Token | , |
Token ID | 11 |
Feature activation | +4.42e-01 |
Pos logit contributions | |
subt | +0.056 |
Deal | +0.055 |
math | +0.055 |
cream | +0.054 |
topp | +0.054 |
Neg logit contributions | |
en | -0.029 |
dot | -0.028 |
el | -0.028 |
fly | -0.027 |
ening | -0.026 |
Token | fifteen |
Token ID | 17280 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
subt | +0.011 |
Deal | +0.011 |
cream | +0.010 |
math | +0.010 |
topp | +0.010 |
Neg logit contributions | |
dot | -0.014 |
en | -0.014 |
el | -0.014 |
fly | -0.013 |
ir | -0.013 |
Token | , |
Token ID | 11 |
Feature activation | +6.50e-01 |
Pos logit contributions | |
subt | +0.059 |
Deal | +0.059 |
math | +0.056 |
cream | +0.056 |
Fif | +0.056 |
Neg logit contributions | |
en | -0.034 |
el | -0.033 |
dot | -0.032 |
ening | -0.030 |
le | -0.030 |
Token | sixteen |
Token ID | 27913 |
Feature activation | +2.53e+00 |
Pos logit contributions | |
subt | +0.017 |
Deal | +0.017 |
math | +0.017 |
cream | +0.017 |
topp | +0.016 |
Neg logit contributions | |
dot | -0.019 |
en | -0.019 |
el | -0.019 |
fly | -0.018 |
ir | -0.018 |
Token | , |
Token ID | 11 |
Feature activation | +8.50e-01 |
Pos logit contributions | |
Deal | +0.100 |
subt | +0.099 |
cream | +0.096 |
math | +0.095 |
bek | +0.094 |
Neg logit contributions | |
en | -0.044 |
el | -0.043 |
dot | -0.042 |
ening | -0.038 |
ir | -0.038 |
Token | seventeen |
Token ID | 38741 |
Feature activation | +4.53e+00 |
Pos logit contributions | |
subt | +0.023 |
Deal | +0.023 |
math | +0.023 |
cream | +0.022 |
subtract | +0.022 |
Neg logit contributions | |
dot | -0.025 |
en | -0.025 |
el | -0.024 |
fly | -0.024 |
ir | -0.023 |
Token | , |
Token ID | 11 |
Feature activation | +8.06e-01 |
Pos logit contributions | |
subt | +0.199 |
bek | +0.193 |
Deal | +0.193 |
Fif | +0.188 |
charity | +0.187 |
Neg logit contributions | |
en | -0.063 |
el | -0.063 |
inkle | -0.052 |
le | -0.051 |
ink | -0.050 |
Token | eighteen |
Token ID | 29095 |
Feature activation | +2.94e+00 |
Pos logit contributions | |
subt | +0.021 |
Deal | +0.021 |
math | +0.021 |
cream | +0.020 |
topp | +0.020 |
Neg logit contributions | |
dot | -0.024 |
en | -0.024 |
el | -0.024 |
fly | -0.023 |
ir | -0.023 |
Token | , |
Token ID | 11 |
Feature activation | +6.72e-01 |
Pos logit contributions | |
subt | +0.119 |
Deal | +0.117 |
math | +0.113 |
cream | +0.113 |
Fif | +0.112 |
Neg logit contributions | |
en | -0.048 |
el | -0.047 |
dot | -0.045 |
inkle | -0.041 |
ir | -0.041 |
Token | nineteen |
Token ID | 43063 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
subt | +0.017 |
Deal | +0.017 |
math | +0.017 |
cream | +0.017 |
topp | +0.016 |
Neg logit contributions | |
dot | -0.021 |
en | -0.020 |
el | -0.020 |
fly | -0.020 |
ir | -0.019 |
Token | , |
Token ID | 11 |
Feature activation | +5.82e-01 |
Pos logit contributions | |
subt | +0.045 |
Deal | +0.043 |
cream | +0.042 |
subtract | +0.042 |
math | +0.042 |
Neg logit contributions | |
en | -0.019 |
el | -0.019 |
dot | -0.017 |
inkle | -0.017 |
le | -0.017 |
Token | thought |
Token ID | 1807 |
Feature activation | +6.16e-01 |
Pos logit contributions | |
subt | +0.015 |
subtract | +0.015 |
math | +0.014 |
cream | +0.014 |
Deal | +0.013 |
Neg logit contributions | |
ir | -0.024 |
fly | -0.023 |
py | -0.023 |
predators | -0.023 |
ays | -0.023 |
Token | they |
Token ID | 484 |
Feature activation | +4.84e-01 |
Pos logit contributions | |
math | +0.020 |
subt | +0.016 |
cakes | +0.016 |
cream | +0.015 |
homework | +0.015 |
Neg logit contributions | |
fly | -0.021 |
ley | -0.020 |
dot | -0.019 |
an | -0.018 |
ga | -0.018 |
Token | looked |
Token ID | 3114 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
Deal | +0.012 |
topp | +0.011 |
y | +0.011 |
subtract | +0.011 |
cream | +0.011 |
Neg logit contributions | |
ug | -0.016 |
ir | -0.016 |
ga | -0.016 |
inkle | -0.015 |
ays | -0.015 |
Token | like |
Token ID | 588 |
Feature activation | +7.77e-01 |
Pos logit contributions | |
subt | +0.030 |
math | +0.030 |
cream | +0.029 |
topp | +0.029 |
Deal | +0.029 |
Neg logit contributions | |
en | -0.035 |
dot | -0.035 |
el | -0.034 |
fly | -0.034 |
ug | -0.034 |
Token | ac |
Token ID | 936 |
Feature activation | +3.40e-01 |
Pos logit contributions | |
math | +0.018 |
subt | +0.017 |
cream | +0.017 |
topp | +0.016 |
subtract | +0.016 |
Neg logit contributions | |
en | -0.027 |
el | -0.027 |
dot | -0.027 |
fly | -0.026 |
ir | -0.026 |
Token | rob |
Token ID | 22609 |
Feature activation | +4.32e+00 |
Pos logit contributions | |
cream | +0.008 |
Deal | +0.007 |
subt | +0.007 |
math | +0.007 |
topp | +0.007 |
Neg logit contributions | |
dot | -0.012 |
en | -0.012 |
fly | -0.012 |
el | -0.011 |
ug | -0.011 |
Token | ats |
Token ID | 1381 |
Feature activation | +5.80e-02 |
Pos logit contributions | |
subt | +0.146 |
math | +0.141 |
cream | +0.140 |
Deal | +0.139 |
subtract | +0.139 |
Neg logit contributions | |
en | -0.099 |
el | -0.098 |
ir | -0.096 |
ug | -0.094 |
le | -0.094 |
Token | . |
Token ID | 13 |
Feature activation | +2.45e-01 |
Pos logit contributions | |
cream | +0.002 |
math | +0.002 |
cakes | +0.002 |
pizza | +0.002 |
cookies | +0.002 |
Neg logit contributions | |
inkle | -0.002 |
dot | -0.002 |
ug | -0.002 |
ir | -0.002 |
el | -0.001 |
Token | She |
Token ID | 1375 |
Feature activation | +5.01e-01 |
Pos logit contributions | |
subt | +0.008 |
cream | +0.008 |
math | +0.008 |
topp | +0.008 |
Deal | +0.008 |
Neg logit contributions | |
en | -0.006 |
dot | -0.006 |
el | -0.006 |
fly | -0.006 |
ug | -0.005 |
Token | enjoyed |
Token ID | 8359 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
subt | +0.012 |
cream | +0.011 |
subtract | +0.011 |
math | +0.011 |
topp | +0.011 |
Neg logit contributions | |
ir | -0.017 |
fly | -0.017 |
ug | -0.017 |
ays | -0.017 |
dot | -0.017 |
Token | the |
Token ID | 262 |
Feature activation | +5.45e-01 |
Pos logit contributions | |
en | +0.008 |
dot | +0.008 |
el | +0.008 |
fly | +0.007 |
ug | +0.007 |
Neg logit contributions | |
subt | -0.007 |
math | -0.007 |
cream | -0.007 |
Deal | -0.007 |
topp | -0.007 |
Token | , |
Token ID | 11 |
Feature activation | +3.70e-01 |
Pos logit contributions | |
subt | +0.048 |
cream | +0.047 |
math | +0.047 |
topp | +0.047 |
subtract | +0.046 |
Neg logit contributions | |
dot | -0.028 |
en | -0.028 |
fly | -0.028 |
ug | -0.027 |
ga | -0.026 |
Token | fifteen |
Token ID | 17280 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
subt | +0.009 |
math | +0.009 |
Deal | +0.009 |
cream | +0.009 |
topp | +0.009 |
Neg logit contributions | |
en | -0.012 |
dot | -0.012 |
el | -0.012 |
fly | -0.011 |
ug | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | +5.09e-01 |
Pos logit contributions | |
subt | +0.046 |
math | +0.046 |
cream | +0.046 |
topp | +0.045 |
subtract | +0.044 |
Neg logit contributions | |
dot | -0.027 |
en | -0.026 |
fly | -0.026 |
ug | -0.026 |
el | -0.025 |
Token | sixteen |
Token ID | 27913 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
subt | +0.013 |
math | +0.013 |
cream | +0.013 |
Deal | +0.013 |
topp | +0.013 |
Neg logit contributions | |
en | -0.015 |
dot | -0.015 |
el | -0.015 |
fly | -0.015 |
ug | -0.015 |
Token | , |
Token ID | 11 |
Feature activation | +6.76e-01 |
Pos logit contributions | |
subt | +0.073 |
math | +0.073 |
cream | +0.073 |
topp | +0.072 |
subtract | +0.071 |
Neg logit contributions | |
fly | -0.045 |
dot | -0.045 |
ug | -0.044 |
ley | -0.043 |
en | -0.043 |
Token | seventeen |
Token ID | 38741 |
Feature activation | +4.32e+00 |
Pos logit contributions | |
subt | +0.018 |
math | +0.017 |
cream | +0.017 |
Deal | +0.017 |
topp | +0.017 |
Neg logit contributions | |
en | -0.020 |
dot | -0.020 |
el | -0.020 |
fly | -0.020 |
ug | -0.019 |
Token | , |
Token ID | 11 |
Feature activation | +7.05e-01 |
Pos logit contributions | |
subt | +0.180 |
Deal | +0.174 |
math | +0.173 |
subtract | +0.170 |
Fif | +0.169 |
Neg logit contributions | |
en | -0.068 |
el | -0.066 |
dot | -0.060 |
le | -0.059 |
ir | -0.058 |
Token | eighteen |
Token ID | 29095 |
Feature activation | +2.77e+00 |
Pos logit contributions | |
subt | +0.018 |
math | +0.018 |
cream | +0.017 |
Deal | +0.017 |
topp | +0.017 |
Neg logit contributions | |
en | -0.022 |
dot | -0.022 |
el | -0.021 |
fly | -0.021 |
ug | -0.021 |
Token | , |
Token ID | 11 |
Feature activation | +6.18e-01 |
Pos logit contributions | |
subt | +0.105 |
math | +0.103 |
cream | +0.103 |
Deal | +0.102 |
topp | +0.102 |
Neg logit contributions | |
en | -0.051 |
dot | -0.050 |
el | -0.049 |
fly | -0.048 |
ug | -0.047 |
Token | nineteen |
Token ID | 43063 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
subt | +0.016 |
math | +0.015 |
Deal | +0.015 |
cream | +0.015 |
topp | +0.015 |
Neg logit contributions | |
en | -0.019 |
dot | -0.019 |
el | -0.019 |
fly | -0.019 |
ug | -0.018 |
Token | , |
Token ID | 11 |
Feature activation | +5.50e-01 |
Pos logit contributions | |
subt | +0.045 |
math | +0.044 |
cream | +0.044 |
Deal | +0.044 |
topp | +0.044 |
Neg logit contributions | |
en | -0.023 |
dot | -0.023 |
el | -0.022 |
fly | -0.022 |
ug | -0.022 |
Token | sund |
Token ID | 37437 |
Feature activation | -4.21e-01 |
Pos logit contributions | |
subt | +0.023 |
Deal | +0.022 |
subtract | +0.021 |
math | +0.020 |
Fif | +0.019 |
Neg logit contributions | |
dot | -0.026 |
el | -0.024 |
wil | -0.024 |
fly | -0.024 |
inkle | -0.024 |
Token | ae |
Token ID | 3609 |
Feature activation | -6.19e-01 |
Pos logit contributions | |
le | +0.013 |
ink | +0.013 |
el | +0.013 |
ie | +0.012 |
ug | +0.012 |
Neg logit contributions | |
subt | -0.013 |
disl | -0.012 |
math | -0.012 |
subtract | -0.012 |
Pro | -0.011 |
Token | with |
Token ID | 351 |
Feature activation | +6.53e-01 |
Pos logit contributions | |
en | +0.014 |
dot | +0.014 |
el | +0.014 |
fly | +0.014 |
le | +0.013 |
Neg logit contributions | |
subt | -0.021 |
math | -0.020 |
Deal | -0.020 |
cream | -0.020 |
topp | -0.020 |
Token | lots |
Token ID | 6041 |
Feature activation | -6.63e-02 |
Pos logit contributions | |
cream | +0.014 |
subt | +0.014 |
math | +0.014 |
topp | +0.014 |
subtract | +0.013 |
Neg logit contributions | |
en | -0.023 |
fly | -0.023 |
dot | -0.023 |
ug | -0.023 |
le | -0.022 |
Token | of |
Token ID | 286 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
fly | +0.002 |
dot | +0.002 |
en | +0.002 |
el | +0.001 |
ug | +0.001 |
Neg logit contributions | |
cream | -0.002 |
subt | -0.002 |
math | -0.002 |
topp | -0.002 |
subtract | -0.002 |
Token | topp |
Token ID | 23126 |
Feature activation | +4.28e+00 |
Pos logit contributions | |
subt | +0.021 |
Deal | +0.019 |
math | +0.019 |
subtract | +0.018 |
Fif | +0.018 |
Neg logit contributions | |
el | -0.038 |
en | -0.037 |
le | -0.036 |
ening | -0.036 |
dot | -0.035 |
Token | ings |
Token ID | 654 |
Feature activation | -4.22e-01 |
Pos logit contributions | |
cream | +0.174 |
topp | +0.173 |
subt | +0.172 |
math | +0.172 |
Deal | +0.169 |
Neg logit contributions | |
dot | -0.066 |
en | -0.064 |
fly | -0.062 |
ug | -0.060 |
el | -0.060 |
Token | . |
Token ID | 13 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
dot | +0.011 |
en | +0.011 |
fly | +0.011 |
el | +0.010 |
ug | +0.010 |
Neg logit contributions | |
cream | -0.013 |
subt | -0.013 |
math | -0.013 |
topp | -0.012 |
Deal | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | -1.76e-01 |
Pos logit contributions | |
en | +0.003 |
dot | +0.003 |
el | +0.003 |
fly | +0.003 |
ug | +0.003 |
Neg logit contributions | |
subt | -0.003 |
cream | -0.003 |
math | -0.003 |
topp | -0.003 |
Deal | -0.003 |
Token |
|
Token ID | 198 |
Feature activation | -2.89e-01 |
Pos logit contributions | |
fly | +0.007 |
dot | +0.007 |
ening | +0.007 |
en | +0.007 |
el | +0.007 |
Neg logit contributions | |
math | -0.003 |
subt | -0.003 |
cream | -0.003 |
subtract | -0.003 |
Deal | -0.003 |
Token | They |
Token ID | 2990 |
Feature activation | +2.43e-01 |
Pos logit contributions | |
en | +0.008 |
dot | +0.008 |
ga | +0.008 |
ug | +0.008 |
ches | +0.008 |
Neg logit contributions | |
topp | -0.008 |
subt | -0.008 |
cream | -0.007 |
Deal | -0.007 |
math | -0.007 |
Token | flavors |
Token ID | 17361 |
Feature activation | +7.07e-01 |
Pos logit contributions | |
cream | +0.046 |
topp | +0.044 |
subt | +0.044 |
math | +0.044 |
Deal | +0.044 |
Neg logit contributions | |
dot | -0.029 |
en | -0.029 |
fly | -0.028 |
ug | -0.028 |
el | -0.028 |
Token | they |
Token ID | 484 |
Feature activation | +2.44e-01 |
Pos logit contributions | |
cream | +0.023 |
topp | +0.022 |
Deal | +0.021 |
math | +0.021 |
subt | +0.021 |
Neg logit contributions | |
dot | -0.018 |
en | -0.018 |
fly | -0.018 |
py | -0.017 |
ir | -0.017 |
Token | liked |
Token ID | 8288 |
Feature activation | +2.73e-01 |
Pos logit contributions | |
cream | +0.007 |
topp | +0.007 |
Deal | +0.007 |
subt | +0.007 |
math | +0.007 |
Neg logit contributions | |
dot | -0.007 |
en | -0.007 |
ug | -0.007 |
el | -0.007 |
fly | -0.007 |
Token | and |
Token ID | 290 |
Feature activation | +3.30e-01 |
Pos logit contributions | |
cream | +0.008 |
math | +0.008 |
topp | +0.008 |
subt | +0.008 |
cakes | +0.008 |
Neg logit contributions | |
dot | -0.008 |
en | -0.007 |
fly | -0.007 |
ug | -0.007 |
el | -0.007 |
Token | what |
Token ID | 644 |
Feature activation | +7.14e-01 |
Pos logit contributions | |
subt | +0.009 |
math | +0.009 |
cream | +0.009 |
Deal | +0.009 |
topp | +0.009 |
Neg logit contributions | |
en | -0.010 |
dot | -0.010 |
el | -0.009 |
fly | -0.009 |
ug | -0.009 |
Token | topp |
Token ID | 23126 |
Feature activation | +4.23e+00 |
Pos logit contributions | |
subt | +0.022 |
Deal | +0.021 |
subtract | +0.020 |
Fif | +0.020 |
cream | +0.020 |
Neg logit contributions | |
el | -0.018 |
en | -0.018 |
dot | -0.017 |
fly | -0.017 |
wil | -0.017 |
Token | ings |
Token ID | 654 |
Feature activation | -3.16e-01 |
Pos logit contributions | |
cream | +0.146 |
topp | +0.146 |
math | +0.142 |
subt | +0.142 |
Deal | +0.140 |
Neg logit contributions | |
dot | -0.094 |
en | -0.090 |
fly | -0.089 |
ug | -0.087 |
el | -0.086 |
Token | would |
Token ID | 561 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
dot | +0.009 |
en | +0.009 |
fly | +0.009 |
ug | +0.009 |
el | +0.009 |
Neg logit contributions | |
cream | -0.009 |
math | -0.009 |
subt | -0.009 |
topp | -0.009 |
Deal | -0.008 |
Token | go |
Token ID | 467 |
Feature activation | +2.78e-01 |
Pos logit contributions | |
dot | +0.054 |
en | +0.054 |
el | +0.053 |
fly | +0.053 |
ug | +0.052 |
Neg logit contributions | |
subt | -0.037 |
cream | -0.037 |
math | -0.037 |
topp | -0.036 |
Deal | -0.036 |
Token | best |
Token ID | 1266 |
Feature activation | -7.10e-01 |
Pos logit contributions | |
cream | +0.008 |
math | +0.008 |
subt | +0.008 |
topp | +0.007 |
Deal | +0.007 |
Neg logit contributions | |
dot | -0.008 |
fly | -0.008 |
py | -0.007 |
ug | -0.007 |
en | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | +3.64e-01 |
Pos logit contributions | |
en | +0.019 |
el | +0.018 |
dot | +0.018 |
ug | +0.018 |
le | +0.018 |
Neg logit contributions | |
subt | -0.022 |
Deal | -0.021 |
math | -0.021 |
subtract | -0.020 |
topp | -0.020 |
Token | friends |
Token ID | 2460 |
Feature activation | +7.31e-01 |
Pos logit contributions | |
cream | +0.006 |
math | +0.006 |
subt | +0.006 |
cakes | +0.006 |
topp | +0.006 |
Neg logit contributions | |
fly | -0.015 |
dot | -0.015 |
en | -0.015 |
ga | -0.014 |
inkle | -0.014 |
Token | didn |
Token ID | 1422 |
Feature activation | -6.86e-01 |
Pos logit contributions | |
subt | +0.017 |
cream | +0.017 |
Deal | +0.016 |
subtract | +0.016 |
math | +0.016 |
Neg logit contributions | |
dot | -0.024 |
inkle | -0.023 |
ays | -0.023 |
fly | -0.023 |
unks | -0.022 |
Token | 't |
Token ID | 470 |
Feature activation | +7.61e-01 |
Pos logit contributions | |
dot | +0.015 |
en | +0.015 |
fly | +0.015 |
ley | +0.015 |
el | +0.015 |
Neg logit contributions | |
math | -0.024 |
subt | -0.023 |
cream | -0.023 |
subtract | -0.022 |
topp | -0.022 |
Token | stop |
Token ID | 2245 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
subt | +0.024 |
topp | +0.024 |
math | +0.023 |
Deal | +0.023 |
cream | +0.023 |
Neg logit contributions | |
en | -0.019 |
dot | -0.019 |
le | -0.018 |
fly | -0.018 |
ga | -0.018 |
Token | qu |
Token ID | 627 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
math | +0.025 |
subt | +0.025 |
subtract | +0.024 |
cream | +0.024 |
topp | +0.023 |
Neg logit contributions | |
ches | -0.031 |
fly | -0.030 |
dot | -0.030 |
ley | -0.030 |
ug | -0.030 |
Token | arre |
Token ID | 9624 |
Feature activation | +4.21e+00 |
Pos logit contributions | |
subt | +0.033 |
math | +0.032 |
cream | +0.031 |
topp | +0.031 |
subtract | +0.030 |
Neg logit contributions | |
dot | -0.103 |
en | -0.102 |
fly | -0.102 |
el | -0.101 |
ug | -0.099 |
Token | ling |
Token ID | 1359 |
Feature activation | +5.29e-01 |
Pos logit contributions | |
subt | +0.173 |
math | +0.171 |
Deal | +0.169 |
cream | +0.169 |
topp | +0.168 |
Neg logit contributions | |
en | -0.068 |
dot | -0.064 |
el | -0.062 |
le | -0.061 |
ug | -0.061 |
Token | and |
Token ID | 290 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
math | +0.016 |
cream | +0.016 |
subt | +0.015 |
cakes | +0.015 |
topp | +0.015 |
Neg logit contributions | |
dot | -0.014 |
inkle | -0.013 |
el | -0.013 |
fly | -0.013 |
ley | -0.013 |
Token | were |
Token ID | 547 |
Feature activation | +7.97e-01 |
Pos logit contributions | |
dot | +0.004 |
inkle | +0.004 |
el | +0.004 |
ug | +0.003 |
ir | +0.003 |
Neg logit contributions | |
subt | -0.003 |
cream | -0.003 |
subtract | -0.003 |
math | -0.003 |
Deal | -0.003 |
Token | getting |
Token ID | 1972 |
Feature activation | +3.30e-01 |
Pos logit contributions | |
cream | +0.021 |
math | +0.021 |
subt | +0.020 |
Deal | +0.019 |
subtract | +0.019 |
Neg logit contributions | |
dot | -0.025 |
el | -0.024 |
en | -0.024 |
ley | -0.024 |
fly | -0.023 |
Token | louder |
Token ID | 27089 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
subt | +0.009 |
cream | +0.009 |
math | +0.009 |
topp | +0.009 |
subtract | +0.009 |
Neg logit contributions | |
en | -0.009 |
dot | -0.009 |
fly | -0.009 |
el | -0.009 |
ug | -0.009 |
Token | said |
Token ID | 531 |
Feature activation | -7.55e-01 |
Pos logit contributions | |
dot | +0.023 |
en | +0.023 |
el | +0.022 |
fly | +0.022 |
ug | +0.022 |
Neg logit contributions | |
subt | -0.023 |
cream | -0.022 |
math | -0.022 |
topp | -0.022 |
Deal | -0.022 |
Token | , |
Token ID | 11 |
Feature activation | +4.00e-01 |
Pos logit contributions | |
dot | +0.015 |
en | +0.015 |
el | +0.015 |
fly | +0.015 |
ug | +0.014 |
Neg logit contributions | |
subt | -0.027 |
math | -0.027 |
cream | -0.027 |
topp | -0.027 |
Deal | -0.026 |
Token | " |
Token ID | 366 |
Feature activation | -3.65e-02 |
Pos logit contributions | |
subt | +0.011 |
Deal | +0.011 |
topp | +0.011 |
math | +0.011 |
cream | +0.011 |
Neg logit contributions | |
dot | -0.011 |
en | -0.011 |
el | -0.011 |
fly | -0.011 |
ga | -0.011 |
Token | Stop |
Token ID | 19485 |
Feature activation | +1.39e+00 |
Pos logit contributions | |
en | +0.001 |
el | +0.001 |
dot | +0.001 |
ga | +0.001 |
ir | +0.001 |
Neg logit contributions | |
math | -0.001 |
subt | -0.001 |
topp | -0.001 |
cream | -0.001 |
Deal | -0.001 |
Token | qu |
Token ID | 627 |
Feature activation | +2.34e+00 |
Pos logit contributions | |
math | +0.035 |
subt | +0.034 |
subtract | +0.034 |
cream | +0.034 |
topp | +0.033 |
Neg logit contributions | |
el | -0.043 |
fly | -0.043 |
ley | -0.043 |
ir | -0.043 |
dot | -0.043 |
Token | arre |
Token ID | 9624 |
Feature activation | +3.97e+00 |
Pos logit contributions | |
subt | +0.037 |
math | +0.035 |
Deal | +0.035 |
cream | +0.035 |
topp | +0.034 |
Neg logit contributions | |
en | -0.097 |
el | -0.095 |
dot | -0.095 |
ug | -0.094 |
ir | -0.093 |
Token | ling |
Token ID | 1359 |
Feature activation | +5.90e-01 |
Pos logit contributions | |
subt | +0.172 |
math | +0.169 |
Deal | +0.169 |
cream | +0.167 |
topp | +0.167 |
Neg logit contributions | |
en | -0.057 |
el | -0.051 |
dot | -0.051 |
le | -0.050 |
ug | -0.050 |
Token | , |
Token ID | 11 |
Feature activation | +7.44e-01 |
Pos logit contributions | |
math | +0.019 |
cream | +0.019 |
subt | +0.018 |
job | +0.018 |
topp | +0.018 |
Neg logit contributions | |
dot | -0.014 |
fly | -0.014 |
wil | -0.014 |
inkle | -0.013 |
el | -0.013 |
Token | boys |
Token ID | 6510 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
subt | +0.020 |
math | +0.020 |
Deal | +0.019 |
topp | +0.019 |
cream | +0.019 |
Neg logit contributions | |
en | -0.022 |
dot | -0.022 |
el | -0.022 |
ug | -0.021 |
fly | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | +4.65e-01 |
Pos logit contributions | |
math | +0.034 |
cream | +0.034 |
Deal | +0.033 |
subt | +0.032 |
subtract | +0.032 |
Neg logit contributions | |
dot | -0.035 |
fly | -0.034 |
en | -0.033 |
py | -0.033 |
el | -0.033 |
Token | You |
Token ID | 921 |
Feature activation | -8.13e-02 |
Pos logit contributions | |
subt | +0.013 |
math | +0.013 |
cream | +0.013 |
topp | +0.013 |
Deal | +0.013 |
Neg logit contributions | |
en | -0.013 |
dot | -0.013 |
el | -0.013 |
fly | -0.013 |
le | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | +2.61e-01 |
Pos logit contributions | |
dot | +0.018 |
ir | +0.018 |
ug | +0.018 |
py | +0.018 |
en | +0.018 |
Neg logit contributions | |
cream | -0.018 |
topp | -0.016 |
math | -0.016 |
cakes | -0.015 |
pizza | -0.015 |
Token | The |
Token ID | 383 |
Feature activation | +8.90e-02 |
Pos logit contributions | |
subt | +0.007 |
math | +0.007 |
cream | +0.007 |
Deal | +0.007 |
topp | +0.007 |
Neg logit contributions | |
en | -0.008 |
dot | -0.008 |
el | -0.008 |
fly | -0.008 |
ug | -0.007 |
Token | bears |
Token ID | 13062 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
subt | +0.002 |
math | +0.002 |
cream | +0.002 |
Deal | +0.002 |
topp | +0.002 |
Neg logit contributions | |
dot | -0.003 |
en | -0.003 |
el | -0.003 |
fly | -0.003 |
ug | -0.003 |
Token | stopped |
Token ID | 5025 |
Feature activation | +6.42e-01 |
Pos logit contributions | |
cream | +0.033 |
topp | +0.033 |
Deal | +0.032 |
subtract | +0.031 |
subt | +0.031 |
Neg logit contributions | |
dot | -0.038 |
en | -0.037 |
ays | -0.037 |
py | -0.037 |
ug | -0.037 |
Token | qu |
Token ID | 627 |
Feature activation | +2.02e+00 |
Pos logit contributions | |
math | +0.018 |
subt | +0.018 |
cream | +0.018 |
subtract | +0.018 |
topp | +0.017 |
Neg logit contributions | |
ches | -0.017 |
dot | -0.017 |
le | -0.017 |
ley | -0.017 |
ug | -0.017 |
Token | arre |
Token ID | 9624 |
Feature activation | +3.95e+00 |
