Token | Soon |
Token ID | 15894 |
Feature activation | -1.67e-02 |
Pos logit contributions | |
goodbye | +0.005 |
Amen | +0.004 |
yes | +0.004 |
| +0.004 |
fs | +0.004 |
Neg logit contributions | |
udge | -0.006 |
ipples | -0.005 |
eter | -0.005 |
ill | -0.005 |
elly | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -1.10e-02 |
Pos logit contributions | |
goodbye | +0.021 |
fs | +0.013 |
yes | +0.013 |
Amen | +0.013 |
medal | +0.012 |
Neg logit contributions | |
udge | -0.048 |
eter | -0.043 |
ipples | -0.043 |
urious | -0.042 |
elly | -0.041 |
Token | leaves |
Token ID | 5667 |
Feature activation | -7.85e-03 |
Pos logit contributions | |
goodbye | +0.022 |
fs | +0.018 |
Amen | +0.018 |
yes | +0.017 |
saddle | +0.017 |
Neg logit contributions | |
udge | -0.022 |
eter | -0.020 |
ipples | -0.019 |
ill | -0.018 |
urious | -0.018 |
Token | were |
Token ID | 547 |
Feature activation | -9.57e-03 |
Pos logit contributions | |
goodbye | +0.015 |
Amen | +0.013 |
fs | +0.012 |
forehead | +0.012 |
saddle | +0.012 |
Neg logit contributions | |
udge | -0.016 |
ipples | -0.014 |
eter | -0.014 |
ill | -0.014 |
ased | -0.013 |
Token | all |
Token ID | 477 |
Feature activation | +1.98e-02 |
Pos logit contributions | |
goodbye | +0.018 |
Amen | +0.015 |
fs | +0.015 |
forehead | +0.013 |
saddle | +0.013 |
Neg logit contributions | |
udge | -0.020 |
eter | -0.018 |
ipples | -0.018 |
ill | -0.017 |
ased | -0.017 |
Token | r |
Token ID | 374 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
udge | +0.041 |
eter | +0.036 |
ipples | +0.035 |
ill | +0.034 |
urious | +0.033 |
Neg logit contributions | |
goodbye | -0.039 |
fs | -0.032 |
Amen | -0.032 |
yes | -0.031 |
saddle | -0.030 |
Token | aked |
Token ID | 4335 |
Feature activation | +8.50e-03 |
Pos logit contributions | |
udge | +0.142 |
Like | +0.131 |
Who | +0.129 |
Who | +0.125 |
urious | +0.123 |
Neg logit contributions | |
fs | -0.182 |
goodbye | -0.174 |
agging | -0.173 |
ons | -0.171 |
on | -0.170 |
Token | up |
Token ID | 510 |
Feature activation | +4.05e-02 |
Pos logit contributions | |
udge | +0.021 |
ipples | +0.019 |
eter | +0.019 |
urious | +0.019 |
elly | +0.018 |
Neg logit contributions | |
goodbye | -0.013 |
fs | -0.009 |
yes | -0.009 |
saddle | -0.009 |
forehead | -0.008 |
Token | and |
Token ID | 290 |
Feature activation | -2.80e-02 |
Pos logit contributions | |
udge | +0.103 |
eter | +0.093 |
ipples | +0.092 |
urious | +0.091 |
elly | +0.088 |
Neg logit contributions | |
goodbye | -0.059 |
fs | -0.042 |
yes | -0.041 |
| -0.039 |
blanket | -0.038 |
Token | Bob |
Token ID | 5811 |
Feature activation | +2.24e-03 |
Pos logit contributions | |
goodbye | +0.050 |
fs | +0.039 |
Amen | +0.039 |
yes | +0.037 |
saddle | +0.036 |
Neg logit contributions | |
udge | -0.064 |
eter | -0.056 |
ipples | -0.056 |
ill | -0.053 |
urious | -0.053 |
Token | was |
Token ID | 373 |
Feature activation | -1.46e-02 |
Pos logit contributions | |
udge | +0.005 |
eter | +0.004 |
ipples | +0.004 |
urious | +0.004 |
elly | +0.004 |
