TOP ACTIVATIONS
MAX = 4.696

was very pleased with her honest and hard work . <|endoftext|>
in excitement . M um said to C eline
new hair brush on her long hair .
. But they did not have enough money to cover all
But Tom was too busy having fun and wanted to stay
life , with or without y ummy Ice - C ream
and fell onto the floor . He was feeling very ashamed
mom asked him why he had worked so hard to clean
store . Mama was looking for something and Ros y was
he remembered that his mom had some cleaning spray .
her blocks , but her brother wanted to play with them
left . M um and Sister were very
to walk ." M um and Dad told Little
told Louise that the oil had run out very quickly .
at her dad with an pleading expression on her face and
to prevent her mom from having to clean up the mud
smell and was filled with y ucky water .
go outside ." M um and Dad smiled and
She remembered how the lady had helped her and decided to
, Tommy 's trumpet breaks . He tries to blow ,

MIN ACTIVATIONS
MIN = -5.574

They watched the moon together until they fell asleep
. She would fly up and up , then come back
would fly over the hills and forests . One day ,
adventure . He watched the birds flying in the sky and
friends . They watch the birds and the flowers . They
Tom watched it go higher and higher until he could not
dad would watch the baby birds grow from a safe distance
up and down . The bird sits on the middle and
the other birds . The birds sing and thank the man
every night to watch the moon together . They became the
They watched the comet for a while , until
scary !" They swing higher and higher , pretending to fly
, Lily would visit the beetle every day and make sure
stars . She would watch them tw inkle in the sky
. He would make the rocket fly high in the sky
, I can make the moon go around the earth !"
the little girl visited the dragon every day and they would
. They would watch the planes take off and land .
dad . They would watch planes take off and land .
They both drew frogs and colored them . They

INTERVAL 2.128 - 4.696
CONTAINS 1.036%

� � I 'm not hungry , I want to sing
him that he had been naughty . They told him that
mom wrapped Lucy in her warm blanket and gave her a
she said , " I have a y ummy muff in
what she wanted to do to make her room better .

INTERVAL -0.439 - 2.128
CONTAINS 67.992%

Mom my , my thumb hurts ." His mom
your hat . It is very warm and cozy . Thank
. From then on , the friend visited the drum every
and didn 't pay attention to what he was doing .
playing in the garden . He loved to dig holes with

INTERVAL -3.007 - -0.439
CONTAINS 30.791%

. M ummy asked Tim my to pass her a h
continue the fun . <|endoftext|> Once upon a time there was
ie . The cloth is soaked in cold water . It
<|endoftext|> Once upon a time , there were two friends called
and less clumsy . And he learned that it 's not

INTERVAL -5.574 - -3.007
CONTAINS 0.181%

. You can put the butter on it . And then
it . She followed the butterfly to a pond where she
" Mom , can we come back to the park tomorrow
the bird soar above the trees and he wanted to fly
it . He asked the bird if he could try it
Token was
Token ID373
Feature activation+1.06e+00

Pos logit contributions
 Sky+0.001
fly+0.001
 tweet+0.001
 Saturn+0.001
 globe+0.001
Neg logit contributions
gh-0.001
 purpose-0.001
array-0.001
 behaviour-0.001
 Sister-0.001
Token very
Token ID845
Feature activation+1.35e+00

Pos logit contributions
gh+0.061
 behaviour+0.058
 purpose+0.058
array+0.057
urance+0.056
Neg logit contributions
 Sky-0.030
fly-0.028
 higher-0.027
 below-0.026
 tweet-0.026
Token pleased
Token ID10607
Feature activation+5.46e-02

Pos logit contributions
gh+0.087
array+0.084
 purpose+0.082
urance+0.082
 behaviour+0.081
Neg logit contributions
 Sky-0.028
 higher-0.025
 together-0.025
fly-0.025
 below-0.024
Token with
Token ID351
Feature activation+8.49e-01

Pos logit contributions
gh+0.003
 purpose+0.003
array+0.003
 behaviour+0.003
 Sister+0.003
Neg logit contributions
 Sky-0.002
fly-0.002
 tweet-0.001
 Saturn-0.001
 globe-0.001
Token her
Token ID607
Feature activation+1.83e+00

Pos logit contributions
gh+0.043
array+0.040
 behaviour+0.038
 purpose+0.038
omm+0.038
Neg logit contributions
 Sky-0.033
fly-0.030
 tweet-0.028
 Saturn-0.027
 Jupiter-0.027
Token honest
Token ID5508
Feature activation+4.70e+00

Pos logit contributions
gh+0.095
array+0.086
scar+0.084
David+0.084
omm+0.082
Neg logit contributions
fly-0.059
 Sky-0.059
 together-0.058
 around-0.057
 tweet-0.057
Token and
Token ID290
Feature activation+1.58e-01

Pos logit contributions
gh+0.220
liv+0.197
uf+0.191
omm+0.191
scar+0.191
Neg logit contributions
 flap-0.149
 around-0.147
 for-0.143
 and-0.142
 fly-0.140
Token hard
Token ID1327
Feature activation+2.81e+00

Pos logit contributions
gh+0.007
 purpose+0.006
array+0.006
 behaviour+0.006
omm+0.006
Neg logit contributions
 Sky-0.007
fly-0.007
 tweet-0.006
 Saturn-0.006
 globe-0.006
Token work
Token ID670
Feature activation+2.44e+00

Pos logit contributions
gh+0.149
uf+0.146
omm+0.144
David+0.140
scar+0.133
Neg logit contributions
 around-0.084
 together-0.079
 down-0.078
 below-0.075
 on-0.073
Token.
Token ID13
Feature activation-3.00e-01

Pos logit contributions
uf+0.135
array+0.130
omm+0.129
 mischief+0.128
gh+0.128
Neg logit contributions
 together-0.074
 around-0.069
 Sky-0.068
 below-0.065
fly-0.064
Token<|endoftext|>
Token ID50256
Feature activation-1.01e+00

Pos logit contributions
fly+0.011
 Sky+0.011
 tweet+0.010
 globe+0.009
 higher+0.009
Neg logit contributions
gh-0.015
 purpose-0.014
omm-0.014
 Sister-0.014
array-0.014
Token in
Token ID287
Feature activation+1.43e+00

Pos logit contributions
 Sky+0.005
fly+0.005
 tweet+0.005
 Saturn+0.004
 globe+0.004
Neg logit contributions
gh-0.008
 purpose-0.008
array-0.008
 behaviour-0.008
 Sister-0.008
Token excitement
Token ID14067
Feature activation+1.81e+00

Pos logit contributions
gh+0.075
omm+0.071
scar+0.068
addy+0.066
array+0.066
Neg logit contributions
 Sky-0.051
 tweet-0.049
 below-0.047
fly-0.047
 higher-0.047
Token.
Token ID13
Feature activation+2.37e+00

Pos logit contributions
gh+0.095
array+0.091
omm+0.090
 mischief+0.088
scar+0.088
Neg logit contributions
fly-0.059
 Sky-0.059
 below-0.055
 together-0.052
 higher-0.050
Token
Token ID198
Feature activation+1.44e+00

Pos logit contributions
gh+0.162
 purpose+0.146
 behaviour+0.146
 supposed+0.140
 consequences+0.139
Neg logit contributions
 Sky-0.043
 below-0.043
 down-0.039
 higher-0.039
als-0.038
Token
Token ID198
Feature activation+2.55e+00

Pos logit contributions
gh+0.093
 behaviour+0.092
array+0.090
urance+0.089
 purpose+0.089
Neg logit contributions
 Sky-0.031
fly-0.025
 Saturn-0.022
 below-0.022
 higher-0.021
TokenM
Token ID44
Feature activation+4.29e+00

Pos logit contributions
 behaviour+0.141
gh+0.140
urance+0.137
 purpose+0.133
 consequences+0.132
Neg logit contributions
 higher-0.066
 together-0.062
 Sky-0.062
 south-0.061
 below-0.058
Tokenum
Token ID388
Feature activation+3.92e-01

Pos logit contributions
gh+0.197
array+0.181
 purpose+0.178
 mischief+0.177
urance+0.177
Neg logit contributions
als-0.141
 below-0.132
ite-0.128
ocular-0.116
ls-0.115
Token said
Token ID531
Feature activation+1.57e+00

Pos logit contributions
gh+0.020
 purpose+0.018
array+0.018
 behaviour+0.017
urance+0.017
Neg logit contributions
fly-0.015
 Sky-0.015
 tweet-0.013
 Saturn-0.012
 below-0.012
Token to
Token ID284
Feature activation+3.05e+00

Pos logit contributions
array+0.089
gh+0.087
 purpose+0.085
urance+0.085
 mischief+0.084
Neg logit contributions
 Sky-0.048
 tweet-0.043
fly-0.042
als-0.038
 higher-0.038
Token C
Token ID327
Feature activation+1.65e+00

Pos logit contributions
gh+0.147
 purpose+0.136
array+0.135
 behaviour+0.134
 policy+0.130
Neg logit contributions
 tweet-0.100
 Sky-0.099
 Flor-0.095
 Saturn-0.094
 fly-0.093
Tokeneline
Token ID4470
Feature activation-3.83e-01

Pos logit contributions
gh+0.066
 purpose+0.060
 mischief+0.059
 punishment+0.059
 attempts+0.058
Neg logit contributions
fly-0.074
 below-0.072
 Sky-0.072
 globe-0.069
als-0.069
Token new
Token ID649
Feature activation+1.43e+00

Pos logit contributions
gh+0.069
array+0.067
urance+0.064
David+0.064
 behaviour+0.063
Neg logit contributions
 Sky-0.049
 tweet-0.043
fly-0.043
 flap-0.042
 higher-0.041
Token hair
Token ID4190
Feature activation+4.76e-01

Pos logit contributions
gh+0.059
array+0.057
urance+0.055
 forgotten+0.054
David+0.054
Neg logit contributions
 Sky-0.062
fly-0.059
 tweet-0.058
 flap-0.057
 higher-0.056
Tokenbrush
Token ID32680
Feature activation+8.99e-01

Pos logit contributions
gh+0.022
array+0.021
 purpose+0.020
 mischief+0.020
 Sister+0.020
Neg logit contributions
fly-0.019
 Sky-0.019
 tweet-0.017
 globe-0.016
 below-0.016
Token on
Token ID319
Feature activation+1.82e+00

Pos logit contributions
gh+0.043
omm+0.041
array+0.041
 mischief+0.040
 Sister+0.040
Neg logit contributions
fly-0.034
 Sky-0.033
 higher-0.029
 tweet-0.029
 below-0.029
Token her
Token ID607
Feature activation+1.52e+00

Pos logit contributions
 behaviour+0.110
gh+0.109
array+0.107
urance+0.106
scar+0.104
Neg logit contributions
 Sky-0.047
 Saturn-0.041
 Jupiter-0.036
 Bloom-0.036
 Mars-0.035
Token long
Token ID890
Feature activation+4.22e+00

Pos logit contributions
array+0.066
gh+0.065
David+0.063
scar+0.063
 Sister+0.060
Neg logit contributions
 Sky-0.062
fly-0.053
 flap-0.053
 globe-0.052
 tweet-0.052
Token hair
Token ID4190
Feature activation+9.97e-01

Pos logit contributions
Enough+0.226
scar+0.210
 supposed+0.208
David+0.208
udging+0.207
Neg logit contributions
 flap-0.127
 Sky-0.119
 fly-0.118
 soar-0.114
 tweet-0.107
Token.
Token ID13
Feature activation-9.90e-02

Pos logit contributions
gh+0.047
 mischief+0.045
array+0.045
 Sister+0.044
 behaviour+0.044
Neg logit contributions
fly-0.037
 Sky-0.037
 below-0.032
 globe-0.032
als-0.032
Token
Token ID220
Feature activation-1.20e+00

Pos logit contributions
fly+0.004
 Sky+0.003
 tweet+0.003
 globe+0.003
 Saturn+0.003
Neg logit contributions
gh-0.005
 purpose-0.005
array-0.005
 Sister-0.005
omm-0.005
Token
Token ID198
Feature activation-1.40e+00

Pos logit contributions
fly+0.050
 Sky+0.046
 tweet+0.045
flies+0.042
 globe+0.041
Neg logit contributions
gh-0.056
 purpose-0.051
omm-0.050
addy-0.050
em-0.049
Token
Token ID198
Feature activation+6.37e-01

Pos logit contributions
fly+0.066
 Sky+0.058
 tweet+0.058
flies+0.057
als+0.052
Neg logit contributions
gh-0.061
 purpose-0.055
addy-0.054
array-0.052
em-0.052
Token.
Token ID13
Feature activation+1.84e+00

Pos logit contributions
omm+0.214
addy+0.208
omo+0.201
innie+0.199
scar+0.196
Neg logit contributions
 together-0.091
 around-0.084
 below-0.078
 on-0.075
 higher-0.075
Token But
Token ID887
Feature activation+3.60e-01

Pos logit contributions
gh+0.087
 behaviour+0.072
illy+0.070
addy+0.069
 purpose+0.068
Neg logit contributions
 Sky-0.082
 below-0.073
fly-0.071
 Jupiter-0.070
 together-0.069
Token they
Token ID484
Feature activation+6.92e-01

Pos logit contributions
gh+0.013
 purpose+0.011
array+0.011
 behaviour+0.011
omm+0.011
Neg logit contributions
 Sky-0.019
fly-0.019
 tweet-0.017
 below-0.016
 globe-0.016
Token did
Token ID750
Feature activation+2.49e+00

Pos logit contributions
gh+0.033
 purpose+0.031
array+0.030
 mischief+0.029
omm+0.029
Neg logit contributions
fly-0.027
 Sky-0.027
 tweet-0.024
 below-0.024
 globe-0.023
Token not
Token ID407
Feature activation+2.47e+00

Pos logit contributions
omm+0.132
array+0.131
illy+0.130
kids+0.129
child+0.128
Neg logit contributions
 together-0.078
 higher-0.069
 Sky-0.066
 closer-0.062
 soar-0.061
Token have
Token ID423
Feature activation+4.14e+00

Pos logit contributions
omm+0.148
addy+0.140
gh+0.139
liv+0.137
innie+0.135
Neg logit contributions
fly-0.067
 fly-0.066
 flap-0.062
 soar-0.061
 tweet-0.061
Token enough
Token ID1576
Feature activation+3.21e+00

Pos logit contributions
omm+0.252
innie+0.236
addy+0.225
array+0.224
urance+0.222
Neg logit contributions
 together-0.104
 the-0.096
 on-0.092
 higher-0.090
 all-0.089
Token money
Token ID1637
Feature activation+2.44e+00

Pos logit contributions
gh+0.156
omm+0.150
scar+0.144
liv+0.144
innie+0.143
Neg logit contributions
 together-0.135
 below-0.123
 higher-0.118
 around-0.118
 on-0.117
Token to
Token ID284
Feature activation+3.28e+00

Pos logit contributions
array+0.137
omm+0.137
gh+0.137
 mischief+0.132
urance+0.131
Neg logit contributions
 below-0.063
 together-0.062
fly-0.062
 higher-0.061
 Sky-0.060
Token cover
Token ID3002
Feature activation+4.50e-01

Pos logit contributions
omm+0.183
gh+0.169
innie+0.168
addy+0.162
scar+0.160
Neg logit contributions
 fly-0.103
 soar-0.097
 flap-0.091
 below-0.091
 together-0.089
Token all
Token ID477
Feature activation+6.47e-01

Pos logit contributions
gh+0.025
array+0.023
 purpose+0.023
 behaviour+0.022
omm+0.022
Neg logit contributions
 Sky-0.014
fly-0.014
 tweet-0.012
 globe-0.012
 below-0.012
Token But
Token ID887
Feature activation+1.86e+00

Pos logit contributions
gh+0.204
 behaviour+0.189
 illness+0.184
oken+0.184
 supposed+0.181
Neg logit contributions
 down-0.068
 below-0.063
 higher-0.061
 together-0.060
 Sky-0.060
Token Tom
Token ID4186
Feature activation+2.80e+00

Pos logit contributions
gh+0.096
 mischief+0.092
 purpose+0.091
urance+0.091
array+0.088
Neg logit contributions
 Sky-0.058
 together-0.058
 higher-0.057
 down-0.054
 below-0.053
Token was
Token ID373
Feature activation+2.43e+00

Pos logit contributions
 purpose+0.146
 mischief+0.141
urance+0.139
array+0.135
gh+0.133
Neg logit contributions
 around-0.083
fly-0.082
bug-0.080
 Sky-0.080
 together-0.075
Token too
Token ID1165
Feature activation+3.14e+00

Pos logit contributions
urance+0.141
 purpose+0.141
 behaviour+0.141
gh+0.139
kids+0.138
Neg logit contributions
 around-0.064
 higher-0.061
 together-0.056
 on-0.056
 closer-0.056
Token busy
Token ID8179
Feature activation+3.42e+00

Pos logit contributions
gh+0.200
 occasions+0.194
urance+0.186
kids+0.180
child+0.170
Neg logit contributions
 together-0.076
 high-0.074
 higher-0.073
 below-0.072
 closer-0.071
Token having
Token ID1719
Feature activation+4.14e+00

Pos logit contributions
omm+0.183
uf+0.180
 Sister+0.176
gh+0.171
David+0.170
Neg logit contributions
 together-0.109
 around-0.107
 below-0.094
 higher-0.093
 on-0.093
Token fun
Token ID1257
Feature activation+3.20e+00

Pos logit contributions
uf+0.267
omm+0.257
urance+0.253
scar+0.250
ailed+0.248
Neg logit contributions
 together-0.084
 around-0.076
 higher-0.072
 on-0.069
 all-0.065
Token and
Token ID290
Feature activation+2.16e+00

Pos logit contributions
uf+0.195
 Sister+0.193
omm+0.191
David+0.185
child+0.181
Neg logit contributions
 together-0.095
 around-0.081
 on-0.067
 under-0.066
 higher-0.064
Token wanted
Token ID2227
Feature activation+2.69e+00

Pos logit contributions
gh+0.135
omm+0.133
uf+0.130
urance+0.130
 purpose+0.128
Neg logit contributions
 Sky-0.042
 together-0.040
 around-0.039
 higher-0.038
 down-0.035
Token to
Token ID284
Feature activation+2.70e+00

Pos logit contributions
urance+0.181
omm+0.175
uf+0.174
ailed+0.173
gh+0.172
Neg logit contributions
 around-0.048
 together-0.046
 higher-0.045
 closer-0.044
 Sky-0.041
Token stay
Token ID2652
Feature activation+1.53e+00

Pos logit contributions
omm+0.157
gh+0.156
kids+0.149
uf+0.149
child+0.148
Neg logit contributions
 fly-0.071
 soar-0.068
 flap-0.066
 Sky-0.057
 around-0.055
Token life
Token ID1204
Feature activation+6.03e-02

Pos logit contributions
gh+0.062
omm+0.061
array+0.060
urance+0.060
 behaviour+0.058
Neg logit contributions
 Sky-0.035
fly-0.032
 tweet-0.029
 below-0.029
 higher-0.029
Token,
Token ID11
Feature activation-8.23e-01

Pos logit contributions
gh+0.003
 purpose+0.003
array+0.003
 behaviour+0.003
omm+0.003
Neg logit contributions
 Sky-0.002
fly-0.002
 tweet-0.002
 globe-0.002
 Saturn-0.002
Token with
Token ID351
Feature activation+1.19e+00

Pos logit contributions
fly+0.034
 Sky+0.034
 tweet+0.030
 globe+0.029
flies+0.028
Neg logit contributions
gh-0.038
 purpose-0.035
 forgotten-0.035
 Sister-0.034
 behaviour-0.034
Token or
Token ID393
Feature activation+1.08e-01

Pos logit contributions
gh+0.062
array+0.060
omm+0.058
urance+0.057
 purpose+0.056
Neg logit contributions
 Sky-0.040
fly-0.038
 tweet-0.034
 Saturn-0.034
 Jupiter-0.033
Token without
Token ID1231
Feature activation+2.64e+00

Pos logit contributions
gh+0.005
 purpose+0.004
array+0.004
 behaviour+0.004
omm+0.004
Neg logit contributions
 Sky-0.005
fly-0.005
 tweet-0.004
 Saturn-0.004
 globe-0.004
Token y
Token ID331
Feature activation+4.10e+00

Pos logit contributions
omm+0.118
gh+0.116
liv+0.109
addy+0.108
innie+0.107
Neg logit contributions
 Sky-0.101
fly-0.101
 tweet-0.096
 higher-0.095
 below-0.092
Tokenummy
Token ID13513
Feature activation+2.94e+00

Pos logit contributions
liv+0.182
 virus+0.179
David+0.178
child+0.177
 forgotten+0.174
Neg logit contributions
als-0.127
 below-0.118
 higher-0.115
 Sky-0.110
 down-0.109
Token Ice
Token ID6663
Feature activation+2.44e+00

Pos logit contributions
gh+0.151
omm+0.151
liv+0.150
David+0.147
array+0.143
Neg logit contributions
 around-0.086
 below-0.084
als-0.082
 on-0.080
 higher-0.080
Token-
Token ID12
Feature activation+1.03e+00

Pos logit contributions
array+0.111
urance+0.109
gh+0.106
 mischief+0.105
ailed+0.101
Neg logit contributions
fly-0.093
 globe-0.088
 Sky-0.088
 below-0.087
als-0.087
TokenC
Token ID34
Feature activation+3.11e+00

Pos logit contributions
gh+0.045
array+0.043
urance+0.042
 purpose+0.041
omm+0.041
Neg logit contributions
fly-0.043
 Sky-0.043
 tweet-0.039
als-0.038
 below-0.038
Tokenream
Token ID1476
Feature activation+1.91e+00

Pos logit contributions
 virus+0.114
urance+0.114
 expression+0.112
array+0.110
 babys+0.110
Neg logit contributions
fly-0.121
ites-0.117
als-0.115
 Sky-0.112
flies-0.110
Token and
Token ID290
Feature activation+1.57e+00

Pos logit contributions
gh+0.034
 purpose+0.032
 behaviour+0.032
array+0.031
urance+0.031
Neg logit contributions
 Sky-0.012
fly-0.012
 tweet-0.010
 higher-0.010
 below-0.009
Token fell
Token ID3214
Feature activation+7.14e-01

Pos logit contributions
gh+0.078
omm+0.077
 purpose+0.076
urance+0.075
 behaviour+0.075
Neg logit contributions
 Sky-0.055
fly-0.049
 higher-0.049
 flap-0.048
 tweet-0.047
Token onto
Token ID4291
Feature activation+2.23e+00

Pos logit contributions
gh+0.046
array+0.043
 purpose+0.043
 behaviour+0.043
omm+0.042
Neg logit contributions
 Sky-0.017
fly-0.017
 higher-0.014
 tweet-0.014
 below-0.013
Token the
Token ID262
Feature activation+4.90e-01

Pos logit contributions
gh+0.158
 behaviour+0.153
 supposed+0.149
urance+0.146
array+0.145
Neg logit contributions
 Sky-0.040
 Saturn-0.029
 Jupiter-0.025
 higher-0.025
 below-0.023
Token floor
Token ID4314
Feature activation+2.27e+00

Pos logit contributions
gh+0.023
 purpose+0.021
array+0.021
omm+0.021
 behaviour+0.021
Neg logit contributions
 Sky-0.020
fly-0.019
 tweet-0.017
 globe-0.017
 below-0.017
Token.
Token ID13
Feature activation+4.04e+00

Pos logit contributions
kids+0.126
 Sister+0.126
gh+0.122
omm+0.121
child+0.120
Neg logit contributions
 below-0.067
 higher-0.066
 Sky-0.066
fly-0.065
 around-0.060
Token He
Token ID679
Feature activation+1.02e+00

Pos logit contributions
gh+0.254
 behaviour+0.241
illy+0.240
 supposed+0.236
 occasions+0.236
Neg logit contributions
 down-0.074
 Sky-0.066
 around-0.059
 below-0.059
 higher-0.050
Token was
Token ID373
Feature activation+1.22e+00

Pos logit contributions
gh+0.044
omm+0.042
 purpose+0.042
urance+0.041
 behaviour+0.040
Neg logit contributions
 Sky-0.044
fly-0.042
 tweet-0.039
 flap-0.038
 higher-0.037
Token feeling
Token ID4203
Feature activation+1.18e+00

Pos logit contributions
gh+0.061
 behaviour+0.057
 purpose+0.057
kids+0.056
array+0.056
Neg logit contributions
 Sky-0.045
fly-0.043
 higher-0.040
 below-0.038
 flap-0.038
Token very
Token ID845
Feature activation+1.52e+00

Pos logit contributions
gh+0.062
kids+0.057
array+0.057
urance+0.057
 purpose+0.056
Neg logit contributions
 Sky-0.041
fly-0.038
 higher-0.038
 below-0.035
 tweet-0.035
Token ashamed
Token ID22461
Feature activation+1.76e+00