Pos logit contributions | |
subt | +0.048 |
cream | +0.048 |
topp | +0.047 |
math | +0.047 |
subtract | +0.046 |
Neg logit contributions | |
dot | -0.064 |
fly | -0.063 |
en | -0.063 |
ug | -0.061 |
el | -0.061 |
Token | ling |
Token ID | 1359 |
Feature activation | +3.48e-01 |
Pos logit contributions | |
subt | +0.175 |
math | +0.173 |
cream | +0.171 |
Deal | +0.171 |
topp | +0.170 |
Neg logit contributions | |
en | -0.050 |
dot | -0.047 |
el | -0.046 |
fly | -0.045 |
ug | -0.043 |
Token | and |
Token ID | 290 |
Feature activation | +3.05e-01 |
Pos logit contributions | |
cream | +0.013 |
math | +0.013 |
subt | +0.012 |
topp | +0.012 |
pizza | +0.012 |
Neg logit contributions | |
dot | -0.007 |
wil | -0.006 |
ug | -0.006 |
inkle | -0.006 |
fly | -0.006 |
Token | asked |
Token ID | 1965 |
Feature activation | +1.84e-01 |
Pos logit contributions | |
subt | +0.009 |
math | +0.009 |
cream | +0.009 |
topp | +0.009 |
Deal | +0.009 |
Neg logit contributions | |
en | -0.008 |
dot | -0.008 |
el | -0.008 |
fly | -0.008 |
ug | -0.008 |
Token | him |
Token ID | 683 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
math | +0.005 |
subt | +0.005 |
cream | +0.005 |
Deal | +0.005 |
topp | +0.005 |
Neg logit contributions | |
dot | -0.005 |
en | -0.005 |
el | -0.005 |
fly | -0.005 |
wil | -0.005 |
Token | for |
Token ID | 329 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
subt | +0.004 |
cream | +0.004 |
math | +0.004 |
Deal | +0.004 |
topp | +0.004 |
Neg logit contributions | |
en | -0.004 |
dot | -0.004 |
fly | -0.004 |
el | -0.004 |
ug | -0.004 |
Token | on |
Token ID | 319 |
Feature activation | +4.13e-01 |
Pos logit contributions | |
subt | +0.020 |
math | +0.020 |
Deal | +0.020 |
cream | +0.020 |
topp | +0.019 |
Neg logit contributions | |
en | -0.019 |
dot | -0.019 |
fly | -0.018 |
el | -0.018 |
wil | -0.018 |
Token | , |
Token ID | 11 |
Feature activation | +2.31e-01 |
Pos logit contributions | |
math | +0.012 |
job | +0.011 |
subt | +0.011 |
cream | +0.010 |
pizza | +0.010 |
Neg logit contributions | |
dot | -0.014 |
fly | -0.013 |
ga | -0.013 |
wil | -0.012 |
ala | -0.012 |
Token | they |
Token ID | 484 |
Feature activation | +6.41e-01 |
Pos logit contributions | |
subt | +0.006 |
cream | +0.006 |
Deal | +0.006 |
topp | +0.006 |
math | +0.006 |
Neg logit contributions | |
en | -0.007 |
dot | -0.007 |
el | -0.006 |
fly | -0.006 |
ug | -0.006 |
Token | never |
Token ID | 1239 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
Deal | +0.018 |
subt | +0.018 |
math | +0.017 |
cream | +0.017 |
topp | +0.016 |
Neg logit contributions | |
dot | -0.018 |
ening | -0.018 |
en | -0.017 |
fly | -0.017 |
el | -0.017 |
Token | qu |
Token ID | 627 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
en | +0.004 |
dot | +0.004 |
fly | +0.004 |
el | +0.004 |
ening | +0.004 |
Neg logit contributions | |
subt | -0.004 |
math | -0.004 |
cream | -0.004 |
Deal | -0.004 |
topp | -0.004 |
Token | arre |
Token ID | 9624 |
Feature activation | +3.95e+00 |
Pos logit contributions | |
subt | +0.020 |
math | +0.020 |
topp | +0.019 |
cream | +0.019 |
subtract | +0.018 |
Neg logit contributions | |
dot | -0.081 |
fly | -0.081 |
en | -0.080 |
ley | -0.079 |
el | -0.079 |
Token | led |
Token ID | 992 |
Feature activation | +5.92e-01 |
Pos logit contributions | |
subt | +0.162 |
math | +0.159 |
cream | +0.158 |
Deal | +0.158 |
topp | +0.157 |
Neg logit contributions | |
en | -0.062 |
dot | -0.060 |
el | -0.058 |
fly | -0.057 |
ug | -0.056 |
Token | again |
Token ID | 757 |
Feature activation | +3.05e-01 |
Pos logit contributions | |
math | +0.018 |
cream | +0.016 |
subt | +0.016 |
Deal | +0.016 |
topp | +0.015 |
Neg logit contributions | |
dot | -0.016 |
el | -0.016 |
ley | -0.016 |
ug | -0.016 |
fly | -0.016 |
Token | over |
Token ID | 625 |
Feature activation | +1.65e-01 |
Pos logit contributions | |
subt | +0.011 |
math | +0.011 |
cream | +0.011 |
topp | +0.010 |
Deal | +0.010 |
Neg logit contributions | |
dot | -0.006 |
en | -0.006 |
el | -0.006 |
fly | -0.006 |
ug | -0.006 |
Token | clothes |
Token ID | 8242 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
math | +0.004 |
subt | +0.004 |
cream | +0.004 |
topp | +0.004 |
subtract | +0.004 |
Neg logit contributions | |
dot | -0.005 |
ches | -0.005 |
fly | -0.005 |
le | -0.005 |
ley | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | -4.54e-02 |
Pos logit contributions | |
cream | +0.051 |
math | +0.050 |
subt | +0.049 |
Deal | +0.048 |
topp | +0.048 |
Neg logit contributions | |
dot | -0.034 |
py | -0.034 |
el | -0.033 |
fly | -0.033 |
ug | -0.033 |
Token | â |
Token ID | 22940 |
Feature activation | +3.33e-01 |
Pos logit contributions | |
math | +0.006 |
subt | +0.005 |
cream | +0.005 |
subtract | +0.005 |
Fif | +0.005 |
Neg logit contributions | |
fly | -0.005 |
en | -0.005 |
dot | -0.005 |
ley | -0.005 |
ening | -0.005 |
Token | € |
Token ID | 26391 |
Feature activation | +8.08e-01 |
Pos logit contributions | |
math | +0.008 |
topp | +0.008 |
subt | +0.008 |
cream | +0.008 |
cakes | +0.008 |
Neg logit contributions | |
en | -0.011 |
el | -0.010 |
le | -0.010 |
ga | -0.010 |
ug | -0.010 |
Token | ™ |
Token ID | 8151 |
Feature activation | +2.73e+00 |
Pos logit contributions | |
subt | +0.022 |
math | +0.021 |
cream | +0.021 |
Deal | +0.020 |
subtract | +0.020 |
Neg logit contributions | |
en | -0.024 |
dot | -0.024 |
el | -0.024 |
fly | -0.024 |
ening | -0.023 |
Token | t |
Token ID | 83 |
Feature activation | +7.89e-01 |
Pos logit contributions | |
cream | +0.076 |
subt | +0.075 |
math | +0.075 |
topp | +0.072 |
Fif | +0.072 |
Neg logit contributions | |
dot | -0.077 |
en | -0.077 |
fly | -0.073 |
ening | -0.072 |
el | -0.070 |
Token | qu |
Token ID | 627 |
Feature activation | +2.19e+00 |
Pos logit contributions | |
subt | +0.021 |
cream | +0.020 |
math | +0.020 |
subtract | +0.020 |
topp | +0.020 |
Neg logit contributions | |
el | -0.023 |
dot | -0.023 |
en | -0.023 |
fly | -0.023 |
ug | -0.022 |
Token | arre |
Token ID | 9624 |
Feature activation | +3.94e+00 |
Pos logit contributions | |
math | +0.024 |
subt | +0.024 |
cream | +0.023 |
subtract | +0.023 |
topp | +0.023 |
Neg logit contributions | |
dot | -0.096 |
fly | -0.095 |
ley | -0.094 |
en | -0.093 |
ga | -0.093 |
Token | ling |
Token ID | 1359 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
cream | +0.149 |
subt | +0.147 |
topp | +0.146 |
math | +0.145 |
subtract | +0.142 |
Neg logit contributions | |
dot | -0.072 |
fly | -0.069 |
el | -0.067 |
ir | -0.065 |
en | -0.064 |
Token | , |
Token ID | 11 |
Feature activation | +4.23e-01 |
Pos logit contributions | |
math | +0.005 |
cream | +0.004 |
job | +0.004 |
cakes | +0.004 |
subt | +0.004 |
Neg logit contributions | |
dot | -0.005 |
fly | -0.005 |
el | -0.005 |
ley | -0.005 |
inkle | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | +4.19e-01 |
Pos logit contributions | |
cream | +0.011 |
subt | +0.010 |
math | +0.010 |
cakes | +0.010 |
subtract | +0.010 |
Neg logit contributions | |
fly | -0.014 |
py | -0.013 |
dot | -0.013 |
en | -0.013 |
el | -0.013 |
Token | furniture |
Token ID | 13091 |
Feature activation | +2.04e-01 |
Pos logit contributions | |
subt | +0.011 |
math | +0.011 |
cream | +0.011 |
Deal | +0.011 |
topp | +0.011 |
Neg logit contributions | |
en | -0.013 |
dot | -0.012 |
el | -0.012 |
fly | -0.012 |
ug | -0.012 |
Token | stayed |
Token ID | 9658 |
Feature activation | +6.05e-02 |
Pos logit contributions | |
cream | +0.006 |
math | +0.006 |
cakes | +0.006 |
subt | +0.005 |
topp | +0.005 |
Neg logit contributions | |
dot | -0.006 |
en | -0.006 |
ir | -0.006 |
el | -0.005 |
ga | -0.005 |
Token | together |
Token ID | 1978 |
Feature activation | -8.41e-01 |
Pos logit contributions | |
en | +0.037 |
dot | +0.037 |
el | +0.036 |
fly | +0.036 |
ug | +0.035 |
Neg logit contributions | |
subt | -0.031 |
math | -0.030 |
cream | -0.030 |
topp | -0.030 |
Deal | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | +1.92e-01 |
Pos logit contributions | |
dot | +0.022 |
el | +0.022 |
en | +0.022 |
fly | +0.022 |
ug | +0.021 |
Neg logit contributions | |
math | -0.025 |
subt | -0.025 |
cream | -0.024 |
topp | -0.024 |
Deal | -0.024 |
Token | They |
Token ID | 1119 |
Feature activation | +1.76e-01 |
Pos logit contributions | |
cream | +0.006 |
subt | +0.006 |
math | +0.006 |
Deal | +0.006 |
topp | +0.006 |
Neg logit contributions | |
en | -0.005 |
dot | -0.005 |
el | -0.005 |
fly | -0.005 |
ug | -0.005 |
Token | never |
Token ID | 1239 |
Feature activation | -4.32e-01 |
Pos logit contributions | |
subt | +0.005 |
Deal | +0.005 |
math | +0.005 |
cream | +0.005 |
subtract | +0.005 |
Neg logit contributions | |
dot | -0.005 |
en | -0.005 |
el | -0.005 |
fly | -0.005 |
inkle | -0.005 |
Token | qu |
Token ID | 627 |
Feature activation | +1.63e+00 |
Pos logit contributions | |
en | +0.011 |
dot | +0.011 |
fly | +0.011 |
el | +0.010 |
le | +0.010 |
Neg logit contributions | |
cream | -0.013 |
subt | -0.013 |
math | -0.013 |
Deal | -0.013 |
topp | -0.013 |
Token | arre |
Token ID | 9624 |
Feature activation | +3.94e+00 |
Pos logit contributions | |
subt | +0.017 |
math | +0.017 |
topp | +0.017 |
cream | +0.016 |
subtract | +0.016 |
Neg logit contributions | |
dot | -0.074 |
fly | -0.073 |
ley | -0.072 |
en | -0.072 |
el | -0.072 |
Token | led |
Token ID | 992 |
Feature activation | +7.72e-01 |
Pos logit contributions | |
subt | +0.156 |
math | +0.154 |
cream | +0.153 |
topp | +0.152 |
Deal | +0.152 |
Neg logit contributions | |
en | -0.066 |
dot | -0.066 |
el | -0.064 |
fly | -0.063 |
ug | -0.061 |
Token | again |
Token ID | 757 |
Feature activation | +2.83e-01 |
Pos logit contributions | |
math | +0.023 |
cream | +0.022 |
topp | +0.021 |
pizza | +0.021 |
subt | +0.020 |
Neg logit contributions | |
el | -0.021 |
dot | -0.021 |
ug | -0.021 |
ley | -0.021 |
inkle | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | +3.49e-01 |
Pos logit contributions | |
subt | +0.011 |
math | +0.011 |
cream | +0.010 |
topp | +0.010 |
subtract | +0.010 |
Neg logit contributions | |
dot | -0.005 |
en | -0.005 |
el | -0.005 |
fly | -0.005 |
ug | -0.005 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +3.75e-01 |
Pos logit contributions | |
cream | +0.010 |
subt | +0.010 |
math | +0.010 |
topp | +0.010 |
Deal | +0.010 |
Neg logit contributions | |
en | -0.010 |
dot | -0.009 |
el | -0.009 |
fly | -0.009 |
ug | -0.009 |
Token | Once |
Token ID | 7454 |
Feature activation | +1.60e-01 |
Pos logit contributions | |
math | +0.007 |
subt | +0.007 |
cream | +0.007 |
topp | +0.007 |
las | +0.007 |
Neg logit contributions | |
wil | -0.014 |
dot | -0.014 |
ening | -0.014 |
en | -0.014 |
ug | -0.013 |
Token | He |
Token ID | 679 |
Feature activation | +4.85e-01 |
Pos logit contributions | |
subt | +0.007 |
cream | +0.007 |
Deal | +0.007 |
topp | +0.007 |
subtract | +0.007 |
Neg logit contributions | |
dot | -0.006 |
en | -0.006 |
el | -0.006 |
ug | -0.005 |
ays | -0.005 |
Token | smells |
Token ID | 25760 |
Feature activation | +5.43e-01 |
Pos logit contributions | |
topp | +0.014 |
math | +0.014 |
subt | +0.013 |
cream | +0.013 |
las | +0.013 |
Neg logit contributions | |
ug | -0.014 |
ir | -0.014 |
dot | -0.014 |
fly | -0.013 |
el | -0.013 |
Token | the |
Token ID | 262 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
subt | +0.014 |
math | +0.014 |
cream | +0.014 |
topp | +0.014 |
Deal | +0.014 |
Neg logit contributions | |
en | -0.016 |
dot | -0.016 |
el | -0.016 |
fly | -0.016 |
ug | -0.016 |
Token | y |
Token ID | 331 |
Feature activation | +1.51e+00 |
Pos logit contributions | |
subt | +0.006 |
Deal | +0.005 |
subtract | +0.005 |
math | +0.005 |
bek | +0.005 |
Neg logit contributions | |
dot | -0.012 |
el | -0.011 |
fly | -0.011 |
en | -0.011 |
le | -0.011 |
Token | ummy |
Token ID | 13513 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
subt | +0.054 |
math | +0.052 |
Deal | +0.052 |
subtract | +0.051 |
cream | +0.050 |
Neg logit contributions | |
en | -0.034 |
el | -0.034 |
dot | -0.033 |
le | -0.031 |
ening | -0.031 |
Token | topp |
Token ID | 23126 |
Feature activation | +3.94e+00 |
Pos logit contributions | |
subt | +0.023 |
Deal | +0.021 |
subtract | +0.020 |
math | +0.019 |
Fif | +0.019 |
Neg logit contributions | |
le | -0.046 |
en | -0.046 |
el | -0.045 |
dot | -0.045 |
ening | -0.044 |
Token | ings |
Token ID | 654 |
Feature activation | -4.24e-01 |
Pos logit contributions | |
subt | +0.173 |
math | +0.168 |
Deal | +0.167 |
subtract | +0.165 |
cream | +0.163 |
Neg logit contributions | |
en | -0.054 |
el | -0.053 |
dot | -0.051 |
fly | -0.049 |
le | -0.048 |
Token | . |
Token ID | 13 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
dot | +0.011 |
en | +0.010 |
fly | +0.010 |
el | +0.010 |
ug | +0.010 |
Neg logit contributions | |
cream | -0.013 |
math | -0.013 |
subt | -0.013 |
topp | -0.013 |
Deal | -0.013 |
Token | He |
Token ID | 679 |
Feature activation | +4.38e-01 |
Pos logit contributions | |
subt | +0.004 |
Deal | +0.004 |
cream | +0.004 |
topp | +0.004 |
subtract | +0.004 |
Neg logit contributions | |
dot | -0.003 |
en | -0.003 |
el | -0.003 |
ays | -0.003 |
ug | -0.003 |
Token | wants |
Token ID | 3382 |
Feature activation | -2.35e-01 |
Pos logit contributions | |
topp | +0.013 |
cream | +0.012 |
math | +0.012 |
subt | +0.012 |
las | +0.012 |
Neg logit contributions | |
ug | -0.013 |
dot | -0.012 |
ir | -0.012 |
fly | -0.012 |
el | -0.012 |
Token | to |
Token ID | 284 |
Feature activation | +6.60e-01 |
Pos logit contributions | |
en | +0.007 |
dot | +0.007 |
el | +0.007 |
fly | +0.007 |
le | +0.007 |
Neg logit contributions | |
subt | -0.006 |
Deal | -0.006 |
math | -0.006 |
cream | -0.006 |
topp | -0.006 |
Token | he |
Token ID | 339 |
Feature activation | -4.35e-01 |
Pos logit contributions | |
math | +0.007 |
cream | +0.007 |
subt | +0.007 |
topp | +0.007 |
Deal | +0.007 |
Neg logit contributions | |
dot | -0.011 |
en | -0.011 |
ir | -0.011 |
ug | -0.011 |
ening | -0.011 |
Token | would |
Token ID | 561 |
Feature activation | -3.37e-01 |
Pos logit contributions | |
dot | +0.014 |
ir | +0.013 |
ays | +0.013 |
ug | +0.013 |
el | +0.013 |
Neg logit contributions | |
Deal | -0.011 |
math | -0.011 |
cream | -0.011 |
topp | -0.011 |
subt | -0.010 |
Token | take |
Token ID | 1011 |
Feature activation | +9.38e-01 |
Pos logit contributions | |
dot | +0.012 |
ays | +0.012 |
ug | +0.012 |
el | +0.012 |
ening | +0.011 |
Neg logit contributions | |
math | -0.007 |
cream | -0.007 |
subt | -0.007 |
subtract | -0.007 |
topp | -0.007 |
Token | the |
Token ID | 262 |
Feature activation | +3.12e-01 |
Pos logit contributions | |
subt | +0.024 |
math | +0.024 |
cream | +0.023 |
topp | +0.023 |
subtract | +0.022 |
Neg logit contributions | |
dot | -0.029 |
en | -0.029 |
fly | -0.029 |
el | -0.028 |
ug | -0.028 |
Token | av |
Token ID | 1196 |
Feature activation | +3.77e+00 |
Pos logit contributions | |
subt | +0.008 |
math | +0.008 |
cream | +0.008 |
topp | +0.008 |
Deal | +0.008 |
Neg logit contributions | |
en | -0.010 |
dot | -0.009 |
el | -0.009 |
fly | -0.009 |
ug | -0.009 |
Token | oc |
Token ID | 420 |
Feature activation | +3.93e+00 |
Pos logit contributions | |
subt | +0.117 |
math | +0.117 |
Deal | +0.115 |
cream | +0.115 |
cakes | +0.112 |
Neg logit contributions | |
en | -0.101 |
el | -0.100 |
fly | -0.094 |
le | -0.093 |
ir | -0.093 |
Token | ados |
Token ID | 22484 |
Feature activation | +1.65e+00 |
Pos logit contributions | |
Deal | +0.110 |
subt | +0.110 |
math | +0.110 |
cream | +0.107 |
Fif | +0.107 |
Neg logit contributions | |
el | -0.117 |
en | -0.111 |
fly | -0.106 |
ors | -0.105 |
ug | -0.105 |
Token | and |
Token ID | 290 |
Feature activation | +8.36e-01 |
Pos logit contributions | |
subt | +0.045 |
math | +0.044 |
cream | +0.044 |
Deal | +0.043 |
topp | +0.043 |
Neg logit contributions | |
en | -0.048 |
dot | -0.047 |
el | -0.047 |
fly | -0.046 |
ug | -0.045 |
Token | scatter |
Token ID | 41058 |
Feature activation | -3.80e-01 |
Pos logit contributions | |
Deal | +0.021 |
subt | +0.020 |
math | +0.020 |
Fif | +0.019 |
cream | +0.019 |
Neg logit contributions | |
en | -0.026 |
fly | -0.025 |
el | -0.025 |
dot | -0.025 |
ley | -0.024 |
Token | them |
Token ID | 606 |
Feature activation | +8.06e-02 |
Pos logit contributions | |
dot | +0.008 |
fly | +0.008 |
ug | +0.008 |
ga | +0.007 |
en | +0.007 |
Neg logit contributions | |
cream | -0.014 |
topp | -0.014 |
math | -0.014 |
subt | -0.013 |
cakes | -0.013 |
Token | around |
Token ID | 1088 |
Feature activation | -1.00e+00 |
Pos logit contributions | |
cream | +0.003 |
math | +0.003 |
topp | +0.002 |
subt | +0.002 |
Deal | +0.002 |
Neg logit contributions | |
dot | -0.002 |
en | -0.002 |
ug | -0.002 |
el | -0.002 |
ga | -0.002 |
Token | lively |
Token ID | 29696 |
Feature activation | +2.13e-01 |
Pos logit contributions | |
subt | +0.012 |
math | +0.012 |
cream | +0.012 |
topp | +0.012 |
Deal | +0.012 |
Neg logit contributions | |
en | -0.016 |
dot | -0.016 |
el | -0.016 |
fly | -0.015 |
ug | -0.015 |
Token | sound |
Token ID | 2128 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
math | +0.005 |
cream | +0.005 |
subt | +0.005 |
topp | +0.005 |
subtract | +0.005 |
Neg logit contributions | |
el | -0.007 |
fly | -0.007 |
pt | -0.006 |
en | -0.006 |
ir | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | +1.59e-01 |
Pos logit contributions | |
el | +0.007 |
ley | +0.007 |
fly | +0.007 |
dot | +0.007 |
ir | +0.007 |
Neg logit contributions | |
cream | -0.008 |
math | -0.008 |
subt | -0.008 |
topp | -0.007 |
subtract | -0.007 |
Token | The |
Token ID | 383 |
Feature activation | +5.56e-01 |
Pos logit contributions | |
subt | +0.004 |
math | +0.004 |
subtract | +0.004 |
topp | +0.004 |
Deal | +0.004 |
Neg logit contributions | |
en | -0.005 |
dot | -0.005 |
le | -0.005 |
el | -0.005 |
fly | -0.005 |
Token | qu |
Token ID | 627 |
Feature activation | +1.41e+00 |
Pos logit contributions | |
subt | +0.016 |
math | +0.016 |
cream | +0.016 |
Deal | +0.015 |
topp | +0.015 |
Neg logit contributions | |
en | -0.015 |
dot | -0.015 |
el | -0.015 |
fly | -0.015 |
ug | -0.015 |
Token | arre |
Token ID | 9624 |
Feature activation | +3.91e+00 |
Pos logit contributions | |
topp | +0.039 |
cream | +0.039 |
subtract | +0.038 |
subt | +0.038 |
math | +0.037 |
Neg logit contributions | |
ley | -0.038 |
fly | -0.037 |
dot | -0.037 |
ala | -0.037 |
wil | -0.037 |
Token | ling |
Token ID | 1359 |
Feature activation | +2.44e-01 |
Pos logit contributions | |
cream | +0.152 |
topp | +0.151 |
subt | +0.148 |
math | +0.146 |
Deal | +0.144 |
Neg logit contributions | |
dot | -0.066 |
fly | -0.063 |
wil | -0.061 |
el | -0.061 |
ala | -0.059 |
Token | stopped |
Token ID | 5025 |
Feature activation | +8.66e-01 |
Pos logit contributions | |
cream | +0.007 |
topp | +0.006 |
math | +0.006 |
subt | +0.006 |
job | +0.006 |
Neg logit contributions | |
unks | -0.007 |
ays | -0.007 |
inkle | -0.006 |
ir | -0.006 |
wil | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | +3.83e-01 |
Pos logit contributions | |
math | +0.024 |
subtract | +0.024 |
subt | +0.024 |
cream | +0.023 |
topp | +0.022 |
Neg logit contributions | |
ches | -0.024 |
ley | -0.024 |
le | -0.023 |
wil | -0.022 |
ays | -0.022 |
Token | they |
Token ID | 484 |
Feature activation | +2.82e-01 |
Pos logit contributions | |
subt | +0.011 |
math | +0.011 |
cream | +0.011 |
topp | +0.011 |
Deal | +0.011 |
Neg logit contributions | |
en | -0.010 |
dot | -0.010 |
el | -0.010 |
fly | -0.010 |
ug | -0.010 |
Token | heard |
Token ID | 2982 |
Feature activation | +6.98e-01 |
Pos logit contributions | |
subt | +0.007 |
cream | +0.007 |
Deal | +0.007 |
topp | +0.007 |
subtract | +0.007 |
Neg logit contributions | |
el | -0.008 |
en | -0.008 |
dot | -0.008 |
ening | -0.008 |
fly | -0.008 |
Token | � |
Token ID | 129 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
subt | +0.030 |
math | +0.029 |
cream | +0.029 |
topp | +0.029 |
Deal | +0.029 |
Neg logit contributions | |
en | -0.031 |
dot | -0.031 |
el | -0.031 |
fly | -0.030 |
ug | -0.030 |
Token | � |
Token ID | 241 |
Feature activation | +9.60e-01 |
Pos logit contributions | |
subt | +0.028 |
math | +0.028 |
cream | +0.027 |
Deal | +0.027 |
subtract | +0.027 |
Neg logit contributions | |
dot | -0.029 |
fly | -0.029 |
en | -0.028 |
el | -0.028 |
ug | -0.028 |
Token | That |
Token ID | 2504 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
subt | +0.024 |
math | +0.024 |
cream | +0.023 |
Deal | +0.023 |
topp | +0.023 |
Neg logit contributions | |
en | -0.030 |
dot | -0.030 |
el | -0.029 |
fly | -0.029 |
ug | -0.029 |
Token | â |
Token ID | 22940 |
Feature activation | +5.88e-02 |
Pos logit contributions | |
math | +0.033 |
subt | +0.033 |
cream | +0.033 |
topp | +0.033 |
Deal | +0.032 |
Neg logit contributions | |
dot | -0.025 |
en | -0.024 |
ir | -0.024 |
ug | -0.024 |
el | -0.024 |
Token | € |
Token ID | 26391 |
Feature activation | +8.28e-01 |
Pos logit contributions | |
math | +0.001 |
topp | +0.001 |
subt | +0.001 |
cream | +0.001 |
cakes | +0.001 |
Neg logit contributions | |
en | -0.002 |
el | -0.002 |
le | -0.002 |
ug | -0.002 |
dot | -0.002 |
Token | ™ |
Token ID | 8151 |
Feature activation | +3.89e+00 |
Pos logit contributions | |
subt | +0.023 |
math | +0.023 |
cream | +0.023 |
topp | +0.023 |
Deal | +0.023 |
Neg logit contributions | |
en | -0.023 |
dot | -0.023 |
el | -0.023 |
fly | -0.022 |
ug | -0.022 |
Token | s |
Token ID | 82 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
Deal | +0.111 |
cream | +0.111 |
subt | +0.110 |
math | +0.109 |
topp | +0.109 |
Neg logit contributions | |
dot | -0.112 |
en | -0.107 |
ening | -0.103 |
fly | -0.102 |
py | -0.101 |
Token | ok |
Token ID | 12876 |
Feature activation | +8.96e-01 |
Pos logit contributions | |
math | +0.041 |
Deal | +0.040 |
cream | +0.040 |
subt | +0.040 |
topp | +0.039 |
Neg logit contributions | |
dot | -0.035 |
en | -0.034 |
ir | -0.033 |
fly | -0.033 |
ug | -0.033 |
Token | , |
Token ID | 11 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
math | +0.028 |
cream | +0.028 |
subt | +0.028 |
Deal | +0.027 |
topp | +0.027 |
Neg logit contributions | |
dot | -0.023 |
en | -0.022 |
wil | -0.022 |
ir | -0.021 |
fly | -0.021 |
Token | you |
Token ID | 345 |
Feature activation | +9.45e-01 |
Pos logit contributions | |
math | +0.030 |
subt | +0.030 |
cream | +0.029 |
topp | +0.029 |
subtract | +0.028 |
Neg logit contributions | |
dot | -0.030 |
en | -0.030 |
el | -0.030 |
fly | -0.030 |
ir | -0.029 |
Token | â |
Token ID | 22940 |
Feature activation | +1.55e-01 |
Pos logit contributions | |
job | +0.021 |
Deal | +0.020 |
