231
 treadmill+0.229
ussy+0.228
fighters+0.219
Neg logit contributions
el-0.072
 mouth-0.072
c-0.071
 wing-0.066
 its-0.063
Token unite
Token ID24558
Feature activation+8.76e-02

Pos logit contributions
 mouth+0.133
c+0.128
el+0.126
 wing+0.125
 its+0.119
Neg logit contributions
 Hood-0.166
ner-0.150
 treadmill-0.148
ussy-0.148
fighters-0.143
Token our
Token ID674
Feature activation-1.41e-01

Pos logit contributions
 Hood+0.153
 treadmill+0.139
ner+0.138
ussy+0.136
fighters+0.134
Neg logit contributions
 mouth-0.113
c-0.111
 wing-0.108
el-0.107
light-0.100
Token some
Token ID617
Feature activation-1.77e-02

Pos logit contributions
 Hood+0.002
ner+0.002
ussy+0.002
 treadmill+0.002
fighters+0.002
Neg logit contributions
 mouth-0.001
c-0.001
 wing-0.001
el-0.001
 its-0.001
Token glue
Token ID22749
Feature activation+6.61e-02

Pos logit contributions
 mouth+0.028
c+0.026
 wing+0.026
el+0.026
 feathers+0.025
Neg logit contributions
 Hood-0.028
ner-0.025
ussy-0.025
 treadmill-0.024
assy-0.024
Token and
Token ID290
Feature activation+1.80e-02

Pos logit contributions
 Hood+0.119
 treadmill+0.112
ner+0.112
ussy+0.109
assy+0.106
Neg logit contributions
 mouth-0.085
c-0.082
 wing-0.080
el-0.079
 its-0.076
Token some
Token ID617
Feature activation-1.89e-02

Pos logit contributions
 Hood+0.029
ussy+0.026
ner+0.026
 treadmill+0.026
assy+0.025
Neg logit contributions
 mouth-0.027
 wing-0.026
c-0.026
el-0.025
 its-0.025
Token tape
Token ID9154
Feature activation+4.52e-02

Pos logit contributions
 mouth+0.029
c+0.028
 wing+0.028
el+0.028
 its+0.027
Neg logit contributions
 Hood-0.029
ner-0.026
ussy-0.026
 treadmill-0.026
fighters-0.024
Token.
Token ID13
Feature activation-2.59e-03

Pos logit contributions
 Hood+0.077
 treadmill+0.070
ner+0.070
ussy+0.069
assy+0.067
Neg logit contributions
 mouth-0.063
c-0.061
 wing-0.059
el-0.059
 its-0.057
Token He
Token ID679
Feature activation+1.06e-02

Pos logit contributions
 mouth+0.003
c+0.003
el+0.003
 wing+0.003
 its+0.003
Neg logit contributions
 Hood-0.005
ner-0.004
 treadmill-0.004
ussy-0.004
fighters-0.004
Token will
Token ID481
Feature activation+8.44e-02

Pos logit contributions
 Hood+0.019
 treadmill+0.019
ner+0.018
ussy+0.018
assy+0.017
Neg logit contributions
 mouth-0.014
c-0.014
el-0.013
 wing-0.012
 feathers-0.012
Token be
Token ID307
Feature activation-2.75e-01

Pos logit contributions
 Hood+0.159
ner+0.147
 treadmill+0.145
ussy+0.144
fighters+0.141
Neg logit contributions
 mouth-0.099
c-0.097
el-0.093
 wing-0.091
light-0.086
Token good
Token ID922
Feature activation+1.28e-01

Pos logit contributions
 mouth+0.375
c+0.359
 wing+0.352
 its+0.341
el+0.341
Neg logit contributions
 Hood-0.488
ner-0.449
 treadmill-0.442
ussy-0.434
fighters-0.429
Token as
Token ID355
Feature activation-5.53e-02

