Token | she |
Token ID | 673 |
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 ID | 815 |
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 ID | 1464 |
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 ID | 3774 |
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 ID | 607 |
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 ID | 12500 |
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 ID | 290 |
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 ID | 50256 |
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 |
Token | One |
Token ID | 3198 |
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 ID | 1110 |
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 ID | 11 |
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 ID | 41998 |
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 ID | 546 |
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 ID | 508 |
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 ID | 373 |
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 ID | 8290 |
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 |
Token | ver |
Token ID | 332 |
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 ID | 13 |
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 ID | 383 |
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 ID | 21657 |
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 ID | 531 |
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 ID | 339 |
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 ID | 815 |
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 ID | 1464 |
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 ID | 6004 |
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 ID | 284 |
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 ID | 465 |
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 ID | 12500 |
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 ID | 290 |
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 ID | 407 |
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 ID | 466 |
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 ID | 1997 |
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 ID | 1165 |
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 ID | 1088 |
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 ID | 290 |
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 ID | 788 |
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 ID | 2497 |
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 ID | 22772 |
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 ID | 21804 |
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 ID | 1028 |
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 ID | 262 |
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 ID | 25724 |
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 ID | 7988 |
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 ID | 287 |
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 ID | 8161 |
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 ID | 290 |
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 ID | 6004 |
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 ID | 284 |
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 ID | 534 |
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 ID | 12500 |
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 ID | 526 |
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 ID | 198 |
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 ID | 198 |
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 |
Token | Tom |
Token ID | 13787 |
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 ID | 290 |
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 ID | 339 |
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 ID | 2497 |
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 ID | 257 |
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 ID | 1263 |
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 ID | 3085 |
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 ID | 21804 |
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 ID | 1028 |
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 ID | 257 |
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 ID | 5509 |
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 ID | 13 |
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 ID | 220 |
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 ID | 14284 |
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 ID | 625 |
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 ID | 17259 |
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 ID | 11 |
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 ID | 256 |
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 |
Token | ipt |
Token ID | 10257 |
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 |
Token | oe |
Token ID | 2577 |
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 |
Token | ing |
Token ID | 278 |
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 ID | 1088 |
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 ID | 12586 |
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 ID | 11 |
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 ID | 284 |
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 ID | 1282 |
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 ID | 290 |
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 ID | 766 |
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 ID | 262 |
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 ID | 5361 |
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 ID | 21421 |
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 ID | 0 |
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 ID | 679 |
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 ID | 373 |
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 ID | 845 |
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 ID | 2936 |
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 ID | 257 |
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 ID | 1310 |
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 ID | 1643 |
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 ID | 8290 |
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 |
Token | ver |
Token ID | 332 |
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 ID | 13 |
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 ID | 1375 |
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 ID | 6364 |
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 ID | 10764 |
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 ID | 5699 |
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 ID | 4186 |
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 ID | 6810 |
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 ID | 257 |
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 ID | 1263 |
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 ID | 33009 |
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 ID | 21804 |
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 ID | 1028 |
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 ID | 257 |
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 ID | 5509 |
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 ID | 13 |
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 ID | 366 |
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 ID | 467 |
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 ID | 673 |
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 ID | 2936 |
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 ID | 881 |
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 ID | 8290 |
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 |
Token | ver |
Token ID | 332 |
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 ID | 13 |
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 ID | 843 |
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 ID | 673 |
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 ID | 1239 |
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 ID | 2936 |
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 ID | 340 |
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 ID | 925 |
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 ID | 607 |
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 ID | 772 |
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 ID | 8290 |
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 |
Token | ver |
Token ID | 332 |
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 ID | 13 |
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 ID | 50256 |
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 |
Token | M |
Token ID | 44 |
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 |
Token | olly |
Token ID | 5098 |
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 ID | 373 |
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 ID | 2497 |
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 ID | 257 |
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 ID | 18002 |
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 ID | 326 |
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 ID | 373 |
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 ID | 21804 |
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 ID | 1028 |
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 ID | 257 |
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 ID | 5509 |
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 ID | 13 |
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 ID | 679 |
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 ID | 198 |
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 ID | 198 |
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 |
Token | The |
Token ID | 464 |
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 ID | 12967 |
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 ID | 373 |
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 ID | 21804 |
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 ID | 1028 |
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 ID | 257 |
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 ID | 1263 |
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 ID | 5509 |
