TOP ACTIVATIONS
MAX = 0.220

cheese and ol ives and ham . They see mom 's
. He likes cheese and ham and lettuce too . He
put the cheese and the ham on the bread . Mom
because I like cheese and ham and yellow and red .
the ham . I like ham ." Ben said .
put the cheese and the ham on the bread ?" Mom
. It had cheese and ham and lettuce . She did
I want to put the ham . I like ham ."
cheeks . That is a ham ster . A rabbit has
. It had cheese and ham and bread . He also
Do you like cheese and ham ?" L ila
and the cheese and the ham . They put them on
. I love cheese and ham . Thank you , boy
do not like cheese and ham . They are too salty
" I want cheese and ham and pineapple ," Tom said
. It is cheese and ham . Do you like cheese
tasty sandwich with cheese and ham . She feels hungry and
ol ives and Ben likes ham . Mom has
. They were cheese and ham and lettuce . They were
ham ster cage . Their ham ster , Fl uffy ,

MIN ACTIVATIONS
MIN = -0.250

to put stones and sand bags around the lake . But
he found a rabbit qu arre lling with a squirrel .
the other kn obs qu arre led over who should be
but today they were qu arre ling . Jenna wanted to
friends didn 't stop qu arre ling and were getting louder
Please , no more qu arre ling ! Let 's find
and saw two birds qu arre ling ! They were squ
both the friends stopped qu arre ling . They decided to
little birds that were qu arre ling . The
and their children , qu arre ling .
told them to stop qu arre lling and to share the
other . They starting qu arre ling again and their voices
two squirrel s kept qu arre ling until finally Sammy said
â t qu arre ling , the furniture stayed
peaceful everyone just stopped qu arre ling .
and they both stopped qu arre lling to look at it
One day they were qu arre ling over who should have
the birds were only qu arre ling like little children do
but today they were qu arre ling . They were arguing
, why are you qu arre ling ?" But the birds

INTERVAL 0.102 - 0.220
CONTAINS 0.352%

He sp its out his sandwich . He says o uch
pirate captain . The par rot would act his part too
friends found him on the branch . They tried to wake
with her friends . The sun was bright and warm ,
again ." The end . <|endoftext|> Tom and Mia

INTERVAL -0.015 - 0.102
CONTAINS 66.767%

did it with love ." The pair hugged and
ig zag was so happy ! It ran around the room
One day , she found a box full of screws and
and started to sing . As she sang her song ,
Tim began to jog . He could feel the wind blowing

INTERVAL -0.133 - -0.015
CONTAINS 32.820%

But you have to follow my rules . You have to
little girl was not so scared any more . The old
Lily should not have been naughty . She should not have
to be safe and happy ." They waited for
tell you that there is nothing to be scared of ".

INTERVAL -0.250 - -0.133
CONTAINS 0.060%

was playing near the f auc et with her friends .
it had made a huge mess . She tried to clean
had made such a huge mess and made him clean it
better . J os ie took the medicine and
store was full of all kinds of snacks , some were
Token cheese
Token ID9891
Feature activation+1.51e-01

Pos logit contributions
+0.023
sed+0.022
 gust+0.022
 stem+0.021
 cheeks+0.020
Neg logit contributions
and-0.019
as-0.019
ying-0.018
pt-0.018
ful-0.018
Token and
Token ID290
Feature activation+1.76e-02

Pos logit contributions
 gust+0.294
sed+0.283
+0.276
 stem+0.270
 cheeks+0.266
Neg logit contributions
and-0.237
as-0.230
ying-0.230
os-0.210
adows-0.206
Token ol
Token ID25776
Feature activation+3.94e-02

Pos logit contributions
+0.031
 gust+0.027
sed+0.027
 sidelines+0.021
 Luckily+0.021
Neg logit contributions
and-0.029
as-0.029
 if-0.028
pt-0.027
 sorts-0.027
Tokenives
Token ID1083
Feature activation-1.02e-01

Pos logit contributions
 gust+0.114
+0.112
sed+0.111
 stem+0.111
 mouth+0.111
Neg logit contributions
and-0.025
ying-0.020
as-0.020
adows-0.020
inker-0.017
Token and
Token ID290
Feature activation+1.76e-02

Pos logit contributions
and+0.162
as+0.157
ying+0.156
pt+0.145
os+0.144
Neg logit contributions
 gust-0.196
sed-0.194
-0.188
 stem-0.186
 mouth-0.180
Token ham
Token ID8891
Feature activation+2.20e-01

Pos logit contributions
+0.028
 gust+0.026
sed+0.025
 stem+0.020
 sidelines+0.020
Neg logit contributions
and-0.032
as-0.032
pt-0.030
 sorts-0.030
smart-0.030
Token.
Token ID13
Feature activation+5.35e-02

Pos logit contributions
 gust+0.445
+0.408
sed+0.401
 stem+0.390
 cheeks+0.380
Neg logit contributions
as-0.351
and-0.350
ying-0.333
adows-0.310
os-0.308
Token They
Token ID1119
Feature activation+5.08e-02

Pos logit contributions
 gust+0.096
sed+0.093
 stem+0.090
 cheeks+0.086
+0.084
Neg logit contributions
and-0.098
as-0.088
ying-0.083
os-0.082
udd-0.081
Token see
Token ID766
Feature activation+1.31e-02

Pos logit contributions
 gust+0.092
sed+0.090
+0.088
 mouth+0.085
 stem+0.085
Neg logit contributions
and-0.084
as-0.082
ying-0.080
adows-0.078
pt-0.077
Token mom
Token ID1995
Feature activation+4.13e-02

Pos logit contributions
+0.025
sed+0.024
 gust+0.024
 stem+0.023
 cheeks+0.022
Neg logit contributions
and-0.021
as-0.020
ying-0.019
pt-0.019
os-0.019
Token's
Token ID338
Feature activation+5.01e-02

Pos logit contributions
 gust+0.086
sed+0.084
+0.081
 stem+0.078
 cheeks+0.075
Neg logit contributions
and-0.064
as-0.062
ying-0.061
os-0.058
adows-0.056
Token.
Token ID13
Feature activation+5.36e-02

Pos logit contributions
 mouth+0.137
 gust+0.134
sed+0.130
 stem+0.128
+0.128
Neg logit contributions
and-0.201
as-0.191
adows-0.189
ying-0.188
inker-0.186
Token He
Token ID679
Feature activation+4.24e-02

Pos logit contributions
 gust+0.082
sed+0.079
 stem+0.077
 cheeks+0.073
+0.072
Neg logit contributions
and-0.114
as-0.103
ying-0.102
adows-0.101
pt-0.096
Token likes
Token ID7832
Feature activation-4.39e-03

Pos logit contributions
 gust+0.077
sed+0.073
+0.073
 stem+0.072
 mouth+0.071
Neg logit contributions
and-0.072
as-0.070
ying-0.069
adows-0.067
udd-0.066
Token cheese
Token ID9891
Feature activation+1.50e-01

Pos logit contributions
and+0.009
as+0.008
ying+0.008
os+0.008
udd+0.008
Neg logit contributions
-0.007
 gust-0.007
sed-0.007
 stem-0.006
 cheeks-0.006
Token and
Token ID290
Feature activation+2.32e-02

Pos logit contributions
 gust+0.250
sed+0.242
 stem+0.237
 mouth+0.237
+0.236
Neg logit contributions
and-0.279
as-0.263
ying-0.263
adows-0.258
os-0.246
Token ham
Token ID8891
Feature activation+2.18e-01

Pos logit contributions
+0.035
 gust+0.034
sed+0.032
 stem+0.028
 sidelines+0.027
Neg logit contributions
and-0.044
as-0.042
 sorts-0.041
pt-0.041
smart-0.040
Token and
Token ID290
Feature activation+2.33e-02

Pos logit contributions
 gust+0.428
+0.397
sed+0.390
 stem+0.383
 mouth+0.369
Neg logit contributions
and-0.362
as-0.352
ying-0.339
adows-0.324
pt-0.314
Token lettuce
Token ID39406
Feature activation+1.18e-01

Pos logit contributions
+0.035
 gust+0.032
sed+0.029
 sidelines+0.025
 stem+0.025
Neg logit contributions
and-0.044
pt-0.043
as-0.043
smart-0.042
ful-0.041
Token too
Token ID1165
Feature activation+6.99e-02

Pos logit contributions
 mouth+0.181
 gust+0.181
sed+0.179
 stem+0.177
+0.172
Neg logit contributions
and-0.232
ying-0.219
adows-0.218
as-0.217
inker-0.211
Token.
Token ID13
Feature activation+5.63e-02

Pos logit contributions
 gust+0.121
sed+0.117
 stem+0.114
+0.114
 mouth+0.114
Neg logit contributions
and-0.128
as-0.121
ying-0.120
adows-0.117
os-0.112
Token He
Token ID679
Feature activation+4.61e-02

Pos logit contributions
 gust+0.084
sed+0.080
 stem+0.078
+0.075
 cheeks+0.074
Neg logit contributions
and-0.122
as-0.112
ying-0.110
adows-0.109
udd-0.105
Token put
Token ID1234
Feature activation-5.67e-03

Pos logit contributions
 gust+0.029
sed+0.028
+0.028
 stem+0.027
 mouth+0.027
Neg logit contributions
and-0.032
as-0.032
pt-0.030
ying-0.030
adows-0.030
Token the
Token ID262
Feature activation+3.62e-02

Pos logit contributions
and+0.007
pt+0.007
os+0.007
as+0.007
ying+0.006
Neg logit contributions
 gust-0.013
sed-0.012
-0.012
 stem-0.012
 cheeks-0.011
Token cheese
Token ID9891
Feature activation+1.49e-01

Pos logit contributions
+0.059
 gust+0.059
 sidelines+0.046
 cheeks+0.043
sed+0.043
Neg logit contributions
as-0.064
and-0.064
ont-0.063
aky-0.062
pt-0.062
Token and
Token ID290
Feature activation+9.60e-03

Pos logit contributions
 gust+0.290
sed+0.282
 stem+0.277
+0.276
 mouth+0.276
Neg logit contributions
and-0.236
as-0.221
ying-0.220
adows-0.215
os-0.202
Token the
Token ID262
Feature activation+3.58e-02

Pos logit contributions
+0.019
 gust+0.018
sed+0.017
 sidelines+0.016
sing+0.015
Neg logit contributions
pt-0.014
 if-0.012
os-0.012
ful-0.012
as-0.011
Token ham
Token ID8891
Feature activation+2.17e-01

Pos logit contributions
+0.056
 gust+0.050
 sidelines+0.038
sed+0.035
 Luckily+0.035
Neg logit contributions
and-0.064
as-0.064
pt-0.063
aky-0.060
 if-0.058
Token on
Token ID319
Feature activation-3.22e-03

Pos logit contributions
 gust+0.448
+0.406
 sidelines+0.386
sed+0.369
Really+0.359
Neg logit contributions
as-0.367
and-0.357
ying-0.326
pt-0.322
adows-0.313
Token the
Token ID262
Feature activation+3.43e-02

Pos logit contributions
and+0.004
ful+0.004
as+0.004
pt+0.003
 sorts+0.003
Neg logit contributions
sed-0.007
 gust-0.007
-0.007
 cheeks-0.006
 sidelines-0.006
Token bread
Token ID8509
Feature activation+1.29e-01

Pos logit contributions
+0.059
 gust+0.055
Really+0.043
sed+0.039
 avail+0.037
Neg logit contributions
and-0.061
ows-0.055
pt-0.055
as-0.055
adows-0.054
Token.
Token ID13
Feature activation+3.79e-02

Pos logit contributions
 gust+0.208
 mouth+0.207
sed+0.205
 stem+0.205
+0.199
Neg logit contributions
and-0.243
ying-0.231
as-0.229
adows-0.226
inker-0.218
Token Mom
Token ID11254
Feature activation+5.37e-02

Pos logit contributions
 gust+0.063
sed+0.061
 stem+0.060
+0.056
 cheeks+0.056
Neg logit contributions
and-0.071
as-0.065
adows-0.062
udd-0.061
ying-0.060
Token because
Token ID780
Feature activation+1.11e-03

Pos logit contributions
 mouth+0.034
 stem+0.033
 gust+0.032
sed+0.031
 cheeks+0.030
Neg logit contributions
and-0.052
ying-0.052
inker-0.052
iam-0.051
asc-0.051
Token I
Token ID314
Feature activation+3.12e-02

Pos logit contributions
 gust+0.001
sed+0.001
 stem+0.001
+0.001
 mouth+0.001
Neg logit contributions
and-0.002
as-0.002
ying-0.002
adows-0.002
os-0.002
Token like
Token ID588
Feature activation+1.11e-02

Pos logit contributions
+0.045
sed+0.045
 mouth+0.044
 stem+0.042
 tongue+0.041
Neg logit contributions
and-0.066
ying-0.064
adows-0.058
inker-0.057
as-0.056
Token cheese
Token ID9891
Feature activation+1.49e-01

Pos logit contributions
 gust+0.019
+0.019
sed+0.018
 stem+0.018
 cheeks+0.017
Neg logit contributions
and-0.020
as-0.018
ying-0.018
pt-0.017
os-0.017
Token and
Token ID290
Feature activation+1.45e-02

Pos logit contributions
 gust+0.226
 mouth+0.226
 stem+0.221
sed+0.219
+0.215
Neg logit contributions
and-0.290
adows-0.274
ying-0.273
as-0.272
inker-0.269
Token ham
Token ID8891
Feature activation+2.17e-01

Pos logit contributions
+0.021
 gust+0.021
sed+0.019
 stem+0.016
 timer+0.015
Neg logit contributions
and-0.028
as-0.026
pt-0.026
smart-0.025
ful-0.025
Token and
Token ID290
Feature activation+1.50e-02

Pos logit contributions
 gust+0.396
sed+0.380
+0.374
 stem+0.374
 mouth+0.371
Neg logit contributions
and-0.374
as-0.352
ying-0.350
adows-0.341
os-0.324
Token yellow
Token ID7872
Feature activation+5.55e-02

Pos logit contributions
+0.021
 gust+0.020
sed+0.018
 stem+0.016
 timer+0.015
Neg logit contributions
and-0.030
pt-0.028
as-0.028
smart-0.027
ful-0.027
Token and
Token ID290
Feature activation+1.42e-02

Pos logit contributions
 gust+0.095
+0.091
sed+0.089
 stem+0.083
 mouth+0.079
Neg logit contributions
and-0.101
as-0.095
pt-0.094
adows-0.092
ying-0.090
Token red
Token ID2266
Feature activation+5.28e-02

Pos logit contributions
+0.020
 gust+0.019
sed+0.017
 sidelines+0.014
 stem+0.014
Neg logit contributions
and-0.028
as-0.027
smart-0.027
pt-0.026
ful-0.026
Token.
Token ID13
Feature activation+4.73e-02

Pos logit contributions
 gust+0.097
+0.094
sed+0.091
 stem+0.086
 mouth+0.081
Neg logit contributions
and-0.089
pt-0.083
as-0.082
adows-0.079
ying-0.078
Token the
Token ID262
Feature activation+3.19e-02

Pos logit contributions
and+0.009
as+0.008
ying+0.008
pt+0.008
os+0.008
Neg logit contributions
 gust-0.011
sed-0.011
-0.010
 stem-0.010
 cheeks-0.010
Token ham
Token ID8891
Feature activation+2.13e-01

Pos logit contributions
+0.052
 gust+0.050
sed+0.044
 cheeks+0.041
 sidelines+0.040
Neg logit contributions
and-0.056
as-0.056
pt-0.054
reat-0.053
ont-0.052
Token.
Token ID13
Feature activation+3.83e-02

Pos logit contributions
 gust+0.386
+0.359
sed+0.354
 stem+0.337
 mouth+0.331
Neg logit contributions
and-0.390
as-0.377
ying-0.363
adows-0.348
pt-0.342
Token I
Token ID314
Feature activation+3.09e-02

Pos logit contributions
 gust+0.059
sed+0.053
+0.052
 stem+0.051
 cheeks+0.047
Neg logit contributions
and-0.080
as-0.073
adows-0.072
ying-0.069
pt-0.069
Token like
Token ID588
Feature activation+1.54e-02

Pos logit contributions
+0.052
 mouth+0.052
 gust+0.051
sed+0.051
 stem+0.051
Neg logit contributions
and-0.057
ying-0.055
inker-0.051
adows-0.050
as-0.049
Token ham
Token ID8891
Feature activation+2.17e-01

Pos logit contributions
+0.029
 gust+0.028
sed+0.027
 stem+0.025
 cheeks+0.025
Neg logit contributions
and-0.027
as-0.024
pt-0.024
ying-0.023
os-0.023
Token."
Token ID526
Feature activation-8.07e-03

Pos logit contributions
 gust+0.440
+0.408
sed+0.401
 stem+0.391
 mouth+0.379
Neg logit contributions
and-0.349
as-0.339
ying-0.324
adows-0.312
pt-0.306
Token Ben
Token ID3932
Feature activation+1.38e-02

Pos logit contributions
os+0.011
and+0.011
pt+0.011
 if+0.010
old+0.010
Neg logit contributions
 stem-0.015
sed-0.015
-0.015
 gust-0.014
 timer-0.014
Token said
Token ID531
Feature activation+6.58e-02

Pos logit contributions
 gust+0.023
sed+0.022
+0.021
 stem+0.021
 mouth+0.021
Neg logit contributions
and-0.026
as-0.025
ying-0.024
adows-0.024
os-0.023
Token.
Token ID13
Feature activation+3.82e-02

Pos logit contributions
 gust+0.125
sed+0.121
 stem+0.120
+0.119
 mouth+0.114
Neg logit contributions
and-0.113
ying-0.104
as-0.102
adows-0.098
pt-0.097
Token
Token ID198
Feature activation+1.80e-02

Pos logit contributions
sed+0.072
 stem+0.071
 gust+0.071
 cheeks+0.069
+0.066
Neg logit contributions
and-0.066
as-0.059
smart-0.056
os-0.053
ying-0.052
Token put
Token ID1234
Feature activation-6.65e-03

Pos logit contributions
sed+0.034
 gust+0.033
+0.029
 cheeks+0.029
Really+0.029
Neg logit contributions
and-0.026
pt-0.026
as-0.025
os-0.025
reat-0.025
Token the
Token ID262
Feature activation+3.73e-02

Pos logit contributions
and+0.011
ying+0.010
as+0.010
pt+0.010
os+0.009
Neg logit contributions
 gust-0.013
sed-0.012
-0.012
 cheeks-0.011
 stem-0.011
Token cheese
Token ID9891
Feature activation+1.50e-01

Pos logit contributions
+0.059
 gust+0.057
 cheeks+0.049
sed+0.048
 sidelines+0.046
Neg logit contributions
as-0.069
and-0.066
reat-0.062
ont-0.060
aking-0.060
Token and
Token ID290
Feature activation+1.70e-02

Pos logit contributions
 gust+0.265
 mouth+0.259
 stem+0.257
sed+0.254
+0.253
Neg logit contributions
and-0.263
adows-0.249
as-0.246
ying-0.244
inker-0.234
Token the
Token ID262
Feature activation+3.71e-02

Pos logit contributions
+0.032
sed+0.031
 gust+0.031
 cheeks+0.027
 stem+0.026
Neg logit contributions
and-0.024
ful-0.023
 if-0.023
pt-0.022
as-0.022
Token ham
Token ID8891
Feature activation+2.15e-01

Pos logit contributions
+0.057
 gust+0.052
 cheeks+0.044
sed+0.043
 Luckily+0.041
Neg logit contributions
as-0.068
and-0.067
reat-0.061
pt-0.060
adows-0.059
Token on
Token ID319
Feature activation-1.07e-03

Pos logit contributions
 gust+0.400
+0.365
sed+0.364
 cheeks+0.350
 stem+0.344
Neg logit contributions
and-0.371
as-0.366
ying-0.360
pt-0.333
adows-0.328
Token the
Token ID262
Feature activation+3.67e-02

Pos logit contributions
and+0.002
as+0.002
ying+0.002
adows+0.002
os+0.002
Neg logit contributions
 gust-0.002
sed-0.002
-0.002
 stem-0.002
 cheeks-0.002
Token bread
Token ID8509
Feature activation+1.33e-01

Pos logit contributions
+0.060
 gust+0.057
Really+0.048
sed+0.047
 cheeks+0.042
Neg logit contributions
and-0.067
as-0.060
adows-0.058
washer-0.057
udd-0.056
Token?"
Token ID1701
Feature activation+8.83e-03

Pos logit contributions
 mouth+0.254
 stem+0.246
 gust+0.243
sed+0.241
+0.236
Neg logit contributions
and-0.215
adows-0.199
as-0.199
ying-0.199
inker-0.193
Token Mom
Token ID11254
Feature activation+5.79e-02

Pos logit contributions
sed+0.017
+0.016
 stem+0.016
 cheeks+0.015
 timer+0.014
Neg logit contributions
 if-0.015
and-0.014
pt-0.013
iam-0.013
 sorts-0.013
Token.
Token ID13
Feature activation+3.43e-02

Pos logit contributions
 gust+0.229
sed+0.220
+0.217
 stem+0.214
 mouth+0.209
Neg logit contributions
and-0.211
ying-0.200
as-0.199
adows-0.190
pt-0.185
Token It
Token ID632
Feature activation+6.24e-02

Pos logit contributions
 gust+0.050
sed+0.049
 stem+0.048
 cheeks+0.046
 mouth+0.045
Neg logit contributions
and-0.071
as-0.064
adows-0.064
ying-0.064
os-0.062
Token had
Token ID550
Feature activation+5.77e-02

Pos logit contributions
 gust+0.124
+0.121
sed+0.116
 stem+0.116
 cheeks+0.113
Neg logit contributions
pt-0.087
and-0.086
as-0.085
os-0.083
 sorts-0.081
Token cheese
Token ID9891
Feature activation+1.43e-01

Pos logit contributions
 gust+0.097
+0.096
sed+0.092
 stem+0.090
 mouth+0.088
Neg logit contributions
and-0.103
as-0.098
pt-0.095
adows-0.093
ying-0.093
Token and
Token ID290
Feature activation+4.19e-03

Pos logit contributions
 gust+0.259
sed+0.249
+0.245
 stem+0.244
 mouth+0.241
Neg logit contributions
and-0.248
as-0.235
ying-0.233
adows-0.226
pt-0.216
Token ham
Token ID8891
Feature activation+2.14e-01

Pos logit contributions
+0.007
 gust+0.006
 sidelines+0.005
sed+0.005
 timer+0.005
Neg logit contributions
pt-0.007
ful-0.007
as-0.007
and-0.007
 sorts-0.007
Token and
Token ID290
Feature activation+5.24e-03

Pos logit contributions
 gust+0.450
+0.411
 sidelines+0.386
sed+0.385
 stem+0.384
Neg logit contributions
as-0.331
and-0.322
ying-0.304
pt-0.291
adows-0.287
Token lettuce
Token ID39406
Feature activation+1.12e-01

