TOP ACTIVATIONS
MAX = 11.541

to wear her cozy bath robe after taking a bath .
Tim saw that his bath robe was too tight . He
they always threw the bath robe in the air when they
- a brand new bath robe . It was soft ,
; he has a bath robe too ." Her
are you wearing the bath robe ?" Cl are
happy and wore the bath robe all day .
And they loved their bath robe , too . <|endoftext|> Anna
. They used the bath robe to pretend they were superheroes
to mark where the bath robe should be cut . He
too full . The bath robe was too big .
had a lovely blue bath robe . Lucy loved to play
about we use the bath robe to make a tent in
he took off his bath robe and took a bath .
to take off her bath robe and wear her p aj
his bag and his bath robe . " Ben
anyone had lost a bath robe . After asking around ,
They see the pink bath robe . They laugh and shake
always wore the orange bath robe . Lily tried to explain
was a big red bath robe that belonged to their dad

MIN ACTIVATIONS
MIN = -10.707

tree with leaves and a coconut ," Tom said . He
reached out and grabbed a coconut . He was so happy
and showed his mom the coconut . She was proud of
, he saw a big coconut tree with lots of coc
the tree to get a coconut , but she was scared
how can we get the coconut ?" Tom had
picked a coconut . The coconut was heavy , but she
the tree and picked a coconut . The coconut was heavy
food . He found some coconut on the beach which he
, it 's a real coconut !" Tom said . He
had to know all about microsc opes .
if she could talk about microsc opes . Her
are welcome , kids . Enjoy your ol ives . But
are sweet and juicy . Enjoy them and stay warm ,"
you go , Anna . Enjoy your tuna !" Mom says
. Fresh and healthy . Enjoy ," he says .
Here are your toys . Enjoy them , but remember to
you go , Spot . Enjoy your meal ," Ben says
This is your reward . Enjoy !" Tim and
. Here you go . Enjoy !" The

INTERVAL 5.979 - 11.541
CONTAINS 0.177%

were dressed in their best clothes . There were balloons and
. Tom and Mia wear hats and coats , but they
They took off the band age and let Lily run in
can take her with you wherever you go ."
Sam noticed a navy sail boat in the distance . They

INTERVAL 0.417 - 5.979
CONTAINS 33.513%

The little girl was full of energy and happy to
. He thought that it would be fun to carry them
" It 's okay , my loves . I
work together . They can help each other and the people
the park and she heard a strange language . She was

INTERVAL -5.145 - 0.417
CONTAINS 65.988%

mom my disappeared again . The girl felt sad and wanted
mom made him a healthy dinner and they talked about how
it was a rainbow and it reminded her to be happy
faster and louder . He went towards L ila . He
. The screen lit up , and his show began to

INTERVAL -10.707 - -5.145
CONTAINS 0.323%

he knew he shouldn 't touch things that aren 't his
The little girl saw a pe b ble . She decided
M oral of the story : It is better to be
the kitchen to get some yogurt . They each got a
pond before , so he pe ered in , blinking in
Token to
Token ID284
Feature activation-1.04e+00

Pos logit contributions
 coconut+0.148
 Enjoy+0.140
ours+0.124
oral+0.123
iw+0.122
Neg logit contributions
aces-0.150
coat-0.149
robe-0.146
amas-0.143
brush-0.140
Token wear
Token ID5806
Feature activation+1.37e+00

Pos logit contributions
 coconut+0.067
 Enjoy+0.061
ours+0.055
grown+0.055
oral+0.054
Neg logit contributions
aces-0.086
coat-0.085
robe-0.084
amas-0.082
brush-0.082
Token her
Token ID607
Feature activation+1.31e+00

Pos logit contributions
amas+0.125
aces+0.123
coat+0.123
bridge+0.120
robe+0.119
Neg logit contributions
 Enjoy-0.069
 coconut-0.067
 shrimp-0.063
 whose-0.061
grown-0.059
Token cozy
Token ID37438
Feature activation+3.40e+00

Pos logit contributions
aces+0.098
amas+0.097
bridge+0.095
coat+0.095
robe+0.092
Neg logit contributions
 coconut-0.091
 Enjoy-0.082
 shrimp-0.081
 whose-0.080
 temple-0.079
Token bath
Token ID7837
Feature activation+2.52e+00

Pos logit contributions
aces+0.218
bridge+0.210
amas+0.198
coat+0.197
brush+0.191
Neg logit contributions
 coconut-0.281
 Enjoy-0.253
 shrimp-0.249
 volcano-0.247
 temple-0.245
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
bridge+0.130
aces+0.116
 click+0.115
pared+0.113
brush+0.112
Neg logit contributions
 shrimp-0.199
arium-0.191
 restaurant-0.191
:-0.190
 swamp-0.188
Token after
Token ID706
Feature activation-1.83e+00

Pos logit contributions
amas+1.175
bridge+1.164
aces+1.161
drops+1.136
coat+1.108
Neg logit contributions
 Enjoy-0.350
:-0.344
arium-0.328
 shrimp-0.307
 coconut-0.306
Token taking
Token ID2263
Feature activation-6.55e-01

Pos logit contributions
 coconut+0.113
 Enjoy+0.103
ours+0.090
grown+0.089
 temple+0.089
Neg logit contributions
aces-0.160
coat-0.157
robe-0.156
amas-0.154
brush-0.151
Token a
Token ID257
Feature activation+5.71e-01

Pos logit contributions
 Enjoy+0.032
 coconut+0.029
grown+0.028
 shrimp+0.028
 oy+0.026
Neg logit contributions
aces-0.061
bridge-0.061
amas-0.061
coat-0.061
robe-0.060
Token bath
Token ID7837
Feature activation+2.49e+00

Pos logit contributions
aces+0.050
bridge+0.045
amas+0.045
robe+0.044
coat+0.043
Neg logit contributions
 coconut-0.038
 Enjoy-0.035
 shrimp-0.032
 volcano-0.031
 temple-0.031
Token.
Token ID13
Feature activation-2.68e-02

Pos logit contributions
aces+0.249
bridge+0.246
brush+0.224
amas+0.222
 click+0.217
Neg logit contributions
 shrimp-0.084
 Enjoy-0.084
 restaurant-0.079
:-0.078
 swamp-0.077
Token Tim
Token ID5045
Feature activation-2.05e+00

Pos logit contributions
 coconut+0.024
 Enjoy+0.021
ours+0.019
 temple+0.019
 volcano+0.018
Neg logit contributions
aces-0.036
coat-0.035
robe-0.035
amas-0.034
brush-0.034
Token saw
Token ID2497
Feature activation+7.49e-02

Pos logit contributions
 coconut+0.130
 Enjoy+0.120
 asks+0.106
grown+0.106
ours+0.105
Neg logit contributions
aces-0.175
robe-0.170
amas-0.168
coat-0.168
bridge-0.165
Token that
Token ID326
Feature activation+8.78e-01

Pos logit contributions
aces+0.006
robe+0.006
coat+0.006
bridge+0.006
brush+0.006
Neg logit contributions
 coconut-0.005
 Enjoy-0.004
oral-0.004
iw-0.004
ours-0.004
Token his
Token ID465
Feature activation-5.67e-01

Pos logit contributions
robe+0.076
coat+0.076
aces+0.075
brush+0.075
bridge+0.073
Neg logit contributions
 coconut-0.052
 Enjoy-0.050
ours-0.044
oral-0.043
iw-0.042
Token bath
Token ID7837
Feature activation+2.51e+00

Pos logit contributions
 coconut+0.038
 Enjoy+0.037
ours+0.033
grown+0.033
oral+0.032
Neg logit contributions
aces-0.045
robe-0.045
coat-0.045
brush-0.043
amas-0.043
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
bridge+0.140
aces+0.140
brush+0.126
amas+0.120
coat+0.118
Neg logit contributions
 shrimp-0.191
 coconut-0.186
 arose-0.182
 crab-0.181
 swamp-0.180
Token was
Token ID373
Feature activation-3.68e-01

Pos logit contributions
aces+1.320
amas+1.288
bridge+1.258
coat+1.255
brush+1.251
Neg logit contributions
 coconut-0.307
:-0.237
 temple-0.231
 Enjoy-0.222
 volcano-0.219
Token too
Token ID1165
Feature activation-1.16e+00

Pos logit contributions
 coconut+0.020
 stealing+0.017
grown+0.017
 shrimp+0.016
 oy+0.016
Neg logit contributions
aces-0.035
amas-0.032
coat-0.031
robe-0.031
drops-0.031
Token tight
Token ID5381
Feature activation-2.07e+00

Pos logit contributions
 coconut+0.076
 Enjoy+0.068
grown+0.062
ours+0.061
 shrimp+0.061
Neg logit contributions
aces-0.100
bridge-0.096
coat-0.095
robe-0.095
amas-0.094
Token.
Token ID13
Feature activation-2.68e-02

Pos logit contributions
:+0.056
 Enjoy+0.048
 coconut+0.047
 stealing+0.045
grown+0.044
Neg logit contributions
bridge-0.238
aces-0.226
brush-0.223
robe-0.216
amas-0.216
Token He
Token ID679
Feature activation+4.06e-01

Pos logit contributions
 coconut+0.002
 Enjoy+0.002
oral+0.002
 temple+0.002
grown+0.002
Neg logit contributions
amas-0.002
aces-0.002
coat-0.002
robe-0.002
brush-0.002
Token they
Token ID484
Feature activation-2.67e-01

Pos logit contributions
aces+0.016
coat+0.016
robe+0.016
amas+0.016
bridge+0.015
Neg logit contributions
 coconut-0.009
 Enjoy-0.008
 temple-0.007
 volcano-0.007
grown-0.007
Token always
Token ID1464
Feature activation+1.26e+00

Pos logit contributions
 coconut+0.017
 Enjoy+0.015
 temple+0.013
ours+0.013
grown+0.013
Neg logit contributions
aces-0.023
coat-0.023
robe-0.023
amas-0.022
bridge-0.022
Token threw
Token ID9617
Feature activation-2.80e-01

Pos logit contributions
aces+0.111
coat+0.108
robe+0.108
amas+0.107
bridge+0.105
Neg logit contributions
 coconut-0.077
 Enjoy-0.071
ours-0.062
 temple-0.061
grown-0.061
Token the
Token ID262
Feature activation+1.33e-01

Pos logit contributions
 coconut+0.016
 Enjoy+0.014
ours+0.012
oral+0.012
grown+0.012
Neg logit contributions
aces-0.026
coat-0.025
robe-0.025
amas-0.025
brush-0.024
Token bath
Token ID7837
Feature activation+2.49e+00

Pos logit contributions
aces+0.010
amas+0.010
coat+0.010
robe+0.010
bridge+0.010
Neg logit contributions
 coconut-0.010
 Enjoy-0.009
 volcano-0.008
 temple-0.008
grown-0.008
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
amas+0.185
bridge+0.181
aces+0.178
coat+0.164
brush+0.163
Neg logit contributions
 coconut-0.156
 shrimp-0.152
 crab-0.146
 volcano-0.139
 Enjoy-0.136
Token in
Token ID287
Feature activation-1.33e+00

Pos logit contributions
amas+1.087
aces+1.074
bridge+1.045
robe+1.008
coat+1.005
Neg logit contributions
 coconut-0.554
 Enjoy-0.513
 shrimp-0.469
:-0.463
grown-0.457
Token the
Token ID262
Feature activation+1.22e-01

Pos logit contributions
 coconut+0.067
 Enjoy+0.060
ours+0.051
oral+0.050
 temple+0.050
Neg logit contributions
aces-0.131
coat-0.129
robe-0.128
amas-0.126
brush-0.125
Token air
Token ID1633
Feature activation+3.20e+00

Pos logit contributions
aces+0.010
coat+0.009
robe+0.009
amas+0.009
brush+0.009
Neg logit contributions
 coconut-0.009
 Enjoy-0.008
ours-0.007
grown-0.007
 temple-0.007
Token when
Token ID618
Feature activation+2.04e+00

Pos logit contributions
aces+0.321
amas+0.315
robe+0.309
bridge+0.308
coat+0.303
Neg logit contributions
 coconut-0.149
 Enjoy-0.121
 shrimp-0.113
 temple-0.113
 volcano-0.113
Token they
Token ID484
Feature activation-3.25e-01

Pos logit contributions
robe+0.181
aces+0.179
coat+0.178
amas+0.177
brush+0.175
Neg logit contributions
 coconut-0.117
 Enjoy-0.106
iw-0.095
oral-0.093
ours-0.092
Token -
Token ID532
Feature activation-1.67e+00

Pos logit contributions
aces+0.080
amas+0.080
coat+0.079
bridge+0.076
robe+0.076
Neg logit contributions
 coconut-0.039
 Enjoy-0.036
 temple-0.032
 whose-0.031
 volcano-0.030
Token a
Token ID257
Feature activation+5.83e-01

Pos logit contributions
 coconut+0.101
 Enjoy+0.092
ours+0.080
 temple+0.079
grown+0.079
Neg logit contributions
aces-0.149
coat-0.145
robe-0.144
amas-0.143
bridge-0.140
Token brand
Token ID4508
Feature activation+4.91e+00

Pos logit contributions
coat+0.043
robe+0.043
aces+0.043
brush+0.042
amas+0.042
Neg logit contributions
 coconut-0.042
 Enjoy-0.039
ours-0.036
grown-0.035
oral-0.035
Token new
Token ID649
Feature activation-1.42e+00

Pos logit contributions
robe+0.370
brush+0.365
coat+0.357
amas+0.347
aces+0.343
Neg logit contributions
 coconut-0.345
 Enjoy-0.311
grown-0.302
ours-0.290
 colony-0.288
Token bath
Token ID7837
Feature activation+2.49e+00

Pos logit contributions
 coconut+0.100
 Enjoy+0.094
ours+0.088
oral+0.084
grown+0.083
Neg logit contributions
coat-0.109
robe-0.109
aces-0.106
brush-0.105
amas-0.102
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
aces+0.183
amas+0.173
bridge+0.171
brush+0.167
coat+0.166
Neg logit contributions
 coconut-0.163
 Enjoy-0.157
 volcano-0.151
 shrimp-0.150
 temple-0.150
Token.
Token ID13
Feature activation-2.68e-02

Pos logit contributions
aces+1.150
amas+1.145
coat+1.088
bridge+1.084
brush+1.076
Neg logit contributions
 Enjoy-0.472
 coconut-0.468
grown-0.411
 temple-0.401
:-0.390
Token It
Token ID632
Feature activation+5.38e-01

Pos logit contributions
 coconut+0.002
 Enjoy+0.001
ours+0.001
 temple+0.001
 volcano+0.001
Neg logit contributions
aces-0.002
robe-0.002
coat-0.002
amas-0.002
brush-0.002
Token was
Token ID373
Feature activation-3.87e-01

Pos logit contributions
aces+0.061
amas+0.061
coat+0.059
bridge+0.058
robe+0.057
Neg logit contributions
 coconut-0.017
 Enjoy-0.016
 asks-0.014
 whose-0.013
 temple-0.013
Token soft
Token ID2705
Feature activation+3.59e+00

Pos logit contributions
 coconut+0.023
 Enjoy+0.022
grown+0.020
 whose+0.019
 oy+0.019
Neg logit contributions
aces-0.034
amas-0.034
coat-0.033
robe-0.032
bridge-0.032
Token,
Token ID11
Feature activation+8.25e-02

Pos logit contributions
aces+0.393
bridge+0.378
amas+0.372
robe+0.368
coat+0.362
Neg logit contributions
 coconut-0.126
 Enjoy-0.116
 shrimp-0.109
grown-0.106
 oy-0.100
Token;
Token ID26
Feature activation-1.27e+00

Pos logit contributions
aces+0.004
coat+0.004
amas+0.004
bridge+0.004
brush+0.004
Neg logit contributions
 coconut-0.002
 Enjoy-0.002
 volcano-0.002
 temple-0.002
 shrimp-0.002
Token he
Token ID339
Feature activation+5.82e-01

Pos logit contributions
 coconut+0.076
 Enjoy+0.068
oral+0.063
ours+0.061
grown+0.061
Neg logit contributions
aces-0.111
robe-0.109
coat-0.108
brush-0.107
bridge-0.105
Token has
Token ID468
Feature activation-8.83e-01

Pos logit contributions
aces+0.053
robe+0.052
coat+0.051
amas+0.050
brush+0.050
Neg logit contributions
 coconut-0.034
 Enjoy-0.031
ours-0.027
oral-0.027
grown-0.026
Token a
Token ID257
Feature activation+5.71e-01

Pos logit contributions
 coconut+0.052
 Enjoy+0.047
ours+0.041
grown+0.040
 temple+0.040
Neg logit contributions
aces-0.080
coat-0.078
robe-0.078
amas-0.077
brush-0.076
Token bath
Token ID7837
Feature activation+2.49e+00

Pos logit contributions
coat+0.041
robe+0.041
brush+0.040
aces+0.039
amas+0.038
Neg logit contributions
 coconut-0.042
 Enjoy-0.040
ours-0.038
oral-0.037
grown-0.036
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
aces+0.175
amas+0.171
bridge+0.165
brush+0.157
coat+0.155
Neg logit contributions
 coconut-0.170
 shrimp-0.162
 Enjoy-0.159
 crab-0.156
 oy-0.152
Token too
Token ID1165
Feature activation-1.17e+00

Pos logit contributions
aces+1.141
amas+1.141
coat+1.071
bridge+1.055
robe+1.050
Neg logit contributions
 coconut-0.499
 Enjoy-0.493
grown-0.419
 shrimp-0.411
 temple-0.406
Token."
Token ID526
Feature activation+1.55e+00

Pos logit contributions
 coconut+0.076
 Enjoy+0.069
oral+0.062
ours+0.061
iw+0.061
Neg logit contributions
aces-0.100
coat-0.098
robe-0.097
amas-0.095
brush-0.093
Token
Token ID198
Feature activation+1.28e-01

Pos logit contributions
aces+0.113
amas+0.112
robe+0.111
stand+0.110
coat+0.110
Neg logit contributions
 coconut-0.114
ours-0.094
 temple-0.093
grown-0.092
 Enjoy-0.092
Token
Token ID198
Feature activation+1.23e-01

Pos logit contributions
coat+0.009
stand+0.009
robe+0.009
aces+0.009
amas+0.009
Neg logit contributions
 coconut-0.009
 Enjoy-0.008
ours-0.008
 volcano-0.007
 oy-0.007
TokenHer
Token ID9360
Feature activation-3.97e-02

Pos logit contributions
coat+0.008
brush+0.008
bridge+0.008
aces+0.008
robe+0.008
Neg logit contributions
 coconut-0.010
ours-0.009
 Enjoy-0.009
 volcano-0.009
 oy-0.008
Token are
Token ID389
Feature activation+1.49e+00

Pos logit contributions
aces+0.004
coat+0.004
robe+0.004
amas+0.004
brush+0.004
Neg logit contributions
 coconut-0.003
 Enjoy-0.003
ours-0.002
grown-0.002
 temple-0.002
Token you
Token ID345
Feature activation-3.73e-01

Pos logit contributions
aces+0.107
robe+0.099
coat+0.099
amas+0.098
brush+0.097
Neg logit contributions
 coconut-0.113
 Enjoy-0.103
oral-0.098
 exhibit-0.095
grown-0.095
Token wearing
Token ID5762
Feature activation+9.60e-01

Pos logit contributions
 coconut+0.023
 Enjoy+0.021
ours+0.019
grown+0.019
 temple+0.018
Neg logit contributions
aces-0.033
coat-0.032
robe-0.031
bridge-0.031
amas-0.031
Token the
Token ID262
Feature activation+1.07e-01

Pos logit contributions
aces+0.107
bridge+0.104
coat+0.102
amas+0.102
drops+0.101
Neg logit contributions
 Enjoy-0.031
 coconut-0.030
 shrimp-0.030
 whose-0.028
 what-0.027
Token bath
Token ID7837
Feature activation+2.47e+00

Pos logit contributions
aces+0.011
bridge+0.011
drops+0.010
brush+0.010
amas+0.010
Neg logit contributions
 shrimp-0.005
 oy-0.004
 coconut-0.004
 crab-0.004
 fish-0.004
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
bridge+0.112
aces+0.111
brush+0.099
coat+0.095
drops+0.089
Neg logit contributions
 Enjoy-0.225
 coconut-0.223
 shrimp-0.222
ours-0.219
arium-0.219
Token?"
Token ID1701
Feature activation-2.20e+00

Pos logit contributions
aces+1.344
bridge+1.301
coat+1.266
amas+1.255
robe+1.235
Neg logit contributions
 Enjoy-0.307
 coconut-0.304
 temple-0.237
arium-0.233
 museum-0.225
Token
Token ID198
Feature activation+1.27e-01

Pos logit contributions
 coconut+0.150
 Enjoy+0.124
 volcano+0.121
iw+0.119
grown+0.117
Neg logit contributions
aces-0.183
amas-0.178
robe-0.176
coat-0.172
brush-0.169
Token
Token ID198
Feature activation+1.31e-01

Pos logit contributions
aces+0.011
amas+0.011
robe+0.011
coat+0.011
brush+0.011
Neg logit contributions
 coconut-0.008
 Enjoy-0.006
iw-0.006
 volcano-0.006
ours-0.006
TokenCl
Token ID2601
Feature activation-1.37e+00

Pos logit contributions
aces+0.011
coat+0.011
amas+0.011
robe+0.011
brush+0.010
Neg logit contributions
 coconut-0.009
 Enjoy-0.008
ours-0.007
 volcano-0.007
iw-0.007
Tokenare
Token ID533
Feature activation-8.26e-01

Pos logit contributions
 coconut+0.054
 Enjoy+0.044
 volcano+0.036
ours+0.035
 temple+0.035
Neg logit contributions
aces-0.152
coat-0.149
robe-0.148
amas-0.148
brush-0.145
Token happy
Token ID3772
Feature activation+1.17e+00

Pos logit contributions
 coconut+0.154
 Enjoy+0.142
ours+0.127
grown+0.126
 temple+0.126
Neg logit contributions
aces-0.160
coat-0.156
robe-0.156
amas-0.153
brush-0.150
Token and
Token ID290
Feature activation-6.88e-02

Pos logit contributions
aces+0.100
coat+0.098
robe+0.097
amas+0.096
brush+0.094
Neg logit contributions
 coconut-0.075
 Enjoy-0.068
ours-0.061
oral-0.060
grown-0.059
Token wore
Token ID12408
Feature activation+1.31e+00

Pos logit contributions
 coconut+0.003
 Enjoy+0.003
 oy+0.003
 shrimp+0.003
 Crab+0.003
Neg logit contributions
aces-0.007
coat-0.007
amas-0.007
bridge-0.007
brush-0.007
Token the
Token ID262
Feature activation+1.05e-01

Pos logit contributions
coat+0.168
amas+0.166
drops+0.164
aces+0.164
robe+0.161
Neg logit contributions
 what-0.021
 Enjoy-0.019
 coconut-0.019
grown-0.011
 shrimp-0.011
Token bath
Token ID7837
Feature activation+2.48e+00

Pos logit contributions
aces+0.011
bridge+0.011
drops+0.010
brush+0.010
ring+0.010
Neg logit contributions
 shrimp-0.004
 fish-0.004
 oy-0.003
 crab-0.003
 swamp-0.003
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
bridge+0.096
amas+0.085
aces+0.084
brush+0.080
pared+0.079
Neg logit contributions
 swamp-0.221
:-0.219
arium-0.219
 volcano-0.218
 shrimp-0.217
Token all
Token ID477
Feature activation-2.13e+00

Pos logit contributions
amas+1.196
aces+1.170
bridge+1.160
coat+1.140
drops+1.096
Neg logit contributions
 coconut-0.437
 Enjoy-0.406
:-0.368
 exhibit-0.363
 temple-0.353
Token day
Token ID1110
Feature activation+1.32e-01

Pos logit contributions
 coconut+0.115
 Enjoy+0.104
ours+0.089
grown+0.088
 temple+0.088
Neg logit contributions
aces-0.204
coat-0.201
robe-0.199
amas-0.197
brush-0.194
Token.
Token ID13
Feature activation-4.54e-02

Pos logit contributions
bridge+0.012
robe+0.012
aces+0.012
coat+0.012
amas+0.012
Neg logit contributions
 coconut-0.007
 Enjoy-0.006
 shrimp-0.006
 temple-0.006
 exhibit-0.006
Token
Token ID198
Feature activation+1.45e-01