Pos logit contributions
 treadmill+0.285
 Hood+0.278
ussy+0.268
ival+0.267
ilda+0.261
Neg logit contributions
 mouth-0.106
 wing-0.101
 feathers-0.085
el-0.083
 be-0.079
Token furry
Token ID40857
Feature activation-7.25e-02

Pos logit contributions
 mouth+0.015
 wing+0.015
 feather+0.014
 nest+0.014
 feathers+0.014
Neg logit contributions
 Hood-0.020
ussy-0.018
ner-0.018
assy-0.018
ilda-0.018
Token bear
Token ID6842
Feature activation+1.13e-01

Pos logit contributions
 mouth+0.103
c+0.099
el+0.097
 wing+0.097
 its+0.092
Neg logit contributions
 Hood-0.121
ner-0.109
 treadmill-0.108
ussy-0.108
fighters-0.104
Token was
Token ID373
Feature activation+1.70e-03

Pos logit contributions
 Hood+0.177
ner+0.157
 treadmill+0.154
ussy+0.152
fighters+0.149
Neg logit contributions
 mouth-0.163
el-0.158
 wing-0.158
c-0.156
rys-0.149
Token happy
Token ID3772
Feature activation+2.95e-03

Pos logit contributions
 Hood+0.003
ussy+0.002
ner+0.002
 treadmill+0.002
assy+0.002
Neg logit contributions
el-0.002
c-0.002
 mouth-0.002
rys-0.002
 wing-0.002
Token in
Token ID287
Feature activation-2.78e-02

Pos logit contributions
 Hood+0.004
ner+0.003
boards+0.003
 treadmill+0.003
ilda+0.003
Neg logit contributions
c-0.005
el-0.005
 wing-0.005
rys-0.005
 pouch-0.005
Token his
Token ID465
Feature activation-8.84e-02

Pos logit contributions
c+0.052
light+0.049
 wing+0.049
 cardinal+0.045
el+0.045
Neg logit contributions
 Hood-0.030
 treadmill-0.026
ner-0.026
 tables-0.024
 downstairs-0.024
Token shade
Token ID17979
Feature activation-9.37e-02

Pos logit contributions
cr+0.118
el+0.117
c+0.117
rys+0.116
 oval+0.115
Neg logit contributions
 Hood-0.130
 treadmill-0.115
ussy-0.104
assy-0.101
ilda-0.100
Token wrap
Token ID14441
Feature activation-3.16e-02

Pos logit contributions
 mouth+0.141
 wing+0.138
c+0.137
el+0.137
rys+0.135
Neg logit contributions
 Hood-0.140
ner-0.124
 treadmill-0.121
ussy-0.119
boards-0.117
Token.
Token ID13
Feature activation+1.04e-02

Pos logit contributions
 mouth+0.045
c+0.044
el+0.043
 wing+0.043
 oval+0.041
Neg logit contributions
 Hood-0.051
 treadmill-0.045
ner-0.045
ussy-0.044
fighters-0.043
Token He
Token ID679
Feature activation+1.83e-02

Pos logit contributions
 Hood+0.016
ner+0.015
 treadmill+0.015
ussy+0.014
fighters+0.014
Neg logit contributions
 mouth-0.015
c-0.015
 wing-0.015
el-0.015
light-0.014
Token could
Token ID714
Feature activation+4.60e-02

Pos logit contributions
 Hood+0.030
ner+0.027
 treadmill+0.026
ussy+0.026
fighters+0.025
Neg logit contributions
 mouth-0.026
c-0.025
el-0.025
 wing-0.025
 its-0.023
Token He
Token ID679
Feature activation+7.52e-03

Pos logit contributions
 mouth+0.009
c+0.009
 wing+0.009
el+0.009
light+0.008
Neg logit contributions
 Hood-0.010
ner-0.009
 treadmill-0.009
ussy-0.009
fighters-0.009
Token tried
Token ID3088
Feature activation+3.24e-02

Pos logit contributions
 Hood+0.013
ner+0.012
ussy+0.012
 treadmill+0.012
assy+0.011
Neg logit contributions
 mouth-0.010
c-0.010
el-0.010
 wing-0.009
 its-0.009
Token to
Token ID284
Feature activation-1.62e-02

