Token | must |
Token ID | 1276 |
Feature activation | -3.90e-01 |
Pos logit contributions | |
forehead | +0.019 |
hand | +0.019 |
cheek | +0.018 |
lap | +0.018 |
tight | +0.018 |
Neg logit contributions | |
pand | -0.027 |
pill | -0.025 |
add | -0.025 |
hawk | -0.025 |
otion | -0.024 |
Token | have |
Token ID | 423 |
Feature activation | +2.61e-01 |
Pos logit contributions | |
pand | +0.017 |
add | +0.015 |
otion | +0.015 |
hawk | +0.015 |
die | +0.015 |
Neg logit contributions | |
forehead | -0.011 |
bek | -0.011 |
hand | -0.011 |
cheek | -0.010 |
lap | -0.010 |
Token | mixed |
Token ID | 7668 |
Feature activation | +5.53e-01 |
Pos logit contributions | |
forehead | +0.008 |
hand | +0.007 |
lap | +0.007 |
cheek | +0.007 |
bek | +0.007 |
Neg logit contributions | |
pand | -0.011 |
pill | -0.010 |
hawk | -0.010 |
add | -0.010 |
otion | -0.010 |
Token | up |
Token ID | 510 |
Feature activation | +3.13e-01 |
Pos logit contributions | |
forehead | +0.019 |
lap | +0.017 |
cheek | +0.017 |
hand | +0.017 |
bek | +0.017 |
Neg logit contributions | |
pand | -0.023 |
pill | -0.020 |
add | -0.019 |
storms | -0.019 |
hawk | -0.019 |
Token | the |
Token ID | 262 |
Feature activation | +4.88e-01 |
Pos logit contributions | |
forehead | +0.009 |
lap | +0.009 |
hand | +0.009 |
bek | +0.009 |
cheek | +0.008 |
Neg logit contributions | |
pand | -0.014 |
pill | -0.012 |
storms | -0.012 |
add | -0.012 |
hawk | -0.011 |
Token | addresses |
Token ID | 9405 |
Feature activation | +4.82e+00 |
Pos logit contributions | |
forehead | +0.013 |
bek | +0.013 |
ae | +0.013 |
ossible | +0.013 |
lap | +0.012 |
Neg logit contributions | |
pand | -0.021 |
aliens | -0.019 |
hawk | -0.019 |
otion | -0.018 |
die | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | -5.52e-01 |
Pos logit contributions | |
forehead | +0.157 |
bek | +0.153 |
lap | +0.143 |
list | +0.141 |
cheek | +0.137 |
Neg logit contributions | |
otion | -0.174 |
pand | -0.171 |
storms | -0.158 |
aliens | -0.156 |
pill | -0.150 |
Token | I |
Token ID | 314 |
Feature activation | -5.17e-02 |
Pos logit contributions | |
pand | +0.021 |
pill | +0.019 |
otion | +0.018 |
add | +0.018 |
hawk | +0.018 |
Neg logit contributions | |
forehead | -0.019 |
hand | -0.018 |
cheek | -0.018 |
lap | -0.017 |
tight | -0.017 |
Token | 'm |
Token ID | 1101 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
pand | +0.002 |
pill | +0.002 |
hawk | +0.002 |
add | +0.002 |
otion | +0.002 |
Neg logit contributions | |
forehead | -0.002 |
hand | -0.001 |
lap | -0.001 |
cheek | -0.001 |
tight | -0.001 |
Token | sorry |
Token ID | 7926 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
pand | +0.039 |
pill | +0.035 |
hawk | +0.034 |
add | +0.034 |
otion | +0.034 |
Neg logit contributions | |
forehead | -0.043 |
hand | -0.041 |
cheek | -0.040 |
lap | -0.040 |
tight | -0.040 |
Token | , |
Token ID | 11 |
Feature activation | +3.48e-01 |
Pos logit contributions | |
pand | +0.001 |
pill | +0.001 |
hawk | +0.001 |
add | +0.001 |
otion | +0.001 |
Neg logit contributions | |
forehead | -0.000 |
hand | -0.000 |
cheek | -0.000 |
tight | -0.000 |
lap | -0.000 |
Token | not |
Token ID | 407 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
forehead | +0.040 |
lap | +0.038 |
cheek | +0.037 |
bek | +0.037 |
hand | +0.037 |
Neg logit contributions | |
pand | -0.054 |
hawk | -0.048 |
add | -0.047 |
die | -0.047 |
pill | -0.046 |
Token | tried |
Token ID | 3088 |
Feature activation | +5.49e-01 |
Pos logit contributions | |
forehead | +0.038 |
hand | +0.035 |
lap | +0.035 |
cheek | +0.035 |
bek | +0.034 |
Neg logit contributions | |
pand | -0.055 |
pill | -0.050 |
add | -0.049 |
hawk | -0.049 |
otion | -0.048 |
Token | to |
Token ID | 284 |
Feature activation | +3.04e-01 |
Pos logit contributions | |
forehead | +0.009 |
hand | +0.009 |
cheek | +0.008 |
lap | +0.008 |
tight | +0.008 |
Neg logit contributions | |
pand | -0.030 |
pill | -0.028 |
add | -0.028 |
hawk | -0.028 |
otion | -0.028 |
Token | get |
Token ID | 651 |
Feature activation | +4.10e-01 |
Pos logit contributions | |
forehead | +0.008 |
hand | +0.008 |
lap | +0.008 |
cheek | +0.008 |
bek | +0.008 |
Neg logit contributions | |
pand | -0.013 |
pill | -0.012 |
hawk | -0.012 |
add | -0.012 |
otion | -0.012 |
Token | the |
Token ID | 262 |
Feature activation | +4.65e-01 |
Pos logit contributions | |
forehead | +0.012 |
lap | +0.011 |
hand | +0.011 |
bek | +0.011 |
cheek | +0.011 |
Neg logit contributions | |
pand | -0.018 |
add | -0.016 |
pill | -0.016 |
hawk | -0.016 |
otion | -0.016 |
Token | cream |
Token ID | 8566 |
Feature activation | +4.02e+00 |
Pos logit contributions | |
bek | +0.015 |
forehead | +0.014 |
lap | +0.014 |
lett | +0.013 |
> | +0.013 |
Neg logit contributions | |
pand | -0.019 |
hawk | -0.017 |
otion | -0.016 |
die | -0.016 |
pill | -0.016 |
Token | . |
Token ID | 13 |
Feature activation | -3.73e-01 |
Pos logit contributions | |
forehead | +0.116 |
bek | +0.110 |
lap | +0.105 |
ossible | +0.103 |
cheek | +0.101 |
Neg logit contributions | |
pand | -0.165 |
hawk | -0.151 |
otion | -0.150 |
storms | -0.150 |
die | -0.149 |
Token | She |
Token ID | 1375 |
Feature activation | -9.17e-03 |
Pos logit contributions | |
pand | +0.013 |
pill | +0.012 |
hawk | +0.012 |
add | +0.012 |
otion | +0.012 |
Neg logit contributions | |
forehead | -0.013 |
hand | -0.013 |
lap | -0.013 |
cheek | -0.013 |
tight | -0.012 |
Token | wished |
Token ID | 16555 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
pand | +0.000 |
pill | +0.000 |
hawk | +0.000 |
add | +0.000 |
storms | +0.000 |
Neg logit contributions | |
forehead | -0.000 |
hand | -0.000 |
cheek | -0.000 |
lap | -0.000 |
bek | -0.000 |
Token | her |
Token ID | 607 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
pand | +0.045 |
pill | +0.043 |
hawk | +0.042 |
add | +0.041 |
storms | +0.041 |
Neg logit contributions | |
forehead | -0.039 |
hand | -0.038 |
cheek | -0.037 |
tight | -0.036 |
lap | -0.036 |
Token | mom |
Token ID | 1995 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
bek | +0.033 |
forehead | +0.029 |
lett | +0.028 |
lap | +0.027 |
ae | +0.027 |
Neg logit contributions | |
pand | -0.048 |
nightmares | -0.043 |
hawk | -0.042 |
mosquitoes | -0.041 |
die | -0.040 |
Token | elephant |
Token ID | 20950 |
Feature activation | +9.38e-01 |
Pos logit contributions | |
forehead | +0.014 |
lap | +0.014 |
bek | +0.014 |
hand | +0.014 |
cheek | +0.013 |
Neg logit contributions | |
pand | -0.009 |
hawk | -0.008 |
add | -0.007 |
pill | -0.007 |
Hipp | -0.007 |
Token | always |
Token ID | 1464 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
hand | +0.021 |
forehead | +0.019 |
tight | +0.019 |
whole | +0.019 |
lap | +0.019 |
Neg logit contributions | |
pand | -0.045 |
otion | -0.044 |
add | -0.043 |
pill | -0.043 |
hawk | -0.042 |
Token | looked |
Token ID | 3114 |
Feature activation | -3.67e-01 |
Pos logit contributions | |
forehead | +0.004 |
hand | +0.004 |
cheek | +0.004 |
tight | +0.004 |
lap | +0.004 |
Neg logit contributions | |
pand | -0.008 |
pill | -0.008 |
add | -0.008 |
otion | -0.008 |
hawk | -0.008 |
Token | at |
Token ID | 379 |
Feature activation | -1.95e-01 |
Pos logit contributions | |
pand | +0.018 |
add | +0.017 |
hawk | +0.017 |
Tweet | +0.016 |
aliens | +0.016 |
Neg logit contributions | |
forehead | -0.008 |
cheek | -0.008 |
bek | -0.007 |
lap | -0.007 |
hand | -0.007 |
Token | the |
Token ID | 262 |
Feature activation | +5.68e-01 |
Pos logit contributions | |
pill | +0.010 |
pand | +0.010 |
otion | +0.010 |
add | +0.009 |
storms | +0.009 |
Neg logit contributions | |
hand | -0.004 |
tight | -0.004 |
forehead | -0.003 |
whole | -0.003 |
hands | -0.003 |
Token | labels |
Token ID | 14722 |
Feature activation | +3.99e+00 |
Pos logit contributions | |
forehead | +0.023 |
bek | +0.022 |
lap | +0.022 |
hand | +0.021 |
cheek | +0.021 |
Neg logit contributions | |
pand | -0.019 |
hawk | -0.017 |
pill | -0.016 |
flashes | -0.015 |
Hipp | -0.015 |
Token | before |
Token ID | 878 |
Feature activation | -3.57e-01 |
Pos logit contributions | |
forehead | +0.097 |
hand | +0.090 |
lap | +0.087 |
cheek | +0.087 |
bek | +0.085 |
Neg logit contributions | |
pand | -0.188 |
pill | -0.175 |
otion | -0.174 |
hawk | -0.173 |
add | -0.172 |
Token | he |
Token ID | 339 |
Feature activation | +6.42e-02 |
Pos logit contributions | |
pand | +0.014 |
pill | +0.014 |
add | +0.014 |
otion | +0.013 |
ican | +0.013 |
Neg logit contributions | |
hand | -0.011 |
tight | -0.011 |
front | -0.010 |
whole | -0.010 |
forehead | -0.010 |
Token | went |
Token ID | 1816 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
forehead | +0.002 |
hand | +0.002 |
lap | +0.002 |
cheek | +0.002 |
tight | +0.002 |
Neg logit contributions | |
pand | -0.003 |
pill | -0.003 |
hawk | -0.003 |
add | -0.003 |
otion | -0.003 |
Token | somewhere |
Token ID | 7382 |
Feature activation | -1.83e+00 |
Pos logit contributions | |
pand | +0.066 |
pill | +0.063 |
otion | +0.062 |
add | +0.062 |
hawk | +0.062 |
Neg logit contributions | |
forehead | -0.024 |
hand | -0.023 |
lap | -0.022 |
tight | -0.021 |
cheek | -0.021 |
Token | new |
Token ID | 649 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
pand | +0.065 |
pill | +0.060 |
add | +0.059 |
otion | +0.059 |
hawk | +0.058 |
Neg logit contributions | |
forehead | -0.066 |
hand | -0.064 |
cheek | -0.062 |
lap | -0.062 |
tight | -0.062 |
Token | everywhere |
Token ID | 8347 |
Feature activation | +7.77e-01 |
Pos logit contributions | |
pand | +0.007 |
add | +0.006 |
pill | +0.006 |
aliens | +0.006 |
storms | +0.006 |
Neg logit contributions | |
forehead | -0.004 |
bek | -0.004 |
lap | -0.003 |
cheek | -0.003 |
hand | -0.003 |
Token | until |
Token ID | 1566 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
bek | +0.028 |
forehead | +0.027 |
ossible | +0.026 |
lap | +0.025 |
bara | +0.024 |
Neg logit contributions | |
pand | -0.027 |
Tweet | -0.024 |
storms | -0.024 |
aliens | -0.023 |
Hipp | -0.023 |
Token | he |
Token ID | 339 |
Feature activation | +5.36e-01 |
Pos logit contributions | |
forehead | +0.003 |
cheek | +0.003 |
hand | +0.003 |
lap | +0.003 |
bek | +0.003 |
Neg logit contributions | |
pand | -0.004 |
hawk | -0.004 |
otion | -0.004 |
pill | -0.004 |
add | -0.004 |
Token | found |
Token ID | 1043 |
Feature activation | +6.05e-01 |
Pos logit contributions | |
hand | +0.012 |
whole | +0.011 |
tight | +0.010 |
forehead | +0.010 |
plate | +0.010 |
Neg logit contributions | |
pand | -0.026 |
pill | -0.025 |
hawk | -0.025 |
otion | -0.025 |
add | -0.024 |
Token | the |
Token ID | 262 |
Feature activation | +7.65e-01 |
Pos logit contributions | |
forehead | +0.018 |
cheek | +0.017 |
lap | +0.017 |
bek | +0.016 |
hand | +0.016 |
Neg logit contributions | |
pand | -0.026 |
pill | -0.024 |
hawk | -0.023 |
add | -0.023 |
otion | -0.023 |
Token | labels |
Token ID | 14722 |
Feature activation | +3.93e+00 |
Pos logit contributions | |
forehead | +0.025 |
bek | +0.024 |
lap | +0.023 |
ossible | +0.023 |
cheek | +0.022 |
Neg logit contributions | |
pand | -0.032 |
pill | -0.027 |
aliens | -0.026 |
lions | -0.026 |
hawk | -0.026 |
Token | . |
Token ID | 13 |
Feature activation | +9.59e-01 |
Pos logit contributions | |
forehead | +0.112 |
bek | +0.107 |
cheek | +0.100 |
lap | +0.100 |
ossible | +0.098 |
Neg logit contributions | |
pand | -0.166 |
otion | -0.163 |
storms | -0.153 |
pill | -0.150 |
hawk | -0.150 |
Token | |
Token ID | 220 |
Feature activation | +4.79e-01 |
Pos logit contributions | |
forehead | +0.030 |
cheek | +0.027 |
bek | +0.027 |
hand | +0.026 |
lap | +0.025 |
Neg logit contributions | |
pand | -0.039 |
pill | -0.038 |
add | -0.037 |
storms | -0.036 |
otion | -0.036 |
Token |
|
Token ID | 198 |
Feature activation | +8.42e-01 |
Pos logit contributions | |
forehead | +0.016 |
bek | +0.014 |
cheek | +0.014 |
hand | +0.013 |
plate | +0.013 |
Neg logit contributions | |
pand | -0.019 |
otion | -0.018 |
add | -0.018 |
pill | -0.018 |
storms | -0.017 |
Token |
|
Token ID | 198 |
Feature activation | +5.89e-01 |
Pos logit contributions | |
forehead | +0.028 |
cheek | +0.025 |
bek | +0.025 |
hand | +0.024 |
lap | +0.023 |
Neg logit contributions | |
pand | -0.033 |
add | -0.032 |
pill | -0.032 |
otion | -0.031 |
storms | -0.030 |
Token | One |
Token ID | 3198 |
Feature activation | +8.87e-01 |
Pos logit contributions | |
forehead | +0.026 |
cheek | +0.025 |
hand | +0.024 |
bek | +0.023 |
lap | +0.023 |
Neg logit contributions | |
pand | -0.017 |
pill | -0.016 |
add | -0.015 |
hawk | -0.014 |
storms | -0.014 |
Token | She |
Token ID | 1375 |
Feature activation | -5.45e-01 |
Pos logit contributions | |
pand | +0.008 |
pill | +0.008 |
otion | +0.008 |
storms | +0.008 |
add | +0.007 |
Neg logit contributions | |
forehead | -0.008 |
cheek | -0.008 |
bek | -0.008 |
hand | -0.008 |
lap | -0.007 |
Token | would |
Token ID | 561 |
Feature activation | -2.98e-01 |
Pos logit contributions | |
pand | +0.026 |
otion | +0.025 |
pill | +0.024 |
hawk | +0.024 |
add | +0.024 |
Neg logit contributions | |
forehead | -0.013 |
hand | -0.012 |
tight | -0.012 |
lap | -0.012 |
cheek | -0.012 |
Token | always |
Token ID | 1464 |
Feature activation | +3.22e-01 |
Pos logit contributions | |
pand | +0.013 |
pill | +0.013 |
add | +0.013 |
otion | +0.013 |
hawk | +0.013 |
Neg logit contributions | |
forehead | -0.008 |
hand | -0.008 |
tight | -0.007 |
cheek | -0.007 |
lap | -0.007 |
Token | check |
Token ID | 2198 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
forehead | +0.008 |
hand | +0.008 |
lap | +0.008 |
cheek | +0.008 |
tight | +0.007 |
Neg logit contributions | |
pand | -0.015 |
pill | -0.014 |
hawk | -0.014 |
add | -0.014 |
otion | -0.013 |
Token | the |
Token ID | 262 |
Feature activation | +8.21e-02 |
Pos logit contributions | |
forehead | +0.032 |
lap | +0.030 |
cheek | +0.030 |
bek | +0.030 |
hand | +0.029 |
Neg logit contributions | |
pand | -0.043 |
otion | -0.040 |
pill | -0.038 |
add | -0.037 |
hawk | -0.037 |
Token | labels |
Token ID | 14722 |
Feature activation | +3.92e+00 |
Pos logit contributions | |
forehead | +0.003 |
hand | +0.003 |
cheek | +0.003 |
lap | +0.003 |
tight | +0.003 |
Neg logit contributions | |
pand | -0.003 |
pill | -0.003 |
otion | -0.003 |
add | -0.003 |
hawk | -0.003 |
Token | before |
Token ID | 878 |
Feature activation | -5.07e-01 |
Pos logit contributions | |
forehead | +0.099 |
hand | +0.090 |
lap | +0.089 |
cheek | +0.089 |
bek | +0.088 |
Neg logit contributions | |
pand | -0.179 |
otion | -0.168 |
pill | -0.164 |
hawk | -0.163 |
storms | -0.162 |
Token | she |
Token ID | 673 |
Feature activation | -1.00e+00 |
Pos logit contributions | |
pand | +0.021 |
add | +0.020 |
pill | +0.020 |
hawk | +0.020 |
otion | +0.019 |
Neg logit contributions | |
forehead | -0.015 |
hand | -0.015 |
tight | -0.014 |
cheek | -0.014 |
lap | -0.014 |
Token | bat |
Token ID | 7365 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
pand | +0.046 |
hawk | +0.044 |
otion | +0.043 |
pill | +0.042 |
add | +0.041 |
Neg logit contributions | |
hand | -0.026 |
forehead | -0.026 |
lap | -0.024 |
tight | -0.024 |
cheek | -0.023 |
Token | hed |
Token ID | 704 |
Feature activation | +6.03e-02 |
Pos logit contributions | |
pand | +0.070 |
otion | +0.066 |
pill | +0.066 |
add | +0.065 |
storms | +0.064 |
Neg logit contributions | |
forehead | -0.031 |
hand | -0.029 |
lap | -0.028 |
cheek | -0.027 |
bek | -0.027 |
Token | her |
Token ID | 607 |
Feature activation | -2.14e-01 |
Pos logit contributions | |
forehead | +0.001 |
lap | +0.001 |
hand | +0.001 |
cheek | +0.001 |
bek | +0.001 |
Neg logit contributions | |
pand | -0.003 |
pill | -0.003 |
otion | -0.003 |
add | -0.003 |
hawk | -0.003 |
Token | want |
Token ID | 765 |
Feature activation | +8.96e-01 |
Pos logit contributions | |
pand | +0.011 |
pill | +0.010 |
hawk | +0.010 |
add | +0.010 |
otion | +0.010 |
Neg logit contributions | |
forehead | -0.007 |
hand | -0.007 |
lap | -0.007 |
cheek | -0.007 |
bek | -0.007 |
Token | to |
Token ID | 284 |
Feature activation | +2.94e-01 |
Pos logit contributions | |
forehead | +0.022 |
hand | +0.021 |
lap | +0.020 |
cheek | +0.020 |
tight | +0.019 |
Neg logit contributions | |
pand | -0.043 |
pill | -0.040 |
hawk | -0.039 |
add | -0.039 |
otion | -0.039 |
Token | buy |
Token ID | 2822 |
Feature activation | +1.59e+00 |
Pos logit contributions | |
forehead | +0.007 |
hand | +0.007 |
cheek | +0.007 |
lap | +0.007 |
tight | +0.007 |
Neg logit contributions | |
pand | -0.014 |
pill | -0.013 |
add | -0.013 |
otion | -0.013 |
hawk | -0.013 |
Token | an |
Token ID | 281 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
forehead | +0.047 |
hand | +0.045 |
lap | +0.043 |
cheek | +0.043 |
tight | +0.042 |
Neg logit contributions | |
pand | -0.067 |
pill | -0.062 |
hawk | -0.061 |
add | -0.061 |
otion | -0.060 |
Token | ice |
Token ID | 4771 |
Feature activation | +2.70e-01 |
Pos logit contributions | |
pand | +0.004 |
add | +0.004 |
hawk | +0.004 |
pill | +0.004 |
otion | +0.004 |
Neg logit contributions | |
forehead | -0.004 |
lap | -0.004 |
bek | -0.004 |
cheek | -0.004 |
hand | -0.004 |
Token | cream |
Token ID | 8566 |
Feature activation | +3.87e+00 |
Pos logit contributions | |
forehead | +0.007 |
bek | +0.006 |
lap | +0.006 |
hand | +0.006 |
doorway | +0.006 |
Neg logit contributions | |
pand | -0.013 |
Hipp | -0.011 |
add | -0.011 |
hon | -0.011 |
hawk | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | +6.99e-01 |
Pos logit contributions | |
forehead | +0.126 |
bek | +0.113 |
lap | +0.112 |
hand | +0.107 |
cheek | +0.106 |
Neg logit contributions | |
pand | -0.155 |
Hipp | -0.138 |
hawk | -0.135 |
otion | -0.132 |
storms | -0.132 |
Token | then |
Token ID | 788 |
Feature activation | +5.81e-01 |
Pos logit contributions | |
forehead | +0.020 |
hand | +0.020 |
tight | +0.019 |
cheek | +0.019 |
lap | +0.019 |
Neg logit contributions | |
pand | -0.029 |
pill | -0.028 |
add | -0.027 |
otion | -0.027 |
storms | -0.027 |
Token | let |
Token ID | 1309 |
Feature activation | -6.76e-01 |
Pos logit contributions | |
forehead | +0.018 |
hand | +0.017 |
lap | +0.016 |
cheek | +0.016 |
tight | +0.016 |
Neg logit contributions | |
pand | -0.024 |
pill | -0.022 |
hawk | -0.022 |
add | -0.022 |
otion | -0.022 |
Token | 's |
Token ID | 338 |
Feature activation | -8.23e-01 |
Pos logit contributions | |
pand | +0.032 |
pill | +0.030 |
add | +0.029 |
otion | +0.029 |
hawk | +0.029 |
Neg logit contributions | |
forehead | -0.017 |
hand | -0.017 |
tight | -0.016 |
cheek | -0.016 |
lap | -0.015 |
Token | give |
Token ID | 1577 |
Feature activation | +2.15e-01 |
Pos logit contributions | |
pand | +0.034 |
hawk | +0.031 |
add | +0.031 |
otion | +0.030 |
pill | +0.030 |
Neg logit contributions | |
forehead | -0.025 |
lap | -0.023 |
cheek | -0.023 |
hand | -0.023 |
bek | -0.023 |
Token | They |
Token ID | 1119 |
Feature activation | -5.42e-01 |
Pos logit contributions | |
pand | +0.027 |
hawk | +0.025 |
pill | +0.025 |
add | +0.024 |
otion | +0.024 |
Neg logit contributions | |
hand | -0.024 |
forehead | -0.024 |
lap | -0.023 |
tight | -0.023 |
cheek | -0.022 |
Token | never |
Token ID | 1239 |
Feature activation | +2.33e-01 |
Pos logit contributions | |
pand | +0.023 |
pill | +0.022 |
add | +0.021 |
hawk | +0.021 |
otion | +0.021 |
Neg logit contributions | |
forehead | -0.015 |
hand | -0.015 |
cheek | -0.014 |
lap | -0.014 |
bek | -0.013 |
Token | left |
Token ID | 1364 |
Feature activation | +1.64e-01 |
Pos logit contributions | |
forehead | +0.006 |
hand | +0.006 |
cheek | +0.005 |
lap | +0.005 |
tight | +0.005 |
Neg logit contributions | |
pand | -0.011 |
pill | -0.010 |
otion | -0.010 |
add | -0.010 |
hawk | -0.010 |
Token | the |
Token ID | 262 |
Feature activation | +9.56e-01 |
Pos logit contributions | |
forehead | +0.005 |
hand | +0.005 |
lap | +0.005 |
tight | +0.005 |
cheek | +0.005 |
Neg logit contributions | |
pand | -0.007 |
pill | -0.006 |
hawk | -0.006 |
add | -0.006 |
otion | -0.006 |
Token | ice |
Token ID | 4771 |
Feature activation | +5.44e-01 |
Pos logit contributions | |
bek | +0.029 |
forehead | +0.029 |
lap | +0.027 |
hand | +0.026 |
ae | +0.026 |
Neg logit contributions | |
pand | -0.039 |
hawk | -0.035 |
pill | -0.033 |
flashes | -0.033 |
aliens | -0.033 |
Token | cream |
Token ID | 8566 |
Feature activation | +3.86e+00 |
Pos logit contributions | |
forehead | +0.012 |
bek | +0.011 |
lap | +0.011 |
hand | +0.010 |
ossible | +0.010 |
Neg logit contributions | |
pand | -0.027 |
add | -0.024 |
otion | -0.024 |
hawk | -0.024 |
storms | -0.024 |
Token | out |
Token ID | 503 |
Feature activation | +2.76e-01 |
Pos logit contributions | |
forehead | +0.157 |
bek | +0.149 |
ossible | +0.144 |
lap | +0.142 |
cheek | +0.141 |
Neg logit contributions | |
pand | -0.117 |
hawk | -0.101 |
otion | -0.100 |
storms | -0.098 |
anymore | -0.097 |
Token | of |
Token ID | 286 |
Feature activation | -6.83e-01 |
Pos logit contributions | |
forehead | +0.009 |
cheek | +0.008 |
bek | +0.008 |
lap | +0.008 |
list | +0.008 |
Neg logit contributions | |
pand | -0.011 |
otion | -0.010 |
pill | -0.010 |
add | -0.010 |
hawk | -0.010 |
Token | the |
Token ID | 262 |
Feature activation | +1.80e-01 |
Pos logit contributions | |
pand | +0.030 |
hawk | +0.027 |
otion | +0.026 |
add | +0.026 |
pill | +0.026 |
Neg logit contributions | |
forehead | -0.020 |
bek | -0.019 |
cheek | -0.018 |
lap | -0.018 |
hand | -0.017 |
Token | freezer |
Token ID | 30967 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
forehead | +0.006 |
bek | +0.006 |
ae | +0.005 |
cheek | +0.005 |
lap | +0.005 |
Neg logit contributions | |
pand | -0.007 |
hawk | -0.007 |
spr | -0.006 |
die | -0.006 |
flashes | -0.006 |
Token | again |
Token ID | 757 |
Feature activation | -1.49e+00 |
Pos logit contributions | |
forehead | +0.062 |
bek | +0.058 |
ossible | +0.058 |
cheek | +0.053 |
lap | +0.053 |
Neg logit contributions | |
pand | -0.069 |
storms | -0.064 |
otion | -0.063 |
anymore | -0.062 |
spr | -0.062 |
Token | ! |
Token ID | 0 |
Feature activation | -9.64e-01 |
Pos logit contributions | |
lap | +0.014 |
forehead | +0.014 |
bek | +0.013 |
cheek | +0.012 |
hand | +0.012 |
Neg logit contributions | |
pand | -0.017 |
hawk | -0.015 |
otion | -0.015 |
happens | -0.015 |
scares | -0.014 |
Token | This |
Token ID | 770 |
Feature activation | -6.61e-01 |
Pos logit contributions | |
pand | +0.042 |
hawk | +0.040 |
add | +0.039 |
aliens | +0.038 |
pill | +0.038 |
Neg logit contributions | |
hand | -0.027 |
forehead | -0.026 |
cheek | -0.025 |
lap | -0.025 |
plate | -0.024 |
Token | is |
Token ID | 318 |
Feature activation | +1.83e-02 |
Pos logit contributions | |
pand | +0.026 |
pill | +0.024 |
otion | +0.024 |
storms | +0.024 |
add | +0.024 |
Neg logit contributions | |
hand | -0.020 |
forehead | -0.020 |
whole | -0.020 |
tight | -0.019 |
cheek | -0.019 |
Token | not |
Token ID | 407 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
forehead | +0.001 |
hand | +0.001 |
lap | +0.001 |
cheek | +0.001 |
bek | +0.001 |
Neg logit contributions | |
pand | -0.001 |
pill | -0.001 |
hawk | -0.001 |
add | -0.001 |
otion | -0.001 |
Token | ice |
Token ID | 4771 |
Feature activation | +6.27e-01 |
Pos logit contributions | |
forehead | +0.048 |
lap | +0.045 |
hand | +0.044 |
cheek | +0.043 |
bek | +0.042 |
Neg logit contributions | |
pand | -0.069 |
hawk | -0.062 |
pill | -0.062 |
add | -0.061 |
otion | -0.060 |
Token | cream |
Token ID | 8566 |
Feature activation | +3.86e+00 |
Pos logit contributions | |