Pos logit contributions
gh+0.077
kids+0.075
child+0.073
em+0.072
omm+0.072
Neg logit contributions
 Sky-0.052
 higher-0.051
 below-0.049
 soar-0.047
fly-0.047
Token mom
Token ID1995
Feature activation+7.57e-01

Pos logit contributions
em+0.136
gh+0.136
array+0.132
David+0.128
 supposed+0.126
Neg logit contributions
 Sky-0.053
 fly-0.043
 rocket-0.043
 globe-0.042
 planet-0.041
Token asked
Token ID1965
Feature activation+5.28e-01

Pos logit contributions
gh+0.035
 purpose+0.034
 behaviour+0.032
array+0.031
urance+0.031
Neg logit contributions
 Sky-0.031
fly-0.031
 tweet-0.028
 below-0.026
 Saturn-0.026
Token him
Token ID683
Feature activation-2.92e-01

Pos logit contributions
gh+0.024
array+0.021
 purpose+0.021
addy+0.020
omm+0.020
Neg logit contributions
 Sky-0.023
fly-0.022
 tweet-0.021
 below-0.019
 Saturn-0.019
Token why
Token ID1521
Feature activation+1.86e+00

Pos logit contributions
fly+0.012
 Sky+0.012
 tweet+0.011
 globe+0.010
 Saturn+0.010
Neg logit contributions
gh-0.014
 purpose-0.012
 behaviour-0.012
 Sister-0.012
omm-0.012
Token he
Token ID339
Feature activation+2.32e+00

Pos logit contributions
gh+0.119
array+0.115
urance+0.114
omm+0.112
 purpose+0.111
Neg logit contributions
 Sky-0.039
fly-0.035
 below-0.033
 together-0.030
 tweet-0.028
Token had
Token ID550
Feature activation+3.97e+00

Pos logit contributions
 purpose+0.129
gh+0.124
omm+0.120
array+0.118
 mischief+0.116
Neg logit contributions
fly-0.073
bug-0.062
 Sky-0.062
 below-0.060
 tweet-0.060
Token worked
Token ID3111
Feature activation+2.75e+00

Pos logit contributions
omm+0.207
gh+0.199
scar+0.193
urance+0.193
innie+0.190
Neg logit contributions
 around-0.114
 together-0.114
 on-0.113
 in-0.106
 down-0.105
Token so
Token ID523
Feature activation+2.05e+00

Pos logit contributions
urance+0.163
uf+0.162
omm+0.159
kids+0.159
child+0.155
Neg logit contributions
 together-0.075
 higher-0.069
 around-0.067
 down-0.057
 on-0.056
Token hard
Token ID1327
Feature activation+3.07e+00

Pos logit contributions
kids+0.106
gh+0.105
urance+0.104
array+0.104
child+0.101
Neg logit contributions
 higher-0.066
 together-0.064
 Sky-0.063
 below-0.062
 around-0.059
Token to
Token ID284
Feature activation+2.39e+00

Pos logit contributions
omm+0.185
gh+0.181
OW+0.174
uf+0.174
urance+0.172
Neg logit contributions
 together-0.075
 around-0.069
 on-0.063
 below-0.061
 down-0.060
Token clean
Token ID3424
Feature activation+1.32e+00

Pos logit contributions
omm+0.131
gh+0.130
kids+0.128
 behaviour+0.124
child+0.123
Neg logit contributions
 fly-0.067
 flap-0.066
 soar-0.064
 higher-0.059
 Sky-0.057
Token store
Token ID3650
Feature activation+2.01e+00

Pos logit contributions
gh+0.024
 behaviour+0.022
 purpose+0.021
urance+0.021
omm+0.021
Neg logit contributions
 Sky-0.020
fly-0.020
 tweet-0.018
 globe-0.017
 Saturn-0.017
Token.
Token ID13
Feature activation+1.01e-01

Pos logit contributions
gh+0.126
 purpose+0.116
 mischief+0.114
omm+0.113
 behaviour+0.112
Neg logit contributions
 together-0.051
fly-0.048
 below-0.048
 Sky-0.048
 around-0.042
Token Mama
Token ID40084
Feature activation+1.21e+00

Pos logit contributions
gh+0.005
 purpose+0.004
array+0.004
 behaviour+0.004
omm+0.004
Neg logit contributions
 Sky-0.004
fly-0.004
 tweet-0.004
 Saturn-0.003
 globe-0.003
Token was
Token ID373
Feature activation+2.08e+00

Pos logit contributions
gh+0.067
 purpose+0.063
 behaviour+0.059
 mischief+0.059
urance+0.059
Neg logit contributions
fly-0.041
 Sky-0.041
 tweet-0.034
 below-0.034
 globe-0.033
Token looking
Token ID2045
Feature activation+8.67e-01

Pos logit contributions
gh+0.116
 behaviour+0.113
 purpose+0.110
urance+0.109
OW+0.105
Neg logit contributions
 Sky-0.060
 together-0.059
 higher-0.057
 below-0.056
 around-0.054
Token for
Token ID329
Feature activation+3.93e+00

Pos logit contributions
gh+0.058
 purpose+0.053
array+0.052
urance+0.052
omm+0.052
Neg logit contributions
 Sky-0.018
fly-0.017
 below-0.015
 higher-0.015
 tweet-0.015
Token something
Token ID1223
Feature activation+2.30e+00

Pos logit contributions
urance+0.252
gh+0.252
omm+0.239
aned+0.237
istent+0.237
Neg logit contributions
 together-0.068
 flowers-0.064
 mountains-0.064
 Sky-0.062
 the-0.059
Token and
Token ID290
Feature activation+1.49e+00

Pos logit contributions
urance+0.129
omm+0.126
array+0.124
child+0.122
gh+0.122
Neg logit contributions
 Sky-0.067
 together-0.067
 around-0.065
 closer-0.064
 higher-0.064
Token Ros
Token ID10018
Feature activation+1.85e+00

Pos logit contributions
gh+0.081
 purpose+0.074
urance+0.073
omm+0.072
 behaviour+0.070
Neg logit contributions
 Sky-0.048
fly-0.047
 together-0.045
 tweet-0.043
 Saturn-0.042
Tokeny
Token ID88
Feature activation+2.75e+00

Pos logit contributions
gh+0.118
 purpose+0.115
 punishment+0.114
 mischief+0.113
omm+0.110
Neg logit contributions
fly-0.045
 Sky-0.035
flies-0.034
 together-0.030
als-0.030
Token was
Token ID373
Feature activation+2.10e+00

Pos logit contributions
gh+0.137
omm+0.134
 purpose+0.133
array+0.126
urance+0.125
Neg logit contributions
fly-0.098
 Sky-0.093
 below-0.087
 together-0.083
 around-0.082
Token he
Token ID339
Feature activation+1.36e+00

Pos logit contributions
gh+0.011
 purpose+0.010
array+0.010
 behaviour+0.010
omm+0.010
Neg logit contributions
 Sky-0.007
fly-0.007
 tweet-0.007
 Saturn-0.006
 globe-0.006
Token remembered
Token ID12086
Feature activation+1.80e+00

Pos logit contributions
 purpose+0.071
urance+0.068
gh+0.067
omm+0.065
 behaviour+0.065
Neg logit contributions
fly-0.047
 Sky-0.044
bug-0.041
 tweet-0.039
 closer-0.039
Token that
Token ID326
Feature activation+4.74e-01

Pos logit contributions
gh+0.116
omm+0.113
urance+0.112
kids+0.109
array+0.108
Neg logit contributions
 Sky-0.034
fly-0.032
 Jupiter-0.030
 together-0.030
 Saturn-0.028
Token his
Token ID465
Feature activation+1.51e+00

Pos logit contributions
gh+0.026
 purpose+0.024
omm+0.024
urance+0.024
array+0.024
Neg logit contributions
 Sky-0.015
fly-0.015
 tweet-0.013
 Saturn-0.013
 globe-0.013
Token mom
Token ID1995
Feature activation+1.03e+00

Pos logit contributions
gh+0.073
array+0.069
em+0.069
urance+0.067
omm+0.067
Neg logit contributions
 Sky-0.056
 globe-0.049
 tweet-0.048
 below-0.048
 deer-0.048
Token had
Token ID550
Feature activation+3.93e+00

Pos logit contributions
gh+0.055
 purpose+0.053
 mischief+0.050
omm+0.050
urance+0.049
Neg logit contributions
fly-0.036
 Sky-0.033
 tweet-0.030
bug-0.029
flies-0.028
Token some
Token ID617
Feature activation+3.37e+00

Pos logit contributions
omm+0.208
urance+0.196
uf+0.193
becca+0.186
kids+0.185
Neg logit contributions
 together-0.119
 on-0.112
 in-0.100
 fly-0.099
 fl-0.098
Token cleaning
Token ID12724
Feature activation+1.30e+00

Pos logit contributions
gh+0.172
omm+0.172
urance+0.169
David+0.168
oph+0.162
Neg logit contributions
 around-0.101
 fly-0.100
 flies-0.097
 on-0.097
 flap-0.097
Token spray
Token ID11662
Feature activation+1.16e+00

Pos logit contributions
gh+0.076
urance+0.073
omm+0.070
array+0.070
 behaviour+0.070
Neg logit contributions
fly-0.034
 Sky-0.034
 globe-0.030
 higher-0.030
 below-0.029
Token.
Token ID13
Feature activation+6.40e-01

Pos logit contributions
gh+0.062
urance+0.057
omm+0.057
array+0.057
 purpose+0.056
Neg logit contributions
fly-0.039
 Sky-0.038
 higher-0.034
 tweet-0.034
 below-0.033
Token
Token ID220
Feature activation-3.20e-01

Pos logit contributions
gh+0.035
 purpose+0.031
 behaviour+0.031
 forgotten+0.030
array+0.030
Neg logit contributions
 Sky-0.023
fly-0.021
 tweet-0.019
 below-0.018
 Saturn-0.018
Token her
Token ID607
Feature activation-6.17e-01

Pos logit contributions
gh+0.030
 behaviour+0.027
urance+0.027
 Sister+0.027
array+0.027
Neg logit contributions
 Sky-0.025
fly-0.023
 tweet-0.022
 Saturn-0.021
 higher-0.021
Token blocks
Token ID7021
Feature activation+2.42e-01

Pos logit contributions
fly+0.028
 Sky+0.027
 tweet+0.026
 Saturn+0.024
 Jupiter+0.024
Neg logit contributions
gh-0.026
 purpose-0.024
 behaviour-0.023
 forgotten-0.022
array-0.022
Token,
Token ID11
Feature activation-4.05e-01

Pos logit contributions
gh+0.011
 purpose+0.010
array+0.010
 behaviour+0.010
 Sister+0.010
Neg logit contributions
fly-0.010
 Sky-0.010
 tweet-0.009
 below-0.009
 globe-0.009
Token but
Token ID475
Feature activation-3.69e-01

Pos logit contributions
fly+0.017
 Sky+0.016
 tweet+0.014
 globe+0.014
flies+0.013
Neg logit contributions
gh-0.019
array-0.018
 purpose-0.017
 Sister-0.017
 behaviour-0.017
Token her
Token ID607
Feature activation+5.36e-01

Pos logit contributions
fly+0.016
 Sky+0.015
 tweet+0.014
 globe+0.014
flies+0.013
Neg logit contributions
gh-0.017
 purpose-0.015
array-0.015
 forgotten-0.015
omm-0.014
Token brother
Token ID3956
Feature activation+3.92e+00

Pos logit contributions
gh+0.026
array+0.024
 purpose+0.023
 Sister+0.023
urance+0.023
Neg logit contributions
 Sky-0.021
fly-0.020
 tweet-0.018
 globe-0.018
 Saturn-0.017
Token wanted
Token ID2227
Feature activation+1.87e+00

Pos logit contributions
 mischief+0.238
 purpose+0.229
oken+0.228
array+0.227
igsaw+0.225
Neg logit contributions
 Sky-0.077
 together-0.077
 around-0.077
fly-0.076
 could-0.067
Token to
Token ID284
Feature activation+1.47e+00

Pos logit contributions
urance+0.115
uf+0.112
gh+0.111
 behaviour+0.109
array+0.107
Neg logit contributions
 Sky-0.049
 higher-0.038
 together-0.038
fly-0.038
 Saturn-0.038
Token play
Token ID711
Feature activation-7.04e-01

Pos logit contributions
gh+0.085
 Sister+0.082
omm+0.082
 behaviour+0.081
array+0.079
Neg logit contributions
 Sky-0.038
 soar-0.038
fly-0.037
 fly-0.036
 flap-0.035
Token with
Token ID351
Feature activation+4.87e-01

Pos logit contributions
fly+0.027
 Sky+0.026
 tweet+0.023
flies+0.022
als+0.021
Neg logit contributions
gh-0.037
 purpose-0.034
array-0.032
 mischief-0.031
 forgotten-0.031
Token them
Token ID606
Feature activation+1.61e-02

Pos logit contributions
gh+0.024
 behaviour+0.022
array+0.022
 purpose+0.022
urance+0.022
Neg logit contributions
 Sky-0.018
fly-0.018
 tweet-0.016
 Saturn-0.015
 higher-0.015
Token left
Token ID1364
Feature activation+1.47e+00

Pos logit contributions
 Sky+0.014
fly+0.014
 tweet+0.013
 Saturn+0.012
 globe+0.012
Neg logit contributions
gh-0.013
 purpose-0.011
array-0.011
 forgotten-0.011
 behaviour-0.011
Token.
Token ID13
Feature activation+8.62e-01

Pos logit contributions
gh+0.086
omm+0.081
array+0.081
 purpose+0.079
urance+0.078
Neg logit contributions
 Sky-0.040
fly-0.039
 below-0.038
 globe-0.034
 together-0.034
Token
Token ID220
Feature activation+4.95e-01

Pos logit contributions
gh+0.052
 purpose+0.047
 behaviour+0.046
array+0.046
 forgotten+0.044
Neg logit contributions
 Sky-0.025
fly-0.023
 Saturn-0.020
 below-0.019
 globe-0.019
Token
Token ID198
Feature activation+3.27e-01

Pos logit contributions
gh+0.029
 purpose+0.027
array+0.027
 behaviour+0.027
 forgotten+0.026
Neg logit contributions
 Sky-0.015
fly-0.014
 tweet-0.012
 Saturn-0.011
 below-0.011
Token
Token ID198
Feature activation+1.21e+00

Pos logit contributions
gh+0.019
 purpose+0.018
array+0.018
 behaviour+0.017
omm+0.017
Neg logit contributions
 Sky-0.010
fly-0.009
 tweet-0.008
 Saturn-0.008
 globe-0.008
TokenM
Token ID44
Feature activation+3.92e+00

Pos logit contributions
gh+0.059
 behaviour+0.058
 purpose+0.057
array+0.057
urance+0.056
Neg logit contributions
 Sky-0.040
fly-0.038
 together-0.035
 Saturn-0.035
 globe-0.035
Tokenum
Token ID388
Feature activation+4.69e-01

Pos logit contributions
 blond+0.177
 blonde+0.176
gh+0.175
array+0.170
 purpose+0.169
Neg logit contributions
ite-0.098
s-0.096
 together-0.091
ites-0.090
fish-0.088
Token and
Token ID290
Feature activation+2.14e+00

Pos logit contributions
gh+0.027
 purpose+0.025
array+0.025
 behaviour+0.024
 mischief+0.024
Neg logit contributions
fly-0.014
 Sky-0.014
 tweet-0.012
 globe-0.011
 Saturn-0.011
Token Sister
Token ID26553
Feature activation+5.99e-01

Pos logit contributions
gh+0.110
 purpose+0.091
 supposed+0.090
array+0.087
 policy+0.085
Neg logit contributions
 Sky-0.076
 down-0.076
 deer-0.075
als-0.074
 below-0.073
Token were
Token ID547
Feature activation+5.49e-01

Pos logit contributions
gh+0.029
 purpose+0.027
array+0.026
 behaviour+0.025
omm+0.025
Neg logit contributions
 Sky-0.024
fly-0.024
 tweet-0.022
 globe-0.020
 below-0.020
Token very
Token ID845
Feature activation+3.38e-01

Pos logit contributions
gh+0.029
 purpose+0.027
array+0.027
 behaviour+0.027
omm+0.026
Neg logit contributions
 Sky-0.019
fly-0.018
 tweet-0.016
 below-0.016
 higher-0.015
Token to
Token ID284
Feature activation+1.51e+00

Pos logit contributions
gh+0.070
urance+0.066
omm+0.065
array+0.065
 behaviour+0.063
Neg logit contributions
 Sky-0.041
fly-0.038
 higher-0.035
 tweet-0.034
 Saturn-0.034
Token walk
Token ID2513
Feature activation-7.10e-01

Pos logit contributions
gh+0.083
omm+0.079
child+0.077
array+0.077
em+0.076
Neg logit contributions
fly-0.046
 Sky-0.046
 soar-0.044
 tweet-0.043
 flap-0.042
Token."
Token ID526
Feature activation+2.46e+00

Pos logit contributions
fly+0.024
 Sky+0.024
 tweet+0.022
 Saturn+0.020
 globe+0.019
Neg logit contributions
gh-0.038
 purpose-0.036
array-0.034
 forgotten-0.034
 behaviour-0.034
Token
Token ID198
Feature activation+1.62e+00

Pos logit contributions
gh+0.167
 supposed+0.149
 purpose+0.147
 behaviour+0.144
kids+0.142
Neg logit contributions
 together-0.046
 down-0.042
 Sky-0.042
 higher-0.038
 below-0.036
Token
Token ID198
Feature activation+2.40e+00

Pos logit contributions
gh+0.108
 behaviour+0.107
array+0.107
 purpose+0.104
 forgotten+0.103
Neg logit contributions
 Sky-0.028
fly-0.021
 together-0.020
 Fly-0.018
 Saturn-0.018
TokenM
Token ID44
Feature activation+3.89e+00

Pos logit contributions
urance+0.127
 behaviour+0.126
 supposed+0.125
gh+0.124
 purpose+0.122
Neg logit contributions
 together-0.063
 Sky-0.054
 higher-0.053
 closer-0.053
 Fly-0.051
Tokenum
Token ID388
Feature activation+8.80e-01

Pos logit contributions
 supposed+0.172
array+0.171
 mischief+0.170
 blonde+0.170
 virus+0.165
Neg logit contributions
ite-0.107
cast-0.100
 together-0.097
fish-0.095
 below-0.091
Token and
Token ID290
Feature activation+2.22e+00

Pos logit contributions
gh+0.051
 purpose+0.049
 mischief+0.047
array+0.046
 behaviour+0.046
Neg logit contributions
fly-0.027
 Sky-0.026
 tweet-0.023
 below-0.021
 globe-0.021
Token Dad
Token ID17415
Feature activation+9.60e-01

Pos logit contributions
gh+0.120
 purpose+0.106
 supposed+0.101
urance+0.100
 policy+0.098
Neg logit contributions
 together-0.075
 down-0.073
 closer-0.072
 below-0.072
 Sky-0.071
Token told
Token ID1297
Feature activation+2.58e+00

Pos logit contributions
gh+0.046
 purpose+0.046
 mischief+0.041
omm+0.041
array+0.041
Neg logit contributions
 Sky-0.037
fly-0.037
 tweet-0.035
 below-0.033
 flap-0.032
Token Little
Token ID7703
Feature activation+3.02e+00

Pos logit contributions
gh+0.132
 supposed+0.128
 policy+0.121
omm+0.121
 forgotten+0.119
Neg logit contributions
 together-0.082
 Sky-0.079
 around-0.076
 down-0.072
 Jupiter-0.071
Token told
Token ID1297
Feature activation-1.21e+00

Pos logit contributions
gh+0.036
 purpose+0.034
array+0.032
 behaviour+0.032
urance+0.032
Neg logit contributions
fly-0.032
 Sky-0.031
 tweet-0.028
 below-0.027
 Saturn-0.026
Token Louise
Token ID31774
Feature activation+4.68e-01

Pos logit contributions
fly+0.044
 globe+0.041
 Sky+0.038
 higher+0.038
 tweet+0.038
Neg logit contributions
gh-0.058
 purpose-0.055
 Sister-0.054
 behaviour-0.054
illy-0.053
Token that
Token ID326
Feature activation+3.30e-01

Pos logit contributions
gh+0.024
 purpose+0.022
array+0.022
omm+0.022
 behaviour+0.022
Neg logit contributions
 Sky-0.017
fly-0.017
 tweet-0.015
 Saturn-0.014
 Jupiter-0.013
Token the
Token ID262
Feature activation+1.98e+00

Pos logit contributions
gh+0.018
 purpose+0.016
array+0.016
omm+0.016
urance+0.016
Neg logit contributions
 Sky-0.011
fly-0.011
 tweet-0.010
 Saturn-0.009
 globe-0.009
Token oil
Token ID3056
Feature activation+2.24e+00

Pos logit contributions
gh+0.093
David+0.086
anked+0.081
oph+0.080
em+0.080
Neg logit contributions
 Sky-0.071
 flap-0.070
 deer-0.069
 globe-0.069
fly-0.066
Token had
Token ID550
Feature activation+3.89e+00

Pos logit contributions
gh+0.113
array+0.108
 purpose+0.106
 mischief+0.106
urance+0.105
Neg logit contributions
fly-0.079
 Sky-0.074
 globe-0.070
als-0.067
 below-0.066
Token run
Token ID1057
Feature activation+2.49e+00

Pos logit contributions
omm+0.169
array+0.160
gh+0.156
innie+0.156
urance+0.154
Neg logit contributions
 Sky-0.146
 fly-0.139
fly-0.138
 flap-0.135
 around-0.134
Token out
Token ID503
Feature activation+5.76e-01

Pos logit contributions
omm+0.162
addy+0.157
gh+0.156
innie+0.151
array+0.151
Neg logit contributions
 around-0.051
 higher-0.050
 Sky-0.047
fly-0.046
 together-0.044
Token very
Token ID845
Feature activation+1.63e+00

Pos logit contributions
gh+0.031
 purpose+0.029
omm+0.029
array+0.029
 Sister+0.028
Neg logit contributions
 Sky-0.019
fly-0.019
 tweet-0.017
 below-0.016
 Saturn-0.015
Token quickly
Token ID2952
Feature activation+3.14e+00

Pos logit contributions
gh+0.092
urance+0.091
omm+0.089
array+0.089
child+0.088
Neg logit contributions
 Sky-0.046
 higher-0.046
 below-0.043
fly-0.040
 closer-0.039
Token.
Token ID13
Feature activation-2.11e+00

Pos logit contributions
omm+0.183
gh+0.182
innie+0.171
scar+0.170
addy+0.170
Neg logit contributions
fly-0.077
 below-0.077
 around-0.075
 together-0.073
 Sky-0.073
Token at
Token ID379
Feature activation-1.10e+00

Pos logit contributions
fly+0.036
 Sky+0.034
 tweet+0.030
 globe+0.029
flies+0.027
Neg logit contributions
 forgotten-0.056
gh-0.056
array-0.055
 behaviour-0.055
 mischief-0.055
Token her
Token ID607
Feature activation+7.20e-01

Pos logit contributions
fly+0.046
 Sky+0.043
 tweet+0.042
 globe+0.040
flies+0.038
Neg logit contributions
gh-0.046
 purpose-0.046
addy-0.045
 Sister-0.045
 mischief-0.043
Token dad
Token ID9955
Feature activation+1.12e-01

Pos logit contributions
gh+0.032
array+0.028
omm+0.028
 purpose+0.028
urance+0.028
Neg logit contributions
 Sky-0.031
fly-0.031
 tweet-0.028
 below-0.028
 globe-0.028
Token with
Token ID351
Feature activation+1.33e+00

Pos logit contributions
gh+0.006
 purpose+0.005
array+0.005
 behaviour+0.005
omm+0.005
Neg logit contributions
 Sky-0.004
fly-0.004
 tweet-0.003
 Saturn-0.003
 globe-0.003
Token an
Token ID281
Feature activation+2.11e+00

Pos logit contributions
gh+0.070
omm+0.066
array+0.064
addy+0.063
 purpose+0.063
Neg logit contributions
 Sky-0.047
fly-0.043
 tweet-0.042
 flap-0.040
 below-0.038
Token pleading
Token ID30279
Feature activation+3.89e+00

Pos logit contributions
gh+0.093
 purpose+0.085
 Sister+0.083
array+0.082
 forgotten+0.082
Neg logit contributions
 Sky-0.081
 flap-0.078
fly-0.077
 tweet-0.077
 Jupiter-0.076
Token expression
Token ID5408
Feature activation+6.15e-01

Pos logit contributions
liv+0.172
array+0.172
scar+0.170
Enough+0.168
uf+0.166
Neg logit contributions
 flap-0.134
fly-0.130
 globe-0.127
 tweet-0.122
 Sky-0.119
Token on
Token ID319
Feature activation+1.15e+00

