Token | asked |
Token ID | 1965 |
Feature activation | +7.15e-01 |
Pos logit contributions | |
pole | +0.003 |
sharing | +0.003 |
isk | +0.003 |
Freddie | +0.003 |
ole | +0.003 |
Neg logit contributions | |
roller | -0.003 |
scrap | -0.002 |
squee | -0.002 |
steps | -0.002 |
levers | -0.002 |
Token | him |
Token ID | 683 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
pole | +0.017 |
isk | +0.016 |
ole | +0.016 |
sharing | +0.015 |
melon | +0.015 |
Neg logit contributions | |
roller | -0.014 |
steps | -0.013 |
pine | -0.012 |
palm | -0.012 |
bars | -0.012 |
Token | , |
Token ID | 11 |
Feature activation | +6.70e-01 |
Pos logit contributions | |
pole | +0.041 |
isk | +0.041 |
sharing | +0.039 |
ole | +0.039 |
Freddie | +0.037 |
Neg logit contributions | |
roller | -0.047 |
steps | -0.043 |
palm | -0.043 |
squee | -0.042 |
bars | -0.042 |
Token | " |
Token ID | 366 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
isk | +0.015 |
pole | +0.015 |
ole | +0.015 |
py | +0.014 |
ingu | +0.014 |
Neg logit contributions | |
roller | -0.013 |
squee | -0.012 |
palm | -0.012 |
wonder | -0.012 |
rooft | -0.012 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.76e+00 |
Pos logit contributions | |
roller | +0.034 |
scrap | +0.029 |
squee | +0.029 |
steps | +0.029 |
levers | +0.029 |
Neg logit contributions | |
pole | -0.036 |
sharing | -0.035 |
isk | -0.035 |
Freddie | -0.033 |
ole | -0.033 |
Token | are |
Token ID | 389 |
Feature activation | +7.23e+00 |
Pos logit contributions | |
roller | +0.061 |
scrap | +0.054 |
squee | +0.053 |
steps | +0.053 |
levers | +0.053 |
Neg logit contributions | |
pole | -0.058 |
sharing | -0.057 |
isk | -0.056 |
Freddie | -0.054 |
ole | -0.054 |
Token | you |
Token ID | 345 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
isk | +0.160 |
pole | +0.140 |
End | +0.126 |
ility | +0.124 |
ending | +0.123 |
Neg logit contributions | |
roller | -0.169 |
steps | -0.164 |
palm | -0.160 |
pine | -0.156 |
da | -0.154 |
Token | going |
Token ID | 1016 |
Feature activation | +2.82e+00 |
Pos logit contributions | |
isk | +0.042 |
pole | +0.038 |
ole | +0.038 |
ingu | +0.037 |
Freddie | +0.036 |
Neg logit contributions | |
roller | -0.035 |
wonder | -0.034 |
palm | -0.034 |
steps | -0.033 |
da | -0.031 |
Token | , |
Token ID | 11 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
pole | +0.062 |
isk | +0.061 |
sharing | +0.058 |
Freddie | +0.057 |
ole | +0.056 |
Neg logit contributions | |
roller | -0.061 |
steps | -0.054 |
squee | -0.053 |
palm | -0.052 |
rooft | -0.052 |
Token | Mouse |
Token ID | 21839 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
isk | +0.023 |
pole | +0.022 |
ole | +0.021 |
sharing | +0.021 |
melon | +0.020 |
Neg logit contributions | |
roller | -0.021 |
palm | -0.020 |
wonder | -0.019 |
pine | -0.019 |
squee | -0.019 |
Token | ?" |
Token ID | 1701 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
pole | +0.039 |
sharing | +0.036 |
End | +0.036 |
isk | +0.036 |
Freddie | +0.036 |
Neg logit contributions | |
roller | -0.037 |
palm | -0.033 |
squee | -0.033 |
wonder | -0.032 |
pine | -0.032 |
Token | and |
Token ID | 290 |
Feature activation | -6.75e-01 |
Pos logit contributions | |
pole | +0.004 |
sharing | +0.004 |
isk | +0.004 |
ole | +0.004 |
Freddie | +0.004 |
Neg logit contributions | |
roller | -0.005 |
scrap | -0.005 |
squee | -0.005 |
palm | -0.005 |
rooft | -0.005 |
Token | asked |
Token ID | 1965 |
Feature activation | +4.98e-01 |
Pos logit contributions | |
roller | +0.014 |
scrap | +0.012 |
steps | +0.012 |
squee | +0.012 |
levers | +0.012 |
Neg logit contributions | |
pole | -0.016 |
isk | -0.015 |
ole | -0.015 |
sharing | -0.015 |
Freddie | -0.014 |
Token | , |
Token ID | 11 |
Feature activation | +2.20e-01 |
Pos logit contributions | |
sharing | +0.008 |
pole | +0.008 |
isk | +0.008 |
being | +0.007 |
Freddie | +0.007 |
Neg logit contributions | |
roller | -0.014 |
scrap | -0.012 |
rooft | -0.012 |
palm | -0.012 |
squee | -0.012 |
Token | " |
Token ID | 366 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
pole | +0.004 |
isk | +0.004 |
sharing | +0.004 |
ole | +0.004 |
py | +0.004 |
Neg logit contributions | |
roller | -0.005 |
squee | -0.005 |
rooft | -0.005 |
levers | -0.005 |
steps | -0.005 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.70e+00 |
Pos logit contributions | |
roller | +0.040 |
scrap | +0.037 |
steps | +0.036 |
palm | +0.036 |
squee | +0.036 |
Neg logit contributions | |
pole | -0.028 |
sharing | -0.028 |
isk | -0.027 |
being | -0.026 |
Freddie | -0.025 |
Token | are |
Token ID | 389 |
Feature activation | +7.21e+00 |
Pos logit contributions | |
roller | +0.064 |
rooft | +0.058 |
levers | +0.057 |
scrap | +0.056 |
palm | +0.056 |
Neg logit contributions | |
sharing | -0.053 |
pole | -0.051 |
isk | -0.049 |
being | -0.048 |
Freddie | -0.048 |
Token | you |
Token ID | 345 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
pole | +0.139 |
isk | +0.139 |
ole | +0.117 |
ility | +0.114 |
End | +0.113 |
Neg logit contributions | |
roller | -0.184 |
steps | -0.172 |
umb | -0.166 |
squee | -0.165 |
palm | -0.164 |
Token | going |
Token ID | 1016 |
Feature activation | +2.76e+00 |
Pos logit contributions | |
pole | +0.044 |
isk | +0.043 |
ole | +0.041 |
End | +0.037 |
ingu | +0.037 |
Neg logit contributions | |
roller | -0.033 |
steps | -0.029 |
palm | -0.029 |
buck | -0.028 |
wonder | -0.028 |
Token | , |
Token ID | 11 |
Feature activation | +8.19e-01 |
Pos logit contributions | |
pole | +0.062 |
isk | +0.060 |
Freddie | +0.056 |
sharing | +0.056 |
ole | +0.055 |
Neg logit contributions | |
roller | -0.061 |
steps | -0.052 |
squee | -0.051 |
scrap | -0.049 |
palm | -0.049 |
Token | Lisa |
Token ID | 15378 |
Feature activation | +2.19e+00 |
Pos logit contributions | |
sharing | +0.017 |
pole | +0.017 |
isk | +0.016 |
Freddie | +0.015 |
being | +0.015 |
Neg logit contributions | |
roller | -0.019 |
scrap | -0.017 |
rooft | -0.016 |
palm | -0.016 |
steps | -0.016 |
Token | ?" |
Token ID | 1701 |
Feature activation | -1.91e+00 |
Pos logit contributions | |
isk | +0.048 |
pole | +0.047 |
End | +0.046 |
ole | +0.046 |
Freddie | +0.045 |
Neg logit contributions | |
roller | -0.041 |
squee | -0.038 |
scrap | -0.037 |
wonder | -0.036 |
palm | -0.036 |
Token | and |
Token ID | 290 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
roller | +0.005 |
steps | +0.004 |
squee | +0.004 |
bars | +0.004 |
seats | +0.004 |
Neg logit contributions | |
pole | -0.008 |
isk | -0.008 |
ole | -0.007 |
py | -0.007 |
End | -0.007 |
Token | asked |
Token ID | 1965 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
pole | +0.003 |
isk | +0.003 |
sharing | +0.002 |
ole | +0.002 |
py | +0.002 |
Neg logit contributions | |
roller | -0.002 |
scrap | -0.002 |
steps | -0.002 |
squee | -0.002 |
palm | -0.002 |
Token | , |
Token ID | 11 |
Feature activation | -4.07e-01 |
Pos logit contributions | |
roller | +0.006 |
squee | +0.005 |
steps | +0.005 |
palm | +0.005 |
scrap | +0.005 |
Neg logit contributions | |
pole | -0.006 |
isk | -0.005 |
sharing | -0.005 |
ole | -0.005 |
ility | -0.005 |
Token | " |
Token ID | 366 |
Feature activation | -2.05e+00 |
Pos logit contributions | |
roller | +0.008 |
squee | +0.007 |
levers | +0.007 |
palm | +0.007 |
wonder | +0.007 |
Neg logit contributions | |
isk | -0.010 |
pole | -0.009 |
py | -0.009 |
ole | -0.009 |
ending | -0.009 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.70e+00 |
Pos logit contributions | |
roller | +0.040 |
scrap | +0.034 |
squee | +0.033 |
rooft | +0.033 |
palm | +0.033 |
Neg logit contributions | |
pole | -0.049 |
sharing | -0.048 |
isk | -0.047 |
ole | -0.046 |
Freddie | -0.046 |
Token | are |
Token ID | 389 |
Feature activation | +7.07e+00 |
Pos logit contributions | |
roller | +0.056 |
steps | +0.048 |
squee | +0.048 |
scrap | +0.048 |
palm | +0.047 |
Neg logit contributions | |
pole | -0.062 |
isk | -0.060 |
sharing | -0.059 |
ole | -0.058 |
Freddie | -0.057 |
Token | we |
Token ID | 356 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
isk | +0.157 |
pole | +0.150 |
py | +0.138 |
ility | +0.134 |
End | +0.125 |
Neg logit contributions | |
roller | -0.171 |
seats | -0.156 |
steps | -0.153 |
umb | -0.147 |
palm | -0.145 |
Token | going |
Token ID | 1016 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
pole | +0.054 |
py | +0.052 |
isk | +0.051 |
ility | +0.046 |
ole | +0.045 |
Neg logit contributions | |
roller | -0.037 |
buck | -0.034 |
wonder | -0.033 |
steps | -0.033 |
seats | -0.031 |
Token | ?" |
Token ID | 1701 |
Feature activation | -1.85e+00 |
Pos logit contributions | |
pole | +0.047 |
isk | +0.046 |
sharing | +0.043 |
Freddie | +0.043 |
ole | +0.043 |
Neg logit contributions | |
roller | -0.042 |
steps | -0.035 |
seats | -0.035 |
squee | -0.035 |
buck | -0.034 |
Token | His |
Token ID | 2399 |
Feature activation | +2.09e-01 |
Pos logit contributions | |
roller | +0.042 |
palm | +0.038 |
scrap | +0.037 |
wonder | +0.037 |
squee | +0.037 |
Neg logit contributions | |
pole | -0.036 |
isk | -0.036 |
sharing | -0.036 |
ole | -0.035 |
py | -0.034 |
Token | mom |
Token ID | 1995 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
sharing | +0.003 |
isk | +0.003 |
ole | +0.003 |
ending | +0.003 |
Mickey | +0.003 |
Neg logit contributions | |
roller | -0.005 |
leaf | -0.005 |
levers | -0.005 |
bars | -0.005 |
rooft | -0.005 |
Token | Mom |
Token ID | 11254 |
Feature activation | +4.84e-01 |
Pos logit contributions | |
sharing | +0.006 |
pole | +0.006 |
isk | +0.006 |
Freddie | +0.006 |
ole | +0.006 |
Neg logit contributions | |
roller | -0.010 |
rooft | -0.009 |
scrap | -0.009 |
squee | -0.009 |
steps | -0.009 |
Token | my |
Token ID | 1820 |
Feature activation | +3.56e-01 |
Pos logit contributions | |
pole | +0.007 |
Freddie | +0.007 |
ending | +0.007 |
isk | +0.007 |
ole | +0.007 |
Neg logit contributions | |
roller | -0.013 |
palm | -0.012 |
pine | -0.012 |
squee | -0.012 |
steps | -0.012 |
Token | , |
Token ID | 11 |
Feature activation | +9.80e-01 |
Pos logit contributions | |
pole | +0.008 |
isk | +0.008 |
ole | +0.007 |
py | +0.007 |
sharing | +0.007 |
Neg logit contributions | |
roller | -0.008 |
steps | -0.007 |
palm | -0.007 |
squee | -0.007 |
wonder | -0.007 |
Token | " |
Token ID | 366 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
pole | +0.017 |
isk | +0.017 |
sharing | +0.016 |
ole | +0.016 |
Freddie | +0.016 |
Neg logit contributions | |
roller | -0.025 |
squee | -0.023 |
scrap | -0.022 |
palm | -0.022 |
steps | -0.022 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.56e+00 |
Pos logit contributions | |
roller | +0.039 |
steps | +0.038 |
scrap | +0.036 |
levers | +0.036 |
squee | +0.036 |
Neg logit contributions | |
pole | -0.036 |
sharing | -0.035 |
isk | -0.035 |
Freddie | -0.032 |
ole | -0.032 |
Token | are |
Token ID | 389 |
Feature activation | +7.05e+00 |
Pos logit contributions | |
roller | +0.058 |
rooft | +0.052 |
scrap | +0.052 |
levers | +0.052 |
steps | +0.051 |
Neg logit contributions | |
sharing | -0.051 |
pole | -0.051 |
isk | -0.049 |
Freddie | -0.048 |
ole | -0.047 |
Token | we |
Token ID | 356 |
Feature activation | +2.13e+00 |
Pos logit contributions | |
isk | +0.136 |
pole | +0.133 |
ility | +0.119 |
py | +0.118 |
ole | +0.114 |
Neg logit contributions | |
roller | -0.184 |
steps | -0.163 |
palm | -0.163 |
seats | -0.160 |
pine | -0.159 |
Token | going |
Token ID | 1016 |
Feature activation | +2.28e+00 |
Pos logit contributions | |
isk | +0.057 |
pole | +0.056 |
py | +0.054 |
ility | +0.051 |
ole | +0.050 |
Neg logit contributions | |
roller | -0.038 |
palm | -0.033 |
wonder | -0.032 |
buck | -0.031 |
steps | -0.030 |
Token | ?" |
Token ID | 1701 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
pole | +0.044 |
isk | +0.042 |
sharing | +0.041 |
Freddie | +0.040 |
ole | +0.039 |
Neg logit contributions | |
roller | -0.056 |
steps | -0.049 |
squee | -0.049 |
scrap | -0.048 |
palm | -0.048 |
Token | Mom |
Token ID | 11254 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
roller | +0.048 |
rooft | +0.045 |
levers | +0.045 |
steps | +0.044 |
bars | +0.044 |
Neg logit contributions | |
sharing | -0.038 |
pole | -0.037 |
isk | -0.036 |
Freddie | -0.036 |
Mickey | -0.036 |
Token | my |
Token ID | 1820 |
Feature activation | -1.99e+00 |
Pos logit contributions | |
roller | +0.049 |
scrap | +0.045 |
squee | +0.045 |
steps | +0.044 |
palm | +0.044 |
Neg logit contributions | |
pole | -0.016 |
sharing | -0.016 |
isk | -0.016 |
ole | -0.015 |
Freddie | -0.015 |
Token | big |
Token ID | 1263 |
Feature activation | -1.56e+00 |
Pos logit contributions | |
roller | +0.047 |
scrap | +0.041 |
steps | +0.041 |
squee | +0.041 |
palm | +0.040 |
Neg logit contributions | |
pole | -0.043 |
sharing | -0.042 |
isk | -0.041 |
ole | -0.040 |
Freddie | -0.040 |
Token | adventure |
Token ID | 8855 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
roller | +0.038 |
scrap | +0.034 |
steps | +0.033 |
squee | +0.033 |
palm | +0.033 |
Neg logit contributions | |
pole | -0.030 |
sharing | -0.029 |
isk | -0.028 |
ole | -0.027 |
Freddie | -0.027 |
Token | . |
Token ID | 13 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
pole | +0.038 |
isk | +0.037 |
sharing | +0.037 |
Freddie | +0.035 |
ole | +0.035 |
Neg logit contributions | |
roller | -0.038 |
steps | -0.033 |
squee | -0.032 |
scrap | -0.032 |
levers | -0.032 |
Token | " |
Token ID | 366 |
Feature activation | -2.82e+00 |
Pos logit contributions | |
pole | +0.005 |
isk | +0.005 |
sharing | +0.005 |
ole | +0.004 |
py | +0.004 |
Neg logit contributions | |
roller | -0.005 |
steps | -0.004 |
squee | -0.004 |
wonder | -0.004 |
palm | -0.004 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.66e+00 |
Pos logit contributions | |
roller | +0.046 |
steps | +0.041 |
seats | +0.040 |
levers | +0.039 |
scrap | +0.039 |
Neg logit contributions | |
pole | -0.076 |
sharing | -0.075 |
isk | -0.074 |
Freddie | -0.071 |
ole | -0.070 |
Token | are |
Token ID | 389 |
Feature activation | +7.04e+00 |
Pos logit contributions | |
roller | +0.060 |
rooft | +0.054 |
scrap | +0.053 |
levers | +0.053 |
steps | +0.052 |
Neg logit contributions | |
sharing | -0.055 |
pole | -0.054 |
isk | -0.052 |
Freddie | -0.051 |
ole | -0.050 |
Token | you |
Token ID | 345 |
Feature activation | +1.65e+00 |
Pos logit contributions | |
isk | +0.140 |
pole | +0.128 |
ility | +0.113 |
ole | +0.111 |
py | +0.110 |
Neg logit contributions | |
roller | -0.184 |
steps | -0.168 |
palm | -0.163 |
seats | -0.163 |
desks | -0.159 |
Token | going |
Token ID | 1016 |
Feature activation | +2.78e+00 |
Pos logit contributions | |
isk | +0.038 |
pole | +0.038 |
ole | +0.036 |
Freddie | +0.034 |
melon | +0.034 |
Neg logit contributions | |
roller | -0.034 |
palm | -0.030 |
steps | -0.030 |
wonder | -0.030 |
buck | -0.029 |
Token | ?", |
Token ID | 35379 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
pole | +0.061 |
isk | +0.060 |
sharing | +0.056 |
Freddie | +0.056 |
ole | +0.055 |
Neg logit contributions | |
roller | -0.062 |
steps | -0.053 |
squee | -0.052 |
palm | -0.051 |
rooft | -0.051 |
Token | asked |
Token ID | 1965 |
Feature activation | -9.69e-01 |
Pos logit contributions | |
roller | +0.041 |
squee | +0.039 |
wonder | +0.038 |
steps | +0.038 |
levers | +0.038 |
Neg logit contributions | |
isk | -0.031 |
pole | -0.029 |
ole | -0.027 |
sharing | -0.027 |
ending | -0.026 |
Token | Jill |
Token ID | 22772 |
Feature activation | -1.94e-01 |
Pos logit contributions | |
roller | +0.020 |
squee | +0.017 |
levers | +0.017 |
steps | +0.017 |
scrap | +0.017 |
Neg logit contributions | |
isk | -0.022 |
pole | -0.022 |
sharing | -0.021 |
ending | -0.020 |
py | -0.020 |
Token | the |
Token ID | 262 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
pole | +0.007 |
sharing | +0.006 |
isk | +0.006 |
ole | +0.006 |
Freddie | +0.006 |
Neg logit contributions | |
roller | -0.009 |
steps | -0.009 |
scrap | -0.008 |
squee | -0.008 |
palm | -0.008 |
Token | pear |
Token ID | 25286 |
Feature activation | -5.98e-02 |
Pos logit contributions | |
sharing | +0.029 |
pole | +0.029 |
ole | +0.027 |
isk | +0.027 |
ending | +0.027 |
Neg logit contributions | |
roller | -0.031 |
levers | -0.028 |
obs | -0.027 |
squee | -0.027 |
scrap | -0.027 |
Token | , |
Token ID | 11 |
Feature activation | +5.54e-01 |
Pos logit contributions | |
roller | +0.001 |
steps | +0.001 |
palm | +0.001 |
bars | +0.001 |
pine | +0.001 |
Neg logit contributions | |
pole | -0.001 |
sharing | -0.001 |
isk | -0.001 |
Freddie | -0.001 |
Mickey | -0.001 |
Token | " |
Token ID | 366 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
isk | +0.012 |
pole | +0.011 |
ole | +0.011 |
sharing | +0.011 |
ending | +0.010 |
Neg logit contributions | |
roller | -0.012 |
squee | -0.011 |
palm | -0.011 |
levers | -0.011 |
steps | -0.011 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.67e+00 |
Pos logit contributions | |
roller | +0.034 |
seats | +0.031 |
steps | +0.030 |
levers | +0.030 |
squee | +0.030 |
Neg logit contributions | |
pole | -0.029 |
isk | -0.028 |
sharing | -0.027 |
Freddie | -0.026 |
ole | -0.026 |
Token | are |
Token ID | 389 |
Feature activation | +7.02e+00 |
Pos logit contributions | |
roller | +0.059 |
scrap | +0.052 |
rooft | +0.052 |
levers | +0.052 |
steps | +0.052 |
Neg logit contributions | |
pole | -0.056 |
sharing | -0.055 |
isk | -0.054 |
Freddie | -0.052 |
ole | -0.052 |
Token | you |
Token ID | 345 |
Feature activation | +1.65e+00 |
Pos logit contributions | |
isk | +0.143 |
pole | +0.135 |
ility | +0.122 |
ending | +0.120 |
py | +0.119 |
Neg logit contributions | |
roller | -0.172 |
palm | -0.162 |
steps | -0.161 |
pine | -0.159 |
scrap | -0.156 |
Token | going |
Token ID | 1016 |
Feature activation | +2.54e+00 |
Pos logit contributions | |
pole | +0.035 |
isk | +0.033 |
ole | +0.032 |
Freddie | +0.030 |
py | +0.029 |
Neg logit contributions | |
roller | -0.038 |
palm | -0.035 |
squee | -0.034 |
wonder | -0.034 |
steps | -0.033 |
Token | ?". |
Token ID | 43634 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
pole | +0.053 |
isk | +0.051 |
sharing | +0.049 |
Freddie | +0.048 |
ole | +0.048 |
Neg logit contributions | |
roller | -0.058 |
steps | -0.051 |
squee | -0.050 |
scrap | -0.050 |
palm | -0.049 |
Token | The |
Token ID | 383 |
Feature activation | -2.10e-01 |
Pos logit contributions | |
roller | +0.037 |
scrap | +0.033 |
squee | +0.033 |
steps | +0.033 |
palm | +0.033 |
Neg logit contributions | |
pole | -0.029 |
sharing | -0.028 |
isk | -0.027 |
ole | -0.026 |
Freddie | -0.026 |
Token | pear |
Token ID | 25286 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
roller | +0.004 |
levers | +0.003 |
squee | +0.003 |
obs | +0.003 |
rooft | +0.003 |
Neg logit contributions | |
pole | -0.005 |
sharing | -0.005 |
ole | -0.005 |
isk | -0.005 |
End | -0.005 |
Token | wise |
Token ID | 10787 |
Feature activation | +3.75e-01 |
Pos logit contributions | |
sharing | +0.018 |
pole | +0.018 |
Freddie | +0.018 |
Mickey | +0.017 |
ole | +0.017 |
Neg logit contributions | |
roller | -0.016 |
levers | -0.014 |
scrap | -0.014 |
desks | -0.014 |
steps | -0.014 |
Token | turtle |
Token ID | 28699 |
Feature activation | +8.08e-01 |
Pos logit contributions | |
pole | +0.008 |
isk | +0.008 |
sharing | +0.008 |
Freddie | +0.007 |
ole | +0.007 |
Neg logit contributions | |
roller | -0.008 |
palm | -0.007 |
steps | -0.007 |
squee | -0.007 |
wonder | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | +7.60e-01 |
Pos logit contributions | |
isk | +0.017 |
pole | +0.017 |
sharing | +0.017 |
ending | +0.017 |
Freddie | +0.017 |
Neg logit contributions | |
roller | -0.016 |
bars | -0.015 |
palm | -0.015 |
squee | -0.015 |
steps | -0.014 |
Token | " |
Token ID | 366 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
isk | +0.018 |
pole | +0.018 |
sharing | +0.017 |
ator | +0.017 |
ole | +0.017 |
Neg logit contributions | |
roller | -0.014 |
steps | -0.014 |
squee | -0.013 |
wonder | -0.013 |
palm | -0.013 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.73e+00 |
Pos logit contributions | |
roller | +0.051 |
scrap | +0.044 |
squee | +0.043 |
palm | +0.043 |
rooft | +0.043 |
Neg logit contributions | |
pole | -0.070 |
sharing | -0.069 |
isk | -0.068 |
ole | -0.066 |
Freddie | -0.066 |
Token | are |
Token ID | 389 |
Feature activation | +7.01e+00 |
Pos logit contributions | |
roller | +0.058 |
scrap | +0.050 |
steps | +0.050 |
squee | +0.050 |
palm | +0.050 |
Neg logit contributions | |
pole | -0.061 |
isk | -0.059 |
sharing | -0.058 |
ole | -0.057 |
Freddie | -0.056 |
Token | you |
Token ID | 345 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
isk | +0.159 |
pole | +0.142 |
ility | +0.132 |
End | +0.130 |
ator | +0.125 |
Neg logit contributions | |
roller | -0.157 |
steps | -0.154 |
palm | -0.146 |
squee | -0.143 |
pine | -0.142 |
Token | going |
Token ID | 1016 |
Feature activation | +2.76e+00 |
Pos logit contributions | |
isk | +0.046 |
ole | +0.045 |
pole | +0.044 |
ingu | +0.042 |
penter | +0.041 |
Neg logit contributions | |
steps | -0.031 |
wonder | -0.031 |
roller | -0.031 |
pe | -0.030 |
palm | -0.030 |
Token | , |
Token ID | 11 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
pole | +0.071 |
isk | +0.070 |
End | +0.066 |
ending | +0.065 |
sharing | +0.064 |
Neg logit contributions | |
roller | -0.049 |
steps | -0.043 |
squee | -0.041 |
palm | -0.040 |
bars | -0.039 |
Token | Fred |
Token ID | 8559 |
Feature activation | +3.24e+00 |
Pos logit contributions | |
isk | +0.023 |
pole | +0.020 |
ingu | +0.019 |
ole | +0.019 |
ator | +0.019 |
Neg logit contributions | |
wonder | -0.023 |
palm | -0.022 |
roller | -0.022 |
steps | -0.022 |
squee | -0.021 |
Token | ?" |
Token ID | 1701 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
isk | +0.076 |
pole | +0.076 |
End | +0.074 |
ole | +0.072 |
ending | +0.072 |
Neg logit contributions | |
roller | -0.057 |
squee | -0.052 |
scrap | -0.050 |
pine | -0.049 |
palm | -0.049 |
Token | her |
Token ID | 607 |
Feature activation | +2.33e+00 |
Pos logit contributions | |
sharing | +0.012 |
isk | +0.012 |
pole | +0.012 |
Freddie | +0.011 |
Mickey | +0.011 |
Neg logit contributions | |
roller | -0.015 |
levers | -0.013 |
scrap | -0.013 |
rooft | -0.013 |
squee | -0.013 |
Token | mum |
Token ID | 25682 |
Feature activation | +6.62e-02 |
Pos logit contributions | |
sharing | +0.035 |
Mickey | +0.033 |
ending | +0.032 |
Freddie | +0.031 |
being | +0.031 |
Neg logit contributions | |
roller | -0.066 |
levers | -0.061 |
bang | -0.060 |
rooft | -0.060 |
scrap | -0.060 |
Token | , |
Token ID | 11 |
Feature activation | +9.40e-01 |
Pos logit contributions | |
pole | +0.001 |
isk | +0.001 |
ole | +0.001 |
Freddie | +0.001 |
py | +0.001 |
Neg logit contributions | |
roller | -0.001 |
steps | -0.001 |
palm | -0.001 |
squee | -0.001 |
wonder | -0.001 |
Token | " |
Token ID | 366 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
pole | +0.020 |
isk | +0.020 |
ole | +0.019 |
py | +0.019 |
ending | +0.018 |
Neg logit contributions | |
roller | -0.020 |
squee | -0.018 |
palm | -0.018 |
wonder | -0.018 |
scrap | -0.017 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.63e+00 |
Pos logit contributions | |
roller | +0.042 |
steps | +0.039 |
levers | +0.039 |
seats | +0.039 |
scrap | +0.038 |
Neg logit contributions | |
pole | -0.037 |
sharing | -0.036 |
isk | -0.036 |
Freddie | -0.034 |
Mickey | -0.033 |
Token | are |
Token ID | 389 |
Feature activation | +7.01e+00 |
Pos logit contributions | |
roller | +0.065 |
rooft | +0.062 |
levers | +0.061 |
scrap | +0.058 |
obs | +0.058 |
Neg logit contributions | |
sharing | -0.050 |
pole | -0.046 |
being | -0.045 |
isk | -0.043 |
Freddie | -0.043 |
Token | we |