forehead | +0.016 |
lap | +0.015 |
hand | +0.014 |
bek | +0.014 |
ossible | +0.013 |
Neg logit contributions | |
pand | -0.030 |
pill | -0.026 |
add | -0.026 |
storms | -0.026 |
otion | -0.026 |
Token | !" |
Token ID | 2474 |
Feature activation | -4.72e-01 |
Pos logit contributions | |
forehead | +0.093 |
lap | +0.085 |
hand | +0.084 |
bek | +0.081 |
cheek | +0.078 |
Neg logit contributions | |
pand | -0.187 |
pill | -0.170 |
hawk | -0.169 |
otion | -0.168 |
storms | -0.168 |
Token | Lily |
Token ID | 20037 |
Feature activation | -6.09e-01 |
Pos logit contributions | |
pand | +0.017 |
pill | +0.015 |
hawk | +0.015 |
otion | +0.015 |
add | +0.015 |
Neg logit contributions | |
forehead | -0.017 |
cheek | -0.016 |
bek | -0.016 |
hand | -0.016 |
lap | -0.016 |
Token | cried |
Token ID | 16896 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
pand | +0.027 |
pill | +0.025 |
storms | +0.024 |
hawk | +0.024 |
flashes | +0.024 |
Neg logit contributions | |
forehead | -0.016 |
bek | -0.015 |
whole | -0.015 |
lap | -0.015 |
cheek | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | -3.19e-01 |
Pos logit contributions | |
pill | +0.079 |
pand | +0.078 |
hawk | +0.076 |
otion | +0.074 |
add | +0.073 |
Neg logit contributions | |
hand | -0.018 |
forehead | -0.017 |
tight | -0.016 |
cheek | -0.014 |
lap | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -5.97e-02 |
Pos logit contributions | |
pand | +0.016 |
pill | +0.015 |
add | +0.015 |
hawk | +0.015 |
otion | +0.015 |
Neg logit contributions | |
hand | -0.007 |
forehead | -0.006 |
lap | -0.006 |
tight | -0.006 |
cheek | -0.006 |
Token | they |
Token ID | 484 |
Feature activation | +9.67e-01 |
Pos logit contributions | |
pand | +0.029 |
Tweet | +0.023 |
add | +0.022 |
hawk | +0.022 |
otion | +0.022 |
Neg logit contributions | |
forehead | -0.053 |
cheek | -0.050 |
lap | -0.050 |
hand | -0.049 |
bek | -0.048 |
Token | arrived |
Token ID | 5284 |
Feature activation | +8.99e-01 |
Pos logit contributions | |
forehead | +0.036 |
cheek | +0.034 |
bek | +0.031 |
lap | +0.031 |
hand | +0.029 |
Neg logit contributions | |
pand | -0.037 |
add | -0.035 |
otion | -0.032 |
hon | -0.030 |
Hipp | -0.030 |
Token | at |
Token ID | 379 |
Feature activation | +9.26e-02 |
Pos logit contributions | |
forehead | +0.033 |
cheek | +0.031 |
bek | +0.030 |
lap | +0.030 |
plate | +0.028 |
Neg logit contributions | |
pand | -0.033 |
add | -0.031 |
otion | -0.029 |
hawk | -0.028 |
Tweet | -0.027 |
Token | the |
Token ID | 262 |
Feature activation | +7.06e-01 |
Pos logit contributions | |
forehead | +0.003 |
cheek | +0.003 |
bek | +0.003 |
lap | +0.003 |
hips | +0.003 |
Neg logit contributions | |
pand | -0.004 |
die | -0.003 |
Tweet | -0.003 |
add | -0.003 |
hawk | -0.003 |
Token | ice |
Token ID | 4771 |
Feature activation | +5.35e-01 |
Pos logit contributions | |
forehead | +0.024 |
cheek | +0.022 |
hand | +0.022 |
lap | +0.022 |
bek | +0.022 |
Neg logit contributions | |
pand | -0.028 |
hawk | -0.024 |
pill | -0.024 |
add | -0.024 |
otion | -0.024 |
Token | cream |
Token ID | 8566 |
Feature activation | +3.86e+00 |
Pos logit contributions | |
forehead | +0.014 |
hand | +0.012 |
bek | +0.012 |
cheek | +0.011 |
lap | +0.011 |
Neg logit contributions | |
pand | -0.026 |
otion | -0.023 |
add | -0.023 |
pill | -0.023 |
ican | -0.022 |
Token | shop |
Token ID | 6128 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
forehead | +0.124 |
cheek | +0.111 |
bek | +0.107 |
hand | +0.107 |
lap | +0.106 |
Neg logit contributions | |
pand | -0.160 |
pill | -0.142 |
otion | -0.142 |
hawk | -0.141 |
storms | -0.138 |
Token | , |
Token ID | 11 |
Feature activation | +5.99e-01 |
Pos logit contributions | |
forehead | +0.068 |
lap | +0.063 |
cheek | +0.062 |
bek | +0.061 |
hand | +0.061 |
Neg logit contributions | |
pand | -0.068 |
otion | -0.060 |
pill | -0.058 |
storms | -0.058 |
add | -0.058 |
Token | Lily |
Token ID | 20037 |
Feature activation | +4.34e-01 |
Pos logit contributions | |
forehead | +0.024 |
cheek | +0.023 |
lap | +0.023 |
bek | +0.022 |
hand | +0.022 |
Neg logit contributions | |
pand | -0.019 |
otion | -0.017 |
add | -0.016 |
hawk | -0.016 |
Tweet | -0.016 |
Token | saw |
Token ID | 2497 |
Feature activation | +4.38e-01 |
Pos logit contributions | |
forehead | +0.014 |
cheek | +0.013 |
bek | +0.013 |
attention | +0.013 |
lap | +0.012 |
Neg logit contributions | |
pand | -0.017 |
pill | -0.016 |
add | -0.015 |
storms | -0.015 |
spr | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | +2.59e-01 |
Pos logit contributions | |
hand | +0.009 |
forehead | +0.009 |
tight | +0.009 |
whole | +0.008 |
plate | +0.008 |
Neg logit contributions | |
pand | -0.021 |
pill | -0.021 |
storms | -0.021 |
otion | -0.021 |
add | -0.021 |
Token | We |
Token ID | 775 |
Feature activation | +5.82e-01 |
Pos logit contributions | |
pand | +0.030 |
hawk | +0.028 |
pill | +0.028 |
add | +0.028 |
otion | +0.027 |
Neg logit contributions | |
forehead | -0.021 |
hand | -0.021 |
lap | -0.020 |
cheek | -0.020 |
tight | -0.020 |
Token | have |
Token ID | 423 |
Feature activation | +1.74e-01 |
Pos logit contributions | |
forehead | +0.021 |
hand | +0.020 |
lap | +0.019 |
cheek | +0.019 |
tight | +0.019 |
Neg logit contributions | |
pand | -0.021 |
pill | -0.019 |
hawk | -0.019 |
add | -0.019 |
otion | -0.019 |
Token | to |
Token ID | 284 |
Feature activation | -3.78e-01 |
Pos logit contributions | |
forehead | +0.005 |
cheek | +0.005 |
lap | +0.005 |
bek | +0.005 |
chin | +0.004 |
Neg logit contributions | |
pand | -0.008 |
pill | -0.007 |
add | -0.007 |
hawk | -0.007 |
conflicts | -0.007 |
Token | pay |
Token ID | 1414 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
pand | +0.015 |
die | +0.014 |
hawk | +0.014 |
add | +0.014 |
otion | +0.013 |
Neg logit contributions | |
forehead | -0.011 |
bek | -0.010 |
lap | -0.010 |
cheek | -0.010 |
hand | -0.010 |
Token | the |
Token ID | 262 |
Feature activation | +2.50e-01 |
Pos logit contributions | |
bek | +0.039 |
lap | +0.039 |
forehead | +0.038 |
doorway | +0.037 |
pron | +0.037 |
Neg logit contributions | |
pand | -0.050 |
add | -0.044 |
happens | -0.044 |
die | -0.044 |
them | -0.043 |
Token | bills |
Token ID | 9024 |
Feature activation | +3.84e+00 |
Pos logit contributions | |
forehead | +0.008 |
lap | +0.008 |
bek | +0.008 |
ae | +0.008 |
hips | +0.007 |
Neg logit contributions | |
pand | -0.010 |
hawk | -0.009 |
helicopter | -0.008 |
die | -0.008 |
Hipp | -0.008 |
Token | first |
Token ID | 717 |
Feature activation | +2.96e-01 |
Pos logit contributions | |
forehead | +0.123 |
lap | +0.121 |
bek | +0.115 |
ossible | +0.108 |
doorway | +0.107 |
Neg logit contributions | |
pand | -0.134 |
otion | -0.132 |
anymore | -0.121 |
die | -0.119 |
storms | -0.118 |
Token | ." |
Token ID | 526 |
Feature activation | -9.57e-02 |
Pos logit contributions | |
forehead | +0.009 |
bek | +0.009 |
cheek | +0.008 |
lap | +0.008 |
hand | +0.008 |
Neg logit contributions | |
pand | -0.012 |
storms | -0.011 |
hawk | -0.010 |
otion | -0.010 |
die | -0.010 |
Token |
|
Token ID | 198 |
Feature activation | -6.79e-01 |
Pos logit contributions | |
pand | +0.004 |
pill | +0.003 |
add | +0.003 |
storms | +0.003 |
otion | +0.003 |
Neg logit contributions | |
forehead | -0.003 |
hand | -0.003 |
bek | -0.003 |
cheek | -0.003 |
lap | -0.003 |
Token |
|
Token ID | 198 |
Feature activation | -8.30e-01 |
Pos logit contributions | |
pand | +0.026 |
add | +0.025 |
pill | +0.024 |
otion | +0.024 |
Tweet | +0.024 |
Neg logit contributions | |
forehead | -0.022 |
bek | -0.021 |
cheek | -0.020 |
hand | -0.020 |
lap | -0.019 |
Token | L |
Token ID | 43 |
Feature activation | +4.46e-02 |
Pos logit contributions | |
pand | +0.032 |
pill | +0.030 |
add | +0.029 |
otion | +0.028 |
hawk | +0.028 |
Neg logit contributions | |
forehead | -0.027 |
hand | -0.026 |
lap | -0.025 |
cheek | -0.025 |
plate | -0.024 |
Token | they |
Token ID | 484 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
pand | +0.065 |
pill | +0.057 |
hawk | +0.057 |
otion | +0.057 |
add | +0.056 |
Neg logit contributions | |
forehead | -0.066 |
hand | -0.063 |
lap | -0.062 |
cheek | -0.061 |
bek | -0.060 |
Token | don |
Token ID | 836 |
Feature activation | -8.21e-01 |
Pos logit contributions | |
pand | +0.005 |
pill | +0.005 |
hawk | +0.004 |
add | +0.004 |
otion | +0.004 |
Neg logit contributions | |
forehead | -0.003 |
hand | -0.003 |
cheek | -0.003 |
tight | -0.003 |
lap | -0.003 |
Token | 't |
Token ID | 470 |
Feature activation | -5.92e-01 |
Pos logit contributions | |
pand | +0.032 |
pill | +0.029 |
hawk | +0.028 |
storms | +0.028 |
add | +0.027 |
Neg logit contributions | |
hand | -0.027 |
forehead | -0.027 |
tight | -0.026 |
whole | -0.026 |
cheek | -0.025 |
Token | get |
Token ID | 651 |
Feature activation | +3.86e-01 |
Pos logit contributions | |
pand | +0.026 |
pill | +0.026 |
add | +0.025 |
otion | +0.025 |
hawk | +0.025 |
Neg logit contributions | |
hand | -0.016 |
forehead | -0.016 |
cheek | -0.015 |
tight | -0.015 |
lap | -0.014 |
Token | the |
Token ID | 262 |
Feature activation | +4.35e-01 |
Pos logit contributions | |
forehead | +0.012 |
hand | +0.012 |
lap | +0.011 |
cheek | +0.011 |
tight | +0.011 |
Neg logit contributions | |
pand | -0.015 |
pill | -0.014 |
hawk | -0.014 |
add | -0.014 |
otion | -0.014 |
Token | letters |
Token ID | 7475 |
Feature activation | +3.84e+00 |
Pos logit contributions | |
bek | +0.015 |
ae | +0.015 |
forehead | +0.015 |
lap | +0.014 |
cheek | +0.013 |
Neg logit contributions | |
pand | -0.015 |
hawk | -0.015 |
aliens | -0.014 |
flashes | -0.014 |
helicopter | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -4.10e-01 |
Pos logit contributions | |
forehead | +0.106 |
bek | +0.101 |
lap | +0.098 |
hand | +0.094 |
cheek | +0.093 |
Neg logit contributions | |
pand | -0.165 |
otion | -0.151 |
hawk | -0.149 |
storms | -0.149 |
die | -0.147 |
Token | You |
Token ID | 921 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
pand | +0.016 |
hawk | +0.015 |
pill | +0.015 |
add | +0.015 |
otion | +0.014 |
Neg logit contributions | |
forehead | -0.013 |
hand | -0.013 |
lap | -0.013 |
cheek | -0.012 |
tight | -0.012 |
Token | should |
Token ID | 815 |
Feature activation | +2.02e-01 |
Pos logit contributions | |
hand | +0.030 |
forehead | +0.030 |
tight | +0.029 |
whole | +0.028 |
cheek | +0.028 |
Neg logit contributions | |
pand | -0.042 |
pill | -0.040 |
add | -0.038 |
otion | -0.038 |
hawk | -0.038 |
Token | be |
Token ID | 307 |
Feature activation | -6.58e-01 |
Pos logit contributions | |
forehead | +0.005 |
hand | +0.005 |
cheek | +0.005 |
lap | +0.005 |
tight | +0.005 |
Neg logit contributions | |
pand | -0.009 |
pill | -0.009 |
hawk | -0.008 |
add | -0.008 |
otion | -0.008 |
Token | careful |
Token ID | 8161 |
Feature activation | +2.16e+00 |
Pos logit contributions | |
pand | +0.030 |
hawk | +0.027 |
add | +0.027 |
pill | +0.027 |
otion | +0.026 |
Neg logit contributions | |
forehead | -0.017 |
lap | -0.016 |
cheek | -0.016 |
bek | -0.016 |
hand | -0.015 |
Token | the |
Token ID | 262 |
Feature activation | +8.00e-01 |
Pos logit contributions | |
pand | +0.010 |
pill | +0.009 |
add | +0.009 |
hawk | +0.009 |
otion | +0.009 |
Neg logit contributions | |
forehead | -0.006 |
hand | -0.006 |
cheek | -0.005 |
lap | -0.005 |
tight | -0.005 |
Token | machine |
Token ID | 4572 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
bek | +0.025 |
forehead | +0.024 |
lap | +0.024 |
hand | +0.023 |
ossible | +0.023 |
Neg logit contributions | |
pand | -0.032 |
hawk | -0.030 |
Hipp | -0.027 |
otion | -0.027 |
storms | -0.027 |
Token | printing |
Token ID | 13570 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
forehead | +0.051 |
lap | +0.049 |
bek | +0.049 |
hand | +0.049 |
cheek | +0.046 |
Neg logit contributions | |
pand | -0.069 |
otion | -0.064 |
hawk | -0.063 |
storms | -0.063 |
pill | -0.061 |
Token | out |
Token ID | 503 |
Feature activation | +6.65e-01 |
Pos logit contributions | |
forehead | +0.057 |
lap | +0.053 |
hand | +0.053 |
cheek | +0.052 |
bek | +0.052 |
Neg logit contributions | |
pand | -0.082 |
hawk | -0.075 |
pill | -0.075 |
otion | -0.074 |
storms | -0.073 |
Token | a |
Token ID | 257 |
Feature activation | +3.97e-01 |
Pos logit contributions | |
forehead | +0.022 |
lap | +0.020 |
cheek | +0.020 |
hand | +0.020 |
bek | +0.019 |
Neg logit contributions | |
pand | -0.026 |
hawk | -0.024 |
pill | -0.023 |
otion | -0.023 |
aliens | -0.023 |
Token | receipt |
Token ID | 14507 |
Feature activation | +3.83e+00 |
Pos logit contributions | |
forehead | +0.013 |
lap | +0.012 |
ossible | +0.012 |
hand | +0.012 |
cheek | +0.012 |
Neg logit contributions | |
pand | -0.015 |
hawk | -0.014 |
pill | -0.013 |
Hipp | -0.013 |
storms | -0.013 |
Token | for |
Token ID | 329 |
Feature activation | +1.39e+00 |
Pos logit contributions | |
forehead | +0.110 |
bek | +0.110 |
lap | +0.110 |
tight | +0.102 |
hand | +0.100 |
Neg logit contributions | |
pand | -0.152 |
otion | -0.146 |
hawk | -0.143 |
storms | -0.135 |
pill | -0.132 |
Token | the |
Token ID | 262 |
Feature activation | +4.06e-01 |
Pos logit contributions | |
forehead | +0.042 |
lap | +0.040 |
bek | +0.040 |
cheek | +0.039 |
hand | +0.039 |
Neg logit contributions | |
pand | -0.058 |
add | -0.051 |
otion | -0.051 |
hawk | -0.051 |
pill | -0.051 |
Token | don |
Token ID | 836 |
Feature activation | +5.71e-01 |
Pos logit contributions | |
bek | +0.011 |
ossible | +0.011 |
lap | +0.011 |
ae | +0.010 |
forehead | +0.010 |
Neg logit contributions | |
pand | -0.017 |
hawk | -0.016 |
Hipp | -0.015 |
aliens | -0.015 |
helicopter | -0.015 |
Token | ut |
Token ID | 315 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
forehead | +0.021 |
bek | +0.020 |
hand | +0.020 |
lap | +0.019 |
cheek | +0.019 |
Neg logit contributions | |
pand | -0.020 |
otion | -0.019 |
pill | -0.018 |
hawk | -0.018 |
add | -0.018 |
Token | ," |
Token ID | 553 |
Feature activation | +6.16e-01 |
Pos logit contributions | |
forehead | +0.044 |
lap | +0.042 |
hand | +0.042 |
bek | +0.041 |
cheek | +0.040 |
Neg logit contributions | |
pand | -0.061 |
otion | -0.059 |
hawk | -0.057 |
add | -0.056 |
pill | -0.056 |
Token | money |
Token ID | 1637 |
Feature activation | +2.18e+00 |
Pos logit contributions | |
forehead | +0.029 |
bek | +0.028 |
cheek | +0.028 |
lap | +0.026 |
ae | +0.025 |
Neg logit contributions | |
pand | -0.037 |
add | -0.034 |
pill | -0.032 |
die | -0.031 |
otion | -0.031 |
Token | to |
Token ID | 284 |
Feature activation | +5.89e-01 |
Pos logit contributions | |
forehead | +0.059 |
lap | +0.055 |
cheek | +0.053 |
bek | +0.053 |
hand | +0.052 |
Neg logit contributions | |
pand | -0.098 |
otion | -0.088 |
pill | -0.088 |
storms | -0.087 |
add | -0.086 |
Token | buy |
Token ID | 2822 |
Feature activation | +2.02e+00 |
Pos logit contributions | |
forehead | +0.015 |
cheek | +0.013 |
lap | +0.013 |
bek | +0.013 |
hand | +0.013 |
Neg logit contributions | |
pand | -0.027 |
add | -0.025 |
die | -0.025 |
otion | -0.024 |
hawk | -0.024 |
Token | an |
Token ID | 281 |
Feature activation | +8.16e-02 |
Pos logit contributions | |
forehead | +0.065 |
cheek | +0.060 |
lap | +0.060 |
bek | +0.060 |
hand | +0.059 |
Neg logit contributions | |
pand | -0.083 |
pill | -0.074 |
add | -0.073 |
otion | -0.071 |
hawk | -0.070 |
Token | ice |
Token ID | 4771 |
Feature activation | +4.31e-01 |
Pos logit contributions | |
forehead | +0.003 |
bek | +0.003 |
lap | +0.003 |
cheek | +0.003 |
hand | +0.003 |
Neg logit contributions | |
pand | -0.003 |
add | -0.003 |
hawk | -0.002 |
otion | -0.002 |
Hipp | -0.002 |
Token | cream |
Token ID | 8566 |
Feature activation | +3.83e+00 |
Pos logit contributions | |
forehead | +0.012 |
bek | +0.011 |
lap | +0.011 |
ossible | +0.010 |
hand | +0.010 |
Neg logit contributions | |
pand | -0.019 |
add | -0.016 |
Hipp | -0.016 |
pill | -0.016 |
hawk | -0.016 |
Token | . |
Token ID | 13 |
Feature activation | +6.58e-01 |
Pos logit contributions | |
forehead | +0.126 |
ossible | +0.113 |
bek | +0.110 |
lap | +0.109 |
cheek | +0.108 |
Neg logit contributions | |
pand | -0.147 |
hawk | -0.129 |
storms | -0.126 |
spr | -0.124 |
anymore | -0.123 |
Token | She |
Token ID | 1375 |
Feature activation | +4.32e-01 |
Pos logit contributions | |
bek | +0.028 |
forehead | +0.026 |
cheek | +0.026 |
cle | +0.025 |
entire | +0.024 |
Neg logit contributions | |
pill | -0.018 |
pand | -0.018 |
otion | -0.017 |
storms | -0.017 |
flashes | -0.015 |
Token | was |
Token ID | 373 |
Feature activation | +2.24e-01 |
Pos logit contributions | |
forehead | +0.015 |
cheek | +0.014 |
lap | +0.014 |
attention | +0.014 |
hand | +0.013 |
Neg logit contributions | |
pand | -0.016 |
pill | -0.014 |
add | -0.014 |
spr | -0.013 |
hawk | -0.013 |
Token | so |
Token ID | 523 |
Feature activation | +5.19e-01 |
Pos logit contributions | |
bek | +0.008 |
forehead | +0.007 |
cheek | +0.007 |
lap | +0.007 |
ossible | +0.007 |
Neg logit contributions | |
pand | -0.008 |
otion | -0.007 |
add | -0.007 |
flashes | -0.006 |
Tweet | -0.006 |
Token | happy |
Token ID | 3772 |
Feature activation | +7.51e-01 |
Pos logit contributions | |
forehead | +0.016 |
lap | +0.016 |
cheek | +0.015 |
bek | +0.015 |
doorstep | +0.015 |
Neg logit contributions | |
pand | -0.019 |
add | -0.019 |
otion | -0.018 |
die | -0.017 |
pill | -0.016 |
Token |
|
Token ID | 198 |
Feature activation | -2.50e-01 |
Pos logit contributions | |
pand | +0.003 |
pill | +0.002 |
add | +0.002 |
otion | +0.002 |
hawk | +0.002 |
Neg logit contributions | |
forehead | -0.001 |
hand | -0.001 |
cheek | -0.001 |
lap | -0.001 |
bek | -0.001 |
Token | They |
Token ID | 2990 |
Feature activation | +3.99e-01 |
Pos logit contributions | |
pand | +0.009 |
hon | +0.009 |
pill | +0.008 |
Hon | +0.008 |
hawk | +0.008 |
Neg logit contributions | |
forehead | -0.009 |
cheek | -0.008 |
hand | -0.008 |
lap | -0.008 |
hands | -0.008 |
Token | paid |
Token ID | 3432 |
Feature activation | +3.84e-01 |
Pos logit contributions | |
forehead | +0.012 |
cheek | +0.012 |
lap | +0.011 |
hand | +0.011 |
bek | +0.011 |
Neg logit contributions | |
pand | -0.017 |
add | -0.015 |
pill | -0.015 |
hawk | -0.014 |
spr | -0.014 |
Token | the |
Token ID | 262 |
Feature activation | +6.43e-01 |
Pos logit contributions | |
bek | +0.009 |
forehead | +0.009 |
lap | +0.008 |
pron | +0.008 |
entire | +0.008 |
Neg logit contributions | |
pand | -0.019 |
add | -0.017 |
aliens | -0.017 |
hon | -0.016 |
die | -0.016 |
Token | ice |
Token ID | 4771 |
Feature activation | +8.17e-01 |
Pos logit contributions | |
forehead | +0.021 |
bek | +0.021 |
lap | +0.020 |
cheek | +0.019 |
hand | +0.019 |
Neg logit contributions | |
pand | -0.026 |
hawk | -0.022 |
add | -0.022 |
pill | -0.021 |
Hipp | -0.021 |
Token | cream |
Token ID | 8566 |
Feature activation | +3.81e+00 |
Pos logit contributions | |
forehead | +0.021 |
bek | +0.018 |
lap | +0.017 |
hand | +0.017 |
satisfied | +0.016 |
Neg logit contributions | |
pand | -0.040 |
add | -0.035 |
Hipp | -0.034 |
hon | -0.033 |
pill | -0.033 |
Token | man |
Token ID | 582 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
forehead | +0.118 |
bek | +0.104 |
cheek | +0.102 |
lap | +0.101 |
entire | +0.099 |
Neg logit contributions | |
pand | -0.162 |
Hipp | -0.138 |
hawk | -0.136 |
add | -0.136 |
spr | -0.135 |
Token | and |
Token ID | 290 |
Feature activation | -4.96e-01 |
Pos logit contributions | |
forehead | +0.039 |
cheek | +0.035 |
bek | +0.035 |
lap | +0.034 |
hand | +0.034 |
Neg logit contributions | |
pand | -0.064 |
pill | -0.058 |
otion | -0.056 |
storms | -0.055 |
add | -0.055 |
Token | l |
Token ID | 300 |
Feature activation | -3.67e-01 |
Pos logit contributions | |
pand | +0.020 |
spr | +0.017 |
flashes | +0.017 |
hawk | +0.017 |
Tweet | +0.017 |
Neg logit contributions | |
forehead | -0.016 |
cheek | -0.015 |
entire | -0.015 |
bek | -0.014 |
whole | -0.014 |
Token | icked |
Token ID | 9484 |
Feature activation | -5.02e-01 |
Pos logit contributions | |
pand | +0.012 |
pill | +0.011 |
hawk | +0.010 |
add | +0.010 |
otion | +0.010 |
Neg logit contributions | |
forehead | -0.015 |
hand | -0.014 |
lap | -0.014 |
cheek | -0.014 |
tight | -0.014 |
Token | their |
Token ID | 511 |
Feature activation | -4.34e-01 |
Pos logit contributions | |
pand | +0.018 |
pill | +0.017 |
add | +0.016 |
hawk | +0.016 |
otion | +0.016 |
Neg logit contributions | |
forehead | -0.018 |
hand | -0.017 |
cheek | -0.017 |
lap | -0.017 |
tight | -0.016 |
Token | Where |
Token ID | 8496 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
pand | +0.046 |
pill | +0.043 |
hawk | +0.041 |
otion | +0.041 |
add | +0.041 |
Neg logit contributions | |
forehead | -0.041 |
hand | -0.038 |
lap | -0.038 |
cheek | -0.037 |
tight | -0.036 |
Token | did |
Token ID | 750 |
Feature activation | +4.79e-01 |
Pos logit contributions | |
pand | +0.043 |
otion | +0.040 |
pill | +0.040 |
add | +0.039 |
hawk | +0.039 |
Neg logit contributions | |
forehead | -0.040 |
hand | -0.037 |
cheek | -0.037 |
lap | -0.037 |
tight | -0.035 |
Token | you |
Token ID | 345 |
Feature activation | +9.88e-01 |
Pos logit contributions | |
forehead | +0.013 |
hand | +0.012 |
cheek | +0.012 |
lap | +0.012 |
tight | +0.012 |
Neg logit contributions | |
pand | -0.021 |
pill | -0.020 |
add | -0.020 |
hawk | -0.019 |
otion | -0.019 |
Token | get |
Token ID | 651 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
forehead | +0.028 |
hand | +0.027 |
lap | +0.026 |
cheek | +0.026 |
tight | +0.025 |
Neg logit contributions | |
pand | -0.043 |
pill | -0.040 |
otion | -0.039 |
add | -0.039 |
hawk | -0.039 |
Token | this |
Token ID | 428 |
Feature activation | +3.61e-01 |
Pos logit contributions | |
forehead | +0.031 |
lap | +0.029 |
cheek | +0.028 |
bek | +0.028 |
hand | +0.028 |
Neg logit contributions | |
pand | -0.044 |
add | -0.040 |
pill | -0.039 |
hawk | -0.039 |
otion | -0.039 |
Token | money |
Token ID | 1637 |
Feature activation | +3.78e+00 |
Pos logit contributions | |
forehead | +0.012 |
bek | +0.012 |
lap | +0.011 |
hand | +0.011 |
cheek | +0.011 |
Neg logit contributions | |
pand | -0.014 |
die | -0.012 |
hawk | -0.012 |
otion | -0.012 |
Tweet | -0.011 |
Token | ? |
Token ID | 30 |
Feature activation | -6.71e-01 |
Pos logit contributions | |
forehead | +0.153 |
lap | +0.143 |
bek | +0.143 |
cheek | +0.136 |
hand | +0.132 |
Neg logit contributions | |
pand | -0.115 |
otion | -0.108 |
die | -0.103 |
hawk | -0.099 |
storms | -0.096 |
Token | This |
Token ID | 770 |
Feature activation | -7.07e-01 |
Pos logit contributions | |
pand | +0.025 |
pill | +0.023 |
otion | +0.023 |
add | +0.022 |
hawk | +0.022 |
Neg logit contributions | |
forehead | -0.023 |
hand | -0.022 |
cheek | -0.021 |
lap | -0.021 |
whole | -0.021 |
Token | is |
Token ID | 318 |
Feature activation | +9.68e-02 |
Pos logit contributions | |
pand | +0.024 |
pill | +0.023 |
hawk | +0.022 |
otion | +0.022 |
storms | +0.020 |
Neg logit contributions | |
forehead | -0.027 |
cheek | -0.025 |
lap | -0.025 |
hand | -0.025 |
bek | -0.024 |
Token | not |
Token ID | 407 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
forehead | +0.003 |
lap | +0.003 |
cheek | +0.003 |
hand | +0.003 |
bek | +0.003 |
Neg logit contributions | |
pand | -0.004 |
hawk | -0.003 |
pill | -0.003 |
die | -0.003 |
otion | -0.003 |
Token | your |
Token ID | 534 |
Feature activation | +7.63e-01 |
Pos logit contributions | |
forehead | +0.057 |
lap | +0.056 |
cheek | +0.052 |
hand | +0.051 |
bek | +0.050 |
Neg logit contributions | |