math | +0.020 |
pay | +0.020 |
list | +0.020 |
Neg logit contributions | |
ening | -0.030 |
dot | -0.030 |
ays | -0.029 |
eds | -0.029 |
ga | -0.029 |
Token | � |
Token ID | 129 |
Feature activation | +7.85e-01 |
Pos logit contributions | |
subt | +0.029 |
math | +0.028 |
cream | +0.028 |
Deal | +0.027 |
topp | +0.027 |
Neg logit contributions | |
en | -0.031 |
dot | -0.031 |
el | -0.031 |
fly | -0.030 |
ug | -0.030 |
Token | � |
Token ID | 241 |
Feature activation | +4.34e-01 |
Pos logit contributions | |
subt | +0.022 |
math | +0.021 |
cream | +0.021 |
Deal | +0.021 |
subtract | +0.021 |
Neg logit contributions | |
dot | -0.023 |
fly | -0.022 |
en | -0.022 |
el | -0.022 |
ug | -0.021 |
Token | That |
Token ID | 2504 |
Feature activation | +8.09e-01 |
Pos logit contributions | |
subt | +0.011 |
math | +0.010 |
cream | +0.010 |
Deal | +0.010 |
topp | +0.010 |
Neg logit contributions | |
en | -0.014 |
dot | -0.014 |
el | -0.014 |
fly | -0.014 |
ug | -0.013 |
Token | â |
Token ID | 22940 |
Feature activation | -1.83e-02 |
Pos logit contributions | |
math | +0.024 |
cream | +0.024 |
subt | +0.024 |
topp | +0.024 |
Deal | +0.023 |
Neg logit contributions | |
ir | -0.022 |
dot | -0.021 |
ug | -0.021 |
en | -0.021 |
fly | -0.021 |
Token | € |
Token ID | 26391 |
Feature activation | +8.75e-01 |
Pos logit contributions | |
en | +0.001 |
el | +0.001 |
le | +0.001 |
ug | +0.001 |
dot | +0.001 |
Neg logit contributions | |
math | -0.000 |
subt | -0.000 |
topp | -0.000 |
cream | -0.000 |
cakes | -0.000 |
Token | ™ |
Token ID | 8151 |
Feature activation | +3.89e+00 |
Pos logit contributions | |
subt | +0.025 |
math | +0.024 |
cream | +0.024 |
topp | +0.024 |
Deal | +0.024 |
Neg logit contributions | |
en | -0.024 |
dot | -0.024 |
el | -0.024 |
fly | -0.024 |
ug | -0.023 |
Token | s |
Token ID | 82 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
cream | +0.114 |
Deal | +0.113 |
math | +0.112 |
subt | +0.111 |
topp | +0.111 |
Neg logit contributions | |
dot | -0.110 |
en | -0.105 |
ening | -0.101 |
fly | -0.100 |
py | -0.099 |
Token | ok |
Token ID | 12876 |
Feature activation | +8.22e-01 |
Pos logit contributions | |
Deal | +0.036 |
math | +0.036 |
cream | +0.035 |
subt | +0.035 |
topp | +0.034 |
Neg logit contributions | |
dot | -0.031 |
en | -0.030 |
ir | -0.029 |
fly | -0.029 |
ug | -0.029 |
Token | . |
Token ID | 13 |
Feature activation | +3.60e-01 |
Pos logit contributions | |
math | +0.027 |
subt | +0.026 |
cream | +0.026 |
Deal | +0.026 |
topp | +0.026 |
Neg logit contributions | |
dot | -0.020 |
en | -0.019 |
wil | -0.019 |
fly | -0.019 |
el | -0.019 |
Token | We |
Token ID | 775 |
Feature activation | +8.43e-01 |
Pos logit contributions | |
subt | +0.010 |
math | +0.010 |
cream | +0.010 |
Deal | +0.010 |
topp | +0.010 |
Neg logit contributions | |
en | -0.010 |
dot | -0.010 |
el | -0.010 |
fly | -0.010 |
ug | -0.009 |
Token | all |
Token ID | 477 |
Feature activation | -2.60e-01 |
Pos logit contributions | |
Deal | +0.021 |
cream | +0.019 |
subt | +0.019 |
job | +0.019 |
topp | +0.019 |
Neg logit contributions | |
ening | -0.025 |
dot | -0.025 |
en | -0.024 |
ays | -0.024 |
predators | -0.024 |
Token | and |
Token ID | 290 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
en | +0.018 |
dot | +0.018 |
py | +0.017 |
fly | +0.017 |
ir | +0.017 |
Neg logit contributions | |
cream | -0.019 |
math | -0.017 |
topp | -0.017 |
pizza | -0.017 |
subtract | -0.017 |
Token | wearing |
Token ID | 5762 |
Feature activation | +2.20e-02 |
Pos logit contributions | |
en | +0.008 |
dot | +0.008 |
el | +0.008 |
fly | +0.008 |
ug | +0.008 |
Neg logit contributions | |
subt | -0.007 |
math | -0.007 |
Deal | -0.007 |
cream | -0.006 |
topp | -0.006 |
Token | gloves |
Token ID | 18051 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
subt | +0.001 |
Deal | +0.001 |
math | +0.001 |
subtract | +0.001 |
cream | +0.001 |
Neg logit contributions | |
en | -0.001 |
el | -0.001 |
dot | -0.001 |
le | -0.001 |
ening | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
dot | +0.032 |
ir | +0.031 |
inkle | +0.031 |
el | +0.031 |
en | +0.031 |
Neg logit contributions | |
cream | -0.026 |
math | -0.024 |
topp | -0.024 |
pizza | -0.023 |
Deal | -0.023 |
Token | The |
Token ID | 383 |
Feature activation | +8.82e-01 |
Pos logit contributions | |
cream | +0.039 |
Deal | +0.038 |
subt | +0.038 |
math | +0.037 |
topp | +0.036 |
Neg logit contributions | |
en | -0.031 |
el | -0.031 |
dot | -0.030 |
fly | -0.029 |
ening | -0.028 |
Token | paramed |
Token ID | 34581 |
Feature activation | +3.87e+00 |
Pos logit contributions | |
subt | +0.024 |
math | +0.023 |
Deal | +0.023 |
topp | +0.022 |
subtract | +0.022 |
Neg logit contributions | |
el | -0.026 |
wil | -0.026 |
ga | -0.025 |
dot | -0.025 |
ley | -0.024 |
Token | ic |
Token ID | 291 |
Feature activation | +5.37e-01 |
Pos logit contributions | |
math | +0.157 |
subt | +0.145 |
topp | +0.145 |
las | +0.144 |
Fif | +0.141 |
Neg logit contributions | |
en | -0.080 |
el | -0.078 |
ie | -0.073 |
le | -0.070 |
ening | -0.070 |
Token | was |
Token ID | 373 |
Feature activation | +2.46e-02 |
Pos logit contributions | |
subt | +0.017 |
cream | +0.017 |
topp | +0.017 |
math | +0.017 |
Deal | +0.017 |
Neg logit contributions | |
dot | -0.013 |
en | -0.013 |
fly | -0.013 |
el | -0.012 |
ug | -0.012 |
Token | very |
Token ID | 845 |
Feature activation | +6.51e-01 |
Pos logit contributions | |
cream | +0.001 |
math | +0.001 |
subt | +0.001 |
topp | +0.001 |
subtract | +0.001 |
Neg logit contributions | |
dot | -0.001 |
en | -0.001 |
fly | -0.001 |
ley | -0.001 |
ir | -0.001 |
Token | polite |
Token ID | 23507 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
subt | +0.016 |
math | +0.015 |
subtract | +0.015 |
Deal | +0.015 |
topp | +0.015 |
Neg logit contributions | |
wil | -0.020 |
inkle | -0.020 |
el | -0.020 |
ug | -0.020 |
en | -0.020 |
Token | and |
Token ID | 290 |
Feature activation | +6.86e-01 |
Pos logit contributions | |
math | +0.040 |
cream | +0.040 |
subt | +0.039 |
topp | +0.039 |
subtract | +0.038 |
Neg logit contributions | |
ir | -0.021 |
dot | -0.020 |
inkle | -0.020 |
el | -0.020 |
fly | -0.020 |
Token | together |
Token ID | 1978 |
Feature activation | +2.91e-02 |
Pos logit contributions | |
en | +0.007 |
ug | +0.006 |
dot | +0.006 |
le | +0.006 |
freely | +0.006 |
Neg logit contributions | |
subt | -0.009 |
topp | -0.009 |
las | -0.009 |
subtract | -0.009 |
Fif | -0.008 |
Token | happily |
Token ID | 18177 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
subt | +0.001 |
cream | +0.001 |
math | +0.001 |
topp | +0.001 |
Deal | +0.001 |
Neg logit contributions | |
en | -0.001 |
dot | -0.001 |
el | -0.001 |
fly | -0.001 |
ug | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | +1.40e-01 |
Pos logit contributions | |
en | +0.005 |
dot | +0.005 |
fly | +0.005 |
ug | +0.005 |
el | +0.004 |
Neg logit contributions | |
subt | -0.010 |
topp | -0.010 |
math | -0.010 |
cream | -0.010 |
las | -0.010 |
Token | never |
Token ID | 1239 |
Feature activation | -3.06e-01 |
Pos logit contributions | |
cream | +0.005 |
subt | +0.005 |
math | +0.005 |
topp | +0.005 |
Deal | +0.005 |
Neg logit contributions | |
en | -0.003 |
dot | -0.003 |
fly | -0.003 |
el | -0.003 |
ga | -0.003 |
Token | qu |
Token ID | 627 |
Feature activation | +1.89e+00 |
Pos logit contributions | |
en | +0.008 |
dot | +0.007 |
fly | +0.007 |
pt | +0.007 |
ening | +0.007 |
Neg logit contributions | |
math | -0.010 |
Deal | -0.010 |
cream | -0.010 |
subt | -0.010 |
las | -0.010 |
Token | arre |
Token ID | 9624 |
Feature activation | +3.85e+00 |
Pos logit contributions | |
subt | +0.027 |
math | +0.026 |
cream | +0.026 |
topp | +0.025 |
Deal | +0.025 |
Neg logit contributions | |
en | -0.079 |
dot | -0.079 |
el | -0.078 |
fly | -0.078 |
ug | -0.077 |
Token | lled |
Token ID | 3353 |
Feature activation | +7.04e-01 |
Pos logit contributions | |
subt | +0.166 |
math | +0.163 |
Deal | +0.162 |
cream | +0.160 |
subtract | +0.159 |
Neg logit contributions | |
en | -0.057 |
dot | -0.051 |
le | -0.051 |
ug | -0.051 |
el | -0.050 |
Token | again |
Token ID | 757 |
Feature activation | +2.21e-01 |
Pos logit contributions | |
math | +0.024 |
cream | +0.024 |
subt | +0.024 |
topp | +0.024 |
Deal | +0.024 |
Neg logit contributions | |
dot | -0.015 |
fly | -0.015 |
el | -0.015 |
en | -0.015 |
ug | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | +2.43e-01 |
Pos logit contributions | |
subt | +0.008 |
Deal | +0.007 |
cream | +0.007 |
math | +0.007 |
topp | +0.007 |
Neg logit contributions | |
en | -0.005 |
dot | -0.005 |
el | -0.005 |
fly | -0.005 |
le | -0.005 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.94e-01 |
Pos logit contributions | |
subt | +0.007 |
cream | +0.007 |
topp | +0.007 |
math | +0.007 |
Deal | +0.007 |
Neg logit contributions | |
en | -0.006 |
dot | -0.006 |
el | -0.006 |
fly | -0.006 |
ug | -0.006 |
Token | Once |
Token ID | 7454 |
Feature activation | +1.49e-01 |
Pos logit contributions | |
math | +0.006 |
cream | +0.006 |
subt | +0.006 |
topp | +0.006 |
Deal | +0.006 |
Neg logit contributions | |
dot | -0.010 |
en | -0.010 |
fly | -0.010 |
ug | -0.010 |
ening | -0.010 |
Token | both |
Token ID | 1111 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
en | +0.002 |
dot | +0.002 |
el | +0.002 |
ches | +0.002 |
fly | +0.002 |
Neg logit contributions | |
math | -0.002 |
subt | -0.002 |
cream | -0.002 |
topp | -0.002 |
Deal | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | -4.57e-01 |
Pos logit contributions | |
cream | +0.003 |
math | +0.003 |
subt | +0.003 |
topp | +0.003 |
cakes | +0.003 |
Neg logit contributions | |
en | -0.003 |
dot | -0.003 |
ga | -0.003 |
el | -0.003 |
ir | -0.003 |
Token | friends |
Token ID | 2460 |
Feature activation | +9.71e-01 |
Pos logit contributions | |
en | +0.015 |
dot | +0.015 |
el | +0.015 |
fly | +0.015 |
ug | +0.015 |
Neg logit contributions | |
subt | -0.010 |
math | -0.010 |
cream | -0.010 |
topp | -0.010 |
Deal | -0.010 |
Token | stopped |
Token ID | 5025 |
Feature activation | +4.60e-01 |
Pos logit contributions | |
cream | +0.021 |
subt | +0.021 |
subtract | +0.020 |
Deal | +0.020 |
cakes | +0.019 |
Neg logit contributions | |
dot | -0.031 |
ays | -0.031 |
py | -0.030 |
unks | -0.030 |
inkle | -0.029 |
Token | qu |
Token ID | 627 |
Feature activation | +1.79e+00 |
Pos logit contributions | |
subt | +0.013 |
math | +0.013 |
subtract | +0.012 |
cream | +0.012 |
topp | +0.012 |
Neg logit contributions | |
ches | -0.013 |
ley | -0.013 |
le | -0.012 |
dot | -0.012 |
ug | -0.012 |
Token | arre |
Token ID | 9624 |
Feature activation | +3.84e+00 |
Pos logit contributions | |
subt | +0.040 |
cream | +0.039 |
subtract | +0.039 |
topp | +0.039 |
math | +0.039 |
Neg logit contributions | |
fly | -0.057 |
dot | -0.057 |
ley | -0.057 |
wil | -0.055 |
ug | -0.055 |
Token | ling |
Token ID | 1359 |
Feature activation | -6.57e-02 |
Pos logit contributions | |
subt | +0.163 |
cream | +0.162 |
math | +0.160 |
topp | +0.160 |
Deal | +0.158 |
Neg logit contributions | |
dot | -0.052 |
en | -0.050 |
el | -0.049 |
fly | -0.049 |
ug | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
dot | +0.002 |
unks | +0.001 |
ug | +0.001 |
inkle | +0.001 |
ley | +0.001 |
Neg logit contributions | |
cream | -0.002 |
math | -0.002 |
subt | -0.002 |
topp | -0.002 |
pizza | -0.002 |
Token | They |
Token ID | 1119 |
Feature activation | +3.35e-01 |
Pos logit contributions | |
subt | +0.004 |
math | +0.004 |
subtract | +0.004 |
topp | +0.004 |
Deal | +0.003 |
Neg logit contributions | |
le | -0.005 |
en | -0.005 |
ie | -0.005 |
inkle | -0.005 |
dot | -0.005 |
Token | decided |
Token ID | 3066 |
Feature activation | -3.99e-01 |
Pos logit contributions | |
cream | +0.008 |
subt | +0.008 |
Deal | +0.008 |
subtract | +0.007 |
topp | +0.007 |
Neg logit contributions | |
dot | -0.011 |
ays | -0.010 |
en | -0.010 |
ug | -0.010 |
el | -0.010 |
Token | to |
Token ID | 284 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
en | +0.013 |
dot | +0.012 |
el | +0.012 |
fly | +0.012 |
ug | +0.012 |
Neg logit contributions | |
subt | -0.010 |
math | -0.010 |
cream | -0.010 |
Deal | -0.010 |
topp | -0.010 |
Token | of |
Token ID | 286 |
Feature activation | +1.71e+00 |
Pos logit contributions | |
dot | +0.029 |
fly | +0.028 |
en | +0.028 |
el | +0.027 |
ley | +0.027 |
Neg logit contributions | |
cream | -0.037 |
math | -0.036 |
topp | -0.036 |
subt | -0.035 |
subtract | -0.034 |
Token | cheese |
Token ID | 9891 |
Feature activation | -3.34e-02 |
Pos logit contributions | |
subt | +0.027 |
cream | +0.027 |
math | +0.027 |
topp | +0.026 |
Deal | +0.025 |
Neg logit contributions | |
dot | -0.070 |
en | -0.070 |
fly | -0.069 |
el | -0.069 |
ug | -0.068 |
Token | and |
Token ID | 290 |
Feature activation | -5.18e-01 |
Pos logit contributions | |
en | +0.001 |
el | +0.001 |
dot | +0.001 |
ir | +0.001 |
ening | +0.001 |
Neg logit contributions | |
cream | -0.001 |
pizza | -0.001 |
topp | -0.001 |
math | -0.001 |
cakes | -0.001 |
Token | sauce |
Token ID | 10746 |
Feature activation | -5.96e-02 |
Pos logit contributions | |
en | +0.018 |
dot | +0.018 |
el | +0.018 |
fly | +0.018 |
ug | +0.017 |
Neg logit contributions | |
subt | -0.011 |
math | -0.010 |
Deal | -0.010 |
cream | -0.010 |
subtract | -0.010 |
Token | and |
Token ID | 290 |
Feature activation | -4.83e-01 |
Pos logit contributions | |
dot | +0.002 |
en | +0.002 |
el | +0.002 |
fly | +0.002 |
ir | +0.002 |
Neg logit contributions | |
cream | -0.002 |
pizza | -0.002 |
topp | -0.002 |
math | -0.002 |
cakes | -0.002 |
Token | topp |
Token ID | 23126 |
Feature activation | +3.84e+00 |
Pos logit contributions | |
en | +0.017 |
dot | +0.017 |
el | +0.017 |
fly | +0.017 |
ug | +0.016 |
Neg logit contributions | |
subt | -0.010 |
math | -0.009 |
Deal | -0.009 |
cream | -0.009 |
topp | -0.009 |
Token | ings |
Token ID | 654 |
Feature activation | -9.70e-01 |
Pos logit contributions | |
cream | +0.152 |
topp | +0.151 |
math | +0.144 |
cakes | +0.143 |
Deal | +0.141 |
Neg logit contributions | |
dot | -0.070 |
fly | -0.067 |
en | -0.065 |
wil | -0.064 |
ga | -0.062 |
Token | . |
Token ID | 13 |
Feature activation | +7.82e-02 |
Pos logit contributions | |
dot | +0.026 |
en | +0.025 |
fly | +0.025 |
el | +0.024 |
ir | +0.024 |
Neg logit contributions | |
cream | -0.030 |
topp | -0.029 |
math | -0.029 |
subt | -0.028 |
cakes | -0.028 |
Token | The |
Token ID | 383 |
Feature activation | -1.66e-01 |
Pos logit contributions | |
subt | +0.002 |
cream | +0.002 |
math | +0.002 |
Deal | +0.002 |
topp | +0.002 |
Neg logit contributions | |
dot | -0.002 |
en | -0.002 |
el | -0.002 |
fly | -0.002 |
ug | -0.002 |
Token | people |
Token ID | 661 |
Feature activation | +2.14e+00 |
Pos logit contributions | |
dot | +0.006 |
en | +0.006 |
el | +0.006 |
fly | +0.006 |
ug | +0.006 |
Neg logit contributions | |
subt | -0.004 |
math | -0.003 |
Deal | -0.003 |
cream | -0.003 |
subtract | -0.003 |
Token | who |
Token ID | 508 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
cream | +0.058 |
topp | +0.056 |
math | +0.054 |
Deal | +0.053 |
cakes | +0.052 |
Neg logit contributions | |
dot | -0.064 |
en | -0.063 |
el | -0.061 |
ug | -0.061 |
py | -0.061 |
Token | new |
Token ID | 649 |
Feature activation | -5.82e-01 |
Pos logit contributions | |
en | +0.051 |
dot | +0.050 |
el | +0.049 |
fly | +0.049 |
ug | +0.048 |
Neg logit contributions | |
subt | -0.037 |
math | -0.036 |
Deal | -0.036 |
cream | -0.036 |
topp | -0.036 |
Token | things |
Token ID | 1243 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
en | +0.020 |
dot | +0.020 |
el | +0.020 |
fly | +0.020 |
ug | +0.020 |
Neg logit contributions | |
subt | -0.013 |
math | -0.012 |
cream | -0.012 |
Deal | -0.012 |
topp | -0.012 |
Token | with |
Token ID | 351 |
Feature activation | -1.99e-01 |
Pos logit contributions | |
dot | +0.038 |
en | +0.037 |
el | +0.037 |
ug | +0.036 |
fly | +0.036 |
Neg logit contributions | |
math | -0.043 |
cream | -0.043 |
subt | -0.043 |
Deal | -0.042 |
topp | -0.042 |
Token | his |
Token ID | 465 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
dot | +0.007 |
ga | +0.006 |
fly | +0.006 |
ug | +0.006 |
ays | +0.006 |
Neg logit contributions | |
math | -0.005 |
cream | -0.005 |
topp | -0.005 |
pizza | -0.005 |
cakes | -0.005 |
Token | friend |
Token ID | 1545 |
Feature activation | -9.94e-01 |
Pos logit contributions | |
en | +0.063 |
le | +0.062 |
ug | +0.061 |
dot | +0.061 |
ga | +0.060 |
Neg logit contributions | |
cream | -0.040 |
math | -0.040 |
topp | -0.039 |
Deal | -0.038 |
subt | -0.038 |
Token | Ch |
Token ID | 609 |
Feature activation | -4.11e+00 |
Pos logit contributions | |
dot | +0.026 |
en | +0.026 |
el | +0.025 |
ir | +0.025 |
ug | +0.025 |
Neg logit contributions | |
math | -0.030 |
subt | -0.029 |
cream | -0.029 |
topp | -0.029 |
Deal | -0.029 |
Token | ir |
Token ID | 343 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
dot | +0.080 |
lands | +0.074 |
landed | +0.067 |
ley | +0.061 |
ening | +0.061 |
Neg logit contributions | |
subtract | -0.135 |
igsaw | -0.134 |
subt | -0.132 |
topp | -0.130 |
Deal | -0.129 |
Token | py |
Token ID | 9078 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
dot | +0.020 |
ays | +0.015 |
wil | +0.015 |
inkle | +0.015 |
ink | +0.014 |
Neg logit contributions | |
math | -0.067 |
cream | -0.066 |
topp | -0.066 |
subtract | -0.066 |
cakes | -0.064 |
Token | . |
Token ID | 13 |
Feature activation | -6.47e-01 |
Pos logit contributions | |
en | +0.027 |
dot | +0.027 |
el | +0.027 |
fly | +0.026 |
ug | +0.026 |
Neg logit contributions | |
subt | -0.034 |
math | -0.034 |
cream | -0.033 |
Deal | -0.033 |
topp | -0.033 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.39e-02 |
Pos logit contributions | |
en | +0.020 |
dot | +0.020 |
el | +0.019 |
ening | +0.019 |
le | +0.019 |
Neg logit contributions | |
subt | -0.016 |
math | -0.016 |
Deal | -0.016 |
cream | -0.016 |
topp | -0.016 |
Token | Once |
Token ID | 7454 |
Feature activation | -3.80e-01 |
Pos logit contributions | |
ening | +0.001 |
dot | +0.001 |
wil | +0.001 |
en | +0.001 |
fly | +0.001 |
Neg logit contributions | |
math | -0.000 |
cream | -0.000 |
subt | -0.000 |
topp | -0.000 |
las | -0.000 |
Token | deep |
Token ID | 2769 |
Feature activation | -3.37e-01 |
Pos logit contributions | |
subt | +0.009 |
math | +0.009 |
cream | +0.009 |
topp | +0.009 |
Deal | +0.009 |
Neg logit contributions | |
en | -0.012 |
dot | -0.012 |
el | -0.012 |
fly | -0.012 |
ug | -0.011 |
Token | paws |
Token ID | 49193 |
Feature activation | -3.97e-01 |
Pos logit contributions | |
fly | +0.009 |
ches | +0.009 |
dot | +0.009 |
ays | +0.009 |
ala | +0.009 |
Neg logit contributions | |
math | -0.010 |
subt | -0.010 |
job | -0.010 |
cream | -0.009 |
topp | -0.009 |
Token | , |
Token ID | 11 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
inkle | +0.012 |
dot | +0.012 |
en | +0.012 |
py | +0.012 |
fly | +0.012 |
Neg logit contributions | |
cream | -0.010 |
math | -0.010 |
subtract | -0.010 |
topp | -0.009 |
subt | -0.009 |
Token | to |
Token ID | 284 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
subt | +0.007 |
math | +0.007 |
cream | +0.007 |
topp | +0.007 |
subtract | +0.007 |
Neg logit contributions | |
en | -0.006 |
dot | -0.006 |
el | -0.005 |
fly | -0.005 |
ir | -0.005 |
Token | no |
Token ID | 645 |
Feature activation | +5.55e-02 |
Pos logit contributions | |
subt | +0.032 |
Deal | +0.032 |
topp | +0.032 |
cream | +0.031 |
las | +0.030 |
Neg logit contributions | |
en | -0.034 |
dot | -0.033 |
el | -0.033 |
fly | -0.032 |
le | -0.031 |
Token | avail |
Token ID | 29107 |
Feature activation | -3.91e+00 |
Pos logit contributions | |
math | +0.001 |
subt | +0.001 |
cream | +0.001 |
topp | +0.001 |
subtract | +0.001 |
Neg logit contributions | |
fly | -0.002 |
dot | -0.002 |
en | -0.002 |
ug | -0.002 |
el | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +2.19e-01 |
Pos logit contributions | |
dot | +0.129 |
wil | +0.124 |
inkle | +0.122 |
ga | +0.120 |
ches | +0.119 |
Neg logit contributions | |
math | -0.084 |
topp | -0.081 |
cream | -0.080 |
Deal | -0.078 |
job | -0.075 |
Token | He |
Token ID | 679 |
Feature activation | +6.07e-01 |
Pos logit contributions | |
subt | +0.006 |
cream | +0.006 |
math | +0.006 |
topp | +0.006 |
Deal | +0.006 |
Neg logit contributions | |
en | -0.006 |
dot | -0.006 |
el | -0.006 |
fly | -0.006 |
ug | -0.006 |
Token | finally |
Token ID | 3443 |
Feature activation | -7.78e-01 |
Pos logit contributions | |
subt | +0.016 |
Deal | +0.016 |
cream | +0.015 |
math | +0.015 |
topp | +0.015 |
Neg logit contributions | |
dot | -0.018 |
ir | -0.018 |
ays | -0.018 |
ug | -0.018 |
en | -0.017 |
Token | thought |
Token ID | 1807 |
Feature activation | +2.52e-01 |
Pos logit contributions | |
en | +0.020 |
dot | +0.019 |
el | +0.019 |
fly | +0.019 |
ug | +0.019 |
Neg logit contributions | |
subt | -0.024 |
math | -0.024 |
cream | -0.024 |
Deal | -0.024 |
topp | -0.024 |
Token | of |
Token ID | 286 |
Feature activation | +9.02e-01 |
Pos logit contributions | |
math | +0.007 |
cream | +0.006 |
topp | +0.006 |
pizza | +0.006 |
subt | +0.006 |
Neg logit contributions | |
fly | -0.009 |
dot | -0.009 |
en | -0.008 |
ley | -0.008 |
el | -0.008 |
Token | times |
Token ID | 1661 |
Feature activation | -6.43e-01 |
Pos logit contributions | |
math | +0.005 |
cream | +0.005 |
subt | +0.005 |
topp | +0.005 |
subtract | +0.005 |
Neg logit contributions | |
fly | -0.004 |
dot | -0.004 |
en | -0.004 |
el | -0.004 |
ley | -0.004 |
Token | , |
Token ID | 11 |
Feature activation | -5.63e-02 |
Pos logit contributions | |
dot | +0.018 |
fly | +0.018 |
py | +0.018 |
eds | +0.018 |
en | +0.018 |
Neg logit contributions | |
math | -0.019 |
subt | -0.018 |
subtract | -0.018 |
cream | -0.017 |
job | -0.016 |
Token | but |
Token ID | 475 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
fly | +0.001 |
dot | +0.001 |
en | +0.001 |
ug | +0.001 |
el | +0.001 |
Neg logit contributions | |
math | -0.002 |
subt | -0.002 |
cream | -0.002 |
topp | -0.002 |
subtract | -0.002 |
Token | to |
Token ID | 284 |
Feature activation | +1.41e+00 |
Pos logit contributions | |
en | +0.037 |
dot | +0.037 |
el | +0.036 |
fly | +0.036 |
ug | +0.035 |
Neg logit contributions | |
subt | -0.046 |
math | -0.046 |
cream | -0.045 |
Deal | -0.045 |
topp | -0.044 |
Token | no |
Token ID | 645 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
topp | +0.043 |
subt | +0.043 |
Deal | +0.041 |
cream | +0.041 |
las | +0.041 |
Neg logit contributions | |
en | -0.038 |
el | -0.036 |
dot | -0.036 |
fly | -0.035 |
le | -0.035 |