Token ID | 356 |
Feature activation | +2.07e+00 |
Pos logit contributions | |
pole | +0.122 |
isk | +0.121 |
sharing | +0.102 |
ole | +0.102 |
ility | +0.101 |
Neg logit contributions | |
roller | -0.188 |
steps | -0.174 |
palm | -0.173 |
squee | -0.169 |
seats | -0.169 |
Token | going |
Token ID | 1016 |
Feature activation | +2.36e+00 |
Pos logit contributions | |
pole | +0.054 |
isk | +0.052 |
py | +0.051 |
Freddie | +0.047 |
ole | +0.047 |
Neg logit contributions | |
roller | -0.036 |
palm | -0.034 |
wonder | -0.034 |
da | -0.031 |
pe | -0.030 |
Token | today |
Token ID | 1909 |
Feature activation | -6.35e-01 |
Pos logit contributions | |
pole | +0.054 |
isk | +0.052 |
sharing | +0.051 |
Freddie | +0.050 |
ole | +0.050 |
Neg logit contributions | |
roller | -0.049 |
steps | -0.042 |
squee | -0.042 |
scrap | -0.042 |
palm | -0.041 |
Token | ?" |
Token ID | 1701 |
Feature activation | -2.00e+00 |
Pos logit contributions | |
roller | +0.016 |
steps | +0.014 |
palm | +0.014 |
squee | +0.014 |
seats | +0.014 |
Neg logit contributions | |
pole | -0.012 |
isk | -0.011 |
sharing | -0.011 |
Freddie | -0.011 |
ole | -0.011 |
Token | Her |
Token ID | 2332 |
Feature activation | +1.05e-02 |
Pos logit contributions | |
roller | +0.052 |
rooft | +0.047 |
levers | +0.047 |
scrap | +0.046 |
steps | +0.046 |
Neg logit contributions | |
sharing | -0.036 |
pole | -0.035 |
Mickey | -0.033 |
Freddie | -0.033 |
being | -0.033 |
Token | called |
Token ID | 1444 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
roller | +0.022 |
rooft | +0.020 |
levers | +0.019 |
steps | +0.019 |
scrap | +0.019 |
Neg logit contributions | |
pole | -0.026 |
sharing | -0.026 |
Freddie | -0.025 |
isk | -0.025 |
being | -0.024 |
Token | out |
Token ID | 503 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
roller | +0.003 |
steps | +0.003 |
scrap | +0.003 |
squee | +0.003 |
palm | +0.003 |
Neg logit contributions | |
pole | -0.003 |
isk | -0.003 |
sharing | -0.003 |
ole | -0.003 |
Mickey | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
sharing | +0.037 |
pole | +0.036 |
ole | +0.035 |
Freddie | +0.034 |
isk | +0.034 |
Neg logit contributions | |
roller | -0.052 |
scrap | -0.046 |
rooft | -0.046 |
squee | -0.045 |
levers | -0.045 |
Token | " |
Token ID | 366 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
roller | +0.003 |
scrap | +0.002 |
squee | +0.002 |
levers | +0.002 |
wonder | +0.002 |
Neg logit contributions | |
isk | -0.002 |
pole | -0.002 |
ole | -0.002 |
sharing | -0.002 |
py | -0.002 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.58e+00 |
Pos logit contributions | |
roller | +0.027 |
scrap | +0.023 |
palm | +0.023 |
rooft | +0.023 |
wonder | +0.022 |
Neg logit contributions | |
pole | -0.039 |
sharing | -0.039 |
isk | -0.038 |
ole | -0.038 |
Mickey | -0.037 |
Token | are |
Token ID | 389 |
Feature activation | +7.00e+00 |
Pos logit contributions | |
roller | +0.060 |
rooft | +0.053 |
scrap | +0.053 |
levers | +0.053 |
squee | +0.052 |
Neg logit contributions | |
sharing | -0.051 |
pole | -0.051 |
isk | -0.049 |
Freddie | -0.047 |
ole | -0.047 |
Token | you |
Token ID | 345 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
isk | +0.150 |
pole | +0.143 |
ility | +0.129 |
End | +0.125 |
py | +0.123 |
Neg logit contributions | |
roller | -0.170 |
steps | -0.153 |
umb | -0.150 |
squee | -0.147 |
seats | -0.147 |
Token | all |
Token ID | 477 |
Feature activation | +3.19e+00 |
Pos logit contributions | |
pole | +0.047 |
isk | +0.045 |
ole | +0.042 |
End | +0.040 |
penter | +0.040 |
Neg logit contributions | |
roller | -0.036 |
mo | -0.031 |
palm | -0.030 |
squee | -0.030 |
steps | -0.030 |
Token | ?" |
Token ID | 1701 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
pole | +0.064 |
isk | +0.064 |
ole | +0.061 |
End | +0.060 |
ingu | +0.057 |
Neg logit contributions | |
roller | -0.076 |
steps | -0.069 |
mo | -0.065 |
bars | -0.065 |
squee | -0.065 |
Token |
|
Token ID | 198 |
Feature activation | +2.79e-01 |
Pos logit contributions | |
roller | +0.033 |
scrap | +0.030 |
steps | +0.029 |
palm | +0.029 |
squee | +0.029 |
Neg logit contributions | |
pole | -0.032 |
isk | -0.031 |
sharing | -0.031 |
ole | -0.029 |
Freddie | -0.029 |
Token |
|
Token ID | 198 |
Feature activation | +1.69e+00 |
Pos logit contributions | |
pole | +0.005 |
sharing | +0.005 |
ole | +0.005 |
isk | +0.005 |
Freddie | +0.005 |
Neg logit contributions | |
roller | -0.007 |
levers | -0.006 |
squee | -0.006 |
scrap | -0.006 |
steps | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | +5.82e-01 |
Pos logit contributions | |
roller | +0.004 |
steps | +0.003 |
wonder | +0.003 |
seats | +0.003 |
squee | +0.003 |
Neg logit contributions | |
isk | -0.006 |
pole | -0.006 |
ole | -0.006 |
py | -0.006 |
Freddie | -0.006 |
Token | asked |
Token ID | 1965 |
Feature activation | +4.44e-01 |
Pos logit contributions | |
pole | +0.013 |
sharing | +0.013 |
isk | +0.012 |
Freddie | +0.012 |
being | +0.012 |
Neg logit contributions | |
roller | -0.012 |
rooft | -0.011 |
scrap | -0.010 |
levers | -0.010 |
squee | -0.010 |
Token | , |
Token ID | 11 |
Feature activation | -3.42e-01 |
Pos logit contributions | |
pole | +0.008 |
sharing | +0.008 |
isk | +0.008 |
Freddie | +0.008 |
ole | +0.008 |
Neg logit contributions | |
roller | -0.011 |
scrap | -0.010 |
squee | -0.010 |
steps | -0.010 |
rooft | -0.010 |
Token | " |
Token ID | 366 |
Feature activation | -1.95e+00 |
Pos logit contributions | |
roller | +0.008 |
palm | +0.007 |
squee | +0.007 |
wonder | +0.007 |
levers | +0.007 |
Neg logit contributions | |
isk | -0.007 |
pole | -0.007 |
py | -0.006 |
sharing | -0.006 |
ole | -0.006 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.83e+00 |
Pos logit contributions | |
roller | +0.041 |
steps | +0.038 |
scrap | +0.038 |
squee | +0.038 |
levers | +0.037 |
Neg logit contributions | |
pole | -0.042 |
sharing | -0.041 |
isk | -0.040 |
Freddie | -0.038 |
ole | -0.038 |
Token | are |
Token ID | 389 |
Feature activation | +6.99e+00 |
Pos logit contributions | |
roller | +0.067 |
rooft | +0.061 |
levers | +0.060 |
scrap | +0.060 |
steps | +0.058 |
Neg logit contributions | |
sharing | -0.055 |
pole | -0.054 |
isk | -0.051 |
Freddie | -0.050 |
being | -0.050 |
Token | we |
Token ID | 356 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
isk | +0.134 |
pole | +0.124 |
py | +0.115 |
ility | +0.114 |
End | +0.111 |
Neg logit contributions | |
roller | -0.187 |
seats | -0.167 |
steps | -0.166 |
umb | -0.160 |
squee | -0.159 |
Token | going |
Token ID | 1016 |
Feature activation | +2.30e+00 |
Pos logit contributions | |
isk | +0.049 |
py | +0.045 |
pole | +0.044 |
Freddie | +0.043 |
ole | +0.042 |
Neg logit contributions | |
roller | -0.042 |
buck | -0.037 |
steps | -0.035 |
seats | -0.035 |
wonder | -0.034 |
Token | ?" |
Token ID | 1701 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
pole | +0.050 |
isk | +0.049 |
sharing | +0.048 |
Freddie | +0.047 |
ole | +0.046 |
Neg logit contributions | |
roller | -0.051 |
steps | -0.044 |
squee | -0.043 |
scrap | -0.043 |
rooft | -0.043 |
Token |
|
Token ID | 198 |
Feature activation | -7.62e-02 |
Pos logit contributions | |
roller | +0.043 |
scrap | +0.038 |
squee | +0.037 |
palm | +0.037 |
steps | +0.037 |
Neg logit contributions | |
pole | -0.047 |
sharing | -0.046 |
isk | -0.046 |
ole | -0.044 |
Freddie | -0.044 |
Token |
|
Token ID | 198 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
roller | +0.002 |
scrap | +0.001 |
squee | +0.001 |
steps | +0.001 |
levers | +0.001 |
Neg logit contributions | |
pole | -0.002 |
sharing | -0.002 |
isk | -0.002 |
ole | -0.001 |
Freddie | -0.001 |
Token | his |
Token ID | 465 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
pole | +0.003 |
isk | +0.003 |
sharing | +0.003 |
Freddie | +0.003 |
ility | +0.003 |
Neg logit contributions | |
roller | -0.004 |
steps | -0.003 |
seats | -0.003 |
wonder | -0.003 |
levers | -0.003 |
Token | mom |
Token ID | 1995 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
sharing | +0.026 |
ole | +0.024 |
pole | +0.024 |
ending | +0.024 |
isk | +0.024 |
Neg logit contributions | |
roller | -0.037 |
rooft | -0.034 |
scrap | -0.034 |
bars | -0.033 |
levers | -0.033 |
Token | , |
Token ID | 11 |
Feature activation | +6.80e-01 |
Pos logit contributions | |
pole | +0.004 |
isk | +0.003 |
Freddie | +0.003 |
py | +0.003 |
ole | +0.003 |
Neg logit contributions | |
roller | -0.004 |
steps | -0.003 |
squee | -0.003 |
palm | -0.003 |
seats | -0.003 |
Token | " |
Token ID | 366 |
Feature activation | -2.16e+00 |
Pos logit contributions | |
pole | +0.016 |
isk | +0.015 |
py | +0.015 |
ole | +0.015 |
ility | +0.015 |
Neg logit contributions | |
roller | -0.014 |
palm | -0.012 |
squee | -0.012 |
wonder | -0.011 |
buck | -0.011 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
roller | +0.044 |
scrap | +0.040 |
steps | +0.039 |
squee | +0.039 |
levers | +0.038 |
Neg logit contributions | |
pole | -0.049 |
sharing | -0.048 |
isk | -0.047 |
ole | -0.045 |
Freddie | -0.045 |
Token | are |
Token ID | 389 |
Feature activation | +6.97e+00 |
Pos logit contributions | |
roller | +0.061 |
scrap | +0.054 |
squee | +0.053 |
steps | +0.053 |
levers | +0.052 |
Neg logit contributions | |
pole | -0.060 |
sharing | -0.059 |
isk | -0.058 |
Freddie | -0.056 |
ole | -0.056 |
Token | we |
Token ID | 356 |
Feature activation | +1.89e+00 |
Pos logit contributions | |
isk | +0.154 |
pole | +0.139 |
py | +0.131 |
ility | +0.128 |
ole | +0.119 |
Neg logit contributions | |
roller | -0.175 |
seats | -0.155 |
steps | -0.153 |
buck | -0.146 |
desks | -0.146 |
Token | going |
Token ID | 1016 |
Feature activation | +2.16e+00 |
Pos logit contributions | |
pole | +0.052 |
isk | +0.052 |
py | +0.051 |
ility | +0.046 |
ole | +0.046 |
Neg logit contributions | |
roller | -0.034 |
buck | -0.029 |
wonder | -0.027 |
steps | -0.026 |
seats | -0.026 |
Token | on |
Token ID | 319 |
Feature activation | +1.91e+00 |
Pos logit contributions | |
pole | +0.055 |
isk | +0.054 |
Freddie | +0.051 |
sharing | +0.050 |
ole | +0.050 |
Neg logit contributions | |
roller | -0.041 |
steps | -0.034 |
squee | -0.033 |
seats | -0.033 |
buck | -0.032 |
Token | vacation |
Token ID | 14600 |
Feature activation | +2.11e+00 |
Pos logit contributions | |
pole | +0.040 |
isk | +0.039 |
sharing | +0.038 |
ole | +0.037 |
Freddie | +0.036 |
Neg logit contributions | |
roller | -0.044 |
steps | -0.037 |
scrap | -0.037 |
squee | -0.037 |
rooft | -0.037 |
Token | ?" |
Token ID | 1701 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
pole | +0.040 |
isk | +0.039 |
sharing | +0.038 |
py | +0.037 |
Freddie | +0.037 |
Neg logit contributions | |
roller | -0.053 |
steps | -0.045 |
seats | -0.045 |
squee | -0.045 |
bars | -0.044 |
Token | and |
Token ID | 290 |
Feature activation | -2.13e-01 |
Pos logit contributions | |
roller | +0.018 |
steps | +0.017 |
bars | +0.017 |
squee | +0.016 |
scrap | +0.016 |
Neg logit contributions | |
pole | -0.019 |
isk | -0.019 |
ole | -0.018 |
sharing | -0.018 |
ator | -0.017 |
Token | said |
Token ID | 531 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
roller | +0.005 |
rooft | +0.005 |
squee | +0.005 |
steps | +0.005 |
leaf | +0.005 |
Neg logit contributions | |
pole | -0.004 |
sharing | -0.004 |
Mickey | -0.004 |
isk | -0.004 |
Freddie | -0.004 |
Token | , |
Token ID | 11 |
Feature activation | +8.74e-01 |
Pos logit contributions | |
roller | +0.028 |
scrap | +0.025 |
rooft | +0.025 |
levers | +0.025 |
steps | +0.025 |
Neg logit contributions | |
pole | -0.017 |
sharing | -0.017 |
isk | -0.016 |
Freddie | -0.016 |
ole | -0.015 |
Token | " |
Token ID | 366 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
pole | +0.013 |
sharing | +0.013 |
isk | +0.013 |
Freddie | +0.012 |
ole | +0.012 |
Neg logit contributions | |
roller | -0.024 |
scrap | -0.022 |
squee | -0.022 |
steps | -0.022 |
palm | -0.022 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.70e+00 |
Pos logit contributions | |
roller | +0.028 |
steps | +0.025 |
squee | +0.024 |
levers | +0.024 |
seats | +0.024 |
Neg logit contributions | |
pole | -0.029 |
isk | -0.029 |
sharing | -0.027 |
Freddie | -0.027 |
Mickey | -0.027 |
Token | are |
Token ID | 389 |
Feature activation | +6.96e+00 |
Pos logit contributions | |
roller | +0.060 |
scrap | +0.053 |
squee | +0.052 |
steps | +0.052 |
levers | +0.052 |
Neg logit contributions | |
pole | -0.056 |
sharing | -0.055 |
isk | -0.054 |
Freddie | -0.052 |
ole | -0.052 |
Token | you |
Token ID | 345 |
Feature activation | +1.89e+00 |
Pos logit contributions | |
pole | +0.146 |
isk | +0.133 |
ility | +0.124 |
ole | +0.123 |
ator | +0.123 |
Neg logit contributions | |
roller | -0.170 |
scrap | -0.158 |
squee | -0.156 |
steps | -0.156 |
pine | -0.152 |
Token | , |
Token ID | 11 |
Feature activation | +1.63e+00 |
Pos logit contributions | |
pole | +0.047 |
ole | +0.045 |
isk | +0.042 |
melon | +0.042 |
ters | +0.040 |
Neg logit contributions | |
roller | -0.036 |
wonder | -0.035 |
steps | -0.034 |
palm | -0.033 |
squee | -0.033 |
Token | Snow |
Token ID | 7967 |
Feature activation | +2.86e+00 |
Pos logit contributions | |
pole | +0.037 |
isk | +0.035 |
ole | +0.035 |
sharing | +0.035 |
melon | +0.034 |
Neg logit contributions | |
roller | -0.034 |
squee | -0.030 |
scrap | -0.030 |
palm | -0.029 |
lace | -0.029 |
Token | ball |
Token ID | 1894 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
End | +0.063 |
pole | +0.061 |
sharing | +0.059 |
ility | +0.057 |
being | +0.057 |
Neg logit contributions | |
roller | -0.052 |
squee | -0.046 |
obs | -0.045 |
bang | -0.044 |
zzy | -0.043 |
Token | ? |
Token ID | 30 |
Feature activation | +3.27e-01 |
Pos logit contributions | |
pole | +0.028 |
ole | +0.026 |
ility | +0.026 |
End | +0.025 |
sharing | +0.025 |
Neg logit contributions | |
roller | -0.021 |
scrap | -0.019 |
steps | -0.018 |
squee | -0.018 |
bars | -0.018 |
Token | named |
Token ID | 3706 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
ending | +0.012 |
pole | +0.012 |
isk | +0.011 |
ole | +0.011 |
sharing | +0.011 |
Neg logit contributions | |
roller | -0.010 |
palm | -0.009 |
squee | -0.009 |
pine | -0.009 |
bars | -0.009 |
Token | Lily |
Token ID | 20037 |
Feature activation | +2.02e+00 |
Pos logit contributions | |
ending | +0.030 |
pole | +0.030 |
End | +0.030 |
ole | +0.029 |
sharing | +0.027 |
Neg logit contributions | |
roller | -0.037 |
squee | -0.035 |
palm | -0.034 |
scrap | -0.033 |
pine | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | +6.33e-01 |
Pos logit contributions | |
ending | +0.057 |
pole | +0.056 |
ole | +0.055 |
involve | +0.055 |
isk | +0.054 |
Neg logit contributions | |
roller | -0.024 |
palm | -0.023 |
squee | -0.022 |
pine | -0.022 |
bars | -0.022 |
Token | " |
Token ID | 366 |
Feature activation | -2.76e+00 |
Pos logit contributions | |
pole | +0.013 |
py | +0.012 |
ole | +0.012 |
ator | +0.011 |
ility | +0.011 |
Neg logit contributions | |
roller | -0.014 |
squee | -0.014 |
scrap | -0.013 |
steps | -0.013 |
wonder | -0.013 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.82e+00 |
Pos logit contributions | |
roller | +0.053 |
scrap | +0.046 |
steps | +0.046 |
squee | +0.045 |
levers | +0.045 |
Neg logit contributions | |
pole | -0.066 |
sharing | -0.065 |
isk | -0.064 |
Freddie | -0.062 |
ole | -0.062 |
Token | are |
Token ID | 389 |
Feature activation | +6.95e+00 |
Pos logit contributions | |
roller | +0.062 |
scrap | +0.054 |
squee | +0.054 |
steps | +0.054 |
palm | +0.053 |
Neg logit contributions | |
pole | -0.061 |
sharing | -0.059 |
isk | -0.058 |
ole | -0.057 |
Freddie | -0.056 |
Token | you |
Token ID | 345 |
Feature activation | +1.71e+00 |
Pos logit contributions | |
isk | +0.141 |
pole | +0.138 |
End | +0.123 |
ility | +0.122 |
ending | +0.121 |
Neg logit contributions | |
roller | -0.165 |
steps | -0.155 |
squee | -0.154 |
da | -0.153 |
palm | -0.150 |
Token | going |
Token ID | 1016 |
Feature activation | +2.35e+00 |
Pos logit contributions | |
pole | +0.041 |
ole | +0.040 |
isk | +0.038 |
melon | +0.037 |
ters | +0.036 |
Neg logit contributions | |
roller | -0.033 |
wonder | -0.031 |
steps | -0.030 |
palm | -0.030 |
mo | -0.029 |
Token | , |
Token ID | 11 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
pole | +0.061 |
isk | +0.058 |
sharing | +0.056 |
End | +0.056 |
ole | +0.055 |
Neg logit contributions | |
roller | -0.042 |
steps | -0.036 |
squee | -0.035 |
bars | -0.034 |
palm | -0.034 |
Token | Ben |
Token ID | 3932 |
Feature activation | +2.35e+00 |
Pos logit contributions | |
isk | +0.024 |
pole | +0.024 |
ole | +0.023 |
ator | +0.022 |
ility | +0.022 |
Neg logit contributions | |
roller | -0.024 |
squee | -0.022 |
da | -0.021 |
wonder | -0.020 |
steps | -0.020 |
Token | ?" |
Token ID | 1701 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
pole | +0.057 |
End | +0.057 |
isk | +0.056 |
ility | +0.056 |
ole | +0.055 |
Neg logit contributions | |
roller | -0.038 |
squee | -0.036 |
scrap | -0.034 |
wonder | -0.033 |
pine | -0.031 |
Token | voice |
Token ID | 3809 |
Feature activation | -3.12e-01 |
Pos logit contributions | |
roller | +0.033 |
obs | +0.032 |
col | +0.031 |
bars | +0.031 |
desks | +0.031 |
Neg logit contributions | |
pole | -0.031 |
sharing | -0.031 |
ole | -0.030 |
Mickey | -0.030 |
ending | -0.029 |
Token | spoke |
Token ID | 5158 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
roller | +0.007 |
scrap | +0.006 |
rooft | +0.006 |
palm | +0.006 |
squee | +0.006 |
Neg logit contributions | |
pole | -0.006 |
sharing | -0.006 |
being | -0.006 |
ole | -0.006 |
isk | -0.006 |
Token | , |
Token ID | 11 |
Feature activation | +6.14e-01 |
Pos logit contributions | |
pole | +0.023 |
sharing | +0.023 |
ole | +0.022 |
isk | +0.022 |
Freddie | +0.021 |
Neg logit contributions | |
roller | -0.032 |
scrap | -0.028 |
squee | -0.028 |
palm | -0.028 |
bars | -0.028 |
Token | " |
Token ID | 366 |
Feature activation | -2.68e+00 |
Pos logit contributions | |
pole | +0.010 |
sharing | +0.010 |
isk | +0.010 |
ole | +0.009 |
Freddie | +0.009 |
Neg logit contributions | |
roller | -0.016 |
scrap | -0.015 |
squee | -0.014 |
steps | -0.014 |
levers | -0.014 |
Token | Where |
Token ID | 8496 |
Feature activation | -3.17e+00 |
Pos logit contributions | |
roller | +0.052 |
scrap | +0.044 |
squee | +0.044 |
steps | +0.044 |
levers | +0.043 |
Neg logit contributions | |
pole | -0.064 |
sharing | -0.063 |
isk | -0.062 |
ole | -0.060 |
Freddie | -0.060 |
Token | are |
Token ID | 389 |
Feature activation | +6.92e+00 |
Pos logit contributions | |
roller | +0.073 |
scrap | +0.064 |
rooft | +0.064 |
levers | +0.064 |
squee | +0.063 |
Neg logit contributions | |
pole | -0.064 |
sharing | -0.063 |
isk | -0.061 |
Freddie | -0.059 |
ole | -0.059 |
Token | you |
Token ID | 345 |
Feature activation | +1.62e+00 |
Pos logit contributions | |
isk | +0.138 |
pole | +0.133 |
End | +0.121 |
ending | +0.116 |
ility | +0.116 |
Neg logit contributions | |
roller | -0.170 |
steps | -0.159 |
umb | -0.153 |
seats | -0.152 |
palm | -0.150 |
Token | going |
Token ID | 1016 |
Feature activation | +2.68e+00 |
Pos logit contributions | |
pole | +0.035 |
isk | +0.034 |
ole | +0.033 |
Freddie | +0.031 |
Mickey | +0.030 |
Neg logit contributions | |
roller | -0.036 |
palm | -0.033 |
steps | -0.033 |
mo | -0.032 |
wonder | -0.031 |
Token | ?". |
Token ID | 43634 |
Feature activation | -1.98e+00 |
Pos logit contributions | |
pole | +0.057 |
isk | +0.055 |
sharing | +0.054 |
Freddie | +0.053 |
ole | +0.052 |
Neg logit contributions | |
roller | -0.060 |
steps | -0.053 |
squee | -0.052 |
scrap | -0.051 |
rooft | -0.051 |
Token | Jill |
Token ID | 22772 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
roller | +0.053 |
scrap | +0.048 |
squee | +0.046 |
bars | +0.046 |
seats | +0.045 |
Neg logit contributions | |
being | -0.032 |
sharing | -0.031 |
pole | -0.031 |
Mickey | -0.030 |
Freddie | -0.030 |
Token | was |
Token ID | 373 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
roller | +0.030 |
scrap | +0.026 |
squee | +0.026 |
steps | +0.026 |
levers | +0.025 |
Neg logit contributions | |
pole | -0.036 |
sharing | -0.035 |
isk | -0.035 |
Freddie | -0.034 |
ole | -0.034 |
Token | sad |
Token ID | 6507 |
Feature activation | +7.11e-02 |
Pos logit contributions | |
roller | +0.044 |
scrap | +0.040 |
steps | +0.040 |
wonder | +0.039 |
squee | +0.039 |
Neg logit contributions | |
pole | -0.057 |
isk | -0.056 |
sharing | -0.054 |
ole | -0.052 |
Freddie | -0.052 |
Token | and |
Token ID | 290 |
Feature activation | +7.59e-01 |
Pos logit contributions | |
pole | +0.001 |
isk | +0.001 |
sharing | +0.001 |
ole | +0.001 |
Freddie | +0.001 |
Neg logit contributions | |
roller | -0.002 |
steps | -0.002 |
squee | -0.002 |
scrap | -0.001 |
levers | -0.001 |
Token | asked |
Token ID | 1965 |
Feature activation | +1.67e-02 |
Pos logit contributions | |
pole | +0.016 |
isk | +0.016 |
sharing | +0.016 |
ole | +0.015 |
Freddie | +0.015 |
Neg logit contributions | |
roller | -0.017 |
scrap | -0.015 |
steps | -0.015 |
squee | -0.015 |
palm | -0.014 |
Token | " |
Token ID | 366 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
pole | +0.000 |
sharing | +0.000 |
isk | +0.000 |
ole | +0.000 |
Freddie | +0.000 |
Neg logit contributions | |
roller | -0.000 |
scrap | -0.000 |
steps | -0.000 |
squee | -0.000 |
levers | -0.000 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.92e+00 |
Pos logit contributions | |
roller | +0.037 |
scrap | +0.032 |
squee | +0.032 |
steps | +0.032 |
levers | +0.031 |
Neg logit contributions | |
pole | -0.041 |
sharing | -0.040 |
isk | -0.040 |
Freddie | -0.038 |
ole | -0.038 |
Token | are |
Token ID | 389 |
Feature activation | +6.91e+00 |
Pos logit contributions | |
roller | +0.065 |
scrap | +0.057 |
squee | +0.056 |
steps | +0.056 |
levers | +0.056 |
Neg logit contributions | |
pole | -0.061 |
sharing | -0.060 |
isk | -0.059 |
Freddie | -0.057 |
ole | -0.057 |
Token | you |
Token ID | 345 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
isk | +0.155 |
pole | +0.152 |
py | +0.134 |
ility | +0.133 |
End | +0.129 |
Neg logit contributions | |
roller | -0.163 |
steps | -0.146 |
umb | -0.145 |
seats | -0.142 |
squee | -0.139 |
Token | going |
Token ID | 1016 |
Feature activation | +2.37e+00 |
Pos logit contributions | |
pole | +0.040 |
isk | +0.039 |
ole | +0.036 |
End | +0.036 |
py | +0.035 |
Neg logit contributions | |
roller | -0.025 |
wonder | -0.021 |
steps | -0.021 |
buck | -0.020 |
squee | -0.020 |
Token | ?" |
Token ID | 1701 |
Feature activation | -2.39e+00 |
Pos logit contributions | |
pole | +0.055 |
isk | +0.054 |
sharing | +0.052 |
Freddie | +0.051 |
ole | +0.050 |
Neg logit contributions | |
roller | -0.049 |
steps | -0.042 |
squee | -0.041 |
rooft | -0.040 |
scrap | -0.040 |
Token | Mom |
Token ID | 11254 |
Feature activation | -1.77e+00 |
Pos logit contributions | |
roller | +0.054 |
scrap | +0.047 |
steps | +0.047 |
squee | +0.047 |
levers | +0.047 |
Neg logit contributions | |
pole | -0.050 |
sharing | -0.049 |
isk | -0.048 |
Freddie | -0.046 |
ole | -0.046 |
Token | my |
Token ID | 1820 |
Feature activation | -2.11e+00 |
Pos logit contributions | |
roller | +0.049 |
scrap | +0.046 |
squee | +0.046 |
wonder | +0.045 |
buck | +0.045 |
Neg logit contributions | |
pole | -0.024 |
isk | -0.023 |
ole | -0.023 |
sharing | -0.023 |
ending | -0.023 |
Token | the |
Token ID | 262 |
Feature activation | -6.05e-01 |
Pos logit contributions | |
roller | +0.012 |
rooft | +0.011 |
bars | +0.011 |
scrap | +0.011 |
levers | +0.011 |
Neg logit contributions | |
sharing | -0.008 |
pole | -0.007 |
Freddie | -0.007 |
Mickey | -0.007 |
isk | -0.007 |
Token | car |
Token ID | 1097 |
Feature activation | +3.96e-02 |