pand | -0.072 |
hawk | -0.063 |
die | -0.061 |
aliens | -0.061 |
otion | -0.060 |
Token | was |
Token ID | 373 |
Feature activation | -4.91e-01 |
Pos logit contributions | |
forehead | +0.010 |
hand | +0.009 |
cheek | +0.009 |
lap | +0.009 |
tight | +0.009 |
Neg logit contributions | |
pand | -0.011 |
pill | -0.010 |
add | -0.010 |
hawk | -0.010 |
otion | -0.010 |
Token | able |
Token ID | 1498 |
Feature activation | +7.64e-01 |
Pos logit contributions | |
pand | +0.020 |
otion | +0.018 |
hawk | +0.018 |
pill | +0.017 |
add | +0.017 |
Neg logit contributions | |
forehead | -0.015 |
lap | -0.014 |
cheek | -0.014 |
bek | -0.014 |
hand | -0.014 |
Token | to |
Token ID | 284 |
Feature activation | +4.29e-01 |
Pos logit contributions | |
forehead | +0.024 |
bek | +0.023 |
lap | +0.022 |
cheek | +0.021 |
plate | +0.021 |
Neg logit contributions | |
pand | -0.031 |
hawk | -0.027 |
pill | -0.027 |
Tweet | -0.026 |
die | -0.026 |
Token | gain |
Token ID | 4461 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
forehead | +0.012 |
hand | +0.012 |
lap | +0.011 |
cheek | +0.011 |
tight | +0.011 |
Neg logit contributions | |
pand | -0.018 |
pill | -0.017 |
hawk | -0.017 |
add | -0.017 |
otion | -0.017 |
Token | this |
Token ID | 428 |
Feature activation | -7.57e-01 |
Pos logit contributions | |
pand | +0.011 |
pill | +0.010 |
hawk | +0.010 |
add | +0.010 |
otion | +0.010 |
Neg logit contributions | |
forehead | -0.008 |
hand | -0.008 |
cheek | -0.008 |
lap | -0.008 |
bek | -0.008 |
Token | badge |
Token ID | 23009 |
Feature activation | +3.78e+00 |
Pos logit contributions | |
pand | +0.031 |
hawk | +0.029 |
storms | +0.028 |
scares | +0.027 |
aliens | +0.027 |
Neg logit contributions | |
forehead | -0.021 |
bek | -0.021 |
ossible | -0.020 |
lap | -0.019 |
hand | -0.019 |
Token | and |
Token ID | 290 |
Feature activation | -4.12e-01 |
Pos logit contributions | |
forehead | +0.108 |
bek | +0.105 |
lap | +0.099 |
hand | +0.099 |
cheek | +0.098 |
Neg logit contributions | |
pand | -0.155 |
storms | -0.144 |
otion | -0.140 |
pill | -0.138 |
hawk | -0.138 |
Token | become |
Token ID | 1716 |
Feature activation | -1.97e-01 |
Pos logit contributions | |
pand | +0.017 |
pill | +0.015 |
hawk | +0.015 |
aliens | +0.015 |
scares | +0.014 |
Neg logit contributions | |
forehead | -0.013 |
cheek | -0.012 |
bek | -0.012 |
lap | -0.012 |
hand | -0.011 |
Token | a |
Token ID | 257 |
Feature activation | +8.84e-01 |
Pos logit contributions | |
pand | +0.008 |
hawk | +0.007 |
pill | +0.007 |
aliens | +0.007 |
conflicts | +0.007 |
Neg logit contributions | |
forehead | -0.006 |
lap | -0.006 |
cheek | -0.006 |
hand | -0.006 |
bek | -0.005 |
Token | cop |
Token ID | 2243 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
forehead | +0.029 |
lap | +0.027 |
ossible | +0.027 |
cheek | +0.026 |
hand | +0.026 |
Neg logit contributions | |
pand | -0.033 |
hawk | -0.032 |
helicopter | -0.028 |
aliens | -0.028 |
Hipp | -0.028 |
Token | ." |
Token ID | 526 |
Feature activation | +8.56e-01 |
Pos logit contributions | |
forehead | +0.049 |
bek | +0.046 |
hand | +0.045 |
lap | +0.045 |
cheek | +0.044 |
Neg logit contributions | |
pand | -0.083 |
storms | -0.075 |
hawk | -0.075 |
pill | -0.074 |
aliens | -0.072 |
Token |
|
Token ID | 198 |
Feature activation | -3.83e-02 |
Pos logit contributions | |
pand | +0.000 |
pill | +0.000 |
hawk | +0.000 |
add | +0.000 |
otion | +0.000 |
Neg logit contributions | |
forehead | -0.000 |
hand | -0.000 |
cheek | -0.000 |
lap | -0.000 |
bek | -0.000 |
Token | Tom |
Token ID | 13787 |
Feature activation | +7.90e-01 |
Pos logit contributions | |
pand | +0.002 |
pill | +0.002 |
hawk | +0.002 |
add | +0.002 |
otion | +0.001 |
Neg logit contributions | |
forehead | -0.001 |
hand | -0.001 |
cheek | -0.001 |
lap | -0.001 |
tight | -0.001 |
Token | looked |
Token ID | 3114 |
Feature activation | -4.32e-02 |
Pos logit contributions | |
forehead | +0.023 |
bek | +0.022 |
hand | +0.022 |
cheek | +0.021 |
lap | +0.021 |
Neg logit contributions | |
pand | -0.032 |
pill | -0.031 |
storms | -0.029 |
add | -0.029 |
otion | -0.029 |
Token | at |
Token ID | 379 |
Feature activation | -5.10e-02 |
Pos logit contributions | |
pand | +0.002 |
pill | +0.002 |
add | +0.002 |
hawk | +0.002 |
otion | +0.002 |
Neg logit contributions | |
forehead | -0.001 |
bek | -0.001 |
hand | -0.001 |
lap | -0.001 |
cheek | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
pand | +0.002 |
hawk | +0.002 |
pill | +0.002 |
add | +0.002 |
otion | +0.002 |
Neg logit contributions | |
forehead | -0.001 |
cheek | -0.001 |
hand | -0.001 |
lap | -0.001 |
bek | -0.001 |
Token | stickers |
Token ID | 28568 |
Feature activation | +3.77e+00 |
Pos logit contributions | |
bek | +0.035 |
forehead | +0.033 |
lap | +0.032 |
ossible | +0.031 |
ae | +0.031 |
Neg logit contributions | |
pand | -0.043 |
hawk | -0.039 |
pill | -0.037 |
flashes | -0.037 |
otion | -0.036 |
Token | . |
Token ID | 13 |
Feature activation | -1.94e-01 |
Pos logit contributions | |
forehead | +0.102 |
bek | +0.096 |
lap | +0.092 |
hand | +0.091 |
cheek | +0.090 |
Neg logit contributions | |
pand | -0.161 |
otion | -0.156 |
hawk | -0.150 |
pill | -0.147 |
storms | -0.147 |
Token | He |
Token ID | 679 |
Feature activation | +5.45e-01 |
Pos logit contributions | |
pand | +0.009 |
hawk | +0.009 |
pill | +0.008 |
add | +0.008 |
otion | +0.008 |
Neg logit contributions | |
hand | -0.005 |
lap | -0.005 |
forehead | -0.005 |
tight | -0.004 |
cheek | -0.004 |
Token | saw |
Token ID | 2497 |
Feature activation | -3.54e-01 |
Pos logit contributions | |
forehead | +0.014 |
hand | +0.013 |
lap | +0.013 |
cheek | +0.013 |
bek | +0.012 |
Neg logit contributions | |
pand | -0.025 |
pill | -0.023 |
hawk | -0.023 |
add | -0.023 |
otion | -0.023 |
Token | one |
Token ID | 530 |
Feature activation | +9.73e-01 |
Pos logit contributions | |
pand | +0.017 |
add | +0.017 |
pill | +0.017 |
storms | +0.017 |
otion | +0.017 |
Neg logit contributions | |
hand | -0.007 |
tight | -0.007 |
forehead | -0.007 |
plate | -0.006 |
front | -0.006 |
Token | with |
Token ID | 351 |
Feature activation | +4.17e-01 |
Pos logit contributions | |
forehead | +0.031 |
bek | +0.029 |
lap | +0.028 |
cheek | +0.028 |
hand | +0.028 |
Neg logit contributions | |
pand | -0.039 |
hawk | -0.035 |
otion | -0.035 |
pill | -0.034 |
add | -0.034 |
Token | have |
Token ID | 423 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
forehead | +0.050 |
hand | +0.049 |
cheek | +0.046 |
lap | +0.046 |
tight | +0.046 |
Neg logit contributions | |
pand | -0.071 |
pill | -0.066 |
add | -0.065 |
hawk | -0.065 |
otion | -0.064 |
Token | to |
Token ID | 284 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
forehead | +0.029 |
cheek | +0.026 |
bek | +0.026 |
lap | +0.026 |
hand | +0.025 |
Neg logit contributions | |
pand | -0.048 |
pill | -0.044 |
hawk | -0.043 |
add | -0.043 |
storms | -0.042 |
Token | look |
Token ID | 804 |
Feature activation | -7.46e-01 |
Pos logit contributions | |
forehead | +0.004 |
hand | +0.003 |
lap | +0.003 |
cheek | +0.003 |
bek | +0.003 |
Neg logit contributions | |
pand | -0.006 |
hawk | -0.005 |
add | -0.005 |
otion | -0.005 |
pill | -0.005 |
Token | at |
Token ID | 379 |
Feature activation | +1.49e-01 |
Pos logit contributions | |
pand | +0.030 |
aliens | +0.026 |
add | +0.026 |
otion | +0.025 |
die | +0.025 |
Neg logit contributions | |
forehead | -0.023 |
bek | -0.023 |
lap | -0.022 |
cheek | -0.021 |
hand | -0.021 |
Token | the |
Token ID | 262 |
Feature activation | +9.02e-01 |
Pos logit contributions | |
forehead | +0.004 |
hand | +0.004 |
lap | +0.004 |
cheek | +0.004 |
bek | +0.004 |
Neg logit contributions | |
pand | -0.006 |
pill | -0.006 |
hawk | -0.006 |
add | -0.006 |
otion | -0.006 |
Token | letters |
Token ID | 7475 |
Feature activation | +3.75e+00 |
Pos logit contributions | |
forehead | +0.029 |
bek | +0.028 |
lap | +0.027 |
tight | +0.026 |
satisfied | +0.026 |
Neg logit contributions | |
pand | -0.036 |
hawk | -0.032 |
aliens | -0.032 |
pill | -0.031 |
flashes | -0.031 |
Token | . |
Token ID | 13 |
Feature activation | -3.00e-01 |
Pos logit contributions | |
forehead | +0.102 |
bek | +0.096 |
lap | +0.092 |
hand | +0.090 |
cheek | +0.089 |
Neg logit contributions | |
pand | -0.162 |
otion | -0.151 |
storms | -0.147 |
hawk | -0.146 |
pill | -0.144 |
Token | They |
Token ID | 1119 |
Feature activation | +1.52e-01 |
Pos logit contributions | |
pand | +0.012 |
hawk | +0.011 |
add | +0.011 |
pill | +0.011 |
otion | +0.011 |
Neg logit contributions | |
forehead | -0.009 |
hand | -0.009 |
lap | -0.009 |
cheek | -0.008 |
tight | -0.008 |
Token | make |
Token ID | 787 |
Feature activation | -1.42e-01 |
Pos logit contributions | |
hand | +0.005 |
forehead | +0.004 |
tight | +0.004 |
cheek | +0.004 |
whole | +0.004 |
Neg logit contributions | |
pand | -0.006 |
pill | -0.006 |
add | -0.006 |
hawk | -0.006 |
otion | -0.006 |
Token | sounds |
Token ID | 5238 |
Feature activation | -1.72e+00 |
Pos logit contributions | |
pand | +0.006 |
pill | +0.006 |
hawk | +0.006 |
storms | +0.006 |
otion | +0.006 |
Neg logit contributions | |
hand | -0.004 |
forehead | -0.004 |
tight | -0.004 |
cheek | -0.003 |
whole | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | -4.05e-01 |
Pos logit contributions | |
pand | +0.087 |
pill | +0.084 |
add | +0.083 |
hawk | +0.081 |
otion | +0.081 |
Neg logit contributions | |
hand | -0.034 |
forehead | -0.034 |
tight | -0.033 |
cheek | -0.032 |
lap | -0.031 |
Token | words |
Token ID | 2456 |
Feature activation | +1.62e+00 |
Pos logit contributions | |
forehead | +0.027 |
bek | +0.026 |
pron | +0.025 |
ae | +0.025 |
nie | +0.025 |
Neg logit contributions | |
pand | -0.029 |
hawk | -0.027 |
flashes | -0.026 |
lions | -0.024 |
aliens | -0.024 |
Token | , |
Token ID | 11 |
Feature activation | +6.22e-01 |
Pos logit contributions | |
forehead | +0.039 |
hand | +0.037 |
lap | +0.035 |
cheek | +0.035 |
tight | +0.034 |
Neg logit contributions | |
pand | -0.077 |
pill | -0.072 |
add | -0.071 |
hawk | -0.071 |
otion | -0.070 |
Token | " |
Token ID | 366 |
Feature activation | -6.47e-01 |
Pos logit contributions | |
forehead | +0.015 |
hand | +0.015 |
lap | +0.014 |
tight | +0.014 |
cheek | +0.014 |
Neg logit contributions | |
pand | -0.029 |
pill | -0.028 |
add | -0.028 |
hawk | -0.027 |
otion | -0.027 |
Token | Welcome |
Token ID | 14618 |
Feature activation | +9.64e-02 |
Pos logit contributions | |
pand | +0.023 |
pill | +0.022 |
add | +0.021 |
hawk | +0.021 |
otion | +0.020 |
Neg logit contributions | |
forehead | -0.024 |
hand | -0.022 |
lap | -0.022 |
cheek | -0.022 |
tight | -0.021 |
Token | " |
Token ID | 1 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
forehead | +0.002 |
lap | +0.002 |
cheek | +0.002 |
hand | +0.002 |
bek | +0.002 |
Neg logit contributions | |
pand | -0.005 |
hawk | -0.004 |
storms | -0.004 |
pill | -0.004 |
add | -0.004 |
Token | written |
Token ID | 3194 |
Feature activation | +3.75e+00 |
Pos logit contributions | |
forehead | +0.058 |
bek | +0.053 |
lap | +0.053 |
list | +0.052 |
hand | +0.050 |
Neg logit contributions | |
pand | -0.074 |
storms | -0.073 |
hon | -0.069 |
Tweet | -0.069 |
add | -0.068 |
Token | on |
Token ID | 319 |
Feature activation | +9.04e-01 |
Pos logit contributions | |
forehead | +0.112 |
lap | +0.102 |
cheek | +0.100 |
hand | +0.098 |
bek | +0.098 |
Neg logit contributions | |
pand | -0.160 |
hawk | -0.147 |
otion | -0.146 |
pill | -0.144 |
add | -0.141 |
Token | it |
Token ID | 340 |
Feature activation | -8.20e-01 |
Pos logit contributions | |
forehead | +0.027 |
cheek | +0.025 |
lap | +0.024 |
hand | +0.024 |
bek | +0.024 |
Neg logit contributions | |
pand | -0.038 |
pill | -0.035 |
hawk | -0.035 |
otion | -0.034 |
storms | -0.034 |
Token | . |
Token ID | 13 |
Feature activation | +3.11e-01 |
Pos logit contributions | |
pand | +0.037 |
pill | +0.034 |
otion | +0.034 |
storms | +0.034 |
hawk | +0.034 |
Neg logit contributions | |
forehead | -0.021 |
hand | -0.019 |
cheek | -0.019 |
lap | -0.019 |
bek | -0.019 |
Token | It |
Token ID | 632 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
forehead | +0.009 |
hand | +0.008 |
cheek | +0.008 |
lap | +0.008 |
bek | +0.008 |
Neg logit contributions | |
pand | -0.014 |
pill | -0.013 |
otion | -0.012 |
add | -0.012 |
storms | -0.012 |
Token | was |
Token ID | 373 |
Feature activation | +2.18e-01 |
Pos logit contributions | |
pand | +0.070 |
add | +0.068 |
pill | +0.066 |
otion | +0.066 |
hawk | +0.066 |
Neg logit contributions | |
hand | -0.032 |
forehead | -0.032 |
tight | -0.031 |
whole | -0.029 |
cheek | -0.029 |
Token | We |
Token ID | 1135 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
pand | +0.041 |
pill | +0.037 |
otion | +0.036 |
add | +0.036 |
hawk | +0.036 |
Neg logit contributions | |
forehead | -0.024 |
hand | -0.024 |
whole | -0.023 |
cheek | -0.023 |
lap | -0.022 |
Token | can |
Token ID | 460 |
Feature activation | -7.09e-01 |
Pos logit contributions | |
hand | +0.032 |
forehead | +0.032 |
tight | +0.030 |
whole | +0.030 |
cheek | +0.030 |
Neg logit contributions | |
pand | -0.041 |
pill | -0.039 |
add | -0.037 |
hawk | -0.037 |
storms | -0.036 |
Token | 't |
Token ID | 470 |
Feature activation | -1.57e-01 |
Pos logit contributions | |
pand | +0.030 |
pill | +0.030 |
add | +0.028 |
storms | +0.028 |
hawk | +0.028 |
Neg logit contributions | |
hand | -0.021 |
forehead | -0.020 |
tight | -0.020 |
lap | -0.019 |
cheek | -0.019 |
Token | get |
Token ID | 651 |
Feature activation | +7.91e-01 |
Pos logit contributions | |
pill | +0.008 |
pand | +0.008 |
otion | +0.007 |
add | +0.007 |
storms | +0.007 |
Neg logit contributions | |
hand | -0.004 |
forehead | -0.004 |
cheek | -0.003 |
tight | -0.003 |
whole | -0.003 |
Token | ice |
Token ID | 4771 |
Feature activation | +4.94e-01 |
Pos logit contributions | |
forehead | +0.025 |
hand | +0.025 |
tight | +0.024 |
cheek | +0.024 |
whole | +0.023 |
Neg logit contributions | |
pand | -0.031 |
pill | -0.030 |
otion | -0.029 |
storms | -0.028 |
add | -0.028 |
Token | cream |
Token ID | 8566 |
Feature activation | +3.75e+00 |
Pos logit contributions | |
forehead | +0.011 |
hand | +0.010 |
lap | +0.010 |
bek | +0.010 |
cheek | +0.009 |
Neg logit contributions | |
pand | -0.025 |
add | -0.023 |
Hipp | -0.022 |
otion | -0.022 |
hon | -0.022 |
Token | ." |
Token ID | 526 |
Feature activation | +7.47e-01 |
Pos logit contributions | |
forehead | +0.122 |
bek | +0.117 |
lap | +0.117 |
hand | +0.115 |
cheek | +0.109 |
Neg logit contributions | |
pand | -0.145 |
hawk | -0.130 |
otion | -0.129 |
storms | -0.129 |
die | -0.125 |
Token | |
Token ID | 220 |
Feature activation | -6.03e-02 |
Pos logit contributions | |
bek | +0.019 |
forehead | +0.019 |
entire | +0.018 |
whole | +0.017 |
cheek | +0.017 |
Neg logit contributions | |
pand | -0.033 |
Tweet | -0.031 |
Hipp | -0.031 |
add | -0.031 |
storms | -0.030 |
Token |
|
Token ID | 198 |
Feature activation | +7.00e-02 |
Pos logit contributions | |
pand | +0.003 |
pill | +0.003 |
add | +0.003 |
otion | +0.003 |
hawk | +0.002 |
Neg logit contributions | |
forehead | -0.002 |
hand | -0.001 |
cheek | -0.001 |
bek | -0.001 |
lap | -0.001 |
Token |
|
Token ID | 198 |
Feature activation | -7.26e-02 |
Pos logit contributions | |
forehead | +0.002 |
bek | +0.002 |
hand | +0.002 |
plate | +0.002 |
cheek | +0.002 |
Neg logit contributions | |
pand | -0.003 |
add | -0.002 |
Tweet | -0.002 |
Hipp | -0.002 |
pill | -0.002 |
Token | Just |
Token ID | 5703 |
Feature activation | +5.88e-02 |
Pos logit contributions | |
pand | +0.002 |
hawk | +0.002 |
add | +0.002 |
pill | +0.002 |
Hipp | +0.002 |
Neg logit contributions | |
forehead | -0.003 |
hand | -0.003 |
cheek | -0.002 |
lap | -0.002 |
plate | -0.002 |
Token | a |
Token ID | 257 |
Feature activation | -6.83e-02 |
Pos logit contributions | |
forehead | +0.015 |
cheek | +0.014 |
bek | +0.013 |
lap | +0.013 |
hand | +0.012 |
Neg logit contributions | |
pand | -0.020 |
add | -0.018 |
pill | -0.018 |
otion | -0.017 |
hawk | -0.017 |
Token | needle |
Token ID | 17598 |
Feature activation | -5.00e-01 |
Pos logit contributions | |
pand | +0.003 |
hawk | +0.003 |
pill | +0.002 |
otion | +0.002 |
die | +0.002 |
Neg logit contributions | |
forehead | -0.002 |
cheek | -0.002 |
bek | -0.002 |
lap | -0.002 |
ae | -0.002 |
Token | and |
Token ID | 290 |
Feature activation | -7.13e-01 |
Pos logit contributions | |
pand | +0.023 |
pill | +0.021 |
hawk | +0.021 |
otion | +0.021 |
storms | +0.021 |
Neg logit contributions | |
forehead | -0.013 |
hand | -0.012 |
lap | -0.012 |
cheek | -0.012 |
bek | -0.011 |
Token | thread |
Token ID | 4704 |
Feature activation | -1.93e+00 |
Pos logit contributions | |
pand | +0.029 |
pill | +0.026 |
hawk | +0.026 |
add | +0.026 |
otion | +0.025 |
Neg logit contributions | |
forehead | -0.022 |
cheek | -0.021 |
lap | -0.021 |
hand | -0.020 |
bek | -0.020 |
Token | and |
Token ID | 290 |
Feature activation | -8.46e-01 |
Pos logit contributions | |
pand | +0.099 |
pill | +0.093 |
hawk | +0.093 |
add | +0.092 |
otion | +0.091 |
Neg logit contributions | |
hand | -0.039 |
forehead | -0.038 |
tight | -0.036 |
cheek | -0.034 |
lap | -0.034 |
Token | se |
Token ID | 384 |
Feature activation | -4.41e+00 |
Pos logit contributions | |
pand | +0.036 |
pill | +0.033 |
hawk | +0.032 |
add | +0.032 |
otion | +0.032 |
Neg logit contributions | |
forehead | -0.025 |
hand | -0.024 |
lap | -0.023 |
cheek | -0.023 |
bek | -0.022 |
Token | wed |
Token ID | 19103 |
Feature activation | +2.90e-01 |
Pos logit contributions | |
pand | +0.185 |
pill | +0.171 |
hawk | +0.169 |
otion | +0.166 |
add | +0.165 |
Neg logit contributions | |
forehead | -0.132 |
hand | -0.126 |
lap | -0.122 |
cheek | -0.121 |
tight | -0.120 |
Token | the |
Token ID | 262 |
Feature activation | -3.00e-01 |
Pos logit contributions | |
forehead | +0.009 |
hand | +0.009 |
lap | +0.009 |
cheek | +0.009 |
tight | +0.009 |
Neg logit contributions | |
pand | -0.011 |
pill | -0.010 |
hawk | -0.010 |
add | -0.010 |
otion | -0.010 |
Token | t |
Token ID | 256 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
pand | +0.012 |
pill | +0.011 |
hawk | +0.010 |
add | +0.010 |
otion | +0.010 |
Neg logit contributions | |
forehead | -0.010 |
hand | -0.010 |
lap | -0.009 |
cheek | -0.009 |
tight | -0.009 |
Token | eddy |
Token ID | 21874 |
Feature activation | +4.41e-01 |
Pos logit contributions | |
pand | +0.006 |
pill | +0.006 |
add | +0.006 |
otion | +0.005 |
hawk | +0.005 |
Neg logit contributions | |
forehead | -0.007 |
hand | -0.007 |
lap | -0.007 |
cheek | -0.007 |
list | -0.007 |
Token | bear |
Token ID | 6842 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
ae | +0.019 |
list | +0.019 |
bek | +0.019 |
forehead | +0.019 |
ossible | +0.019 |
Neg logit contributions | |
pand | -0.012 |
Tweet | -0.010 |
hawk | -0.010 |
pill | -0.010 |
storms | -0.010 |
Token | a |
Token ID | 257 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
pand | +0.017 |
hawk | +0.016 |
pill | +0.016 |
otion | +0.016 |
storms | +0.016 |
Neg logit contributions | |
hand | -0.010 |
forehead | -0.010 |
tight | -0.010 |
lap | -0.009 |
whole | -0.009 |
Token | needle |
Token ID | 17598 |
Feature activation | -5.34e-01 |
Pos logit contributions | |
forehead | +0.003 |
lap | +0.003 |
hand | +0.003 |
bek | +0.003 |
cheek | +0.003 |
Neg logit contributions | |
pand | -0.005 |
hawk | -0.005 |
pill | -0.004 |
otion | -0.004 |
add | -0.004 |
Token | and |
Token ID | 290 |
Feature activation | -7.99e-01 |
Pos logit contributions | |
pand | +0.025 |
pill | +0.023 |
hawk | +0.023 |
otion | +0.023 |
add | +0.023 |
Neg logit contributions | |
forehead | -0.013 |
hand | -0.012 |
cheek | -0.012 |
lap | -0.012 |
bek | -0.011 |
Token | thread |
Token ID | 4704 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
pand | +0.031 |
pill | +0.028 |
hawk | +0.028 |
add | +0.027 |
Tweet | +0.027 |
Neg logit contributions | |
forehead | -0.026 |
bek | -0.025 |
cheek | -0.025 |
lap | -0.024 |
hand | -0.024 |
Token | and |
Token ID | 290 |
Feature activation | -8.51e-01 |
Pos logit contributions | |
pand | +0.090 |
pill | +0.085 |
add | +0.084 |
hawk | +0.084 |
otion | +0.083 |
Neg logit contributions | |
hand | -0.027 |
forehead | -0.027 |
tight | -0.025 |
cheek | -0.023 |
lap | -0.023 |
Token | se |
Token ID | 384 |
Feature activation | -4.40e+00 |
Pos logit contributions | |
pand | +0.034 |
pill | +0.031 |
hawk | +0.031 |
add | +0.031 |
otion | +0.030 |
Neg logit contributions | |
forehead | -0.026 |
cheek | -0.025 |
hand | -0.024 |
lap | -0.024 |
bek | -0.024 |
Token | wed |
Token ID | 19103 |
Feature activation | +5.89e-01 |
Pos logit contributions | |
pand | +0.197 |
hawk | +0.184 |
pill | +0.183 |
otion | +0.179 |
storms | +0.176 |
Neg logit contributions | |
forehead | -0.120 |
hand | -0.114 |
tight | -0.110 |
lap | -0.109 |
cheek | -0.108 |
Token | Teddy |
Token ID | 29345 |
Feature activation | -6.72e-01 |
Pos logit contributions | |
forehead | +0.015 |
hand | +0.014 |
cheek | +0.014 |
lap | +0.014 |
tight | +0.013 |
Neg logit contributions | |
pand | -0.027 |
pill | -0.026 |
hawk | -0.025 |
add | -0.025 |
otion | -0.025 |
Token | 's |
Token ID | 338 |
Feature activation | -8.04e-01 |
Pos logit contributions | |
pand | +0.032 |
pill | +0.030 |
hawk | +0.030 |
add | +0.030 |
otion | +0.030 |
Neg logit contributions | |
forehead | -0.016 |
hand | -0.015 |
cheek | -0.015 |
lap | -0.014 |
tight | -0.014 |
Token | eye |
Token ID | 4151 |
Feature activation | +6.30e-02 |
Pos logit contributions | |
pand | +0.028 |
hawk | +0.024 |
add | +0.023 |
Tweet | +0.023 |
pill | +0.023 |
Neg logit contributions | |
bek | -0.030 |
ae | -0.027 |
forehead | -0.027 |
ossible | -0.026 |
list | -0.026 |
Token | back |
Token ID | 736 |
Feature activation | -5.08e-01 |
Pos logit contributions | |
forehead | +0.001 |
bek | +0.001 |
hand | +0.001 |
lap | +0.001 |
ossible | +0.001 |
Neg logit contributions | |
pand | -0.003 |
pill | -0.003 |
storms | -0.003 |
add | -0.003 |
otion | -0.003 |
Token | the |
Token ID | 262 |
Feature activation | -1.66e-01 |
Pos logit contributions | |
pand | +0.031 |
pill | +0.029 |
hawk | +0.029 |
otion | +0.028 |
add | +0.028 |
Neg logit contributions | |
forehead | -0.023 |
hand | -0.022 |
tight | -0.021 |
lap | -0.021 |
cheek | -0.021 |
Token | t |
Token ID | 256 |
Feature activation | +3.68e-01 |
Pos logit contributions | |
pand | +0.007 |
hawk | +0.006 |
pill | +0.006 |
add | +0.006 |
otion | +0.006 |
Neg logit contributions | |
forehead | -0.005 |
bek | -0.005 |
hand | -0.005 |
lap | -0.005 |
cheek | -0.005 |
Token | eddy |
Token ID | 21874 |
Feature activation | +7.26e-01 |
Pos logit contributions | |
forehead | +0.015 |
hand | +0.014 |
lap | +0.014 |
cheek | +0.014 |
list | +0.014 |
Neg logit contributions | |
pand | -0.011 |
pill | -0.010 |
add | -0.010 |
hawk | -0.010 |
otion | -0.010 |
Token | bear |
Token ID | 6842 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
bek | +0.032 |
list | +0.031 |
forehead | +0.031 |
ae | +0.031 |
lap | +0.031 |
Neg logit contributions | |
pand | -0.020 |
hawk | -0.017 |
Tweet | -0.016 |
Hipp | -0.016 |
aliens | -0.016 |
Token | and |
Token ID | 290 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
pand | +0.044 |
Tweet | +0.039 |
pill | +0.039 |
hawk | +0.039 |
storms | +0.039 |
Neg logit contributions | |
forehead | -0.028 |
bek | -0.026 |
cheek | -0.026 |
lap | -0.025 |
hand | -0.024 |
Token | se |
Token ID | 384 |
Feature activation | -4.36e+00 |
Pos logit contributions | |