Pos logit contributions
 coconut+0.003
 temple+0.003
 oy+0.003
grown+0.003
 crab+0.003
Neg logit contributions
amas-0.004
aces-0.004
robe-0.004
coat-0.003
bridge-0.003
Token
Token ID198
Feature activation+1.41e-01

Pos logit contributions
amas+0.011
robe+0.011
aces+0.011
coat+0.011
stand+0.010
Neg logit contributions
 coconut-0.010
ours-0.009
 Enjoy-0.009
 oy-0.008
 shrimp-0.008
Token And
Token ID843
Feature activation-1.21e+00

Pos logit contributions
 coconut+0.002
 Enjoy+0.002
 temple+0.002
ours+0.002
 volcano+0.002
Neg logit contributions
robe-0.004
aces-0.004
coat-0.004
amas-0.004
brush-0.004
Token they
Token ID484
Feature activation-3.08e-01

Pos logit contributions
 coconut+0.060
 Enjoy+0.053
ours+0.045
grown+0.044
 temple+0.044
Neg logit contributions
aces-0.121
coat-0.119
robe-0.119
amas-0.117
brush-0.115
Token loved
Token ID6151
Feature activation+4.57e-01

Pos logit contributions
 coconut+0.018
 Enjoy+0.016
ours+0.014
grown+0.014
 temple+0.014
Neg logit contributions
aces-0.028
coat-0.027
robe-0.027
amas-0.027
brush-0.026
Token their
Token ID511
Feature activation-1.24e+00

Pos logit contributions
robe+0.035
aces+0.034
coat+0.034
amas+0.032
stand+0.032
Neg logit contributions
 coconut-0.033
oral-0.027
 Enjoy-0.026
 exhibit-0.025
iw-0.025
Token bath
Token ID7837
Feature activation+2.48e+00

Pos logit contributions
 coconut+0.076
 Enjoy+0.070
 shrimp+0.064
 temple+0.064
 volcano+0.063
Neg logit contributions
aces-0.111
bridge-0.105
coat-0.103
amas-0.103
brush-0.102
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
bridge+0.201
aces+0.196
brush+0.191
coat+0.180
amas+0.177
Neg logit contributions
 shrimp-0.130
 swamp-0.129
 Enjoy-0.129
 volcano-0.128
 restaurant-0.128
Token,
Token ID11
Feature activation+1.04e-01

Pos logit contributions
amas+1.249
aces+1.233
bridge+1.214
coat+1.191
brush+1.156
Neg logit contributions
 Enjoy-0.362
 coconut-0.356
 volcano-0.318
 restaurant-0.304
 temple-0.296
Token too
Token ID1165
Feature activation-1.17e+00

Pos logit contributions
aces+0.008
coat+0.008
robe+0.008
amas+0.008
bridge+0.008
Neg logit contributions
 coconut-0.007
 Enjoy-0.007
grown-0.006
ours-0.006
 temple-0.006
Token.
Token ID13
Feature activation-4.54e-02

Pos logit contributions
 coconut+0.077
 Enjoy+0.070
ours+0.061
iw+0.061
oral+0.060
Neg logit contributions
aces-0.098
coat-0.096
robe-0.095
amas-0.095
brush-0.092
Token<|endoftext|>
Token ID50256
Feature activation-4.34e-01

Pos logit contributions
 coconut+0.002
 Enjoy+0.002
 temple+0.002
ours+0.002
grown+0.002
Neg logit contributions
aces-0.004
robe-0.004
amas-0.004
coat-0.004
brush-0.004
TokenAnna
Token ID31160
Feature activation-1.80e-01

Pos logit contributions
 coconut+0.021
 Enjoy+0.016
ours+0.014
 temple+0.014
 mistake+0.014
Neg logit contributions
robe-0.046
aces-0.045
amas-0.044
coat-0.044
bridge-0.043
Token.
Token ID13
Feature activation-4.94e-02

Pos logit contributions
aces+0.151
coat+0.148
robe+0.147
amas+0.145
bridge+0.143
Neg logit contributions
 coconut-0.093
 Enjoy-0.085
ours-0.074
grown-0.073
 temple-0.072
Token They
Token ID1119
Feature activation-1.23e+00

Pos logit contributions
 coconut+0.003
 Enjoy+0.003
 oy+0.003
 mole+0.003
elia+0.003
Neg logit contributions
coat-0.004
aces-0.004
amas-0.004
bridge-0.004
robe-0.004
Token used
Token ID973
Feature activation-1.28e+00

Pos logit contributions
 coconut+0.082
 Enjoy+0.075
 volcano+0.067
oral+0.067
ours+0.067
Neg logit contributions
coat-0.103
aces-0.102
robe-0.099
brush-0.098
bridge-0.094
Token the
Token ID262
Feature activation+1.06e-01

Pos logit contributions
 coconut+0.059
 Enjoy+0.052
grown+0.044
ours+0.044
 shrimp+0.044
Neg logit contributions
aces-0.131
amas-0.129
coat-0.129
robe-0.128
bridge-0.126
Token bath
Token ID7837
Feature activation+2.47e+00

Pos logit contributions
aces+0.010
bridge+0.010
amas+0.009
coat+0.009
brush+0.009
Neg logit contributions
 shrimp-0.006
 coconut-0.006
 question-0.005
 oy-0.005
 swamp-0.005
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
bridge+0.129
amas+0.118
aces+0.111
brush+0.104
iver+0.102
Neg logit contributions
 shrimp-0.211
ours-0.209
 coconut-0.209
 Enjoy-0.204
:-0.204
Token to
Token ID284
Feature activation-5.48e-01

Pos logit contributions
amas+1.403
aces+1.356
bridge+1.349
coat+1.320
robe+1.280
Neg logit contributions
 coconut-0.304
 Enjoy-0.251
 shrimp-0.209
:-0.208
 exhibit-0.196
Token pretend
Token ID16614
Feature activation+3.60e+00

Pos logit contributions
 coconut+0.031
 Enjoy+0.030
 whose+0.026
 investigate+0.026
 shrimp+0.026
Neg logit contributions
aces-0.051
amas-0.049
bridge-0.048
robe-0.046
coat-0.045
Token they
Token ID484
Feature activation-3.26e-01

Pos logit contributions
aces+0.378
coat+0.370
amas+0.367
robe+0.364
bridge+0.358
Neg logit contributions
 coconut-0.158
 Enjoy-0.136
 volcano-0.120
 temple-0.119
grown-0.115
Token were
Token ID547
Feature activation+1.11e+00

Pos logit contributions
 coconut+0.014
 Enjoy+0.013
grown+0.012
 whose+0.011
 temple+0.011
Neg logit contributions
amas-0.034
aces-0.033
bridge-0.032
coat-0.032
brush-0.032
Token superheroes
Token ID40896
Feature activation+2.22e+00

Pos logit contributions
amas+0.104
aces+0.104
brush+0.103
coat+0.102
bridge+0.099
Neg logit contributions
 coconut-0.053
 Enjoy-0.053
 shrimp-0.051
 whose-0.049
grown-0.048
Token to
Token ID284
Feature activation-5.53e-01

Pos logit contributions
 coconut+0.039
 Enjoy+0.033
grown+0.027
 whose+0.024
 exhibit+0.023
Neg logit contributions
amas-0.143
bridge-0.142
coat-0.140
robe-0.138
aces-0.138
Token mark
Token ID1317
Feature activation-1.72e+00

Pos logit contributions
 coconut+0.036
 Enjoy+0.033
 oy+0.029
 temple+0.028
 whose+0.028
Neg logit contributions
aces-0.047
amas-0.046
bridge-0.045
robe-0.045
coat-0.044
Token where
Token ID810
Feature activation+1.26e+00

Pos logit contributions
 coconut+0.091
 Enjoy+0.082
ours+0.070
grown+0.069
 temple+0.069
Neg logit contributions
aces-0.165
coat-0.162
robe-0.161
amas-0.159
brush-0.156
Token the
Token ID262
Feature activation+1.06e-01

Pos logit contributions
aces+0.116
robe+0.113
brush+0.111
coat+0.110
bridge+0.110
Neg logit contributions
 coconut-0.070
 Enjoy-0.065
oral-0.056
ours-0.056
iw-0.056
Token bath
Token ID7837
Feature activation+2.47e+00

Pos logit contributions
coat+0.008
aces+0.008
amas+0.008
bridge+0.008
robe+0.008
Neg logit contributions
 coconut-0.008
 Enjoy-0.006
 oy-0.006
 shrimp-0.006
 temple-0.006
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
bridge+0.073
aces+0.068
 forehead+0.066
coat+0.066
amas+0.063
Neg logit contributions
 shrimp-0.236
 arose-0.235
 swamp-0.234
 coconut-0.234
 volcano-0.233
Token should
Token ID815
Feature activation-4.52e+00

Pos logit contributions
amas+1.237
coat+1.216
aces+1.198
bridge+1.168
robe+1.165
Neg logit contributions
 coconut-0.405
 temple-0.328
 Enjoy-0.313
 volcano-0.311
:-0.302
Token be
Token ID307
Feature activation+8.01e-01

Pos logit contributions
 coconut+0.293
 Enjoy+0.265
 temple+0.235
 volcano+0.229
 crab+0.224
Neg logit contributions
amas-0.380
aces-0.378
coat-0.373
robe-0.366
bridge-0.356
Token cut
Token ID2005
Feature activation-1.58e-02

Pos logit contributions
amas+0.071
robe+0.069
coat+0.068
aces+0.067
bridge+0.067
Neg logit contributions
 coconut-0.049
 Enjoy-0.042
grown-0.040
 shrimp-0.039
 whose-0.039
Token.
Token ID13
Feature activation-4.15e-02

Pos logit contributions
 coconut+0.001
 Enjoy+0.001
grown+0.001
 oy+0.001
 shrimp+0.001
Neg logit contributions
coat-0.002
amas-0.001
bridge-0.001
robe-0.001
aces-0.001
Token He
Token ID679
Feature activation+3.98e-01

Pos logit contributions
 coconut+0.003
iw+0.003
grown+0.003
 temple+0.003
 crab+0.003
Neg logit contributions
amas-0.003
aces-0.003
robe-0.003
bridge-0.003
brush-0.003
Token too
Token ID1165
Feature activation-1.16e+00

Pos logit contributions
 coconut+0.022
 Enjoy+0.019
grown+0.018
 volcano+0.017
 stealing+0.017
Neg logit contributions
aces-0.031
robe-0.031
amas-0.031
coat-0.030
bridge-0.029
Token full
Token ID1336
Feature activation+9.61e-01

Pos logit contributions
 coconut+0.087
 Enjoy+0.082
ours+0.074
 temple+0.072
oral+0.072
Neg logit contributions
aces-0.085
coat-0.084
robe-0.083
amas-0.082
brush-0.081
Token.
Token ID13
Feature activation-2.52e-02

Pos logit contributions
aces+0.100
amas+0.100
bridge+0.099
robe+0.099
coat+0.099
Neg logit contributions
 coconut-0.038
 Enjoy-0.032
:-0.030
grown-0.029
 oy-0.028
Token The
Token ID383
Feature activation-5.58e-01

Pos logit contributions
 coconut+0.002
 temple+0.002
 Enjoy+0.002
 whale+0.001
ours+0.001
Neg logit contributions
aces-0.002
amas-0.002
robe-0.002
coat-0.002
brush-0.002
Token bath
Token ID7837
Feature activation+2.48e+00

Pos logit contributions
 coconut+0.039
 Enjoy+0.036
ours+0.033
grown+0.032
iw+0.032
Neg logit contributions
aces-0.044
coat-0.043
robe-0.043
amas-0.042
brush-0.042
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
aces+0.198
bridge+0.183
amas+0.181
coat+0.180
brush+0.175
Neg logit contributions
 coconut-0.153
 volcano-0.138
 crab-0.137
 shrimp-0.137
 arose-0.133
Token was
Token ID373
Feature activation-3.88e-01

Pos logit contributions
aces+1.206
amas+1.190
coat+1.160
robe+1.139
bridge+1.136
Neg logit contributions
 coconut-0.473
 crab-0.355
 temple-0.352
 volcano-0.344
 Enjoy-0.343
Token too
Token ID1165
Feature activation-1.19e+00

Pos logit contributions
 coconut+0.023
grown+0.020
 Enjoy+0.019
 stealing+0.019
 shrimp+0.018
Neg logit contributions
aces-0.034
amas-0.034
robe-0.033
coat-0.032
bridge-0.032
Token big
Token ID1263
Feature activation+1.29e+00

Pos logit contributions
 coconut+0.086
 Enjoy+0.079
ours+0.071
grown+0.070
 temple+0.070
Neg logit contributions
aces-0.093
coat-0.090
robe-0.089
amas-0.088
bridge-0.086
Token.
Token ID13
Feature activation-5.93e-02

Pos logit contributions
robe+0.138
amas+0.138
bridge+0.138
coat+0.136
aces+0.135
Neg logit contributions
 coconut-0.042
 Enjoy-0.041
grown-0.039
 stealing-0.038
:-0.037
Token
Token ID198
Feature activation+1.11e-01

Pos logit contributions
 coconut+0.004
 Enjoy+0.003
 temple+0.003
ours+0.003
grown+0.003
Neg logit contributions
aces-0.005
amas-0.005
robe-0.005
coat-0.005
brush-0.005
Token had
Token ID550
Feature activation-9.72e-02

Pos logit contributions
coat+0.031
aces+0.031
robe+0.031
brush+0.030
amas+0.029
Neg logit contributions
 coconut-0.025
 Enjoy-0.022
ours-0.020
iw-0.019
oral-0.019
Token a
Token ID257
Feature activation+5.71e-01

Pos logit contributions
 coconut+0.005
 Enjoy+0.004
grown+0.004
 temple+0.004
 whose+0.004
Neg logit contributions
amas-0.009
robe-0.009
aces-0.009
coat-0.009
bridge-0.009
Token lovely
Token ID14081
Feature activation+6.93e-01

Pos logit contributions
coat+0.040
robe+0.038
brush+0.037
amas+0.035
aces+0.035
Neg logit contributions
 coconut-0.045
 Enjoy-0.043
ours-0.041
oral-0.039
grown-0.039
Token blue
Token ID4171
Feature activation+1.46e+00

Pos logit contributions
aces+0.052
coat+0.052
robe+0.051
amas+0.050
brush+0.049
Neg logit contributions
 coconut-0.051
 Enjoy-0.047
ours-0.043
grown-0.042
oral-0.042
Token bath
Token ID7837
Feature activation+2.48e+00

Pos logit contributions
coat+0.088
robe+0.088
aces+0.086
brush+0.084
amas+0.082
Neg logit contributions
 coconut-0.128
 Enjoy-0.122
ours-0.113
grown-0.112
oral-0.111
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
aces+0.111
bridge+0.110
amas+0.110
pared+0.101
rell+0.095
Neg logit contributions
 shrimp-0.215
:-0.212
 coconut-0.210
 Enjoy-0.207
 whose-0.207
Token.
Token ID13
Feature activation-4.70e-02

Pos logit contributions
amas+1.242
aces+1.218
bridge+1.180
coat+1.129
drops+1.127
Neg logit contributions
 Enjoy-0.389
 coconut-0.342
:-0.339
grown-0.316
 whose-0.316
Token Lucy
Token ID22162
Feature activation-1.37e+00

Pos logit contributions
 coconut+0.003
 Enjoy+0.003
oral+0.003
 temple+0.003
 oy+0.003
Neg logit contributions
aces-0.004
coat-0.004
amas-0.004
robe-0.004
brush-0.004
Token loved
Token ID6151
Feature activation+4.49e-01

Pos logit contributions
 coconut+0.068
 Enjoy+0.062
ours+0.053
grown+0.053
 whose+0.052
Neg logit contributions
aces-0.134
robe-0.130
coat-0.130
amas-0.130
bridge-0.128
Token to
Token ID284
Feature activation-5.63e-01

Pos logit contributions
aces+0.042
coat+0.041
robe+0.041
amas+0.040
brush+0.039
Neg logit contributions
 coconut-0.026
 Enjoy-0.023
ours-0.020
oral-0.020
 temple-0.019
Token play
Token ID711
Feature activation+1.05e-01

Pos logit contributions
 Enjoy+0.030
 coconut+0.029
 shrimp+0.025
ours+0.024
 whose+0.024
Neg logit contributions
aces-0.053
amas-0.051
bridge-0.050
coat-0.049
robe-0.049
Token about
Token ID546
Feature activation-1.06e+00

Pos logit contributions
 coconut+0.003
 Enjoy+0.002
iw+0.002
 Crab+0.002
grown+0.002
Neg logit contributions
aces-0.003
brush-0.003
coat-0.003
robe-0.003
bridge-0.003
Token we
Token ID356
Feature activation-4.00e-01

Pos logit contributions
 coconut+0.079
 Enjoy+0.072
oral+0.069
iw+0.067
grown+0.064
Neg logit contributions
amas-0.075
aces-0.074
brush-0.073
robe-0.073
bridge-0.071
Token use
Token ID779
Feature activation-3.49e+00

Pos logit contributions
 coconut+0.027
 Enjoy+0.024
ours+0.022
grown+0.022
iw+0.022
Neg logit contributions
aces-0.033
coat-0.033
robe-0.032
amas-0.032
brush-0.032
Token the
Token ID262
Feature activation+9.70e-02

Pos logit contributions
 coconut+0.195
 Enjoy+0.175
oral+0.151
ours+0.148
grown+0.147
Neg logit contributions
aces-0.327
coat-0.319
robe-0.317
amas-0.313
brush-0.311
Token bath
Token ID7837
Feature activation+2.48e+00

Pos logit contributions
aces+0.008
bridge+0.007
amed+0.007
amas+0.007
coat+0.007
Neg logit contributions
 coconut-0.006
 oy-0.006
 shrimp-0.006
 museum-0.006
 temple-0.006
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
aces+0.118
bridge+0.115
coat+0.106
brush+0.102
amas+0.101
Neg logit contributions
 Enjoy-0.225
 coconut-0.224
ours-0.220
arium-0.214
 volcano-0.212
Token to
Token ID284
Feature activation-5.49e-01

Pos logit contributions
aces+1.215
coat+1.174
amas+1.169
bridge+1.162
robe+1.143
Neg logit contributions
 coconut-0.471
 Enjoy-0.450
ours-0.362
 shrimp-0.347
 temple-0.347
Token make
Token ID787
Feature activation+3.57e+00

Pos logit contributions
 coconut+0.037
 Enjoy+0.035
 temple+0.031
ours+0.031
 whose+0.031
Neg logit contributions
aces-0.046
bridge-0.043
coat-0.043
amas-0.043
robe-0.042
Token a
Token ID257
Feature activation+5.57e-01

Pos logit contributions
aces+0.324
coat+0.316
robe+0.316
amas+0.312
brush+0.306
Neg logit contributions
 coconut-0.210
 Enjoy-0.190
ours-0.164
grown-0.163
 temple-0.162
Token tent
Token ID11105
Feature activation+7.70e-01

Pos logit contributions
robe+0.039
coat+0.037
amas+0.037
brush+0.037
bridge+0.034
Neg logit contributions
 coconut-0.043
 Enjoy-0.040
grown-0.039
ours-0.038
iw-0.037
Token in
Token ID287
Feature activation-1.35e+00

Pos logit contributions
aces+0.088
coat+0.084
amas+0.083
bridge+0.081
robe+0.080
Neg logit contributions
 coconut-0.025
 Enjoy-0.025
 temple-0.020
arium-0.020
 volcano-0.020
Token he
Token ID339
Feature activation+5.45e-01

Pos logit contributions
aces+0.202
brush+0.197
robe+0.197
coat+0.194
amas+0.192
Neg logit contributions
 coconut-0.197
 Enjoy-0.192
ours-0.175
iw-0.169
oral-0.169
Token took
Token ID1718
Feature activation-1.75e-01

Pos logit contributions
aces+0.047
coat+0.046
robe+0.046
amas+0.045
brush+0.045
Neg logit contributions
 coconut-0.034
 Enjoy-0.031
ours-0.028
grown-0.027
 temple-0.027
Token off
Token ID572
Feature activation+7.23e-01

Pos logit contributions
 coconut+0.010
 Enjoy+0.009
ours+0.008
grown+0.008
 temple+0.007
Neg logit contributions
aces-0.016
coat-0.016
robe-0.016
amas-0.016
bridge-0.015
Token his
Token ID465
Feature activation-6.24e-01

Pos logit contributions
aces+0.062
coat+0.062
robe+0.062
amas+0.060
brush+0.060
Neg logit contributions
 coconut-0.046
 Enjoy-0.041
ours-0.038
oral-0.037
 temple-0.036
Token bath
Token ID7837
Feature activation+2.48e+00

Pos logit contributions
 coconut+0.035
 shrimp+0.034
 Enjoy+0.033
ours+0.031
 oy+0.030
Neg logit contributions
aces-0.056
bridge-0.056
brush-0.053
coat-0.052
amas-0.050
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
bridge+0.105
aces+0.085
brush+0.072
pared+0.068
mic+0.066
Neg logit contributions
 shrimp-0.238
:-0.233
ours-0.231
 restaurant-0.228
 Enjoy-0.228
Token and
Token ID290
Feature activation-8.32e-02

Pos logit contributions
bridge+1.179
aces+1.169
amas+1.141
coat+1.136
robe+1.107
Neg logit contributions
 coconut-0.450
 Enjoy-0.400
:-0.352
 whose-0.351
 exhibit-0.349
Token took
Token ID1718
Feature activation-1.69e-01

Pos logit contributions
 coconut+0.005
 Enjoy+0.004
 shrimp+0.004
 whose+0.004
 oy+0.004
Neg logit contributions
aces-0.008
amas-0.007
bridge-0.007
coat-0.007
robe-0.007
Token a
Token ID257
Feature activation+5.57e-01

Pos logit contributions
 coconut+0.009
 Enjoy+0.008
grown+0.007
 shrimp+0.007
ours+0.007
Neg logit contributions
coat-0.017
aces-0.016
robe-0.016
amas-0.016
bridge-0.015
Token bath
Token ID7837
Feature activation+2.47e+00

Pos logit contributions
aces+0.039
coat+0.039
robe+0.038
amas+0.038
brush+0.037
Neg logit contributions
 coconut-0.044
 Enjoy-0.041
ours-0.037
grown-0.036
 temple-0.036
Token.
Token ID13
Feature activation-4.15e-02

Pos logit contributions
bridge+0.247
aces+0.246
amas+0.228
brush+0.223
rell+0.219
Neg logit contributions
 Enjoy-0.087
 shrimp-0.085
 coconut-0.082
ours-0.077
 crab-0.076
Token to
Token ID284
Feature activation-5.47e-01

Pos logit contributions
robe+0.093
brush+0.092
aces+0.089
coat+0.088
 cartridge+0.084
Neg logit contributions
 coconut-0.088
iw-0.076
 Enjoy-0.075
 volcano-0.071
rill-0.070
Token take
Token ID1011
Feature activation-2.18e+00

Pos logit contributions
 Enjoy+0.033
 coconut+0.031
 shrimp+0.027
 investigate+0.027
 whose+0.026
Neg logit contributions
aces-0.053
bridge-0.048
robe-0.046
amas-0.046
coat-0.046
Token off
Token ID572
Feature activation+7.34e-01

Pos logit contributions
 coconut+0.132
 Enjoy+0.126
ours+0.107
grown+0.107
 temple+0.105
Neg logit contributions
aces-0.192
coat-0.188
robe-0.187
amas-0.184
bridge-0.182
Token her
Token ID607
Feature activation+1.23e+00

Pos logit contributions
aces+0.061
coat+0.061
robe+0.060
amas+0.059
brush+0.058
Neg logit contributions
 coconut-0.049
 Enjoy-0.044
ours-0.039
 volcano-0.039
oral-0.038
Token bath
Token ID7837
Feature activation+2.48e+00

Pos logit contributions
aces+0.116
bridge+0.113
brush+0.108
amas+0.105
coat+0.105
Neg logit contributions
 shrimp-0.063
 coconut-0.063
 Enjoy-0.061
ours-0.058
 oy-0.057
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
bridge+0.159
aces+0.153
brush+0.133
drops+0.127
pared+0.125
Neg logit contributions
 restaurant-0.183
:-0.182
 shrimp-0.182
 Enjoy-0.181
 swamp-0.180
Token and
Token ID290
Feature activation-7.98e-02