Neg logit contributions | |
goodbye | -0.004 |
fs | -0.003 |
yes | -0.003 |
Amen | -0.003 |
saddle | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | +1.02e-03 |
Pos logit contributions | |
goodbye | +0.026 |
fs | +0.023 |
pron | +0.022 |
forehead | +0.022 |
Amen | +0.022 |
Neg logit contributions | |
udge | -0.021 |
ipples | -0.018 |
elly | -0.017 |
ill | -0.016 |
eter | -0.016 |
Token | One |
Token ID | 1881 |
Feature activation | +3.55e-02 |
Pos logit contributions | |
udge | +0.002 |
eter | +0.002 |
ipples | +0.002 |
urious | +0.002 |
ased | +0.002 |
Neg logit contributions | |
goodbye | -0.002 |
Amen | -0.001 |
yes | -0.001 |
fs | -0.001 |
| -0.001 |
Token | day |
Token ID | 1110 |
Feature activation | -1.23e-02 |
Pos logit contributions | |
udge | +0.100 |
eter | +0.095 |
ased | +0.089 |
ill | +0.086 |
Like | +0.085 |
Neg logit contributions | |
goodbye | -0.045 |
yes | -0.026 |
Amen | -0.026 |
medal | -0.026 |
blanket | -0.025 |
Token | , |
Token ID | 11 |
Feature activation | +1.03e-02 |
Pos logit contributions | |
goodbye | +0.017 |
Amen | +0.012 |
fs | +0.012 |
forehead | +0.011 |
saddle | +0.011 |
Neg logit contributions | |
udge | -0.033 |
eter | -0.029 |
ipples | -0.029 |
ill | -0.028 |
elly | -0.028 |
Token | Cr |
Token ID | 3864 |
Feature activation | +6.02e-02 |
Pos logit contributions | |
udge | +0.021 |
eter | +0.018 |
ipples | +0.018 |
ill | +0.017 |
elly | +0.017 |
Neg logit contributions | |
goodbye | -0.020 |
Amen | -0.017 |
fs | -0.017 |
forehead | -0.016 |
pron | -0.016 |
Token | ust |
Token ID | 436 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
udge | +0.068 |
Esk | +0.064 |
ased | +0.062 |
eter | +0.056 |
ipples | +0.055 |
Neg logit contributions | |
goodbye | -0.131 |
magazine | -0.119 |
fs | -0.117 |
medal | -0.115 |
isy | -0.113 |
Token | y |
Token ID | 88 |
Feature activation | +1.02e-02 |
Pos logit contributions | |
udge | +0.196 |
eter | +0.168 |
ierce | +0.150 |
aught | +0.148 |
icy | +0.144 |
Neg logit contributions | |
goodbye | -0.231 |
medal | -0.180 |
helicopter | -0.174 |
Amen | -0.174 |
off | -0.171 |
Token | found |
Token ID | 1043 |
Feature activation | +2.19e-02 |
Pos logit contributions | |
udge | +0.021 |
eter | +0.019 |
ipples | +0.018 |
ased | +0.017 |
ill | +0.017 |
Neg logit contributions | |
goodbye | -0.020 |
fs | -0.015 |
Amen | -0.015 |
yes | -0.015 |
medal | -0.015 |
Token | a |
Token ID | 257 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
udge | +0.052 |
eter | +0.046 |
ipples | +0.045 |
ill | +0.044 |
urious | +0.043 |
Neg logit contributions | |
goodbye | -0.037 |
fs | -0.029 |
Amen | -0.029 |
yes | -0.027 |
saddle | -0.027 |
Token | shiny |
Token ID | 22441 |
Feature activation | -2.35e-03 |
Pos logit contributions | |
goodbye | +0.033 |
fs | +0.027 |
Amen | +0.027 |
ishly | +0.025 |
forehead | +0.025 |
Neg logit contributions | |
udge | -0.036 |
ipples | -0.030 |
eter | -0.030 |
elly | -0.029 |
urious | -0.028 |
Token | shell |
Token ID | 7582 |
Feature activation | -1.99e-02 |
Pos logit contributions | |