Pos logit contributions
gh+0.033
 purpose+0.031
array+0.031
omm+0.030
 mischief+0.030
Neg logit contributions
fly-0.021
 Sky-0.020
 tweet-0.018
 below-0.017
 higher-0.017
Token her
Token ID607
Feature activation+1.59e+00

Pos logit contributions
array+0.057
gh+0.056
 supposed+0.055
 forgotten+0.055
 behaviour+0.054
Neg logit contributions
 Sky-0.044
fly-0.040
 tweet-0.035
 Saturn-0.035
 below-0.034
Token face
Token ID1986
Feature activation+1.36e+00

Pos logit contributions
array+0.068
gh+0.065
omm+0.064
 forgotten+0.062
scar+0.062
Neg logit contributions
 Sky-0.063
fly-0.062
 tweet-0.058
 globe-0.057
 flap-0.057
Token and
Token ID290
Feature activation+2.92e-01

Pos logit contributions
gh+0.067
array+0.064
 purpose+0.063
 mischief+0.062
 Sister+0.061
Neg logit contributions
fly-0.051
 Sky-0.050
 tweet-0.043
 below-0.043
flies-0.042
Token to
Token ID284
Feature activation+1.54e+00

Pos logit contributions
gh+0.032
 purpose+0.029
omm+0.029
array+0.029
 behaviour+0.029
Neg logit contributions
 Sky-0.022
fly-0.022
 tweet-0.019
 below-0.019
 higher-0.019
Token prevent
Token ID2948
Feature activation+1.89e+00

Pos logit contributions
gh+0.083
omm+0.080
kids+0.077
 behaviour+0.077
child+0.076
Neg logit contributions
 Sky-0.048
 flap-0.047
 fly-0.047
 soar-0.047
fly-0.044
Token her
Token ID607
Feature activation+1.98e+00

Pos logit contributions
gh+0.114
array+0.104
omm+0.104
urance+0.104
David+0.103
Neg logit contributions
 Sky-0.055
fly-0.043
 higher-0.042
 below-0.041
 Bloom-0.040
Token mom
Token ID1995
Feature activation+1.07e+00

Pos logit contributions
gh+0.113
David+0.107
omm+0.104
array+0.101
urance+0.101
Neg logit contributions
fly-0.056
 Sky-0.053
 higher-0.050
 together-0.049
flies-0.049
Token from
Token ID422
Feature activation+1.62e+00

Pos logit contributions
gh+0.060
omm+0.056
 purpose+0.055
 behaviour+0.054
 Sister+0.053
Neg logit contributions
fly-0.034
 Sky-0.032
 below-0.029
 higher-0.028
 tweet-0.028
Token having
Token ID1719
Feature activation+3.88e+00

Pos logit contributions
gh+0.073
omm+0.070
urance+0.066
addy+0.066
 behaviour+0.065
Neg logit contributions
 Sky-0.065
fly-0.062
 below-0.060
 flap-0.059
 higher-0.059
Token to
Token ID284
Feature activation+3.39e+00

Pos logit contributions
omm+0.247
scar+0.242
uf+0.240
David+0.238
child+0.237
Neg logit contributions
 together-0.088
 higher-0.065
 on-0.063
 around-0.061
 Sky-0.060
Token clean
Token ID3424
Feature activation+1.71e+00

Pos logit contributions
omm+0.207
child+0.194
kids+0.190
gh+0.188
 behaviour+0.183
Neg logit contributions
 fly-0.099
 soar-0.090
 flap-0.085
 float-0.078
 sing-0.075
Token up
Token ID510
Feature activation+2.10e+00

Pos logit contributions
gh+0.117
urance+0.111
David+0.109
omm+0.109
OW+0.108
Neg logit contributions
 higher-0.028
 around-0.026
 Sky-0.026
 down-0.025
 together-0.025
Token the
Token ID262
Feature activation+2.15e+00

Pos logit contributions
gh+0.121
omm+0.118
urance+0.115
David+0.113
array+0.112
Neg logit contributions
 Sky-0.055
 higher-0.052
 together-0.052
 below-0.052
fly-0.048
Token mud
Token ID17492
Feature activation+3.29e-01

Pos logit contributions
gh+0.108
David+0.103
urance+0.102
 Samantha+0.100
uf+0.100
Neg logit contributions
 higher-0.071
 around-0.066
 globe-0.065
 flap-0.064
fl-0.063
Token smell
Token ID8508
Feature activation+2.43e+00

Pos logit contributions
David+0.146
child+0.134
em+0.133
kids+0.133
urance+0.131
Neg logit contributions
fly-0.116
 globe-0.105
 dragon-0.102
 flap-0.101
 higher-0.100
Token and
Token ID290
Feature activation+1.60e+00

Pos logit contributions
omm+0.159
urance+0.156
 Sister+0.155
array+0.154
 behaviour+0.153
Neg logit contributions
 below-0.043
 around-0.041
fly-0.040
 Sky-0.039
 globe-0.036
Token was
Token ID373
Feature activation+1.83e+00

Pos logit contributions
gh+0.095
urance+0.088
omm+0.087
 behaviour+0.085
array+0.085
Neg logit contributions
 Sky-0.045
fly-0.043
 below-0.038
 globe-0.036
 tweet-0.035
Token filled
Token ID5901
Feature activation+6.79e-01

Pos logit contributions
gh+0.087
urance+0.080
 behaviour+0.079
array+0.079
kids+0.079
Neg logit contributions
fly-0.074
 Sky-0.073
 below-0.065
 higher-0.061
 tweet-0.061
Token with
Token ID351
Feature activation+2.24e+00

Pos logit contributions
gh+0.045
 behaviour+0.043
omm+0.043
urance+0.043
 Sister+0.043
Neg logit contributions
 Sky-0.013
fly-0.013
 below-0.010
 globe-0.010
 higher-0.009
Token y
Token ID331
Feature activation+3.86e+00

Pos logit contributions
gh+0.113
omm+0.103
urance+0.100
array+0.097
uf+0.097
Neg logit contributions
 Sky-0.081
fly-0.078
 below-0.074
 higher-0.072
 around-0.072
Tokenucky
Token ID5309
Feature activation+1.42e+00

Pos logit contributions
gh+0.168
David+0.158
 forgotten+0.158
liv+0.157
 virus+0.154
Neg logit contributions
als-0.143
 together-0.128
 below-0.126
 higher-0.121
 Sky-0.120
Token water
Token ID1660
Feature activation+1.28e+00

Pos logit contributions
gh+0.072
omm+0.064
array+0.064
urance+0.064
David+0.064
Neg logit contributions
fly-0.053
 Sky-0.052
flies-0.046
 globe-0.046
als-0.045
Token.
Token ID13
Feature activation+1.52e+00

Pos logit contributions
gh+0.064
omm+0.059
array+0.058
 purpose+0.058
urance+0.057
Neg logit contributions
fly-0.049
 Sky-0.046
 below-0.042
 tweet-0.041
flies-0.041
Token
Token ID220
Feature activation+1.92e-01

Pos logit contributions
gh+0.076
 behaviour+0.066
 purpose+0.065
 forgotten+0.063
 virus+0.062
Neg logit contributions
 Sky-0.062
 below-0.055
fly-0.052
 globe-0.051
 Jupiter-0.050
Token
Token ID198
Feature activation+1.08e+00

Pos logit contributions
gh+0.011
 purpose+0.010
array+0.010
 behaviour+0.010
 Sister+0.010
Neg logit contributions
 Sky-0.006
fly-0.006
 tweet-0.005
 Saturn-0.005
 globe-0.005
Token go
Token ID467
Feature activation-2.75e-01

Pos logit contributions
gh+0.010
 purpose+0.009
array+0.009
omm+0.009
 behaviour+0.009
Neg logit contributions
 Sky-0.010
fly-0.010
 tweet-0.009
 Saturn-0.009
 globe-0.009
Token outside
Token ID2354
Feature activation-1.21e+00

Pos logit contributions
fly+0.009
 Sky+0.009
 tweet+0.008
 globe+0.008
 Saturn+0.008
Neg logit contributions
gh-0.015
 purpose-0.014
array-0.014
 forgotten-0.013
 Sister-0.013
Token."
Token ID526
Feature activation+1.63e+00

Pos logit contributions
fly+0.052
 Sky+0.052
 tweet+0.048
 globe+0.044
 Saturn+0.044
Neg logit contributions
gh-0.052
 purpose-0.049
array-0.048
 forgotten-0.047
 Sister-0.047
Token
Token ID198
Feature activation+8.80e-01

Pos logit contributions
gh+0.107
 purpose+0.097
 behaviour+0.094
 forgotten+0.094
 supposed+0.093
Neg logit contributions
 Sky-0.035
 together-0.032
fly-0.031
 Fly-0.027
 tweet-0.027
Token
Token ID198
Feature activation+1.37e+00

Pos logit contributions
gh+0.054
 behaviour+0.051
 purpose+0.051
array+0.051
 forgotten+0.050
Neg logit contributions
 Sky-0.022
fly-0.020
 tweet-0.017
 Saturn-0.016
 globe-0.016
TokenM
Token ID44
Feature activation+3.84e+00

Pos logit contributions
gh+0.068
 behaviour+0.066
 purpose+0.065
urance+0.064
omm+0.063
Neg logit contributions
 Sky-0.045
fly-0.043
 together-0.041
 tweet-0.040
 higher-0.039
Tokenum
Token ID388
Feature activation+1.46e-01

Pos logit contributions
 blonde+0.162
 supposed+0.157
 purpose+0.156
 blond+0.156
 mischief+0.156
Neg logit contributions
ite-0.111
 together-0.105
cast-0.105
ites-0.095
 tweet-0.095
Token and
Token ID290
Feature activation+2.02e+00

Pos logit contributions
gh+0.008
 purpose+0.008
array+0.008
 behaviour+0.007
omm+0.007
Neg logit contributions
 Sky-0.005
fly-0.005
 tweet-0.004
 globe-0.004
 Saturn-0.004
Token Dad
Token ID17415
Feature activation+4.11e-01

Pos logit contributions
gh+0.100
 purpose+0.087
 supposed+0.084
urance+0.080
 policy+0.080
Neg logit contributions
 together-0.076
 Sky-0.075
 down-0.075
 tweet-0.073
 higher-0.072
Token smiled
Token ID13541
Feature activation-1.88e-01

Pos logit contributions
gh+0.019
 purpose+0.018
array+0.017
 behaviour+0.016
omm+0.016
Neg logit contributions
 Sky-0.017
fly-0.017
 tweet-0.016
 below-0.015
 globe-0.015
Token and
Token ID290
Feature activation-2.40e-01

Pos logit contributions
 Sky+0.007
fly+0.006
 tweet+0.006
 Saturn+0.005
 globe+0.005
Neg logit contributions
gh-0.010
 purpose-0.009
 behaviour-0.009
array-0.009
 Sister-0.009
Token She
Token ID1375
Feature activation+1.14e+00

Pos logit contributions
gh+0.103
 behaviour+0.089
scar+0.088
 supposed+0.087
dad+0.087
Neg logit contributions
 Sky-0.045
 Bloom-0.035
als-0.034
 Jupiter-0.033
 Sparkle-0.033
Token remembered
Token ID12086
Feature activation+1.41e+00

Pos logit contributions
gh+0.057
 purpose+0.054
urance+0.052
 behaviour+0.051
 mischief+0.050
Neg logit contributions
fly-0.044
 Sky-0.044
 tweet-0.038
 flap-0.036
 deer-0.036
Token how
Token ID703
Feature activation+1.86e+00

Pos logit contributions
gh+0.091
omm+0.083
urance+0.082
array+0.081
 purpose+0.081
Neg logit contributions
 Sky-0.034
fly-0.031
 tweet-0.028
 below-0.028
 Saturn-0.027
Token the
Token ID262
Feature activation+1.45e+00

Pos logit contributions
gh+0.104
urance+0.100
array+0.099
omm+0.098
kids+0.097
Neg logit contributions
 together-0.058
 Sky-0.052
 around-0.051
 below-0.046
 higher-0.046
Token lady
Token ID10846
Feature activation-3.09e-01

Pos logit contributions
gh+0.088
urance+0.080
omm+0.079
uf+0.078
anked+0.076
Neg logit contributions
 higher-0.037
 Sky-0.037
 below-0.035
 fly-0.034
 globe-0.034
Token had
Token ID550
Feature activation+3.83e+00

Pos logit contributions
 Sky+0.014
fly+0.014
 tweet+0.013
 Saturn+0.012
 globe+0.012
Neg logit contributions
gh-0.013
 purpose-0.011
array-0.011
 forgotten-0.011
 behaviour-0.011
Token helped
Token ID4193
Feature activation+1.50e+00

Pos logit contributions
omm+0.207
uf+0.188
gh+0.185
urance+0.184
innie+0.182
Neg logit contributions
 together-0.125
 fly-0.109
 on-0.108
 around-0.107
 in-0.106
Token her
Token ID607
Feature activation+1.22e+00

Pos logit contributions
gh+0.077
array+0.073
omm+0.071
omo+0.070
scar+0.070
Neg logit contributions
 Sky-0.053
fly-0.049
 tweet-0.045
 below-0.044
 higher-0.043
Token and
Token ID290
Feature activation+6.87e-01

Pos logit contributions
gh+0.065
array+0.060
omm+0.059
David+0.058
kids+0.058
Neg logit contributions
fly-0.041
 Sky-0.040
 below-0.036
 flap-0.035
 fly-0.035
Token decided
Token ID3066
Feature activation+1.38e+00

Pos logit contributions
gh+0.033
 purpose+0.030
array+0.029
urance+0.029
omm+0.029
Neg logit contributions
fly-0.028
 Sky-0.028
 tweet-0.025
 below-0.023
 Saturn-0.023
Token to
Token ID284
Feature activation+1.66e+00

Pos logit contributions
gh+0.079
omm+0.076
array+0.075
urance+0.074
 mischief+0.073
Neg logit contributions
 Sky-0.039
fly-0.035
 below-0.034
 higher-0.033
 together-0.033
Token,
Token ID11
Feature activation+2.08e+00

Pos logit contributions
urance+0.165
array+0.165
 mischief+0.164
gh+0.163
child+0.163
Neg logit contributions
 together-0.044
 Sky-0.039
 higher-0.035
 on-0.035
 aboard-0.035
Token Tommy
Token ID19919
Feature activation+1.46e-01

Pos logit contributions
gh+0.096
urance+0.091
 behaviour+0.089
omm+0.086
array+0.084
Neg logit contributions
 together-0.080
 Sky-0.080
 higher-0.078
 below-0.075
 down-0.074
Token's
Token ID338
Feature activation-8.14e-02

Pos logit contributions
gh+0.008
 purpose+0.007
array+0.007
 behaviour+0.007
omm+0.007
Neg logit contributions
fly-0.005
 Sky-0.005
 tweet-0.005
 Saturn-0.004
 globe-0.004
Token trumpet
Token ID40334
Feature activation+1.79e-01

Pos logit contributions
fly+0.004
 Sky+0.004
 tweet+0.003
 Saturn+0.003
 globe+0.003
Neg logit contributions
gh-0.003
 purpose-0.003
 behaviour-0.003
array-0.003
 Sister-0.003
Token breaks
Token ID9457
Feature activation+1.11e+00

Pos logit contributions
gh+0.008
 purpose+0.007
array+0.007
 behaviour+0.007
omm+0.007
Neg logit contributions
fly-0.008
 Sky-0.008
 tweet-0.007
 globe-0.006
 Saturn-0.006
Token.
Token ID13
Feature activation+3.79e+00

Pos logit contributions
gh+0.066
 behaviour+0.062
urance+0.061
array+0.060
 Sister+0.060
Neg logit contributions
fly-0.031
 Sky-0.031
 higher-0.026
 below-0.026
 Saturn-0.025
Token He
Token ID679
Feature activation-5.43e-01

Pos logit contributions
gh+0.257
illy+0.242
 injuries+0.236
 supposed+0.235
liv+0.232
Neg logit contributions
 down-0.058
 Sky-0.058
 Twist-0.043
 around-0.043
 below-0.042
Token tries
Token ID8404
Feature activation+5.07e-01

Pos logit contributions
fly+0.021
 Sky+0.021
 tweet+0.019
 globe+0.018
 below+0.018
Neg logit contributions
gh-0.026
array-0.024
 purpose-0.023
 forgotten-0.023
 behaviour-0.023
Token to
Token ID284
Feature activation-4.37e-01

Pos logit contributions
gh+0.025
array+0.023
omm+0.022
 behaviour+0.022
urance+0.022
Neg logit contributions
 Sky-0.019
fly-0.019
 tweet-0.017
 higher-0.016
 Saturn-0.016
Token blow
Token ID6611
Feature activation-7.07e-01

Pos logit contributions
 Sky+0.019
fly+0.019
 tweet+0.017
 globe+0.017
 Saturn+0.016
Neg logit contributions
gh-0.019
 purpose-0.018
array-0.017
urance-0.016
 behaviour-0.016
Token,
Token ID11
Feature activation-3.05e-01

Pos logit contributions
fly+0.029
 Sky+0.027
 tweet+0.025
flies+0.024
 globe+0.023
Neg logit contributions
gh-0.035
 purpose-0.033
 mischief-0.030
array-0.030
 behaviour-0.030
Token
Token ID198
Feature activation-1.24e+00

Pos logit contributions
fly+0.027
 Sky+0.024
flies+0.023
 globe+0.022
 tweet+0.022
Neg logit contributions
gh-0.026
 purpose-0.025
 Sister-0.025
addy-0.024
 punishment-0.024
Token
Token ID198
Feature activation-3.29e-01

Pos logit contributions
fly+0.056
 Sky+0.050
flies+0.048
 tweet+0.048
als+0.047
Neg logit contributions
gh-0.055
 purpose-0.050
addy-0.048
anked-0.047
David-0.047
TokenThey
Token ID2990
Feature activation-3.18e-01

Pos logit contributions
fly+0.017
 Sky+0.016
 tweet+0.015
flies+0.015
 Saturn+0.014
Neg logit contributions
gh-0.012
 purpose-0.011
array-0.011
addy-0.010
 Sister-0.010
Token watched
Token ID7342
Feature activation-1.34e+00

Pos logit contributions
fly+0.014
 Sky+0.014
 tweet+0.013
 globe+0.012
 Saturn+0.012
Neg logit contributions
gh-0.013
array-0.012
 purpose-0.012
 Sister-0.012
 forgotten-0.012
Token the
Token ID262
Feature activation-2.25e+00

Pos logit contributions
fly+0.059
 Sky+0.052
 globe+0.050
flies+0.048
 tweet+0.048
Neg logit contributions
gh-0.057
 purpose-0.056
 Sister-0.055
 punishment-0.054
anked-0.054
Token moon
Token ID8824
Feature activation-5.57e+00

Pos logit contributions
fly+0.076
 Sky+0.076
bug+0.072
 Bloom+0.066
 Flor+0.064
Neg logit contributions
 purpose-0.107
gh-0.101
array-0.099
child-0.098
 mischief-0.097
Token together
Token ID1978
Feature activation-1.74e+00

Pos logit contributions
 Asia+0.201
 Sky+0.198
 Bloom+0.190
 Fly+0.187
 globe+0.186
Neg logit contributions
gh-0.197
 efforts-0.187
 behaviour-0.176
 purpose-0.174
anked-0.171
Token until
Token ID1566
Feature activation-6.05e-01

Pos logit contributions
fly+0.084
 Sky+0.083
 tweet+0.075
 globe+0.075
 Jupiter+0.075
Neg logit contributions
gh-0.063
 purpose-0.058
 behaviour-0.056
 punishment-0.054
urance-0.052
Token they
Token ID484
Feature activation-1.48e+00

Pos logit contributions
fly+0.023
 Sky+0.022
 tweet+0.020
flies+0.019
 globe+0.018
Neg logit contributions
gh-0.030
 purpose-0.029
 Sister-0.028
 forgotten-0.028
 punishment-0.027
Token fell
Token ID3214
Feature activation-1.38e+00

Pos logit contributions
 Sky+0.079
fly+0.077
 tweet+0.073
 globe+0.070
 Jupiter+0.070
Neg logit contributions
 forgotten-0.048
gh-0.047
 disob-0.041
 happened-0.040
addy-0.039
Token asleep
Token ID16039
Feature activation-1.33e+00

Pos logit contributions
 Sky+0.049
fly+0.049
 tweet+0.045
 Saturn+0.043
 Bloom+0.042
Neg logit contributions
gh-0.065
 purpose-0.063
array-0.060
 forgotten-0.059
anked-0.059
Token.
Token ID13
Feature activation-1.76e+00

Pos logit contributions
 Sky+0.116
fly+0.114
 tweet+0.114
 globe+0.107
 Fly+0.106
Neg logit contributions
gh-0.048
 purpose-0.045
 time-0.041
 business-0.036
 supposed-0.036
Token She
Token ID1375
Feature activation-3.05e+00

Pos logit contributions
fly+0.093
 Sky+0.080
flies+0.080
 globe+0.078
 tweet+0.077
Neg logit contributions
gh-0.060
 purpose-0.060
 Sister-0.060
urance-0.056
addy-0.056
Token would
Token ID561
Feature activation-3.09e+00

Pos logit contributions
fly+0.171
 Sky+0.168
 tweet+0.158
stack+0.156
 south+0.153
Neg logit contributions
gh-0.076
 happened-0.073
 hated-0.072
 been-0.069
 forgotten-0.066
Token fly
Token ID6129
Feature activation-3.75e+00

Pos logit contributions
fly+0.116
 Sky+0.106
flies+0.106
rots+0.105
als+0.103
Neg logit contributions
gh-0.133
 disob-0.126
 purpose-0.121
 time-0.120
 punish-0.117
Token up
Token ID510
Feature activation-2.68e+00

Pos logit contributions
fly+0.175
stack+0.168
 globe+0.160
 launcher+0.157
rots+0.154
Neg logit contributions
 time-0.124
 purpose-0.114
gh-0.113
 forgotten-0.103
 gotten-0.090
Token and
Token ID290
Feature activation-5.39e+00

Pos logit contributions
fly+0.117
 tweet+0.100
 globe+0.098
 Sky+0.097
flies+0.094
Neg logit contributions
 purpose-0.101
gh-0.101
 time-0.099
 forgotten-0.099
 trouble-0.094
Token up
Token ID510
Feature activation-1.74e+00

Pos logit contributions
rots+0.239
ectar+0.215
ang+0.209
fly+0.206
 Sky+0.203
Neg logit contributions
 quit-0.155
 gotten-0.136
 disob-0.133
 forgotten-0.133
 delay-0.132
Token,
Token ID11
Feature activation-4.45e+00

Pos logit contributions
fly+0.078
 Sky+0.068
 globe+0.067
 tweet+0.066
 Saturn+0.064
Neg logit contributions
gh-0.070
 purpose-0.067
 forgotten-0.065
 punishment-0.061
 time-0.060
Token then
Token ID788
Feature activation-2.89e+00

Pos logit contributions
fly+0.195
rots+0.186
 globe+0.173
fish+0.171
girls+0.171
Neg logit contributions
 expression-0.131
 complaining-0.125
 forgotten-0.123
 grown-0.122
 refused-0.122
Token come
Token ID1282
Feature activation-4.01e+00

Pos logit contributions
fly+0.113
 globe+0.108
 Sky+0.096
pole+0.092
 Saturn+0.091
Neg logit contributions
gh-0.115
 behaviour-0.112
 happened-0.107
 disob-0.105
 forgotten-0.105
Token back
Token ID736
Feature activation-3.06e+00

Pos logit contributions
rob+0.188
 Sky+0.176
fish+0.170
 Blink+0.167
igator+0.164
Neg logit contributions
 supposed-0.125
gh-0.115
 business-0.112
 forgotten-0.106
 purpose-0.105
Token would
Token ID561
Feature activation-1.86e+00

Pos logit contributions
 Sky+0.179
 Saturn+0.160
stack+0.153
 Jupiter+0.153
 higher+0.153
Neg logit contributions
gh-0.111
 happened-0.107
 disob-0.099
 hated-0.097
 forgotten-0.095
Token fly
Token ID6129
Feature activation-3.87e+00

Pos logit contributions
 Sky+0.061
fly+0.059
 globe+0.054
 Saturn+0.051
als+0.051
Neg logit contributions
gh-0.097
 purpose-0.091
 disob-0.089
 punish-0.087
 forgotten-0.085
Token over
Token ID625
Feature activation-1.97e+00

Pos logit contributions
stack+0.176
rots+0.165
 globe+0.165
 Sky+0.164
fly+0.162
Neg logit contributions
 time-0.132
gh-0.125
 purpose-0.118
 forgotten-0.100
 gotten-0.091
Token the
Token ID262
Feature activation-2.44e+00

Pos logit contributions
fly+0.088
 Sky+0.079
 globe+0.075
 tweet+0.074
pole+0.074
Neg logit contributions
 purpose-0.080
 mischief-0.079
 trouble-0.078
gh-0.077
 time-0.073
Token hills
Token ID18639
Feature activation-2.98e+00