Pos logit contributions | |
roller | +0.013 |
scrap | +0.011 |
levers | +0.011 |
squee | +0.011 |
rooft | +0.011 |
Neg logit contributions | |
pole | -0.013 |
sharing | -0.013 |
isk | -0.012 |
ole | -0.012 |
Freddie | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | +2.95e-01 |
Pos logit contributions | |
isk | +0.001 |
sharing | +0.001 |
Freddie | +0.001 |
py | +0.001 |
pole | +0.001 |
Neg logit contributions | |
roller | -0.001 |
bars | -0.001 |
steps | -0.001 |
levers | -0.001 |
seats | -0.001 |
Token | " |
Token ID | 366 |
Feature activation | -3.07e+00 |
Pos logit contributions | |
pole | +0.006 |
sharing | +0.005 |
ole | +0.005 |
isk | +0.005 |
Mickey | +0.005 |
Neg logit contributions | |
roller | -0.007 |
scrap | -0.007 |
squee | -0.007 |
lace | -0.006 |
rooft | -0.006 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.74e+00 |
Pos logit contributions | |
steps | +0.067 |
roller | +0.065 |
squee | +0.064 |
scrap | +0.063 |
seats | +0.061 |
Neg logit contributions | |
pole | -0.067 |
sharing | -0.061 |
ole | -0.058 |
isk | -0.057 |
being | -0.055 |
Token | are |
Token ID | 389 |
Feature activation | +6.91e+00 |
Pos logit contributions | |
roller | +0.064 |
rooft | +0.059 |
scrap | +0.058 |
levers | +0.058 |
squee | +0.057 |
Neg logit contributions | |
sharing | -0.054 |
pole | -0.053 |
isk | -0.050 |
Freddie | -0.049 |
being | -0.049 |
Token | we |
Token ID | 356 |
Feature activation | +2.12e+00 |
Pos logit contributions | |
isk | +0.151 |
pole | +0.128 |
py | +0.128 |
ility | +0.126 |
ending | +0.120 |
Neg logit contributions | |
roller | -0.176 |
seats | -0.158 |
steps | -0.155 |
umb | -0.151 |
buck | -0.149 |
Token | going |
Token ID | 1016 |
Feature activation | +2.31e+00 |
Pos logit contributions | |
isk | +0.056 |
py | +0.054 |
pole | +0.052 |
Freddie | +0.052 |
acky | +0.050 |
Neg logit contributions | |
roller | -0.040 |
buck | -0.034 |
wonder | -0.033 |
palm | -0.032 |
mo | -0.031 |
Token | , |
Token ID | 11 |
Feature activation | +6.49e-01 |
Pos logit contributions | |
pole | +0.052 |
isk | +0.051 |
sharing | +0.050 |
Freddie | +0.049 |
ole | +0.048 |
Neg logit contributions | |
roller | -0.049 |
steps | -0.041 |
squee | -0.041 |
scrap | -0.041 |
rooft | -0.041 |
Token | Dad |
Token ID | 17415 |
Feature activation | +2.24e+00 |
Pos logit contributions | |
pole | +0.014 |
sharing | +0.014 |
isk | +0.014 |
ole | +0.013 |
Freddie | +0.013 |
Neg logit contributions | |
roller | -0.014 |
scrap | -0.012 |
squee | -0.012 |
steps | -0.012 |
rooft | -0.012 |
Token | ?" |
Token ID | 1701 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
End | +0.054 |
isk | +0.054 |
py | +0.054 |
Freddie | +0.052 |
ending | +0.052 |
Neg logit contributions | |
roller | -0.042 |
buck | -0.037 |
squee | -0.034 |
steps | -0.033 |
wonder | -0.031 |
Token | asking |
Token ID | 4737 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
roller | +0.014 |
scrap | +0.013 |
squee | +0.013 |
rooft | +0.013 |
lace | +0.013 |
Neg logit contributions | |
pole | -0.008 |
sharing | -0.008 |
being | -0.007 |
isk | -0.007 |
ole | -0.007 |
Token | him |
Token ID | 683 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
pole | +0.003 |
isk | +0.003 |
sharing | +0.003 |
ole | +0.003 |
melon | +0.003 |
Neg logit contributions | |
roller | -0.004 |
steps | -0.004 |
levers | -0.003 |
squee | -0.003 |
scrap | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | +5.62e-01 |
Pos logit contributions | |
pole | +0.026 |
isk | +0.025 |
sharing | +0.024 |
ole | +0.023 |
Freddie | +0.023 |
Neg logit contributions | |
roller | -0.031 |
steps | -0.028 |
palm | -0.028 |
squee | -0.028 |
levers | -0.028 |
Token | " |
Token ID | 366 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
pole | +0.010 |
isk | +0.010 |
sharing | +0.010 |
ole | +0.010 |
Freddie | +0.009 |
Neg logit contributions | |
roller | -0.014 |
squee | -0.012 |
scrap | -0.012 |
levers | -0.012 |
rooft | -0.012 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.93e+00 |
Pos logit contributions | |
roller | +0.037 |
scrap | +0.033 |
squee | +0.031 |
steps | +0.031 |
rooft | +0.031 |
Neg logit contributions | |
pole | -0.050 |
sharing | -0.050 |
isk | -0.048 |
ole | -0.048 |
melon | -0.047 |
Token | are |
Token ID | 389 |
Feature activation | +6.90e+00 |
Pos logit contributions | |
roller | +0.067 |
scrap | +0.059 |
rooft | +0.059 |
levers | +0.058 |
squee | +0.058 |
Neg logit contributions | |
pole | -0.059 |
sharing | -0.059 |
isk | -0.057 |
Freddie | -0.055 |
ole | -0.055 |
Token | you |
Token ID | 345 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
isk | +0.138 |
pole | +0.129 |
End | +0.114 |
ility | +0.113 |
sharing | +0.111 |
Neg logit contributions | |
roller | -0.173 |
steps | -0.165 |
seats | -0.156 |
scrap | -0.156 |
palm | -0.155 |
Token | going |
Token ID | 1016 |
Feature activation | +2.48e+00 |
Pos logit contributions | |
pole | +0.038 |
isk | +0.037 |
ole | +0.037 |
melon | +0.034 |
Freddie | +0.034 |
Neg logit contributions | |
roller | -0.034 |
palm | -0.032 |
wonder | -0.031 |
steps | -0.031 |
buck | -0.030 |
Token | ?". |
Token ID | 43634 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
pole | +0.054 |
isk | +0.052 |
sharing | +0.052 |
Freddie | +0.050 |
ole | +0.049 |
Neg logit contributions | |
roller | -0.054 |
steps | -0.048 |
squee | -0.047 |
scrap | -0.046 |
rooft | -0.046 |
Token | Jack |
Token ID | 3619 |
Feature activation | -7.47e-01 |
Pos logit contributions | |
roller | +0.044 |
squee | +0.040 |
scrap | +0.040 |
steps | +0.040 |
wonder | +0.040 |
Neg logit contributions | |
pole | -0.023 |
isk | -0.023 |
sharing | -0.021 |
ole | -0.021 |
py | -0.020 |
Token | looked |
Token ID | 3114 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
roller | +0.013 |
scrap | +0.012 |
squee | +0.011 |
palm | +0.011 |
steps | +0.011 |
Neg logit contributions | |
sharing | -0.019 |
pole | -0.019 |
isk | -0.018 |
ole | -0.018 |
Freddie | -0.017 |
Token | and |
Token ID | 290 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
roller | +0.014 |
palm | +0.012 |
scrap | +0.012 |
steps | +0.012 |
levers | +0.012 |
Neg logit contributions | |
pole | -0.015 |
sharing | -0.014 |
isk | -0.014 |
Freddie | -0.014 |
ole | -0.014 |
Token | said |
Token ID | 531 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
roller | +0.020 |
rooft | +0.017 |
steps | +0.017 |
scrap | +0.017 |
squee | +0.016 |
Neg logit contributions | |
pole | -0.024 |
sharing | -0.024 |
isk | -0.023 |
Freddie | -0.023 |
Mickey | -0.023 |
Token | , |
Token ID | 11 |
Feature activation | +5.61e-01 |
Pos logit contributions | |
roller | +0.035 |
scrap | +0.031 |
steps | +0.031 |
squee | +0.031 |
levers | +0.031 |
Neg logit contributions | |
pole | -0.022 |
sharing | -0.021 |
isk | -0.021 |
Freddie | -0.020 |
ole | -0.020 |
Token | " |
Token ID | 366 |
Feature activation | -1.99e+00 |
Pos logit contributions | |
isk | +0.011 |
pole | +0.011 |
ole | +0.011 |
sharing | +0.011 |
py | +0.010 |
Neg logit contributions | |
roller | -0.012 |
squee | -0.011 |
scrap | -0.011 |
steps | -0.011 |
rooft | -0.011 |
Token | Where |
Token ID | 8496 |
Feature activation | -3.06e+00 |
Pos logit contributions | |
roller | +0.038 |
scrap | +0.032 |
squee | +0.032 |
steps | +0.032 |
levers | +0.031 |
Neg logit contributions | |
pole | -0.049 |
sharing | -0.048 |
isk | -0.047 |
Freddie | -0.046 |
ole | -0.046 |
Token | are |
Token ID | 389 |
Feature activation | +6.89e+00 |
Pos logit contributions | |
roller | +0.070 |
rooft | +0.062 |
scrap | +0.062 |
levers | +0.062 |
squee | +0.061 |
Neg logit contributions | |
pole | -0.061 |
sharing | -0.061 |
isk | -0.059 |
Freddie | -0.057 |
ole | -0.057 |
Token | you |
Token ID | 345 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
pole | +0.135 |
isk | +0.133 |
ole | +0.118 |
ility | +0.116 |
End | +0.115 |
Neg logit contributions | |
roller | -0.167 |
steps | -0.160 |
scrap | -0.156 |
pine | -0.153 |
squee | -0.153 |
Token | going |
Token ID | 1016 |
Feature activation | +2.54e+00 |
Pos logit contributions | |
pole | +0.046 |
isk | +0.044 |
ole | +0.043 |
End | +0.041 |
ters | +0.040 |
Neg logit contributions | |
roller | -0.030 |
palm | -0.029 |
wonder | -0.029 |
steps | -0.028 |
pe | -0.028 |
Token | , |
Token ID | 11 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
pole | +0.059 |
isk | +0.056 |
sharing | +0.054 |
End | +0.053 |
Freddie | +0.053 |
Neg logit contributions | |
roller | -0.054 |
steps | -0.046 |
squee | -0.045 |
palm | -0.044 |
rooft | -0.043 |
Token | little |
Token ID | 1310 |
Feature activation | +2.22e+00 |
Pos logit contributions | |
pole | +0.022 |
ility | +0.022 |
isk | +0.021 |
imming | +0.021 |
ole | +0.021 |
Neg logit contributions | |
roller | -0.023 |
pine | -0.021 |
scrap | -0.021 |
wonder | -0.020 |
squee | -0.020 |
Token | girl |
Token ID | 2576 |
Feature activation | +2.29e+00 |
Pos logit contributions | |
pole | +0.050 |
isk | +0.049 |
sharing | +0.049 |
ole | +0.046 |
melon | +0.045 |
Neg logit contributions | |
roller | -0.046 |
steps | -0.041 |
scrap | -0.041 |
squee | -0.041 |
palm | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | +8.89e-01 |
Pos logit contributions | |
pole | +0.028 |
isk | +0.028 |
ole | +0.027 |
sharing | +0.027 |
Freddie | +0.027 |
Neg logit contributions | |
roller | -0.021 |
bars | -0.019 |
steps | -0.018 |
palm | -0.018 |
seats | -0.018 |
Token |
|
Token ID | 198 |
Feature activation | +5.74e-01 |
Pos logit contributions | |
pole | +0.017 |
sharing | +0.017 |
isk | +0.017 |
Freddie | +0.016 |
Mickey | +0.016 |
Neg logit contributions | |
roller | -0.021 |
steps | -0.019 |
levers | -0.019 |
scrap | -0.018 |
rooft | -0.018 |
Token |
|
Token ID | 198 |
Feature activation | +2.69e+00 |
Pos logit contributions | |
pole | +0.009 |
sharing | +0.008 |
isk | +0.008 |
Freddie | +0.008 |
ole | +0.008 |
Neg logit contributions | |
roller | -0.015 |
levers | -0.015 |
scrap | -0.015 |
palm | -0.014 |
rooft | -0.014 |
Token | " |
Token ID | 1 |
Feature activation | -1.92e+00 |
Pos logit contributions | |
pole | +0.048 |
sharing | +0.047 |
isk | +0.046 |
ole | +0.044 |
Freddie | +0.044 |
Neg logit contributions | |
roller | -0.068 |
scrap | -0.061 |
steps | -0.061 |
squee | -0.061 |
levers | -0.060 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.19e+00 |
Pos logit contributions | |
roller | +0.037 |
steps | +0.032 |
scrap | +0.032 |
squee | +0.031 |
levers | +0.031 |
Neg logit contributions | |
pole | -0.047 |
sharing | -0.045 |
isk | -0.045 |
Freddie | -0.043 |
Mickey | -0.043 |
Token | are |
Token ID | 389 |
Feature activation | +6.87e+00 |
Pos logit contributions | |
roller | +0.050 |
scrap | +0.044 |
rooft | +0.044 |
levers | +0.044 |
steps | +0.044 |
Neg logit contributions | |
pole | -0.044 |
sharing | -0.044 |
isk | -0.043 |
Freddie | -0.041 |
ole | -0.041 |
Token | we |
Token ID | 356 |
Feature activation | +2.07e+00 |
Pos logit contributions | |
pole | +0.131 |
isk | +0.123 |
ole | +0.112 |
sharing | +0.112 |
ility | +0.109 |
Neg logit contributions | |
roller | -0.179 |
squee | -0.164 |
steps | -0.164 |
pine | -0.163 |
scrap | -0.162 |
Token | going |
Token ID | 1016 |
Feature activation | +2.35e+00 |
Pos logit contributions | |
pole | +0.057 |
isk | +0.056 |
py | +0.054 |
Freddie | +0.052 |
ility | +0.051 |
Neg logit contributions | |
roller | -0.030 |
palm | -0.030 |
pe | -0.029 |
wonder | -0.029 |
pine | -0.027 |
Token | , |
Token ID | 11 |
Feature activation | +5.86e-01 |
Pos logit contributions | |
pole | +0.047 |
isk | +0.046 |
sharing | +0.045 |
Freddie | +0.044 |
ole | +0.043 |
Neg logit contributions | |
roller | -0.056 |
steps | -0.048 |
squee | -0.048 |
scrap | -0.048 |
palm | -0.048 |
Token | mom |
Token ID | 1995 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
pole | +0.013 |
isk | +0.012 |
ole | +0.012 |
sharing | +0.012 |
melon | +0.012 |
Neg logit contributions | |
roller | -0.013 |
squee | -0.011 |
seats | -0.011 |
scrap | -0.011 |
lace | -0.011 |
Token | ?" |
Token ID | 1701 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
End | +0.038 |
isk | +0.037 |
Freddie | +0.037 |
ending | +0.036 |
ole | +0.036 |
Neg logit contributions | |
roller | -0.032 |
buck | -0.029 |
squee | -0.027 |
seats | -0.026 |
pine | -0.026 |
Token | . |
Token ID | 13 |
Feature activation | +3.40e-01 |
Pos logit contributions | |
pole | +0.011 |
sharing | +0.011 |
isk | +0.011 |
Freddie | +0.010 |
ole | +0.010 |
Neg logit contributions | |
roller | -0.011 |
steps | -0.010 |
levers | -0.010 |
squee | -0.010 |
scrap | -0.010 |
Token |
|
Token ID | 198 |
Feature activation | +6.60e-01 |
Pos logit contributions | |
pole | +0.006 |
Mickey | +0.006 |
sharing | +0.006 |
Freddie | +0.005 |
isk | +0.005 |
Neg logit contributions | |
roller | -0.008 |
scrap | -0.008 |
palm | -0.008 |
ti | -0.008 |
squee | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | +2.58e+00 |
Pos logit contributions | |
pole | +0.009 |
Freddie | +0.009 |
sharing | +0.008 |
isk | +0.008 |
Mickey | +0.008 |
Neg logit contributions | |
roller | -0.018 |
scrap | -0.018 |
squee | -0.017 |
palm | -0.017 |
lace | -0.017 |
Token | " |
Token ID | 1 |
Feature activation | -2.29e+00 |
Pos logit contributions | |
pole | +0.044 |
sharing | +0.042 |
being | +0.039 |
isk | +0.039 |
Mickey | +0.039 |
Neg logit contributions | |
roller | -0.061 |
scrap | -0.060 |
squee | -0.060 |
steps | -0.058 |
palm | -0.058 |
Token | Where |
Token ID | 8496 |
Feature activation | -2.57e+00 |
Pos logit contributions | |
roller | +0.049 |
steps | +0.046 |
palm | +0.044 |
squee | +0.044 |
scrap | +0.044 |
Neg logit contributions | |
pole | -0.049 |
sharing | -0.047 |
isk | -0.047 |
Freddie | -0.045 |
being | -0.044 |
Token | are |
Token ID | 389 |
Feature activation | +6.86e+00 |
Pos logit contributions | |
roller | +0.061 |
rooft | +0.056 |
levers | +0.054 |
scrap | +0.054 |
palm | +0.053 |
Neg logit contributions | |
sharing | -0.050 |
pole | -0.049 |
isk | -0.047 |
being | -0.045 |
Freddie | -0.045 |
Token | you |
Token ID | 345 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
isk | +0.119 |
pole | +0.117 |
ility | +0.100 |
ole | +0.099 |
sharing | +0.098 |
Neg logit contributions | |
roller | -0.192 |
steps | -0.173 |
seats | -0.171 |
squee | -0.169 |
scrap | -0.168 |
Token | going |
Token ID | 1016 |
Feature activation | +2.60e+00 |
Pos logit contributions | |
pole | +0.041 |
isk | +0.038 |
ole | +0.036 |
Freddie | +0.035 |
py | +0.034 |
Neg logit contributions | |
roller | -0.038 |
buck | -0.032 |
mo | -0.032 |
steps | -0.032 |
squee | -0.032 |
Token | , |
Token ID | 11 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
pole | +0.057 |
isk | +0.055 |
sharing | +0.054 |
Freddie | +0.052 |
ole | +0.052 |
Neg logit contributions | |
roller | -0.058 |
steps | -0.049 |
squee | -0.049 |
scrap | -0.048 |
rooft | -0.047 |
Token | Mister |
Token ID | 43438 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
pole | +0.030 |
sharing | +0.029 |
isk | +0.028 |
being | +0.027 |
ole | +0.027 |
Neg logit contributions | |
roller | -0.030 |
scrap | -0.029 |
squee | -0.027 |
rooft | -0.027 |
palm | -0.027 |
Token | Tom |
Token ID | 4186 |
Feature activation | +2.38e+00 |
Pos logit contributions | |
pole | +0.033 |
sharing | +0.032 |
isk | +0.032 |
ole | +0.030 |
Freddie | +0.030 |
Neg logit contributions | |
roller | -0.050 |
scrap | -0.046 |
steps | -0.045 |
squee | -0.045 |
levers | -0.045 |
Token | When |
Token ID | 1649 |
Feature activation | -6.14e-01 |
Pos logit contributions | |
roller | +0.008 |
squee | +0.007 |
rooft | +0.007 |
steps | +0.007 |
seats | +0.007 |
Neg logit contributions | |
pole | -0.009 |
sharing | -0.008 |
isk | -0.008 |
py | -0.008 |
ole | -0.008 |
Token | she |
Token ID | 673 |
Feature activation | +4.58e-01 |
Pos logit contributions | |
roller | +0.014 |
steps | +0.012 |
scrap | +0.012 |
squee | +0.012 |
palm | +0.012 |
Neg logit contributions | |
pole | -0.013 |
isk | -0.013 |
ole | -0.012 |
sharing | -0.012 |
py | -0.012 |
Token | sat |
Token ID | 3332 |
Feature activation | -1.94e-01 |
Pos logit contributions | |
pole | +0.010 |
sharing | +0.010 |
isk | +0.010 |
Freddie | +0.009 |
ole | +0.009 |
Neg logit contributions | |
roller | -0.010 |
levers | -0.009 |
rooft | -0.009 |
scrap | -0.009 |
steps | -0.009 |
Token | in |
Token ID | 287 |
Feature activation | -9.84e-01 |
Pos logit contributions | |
roller | +0.005 |
levers | +0.004 |
rooft | +0.004 |
lace | +0.004 |
bars | +0.004 |
Neg logit contributions | |
sharing | -0.004 |
pole | -0.004 |
ole | -0.004 |
isk | -0.003 |
Freddie | -0.003 |
Token | the |
Token ID | 262 |
Feature activation | -2.50e+00 |
Pos logit contributions | |
roller | +0.025 |
levers | +0.022 |
rooft | +0.022 |
scrap | +0.022 |
obs | +0.022 |
Neg logit contributions | |
sharing | -0.017 |
pole | -0.016 |
isk | -0.016 |
Freddie | -0.016 |
being | -0.016 |
Token | bar |
Token ID | 2318 |
Feature activation | -7.20e+00 |
Pos logit contributions | |
levers | +0.041 |
roller | +0.039 |
col | +0.038 |
bang | +0.038 |
leaf | +0.038 |
Neg logit contributions | |
end | -0.059 |
being | -0.057 |
End | -0.057 |
sharing | -0.054 |
ending | -0.054 |
Token | ber |
Token ID | 527 |
Feature activation | -4.61e+00 |
Pos logit contributions | |
roller | +0.151 |
scrap | +0.141 |
rooft | +0.138 |
squee | +0.134 |
lace | +0.132 |
Neg logit contributions | |
pole | -0.150 |
sharing | -0.149 |
ole | -0.144 |
being | -0.142 |
isk | -0.140 |
Token | chair |
Token ID | 5118 |
Feature activation | -3.55e+00 |
Pos logit contributions | |
roller | +0.109 |
scrap | +0.097 |
squee | +0.096 |
rooft | +0.096 |
steps | +0.096 |
Neg logit contributions | |
pole | -0.089 |
sharing | -0.087 |
isk | -0.086 |
Freddie | -0.082 |
ole | -0.082 |
Token | , |
Token ID | 11 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
roller | +0.086 |
steps | +0.076 |
scrap | +0.074 |
squee | +0.074 |
bars | +0.074 |
Neg logit contributions | |
pole | -0.068 |
sharing | -0.067 |
isk | -0.066 |
ole | -0.063 |
Freddie | -0.063 |
Token | she |
Token ID | 673 |
Feature activation | +3.28e-01 |
Pos logit contributions | |
roller | +0.035 |
scrap | +0.031 |
squee | +0.031 |
steps | +0.031 |
rooft | +0.031 |
Neg logit contributions | |
pole | -0.031 |
sharing | -0.031 |
isk | -0.030 |
Freddie | -0.029 |
ole | -0.029 |
Token | was |
Token ID | 373 |
Feature activation | -9.83e-01 |
Pos logit contributions | |
pole | +0.007 |
sharing | +0.007 |
isk | +0.007 |
Freddie | +0.007 |
being | +0.007 |
Neg logit contributions | |
roller | -0.007 |
rooft | -0.006 |
levers | -0.006 |
palm | -0.006 |
steps | -0.006 |
Token | through |
Token ID | 832 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
sharing | +0.056 |
pole | +0.055 |
isk | +0.055 |
Freddie | +0.053 |
ole | +0.052 |
Neg logit contributions | |
roller | -0.046 |
steps | -0.040 |
palm | -0.039 |
rooft | -0.039 |
squee | -0.039 |
Token | the |
Token ID | 262 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
roller | +0.038 |
scrap | +0.034 |
squee | +0.034 |
levers | +0.033 |
lace | +0.033 |
Neg logit contributions | |
sharing | -0.023 |
pole | -0.023 |
being | -0.022 |
Freddie | -0.021 |
isk | -0.021 |
Token | sky |
Token ID | 6766 |
Feature activation | +6.11e-02 |
Pos logit contributions | |
roller | +0.037 |
scrap | +0.033 |
levers | +0.033 |
squee | +0.033 |
lace | +0.032 |
Neg logit contributions | |
pole | -0.035 |
ole | -0.033 |
sharing | -0.033 |
End | -0.032 |
ending | -0.032 |
Token | and |
Token ID | 290 |
Feature activation | -4.39e-01 |
Pos logit contributions | |
isk | +0.001 |
pole | +0.001 |
sharing | +0.001 |
Freddie | +0.001 |
Mickey | +0.001 |
Neg logit contributions | |
roller | -0.001 |
steps | -0.001 |
palm | -0.001 |
rooft | -0.001 |
squee | -0.001 |
Token | measuring |
Token ID | 15964 |
Feature activation | -5.33e-01 |
Pos logit contributions | |
roller | +0.010 |
scrap | +0.009 |
squee | +0.009 |
steps | +0.009 |
levers | +0.009 |
Neg logit contributions | |
pole | -0.009 |
sharing | -0.008 |
isk | -0.008 |
Freddie | -0.008 |
ole | -0.008 |
Token | how |
Token ID | 703 |
Feature activation | -6.75e+00 |
Pos logit contributions | |
roller | +0.013 |
scrap | +0.012 |
squee | +0.011 |
rooft | +0.011 |
leaf | +0.011 |
Neg logit contributions | |
pole | -0.010 |
sharing | -0.010 |
ole | -0.009 |
being | -0.009 |
isk | -0.009 |
Token | far |
Token ID | 1290 |
Feature activation | +2.16e+00 |
Pos logit contributions | |
roller | +0.133 |
lace | +0.115 |
scrap | +0.114 |
bisc | +0.114 |
squee | +0.112 |
Neg logit contributions | |
pole | -0.157 |
sharing | -0.154 |
ole | -0.148 |
being | -0.147 |
Freddie | -0.142 |
Token | it |
Token ID | 340 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
isk | +0.054 |
sharing | +0.053 |
pole | +0.053 |
ole | +0.050 |
Freddie | +0.049 |
Neg logit contributions | |
roller | -0.040 |
steps | -0.034 |
scrap | -0.034 |
levers | -0.034 |
squee | -0.034 |
Token | could |
Token ID | 714 |
Feature activation | -2.07e+00 |
Pos logit contributions | |
sharing | +0.033 |
isk | +0.032 |
ole | +0.032 |
Freddie | +0.032 |
pole | +0.032 |
Neg logit contributions | |
roller | -0.023 |
steps | -0.022 |
scrap | -0.021 |
squee | -0.020 |
wonder | -0.020 |
Token | go |
Token ID | 467 |
Feature activation | +2.75e+00 |
Pos logit contributions | |
roller | +0.049 |
steps | +0.044 |
palm | +0.044 |
wonder | +0.044 |
bars | +0.043 |
Neg logit contributions | |
sharing | -0.039 |
isk | -0.038 |
Freddie | -0.038 |
pole | -0.038 |
ole | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | -5.29e-01 |
Pos logit contributions | |
isk | +0.067 |
sharing | +0.066 |
ingu | +0.064 |
Freddie | +0.063 |
pole | +0.062 |
Neg logit contributions | |
roller | -0.053 |
steps | -0.047 |
rooft | -0.046 |
bars | -0.045 |
palm | -0.044 |
Token | her |
Token ID | 607 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
pole | +0.005 |
isk | +0.005 |
sharing | +0.005 |
ole | +0.005 |
Freddie | +0.005 |
Neg logit contributions | |
roller | -0.005 |
scrap | -0.005 |
palm | -0.005 |
steps | -0.005 |
squee | -0.005 |
Token | mom |
Token ID | 1995 |
Feature activation | -2.22e+00 |
Pos logit contributions | |
roller | +0.038 |
col | +0.036 |
levers | +0.035 |
leaf | +0.035 |
obs | +0.035 |
Neg logit contributions | |
sharing | -0.021 |
ending | -0.020 |
being | -0.019 |
pole | -0.019 |
end | -0.018 |
Token | both |
Token ID | 1111 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
roller | +0.047 |
scrap | +0.041 |
squee | +0.040 |
steps | +0.040 |
rooft | +0.040 |
Neg logit contributions | |
pole | -0.050 |
sharing | -0.049 |
isk | -0.048 |
Freddie | -0.046 |
ole | -0.045 |
Token | laughed |
Token ID | 13818 |
Feature activation | +1.76e-01 |
Pos logit contributions | |
roller | +0.027 |
rooft | +0.023 |
scrap | +0.023 |
levers | +0.023 |
squee | +0.023 |
Neg logit contributions | |
pole | -0.032 |
sharing | -0.031 |
isk | -0.030 |
Freddie | -0.030 |
ole | -0.029 |
Token | at |
Token ID | 379 |
Feature activation | -2.30e+00 |
Pos logit contributions | |
isk | +0.004 |
pole | +0.004 |
ole | +0.004 |
sharing | +0.004 |
Mickey | +0.004 |
Neg logit contributions | |
roller | -0.004 |
squee | -0.003 |
wonder | -0.003 |
steps | -0.003 |
levers | -0.003 |
Token | how |
Token ID | 703 |
Feature activation | -6.70e+00 |
Pos logit contributions | |
roller | +0.065 |
scrap | +0.059 |
obs | +0.058 |
lace | +0.058 |
levers | +0.058 |
Neg logit contributions | |
sharing | -0.035 |
being | -0.033 |
pole | -0.032 |
Freddie | -0.032 |
isk | -0.031 |
Token | much |
Token ID | 881 |
Feature activation | -7.24e-01 |
Pos logit contributions | |
roller | +0.146 |
lace | +0.124 |
levers | +0.123 |
scrap | +0.122 |
rooft | +0.120 |
Neg logit contributions | |
sharing | -0.146 |