goodbye | +0.004 |
Amen | +0.004 |
fs | +0.004 |
yes | +0.003 |
Midnight | +0.003 |
Neg logit contributions | |
udge | -0.005 |
ipples | -0.004 |
eter | -0.004 |
elly | -0.004 |
ill | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -1.08e-02 |
Pos logit contributions | |
goodbye | +0.036 |
Amen | +0.032 |
fs | +0.029 |
forehead | +0.029 |
headset | +0.028 |
Neg logit contributions | |
udge | -0.031 |
eter | -0.027 |
ill | -0.025 |
ipples | -0.025 |
elly | -0.025 |
Token | It |
Token ID | 632 |
Feature activation | -2.21e-03 |
Pos logit contributions | |
goodbye | +0.020 |
fs | +0.017 |
Amen | +0.017 |
yes | +0.016 |
saddle | +0.016 |
Neg logit contributions | |
udge | -0.023 |
eter | -0.020 |
ipples | -0.019 |
urious | -0.019 |
ill | -0.019 |
Token | was |
Token ID | 373 |
Feature activation | +4.28e-03 |
Pos logit contributions | |
goodbye | +0.003 |
| +0.003 |
yes | +0.003 |
fs | +0.003 |
blanket | +0.002 |
Neg logit contributions | |
udge | -0.006 |
eter | -0.005 |
ipples | -0.005 |
elly | -0.005 |
ierce | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | -8.53e-03 |
Pos logit contributions | |
udge | +0.010 |
eter | +0.009 |
ipples | +0.009 |
ill | +0.009 |
urious | +0.009 |
Neg logit contributions | |
goodbye | -0.007 |
Amen | -0.005 |
fs | -0.005 |
yes | -0.005 |
forehead | -0.005 |
Token | sun |
Token ID | 4252 |
Feature activation | -8.56e-03 |
Pos logit contributions | |
goodbye | +0.017 |
fs | +0.014 |
Amen | +0.014 |
ishly | +0.013 |
forehead | +0.013 |
Neg logit contributions | |
udge | -0.017 |
eter | -0.014 |
ipples | -0.014 |
elly | -0.014 |
cy | -0.013 |
Token | sh |
Token ID | 1477 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
goodbye | +0.020 |
Amen | +0.017 |
fs | +0.017 |
headset | +0.016 |
medal | +0.016 |
Neg logit contributions | |
udge | -0.014 |
eter | -0.013 |
ill | -0.012 |
cy | -0.012 |
ipples | -0.012 |
Token | iny |
Token ID | 3541 |
Feature activation | +1.59e-02 |
Pos logit contributions | |
udge | +0.183 |
eter | +0.158 |
Like | +0.147 |
aught | +0.146 |
urious | +0.143 |
Neg logit contributions | |
goodbye | -0.197 |
fs | -0.176 |
pron | -0.165 |
Amen | -0.159 |
saddle | -0.159 |
Token | day |
Token ID | 1110 |
Feature activation | +8.34e-04 |
Pos logit contributions | |
udge | +0.033 |
eter | +0.029 |
ipples | +0.028 |
ill | +0.027 |
elly | +0.027 |
Neg logit contributions | |
goodbye | -0.030 |
Amen | -0.026 |
fs | -0.025 |
forehead | -0.023 |
pron | -0.023 |
Token | and |
Token ID | 290 |
Feature activation | -1.94e-02 |
Pos logit contributions | |
udge | +0.001 |
ill | +0.001 |
eter | +0.001 |
ipples | +0.001 |
elly | +0.001 |
Neg logit contributions | |
goodbye | -0.002 |
Amen | -0.002 |
fs | -0.002 |
saddle | -0.001 |
medal | -0.001 |
Token | there |
Token ID | 612 |
Feature activation | -1.52e-02 |
Pos logit contributions | |
goodbye | +0.032 |
Amen | +0.026 |
fs | +0.024 |
saddle | +0.023 |
forehead | +0.023 |
Neg logit contributions | |
udge | -0.044 |
eter | -0.040 |
ipples | -0.039 |
ill | -0.038 |