Pos logit contributions
+0.008
 gust+0.007
 sidelines+0.006
sed+0.006
 timer+0.005
Neg logit contributions
pt-0.010
ful-0.009
as-0.009
 if-0.009
and-0.009
Token.
Token ID13
Feature activation+3.78e-02

Pos logit contributions
 gust+0.185
sed+0.181
 mouth+0.179
 stem+0.178
+0.176
Neg logit contributions
and-0.209
as-0.196
ying-0.196
adows-0.194
inker-0.184
Token She
Token ID1375
Feature activation+2.39e-02

Pos logit contributions
 gust+0.053
sed+0.053
 stem+0.051
 mouth+0.049
+0.048
Neg logit contributions
and-0.082
as-0.075
ying-0.075
adows-0.073
udd-0.071
Token did
Token ID750
Feature activation+3.01e-02

Pos logit contributions
 gust+0.048
+0.048
 stem+0.045
sed+0.044
 sidelines+0.042
Neg logit contributions
as-0.034
pt-0.033
and-0.033
udd-0.032
os-0.030
Token I
Token ID314
Feature activation+3.18e-02

Pos logit contributions
 gust+0.065
 stem+0.055
sed+0.054
 timer+0.054
+0.053
Neg logit contributions
and-0.072
as-0.067
os-0.067
adows-0.066
ying-0.063
Token want
Token ID765
Feature activation-2.40e-02

Pos logit contributions
+0.051
 mouth+0.050
sed+0.049
 gust+0.048
 stem+0.048
Neg logit contributions
and-0.063
ying-0.062
inker-0.056
as-0.054
adows-0.053
Token to
Token ID284
Feature activation+1.75e-02

Pos logit contributions
and+0.033
pt+0.028
as+0.027
ful+0.026
os+0.026
Neg logit contributions
 gust-0.054
-0.053
sed-0.052
 stem-0.050
 cheeks-0.048
Token put
Token ID1234
Feature activation-5.69e-03

Pos logit contributions
 gust+0.031
sed+0.030
+0.027
Really+0.026
 mouth+0.026
Neg logit contributions
and-0.030
pt-0.030
adows-0.029
as-0.029
reat-0.027
Token the
Token ID262
Feature activation+3.19e-02

Pos logit contributions
and+0.009
as+0.008
ying+0.008
pt+0.008
os+0.008
Neg logit contributions
 gust-0.011
sed-0.011
-0.010
 stem-0.010
 cheeks-0.010
Token ham
Token ID8891
Feature activation+2.13e-01

Pos logit contributions
+0.052
 gust+0.050
sed+0.044
 cheeks+0.041
 sidelines+0.040
Neg logit contributions
and-0.056
as-0.056
pt-0.054
reat-0.053
ont-0.052
Token.
Token ID13
Feature activation+3.83e-02

Pos logit contributions
 gust+0.386
+0.359
sed+0.354
 stem+0.337
 mouth+0.331
Neg logit contributions
and-0.390
as-0.377
ying-0.363
adows-0.348
pt-0.342
Token I
Token ID314
Feature activation+3.09e-02

Pos logit contributions
 gust+0.059
sed+0.053
+0.052
 stem+0.051
 cheeks+0.047
Neg logit contributions
and-0.080
as-0.073
adows-0.072
ying-0.069
pt-0.069
Token like
Token ID588
Feature activation+1.54e-02

Pos logit contributions
+0.052
 mouth+0.052
 gust+0.051
sed+0.051
 stem+0.051
Neg logit contributions
and-0.057
ying-0.055
inker-0.051
adows-0.050
as-0.049
Token ham
Token ID8891
Feature activation+2.17e-01

Pos logit contributions
+0.029
 gust+0.028
sed+0.027
 stem+0.025
 cheeks+0.025
Neg logit contributions
and-0.027
as-0.024
pt-0.024
ying-0.023
os-0.023
Token."
Token ID526
Feature activation-8.07e-03

Pos logit contributions
 gust+0.440
+0.408
sed+0.401
 stem+0.391
 mouth+0.379
Neg logit contributions
and-0.349
as-0.339
ying-0.324
adows-0.312
pt-0.306
Token cheeks
Token ID25839
Feature activation+1.29e-01

Pos logit contributions
pt+0.025
as+0.024
 outage+0.023
adows+0.023
 sorts+0.023
Neg logit contributions
-0.016
 gust-0.012
Really-0.011
sed-0.011
 timer-0.010
Token.
Token ID13
Feature activation+4.21e-02

Pos logit contributions
 gust+0.190
 mouth+0.186
 stem+0.184
+0.177
sed+0.175
Neg logit contributions
and-0.270
adows-0.253
ying-0.247
as-0.244
iam-0.239
Token That
Token ID1320
Feature activation+2.34e-02

Pos logit contributions
 gust+0.054
sed+0.049
+0.049
 stem+0.048
 mouth+0.045
Neg logit contributions
and-0.095
as-0.090
ying-0.088
adows-0.088
os-0.086
Token is
Token ID318
Feature activation-2.21e-02

Pos logit contributions
 gust+0.045
+0.045
sed+0.043
 stem+0.042
 cheeks+0.041
Neg logit contributions
and-0.035
as-0.034
ying-0.033
os-0.033
 sorts-0.033
Token a
Token ID257
Feature activation+4.07e-02

Pos logit contributions
and+0.034
as+0.034
ying+0.032
adows+0.030
iam+0.030
Neg logit contributions
-0.044
 gust-0.041
 cheeks-0.039
 stem-0.039
sed-0.038
Token ham
Token ID8891
Feature activation+2.12e-01

Pos logit contributions
+0.072
 gust+0.066
 cheeks+0.066
sed+0.060
 mouths+0.059
Neg logit contributions
as-0.059
and-0.059
 outage-0.057
pt-0.056
ful-0.055
Tokenster
Token ID1706
Feature activation+5.87e-02

Pos logit contributions
 gust+0.375
+0.336
 cheeks+0.313
sed+0.312
 stem+0.307
Neg logit contributions
as-0.395
and-0.381
ying-0.375
adows-0.351
ful-0.346
Token.
Token ID13
Feature activation+4.18e-02

Pos logit contributions
 mouth+0.095
 gust+0.095
sed+0.094
 stem+0.094
+0.090
Neg logit contributions
and-0.110
ying-0.103
as-0.102
adows-0.102
inker-0.099
Token A
Token ID317
Feature activation+6.55e-02

Pos logit contributions
 gust+0.062
sed+0.056
+0.054
 stem+0.054
 cheeks+0.051
Neg logit contributions
and-0.088
as-0.081
os-0.080
adows-0.079
ying-0.079
Token rabbit
Token ID22746
Feature activation+5.59e-02

Pos logit contributions
+0.110
 gust+0.094
 cheeks+0.089
sed+0.087
 sidelines+0.085
Neg logit contributions
as-0.114
and-0.109
 outage-0.106
pt-0.104
ont-0.104
Token has
Token ID468
Feature activation+2.49e-02

Pos logit contributions
sed+0.090
 gust+0.089
 mouth+0.089
 stem+0.087
+0.084
Neg logit contributions
and-0.107
ying-0.100
adows-0.099
as-0.099
inker-0.095
Token.
Token ID13
Feature activation+3.19e-02

Pos logit contributions
 gust+0.072
sed+0.069
 stem+0.068
 mouth+0.068
+0.067
Neg logit contributions
and-0.080
as-0.076
ying-0.075
adows-0.074
os-0.070
Token It
Token ID632
Feature activation+6.19e-02

Pos logit contributions
 gust+0.047
sed+0.045
 stem+0.044
 cheeks+0.042
+0.041
Neg logit contributions
and-0.069
as-0.062
ying-0.061
adows-0.061
os-0.059
Token had
Token ID550
Feature activation+5.59e-02

Pos logit contributions
 gust+0.127
+0.122
sed+0.116
 stem+0.115
 sidelines+0.113
Neg logit contributions
and-0.085
as-0.083
pt-0.083
os-0.082
 sorts-0.079
Token cheese
Token ID9891
Feature activation+1.43e-01

Pos logit contributions
 gust+0.092
+0.091
sed+0.086
 stem+0.084
 mouth+0.081
Neg logit contributions
and-0.104
as-0.099
adows-0.095
pt-0.094
ying-0.094
Token and
Token ID290
Feature activation+6.29e-03

Pos logit contributions
 gust+0.250
sed+0.241
 stem+0.237
+0.236
 mouth+0.236
Neg logit contributions
and-0.256
as-0.241
ying-0.241
adows-0.235
os-0.224
Token ham
Token ID8891
Feature activation+2.12e-01

Pos logit contributions
+0.010
 gust+0.009
sed+0.008
 sidelines+0.008
 timer+0.007
Neg logit contributions
and-0.011
ful-0.011
as-0.011
pt-0.011
 sorts-0.011
Token and
Token ID290
Feature activation+6.98e-03

Pos logit contributions
 gust+0.421
+0.388
 stem+0.374
sed+0.373
 cheeks+0.358
Neg logit contributions
and-0.341
as-0.337
ying-0.319
adows-0.308
pt-0.302
Token bread
Token ID8509
Feature activation+1.28e-01

Pos logit contributions
+0.010
 gust+0.009
sed+0.008
 sidelines+0.008
 timer+0.007
Neg logit contributions
pt-0.013
and-0.013
ful-0.013
as-0.013
smart-0.012
Token.
Token ID13
Feature activation+3.71e-02

Pos logit contributions
 gust+0.225
sed+0.216
+0.212
 stem+0.211
 mouth+0.209
Neg logit contributions
and-0.227
as-0.215
ying-0.213
adows-0.208
pt-0.199
Token He
Token ID679
Feature activation+3.28e-02

Pos logit contributions
 gust+0.051
sed+0.049
 stem+0.048
+0.045
 cheeks+0.045
Neg logit contributions
and-0.083
as-0.077
ying-0.076
adows-0.075
os-0.073
Token also
Token ID635
Feature activation+2.95e-02

Pos logit contributions
 gust+0.062
+0.060
 stem+0.057
sed+0.057
 mouth+0.054
Neg logit contributions
and-0.052
as-0.052
pt-0.049
udd-0.048
os-0.048
Token Do
Token ID2141
Feature activation+4.72e-02

Pos logit contributions
 gust+0.062
sed+0.056
+0.054
 stem+0.054
 cheeks+0.051
Neg logit contributions
and-0.088
as-0.081
adows-0.080
ying-0.077
os-0.077
Token you
Token ID345
Feature activation+3.07e-02

Pos logit contributions
 gust+0.087
+0.086
sed+0.083
 stem+0.083
 timer+0.079
Neg logit contributions
and-0.083
as-0.079
adows-0.074
ying-0.073
os-0.070
Token like
Token ID588
Feature activation+9.65e-03

Pos logit contributions
 gust+0.067
sed+0.064
 stem+0.064
 mouth+0.064
+0.063
Neg logit contributions
and-0.042
ying-0.039
as-0.038
adows-0.037
os-0.035
Token cheese
Token ID9891
Feature activation+1.45e-01

Pos logit contributions
+0.018
 gust+0.017
sed+0.017
 cheeks+0.016
 stem+0.015
Neg logit contributions
and-0.017
as-0.015
os-0.015
smart-0.015
ful-0.015
Token and
Token ID290
Feature activation+9.73e-03

Pos logit contributions
 gust+0.291
 mouth+0.283
sed+0.282
 stem+0.281
+0.279
Neg logit contributions
and-0.221
as-0.204
ying-0.202
adows-0.201
os-0.187
Token ham
Token ID8891
Feature activation+2.11e-01

Pos logit contributions
+0.016
 gust+0.014
sed+0.013
 sidelines+0.011
 cheeks+0.011
Neg logit contributions
and-0.018
 if-0.017
ful-0.017
as-0.017
smart-0.017
Token?"
Token ID1701
Feature activation-1.54e-02

Pos logit contributions
 gust+0.426
sed+0.403
+0.398
 stem+0.395
 mouth+0.389
Neg logit contributions
and-0.330
as-0.313
ying-0.310
adows-0.297
os-0.283
Token
Token ID198
Feature activation+1.36e-02

Pos logit contributions
and+0.031
as+0.029
ying+0.028
adows+0.028
os+0.028
Neg logit contributions
 gust-0.024
sed-0.024
 stem-0.023
-0.022
 mouth-0.022
Token
Token ID198
Feature activation+1.31e-02

Pos logit contributions
 gust+0.018
sed+0.017
 mouth+0.017
 stem+0.017
+0.017
Neg logit contributions
and-0.030
ying-0.029
as-0.028
adows-0.028
os-0.027
TokenL
Token ID43
Feature activation-2.18e-02

Pos logit contributions
 gust+0.018
 mouth+0.018
 stem+0.017
+0.017
sed+0.017
Neg logit contributions
ying-0.027
and-0.027
adows-0.026
as-0.025
 outage-0.025
Tokenila
Token ID10102
Feature activation-7.41e-02

Pos logit contributions
and+0.039
adows+0.036
 outage+0.035
pt+0.035
 sorts+0.035
Neg logit contributions
 gust-0.034
sed-0.033
 stem-0.033
 mouth-0.032
-0.032
Token and
Token ID290
Feature activation+1.09e-02

Pos logit contributions
 gust+0.250
sed+0.241
+0.236
 stem+0.233
 mouth+0.232
Neg logit contributions
and-0.214
as-0.201
ying-0.200
adows-0.193
os-0.185
Token the
Token ID262
Feature activation+3.63e-02

Pos logit contributions
+0.020
sed+0.019
 gust+0.019
 sidelines+0.017
 stem+0.017
Neg logit contributions
and-0.016
as-0.016
pt-0.015
os-0.015
ful-0.015
Token cheese
Token ID9891
Feature activation+1.47e-01

Pos logit contributions
+0.059
 gust+0.053
sed+0.044
 Luckily+0.042
Really+0.041
Neg logit contributions
as-0.063
and-0.060
pt-0.056
reat-0.056
udd-0.054
Token and
Token ID290
Feature activation+1.18e-02

Pos logit contributions
 gust+0.284
sed+0.271
+0.265
 stem+0.256
 mouth+0.254
Neg logit contributions
and-0.241
as-0.228
ying-0.226
adows-0.212
os-0.208
Token the
Token ID262
Feature activation+3.62e-02

Pos logit contributions
+0.024
sed+0.022
 gust+0.022
 sidelines+0.020
 stem+0.020
Neg logit contributions
and-0.016
os-0.015
as-0.015
pt-0.015
ful-0.015
Token ham
Token ID8891
Feature activation+2.11e-01

Pos logit contributions
+0.056
 gust+0.049
sed+0.042
 Luckily+0.040
 cheeks+0.038
Neg logit contributions
as-0.065
and-0.063
pt-0.059
reat-0.058
adows-0.058
Token.
Token ID13
Feature activation+4.32e-02

Pos logit contributions
 gust+0.429
+0.386
 sidelines+0.367
sed+0.365
Really+0.356
Neg logit contributions
as-0.351
and-0.338
ying-0.318
ful-0.301
adows-0.298
Token They
Token ID1119
Feature activation+3.70e-02

Pos logit contributions
sed+0.070
 gust+0.070
 stem+0.068
 cheeks+0.065
 timer+0.062
Neg logit contributions
and-0.083
as-0.075
udd-0.072
ying-0.072
adows-0.070
Token put
Token ID1234
Feature activation-7.67e-03

Pos logit contributions
 gust+0.062
sed+0.060
 stem+0.058
+0.058
 mouth+0.058
Neg logit contributions
and-0.069
as-0.065
ying-0.065
adows-0.064
os-0.061
Token them
Token ID606
Feature activation-8.69e-03

Pos logit contributions
and+0.010
as+0.010
ying+0.010
os+0.010
pt+0.010
Neg logit contributions
sed-0.016
 gust-0.016
 cheeks-0.015
-0.015
 stem-0.015
Token on
Token ID319
Feature activation-3.69e-03

Pos logit contributions
and+0.015
ying+0.014
as+0.014
adows+0.014
inker+0.013
Neg logit contributions
 gust-0.016
sed-0.015
 mouth-0.015
 stem-0.015
-0.015
Token.
Token ID13
Feature activation+4.02e-02

Pos logit contributions
 gust+0.112
 mouth+0.111
 stem+0.105
sed+0.105
 cheeks+0.103
Neg logit contributions
and-0.168
adows-0.155
ying-0.153
as-0.153
 outage-0.149
Token I
Token ID314
Feature activation+2.70e-02

Pos logit contributions
 gust+0.068
 stem+0.056
 timer+0.056
sed+0.056
 sidelines+0.053
Neg logit contributions
and-0.076
os-0.071
adows-0.070
ibr-0.068
asc-0.067
Token love
Token ID1842
Feature activation-2.99e-02

Pos logit contributions
+0.050
 mouth+0.049
sed+0.049
 stem+0.049
 gust+0.048
Neg logit contributions
and-0.045
ying-0.044
inker-0.040
adows-0.039
as-0.039
Token cheese
Token ID9891
Feature activation+1.45e-01

Pos logit contributions
and+0.056
os+0.050
ful+0.050
as+0.049
 sorts+0.048
Neg logit contributions
 gust-0.053
-0.051
sed-0.049
 cheeks-0.044
 stem-0.043
Token and
Token ID290
Feature activation+9.16e-03

Pos logit contributions
 mouth+0.214
 stem+0.209
 gust+0.205
sed+0.204
+0.199
Neg logit contributions
and-0.285
ying-0.274
adows-0.274
as-0.268
inker-0.268
Token ham
Token ID8891
Feature activation+2.10e-01

Pos logit contributions
 gust+0.013
+0.013
sed+0.012
 sidelines+0.010
 timer+0.010
Neg logit contributions
and-0.017
 if-0.017
ful-0.017
 sorts-0.016
smart-0.016
Token.
Token ID13
Feature activation+4.24e-02

Pos logit contributions
 gust+0.394
sed+0.369
+0.366
 stem+0.361
 mouth+0.355
Neg logit contributions
and-0.362
as-0.345
ying-0.339
adows-0.329
os-0.315
Token Thank
Token ID6952
Feature activation+1.03e-01

Pos logit contributions
 gust+0.068
sed+0.059
 stem+0.058
+0.055
 cheeks+0.054
Neg logit contributions
and-0.085
adows-0.078
as-0.077
os-0.075
ibr-0.075
Token you
Token ID345
Feature activation+3.36e-02

Pos logit contributions
sed+0.213
 gust+0.212
 stem+0.200
 mouth+0.190
sing+0.190
Neg logit contributions
and-0.171
as-0.151
ful-0.146
smart-0.144
os-0.143
Token,
Token ID11
Feature activation+5.20e-02

Pos logit contributions
 mouth+0.049
 gust+0.047
sed+0.045
 stem+0.044
+0.044
Neg logit contributions
ying-0.068
and-0.067
adows-0.065
as-0.065
 outage-0.063
Token boy
Token ID2933
Feature activation+5.90e-02

Pos logit contributions
 gust+0.073
sed+0.072
+0.072
 stem+0.071
 mouth+0.067
Neg logit contributions
and-0.107
as-0.103
ying-0.101
adows-0.098
os-0.098
Token do
Token ID466
Feature activation+4.32e-02

Pos logit contributions
sed+0.044
 mouth+0.044
 gust+0.043
+0.043
 stem+0.042
Neg logit contributions
and-0.049
ying-0.047
adows-0.045
as-0.043
inker-0.043
Token not
Token ID407
Feature activation-1.79e-03

Pos logit contributions
 gust+0.070
sed+0.066
+0.064
 stem+0.064
 mouth+0.063
Neg logit contributions
and-0.085
as-0.081
ying-0.080
adows-0.078
udd-0.075
Token like
Token ID588
Feature activation+7.61e-03

Pos logit contributions
and+0.003
ying+0.002
as+0.002
adows+0.002
os+0.002
Neg logit contributions
 gust-0.004
sed-0.004
 stem-0.004
 mouth-0.004
-0.004
Token cheese
Token ID9891
Feature activation+1.44e-01

Pos logit contributions
+0.013
sed+0.012
 gust+0.012
 stem+0.011
 cheeks+0.011
Neg logit contributions
and-0.014
ying-0.013
as-0.013
ful-0.012
reat-0.012
Token and
Token ID290
Feature activation+2.31e-03

Pos logit contributions
 gust+0.219
 mouth+0.218
 stem+0.214
sed+0.211
+0.209
Neg logit contributions
and-0.280
adows-0.270
ying-0.263
as-0.263
inker-0.258
Token ham
Token ID8891
Feature activation+2.10e-01

Pos logit contributions
+0.004
 gust+0.003
sed+0.003
 stem+0.003
 timer+0.003
Neg logit contributions
and-0.004
as-0.004
pt-0.004
ful-0.004
ying-0.004
Token.
Token ID13
Feature activation+2.87e-02

Pos logit contributions
 gust+0.378
sed+0.363
+0.357
 stem+0.356
 mouth+0.353
Neg logit contributions
and-0.366
as-0.345
ying-0.343
adows-0.335
os-0.318
Token They
Token ID1119
Feature activation+2.97e-02

Pos logit contributions
 gust+0.044
+0.039
sed+0.038
 stem+0.037
 timer+0.037
Neg logit contributions
and-0.059
as-0.054
os-0.053
ibr-0.051
smart-0.051
Token are
Token ID389
Feature activation-4.20e-02

Pos logit contributions
 gust+0.063
sed+0.061
+0.060
 stem+0.060
 mouth+0.059
Neg logit contributions
and-0.042
as-0.039
ying-0.039
adows-0.038
pt-0.036
Token too
Token ID1165
Feature activation+6.04e-02

Pos logit contributions
and+0.069
as+0.068
ying+0.065
pt+0.063
adows+0.061
Neg logit contributions
-0.078
 gust-0.075
sed-0.070
 cheeks-0.068
 stem-0.068
Token salty
Token ID36021
Feature activation+1.39e-02

Pos logit contributions
 gust+0.089
sed+0.086
+0.085
 stem+0.083
 mouth+0.082
Neg logit contributions
and-0.125
as-0.120
ying-0.118
adows-0.115
udd-0.111
Token"
Token ID1
Feature activation-3.38e-02

Pos logit contributions
 gust+0.024
sed+0.023
 stem+0.023
+0.023
 mouth+0.023
Neg logit contributions
and-0.019
ying-0.018
as-0.017
adows-0.017
pt-0.016
TokenI
Token ID40
Feature activation+3.92e-02

Pos logit contributions
and+0.065
as+0.061
ying+0.057
iam+0.056
os+0.055
Neg logit contributions
 gust-0.060
sed-0.058
 mouth-0.056
 stem-0.055
 cheeks-0.054
Token want
Token ID765
Feature activation-3.28e-02

Pos logit contributions
+0.061
 mouth+0.060
sed+0.060
 gust+0.060
 stem+0.056
Neg logit contributions
and-0.078
ying-0.078
adows-0.068
inker-0.067
as-0.067
Token cheese
Token ID9891
Feature activation+1.44e-01

Pos logit contributions
and+0.068
as+0.064
ying+0.063
adows+0.061
os+0.060
Neg logit contributions
 gust-0.049
sed-0.048
-0.047
 stem-0.046
 mouth-0.044
Token and
Token ID290
Feature activation+5.33e-03

