TOP ACTIVATIONS
MAX = 0.085

she should always trust her gut and <|endoftext|> One day ,
quarrel about who was bra ver . The monkey said he
should always listen to his gut and not do anything too
around and then saw Jill leaning against the laundry basket in
careful and listen to your gut ." Tom and
he saw a big staff leaning against a tree .
jumping over logs , t ipt oe ing around rocks ,
to come and see the operating theatre ! He was very
felt a little bit bra ver . She slowly stepped closer
Tom noticed a big cane leaning against a tree . "
go she felt much bra ver . And she never felt
it made her even bra ver . <|endoftext|> M olly was
saw a ladder that was leaning against a tree . He
The pig was leaning against a big tree when
Billy decided to take his leaning style to a new level
. Today , she was leaning against the wall in the
, they saw the cap leaning against a tree . Billy
stick the mustache on her upper lip with some tape .
higher , he felt bra ver and bra ver . Soon
M olly was a very careful girl . She loved to

MIN ACTIVATIONS
MIN = -0.079

" WHAT ARE YOU DO ING ?" Mom yells .
and bubbles and sounds and smells . Lily was scared .
. She liked building sand cast les and collecting seas hell
the beach and build sand cast les .
, laughed and built sand cast les . Soon , they
is soft and brown and smells good . One
. They would build sand cast les and play in the
the ocean and build sand cast les . When
the sand and building sand cast les .
She picks the flower and smells it . It has a
the beach and build sand cast les . But one day
on , they built sand cast les together every weekend .
. She would build sand cast les and collect pretty seas
play . They built sand cast les and played tag .
She liked to build sand cast les and collect seas hell
the beach and building sand cast les . One day ,
beach . They made sand cast les and ran through the
She liked to build sand cast les and collect seas hell
while , then built sand cast les with his mom .
bucket . She made sand cast les and sand pies .

INTERVAL 0.044 - 0.085
CONTAINS 0.764%

played in the lovely tent every time they came to visit
the fish said . He began to believe in himself and
, � � € � � Oh , wow . That
Tim my 's toys were all over the place because he
lower it . Jack was careful not to make it too

INTERVAL 0.003 - 0.044
CONTAINS 41.998%

in the long run . <|endoftext|> Once upon a time there
the kitchen . The bottle was full of a red sauce
says . She knows he is right . She does not
and Jack were so excited to try it out .
the man shouts . " This is not a toy ,

INTERVAL -0.038 - 0.003
CONTAINS 55.800%

Ben were sad . They wanted to play with Spot in
it into slices and shared it with each other . It
her backpack and they set off . When
trail again . K isa learnt her lesson -
smiled and hugged . Tom learned to be kind and respectful

INTERVAL -0.079 - -0.038
CONTAINS 1.438%

go on it , Sara ," Ben said . " It
spark ly . You will look adorable in them !" Lily
little better . " Next time , be more
to another . " Look , Ben , a
fast as they could to see who could reach the other
Token she
Token ID673
Feature activation+1.39e-02

Pos logit contributions
 outage+0.108
bars+0.106
 hill+0.105
 directions+0.102
 direction+0.101
Neg logit contributions
hire-0.075
 Princess-0.068
 Queen-0.067
ocket-0.061
 robber-0.060
Token should
Token ID815
Feature activation+2.05e-02

Pos logit contributions
bars+0.015
 outage+0.015
 hill+0.015
 directions+0.014
 direction+0.014
Neg logit contributions
hire-0.029
 Princess-0.027
 Queen-0.026
ocket-0.026
 robber-0.025
Token always
Token ID1464
Feature activation+1.76e-02

Pos logit contributions
bars+0.031
 outage+0.030
spe+0.029
 hill+0.029
 direction+0.029
Neg logit contributions
hire-0.036
 Princess-0.031
 Queen-0.031
 robber-0.030
ocket-0.029
Token trust
Token ID3774
Feature activation+1.59e-02

Pos logit contributions
bars+0.029
 outage+0.028
spe+0.028
 hill+0.026
ipt+0.025
Neg logit contributions
hire-0.028
 Princess-0.024
 Queen-0.024
 robber-0.024
ocket-0.023
Token her
Token ID607
Feature activation+5.01e-03

Pos logit contributions
bars+0.027
 directions+0.027
 direction+0.027
 outage+0.027
 hill+0.026
Neg logit contributions
hire-0.024
 Princess-0.020
ocket-0.020
 Queen-0.020
 robber-0.019
Token gut
Token ID12500
Feature activation+8.54e-02

Pos logit contributions
bars+0.009
 outage+0.009
 hill+0.008
 directions+0.008
 direction+0.008
Neg logit contributions
hire-0.007
 Queen-0.006
 Princess-0.006
ocket-0.006
 robber-0.006
Token and
Token ID290
Feature activation+1.45e-02

Pos logit contributions
bars+0.125
 hill+0.123
 outage+0.120
 kay+0.116
ipt+0.113
Neg logit contributions
hire-0.141
 Princess-0.132
 Queen-0.131
ocket-0.117
 robber-0.117
Token<|endoftext|>
Token ID50256
Feature activation+2.14e-02

Pos logit contributions
bars+0.023
 outage+0.022
 hill+0.022
spe+0.021
 angle+0.021
Neg logit contributions
hire-0.023
 Queen-0.021
 Princess-0.020
 robber-0.019
ocket-0.019
TokenOne
Token ID3198
Feature activation+4.02e-02

Pos logit contributions
 outage+0.032
bars+0.032
spe+0.030
 directions+0.029
 hill+0.029
Neg logit contributions
 Princess-0.035
hire-0.035
 Queen-0.033
 Prince-0.032
adies-0.032
Token day
Token ID1110
Feature activation-1.09e-02

Pos logit contributions
 outage+0.048
bars+0.048
 Mechan+0.042
bike+0.042
spe+0.041
Neg logit contributions
hire-0.074
 Princess-0.071
ily-0.070
adies-0.067
 mushrooms-0.067
Token,
Token ID11
Feature activation-1.33e-02

Pos logit contributions
hire+0.028
 Queen+0.025
ocket+0.025
 Princess+0.025
 robber+0.024
Neg logit contributions
 blowing-0.007
 kay-0.007
 direction-0.007
 hill-0.007
bars-0.007
Token quarrel
Token ID41998
Feature activation+1.14e-02

Pos logit contributions
bars+0.044
 outage+0.042
spe+0.038
ipt+0.037
ney+0.037
Neg logit contributions
hire-0.047
 Princess-0.044
 Queen-0.043
 robber-0.041
 mushrooms-0.040
Token about
Token ID546
Feature activation+3.82e-03

Pos logit contributions
 hill+0.018
 outage+0.018
bars+0.018
 kay+0.018
spe+0.017
Neg logit contributions
hire-0.015
 as-0.013
ily-0.013
 like-0.013
 Princess-0.013
Token who
Token ID508
Feature activation+4.13e-02

Pos logit contributions
bars+0.006
 outage+0.006
 hill+0.006
 directions+0.006
 direction+0.006
Neg logit contributions
hire-0.006
 Princess-0.005
 Queen-0.005
ocket-0.005
 robber-0.005
Token was
Token ID373
Feature activation+8.81e-03

Pos logit contributions
bars+0.049
 outage+0.048
 hill+0.048
 directions+0.045
 direction+0.045
Neg logit contributions
hire-0.085
 Princess-0.075
ocket-0.074
 Queen-0.074
adies-0.071
Token bra
Token ID8290
Feature activation-7.11e-03

Pos logit contributions
 outage+0.010
bars+0.010
 hill+0.010
 directions+0.010
 direction+0.010
Neg logit contributions
hire-0.018
 Princess-0.016
 Queen-0.016
ocket-0.016
 robber-0.015
Tokenver
Token ID332
Feature activation+8.36e-02

Pos logit contributions
hire+0.013
 Princess+0.011
 Queen+0.011
ocket+0.011
 robber+0.011
Neg logit contributions
 hill-0.010
 outage-0.010
bars-0.010
 directions-0.010
 direction-0.010
Token.
Token ID13
Feature activation-3.43e-03

Pos logit contributions
bars+0.131
 hill+0.124
 outage+0.124
 directions+0.120
 Mechan+0.118
Neg logit contributions
hire-0.127
 Queen-0.113
 Princess-0.110
 robber-0.108
ns-0.108
Token The
Token ID383
Feature activation+2.06e-02

Pos logit contributions
hire+0.008
ocket+0.007
vell+0.007
adies+0.007
 Queen+0.007
Neg logit contributions
 hill-0.003
bars-0.003
 direction-0.003
 directions-0.003
 outage-0.003
Token monkey
Token ID21657
Feature activation-4.61e-03

Pos logit contributions
 hill+0.029
 outage+0.028
bars+0.027
 directions+0.027
 direction+0.027
Neg logit contributions
hire-0.038
 Princess-0.033
ocket-0.033
 Queen-0.033
adies-0.032
Token said
Token ID531
Feature activation-2.34e-02

Pos logit contributions
hire+0.009
 Princess+0.008
 Queen+0.008
ocket+0.008
 robber+0.008
Neg logit contributions
bars-0.006
 outage-0.006
 hill-0.006
 directions-0.005
 direction-0.005
Token he
Token ID339
Feature activation+1.25e-02

Pos logit contributions
hire+0.054
 Princess+0.048
ocket+0.048
 Queen+0.048
adies+0.046
Neg logit contributions
 outage-0.021
bars-0.020
 hill-0.020
 directions-0.019
 blowing-0.018
Token should
Token ID815
Feature activation+2.52e-02

Pos logit contributions
bars+0.014
 outage+0.014
 hill+0.014
 directions+0.014
 direction+0.014
Neg logit contributions
hire-0.020
 Princess-0.018
 Queen-0.018
ocket-0.018
 robber-0.018
Token always
Token ID1464
Feature activation+2.16e-02

Pos logit contributions
bars+0.035
 outage+0.034
 direction+0.032
spe+0.032
 directions+0.032
Neg logit contributions
hire-0.046
 Princess-0.041
 Queen-0.040
 robber-0.040
ocket-0.038
Token listen
Token ID6004
Feature activation+3.01e-02

Pos logit contributions
bars+0.036
 outage+0.035
spe+0.034
 hill+0.033
ipt+0.032
Neg logit contributions
hire-0.033
 Princess-0.030
 Queen-0.029
 robber-0.029
ocket-0.028
Token to
Token ID284
Feature activation+1.36e-02

Pos logit contributions
bars+0.028
 hill+0.027
 outage+0.027
 direction+0.024
 directions+0.024
Neg logit contributions
hire-0.068
 Princess-0.062
 Queen-0.062
ocket-0.060
 robber-0.059
Token his
Token ID465
Feature activation+1.16e-02

Pos logit contributions
bars+0.018
 outage+0.018
 hill+0.017
 directions+0.017
 direction+0.017
Neg logit contributions
hire-0.026
 Princess-0.023
 Queen-0.023
ocket-0.022
 robber-0.022
Token gut
Token ID12500
Feature activation+8.28e-02

Pos logit contributions
bars+0.018
 outage+0.017
 hill+0.016
 blowing+0.016
spe+0.016
Neg logit contributions
hire-0.019
 Princess-0.018
 Queen-0.018
 robber-0.018
 Prince-0.017
Token and
Token ID290
Feature activation+1.29e-02

Pos logit contributions
bars+0.122
 outage+0.116
 hill+0.115
spe+0.108
 kay+0.108
Neg logit contributions
hire-0.133
 Princess-0.129
 Queen-0.124
 Prince-0.113
ocket-0.113
Token not
Token ID407
Feature activation+2.22e-02

Pos logit contributions
bars+0.020
 outage+0.019
 hill+0.019
spe+0.018
ney+0.018
Neg logit contributions
hire-0.021
 Queen-0.019
 Princess-0.019
 robber-0.018
ocket-0.017
Token do
Token ID466
Feature activation-1.09e-02

Pos logit contributions
bars+0.033
 outage+0.032
 directions+0.031
 hill+0.030
ipt+0.030
Neg logit contributions
hire-0.037
 robber-0.033
 Queen-0.032
 Princess-0.032
ocket-0.030
Token anything
Token ID1997
Feature activation-2.74e-02

Pos logit contributions
hire+0.016
 Princess+0.014
 Queen+0.014
ocket+0.013
 robber+0.013
Neg logit contributions
bars-0.019
 outage-0.019
 hill-0.019
 directions-0.018
 direction-0.018
Token too
Token ID1165
Feature activation-2.03e-04

Pos logit contributions
hire+0.050
 Princess+0.044
 Queen+0.043
ocket+0.042
 robber+0.042
Neg logit contributions
bars-0.038
 outage-0.038
 hill-0.037
 directions-0.036
 direction-0.036
Token around
Token ID1088
Feature activation+2.37e-02

Pos logit contributions
 outage+0.067
bars+0.063
 hill+0.061
bike+0.061
spe+0.061
Neg logit contributions
hire-0.064
 Princess-0.060
 Queen-0.056
 mushrooms-0.052
 robber-0.051
Token and
Token ID290
Feature activation-1.88e-03

Pos logit contributions
 hill+0.019
 directions+0.018
 direction+0.017
 blowing+0.016
 way+0.016
Neg logit contributions
hire-0.060
ocket-0.053
 Queen-0.051
adies-0.051
 robber-0.051
Token then
Token ID788
Feature activation+2.94e-05

Pos logit contributions
hire+0.004
adies+0.004
ocket+0.003
ocol+0.003
 robber+0.003
Neg logit contributions
 directions-0.002
 way-0.002
 direction-0.002
 was-0.002
 hill-0.002
Token saw
Token ID2497
Feature activation-1.34e-02

Pos logit contributions
 was+0.000
 directions+0.000
 way+0.000
 direction+0.000
 hill+0.000
Neg logit contributions
hire-0.000
ocket-0.000
adies-0.000
 robber-0.000
reath-0.000
Token Jill
Token ID22772
Feature activation-6.69e-03

Pos logit contributions
hire+0.028
pper+0.024
ocket+0.024
 robber+0.024
ns+0.024
Neg logit contributions
 directions-0.016
 direction-0.015
 was-0.014
 tube-0.014
 hill-0.014
Token leaning
Token ID21804
Feature activation+8.19e-02

Pos logit contributions
hire+0.013
 Princess+0.012
ocket+0.012
 Queen+0.012
 robber+0.011
Neg logit contributions
 outage-0.008
bars-0.008
 hill-0.008
 directions-0.008
 direction-0.007
Token against
Token ID1028
Feature activation+2.83e-02

Pos logit contributions
 outage+0.105
bars+0.102
 hill+0.100
 directions+0.097
 direction+0.094
Neg logit contributions
hire-0.158
 Princess-0.143
 Queen-0.142
ocket-0.138
 robber-0.137
Token the
Token ID262
Feature activation-1.19e-02

Pos logit contributions
 directions+0.034
 direction+0.033
 hill+0.032
 way+0.029
 kay+0.028
Neg logit contributions
hire-0.060
ocket-0.051
adies-0.051
 robber-0.050
vell-0.050
Token laundry
Token ID25724
Feature activation+2.29e-02

Pos logit contributions
hire+0.025
 Princess+0.022
 Queen+0.022
ocket+0.021
 robber+0.021
Neg logit contributions
bars-0.014
 outage-0.014
 hill-0.013
 directions-0.013
 direction-0.013
Token basket
Token ID7988
Feature activation+1.12e-02

Pos logit contributions
bars+0.026
 outage+0.026
 directions+0.024
spe+0.024
 blowing+0.024
Neg logit contributions
hire-0.044
 Princess-0.041
 Queen-0.040
 robber-0.039
ocket-0.039
Token in
Token ID287
Feature activation+1.51e-02

Pos logit contributions
 outage+0.015
bars+0.015
 directions+0.014
 kay+0.014
 hill+0.014
Neg logit contributions
hire-0.020
 Princess-0.017
ocket-0.017
 Queen-0.017
 robber-0.016
Token careful
Token ID8161
Feature activation+5.29e-02

Pos logit contributions
hire+0.012
 Princess+0.010
 Queen+0.010
ocket+0.010
 robber+0.010
Neg logit contributions
bars-0.008
 outage-0.008
 hill-0.008
 directions-0.008
 direction-0.008
Token and
Token ID290
Feature activation+6.70e-03

Pos logit contributions
 hill+0.078
bars+0.076
 kay+0.075
 outage+0.074
 Mechan+0.073
Neg logit contributions
hire-0.090
 Queen-0.076
 Princess-0.075
adies-0.072
ured-0.071
Token listen
Token ID6004
Feature activation+1.33e-02

Pos logit contributions
bars+0.010
 outage+0.010
 hill+0.010
 direction+0.009
spe+0.009
Neg logit contributions
hire-0.011
 Princess-0.010
 Queen-0.010
 robber-0.009
ocket-0.009
Token to
Token ID284
Feature activation-8.21e-04

Pos logit contributions
bars+0.012
 outage+0.011
 hill+0.011
 direction+0.010
 directions+0.010
Neg logit contributions
hire-0.031
 Princess-0.028
 Queen-0.028
ocket-0.027
 robber-0.027
Token your
Token ID534
Feature activation+8.19e-03

Pos logit contributions
hire+0.002
ocket+0.001
adies+0.001
 Princess+0.001
 Queen+0.001
Neg logit contributions
 directions-0.001
 direction-0.001
 hill-0.001
 outage-0.001
bars-0.001
Token gut
Token ID12500
Feature activation+8.16e-02

Pos logit contributions
bars+0.011
 outage+0.010
 hill+0.010
 blowing+0.009
spe+0.009
Neg logit contributions
hire-0.016
 Queen-0.015
 Princess-0.014
 robber-0.014
ocket-0.014
Token."
Token ID526
Feature activation+4.36e-03

Pos logit contributions
bars+0.120
 hill+0.114
 outage+0.114
 Mechan+0.107
 kay+0.107
Neg logit contributions
hire-0.135
 Princess-0.127
 Queen-0.124
ocket-0.110
ured-0.110
Token
Token ID198
Feature activation+1.27e-02

Pos logit contributions
 outage+0.005
bars+0.004
 hill+0.004
 directions+0.004
 direction+0.004
Neg logit contributions
hire-0.009
 Princess-0.009
 Queen-0.009
 Prince-0.008
ocket-0.008
Token
Token ID198
Feature activation-1.55e-02

Pos logit contributions
bars+0.016
 outage+0.016
 hill+0.015
 direction+0.014
 directions+0.014
Neg logit contributions
hire-0.025
 Princess-0.024
 Queen-0.022
 Prince-0.022
ocket-0.021
TokenTom
Token ID13787
Feature activation-2.12e-03

Pos logit contributions
hire+0.021
 Princess+0.018
 Queen+0.017
 robber+0.016
ocket+0.015
Neg logit contributions
 outage-0.030
bars-0.029
 direction-0.028
 hill-0.028
 directions-0.028
Token and
Token ID290
Feature activation-3.13e-02

Pos logit contributions
hire+0.004
 Princess+0.004
 Queen+0.004
ocket+0.004
 robber+0.004
Neg logit contributions
 outage-0.003
bars-0.003
 hill-0.003
 directions-0.002
 direction-0.002
Token he
Token ID339
Feature activation-2.32e-02

Pos logit contributions
bars+0.055
 outage+0.055
 hill+0.054
 directions+0.053
 direction+0.052
Neg logit contributions
hire-0.063
 Princess-0.055
 Queen-0.055
ocket-0.053
 robber-0.052
Token saw
Token ID2497
Feature activation-1.09e-02

Pos logit contributions
hire+0.044
 Queen+0.038
 Princess+0.037
 robber+0.037
ocket+0.036
Neg logit contributions
 hill-0.031
 direction-0.031
 kay-0.030
bars-0.029
 way-0.029
Token a
Token ID257
Feature activation-1.29e-02

Pos logit contributions
hire+0.020
ocket+0.017
adies+0.016
 Princess+0.016
 Queen+0.016
Neg logit contributions
 hill-0.015
 directions-0.015
 kay-0.015
 direction-0.015
 outage-0.015
Token big
Token ID1263
Feature activation-2.08e-03

Pos logit contributions
hire+0.024
 Princess+0.021
 Queen+0.021
 robber+0.020
ocket+0.020
Neg logit contributions
bars-0.018
 outage-0.018
 directions-0.017
 hill-0.017
 direction-0.016
Token staff
Token ID3085
Feature activation-9.87e-03

Pos logit contributions
hire+0.004
 Princess+0.004
 Queen+0.004
 robber+0.004
 mushrooms+0.003
Neg logit contributions
bars-0.003
 outage-0.003
 directions-0.002
spe-0.002
 direction-0.002
Token leaning
Token ID21804
Feature activation+8.15e-02

Pos logit contributions
hire+0.017
ocket+0.015
 Princess+0.015
 Queen+0.015
adies+0.014
Neg logit contributions
 directions-0.014
 outage-0.014
bars-0.014
spe-0.013
 direction-0.013
Token against
Token ID1028
Feature activation+2.79e-02

Pos logit contributions
 outage+0.120
 directions+0.115
bars+0.113
 hill+0.110
 direction+0.107
Neg logit contributions
hire-0.147
 Queen-0.125
 Princess-0.122
ocket-0.121
adies-0.121
Token a
Token ID257
Feature activation-2.06e-02

Pos logit contributions
 hill+0.037
bars+0.037
 outage+0.037
 directions+0.036
 direction+0.036
Neg logit contributions
hire-0.052
 Princess-0.046
 Queen-0.045
ocket-0.045
 robber-0.044
Token tree
Token ID5509
Feature activation+3.91e-03

Pos logit contributions
hire+0.045
 Princess+0.041
 Queen+0.040
ocket+0.039
 robber+0.039
Neg logit contributions
bars-0.021
 outage-0.021
 hill-0.020
 directions-0.019
 direction-0.019
Token.
Token ID13
Feature activation+1.46e-02

Pos logit contributions
bars+0.007
 Mechan+0.007
 directions+0.007
 outage+0.006
spe+0.006
Neg logit contributions
hire-0.005
 like-0.004
 as-0.004
 mushrooms-0.004
 Princess-0.004
Token
Token ID220
Feature activation+3.94e-02

Pos logit contributions
 outage+0.021
bars+0.020
 hill+0.019
 directions+0.019
spe+0.019
Neg logit contributions
hire-0.027
 Princess-0.025
 Queen-0.024
 robber-0.023
ocket-0.022
Token jumping
Token ID14284
Feature activation+4.00e-03

Pos logit contributions
hire+0.001
 Princess+0.001
ocket+0.001
 Queen+0.001
adies+0.001
Neg logit contributions
 hill-0.001
bars-0.001
 outage-0.001
 direction-0.001
 directions-0.001
Token over
Token ID625
Feature activation+2.75e-03

Pos logit contributions
 hill+0.004
 outage+0.004
bars+0.004
 directions+0.004
 direction+0.004
Neg logit contributions
hire-0.009
 Princess-0.008
ocket-0.008
 Queen-0.008
 robber-0.007
Token logs
Token ID17259
Feature activation+2.34e-02

Pos logit contributions
 outage+0.004
bars+0.003
 hill+0.003
 direction+0.003
 directions+0.003
Neg logit contributions
hire-0.005
 Princess-0.005
 Queen-0.005
 mushrooms-0.005
ocket-0.005
Token,
Token ID11
Feature activation-8.49e-03

Pos logit contributions
 outage+0.037
bars+0.035
 Mechan+0.034
icycle+0.034
 hurts+0.034
Neg logit contributions
hire-0.034
 as-0.030
 Queen-0.030
 Princess-0.029
ured-0.029
Token t
Token ID256
Feature activation+5.92e-03

Pos logit contributions
hire+0.018
 Princess+0.016
 Queen+0.016
ocket+0.016
 robber+0.015
Neg logit contributions
bars-0.009
 outage-0.009
 hill-0.009
 directions-0.009
 direction-0.009
Tokenipt
Token ID10257
Feature activation+8.14e-02

Pos logit contributions
 outage+0.007
 hill+0.007
bars+0.007
 direction+0.006
 directions+0.006
Neg logit contributions
hire-0.012
 Queen-0.011
 Princess-0.011
ocket-0.011
adies-0.010
Tokenoe
Token ID2577
Feature activation-1.32e-03

Pos logit contributions
 outage+0.077
 hill+0.077
bars+0.070
 direction+0.069
 directions+0.068
Neg logit contributions
hire-0.184
adies-0.161
 Princess-0.160
 Queen-0.160
ocket-0.155
Tokening
Token ID278
Feature activation+2.00e-02

Pos logit contributions
hire+0.002
 Princess+0.002
 Queen+0.002
ocket+0.002
adies+0.002
Neg logit contributions
 outage-0.002
 hill-0.002
bars-0.002
 directions-0.002
 direction-0.002
Token around
Token ID1088
Feature activation+2.02e-02