Pos logit contributions
fly+0.105
 Sky+0.101
 tweet+0.095
 together+0.093
bug+0.090
Neg logit contributions
 purpose-0.098
 forgotten-0.090
gh-0.087
 situation-0.087
 mischief-0.087
Token and
Token ID290
Feature activation-5.34e+00

Pos logit contributions
 Sky+0.149
 tweet+0.147
fly+0.141
 Fly+0.133
 Flor+0.131
Neg logit contributions
gh-0.089
 purpose-0.081
 time-0.081
 forgotten-0.077
 business-0.075
Token forests
Token ID17039
Feature activation-2.57e+00

Pos logit contributions
rots+0.235
 globe+0.206
ang+0.203
 Sky+0.198
ites+0.196
Neg logit contributions
 quit-0.156
 disob-0.155
 punish-0.151
 grown-0.149
 gotten-0.147
Token.
Token ID13
Feature activation-1.77e+00

Pos logit contributions
fly+0.140
 Sky+0.139
 tweet+0.138
 Fly+0.130
 Saturn+0.130
Neg logit contributions
gh-0.074
 forgotten-0.060
 purpose-0.057
 business-0.053
 time-0.052
Token One
Token ID1881
Feature activation-1.39e+00

Pos logit contributions
fly+0.096
flies+0.085
fish+0.084
 tweet+0.084
 Sky+0.083
Neg logit contributions
 purpose-0.053
gh-0.050
urance-0.048
 time-0.047
array-0.047
Token day
Token ID1110
Feature activation+3.98e-01

Pos logit contributions
 Sky+0.079
fly+0.078
 tweet+0.073
flies+0.072
als+0.072
Neg logit contributions
gh-0.038
 time-0.037
 purpose-0.036
 punishment-0.033
 behaviour-0.032
Token,
Token ID11
Feature activation-1.57e+00

Pos logit contributions
gh+0.020
array+0.018
 purpose+0.018
omm+0.018
 behaviour+0.018
Neg logit contributions
 Sky-0.015
fly-0.015
 tweet-0.014
 higher-0.013
 globe-0.013
Token adventure
Token ID8855
Feature activation+5.53e-01

Pos logit contributions
gh+0.062
 purpose+0.055
omm+0.055
 Sister+0.055
 behaviour+0.055
Neg logit contributions
 Sky-0.054
fly-0.051
 Saturn-0.048
 Jupiter-0.048
 tweet-0.047
Token.
Token ID13
Feature activation+1.15e+00

Pos logit contributions
gh+0.031
omm+0.029
 purpose+0.029
 behaviour+0.029
 Sister+0.029
Neg logit contributions
 Sky-0.017
fly-0.017
 below-0.014
 tweet-0.014
 higher-0.014
Token He
Token ID679
Feature activation-1.54e-02

Pos logit contributions
gh+0.069
 purpose+0.061
 behaviour+0.061
 forgotten+0.061
addy+0.061
Neg logit contributions
 Sky-0.033
fly-0.030
 below-0.028
 globe-0.027
 Jupiter-0.026
Token watched
Token ID7342
Feature activation-1.21e-01

Pos logit contributions
 Sky+0.001
fly+0.001
 tweet+0.001
 globe+0.001
 Saturn+0.001
Neg logit contributions
gh-0.001
 purpose-0.001
array-0.001
 behaviour-0.001
 Sister-0.001
Token the
Token ID262
Feature activation-6.30e-01

Pos logit contributions
fly+0.005
 Sky+0.005
 tweet+0.004
 globe+0.004
 Saturn+0.004
Neg logit contributions
gh-0.006
 purpose-0.005
array-0.005
 behaviour-0.005
 Sister-0.005
Token birds
Token ID10087
Feature activation-5.17e+00

Pos logit contributions
fly+0.022
 Sky+0.022
 tweet+0.019
 Saturn+0.018
flies+0.018
Neg logit contributions
gh-0.032
 purpose-0.031
array-0.030
 behaviour-0.030
 Sister-0.029
Token flying
Token ID7348
Feature activation-2.75e+00

Pos logit contributions
 globe+0.201
 Sky+0.197
 Dragon+0.190
 Asia+0.182
 Lia+0.178
Neg logit contributions
 time-0.150
 cleaning-0.136
 grown-0.136
array-0.135
gh-0.135
Token in
Token ID287
Feature activation-6.93e-01

Pos logit contributions
 tweet+0.094
 Sky+0.091
fly+0.089
 Dragon+0.086
bug+0.080
Neg logit contributions
gh-0.118
 purpose-0.116
 time-0.116
 forgotten-0.101
anked-0.100
Token the
Token ID262
Feature activation-1.23e+00

Pos logit contributions
fly+0.025
 Sky+0.024
 tweet+0.022
flies+0.020
als+0.019
Neg logit contributions
gh-0.035
 purpose-0.034
array-0.033
 mischief-0.032
 Sister-0.031
Token sky
Token ID6766
Feature activation-2.14e+00

Pos logit contributions
fly+0.049
 Sky+0.048
 tweet+0.044
flies+0.041
 globe+0.040
Neg logit contributions
array-0.054
 purpose-0.054
gh-0.053
 Sister-0.051
 forgotten-0.051
Token and
Token ID290
Feature activation-3.57e+00

Pos logit contributions
 Sky+0.116
fly+0.111
 tweet+0.111
 globe+0.107
 Saturn+0.103
Neg logit contributions
gh-0.058
 purpose-0.056
 time-0.050
em-0.048
array-0.048
Token friends
Token ID2460
Feature activation-1.04e+00

Pos logit contributions
fly+0.037
 Sky+0.035
 tweet+0.033
 globe+0.031
 Jupiter+0.030
Neg logit contributions
gh-0.033
 purpose-0.030
 forgotten-0.029
 mischief-0.029
 Sister-0.028
Token.
Token ID13
Feature activation-5.22e-01

Pos logit contributions
fly+0.056
 Sky+0.055
 tweet+0.052
 Saturn+0.050
 Jupiter+0.050
Neg logit contributions
gh-0.034
 purpose-0.029
 behaviour-0.029
 forgotten-0.028
 Sister-0.027
Token They
Token ID1119
Feature activation-2.03e+00

Pos logit contributions
fly+0.021
 Sky+0.020
 tweet+0.019
 globe+0.017
 flap+0.017
Neg logit contributions
gh-0.024
 purpose-0.023
 Sister-0.023
array-0.022
urance-0.022
Token watch
Token ID2342
Feature activation-2.31e+00

Pos logit contributions
fly+0.093
 Sky+0.088
flies+0.080
 Jupiter+0.079
 globe+0.078
Neg logit contributions
gh-0.077
 forgotten-0.076
 disob-0.073
 punish-0.071
 Sister-0.070
Token the
Token ID262
Feature activation-1.39e+00

Pos logit contributions
fly+0.108
 Sky+0.091
 globe+0.091
stack+0.089
flies+0.087
Neg logit contributions
gh-0.090
 purpose-0.088
 time-0.083
 mischief-0.080
 Tommy-0.079
Token birds
Token ID10087
Feature activation-5.11e+00

Pos logit contributions
fly+0.055
 Sky+0.054
 tweet+0.046
 Saturn+0.045
 Jupiter+0.045
Neg logit contributions
gh-0.063
 purpose-0.061
 behaviour-0.059
 punishment-0.058
 mischief-0.057
Token and
Token ID290
Feature activation-4.19e+00

Pos logit contributions
 globe+0.197
 Sky+0.181
 Asia+0.179
 Saturn+0.178
 Dragon+0.177
Neg logit contributions
 time-0.162
gh-0.153
 punish-0.153
 grown-0.144
urance-0.144
Token the
Token ID262
Feature activation-1.82e+00

Pos logit contributions
fly+0.202
pole+0.200
 globe+0.198
rots+0.192
pp+0.181
Neg logit contributions
 mischief-0.109
 Sister-0.109
 grown-0.104
 trouble-0.102
 Tommy-0.100
Token flowers
Token ID12734
Feature activation-4.19e+00

Pos logit contributions
fly+0.062
 Sky+0.060
 Saturn+0.051
 Jupiter+0.050
 Fly+0.048
Neg logit contributions
gh-0.091
 purpose-0.089
 behaviour-0.085
 mischief-0.084
 Sister-0.082
Token.
Token ID13
Feature activation-7.89e-01

Pos logit contributions
fly+0.214
 Sky+0.214
 Jupiter+0.208
 Fly+0.206
 Dragon+0.206
Neg logit contributions
gh-0.091
 time-0.084
 purpose-0.078
 forgotten-0.074
 grown-0.069
Token They
Token ID1119
Feature activation-3.15e+00

Pos logit contributions
fly+0.036
 Sky+0.032
 tweet+0.032
 flap+0.030
 globe+0.030
Neg logit contributions
gh-0.032
 purpose-0.032
 Sister-0.031
urance-0.030
array-0.029
Token Tom
Token ID4186
Feature activation-1.26e+00

Pos logit contributions
gh+0.094
 behaviour+0.087
illy+0.084
 supposed+0.084
 purpose+0.083
Neg logit contributions
 Sky-0.049
 below-0.041
 higher-0.041
 Saturn-0.038
 mountains-0.036
Token watched
Token ID7342
Feature activation-8.48e-01

Pos logit contributions
fly+0.061
 Sky+0.061
 tweet+0.057
 globe+0.054
 Saturn+0.053
Neg logit contributions
gh-0.046
 forgotten-0.041
array-0.041
 Sister-0.039
addy-0.038
Token it
Token ID340
Feature activation-3.02e+00

Pos logit contributions
fly+0.035
 Sky+0.032
 tweet+0.030
flies+0.029
 globe+0.029
Neg logit contributions
gh-0.041
 purpose-0.038
 behaviour-0.036
 punishment-0.036
 mischief-0.035
Token go
Token ID467
Feature activation-1.59e+00

Pos logit contributions
fly+0.122
 Sky+0.116
 globe+0.114
als+0.112
rots+0.111
Neg logit contributions
gh-0.129
 purpose-0.114
urance-0.108
 behaviour-0.106
 punishment-0.106
Token higher
Token ID2440
Feature activation-2.58e+00

Pos logit contributions
fly+0.058
 Sky+0.058
 tweet+0.054
bug+0.049
flies+0.049
Neg logit contributions
gh-0.080
 forgotten-0.071
 purpose-0.071
array-0.068
 business-0.064
Token and
Token ID290
Feature activation-5.01e+00

Pos logit contributions
 tweet+0.131
 Sky+0.127
fly+0.124
bug+0.113
 Fly+0.111
Neg logit contributions
 purpose-0.087
gh-0.078
 time-0.077
array-0.072
anked-0.071
Token higher
Token ID2440
Feature activation-2.04e+00

Pos logit contributions
rots+0.182
 Sky+0.175
fly+0.174
 globe+0.163
stack+0.157
Neg logit contributions
 forgotten-0.197
 complained-0.175
 quit-0.171
array-0.168
 disaster-0.168
Token until
Token ID1566
Feature activation-2.38e-01

Pos logit contributions
 Sky+0.105
 tweet+0.103
fly+0.102
bug+0.094
 Jupiter+0.094
Neg logit contributions
 purpose-0.063
gh-0.062
anked-0.054
 forgotten-0.052
 behaviour-0.052
Token he
Token ID339
Feature activation-1.26e+00

Pos logit contributions
fly+0.008
 Sky+0.008
 tweet+0.007
 globe+0.007
flies+0.007
Neg logit contributions
gh-0.013
 purpose-0.012
array-0.011
 forgotten-0.011
 Sister-0.011
Token could
Token ID714
Feature activation-1.56e+00

Pos logit contributions
 Sky+0.062
fly+0.061
 tweet+0.058
 globe+0.055
 deer+0.053
Neg logit contributions
gh-0.048
 forgotten-0.043
addy-0.039
 purpose-0.038
array-0.038
Token not
Token ID407
Feature activation-1.81e+00

Pos logit contributions
fly+0.073
 Sky+0.072
als+0.066
 globe+0.066
 tweet+0.065
Neg logit contributions
gh-0.061
 forgotten-0.056
 purpose-0.055
 supposed-0.051
urance-0.050
Token dad
Token ID9955
Feature activation-1.05e+00

Pos logit contributions
gh+0.041
array+0.038
 purpose+0.035
urance+0.035
anked+0.035
Neg logit contributions
 Sky-0.045
fly-0.041
 deer-0.040
 tweet-0.040
 below-0.040
Token would
Token ID561
Feature activation-1.27e+00

Pos logit contributions
 Sky+0.051
fly+0.049
 tweet+0.046
 Saturn+0.045
 Jupiter+0.044
Neg logit contributions
gh-0.040
 forgotten-0.036
 Sister-0.035
 behaviour-0.034
array-0.034
Token watch
Token ID2342
Feature activation-2.96e+00

Pos logit contributions
 Sky+0.064
fly+0.063
 Saturn+0.057
 tweet+0.057
 globe+0.057
Neg logit contributions
gh-0.044
 purpose-0.041
urance-0.038
array-0.037
 forgotten-0.036
Token the
Token ID262
Feature activation-6.39e-01

Pos logit contributions
fly+0.128
 globe+0.117
 Sky+0.115
fish+0.109
als+0.109
Neg logit contributions
 purpose-0.112
gh-0.111
 mischief-0.106
 behaviour-0.104
 time-0.101
Token baby
Token ID5156
Feature activation-3.58e+00

Pos logit contributions
fly+0.029
 Sky+0.028
 tweet+0.026
 Saturn+0.025
 globe+0.024
Neg logit contributions
 purpose-0.026
gh-0.026
 behaviour-0.025
 Sister-0.024
 mischief-0.024
Token birds
Token ID10087
Feature activation-5.00e+00

Pos logit contributions
 Saturn+0.134
 Jupiter+0.123
 globe+0.118
fly+0.118
 Mars+0.117
Neg logit contributions
 behaviour-0.143
 trouble-0.132
 punish-0.131
 forgotten-0.129
 mischief-0.129
Token grow
Token ID1663
Feature activation-2.41e+00

Pos logit contributions
 globe+0.181
 Saturn+0.178
 Asia+0.173
 Sky+0.167
 Dragon+0.165
Neg logit contributions
 punish-0.176
gh-0.167
urance-0.163
 behaviour-0.162
 forgotten-0.160
Token from
Token ID422
Feature activation-1.52e+00

Pos logit contributions
fly+0.088
 Sky+0.086
 globe+0.082
stack+0.080
 tweet+0.080
Neg logit contributions
gh-0.107
 purpose-0.101
 behaviour-0.100
 mischief-0.099
anked-0.098
Token a
Token ID257
Feature activation-1.29e+00

Pos logit contributions
 Sky+0.064
fly+0.063
 tweet+0.057
als+0.054
flies+0.053
Neg logit contributions
 purpose-0.065
 mischief-0.064
 trouble-0.061
 forgotten-0.061
gh-0.061
Token safe
Token ID3338
Feature activation-7.44e-01

Pos logit contributions
fly+0.051
 Sky+0.050
flies+0.045
 tweet+0.044
 Saturn+0.043
Neg logit contributions
gh-0.060
 purpose-0.058
 forgotten-0.057
 mischief-0.053
 Sister-0.053
Token distance
Token ID5253
Feature activation-9.31e-01

Pos logit contributions
 Sky+0.031
fly+0.030
 tweet+0.028
 Saturn+0.027
 Jupiter+0.026
Neg logit contributions
 purpose-0.032
gh-0.032
array-0.030
 forgotten-0.030
 behaviour-0.030
Token up
Token ID510
Feature activation-1.38e+00

Pos logit contributions
fly+0.081
 Sky+0.075
 Jupiter+0.071
flies+0.068
 tweet+0.068
Neg logit contributions
gh-0.089
 purpose-0.082
 time-0.075
 trouble-0.074
 behaviour-0.072
Token and
Token ID290
Feature activation-4.38e+00

Pos logit contributions
fly+0.066
 Sky+0.060
 tweet+0.058
flies+0.055
 globe+0.054
Neg logit contributions
gh-0.052
 purpose-0.051
 trouble-0.047
 forgotten-0.046
 mischief-0.046
Token down
Token ID866
Feature activation-1.26e+00

Pos logit contributions
fly+0.175
 Jupiter+0.170
rots+0.165
 Saturn+0.154
 Sky+0.153
Neg logit contributions
 forgotten-0.156
 Tommy-0.149
 complained-0.148
 whine-0.142
 Sister-0.138
Token.
Token ID13
Feature activation-1.55e+00

Pos logit contributions
fly+0.070
 Sky+0.066
 tweet+0.064
 Jupiter+0.060
flies+0.060
Neg logit contributions
gh-0.037
 purpose-0.037
 trouble-0.034
 forgotten-0.034
 mischief-0.033
Token The
Token ID383
Feature activation-1.73e+00

Pos logit contributions
fly+0.084
 tweet+0.075
pole+0.072
flies+0.071
 globe+0.070
Neg logit contributions
 Sister-0.051
 purpose-0.050
 mischief-0.048
 Tommy-0.047
urance-0.045
Token bird
Token ID6512
Feature activation-4.99e+00

Pos logit contributions
fly+0.060
 Sky+0.060
 Sparkle+0.053
 Jupiter+0.051
 Saturn+0.051
Neg logit contributions
 purpose-0.081
 trouble-0.079
 punishment-0.078
gh-0.076
 accident-0.076
Token sits
Token ID10718
Feature activation-2.76e+00

Pos logit contributions
stack+0.215
 globe+0.204
fly+0.203
 Jupiter+0.197
chard+0.192
Neg logit contributions
 time-0.148
 mistake-0.137
 complained-0.136
 trouble-0.135
 hated-0.134
Token on
Token ID319
Feature activation-5.26e-01

Pos logit contributions
fly+0.130
 Sky+0.129
 tweet+0.117
flies+0.115
 Jupiter+0.112
Neg logit contributions
gh-0.093
 forgotten-0.085
dad-0.085
 badly-0.084
addy-0.083
Token the
Token ID262
Feature activation-5.53e-01

Pos logit contributions
fly+0.020
 Sky+0.018
 tweet+0.017
flies+0.016
 globe+0.016
Neg logit contributions
gh-0.027
 purpose-0.025
 Sister-0.024
array-0.024
 mischief-0.024
Token middle
Token ID3504
Feature activation-1.04e+00

Pos logit contributions
fly+0.023
 Sky+0.022
 tweet+0.019
flies+0.019
 Jupiter+0.018
Neg logit contributions
gh-0.027
 purpose-0.024
array-0.023
 forgotten-0.023
 behaviour-0.023
Token and
Token ID290
Feature activation-4.54e+00

Pos logit contributions
fly+0.048
 Sky+0.047
 tweet+0.043
 Jupiter+0.041
flies+0.040
Neg logit contributions
gh-0.043
 purpose-0.039
em-0.037
 behaviour-0.037
addy-0.037
Token the
Token ID262
Feature activation-6.02e-01

Pos logit contributions
gh+0.011
 purpose+0.010
array+0.010
omm+0.010
 behaviour+0.010
Neg logit contributions
 Sky-0.008
fly-0.008
 tweet-0.007
 Saturn-0.007
 below-0.007
Token other
Token ID584
Feature activation-6.85e-01

Pos logit contributions
fly+0.018
 Sky+0.018
 tweet+0.014
 Saturn+0.014
 Jupiter+0.014
Neg logit contributions
gh-0.034
 purpose-0.033
 behaviour-0.032
array-0.031
 mischief-0.031
Token birds
Token ID10087
Feature activation-3.43e+00

Pos logit contributions
fly+0.023
 Sky+0.023
 tweet+0.019
 Saturn+0.019
 Jupiter+0.019
Neg logit contributions
gh-0.034
 purpose-0.033
 behaviour-0.033
 mischief-0.032
em-0.032
Token.
Token ID13
Feature activation-1.20e+00

Pos logit contributions
 Sky+0.180
fly+0.178
 Jupiter+0.177
 Saturn+0.176
 globe+0.173
Neg logit contributions
gh-0.075
 forgotten-0.072
 supposed-0.068
 behaviour-0.066
 disob-0.064
Token The
Token ID383
Feature activation-8.80e-01

Pos logit contributions
fly+0.062
 tweet+0.054
flies+0.054
 Sky+0.053
 globe+0.053
Neg logit contributions
 purpose-0.040
gh-0.040
 Sister-0.040
 mischief-0.039
 Tommy-0.039
Token birds
Token ID10087
Feature activation-4.98e+00

Pos logit contributions
fly+0.028
 Sky+0.027
flies+0.022
 Saturn+0.022
 Jupiter+0.022
Neg logit contributions
gh-0.046
 purpose-0.046
 punishment-0.045
 behaviour-0.044
 forgotten-0.044
Token sing
Token ID1702
Feature activation-2.60e+00

Pos logit contributions
 Jupiter+0.189
fly+0.187
 Sky+0.183
 Saturn+0.182
 Pluto+0.177
Neg logit contributions
 punish-0.162
 disob-0.160
 time-0.158
 grown-0.154
 pain-0.147
Token and
Token ID290
Feature activation-3.75e+00

Pos logit contributions
fly+0.116
stack+0.113
 globe+0.110
flies+0.104
 Sky+0.101
Neg logit contributions
gh-0.097
 purpose-0.087
 time-0.084
 behaviour-0.081
 punishment-0.078
Token thank
Token ID5875
Feature activation-1.17e+00

Pos logit contributions
pole+0.154
 globe+0.140
rots+0.138
fly+0.134
ite+0.127
Neg logit contributions
 trouble-0.126
 pain-0.125
 punish-0.125
 purpose-0.125
 mischief-0.125
Token the
Token ID262
Feature activation-5.41e-01

Pos logit contributions
fly+0.061
 Sky+0.054
 globe+0.053
 tweet+0.051
 Jupiter+0.051
Neg logit contributions
gh-0.045
 purpose-0.037
 Sister-0.037
 behaviour-0.037
em-0.036
Token man
Token ID582
Feature activation-2.51e+00

Pos logit contributions
fly+0.017
 Sky+0.017
 Saturn+0.014
 tweet+0.014
 Jupiter+0.013
Neg logit contributions
gh-0.029
 purpose-0.028
 behaviour-0.027
array-0.027
 Sister-0.027
Token every
Token ID790
Feature activation-1.30e+00

Pos logit contributions
stack+0.217
fly+0.217
pp+0.208
pole+0.195
cher+0.191
Neg logit contributions
 trouble-0.127
 time-0.118
 happened-0.117
 purpose-0.111
 mistakes-0.111
Token night
Token ID1755
Feature activation-1.80e+00

Pos logit contributions
 Sky+0.064
fly+0.062
flies+0.058
 tweet+0.057
als+0.056
Neg logit contributions
gh-0.046
 purpose-0.043
 punishment-0.042
 behaviour-0.041
 forgotten-0.040
Token to
Token ID284
Feature activation-2.96e+00

Pos logit contributions
 Sky+0.080
fly+0.078
 tweet+0.076
 globe+0.071
 flap+0.071
Neg logit contributions
gh-0.071
 purpose-0.067
 forgotten-0.066
array-0.065
 behaviour-0.065
Token watch
Token ID2342
Feature activation-2.78e+00

Pos logit contributions
fly+0.143
 globe+0.139
als+0.137
 Sky+0.136
pole+0.135
Neg logit contributions
 purpose-0.095
gh-0.087
 mischief-0.085
 punish-0.085
 grown-0.081
Token the
Token ID262
Feature activation-2.49e+00

Pos logit contributions
fly+0.122
stack+0.115
 Sky+0.112
 globe+0.112
 tweet+0.105
Neg logit contributions
gh-0.103
 purpose-0.100
 behaviour-0.094
 mischief-0.093
 punishment-0.089
Token moon
Token ID8824
Feature activation-4.97e+00

Pos logit contributions
fly+0.089
 Sky+0.085
stack+0.081
bug+0.081
 Flor+0.074
Neg logit contributions
 purpose-0.115
 time-0.111
 mischief-0.106
 happened-0.106
array-0.106
Token together
Token ID1978
Feature activation-1.80e+00

Pos logit contributions
 Sky+0.195
 Asia+0.189
 globe+0.182
 Bloom+0.181
 Fly+0.178
Neg logit contributions
gh-0.173
 behaviour-0.147
 purpose-0.147
urance-0.142
 time-0.141
Token.
Token ID13
Feature activation-1.21e+00

Pos logit contributions
 Sky+0.094
fly+0.092
 tweet+0.087
 globe+0.085
 Saturn+0.085
Neg logit contributions
gh-0.058
 purpose-0.051
 behaviour-0.047
 forgotten-0.046
urance-0.044
Token They
Token ID1119
Feature activation-2.71e+00

Pos logit contributions
fly+0.053
 Sky+0.048
 flap+0.046
 tweet+0.046
 higher+0.046
Neg logit contributions
gh-0.050
 purpose-0.049
omm-0.048
urance-0.047
 Sister-0.046
Token became
Token ID2627
Feature activation-1.84e+00