ased | -0.037 |
Token | were |
Token ID | 547 |
Feature activation | +1.36e-03 |
Pos logit contributions | |
goodbye | +0.030 |
fs | +0.024 |
Amen | +0.024 |
yes | +0.023 |
saddle | +0.023 |
Neg logit contributions | |
udge | -0.031 |
eter | -0.027 |
ipples | -0.027 |
urious | -0.026 |
ill | -0.025 |
Token |
|
Token ID | 198 |
Feature activation | -4.61e-03 |
Pos logit contributions | |
udge | +0.007 |
ipples | +0.006 |
eter | +0.006 |
urious | +0.006 |
ased | +0.006 |
Neg logit contributions | |
goodbye | -0.004 |
Amen | -0.003 |
| -0.003 |
fs | -0.003 |
yes | -0.003 |
Token |
|
Token ID | 198 |
Feature activation | -9.35e-04 |
Pos logit contributions | |
goodbye | +0.008 |
fs | +0.007 |
Amen | +0.007 |
yes | +0.006 |
saddle | +0.006 |
Neg logit contributions | |
udge | -0.010 |
eter | -0.009 |
ipples | -0.009 |
ill | -0.008 |
urious | -0.008 |
Token | The |
Token ID | 464 |
Feature activation | -1.51e-02 |
Pos logit contributions | |
goodbye | +0.002 |
fs | +0.001 |
Amen | +0.001 |
yes | +0.001 |
saddle | +0.001 |
Neg logit contributions | |
udge | -0.002 |
eter | -0.002 |
ipples | -0.002 |
ill | -0.002 |
urious | -0.002 |
Token | cat |
Token ID | 3797 |
Feature activation | -5.76e-03 |
Pos logit contributions | |
goodbye | +0.031 |
fs | +0.025 |
Amen | +0.023 |
saddle | +0.023 |
chief | +0.023 |
Neg logit contributions | |
udge | -0.029 |
urious | -0.026 |
eter | -0.026 |
ipples | -0.025 |
Like | -0.024 |
Token | grinned |
Token ID | 36268 |
Feature activation | +2.27e-02 |
Pos logit contributions | |
goodbye | +0.012 |
fs | +0.010 |
Amen | +0.010 |
pron | +0.010 |
saddle | +0.009 |
Neg logit contributions | |
udge | -0.011 |
eter | -0.009 |
ipples | -0.009 |
ill | -0.009 |
urious | -0.009 |
Token | mis |
Token ID | 2984 |
Feature activation | +1.00e-01 |
Pos logit contributions | |
udge | +0.053 |
eter | +0.047 |
ipples | +0.046 |
urious | +0.044 |
ill | +0.044 |
Neg logit contributions | |
goodbye | -0.040 |
Amen | -0.030 |
fs | -0.029 |
yes | -0.029 |
forehead | -0.028 |
Token | chie |
Token ID | 3043 |
Feature activation | +4.80e-02 |
Pos logit contributions | |
udge | +0.152 |
ipples | +0.151 |
ill | +0.146 |
Esk | +0.142 |
eter | +0.138 |
Neg logit contributions | |
goodbye | -0.171 |
ishly | -0.163 |
saddle | -0.156 |
forehead | -0.155 |
Amen | -0.152 |
Token | v |
Token ID | 85 |
Feature activation | -1.69e-02 |
Pos logit contributions | |
udge | +0.057 |
eter | +0.045 |
ipples | +0.044 |
ill | +0.039 |
urious | +0.039 |
Neg logit contributions | |
goodbye | -0.136 |
fs | -0.118 |
Amen | -0.117 |
yes | -0.115 |
saddle | -0.113 |
Token | ously |
Token ID | 3481 |
Feature activation | +1.22e-02 |
Pos logit contributions | |
goodbye | +0.037 |
fs | +0.031 |
Amen | +0.030 |
yes | +0.029 |
forehead | +0.029 |
Neg logit contributions | |
udge | -0.030 |
eter | -0.026 |
ipples | -0.026 |
urious | -0.024 |
ased | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | +1.08e-02 |
Pos logit contributions | |
udge | +0.022 |
eter | +0.020 |
urious | +0.018 |
ased | +0.018 |