Pos logit contributions
 gust+0.228
sed+0.220
 mouth+0.216
 stem+0.216
+0.215
Neg logit contributions
and-0.282
ying-0.266
as-0.266
adows-0.262
pt-0.249
Token ham
Token ID8891
Feature activation+2.10e-01

Pos logit contributions
 gust+0.008
+0.008
sed+0.007
 stem+0.006
 sidelines+0.006
Neg logit contributions
and-0.010
 sorts-0.009
as-0.009
 if-0.009
os-0.009
Token and
Token ID290
Feature activation+6.56e-03

Pos logit contributions
 gust+0.418
+0.384
sed+0.383
 stem+0.376
 mouth+0.362
Neg logit contributions
and-0.339
as-0.330
ying-0.320
adows-0.304
os-0.297
Token pineapple
Token ID45540
Feature activation+1.30e-01

Pos logit contributions
 gust+0.009
+0.009
sed+0.008
 stem+0.007
 sidelines+0.007
Neg logit contributions
and-0.012
as-0.012
smart-0.012
 if-0.011
ful-0.011
Token,"
Token ID553
Feature activation+6.43e-02

Pos logit contributions
 gust+0.241
sed+0.233
+0.229
 stem+0.227
 mouth+0.224
Neg logit contributions
and-0.221
as-0.210
ying-0.207
adows-0.200
os-0.193
Token Tom
Token ID4186
Feature activation+3.96e-02

Pos logit contributions
+0.113
sed+0.110
 stem+0.107
 gust+0.098
 tongue+0.092
Neg logit contributions
iam-0.105
 if-0.101
as-0.100
and-0.099
 sorts-0.097
Token said
Token ID531
Feature activation+6.25e-02

Pos logit contributions
 gust+0.056
sed+0.054
 mouth+0.053
 stem+0.052
+0.051
Neg logit contributions
and-0.083
ying-0.080
as-0.079
adows-0.079
inker-0.075
Token.
Token ID13
Feature activation+3.83e-02

Pos logit contributions
 mouth+0.166
sed+0.157
 stem+0.155
+0.153
 gust+0.146
Neg logit contributions
and-0.251
adows-0.246
inker-0.246
ying-0.244
 outage-0.236
Token It
Token ID632
Feature activation+6.03e-02

Pos logit contributions
 gust+0.062
sed+0.054
 stem+0.053
+0.051
 timer+0.051
Neg logit contributions
and-0.078
adows-0.069
as-0.069
ibr-0.068
os-0.068
Token is
Token ID318
Feature activation-2.66e-02

Pos logit contributions
 gust+0.116
sed+0.110
+0.108
 stem+0.108
 mouth+0.106
Neg logit contributions
and-0.099
as-0.094
ying-0.093
adows-0.090
os-0.088
Token cheese
Token ID9891
Feature activation+1.42e-01

Pos logit contributions
and+0.041
as+0.040
ying+0.038
adows+0.037
pt+0.037
Neg logit contributions
-0.052
 gust-0.051
sed-0.049
 cheeks-0.047
 stem-0.046
Token and
Token ID290
Feature activation+6.26e-03

Pos logit contributions
 mouth+0.203
 stem+0.197
sed+0.194
 gust+0.190
+0.187
Neg logit contributions
and-0.288
inker-0.276
ying-0.275
adows-0.273
as-0.268
Token ham
Token ID8891
Feature activation+2.09e-01

Pos logit contributions
+0.009
 gust+0.009
sed+0.008
 sidelines+0.007
 cheeks+0.007
Neg logit contributions
and-0.011
ful-0.011
smart-0.011
pt-0.011
as-0.011
Token.
Token ID13
Feature activation+4.11e-02

Pos logit contributions
 gust+0.370
sed+0.356
 stem+0.349
+0.349
 mouth+0.347
Neg logit contributions
and-0.370
as-0.349
ying-0.347
adows-0.339
os-0.322
Token Do
Token ID2141
Feature activation+4.72e-02

Pos logit contributions
 gust+0.062
sed+0.056
+0.054
 stem+0.054
 cheeks+0.051
Neg logit contributions
and-0.088
as-0.081
adows-0.080
ying-0.077
os-0.077
Token you
Token ID345
Feature activation+3.07e-02

Pos logit contributions
 gust+0.087
+0.086
sed+0.083
 stem+0.083
 timer+0.079
Neg logit contributions
and-0.083
as-0.079
adows-0.074
ying-0.073
os-0.070
Token like
Token ID588
Feature activation+9.65e-03

Pos logit contributions
 gust+0.067
sed+0.064
 stem+0.064
 mouth+0.064
+0.063
Neg logit contributions
and-0.042
ying-0.039
as-0.038
adows-0.037
os-0.035
Token cheese
Token ID9891
Feature activation+1.45e-01

Pos logit contributions
+0.018
 gust+0.017
sed+0.017
 cheeks+0.016
 stem+0.015
Neg logit contributions
and-0.017
as-0.015
os-0.015
smart-0.015
ful-0.015
Token tasty
Token ID25103
Feature activation+2.77e-02

Pos logit contributions
+0.059
sed+0.056
 gust+0.054
 cheeks+0.054
 stem+0.050
Neg logit contributions
as-0.076
and-0.073
 outage-0.070
udd-0.069
ying-0.066
Token sandwich
Token ID20433
Feature activation+1.21e-01

Pos logit contributions
+0.042
sed+0.042
Really+0.038
 mouths+0.038
 cheeks+0.037
Neg logit contributions
 sorts-0.045
reat-0.043
aking-0.041
and-0.041
udd-0.041
Token with
Token ID351
Feature activation+7.46e-02

Pos logit contributions
 gust+0.232
sed+0.223
+0.220
 stem+0.219
 mouth+0.217
Neg logit contributions
and-0.198
as-0.186
ying-0.186
adows-0.178
os-0.170
Token cheese
Token ID9891
Feature activation+1.40e-01

Pos logit contributions
+0.139
 gust+0.136
sed+0.130
 stem+0.124
 cheeks+0.121
Neg logit contributions
and-0.115
pt-0.110
as-0.109
ful-0.102
col-0.102
Token and
Token ID290
Feature activation+8.72e-04

Pos logit contributions
 gust+0.260
sed+0.249
+0.246
 stem+0.244
 mouth+0.242
Neg logit contributions
and-0.236
as-0.224
ying-0.222
adows-0.214
pt-0.206
Token ham
Token ID8891
Feature activation+2.09e-01

Pos logit contributions
+0.001
 gust+0.001
sed+0.001
 sidelines+0.001
 stem+0.001
Neg logit contributions
and-0.001
pt-0.001
as-0.001
 if-0.001
 sorts-0.001
Token.
Token ID13
Feature activation+3.59e-02

Pos logit contributions
 gust+0.432
+0.397
sed+0.376
 stem+0.375
 cheeks+0.367
Neg logit contributions
as-0.327
and-0.318
ying-0.302
pt-0.286
adows-0.285
Token She
Token ID1375
Feature activation+2.30e-02

Pos logit contributions
 gust+0.059
 timer+0.058
sed+0.058
 stem+0.057
 cheeks+0.057
Neg logit contributions
and-0.068
as-0.060
ying-0.059
os-0.058
adows-0.058
Token feels
Token ID5300
Feature activation-4.93e-03

Pos logit contributions
 gust+0.046
+0.045
 stem+0.042
sed+0.041
 cheeks+0.040
Neg logit contributions
as-0.033
and-0.032
udd-0.031
pt-0.031
ying-0.030
Token hungry
Token ID14720
Feature activation+3.23e-02

Pos logit contributions
and+0.009
pt+0.008
ying+0.008
as+0.008
adows+0.008
Neg logit contributions
-0.009
sed-0.009
 gust-0.009
 stem-0.008
 cheeks-0.008
Token and
Token ID290
Feature activation+4.61e-03

Pos logit contributions
 gust+0.052
 mouth+0.051
 stem+0.049
sed+0.049
+0.047
Neg logit contributions
and-0.061
as-0.059
ying-0.059
adows-0.058
inker-0.056
Token ol
Token ID25776
Feature activation+3.05e-02

Pos logit contributions
and+0.031
as+0.029
ying+0.029
os+0.028
 sorts+0.027
Neg logit contributions
-0.025
 gust-0.025
sed-0.024
 cheeks-0.022
 stem-0.022
Tokenives
Token ID1083
Feature activation-1.10e-01

Pos logit contributions
+0.082
 gust+0.079
 mouth+0.076
 grape+0.076
 stem+0.076
Neg logit contributions
and-0.023
ying-0.022
inker-0.021
adows-0.021
pt-0.021
Token and
Token ID290
Feature activation-6.22e-03

Pos logit contributions
and+0.214
ying+0.200
as+0.199
adows+0.198
 outage+0.188
Neg logit contributions
 gust-0.177
 mouth-0.170
sed-0.169
-0.168
 stem-0.167
Token Ben
Token ID3932
Feature activation-4.44e-03

Pos logit contributions
as+0.012
and+0.011
 sorts+0.011
ying+0.011
os+0.011
Neg logit contributions
 gust-0.010
-0.009
sed-0.009
 stem-0.008
 mouth-0.008
Token likes
Token ID7832
Feature activation-1.43e-02

Pos logit contributions
and+0.008
ying+0.008
as+0.008
adows+0.007
inker+0.007
Neg logit contributions
 gust-0.008
sed-0.007
 mouth-0.007
 stem-0.007
-0.007
Token ham
Token ID8891
Feature activation+2.08e-01

Pos logit contributions
and+0.027
as+0.025
ying+0.024
os+0.024
adows+0.023
Neg logit contributions
 gust-0.024
sed-0.024
-0.023
 stem-0.022
 cheeks-0.022
Token.
Token ID13
Feature activation+3.94e-03

Pos logit contributions
 gust+0.358
sed+0.346
 stem+0.339
+0.338
 mouth+0.338
Neg logit contributions
and-0.379
as-0.357
ying-0.356
adows-0.349
os-0.332
Token
Token ID198
Feature activation-1.62e-02

Pos logit contributions
 gust+0.007
sed+0.006
 stem+0.006
 timer+0.006
 cheeks+0.006
Neg logit contributions
and-0.007
as-0.007
ying-0.007
udd-0.007
os-0.007
Token
Token ID198
Feature activation-1.77e-02

Pos logit contributions
and+0.036
ying+0.035
as+0.035
adows+0.035
pt+0.033
Neg logit contributions
 gust-0.020
-0.019
sed-0.019
 mouth-0.019
 stem-0.018
TokenMom
Token ID29252
Feature activation+6.97e-02

Pos logit contributions
and+0.033
as+0.031
ying+0.031
adows+0.031
os+0.029
Neg logit contributions
 gust-0.030
sed-0.029
 stem-0.028
 mouth-0.028
-0.028
Token has
Token ID468
Feature activation+1.83e-02

Pos logit contributions
 gust+0.110
sed+0.107
 mouth+0.107
 stem+0.105
+0.105
Neg logit contributions
and-0.138
adows-0.127
as-0.126
ying-0.125
pt-0.119
Token.
Token ID13
Feature activation+3.13e-02

Pos logit contributions
 mouth+0.078
 gust+0.078
 stem+0.075
+0.074
sed+0.074
Neg logit contributions
and-0.107
adows-0.103
as-0.103
ying-0.102
inker-0.101
Token They
Token ID1119
Feature activation+2.82e-02

Pos logit contributions
 gust+0.046
sed+0.045
 stem+0.044
 cheeks+0.041
+0.041
Neg logit contributions
and-0.065
ying-0.060
as-0.060
udd-0.058
adows-0.057
Token were
Token ID547
Feature activation+4.92e-03

Pos logit contributions
 gust+0.041
sed+0.040
 mouth+0.040
 stem+0.039
+0.039
Neg logit contributions
and-0.060
ying-0.057
adows-0.055
as-0.054
inker-0.054
Token cheese
Token ID9891
Feature activation+1.42e-01

Pos logit contributions
+0.009
 gust+0.009
sed+0.009
 stem+0.008
 cheeks+0.008
Neg logit contributions
and-0.008
as-0.008
ying-0.008
adows-0.007
pt-0.007
Token and
Token ID290
Feature activation+2.49e-03

Pos logit contributions
 gust+0.225
sed+0.221
 mouth+0.219
 stem+0.217
+0.215
Neg logit contributions
and-0.274
ying-0.259
as-0.255
adows-0.253
inker-0.243
Token ham
Token ID8891
Feature activation+2.08e-01

Pos logit contributions
+0.004
 gust+0.004
sed+0.003
 sidelines+0.003
 Luckily+0.003
Neg logit contributions
ful-0.004
and-0.004
 sorts-0.004
as-0.004
pt-0.004
Token and
Token ID290
Feature activation+3.52e-03

Pos logit contributions
 gust+0.399
+0.370
sed+0.368
 stem+0.366
 mouth+0.357
Neg logit contributions
and-0.347
as-0.333
ying-0.322
adows-0.317
os-0.301
Token lettuce
Token ID39406
Feature activation+1.08e-01

Pos logit contributions
+0.005
 gust+0.005
sed+0.004
 stem+0.004
 sidelines+0.004
Neg logit contributions
and-0.007
pt-0.006
ful-0.006
as-0.006
smart-0.006
Token.
Token ID13
Feature activation+3.57e-02

Pos logit contributions
sed+0.166
 mouth+0.166
 gust+0.163
 stem+0.162
+0.159
Neg logit contributions
and-0.214
ying-0.203
as-0.199
adows-0.199
inker-0.195
Token They
Token ID1119
Feature activation+3.17e-02

Pos logit contributions
 gust+0.054
sed+0.052
 stem+0.051
 mouth+0.048
 cheeks+0.048
Neg logit contributions
and-0.075
as-0.069
ying-0.069
adows-0.066
udd-0.065
Token were
Token ID547
Feature activation+7.10e-03

Pos logit contributions
 gust+0.050
sed+0.048
 stem+0.047
+0.047
 mouth+0.047
Neg logit contributions
and-0.062
as-0.059
ying-0.058
adows-0.057
os-0.055
Token ham
Token ID8891
Feature activation+2.06e-01

Pos logit contributions
and+0.014
as+0.014
ying+0.013
adows+0.013
pt+0.013
Neg logit contributions
 gust-0.009
-0.008
sed-0.008
 stem-0.008
 mouth-0.008
Tokenster
Token ID1706
Feature activation+5.14e-02

Pos logit contributions
 gust+0.285
+0.256
sed+0.242
 stem+0.233
 mouth+0.220
Neg logit contributions
as-0.454
and-0.444
adows-0.431
ying-0.428
pt-0.413
Token cage
Token ID17186
Feature activation+7.66e-02

Pos logit contributions
 gust+0.086
sed+0.082
+0.081
 stem+0.076
 mouth+0.073
Neg logit contributions
and-0.094
as-0.094
adows-0.091
ying-0.090
pt-0.088
Token.
Token ID13
Feature activation+3.39e-02

Pos logit contributions
 gust+0.118
 mouth+0.118
 stem+0.114
sed+0.113
 cheeks+0.110
Neg logit contributions
and-0.157
ying-0.146
adows-0.140
iam-0.139
as-0.139
Token Their
Token ID5334
Feature activation+3.88e-02

Pos logit contributions
 gust+0.054
sed+0.053
 stem+0.052
 cheeks+0.050
+0.047
Neg logit contributions
and-0.069
as-0.061
ying-0.061
smart-0.059
os-0.059
Token ham
Token ID8891
Feature activation+2.08e-01

Pos logit contributions
+0.051
 gust+0.051
 stem+0.036
 sidelines+0.035
sed+0.035
Neg logit contributions
and-0.086
as-0.085
pt-0.082
adows-0.080
os-0.080
Tokenster
Token ID1706
Feature activation+5.36e-02

Pos logit contributions
 gust+0.336
+0.306
sed+0.279
 stem+0.279
 sidelines+0.271
Neg logit contributions
as-0.413
and-0.401
ying-0.386
adows-0.386
pt-0.378
Token,
Token ID11
Feature activation+4.05e-02

Pos logit contributions
 gust+0.099
+0.095
sed+0.088
 stem+0.086
 sidelines+0.084
Neg logit contributions
as-0.090
and-0.087
pt-0.084
os-0.083
adows-0.083
Token Fl
Token ID1610
Feature activation+3.88e-02

Pos logit contributions
 gust+0.069
+0.068
sed+0.066
 stem+0.065
 mouth+0.062
Neg logit contributions
and-0.073
as-0.069
ying-0.068
pt-0.066
adows-0.065
Tokenuffy
Token ID15352
Feature activation-1.99e-02

Pos logit contributions
 gust+0.084
 sidelines+0.075
 timer+0.072
 stem+0.071
+0.068
Neg logit contributions
as-0.063
udd-0.059
adows-0.059
adow-0.056
os-0.055
Token,
Token ID11
Feature activation+4.12e-02

Pos logit contributions
and+0.035
as+0.035
adows+0.033
ying+0.033
os+0.033
Neg logit contributions
 gust-0.034
-0.032
sed-0.031
 stem-0.031
 mouth-0.030
Token to
Token ID284
Feature activation-1.60e-02

Pos logit contributions
 mouth+0.044
 gust+0.043
sed+0.042
 stem+0.042
+0.041
Neg logit contributions
and-0.052
as-0.050
ying-0.050
adows-0.048
udd-0.047
Token put
Token ID1234
Feature activation-3.29e-02

Pos logit contributions
and+0.038
as+0.036
ying+0.036
udd+0.033
adows+0.033
Neg logit contributions
 stem-0.021
 gust-0.020
 mouth-0.020
-0.019
sed-0.019
Token stones
Token ID14966
Feature activation-2.90e-02

Pos logit contributions
and+0.068
as+0.063
ying+0.061
 outage+0.061
adows+0.061
Neg logit contributions
 gust-0.051
 mouth-0.051
-0.048
 stem-0.048
sed-0.047
Token and
Token ID290
Feature activation-2.96e-02

Pos logit contributions
inker+0.055
and+0.055
asc+0.055
iam+0.054
ibr+0.052
Neg logit contributions
 stem-0.043
 mouth-0.042
 gust-0.041
-0.039
sed-0.039
Token sand
Token ID6450
Feature activation+1.50e-02

Pos logit contributions
and+0.060
ying+0.057
as+0.057
adows+0.055
udd+0.053
Neg logit contributions
 gust-0.045
sed-0.044
 stem-0.043
 mouth-0.043
-0.042
Tokenbags
Token ID34005
Feature activation-2.50e-01

Pos logit contributions
sed+0.024
 mouth+0.024
 stem+0.023
 cheeks+0.023
+0.022
Neg logit contributions
and-0.028
 outage-0.027
 sorts-0.026
iam-0.026
ying-0.026
Token around
Token ID1088
Feature activation-5.49e-02

Pos logit contributions
and+0.425
iam+0.414
asc+0.414
ying+0.406
inker+0.403
Neg logit contributions
 gust-0.432
-0.411
sed-0.409
 mouth-0.408
 cheeks-0.405
Token the
Token ID262
Feature activation-3.94e-04

Pos logit contributions
and+0.112
as+0.107
ying+0.106
adows+0.104
udd+0.102
Neg logit contributions
 gust-0.083
 stem-0.079
-0.078
 mouth-0.078
sed-0.078
Token lake
Token ID13546
Feature activation+3.40e-03

Pos logit contributions
and+0.001
as+0.001
ying+0.001
adows+0.001
os+0.001
Neg logit contributions
 gust-0.001
sed-0.001
 stem-0.001
 mouth-0.001
-0.001
Token.
Token ID13
Feature activation-7.38e-03

Pos logit contributions
 mouth+0.005
 stem+0.005
 gust+0.005
sed+0.005
+0.005
Neg logit contributions
and-0.007
ying-0.007
as-0.007
inker-0.006
asc-0.006
Token But
Token ID887
Feature activation+4.13e-02

Pos logit contributions
and+0.016
as+0.015
adows+0.015
ying+0.015
os+0.015
Neg logit contributions
 gust-0.010
sed-0.010
 stem-0.010
 mouth-0.009
 cheeks-0.009
Token he
Token ID339
Feature activation-2.35e-02

Pos logit contributions
and+0.006
as+0.006
ying+0.006
udd+0.006
 outage+0.006
Neg logit contributions
 gust-0.002
 mouth-0.002
sed-0.002
 stem-0.002
-0.002
Token found
Token ID1043
Feature activation-1.10e-02

Pos logit contributions
and+0.050
ying+0.048
as+0.047
adows+0.046
 outage+0.044
Neg logit contributions
 gust-0.034
sed-0.032
 mouth-0.032
 stem-0.031
-0.031
Token a
Token ID257
Feature activation+7.55e-03

Pos logit contributions
and+0.023
as+0.022
ying+0.022
adows+0.021
udd+0.021
Neg logit contributions
 gust-0.016
 mouth-0.016
sed-0.016
 stem-0.015
-0.015
Token rabbit
Token ID22746
Feature activation+1.79e-02

Pos logit contributions
 cheeks+0.012
+0.012
sed+0.012
 gust+0.012
 mouths+0.011
Neg logit contributions
and-0.013
as-0.012
 outage-0.012
 sorts-0.012
udd-0.011
Token qu
Token ID627
Feature activation-2.35e-02

Pos logit contributions
 gust+0.032
sed+0.031
 stem+0.030
+0.030
 mouth+0.029
Neg logit contributions
and-0.032
as-0.030
ying-0.030
adows-0.029
pt-0.028
Tokenarre
Token ID9624
Feature activation-2.46e-01

Pos logit contributions
and+0.033
ying+0.030
as+0.030
adows+0.029
os+0.027
Neg logit contributions
 gust-0.050
sed-0.048
 stem-0.048
-0.048
 mouth-0.048
Tokenlling
Token ID2680
Feature activation-1.30e-01

Pos logit contributions
and+0.390
as+0.376
adows+0.374
os+0.370
udd+0.365
Neg logit contributions
 gust-0.456
 stem-0.426
 timer-0.399
 cheeks-0.397
 mouth-0.391
Token with
Token ID351
Feature activation+4.77e-02

Pos logit contributions
and+0.234
as+0.220
adows+0.217
ying+0.215
udd+0.208
Neg logit contributions
 gust-0.229
sed-0.222
 mouth-0.220
 stem-0.215
-0.204
Token a
Token ID257
Feature activation+7.41e-03

Pos logit contributions
 gust+0.085
sed+0.085
+0.083
 stem+0.082
 mouth+0.079
Neg logit contributions
and-0.080
as-0.074
pt-0.072
adows-0.071
ying-0.071
Token squirrel
Token ID33039
Feature activation+5.27e-02

Pos logit contributions
sed+0.011
+0.010
 gust+0.010
 cheeks+0.010
 stem+0.010
Neg logit contributions
as-0.013
and-0.013
pt-0.013
 outage-0.013
 sorts-0.013
Token.
Token ID13
Feature activation-3.34e-02

Pos logit contributions
 gust+0.086
 mouth+0.085
sed+0.084
 stem+0.083
+0.082
Neg logit contributions
and-0.098
ying-0.093
as-0.091
adows-0.091
inker-0.087
Token the
Token ID262
Feature activation-1.35e-02