Pos logit contributions
 outage+0.027
bars+0.027
 hill+0.026
 directions+0.025
 direction+0.025
Neg logit contributions
hire-0.037
 Princess-0.033
 Queen-0.033
ocket-0.032
 robber-0.031
Token rocks
Token ID12586
Feature activation+1.59e-02

Pos logit contributions
bars+0.026
 outage+0.026
 hill+0.024
 directions+0.023
 direction+0.023
Neg logit contributions
hire-0.039
 Princess-0.036
 Queen-0.035
ocket-0.034
 robber-0.034
Token,
Token ID11
Feature activation-3.97e-03

Pos logit contributions
 outage+0.023
bars+0.021
 direction+0.021
 directions+0.021
 Mechan+0.021
Neg logit contributions
hire-0.027
 Princess-0.023
 Queen-0.023
 mushrooms-0.022
ocket-0.022
Token to
Token ID284
Feature activation+1.32e-02

Pos logit contributions
bars+0.005
 directions+0.005
 outage+0.005
 hill+0.005
 direction+0.005
Neg logit contributions
hire-0.006
ocket-0.005
 Princess-0.005
 robber-0.005
 Queen-0.005
Token come
Token ID1282
Feature activation-1.85e-03

Pos logit contributions
 outage+0.017
bars+0.017
 hill+0.017
 directions+0.016
 direction+0.016
Neg logit contributions
hire-0.025
 Princess-0.022
 Queen-0.022
ocket-0.022
 robber-0.021
Token and
Token ID290
Feature activation+2.21e-02

Pos logit contributions
hire+0.004
 Princess+0.003
 Queen+0.003
ocket+0.003
 robber+0.003
Neg logit contributions
 outage-0.002
bars-0.002
 hill-0.002
 directions-0.002
 direction-0.002
Token see
Token ID766
Feature activation-5.40e-03

Pos logit contributions
 outage+0.028
bars+0.028
 hill+0.027
 directions+0.026
 direction+0.026
Neg logit contributions
hire-0.043
 Princess-0.038
 Queen-0.038
ocket-0.037
 robber-0.036
Token the
Token ID262
Feature activation+1.28e-02

Pos logit contributions
hire+0.010
ocket+0.009
 robber+0.009
 Princess+0.009
adies+0.008
Neg logit contributions
bars-0.007
 directions-0.007
 hill-0.007
 direction-0.007
 outage-0.007
Token operating
Token ID5361
Feature activation+8.11e-02

Pos logit contributions
bars+0.017
 outage+0.017
 hill+0.016
 directions+0.016
 direction+0.015
Neg logit contributions
hire-0.025
 Princess-0.022
 Queen-0.022
ocket-0.021
 robber-0.021
Token theatre
Token ID21421
Feature activation+1.99e-03

Pos logit contributions
bars+0.085
 outage+0.078
umbles+0.074
bike+0.069
 blowing+0.069
Neg logit contributions
 Princess-0.160
 Queen-0.160
hire-0.159
ocket-0.142
 queen-0.141
Token!
Token ID0
Feature activation+1.56e-02

Pos logit contributions
 outage+0.002
bars+0.002
 hill+0.002
 directions+0.002
 direction+0.002
Neg logit contributions
hire-0.004
 Princess-0.004
 Queen-0.004
ocket-0.004
 robber-0.004
Token He
Token ID679
Feature activation+3.17e-03

Pos logit contributions
 outage+0.024
bars+0.023
 blowing+0.023
 hill+0.023
 directions+0.022
Neg logit contributions
hire-0.025
 Princess-0.023
 Queen-0.023
 robber-0.021
adies-0.021
Token was
Token ID373
Feature activation+2.32e-02

Pos logit contributions
bars+0.004
 outage+0.004
 hill+0.004
 directions+0.003
 direction+0.003
Neg logit contributions
hire-0.007
 Princess-0.006
 Queen-0.006
ocket-0.006
 robber-0.006
Token very
Token ID845
Feature activation+8.54e-03

Pos logit contributions
bars+0.034
 outage+0.034
 hill+0.034
 directions+0.032
 direction+0.032
Neg logit contributions
hire-0.040
 Princess-0.035
 Queen-0.035
ocket-0.034
 robber-0.033
Token felt
Token ID2936
Feature activation-1.51e-02

Pos logit contributions
bars+0.006
 hill+0.006
 direction+0.006
 blowing+0.006
 outage+0.006
Neg logit contributions
hire-0.012
 Princess-0.011
ocket-0.011
 Queen-0.010
adies-0.010
Token a
Token ID257
Feature activation+8.48e-03

Pos logit contributions
hire+0.029
 robber+0.026
ocket+0.025
 Queen+0.025
 Princess+0.025
Neg logit contributions
 directions-0.019
 direction-0.019
 outage-0.018
 kay-0.018
 blowing-0.018
Token little
Token ID1310
Feature activation-8.38e-03

Pos logit contributions
bars+0.013
 outage+0.013
 hill+0.013
 directions+0.013
 direction+0.012
Neg logit contributions
hire-0.014
 Princess-0.012
 Queen-0.012
ocket-0.012
 robber-0.011
Token bit
Token ID1643
Feature activation+1.08e-02

Pos logit contributions
hire+0.013
 Princess+0.012
 Queen+0.012
 robber+0.011
 Prince+0.010
Neg logit contributions
bars-0.014
 outage-0.014
 directions-0.013
spe-0.013
 hill-0.013
Token bra
Token ID8290
Feature activation-9.54e-03

Pos logit contributions
bars+0.016
 outage+0.016
 hill+0.015
 directions+0.015
 direction+0.014
Neg logit contributions
hire-0.019
 Princess-0.017
 Queen-0.017
ocket-0.016
 robber-0.016
Tokenver
Token ID332
Feature activation+8.08e-02

Pos logit contributions
hire+0.017
ocket+0.015
 Princess+0.015
 Queen+0.015
adies+0.014
Neg logit contributions
 hill-0.014
 outage-0.013
bars-0.013
 directions-0.013
 direction-0.013
Token.
Token ID13
Feature activation+7.19e-03

Pos logit contributions
 outage+0.103
bars+0.103
 hill+0.099
 directions+0.095
 direction+0.093
Neg logit contributions
hire-0.156
 Princess-0.140
 Queen-0.138
ocket-0.133
adies-0.131
Token She
Token ID1375
Feature activation-4.33e-04

Pos logit contributions
bars+0.008
 direction+0.008
 hill+0.008
 directions+0.008
spe+0.007
Neg logit contributions
hire-0.015
ocket-0.014
vell-0.014
 robber-0.013
pper-0.013
Token slowly
Token ID6364
Feature activation+1.80e-02

Pos logit contributions
hire+0.001
 Princess+0.001
 Queen+0.001
ocket+0.001
 robber+0.001
Neg logit contributions
bars-0.001
 hill-0.001
 outage-0.001
 direction-0.001
 directions-0.000
Token stepped
Token ID10764
Feature activation+1.72e-02

Pos logit contributions
bars+0.027
 outage+0.027
 hill+0.027
 directions+0.026
 direction+0.026
Neg logit contributions
hire-0.030
 Princess-0.027
 Queen-0.026
ocket-0.026
 robber-0.025
Token closer
Token ID5699
Feature activation-4.98e-04

Pos logit contributions
 hill+0.021
bars+0.020
 direction+0.020
 directions+0.020
 kay+0.019
Neg logit contributions
hire-0.035
 Princess-0.031
 Queen-0.031
ocket-0.030
 robber-0.030
Token Tom
Token ID4186
Feature activation-5.53e-03

Pos logit contributions
hire+0.010
 Princess+0.009
 Queen+0.009
ocket+0.008
 robber+0.008
Neg logit contributions
 outage-0.010
bars-0.010
 hill-0.010
 directions-0.010
 direction-0.010
Token noticed
Token ID6810
Feature activation+2.17e-03

Pos logit contributions
hire+0.009
 Princess+0.008
 Queen+0.008
ocket+0.008
 robber+0.008
Neg logit contributions
 outage-0.009
bars-0.008
 directions-0.008
 hill-0.008
 direction-0.008
Token a
Token ID257
Feature activation-5.36e-03

Pos logit contributions
bars+0.003
 outage+0.003
 hill+0.003
 direction+0.003
 directions+0.003
Neg logit contributions
hire-0.004
 Princess-0.003
 Queen-0.003
ocket-0.003
 robber-0.003
Token big
Token ID1263
Feature activation+1.43e-03

Pos logit contributions
hire+0.010
 Princess+0.009
 Queen+0.009
 robber+0.008
ocket+0.008
Neg logit contributions
bars-0.008
 outage-0.007
 directions-0.007
 direction-0.007
 hill-0.007
Token cane
Token ID33009
Feature activation+2.60e-02

Pos logit contributions
bars+0.002
 outage+0.002
 directions+0.002
spe+0.002
 direction+0.002
Neg logit contributions
hire-0.003
 Princess-0.003
 Queen-0.003
 robber-0.003
ocket-0.002
Token leaning
Token ID21804
Feature activation+8.07e-02

Pos logit contributions
 directions+0.042
 outage+0.039
spe+0.038
bars+0.038
 direction+0.037
Neg logit contributions
hire-0.041
 Princess-0.037
 Queen-0.037
ocket-0.037
 robber-0.037
Token against
Token ID1028
Feature activation+2.52e-02

Pos logit contributions
 outage+0.118
 directions+0.114
 hill+0.106
bars+0.105
 direction+0.103
Neg logit contributions
hire-0.146
 Queen-0.125
 robber-0.123
 Princess-0.122
ocket-0.119
Token a
Token ID257
Feature activation-1.98e-02

Pos logit contributions
 hill+0.028
 directions+0.028
 direction+0.028
bars+0.027
 kay+0.026
Neg logit contributions
hire-0.053
ocket-0.047
 Princess-0.047
 Queen-0.046
adies-0.045
Token tree
Token ID5509
Feature activation+5.61e-03

Pos logit contributions
hire+0.041
 Princess+0.037
 Queen+0.037
 robber+0.036
ocket+0.035
Neg logit contributions
bars-0.023
 outage-0.022
 directions-0.021
spe-0.021
 direction-0.020
Token.
Token ID13
Feature activation+7.02e-03

Pos logit contributions
 Mechan+0.010
 directions+0.010
bars+0.010
 outage+0.010
 mistakes+0.009
Neg logit contributions
hire-0.006
 like-0.006
 as-0.006
 Princess-0.005
ocket-0.005
Token "
Token ID366
Feature activation-2.64e-02

Pos logit contributions
 outage+0.013
 blowing+0.011
 directions+0.011
spe+0.011
bars+0.011
Neg logit contributions
hire-0.010
 Princess-0.009
 Queen-0.009
 robber-0.008
adies-0.008
Token go
Token ID467
Feature activation+2.16e-03

Pos logit contributions
hire+0.016
ocket+0.014
adies+0.013
vell+0.013
 robber+0.013
Neg logit contributions
 directions-0.009
 hill-0.009
 way-0.009
 direction-0.009
 kay-0.008
Token she
Token ID673
Feature activation-6.33e-03

Pos logit contributions
bars+0.003
 outage+0.003
 hill+0.002
spe+0.002
 directions+0.002
Neg logit contributions
hire-0.004
 Princess-0.004
 Queen-0.004
ocket-0.004
 robber-0.004
Token felt
Token ID2936
Feature activation-1.85e-02

Pos logit contributions
hire+0.013
 Princess+0.012
 Queen+0.012
ocket+0.012
 robber+0.011
Neg logit contributions
 outage-0.007
bars-0.007
 hill-0.007
 directions-0.007
 direction-0.006
Token much
Token ID881
Feature activation-1.02e-02

Pos logit contributions
hire+0.037
 Queen+0.032
ocket+0.032
 robber+0.032
 Princess+0.031
Neg logit contributions
 direction-0.022
 outage-0.022
 directions-0.022
bars-0.021
 hill-0.021
Token bra
Token ID8290
Feature activation-2.85e-03

Pos logit contributions
hire+0.018
 Princess+0.017
 Queen+0.016
ocket+0.015
 Prince+0.015
Neg logit contributions
bars-0.015
 outage-0.015
 hill-0.014
spe-0.014
ney-0.014
Tokenver
Token ID332
Feature activation+8.06e-02

Pos logit contributions
hire+0.005
 Princess+0.005
ocket+0.005
 Queen+0.004
adies+0.004
Neg logit contributions
 hill-0.004
bars-0.004
 outage-0.004
 directions-0.004
 direction-0.004
Token.
Token ID13
Feature activation+2.30e-02

Pos logit contributions
 outage+0.115
bars+0.114
 hill+0.111
 directions+0.105
ney+0.105
Neg logit contributions
hire-0.143
 Princess-0.129
 Queen-0.126
ocket-0.119
adies-0.117
Token And
Token ID843
Feature activation+1.37e-03

Pos logit contributions
bars+0.025
 direction+0.024
 hill+0.024
 directions+0.024
 outage+0.022
Neg logit contributions
hire-0.049
 robber-0.044
ocket-0.043
 Queen-0.043
 Princess-0.042
Token she
Token ID673
Feature activation+2.72e-03

Pos logit contributions
bars+0.002
 outage+0.002
 hill+0.002
 directions+0.002
 direction+0.002
Neg logit contributions
hire-0.002
 Princess-0.002
 Queen-0.002
ocket-0.002
 robber-0.002
Token never
Token ID1239
Feature activation+1.64e-02

Pos logit contributions
 outage+0.003
bars+0.003
 hill+0.003
 directions+0.003
 direction+0.003
Neg logit contributions
hire-0.005
 Princess-0.005
 Queen-0.005
ocket-0.005
 robber-0.005
Token felt
Token ID2936
Feature activation-1.37e-02

Pos logit contributions
bars+0.029
 outage+0.029
 hill+0.028
 directions+0.028
 direction+0.027
Neg logit contributions
hire-0.023
 Princess-0.021
 Queen-0.020
ocket-0.019
 robber-0.019
Token it
Token ID340
Feature activation+4.34e-03

Pos logit contributions
 outage+0.016
bars+0.016
 hill+0.015
 directions+0.015
 direction+0.015
Neg logit contributions
hire-0.014
 Princess-0.012
 Queen-0.012
ocket-0.012
 robber-0.011
Token made
Token ID925
Feature activation+4.23e-03

Pos logit contributions
bars+0.004
 outage+0.004
 hill+0.004
 directions+0.003
 direction+0.003
Neg logit contributions
hire-0.010
 Princess-0.009
 Queen-0.009
ocket-0.009
 robber-0.009
Token her
Token ID607
Feature activation+9.34e-03

Pos logit contributions
bars+0.008
 outage+0.008
 directions+0.008
 hill+0.007
 direction+0.007
Neg logit contributions
hire-0.006
ocket-0.005
 Queen-0.005
 Princess-0.005
 robber-0.005
Token even
Token ID772
Feature activation+2.23e-02

Pos logit contributions
bars+0.015
 outage+0.015
ipt+0.015
enn+0.014
 Mechan+0.014
Neg logit contributions
hire-0.014
 Princess-0.013
 Queen-0.013
ocket-0.012
 Prince-0.012
Token bra
Token ID8290
Feature activation+8.74e-03

Pos logit contributions
 outage+0.033
 hill+0.033
bars+0.033
ney+0.031
bike+0.030
Neg logit contributions
hire-0.038
 Princess-0.035
 Queen-0.034
 Prince-0.031
 robber-0.031
Tokenver
Token ID332
Feature activation+8.00e-02

Pos logit contributions
 hill+0.013
 directions+0.012
 direction+0.012
bars+0.012
 outage+0.012
Neg logit contributions
hire-0.015
ocket-0.014
 Princess-0.013
 Queen-0.013
adies-0.013
Token.
Token ID13
Feature activation+2.26e-02

Pos logit contributions
 outage+0.117
 hill+0.116
bars+0.114
 motor+0.108
ney+0.107
Neg logit contributions
hire-0.132
 Princess-0.121
 Queen-0.116
ocket-0.109
 Miranda-0.106
Token<|endoftext|>
Token ID50256
Feature activation+3.19e-02

Pos logit contributions
 outage+0.026
 hill+0.025
bars+0.024
 blowing+0.023
 directions+0.023
Neg logit contributions
hire-0.047
 Princess-0.045
 Queen-0.044
ocket-0.041
 Prince-0.041
TokenM
Token ID44
Feature activation+3.51e-03

Pos logit contributions
 outage+0.050
bars+0.049
spe+0.046
 directions+0.046
 blowing+0.045
Neg logit contributions
 Princess-0.051
hire-0.049
 Queen-0.047
 Prince-0.046
adies-0.044
Tokenolly
Token ID5098
Feature activation-5.08e-03

Pos logit contributions
 outage+0.006
bars+0.005
 hill+0.005
 directions+0.005
 direction+0.005
Neg logit contributions
hire-0.006
 Princess-0.005
ocket-0.005
 Queen-0.005
adies-0.005
Token was
Token ID373
Feature activation+1.24e-02

Pos logit contributions
hire+0.012
ocket+0.010
 robber+0.010
 Queen+0.010
 Princess+0.010
Neg logit contributions
 hill-0.005
bars-0.004
 outage-0.004
 direction-0.004
 blowing-0.004
Token saw
Token ID2497
Feature activation-1.77e-02

Pos logit contributions
hire+0.019
 Princess+0.016
 Queen+0.016
ocket+0.016
 robber+0.015
Neg logit contributions
 outage-0.020
 hill-0.020
bars-0.020
 direction-0.019
 directions-0.019
Token a
Token ID257
Feature activation-1.60e-02

Pos logit contributions
hire+0.026
 Princess+0.022
 Queen+0.022
ocket+0.021
 robber+0.021
Neg logit contributions
 outage-0.031
bars-0.031
 hill-0.031
 directions-0.030
 direction-0.030
Token ladder
Token ID18002
Feature activation+2.35e-02

Pos logit contributions
hire+0.029
 Princess+0.026
 Queen+0.026
 robber+0.025
ocket+0.024
Neg logit contributions
bars-0.023
 outage-0.022
 directions-0.022
 direction-0.021
spe-0.020
Token that
Token ID326
Feature activation+2.04e-02

Pos logit contributions
 directions+0.042
 outage+0.041
 tactic+0.040
 sibling+0.040
 limits+0.039
Neg logit contributions
 like-0.027
hire-0.026
 as-0.025
 behind-0.021
eh-0.020
Token was
Token ID373
Feature activation+4.50e-03

Pos logit contributions
 outage+0.027
bars+0.024
enn+0.024
 perspective+0.024
 mistakes+0.023
Neg logit contributions
hire-0.035
 Queen-0.033
 Princess-0.033
 mushrooms-0.030
adies-0.030
Token leaning
Token ID21804
Feature activation+8.00e-02

Pos logit contributions
bars+0.007
 outage+0.007
 directions+0.007
 hill+0.007
 direction+0.007
Neg logit contributions
hire-0.007
 Queen-0.006
 Princess-0.006
 robber-0.006
 queen-0.005
Token against
Token ID1028
Feature activation+1.85e-02

Pos logit contributions
 outage+0.114
 directions+0.114
 hill+0.110
 direction+0.107
bars+0.104
Neg logit contributions
hire-0.142
 Queen-0.118
 Princess-0.117
 robber-0.117
adies-0.116
Token a
Token ID257
Feature activation-2.51e-02

Pos logit contributions
 outage+0.027
bars+0.026
 hill+0.025
 directions+0.025
 direction+0.024
Neg logit contributions
hire-0.033
 Princess-0.029
 Queen-0.029
ocket-0.028
 robber-0.028
Token tree
Token ID5509
Feature activation+2.82e-03

Pos logit contributions
hire+0.047
 Princess+0.043
 Queen+0.043
 robber+0.042
 mushrooms+0.039
Neg logit contributions
bars-0.033
 outage-0.031
 directions-0.030
spe-0.030
umbles-0.029
Token.
Token ID13
Feature activation+3.68e-03

Pos logit contributions
 directions+0.005
 Mechan+0.005
bars+0.005
 outage+0.005
ipt+0.005
Neg logit contributions
hire-0.003
 like-0.003
 as-0.003
 Princess-0.002
 Queen-0.002
Token He
Token ID679
Feature activation-8.82e-03

Pos logit contributions
 outage+0.006
bars+0.006
 directions+0.006
spe+0.006
 blowing+0.006
Neg logit contributions
hire-0.005
 Princess-0.005
 Queen-0.005
 robber-0.005
 mushrooms-0.004
Token
Token ID198
Feature activation+1.28e-02

Pos logit contributions
 outage+0.001
bars+0.001
 hill+0.001
 directions+0.001
 direction+0.001
Neg logit contributions
hire-0.001
 Princess-0.001
 Queen-0.001
ocket-0.001
 robber-0.001
Token
Token ID198
Feature activation-1.74e-02

Pos logit contributions
bars+0.020
 outage+0.019
spe+0.018
bike+0.017
 hill+0.017
Neg logit contributions
hire-0.020
 Princess-0.020
ily-0.018
 Queen-0.018
adies-0.018
TokenThe
Token ID464
Feature activation+1.64e-02

Pos logit contributions
hire+0.032
 Princess+0.028
 Queen+0.028
 robber+0.027
ocket+0.026
Neg logit contributions
bars-0.025
 outage-0.024
 hill-0.023
 direction-0.023
 directions-0.022
Token pig
Token ID12967
Feature activation+9.71e-03

Pos logit contributions
 hill+0.025
bars+0.025
 outage+0.025
 directions+0.024
 direction+0.024
Neg logit contributions
hire-0.027
 Princess-0.024
ocket-0.023
 Queen-0.023
adies-0.022
Token was
Token ID373
Feature activation+1.50e-02

Pos logit contributions
 hill+0.012
 outage+0.012
bars+0.012
 direction+0.012
 directions+0.012
Neg logit contributions
hire-0.018
 Princess-0.016
 Queen-0.016
ocket-0.016
 robber-0.016
Token leaning
Token ID21804
Feature activation+7.99e-02

Pos logit contributions
bars+0.024
 outage+0.023
 hill+0.023
 directions+0.021
spe+0.021
Neg logit contributions
hire-0.025
 Princess-0.024
 Queen-0.023
ocket-0.022
 robber-0.022
Token against
Token ID1028
Feature activation+1.94e-02

Pos logit contributions
 outage+0.103
bars+0.098
 hill+0.097
 directions+0.094
 direction+0.091
Neg logit contributions
hire-0.158
 Princess-0.141
 Queen-0.141
ocket-0.135
 robber-0.134
Token a
Token ID257
Feature activation-2.52e-02

Pos logit contributions
 hill+0.025
 directions+0.025
 direction+0.024
bars+0.024
 outage+0.023
Neg logit contributions
hire-0.037
ocket-0.032
 Princess-0.032
 Queen-0.032
adies-0.031
Token big
Token ID1263
Feature activation-1.02e-02

Pos logit contributions
hire+0.046
 Princess+0.043
 Queen+0.042
 robber+0.041
 mushrooms+0.039
Neg logit contributions
bars-0.037
 outage-0.033
spe-0.033
 directions-0.031
ipt-0.031
Token tree
Token ID5509
Feature activation+4.70e-03

Pos logit contributions
hire+0.020
 Princess+0.018
 Queen+0.018
 robber+0.018
 mushrooms+0.017
Neg logit contributions
bars-0.014
 outage-0.013
spe-0.012
 directions-0.012
 blowing-0.011
Token when
Token ID618
Feature activation+4.86e-02

Pos logit contributions
bars+0.008
 outage+0.008
 directions+0.008
 Mechan+0.007
spe+0.007
Neg logit contributions
hire-0.006
 like-0.006
 as-0.006
 Princess-0.006
ocket-0.005
Token Billy
Token ID15890
Feature activation-4.13e-03

Pos logit contributions
hire+0.034
ocket+0.032
 Princess+0.031
 Queen+0.030
adies+0.030
Neg logit contributions
 hill-0.024
bars-0.024
 direction-0.023
 motor-0.023
 blowing-0.023
Token decided
Token ID3066
Feature activation-1.23e-02

Pos logit contributions
hire+0.008
 Princess+0.007
 Queen+0.007
ocket+0.007
 robber+0.007
Neg logit contributions
bars-0.005
 outage-0.005
 hill-0.005
 directions-0.005
 direction-0.005
Token to
Token ID284
Feature activation+2.19e-02

Pos logit contributions
hire+0.029
 Princess+0.027
 Queen+0.026
 robber+0.026
ocket+0.026
Neg logit contributions
bars-0.010
 outage-0.010
 hill-0.010
 directions-0.009
 direction-0.009
Token take
Token ID1011
Feature activation-1.25e-02