ipples | +0.018 |
Neg logit contributions | |
fs | -0.022 |
Amen | -0.022 |
pron | -0.021 |
goodbye | -0.021 |
medal | -0.020 |
Token | " |
Token ID | 366 |
Feature activation | +4.91e-02 |
Pos logit contributions | |
udge | +0.026 |
ipples | +0.024 |
eter | +0.023 |
urious | +0.022 |
ill | +0.022 |
Neg logit contributions | |
goodbye | -0.018 |
Amen | -0.013 |
| -0.013 |
yes | -0.012 |
fs | -0.012 |
Token | books |
Token ID | 3835 |
Feature activation | -3.24e-03 |
Pos logit contributions | |
goodbye | +0.019 |
fs | +0.016 |
Amen | +0.016 |
pron | +0.015 |
yes | +0.015 |
Neg logit contributions | |
udge | -0.019 |
eter | -0.018 |
ipples | -0.018 |
ill | -0.017 |
ased | -0.016 |
Token | and |
Token ID | 290 |
Feature activation | -5.98e-03 |
Pos logit contributions | |
goodbye | +0.007 |
Amen | +0.006 |
forehead | +0.006 |
fs | +0.006 |
chief | +0.006 |
Neg logit contributions | |
udge | -0.005 |
ill | -0.005 |
eter | -0.004 |
ipples | -0.004 |
elly | -0.004 |
Token | cakes |
Token ID | 30849 |
Feature activation | -1.43e-03 |
Pos logit contributions | |
goodbye | +0.011 |
fs | +0.010 |
Amen | +0.009 |
pron | +0.009 |
yes | +0.009 |
Neg logit contributions | |
udge | -0.012 |
eter | -0.011 |
ipples | -0.011 |
ill | -0.010 |
ased | -0.010 |
Token | and |
Token ID | 290 |
Feature activation | +9.15e-03 |
Pos logit contributions | |
goodbye | +0.003 |
Amen | +0.003 |
fs | +0.003 |
forehead | +0.003 |
pron | +0.003 |
Neg logit contributions | |
udge | -0.002 |
eter | -0.002 |
ill | -0.002 |
ipples | -0.002 |
ased | -0.002 |
Token | ice |
Token ID | 4771 |
Feature activation | +2.29e-02 |
Pos logit contributions | |
udge | +0.019 |
eter | +0.017 |
ipples | +0.017 |
ill | +0.016 |
ased | +0.016 |
Neg logit contributions | |
goodbye | -0.017 |
fs | -0.014 |
Amen | -0.014 |
yes | -0.013 |
pron | -0.013 |
Token | cre |
Token ID | 1126 |
Feature activation | +1.00e-01 |
Pos logit contributions | |
udge | +0.050 |
eter | +0.044 |
ipples | +0.043 |
urious | +0.042 |
ill | +0.042 |
Neg logit contributions | |
goodbye | -0.042 |
fs | -0.034 |
Amen | -0.033 |
yes | -0.032 |
saddle | -0.032 |
Token | ams |
Token ID | 4105 |
Feature activation | -3.96e-03 |
Pos logit contributions | |
udge | +0.200 |
Like | +0.182 |
eter | +0.181 |
ierce | +0.174 |
hed | +0.171 |
Neg logit contributions | |
goodbye | -0.142 |
Amen | -0.122 |
fs | -0.122 |
on | -0.120 |
pron | -0.120 |
Token | and |
Token ID | 290 |
Feature activation | -1.65e-02 |
Pos logit contributions | |
goodbye | +0.008 |
Amen | +0.008 |
fs | +0.007 |
Midnight | +0.007 |
pron | +0.007 |
Neg logit contributions | |
udge | -0.007 |
eter | -0.006 |
ipples | -0.006 |
ill | -0.006 |
elly | -0.006 |
Token | cand |
Token ID | 2658 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
goodbye | +0.029 |
fs | +0.024 |
Amen | +0.023 |
yes | +0.022 |
pron | +0.022 |
Neg logit contributions | |
udge | -0.036 |
eter | -0.033 |
ipples | -0.032 |
ill | -0.030 |
ased | -0.030 |
Token | ies |
Token ID | 444 |
Feature activation | -9.29e-03 |