Pos logit contributions
bridge+1.270
aces+1.269
drops+1.175
amas+1.165
coat+1.141
Neg logit contributions
 Enjoy-0.363
 coconut-0.335
:-0.304
arium-0.293
 exhibit-0.279
Token wear
Token ID5806
Feature activation+1.42e+00

Pos logit contributions
 coconut+0.005
 Enjoy+0.004
 temple+0.004
 whose+0.004
 shrimp+0.004
Neg logit contributions
aces-0.008
bridge-0.007
amas-0.007
coat-0.007
robe-0.007
Token her
Token ID607
Feature activation+1.24e+00

Pos logit contributions
aces+0.179
amas+0.173
drops+0.173
bridge+0.171
coat+0.170
Neg logit contributions
 what-0.027
 Enjoy-0.022
 it-0.017
 their-0.015
 shrimp-0.014
Token p
Token ID279
Feature activation-3.31e+00

Pos logit contributions
bridge+0.120
aces+0.118
brush+0.111
drops+0.111
amas+0.111
Neg logit contributions
 shrimp-0.059
 whose-0.054
 Enjoy-0.052
 coconut-0.051
 what-0.050
Tokenaj
Token ID1228
Feature activation+6.05e+00

Pos logit contributions
 coconut+0.262
ours+0.252
 Enjoy+0.242
grown+0.231
iw+0.230
Neg logit contributions
aces-0.217
bridge-0.216
brush-0.199
amas-0.195
robe-0.195
Token his
Token ID465
Feature activation-6.15e-01

Pos logit contributions
 coconut+0.006
 Enjoy+0.005
oral+0.005
ours+0.004
grown+0.004
Neg logit contributions
aces-0.008
robe-0.008
coat-0.008
brush-0.008
bridge-0.007
Token bag
Token ID6131
Feature activation+3.29e+00

Pos logit contributions
 coconut+0.039
 Enjoy+0.033
 shrimp+0.032
 volcano+0.032
 crab+0.032
Neg logit contributions
aces-0.054
bridge-0.052
amas-0.051
coat-0.050
brush-0.050
Token and
Token ID290
Feature activation-8.32e-02

Pos logit contributions
amas+0.346
 cartridge+0.333
aces+0.324
coat+0.320
uster+0.318
Neg logit contributions
 coconut-0.103
grown-0.100
:-0.095
 temple-0.081
 what-0.078
Token his
Token ID465
Feature activation-6.16e-01

Pos logit contributions
 coconut+0.005
 Enjoy+0.004
 temple+0.004
ours+0.004
grown+0.004
Neg logit contributions
aces-0.008
coat-0.008
amas-0.008
robe-0.008
bridge-0.007
Token bath
Token ID7837
Feature activation+2.48e+00

Pos logit contributions
 coconut+0.043
 shrimp+0.038
 Enjoy+0.037
 volcano+0.037
 crab+0.037
Neg logit contributions
aces-0.049
bridge-0.047
amas-0.046
coat-0.045
brush-0.044
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
bridge+0.080
aces+0.076
pared+0.072
mic+0.071
brush+0.065
Neg logit contributions
 shrimp-0.234
ours-0.229
 arose-0.229
:-0.226
 coconut-0.223
Token.
Token ID13
Feature activation-4.15e-02

Pos logit contributions
amas+1.309
coat+1.240
aces+1.238
uster+1.229
robe+1.217
Neg logit contributions
 coconut-0.291
:-0.261
 temple-0.259
grown-0.248
 Enjoy-0.208
Token
Token ID198
Feature activation+1.30e-01

Pos logit contributions
 coconut+0.003
 temple+0.003
oral+0.003
 exhibit+0.003
 Enjoy+0.003
Neg logit contributions
robe-0.003
aces-0.003
coat-0.003
amas-0.003
brush-0.003
Token
Token ID198
Feature activation+1.40e-01

Pos logit contributions
robe+0.011
amas+0.011
aces+0.011
coat+0.011
brush+0.011
Neg logit contributions
 coconut-0.008
 Enjoy-0.007
iw-0.006
ours-0.006
 shrimp-0.006
Token"
Token ID1
Feature activation+1.83e+00

Pos logit contributions
aces+0.013
coat+0.012
robe+0.012
amas+0.012
bridge+0.012
Neg logit contributions
 coconut-0.009
 Enjoy-0.008
ours-0.007
oral-0.007
grown-0.007
TokenBen
Token ID11696
Feature activation+4.02e-01

Pos logit contributions
aces+0.184
coat+0.178
amas+0.177
robe+0.175
brush+0.166
Neg logit contributions
 coconut-0.094
 Enjoy-0.091
oral-0.073
 shrimp-0.071
iw-0.070
Token anyone
Token ID2687
Feature activation-3.41e+00

Pos logit contributions
 coconut+0.018
 Enjoy+0.016
iw+0.014
 volcano+0.014
ours+0.014
Neg logit contributions
robe-0.023
aces-0.022
coat-0.022
brush-0.021
bridge-0.021
Token had
Token ID550
Feature activation-1.12e-01

Pos logit contributions
 coconut+0.238
 Enjoy+0.225
oral+0.217
iw+0.205
ours+0.204
Neg logit contributions
aces-0.264
coat-0.251
brush-0.242
robe-0.241
bridge-0.220
Token lost
Token ID2626
Feature activation-2.94e+00

Pos logit contributions
 coconut+0.006
 Enjoy+0.006
grown+0.005
 temple+0.005
ours+0.005
Neg logit contributions
aces-0.010
coat-0.010
robe-0.010
amas-0.010
brush-0.010
Token a
Token ID257
Feature activation+5.62e-01

Pos logit contributions
 coconut+0.131
 Enjoy+0.108
 shrimp+0.099
grown+0.097
 oy+0.097
Neg logit contributions
aces-0.305
coat-0.304
amas-0.303
robe-0.295
bridge-0.295
Token bath
Token ID7837
Feature activation+2.49e+00

Pos logit contributions
aces+0.043
coat+0.041
amas+0.040
robe+0.040
bridge+0.040
Neg logit contributions
 coconut-0.042
 Enjoy-0.039
ours-0.035
grown-0.034
 temple-0.034
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
aces+0.196
bridge+0.195
amas+0.190
brush+0.177
amed+0.176
Neg logit contributions
 shrimp-0.144
 coconut-0.138
 crab-0.133
 Enjoy-0.132
 oy-0.130
Token.
Token ID13
Feature activation-4.41e-02

Pos logit contributions
amas+1.274
aces+1.263
bridge+1.228
coat+1.196
robe+1.176
Neg logit contributions
 coconut-0.362
 Enjoy-0.351
 shrimp-0.313
:-0.301
grown-0.297
Token After
Token ID2293
Feature activation-4.18e+00

Pos logit contributions
 coconut+0.003
grown+0.002
 volcano+0.002
 Enjoy+0.002
iw+0.002
Neg logit contributions
robe-0.004
amas-0.004
aces-0.004
coat-0.003
 cartridge-0.003
Token asking
Token ID4737
Feature activation-5.13e-01

Pos logit contributions
 coconut+0.288
 Enjoy+0.252
ours+0.238
iw+0.236
oral+0.231
Neg logit contributions
robe-0.331
aces-0.330
coat-0.322
amas-0.315
bridge-0.314
Token around
Token ID1088
Feature activation+4.36e+00

Pos logit contributions
 coconut+0.033
 Enjoy+0.030
ours+0.026
grown+0.026
 temple+0.026
Neg logit contributions
aces-0.044
coat-0.043
robe-0.043
amas-0.042
brush-0.042
Token,
Token ID11
Feature activation+1.06e-01

Pos logit contributions
aces+0.397
coat+0.389
robe+0.386
amas+0.381
brush+0.375
Neg logit contributions
 coconut-0.255
 Enjoy-0.231
ours-0.200
grown-0.197
 temple-0.197
Token They
Token ID1119
Feature activation-1.23e+00

Pos logit contributions
 coconut+0.003
 Enjoy+0.003
iw+0.003
 temple+0.002
oral+0.002
Neg logit contributions
robe-0.005
aces-0.005
amas-0.005
coat-0.005
brush-0.004
Token see
Token ID766
Feature activation+1.36e+00

Pos logit contributions
 coconut+0.080
 Enjoy+0.073
ours+0.065
grown+0.064
oral+0.063
Neg logit contributions
aces-0.104
coat-0.102
robe-0.101
amas-0.099
brush-0.098
Token the
Token ID262
Feature activation+9.42e-02

Pos logit contributions
aces+0.137
coat+0.135
robe+0.134
amas+0.133
brush+0.130
Neg logit contributions
 coconut-0.066
 Enjoy-0.058
ours-0.049
 temple-0.049
grown-0.049
Token pink
Token ID11398
Feature activation+2.19e+00

Pos logit contributions
aces+0.009
coat+0.008
robe+0.008
amas+0.008
brush+0.008
Neg logit contributions
 coconut-0.005
 Enjoy-0.005
 temple-0.004
 volcano-0.004
grown-0.004
Token bath
Token ID7837
Feature activation+2.46e+00

Pos logit contributions
aces+0.178
coat+0.172
robe+0.169
amas+0.166
brush+0.162
Neg logit contributions
 coconut-0.151
 Enjoy-0.138
 temple-0.128
 volcano-0.126
ours-0.125
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
aces+0.176
coat+0.163
amas+0.156
uster+0.155
 brid+0.153
Neg logit contributions
 volcano-0.161
 temple-0.158
 arose-0.152
ours-0.152
 museum-0.148
Token.
Token ID13
Feature activation-5.60e-02

Pos logit contributions
amas+1.184
aces+1.180
coat+1.169
robe+1.135
brush+1.098
Neg logit contributions
 coconut-0.463
 Enjoy-0.418
 temple-0.407
:-0.374
grown-0.373
Token They
Token ID1119
Feature activation-1.23e+00

Pos logit contributions
 coconut+0.004
 Enjoy+0.003
iw+0.003
 temple+0.003
oral+0.003
Neg logit contributions
robe-0.005
aces-0.005
amas-0.005
coat-0.005
brush-0.005
Token laugh
Token ID6487
Feature activation+5.01e-02

Pos logit contributions
 coconut+0.085
 Enjoy+0.078
ours+0.070
grown+0.069
oral+0.068
Neg logit contributions
aces-0.099
coat-0.096
robe-0.095
amas-0.094
brush-0.093
Token and
Token ID290
Feature activation-8.52e-02

Pos logit contributions
coat+0.006
aces+0.006
robe+0.006
amas+0.006
brush+0.006
Neg logit contributions
:-0.001
 what-0.001
 whose-0.001
 Enjoy-0.001
 asks-0.001
Token shake
Token ID13279
Feature activation+5.43e-02

Pos logit contributions
 coconut+0.005
 Enjoy+0.005
 asks+0.004
 shrimp+0.004
 temple+0.004
Neg logit contributions
aces-0.008
amas-0.008
bridge-0.008
robe-0.008
coat-0.008
Token always
Token ID1464
Feature activation+1.20e+00

Pos logit contributions
aces+0.147
coat+0.144
robe+0.143
amas+0.142
bridge+0.140
Neg logit contributions
 coconut-0.078
 Enjoy-0.070
ours-0.059
grown-0.059
 temple-0.058
Token wore
Token ID12408
Feature activation+1.29e+00

Pos logit contributions
aces+0.106
bridge+0.103
amas+0.102
robe+0.100
brush+0.099
Neg logit contributions
 coconut-0.071
 Enjoy-0.070
 asks-0.060
ours-0.060
 shrimp-0.059
Token the
Token ID262
Feature activation+9.46e-02

Pos logit contributions
aces+0.143
amas+0.142
bridge+0.142
drops+0.141
coat+0.138
Neg logit contributions
 shrimp-0.040
 what-0.039
 Enjoy-0.037
 whose-0.036
 coconut-0.035
Token orange
Token ID10912
Feature activation-5.52e-01

Pos logit contributions
aces+0.009
bridge+0.009
amas+0.009
drops+0.009
brush+0.009
Neg logit contributions
 shrimp-0.005
 coconut-0.005
 oy-0.004
 crab-0.004
 whale-0.004
Token bath
Token ID7837
Feature activation+2.46e+00

Pos logit contributions
 coconut+0.034
 shrimp+0.032
 crab+0.031
 Enjoy+0.030
 oy+0.029
Neg logit contributions
aces-0.046
bridge-0.044
amas-0.043
coat-0.043
drops-0.041
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
bridge+0.093
aces+0.086
amas+0.075
 click+0.071
brush+0.070
Neg logit contributions
 shrimp-0.245
 Enjoy-0.232
:-0.231
arium-0.231
 restaurant-0.231
Token.
Token ID13
Feature activation-5.47e-02

Pos logit contributions
aces+1.182
bridge+1.166
amas+1.157
coat+1.127
robe+1.099
Neg logit contributions
 coconut-0.457
 Enjoy-0.443
 shrimp-0.372
grown-0.360
 temple-0.359
Token Lily
Token ID20037
Feature activation-5.01e-01

Pos logit contributions
 coconut+0.003
 Enjoy+0.003
 temple+0.002
grown+0.002
ours+0.002
Neg logit contributions
aces-0.005
robe-0.005
coat-0.005
amas-0.005
brush-0.005
Token tried
Token ID3088
Feature activation-2.25e+00

Pos logit contributions
 coconut+0.037
 Enjoy+0.034
grown+0.031
iw+0.031
 volcano+0.030
Neg logit contributions
coat-0.038
amas-0.037
robe-0.037
aces-0.037
brush-0.036
Token to
Token ID284
Feature activation-5.53e-01

Pos logit contributions
 coconut+0.134
 Enjoy+0.120
ours+0.104
oral+0.104
 volcano+0.104
Neg logit contributions
coat-0.204
aces-0.204
robe-0.201
brush-0.196
amas-0.195
Token explain
Token ID4727
Feature activation-2.95e+00

Pos logit contributions
 investigate+0.024
 Enjoy+0.024
 coconut+0.023
 question+0.022
 mistake+0.021
Neg logit contributions
bridge-0.057
aces-0.057
amas-0.053
robe-0.052
brush-0.049
Token was
Token ID373
Feature activation-4.00e-01

Pos logit contributions
robe+0.240
coat+0.235
aces+0.229
brush+0.219
amas+0.216
Neg logit contributions
 coconut-0.162
 Enjoy-0.158
ours-0.141
oral-0.139
iw-0.135
Token a
Token ID257
Feature activation+5.34e-01

Pos logit contributions
 coconut+0.024
 Enjoy+0.022
ours+0.019
oral+0.019
 temple+0.019
Neg logit contributions
aces-0.036
coat-0.035
robe-0.035
amas-0.034
brush-0.034
Token big
Token ID1263
Feature activation+1.29e+00

Pos logit contributions
coat+0.037
robe+0.036
brush+0.035
amas+0.033
bridge+0.032
Neg logit contributions
 coconut-0.041
 Enjoy-0.040
ours-0.039
grown-0.036
oral-0.036
Token red
Token ID2266
Feature activation+1.88e+00

Pos logit contributions
coat+0.097
robe+0.094
brush+0.091
aces+0.091
amas+0.087
Neg logit contributions
 coconut-0.093
 Enjoy-0.090
ours-0.082
oral-0.081
grown-0.079
Token bath
Token ID7837
Feature activation+2.46e+00

Pos logit contributions
coat+0.129
robe+0.128
brush+0.120
aces+0.118
amas+0.118
Neg logit contributions
 coconut-0.146
 Enjoy-0.142
ours-0.134
grown-0.131
 asks-0.130
Tokenrobe
Token ID25481
Feature activation+1.15e+01

Pos logit contributions
aces+0.145
amas+0.138
bridge+0.136
brush+0.129
coat+0.124
Neg logit contributions
 coconut-0.193
 shrimp-0.186
 volcano-0.186
 swamp-0.184
 Enjoy-0.184
Token that
Token ID326
Feature activation+8.13e-01

Pos logit contributions
amas+1.286
aces+1.265
bridge+1.193
coat+1.190
amed+1.164
Neg logit contributions
 Enjoy-0.359
 coconut-0.322
 whose-0.302
:-0.300
 museum-0.298
Token belonged
Token ID19611
Feature activation+2.23e+00

Pos logit contributions
aces+0.082
amas+0.082
coat+0.080
robe+0.077
bridge+0.077
Neg logit contributions
 coconut-0.039
 Enjoy-0.033
 temple-0.030
 volcano-0.030
ours-0.029
Token to
Token ID284
Feature activation-5.59e-01

Pos logit contributions
amas+0.252
aces+0.246
bridge+0.244
brush+0.244
coat+0.238
Neg logit contributions
 coconut-0.072
 Enjoy-0.063
grown-0.061
 whose-0.053
 shrimp-0.050
Token their
Token ID511
Feature activation-1.25e+00

Pos logit contributions
 coconut+0.034
 Enjoy+0.029
ours+0.027
iw+0.026
oral+0.026
Neg logit contributions
aces-0.050
robe-0.050
coat-0.049
amas-0.048
bridge-0.047
Token dad
Token ID9955
Feature activation+1.64e+00

Pos logit contributions
 coconut+0.095
 Enjoy+0.090
ours+0.084
 inspection+0.080
grown+0.078
Neg logit contributions
robe-0.088
coat-0.081
amas-0.074
brush-0.073
aces-0.072
Token tree
Token ID5509
Feature activation-2.21e+00

Pos logit contributions
 coconut+0.403
 Enjoy+0.362
 temple+0.320
ours+0.320
 volcano+0.318
Neg logit contributions
aces-0.560
coat-0.549
robe-0.546
amas-0.539
brush-0.524
Token with
Token ID351
Feature activation+1.19e+00

Pos logit contributions
 coconut+0.132
 Enjoy+0.115
 temple+0.105
 volcano+0.103
ours+0.103
Neg logit contributions
aces-0.201
robe-0.197
coat-0.197
amas-0.195
bridge-0.186
Token leaves
Token ID5667
Feature activation-1.63e-01

Pos logit contributions
aces+0.098
brush+0.092
bridge+0.091
coat+0.088
robe+0.087
Neg logit contributions
 Enjoy-0.070
 coconut-0.070
oral-0.069
 exhibit-0.062
ours-0.062
Token and
Token ID290
Feature activation-1.76e-01

Pos logit contributions
 coconut+0.009
 Enjoy+0.008
ours+0.007
 temple+0.007
grown+0.007
Neg logit contributions
aces-0.015
coat-0.015
robe-0.015
amas-0.015
bridge-0.014
Token a
Token ID257
Feature activation+4.51e-01

Pos logit contributions
 coconut+0.012
 Enjoy+0.012
oral+0.011
ours+0.011
grown+0.011
Neg logit contributions
aces-0.013
brush-0.013
coat-0.012
bridge-0.012
robe-0.012
Token coconut
Token ID20132
Feature activation-1.07e+01

Pos logit contributions
coat+0.034
brush+0.034
aces+0.033
robe+0.033
bridge+0.032
Neg logit contributions
 Enjoy-0.030
 coconut-0.030
ours-0.028
grown-0.027
oral-0.027
Token,"
Token ID553
Feature activation-7.59e-03

Pos logit contributions
 coconut+0.663
 Enjoy+0.618
grown+0.534
ours+0.534
oral+0.531
Neg logit contributions
aces-0.931
coat-0.904
robe-0.896
amas-0.882
brush-0.880
Token Tom
Token ID4186
Feature activation+6.11e-03

Pos logit contributions
 coconut+0.001
 Enjoy+0.001
oral+0.000
grown+0.000
ours+0.000
Neg logit contributions
aces-0.001
brush-0.001
coat-0.001
robe-0.000
bridge-0.000
Token said
Token ID531
Feature activation+5.31e-01

Pos logit contributions
aces+0.001
coat+0.001
robe+0.001
amas+0.001
brush+0.001
Neg logit contributions
 coconut-0.000
 Enjoy-0.000
ours-0.000
grown-0.000
 temple-0.000
Token.
Token ID13
Feature activation-1.46e-01

Pos logit contributions
coat+0.054
aces+0.054
amas+0.054
bridge+0.053
robe+0.053
Neg logit contributions
 coconut-0.023
 Enjoy-0.021
ours-0.018
 whose-0.018
 shrimp-0.017
Token He
Token ID679
Feature activation+2.98e-01

Pos logit contributions
 coconut+0.009
 Enjoy+0.008
 temple+0.007
grown+0.007
ours+0.007
Neg logit contributions
aces-0.013
robe-0.013
coat-0.012
amas-0.012
brush-0.012
Token reached
Token ID4251
Feature activation+5.95e-01

Pos logit contributions
aces+0.027
coat+0.026
robe+0.026
amas+0.026
brush+0.025
Neg logit contributions
 coconut-0.019
 Enjoy-0.018
ours-0.015
grown-0.015
 temple-0.015
Token out
Token ID503
Feature activation+1.52e+00

Pos logit contributions
aces+0.048
coat+0.047
robe+0.047
amas+0.046
brush+0.045
Neg logit contributions
 coconut-0.041
 Enjoy-0.038
ours-0.033
oral-0.033
grown-0.033
Token and
Token ID290
Feature activation-1.79e-01

Pos logit contributions
aces+0.135
coat+0.131
brush+0.129
robe+0.128
bridge+0.127
Neg logit contributions
 coconut-0.091
 Enjoy-0.085
ours-0.074
oral-0.073
iw-0.073
Token grabbed
Token ID13176
Feature activation-5.66e-01

Pos logit contributions
 coconut+0.011
 Enjoy+0.010
ours+0.009
grown+0.009
 temple+0.009
Neg logit contributions
aces-0.016
coat-0.015
robe-0.015
amas-0.015
brush-0.015
Token a
Token ID257
Feature activation+4.51e-01

Pos logit contributions
 coconut+0.033
 Enjoy+0.031
oral+0.029
 temple+0.027
ours+0.027
Neg logit contributions
aces-0.051
coat-0.048
brush-0.048
robe-0.048
bridge-0.047
Token coconut
Token ID20132
Feature activation-1.07e+01

Pos logit contributions
aces+0.035
coat+0.034
robe+0.034
amas+0.033
brush+0.033
Neg logit contributions
 coconut-0.032
 Enjoy-0.030
ours-0.027
grown-0.027
oral-0.027
Token.
Token ID13
Feature activation-1.40e-01

Pos logit contributions
 coconut+0.658
 Enjoy+0.568
ours+0.503
iw+0.502
 volcano+0.502
Neg logit contributions
aces-0.949
robe-0.939
coat-0.939
amas-0.931
bridge-0.902
Token He
Token ID679
Feature activation+3.02e-01

Pos logit contributions
 coconut+0.010
 Enjoy+0.008
 temple+0.008
oral+0.008
grown+0.008
Neg logit contributions
aces-0.012
robe-0.011
coat-0.011
amas-0.011
brush-0.011
Token was
Token ID373
Feature activation-4.79e-01

Pos logit contributions
aces+0.024
coat+0.023
brush+0.023
robe+0.023
bridge+0.022
Neg logit contributions
 Enjoy-0.020
 coconut-0.020
oral-0.018
grown-0.018
ours-0.017
Token so
Token ID523
Feature activation+2.66e+00

Pos logit contributions
 coconut+0.033
 Enjoy+0.029
grown+0.026
 stealing+0.026
 oy+0.026
Neg logit contributions
robe-0.041
amas-0.040
coat-0.039
aces-0.038
brush-0.037
Token happy
Token ID3772
Feature activation+1.06e+00

Pos logit contributions
aces+0.196
coat+0.189
robe+0.187
amas+0.183
brush+0.181
Neg logit contributions
 coconut-0.201
 Enjoy-0.188
ours-0.170
oral-0.170
grown-0.168
Token and
Token ID290
Feature activation-1.78e-01

Pos logit contributions
 coconut+0.126
 Enjoy+0.109
 exhibit+0.100
 temple+0.099
 volcano+0.098
Neg logit contributions
aces-0.203
robe-0.200
amas-0.198
coat-0.196
bridge-0.196
Token showed
Token ID3751
Feature activation-1.51e-01