Pos logit contributions
 Hood+0.053
 treadmill+0.048
ner+0.048
ussy+0.047
fighters+0.045
Neg logit contributions
c-0.046
 mouth-0.046
el-0.045
 wing-0.045
light-0.043
Token grab
Token ID5552
Feature activation-2.85e-02

Pos logit contributions
 mouth+0.020
el+0.019
c+0.019
 wing+0.019
 its+0.019
Neg logit contributions
 Hood-0.030
ner-0.027
ussy-0.027
 treadmill-0.026
assy-0.026
Token the
Token ID262
Feature activation+1.68e-02

Pos logit contributions
 mouth+0.038
c+0.037
 wing+0.036
el+0.036
light+0.034
Neg logit contributions
 Hood-0.049
ner-0.045
 treadmill-0.044
ussy-0.044
fighters-0.042
Token chain
Token ID6333
Feature activation-1.92e-01

Pos logit contributions
 Hood+0.037
ner+0.034
 treadmill+0.034
ussy+0.033
fighters+0.033
Neg logit contributions
 mouth-0.014
c-0.014
el-0.013
 wing-0.013
light-0.013
Token from
Token ID422
Feature activation-1.14e-01

Pos logit contributions
 mouth+0.256
c+0.247
el+0.242
 wing+0.242
 its+0.228
Neg logit contributions
 Hood-0.335
ner-0.304
 treadmill-0.300
ussy-0.300
fighters-0.289
Token them
Token ID606
Feature activation+2.30e-02

Pos logit contributions
 its+0.140
 mouth+0.137
el+0.133
c+0.122
 feathers+0.121
Neg logit contributions
 Hood-0.217
ussy-0.203
ner-0.203
assy-0.202
ives-0.202
Token.
Token ID13
Feature activation-6.87e-03

Pos logit contributions
 Hood+0.034
ner+0.031
 treadmill+0.031
ussy+0.030
fighters+0.028
Neg logit contributions
c-0.035
 mouth-0.035
 wing-0.035
el-0.034
light-0.033
Token
Token ID198
Feature activation-1.77e-02

Pos logit contributions
 mouth+0.009
c+0.008
el+0.008
 wing+0.008
 its+0.008
Neg logit contributions
 Hood-0.013
ner-0.012
 treadmill-0.011
ussy-0.011
fighters-0.011
Token
Token ID198
Feature activation-1.73e-02

Pos logit contributions
 its+0.013
 mouth+0.011
c+0.011
el+0.011
 be+0.009
Neg logit contributions
 Hood-0.043
ussy-0.040
 treadmill-0.039
working-0.039
ner-0.039
Token It
Token ID632
Feature activation-9.10e-02

Pos logit contributions
 Hood+0.008
ner+0.008
 treadmill+0.007
ussy+0.007
fighters+0.007
Neg logit contributions
 wing-0.008
el-0.008
c-0.008
 mouth-0.008
 feathers-0.008
Token's
Token ID338
Feature activation-8.75e-02

Pos logit contributions
 mouth+0.132
el+0.128
 wing+0.128
c+0.125
light+0.119
Neg logit contributions
 Hood-0.147
 treadmill-0.127
ussy-0.127
ner-0.126
fighters-0.125
Token purple
Token ID14032
Feature activation-7.52e-02

Pos logit contributions
el+0.145
c+0.143
rys+0.143
light+0.138
 cardinal+0.130
Neg logit contributions
 Hood-0.106
 treadmill-0.093
ussy-0.092
fighters-0.087
 scientists-0.084
Token!"
Token ID2474
Feature activation+1.78e-01

Pos logit contributions
c+0.121
rys+0.118
el+0.116
 mouth+0.114
 wing+0.114
Neg logit contributions
 Hood-0.094
ner-0.084
 treadmill-0.080
oons-0.079
boards-0.079
Token Tom
Token ID4186
Feature activation+5.07e-02