Pos logit contributions | |
udge | +0.016 |
eter | +0.013 |
ipples | +0.013 |
ill | +0.012 |
urious | +0.012 |
Neg logit contributions | |
goodbye | -0.031 |
Amen | -0.026 |
fs | -0.026 |
yes | -0.025 |
forehead | -0.025 |
Token | and |
Token ID | 290 |
Feature activation | -1.99e-02 |
Pos logit contributions | |
goodbye | +0.021 |
Amen | +0.020 |
fs | +0.020 |
forehead | +0.019 |
pron | +0.019 |
Neg logit contributions | |
udge | -0.012 |
ipples | -0.011 |
eter | -0.011 |
ill | -0.011 |
than | -0.010 |
Token | cakes |
Token ID | 30849 |
Feature activation | -2.01e-02 |
Pos logit contributions | |
goodbye | +0.012 |
Amen | +0.010 |
fs | +0.010 |
forehead | +0.009 |
yes | +0.009 |
Neg logit contributions | |
udge | -0.015 |
eter | -0.014 |
ipples | -0.014 |
ill | -0.013 |
ased | -0.013 |
Token | and |
Token ID | 290 |
Feature activation | +2.70e-03 |
Pos logit contributions | |
goodbye | +0.033 |
Amen | +0.028 |
fs | +0.026 |
forehead | +0.025 |
saddle | +0.024 |
Neg logit contributions | |
udge | -0.047 |
eter | -0.042 |
ipples | -0.042 |
ill | -0.041 |
ased | -0.040 |
Token | cookies |
Token ID | 14746 |
Feature activation | -1.95e-02 |
Pos logit contributions | |
udge | +0.006 |
eter | +0.006 |
ipples | +0.006 |
ill | +0.005 |
ased | +0.005 |
Neg logit contributions | |
goodbye | -0.004 |
fs | -0.003 |
Amen | -0.003 |
forehead | -0.003 |
saddle | -0.003 |
Token | and |
Token ID | 290 |
Feature activation | -5.07e-03 |
Pos logit contributions | |
goodbye | +0.036 |
Amen | +0.032 |
fs | +0.029 |
forehead | +0.029 |
saddle | +0.028 |
Neg logit contributions | |
udge | -0.041 |
ipples | -0.036 |
ill | -0.036 |
eter | -0.035 |
ased | -0.034 |
Token | ch |
Token ID | 442 |
Feature activation | +1.99e-02 |
Pos logit contributions | |
goodbye | +0.008 |
fs | +0.007 |
Amen | +0.007 |
saddle | +0.006 |
pron | +0.006 |
Neg logit contributions | |
udge | -0.011 |
eter | -0.011 |
ipples | -0.010 |
ill | -0.010 |
ased | -0.010 |
Token | ocol |
Token ID | 4668 |
Feature activation | +1.00e-01 |
Pos logit contributions | |
udge | +0.036 |
eter | +0.031 |
urious | +0.031 |
Like | +0.031 |
ill | +0.029 |
Neg logit contributions | |
goodbye | -0.039 |
fs | -0.031 |
saddle | -0.029 |
chief | -0.029 |
on | -0.029 |
Token | ates |
Token ID | 689 |
Feature activation | -2.55e-02 |
Pos logit contributions | |
udge | +0.239 |
eter | +0.213 |
urious | +0.211 |
ipples | +0.205 |
ierce | +0.202 |
Neg logit contributions | |
goodbye | -0.140 |
fs | -0.120 |
Amen | -0.108 |
saddle | -0.106 |
on | -0.103 |
Token | and |
Token ID | 290 |
Feature activation | -1.37e-02 |
Pos logit contributions | |
goodbye | +0.052 |
Amen | +0.048 |
fs | +0.043 |
forehead | +0.042 |
saddle | +0.042 |
Neg logit contributions | |
udge | -0.049 |
ill | -0.044 |
ipples | -0.044 |
eter | -0.043 |
ased | -0.040 |
Token | ice |
Token ID | 4771 |
Feature activation | +1.54e-02 |
Pos logit contributions | |
goodbye | +0.023 |
fs | +0.019 |
Amen | +0.019 |
saddle | +0.017 |
pron | +0.017 |
Neg logit contributions | |