Pos logit contributions
 Sky+0.143
fly+0.139
 higher+0.139
 aboard+0.134
 south+0.133
Neg logit contributions
 forgotten-0.076
gh-0.073
 disagreed-0.066
 disob-0.064
 been-0.064
Token the
Token ID262
Feature activation-1.75e+00

Pos logit contributions
fly+0.101
 tweet+0.090
 Sky+0.090
 globe+0.089
als+0.088
Neg logit contributions
 purpose-0.056
 Sister-0.055
 mischief-0.054
gh-0.053
 supposed-0.052
Token
Token ID198
Feature activation-4.28e-01

Pos logit contributions
fly+0.003
 Sky+0.003
 tweet+0.002
 globe+0.002
 Saturn+0.002
Neg logit contributions
gh-0.004
 purpose-0.004
array-0.004
 Sister-0.004
 behaviour-0.004
Token
Token ID198
Feature activation+3.80e-01

Pos logit contributions
fly+0.016
 Sky+0.015
 tweet+0.014
flies+0.013
 globe+0.013
Neg logit contributions
gh-0.022
 purpose-0.020
 Sister-0.020
array-0.019
addy-0.019
TokenThey
Token ID2990
Feature activation-4.10e-01

Pos logit contributions
gh+0.017
 purpose+0.015
 behaviour+0.015
array+0.015
omm+0.015
Neg logit contributions
 Sky-0.016
fly-0.016
 tweet-0.015
 Saturn-0.014
 globe-0.014
Token watched
Token ID7342
Feature activation-1.03e+00

Pos logit contributions
fly+0.018
 Sky+0.017
 tweet+0.016
 globe+0.015
 Saturn+0.015
Neg logit contributions
gh-0.018
 purpose-0.016
array-0.016
 forgotten-0.016
 Sister-0.016
Token the
Token ID262
Feature activation-1.02e+00

Pos logit contributions
fly+0.042
 Sky+0.038
 globe+0.036
 tweet+0.036
flies+0.034
Neg logit contributions
gh-0.047
 purpose-0.045
 Sister-0.044
 behaviour-0.043
urance-0.043
Token comet
Token ID31733
Feature activation-4.96e+00

Pos logit contributions
fly+0.040
 Sky+0.040
 tweet+0.035
 Saturn+0.033
 Jupiter+0.033
Neg logit contributions
gh-0.047
 purpose-0.046
 behaviour-0.044
 Sister-0.043
 punishment-0.043
Token for
Token ID329
Feature activation-9.98e-02

Pos logit contributions
 globe+0.175
 Sky+0.171
 Asia+0.165
fly+0.163
als+0.162
Neg logit contributions
gh-0.192
 behaviour-0.181
 purpose-0.178
 efforts-0.174
urance-0.169
Token a
Token ID257
Feature activation-2.16e-01

Pos logit contributions
fly+0.004
 Sky+0.004
 tweet+0.004
 globe+0.003
 Saturn+0.003
Neg logit contributions
gh-0.005
 purpose-0.004
array-0.004
 behaviour-0.004
 Sister-0.004
Token while
Token ID981
Feature activation+1.16e-01

Pos logit contributions
fly+0.010
 Sky+0.010
 tweet+0.009
 Saturn+0.008
 below+0.008
Neg logit contributions
gh-0.009
 purpose-0.008
array-0.008
 behaviour-0.008
 Sister-0.008
Token,
Token ID11
Feature activation-1.50e+00

Pos logit contributions
gh+0.006
 purpose+0.005
array+0.005
 behaviour+0.005
omm+0.005
Neg logit contributions
 Sky-0.004
fly-0.004
 tweet-0.004
 Saturn-0.004
 globe-0.004
Token until
Token ID1566
Feature activation-7.96e-01

Pos logit contributions
fly+0.068
 Sky+0.067
 tweet+0.060
als+0.059
 globe+0.059
Neg logit contributions
 forgotten-0.058
gh-0.057
array-0.057
 Sister-0.057
 mischief-0.054
Token scary
Token ID14343
Feature activation-1.38e+00

Pos logit contributions
fly+0.218
als+0.208
ites+0.204
stack+0.203
 Sky+0.202
Neg logit contributions
 unfair-0.106
 forgotten-0.098
 trouble-0.095
 expensive-0.092
 embarrassing-0.088
Token!"
Token ID2474
Feature activation+1.57e-01

Pos logit contributions
fly+0.064
 Sky+0.063
 tweet+0.060
 Saturn+0.057
 globe+0.056
Neg logit contributions
gh-0.052
 purpose-0.048
 behaviour-0.045
array-0.044
 supposed-0.044
Token They
Token ID1119
Feature activation-1.14e+00

Pos logit contributions
gh+0.008
 purpose+0.008
array+0.008
 behaviour+0.007
omm+0.007
Neg logit contributions
 Sky-0.006
fly-0.005
 tweet-0.005
 Saturn-0.005
 globe-0.004
Token swing
Token ID9628
Feature activation-3.18e+00

Pos logit contributions
fly+0.051
 Sky+0.049
 globe+0.044
 Saturn+0.042
 Jupiter+0.042
Neg logit contributions
gh-0.045
 Sister-0.043
 forgotten-0.042
 disob-0.042
array-0.042
Token higher
Token ID2440
Feature activation-2.86e+00

Pos logit contributions
fly+0.127
 Dragon+0.123
 reef+0.116
fish+0.114
 globe+0.112
Neg logit contributions
anked-0.109
 trouble-0.105
gh-0.104
 Tommy-0.102
 purpose-0.099
Token and
Token ID290
Feature activation-4.95e+00

Pos logit contributions
 Sky+0.146
 tweet+0.144
fly+0.140
 Sparkle+0.139
 Dragon+0.137
Neg logit contributions
 purpose-0.079
anked-0.079
 time-0.074
gh-0.070
em-0.070
Token higher
Token ID2440
Feature activation-2.42e+00

Pos logit contributions
rots+0.179
fly+0.174
amer+0.161
 globe+0.158
 Dragon+0.154
Neg logit contributions
 forgotten-0.186
 complained-0.171
 refused-0.168
 disagreed-0.164
 Tommy-0.163
Token,
Token ID11
Feature activation-3.03e+00

Pos logit contributions
 Sky+0.124
 tweet+0.122
fly+0.120
 Jupiter+0.115
 Sparkle+0.115
Neg logit contributions
 purpose-0.069
anked-0.065
gh-0.064
 time-0.059
em-0.058
Token pretending
Token ID23662
Feature activation-9.82e-01

Pos logit contributions
fly+0.142
 Sky+0.135
rots+0.130
 Dragon+0.125
 globe+0.125
Neg logit contributions
 forgotten-0.108
array-0.102
 accident-0.098
gh-0.096
 supposed-0.095
Token to
Token ID284
Feature activation-2.62e+00

Pos logit contributions
fly+0.046
 Sky+0.045
 tweet+0.042
 Saturn+0.040
 globe+0.040
Neg logit contributions
gh-0.040
 purpose-0.035
 forgotten-0.033
 behaviour-0.033
array-0.032
Token fly
Token ID6129
Feature activation-3.07e+00

Pos logit contributions
 globe+0.112
rots+0.111
 Sky+0.110
fly+0.108
als+0.108
Neg logit contributions
 purpose-0.095
gh-0.092
 disob-0.083
 reason-0.083
 mischief-0.081
Token,
Token ID11
Feature activation-1.25e+00

Pos logit contributions
gh+0.024
array+0.023
 purpose+0.022
 behaviour+0.022
urance+0.022
Neg logit contributions
 Sky-0.018
fly-0.017
 tweet-0.015
 Bloom-0.014
 globe-0.014
Token Lily
Token ID20037
Feature activation-6.88e-01

Pos logit contributions
fly+0.075
 Sky+0.065
 globe+0.065
 tweet+0.064
flies+0.064
Neg logit contributions
 purpose-0.035
gh-0.034
 punishment-0.033
 Sister-0.032
 forgotten-0.032
Token would
Token ID561
Feature activation-8.08e-01

Pos logit contributions
 Sky+0.033
fly+0.032
 tweet+0.030
 globe+0.028
 Saturn+0.028
Neg logit contributions
gh-0.027
 forgotten-0.024
array-0.024
 purpose-0.024
 Sister-0.024
Token visit
Token ID3187
Feature activation-2.62e+00

Pos logit contributions
 Sky+0.039
fly+0.038
 globe+0.035
 below+0.034
 tweet+0.034
Neg logit contributions
gh-0.031
 purpose-0.028
array-0.027
 behaviour-0.027
urance-0.026
Token the
Token ID262
Feature activation-1.07e+00

Pos logit contributions
fly+0.136
 globe+0.125
stack+0.116
ite+0.112
pole+0.111
Neg logit contributions
 purpose-0.088
gh-0.082
 mischief-0.081
 forgotten-0.080
 trouble-0.078
Token beetle
Token ID45489
Feature activation-4.94e+00

Pos logit contributions
fly+0.041
 Sky+0.039
 tweet+0.033
flies+0.033
 Jupiter+0.032
Neg logit contributions
gh-0.053
 purpose-0.053
 forgotten-0.050
 behaviour-0.050
 mischief-0.050
Token every
Token ID790
Feature activation-6.74e-01

Pos logit contributions
fly+0.199
 globe+0.195
 Sky+0.188
 Saturn+0.186
 Pluto+0.172
Neg logit contributions
gh-0.169
 time-0.159
 business-0.158
 behaviour-0.153
 forgotten-0.149
Token day
Token ID1110
Feature activation-1.68e+00

Pos logit contributions
 Sky+0.033
fly+0.033
 tweet+0.030
flies+0.030
 globe+0.030
Neg logit contributions
gh-0.025
 purpose-0.023
 behaviour-0.022
 forgotten-0.022
 mischief-0.022
Token and
Token ID290
Feature activation-2.88e+00

Pos logit contributions
fly+0.076
 Sky+0.075
 tweet+0.070
 globe+0.070
 flap+0.065
Neg logit contributions
gh-0.068
 forgotten-0.061
 behaviour-0.059
 purpose-0.059
 supposed-0.059
Token make
Token ID787
Feature activation-1.80e+00

Pos logit contributions
 globe+0.130
fly+0.122
 Sky+0.122
ites+0.118
pole+0.116
Neg logit contributions
 forgotten-0.101
 purpose-0.093
gh-0.093
 disob-0.092
 punishment-0.092
Token sure
Token ID1654
Feature activation-1.34e+00

Pos logit contributions
fly+0.096
 globe+0.089
stack+0.084
 Sky+0.084
 tweet+0.080
Neg logit contributions
gh-0.067
 mischief-0.064
 purpose-0.064
 behaviour-0.058
 trouble-0.058
Token stars
Token ID5788
Feature activation-2.26e+00

Pos logit contributions
gh+0.045
urance+0.040
omm+0.040
 behaviour+0.040
array+0.040
Neg logit contributions
 Sky-0.034
fly-0.031
 tweet-0.029
 deer-0.027
 Saturn-0.027
Token.
Token ID13
Feature activation-3.52e-01

Pos logit contributions
 Sky+0.121
 tweet+0.119
fly+0.116
 Saturn+0.116
 Fly+0.112
Neg logit contributions
gh-0.065
 purpose-0.053
 forgotten-0.051
 supposed-0.051
 time-0.049
Token She
Token ID1375
Feature activation-9.83e-01

Pos logit contributions
fly+0.016
 Sky+0.015
 tweet+0.014
 globe+0.013
flies+0.013
Neg logit contributions
gh-0.015
 purpose-0.014
omm-0.014
array-0.014
 Sister-0.013
Token would
Token ID561
Feature activation-2.07e+00

Pos logit contributions
 Sky+0.047
fly+0.047
 tweet+0.044
 higher+0.042
 below+0.042
Neg logit contributions
gh-0.040
 forgotten-0.033
 purpose-0.032
array-0.031
 supposed-0.031
Token watch
Token ID2342
Feature activation-4.57e+00

Pos logit contributions
 Sky+0.095
fly+0.093
flies+0.085
 below+0.083
als+0.083
Neg logit contributions
gh-0.080
 purpose-0.070
 forgotten-0.065
 disob-0.064
array-0.064
Token them
Token ID606
Feature activation-4.94e+00

Pos logit contributions
stack+0.237
ite+0.226
fly+0.219
 Follow+0.219
bug+0.215
Neg logit contributions
gh-0.135
 purpose-0.133
 time-0.126
 grown-0.115
 mischief-0.102
Token tw
Token ID665
Feature activation-2.11e+00

Pos logit contributions
stack+0.196
 Follow+0.179
 globe+0.178
 Sky+0.168
fly+0.167
Neg logit contributions
gh-0.170
 time-0.167
 purpose-0.156
 grown-0.153
 chores-0.152
Tokeninkle
Token ID19894
Feature activation-2.35e+00

Pos logit contributions
 Sky+0.117
fly+0.113
 tweet+0.111
 globe+0.108
 deer+0.106
Neg logit contributions
gh-0.068
addy-0.056
 purpose-0.055
anked-0.055
omm-0.052
Token in
Token ID287
Feature activation-6.24e-01

Pos logit contributions
 Sky+0.086
stack+0.085
fly+0.085
 tweet+0.083
 globe+0.079
Neg logit contributions
gh-0.102
 purpose-0.098
 time-0.090
omm-0.089
 behaviour-0.088
Token the
Token ID262
Feature activation-1.57e+00

Pos logit contributions
fly+0.027
 Sky+0.026
 tweet+0.023
flies+0.023
 globe+0.022
Neg logit contributions
gh-0.027
 purpose-0.027
 mischief-0.025
array-0.025
 Sister-0.024
Token sky
Token ID6766
Feature activation-2.50e+00

Pos logit contributions
fly+0.066
 Sky+0.065
flies+0.058
 tweet+0.056
stack+0.055
Neg logit contributions
gh-0.067
 purpose-0.066
array-0.064
 mischief-0.062
 Sister-0.061
Token.
Token ID13
Feature activation-6.89e-01

Pos logit contributions
fly+0.002
 Sky+0.002
 tweet+0.002
 Saturn+0.002
 globe+0.002
Neg logit contributions
gh-0.002
 purpose-0.002
array-0.002
 behaviour-0.002
 Sister-0.002
Token He
Token ID679
Feature activation-1.25e+00

Pos logit contributions
fly+0.034
 Sky+0.032
 tweet+0.031
flies+0.029
 globe+0.029
Neg logit contributions
gh-0.025
 purpose-0.024
omm-0.024
 Sister-0.024
array-0.023
Token would
Token ID561
Feature activation-1.84e+00

Pos logit contributions
 Sky+0.060
fly+0.057
 tweet+0.056
 higher+0.052
 below+0.051
Neg logit contributions
gh-0.049
 forgotten-0.043
 purpose-0.040
 Sister-0.040
array-0.039
Token make
Token ID787
Feature activation-2.59e+00

Pos logit contributions
 Sky+0.079
fly+0.078
als+0.071
flies+0.070
 globe+0.068
Neg logit contributions
gh-0.075
 purpose-0.069
 forgotten-0.065
array-0.065
urance-0.065
Token the
Token ID262
Feature activation-2.86e+00

Pos logit contributions
fly+0.134
stack+0.119
bug+0.116
 tweet+0.114
 globe+0.113
Neg logit contributions
gh-0.098
 purpose-0.093
 mischief-0.091
 trouble-0.087
 business-0.082
Token rocket
Token ID10701
Feature activation-4.92e+00

Pos logit contributions
fly+0.114
bug+0.103
 Sky+0.103
 tweet+0.099
flies+0.096
Neg logit contributions
 purpose-0.139
 time-0.135
 trouble-0.129
 mischief-0.125
 consequences-0.125
Token fly
Token ID6129
Feature activation-3.21e+00

Pos logit contributions
 Sky+0.188
stack+0.184
als+0.182
 Flor+0.179
 Fly+0.176
Neg logit contributions
gh-0.179
 purpose-0.176
 product-0.151
 business-0.151
 time-0.151
Token high
Token ID1029
Feature activation-2.22e+00

Pos logit contributions
 Sky+0.123
 tweet+0.122
fly+0.112
 Flor+0.108
bug+0.107
Neg logit contributions
gh-0.146
 purpose-0.135
 time-0.130
 forgotten-0.120
 business-0.113
Token in
Token ID287
Feature activation-5.14e-01

Pos logit contributions
 tweet+0.087
 Sky+0.084
fly+0.081
 Saturn+0.071
 Fly+0.070
Neg logit contributions
gh-0.105
 purpose-0.099
array-0.090
 forgotten-0.090
anked-0.087
Token the
Token ID262
Feature activation-6.44e-01

Pos logit contributions
fly+0.018
 Sky+0.018
 tweet+0.017
flies+0.015
als+0.015
Neg logit contributions
gh-0.027
 purpose-0.026
array-0.024
 mischief-0.024
 forgotten-0.024
Token sky
Token ID6766
Feature activation-2.22e+00

Pos logit contributions
fly+0.026
 Sky+0.026
 tweet+0.023
flies+0.022
 Bloom+0.021
Neg logit contributions
gh-0.031
 purpose-0.029
array-0.029
 forgotten-0.027
 mischief-0.027
Token,
Token ID11
Feature activation-1.92e+00

Pos logit contributions
 Sky+0.003
fly+0.003
 tweet+0.002
 Saturn+0.002
 globe+0.002
Neg logit contributions
gh-0.003
 purpose-0.003
array-0.003
 behaviour-0.003
 Sister-0.002
Token I
Token ID314
Feature activation-1.01e-02

Pos logit contributions
fly+0.091
 Sky+0.081
 below+0.079
flies+0.077
 globe+0.077
Neg logit contributions
gh-0.068
 Tommy-0.065
 purpose-0.065
 punishment-0.062
 disob-0.061
Token can
Token ID460
Feature activation-1.45e+00

Pos logit contributions
 Sky+0.000
fly+0.000
 tweet+0.000
 Saturn+0.000
 globe+0.000
Neg logit contributions
gh-0.000
 purpose-0.000
array-0.000
 behaviour-0.000
 Sister-0.000
Token make
Token ID787
Feature activation-2.06e+00

Pos logit contributions
fly+0.061
 Sky+0.060
 tweet+0.054
flies+0.054
 globe+0.053
Neg logit contributions
gh-0.062
 purpose-0.056
 punish-0.051
 disob-0.050
urance-0.050
Token the
Token ID262
Feature activation-2.17e+00

Pos logit contributions
fly+0.099
 globe+0.088
 Sky+0.086
stack+0.086
 tweet+0.083
Neg logit contributions
gh-0.081
 purpose-0.080
 mischief-0.075
 trouble-0.071
array-0.069
Token moon
Token ID8824
Feature activation-4.91e+00

Pos logit contributions
fly+0.089
 Sky+0.085
bug+0.078
 tweet+0.070
 Fly+0.070
Neg logit contributions
 purpose-0.098
 mischief-0.088
 trouble-0.086
 laundry-0.085
array-0.085
Token go
Token ID467
Feature activation-2.36e+00

Pos logit contributions
 Sky+0.189
 Fly+0.189
stack+0.187
 Bloom+0.182
 Saturn+0.177
Neg logit contributions
gh-0.188
 purpose-0.173
 product-0.149
 punish-0.140
 behaviour-0.140
Token around
Token ID1088
Feature activation-1.89e+00

Pos logit contributions
 Sky+0.083
 tweet+0.082
fly+0.080
stack+0.076
 Dragon+0.074
Neg logit contributions
gh-0.107
 purpose-0.104
 forgotten-0.096
 time-0.096
array-0.092
Token the
Token ID262
Feature activation-1.32e+00

Pos logit contributions
fly+0.079
 tweet+0.075
 Sky+0.074
flies+0.067
 globe+0.066
Neg logit contributions
 purpose-0.082
gh-0.081
 mischief-0.075
array-0.074
 trouble-0.072
Token earth
Token ID4534
Feature activation-3.80e+00

Pos logit contributions
fly+0.051
 Sky+0.050
 tweet+0.046
flies+0.042
 Fly+0.040
Neg logit contributions
 purpose-0.063
gh-0.061
array-0.060
 mischief-0.058
 Sister-0.056
Token!"
Token ID2474
Feature activation-1.99e-01

Pos logit contributions
 Sky+0.190
 tweet+0.183
 Fly+0.176
 Saturn+0.174
 Jupiter+0.169
Neg logit contributions
 purpose-0.115
gh-0.110
array-0.096
em-0.092
 forgotten-0.091
Token the
Token ID262
Feature activation-2.05e+00

Pos logit contributions
fly+0.142
 tweet+0.122
 flap+0.120
flies+0.118
 globe+0.117
Neg logit contributions
 Sister-0.094
 Tommy-0.092
 time-0.089
 purpose-0.089
 punishment-0.088
Token little
Token ID1310
Feature activation-2.69e+00

Pos logit contributions
fly+0.131
flies+0.122
ites+0.121
als+0.120
stack+0.120
Neg logit contributions
 time-0.047
 punishment-0.041
 purpose-0.041
 trouble-0.036
gh-0.035
Token girl
Token ID2576
Feature activation-1.84e+00

Pos logit contributions
stack+0.147
fly+0.145
girls+0.141
 Saturn+0.140
ites+0.137
Neg logit contributions
 trouble-0.061
 time-0.058
 purpose-0.056
 Sister-0.054
 punishment-0.053
Token visited
Token ID8672
Feature activation-3.31e+00

Pos logit contributions
 Sky+0.082
fly+0.076
 tweet+0.073
 globe+0.072
 higher+0.071
Neg logit contributions
gh-0.073
 forgotten-0.069
 behaviour-0.067
 Sister-0.066
 supposed-0.065
Token the
Token ID262
Feature activation-2.25e+00

Pos logit contributions
fly+0.169
ite+0.153
flies+0.148
 globe+0.148
als+0.148
Neg logit contributions
 purpose-0.113
 time-0.106
gh-0.101
 business-0.092
 trouble-0.091
Token dragon
Token ID10441
Feature activation-4.90e+00

Pos logit contributions
fly+0.103
 tweet+0.097
ites+0.094
flies+0.092
als+0.092
Neg logit contributions
 purpose-0.088
 time-0.087
 situation-0.082
 grown-0.081
 forgotten-0.080
Token every
Token ID790
Feature activation-1.56e+00

Pos logit contributions
 higher+0.185
 globe+0.185
 Sky+0.183
 Saturn+0.182
fly+0.177
Neg logit contributions
gh-0.161
 business-0.160
 time-0.148
 behaviour-0.146
 efforts-0.141
Token day
Token ID1110
Feature activation-1.71e+00

Pos logit contributions
 Sky+0.085
fly+0.084
flies+0.082
als+0.080
 tweet+0.079
Neg logit contributions
gh-0.045
 time-0.044
 purpose-0.041
 behaviour-0.039
 punishment-0.039
Token and
Token ID290
Feature activation-3.12e+00

Pos logit contributions
fly+0.082
 Sky+0.079
 tweet+0.078
 globe+0.075
 flap+0.071
Neg logit contributions
gh-0.065
 forgotten-0.056
 purpose-0.056
 supposed-0.055
 behaviour-0.054
Token they
Token ID484
Feature activation-3.55e+00

Pos logit contributions
 globe+0.137
fly+0.135
 Sky+0.133
 higher+0.130
 south+0.130
Neg logit contributions
 happened-0.107
 forgotten-0.106
gh-0.101
 complained-0.099
 Sister-0.095
Token would
Token ID561
Feature activation-3.01e+00

Pos logit contributions
 Sky+0.184
 higher+0.180
fly+0.172
 globe+0.169
 Follow+0.169
Neg logit contributions
 forgotten-0.092
gh-0.091
 disob-0.082
 disagreed-0.081
 happened-0.080
Token.
Token ID13
Feature activation-2.94e-01

Pos logit contributions
gh+0.034
omm+0.032
 Sister+0.032
array+0.031
 purpose+0.031
Neg logit contributions
 Sky-0.026
fly-0.025
 tweet-0.022
 Saturn-0.022
 globe-0.022
Token They
Token ID1119
Feature activation-5.39e-01

Pos logit contributions
fly+0.011
 Sky+0.011
 tweet+0.010
flies+0.009
 globe+0.009
Neg logit contributions
gh-0.014
 purpose-0.014
array-0.013
 Sister-0.013
omm-0.013
Token would
Token ID561
Feature activation-1.34e+00

Pos logit contributions
fly+0.026
 Sky+0.026
 tweet+0.023
 globe+0.022
flies+0.022
Neg logit contributions
gh-0.021
 forgotten-0.019
 Sister-0.019
array-0.019
 behaviour-0.018
Token watch
Token ID2342
Feature activation-2.09e+00