Pos logit contributions
 Hood+0.201
ner+0.199
 treadmill+0.186
assy+0.176
cakes+0.168
Neg logit contributions
 wing-0.313
el-0.311
c-0.309
light-0.301
 feather-0.297
Token replied
Token ID8712
Feature activation-2.22e-01

Pos logit contributions
 Hood+0.102
ner+0.094
 treadmill+0.092
ussy+0.091
fighters+0.091
Neg logit contributions
 mouth-0.058
c-0.054
el-0.052
 its-0.049
 wing-0.049
Token,
Token ID11
Feature activation+1.27e-02

Pos logit contributions
 mouth+0.308
c+0.293
 wing+0.288
el+0.287
 its+0.277
Neg logit contributions
 Hood-0.384
ner-0.345
ussy-0.342
 treadmill-0.342
fighters-0.331
Token "
Token ID366
Feature activation+5.77e-02

Pos logit contributions
 Hood+0.019
ner+0.017
 treadmill+0.017
ussy+0.016
fighters+0.016
Neg logit contributions
 mouth-0.020
c-0.020
el-0.020
 wing-0.019
light-0.018
TokenIt
Token ID1026
Feature activation-9.54e-02

Pos logit contributions
ner+0.066
 Hood+0.061
ussy+0.059
fighters+0.059
ives+0.057
Neg logit contributions
 mouth-0.110
 eyes-0.104
 wing-0.101
 mouths-0.100
el-0.100
Token looks
Token ID3073
Feature activation-1.85e-02

Pos logit contributions
 mouth+0.109
c+0.104
 wing+0.103
el+0.103
 its+0.095
Neg logit contributions
 Hood-0.184
ner-0.167
ussy-0.166
 treadmill-0.165
fighters-0.161
Token like
Token ID588
Feature activation-4.93e-02

Pos logit contributions
c+0.027
el+0.027
 wing+0.027
 mouth+0.026
light+0.026
Neg logit contributions
 Hood-0.027
 treadmill-0.025
ussy-0.025
ner-0.024
fighters-0.023
TokenThe
Token ID464
Feature activation-8.41e-03

Pos logit contributions
 Hood+0.006
ner+0.005
ussy+0.005
 treadmill+0.005
fighters+0.005
Neg logit contributions
 mouth-0.005
c-0.005
el-0.005
 wing-0.005
 its-0.004
Token animals
Token ID4695
Feature activation+5.31e-02

Pos logit contributions
 mouth+0.012
 wing+0.011
 fin+0.010
 feather+0.010
 feathers+0.010
Neg logit contributions
 treadmill-0.014
ussy-0.014
 Hood-0.014
drivers-0.014
ives-0.014
Token all
Token ID477
Feature activation+1.02e-01

Pos logit contributions
 Hood+0.084
ner+0.076
fighters+0.071
 treadmill+0.070
ussy+0.069
Neg logit contributions
 mouth-0.081
c-0.077
 wing-0.075
el-0.074
 oval-0.073
Token started
Token ID2067
Feature activation+7.88e-02

Pos logit contributions
 Hood+0.155
ner+0.134
fighters+0.130
ussy+0.124
 treadmill+0.122
Neg logit contributions
 mouth-0.165
 wing-0.156
el-0.149
c-0.147
 tongue-0.145
Token to
Token ID284
Feature activation-1.14e-03

Pos logit contributions
 Hood+0.108
 treadmill+0.091
ner+0.090
fighters+0.088
 scientists+0.086
Neg logit contributions
c-0.128
el-0.124
 mouth-0.122
 pouch-0.120
 wing-0.119
Token trem
Token ID10023
Feature activation-2.24e-01

Pos logit contributions
 mouth+0.002
c+0.001
el+0.001
 wing+0.001
light+0.001
Neg logit contributions
 Hood-0.002
ner-0.002
 treadmill-0.002
ussy-0.002
fighters-0.002
Tokenble
Token ID903
Feature activation-9.38e-03