Token | avail |
Token ID | 29107 |
Feature activation | -3.81e+00 |
Pos logit contributions | |
math | +0.010 |
subt | +0.010 |
cream | +0.009 |
topp | +0.009 |
Deal | +0.009 |
Neg logit contributions | |
dot | -0.013 |
en | -0.013 |
fly | -0.013 |
ug | -0.013 |
el | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | +5.17e-01 |
Pos logit contributions | |
dot | +0.123 |
ga | +0.116 |
inkle | +0.116 |
hawk | +0.116 |
ug | +0.116 |
Neg logit contributions | |
math | -0.087 |
Deal | -0.086 |
cream | -0.080 |
topp | -0.079 |
job | -0.077 |
Token | |
Token ID | 220 |
Feature activation | +1.57e+00 |
Pos logit contributions | |
subt | +0.014 |
cream | +0.014 |
topp | +0.014 |
math | +0.014 |
Deal | +0.014 |
Neg logit contributions | |
en | -0.015 |
el | -0.015 |
dot | -0.015 |
fly | -0.014 |
ir | -0.014 |
Token |
|
Token ID | 628 |
Feature activation | +6.21e-01 |
Pos logit contributions | |
math | +0.040 |
subt | +0.039 |
Deal | +0.038 |
cream | +0.038 |
subtract | +0.037 |
Neg logit contributions | |
dot | -0.050 |
en | -0.050 |
fly | -0.048 |
el | -0.048 |
ening | -0.048 |
Token | Jane |
Token ID | 12091 |
Feature activation | +2.39e-01 |
Pos logit contributions | |
subt | +0.020 |
topp | +0.020 |
subtract | +0.019 |
cream | +0.019 |
Deal | +0.019 |
Neg logit contributions | |
dot | -0.014 |
wil | -0.014 |
el | -0.014 |
ir | -0.014 |
en | -0.014 |
Token | 's |
Token ID | 338 |
Feature activation | +9.86e-01 |
Pos logit contributions | |
subt | +0.005 |
cream | +0.004 |
math | +0.004 |
subtract | +0.004 |
Deal | +0.004 |
Neg logit contributions | |
ays | -0.008 |
dot | -0.008 |
py | -0.008 |
ug | -0.008 |
ir | -0.008 |
Token | again |
Token ID | 757 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
ening | +0.011 |
py | +0.011 |
dot | +0.011 |
ley | +0.011 |
wil | +0.011 |
Neg logit contributions | |
math | -0.011 |
cream | -0.010 |
Deal | -0.009 |
quiz | -0.009 |
homework | -0.009 |
Token | ?" |
Token ID | 1701 |
Feature activation | -7.39e-01 |
Pos logit contributions | |
dot | +0.028 |
py | +0.028 |
ening | +0.028 |
fly | +0.027 |
wil | +0.027 |
Neg logit contributions | |
math | -0.033 |
cream | -0.032 |
job | -0.030 |
dessert | -0.029 |
Deal | -0.029 |
Token |
|
Token ID | 198 |
Feature activation | -9.00e-01 |
Pos logit contributions | |
en | +0.026 |
dot | +0.026 |
ening | +0.025 |
ches | +0.025 |
le | +0.025 |
Neg logit contributions | |
math | -0.017 |
subt | -0.016 |
Deal | -0.016 |
cream | -0.015 |
subtract | -0.015 |
Token |
|
Token ID | 198 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
en | +0.035 |
ening | +0.035 |
dot | +0.034 |
fly | +0.033 |
wil | +0.032 |
Neg logit contributions | |
math | -0.017 |
cream | -0.017 |
subt | -0.016 |
Deal | -0.016 |
subtract | -0.015 |
Token | Spot |
Token ID | 32565 |
Feature activation | -1.56e+00 |
Pos logit contributions | |
en | +0.042 |
dot | +0.042 |
el | +0.042 |
fly | +0.042 |
ug | +0.041 |
Neg logit contributions | |
subt | -0.030 |
math | -0.029 |
cream | -0.029 |
topp | -0.029 |
Deal | -0.029 |
Token | n |
Token ID | 299 |
Feature activation | -3.67e+00 |
Pos logit contributions | |
dot | +0.051 |
ug | +0.051 |
ir | +0.050 |
el | +0.049 |
en | +0.049 |
Neg logit contributions | |
Deal | -0.037 |
cream | -0.037 |
subt | -0.037 |
math | -0.037 |
topp | -0.036 |
Token | uzz |
Token ID | 4715 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
dot | +0.101 |
ches | +0.099 |
ening | +0.098 |
en | +0.093 |
predators | +0.091 |
Neg logit contributions | |
Deal | -0.104 |
cream | -0.103 |
math | -0.099 |
topp | -0.098 |
Fif | -0.096 |
Token | led |
Token ID | 992 |
Feature activation | -4.06e-01 |
Pos logit contributions | |
dot | +0.030 |
ug | +0.029 |
predators | +0.028 |
ga | +0.028 |
ir | +0.027 |
Neg logit contributions | |
cream | -0.045 |
Deal | -0.042 |
topp | -0.042 |
math | -0.039 |
las | -0.038 |
Token | Tom |
Token ID | 4186 |
Feature activation | -4.98e-01 |
Pos logit contributions | |
dot | +0.011 |
ug | +0.011 |
fly | +0.011 |
wil | +0.011 |
en | +0.011 |
Neg logit contributions | |
cream | -0.012 |
math | -0.011 |
subt | -0.011 |
topp | -0.011 |
las | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | -9.59e-01 |
Pos logit contributions | |
dot | +0.013 |
ir | +0.013 |
ug | +0.013 |
inkle | +0.012 |
wil | +0.012 |
Neg logit contributions | |
cream | -0.015 |
subtract | -0.015 |
topp | -0.015 |
subt | -0.015 |
math | -0.014 |
Token | He |
Token ID | 679 |
Feature activation | +7.24e-02 |
Pos logit contributions | |
en | +0.028 |
dot | +0.028 |
el | +0.028 |
fly | +0.027 |
ug | +0.027 |
Neg logit contributions | |
subt | -0.026 |
math | -0.025 |
cream | -0.025 |
Deal | -0.025 |
topp | -0.025 |
Token | out |
Token ID | 503 |
Feature activation | +2.40e-01 |
Pos logit contributions | |
ink | +0.038 |
inkle | +0.038 |
wil | +0.037 |
fly | +0.037 |
dot | +0.036 |
Neg logit contributions | |
math | -0.025 |
cream | -0.024 |
job | -0.024 |
subtract | -0.020 |
ready | -0.020 |
Token | , |
Token ID | 11 |
Feature activation | +3.51e-02 |
Pos logit contributions | |
math | +0.006 |
subt | +0.006 |
subtract | +0.006 |
cream | +0.006 |
topp | +0.006 |
Neg logit contributions | |
ga | -0.007 |
dot | -0.007 |
ley | -0.007 |
inkle | -0.007 |
ays | -0.007 |
Token | but |
Token ID | 475 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
math | +0.001 |
cream | +0.001 |
subt | +0.001 |
subtract | +0.001 |
topp | +0.001 |
Neg logit contributions | |
el | -0.001 |
ug | -0.001 |
ir | -0.001 |
dot | -0.001 |
en | -0.001 |
Token | to |
Token ID | 284 |
Feature activation | +1.60e+00 |
Pos logit contributions | |
dot | +0.047 |
en | +0.047 |
el | +0.046 |
fly | +0.045 |
ir | +0.045 |
Neg logit contributions | |
math | -0.032 |
cream | -0.031 |
subt | -0.031 |
Deal | -0.030 |
topp | -0.029 |
Token | no |
Token ID | 645 |
Feature activation | +2.90e-01 |
Pos logit contributions | |
subt | +0.042 |
math | +0.041 |
cream | +0.041 |
Deal | +0.040 |
topp | +0.040 |
Neg logit contributions | |
en | -0.048 |
dot | -0.048 |
el | -0.047 |
fly | -0.046 |
ug | -0.046 |
Token | avail |
Token ID | 29107 |
Feature activation | -3.64e+00 |
Pos logit contributions | |
math | +0.007 |
cream | +0.007 |
subt | +0.007 |
topp | +0.007 |
subtract | +0.007 |
Neg logit contributions | |
fly | -0.009 |
dot | -0.009 |
ug | -0.009 |
ga | -0.009 |
el | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | +1.89e-01 |
Pos logit contributions | |
dot | +0.117 |
wil | +0.112 |
inkle | +0.111 |
ga | +0.110 |
ir | +0.108 |
Neg logit contributions | |
math | -0.082 |
topp | -0.079 |
Deal | -0.079 |
cream | -0.078 |
job | -0.073 |
Token | He |
Token ID | 679 |
Feature activation | +3.64e-01 |
Pos logit contributions | |
subt | +0.005 |
cream | +0.005 |
math | +0.005 |
topp | +0.005 |
Deal | +0.005 |
Neg logit contributions | |
en | -0.006 |
dot | -0.006 |
el | -0.006 |
fly | -0.005 |
ug | -0.005 |
Token | felt |
Token ID | 2936 |
Feature activation | -9.35e-01 |
Pos logit contributions | |
cream | +0.009 |
subt | +0.009 |
Deal | +0.008 |
topp | +0.008 |
y | +0.008 |
Neg logit contributions | |
ays | -0.012 |
dot | -0.011 |
ir | -0.011 |
inkle | -0.011 |
ors | -0.011 |
Token | sad |
Token ID | 6507 |
Feature activation | +5.46e-01 |
Pos logit contributions | |
fly | +0.033 |
ley | +0.032 |
dot | +0.032 |
el | +0.031 |
en | +0.031 |
Neg logit contributions | |
math | -0.021 |
cream | -0.020 |
y | -0.019 |
subt | -0.019 |
las | -0.018 |
Token | and |
Token ID | 290 |
Feature activation | +2.50e-01 |
Pos logit contributions | |
math | +0.017 |
cream | +0.015 |
subt | +0.015 |
job | +0.015 |
subtract | +0.014 |
Neg logit contributions | |
ir | -0.015 |
el | -0.014 |
ga | -0.014 |
fly | -0.014 |
inkle | -0.014 |
Token | had |
Token ID | 550 |
Feature activation | -4.23e-01 |
Pos logit contributions | |
subt | +0.005 |
cream | +0.005 |
math | +0.005 |
Deal | +0.005 |
topp | +0.005 |
Neg logit contributions | |
en | -0.005 |
dot | -0.005 |
el | -0.005 |
fly | -0.005 |
ug | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | +1.71e-01 |
Pos logit contributions | |
dot | +0.012 |
el | +0.012 |
wil | +0.012 |
en | +0.012 |
ga | +0.011 |
Neg logit contributions | |
subt | -0.012 |
Deal | -0.011 |
subtract | -0.011 |
topp | -0.011 |
cream | -0.011 |
Token | pet |
Token ID | 4273 |
Feature activation | -3.27e-01 |
Pos logit contributions | |
subt | +0.005 |
subtract | +0.005 |
topp | +0.005 |
Fif | +0.005 |
Deal | +0.005 |
Neg logit contributions | |
dot | -0.005 |
wil | -0.004 |
fly | -0.004 |
el | -0.004 |
inkle | -0.004 |
Token | rabbit |
Token ID | 22746 |
Feature activation | -7.57e-01 |
Pos logit contributions | |
en | +0.010 |
ening | +0.010 |
inkle | +0.010 |
ug | +0.009 |
ir | +0.009 |
Neg logit contributions | |
math | -0.009 |
cream | -0.009 |
Deal | -0.009 |
subt | -0.008 |
pizza | -0.008 |
Token | named |
Token ID | 3706 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
ening | +0.022 |
en | +0.022 |
inkle | +0.021 |
dot | +0.021 |
ays | +0.021 |
Neg logit contributions | |
math | -0.022 |
cream | -0.021 |
pizza | -0.021 |
Deal | -0.020 |
cakes | -0.019 |
Token | Snow |
Token ID | 7967 |
Feature activation | -3.56e+00 |
Pos logit contributions | |
ening | +0.040 |
dot | +0.040 |
ug | +0.040 |
en | +0.040 |
le | +0.040 |
Neg logit contributions | |
math | -0.029 |
subt | -0.026 |
pizza | -0.025 |
subtract | -0.025 |
cream | -0.025 |
Token | y |
Token ID | 88 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
lands | +0.066 |
dot | +0.064 |
predators | +0.064 |
ening | +0.063 |
ug | +0.063 |
Neg logit contributions | |
cream | -0.120 |
Deal | -0.114 |
topp | -0.107 |
pizza | -0.105 |
las | -0.105 |
Token | . |
Token ID | 13 |
Feature activation | -4.02e-01 |
Pos logit contributions | |
en | +0.038 |
ening | +0.037 |
ir | +0.037 |
ug | +0.037 |
dot | +0.037 |
Neg logit contributions | |
math | -0.027 |
Deal | -0.027 |
cream | -0.026 |
pizza | -0.025 |
job | -0.024 |
Token | Lily |
Token ID | 20037 |
Feature activation | -4.58e-01 |
Pos logit contributions | |
fly | +0.014 |
ening | +0.013 |
ug | +0.013 |
le | +0.013 |
ir | +0.013 |
Neg logit contributions | |
snack | -0.009 |
math | -0.009 |
Cere | -0.009 |
y | -0.009 |
cookies | -0.009 |
Token | loved |
Token ID | 6151 |
Feature activation | +7.57e-03 |
Pos logit contributions | |
ir | +0.022 |
ays | +0.022 |
ug | +0.021 |
dot | +0.021 |
predators | +0.020 |
Neg logit contributions | |
Deal | -0.006 |
y | -0.006 |
cream | -0.006 |
responsible | -0.004 |
creative | -0.004 |
Token | to |
Token ID | 284 |
Feature activation | -5.50e-02 |
Pos logit contributions | |
subt | +0.000 |
math | +0.000 |
cream | +0.000 |
topp | +0.000 |
Deal | +0.000 |
Neg logit contributions | |
en | -0.000 |
dot | -0.000 |
el | -0.000 |
fly | -0.000 |
ug | -0.000 |
Token | ens |
Token ID | 641 |
Feature activation | -2.03e+00 |
Pos logit contributions | |
dot | +0.056 |
en | +0.056 |
el | +0.055 |
fly | +0.055 |
ug | +0.053 |
Neg logit contributions | |
subt | -0.059 |
Deal | -0.059 |
cream | -0.059 |
math | -0.059 |
topp | -0.057 |
Token | and |
Token ID | 290 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
dot | +0.069 |
inkle | +0.066 |
en | +0.065 |
el | +0.064 |
ening | +0.064 |
Neg logit contributions | |
math | -0.046 |
pizza | -0.044 |
Deal | -0.043 |
cream | -0.043 |
cakes | -0.042 |
Token | a |
Token ID | 257 |
Feature activation | -3.66e-01 |
Pos logit contributions | |
fly | +0.003 |
dot | +0.003 |
en | +0.003 |
py | +0.003 |
pt | +0.003 |
Neg logit contributions | |
subt | -0.004 |
topp | -0.004 |
math | -0.004 |
subtract | -0.004 |
Deal | -0.004 |
Token | rabbit |
Token ID | 22746 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
dot | +0.009 |
fly | +0.009 |
wil | +0.008 |
en | +0.008 |
el | +0.008 |
Neg logit contributions | |
subt | -0.012 |
Deal | -0.011 |
topp | -0.011 |
subtract | -0.011 |
math | -0.011 |
Token | named |
Token ID | 3706 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
en | +0.044 |
ening | +0.043 |
dot | +0.043 |
inkle | +0.043 |
el | +0.042 |
Neg logit contributions | |
math | -0.044 |
Deal | -0.043 |
cream | -0.042 |
subt | -0.041 |
topp | -0.041 |
Token | Ho |
Token ID | 9544 |
Feature activation | -3.55e+00 |
Pos logit contributions | |
en | +0.036 |
dot | +0.036 |
el | +0.035 |
fly | +0.035 |
ug | +0.034 |
Neg logit contributions | |
subt | -0.034 |
math | -0.034 |
cream | -0.034 |
topp | -0.033 |
Deal | -0.033 |
Token | ppy |
Token ID | 14097 |
Feature activation | -2.14e+00 |
Pos logit contributions | |
ening | +0.065 |
dot | +0.065 |
wil | +0.062 |
en | +0.061 |
el | +0.060 |
Neg logit contributions | |
Deal | -0.135 |
topp | -0.124 |
cream | -0.123 |
math | -0.122 |
subt | -0.121 |
Token | , |
Token ID | 11 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
dot | +0.069 |
en | +0.067 |
inkle | +0.067 |
el | +0.067 |
ening | +0.066 |
Neg logit contributions | |
math | -0.050 |
Deal | -0.047 |
cream | -0.047 |
subt | -0.047 |
pizza | -0.046 |
Token | getting |
Token ID | 1972 |
Feature activation | +2.13e-01 |
Pos logit contributions | |
inkle | +0.047 |
dot | +0.045 |
ir | +0.045 |
ug | +0.045 |
le | +0.045 |
Neg logit contributions | |
y | -0.027 |
hamb | -0.026 |
cream | -0.026 |
pizza | -0.025 |
math | -0.024 |
Token | ready |
Token ID | 3492 |
Feature activation | -2.40e-01 |
Pos logit contributions | |
subt | +0.006 |
math | +0.006 |
Deal | +0.006 |
topp | +0.006 |
cream | +0.006 |
Neg logit contributions | |
en | -0.006 |
el | -0.006 |
dot | -0.006 |
fly | -0.006 |
ug | -0.006 |
Token | for |
Token ID | 329 |
Feature activation | +1.93e-02 |
Pos logit contributions | |
dot | +0.008 |
en | +0.008 |
fly | +0.008 |
el | +0.008 |
ug | +0.008 |
Neg logit contributions | |
math | -0.005 |
cream | -0.005 |
subt | -0.005 |
topp | -0.005 |
las | -0.005 |
Token | met |
Token ID | 1138 |
Feature activation | +3.95e-01 |
Pos logit contributions | |
ays | +0.013 |
ir | +0.013 |
inkle | +0.013 |
dot | +0.013 |
ug | +0.013 |
Neg logit contributions | |
subt | -0.007 |
cream | -0.006 |
math | -0.006 |
subtract | -0.006 |
Deal | -0.006 |
Token | a |
Token ID | 257 |
Feature activation | -6.06e-01 |
Pos logit contributions | |
math | +0.007 |
cream | +0.007 |
Deal | +0.007 |
subt | +0.006 |
topp | +0.006 |
Neg logit contributions | |
dot | -0.016 |
en | -0.015 |
ga | -0.015 |
ening | -0.015 |
ug | -0.015 |
Token | kind |
Token ID | 1611 |
Feature activation | -7.48e-01 |
Pos logit contributions | |
en | +0.019 |
dot | +0.019 |
el | +0.019 |
fly | +0.019 |
ug | +0.019 |
Neg logit contributions | |
subt | -0.015 |
math | -0.014 |
cream | -0.014 |
Deal | -0.014 |
topp | -0.014 |
Token | bird |
Token ID | 6512 |
Feature activation | -4.86e-01 |
Pos logit contributions | |
ays | +0.023 |
ches | +0.023 |
unks | +0.023 |
ening | +0.022 |
en | +0.022 |
Neg logit contributions | |
cream | -0.019 |
math | -0.019 |
pizza | -0.018 |
customer | -0.018 |
job | -0.017 |
Token | named |
Token ID | 3706 |
Feature activation | -5.37e-01 |
Pos logit contributions | |
dot | +0.013 |
en | +0.013 |
ga | +0.013 |
ening | +0.013 |
inkle | +0.013 |
Neg logit contributions | |
math | -0.013 |
cream | -0.013 |
Deal | -0.013 |
pizza | -0.012 |
topp | -0.012 |
Token | Ch |
Token ID | 609 |
Feature activation | -3.53e+00 |
Pos logit contributions | |
en | +0.012 |
dot | +0.012 |
el | +0.012 |
ug | +0.012 |
le | +0.012 |
Neg logit contributions | |
math | -0.018 |
subt | -0.018 |
cream | -0.018 |
topp | -0.017 |
Deal | -0.017 |
Token | ir |
Token ID | 343 |
Feature activation | -6.38e-01 |
Pos logit contributions | |
dot | +0.059 |
landed | +0.057 |
prope | +0.056 |
lands | +0.055 |
ley | +0.055 |
Neg logit contributions | |
math | -0.114 |
igsaw | -0.113 |
Fif | -0.107 |
topp | -0.106 |
las | -0.106 |
Token | py |
Token ID | 9078 |
Feature activation | -8.46e-01 |
Pos logit contributions | |
dot | +0.007 |
inkle | +0.007 |
ley | +0.007 |
wil | +0.006 |
ga | +0.006 |
Neg logit contributions | |
math | -0.027 |
cream | -0.026 |
subtract | -0.026 |
cakes | -0.026 |
topp | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
inkle | +0.027 |
dot | +0.025 |
ga | +0.025 |
ink | +0.025 |
en | +0.025 |
Neg logit contributions | |
math | -0.020 |
Deal | -0.019 |
cream | -0.019 |
topp | -0.019 |
subt | -0.019 |
Token | Ch |
Token ID | 609 |
Feature activation | -2.97e+00 |
Pos logit contributions | |
en | +0.025 |
el | +0.024 |
le | +0.024 |
wil | +0.024 |
dot | +0.024 |
Neg logit contributions | |
subt | -0.033 |
math | -0.032 |
Deal | -0.031 |
subtract | -0.031 |
topp | -0.031 |
Token | ir |
Token ID | 343 |
Feature activation | -7.92e-01 |
Pos logit contributions | |
dot | +0.054 |
hole | +0.052 |
lands | +0.051 |
landed | +0.050 |
ley | +0.050 |
Neg logit contributions | |
igsaw | -0.098 |
math | -0.093 |
topp | -0.088 |
dessert | -0.087 |
Fif | -0.086 |
Token | him |
Token ID | 683 |
Feature activation | +2.93e-02 |
Pos logit contributions | |
dot | +0.011 |
fly | +0.011 |
en | +0.011 |
ug | +0.011 |
inkle | +0.011 |
Neg logit contributions | |
math | -0.013 |
cream | -0.012 |
subt | -0.012 |
topp | -0.012 |
subtract | -0.012 |
Token | a |
Token ID | 257 |
Feature activation | -5.80e-01 |
Pos logit contributions | |
subt | +0.001 |
subtract | +0.001 |
Deal | +0.001 |
topp | +0.001 |
bek | +0.001 |
Neg logit contributions | |
el | -0.001 |
ches | -0.001 |
en | -0.001 |
pt | -0.001 |
ley | -0.001 |
Token | lick |
Token ID | 36728 |
Feature activation | +1.63e-01 |
Pos logit contributions | |
dot | +0.015 |
fly | +0.014 |
wil | +0.014 |
pt | +0.014 |
ches | +0.014 |
Neg logit contributions | |
subt | -0.019 |
Deal | -0.018 |
subtract | -0.017 |
topp | -0.017 |
Fif | -0.016 |
Token | and |
Token ID | 290 |
Feature activation | -7.27e-01 |
Pos logit contributions | |
cream | +0.006 |
math | +0.005 |
topp | +0.005 |
pizza | +0.005 |
las | +0.005 |
Neg logit contributions | |
inkle | -0.004 |
en | -0.004 |
fly | -0.004 |
py | -0.004 |
dot | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | -2.71e-01 |
Pos logit contributions | |
dot | +0.020 |
en | +0.019 |
fly | +0.019 |
wil | +0.019 |
el | +0.019 |
Neg logit contributions | |
Deal | -0.021 |
subt | -0.020 |
math | -0.020 |
Fif | -0.020 |
subtract | -0.019 |
Token | n |
Token ID | 299 |
Feature activation | -3.52e+00 |
Pos logit contributions | |
dot | +0.008 |
fly | +0.007 |
wing | +0.007 |
wil | +0.007 |
inkle | +0.006 |
Neg logit contributions | |
subt | -0.009 |
Deal | -0.008 |
subtract | -0.008 |
Fif | -0.007 |
topp | -0.007 |
Token | uzzle |
Token ID | 9625 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
ening | +0.090 |
ches | +0.088 |
dot | +0.086 |
en | +0.086 |
predators | +0.082 |
Neg logit contributions | |
cream | -0.100 |
Deal | -0.099 |
topp | -0.097 |
math | -0.095 |
bek | -0.093 |
Token | . |
Token ID | 13 |
Feature activation | -3.56e-01 |
Pos logit contributions | |
dot | +0.044 |
inkle | +0.043 |
unks | +0.041 |
en | +0.041 |
predators | +0.041 |
Neg logit contributions | |
cream | -0.036 |
math | -0.032 |
cakes | -0.031 |
topp | -0.031 |
pizza | -0.031 |
Token | She |
Token ID | 1375 |
Feature activation | -7.20e-03 |
Pos logit contributions | |
en | +0.010 |
le | +0.010 |
fly | +0.010 |
el | +0.010 |
dot | +0.010 |
Neg logit contributions | |
math | -0.010 |
subt | -0.010 |
subtract | -0.010 |
Deal | -0.009 |
cream | -0.009 |
Token | learned |
Token ID | 4499 |
Feature activation | -9.53e-01 |
Pos logit contributions | |
dot | +0.000 |
ening | +0.000 |
ug | +0.000 |
en | +0.000 |
ir | +0.000 |
Neg logit contributions | |
cream | -0.000 |
Deal | -0.000 |
subt | -0.000 |
topp | -0.000 |
math | -0.000 |
Token | her |
Token ID | 607 |
Feature activation | +6.51e-01 |
Pos logit contributions | |
en | +0.022 |
el | +0.021 |
dot | +0.020 |
pt | +0.020 |
fly | +0.020 |
Neg logit contributions | |
subt | -0.032 |
Deal | -0.031 |
topp | -0.031 |
subtract | -0.031 |
Fif | -0.030 |
Token | to |
Token ID | 284 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
cream | +0.019 |
math | +0.019 |
topp | +0.018 |
subt | +0.018 |
Deal | +0.018 |
Neg logit contributions | |
ug | -0.017 |
dot | -0.016 |
el | -0.016 |
ley | -0.016 |
ir | -0.016 |
Token | Ben |
Token ID | 3932 |
Feature activation | -1.13e+00 |
Pos logit contributions | |
ug | +0.006 |
dot | +0.005 |
ening | +0.005 |
en | +0.005 |
el | +0.005 |
Neg logit contributions | |
math | -0.007 |
cream | -0.007 |
subt | -0.007 |
las | -0.007 |
topp | -0.007 |
Token | and |
Token ID | 290 |
Feature activation | -8.36e-01 |
Pos logit contributions | |
ug | +0.028 |
dot | +0.027 |
wil | +0.027 |
ir | +0.027 |
py | +0.026 |
Neg logit contributions | |
cream | -0.034 |
math | -0.034 |
pizza | -0.033 |
topp | -0.033 |
list | -0.031 |
Token | Lily |
Token ID | 20037 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
fly | +0.022 |
dot | +0.022 |
en | +0.022 |
ches | +0.021 |
wil | +0.021 |
Neg logit contributions | |
subt | -0.025 |
Deal | -0.025 |
math | -0.025 |
topp | -0.024 |
subtract | -0.024 |
Token | and |
Token ID | 290 |
Feature activation | +1.59e-01 |
Pos logit contributions | |
ug | +0.004 |
dot | +0.004 |
ir | +0.003 |
el | +0.003 |
py | +0.003 |
Neg logit contributions | |
cream | -0.003 |
topp | -0.003 |
math | -0.003 |
subtract | -0.003 |
subt | -0.003 |
Token | n |
Token ID | 299 |
Feature activation | -3.50e+00 |
Pos logit contributions | |
subt | +0.005 |
Deal | +0.005 |
math | +0.005 |
subtract | +0.005 |
topp | +0.005 |
Neg logit contributions | |
fly | -0.004 |
en | -0.004 |
landed | -0.004 |
dot | -0.004 |
wil | -0.004 |
Token | uzz |
Token ID | 4715 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
ches | +0.088 |
dot | +0.088 |
ening | +0.084 |
en | +0.080 |
hawk | +0.079 |
Neg logit contributions | |
Deal | -0.103 |
cream | -0.103 |
topp | -0.101 |
bek | -0.098 |
Fif | -0.097 |
Token | led |
Token ID | 992 |
Feature activation | -8.81e-01 |
Pos logit contributions | |
dot | +0.031 |
ug | +0.029 |
py | +0.028 |
inkle | +0.028 |
ga | +0.028 |
Neg logit contributions | |
cream | -0.049 |
topp | -0.047 |
Deal | -0.044 |
cakes | -0.042 |
pizza | -0.042 |
Token | their |
Token ID | 511 |
Feature activation | +3.36e-01 |
Pos logit contributions | |
dot | +0.023 |
en | +0.023 |
ug | +0.023 |
fly | +0.022 |
el | +0.022 |
Neg logit contributions | |
cream | -0.026 |
math | -0.026 |
subt | -0.026 |
topp | -0.025 |
Deal | -0.025 |
Token | hands |
Token ID | 2832 |
Feature activation | -9.41e-01 |
Pos logit contributions | |
subt | +0.011 |
math | +0.011 |
cream | +0.011 |
topp | +0.011 |
Deal | +0.011 |
Neg logit contributions | |
en | -0.008 |
dot | -0.008 |
el | -0.008 |
fly | -0.008 |