Pos logit contributions
fly+0.066
 Sky+0.063
flies+0.058
 Dragon+0.057
als+0.057
Neg logit contributions
gh-0.049
 purpose-0.045
 forgotten-0.045
array-0.044
urance-0.044
Token the
Token ID262
Feature activation-9.61e-01

Pos logit contributions
fly+0.107
 Sky+0.090
flies+0.089
 tweet+0.088
bug+0.088
Neg logit contributions
gh-0.080
 purpose-0.077
 mischief-0.076
 Tommy-0.074
array-0.073
Token planes
Token ID13016
Feature activation-4.90e+00

Pos logit contributions
fly+0.039
 Sky+0.036
 tweet+0.032
flies+0.031
bug+0.030
Neg logit contributions
gh-0.047
 purpose-0.044
array-0.044
 forgotten-0.043
 behaviour-0.043
Token take
Token ID1011
Feature activation+2.39e-01

Pos logit contributions
 Sky+0.198
fly+0.189
 Flor+0.184
 Dragon+0.182
Sky+0.179
Neg logit contributions
 punish-0.166
 time-0.164
 delay-0.157
 afford-0.156
gh-0.156
Token off
Token ID572
Feature activation-8.29e-01

Pos logit contributions
gh+0.013
 purpose+0.012
array+0.012
omm+0.011
 behaviour+0.011
Neg logit contributions
 Sky-0.008
fly-0.008
 tweet-0.007
 Saturn-0.007
 globe-0.007
Token and
Token ID290
Feature activation-2.76e+00

Pos logit contributions
fly+0.031
 Sky+0.029
 tweet+0.028
flies+0.025
 globe+0.024
Neg logit contributions
gh-0.041
 purpose-0.040
 forgotten-0.037
array-0.037
 Sister-0.037
Token land
Token ID1956
Feature activation-1.26e+00

Pos logit contributions
fly+0.131
 Sky+0.126
rots+0.120
 globe+0.114
 tweet+0.112
Neg logit contributions
 forgotten-0.094
gh-0.091
 punish-0.088
 purpose-0.088
 Sister-0.087
Token.
Token ID13
Feature activation-7.17e-01

Pos logit contributions
fly+0.048
 Sky+0.046
 tweet+0.044
 Fly+0.038
bug+0.037
Neg logit contributions
gh-0.062
 purpose-0.057
 forgotten-0.057
 behaviour-0.054
 punishment-0.054
Token dad
Token ID9955
Feature activation+2.88e-01

Pos logit contributions
fly+0.025
 Sky+0.024
 tweet+0.023
 globe+0.022
 Saturn+0.022
Neg logit contributions
gh-0.011
 Sister-0.009
 purpose-0.009
array-0.009
 behaviour-0.009
Token.
Token ID13
Feature activation-1.45e-01

Pos logit contributions
gh+0.013
 purpose+0.012
array+0.012
 behaviour+0.011
omm+0.011
Neg logit contributions
fly-0.012
 Sky-0.012
 tweet-0.011
 globe-0.011
 below-0.010
Token They
Token ID1119
Feature activation-8.89e-01

Pos logit contributions
fly+0.006
 Sky+0.006
 tweet+0.005
 globe+0.005
 Saturn+0.005
Neg logit contributions
gh-0.007
 purpose-0.006
array-0.006
 Sister-0.006
omm-0.006
Token would
Token ID561
Feature activation-1.57e+00

Pos logit contributions
fly+0.041
 Sky+0.040
 tweet+0.036
flies+0.035
 higher+0.034
Neg logit contributions
gh-0.036
 forgotten-0.033
 Sister-0.033
array-0.031
kids-0.031
Token watch
Token ID2342
Feature activation-2.09e+00

Pos logit contributions
fly+0.081
 Sky+0.075
flies+0.071
 Dragon+0.070
als+0.069
Neg logit contributions
gh-0.054
 forgotten-0.048
array-0.048
urance-0.048
 purpose-0.047
Token planes
Token ID13016
Feature activation-4.89e+00

Pos logit contributions
fly+0.108
flies+0.087
 Sky+0.087
stack+0.086
 tweet+0.084
Neg logit contributions
gh-0.078
 purpose-0.077
 mischief-0.075
array-0.075
 time-0.075
Token take
Token ID1011
Feature activation-7.51e-02

Pos logit contributions
fly+0.222
 Flor+0.214
 bead+0.203
 Sky+0.202
bug+0.200
Neg logit contributions
 time-0.141
gh-0.128
 punish-0.123
 sale-0.121
 done-0.121
Token off
Token ID572
Feature activation-1.17e+00

Pos logit contributions
fly+0.003
 Sky+0.003
 tweet+0.003
 globe+0.003
 Saturn+0.002
Neg logit contributions
gh-0.004
 purpose-0.003
array-0.003
 behaviour-0.003
 Sister-0.003
Token and
Token ID290
Feature activation-2.91e+00

Pos logit contributions
fly+0.047
 Sky+0.041
 tweet+0.041
flies+0.037
 globe+0.037
Neg logit contributions
gh-0.054
 purpose-0.053
 forgotten-0.050
 Sister-0.050
array-0.050
Token land
Token ID1956
Feature activation-1.53e+00

Pos logit contributions
fly+0.140
 Sky+0.130
rots+0.129
 globe+0.120
fish+0.118
Neg logit contributions
 forgotten-0.096
gh-0.092
 punish-0.090
array-0.088
 purpose-0.088
Token.
Token ID13
Feature activation-1.00e+00

Pos logit contributions
fly+0.061
 Sky+0.054
 tweet+0.054
 Fly+0.048
bug+0.046
Neg logit contributions
gh-0.071
 purpose-0.068
 forgotten-0.067
em-0.063
 business-0.063
Token
Token ID198
Feature activation-1.52e+00

Pos logit contributions
fly+0.072
 tweet+0.063
flies+0.062
 below+0.061
fish+0.061
Neg logit contributions
 Sister-0.049
 purpose-0.047
gh-0.047
 Tommy-0.047
omm-0.046
Token
Token ID198
Feature activation-6.70e-01

Pos logit contributions
fly+0.072
 tweet+0.065
 Sky+0.064
flies+0.064
als+0.062
Neg logit contributions
gh-0.063
 purpose-0.058
addy-0.055
David-0.054
child-0.053
TokenThey
Token ID2990
Feature activation-1.24e+00

Pos logit contributions
fly+0.032
 Sky+0.031
 tweet+0.029
flies+0.028
 below+0.028
Neg logit contributions
gh-0.027
 purpose-0.024
array-0.024
 Sister-0.023
David-0.023
Token both
Token ID1111
Feature activation-6.00e-01

Pos logit contributions
 Sky+0.063
fly+0.062
 tweet+0.057
 globe+0.055
 Fly+0.055
Neg logit contributions
gh-0.043
 forgotten-0.040
array-0.037
 Sister-0.036
addy-0.036
Token drew
Token ID9859
Feature activation-3.51e+00

Pos logit contributions
 Sky+0.028
fly+0.027
 tweet+0.025
 globe+0.024
 Saturn+0.024
Neg logit contributions
gh-0.024
 forgotten-0.022
array-0.021
 purpose-0.021
 Sister-0.021
Token frogs
Token ID37475
Feature activation-4.89e+00

Pos logit contributions
stack+0.193
fly+0.191
 tweet+0.169
 globe+0.167
 Follow+0.166
Neg logit contributions
gh-0.085
 trouble-0.082
 time-0.081
 purpose-0.077
 Tommy-0.072
Token and
Token ID290
Feature activation-2.26e+00

Pos logit contributions
Flash+0.258
rob+0.236
 globe+0.228
 Asia+0.227
stack+0.224
Neg logit contributions
gh-0.081
 grown-0.075
 time-0.074
:-0.073
 --0.070
Token colored
Token ID16396
Feature activation-3.35e+00

Pos logit contributions
fly+0.080
 Sky+0.075
 globe+0.073
 below+0.073
 higher+0.068
Neg logit contributions
 Sister-0.098
 purpose-0.097
 mischief-0.094
gh-0.094
 forgotten-0.093
Token them
Token ID606
Feature activation-3.48e+00

Pos logit contributions
stack+0.146
 Follow+0.141
fly+0.141
 Sky+0.132
 tweet+0.130
Neg logit contributions
 purpose-0.107
gh-0.103
 time-0.095
anked-0.094
 Tommy-0.092
Token.
Token ID13
Feature activation-1.29e+00

Pos logit contributions
stack+0.156
 Follow+0.150
fly+0.145
 tweet+0.140
 south+0.140
Neg logit contributions
gh-0.117
 purpose-0.103
 time-0.099
 happened-0.089
 sale-0.087
Token They
Token ID1119
Feature activation-2.12e+00

Pos logit contributions
fly+0.063
 tweet+0.057
 Sky+0.056
fish+0.054
 globe+0.053
Neg logit contributions
 purpose-0.047
 Sister-0.047
urance-0.046
gh-0.045
array-0.044
Token�
Token ID129
Feature activation+1.70e+00

Pos logit contributions
gh+0.017
 purpose+0.016
array+0.016
 behaviour+0.015
 Sister+0.015
Neg logit contributions
 Sky-0.021
fly-0.021
 tweet-0.019
 globe-0.018
 Saturn-0.018
Token�
Token ID241
Feature activation+2.63e+00

Pos logit contributions
gh+0.058
 behaviour+0.056
 punishment+0.055
 policy+0.054
 purpose+0.054
Neg logit contributions
 Sky-0.082
fly-0.077
 below-0.076
 Saturn-0.076
 Jupiter-0.075
TokenI
Token ID40
Feature activation+1.06e+00

Pos logit contributions
gh+0.140
omm+0.138
array+0.136
urance+0.136
 behaviour+0.132
Neg logit contributions
 Sky-0.074
fly-0.072
 below-0.068
 tweet-0.067
 globe-0.065
Token'm
Token ID1101
Feature activation+8.87e-01

Pos logit contributions
gh+0.054
array+0.052
 purpose+0.052
omm+0.051
urance+0.050
Neg logit contributions
fly-0.037
 Sky-0.035
 tweet-0.033
 soar-0.030
 below-0.029
Token not
Token ID407
Feature activation+1.56e+00

Pos logit contributions
gh+0.044
 purpose+0.042
array+0.041
 behaviour+0.041
urance+0.039
Neg logit contributions
 Sky-0.033
fly-0.031
 tweet-0.029
 Saturn-0.027
 higher-0.027
Token hungry
Token ID14720
Feature activation+2.37e+00

Pos logit contributions
gh+0.079
omm+0.071
anked+0.071
 purpose+0.070
urance+0.070
Neg logit contributions
fly-0.055
 Sky-0.053
 tweet-0.049
 higher-0.049
 below-0.049
Token,
Token ID11
Feature activation+3.32e-01

Pos logit contributions
gh+0.130
array+0.125
omm+0.125
innie+0.124
David+0.124
Neg logit contributions
 Sky-0.065
 higher-0.065
fly-0.062
 below-0.060
 around-0.058
Token I
Token ID314
Feature activation+1.52e+00

Pos logit contributions
gh+0.018
 purpose+0.016
array+0.016
 behaviour+0.016
urance+0.016
Neg logit contributions
 Sky-0.011
fly-0.011
 tweet-0.010
 Saturn-0.009
 globe-0.009
Token want
Token ID765
Feature activation+1.59e+00

Pos logit contributions
gh+0.080
array+0.077
 purpose+0.077
omm+0.076
urance+0.074
Neg logit contributions
fly-0.051
 Sky-0.045
 tweet-0.044
 soar-0.039
flies-0.039
Token to
Token ID284
Feature activation+1.82e+00

Pos logit contributions
gh+0.089
urance+0.087
array+0.085
anked+0.083
omm+0.082
Neg logit contributions
 Sky-0.050
fly-0.043
 Saturn-0.039
 tweet-0.038
als-0.038
Token sing
Token ID1702
Feature activation-6.06e-01

Pos logit contributions
gh+0.100
omm+0.096
array+0.095
em+0.094
anked+0.094
Neg logit contributions
 Sky-0.053
fly-0.052
 soar-0.049
 tweet-0.048
 flap-0.047
Token him
Token ID683
Feature activation+1.32e+00

Pos logit contributions
gh+0.066
array+0.060
 supposed+0.058
 purpose+0.057
 forgotten+0.057
Neg logit contributions
 Sky-0.046
fly-0.040
 tweet-0.038
 Saturn-0.036
 Jupiter-0.035
Token that
Token ID326
Feature activation-2.90e-01

Pos logit contributions
gh+0.073
array+0.071
omm+0.068
 purpose+0.068
urance+0.067
Neg logit contributions
 Sky-0.043
fly-0.040
 tweet-0.035
 Jupiter-0.033
flies-0.033
Token he
Token ID339
Feature activation+6.44e-01

Pos logit contributions
fly+0.010
 Sky+0.010
 tweet+0.009
 globe+0.008
 Saturn+0.008
Neg logit contributions
gh-0.015
 purpose-0.014
 behaviour-0.014
array-0.014
 Sister-0.014
Token had
Token ID550
Feature activation+2.87e+00

Pos logit contributions
gh+0.031
 purpose+0.029
array+0.029
omm+0.028
urance+0.028
Neg logit contributions
fly-0.025
 Sky-0.024
 tweet-0.022
flies-0.021
 globe-0.020
Token been
Token ID587
Feature activation+2.13e+00

Pos logit contributions
omm+0.168
gh+0.155
innie+0.155
addy+0.151
urance+0.151
Neg logit contributions
 Sky-0.073
 together-0.069
fly-0.068
 tweet-0.065
 flies-0.062
Token naughty
Token ID45128
Feature activation+2.15e+00

Pos logit contributions
urance+0.102
omm+0.102
gh+0.101
array+0.099
 purpose+0.097
Neg logit contributions
 together-0.073
 Sky-0.072
fly-0.070
 higher-0.069
 around-0.066
Token.
Token ID13
Feature activation+1.04e+00

Pos logit contributions
gh+0.147
omm+0.136
 purpose+0.134
liv+0.134
addy+0.132
Neg logit contributions
fly-0.039
 Sky-0.038
 below-0.035
 around-0.035
 together-0.034
Token They
Token ID1119
Feature activation-1.69e-02

Pos logit contributions
gh+0.061
 behaviour+0.054
 purpose+0.053
array+0.053
 forgotten+0.053
Neg logit contributions
 Sky-0.034
fly-0.031
 Saturn-0.027
 Jupiter-0.027
 tweet-0.025
Token told
Token ID1297
Feature activation+1.25e+00

Pos logit contributions
 Sky+0.001
fly+0.001
 tweet+0.001
 globe+0.001
 Saturn+0.001
Neg logit contributions
gh-0.001
 purpose-0.001
array-0.001
 behaviour-0.001
 Sister-0.001
Token him
Token ID683
Feature activation+1.55e+00

Pos logit contributions
gh+0.066
array+0.059
 purpose+0.058
 supposed+0.057
urance+0.057
Neg logit contributions
 Sky-0.048
fly-0.041
 Saturn-0.039
 tweet-0.038
 Jupiter-0.037
Token that
Token ID326
Feature activation+1.36e-01

Pos logit contributions
gh+0.085
array+0.085
omm+0.081
 purpose+0.080
urance+0.080
Neg logit contributions
 Sky-0.049
fly-0.046
 tweet-0.039
flies-0.037
 Jupiter-0.037
Token mom
Token ID1995
Feature activation+1.28e+00

Pos logit contributions
gh+0.083
array+0.082
em+0.078
anked+0.076
urance+0.076
Neg logit contributions
 Sky-0.048
fly-0.041
 tweet-0.038
 deer-0.037
 dragon-0.036
Token wrapped
Token ID12908
Feature activation-8.18e-01

Pos logit contributions
 purpose+0.067
gh+0.065
 mischief+0.064
 behaviour+0.061
array+0.060
Neg logit contributions
fly-0.046
 Sky-0.046
 tweet-0.042
bug-0.039
 deer-0.037
Token Lucy
Token ID22162
Feature activation-1.31e-01

Pos logit contributions
fly+0.038
 Sky+0.035
 tweet+0.035
 soar+0.033
 Saturn+0.032
Neg logit contributions
gh-0.032
 purpose-0.031
 Sister-0.030
omm-0.029
array-0.029
Token in
Token ID287
Feature activation+1.82e+00

Pos logit contributions
fly+0.003
 Sky+0.003
 tweet+0.003
 Saturn+0.003
 globe+0.003
Neg logit contributions
gh-0.008
 purpose-0.008
array-0.007
 behaviour-0.007
 Sister-0.007
Token her
Token ID607
Feature activation+1.22e+00

Pos logit contributions
gh+0.113
 behaviour+0.105
omm+0.103
scar+0.103
urance+0.102
Neg logit contributions
 Sky-0.048
 flap-0.039
 Saturn-0.038
 Jupiter-0.036
 higher-0.036
Token warm
Token ID5814
Feature activation+2.19e+00

Pos logit contributions
gh+0.054
array+0.050
omm+0.050
urance+0.050
David+0.049
Neg logit contributions
 Sky-0.050
fly-0.047
 tweet-0.045
 globe-0.044
 flap-0.044
Token blanket
Token ID18447
Feature activation-6.43e-01

Pos logit contributions
 supposed+0.098
gh+0.098
 forgotten+0.095
David+0.093
 difficulty+0.090
Neg logit contributions
 Sky-0.083
 flap-0.078
 tweet-0.077
 globe-0.077
fly-0.076
Token and
Token ID290
Feature activation+5.89e-02

Pos logit contributions
fly+0.023
 Sky+0.022
 tweet+0.021
 Saturn+0.019
 globe+0.018
Neg logit contributions
gh-0.033
 purpose-0.031
array-0.030
 forgotten-0.030
 Sister-0.030
Token gave
Token ID2921
Feature activation+1.73e-02

Pos logit contributions
gh+0.003
 purpose+0.002
array+0.002
 behaviour+0.002
omm+0.002
Neg logit contributions
 Sky-0.003
fly-0.003
 tweet-0.002
 Saturn-0.002
 globe-0.002
Token her
Token ID607
Feature activation+1.10e+00

Pos logit contributions
gh+0.001
 purpose+0.001
array+0.001
 behaviour+0.001
 Sister+0.001
Neg logit contributions
fly-0.001
 Sky-0.001
 tweet-0.001
 Saturn-0.001
 globe-0.001
Token a
Token ID257
Feature activation+7.46e-01

Pos logit contributions
gh+0.065
array+0.060
omm+0.059
urance+0.058
 purpose+0.058
Neg logit contributions
 Sky-0.034
fly-0.030
 tweet-0.027
 higher-0.026
 Saturn-0.026
Token she
Token ID673
Feature activation-3.43e-01

Pos logit contributions
gh+0.100
array+0.098
omm+0.097
urance+0.096
 purpose+0.094
Neg logit contributions
 Sky-0.030
fly-0.027
 below-0.025
 higher-0.023
 deer-0.022
Token said
Token ID531
Feature activation+9.73e-01

Pos logit contributions
 Sky+0.016
fly+0.016
 tweet+0.015
 globe+0.014
 Saturn+0.014
Neg logit contributions
gh-0.014
 Sister-0.012
 purpose-0.012
 forgotten-0.012
array-0.012
Token,
Token ID11
Feature activation+2.83e-02

Pos logit contributions
array+0.051
gh+0.049
 mischief+0.048
 purpose+0.048
kids+0.047
Neg logit contributions
 Sky-0.035
fly-0.032
 tweet-0.031
 higher-0.028
 below-0.027
Token "
Token ID366
Feature activation+2.87e-01

Pos logit contributions
gh+0.001
 purpose+0.001
array+0.001
 behaviour+0.001
 Sister+0.001
Neg logit contributions
 Sky-0.001
fly-0.001
 tweet-0.001
 Saturn-0.001
 globe-0.001
TokenI
Token ID40
Feature activation+7.51e-01

Pos logit contributions
gh+0.012
 purpose+0.011
array+0.011
 behaviour+0.011
omm+0.011
Neg logit contributions
fly-0.013
 Sky-0.013
 tweet-0.012
 below-0.011
 globe-0.011
Token have
Token ID423
Feature activation+2.40e+00

Pos logit contributions
gh+0.040
array+0.038
 purpose+0.038
omm+0.037
urance+0.037
Neg logit contributions
fly-0.026
 Sky-0.024
 tweet-0.023
 flap-0.020
 higher-0.020
Token a
Token ID257
Feature activation+1.43e+00

Pos logit contributions
omm+0.132
gh+0.131
urance+0.130
array+0.125
anked+0.125
Neg logit contributions
 Sky-0.068
 higher-0.059
 tweet-0.058
 together-0.057
fly-0.057
Token y
Token ID331
Feature activation+2.62e+00

Pos logit contributions
gh+0.065
array+0.064
anked+0.061
em+0.061
urance+0.061
Neg logit contributions
 Sky-0.058
fly-0.057
 tweet-0.056
 higher-0.053
 flap-0.052
Tokenummy
Token ID13513
Feature activation+1.72e+00

Pos logit contributions
David+0.096
gh+0.092
 forgotten+0.089
array+0.087
 Sister+0.087
Neg logit contributions
fly-0.114
 Sky-0.109
 higher-0.107
 tweet-0.106
 flap-0.106
Token muff
Token ID27563
Feature activation-8.34e-01

Pos logit contributions
gh+0.077
David+0.074
array+0.074
 Dylan+0.072
omm+0.072
Neg logit contributions
 tweet-0.068
 Sky-0.066
fly-0.063
 fly-0.060
 soar-0.060
Tokenin
Token ID259
Feature activation-5.43e-01

Pos logit contributions
 Sky+0.032
fly+0.031
 Jupiter+0.029
 Saturn+0.029
 tweet+0.028
Neg logit contributions
gh-0.042
addy-0.037
em-0.037
 behaviour-0.037
omm-0.036
Token what
Token ID644
Feature activation+2.08e+00

Pos logit contributions
fly+0.018
 Sky+0.018
 tweet+0.016
 globe+0.015
 Saturn+0.015
Neg logit contributions
gh-0.026
 purpose-0.024
 behaviour-0.023
 Sister-0.023
array-0.023
Token she
Token ID673
Feature activation+1.20e+00

Pos logit contributions
urance+0.128
array+0.122
 behaviour+0.118
omm+0.117
kids+0.117
Neg logit contributions
 Sky-0.054
fly-0.042
 Bloom-0.042
 together-0.042
 Saturn-0.041
Token wanted
Token ID2227
Feature activation+2.33e+00

Pos logit contributions
gh+0.062
 purpose+0.060
urance+0.056
omm+0.056
 behaviour+0.055
Neg logit contributions
fly-0.043
 Sky-0.040
 tweet-0.037
 below-0.036
flies-0.035
Token to
Token ID284
Feature activation+2.59e+00

Pos logit contributions
uf+0.140
 Sister+0.139
urance+0.137
gh+0.135
David+0.135
Neg logit contributions
 together-0.052
 around-0.051
 Sky-0.049
fly-0.046
 under-0.043
Token do
Token ID466
Feature activation+2.19e+00

Pos logit contributions
uf+0.150
gh+0.150
 Sister+0.149
omm+0.148
 forgotten+0.146
Neg logit contributions
 fly-0.057
 together-0.056
 Sky-0.054
 flap-0.054
 soar-0.052
Token to
Token ID284
Feature activation+2.66e+00

Pos logit contributions
 Sister+0.127
omm+0.120
uf+0.120
array+0.120
urance+0.120
Neg logit contributions
 together-0.061
 Sky-0.055
 around-0.052
fly-0.049
 higher-0.047
Token make
Token ID787
Feature activation+2.05e+00

Pos logit contributions
omm+0.156
gh+0.152
 behaviour+0.150
 Sister+0.149
uf+0.149
Neg logit contributions
 fly-0.068
 soar-0.063
 flap-0.063
 sing-0.057
 Sky-0.056
Token her
Token ID607
Feature activation+5.42e-01

Pos logit contributions
gh+0.114
urance+0.113
scar+0.108
uf+0.108
 forgotten+0.108
Neg logit contributions
 together-0.058
 Sky-0.053
 higher-0.049
 Saturn-0.047
 around-0.045
Token room
Token ID2119
Feature activation+1.13e+00

Pos logit contributions
gh+0.026
 purpose+0.023
array+0.023
omm+0.023
 Sister+0.022
Neg logit contributions
 Sky-0.022
fly-0.021
 tweet-0.019
 higher-0.018
 Saturn-0.018
Token better
Token ID1365
Feature activation+9.88e-01

Pos logit contributions
gh+0.050
 Sister+0.045
omm+0.045
 behaviour+0.044
urance+0.044
Neg logit contributions
 Sky-0.046
fly-0.046
 higher-0.044
 tweet-0.043
 below-0.042
Token.
Token ID13
Feature activation+1.38e+00

Pos logit contributions
gh+0.049
array+0.046
omm+0.045
urance+0.045
 Sister+0.045
Neg logit contributions
fly-0.035
 Sky-0.034
 together-0.031
 tweet-0.030
 higher-0.029
TokenMom
Token ID29252
Feature activation+1.20e+00