Pos logit contributions
 coconut+0.012
 Enjoy+0.011
oral+0.011
ours+0.010
grown+0.010
Neg logit contributions
aces-0.014
coat-0.014
robe-0.014
brush-0.013
amas-0.013
Token his
Token ID465
Feature activation-7.19e-01

Pos logit contributions
 coconut+0.008
 Enjoy+0.008
oral+0.007
ours+0.007
grown+0.007
Neg logit contributions
aces-0.014
robe-0.013
brush-0.013
coat-0.013
bridge-0.013
Token mom
Token ID1995
Feature activation+6.15e-01

Pos logit contributions
 Enjoy+0.054
ours+0.047
grown+0.046
arium+0.045
 whose+0.044
Neg logit contributions
aces-0.052
brush-0.049
robe-0.047
coat-0.046
bridge-0.045
Token the
Token ID262
Feature activation+1.16e-02

Pos logit contributions
aces+0.051
coat+0.050
robe+0.049
brush+0.049
bridge+0.047
Neg logit contributions
 coconut-0.038
 Enjoy-0.035
oral-0.033
iw-0.031
ours-0.031
Token coconut
Token ID20132
Feature activation-1.07e+01

Pos logit contributions
aces+0.001
brush+0.001
bridge+0.001
coat+0.001
amas+0.001
Neg logit contributions
 Enjoy-0.001
ours-0.001
arium-0.001
 whose-0.001
grown-0.001
Token.
Token ID13
Feature activation-1.33e-01

Pos logit contributions
 coconut+0.661
 Enjoy+0.649
oral+0.562
grown+0.559
ours+0.553
Neg logit contributions
aces-0.922
robe-0.873
coat-0.871
brush-0.864
bridge-0.852
Token She
Token ID1375
Feature activation+2.51e-01

Pos logit contributions
 coconut+0.009
 Enjoy+0.008
 temple+0.007
grown+0.007
 whale+0.007
Neg logit contributions
robe-0.011
amas-0.011
aces-0.011
coat-0.011
brush-0.010
Token was
Token ID373
Feature activation-4.76e-01

Pos logit contributions
aces+0.021
coat+0.020
brush+0.020
robe+0.020
bridge+0.019
Neg logit contributions
 coconut-0.016
 Enjoy-0.016
grown-0.014
ours-0.014
oral-0.014
Token proud
Token ID6613
Feature activation+4.04e-01

Pos logit contributions
 coconut+0.032
 Enjoy+0.029
grown+0.026
 temple+0.025
 Crab+0.025
Neg logit contributions
amas-0.041
robe-0.040
aces-0.040
coat-0.039
brush-0.037
Token of
Token ID286
Feature activation+4.53e-01

Pos logit contributions
aces+0.032
coat+0.031
robe+0.030
brush+0.030
amas+0.030
Neg logit contributions
 coconut-0.028
 Enjoy-0.026
ours-0.023
oral-0.023
 temple-0.023
Token,
Token ID11
Feature activation-6.83e-03

Pos logit contributions
 coconut+0.105
 Enjoy+0.089
 temple+0.082
 volcano+0.081
 exhibit+0.077
Neg logit contributions
aces-0.206
coat-0.201
amas-0.200
robe-0.199
bridge-0.195
Token he
Token ID339
Feature activation+4.61e-01

Pos logit contributions
 coconut+0.000
 Enjoy+0.000
ours+0.000
oral+0.000
iw+0.000
Neg logit contributions
aces-0.001
robe-0.001
coat-0.001
amas-0.001
bridge-0.001
Token saw
Token ID2497
Feature activation-1.07e-01

Pos logit contributions
aces+0.044
coat+0.043
robe+0.043
amas+0.042
bridge+0.042
Neg logit contributions
 coconut-0.024
 Enjoy-0.023
ours-0.019
oral-0.019
iw-0.019
Token a
Token ID257
Feature activation+4.50e-01

Pos logit contributions
 coconut+0.007
 Enjoy+0.007
ours+0.006
iw+0.006
oral+0.006
Neg logit contributions
aces-0.008
bridge-0.008
robe-0.008
brush-0.008
coat-0.008
Token big
Token ID1263
Feature activation+1.21e+00

Pos logit contributions
coat+0.032
brush+0.032
robe+0.031
amas+0.031
bridge+0.030
Neg logit contributions
 Enjoy-0.032
 coconut-0.031
ours-0.030
grown-0.028
iw-0.027
Token coconut
Token ID20132
Feature activation-1.07e+01

Pos logit contributions
coat+0.097
brush+0.096
robe+0.093
amas+0.093
bridge+0.093
Neg logit contributions
 Enjoy-0.073
 coconut-0.073
oral-0.066
ours-0.066
grown-0.062
Token tree
Token ID5509
Feature activation-2.20e+00

Pos logit contributions
 coconut+0.754
 Enjoy+0.747
oral+0.654
ours+0.654
grown+0.639
Neg logit contributions
aces-0.817
coat-0.789
brush-0.781
robe-0.778
bridge-0.772
Token with
Token ID351
Feature activation+1.18e+00

Pos logit contributions
 coconut+0.150
 Enjoy+0.144
oral+0.127
ours+0.127
iw+0.121
Neg logit contributions
aces-0.176
coat-0.173
robe-0.169
brush-0.167
amas-0.166
Token lots
Token ID6041
Feature activation-9.57e-01

Pos logit contributions
aces+0.096
bridge+0.093
brush+0.090
robe+0.089
coat+0.089
Neg logit contributions
 coconut-0.072
 Enjoy-0.072
oral-0.068
ours-0.067
arium-0.064
Token of
Token ID286
Feature activation+4.43e-01

Pos logit contributions
 Enjoy+0.070
 coconut+0.068
ours+0.063
oral+0.062
arium+0.062
Neg logit contributions
aces-0.072
brush-0.067
bridge-0.065
coat-0.064
amas-0.060
Token coc
Token ID8954
Feature activation-3.99e+00

Pos logit contributions
aces+0.032
bridge+0.030
brush+0.030
coat+0.027
amas+0.025
Neg logit contributions
arium-0.032
oral-0.030
 Enjoy-0.030
ours-0.029
 exhibit-0.028
Token the
Token ID262
Feature activation+2.85e-02

Pos logit contributions
aces+0.011
coat+0.011
brush+0.010
robe+0.010
bridge+0.010
Neg logit contributions
 coconut-0.008
 Enjoy-0.008
ours-0.007
oral-0.007
 aquarium-0.006
Token tree
Token ID5509
Feature activation-2.19e+00

Pos logit contributions
coat+0.002
aces+0.002
robe+0.002
brush+0.002
amas+0.002
Neg logit contributions
 Enjoy-0.002
ours-0.002
 coconut-0.002
 whose-0.002
oral-0.002
Token to
Token ID284
Feature activation-6.33e-01

Pos logit contributions
 coconut+0.135
 Enjoy+0.121
 temple+0.107
ours+0.107
grown+0.106
Neg logit contributions
aces-0.192
coat-0.187
robe-0.187
amas-0.185
bridge-0.182
Token get
Token ID651
Feature activation-1.27e+00

Pos logit contributions
 coconut+0.045
oral+0.044
ours+0.043
grown+0.041
 Enjoy+0.040
Neg logit contributions
aces-0.042
coat-0.041
bridge-0.041
brush-0.041
robe-0.039
Token a
Token ID257
Feature activation+4.68e-01

Pos logit contributions
 coconut+0.079
 Enjoy+0.072
oral+0.070
 exhibit+0.067
ours+0.066
Neg logit contributions
aces-0.105
coat-0.103
brush-0.101
robe-0.100
bridge-0.099
Token coconut
Token ID20132
Feature activation-1.07e+01

Pos logit contributions
coat+0.036
brush+0.035
robe+0.032
amas+0.031
aces+0.030
Neg logit contributions
 Enjoy-0.030
 coconut-0.029
ours-0.029
grown-0.029
oral-0.028
Token,
Token ID11
Feature activation+1.10e-02

Pos logit contributions
 coconut+0.615
 Enjoy+0.471
 oy+0.455
 crab+0.451
 volcano+0.445
Neg logit contributions
robe-1.015
amas-1.011
coat-1.006
aces-1.002
bridge-0.969
Token but
Token ID475
Feature activation+2.26e+00

Pos logit contributions
aces+0.001
robe+0.001
coat+0.001
amas+0.001
bridge+0.001
Neg logit contributions
 coconut-0.001
 Enjoy-0.001
ours-0.001
 temple-0.001
grown-0.001
Token she
Token ID673
Feature activation+1.45e+00

Pos logit contributions
aces+0.194
robe+0.189
coat+0.188
brush+0.185
amas+0.183
Neg logit contributions
 coconut-0.142
 Enjoy-0.133
oral-0.116
ours-0.114
grown-0.113
Token was
Token ID373
Feature activation-4.55e-01

Pos logit contributions
aces+0.103
coat+0.097
brush+0.096
robe+0.095
amas+0.093
Neg logit contributions
 Enjoy-0.117
 coconut-0.107
 aquarium-0.100
 museum-0.098
 seafood-0.097
Token scared
Token ID12008
Feature activation-3.56e-01

Pos logit contributions
 coconut+0.035
 Enjoy+0.033
oral+0.030
ours+0.029
 temple+0.029
Neg logit contributions
aces-0.034
coat-0.032
robe-0.031
amas-0.030
brush-0.030
Token how
Token ID703
Feature activation-2.73e-01

Pos logit contributions
 coconut+0.089
 Enjoy+0.084
oral+0.080
grown+0.079
iw+0.074
Neg logit contributions
aces-0.108
brush-0.104
robe-0.103
bridge-0.100
coat-0.099
Token can
Token ID460
Feature activation-2.06e-01

Pos logit contributions
 coconut+0.017
 Enjoy+0.016
iw+0.014
grown+0.014
 Crab+0.013
Neg logit contributions
aces-0.024
robe-0.023
coat-0.023
amas-0.023
brush-0.023
Token we
Token ID356
Feature activation-4.58e-01

Pos logit contributions
 coconut+0.012
 Enjoy+0.011
ours+0.009
 temple+0.009
grown+0.009
Neg logit contributions
aces-0.019
coat-0.019
robe-0.019
amas-0.018
brush-0.018
Token get
Token ID651
Feature activation-1.25e+00

Pos logit contributions
 coconut+0.031
 Enjoy+0.029
ours+0.027
grown+0.027
oral+0.026
Neg logit contributions
aces-0.036
coat-0.036
brush-0.035
robe-0.035
amas-0.034
Token the
Token ID262
Feature activation+3.21e-02

Pos logit contributions
 coconut+0.082
 Enjoy+0.080
oral+0.079
 exhibit+0.075
arium+0.072
Neg logit contributions
brush-0.091
aces-0.090
coat-0.086
amas-0.084
bridge-0.084
Token coconut
Token ID20132
Feature activation-1.07e+01

Pos logit contributions
aces+0.003
coat+0.003
brush+0.003
robe+0.002
bridge+0.002
Neg logit contributions
 Enjoy-0.002
ours-0.002
grown-0.002
 coconut-0.002
 whose-0.002
Token?"
Token ID1701
Feature activation-2.28e+00

Pos logit contributions
 coconut+0.648
 volcano+0.512
 Enjoy+0.503
 temple+0.500
 crab+0.497
Neg logit contributions
aces-0.967
coat-0.966
robe-0.948
amas-0.945
bridge-0.909
Token
Token ID198
Feature activation+6.72e-02

Pos logit contributions
 coconut+0.168
 Enjoy+0.138
 oy+0.135
 whale+0.133
 clam+0.133
Neg logit contributions
aces-0.179
coat-0.173
amas-0.171
robe-0.164
brush-0.162
Token
Token ID198
Feature activation+7.64e-02

Pos logit contributions
aces+0.006
amas+0.006
robe+0.006
coat+0.006
brush+0.006
Neg logit contributions
 coconut-0.004
 Enjoy-0.003
 temple-0.003
 shrimp-0.003
iw-0.003
TokenTom
Token ID13787
Feature activation+1.37e+00

Pos logit contributions
aces+0.007
robe+0.007
coat+0.007
amas+0.007
bridge+0.007
Neg logit contributions
 coconut-0.004
 Enjoy-0.004
oral-0.003
grown-0.003
 temple-0.003
Token had
Token ID550
Feature activation-1.79e-01

Pos logit contributions
aces+0.117
coat+0.114
robe+0.111
amas+0.110
brush+0.109
Neg logit contributions
 coconut-0.090
 Enjoy-0.084
 temple-0.072
grown-0.072
ours-0.072
Token picked
Token ID6497
Feature activation-2.50e+00

Pos logit contributions
 coconut+0.009
 Enjoy+0.008
oral+0.008
ours+0.007
grown+0.007
Neg logit contributions
aces-0.010
coat-0.010
brush-0.010
robe-0.010
bridge-0.009
Token a
Token ID257
Feature activation+5.12e-01

Pos logit contributions
 coconut+0.193
 Enjoy+0.181
oral+0.178
grown+0.170
 volcano+0.169
Neg logit contributions
aces-0.176
brush-0.160
robe-0.160
coat-0.158
bridge-0.157
Token coconut
Token ID20132
Feature activation-1.07e+01

Pos logit contributions
brush+0.034
coat+0.033
amas+0.031
bridge+0.030
aces+0.030
Neg logit contributions
 Enjoy-0.037
ours-0.036
grown-0.034
oral-0.033
 asks-0.033
Token.
Token ID13
Feature activation-9.05e-02

Pos logit contributions
 coconut+0.651
 Enjoy+0.640
oral+0.555
grown+0.550
ours+0.546
Neg logit contributions
aces-0.918
brush-0.879
coat-0.879
robe-0.871
bridge-0.855
Token The
Token ID383
Feature activation-6.16e-01

Pos logit contributions
 coconut+0.007
 Enjoy+0.006
oral+0.006
 whale+0.005
grown+0.005
Neg logit contributions
aces-0.007
amas-0.007
coat-0.007
robe-0.007
brush-0.007
Token coconut
Token ID20132
Feature activation-1.07e+01

Pos logit contributions
 Enjoy+0.047
ours+0.042
arium+0.041
 arose+0.040
 whose+0.040
Neg logit contributions
aces-0.040
brush-0.039
coat-0.038
robe-0.037
bridge-0.037
Token was
Token ID373
Feature activation-4.24e-01

Pos logit contributions
 coconut+0.546
 Enjoy+0.458
 volcano+0.399
 temple+0.399
ours+0.397
Neg logit contributions
aces-1.054
coat-1.043
robe-1.036
amas-1.028
bridge-0.998
Token heavy
Token ID4334
Feature activation+1.25e+00

Pos logit contributions
 coconut+0.029
 Enjoy+0.023
 oy+0.022
grown+0.022
iw+0.022
Neg logit contributions
amas-0.037
robe-0.037
aces-0.036
coat-0.036
bridge-0.034
Token,
Token ID11
Feature activation+5.42e-02

Pos logit contributions
aces+0.114
robe+0.113
coat+0.112
amas+0.111
bridge+0.110
Neg logit contributions
 coconut-0.071
 Enjoy-0.062
grown-0.056
 temple-0.055
 volcano-0.055
Token but
Token ID475
Feature activation+2.30e+00

Pos logit contributions
robe+0.005
amas+0.005
aces+0.005
bridge+0.005
coat+0.005
Neg logit contributions
 coconut-0.003
 stealing-0.003
 Enjoy-0.003
 whose-0.003
grown-0.003
Token she
Token ID673
Feature activation+1.48e+00

Pos logit contributions
aces+0.195
robe+0.189
coat+0.189
brush+0.184
amas+0.183
Neg logit contributions
 coconut-0.148
 Enjoy-0.137
oral-0.121
ours-0.121
grown-0.118
Token the
Token ID262
Feature activation+5.70e-02

Pos logit contributions
aces+0.039
coat+0.038
brush+0.038
robe+0.037
bridge+0.036
Neg logit contributions
 Enjoy-0.030
 coconut-0.030
oral-0.028
ours-0.028
 exhibit-0.027
Token tree
Token ID5509
Feature activation-2.15e+00

Pos logit contributions
brush+0.004
coat+0.004
aces+0.004
drops+0.003
amas+0.003
Neg logit contributions
 Enjoy-0.004
ours-0.004
 whose-0.004
 arose-0.004
 asks-0.003
Token and
Token ID290
Feature activation-1.27e-01

Pos logit contributions
 coconut+0.121
 Enjoy+0.106
 temple+0.097
 exhibit+0.095
 volcano+0.095
Neg logit contributions
aces-0.199
robe-0.197
coat-0.192
amas-0.192
bridge-0.191
Token picked
Token ID6497
Feature activation-2.50e+00

Pos logit contributions
 coconut+0.009
 Enjoy+0.008
oral+0.008
ours+0.007
grown+0.007
Neg logit contributions
aces-0.010
coat-0.010
brush-0.010
robe-0.010
bridge-0.009
Token a
Token ID257
Feature activation+5.12e-01

Pos logit contributions
 coconut+0.193
 Enjoy+0.181
oral+0.178
grown+0.170
 volcano+0.169
Neg logit contributions
aces-0.176
brush-0.160
robe-0.160
coat-0.158
bridge-0.157
Token coconut
Token ID20132
Feature activation-1.07e+01

Pos logit contributions
brush+0.034
coat+0.033
amas+0.031
bridge+0.030
aces+0.030
Neg logit contributions
 Enjoy-0.037
ours-0.036
grown-0.034
oral-0.033
 asks-0.033
Token.
Token ID13
Feature activation-9.05e-02

Pos logit contributions
 coconut+0.651
 Enjoy+0.640
oral+0.555
grown+0.550
ours+0.546
Neg logit contributions
aces-0.918
brush-0.879
coat-0.879
robe-0.871
bridge-0.855
Token The
Token ID383
Feature activation-6.16e-01

Pos logit contributions
 coconut+0.007
 Enjoy+0.006
oral+0.006
 whale+0.005
grown+0.005
Neg logit contributions
aces-0.007
amas-0.007
coat-0.007
robe-0.007
brush-0.007
Token coconut
Token ID20132
Feature activation-1.07e+01

Pos logit contributions
 Enjoy+0.047
ours+0.042
arium+0.041
 arose+0.040
 whose+0.040
Neg logit contributions
aces-0.040
brush-0.039
coat-0.038
robe-0.037
bridge-0.037
Token was
Token ID373
Feature activation-4.24e-01

Pos logit contributions
 coconut+0.546
 Enjoy+0.458
 volcano+0.399
 temple+0.399
ours+0.397
Neg logit contributions
aces-1.054
coat-1.043
robe-1.036
amas-1.028
bridge-0.998
Token heavy
Token ID4334
Feature activation+1.25e+00

Pos logit contributions
 coconut+0.029
 Enjoy+0.023
 oy+0.022
grown+0.022
iw+0.022
Neg logit contributions
amas-0.037
robe-0.037
aces-0.036
coat-0.036
bridge-0.034
Token food
Token ID2057
Feature activation-3.32e-01

Pos logit contributions
aces+0.057
bridge+0.054
robe+0.053
coat+0.053
brush+0.052
Neg logit contributions
 coconut-0.051
oral-0.050
ours-0.050
 Enjoy-0.050
iw-0.047
Token.
Token ID13
Feature activation-4.50e-02

Pos logit contributions
 coconut+0.017
 Enjoy+0.015
 shrimp+0.013
grown+0.013
 temple+0.013
Neg logit contributions
aces-0.032
amas-0.032
coat-0.031
robe-0.031
bridge-0.031
Token He
Token ID679
Feature activation+3.99e-01

Pos logit contributions
 coconut+0.003
 Enjoy+0.003
iw+0.003
grown+0.003
 temple+0.003
Neg logit contributions
aces-0.003
robe-0.003
amas-0.003
coat-0.003
brush-0.003
Token found
Token ID1043
Feature activation+2.37e+00

Pos logit contributions
coat+0.032
aces+0.032
robe+0.032
amas+0.030
brush+0.030
Neg logit contributions
 coconut-0.026
 Enjoy-0.025
oral-0.023
iw-0.022
ours-0.022
Token some
Token ID617
Feature activation-4.53e-01

Pos logit contributions
aces+0.202
robe+0.197
coat+0.193
bridge+0.192
brush+0.190
Neg logit contributions
 coconut-0.146
 Enjoy-0.135
oral-0.123
ours-0.122
 temple-0.119
Token coconut
Token ID20132
Feature activation-1.06e+01

Pos logit contributions
 coconut+0.027
 Enjoy+0.027
ours+0.024
oral+0.024
iw+0.023
Neg logit contributions
aces-0.037
robe-0.036
coat-0.036
brush-0.035
bridge-0.035
Token on
Token ID319
Feature activation-6.18e-01

Pos logit contributions
 Enjoy+0.674
 coconut+0.654
oral+0.593
ours+0.570
grown+0.564
Neg logit contributions
aces-0.890
robe-0.847
coat-0.846
brush-0.840
bridge-0.816
Token the
Token ID262
Feature activation+1.14e-01

Pos logit contributions
 Enjoy+0.046
ours+0.046
oral+0.043
 coconut+0.042
grown+0.041
Neg logit contributions
aces-0.038
brush-0.038
bridge-0.037
robe-0.037
coat-0.036
Token beach
Token ID10481
Feature activation-1.93e+00

Pos logit contributions
aces+0.009
robe+0.009
coat+0.009
amas+0.009
bridge+0.009
Neg logit contributions
 coconut-0.007
 Enjoy-0.007
ours-0.006
grown-0.006
oral-0.006
Token which
Token ID543
Feature activation+1.16e+00

Pos logit contributions
 coconut+0.114
 Enjoy+0.099
 temple+0.090
 volcano+0.089
 exhibit+0.088
Neg logit contributions
aces-0.177
coat-0.172
amas-0.172
robe-0.171
bridge-0.167
Token he
Token ID339
Feature activation+5.54e-01

Pos logit contributions
robe+0.097
aces+0.095
coat+0.093
brush+0.091
bridge+0.091
Neg logit contributions
 Enjoy-0.069
 coconut-0.068
ours-0.062
oral-0.061
grown-0.060
Token,
Token ID11
Feature activation+9.07e-02

Pos logit contributions
 coconut+0.194
 Enjoy+0.179
oral+0.160
grown+0.158
iw+0.154
Neg logit contributions
aces-0.272
coat-0.263
amas-0.260
brush-0.257
robe-0.256
Token it
Token ID340
Feature activation-1.23e+00

Pos logit contributions
aces+0.008
coat+0.008
robe+0.008
brush+0.008
bridge+0.008
Neg logit contributions
 coconut-0.005
 Enjoy-0.005
oral-0.004
grown-0.004
ours-0.004
Token's
Token ID338
Feature activation+1.50e+00

Pos logit contributions
 coconut+0.073
 Enjoy+0.064
 volcano+0.060
ours+0.058
 temple+0.058
Neg logit contributions
aces-0.111
coat-0.110
robe-0.108
amas-0.108
brush-0.104
Token a
Token ID257
Feature activation+5.55e-01

Pos logit contributions
aces+0.130
coat+0.129
robe+0.127
amas+0.124
brush+0.122
Neg logit contributions
 coconut-0.097
 Enjoy-0.086
 volcano-0.077
 temple-0.076
 oy-0.076
Token real
Token ID1103
Feature activation-1.10e+00

Pos logit contributions
coat+0.043
robe+0.042
brush+0.042
bridge+0.040
amas+0.039
Neg logit contributions
 Enjoy-0.035
 coconut-0.035
ours-0.035
oral-0.033
grown-0.033
Token coconut
Token ID20132
Feature activation-1.06e+01

Pos logit contributions
oral+0.064
 Enjoy+0.063
 coconut+0.062
ours+0.062
arium+0.059
Neg logit contributions
coat-0.087
robe-0.085
aces-0.085
brush-0.083
bridge-0.081
Token!"
Token ID2474
Feature activation+1.49e+00

Pos logit contributions
 coconut+0.701
 Enjoy+0.631
ours+0.563
 temple+0.559
 volcano+0.558
Neg logit contributions
aces-0.880
coat-0.866
robe-0.862
amas-0.848
bridge-0.827
Token Tom
Token ID4186
Feature activation+9.88e-02

Pos logit contributions
aces+0.131
robe+0.126
coat+0.125
brush+0.124
amas+0.124
Neg logit contributions
 coconut-0.092
 Enjoy-0.082
ours-0.073
grown-0.072
oral-0.072
Token said
Token ID531
Feature activation+6.26e-01