Pos logit contributions
 mouth+0.268
el+0.266
 wing+0.262
c+0.255
 its+0.245
Neg logit contributions
 Hood-0.422
 treadmill-0.386
ner-0.384
ussy-0.377
assy-0.371
Token.
Token ID13
Feature activation+1.93e-02

Pos logit contributions
 mouth+0.014
c+0.014
 wing+0.014
el+0.014
rys+0.013
Neg logit contributions
 Hood-0.014
ner-0.012
 treadmill-0.012
ussy-0.012
fighters-0.012
Token But
Token ID887
Feature activation-6.48e-02

Pos logit contributions
 Hood+0.032
ner+0.030
 treadmill+0.029
ussy+0.029
fighters+0.028
Neg logit contributions
 mouth-0.026
c-0.026
 wing-0.026
el-0.026
 feather-0.024
Token the
Token ID262
Feature activation+3.00e-02

Pos logit contributions
c+0.102
 mouth+0.099
 wing+0.099
el+0.098
rys+0.094
Neg logit contributions
 Hood-0.095
ner-0.086
fighters-0.082
 treadmill-0.080
ussy-0.077
Token little
Token ID1310
Feature activation+1.28e-01

Pos logit contributions
 Hood+0.048
ussy+0.045
 treadmill+0.045
ner+0.045
assy+0.044
Neg logit contributions
 mouth-0.045
 wing-0.042
c-0.041
el-0.040
 feathers-0.040
Token and
Token ID290
Feature activation+2.68e-02

Pos logit contributions
 Hood+0.224
ner+0.197
 treadmill+0.194
assy+0.188
ilda+0.188
Neg logit contributions
c-0.293
 mouth-0.291
 wing-0.284
el-0.276
light-0.273
Token held
Token ID2714
Feature activation-5.65e-02

Pos logit contributions
 Hood+0.048
ner+0.043
 treadmill+0.043
ussy+0.043
fighters+0.041
Neg logit contributions
 mouth-0.035
c-0.034
el-0.033
 wing-0.033
 its-0.031
Token it
Token ID340
Feature activation-4.77e-02

Pos logit contributions
 mouth+0.067
c+0.065
 wing+0.063
el+0.062
light+0.059
Neg logit contributions
 Hood-0.107
ner-0.098
 treadmill-0.097
ussy-0.096
fighters-0.093
Token in
Token ID287
Feature activation-3.06e-02

Pos logit contributions
ric+0.088
c+0.086
 cardinal+0.084
 mouth+0.083
rys+0.082
Neg logit contributions
 Hood-0.052
 treadmill-0.047
ner-0.043
cases-0.043
ilda-0.042
Token her
Token ID607
Feature activation-4.98e-02

Pos logit contributions
c+0.062
ric+0.062
light+0.058
fly+0.058
 wing+0.057
Neg logit contributions
 downstairs-0.023
ner-0.023
 work-0.023
 treadmill-0.022
assy-0.021
Token hand
Token ID1021
Feature activation-2.53e-01

Pos logit contributions
ectar+0.070
rim+0.067
 peac+0.067
fly+0.066
 cardinal+0.066
Neg logit contributions
 Hood-0.061
 hugs-0.057
ilda-0.054
 treadmill-0.053
assy-0.052
Token.
Token ID13
Feature activation+1.24e-02

Pos logit contributions
rys+0.469
 cardinal+0.444
c+0.442
el+0.438
 wing+0.438
Neg logit contributions
 Hood-0.249
ner-0.243
ilda-0.240
 treadmill-0.236
cases-0.233
Token
Token ID198
Feature activation-6.34e-03

Pos logit contributions
 Hood+0.026
ussy+0.025
fighters+0.024
ner+0.024
 treadmill+0.023
Neg logit contributions
 mouth-0.014
 its-0.012
c-0.011
 wing-0.011
el-0.011
Token
Token ID198
Feature activation-6.01e-03