pand | +0.059 |
pill | +0.054 |
hawk | +0.053 |
add | +0.052 |
otion | +0.052 |
Neg logit contributions | |
forehead | -0.039 |
hand | -0.037 |
cheek | -0.036 |
lap | -0.036 |
bek | -0.035 |
Token | wed |
Token ID | 19103 |
Feature activation | +5.75e-02 |
Pos logit contributions | |
pand | +0.180 |
pill | +0.167 |
hawk | +0.165 |
otion | +0.162 |
add | +0.161 |
Neg logit contributions | |
forehead | -0.133 |
hand | -0.127 |
lap | -0.123 |
cheek | -0.122 |
tight | -0.121 |
Token | up |
Token ID | 510 |
Feature activation | -5.35e-01 |
Pos logit contributions | |
forehead | +0.002 |
hand | +0.002 |
lap | +0.002 |
cheek | +0.002 |
tight | +0.002 |
Neg logit contributions | |
pand | -0.002 |
pill | -0.002 |
hawk | -0.002 |
add | -0.002 |
otion | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | -3.23e-01 |
Pos logit contributions | |
pand | +0.026 |
otion | +0.025 |
hawk | +0.025 |
pill | +0.025 |
add | +0.025 |
Neg logit contributions | |
tight | -0.012 |
forehead | -0.012 |
hand | -0.011 |
lap | -0.010 |
whole | -0.010 |
Token | hole |
Token ID | 7604 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
pand | +0.012 |
pill | +0.011 |
hawk | +0.011 |
add | +0.011 |
otion | +0.011 |
Neg logit contributions | |
forehead | -0.011 |
hand | -0.010 |
bek | -0.010 |
lap | -0.010 |
cheek | -0.010 |
Token | in |
Token ID | 287 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
pand | +0.055 |
otion | +0.051 |
add | +0.050 |
storms | +0.050 |
pill | +0.050 |
Neg logit contributions | |
forehead | -0.029 |
bek | -0.027 |
hand | -0.026 |
lap | -0.026 |
cheek | -0.026 |
Token | jumping |
Token ID | 14284 |
Feature activation | +5.55e-02 |
Pos logit contributions | |
add | +0.015 |
pand | +0.015 |
pill | +0.015 |
hawk | +0.014 |
otion | +0.014 |
Neg logit contributions | |
hand | -0.008 |
tight | -0.008 |
forehead | -0.007 |
whole | -0.007 |
lap | -0.007 |
Token | in |
Token ID | 287 |
Feature activation | +3.84e-01 |
Pos logit contributions | |
tight | +0.001 |
hand | +0.001 |
forehead | +0.001 |
whole | +0.001 |
cheek | +0.001 |
Neg logit contributions | |
pand | -0.003 |
pill | -0.003 |
otion | -0.003 |
add | -0.003 |
hawk | -0.003 |
Token | the |
Token ID | 262 |
Feature activation | +2.30e-01 |
Pos logit contributions | |
forehead | +0.009 |
hand | +0.008 |
cheek | +0.008 |
lap | +0.008 |
tight | +0.008 |
Neg logit contributions | |
pand | -0.019 |
pill | -0.018 |
hawk | -0.017 |
add | -0.017 |
otion | -0.017 |
Token | water |
Token ID | 1660 |
Feature activation | +7.22e-01 |
Pos logit contributions | |
forehead | +0.008 |
bek | +0.008 |
cheek | +0.008 |
lap | +0.008 |
hand | +0.008 |
Neg logit contributions | |
pand | -0.008 |
hawk | -0.008 |
pill | -0.007 |
add | -0.007 |
flashes | -0.007 |
Token | and |
Token ID | 290 |
Feature activation | +3.19e-01 |
Pos logit contributions | |
forehead | +0.016 |
hand | +0.015 |
lap | +0.014 |
cheek | +0.014 |
tight | +0.014 |
Neg logit contributions | |
pand | -0.036 |
pill | -0.034 |
add | -0.033 |
hawk | -0.033 |
otion | -0.033 |
Token | se |
Token ID | 384 |
Feature activation | -4.18e+00 |
Pos logit contributions | |
forehead | +0.010 |
hand | +0.009 |
cheek | +0.009 |
lap | +0.009 |
tight | +0.009 |
Neg logit contributions | |
pand | -0.013 |
pill | -0.012 |
add | -0.012 |
otion | -0.012 |
hawk | -0.012 |
Token | ag |
Token ID | 363 |
Feature activation | -8.33e-02 |
Pos logit contributions | |
pand | +0.175 |
hawk | +0.160 |
pill | +0.160 |
otion | +0.157 |
storms | +0.152 |
Neg logit contributions | |
forehead | -0.127 |
hand | -0.120 |
lap | -0.119 |
bek | -0.119 |
tight | -0.117 |
Token | ull |
Token ID | 724 |
Feature activation | -2.27e+00 |
Pos logit contributions | |
pand | +0.002 |
pill | +0.002 |
hawk | +0.002 |
storms | +0.002 |
add | +0.002 |
Neg logit contributions | |
hand | -0.004 |
forehead | -0.004 |
whole | -0.004 |
lap | -0.004 |
tight | -0.004 |
Token | s |
Token ID | 82 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
pand | +0.106 |
otion | +0.101 |
add | +0.099 |
pill | +0.099 |
hawk | +0.098 |
Neg logit contributions | |
forehead | -0.054 |
hand | -0.054 |
tight | -0.052 |
cheek | -0.051 |
whole | -0.050 |
Token | soaring |
Token ID | 32376 |
Feature activation | -5.27e-01 |
Pos logit contributions | |
pand | +0.077 |
pill | +0.073 |
add | +0.071 |
otion | +0.070 |
hawk | +0.070 |
Neg logit contributions | |
forehead | -0.046 |
hand | -0.045 |
cheek | -0.042 |
lap | -0.042 |
tight | -0.042 |
Token | in |
Token ID | 287 |
Feature activation | +4.48e-01 |
Pos logit contributions | |
pand | +0.023 |
hawk | +0.022 |
add | +0.022 |
storms | +0.021 |
pill | +0.021 |
Neg logit contributions | |
forehead | -0.015 |
lap | -0.013 |
hand | -0.013 |
cheek | -0.013 |
plate | -0.013 |
Token | the |
Token ID | 262 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
pand | +0.001 |
add | +0.001 |
pill | +0.001 |
otion | +0.001 |
hawk | +0.001 |
Neg logit contributions | |
forehead | -0.001 |
cheek | -0.001 |
lap | -0.001 |
bek | -0.001 |
hand | -0.001 |
Token | needle |
Token ID | 17598 |
Feature activation | -1.26e-03 |
Pos logit contributions | |
bek | +0.004 |
ae | +0.004 |
forehead | +0.004 |
satisfied | +0.004 |
cle | +0.003 |
Neg logit contributions | |
pand | -0.004 |
pill | -0.004 |
otion | -0.004 |
add | -0.004 |
hawk | -0.004 |
Token | and |
Token ID | 290 |
Feature activation | -9.55e-01 |
Pos logit contributions | |
pand | +0.000 |
pill | +0.000 |
add | +0.000 |
hawk | +0.000 |
otion | +0.000 |
Neg logit contributions | |
forehead | -0.000 |
hand | -0.000 |
lap | -0.000 |
cheek | -0.000 |
tight | -0.000 |
Token | thread |
Token ID | 4704 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
pand | +0.037 |
flashes | +0.032 |
Tweet | +0.032 |
spr | +0.031 |
hawk | +0.031 |
Neg logit contributions | |
forehead | -0.032 |
bek | -0.031 |
cheek | -0.031 |
ae | -0.029 |
lap | -0.029 |
Token | and |
Token ID | 290 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
pand | +0.076 |
pill | +0.072 |
hawk | +0.071 |
add | +0.071 |
storms | +0.070 |
Neg logit contributions | |
hand | -0.026 |
tight | -0.025 |
forehead | -0.024 |
lap | -0.023 |
cheek | -0.022 |
Token | se |
Token ID | 384 |
Feature activation | -4.17e+00 |
Pos logit contributions | |
pand | +0.045 |
flashes | +0.039 |
pill | +0.039 |
hawk | +0.039 |
Tweet | +0.039 |
Neg logit contributions | |
forehead | -0.035 |
cheek | -0.034 |
bek | -0.033 |
lap | -0.032 |
ae | -0.031 |
Token | wed |
Token ID | 19103 |
Feature activation | +4.12e-01 |
Pos logit contributions | |
pand | +0.174 |
pill | +0.161 |
hawk | +0.159 |
otion | +0.157 |
add | +0.155 |
Neg logit contributions | |
forehead | -0.125 |
hand | -0.120 |
lap | -0.116 |
cheek | -0.116 |
tight | -0.114 |
Token | the |
Token ID | 262 |
Feature activation | -2.62e-01 |
Pos logit contributions | |
forehead | +0.012 |
hand | +0.012 |
lap | +0.012 |
cheek | +0.012 |
tight | +0.011 |
Neg logit contributions | |
pand | -0.017 |
pill | -0.016 |
hawk | -0.016 |
add | -0.015 |
otion | -0.015 |
Token | hole |
Token ID | 7604 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
pand | +0.008 |
pill | +0.007 |
hawk | +0.007 |
add | +0.006 |
otion | +0.006 |
Neg logit contributions | |
bek | -0.011 |
ae | -0.010 |
forehead | -0.010 |
lap | -0.010 |
satisfied | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
pand | +0.067 |
pill | +0.062 |
add | +0.062 |
otion | +0.061 |
hawk | +0.061 |
Neg logit contributions | |
forehead | -0.028 |
hand | -0.026 |
lap | -0.025 |
cheek | -0.025 |
bek | -0.024 |
Token | She |
Token ID | 1375 |
Feature activation | -8.44e-01 |
Pos logit contributions | |
pand | +0.034 |
pill | +0.031 |
otion | +0.031 |
storms | +0.030 |
add | +0.030 |
Neg logit contributions | |
forehead | -0.041 |
cheek | -0.039 |
bek | -0.038 |
hand | -0.038 |
lap | -0.037 |
Token | a |
Token ID | 257 |
Feature activation | +5.52e-01 |
Pos logit contributions | |
forehead | +0.006 |
hand | +0.005 |
lap | +0.005 |
cheek | +0.005 |
tight | +0.005 |
Neg logit contributions | |
pand | -0.010 |
pill | -0.009 |
hawk | -0.009 |
add | -0.009 |
otion | -0.009 |
Token | needle |
Token ID | 17598 |
Feature activation | -9.33e-02 |
Pos logit contributions | |
forehead | +0.017 |
lap | +0.016 |
cheek | +0.015 |
hand | +0.015 |
bek | +0.015 |
Neg logit contributions | |
pand | -0.022 |
hawk | -0.020 |
pill | -0.019 |
add | -0.019 |
die | -0.019 |
Token | and |
Token ID | 290 |
Feature activation | -6.44e-01 |
Pos logit contributions | |
pand | +0.005 |
pill | +0.005 |
add | +0.004 |
hawk | +0.004 |
otion | +0.004 |
Neg logit contributions | |
forehead | -0.002 |
hand | -0.002 |
tight | -0.002 |
cheek | -0.002 |
lap | -0.002 |
Token | thread |
Token ID | 4704 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
pand | +0.024 |
pill | +0.022 |
hawk | +0.022 |
add | +0.021 |
flashes | +0.021 |
Neg logit contributions | |
forehead | -0.021 |
cheek | -0.020 |
lap | -0.020 |
bek | -0.020 |
hand | -0.020 |
Token | and |
Token ID | 290 |
Feature activation | -8.01e-01 |
Pos logit contributions | |
pand | +0.077 |
pill | +0.073 |
hawk | +0.072 |
add | +0.071 |
otion | +0.069 |
Neg logit contributions | |
hand | -0.028 |
forehead | -0.027 |
tight | -0.027 |
cheek | -0.025 |
lap | -0.024 |
Token | se |
Token ID | 384 |
Feature activation | -4.16e+00 |
Pos logit contributions | |
pand | +0.032 |
pill | +0.029 |
hawk | +0.029 |
add | +0.028 |
otion | +0.028 |
Neg logit contributions | |
forehead | -0.025 |
cheek | -0.024 |
lap | -0.024 |
hand | -0.023 |
bek | -0.023 |
Token | wed |
Token ID | 19103 |
Feature activation | +5.55e-01 |
Pos logit contributions | |
pand | +0.179 |
pill | +0.165 |
hawk | +0.164 |
otion | +0.161 |
storms | +0.159 |
Neg logit contributions | |
forehead | -0.120 |
hand | -0.116 |
lap | -0.111 |
tight | -0.110 |
cheek | -0.110 |
Token | a |
Token ID | 257 |
Feature activation | +6.04e-02 |
Pos logit contributions | |
forehead | +0.016 |
hand | +0.015 |
lap | +0.015 |
cheek | +0.015 |
bek | +0.014 |
Neg logit contributions | |
pand | -0.024 |
pill | -0.022 |
otion | -0.022 |
add | -0.022 |
hawk | -0.022 |
Token | patch |
Token ID | 8529 |
Feature activation | -9.43e-01 |
Pos logit contributions | |
forehead | +0.002 |
bek | +0.002 |
ossible | +0.002 |
lap | +0.002 |
ae | +0.002 |
Neg logit contributions | |
pand | -0.002 |
hawk | -0.002 |
pill | -0.002 |
Tweet | -0.002 |
die | -0.002 |
Token | on |
Token ID | 319 |
Feature activation | +3.95e-02 |
Pos logit contributions | |
pand | +0.038 |
otion | +0.037 |
add | +0.035 |
storms | +0.035 |
hawk | +0.035 |
Neg logit contributions | |
forehead | -0.027 |
bek | -0.026 |
lap | -0.025 |
hand | -0.025 |
cheek | -0.024 |
Token | Patch |
Token ID | 17106 |
Feature activation | -6.91e-01 |
Pos logit contributions | |
forehead | +0.001 |
cheek | +0.001 |
hand | +0.001 |
lap | +0.001 |
bek | +0.001 |
Neg logit contributions | |
pand | -0.002 |
pill | -0.001 |
hawk | -0.001 |
otion | -0.001 |
storms | -0.001 |
Token | , |
Token ID | 11 |
Feature activation | +2.84e-01 |
Pos logit contributions | |
pand | +0.013 |
add | +0.013 |
pill | +0.013 |
hawk | +0.013 |
storms | +0.013 |
Neg logit contributions | |
hand | -0.005 |
tight | -0.004 |
full | -0.004 |
whole | -0.004 |
forehead | -0.004 |
Token | sand |
Token ID | 6450 |
Feature activation | -2.08e-01 |
Pos logit contributions | |
forehead | +0.011 |
hand | +0.011 |
lap | +0.010 |
tight | +0.010 |
cheek | +0.010 |
Neg logit contributions | |
pand | -0.009 |
otion | -0.009 |
pill | -0.009 |
hawk | -0.009 |
add | -0.009 |
Token | cast |
Token ID | 2701 |
Feature activation | -3.12e-01 |
Pos logit contributions | |
pand | +0.010 |
pill | +0.010 |
hawk | +0.009 |
add | +0.009 |
otion | +0.009 |
Neg logit contributions | |
hand | -0.005 |
forehead | -0.004 |
tight | -0.004 |
hands | -0.004 |
whole | -0.004 |
Token | les |
Token ID | 829 |
Feature activation | +2.34e-01 |
Pos logit contributions | |
pand | +0.008 |
pill | +0.007 |
hawk | +0.007 |
add | +0.007 |
otion | +0.007 |
Neg logit contributions | |
forehead | -0.014 |
hand | -0.013 |
lap | -0.013 |
cheek | -0.013 |
tight | -0.013 |
Token | and |
Token ID | 290 |
Feature activation | +3.88e-01 |
Pos logit contributions | |
hand | +0.005 |
forehead | +0.004 |
tight | +0.004 |
lap | +0.004 |
whole | +0.004 |
Neg logit contributions | |
pand | -0.012 |
add | -0.012 |
hawk | -0.011 |
pill | -0.011 |
otion | -0.011 |
Token | se |
Token ID | 384 |
Feature activation | -4.11e+00 |
Pos logit contributions | |
forehead | +0.013 |
hand | +0.012 |
lap | +0.012 |
tight | +0.012 |
cheek | +0.012 |
Neg logit contributions | |
pand | -0.015 |
hawk | -0.014 |
otion | -0.014 |
add | -0.014 |
pill | -0.014 |
Token | ag |
Token ID | 363 |
Feature activation | -2.30e-01 |
Pos logit contributions | |
pand | +0.173 |
hawk | +0.162 |
pill | +0.160 |
otion | +0.158 |
storms | +0.154 |
Neg logit contributions | |
forehead | -0.120 |
hand | -0.115 |
lap | -0.112 |
tight | -0.110 |
cheek | -0.110 |
Token | ull |
Token ID | 724 |
Feature activation | -2.54e+00 |
Pos logit contributions | |
pand | +0.006 |
pill | +0.006 |
hawk | +0.006 |
add | +0.006 |
otion | +0.005 |
Neg logit contributions | |
forehead | -0.010 |
hand | -0.010 |
lap | -0.009 |
cheek | -0.009 |
tight | -0.009 |
Token | s |
Token ID | 82 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
pand | +0.113 |
otion | +0.108 |
hawk | +0.107 |
add | +0.106 |
pill | +0.106 |
Neg logit contributions | |
forehead | -0.065 |
hand | -0.064 |
tight | -0.061 |
cheek | -0.061 |
lap | -0.060 |
Token | . |
Token ID | 13 |
Feature activation | +4.94e-01 |
Pos logit contributions | |
pand | +0.066 |
pill | +0.061 |
hawk | +0.060 |
add | +0.060 |
otion | +0.060 |
Neg logit contributions | |
forehead | -0.042 |
hand | -0.038 |
cheek | -0.038 |
bek | -0.038 |
lap | -0.038 |
Token | Jane |
Token ID | 12091 |
Feature activation | +7.62e-01 |
Pos logit contributions | |
forehead | +0.015 |
hand | +0.014 |
cheek | +0.014 |
lap | +0.014 |
bek | +0.014 |
Neg logit contributions | |
pand | -0.020 |
pill | -0.019 |
add | -0.019 |
otion | -0.018 |
storms | -0.018 |
Token | thread |
Token ID | 4704 |
Feature activation | -2.74e-01 |
Pos logit contributions | |
forehead | +0.004 |
hand | +0.004 |
lap | +0.003 |
cheek | +0.003 |
tight | +0.003 |
Neg logit contributions | |
pand | -0.005 |
pill | -0.004 |
hawk | -0.004 |
add | -0.004 |
otion | -0.004 |
Token | and |
Token ID | 290 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
pand | +0.014 |
pill | +0.014 |
hawk | +0.013 |
add | +0.013 |
otion | +0.013 |
Neg logit contributions | |
forehead | -0.005 |
hand | -0.005 |
tight | -0.005 |
cheek | -0.005 |
lap | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | -9.41e-01 |
Pos logit contributions | |
pand | +0.006 |
pill | +0.005 |
hawk | +0.005 |
add | +0.005 |
otion | +0.005 |
Neg logit contributions | |
forehead | -0.004 |
hand | -0.004 |
lap | -0.003 |
cheek | -0.003 |
tight | -0.003 |
Token | needle |
Token ID | 17598 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
pand | +0.035 |
hawk | +0.032 |
pill | +0.031 |
die | +0.030 |
add | +0.030 |
Neg logit contributions | |
forehead | -0.030 |
bek | -0.028 |
ae | -0.028 |
lap | -0.027 |
cheek | -0.027 |
Token | and |
Token ID | 290 |
Feature activation | -6.30e-01 |
Pos logit contributions | |
pand | +0.063 |
pill | +0.060 |
hawk | +0.059 |
add | +0.058 |
otion | +0.057 |
Neg logit contributions | |
hand | -0.024 |
forehead | -0.024 |
tight | -0.022 |
cheek | -0.022 |
lap | -0.021 |
Token | se |
Token ID | 384 |
Feature activation | -4.09e+00 |
Pos logit contributions | |
pand | +0.025 |
pill | +0.023 |
hawk | +0.022 |
add | +0.022 |
otion | +0.022 |
Neg logit contributions | |
forehead | -0.020 |
cheek | -0.019 |
hand | -0.019 |
lap | -0.019 |
bek | -0.018 |
Token | wed |
Token ID | 19103 |
Feature activation | +5.16e-01 |
Pos logit contributions | |
pand | +0.170 |
pill | +0.157 |
hawk | +0.155 |
otion | +0.153 |
add | +0.151 |
Neg logit contributions | |
forehead | -0.124 |
hand | -0.118 |
lap | -0.114 |
cheek | -0.114 |
tight | -0.112 |
Token | up |
Token ID | 510 |
Feature activation | -5.57e-02 |
Pos logit contributions | |
forehead | +0.017 |
hand | +0.017 |
tight | +0.016 |
cheek | +0.016 |
lap | +0.016 |
Neg logit contributions | |
pand | -0.020 |
pill | -0.019 |
hawk | -0.018 |
otion | -0.018 |
add | -0.018 |
Token | the |
Token ID | 262 |
Feature activation | -1.79e-01 |
Pos logit contributions | |
pand | +0.003 |
otion | +0.003 |
pill | +0.003 |
hawk | +0.003 |
add | +0.003 |
Neg logit contributions | |
tight | -0.001 |
hand | -0.001 |
forehead | -0.001 |
whole | -0.001 |
lap | -0.001 |
Token | tear |
Token ID | 11626 |
Feature activation | +2.60e-01 |
Pos logit contributions | |
pand | +0.007 |
pill | +0.007 |
add | +0.007 |
otion | +0.007 |
hawk | +0.007 |
Neg logit contributions | |
forehead | -0.006 |
hand | -0.005 |
cheek | -0.005 |
lap | -0.005 |
tight | -0.005 |
Token | in |
Token ID | 287 |
Feature activation | +7.19e-02 |
Pos logit contributions | |
forehead | +0.005 |
hand | +0.005 |
tight | +0.005 |
cheek | +0.005 |
whole | +0.005 |
Neg logit contributions | |
pand | -0.014 |
pill | -0.013 |
hawk | -0.013 |
add | -0.013 |
otion | -0.013 |
Token | needle |
Token ID | 17598 |
Feature activation | -4.92e-01 |
Pos logit contributions | |
forehead | +0.007 |
bek | +0.007 |
ae | +0.007 |
cheek | +0.007 |
lap | +0.007 |
Neg logit contributions | |
pand | -0.010 |
hawk | -0.009 |
pill | -0.009 |
die | -0.008 |
otion | -0.008 |
Token | and |
Token ID | 290 |
Feature activation | -1.56e-01 |
Pos logit contributions | |
pand | +0.024 |
pill | +0.022 |
hawk | +0.022 |
add | +0.022 |
otion | +0.022 |
Neg logit contributions | |
forehead | -0.012 |
hand | -0.011 |
lap | -0.011 |
cheek | -0.011 |
tight | -0.010 |
Token | thread |
Token ID | 4704 |
Feature activation | -1.90e+00 |
Pos logit contributions | |
pand | +0.006 |
pill | +0.005 |
hawk | +0.005 |
flashes | +0.005 |
Tweet | +0.005 |
Neg logit contributions | |
forehead | -0.005 |
cheek | -0.005 |
bek | -0.005 |
lap | -0.005 |
whole | -0.005 |
Token | and |
Token ID | 290 |
Feature activation | -3.90e-01 |
Pos logit contributions | |
pand | +0.096 |
pill | +0.091 |
hawk | +0.089 |
add | +0.089 |
otion | +0.088 |
Neg logit contributions | |
forehead | -0.040 |
hand | -0.039 |
tight | -0.037 |
cheek | -0.036 |
lap | -0.036 |
Token | carefully |
Token ID | 7773 |
Feature activation | -9.26e-01 |
Pos logit contributions | |
pand | +0.015 |
pill | +0.014 |
hawk | +0.013 |
flashes | +0.013 |
Tweet | +0.013 |
Neg logit contributions | |
forehead | -0.012 |
cheek | -0.012 |
lap | -0.012 |
bek | -0.012 |
hand | -0.011 |
Token | se |
Token ID | 384 |
Feature activation | -4.08e+00 |
Pos logit contributions | |
pand | +0.041 |
pill | +0.038 |
add | +0.037 |
otion | +0.037 |
hawk | +0.037 |
Neg logit contributions | |
forehead | -0.025 |
hand | -0.024 |
cheek | -0.023 |
lap | -0.023 |
bek | -0.022 |
Token | wed |
Token ID | 19103 |
Feature activation | +8.24e-01 |
Pos logit contributions | |
pand | +0.175 |
hawk | +0.161 |
pill | +0.160 |
otion | +0.157 |
storms | +0.155 |
Neg logit contributions | |
forehead | -0.119 |
hand | -0.114 |
lap | -0.110 |
cheek | -0.109 |
tight | -0.109 |
Token | Mr |
Token ID | 1770 |
Feature activation | +5.85e-01 |
Pos logit contributions | |
forehead | +0.026 |
hand | +0.024 |
lap | +0.024 |
cheek | +0.024 |
bek | +0.023 |
Neg logit contributions | |
pand | -0.034 |
pill | -0.031 |
add | -0.030 |
hawk | -0.030 |
otion | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | +4.83e-01 |
Pos logit contributions | |
forehead | +0.010 |
hand | +0.010 |
cheek | +0.009 |
tight | +0.009 |
lap | +0.008 |
Neg logit contributions | |
pand | -0.032 |
add | -0.031 |
pill | -0.031 |
otion | -0.030 |
hawk | -0.030 |
Token | Fl |
Token ID | 1610 |
Feature activation | -2.38e-01 |
Pos logit contributions | |
forehead | +0.019 |
hand | +0.018 |
lap | +0.018 |
cheek | +0.018 |
tight | +0.017 |
Neg logit contributions | |
pand | -0.016 |
pill | -0.014 |
add | -0.014 |
hawk | -0.014 |
otion | -0.014 |
Token | uffy |
Token ID | 15352 |
Feature activation | -8.12e-02 |
Pos logit contributions | |
pand | +0.009 |
hawk | +0.008 |
pill | +0.008 |
otion | +0.008 |
flashes | +0.008 |
Neg logit contributions | |
forehead | -0.008 |
hand | -0.008 |
cheek | -0.008 |
tight | -0.008 |
whole | -0.008 |
Token | of |
Token ID | 286 |
Feature activation | -6.48e-01 |
Pos logit contributions | |
pand | +0.073 |
pill | +0.072 |
add | +0.071 |
hawk | +0.071 |
otion | +0.070 |
Neg logit contributions | |
hand | -0.030 |
tight | -0.029 |
forehead | -0.027 |
whole | -0.027 |
lap | -0.027 |
Token | waves |
Token ID | 9813 |
Feature activation | +1.61e-01 |
Pos logit contributions | |
otion | +0.028 |
storms | +0.027 |
pill | +0.027 |
ican | +0.027 |
pand | +0.026 |
Neg logit contributions | |
hand | -0.019 |
tight | -0.018 |
whole | -0.017 |
forehead | -0.017 |
hands | -0.017 |
Token | and |
Token ID | 290 |
Feature activation | -2.50e-01 |
Pos logit contributions | |
hand | +0.003 |
forehead | +0.003 |
tight | +0.003 |
cheek | +0.003 |
lap | +0.003 |
Neg logit contributions | |
pand | -0.008 |
pill | -0.008 |
add | -0.008 |
otion | -0.008 |
hawk | -0.008 |
Token | wind |
Token ID | 2344 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
pand | +0.009 |
pill | +0.009 |
hawk | +0.009 |
add | +0.009 |
otion | +0.009 |
Neg logit contributions | |
forehead | -0.008 |
hand | -0.008 |
lap | -0.008 |
cheek | -0.008 |
tight | -0.008 |
Token | and |
Token ID | 290 |
Feature activation | -7.02e-01 |
Pos logit contributions | |
pand | +0.045 |
pill | +0.043 |
hawk | +0.042 |
add | +0.041 |
otion | +0.041 |
Neg logit contributions | |
forehead | -0.031 |
hand | -0.029 |
cheek | -0.029 |
tight | -0.028 |
lap | -0.028 |
Token | se |
Token ID | 384 |
Feature activation | -4.08e+00 |
Pos logit contributions | |
pand | +0.027 |
pill | +0.025 |
add | +0.025 |
hawk | +0.025 |
otion | +0.025 |
Neg logit contributions | |
forehead | -0.023 |
hand | -0.022 |
lap | -0.021 |
cheek | -0.021 |
tight | -0.021 |
Token | ag |
Token ID | 363 |
Feature activation | -3.89e-01 |
Pos logit contributions | |
pand | +0.183 |
hawk | +0.173 |
pill | +0.169 |
otion | +0.168 |
Hipp | +0.162 |
Neg logit contributions | |
forehead | -0.109 |
hand | -0.103 |
lap | -0.100 |
tight | -0.099 |
cheek | -0.099 |
Token | ull |
Token ID | 724 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
pand | +0.011 |
pill | +0.009 |
hawk | +0.009 |
add | +0.009 |
otion | +0.009 |
Neg logit contributions | |
forehead | -0.017 |
hand | -0.017 |
lap | -0.016 |
cheek | -0.016 |
tight | -0.016 |
Token | s |
Token ID | 82 |
Feature activation | -2.03e+00 |
Pos logit contributions | |
pand | +0.131 |
otion | +0.125 |
pill | +0.124 |
hawk | +0.123 |
add | +0.122 |
Neg logit contributions | |
forehead | -0.068 |
hand | -0.068 |
tight | -0.064 |
cheek | -0.064 |
lap | -0.063 |
Token | . |
Token ID | 13 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
pand | +0.100 |
pill | +0.095 |
add | +0.093 |
hawk | +0.092 |
otion | +0.092 |
Neg logit contributions | |
forehead | -0.046 |
hand | -0.045 |
lap | -0.042 |
cheek | -0.042 |
tight | -0.041 |
Token |
|
Token ID | 198 |
Feature activation | +2.30e-01 |
Pos logit contributions | |
forehead | +0.004 |
hand | +0.004 |
lap | +0.004 |
cheek | +0.004 |
tight | +0.004 |
Neg logit contributions | |
pand | -0.008 |
pill | -0.007 |
hawk | -0.007 |
add | -0.007 |