ug | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
en | +0.027 |
dot | +0.027 |
ug | +0.027 |
fly | +0.026 |
py | +0.026 |
Neg logit contributions | |
cream | -0.025 |
topp | -0.025 |
math | -0.025 |
subt | -0.024 |
Deal | -0.023 |
Token | yt |
Token ID | 20760 |
Feature activation | +5.95e-01 |
Pos logit contributions | |
el | +0.031 |
ches | +0.030 |
en | +0.030 |
ug | +0.030 |
ir | +0.030 |
Neg logit contributions | |
cream | -0.032 |
math | -0.031 |
topp | -0.030 |
subt | -0.029 |
job | -0.029 |
Token | ale |
Token ID | 1000 |
Feature activation | -9.68e-01 |
Pos logit contributions | |
math | +0.020 |
cream | +0.020 |
topp | +0.020 |
subt | +0.019 |
Deal | +0.019 |
Neg logit contributions | |
dot | -0.012 |
wil | -0.012 |
fly | -0.012 |
en | -0.011 |
ink | -0.011 |
Token | creature |
Token ID | 7185 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
en | +0.024 |
dot | +0.024 |
el | +0.024 |
ug | +0.023 |
ir | +0.023 |
Neg logit contributions | |
cream | -0.029 |
math | -0.029 |
topp | -0.029 |
subt | -0.028 |
subtract | -0.027 |
Token | called |
Token ID | 1444 |
Feature activation | -9.08e-01 |
Pos logit contributions | |
ug | +0.032 |
en | +0.032 |
ir | +0.031 |
dot | +0.031 |
el | +0.031 |
Neg logit contributions | |
math | -0.028 |
cream | -0.028 |
topp | -0.027 |
pizza | -0.026 |
subt | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
en | +0.027 |
dot | +0.027 |
el | +0.026 |
le | +0.026 |
wil | +0.026 |
Neg logit contributions | |
math | -0.024 |
subt | -0.023 |
cream | -0.023 |
topp | -0.023 |
las | -0.022 |
Token | n |
Token ID | 299 |
Feature activation | -3.49e+00 |
Pos logit contributions | |
en | +0.035 |
el | +0.034 |
ug | +0.033 |
dot | +0.033 |
le | +0.033 |
Neg logit contributions | |
math | -0.034 |
cream | -0.033 |
subt | -0.033 |
topp | -0.033 |
las | -0.032 |
Token | ymph |
Token ID | 20896 |
Feature activation | -1.74e+00 |
Pos logit contributions | |
ening | +0.086 |
ches | +0.083 |
landed | +0.081 |
wil | +0.079 |
dot | +0.077 |
Neg logit contributions | |
bek | -0.096 |
Fif | -0.091 |
las | -0.090 |
subt | -0.087 |
Deal | -0.087 |
Token | ! |
Token ID | 0 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
el | +0.048 |
ug | +0.047 |
dot | +0.047 |
en | +0.047 |
ays | +0.047 |
Neg logit contributions | |
math | -0.045 |
cream | -0.045 |
topp | -0.045 |
las | -0.044 |
subt | -0.044 |
Token |
|
Token ID | 198 |
Feature activation | +2.61e-01 |
Pos logit contributions | |
en | +0.005 |
dot | +0.005 |
el | +0.005 |
fly | +0.005 |
le | +0.005 |
Neg logit contributions | |
subt | -0.005 |
math | -0.005 |
cream | -0.005 |
topp | -0.005 |
Deal | -0.005 |
Token |
|
Token ID | 198 |
Feature activation | -5.22e-02 |
Pos logit contributions | |
subt | +0.005 |
math | +0.005 |
cream | +0.004 |
subtract | +0.004 |
Fif | +0.004 |
Neg logit contributions | |
en | -0.010 |
ening | -0.010 |
dot | -0.010 |
fly | -0.010 |
wil | -0.010 |
Token | The |
Token ID | 464 |
Feature activation | +4.52e-01 |
Pos logit contributions | |
en | +0.002 |
dot | +0.002 |
el | +0.002 |
fly | +0.002 |
ug | +0.002 |
Neg logit contributions | |
subt | -0.001 |
math | -0.001 |
cream | -0.001 |
topp | -0.001 |
Deal | -0.001 |
Token | , |
Token ID | 11 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
dot | +0.014 |
wil | +0.013 |
inkle | +0.013 |
ening | +0.013 |
ir | +0.013 |
Neg logit contributions | |
math | -0.013 |
subt | -0.013 |
cream | -0.013 |
pizza | -0.013 |
las | -0.012 |
Token | a |
Token ID | 257 |
Feature activation | +2.23e-01 |
Pos logit contributions | |
en | +0.044 |
dot | +0.044 |
el | +0.043 |
fly | +0.043 |
ug | +0.042 |
Neg logit contributions | |
subt | -0.031 |
math | -0.030 |
cream | -0.030 |
Deal | -0.030 |
topp | -0.030 |
Token | little |
Token ID | 1310 |
Feature activation | -6.83e-01 |
Pos logit contributions | |
subt | +0.006 |
Deal | +0.006 |
topp | +0.006 |
subtract | +0.006 |
math | +0.006 |
Neg logit contributions | |
dot | -0.007 |
fly | -0.006 |
ug | -0.006 |
hawk | -0.006 |
el | -0.006 |
Token | bird |
Token ID | 6512 |
Feature activation | -8.59e-01 |
Pos logit contributions | |
en | +0.017 |
dot | +0.016 |
el | +0.016 |
fly | +0.016 |
ug | +0.016 |
Neg logit contributions | |
subt | -0.022 |
math | -0.022 |
cream | -0.021 |
topp | -0.021 |
Deal | -0.021 |
Token | named |
Token ID | 3706 |
Feature activation | -6.56e-01 |
Pos logit contributions | |
dot | +0.021 |
en | +0.021 |
ening | +0.021 |
ga | +0.020 |
ays | +0.020 |
Neg logit contributions | |
cream | -0.025 |
math | -0.024 |
Deal | -0.024 |
topp | -0.023 |
subt | -0.023 |
Token | Ch |
Token ID | 609 |
Feature activation | -3.48e+00 |
Pos logit contributions | |
en | +0.016 |
dot | +0.016 |
el | +0.016 |
fly | +0.015 |
ug | +0.015 |
Neg logit contributions | |
subt | -0.021 |
math | -0.021 |
cream | -0.020 |
Deal | -0.020 |
topp | -0.020 |
Token | ir |
Token ID | 343 |
Feature activation | -7.32e-01 |
Pos logit contributions | |
wil | +0.059 |
dot | +0.053 |
ening | +0.052 |
lands | +0.052 |
landed | +0.047 |
Neg logit contributions | |
math | -0.116 |
igsaw | -0.111 |
Fif | -0.111 |
Deal | -0.111 |
subtract | -0.111 |
Token | py |
Token ID | 9078 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
dot | +0.006 |
wil | +0.006 |
inkle | +0.006 |
ley | +0.005 |
en | +0.005 |
Neg logit contributions | |
math | -0.033 |
cream | -0.033 |
subtract | -0.032 |
subt | -0.032 |
topp | -0.031 |
Token | , |
Token ID | 11 |
Feature activation | -7.11e-01 |
Pos logit contributions | |
inkle | +0.036 |
dot | +0.036 |
ays | +0.035 |
ga | +0.035 |
wil | +0.034 |
Neg logit contributions | |
math | -0.032 |
subt | -0.031 |
cream | -0.030 |
las | -0.029 |
topp | -0.029 |
Token | flew |
Token ID | 13112 |
Feature activation | -8.48e-01 |
Pos logit contributions | |
inkle | +0.027 |
dot | +0.025 |
le | +0.025 |
en | +0.024 |
ir | +0.024 |
Neg logit contributions | |
cream | -0.014 |
hamb | -0.014 |
subt | -0.014 |
las | -0.014 |
y | -0.014 |
Token | by |
Token ID | 416 |
Feature activation | +5.94e-01 |
Pos logit contributions | |
dot | +0.025 |
en | +0.025 |
inkle | +0.024 |
el | +0.024 |
fly | +0.024 |
Neg logit contributions | |
cream | -0.023 |
math | -0.022 |
subt | -0.022 |
topp | -0.021 |
Deal | -0.021 |
Token | him |
Token ID | 683 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
math | +0.007 |
topp | +0.007 |
cream | +0.006 |
subt | +0.006 |
cakes | +0.006 |
Neg logit contributions | |
dot | -0.011 |
ga | -0.011 |
ley | -0.011 |
le | -0.011 |
fly | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +3.27e-01 |
Pos logit contributions | |
dot | +0.008 |
en | +0.008 |
ga | +0.007 |
ug | +0.007 |
el | +0.007 |
Neg logit contributions | |
math | -0.007 |
cream | -0.006 |
topp | -0.006 |
subt | -0.006 |
Deal | -0.006 |
Token | But |
Token ID | 887 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
subt | +0.009 |
math | +0.008 |
cream | +0.008 |
topp | +0.008 |
Deal | +0.008 |
Neg logit contributions | |
en | -0.010 |
dot | -0.010 |
el | -0.010 |
fly | -0.010 |
ug | -0.009 |
Token | to |
Token ID | 284 |
Feature activation | +1.92e+00 |
Pos logit contributions | |
dot | +0.029 |
en | +0.029 |
el | +0.028 |
fly | +0.028 |
ir | +0.028 |
Neg logit contributions | |
math | -0.028 |
subt | -0.028 |
cream | -0.027 |
topp | -0.027 |
Deal | -0.027 |
Token | no |
Token ID | 645 |
Feature activation | +4.43e-01 |
Pos logit contributions | |
topp | +0.055 |
subt | +0.055 |
Deal | +0.055 |
cream | +0.054 |
las | +0.051 |
Neg logit contributions | |
en | -0.054 |
el | -0.052 |
fly | -0.050 |
ir | -0.050 |
dot | -0.050 |
Token | avail |
Token ID | 29107 |
Feature activation | -3.48e+00 |
Pos logit contributions | |
math | +0.010 |
subt | +0.010 |
cream | +0.010 |
topp | +0.010 |
Deal | +0.009 |
Neg logit contributions | |
dot | -0.015 |
en | -0.015 |
fly | -0.015 |
el | -0.015 |
ug | -0.015 |
Token | , |
Token ID | 11 |
Feature activation | -1.97e-01 |
Pos logit contributions | |
dot | +0.109 |
wil | +0.105 |
horizon | +0.103 |
ga | +0.103 |
hawk | +0.102 |
Neg logit contributions | |
math | -0.079 |
topp | -0.078 |
Deal | -0.077 |
cream | -0.073 |
job | -0.072 |
Token | they |
Token ID | 484 |
Feature activation | +1.75e-01 |
Pos logit contributions | |
dot | +0.006 |
en | +0.006 |
el | +0.006 |
fly | +0.006 |
ir | +0.006 |
Neg logit contributions | |
math | -0.005 |
subt | -0.005 |
cream | -0.005 |
topp | -0.005 |
Deal | -0.005 |
Token | were |
Token ID | 547 |
Feature activation | +4.01e-01 |
Pos logit contributions | |
Deal | +0.004 |
cream | +0.004 |
y | +0.004 |
math | +0.004 |
topp | +0.004 |
Neg logit contributions | |
dot | -0.005 |
ays | -0.005 |
ug | -0.005 |
ga | -0.005 |
en | -0.005 |
Token | too |
Token ID | 1165 |
Feature activation | +4.17e-01 |
Pos logit contributions | |
math | +0.011 |
cream | +0.010 |
subt | +0.010 |
topp | +0.009 |
pizza | +0.009 |
Neg logit contributions | |
dot | -0.013 |
el | -0.012 |
ley | -0.012 |
en | -0.012 |
ug | -0.012 |
Token | scared |
Token ID | 12008 |
Feature activation | +6.64e-01 |
Pos logit contributions | |
cream | +0.010 |
math | +0.009 |
subt | +0.009 |
topp | +0.009 |
job | +0.009 |
Neg logit contributions | |
en | -0.014 |
dot | -0.014 |
fly | -0.014 |
el | -0.013 |
py | -0.013 |
Token | Where |
Token ID | 6350 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
subt | +0.019 |
math | +0.019 |
topp | +0.019 |
cream | +0.019 |
subtract | +0.018 |
Neg logit contributions | |
en | -0.023 |
dot | -0.022 |
el | -0.022 |
ug | -0.022 |
fly | -0.022 |
Token | are |
Token ID | 389 |
Feature activation | +1.45e+00 |
Pos logit contributions | |
en | +0.026 |
dot | +0.026 |
el | +0.026 |
fly | +0.025 |
ug | +0.025 |
Neg logit contributions | |
subt | -0.036 |
math | -0.035 |
cream | -0.035 |
Deal | -0.035 |
topp | -0.035 |
Token | you |
Token ID | 345 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
Deal | +0.057 |
subt | +0.056 |
subtract | +0.053 |
bek | +0.053 |
Fif | +0.051 |
Neg logit contributions | |
el | -0.029 |
us | -0.024 |
ie | -0.024 |
pt | -0.023 |
en | -0.023 |
Token | ?" |
Token ID | 1701 |
Feature activation | -9.31e-01 |
Pos logit contributions | |
dot | +0.004 |
ug | +0.004 |
fly | +0.003 |
ley | +0.003 |
ga | +0.003 |
Neg logit contributions | |
math | -0.003 |
cream | -0.003 |
las | -0.003 |
subt | -0.003 |
topp | -0.003 |
Token | their |
Token ID | 511 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
en | +0.035 |
dot | +0.035 |
ug | +0.034 |
fly | +0.034 |
ening | +0.034 |
Neg logit contributions | |
math | -0.019 |
subt | -0.018 |
cream | -0.017 |
subtract | -0.016 |
las | -0.016 |
Token | n |
Token ID | 299 |
Feature activation | -3.47e+00 |
Pos logit contributions | |
math | +0.027 |
cream | +0.026 |
subt | +0.025 |
cakes | +0.023 |
topp | +0.023 |
Neg logit contributions | |
en | -0.055 |
ug | -0.053 |
le | -0.053 |
ches | -0.053 |
ening | -0.053 |
Token | anny |
Token ID | 7737 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
ening | +0.104 |
dot | +0.104 |
ches | +0.103 |
hawk | +0.099 |
en | +0.098 |
Neg logit contributions | |
subt | -0.081 |
math | -0.080 |
cream | -0.080 |
topp | -0.077 |
Fif | -0.077 |
Token | shouted |
Token ID | 17293 |
Feature activation | -3.87e-01 |
Pos logit contributions | |
dot | +0.036 |
ug | +0.036 |
ga | +0.035 |
ir | +0.035 |
predators | +0.034 |
Neg logit contributions | |
subt | -0.025 |
subtract | -0.024 |
math | -0.024 |
cream | -0.024 |
topp | -0.023 |
Token | . |
Token ID | 13 |
Feature activation | -3.95e-01 |
Pos logit contributions | |
en | +0.010 |
dot | +0.010 |
fly | +0.010 |
el | +0.010 |
ug | +0.010 |
Neg logit contributions | |
subt | -0.012 |
math | -0.011 |
cream | -0.011 |
subtract | -0.011 |
topp | -0.011 |
Token | She |
Token ID | 1375 |
Feature activation | -7.62e-01 |
Pos logit contributions | |
en | +0.015 |
le | +0.015 |
ening | +0.014 |
ug | +0.014 |
el | +0.014 |
Neg logit contributions | |
math | -0.008 |
subt | -0.008 |
subtract | -0.007 |
las | -0.007 |
cream | -0.007 |
Token | saw |
Token ID | 2497 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
ug | +0.024 |
dot | +0.023 |
ir | +0.023 |
en | +0.023 |
el | +0.023 |
Neg logit contributions | |
cream | -0.020 |
subt | -0.020 |
topp | -0.019 |
math | -0.019 |
las | -0.018 |
Token | clothes |
Token ID | 8242 |
Feature activation | +7.97e-01 |
Pos logit contributions | |
subt | +0.023 |
math | +0.023 |
cream | +0.023 |
Deal | +0.023 |
topp | +0.022 |
Neg logit contributions | |
dot | -0.019 |
fly | -0.019 |
en | -0.019 |
el | -0.019 |
ug | -0.019 |
Token | , |
Token ID | 11 |
Feature activation | -6.08e-01 |
Pos logit contributions | |
cream | +0.027 |
math | +0.026 |
subt | +0.026 |
Deal | +0.025 |
subtract | +0.025 |
Neg logit contributions | |
fly | -0.019 |
dot | -0.018 |
ug | -0.018 |
py | -0.018 |
ir | -0.018 |
Token | like |
Token ID | 588 |
Feature activation | +9.12e-01 |
Pos logit contributions | |
en | +0.018 |
dot | +0.018 |
fly | +0.017 |
ug | +0.017 |
el | +0.017 |
Neg logit contributions | |
math | -0.017 |
subt | -0.017 |
cream | -0.016 |
topp | -0.016 |
subtract | -0.016 |
Token | her |
Token ID | 607 |
Feature activation | -6.33e-02 |
Pos logit contributions | |
math | +0.023 |
subt | +0.022 |
cream | +0.022 |
subtract | +0.021 |
cakes | +0.021 |
Neg logit contributions | |
fly | -0.030 |
dot | -0.030 |
inkle | -0.030 |
en | -0.029 |
ley | -0.029 |
Token | sister |
Token ID | 6621 |
Feature activation | -5.62e-01 |
Pos logit contributions | |
fly | +0.002 |
ug | +0.002 |
en | +0.002 |
ches | +0.002 |
ir | +0.002 |
Neg logit contributions | |
math | -0.001 |
cream | -0.001 |
subt | -0.001 |
job | -0.001 |
cakes | -0.001 |
Token | L |
Token ID | 406 |
Feature activation | -3.47e+00 |
Pos logit contributions | |
ir | +0.016 |
ug | +0.016 |
dot | +0.016 |
fly | +0.016 |
hawk | +0.016 |
Neg logit contributions | |
math | -0.016 |
subt | -0.015 |
cream | -0.015 |
cakes | -0.015 |
subtract | -0.014 |
Token | ila |
Token ID | 10102 |
Feature activation | -8.71e-01 |
Pos logit contributions | |
dot | +0.092 |
fly | +0.087 |
ening | +0.084 |
en | +0.081 |
ley | +0.080 |
Neg logit contributions | |
cream | -0.109 |
Deal | -0.109 |
math | -0.109 |
subt | -0.103 |
cakes | -0.102 |
Token | . |
Token ID | 13 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
ir | +0.029 |
hawk | +0.028 |
ug | +0.028 |
dot | +0.028 |
el | +0.028 |
Neg logit contributions | |
math | -0.021 |
cream | -0.019 |
cakes | -0.019 |
homework | -0.018 |
subt | -0.018 |
Token |
|
Token ID | 198 |
Feature activation | -1.76e-01 |
Pos logit contributions | |
en | +0.005 |
dot | +0.005 |
fly | +0.005 |
el | +0.005 |
ug | +0.005 |
Neg logit contributions | |
subt | -0.006 |
math | -0.006 |
Deal | -0.006 |
cream | -0.006 |
subtract | -0.006 |
Token |
|
Token ID | 198 |
Feature activation | -5.45e-01 |
Pos logit contributions | |
ening | +0.007 |
dot | +0.007 |
fly | +0.007 |
en | +0.007 |
ley | +0.007 |
Neg logit contributions | |
math | -0.003 |
Deal | -0.003 |
subt | -0.003 |
cream | -0.002 |
Fif | -0.002 |
Token | One |
Token ID | 3198 |
Feature activation | +3.80e-01 |
Pos logit contributions | |
en | +0.015 |
el | +0.015 |
dot | +0.015 |
ir | +0.015 |
ug | +0.015 |
Neg logit contributions | |
subt | -0.016 |
topp | -0.015 |
cream | -0.015 |
math | -0.015 |
subtract | -0.015 |
Token | Tom |
Token ID | 4186 |
Feature activation | +1.49e-01 |
Pos logit contributions | |
dot | +0.003 |
fly | +0.003 |
inkle | +0.003 |
ink | +0.003 |
wil | +0.003 |
Neg logit contributions | |
math | -0.002 |
cream | -0.002 |
subt | -0.002 |
pizza | -0.002 |
topp | -0.002 |
Token | a |
Token ID | 257 |
Feature activation | -3.14e-01 |
Pos logit contributions | |
subt | +0.003 |
math | +0.003 |
cream | +0.003 |
topp | +0.003 |
Deal | +0.003 |
Neg logit contributions | |
dot | -0.005 |
en | -0.005 |
el | -0.005 |
fly | -0.005 |
ug | -0.005 |
Token | lick |
Token ID | 36728 |
Feature activation | +1.81e-01 |
Pos logit contributions | |
dot | +0.009 |
fly | +0.009 |
ches | +0.008 |
pt | +0.008 |
inkle | +0.008 |
Neg logit contributions | |
subt | -0.009 |
Deal | -0.009 |
subtract | -0.008 |
topp | -0.008 |
Fif | -0.008 |
Token | and |
Token ID | 290 |
Feature activation | -3.43e-01 |
Pos logit contributions | |
cream | +0.006 |
math | +0.006 |
topp | +0.006 |
pizza | +0.005 |
las | +0.005 |
Neg logit contributions | |
inkle | -0.004 |
dot | -0.004 |
wil | -0.004 |
en | -0.004 |
fly | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | -8.53e-02 |
Pos logit contributions | |
en | +0.009 |
dot | +0.009 |
fly | +0.009 |
el | +0.009 |
le | +0.008 |
Neg logit contributions | |
Deal | -0.010 |
subt | -0.010 |
bek | -0.010 |
cream | -0.010 |
Fif | -0.009 |
Token | n |
Token ID | 299 |
Feature activation | -3.46e+00 |
Pos logit contributions | |
dot | +0.002 |
fly | +0.002 |
wil | +0.002 |
inkle | +0.002 |
wing | +0.002 |
Neg logit contributions | |
subt | -0.003 |
Deal | -0.003 |
subtract | -0.002 |
math | -0.002 |
Fif | -0.002 |
Token | udge |
Token ID | 12587 |
Feature activation | -7.44e-01 |
Pos logit contributions | |
ening | +0.093 |
dot | +0.089 |
ches | +0.088 |
en | +0.087 |
predators | +0.085 |
Neg logit contributions | |
Deal | -0.097 |
cream | -0.096 |
math | -0.093 |
topp | -0.091 |
Fif | -0.089 |
Token | . |
Token ID | 13 |
Feature activation | -6.29e-01 |
Pos logit contributions | |
en | +0.019 |
dot | +0.018 |
el | +0.018 |
fly | +0.018 |
ug | +0.018 |
Neg logit contributions | |
cream | -0.023 |
subt | -0.023 |
math | -0.023 |
topp | -0.023 |
Deal | -0.022 |
Token | Tom |
Token ID | 4186 |
Feature activation | -4.05e-02 |
Pos logit contributions | |
dot | +0.018 |
en | +0.017 |
el | +0.017 |
fly | +0.017 |
ug | +0.017 |
Neg logit contributions | |
subt | -0.018 |
cream | -0.017 |
math | -0.017 |
Deal | -0.017 |
topp | -0.017 |
Token | smiles |
Token ID | 21845 |
Feature activation | +4.25e-01 |
Pos logit contributions | |
ug | +0.001 |
ir | +0.001 |
dot | +0.001 |
el | +0.001 |
fly | +0.001 |
Neg logit contributions | |
math | -0.001 |
cream | -0.001 |
subt | -0.001 |
topp | -0.001 |
subtract | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | +4.45e-01 |
Pos logit contributions | |
subt | +0.018 |
math | +0.017 |
cream | +0.017 |
Deal | +0.017 |
topp | +0.017 |
Neg logit contributions | |
en | -0.006 |
dot | -0.006 |
el | -0.006 |
fly | -0.006 |
ug | -0.006 |
Token | some |
Token ID | 617 |
Feature activation | -7.01e-01 |
Pos logit contributions | |
wil | +0.032 |
landed | +0.031 |
pt | +0.030 |
en | +0.030 |
fly | +0.030 |
Neg logit contributions | |
subt | -0.042 |
topp | -0.040 |
las | -0.040 |
subtract | -0.039 |
igsaw | -0.039 |
Token | friends |
Token ID | 2460 |
Feature activation | -2.91e-01 |
Pos logit contributions | |
en | +0.028 |
fly | +0.028 |
dot | +0.028 |
el | +0.027 |
ening | +0.027 |
Neg logit contributions | |
cream | -0.013 |
math | -0.012 |
Deal | -0.012 |
topp | -0.012 |
subt | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | -1.99e-01 |
Pos logit contributions | |
dot | +0.009 |
en | +0.009 |
el | +0.008 |
ug | +0.008 |
ir | +0.008 |
Neg logit contributions | |
cream | -0.008 |
math | -0.008 |
Deal | -0.008 |
subt | -0.008 |
topp | -0.007 |
Token | She |
Token ID | 1375 |
Feature activation | -4.54e-01 |
Pos logit contributions | |
en | +0.007 |
le | +0.007 |
ening | +0.007 |
ir | +0.007 |
inkle | +0.007 |
Neg logit contributions | |
math | -0.004 |
subt | -0.004 |
Deal | -0.004 |
subtract | -0.004 |
Fif | -0.004 |
Token | wished |
Token ID | 16555 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
ir | +0.016 |
ug | +0.016 |
ening | +0.016 |
ays | +0.016 |
ors | +0.016 |
Neg logit contributions | |
Deal | -0.010 |
cream | -0.010 |
y | -0.009 |
subt | -0.009 |
cash | -0.009 |
Token | Snow |
Token ID | 7967 |
Feature activation | -3.45e+00 |
Pos logit contributions | |
dot | +0.049 |
en | +0.049 |
fly | +0.048 |
le | +0.047 |
ening | +0.047 |
Neg logit contributions | |
math | -0.020 |
cream | -0.019 |
subt | -0.018 |
cakes | -0.017 |
las | -0.017 |
Token | y |
Token ID | 88 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
predators | +0.079 |
prope | +0.071 |
inkle | +0.071 |
ening | +0.070 |
dot | +0.070 |
Neg logit contributions | |
cream | -0.107 |
Deal | -0.099 |
topp | -0.091 |
las | -0.087 |
Un | -0.086 |
Token | was |
Token ID | 373 |
Feature activation | -1.38e+00 |
Pos logit contributions | |
ir | +0.034 |
dot | +0.033 |
ug | +0.033 |
inkle | +0.032 |
ga | +0.032 |
Neg logit contributions | |
cream | -0.028 |
math | -0.026 |
Deal | -0.026 |
job | -0.024 |
topp | -0.024 |
Token | alive |
Token ID | 6776 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
en | +0.046 |
dot | +0.046 |
el | +0.045 |
fly | +0.045 |
ug | +0.044 |
Neg logit contributions | |
math | -0.031 |
subt | -0.031 |
cream | -0.031 |
Deal | -0.030 |
topp | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | -4.77e-01 |
Pos logit contributions | |
ir | +0.041 |
dot | +0.041 |
el | +0.040 |
unks | +0.040 |
inkle | +0.040 |
Neg logit contributions | |
math | -0.024 |
cream | -0.023 |
Deal | -0.022 |
y | -0.020 |
pizza | -0.020 |
Token | She |
Token ID | 1375 |
Feature activation | -4.79e-01 |
Pos logit contributions | |
en | +0.017 |
ening | +0.016 |
ir | +0.016 |
le | +0.016 |
fly | +0.016 |
Neg logit contributions | |
Deal | -0.011 |
math | -0.011 |
subt | -0.011 |
subtract | -0.010 |
Fif | -0.009 |
Token | , |
Token ID | 11 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
dot | +0.045 |
ening | +0.043 |
wil | +0.042 |
en | +0.042 |
ga | +0.041 |
Neg logit contributions | |
math | -0.049 |
cream | -0.048 |
pizza | -0.047 |
cakes | -0.046 |
las | -0.046 |
Token | a |
Token ID | 257 |
Feature activation | -6.05e-01 |
Pos logit contributions | |
en | +0.078 |
dot | +0.076 |
el | +0.075 |
ening | +0.074 |
ug | +0.074 |
Neg logit contributions | |
math | -0.056 |
subt | -0.056 |
cream | -0.056 |
Deal | -0.055 |
topp | -0.054 |
Token | little |
Token ID | 1310 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
en | +0.019 |
dot | +0.019 |
el | +0.019 |
fly | +0.018 |
ug | +0.018 |
Neg logit contributions | |
subt | -0.015 |
math | -0.015 |
cream | -0.015 |
Deal | -0.014 |
topp | -0.014 |
Token | chick |
Token ID | 17943 |
Feature activation | -1.87e+00 |
Pos logit contributions | |
en | +0.035 |
inkle | +0.034 |
ches | +0.033 |
ening | +0.033 |
ink | +0.033 |
Neg logit contributions | |
cream | -0.027 |
math | -0.026 |
Deal | -0.025 |
pizza | -0.025 |
topp | -0.025 |
Token | named |
Token ID | 3706 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
en | +0.049 |
ening | +0.049 |
dot | +0.047 |
el | +0.046 |
inkle | +0.046 |