Pos logit contributions
robe+0.009
aces+0.009
coat+0.009
amas+0.009
bridge+0.009
Neg logit contributions
 coconut-0.005
 Enjoy-0.004
 whose-0.004
 asks-0.004
ours-0.004
Token.
Token ID13
Feature activation-4.07e-02

Pos logit contributions
coat+0.063
aces+0.062
amas+0.062
robe+0.062
bridge+0.061
Neg logit contributions
 coconut-0.029
 Enjoy-0.025
ours-0.022
 whose-0.021
 shrimp-0.021
Token He
Token ID679
Feature activation+4.00e-01

Pos logit contributions
 coconut+0.003
 Enjoy+0.002
 temple+0.002
ours+0.002
grown+0.002
Neg logit contributions
aces-0.003
robe-0.003
coat-0.003
amas-0.003
brush-0.003
Token had
Token ID550
Feature activation-1.90e-01

Pos logit contributions
aces+0.127
coat+0.124
robe+0.123
brush+0.122
bridge+0.120
Neg logit contributions
 coconut-0.092
 Enjoy-0.084
iw-0.071
ours-0.071
oral-0.071
Token to
Token ID284
Feature activation-6.27e-01

Pos logit contributions
 coconut+0.010
 Enjoy+0.009
grown+0.008
 temple+0.008
 whose+0.007
Neg logit contributions
robe-0.018
amas-0.018
aces-0.018
coat-0.018
bridge-0.017
Token know
Token ID760
Feature activation-2.25e+00

Pos logit contributions
 coconut+0.036
 Enjoy+0.033
 temple+0.030
 museum+0.027
 investigate+0.027
Neg logit contributions
aces-0.057
robe-0.056
amas-0.056
coat-0.056
bridge-0.054
Token all
Token ID477
Feature activation-2.24e+00

Pos logit contributions
 coconut+0.093
 Enjoy+0.081
ours+0.065
grown+0.064
 temple+0.063
Neg logit contributions
aces-0.243
coat-0.239
robe-0.237
amas-0.235
brush-0.231
Token about
Token ID546
Feature activation-1.14e+00

Pos logit contributions
 coconut+0.122
 Enjoy+0.109
ours+0.094
oral+0.092
iw+0.092
Neg logit contributions
aces-0.213
coat-0.208
robe-0.207
amas-0.203
brush-0.202
Token microsc
Token ID21145
Feature activation-8.59e+00

Pos logit contributions
 coconut+0.074
 Enjoy+0.068
oral+0.061
ours+0.061
iw+0.060
Neg logit contributions
aces-0.096
robe-0.092
coat-0.092
brush-0.090
amas-0.088
Tokenopes
Token ID13920
Feature activation+1.75e+00

Pos logit contributions
 coconut+0.671
 Enjoy+0.604
ours+0.578
 volcano+0.572
 exhibit+0.570
Neg logit contributions
coat-0.585
robe-0.572
brush-0.563
bridge-0.550
aces-0.524
Token.
Token ID13
Feature activation-1.41e-01

Pos logit contributions
amas+0.166
robe+0.165
coat+0.165
aces+0.164
bridge+0.159
Neg logit contributions
 coconut-0.089
 Enjoy-0.078
 temple-0.073
grown-0.071
 volcano-0.070
Token
Token ID220
Feature activation-1.52e-01

Pos logit contributions
 coconut+0.010
 whale+0.009
 Enjoy+0.008
iw+0.008
 crab+0.008
Neg logit contributions
aces-0.011
amas-0.011
robe-0.011
brush-0.010
coat-0.010
Token
Token ID198
Feature activation+2.57e-02

Pos logit contributions
 coconut+0.005
 Enjoy+0.005
ours+0.004
grown+0.004
oral+0.004
Neg logit contributions
aces-0.017
coat-0.017
robe-0.017
amas-0.016
bridge-0.016
Token
Token ID198
Feature activation+3.59e-02

Pos logit contributions
aces+0.002
amas+0.002
robe+0.002
coat+0.002
brush+0.002
Neg logit contributions
 coconut-0.002
ours-0.001
 Enjoy-0.001
 whale-0.001
 volcano-0.001
Token if
Token ID611
Feature activation-3.83e-01

Pos logit contributions
 coconut+0.122
 Enjoy+0.102
iw+0.097
 volcano+0.096
ours+0.095
Neg logit contributions
aces-0.149
robe-0.145
amas-0.141
coat-0.141
brush-0.139
Token she
Token ID673
Feature activation+1.43e+00

Pos logit contributions
 coconut+0.032
 Enjoy+0.029
ours+0.028
iw+0.027
 exhibit+0.027
Neg logit contributions
robe-0.024
aces-0.024
brush-0.023
coat-0.023
amas-0.022
Token could
Token ID714
Feature activation+1.65e+00

Pos logit contributions
aces+0.092
brush+0.085
robe+0.082
coat+0.081
bridge+0.076
Neg logit contributions
 coconut-0.123
 Enjoy-0.115
ours-0.109
 aquarium-0.109
 volcano-0.108
Token talk
Token ID1561
Feature activation-8.66e-01

Pos logit contributions
aces+0.134
coat+0.131
robe+0.130
amas+0.128
brush+0.126
Neg logit contributions
 coconut-0.113
 Enjoy-0.104
ours-0.092
grown-0.091
 temple-0.091
Token about
Token ID546
Feature activation-1.15e+00

Pos logit contributions
 coconut+0.043
 Enjoy+0.039
grown+0.032
ours+0.032
 temple+0.032
Neg logit contributions
aces-0.086
coat-0.084
robe-0.084
amas-0.083
bridge-0.081
Token microsc
Token ID21145
Feature activation-8.59e+00

Pos logit contributions
 coconut+0.074
 Enjoy+0.070
oral+0.063
ours+0.063
iw+0.060
Neg logit contributions
aces-0.096
robe-0.093
brush-0.091
coat-0.090
bridge-0.090
Tokenopes
Token ID13920
Feature activation+1.77e+00

Pos logit contributions
 coconut+0.662
 Enjoy+0.590
ours+0.566
 volcano+0.563
grown+0.563
Neg logit contributions
coat-0.596
robe-0.576
brush-0.567
bridge-0.555
aces-0.526
Token.
Token ID13
Feature activation-1.32e-01

Pos logit contributions
aces+0.162
coat+0.160
robe+0.159
amas+0.159
bridge+0.154
Neg logit contributions
 coconut-0.098
 Enjoy-0.088
 temple-0.078
grown-0.077
 volcano-0.076
Token
Token ID198
Feature activation+3.71e-02

Pos logit contributions
 coconut+0.010
grown+0.009
iw+0.008
 Enjoy+0.008
 volcano+0.008
Neg logit contributions
amas-0.010
robe-0.009
aces-0.009
brush-0.009
coat-0.009
Token
Token ID198
Feature activation+4.66e-02

Pos logit contributions
amas+0.003
aces+0.003
robe+0.003
brush+0.003
stand+0.003
Neg logit contributions
 coconut-0.003
ours-0.002
 volcano-0.002
 Enjoy-0.002
iw-0.002
TokenHer
Token ID9360
Feature activation-1.21e-01

Pos logit contributions
amas+0.003
aces+0.003
brush+0.003
robe+0.003
coat+0.003
Neg logit contributions
 coconut-0.004
 Enjoy-0.003
ours-0.003
 volcano-0.003
iw-0.003
Token are
Token ID389
Feature activation+1.40e+00

Pos logit contributions
coat+0.041
aces+0.041
amas+0.041
robe+0.040
brush+0.038
Neg logit contributions
 coconut-0.033
 Enjoy-0.031
 temple-0.026
 shrimp-0.026
 volcano-0.026
Token welcome
Token ID7062
Feature activation+9.75e-01

Pos logit contributions
coat+0.124
aces+0.123
robe+0.122
amas+0.122
brush+0.119
Neg logit contributions
 coconut-0.087
 Enjoy-0.082
 temple-0.068
 shrimp-0.068
 volcano-0.068
Token,
Token ID11
Feature activation-1.10e-03

Pos logit contributions
aces+0.094
coat+0.092
robe+0.092
amas+0.090
bridge+0.089
Neg logit contributions
 coconut-0.052
 Enjoy-0.048
grown-0.040
ours-0.040
 temple-0.040
Token kids
Token ID3988
Feature activation+3.57e-01

Pos logit contributions
 coconut+0.000
 Enjoy+0.000
ours+0.000
grown+0.000
 temple+0.000
Neg logit contributions
aces-0.000
coat-0.000
robe-0.000
amas-0.000
brush-0.000
Token.
Token ID13
Feature activation-1.29e-01

Pos logit contributions
aces+0.031
coat+0.031
robe+0.030
amas+0.030
brush+0.030
Neg logit contributions
 coconut-0.022
 Enjoy-0.020
ours-0.018
oral-0.017
 temple-0.017
Token Enjoy
Token ID18179
Feature activation-8.45e+00

Pos logit contributions
 coconut+0.010
 Enjoy+0.008
 temple+0.008
oral+0.008
 volcano+0.008
Neg logit contributions
aces-0.010
robe-0.010
coat-0.009
amas-0.009
bridge-0.009
Token your
Token ID534
Feature activation-1.34e+00

Pos logit contributions
 coconut+0.499
 Enjoy+0.436
grown+0.390
oral+0.388
 volcano+0.386
Neg logit contributions
aces-0.760
coat-0.742
robe-0.738
amas-0.730
brush-0.711
Token ol
Token ID25776
Feature activation-5.23e+00

Pos logit contributions
 coconut+0.090
 Enjoy+0.084
grown+0.078
ours+0.078
iw+0.076
Neg logit contributions
aces-0.106
coat-0.103
robe-0.102
amas-0.099
brush-0.097
Tokenives
Token ID1083
Feature activation+4.62e+00

Pos logit contributions
 Enjoy+0.445
 coconut+0.436
iw+0.398
 shrimp+0.392
ours+0.375
Neg logit contributions
amas-0.310
coat-0.300
bridge-0.298
robe-0.296
aces-0.281
Token.
Token ID13
Feature activation-1.39e-01

Pos logit contributions
aces+0.455
coat+0.454
bridge+0.453
amas+0.442
robe+0.441
Neg logit contributions
 coconut-0.218
 Enjoy-0.202
 shrimp-0.182
 oy-0.170
 Crab-0.165
Token But
Token ID887
Feature activation-1.03e+00

Pos logit contributions
 coconut+0.008
 temple+0.007
grown+0.007
 volcano+0.006
ours+0.006
Neg logit contributions
aces-0.013
robe-0.012
coat-0.012
amas-0.012
brush-0.012
Token are
Token ID389
Feature activation+1.40e+00

Pos logit contributions
 coconut+0.088
 Enjoy+0.082
 temple+0.071
iw+0.071
ours+0.071
Neg logit contributions
aces-0.107
coat-0.106
robe-0.106
amas-0.105
bridge-0.102
Token sweet
Token ID6029
Feature activation+2.73e+00

Pos logit contributions
aces+0.115
coat+0.113
robe+0.113
amas+0.111
brush+0.108
Neg logit contributions
 coconut-0.094
 Enjoy-0.087
ours-0.076
 temple-0.076
grown-0.075
Token and
Token ID290
Feature activation-1.74e-01

Pos logit contributions
aces+0.232
coat+0.227
robe+0.225
amas+0.222
brush+0.218
Neg logit contributions
 coconut-0.176
 Enjoy-0.161
ours-0.142
grown-0.140
 temple-0.140
Token juicy
Token ID35075
Feature activation-3.62e+00

Pos logit contributions
 coconut+0.012
 Enjoy+0.011
ours+0.009
iw+0.009
 temple+0.009
Neg logit contributions
aces-0.014
coat-0.014
robe-0.014
amas-0.014
bridge-0.014
Token.
Token ID13
Feature activation-1.29e-01

Pos logit contributions
 coconut+0.230
 Enjoy+0.213
ours+0.187
oral+0.187
grown+0.185
Neg logit contributions
aces-0.314
coat-0.301
robe-0.296
brush-0.293
amas-0.292
Token Enjoy
Token ID18179
Feature activation-8.45e+00

Pos logit contributions
 coconut+0.008
 Enjoy+0.007
 temple+0.007
 volcano+0.006
grown+0.006
Neg logit contributions
aces-0.011
robe-0.011
coat-0.011
brush-0.011
amas-0.011
Token them
Token ID606
Feature activation+1.22e+00

Pos logit contributions
 coconut+0.486
 Enjoy+0.430
grown+0.383
oral+0.377
 volcano+0.377
Neg logit contributions
aces-0.772
coat-0.752
robe-0.743
amas-0.739
brush-0.733
Token and
Token ID290
Feature activation-1.71e-01

Pos logit contributions
aces+0.112
coat+0.111
robe+0.110
amas+0.108
bridge+0.108
Neg logit contributions
 coconut-0.069
 Enjoy-0.062
ours-0.053
iw-0.053
 temple-0.053
Token stay
Token ID2652
Feature activation+1.76e+00

Pos logit contributions
 coconut+0.011
 Enjoy+0.009
grown+0.009
oral+0.009
ours+0.009
Neg logit contributions
aces-0.014
coat-0.014
brush-0.014
robe-0.014
amas-0.013
Token warm
Token ID5814
Feature activation+3.53e+00

Pos logit contributions
aces+0.154
robe+0.149
coat+0.148
amas+0.146
bridge+0.145
Neg logit contributions
 coconut-0.110
 Enjoy-0.104
 shrimp-0.092
grown-0.090
iw-0.090
Token,"
Token ID553
Feature activation-8.41e-03

Pos logit contributions
aces+0.332
coat+0.327
robe+0.327
amas+0.321
bridge+0.320
Neg logit contributions
 coconut-0.186
 Enjoy-0.170
grown-0.148
 shrimp-0.146
iw-0.145
Token you
Token ID345
Feature activation-4.79e-01

Pos logit contributions
 coconut+0.011
 Enjoy+0.010
grown+0.009
ours+0.009
oral+0.009
Neg logit contributions
aces-0.015
coat-0.014
robe-0.014
amas-0.014
brush-0.014
Token go
Token ID467
Feature activation+1.85e+00

Pos logit contributions
 coconut+0.043
grown+0.039
iw+0.039
 Enjoy+0.039
 temple+0.037
Neg logit contributions
aces-0.028
brush-0.027
robe-0.026
bridge-0.025
amas-0.025
Token,
Token ID11
Feature activation-6.01e-03

Pos logit contributions
aces+0.158
coat+0.157
robe+0.155
amas+0.152
brush+0.150
Neg logit contributions
 coconut-0.117
 Enjoy-0.106
ours-0.095
oral-0.093
grown-0.092
Token Anna
Token ID11735
Feature activation+5.34e-01

Pos logit contributions
 coconut+0.000
 Enjoy+0.000
ours+0.000
grown+0.000
 temple+0.000
Neg logit contributions
aces-0.001
coat-0.001
robe-0.001
amas-0.001
brush-0.001
Token.
Token ID13
Feature activation-1.39e-01

Pos logit contributions
aces+0.050
robe+0.049
coat+0.049
amas+0.049
bridge+0.048
Neg logit contributions
 coconut-0.029
 Enjoy-0.026
 temple-0.023
grown-0.023
iw-0.023
Token Enjoy
Token ID18179
Feature activation-8.45e+00

Pos logit contributions
 coconut+0.009
 temple+0.008
grown+0.008
 Enjoy+0.008
oral+0.007
Neg logit contributions
aces-0.012
robe-0.011
amas-0.011
coat-0.011
brush-0.011
Token your
Token ID534
Feature activation-1.34e+00

Pos logit contributions
 coconut+0.496
 Enjoy+0.440
grown+0.393
oral+0.387
 volcano+0.385
Neg logit contributions
aces-0.758
coat-0.739
robe-0.736
amas-0.734
brush-0.721
Token tuna
Token ID38883
Feature activation-3.17e+00

Pos logit contributions
 coconut+0.072
 Enjoy+0.063
grown+0.058
 temple+0.056
ours+0.056
Neg logit contributions
aces-0.126
robe-0.124
amas-0.123
coat-0.122
brush-0.118
Token!"
Token ID2474
Feature activation+1.38e+00

Pos logit contributions
 coconut+0.190
 Enjoy+0.171
 shrimp+0.148
ours+0.148
 temple+0.147
Neg logit contributions
aces-0.283
coat-0.279
robe-0.274
amas-0.274
bridge-0.271
Token Mom
Token ID11254
Feature activation-1.23e+00

Pos logit contributions
aces+0.111
amas+0.108
brush+0.108
robe+0.107
coat+0.105
Neg logit contributions
 coconut-0.095
 Enjoy-0.082
grown-0.076
iw-0.074
 volcano-0.073
Token says
Token ID1139
Feature activation-9.05e-01

Pos logit contributions
 coconut+0.073
 Enjoy+0.062
grown+0.058
 temple+0.055
oral+0.055
Neg logit contributions
aces-0.112
coat-0.109
amas-0.108
brush-0.106
robe-0.106
Token.
Token ID13
Feature activation-1.40e-01

Pos logit contributions
 coconut+0.012
 Enjoy+0.010
 oy+0.009
 temple+0.009
 shrimp+0.009
Neg logit contributions
amas-0.031
coat-0.031
bridge-0.030
robe-0.030
aces-0.029
Token Fresh
Token ID20138
Feature activation-4.36e+00

Pos logit contributions
 coconut+0.010
 Enjoy+0.008
 temple+0.008
grown+0.008
 volcano+0.008
Neg logit contributions
aces-0.011
robe-0.011
coat-0.011
amas-0.011
brush-0.011
Token and
Token ID290
Feature activation-1.85e-01

Pos logit contributions
 coconut+0.282
 Enjoy+0.257
ours+0.226
grown+0.224
 temple+0.223
Neg logit contributions
aces-0.370
coat-0.364
robe-0.361
amas-0.357
brush-0.349
Token healthy
Token ID5448
Feature activation+1.08e+00

Pos logit contributions
 coconut+0.013
 Enjoy+0.012
oral+0.011
grown+0.011
ours+0.011
Neg logit contributions
aces-0.015
coat-0.014
robe-0.014
brush-0.014
amas-0.013
Token.
Token ID13
Feature activation-1.37e-01

Pos logit contributions
aces+0.092
coat+0.088
robe+0.088
brush+0.086
amas+0.086
Neg logit contributions
 coconut-0.068
 Enjoy-0.064
ours-0.057
oral-0.056
grown-0.055
Token Enjoy
Token ID18179
Feature activation-8.45e+00

Pos logit contributions
 coconut+0.009
 Enjoy+0.008
 temple+0.008
grown+0.007
 volcano+0.007
Neg logit contributions
aces-0.012
robe-0.011
coat-0.011
amas-0.011
brush-0.011
Token,"
Token ID553
Feature activation-4.05e-04

Pos logit contributions
 coconut+0.495
 Enjoy+0.444
grown+0.391
oral+0.389
 volcano+0.383
Neg logit contributions
aces-0.764
coat-0.742
robe-0.739
amas-0.728
brush-0.724
Token he
Token ID339
Feature activation+4.61e-01

Pos logit contributions
 coconut+0.000
 Enjoy+0.000
 colony+0.000
arium+0.000
 exhibit+0.000
Neg logit contributions
aces-0.000
robe-0.000
brush-0.000
bridge-0.000
coat-0.000
Token says
Token ID1139
Feature activation-8.87e-01

Pos logit contributions
aces+0.049
coat+0.049
robe+0.048
amas+0.048
brush+0.047
Neg logit contributions
 coconut-0.020
 Enjoy-0.017
ours-0.014
 temple-0.014
grown-0.013
Token.
Token ID13
Feature activation-1.34e-01

Pos logit contributions
 coconut+0.039
 Enjoy+0.036
ours+0.031
 whose+0.030
grown+0.030
Neg logit contributions
coat-0.090
aces-0.088
amas-0.088
robe-0.088
bridge-0.086
Token
Token ID198
Feature activation+4.25e-02

Pos logit contributions
 coconut+0.007
 Enjoy+0.006
 temple+0.006
 volcano+0.005
ours+0.005
Neg logit contributions
aces-0.013
robe-0.013
coat-0.013
brush-0.012
amas-0.012
Token Here
Token ID3423
Feature activation-2.24e+00

Pos logit contributions
 coconut+0.008
 temple+0.007
 Enjoy+0.007
 volcano+0.006
grown+0.006
Neg logit contributions
aces-0.009
robe-0.009
coat-0.009
amas-0.009
brush-0.009
Token are
Token ID389
Feature activation+1.41e+00

Pos logit contributions
 coconut+0.114
 Enjoy+0.103
 temple+0.090
ours+0.086
 volcano+0.086
Neg logit contributions
aces-0.219
coat-0.217
robe-0.216
amas-0.214
bridge-0.210
Token your
Token ID534
Feature activation-1.32e+00

Pos logit contributions
aces+0.111
robe+0.105
coat+0.105
amas+0.102
brush+0.101
Neg logit contributions
 coconut-0.101
 Enjoy-0.086
grown-0.082
ours-0.082
oral-0.081
Token toys
Token ID14958
Feature activation+1.08e+00

Pos logit contributions
 coconut+0.092
 Enjoy+0.085
ours+0.076
grown+0.075
 temple+0.074
Neg logit contributions
aces-0.105
coat-0.103
robe-0.102
amas-0.101
brush-0.099
Token.
Token ID13
Feature activation-1.28e-01

Pos logit contributions
aces+0.111
amas+0.111
bridge+0.109
coat+0.107
robe+0.106
Neg logit contributions
 coconut-0.042
 Enjoy-0.040
 temple-0.037
 museum-0.034
 shrimp-0.033
Token Enjoy
Token ID18179
Feature activation-8.44e+00

Pos logit contributions
 coconut+0.009
 temple+0.007
 Enjoy+0.007
 volcano+0.007
grown+0.007
Neg logit contributions
aces-0.011
robe-0.010
coat-0.010
amas-0.010
brush-0.010
Token them
Token ID606
Feature activation+1.22e+00

Pos logit contributions
 coconut+0.446
 Enjoy+0.401
ours+0.341
 temple+0.335
grown+0.335
Neg logit contributions
aces-0.818
coat-0.802
robe-0.798
amas-0.788
brush-0.775
Token,
Token ID11
Feature activation-6.26e-03

Pos logit contributions
bridge+0.117
coat+0.116
aces+0.116
robe+0.114
brush+0.113
Neg logit contributions
 coconut-0.059
 Enjoy-0.057
 shrimp-0.047
 temple-0.046
 whose-0.046
Token but
Token ID475
Feature activation+2.25e+00

Pos logit contributions
 coconut+0.000
 Enjoy+0.000
 temple+0.000
 shrimp+0.000
ours+0.000
Neg logit contributions
aces-0.001
coat-0.001
robe-0.001
amas-0.001
bridge-0.001
Token remember
Token ID3505
Feature activation-1.97e+00

Pos logit contributions
robe+0.170
aces+0.168
brush+0.164
coat+0.163
amas+0.157
Neg logit contributions
 coconut-0.158
grown-0.132
 Enjoy-0.132
ours-0.128
 exhibit-0.128
Token to
Token ID284
Feature activation-6.43e-01

Pos logit contributions
 coconut+0.134
 Enjoy+0.110
iw+0.109
oral+0.104
rust+0.104
Neg logit contributions
aces-0.154
robe-0.151
coat-0.150
brush-0.148
amas-0.146
Token you
Token ID345
Feature activation-4.73e-01

Pos logit contributions
 coconut+0.010
 Enjoy+0.009
grown+0.008
ours+0.008
oral+0.008
Neg logit contributions
aces-0.014
coat-0.014
robe-0.014
amas-0.014
brush-0.013
Token go
Token ID467
Feature activation+1.86e+00

Pos logit contributions
 coconut+0.036
 Enjoy+0.034
ours+0.030
grown+0.030
 temple+0.030
Neg logit contributions
aces-0.035
coat-0.034
robe-0.033
amas-0.033
brush-0.032
Token,
Token ID11
Feature activation-6.26e-03

Pos logit contributions
aces+0.177
coat+0.172
robe+0.172
amas+0.170
bridge+0.168
Neg logit contributions
 coconut-0.102
 Enjoy-0.091
 temple-0.078
ours-0.078
grown-0.078
Token Spot
Token ID15899
Feature activation+1.66e+00