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 ID | 618 |
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 ID | 15890 |
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 ID | 3066 |
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 ID | 284 |
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 ID | 1011 |
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 ID | 465 |
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 ID | 21804 |
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 ID | 3918 |
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 ID | 284 |
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 ID | 257 |
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 ID | 649 |
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 ID | 1241 |
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 ID | 13 |
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 ID | 6288 |
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 ID | 11 |
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 ID | 673 |
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 ID | 373 |
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 ID | 21804 |
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 ID | 1028 |
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 ID | 262 |
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 ID | 3355 |
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 ID | 287 |
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 ID | 262 |
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 ID | 11 |
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 ID | 484 |
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 ID | 2497 |
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 ID | 262 |
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 ID | 1451 |
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 ID | 21804 |
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 ID | 1028 |
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 ID | 257 |
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 ID | 5509 |
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 ID | 13 |
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 ID | 15890 |
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 ID | 4859 |
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 ID | 262 |
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 ID | 49303 |
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 ID | 319 |
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 ID | 607 |
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 ID | 6727 |
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 ID | 10645 |
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 ID | 351 |
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 ID | 617 |
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 ID | 9154 |
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 ID | 13 |
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 ID | 2440 |
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 ID | 11 |
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 ID | 339 |
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 ID | 2936 |
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 ID | 8290 |
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 |
Token | ver |
Token ID | 332 |
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 ID | 290 |
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 ID | 8290 |
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 |
Token | ver |
Token ID | 332 |
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 ID | 13 |
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 ID | 15894 |
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 |
Token | M |
Token ID | 44 |
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 |
Token | olly |
Token ID | 5098 |
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 ID | 373 |
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 ID | 257 |
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 ID | 845 |
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 ID | 8161 |
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 ID | 2576 |
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 ID | 13 |
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 ID | 1375 |
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 ID | 6151 |
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 ID | 284 |
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 ID | 198 |
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 ID | 1 |
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 |
Token | WHAT |
Token ID | 32971 |
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 ID | 15986 |
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 ID | 7013 |
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 ID | 8410 |
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 |
Token | ING |
Token ID | 2751 |
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 ID | 1701 |
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 ID | 11254 |
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 ID | 42675 |
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 ID | 13 |
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 ID | 290 |
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 ID | 25037 |
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 ID | 290 |
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 ID | 5238 |
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 ID | 290 |
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 ID | 25760 |
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 ID | 13 |
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 ID | 20037 |
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 ID | 373 |
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 ID | 12008 |
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 ID | 13 |
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 ID | 13 |
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 ID | 1375 |
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 ID | 8288 |
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 ID | 2615 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 290 |
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 ID | 13157 |
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 ID | 21547 |
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 |
Token | hell |
Token ID | 12758 |
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 ID | 262 |
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 ID | 10481 |
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 ID | 290 |
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 ID | 1382 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 13 |
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 ID | 220 |
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 ID | 198 |
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 ID | 198 |
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 ID | 11 |
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 ID | 13818 |
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 ID | 290 |
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 ID | 3170 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 13 |
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 ID | 15894 |
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 ID | 11 |
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 ID | 484 |
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 ID | 318 |
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 ID | 2705 |
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 ID | 290 |
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 ID | 7586 |
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 ID | 290 |
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 ID | 25760 |
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 ID | 922 |
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 ID | 13 |
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 ID | 198 |
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 ID | 198 |
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 |
Token | One |
Token ID | 3198 |
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 ID | 13 |
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 ID | 1119 |
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 ID | 561 |
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 ID | 1382 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 290 |
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 ID | 711 |
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 ID | 287 |
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 ID | 262 |
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 ID | 262 |
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 ID | 9151 |
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 ID | 290 |
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 ID | 1382 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 13 |
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 ID | 198 |
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 ID | 198 |
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 |
Token | When |
Token ID | 2215 |
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 ID | 262 |
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 ID | 6450 |
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 ID | 290 |
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 ID | 2615 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 13 |
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 ID | 220 |
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 ID | 198 |
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 ID | 198 |
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 ID | 1375 |
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 ID | 11103 |
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 ID | 262 |
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 ID | 15061 |
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 ID | 290 |
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 ID | 25760 |