Pos logit contributions
 outage+0.033
bars+0.032
spe+0.031
 directions+0.030
 direction+0.030
Neg logit contributions
hire-0.036
 Princess-0.034
 Queen-0.033
 robber-0.030
ocket-0.030
Token his
Token ID465
Feature activation+2.49e-05

Pos logit contributions
hire+0.023
ocket+0.020
 Princess+0.020
 Queen+0.020
adies+0.019
Neg logit contributions
bars-0.017
 hill-0.017
 directions-0.017
 direction-0.017
 outage-0.016
Token leaning
Token ID21804
Feature activation+7.98e-02

Pos logit contributions
 outage+0.000
bars+0.000
spe+0.000
 directions+0.000
 hill+0.000
Neg logit contributions
hire-0.000
 Princess-0.000
 Queen-0.000
 robber-0.000
ocket-0.000
Token style
Token ID3918
Feature activation+9.58e-03

Pos logit contributions
 outage+0.102
 directions+0.098
 hill+0.088
spe+0.086
 difference+0.086
Neg logit contributions
hire-0.143
 Queen-0.130
 Princess-0.129
 robber-0.125
ocket-0.120
Token to
Token ID284
Feature activation-4.96e-03

Pos logit contributions
 outage+0.012
 directions+0.012
 hill+0.012
 direction+0.011
spe+0.011
Neg logit contributions
hire-0.018
 Princess-0.016
 Queen-0.015
 robber-0.015
ocket-0.015
Token a
Token ID257
Feature activation-2.40e-02

Pos logit contributions
hire+0.009
 Princess+0.008
 Queen+0.008
ocket+0.008
 robber+0.008
Neg logit contributions
 outage-0.007
bars-0.007
 hill-0.006
 directions-0.006
 direction-0.006
Token new
Token ID649
Feature activation+1.11e-03

Pos logit contributions
hire+0.044
 Princess+0.042
 Queen+0.040
 robber+0.039
ocket+0.037
Neg logit contributions
bars-0.033
 outage-0.032
 directions-0.030
 direction-0.029
 hill-0.029
Token level
Token ID1241
Feature activation+3.56e-02

Pos logit contributions
bars+0.001
 outage+0.001
umbles+0.001
 directions+0.001
umbling+0.001
Neg logit contributions
 Princess-0.002
 Queen-0.002
hire-0.002
 queen-0.002
 robber-0.002
Token.
Token ID13
Feature activation+1.47e-02

Pos logit contributions
 outage+0.032
bars+0.032
 hill+0.031
 directions+0.030
 direction+0.030
Neg logit contributions
hire-0.054
 Princess-0.048
 Queen-0.048
ocket-0.047
 robber-0.046
Token Today
Token ID6288
Feature activation+1.27e-02

Pos logit contributions
bars+0.020
 hill+0.019
 outage+0.019
 direction+0.019
 directions+0.018
Neg logit contributions
hire-0.028
ocket-0.025
adies-0.024
 Queen-0.023
vell-0.023
Token,
Token ID11
Feature activation-2.37e-03

Pos logit contributions
bars+0.011
 outage+0.011
 hill+0.011
 direction+0.010
 directions+0.010
Neg logit contributions
hire-0.030
ocket-0.027
 Queen-0.027
 Princess-0.027
 robber-0.026
Token she
Token ID673
Feature activation+3.33e-03

Pos logit contributions
hire+0.005
ocket+0.004
adies+0.004
ilies+0.004
vell+0.004
Neg logit contributions
 hill-0.003
bars-0.003
 way-0.003
 direction-0.003
 blowing-0.003
Token was
Token ID373
Feature activation+1.59e-02

Pos logit contributions
bars+0.003
 outage+0.003
 hill+0.003
 direction+0.003
spe+0.003
Neg logit contributions
hire-0.008
 Princess-0.007
 Queen-0.007
ocket-0.007
adies-0.007
Token leaning
Token ID21804
Feature activation+7.97e-02

Pos logit contributions
 outage+0.019
bars+0.019
 hill+0.019
 directions+0.018
 direction+0.018
Neg logit contributions
hire-0.032
 Princess-0.029
 Queen-0.028
ocket-0.028
 robber-0.027
Token against
Token ID1028
Feature activation+2.31e-02

Pos logit contributions
 outage+0.097
 hill+0.094
bars+0.092
 directions+0.090
 direction+0.088
Neg logit contributions
hire-0.160
 Princess-0.144
 Queen-0.143
ocket-0.138
 robber-0.137
Token the
Token ID262
Feature activation-1.91e-02

Pos logit contributions
bars+0.034
 hill+0.033
 directions+0.033
 direction+0.033
 outage+0.032
Neg logit contributions
hire-0.041
ocket-0.035
 Princess-0.034
adies-0.034
 Queen-0.034
Token wall
Token ID3355
Feature activation+2.67e-02

Pos logit contributions
hire+0.039
 Princess+0.036
 Queen+0.036
 robber+0.034
ocket+0.034
Neg logit contributions
 outage-0.022
bars-0.021
 hill-0.020
 directions-0.020
 blowing-0.020
Token in
Token ID287
Feature activation+2.05e-02

Pos logit contributions
 outage+0.032
 hill+0.032
 kay+0.031
bars+0.031
 directions+0.030
Neg logit contributions
hire-0.050
 Princess-0.045
 Queen-0.045
ocket-0.043
adies-0.042
Token the
Token ID262
Feature activation-1.78e-02

Pos logit contributions
 outage+0.031
bars+0.030
 hill+0.030
 blowing+0.028
ney+0.028
Neg logit contributions
hire-0.034
 Princess-0.031
 Queen-0.031
ocket-0.029
 robber-0.029
Token,
Token ID11
Feature activation-2.12e-02

Pos logit contributions
 hill+0.001
 kay+0.001
 way+0.001
 direction+0.001
 directions+0.001
Neg logit contributions
hire-0.005
ocket-0.005
 Queen-0.004
ns-0.004
 Princess-0.004
Token they
Token ID484
Feature activation-2.95e-02

Pos logit contributions
hire+0.041
ocket+0.038
 Princess+0.035
izards+0.035
adies+0.034
Neg logit contributions
 hill-0.029
 blowing-0.028
 kay-0.027
 direction-0.027
 motor-0.027
Token saw
Token ID2497
Feature activation-1.72e-02

Pos logit contributions
hire+0.052
 Princess+0.047
adies+0.046
 Queen+0.044
ured+0.043
Neg logit contributions
 hill-0.039
 kay-0.039
 way-0.038
 direction-0.038
 was-0.037
Token the
Token ID262
Feature activation-2.03e-03

Pos logit contributions
hire+0.029
 Princess+0.025
ocket+0.024
 Queen+0.024
adies+0.023
Neg logit contributions
bars-0.026
 hill-0.026
 outage-0.026
 directions-0.026
 direction-0.026
Token cap
Token ID1451
Feature activation-1.45e-02

Pos logit contributions
hire+0.003
 Princess+0.003
 robber+0.003
 Queen+0.003
 prince+0.003
Neg logit contributions
 outage-0.003
bars-0.003
spe-0.003
 directions-0.003
ney-0.003
Token leaning
Token ID21804
Feature activation+7.97e-02

Pos logit contributions
hire+0.028
 Princess+0.025
 Queen+0.025
ocket+0.024
 robber+0.024
Neg logit contributions
 outage-0.018
bars-0.018
 hill-0.018
 directions-0.017
 direction-0.017
Token against
Token ID1028
Feature activation+2.03e-02

Pos logit contributions
 outage+0.118
 directions+0.109
 hill+0.106
bars+0.106
 direction+0.102
Neg logit contributions
hire-0.142
 Queen-0.124
 Princess-0.123
 robber-0.121
ocket-0.117
Token a
Token ID257
Feature activation-2.51e-02

Pos logit contributions
 hill+0.027
 direction+0.026
 directions+0.026
bars+0.025
 blowing+0.025
Neg logit contributions
hire-0.039
ocket-0.035
 Princess-0.035
 Queen-0.034
adies-0.034
Token tree
Token ID5509
Feature activation+4.44e-03

Pos logit contributions
hire+0.049
 Princess+0.044
 Queen+0.044
 robber+0.043
ocket+0.041
Neg logit contributions
bars-0.033
 outage-0.031
spe-0.029
 directions-0.029
umbles-0.028
Token.
Token ID13
Feature activation-1.63e-03

Pos logit contributions
bars+0.007
 directions+0.007
 outage+0.007
 Mechan+0.007
iff+0.007
Neg logit contributions
hire-0.006
 like-0.005
 as-0.005
 Princess-0.005
 Queen-0.005
Token Billy
Token ID15890
Feature activation-1.23e-03

Pos logit contributions
hire+0.003
ocket+0.003
 Queen+0.002
 Princess+0.002
adies+0.002
Neg logit contributions
 hill-0.002
bars-0.002
 motor-0.002
 outage-0.002
 direction-0.002
Token stick
Token ID4859
Feature activation-2.35e-03

Pos logit contributions
hire+0.018
 Princess+0.016
 Queen+0.016
ocket+0.015
 robber+0.015
Neg logit contributions
bars-0.011
 outage-0.011
 hill-0.010
spe-0.010
 directions-0.010
Token the
Token ID262
Feature activation+5.87e-03

Pos logit contributions
hire+0.005
 Princess+0.004
 Queen+0.004
ocket+0.004
adies+0.004
Neg logit contributions
 hill-0.003
bars-0.003
 directions-0.003
 direction-0.003
 outage-0.003
Token mustache
Token ID49303
Feature activation+6.89e-04

Pos logit contributions
 outage+0.007
bars+0.007
 hill+0.007
 directions+0.007
 direction+0.006
Neg logit contributions
hire-0.012
 Princess-0.011
 Queen-0.010
ocket-0.010
 robber-0.010
Token on
Token ID319
Feature activation-1.67e-02

Pos logit contributions
 outage+0.001
bars+0.001
 hill+0.001
 directions+0.001
 direction+0.001
Neg logit contributions
hire-0.001
 Princess-0.001
 Queen-0.001
ocket-0.001
 robber-0.001
Token her
Token ID607
Feature activation-2.34e-02

Pos logit contributions
hire+0.035
ocket+0.032
adies+0.032
 Princess+0.030
 Queen+0.030
Neg logit contributions
 hill-0.019
 directions-0.018
 direction-0.018
bars-0.017
 way-0.017
Token upper
Token ID6727
Feature activation+7.96e-02

Pos logit contributions
hire+0.045
 Princess+0.040
 Queen+0.040
ocket+0.040
 robber+0.038
Neg logit contributions
 outage-0.031
bars-0.030
 hill-0.028
 directions-0.028
spe-0.028
Token lip
Token ID10645
Feature activation-9.60e-04

Pos logit contributions
 outage+0.102
 hill+0.095
bars+0.094
spe+0.092
 directions+0.090
Neg logit contributions
hire-0.159
 Queen-0.139
ocket-0.139
 Princess-0.138
 robber-0.133
Token with
Token ID351
Feature activation-2.30e-02

Pos logit contributions
hire+0.002
 Princess+0.002
 Queen+0.002
ocket+0.002
 robber+0.002
Neg logit contributions
 outage-0.001
bars-0.001
 hill-0.001
 directions-0.001
 direction-0.001
Token some
Token ID617
Feature activation-1.05e-02

Pos logit contributions
hire+0.040
 Princess+0.036
 Queen+0.035
ocket+0.035
adies+0.035
Neg logit contributions
 directions-0.033
bars-0.033
 hill-0.032
 direction-0.032
 outage-0.031
Token tape
Token ID9154
Feature activation+2.72e-02

Pos logit contributions
hire+0.018
 Princess+0.016
 Queen+0.016
ocket+0.015
 robber+0.015
Neg logit contributions
 outage-0.016
bars-0.016
 hill-0.015
 directions-0.015
 direction-0.015
Token.
Token ID13
Feature activation+1.62e-03

Pos logit contributions
 outage+0.031
bars+0.030
 hill+0.029
 directions+0.029
 direction+0.028
Neg logit contributions
hire-0.056
 Princess-0.050
 Queen-0.050
ocket-0.049
 robber-0.048
Token higher
Token ID2440
Feature activation+1.67e-02

Pos logit contributions
hire+0.014
ocket+0.012
 Queen+0.012
 Princess+0.012
ns+0.012
Neg logit contributions
 hill-0.006
 directions-0.006
 direction-0.006
 kay-0.006
bars-0.005
Token,
Token ID11
Feature activation-2.63e-03

Pos logit contributions
 outage+0.024
bars+0.022
spe+0.020
ney+0.020
 directions+0.020
Neg logit contributions
hire-0.031
 Princess-0.027
 Queen-0.027
 mushrooms-0.025
 robber-0.025
Token he
Token ID339
Feature activation-6.85e-03

Pos logit contributions
hire+0.006
ocket+0.005
 Queen+0.005
adies+0.005
 Princess+0.005
Neg logit contributions
 hill-0.003
bars-0.003
 kay-0.003
 direction-0.003
 directions-0.003
Token felt
Token ID2936
Feature activation-2.28e-02

Pos logit contributions
hire+0.013
 Princess+0.011
 Queen+0.011
ocket+0.011
 robber+0.011
Neg logit contributions
bars-0.009
 outage-0.009
 hill-0.009
 directions-0.009
 direction-0.009
Token bra
Token ID8290
Feature activation-1.16e-02

Pos logit contributions
hire+0.042
 Queen+0.037
 Princess+0.036
ocket+0.036
 robber+0.036
Neg logit contributions
bars-0.031
 outage-0.031
 directions-0.031
 direction-0.031
 hill-0.030
Tokenver
Token ID332
Feature activation+7.96e-02

Pos logit contributions
hire+0.022
 Princess+0.019
 Queen+0.019
ocket+0.019
 robber+0.018
Neg logit contributions
 outage-0.015
bars-0.015
 hill-0.015
 directions-0.015
 direction-0.014
Token and
Token ID290
Feature activation+5.47e-03

Pos logit contributions
 outage+0.126
bars+0.121
 hill+0.118
 Mechan+0.114
 directions+0.114
Neg logit contributions
hire-0.126
 Princess-0.106
 Queen-0.103
 robber-0.101
 as-0.099
Token bra
Token ID8290
Feature activation-1.51e-02

Pos logit contributions
 outage+0.008
bars+0.008
 hill+0.008
ney+0.008
 blowing+0.008
Neg logit contributions
hire-0.009
 Princess-0.008
 Queen-0.008
ocket-0.008
 robber-0.007
Tokenver
Token ID332
Feature activation+7.58e-02

Pos logit contributions
hire+0.027
 Princess+0.024
 Queen+0.024
ocket+0.023
 robber+0.022
Neg logit contributions
bars-0.022
 outage-0.022
 hill-0.021
 directions-0.020
 direction-0.020
Token.
Token ID13
Feature activation+2.00e-02

Pos logit contributions
 outage+0.118
bars+0.110
 hill+0.107
ney+0.103
 directions+0.102
Neg logit contributions
hire-0.130
 Princess-0.115
 Queen-0.111
 robber-0.108
ocket-0.104
Token Soon
Token ID15894
Feature activation-1.16e-02

Pos logit contributions
 outage+0.025
bars+0.024
 hill+0.023
 directions+0.022
 blowing+0.022
Neg logit contributions
hire-0.040
 Princess-0.036
 Queen-0.035
ocket-0.034
 robber-0.034
TokenM
Token ID44
Feature activation-9.04e-04

Pos logit contributions
 outage+0.052
bars+0.052
spe+0.047
 directions+0.046
 blowing+0.045
Neg logit contributions
 Princess-0.063
hire-0.060
 Queen-0.058
 Prince-0.058
adies-0.055
Tokenolly
Token ID5098
Feature activation-8.59e-03

Pos logit contributions
hire+0.002
 Princess+0.002
ocket+0.002
 Queen+0.002
 robber+0.001
Neg logit contributions
 outage-0.001
bars-0.001
 directions-0.001
 direction-0.001
 hill-0.001
Token was
Token ID373
Feature activation+7.12e-03

Pos logit contributions
hire+0.021
ocket+0.018
 Princess+0.018
 Queen+0.018
 robber+0.018
Neg logit contributions
 hill-0.007
 outage-0.007
bars-0.007
 direction-0.006
 way-0.006
Token a
Token ID257
Feature activation+1.99e-02

Pos logit contributions
bars+0.009
 outage+0.009
 hill+0.009
 blowing+0.009
 direction+0.009
Neg logit contributions
hire-0.014
 Queen-0.012
 Princess-0.012
ocket-0.012
 robber-0.012
Token very
Token ID845
Feature activation-1.42e-02

Pos logit contributions
 hill+0.024
 kay+0.022
 motor+0.021
 room+0.020
 way+0.019
Neg logit contributions
hire-0.042
ocket-0.038
adies-0.038
izards-0.036
ilies-0.036
Token careful
Token ID8161
Feature activation+7.95e-02

Pos logit contributions
hire+0.030
ocket+0.026
 Queen+0.025
adies+0.025
izards+0.025
Neg logit contributions
 hill-0.016
 kay-0.015
 day-0.015
 way-0.015
spe-0.015
Token girl
Token ID2576
Feature activation-5.85e-03

Pos logit contributions
bars+0.094
 outage+0.093
 directions+0.089
 hill+0.087
 direction+0.086
Neg logit contributions
hire-0.161
 Princess-0.145
 Queen-0.144
 robber-0.141
ocket-0.138
Token.
Token ID13
Feature activation+2.26e-03

Pos logit contributions
hire+0.012
 Princess+0.011
 Queen+0.011
ocket+0.010
 robber+0.010
Neg logit contributions
 outage-0.007
bars-0.007
 hill-0.007
 directions-0.007
 direction-0.006
Token She
Token ID1375
Feature activation+8.59e-03

Pos logit contributions
bars+0.004
 outage+0.003
 hill+0.003
 directions+0.003
 direction+0.003
Neg logit contributions
hire-0.004
ocket-0.003
 Queen-0.003
 Princess-0.003
 robber-0.003
Token loved
Token ID6151
Feature activation+1.71e-02

Pos logit contributions
bars+0.009
 outage+0.009
 hill+0.009
 direction+0.008
 directions+0.008
Neg logit contributions
hire-0.019
 Princess-0.017
 Queen-0.017
ocket-0.017
 robber-0.016
Token to
Token ID284
Feature activation+1.14e-02

Pos logit contributions
 outage+0.017
bars+0.017
 hill+0.016
 direction+0.016
 directions+0.015
Neg logit contributions
hire-0.037
 Princess-0.035
 Queen-0.034
 robber-0.033
ocket-0.033
Token
Token ID198
Feature activation-3.01e-02

Pos logit contributions
hire+0.005
 Princess+0.005
 Queen+0.005
ocket+0.004
ily+0.004
Neg logit contributions
bars-0.003
 hill-0.003
 outage-0.003
spe-0.003
bike-0.003
Token"
Token ID1
Feature activation-5.53e-02

Pos logit contributions
hire+0.055
 Princess+0.049
 Queen+0.049
ocket+0.047
 robber+0.046
Neg logit contributions
bars-0.042
 outage-0.042
 hill-0.040
 directions-0.039
 direction-0.039
TokenWHAT
Token ID32971
Feature activation-5.32e-02

Pos logit contributions
hire+0.101
ocket+0.091
 Princess+0.088
 Queen+0.087
 robber+0.086
Neg logit contributions
 outage-0.075
 hill-0.074
 directions-0.074
bars-0.074
 direction-0.072
Token ARE
Token ID15986
Feature activation-5.25e-02

Pos logit contributions
hire+0.111
 Queen+0.094
 Princess+0.093
ocket+0.091
 robber+0.088
Neg logit contributions
 outage-0.065
 hill-0.063
bars-0.063
 story-0.060
 direction-0.058
Token YOU
Token ID7013
Feature activation-3.28e-02

Pos logit contributions
hire+0.100
ocket+0.084
 Queen+0.082
ns+0.080
vell+0.077
Neg logit contributions
 story-0.069
 hill-0.066
bars-0.064
 kay-0.063
 motor-0.062
Token DO
Token ID8410
Feature activation-7.92e-02

Pos logit contributions
hire+0.068
 Princess+0.060
 Queen+0.060
ocket+0.059
 robber+0.057
Neg logit contributions
 hill-0.037
 outage-0.037
bars-0.036
 direction-0.035
 directions-0.034
TokenING
Token ID2751
Feature activation+5.16e-03

Pos logit contributions
hire+0.151
 Queen+0.134
 mushrooms+0.132
ocket+0.129
 queen+0.127
Neg logit contributions
 story-0.101
 outage-0.098
s-0.091
umbles-0.091
 hill-0.090
Token?"
Token ID1701
Feature activation-3.77e-02

Pos logit contributions
 hill+0.005
 outage+0.005
 arm+0.005
 blowing+0.005
enn+0.005
Neg logit contributions
hire-0.011
ocket-0.010
vell-0.010
 Princess-0.009
 mushrooms-0.009
Token Mom
Token ID11254
Feature activation+2.36e-02

Pos logit contributions
hire+0.068
ocket+0.057
 robber+0.055
vell+0.054
 Queen+0.053
Neg logit contributions
 outage-0.055
 direction-0.054
bars-0.054
 hill-0.054
 directions-0.053
Token yells
Token ID42675
Feature activation-3.91e-02

Pos logit contributions
 outage+0.033
spe+0.033
 hill+0.033
bars+0.032
 kay+0.031
Neg logit contributions
hire-0.038
 Princess-0.037
 Queen-0.037
adies-0.035
 Prince-0.034
Token.
Token ID13
Feature activation+6.25e-03

Pos logit contributions
hire+0.095
ocket+0.084
 Princess+0.084
 Queen+0.084
 robber+0.082
Neg logit contributions
bars-0.031
 outage-0.031
 directions-0.029
 direction-0.029
 hill-0.029
Token and
Token ID290
Feature activation-4.09e-02

Pos logit contributions
hire+0.002
 Princess+0.002
 Queen+0.001
ocket+0.001
 robber+0.001
Neg logit contributions
bars-0.001
 outage-0.001
 hill-0.001
 directions-0.001
 direction-0.001
Token bubbles
Token ID25037
Feature activation-9.77e-03

Pos logit contributions
hire+0.074
 Princess+0.065
ocket+0.063
 Queen+0.062
adies+0.061
Neg logit contributions
 outage-0.059
bars-0.057
 hill-0.057
 directions-0.056
 direction-0.056
Token and
Token ID290
Feature activation-3.81e-02

Pos logit contributions
hire+0.021
 Princess+0.019
 Queen+0.019
ocket+0.019
 robber+0.019
Neg logit contributions
 outage-0.010
bars-0.010
 hill-0.010
 directions-0.009
 direction-0.009
Token sounds
Token ID5238
Feature activation-2.63e-02

Pos logit contributions
hire+0.068
 Princess+0.060
ocket+0.059
 Queen+0.058
adies+0.057
Neg logit contributions
 outage-0.055
bars-0.054
 hill-0.053
 directions-0.052
 direction-0.051
Token and
Token ID290
Feature activation-2.14e-02

Pos logit contributions
hire+0.067
ocket+0.064
 Queen+0.062
adies+0.061
 Princess+0.061
Neg logit contributions
 outage-0.016
 blowing-0.016
 directions-0.015
 hill-0.015
 direction-0.014
Token smells
Token ID25760
Feature activation-7.89e-02

Pos logit contributions
hire+0.038
 Princess+0.034
 Queen+0.033
ocket+0.032
 robber+0.032
Neg logit contributions
bars-0.031
 outage-0.031
 hill-0.030
 directions-0.029
 direction-0.029
Token.
Token ID13
Feature activation-2.20e-02

Pos logit contributions
hire+0.184
ocket+0.168
 robber+0.166
 Queen+0.165
adies+0.165
Neg logit contributions
 directions-0.065
 direction-0.063
 outage-0.063
 hill-0.062
bars-0.060
Token Lily
Token ID20037
Feature activation+5.13e-03

Pos logit contributions
hire+0.044
ocket+0.039
vell+0.038
pper+0.037
ils+0.037
Neg logit contributions
bars-0.027
 outage-0.026
 hill-0.025
 direction-0.025
 directions-0.025
Token was
Token ID373
Feature activation+1.16e-02

Pos logit contributions
bars+0.006
 outage+0.006
 hill+0.005
 directions+0.005
 direction+0.005
Neg logit contributions
hire-0.011
 Princess-0.010
 Queen-0.010
ocket-0.009
 robber-0.009
Token scared
Token ID12008
Feature activation+2.27e-02

Pos logit contributions
 outage+0.015
bars+0.015
 hill+0.015
 directions+0.015
 direction+0.014
Neg logit contributions
hire-0.022
 Princess-0.019
 Queen-0.019
ocket-0.019
 robber-0.018
Token.
Token ID13
Feature activation-1.64e-02