udge | -0.030 |
eter | -0.028 |
ipples | -0.027 |
ill | -0.026 |
ased | -0.025 |
Token | cream |
Token ID | 8566 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
udge | +0.039 |
eter | +0.035 |
ipples | +0.034 |
urious | +0.033 |
ill | +0.033 |
Neg logit contributions | |
goodbye | -0.023 |
fs | -0.017 |
Amen | -0.017 |
yes | -0.016 |
saddle | -0.016 |
Token | . |
Token ID | 13 |
Feature activation | +1.01e-02 |
Pos logit contributions | |
goodbye | +0.034 |
Amen | +0.029 |
fs | +0.027 |
forehead | +0.026 |
yes | +0.026 |
Neg logit contributions | |
udge | -0.035 |
eter | -0.031 |
ipples | -0.030 |
ill | -0.030 |
ased | -0.028 |
Token | Cr |
Token ID | 3864 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
goodbye | +0.035 |
fs | +0.031 |
Amen | +0.031 |
forehead | +0.030 |
saddle | +0.029 |
Neg logit contributions | |
udge | -0.019 |
eter | -0.016 |
ipples | -0.016 |
ill | -0.015 |
ierce | -0.014 |
Token | ust |
Token ID | 436 |
Feature activation | +9.89e-02 |
Pos logit contributions | |
udge | +0.056 |
eter | +0.050 |
ased | +0.050 |
Esk | +0.047 |
ipples | +0.045 |
Neg logit contributions | |
goodbye | -0.114 |
fs | -0.104 |
pron | -0.103 |
magazine | -0.101 |
medal | -0.099 |
Token | y |
Token ID | 88 |
Feature activation | +5.66e-03 |
Pos logit contributions | |
udge | +0.167 |
eter | +0.140 |
ipples | +0.124 |
ased | +0.122 |
urious | +0.121 |
Neg logit contributions | |
goodbye | -0.236 |
medal | -0.191 |
Amen | -0.185 |
helicopter | -0.185 |
saddle | -0.180 |
Token | . |
Token ID | 13 |
Feature activation | +2.34e-02 |
Pos logit contributions | |
udge | +0.011 |
eter | +0.010 |
ipples | +0.010 |
ill | +0.009 |
elly | +0.009 |
Neg logit contributions | |
goodbye | -0.011 |
fs | -0.010 |
Amen | -0.009 |
yes | -0.009 |
forehead | -0.009 |
Token | Cr |
Token ID | 3864 |
Feature activation | +5.06e-02 |
Pos logit contributions | |
udge | +0.045 |
eter | +0.038 |
ipples | +0.038 |
urious | +0.036 |
ill | +0.035 |
Neg logit contributions | |
goodbye | -0.050 |
Amen | -0.040 |
fs | -0.040 |
yes | -0.040 |
saddle | -0.038 |
Token | ust |
Token ID | 436 |
Feature activation | +9.97e-02 |
Pos logit contributions | |
udge | +0.054 |
Esk | +0.052 |
ased | +0.051 |
eter | +0.048 |
ipples | +0.045 |
Neg logit contributions | |
goodbye | -0.108 |
fs | -0.102 |
magazine | -0.102 |
medal | -0.099 |
on | -0.097 |
Token | y |
Token ID | 88 |
Feature activation | +5.02e-03 |
Pos logit contributions | |
udge | +0.170 |
eter | +0.144 |
ierce | +0.129 |
ipples | +0.129 |
ill | +0.126 |
Neg logit contributions | |
goodbye | -0.231 |
Amen | -0.188 |
medal | -0.186 |
helicopter | -0.180 |
saddle | -0.177 |
Token | lived |
Token ID | 5615 |
Feature activation | +2.23e-02 |
Pos logit contributions | |
udge | +0.009 |
eter | +0.008 |
ipples | +0.008 |
urious | +0.007 |
ill | +0.007 |
Neg logit contributions | |
goodbye | -0.011 |
fs | -0.009 |
Amen | -0.009 |
yes | -0.009 |
forehead | -0.009 |
Token | in |
Token ID | 287 |
Feature activation | +3.55e-02 |
Pos logit contributions | |