Pos logit contributions
 its+0.005
c+0.005
 mouth+0.005
el+0.004
light+0.004
Neg logit contributions
 Hood-0.014
ussy-0.014
 treadmill-0.014
fighters-0.013
ner-0.013
Tokenâ
Token ID22940
Feature activation+3.14e-02

Pos logit contributions
 its+0.007
light+0.007
 mouth+0.007
c+0.007
 wing+0.007
Neg logit contributions
 Hood-0.011
ussy-0.010
drivers-0.010
 treadmill-0.010
ner-0.010
Token
Token ID26391
Feature activation+2.37e-02

Pos logit contributions
 Hood+0.048
ner+0.043
 treadmill+0.042
ussy+0.042
fighters+0.040
Neg logit contributions
 mouth-0.048
c-0.048
el-0.047
 wing-0.046
light-0.044
Token boy
Token ID2933
Feature activation+5.06e-02

Pos logit contributions
 Hood+0.026
ussy+0.026
ilda+0.025
 treadmill+0.024
oons+0.024
Neg logit contributions
 mouth-0.013
 be-0.012
 color-0.011
c-0.011
 wing-0.010
Token said
Token ID531
Feature activation-1.02e-01

Pos logit contributions
 Hood+0.080
ner+0.069
ussy+0.068
fighters+0.068
 treadmill+0.068
Neg logit contributions
rys-0.073
 wing-0.071
el-0.070
light-0.070
ric-0.069
Token in
Token ID287
Feature activation-3.76e-02

Pos logit contributions
 mouth+0.141
c+0.135
 wing+0.133
el+0.132
 its+0.126
Neg logit contributions
 Hood-0.176
ner-0.159
 treadmill-0.158
ussy-0.158
fighters-0.152
Token a
Token ID257
Feature activation+4.33e-02

Pos logit contributions
c+0.069
light+0.068
 mouth+0.067
ric+0.067
 wing+0.066
Neg logit contributions
 Hood-0.041
assy-0.035
ner-0.035
 treadmill-0.034
ilda-0.031
Token rude
Token ID22066
Feature activation-2.75e-02

Pos logit contributions
 treadmill+0.043
 Hood+0.035
 firefighter+0.032
 cop+0.031
 manager+0.031
Neg logit contributions
ectar-0.089
light-0.088
el-0.086
ric-0.084
cr-0.081
Token voice
Token ID3809
Feature activation-1.92e-01

Pos logit contributions
c+0.029
light+0.028
el+0.027
 mouth+0.026
 wing+0.024
Neg logit contributions
 Hood-0.053
 treadmill-0.050
ner-0.048
ussy-0.047
ilda-0.047
Token.
Token ID13
Feature activation+1.58e-03

Pos logit contributions
 mouth+0.296
c+0.287
el+0.282
 wing+0.281
 its+0.267
Neg logit contributions
 Hood-0.297
ner-0.267
 treadmill-0.262
ussy-0.261
fighters-0.252
Token
Token ID220
Feature activation-9.19e-02

Pos logit contributions
 Hood+0.003
ner+0.003
ussy+0.003
 treadmill+0.003
fighters+0.003
Neg logit contributions
 mouth-0.002
c-0.002
el-0.002
 wing-0.002
 its-0.002
Token
Token ID198
Feature activation-1.22e-02

Pos logit contributions
 mouth+0.101
c+0.098
el+0.094
 its+0.092
 wing+0.089
Neg logit contributions
 Hood-0.182
 treadmill-0.169
ner-0.167
ussy-0.164
fighters-0.160
Token
Token ID198
Feature activation-1.24e-02

Pos logit contributions
 mouth+0.017
 wing+0.017
c+0.017
el+0.016
 feathers+0.016
Neg logit contributions
 Hood-0.020
ner-0.018
 treadmill-0.018
ussy-0.018
fighters-0.017
TokenL
Token ID43
Feature activation+1.06e-02

Pos logit contributions
 mouth+0.019
c+0.018
el+0.018
 wing+0.018
 its+0.017
Neg logit contributions
 Hood-0.020
ner-0.018
 treadmill-0.018
ussy-0.017
fighters-0.017