pole | -0.145 |
Freddie | -0.140 |
ole | -0.138 |
being | -0.137 |
Token | the |
Token ID | 262 |
Feature activation | +2.69e-01 |
Pos logit contributions | |
roller | +0.018 |
steps | +0.017 |
squee | +0.016 |
wonder | +0.016 |
seats | +0.016 |
Neg logit contributions | |
isk | -0.014 |
pole | -0.014 |
sharing | -0.013 |
ole | -0.012 |
ingu | -0.012 |
Token | bee |
Token ID | 20697 |
Feature activation | -7.57e-01 |
Pos logit contributions | |
pole | +0.006 |
sharing | +0.006 |
being | +0.006 |
ending | +0.005 |
Freddie | +0.005 |
Neg logit contributions | |
roller | -0.006 |
levers | -0.005 |
squee | -0.005 |
scrap | -0.005 |
rooft | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | -7.50e-01 |
Pos logit contributions | |
roller | +0.014 |
steps | +0.012 |
wonder | +0.012 |
scrap | +0.012 |
squee | +0.012 |
Neg logit contributions | |
isk | -0.018 |
pole | -0.018 |
sharing | -0.018 |
ole | -0.018 |
Mickey | -0.017 |
Token | flying |
Token ID | 7348 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
roller | +0.016 |
steps | +0.014 |
palm | +0.014 |
wonder | +0.014 |
bars | +0.014 |
Neg logit contributions | |
isk | -0.016 |
pole | -0.016 |
ole | -0.015 |
ingu | -0.015 |
Mickey | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | -1.84e-01 |
Pos logit contributions | |
roller | +0.014 |
steps | +0.013 |
bars | +0.013 |
palm | +0.013 |
scrap | +0.013 |
Neg logit contributions | |
pole | -0.009 |
isk | -0.009 |
sharing | -0.009 |
ole | -0.009 |
Freddie | -0.008 |
Token | It |
Token ID | 632 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
roller | +0.004 |
steps | +0.004 |
scrap | +0.004 |
squee | +0.004 |
palm | +0.004 |
Neg logit contributions | |
pole | -0.004 |
sharing | -0.004 |
isk | -0.004 |
ole | -0.004 |
Freddie | -0.004 |
Token | jumped |
Token ID | 11687 |
Feature activation | +2.52e-01 |
Pos logit contributions | |
isk | +0.026 |
pole | +0.026 |
sharing | +0.024 |
Freddie | +0.024 |
being | +0.024 |
Neg logit contributions | |
roller | -0.031 |
rooft | -0.027 |
obs | -0.027 |
levers | -0.027 |
leaf | -0.026 |
Token | onto |
Token ID | 4291 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
pole | +0.005 |
sharing | +0.005 |
isk | +0.005 |
ole | +0.004 |
Freddie | +0.004 |
Neg logit contributions | |
roller | -0.006 |
scrap | -0.005 |
levers | -0.005 |
palm | -0.005 |
squee | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -1.13e+00 |
Pos logit contributions | |
roller | +0.031 |
steps | +0.027 |
palm | +0.027 |
squee | +0.027 |
scrap | +0.027 |
Neg logit contributions | |
pole | -0.028 |
sharing | -0.028 |
isk | -0.027 |
ole | -0.026 |
Freddie | -0.025 |
Token | bar |
Token ID | 2318 |
Feature activation | -6.68e+00 |
Pos logit contributions | |
col | +0.021 |
wonder | +0.019 |
icol | +0.019 |
zzy | +0.019 |
roller | +0.018 |
Neg logit contributions | |
being | -0.026 |
end | -0.025 |
pole | -0.024 |
sharing | -0.024 |
ter | -0.023 |
Token | ber |
Token ID | 527 |
Feature activation | -3.64e+00 |
Pos logit contributions | |
roller | +0.149 |
scrap | +0.145 |
squee | +0.138 |
umb | +0.138 |
rooft | +0.136 |
Neg logit contributions | |
pole | -0.132 |
ole | -0.124 |
sharing | -0.124 |
isk | -0.119 |
being | -0.118 |
Token | â |
Token ID | 22940 |
Feature activation | +4.83e-01 |
Pos logit contributions | |
roller | +0.095 |
scrap | +0.086 |
squee | +0.085 |
rooft | +0.085 |
levers | +0.085 |
Neg logit contributions | |
pole | -0.061 |
sharing | -0.059 |
isk | -0.058 |
Freddie | -0.055 |
ole | -0.055 |
Token | € |
Token ID | 26391 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
sharing | +0.011 |
pole | +0.011 |
ole | +0.010 |
Freddie | +0.010 |
ingu | +0.010 |
Neg logit contributions | |
roller | -0.010 |
bars | -0.009 |
scrap | -0.008 |
squee | -0.008 |
lace | -0.008 |
Token | ™ |
Token ID | 8151 |
Feature activation | -4.50e+00 |
Pos logit contributions | |
roller | +0.031 |
levers | +0.028 |
scrap | +0.028 |
squee | +0.028 |
rooft | +0.027 |
Neg logit contributions | |
sharing | -0.036 |
isk | -0.036 |
pole | -0.036 |
Freddie | -0.034 |
being | -0.033 |
Token | s |
Token ID | 82 |
Feature activation | -2.44e+00 |
Pos logit contributions | |
roller | +0.092 |
palm | +0.081 |
steps | +0.079 |
bars | +0.079 |
scrap | +0.076 |
Neg logit contributions | |
isk | -0.098 |
pole | -0.097 |
sharing | -0.097 |
ending | -0.093 |
Freddie | -0.093 |
Token | down |
Token ID | 866 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
pole | +0.027 |
sharing | +0.027 |
isk | +0.027 |
Freddie | +0.026 |
ole | +0.026 |
Neg logit contributions | |
roller | -0.028 |
scrap | -0.024 |
steps | -0.024 |
squee | -0.024 |
levers | -0.024 |
Token | and |
Token ID | 290 |
Feature activation | -8.10e-01 |
Pos logit contributions | |
roller | +0.034 |
scrap | +0.031 |
levers | +0.029 |
squee | +0.029 |
palm | +0.029 |
Neg logit contributions | |
pole | -0.035 |
sharing | -0.035 |
isk | -0.033 |
ole | -0.033 |
Freddie | -0.033 |
Token | perched |
Token ID | 49264 |
Feature activation | -5.06e-01 |
Pos logit contributions | |
roller | +0.017 |
levers | +0.014 |
rooft | +0.014 |
steps | +0.014 |
bars | +0.014 |
Neg logit contributions | |
pole | -0.018 |
sharing | -0.018 |
Freddie | -0.017 |
isk | -0.017 |
Mickey | -0.017 |
Token | atop |
Token ID | 20156 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
roller | +0.011 |
scrap | +0.009 |
squee | +0.009 |
steps | +0.009 |
levers | +0.009 |
Neg logit contributions | |
pole | -0.011 |
sharing | -0.011 |
isk | -0.011 |
ole | -0.010 |
Freddie | -0.010 |
Token | a |
Token ID | 257 |
Feature activation | -5.29e+00 |
Pos logit contributions | |
roller | +0.050 |
scrap | +0.045 |
squee | +0.044 |
steps | +0.044 |
levers | +0.043 |
Neg logit contributions | |
pole | -0.040 |
sharing | -0.040 |
Freddie | -0.038 |
isk | -0.037 |
ole | -0.037 |
Token | blade |
Token ID | 11865 |
Feature activation | -6.59e+00 |
Pos logit contributions | |
roller | +0.093 |
scrap | +0.088 |
squee | +0.086 |
desks | +0.085 |
obs | +0.085 |
Neg logit contributions | |
pole | -0.132 |
ole | -0.123 |
sharing | -0.118 |
end | -0.117 |
Freddie | -0.117 |
Token | of |
Token ID | 286 |
Feature activation | -1.67e+00 |
Pos logit contributions | |
roller | +0.170 |
scrap | +0.156 |
squee | +0.154 |
rooft | +0.153 |
levers | +0.150 |
Neg logit contributions | |
pole | -0.122 |
sharing | -0.110 |
ole | -0.105 |
Freddie | -0.100 |
being | -0.100 |
Token | grass |
Token ID | 8701 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
roller | +0.034 |
op | +0.033 |
obs | +0.033 |
zzy | +0.033 |
Jo | +0.032 |
Neg logit contributions | |
end | -0.034 |
pole | -0.033 |
being | -0.032 |
sharing | -0.031 |
End | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | -7.24e-01 |
Pos logit contributions | |
roller | +0.023 |
steps | +0.020 |
palm | +0.020 |
squee | +0.020 |
levers | +0.020 |
Neg logit contributions | |
pole | -0.022 |
isk | -0.022 |
sharing | -0.022 |
Freddie | -0.021 |
Mickey | -0.021 |
Token | It |
Token ID | 632 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
roller | +0.015 |
scrap | +0.013 |
squee | +0.013 |
steps | +0.013 |
levers | +0.012 |
Neg logit contributions | |
pole | -0.017 |
sharing | -0.016 |
isk | -0.016 |
ole | -0.016 |
Freddie | -0.016 |
Token | looked |
Token ID | 3114 |
Feature activation | +1.57e+00 |
Pos logit contributions | |
pole | +0.024 |
isk | +0.022 |
sharing | +0.022 |
Freddie | +0.021 |
being | +0.021 |
Neg logit contributions | |
roller | -0.024 |
rooft | -0.020 |
desks | -0.020 |
levers | -0.020 |
squee | -0.020 |
Token | led |
Token ID | 992 |
Feature activation | +1.27e-01 |
Pos logit contributions | |
roller | +0.053 |
rooft | +0.046 |
squee | +0.046 |
scrap | +0.046 |
seats | +0.045 |
Neg logit contributions | |
pole | -0.035 |
sharing | -0.034 |
ole | -0.033 |
isk | -0.033 |
Freddie | -0.032 |
Token | up |
Token ID | 510 |
Feature activation | +3.55e-01 |
Pos logit contributions | |
sharing | +0.002 |
Freddie | +0.002 |
pole | +0.002 |
being | +0.002 |
Mickey | +0.002 |
Neg logit contributions | |
roller | -0.003 |
rooft | -0.003 |
seats | -0.003 |
scrap | -0.003 |
leaf | -0.003 |
Token | her |
Token ID | 607 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
sharing | +0.006 |
being | +0.005 |
ole | +0.005 |
Freddie | +0.005 |
pole | +0.005 |
Neg logit contributions | |
roller | -0.010 |
leaf | -0.008 |
levers | -0.008 |
scrap | -0.008 |
obs | -0.008 |
Token | nose |
Token ID | 9686 |
Feature activation | -3.08e+00 |
Pos logit contributions | |
roller | +0.042 |
rooft | +0.035 |
scrap | +0.035 |
squee | +0.035 |
steps | +0.035 |
Neg logit contributions | |
pole | -0.048 |
sharing | -0.048 |
isk | -0.046 |
ole | -0.046 |
Freddie | -0.046 |
Token | at |
Token ID | 379 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
roller | +0.073 |
scrap | +0.064 |
rooft | +0.064 |
squee | +0.064 |
steps | +0.063 |
Neg logit contributions | |
pole | -0.061 |
sharing | -0.059 |
isk | -0.057 |
Freddie | -0.055 |
ole | -0.055 |
Token | how |
Token ID | 703 |
Feature activation | -6.56e+00 |
Pos logit contributions | |
roller | +0.032 |
scrap | +0.029 |
squee | +0.028 |
levers | +0.028 |
steps | +0.028 |
Neg logit contributions | |
pole | -0.027 |
sharing | -0.027 |
isk | -0.026 |
Freddie | -0.025 |
ole | -0.025 |
Token | it |
Token ID | 340 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
roller | +0.168 |
levers | +0.147 |
rooft | +0.146 |
seats | +0.145 |
bars | +0.145 |
Neg logit contributions | |
sharing | -0.119 |
being | -0.115 |
Freddie | -0.109 |
ole | -0.109 |
melon | -0.104 |
Token | smelled |
Token ID | 37609 |
Feature activation | +1.25e-02 |
Pos logit contributions | |
pole | +0.032 |
sharing | +0.031 |
isk | +0.031 |
ole | +0.030 |
Freddie | +0.030 |
Neg logit contributions | |
roller | -0.028 |
scrap | -0.024 |
squee | -0.024 |
steps | -0.024 |
levers | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | -5.45e-01 |
Pos logit contributions | |
isk | +0.000 |
pole | +0.000 |
py | +0.000 |
sharing | +0.000 |
ole | +0.000 |
Neg logit contributions | |
roller | -0.000 |
squee | -0.000 |
scrap | -0.000 |
steps | -0.000 |
wonder | -0.000 |
Token | But |
Token ID | 887 |
Feature activation | -3.53e-01 |
Pos logit contributions | |
roller | +0.013 |
scrap | +0.011 |
levers | +0.011 |
steps | +0.011 |
palm | +0.011 |
Neg logit contributions | |
pole | -0.011 |
sharing | -0.011 |
isk | -0.011 |
Freddie | -0.011 |
Mickey | -0.010 |
Token | then |
Token ID | 788 |
Feature activation | +7.47e-01 |
Pos logit contributions | |
roller | +0.008 |
scrap | +0.007 |
palm | +0.007 |
squee | +0.007 |
steps | +0.007 |
Neg logit contributions | |
pole | -0.008 |
isk | -0.007 |
sharing | -0.007 |
ole | -0.007 |
Freddie | -0.007 |
Token | his |
Token ID | 465 |
Feature activation | -1.98e+00 |
Pos logit contributions | |
roller | +0.037 |
scrap | +0.034 |
rooft | +0.034 |
palm | +0.034 |
levers | +0.034 |
Neg logit contributions | |
sharing | -0.020 |
pole | -0.019 |
Freddie | -0.018 |
isk | -0.018 |
being | -0.018 |
Token | breakfast |
Token ID | 12607 |
Feature activation | -3.17e+00 |
Pos logit contributions | |
roller | +0.035 |
rooft | +0.031 |
levers | +0.031 |
squee | +0.030 |
scrap | +0.029 |
Neg logit contributions | |
sharing | -0.048 |
pole | -0.047 |
isk | -0.046 |
being | -0.046 |
ole | -0.045 |
Token | and |
Token ID | 290 |
Feature activation | -2.33e-01 |
Pos logit contributions | |
roller | +0.066 |
steps | +0.057 |
bars | +0.056 |
seats | +0.056 |
squee | +0.056 |
Neg logit contributions | |
pole | -0.071 |
isk | -0.069 |
sharing | -0.069 |
Freddie | -0.067 |
ole | -0.066 |
Token | brushed |
Token ID | 31782 |
Feature activation | -2.18e+00 |
Pos logit contributions | |
roller | +0.005 |
rooft | +0.005 |
obs | +0.004 |
squee | +0.004 |
levers | +0.004 |
Neg logit contributions | |
sharing | -0.005 |
being | -0.005 |
pole | -0.004 |
Freddie | -0.004 |
isk | -0.004 |
Token | his |
Token ID | 465 |
Feature activation | -3.22e+00 |
Pos logit contributions | |
roller | +0.063 |
palm | +0.057 |
rooft | +0.057 |
scrap | +0.057 |
squee | +0.056 |
Neg logit contributions | |
sharing | -0.030 |
pole | -0.030 |
isk | -0.029 |
Freddie | -0.028 |
ole | -0.028 |
Token | teeth |
Token ID | 9941 |
Feature activation | -6.55e+00 |
Pos logit contributions | |
roller | +0.057 |
rooft | +0.049 |
col | +0.048 |
levers | +0.048 |
leaf | +0.048 |
Neg logit contributions | |
sharing | -0.077 |
isk | -0.075 |
ending | -0.075 |
being | -0.075 |
pole | -0.074 |
Token | . |
Token ID | 13 |
Feature activation | +3.16e-02 |
Pos logit contributions | |
roller | +0.161 |
scrap | +0.145 |
squee | +0.144 |
rooft | +0.142 |
palm | +0.142 |
Neg logit contributions | |
pole | -0.121 |
sharing | -0.119 |
isk | -0.114 |
ole | -0.112 |
being | -0.108 |
Token | He |
Token ID | 679 |
Feature activation | +1.49e-01 |
Pos logit contributions | |
pole | +0.001 |
sharing | +0.000 |
isk | +0.000 |
Freddie | +0.000 |
ole | +0.000 |
Neg logit contributions | |
roller | -0.001 |
scrap | -0.001 |
steps | -0.001 |
squee | -0.001 |
palm | -0.001 |
Token | hugged |
Token ID | 41130 |
Feature activation | -3.78e-01 |
Pos logit contributions | |
sharing | +0.003 |
pole | +0.003 |
being | +0.003 |
isk | +0.003 |
Freddie | +0.002 |
Neg logit contributions | |
roller | -0.003 |
rooft | -0.003 |
levers | -0.003 |
obs | -0.003 |
palm | -0.003 |
Token | his |
Token ID | 465 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
roller | +0.012 |
squee | +0.011 |
rooft | +0.011 |
levers | +0.011 |
leaf | +0.011 |
Neg logit contributions | |
sharing | -0.004 |
pole | -0.004 |
isk | -0.004 |
Freddie | -0.004 |
being | -0.004 |
Token | mom |
Token ID | 1995 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
roller | +0.037 |
rooft | +0.034 |
col | +0.033 |
Jo | +0.033 |
bars | +0.033 |
Neg logit contributions | |
sharing | -0.030 |
ole | -0.028 |
ending | -0.028 |
pole | -0.028 |
being | -0.027 |
Token | , |
Token ID | 11 |
Feature activation | -9.80e-01 |
Pos logit contributions | |
pole | +0.055 |
isk | +0.054 |
sharing | +0.050 |
py | +0.049 |
Freddie | +0.048 |
Neg logit contributions | |
roller | -0.057 |
steps | -0.051 |
squee | -0.051 |
scrap | -0.050 |
wonder | -0.049 |
Token | Jimmy |
Token ID | 12963 |
Feature activation | -6.22e-01 |
Pos logit contributions | |
roller | +0.021 |
seats | +0.018 |
steps | +0.018 |
scrap | +0.018 |
wonder | +0.017 |
Neg logit contributions | |
isk | -0.025 |
pole | -0.024 |
py | -0.023 |
pper | -0.021 |
ole | -0.021 |
Token | looked |
Token ID | 3114 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
roller | +0.012 |
scrap | +0.011 |
steps | +0.011 |
squee | +0.011 |
palm | +0.011 |
Neg logit contributions | |
pole | -0.015 |
isk | -0.014 |
ole | -0.014 |
sharing | -0.014 |
Freddie | -0.013 |
Token | around |
Token ID | 1088 |
Feature activation | +4.72e-01 |
Pos logit contributions | |
pole | +0.032 |
isk | +0.032 |
sharing | +0.031 |
py | +0.030 |
ole | +0.029 |
Neg logit contributions | |
roller | -0.035 |
scrap | -0.031 |
steps | -0.030 |
squee | -0.030 |
palm | -0.030 |
Token | at |
Token ID | 379 |
Feature activation | -2.08e+00 |
Pos logit contributions | |
pole | +0.010 |
isk | +0.010 |
sharing | +0.009 |
ole | +0.009 |
py | +0.009 |
Neg logit contributions | |
roller | -0.011 |
wonder | -0.009 |
squee | -0.009 |
steps | -0.009 |
scrap | -0.009 |
Token | how |
Token ID | 703 |
Feature activation | -6.50e+00 |
Pos logit contributions | |
roller | +0.051 |
scrap | +0.046 |
steps | +0.045 |
palm | +0.045 |
squee | +0.045 |
Neg logit contributions | |
pole | -0.040 |
sharing | -0.038 |
isk | -0.038 |
ole | -0.036 |
Freddie | -0.036 |
Token | messy |
Token ID | 24097 |
Feature activation | -4.24e-01 |
Pos logit contributions | |
roller | +0.139 |
scrap | +0.122 |
steps | +0.121 |
squee | +0.121 |
levers | +0.120 |
Neg logit contributions | |
pole | -0.142 |
sharing | -0.138 |
isk | -0.137 |
ole | -0.131 |
Freddie | -0.131 |
Token | the |
Token ID | 262 |
Feature activation | +9.10e-01 |
Pos logit contributions | |
roller | +0.009 |
steps | +0.008 |
seats | +0.007 |
levers | +0.007 |
squee | +0.007 |
Neg logit contributions | |
isk | -0.010 |
pole | -0.010 |
ole | -0.009 |
py | -0.009 |
ility | -0.009 |
Token | house |
Token ID | 2156 |
Feature activation | -6.94e-01 |
Pos logit contributions | |
isk | +0.018 |
pole | +0.018 |
py | +0.016 |
pper | +0.016 |
Freddie | +0.016 |
Neg logit contributions | |
roller | -0.022 |
steps | -0.021 |
seats | -0.020 |
wonder | -0.020 |
rooft | -0.019 |
Token | was |
Token ID | 373 |
Feature activation | +1.59e-01 |
Pos logit contributions | |
roller | +0.012 |
steps | +0.010 |
seats | +0.010 |
scrap | +0.010 |
levers | +0.010 |
Neg logit contributions | |
isk | -0.018 |
pole | -0.018 |
Freddie | -0.017 |
sharing | -0.017 |
ole | -0.017 |
Token | , |
Token ID | 11 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
isk | +0.004 |
py | +0.004 |
pole | +0.004 |
ingu | +0.004 |
Freddie | +0.004 |
Neg logit contributions | |
roller | -0.003 |
steps | -0.002 |
scrap | -0.002 |
wonder | -0.002 |
squee | -0.002 |
Token | hair |
Token ID | 4190 |
Feature activation | -2.83e+00 |
Pos logit contributions | |
roller | +0.028 |
rooft | +0.025 |
levers | +0.024 |
scrap | +0.024 |
steps | +0.024 |
Neg logit contributions | |
sharing | -0.030 |
pole | -0.030 |
Freddie | -0.029 |
ole | -0.029 |
being | -0.028 |
Token | tied |
Token ID | 8165 |
Feature activation | +8.36e-01 |
Pos logit contributions | |
roller | +0.070 |
rooft | +0.063 |
Ah | +0.062 |
wonder | +0.062 |
squee | +0.062 |
Neg logit contributions | |
sharing | -0.051 |
pole | -0.051 |
being | -0.047 |
ole | -0.047 |
isk | -0.046 |
Token | in |
Token ID | 287 |
Feature activation | +2.03e-01 |
Pos logit contributions | |
pole | +0.017 |
isk | +0.017 |
sharing | +0.017 |
Freddie | +0.016 |
Mickey | +0.016 |
Neg logit contributions | |
roller | -0.019 |
steps | -0.017 |
scrap | -0.016 |
squee | -0.016 |
levers | -0.016 |
Token | two |
Token ID | 734 |
Feature activation | -2.23e+00 |
Pos logit contributions | |
sharing | +0.004 |
being | +0.004 |
Freddie | +0.003 |
Mickey | +0.003 |
ole | +0.003 |
Neg logit contributions | |
roller | -0.005 |
obs | -0.005 |
leaf | -0.005 |
scrap | -0.004 |
levers | -0.004 |
Token | pig |
Token ID | 12967 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
roller | +0.039 |
rooft | +0.033 |
scrap | +0.032 |
leaf | +0.032 |
squee | +0.031 |
Neg logit contributions | |
pole | -0.055 |
sharing | -0.054 |
ole | -0.052 |
Freddie | -0.052 |
being | -0.051 |
Token | tails |
Token ID | 26404 |
Feature activation | -6.46e+00 |
Pos logit contributions | |
pole | +0.027 |
sharing | +0.027 |
isk | +0.027 |
Freddie | +0.026 |
ole | +0.026 |
Neg logit contributions | |
roller | -0.019 |
steps | -0.016 |
scrap | -0.016 |
palm | -0.016 |
squee | -0.016 |
Token | . |
Token ID | 13 |
Feature activation | +1.87e-01 |
Pos logit contributions | |
roller | +0.144 |
scrap | +0.127 |
steps | +0.126 |
squee | +0.126 |
palm | +0.125 |
Neg logit contributions | |
pole | -0.134 |
sharing | -0.131 |
isk | -0.130 |
Freddie | -0.124 |
ole | -0.124 |
Token | " |
Token ID | 366 |
Feature activation | -3.34e+00 |
Pos logit contributions | |
pole | +0.004 |
sharing | +0.004 |
isk | +0.004 |
ole | +0.004 |
Freddie | +0.004 |
Neg logit contributions | |
roller | -0.004 |
squee | -0.003 |
palm | -0.003 |
scrap | -0.003 |
steps | -0.003 |
Token | M |
Token ID | 44 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
roller | +0.073 |
steps | +0.071 |
seats | +0.068 |
levers | +0.066 |
desks | +0.065 |
Neg logit contributions | |
pole | -0.073 |
sharing | -0.068 |
isk | -0.066 |
ole | -0.064 |
Freddie | -0.064 |
Token | ummy |
Token ID | 13513 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
roller | +0.041 |
palm | +0.036 |
squee | +0.036 |
steps | +0.036 |
pine | +0.035 |
Neg logit contributions | |
pole | -0.033 |
isk | -0.031 |
Freddie | -0.031 |
sharing | -0.031 |
Mickey | -0.030 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
isk | +0.038 |
pole | +0.038 |
sharing | +0.037 |
ole | +0.036 |
Freddie | +0.036 |
Neg logit contributions | |
roller | -0.033 |
squee | -0.029 |
steps | -0.029 |
bars | -0.028 |
palm | -0.027 |
Token | Then |
Token ID | 3244 |
Feature activation | +2.30e-01 |
Pos logit contributions | |
roller | +0.010 |
steps | +0.009 |
squee | +0.009 |
rooft | +0.009 |
bars | +0.009 |
Neg logit contributions | |
pole | -0.009 |
sharing | -0.009 |
isk | -0.009 |
ole | -0.009 |
py | -0.008 |
Token | he |
Token ID | 339 |
Feature activation | +6.81e-01 |
Pos logit contributions | |
pole | +0.004 |
sharing | +0.004 |
isk | +0.004 |
ole | +0.004 |
Freddie | +0.004 |
Neg logit contributions | |
roller | -0.006 |
scrap | -0.005 |
steps | -0.005 |
squee | -0.005 |
levers | -0.005 |
Token | put |
Token ID | 1234 |
Feature activation | -7.05e-02 |
Pos logit contributions | |
pole | +0.014 |
isk | +0.013 |
sharing | +0.013 |
Freddie | +0.013 |
ole | +0.013 |
Neg logit contributions | |
roller | -0.016 |
rooft | -0.014 |
scrap | -0.014 |
palm | -0.014 |
levers | -0.014 |
Token | in |
Token ID | 287 |
Feature activation | +2.93e-02 |
Pos logit contributions | |
roller | +0.002 |
scrap | +0.001 |
squee | +0.001 |
steps | +0.001 |
levers | +0.001 |
Neg logit contributions | |
pole | -0.001 |
sharing | -0.001 |
isk | -0.001 |
Freddie | -0.001 |
ole | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | -5.09e+00 |
Pos logit contributions | |
sharing | +0.000 |
being | +0.000 |
pole | +0.000 |
ole | +0.000 |
Freddie | +0.000 |
Neg logit contributions | |
roller | -0.001 |
rooft | -0.001 |
scrap | -0.001 |
squee | -0.001 |
obs | -0.001 |
Token | bar |
Token ID | 2318 |
Feature activation | -6.45e+00 |
Pos logit contributions | |
roller | +0.096 |
rooft | +0.092 |
levers | +0.088 |
obs | +0.088 |
leaf | +0.088 |
Neg logit contributions | |
pole | -0.113 |
ole | -0.110 |
sharing | -0.109 |
Freddie | -0.106 |
Mickey | -0.106 |
Token | of |
Token ID | 286 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
roller | +0.157 |
scrap | +0.152 |
rooft | +0.144 |
squee | +0.139 |
umb | +0.139 |
Neg logit contributions | |
pole | -0.118 |
ole | -0.113 |
sharing | -0.108 |
being | -0.103 |
mate | -0.103 |
Token | soap |
Token ID | 19533 |
Feature activation | -2.58e+00 |
Pos logit contributions | |
roller | +0.026 |
rooft | +0.024 |
scrap | +0.023 |
steps | +0.023 |
squee | +0.023 |
Neg logit contributions | |
sharing | -0.022 |
pole | -0.022 |
Freddie | -0.021 |
ole | -0.020 |
Mickey | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | -5.55e-01 |
Pos logit contributions | |
roller | +0.052 |
steps | +0.045 |
squee | +0.045 |
palm | +0.045 |
scrap | +0.045 |
Neg logit contributions | |
pole | -0.059 |
isk | -0.058 |
sharing | -0.058 |
Freddie | -0.055 |
ole | -0.055 |
Token |
|
Token ID | 198 |
Feature activation | -2.85e-01 |
Pos logit contributions | |
roller | +0.013 |
steps | +0.011 |
squee | +0.011 |
levers | +0.011 |
rooft | +0.011 |
Neg logit contributions | |
pole | -0.011 |
sharing | -0.011 |
isk | -0.011 |
ole | -0.011 |
Freddie | -0.010 |
Token |
|
Token ID | 198 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
roller | +0.007 |
scrap | +0.006 |
levers | +0.006 |
palm | +0.006 |
squee | +0.006 |
Neg logit contributions | |
pole | -0.005 |
sharing | -0.005 |
isk | -0.005 |
Freddie | -0.005 |
ole | -0.005 |
Token | carefully |
Token ID | 7773 |
Feature activation | -2.23e+00 |
Pos logit contributions | |
pole | +0.002 |
sharing | +0.001 |
being | +0.001 |
isk | +0.001 |
always | +0.001 |
Neg logit contributions | |
roller | -0.002 |
rooft | -0.002 |
levers | -0.002 |
steps | -0.002 |
leaf | -0.002 |
Token | tied |
Token ID | 8165 |