Pos logit contributions
gh+0.022
 behaviour+0.020
 purpose+0.020
urance+0.020
omm+0.019
Neg logit contributions
fly-0.022
 Sky-0.021
 tweet-0.020
 globe-0.019
 below-0.018
Tokenmy
Token ID1820
Feature activation+1.94e+00

Pos logit contributions
gh+0.070
 mischief+0.066
 purpose+0.066
 behaviour+0.065
array+0.064
Neg logit contributions
fly-0.036
 Sky-0.033
 tweet-0.031
 globe-0.026
flies-0.026
Token,
Token ID11
Feature activation-3.03e-01

Pos logit contributions
gh+0.116
 mischief+0.114
ern+0.111
 purpose+0.110
array+0.110
Neg logit contributions
fly-0.052
 tweet-0.047
 Sky-0.046
 together-0.037
 globe-0.037
Token my
Token ID616
Feature activation+1.57e+00

Pos logit contributions
fly+0.011
 Sky+0.011
 tweet+0.010
 below+0.009
 Saturn+0.009
Neg logit contributions
gh-0.015
 purpose-0.014
 Sister-0.014
array-0.014
 behaviour-0.014
Token thumb
Token ID15683
Feature activation+1.34e+00

Pos logit contributions
gh+0.070
array+0.069
kids+0.066
urance+0.066
ern+0.065
Neg logit contributions
fly-0.060
 tweet-0.056
 globe-0.056
 Sky-0.055
 flap-0.052
Token hurts
Token ID20406
Feature activation+9.97e-01

Pos logit contributions
gh+0.073
 mischief+0.070
array+0.069
 Sister+0.069
 purpose+0.069
Neg logit contributions
fly-0.045
 Sky-0.040
 tweet-0.038
 globe-0.035
flies-0.035
Token."
Token ID526
Feature activation+1.33e+00

Pos logit contributions
gh+0.057
 behaviour+0.053
array+0.053
addy+0.052
omm+0.052
Neg logit contributions
fly-0.031
 Sky-0.029
 tweet-0.026
 higher-0.024
 globe-0.023
Token
Token ID198
Feature activation+1.31e-01

Pos logit contributions
gh+0.081
 behaviour+0.078
 forgotten+0.077
 supposed+0.077
 mischief+0.077
Neg logit contributions
 Sky-0.034
fly-0.030
 together-0.025
 tweet-0.025
 globe-0.024
Token
Token ID198
Feature activation+1.84e+00

Pos logit contributions
gh+0.007
 purpose+0.007
array+0.007
 behaviour+0.007
 Sister+0.007
Neg logit contributions
 Sky-0.004
fly-0.004
 tweet-0.004
 Saturn-0.003
 globe-0.003
TokenHis
Token ID6653
Feature activation+8.88e-01

Pos logit contributions
 behaviour+0.098
 mischief+0.097
 difficulty+0.096
 consequences+0.094
oken+0.094
Neg logit contributions
 Sky-0.056
fly-0.052
 together-0.051
 globe-0.049
pole-0.046
Token mom
Token ID1995
Feature activation+5.14e-01

Pos logit contributions
gh+0.042
array+0.039
em+0.038
urance+0.038
 purpose+0.037
Neg logit contributions
 Sky-0.035
fly-0.033
 tweet-0.030
 globe-0.030
 deer-0.029
Token your
Token ID534
Feature activation-8.57e-01

Pos logit contributions
fly+0.044
 Sky+0.040
 tweet+0.037
flies+0.037
 below+0.036
Neg logit contributions
gh-0.044
 purpose-0.040
addy-0.037
 Sister-0.037
array-0.036
Token hat
Token ID6877
Feature activation-1.29e+00

Pos logit contributions
fly+0.038
 Sky+0.037
 below+0.032
 tweet+0.032
flies+0.032
Neg logit contributions
gh-0.039
 purpose-0.035
 behaviour-0.033
 Sister-0.033
array-0.032
Token.
Token ID13
Feature activation-1.29e+00

Pos logit contributions
fly+0.063
 Sky+0.063
 tweet+0.060
 Fly+0.056
 Saturn+0.056
Neg logit contributions
gh-0.050
 purpose-0.045
 forgotten-0.040
 behaviour-0.040
addy-0.040
Token It
Token ID632
Feature activation-2.44e+00

Pos logit contributions
fly+0.061
 Sky+0.057
flies+0.053
 tweet+0.052
 below+0.052
Neg logit contributions
gh-0.050
 Sister-0.047
omm-0.047
 purpose-0.047
urance-0.046
Token is
Token ID318
Feature activation-3.29e-01

Pos logit contributions
 below+0.119
fly+0.116
 Saturn+0.112
 Jupiter+0.109
 Sky+0.109
Neg logit contributions
gh-0.076
 disob-0.071
 been-0.070
 forgotten-0.066
 laundry-0.063
Token very
Token ID845
Feature activation-9.51e-02

Pos logit contributions
fly+0.014
 Sky+0.014
 tweet+0.013
 globe+0.012
 below+0.012
Neg logit contributions
gh-0.015
 purpose-0.013
array-0.013
 forgotten-0.013
 Sister-0.013
Token warm
Token ID5814
Feature activation-8.95e-01

Pos logit contributions
fly+0.004
 Sky+0.004
 tweet+0.004
 globe+0.004
 Saturn+0.003
Neg logit contributions
gh-0.004
 purpose-0.004
array-0.004
 behaviour-0.004
 Sister-0.004
Token and
Token ID290
Feature activation-1.71e+00

Pos logit contributions
fly+0.039
 Sky+0.039
 tweet+0.037
 Saturn+0.034
 Jupiter+0.033
Neg logit contributions
gh-0.038
 purpose-0.036
array-0.036
 behaviour-0.034
 Sister-0.034
Token cozy
Token ID37438
Feature activation-9.39e-01

Pos logit contributions
fly+0.088
 Sky+0.085
 tweet+0.077
flies+0.075
 below+0.074
Neg logit contributions
gh-0.062
 forgotten-0.057
 purpose-0.056
 Sister-0.054
 supposed-0.053
Token.
Token ID13
Feature activation-1.85e+00

Pos logit contributions
fly+0.046
 Sky+0.046
 tweet+0.044
 Saturn+0.041
 Jupiter+0.040
Neg logit contributions
gh-0.034
 purpose-0.033
array-0.031
 forgotten-0.031
 behaviour-0.030
Token Thank
Token ID6952
Feature activation-2.32e-01

Pos logit contributions
fly+0.089
 Sky+0.079
 below+0.077
 globe+0.077
flies+0.076
Neg logit contributions
 Sister-0.069
omm-0.067
gh-0.066
urance-0.064
 purpose-0.064
Token.
Token ID13
Feature activation-2.04e-01

Pos logit contributions
fly+0.031
 Sky+0.029
 tweet+0.028
flies+0.025
 Saturn+0.024
Neg logit contributions
gh-0.052
 purpose-0.049
array-0.048
 forgotten-0.048
 Sister-0.047
Token From
Token ID3574
Feature activation-7.75e-02

Pos logit contributions
fly+0.008
 Sky+0.008
 tweet+0.007
 globe+0.007
 Saturn+0.007
Neg logit contributions
gh-0.010
 purpose-0.009
array-0.009
 Sister-0.009
omm-0.009
Token then
Token ID788
Feature activation+2.31e-01

Pos logit contributions
fly+0.003
 Sky+0.003
 tweet+0.002
 globe+0.002
 Saturn+0.002
Neg logit contributions
gh-0.004
 purpose-0.004
array-0.004
 behaviour-0.004
 Sister-0.004
Token on
Token ID319
Feature activation+6.84e-01

Pos logit contributions
gh+0.016
 purpose+0.015
array+0.015
omm+0.015
 behaviour+0.015
Neg logit contributions
 Sky-0.004
fly-0.004
 tweet-0.003
 Saturn-0.003
 globe-0.003
Token,
Token ID11
Feature activation-1.15e+00

Pos logit contributions
gh+0.034
array+0.032
 purpose+0.031
omm+0.031
urance+0.031
Neg logit contributions
 Sky-0.026
fly-0.024
 tweet-0.022
 globe-0.021
 Saturn-0.020
Token the
Token ID262
Feature activation-2.68e-01

Pos logit contributions
fly+0.048
 Sky+0.043
 tweet+0.042
 globe+0.041
 flap+0.040
Neg logit contributions
gh-0.052
 purpose-0.049
 Sister-0.048
 behaviour-0.047
 forgotten-0.046
Token friend
Token ID1545
Feature activation+6.04e-01

Pos logit contributions
 Sky+0.011
fly+0.011
 tweet+0.010
 Saturn+0.009
 globe+0.009
Neg logit contributions
gh-0.012
 purpose-0.011
 behaviour-0.011
 Sister-0.011
array-0.011
Token visited
Token ID8672
Feature activation-4.64e-01

Pos logit contributions
gh+0.028
 purpose+0.026
array+0.025
omm+0.025
urance+0.025
Neg logit contributions
 Sky-0.025
fly-0.025
 tweet-0.022
 below-0.021
 Saturn-0.021
Token the
Token ID262
Feature activation-7.60e-01

Pos logit contributions
fly+0.016
 Sky+0.015
 tweet+0.014
 globe+0.013
 Saturn+0.013
Neg logit contributions
gh-0.025
 purpose-0.023
 behaviour-0.022
 Sister-0.022
array-0.022
Token drum
Token ID13026
Feature activation-1.95e+00

Pos logit contributions
 Sky+0.031
fly+0.031
 tweet+0.028
 Saturn+0.026
 globe+0.026
Neg logit contributions
gh-0.034
 purpose-0.033
 behaviour-0.031
array-0.031
 Sister-0.031
Token every
Token ID790
Feature activation+8.62e-01

Pos logit contributions
 Sky+0.066
fly+0.066
 tweet+0.059
 Saturn+0.056
 globe+0.056
Neg logit contributions
gh-0.101
 purpose-0.092
 behaviour-0.091
urance-0.088
array-0.088
Token and
Token ID290
Feature activation+1.72e-01

Pos logit contributions
gh+0.038
 purpose+0.035
omm+0.035
array+0.035
 Sister+0.034
Neg logit contributions
 Sky-0.024
fly-0.024
 tweet-0.021
 below-0.021
 higher-0.020
Token didn
Token ID1422
Feature activation+1.50e+00

Pos logit contributions
gh+0.008
 purpose+0.007
array+0.007
omm+0.007
 behaviour+0.007
Neg logit contributions
 Sky-0.007
fly-0.007
 tweet-0.006
 Saturn-0.006
 globe-0.006
Token't
Token ID470
Feature activation+5.94e-01

Pos logit contributions
array+0.072
gh+0.071
 behaviour+0.070
kids+0.069
child+0.067
Neg logit contributions
 Sky-0.059
fly-0.053
 soar-0.048
 higher-0.048
 Fly-0.047
Token pay
Token ID1414
Feature activation+6.37e-01

Pos logit contributions
gh+0.030
omm+0.028
array+0.028
 purpose+0.027
addy+0.027
Neg logit contributions
fly-0.022
 Sky-0.022
 tweet-0.020
 flap-0.019
 soar-0.019
Token attention
Token ID3241
Feature activation+1.15e+00

Pos logit contributions
gh+0.035
array+0.033
omm+0.033
 purpose+0.033
urance+0.032
Neg logit contributions
 Sky-0.020
fly-0.019
 tweet-0.017
 higher-0.017
 below-0.016
Token to
Token ID284
Feature activation+6.68e-01

Pos logit contributions
gh+0.056
array+0.055
omm+0.053
 purpose+0.053
 Sister+0.052
Neg logit contributions
 Sky-0.042
fly-0.042
 below-0.037
 tweet-0.037
 higher-0.036
Token what
Token ID644
Feature activation+1.64e+00

Pos logit contributions
gh+0.034
array+0.031
omm+0.031
 purpose+0.030
 behaviour+0.030
Neg logit contributions
 Sky-0.025
fly-0.024
 tweet-0.022
 Saturn-0.021
 below-0.021
Token he
Token ID339
Feature activation+6.31e-01

Pos logit contributions
omm+0.094
urance+0.092
array+0.092
addy+0.088
gh+0.087
Neg logit contributions
 Sky-0.051
fly-0.047
 tweet-0.039
 below-0.038
 soar-0.037
Token was
Token ID373
Feature activation+1.29e+00

Pos logit contributions
gh+0.029
 purpose+0.027
omm+0.026
array+0.026
 Sister+0.026
Neg logit contributions
fly-0.027
 Sky-0.026
 tweet-0.024
 below-0.022
flies-0.022
Token doing
Token ID1804
Feature activation+1.78e+00

Pos logit contributions
omm+0.054
gh+0.053
 purpose+0.051
 Sister+0.051
addy+0.049
Neg logit contributions
fly-0.055
 Sky-0.052
 below-0.052
 higher-0.051
 tweet-0.050
Token.
Token ID13
Feature activation+7.91e-01

Pos logit contributions
omm+0.089
 Sister+0.089
gh+0.088
array+0.086
addy+0.086
Neg logit contributions
fly-0.056
 Sky-0.055
 together-0.052
 below-0.052
 higher-0.050
Token playing
Token ID2712
Feature activation-4.86e-01

Pos logit contributions
gh+0.023
 purpose+0.021
 behaviour+0.021
array+0.021
omm+0.020
Neg logit contributions
 Sky-0.023
fly-0.023
 tweet-0.021
 Jupiter-0.020
 Saturn-0.020
Token in
Token ID287
Feature activation+4.53e-01

Pos logit contributions
fly+0.016
 Sky+0.015
 tweet+0.014
 Saturn+0.013
flies+0.012
Neg logit contributions
gh-0.027
 purpose-0.025
array-0.025
 mischief-0.024
 behaviour-0.024
Token the
Token ID262
Feature activation+2.06e-01

Pos logit contributions
gh+0.026
 behaviour+0.024
omm+0.024
 purpose+0.024
urance+0.024
Neg logit contributions
 Sky-0.014
fly-0.013
 tweet-0.011
 Saturn-0.011
 globe-0.011
Token garden
Token ID11376
Feature activation+5.36e-01

Pos logit contributions
gh+0.010
 purpose+0.009
 behaviour+0.009
array+0.009
omm+0.009
Neg logit contributions
 Sky-0.008
fly-0.008
 tweet-0.007
 globe-0.007
 Saturn-0.007
Token.
Token ID13
Feature activation+8.71e-01

Pos logit contributions
gh+0.025
 purpose+0.023
omm+0.023
 Sister+0.023
 behaviour+0.023
Neg logit contributions
 Sky-0.021
fly-0.021
 tweet-0.019
 below-0.019
 globe-0.018
Token He
Token ID679
Feature activation+5.20e-01

Pos logit contributions
gh+0.050
 purpose+0.045
 behaviour+0.044
omm+0.044
 forgotten+0.043
Neg logit contributions
 Sky-0.027
fly-0.025
 below-0.023
 Saturn-0.023
 Jupiter-0.022
Token loved
Token ID6151
Feature activation+2.01e-01

Pos logit contributions
gh+0.026
 purpose+0.025
omm+0.024
urance+0.024
 behaviour+0.024
Neg logit contributions
 Sky-0.019
fly-0.019
 tweet-0.017
 globe-0.016
 Saturn-0.016
Token to
Token ID284
Feature activation+7.56e-01

Pos logit contributions
gh+0.009
 purpose+0.008
array+0.008
omm+0.008
 behaviour+0.008
Neg logit contributions
 Sky-0.008
fly-0.008
 tweet-0.008
 Saturn-0.007
 globe-0.007
Token dig
Token ID3100
Feature activation-1.13e+00

Pos logit contributions
gh+0.043
omm+0.041
 purpose+0.039
 behaviour+0.039
 Sister+0.039
Neg logit contributions
 Sky-0.023
fly-0.022
 tweet-0.021
 flap-0.020
 soar-0.019
Token holes
Token ID10421
Feature activation-1.89e+00

Pos logit contributions
fly+0.043
 Sky+0.041
 tweet+0.038
flies+0.033
 Saturn+0.033
Neg logit contributions
gh-0.059
 purpose-0.054
array-0.052
 mischief-0.052
anked-0.051
Token with
Token ID351
Feature activation-1.04e-01

Pos logit contributions
 Sky+0.080
fly+0.080
 tweet+0.077
 Saturn+0.068
 Jupiter+0.066
Neg logit contributions
gh-0.083
 purpose-0.072
anked-0.071
 forgotten-0.071
 time-0.069
Token.
Token ID13
Feature activation+1.57e+00

Pos logit contributions
urance+0.071
gh+0.071
array+0.071
omm+0.070
 Sister+0.068
Neg logit contributions
 Sky-0.050
fly-0.048
 higher-0.043
 below-0.043
 together-0.043
Token M
Token ID337
Feature activation+2.44e+00

Pos logit contributions
gh+0.102
 behaviour+0.089
 purpose+0.088
 disob+0.086
 forgotten+0.085
Neg logit contributions
 Sky-0.042
fly-0.040
 below-0.035
 Jupiter-0.034
 Saturn-0.033
Tokenummy
Token ID13513
Feature activation+1.22e+00

Pos logit contributions
gh+0.098
 forgotten+0.097
 purpose+0.090
 difficulty+0.089
 mischief+0.087
Neg logit contributions
fly-0.108
 Sky-0.103
flies-0.101
bug-0.093
 globe-0.092
Token asked
Token ID1965
Feature activation+1.64e+00

Pos logit contributions
gh+0.063
 purpose+0.059
urance+0.056
array+0.056
 behaviour+0.055
Neg logit contributions
fly-0.046
 Sky-0.045
 tweet-0.039
 below-0.038
 deer-0.038
Token Tim
Token ID5045
Feature activation+5.58e-02

Pos logit contributions
gh+0.090
array+0.087
 purpose+0.085
urance+0.083
 supposed+0.082
Neg logit contributions
 Sky-0.052
fly-0.048
 together-0.045
 tweet-0.044
 Jupiter-0.041
Tokenmy
Token ID1820
Feature activation-4.80e-01

Pos logit contributions
gh+0.003
 purpose+0.003
array+0.002
 behaviour+0.002
 Sister+0.002
Neg logit contributions
 Sky-0.002
fly-0.002
 tweet-0.002
 Saturn-0.002
 globe-0.002
Token to
Token ID284
Feature activation+1.92e+00

Pos logit contributions
fly+0.019
 Sky+0.019
 tweet+0.017
 globe+0.016
 Saturn+0.016
Neg logit contributions
gh-0.023
 purpose-0.021
 behaviour-0.020
omm-0.020
 Sister-0.020
Token pass
Token ID1208
Feature activation+3.08e-01

Pos logit contributions
gh+0.103
array+0.097
omm+0.097
 behaviour+0.095
 purpose+0.093
Neg logit contributions
 Sky-0.061
fly-0.056
 flap-0.055
 soar-0.052
 fly-0.052
Token her
Token ID607
Feature activation+3.76e-01

Pos logit contributions
gh+0.016
 purpose+0.015
array+0.014
 behaviour+0.014
urance+0.014
Neg logit contributions
 Sky-0.011
fly-0.011
 tweet-0.010
 Saturn-0.009
 below-0.009
Token a
Token ID257
Feature activation+4.58e-01

Pos logit contributions
gh+0.018
array+0.017
 purpose+0.017
omm+0.016
 behaviour+0.016
Neg logit contributions
 Sky-0.014
fly-0.014
 tweet-0.013
 globe-0.012
 Saturn-0.012
Token h
Token ID289
Feature activation+1.60e+00

Pos logit contributions
gh+0.021
array+0.020
 purpose+0.019
omm+0.019
 behaviour+0.019
Neg logit contributions
fly-0.018
 Sky-0.018
 tweet-0.017
 globe-0.016
 Saturn-0.015
Token continue
Token ID2555
Feature activation+6.70e-01

Pos logit contributions
fly+0.004
 Sky+0.004
 tweet+0.004
 globe+0.004
 Saturn+0.004
Neg logit contributions
gh-0.005
 purpose-0.005
array-0.005
 behaviour-0.005
 Sister-0.005
Token the
Token ID262
Feature activation+1.86e-01

Pos logit contributions
gh+0.035
omm+0.034
array+0.033
 purpose+0.033
 behaviour+0.032
Neg logit contributions
 Sky-0.022
fly-0.022
 tweet-0.019
 Saturn-0.018
 below-0.018
Token fun
Token ID1257
Feature activation+8.61e-01

Pos logit contributions
gh+0.009
 purpose+0.008
array+0.008
 behaviour+0.008
omm+0.008
Neg logit contributions
 Sky-0.007
fly-0.007
 tweet-0.007
 globe-0.006
 Saturn-0.006
Token.
Token ID13
Feature activation-2.51e-02

Pos logit contributions
gh+0.042
omm+0.040
 Sister+0.039
array+0.038
urance+0.038
Neg logit contributions
 Sky-0.032
fly-0.031
 below-0.029
 globe-0.028
 higher-0.027
Token<|endoftext|>
Token ID50256
Feature activation-7.02e-01

Pos logit contributions
fly+0.001
 Sky+0.001
 tweet+0.001
 globe+0.001
 Saturn+0.001
Neg logit contributions
gh-0.001
 purpose-0.001
array-0.001
 Sister-0.001
 behaviour-0.001
TokenOnce
Token ID7454
Feature activation-8.86e-01

Pos logit contributions
fly+0.026
 Sky+0.025
 tweet+0.023
flies+0.021
 Saturn+0.021
Neg logit contributions
gh-0.035
 purpose-0.032
array-0.032
 Sister-0.031
omm-0.031
Token upon
Token ID2402
Feature activation-2.49e-01

Pos logit contributions
fly+0.034
 Sky+0.033
 tweet+0.031
flies+0.028
als+0.028
Neg logit contributions
gh-0.044
 purpose-0.040
addy-0.039
 behaviour-0.039
 Sister-0.039
Token a
Token ID257
Feature activation+2.11e+00

Pos logit contributions
fly+0.010
 Sky+0.010
 tweet+0.009
 globe+0.009
 Saturn+0.009
Neg logit contributions
gh-0.012
 purpose-0.011
array-0.010
 behaviour-0.010
 forgotten-0.010
Token time
Token ID640
Feature activation+4.58e-02

Pos logit contributions
omm+0.050
array+0.050
 behaviour+0.050
 Sister+0.049
gh+0.048
Neg logit contributions
 Sky-0.121
fly-0.113
 tweet-0.111
 higher-0.109
 globe-0.107
Token there
Token ID612
Feature activation+7.13e-01

Pos logit contributions
gh+0.002
 purpose+0.002
array+0.002
 behaviour+0.002
 Sister+0.002
Neg logit contributions
 Sky-0.002
fly-0.002
 tweet-0.002
 Saturn-0.002
 globe-0.002
Token was
Token ID373
Feature activation+1.25e-01

Pos logit contributions
gh+0.035
array+0.033
 purpose+0.033
 behaviour+0.032
urance+0.032
Neg logit contributions
 Sky-0.027
fly-0.027
 tweet-0.023
 globe-0.023
 below-0.022
Tokenie
Token ID494
Feature activation-1.06e+00

Pos logit contributions
gh+0.038
array+0.033
urance+0.033
omm+0.033
 purpose+0.033
Neg logit contributions
fly-0.027
 Sky-0.027
 tweet-0.025
 below-0.023
 higher-0.023
Token.
Token ID13
Feature activation-1.05e+00

Pos logit contributions
 Sky+0.044
fly+0.043
 tweet+0.040
 Saturn+0.039
 globe+0.038
Neg logit contributions
gh-0.047
 purpose-0.044
 behaviour-0.042
 Sister-0.041
 forgotten-0.041
Token The
Token ID383
Feature activation+4.33e-01

Pos logit contributions
fly+0.047
 Sky+0.045
 below+0.042
 globe+0.041
 higher+0.041
Neg logit contributions
gh-0.042
omm-0.041
 Sister-0.041
 purpose-0.040
urance-0.040
Token cloth
Token ID16270
Feature activation-1.75e+00

Pos logit contributions
gh+0.021
array+0.018
 purpose+0.018
urance+0.018
 behaviour+0.018
Neg logit contributions
 Sky-0.017
fly-0.017
 tweet-0.016
 globe-0.015
 higher-0.015
Token is
Token ID318
Feature activation+8.06e-03

Pos logit contributions
 Sky+0.063
fly+0.058
 Saturn+0.055
 tweet+0.054
 Jupiter+0.053
Neg logit contributions
gh-0.087
 purpose-0.081
 behaviour-0.080
 forgotten-0.079
array-0.077
Token soaked
Token ID33657
Feature activation-1.15e+00

Pos logit contributions
gh+0.000
 purpose+0.000
array+0.000
 behaviour+0.000
 Sister+0.000
Neg logit contributions
 Sky-0.000
fly-0.000
 tweet-0.000
 globe-0.000
 Saturn-0.000
Token in
Token ID287
Feature activation+7.38e-01