Pos logit contributions
and+0.199
as+0.188
adows+0.184
ying+0.184
inker+0.182
Neg logit contributions
 gust-0.122
 mouth-0.115
 stem-0.112
sed-0.105
-0.101
Token other
Token ID584
Feature activation-1.65e-02

Pos logit contributions
and+0.030
ying+0.029
as+0.029
adows+0.028
os+0.027
Neg logit contributions
 gust-0.017
 stem-0.017
sed-0.016
 mouth-0.016
-0.016
Token kn
Token ID638
Feature activation-2.31e-02

Pos logit contributions
and+0.035
pt+0.032
as+0.032
ying+0.031
 outage+0.031
Neg logit contributions
 gust-0.023
 stem-0.022
 mouth-0.021
 cheeks-0.020
sed-0.020
Tokenobs
Token ID8158
Feature activation-7.91e-02

Pos logit contributions
and+0.032
ying+0.030
as+0.029
adows+0.029
pt+0.028
Neg logit contributions
 gust-0.049
sed-0.047
 stem-0.047
 mouth-0.047
-0.046
Token qu
Token ID627
Feature activation-2.69e-02

Pos logit contributions
and+0.157
as+0.143
ying+0.143
adows+0.143
inker+0.140
Neg logit contributions
 gust-0.120
 mouth-0.120
sed-0.119
 stem-0.118
-0.114
Tokenarre
Token ID9624
Feature activation-2.45e-01

Pos logit contributions
and+0.039
ying+0.036
 outage+0.036
adows+0.035
iam+0.034
Neg logit contributions
 gust-0.054
 stem-0.053
-0.052
 mouth-0.051
sed-0.049
Tokenled
Token ID992
Feature activation-4.79e-03

Pos logit contributions
and+0.432
adows+0.406
iam+0.402
ibr+0.402
as+0.389
Neg logit contributions
 gust-0.415
 stem-0.380
 cheeks-0.369
 mouth-0.357
 timer-0.349
Token over
Token ID625
Feature activation-5.05e-02

Pos logit contributions
 outage+0.010
ying+0.010
and+0.010
adows+0.010
as+0.010
Neg logit contributions
 mouth-0.007
 gust-0.006
 stem-0.006
sed-0.006
 cheeks-0.006
Token who
Token ID508
Feature activation-8.08e-03

Pos logit contributions
and+0.105
ying+0.100
adows+0.100
as+0.099
 outage+0.096
Neg logit contributions
 gust-0.073
 mouth-0.070
 stem-0.069
sed-0.069
-0.068
Token should
Token ID815
Feature activation+9.26e-04

Pos logit contributions
and+0.016
ying+0.015
adows+0.014
as+0.014
inker+0.014
Neg logit contributions
 gust-0.013
-0.012
sed-0.012
 mouth-0.012
 stem-0.012
Token be
Token ID307
Feature activation+5.05e-02

Pos logit contributions
 gust+0.001
sed+0.001
+0.001
 stem+0.001
 mouth+0.001
Neg logit contributions
and-0.002
as-0.002
ying-0.002
adows-0.002
pt-0.002
Token but
Token ID475
Feature activation-1.17e-02

Pos logit contributions
and+0.010
as+0.009
ying+0.009
adows+0.009
pt+0.009
Neg logit contributions
 gust-0.008
 mouth-0.007
sed-0.007
 stem-0.007
-0.007
Token today
Token ID1909
Feature activation-3.30e-02

Pos logit contributions
and+0.025
as+0.024
ying+0.024
adows+0.023
udd+0.022
Neg logit contributions
 gust-0.017
sed-0.016
 mouth-0.016
 stem-0.016
-0.015
Token they
Token ID484
Feature activation-4.31e-02

Pos logit contributions
and+0.080
as+0.076
adows+0.075
ying+0.075
aking+0.071
Neg logit contributions
 gust-0.039
 mouth-0.036
-0.035
sed-0.034
 stem-0.034
Token were
Token ID547
Feature activation-2.45e-02

Pos logit contributions
and+0.098
ying+0.087
adows+0.087
as+0.085
 outage+0.084
Neg logit contributions
 gust-0.062
 mouth-0.057
-0.056
sed-0.056
 stem-0.055
Token qu
Token ID627
Feature activation-2.45e-02

Pos logit contributions
and+0.046
ying+0.043
as+0.043
adows+0.042
inker+0.040
Neg logit contributions
 gust-0.042
sed-0.040
 mouth-0.040
 stem-0.039
-0.039
Tokenarre
Token ID9624
Feature activation-2.45e-01

Pos logit contributions
and+0.022
ying+0.020
 outage+0.019
adows+0.018
pt+0.018
Neg logit contributions
 gust-0.063
-0.062
 stem-0.060
 mouth-0.059
sed-0.058
Tokenling
Token ID1359
Feature activation-3.56e-02

Pos logit contributions
and+0.373
 outage+0.351
adows+0.350
pt+0.348
iam+0.347
Neg logit contributions
 gust-0.453
 timer-0.402
 cheeks-0.400
 stem-0.395
 gu-0.380
Token.
Token ID13
Feature activation-2.43e-02

Pos logit contributions
and+0.070
ying+0.068
as+0.067
adows+0.066
 outage+0.065
Neg logit contributions
 gust-0.054
 mouth-0.053
sed-0.050
 cheeks-0.049
 stem-0.049
Token Jenna
Token ID45377
Feature activation-6.79e-02

Pos logit contributions
and+0.050
as+0.047
ying+0.047
adows+0.046
os+0.044
Neg logit contributions
 gust-0.037
sed-0.035
 stem-0.035
-0.034
 mouth-0.034
Token wanted
Token ID2227
Feature activation-2.36e-02

Pos logit contributions
and+0.145
ying+0.136
as+0.134
adows+0.133
inker+0.128
Neg logit contributions
 gust-0.098
sed-0.093
 mouth-0.092
 stem-0.092
-0.091
Token to
Token ID284
Feature activation-2.10e-02

Pos logit contributions
and+0.040
ying+0.038
as+0.038
adows+0.038
udd+0.035
Neg logit contributions
 gust-0.043
 mouth-0.042
sed-0.041
 stem-0.041
-0.040
Token friends
Token ID2460
Feature activation-6.47e-02

Pos logit contributions
and+0.082
as+0.081
ying+0.078
adows+0.075
os+0.072
Neg logit contributions
-0.076
sed-0.076
 gust-0.074
 stem-0.074
 mouth-0.071
Token didn
Token ID1422
Feature activation-2.11e-02

Pos logit contributions
and+0.136
adows+0.125
ying+0.124
as+0.122
 outage+0.121
Neg logit contributions
 gust-0.099
 mouth-0.090
sed-0.090
 cheeks-0.089
 stem-0.087
Token't
Token ID470
Feature activation-4.40e-02

Pos logit contributions
and+0.057
adows+0.056
as+0.056
ying+0.055
udd+0.053
Neg logit contributions
-0.019
 gust-0.017
 mouth-0.015
sed-0.015
 stem-0.014
Token stop
Token ID2245
Feature activation-2.41e-02

Pos logit contributions
ying+0.079
os+0.077
adows+0.077
udd+0.076
ibr+0.076
Neg logit contributions
 gust-0.073
 stem-0.064
 be-0.064
 mouth-0.063
 cheeks-0.062
Token qu
Token ID627
Feature activation-1.60e-02

Pos logit contributions
and+0.041
ying+0.039
as+0.038
pt+0.037
adows+0.036
Neg logit contributions
 gust-0.042
sed-0.042
 stem-0.042
-0.041
 mouth-0.040
Tokenarre
Token ID9624
Feature activation-2.43e-01

Pos logit contributions
and+0.012
ying+0.011
adows+0.010
as+0.010
 outage+0.009
Neg logit contributions
 gust-0.044
-0.043
 mouth-0.043
 stem-0.042
sed-0.042
Tokenling
Token ID1359
Feature activation-3.17e-02

Pos logit contributions
adows+0.327
and+0.319
ibr+0.305
as+0.304
iam+0.303
Neg logit contributions
 gust-0.501
 cheeks-0.457
 timer-0.445
 mouth-0.427
 stem-0.423
Token and
Token ID290
Feature activation-4.01e-02

Pos logit contributions
and+0.065
as+0.064
adows+0.063
ying+0.062
asc+0.061
Neg logit contributions
 gust-0.046
 mouth-0.044
 cheeks-0.042
sed-0.041
 tongue-0.040
Token were
Token ID547
Feature activation-2.13e-02

Pos logit contributions
and+0.080
as+0.076
ying+0.076
adows+0.074
os+0.071
Neg logit contributions
 gust-0.063
sed-0.060
 stem-0.059
 mouth-0.059
-0.059
Token getting
Token ID1972
Feature activation-6.70e-03

Pos logit contributions
and+0.042
ying+0.040
adows+0.040
as+0.039
inker+0.038
Neg logit contributions
 gust-0.035
sed-0.033
 mouth-0.033
 stem-0.031
-0.031
Token louder
Token ID27089
Feature activation+8.83e-03

Pos logit contributions
and+0.013
as+0.012
ying+0.012
adows+0.012
os+0.011
Neg logit contributions
 gust-0.011
sed-0.011
 stem-0.010
-0.010
 mouth-0.010
TokenPlease
Token ID5492
Feature activation+9.33e-02

Pos logit contributions
and+0.100
as+0.094
ying+0.093
adows+0.092
os+0.088
Neg logit contributions
 gust-0.074
sed-0.073
 mouth-0.071
 stem-0.070
-0.069
Token,
Token ID11
Feature activation+3.52e-04

Pos logit contributions
 gust+0.136
 mouth+0.131
+0.131
sed+0.130
 stem+0.128
Neg logit contributions
and-0.202
ying-0.188
as-0.185
adows-0.182
udd-0.174
Token no
Token ID645
Feature activation+2.86e-02

Pos logit contributions
 gust+0.000
 mouth+0.000
 stem+0.000
sed+0.000
+0.000
Neg logit contributions
and-0.001
as-0.001
adows-0.001
ying-0.001
udd-0.001
Token more
Token ID517
Feature activation+3.60e-02

Pos logit contributions
sed+0.059
 gust+0.057
+0.057
 stem+0.056
 mouth+0.055
Neg logit contributions
and-0.042
as-0.039
ying-0.039
adows-0.038
inker-0.036
Token qu
Token ID627
Feature activation-2.22e-02

Pos logit contributions
 gust+0.060
 mouth+0.055
 stem+0.054
sed+0.053
 cheeks+0.052
Neg logit contributions
and-0.073
adows-0.067
as-0.067
ying-0.066
udd-0.064
Tokenarre
Token ID9624
Feature activation-2.41e-01

Pos logit contributions
and+0.023
adows+0.019
ying+0.019
as+0.018
 outage+0.018
Neg logit contributions
 gust-0.056
-0.055
 mouth-0.053
 stem-0.053
sed-0.051
Tokenling
Token ID1359
Feature activation-2.81e-02

Pos logit contributions
and+0.386
adows+0.359
ibr+0.346
iam+0.343
udd+0.343
Neg logit contributions
 gust-0.462
 stem-0.403
 cheeks-0.401
 timer-0.394
 mouth-0.389
Token!
Token ID0
Feature activation-9.96e-03

Pos logit contributions
and+0.053
adows+0.050
udd+0.049
iam+0.048
ying+0.048
Neg logit contributions
 gust-0.044
 mouth-0.044
 stem-0.041
 top-0.039
 tongue-0.039
Token Let
Token ID3914
Feature activation-2.50e-02

Pos logit contributions
and+0.022
as+0.021
ying+0.021
os+0.021
adows+0.020
Neg logit contributions
 gust-0.013
 stem-0.012
sed-0.012
-0.011
 mouth-0.011
Token's
Token ID338
Feature activation+1.08e-02

Pos logit contributions
and+0.057
as+0.053
adows+0.053
ying+0.052
 outage+0.049
Neg logit contributions
 gust-0.035
 mouth-0.033
sed-0.031
 stem-0.031
-0.031
Token find
Token ID1064
Feature activation-1.97e-02

Pos logit contributions
 gust+0.015
sed+0.015
 stem+0.014
+0.014
 mouth+0.014
Neg logit contributions
and-0.023
as-0.022
ying-0.022
adows-0.021
os-0.021
Token and
Token ID290
Feature activation-3.43e-02

Pos logit contributions
 gust+0.035
sed+0.034
+0.034
 stem+0.034
 mouth+0.033
Neg logit contributions
and-0.028
ying-0.026
as-0.026
adows-0.025
pt-0.024
Token saw
Token ID2497
Feature activation-2.16e-02

Pos logit contributions
and+0.072
as+0.069
ying+0.069
adows+0.067
os+0.064
Neg logit contributions
 gust-0.050
sed-0.047
 stem-0.046
 mouth-0.046
-0.046
Token two
Token ID734
Feature activation-4.36e-02

Pos logit contributions
and+0.049
as+0.047
ying+0.047
adows+0.045
aking+0.045
Neg logit contributions
 gust-0.029
 mouth-0.028
sed-0.027
 stem-0.025
 cheeks-0.025
Token birds
Token ID10087
Feature activation+2.71e-02

Pos logit contributions
as+0.083
and+0.082
adows+0.079
ying+0.078
pt+0.075
Neg logit contributions
-0.068
sed-0.067
 stem-0.065
 gust-0.064
 mouth-0.061
Token qu
Token ID627
Feature activation-1.66e-02

Pos logit contributions
 gust+0.041
sed+0.040
 stem+0.039
+0.039
 mouth+0.039
Neg logit contributions
and-0.055
as-0.052
ying-0.052
adows-0.051
os-0.048
Tokenarre
Token ID9624
Feature activation-2.40e-01

Pos logit contributions
and+0.023
ying+0.021
as+0.020
adows+0.020
 outage+0.019
Neg logit contributions
 gust-0.036
 stem-0.034
sed-0.034
-0.034
 mouth-0.034
Tokenling
Token ID1359
Feature activation-3.15e-02

Pos logit contributions
and+0.360
adows+0.334
udd+0.332
iam+0.331
as+0.328
Neg logit contributions
 gust-0.460
 stem-0.426
 cheeks-0.421
 timer-0.417
 mouth-0.402
Token!
Token ID0
Feature activation+3.57e-03

Pos logit contributions
and+0.058
ying+0.055
as+0.055
adows+0.054
udd+0.052
Neg logit contributions
 gust-0.053
sed-0.051
 mouth-0.051
 stem-0.050
-0.049
Token They
Token ID1119
Feature activation-9.45e-03

Pos logit contributions
 gust+0.005
 stem+0.005
sed+0.005
+0.005
 cheeks+0.004
Neg logit contributions
and-0.007
os-0.007
as-0.007
adows-0.007
ying-0.007
Token were
Token ID547
Feature activation-1.57e-02

Pos logit contributions
and+0.018
ying+0.017
adows+0.017
as+0.017
 outage+0.016
Neg logit contributions
 gust-0.016
sed-0.015
 stem-0.015
 mouth-0.015
-0.014
Token squ
Token ID2809
Feature activation+2.17e-02

Pos logit contributions
and+0.028
ying+0.026
as+0.026
adows+0.025
os+0.024
Neg logit contributions
 gust-0.029
sed-0.028
 mouth-0.027
 stem-0.027
-0.027
Token both
Token ID1111
Feature activation+3.10e-02

Pos logit contributions
 gust+0.010
sed+0.010
+0.009
 stem+0.009
 mouth+0.009
Neg logit contributions
and-0.013
ying-0.013
as-0.013
adows-0.012
os-0.012
Token the
Token ID262
Feature activation-5.05e-03

Pos logit contributions
 gust+0.058
 mouth+0.055
 stem+0.054
sed+0.053
 cheeks+0.051
Neg logit contributions
and-0.055
ying-0.050
as-0.050
adows-0.050
inker-0.048
Token friends
Token ID2460
Feature activation-5.92e-02

Pos logit contributions
as+0.009
and+0.009
ying+0.008
adows+0.008
pt+0.008
Neg logit contributions
-0.008
 gust-0.008
sed-0.008
 stem-0.007
 mouth-0.007
Token stopped
Token ID5025
Feature activation+7.73e-03

Pos logit contributions
and+0.124
ying+0.113
adows+0.112
as+0.112
inker+0.110
Neg logit contributions
 gust-0.091
 mouth-0.085
sed-0.083
 stem-0.082
-0.082
Token qu
Token ID627
Feature activation-1.71e-02

Pos logit contributions
 gust+0.014
sed+0.013
 stem+0.013
+0.013
 mouth+0.013
Neg logit contributions
and-0.014
ying-0.013
as-0.013
adows-0.012
pt-0.012
Tokenarre
Token ID9624
Feature activation-2.40e-01

Pos logit contributions
and+0.018
as+0.016
ying+0.016
adows+0.015
os+0.014
Neg logit contributions
 gust-0.043
sed-0.042
 stem-0.041
-0.041
 mouth-0.041
Tokenling
Token ID1359
Feature activation-2.85e-02

Pos logit contributions
and+0.348
as+0.329
adows+0.327
pt+0.308
iam+0.306
Neg logit contributions
 gust-0.496
 stem-0.455
 cheeks-0.453
 mouth-0.445
 timer-0.444
Token.
Token ID13
Feature activation-1.17e-02

Pos logit contributions
and+0.054
as+0.052
ying+0.051
adows+0.050
inker+0.049
Neg logit contributions
 gust-0.047
 mouth-0.045
sed-0.044
 stem-0.043
-0.043
Token They
Token ID1119
Feature activation-8.44e-03

Pos logit contributions
and+0.024
as+0.022
os+0.022
ying+0.022
adows+0.021
Neg logit contributions
 gust-0.017
 stem-0.016
sed-0.016
 timer-0.015
 mouth-0.015
Token decided
Token ID3066
Feature activation-2.87e-02

Pos logit contributions
and+0.018
ying+0.017
adows+0.017
inker+0.016
 outage+0.016
Neg logit contributions
 gust-0.013
 mouth-0.012
 stem-0.012
sed-0.012
 cheeks-0.011
Token to
Token ID284
Feature activation-1.53e-02

Pos logit contributions
as+0.060
and+0.059
adows+0.059
ibr+0.057
os+0.056
Neg logit contributions
 mouth-0.040
 gust-0.040
sed-0.038
 cheeks-0.037
 stem-0.036
Token little
Token ID1310
Feature activation+8.41e-03

Pos logit contributions
as+0.081
and+0.081
ying+0.077
adows+0.076
udd+0.075
Neg logit contributions
sed-0.063
-0.063
 gust-0.061
 stem-0.061
 mouth-0.057
Token birds
Token ID10087
Feature activation+3.19e-02

Pos logit contributions
+0.012
 gust+0.012
sed+0.012
 stem+0.011
 mouth+0.011
Neg logit contributions
and-0.017
as-0.017
adows-0.016
ying-0.016
pt-0.016
Token that
Token ID326
Feature activation-5.46e-03

Pos logit contributions
 gust+0.050
sed+0.048
 stem+0.047
+0.047
 mouth+0.047
Neg logit contributions
and-0.063
as-0.060
ying-0.060
adows-0.059
pt-0.057
Token were
Token ID547
Feature activation-2.08e-02

Pos logit contributions
and+0.011
ying+0.010
as+0.010
adows+0.010
iam+0.010
Neg logit contributions
 gust-0.009
sed-0.008
 mouth-0.008
 stem-0.008
-0.008
Token qu
Token ID627
Feature activation-1.94e-02

Pos logit contributions
and+0.045
ying+0.043
as+0.042
adows+0.042
os+0.040
Neg logit contributions
 gust-0.029
sed-0.028
 stem-0.027
 mouth-0.027
-0.027
Tokenarre
Token ID9624
Feature activation-2.40e-01

Pos logit contributions
and+0.022
ying+0.020
adows+0.020
 outage+0.019
as+0.019
Neg logit contributions
 gust-0.046
-0.044
 stem-0.044
 mouth-0.043
sed-0.042
Tokenling
Token ID1359
Feature activation-2.94e-02

Pos logit contributions
and+0.382
adows+0.361
iam+0.351
as+0.344
ibr+0.344
Neg logit contributions
 gust-0.455
 stem-0.414
 cheeks-0.401
 timer-0.388
 mouth-0.386
Token.
Token ID13
Feature activation-2.19e-02

Pos logit contributions
ying+0.063
and+0.062
iam+0.062
adows+0.061
as+0.061
Neg logit contributions
 gust-0.039
 mouth-0.037
sed-0.034
 stem-0.034
 cheeks-0.033
Token
Token ID198
Feature activation-3.30e-02

Pos logit contributions
and+0.043
os+0.040
as+0.039
adows+0.037
ying+0.037
Neg logit contributions
 gust-0.034
sed-0.033
 stem-0.033
-0.031
 tongue-0.030
Token
Token ID198
Feature activation-3.31e-02

Pos logit contributions
and+0.075
ying+0.074
as+0.073
adows+0.073
pt+0.071
Neg logit contributions
 gust-0.038
-0.037
sed-0.035
 stem-0.035
 mouth-0.035
TokenThe
Token ID464
Feature activation+4.32e-02

Pos logit contributions
and+0.065
ying+0.062
as+0.061
adows+0.061
pt+0.058
Neg logit contributions
 gust-0.053
 stem-0.050
-0.049
sed-0.049
 mouth-0.048
Token and
Token ID290
Feature activation-3.90e-02

Pos logit contributions
and+0.150
ying+0.139
as+0.139
adows+0.139
inker+0.136
Neg logit contributions
 gust-0.105
 mouth-0.104
-0.101
sed-0.099
 stem-0.099
Token their
Token ID511
Feature activation-3.95e-02

Pos logit contributions
and+0.078
ying+0.075
as+0.073
adows+0.072
udd+0.069
Neg logit contributions
 gust-0.061
sed-0.058
 mouth-0.057
 stem-0.056
-0.055
Token children
Token ID1751
Feature activation-5.66e-02

Pos logit contributions
and+0.077
as+0.075
ying+0.071
adows+0.071
udd+0.069
Neg logit contributions
 gust-0.062
-0.059
 stem-0.058
sed-0.057
 mouth-0.055
Token,
Token ID11
Feature activation+3.86e-03

Pos logit contributions
and+0.118
as+0.110
ying+0.110
adows+0.108
inker+0.107
Neg logit contributions
 gust-0.080
 mouth-0.079
sed-0.076
-0.076
 stem-0.074
Token qu
Token ID627
Feature activation-1.75e-02

Pos logit contributions
 gust+0.006
 mouth+0.006
sed+0.006
+0.006
 stem+0.006
Neg logit contributions
and-0.008
ying-0.007
as-0.007
adows-0.007
inker-0.007
Tokenarre
Token ID9624
Feature activation-2.40e-01

Pos logit contributions
and+0.020
ying+0.018
adows+0.018
 outage+0.017
as+0.017
Neg logit contributions
 gust-0.041
 mouth-0.040
-0.039
 stem-0.039
sed-0.039
Tokenling
Token ID1359
Feature activation-2.80e-02