Pos logit contributions
 outage+0.025
bars+0.025
 hill+0.024
 directions+0.023
 direction+0.023
Neg logit contributions
hire-0.048
 Princess-0.044
 Queen-0.043
ocket-0.041
 robber-0.041
Token.
Token ID13
Feature activation+4.33e-03

Pos logit contributions
hire+0.002
 Princess+0.002
 Queen+0.002
 mushrooms+0.002
ocket+0.002
Neg logit contributions
 outage-0.002
bars-0.002
 hill-0.002
 directions-0.002
spe-0.002
Token She
Token ID1375
Feature activation-8.06e-03

Pos logit contributions
bars+0.005
 hill+0.004
 outage+0.004
 direction+0.004
 directions+0.004
Neg logit contributions
hire-0.010
ocket-0.009
 Queen-0.008
 robber-0.008
adies-0.008
Token liked
Token ID8288
Feature activation-2.84e-03

Pos logit contributions
hire+0.015
 Princess+0.014
 Queen+0.013
ocket+0.013
 robber+0.013
Neg logit contributions
bars-0.011
 outage-0.010
 hill-0.010
 directions-0.010
 direction-0.010
Token building
Token ID2615
Feature activation-3.84e-03

Pos logit contributions
hire+0.006
 Princess+0.005
 Queen+0.005
ocket+0.005
 mushrooms+0.005
Neg logit contributions
 outage-0.003
bars-0.003
 hill-0.003
 direction-0.003
 directions-0.003
Token sand
Token ID6450
Feature activation-3.34e-02

Pos logit contributions
hire+0.008
 robber+0.007
ocket+0.007
 Queen+0.007
 Princess+0.007
Neg logit contributions
 directions-0.005
 hill-0.004
 direction-0.004
bars-0.004
 blowing-0.004
Tokencast
Token ID2701
Feature activation-7.82e-02

Pos logit contributions
hire+0.071
 Princess+0.065
 Queen+0.064
 robber+0.064
ocket+0.063
Neg logit contributions
 hill-0.039
bars-0.036
 blowing-0.034
 directions-0.034
 outage-0.033
Tokenles
Token ID829
Feature activation-1.70e-02

Pos logit contributions
hire+0.136
 Queen+0.129
 Princess+0.126
ocket+0.120
 robber+0.120
Neg logit contributions
bars-0.114
 hill-0.110
s-0.108
 blowing-0.104
 kay-0.101
Token and
Token ID290
Feature activation-2.28e-02

Pos logit contributions
hire+0.039
 Princess+0.034
 Queen+0.034
 robber+0.034
ocket+0.034
Neg logit contributions
 hill-0.016
bars-0.016
 outage-0.016
 blowing-0.015
 directions-0.015
Token collecting
Token ID13157
Feature activation-3.56e-02

Pos logit contributions
hire+0.043
 Princess+0.038
 Queen+0.038
ocket+0.037
 robber+0.036
Neg logit contributions
 outage-0.030
bars-0.030
 hill-0.030
 directions-0.029
 direction-0.028
Token seas
Token ID21547
Feature activation-4.71e-02

Pos logit contributions
hire+0.069
 robber+0.066
 Queen+0.064
ocket+0.062
 Princess+0.062
Neg logit contributions
 directions-0.047
 direction-0.045
 kay-0.044
 hill-0.044
 way-0.042
Tokenhell
Token ID12758
Feature activation-6.94e-02

Pos logit contributions
hire+0.064
 Queen+0.061
 Princess+0.059
 robber+0.059
 Prince+0.054
Neg logit contributions
 outage-0.083
s-0.082
 directions-0.081
bars-0.080
 hill-0.080
Token the
Token ID262
Feature activation-1.23e-02

Pos logit contributions
hire+0.014
 Princess+0.014
 Queen+0.013
 robber+0.012
 Prince+0.012
Neg logit contributions
bars-0.012
 outage-0.012
spe-0.011
ney-0.011
umbling-0.011
Token beach
Token ID10481
Feature activation-9.63e-03

Pos logit contributions
hire+0.026
 Princess+0.024
 Queen+0.023
ocket+0.023
 robber+0.022
Neg logit contributions
bars-0.013
 outage-0.013
 hill-0.013
 directions-0.012
 direction-0.012
Token and
Token ID290
Feature activation-4.04e-03

Pos logit contributions
hire+0.019
 Princess+0.017
 Queen+0.017
ocket+0.016
 robber+0.016
Neg logit contributions
 outage-0.012
bars-0.012
 directions-0.011
 hill-0.011
 direction-0.011
Token build
Token ID1382
Feature activation-5.72e-03

Pos logit contributions
hire+0.008
 Princess+0.007
 Queen+0.007
ocket+0.007
 robber+0.007
Neg logit contributions
bars-0.005
 outage-0.005
 hill-0.005
 directions-0.005
 direction-0.005
Token sand
Token ID6450
Feature activation-3.02e-02

Pos logit contributions
hire+0.012
 robber+0.010
 Queen+0.010
ocket+0.010
 Princess+0.010
Neg logit contributions
 directions-0.007
 hill-0.007
 direction-0.007
 way-0.007
 kay-0.006
Tokencast
Token ID2701
Feature activation-7.76e-02

Pos logit contributions
hire+0.051
 Princess+0.045
 Queen+0.045
ocket+0.043
 robber+0.043
Neg logit contributions
 hill-0.046
bars-0.046
 outage-0.045
 directions-0.044
 direction-0.043
Tokenles
Token ID829
Feature activation-1.24e-02

Pos logit contributions
hire+0.136
 Queen+0.128
 Princess+0.126
 robber+0.119
ocket+0.119
Neg logit contributions
bars-0.112
 hill-0.108
s-0.104
 directions-0.098
 direction-0.098
Token.
Token ID13
Feature activation+7.95e-04

Pos logit contributions
hire+0.027
 Princess+0.025
 Queen+0.025
ocket+0.024
 robber+0.024
Neg logit contributions
bars-0.012
 outage-0.012
 hill-0.012
 directions-0.011
 direction-0.011
Token
Token ID220
Feature activation+2.75e-02

Pos logit contributions
 hill+0.001
bars+0.001
 direction+0.001
 directions+0.001
 outage+0.001
Neg logit contributions
hire-0.002
ocket-0.002
vell-0.002
 Queen-0.002
adies-0.002
Token
Token ID198
Feature activation+8.21e-03

Pos logit contributions
 outage+0.023
bars+0.022
 hill+0.022
 directions+0.020
 direction+0.020
Neg logit contributions
hire-0.065
 Princess-0.060
 Queen-0.059
ocket-0.058
adies-0.057
Token
Token ID198
Feature activation-1.75e-02

Pos logit contributions
 outage+0.007
bars+0.007
 hill+0.007
 direction+0.006
 directions+0.006
Neg logit contributions
hire-0.019
 Princess-0.018
 Queen-0.017
ocket-0.017
 robber-0.017
Token,
Token ID11
Feature activation-1.18e-02

Pos logit contributions
 outage+0.011
bars+0.011
 hill+0.011
 directions+0.011
 direction+0.011
Neg logit contributions
hire-0.020
 Princess-0.018
 Queen-0.018
ocket-0.017
 robber-0.017
Token laughed
Token ID13818
Feature activation-1.78e-02

Pos logit contributions
hire+0.023
 Princess+0.021
 Queen+0.020
 robber+0.020
ocket+0.019
Neg logit contributions
 outage-0.015
bars-0.015
 hill-0.014
 directions-0.014
 direction-0.014
Token and
Token ID290
Feature activation+3.53e-03

Pos logit contributions
hire+0.039
ocket+0.037
 Queen+0.035
 Princess+0.035
adies+0.034
Neg logit contributions
 blowing-0.018
 hill-0.017
bars-0.017
 directions-0.017
 direction-0.016
Token built
Token ID3170
Feature activation-4.40e-03

Pos logit contributions
bars+0.005
 outage+0.005
 hill+0.005
 directions+0.004
 direction+0.004
Neg logit contributions
hire-0.007
 Princess-0.006
 Queen-0.006
ocket-0.006
adies-0.006
Token sand
Token ID6450
Feature activation-3.23e-02

Pos logit contributions
hire+0.008
ocket+0.007
 Queen+0.007
 Princess+0.007
adies+0.007
Neg logit contributions
 directions-0.006
 hill-0.006
 direction-0.006
bars-0.006
 outage-0.006
Tokencast
Token ID2701
Feature activation-7.62e-02

Pos logit contributions
hire+0.059
 Princess+0.053
 Queen+0.052
ocket+0.052
 robber+0.050
Neg logit contributions
 hill-0.046
bars-0.045
 directions-0.043
 outage-0.043
 direction-0.042
Tokenles
Token ID829
Feature activation-1.58e-02

Pos logit contributions
hire+0.132
 Queen+0.123
 Princess+0.120
ocket+0.116
 robber+0.111
Neg logit contributions
bars-0.114
 hill-0.110
s-0.105
 directions-0.103
 blowing-0.102
Token.
Token ID13
Feature activation-4.42e-05

Pos logit contributions
hire+0.034
 Princess+0.031
 Queen+0.030
ocket+0.030
 robber+0.029
Neg logit contributions
bars-0.017
 outage-0.017
 hill-0.016
 directions-0.016
 direction-0.015
Token Soon
Token ID15894
Feature activation-1.25e-02

Pos logit contributions
hire+0.000
ocket+0.000
 Queen+0.000
 Princess+0.000
adies+0.000
Neg logit contributions
bars-0.000
 hill-0.000
 outage-0.000
 direction-0.000
 directions-0.000
Token,
Token ID11
Feature activation-1.31e-02

Pos logit contributions
hire+0.029
ocket+0.027
 Queen+0.025
adies+0.025
 robber+0.025
Neg logit contributions
 hill-0.011
 directions-0.011
 way-0.011
 direction-0.011
 kay-0.011
Token they
Token ID484
Feature activation-1.76e-02

Pos logit contributions
hire+0.027
ocket+0.024
adies+0.023
 Princess+0.022
 Queen+0.022
Neg logit contributions
 hill-0.016
bars-0.015
 direction-0.015
 directions-0.015
 kay-0.015
Token is
Token ID318
Feature activation-6.72e-04

Pos logit contributions
 outage+0.000
bars+0.000
spe+0.000
bike+0.000
 direction+0.000
Neg logit contributions
hire-0.000
 Princess-0.000
 Queen-0.000
 mushrooms-0.000
ocket-0.000
Token soft
Token ID2705
Feature activation+6.26e-03

Pos logit contributions
hire+0.001
 Princess+0.001
 Queen+0.001
ocket+0.001
 robber+0.001
Neg logit contributions
 blowing-0.001
 hill-0.001
 directions-0.001
 direction-0.001
 outage-0.001
Token and
Token ID290
Feature activation-2.93e-03

Pos logit contributions
 hill+0.005
 outage+0.005
bars+0.005
 direction+0.005
 directions+0.005
Neg logit contributions
hire-0.015
 Princess-0.013
 Queen-0.013
ocket-0.013
 robber-0.013
Token brown
Token ID7586
Feature activation-4.74e-03

Pos logit contributions
hire+0.005
 Princess+0.005
 Queen+0.005
ocket+0.005
 robber+0.004
Neg logit contributions
 outage-0.004
bars-0.004
 hill-0.004
 directions-0.004
 direction-0.004
Token and
Token ID290
Feature activation-1.05e-02

Pos logit contributions
hire+0.011
 Princess+0.010
 Queen+0.010
ocket+0.010
 robber+0.009
Neg logit contributions
 outage-0.004
bars-0.004
 hill-0.004
 directions-0.004
 direction-0.004
Token smells
Token ID25760
Feature activation-7.60e-02

Pos logit contributions
hire+0.021
 Princess+0.018
 Queen+0.018
ocket+0.018
 robber+0.018
Neg logit contributions
bars-0.013
 hill-0.013
 outage-0.013
 directions-0.012
 direction-0.012
Token good
Token ID922
Feature activation-2.65e-02

Pos logit contributions
hire+0.178
 robber+0.166
ocket+0.165
 Queen+0.163
adies+0.159
Neg logit contributions
 directions-0.061
 direction-0.060
 way-0.060
 hill-0.059
 kay-0.057
Token.
Token ID13
Feature activation+5.36e-04

Pos logit contributions
hire+0.066
 Princess+0.059
 Queen+0.059
ocket+0.059
 robber+0.058
Neg logit contributions
 outage-0.018
 hill-0.018
bars-0.018
 directions-0.018
 blowing-0.018
Token
Token ID198
Feature activation+1.66e-02

Pos logit contributions
bars+0.001
 outage+0.001
 hill+0.001
 directions+0.001
 direction+0.001
Neg logit contributions
hire-0.001
 Princess-0.001
 Queen-0.001
ocket-0.001
 robber-0.001
Token
Token ID198
Feature activation-1.64e-02

Pos logit contributions
bars+0.022
 outage+0.021
spe+0.021
bike+0.020
 directions+0.020
Neg logit contributions
hire-0.030
 Princess-0.028
adies-0.028
ily-0.027
 Queen-0.026
TokenOne
Token ID3198
Feature activation+2.20e-02

Pos logit contributions
hire+0.030
 Princess+0.027
 Queen+0.027
ocket+0.027
 robber+0.026
Neg logit contributions
bars-0.023
 hill-0.022
 outage-0.022
 directions-0.021
 direction-0.021
Token.
Token ID13
Feature activation+4.88e-03

Pos logit contributions
 outage+0.010
 hill+0.010
bars+0.010
 direction+0.010
 directions+0.010
Neg logit contributions
hire-0.013
 Princess-0.012
 Queen-0.012
ocket-0.011
 robber-0.011
Token They
Token ID1119
Feature activation-1.09e-02

Pos logit contributions
bars+0.006
 hill+0.005
 outage+0.005
 direction+0.005
 directions+0.005
Neg logit contributions
hire-0.010
ocket-0.009
 Queen-0.009
 Princess-0.009
adies-0.009
Token would
Token ID561
Feature activation+1.59e-02

Pos logit contributions
hire+0.022
 Princess+0.020
ocket+0.020
adies+0.019
 Queen+0.019
Neg logit contributions
 hill-0.013
bars-0.013
spe-0.012
 kay-0.012
 outage-0.012
Token build
Token ID1382
Feature activation+1.19e-03

Pos logit contributions
bars+0.019
 outage+0.019
 hill+0.019
 directions+0.018
 direction+0.017
Neg logit contributions
hire-0.032
 Princess-0.029
 Queen-0.028
ocket-0.028
adies-0.027
Token sand
Token ID6450
Feature activation-3.01e-02

Pos logit contributions
 hill+0.001
 directions+0.001
 direction+0.001
 way+0.001
bars+0.001
Neg logit contributions
hire-0.002
ocket-0.002
 Queen-0.002
 Princess-0.002
adies-0.002
Tokencast
Token ID2701
Feature activation-7.56e-02

Pos logit contributions
hire+0.057
 Princess+0.052
 Queen+0.051
ocket+0.050
 robber+0.049
Neg logit contributions
 hill-0.041
bars-0.040
 outage-0.038
 directions-0.038
 blowing-0.037
Tokenles
Token ID829
Feature activation-1.15e-02

Pos logit contributions
hire+0.131
 Queen+0.122
 Princess+0.120
ocket+0.116
 robber+0.111
Neg logit contributions
bars-0.113
 hill-0.109
s-0.105
 blowing-0.101
 kay-0.099
Token and
Token ID290
Feature activation-1.64e-02

Pos logit contributions
hire+0.025
 Princess+0.023
 Queen+0.023
ocket+0.022
 robber+0.022
Neg logit contributions
bars-0.011
 outage-0.011
 hill-0.011
 directions-0.010
 direction-0.010
Token play
Token ID711
Feature activation-1.80e-02

Pos logit contributions
hire+0.031
 Princess+0.028
 Queen+0.027
ocket+0.027
 robber+0.026
Neg logit contributions
bars-0.021
 outage-0.021
 hill-0.021
 directions-0.020
 direction-0.020
Token in
Token ID287
Feature activation+3.77e-03

Pos logit contributions
hire+0.040
 Queen+0.035
ocket+0.035
 Princess+0.035
 robber+0.034
Neg logit contributions
 hill-0.020
 directions-0.017
 outage-0.017
 kay-0.017
 direction-0.017
Token the
Token ID262
Feature activation-1.77e-02

Pos logit contributions
bars+0.006
 outage+0.006
umbling+0.006
ney+0.005
bike+0.005
Neg logit contributions
hire-0.006
 Princess-0.006
 Queen-0.005
 mushrooms-0.005
 Prince-0.005
Token the
Token ID262
Feature activation-1.69e-02

Pos logit contributions
 outage+0.026
bars+0.025
ney+0.024
spe+0.024
bike+0.023
Neg logit contributions
hire-0.024
 Princess-0.023
 Queen-0.022
adies-0.021
 mushrooms-0.021
Token ocean
Token ID9151
Feature activation-1.14e-02

Pos logit contributions
hire+0.025
 Princess+0.023
 Queen+0.022
adies+0.021
 Prince+0.020
Neg logit contributions
 outage-0.030
bars-0.028
spe-0.027
 blowing-0.027
ney-0.026
Token and
Token ID290
Feature activation-6.07e-03

Pos logit contributions
hire+0.021
 Princess+0.019
 Queen+0.019
adies+0.018
ocket+0.018
Neg logit contributions
 outage-0.015
bars-0.015
spe-0.014
 directions-0.014
 direction-0.013
Token build
Token ID1382
Feature activation-2.67e-03

Pos logit contributions
hire+0.011
 Princess+0.010
 Queen+0.010
ocket+0.009
 robber+0.009
Neg logit contributions
 outage-0.009
bars-0.009
 directions-0.008
 hill-0.008
 direction-0.008
Token sand
Token ID6450
Feature activation-2.44e-02

Pos logit contributions
hire+0.005
 robber+0.005
 Queen+0.005
ocket+0.005
 Princess+0.005
Neg logit contributions
 directions-0.003
 hill-0.003
 direction-0.003
 kay-0.003
 way-0.003
Tokencast
Token ID2701
Feature activation-7.54e-02

Pos logit contributions
hire+0.041
 Princess+0.036
 Queen+0.036
ocket+0.035
 robber+0.035
Neg logit contributions
 hill-0.039
bars-0.037
 outage-0.036
 directions-0.036
 kay-0.035
Tokenles
Token ID829
Feature activation-1.42e-02

Pos logit contributions
hire+0.131
 Queen+0.124
 Princess+0.122
 robber+0.115
ocket+0.115
Neg logit contributions
bars-0.110
 hill-0.105
s-0.104
 kay-0.098
 blowing-0.097
Token.
Token ID13
Feature activation+1.28e-02

Pos logit contributions
hire+0.034
 Princess+0.031
 Queen+0.031
 robber+0.030
ocket+0.030
Neg logit contributions
 hill-0.012
bars-0.011
 outage-0.011
 directions-0.011
 blowing-0.011
Token
Token ID198
Feature activation+2.00e-02

Pos logit contributions
 outage+0.012
bars+0.012
 hill+0.011
 directions+0.011
spe+0.011
Neg logit contributions
hire-0.029
 Princess-0.027
 Queen-0.027
ocket-0.025
adies-0.025
Token
Token ID198
Feature activation-1.25e-02

Pos logit contributions
bars+0.031
spe+0.028
bike+0.028
 outage+0.028
 hill+0.027
Neg logit contributions
hire-0.032
 Princess-0.032
ily-0.030
 Queen-0.029
adies-0.029
TokenWhen
Token ID2215
Feature activation+5.17e-03

Pos logit contributions
hire+0.024
ocket+0.022
 Queen+0.022
 Princess+0.022
 robber+0.021
Neg logit contributions
 hill-0.016
 outage-0.015
bars-0.015
 direction-0.015
 directions-0.015
Token the
Token ID262
Feature activation-1.54e-02

Pos logit contributions
 outage+0.028
bars+0.027
spe+0.025
ney+0.024
umbling+0.024
Neg logit contributions
hire-0.025
 Princess-0.023
 Queen-0.022
 mushrooms-0.021
adies-0.021
Token sand
Token ID6450
Feature activation-1.77e-02

Pos logit contributions
hire+0.021
 Princess+0.018
 Queen+0.017
adies+0.016
 mushrooms+0.016
Neg logit contributions
 outage-0.030
bars-0.029
spe-0.027
 blowing-0.027
bike-0.026
Token and
Token ID290
Feature activation-5.44e-03

Pos logit contributions
hire+0.039
 Princess+0.035
 Queen+0.034
ocket+0.034
 robber+0.033
Neg logit contributions
bars-0.018
 outage-0.018
 hill-0.018
 directions-0.017
 direction-0.017
Token building
Token ID2615
Feature activation-3.74e-03

Pos logit contributions
hire+0.009
 Princess+0.008
 Queen+0.008
 mushrooms+0.007
adies+0.007
Neg logit contributions
 outage-0.009
bars-0.009
spe-0.008
ney-0.008
 direction-0.008
Token sand
Token ID6450
Feature activation-2.58e-02

Pos logit contributions
hire+0.007
 Queen+0.006
 Princess+0.006
ocket+0.006
 robber+0.006
Neg logit contributions
 hill-0.005
 directions-0.005
bars-0.005
 direction-0.005
 outage-0.005
Tokencast
Token ID2701
Feature activation-7.53e-02

Pos logit contributions
hire+0.045
 Princess+0.041
 Queen+0.040
 robber+0.039
ocket+0.039
Neg logit contributions
 hill-0.039
bars-0.037
 directions-0.036
 outage-0.036
 blowing-0.035
Tokenles
Token ID829
Feature activation-1.43e-02

Pos logit contributions
hire+0.132
 Queen+0.127
 Princess+0.125
 robber+0.118
ocket+0.118
Neg logit contributions
bars-0.107
s-0.103
 hill-0.101
 blowing-0.094
 kay-0.093
Token.
Token ID13
Feature activation+7.53e-03

Pos logit contributions
hire+0.033
 Princess+0.030
 Queen+0.029
ocket+0.029
 robber+0.028
Neg logit contributions
 hill-0.013
bars-0.013
 outage-0.013
 directions-0.012
 direction-0.012
Token
Token ID220
Feature activation+3.57e-02

Pos logit contributions
 outage+0.012
bars+0.011
 directions+0.010
spe+0.010
 blowing+0.010
Neg logit contributions
hire-0.012
 Princess-0.012
 Queen-0.011
 Pink-0.011
 robber-0.011
Token
Token ID198
Feature activation+1.14e-02

Pos logit contributions
 outage+0.044
 hill+0.043
bars+0.043
spe+0.040
 blowing+0.040
Neg logit contributions
hire-0.069
 Princess-0.062
ily-0.059
 Queen-0.059
adies-0.059
Token
Token ID198
Feature activation-2.10e-02

Pos logit contributions
bars+0.018
 outage+0.018
bike+0.017
spe+0.016
 directions+0.016
Neg logit contributions
 Princess-0.016
hire-0.016
ily-0.015
 Queen-0.014
adies-0.014
Token She
Token ID1375
Feature activation-1.24e-02

Pos logit contributions
hire+0.023
ocket+0.020
 Queen+0.020
 Princess+0.020
 robber+0.019
Neg logit contributions
bars-0.012
 outage-0.012
 hill-0.012
 directions-0.012
 direction-0.012
Token picks
Token ID11103
Feature activation-3.23e-02

Pos logit contributions
hire+0.027
 Queen+0.024
 Princess+0.024
ocket+0.023
 robber+0.023
Neg logit contributions
bars-0.013
 hill-0.013
 outage-0.013
 direction-0.012
 directions-0.012
Token the
Token ID262
Feature activation+4.43e-03

Pos logit contributions
hire+0.080
ocket+0.075
 robber+0.075
vell+0.073
adies+0.073
Neg logit contributions
 directions-0.020
bars-0.019
 direction-0.019
 blowing-0.018
 hill-0.018
Token flower
Token ID15061
Feature activation-6.32e-04

Pos logit contributions
 hill+0.006
 direction+0.006
 way+0.006
 kay+0.006
 outage+0.006
Neg logit contributions
hire-0.008
adies-0.007
 Princess-0.007
 Queen-0.007
ocket-0.007
Token and
Token ID290
Feature activation-3.48e-02

Pos logit contributions
hire+0.001
 Princess+0.001
 Queen+0.001
ocket+0.001
 robber+0.001
Neg logit contributions
 outage-0.001
bars-0.001
 hill-0.001
 directions-0.001
 direction-0.001
Token smells
Token ID25760
Feature activation-7.52e-02