otion | -0.007 |
Token | needle |
Token ID | 17598 |
Feature activation | -7.74e-01 |
Pos logit contributions | |
pand | +0.006 |
pill | +0.006 |
add | +0.006 |
hawk | +0.006 |
otion | +0.005 |
Neg logit contributions | |
forehead | -0.005 |
hand | -0.005 |
cheek | -0.005 |
lap | -0.005 |
bek | -0.005 |
Token | and |
Token ID | 290 |
Feature activation | -6.38e-01 |
Pos logit contributions | |
pand | +0.038 |
pill | +0.036 |
add | +0.035 |
hawk | +0.035 |
otion | +0.035 |
Neg logit contributions | |
forehead | -0.017 |
hand | -0.016 |
cheek | -0.015 |
lap | -0.015 |
tight | -0.015 |
Token | thread |
Token ID | 4704 |
Feature activation | -1.91e+00 |
Pos logit contributions | |
pand | +0.025 |
pill | +0.023 |
hawk | +0.022 |
Tweet | +0.022 |
flashes | +0.022 |
Neg logit contributions | |
forehead | -0.020 |
cheek | -0.019 |
bek | -0.019 |
lap | -0.018 |
whole | -0.018 |
Token | and |
Token ID | 290 |
Feature activation | -7.24e-01 |
Pos logit contributions | |
pand | +0.098 |
pill | +0.093 |
hawk | +0.091 |
otion | +0.091 |
add | +0.091 |
Neg logit contributions | |
hand | -0.038 |
forehead | -0.038 |
tight | -0.036 |
cheek | -0.034 |
lap | -0.034 |
Token | carefully |
Token ID | 7773 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
pand | +0.030 |
pill | +0.027 |
hawk | +0.026 |
Tweet | +0.026 |
add | +0.026 |
Neg logit contributions | |
forehead | -0.021 |
cheek | -0.020 |
bek | -0.020 |
lap | -0.020 |
hand | -0.020 |
Token | se |
Token ID | 384 |
Feature activation | -4.07e+00 |
Pos logit contributions | |
pand | +0.048 |
pill | +0.045 |
add | +0.044 |
hawk | +0.044 |
otion | +0.044 |
Neg logit contributions | |
forehead | -0.028 |
hand | -0.026 |
cheek | -0.025 |
lap | -0.025 |
bek | -0.024 |
Token | wed |
Token ID | 19103 |
Feature activation | +7.79e-01 |
Pos logit contributions | |
pand | +0.176 |
pill | +0.162 |
hawk | +0.162 |
otion | +0.159 |
storms | +0.156 |
Neg logit contributions | |
forehead | -0.117 |
hand | -0.112 |
lap | -0.108 |
tight | -0.108 |
cheek | -0.108 |
Token | the |
Token ID | 262 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
forehead | +0.022 |
hand | +0.020 |
lap | +0.020 |
cheek | +0.020 |
bek | +0.019 |
Neg logit contributions | |
pand | -0.034 |
pill | -0.032 |
add | -0.031 |
hawk | -0.031 |
otion | -0.031 |
Token | hole |
Token ID | 7604 |
Feature activation | -8.47e-01 |
Pos logit contributions | |
bek | +0.000 |
forehead | +0.000 |
ae | +0.000 |
hand | +0.000 |
lap | +0.000 |
Neg logit contributions | |
pand | -0.000 |
pill | -0.000 |
hawk | -0.000 |
storms | -0.000 |
otion | -0.000 |
Token | shut |
Token ID | 4423 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
pand | +0.039 |
otion | +0.036 |
add | +0.035 |
storms | +0.035 |
pill | +0.035 |
Neg logit contributions | |
forehead | -0.021 |
bek | -0.019 |
lap | -0.019 |
hand | -0.019 |
cheek | -0.019 |
Token | . |
Token ID | 13 |
Feature activation | +4.63e-01 |
Pos logit contributions | |
pand | +0.005 |
hawk | +0.004 |
storms | +0.004 |
add | +0.004 |
Tweet | +0.004 |
Neg logit contributions | |
forehead | -0.003 |
bek | -0.003 |
ossible | -0.003 |
cheek | -0.003 |
lap | -0.003 |
Token | You |
Token ID | 921 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
pand | +0.015 |
pill | +0.014 |
hawk | +0.014 |
add | +0.014 |
otion | +0.014 |
Neg logit contributions | |
forehead | -0.013 |
hand | -0.013 |
lap | -0.012 |
cheek | -0.012 |
tight | -0.012 |
Token | must |
Token ID | 1276 |
Feature activation | -3.66e-01 |
Pos logit contributions | |
forehead | +0.040 |
hand | +0.038 |
lap | +0.037 |
cheek | +0.037 |
tight | +0.036 |
Neg logit contributions | |
pand | -0.047 |
pill | -0.044 |
hawk | -0.043 |
add | -0.043 |
otion | -0.042 |
Token | have |
Token ID | 423 |
Feature activation | +3.73e-01 |
Pos logit contributions | |
pand | +0.016 |
pill | +0.015 |
hawk | +0.015 |
add | +0.015 |
otion | +0.015 |
Neg logit contributions | |
forehead | -0.010 |
hand | -0.010 |
cheek | -0.009 |
lap | -0.009 |
tight | -0.009 |
Token | let |
Token ID | 1309 |
Feature activation | -5.96e-01 |
Pos logit contributions | |
forehead | +0.010 |
hand | +0.009 |
lap | +0.009 |
cheek | +0.009 |
tight | +0.009 |
Neg logit contributions | |
pand | -0.017 |
pill | -0.016 |
hawk | -0.015 |
add | -0.015 |
otion | -0.015 |
Token | your |
Token ID | 534 |
Feature activation | -8.73e-01 |
Pos logit contributions | |
pand | +0.025 |
pill | +0.023 |
hawk | +0.022 |
add | +0.022 |
otion | +0.022 |
Neg logit contributions | |
forehead | -0.018 |
hand | -0.017 |
lap | -0.017 |
cheek | -0.017 |
bek | -0.016 |
Token | bug |
Token ID | 5434 |
Feature activation | -4.07e+00 |
Pos logit contributions | |
pand | +0.038 |
pill | +0.035 |
hawk | +0.035 |
add | +0.034 |
otion | +0.034 |
Neg logit contributions | |
forehead | -0.025 |
hand | -0.023 |
cheek | -0.022 |
lap | -0.022 |
bek | -0.022 |
Token | go |
Token ID | 467 |
Feature activation | -8.85e-01 |
Pos logit contributions | |
pand | +0.181 |
pill | +0.171 |
add | +0.167 |
hawk | +0.167 |
otion | +0.167 |
Neg logit contributions | |
forehead | -0.110 |
hand | -0.108 |
lap | -0.101 |
cheek | -0.101 |
tight | -0.098 |
Token | , |
Token ID | 11 |
Feature activation | +8.78e-01 |
Pos logit contributions | |
pand | +0.044 |
pill | +0.040 |
hawk | +0.040 |
add | +0.040 |
otion | +0.039 |
Neg logit contributions | |
forehead | -0.020 |
hand | -0.018 |
lap | -0.018 |
cheek | -0.018 |
bek | -0.018 |
Token | John |
Token ID | 1757 |
Feature activation | +5.07e-01 |
Pos logit contributions | |
forehead | +0.027 |
lap | +0.025 |
hand | +0.025 |
cheek | +0.025 |
bek | +0.024 |
Neg logit contributions | |
pand | -0.036 |
pill | -0.033 |
otion | -0.032 |
hawk | -0.032 |
add | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | -2.96e-01 |
Pos logit contributions | |
forehead | +0.013 |
hand | +0.012 |
lap | +0.012 |
cheek | +0.012 |
tight | +0.011 |
Neg logit contributions | |
pand | -0.023 |
pill | -0.022 |
hawk | -0.021 |
otion | -0.021 |
add | -0.021 |
Token | You |
Token ID | 921 |
Feature activation | +8.18e-01 |
Pos logit contributions | |
pand | +0.011 |
pill | +0.010 |
add | +0.010 |
otion | +0.010 |
hawk | +0.010 |
Neg logit contributions | |
forehead | -0.010 |
hand | -0.009 |
cheek | -0.009 |
lap | -0.009 |
whole | -0.009 |
Token | and |
Token ID | 290 |
Feature activation | +3.54e-01 |
Pos logit contributions | |
pand | +0.001 |
pill | +0.001 |
hawk | +0.001 |
add | +0.001 |
otion | +0.001 |
Neg logit contributions | |
forehead | -0.000 |
hand | -0.000 |
lap | -0.000 |
cheek | -0.000 |
tight | -0.000 |
Token | white |
Token ID | 2330 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
forehead | +0.010 |
hand | +0.010 |
cheek | +0.010 |
lap | +0.009 |
bek | +0.009 |
Neg logit contributions | |
pand | -0.015 |
pill | -0.014 |
add | -0.014 |
hawk | -0.014 |
otion | -0.013 |
Token | cloth |
Token ID | 16270 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
otion | +0.050 |
pand | +0.046 |
hawk | +0.045 |
storms | +0.044 |
die | +0.044 |
Neg logit contributions | |
hand | -0.046 |
hands | -0.045 |
forehead | -0.042 |
lap | -0.042 |
tight | -0.041 |
Token | and |
Token ID | 290 |
Feature activation | +4.11e-02 |
Pos logit contributions | |
pand | +0.050 |
pill | +0.047 |
hawk | +0.047 |
add | +0.047 |
otion | +0.047 |
Neg logit contributions | |
forehead | -0.023 |
hand | -0.022 |
tight | -0.021 |
cheek | -0.021 |
lap | -0.021 |
Token | carefully |
Token ID | 7773 |
Feature activation | -8.81e-01 |
Pos logit contributions | |
forehead | +0.001 |
cheek | +0.001 |
hand | +0.001 |
lap | +0.001 |
bek | +0.001 |
Neg logit contributions | |
pand | -0.002 |
pill | -0.002 |
add | -0.002 |
hawk | -0.002 |
storms | -0.002 |
Token | se |
Token ID | 384 |
Feature activation | -4.06e+00 |
Pos logit contributions | |
pand | +0.041 |
hawk | +0.039 |
pill | +0.039 |
otion | +0.038 |
storms | +0.038 |
Neg logit contributions | |
hand | -0.022 |
forehead | -0.021 |
tight | -0.020 |
lap | -0.020 |
cheek | -0.019 |
Token | wed |
Token ID | 19103 |
Feature activation | +9.04e-01 |
Pos logit contributions | |
pand | +0.185 |
hawk | +0.173 |
pill | +0.170 |
otion | +0.169 |
storms | +0.164 |
Neg logit contributions | |
forehead | -0.105 |
hand | -0.101 |
tight | -0.098 |
lap | -0.098 |
bek | -0.097 |
Token | them |
Token ID | 606 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
hand | +0.032 |
forehead | +0.032 |
tight | +0.032 |
lap | +0.030 |
whole | +0.030 |
Neg logit contributions | |
pand | -0.032 |
pill | -0.029 |
hawk | -0.029 |
otion | -0.029 |
storms | -0.028 |
Token | together |
Token ID | 1978 |
Feature activation | +6.50e-01 |
Pos logit contributions | |
forehead | +0.003 |
hand | +0.003 |
cheek | +0.003 |
lap | +0.003 |
bek | +0.003 |
Neg logit contributions | |
pand | -0.005 |
add | -0.005 |
pill | -0.005 |
storms | -0.005 |
otion | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | +8.39e-01 |
Pos logit contributions | |
forehead | +0.016 |
bek | +0.015 |
cheek | +0.015 |
hand | +0.014 |
lap | +0.014 |
Neg logit contributions | |
pand | -0.030 |
pill | -0.028 |
storms | -0.027 |
add | -0.027 |
otion | -0.027 |
Token |
|
Token ID | 198 |
Feature activation | +8.06e-01 |
Pos logit contributions | |
forehead | +0.030 |
bek | +0.029 |
cheek | +0.028 |
entire | +0.027 |
whole | +0.026 |
Neg logit contributions | |
pand | -0.030 |
pill | -0.028 |
otion | -0.027 |
Hipp | -0.027 |
add | -0.027 |
Token | fish |
Token ID | 5916 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
pand | +0.006 |
otion | +0.006 |
pill | +0.006 |
hawk | +0.006 |
storms | +0.006 |
Neg logit contributions | |
hand | -0.003 |
forehead | -0.002 |
tight | -0.002 |
whole | -0.002 |
plate | -0.002 |
Token | in |
Token ID | 287 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
pand | +0.047 |
pill | +0.046 |
add | +0.046 |
storms | +0.046 |
hawk | +0.045 |
Neg logit contributions | |
hand | -0.024 |
tight | -0.022 |
whole | -0.021 |
plate | -0.020 |
full | -0.020 |
Token | the |
Token ID | 262 |
Feature activation | +1.59e-01 |
Pos logit contributions | |
forehead | +0.004 |
cheek | +0.004 |
hand | +0.003 |
lap | +0.003 |
bek | +0.003 |
Neg logit contributions | |
pand | -0.007 |
pill | -0.007 |
hawk | -0.007 |
add | -0.006 |
otion | -0.006 |
Token | ocean |
Token ID | 9151 |
Feature activation | +4.74e-01 |
Pos logit contributions | |
forehead | +0.006 |
bek | +0.006 |
ae | +0.006 |
cheek | +0.006 |
lap | +0.006 |
Neg logit contributions | |
pand | -0.005 |
hawk | -0.005 |
helicopter | -0.004 |
add | -0.004 |
aliens | -0.004 |
Token | and |
Token ID | 290 |
Feature activation | +3.19e-01 |
Pos logit contributions | |
forehead | +0.013 |
bek | +0.012 |
hand | +0.012 |
cheek | +0.012 |
lap | +0.012 |
Neg logit contributions | |
pand | -0.021 |
pill | -0.019 |
otion | -0.019 |
storms | -0.019 |
hawk | -0.019 |
Token | se |
Token ID | 384 |
Feature activation | -4.02e+00 |
Pos logit contributions | |
forehead | +0.009 |
hand | +0.009 |
cheek | +0.009 |
lap | +0.009 |
tight | +0.008 |
Neg logit contributions | |
pand | -0.013 |
pill | -0.012 |
hawk | -0.012 |
add | -0.012 |
otion | -0.012 |
Token | ag |
Token ID | 363 |
Feature activation | +2.16e-02 |
Pos logit contributions | |
pand | +0.169 |
hawk | +0.158 |
pill | +0.156 |
otion | +0.152 |
storms | +0.149 |
Neg logit contributions | |
forehead | -0.118 |
hand | -0.114 |
bek | -0.111 |
lap | -0.111 |
tight | -0.110 |
Token | ull |
Token ID | 724 |
Feature activation | -2.06e+00 |
Pos logit contributions | |
forehead | +0.001 |
hand | +0.001 |
cheek | +0.001 |
lap | +0.001 |
tight | +0.001 |
Neg logit contributions | |
pand | -0.000 |
pill | -0.000 |
add | -0.000 |
otion | -0.000 |
hawk | -0.000 |
Token | s |
Token ID | 82 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
pand | +0.093 |
otion | +0.089 |
pill | +0.088 |
hawk | +0.088 |
add | +0.087 |
Neg logit contributions | |
forehead | -0.052 |
hand | -0.051 |
tight | -0.049 |
cheek | -0.049 |
lap | -0.048 |
Token | swoop |
Token ID | 38527 |
Feature activation | +7.93e-01 |
Pos logit contributions | |
pand | +0.075 |
pill | +0.069 |
hawk | +0.068 |
add | +0.068 |
otion | +0.067 |
Neg logit contributions | |
forehead | -0.046 |
hand | -0.043 |
cheek | -0.042 |
lap | -0.042 |
bek | -0.041 |
Token | ing |
Token ID | 278 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
forehead | +0.022 |
lap | +0.021 |
hand | +0.021 |
cheek | +0.020 |
bek | +0.020 |
Neg logit contributions | |
pand | -0.034 |
pill | -0.032 |
add | -0.032 |
hawk | -0.031 |
otion | -0.031 |
Token | sea |
Token ID | 5417 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
forehead | +0.024 |
hand | +0.023 |
lap | +0.022 |
cheek | +0.022 |
tight | +0.022 |
Neg logit contributions | |
pand | -0.024 |
pill | -0.022 |
hawk | -0.021 |
add | -0.021 |
otion | -0.021 |
Token | , |
Token ID | 11 |
Feature activation | +8.26e-01 |
Pos logit contributions | |
forehead | +0.003 |
hand | +0.003 |
tight | +0.003 |
lap | +0.003 |
cheek | +0.003 |
Neg logit contributions | |
pand | -0.006 |
add | -0.006 |
pill | -0.006 |
hawk | -0.006 |
otion | -0.006 |
Token | sandy |
Token ID | 44039 |
Feature activation | -7.95e-01 |
Pos logit contributions | |
forehead | +0.021 |
hand | +0.020 |
tight | +0.019 |
lap | +0.019 |
cheek | +0.019 |
Neg logit contributions | |
pand | -0.038 |
add | -0.036 |
pill | -0.036 |
otion | -0.036 |
hawk | -0.036 |
Token | beach |
Token ID | 10481 |
Feature activation | +1.82e-02 |
Pos logit contributions | |
pand | +0.028 |
hawk | +0.024 |
pill | +0.024 |
add | +0.024 |
storms | +0.024 |
Neg logit contributions | |
forehead | -0.030 |
hand | -0.028 |
bek | -0.028 |
lap | -0.028 |
cheek | -0.027 |
Token | and |
Token ID | 290 |
Feature activation | +2.62e-01 |
Pos logit contributions | |
forehead | +0.001 |
hand | +0.000 |
bek | +0.000 |
lap | +0.000 |
cheek | +0.000 |
Neg logit contributions | |
pand | -0.001 |
pill | -0.001 |
storms | -0.001 |
hawk | -0.001 |
otion | -0.001 |
Token | se |
Token ID | 384 |
Feature activation | -3.99e+00 |
Pos logit contributions | |
forehead | +0.008 |
cheek | +0.008 |
hand | +0.008 |
lap | +0.007 |
bek | +0.007 |
Neg logit contributions | |
pand | -0.011 |
pill | -0.010 |
hawk | -0.010 |
otion | -0.009 |
add | -0.009 |
Token | ag |
Token ID | 363 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
pand | +0.164 |
hawk | +0.151 |
pill | +0.150 |
otion | +0.147 |
storms | +0.144 |
Neg logit contributions | |
forehead | -0.123 |
hand | -0.117 |
lap | -0.114 |
bek | -0.113 |
cheek | -0.113 |
Token | ull |
Token ID | 724 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
pand | +0.003 |
add | +0.003 |
otion | +0.003 |
pill | +0.002 |
hawk | +0.002 |
Neg logit contributions | |
forehead | -0.007 |
cheek | -0.007 |
lap | -0.007 |
hand | -0.007 |
list | -0.007 |
Token | s |
Token ID | 82 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
pand | +0.101 |
pill | +0.093 |
hawk | +0.092 |
add | +0.092 |
otion | +0.092 |
Neg logit contributions | |
forehead | -0.069 |
hand | -0.066 |
cheek | -0.064 |
lap | -0.064 |
tight | -0.062 |
Token | in |
Token ID | 287 |
Feature activation | -2.62e-01 |
Pos logit contributions | |
pand | +0.077 |
pill | +0.071 |
hawk | +0.071 |
add | +0.070 |
otion | +0.070 |
Neg logit contributions | |
forehead | -0.042 |
hand | -0.040 |
lap | -0.038 |
cheek | -0.038 |
bek | -0.038 |
Token | the |
Token ID | 262 |
Feature activation | -3.16e-03 |
Pos logit contributions | |
pand | +0.012 |
pill | +0.012 |
hawk | +0.011 |
add | +0.011 |
otion | +0.011 |
Neg logit contributions | |
forehead | -0.007 |
hand | -0.006 |
cheek | -0.006 |
lap | -0.006 |
bek | -0.006 |
Token | jumping |
Token ID | 14284 |
Feature activation | +2.33e-01 |
Pos logit contributions | |
add | +0.004 |
pand | +0.004 |
pill | +0.004 |
hawk | +0.004 |
storms | +0.004 |
Neg logit contributions | |
hand | -0.002 |
tight | -0.002 |
full | -0.002 |
forehead | -0.002 |
between | -0.002 |
Token | in |
Token ID | 287 |
Feature activation | +4.28e-01 |
Pos logit contributions | |
tight | +0.004 |
hand | +0.004 |
whole | +0.004 |
forehead | +0.004 |
cheek | +0.003 |
Neg logit contributions | |
pand | -0.013 |
otion | -0.012 |
pill | -0.012 |
add | -0.012 |
hawk | -0.012 |
Token | the |
Token ID | 262 |
Feature activation | +1.84e-01 |
Pos logit contributions | |
hand | +0.009 |
forehead | +0.009 |
tight | +0.008 |
lap | +0.008 |
cheek | +0.008 |
Neg logit contributions | |
pand | -0.021 |
pill | -0.020 |
otion | -0.020 |
add | -0.020 |
hawk | -0.020 |
Token | water |
Token ID | 1660 |
Feature activation | +9.14e-01 |
Pos logit contributions | |
forehead | +0.007 |
bek | +0.006 |
hand | +0.006 |
cheek | +0.006 |
lap | +0.006 |
Neg logit contributions | |
pand | -0.007 |
hawk | -0.006 |
pill | -0.006 |
add | -0.006 |
storms | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | +3.90e-01 |
Pos logit contributions | |
forehead | +0.022 |
hand | +0.021 |
lap | +0.020 |
cheek | +0.020 |
tight | +0.020 |
Neg logit contributions | |
pand | -0.043 |
pill | -0.040 |
add | -0.039 |
hawk | -0.039 |
otion | -0.039 |
Token | se |
Token ID | 384 |
Feature activation | -3.99e+00 |
Pos logit contributions | |
forehead | +0.011 |
hand | +0.011 |
lap | +0.011 |
tight | +0.011 |
cheek | +0.011 |
Neg logit contributions | |
pand | -0.016 |
add | -0.015 |
otion | -0.015 |
pill | -0.015 |
hawk | -0.015 |
Token | ag |
Token ID | 363 |
Feature activation | +2.89e-02 |
Pos logit contributions | |
pand | +0.167 |
hawk | +0.156 |
pill | +0.154 |
otion | +0.151 |
storms | +0.147 |
Neg logit contributions | |
forehead | -0.118 |
hand | -0.113 |
bek | -0.111 |
lap | -0.111 |
tight | -0.110 |
Token | ull |
Token ID | 724 |
Feature activation | -2.19e+00 |
Pos logit contributions | |
hand | +0.001 |
forehead | +0.001 |
whole | +0.001 |
lap | +0.001 |
tight | +0.001 |
Neg logit contributions | |
pand | -0.001 |
pill | -0.001 |
hawk | -0.001 |
add | -0.001 |
otion | -0.001 |
Token | s |
Token ID | 82 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
pand | +0.105 |
otion | +0.103 |
hawk | +0.100 |
add | +0.100 |
pill | +0.099 |
Neg logit contributions | |
hand | -0.048 |
tight | -0.048 |
forehead | -0.047 |
cheek | -0.046 |
whole | -0.045 |
Token | flying |
Token ID | 7348 |
Feature activation | -7.49e-01 |
Pos logit contributions | |
pand | +0.074 |
pill | +0.071 |
add | +0.069 |
otion | +0.068 |
hawk | +0.068 |
Neg logit contributions | |
forehead | -0.042 |
hand | -0.041 |
cheek | -0.039 |
lap | -0.038 |
tight | -0.038 |
Token | in |
Token ID | 287 |
Feature activation | +2.32e-01 |
Pos logit contributions | |
pand | +0.037 |
pill | +0.035 |
otion | +0.034 |
add | +0.034 |
hawk | +0.034 |
Neg logit contributions | |
forehead | -0.016 |
hand | -0.016 |
cheek | -0.015 |
lap | -0.015 |
tight | -0.015 |
Token | and |
Token ID | 290 |
Feature activation | +6.74e-02 |
Pos logit contributions | |
hand | +0.010 |
tight | +0.009 |
forehead | +0.009 |
lap | +0.009 |
whole | +0.009 |
Neg logit contributions | |
pand | -0.026 |
hawk | -0.025 |
add | -0.025 |
pill | -0.025 |
otion | -0.025 |
Token | circles |
Token ID | 13332 |
Feature activation | -8.96e-01 |
Pos logit contributions | |
forehead | +0.003 |
hand | +0.002 |
lap | +0.002 |
cheek | +0.002 |
tight | +0.002 |
Neg logit contributions | |
pand | -0.002 |
pill | -0.002 |
otion | -0.002 |
add | -0.002 |
hawk | -0.002 |
Token | and |
Token ID | 290 |
Feature activation | -2.45e-01 |
Pos logit contributions | |
pand | +0.046 |
hawk | +0.046 |
pill | +0.044 |
add | +0.044 |
ican | +0.043 |
Neg logit contributions | |
tight | -0.017 |
hand | -0.017 |
forehead | -0.015 |
whole | -0.015 |
lap | -0.015 |
Token | triangles |
Token ID | 44360 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
pand | +0.009 |
add | +0.008 |
otion | +0.008 |
hawk | +0.008 |
pill | +0.008 |
Neg logit contributions | |
forehead | -0.009 |
hand | -0.008 |
lap | -0.008 |
cheek | -0.008 |
tight | -0.008 |
Token | and |
Token ID | 290 |
Feature activation | -6.49e-01 |
Pos logit contributions | |
pand | +0.086 |
hawk | +0.082 |
pill | +0.081 |
add | +0.080 |
otion | +0.079 |
Neg logit contributions | |
hand | -0.031 |
forehead | -0.030 |
tight | -0.029 |
whole | -0.028 |
lap | -0.028 |
Token | rect |
Token ID | 13621 |
Feature activation | -3.98e+00 |
Pos logit contributions | |
pand | +0.024 |
add | +0.023 |
pill | +0.022 |
otion | +0.022 |
hawk | +0.022 |
Neg logit contributions | |
forehead | -0.022 |
hand | -0.022 |
lap | -0.021 |
cheek | -0.020 |
tight | -0.020 |
Token | angles |
Token ID | 27787 |
Feature activation | -8.13e-02 |
Pos logit contributions | |
pand | +0.150 |
hawk | +0.137 |
pill | +0.137 |
otion | +0.133 |
storms | +0.132 |
Neg logit contributions | |
forehead | -0.134 |
hand | -0.130 |
lap | -0.125 |
cheek | -0.124 |
tight | -0.124 |
Token | . |
Token ID | 13 |
Feature activation | +6.09e-02 |
Pos logit contributions | |
pand | +0.004 |
pill | +0.004 |
hawk | +0.004 |
add | +0.004 |
otion | +0.004 |
Neg logit contributions | |
hand | -0.002 |
forehead | -0.002 |
tight | -0.001 |
lap | -0.001 |
cheek | -0.001 |
Token | She |
Token ID | 1375 |
Feature activation | +7.79e-02 |
Pos logit contributions | |
hand | +0.002 |
forehead | +0.002 |
lap | +0.002 |
tight | +0.002 |
cheek | +0.002 |
Neg logit contributions | |
pand | -0.002 |
hawk | -0.002 |
pill | -0.002 |
otion | -0.002 |
add | -0.002 |
Token | made |
Token ID | 925 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
forehead | +0.002 |
hand | +0.002 |
lap | +0.002 |
cheek | +0.002 |
tight | +0.002 |
Neg logit contributions | |
pand | -0.004 |
pill | -0.003 |
hawk | -0.003 |
add | -0.003 |
otion | -0.003 |
Token | her |
Token ID | 607 |
Feature activation | +7.17e-01 |
Pos logit contributions | |
hand | +0.025 |
tight | +0.025 |
forehead | +0.025 |
whole | +0.024 |
lap | +0.023 |
Neg logit contributions | |
pand | -0.053 |
hawk | -0.053 |
pill | -0.051 |
otion | -0.051 |
storms | -0.050 |
Token | and |
Token ID | 290 |
Feature activation | -5.49e-01 |
Pos logit contributions | |
forehead | +0.003 |
bek | +0.002 |
cheek | +0.002 |
hand | +0.002 |
lap | +0.002 |
Neg logit contributions | |
pand | -0.004 |
pill | -0.004 |
storms | -0.004 |
hawk | -0.004 |
otion | -0.004 |
Token | learned |
Token ID | 4499 |
Feature activation | -3.46e-01 |
Pos logit contributions | |
pand | +0.020 |
Tweet | +0.018 |
pill | +0.017 |
die | +0.017 |
hawk | +0.017 |
Neg logit contributions | |
forehead | -0.018 |
cheek | -0.018 |
lap | -0.018 |
hand | -0.017 |
entire | -0.017 |
Token | something |
Token ID | 1223 |
Feature activation | -2.31e+00 |
Pos logit contributions | |
pand | +0.014 |
pill | +0.013 |
hawk | +0.012 |
add | +0.012 |
otion | +0.012 |
Neg logit contributions | |
forehead | -0.011 |
hand | -0.011 |
lap | -0.010 |
cheek | -0.010 |
tight | -0.010 |
Token | new |
Token ID | 649 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
pand | +0.105 |
pill | +0.099 |
hawk | +0.098 |
add | +0.098 |
otion | +0.097 |
Neg logit contributions | |
forehead | -0.059 |
hand | -0.056 |
tight | -0.054 |
cheek | -0.054 |
lap | -0.053 |
Token | about |
Token ID | 546 |
Feature activation | +1.20e-01 |
Pos logit contributions | |
pand | +0.062 |
storms | +0.056 |
pill | +0.056 |
otion | +0.055 |
aliens | +0.055 |
Neg logit contributions | |
forehead | -0.043 |
cheek | -0.039 |
bek | -0.038 |
hand | -0.038 |
lap | -0.038 |
Token | se |
Token ID | 384 |
Feature activation | -3.92e+00 |
Pos logit contributions | |
hand | +0.003 |
forehead | +0.003 |