Pos logit contributions
and+0.435
adows+0.418
ibr+0.410
pt+0.406
iam+0.400
Neg logit contributions
 gust-0.364
 timer-0.339
 cheeks-0.336
 stem-0.329
 lips-0.309
Token.
Token ID13
Feature activation-2.73e-02

Pos logit contributions
and+0.057
ying+0.054
as+0.054
adows+0.053
 outage+0.052
Neg logit contributions
 gust-0.042
 mouth-0.041
 cheeks-0.040
sed-0.039
 stem-0.039
Token
Token ID220
Feature activation+3.68e-02

Pos logit contributions
and+0.055
as+0.050
os+0.050
ying+0.049
adows+0.049
Neg logit contributions
sed-0.042
 gust-0.042
 stem-0.041
 tongue-0.037
 cheeks-0.037
Token
Token ID198
Feature activation-4.11e-02

Pos logit contributions
+0.039
 gust+0.039
 mouth+0.038
sed+0.037
 stem+0.035
Neg logit contributions
and-0.085
ying-0.083
adows-0.083
as-0.082
pt-0.081
Token
Token ID198
Feature activation-4.02e-02

Pos logit contributions
ying+0.098
adows+0.097
and+0.095
pt+0.094
as+0.094
Neg logit contributions
-0.042
 gust-0.041
 mouth-0.037
 stem-0.037
sed-0.035
Token told
Token ID1297
Feature activation-6.52e-02

Pos logit contributions
and+0.060
adows+0.058
ying+0.056
as+0.056
inker+0.054
Neg logit contributions
 gust-0.037
 mouth-0.034
sed-0.034
-0.033
 stem-0.033
Token them
Token ID606
Feature activation-3.00e-02

Pos logit contributions
and+0.125
as+0.119
ying+0.117
adows+0.117
os+0.110
Neg logit contributions
 gust-0.109
sed-0.102
 mouth-0.101
 stem-0.101
-0.101
Token to
Token ID284
Feature activation-8.10e-03

Pos logit contributions
and+0.060
adows+0.059
as+0.058
aking+0.057
asc+0.057
Neg logit contributions
 mouth-0.046
 gust-0.044
 cheeks-0.043
sed-0.041
 mouths-0.040
Token stop
Token ID2245
Feature activation-1.17e-02

Pos logit contributions
and+0.018
as+0.018
ying+0.018
adows+0.017
pt+0.017
Neg logit contributions
 gust-0.010
sed-0.010
 stem-0.009
-0.009
 mouth-0.009
Token qu
Token ID627
Feature activation-7.65e-03

Pos logit contributions
and+0.022
ying+0.021
as+0.021
adows+0.020
pt+0.019
Neg logit contributions
 gust-0.019
sed-0.019
-0.018
 stem-0.018
 mouth-0.018
Tokenarre
Token ID9624
Feature activation-2.38e-01

Pos logit contributions
and+0.006
ying+0.005
as+0.005
adows+0.005
 outage+0.004
Neg logit contributions
 gust-0.021
 mouth-0.021
-0.021
sed-0.020
 stem-0.020
Tokenlling
Token ID2680
Feature activation-1.20e-01

Pos logit contributions
adows+0.418
and+0.402
pt+0.395
 outage+0.394
iam+0.393
Neg logit contributions
 gust-0.355
 timer-0.344
 cheeks-0.338
 lips-0.316
 stem-0.313
Token and
Token ID290
Feature activation-2.63e-02

Pos logit contributions
adows+0.276
and+0.273
asc+0.270
inker+0.269
iam+0.267
Neg logit contributions
 mouth-0.140
 gust-0.129
 cheeks-0.129
 tongue-0.121
sed-0.120
Token to
Token ID284
Feature activation-6.55e-03

Pos logit contributions
and+0.052
ying+0.049
as+0.049
adows+0.048
pt+0.046
Neg logit contributions
 gust-0.041
sed-0.040
-0.039
 stem-0.039
 mouth-0.038
Token share
Token ID2648
Feature activation+2.59e-02

Pos logit contributions
and+0.014
as+0.013
ying+0.013
adows+0.013
os+0.012
Neg logit contributions
 gust-0.010
sed-0.009
 stem-0.009
 mouth-0.009
-0.009
Token the
Token ID262
Feature activation+7.60e-03

Pos logit contributions
 mouth+0.037
 stem+0.035
 gust+0.035
+0.033
 cheeks+0.033
Neg logit contributions
and-0.053
adows-0.052
ying-0.051
as-0.051
 outage-0.051
Token other
Token ID584
Feature activation-1.12e-02

Pos logit contributions
 gust+0.001
 mouth+0.001
 stem+0.001
+0.001
sed+0.001
Neg logit contributions
and-0.002
as-0.002
adows-0.002
ying-0.001
udd-0.001
Token.
Token ID13
Feature activation-2.29e-02

Pos logit contributions
and+0.024
ying+0.023
asc+0.023
as+0.023
adows+0.023
Neg logit contributions
 mouth-0.015
 gust-0.015
 cheeks-0.014
sed-0.014
 tongue-0.014
Token They
Token ID1119
Feature activation-1.02e-02

Pos logit contributions
and+0.047
os+0.042
as+0.040
ying+0.039
 sorts+0.039
Neg logit contributions
 gust-0.037
 stem-0.035
sed-0.033
 timer-0.031
 cheeks-0.031
Token starting
Token ID3599
Feature activation+5.13e-02

Pos logit contributions
and+0.024
ying+0.022
adows+0.022
as+0.021
asc+0.021
Neg logit contributions
 gust-0.014
 mouth-0.013
 cheeks-0.013
sed-0.013
 stem-0.012
Token qu
Token ID627
Feature activation-1.50e-02

Pos logit contributions
 gust+0.090
 mouth+0.089
sed+0.089
+0.086
 stem+0.086
Neg logit contributions
and-0.091
as-0.087
ying-0.087
adows-0.086
udd-0.081
Tokenarre
Token ID9624
Feature activation-2.37e-01

Pos logit contributions
and+0.013
ying+0.012
 outage+0.011
adows+0.011
as+0.010
Neg logit contributions
 gust-0.040
 mouth-0.038
-0.038
 stem-0.037
sed-0.037
Tokenling
Token ID1359
Feature activation-2.43e-02

Pos logit contributions
adows+0.375
and+0.370
iam+0.359
ibr+0.354
 outage+0.352
Neg logit contributions
 gust-0.422
 cheeks-0.378
 timer-0.356
 mouth-0.353
 stem-0.352
Token again
Token ID757
Feature activation+1.35e-02

Pos logit contributions
and+0.049
as+0.049
adows+0.049
ying+0.049
asc+0.048
Neg logit contributions
 gust-0.035
 mouth-0.035
 cheeks-0.033
sed-0.032
 tongue-0.031
Token and
Token ID290
Feature activation-3.42e-02

Pos logit contributions
 gust+0.019
 mouth+0.019
 stem+0.018
+0.017
sed+0.017
Neg logit contributions
ying-0.028
and-0.028
as-0.028
asc-0.027
iam-0.027
Token their
Token ID511
Feature activation-3.35e-02

Pos logit contributions
and+0.063
ying+0.060
as+0.060
adows+0.058
os+0.056
Neg logit contributions
 gust-0.058
sed-0.056
-0.055
 stem-0.055
 mouth-0.054
Token voices
Token ID10839
Feature activation-4.00e-02

Pos logit contributions
and+0.062
as+0.061
pt+0.057
ying+0.057
adows+0.057
Neg logit contributions
 gust-0.051
 stem-0.050
-0.050
 mouth-0.048
sed-0.048
Token two
Token ID734
Feature activation-4.51e-02

Pos logit contributions
 gust+0.052
sed+0.051
 stem+0.050
 mouth+0.049
+0.049
Neg logit contributions
and-0.081
as-0.077
ying-0.077
adows-0.075
pt-0.072
Token squirrel
Token ID33039
Feature activation+6.29e-02

Pos logit contributions
and+0.093
ying+0.087
adows+0.086
as+0.086
 outage+0.082
Neg logit contributions
 gust-0.069
sed-0.064
 mouth-0.063
-0.063
 stem-0.062
Tokens
Token ID82
Feature activation-3.03e-03

Pos logit contributions
sed+0.077
 mouth+0.077
 gust+0.076
+0.073
 cheeks+0.069
Neg logit contributions
and-0.139
ying-0.127
 outage-0.127
adows-0.126
inker-0.125
Token kept
Token ID4030
Feature activation+1.45e-02

Pos logit contributions
and+0.006
ying+0.006
as+0.006
adows+0.005
 outage+0.005
Neg logit contributions
 gust-0.005
sed-0.005
 mouth-0.005
-0.005
 stem-0.004
Token qu
Token ID627
Feature activation-1.34e-02

Pos logit contributions
 gust+0.022
sed+0.021
 mouth+0.021
+0.021
 stem+0.021
Neg logit contributions
and-0.030
as-0.028
ying-0.028
adows-0.028
udd-0.026
Tokenarre
Token ID9624
Feature activation-2.37e-01

Pos logit contributions
 outage+0.017
and+0.017
ying+0.016
adows+0.015
pt+0.015
Neg logit contributions
 gust-0.028
-0.028
 mouth-0.027
 stem-0.026
 lips-0.025
Tokenling
Token ID1359
Feature activation-2.72e-02

Pos logit contributions
pt+0.411
 outage+0.408
ibr+0.404
and+0.403
asc+0.402
Neg logit contributions
 gust-0.337
 cheeks-0.324
 timer-0.308
 lips-0.297
 stem-0.292
Token until
Token ID1566
Feature activation+1.41e-02

Pos logit contributions
ying+0.054
 outage+0.053
and+0.053
adows+0.052
as+0.052
Neg logit contributions
 mouth-0.039
 gust-0.039
 cheeks-0.038
 stem-0.036
sed-0.036
Token finally
Token ID3443
Feature activation+4.25e-02

Pos logit contributions
sed+0.022
 gust+0.021
 stem+0.021
 mouth+0.020
+0.020
Neg logit contributions
and-0.027
ying-0.026
as-0.026
adows-0.025
pt-0.025
Token Sammy
Token ID44067
Feature activation-3.83e-02

Pos logit contributions
 gust+0.046
 mouth+0.043
 cheeks+0.039
sed+0.038
+0.037
Neg logit contributions
and-0.110
as-0.104
adows-0.103
 outage-0.101
ying-0.101
Token said
Token ID531
Feature activation+1.79e-02

Pos logit contributions
and+0.085
ying+0.082
adows+0.080
as+0.079
 outage+0.077
Neg logit contributions
 gust-0.051
 mouth-0.048
-0.047
sed-0.047
 cheeks-0.044
Tokenâ
Token ID22940
Feature activation-1.07e-01

Pos logit contributions
and+0.091
adows+0.089
as+0.088
ying+0.087
udd+0.085
Neg logit contributions
 gust-0.029
-0.029
 stem-0.022
sed-0.022
 mouth-0.022
Token
Token ID26391
Feature activation+1.03e-02

Pos logit contributions
and+0.182
ying+0.171
as+0.171
adows+0.167
os+0.158
Neg logit contributions
 gust-0.199
sed-0.193
 stem-0.189
 mouth-0.189
-0.187
Token
Token ID8151
Feature activation-9.55e-02

Pos logit contributions
+0.019
 gust+0.019
 mouth+0.018
sed+0.018
 cheeks+0.017
Neg logit contributions
ying-0.017
and-0.017
pt-0.016
as-0.016
 sorts-0.016
Tokent
Token ID83
Feature activation-3.46e-02

Pos logit contributions
and+0.169
adows+0.163
ying+0.163
as+0.158
inker+0.156
Neg logit contributions
 gust-0.161
-0.157
 mouth-0.152
sed-0.150
 stem-0.145
Token qu
Token ID627
Feature activation-1.55e-02

Pos logit contributions
and+0.067
adows+0.065
ying+0.064
udd+0.063
inker+0.063
Neg logit contributions
 gust-0.051
 mouth-0.046
 stem-0.044
 tongue-0.044
 cheeks-0.040
Tokenarre
Token ID9624
Feature activation-2.37e-01

Pos logit contributions
and+0.014
ying+0.012
adows+0.012
 outage+0.012
as+0.011
Neg logit contributions
 gust-0.041
-0.040
 mouth-0.039
 stem-0.038
sed-0.037
Tokenling
Token ID1359
Feature activation-2.23e-02

Pos logit contributions
and+0.405
adows+0.380
vern+0.362
ibr+0.362
as+0.355
Neg logit contributions
 gust-0.429
 cheeks-0.390
 stem-0.351
 mouth-0.349
 timer-0.349
Token,
Token ID11
Feature activation+1.09e-02

Pos logit contributions
and+0.050
as+0.048
adows+0.048
aking+0.046
ying+0.046
Neg logit contributions
 gust-0.029
 mouth-0.029
 cheeks-0.027
sed-0.025
 stem-0.025
Token the
Token ID262
Feature activation-2.64e-03

Pos logit contributions
 gust+0.014
 mouth+0.013
+0.012
sed+0.012
 stem+0.012
Neg logit contributions
and-0.026
adows-0.024
as-0.024
ying-0.024
inker-0.023
Token furniture
Token ID13091
Feature activation-1.01e-01

Pos logit contributions
and+0.006
ying+0.005
as+0.005
adows+0.005
os+0.005
Neg logit contributions
 gust-0.004
sed-0.004
 mouth-0.004
 stem-0.004
-0.004
Token stayed
Token ID9658
Feature activation+1.33e-02

Pos logit contributions
and+0.212
ying+0.195
as+0.195
adows+0.195
inker+0.186
Neg logit contributions
 gust-0.151
sed-0.143
 mouth-0.143
-0.141
 stem-0.139
Token peaceful
Token ID12309
Feature activation-2.87e-02

Pos logit contributions
and+0.019
ying+0.018
adows+0.018
as+0.018
iam+0.017
Neg logit contributions
 gust-0.014
 mouth-0.012
 stem-0.012
sed-0.012
-0.012
Token everyone
Token ID2506
Feature activation-6.77e-02

Pos logit contributions
as+0.062
and+0.062
iam+0.058
adows+0.058
ibr+0.058
Neg logit contributions
 gust-0.039
sed-0.037
 mouth-0.036
-0.036
 stem-0.036
Token just
Token ID655
Feature activation+3.29e-02

Pos logit contributions
and+0.133
inker+0.132
adows+0.127
as+0.126
iam+0.125
Neg logit contributions
 gust-0.098
 mouth-0.094
sed-0.090
 stem-0.089
 cheeks-0.088
Token stopped
Token ID5025
Feature activation+6.42e-03

Pos logit contributions
 gust+0.055
sed+0.053
 mouth+0.052
 stem+0.051
+0.050
Neg logit contributions
and-0.063
as-0.060
ying-0.059
adows-0.057
inker-0.056
Token qu
Token ID627
Feature activation-1.70e-02

Pos logit contributions
 gust+0.011
sed+0.011
+0.011
 stem+0.011
 mouth+0.010
Neg logit contributions
and-0.011
ying-0.011
as-0.011
adows-0.010
os-0.010
Tokenarre
Token ID9624
Feature activation-2.37e-01

Pos logit contributions
and+0.018
ying+0.016
as+0.016
adows+0.015
 outage+0.014
Neg logit contributions
 gust-0.043
-0.041
 mouth-0.041
 stem-0.041
sed-0.040
Tokenling
Token ID1359
Feature activation-2.68e-02

Pos logit contributions
and+0.401
adows+0.381
iam+0.375
as+0.374
ibr+0.362
Neg logit contributions
 gust-0.419
 cheeks-0.372
 timer-0.369
 stem-0.365
 mouth-0.348
Token.
Token ID13
Feature activation-2.27e-02

Pos logit contributions
and+0.056
as+0.055
ying+0.055
adows+0.055
udd+0.054
Neg logit contributions
 gust-0.036
 mouth-0.035
 stem-0.033
 cheeks-0.032
sed-0.031
Token
Token ID220
Feature activation+4.02e-02

Pos logit contributions
and+0.052
as+0.049
ying+0.049
adows+0.048
os+0.046
Neg logit contributions
 gust-0.029
sed-0.028
 stem-0.027
-0.027
 mouth-0.027
Token
Token ID198
Feature activation-3.14e-02

Pos logit contributions
 gust+0.050
+0.049
sed+0.048
 mouth+0.047
 stem+0.046
Neg logit contributions
and-0.090
ying-0.086
as-0.086
adows-0.085
pt-0.082
Token
Token ID198
Feature activation-3.06e-02

Pos logit contributions
ying+0.080
adows+0.078
as+0.077
pt+0.076
and+0.076
Neg logit contributions
-0.026
 gust-0.025
 mouth-0.024
sed-0.022
 stem-0.022
Token and
Token ID290
Feature activation-3.11e-02

Pos logit contributions
 gust+0.001
 mouth+0.001
sed+0.001
 cheeks+0.001
 tongue+0.001
Neg logit contributions
and-0.002
adows-0.002
as-0.002
ying-0.002
ibr-0.002
Token they
Token ID484
Feature activation-3.23e-02

Pos logit contributions
and+0.069
as+0.065
ying+0.065
adows+0.064
os+0.061
Neg logit contributions
 gust-0.043
sed-0.041
 mouth-0.040
 stem-0.039
-0.039
Token both
Token ID1111
Feature activation+3.12e-02

Pos logit contributions
and+0.075
ying+0.070
adows+0.068
as+0.065
udd+0.064
Neg logit contributions
 gust-0.046
sed-0.042
 mouth-0.041
 cheeks-0.040
 stem-0.039
Token stopped
Token ID5025
Feature activation+7.94e-03

Pos logit contributions
 gust+0.044
 mouth+0.041
sed+0.041
 cheeks+0.039
 stem+0.038
Neg logit contributions
and-0.071
ying-0.068
adows-0.064
as-0.063
inker-0.063
Token qu
Token ID627
Feature activation-1.31e-02

Pos logit contributions
 gust+0.014
sed+0.013
 stem+0.013
+0.013
 mouth+0.013
Neg logit contributions
and-0.015
as-0.014
ying-0.014
adows-0.013
os-0.013
Tokenarre
Token ID9624
Feature activation-2.36e-01

Pos logit contributions
and+0.012
ying+0.011
as+0.010
adows+0.010
 outage+0.009
Neg logit contributions
 gust-0.034
 mouth-0.033
-0.033
sed-0.033
 stem-0.033
Tokenlling
Token ID2680
Feature activation-1.22e-01

Pos logit contributions
and+0.341
adows+0.331
as+0.319
iam+0.309
pt+0.302
Neg logit contributions
 gust-0.478
 cheeks-0.434
 timer-0.429
 stem-0.428
 mouth-0.419
Token to
Token ID284
Feature activation-1.37e-02

Pos logit contributions
and+0.265
adows+0.260
as+0.255
ying+0.254
iam+0.251
Neg logit contributions
 gust-0.161
 mouth-0.160
sed-0.154
 cheeks-0.145
 tongue-0.141
Token look
Token ID804
Feature activation-3.19e-02

Pos logit contributions
and+0.025
as+0.023
ying+0.023
adows+0.023
pt+0.022
Neg logit contributions
 gust-0.024
sed-0.023
-0.023
 stem-0.023
 mouth-0.023
Token at
Token ID379
Feature activation+3.14e-02

Pos logit contributions
and+0.073
adows+0.070
as+0.069
iam+0.067
inker+0.067
Neg logit contributions
 gust-0.044
 mouth-0.042
sed-0.035
 lips-0.034
 cheeks-0.034
Token it
Token ID340
Feature activation+1.75e-02

Pos logit contributions
 gust+0.042
 mouth+0.041
sed+0.040
 stem+0.038
+0.038
Neg logit contributions
and-0.071
as-0.066
adows-0.066
ying-0.065
aking-0.064
TokenOne
Token ID3198
Feature activation+3.40e-02

Pos logit contributions
and+0.080
ying+0.080
adows+0.078
pt+0.078
as+0.078
Neg logit contributions
 gust-0.059
-0.056
 stem-0.053
 mouth-0.052
 timer-0.050
Token day
Token ID1110
Feature activation+3.07e-02

Pos logit contributions
 gust+0.049
sed+0.048
+0.046
 stem+0.046
 mouth+0.045
Neg logit contributions
and-0.072
as-0.068
ying-0.068
adows-0.067
os-0.065
Token they
Token ID484
Feature activation-3.69e-02

Pos logit contributions
 gust+0.043
sed+0.040
 mouth+0.040
+0.040
 stem+0.039
Neg logit contributions
and-0.066
as-0.063
ying-0.063
adows-0.062
udd-0.059
Token were
Token ID547
Feature activation-2.05e-02

Pos logit contributions
and+0.079
adows+0.076
 outage+0.076
ying+0.076
vern+0.073
Neg logit contributions
 gust-0.052
 mouth-0.047
-0.046
 stem-0.044
sed-0.043
Token qu
Token ID627
Feature activation-1.49e-02

Pos logit contributions
and+0.039
as+0.038
ying+0.037
adows+0.036
os+0.035
Neg logit contributions
 gust-0.033
 stem-0.032
-0.032
sed-0.031
 mouth-0.030
Tokenarre
Token ID9624
Feature activation-2.36e-01

Pos logit contributions
and+0.011
ying+0.009
as+0.009
adows+0.009
pt+0.008
Neg logit contributions
 gust-0.042
-0.040
 stem-0.040
 mouth-0.040
sed-0.040
Tokenling
Token ID1359
Feature activation-2.52e-02

Pos logit contributions
and+0.398
adows+0.378
as+0.362
vern+0.358
pt+0.356
Neg logit contributions
 gust-0.423
 cheeks-0.371
 stem-0.370
 timer-0.367
 mouth-0.356
Token over
Token ID625
Feature activation-3.93e-02

Pos logit contributions
and+0.049
as+0.047
ying+0.047
adows+0.046
 outage+0.044
Neg logit contributions
 gust-0.040
 mouth-0.038
sed-0.038
-0.037
 stem-0.036
Token who
Token ID508
Feature activation+2.15e-03

Pos logit contributions
and+0.074
as+0.069
ying+0.069
adows+0.068
os+0.065
Neg logit contributions
 gust-0.066
sed-0.064
 stem-0.063
-0.062
 mouth-0.062
Token should
Token ID815
Feature activation+9.63e-03

Pos logit contributions
 gust+0.003
sed+0.003
 mouth+0.003
+0.003
 stem+0.003
Neg logit contributions
and-0.005
ying-0.004
adows-0.004
as-0.004
inker-0.004
Token have
Token ID423
Feature activation+1.15e-02

Pos logit contributions
 gust+0.014
sed+0.013
 stem+0.013
+0.013
 mouth+0.013
Neg logit contributions
and-0.020
as-0.019
ying-0.019
adows-0.019
pt-0.018
Token the
Token ID262
Feature activation+1.49e-03

Pos logit contributions
and+0.004
as+0.004
ying+0.004
adows+0.004
udd+0.004
Neg logit contributions
 gust-0.003
 mouth-0.003
 stem-0.003
sed-0.003
-0.002
Token birds
Token ID10087
Feature activation+3.24e-02