Feature activation | -5.95e-01 |
Pos logit contributions | |
roller | +0.050 |
rooft | +0.045 |
scrap | +0.044 |
palm | +0.043 |
levers | +0.043 |
Neg logit contributions | |
pole | -0.045 |
sharing | -0.044 |
isk | -0.042 |
being | -0.041 |
ole | -0.040 |
Token | the |
Token ID | 262 |
Feature activation | -9.54e-01 |
Pos logit contributions | |
roller | +0.014 |
scrap | +0.012 |
squee | +0.012 |
palm | +0.012 |
leaf | +0.012 |
Neg logit contributions | |
pole | -0.012 |
sharing | -0.011 |
ole | -0.011 |
being | -0.011 |
Freddie | -0.011 |
Token | ribbon |
Token ID | 29092 |
Feature activation | -2.99e+00 |
Pos logit contributions | |
roller | +0.017 |
levers | +0.014 |
Jo | +0.014 |
squee | +0.014 |
rooft | +0.014 |
Neg logit contributions | |
pole | -0.024 |
sharing | -0.023 |
ole | -0.022 |
being | -0.022 |
Freddie | -0.022 |
Token | in |
Token ID | 287 |
Feature activation | -7.65e-01 |
Pos logit contributions | |
roller | +0.069 |
scrap | +0.062 |
squee | +0.060 |
rooft | +0.059 |
levers | +0.059 |
Neg logit contributions | |
pole | -0.064 |
sharing | -0.058 |
ole | -0.058 |
Freddie | -0.056 |
isk | -0.055 |
Token | a |
Token ID | 257 |
Feature activation | -6.45e+00 |
Pos logit contributions | |
roller | +0.020 |
obs | +0.018 |
scrap | +0.018 |
leaf | +0.018 |
squee | +0.018 |
Neg logit contributions | |
sharing | -0.013 |
pole | -0.013 |
ole | -0.012 |
being | -0.012 |
Mickey | -0.012 |
Token | knot |
Token ID | 29654 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
roller | +0.124 |
leaf | +0.116 |
obs | +0.115 |
levers | +0.114 |
bang | +0.114 |
Neg logit contributions | |
pole | -0.151 |
ole | -0.134 |
end | -0.132 |
sharing | -0.130 |
Mickey | -0.126 |
Token | and |
Token ID | 290 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
roller | +0.037 |
scrap | +0.033 |
squee | +0.033 |
steps | +0.032 |
levers | +0.032 |
Neg logit contributions | |
pole | -0.032 |
sharing | -0.031 |
isk | -0.031 |
ole | -0.030 |
Freddie | -0.030 |
Token | str |
Token ID | 965 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
roller | +0.038 |
rooft | +0.035 |
palm | +0.034 |
scrap | +0.034 |
leaf | +0.033 |
Neg logit contributions | |
sharing | -0.027 |
pole | -0.027 |
being | -0.026 |
Mickey | -0.026 |
Freddie | -0.025 |
Token | ung |
Token ID | 2150 |
Feature activation | -1.96e+00 |
Pos logit contributions | |
roller | +0.036 |
scrap | +0.032 |
squee | +0.032 |
steps | +0.032 |
levers | +0.032 |
Neg logit contributions | |
pole | -0.030 |
sharing | -0.030 |
isk | -0.029 |
Freddie | -0.028 |
Mickey | -0.027 |
Token | the |
Token ID | 262 |
Feature activation | -1.70e-01 |
Pos logit contributions | |
roller | +0.050 |
scrap | +0.044 |
palm | +0.043 |
seats | +0.043 |
squee | +0.042 |
Neg logit contributions | |
pole | -0.034 |
sharing | -0.034 |
Freddie | -0.034 |
ole | -0.033 |
Mickey | -0.033 |
Token | milk |
Token ID | 7545 |
Feature activation | -1.72e+00 |
Pos logit contributions | |
roller | +0.032 |
rooft | +0.028 |
levers | +0.028 |
steps | +0.027 |
scrap | +0.027 |
Neg logit contributions | |
sharing | -0.034 |
pole | -0.033 |
isk | -0.032 |
Freddie | -0.032 |
being | -0.031 |
Token | . |
Token ID | 13 |
Feature activation | +2.54e-01 |
Pos logit contributions | |
roller | +0.030 |
bars | +0.027 |
squee | +0.027 |
steps | +0.027 |
scrap | +0.026 |
Neg logit contributions | |
pole | -0.043 |
isk | -0.042 |
py | -0.041 |
sharing | -0.041 |
Freddie | -0.040 |
Token | He |
Token ID | 679 |
Feature activation | -1.21e-02 |
Pos logit contributions | |
sharing | +0.004 |
pole | +0.004 |
isk | +0.004 |
Freddie | +0.004 |
Mickey | +0.003 |
Neg logit contributions | |
roller | -0.007 |
scrap | -0.006 |
steps | -0.006 |
palm | -0.006 |
levers | -0.006 |
Token | brushed |
Token ID | 31782 |
Feature activation | -2.13e+00 |
Pos logit contributions | |
roller | +0.000 |
rooft | +0.000 |
levers | +0.000 |
steps | +0.000 |
palm | +0.000 |
Neg logit contributions | |
sharing | -0.000 |
pole | -0.000 |
isk | -0.000 |
Freddie | -0.000 |
being | -0.000 |
Token | his |
Token ID | 465 |
Feature activation | -3.18e+00 |
Pos logit contributions | |
roller | +0.060 |
scrap | +0.053 |
rooft | +0.053 |
steps | +0.053 |
squee | +0.053 |
Neg logit contributions | |
pole | -0.032 |
sharing | -0.032 |
isk | -0.031 |
Freddie | -0.030 |
ole | -0.030 |
Token | teeth |
Token ID | 9941 |
Feature activation | -6.44e+00 |
Pos logit contributions | |
roller | +0.055 |
levers | +0.045 |
Jo | +0.045 |
leaf | +0.045 |
col | +0.044 |
Neg logit contributions | |
sharing | -0.080 |
isk | -0.080 |
pole | -0.079 |
being | -0.078 |
Mickey | -0.078 |
Token | every |
Token ID | 790 |
Feature activation | -2.54e+00 |
Pos logit contributions | |
roller | +0.156 |
scrap | +0.138 |
levers | +0.137 |
squee | +0.137 |
rooft | +0.136 |
Neg logit contributions | |
pole | -0.123 |
sharing | -0.120 |
isk | -0.117 |
ole | -0.113 |
Freddie | -0.112 |
Token | day |
Token ID | 1110 |
Feature activation | -9.52e-01 |
Pos logit contributions | |
roller | +0.071 |
levers | +0.067 |
rooft | +0.066 |
squee | +0.064 |
scrap | +0.064 |
Neg logit contributions | |
sharing | -0.036 |
pole | -0.035 |
ending | -0.034 |
Freddie | -0.033 |
isk | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | +2.52e-01 |
Pos logit contributions | |
roller | +0.020 |
steps | +0.017 |
palm | +0.017 |
bars | +0.017 |
squee | +0.017 |
Neg logit contributions | |
pole | -0.022 |
isk | -0.021 |
sharing | -0.021 |
Freddie | -0.020 |
ole | -0.020 |
Token | He |
Token ID | 679 |
Feature activation | +2.97e-01 |
Pos logit contributions | |
sharing | +0.005 |
pole | +0.004 |
isk | +0.004 |
Freddie | +0.004 |
Mickey | +0.004 |
Neg logit contributions | |
roller | -0.006 |
scrap | -0.006 |
steps | -0.006 |
levers | -0.006 |
squee | -0.006 |
Token | stayed |
Token ID | 9658 |
Feature activation | +2.62e+00 |
Pos logit contributions | |
sharing | +0.006 |
pole | +0.006 |
isk | +0.006 |
Freddie | +0.006 |
being | +0.006 |
Neg logit contributions | |
roller | -0.007 |
rooft | -0.006 |
levers | -0.006 |
steps | -0.006 |
palm | -0.006 |
Token | it |
Token ID | 340 |
Feature activation | -8.58e-01 |
Pos logit contributions | |
pole | +0.001 |
sharing | +0.001 |
Freddie | +0.001 |
isk | +0.001 |
Mickey | +0.001 |
Neg logit contributions | |
roller | -0.001 |
scrap | -0.001 |
levers | -0.001 |
steps | -0.001 |
rooft | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +4.98e-01 |
Pos logit contributions | |
roller | +0.016 |
bars | +0.015 |
steps | +0.015 |
squee | +0.014 |
scrap | +0.014 |
Neg logit contributions | |
isk | -0.020 |
pole | -0.020 |
sharing | -0.020 |
Freddie | -0.019 |
ole | -0.019 |
Token | She |
Token ID | 1375 |
Feature activation | +1.98e-02 |
Pos logit contributions | |
sharing | +0.009 |
Mickey | +0.008 |
Freddie | +0.008 |
isk | +0.008 |
Blue | +0.008 |
Neg logit contributions | |
roller | -0.013 |
scrap | -0.012 |
levers | -0.011 |
palm | -0.011 |
leaf | -0.011 |
Token | brushed |
Token ID | 31782 |
Feature activation | -2.46e+00 |
Pos logit contributions | |
sharing | +0.000 |
isk | +0.000 |
pole | +0.000 |
being | +0.000 |
Freddie | +0.000 |
Neg logit contributions | |
roller | -0.000 |
rooft | -0.000 |
levers | -0.000 |
steps | -0.000 |
obs | -0.000 |
Token | her |
Token ID | 607 |
Feature activation | -3.41e+00 |
Pos logit contributions | |
roller | +0.059 |
rooft | +0.051 |
palm | +0.050 |
leaf | +0.050 |
steps | +0.050 |
Neg logit contributions | |
pole | -0.047 |
sharing | -0.047 |
isk | -0.046 |
Freddie | -0.046 |
Mickey | -0.046 |
Token | teeth |
Token ID | 9941 |
Feature activation | -6.43e+00 |
Pos logit contributions | |
roller | +0.069 |
leaf | +0.061 |
col | +0.058 |
Jo | +0.057 |
levers | +0.057 |
Neg logit contributions | |
isk | -0.077 |
being | -0.075 |
sharing | -0.075 |
pole | -0.073 |
Mickey | -0.073 |
Token | carefully |
Token ID | 7773 |
Feature activation | -1.88e+00 |
Pos logit contributions | |
roller | +0.157 |
scrap | +0.140 |
squee | +0.137 |
rooft | +0.137 |
levers | +0.137 |
Neg logit contributions | |
pole | -0.122 |
sharing | -0.119 |
isk | -0.115 |
ole | -0.112 |
Freddie | -0.110 |
Token | . |
Token ID | 13 |
Feature activation | +4.22e-01 |
Pos logit contributions | |
roller | +0.041 |
scrap | +0.036 |
steps | +0.036 |
squee | +0.036 |
levers | +0.036 |
Neg logit contributions | |
pole | -0.040 |
sharing | -0.039 |
isk | -0.038 |
Freddie | -0.037 |
ole | -0.037 |
Token | She |
Token ID | 1375 |
Feature activation | +4.89e-02 |
Pos logit contributions | |
sharing | +0.008 |
Mickey | +0.007 |
Freddie | +0.007 |
isk | +0.007 |
pole | +0.007 |
Neg logit contributions | |
roller | -0.011 |
scrap | -0.009 |
palm | -0.009 |
levers | -0.009 |
steps | -0.009 |
Token | spit |
Token ID | 27591 |
Feature activation | -2.55e+00 |
Pos logit contributions | |
sharing | +0.001 |
isk | +0.001 |
pole | +0.001 |
being | +0.001 |
always | +0.001 |
Neg logit contributions | |
roller | -0.001 |
rooft | -0.001 |
levers | -0.001 |
steps | -0.001 |
leaf | -0.001 |
Token | out |
Token ID | 503 |
Feature activation | +9.05e-01 |
Pos logit contributions | |
roller | +0.061 |
rooft | +0.056 |
seats | +0.055 |
palm | +0.055 |
desks | +0.055 |
Neg logit contributions | |
Freddie | -0.047 |
pole | -0.046 |
end | -0.045 |
Mickey | -0.045 |
sharing | -0.045 |
Token | toy |
Token ID | 13373 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
sharing | +0.034 |
being | +0.033 |
ending | +0.032 |
ole | +0.030 |
pole | +0.030 |
Neg logit contributions | |
roller | -0.052 |
rooft | -0.048 |
Ah | -0.047 |
leaf | -0.047 |
scrap | -0.045 |
Token | . |
Token ID | 13 |
Feature activation | -7.83e-02 |
Pos logit contributions | |
pole | +0.025 |
isk | +0.025 |
sharing | +0.025 |
ole | +0.024 |
Freddie | +0.024 |
Neg logit contributions | |
roller | -0.022 |
steps | -0.019 |
bars | -0.019 |
levers | -0.019 |
squee | -0.019 |
Token | They |
Token ID | 1119 |
Feature activation | +6.49e-01 |
Pos logit contributions | |
roller | +0.002 |
squee | +0.001 |
seats | +0.001 |
rooft | +0.001 |
wonder | +0.001 |
Neg logit contributions | |
pole | -0.002 |
isk | -0.002 |
ole | -0.002 |
sharing | -0.002 |
py | -0.002 |
Token | laughed |
Token ID | 13818 |
Feature activation | +2.97e-01 |
Pos logit contributions | |
pole | +0.013 |
sharing | +0.013 |
isk | +0.012 |
Freddie | +0.012 |
ole | +0.012 |
Neg logit contributions | |
roller | -0.015 |
scrap | -0.013 |
squee | -0.013 |
steps | -0.013 |
levers | -0.013 |
Token | at |
Token ID | 379 |
Feature activation | -1.85e+00 |
Pos logit contributions | |
isk | +0.008 |
pole | +0.008 |
ole | +0.007 |
melon | +0.007 |
py | +0.007 |
Neg logit contributions | |
roller | -0.005 |
squee | -0.004 |
wonder | -0.004 |
levers | -0.004 |
steps | -0.004 |
Token | how |
Token ID | 703 |
Feature activation | -6.43e+00 |
Pos logit contributions | |
roller | +0.047 |
scrap | +0.042 |
steps | +0.042 |
squee | +0.042 |
rooft | +0.041 |
Neg logit contributions | |
pole | -0.033 |
sharing | -0.032 |
isk | -0.031 |
Freddie | -0.030 |
ole | -0.030 |
Token | clever |
Token ID | 14169 |
Feature activation | -1.38e+00 |
Pos logit contributions | |
roller | +0.141 |
rooft | +0.124 |
levers | +0.123 |
palm | +0.123 |
bars | +0.121 |
Neg logit contributions | |
sharing | -0.137 |
pole | -0.131 |
Freddie | -0.129 |
isk | -0.127 |
being | -0.126 |
Token | he |
Token ID | 339 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
roller | +0.029 |
steps | +0.026 |
levers | +0.026 |
squee | +0.026 |
lace | +0.025 |
Neg logit contributions | |
isk | -0.032 |
pole | -0.031 |
ole | -0.030 |
sharing | -0.029 |
melon | -0.029 |
Token | was |
Token ID | 373 |
Feature activation | -5.30e-01 |
Pos logit contributions | |
pole | +0.041 |
sharing | +0.040 |
isk | +0.039 |
ole | +0.039 |
Freddie | +0.038 |
Neg logit contributions | |
roller | -0.036 |
scrap | -0.031 |
squee | -0.031 |
steps | -0.031 |
levers | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | -4.36e-01 |
Pos logit contributions | |
roller | +0.010 |
bars | +0.008 |
steps | +0.008 |
palm | +0.008 |
scrap | +0.008 |
Neg logit contributions | |
pole | -0.013 |
isk | -0.013 |
ole | -0.013 |
End | -0.013 |
Freddie | -0.012 |
Token | Ace |
Token ID | 17102 |
Feature activation | -3.91e-01 |
Pos logit contributions | |
roller | +0.009 |
squee | +0.008 |
seats | +0.008 |
steps | +0.008 |
scrap | +0.008 |
Neg logit contributions | |
pole | -0.010 |
ole | -0.009 |
isk | -0.009 |
sharing | -0.009 |
py | -0.009 |
Token | She |
Token ID | 1375 |
Feature activation | +1.52e-02 |
Pos logit contributions | |
sharing | +0.008 |
isk | +0.008 |
Freddie | +0.008 |
pole | +0.008 |
Mickey | +0.007 |
Neg logit contributions | |
roller | -0.011 |
levers | -0.009 |
scrap | -0.009 |
rooft | -0.009 |
steps | -0.009 |
Token | poured |
Token ID | 19036 |
Feature activation | -2.77e+00 |
Pos logit contributions | |
isk | +0.000 |
sharing | +0.000 |
pole | +0.000 |
Freddie | +0.000 |
being | +0.000 |
Neg logit contributions | |
roller | -0.000 |
rooft | -0.000 |
levers | -0.000 |
steps | -0.000 |
palm | -0.000 |
Token | some |
Token ID | 617 |
Feature activation | -3.36e+00 |
Pos logit contributions | |
roller | +0.073 |
Jo | +0.066 |
levers | +0.064 |
scrap | +0.064 |
leaf | +0.064 |
Neg logit contributions | |
melon | -0.047 |
sharing | -0.045 |
Freddie | -0.043 |
ole | -0.043 |
ter | -0.042 |
Token | milk |
Token ID | 7545 |
Feature activation | -1.49e+00 |
Pos logit contributions | |
roller | +0.079 |
rooft | +0.077 |
Jo | +0.076 |
icol | +0.073 |
amaz | +0.072 |
Neg logit contributions | |
end | -0.060 |
ole | -0.060 |
melon | -0.058 |
ters | -0.056 |
sharing | -0.056 |
Token | in |
Token ID | 287 |
Feature activation | -9.59e-01 |
Pos logit contributions | |
roller | +0.030 |
steps | +0.027 |
squee | +0.027 |
bars | +0.026 |
palm | +0.026 |
Neg logit contributions | |
pole | -0.034 |
isk | -0.033 |
sharing | -0.033 |
Freddie | -0.032 |
Mickey | -0.031 |
Token | a |
Token ID | 257 |
Feature activation | -6.42e+00 |
Pos logit contributions | |
roller | +0.026 |
scrap | +0.023 |
levers | +0.023 |
steps | +0.023 |
leaf | +0.023 |
Neg logit contributions | |
sharing | -0.016 |
pole | -0.015 |
Freddie | -0.015 |
ole | -0.015 |
being | -0.014 |
Token | purple |
Token ID | 14032 |
Feature activation | -4.25e+00 |
Pos logit contributions | |
roller | +0.127 |
levers | +0.121 |
rooft | +0.118 |
leaf | +0.115 |
Jo | +0.114 |
Neg logit contributions | |
pole | -0.140 |
ole | -0.136 |
melon | -0.134 |
end | -0.134 |
sharing | -0.133 |
Token | cup |
Token ID | 6508 |
Feature activation | -4.61e+00 |
Pos logit contributions | |
roller | +0.073 |
levers | +0.066 |
scrap | +0.064 |
rooft | +0.064 |
leaf | +0.061 |
Neg logit contributions | |
pole | -0.107 |
sharing | -0.106 |
ole | -0.104 |
melon | -0.104 |
isk | -0.103 |
Token | and |
Token ID | 290 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
roller | +0.105 |
scrap | +0.094 |
levers | +0.092 |
squee | +0.091 |
rooft | +0.091 |
Neg logit contributions | |
pole | -0.094 |
sharing | -0.092 |
isk | -0.090 |
ole | -0.087 |
Freddie | -0.086 |
Token | added |
Token ID | 2087 |
Feature activation | -2.35e+00 |
Pos logit contributions | |
roller | +0.026 |
rooft | +0.023 |
levers | +0.023 |
obs | +0.022 |
scrap | +0.022 |
Neg logit contributions | |
sharing | -0.019 |
Freddie | -0.019 |
being | -0.018 |
Mickey | -0.018 |
pole | -0.017 |
Token | some |
Token ID | 617 |
Feature activation | -3.28e+00 |
Pos logit contributions | |
roller | +0.055 |
scrap | +0.048 |
levers | +0.047 |
rooft | +0.047 |
Jo | +0.046 |
Neg logit contributions | |
sharing | -0.046 |
pole | -0.044 |
ole | -0.044 |
being | -0.043 |
Freddie | -0.043 |
Token | Everyone |
Token ID | 11075 |
Feature activation | -3.54e-01 |
Pos logit contributions | |
roller | +0.006 |
scrap | +0.005 |
levers | +0.005 |
squee | +0.005 |
steps | +0.005 |
Neg logit contributions | |
pole | -0.005 |
sharing | -0.005 |
isk | -0.005 |
Freddie | -0.005 |
Mickey | -0.005 |
Token | compliment |
Token ID | 19370 |
Feature activation | +3.58e-01 |
Pos logit contributions | |
roller | +0.007 |
scrap | +0.006 |
steps | +0.006 |
squee | +0.006 |
wonder | +0.006 |
Neg logit contributions | |
isk | -0.009 |
pole | -0.008 |
sharing | -0.008 |
ole | -0.008 |
Freddie | -0.008 |
Token | ed |
Token ID | 276 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
pole | +0.011 |
sharing | +0.011 |
isk | +0.011 |
Freddie | +0.010 |
ole | +0.010 |
Neg logit contributions | |
roller | -0.005 |
scrap | -0.004 |
rooft | -0.004 |
squee | -0.004 |
palm | -0.004 |
Token | her |
Token ID | 607 |
Feature activation | -5.90e-01 |
Pos logit contributions | |
pole | +0.021 |
sharing | +0.021 |
isk | +0.020 |
Freddie | +0.020 |
ole | +0.020 |
Neg logit contributions | |
roller | -0.024 |
scrap | -0.021 |
rooft | -0.021 |
palm | -0.021 |
squee | -0.020 |
Token | on |
Token ID | 319 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
roller | +0.014 |
scrap | +0.012 |
steps | +0.012 |
squee | +0.012 |
levers | +0.012 |
Neg logit contributions | |
pole | -0.012 |
sharing | -0.011 |
isk | -0.011 |
Freddie | -0.011 |
ole | -0.011 |
Token | how |
Token ID | 703 |
Feature activation | -6.41e+00 |
Pos logit contributions | |
pole | +0.023 |
sharing | +0.023 |
isk | +0.023 |
Freddie | +0.022 |
ole | +0.022 |
Neg logit contributions | |
roller | -0.026 |
scrap | -0.023 |
squee | -0.023 |
steps | -0.023 |
rooft | -0.023 |
Token | special |
Token ID | 2041 |
Feature activation | +5.82e-01 |
Pos logit contributions | |
roller | +0.153 |
bars | +0.133 |
bisc | +0.132 |
levers | +0.132 |
palm | +0.132 |
Neg logit contributions | |
sharing | -0.128 |
being | -0.119 |
pole | -0.116 |
ole | -0.115 |
Freddie | -0.113 |
Token | she |
Token ID | 673 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
pole | +0.013 |
isk | +0.013 |
sharing | +0.013 |
ole | +0.012 |
Freddie | +0.012 |
Neg logit contributions | |
roller | -0.012 |
steps | -0.010 |
squee | -0.010 |
scrap | -0.010 |
levers | -0.010 |
Token | looked |
Token ID | 3114 |
Feature activation | +2.07e+00 |
Pos logit contributions | |
pole | +0.047 |
sharing | +0.045 |
ole | +0.045 |
isk | +0.045 |
Freddie | +0.044 |
Neg logit contributions | |
roller | -0.039 |
scrap | -0.034 |
steps | -0.034 |
squee | -0.034 |
wonder | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | -4.69e-01 |
Pos logit contributions | |
pole | +0.041 |
isk | +0.040 |
sharing | +0.040 |
Freddie | +0.038 |
ole | +0.038 |
Neg logit contributions | |
roller | -0.048 |
scrap | -0.042 |
steps | -0.042 |
squee | -0.042 |
levers | -0.042 |
Token |
|
Token ID | 198 |
Feature activation | +7.30e-01 |
Pos logit contributions | |
roller | +0.010 |
scrap | +0.009 |
squee | +0.009 |
steps | +0.009 |
levers | +0.009 |
Neg logit contributions | |
pole | -0.010 |
sharing | -0.010 |
isk | -0.010 |
Freddie | -0.009 |
ole | -0.009 |
Token | hit |
Token ID | 2277 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
roller | +0.027 |
rooft | +0.026 |
wonder | +0.025 |
leaf | +0.025 |
amaz | +0.024 |
Neg logit contributions | |
sharing | -0.027 |
isk | -0.027 |
being | -0.027 |
pole | -0.026 |
Mickey | -0.025 |
Token | her |
Token ID | 607 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
roller | +0.026 |
obs | +0.024 |
squee | +0.024 |
cu | +0.023 |
leaf | +0.023 |
Neg logit contributions | |
being | -0.017 |
sharing | -0.017 |
pole | -0.016 |
isk | -0.016 |
Mickey | -0.016 |
Token | face |
Token ID | 1986 |
Feature activation | -9.57e-01 |
Pos logit contributions | |
roller | +0.025 |
rooft | +0.023 |
leaf | +0.023 |
Ah | +0.023 |
squee | +0.023 |
Neg logit contributions | |
pole | -0.021 |
sharing | -0.020 |
ole | -0.019 |
being | -0.019 |
isk | -0.019 |
Token | on |
Token ID | 319 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
roller | +0.024 |
leaf | +0.022 |
seats | +0.022 |
rooft | +0.022 |
wonder | +0.021 |
Neg logit contributions | |
pole | -0.017 |
sharing | -0.016 |
being | -0.015 |
ole | -0.015 |
isk | -0.015 |
Token | the |
Token ID | 262 |
Feature activation | -5.45e-01 |
Pos logit contributions | |
Mickey | +0.002 |
being | +0.002 |
sharing | +0.002 |
Freddie | +0.002 |
isk | +0.002 |
Neg logit contributions | |
roller | -0.003 |
leaf | -0.003 |
obs | -0.003 |
squee | -0.003 |
levers | -0.003 |
Token | bar |
Token ID | 2318 |
Feature activation | -6.41e+00 |
Pos logit contributions | |
roller | +0.011 |
Jo | +0.011 |
wonder | +0.011 |
squee | +0.011 |
amaz | +0.011 |
Neg logit contributions | |
pole | -0.011 |
being | -0.010 |
end | -0.010 |
ter | -0.010 |
sharing | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | -6.15e-01 |
Pos logit contributions | |
roller | +0.152 |
scrap | +0.150 |
squee | +0.142 |
umb | +0.141 |
rooft | +0.141 |
Neg logit contributions | |
pole | -0.125 |
ole | -0.118 |
sharing | -0.106 |
melon | -0.103 |
pper | -0.103 |
Token | Her |
Token ID | 2332 |
Feature activation | -2.70e-02 |
Pos logit contributions | |
roller | +0.015 |
levers | +0.013 |
rooft | +0.013 |
steps | +0.013 |
leaf | +0.013 |
Neg logit contributions | |
sharing | -0.011 |
Mickey | -0.011 |
Freddie | -0.011 |
pole | -0.011 |
isk | -0.011 |
Token | parents |
Token ID | 3397 |
Feature activation | -1.73e+00 |
Pos logit contributions | |
leaf | +0.001 |
rooft | +0.001 |
roller | +0.001 |
Ah | +0.001 |
col | +0.001 |
Neg logit contributions | |
sharing | -0.000 |
being | -0.000 |
ending | -0.000 |
okay | -0.000 |
not | -0.000 |
Token | were |
Token ID | 547 |
Feature activation | -3.41e-01 |
Pos logit contributions | |
roller | +0.038 |
rooft | +0.034 |
steps | +0.033 |
squee | +0.033 |
palm | +0.032 |
Neg logit contributions | |
pole | -0.037 |
sharing | -0.037 |
isk | -0.035 |
being | -0.034 |
Freddie | -0.034 |
Token | worried |
Token ID | 7960 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
roller | +0.007 |
scrap | +0.007 |
steps | +0.006 |
squee | +0.006 |
palm | +0.006 |
Neg logit contributions | |
pole | -0.008 |
isk | -0.007 |
sharing | -0.007 |
Freddie | -0.007 |
ole | -0.007 |
Token | warrior |
Token ID | 16491 |
Feature activation | +2.24e+00 |
Pos logit contributions | |
roller | +0.006 |
levers | +0.006 |
desks | +0.005 |
rooft | +0.005 |
scrap | +0.005 |
Neg logit contributions | |
pole | -0.005 |
sharing | -0.005 |
ole | -0.005 |
ending | -0.005 |
Freddie | -0.005 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.78e+00 |
Pos logit contributions | |
isk | +0.054 |
pole | +0.054 |
sharing | +0.053 |
ole | +0.051 |
Freddie | +0.051 |
Neg logit contributions | |
roller | -0.042 |
steps | -0.035 |
squee | -0.035 |
bars | -0.035 |
palm | -0.035 |
Token | She |
Token ID | 1375 |
Feature activation | +8.29e-01 |
Pos logit contributions | |
roller | +0.040 |
scrap | +0.036 |
levers | +0.036 |
rooft | +0.035 |
squee | +0.035 |
Neg logit contributions | |
pole | -0.036 |
sharing | -0.035 |
isk | -0.035 |
being | -0.033 |
Freddie | -0.033 |
Token | spun |
Token ID | 26843 |
Feature activation | -9.12e-01 |
Pos logit contributions | |
pole | +0.017 |
isk | +0.016 |
sharing | +0.016 |
being | +0.015 |
Freddie | +0.015 |
Neg logit contributions | |
roller | -0.018 |
rooft | -0.017 |
obs | -0.016 |
desks | -0.016 |
levers | -0.016 |
Token | in |
Token ID | 287 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
roller | +0.024 |
obs | +0.023 |
desks | +0.023 |
palm | +0.023 |
scrap | +0.022 |
Neg logit contributions | |
pole | -0.014 |
being | -0.013 |
Freddie | -0.013 |
sharing | -0.013 |
ole | -0.012 |
Token | a |
Token ID | 257 |
Feature activation | -6.41e+00 |
Pos logit contributions | |
obs | +0.047 |
leaf | +0.043 |
Boo | +0.042 |
roller | +0.042 |
itten | +0.042 |
Neg logit contributions | |
being | -0.028 |
sharing | -0.026 |
end | -0.024 |
Freddie | -0.021 |