Neg logit contributions | |
cream | -0.054 |
Deal | -0.053 |
math | -0.051 |
pizza | -0.050 |
job | -0.050 |
Token | Ch |
Token ID | 609 |
Feature activation | -3.43e+00 |
Pos logit contributions | |
en | +0.049 |
ening | +0.049 |
le | +0.047 |
dot | +0.047 |
inkle | +0.046 |
Neg logit contributions | |
math | -0.043 |
pizza | -0.041 |
las | -0.041 |
y | -0.041 |
cream | -0.041 |
Token | ir |
Token ID | 343 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
lands | +0.070 |
ening | +0.067 |
dot | +0.061 |
fly | +0.060 |
landed | +0.059 |
Neg logit contributions | |
igsaw | -0.099 |
lunch | -0.097 |
dinner | -0.094 |
math | -0.091 |
dessert | -0.089 |
Token | py |
Token ID | 9078 |
Feature activation | -1.49e+00 |
Pos logit contributions | |
inkle | +0.017 |
dot | +0.017 |
wil | +0.015 |
ays | +0.015 |
ening | +0.014 |
Neg logit contributions | |
cream | -0.051 |
math | -0.050 |
subtract | -0.049 |
cakes | -0.048 |
pies | -0.048 |
Token | . |
Token ID | 13 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
inkle | +0.046 |
dot | +0.045 |
en | +0.045 |
ening | +0.044 |
el | +0.044 |
Neg logit contributions | |
Deal | -0.036 |
math | -0.035 |
cream | -0.035 |
topp | -0.033 |
subt | -0.033 |
Token | They |
Token ID | 1119 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
wil | +0.036 |
le | +0.035 |
ening | +0.034 |
en | +0.033 |
inkle | +0.033 |
Neg logit contributions | |
math | -0.028 |
subt | -0.027 |
las | -0.026 |
Grad | -0.026 |
subtract | -0.026 |
Token | played |
Token ID | 2826 |
Feature activation | -9.91e-01 |
Pos logit contributions | |
ays | +0.008 |
inkle | +0.007 |
en | +0.007 |
ening | +0.007 |
el | +0.007 |
Neg logit contributions | |
Deal | -0.004 |
y | -0.003 |
cream | -0.003 |
topp | -0.003 |
cakes | -0.003 |
Token | to |
Token ID | 284 |
Feature activation | +6.75e-01 |
Pos logit contributions | |
subt | +0.026 |
Deal | +0.025 |
bek | +0.024 |
subtract | +0.024 |
Fif | +0.024 |
Neg logit contributions | |
dot | -0.018 |
fly | -0.017 |
en | -0.016 |
ga | -0.016 |
el | -0.016 |
Token | play |
Token ID | 711 |
Feature activation | -9.14e-01 |
Pos logit contributions | |
subt | +0.013 |
math | +0.013 |
cream | +0.013 |
Deal | +0.013 |
topp | +0.013 |
Neg logit contributions | |
en | -0.025 |
dot | -0.025 |
el | -0.024 |
fly | -0.024 |
ening | -0.024 |
Token | in |
Token ID | 287 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
en | +0.025 |
dot | +0.025 |
el | +0.024 |
fly | +0.024 |
ug | +0.024 |
Neg logit contributions | |
subt | -0.027 |
math | -0.026 |
topp | -0.026 |
cream | -0.026 |
subtract | -0.026 |
Token | the |
Token ID | 262 |
Feature activation | -1.00e+00 |
Pos logit contributions | |
en | +0.032 |
dot | +0.032 |
el | +0.031 |
fly | +0.031 |
ug | +0.031 |
Neg logit contributions | |
subt | -0.028 |
math | -0.028 |
cream | -0.028 |
topp | -0.027 |
Deal | -0.027 |
Token | park |
Token ID | 3952 |
Feature activation | -8.06e-01 |
Pos logit contributions | |
dot | +0.028 |
en | +0.028 |
el | +0.028 |
fly | +0.026 |
wil | +0.026 |
Neg logit contributions | |
subt | -0.029 |
Deal | -0.028 |
topp | -0.028 |
subtract | -0.028 |
cream | -0.026 |
Token | every |
Token ID | 790 |
Feature activation | -3.41e+00 |
Pos logit contributions | |
en | +0.021 |
dot | +0.021 |
el | +0.020 |
fly | +0.020 |
ening | +0.020 |
Neg logit contributions | |
math | -0.024 |
Deal | -0.024 |
cream | -0.024 |
subt | -0.024 |
topp | -0.024 |
Token | day |
Token ID | 1110 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
en | +0.131 |
dot | +0.129 |
fly | +0.128 |
el | +0.127 |
ug | +0.127 |
Neg logit contributions | |
cream | -0.062 |
math | -0.061 |
subt | -0.058 |
Deal | -0.056 |
topp | -0.056 |
Token | . |
Token ID | 13 |
Feature activation | -1.50e+00 |
Pos logit contributions | |
fly | +0.040 |
en | +0.040 |
dot | +0.039 |
ley | +0.038 |
py | +0.038 |
Neg logit contributions | |
math | -0.034 |
cream | -0.032 |
subt | -0.032 |
topp | -0.031 |
Deal | -0.030 |
Token | True |
Token ID | 6407 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
en | +0.043 |
el | +0.042 |
ug | +0.042 |
dot | +0.041 |
fly | +0.041 |
Neg logit contributions | |
math | -0.043 |
subt | -0.042 |
Deal | -0.041 |
topp | -0.040 |
cream | -0.040 |
Token | is |
Token ID | 318 |
Feature activation | -7.36e-01 |
Pos logit contributions | |
ir | +0.059 |
ug | +0.057 |
dot | +0.055 |
el | +0.053 |
py | +0.050 |
Neg logit contributions | |
Deal | -0.039 |
math | -0.039 |
cream | -0.035 |
pizza | -0.033 |
topp | -0.033 |
Token | a |
Token ID | 257 |
Feature activation | -4.18e-01 |
Pos logit contributions | |
en | +0.023 |
dot | +0.023 |
el | +0.023 |
fly | +0.023 |
ug | +0.022 |
Neg logit contributions | |
subt | -0.018 |
math | -0.018 |
cream | -0.018 |
topp | -0.018 |
Deal | -0.018 |
Token | to |
Token ID | 284 |
Feature activation | +8.60e-01 |
Pos logit contributions | |
Deal | +0.041 |
subt | +0.041 |
bek | +0.040 |
subtract | +0.038 |
Fif | +0.038 |
Neg logit contributions | |
dot | -0.024 |
el | -0.022 |
en | -0.022 |
fly | -0.021 |
ga | -0.021 |
Token | play |
Token ID | 711 |
Feature activation | -5.94e-01 |
Pos logit contributions | |
subt | +0.017 |
math | +0.017 |
cream | +0.017 |
topp | +0.016 |
Deal | +0.016 |
Neg logit contributions | |
en | -0.031 |
dot | -0.031 |
el | -0.030 |
fly | -0.030 |
ug | -0.030 |
Token | in |
Token ID | 287 |
Feature activation | -7.76e-01 |
Pos logit contributions | |
en | +0.013 |
dot | +0.013 |
le | +0.013 |
ug | +0.013 |
el | +0.013 |
Neg logit contributions | |
subt | -0.021 |
subtract | -0.020 |
topp | -0.019 |
Fif | -0.019 |
Deal | -0.019 |
Token | the |
Token ID | 262 |
Feature activation | -7.35e-01 |
Pos logit contributions | |
en | +0.021 |
dot | +0.021 |
el | +0.021 |
fly | +0.020 |
ug | +0.020 |
Neg logit contributions | |
subt | -0.023 |
Deal | -0.023 |
topp | -0.022 |
cream | -0.022 |
math | -0.022 |
Token | park |
Token ID | 3952 |
Feature activation | -5.23e-01 |
Pos logit contributions | |
el | +0.018 |
dot | +0.018 |
en | +0.017 |
wil | +0.016 |
fly | +0.016 |
Neg logit contributions | |
subt | -0.024 |
Deal | -0.024 |
subtract | -0.024 |
topp | -0.023 |
bek | -0.022 |
Token | every |
Token ID | 790 |
Feature activation | -3.40e+00 |
Pos logit contributions | |
en | +0.012 |
dot | +0.012 |
el | +0.012 |
fly | +0.012 |
ug | +0.011 |
Neg logit contributions | |
subt | -0.018 |
math | -0.017 |
cream | -0.017 |
topp | -0.017 |
Deal | -0.017 |
Token | day |
Token ID | 1110 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
en | +0.125 |
dot | +0.124 |
fly | +0.122 |
el | +0.121 |
ug | +0.120 |
Neg logit contributions | |
cream | -0.066 |
math | -0.066 |
subt | -0.065 |
Deal | -0.062 |
topp | -0.062 |
Token | . |
Token ID | 13 |
Feature activation | -7.63e-01 |
Pos logit contributions | |
fly | +0.031 |
en | +0.031 |
dot | +0.031 |
ley | +0.030 |
py | +0.030 |
Neg logit contributions | |
math | -0.030 |
subt | -0.030 |
cream | -0.030 |
topp | -0.029 |
Deal | -0.028 |
Token | They |
Token ID | 1119 |
Feature activation | -5.25e-01 |
Pos logit contributions | |
en | +0.018 |
dot | +0.018 |
el | +0.017 |
fly | +0.017 |
le | +0.016 |
Neg logit contributions | |
subt | -0.025 |
cream | -0.025 |
Deal | -0.025 |
math | -0.025 |
topp | -0.024 |
Token | swing |
Token ID | 9628 |
Feature activation | -8.56e-01 |
Pos logit contributions | |
el | +0.018 |
en | +0.018 |
dot | +0.018 |
ening | +0.018 |
fly | +0.018 |
Neg logit contributions | |
Deal | -0.012 |
math | -0.011 |
subt | -0.011 |
cream | -0.011 |
topp | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
dot | +0.026 |
ening | +0.025 |
en | +0.024 |
inkle | +0.024 |
fly | +0.024 |
Neg logit contributions | |
cream | -0.026 |
math | -0.025 |
Deal | -0.024 |
cakes | -0.023 |
pizza | -0.022 |
Token | � |
Token ID | 129 |
Feature activation | +4.33e-01 |
Pos logit contributions | |
subt | +0.019 |
cream | +0.018 |
math | +0.018 |
topp | +0.018 |
Deal | +0.018 |
Neg logit contributions | |
dot | -0.017 |
en | -0.017 |
fly | -0.017 |
el | -0.017 |
ug | -0.017 |
Token | � |
Token ID | 241 |
Feature activation | +4.61e-01 |
Pos logit contributions | |
subt | +0.012 |
math | +0.012 |
cream | +0.012 |
subtract | +0.012 |
Deal | +0.012 |
Neg logit contributions | |
dot | -0.012 |
fly | -0.012 |
en | -0.012 |
el | -0.012 |
ug | -0.012 |
Token | Don |
Token ID | 3987 |
Feature activation | -7.78e-01 |
Pos logit contributions | |
subt | +0.011 |
math | +0.011 |
cream | +0.011 |
Deal | +0.011 |
topp | +0.011 |
Neg logit contributions | |
en | -0.015 |
dot | -0.015 |
el | -0.015 |
fly | -0.014 |
ug | -0.014 |
Token | â |
Token ID | 22940 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
dot | +0.016 |
en | +0.016 |
fly | +0.016 |
el | +0.015 |
ening | +0.015 |
Neg logit contributions | |
subt | -0.028 |
math | -0.028 |
cream | -0.027 |
Deal | -0.027 |
subtract | -0.027 |
Token | € |
Token ID | 26391 |
Feature activation | +8.64e-01 |
Pos logit contributions | |
math | +0.008 |
topp | +0.008 |
subt | +0.008 |
cakes | +0.007 |
cream | +0.007 |
Neg logit contributions | |
en | -0.010 |
el | -0.009 |
le | -0.009 |
ga | -0.009 |
ug | -0.009 |
Token | ™ |
Token ID | 8151 |
Feature activation | +3.64e+00 |
Pos logit contributions | |
subt | +0.025 |
cream | +0.025 |
math | +0.025 |
topp | +0.025 |
Deal | +0.025 |
Neg logit contributions | |
dot | -0.023 |
fly | -0.022 |
en | -0.022 |
el | -0.022 |
ug | -0.022 |
Token | t |
Token ID | 83 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
math | +0.101 |
subt | +0.101 |
Deal | +0.100 |
cream | +0.100 |
topp | +0.098 |
Neg logit contributions | |
dot | -0.105 |
en | -0.104 |
ening | -0.099 |
le | -0.097 |
fly | -0.096 |
Token | be |
Token ID | 307 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
math | +0.030 |
subtract | +0.029 |
subt | +0.029 |
cream | +0.028 |
job | +0.028 |
Neg logit contributions | |
ays | -0.038 |
dot | -0.038 |
fly | -0.038 |
ley | -0.037 |
el | -0.037 |
Token | scared |
Token ID | 12008 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
subt | +0.030 |
math | +0.030 |
cream | +0.029 |
subtract | +0.029 |
topp | +0.028 |
Neg logit contributions | |
dot | -0.032 |
el | -0.031 |
en | -0.031 |
fly | -0.031 |
ley | -0.031 |
Token | , |
Token ID | 11 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
math | +0.030 |
cream | +0.029 |
job | +0.027 |
subt | +0.026 |
subtract | +0.026 |
Neg logit contributions | |
fly | -0.036 |
dot | -0.036 |
inkle | -0.035 |
wil | -0.035 |
ittered | -0.034 |
Token | I |
Token ID | 314 |
Feature activation | -4.19e-01 |
Pos logit contributions | |
math | +0.031 |
subt | +0.031 |
cream | +0.031 |
subtract | +0.029 |
topp | +0.029 |
Neg logit contributions | |
fly | -0.031 |
en | -0.031 |
el | -0.031 |
dot | -0.031 |
ening | -0.030 |
Token | � |
Token ID | 129 |
Feature activation | +5.43e-01 |
Pos logit contributions | |
subt | +0.019 |
math | +0.019 |
cream | +0.019 |
topp | +0.018 |
Deal | +0.018 |
Neg logit contributions | |
en | -0.021 |
dot | -0.021 |
el | -0.020 |
fly | -0.020 |
ug | -0.020 |
Token | � |
Token ID | 241 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
subt | +0.015 |
math | +0.014 |
cream | +0.014 |
Deal | +0.014 |
subtract | +0.014 |
Neg logit contributions | |
dot | -0.016 |
fly | -0.016 |
en | -0.016 |
el | -0.016 |
ug | -0.015 |
Token | That |
Token ID | 2504 |
Feature activation | +7.61e-01 |
Pos logit contributions | |
subt | +0.004 |
math | +0.003 |
cream | +0.003 |
topp | +0.003 |
Deal | +0.003 |
Neg logit contributions | |
en | -0.004 |
dot | -0.004 |
el | -0.004 |
fly | -0.004 |
ug | -0.004 |
Token | â |
Token ID | 22940 |
Feature activation | -2.64e-01 |
Pos logit contributions | |
math | +0.022 |
cream | +0.022 |
subt | +0.022 |
topp | +0.022 |
subtract | +0.021 |
Neg logit contributions | |
ir | -0.021 |
dot | -0.021 |
ug | -0.021 |
fly | -0.020 |
en | -0.020 |
Token | € |
Token ID | 26391 |
Feature activation | +6.20e-01 |
Pos logit contributions | |
en | +0.008 |
el | +0.008 |
le | +0.008 |
ug | +0.008 |
ches | +0.008 |
Neg logit contributions | |
math | -0.007 |
subt | -0.006 |
topp | -0.006 |
cream | -0.006 |
subtract | -0.006 |
Token | ™ |
Token ID | 8151 |
Feature activation | +3.58e+00 |
Pos logit contributions | |
subt | +0.017 |
math | +0.017 |
cream | +0.017 |
Deal | +0.017 |
topp | +0.017 |
Neg logit contributions | |
en | -0.018 |
dot | -0.018 |
el | -0.017 |
fly | -0.017 |
ug | -0.017 |
Token | s |
Token ID | 82 |
Feature activation | +9.66e-01 |
Pos logit contributions | |
cream | +0.106 |
Deal | +0.106 |
math | +0.106 |
subt | +0.105 |
topp | +0.104 |
Neg logit contributions | |
dot | -0.100 |
en | -0.095 |
ening | -0.090 |
fly | -0.090 |
predators | -0.089 |
Token | an |
Token ID | 281 |
Feature activation | +7.60e-01 |
Pos logit contributions | |
math | +0.030 |
Deal | +0.030 |
cream | +0.030 |
subt | +0.029 |
topp | +0.029 |
Neg logit contributions | |
dot | -0.025 |
en | -0.024 |
fly | -0.024 |
ir | -0.024 |
ug | -0.023 |
Token | elevator |
Token ID | 20932 |
Feature activation | +1.41e+00 |
Pos logit contributions | |
math | +0.017 |
cream | +0.016 |
job | +0.016 |
subt | +0.016 |
list | +0.016 |
Neg logit contributions | |
fly | -0.028 |
dot | -0.026 |
ays | -0.025 |
en | -0.025 |
ches | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | +5.33e-01 |
Pos logit contributions | |
cream | +0.045 |
subt | +0.045 |
math | +0.045 |
Deal | +0.043 |
topp | +0.043 |
Neg logit contributions | |
dot | -0.035 |
en | -0.034 |
el | -0.033 |
ir | -0.033 |
ug | -0.033 |
Token | It |
Token ID | 632 |
Feature activation | -8.29e-02 |
Pos logit contributions | |
subt | +0.015 |
math | +0.015 |
subtract | +0.015 |
Deal | +0.014 |
topp | +0.014 |
Neg logit contributions | |
en | -0.015 |
fly | -0.015 |
dot | -0.015 |
wil | -0.015 |
le | -0.015 |
Token | on |
Token ID | 319 |
Feature activation | +5.38e-01 |
Pos logit contributions | |
subt | +0.033 |
topp | +0.033 |
cream | +0.033 |
math | +0.033 |
subtract | +0.033 |
Neg logit contributions | |
el | -0.021 |
ug | -0.021 |
en | -0.020 |
dot | -0.020 |
ir | -0.019 |
Token | , |
Token ID | 11 |
Feature activation | +6.11e-01 |
Pos logit contributions | |
math | +0.017 |
cream | +0.016 |
subt | +0.016 |
job | +0.016 |
pizza | +0.015 |
Neg logit contributions | |
dot | -0.015 |
fly | -0.015 |
en | -0.014 |
wil | -0.014 |
ga | -0.014 |
Token | they |
Token ID | 484 |
Feature activation | +4.61e-01 |
Pos logit contributions | |
subt | +0.022 |
Deal | +0.022 |
subtract | +0.021 |
bek | +0.021 |
Fif | +0.021 |
Neg logit contributions | |
en | -0.012 |
or | -0.011 |
el | -0.011 |
dot | -0.011 |
ug | -0.010 |
Token | never |
Token ID | 1239 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
subt | +0.014 |
math | +0.013 |
cream | +0.013 |
Deal | +0.013 |
topp | +0.013 |
Neg logit contributions | |
en | -0.012 |
dot | -0.012 |
el | -0.012 |
fly | -0.012 |
ug | -0.012 |
Token | qu |
Token ID | 627 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
en | +0.008 |
dot | +0.007 |
fly | +0.007 |
el | +0.007 |
le | +0.007 |
Neg logit contributions | |
math | -0.009 |
subt | -0.009 |
cream | -0.009 |
Deal | -0.008 |
topp | -0.008 |
Token | arre |
Token ID | 9624 |
Feature activation | +3.78e+00 |
Pos logit contributions | |
subt | +0.021 |
math | +0.020 |
cream | +0.020 |
topp | +0.020 |
subtract | +0.019 |
Neg logit contributions | |
dot | -0.078 |
en | -0.077 |
fly | -0.077 |
el | -0.076 |
ley | -0.076 |
Token | led |
Token ID | 992 |
Feature activation | +5.93e-01 |
Pos logit contributions | |
subt | +0.160 |
math | +0.157 |
cream | +0.156 |
Deal | +0.155 |
topp | +0.155 |
Neg logit contributions | |
en | -0.056 |
dot | -0.053 |
el | -0.052 |
ug | -0.050 |
le | -0.050 |
Token | again |
Token ID | 757 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
math | +0.018 |
cream | +0.017 |
Deal | +0.016 |
subt | +0.016 |
topp | +0.016 |
Neg logit contributions | |
dot | -0.016 |
el | -0.016 |
ley | -0.016 |
ir | -0.016 |
inkle | -0.016 |
Token | and |
Token ID | 290 |
Feature activation | +2.26e-01 |
Pos logit contributions | |
subt | +0.012 |
math | +0.011 |
cream | +0.011 |
topp | +0.011 |
Deal | +0.011 |
Neg logit contributions | |
dot | -0.006 |
en | -0.006 |
el | -0.006 |
fly | -0.006 |
ug | -0.005 |
Token | played |
Token ID | 2826 |
Feature activation | -5.63e-01 |
Pos logit contributions | |
subt | +0.007 |
topp | +0.007 |
cream | +0.007 |
math | +0.007 |
Deal | +0.007 |
Neg logit contributions | |
en | -0.006 |
fly | -0.005 |
el | -0.005 |
dot | -0.005 |
wil | -0.005 |
Token | together |
Token ID | 1978 |
Feature activation | -7.05e-01 |
Pos logit contributions | |
en | +0.015 |
le | +0.014 |
dot | +0.014 |
ug | +0.014 |
inkle | +0.013 |
Neg logit contributions | |
subt | -0.017 |
subtract | -0.017 |
topp | -0.017 |
las | -0.016 |
Fif | -0.016 |
Token | thought |
Token ID | 1807 |
Feature activation | +6.16e-01 |
Pos logit contributions | |
subt | +0.015 |
subtract | +0.015 |
math | +0.014 |
cream | +0.014 |
Deal | +0.013 |
Neg logit contributions | |
ir | -0.024 |
fly | -0.023 |
py | -0.023 |
predators | -0.023 |
ays | -0.023 |
Token | they |
Token ID | 484 |
Feature activation | +4.84e-01 |
Pos logit contributions | |
math | +0.020 |
subt | +0.016 |
cakes | +0.016 |
cream | +0.015 |
homework | +0.015 |
Neg logit contributions | |
fly | -0.021 |
ley | -0.020 |
dot | -0.019 |
an | -0.018 |
ga | -0.018 |
Token | looked |
Token ID | 3114 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
Deal | +0.012 |
topp | +0.011 |
y | +0.011 |
subtract | +0.011 |
cream | +0.011 |
Neg logit contributions | |
ug | -0.016 |
ir | -0.016 |
ga | -0.016 |
inkle | -0.015 |
ays | -0.015 |
Token | like |
Token ID | 588 |
Feature activation | +7.77e-01 |
Pos logit contributions | |
subt | +0.030 |
math | +0.030 |
cream | +0.029 |
topp | +0.029 |
Deal | +0.029 |
Neg logit contributions | |
en | -0.035 |
dot | -0.035 |
el | -0.034 |
fly | -0.034 |
ug | -0.034 |
Token | ac |
Token ID | 936 |
Feature activation | +3.40e-01 |
Pos logit contributions | |
math | +0.018 |
subt | +0.017 |
cream | +0.017 |
topp | +0.016 |
subtract | +0.016 |
Neg logit contributions | |
en | -0.027 |
el | -0.027 |
dot | -0.027 |
fly | -0.026 |
ir | -0.026 |
Token | rob |
Token ID | 22609 |
Feature activation | +4.32e+00 |
Pos logit contributions | |
cream | +0.008 |
Deal | +0.007 |
subt | +0.007 |
math | +0.007 |
topp | +0.007 |
Neg logit contributions | |
dot | -0.012 |
en | -0.012 |
fly | -0.012 |
el | -0.011 |
ug | -0.011 |
Token | ats |
Token ID | 1381 |
Feature activation | +5.80e-02 |
Pos logit contributions | |
subt | +0.146 |
math | +0.141 |
cream | +0.140 |
Deal | +0.139 |
subtract | +0.139 |
Neg logit contributions | |
en | -0.099 |
el | -0.098 |
ir | -0.096 |
ug | -0.094 |
le | -0.094 |
Token | . |
Token ID | 13 |
Feature activation | +2.45e-01 |
Pos logit contributions | |
cream | +0.002 |
math | +0.002 |
cakes | +0.002 |
pizza | +0.002 |
cookies | +0.002 |
Neg logit contributions | |
inkle | -0.002 |
dot | -0.002 |
ug | -0.002 |
ir | -0.002 |
el | -0.001 |
Token | She |
Token ID | 1375 |
Feature activation | +5.01e-01 |
Pos logit contributions | |
subt | +0.008 |
cream | +0.008 |
math | +0.008 |
topp | +0.008 |
Deal | +0.008 |
Neg logit contributions | |
en | -0.006 |
dot | -0.006 |
el | -0.006 |
fly | -0.006 |
ug | -0.005 |
Token | enjoyed |
Token ID | 8359 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
subt | +0.012 |
cream | +0.011 |
subtract | +0.011 |
math | +0.011 |
topp | +0.011 |
Neg logit contributions | |
ir | -0.017 |
fly | -0.017 |
ug | -0.017 |
ays | -0.017 |
dot | -0.017 |
Token | the |
Token ID | 262 |
Feature activation | +5.45e-01 |
Pos logit contributions | |
en | +0.008 |
dot | +0.008 |
el | +0.008 |
fly | +0.007 |
ug | +0.007 |
Neg logit contributions | |
subt | -0.007 |
math | -0.007 |
cream | -0.007 |
Deal | -0.007 |
topp | -0.007 |
Token | � |
Token ID | 129 |
Feature activation | +8.28e-01 |
Pos logit contributions | |
subt | +0.022 |
math | +0.022 |
cream | +0.021 |
Deal | +0.021 |
topp | +0.021 |
Neg logit contributions | |
en | -0.026 |
dot | -0.026 |
el | -0.026 |
fly | -0.025 |
ening | -0.025 |
Token | � |
Token ID | 241 |
Feature activation | +6.53e-01 |
Pos logit contributions | |
subt | +0.024 |
math | +0.023 |
cream | +0.023 |
topp | +0.023 |
Deal | +0.023 |
Neg logit contributions | |
en | -0.023 |
dot | -0.023 |
el | -0.022 |
fly | -0.022 |
ug | -0.022 |
Token | That |
Token ID | 2504 |
Feature activation | +8.96e-01 |
Pos logit contributions | |
subt | +0.016 |
math | +0.016 |
cream | +0.016 |
topp | +0.016 |
Deal | +0.016 |
Neg logit contributions | |
en | -0.020 |
dot | -0.020 |
el | -0.020 |
fly | -0.020 |
ug | -0.019 |
Token | â |
Token ID | 22940 |
Feature activation | -2.07e-01 |
Pos logit contributions | |
math | +0.028 |
cream | +0.028 |
subt | +0.028 |
topp | +0.028 |
Deal | +0.027 |
Neg logit contributions | |
dot | -0.022 |
en | -0.022 |
ir | -0.022 |
fly | -0.022 |
el | -0.022 |
Token | € |
Token ID | 26391 |
Feature activation | +7.10e-01 |
Pos logit contributions | |
en | +0.007 |
el | +0.006 |
le | +0.006 |
ug | +0.006 |
ening | +0.006 |
Neg logit contributions | |
math | -0.005 |
topp | -0.005 |
subt | -0.005 |
cream | -0.005 |
cakes | -0.005 |
Token | ™ |
Token ID | 8151 |
Feature activation | +3.55e+00 |
Pos logit contributions | |
subt | +0.020 |
math | +0.020 |
cream | +0.020 |
Deal | +0.019 |
topp | +0.019 |
Neg logit contributions | |
en | -0.020 |
dot | -0.020 |
el | -0.019 |
fly | -0.019 |
ug | -0.019 |
Token | s |
Token ID | 82 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
Deal | +0.106 |
cream | +0.104 |
subt | +0.103 |
math | +0.103 |
topp | +0.102 |
Neg logit contributions | |
dot | -0.099 |
en | -0.095 |
fly | -0.092 |
ening | -0.091 |
py | -0.090 |
Token | the |
Token ID | 262 |
Feature activation | +4.58e-01 |
Pos logit contributions | |
math | +0.035 |
subt | +0.035 |
cream | +0.035 |
Deal | +0.035 |
topp | +0.034 |
Neg logit contributions | |
dot | -0.032 |
en | -0.031 |
fly | -0.030 |
el | -0.030 |
ir | -0.030 |
Token | bird |
Token ID | 6512 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
subt | +0.011 |
math | +0.011 |
Deal | +0.011 |
topp | +0.011 |
cream | +0.011 |
Neg logit contributions | |
dot | -0.014 |
en | -0.014 |
el | -0.014 |
fly | -0.014 |
ug | -0.013 |
Token | that |
Token ID | 326 |
Feature activation | +9.47e-01 |
Pos logit contributions | |
cream | +0.034 |
Deal | +0.033 |
topp | +0.032 |
math | +0.032 |
y | +0.031 |
Neg logit contributions | |
dot | -0.033 |
ga | -0.031 |
ir | -0.030 |
ug | -0.030 |
en | -0.030 |
Token | was |
Token ID | 373 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