Pos logit contributions
 gust+0.002
+0.002
sed+0.002
 mouth+0.002
 stem+0.002
Neg logit contributions
and-0.003
as-0.003
udd-0.003
adows-0.003
ying-0.003
Token were
Token ID547
Feature activation-1.68e-02

Pos logit contributions
 gust+0.051
sed+0.047
 mouth+0.047
+0.047
 stem+0.046
Neg logit contributions
and-0.066
as-0.059
iam-0.059
adows-0.059
inker-0.058
Token only
Token ID691
Feature activation+5.13e-02

Pos logit contributions
and+0.035
as+0.032
adows+0.032
ying+0.032
os+0.032
Neg logit contributions
 gust-0.026
sed-0.025
 mouth-0.025
 stem-0.024
 tongue-0.022
Token qu
Token ID627
Feature activation-1.59e-02

Pos logit contributions
 mouth+0.078
 gust+0.077
sed+0.076
 stem+0.073
 tongue+0.069
Neg logit contributions
and-0.103
ying-0.099
inker-0.095
as-0.095
adows-0.095
Tokenarre
Token ID9624
Feature activation-2.36e-01

Pos logit contributions
and+0.014
ying+0.013
 outage+0.013
adows+0.012
iam+0.011
Neg logit contributions
 gust-0.041
-0.040
 mouth-0.039
 stem-0.039
sed-0.038
Tokenling
Token ID1359
Feature activation-2.51e-02

Pos logit contributions
and+0.418
iam+0.398
adows+0.395
ibr+0.393
asc+0.381
Neg logit contributions
 gust-0.407
 cheeks-0.355
 stem-0.350
 mouth-0.339
 tongue-0.331
Token like
Token ID588
Feature activation-1.48e-02

Pos logit contributions
iam+0.052
and+0.052
inker+0.050
ying+0.050
asc+0.050
Neg logit contributions
 gust-0.034
 mouth-0.033
sed-0.030
 stem-0.030
 cheeks-0.030
Token little
Token ID1310
Feature activation+1.45e-02

Pos logit contributions
and+0.030
as+0.028
ying+0.028
adows+0.028
os+0.027
Neg logit contributions
 gust-0.022
sed-0.021
 stem-0.021
-0.021
 mouth-0.021
Token children
Token ID1751
Feature activation-4.76e-02

Pos logit contributions
 gust+0.020
sed+0.019
+0.019
 stem+0.019
 mouth+0.018
Neg logit contributions
and-0.031
as-0.030
ying-0.030
adows-0.029
pt-0.028
Token do
Token ID466
Feature activation+2.25e-02

Pos logit contributions
and+0.103
inker+0.100
iam+0.098
ying+0.097
asc+0.096
Neg logit contributions
 mouth-0.062
 gust-0.060
sed-0.057
 cheeks-0.056
-0.056
Token but
Token ID475
Feature activation-9.59e-03

Pos logit contributions
 gust+0.011
sed+0.011
 mouth+0.011
 stem+0.011
+0.011
Neg logit contributions
and-0.012
ying-0.011
as-0.011
adows-0.011
pt-0.011
Token today
Token ID1909
Feature activation-2.88e-02

Pos logit contributions
and+0.016
as+0.015
ying+0.015
os+0.015
 sorts+0.015
Neg logit contributions
 gust-0.016
-0.016
 stem-0.016
sed-0.016
 cheeks-0.016
Token they
Token ID484
Feature activation-3.12e-02

Pos logit contributions
and+0.056
ying+0.054
as+0.053
adows+0.052
os+0.051
Neg logit contributions
 gust-0.045
sed-0.045
 stem-0.044
-0.042
 mouth-0.041
Token were
Token ID547
Feature activation-1.87e-02

Pos logit contributions
and+0.069
ying+0.061
adows+0.061
as+0.060
 outage+0.058
Neg logit contributions
 gust-0.047
-0.044
sed-0.044
 mouth-0.043
 stem-0.043
Token qu
Token ID627
Feature activation-1.60e-02

Pos logit contributions
and+0.034
as+0.032
ying+0.032
adows+0.031
os+0.030
Neg logit contributions
 gust-0.032
sed-0.031
-0.031
 stem-0.031
 mouth-0.030
Tokenarre
Token ID9624
Feature activation-2.35e-01

Pos logit contributions
and+0.014
ying+0.012
adows+0.012
 outage+0.011
as+0.011
Neg logit contributions
 gust-0.042
-0.041
 stem-0.040
 mouth-0.040
sed-0.039
Tokenling
Token ID1359
Feature activation-2.38e-02

Pos logit contributions
and+0.418
adows+0.390
iam+0.380
as+0.378
pt+0.375
Neg logit contributions
 gust-0.399
 stem-0.348
 cheeks-0.344
 timer-0.339
 mouth-0.324
Token.
Token ID13
Feature activation-2.79e-02

Pos logit contributions
and+0.050
ying+0.048
as+0.047
inker+0.046
adows+0.046
Neg logit contributions
 mouth-0.033
 gust-0.033
sed-0.031
 cheeks-0.030
 stem-0.029
Token They
Token ID1119
Feature activation-8.43e-03

Pos logit contributions
and+0.047
os+0.046
as+0.045
adows+0.045
 sorts+0.042
Neg logit contributions
 gust-0.050
 stem-0.047
-0.046
sed-0.045
 timer-0.044
Token were
Token ID547
Feature activation-1.81e-02

Pos logit contributions
and+0.017
ying+0.015
as+0.015
adows+0.015
 outage+0.014
Neg logit contributions
 gust-0.014
sed-0.013
 stem-0.013
 mouth-0.013
-0.013
Token arguing
Token ID11810
Feature activation-5.74e-02

Pos logit contributions
and+0.033
as+0.032
ying+0.031
adows+0.031
pt+0.029
Neg logit contributions
 gust-0.031
sed-0.030
-0.030
 stem-0.029
 mouth-0.029
Token,
Token ID11
Feature activation+1.24e-02

Pos logit contributions
and+0.002
as+0.002
adows+0.002
ying+0.002
os+0.002
Neg logit contributions
 gust-0.002
 mouth-0.002
sed-0.002
 stem-0.002
-0.002
Token why
Token ID1521
Feature activation-5.20e-02

Pos logit contributions
 gust+0.015
 mouth+0.014
sed+0.014
 stem+0.013
+0.013
Neg logit contributions
and-0.030
adows-0.028
as-0.028
ying-0.028
pt-0.027
Token are
Token ID389
Feature activation-6.61e-02

Pos logit contributions
and+0.116
adows+0.099
year+0.098
col+0.096
iam+0.096
Neg logit contributions
 mouth-0.080
 gust-0.076
sed-0.070
 cheeks-0.068
 stem-0.067
Token you
Token ID345
Feature activation+9.87e-03

Pos logit contributions
and+0.146
as+0.139
ying+0.139
adows+0.136
os+0.131
Neg logit contributions
 gust-0.087
sed-0.084
 stem-0.082
-0.082
 mouth-0.081
Token qu
Token ID627
Feature activation-1.41e-02

Pos logit contributions
 gust+0.015
 mouth+0.015
sed+0.014
 stem+0.014
+0.014
Neg logit contributions
and-0.021
adows-0.019
as-0.019
ying-0.019
 sorts-0.018
Tokenarre
Token ID9624
Feature activation-2.35e-01

Pos logit contributions
and+0.012
ying+0.010
adows+0.010
as+0.009
 outage+0.009
Neg logit contributions
 gust-0.038
-0.037
 mouth-0.037
 stem-0.036
sed-0.036
Tokenling
Token ID1359
Feature activation-2.56e-02

Pos logit contributions
and+0.409
adows+0.377
ibr+0.375
iam+0.365
pt+0.358
Neg logit contributions
 gust-0.393
 cheeks-0.346
 stem-0.345
 lips-0.337
 mouth-0.334
Token?"
Token ID1701
Feature activation-5.14e-02

Pos logit contributions
and+0.050
as+0.045
adows+0.045
ibr+0.044
ying+0.043
Neg logit contributions
 mouth-0.042
 gust-0.041
 stem-0.039
sed-0.037
 grape-0.036
Token But
Token ID887
Feature activation+3.98e-02

Pos logit contributions
and+0.107
os+0.099
adows+0.098
smart+0.094
 sorts+0.093
Neg logit contributions
sed-0.076
 gust-0.075
 stem-0.073
 timer-0.064
 cheeks-0.063
Token the
Token ID262
Feature activation-2.85e-04

Pos logit contributions
 gust+0.067
sed+0.067
 stem+0.066
+0.065
 cheeks+0.061
Neg logit contributions
and-0.068
ying-0.067
as-0.064
adows-0.064
pt-0.062
Token birds
Token ID10087
Feature activation+3.15e-02

Pos logit contributions
as+0.001
and+0.001
adows+0.000
ying+0.000
aking+0.000
Neg logit contributions
-0.000
sed-0.000
 gust-0.000
 mouth-0.000
 stem-0.000
Token He
Token ID679
Feature activation+4.13e-02

Pos logit contributions
 gust+0.069
sed+0.067
 stem+0.065
 cheeks+0.061
 timer+0.060
Neg logit contributions
and-0.102
ying-0.092
as-0.091
os-0.090
iam-0.089
Token sp
Token ID599
Feature activation-9.25e-02

Pos logit contributions
 gust+0.076
+0.075
 stem+0.069
sed+0.067
 cheeks+0.064
Neg logit contributions
as-0.065
pt-0.064
and-0.063
ying-0.061
os-0.061
Tokenits
Token ID896
Feature activation-2.71e-02

Pos logit contributions
and+0.169
adows+0.159
 sorts+0.151
 outage+0.149
smart+0.148
Neg logit contributions
 gust-0.150
 stem-0.143
sed-0.141
 mouth-0.139
 cheeks-0.138
Token out
Token ID503
Feature activation+1.40e-02

Pos logit contributions
and+0.062
adows+0.060
as+0.058
ying+0.058
inker+0.057
Neg logit contributions
 gust-0.035
 mouth-0.034
 stem-0.032
sed-0.031
 tongue-0.031
Token his
Token ID465
Feature activation+3.39e-03

Pos logit contributions
 mouth+0.017
 gust+0.017
 stem+0.016
 tongue+0.015
 grape+0.015
Neg logit contributions
and-0.034
adows-0.033
as-0.032
ying-0.032
inker-0.031
Token sandwich
Token ID20433
Feature activation+1.26e-01

Pos logit contributions
 gust+0.005
Really+0.004
+0.004
 sidelines+0.004
 stem+0.004
Neg logit contributions
and-0.007
as-0.006
reat-0.006
pt-0.006
ont-0.006
Token.
Token ID13
Feature activation+4.65e-02

Pos logit contributions
 gust+0.216
sed+0.209
 stem+0.205
 mouth+0.205
+0.204
Neg logit contributions
and-0.231
as-0.217
ying-0.216
adows-0.213
os-0.202
Token He
Token ID679
Feature activation+3.89e-02

Pos logit contributions
 gust+0.075
sed+0.072
 stem+0.070
 cheeks+0.066
+0.066
Neg logit contributions
and-0.094
ying-0.082
as-0.081
os-0.080
iam-0.079
Token says
Token ID1139
Feature activation+7.11e-02

Pos logit contributions
 gust+0.072
+0.072
 stem+0.066
sed+0.065
 cheeks+0.062
Neg logit contributions
as-0.060
pt-0.059
and-0.059
ying-0.058
udd-0.057
Token o
Token ID267
Feature activation+2.21e-03

Pos logit contributions
 gust+0.111
+0.109
sed+0.108
 stem+0.107
 cheeks+0.100
Neg logit contributions
and-0.136
ying-0.129
as-0.128
pt-0.126
os-0.123
Tokenuch
Token ID794
Feature activation+5.13e-02

Pos logit contributions
 gust+0.004
sed+0.004
 stem+0.004
 mouth+0.004
 cheeks+0.004
Neg logit contributions
and-0.004
as-0.004
ying-0.003
adows-0.003
os-0.003
Token pirate
Token ID25868
Feature activation+7.06e-04

Pos logit contributions
+0.021
 gust+0.021
 stem+0.021
sed+0.021
 mouth+0.020
Neg logit contributions
and-0.020
as-0.019
adows-0.019
ying-0.018
udd-0.018
Token captain
Token ID10654
Feature activation-6.20e-02

Pos logit contributions
 mouth+0.001
 gust+0.001
sed+0.001
 stem+0.001
+0.001
Neg logit contributions
and-0.001
ying-0.001
as-0.001
adows-0.001
udd-0.001
Token.
Token ID13
Feature activation-1.79e-03

Pos logit contributions
and+0.121
ying+0.119
as+0.115
udd+0.114
adows+0.112
Neg logit contributions
 mouth-0.089
 gust-0.088
sed-0.084
 stem-0.083
-0.081
Token The
Token ID383
Feature activation+4.65e-02

Pos logit contributions
and+0.004
as+0.003
ying+0.003
adows+0.003
os+0.003
Neg logit contributions
 gust-0.003
sed-0.003
 stem-0.003
 cheeks-0.002
 mouth-0.002
Token par
Token ID1582
Feature activation+7.13e-02

Pos logit contributions
 gust+0.080
+0.080
sed+0.078
 stem+0.076
 cheeks+0.073
Neg logit contributions
and-0.082
as-0.078
iam-0.076
ying-0.075
adows-0.075
Tokenrot
Token ID10599
Feature activation+1.05e-01

Pos logit contributions
 gust+0.127
 stem+0.118
sed+0.117
 mouth+0.116
 cheeks+0.116
Neg logit contributions
and-0.135
as-0.133
ying-0.123
adows-0.119
inker-0.119
Token would
Token ID561
Feature activation-2.67e-02

Pos logit contributions
 mouth+0.137
sed+0.136
 gust+0.131
 stem+0.128
 cheeks+0.123
Neg logit contributions
and-0.236
ying-0.226
inker-0.215
adows-0.213
asc-0.212
Token act
Token ID719
Feature activation-3.54e-02

Pos logit contributions
and+0.061
ying+0.060
as+0.058
adows+0.056
udd+0.055
Neg logit contributions
 gust-0.034
 stem-0.034
sed-0.033
 mouth-0.033
-0.031
Token his
Token ID465
Feature activation-2.34e-02

Pos logit contributions
asc+0.071
ying+0.069
os+0.065
beh+0.064
and+0.064
Neg logit contributions
 gust-0.052
 mouth-0.048
 face-0.045
 cheeks-0.043
 lips-0.042
Token part
Token ID636
Feature activation-3.76e-02

Pos logit contributions
and+0.049
ying+0.047
as+0.046
adows+0.044
os+0.044
Neg logit contributions
 gust-0.035
sed-0.035
 mouth-0.034
 stem-0.033
 cheeks-0.032
Token too
Token ID1165
Feature activation+3.85e-02

Pos logit contributions
and+0.064
asc+0.064
ying+0.063
inker+0.062
as+0.062
Neg logit contributions
 mouth-0.064
 gust-0.062
 stem-0.060
sed-0.059
 cheeks-0.058
Token friends
Token ID2460
Feature activation-5.33e-02

Pos logit contributions
+0.027
 gust+0.027
 stem+0.025
sed+0.024
 mouth+0.022
Neg logit contributions
and-0.047
as-0.046
adows-0.044
pt-0.044
ying-0.044
Token found
Token ID1043
Feature activation+2.13e-03

Pos logit contributions
and+0.120
ying+0.111
as+0.111
inker+0.108
 outage+0.108
Neg logit contributions
 gust-0.072
 mouth-0.068
sed-0.065
-0.063
 cheeks-0.062
Token him
Token ID683
Feature activation+4.62e-02

Pos logit contributions
 gust+0.003
 mouth+0.003
sed+0.003
 stem+0.003
+0.003
Neg logit contributions
and-0.004
as-0.004
adows-0.004
ying-0.004
 outage-0.004
Token on
Token ID319
Feature activation-2.73e-02

Pos logit contributions
 mouth+0.074
 gust+0.072
sed+0.071
 stem+0.068
+0.068
Neg logit contributions
and-0.087
as-0.086
ying-0.085
inker-0.082
adows-0.080
Token the
Token ID262
Feature activation-1.67e-03

Pos logit contributions
and+0.059
as+0.058
ying+0.057
adows+0.056
os+0.053
Neg logit contributions
 gust-0.039
 mouth-0.037
-0.036
 stem-0.036
sed-0.036
Token branch
Token ID8478
Feature activation+1.04e-01

Pos logit contributions
and+0.003
adows+0.003
 outage+0.003
as+0.003
udd+0.003
Neg logit contributions
-0.002
 gust-0.002
sed-0.002
 mouth-0.002
Really-0.002
Token.
Token ID13
Feature activation-1.10e-02

Pos logit contributions
 gust+0.156
 mouth+0.155
+0.150
 stem+0.148
sed+0.147
Neg logit contributions
ying-0.200
and-0.199
as-0.196
inker-0.192
iam-0.192
Token They
Token ID1119
Feature activation-4.79e-03

Pos logit contributions
and+0.025
as+0.023
ying+0.023
adows+0.023
os+0.022
Neg logit contributions
 gust-0.015
sed-0.014
 stem-0.014
-0.013
 mouth-0.013
Token tried
Token ID3088
Feature activation+4.77e-02

Pos logit contributions
and+0.011
ying+0.010
inker+0.010
 outage+0.010
adows+0.009
Neg logit contributions
 gust-0.007
 mouth-0.007
 stem-0.006
sed-0.006
 cheeks-0.006
Token to
Token ID284
Feature activation-8.70e-03

Pos logit contributions
 mouth+0.084
 gust+0.083
sed+0.081
 stem+0.079
+0.078
Neg logit contributions
and-0.083
as-0.083
ying-0.081
adows-0.081
inker-0.075
Token wake
Token ID7765
Feature activation-5.42e-02

Pos logit contributions
and+0.018
as+0.017
ying+0.017
adows+0.017
pt+0.016
Neg logit contributions
 gust-0.013
sed-0.012
 stem-0.012
-0.012
 mouth-0.012
Token with
Token ID351
Feature activation+5.86e-02

Pos logit contributions
 gust+0.051
 mouth+0.049
+0.048
 stem+0.047
sed+0.046
Neg logit contributions
and-0.070
as-0.070
ying-0.068
adows-0.067
inker-0.065
Token her
Token ID607
Feature activation-4.68e-03

Pos logit contributions
 gust+0.084
 mouth+0.083
 stem+0.080
+0.079
sed+0.077
Neg logit contributions
and-0.122
ying-0.120
as-0.116
adows-0.114
 outage-0.113
Token friends
Token ID2460
Feature activation-4.80e-02

Pos logit contributions
and+0.010
as+0.009
ying+0.009
adows+0.009
pt+0.009
Neg logit contributions
 gust-0.007
sed-0.007
 stem-0.006
-0.006
 mouth-0.006
Token.
Token ID13
Feature activation-5.68e-03

Pos logit contributions
and+0.100
ying+0.096
as+0.096
inker+0.095
adows+0.094
Neg logit contributions
 gust-0.067
 mouth-0.066
-0.066
 stem-0.063
sed-0.062
Token The
Token ID383
Feature activation+4.31e-02

Pos logit contributions
ying+0.012
and+0.012
as+0.012
adows+0.011
 outage+0.011
Neg logit contributions
 mouth-0.008
 gust-0.008
-0.007
 stem-0.007
sed-0.007
Token sun
Token ID4252
Feature activation+1.07e-01

Pos logit contributions
sed+0.066
 gust+0.065
+0.063
 stem+0.062
 mouth+0.061
Neg logit contributions
as-0.081
and-0.081
ying-0.078
adows-0.078
iam-0.076
Token was
Token ID373
Feature activation+1.59e-03

Pos logit contributions
 gust+0.174
sed+0.168
 stem+0.164
+0.164
 mouth+0.163
Neg logit contributions
and-0.205
as-0.195
ying-0.195
adows-0.190
os-0.183
Token bright
Token ID6016
Feature activation+1.43e-02

Pos logit contributions
 gust+0.002
sed+0.002
 stem+0.002
 mouth+0.002
+0.002
Neg logit contributions
and-0.003
as-0.003
ying-0.003
adows-0.003
os-0.003
Token and
Token ID290
Feature activation-2.52e-02

Pos logit contributions
 gust+0.025
 stem+0.023
 mouth+0.023
+0.023
sed+0.023
Neg logit contributions
and-0.027
as-0.025
ying-0.025
adows-0.024
inker-0.024
Token warm
Token ID5814
Feature activation-5.81e-03

Pos logit contributions
and+0.056
ying+0.054
as+0.053
adows+0.052
inker+0.051
Neg logit contributions
 gust-0.036
 mouth-0.033
 stem-0.032
sed-0.031
-0.030
Token,
Token ID11
Feature activation+1.62e-02

Pos logit contributions
and+0.011
as+0.010
ying+0.010
adows+0.010
os+0.009
Neg logit contributions
 gust-0.010
sed-0.010
-0.009
 stem-0.009
 mouth-0.009
Token again
Token ID757
Feature activation+4.68e-02

Pos logit contributions
 mouth+0.044
 gust+0.043
sed+0.041
 stem+0.040
+0.040
Neg logit contributions
and-0.059
ying-0.059
as-0.057
iam-0.057
inker-0.056
Token."
Token ID526
Feature activation-2.03e-02

Pos logit contributions
 mouth+0.077
 gust+0.077
 stem+0.074
+0.072
sed+0.071
Neg logit contributions
and-0.085
ying-0.084
as-0.080
iam-0.078
inker-0.078
Token
Token ID198
Feature activation+1.24e-02

Pos logit contributions
and+0.042
ying+0.036
adows+0.036
as+0.036
os+0.036
Neg logit contributions
 gust-0.032
sed-0.031
 stem-0.031
-0.028
 mouth-0.028
Token
Token ID198
Feature activation+1.19e-02

Pos logit contributions
 gust+0.018
sed+0.017
 mouth+0.017
+0.017
 stem+0.017
Neg logit contributions
and-0.025
ying-0.025
as-0.024
adows-0.024
os-0.023
TokenThe
Token ID464
Feature activation+6.78e-02

Pos logit contributions
 gust+0.020
+0.019
 stem+0.019
 mouth+0.018
sed+0.017
Neg logit contributions
ying-0.021
adows-0.021
and-0.021
as-0.020
os-0.020
Token end
Token ID886
Feature activation+1.11e-01

Pos logit contributions
sed+0.096
+0.093
 gust+0.093
 stem+0.092
 mouth+0.092
Neg logit contributions
and-0.134
ying-0.132
as-0.131
adows-0.126
pt-0.124
Token.
Token ID13
Feature activation+2.53e-02

Pos logit contributions
 gust+0.145
 mouth+0.143
sed+0.133
+0.128
 cheeks+0.125
Neg logit contributions
and-0.243
ying-0.237
as-0.232
adows-0.228
udd-0.224
Token<|endoftext|>
Token ID50256
Feature activation+1.75e-04

Pos logit contributions
 gust+0.038
sed+0.037
 stem+0.037
 mouth+0.035
 cheeks+0.035
Neg logit contributions
and-0.052
as-0.048
ying-0.048
adows-0.047
os-0.046
TokenTom
Token ID13787
Feature activation-4.41e-03