Pos logit contributions
hire+0.076
ocket+0.068
 Princess+0.066
 Queen+0.066
 robber+0.066
Neg logit contributions
bars-0.037
 hill-0.037
 direction-0.034
 directions-0.034
 kay-0.033
Token it
Token ID340
Feature activation-8.09e-03

Pos logit contributions
hire+0.161
 robber+0.152
adies+0.146
ocket+0.145
 Queen+0.144
Neg logit contributions
 directions-0.073
bars-0.071
 direction-0.070
 kay-0.070
 blowing-0.068
Token.
Token ID13
Feature activation-1.10e-02

Pos logit contributions
hire+0.019
 Princess+0.017
 Queen+0.017
 robber+0.017
ocket+0.017
Neg logit contributions
bars-0.007
 hill-0.007
 outage-0.007
 directions-0.007
 direction-0.007
Token It
Token ID632
Feature activation-2.22e-04

Pos logit contributions
hire+0.023
ocket+0.020
 Queen+0.019
 Princess+0.019
 robber+0.019
Neg logit contributions
bars-0.013
 hill-0.013
 outage-0.013
 directions-0.012
 direction-0.012
Token has
Token ID468
Feature activation-2.10e-02

Pos logit contributions
hire+0.001
 Princess+0.001
 Queen+0.001
ocket+0.001
 robber+0.001
Neg logit contributions
 hill-0.000
bars-0.000
 outage-0.000
 directions-0.000
 direction-0.000
Token a
Token ID257
Feature activation-1.19e-02

Pos logit contributions
hire+0.040
 robber+0.035
 Princess+0.034
ocket+0.034
adies+0.034
Neg logit contributions
 directions-0.028
 direction-0.027
 way-0.027
 outage-0.026
bars-0.025
Token the
Token ID262
Feature activation-1.85e-02

Pos logit contributions
hire+0.005
 Princess+0.004
 Queen+0.004
 robber+0.004
 mushrooms+0.004
Neg logit contributions
 outage-0.004
bars-0.004
 direction-0.003
spe-0.003
 directions-0.003
Token beach
Token ID10481
Feature activation-1.77e-02

Pos logit contributions
hire+0.035
 Princess+0.032
 Queen+0.031
 robber+0.030
ocket+0.030
Neg logit contributions
 outage-0.025
bars-0.024
spe-0.023
 direction-0.023
 directions-0.022
Token and
Token ID290
Feature activation-7.90e-03

Pos logit contributions
hire+0.037
 Princess+0.034
 Queen+0.033
ocket+0.033
 robber+0.032
Neg logit contributions
 outage-0.020
bars-0.020
 hill-0.019
 directions-0.018
 direction-0.018
Token build
Token ID1382
Feature activation-1.56e-03

Pos logit contributions
hire+0.014
 Princess+0.013
 Queen+0.013
ocket+0.012
 robber+0.012
Neg logit contributions
 outage-0.011
bars-0.011
 hill-0.011
 directions-0.011
 direction-0.010
Token sand
Token ID6450
Feature activation-2.34e-02

Pos logit contributions
hire+0.003
 robber+0.003
 Queen+0.003
ocket+0.003
 Princess+0.003
Neg logit contributions
 directions-0.002
 hill-0.002
 direction-0.002
 kay-0.002
 room-0.002
Tokencast
Token ID2701
Feature activation-7.52e-02

Pos logit contributions
hire+0.042
 Princess+0.038
 Queen+0.037
ocket+0.037
 robber+0.037
Neg logit contributions
 hill-0.035
bars-0.032
 directions-0.031
 kay-0.030
 outage-0.030
Tokenles
Token ID829
Feature activation-1.27e-02

Pos logit contributions
hire+0.134
 Queen+0.130
 Princess+0.127
 robber+0.121
ocket+0.120
Neg logit contributions
bars-0.104
s-0.101
 hill-0.097
 blowing-0.091
 kay-0.089
Token.
Token ID13
Feature activation+8.60e-04

Pos logit contributions
hire+0.031
 Princess+0.028
 Queen+0.028
 robber+0.028
ocket+0.027
Neg logit contributions
 hill-0.010
bars-0.009
 outage-0.009
 directions-0.009
 blowing-0.009
Token But
Token ID887
Feature activation+3.20e-02

Pos logit contributions
 outage+0.001
bars+0.001
 hill+0.001
 directions+0.001
spe+0.001
Neg logit contributions
hire-0.001
 Princess-0.001
 Queen-0.001
ocket-0.001
 robber-0.001
Token one
Token ID530
Feature activation+3.10e-02

Pos logit contributions
bars+0.050
 hill+0.049
 outage+0.049
 direction+0.047
 directions+0.047
Neg logit contributions
hire-0.054
 Princess-0.046
 Queen-0.046
ocket-0.045
 robber-0.044
Token day
Token ID1110
Feature activation-1.88e-02

Pos logit contributions
bars+0.027
 outage+0.023
 directions+0.022
 hill+0.020
bike+0.020
Neg logit contributions
hire-0.075
 Princess-0.071
 Queen-0.070
ocket-0.067
 robber-0.067
Token on
Token ID319
Feature activation+1.56e-02

Pos logit contributions
 outage+0.025
bars+0.023
spe+0.022
 directions+0.022
 hill+0.021
Neg logit contributions
hire-0.025
 Princess-0.023
 robber-0.022
 Queen-0.022
adies-0.021
Token,
Token ID11
Feature activation+1.43e-03

Pos logit contributions
 outage+0.018
bars+0.018
 hill+0.017
 directions+0.017
 direction+0.016
Neg logit contributions
hire-0.032
 Princess-0.030
 Queen-0.029
ocket-0.028
 robber-0.028
Token they
Token ID484
Feature activation-2.14e-03

Pos logit contributions
 hill+0.002
 way+0.002
 day+0.002
 blowing+0.002
spe+0.002
Neg logit contributions
hire-0.002
ocket-0.002
adies-0.002
vell-0.002
ured-0.002
Token built
Token ID3170
Feature activation+1.36e-03

Pos logit contributions
hire+0.004
 Princess+0.004
adies+0.004
ocket+0.004
 Queen+0.004
Neg logit contributions
 hill-0.003
bars-0.003
 outage-0.002
 blowing-0.002
 directions-0.002
Token sand
Token ID6450
Feature activation-2.94e-02

Pos logit contributions
 hill+0.002
 directions+0.002
 direction+0.002
 way+0.002
 kay+0.001
Neg logit contributions
hire-0.003
ocket-0.003
adies-0.002
 robber-0.002
 Queen-0.002
Tokencast
Token ID2701
Feature activation-7.49e-02

Pos logit contributions
hire+0.058
 Princess+0.051
ocket+0.051
 robber+0.050
 Queen+0.050
Neg logit contributions
 hill-0.039
bars-0.035
 directions-0.035
ney-0.034
spe-0.034
Tokenles
Token ID829
Feature activation-2.36e-02

Pos logit contributions
hire+0.132
 Queen+0.124
 Princess+0.121
ocket+0.119
 robber+0.112
Neg logit contributions
bars-0.107
s-0.104
 hill-0.101
 blowing-0.096
 directions-0.094
Token together
Token ID1978
Feature activation-1.05e-02

Pos logit contributions
hire+0.053
 Princess+0.048
 Queen+0.047
ocket+0.047
 robber+0.046
Neg logit contributions
bars-0.022
 hill-0.022
 outage-0.022
 directions-0.021
 direction-0.021
Token every
Token ID790
Feature activation+5.31e-02

Pos logit contributions
hire+0.021
 Princess+0.019
 Queen+0.019
ocket+0.019
 robber+0.018
Neg logit contributions
 outage-0.013
bars-0.013
 hill-0.012
 directions-0.011
 direction-0.011
Token weekend
Token ID5041
Feature activation-1.22e-02

Pos logit contributions
bars+0.061
 outage+0.060
bike+0.049
umbling+0.048
ney+0.047
Neg logit contributions
hire-0.107
 Princess-0.105
 Queen-0.104
 mushrooms-0.100
 queen-0.100
Token.
Token ID13
Feature activation+1.52e-02

Pos logit contributions
hire+0.027
 Princess+0.024
 Queen+0.024
ocket+0.023
 robber+0.023
Neg logit contributions
bars-0.012
 outage-0.012
 hill-0.012
 directions-0.012
 direction-0.011
Token.
Token ID13
Feature activation+4.61e-03

Pos logit contributions
hire+0.037
 Princess+0.033
 Queen+0.033
ocket+0.032
 robber+0.031
Neg logit contributions
 outage-0.020
bars-0.020
 hill-0.019
 directions-0.019
 direction-0.018
Token She
Token ID1375
Feature activation-6.37e-03

Pos logit contributions
bars+0.005
 hill+0.005
 direction+0.004
 outage+0.004
spe+0.004
Neg logit contributions
hire-0.010
ocket-0.009
vell-0.009
 Queen-0.009
ils-0.009
Token would
Token ID561
Feature activation+1.11e-02

Pos logit contributions
hire+0.012
 Princess+0.011
 Queen+0.011
ocket+0.011
 robber+0.010
Neg logit contributions
bars-0.008
 outage-0.008
 hill-0.008
 directions-0.008
 direction-0.008
Token build
Token ID1382
Feature activation+5.38e-05

Pos logit contributions
 hill+0.014
bars+0.013
 outage+0.013
 directions+0.012
 direction+0.012
Neg logit contributions
hire-0.022
 Princess-0.020
ocket-0.020
 Queen-0.019
adies-0.019
Token sand
Token ID6450
Feature activation-2.94e-02

Pos logit contributions
 directions+0.000
 direction+0.000
 hill+0.000
 way+0.000
 room+0.000
Neg logit contributions
hire-0.000
 robber-0.000
ocket-0.000
 Queen-0.000
pper-0.000
Tokencast
Token ID2701
Feature activation-7.49e-02

Pos logit contributions
hire+0.056
 Princess+0.051
 Queen+0.050
ocket+0.050
 robber+0.049
Neg logit contributions
 hill-0.041
bars-0.038
 directions-0.037
 kay-0.036
 blowing-0.036
Tokenles
Token ID829
Feature activation-1.56e-02

Pos logit contributions
hire+0.134
 Queen+0.127
 Princess+0.123
ocket+0.119
 robber+0.115
Neg logit contributions
bars-0.107
 hill-0.104
s-0.102
 blowing-0.094
 kay-0.091
Token and
Token ID290
Feature activation-1.82e-02

Pos logit contributions
hire+0.037
 Princess+0.033
 Queen+0.033
 robber+0.033
ocket+0.032
Neg logit contributions
 hill-0.013
bars-0.013
 outage-0.013
 blowing-0.012
 directions-0.012
Token collect
Token ID2824
Feature activation-4.02e-02

Pos logit contributions
hire+0.034
 Princess+0.030
 Queen+0.030
ocket+0.029
 robber+0.029
Neg logit contributions
bars-0.024
 outage-0.024
 hill-0.024
 directions-0.023
 direction-0.023
Token pretty
Token ID2495
Feature activation-3.91e-02

Pos logit contributions
hire+0.082
 robber+0.081
 Queen+0.076
pper+0.074
 Princess+0.073
Neg logit contributions
 directions-0.048
 kay-0.047
 direction-0.046
 hill-0.043
 way-0.042
Token seas
Token ID21547
Feature activation-4.62e-02

Pos logit contributions
hire+0.068
 robber+0.061
 Queen+0.061
 Princess+0.060
ocket+0.059
Neg logit contributions
 hill-0.058
 directions-0.058
bars-0.056
 direction-0.056
 kay-0.055
Token play
Token ID711
Feature activation-1.38e-02

Pos logit contributions
hire+0.011
 Princess+0.010
 Queen+0.010
ocket+0.010
 robber+0.010
Neg logit contributions
bars-0.007
 outage-0.007
 hill-0.006
 directions-0.006
 direction-0.006
Token.
Token ID13
Feature activation+8.14e-03

Pos logit contributions
hire+0.030
 Princess+0.027
 Queen+0.026
ocket+0.026
 robber+0.025
Neg logit contributions
 outage-0.015
bars-0.015
 hill-0.015
 directions-0.014
 direction-0.014
Token They
Token ID1119
Feature activation-1.13e-02

Pos logit contributions
bars+0.011
 outage+0.011
 hill+0.011
 direction+0.011
 directions+0.011
Neg logit contributions
hire-0.015
ocket-0.013
 Princess-0.013
 Queen-0.013
 robber-0.012
Token built
Token ID3170
Feature activation-6.26e-03

Pos logit contributions
hire+0.022
 Princess+0.020
 Queen+0.020
ocket+0.019
 robber+0.019
Neg logit contributions
bars-0.014
 hill-0.014
 outage-0.013
 directions-0.013
 blowing-0.013
Token sand
Token ID6450
Feature activation-3.16e-02

Pos logit contributions
hire+0.011
ocket+0.010
 Queen+0.010
 robber+0.010
 Princess+0.010
Neg logit contributions
 hill-0.009
 directions-0.009
 direction-0.009
bars-0.008
 way-0.008
Tokencast
Token ID2701
Feature activation-7.48e-02

Pos logit contributions
hire+0.059
 Princess+0.053
ocket+0.052
 Queen+0.052
 robber+0.051
Neg logit contributions
 hill-0.046
bars-0.042
 directions-0.041
 blowing-0.040
 outage-0.040
Tokenles
Token ID829
Feature activation-2.29e-02

Pos logit contributions
hire+0.134
 Queen+0.125
 Princess+0.123
ocket+0.120
 robber+0.115
Neg logit contributions
bars-0.108
 hill-0.105
s-0.103
 blowing-0.097
 outage-0.095
Token and
Token ID290
Feature activation-1.39e-02

Pos logit contributions
hire+0.049
 Princess+0.045
 Queen+0.044
ocket+0.043
 robber+0.042
Neg logit contributions
bars-0.024
 outage-0.024
 hill-0.023
 directions-0.022
 direction-0.022
Token played
Token ID2826
Feature activation-1.61e-02

Pos logit contributions
hire+0.026
 Princess+0.023
ocket+0.023
 Queen+0.023
 robber+0.022
Neg logit contributions
bars-0.019
 hill-0.019
 outage-0.018
 directions-0.018
 blowing-0.017
Token tag
Token ID7621
Feature activation-3.03e-02

Pos logit contributions
hire+0.033
 Queen+0.029
 Princess+0.029
ocket+0.029
 robber+0.028
Neg logit contributions
 hill-0.019
 directions-0.018
 outage-0.018
 direction-0.018
bars-0.017
Token.
Token ID13
Feature activation+6.14e-04

Pos logit contributions
hire+0.069
 Princess+0.062
 Queen+0.062
ocket+0.061
 robber+0.060
Neg logit contributions
 hill-0.029
 outage-0.028
bars-0.028
 directions-0.027
 direction-0.027
Token She
Token ID1375
Feature activation-1.21e-02

Pos logit contributions
hire+0.017
ocket+0.015
 Queen+0.015
 Princess+0.015
 robber+0.014
Neg logit contributions
bars-0.016
 hill-0.015
 outage-0.015
 directions-0.015
 direction-0.015
Token liked
Token ID8288
Feature activation-5.96e-04

Pos logit contributions
hire+0.018
 Princess+0.016
ns+0.016
adies+0.016
 Queen+0.015
Neg logit contributions
 outage-0.020
bars-0.019
 directions-0.019
 direction-0.019
umbling-0.019
Token to
Token ID284
Feature activation+3.07e-03

Pos logit contributions
hire+0.001
 Princess+0.001
 Queen+0.001
 mushrooms+0.001
 robber+0.001
Neg logit contributions
 outage-0.001
bars-0.001
ney-0.001
 direction-0.001
spe-0.001
Token build
Token ID1382
Feature activation-2.03e-03

Pos logit contributions
 hill+0.004
bars+0.004
 outage+0.004
 directions+0.003
 direction+0.003
Neg logit contributions
hire-0.006
 Princess-0.006
 Queen-0.006
ocket-0.005
 robber-0.005
Token sand
Token ID6450
Feature activation-2.68e-02

Pos logit contributions
hire+0.005
ocket+0.004
 Queen+0.004
 robber+0.004
pper+0.004
Neg logit contributions
 directions-0.002
 hill-0.002
 direction-0.002
 kay-0.002
 way-0.002
Tokencast
Token ID2701
Feature activation-7.48e-02

Pos logit contributions
hire+0.051
 Princess+0.047
ocket+0.047
 Queen+0.046
 robber+0.045
Neg logit contributions
 hill-0.038
bars-0.033
s-0.033
 kay-0.032
 directions-0.032
Tokenles
Token ID829
Feature activation-1.40e-02

Pos logit contributions
hire+0.135
 Queen+0.133
 Princess+0.129
ocket+0.123
 robber+0.121
Neg logit contributions
s-0.100
bars-0.100
 hill-0.093
 blowing-0.087
 kay-0.086
Token and
Token ID290
Feature activation-1.79e-02

Pos logit contributions
hire+0.033
 Princess+0.030
 Queen+0.030
ocket+0.029
 robber+0.029
Neg logit contributions
 hill-0.012
bars-0.012
 outage-0.012
 blowing-0.011
 directions-0.011
Token collect
Token ID2824
Feature activation-3.36e-02

Pos logit contributions
hire+0.031
 Princess+0.028
 Queen+0.027
ocket+0.026
 robber+0.026
Neg logit contributions
 outage-0.027
bars-0.026
 hill-0.025
 direction-0.025
 directions-0.025
Token seas
Token ID21547
Feature activation-4.31e-02

Pos logit contributions
hire+0.067
 robber+0.067
pper+0.065
 Queen+0.065
ocket+0.063
Neg logit contributions
 kay-0.039
 directions-0.039
 direction-0.037
 hill-0.035
s-0.034
Tokenhell
Token ID12758
Feature activation-6.64e-02

Pos logit contributions
 Queen+0.064
hire+0.064
 Princess+0.062
 robber+0.059
 Prince+0.057
Neg logit contributions
s-0.074
 story-0.068
 outage-0.065
 hill-0.065
 directions-0.065
Token the
Token ID262
Feature activation-1.22e-02

Pos logit contributions
 outage+0.036
bars+0.035
spe+0.034
 direction+0.032
umbling+0.032
Neg logit contributions
hire-0.035
 Princess-0.035
 Queen-0.033
 Prince-0.030
 queen-0.030
Token beach
Token ID10481
Feature activation-1.29e-02

Pos logit contributions
hire+0.025
 Princess+0.023
 Queen+0.023
 robber+0.022
ocket+0.022
Neg logit contributions
 outage-0.014
bars-0.014
 directions-0.013
 direction-0.013
 hill-0.013
Token and
Token ID290
Feature activation-7.26e-05

Pos logit contributions
hire+0.026
 Princess+0.023
 Queen+0.023
ocket+0.022
adies+0.022
Neg logit contributions
 outage-0.016
bars-0.016
 directions-0.015
 hill-0.015
 direction-0.014
Token building
Token ID2615
Feature activation-1.67e-03

Pos logit contributions
hire+0.000
 Princess+0.000
 Queen+0.000
adies+0.000
ocket+0.000
Neg logit contributions
 outage-0.000
bars-0.000
 directions-0.000
 direction-0.000
 hill-0.000
Token sand
Token ID6450
Feature activation-2.42e-02

Pos logit contributions
hire+0.003
 robber+0.003
 Queen+0.003
ocket+0.003
 Princess+0.003
Neg logit contributions
 directions-0.002
 hill-0.002
bars-0.002
 direction-0.002
 kay-0.002
Tokencast
Token ID2701
Feature activation-7.44e-02

Pos logit contributions
hire+0.045
 Princess+0.041
 Queen+0.040
 robber+0.039
ocket+0.039
Neg logit contributions
 hill-0.034
bars-0.033
 outage-0.031
 directions-0.031
 direction-0.031
Tokenles
Token ID829
Feature activation-1.43e-02

Pos logit contributions
hire+0.132
 Queen+0.127
 Princess+0.124
 robber+0.118
ocket+0.117
Neg logit contributions
bars-0.105
s-0.101
 hill-0.098
 blowing-0.092
 direction-0.092
Token.
Token ID13
Feature activation+2.19e-03

Pos logit contributions
hire+0.034
 Princess+0.031
 Queen+0.031
 robber+0.030
ocket+0.030
Neg logit contributions
 hill-0.012
bars-0.012
 outage-0.011
 directions-0.011
 blowing-0.011
Token One
Token ID1881
Feature activation+2.54e-02

Pos logit contributions
 outage+0.003
bars+0.003
 hill+0.003
 directions+0.003
 direction+0.003
Neg logit contributions
hire-0.004
 Princess-0.003
 Queen-0.003
ocket-0.003
 robber-0.003
Token day
Token ID1110
Feature activation-1.98e-02

Pos logit contributions
bars+0.031
 outage+0.028
spe+0.025
bike+0.025
 directions+0.025
Neg logit contributions
hire-0.053
 Princess-0.047
adies-0.046
ily-0.046
 Queen-0.045
Token,
Token ID11
Feature activation-2.63e-02

Pos logit contributions
hire+0.047
 Queen+0.042
 Princess+0.042
ocket+0.042
 robber+0.041
Neg logit contributions
bars-0.017
 outage-0.016
 hill-0.016
 direction-0.015
 blowing-0.015
Token beach
Token ID10481
Feature activation-1.15e-02

Pos logit contributions
hire+0.060
ocket+0.054
adies+0.051
izards+0.051
imes+0.050
Neg logit contributions
 hill-0.024
 kay-0.021
 direction-0.020
 way-0.020
 directions-0.020
Token.
Token ID13
Feature activation-1.17e-02

Pos logit contributions
hire+0.021
 Princess+0.020
 Queen+0.019
ocket+0.018
adies+0.018
Neg logit contributions
 outage-0.015
bars-0.015
 directions-0.013
 direction-0.013
 motor-0.013
Token They
Token ID1119
Feature activation-1.25e-02

Pos logit contributions
hire+0.026
ocket+0.023
vell+0.022
pper+0.021
 Queen+0.021
Neg logit contributions
 hill-0.013
bars-0.013
 outage-0.012
 direction-0.012
 directions-0.011
Token made
Token ID925
Feature activation+4.96e-03

Pos logit contributions
hire+0.023
 Princess+0.021
 Queen+0.020
ocket+0.020
 robber+0.020
Neg logit contributions
 hill-0.018
bars-0.017
 kay-0.017
 blowing-0.016
 direction-0.016
Token sand
Token ID6450
Feature activation-2.27e-02

Pos logit contributions
 directions+0.007
 hill+0.007
bars+0.007
 direction+0.007
 outage+0.007
Neg logit contributions
hire-0.009
 Princess-0.008
ocket-0.008
 Queen-0.007
 robber-0.007
Tokencast
Token ID2701
Feature activation-7.43e-02

Pos logit contributions
hire+0.039
 Princess+0.035
 Queen+0.034
ocket+0.034
 robber+0.033
Neg logit contributions
 hill-0.034
bars-0.034
 directions-0.033
 outage-0.033
 direction-0.032
Tokenles
Token ID829
Feature activation-1.75e-02

Pos logit contributions
hire+0.130
 Queen+0.124
 Princess+0.122
ocket+0.115
 robber+0.112
Neg logit contributions
bars-0.107
 hill-0.102
s-0.102
 blowing-0.096
 kay-0.094
Token and
Token ID290
Feature activation-1.99e-02

Pos logit contributions
hire+0.035
 Princess+0.031
 Queen+0.031
ocket+0.030
 robber+0.029
Neg logit contributions
bars-0.021
 outage-0.021
 hill-0.020
 directions-0.019
 direction-0.019
Token ran
Token ID4966
Feature activation-2.07e-02

Pos logit contributions
hire+0.037
 Princess+0.032
ocket+0.032
 Queen+0.031
 robber+0.031
Neg logit contributions
 hill-0.028
bars-0.027
 outage-0.026
 directions-0.026
 direction-0.026
Token through
Token ID832
Feature activation+7.50e-03

Pos logit contributions
hire+0.043
 Princess+0.038
 Queen+0.038
ocket+0.037
 robber+0.036
Neg logit contributions
bars-0.024
 outage-0.024
 hill-0.024
 directions-0.023
 direction-0.023
Token the
Token ID262
Feature activation-1.97e-02

Pos logit contributions
 outage+0.013
bars+0.012
ney+0.011
bike+0.011
umbling+0.011
Neg logit contributions
hire-0.012
 Princess-0.011
 Queen-0.010
 mushrooms-0.010
 wow-0.010
Token She
Token ID1375
Feature activation-8.41e-03

Pos logit contributions
hire+0.017
ocket+0.015
 Queen+0.015
 Princess+0.014
 robber+0.014
Neg logit contributions
bars-0.015
 hill-0.015
 outage-0.015
 direction-0.014
 directions-0.014
Token liked
Token ID8288
Feature activation+1.61e-03