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 ID | 340 |
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 ID | 13 |
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 ID | 632 |
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 ID | 468 |
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 ID | 257 |
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 ID | 262 |
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 ID | 10481 |
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 ID | 290 |
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 ID | 1382 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 13 |
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 ID | 887 |
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 ID | 530 |
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 ID | 1110 |
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 ID | 319 |
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 ID | 11 |
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 ID | 484 |
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 ID | 3170 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 1978 |
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 ID | 790 |
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 ID | 5041 |
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 ID | 13 |
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 ID | 13 |
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 ID | 1375 |
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 ID | 561 |
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 ID | 1382 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 290 |
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 ID | 2824 |
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 ID | 2495 |
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 ID | 21547 |
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 ID | 711 |
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 ID | 13 |
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 ID | 1119 |
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 ID | 3170 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 290 |
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 ID | 2826 |
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 ID | 7621 |
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 ID | 13 |
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 ID | 1375 |
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 ID | 8288 |
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 ID | 284 |
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 ID | 1382 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 290 |
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 ID | 2824 |
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 ID | 21547 |
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 |
Token | hell |
Token ID | 12758 |
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 ID | 262 |
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 ID | 10481 |
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 ID | 290 |
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 ID | 2615 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 13 |
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 ID | 1881 |
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 ID | 1110 |
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 ID | 11 |
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 ID | 10481 |
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 ID | 13 |
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 ID | 1119 |
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 ID | 925 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 290 |
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 ID | 4966 |
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 ID | 832 |
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 ID | 262 |
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 ID | 1375 |
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 ID | 8288 |
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 ID | 284 |
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 ID | 1382 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 290 |
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 ID | 2824 |
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 ID | 21547 |
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 |
Token | hell |
Token ID | 12758 |
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 ID | 981 |
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 ID | 11 |
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 ID | 788 |
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 ID | 3170 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 351 |
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 ID | 465 |
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 ID | 1995 |
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 ID | 13 |
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 ID | 19236 |
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 ID | 13 |
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 ID | 1375 |
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 ID | 925 |
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 ID | 6450 |
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 |
Token | cast |
Token ID | 2701 |
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 |
Token | les |
Token ID | 829 |
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 ID | 290 |
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 ID | 6450 |
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 ID | 47199 |
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 ID | 13 |
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 ID | 2826 |
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 ID | 287 |
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 ID | 262 |
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 ID | 14081 |
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 ID | 11105 |
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 ID | 790 |
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 ID | 640 |
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 ID | 484 |
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 ID | 1625 |
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 ID | 284 |
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 ID | 3187 |
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 ID | 262 |
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 ID | 5916 |
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 ID | 531 |
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 ID | 13 |
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 ID | 679 |
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 ID | 2540 |
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 ID | 284 |
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 ID | 1975 |
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 ID | 287 |
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 ID | 2241 |
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 ID | 290 |
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 ID | 11 |
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 ID | 6184 |
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 ID | 95 |
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 ID | 26391 |
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 ID | 129 |
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 ID | 241 |
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 |
Token | Oh |
Token ID | 5812 |
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 ID | 11 |
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 ID | 28796 |
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 ID | 13 |
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 ID | 1320 |
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 ID | 5045 |
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 |
Token | my |
Token ID | 1820 |
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 ID | 338 |
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 ID | 14958 |
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 ID | 547 |
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 ID | 477 |
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 ID | 625 |
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 ID | 262 |
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 ID | 1295 |
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 ID | 780 |
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 ID | 339 |
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 ID | 2793 |
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 ID | 340 |
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 ID | 13 |
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 ID | 3619 |
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 ID | 373 |
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 ID | 8161 |
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 ID | 407 |
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 ID | 284 |
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 ID | 787 |
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 ID | 340 |
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 ID | 1165 |
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 ID | 287 |
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 ID | 262 |
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 ID | 890 |
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 ID | 1057 |
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 ID | 13 |
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 ID | 50256 |
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 |
Token | Once |
Token ID | 7454 |
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 ID | 2402 |
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 ID | 257 |
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 ID | 640 |
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 ID | 612 |
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 ID | 262 |