Pos logit contributions
 coconut+0.000
 Enjoy+0.000
grown+0.000
iw+0.000
oral+0.000
Neg logit contributions
aces-0.001
robe-0.001
coat-0.001
amas-0.001
bridge-0.001
Token.
Token ID13
Feature activation-1.49e-01

Pos logit contributions
aces+0.157
coat+0.153
robe+0.152
amas+0.151
bridge+0.149
Neg logit contributions
 coconut-0.091
 Enjoy-0.084
 volcano-0.071
 temple-0.071
 exhibit-0.071
Token Enjoy
Token ID18179
Feature activation-8.44e+00

Pos logit contributions
 coconut+0.010
 Enjoy+0.009
 temple+0.008
grown+0.008
oral+0.008
Neg logit contributions
aces-0.012
robe-0.012
coat-0.012
amas-0.012
brush-0.012
Token your
Token ID534
Feature activation-1.33e+00

Pos logit contributions
 coconut+0.462
 Enjoy+0.414
ours+0.355
grown+0.352
 temple+0.348
Neg logit contributions
aces-0.801
coat-0.785
robe-0.781
amas-0.772
brush-0.759
Token meal
Token ID9799
Feature activation-4.71e-01

Pos logit contributions
 coconut+0.082
 Enjoy+0.075
ours+0.066
 temple+0.065
 shrimp+0.065
Neg logit contributions
aces-0.118
coat-0.115
robe-0.113
amas-0.112
bridge-0.112
Token,"
Token ID553
Feature activation-6.07e-03

Pos logit contributions
 coconut+0.020
 Enjoy+0.018
 shrimp+0.015
 oy+0.015
 temple+0.015
Neg logit contributions
bridge-0.048
coat-0.047
aces-0.047
robe-0.047
amas-0.046
Token Ben
Token ID3932
Feature activation+7.62e-01

Pos logit contributions
 coconut+0.000
 Enjoy+0.000
grown+0.000
 exhibit+0.000
ours+0.000
Neg logit contributions
aces-0.000
robe-0.000
amas-0.000
coat-0.000
brush-0.000
Token says
Token ID1139
Feature activation-8.89e-01

Pos logit contributions
aces+0.078
coat+0.076
robe+0.076
amas+0.075
brush+0.074
Neg logit contributions
 coconut-0.036
 Enjoy-0.032
grown-0.027
 temple-0.026
ours-0.026
TokenThis
Token ID1212
Feature activation+8.25e-01

Pos logit contributions
aces+0.106
coat+0.104
robe+0.104
amas+0.104
brush+0.099
Neg logit contributions
 coconut-0.082
 Enjoy-0.080
iw-0.068
oral-0.067
 Crab-0.065
Token is
Token ID318
Feature activation-5.44e-01

Pos logit contributions
aces+0.082
coat+0.081
robe+0.080
amas+0.079
brush+0.078
Neg logit contributions
 coconut-0.041
 Enjoy-0.036
ours-0.031
grown-0.031
 temple-0.030
Token your
Token ID534
Feature activation-1.34e+00

Pos logit contributions
 coconut+0.035
 Enjoy+0.031
ours+0.028
grown+0.028
 temple+0.027
Neg logit contributions
aces-0.047
coat-0.046
robe-0.045
amas-0.045
brush-0.044
Token reward
Token ID6721
Feature activation-6.04e-01

Pos logit contributions
 coconut+0.089
 Enjoy+0.082
ours+0.076
grown+0.075
oral+0.074
Neg logit contributions
aces-0.109
robe-0.107
coat-0.107
amas-0.102
brush-0.102
Token.
Token ID13
Feature activation-1.35e-01

Pos logit contributions
 coconut+0.028
 Enjoy+0.026
 whose+0.022
 exhibit+0.022
 shrimp+0.022
Neg logit contributions
amas-0.059
aces-0.058
coat-0.058
bridge-0.058
brush-0.057
Token Enjoy
Token ID18179
Feature activation-8.44e+00

Pos logit contributions
 coconut+0.009
 temple+0.008
grown+0.008
 Enjoy+0.008
oral+0.007
Neg logit contributions
aces-0.011
robe-0.011
amas-0.010
coat-0.010
bridge-0.010
Token!"
Token ID2474
Feature activation+1.40e+00

Pos logit contributions
 coconut+0.477
 Enjoy+0.427
ours+0.370
grown+0.368
oral+0.366
Neg logit contributions
aces-0.785
coat-0.768
robe-0.764
amas-0.756
brush-0.742
Token
Token ID198
Feature activation+4.59e-02

Pos logit contributions
aces+0.130
coat+0.126
robe+0.125
amas+0.125
bridge+0.123
Neg logit contributions
 coconut-0.079
 Enjoy-0.069
 temple-0.063
ours-0.063
grown-0.061
Token
Token ID198
Feature activation+4.12e-02

Pos logit contributions
amas+0.004
robe+0.004
aces+0.004
coat+0.004
bridge+0.004
Neg logit contributions
 coconut-0.003
 Enjoy-0.003
ours-0.002
 temple-0.002
 oy-0.002
TokenTim
Token ID14967
Feature activation+9.08e-01

Pos logit contributions
coat+0.002
bridge+0.002
brush+0.002
aces+0.002
robe+0.002
Neg logit contributions
 coconut-0.003
 Enjoy-0.003
ours-0.003
 temple-0.003
 volcano-0.003
Token and
Token ID290
Feature activation-1.63e-01

Pos logit contributions
aces+0.078
coat+0.077
robe+0.077
amas+0.076
bridge+0.074
Neg logit contributions
 coconut-0.057
 Enjoy-0.052
ours-0.046
 temple-0.045
grown-0.045
Token.
Token ID13
Feature activation-1.45e-01

Pos logit contributions
 coconut+0.015
 Enjoy+0.013
grown+0.013
ours+0.013
 temple+0.012
Neg logit contributions
aces-0.019
coat-0.019
brush-0.019
robe-0.019
amas-0.019
Token Here
Token ID3423
Feature activation-2.26e+00

Pos logit contributions
 coconut+0.010
 temple+0.008
grown+0.008
 Enjoy+0.008
 volcano+0.008
Neg logit contributions
aces-0.012
robe-0.011
coat-0.011
amas-0.011
brush-0.011
Token you
Token ID345
Feature activation-4.73e-01

Pos logit contributions
 coconut+0.127
 Enjoy+0.115
 temple+0.099
ours+0.099
 volcano+0.097
Neg logit contributions
aces-0.209
coat-0.207
robe-0.206
amas-0.203
bridge-0.199
Token go
Token ID467
Feature activation+1.85e+00

Pos logit contributions
 coconut+0.035
 Enjoy+0.031
grown+0.030
iw+0.029
ours+0.029
Neg logit contributions
aces-0.035
robe-0.034
brush-0.034
coat-0.033
bridge-0.033
Token.
Token ID13
Feature activation-1.38e-01

Pos logit contributions
aces+0.146
coat+0.145
robe+0.143
amas+0.141
brush+0.138
Neg logit contributions
 coconut-0.129
 Enjoy-0.113
oral-0.105
ours-0.104
grown-0.103
Token Enjoy
Token ID18179
Feature activation-8.44e+00

Pos logit contributions
 coconut+0.009
grown+0.007
 temple+0.007
 volcano+0.007
oral+0.007
Neg logit contributions
aces-0.012
robe-0.011
amas-0.011
brush-0.011
coat-0.011
Token!"
Token ID2474
Feature activation+1.38e+00

Pos logit contributions
 coconut+0.478
 Enjoy+0.423
grown+0.369
ours+0.369
oral+0.366
Neg logit contributions
aces-0.781
coat-0.765
robe-0.760
amas-0.754
brush-0.740
Token
Token ID220
Feature activation-1.57e-01

Pos logit contributions
aces+0.131
coat+0.127
robe+0.127
amas+0.126
brush+0.123
Neg logit contributions
 coconut-0.076
 Enjoy-0.067
ours-0.059
grown-0.059
 temple-0.058
Token
Token ID198
Feature activation+3.79e-02

Pos logit contributions
 coconut+0.007
 Enjoy+0.006
 temple+0.005
grown+0.005
 volcano+0.005
Neg logit contributions
aces-0.016
robe-0.016
coat-0.016
amas-0.016
brush-0.015
Token
Token ID198
Feature activation+3.93e-02

Pos logit contributions
aces+0.004
coat+0.004
robe+0.004
amas+0.004
brush+0.004
Neg logit contributions
 coconut-0.002
 Enjoy-0.002
ours-0.001
grown-0.001
 temple-0.001
TokenThe
Token ID464
Feature activation+1.03e+00

Pos logit contributions
aces+0.003
robe+0.003
coat+0.003
amas+0.003
brush+0.003
Neg logit contributions
 coconut-0.003
 Enjoy-0.003
oral-0.002
iw-0.002
 temple-0.002
Token were
Token ID547
Feature activation+1.09e+00

Pos logit contributions
aces+0.179
amas+0.176
coat+0.175
robe+0.174
brush+0.168
Neg logit contributions
 coconut-0.108
 Enjoy-0.098
 temple-0.089
grown-0.088
 whose-0.088
Token dressed
Token ID12049
Feature activation+1.81e+00

Pos logit contributions
aces+0.098
coat+0.098
robe+0.097
amas+0.096
brush+0.091
Neg logit contributions
 coconut-0.064
 Enjoy-0.059
 temple-0.056
grown-0.055
 whale-0.054
Token in
Token ID287
Feature activation-1.36e+00

Pos logit contributions
aces+0.187
coat+0.180
brush+0.178
 cartridge+0.176
amas+0.174
Neg logit contributions
 coconut-0.071
 shrimp-0.065
 whose-0.064
grown-0.064
 temple-0.063
Token their
Token ID511
Feature activation-1.26e+00

Pos logit contributions
 coconut+0.078
 Enjoy+0.070
ours+0.062
iw+0.060
oral+0.060
Neg logit contributions
aces-0.124
coat-0.122
robe-0.121
amas-0.120
bridge-0.119
Token best
Token ID1266
Feature activation+2.25e+00

Pos logit contributions
 coconut+0.087
 Enjoy+0.080
ours+0.071
 temple+0.070
grown+0.070
Neg logit contributions
aces-0.102
coat-0.100
robe-0.099
amas-0.098
brush-0.096
Token clothes
Token ID8242
Feature activation+6.14e+00

Pos logit contributions
aces+0.205
coat+0.199
robe+0.198
amas+0.196
bridge+0.192
Neg logit contributions
 coconut-0.132
 Enjoy-0.119
 temple-0.106
 volcano-0.104
grown-0.103
Token.
Token ID13
Feature activation-5.40e-02

Pos logit contributions
amas+0.660
aces+0.653
bridge+0.643
coat+0.629
robe+0.623
Neg logit contributions
 coconut-0.200
 Enjoy-0.171
 temple-0.171
grown-0.165
 whose-0.162
Token There
Token ID1318
Feature activation+9.79e-01

Pos logit contributions
 coconut+0.004
iw+0.003
 Enjoy+0.003
 crab+0.003
 oy+0.003
Neg logit contributions
aces-0.004
robe-0.004
amas-0.004
coat-0.004
bridge-0.004
Token were
Token ID547
Feature activation+1.10e+00

Pos logit contributions
aces+0.090
coat+0.089
robe+0.088
amas+0.087
brush+0.085
Neg logit contributions
 coconut-0.056
 Enjoy-0.051
 temple-0.044
ours-0.044
 volcano-0.044
Token balloons
Token ID35485
Feature activation+1.31e+00

Pos logit contributions
aces+0.099
coat+0.099
amas+0.098
robe+0.097
brush+0.093
Neg logit contributions
 coconut-0.065
 Enjoy-0.059
 temple-0.053
 volcano-0.053
grown-0.052
Token and
Token ID290
Feature activation-9.19e-02

Pos logit contributions
aces+0.135
amas+0.132
coat+0.129
robe+0.128
bridge+0.123
Neg logit contributions
 coconut-0.062
 temple-0.051
 volcano-0.050
 shrimp-0.049
 Enjoy-0.049
Token.
Token ID13
Feature activation-7.70e-02

Pos logit contributions
 coconut+0.031
 Enjoy+0.027
ours+0.024
 temple+0.024
grown+0.024
Neg logit contributions
aces-0.056
coat-0.055
robe-0.055
amas-0.054
bridge-0.054
Token Tom
Token ID4186
Feature activation+6.12e-02

Pos logit contributions
 coconut+0.005
 Enjoy+0.004
 volcano+0.004
 temple+0.004
iw+0.004
Neg logit contributions
aces-0.007
amas-0.007
coat-0.007
robe-0.006
brush-0.006
Token and
Token ID290
Feature activation-1.22e-01

Pos logit contributions
aces+0.005
coat+0.005
robe+0.005
amas+0.005
brush+0.005
Neg logit contributions
 coconut-0.004
 Enjoy-0.004
oral-0.003
grown-0.003
iw-0.003
Token Mia
Token ID32189
Feature activation-2.92e+00

Pos logit contributions
 coconut+0.006
 Enjoy+0.005
oral+0.005
iw+0.005
grown+0.005
Neg logit contributions
aces-0.012
coat-0.012
robe-0.012
brush-0.011
bridge-0.011
Token wear
Token ID5806
Feature activation+1.39e+00

Pos logit contributions
 coconut+0.208
 Enjoy+0.194
oral+0.180
iw+0.177
grown+0.177
Neg logit contributions
aces-0.227
coat-0.221
brush-0.211
robe-0.209
amas-0.204
Token hats
Token ID23910
Feature activation+7.52e+00

Pos logit contributions
bridge+0.156
amas+0.156
aces+0.155
robe+0.149
 cartridge+0.148
Neg logit contributions
 coconut-0.043
 shrimp-0.035
 whose-0.035
 Enjoy-0.033
 their-0.031
Token and
Token ID290
Feature activation-1.18e-01

Pos logit contributions
amas+0.844
bridge+0.818
aces+0.808
robe+0.787
 cartridge+0.767
Neg logit contributions
 coconut-0.212
:-0.202
 whose-0.187
grown-0.178
 shrimp-0.178
Token coats
Token ID30720
Feature activation+5.10e+00

Pos logit contributions
 coconut+0.010
 Enjoy+0.009
 shrimp+0.009
 oy+0.009
 crab+0.009
Neg logit contributions
amas-0.007
aces-0.007
robe-0.007
bridge-0.007
coat-0.007
Token,
Token ID11
Feature activation+5.98e-02

Pos logit contributions
amas+0.573
aces+0.559
bridge+0.555
robe+0.538
 cartridge+0.533
Neg logit contributions
 coconut-0.151
:-0.129
 whose-0.127
 Enjoy-0.120
grown-0.113
Token but
Token ID475
Feature activation+2.30e+00

Pos logit contributions
amas+0.005
robe+0.005
aces+0.005
bridge+0.005
amed+0.005
Neg logit contributions
 coconut-0.004
 shrimp-0.003
 whose-0.003
 oy-0.003
 crab-0.003
Token they
Token ID484
Feature activation-3.43e-01

Pos logit contributions
aces+0.233
coat+0.228
robe+0.227
amas+0.224
brush+0.221
Neg logit contributions
 coconut-0.111
 Enjoy-0.099
ours-0.083
grown-0.081
 temple-0.081
Token They
Token ID1119
Feature activation-1.24e+00

Pos logit contributions
 coconut+0.004
 Enjoy+0.003
 temple+0.003
 volcano+0.003
iw+0.003
Neg logit contributions
aces-0.007
robe-0.007
coat-0.007
amas-0.007
brush-0.007
Token took
Token ID1718
Feature activation-1.95e-01

Pos logit contributions
 coconut+0.087
 Enjoy+0.080
ours+0.073
grown+0.071
 volcano+0.071
Neg logit contributions
aces-0.098
coat-0.096
robe-0.096
brush-0.093
amas-0.092
Token off
Token ID572
Feature activation+7.10e-01

Pos logit contributions
 coconut+0.010
 Enjoy+0.010
ours+0.008
 shrimp+0.008
grown+0.008
Neg logit contributions
aces-0.019
coat-0.019
amas-0.018
robe-0.018
bridge-0.018
Token the
Token ID262
Feature activation+7.81e-02

Pos logit contributions
aces+0.062
coat+0.062
robe+0.061
brush+0.060
amas+0.059
Neg logit contributions
 coconut-0.044
 Enjoy-0.040
ours-0.036
oral-0.034
 volcano-0.034
Token band
Token ID4097
Feature activation+2.18e+00

Pos logit contributions
aces+0.006
bridge+0.006
robe+0.006
amas+0.006
coat+0.006
Neg logit contributions
 coconut-0.005
 Enjoy-0.005
 temple-0.005
 volcano-0.005
oral-0.005
Tokenage
Token ID496
Feature activation+7.80e+00

Pos logit contributions
amas+0.128
bridge+0.119
robe+0.117
aces+0.117
 cartridge+0.115
Neg logit contributions
 exhibit-0.172
 coconut-0.170
 Enjoy-0.162
 named-0.160
:-0.158
Token and
Token ID290
Feature activation-1.01e-01

Pos logit contributions
bridge+0.862
amas+0.845
aces+0.839
coat+0.822
robe+0.816
Neg logit contributions
 coconut-0.239
 temple-0.212
:-0.208
 exhibit-0.193
 Enjoy-0.186
Token let
Token ID1309
Feature activation+4.84e-02

Pos logit contributions
 coconut+0.006
 Enjoy+0.006
 temple+0.005
 oy+0.005
 shrimp+0.005
Neg logit contributions
aces-0.009
amas-0.009
robe-0.009
coat-0.009
bridge-0.009
Token Lily
Token ID20037
Feature activation-5.05e-01

Pos logit contributions
aces+0.005
coat+0.005
amas+0.005
bridge+0.005
robe+0.005
Neg logit contributions
 coconut-0.002
 Enjoy-0.002
ours-0.001
 temple-0.001
 shrimp-0.001
Token run
Token ID1057
Feature activation+4.82e-02

Pos logit contributions
 coconut+0.026
 Enjoy+0.026
 whose+0.022
 exhibit+0.022
ours+0.021
Neg logit contributions
aces-0.047
amas-0.046
bridge-0.046
robe-0.046
coat-0.043
Token in
Token ID287
Feature activation-1.36e+00

Pos logit contributions
aces+0.005
amas+0.005
bridge+0.004
robe+0.004
brush+0.004
Neg logit contributions
 Enjoy-0.002
 coconut-0.002
 exhibit-0.002
grown-0.002
 investigate-0.002
Token can
Token ID460
Feature activation-1.69e-01

Pos logit contributions
aces+0.025
coat+0.025
amas+0.025
robe+0.025
bridge+0.024
Neg logit contributions
 coconut-0.015
 Enjoy-0.014
 temple-0.012
 volcano-0.012
ours-0.011
Token take
Token ID1011
Feature activation-2.20e+00

Pos logit contributions
 coconut+0.011
 Enjoy+0.010
ours+0.009
 temple+0.009
grown+0.009
Neg logit contributions
aces-0.014
coat-0.014
robe-0.014
amas-0.014
bridge-0.013
Token her
Token ID607
Feature activation+1.21e+00

Pos logit contributions
 coconut+0.106
 Enjoy+0.104
grown+0.082
 shrimp+0.081
 temple+0.080
Neg logit contributions
coat-0.217
robe-0.217
aces-0.214
amas-0.214
bridge-0.210
Token with
Token ID351
Feature activation+1.26e+00

Pos logit contributions
aces+0.109
amas+0.106
bridge+0.106
coat+0.105
robe+0.104
Neg logit contributions
 coconut-0.070
 Enjoy-0.063
 temple-0.057
 shrimp-0.056
 whose-0.056
Token you
Token ID345
Feature activation-4.09e-01

Pos logit contributions
aces+0.121
coat+0.121
robe+0.120
bridge+0.118
amas+0.118
Neg logit contributions
 coconut-0.064
 Enjoy-0.060
 shrimp-0.050
grown-0.050
 temple-0.049
Token wherever
Token ID14530
Feature activation+6.57e+00

Pos logit contributions
 coconut+0.024
 Enjoy+0.022
ours+0.019
grown+0.019
 temple+0.019
Neg logit contributions
aces-0.037
coat-0.036
robe-0.036
amas-0.035
bridge-0.035
Token you
Token ID345
Feature activation-3.99e-01

Pos logit contributions
aces+0.620
coat+0.618
robe+0.604
bridge+0.596
amas+0.593
Neg logit contributions
 coconut-0.343
 Enjoy-0.333
 temple-0.286
ours-0.279
grown-0.275
Token go
Token ID467
Feature activation+1.92e+00

Pos logit contributions
 coconut+0.027
 Enjoy+0.024
ours+0.022
grown+0.022
 volcano+0.022
Neg logit contributions
aces-0.033
coat-0.032
robe-0.032
amas-0.031
brush-0.031
Token."
Token ID526
Feature activation+1.53e+00

Pos logit contributions
aces+0.183
coat+0.178
bridge+0.178
robe+0.177
amas+0.176
Neg logit contributions
 coconut-0.095
 Enjoy-0.092
grown-0.081
 temple-0.079
ours-0.078
Token
Token ID198
Feature activation+1.04e-01

Pos logit contributions
aces+0.150
robe+0.146
amas+0.145
coat+0.144
brush+0.141
Neg logit contributions
 coconut-0.083
 Enjoy-0.068
 volcano-0.063
 temple-0.062
grown-0.062
Token
Token ID198
Feature activation+1.07e-01

Pos logit contributions
aces+0.009
robe+0.009
coat+0.009
amas+0.009
brush+0.009
Neg logit contributions
 coconut-0.006
 Enjoy-0.005
 volcano-0.005
grown-0.005
 temple-0.005
Token Sam
Token ID3409
Feature activation+6.62e-02

Pos logit contributions
 coconut+0.007
 Enjoy+0.006
oral+0.006
ours+0.006
iw+0.006
Neg logit contributions
aces-0.011
coat-0.011
robe-0.011
amas-0.011
brush-0.011
Token noticed
Token ID6810
Feature activation-6.56e-02

Pos logit contributions
aces+0.006
coat+0.005
robe+0.005
amas+0.005
brush+0.005
Neg logit contributions
 coconut-0.004
 Enjoy-0.004
oral-0.003
 volcano-0.003
ours-0.003
Token a
Token ID257
Feature activation+5.13e-01

Pos logit contributions
 coconut+0.004
 Enjoy+0.004
iw+0.004
oral+0.004
ours+0.004
Neg logit contributions
aces-0.005
robe-0.005
bridge-0.005
brush-0.005
coat-0.005
Token navy
Token ID23956
Feature activation+4.81e-01

Pos logit contributions
coat+0.031
robe+0.030
brush+0.030
amas+0.029
stand+0.026
Neg logit contributions
 Enjoy-0.041
ours-0.040
 asks-0.038
 coconut-0.038
grown-0.037
Token sail
Token ID14936
Feature activation+1.63e+00

Pos logit contributions
aces+0.039
coat+0.038
robe+0.038
amas+0.037
brush+0.037
Neg logit contributions
 coconut-0.033
 Enjoy-0.030
ours-0.027
grown-0.026
iw-0.026
Tokenboat
Token ID24482
Feature activation+7.43e+00

Pos logit contributions
aces+0.144
amas+0.141
coat+0.137
brush+0.133
bridge+0.133
Neg logit contributions
 coconut-0.100
 Enjoy-0.089
 volcano-0.082
 temple-0.082
 whale-0.080
Token in
Token ID287
Feature activation-1.39e+00

Pos logit contributions
aces+0.679
coat+0.662
amas+0.658
robe+0.656
brush+0.641
Neg logit contributions
 coconut-0.430
 Enjoy-0.386
 temple-0.340
grown-0.338
 volcano-0.334
Token the
Token ID262
Feature activation+6.47e-02

Pos logit contributions
 coconut+0.080
 Enjoy+0.073
ours+0.067
oral+0.064
iw+0.063
Neg logit contributions
aces-0.123
coat-0.122
robe-0.122
brush-0.120
bridge-0.120
Token distance
Token ID5253
Feature activation+3.06e+00

Pos logit contributions
aces+0.007
coat+0.006
robe+0.006
amas+0.006
brush+0.006
Neg logit contributions
 coconut-0.003
 Enjoy-0.003
 temple-0.003
 volcano-0.002
 whale-0.002
Token.
Token ID13
Feature activation-8.02e-02