math | +0.028 |
cream | +0.028 |
subt | +0.027 |
topp | +0.027 |
Deal | +0.027 |
Neg logit contributions | |
dot | -0.025 |
en | -0.025 |
ir | -0.025 |
fly | -0.025 |
el | -0.025 |
Token | to |
Token ID | 284 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
cream | +0.014 |
math | +0.014 |
subt | +0.014 |
topp | +0.013 |
job | +0.013 |
Neg logit contributions | |
dot | -0.028 |
en | -0.027 |
fly | -0.027 |
ug | -0.026 |
py | -0.026 |
Token | go |
Token ID | 467 |
Feature activation | +7.04e-01 |
Pos logit contributions | |
subt | +0.050 |
Deal | +0.049 |
topp | +0.048 |
math | +0.047 |
subtract | +0.047 |
Neg logit contributions | |
en | -0.042 |
el | -0.041 |
ir | -0.039 |
dot | -0.038 |
le | -0.038 |
Token | , |
Token ID | 11 |
Feature activation | +4.10e-01 |
Pos logit contributions | |
cream | +0.018 |
math | +0.017 |
job | +0.016 |
pizza | +0.015 |
subt | +0.015 |
Neg logit contributions | |
dot | -0.024 |
fly | -0.023 |
ening | -0.022 |
wil | -0.022 |
en | -0.022 |
Token | Lucy |
Token ID | 22162 |
Feature activation | -7.75e-02 |
Pos logit contributions | |
subt | +0.013 |
topp | +0.013 |
subtract | +0.013 |
Deal | +0.013 |
math | +0.012 |
Neg logit contributions | |
en | -0.010 |
el | -0.010 |
dot | -0.010 |
fly | -0.009 |
le | -0.009 |
Token | and |
Token ID | 290 |
Feature activation | +2.82e+00 |
Pos logit contributions | |
ug | +0.002 |
dot | +0.002 |
ir | +0.002 |
py | +0.002 |
ening | +0.002 |
Neg logit contributions | |
cream | -0.002 |
subt | -0.002 |
topp | -0.002 |
hamb | -0.002 |
las | -0.002 |
Token | her |
Token ID | 607 |
Feature activation | +1.69e+00 |
Pos logit contributions | |
subt | +0.111 |
topp | +0.108 |
Deal | +0.105 |
subtract | +0.104 |
Fif | +0.103 |
Neg logit contributions | |
el | -0.064 |
en | -0.060 |
ie | -0.054 |
le | -0.052 |
i | -0.050 |
Token | Dad |
Token ID | 17415 |
Feature activation | +7.08e-01 |
Pos logit contributions | |
subt | +0.045 |
subtract | +0.042 |
Deal | +0.042 |
math | +0.041 |
topp | +0.041 |
Neg logit contributions | |
el | -0.053 |
en | -0.051 |
wil | -0.050 |
dot | -0.049 |
pt | -0.048 |
Token | brought |
Token ID | 3181 |
Feature activation | +4.14e-01 |
Pos logit contributions | |
cream | +0.018 |
subt | +0.018 |
Deal | +0.018 |
topp | +0.018 |
math | +0.017 |
Neg logit contributions | |
dot | -0.022 |
en | -0.021 |
fly | -0.021 |
ening | -0.021 |
el | -0.020 |
Token | the |
Token ID | 262 |
Feature activation | -1.13e+00 |
Pos logit contributions | |
subt | +0.012 |
cream | +0.011 |
math | +0.011 |
topp | +0.011 |
Deal | +0.011 |
Neg logit contributions | |
dot | -0.012 |
en | -0.012 |
fly | -0.012 |
el | -0.011 |
ug | -0.011 |
Token | cooler |
Token ID | 19346 |
Feature activation | +6.81e-01 |
Pos logit contributions | |
dot | +0.044 |
el | +0.044 |
en | +0.044 |
fly | +0.044 |
le | +0.043 |
Neg logit contributions | |
subt | -0.020 |
subtract | -0.018 |
math | -0.017 |
topp | -0.016 |
Deal | -0.016 |
Token | back |
Token ID | 736 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
subt | +0.022 |
math | +0.021 |
topp | +0.021 |
Deal | +0.021 |
subtract | +0.021 |
Neg logit contributions | |
en | -0.017 |
dot | -0.016 |
el | -0.016 |
fly | -0.016 |
ug | -0.016 |
Token | you |
Token ID | 345 |
Feature activation | +2.43e-01 |
Pos logit contributions | |
en | +0.031 |
dot | +0.031 |
el | +0.031 |
ug | +0.030 |
fly | +0.030 |
Neg logit contributions | |
subt | -0.040 |
Deal | -0.039 |
topp | -0.038 |
math | -0.038 |
cream | -0.038 |
Token | are |
Token ID | 389 |
Feature activation | +2.69e-01 |
Pos logit contributions | |
topp | +0.006 |
cream | +0.006 |
subtract | +0.006 |
Deal | +0.006 |
math | +0.006 |
Neg logit contributions | |
el | -0.008 |
dot | -0.008 |
en | -0.008 |
ening | -0.007 |
fly | -0.007 |
Token | using |
Token ID | 1262 |
Feature activation | -6.53e-02 |
Pos logit contributions | |
subt | +0.007 |
Deal | +0.006 |
topp | +0.006 |
math | +0.006 |
cream | +0.006 |
Neg logit contributions | |
en | -0.009 |
dot | -0.009 |
el | -0.008 |
fly | -0.008 |
ug | -0.008 |
Token | the |
Token ID | 262 |
Feature activation | -2.42e-01 |
Pos logit contributions | |
fly | +0.002 |
dot | +0.002 |
py | +0.002 |
le | +0.002 |
ley | +0.002 |
Neg logit contributions | |
math | -0.002 |
cream | -0.002 |
subt | -0.002 |
topp | -0.002 |
subtract | -0.001 |
Token | right |
Token ID | 826 |
Feature activation | -6.30e-01 |
Pos logit contributions | |
en | +0.008 |
dot | +0.008 |
el | +0.007 |
fly | +0.007 |
ug | +0.007 |
Neg logit contributions | |
subt | -0.006 |
math | -0.006 |
cream | -0.006 |
topp | -0.006 |
Deal | -0.006 |
Token | numbers |
Token ID | 3146 |
Feature activation | +1.59e+00 |
Pos logit contributions | |
fly | +0.020 |
dot | +0.020 |
ir | +0.020 |
el | +0.019 |
en | +0.019 |
Neg logit contributions | |
math | -0.016 |
cream | -0.015 |
topp | -0.015 |
subt | -0.015 |
subtract | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -5.20e-02 |
Pos logit contributions | |
math | +0.047 |
cream | +0.047 |
subt | +0.046 |
topp | +0.046 |
subtract | +0.045 |
Neg logit contributions | |
dot | -0.042 |
en | -0.042 |
fly | -0.041 |
el | -0.041 |
ir | -0.041 |
Token | If |
Token ID | 1002 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
dot | +0.001 |
en | +0.001 |
el | +0.001 |
ug | +0.001 |
fly | +0.001 |
Neg logit contributions | |
cream | -0.002 |
subt | -0.002 |
math | -0.002 |
topp | -0.002 |
Deal | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | +7.75e-01 |
Pos logit contributions | |
en | +0.033 |
dot | +0.033 |
el | +0.032 |
fly | +0.032 |
ug | +0.032 |
Neg logit contributions | |
subt | -0.031 |
math | -0.031 |
cream | -0.031 |
topp | -0.030 |
Deal | -0.030 |
Token | numbers |
Token ID | 3146 |
Feature activation | +9.94e-01 |
Pos logit contributions | |
subt | +0.021 |
Deal | +0.021 |
topp | +0.019 |
Fif | +0.019 |
cream | +0.019 |
Neg logit contributions | |
dot | -0.023 |
el | -0.021 |
en | -0.021 |
fly | -0.021 |
ug | -0.021 |
Token | match |
Token ID | 2872 |
Feature activation | +3.74e-01 |
Pos logit contributions | |
math | +0.029 |
cream | +0.029 |
subtract | +0.028 |
topp | +0.028 |
subt | +0.027 |
Neg logit contributions | |
dot | -0.027 |
en | -0.027 |
py | -0.027 |
fly | -0.027 |
ley | -0.026 |
Token | . |
Token ID | 13 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
en | +0.003 |
dot | +0.003 |
el | +0.002 |
fly | +0.002 |
ug | +0.002 |
Neg logit contributions | |
subt | -0.004 |
math | -0.004 |
cream | -0.004 |
topp | -0.003 |
Deal | -0.003 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
el | +0.009 |
dot | +0.009 |
en | +0.009 |
ug | +0.009 |
fly | +0.009 |
Neg logit contributions | |
subt | -0.013 |
cream | -0.012 |
math | -0.012 |
topp | -0.012 |
Deal | -0.012 |
Token | One |
Token ID | 3198 |
Feature activation | -1.95e-01 |
Pos logit contributions | |
math | +0.003 |
cream | +0.002 |
subt | +0.002 |
topp | +0.002 |
Deal | +0.002 |
Neg logit contributions | |
en | -0.004 |
dot | -0.004 |
fly | -0.003 |
ening | -0.003 |
el | -0.003 |
Token | day |
Token ID | 1110 |
Feature activation | +3.47e-01 |
Pos logit contributions | |
el | +0.005 |
ie | +0.005 |
le | +0.005 |
y | +0.005 |
dot | +0.005 |
Neg logit contributions | |
subt | -0.007 |
subtract | -0.006 |
topp | -0.006 |
Deal | -0.006 |
Fif | -0.006 |
Token | , |
Token ID | 11 |
Feature activation | -1.78e-01 |
Pos logit contributions | |
subt | +0.014 |
Deal | +0.014 |
subtract | +0.014 |
math | +0.014 |
topp | +0.014 |
Neg logit contributions | |
el | -0.005 |
en | -0.005 |
dot | -0.005 |
ug | -0.005 |
ir | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
en | +0.006 |
dot | +0.006 |
ening | +0.006 |
ga | +0.006 |
fly | +0.006 |
Neg logit contributions | |
cream | -0.004 |
math | -0.004 |
Deal | -0.004 |
topp | -0.004 |
subt | -0.003 |
Token | little |
Token ID | 1310 |
Feature activation | +2.49e-01 |
Pos logit contributions | |
subt | +0.046 |
subtract | +0.044 |
topp | +0.041 |
math | +0.039 |
Fif | +0.038 |
Neg logit contributions | |
dot | -0.035 |
fly | -0.031 |
ug | -0.030 |
hawk | -0.029 |
le | -0.029 |
Token | boy |
Token ID | 2933 |
Feature activation | -6.03e-01 |
Pos logit contributions | |
subt | +0.007 |
math | +0.006 |
cream | +0.006 |
topp | +0.006 |
Deal | +0.006 |
Neg logit contributions | |
en | -0.008 |
dot | -0.008 |
el | -0.007 |
fly | -0.007 |
ug | -0.007 |
Token | named |
Token ID | 3706 |
Feature activation | -3.89e-01 |
Pos logit contributions | |
en | +0.011 |
dot | +0.011 |
fly | +0.011 |
el | +0.010 |
ug | +0.010 |
Neg logit contributions | |
subt | -0.023 |
math | -0.022 |
cream | -0.022 |
topp | -0.022 |
Deal | -0.022 |
Token | Tim |
Token ID | 5045 |
Feature activation | -1.38e+00 |
Pos logit contributions | |
el | +0.008 |
pt | +0.007 |
dot | +0.007 |
fly | +0.007 |
ug | +0.007 |
Neg logit contributions | |
subt | -0.015 |
subtract | -0.014 |
Deal | -0.014 |
cooperative | -0.013 |
math | -0.013 |
Token | went |
Token ID | 1816 |
Feature activation | -1.56e-01 |
Pos logit contributions | |
dot | +0.043 |
ug | +0.042 |
fly | +0.042 |
ala | +0.042 |
ir | +0.042 |
Neg logit contributions | |
cream | -0.035 |
topp | -0.032 |
Deal | -0.032 |
math | -0.032 |
las | -0.031 |
Token | it |
Token ID | 340 |
Feature activation | -2.42e-01 |
Pos logit contributions | |
subt | +0.013 |
Deal | +0.013 |
cream | +0.013 |
math | +0.013 |
topp | +0.012 |
Neg logit contributions | |
en | -0.013 |
dot | -0.012 |
el | -0.012 |
ug | -0.012 |
le | -0.012 |
Token | look |
Token ID | 804 |
Feature activation | -4.76e-02 |
Pos logit contributions | |
en | +0.007 |
dot | +0.007 |
el | +0.007 |
fly | +0.007 |
le | +0.007 |
Neg logit contributions | |
subt | -0.007 |
Deal | -0.007 |
topp | -0.006 |
cream | -0.006 |
math | -0.006 |
Token | nice |
Token ID | 3621 |
Feature activation | +2.68e-01 |
Pos logit contributions | |
en | +0.002 |
dot | +0.002 |
el | +0.001 |
le | +0.001 |
ening | +0.001 |
Neg logit contributions | |
subt | -0.001 |
Deal | -0.001 |
cream | -0.001 |
math | -0.001 |
topp | -0.001 |
Token | again |
Token ID | 757 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
subt | +0.009 |
math | +0.009 |
cream | +0.009 |
Deal | +0.009 |
topp | +0.009 |
Neg logit contributions | |
en | -0.006 |
dot | -0.006 |
el | -0.006 |
fly | -0.006 |
ug | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | +8.71e-01 |
Pos logit contributions | |
en | +0.002 |
el | +0.002 |
le | +0.002 |
dot | +0.002 |
ening | +0.002 |
Neg logit contributions | |
subt | -0.005 |
Deal | -0.005 |
subtract | -0.005 |
topp | -0.005 |
Fif | -0.005 |
Token | |
Token ID | 220 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
cream | +0.025 |
topp | +0.024 |
subt | +0.024 |
math | +0.024 |
Deal | +0.024 |
Neg logit contributions | |
en | -0.024 |
el | -0.024 |
dot | -0.024 |
ug | -0.023 |
fly | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | +8.91e-01 |
Pos logit contributions | |
subt | +0.059 |
cream | +0.058 |
math | +0.057 |
Deal | +0.057 |
topp | +0.057 |
Neg logit contributions | |
en | -0.052 |
dot | -0.052 |
el | -0.051 |
ug | -0.051 |
fly | -0.050 |
Token | Paul |
Token ID | 12041 |
Feature activation | +6.42e-01 |
Pos logit contributions | |
subt | +0.023 |
cream | +0.023 |
topp | +0.023 |
math | +0.023 |
Deal | +0.022 |
Neg logit contributions | |
en | -0.027 |
dot | -0.027 |
el | -0.026 |
ug | -0.026 |
fly | -0.026 |
Token | was |
Token ID | 373 |
Feature activation | -1.11e-01 |
Pos logit contributions | |
subt | +0.018 |
cream | +0.017 |
topp | +0.017 |
subtract | +0.017 |
math | +0.017 |
Neg logit contributions | |
dot | -0.019 |
ir | -0.018 |
fly | -0.018 |
predators | -0.018 |
py | -0.018 |
Token | very |
Token ID | 845 |
Feature activation | +2.86e-01 |
Pos logit contributions | |
en | +0.003 |
dot | +0.003 |
el | +0.003 |
fly | +0.003 |
ug | +0.003 |
Neg logit contributions | |
subt | -0.003 |
topp | -0.003 |
Deal | -0.003 |
math | -0.003 |
cream | -0.003 |
Token | proud |
Token ID | 6613 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
topp | +0.009 |
subt | +0.009 |
subtract | +0.009 |
las | +0.008 |
Deal | +0.008 |
Neg logit contributions | |
wil | -0.007 |
inkle | -0.006 |
ug | -0.006 |
trapped | -0.006 |
ubby | -0.006 |
Token | € |
Token ID | 26391 |
Feature activation | +5.52e-01 |
Pos logit contributions | |
en | +0.010 |
le | +0.010 |
el | +0.010 |
fly | +0.010 |
ug | +0.010 |
Neg logit contributions | |
math | -0.008 |
topp | -0.008 |
subt | -0.008 |
cream | -0.007 |
subtract | -0.007 |
Token |
|
Token ID | 198 |
Feature activation | +1.21e-02 |
Pos logit contributions | |
subt | +0.015 |
math | +0.015 |
cream | +0.015 |
Deal | +0.015 |
topp | +0.014 |
Neg logit contributions | |
en | -0.016 |
dot | -0.016 |
el | -0.016 |
fly | -0.015 |
ug | -0.015 |
Token |
|
Token ID | 198 |
Feature activation | -2.96e-02 |
Pos logit contributions | |
subt | +0.000 |
math | +0.000 |
cream | +0.000 |
Deal | +0.000 |
subtract | +0.000 |
Neg logit contributions | |
dot | -0.000 |
en | -0.000 |
fly | -0.000 |
ening | -0.000 |
el | -0.000 |
Token | The |
Token ID | 464 |
Feature activation | +7.54e-01 |
Pos logit contributions | |
en | +0.001 |
ug | +0.001 |
el | +0.001 |
dot | +0.001 |
ga | +0.001 |
Neg logit contributions | |
topp | -0.001 |
math | -0.001 |
subt | -0.001 |
cream | -0.001 |
Deal | -0.001 |
Token | man |
Token ID | 582 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
subt | +0.021 |
math | +0.020 |
topp | +0.020 |
cream | +0.020 |
Deal | +0.020 |
Neg logit contributions | |
dot | -0.022 |
en | -0.022 |
el | -0.021 |
fly | -0.021 |
ug | -0.021 |
Token | and |
Token ID | 290 |
Feature activation | +1.93e+00 |
Pos logit contributions | |
cream | +0.030 |
subt | +0.029 |
topp | +0.029 |
Deal | +0.028 |
math | +0.028 |
Neg logit contributions | |
dot | -0.041 |
en | -0.038 |
ir | -0.038 |
ays | -0.038 |
ening | -0.038 |
Token | his |
Token ID | 465 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
topp | +0.061 |
Deal | +0.060 |
subt | +0.060 |
math | +0.059 |
cream | +0.058 |
Neg logit contributions | |
en | -0.049 |
el | -0.048 |
ley | -0.047 |
dot | -0.046 |
ug | -0.046 |
Token | wife |
Token ID | 3656 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
subt | +0.025 |
topp | +0.024 |
Deal | +0.024 |
math | +0.024 |
subtract | +0.023 |
Neg logit contributions | |
en | -0.049 |
dot | -0.048 |
el | -0.048 |
ug | -0.047 |
fly | -0.047 |
Token | agreed |
Token ID | 4987 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
cream | +0.026 |
Deal | +0.025 |
subt | +0.025 |
topp | +0.024 |
cakes | +0.024 |
Neg logit contributions | |
dot | -0.033 |
ir | -0.031 |
ays | -0.031 |
ening | -0.031 |
en | -0.030 |
Token | and |
Token ID | 290 |
Feature activation | +2.73e-01 |
Pos logit contributions | |
dot | +0.032 |
fly | +0.031 |
en | +0.031 |
el | +0.030 |
ir | +0.030 |
Neg logit contributions | |
cream | -0.034 |
math | -0.033 |
subt | -0.033 |
topp | -0.032 |
Deal | -0.032 |
Token | so |
Token ID | 523 |
Feature activation | +5.26e-01 |
Pos logit contributions | |
math | +0.008 |
subt | +0.008 |
Deal | +0.008 |
topp | +0.008 |
cream | +0.008 |
Neg logit contributions | |
en | -0.007 |
el | -0.007 |
ug | -0.007 |
fly | -0.007 |
dot | -0.007 |
Token | their |
Token ID | 511 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
subt | +0.016 |
math | +0.016 |
cream | +0.016 |
Deal | +0.016 |
topp | +0.016 |
Neg logit contributions | |
dot | -0.017 |
en | -0.017 |
fly | -0.017 |
el | -0.017 |
ug | -0.016 |
Token | mom |
Token ID | 1995 |
Feature activation | -5.55e-02 |
Pos logit contributions | |
subt | +0.037 |
Deal | +0.037 |
topp | +0.036 |
subtract | +0.035 |
math | +0.035 |
Neg logit contributions | |
dot | -0.038 |
el | -0.037 |
en | -0.037 |
fly | -0.036 |
wil | -0.035 |
Token | , |
Token ID | 11 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
dot | +0.001 |
ug | +0.001 |
en | +0.001 |
fly | +0.001 |
wil | +0.001 |
Neg logit contributions | |
subt | -0.002 |
math | -0.002 |
cream | -0.002 |
topp | -0.002 |
subtract | -0.002 |
Token | who |
Token ID | 508 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
cream | +0.003 |
math | +0.003 |
Deal | +0.003 |
subt | +0.003 |
cakes | +0.003 |
Neg logit contributions | |
dot | -0.005 |
en | -0.005 |
ug | -0.005 |
ir | -0.005 |
inkle | -0.005 |
Token | is |
Token ID | 318 |
Feature activation | -2.20e-02 |
Pos logit contributions | |
cream | +0.038 |
subt | +0.038 |
math | +0.038 |
Deal | +0.038 |
topp | +0.037 |
Neg logit contributions | |
dot | -0.023 |
en | -0.023 |
fly | -0.022 |
ug | -0.022 |
el | -0.022 |
Token | in |
Token ID | 287 |
Feature activation | +2.49e-01 |
Pos logit contributions | |
dot | +0.001 |
en | +0.001 |
fly | +0.001 |
ley | +0.001 |
ug | +0.001 |
Neg logit contributions | |
cream | -0.001 |
math | -0.001 |
subt | -0.001 |
cakes | -0.001 |
topp | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | +7.23e-01 |
Pos logit contributions | |
math | +0.006 |
subt | +0.006 |
cream | +0.006 |
topp | +0.006 |
subtract | +0.006 |
Neg logit contributions | |
en | -0.008 |
dot | -0.008 |
fly | -0.008 |
ug | -0.008 |
el | -0.008 |
Token | kitchen |
Token ID | 9592 |
Feature activation | +1.35e+00 |
Pos logit contributions | |
subt | +0.022 |
subtract | +0.021 |
topp | +0.021 |
Deal | +0.021 |
Fif | +0.020 |
Neg logit contributions | |
el | -0.019 |
wil | -0.019 |
dot | -0.019 |
wing | -0.017 |
ga | -0.017 |
Token | . |
Token ID | 13 |
Feature activation | -1.86e-01 |
Pos logit contributions | |
cream | +0.041 |
math | +0.041 |
subt | +0.040 |
topp | +0.040 |
Deal | +0.039 |
Neg logit contributions | |
dot | -0.035 |
en | -0.035 |
fly | -0.034 |
ug | -0.034 |
el | -0.034 |
Token |
|
Token ID | 198 |
Feature activation | +4.20e-02 |
Pos logit contributions | |
en | +0.006 |
le | +0.006 |
dot | +0.006 |
ug | +0.006 |
fly | +0.006 |
Neg logit contributions | |
math | -0.004 |
subt | -0.004 |
Deal | -0.004 |
cream | -0.004 |
subtract | -0.004 |
Token |
|
Token ID | 198 |
Feature activation | -3.57e-01 |
Pos logit contributions | |
math | +0.001 |
cream | +0.001 |
subt | +0.001 |
pizza | +0.001 |
subtract | +0.001 |
Neg logit contributions | |
en | -0.002 |
ening | -0.002 |
dot | -0.002 |
fly | -0.002 |
le | -0.002 |
Token | Lucy |
Token ID | 22162 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
en | +0.003 |
dot | +0.003 |
el | +0.003 |
fly | +0.003 |
ir | +0.003 |
Neg logit contributions | |
subt | -0.003 |
topp | -0.003 |
cream | -0.003 |
math | -0.003 |
Deal | -0.003 |
Token | was |
Token ID | 373 |
Feature activation | -3.69e-01 |
Pos logit contributions | |
ug | +0.052 |
dot | +0.051 |
ir | +0.051 |
fly | +0.050 |
ches | +0.050 |
Neg logit contributions | |
cream | -0.038 |
Deal | -0.036 |
las | -0.035 |
topp | -0.034 |
subt | -0.034 |
Token | playing |
Token ID | 2712 |
Feature activation | -2.62e-01 |
Pos logit contributions | |
el | +0.009 |
ug | +0.009 |
en | +0.009 |
dot | +0.009 |
ga | +0.009 |
Neg logit contributions | |
topp | -0.012 |
subt | -0.011 |
Deal | -0.011 |
subtract | -0.011 |
cream | -0.011 |
Token | in |
Token ID | 287 |
Feature activation | -8.59e-01 |
Pos logit contributions | |
dot | +0.004 |
le | +0.004 |
en | +0.004 |
bed | +0.003 |
el | +0.003 |
Neg logit contributions | |
subt | -0.012 |
topp | -0.011 |
subtract | -0.011 |
bek | -0.011 |
hamb | -0.010 |
Token | her |
Token ID | 607 |
Feature activation | -8.18e-01 |
Pos logit contributions | |
en | +0.023 |
el | +0.023 |
dot | +0.023 |
ir | +0.021 |
fly | +0.021 |
Neg logit contributions | |
Deal | -0.026 |
subt | -0.026 |
topp | -0.025 |
cream | -0.024 |
subtract | -0.024 |
Token | yard |
Token ID | 12699 |
Feature activation | -4.55e-01 |
Pos logit contributions | |
dot | +0.020 |
el | +0.020 |
en | +0.018 |
wil | +0.018 |
wing | +0.018 |
Neg logit contributions | |
topp | -0.028 |
subt | -0.028 |
Deal | -0.027 |
subtract | -0.027 |
bek | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
en | +0.011 |
dot | +0.010 |
el | +0.010 |
fly | +0.010 |
ug | +0.010 |
Neg logit contributions | |
subt | -0.015 |
math | -0.015 |
topp | -0.015 |
Deal | -0.015 |
cream | -0.015 |
Token | She |
Token ID | 1375 |
Feature activation | +2.09e-01 |
Pos logit contributions | |
en | +0.003 |
dot | +0.003 |
fly | +0.003 |
ug | +0.003 |
el | +0.003 |
Neg logit contributions | |
math | -0.003 |
subt | -0.003 |
cream | -0.003 |
Deal | -0.003 |
topp | -0.003 |
Token | saw |
Token ID | 2497 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
cream | +0.005 |
Deal | +0.004 |
math | +0.004 |
subt | +0.004 |
topp | +0.004 |
Neg logit contributions | |
ir | -0.008 |
en | -0.008 |
fly | -0.008 |
dot | -0.007 |
ening | -0.007 |
Token | a |
Token ID | 257 |
Feature activation | -3.71e-01 |
Pos logit contributions | |
math | +0.024 |
subt | +0.024 |
cream | +0.024 |
topp | +0.023 |
Deal | +0.023 |
Neg logit contributions | |
en | -0.040 |
dot | -0.040 |
fly | -0.039 |
el | -0.039 |
ug | -0.038 |
Token | small |
Token ID | 1402 |
Feature activation | -7.00e-01 |
Pos logit contributions | |
dot | +0.011 |
fly | +0.010 |
en | +0.010 |
el | +0.010 |
wil | +0.010 |
Neg logit contributions | |
subt | -0.010 |
Deal | -0.010 |
subtract | -0.010 |
topp | -0.010 |
math | -0.010 |
Token | big |
Token ID | 1263 |
Feature activation | -6.27e-01 |
Pos logit contributions | |
en | +0.023 |
el | +0.023 |
dot | +0.023 |
ug | +0.022 |
ir | +0.022 |
Neg logit contributions | |
math | -0.018 |
cream | -0.018 |
subt | -0.018 |
topp | -0.017 |
Deal | -0.017 |
Token | cave |
Token ID | 11527 |
Feature activation | -7.05e-01 |
Pos logit contributions | |
ays | +0.018 |
ala | +0.018 |
ug | +0.018 |
ga | +0.018 |
en | +0.018 |
Neg logit contributions | |
math | -0.018 |
cream | -0.018 |
job | -0.017 |
pizza | -0.017 |
topp | -0.016 |
Token | in |
Token ID | 287 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
en | +0.017 |
dot | +0.016 |
el | +0.016 |
ug | +0.016 |
fly | +0.016 |
Neg logit contributions | |
cream | -0.023 |
math | -0.023 |
subt | -0.023 |
topp | -0.022 |
Deal | -0.022 |
Token | the |
Token ID | 262 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
dot | +0.050 |
ug | +0.050 |
en | +0.050 |
fly | +0.050 |
el | +0.049 |
Neg logit contributions | |
math | -0.039 |
cream | -0.038 |
subt | -0.038 |
topp | -0.037 |
subtract | -0.036 |
Token | distance |
Token ID | 5253 |
Feature activation | +2.31e-01 |
Pos logit contributions | |
en | +0.028 |
dot | +0.027 |
el | +0.027 |
ug | +0.027 |
fly | +0.027 |
Neg logit contributions | |
math | -0.041 |
subt | -0.041 |
cream | -0.041 |
Deal | -0.040 |
topp | -0.040 |
Token | . |
Token ID | 13 |
Feature activation | -4.83e-01 |