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 ID | 9592 |
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 ID | 13 |
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 ID | 383 |
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 ID | 9294 |
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 ID | 373 |
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 ID | 1336 |
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 ID | 286 |
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 ID | 257 |
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 ID | 2266 |
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 ID | 10746 |
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 ID | 1139 |
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 ID | 13 |
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 ID | 1375 |
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 ID | 4206 |
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 ID | 339 |
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 ID | 318 |
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 ID | 826 |
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 ID | 13 |
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 ID | 1375 |
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 ID | 857 |
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 ID | 407 |
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 ID | 290 |
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 ID | 3619 |
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 ID | 547 |
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 ID | 523 |
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 ID | 6568 |
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 ID | 284 |
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 ID | 1949 |
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 ID | 340 |
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 ID | 503 |
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 ID | 13 |
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 ID | 198 |
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 ID | 262 |
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 ID | 582 |
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 ID | 34757 |
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 ID | 13 |
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 ID | 366 |
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 |
Token | This |
Token ID | 1212 |
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 ID | 318 |
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 ID | 407 |
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 ID | 257 |
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 ID | 13373 |
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 ID | 11 |
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 ID | 3932 |
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 ID | 547 |
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 ID | 6507 |
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 ID | 13 |
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 ID | 1119 |
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 ID | 2227 |
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 ID | 284 |
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 ID | 711 |
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 ID | 351 |
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 ID | 15899 |
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 ID | 287 |
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 ID | 340 |
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 ID | 656 |
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 ID | 24314 |
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 ID | 290 |
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 ID | 4888 |
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 ID | 340 |
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 ID | 351 |
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 ID | 1123 |
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 ID | 584 |
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 ID | 13 |
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 ID | 632 |
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 ID | 607 |
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 ID | 22526 |
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 ID | 290 |
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 ID | 484 |
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 ID | 900 |
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 ID | 572 |
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 ID | 13 |
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 ID | 220 |
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 ID | 198 |
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 ID | 198 |
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 |
Token | When |
Token ID | 2215 |
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 ID | 8025 |
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 ID | 757 |
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 ID | 13 |
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 ID | 198 |
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 ID | 198 |
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 |
Token | K |
Token ID | 42 |
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 |
Token | isa |
Token ID | 9160 |
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 ID | 26338 |
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 ID | 607 |
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 ID | 11483 |
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 ID | 532 |
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 ID | 13541 |
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 ID | 290 |
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 ID | 41130 |
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 ID | 13 |
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 ID | 4186 |
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 ID | 4499 |
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 ID | 284 |
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 ID | 307 |
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 ID | 1611 |
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 ID | 290 |
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 ID | 25923 |
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 ID | 467 |
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 ID | 319 |
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 ID | 340 |
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 ID | 11 |
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 ID | 24799 |
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 ID | 553 |
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 ID | 3932 |
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 ID | 531 |
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 ID | 13 |
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 ID | 366 |
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 |
Token | It |
Token ID | 1026 |
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 ID | 9009 |
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 |
Token | ly |
Token ID | 306 |
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 ID | 13 |
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 ID | 921 |
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 ID | 481 |
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 ID | 804 |
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 ID | 29012 |
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 ID | 287 |
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 ID | 606 |
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 ID | 2474 |
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 ID | 20037 |
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 ID | 1310 |
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 ID | 1365 |
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 ID | 13 |
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 ID | 198 |
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 ID | 198 |
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 ID | 1 |
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 |
Token | Next |
Token ID | 10019 |
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 ID | 640 |
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 ID | 11 |
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 ID | 307 |
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 ID | 517 |
Feature activation | +2.13e-02 |
Token | to |
Token ID | 284 |
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 ID | 1194 |
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 ID | 13 |
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 ID | 198 |
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 ID | 198 |
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 ID | 1 |
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 |
Token | Look |
Token ID | 8567 |
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 ID | 11 |
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 ID | 3932 |
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 ID | 11 |
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 ID | 257 |
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 ID | 3049 |
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 ID | 355 |
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 ID | 484 |
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 ID | 714 |
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 ID | 284 |
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 ID | 766 |
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 ID | 508 |
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 ID | 714 |
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 ID | 3151 |
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 ID | 262 |
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 ID | 584 |
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 |