Pos logit contributions
 gust+0.000
+0.000
 mouth+0.000
sed+0.000
 stem+0.000
Neg logit contributions
and-0.000
as-0.000
ying-0.000
adows-0.000
udd-0.000
Token and
Token ID290
Feature activation-1.60e-02

Pos logit contributions
and+0.009
 outage+0.008
ying+0.008
as+0.008
vell+0.007
Neg logit contributions
 mouth-0.007
sed-0.007
 ham-0.007
 stem-0.006
 tongue-0.006
Token Mia
Token ID32189
Feature activation-4.92e-02

Pos logit contributions
and+0.032
as+0.030
ying+0.030
adows+0.029
udd+0.028
Neg logit contributions
 gust-0.025
sed-0.024
 mouth-0.024
 stem-0.024
-0.024
Token did
Token ID750
Feature activation+1.40e-02

Pos logit contributions
and+0.049
ying+0.044
adows+0.041
inker+0.040
 outage+0.040
Neg logit contributions
 stem-0.034
-0.033
sed-0.033
 gust-0.032
 mouth-0.032
Token it
Token ID340
Feature activation+3.10e-02

Pos logit contributions
 gust+0.019
 mouth+0.018
 stem+0.018
sed+0.018
+0.018
Neg logit contributions
and-0.031
ying-0.029
as-0.029
adows-0.029
os-0.028
Token with
Token ID351
Feature activation+6.73e-02

Pos logit contributions
 mouth+0.050
 gust+0.049
+0.048
 top+0.048
 stem+0.047
Neg logit contributions
and-0.054
adows-0.052
as-0.051
 outage-0.051
ying-0.051
Token love
Token ID1842
Feature activation-4.27e-02

Pos logit contributions
 gust+0.112
 mouth+0.108
sed+0.107
 stem+0.106
+0.103
Neg logit contributions
and-0.130
ying-0.124
as-0.122
adows-0.121
os-0.113
Token."
Token ID526
Feature activation-4.75e-02

Pos logit contributions
adows+0.076
ying+0.075
and+0.075
iam+0.073
pt+0.072
Neg logit contributions
 mouth-0.068
 gust-0.068
 stem-0.066
-0.064
sed-0.063
Token
Token ID198
Feature activation-1.42e-02

Pos logit contributions
and+0.095
ying+0.085
as+0.084
adows+0.083
udd+0.083
Neg logit contributions
 gust-0.078
sed-0.075
 stem-0.074
 mouth-0.069
-0.067
Token
Token ID198
Feature activation-1.41e-02

Pos logit contributions
and+0.031
as+0.029
ying+0.029
adows+0.029
os+0.027
Neg logit contributions
 gust-0.020
sed-0.019
 stem-0.018
-0.018
 mouth-0.018
TokenThe
Token ID464
Feature activation+5.74e-02

Pos logit contributions
and+0.024
ying+0.023
adows+0.023
as+0.023
pt+0.022
Neg logit contributions
 gust-0.026
 stem-0.024
-0.024
sed-0.024
 mouth-0.024
Token pair
Token ID5166
Feature activation-6.23e-02

Pos logit contributions
sed+0.091
+0.091
 gust+0.090
 mouth+0.087
 tongue+0.086
Neg logit contributions
and-0.102
as-0.099
ying-0.098
 sorts-0.097
inker-0.095
Token hugged
Token ID41130
Feature activation+9.79e-03

Pos logit contributions
and+0.104
as+0.097
ying+0.097
adows+0.095
os+0.090
Neg logit contributions
 gust-0.116
sed-0.113
 stem-0.111
-0.111
 mouth-0.110
Token and
Token ID290
Feature activation-1.72e-02

Pos logit contributions
 gust+0.017
sed+0.017
 stem+0.017
+0.016
 mouth+0.016
Neg logit contributions
and-0.018
ying-0.016
as-0.016
adows-0.016
pt-0.015
Tokenig
Token ID328
Feature activation-1.82e-02

Pos logit contributions
and+0.062
as+0.061
ying+0.060
adows+0.059
inker+0.055
Neg logit contributions
 gust-0.083
 mouth-0.080
sed-0.080
 stem-0.079
-0.078
Tokenzag
Token ID50183
Feature activation-1.05e-01

Pos logit contributions
and+0.028
ying+0.026
as+0.025
adows+0.025
pt+0.024
Neg logit contributions
 gust-0.035
sed-0.035
 stem-0.035
 mouth-0.035
-0.034
Token was
Token ID373
Feature activation-1.80e-03

Pos logit contributions
and+0.206
as+0.197
ying+0.194
adows+0.193
os+0.184
Neg logit contributions
 gust-0.162
sed-0.156
-0.156
 stem-0.153
 mouth-0.151
Token so
Token ID523
Feature activation-7.36e-03

Pos logit contributions
and+0.003
ying+0.003
as+0.003
adows+0.003
inker+0.003
Neg logit contributions
 gust-0.003
 mouth-0.003
 stem-0.003
sed-0.003
 tongue-0.003
Token happy
Token ID3772
Feature activation-2.68e-03

Pos logit contributions
and+0.016
ying+0.015
as+0.015
adows+0.015
 outage+0.015
Neg logit contributions
 gust-0.010
 mouth-0.009
 stem-0.009
sed-0.009
-0.009
Token!
Token ID0
Feature activation+9.00e-03

Pos logit contributions
and+0.005
as+0.005
ying+0.005
adows+0.005
udd+0.005
Neg logit contributions
 gust-0.004
 mouth-0.004
sed-0.004
-0.004
 stem-0.004
Token It
Token ID632
Feature activation+3.91e-02

Pos logit contributions
 gust+0.012
sed+0.012
 stem+0.011
 mouth+0.011
+0.011
Neg logit contributions
and-0.020
ying-0.019
as-0.019
adows-0.018
os-0.018
Token ran
Token ID4966
Feature activation-2.23e-02

Pos logit contributions
 gust+0.046
 mouth+0.043
sed+0.043
 stem+0.042
 cheeks+0.038
Neg logit contributions
and-0.095
ying-0.092
as-0.089
adows-0.088
inker-0.086
Token around
Token ID1088
Feature activation-5.54e-02

Pos logit contributions
as+0.039
adows+0.039
ying+0.038
and+0.038
os+0.037
Neg logit contributions
 gust-0.040
 mouth-0.037
sed-0.036
 top-0.035
 stem-0.035
Token the
Token ID262
Feature activation-1.48e-03

Pos logit contributions
and+0.114
as+0.112
ying+0.110
adows+0.108
udd+0.105
Neg logit contributions
 gust-0.082
sed-0.078
 mouth-0.077
 stem-0.077
-0.076
Token room
Token ID2119
Feature activation-3.87e-02

Pos logit contributions
and+0.003
as+0.003
ying+0.003
adows+0.003
os+0.003
Neg logit contributions
 gust-0.002
sed-0.002
 stem-0.002
 mouth-0.002
-0.002
TokenOne
Token ID3198
Feature activation+3.46e-02

Pos logit contributions
and+0.068
as+0.064
ying+0.063
adows+0.062
os+0.058
Neg logit contributions
 gust-0.066
sed-0.066
 mouth-0.064
-0.063
 stem-0.063
Token day
Token ID1110
Feature activation+2.74e-02

Pos logit contributions
sed+0.054
 gust+0.053
+0.053
 stem+0.051
 mouth+0.050
Neg logit contributions
and-0.070
as-0.066
ying-0.065
adows-0.064
 sorts-0.062
Token,
Token ID11
Feature activation+2.98e-03

Pos logit contributions
sed+0.057
 stem+0.055
+0.054
 timer+0.053
 gust+0.052
Neg logit contributions
ying-0.036
and-0.035
pt-0.034
as-0.033
adows-0.033
Token she
Token ID673
Feature activation-1.67e-02

Pos logit contributions
sed+0.007
+0.007
 stem+0.006
 sidelines+0.006
 tongue+0.006
Neg logit contributions
 if-0.003
 sorts-0.003
and-0.003
 along-0.003
 sure-0.002
Token found
Token ID1043
Feature activation-9.79e-03

Pos logit contributions
and+0.036
ying+0.035
adows+0.034
as+0.033
 outage+0.032
Neg logit contributions
 mouth-0.023
 gust-0.023
sed-0.022
 stem-0.022
-0.021
Token a
Token ID257
Feature activation+5.39e-03

Pos logit contributions
and+0.020
as+0.019
ying+0.019
adows+0.018
os+0.017
Neg logit contributions
 gust-0.015
sed-0.015
 mouth-0.014
 stem-0.014
-0.014
Token box
Token ID3091
Feature activation-3.83e-02

Pos logit contributions
+0.009
sed+0.009
 cheeks+0.009
 gust+0.009
 stem+0.008
Neg logit contributions
and-0.009
as-0.008
 outage-0.008
 sorts-0.008
washer-0.008
Token full
Token ID1336
Feature activation+4.95e-03

Pos logit contributions
and+0.062
as+0.057
ying+0.057
adows+0.056
os+0.053
Neg logit contributions
 gust-0.074
sed-0.072
 stem-0.071
-0.071
 mouth-0.070
Token of
Token ID286
Feature activation-5.40e-03

Pos logit contributions
sed+0.011
 gust+0.011
+0.011
 stem+0.011
 mouth+0.011
Neg logit contributions
and-0.006
as-0.006
ying-0.006
adows-0.005
pt-0.005
Token screws
Token ID23742
Feature activation-6.66e-02

Pos logit contributions
and+0.010
ying+0.010
as+0.010
adows+0.010
os+0.009
Neg logit contributions
 gust-0.009
sed-0.008
 mouth-0.008
 stem-0.008
-0.008
Token and
Token ID290
Feature activation-3.82e-02

Pos logit contributions
and+0.146
asc+0.143
adows+0.143
inker+0.141
iam+0.139
Neg logit contributions
 mouth-0.080
 stem-0.079
 gust-0.077
-0.076
 cheeks-0.071
Token and
Token ID290
Feature activation-3.60e-02

Pos logit contributions
 gust+0.015
 mouth+0.013
+0.013
 stem+0.012
 tongue+0.011
Neg logit contributions
asc-0.028
ibr-0.028
inker-0.027
adows-0.027
as-0.027
Token started
Token ID2067
Feature activation+2.59e-02

Pos logit contributions
and+0.077
ying+0.073
as+0.073
adows+0.071
inker+0.068
Neg logit contributions
 gust-0.053
sed-0.050
 mouth-0.049
 stem-0.049
-0.047
Token to
Token ID284
Feature activation-1.37e-02

Pos logit contributions
 gust+0.046
sed+0.044
 mouth+0.044
 stem+0.043
+0.042
Neg logit contributions
and-0.046
as-0.044
ying-0.044
adows-0.042
iam-0.041
Token sing
Token ID1702
Feature activation-2.75e-02

Pos logit contributions
and+0.033
as+0.032
ying+0.032
adows+0.031
os+0.030
Neg logit contributions
 gust-0.015
sed-0.015
 stem-0.014
 mouth-0.014
-0.014
Token.
Token ID13
Feature activation-2.05e-02

Pos logit contributions
ying+0.058
as+0.056
and+0.056
iam+0.056
 outage+0.055
Neg logit contributions
 gust-0.040
 mouth-0.036
 stem-0.035
 cheeks-0.033
 lips-0.033
Token As
Token ID1081
Feature activation+3.81e-02

Pos logit contributions
and+0.041
adows+0.037
 sorts+0.036
os+0.036
ying+0.035
Neg logit contributions
sed-0.031
 gust-0.030
 timer-0.029
 mouth-0.028
 stem-0.028
Token she
Token ID673
Feature activation-1.53e-02

Pos logit contributions
sed+0.051
 gust+0.049
 mouth+0.048
 stem+0.048
+0.048
Neg logit contributions
and-0.083
ying-0.079
as-0.079
adows-0.077
pt-0.074
Token sang
Token ID25889
Feature activation-1.47e-02

Pos logit contributions
and+0.030
as+0.029
ying+0.029
adows+0.028
os+0.027
Neg logit contributions
 gust-0.024
sed-0.023
 stem-0.023
-0.022
 mouth-0.022
Token her
Token ID607
Feature activation-7.21e-03

Pos logit contributions
and+0.034
as+0.034
ying+0.034
iam+0.033
adows+0.032
Neg logit contributions
 gust-0.019
 mouth-0.016
 stem-0.015
 cheeks-0.015
sed-0.014
Token song
Token ID3496
Feature activation+2.66e-02

Pos logit contributions
and+0.015
as+0.014
adows+0.014
ying+0.014
pt+0.014
Neg logit contributions
 gust-0.010
-0.010
sed-0.010
 stem-0.010
 mouth-0.009
Token,
Token ID11
Feature activation+4.24e-03

Pos logit contributions
 gust+0.039
sed+0.036
 mouth+0.036
 stem+0.035
+0.035
Neg logit contributions
and-0.056
ying-0.053
as-0.053
adows-0.052
iam-0.051
Token Tim
Token ID5045
Feature activation+8.72e-03

Pos logit contributions
sed+0.057
 gust+0.056
 stem+0.055
+0.053
 mouth+0.051
Neg logit contributions
and-0.071
ying-0.067
as-0.066
adows-0.063
pt-0.063
Token began
Token ID2540
Feature activation+2.35e-02

Pos logit contributions
 mouth+0.008
 gust+0.008
 cheeks+0.007
 be+0.007
 head+0.007
Neg logit contributions
and-0.021
 outage-0.021
auc-0.021
vell-0.021
ying-0.020
Token to
Token ID284
Feature activation-1.68e-02

Pos logit contributions
 mouth+0.032
 gust+0.032
 cheeks+0.028
 tongue+0.028
+0.028
Neg logit contributions
as-0.049
vell-0.048
ying-0.048
and-0.047
auc-0.046
Token jog
Token ID48342
Feature activation-8.07e-02

Pos logit contributions
and+0.038
as+0.037
ying+0.037
auc+0.035
os+0.035
Neg logit contributions
 gust-0.022
 mouth-0.021
-0.021
 stem-0.020
 cheeks-0.020
Token.
Token ID13
Feature activation-2.23e-02

Pos logit contributions
as+0.165
ying+0.162
ibr+0.159
os+0.157
and+0.156
Neg logit contributions
 gust-0.108
-0.104
 top-0.104
 mouth-0.100
 cheeks-0.092
Token He
Token ID679
Feature activation-1.30e-02

Pos logit contributions
and+0.047
adows+0.042
ying+0.041
as+0.041
os+0.040
Neg logit contributions
 gust-0.034
sed-0.033
 stem-0.033
-0.031
 mouth-0.030
Token could
Token ID714
Feature activation+1.19e-02

Pos logit contributions
and+0.030
ying+0.029
 outage+0.028
asc+0.027
as+0.027
Neg logit contributions
 gust-0.016
 mouth-0.016
sed-0.015
 cheeks-0.014
 stem-0.013
Token feel
Token ID1254
Feature activation-3.33e-02

Pos logit contributions
 gust+0.017
sed+0.016
+0.016
 mouth+0.016
 stem+0.015
Neg logit contributions
and-0.026
ying-0.025
as-0.025
adows-0.024
os-0.023
Token the
Token ID262
Feature activation-6.31e-03

Pos logit contributions
as+0.087
and+0.086
os+0.085
asc+0.084
auc+0.083
Neg logit contributions
 gust-0.034
 mouth-0.029
 fle-0.026
 top-0.022
 sun-0.021
Token wind
Token ID2344
Feature activation+4.23e-02

Pos logit contributions
and+0.017
ying+0.016
as+0.016
adows+0.016
os+0.015
Neg logit contributions
 gust-0.006
 mouth-0.006
sed-0.006
 stem-0.006
-0.005
Token blowing
Token ID19280
Feature activation-5.96e-05

Pos logit contributions
 gust+0.072
 stem+0.066
 mouth+0.066
+0.065
sed+0.063
Neg logit contributions
and-0.080
as-0.074
adows-0.073
inker-0.071
ying-0.071
Token But
Token ID887
Feature activation+5.77e-02

Pos logit contributions
 gust+0.021
 stem+0.019
 timer+0.018
sed+0.018
+0.018
Neg logit contributions
and-0.024
os-0.021
as-0.021
adows-0.021
smart-0.020
Token you
Token ID345
Feature activation+2.83e-02

Pos logit contributions
 gust+0.083
+0.080
sed+0.076
 stem+0.072
 cheeks+0.070
Neg logit contributions
and-0.117
ying-0.116
as-0.111
reat-0.108
 if-0.107
Token have
Token ID423
Feature activation+3.53e-02

Pos logit contributions
 gust+0.038
sed+0.036
 mouth+0.036
+0.036
 stem+0.035
Neg logit contributions
and-0.063
ying-0.060
as-0.060
adows-0.059
udd-0.056
Token to
Token ID284
Feature activation+8.94e-03

Pos logit contributions
 gust+0.066
 mouth+0.065
sed+0.065
 stem+0.063
+0.061
Neg logit contributions
and-0.059
ying-0.058
adows-0.056
as-0.056
inker-0.053
Token follow
Token ID1061
Feature activation-4.41e-02

Pos logit contributions
 gust+0.011
sed+0.010
 stem+0.010
+0.010
 mouth+0.010
Neg logit contributions
and-0.021
as-0.020
ying-0.020
adows-0.020
os-0.019
Token my
Token ID616
Feature activation-2.60e-02

Pos logit contributions
and+0.102
as+0.098
ying+0.098
adows+0.097
udd+0.092
Neg logit contributions
 gust-0.054
 mouth-0.051
sed-0.051
 stem-0.050
-0.049
Token rules
Token ID3173
Feature activation-7.59e-02

Pos logit contributions
and+0.060
as+0.058
ying+0.057
adows+0.056
inker+0.055
Neg logit contributions
 gust-0.032
-0.030
sed-0.029
 stem-0.029
 mouth-0.027
Token.
Token ID13
Feature activation+1.63e-02

Pos logit contributions
beh+0.182
adows+0.176
inker+0.175
vell+0.174
lled+0.173
Neg logit contributions
 mouth-0.066
 top-0.054
 stem-0.052
 mom-0.052
-0.051
Token You
Token ID921
Feature activation+3.26e-02

Pos logit contributions
 gust+0.024
sed+0.023
 stem+0.023
+0.021
 cheeks+0.021
Neg logit contributions
and-0.034
as-0.031
adows-0.031
os-0.030
ying-0.030
Token have
Token ID423
Feature activation+3.68e-02

Pos logit contributions
 gust+0.042
 mouth+0.042
+0.042
sed+0.040
 stem+0.039
Neg logit contributions
and-0.075
ying-0.070
adows-0.068
iam-0.066
udd-0.066
Token to
Token ID284
Feature activation+9.83e-03

Pos logit contributions
 mouth+0.061
sed+0.058
 gust+0.055
 stem+0.054
 tongue+0.052
Neg logit contributions
ying-0.074
and-0.070
adows-0.069
inker-0.069
as-0.066
Token little
Token ID1310
Feature activation+7.01e-03

Pos logit contributions
+0.051
sed+0.050
 gust+0.044
 mouths+0.044
 tongue+0.044
Neg logit contributions
ying-0.049
and-0.048
as-0.045
aking-0.044
adows-0.044
Token girl
Token ID2576
Feature activation+2.76e-02

Pos logit contributions
sed+0.012
+0.011
 gust+0.010
 cheeks+0.010
 tongue+0.010
Neg logit contributions
ying-0.013
and-0.013
aking-0.012
 sorts-0.012
as-0.012
Token was
Token ID373
Feature activation-1.01e-02

Pos logit contributions
 gust+0.048
+0.046
sed+0.046
 stem+0.045
 mouth+0.043
Neg logit contributions
and-0.048
as-0.046
adows-0.044
os-0.044
ying-0.044
Token not
Token ID407
Feature activation-3.20e-02

Pos logit contributions
and+0.019
ying+0.018
as+0.018
adows+0.018
inker+0.017
Neg logit contributions
 gust-0.017
 mouth-0.016
sed-0.016
 stem-0.015
-0.015
Token so
Token ID523
Feature activation-1.07e-02

Pos logit contributions
and+0.059
udd+0.057
as+0.056
ying+0.055
os+0.055
Neg logit contributions
 gust-0.057
 mouth-0.051
 stem-0.050
sed-0.049
 tongue-0.047
Token scared
Token ID12008
Feature activation-5.49e-02

Pos logit contributions
and+0.024
as+0.023
inker+0.022
iam+0.022
adows+0.022
Neg logit contributions
 gust-0.017
 mouth-0.013
 grape-0.012
 stem-0.012
sed-0.012
Token any
Token ID597
Feature activation-4.60e-02

Pos logit contributions
and+0.117
adows+0.116
as+0.116
iam+0.111
inker+0.110
Neg logit contributions
 gust-0.078
 mouth-0.068
 stem-0.064
-0.063
 cheeks-0.063
Token more
Token ID517
Feature activation+3.65e-02

Pos logit contributions
and+0.081
ying+0.077
as+0.076
adows+0.074
 outage+0.071
Neg logit contributions
sed-0.082
-0.080
 gust-0.080
 stem-0.079
 mouth-0.077
Token.
Token ID13
Feature activation-2.43e-02

Pos logit contributions
 gust+0.057
 mouth+0.052
 stem+0.050
sed+0.050
 cheeks+0.050
Neg logit contributions
and-0.076
as-0.071
adows-0.069
asc-0.068
udd-0.068
Token The
Token ID383
Feature activation+3.05e-02

Pos logit contributions
and+0.051
ying+0.046
adows+0.046
as+0.045
os+0.044
Neg logit contributions
 gust-0.038
sed-0.036
 stem-0.035
-0.033
 timer-0.033
Token old
Token ID1468
Feature activation+2.15e-02

Pos logit contributions
+0.055
sed+0.054
 tongue+0.050
 mouths+0.049
 gust+0.049
Neg logit contributions
ying-0.052
aking-0.044
adows-0.044
as-0.043
udd-0.043
Token Lily
Token ID20037
Feature activation+1.26e-02

Pos logit contributions
 gust+0.056
sed+0.053
 stem+0.052
+0.050
 mouth+0.049
Neg logit contributions
and-0.079
as-0.073
ying-0.072
adows-0.071
os-0.069
Token should
Token ID815
Feature activation+3.10e-02

Pos logit contributions
 mouth+0.018
sed+0.018
 gust+0.017
 cheeks+0.016
 stem+0.016
Neg logit contributions
and-0.027
ying-0.026
adows-0.025
as-0.025
 sorts-0.024
Token not
Token ID407
Feature activation-1.56e-03

Pos logit contributions
 gust+0.040
 mouth+0.040
 stem+0.040
sed+0.040
+0.039
Neg logit contributions
and-0.072
as-0.066
ying-0.065
adows-0.064
os-0.063
Token have
Token ID423
Feature activation+3.68e-02

Pos logit contributions
and+0.003
as+0.003
ying+0.003
adows+0.003
pt+0.003
Neg logit contributions
 gust-0.002
sed-0.002
-0.002
 mouth-0.002
 stem-0.002
Token been
Token ID587
Feature activation+8.12e-02