End | -0.020 |
Token | circle |
Token ID | 9197 |
Feature activation | -2.73e+00 |
Pos logit contributions | |
obs | +0.121 |
levers | +0.113 |
buck | +0.111 |
roller | +0.110 |
leaf | +0.109 |
Neg logit contributions | |
pole | -0.145 |
end | -0.139 |
ole | -0.130 |
End | -0.127 |
sharing | -0.127 |
Token | and |
Token ID | 290 |
Feature activation | -1.85e+00 |
Pos logit contributions | |
roller | +0.064 |
scrap | +0.057 |
rooft | +0.056 |
squee | +0.056 |
levers | +0.055 |
Neg logit contributions | |
pole | -0.055 |
sharing | -0.053 |
isk | -0.051 |
ole | -0.050 |
Freddie | -0.049 |
Token | sm |
Token ID | 895 |
Feature activation | -2.82e+00 |
Pos logit contributions | |
roller | +0.039 |
rooft | +0.037 |
palm | +0.036 |
obs | +0.035 |
wonder | +0.035 |
Neg logit contributions | |
pole | -0.037 |
sharing | -0.036 |
Freddie | -0.036 |
being | -0.036 |
Mickey | -0.034 |
Token | acked |
Token ID | 6021 |
Feature activation | -7.61e-01 |
Pos logit contributions | |
roller | +0.061 |
scrap | +0.055 |
levers | +0.054 |
seats | +0.053 |
steps | +0.052 |
Neg logit contributions | |
ole | -0.061 |
sharing | -0.061 |
pole | -0.060 |
isk | -0.060 |
py | -0.057 |
Token | her |
Token ID | 607 |
Feature activation | -2.65e+00 |
Pos logit contributions | |
roller | +0.021 |
leaf | +0.019 |
scrap | +0.019 |
rooft | +0.019 |
seats | +0.019 |
Neg logit contributions | |
sharing | -0.011 |
pole | -0.011 |
ole | -0.011 |
being | -0.011 |
Freddie | -0.010 |
Token | to |
Token ID | 284 |
Feature activation | +1.89e-01 |
Pos logit contributions | |
roller | +0.047 |
rooft | +0.043 |
leaf | +0.042 |
levers | +0.042 |
Ah | +0.042 |
Neg logit contributions | |
sharing | -0.047 |
being | -0.044 |
Freddie | -0.040 |
Mickey | -0.039 |
isk | -0.039 |
Token | rub |
Token ID | 6437 |
Feature activation | -2.18e+00 |
Pos logit contributions | |
pole | +0.004 |
sharing | +0.004 |
isk | +0.004 |
Freddie | +0.004 |
ole | +0.004 |
Neg logit contributions | |
roller | -0.004 |
scrap | -0.004 |
rooft | -0.004 |
steps | -0.004 |
squee | -0.004 |
Token | dirt |
Token ID | 13647 |
Feature activation | -3.16e-01 |
Pos logit contributions | |
roller | +0.052 |
rooft | +0.045 |
leaf | +0.045 |
seats | +0.044 |
bars | +0.044 |
Neg logit contributions | |
Freddie | -0.042 |
sharing | -0.041 |
ole | -0.041 |
Mickey | -0.041 |
pole | -0.040 |
Token | onto |
Token ID | 4291 |
Feature activation | -4.33e-01 |
Pos logit contributions | |
roller | +0.007 |
scrap | +0.007 |
squee | +0.007 |
steps | +0.006 |
levers | +0.006 |
Neg logit contributions | |
pole | -0.006 |
sharing | -0.006 |
isk | -0.006 |
Freddie | -0.006 |
ole | -0.006 |
Token | her |
Token ID | 607 |
Feature activation | -2.71e+00 |
Pos logit contributions | |
roller | +0.010 |
scrap | +0.009 |
squee | +0.009 |
steps | +0.008 |
levers | +0.008 |
Neg logit contributions | |
pole | -0.009 |
sharing | -0.009 |
isk | -0.009 |
Freddie | -0.008 |
ole | -0.008 |
Token | teeth |
Token ID | 9941 |
Feature activation | -6.41e+00 |
Pos logit contributions | |
roller | +0.044 |
rooft | +0.036 |
levers | +0.036 |
squee | +0.035 |
scrap | +0.035 |
Neg logit contributions | |
pole | -0.072 |
sharing | -0.072 |
isk | -0.071 |
ole | -0.069 |
Freddie | -0.069 |
Token | . |
Token ID | 13 |
Feature activation | -8.55e-01 |
Pos logit contributions | |
roller | +0.152 |
scrap | +0.135 |
squee | +0.134 |
rooft | +0.133 |
steps | +0.133 |
Neg logit contributions | |
pole | -0.124 |
sharing | -0.121 |
isk | -0.118 |
ole | -0.115 |
Freddie | -0.114 |
Token | But |
Token ID | 887 |
Feature activation | -6.16e-01 |
Pos logit contributions | |
roller | +0.020 |
scrap | +0.018 |
levers | +0.018 |
palm | +0.017 |
seats | +0.017 |
Neg logit contributions | |
sharing | -0.017 |
pole | -0.016 |
isk | -0.016 |
Freddie | -0.016 |
Mickey | -0.016 |
Token | the |
Token ID | 262 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
roller | +0.014 |
scrap | +0.012 |
steps | +0.012 |
palm | +0.012 |
squee | +0.012 |
Neg logit contributions | |
pole | -0.013 |
isk | -0.012 |
sharing | -0.012 |
ole | -0.012 |
Freddie | -0.012 |
Token | dirt |
Token ID | 13647 |
Feature activation | -3.66e-01 |
Pos logit contributions | |
sharing | +0.025 |
ending | +0.024 |
being | +0.024 |
Freddie | +0.023 |
pole | +0.023 |
Neg logit contributions | |
roller | -0.024 |
rooft | -0.021 |
levers | -0.021 |
leaf | -0.021 |
bars | -0.020 |
Token | wouldn |
Token ID | 3636 |
Feature activation | -3.79e+00 |
Pos logit contributions | |
roller | +0.008 |
rooft | +0.007 |
scrap | +0.007 |
squee | +0.007 |
levers | +0.007 |
Neg logit contributions | |
pole | -0.008 |
sharing | -0.008 |
isk | -0.007 |
ole | -0.007 |
Freddie | -0.007 |
Token |
|
Token ID | 198 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
sharing | +0.001 |
pole | +0.001 |
Freddie | +0.001 |
isk | +0.001 |
ole | +0.001 |
Neg logit contributions | |
roller | -0.001 |
levers | -0.001 |
scrap | -0.001 |
lace | -0.001 |
squee | -0.001 |
Token | Jane |
Token ID | 41083 |
Feature activation | -7.38e-01 |
Pos logit contributions | |
pole | +0.045 |
sharing | +0.044 |
isk | +0.043 |
Freddie | +0.042 |
ole | +0.042 |
Neg logit contributions | |
roller | -0.034 |
scrap | -0.029 |
squee | -0.029 |
steps | -0.029 |
levers | -0.028 |
Token | tried |
Token ID | 3088 |
Feature activation | -2.24e+00 |
Pos logit contributions | |
roller | +0.016 |
rooft | +0.014 |
levers | +0.014 |
scrap | +0.014 |
steps | +0.014 |
Neg logit contributions | |
pole | -0.016 |
sharing | -0.015 |
isk | -0.015 |
Freddie | -0.015 |
ole | -0.014 |
Token | brushing |
Token ID | 42798 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
roller | +0.053 |
rooft | +0.048 |
levers | +0.048 |
seats | +0.047 |
leaf | +0.047 |
Neg logit contributions | |
sharing | -0.046 |
being | -0.043 |
Freddie | -0.041 |
Mickey | -0.039 |
isk | -0.039 |
Token | her |
Token ID | 607 |
Feature activation | -2.70e+00 |
Pos logit contributions | |
roller | +0.035 |
scrap | +0.030 |
steps | +0.030 |
rooft | +0.030 |
squee | +0.030 |
Neg logit contributions | |
pole | -0.031 |
sharing | -0.030 |
isk | -0.030 |
Freddie | -0.029 |
ole | -0.029 |
Token | teeth |
Token ID | 9941 |
Feature activation | -6.40e+00 |
Pos logit contributions | |
roller | +0.050 |
leaf | +0.046 |
col | +0.045 |
rooft | +0.044 |
Jo | +0.042 |
Neg logit contributions | |
isk | -0.064 |
being | -0.064 |
sharing | -0.063 |
Mickey | -0.062 |
ending | -0.060 |
Token | and |
Token ID | 290 |
Feature activation | -7.81e-01 |
Pos logit contributions | |
roller | +0.162 |
scrap | +0.146 |
levers | +0.144 |
squee | +0.144 |
rooft | +0.144 |
Neg logit contributions | |
pole | -0.114 |
sharing | -0.111 |
isk | -0.108 |
ole | -0.105 |
Freddie | -0.104 |
Token | she |
Token ID | 673 |
Feature activation | -1.29e-01 |
Pos logit contributions | |
roller | +0.019 |
rooft | +0.016 |
scrap | +0.016 |
palm | +0.016 |
steps | +0.016 |
Neg logit contributions | |
sharing | -0.015 |
pole | -0.015 |
isk | -0.014 |
Freddie | -0.014 |
being | -0.014 |
Token | was |
Token ID | 373 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
roller | +0.003 |
rooft | +0.003 |
levers | +0.003 |
leaf | +0.003 |
palm | +0.003 |
Neg logit contributions | |
pole | -0.002 |
sharing | -0.002 |
isk | -0.002 |
being | -0.002 |
Freddie | -0.002 |
Token | amazed |
Token ID | 24562 |
Feature activation | +6.97e-01 |
Pos logit contributions | |
roller | +0.033 |
scrap | +0.030 |
squee | +0.030 |
steps | +0.030 |
levers | +0.029 |
Neg logit contributions | |
pole | -0.023 |
sharing | -0.022 |
isk | -0.022 |
ole | -0.021 |
Freddie | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | -5.62e-01 |
Pos logit contributions | |
isk | +0.016 |
pole | +0.015 |
sharing | +0.014 |
Freddie | +0.014 |
py | +0.014 |
Neg logit contributions | |
roller | -0.014 |
squee | -0.013 |
steps | -0.013 |
wonder | -0.013 |
rooft | -0.013 |
Token | , |
Token ID | 11 |
Feature activation | +5.28e-01 |
Pos logit contributions | |
sharing | +0.055 |
pole | +0.053 |
isk | +0.052 |
Freddie | +0.051 |
ole | +0.051 |
Neg logit contributions | |
roller | -0.062 |
scrap | -0.055 |
levers | -0.053 |
rooft | -0.053 |
squee | -0.053 |
Token | " |
Token ID | 366 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
pole | +0.009 |
isk | +0.009 |
sharing | +0.009 |
ole | +0.009 |
Freddie | +0.008 |
Neg logit contributions | |
roller | -0.013 |
scrap | -0.012 |
squee | -0.012 |
steps | -0.012 |
rooft | -0.012 |
Token | Mill |
Token ID | 22603 |
Feature activation | +3.89e+00 |
Pos logit contributions | |
roller | +0.028 |
scrap | +0.024 |
steps | +0.024 |
squee | +0.024 |
levers | +0.024 |
Neg logit contributions | |
pole | -0.036 |
isk | -0.035 |
sharing | -0.035 |
ole | -0.033 |
Freddie | -0.033 |
Token | ie |
Token ID | 494 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
sharing | +0.034 |
pole | +0.034 |
isk | +0.032 |
Freddie | +0.030 |
ole | +0.028 |
Neg logit contributions | |
roller | -0.130 |
palm | -0.121 |
steps | -0.121 |
rooft | -0.121 |
squee | -0.120 |
Token | , |
Token ID | 11 |
Feature activation | +1.68e+00 |
Pos logit contributions | |
pole | +0.043 |
sharing | +0.040 |
isk | +0.039 |
ole | +0.039 |
ingu | +0.038 |
Neg logit contributions | |
roller | -0.038 |
squee | -0.034 |
steps | -0.034 |
palm | -0.033 |
bars | -0.032 |
Token | come |
Token ID | 1282 |
Feature activation | +4.04e+00 |
Pos logit contributions | |
pole | +0.034 |
ole | +0.032 |
isk | +0.032 |
sharing | +0.031 |
Freddie | +0.031 |
Neg logit contributions | |
roller | -0.039 |
steps | -0.035 |
squee | -0.035 |
palm | -0.035 |
wonder | -0.035 |
Token | home |
Token ID | 1363 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
pole | +0.081 |
sharing | +0.079 |
isk | +0.078 |
Freddie | +0.075 |
ole | +0.075 |
Neg logit contributions | |
roller | -0.094 |
scrap | -0.083 |
squee | -0.082 |
steps | -0.082 |
levers | -0.081 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
pole | +0.029 |
isk | +0.027 |
sharing | +0.027 |
Freddie | +0.026 |
ole | +0.026 |
Neg logit contributions | |
roller | -0.029 |
steps | -0.026 |
squee | -0.025 |
rooft | -0.025 |
palm | -0.025 |
Token | But |
Token ID | 887 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
roller | +0.042 |
scrap | +0.037 |
squee | +0.035 |
lace | +0.035 |
seats | +0.035 |
Neg logit contributions | |
pole | -0.030 |
Mickey | -0.029 |
sharing | -0.029 |
being | -0.028 |
isk | -0.027 |
Token | Mill |
Token ID | 9212 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
roller | +0.002 |
wonder | +0.002 |
palm | +0.002 |
rooft | +0.002 |
levers | +0.002 |
Neg logit contributions | |
isk | -0.004 |
pole | -0.004 |
sharing | -0.004 |
ole | -0.003 |
ingu | -0.003 |
Token | ie |
Token ID | 494 |
Feature activation | -8.72e-01 |
Pos logit contributions | |
sharing | +0.017 |
pole | +0.016 |
isk | +0.015 |
Freddie | +0.014 |
ole | +0.014 |
Neg logit contributions | |
roller | -0.068 |
scrap | -0.064 |
palm | -0.063 |
steps | -0.063 |
squee | -0.063 |
Token | explore |
Token ID | 7301 |
Feature activation | +2.31e+00 |
Pos logit contributions | |
sharing | +0.019 |
Freddie | +0.017 |
Mickey | +0.017 |
being | +0.017 |
isk | +0.016 |
Neg logit contributions | |
roller | -0.025 |
rooft | -0.024 |
levers | -0.024 |
scrap | -0.023 |
obs | -0.023 |
Token | and |
Token ID | 290 |
Feature activation | -2.45e-01 |
Pos logit contributions | |
sharing | +0.036 |
being | +0.034 |
Mickey | +0.031 |
pole | +0.030 |
Freddie | +0.030 |
Neg logit contributions | |
roller | -0.064 |
scrap | -0.061 |
levers | -0.058 |
Ah | -0.058 |
squee | -0.058 |
Token | always |
Token ID | 1464 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
roller | +0.006 |
obs | +0.006 |
levers | +0.006 |
bisc | +0.005 |
squee | +0.005 |
Neg logit contributions | |
sharing | -0.005 |
being | -0.005 |
ending | -0.004 |
always | -0.004 |
end | -0.004 |
Token | wondered |
Token ID | 14028 |
Feature activation | +9.00e-01 |
Pos logit contributions | |
sharing | +0.035 |
being | +0.033 |
pole | +0.032 |
isk | +0.032 |
ending | +0.031 |
Neg logit contributions | |
roller | -0.038 |
rooft | -0.034 |
levers | -0.034 |
bars | -0.033 |
squee | -0.032 |
Token | what |
Token ID | 644 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
pole | +0.014 |
sharing | +0.013 |
isk | +0.013 |
Freddie | +0.012 |
ole | +0.012 |
Neg logit contributions | |
roller | -0.025 |
scrap | -0.023 |
squee | -0.022 |
steps | -0.022 |
levers | -0.022 |
Token | it |
Token ID | 340 |
Feature activation | +3.92e+00 |
Pos logit contributions | |
roller | +0.029 |
rooft | +0.026 |
levers | +0.025 |
obs | +0.025 |
squee | +0.025 |
Neg logit contributions | |
sharing | -0.025 |
being | -0.024 |
pole | -0.023 |
Mickey | -0.022 |
Freddie | -0.022 |
Token | was |
Token ID | 373 |
Feature activation | +3.07e+00 |
Pos logit contributions | |
pole | +0.091 |
sharing | +0.089 |
ole | +0.088 |
isk | +0.087 |
Freddie | +0.085 |
Neg logit contributions | |
roller | -0.078 |
scrap | -0.070 |
steps | -0.069 |
squee | -0.068 |
wonder | -0.067 |
Token | like |
Token ID | 588 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
isk | +0.077 |
pole | +0.074 |
Freddie | +0.074 |
ole | +0.072 |
Mickey | +0.070 |
Neg logit contributions | |
roller | -0.055 |
wonder | -0.049 |
steps | -0.049 |
palm | -0.047 |
seats | -0.046 |
Token | to |
Token ID | 284 |
Feature activation | +9.91e-01 |
Pos logit contributions | |
sharing | +0.028 |
pole | +0.028 |
isk | +0.027 |
being | +0.026 |
Freddie | +0.025 |
Neg logit contributions | |
roller | -0.029 |
scrap | -0.027 |
levers | -0.026 |
squee | -0.026 |
rooft | -0.025 |
Token | travel |
Token ID | 3067 |
Feature activation | +2.24e+00 |
Pos logit contributions | |
end | +0.017 |
sharing | +0.016 |
being | +0.016 |
ending | +0.015 |
pole | +0.015 |
Neg logit contributions | |
rooft | -0.025 |
roller | -0.025 |
squee | -0.024 |
obs | -0.024 |
col | -0.024 |
Token | in |
Token ID | 287 |
Feature activation | +7.16e-01 |
Pos logit contributions | |
pole | +0.045 |
sharing | +0.044 |
isk | +0.043 |
ole | +0.042 |
Freddie | +0.041 |
Neg logit contributions | |
roller | -0.052 |
scrap | -0.046 |
squee | -0.046 |
palm | -0.045 |
steps | -0.045 |
Token | brave |
Token ID | 14802 |
Feature activation | +2.84e-01 |
Pos logit contributions | |
roller | +0.003 |
squee | +0.003 |
scrap | +0.003 |
steps | +0.003 |
palm | +0.003 |
Neg logit contributions | |
pole | -0.003 |
isk | -0.003 |
sharing | -0.003 |
ole | -0.002 |
Freddie | -0.002 |
Token | dog |
Token ID | 3290 |
Feature activation | +2.03e+00 |
Pos logit contributions | |
pole | +0.006 |
isk | +0.006 |
sharing | +0.006 |
ole | +0.006 |
Freddie | +0.006 |
Neg logit contributions | |
roller | -0.006 |
squee | -0.005 |
steps | -0.005 |
scrap | -0.005 |
palm | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | +1.58e+00 |
Pos logit contributions | |
pole | +0.051 |
isk | +0.051 |
sharing | +0.050 |
ole | +0.049 |
ending | +0.048 |
Neg logit contributions | |
roller | -0.035 |
bars | -0.030 |
squee | -0.030 |
steps | -0.030 |
scrap | -0.029 |
Token | You |
Token ID | 921 |
Feature activation | -6.19e-01 |
Pos logit contributions | |
pole | +0.026 |
ole | +0.025 |
isk | +0.025 |
sharing | +0.024 |
melon | +0.022 |
Neg logit contributions | |
roller | -0.042 |
squee | -0.039 |
scrap | -0.038 |
lace | -0.038 |
rooft | -0.038 |
Token | are |
Token ID | 389 |
Feature activation | +2.50e+00 |
Pos logit contributions | |
roller | +0.012 |
wonder | +0.011 |
squee | +0.011 |
scrap | +0.011 |
steps | +0.011 |
Neg logit contributions | |
pole | -0.015 |
isk | -0.014 |
sharing | -0.014 |
ole | -0.014 |
Freddie | -0.014 |
Token | almost |
Token ID | 2048 |
Feature activation | +3.70e+00 |
Pos logit contributions | |
isk | +0.074 |
ole | +0.071 |
ending | +0.070 |
pole | +0.069 |
icking | +0.069 |
Neg logit contributions | |
roller | -0.034 |
g | -0.030 |
squee | -0.030 |
wonder | -0.029 |
mo | -0.029 |
Token | done |
Token ID | 1760 |
Feature activation | +1.37e+00 |
Pos logit contributions | |
pole | +0.076 |
isk | +0.074 |
End | +0.070 |
sharing | +0.069 |
ending | +0.069 |
Neg logit contributions | |
roller | -0.085 |
scrap | -0.074 |
squee | -0.073 |
mo | -0.071 |
steps | -0.070 |
Token | . |
Token ID | 13 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
sharing | +0.027 |
pole | +0.025 |
Freddie | +0.025 |
being | +0.025 |
Mickey | +0.024 |
Neg logit contributions | |
roller | -0.032 |
palm | -0.030 |
steps | -0.030 |
levers | -0.029 |
scrap | -0.029 |
Token | Just |
Token ID | 2329 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
pole | +0.024 |
ole | +0.023 |
isk | +0.022 |
sharing | +0.022 |
Freddie | +0.020 |
Neg logit contributions | |
roller | -0.038 |
squee | -0.035 |
scrap | -0.034 |
rooft | -0.034 |
steps | -0.034 |
Token | a |
Token ID | 257 |
Feature activation | -5.57e-01 |
Pos logit contributions | |
pole | +0.030 |
ole | +0.029 |
isk | +0.028 |
ator | +0.028 |
ingu | +0.027 |
Neg logit contributions | |
roller | -0.025 |
squee | -0.023 |
wonder | -0.023 |
steps | -0.021 |
scrap | -0.021 |
Token | little |
Token ID | 1310 |
Feature activation | +2.66e-01 |
Pos logit contributions | |
roller | +0.012 |
squee | +0.011 |
scrap | +0.011 |
steps | +0.011 |
palm | +0.011 |
Neg logit contributions | |
pole | -0.012 |
isk | -0.012 |
sharing | -0.011 |
Freddie | -0.011 |
ole | -0.011 |
Token | collar |
Token ID | 19908 |
Feature activation | -1.83e+00 |
Pos logit contributions | |
roller | +0.008 |
leaf | +0.008 |
Jo | +0.007 |
rooft | +0.007 |
col | +0.007 |
Neg logit contributions | |
sharing | -0.008 |
pole | -0.008 |
ending | -0.008 |
ole | -0.008 |
isk | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | -6.47e-03 |
Pos logit contributions | |
roller | +0.040 |
scrap | +0.035 |
steps | +0.035 |
squee | +0.035 |
levers | +0.035 |
Neg logit contributions | |
pole | -0.039 |
sharing | -0.038 |
isk | -0.038 |
Freddie | -0.036 |
ole | -0.036 |
Token |
|
Token ID | 198 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
roller | +0.000 |
scrap | +0.000 |
steps | +0.000 |
levers | +0.000 |
rooft | +0.000 |
Neg logit contributions | |
sharing | -0.000 |
pole | -0.000 |
Freddie | -0.000 |
isk | -0.000 |
ole | -0.000 |
Token |
|
Token ID | 198 |
Feature activation | +2.77e+00 |
Pos logit contributions | |
sharing | +0.013 |
ole | +0.012 |
isk | +0.012 |
Freddie | +0.012 |
Mickey | +0.011 |
Neg logit contributions | |
roller | -0.031 |
lace | -0.028 |
levers | -0.028 |
scrap | -0.027 |
rooft | -0.027 |
Token | " |
Token ID | 1 |
Feature activation | -2.67e+00 |
Pos logit contributions | |
pole | +0.055 |
sharing | +0.054 |
isk | +0.053 |
ole | +0.051 |
Freddie | +0.051 |
Neg logit contributions | |
roller | -0.065 |
scrap | -0.058 |
squee | -0.057 |
steps | -0.057 |
levers | -0.057 |
Token | Uh |
Token ID | 34653 |
Feature activation | +4.11e+00 |
Pos logit contributions | |
roller | +0.050 |
steps | +0.043 |
scrap | +0.042 |
squee | +0.042 |
levers | +0.042 |
Neg logit contributions | |
pole | -0.066 |
sharing | -0.065 |
isk | -0.064 |
Freddie | -0.062 |
ole | -0.062 |
Token | - |
Token ID | 12 |
Feature activation | -3.87e-01 |
Pos logit contributions | |
pole | +0.102 |
sharing | +0.102 |
isk | +0.099 |
ole | +0.097 |
Freddie | +0.095 |
Neg logit contributions | |
roller | -0.075 |
rooft | -0.065 |
scrap | -0.065 |
levers | -0.064 |
squee | -0.064 |
Token | oh |
Token ID | 1219 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
wonder | +0.010 |
roller | +0.010 |
amaz | +0.009 |
levers | +0.009 |
seats | +0.009 |
Neg logit contributions | |
ole | -0.007 |
isk | -0.007 |
pole | -0.007 |
ter | -0.006 |
sharing | -0.006 |
Token | ," |
Token ID | 553 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
pole | +0.021 |
sharing | +0.020 |
isk | +0.020 |
ole | +0.020 |
Freddie | +0.019 |
Neg logit contributions | |
roller | -0.024 |
levers | -0.021 |
rooft | -0.021 |
scrap | -0.021 |
palm | -0.021 |
Token | Tom |
Token ID | 4186 |
Feature activation | -2.25e+00 |
Pos logit contributions | |
roller | +0.032 |
rooft | +0.028 |
scrap | +0.028 |
squee | +0.027 |
steps | +0.027 |
Neg logit contributions | |
sharing | -0.030 |
pole | -0.030 |
Freddie | -0.029 |
ole | -0.028 |
isk | -0.028 |
Token | says |
Token ID | 1139 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
roller | +0.044 |
rooft | +0.037 |
scrap | +0.037 |
palm | +0.036 |
col | +0.035 |
Neg logit contributions | |
pole | -0.054 |
sharing | -0.052 |
isk | -0.051 |
Freddie | -0.051 |
being | -0.050 |
Token | while |
Token ID | 981 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
pole | +0.001 |
sharing | +0.001 |
isk | +0.001 |
Freddie | +0.001 |
ole | +0.001 |
Neg logit contributions | |
roller | -0.001 |
scrap | -0.001 |
squee | -0.001 |
levers | -0.001 |
steps | -0.001 |
Token | , |
Token ID | 11 |
Feature activation | -5.66e-01 |
Pos logit contributions | |
roller | +0.029 |
steps | +0.025 |
seats | +0.024 |
squee | +0.024 |
scrap | +0.024 |
Neg logit contributions | |
pole | -0.031 |
isk | -0.030 |
melon | -0.028 |
Freddie | -0.028 |
ole | -0.028 |
Token | it |
Token ID | 340 |
Feature activation | +2.12e+00 |
Pos logit contributions | |
roller | +0.013 |
scrap | +0.011 |
squee | +0.011 |
steps | +0.011 |
levers | +0.011 |
Neg logit contributions | |
pole | -0.012 |
isk | -0.011 |
sharing | -0.011 |
ole | -0.011 |
Freddie | -0.011 |
Token | would |
Token ID | 561 |
Feature activation | +2.43e-01 |
Pos logit contributions | |
sharing | +0.050 |
pole | +0.050 |
ole | +0.049 |
isk | +0.049 |
Freddie | +0.048 |
Neg logit contributions | |
roller | -0.040 |
steps | -0.037 |
scrap | -0.037 |
squee | -0.037 |
wonder | -0.036 |
Token | always |
Token ID | 1464 |
Feature activation | +2.19e+00 |
Pos logit contributions | |
pole | +0.005 |
ole | +0.005 |
Freddie | +0.005 |
isk | +0.005 |
sharing | +0.005 |
Neg logit contributions | |
roller | -0.005 |
wonder | -0.005 |
scrap | -0.005 |
palm | -0.005 |
squee | -0.005 |
Token | come |
Token ID | 1282 |
Feature activation | +4.31e+00 |
Pos logit contributions | |
pole | +0.041 |
ole | +0.040 |
isk | +0.039 |
Freddie | +0.039 |
sharing | +0.039 |
Neg logit contributions | |
roller | -0.053 |
wonder | -0.050 |
scrap | -0.049 |
steps | -0.048 |
squee | -0.048 |
Token | back |
Token ID | 736 |
Feature activation | +2.68e+00 |
Pos logit contributions | |
pole | +0.085 |
sharing | +0.083 |
isk | +0.083 |
Freddie | +0.080 |
ole | +0.078 |
Neg logit contributions | |
roller | -0.100 |
rooft | -0.088 |
scrap | -0.088 |
palm | -0.088 |
squee | -0.088 |
Token | to |
Token ID | 284 |
Feature activation | +9.14e-02 |
Pos logit contributions | |
pole | +0.055 |
sharing | +0.054 |
isk | +0.053 |
Freddie | +0.051 |
ole | +0.051 |
Neg logit contributions | |
roller | -0.061 |
scrap | -0.054 |
squee | -0.053 |
steps | -0.053 |
levers | -0.053 |
Token | make |
Token ID | 787 |
Feature activation | -4.29e-01 |
Pos logit contributions | |
pole | +0.002 |
sharing | +0.002 |
isk | +0.002 |
ole | +0.002 |
Freddie | +0.002 |
Neg logit contributions | |
roller | -0.002 |
scrap | -0.002 |
squee | -0.002 |
steps | -0.002 |
rooft | -0.002 |
Token | her |
Token ID | 607 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
roller | +0.010 |
levers | +0.009 |
scrap | +0.009 |
palm | +0.009 |
rooft | +0.009 |
Neg logit contributions | |
sharing | -0.008 |
pole | -0.008 |
being | -0.008 |
isk | -0.008 |
Freddie | -0.008 |
Token | warm |
Token ID | 5814 |
Feature activation | +5.26e-01 |
Pos logit contributions | |
roller | +0.028 |
levers | +0.025 |
squee | +0.024 |
scrap | +0.024 |
rooft | +0.024 |
Neg logit contributions | |
sharing | -0.020 |
pole | -0.020 |
being | -0.018 |
isk | -0.018 |
ending | -0.018 |
Token |
|