Pos logit contributions
aces+0.277
coat+0.272
robe+0.270
amas+0.266
brush+0.262
Neg logit contributions
 coconut-0.181
 Enjoy-0.164
ours-0.143
grown-0.140
 temple-0.140
Token They
Token ID1119
Feature activation-1.25e+00

Pos logit contributions
 coconut+0.005
 Enjoy+0.004
 temple+0.004
grown+0.004
ours+0.004
Neg logit contributions
aces-0.007
coat-0.007
robe-0.007
amas-0.007
brush-0.007
Token
Token ID198
Feature activation+1.21e-01

Pos logit contributions
robe+0.010
amas+0.010
coat+0.010
aces+0.010
bridge+0.010
Neg logit contributions
 coconut-0.008
 Enjoy-0.006
ours-0.006
iw-0.006
 oy-0.006
TokenThe
Token ID464
Feature activation+1.11e+00

Pos logit contributions
aces+0.009
coat+0.009
robe+0.009
amas+0.009
brush+0.009
Neg logit contributions
 coconut-0.009
 Enjoy-0.008
ours-0.007
grown-0.007
 temple-0.007
Token little
Token ID1310
Feature activation-8.56e-01

Pos logit contributions
aces+0.087
coat+0.086
robe+0.085
amas+0.083
bridge+0.082
Neg logit contributions
 coconut-0.078
 Enjoy-0.075
ours-0.067
iw-0.065
grown-0.065
Token girl
Token ID2576
Feature activation-1.03e-01

Pos logit contributions
 coconut+0.054
 Enjoy+0.049
 temple+0.043
ours+0.043
grown+0.043
Neg logit contributions
aces-0.074
coat-0.072
robe-0.072
amas-0.072
bridge-0.070
Token was
Token ID373
Feature activation-3.94e-01

Pos logit contributions
 coconut+0.007
 Enjoy+0.006
ours+0.005
grown+0.005
iw+0.005
Neg logit contributions
aces-0.009
coat-0.009
robe-0.009
amas-0.008
brush-0.008
Token full
Token ID1336
Feature activation+9.37e-01

Pos logit contributions
 coconut+0.022
 Enjoy+0.020
 temple+0.018
grown+0.018
 swamp+0.018
Neg logit contributions
aces-0.037
amas-0.036
coat-0.035
robe-0.035
bridge-0.034
Token of
Token ID286
Feature activation+5.30e-01

Pos logit contributions
aces+0.090
coat+0.088
robe+0.088
amas+0.087
bridge+0.085
Neg logit contributions
 coconut-0.050
 Enjoy-0.045
 temple-0.039
grown-0.039
 oy-0.039
Token energy
Token ID2568
Feature activation+5.25e-01

Pos logit contributions
aces+0.038
coat+0.037
robe+0.036
brush+0.035
bridge+0.035
Neg logit contributions
 coconut-0.040
 Enjoy-0.037
ours-0.035
oral-0.034
iw-0.034
Token and
Token ID290
Feature activation-8.09e-02

Pos logit contributions
aces+0.051
amas+0.050
robe+0.050
bridge+0.050
coat+0.049
Neg logit contributions
 coconut-0.026
 Enjoy-0.022
 temple-0.021
grown-0.021
 volcano-0.020
Token happy
Token ID3772
Feature activation+1.16e+00

Pos logit contributions
 coconut+0.004
 Enjoy+0.004
 temple+0.004
grown+0.004
 whose+0.004
Neg logit contributions
aces-0.008
amas-0.008
coat-0.007
robe-0.007
bridge-0.007
Token to
Token ID284
Feature activation-5.49e-01

Pos logit contributions
aces+0.101
robe+0.099
coat+0.099
amas+0.097
brush+0.095
Neg logit contributions
 coconut-0.072
 Enjoy-0.066
ours-0.058
iw-0.056
oral-0.056
Token.
Token ID13
Feature activation-7.17e-02

Pos logit contributions
 coconut+0.176
 Enjoy+0.165
ours+0.144
oral+0.144
iw+0.139
Neg logit contributions
aces-0.238
coat-0.232
robe-0.231
amas-0.224
brush-0.222
Token He
Token ID679
Feature activation+3.77e-01

Pos logit contributions
 coconut+0.005
oral+0.005
 Enjoy+0.005
 temple+0.005
grown+0.005
Neg logit contributions
aces-0.005
amas-0.005
coat-0.005
robe-0.005
drops-0.004
Token thought
Token ID1807
Feature activation-2.40e+00

Pos logit contributions
coat+0.028
aces+0.027
brush+0.027
robe+0.025
amed+0.025
Neg logit contributions
 Enjoy-0.028
 coconut-0.026
oral-0.024
 temple-0.024
grown-0.023
Token that
Token ID326
Feature activation+7.99e-01

Pos logit contributions
 coconut+0.154
 Enjoy+0.149
iw+0.132
oral+0.128
grown+0.123
Neg logit contributions
aces-0.208
coat-0.191
brush-0.190
robe-0.189
bridge-0.188
Token it
Token ID340
Feature activation-1.26e+00

Pos logit contributions
aces+0.065
robe+0.062
brush+0.061
coat+0.059
bridge+0.059
Neg logit contributions
 Enjoy-0.047
 coconut-0.047
oral-0.042
grown-0.042
ours-0.041
Token would
Token ID561
Feature activation+1.35e+00

Pos logit contributions
 coconut+0.068
 Enjoy+0.061
ours+0.053
grown+0.052
 temple+0.052
Neg logit contributions
aces-0.120
coat-0.118
robe-0.117
amas-0.116
bridge-0.114
Token be
Token ID307
Feature activation+7.75e-01

Pos logit contributions
aces+0.101
coat+0.099
robe+0.098
amas+0.095
brush+0.095
Neg logit contributions
 coconut-0.100
 Enjoy-0.093
ours-0.085
grown-0.084
 temple-0.082
Token fun
Token ID1257
Feature activation+1.50e+00

Pos logit contributions
aces+0.064
coat+0.063
robe+0.062
amas+0.061
brush+0.060
Neg logit contributions
 coconut-0.052
 Enjoy-0.048
ours-0.042
grown-0.042
 temple-0.042
Token to
Token ID284
Feature activation-5.64e-01

Pos logit contributions
aces+0.158
amas+0.156
robe+0.156
coat+0.155
bridge+0.152
Neg logit contributions
 coconut-0.065
 Enjoy-0.056
grown-0.047
 temple-0.046
ours-0.045
Token carry
Token ID3283
Feature activation+1.97e+00

Pos logit contributions
 coconut+0.040
oral+0.038
grown+0.038
ours+0.037
 Enjoy+0.037
Neg logit contributions
coat-0.043
brush-0.041
aces-0.039
robe-0.037
drops-0.037
Token them
Token ID606
Feature activation+1.29e+00

Pos logit contributions
aces+0.173
robe+0.165
coat+0.165
brush+0.159
bridge+0.157
Neg logit contributions
 coconut-0.122
 Enjoy-0.114
oral-0.103
ours-0.102
 temple-0.098
Token
Token ID198
Feature activation+3.26e-02

Pos logit contributions
 coconut+0.009
 Enjoy+0.007
 temple+0.007
 exhibit+0.007
 volcano+0.007
Neg logit contributions
aces-0.013
robe-0.013
coat-0.013
amas-0.013
brush-0.012
Token
Token ID198
Feature activation+3.50e-02

Pos logit contributions
aces+0.003
amas+0.003
robe+0.002
stand+0.002
coat+0.002
Neg logit contributions
 coconut-0.002
 Enjoy-0.002
 whale-0.002
 volcano-0.002
 crab-0.002
Token"
Token ID1
Feature activation+1.72e+00

Pos logit contributions
aces+0.003
coat+0.003
robe+0.003
amas+0.003
brush+0.003
Neg logit contributions
 coconut-0.003
 Enjoy-0.002
 volcano-0.002
ours-0.002
 temple-0.002
TokenIt
Token ID1026
Feature activation-4.40e-01

Pos logit contributions
aces+0.147
coat+0.144
robe+0.143
amas+0.141
brush+0.139
Neg logit contributions
 coconut-0.111
 Enjoy-0.102
ours-0.090
 temple-0.088
grown-0.088
Token's
Token ID338
Feature activation+1.41e+00

Pos logit contributions
 coconut+0.042
 Enjoy+0.039
grown+0.037
 temple+0.036
oral+0.036
Neg logit contributions
aces-0.023
robe-0.023
coat-0.023
amas-0.023
brush-0.021
Token okay
Token ID8788
Feature activation+2.71e+00

Pos logit contributions
aces+0.069
coat+0.067
amas+0.065
robe+0.065
brush+0.063
Neg logit contributions
 coconut-0.142
 Enjoy-0.133
ours-0.124
 temple-0.124
 volcano-0.124
Token,
Token ID11
Feature activation-5.93e-03

Pos logit contributions
coat+0.196
aces+0.192
amas+0.191
brush+0.185
robe+0.185
Neg logit contributions
 coconut-0.210
 Enjoy-0.190
iw-0.169
ours-0.167
 aquarium-0.167
Token my
Token ID616
Feature activation+1.63e-01

Pos logit contributions
 coconut+0.000
 Enjoy+0.000
 shrimp+0.000
ours+0.000
 crab+0.000
Neg logit contributions
aces-0.001
robe-0.001
coat-0.001
amas-0.001
bridge-0.001
Token loves
Token ID10408
Feature activation-1.32e+00

Pos logit contributions
coat+0.012
robe+0.012
aces+0.012
amas+0.012
brush+0.011
Neg logit contributions
 coconut-0.012
 Enjoy-0.011
 Crab-0.010
grown-0.010
rill-0.010
Token.
Token ID13
Feature activation-1.38e-01

Pos logit contributions
 coconut+0.085
 Enjoy+0.077
ours+0.068
 volcano+0.066
 temple+0.066
Neg logit contributions
aces-0.115
coat-0.112
robe-0.110
amas-0.109
brush-0.108
Token I
Token ID314
Feature activation+4.43e-01

Pos logit contributions
 coconut+0.010
 Enjoy+0.008
 temple+0.008
oral+0.008
grown+0.008
Neg logit contributions
aces-0.011
robe-0.011
coat-0.011
amas-0.011
brush-0.010
Token work
Token ID670
Feature activation-5.38e-01

Pos logit contributions
 coconut+0.014
 Enjoy+0.013
 temple+0.011
oral+0.011
 volcano+0.011
Neg logit contributions
aces-0.017
coat-0.017
robe-0.017
amas-0.017
bridge-0.016
Token together
Token ID1978
Feature activation-1.29e+00

Pos logit contributions
 coconut+0.025
 Enjoy+0.025
grown+0.021
ours+0.020
oral+0.020
Neg logit contributions
coat-0.052
aces-0.052
robe-0.051
amas-0.051
bridge-0.051
Token.
Token ID13
Feature activation-1.09e-01

Pos logit contributions
 coconut+0.074
 Enjoy+0.067
ours+0.058
 temple+0.058
grown+0.058
Neg logit contributions
aces-0.116
coat-0.115
robe-0.114
amas-0.113
bridge-0.111
Token They
Token ID1119
Feature activation-1.29e+00

Pos logit contributions
 coconut+0.008
 temple+0.006
grown+0.006
 crab+0.006
 volcano+0.006
Neg logit contributions
aces-0.009
amas-0.009
robe-0.009
coat-0.009
brush-0.008
Token can
Token ID460
Feature activation-2.02e-01

Pos logit contributions
 coconut+0.070
 Enjoy+0.066
 temple+0.056
ours+0.055
 volcano+0.055
Neg logit contributions
aces-0.121
coat-0.119
robe-0.119
amas-0.119
bridge-0.116
Token help
Token ID1037
Feature activation+1.82e+00

Pos logit contributions
 coconut+0.014
 Enjoy+0.013
ours+0.012
 temple+0.012
grown+0.012
Neg logit contributions
aces-0.016
coat-0.015
robe-0.015
amas-0.015
bridge-0.015
Token each
Token ID1123
Feature activation-2.33e-01

Pos logit contributions
aces+0.171
coat+0.167
robe+0.167
amas+0.164
bridge+0.162
Neg logit contributions
 coconut-0.101
 Enjoy-0.092
ours-0.079
 temple-0.078
grown-0.078
Token other
Token ID584
Feature activation-4.84e+00

Pos logit contributions
 coconut+0.015
 Enjoy+0.014
ours+0.013
iw+0.012
grown+0.012
Neg logit contributions
aces-0.019
robe-0.019
bridge-0.019
brush-0.018
coat-0.018
Token and
Token ID290
Feature activation-1.40e-01

Pos logit contributions
 coconut+0.239
 Enjoy+0.218
 temple+0.201
 volcano+0.197
 whose+0.195
Neg logit contributions
coat-0.466
aces-0.458
robe-0.455
amas-0.446
bridge-0.438
Token the
Token ID262
Feature activation+4.04e-02

Pos logit contributions
 coconut+0.009
 Enjoy+0.008
ours+0.008
grown+0.007
iw+0.007
Neg logit contributions
aces-0.012
coat-0.011
robe-0.011
brush-0.011
amas-0.011
Token people
Token ID661
Feature activation+1.90e+00

Pos logit contributions
aces+0.003
robe+0.003
coat+0.003
amas+0.003
brush+0.003
Neg logit contributions
 coconut-0.003
 Enjoy-0.003
ours-0.002
grown-0.002
iw-0.002
Token the
Token ID262
Feature activation+8.13e-02

Pos logit contributions
 coconut+0.087
 Enjoy+0.079
ours+0.073
oral+0.070
iw+0.068
Neg logit contributions
coat-0.115
aces-0.114
robe-0.114
amas-0.109
brush-0.109
Token park
Token ID3952
Feature activation-2.16e+00

Pos logit contributions
coat+0.007
robe+0.007
amas+0.007
brush+0.007
aces+0.007
Neg logit contributions
 coconut-0.005
 Enjoy-0.004
ours-0.004
oral-0.004
iw-0.004
Token and
Token ID290
Feature activation-1.10e-01

Pos logit contributions
 coconut+0.165
 Enjoy+0.152
ours+0.139
oral+0.138
iw+0.135
Neg logit contributions
aces-0.157
coat-0.156
robe-0.153
amas-0.148
bridge-0.145
Token she
Token ID673
Feature activation+1.50e+00

Pos logit contributions
 coconut+0.007
 Enjoy+0.006
oral+0.005
ours+0.005
iw+0.005
Neg logit contributions
aces-0.010
coat-0.010
robe-0.010
amas-0.009
bridge-0.009
Token heard
Token ID2982
Feature activation-1.71e+00

Pos logit contributions
aces+0.160
coat+0.157
robe+0.156
amas+0.155
brush+0.153
Neg logit contributions
 coconut-0.064
 Enjoy-0.056
ours-0.045
grown-0.045
 temple-0.045
Token a
Token ID257
Feature activation+5.28e-01

Pos logit contributions
 coconut+0.113
 Enjoy+0.104
iw+0.094
ours+0.094
oral+0.091
Neg logit contributions
aces-0.143
robe-0.138
coat-0.138
bridge-0.136
brush-0.135
Token strange
Token ID6283
Feature activation-1.24e+00

Pos logit contributions
coat+0.028
robe+0.028
amas+0.027
brush+0.027
bridge+0.026
Neg logit contributions
 coconut-0.047
 Enjoy-0.046
ours-0.044
iw-0.042
grown-0.042
Token language
Token ID3303
Feature activation-1.94e+00

Pos logit contributions
 coconut+0.125
 Enjoy+0.122
ours+0.113
iw+0.112
grown+0.112
Neg logit contributions
robe-0.056
coat-0.055
aces-0.054
amas-0.054
brush-0.051
Token.
Token ID13
Feature activation-6.46e-02

Pos logit contributions
 coconut+0.116
 Enjoy+0.105
ours+0.091
grown+0.090
 temple+0.089
Neg logit contributions
aces-0.175
coat-0.171
robe-0.170
amas-0.168
brush-0.165
Token She
Token ID1375
Feature activation+3.25e-01

Pos logit contributions
 coconut+0.004
 Enjoy+0.004
ours+0.004
grown+0.004
 temple+0.004
Neg logit contributions
aces-0.005
coat-0.005
robe-0.005
amas-0.005
brush-0.005
Token was
Token ID373
Feature activation-4.01e-01

Pos logit contributions
aces+0.023
coat+0.023
robe+0.022
brush+0.022
bridge+0.022
Neg logit contributions
 Enjoy-0.024
 coconut-0.024
iw-0.022
ours-0.021
oral-0.020
Token mom
Token ID1995
Feature activation+6.87e-01

Pos logit contributions
aces+0.008
coat+0.008
amas+0.007
robe+0.007
bridge+0.007
Neg logit contributions
 coconut-0.006
 Enjoy-0.005
 temple-0.005
 volcano-0.005
oral-0.005
Tokenmy
Token ID1820
Feature activation+9.41e-01

Pos logit contributions
aces+0.066
robe+0.063
amas+0.062
bridge+0.061
coat+0.060
Neg logit contributions
 coconut-0.035
grown-0.031
 whose-0.031
 Enjoy-0.029
ours-0.028
Token disappeared
Token ID12120
Feature activation+1.06e+00

Pos logit contributions
aces+0.096
robe+0.092
amas+0.091
coat+0.089
bridge+0.088
Neg logit contributions
 coconut-0.044
grown-0.036
 whose-0.036
 whale-0.034
 crab-0.034
Token again
Token ID757
Feature activation-3.84e+00

Pos logit contributions
aces+0.109
coat+0.104
amas+0.102
brush+0.102
bridge+0.102
Neg logit contributions
 coconut-0.046
grown-0.044
 temple-0.038
 oy-0.037
 volcano-0.035
Token.
Token ID13
Feature activation-5.14e-02

Pos logit contributions
 coconut+0.231
 Enjoy+0.209
ours+0.183
grown+0.180
 temple+0.180
Neg logit contributions
aces-0.343
coat-0.336
robe-0.334
amas-0.329
brush-0.324
Token The
Token ID383
Feature activation-5.86e-01

Pos logit contributions
 coconut+0.004
 Enjoy+0.003
 whale+0.003
iw+0.003
 crab+0.003
Neg logit contributions
amas-0.004
robe-0.004
aces-0.004
coat-0.004
brush-0.004
Token girl
Token ID2576
Feature activation-1.03e-01

Pos logit contributions
 coconut+0.043
 Enjoy+0.041
ours+0.036
iw+0.036
grown+0.035
Neg logit contributions
robe-0.044
coat-0.044
aces-0.043
amas-0.042
brush-0.042
Token felt
Token ID2936
Feature activation+1.71e+00

Pos logit contributions
 coconut+0.006
 Enjoy+0.005
ours+0.004
grown+0.004
oral+0.004
Neg logit contributions
aces-0.010
coat-0.010
robe-0.009
amas-0.009
brush-0.009
Token sad
Token ID6507
Feature activation-5.77e-01

Pos logit contributions
aces+0.152
coat+0.151
amas+0.150
robe+0.145
bridge+0.141
Neg logit contributions
 coconut-0.106
 Enjoy-0.100
grown-0.087
 shrimp-0.086
 oy-0.085
Token and
Token ID290
Feature activation-9.42e-02

Pos logit contributions
 coconut+0.033
 Enjoy+0.030
grown+0.027
ours+0.026
 temple+0.026
Neg logit contributions
aces-0.053
coat-0.052
robe-0.051
amas-0.051
bridge-0.050
Token wanted
Token ID2227
Feature activation-1.53e+00

Pos logit contributions
 coconut+0.006
 Enjoy+0.006
ours+0.005
oral+0.005
grown+0.005
Neg logit contributions
aces-0.008
coat-0.008
robe-0.008
amas-0.007
brush-0.007
Token mom
Token ID1995
Feature activation+6.12e-01

Pos logit contributions
aces+0.118
robe+0.117
coat+0.116
amas+0.116
brush+0.114
Neg logit contributions
 coconut-0.088
 Enjoy-0.083
ours-0.073
iw-0.073
grown-0.073
Token made
Token ID925
Feature activation+4.63e+00

Pos logit contributions
aces+0.060
robe+0.058
coat+0.058
amas+0.057
bridge+0.056
Neg logit contributions
 coconut-0.032
 Enjoy-0.029
ours-0.025
grown-0.025
 temple-0.024
Token him
Token ID683
Feature activation+1.33e+00

Pos logit contributions
aces+0.392
amas+0.390
coat+0.383
robe+0.381
brush+0.373
Neg logit contributions
 coconut-0.294
 Enjoy-0.262
oral-0.236
iw-0.236
 temple-0.235
Token a
Token ID257
Feature activation+4.51e-01

Pos logit contributions
aces+0.117
coat+0.114
robe+0.114
amas+0.113
bridge+0.110
Neg logit contributions
 coconut-0.081
 Enjoy-0.074
ours-0.064
 temple-0.064
grown-0.063
Token healthy
Token ID5448
Feature activation+1.07e+00

Pos logit contributions
coat+0.030
brush+0.029
robe+0.028
amas+0.028
 forehead+0.025
Neg logit contributions
 Enjoy-0.033
 coconut-0.032
grown-0.032
ours-0.032
 asks-0.030
Token dinner
Token ID8073
Feature activation-6.45e-01

Pos logit contributions
coat+0.088
aces+0.086
amas+0.086
robe+0.086
brush+0.084
Neg logit contributions
 coconut-0.069
 Enjoy-0.064
ours-0.059
grown-0.058
oral-0.057
Token and
Token ID290
Feature activation-1.78e-01

Pos logit contributions
 coconut+0.024
 Enjoy+0.022
 temple+0.021
 restaurant+0.020
 volcano+0.019
Neg logit contributions
aces-0.068
coat-0.065
amas-0.065
bridge-0.064
robe-0.064
Token they
Token ID484
Feature activation-4.18e-01

Pos logit contributions
 coconut+0.011
 Enjoy+0.010
oral+0.009
ours+0.009
grown+0.009
Neg logit contributions
robe-0.015
coat-0.015
aces-0.015
brush-0.014
amas-0.014
Token talked
Token ID6619
Feature activation-2.29e+00

Pos logit contributions
 coconut+0.029
 Enjoy+0.027
ours+0.025
oral+0.024
iw+0.024
Neg logit contributions
aces-0.032
robe-0.032
brush-0.031
coat-0.031
bridge-0.031
Token about
Token ID546
Feature activation-1.15e+00

Pos logit contributions
 coconut+0.111
 Enjoy+0.101
grown+0.084
 volcano+0.083
 temple+0.082
Neg logit contributions
aces-0.229
coat-0.226
robe-0.223
amas-0.221
bridge-0.217
Token how
Token ID703
Feature activation-3.05e-01

Pos logit contributions
 coconut+0.074
oral+0.071
ours+0.068
iw+0.068
 Enjoy+0.066
Neg logit contributions
aces-0.080
amas-0.079
robe-0.077
bridge-0.076
brush-0.076
Token it
Token ID340
Feature activation-1.33e+00

Pos logit contributions
aces+0.101
brush+0.098
robe+0.097
coat+0.095
bridge+0.093
Neg logit contributions
 coconut-0.067
 Enjoy-0.061
iw-0.057
ours-0.053
oral-0.052
Token was
Token ID373
Feature activation-4.85e-01

Pos logit contributions
 coconut+0.065
 Enjoy+0.057
ours+0.048
 temple+0.048
grown+0.048
Neg logit contributions
aces-0.133
coat-0.130
robe-0.129
amas-0.128
bridge-0.126
Token a
Token ID257
Feature activation+4.56e-01

Pos logit contributions
 coconut+0.032
 Enjoy+0.029
ours+0.026
iw+0.025
oral+0.025
Neg logit contributions
aces-0.041
coat-0.040
robe-0.040
amas-0.039
brush-0.039
Token rainbow
Token ID27223
Feature activation-1.87e+00

Pos logit contributions
robe+0.030
coat+0.030
brush+0.029
bridge+0.027
 cartridge+0.026
Neg logit contributions
ours-0.034
 coconut-0.033
 Enjoy-0.033
grown-0.031
 asks-0.031
Token and
Token ID290
Feature activation-1.85e-01

Pos logit contributions
 coconut+0.093
 temple+0.081
 volcano+0.077
 whale+0.075
 Enjoy+0.075
Neg logit contributions
amas-0.183
aces-0.182
robe-0.174
coat-0.171
bridge-0.168
Token it
Token ID340
Feature activation-1.32e+00