Pos logit contributions
hire+0.012
 Princess+0.011
ns+0.011
adies+0.010
 robber+0.010
Neg logit contributions
 outage-0.014
bars-0.014
 directions-0.014
 direction-0.013
umbling-0.013
Token to
Token ID284
Feature activation+3.45e-03

Pos logit contributions
 outage+0.002
bars+0.002
ney+0.002
 direction+0.002
spe+0.002
Neg logit contributions
hire-0.003
 Princess-0.003
 Queen-0.003
 mushrooms-0.003
 robber-0.003
Token build
Token ID1382
Feature activation-2.46e-03

Pos logit contributions
 hill+0.004
 outage+0.004
bars+0.004
 directions+0.004
 direction+0.004
Neg logit contributions
hire-0.007
 Princess-0.006
 Queen-0.006
ocket-0.006
 robber-0.006
Token sand
Token ID6450
Feature activation-2.58e-02

Pos logit contributions
hire+0.005
 Queen+0.005
ocket+0.005
 robber+0.005
 Princess+0.005
Neg logit contributions
 directions-0.003
 hill-0.003
 direction-0.002
 kay-0.002
 way-0.002
Tokencast
Token ID2701
Feature activation-7.42e-02

Pos logit contributions
hire+0.049
 Princess+0.045
ocket+0.044
 Queen+0.044
 robber+0.043
Neg logit contributions
 hill-0.037
s-0.032
bars-0.032
 kay-0.031
 directions-0.031
Tokenles
Token ID829
Feature activation-1.37e-02

Pos logit contributions
hire+0.132
 Queen+0.129
 Princess+0.126
ocket+0.119
 robber+0.117
Neg logit contributions
bars-0.103
s-0.102
 hill-0.097
 blowing-0.091
 kay-0.090
Token and
Token ID290
Feature activation-1.71e-02

Pos logit contributions
hire+0.032
 Princess+0.029
 Queen+0.029
ocket+0.029
 robber+0.028
Neg logit contributions
 hill-0.012
bars-0.011
 outage-0.011
 blowing-0.011
 directions-0.010
Token collect
Token ID2824
Feature activation-3.33e-02

Pos logit contributions
hire+0.030
 Princess+0.027
 Queen+0.026
 robber+0.025
ocket+0.025
Neg logit contributions
 outage-0.026
bars-0.025
 direction-0.024
 hill-0.024
 directions-0.024
Token seas
Token ID21547
Feature activation-4.21e-02

Pos logit contributions
hire+0.067
 robber+0.067
 Queen+0.065
pper+0.064
 Princess+0.062
Neg logit contributions
 kay-0.039
 directions-0.039
 direction-0.037
 hill-0.035
s-0.034
Tokenhell
Token ID12758
Feature activation-6.53e-02

Pos logit contributions
hire+0.063
 Queen+0.063
 Princess+0.061
 robber+0.058
 Prince+0.056
Neg logit contributions
s-0.073
 story-0.065
 directions-0.063
 outage-0.063
bars-0.063
Token while
Token ID981
Feature activation+3.25e-02

Pos logit contributions
hire+0.042
 Princess+0.037
 Queen+0.036
 robber+0.035
ocket+0.035
Neg logit contributions
bars-0.043
 outage-0.043
 hill-0.041
 directions-0.040
 direction-0.040
Token,
Token ID11
Feature activation-1.59e-02

Pos logit contributions
 outage+0.040
bars+0.039
 hill+0.038
 directions+0.036
 direction+0.036
Neg logit contributions
hire-0.065
 Princess-0.059
 Queen-0.058
ocket-0.056
 robber-0.056
Token then
Token ID788
Feature activation+5.25e-04

Pos logit contributions
hire+0.032
ocket+0.029
 Queen+0.028
 Princess+0.027
adies+0.027
Neg logit contributions
 hill-0.020
 blowing-0.019
bars-0.019
 way-0.018
 direction-0.018
Token built
Token ID3170
Feature activation-2.25e-03

Pos logit contributions
 hill+0.001
bars+0.001
 direction+0.001
 kay+0.001
 directions+0.001
Neg logit contributions
hire-0.001
ocket-0.001
 Princess-0.001
 Queen-0.001
adies-0.001
Token sand
Token ID6450
Feature activation-2.85e-02

Pos logit contributions
hire+0.004
ocket+0.004
 Queen+0.004
 Princess+0.004
 robber+0.003
Neg logit contributions
 hill-0.003
 directions-0.003
 direction-0.003
bars-0.003
 kay-0.003
Tokencast
Token ID2701
Feature activation-7.40e-02

Pos logit contributions
hire+0.049
 Princess+0.044
 Queen+0.043
ocket+0.043
 robber+0.041
Neg logit contributions
 hill-0.044
bars-0.042
 directions-0.041
 outage-0.040
 blowing-0.040
Tokenles
Token ID829
Feature activation-2.07e-02

Pos logit contributions
hire+0.128
 Queen+0.121
 Princess+0.118
ocket+0.112
 robber+0.110
Neg logit contributions
bars-0.111
 hill-0.106
s-0.103
 blowing-0.100
 outage-0.099
Token with
Token ID351
Feature activation-3.23e-02

Pos logit contributions
hire+0.043
 Princess+0.039
 Queen+0.039
ocket+0.038
 robber+0.037
Neg logit contributions
 outage-0.023
bars-0.023
 hill-0.022
 directions-0.021
 direction-0.021
Token his
Token ID465
Feature activation-1.13e-02

Pos logit contributions
hire+0.066
ocket+0.058
 Princess+0.058
adies+0.058
 Queen+0.057
Neg logit contributions
 directions-0.039
 hill-0.038
bars-0.038
 direction-0.037
 kay-0.037
Token mom
Token ID1995
Feature activation+3.67e-05

Pos logit contributions
hire+0.022
 Princess+0.020
 Queen+0.020
ocket+0.019
 robber+0.019
Neg logit contributions
 outage-0.014
bars-0.014
 hill-0.013
 directions-0.013
 direction-0.012
Token.
Token ID13
Feature activation+2.48e-03

Pos logit contributions
 outage+0.000
bars+0.000
 hill+0.000
 directions+0.000
 direction+0.000
Neg logit contributions
hire-0.000
 Princess-0.000
 Queen-0.000
ocket-0.000
 Prince-0.000
Token bucket
Token ID19236
Feature activation+8.88e-03

Pos logit contributions
hire+0.024
 Princess+0.021
 Queen+0.021
ocket+0.020
 robber+0.019
Neg logit contributions
 outage-0.020
bars-0.020
 hill-0.019
 directions-0.018
 blowing-0.018
Token.
Token ID13
Feature activation+4.85e-03

Pos logit contributions
 outage+0.011
bars+0.011
 hill+0.010
bike+0.010
 directions+0.010
Neg logit contributions
hire-0.017
 Princess-0.015
 Queen-0.015
ocket-0.014
 mushrooms-0.014
Token She
Token ID1375
Feature activation-8.87e-03

Pos logit contributions
bars+0.007
 hill+0.007
 outage+0.007
 direction+0.006
 directions+0.006
Neg logit contributions
hire-0.009
ocket-0.008
 robber-0.008
 Princess-0.008
 Queen-0.008
Token made
Token ID925
Feature activation+3.06e-03

Pos logit contributions
hire+0.017
 Princess+0.015
 Queen+0.015
ocket+0.014
 robber+0.014
Neg logit contributions
 outage-0.012
bars-0.012
 hill-0.011
 directions-0.011
 direction-0.011
Token sand
Token ID6450
Feature activation-2.24e-02

Pos logit contributions
 directions+0.004
 blowing+0.004
 direction+0.004
bars+0.004
 way+0.004
Neg logit contributions
hire-0.006
 robber-0.005
ocket-0.005
 Queen-0.005
 Princess-0.005
Tokencast
Token ID2701
Feature activation-7.40e-02

Pos logit contributions
hire+0.042
 Princess+0.040
 robber+0.039
 Queen+0.039
ocket+0.039
Neg logit contributions
 hill-0.030
 directions-0.028
bars-0.027
 kay-0.027
 blowing-0.026
Tokenles
Token ID829
Feature activation-1.72e-02

Pos logit contributions
hire+0.134
 Queen+0.133
 Princess+0.129
 robber+0.121
ocket+0.120
Neg logit contributions
bars-0.102
s-0.098
 hill-0.095
 blowing-0.089
 directions-0.087
Token and
Token ID290
Feature activation-2.00e-02

Pos logit contributions
hire+0.039
 Princess+0.035
 Queen+0.035
ocket+0.034
 robber+0.034
Neg logit contributions
bars-0.016
 outage-0.016
 hill-0.016
 directions-0.015
 blowing-0.015
Token sand
Token ID6450
Feature activation-2.69e-02

Pos logit contributions
hire+0.037
 Princess+0.032
ocket+0.032
 Queen+0.032
 robber+0.032
Neg logit contributions
bars-0.027
 hill-0.027
 directions-0.026
 outage-0.026
 direction-0.026
Token pies
Token ID47199
Feature activation-4.72e-02

Pos logit contributions
hire+0.057
 Princess+0.053
 Queen+0.052
 robber+0.052
ocket+0.051
Neg logit contributions
 hill-0.030
bars-0.028
 directions-0.027
 blowing-0.027
 outage-0.026
Token.
Token ID13
Feature activation+2.16e-03

Pos logit contributions
hire+0.106
 Princess+0.097
 Queen+0.096
ocket+0.094
 robber+0.093
Neg logit contributions
bars-0.046
 hill-0.046
 outage-0.045
 directions-0.043
 blowing-0.042
Token played
Token ID2826
Feature activation-2.11e-03

Pos logit contributions
hire+0.022
 Princess+0.020
 Queen+0.019
ocket+0.019
 robber+0.019
Neg logit contributions
 outage-0.016
bars-0.016
 hill-0.015
 directions-0.015
 direction-0.015
Token in
Token ID287
Feature activation+9.25e-03

Pos logit contributions
hire+0.005
 Queen+0.004
 Princess+0.004
ocket+0.004
 robber+0.004
Neg logit contributions
 hill-0.002
 outage-0.002
 directions-0.002
bars-0.002
 direction-0.002
Token the
Token ID262
Feature activation-1.80e-02

Pos logit contributions
bars+0.015
 outage+0.015
umbling+0.013
spe+0.013
ney+0.013
Neg logit contributions
hire-0.015
 Princess-0.015
 Queen-0.014
 mushrooms-0.013
 robber-0.013
Token lovely
Token ID14081
Feature activation-1.91e-02

Pos logit contributions
hire+0.035
 Princess+0.033
 Queen+0.032
 robber+0.031
adies+0.031
Neg logit contributions
bars-0.023
 outage-0.023
spe-0.021
 blowing-0.020
 hill-0.019
Token tent
Token ID11105
Feature activation-1.33e-02

Pos logit contributions
hire+0.033
 Princess+0.032
 Queen+0.031
adies+0.030
 mushrooms+0.029
Neg logit contributions
bars-0.028
 outage-0.027
spe-0.025
umbles-0.024
 blowing-0.023
Token every
Token ID790
Feature activation+4.73e-02

Pos logit contributions
hire+0.026
 Princess+0.024
 Queen+0.023
ocket+0.023
 robber+0.022
Neg logit contributions
bars-0.017
 outage-0.017
 hill-0.016
 directions-0.015
 direction-0.015
Token time
Token ID640
Feature activation-9.52e-04

Pos logit contributions
bars+0.053
 outage+0.047
spe+0.044
bike+0.042
umbling+0.039
Neg logit contributions
 Princess-0.098
hire-0.096
 Queen-0.094
adies-0.091
 robber-0.089
Token they
Token ID484
Feature activation-1.38e-02

Pos logit contributions
hire+0.002
 Princess+0.002
 Queen+0.002
ocket+0.002
 robber+0.002
Neg logit contributions
bars-0.001
 outage-0.001
 hill-0.001
 directions-0.001
 direction-0.001
Token came
Token ID1625
Feature activation-6.61e-03

Pos logit contributions
hire+0.029
 Princess+0.025
ocket+0.024
 Queen+0.024
adies+0.024
Neg logit contributions
 hill-0.016
 blowing-0.015
 kay-0.015
bars-0.015
 way-0.014
Token to
Token ID284
Feature activation-1.12e-03

Pos logit contributions
hire+0.013
 Princess+0.012
 Queen+0.012
adies+0.012
 robber+0.011
Neg logit contributions
bars-0.008
 outage-0.008
 hill-0.007
spe-0.007
 motor-0.007
Token visit
Token ID3187
Feature activation-2.21e-02

Pos logit contributions
hire+0.002
ocket+0.002
 Princess+0.002
 Queen+0.002
adies+0.002
Neg logit contributions
 hill-0.001
 outage-0.001
 directions-0.001
 direction-0.001
bars-0.001
Token the
Token ID262
Feature activation+2.00e-02

Pos logit contributions
hire+0.017
 Princess+0.017
 Queen+0.016
 Prince+0.015
 mushrooms+0.015
Neg logit contributions
bars-0.025
 outage-0.025
umbling-0.023
 hill-0.023
roller-0.023
Token fish
Token ID5916
Feature activation+4.15e-03

Pos logit contributions
bars+0.031
 outage+0.030
 directions+0.028
 hill+0.028
 blowing+0.027
Neg logit contributions
hire-0.035
 Princess-0.032
 Queen-0.031
ocket-0.029
 robber-0.029
Token said
Token ID531
Feature activation-2.60e-02

Pos logit contributions
 outage+0.005
bars+0.005
 hill+0.005
 directions+0.005
 direction+0.005
Neg logit contributions
hire-0.008
 Princess-0.007
 Queen-0.007
ocket-0.007
 robber-0.007
Token.
Token ID13
Feature activation+2.09e-02

Pos logit contributions
hire+0.050
 Princess+0.046
 Queen+0.045
ocket+0.043
 robber+0.042
Neg logit contributions
 outage-0.034
bars-0.033
 hill-0.032
 direction-0.031
 directions-0.031
Token He
Token ID679
Feature activation-2.13e-03

Pos logit contributions
 outage+0.032
bars+0.030
 directions+0.028
 direction+0.028
 blowing+0.028
Neg logit contributions
hire-0.036
 Princess-0.034
 Queen-0.033
 Prince-0.031
 robber-0.030
Token began
Token ID2540
Feature activation+5.30e-02

Pos logit contributions
hire+0.004
 Princess+0.004
 Queen+0.004
ocket+0.004
 robber+0.003
Neg logit contributions
bars-0.003
 outage-0.003
 hill-0.003
 directions-0.003
 direction-0.003
Token to
Token ID284
Feature activation+2.55e-02

Pos logit contributions
bars+0.083
 outage+0.081
ney+0.077
 hill+0.073
icycle+0.068
Neg logit contributions
 Princess-0.085
hire-0.084
 Queen-0.084
 Prince-0.079
 mushrooms-0.078
Token believe
Token ID1975
Feature activation-3.29e-03

Pos logit contributions
bars+0.037
 outage+0.036
ney+0.033
ipt+0.032
spe+0.032
Neg logit contributions
hire-0.046
 Princess-0.042
 Queen-0.042
 robber-0.039
 Prince-0.039
Token in
Token ID287
Feature activation+1.95e-02

Pos logit contributions
hire+0.007
 robber+0.006
ns+0.006
 Princess+0.006
ocket+0.006
Neg logit contributions
 blowing-0.003
 hill-0.003
 outage-0.003
bars-0.003
 directions-0.003
Token himself
Token ID2241
Feature activation-6.58e-03

Pos logit contributions
 directions+0.021
 hill+0.020
 direction+0.020
bars+0.019
 outage+0.018
Neg logit contributions
hire-0.042
 Princess-0.037
ocket-0.037
adies-0.037
 Queen-0.037
Token and
Token ID290
Feature activation+1.51e-02

Pos logit contributions
hire+0.011
 Princess+0.010
ocket+0.010
 Queen+0.009
 mushrooms+0.009
Neg logit contributions
 outage-0.011
bars-0.010
spe-0.009
 hill-0.009
 directions-0.009
Token,
Token ID11
Feature activation+1.15e-02

Pos logit contributions
hire+0.048
 robber+0.042
ocket+0.042
 Queen+0.042
 Princess+0.042
Neg logit contributions
bars-0.013
 outage-0.012
 hill-0.011
spe-0.011
 directions-0.011
Token �
Token ID6184
Feature activation-2.49e-02

Pos logit contributions
bars+0.013
 hill+0.013
 outage+0.013
 blowing+0.012
 directions+0.012
Neg logit contributions
hire-0.024
ocket-0.021
 Princess-0.021
 Queen-0.021
adies-0.020
Token�
Token ID95
Feature activation+8.12e-03

Pos logit contributions
hire+0.045
 Queen+0.039
 mushrooms+0.038
ocket+0.037
 Princess+0.036
Neg logit contributions
 hill-0.037
 outage-0.036
bars-0.035
 kay-0.034
ney-0.033
Token€
Token ID26391
Feature activation-1.14e-02

Pos logit contributions
 outage+0.012
ney+0.011
 hill+0.011
bars+0.011
 directions+0.011
Neg logit contributions
hire-0.015
 Queen-0.013
 mushrooms-0.012
ocket-0.012
 Princess-0.012
Token�
Token ID129
Feature activation-1.30e-02

Pos logit contributions
hire+0.024
 Princess+0.021
 Queen+0.021
ocket+0.021
 robber+0.020
Neg logit contributions
 outage-0.013
 hill-0.013
bars-0.013
 direction-0.012
 directions-0.012
Token�
Token ID241
Feature activation+4.97e-02

Pos logit contributions
hire+0.021
 Queen+0.018
 Princess+0.018
ocket+0.017
 mushrooms+0.017
Neg logit contributions
 outage-0.021
 hill-0.021
bars-0.021
 directions-0.020
 blowing-0.020
TokenOh
Token ID5812
Feature activation-9.82e-03

Pos logit contributions
bars+0.069
 outage+0.068
 hill+0.067
 directions+0.065
 direction+0.064
Neg logit contributions
hire-0.091
 Princess-0.081
 Queen-0.079
ocket-0.077
 robber-0.076
Token,
Token ID11
Feature activation-1.01e-03

Pos logit contributions
hire+0.024
ocket+0.022
 Queen+0.021
 Princess+0.021
 robber+0.021
Neg logit contributions
bars-0.008
 outage-0.008
 hill-0.008
 directions-0.008
 direction-0.007
Token wow
Token ID28796
Feature activation-2.44e-02

Pos logit contributions
hire+0.002
ocket+0.002
ns+0.002
 Queen+0.002
adies+0.002
Neg logit contributions
bars-0.001
 hill-0.001
 way-0.001
 outage-0.001
 directions-0.001
Token.
Token ID13
Feature activation+1.40e-02

Pos logit contributions
hire+0.059
ocket+0.053
 Queen+0.052
 robber+0.051
 Princess+0.051
Neg logit contributions
 hill-0.021
 way-0.020
 direction-0.019
s-0.019
 directions-0.019
Token That
Token ID1320
Feature activation+8.46e-03

Pos logit contributions
bars+0.017
 outage+0.016
 hill+0.015
 Keeping+0.015
spe+0.014
Neg logit contributions
hire-0.029
ocket-0.026
ns-0.024
ils-0.024
 Queen-0.024
Token Tim
Token ID5045
Feature activation+1.20e-02

Pos logit contributions
 outage+0.049
bars+0.048
 hill+0.046
 directions+0.046
 direction+0.045
Neg logit contributions
hire-0.047
 Princess-0.042
 Queen-0.042
 robber-0.039
ocket-0.038
Tokenmy
Token ID1820
Feature activation+3.31e-03

Pos logit contributions
 outage+0.013
bars+0.012
 hill+0.012
 directions+0.011
 direction+0.011
Neg logit contributions
hire-0.026
 Princess-0.023
 Queen-0.023
 robber-0.023
ocket-0.023
Token's
Token ID338
Feature activation+3.07e-02

Pos logit contributions
 outage+0.003
bars+0.003
 hill+0.003
 directions+0.003
 direction+0.003
Neg logit contributions
hire-0.007
 Princess-0.007
 Queen-0.007
ocket-0.007
 robber-0.006
Token toys
Token ID14958
Feature activation-6.44e-03

Pos logit contributions
bars+0.040
 outage+0.040
 directions+0.036
 hill+0.036
 direction+0.035
Neg logit contributions
hire-0.058
 Queen-0.052
 robber-0.052
 Princess-0.051
ocket-0.049
Token were
Token ID547
Feature activation-1.11e-02

Pos logit contributions
hire+0.011
 Princess+0.010
ocket+0.010
 robber+0.010
 Queen+0.010
Neg logit contributions
 outage-0.009
bars-0.008
 directions-0.008
 direction-0.008
spe-0.008
Token all
Token ID477
Feature activation+4.81e-02

Pos logit contributions
hire+0.022
 Princess+0.020
 Queen+0.020
ocket+0.019
adies+0.019
Neg logit contributions
 outage-0.013
bars-0.013
 hill-0.013
 blowing-0.013
 directions-0.013
Token over
Token ID625
Feature activation+3.53e-03

Pos logit contributions
bars+0.069
 hill+0.068
 outage+0.066
ney+0.064
ipt+0.063
Neg logit contributions
hire-0.081
 Princess-0.072
 Queen-0.070
 robber-0.069
 mushrooms-0.068
Token the
Token ID262
Feature activation-1.03e-02

Pos logit contributions
 outage+0.005
bars+0.005
 hill+0.004
 direction+0.004
 directions+0.004
Neg logit contributions
hire-0.007
 Princess-0.006
 Queen-0.006
ocket-0.006
 robber-0.006
Token place
Token ID1295
Feature activation+3.21e-02

Pos logit contributions
hire+0.020
 Princess+0.019
 Queen+0.018
 robber+0.018
 mushrooms+0.018
Neg logit contributions
bars-0.013
 outage-0.012
 blowing-0.011
spe-0.010
ipt-0.010
Token because
Token ID780
Feature activation+1.78e-02

Pos logit contributions
bars+0.038
 outage+0.038
 hill+0.037
 directions+0.036
 direction+0.035
Neg logit contributions
hire-0.062
 Princess-0.056
 Queen-0.055
ocket-0.054
 robber-0.053
Token he
Token ID339
Feature activation+2.29e-03

Pos logit contributions
 outage+0.026
bars+0.026
 hill+0.025
 directions+0.024
 direction+0.024
Neg logit contributions
hire-0.031
 Princess-0.028
 Queen-0.027
ocket-0.026
 robber-0.026
Token lower
Token ID2793
Feature activation+1.85e-02

Pos logit contributions
bars+0.028
 outage+0.028
spe+0.026
umbling+0.025
ney+0.025
Neg logit contributions
hire-0.026
 Princess-0.023
 Queen-0.022
 wow-0.022
 robber-0.021
Token it
Token ID340
Feature activation+3.37e-03

Pos logit contributions
 outage+0.028
bars+0.027
 hill+0.025
spe+0.025
ney+0.024
Neg logit contributions
hire-0.033
 Princess-0.028
 Queen-0.028
ocket-0.026
adies-0.026
Token.
Token ID13
Feature activation+1.99e-02

Pos logit contributions
 outage+0.005
bars+0.004
 hill+0.004
spe+0.004
 kay+0.004
Neg logit contributions
hire-0.006
 Princess-0.006
 Queen-0.005
ocket-0.005
 robber-0.005
Token Jack
Token ID3619
Feature activation+3.13e-03

Pos logit contributions
bars+0.018
 outage+0.017
 hill+0.017
 directions+0.016
 direction+0.016
Neg logit contributions
hire-0.046
 Princess-0.042
 Queen-0.042
ocket-0.041
 robber-0.040
Token was
Token ID373
Feature activation+1.54e-02

Pos logit contributions
 outage+0.004
bars+0.004
 hill+0.004
 directions+0.004
 direction+0.004
Neg logit contributions
hire-0.006
 Princess-0.005
 Queen-0.005
ocket-0.005
ns-0.005
Token careful
Token ID8161
Feature activation+5.95e-02

Pos logit contributions
bars+0.027
 outage+0.026
 hill+0.026
spe+0.024
 directions+0.024
Neg logit contributions
hire-0.023
 Princess-0.020
 Queen-0.020
ocket-0.019
 robber-0.018
Token not
Token ID407
Feature activation+1.21e-02

Pos logit contributions
 hill+0.097
 outage+0.097
 Mechan+0.094
bars+0.092
 kay+0.091
Neg logit contributions
hire-0.090
ured-0.074
 as-0.072
 Princess-0.068
 Queen-0.068
Token to
Token ID284
Feature activation-8.25e-03