Token ID | 198 |
Feature activation | +2.30e+00 |
Pos logit contributions | |
isk | +0.005 |
Freddie | +0.005 |
sharing | +0.005 |
pole | +0.005 |
Mickey | +0.005 |
Neg logit contributions | |
roller | -0.012 |
levers | -0.012 |
umb | -0.012 |
obs | -0.012 |
lace | -0.011 |
Token | At |
Token ID | 2953 |
Feature activation | +2.81e+00 |
Pos logit contributions | |
pole | +0.051 |
sharing | +0.050 |
isk | +0.049 |
ole | +0.047 |
Freddie | +0.047 |
Neg logit contributions | |
roller | -0.049 |
scrap | -0.043 |
steps | -0.042 |
squee | -0.042 |
levers | -0.042 |
Token | the |
Token ID | 262 |
Feature activation | +3.80e-01 |
Pos logit contributions | |
pole | +0.060 |
isk | +0.057 |
sharing | +0.056 |
ole | +0.054 |
Freddie | +0.053 |
Neg logit contributions | |
roller | -0.065 |
steps | -0.057 |
squee | -0.056 |
scrap | -0.056 |
palm | -0.055 |
Token | party |
Token ID | 2151 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
pole | +0.006 |
Freddie | +0.006 |
ole | +0.005 |
sharing | +0.005 |
isk | +0.005 |
Neg logit contributions | |
roller | -0.011 |
desks | -0.010 |
seats | -0.010 |
wonder | -0.010 |
palm | -0.010 |
Token | , |
Token ID | 11 |
Feature activation | -6.28e-01 |
Pos logit contributions | |
isk | +0.029 |
pole | +0.029 |
py | +0.028 |
ility | +0.028 |
ole | +0.028 |
Neg logit contributions | |
roller | -0.027 |
steps | -0.024 |
seats | -0.024 |
bars | -0.023 |
desks | -0.022 |
Token | the |
Token ID | 262 |
Feature activation | +9.09e-01 |
Pos logit contributions | |
roller | +0.014 |
palm | +0.012 |
steps | +0.012 |
levers | +0.012 |
seats | +0.012 |
Neg logit contributions | |
isk | -0.013 |
pole | -0.013 |
sharing | -0.013 |
ole | -0.012 |
py | -0.012 |
Token | giant |
Token ID | 6175 |
Feature activation | -6.78e-01 |
Pos logit contributions | |
pole | +0.023 |
sharing | +0.022 |
isk | +0.022 |
Freddie | +0.021 |
ole | +0.021 |
Neg logit contributions | |
roller | -0.017 |
scrap | -0.014 |
squee | -0.014 |
levers | -0.014 |
lace | -0.014 |
Token | shared |
Token ID | 4888 |
Feature activation | +7.63e-01 |
Pos logit contributions | |
roller | +0.014 |
bars | +0.012 |
steps | +0.012 |
seats | +0.012 |
scrap | +0.012 |
Neg logit contributions | |
sharing | -0.015 |
isk | -0.015 |
pole | -0.015 |
ole | -0.015 |
Freddie | -0.015 |
Token | all |
Token ID | 477 |
Feature activation | -2.30e-01 |
Pos logit contributions | |
sharing | +0.014 |
pole | +0.013 |
being | +0.012 |
isk | +0.012 |
ole | +0.012 |
Neg logit contributions | |
roller | -0.019 |
scrap | -0.018 |
squee | -0.018 |
rooft | -0.018 |
levers | -0.018 |
Token | the |
Token ID | 262 |
Feature activation | -5.11e-01 |
Pos logit contributions | |
roller | +0.006 |
scrap | +0.005 |
steps | +0.005 |
palm | +0.005 |
squee | +0.005 |
Neg logit contributions | |
pole | -0.004 |
isk | -0.004 |
sharing | -0.004 |
Freddie | -0.004 |
ole | -0.004 |
Token | things |
Token ID | 1243 |
Feature activation | -1.95e+00 |
Pos logit contributions | |
roller | +0.012 |
scrap | +0.011 |
squee | +0.010 |
rooft | +0.010 |
levers | +0.010 |
Neg logit contributions | |
pole | -0.010 |
sharing | -0.010 |
isk | -0.010 |
ole | -0.010 |
Freddie | -0.009 |
Token | jump |
Token ID | 4391 |
Feature activation | +6.10e-01 |
Pos logit contributions | |
pole | +0.023 |
sharing | +0.022 |
isk | +0.022 |
Freddie | +0.022 |
Mickey | +0.021 |
Neg logit contributions | |
roller | -0.028 |
rooft | -0.025 |
scrap | -0.025 |
levers | -0.025 |
obs | -0.024 |
Token | right |
Token ID | 826 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
pole | +0.010 |
Freddie | +0.009 |
sharing | +0.009 |
ole | +0.009 |
isk | +0.009 |
Neg logit contributions | |
roller | -0.016 |
desks | -0.015 |
rooft | -0.015 |
scrap | -0.015 |
seats | -0.015 |
Token | in |
Token ID | 287 |
Feature activation | -3.25e-01 |
Pos logit contributions | |
pole | +0.017 |
ter | +0.016 |
sharing | +0.016 |
being | +0.016 |
ole | +0.016 |
Neg logit contributions | |
roller | -0.031 |
scrap | -0.029 |
desks | -0.028 |
leaf | -0.028 |
levers | -0.028 |
Token | and |
Token ID | 290 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
roller | +0.008 |
obs | +0.008 |
leaf | +0.008 |
levers | +0.008 |
scrap | +0.008 |
Neg logit contributions | |
sharing | -0.005 |
being | -0.005 |
Freddie | -0.005 |
pole | -0.005 |
ole | -0.005 |
Token | splash |
Token ID | 22870 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
pole | +0.003 |
sharing | +0.003 |
Freddie | +0.003 |
Mickey | +0.003 |
isk | +0.003 |
Neg logit contributions | |
roller | -0.004 |
rooft | -0.003 |
obs | -0.003 |
levers | -0.003 |
seats | -0.003 |
Token | around |
Token ID | 1088 |
Feature activation | +2.58e-01 |
Pos logit contributions | |
pole | +0.019 |
sharing | +0.018 |
Freddie | +0.018 |
ole | +0.018 |
being | +0.017 |
Neg logit contributions | |
roller | -0.028 |
rooft | -0.025 |
scrap | -0.024 |
leaf | -0.024 |
squee | -0.024 |
Token | , |
Token ID | 11 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
pole | +0.005 |
sharing | +0.005 |
isk | +0.005 |
ole | +0.005 |
Freddie | +0.005 |
Neg logit contributions | |
roller | -0.006 |
scrap | -0.005 |
steps | -0.005 |
squee | -0.005 |
rooft | -0.005 |
Token | making |
Token ID | 1642 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
roller | +0.034 |
steps | +0.031 |
scrap | +0.030 |
squee | +0.030 |
palm | +0.030 |
Neg logit contributions | |
pole | -0.028 |
isk | -0.027 |
ole | -0.027 |
sharing | -0.027 |
Freddie | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | -2.55e+00 |
Pos logit contributions | |
roller | +0.034 |
rooft | +0.031 |
obs | +0.030 |
scrap | +0.030 |
palm | +0.029 |
Neg logit contributions | |
sharing | -0.026 |
being | -0.025 |
pole | -0.025 |
Freddie | -0.024 |
isk | -0.023 |
Token | big |
Token ID | 1263 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
roller | +0.052 |
leaf | +0.048 |
levers | +0.047 |
rooft | +0.046 |
col | +0.046 |
Neg logit contributions | |
pole | -0.055 |
sharing | -0.055 |
ole | -0.054 |
ending | -0.052 |
isk | -0.051 |
Token | mess |
Token ID | 2085 |
Feature activation | +1.27e+00 |
Pos logit contributions | |
roller | +0.038 |
levers | +0.035 |
scrap | +0.034 |
rooft | +0.034 |
leaf | +0.034 |
Neg logit contributions | |
sharing | -0.047 |
pole | -0.047 |
ole | -0.044 |
Freddie | -0.044 |
isk | -0.044 |
Token | Lily |
Token ID | 20037 |
Feature activation | -9.37e-01 |
Pos logit contributions | |
pole | +0.002 |
isk | +0.002 |
sharing | +0.002 |
ole | +0.002 |
py | +0.002 |
Neg logit contributions | |
roller | -0.002 |
scrap | -0.001 |
palm | -0.001 |
steps | -0.001 |
squee | -0.001 |
Token | pulled |
Token ID | 5954 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
roller | +0.022 |
rooft | +0.020 |
levers | +0.019 |
leaf | +0.019 |
col | +0.019 |
Neg logit contributions | |
pole | -0.018 |
isk | -0.018 |
sharing | -0.018 |
being | -0.017 |
Freddie | -0.017 |
Token | it |
Token ID | 340 |
Feature activation | -8.01e-01 |
Pos logit contributions | |
roller | +0.030 |
leaf | +0.027 |
seats | +0.026 |
rooft | +0.026 |
desks | +0.026 |
Neg logit contributions | |
sharing | -0.023 |
being | -0.023 |
pole | -0.023 |
Freddie | -0.022 |
Mickey | -0.022 |
Token | away |
Token ID | 1497 |
Feature activation | +7.26e-01 |
Pos logit contributions | |
roller | +0.019 |
scrap | +0.017 |
rooft | +0.017 |
squee | +0.017 |
palm | +0.017 |
Neg logit contributions | |
pole | -0.016 |
sharing | -0.015 |
isk | -0.015 |
ole | -0.014 |
Freddie | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | +4.97e-01 |
Pos logit contributions | |
pole | +0.014 |
sharing | +0.014 |
isk | +0.013 |
ole | +0.013 |
Freddie | +0.013 |
Neg logit contributions | |
roller | -0.017 |
scrap | -0.015 |
squee | -0.015 |
steps | -0.015 |
rooft | -0.015 |
Token |
|
Token ID | 198 |
Feature activation | +8.94e-01 |
Pos logit contributions | |
pole | +0.010 |
sharing | +0.010 |
isk | +0.010 |
Freddie | +0.010 |
ole | +0.010 |
Neg logit contributions | |
roller | -0.011 |
scrap | -0.010 |
steps | -0.010 |
levers | -0.010 |
squee | -0.010 |
Token |
|
Token ID | 198 |
Feature activation | +2.67e+00 |
Pos logit contributions | |
pole | +0.016 |
isk | +0.016 |
sharing | +0.016 |
ole | +0.016 |
Freddie | +0.016 |
Neg logit contributions | |
roller | -0.022 |
scrap | -0.020 |
rooft | -0.019 |
levers | -0.019 |
squee | -0.019 |
Token | " |
Token ID | 1 |
Feature activation | -3.01e+00 |
Pos logit contributions | |
pole | +0.048 |
sharing | +0.048 |
isk | +0.047 |
Freddie | +0.045 |
ending | +0.044 |
Neg logit contributions | |
roller | -0.067 |
scrap | -0.060 |
steps | -0.060 |
levers | -0.059 |
palm | -0.059 |
Token | No |
Token ID | 2949 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
roller | +0.067 |
seats | +0.060 |
desks | +0.060 |
steps | +0.059 |
rooft | +0.059 |
Neg logit contributions | |
Freddie | -0.056 |
pole | -0.055 |
End | -0.055 |
isk | -0.054 |
sharing | -0.054 |
Token | , |
Token ID | 11 |
Feature activation | +1.60e+00 |
Pos logit contributions | |
roller | +0.030 |
scrap | +0.027 |
rooft | +0.027 |
levers | +0.027 |
squee | +0.027 |
Neg logit contributions | |
pole | -0.018 |
sharing | -0.018 |
isk | -0.017 |
Freddie | -0.017 |
ole | -0.017 |
Token | this |
Token ID | 428 |
Feature activation | +1.54e-01 |
Pos logit contributions | |
pole | +0.031 |
sharing | +0.030 |
isk | +0.030 |
ole | +0.029 |
Freddie | +0.028 |
Neg logit contributions | |
roller | -0.038 |
scrap | -0.034 |
squee | -0.034 |
steps | -0.034 |
palm | -0.034 |
Token | Don |
Token ID | 3987 |
Feature activation | -8.35e-01 |
Pos logit contributions | |
roller | +0.010 |
scrap | +0.009 |
squee | +0.009 |
wonder | +0.008 |
palm | +0.008 |
Neg logit contributions | |
pole | -0.014 |
sharing | -0.014 |
isk | -0.014 |
ole | -0.013 |
Freddie | -0.013 |
Token | 't |
Token ID | 470 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
roller | +0.019 |
steps | +0.018 |
scrap | +0.017 |
squee | +0.017 |
rooft | +0.016 |
Neg logit contributions | |
pole | -0.017 |
sharing | -0.017 |
isk | -0.016 |
melon | -0.015 |
ility | -0.015 |
Token | cry |
Token ID | 3960 |
Feature activation | +1.70e+00 |
Pos logit contributions | |
pole | +0.025 |
isk | +0.024 |
ator | +0.023 |
ole | +0.023 |
ingu | +0.022 |
Neg logit contributions | |
roller | -0.018 |
scrap | -0.018 |
wonder | -0.018 |
squee | -0.017 |
steps | -0.017 |
Token | , |
Token ID | 11 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
pole | +0.039 |
isk | +0.036 |
Freddie | +0.035 |
sharing | +0.034 |
ole | +0.034 |
Neg logit contributions | |
roller | -0.033 |
squee | -0.031 |
steps | -0.030 |
scrap | -0.029 |
rooft | -0.029 |
Token | Sam |
Token ID | 3409 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
pole | +0.038 |
ole | +0.038 |
isk | +0.037 |
sharing | +0.036 |
melon | +0.034 |
Neg logit contributions | |
roller | -0.039 |
squee | -0.034 |
palm | -0.033 |
steps | -0.033 |
scrap | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | +2.47e+00 |
Pos logit contributions | |
pole | +0.057 |
isk | +0.054 |
sharing | +0.053 |
ole | +0.053 |
ingu | +0.052 |
Neg logit contributions | |
roller | -0.031 |
squee | -0.029 |
steps | -0.028 |
scrap | -0.027 |
bars | -0.026 |
Token | I |
Token ID | 314 |
Feature activation | -1.38e+00 |
Pos logit contributions | |
pole | +0.039 |
isk | +0.037 |
sharing | +0.036 |
ole | +0.036 |
Freddie | +0.033 |
Neg logit contributions | |
roller | -0.068 |
squee | -0.064 |
scrap | -0.063 |
palm | -0.062 |
steps | -0.062 |
Token | can |
Token ID | 460 |
Feature activation | -9.04e-01 |
Pos logit contributions | |
roller | +0.030 |
rooft | +0.026 |
levers | +0.026 |
steps | +0.025 |
palm | +0.025 |
Neg logit contributions | |
sharing | -0.030 |
pole | -0.030 |
isk | -0.030 |
ole | -0.029 |
Freddie | -0.028 |
Token | help |
Token ID | 1037 |
Feature activation | +6.16e-02 |
Pos logit contributions | |
roller | +0.019 |
scrap | +0.018 |
squee | +0.017 |
wonder | +0.017 |
steps | +0.017 |
Neg logit contributions | |
pole | -0.019 |
sharing | -0.019 |
isk | -0.019 |
ole | -0.018 |
ending | -0.018 |
Token | you |
Token ID | 345 |
Feature activation | +3.02e-02 |
Pos logit contributions | |
pole | +0.001 |
isk | +0.001 |
sharing | +0.001 |
ole | +0.001 |
ending | +0.001 |
Neg logit contributions | |
roller | -0.002 |
steps | -0.002 |
squee | -0.002 |
scrap | -0.002 |
levers | -0.002 |
Token | fix |
Token ID | 4259 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
pole | +0.001 |
isk | +0.001 |
sharing | +0.001 |
Freddie | +0.001 |
ending | +0.001 |
Neg logit contributions | |
roller | -0.001 |
wonder | -0.001 |
steps | -0.000 |
squee | -0.000 |
bars | -0.000 |
Token | Just |
Token ID | 2329 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
pole | +0.025 |
isk | +0.023 |
sharing | +0.023 |
ole | +0.023 |
Freddie | +0.022 |
Neg logit contributions | |
roller | -0.036 |
squee | -0.033 |
lace | -0.032 |
steps | -0.032 |
rooft | -0.032 |
Token | try |
Token ID | 1949 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
pole | +0.027 |
isk | +0.026 |
ole | +0.026 |
sharing | +0.026 |
Freddie | +0.025 |
Neg logit contributions | |
roller | -0.025 |
wonder | -0.023 |
squee | -0.023 |
palm | -0.022 |
steps | -0.022 |
Token | your |
Token ID | 534 |
Feature activation | -4.22e-01 |
Pos logit contributions | |
sharing | +0.003 |
pole | +0.002 |
being | +0.002 |
Freddie | +0.002 |
isk | +0.002 |
Neg logit contributions | |
roller | -0.003 |
scrap | -0.002 |
levers | -0.002 |
wonder | -0.002 |
rooft | -0.002 |
Token | best |
Token ID | 1266 |
Feature activation | +5.29e-01 |
Pos logit contributions | |
roller | +0.011 |
rooft | +0.010 |
col | +0.010 |
levers | +0.009 |
Jo | +0.009 |
Neg logit contributions | |
sharing | -0.007 |
being | -0.007 |
pole | -0.007 |
ending | -0.007 |
isk | -0.007 |
Token | and |
Token ID | 290 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
sharing | +0.010 |
pole | +0.009 |
ole | +0.009 |
ending | +0.009 |
being | +0.009 |
Neg logit contributions | |
roller | -0.013 |
rooft | -0.012 |
scrap | -0.012 |
squee | -0.012 |
levers | -0.012 |
Token | have |
Token ID | 423 |
Feature activation | +1.44e+00 |
Pos logit contributions | |
sharing | +0.027 |
pole | +0.027 |
isk | +0.025 |
Freddie | +0.025 |
being | +0.025 |
Neg logit contributions | |
roller | -0.043 |
scrap | -0.039 |
rooft | -0.039 |
squee | -0.038 |
palm | -0.038 |
Token | fun |
Token ID | 1257 |
Feature activation | +8.60e-01 |
Pos logit contributions | |
sharing | +0.029 |
pole | +0.028 |
Freddie | +0.027 |
isk | +0.027 |
being | +0.026 |
Neg logit contributions | |
roller | -0.033 |
scrap | -0.029 |
levers | -0.029 |
rooft | -0.029 |
squee | -0.029 |
Token | . |
Token ID | 13 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
sharing | +0.016 |
pole | +0.015 |
being | +0.015 |
ole | +0.014 |
isk | +0.014 |
Neg logit contributions | |
roller | -0.021 |
scrap | -0.019 |
palm | -0.019 |
rooft | -0.019 |
levers | -0.019 |
Token | Remember |
Token ID | 11436 |
Feature activation | +2.15e-01 |
Pos logit contributions | |
pole | +0.021 |
sharing | +0.020 |
isk | +0.020 |
ole | +0.020 |
Freddie | +0.019 |
Neg logit contributions | |
roller | -0.028 |
squee | -0.025 |
scrap | -0.024 |
levers | -0.024 |
rooft | -0.024 |
Token | , |
Token ID | 11 |
Feature activation | +5.74e-01 |
Pos logit contributions | |
pole | +0.005 |
isk | +0.005 |
Freddie | +0.005 |
Mickey | +0.005 |
ole | +0.004 |
Neg logit contributions | |
roller | -0.004 |
steps | -0.004 |
squee | -0.004 |
seats | -0.004 |
palm | -0.004 |
Token | exams |
Token ID | 26420 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
pole | +0.009 |
isk | +0.009 |
ole | +0.009 |
Freddie | +0.008 |
melon | +0.008 |
Neg logit contributions | |
roller | -0.016 |
steps | -0.014 |
squee | -0.014 |
levers | -0.014 |
lace | -0.014 |
Token | the |
Token ID | 262 |
Feature activation | -5.00e-01 |
Pos logit contributions | |
roller | +0.060 |
rooft | +0.052 |
scrap | +0.052 |
squee | +0.050 |
levers | +0.050 |
Neg logit contributions | |
pole | -0.050 |
sharing | -0.050 |
isk | -0.048 |
Freddie | -0.047 |
ole | -0.047 |
Token | rake |
Token ID | 47885 |
Feature activation | -3.36e+00 |
Pos logit contributions | |
roller | +0.008 |
scrap | +0.007 |
steps | +0.007 |
squee | +0.007 |
levers | +0.007 |
Neg logit contributions | |
pole | -0.013 |
sharing | -0.013 |
isk | -0.013 |
Freddie | -0.012 |
ole | -0.012 |
Token | to |
Token ID | 284 |
Feature activation | -5.43e-01 |
Pos logit contributions | |
roller | +0.055 |
steps | +0.047 |
scrap | +0.046 |
squee | +0.046 |
palm | +0.045 |
Neg logit contributions | |
pole | -0.089 |
sharing | -0.088 |
isk | -0.087 |
Freddie | -0.085 |
ole | -0.084 |
Token | make |
Token ID | 787 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
roller | +0.012 |
scrap | +0.010 |
rooft | +0.010 |
squee | +0.010 |
steps | +0.010 |
Neg logit contributions | |
pole | -0.012 |
sharing | -0.012 |
isk | -0.012 |
ole | -0.011 |
Freddie | -0.011 |
Token | the |
Token ID | 262 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
roller | +0.025 |
scrap | +0.023 |
rooft | +0.022 |
palm | +0.022 |
leaf | +0.022 |
Neg logit contributions | |
sharing | -0.018 |
being | -0.018 |
pole | -0.018 |
isk | -0.018 |
Mickey | -0.017 |
Token | pile |
Token ID | 14540 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
roller | +0.002 |
scrap | +0.002 |
levers | +0.002 |
squee | +0.002 |
rooft | +0.002 |
Neg logit contributions | |
pole | -0.003 |
sharing | -0.003 |
isk | -0.003 |
ole | -0.003 |
Freddie | -0.003 |
Token | of |
Token ID | 286 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
roller | +0.025 |
steps | +0.022 |
palm | +0.022 |
squee | +0.022 |
levers | +0.022 |
Neg logit contributions | |
pole | -0.024 |
isk | -0.023 |
sharing | -0.023 |
Freddie | -0.022 |
Mickey | -0.022 |
Token | leaves |
Token ID | 5667 |
Feature activation | -8.82e-01 |
Pos logit contributions | |
sharing | +0.003 |
pole | +0.003 |
isk | +0.003 |
Freddie | +0.003 |
Mickey | +0.003 |
Neg logit contributions | |
roller | -0.002 |
scrap | -0.001 |
rooft | -0.001 |
levers | -0.001 |
squee | -0.001 |
Token | even |
Token ID | 772 |
Feature activation | +6.16e-02 |
Pos logit contributions | |
roller | +0.018 |
steps | +0.016 |
palm | +0.016 |
squee | +0.016 |
bars | +0.016 |
Neg logit contributions | |
isk | -0.020 |
pole | -0.019 |
sharing | -0.019 |
Freddie | -0.019 |
Mickey | -0.018 |
Token | bigger |
Token ID | 5749 |
Feature activation | +3.76e-01 |
Pos logit contributions | |
pole | +0.001 |
isk | +0.001 |
sharing | +0.001 |
Freddie | +0.001 |
ole | +0.001 |
Neg logit contributions | |
roller | -0.001 |
steps | -0.001 |
squee | -0.001 |
scrap | -0.001 |
levers | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
sharing | +0.007 |
pole | +0.007 |
being | +0.006 |
isk | +0.006 |
ole | +0.006 |
Neg logit contributions | |
roller | -0.009 |
scrap | -0.009 |
leaf | -0.008 |
levers | -0.008 |
wonder | -0.008 |
Token | home |
Token ID | 1363 |
Feature activation | -7.39e-01 |
Pos logit contributions | |
sharing | +0.019 |
pole | +0.019 |
isk | +0.018 |
ole | +0.017 |
Freddie | +0.017 |
Neg logit contributions | |
roller | -0.019 |
scrap | -0.016 |
levers | -0.016 |
squee | -0.016 |
rooft | -0.016 |
Token | and |
Token ID | 290 |
Feature activation | -7.82e-01 |
Pos logit contributions | |
roller | +0.015 |
scrap | +0.013 |
steps | +0.013 |
seats | +0.013 |
squee | +0.013 |
Neg logit contributions | |
pole | -0.017 |
Freddie | -0.016 |
sharing | -0.016 |
isk | -0.016 |
ole | -0.016 |
Token | cried |
Token ID | 16896 |
Feature activation | -7.68e-01 |
Pos logit contributions | |
roller | +0.016 |
scrap | +0.014 |
squee | +0.014 |
steps | +0.014 |
palm | +0.014 |
Neg logit contributions | |
pole | -0.018 |
sharing | -0.017 |
isk | -0.017 |
ole | -0.016 |
Freddie | -0.016 |
Token | to |
Token ID | 284 |
Feature activation | +2.01e-01 |
Pos logit contributions | |
roller | +0.019 |
scrap | +0.017 |
steps | +0.017 |
squee | +0.017 |
levers | +0.017 |
Neg logit contributions | |
pole | -0.015 |
isk | -0.014 |
sharing | -0.014 |
ole | -0.013 |
Freddie | -0.013 |
Token | her |
Token ID | 607 |
Feature activation | +8.89e-02 |
Pos logit contributions | |
pole | +0.005 |
ole | +0.004 |
sharing | +0.004 |
isk | +0.004 |
Freddie | +0.004 |
Neg logit contributions | |
roller | -0.004 |
scrap | -0.004 |
steps | -0.004 |
palm | -0.004 |
wonder | -0.004 |
Token | mom |
Token ID | 1995 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
sharing | +0.002 |
pole | +0.002 |
isk | +0.002 |
ole | +0.002 |
Freddie | +0.002 |
Neg logit contributions | |
roller | -0.002 |
scrap | -0.002 |
rooft | -0.002 |
squee | -0.002 |
levers | -0.002 |
Token | about |
Token ID | 546 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
roller | +0.002 |
steps | +0.002 |
bars | +0.002 |
squee | +0.002 |
wonder | +0.002 |
Neg logit contributions | |
pole | -0.002 |
Freddie | -0.002 |
emo | -0.002 |
ole | -0.002 |
ending | -0.002 |
Token | what |
Token ID | 644 |
Feature activation | -2.72e+00 |
Pos logit contributions | |
pole | +0.030 |
isk | +0.028 |
sharing | +0.028 |
ole | +0.028 |
Freddie | +0.027 |
Neg logit contributions | |
roller | -0.030 |
scrap | -0.026 |
squee | -0.026 |
steps | -0.026 |
palm | -0.025 |
Token | happened |
Token ID | 3022 |
Feature activation | +1.46e+00 |
Pos logit contributions | |
roller | +0.049 |
scrap | +0.045 |
squee | +0.042 |
steps | +0.041 |
palm | +0.041 |
Neg logit contributions | |
pole | -0.072 |
isk | -0.069 |
ole | -0.066 |
sharing | -0.064 |
Freddie | -0.063 |
Token | . |
Token ID | 13 |
Feature activation | -5.87e-01 |
Pos logit contributions | |
pole | +0.036 |
isk | +0.035 |
Freddie | +0.033 |
ole | +0.033 |
Mickey | +0.033 |
Neg logit contributions | |
roller | -0.027 |
scrap | -0.024 |
steps | -0.024 |
squee | -0.023 |
wonder | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | +2.56e-02 |
Pos logit contributions | |
roller | +0.012 |
scrap | +0.010 |
squee | +0.010 |
steps | +0.010 |
wonder | +0.010 |
Neg logit contributions | |
pole | -0.014 |
sharing | -0.013 |
isk | -0.013 |
ole | -0.013 |
Freddie | -0.012 |
Token | breakfast |
Token ID | 12607 |
Feature activation | -2.29e+00 |
Pos logit contributions | |
roller | +0.016 |
steps | +0.014 |
palm | +0.014 |
scrap | +0.014 |
squee | +0.014 |
Neg logit contributions | |
pole | -0.016 |
isk | -0.015 |
sharing | -0.015 |
py | -0.014 |
ole | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -1.78e-02 |
Pos logit contributions | |
roller | +0.048 |
steps | +0.041 |
bars | +0.040 |
seats | +0.040 |
palm | +0.040 |
Neg logit contributions | |
pole | -0.051 |
isk | -0.051 |
sharing | -0.050 |
py | -0.048 |
Freddie | -0.048 |
Token | He |
Token ID | 679 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
roller | +0.000 |
scrap | +0.000 |
squee | +0.000 |
steps | +0.000 |
levers | +0.000 |
Neg logit contributions | |
pole | -0.000 |
sharing | -0.000 |
isk | -0.000 |
ole | -0.000 |
Freddie | -0.000 |
Token | quickly |
Token ID | 2952 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
pole | +0.022 |
sharing | +0.022 |
isk | +0.022 |
Freddie | +0.021 |
Mickey | +0.021 |
Neg logit contributions | |
roller | -0.024 |
levers | -0.021 |
rooft | -0.021 |
steps | -0.021 |
squee | -0.021 |
Token | ran |
Token ID | 4966 |
Feature activation | -7.90e-02 |
Pos logit contributions | |
roller | +0.028 |
rooft | +0.024 |
levers | +0.024 |
scrap | +0.024 |
steps | +0.024 |
Neg logit contributions | |
pole | -0.027 |
sharing | -0.027 |
isk | -0.027 |
Freddie | -0.026 |
ole | -0.025 |
Token | outside |
Token ID | 2354 |
Feature activation | -1.49e+00 |
Pos logit contributions | |
roller | +0.002 |
scrap | +0.001 |
levers | +0.001 |