Pos logit contributions | |
cream | +0.007 |
math | +0.007 |
subt | +0.007 |
topp | +0.007 |
subtract | +0.007 |
Neg logit contributions | |
dot | -0.006 |
en | -0.006 |
fly | -0.006 |
ug | -0.006 |
el | -0.006 |
Token | He |
Token ID | 679 |
Feature activation | +6.42e-02 |
Pos logit contributions | |
dot | +0.013 |
en | +0.013 |
el | +0.013 |
fly | +0.012 |
ug | +0.012 |
Neg logit contributions | |
subt | -0.014 |
cream | -0.014 |
Deal | -0.014 |
math | -0.014 |
topp | -0.014 |
Token | became |
Token ID | 2627 |
Feature activation | -1.88e-01 |
Pos logit contributions | |
subt | +0.002 |
cream | +0.002 |
math | +0.002 |
Deal | +0.002 |
topp | +0.002 |
Neg logit contributions | |
dot | -0.002 |
en | -0.002 |
el | -0.002 |
ir | -0.002 |
fly | -0.002 |
Token | curious |
Token ID | 11040 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
dot | +0.004 |
el | +0.004 |
ug | +0.004 |
fly | +0.004 |
wil | +0.004 |
Neg logit contributions | |
subt | -0.006 |
topp | -0.006 |
subtract | -0.006 |
cream | -0.006 |
Deal | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | +3.27e-02 |
Pos logit contributions | |
math | +0.057 |
cream | +0.054 |
topp | +0.054 |
pizza | +0.052 |
cakes | +0.052 |
Neg logit contributions | |
inkle | -0.049 |
fly | -0.049 |
ir | -0.049 |
ga | -0.048 |
ink | -0.046 |
Token | wanted |
Token ID | 2227 |
Feature activation | -7.03e-01 |
Pos logit contributions | |
subt | +0.001 |
topp | +0.001 |
math | +0.001 |
cream | +0.001 |
subtract | +0.001 |
Neg logit contributions | |
en | -0.001 |
ie | -0.001 |
ley | -0.001 |
landed | -0.001 |
el | -0.001 |
Token | castle |
Token ID | 16669 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
dot | +0.003 |
fly | +0.003 |
en | +0.003 |
el | +0.003 |
wil | +0.003 |
Neg logit contributions | |
subt | -0.004 |
Deal | -0.004 |
subtract | -0.004 |
topp | -0.004 |
cream | -0.004 |
Token | and |
Token ID | 290 |
Feature activation | -3.62e-01 |
Pos logit contributions | |
subt | +0.049 |
math | +0.048 |
cream | +0.048 |
topp | +0.048 |
Deal | +0.048 |
Neg logit contributions | |
dot | -0.023 |
en | -0.023 |
el | -0.022 |
fly | -0.021 |
ug | -0.021 |
Token | told |
Token ID | 1297 |
Feature activation | -5.00e-01 |
Pos logit contributions | |
fly | +0.009 |
en | +0.009 |
dot | +0.009 |
el | +0.008 |
ley | +0.008 |
Neg logit contributions | |
subt | -0.012 |
math | -0.011 |
Deal | -0.011 |
topp | -0.011 |
cream | -0.011 |
Token | the |
Token ID | 262 |
Feature activation | -1.20e-02 |
Pos logit contributions | |
en | +0.011 |
el | +0.011 |
dot | +0.011 |
le | +0.010 |
ir | +0.010 |
Neg logit contributions | |
subt | -0.017 |
Deal | -0.017 |
cream | -0.017 |
topp | -0.017 |
subtract | -0.016 |
Token | knight |
Token ID | 22062 |
Feature activation | +2.17e-01 |
Pos logit contributions | |
dot | +0.000 |
en | +0.000 |
fly | +0.000 |
py | +0.000 |
le | +0.000 |
Neg logit contributions | |
subt | -0.000 |
Deal | -0.000 |
cream | -0.000 |
topp | -0.000 |
subtract | -0.000 |
Token | that |
Token ID | 326 |
Feature activation | +8.27e-01 |
Pos logit contributions | |
subt | +0.006 |
math | +0.006 |
cream | +0.006 |
Deal | +0.006 |
topp | +0.006 |
Neg logit contributions | |
en | -0.006 |
dot | -0.006 |
el | -0.006 |
fly | -0.006 |
ug | -0.006 |
Token | he |
Token ID | 339 |
Feature activation | -1.94e-01 |
Pos logit contributions | |
subt | +0.021 |
Deal | +0.021 |
cream | +0.020 |
topp | +0.020 |
math | +0.020 |
Neg logit contributions | |
en | -0.025 |
dot | -0.025 |
fly | -0.024 |
el | -0.024 |
ug | -0.024 |
Token | needed |
Token ID | 2622 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
ir | +0.007 |
ays | +0.007 |
dot | +0.006 |
ga | +0.006 |
en | +0.006 |
Neg logit contributions | |
Deal | -0.004 |
topp | -0.004 |
math | -0.004 |
cream | -0.004 |
y | -0.004 |
Token | to |
Token ID | 284 |
Feature activation | +9.03e-01 |
Pos logit contributions | |
en | +0.039 |
dot | +0.039 |
fly | +0.038 |
el | +0.038 |
ug | +0.037 |
Neg logit contributions | |
math | -0.020 |
subt | -0.020 |
cream | -0.020 |
topp | -0.019 |
subtract | -0.019 |
Token | prepare |
Token ID | 8335 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
subt | +0.020 |
math | +0.019 |
cream | +0.019 |
Deal | +0.019 |
topp | +0.019 |
Neg logit contributions | |
en | -0.031 |
dot | -0.031 |
el | -0.030 |
fly | -0.030 |
ug | -0.030 |
Token | better |
Token ID | 1365 |
Feature activation | -7.46e-02 |
Pos logit contributions | |
subt | +0.003 |
math | +0.003 |
cream | +0.003 |
topp | +0.003 |
Deal | +0.003 |
Neg logit contributions | |
dot | -0.004 |
en | -0.004 |
fly | -0.004 |
el | -0.004 |
ug | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -4.91e-01 |
Pos logit contributions | |
subt | +0.007 |
math | +0.007 |
cream | +0.007 |
topp | +0.007 |
Deal | +0.007 |
Neg logit contributions | |
en | -0.005 |
dot | -0.005 |
el | -0.004 |
fly | -0.004 |
ug | -0.004 |
Token |
|
Token ID | 198 |
Feature activation | -2.15e-01 |
Pos logit contributions | |
dot | +0.012 |
el | +0.011 |
en | +0.011 |
fly | +0.011 |
ug | +0.011 |
Neg logit contributions | |
cream | -0.016 |
subt | -0.016 |
topp | -0.015 |
Deal | -0.015 |
math | -0.015 |
Token |
|
Token ID | 198 |
Feature activation | -6.29e-01 |
Pos logit contributions | |
en | +0.008 |
dot | +0.008 |
ening | +0.008 |
fly | +0.008 |
ley | +0.008 |
Neg logit contributions | |
math | -0.004 |
cream | -0.004 |
subt | -0.004 |
Deal | -0.004 |
subtract | -0.004 |
Token | " |
Token ID | 1 |
Feature activation | -1.52e-01 |
Pos logit contributions | |
ir | +0.016 |
el | +0.016 |
ug | +0.016 |
en | +0.016 |
dot | +0.015 |
Neg logit contributions | |
topp | -0.019 |
subt | -0.018 |
math | -0.018 |
cream | -0.018 |
subtract | -0.018 |
Token | Can |
Token ID | 6090 |
Feature activation | -4.29e-01 |
Pos logit contributions | |
dot | +0.004 |
le | +0.004 |
ga | +0.004 |
el | +0.004 |
wil | +0.004 |
Neg logit contributions | |
math | -0.004 |
job | -0.004 |
list | -0.004 |
subt | -0.004 |
topp | -0.004 |
Token | we |
Token ID | 356 |
Feature activation | -5.13e-01 |
Pos logit contributions | |
en | +0.013 |
dot | +0.012 |
ening | +0.012 |
fly | +0.012 |
el | +0.012 |
Neg logit contributions | |
math | -0.012 |
subt | -0.012 |
Deal | -0.012 |
cream | -0.012 |
topp | -0.011 |
Token | still |
Token ID | 991 |
Feature activation | -5.80e-01 |
Pos logit contributions | |
en | +0.017 |
dot | +0.017 |
el | +0.017 |
fly | +0.016 |
ug | +0.016 |
Neg logit contributions | |
subt | -0.012 |
topp | -0.012 |
cream | -0.012 |
math | -0.012 |
Deal | -0.012 |
Token | play |
Token ID | 711 |
Feature activation | -3.51e-01 |
Pos logit contributions | |
ening | +0.021 |
ley | +0.021 |
ches | +0.021 |
dot | +0.021 |
en | +0.021 |
Neg logit contributions | |
math | -0.011 |
cream | -0.011 |
subtract | -0.011 |
subt | -0.011 |
Deal | -0.011 |
Token | with |
Token ID | 351 |
Feature activation | +3.94e-01 |
Pos logit contributions | |
en | +0.010 |
dot | +0.010 |
el | +0.010 |
fly | +0.010 |
ening | +0.010 |
Neg logit contributions | |
math | -0.010 |
subt | -0.009 |
cream | -0.009 |
Deal | -0.009 |
topp | -0.009 |
Token | the |
Token ID | 262 |
Feature activation | -8.19e-01 |
Pos logit contributions | |
math | +0.011 |
cream | +0.010 |
topp | +0.010 |
cakes | +0.010 |
pizza | +0.010 |
Neg logit contributions | |
fly | -0.013 |
dot | -0.013 |
py | -0.012 |
en | -0.012 |
ga | -0.012 |
Token | blocks |
Token ID | 7021 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
dot | +0.025 |
en | +0.024 |
el | +0.024 |
fly | +0.023 |
ening | +0.023 |
Neg logit contributions | |
subt | -0.022 |
Deal | -0.021 |
subtract | -0.020 |
topp | -0.020 |
cream | -0.020 |
Token | girl |
Token ID | 2576 |
Feature activation | +6.45e-01 |
Pos logit contributions | |
subt | +0.037 |
math | +0.036 |
cream | +0.035 |
Deal | +0.035 |
topp | +0.035 |
Neg logit contributions | |
en | -0.046 |
dot | -0.046 |
el | -0.045 |
fly | -0.045 |
ug | -0.044 |
Token | trusted |
Token ID | 13467 |
Feature activation | -9.41e-01 |
Pos logit contributions | |
cream | +0.012 |
Deal | +0.012 |
topp | +0.012 |
hamb | +0.011 |
y | +0.011 |
Neg logit contributions | |
ir | -0.024 |
ays | -0.024 |
dot | -0.024 |
ening | -0.023 |
ala | -0.023 |
Token | the |
Token ID | 262 |
Feature activation | +5.51e-01 |
Pos logit contributions | |
en | +0.026 |
el | +0.025 |
fly | +0.024 |
dot | +0.024 |
ir | +0.024 |
Neg logit contributions | |
subt | -0.028 |
topp | -0.027 |
Deal | -0.027 |
cream | -0.027 |
math | -0.027 |
Token | sailor |
Token ID | 43272 |
Feature activation | +7.17e-01 |
Pos logit contributions | |
subt | +0.017 |
math | +0.017 |
Deal | +0.017 |
topp | +0.017 |
cream | +0.017 |
Neg logit contributions | |
dot | -0.014 |
en | -0.014 |
el | -0.014 |
fly | -0.013 |
ug | -0.013 |
Token | and |
Token ID | 290 |
Feature activation | +1.90e-01 |
Pos logit contributions | |
math | +0.020 |
cream | +0.020 |
job | +0.019 |
pizza | +0.019 |
Deal | +0.019 |
Neg logit contributions | |
dot | -0.021 |
ays | -0.020 |
ir | -0.020 |
el | -0.019 |
ug | -0.019 |
Token | felt |
Token ID | 2936 |
Feature activation | -2.00e+00 |
Pos logit contributions | |
subt | +0.006 |
math | +0.006 |
topp | +0.006 |
subtract | +0.006 |
Deal | +0.006 |
Neg logit contributions | |
en | -0.005 |
fly | -0.005 |
ley | -0.004 |
dot | -0.004 |
el | -0.004 |
Token | safe |
Token ID | 3338 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
en | +0.067 |
dot | +0.066 |
el | +0.065 |
fly | +0.065 |
ug | +0.064 |
Neg logit contributions | |
subt | -0.046 |
math | -0.045 |
cream | -0.045 |
Deal | -0.045 |
topp | -0.044 |
Token | with |
Token ID | 351 |
Feature activation | +1.90e-01 |
Pos logit contributions | |
dot | +0.035 |
en | +0.035 |
el | +0.034 |
ir | +0.034 |
ug | +0.033 |
Neg logit contributions | |
cream | -0.042 |
math | -0.041 |
subt | -0.040 |
Deal | -0.040 |
topp | -0.040 |
Token | him |
Token ID | 683 |
Feature activation | +3.69e-01 |
Pos logit contributions | |
math | +0.004 |
cream | +0.004 |
subt | +0.004 |
topp | +0.004 |
Deal | +0.004 |
Neg logit contributions | |
dot | -0.007 |
en | -0.006 |
fly | -0.006 |
ug | -0.006 |
el | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | +3.01e-01 |
Pos logit contributions | |
math | +0.010 |
cream | +0.010 |
Deal | +0.010 |
topp | +0.010 |
subt | +0.010 |
Neg logit contributions | |
en | -0.010 |
dot | -0.010 |
el | -0.010 |
fly | -0.010 |
ir | -0.010 |
Token | They |
Token ID | 1119 |
Feature activation | +2.31e-01 |
Pos logit contributions | |
subt | +0.009 |
math | +0.009 |
cream | +0.009 |
Deal | +0.009 |
topp | +0.009 |
Neg logit contributions | |
en | -0.008 |
dot | -0.007 |
el | -0.007 |
fly | -0.007 |
ug | -0.007 |
Token | she |
Token ID | 673 |
Feature activation | -6.35e-01 |
Pos logit contributions | |
en | +0.002 |
dot | +0.002 |
fly | +0.001 |
el | +0.001 |
le | +0.001 |
Neg logit contributions | |
math | -0.001 |
subt | -0.001 |
Deal | -0.001 |
cream | -0.001 |
subtract | -0.001 |
Token | wouldn |
Token ID | 3636 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
ir | +0.023 |
ays | +0.022 |
dot | +0.022 |
ening | +0.022 |
el | +0.022 |
Neg logit contributions | |
cream | -0.015 |
y | -0.015 |
Deal | -0.014 |
topp | -0.013 |
treat | -0.013 |
Token | 't |
Token ID | 470 |
Feature activation | -3.59e-01 |
Pos logit contributions | |
ening | +0.070 |
ley | +0.067 |
fly | +0.067 |
wil | +0.066 |
maid | +0.065 |
Neg logit contributions | |
pay | -0.032 |
earn | -0.031 |
math | -0.030 |
job | -0.029 |
subtract | -0.028 |
Token | scare |
Token ID | 19437 |
Feature activation | +2.73e-01 |
Pos logit contributions | |
ays | +0.014 |
el | +0.013 |
ening | +0.013 |
ink | +0.013 |
ug | +0.013 |
Neg logit contributions | |
pay | -0.006 |
listen | -0.006 |
job | -0.006 |
cream | -0.006 |
treat | -0.006 |
Token | it |
Token ID | 340 |
Feature activation | -4.80e-01 |
Pos logit contributions | |
math | +0.007 |
cream | +0.007 |
job | +0.006 |
subt | +0.006 |
Fif | +0.006 |
Neg logit contributions | |
dot | -0.008 |
inkle | -0.008 |
ink | -0.008 |
fly | -0.008 |
ittered | -0.008 |
Token | away |
Token ID | 1497 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
ink | +0.015 |
ley | +0.015 |
dot | +0.015 |
inkle | +0.015 |
ug | +0.015 |
Neg logit contributions | |
math | -0.011 |
cream | -0.011 |
job | -0.010 |
y | -0.010 |
subtract | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | -3.52e-01 |
Pos logit contributions | |
dot | +0.047 |
en | +0.046 |
el | +0.046 |
ug | +0.045 |
ir | +0.045 |
Neg logit contributions | |
math | -0.048 |
cream | -0.047 |
subt | -0.046 |
Deal | -0.044 |
topp | -0.044 |
Token | She |
Token ID | 1375 |
Feature activation | +4.24e-02 |
Pos logit contributions | |
en | +0.009 |
el | +0.009 |
dot | +0.009 |
fly | +0.009 |
ir | +0.009 |
Neg logit contributions | |
subt | -0.011 |
cream | -0.011 |
topp | -0.010 |
math | -0.010 |
Deal | -0.010 |
Token | saw |
Token ID | 2497 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
cream | +0.001 |
topp | +0.001 |
Deal | +0.001 |
y | +0.001 |
subt | +0.001 |
Neg logit contributions | |
ays | -0.002 |
dot | -0.002 |
ening | -0.002 |
ir | -0.001 |
inkle | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | -6.43e-01 |
Pos logit contributions | |
math | +0.020 |
cream | +0.019 |
topp | +0.017 |
cakes | +0.017 |
subt | +0.017 |
Neg logit contributions | |
inkle | -0.041 |
en | -0.041 |
ug | -0.040 |
dot | -0.040 |
ches | -0.040 |
Token | bird |
Token ID | 6512 |
Feature activation | +5.29e-01 |
Pos logit contributions | |
en | +0.018 |
ug | +0.018 |
ches | +0.018 |
ors | +0.017 |
ley | +0.017 |
Neg logit contributions | |
cream | -0.020 |
math | -0.019 |
topp | -0.019 |
job | -0.018 |
cakes | -0.018 |
Token | them |
Token ID | 606 |
Feature activation | +8.87e-01 |
Pos logit contributions | |
en | +0.002 |
dot | +0.002 |
el | +0.002 |
fly | +0.002 |
ug | +0.002 |
Neg logit contributions | |
subt | -0.003 |
math | -0.003 |
cream | -0.003 |
topp | -0.003 |
Deal | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | +4.86e-02 |
Pos logit contributions | |
math | +0.024 |
cream | +0.024 |
subt | +0.023 |
Deal | +0.023 |
topp | +0.023 |
Neg logit contributions | |
dot | -0.026 |
en | -0.026 |
fly | -0.025 |
ug | -0.025 |
inkle | -0.025 |
Token | She |
Token ID | 1375 |
Feature activation | -7.04e-01 |
Pos logit contributions | |
subt | +0.002 |
math | +0.002 |
cream | +0.002 |
topp | +0.002 |
Deal | +0.002 |
Neg logit contributions | |
en | -0.001 |
dot | -0.001 |
el | -0.001 |
fly | -0.001 |
ug | -0.001 |
Token | wished |
Token ID | 16555 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
dot | +0.021 |
en | +0.021 |
el | +0.021 |
ug | +0.021 |
fly | +0.020 |
Neg logit contributions | |
subt | -0.019 |
cream | -0.019 |
Deal | -0.018 |
math | -0.018 |
topp | -0.018 |
Token | she |
Token ID | 673 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
dot | +0.042 |
en | +0.041 |
fly | +0.041 |
el | +0.040 |
ug | +0.040 |
Neg logit contributions | |
math | -0.029 |
subt | -0.029 |
cream | -0.028 |
Deal | -0.027 |
subtract | -0.027 |
Token | could |
Token ID | 714 |
Feature activation | -1.74e+00 |
Pos logit contributions | |
dot | +0.037 |
en | +0.037 |
ir | +0.037 |
ening | +0.035 |
el | +0.035 |
Neg logit contributions | |
Deal | -0.043 |
cream | -0.042 |
subt | -0.039 |
math | -0.039 |
topp | -0.039 |
Token | let |
Token ID | 1309 |
Feature activation | -4.58e-01 |
Pos logit contributions | |
ening | +0.064 |
dot | +0.063 |
en | +0.063 |
el | +0.061 |
ug | +0.061 |
Neg logit contributions | |
cream | -0.037 |
Deal | -0.036 |
math | -0.035 |
subt | -0.033 |
job | -0.033 |
Token | them |
Token ID | 606 |
Feature activation | +1.61e-01 |
Pos logit contributions | |
en | +0.009 |
dot | +0.008 |
el | +0.008 |
fly | +0.008 |
ug | +0.008 |
Neg logit contributions | |
subt | -0.017 |
math | -0.017 |
cream | -0.017 |
Deal | -0.017 |
topp | -0.017 |
Token | go |
Token ID | 467 |
Feature activation | -5.67e-01 |
Pos logit contributions | |
subt | +0.005 |
topp | +0.004 |
math | +0.004 |
Deal | +0.004 |
subtract | +0.004 |
Neg logit contributions | |
en | -0.005 |
dot | -0.005 |
el | -0.005 |
le | -0.005 |
fly | -0.005 |
Token | home |
Token ID | 1363 |
Feature activation | +1.85e-01 |
Pos logit contributions | |
dot | +0.020 |
ening | +0.020 |
fly | +0.019 |
wil | +0.019 |
inkle | +0.019 |
Neg logit contributions | |
cream | -0.013 |
math | -0.012 |
job | -0.011 |
pay | -0.011 |
pizza | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
subt | +0.006 |
math | +0.006 |
cream | +0.006 |
Deal | +0.006 |
topp | +0.005 |
Neg logit contributions | |
dot | -0.005 |
en | -0.005 |
el | -0.005 |
fly | -0.005 |
ug | -0.004 |
Token | help |
Token ID | 1037 |
Feature activation | -6.55e-01 |
Pos logit contributions | |
dot | +0.015 |
ening | +0.014 |
en | +0.014 |
el | +0.014 |
fly | +0.014 |
Neg logit contributions | |
cream | -0.010 |
math | -0.010 |
subt | -0.009 |
Deal | -0.009 |
topp | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
dot | +0.020 |
fly | +0.019 |
en | +0.019 |
wil | +0.019 |
ening | +0.019 |
Neg logit contributions | |
math | -0.018 |
cream | -0.017 |
subt | -0.017 |
subtract | -0.017 |
topp | -0.017 |
Token | He |
Token ID | 679 |
Feature activation | +6.37e-02 |
Pos logit contributions | |
en | +0.008 |
dot | +0.008 |
el | +0.008 |
ug | +0.008 |
fly | +0.008 |
Neg logit contributions | |
cream | -0.010 |
Deal | -0.010 |
subt | -0.010 |
topp | -0.010 |
math | -0.010 |
Token | found |
Token ID | 1043 |
Feature activation | +4.81e-02 |
Pos logit contributions | |
cream | +0.002 |
subt | +0.002 |
topp | +0.002 |
Deal | +0.001 |
math | +0.001 |
Neg logit contributions | |
dot | -0.002 |
ir | -0.002 |
el | -0.002 |
en | -0.002 |
ening | -0.002 |
Token | a |
Token ID | 257 |
Feature activation | -9.16e-01 |
Pos logit contributions | |
cream | +0.001 |
math | +0.001 |
topp | +0.001 |
subt | +0.001 |
subtract | +0.001 |
Neg logit contributions | |
dot | -0.002 |
en | -0.002 |
fly | -0.002 |
el | -0.002 |
ir | -0.002 |
Token | way |
Token ID | 835 |
Feature activation | -2.50e+00 |
Pos logit contributions | |
en | +0.028 |
dot | +0.028 |
el | +0.028 |
fly | +0.027 |
ug | +0.027 |
Neg logit contributions | |
subt | -0.023 |
math | -0.023 |
cream | -0.023 |
topp | -0.022 |
Deal | -0.022 |
Token | to |
Token ID | 284 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
inkle | +0.111 |
ink | +0.108 |
le | +0.107 |
dot | +0.105 |
fly | +0.103 |
Neg logit contributions | |
math | -0.035 |
cream | -0.032 |
job | -0.028 |
pizza | -0.025 |
subtract | -0.023 |
Token | clean |
Token ID | 3424 |
Feature activation | -5.44e-01 |
Pos logit contributions | |
en | +0.005 |
dot | +0.005 |
el | +0.005 |
fly | +0.004 |
ug | +0.004 |
Neg logit contributions | |
subt | -0.003 |
math | -0.003 |
cream | -0.003 |
Deal | -0.003 |
topp | -0.003 |
Token | the |
Token ID | 262 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
fly | +0.016 |
dot | +0.016 |
py | +0.015 |
ug | +0.015 |
wil | +0.015 |
Neg logit contributions | |
cream | -0.017 |
math | -0.016 |
topp | -0.015 |
cakes | -0.015 |
pizza | -0.015 |
Token | p |
Token ID | 279 |
Feature activation | +1.90e-01 |
Pos logit contributions | |
ug | +0.033 |
en | +0.033 |
fly | +0.033 |
ley | +0.032 |
el | +0.032 |
Neg logit contributions | |
cream | -0.026 |
math | -0.025 |
topp | -0.024 |
subt | -0.024 |
cakes | -0.023 |
Token | uddle |
Token ID | 24500 |
Feature activation | +2.26e-02 |
Pos logit contributions | |
cream | +0.006 |
math | +0.005 |
subt | +0.005 |
Deal | +0.005 |
topp | +0.005 |
Neg logit contributions | |
dot | -0.005 |
en | -0.005 |
fly | -0.005 |
ening | -0.005 |
le | -0.005 |
Token | it |
Token ID | 340 |
Feature activation | -5.87e-01 |
Pos logit contributions | |
fly | +0.004 |
en | +0.004 |
ley | +0.004 |
ga | +0.004 |
dot | +0.004 |
Neg logit contributions | |
math | -0.005 |
cream | -0.005 |
cakes | -0.005 |
subt | -0.004 |
topp | -0.004 |
Token | was |
Token ID | 373 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
ir | +0.017 |
ug | +0.017 |
en | +0.017 |
dot | +0.017 |
ening | +0.016 |
Neg logit contributions | |
cream | -0.016 |
Deal | -0.016 |
subt | -0.015 |
math | -0.015 |
topp | -0.015 |
Token | very |
Token ID | 845 |
Feature activation | -4.83e-01 |
Pos logit contributions | |
dot | +0.048 |
le | +0.047 |
ug | +0.047 |
fly | +0.047 |
en | +0.046 |
Neg logit contributions | |
cream | -0.027 |
math | -0.026 |
y | -0.025 |
cakes | -0.023 |
topp | -0.022 |
Token | soft |
Token ID | 2705 |
Feature activation | -7.50e-01 |
Pos logit contributions | |
el | +0.014 |
dot | +0.014 |
wil | +0.014 |
en | +0.014 |
ug | +0.014 |
Neg logit contributions | |
subt | -0.013 |
subtract | -0.012 |
topp | -0.012 |
math | -0.012 |
Deal | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
ir | +0.020 |
ug | +0.019 |
inkle | +0.019 |
fly | +0.019 |
ala | +0.018 |
Neg logit contributions | |
cream | -0.025 |
math | -0.024 |
pizza | -0.023 |
cakes | -0.023 |
topp | -0.022 |
Token | fluffy |
Token ID | 39145 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
en | +0.005 |
dot | +0.005 |
el | +0.005 |
fly | +0.005 |
ug | +0.005 |
Neg logit contributions | |
subt | -0.004 |
cream | -0.004 |
math | -0.004 |
Deal | -0.004 |
topp | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | +3.83e-01 |
Pos logit contributions | |
ug | +0.060 |
ir | +0.059 |
inkle | +0.058 |
fly | +0.056 |
ala | +0.055 |
Neg logit contributions | |
cream | -0.035 |
math | -0.033 |
pizza | -0.032 |
dessert | -0.031 |
cakes | -0.031 |
Token | |
Token ID | 220 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
subt | +0.011 |
cream | +0.011 |
math | +0.011 |
topp | +0.011 |
Deal | +0.011 |
Neg logit contributions | |
en | -0.010 |
dot | -0.010 |
el | -0.010 |
fly | -0.010 |
ug | -0.010 |
Token |
|
Token ID | 198 |
Feature activation | +3.51e-01 |
Pos logit contributions | |
math | +0.025 |
Deal | +0.024 |
cream | +0.023 |
subt | +0.023 |
subtract | +0.023 |
Neg logit contributions | |
en | -0.046 |
dot | -0.046 |
ening | -0.045 |
fly | -0.045 |
ley | -0.044 |
Token | Every |
Token ID | 6109 |
Feature activation | +2.63e-01 |
Pos logit contributions | |
math | +0.007 |
Deal | +0.007 |
subt | +0.006 |
cream | +0.006 |
subtract | +0.006 |
Neg logit contributions | |
dot | -0.013 |
fly | -0.013 |
ening | -0.013 |
wil | -0.013 |
en | -0.013 |
Token | day |
Token ID | 1110 |
Feature activation | +1.48e-01 |
Pos logit contributions | |
math | +0.004 |
cream | +0.004 |
Deal | +0.004 |
job | +0.004 |
subt | +0.004 |
Neg logit contributions | |
en | -0.011 |
dot | -0.010 |
ening | -0.010 |
ug | -0.010 |
fly | -0.010 |