Pos logit contributions
bars+0.012
 outage+0.012
 hill+0.011
ney+0.011
spe+0.011
Neg logit contributions
hire-0.026
 robber-0.023
 Princess-0.023
 Queen-0.023
 mushrooms-0.022
Token make
Token ID787
Feature activation+3.69e-03

Pos logit contributions
hire+0.013
 robber+0.012
 Princess+0.011
 Queen+0.011
 like+0.011
Neg logit contributions
bars-0.012
 outage-0.012
spe-0.012
 directions-0.012
ipt-0.011
Token it
Token ID340
Feature activation-3.38e-03

Pos logit contributions
bars+0.005
 hill+0.005
 outage+0.005
 directions+0.005
 direction+0.005
Neg logit contributions
hire-0.007
ocket-0.006
 Princess-0.006
 Queen-0.006
adies-0.006
Token too
Token ID1165
Feature activation+6.30e-03

Pos logit contributions
hire+0.006
 Princess+0.005
 Queen+0.005
 robber+0.005
ocket+0.005
Neg logit contributions
 outage-0.005
bars-0.005
 hill-0.005
 directions-0.004
 direction-0.004
Token in
Token ID287
Feature activation+3.02e-02

Pos logit contributions
bars+0.004
 outage+0.004
 hill+0.004
 motor+0.004
spe+0.004
Neg logit contributions
hire-0.008
 Princess-0.007
 Queen-0.007
ocket-0.007
 robber-0.007
Token the
Token ID262
Feature activation-2.90e-03

Pos logit contributions
bars+0.045
 outage+0.042
 hill+0.040
ney+0.038
bike+0.038
Neg logit contributions
hire-0.053
 Princess-0.049
 Queen-0.049
ocket-0.047
 robber-0.046
Token long
Token ID890
Feature activation+2.39e-02

Pos logit contributions
hire+0.006
 Princess+0.006
 Queen+0.006
 Prince+0.005
ocket+0.005
Neg logit contributions
bars-0.003
 outage-0.003
 blowing-0.003
spe-0.003
bike-0.003
Token run
Token ID1057
Feature activation+2.92e-04

Pos logit contributions
bars+0.028
 outage+0.026
bike+0.024
 blowing+0.024
umbling+0.022
Neg logit contributions
hire-0.047
 Princess-0.046
 Queen-0.044
 mushrooms-0.044
ocket-0.042
Token.
Token ID13
Feature activation+2.96e-02

Pos logit contributions
 outage+0.000
bars+0.000
 hill+0.000
 directions+0.000
 direction+0.000
Neg logit contributions
hire-0.001
 Princess-0.001
 Queen-0.001
ocket-0.001
 robber-0.001
Token<|endoftext|>
Token ID50256
Feature activation+3.08e-02

Pos logit contributions
 outage+0.025
bars+0.025
 hill+0.024
 directions+0.022
 direction+0.022
Neg logit contributions
hire-0.070
 Princess-0.065
 Queen-0.064
ocket-0.062
 robber-0.061
TokenOnce
Token ID7454
Feature activation-2.18e-02

Pos logit contributions
 outage+0.046
bars+0.046
spe+0.042
 directions+0.042
 blowing+0.041
Neg logit contributions
 Princess-0.051
hire-0.050
 Queen-0.047
 Prince-0.046
ocket-0.045
Token upon
Token ID2402
Feature activation+1.44e-02

Pos logit contributions
hire+0.031
 Princess+0.029
ocket+0.027
 Queen+0.027
adies+0.026
Neg logit contributions
bike-0.036
 outage-0.036
bars-0.035
 hill-0.034
 directions-0.034
Token a
Token ID257
Feature activation-3.70e-03

Pos logit contributions
 outage+0.022
 hill+0.022
bars+0.022
 directions+0.021
 direction+0.021
Neg logit contributions
hire-0.023
 Queen-0.020
 Princess-0.020
ocket-0.020
 robber-0.019
Token time
Token ID640
Feature activation+1.38e-02

Pos logit contributions
hire+0.007
 Princess+0.007
 Queen+0.007
ocket+0.007
 robber+0.006
Neg logit contributions
bars-0.004
 outage-0.004
 hill-0.004
 directions-0.004
 direction-0.004
Token there
Token ID612
Feature activation+2.97e-02

Pos logit contributions
bars+0.017
 outage+0.017
 hill+0.016
 directions+0.016
 direction+0.015
Neg logit contributions
hire-0.027
 Princess-0.025
 Queen-0.025
ocket-0.024
adies-0.024
Token the
Token ID262
Feature activation-6.61e-03

Pos logit contributions
 outage+0.020
bars+0.019
spe+0.018
 hill+0.018
ney+0.017
Neg logit contributions
hire-0.026
 Princess-0.023
 Queen-0.022
 mushrooms-0.022
ocket-0.022
Token kitchen
Token ID9592
Feature activation-1.55e-02

Pos logit contributions
hire+0.013
 Princess+0.011
 Queen+0.011
ocket+0.011
 robber+0.011
Neg logit contributions
bars-0.009
 outage-0.009
 hill-0.008
 directions-0.008
spe-0.008
Token.
Token ID13
Feature activation+5.96e-03

Pos logit contributions
hire+0.031
 Princess+0.028
 Queen+0.028
ocket+0.027
 robber+0.027
Neg logit contributions
 outage-0.019
bars-0.018
 hill-0.017
spe-0.017
 directions-0.017
Token The
Token ID383
Feature activation+8.51e-03

Pos logit contributions
 outage+0.010
bars+0.010
spe+0.009
 directions+0.009
 blowing+0.009
Neg logit contributions
hire-0.009
 Princess-0.009
 Queen-0.008
 robber-0.008
 mushrooms-0.008
Token bottle
Token ID9294
Feature activation+5.26e-04

Pos logit contributions
bars+0.011
 outage+0.010
spe+0.010
 hill+0.009
 directions+0.009
Neg logit contributions
hire-0.017
 Princess-0.016
 Queen-0.016
ocket-0.015
 robber-0.015
Token was
Token ID373
Feature activation+9.06e-03

Pos logit contributions
 outage+0.000
bars+0.000
 directions+0.000
 hill+0.000
 direction+0.000
Neg logit contributions
hire-0.001
 Princess-0.001
 Queen-0.001
ocket-0.001
 robber-0.001
Token full
Token ID1336
Feature activation+7.88e-03

Pos logit contributions
bars+0.012
 outage+0.012
 hill+0.012
 directions+0.011
 direction+0.011
Neg logit contributions
hire-0.017
 Princess-0.015
 Queen-0.015
ocket-0.015
 robber-0.014
Token of
Token ID286
Feature activation-6.33e-03

Pos logit contributions
bars+0.008
 outage+0.008
 hill+0.007
spe+0.007
 directions+0.007
Neg logit contributions
hire-0.017
 Princess-0.016
 Queen-0.016
ocket-0.015
 mushrooms-0.015
Token a
Token ID257
Feature activation-1.21e-02

Pos logit contributions
hire+0.011
 Princess+0.010
 Queen+0.010
ocket+0.009
 robber+0.009
Neg logit contributions
 outage-0.009
bars-0.009
 hill-0.009
 directions-0.009
 direction-0.009
Token red
Token ID2266
Feature activation+1.07e-02

Pos logit contributions
hire+0.020
 Princess+0.018
 Queen+0.018
 robber+0.017
ocket+0.017
Neg logit contributions
bars-0.020
 outage-0.018
 directions-0.018
bike-0.017
umbles-0.017
Token sauce
Token ID10746
Feature activation-1.12e-02

Pos logit contributions
bars+0.013
 outage+0.013
 directions+0.013
spe+0.012
ney+0.011
Neg logit contributions
hire-0.021
 Princess-0.019
 Queen-0.019
 robber-0.019
ocket-0.018
Token says
Token ID1139
Feature activation-2.47e-02

Pos logit contributions
hire+0.014
 Princess+0.013
 Queen+0.013
 robber+0.012
ocket+0.012
Neg logit contributions
 outage-0.007
 hill-0.007
bars-0.007
 direction-0.007
 directions-0.006
Token.
Token ID13
Feature activation-6.88e-03

Pos logit contributions
hire+0.035
ocket+0.034
ily+0.032
 Princess+0.031
 Queen+0.030
Neg logit contributions
 hill-0.041
 outage-0.040
 kay-0.040
bars-0.040
icycle-0.040
Token She
Token ID1375
Feature activation-1.03e-02

Pos logit contributions
hire+0.012
ocket+0.010
 Queen+0.010
 Princess+0.010
 robber+0.010
Neg logit contributions
bars-0.010
 hill-0.010
 outage-0.010
 direction-0.010
 directions-0.009
Token knows
Token ID4206
Feature activation-1.58e-02

Pos logit contributions
hire+0.022
 Queen+0.020
 Princess+0.020
ocket+0.019
 robber+0.019
Neg logit contributions
bars-0.011
 hill-0.011
 outage-0.010
 direction-0.010
 directions-0.010
Token he
Token ID339
Feature activation+6.61e-04

Pos logit contributions
hire+0.027
 Princess+0.023
 Queen+0.022
ocket+0.022
 robber+0.022
Neg logit contributions
bars-0.025
 outage-0.024
 hill-0.024
 directions-0.023
 direction-0.023
Token is
Token ID318
Feature activation+3.87e-03

Pos logit contributions
 outage+0.001
bars+0.001
 hill+0.000
 direction+0.000
 directions+0.000
Neg logit contributions
hire-0.002
 Princess-0.001
 Queen-0.001
ocket-0.001
 robber-0.001
Token right
Token ID826
Feature activation+1.25e-02

Pos logit contributions
 blowing+0.004
 directions+0.004
 way+0.004
 living+0.004
 outage+0.004
Neg logit contributions
hire-0.008
ocket-0.007
adies-0.007
als-0.007
 robber-0.006
Token.
Token ID13
Feature activation-8.08e-03

Pos logit contributions
 outage+0.012
bars+0.012
 hill+0.012
 directions+0.011
 direction+0.011
Neg logit contributions
hire-0.028
 Princess-0.026
 Queen-0.025
ocket-0.025
 robber-0.025
Token She
Token ID1375
Feature activation-8.81e-03

Pos logit contributions
hire+0.016
ocket+0.013
 robber+0.013
 Queen+0.013
 Princess+0.013
Neg logit contributions
bars-0.011
 hill-0.010
 outage-0.010
 direction-0.010
 directions-0.010
Token does
Token ID857
Feature activation+7.43e-03

Pos logit contributions
hire+0.019
 Queen+0.017
 Princess+0.017
ocket+0.017
 robber+0.016
Neg logit contributions
bars-0.009
 hill-0.009
 outage-0.009
 direction-0.008
 directions-0.008
Token not
Token ID407
Feature activation+2.02e-02

Pos logit contributions
 outage+0.008
 blowing+0.007
 directions+0.007
bars+0.007
 hill+0.007
Neg logit contributions
hire-0.016
 Queen-0.014
 Princess-0.014
ocket-0.014
 mushrooms-0.014
Token and
Token ID290
Feature activation-9.38e-03

Pos logit contributions
hire+0.004
 Princess+0.004
 Queen+0.004
ocket+0.004
 robber+0.004
Neg logit contributions
 outage-0.003
bars-0.003
 hill-0.003
 directions-0.003
 direction-0.003
Token Jack
Token ID3619
Feature activation-5.19e-03

Pos logit contributions
hire+0.015
ocket+0.012
 Princess+0.012
adies+0.012
 Queen+0.012
Neg logit contributions
bars-0.016
 hill-0.015
 directions-0.015
 outage-0.015
 direction-0.015
Token were
Token ID547
Feature activation-2.33e-03

Pos logit contributions
hire+0.009
 Princess+0.008
 Queen+0.008
ocket+0.008
 robber+0.008
Neg logit contributions
 outage-0.007
bars-0.007
 hill-0.007
 directions-0.007
 direction-0.007
Token so
Token ID523
Feature activation-1.87e-03

Pos logit contributions
hire+0.004
 Princess+0.004
 Queen+0.004
ocket+0.004
 robber+0.004
Neg logit contributions
 outage-0.003
bars-0.003
 hill-0.003
 directions-0.003
 direction-0.003
Token excited
Token ID6568
Feature activation+7.36e-03

Pos logit contributions
hire+0.003
 Princess+0.003
 Queen+0.003
ocket+0.003
 robber+0.003
Neg logit contributions
 outage-0.003
bars-0.003
 hill-0.003
 direction-0.003
 directions-0.003
Token to
Token ID284
Feature activation+1.32e-02

Pos logit contributions
 outage+0.008
bars+0.008
 hill+0.008
 directions+0.008
 direction+0.008
Neg logit contributions
hire-0.016
 Princess-0.014
 Queen-0.014
ocket-0.014
 robber-0.013
Token try
Token ID1949
Feature activation+1.27e-02

Pos logit contributions
bars+0.019
 outage+0.019
 hill+0.018
 direction+0.018
ney+0.017
Neg logit contributions
hire-0.023
 Princess-0.021
 Queen-0.021
 robber-0.020
ocket-0.019
Token it
Token ID340
Feature activation-6.94e-03

Pos logit contributions
bars+0.021
 outage+0.021
ney+0.020
spe+0.020
 hill+0.020
Neg logit contributions
hire-0.018
 Princess-0.017
 Queen-0.016
 wow-0.015
 robber-0.015
Token out
Token ID503
Feature activation-4.08e-03

Pos logit contributions
hire+0.013
 Princess+0.011
 Queen+0.011
ocket+0.011
 robber+0.011
Neg logit contributions
 outage-0.009
bars-0.009
 hill-0.009
 directions-0.009
 direction-0.009
Token.
Token ID13
Feature activation+2.17e-02

Pos logit contributions
hire+0.008
 Princess+0.007
 Queen+0.007
ocket+0.007
 robber+0.006
Neg logit contributions
bars-0.005
 outage-0.005
 hill-0.005
ney-0.005
 direction-0.005
Token
Token ID198
Feature activation+1.55e-02

Pos logit contributions
bars+0.020
 outage+0.020
 hill+0.020
 directions+0.019
 direction+0.019
Neg logit contributions
hire-0.050
 Princess-0.045
 Queen-0.044
ocket-0.044
 robber-0.043
Token the
Token ID262
Feature activation+1.44e-02

Pos logit contributions
hire+0.074
ocket+0.064
 Princess+0.063
 Queen+0.062
 robber+0.061
Neg logit contributions
 hill-0.049
 outage-0.048
bars-0.048
 directions-0.046
 direction-0.046
Token man
Token ID582
Feature activation+1.33e-02

Pos logit contributions
 hill+0.016
bars+0.016
 outage+0.016
 directions+0.015
 direction+0.015
Neg logit contributions
hire-0.030
 Princess-0.027
 Queen-0.027
ocket-0.027
adies-0.026
Token shouts
Token ID34757
Feature activation-2.81e-02

Pos logit contributions
 outage+0.016
bars+0.016
 hill+0.016
 directions+0.015
 direction+0.015
Neg logit contributions
hire-0.026
 Princess-0.024
 Queen-0.024
ocket-0.023
adies-0.022
Token.
Token ID13
Feature activation+8.69e-03

Pos logit contributions
hire+0.071
ocket+0.063
 Princess+0.062
 Queen+0.062
 robber+0.061
Neg logit contributions
 directions-0.019
 direction-0.019
bars-0.019
 outage-0.018
s-0.017
Token "
Token ID366
Feature activation-5.12e-03

Pos logit contributions
bars+0.011
 outage+0.011
 hill+0.011
 directions+0.011
 direction+0.011
Neg logit contributions
hire-0.017
 Princess-0.015
 Queen-0.015
ocket-0.014
 robber-0.014
TokenThis
Token ID1212
Feature activation+3.28e-02

Pos logit contributions
hire+0.009
ocket+0.009
 Princess+0.009
 Queen+0.008
 mushrooms+0.008
Neg logit contributions
 hill-0.007
bars-0.007
 outage-0.007
 directions-0.006
spe-0.006
Token is
Token ID318
Feature activation+6.58e-03

Pos logit contributions
bars+0.027
spe+0.024
 outage+0.024
 directions+0.024
 blowing+0.022
Neg logit contributions
hire-0.078
 robber-0.072
 Queen-0.071
 Princess-0.071
adies-0.069
Token not
Token ID407
Feature activation+3.18e-02

Pos logit contributions
 blowing+0.007
 way+0.007
 kay+0.006
 outage+0.006
 direction+0.006
Neg logit contributions
hire-0.016
ocket-0.014
 Princess-0.013
ilies-0.013
adies-0.013
Token a
Token ID257
Feature activation+1.75e-02

Pos logit contributions
 outage+0.040
 way+0.039
 blowing+0.038
 hill+0.037
 directions+0.037
Neg logit contributions
hire-0.061
ocket-0.057
 Princess-0.056
 Queen-0.054
ilies-0.053
Token toy
Token ID13373
Feature activation-3.11e-03

Pos logit contributions
bars+0.019
 outage+0.019
 directions+0.018
 hill+0.017
 direction+0.017
Neg logit contributions
hire-0.037
 Princess-0.034
 Queen-0.034
 robber-0.033
ocket-0.033
Token,
Token ID11
Feature activation-2.71e-02

Pos logit contributions
hire+0.007
 Princess+0.006
 Queen+0.006
ocket+0.006
adies+0.006
Neg logit contributions
 outage-0.003
bars-0.003
 hill-0.003
 direction-0.003
 directions-0.003
Token Ben
Token ID3932
Feature activation-6.99e-03

Pos logit contributions
hire+0.040
ocket+0.033
 Princess+0.032
 Queen+0.032
adies+0.030
Neg logit contributions
 hill-0.057
bars-0.056
 kay-0.055
 blowing-0.054
 direction-0.053
Token were
Token ID547
Feature activation+6.01e-04

Pos logit contributions
hire+0.014
 Princess+0.012
 Queen+0.012
ocket+0.012
 robber+0.012
Neg logit contributions
bars-0.009
 outage-0.009
 hill-0.008
 directions-0.008
 direction-0.008
Token sad
Token ID6507
Feature activation+1.01e-03

Pos logit contributions
 outage+0.001
bars+0.001
 hill+0.001
 direction+0.001
 directions+0.001
Neg logit contributions
hire-0.001
 Princess-0.001
 Queen-0.001
ocket-0.001
adies-0.001
Token.
Token ID13
Feature activation-3.35e-03

Pos logit contributions
bars+0.002
spe+0.001
 directions+0.001
 outage+0.001
 hill+0.001
Neg logit contributions
hire-0.002
 Princess-0.001
 Queen-0.001
 robber-0.001
 mushrooms-0.001
Token They
Token ID1119
Feature activation-5.42e-03

Pos logit contributions
hire+0.007
 Princess+0.006
 Queen+0.006
ocket+0.006
 robber+0.006
Neg logit contributions
 outage-0.004
bars-0.004
 hill-0.004
 directions-0.004
 direction-0.004
Token wanted
Token ID2227
Feature activation-1.06e-03

Pos logit contributions
hire+0.012
 Princess+0.010
adies+0.010
ocket+0.010
 Queen+0.010
Neg logit contributions
 hill-0.006
 direction-0.005
 outage-0.005
bars-0.005
 way-0.005
Token to
Token ID284
Feature activation+1.17e-02

Pos logit contributions
hire+0.002
 Princess+0.002
 Queen+0.002
 robber+0.002
ocket+0.002
Neg logit contributions
bars-0.001
 outage-0.001
 hill-0.001
spe-0.001
ney-0.001
Token play
Token ID711
Feature activation-1.44e-03

Pos logit contributions
bars+0.015
 outage+0.015
 hill+0.014
 directions+0.014
 direction+0.014
Neg logit contributions
hire-0.023
 Princess-0.020
 Queen-0.020
ocket-0.020
 robber-0.019
Token with
Token ID351
Feature activation-2.68e-02

Pos logit contributions
hire+0.003
 Queen+0.003
ocket+0.003
 Princess+0.003
adies+0.003
Neg logit contributions
 hill-0.001
 outage-0.001
 direction-0.001
 directions-0.001
bars-0.001
Token Spot
Token ID15899
Feature activation-1.80e-02

Pos logit contributions
hire+0.056
ocket+0.049
 Princess+0.049
 Queen+0.048
adies+0.048
Neg logit contributions
 hill-0.031
 directions-0.030
 direction-0.030
bars-0.030
 kay-0.029
Token in
Token ID287
Feature activation+2.20e-02

Pos logit contributions
hire+0.037
 Princess+0.034
 Queen+0.033
ocket+0.032
 robber+0.032
Neg logit contributions
 outage-0.021
bars-0.021
 hill-0.020
 directions-0.019
 direction-0.019
Token it
Token ID340
Feature activation-1.21e-02

Pos logit contributions
 directions+0.010
 direction+0.009
bars+0.009
 hill+0.009
 way+0.008
Neg logit contributions
hire-0.021
ocket-0.020
 robber-0.020
 Princess-0.020
adies-0.019
Token into
Token ID656
Feature activation+4.86e-03

Pos logit contributions
hire+0.026
 Princess+0.024
 Queen+0.023
ocket+0.023
 robber+0.023
Neg logit contributions
 hill-0.012
bars-0.012
 outage-0.012
 directions-0.012
 direction-0.012
Token slices
Token ID24314
Feature activation-1.59e-02

Pos logit contributions
 hill+0.006
 directions+0.006
bars+0.006
 outage+0.006
 direction+0.006
Neg logit contributions
hire-0.009
 Princess-0.009
ocket-0.008
 Queen-0.008
 robber-0.008
Token and
Token ID290
Feature activation-5.63e-03

Pos logit contributions
hire+0.032
 Princess+0.028
ocket+0.028
 Queen+0.028
 robber+0.027
Neg logit contributions
 outage-0.019
bars-0.019
 directions-0.018
 hill-0.018
 direction-0.018
Token shared
Token ID4888
Feature activation-4.35e-02

Pos logit contributions
hire+0.010
adies+0.009
 Princess+0.009
ocket+0.009
 robber+0.009
Neg logit contributions
 hill-0.008
bars-0.008
 blowing-0.007
 directions-0.007
 way-0.007
Token it
Token ID340
Feature activation-1.47e-02

Pos logit contributions
 robber+0.097
hire+0.096
ocket+0.091
adies+0.090
vell+0.089
Neg logit contributions
 directions-0.039
s-0.035
 story-0.034
 way-0.034
 hill-0.034
Token with
Token ID351
Feature activation-2.94e-02

Pos logit contributions
hire+0.036
adies+0.032
 robber+0.032
 Princess+0.032
ocket+0.032
Neg logit contributions
 hill-0.012
 directions-0.011
spe-0.011
bars-0.011
 way-0.010
Token each
Token ID1123
Feature activation+3.24e-02

Pos logit contributions
hire+0.063
adies+0.059
ocket+0.058
 robber+0.057
 Princess+0.056
Neg logit contributions
 directions-0.032
 direction-0.028
 way-0.027
 hill-0.027
 motor-0.027
Token other
Token ID584
Feature activation+1.76e-02

Pos logit contributions
 hill+0.027
 directions+0.027
 outage+0.027
bars+0.027
 direction+0.027
Neg logit contributions
hire-0.076
 Princess-0.069
ocket-0.068
 Queen-0.068
 robber-0.066
Token.
Token ID13
Feature activation+2.61e-02

Pos logit contributions
 outage+0.022
bars+0.021
 hill+0.020
 direction+0.020
 directions+0.020
Neg logit contributions
hire-0.035
 Princess-0.032
 Queen-0.031
ocket-0.030
 robber-0.030
Token It
Token ID632
Feature activation+2.13e-03

Pos logit contributions
 outage+0.033
bars+0.032
 hill+0.030
 directions+0.029
 direction+0.029
Neg logit contributions
hire-0.052
 Princess-0.050
 Queen-0.048
 Prince-0.045
ocket-0.044
Token her
Token ID607
Feature activation-1.94e-02

Pos logit contributions
bars+0.005
 hill+0.005
 directions+0.005
 direction+0.005
 outage+0.005
Neg logit contributions
hire-0.005
ocket-0.004
 Princess-0.004
 Queen-0.004
 robber-0.004
Token backpack
Token ID22526
Feature activation-1.27e-02

Pos logit contributions
hire+0.037
 Princess+0.035
 Queen+0.034
ocket+0.033
 robber+0.032
Neg logit contributions
 outage-0.026
bars-0.025
 blowing-0.023
spe-0.023
 hill-0.022
Token and
Token ID290
Feature activation-1.36e-02

Pos logit contributions
hire+0.027
 Princess+0.024
 Queen+0.024
ocket+0.024
 robber+0.023
Neg logit contributions
 outage-0.014
bars-0.014
 hill-0.013
 directions-0.013
 direction-0.013
Token they
Token ID484
Feature activation-8.39e-03