Pos logit contributions
 Sky+0.044
fly+0.043
 tweet+0.040
 Saturn+0.037
 globe+0.035
Neg logit contributions
gh-0.056
 purpose-0.054
array-0.051
 behaviour-0.050
 Sister-0.050
Token cold
Token ID4692
Feature activation+7.06e-01

Pos logit contributions
gh+0.043
 behaviour+0.039
urance+0.039
omm+0.039
array+0.038
Neg logit contributions
 Sky-0.022
fly-0.022
 tweet-0.019
 Saturn-0.018
 Jupiter-0.018
Token water
Token ID1660
Feature activation-2.90e-01

Pos logit contributions
gh+0.037
omm+0.033
 purpose+0.033
 behaviour+0.033
addy+0.033
Neg logit contributions
 Sky-0.025
fly-0.025
 tweet-0.022
 globe-0.021
 below-0.021
Token.
Token ID13
Feature activation-1.31e+00

Pos logit contributions
 Sky+0.012
fly+0.012
 tweet+0.011
 Saturn+0.010
 globe+0.010
Neg logit contributions
gh-0.013
 purpose-0.012
array-0.012
 behaviour-0.012
 Sister-0.012
Token It
Token ID632
Feature activation-2.06e+00

Pos logit contributions
fly+0.058
 Sky+0.056
 below+0.052
 globe+0.052
 tweet+0.051
Neg logit contributions
 Sister-0.052
omm-0.052
gh-0.052
urance-0.050
 purpose-0.049
Token<|endoftext|>
Token ID50256
Feature activation-4.50e-02

Pos logit contributions
gh+0.006
 purpose+0.005
array+0.005
 behaviour+0.005
 Sister+0.005
Neg logit contributions
 Sky-0.003
fly-0.003
 tweet-0.003
 Saturn-0.003
 globe-0.003
TokenOnce
Token ID7454
Feature activation-7.58e-01

Pos logit contributions
fly+0.001
 Sky+0.001
 tweet+0.001
 Saturn+0.001
 globe+0.001
Neg logit contributions
gh-0.002
 purpose-0.002
array-0.002
 behaviour-0.002
 Sister-0.002
Token upon
Token ID2402
Feature activation+1.13e-01

Pos logit contributions
fly+0.029
 Sky+0.028
 tweet+0.026
flies+0.024
 globe+0.023
Neg logit contributions
gh-0.037
 purpose-0.034
addy-0.034
 behaviour-0.034
 Sister-0.033
Token a
Token ID257
Feature activation+2.10e+00

Pos logit contributions
gh+0.006
 purpose+0.005
array+0.005
 behaviour+0.005
omm+0.005
Neg logit contributions
 Sky-0.004
fly-0.004
 tweet-0.004
 Saturn-0.004
 globe-0.004
Token time
Token ID640
Feature activation+1.41e-01

Pos logit contributions
omm+0.049
 behaviour+0.048
gh+0.048
array+0.048
 Sister+0.048
Neg logit contributions
 Sky-0.121
fly-0.114
 tweet-0.112
 higher-0.108
 globe-0.107
Token,
Token ID11
Feature activation-1.48e+00

Pos logit contributions
gh+0.007
 purpose+0.006
array+0.006
 behaviour+0.006
 Sister+0.006
Neg logit contributions
 Sky-0.006
fly-0.006
 tweet-0.005
 globe-0.005
 Saturn-0.005
Token there
Token ID612
Feature activation+6.58e-01

Pos logit contributions
fly+0.061
 Sky+0.060
als+0.057
 tweet+0.056
flies+0.056
Neg logit contributions
gh-0.064
 time-0.057
 purpose-0.056
 happened-0.056
 Tommy-0.054
Token were
Token ID547
Feature activation+5.46e-01

Pos logit contributions
gh+0.032
array+0.030
 purpose+0.030
 behaviour+0.029
urance+0.029
Neg logit contributions
 Sky-0.025
fly-0.025
 tweet-0.022
 globe-0.021
 below-0.021
Token two
Token ID734
Feature activation+1.18e+00

Pos logit contributions
gh+0.026
 purpose+0.024
 behaviour+0.023
array+0.023
urance+0.023
Neg logit contributions
 Sky-0.022
fly-0.021
 tweet-0.019
 Saturn-0.018
 Jupiter-0.018
Token friends
Token ID2460
Feature activation+8.38e-01

Pos logit contributions
gh+0.064
 purpose+0.059
array+0.058
urance+0.057
 behaviour+0.056
Neg logit contributions
 Sky-0.041
fly-0.037
 tweet-0.035
 deer-0.034
 planets-0.032
Token called
Token ID1444
Feature activation+8.42e-01

Pos logit contributions
gh+0.047
 purpose+0.044
array+0.043
omm+0.042
 behaviour+0.042
Neg logit contributions
fly-0.027
 Sky-0.027
 tweet-0.024
 together-0.022
 Saturn-0.022
Token and
Token ID290
Feature activation-1.71e+00

Pos logit contributions
 Sky+0.010
fly+0.010
 tweet+0.009
 Saturn+0.008
 globe+0.008
Neg logit contributions
gh-0.014
 purpose-0.013
 behaviour-0.013
array-0.013
 Sister-0.013
Token less
Token ID1342
Feature activation+8.10e-01

Pos logit contributions
fly+0.078
 Sky+0.076
 globe+0.071
 tweet+0.068
als+0.067
Neg logit contributions
gh-0.072
 forgotten-0.070
 supposed-0.064
 punishment-0.064
 purpose-0.062
Token clumsy
Token ID39120
Feature activation+7.07e-01

Pos logit contributions
gh+0.040
omm+0.036
 purpose+0.035
array+0.035
urance+0.035
Neg logit contributions
fly-0.032
 Sky-0.032
 below-0.029
 higher-0.028
 tweet-0.028
Token.
Token ID13
Feature activation+1.48e-01

Pos logit contributions
gh+0.035
omm+0.032
array+0.032
 purpose+0.032
 Sister+0.032
Neg logit contributions
 Sky-0.026
fly-0.026
 tweet-0.023
 below-0.022
 higher-0.022
Token And
Token ID843
Feature activation-1.23e+00

Pos logit contributions
gh+0.007
 purpose+0.007
array+0.007
 behaviour+0.007
 Sister+0.007
Neg logit contributions
 Sky-0.006
fly-0.006
 tweet-0.005
 globe-0.005
 Saturn-0.005
Token he
Token ID339
Feature activation-1.09e+00

Pos logit contributions
fly+0.052
 Sky+0.050
 tweet+0.048
 flap+0.045
als+0.044
Neg logit contributions
gh-0.055
 purpose-0.053
 punishment-0.052
 behaviour-0.051
 forgotten-0.051
Token learned
Token ID4499
Feature activation+4.51e-02

Pos logit contributions
 Sky+0.059
fly+0.057
 tweet+0.054
 Saturn+0.052
 south+0.050
Neg logit contributions
gh-0.038
 forgotten-0.032
 purpose-0.031
 Sister-0.031
 behaviour-0.031
Token that
Token ID326
Feature activation-4.92e-01

Pos logit contributions
gh+0.002
 purpose+0.002
array+0.002
 behaviour+0.002
 Sister+0.002
Neg logit contributions
 Sky-0.002
fly-0.002
 tweet-0.001
 Saturn-0.001
 globe-0.001
Token it
Token ID340
Feature activation-1.17e+00

Pos logit contributions
fly+0.020
 Sky+0.019
 tweet+0.017
als+0.016
 globe+0.016
Neg logit contributions
gh-0.024
 purpose-0.022
 behaviour-0.021
array-0.021
 mischief-0.021
Token's
Token ID338
Feature activation+5.57e-03

Pos logit contributions
 Sky+0.052
fly+0.051
 tweet+0.050
 Saturn+0.046
 globe+0.045
Neg logit contributions
gh-0.049
 behaviour-0.043
 purpose-0.043
omm-0.043
 supposed-0.043
Token not
Token ID407
Feature activation+9.61e-01

Pos logit contributions
gh+0.000
 purpose+0.000
array+0.000
 behaviour+0.000
 Sister+0.000
Neg logit contributions
 Sky-0.000
fly-0.000
 tweet-0.000
 globe-0.000
 Saturn-0.000
Token.
Token ID13
Feature activation-1.86e+00

Pos logit contributions
 Sky+0.014
fly+0.013
 tweet+0.012
 globe+0.012
 Saturn+0.012
Neg logit contributions
gh-0.013
 purpose-0.012
array-0.012
 behaviour-0.012
 forgotten-0.012
Token You
Token ID921
Feature activation-1.33e+00

Pos logit contributions
fly+0.088
fish+0.082
 below+0.082
 Sky+0.082
 higher+0.080
Neg logit contributions
omm-0.070
 Sister-0.066
urance-0.063
gh-0.062
 purpose-0.062
Token can
Token ID460
Feature activation-4.05e-01

Pos logit contributions
 Sky+0.051
fly+0.048
 below+0.043
 tweet+0.043
 globe+0.042
Neg logit contributions
gh-0.061
 forgotten-0.061
 disob-0.059
 punish-0.058
 purpose-0.056
Token put
Token ID1234
Feature activation-2.41e+00

Pos logit contributions
 Sky+0.017
fly+0.017
 tweet+0.015
 Saturn+0.015
 globe+0.014
Neg logit contributions
gh-0.019
 purpose-0.017
array-0.016
omm-0.016
 Sister-0.016
Token the
Token ID262
Feature activation-1.06e+00

Pos logit contributions
fly+0.115
 Sky+0.108
flies+0.100
 tweet+0.100
 Jupiter+0.100
Neg logit contributions
gh-0.094
 purpose-0.091
 trouble-0.080
em-0.080
 Sister-0.079
Token butter
Token ID9215
Feature activation-3.45e+00

Pos logit contributions
 Sky+0.047
fly+0.046
 Jupiter+0.042
 Saturn+0.042
 tweet+0.042
Neg logit contributions
gh-0.045
 purpose-0.043
 behaviour-0.040
 Sister-0.039
 forgotten-0.039
Token on
Token ID319
Feature activation-6.29e-01

Pos logit contributions
 Sky+0.133
 Jupiter+0.130
 Saturn+0.129
 Fly+0.122
 Venus+0.120
Neg logit contributions
gh-0.158
 purpose-0.141
em-0.134
 forgotten-0.130
illy-0.126
Token it
Token ID340
Feature activation-2.21e+00

Pos logit contributions
fly+0.023
 Sky+0.022
 tweet+0.020
flies+0.019
 globe+0.019
Neg logit contributions
gh-0.032
 purpose-0.031
 Sister-0.029
array-0.029
omm-0.029
Token.
Token ID13
Feature activation-1.62e+00

Pos logit contributions
fly+0.092
 Sky+0.089
 tweet+0.086
 Fly+0.079
 Saturn+0.079
Neg logit contributions
gh-0.099
 purpose-0.092
 forgotten-0.087
array-0.084
 punishment-0.084
Token And
Token ID843
Feature activation-2.51e+00

Pos logit contributions
fly+0.075
 Sky+0.070
fish+0.068
 below+0.068
 globe+0.068
Neg logit contributions
omm-0.062
 Sister-0.060
gh-0.059
urance-0.058
 purpose-0.058
Token then
Token ID788
Feature activation-9.54e-01

Pos logit contributions
fly+0.116
 globe+0.106
 Sky+0.104
 below+0.097
 planets+0.094
Neg logit contributions
gh-0.094
 purpose-0.091
 punishment-0.089
 disob-0.087
 Sister-0.085
Token it
Token ID340
Feature activation+3.74e-03

Pos logit contributions
fly+0.010
 Sky+0.010
 tweet+0.009
 globe+0.009
flies+0.008
Neg logit contributions
gh-0.015
 purpose-0.014
array-0.013
 behaviour-0.013
 mischief-0.013
Token.
Token ID13
Feature activation+5.22e-01

Pos logit contributions
gh+0.000
 purpose+0.000
array+0.000
 behaviour+0.000
 Sister+0.000
Neg logit contributions
 Sky-0.000
fly-0.000
 tweet-0.000
 Saturn-0.000
 globe-0.000
Token She
Token ID1375
Feature activation-3.84e-01

Pos logit contributions
gh+0.028
 purpose+0.025
 behaviour+0.025
array+0.024
 forgotten+0.024
Neg logit contributions
 Sky-0.019
fly-0.018
 tweet-0.017
 Saturn-0.016
 Bloom-0.015
Token followed
Token ID3940
Feature activation-1.07e+00

Pos logit contributions
fly+0.019
 Sky+0.019
 tweet+0.017
 globe+0.017
 Saturn+0.017
Neg logit contributions
gh-0.014
 Sister-0.012
 forgotten-0.012
array-0.012
 purpose-0.012
Token the
Token ID262
Feature activation-1.90e-01

Pos logit contributions
fly+0.045
 Sky+0.041
flies+0.039
 tweet+0.038
 globe+0.038
Neg logit contributions
gh-0.048
 purpose-0.047
 Sister-0.045
 mischief-0.044
 behaviour-0.043
Token butterfly
Token ID35113
Feature activation-3.10e+00

Pos logit contributions
fly+0.005
 Sky+0.005
 tweet+0.004
 Saturn+0.004
 globe+0.004
Neg logit contributions
gh-0.011
 purpose-0.011
array-0.011
 behaviour-0.011
 Sister-0.010
Token to
Token ID284
Feature activation-1.74e+00

Pos logit contributions
 globe+0.155
fly+0.153
 Dragon+0.151
 Sky+0.151
 Saturn+0.149
Neg logit contributions
 business-0.077
gh-0.075
 behaviour-0.073
 purpose-0.073
 forgotten-0.072
Token a
Token ID257
Feature activation-6.95e-01

Pos logit contributions
fly+0.085
 globe+0.077
 Sky+0.074
flies+0.074
ites+0.072
Neg logit contributions
 purpose-0.070
 mischief-0.065
 trouble-0.061
gh-0.061
 punishment-0.061
Token pond
Token ID16723
Feature activation-6.28e-01

Pos logit contributions
fly+0.026
 Sky+0.026
 tweet+0.023
flies+0.022
 globe+0.022
Neg logit contributions
gh-0.034
 purpose-0.033
 forgotten-0.031
array-0.030
 mischief-0.030
Token where
Token ID810
Feature activation-1.67e+00

Pos logit contributions
 Sky+0.028
fly+0.027
 tweet+0.025
 Saturn+0.024
 globe+0.023
Neg logit contributions
gh-0.027
 purpose-0.025
 forgotten-0.025
array-0.024
 behaviour-0.024
Token she
Token ID673
Feature activation-6.31e-01

Pos logit contributions
fly+0.071
 Sky+0.062
 globe+0.060
 tweet+0.058
flies+0.055
Neg logit contributions
 forgotten-0.071
gh-0.071
 happened-0.069
 purpose-0.068
 mischief-0.068
Token "
Token ID366
Feature activation+1.88e-01

Pos logit contributions
fly+0.022
 Sky+0.022
 tweet+0.019
flies+0.019
 Saturn+0.019
Neg logit contributions
gh-0.022
 Sister-0.020
 purpose-0.020
omm-0.020
 behaviour-0.019
TokenMom
Token ID29252
Feature activation+1.38e+00

Pos logit contributions
gh+0.008
 purpose+0.007
array+0.007
 behaviour+0.007
omm+0.007
Neg logit contributions
 Sky-0.008
fly-0.008
 tweet-0.007
 globe-0.007
 Saturn-0.007
Token,
Token ID11
Feature activation-1.72e+00

Pos logit contributions
gh+0.076
 mischief+0.072
 purpose+0.071
array+0.070
 behaviour+0.069
Neg logit contributions
fly-0.044
 Sky-0.042
 tweet-0.040
 higher-0.035
 globe-0.034
Token can
Token ID460
Feature activation-1.02e+00

Pos logit contributions
fly+0.073
 Sky+0.068
flies+0.065
 below+0.062
 valleys+0.059
Neg logit contributions
gh-0.077
 punishment-0.074
 behaviour-0.069
 forgotten-0.069
 Sister-0.069
Token we
Token ID356
Feature activation-8.10e-01

Pos logit contributions
fly+0.039
 Sky+0.036
 tweet+0.033
 below+0.033
flies+0.032
Neg logit contributions
gh-0.051
 Sister-0.046
 purpose-0.046
omm-0.045
addy-0.044
Token come
Token ID1282
Feature activation-3.19e+00

Pos logit contributions
 Sky+0.037
fly+0.036
 tweet+0.033
 globe+0.031
 Saturn+0.031
Neg logit contributions
gh-0.034
 purpose-0.031
 forgotten-0.031
 Sister-0.030
addy-0.029
Token back
Token ID736
Feature activation-2.10e+00

Pos logit contributions
 Sky+0.153
fly+0.149
stack+0.138
flies+0.136
als+0.134
Neg logit contributions
gh-0.102
 forgotten-0.100
 Tommy-0.100
 supposed-0.096
 purpose-0.093
Token to
Token ID284
Feature activation-1.51e+00

Pos logit contributions
fly+0.101
 Sky+0.096
 tweet+0.094
 globe+0.089
flies+0.087
Neg logit contributions
addy-0.076
 Sister-0.075
 purpose-0.075
gh-0.075
 forgotten-0.074
Token the
Token ID262
Feature activation-1.17e-01

Pos logit contributions
fly+0.069
 Sky+0.063
flies+0.059
 globe+0.058
 tweet+0.058
Neg logit contributions
 purpose-0.062
 mischief-0.061
gh-0.060
 Sister-0.059
 forgotten-0.059
Token park
Token ID3952
Feature activation-2.30e+00

Pos logit contributions
fly+0.005
 Sky+0.005
 tweet+0.005
 Saturn+0.004
flies+0.004
Neg logit contributions
gh-0.005
 purpose-0.005
array-0.005
 behaviour-0.005
 Sister-0.005
Token tomorrow
Token ID9439
Feature activation-3.29e-01

Pos logit contributions
 Sky+0.096
fly+0.092
 globe+0.083
 tweet+0.083
flies+0.082
Neg logit contributions
gh-0.100
 forgotten-0.094
 purpose-0.092
addy-0.091
 punishment-0.090
Token the
Token ID262
Feature activation-1.12e+00

Pos logit contributions
fly+0.077
 Sky+0.071
 globe+0.070
flies+0.067
als+0.064
Neg logit contributions
gh-0.072
anked-0.066
 purpose-0.066
 punishment-0.063
 Sister-0.062
Token bird
Token ID6512
Feature activation-3.89e+00

Pos logit contributions
 Sky+0.036
fly+0.035
 Saturn+0.030
 Jupiter+0.029
 Flor+0.028
Neg logit contributions
 purpose-0.058
gh-0.058
array-0.055
 Sister-0.055
 behaviour-0.054
Token soar
Token ID48701
Feature activation-3.11e+00

Pos logit contributions
 globe+0.166
 Sky+0.154
 Dragon+0.145
rob+0.144
 Asia+0.144
Neg logit contributions
gh-0.130
 behaviour-0.127
 purpose-0.121
 mischief-0.121
 gotten-0.120
Token above
Token ID2029
Feature activation-1.32e+00

Pos logit contributions
 Sky+0.125
 Dragon+0.117
 tweet+0.114
 Dragons+0.113
 globe+0.111
Neg logit contributions
gh-0.121
 time-0.117
 purpose-0.111
 forgotten-0.101
anked-0.099
Token the
Token ID262
Feature activation-1.21e+00

Pos logit contributions
fly+0.062
 Sky+0.059
 tweet+0.055
flies+0.055
 globe+0.053
Neg logit contributions
gh-0.053
 purpose-0.048
 mischief-0.045
 behaviour-0.043
 Sister-0.043
Token trees
Token ID7150
Feature activation-3.38e+00

Pos logit contributions
 Sky+0.045
fly+0.044
 tweet+0.040
 Saturn+0.037
 Jupiter+0.036
Neg logit contributions
 purpose-0.058
array-0.055
 Sister-0.054
gh-0.054
 mischief-0.053
Token and
Token ID290
Feature activation-3.89e+00

Pos logit contributions
 Sky+0.173
 Saturn+0.162
 globe+0.162
 Dragon+0.161
fly+0.156
Neg logit contributions
gh-0.086
array-0.076
 forgotten-0.075
urance-0.074
anked-0.073
Token he
Token ID339
Feature activation-2.07e+00

Pos logit contributions
 globe+0.173
ite+0.165
rots+0.165
ites+0.161
 Sky+0.161
Neg logit contributions
 complained-0.120
 forgotten-0.118
 refused-0.110
 happened-0.110
 gotten-0.107
Token wanted
Token ID2227
Feature activation+1.64e-01

Pos logit contributions
 Sky+0.101
 globe+0.096
fly+0.094
stack+0.088
 tweet+0.088
Neg logit contributions
gh-0.071
 forgotten-0.068
 complained-0.062
 happened-0.061
 disagreed-0.061
Token to
Token ID284
Feature activation-1.58e+00

Pos logit contributions
gh+0.008
 purpose+0.007
array+0.007
omm+0.007
 behaviour+0.007
Neg logit contributions
 Sky-0.007
fly-0.007
 tweet-0.006
 Saturn-0.006
 globe-0.006
Token fly
Token ID6129
Feature activation-2.33e+00

Pos logit contributions
 Sky+0.055
fly+0.055
 globe+0.052
flies+0.049
 Jupiter+0.048
Neg logit contributions
 purpose-0.076
gh-0.076
 punish-0.070
 disob-0.070
 mischief-0.069
Token it
Token ID340
Feature activation-1.47e+00

Pos logit contributions
fly+0.087
 Sky+0.079
flies+0.078
als+0.075
 tweet+0.070
Neg logit contributions
gh-0.095
 purpose-0.082
 mischief-0.074
 time-0.074
 trouble-0.072
Token.
Token ID13
Feature activation+2.27e-01

Pos logit contributions
fly+0.080
 Sky+0.078
 tweet+0.073
 globe+0.073
als+0.072
Neg logit contributions
gh-0.052
 purpose-0.043
 forgotten-0.040
 behaviour-0.038
 supposed-0.038
Token He
Token ID679
Feature activation-1.64e-01

Pos logit contributions
gh+0.012
 purpose+0.011
array+0.011
 behaviour+0.011
omm+0.010
Neg logit contributions
 Sky-0.008
fly-0.008
 tweet-0.007
 Saturn-0.007
 below-0.007
Token asked
Token ID1965
Feature activation-8.37e-01

Pos logit contributions
 Sky+0.008
fly+0.008
 tweet+0.007
 globe+0.007
 Saturn+0.007
Neg logit contributions
gh-0.007
 purpose-0.006
array-0.006
 Sister-0.006
 behaviour-0.006
Token the
Token ID262
Feature activation-9.20e-01

Pos logit contributions
fly+0.042
 Sky+0.037
flies+0.037
 globe+0.036
 tweet+0.035
Neg logit contributions
gh-0.034
 behaviour-0.029
 purpose-0.029
 punishment-0.029
 Sister-0.028
Token bird
Token ID6512
Feature activation-3.17e+00

Pos logit contributions
fly+0.033
 Sky+0.031
flies+0.029
 tweet+0.028
als+0.028
Neg logit contributions
 purpose-0.044
gh-0.044
 Sister-0.041
 disob-0.040
 behaviour-0.040
Token if
Token ID611
Feature activation-1.13e+00

Pos logit contributions
 globe+0.150
als+0.138
 Sky+0.138
 valleys+0.137
 launcher+0.134
Neg logit contributions
 behaviour-0.096
gh-0.096
 punishment-0.089
 adult-0.089
 permission-0.089
Token he
Token ID339
Feature activation-2.20e-01

Pos logit contributions
fly+0.045
 Sky+0.042
 tweet+0.040
 globe+0.039
flies+0.037
Neg logit contributions
gh-0.052
 forgotten-0.048
 Sister-0.047
 punishment-0.047
 purpose-0.047
Token could
Token ID714
Feature activation-4.27e-01

Pos logit contributions
 Sky+0.007
fly+0.007
 tweet+0.006
 globe+0.006
 Saturn+0.006
Neg logit contributions
gh-0.012
 purpose-0.011
 behaviour-0.011
 forgotten-0.011
 Sister-0.011
Token try
Token ID1949
Feature activation-8.96e-01

Pos logit contributions
 Sky+0.017
fly+0.017
 tweet+0.015
 globe+0.015
 Saturn+0.014
Neg logit contributions
gh-0.021
 purpose-0.019
 behaviour-0.018
 forgotten-0.018
 Sister-0.018
Token it
Token ID340
Feature activation-1.09e+00

Pos logit contributions
fly+0.039
 Sky+0.039
 tweet+0.035
 globe+0.034
flies+0.033
Neg logit contributions
gh-0.040
 purpose-0.036
 Sister-0.034
 forgotten-0.034
 mischief-0.034