tight | +0.003 |
lap | +0.003 |
cheek | +0.003 |
Neg logit contributions | |
pand | -0.005 |
add | -0.005 |
pill | -0.005 |
otion | -0.005 |
hawk | -0.005 |
Token | ag |
Token ID | 363 |
Feature activation | +2.20e-01 |
Pos logit contributions | |
pand | +0.126 |
pill | +0.113 |
hawk | +0.112 |
otion | +0.109 |
add | +0.108 |
Neg logit contributions | |
forehead | -0.156 |
hand | -0.149 |
cheek | -0.147 |
lap | -0.146 |
tight | -0.144 |
Token | ull |
Token ID | 724 |
Feature activation | -2.06e+00 |
Pos logit contributions | |
forehead | +0.009 |
hand | +0.009 |
lap | +0.009 |
cheek | +0.009 |
tight | +0.008 |
Neg logit contributions | |
pand | -0.007 |
pill | -0.006 |
otion | -0.006 |
add | -0.006 |
hawk | -0.006 |
Token | s |
Token ID | 82 |
Feature activation | -2.07e+00 |
Pos logit contributions | |
pand | +0.094 |
add | +0.092 |
otion | +0.091 |
hawk | +0.089 |
pill | +0.087 |
Neg logit contributions | |
forehead | -0.049 |
cheek | -0.048 |
hand | -0.048 |
tight | -0.047 |
whole | -0.046 |
Token | . |
Token ID | 13 |
Feature activation | +4.44e-01 |
Pos logit contributions | |
pand | +0.093 |
storms | +0.084 |
hawk | +0.084 |
otion | +0.084 |
pill | +0.084 |
Neg logit contributions | |
forehead | -0.052 |
bek | -0.049 |
lap | -0.047 |
hand | -0.047 |
cheek | -0.046 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -5.30e-01 |
Pos logit contributions | |
forehead | +0.013 |
cheek | +0.012 |
hand | +0.012 |
bek | +0.012 |
lap | +0.011 |
Neg logit contributions | |
pand | -0.019 |
pill | -0.018 |
add | -0.017 |
storms | -0.017 |
otion | -0.017 |
Token | cloth |
Token ID | 16270 |
Feature activation | +4.09e-01 |
Pos logit contributions | |
pand | +0.009 |
hawk | +0.008 |
pill | +0.008 |
add | +0.008 |
flashes | +0.008 |
Neg logit contributions | |
forehead | -0.008 |
bek | -0.008 |
lap | -0.007 |
hand | -0.007 |
cheek | -0.007 |
Token | around |
Token ID | 1088 |
Feature activation | -7.22e-01 |
Pos logit contributions | |
forehead | +0.012 |
bek | +0.011 |
lap | +0.011 |
hand | +0.011 |
cheek | +0.011 |
Neg logit contributions | |
pand | -0.017 |
storms | -0.016 |
otion | -0.015 |
pill | -0.015 |
hawk | -0.015 |
Token | the |
Token ID | 262 |
Feature activation | -2.93e-01 |
Pos logit contributions | |
pand | +0.028 |
add | +0.024 |
pill | +0.024 |
otion | +0.024 |
hawk | +0.024 |
Neg logit contributions | |
forehead | -0.025 |
lap | -0.023 |
cheek | -0.023 |
bek | -0.023 |
hand | -0.022 |
Token | jacket |
Token ID | 15224 |
Feature activation | +6.00e-01 |
Pos logit contributions | |
pand | +0.011 |
hawk | +0.010 |
pill | +0.010 |
add | +0.010 |
storms | +0.010 |
Neg logit contributions | |
forehead | -0.010 |
bek | -0.009 |
lap | -0.009 |
hand | -0.009 |
cheek | -0.009 |
Token | and |
Token ID | 290 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
forehead | +0.019 |
bek | +0.019 |
hand | +0.018 |
ossible | +0.017 |
lap | +0.017 |
Neg logit contributions | |
pand | -0.022 |
storms | -0.021 |
otion | -0.020 |
Tweet | -0.020 |
pill | -0.020 |
Token | se |
Token ID | 384 |
Feature activation | -3.91e+00 |
Pos logit contributions | |
forehead | +0.003 |
cheek | +0.003 |
lap | +0.003 |
hand | +0.003 |
bek | +0.002 |
Neg logit contributions | |
pand | -0.004 |
pill | -0.004 |
hawk | -0.004 |
otion | -0.004 |
add | -0.004 |
Token | wed |
Token ID | 19103 |
Feature activation | +7.50e-01 |
Pos logit contributions | |
pand | +0.170 |
pill | +0.157 |
hawk | +0.156 |
otion | +0.152 |
storms | +0.151 |
Neg logit contributions | |
forehead | -0.112 |
hand | -0.107 |
lap | -0.103 |
cheek | -0.103 |
tight | -0.103 |
Token | it |
Token ID | 340 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
forehead | +0.026 |
lap | +0.024 |
hand | +0.024 |
cheek | +0.024 |
bek | +0.024 |
Neg logit contributions | |
pand | -0.028 |
pill | -0.026 |
add | -0.025 |
otion | -0.025 |
hawk | -0.025 |
Token | up |
Token ID | 510 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
pand | +0.061 |
add | +0.057 |
pill | +0.056 |
otion | +0.056 |
storms | +0.056 |
Neg logit contributions | |
forehead | -0.032 |
hand | -0.030 |
lap | -0.029 |
cheek | -0.028 |
bek | -0.028 |
Token | to |
Token ID | 284 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
pand | +0.010 |
add | +0.009 |
Tweet | +0.009 |
storms | +0.009 |
hawk | +0.009 |
Neg logit contributions | |
forehead | -0.006 |
hand | -0.006 |
lap | -0.006 |
cheek | -0.006 |
bek | -0.005 |
Token | make |
Token ID | 787 |
Feature activation | +2.48e-01 |
Pos logit contributions | |
forehead | +0.003 |
bek | +0.003 |
lap | +0.003 |
ossible | +0.003 |
cheek | +0.003 |
Neg logit contributions | |
pand | -0.004 |
hawk | -0.004 |
die | -0.004 |
otion | -0.004 |
add | -0.003 |
Token | thread |
Token ID | 4704 |
Feature activation | -2.06e+00 |
Pos logit contributions | |
pand | +0.028 |
pill | +0.026 |
hawk | +0.026 |
Tweet | +0.025 |
flashes | +0.025 |
Neg logit contributions | |
forehead | -0.022 |
cheek | -0.021 |
lap | -0.021 |
bek | -0.020 |
hand | -0.020 |
Token | from |
Token ID | 422 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
pand | +0.100 |
pill | +0.093 |
hawk | +0.092 |
add | +0.091 |
otion | +0.090 |
Neg logit contributions | |
forehead | -0.048 |
hand | -0.047 |
tight | -0.045 |
cheek | -0.044 |
lap | -0.043 |
Token | her |
Token ID | 607 |
Feature activation | -9.77e-01 |
Pos logit contributions | |
pand | +0.066 |
hawk | +0.059 |
otion | +0.058 |
Tweet | +0.057 |
storms | +0.056 |
Neg logit contributions | |
forehead | -0.052 |
cheek | -0.049 |
lap | -0.049 |
bek | -0.048 |
hand | -0.046 |
Token | bag |
Token ID | 6131 |
Feature activation | -2.24e-01 |
Pos logit contributions | |
pand | +0.044 |
hawk | +0.039 |
pill | +0.038 |
aliens | +0.038 |
storms | +0.037 |
Neg logit contributions | |
bek | -0.025 |
forehead | -0.024 |
cheek | -0.022 |
ossible | -0.022 |
ae | -0.021 |
Token | and |
Token ID | 290 |
Feature activation | -3.98e-01 |
Pos logit contributions | |
pand | +0.008 |
otion | +0.008 |
Tweet | +0.008 |
storms | +0.008 |
add | +0.007 |
Neg logit contributions | |
bek | -0.007 |
forehead | -0.007 |
cheek | -0.006 |
ae | -0.006 |
lap | -0.006 |
Token | se |
Token ID | 384 |
Feature activation | -3.91e+00 |
Pos logit contributions | |
pand | +0.016 |
pill | +0.015 |
hawk | +0.015 |
otion | +0.014 |
storms | +0.014 |
Neg logit contributions | |
forehead | -0.012 |
lap | -0.011 |
cheek | -0.011 |
hand | -0.011 |
bek | -0.011 |
Token | wed |
Token ID | 19103 |
Feature activation | +4.50e-01 |
Pos logit contributions | |
pand | +0.163 |
pill | +0.149 |
hawk | +0.149 |
otion | +0.145 |
storms | +0.144 |
Neg logit contributions | |
forehead | -0.119 |
hand | -0.114 |
lap | -0.110 |
cheek | -0.110 |
tight | -0.108 |
Token | up |
Token ID | 510 |
Feature activation | -7.23e-02 |
Pos logit contributions | |
forehead | +0.015 |
lap | +0.014 |
cheek | +0.014 |
hand | +0.014 |
bek | +0.013 |
Neg logit contributions | |
pand | -0.018 |
pill | -0.016 |
add | -0.016 |
otion | -0.016 |
hawk | -0.016 |
Token | Lucy |
Token ID | 22162 |
Feature activation | +8.32e-01 |
Pos logit contributions | |
pand | +0.003 |
pill | +0.003 |
hawk | +0.003 |
add | +0.003 |
otion | +0.003 |
Neg logit contributions | |
forehead | -0.002 |
hand | -0.002 |
lap | -0.002 |
tight | -0.002 |
cheek | -0.002 |
Token | 's |
Token ID | 338 |
Feature activation | -4.00e-01 |
Pos logit contributions | |
forehead | +0.023 |
bek | +0.022 |
lap | +0.021 |
cheek | +0.021 |
list | +0.020 |
Neg logit contributions | |
pand | -0.035 |
pill | -0.032 |
storms | -0.031 |
otion | -0.031 |
die | -0.030 |
Token | dress |
Token ID | 6576 |
Feature activation | -2.01e-01 |
Pos logit contributions | |
pand | +0.014 |
hawk | +0.013 |
pill | +0.012 |
die | +0.012 |
add | +0.012 |
Neg logit contributions | |
bek | -0.014 |
ae | -0.013 |
ossible | -0.012 |
forehead | -0.012 |
lap | -0.012 |
Token | Ann |
Token ID | 18858 |
Feature activation | +9.76e-01 |
Pos logit contributions | |
forehead | +0.024 |
hand | +0.022 |
cheek | +0.022 |
lap | +0.022 |
plate | +0.021 |
Neg logit contributions | |
pand | -0.029 |
pill | -0.026 |
otion | -0.026 |
add | -0.026 |
hawk | -0.026 |
Token | kept |
Token ID | 4030 |
Feature activation | +3.99e-01 |
Pos logit contributions | |
forehead | +0.026 |
hand | +0.025 |
cheek | +0.024 |
lap | +0.024 |
tight | +0.023 |
Neg logit contributions | |
pand | -0.044 |
pill | -0.041 |
add | -0.040 |
hawk | -0.040 |
otion | -0.040 |
Token | playing |
Token ID | 2712 |
Feature activation | +8.77e-01 |
Pos logit contributions | |
forehead | +0.012 |
hand | +0.012 |
lap | +0.012 |
tight | +0.012 |
cheek | +0.011 |
Neg logit contributions | |
pand | -0.016 |
pill | -0.015 |
add | -0.015 |
hawk | -0.015 |
otion | -0.015 |
Token | with |
Token ID | 351 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
forehead | +0.021 |
hand | +0.020 |
cheek | +0.019 |
bek | +0.019 |
lap | +0.019 |
Neg logit contributions | |
pand | -0.041 |
otion | -0.038 |
pill | -0.038 |
hawk | -0.038 |
storms | -0.037 |
Token | the |
Token ID | 262 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
hand | +0.030 |
forehead | +0.029 |
tight | +0.028 |
lap | +0.027 |
plate | +0.026 |
Neg logit contributions | |
pand | -0.072 |
pill | -0.069 |
add | -0.068 |
hawk | -0.067 |
otion | -0.067 |
Token | tape |
Token ID | 9154 |
Feature activation | +2.77e+00 |
Pos logit contributions | |
bek | +0.006 |
forehead | +0.006 |
ae | +0.005 |
ossible | +0.005 |
lap | +0.005 |
Neg logit contributions | |
pand | -0.005 |
flashes | -0.004 |
aliens | -0.004 |
otion | -0.004 |
hawk | -0.004 |
Token | and |
Token ID | 290 |
Feature activation | +3.43e-01 |
Pos logit contributions | |
forehead | +0.071 |
hand | +0.066 |
cheek | +0.065 |
lap | +0.064 |
bek | +0.064 |
Neg logit contributions | |
pand | -0.127 |
pill | -0.116 |
otion | -0.116 |
storms | -0.115 |
hawk | -0.115 |
Token | got |
Token ID | 1392 |
Feature activation | +3.77e-01 |
Pos logit contributions | |
forehead | +0.011 |
cheek | +0.010 |
hand | +0.010 |
lap | +0.010 |
bek | +0.009 |
Neg logit contributions | |
pand | -0.014 |
pill | -0.013 |
otion | -0.012 |
hawk | -0.012 |
storms | -0.012 |
Token | lost |
Token ID | 2626 |
Feature activation | +4.09e-01 |
Pos logit contributions | |
forehead | +0.009 |
hand | +0.009 |
tight | +0.008 |
cheek | +0.008 |
lap | +0.008 |
Neg logit contributions | |
pand | -0.018 |
pill | -0.017 |
hawk | -0.017 |
otion | -0.017 |
add | -0.017 |
Token | in |
Token ID | 287 |
Feature activation | +6.59e-01 |
Pos logit contributions | |
forehead | +0.008 |
hand | +0.008 |
lap | +0.008 |
cheek | +0.008 |
tight | +0.007 |
Neg logit contributions | |
pand | -0.021 |
pill | -0.020 |
hawk | -0.019 |
add | -0.019 |
otion | -0.019 |
Token | her |
Token ID | 607 |
Feature activation | -8.06e-02 |
Pos logit contributions | |
forehead | +0.018 |
cheek | +0.017 |
lap | +0.016 |
hand | +0.016 |
bek | +0.016 |
Neg logit contributions | |
pand | -0.030 |
pill | -0.027 |
hawk | -0.027 |
otion | -0.027 |
storms | -0.026 |
Token | . |
Token ID | 13 |
Feature activation | -3.34e-01 |
Pos logit contributions | |
pand | +0.026 |
otion | +0.024 |
pill | +0.024 |
add | +0.024 |
storms | +0.023 |
Neg logit contributions | |
forehead | -0.012 |
hand | -0.011 |
lap | -0.011 |
bek | -0.011 |
cheek | -0.011 |
Token | They |
Token ID | 1119 |
Feature activation | -2.38e-01 |
Pos logit contributions | |
pand | +0.014 |
pill | +0.013 |
hawk | +0.013 |
add | +0.013 |
otion | +0.013 |
Neg logit contributions | |
hand | -0.010 |
forehead | -0.010 |
lap | -0.010 |
tight | -0.009 |
cheek | -0.009 |
Token | take |
Token ID | 1011 |
Feature activation | +5.13e-01 |
Pos logit contributions | |
pand | +0.010 |
pill | +0.009 |
hawk | +0.009 |
add | +0.009 |
otion | +0.009 |
Neg logit contributions | |
forehead | -0.007 |
hand | -0.007 |
lap | -0.007 |
cheek | -0.006 |
bek | -0.006 |
Token | away |
Token ID | 1497 |
Feature activation | -8.13e-03 |
Pos logit contributions | |
forehead | +0.015 |
hand | +0.015 |
lap | +0.014 |
cheek | +0.014 |
tight | +0.014 |
Neg logit contributions | |
pand | -0.022 |
pill | -0.020 |
hawk | -0.020 |
add | -0.019 |
otion | -0.019 |
Token | their |
Token ID | 511 |
Feature activation | +2.12e-02 |
Pos logit contributions | |
pand | +0.000 |
pill | +0.000 |
add | +0.000 |
flashes | +0.000 |
otion | +0.000 |
Neg logit contributions | |
forehead | -0.000 |
cheek | -0.000 |
lap | -0.000 |
bek | -0.000 |
hand | -0.000 |
Token | brushes |
Token ID | 36377 |
Feature activation | +2.52e+00 |
Pos logit contributions | |
forehead | +0.001 |
bek | +0.001 |
lap | +0.001 |
ae | +0.001 |
nie | +0.001 |
Neg logit contributions | |
pand | -0.001 |
die | -0.001 |
pill | -0.001 |
add | -0.001 |
conflicts | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | -7.92e-01 |
Pos logit contributions | |
forehead | +0.069 |
bek | +0.064 |
cheek | +0.062 |
lap | +0.062 |
hand | +0.062 |
Neg logit contributions | |
pand | -0.110 |
otion | -0.100 |
pill | -0.100 |
hawk | -0.100 |
storms | -0.099 |
Token | colors |
Token ID | 7577 |
Feature activation | -3.70e-01 |
Pos logit contributions | |
pand | +0.030 |
die | +0.027 |
flashes | +0.027 |
spr | +0.026 |
pill | +0.026 |
Neg logit contributions | |
forehead | -0.026 |
bek | -0.026 |
cheek | -0.025 |
whole | -0.024 |
entire | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | -2.84e-01 |
Pos logit contributions | |
pand | +0.020 |
pill | +0.019 |
hawk | +0.018 |
add | +0.018 |
otion | +0.018 |
Neg logit contributions | |
forehead | -0.007 |
hand | -0.007 |
tight | -0.006 |
cheek | -0.006 |
lap | -0.006 |
Token | They |
Token ID | 1119 |
Feature activation | -2.59e-01 |
Pos logit contributions | |
pand | +0.012 |
hawk | +0.011 |
pill | +0.011 |
add | +0.011 |
otion | +0.011 |
Neg logit contributions | |
hand | -0.008 |
forehead | -0.008 |
lap | -0.008 |
tight | -0.008 |
cheek | -0.008 |
Token | send |
Token ID | 3758 |
Feature activation | +8.81e-01 |
Pos logit contributions | |
pand | +0.011 |
pill | +0.010 |
add | +0.010 |
hawk | +0.010 |
otion | +0.010 |
Neg logit contributions | |
forehead | -0.008 |
hand | -0.007 |
lap | -0.007 |
cheek | -0.007 |
bek | -0.007 |
Token | saw |
Token ID | 2497 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
forehead | +0.019 |
hand | +0.018 |
cheek | +0.018 |
lap | +0.018 |
tight | +0.017 |
Neg logit contributions | |
pand | -0.035 |
pill | -0.032 |
hawk | -0.032 |
add | -0.032 |
otion | -0.031 |
Token | that |
Token ID | 326 |
Feature activation | +3.81e-01 |
Pos logit contributions | |
hand | +0.002 |
tight | +0.002 |
forehead | +0.002 |
hands | +0.002 |
front | +0.002 |
Neg logit contributions | |
pand | -0.006 |
add | -0.006 |
pill | -0.005 |
storms | -0.005 |
hawk | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | +3.48e-01 |
Pos logit contributions | |
hand | +0.009 |
forehead | +0.009 |
tight | +0.009 |
lap | +0.008 |
whole | +0.008 |
Neg logit contributions | |
pand | -0.018 |
add | -0.017 |
pill | -0.017 |
hawk | -0.016 |
otion | -0.016 |
Token | tower |
Token ID | 10580 |
Feature activation | -5.68e-02 |
Pos logit contributions | |
bek | +0.014 |
forehead | +0.013 |
ossible | +0.012 |
lap | +0.012 |
cle | +0.012 |
Neg logit contributions | |
pand | -0.012 |
flashes | -0.011 |
hawk | -0.011 |
helicopter | -0.010 |
aliens | -0.010 |
Token | was |
Token ID | 373 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
pand | +0.003 |
add | +0.003 |
pill | +0.003 |
hawk | +0.003 |
otion | +0.002 |
Neg logit contributions | |
forehead | -0.001 |
hand | -0.001 |
tight | -0.001 |
lap | -0.001 |
cheek | -0.001 |
Token | just |
Token ID | 655 |
Feature activation | +2.66e+00 |
Pos logit contributions | |
forehead | +0.005 |
cheek | +0.004 |
lap | +0.004 |
hand | +0.004 |
bek | +0.004 |
Neg logit contributions | |
pand | -0.006 |
hawk | -0.005 |
add | -0.005 |
pill | -0.005 |
otion | -0.005 |
Token | an |
Token ID | 281 |
Feature activation | -5.22e-01 |
Pos logit contributions | |
forehead | +0.070 |
hand | +0.065 |
lap | +0.065 |
cheek | +0.064 |
bek | +0.063 |
Neg logit contributions | |
pand | -0.122 |
pill | -0.111 |
hawk | -0.110 |
add | -0.109 |
otion | -0.109 |
Token | ordinary |
Token ID | 8850 |
Feature activation | +6.98e-01 |
Pos logit contributions | |
pand | +0.019 |
hawk | +0.017 |
add | +0.017 |
pill | +0.017 |
otion | +0.017 |
Neg logit contributions | |
forehead | -0.019 |
hand | -0.018 |
lap | -0.018 |
cheek | -0.018 |
bek | -0.017 |
Token | building |
Token ID | 2615 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
forehead | +0.023 |
bek | +0.022 |
cheek | +0.021 |
lap | +0.021 |
ossible | +0.021 |
Neg logit contributions | |
pand | -0.027 |
hawk | -0.024 |
flashes | -0.023 |
storms | -0.023 |
otion | -0.023 |
Token | . |
Token ID | 13 |
Feature activation | +2.70e-01 |
Pos logit contributions | |
forehead | +0.049 |
cheek | +0.044 |
hand | +0.044 |
lap | +0.043 |
bek | +0.043 |
Neg logit contributions | |
pand | -0.091 |
otion | -0.083 |
pill | -0.083 |
storms | -0.082 |
hawk | -0.081 |
Token | It |
Token ID | 632 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
forehead | +0.009 |
hand | +0.009 |
lap | +0.008 |
tight | +0.008 |
cheek | +0.008 |
Neg logit contributions | |
pand | -0.010 |
hawk | -0.010 |
add | -0.009 |
otion | -0.009 |
pill | -0.009 |
Token | smiled |
Token ID | 13541 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
forehead | +0.017 |
hand | +0.016 |
lap | +0.016 |
cheek | +0.015 |
bek | +0.015 |
Neg logit contributions | |
pand | -0.026 |
pill | -0.024 |
hawk | -0.023 |
storms | -0.023 |
add | -0.023 |
Token | . |
Token ID | 13 |
Feature activation | +1.86e-01 |
Pos logit contributions | |
forehead | +0.032 |
hand | +0.029 |
lap | +0.029 |
bek | +0.029 |
cheek | +0.028 |
Neg logit contributions | |
pand | -0.064 |
hawk | -0.058 |
add | -0.058 |
storms | -0.058 |
pill | -0.058 |
Token | He |
Token ID | 679 |
Feature activation | +4.03e-01 |
Pos logit contributions | |
forehead | +0.006 |
cheek | +0.006 |
hand | +0.006 |
bek | +0.006 |
lap | +0.006 |
Neg logit contributions | |
pand | -0.007 |
pill | -0.007 |
add | -0.006 |
otion | -0.006 |
storms | -0.006 |
Token | found |
Token ID | 1043 |
Feature activation | +3.73e-01 |
Pos logit contributions | |
forehead | +0.012 |
cheek | +0.011 |
lap | +0.011 |
hand | +0.011 |
list | +0.011 |
Neg logit contributions | |
pand | -0.016 |
pill | -0.015 |
add | -0.015 |
hawk | -0.015 |
storms | -0.015 |
Token | a |
Token ID | 257 |
Feature activation | +5.63e-01 |
Pos logit contributions | |
forehead | +0.011 |
cheek | +0.010 |
lap | +0.010 |
bek | +0.010 |
hand | +0.009 |
Neg logit contributions | |
pand | -0.016 |
pill | -0.015 |
hawk | -0.015 |
otion | -0.015 |
add | -0.015 |
Token | tape |
Token ID | 9154 |
Feature activation | +2.66e+00 |
Pos logit contributions | |
forehead | +0.017 |
bek | +0.016 |
ossible | +0.016 |
lap | +0.015 |
cheek | +0.015 |
Neg logit contributions | |
pand | -0.021 |
hawk | -0.021 |
flashes | -0.019 |
pill | -0.019 |
aliens | -0.019 |
Token | and |
Token ID | 290 |
Feature activation | +3.33e-01 |
Pos logit contributions | |
forehead | +0.076 |
bek | +0.071 |
lap | +0.070 |
hand | +0.069 |
cheek | +0.069 |
Neg logit contributions | |
pand | -0.110 |
hawk | -0.104 |
otion | -0.101 |
storms | -0.100 |
pill | -0.098 |
Token | put |
Token ID | 1234 |
Feature activation | +4.24e-01 |
Pos logit contributions | |
forehead | +0.010 |
cheek | +0.010 |
lap | +0.009 |
bek | +0.009 |
list | +0.009 |
Neg logit contributions | |
pand | -0.013 |
pill | -0.012 |
flashes | -0.012 |
hawk | -0.012 |
aliens | -0.012 |
Token | it |
Token ID | 340 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
forehead | +0.017 |
bek | +0.017 |
lap | +0.016 |
cheek | +0.016 |
list | +0.016 |
Neg logit contributions | |
pand | -0.012 |
flashes | -0.011 |
hawk | -0.011 |
otion | -0.011 |
Tweet | -0.010 |
Token | into |
Token ID | 656 |
Feature activation | -2.82e-01 |
Pos logit contributions | |
pand | +0.053 |
add | +0.049 |
pill | +0.048 |
otion | +0.048 |
storms | +0.048 |
Neg logit contributions | |
forehead | -0.026 |
hand | -0.024 |
bek | -0.023 |
lap | -0.023 |
cheek | -0.023 |
Token | the |
Token ID | 262 |
Feature activation | -7.55e-02 |
Pos logit contributions | |
pand | +0.011 |
hawk | +0.010 |
aliens | +0.010 |
Tweet | +0.010 |
flashes | +0.010 |
Neg logit contributions | |
forehead | -0.009 |
cheek | -0.008 |
bek | -0.008 |
lap | -0.008 |
ossible | -0.008 |
Token | My |
Token ID | 3666 |
Feature activation | -1.03e-01 |
Pos logit contributions | |
pand | +0.013 |
add | +0.012 |
otion | +0.012 |
hawk | +0.012 |
storms | +0.012 |
Neg logit contributions | |
hand | -0.008 |
forehead | -0.008 |
lap | -0.008 |
whole | -0.008 |
cheek | -0.008 |
Token | favorite |
Token ID | 4004 |
Feature activation | +8.44e-01 |
Pos logit contributions | |
pand | +0.004 |
pill | +0.004 |
hawk | +0.004 |
Hipp | +0.004 |
storms | +0.004 |
Neg logit contributions | |
bek | -0.003 |
forehead | -0.003 |
lap | -0.003 |
cheek | -0.003 |
hand | -0.003 |
Token | food |
Token ID | 2057 |
Feature activation | -4.03e-02 |
Pos logit contributions | |
forehead | +0.026 |
bek | +0.024 |
hand | +0.024 |
cheek | +0.024 |
lap | +0.024 |
Neg logit contributions | |
pand | -0.035 |
pill | -0.031 |
hawk | -0.031 |
otion | -0.030 |
add | -0.030 |
Token | is |
Token ID | 318 |
Feature activation | -3.01e-01 |
Pos logit contributions | |
pand | +0.002 |
pill | +0.001 |
otion | +0.001 |
hawk | +0.001 |
storms | +0.001 |
Neg logit contributions | |
forehead | -0.001 |
bek | -0.001 |
hand | -0.001 |
lap | -0.001 |
cheek | -0.001 |
Token | ice |
Token ID | 4771 |
Feature activation | -9.01e-02 |
Pos logit contributions | |
pand | +0.014 |
pill | +0.013 |
add | +0.012 |
hawk | +0.012 |
otion | +0.012 |
Neg logit contributions | |
forehead | -0.008 |
hand | -0.008 |
lap | -0.007 |
cheek | -0.007 |
tight | -0.007 |
Token | cream |
Token ID | 8566 |
Feature activation | +3.30e+00 |
Pos logit contributions | |
pand | +0.004 |
pill | +0.004 |
Hipp | +0.003 |
storms | +0.003 |
hawk | +0.003 |
Neg logit contributions | |
forehead | -0.003 |
bek | -0.002 |
hand | -0.002 |
lap | -0.002 |
ossible | -0.002 |
Token | ." |
Token ID | 526 |
Feature activation | +6.44e-01 |
Pos logit contributions | |
forehead | +0.094 |
bek | +0.091 |
lap | +0.087 |
hand | +0.086 |
ossible | +0.082 |
Neg logit contributions | |
pand | -0.138 |
storms | -0.126 |
Hipp | -0.123 |
pill | -0.123 |
otion | -0.122 |
Token |
|
Token ID | 198 |
Feature activation | -8.30e-02 |
Pos logit contributions | |
forehead | +0.015 |
hand | +0.014 |
cheek | +0.014 |
lap | +0.014 |
tight | +0.013 |
Neg logit contributions | |
pand | -0.031 |
pill | -0.029 |
hawk | -0.029 |
add | -0.029 |
otion | -0.029 |
Token |
|
Token ID | 198 |
Feature activation | -3.88e-02 |
Pos logit contributions | |
pand | +0.004 |
pill | +0.004 |
hawk | +0.004 |
add | +0.004 |
otion | +0.004 |
Neg logit contributions | |
forehead | -0.002 |
hand | -0.002 |
lap | -0.002 |
cheek | -0.002 |
tight | -0.002 |
Token | Finally |
Token ID | 11158 |
Feature activation | +1.85e-01 |
Pos logit contributions | |
pand | +0.002 |
otion | +0.002 |
add | +0.002 |
pill | +0.002 |
storms | +0.002 |
Neg logit contributions | |
forehead | -0.001 |
hand | -0.001 |
lap | -0.001 |
tight | -0.001 |
cheek | -0.001 |
Token | , |
Token ID | 11 |
Feature activation | +4.17e-02 |
Pos logit contributions | |
forehead | +0.005 |
hand | +0.005 |
tight | +0.005 |
lap | +0.004 |
cheek | +0.004 |