Pos logit contributions
hire+0.028
ocket+0.024
adies+0.024
 Princess+0.024
 robber+0.024
Neg logit contributions
bars-0.016
 hill-0.016
 way-0.015
 kay-0.015
 directions-0.015
Token set
Token ID900
Feature activation+5.89e-03

Pos logit contributions
hire+0.017
 Princess+0.015
 Queen+0.015
ocket+0.014
 robber+0.014
Neg logit contributions
bars-0.010
 hill-0.010
 kay-0.010
 outage-0.010
 directions-0.010
Token off
Token ID572
Feature activation-4.51e-03

Pos logit contributions
 outage+0.008
bars+0.007
 hill+0.007
 directions+0.007
 direction+0.007
Neg logit contributions
hire-0.011
 Princess-0.010
 Queen-0.010
ocket-0.010
 robber-0.010
Token.
Token ID13
Feature activation+1.46e-02

Pos logit contributions
hire+0.008
 Princess+0.008
 Queen+0.007
ocket+0.007
adies+0.007
Neg logit contributions
bars-0.006
 outage-0.006
 hill-0.006
 directions-0.006
 direction-0.006
Token
Token ID220
Feature activation+2.71e-02

Pos logit contributions
bars+0.012
 hill+0.011
 direction+0.011
 directions+0.010
 outage+0.009
Neg logit contributions
hire-0.036
ocket-0.033
adies-0.032
 robber-0.032
 Queen-0.032
Token
Token ID198
Feature activation+1.31e-02

Pos logit contributions
bars+0.019
 outage+0.019
 hill+0.018
 directions+0.017
 direction+0.016
Neg logit contributions
hire-0.068
 Princess-0.062
 Queen-0.062
ocket-0.061
 robber-0.060
Token
Token ID198
Feature activation-1.55e-02

Pos logit contributions
bars+0.010
 outage+0.010
 hill+0.009
 directions+0.009
 direction+0.008
Neg logit contributions
hire-0.032
 Princess-0.030
 Queen-0.029
ocket-0.029
 robber-0.029
TokenWhen
Token ID2215
Feature activation+5.06e-03

Pos logit contributions
hire+0.029
ocket+0.026
 Queen+0.026
 Princess+0.026
 mushrooms+0.025
Neg logit contributions
bars-0.020
 outage-0.020
 hill-0.020
 directions-0.019
spe-0.019
Token trail
Token ID8025
Feature activation-5.85e-03

Pos logit contributions
hire+0.034
 Princess+0.033
 robber+0.032
 Queen+0.031
ocket+0.030
Neg logit contributions
 outage-0.023
bars-0.022
spe-0.021
ney-0.020
umbles-0.019
Token again
Token ID757
Feature activation-4.89e-03

Pos logit contributions
hire+0.013
 Princess+0.012
 Queen+0.011
ocket+0.011
 robber+0.011
Neg logit contributions
 outage-0.006
bars-0.006
 hill-0.006
 directions-0.006
 direction-0.006
Token.
Token ID13
Feature activation+1.86e-02

Pos logit contributions
hire+0.009
 Princess+0.009
 Queen+0.008
ocket+0.008
 robber+0.008
Neg logit contributions
bars-0.006
 outage-0.006
 hill-0.006
 directions-0.006
 direction-0.005
Token
Token ID198
Feature activation+1.02e-02

Pos logit contributions
bars+0.020
 outage+0.019
 hill+0.018
 blowing+0.018
 directions+0.017
Neg logit contributions
hire-0.039
 Princess-0.037
 Queen-0.037
ocket-0.034
 Prince-0.034
Token
Token ID198
Feature activation-1.67e-02

Pos logit contributions
bars+0.013
 outage+0.012
spe+0.011
 hill+0.011
 direction+0.011
Neg logit contributions
hire-0.020
 Princess-0.019
ocket-0.018
 Queen-0.018
adies-0.017
TokenK
Token ID42
Feature activation-6.99e-03

Pos logit contributions
hire+0.028
 Princess+0.025
 Queen+0.024
adies+0.023
ocket+0.023
Neg logit contributions
bars-0.026
 outage-0.025
 hill-0.024
 direction-0.024
 directions-0.023
Tokenisa
Token ID9160
Feature activation+2.38e-02

Pos logit contributions
hire+0.011
 Princess+0.010
 Queen+0.009
ocket+0.009
adies+0.009
Neg logit contributions
 hill-0.011
 outage-0.011
bars-0.011
 motor-0.011
 directions-0.011
Token learnt
Token ID26338
Feature activation-1.21e-02

Pos logit contributions
 outage+0.030
bars+0.029
 hill+0.028
 directions+0.027
 direction+0.026
Neg logit contributions
hire-0.046
 Princess-0.043
 Queen-0.043
ocket-0.041
adies-0.039
Token her
Token ID607
Feature activation+2.85e-03

Pos logit contributions
hire+0.020
 Princess+0.017
 Queen+0.017
ocket+0.017
 robber+0.016
Neg logit contributions
bars-0.019
 outage-0.019
 hill-0.019
 directions-0.018
 direction-0.018
Token lesson
Token ID11483
Feature activation-5.60e-03

Pos logit contributions
bars+0.004
umbling+0.004
 outage+0.003
 hill+0.003
ney+0.003
Neg logit contributions
hire-0.005
 Princess-0.005
 Queen-0.005
ocket-0.005
 robber-0.005
Token -
Token ID532
Feature activation+2.66e-02

Pos logit contributions
hire+0.011
 Princess+0.010
 Queen+0.010
ocket+0.009
adies+0.009
Neg logit contributions
bars-0.007
 hill-0.007
 outage-0.007
 directions-0.007
 direction-0.006
Token smiled
Token ID13541
Feature activation+5.75e-03

Pos logit contributions
 outage+0.002
bars+0.002
 hill+0.002
 directions+0.002
 direction+0.002
Neg logit contributions
hire-0.003
 Princess-0.003
 Queen-0.003
ocket-0.002
 robber-0.002
Token and
Token ID290
Feature activation-9.75e-03

Pos logit contributions
 outage+0.007
bars+0.007
 hill+0.007
 directions+0.006
 direction+0.006
Neg logit contributions
hire-0.012
 Princess-0.011
 Queen-0.011
ocket-0.010
 robber-0.010
Token hugged
Token ID41130
Feature activation-2.56e-02

Pos logit contributions
hire+0.019
 Princess+0.016
adies+0.016
ocket+0.016
 robber+0.016
Neg logit contributions
bars-0.012
 hill-0.012
 outage-0.012
 kay-0.012
 direction-0.012
Token.
Token ID13
Feature activation-7.66e-03

Pos logit contributions
hire+0.044
 Princess+0.040
 Queen+0.039
 robber+0.037
ocket+0.037
Neg logit contributions
 outage-0.039
bars-0.039
 hill-0.037
 direction-0.035
spe-0.035
Token Tom
Token ID4186
Feature activation-4.51e-03

Pos logit contributions
hire+0.014
 Princess+0.013
 Queen+0.013
 robber+0.012
adies+0.012
Neg logit contributions
 outage-0.010
bars-0.010
 hill-0.010
 directions-0.009
 blowing-0.009
Token learned
Token ID4499
Feature activation-2.02e-02

Pos logit contributions
hire+0.009
 Princess+0.008
 Queen+0.008
ocket+0.008
 robber+0.008
Neg logit contributions
bars-0.005
 outage-0.005
 hill-0.005
 directions-0.005
 direction-0.005
Token to
Token ID284
Feature activation+3.58e-04

Pos logit contributions
hire+0.026
 Princess+0.023
 Queen+0.023
 Prince+0.021
 robber+0.021
Neg logit contributions
bars-0.039
 hill-0.038
 outage-0.038
ney-0.037
 direction-0.037
Token be
Token ID307
Feature activation-1.09e-02

Pos logit contributions
bars+0.000
 outage+0.000
spe+0.000
 hill+0.000
 direction+0.000
Neg logit contributions
hire-0.001
 Princess-0.001
 Queen-0.001
 robber-0.001
ocket-0.001
Token kind
Token ID1611
Feature activation+1.07e-02

Pos logit contributions
hire+0.021
 Princess+0.019
 Queen+0.019
ocket+0.019
 robber+0.018
Neg logit contributions
 outage-0.013
bars-0.013
 directions-0.013
 direction-0.013
 hill-0.013
Token and
Token ID290
Feature activation-6.85e-03

Pos logit contributions
bars+0.015
 outage+0.015
 hill+0.014
 kay+0.014
 blowing+0.014
Neg logit contributions
hire-0.019
 Queen-0.018
 Princess-0.018
 robber-0.016
 Prince-0.016
Token respectful
Token ID25923
Feature activation+2.40e-02

Pos logit contributions
hire+0.011
 Queen+0.010
 Princess+0.010
 robber+0.009
ocket+0.009
Neg logit contributions
bars-0.011
 outage-0.011
 hill-0.011
 direction-0.010
icycle-0.010
Token go
Token ID467
Feature activation+7.81e-03

Pos logit contributions
 outage+0.020
 hill+0.020
bars+0.020
 direction+0.018
 kay+0.018
Neg logit contributions
hire-0.034
 Princess-0.031
 Queen-0.030
ocket-0.030
 robber-0.029
Token on
Token ID319
Feature activation+6.10e-03

Pos logit contributions
bars+0.010
 outage+0.010
 hill+0.010
 directions+0.010
 direction+0.010
Neg logit contributions
hire-0.015
 Princess-0.013
 Queen-0.013
ocket-0.013
 robber-0.012
Token it
Token ID340
Feature activation-1.26e-02

Pos logit contributions
 outage+0.009
bars+0.009
 hill+0.008
ney+0.008
 direction+0.008
Neg logit contributions
hire-0.011
 Princess-0.010
 Queen-0.009
ocket-0.009
 robber-0.009
Token,
Token ID11
Feature activation-1.31e-02

Pos logit contributions
hire+0.024
 Princess+0.021
 Queen+0.021
ocket+0.020
 robber+0.020
Neg logit contributions
bars-0.017
 outage-0.017
 hill-0.016
 directions-0.016
 direction-0.016
Token Sara
Token ID24799
Feature activation+4.35e-03

Pos logit contributions
hire+0.026
ocket+0.023
 Queen+0.022
 Princess+0.022
 robber+0.021
Neg logit contributions
 hill-0.017
bars-0.016
 outage-0.016
 kay-0.016
 way-0.015
Token,"
Token ID553
Feature activation-4.45e-02

Pos logit contributions
bars+0.005
 outage+0.005
 hill+0.005
 directions+0.005
 direction+0.005
Neg logit contributions
hire-0.009
 Princess-0.008
 Queen-0.008
ocket-0.008
 robber-0.008
Token Ben
Token ID3932
Feature activation-8.37e-04

Pos logit contributions
hire+0.052
 Princess+0.044
 Queen+0.042
ocket+0.040
 robber+0.039
Neg logit contributions
 outage-0.092
bars-0.091
 hill-0.089
 directions-0.088
 direction-0.087
Token said
Token ID531
Feature activation-1.92e-02

Pos logit contributions
hire+0.002
 Princess+0.001
 Queen+0.001
ocket+0.001
adies+0.001
Neg logit contributions
bars-0.001
 outage-0.001
 hill-0.001
 directions-0.001
 direction-0.001
Token.
Token ID13
Feature activation+1.32e-02

Pos logit contributions
hire+0.035
 Princess+0.033
 Queen+0.032
 robber+0.030
ocket+0.029
Neg logit contributions
 hill-0.026
 outage-0.026
bars-0.026
 direction-0.024
 directions-0.024
Token "
Token ID366
Feature activation-2.90e-03

Pos logit contributions
 outage+0.021
bars+0.020
 hill+0.020
 directions+0.019
 blowing+0.019
Neg logit contributions
hire-0.022
 Princess-0.020
 Queen-0.019
 robber-0.018
 Prince-0.018
TokenIt
Token ID1026
Feature activation+1.72e-02

Pos logit contributions
hire+0.005
 Princess+0.004
 Queen+0.004
ocket+0.004
 robber+0.004
Neg logit contributions
 outage-0.004
bars-0.004
 hill-0.004
 directions-0.004
 direction-0.004
Token spark
Token ID9009
Feature activation-3.61e-02

Pos logit contributions
hire+0.008
adies+0.007
ocket+0.007
 Queen+0.007
 Princess+0.007
Neg logit contributions
bars-0.005
 hill-0.005
 directions-0.005
 outage-0.005
 blowing-0.004
Tokenly
Token ID306
Feature activation-2.99e-02

Pos logit contributions
hire+0.045
 Queen+0.042
ocket+0.041
 Princess+0.041
 robber+0.039
Neg logit contributions
bars-0.070
s-0.070
 outage-0.066
 directions-0.065
 hurts-0.064
Token.
Token ID13
Feature activation-3.55e-03

Pos logit contributions
hire+0.067
 Princess+0.061
 Queen+0.060
ocket+0.060
 robber+0.059
Neg logit contributions
bars-0.028
 outage-0.028
 hill-0.028
 directions-0.027
 direction-0.027
Token You
Token ID921
Feature activation-1.95e-02

Pos logit contributions
hire+0.008
ocket+0.007
vell+0.006
ils+0.006
 Queen+0.006
Neg logit contributions
bars-0.004
 outage-0.004
 directions-0.003
 hill-0.003
 direction-0.003
Token will
Token ID481
Feature activation+1.63e-02

Pos logit contributions
hire+0.046
adies+0.042
 Princess+0.041
ocket+0.041
 Queen+0.041
Neg logit contributions
 hill-0.016
 outage-0.015
bars-0.015
 way-0.013
 kay-0.013
Token look
Token ID804
Feature activation-4.49e-02

Pos logit contributions
 hill+0.024
 outage+0.024
 directions+0.023
 blowing+0.023
 direction+0.023
Neg logit contributions
hire-0.027
adies-0.024
 Queen-0.024
ocket-0.024
 Princess-0.024
Token adorable
Token ID29012
Feature activation-2.75e-02

Pos logit contributions
hire+0.092
pper+0.086
ocket+0.085
 troll+0.083
 robber+0.082
Neg logit contributions
s-0.047
 to-0.044
 again-0.042
 direction-0.042
 directions-0.041
Token in
Token ID287
Feature activation+7.44e-03

Pos logit contributions
hire+0.062
 Princess+0.055
ocket+0.055
 robber+0.055
 Queen+0.054
Neg logit contributions
bars-0.026
 outage-0.025
 hill-0.025
 directions-0.024
 direction-0.023
Token them
Token ID606
Feature activation-3.68e-02

Pos logit contributions
 directions+0.009
 hill+0.008
 direction+0.008
bars+0.008
 outage+0.008
Neg logit contributions
hire-0.015
 Queen-0.014
 Princess-0.013
 robber-0.013
ocket-0.013
Token!"
Token ID2474
Feature activation-2.52e-02

Pos logit contributions
hire+0.075
 Princess+0.068
 Queen+0.067
ocket+0.065
 robber+0.064
Neg logit contributions
 outage-0.043
bars-0.043
 hill-0.041
 directions-0.040
 direction-0.039
Token Lily
Token ID20037
Feature activation+1.68e-02

Pos logit contributions
hire+0.031
 Princess+0.027
 Queen+0.026
ocket+0.024
 robber+0.024
Neg logit contributions
 outage-0.050
bars-0.050
 hill-0.049
 directions-0.048
 direction-0.048
Token little
Token ID1310
Feature activation-1.85e-02

Pos logit contributions
bars+0.003
 outage+0.003
 directions+0.003
 hill+0.003
spe+0.003
Neg logit contributions
hire-0.003
 Princess-0.002
 Queen-0.002
 robber-0.002
ocket-0.002
Token better
Token ID1365
Feature activation-4.39e-03

Pos logit contributions
hire+0.029
 Princess+0.028
 Queen+0.027
 Prince+0.024
 robber+0.024
Neg logit contributions
bars-0.031
 outage-0.029
icycle-0.029
 directions-0.028
spe-0.028
Token.
Token ID13
Feature activation-7.49e-03

Pos logit contributions
hire+0.008
 Princess+0.008
 Queen+0.008
ocket+0.007
 robber+0.007
Neg logit contributions
bars-0.006
 directions-0.006
 hill-0.005
 outage-0.005
 direction-0.005
Token
Token ID198
Feature activation+3.04e-03

Pos logit contributions
hire+0.014
 Princess+0.013
 Queen+0.013
ocket+0.012
 robber+0.012
Neg logit contributions
bars-0.010
 outage-0.010
 hill-0.010
 directions-0.009
 blowing-0.009
Token
Token ID198
Feature activation-2.07e-02

Pos logit contributions
bars+0.005
 outage+0.004
spe+0.004
 hill+0.004
 direction+0.004
Neg logit contributions
hire-0.005
 Princess-0.005
 Queen-0.005
ocket-0.005
 Prince-0.005
Token"
Token ID1
Feature activation-3.91e-02

Pos logit contributions
hire+0.030
 Queen+0.025
 Princess+0.025
adies+0.023
ocket+0.023
Neg logit contributions
bars-0.038
 outage-0.036
 direction-0.035
 hill-0.035
 way-0.035
TokenNext
Token ID10019
Feature activation+3.34e-02

Pos logit contributions
hire+0.068
 Princess+0.060
 Queen+0.059
ocket+0.056
 robber+0.056
Neg logit contributions
bars-0.059
 outage-0.059
 hill-0.058
 direction-0.056
 directions-0.056
Token time
Token ID640
Feature activation+7.76e-03

Pos logit contributions
bars+0.053
 outage+0.050
spe+0.049
bike+0.046
 hill+0.045
Neg logit contributions
 Princess-0.054
hire-0.052
 Queen-0.051
 mushrooms-0.050
adies-0.049
Token,
Token ID11
Feature activation-7.58e-03

Pos logit contributions
bars+0.008
 outage+0.008
 hill+0.008
 directions+0.007
 direction+0.007
Neg logit contributions
hire-0.017
 Princess-0.015
 Queen-0.015
ocket-0.015
 robber-0.015
Token be
Token ID307
Feature activation-6.69e-07

Pos logit contributions
hire+0.013
 Princess+0.011
 Queen+0.011
ocket+0.011
 robber+0.011
Neg logit contributions
 outage-0.011
bars-0.011
 hill-0.011
 directions-0.011
 direction-0.010
Token more
Token ID517
Feature activation+2.13e-02

Feature not active on prev token;
no predictions were affected.
Token to
Token ID284
Feature activation-1.59e-02

Pos logit contributions
hire+0.011
 Princess+0.010
 Queen+0.010
ocket+0.010
 robber+0.009
Neg logit contributions
bars-0.006
 outage-0.006
 hill-0.006
 directions-0.006
 direction-0.005
Token another
Token ID1194
Feature activation+1.26e-02

Pos logit contributions
hire+0.037
vell+0.032
 robber+0.032
ocket+0.031
adies+0.030
Neg logit contributions
 hill-0.016
 direction-0.014
 way-0.014
 kay-0.014
 directions-0.014
Token.
Token ID13
Feature activation-9.03e-04

Pos logit contributions
 hill+0.015
 outage+0.014
 direction+0.014
bars+0.014
 directions+0.013
Neg logit contributions
hire-0.027
ocket-0.023
 Princess-0.023
adies-0.023
 Queen-0.023
Token
Token ID198
Feature activation+9.74e-03

Pos logit contributions
hire+0.002
 Princess+0.001
 Queen+0.001
ocket+0.001
adies+0.001
Neg logit contributions
 outage-0.001
bars-0.001
 hill-0.001
 directions-0.001
 blowing-0.001
Token
Token ID198
Feature activation-1.98e-02

Pos logit contributions
bike+0.015
bars+0.014
 outage+0.014
 blowing+0.013
spe+0.013
Neg logit contributions
 Princess-0.016
hire-0.015
ily-0.015
adies-0.014
 Queen-0.014
Token"
Token ID1
Feature activation-5.28e-02

Pos logit contributions
hire+0.035
 Princess+0.030
 Queen+0.030
ocket+0.029
adies+0.029
Neg logit contributions
 outage-0.029
bars-0.029
 hill-0.029
 direction-0.028
 directions-0.028
TokenLook
Token ID8567
Feature activation-6.22e-02

Pos logit contributions
hire+0.103
ocket+0.091
 Princess+0.089
 robber+0.088
 Queen+0.088
Neg logit contributions
 hill-0.067
bars-0.067
 direction-0.065
 outage-0.064
 directions-0.064
Token,
Token ID11
Feature activation-2.61e-02

Pos logit contributions
hire+0.173
 robber+0.159
ns+0.157
 Queen+0.152
 Princess+0.150
Neg logit contributions
s-0.029
's-0.025
 fan-0.019
 again-0.019
 kay-0.019
Token Ben
Token ID3932
Feature activation-1.82e-03

Pos logit contributions
hire+0.055
ns+0.049
ured+0.046
eh+0.045
ocket+0.044
Neg logit contributions
 hill-0.031
 kay-0.029
 story-0.028
 is-0.028
 way-0.028
Token,
Token ID11
Feature activation-2.39e-02

Pos logit contributions
hire+0.004
ocket+0.004
 Princess+0.004
 Queen+0.004
adies+0.004
Neg logit contributions
 hill-0.001
 outage-0.001
 way-0.001
 direction-0.001
 blowing-0.001
Token a
Token ID257
Feature activation+5.28e-03

Pos logit contributions
hire+0.050
ns+0.044
ured+0.043
ocket+0.042
adies+0.041
Neg logit contributions
 hill-0.028
 kay-0.026
 outage-0.024
 story-0.024
 way-0.024
Token fast
Token ID3049
Feature activation-2.31e-02

Pos logit contributions
bars+0.014
 outage+0.013
 hill+0.013
 direction+0.013
 directions+0.013
Neg logit contributions
hire-0.019
 Princess-0.017
 Queen-0.017
 robber-0.016
adies-0.016
Token as
Token ID355
Feature activation+1.26e-02

Pos logit contributions
hire+0.028
ocket+0.024
 Queen+0.023
 Princess+0.022
adies+0.021
Neg logit contributions
 hill-0.047
bars-0.046
 blowing-0.045
 outage-0.045
 kay-0.045
Token they
Token ID484
Feature activation-3.73e-02

Pos logit contributions
bars+0.016
 outage+0.015
 hill+0.015
 directions+0.015
 direction+0.014
Neg logit contributions
hire-0.025
 Princess-0.022
ocket-0.022
 Queen-0.022
 robber-0.021
Token could
Token ID714
Feature activation+1.76e-02

Pos logit contributions
hire+0.073
 Princess+0.065
 Queen+0.064
ocket+0.063
adies+0.062
Neg logit contributions
bars-0.048
 hill-0.047
 outage-0.046
 direction-0.044
 directions-0.044
Token to
Token ID284
Feature activation-1.53e-02

Pos logit contributions
bars+0.025
 outage+0.025
ney+0.024
 directions+0.023
spe+0.023
Neg logit contributions
hire-0.029
 Princess-0.026
 robber-0.026
 Queen-0.025
adies-0.024
Token see
Token ID766
Feature activation-3.91e-02

Pos logit contributions
hire+0.025
 Princess+0.022
 robber+0.021
 Queen+0.021
 wow+0.020
Neg logit contributions
bars-0.025
 outage-0.024
ney-0.022
spe-0.022
enn-0.022
Token who
Token ID508
Feature activation+3.02e-02

Pos logit contributions
hire+0.078
 Princess+0.069
ocket+0.068
 Queen+0.068
 robber+0.067
Neg logit contributions
bars-0.048
 hill-0.047
 outage-0.047
 directions-0.046
 direction-0.045
Token could
Token ID714
Feature activation+1.84e-02

Pos logit contributions
 outage+0.037
bars+0.036
 hill+0.035
 directions+0.035
 direction+0.034
Neg logit contributions
hire-0.059
 Princess-0.054
 Queen-0.053
ocket-0.051
 robber-0.051
Token reach
Token ID3151
Feature activation+4.62e-03

Pos logit contributions
bars+0.026
 outage+0.026
 direction+0.024
 hill+0.024
 directions+0.024
Neg logit contributions
hire-0.033
 Princess-0.029
 Queen-0.029
 robber-0.028
ocket-0.027
Token the
Token ID262
Feature activation-2.87e-03

Pos logit contributions
bars+0.005
 hill+0.005
 outage+0.005
 directions+0.005
 direction+0.005
Neg logit contributions
hire-0.009
 Princess-0.008
 Queen-0.008
ocket-0.008
 robber-0.008
Token other
Token ID584
Feature activation+1.84e-02

Pos logit contributions
hire+0.006
 Princess+0.005
 Queen+0.005
ocket+0.005
 robber+0.005
Neg logit contributions
 outage-0.003
bars-0.003
 hill-0.003
 directions-0.003
 direction-0.003