Pos logit contributions
 gust+0.072
sed+0.070
 mouth+0.069
 stem+0.069
+0.068
Neg logit contributions
and-0.059
ying-0.055
as-0.055
adows-0.054
os-0.050
Token naughty
Token ID45128
Feature activation-6.45e-02

Pos logit contributions
 gust+0.129
 mouth+0.127
sed+0.122
 stem+0.116
 tongue+0.113
Neg logit contributions
and-0.156
ying-0.151
adows-0.149
inker-0.146
as-0.146
Token.
Token ID13
Feature activation+3.67e-02

Pos logit contributions
adows+0.145
and+0.139
os+0.137
asc+0.137
 sorts+0.136
Neg logit contributions
 gust-0.076
 mouth-0.074
 cheeks-0.071
sed-0.068
 tongue-0.065
Token She
Token ID1375
Feature activation+2.23e-02

Pos logit contributions
 gust+0.048
sed+0.046
 stem+0.045
+0.044
 mouth+0.044
Neg logit contributions
and-0.083
as-0.078
ying-0.078
adows-0.077
os-0.074
Token should
Token ID815
Feature activation+3.19e-02

Pos logit contributions
 gust+0.039
sed+0.039
 mouth+0.038
 stem+0.037
+0.037
Neg logit contributions
and-0.041
ying-0.038
as-0.037
adows-0.037
os-0.035
Token not
Token ID407
Feature activation-1.14e-03

Pos logit contributions
 gust+0.044
sed+0.043
 stem+0.042
 mouth+0.042
+0.042
Neg logit contributions
and-0.070
as-0.066
ying-0.065
adows-0.064
os-0.062
Token have
Token ID423
Feature activation+3.71e-02

Pos logit contributions
and+0.002
as+0.002
ying+0.002
adows+0.002
pt+0.002
Neg logit contributions
 gust-0.002
sed-0.002
-0.002
 stem-0.002
 mouth-0.002
Token to
Token ID284
Feature activation+1.36e-02

Pos logit contributions
 mouth+0.106
sed+0.102
+0.094
 tongue+0.094
 cheeks+0.093
Neg logit contributions
and-0.131
ying-0.131
inker-0.127
adows-0.125
iam-0.124
Token be
Token ID307
Feature activation+8.22e-02

Pos logit contributions
 gust+0.019
sed+0.018
+0.018
 stem+0.017
 mouth+0.017
Neg logit contributions
and-0.029
as-0.028
adows-0.027
ying-0.027
pt-0.026
Token safe
Token ID3338
Feature activation+4.74e-03

Pos logit contributions
 mouth+0.129
 gust+0.124
 stem+0.118
sed+0.116
 tongue+0.115
Neg logit contributions
and-0.165
ying-0.158
inker-0.153
 outage-0.153
as-0.151
Token and
Token ID290
Feature activation+1.42e-02

Pos logit contributions
 mouth+0.006
 stem+0.006
sed+0.005
 gust+0.005
 tongue+0.005
Neg logit contributions
and-0.010
inker-0.010
adows-0.010
ying-0.010
ibr-0.010
Token happy
Token ID3772
Feature activation+2.50e-02

Pos logit contributions
 gust+0.025
+0.024
sed+0.024
 stem+0.024
 mouth+0.023
Neg logit contributions
and-0.025
as-0.024
ying-0.023
adows-0.023
pt-0.022
Token."
Token ID526
Feature activation-1.64e-02

Pos logit contributions
 gust+0.045
sed+0.044
 mouth+0.044
 stem+0.043
+0.042
Neg logit contributions
and-0.043
as-0.041
ying-0.040
adows-0.040
inker-0.038
Token
Token ID198
Feature activation+1.90e-02

Pos logit contributions
and+0.027
as+0.024
os+0.024
ying+0.023
adows+0.022
Neg logit contributions
 gust-0.033
sed-0.031
 stem-0.030
-0.029
 timer-0.028
Token
Token ID198
Feature activation+1.80e-02

Pos logit contributions
 gust+0.031
sed+0.030
 stem+0.029
+0.029
 mouth+0.028
Neg logit contributions
and-0.038
as-0.036
ying-0.035
adows-0.034
os-0.033
TokenThey
Token ID2990
Feature activation+2.02e-02

Pos logit contributions
 gust+0.037
sed+0.036
 mouth+0.035
 stem+0.035
+0.035
Neg logit contributions
and-0.028
as-0.026
ying-0.025
adows-0.024
os-0.023
Token waited
Token ID13488
Feature activation+1.48e-02

Pos logit contributions
 gust+0.029
sed+0.028
 mouth+0.028
 stem+0.028
+0.027
Neg logit contributions
and-0.045
ying-0.042
 outage-0.040
adows-0.040
as-0.040
Token for
Token ID329
Feature activation+4.63e-02

Pos logit contributions
 gust+0.024
sed+0.023
 mouth+0.023
 stem+0.023
+0.022
Neg logit contributions
and-0.028
as-0.027
ying-0.026
adows-0.026
os-0.026
Token tell
Token ID1560
Feature activation-4.48e-02

Pos logit contributions
and+0.021
as+0.020
ying+0.020
adows+0.019
os+0.019
Neg logit contributions
 gust-0.014
 stem-0.013
 mouth-0.013
sed-0.013
-0.013
Token you
Token ID345
Feature activation+1.10e-02

Pos logit contributions
and+0.098
ying+0.094
as+0.094
adows+0.093
aking+0.091
Neg logit contributions
 gust-0.058
 mouth-0.058
 stem-0.054
sed-0.054
 cheeks-0.053
Token that
Token ID326
Feature activation+5.18e-04

Pos logit contributions
 mouth+0.018
 stem+0.016
 top+0.016
 cheeks+0.015
 gust+0.015
Neg logit contributions
iam-0.020
and-0.020
asc-0.020
idget-0.020
auc-0.019
Token there
Token ID612
Feature activation-3.45e-02

Pos logit contributions
 gust+0.001
 mouth+0.001
sed+0.001
 stem+0.001
+0.001
Neg logit contributions
and-0.001
as-0.001
ying-0.001
adows-0.001
udd-0.001
Token is
Token ID318
Feature activation-4.79e-02

Pos logit contributions
and+0.069
iam+0.064
as+0.063
ying+0.063
udd+0.062
Neg logit contributions
 mouth-0.051
 gust-0.050
 stem-0.048
sed-0.048
-0.046
Token nothing
Token ID2147
Feature activation-2.15e-02

Pos logit contributions
and+0.098
ying+0.093
as+0.092
adows+0.089
inker+0.087
Neg logit contributions
 gust-0.073
 mouth-0.071
sed-0.071
 stem-0.069
-0.066
Token to
Token ID284
Feature activation-8.91e-03

Pos logit contributions
inker+0.042
iam+0.041
and+0.041
adows+0.039
asc+0.039
Neg logit contributions
 gust-0.033
 mouth-0.032
 stem-0.029
sed-0.029
 grape-0.028
Token be
Token ID307
Feature activation+6.06e-02

Pos logit contributions
and+0.019
ying+0.018
as+0.018
adows+0.017
pt+0.017
Neg logit contributions
 gust-0.013
sed-0.013
-0.012
 mouth-0.012
 stem-0.012
Token scared
Token ID12008
Feature activation-4.38e-02

Pos logit contributions
 mouth+0.100
 gust+0.098
 stem+0.091
sed+0.090
 tongue+0.087
Neg logit contributions
and-0.118
ying-0.111
iam-0.109
inker-0.107
as-0.105
Token of
Token ID286
Feature activation+3.00e-03

Pos logit contributions
iam+0.081
zzy+0.079
adows+0.078
and+0.078
ibr+0.078
Neg logit contributions
 gust-0.070
 mouth-0.063
 stem-0.059
 cheeks-0.056
 sun-0.056
Token".
Token ID1911
Feature activation-3.50e-02

Pos logit contributions
 gust+0.005
 mouth+0.005
 stem+0.004
sed+0.004
 cheeks+0.004
Neg logit contributions
and-0.006
as-0.006
ying-0.005
adows-0.005
iam-0.005
Token was
Token ID373
Feature activation-1.29e-02

Pos logit contributions
and+0.125
ying+0.115
adows+0.113
as+0.113
 outage+0.111
Neg logit contributions
 gust-0.067
sed-0.064
 mouth-0.063
 stem-0.063
-0.061
Token playing
Token ID2712
Feature activation-5.81e-03

Pos logit contributions
and+0.024
as+0.023
ying+0.023
adows+0.022
aking+0.021
Neg logit contributions
 gust-0.022
sed-0.021
 stem-0.021
-0.020
 mouth-0.020
Token near
Token ID1474
Feature activation+6.69e-02

Pos logit contributions
and+0.011
as+0.011
ying+0.011
adows+0.011
 outage+0.011
Neg logit contributions
 gust-0.009
 stem-0.008
 mouth-0.008
-0.008
sed-0.008
Token the
Token ID262
Feature activation-8.62e-03

Pos logit contributions
sed+0.119
 gust+0.117
+0.113
 stem+0.109
 mouth+0.108
Neg logit contributions
and-0.118
as-0.110
ying-0.110
adows-0.107
os-0.105
Token f
Token ID277
Feature activation-1.54e-01

Pos logit contributions
and+0.018
as+0.017
ying+0.017
adows+0.017
os+0.016
Neg logit contributions
 gust-0.013
sed-0.012
 stem-0.012
-0.012
 mouth-0.012
Tokenauc
Token ID14272
Feature activation-2.14e-01

Pos logit contributions
and+0.138
adows+0.118
ust+0.109
pt+0.107
lled+0.104
Neg logit contributions
-0.390
 stem-0.369
 gust-0.357
 mouth-0.357
sed-0.352
Tokenet
Token ID316
Feature activation-4.53e-02

Pos logit contributions
and+0.366
as+0.342
ying+0.334
adows+0.324
 outage+0.322
Neg logit contributions
 stem-0.378
sed-0.378
 mouth-0.378
 gust-0.364
 cheeks-0.360
Token with
Token ID351
Feature activation+4.76e-02

Pos logit contributions
and+0.086
as+0.081
ying+0.080
adows+0.079
pt+0.075
Neg logit contributions
 gust-0.074
-0.071
sed-0.071
 stem-0.071
 mouth-0.071
Token her
Token ID607
Feature activation-1.60e-02

Pos logit contributions
 gust+0.079
sed+0.075
 stem+0.074
 mouth+0.074
+0.073
Neg logit contributions
and-0.091
ying-0.087
as-0.086
adows-0.085
udd-0.081
Token friends
Token ID2460
Feature activation-6.07e-02

Pos logit contributions
and+0.034
as+0.032
ying+0.032
adows+0.031
pt+0.030
Neg logit contributions
 gust-0.023
sed-0.022
 stem-0.022
-0.022
 mouth-0.021
Token.
Token ID13
Feature activation-3.12e-02

Pos logit contributions
and+0.128
adows+0.121
ying+0.121
as+0.121
inker+0.119
Neg logit contributions
 gust-0.083
 mouth-0.082
 stem-0.079
-0.079
sed-0.078
Token it
Token ID340
Feature activation+1.01e-02

Pos logit contributions
and+0.089
as+0.085
ying+0.085
adows+0.083
pt+0.080
Neg logit contributions
 gust-0.065
sed-0.063
-0.062
 stem-0.061
 mouth-0.060
Token had
Token ID550
Feature activation+1.94e-02

Pos logit contributions
 gust+0.014
 mouth+0.014
 stem+0.014
sed+0.013
+0.013
Neg logit contributions
and-0.022
adows-0.021
as-0.021
ying-0.020
inker-0.020
Token made
Token ID925
Feature activation-3.70e-02

Pos logit contributions
 gust+0.031
sed+0.030
 mouth+0.029
 stem+0.029
+0.028
Neg logit contributions
and-0.038
ying-0.037
as-0.036
adows-0.036
os-0.034
Token a
Token ID257
Feature activation+7.57e-04

Pos logit contributions
and+0.073
as+0.070
ying+0.070
adows+0.069
 outage+0.066
Neg logit contributions
 gust-0.058
 mouth-0.054
 stem-0.054
-0.053
sed-0.052
Token huge
Token ID3236
Feature activation-3.35e-02

Pos logit contributions
+0.001
sed+0.001
 gust+0.001
 cheeks+0.001
 stem+0.001
Neg logit contributions
and-0.001
 outage-0.001
as-0.001
pt-0.001
udd-0.001
Token mess
Token ID2085
Feature activation-1.37e-01

Pos logit contributions
 sorts+0.038
and+0.037
pt+0.037
 outage+0.037
udd+0.037
Neg logit contributions
sed-0.070
-0.070
 gust-0.065
 stem-0.063
Really-0.062
Token.
Token ID13
Feature activation-3.98e-02

Pos logit contributions
adows+0.319
and+0.310
inker+0.298
udd+0.296
as+0.295
Neg logit contributions
 gust-0.149
 mouth-0.135
-0.134
 stem-0.132
 cheeks-0.128
Token She
Token ID1375
Feature activation-2.37e-02

Pos logit contributions
and+0.073
adows+0.060
os+0.059
 sorts+0.057
ying+0.057
Neg logit contributions
sed-0.075
 gust-0.072
 timer-0.071
 stem-0.067
 tongue-0.065
Token tried
Token ID3088
Feature activation+3.60e-02

Pos logit contributions
and+0.043
as+0.041
ying+0.040
adows+0.039
pt+0.039
Neg logit contributions
 gust-0.041
sed-0.039
-0.039
 stem-0.039
 mouth-0.038
Token to
Token ID284
Feature activation-1.85e-02

Pos logit contributions
 gust+0.064
 mouth+0.062
 stem+0.060
+0.060
sed+0.059
Neg logit contributions
adows-0.062
as-0.062
and-0.061
ying-0.060
asc-0.057
Token clean
Token ID3424
Feature activation-6.85e-02

Pos logit contributions
and+0.034
as+0.032
ying+0.032
adows+0.031
os+0.030
Neg logit contributions
 gust-0.032
 stem-0.031
sed-0.031
 mouth-0.030
-0.030
Token had
Token ID550
Feature activation+3.12e-02

Pos logit contributions
and+0.145
ying+0.136
as+0.136
adows+0.135
inker+0.128
Neg logit contributions
 gust-0.106
sed-0.102
 mouth-0.102
-0.100
 stem-0.099
Token made
Token ID925
Feature activation-3.01e-02

Pos logit contributions
 gust+0.053
 mouth+0.052
sed+0.051
 stem+0.050
+0.049
Neg logit contributions
and-0.058
ying-0.057
adows-0.057
as-0.054
inker-0.052
Token such
Token ID884
Feature activation+2.77e-02

Pos logit contributions
and+0.062
as+0.062
adows+0.060
ying+0.059
os+0.057
Neg logit contributions
 gust-0.044
 mouth-0.043
-0.039
sed-0.039
 stem-0.038
Token a
Token ID257
Feature activation+1.25e-02

Pos logit contributions
 gust+0.042
 mouth+0.041
sed+0.040
 stem+0.039
+0.038
Neg logit contributions
and-0.059
ying-0.056
as-0.055
adows-0.054
os-0.053
Token huge
Token ID3236
Feature activation-2.61e-02

Pos logit contributions
sed+0.024
+0.023
 gust+0.023
 cheeks+0.023
 stem+0.022
Neg logit contributions
and-0.019
as-0.018
udd-0.018
ying-0.017
 outage-0.017
Token mess
Token ID2085
Feature activation-1.34e-01

Pos logit contributions
and+0.042
udd+0.041
 sorts+0.040
ying+0.040
as+0.039
Neg logit contributions
sed-0.043
-0.043
 gust-0.042
 stem-0.039
 cheeks-0.038
Token and
Token ID290
Feature activation-3.42e-02

Pos logit contributions
adows+0.296
and+0.283
inker+0.277
as+0.274
asc+0.272
Neg logit contributions
 gust-0.159
-0.154
 mouth-0.152
 cheeks-0.144
 tongue-0.139
Token made
Token ID925
Feature activation-2.93e-02

Pos logit contributions
and+0.076
as+0.073
adows+0.073
ying+0.073
inker+0.069
Neg logit contributions
 gust-0.046
 mouth-0.044
sed-0.043
 stem-0.041
 cheeks-0.040
Token him
Token ID683
Feature activation+4.10e-02

Pos logit contributions
and+0.056
as+0.055
ying+0.054
adows+0.054
os+0.051
Neg logit contributions
 gust-0.047
 mouth-0.045
sed-0.043
-0.043
 stem-0.042
Token clean
Token ID3424
Feature activation-6.54e-02

Pos logit contributions
 gust+0.056
 mouth+0.055
 stem+0.050
 cheeks+0.050
 tongue+0.050
Neg logit contributions
and-0.088
ying-0.087
as-0.087
adows-0.084
inker-0.082
Token it
Token ID340
Feature activation+1.73e-02

Pos logit contributions
as+0.143
and+0.138
adows+0.133
ibr+0.130
ying+0.129
Neg logit contributions
 mouth-0.095
 gust-0.091
 stem-0.086
 cheeks-0.085
 tongue-0.084
Token better
Token ID1365
Feature activation+2.81e-03

Pos logit contributions
and+0.067
as+0.067
inker+0.066
os+0.066
asc+0.063
Neg logit contributions
 gust-0.048
 mouth-0.045
 stem-0.041
-0.040
 tum-0.040
Token.
Token ID13
Feature activation-1.22e-02

Pos logit contributions
 mouth+0.004
+0.003
 cheeks+0.003
 stem+0.003
 gust+0.003
Neg logit contributions
and-0.006
as-0.006
asc-0.006
inker-0.006
ying-0.006
Token
Token ID198
Feature activation-2.44e-02

Pos logit contributions
and+0.021
ying+0.021
as+0.021
adows+0.020
pt+0.020
Neg logit contributions
 mouth-0.020
 gust-0.020
 stem-0.020
-0.019
sed-0.019
Token
Token ID198
Feature activation-2.41e-02

Pos logit contributions
ying+0.060
adows+0.059
as+0.059
and+0.059
asc+0.058
Neg logit contributions
 gust-0.020
-0.020
 mouth-0.019
 stem-0.019
 top-0.016
TokenJ
Token ID41
Feature activation-7.09e-02

Pos logit contributions
ying+0.048
adows+0.047
asc+0.046
and+0.045
pt+0.045
Neg logit contributions
 gust-0.033
 stem-0.032
-0.030
 mouth-0.030
 top-0.029
Tokenos
Token ID418
Feature activation-1.89e-01

Pos logit contributions
 upstream+0.078
beh+0.076
 responsibly+0.075
 outage+0.074
ying+0.074
Neg logit contributions
 stem-0.137
 mouth-0.136
 gust-0.133
 cheeks-0.133
sed-0.132
Tokenie
Token ID494
Feature activation-4.06e-02

Pos logit contributions
and+0.361
as+0.321
adows+0.316
 outage+0.311
inker+0.310
Neg logit contributions
 gust-0.306
 stem-0.296
 cheeks-0.293
 mouth-0.283
-0.281
Token took
Token ID1718
Feature activation-2.38e-02

Pos logit contributions
and+0.081
ying+0.076
as+0.076
adows+0.074
inker+0.071
Neg logit contributions
 gust-0.063
sed-0.061
 mouth-0.060
 stem-0.060
-0.059
Token the
Token ID262
Feature activation-5.79e-03

Pos logit contributions
and+0.047
ying+0.047
 outage+0.046
as+0.046
adows+0.045
Neg logit contributions
 gust-0.035
 mouth-0.034
 stem-0.032
 cheeks-0.032
-0.031
Token medicine
Token ID9007
Feature activation+1.85e-02

Pos logit contributions
and+0.011
as+0.011
ying+0.010
adows+0.010
udd+0.010
Neg logit contributions
 gust-0.009
sed-0.009
-0.009
 stem-0.009
 mouth-0.009
Token and
Token ID290
Feature activation-3.13e-02

Pos logit contributions
 mouth+0.027
 gust+0.026
+0.026
 stem+0.026
sed+0.024
Neg logit contributions
asc-0.037
and-0.037
inker-0.036
adows-0.036
as-0.036
Token store
Token ID3650
Feature activation-5.07e-03

Pos logit contributions
 gust+0.033
sed+0.032
 stem+0.032
+0.031
 mouth+0.028
Neg logit contributions
and-0.051
as-0.048
iam-0.048
washer-0.048
ying-0.046
Token was
Token ID373
Feature activation-1.42e-02

Pos logit contributions
and+0.010
as+0.009
ying+0.009
os+0.009
adows+0.009
Neg logit contributions
 gust-0.008
 stem-0.008
-0.008
sed-0.008
 mouth-0.007
Token full
Token ID1336
Feature activation+4.64e-03

Pos logit contributions
and+0.026
as+0.025
ying+0.025
adows+0.024
os+0.023
Neg logit contributions
 gust-0.024
sed-0.023
 stem-0.023
 mouth-0.023
-0.023
Token of
Token ID286
Feature activation-9.53e-03

Pos logit contributions
 gust+0.009
 mouth+0.009
sed+0.009
 stem+0.009
+0.009
Neg logit contributions
and-0.007
as-0.007
ying-0.007
adows-0.007
inker-0.006
Token all
Token ID477
Feature activation-9.32e-02

Pos logit contributions
and+0.018
as+0.017
ying+0.017
adows+0.017
 sorts+0.016
Neg logit contributions
 gust-0.015
sed-0.015
-0.015
 stem-0.014
 mouth-0.014
Token kinds
Token ID6982
Feature activation-1.39e-01

Pos logit contributions
and+0.181
inker+0.181
adows+0.181
as+0.175
ibr+0.173
Neg logit contributions
 mouth-0.144
 grape-0.139
 gust-0.135
 stem-0.126
 gu-0.126
Token of
Token ID286
Feature activation-1.04e-02

Pos logit contributions
and+0.214
adows+0.211
ying+0.205
as+0.203
inker+0.202
Neg logit contributions
 gust-0.268
 mouth-0.257
-0.253
sed-0.251
 cheeks-0.249
Token snacks
Token ID28487
Feature activation-1.09e-02

Pos logit contributions
and+0.019
ying+0.018
as+0.018
adows+0.017
 sorts+0.017
Neg logit contributions
sed-0.018
 gust-0.018
-0.017
 stem-0.017
 mouth-0.016
Token,
Token ID11
Feature activation-5.47e-03

Pos logit contributions
and+0.023
adows+0.022
as+0.022
inker+0.021
ying+0.021
Neg logit contributions
 mouth-0.015
 gust-0.015
-0.015
sed-0.015
 stem-0.014
Token some
Token ID617
Feature activation-4.18e-02

Pos logit contributions
and+0.011
as+0.010
ying+0.010
adows+0.010
os+0.009
Neg logit contributions
 gust-0.009
sed-0.008
 stem-0.008
 mouth-0.008
-0.008
Token were
Token ID547
Feature activation-2.92e-02

Pos logit contributions
and+0.082
as+0.079
ying+0.077
adows+0.075
pt+0.073
Neg logit contributions
sed-0.064
 gust-0.064
-0.062
 stem-0.062
 mouth-0.059