Neg logit contributions | |
pand | -0.008 |
pill | -0.008 |
add | -0.008 |
hawk | -0.008 |
otion | -0.008 |
Token | sack |
Token ID | 23704 |
Feature activation | +4.40e-01 |
Pos logit contributions | |
forehead | +0.018 |
hand | +0.017 |
lap | +0.017 |
cheek | +0.017 |
bek | +0.017 |
Neg logit contributions | |
pand | -0.024 |
hawk | -0.022 |
pill | -0.022 |
add | -0.021 |
otion | -0.021 |
Token | behind |
Token ID | 2157 |
Feature activation | +1.85e-01 |
Pos logit contributions | |
forehead | +0.013 |
bek | +0.011 |
cheek | +0.011 |
lap | +0.011 |
hand | +0.011 |
Neg logit contributions | |
pand | -0.019 |
hawk | -0.017 |
otion | -0.017 |
pill | -0.017 |
storms | -0.017 |
Token | her |
Token ID | 607 |
Feature activation | -4.55e-01 |
Pos logit contributions | |
hand | +0.004 |
forehead | +0.004 |
lap | +0.004 |
tight | +0.004 |
cheek | +0.004 |
Neg logit contributions | |
pand | -0.009 |
add | -0.009 |
pill | -0.009 |
hawk | -0.009 |
otion | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | +4.21e-01 |
Pos logit contributions | |
pand | +0.022 |
pill | +0.020 |
hawk | +0.020 |
storms | +0.019 |
add | +0.019 |
Neg logit contributions | |
forehead | -0.011 |
bek | -0.010 |
cheek | -0.010 |
lap | -0.010 |
hand | -0.010 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -5.20e-02 |
Pos logit contributions | |
forehead | +0.013 |
hand | +0.013 |
lap | +0.012 |
cheek | +0.012 |
tight | +0.012 |
Neg logit contributions | |
pand | -0.017 |
pill | -0.016 |
hawk | -0.015 |
add | -0.015 |
otion | -0.015 |
Token | Billy |
Token ID | 49640 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
pand | +0.002 |
pill | +0.002 |
add | +0.002 |
hawk | +0.002 |
otion | +0.002 |
Neg logit contributions | |
forehead | -0.002 |
hand | -0.001 |
cheek | -0.001 |
lap | -0.001 |
bek | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | -5.36e-01 |
Pos logit contributions | |
forehead | +0.007 |
hand | +0.007 |
lap | +0.007 |
tight | +0.007 |
cheek | +0.007 |
Neg logit contributions | |
pand | -0.014 |
pill | -0.013 |
add | -0.013 |
hawk | -0.013 |
otion | -0.013 |
Token | sad |
Token ID | 6507 |
Feature activation | -8.35e-01 |
Pos logit contributions | |
pand | +0.023 |
pill | +0.021 |
add | +0.021 |
otion | +0.021 |
hawk | +0.021 |
Neg logit contributions | |
forehead | -0.015 |
hand | -0.014 |
cheek | -0.014 |
lap | -0.014 |
bek | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -8.78e-01 |
Pos logit contributions | |
pand | +0.039 |
pill | +0.036 |
hawk | +0.036 |
otion | +0.036 |
add | +0.036 |
Neg logit contributions | |
forehead | -0.020 |
hand | -0.019 |
lap | -0.018 |
cheek | -0.018 |
bek | -0.018 |
Token | No |
Token ID | 1400 |
Feature activation | -3.95e-01 |
Pos logit contributions | |
pand | +0.040 |
pill | +0.038 |
hawk | +0.037 |
add | +0.036 |
otion | +0.036 |
Neg logit contributions | |
hand | -0.023 |
forehead | -0.023 |
lap | -0.022 |
tight | -0.021 |
cheek | -0.021 |
Token | matter |
Token ID | 2300 |
Feature activation | -6.35e-01 |
Pos logit contributions | |
pill | +0.018 |
otion | +0.018 |
add | +0.018 |
pand | +0.018 |
storms | +0.018 |
Neg logit contributions | |
hand | -0.009 |
tight | -0.008 |
whole | -0.008 |
forehead | -0.008 |
hands | -0.008 |
Token | doll |
Token ID | 3654 |
Feature activation | +2.61e-01 |
Pos logit contributions | |
forehead | +0.017 |
bek | +0.016 |
hand | +0.016 |
lap | +0.016 |
cheek | +0.015 |
Neg logit contributions | |
pand | -0.029 |
pill | -0.026 |
hawk | -0.026 |
add | -0.025 |
otion | -0.025 |
Token | ." |
Token ID | 526 |
Feature activation | +4.81e-01 |
Pos logit contributions | |
forehead | +0.007 |
hand | +0.006 |
bek | +0.006 |
lap | +0.006 |
cheek | +0.006 |
Neg logit contributions | |
pand | -0.012 |
pill | -0.011 |
hawk | -0.011 |
storms | -0.011 |
otion | -0.011 |
Token |
|
Token ID | 198 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
forehead | +0.013 |
hand | +0.012 |
cheek | +0.011 |
bek | +0.011 |
lap | +0.011 |
Neg logit contributions | |
pand | -0.022 |
pill | -0.021 |
add | -0.020 |
otion | -0.020 |
hawk | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
pand | +0.006 |
pill | +0.006 |
add | +0.006 |
hawk | +0.006 |
otion | +0.006 |
Neg logit contributions | |
forehead | -0.004 |
hand | -0.003 |
cheek | -0.003 |
lap | -0.003 |
bek | -0.003 |
Token | He |
Token ID | 1544 |
Feature activation | +8.36e-03 |
Pos logit contributions | |
pand | +0.011 |
pill | +0.011 |
hawk | +0.010 |
add | +0.010 |
otion | +0.010 |
Neg logit contributions | |
forehead | -0.008 |
hand | -0.008 |
cheek | -0.008 |
lap | -0.007 |
plate | -0.007 |
Token | took |
Token ID | 1718 |
Feature activation | +5.75e-01 |
Pos logit contributions | |
forehead | +0.000 |
hand | +0.000 |
cheek | +0.000 |
lap | +0.000 |
tight | +0.000 |
Neg logit contributions | |
pand | -0.000 |
pill | -0.000 |
add | -0.000 |
hawk | -0.000 |
otion | -0.000 |
Token | Lily |
Token ID | 20037 |
Feature activation | +8.96e-01 |
Pos logit contributions | |
forehead | +0.016 |
hand | +0.015 |
lap | +0.015 |
cheek | +0.015 |
tight | +0.014 |
Neg logit contributions | |
pand | -0.025 |
pill | -0.024 |
hawk | -0.023 |
add | -0.023 |
otion | -0.023 |
Token | and |
Token ID | 290 |
Feature activation | -5.31e-01 |
Pos logit contributions | |
forehead | +0.026 |
bek | +0.024 |
cheek | +0.023 |
lap | +0.022 |
satisfied | +0.022 |
Neg logit contributions | |
pand | -0.038 |
otion | -0.034 |
pill | -0.034 |
storms | -0.033 |
hawk | -0.033 |
Token | Rose |
Token ID | 8049 |
Feature activation | -8.34e-02 |
Pos logit contributions | |
pand | +0.022 |
pill | +0.020 |
hawk | +0.020 |
add | +0.019 |
otion | +0.019 |
Neg logit contributions | |
forehead | -0.016 |
cheek | -0.015 |
bek | -0.015 |
lap | -0.015 |
whole | -0.015 |
Token | to |
Token ID | 284 |
Feature activation | -7.57e-01 |
Pos logit contributions | |
pand | +0.004 |
otion | +0.003 |
pill | +0.003 |
storms | +0.003 |
hawk | +0.003 |
Neg logit contributions | |
forehead | -0.002 |
bek | -0.002 |
cheek | -0.002 |
lap | -0.002 |
hand | -0.002 |
Token | his |
Token ID | 465 |
Feature activation | -4.03e-01 |
Pos logit contributions | |
pand | +0.034 |
pill | +0.033 |
hawk | +0.032 |
add | +0.032 |
otion | +0.032 |
Neg logit contributions | |
forehead | -0.020 |
hand | -0.019 |
lap | -0.018 |
cheek | -0.018 |
tight | -0.018 |
Token | look |
Token ID | 804 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
hand | +0.034 |
forehead | +0.033 |
tight | +0.032 |
cheek | +0.031 |
lap | +0.031 |
Neg logit contributions | |
pand | -0.044 |
pill | -0.042 |
add | -0.041 |
hawk | -0.041 |
storms | -0.040 |
Token | the |
Token ID | 262 |
Feature activation | +7.61e-01 |
Pos logit contributions | |
pand | +0.048 |
pill | +0.045 |
hawk | +0.044 |
add | +0.044 |
otion | +0.044 |
Neg logit contributions | |
forehead | -0.030 |
hand | -0.029 |
tight | -0.028 |
cheek | -0.028 |
lap | -0.028 |
Token | same |
Token ID | 976 |
Feature activation | +7.05e-01 |
Pos logit contributions | |
forehead | +0.025 |
bek | +0.025 |
lap | +0.023 |
cheek | +0.023 |
ossible | +0.022 |
Neg logit contributions | |
pand | -0.029 |
hawk | -0.026 |
aliens | -0.026 |
pill | -0.026 |
flashes | -0.025 |
Token | when |
Token ID | 618 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
forehead | +0.019 |
bek | +0.018 |
cheek | +0.017 |
lap | +0.017 |
hand | +0.017 |
Neg logit contributions | |
pand | -0.031 |
pill | -0.028 |
hawk | -0.028 |
otion | -0.028 |
storms | -0.027 |
Token | you |
Token ID | 345 |
Feature activation | +3.30e-01 |
Pos logit contributions | |
forehead | +0.003 |
hand | +0.003 |
cheek | +0.003 |
lap | +0.003 |
tight | +0.003 |
Neg logit contributions | |
pand | -0.005 |
pill | -0.005 |
add | -0.005 |
hawk | -0.004 |
otion | -0.004 |
Token | turn |
Token ID | 1210 |
Feature activation | +2.78e-01 |
Pos logit contributions | |
forehead | +0.010 |
hand | +0.009 |
lap | +0.009 |
cheek | +0.009 |
tight | +0.009 |
Neg logit contributions | |
pand | -0.014 |
pill | -0.013 |
hawk | -0.013 |
add | -0.013 |
otion | -0.013 |
Token | them |
Token ID | 606 |
Feature activation | +8.12e-01 |
Pos logit contributions | |
forehead | +0.007 |
hand | +0.006 |
cheek | +0.006 |
lap | +0.006 |
tight | +0.006 |
Neg logit contributions | |
pand | -0.013 |
pill | -0.012 |
hawk | -0.012 |
add | -0.012 |
otion | -0.012 |
Token | sideways |
Token ID | 35303 |
Feature activation | -1.89e-02 |
Pos logit contributions | |
forehead | +0.018 |
hand | +0.017 |
lap | +0.016 |
cheek | +0.016 |
bek | +0.016 |
Neg logit contributions | |
pand | -0.040 |
pill | -0.037 |
add | -0.037 |
hawk | -0.037 |
otion | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | -4.51e-01 |
Pos logit contributions | |
pand | +0.001 |
pill | +0.001 |
hawk | +0.001 |
add | +0.001 |
otion | +0.001 |
Neg logit contributions | |
forehead | -0.000 |
hand | -0.000 |
tight | -0.000 |
cheek | -0.000 |
lap | -0.000 |
Token | Like |
Token ID | 4525 |
Feature activation | +4.97e-01 |
Pos logit contributions | |
pand | +0.018 |
hawk | +0.017 |
add | +0.017 |
pill | +0.017 |
otion | +0.016 |
Neg logit contributions | |
hand | -0.014 |
forehead | -0.013 |
lap | -0.013 |
tight | -0.013 |
cheek | -0.013 |
Token | three |
Token ID | 1115 |
Feature activation | -8.76e-02 |
Pos logit contributions | |
hand | +0.013 |
forehead | +0.013 |
tight | +0.013 |
lap | +0.012 |
plate | +0.012 |
Neg logit contributions | |
pand | -0.021 |
add | -0.021 |
pill | -0.020 |
hawk | -0.020 |
otion | -0.020 |
Token | judge |
Token ID | 5052 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
forehead | +0.005 |
bek | +0.005 |
lap | +0.005 |
hand | +0.005 |
tight | +0.005 |
Neg logit contributions | |
pand | -0.005 |
hawk | -0.005 |
pill | -0.004 |
aliens | -0.004 |
add | -0.004 |
Token | announced |
Token ID | 3414 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
forehead | +0.031 |
hand | +0.029 |
lap | +0.028 |
cheek | +0.028 |
tight | +0.027 |
Neg logit contributions | |
pand | -0.058 |
pill | -0.054 |
hawk | -0.054 |
add | -0.054 |
otion | -0.053 |
Token | that |
Token ID | 326 |
Feature activation | +3.52e-01 |
Pos logit contributions | |
forehead | +0.052 |
hand | +0.050 |
cheek | +0.048 |
lap | +0.048 |
tight | +0.048 |
Neg logit contributions | |
pand | -0.074 |
pill | -0.069 |
hawk | -0.068 |
add | -0.068 |
otion | -0.068 |
Token | the |
Token ID | 262 |
Feature activation | +2.22e-01 |
Pos logit contributions | |
hand | +0.010 |
forehead | +0.010 |
tight | +0.010 |
cheek | +0.010 |
lap | +0.010 |
Neg logit contributions | |
pand | -0.014 |
add | -0.013 |
hawk | -0.013 |
pill | -0.013 |
otion | -0.013 |
Token | princess |
Token ID | 21752 |
Feature activation | +9.29e-01 |
Pos logit contributions | |
forehead | +0.007 |
lap | +0.007 |
hand | +0.007 |
bek | +0.007 |
cheek | +0.007 |
Neg logit contributions | |
pand | -0.008 |
hawk | -0.008 |
pill | -0.008 |
add | -0.007 |
otion | -0.007 |
Token | had |
Token ID | 550 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
forehead | +0.023 |
hand | +0.022 |
lap | +0.021 |
cheek | +0.021 |
tight | +0.021 |
Neg logit contributions | |
pand | -0.043 |
pill | -0.040 |
add | -0.040 |
hawk | -0.040 |
otion | -0.039 |
Token | won |
Token ID | 1839 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
forehead | +0.045 |
cheek | +0.041 |
lap | +0.041 |
hand | +0.040 |
bek | +0.040 |
Neg logit contributions | |
pand | -0.066 |
pill | -0.061 |
add | -0.059 |
hawk | -0.058 |
storms | -0.058 |
Token | the |
Token ID | 262 |
Feature activation | +2.77e-01 |
Pos logit contributions | |
forehead | +0.024 |
hand | +0.024 |
tight | +0.022 |
whole | +0.022 |
lap | +0.022 |
Neg logit contributions | |
pand | -0.057 |
pill | -0.054 |
hawk | -0.054 |
add | -0.053 |
otion | -0.052 |
Token | contest |
Token ID | 8414 |
Feature activation | +6.99e-01 |
Pos logit contributions | |
forehead | +0.011 |
hand | +0.010 |
lap | +0.010 |
cheek | +0.010 |
bek | +0.010 |
Neg logit contributions | |
pand | -0.009 |
pill | -0.008 |
hawk | -0.008 |
add | -0.008 |
otion | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | +7.88e-01 |
Pos logit contributions | |
forehead | +0.018 |
hand | +0.016 |
cheek | +0.016 |
lap | +0.015 |
bek | +0.015 |
Neg logit contributions | |
pand | -0.033 |
pill | -0.031 |
storms | -0.030 |
otion | -0.029 |
hawk | -0.029 |
Token | He |
Token ID | 679 |
Feature activation | +7.32e-01 |
Pos logit contributions | |
forehead | +0.025 |
hand | +0.024 |
cheek | +0.023 |
bek | +0.023 |
lap | +0.022 |
Neg logit contributions | |
pand | -0.032 |
pill | -0.030 |
add | -0.028 |
otion | -0.028 |
storms | -0.028 |
Token | gave |
Token ID | 2921 |
Feature activation | +7.13e-01 |
Pos logit contributions | |
forehead | +0.021 |
lap | +0.020 |
cheek | +0.020 |
hand | +0.020 |
whole | +0.019 |
Neg logit contributions | |
pand | -0.033 |
pill | -0.030 |
hawk | -0.030 |
otion | -0.030 |
add | -0.029 |
Token | it |
Token ID | 340 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
forehead | +0.030 |
bek | +0.028 |
cheek | +0.028 |
lap | +0.028 |
list | +0.027 |
Neg logit contributions | |
pand | -0.022 |
hawk | -0.019 |
Tweet | -0.019 |
add | -0.019 |
pill | -0.018 |
Token | a |
Token ID | 257 |
Feature activation | +3.18e-01 |
Pos logit contributions | |
pand | +0.063 |
add | +0.055 |
Tweet | +0.054 |
pill | +0.053 |
aliens | +0.052 |
Neg logit contributions | |
forehead | -0.042 |
bek | -0.041 |
cheek | -0.038 |
lap | -0.037 |
pron | -0.036 |
Token | turn |
Token ID | 1210 |
Feature activation | +7.91e-01 |
Pos logit contributions | |
forehead | +0.010 |
bek | +0.009 |
lap | +0.009 |
ae | +0.009 |
ossible | +0.008 |
Neg logit contributions | |
pand | -0.012 |
hawk | -0.011 |
die | -0.011 |
pill | -0.011 |
spr | -0.010 |
Token | ! |
Token ID | 0 |
Feature activation | +5.86e-01 |
Pos logit contributions | |
forehead | +0.020 |
bek | +0.018 |
lap | +0.017 |
cheek | +0.017 |
hand | +0.017 |
Neg logit contributions | |
pand | -0.037 |
pill | -0.034 |
otion | -0.033 |
hawk | -0.033 |
add | -0.033 |
Token | He |
Token ID | 679 |
Feature activation | +5.91e-01 |
Pos logit contributions | |
hand | +0.021 |
tight | +0.020 |
lap | +0.020 |
forehead | +0.020 |
hands | +0.019 |
Neg logit contributions | |
pand | -0.021 |
hawk | -0.019 |
otion | -0.019 |
add | -0.018 |
pill | -0.018 |
Token | was |
Token ID | 373 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
forehead | +0.015 |
hand | +0.014 |
cheek | +0.014 |
lap | +0.014 |
tight | +0.013 |
Neg logit contributions | |
pand | -0.027 |
pill | -0.026 |
add | -0.025 |
hawk | -0.025 |
otion | -0.025 |
Token | very |
Token ID | 845 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
forehead | +0.003 |
cheek | +0.003 |
lap | +0.003 |
bek | +0.003 |
hand | +0.003 |
Neg logit contributions | |
pand | -0.005 |
otion | -0.005 |
add | -0.005 |
pill | -0.005 |
hawk | -0.005 |
Token | surprised |
Token ID | 6655 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
forehead | +0.033 |
lap | +0.031 |
bek | +0.030 |
cheek | +0.029 |
hand | +0.029 |
Neg logit contributions | |
pand | -0.046 |
add | -0.043 |
otion | -0.041 |
die | -0.040 |
hawk | -0.040 |
Token | at |
Token ID | 379 |
Feature activation | +4.31e-01 |
Pos logit contributions | |
pand | +0.012 |
pill | +0.011 |
storms | +0.011 |
add | +0.011 |
hawk | +0.011 |
Neg logit contributions | |
forehead | -0.006 |
lap | -0.005 |
hand | -0.005 |
cheek | -0.005 |
bek | -0.005 |
Token | what |
Token ID | 644 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
forehead | +0.012 |
hand | +0.012 |
lap | +0.011 |
cheek | +0.011 |
tight | +0.011 |
Neg logit contributions | |
pand | -0.019 |
pill | -0.018 |
hawk | -0.017 |
add | -0.017 |
otion | -0.017 |
Token | dog |
Token ID | 3290 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
hand | +0.028 |
forehead | +0.027 |
tight | +0.026 |
cheek | +0.026 |
plate | +0.026 |
Neg logit contributions | |
pand | -0.032 |
pill | -0.032 |
otion | -0.032 |
add | -0.031 |
storms | -0.031 |
Token | named |
Token ID | 3706 |
Feature activation | +2.77e-01 |
Pos logit contributions | |
pand | +0.005 |
add | +0.005 |
pill | +0.005 |
otion | +0.005 |
hawk | +0.005 |
Neg logit contributions | |
hand | -0.005 |
forehead | -0.005 |
lap | -0.005 |
tight | -0.005 |
cheek | -0.004 |
Token | Leo |
Token ID | 19632 |
Feature activation | -4.77e-01 |
Pos logit contributions | |
forehead | +0.009 |
hand | +0.009 |
lap | +0.009 |
cheek | +0.009 |
tight | +0.009 |
Neg logit contributions | |
pand | -0.010 |
pill | -0.009 |
hawk | -0.009 |
add | -0.009 |
otion | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | -7.51e-01 |
Pos logit contributions | |
pand | +0.026 |
pill | +0.025 |
otion | +0.024 |
add | +0.024 |
hawk | +0.024 |
Neg logit contributions | |
forehead | -0.008 |
hand | -0.008 |
tight | -0.007 |
lap | -0.007 |
cheek | -0.007 |
Token | Leo |
Token ID | 19632 |
Feature activation | -3.59e-01 |
Pos logit contributions | |
pand | +0.025 |
pill | +0.023 |
hawk | +0.022 |
storms | +0.022 |
add | +0.022 |
Neg logit contributions | |
forehead | -0.029 |
hand | -0.027 |
cheek | -0.027 |
lap | -0.026 |
bek | -0.026 |
Token | was |
Token ID | 373 |
Feature activation | -3.86e-01 |
Pos logit contributions | |
pand | +0.018 |
otion | +0.017 |
pill | +0.017 |
add | +0.017 |
hawk | +0.017 |
Neg logit contributions | |
hand | -0.007 |
tight | -0.007 |
forehead | -0.007 |
whole | -0.007 |
lap | -0.007 |
Token | very |
Token ID | 845 |
Feature activation | +7.14e-01 |
Pos logit contributions | |
pill | +0.021 |
otion | +0.021 |
pand | +0.021 |
storms | +0.021 |
ican | +0.020 |
Neg logit contributions | |
tight | -0.007 |
hand | -0.007 |
whole | -0.006 |
full | -0.006 |
forehead | -0.006 |
Token | excited |
Token ID | 6568 |
Feature activation | +7.91e-01 |
Pos logit contributions | |
forehead | +0.017 |
hand | +0.016 |
tight | +0.016 |
cheek | +0.016 |
lap | +0.015 |
Neg logit contributions | |
pand | -0.034 |
pill | -0.032 |
hawk | -0.032 |
otion | -0.032 |
storms | -0.031 |
Token | to |
Token ID | 284 |
Feature activation | +2.03e-02 |
Pos logit contributions | |
bek | +0.029 |
forehead | +0.028 |
cheek | +0.026 |
lap | +0.026 |
hand | +0.025 |
Neg logit contributions | |
pand | -0.029 |
storms | -0.027 |
pill | -0.024 |
otion | -0.024 |
Tweet | -0.023 |
Token | go |
Token ID | 467 |
Feature activation | -4.19e-01 |
Pos logit contributions | |
forehead | +0.001 |
hand | +0.001 |
cheek | +0.001 |
lap | +0.001 |
tight | +0.001 |
Neg logit contributions | |
pand | -0.001 |
pill | -0.001 |
hawk | -0.001 |
add | -0.001 |
otion | -0.001 |
Token | on |
Token ID | 319 |
Feature activation | +1.98e-02 |
Pos logit contributions | |
pand | +0.021 |
add | +0.019 |
pill | +0.018 |
hawk | +0.018 |
otion | +0.018 |
Neg logit contributions | |
forehead | -0.010 |
bek | -0.009 |
cheek | -0.009 |
hand | -0.009 |
lap | -0.008 |
Token | He |
Token ID | 679 |
Feature activation | +2.57e-01 |
Pos logit contributions | |
pand | +0.014 |
pill | +0.013 |
hawk | +0.013 |
add | +0.013 |
otion | +0.012 |
Neg logit contributions | |
forehead | -0.010 |
hand | -0.010 |
lap | -0.010 |
cheek | -0.010 |
tight | -0.009 |
Token | tried |
Token ID | 3088 |
Feature activation | +6.46e-01 |
Pos logit contributions | |
forehead | +0.007 |
hand | +0.006 |
lap | +0.006 |
cheek | +0.006 |
tight | +0.006 |
Neg logit contributions | |
pand | -0.012 |
pill | -0.011 |
add | -0.011 |
hawk | -0.011 |
otion | -0.010 |
Token | to |
Token ID | 284 |
Feature activation | +7.41e-01 |
Pos logit contributions | |
forehead | +0.012 |
hand | +0.011 |
lap | +0.010 |
cheek | +0.010 |
tight | +0.010 |
Neg logit contributions | |
pand | -0.035 |
pill | -0.033 |
hawk | -0.033 |
add | -0.032 |
otion | -0.032 |
Token | grab |
Token ID | 5552 |
Feature activation | +8.23e-01 |
Pos logit contributions | |
forehead | +0.022 |
bek | +0.022 |
lap | +0.020 |
hand | +0.020 |
cheek | +0.020 |
Neg logit contributions | |
pand | -0.030 |
hawk | -0.027 |
add | -0.027 |
otion | -0.027 |
die | -0.026 |
Token | it |
Token ID | 340 |
Feature activation | -8.62e-01 |
Pos logit contributions | |
forehead | +0.027 |
bek | +0.025 |
cheek | +0.025 |
lap | +0.024 |
list | +0.023 |
Neg logit contributions | |
pand | -0.033 |
hawk | -0.029 |
add | -0.029 |
pill | -0.029 |
otion | -0.029 |
Token | from |
Token ID | 422 |
Feature activation | -1.99e-01 |
Pos logit contributions | |
pand | +0.042 |
pill | +0.040 |
hawk | +0.039 |
add | +0.039 |
otion | +0.039 |
Neg logit contributions | |
forehead | -0.020 |
hand | -0.020 |
tight | -0.019 |
cheek | -0.018 |
lap | -0.018 |
Token | Lily |
Token ID | 20037 |
Feature activation | +4.43e-01 |
Pos logit contributions | |
pand | +0.008 |
hawk | +0.007 |
Tweet | +0.007 |
add | +0.007 |
Hipp | +0.007 |
Neg logit contributions | |
forehead | -0.007 |
bek | -0.007 |
lap | -0.006 |
cheek | -0.006 |
list | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | -3.97e-01 |
Pos logit contributions | |
forehead | +0.012 |
bek | +0.011 |
cheek | +0.011 |
lap | +0.011 |
whole | +0.010 |
Neg logit contributions | |
pand | -0.020 |
pill | -0.018 |
storms | -0.018 |
otion | -0.018 |
hawk | -0.017 |
Token | But |
Token ID | 887 |
Feature activation | +1.88e-01 |
Pos logit contributions | |
pand | +0.017 |
hawk | +0.016 |
pill | +0.016 |
add | +0.016 |
otion | +0.016 |
Neg logit contributions | |
forehead | -0.011 |
hand | -0.011 |
lap | -0.010 |
tight | -0.010 |
cheek | -0.010 |
Token | Lily |
Token ID | 20037 |
Feature activation | -1.82e-01 |
Pos logit contributions | |
forehead | +0.006 |
hand | +0.006 |
lap | +0.006 |
cheek | +0.006 |
bek | +0.006 |
Neg logit contributions | |
pand | -0.007 |
otion | -0.007 |
pill | -0.007 |
hawk | -0.007 |
add | -0.007 |
Token | did |
Token ID | 750 |
Feature activation | +3.51e-01 |
Pos logit contributions | |
pand | +0.008 |
pill | +0.008 |
add | +0.008 |
otion | +0.008 |
hawk | +0.008 |
Neg logit contributions | |
forehead | -0.005 |
hand | -0.005 |
tight | -0.004 |
lap | -0.004 |
cheek | -0.004 |
Token | Tom |
Token ID | 13787 |
Feature activation | +1.80e-01 |
Pos logit contributions | |
pand | +0.024 |
hawk | +0.022 |
storms | +0.022 |
add | +0.022 |
pill | +0.022 |
Neg logit contributions | |
hand | -0.014 |
whole | -0.014 |
lap | -0.013 |
ossible | -0.012 |
front | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | -7.00e-01 |
Pos logit contributions | |
forehead | +0.005 |
cheek | +0.005 |
lap | +0.005 |
bek | +0.004 |
hand | +0.004 |
Neg logit contributions | |
pand | -0.008 |
pill | -0.007 |
otion | -0.007 |
storms | -0.007 |
hawk | -0.007 |
Token | Lily |
Token ID | 20037 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
pand | +0.029 |
hawk | +0.026 |
add | +0.025 |
otion | +0.025 |
pill | +0.025 |
Neg logit contributions | |
forehead | -0.022 |
cheek | -0.020 |
lap | -0.020 |
bek | -0.020 |
hand | -0.019 |
Token | are |
Token ID | 389 |
Feature activation | -7.37e-01 |
Pos logit contributions | |
pand | +0.062 |
pill | +0.059 |
add | +0.058 |
hawk | +0.058 |
otion | +0.057 |
Neg logit contributions | |
hand | -0.031 |
forehead | -0.030 |
tight | -0.028 |
whole | -0.028 |
lap | -0.028 |
Token | friends |
Token ID | 2460 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
pand | +0.027 |
add | +0.024 |
otion | +0.023 |
pill | +0.022 |
hawk | +0.021 |
Neg logit contributions | |
forehead | -0.028 |
bek | -0.026 |
lap | -0.026 |
cheek | -0.026 |
hand | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | -2.07e+00 |
Pos logit contributions | |
forehead | +0.039 |
bek | +0.038 |