palm | +0.001 |
squee | +0.001 |
Neg logit contributions | |
pole | -0.002 |
sharing | -0.002 |
isk | -0.002 |
Freddie | -0.002 |
ole | -0.002 |
Token | to |
Token ID | 284 |
Feature activation | -1.65e+00 |
Pos logit contributions | |
roller | +0.029 |
scrap | +0.025 |
squee | +0.024 |
steps | +0.024 |
levers | +0.024 |
Neg logit contributions | |
pole | -0.036 |
sharing | -0.035 |
isk | -0.035 |
Freddie | -0.034 |
ole | -0.033 |
Token | play |
Token ID | 711 |
Feature activation | +1.35e+00 |
Pos logit contributions | |
roller | +0.041 |
levers | +0.037 |
obs | +0.037 |
rooft | +0.036 |
lace | +0.036 |
Neg logit contributions | |
sharing | -0.030 |
isk | -0.029 |
Freddie | -0.029 |
pole | -0.029 |
being | -0.028 |
Token | with |
Token ID | 351 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
pole | +0.028 |
isk | +0.027 |
sharing | +0.026 |
ole | +0.026 |
Freddie | +0.025 |
Neg logit contributions | |
roller | -0.031 |
scrap | -0.027 |
steps | -0.027 |
squee | -0.027 |
bars | -0.027 |
Token | his |
Token ID | 465 |
Feature activation | -3.65e-01 |
Pos logit contributions | |
sharing | +0.020 |
Freddie | +0.020 |
pole | +0.020 |
isk | +0.019 |
Mickey | +0.019 |
Neg logit contributions | |
roller | -0.046 |
rooft | -0.042 |
levers | -0.041 |
squee | -0.041 |
scrap | -0.041 |
Token | friends |
Token ID | 2460 |
Feature activation | +2.46e-02 |
Pos logit contributions | |
roller | +0.009 |
scrap | +0.008 |
squee | +0.008 |
steps | +0.008 |
palm | +0.008 |
Neg logit contributions | |
pole | -0.007 |
sharing | -0.006 |
isk | -0.006 |
Freddie | -0.006 |
ole | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | +4.25e-01 |
Pos logit contributions | |
roller | +0.016 |
steps | +0.016 |
bars | +0.015 |
palm | +0.015 |
squee | +0.015 |
Neg logit contributions | |
isk | -0.026 |
py | -0.025 |
pole | -0.024 |
ending | -0.024 |
sharing | -0.024 |
Token | You |
Token ID | 921 |
Feature activation | -1.50e+00 |
Pos logit contributions | |
isk | +0.010 |
pole | +0.009 |
sharing | +0.009 |
ole | +0.009 |
ending | +0.009 |
Neg logit contributions | |
roller | -0.008 |
squee | -0.007 |
palm | -0.007 |
rooft | -0.007 |
wonder | -0.007 |
Token | can |
Token ID | 460 |
Feature activation | -8.87e-01 |
Pos logit contributions | |
roller | +0.027 |
wonder | +0.026 |
scrap | +0.025 |
steps | +0.025 |
bars | +0.025 |
Neg logit contributions | |
sharing | -0.035 |
pole | -0.035 |
isk | -0.035 |
ole | -0.034 |
ending | -0.033 |
Token | each |
Token ID | 1123 |
Feature activation | +4.67e-01 |
Pos logit contributions | |
roller | +0.020 |
wonder | +0.018 |
squee | +0.017 |
palm | +0.017 |
steps | +0.017 |
Neg logit contributions | |
pole | -0.018 |
isk | -0.018 |
sharing | -0.017 |
ending | -0.017 |
ole | -0.017 |
Token | take |
Token ID | 1011 |
Feature activation | +7.04e-02 |
Pos logit contributions | |
isk | +0.011 |
pole | +0.010 |
ility | +0.010 |
sharing | +0.010 |
py | +0.010 |
Neg logit contributions | |
roller | -0.009 |
palm | -0.009 |
steps | -0.008 |
wonder | -0.008 |
pine | -0.008 |
Token | one |
Token ID | 530 |
Feature activation | -2.22e+00 |
Pos logit contributions | |
pole | +0.001 |
isk | +0.001 |
sharing | +0.001 |
ending | +0.001 |
py | +0.001 |
Neg logit contributions | |
roller | -0.002 |
steps | -0.001 |
squee | -0.001 |
scrap | -0.001 |
bars | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | +2.49e-01 |
Pos logit contributions | |
roller | +0.053 |
scrap | +0.046 |
steps | +0.046 |
squee | +0.046 |
palm | +0.046 |
Neg logit contributions | |
pole | -0.043 |
sharing | -0.043 |
isk | -0.042 |
Freddie | -0.040 |
ole | -0.040 |
Token | keep |
Token ID | 1394 |
Feature activation | +7.49e-01 |
Pos logit contributions | |
pole | +0.005 |
isk | +0.005 |
ole | +0.005 |
sharing | +0.005 |
py | +0.005 |
Neg logit contributions | |
roller | -0.005 |
steps | -0.005 |
squee | -0.005 |
palm | -0.005 |
scrap | -0.005 |
Token | it |
Token ID | 340 |
Feature activation | +2.07e+00 |
Pos logit contributions | |
isk | +0.018 |
pole | +0.018 |
ending | +0.017 |
py | +0.016 |
que | +0.016 |
Neg logit contributions | |
roller | -0.014 |
steps | -0.013 |
scrap | -0.012 |
squee | -0.012 |
palm | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
isk | +0.050 |
sharing | +0.045 |
py | +0.045 |
End | +0.044 |
ending | +0.044 |
Neg logit contributions | |
roller | -0.037 |
steps | -0.035 |
ti | -0.035 |
palm | -0.034 |
pine | -0.033 |
Token | But |
Token ID | 887 |
Feature activation | +8.14e-01 |
Pos logit contributions | |
pole | +0.024 |
sharing | +0.023 |
isk | +0.023 |
ole | +0.022 |
Freddie | +0.022 |
Neg logit contributions | |
roller | -0.030 |
squee | -0.027 |
rooft | -0.027 |
scrap | -0.027 |
palm | -0.027 |
Token | But |
Token ID | 1537 |
Feature activation | +5.13e-01 |
Pos logit contributions | |
sharing | +0.047 |
pole | +0.047 |
isk | +0.046 |
Freddie | +0.044 |
ole | +0.044 |
Neg logit contributions | |
roller | -0.050 |
scrap | -0.043 |
squee | -0.042 |
rooft | -0.042 |
steps | -0.042 |
Token | Lily |
Token ID | 20037 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
pole | +0.013 |
isk | +0.013 |
ole | +0.012 |
py | +0.012 |
sharing | +0.012 |
Neg logit contributions | |
roller | -0.010 |
scrap | -0.009 |
palm | -0.008 |
mo | -0.008 |
squee | -0.008 |
Token | and |
Token ID | 290 |
Feature activation | +2.12e+00 |
Pos logit contributions | |
roller | +0.041 |
rooft | +0.038 |
leaf | +0.038 |
levers | +0.037 |
col | +0.036 |
Neg logit contributions | |
isk | -0.029 |
pole | -0.029 |
sharing | -0.028 |
being | -0.028 |
Freddie | -0.027 |
Token | Ben |
Token ID | 3932 |
Feature activation | -2.17e+00 |
Pos logit contributions | |
pole | +0.046 |
ole | +0.042 |
sharing | +0.041 |
isk | +0.041 |
py | +0.038 |
Neg logit contributions | |
roller | -0.049 |
scrap | -0.045 |
squee | -0.044 |
steps | -0.044 |
palm | -0.044 |
Token | did |
Token ID | 750 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
roller | +0.058 |
rooft | +0.052 |
leaf | +0.051 |
col | +0.050 |
squee | +0.050 |
Neg logit contributions | |
pole | -0.035 |
being | -0.033 |
isk | -0.033 |
sharing | -0.033 |
end | -0.032 |
Token | not |
Token ID | 407 |
Feature activation | -7.46e-01 |
Pos logit contributions | |
roller | +0.051 |
steps | +0.047 |
scrap | +0.047 |
squee | +0.046 |
levers | +0.046 |
Neg logit contributions | |
pole | -0.024 |
isk | -0.021 |
Freddie | -0.020 |
py | -0.020 |
ole | -0.020 |
Token | hear |
Token ID | 3285 |
Feature activation | +3.64e-01 |
Pos logit contributions | |
roller | +0.016 |
scrap | +0.014 |
rooft | +0.014 |
palm | +0.014 |
squee | +0.014 |
Neg logit contributions | |
pole | -0.016 |
sharing | -0.016 |
isk | -0.016 |
ole | -0.015 |
Freddie | -0.015 |
Token | her |
Token ID | 607 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
pole | +0.008 |
sharing | +0.007 |
isk | +0.007 |
ole | +0.007 |
Freddie | +0.007 |
Neg logit contributions | |
roller | -0.009 |
scrap | -0.008 |
steps | -0.008 |
levers | -0.007 |
squee | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | +4.20e-01 |
Pos logit contributions | |
roller | +0.029 |
rooft | +0.026 |
scrap | +0.026 |
squee | +0.026 |
palm | +0.025 |
Neg logit contributions | |
sharing | -0.020 |
pole | -0.019 |
isk | -0.019 |
ole | -0.019 |
being | -0.019 |
Token | They |
Token ID | 1119 |
Feature activation | +1.26e-01 |
Pos logit contributions | |
pole | +0.007 |
sharing | +0.006 |
isk | +0.006 |
Freddie | +0.006 |
ole | +0.006 |
Neg logit contributions | |
roller | -0.012 |
scrap | -0.010 |
steps | -0.010 |
squee | -0.010 |
levers | -0.010 |
Token | were |
Token ID | 547 |
Feature activation | +2.58e-03 |
Pos logit contributions | |
isk | +0.002 |
always | +0.002 |
pole | +0.002 |
being | +0.002 |
end | +0.002 |
Neg logit contributions | |
roller | -0.003 |
obs | -0.003 |
rooft | -0.003 |
leaf | -0.003 |
icol | -0.003 |
Token | mom |
Token ID | 1995 |
Feature activation | -2.33e+00 |
Pos logit contributions | |
roller | +0.047 |
rooft | +0.043 |
squee | +0.042 |
steps | +0.042 |
lace | +0.042 |
Neg logit contributions | |
sharing | -0.030 |
pole | -0.030 |
isk | -0.028 |
being | -0.028 |
Freddie | -0.028 |
Token | said |
Token ID | 531 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
roller | +0.039 |
scrap | +0.033 |
steps | +0.033 |
squee | +0.032 |
levers | +0.032 |
Neg logit contributions | |
pole | -0.062 |
sharing | -0.061 |
isk | -0.060 |
Freddie | -0.059 |
ole | -0.058 |
Token | . |
Token ID | 13 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
roller | +0.026 |
scrap | +0.023 |
steps | +0.023 |
squee | +0.023 |
levers | +0.023 |
Neg logit contributions | |
pole | -0.031 |
isk | -0.030 |
sharing | -0.029 |
ole | -0.028 |
Freddie | -0.028 |
Token | " |
Token ID | 366 |
Feature activation | -3.67e+00 |
Pos logit contributions | |
roller | +0.003 |
scrap | +0.003 |
palm | +0.003 |
levers | +0.003 |
squee | +0.003 |
Neg logit contributions | |
sharing | -0.002 |
pole | -0.002 |
isk | -0.002 |
Freddie | -0.002 |
being | -0.002 |
Token | The |
Token ID | 464 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
roller | +0.073 |
scrap | +0.064 |
rooft | +0.063 |
levers | +0.062 |
squee | +0.062 |
Neg logit contributions | |
pole | -0.085 |
sharing | -0.083 |
isk | -0.082 |
Freddie | -0.080 |
ole | -0.080 |
Token | bar |
Token ID | 2318 |
Feature activation | -4.31e+00 |
Pos logit contributions | |
being | +0.037 |
sharing | +0.037 |
end | +0.036 |
ending | +0.036 |
End | +0.035 |
Neg logit contributions | |
col | -0.030 |
roller | -0.030 |
rooft | -0.030 |
umb | -0.029 |
squee | -0.028 |
Token | ber |
Token ID | 527 |
Feature activation | -2.68e+00 |
Pos logit contributions | |
roller | +0.101 |
rooft | +0.092 |
scrap | +0.092 |
lace | +0.089 |
umb | +0.089 |
Neg logit contributions | |
sharing | -0.081 |
pole | -0.080 |
being | -0.076 |
ole | -0.076 |
isk | -0.074 |
Token | is |
Token ID | 318 |
Feature activation | +4.85e-01 |
Pos logit contributions | |
roller | +0.052 |
squee | +0.046 |
steps | +0.046 |
scrap | +0.045 |
bars | +0.045 |
Neg logit contributions | |
pole | -0.065 |
isk | -0.063 |
sharing | -0.062 |
ole | -0.062 |
Freddie | -0.062 |
Token | nice |
Token ID | 3621 |
Feature activation | +6.90e-01 |
Pos logit contributions | |
pole | +0.010 |
isk | +0.010 |
ole | +0.009 |
ator | +0.009 |
ingu | +0.009 |
Neg logit contributions | |
roller | -0.012 |
scrap | -0.011 |
squee | -0.010 |
mo | -0.010 |
buck | -0.010 |
Token | and |
Token ID | 290 |
Feature activation | +6.44e-01 |
Pos logit contributions | |
pole | +0.017 |
isk | +0.017 |
ole | +0.016 |
Freddie | +0.016 |
Mickey | +0.016 |
Neg logit contributions | |
roller | -0.013 |
steps | -0.012 |
squee | -0.011 |
bars | -0.011 |
seats | -0.011 |
Token | he |
Token ID | 339 |
Feature activation | +9.95e-01 |
Pos logit contributions | |
pole | +0.012 |
isk | +0.012 |
sharing | +0.011 |
ole | +0.011 |
Freddie | +0.011 |
Neg logit contributions | |
roller | -0.016 |
scrap | -0.014 |
squee | -0.014 |
steps | -0.014 |
levers | -0.014 |
Token | ," |
Token ID | 553 |
Feature activation | -1.38e+00 |
Pos logit contributions | |
pole | +0.026 |
isk | +0.025 |
sharing | +0.025 |
Freddie | +0.024 |
ole | +0.024 |
Neg logit contributions | |
roller | -0.025 |
steps | -0.022 |
levers | -0.022 |
squee | -0.022 |
scrap | -0.022 |
Token | his |
Token ID | 465 |
Feature activation | +4.19e-02 |
Pos logit contributions | |
roller | +0.027 |
scrap | +0.025 |
squee | +0.024 |
steps | +0.024 |
levers | +0.024 |
Neg logit contributions | |
pole | -0.032 |
isk | -0.032 |
sharing | -0.031 |
ole | -0.030 |
py | -0.030 |
Token | mom |
Token ID | 1995 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
ole | +0.001 |
sharing | +0.001 |
ending | +0.001 |
isk | +0.001 |
End | +0.001 |
Neg logit contributions | |
roller | -0.001 |
rooft | -0.001 |
leaf | -0.001 |
col | -0.001 |
bars | -0.001 |
Token | said |
Token ID | 531 |
Feature activation | -1.13e+00 |
Pos logit contributions | |
roller | +0.037 |
scrap | +0.031 |
squee | +0.031 |
steps | +0.031 |
levers | +0.030 |
Neg logit contributions | |
pole | -0.054 |
sharing | -0.053 |
isk | -0.052 |
Freddie | -0.051 |
ole | -0.050 |
Token | . |
Token ID | 13 |
Feature activation | -1.89e-01 |
Pos logit contributions | |
roller | +0.022 |
steps | +0.020 |
squee | +0.020 |
scrap | +0.019 |
levers | +0.019 |
Neg logit contributions | |
isk | -0.026 |
pole | -0.026 |
sharing | -0.025 |
py | -0.025 |
ole | -0.024 |
Token | " |
Token ID | 366 |
Feature activation | -3.89e+00 |
Pos logit contributions | |
roller | +0.004 |
squee | +0.004 |
steps | +0.004 |
levers | +0.004 |
wonder | +0.004 |
Neg logit contributions | |
isk | -0.004 |
pole | -0.004 |
py | -0.004 |
ole | -0.004 |
sharing | -0.004 |
Token | Don |
Token ID | 3987 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
roller | +0.076 |
scrap | +0.066 |
steps | +0.065 |
squee | +0.065 |
levers | +0.064 |
Neg logit contributions | |
pole | -0.092 |
sharing | -0.090 |
isk | -0.089 |
Freddie | -0.086 |
ole | -0.086 |
Token | 't |
Token ID | 470 |
Feature activation | +3.66e-01 |
Pos logit contributions | |
roller | +0.034 |
steps | +0.031 |
scrap | +0.028 |
squee | +0.028 |
Jo | +0.028 |
Neg logit contributions | |
pole | -0.035 |
sharing | -0.034 |
isk | -0.033 |
melon | -0.032 |
ility | -0.032 |
Token | pick |
Token ID | 2298 |
Feature activation | +8.44e-01 |
Pos logit contributions | |
pole | +0.008 |
isk | +0.008 |
sharing | +0.007 |
Freddie | +0.007 |
ole | +0.007 |
Neg logit contributions | |
roller | -0.008 |
scrap | -0.007 |
palm | -0.007 |
wonder | -0.007 |
squee | -0.007 |
Token | up |
Token ID | 510 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
pole | +0.016 |
sharing | +0.015 |
isk | +0.015 |
ole | +0.015 |
Freddie | +0.015 |
Neg logit contributions | |
roller | -0.020 |
scrap | -0.018 |
wonder | -0.018 |
levers | -0.018 |
squee | -0.018 |
Token | anything |
Token ID | 1997 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
pole | +0.041 |
isk | +0.040 |
sharing | +0.039 |
ole | +0.037 |
Freddie | +0.037 |
Neg logit contributions | |
roller | -0.046 |
steps | -0.041 |
scrap | -0.040 |
squee | -0.040 |
palm | -0.040 |
Token | might |
Token ID | 1244 |
Feature activation | -2.32e-01 |
Pos logit contributions | |
pole | +0.066 |
sharing | +0.066 |
isk | +0.066 |
Freddie | +0.066 |
ole | +0.064 |
Neg logit contributions | |
roller | -0.037 |
wonder | -0.037 |
steps | -0.036 |
bars | -0.035 |
squee | -0.034 |
Token | break |
Token ID | 2270 |
Feature activation | +5.96e-01 |
Pos logit contributions | |
roller | +0.005 |
wonder | +0.004 |
scrap | +0.004 |
steps | +0.004 |
squee | +0.004 |
Neg logit contributions | |
pole | -0.005 |
isk | -0.005 |
ole | -0.005 |
Freddie | -0.005 |
sharing | -0.005 |
Token | ." |
Token ID | 526 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
pole | +0.010 |
sharing | +0.010 |
isk | +0.010 |
ole | +0.010 |
Freddie | +0.010 |
Neg logit contributions | |
roller | -0.015 |
scrap | -0.014 |
squee | -0.014 |
steps | -0.014 |
levers | -0.013 |
Token | Tim |
Token ID | 5045 |
Feature activation | -9.22e-01 |
Pos logit contributions | |
roller | +0.037 |
squee | +0.033 |
bars | +0.033 |
scrap | +0.033 |
palm | +0.032 |
Neg logit contributions | |
pole | -0.030 |
sharing | -0.030 |
isk | -0.029 |
ole | -0.028 |
py | -0.028 |
Token | my |
Token ID | 1820 |
Feature activation | -2.50e+00 |
Pos logit contributions | |
roller | +0.031 |
scrap | +0.029 |
squee | +0.029 |
steps | +0.029 |
levers | +0.029 |
Neg logit contributions | |
pole | -0.009 |
sharing | -0.008 |
isk | -0.008 |
Freddie | -0.007 |
ole | -0.007 |
Token | didn |
Token ID | 1422 |
Feature activation | -4.80e+00 |
Pos logit contributions | |
roller | +0.057 |
scrap | +0.052 |
steps | +0.051 |
squee | +0.051 |
palm | +0.050 |
Neg logit contributions | |
pole | -0.050 |
sharing | -0.049 |
isk | -0.048 |
ole | -0.047 |
Freddie | -0.046 |
Token | 't |
Token ID | 470 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
roller | +0.125 |
steps | +0.112 |
scrap | +0.111 |
squee | +0.110 |
levers | +0.108 |
Neg logit contributions | |
pole | -0.087 |
isk | -0.081 |
sharing | -0.081 |
ole | -0.076 |
Freddie | -0.075 |
Token | listen |
Token ID | 6004 |
Feature activation | -3.99e-02 |
Pos logit contributions | |
roller | +0.022 |
wonder | +0.019 |
scrap | +0.019 |
steps | +0.018 |
levers | +0.018 |
Neg logit contributions | |
pole | -0.035 |
isk | -0.034 |
sharing | -0.033 |
Freddie | -0.033 |
Mickey | -0.032 |
Token | and |
Token ID | 290 |
Feature activation | -3.39e-01 |
Pos logit contributions | |
roller | +0.001 |
steps | +0.001 |
scrap | +0.001 |
levers | +0.001 |
squee | +0.001 |
Neg logit contributions | |
pole | -0.001 |
isk | -0.001 |
sharing | -0.001 |
Freddie | -0.001 |
ole | -0.001 |
Token | continued |
Token ID | 3767 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
roller | +0.007 |
scrap | +0.007 |
steps | +0.007 |
squee | +0.007 |
levers | +0.006 |
Neg logit contributions | |
pole | -0.007 |
isk | -0.007 |
sharing | -0.007 |
ole | -0.007 |
Freddie | -0.007 |
Token | to |
Token ID | 284 |
Feature activation | +4.69e-01 |
Pos logit contributions | |
pole | +0.044 |
isk | +0.042 |
sharing | +0.041 |
ole | +0.041 |
Freddie | +0.041 |
Neg logit contributions | |
roller | -0.028 |
scrap | -0.024 |
steps | -0.024 |
squee | -0.024 |
levers | -0.023 |
Token | saw |
Token ID | 2497 |
Feature activation | -2.22e+00 |
Pos logit contributions | |
roller | +0.028 |
steps | +0.024 |
levers | +0.024 |
rooft | +0.023 |
bars | +0.023 |
Neg logit contributions | |
pole | -0.030 |
sharing | -0.029 |
being | -0.029 |
Freddie | -0.029 |
isk | -0.029 |
Token | a |
Token ID | 257 |
Feature activation | -2.33e+00 |
Pos logit contributions | |
roller | +0.064 |
obs | +0.057 |
scrap | +0.056 |
squee | +0.055 |
bisc | +0.054 |
Neg logit contributions | |
being | -0.032 |
ter | -0.032 |
pole | -0.031 |
Mickey | -0.031 |
sharing | -0.030 |
Token | fairy |
Token ID | 25607 |
Feature activation | +8.01e-01 |
Pos logit contributions | |
roller | +0.048 |
obs | +0.044 |
levers | +0.044 |
scrap | +0.043 |
rooft | +0.043 |
Neg logit contributions | |
sharing | -0.050 |
pole | -0.049 |
ole | -0.048 |
Freddie | -0.045 |
ending | -0.045 |
Token | sitting |
Token ID | 5586 |
Feature activation | +2.22e-01 |
Pos logit contributions | |
pole | +0.016 |
sharing | +0.016 |
isk | +0.015 |
Freddie | +0.015 |
ole | +0.015 |
Neg logit contributions | |
roller | -0.019 |
scrap | -0.017 |
steps | -0.016 |
squee | -0.016 |
levers | -0.016 |
Token | on |
Token ID | 319 |
Feature activation | -7.15e-01 |
Pos logit contributions | |
pole | +0.004 |
sharing | +0.004 |
isk | +0.004 |
ole | +0.004 |
Freddie | +0.004 |
Neg logit contributions | |
roller | -0.005 |
squee | -0.005 |
scrap | -0.005 |
levers | -0.005 |
steps | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | -4.17e+00 |
Pos logit contributions | |
roller | +0.019 |
scrap | +0.018 |
obs | +0.017 |
levers | +0.017 |
steps | +0.017 |
Neg logit contributions | |
Freddie | -0.011 |
Mickey | -0.011 |
sharing | -0.011 |
being | -0.011 |
isk | -0.011 |
Token | blade |
Token ID | 11865 |
Feature activation | -5.80e+00 |
Pos logit contributions | |
roller | +0.070 |
levers | +0.064 |
squee | +0.064 |
scrap | +0.063 |
obs | +0.061 |
Neg logit contributions | |
pole | -0.111 |
sharing | -0.101 |
Freddie | -0.101 |
ole | -0.100 |
isk | -0.099 |
Token | of |
Token ID | 286 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
roller | +0.149 |
squee | +0.135 |
scrap | +0.135 |
rooft | +0.133 |
levers | +0.132 |
Neg logit contributions | |
pole | -0.108 |
sharing | -0.099 |
ole | -0.094 |
isk | -0.094 |
Freddie | -0.091 |
Token | grass |
Token ID | 8701 |
Feature activation | -1.88e-02 |
Pos logit contributions | |
roller | +0.020 |
obs | +0.018 |
squee | +0.017 |
Jo | +0.017 |
bang | +0.016 |
Neg logit contributions | |
pole | -0.024 |
being | -0.023 |
Mickey | -0.023 |
Freddie | -0.023 |
py | -0.022 |
Token | on |
Token ID | 319 |
Feature activation | -2.38e-01 |
Pos logit contributions | |
roller | +0.000 |
scrap | +0.000 |
steps | +0.000 |
squee | +0.000 |
palm | +0.000 |
Neg logit contributions | |
pole | -0.000 |
sharing | -0.000 |
isk | -0.000 |
Freddie | -0.000 |
ole | -0.000 |
Token | the |
Token ID | 262 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
roller | +0.006 |
scrap | +0.005 |
squee | +0.005 |
levers | +0.005 |
steps | +0.005 |
Neg logit contributions | |
pole | -0.004 |
sharing | -0.004 |
Freddie | -0.004 |
Mickey | -0.004 |
isk | -0.004 |
Token | finished |
Token ID | 5201 |
Feature activation | +2.57e+00 |
Pos logit contributions | |
sharing | +0.000 |
pole | +0.000 |
isk | +0.000 |
ole | +0.000 |
Freddie | +0.000 |
Neg logit contributions | |
roller | -0.000 |
rooft | -0.000 |
scrap | -0.000 |
squee | -0.000 |
levers | -0.000 |
Token | your |
Token ID | 534 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
pole | +0.049 |
isk | +0.047 |
ole | +0.044 |
sharing | +0.044 |
Freddie | +0.043 |
Neg logit contributions | |
roller | -0.063 |
squee | -0.057 |
scrap | -0.057 |
steps | -0.056 |
rooft | -0.055 |
Token | dinner |
Token ID | 8073 |
Feature activation | -3.03e-01 |
Pos logit contributions | |
roller | +0.031 |
steps | +0.030 |
squee | +0.028 |
scrap | +0.028 |
seats | +0.027 |
Neg logit contributions | |
pole | -0.040 |
isk | -0.036 |
ole | -0.035 |
py | -0.034 |
sharing | -0.034 |
Token | when |
Token ID | 618 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
roller | +0.007 |
steps | +0.006 |
squee | +0.006 |
bars | +0.006 |
seats | +0.006 |
Neg logit contributions | |
pole | -0.006 |
isk | -0.006 |
sharing | -0.006 |
Freddie | -0.006 |
Mickey | -0.006 |
Token | you |
Token ID | 345 |
Feature activation | -6.30e-01 |
Pos logit contributions | |
roller | +0.037 |
steps | +0.036 |
palm | +0.034 |
scrap | +0.034 |
squee | +0.034 |
Neg logit contributions | |
isk | -0.032 |
pole | -0.032 |
ole | -0.028 |
py | -0.027 |
pper | -0.026 |
Token | didn |
Token ID | 1422 |
Feature activation | -3.69e+00 |
Pos logit contributions | |
roller | +0.014 |
squee | +0.012 |
scrap | +0.012 |
steps | +0.012 |
bars | +0.012 |
Neg logit contributions | |
pole | -0.014 |
sharing | -0.013 |
isk | -0.013 |
ole | -0.013 |
Freddie | -0.013 |
Token | 't |
Token ID | 470 |
Feature activation | -7.26e-01 |
Pos logit contributions | |
roller | +0.107 |
scrap | +0.097 |
steps | +0.096 |
squee | +0.096 |
rooft | +0.095 |
Neg logit contributions | |
pole | -0.053 |
sharing | -0.051 |
isk | -0.050 |
ole | -0.047 |
Freddie | -0.047 |
Token | ," |
Token ID | 553 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
roller | +0.014 |
wonder | +0.013 |
steps | +0.012 |
scrap | +0.012 |
squee | +0.012 |
Neg logit contributions | |
pole | -0.017 |
isk | -0.017 |
Freddie | -0.016 |
ole | -0.016 |
Mickey | -0.015 |
Token | her |
Token ID | 607 |
Feature activation | -3.18e-01 |
Pos logit contributions | |
roller | +0.044 |
rooft | +0.040 |
obs | +0.039 |
levers | +0.038 |
steps | +0.037 |
Neg logit contributions | |
sharing | -0.032 |
being | -0.030 |
pole | -0.030 |
Mickey | -0.030 |
Freddie | -0.029 |
Token | mom |
Token ID | 1995 |
Feature activation | -2.24e+00 |
Pos logit contributions | |
roller | +0.009 |
leaf | +0.009 |
rooft | +0.009 |
col | +0.008 |
levers | +0.008 |
Neg logit contributions | |
sharing | -0.005 |
ending | -0.004 |
isk | -0.004 |
ole | -0.004 |
pole | -0.004 |
Token | said |
Token ID | 531 |
Feature activation | -6.73e-01 |
Pos logit contributions | |
roller | +0.037 |
scrap | +0.032 |
squee | +0.031 |
steps | +0.031 |
levers | +0.031 |
Neg logit contributions | |
pole | -0.059 |
sharing | -0.058 |
isk | -0.057 |
Freddie | -0.056 |
ole | -0.056 |