Pos logit contributions
 coconut+0.011
 Enjoy+0.010
oral+0.010
iw+0.009
 volcano+0.009
Neg logit contributions
aces-0.016
brush-0.016
robe-0.015
coat-0.015
bridge-0.015
Token reminded
Token ID14516
Feature activation+1.99e+00

Pos logit contributions
 coconut+0.066
 Enjoy+0.055
 temple+0.053
 volcano+0.051
ours+0.051
Neg logit contributions
aces-0.130
coat-0.129
amas-0.129
robe-0.127
bridge-0.123
Token her
Token ID607
Feature activation+1.13e+00

Pos logit contributions
aces+0.159
coat+0.155
robe+0.154
amas+0.151
brush+0.149
Neg logit contributions
 coconut-0.139
 Enjoy-0.127
ours-0.114
 temple-0.112
 volcano-0.112
Token to
Token ID284
Feature activation-6.37e-01

Pos logit contributions
aces+0.107
robe+0.104
coat+0.104
brush+0.103
amas+0.100
Neg logit contributions
 coconut-0.063
 Enjoy-0.057
iw-0.049
ours-0.049
 volcano-0.048
Token be
Token ID307
Feature activation+7.03e-01

Pos logit contributions
 coconut+0.044
 Enjoy+0.040
ours+0.036
grown+0.035
 temple+0.035
Neg logit contributions
aces-0.051
coat-0.050
robe-0.050
amas-0.049
brush-0.048
Token happy
Token ID3772
Feature activation+1.07e+00

Pos logit contributions
amas+0.056
aces+0.055
robe+0.055
coat+0.054
bridge+0.052
Neg logit contributions
 coconut-0.049
 Enjoy-0.046
grown-0.043
 whose-0.041
 temple-0.040
Token faster
Token ID5443
Feature activation+3.13e+00

Pos logit contributions
amas+0.164
aces+0.157
bridge+0.157
robe+0.156
coat+0.154
Neg logit contributions
 coconut-0.092
 Enjoy-0.081
 exhibit-0.076
 temple-0.075
 whose-0.074
Token and
Token ID290
Feature activation-1.77e-01

Pos logit contributions
amas+0.297
coat+0.289
robe+0.289
aces+0.285
bridge+0.283
Neg logit contributions
 coconut-0.167
 Enjoy-0.140
grown-0.132
 temple-0.130
 crab-0.127
Token louder
Token ID27089
Feature activation-2.16e-01

Pos logit contributions
 coconut+0.013
 Enjoy+0.012
 asks+0.012
 stealing+0.011
 Crab+0.011
Neg logit contributions
amas-0.014
robe-0.013
aces-0.012
coat-0.012
bridge-0.012
Token.
Token ID13
Feature activation-1.38e-01

Pos logit contributions
 coconut+0.010
 Enjoy+0.009
 temple+0.008
grown+0.008
 whose+0.008
Neg logit contributions
coat-0.022
robe-0.022
amas-0.021
aces-0.021
brush-0.021
Token He
Token ID679
Feature activation+2.97e-01

Pos logit contributions
 coconut+0.009
 Enjoy+0.008
ours+0.007
 temple+0.007
grown+0.007
Neg logit contributions
aces-0.012
robe-0.012
coat-0.011
amas-0.011
bridge-0.011
Token went
Token ID1816
Feature activation+4.09e-01

Pos logit contributions
aces+0.025
coat+0.024
robe+0.024
amas+0.024
brush+0.023
Neg logit contributions
 coconut-0.019
 Enjoy-0.018
ours-0.016
 temple-0.015
grown-0.015
Token towards
Token ID3371
Feature activation-9.18e-01

Pos logit contributions
aces+0.030
coat+0.029
robe+0.029
amas+0.028
brush+0.028
Neg logit contributions
 coconut-0.031
 Enjoy-0.029
ours-0.026
grown-0.026
 temple-0.026
Token L
Token ID406
Feature activation+3.26e-01

Pos logit contributions
 coconut+0.057
 Enjoy+0.054
oral+0.049
ours+0.047
iw+0.046
Neg logit contributions
robe-0.077
aces-0.077
coat-0.074
bridge-0.072
brush-0.071
Tokenila
Token ID10102
Feature activation-2.79e+00

Pos logit contributions
aces+0.034
coat+0.033
robe+0.033
bridge+0.032
amas+0.032
Neg logit contributions
 coconut-0.016
 Enjoy-0.014
 volcano-0.011
iw-0.011
grown-0.011
Token.
Token ID13
Feature activation-1.46e-01

Pos logit contributions
 coconut+0.129
 Enjoy+0.113
 exhibit+0.112
 whose+0.110
 temple+0.105
Neg logit contributions
robe-0.269
amas-0.268
bridge-0.266
aces-0.264
coat-0.262
Token He
Token ID679
Feature activation+2.98e-01

Pos logit contributions
 coconut+0.009
 Enjoy+0.008
 temple+0.007
ours+0.007
 volcano+0.007
Neg logit contributions
aces-0.013
robe-0.012
coat-0.012
amas-0.012
brush-0.012
Token.
Token ID13
Feature activation-8.44e-02

Pos logit contributions
 coconut+0.074
 temple+0.061
 Enjoy+0.061
 volcano+0.061
 exhibit+0.060
Neg logit contributions
coat-0.201
aces-0.201
robe-0.196
amas-0.194
bridge-0.193
Token The
Token ID383
Feature activation-6.12e-01

Pos logit contributions
 coconut+0.006
 Enjoy+0.005
grown+0.005
iw+0.005
 temple+0.005
Neg logit contributions
amas-0.007
aces-0.007
robe-0.007
coat-0.006
brush-0.006
Token screen
Token ID3159
Feature activation-5.72e-01

Pos logit contributions
 coconut+0.044
 Enjoy+0.041
 volcano+0.037
 temple+0.037
ours+0.037
Neg logit contributions
aces-0.048
coat-0.047
robe-0.045
amas-0.045
brush-0.044
Token lit
Token ID6578
Feature activation+3.03e+00

Pos logit contributions
 coconut+0.030
 volcano+0.025
 temple+0.024
 Enjoy+0.024
ours+0.023
Neg logit contributions
coat-0.055
aces-0.055
robe-0.053
amas-0.053
bridge-0.050
Token up
Token ID510
Feature activation+1.77e-01

Pos logit contributions
aces+0.259
coat+0.253
robe+0.251
brush+0.243
amas+0.243
Neg logit contributions
 coconut-0.198
 Enjoy-0.176
 volcano-0.164
 temple-0.161
 whose-0.161
Token,
Token ID11
Feature activation+5.98e-02

Pos logit contributions
aces+0.015
coat+0.015
robe+0.015
amas+0.014
brush+0.014
Neg logit contributions
 coconut-0.011
 Enjoy-0.010
ours-0.009
grown-0.009
 temple-0.009
Token and
Token ID290
Feature activation-1.11e-01

Pos logit contributions
aces+0.006
robe+0.006
coat+0.006
amas+0.006
brush+0.006
Neg logit contributions
 coconut-0.003
 Enjoy-0.002
 whose-0.002
 volcano-0.002
 temple-0.002
Token his
Token ID465
Feature activation-6.38e-01

Pos logit contributions
 coconut+0.005
 Enjoy+0.004
oral+0.003
ours+0.003
grown+0.003
Neg logit contributions
aces-0.012
coat-0.012
robe-0.012
amas-0.011
brush-0.011
Token show
Token ID905
Feature activation-1.47e+00

Pos logit contributions
 coconut+0.042
 Enjoy+0.040
iw+0.035
ours+0.035
grown+0.034
Neg logit contributions
aces-0.053
robe-0.053
coat-0.052
amas-0.051
brush-0.050
Token began
Token ID2540
Feature activation-7.17e-02

Pos logit contributions
 coconut+0.088
 Enjoy+0.079
ours+0.070
 temple+0.070
grown+0.069
Neg logit contributions
aces-0.129
coat-0.128
amas-0.127
robe-0.124
brush-0.121
Token to
Token ID284
Feature activation-5.79e-01

Pos logit contributions
 coconut+0.003
 Enjoy+0.003
ours+0.002
grown+0.002
 temple+0.002
Neg logit contributions
aces-0.007
coat-0.007
robe-0.007
amas-0.007
brush-0.007
Token he
Token ID339
Feature activation+6.34e-01

Pos logit contributions
aces+0.197
coat+0.193
robe+0.192
brush+0.190
amas+0.186
Neg logit contributions
 coconut-0.138
 Enjoy-0.126
oral-0.115
ours-0.112
grown-0.112
Token knew
Token ID2993
Feature activation-2.63e+00

Pos logit contributions
aces+0.052
coat+0.050
robe+0.050
brush+0.049
bridge+0.048
Neg logit contributions
 coconut-0.042
 Enjoy-0.039
grown-0.035
ours-0.035
oral-0.035
Token he
Token ID339
Feature activation+6.04e-01

Pos logit contributions
 coconut+0.175
 Enjoy+0.162
iw+0.148
oral+0.148
ours+0.147
Neg logit contributions
aces-0.215
robe-0.206
brush-0.205
coat-0.202
bridge-0.197
Token shouldn
Token ID6584
Feature activation-3.32e+00

Pos logit contributions
aces+0.052
coat+0.051
robe+0.050
amas+0.049
brush+0.049
Neg logit contributions
 coconut-0.038
 Enjoy-0.035
ours-0.031
iw-0.030
oral-0.030
Token't
Token ID470
Feature activation+1.36e+00

Pos logit contributions
 coconut+0.264
 Enjoy+0.246
ours+0.224
grown+0.221
 temple+0.220
Neg logit contributions
aces-0.233
coat-0.226
robe-0.225
amas-0.221
brush-0.216
Token touch
Token ID3638
Feature activation-5.22e+00

Pos logit contributions
aces+0.100
brush+0.098
coat+0.097
robe+0.094
amas+0.092
Neg logit contributions
 coconut-0.097
ours-0.095
 Enjoy-0.093
iw-0.088
grown-0.088
Token things
Token ID1243
Feature activation+8.71e-02

Pos logit contributions
 coconut+0.297
 Enjoy+0.283
oral+0.257
ours+0.238
iw+0.232
Neg logit contributions
aces-0.483
coat-0.460
robe-0.459
brush-0.451
amas-0.450
Token that
Token ID326
Feature activation+8.42e-01

Pos logit contributions
amas+0.009
coat+0.009
aces+0.009
robe+0.009
bridge+0.009
Neg logit contributions
 coconut-0.004
 Enjoy-0.003
 shrimp-0.003
 whose-0.003
 oy-0.003
Token aren
Token ID3588
Feature activation+2.49e+00

Pos logit contributions
amas+0.079
aces+0.078
coat+0.077
bridge+0.076
robe+0.075
Neg logit contributions
 coconut-0.047
 Enjoy-0.041
 temple-0.037
 volcano-0.037
 oy-0.036
Token't
Token ID470
Feature activation+1.34e+00

Pos logit contributions
coat+0.211
robe+0.201
aces+0.199
amas+0.194
brush+0.190
Neg logit contributions
 coconut-0.160
 Enjoy-0.148
oral-0.136
 volcano-0.132
ours-0.131
Token his
Token ID465
Feature activation-6.19e-01

Pos logit contributions
amas+0.124
robe+0.124
aces+0.123
coat+0.122
brush+0.118
Neg logit contributions
 coconut-0.078
 Enjoy-0.068
 oy-0.061
 shrimp-0.061
 temple-0.060
TokenThe
Token ID464
Feature activation+1.03e+00

Pos logit contributions
coat+0.003
aces+0.003
brush+0.003
bridge+0.003
amas+0.003
Neg logit contributions
 coconut-0.004
 Enjoy-0.004
ours-0.004
 volcano-0.004
 oy-0.004
Token little
Token ID1310
Feature activation-9.30e-01

Pos logit contributions
robe+0.071
coat+0.069
brush+0.068
aces+0.066
amas+0.063
Neg logit contributions
 coconut-0.081
 Enjoy-0.080
ours-0.075
grown-0.072
iw-0.071
Token girl
Token ID2576
Feature activation-1.87e-01

Pos logit contributions
 coconut+0.070
 Enjoy+0.066
ours+0.059
oral+0.058
 volcano+0.057
Neg logit contributions
aces-0.069
coat-0.069
robe-0.068
brush-0.066
amas-0.064
Token saw
Token ID2497
Feature activation-9.68e-02

Pos logit contributions
 coconut+0.012
 Enjoy+0.011
ours+0.010
iw+0.010
 volcano+0.010
Neg logit contributions
aces-0.016
coat-0.016
brush-0.015
robe-0.015
amas-0.015
Token a
Token ID257
Feature activation+4.56e-01

Pos logit contributions
 coconut+0.006
 Enjoy+0.006
oral+0.005
ours+0.005
iw+0.005
Neg logit contributions
aces-0.008
robe-0.008
coat-0.008
brush-0.008
amas-0.008
Token pe
Token ID613
Feature activation-6.43e+00

Pos logit contributions
coat+0.033
brush+0.032
robe+0.031
amas+0.030
bridge+0.028
Neg logit contributions
 Enjoy-0.034
 coconut-0.032
ours-0.031
oral-0.029
 asks-0.029
Tokenb
Token ID65
Feature activation+3.26e+00

Pos logit contributions
 coconut+0.606
 Enjoy+0.580
ours+0.525
 volcano+0.522
oral+0.522
Neg logit contributions
aces-0.356
coat-0.340
robe-0.330
brush-0.327
bridge-0.325
Tokenble
Token ID903
Feature activation-4.15e-01

Pos logit contributions
coat+0.236
amas+0.234
aces+0.232
robe+0.229
bridge+0.217
Neg logit contributions
 coconut-0.251
 Enjoy-0.214
 temple-0.206
 crab-0.204
ours-0.203
Token.
Token ID13
Feature activation-1.33e-01

Pos logit contributions
 coconut+0.027
 Enjoy+0.025
ours+0.022
grown+0.022
 temple+0.022
Neg logit contributions
aces-0.035
coat-0.034
robe-0.034
amas-0.033
brush-0.032
Token She
Token ID1375
Feature activation+2.40e-01

Pos logit contributions
 coconut+0.011
 Enjoy+0.010
oral+0.010
 whale+0.010
 volcano+0.010
Neg logit contributions
aces-0.009
amas-0.008
robe-0.008
coat-0.008
brush-0.008
Token decided
Token ID3066
Feature activation-2.70e+00

Pos logit contributions
aces+0.018
brush+0.018
coat+0.018
amed+0.017
bridge+0.016
Neg logit contributions
 Enjoy-0.017
 coconut-0.016
ours-0.015
 volcano-0.015
iw-0.015
TokenM
Token ID44
Feature activation-1.50e+00

Pos logit contributions
aces+0.004
coat+0.003
robe+0.003
amas+0.003
bridge+0.003
Neg logit contributions
 coconut-0.004
 Enjoy-0.003
ours-0.003
 temple-0.003
grown-0.003
Tokenoral
Token ID6864
Feature activation-7.24e+00

Pos logit contributions
 coconut+0.050
 Enjoy+0.046
ours+0.035
 volcano+0.033
 whose+0.033
Neg logit contributions
aces-0.173
brush-0.169
coat-0.167
amas-0.166
robe-0.164
Token of
Token ID286
Feature activation+4.57e-01

Pos logit contributions
 coconut+0.492
 Enjoy+0.402
ours+0.393
iw+0.383
 crab+0.376
Neg logit contributions
aces-0.609
robe-0.596
brush-0.592
amas-0.563
coat-0.551
Token the
Token ID262
Feature activation+2.42e-02

Pos logit contributions
robe+0.037
aces+0.036
amas+0.035
brush+0.035
stand+0.035
Neg logit contributions
 coconut-0.029
 Enjoy-0.025
ours-0.025
iw-0.023
grown-0.023
Token story
Token ID1621
Feature activation-2.30e+00

Pos logit contributions
aces+0.002
robe+0.002
brush+0.002
amas+0.002
bridge+0.002
Neg logit contributions
 coconut-0.002
 Enjoy-0.001
ours-0.001
grown-0.001
iw-0.001
Token:
Token ID25
Feature activation-5.63e+00

Pos logit contributions
 coconut+0.101
 Enjoy+0.075
iw+0.072
 shrimp+0.069
grown+0.068
Neg logit contributions
robe-0.252
aces-0.249
amas-0.238
brush-0.235
coat-0.229
Token It
Token ID632
Feature activation+4.52e-01

Pos logit contributions
 coconut+0.387
grown+0.295
 temple+0.288
iw+0.288
 crab+0.286
Neg logit contributions
amas-0.445
aces-0.440
robe-0.437
brush-0.420
 cartridge-0.415
Token is
Token ID318
Feature activation-5.34e-01

Pos logit contributions
aces+0.035
robe+0.034
amas+0.034
coat+0.033
brush+0.033
Neg logit contributions
 coconut-0.033
 Enjoy-0.029
grown-0.027
iw-0.026
 temple-0.026
Token better
Token ID1365
Feature activation-2.43e+00

Pos logit contributions
 coconut+0.044
 exhibit+0.038
 temple+0.037
iw+0.037
 volcano+0.037
Neg logit contributions
aces-0.034
robe-0.033
brush-0.033
coat-0.032
 cartridge-0.031
Token to
Token ID284
Feature activation-6.18e-01

Pos logit contributions
 coconut+0.167
 Enjoy+0.141
ours+0.132
 volcano+0.132
 aquarium+0.130
Neg logit contributions
aces-0.190
robe-0.183
coat-0.182
brush-0.178
 cartridge-0.174
Token be
Token ID307
Feature activation+7.19e-01

Pos logit contributions
 coconut+0.045
 Enjoy+0.039
ours+0.038
grown+0.037
iw+0.037
Neg logit contributions
aces-0.047
coat-0.046
robe-0.046
brush-0.046
amas-0.045
Token the
Token ID262
Feature activation+5.73e-02

Pos logit contributions
 coconut+0.046
oral+0.044
iw+0.042
ours+0.041
 Enjoy+0.040
Neg logit contributions
coat-0.037
amas-0.035
robe-0.035
stand-0.035
stick-0.031
Token kitchen
Token ID9592
Feature activation-1.19e+00

Pos logit contributions
robe+0.005
coat+0.005
brush+0.005
amas+0.005
aces+0.005
Neg logit contributions
 coconut-0.004
 Enjoy-0.003
ours-0.003
grown-0.003
oral-0.003
Token to
Token ID284
Feature activation-5.81e-01

Pos logit contributions
 coconut+0.061
 Enjoy+0.055
 temple+0.048
ours+0.048
 volcano+0.047
Neg logit contributions
aces-0.117
robe-0.113
coat-0.113
bridge-0.112
amas-0.111
Token get
Token ID651
Feature activation-1.23e+00

Pos logit contributions
 coconut+0.046
oral+0.041
 Enjoy+0.040
grown+0.040
iw+0.040
Neg logit contributions
robe-0.038
coat-0.037
amas-0.037
brush-0.037
bridge-0.036
Token some
Token ID617
Feature activation-4.90e-01

Pos logit contributions
 coconut+0.091
 Enjoy+0.080
oral+0.077
 exhibit+0.075
iw+0.074
Neg logit contributions
aces-0.090
robe-0.090
coat-0.088
amas-0.088
brush-0.087
Token yogurt
Token ID32132
Feature activation-6.63e+00

Pos logit contributions
 coconut+0.033
 Enjoy+0.031
ours+0.027
grown+0.027
 temple+0.027
Neg logit contributions
aces-0.040
coat-0.039
robe-0.039
amas-0.038
brush-0.037
Token.
Token ID13
Feature activation-8.02e-02

Pos logit contributions
 coconut+0.264
 shrimp+0.207
 oy+0.200
 Enjoy+0.194
ours+0.191
Neg logit contributions
aces-0.695
coat-0.686
amas-0.683
bridge-0.682
robe-0.669
Token They
Token ID1119
Feature activation-1.25e+00

Pos logit contributions
 coconut+0.005
 temple+0.005
 oy+0.004
 Enjoy+0.004
grown+0.004
Neg logit contributions
amas-0.007
robe-0.006
aces-0.006
coat-0.006
brush-0.006
Token each
Token ID1123
Feature activation-2.05e-01

Pos logit contributions
 coconut+0.088
 Enjoy+0.082
oral+0.076
iw+0.072
grown+0.072
Neg logit contributions
aces-0.096
coat-0.096
robe-0.094
brush-0.092
amas-0.090
Token got
Token ID1392
Feature activation+7.27e-01

Pos logit contributions
 coconut+0.015
 Enjoy+0.013
 temple+0.012
oral+0.012
 volcano+0.012
Neg logit contributions
coat-0.016
aces-0.016
robe-0.016
brush-0.015
bridge-0.015
Token a
Token ID257
Feature activation+5.22e-01

Pos logit contributions
aces+0.055
robe+0.053
coat+0.053
amas+0.053
brush+0.052
Neg logit contributions
 coconut-0.053
 Enjoy-0.049
oral-0.044
 exhibit-0.043
 temple-0.043
Token pond
Token ID16723
Feature activation-2.66e+00

Pos logit contributions
coat+0.039
robe+0.038
brush+0.037
bridge+0.035
amas+0.035
Neg logit contributions
 coconut-0.034
 Enjoy-0.034
ours-0.032
grown-0.030
oral-0.030
Token before
Token ID878
Feature activation+7.68e-01

Pos logit contributions
 coconut+0.170
 Enjoy+0.156
ours+0.138
oral+0.136
grown+0.136
Neg logit contributions
aces-0.226
coat-0.222
robe-0.220
amas-0.216
brush-0.213
Token,
Token ID11
Feature activation+1.40e-01

Pos logit contributions
aces+0.062
coat+0.061
amas+0.060
robe+0.060
bridge+0.058
Neg logit contributions
 coconut-0.051
 Enjoy-0.046
ours-0.043
oral-0.042
iw-0.041
Token so
Token ID523
Feature activation+2.79e+00

Pos logit contributions
aces+0.011
coat+0.011
robe+0.011
amas+0.010
brush+0.010
Neg logit contributions
 coconut-0.010
 Enjoy-0.009
ours-0.008
grown-0.008
 temple-0.008
Token he
Token ID339
Feature activation+5.82e-01

Pos logit contributions
aces+0.233
robe+0.228
coat+0.227
amas+0.223
brush+0.223
Neg logit contributions
 coconut-0.175
 Enjoy-0.159
ours-0.147
oral-0.146
iw-0.141
Token pe
Token ID613
Feature activation-6.33e+00

Pos logit contributions
aces+0.046
coat+0.045
robe+0.045
brush+0.044
bridge+0.044
Neg logit contributions
 coconut-0.039
 Enjoy-0.037
oral-0.033
grown-0.033
iw-0.033
Tokenered
Token ID1068
Feature activation-2.39e+00

Pos logit contributions
 coconut+0.468
 Enjoy+0.454
 volcano+0.400
 museum+0.396
oral+0.395
Neg logit contributions
aces-0.480
bridge-0.452
coat-0.452
brush-0.444
robe-0.434
Token in
Token ID287
Feature activation-1.33e+00

Pos logit contributions
 coconut+0.197
 Enjoy+0.190
oral+0.180
ours+0.173
iw+0.171
Neg logit contributions
aces-0.155
coat-0.144
robe-0.143
bridge-0.142
brush-0.142
Token,
Token ID11
Feature activation+1.08e-01

Pos logit contributions
 coconut+0.086
 Enjoy+0.082
ours+0.082
oral+0.079
iw+0.076
Neg logit contributions
aces-0.099
brush-0.096
coat-0.095
robe-0.094
bridge-0.094
Token blinking
Token ID43196
Feature activation-1.03e+00

Pos logit contributions
aces+0.009
robe+0.009
coat+0.009
amas+0.009
bridge+0.009
Neg logit contributions
 coconut-0.007
 Enjoy-0.006
grown-0.006
 temple-0.006
 whose-0.006
Token in
Token ID287
Feature activation-1.35e+00

Pos logit contributions
 coconut+0.058
 Enjoy+0.048
grown+0.046
 oy+0.045
 shrimp+0.045
Neg logit contributions
coat-0.096
aces-0.095
amas-0.094
robe-0.094
brush-0.090