list | +0.035 |
cheek | +0.034 |
hand | +0.033 |
Neg logit contributions | |
pand | -0.056 |
otion | -0.050 |
storms | -0.050 |
hawk | -0.050 |
pill | -0.049 |
Token | They |
Token ID | 1119 |
Feature activation | -5.21e-01 |
Pos logit contributions | |
pand | +0.074 |
pill | +0.068 |
hawk | +0.067 |
add | +0.066 |
otion | +0.066 |
Neg logit contributions | |
forehead | -0.073 |
hand | -0.072 |
lap | -0.070 |
cheek | -0.069 |
tight | -0.069 |
Token | like |
Token ID | 588 |
Feature activation | -5.14e-01 |
Pos logit contributions | |
pand | +0.020 |
pill | +0.019 |
add | +0.018 |
hawk | +0.018 |
otion | +0.018 |
Neg logit contributions | |
forehead | -0.017 |
hand | -0.016 |
lap | -0.016 |
cheek | -0.016 |
tight | -0.015 |
Token | to |
Token ID | 284 |
Feature activation | -6.18e-01 |
Pos logit contributions | |
pand | +0.030 |
otion | +0.029 |
hawk | +0.029 |
add | +0.029 |
pill | +0.029 |
Neg logit contributions | |
hand | -0.006 |
forehead | -0.006 |
lap | -0.005 |
whole | -0.005 |
tight | -0.005 |
Token | play |
Token ID | 711 |
Feature activation | -9.45e-01 |
Pos logit contributions | |
pill | +0.031 |
pand | +0.031 |
hawk | +0.030 |
otion | +0.029 |
storms | +0.029 |
Neg logit contributions | |
hand | -0.014 |
lap | -0.013 |
forehead | -0.013 |
whole | -0.012 |
ossible | -0.012 |
Token | chess |
Token ID | 19780 |
Feature activation | +2.82e-01 |
Pos logit contributions | |
pand | +0.050 |
pill | +0.048 |
otion | +0.047 |
hawk | +0.047 |
storms | +0.046 |
Neg logit contributions | |
hand | -0.018 |
whole | -0.016 |
forehead | -0.016 |
tight | -0.016 |
lap | -0.016 |
Token | " |
Token ID | 366 |
Feature activation | -5.38e-01 |
Pos logit contributions | |
pand | +0.018 |
pill | +0.017 |
add | +0.016 |
hawk | +0.016 |
otion | +0.016 |
Neg logit contributions | |
forehead | -0.010 |
hand | -0.010 |
lap | -0.010 |
cheek | -0.010 |
tight | -0.009 |
Token | Someone |
Token ID | 28211 |
Feature activation | -7.03e-01 |
Pos logit contributions | |
pand | +0.022 |
pill | +0.020 |
add | +0.020 |
hawk | +0.020 |
otion | +0.020 |
Neg logit contributions | |
forehead | -0.017 |
hand | -0.016 |
cheek | -0.015 |
lap | -0.015 |
tight | -0.015 |
Token | , |
Token ID | 11 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
pand | +0.028 |
pill | +0.025 |
hawk | +0.025 |
otion | +0.025 |
add | +0.025 |
Neg logit contributions | |
forehead | -0.023 |
hand | -0.021 |
lap | -0.021 |
cheek | -0.021 |
tight | -0.020 |
Token | help |
Token ID | 1037 |
Feature activation | -3.67e-01 |
Pos logit contributions | |
pand | +0.014 |
hawk | +0.013 |
otion | +0.012 |
add | +0.012 |
pill | +0.012 |
Neg logit contributions | |
forehead | -0.010 |
lap | -0.009 |
cheek | -0.009 |
hand | -0.009 |
bek | -0.009 |
Token | us |
Token ID | 514 |
Feature activation | -3.71e-02 |
Pos logit contributions | |
pand | +0.014 |
hawk | +0.013 |
otion | +0.013 |
add | +0.012 |
storms | +0.012 |
Neg logit contributions | |
forehead | -0.012 |
bek | -0.012 |
lap | -0.012 |
cheek | -0.011 |
satisfied | -0.011 |
Token | !" |
Token ID | 2474 |
Feature activation | -5.89e-01 |
Pos logit contributions | |
pand | +0.002 |
hawk | +0.002 |
otion | +0.002 |
add | +0.002 |
pill | +0.002 |
Neg logit contributions | |
forehead | -0.001 |
hand | -0.001 |
lap | -0.001 |
cheek | -0.001 |
bek | -0.001 |
Token |
|
Token ID | 198 |
Feature activation | -2.43e-01 |
Pos logit contributions | |
pand | +0.028 |
pill | +0.026 |
hawk | +0.026 |
add | +0.026 |
otion | +0.026 |
Neg logit contributions | |
forehead | -0.014 |
hand | -0.013 |
lap | -0.013 |
cheek | -0.013 |
tight | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
pand | +0.012 |
pill | +0.011 |
hawk | +0.011 |
add | +0.011 |
otion | +0.011 |
Neg logit contributions | |
forehead | -0.006 |
hand | -0.006 |
cheek | -0.005 |
lap | -0.005 |
bek | -0.005 |
Token | Their |
Token ID | 14574 |
Feature activation | -7.00e-01 |
Pos logit contributions | |
pand | +0.011 |
hawk | +0.011 |
pill | +0.010 |
add | +0.010 |
otion | +0.010 |
Neg logit contributions | |
forehead | -0.007 |
hand | -0.007 |
cheek | -0.007 |
lap | -0.007 |
plate | -0.006 |
Token | father |
Token ID | 2988 |
Feature activation | +2.32e-01 |
Pos logit contributions | |
pill | +0.041 |
pand | +0.041 |
otion | +0.041 |
storms | +0.039 |
add | +0.039 |
Neg logit contributions | |
hand | -0.009 |
forehead | -0.009 |
tight | -0.008 |
cheek | -0.008 |
hands | -0.007 |
Token | heard |
Token ID | 2982 |
Feature activation | -9.55e-01 |
Pos logit contributions | |
forehead | +0.007 |
hand | +0.007 |
lap | +0.007 |
cheek | +0.007 |
tight | +0.007 |
Neg logit contributions | |
pand | -0.009 |
pill | -0.009 |
add | -0.008 |
otion | -0.008 |
hawk | -0.008 |
Token | outside |
Token ID | 2354 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
pand | +0.024 |
add | +0.021 |
otion | +0.021 |
pill | +0.021 |
hawk | +0.021 |
Neg logit contributions | |
forehead | -0.017 |
bek | -0.016 |
cheek | -0.016 |
lap | -0.015 |
hand | -0.015 |
Token | playing |
Token ID | 2712 |
Feature activation | +1.76e-01 |
Pos logit contributions | |
forehead | +0.035 |
bek | +0.034 |
cheek | +0.033 |
lap | +0.032 |
hand | +0.032 |
Neg logit contributions | |
pand | -0.044 |
otion | -0.038 |
pill | -0.038 |
add | -0.038 |
hawk | -0.038 |
Token | in |
Token ID | 287 |
Feature activation | +2.48e-01 |
Pos logit contributions | |
forehead | +0.004 |
bek | +0.004 |
hand | +0.004 |
cheek | +0.004 |
lap | +0.004 |
Neg logit contributions | |
pand | -0.008 |
pill | -0.008 |
hawk | -0.007 |
add | -0.007 |
otion | -0.007 |
Token | his |
Token ID | 465 |
Feature activation | -6.37e-01 |
Pos logit contributions | |
forehead | +0.007 |
cheek | +0.006 |
bek | +0.006 |
lap | +0.006 |
hand | +0.006 |
Neg logit contributions | |
pand | -0.011 |
pill | -0.010 |
hawk | -0.010 |
otion | -0.010 |
flashes | -0.010 |
Token | backyard |
Token ID | 24296 |
Feature activation | -4.53e-01 |
Pos logit contributions | |
pand | +0.028 |
pill | +0.025 |
hawk | +0.025 |
add | +0.024 |
otion | +0.024 |
Neg logit contributions | |
forehead | -0.019 |
bek | -0.017 |
cheek | -0.017 |
hand | -0.017 |
lap | -0.016 |
Token | . |
Token ID | 13 |
Feature activation | -4.90e-01 |
Pos logit contributions | |
pand | +0.021 |
pill | +0.018 |
otion | +0.018 |
storms | +0.018 |
hawk | +0.018 |
Neg logit contributions | |
forehead | -0.012 |
bek | -0.012 |
cheek | -0.011 |
hand | -0.010 |
lap | -0.010 |
Token | He |
Token ID | 679 |
Feature activation | +2.70e-01 |
Pos logit contributions | |
pand | +0.019 |
pill | +0.017 |
add | +0.017 |
otion | +0.017 |
hawk | +0.016 |
Neg logit contributions | |
forehead | -0.017 |
hand | -0.016 |
cheek | -0.016 |
lap | -0.015 |
bek | -0.015 |
Token | saw |
Token ID | 2497 |
Feature activation | -2.82e-01 |
Pos logit contributions | |
forehead | +0.007 |
hand | +0.006 |
cheek | +0.006 |
lap | +0.006 |
tight | +0.006 |
Neg logit contributions | |
pand | -0.013 |
pill | -0.012 |
hawk | -0.012 |
add | -0.012 |
otion | -0.012 |
Token | a |
Token ID | 257 |
Feature activation | +2.56e-01 |
Pos logit contributions | |
pand | +0.015 |
pill | +0.015 |
storms | +0.014 |
otion | +0.014 |
add | +0.014 |
Neg logit contributions | |
hand | -0.005 |
forehead | -0.005 |
tight | -0.004 |
whole | -0.004 |
lap | -0.004 |
Token | clear |
Token ID | 1598 |
Feature activation | -6.17e-02 |
Pos logit contributions | |
forehead | +0.007 |
hand | +0.007 |
lap | +0.007 |
cheek | +0.007 |
tight | +0.006 |
Neg logit contributions | |
pand | -0.011 |
pill | -0.010 |
hawk | -0.010 |
add | -0.010 |
otion | -0.010 |
Token | sky |
Token ID | 6766 |
Feature activation | -1.42e-01 |
Pos logit contributions | |
pand | +0.002 |
pill | +0.002 |
add | +0.002 |
otion | +0.002 |
storms | +0.002 |
Neg logit contributions | |
forehead | -0.002 |
hand | -0.002 |
cheek | -0.002 |
lap | -0.002 |
tight | -0.002 |
Token | |
Token ID | 220 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
forehead | +0.029 |
cheek | +0.028 |
bek | +0.028 |
hand | +0.026 |
list | +0.025 |
Neg logit contributions | |
pand | -0.026 |
pill | -0.026 |
add | -0.024 |
storms | -0.024 |
otion | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | +4.82e-01 |
Pos logit contributions | |
forehead | +0.005 |
cheek | +0.005 |
hand | +0.005 |
bek | +0.005 |
lap | +0.004 |
Neg logit contributions | |
pand | -0.009 |
pill | -0.008 |
add | -0.008 |
otion | -0.008 |
storms | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | +3.29e-01 |
Pos logit contributions | |
forehead | +0.016 |
cheek | +0.015 |
bek | +0.014 |
hand | +0.014 |
list | +0.013 |
Neg logit contributions | |
pand | -0.018 |
pill | -0.018 |
add | -0.018 |
otion | -0.017 |
storms | -0.016 |
Token | Suddenly |
Token ID | 38582 |
Feature activation | +8.19e-02 |
Pos logit contributions | |
forehead | +0.010 |
cheek | +0.009 |
hand | +0.009 |
lap | +0.008 |
bek | +0.008 |
Neg logit contributions | |
pill | -0.014 |
pand | -0.014 |
add | -0.013 |
otion | -0.013 |
hawk | -0.012 |
Token | , |
Token ID | 11 |
Feature activation | +1.57e-01 |
Pos logit contributions | |
forehead | +0.002 |
hand | +0.002 |
tight | +0.002 |
lap | +0.002 |
cheek | +0.002 |
Neg logit contributions | |
pand | -0.004 |
pill | -0.003 |
hawk | -0.003 |
add | -0.003 |
storms | -0.003 |
Token | something |
Token ID | 1223 |
Feature activation | -2.21e+00 |
Pos logit contributions | |
forehead | +0.005 |
lap | +0.005 |
cheek | +0.005 |
bek | +0.005 |
hand | +0.005 |
Neg logit contributions | |
pand | -0.006 |
otion | -0.005 |
add | -0.005 |
pill | -0.005 |
hawk | -0.005 |
Token | loud |
Token ID | 7812 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
pand | +0.087 |
pill | +0.082 |
otion | +0.079 |
storms | +0.078 |
add | +0.077 |
Neg logit contributions | |
forehead | -0.069 |
tight | -0.068 |
hand | -0.068 |
whole | -0.067 |
satisfied | -0.064 |
Token | happened |
Token ID | 3022 |
Feature activation | -4.78e-01 |
Pos logit contributions | |
pand | +0.045 |
pill | +0.044 |
conflicts | +0.043 |
add | +0.042 |
conclusions | +0.040 |
Neg logit contributions | |
tight | -0.029 |
forehead | -0.029 |
hand | -0.028 |
between | -0.027 |
grabbed | -0.026 |
Token | . |
Token ID | 13 |
Feature activation | +5.06e-01 |
Pos logit contributions | |
pand | +0.024 |
pill | +0.022 |
hawk | +0.022 |
otion | +0.022 |
add | +0.022 |
Neg logit contributions | |
forehead | -0.011 |
hand | -0.010 |
lap | -0.009 |
cheek | -0.009 |
tight | -0.009 |
Token | It |
Token ID | 632 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
forehead | +0.015 |
hand | +0.015 |
lap | +0.014 |
cheek | +0.014 |
tight | +0.014 |
Neg logit contributions | |
pand | -0.021 |
pill | -0.019 |
hawk | -0.019 |
add | -0.019 |
otion | -0.019 |
Token | was |
Token ID | 373 |
Feature activation | +6.81e-02 |
Pos logit contributions | |
pand | +0.064 |
pill | +0.061 |
add | +0.059 |
otion | +0.059 |
hawk | +0.057 |
Neg logit contributions | |
hand | -0.029 |
forehead | -0.029 |
whole | -0.027 |
tight | -0.027 |
lap | -0.026 |
Token | get |
Token ID | 651 |
Feature activation | +3.13e-01 |
Pos logit contributions | |
forehead | +0.010 |
hand | +0.010 |
lap | +0.010 |
cheek | +0.010 |
tight | +0.009 |
Neg logit contributions | |
pand | -0.016 |
pill | -0.015 |
hawk | -0.015 |
add | -0.015 |
otion | -0.015 |
Token | some |
Token ID | 617 |
Feature activation | -2.92e-01 |
Pos logit contributions | |
forehead | +0.009 |
bek | +0.009 |
lap | +0.009 |
cheek | +0.008 |
hand | +0.008 |
Neg logit contributions | |
pand | -0.014 |
add | -0.012 |
hawk | -0.012 |
Hipp | -0.012 |
pill | -0.011 |
Token | clay |
Token ID | 21558 |
Feature activation | +7.50e-01 |
Pos logit contributions | |
pand | +0.013 |
pill | +0.012 |
hawk | +0.012 |
add | +0.012 |
otion | +0.012 |
Neg logit contributions | |
forehead | -0.008 |
hand | -0.008 |
lap | -0.007 |
cheek | -0.007 |
tight | -0.007 |
Token | to |
Token ID | 284 |
Feature activation | -6.58e-02 |
Pos logit contributions | |
forehead | +0.017 |
hand | +0.016 |
lap | +0.016 |
cheek | +0.016 |
tight | +0.015 |
Neg logit contributions | |
pand | -0.036 |
pill | -0.034 |
hawk | -0.034 |
add | -0.034 |
otion | -0.033 |
Token | make |
Token ID | 787 |
Feature activation | -3.55e-01 |
Pos logit contributions | |
pand | +0.003 |
pill | +0.003 |
storms | +0.003 |
add | +0.003 |
otion | +0.003 |
Neg logit contributions | |
forehead | -0.002 |
hand | -0.002 |
lap | -0.001 |
cheek | -0.001 |
tight | -0.001 |
Token | something |
Token ID | 1223 |
Feature activation | -2.39e+00 |
Pos logit contributions | |
pand | +0.016 |
pill | +0.015 |
otion | +0.015 |
hawk | +0.015 |
add | +0.014 |
Neg logit contributions | |
hand | -0.009 |
forehead | -0.009 |
tight | -0.009 |
whole | -0.009 |
cheek | -0.009 |
Token | fun |
Token ID | 1257 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
pand | +0.109 |
pill | +0.102 |
hawk | +0.100 |
add | +0.100 |
otion | +0.100 |
Neg logit contributions | |
forehead | -0.062 |
hand | -0.059 |
cheek | -0.057 |
lap | -0.056 |
tight | -0.055 |
Token | . |
Token ID | 13 |
Feature activation | +2.81e-02 |
Pos logit contributions | |
forehead | +0.003 |
hand | +0.002 |
cheek | +0.002 |
lap | +0.002 |
bek | +0.002 |
Neg logit contributions | |
pand | -0.006 |
pill | -0.005 |
otion | -0.005 |
hawk | -0.005 |
add | -0.005 |
Token | He |
Token ID | 679 |
Feature activation | +2.51e-01 |
Pos logit contributions | |
forehead | +0.001 |
bek | +0.001 |
cheek | +0.001 |
satisfied | +0.001 |
hand | +0.001 |
Neg logit contributions | |
pand | -0.001 |
Hipp | -0.001 |
pill | -0.001 |
Tweet | -0.001 |
storms | -0.001 |
Token | went |
Token ID | 1816 |
Feature activation | -7.15e-01 |
Pos logit contributions | |
forehead | +0.006 |
hand | +0.006 |
tight | +0.006 |
lap | +0.006 |
cheek | +0.006 |
Neg logit contributions | |
pand | -0.012 |
pill | -0.011 |
otion | -0.011 |
hawk | -0.011 |
add | -0.011 |
Token | to |
Token ID | 284 |
Feature activation | +5.71e-01 |
Pos logit contributions | |
pand | +0.035 |
pill | +0.032 |
add | +0.032 |
hawk | +0.031 |
otion | +0.031 |
Neg logit contributions | |
forehead | -0.017 |
bek | -0.015 |
hand | -0.015 |
cheek | -0.015 |
lap | -0.014 |
Token | brother |
Token ID | 3956 |
Feature activation | -6.38e-01 |
Pos logit contributions | |
pand | +0.029 |
pill | +0.027 |
hawk | +0.026 |
add | +0.026 |
otion | +0.025 |
Neg logit contributions | |
forehead | -0.024 |
hand | -0.023 |
tight | -0.022 |
cheek | -0.022 |
lap | -0.022 |
Token | to |
Token ID | 284 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
pand | +0.026 |
pill | +0.024 |
hawk | +0.024 |
add | +0.024 |
otion | +0.023 |
Neg logit contributions | |
forehead | -0.020 |
hand | -0.019 |
tight | -0.019 |
cheek | -0.018 |
lap | -0.018 |
Token | apologize |
Token ID | 16521 |
Feature activation | -4.34e-01 |
Pos logit contributions | |
pand | +0.052 |
add | +0.048 |
hawk | +0.047 |
otion | +0.047 |
die | +0.047 |
Neg logit contributions | |
forehead | -0.036 |
lap | -0.033 |
bek | -0.033 |
hand | -0.032 |
cheek | -0.032 |
Token | and |
Token ID | 290 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
pand | +0.019 |
storms | +0.018 |
otion | +0.018 |
pill | +0.018 |
add | +0.017 |
Neg logit contributions | |
forehead | -0.011 |
hand | -0.011 |
lap | -0.011 |
bek | -0.010 |
cheek | -0.010 |
Token | say |
Token ID | 910 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
pand | +0.044 |
hawk | +0.040 |
pill | +0.039 |
otion | +0.039 |
add | +0.039 |
Neg logit contributions | |
forehead | -0.030 |
lap | -0.028 |
hand | -0.028 |
cheek | -0.028 |
whole | -0.027 |
Token | something |
Token ID | 1223 |
Feature activation | -2.10e+00 |
Pos logit contributions | |
pill | +0.017 |
pand | +0.016 |
add | +0.016 |
hawk | +0.015 |
ican | +0.015 |
Neg logit contributions | |
forehead | -0.011 |
tight | -0.011 |
cheek | -0.011 |
hand | -0.010 |
satisfied | -0.010 |
Token | nice |
Token ID | 3621 |
Feature activation | -8.83e-01 |
Pos logit contributions | |
pand | +0.084 |
pill | +0.078 |
hawk | +0.078 |
add | +0.076 |
otion | +0.075 |
Neg logit contributions | |
forehead | -0.068 |
hand | -0.064 |
cheek | -0.064 |
tight | -0.063 |
lap | -0.061 |
Token | about |
Token ID | 546 |
Feature activation | +1.99e-01 |
Pos logit contributions | |
pand | +0.039 |
pill | +0.037 |
otion | +0.036 |
hawk | +0.036 |
add | +0.036 |
Neg logit contributions | |
forehead | -0.024 |
hand | -0.023 |
lap | -0.022 |
cheek | -0.022 |
tight | -0.021 |
Token | the |
Token ID | 262 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
forehead | +0.006 |
hand | +0.006 |
cheek | +0.006 |
lap | +0.006 |
tight | +0.006 |
Neg logit contributions | |
pand | -0.008 |
pill | -0.008 |
add | -0.008 |
hawk | -0.007 |
otion | -0.007 |
Token | toy |
Token ID | 13373 |
Feature activation | -8.63e-02 |
Pos logit contributions | |
pand | +0.006 |
pill | +0.005 |
hawk | +0.005 |
otion | +0.005 |
add | +0.005 |
Neg logit contributions | |
forehead | -0.004 |
hand | -0.004 |
lap | -0.004 |
cheek | -0.004 |
bek | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -7.49e-01 |
Pos logit contributions | |
pand | +0.004 |
pill | +0.004 |
hawk | +0.004 |
add | +0.004 |
otion | +0.004 |
Neg logit contributions | |
forehead | -0.002 |
hand | -0.002 |
lap | -0.002 |
cheek | -0.002 |
tight | -0.002 |
Token | of |
Token ID | 286 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
pand | +0.043 |
pill | +0.039 |
add | +0.038 |
hawk | +0.037 |
otion | +0.036 |
Neg logit contributions | |
forehead | -0.031 |
hand | -0.028 |
bek | -0.028 |
cheek | -0.028 |
lap | -0.027 |
Token | their |
Token ID | 511 |
Feature activation | -7.82e-01 |
Pos logit contributions | |
pand | +0.039 |
pill | +0.036 |
hawk | +0.036 |
add | +0.036 |
otion | +0.035 |
Neg logit contributions | |
forehead | -0.033 |
hand | -0.032 |
lap | -0.031 |
cheek | -0.031 |
tight | -0.031 |
Token | oats |
Token ID | 47009 |
Feature activation | -5.96e-01 |
Pos logit contributions | |
pand | +0.026 |
pill | +0.023 |
add | +0.023 |
hawk | +0.023 |
otion | +0.022 |
Neg logit contributions | |
forehead | -0.031 |
hand | -0.029 |
cheek | -0.029 |
lap | -0.029 |
tight | -0.028 |
Token | with |
Token ID | 351 |
Feature activation | +5.49e-03 |
Pos logit contributions | |
pand | +0.021 |
Tweet | +0.019 |
die | +0.017 |
pill | +0.017 |
anymore | +0.017 |
Neg logit contributions | |
forehead | -0.021 |
bek | -0.021 |
doorway | -0.019 |
hand | -0.018 |
pron | -0.018 |
Token | the |
Token ID | 262 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
forehead | +0.000 |
lap | +0.000 |
hand | +0.000 |
cheek | +0.000 |
bek | +0.000 |
Neg logit contributions | |
pand | -0.000 |
pill | -0.000 |
add | -0.000 |
hawk | -0.000 |
otion | -0.000 |
Token | bird |
Token ID | 6512 |
Feature activation | -2.49e+00 |
Pos logit contributions | |
pand | +0.034 |
pill | +0.032 |
otion | +0.032 |
storms | +0.031 |
add | +0.030 |
Neg logit contributions | |
hand | -0.038 |
forehead | -0.038 |
whole | -0.036 |
cheek | -0.036 |
tight | -0.036 |
Token | . |
Token ID | 13 |
Feature activation | +2.14e-01 |
Pos logit contributions | |
pand | +0.117 |
pill | +0.108 |
add | +0.104 |
hawk | +0.104 |
storms | +0.103 |
Neg logit contributions | |
forehead | -0.065 |
hand | -0.058 |
cheek | -0.057 |
lap | -0.057 |
bek | -0.057 |
Token | The |
Token ID | 383 |
Feature activation | -1.72e-01 |
Pos logit contributions | |
forehead | +0.008 |
hand | +0.008 |
cheek | +0.008 |
lap | +0.007 |
bek | +0.007 |
Neg logit contributions | |
pand | -0.007 |
pill | -0.007 |
add | -0.007 |
otion | -0.006 |
hawk | -0.006 |
Token | bird |
Token ID | 6512 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
pand | +0.005 |
pill | +0.005 |
storms | +0.005 |
otion | +0.005 |
add | +0.004 |
Neg logit contributions | |
hand | -0.007 |
forehead | -0.007 |
whole | -0.007 |
cheek | -0.007 |
tight | -0.007 |
Token | was |
Token ID | 373 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
pand | +0.080 |
otion | +0.077 |
pill | +0.076 |
add | +0.075 |
hawk | +0.074 |
Neg logit contributions | |
forehead | -0.039 |
hand | -0.039 |
tight | -0.037 |
cheek | -0.036 |
lap | -0.035 |
Token | very |
Token ID | 845 |
Feature activation | +9.83e-01 |
Pos logit contributions | |
forehead | +0.003 |
hand | +0.003 |
tight | +0.003 |
cheek | +0.003 |
lap | +0.003 |
Neg logit contributions | |
pand | -0.006 |
pill | -0.006 |
otion | -0.006 |
add | -0.006 |
hawk | -0.006 |
Token | her |
Token ID | 607 |
Feature activation | -2.78e-01 |
Pos logit contributions | |
pand | +0.012 |
pill | +0.011 |
hawk | +0.011 |
add | +0.011 |
otion | +0.011 |
Neg logit contributions | |
forehead | -0.010 |
hand | -0.010 |
lap | -0.009 |
cheek | -0.009 |
tight | -0.009 |
Token | point |
Token ID | 966 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
pand | +0.012 |
hawk | +0.011 |
otion | +0.011 |
die | +0.011 |
storms | +0.011 |
Neg logit contributions | |
forehead | -0.007 |
bek | -0.007 |
cheek | -0.006 |
entire | -0.006 |
hand | -0.006 |
Token | it |
Token ID | 340 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
forehead | +0.002 |
hand | +0.002 |
lap | +0.002 |
cheek | +0.002 |
bek | +0.002 |
Neg logit contributions | |
pand | -0.005 |
pill | -0.005 |
hawk | -0.004 |
otion | -0.004 |
add | -0.004 |
Token | at |
Token ID | 379 |
Feature activation | -3.78e-01 |
Pos logit contributions | |
pand | +0.063 |
pill | +0.059 |
hawk | +0.058 |
add | +0.058 |
otion | +0.058 |
Neg logit contributions | |
forehead | -0.026 |
hand | -0.025 |
lap | -0.024 |
cheek | -0.024 |
tight | -0.023 |
Token | the |
Token ID | 262 |
Feature activation | +6.56e-01 |
Pos logit contributions | |
pand | +0.017 |
pill | +0.015 |
hawk | +0.015 |
otion | +0.015 |
add | +0.015 |
Neg logit contributions | |
forehead | -0.010 |
hand | -0.010 |
cheek | -0.010 |
lap | -0.010 |
bek | -0.009 |
Token | bird |
Token ID | 6512 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
forehead | +0.021 |
bek | +0.021 |
lap | +0.020 |
satisfied | +0.019 |
ae | +0.019 |
Neg logit contributions | |
pand | -0.025 |
hawk | -0.024 |
aliens | -0.023 |
pill | -0.022 |
storms | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | -1.58e-01 |
Pos logit contributions | |
pand | +0.119 |
add | +0.114 |
pill | +0.113 |
hawk | +0.112 |
otion | +0.112 |
Neg logit contributions | |
forehead | -0.042 |
hand | -0.041 |
lap | -0.038 |
cheek | -0.038 |
tight | -0.037 |
Token | He |
Token ID | 679 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
pand | +0.007 |
hawk | +0.006 |
add | +0.006 |
pill | +0.006 |
otion | +0.006 |
Neg logit contributions | |
hand | -0.004 |
forehead | -0.004 |
lap | -0.004 |
tight | -0.004 |
cheek | -0.004 |
Token | tells |
Token ID | 4952 |
Feature activation | -2.27e-01 |
Pos logit contributions | |
hand | +0.003 |
tight | +0.003 |
head | +0.003 |
forehead | +0.003 |
hands | +0.003 |
Neg logit contributions | |
pand | -0.005 |
otion | -0.005 |
add | -0.005 |
hawk | -0.004 |
storms | -0.004 |
Token | her |
Token ID | 607 |
Feature activation | +4.19e-01 |
Pos logit contributions | |
pill | +0.011 |
pand | +0.010 |
hawk | +0.010 |
storms | +0.010 |
otion | +0.010 |
Neg logit contributions | |
forehead | -0.005 |
hand | -0.005 |
tight | -0.005 |
cheek | -0.005 |
plate | -0.005 |
Token | to |
Token ID | 284 |
Feature activation | -3.00e-01 |
Pos logit contributions | |
hand | +0.011 |
forehead | +0.011 |
tight | +0.011 |
lap | +0.010 |
cheek | +0.010 |
Neg logit contributions | |
pand | -0.018 |
pill | -0.018 |
add | -0.018 |
hawk | -0.017 |
ican | -0.017 |