udge | +0.047 |
eter | +0.041 |
ipples | +0.041 |
ill | +0.039 |
urious | +0.039 |
Neg logit contributions | |
goodbye | -0.043 |
fs | -0.035 |
Amen | -0.034 |
yes | -0.033 |
forehead | -0.032 |
Token | a |
Token ID | 257 |
Feature activation | +7.69e-03 |
Pos logit contributions | |
udge | +0.084 |
eter | +0.074 |
ipples | +0.074 |
ill | +0.071 |
urious | +0.070 |
Neg logit contributions | |
goodbye | -0.059 |
fs | -0.047 |
Amen | -0.045 |
yes | -0.043 |
forehead | -0.043 |
Token | big |
Token ID | 1263 |
Feature activation | +6.40e-03 |
Pos logit contributions | |
udge | +0.016 |
ipples | +0.014 |
urious | +0.013 |
eter | +0.013 |
elly | +0.013 |
Neg logit contributions | |
goodbye | -0.015 |
fs | -0.013 |
yes | -0.012 |
Amen | -0.012 |
forehead | -0.012 |
Token | bled |
Token ID | 9342 |
Feature activation | +3.10e-04 |
Pos logit contributions | |
goodbye | +0.019 |
Amen | +0.017 |
fs | +0.016 |
magazine | +0.015 |
pron | +0.015 |
Neg logit contributions | |
udge | -0.022 |
eter | -0.019 |
ipples | -0.019 |
ook | -0.018 |
ill | -0.018 |
Token | up |
Token ID | 510 |
Feature activation | +1.71e-02 |
Pos logit contributions | |
udge | +0.001 |
eter | +0.001 |
ipples | +0.001 |
ill | +0.001 |
ased | +0.001 |
Neg logit contributions | |
goodbye | -0.001 |
Amen | -0.000 |
fs | -0.000 |
yes | -0.000 |
forehead | -0.000 |
Token | all |
Token ID | 477 |
Feature activation | +3.93e-02 |
Pos logit contributions | |
udge | +0.035 |
eter | +0.032 |
ipples | +0.031 |
ill | +0.030 |
ased | +0.029 |
Neg logit contributions | |
goodbye | -0.030 |
Amen | -0.028 |
fs | -0.026 |
pron | -0.025 |
forehead | -0.024 |
Token | the |
Token ID | 262 |
Feature activation | +1.20e-02 |
Pos logit contributions | |
udge | +0.089 |
eter | +0.079 |
ipples | +0.077 |
ill | +0.076 |
urious | +0.074 |
Neg logit contributions | |
goodbye | -0.069 |
Amen | -0.056 |
fs | -0.056 |
yes | -0.054 |
forehead | -0.052 |
Token | ch |
Token ID | 442 |
Feature activation | +2.44e-02 |
Pos logit contributions | |
udge | +0.029 |
eter | +0.026 |
ipples | +0.026 |
ill | +0.025 |
urious | +0.025 |
Neg logit contributions | |
goodbye | -0.020 |
fs | -0.015 |
Amen | -0.015 |
yes | -0.014 |
saddle | -0.014 |
Token | ocol |
Token ID | 4668 |
Feature activation | +9.95e-02 |
Pos logit contributions | |
udge | +0.040 |
eter | +0.037 |
urious | +0.036 |
ipples | +0.035 |
Like | +0.034 |
Neg logit contributions | |
goodbye | -0.053 |
fs | -0.045 |
saddle | -0.042 |
medal | -0.042 |
on | -0.042 |
Token | ates |
Token ID | 689 |
Feature activation | -2.28e-02 |
Pos logit contributions | |
udge | +0.232 |
eter | +0.210 |
ipples | +0.204 |
urious | +0.200 |
elly | +0.195 |
Neg logit contributions | |
goodbye | -0.156 |
fs | -0.129 |
Amen | -0.120 |
saddle | -0.115 |
yes | -0.114 |
Token | , |
Token ID | 11 |
Feature activation | +1.94e-02 |
Pos logit contributions | |
Amen | +0.047 |
goodbye | +0.046 |
fs | +0.039 |
headset | +0.039 |
saddle | +0.038 |
Neg logit contributions | |
udge | -0.042 |
ill | -0.039 |
eter | -0.037 |
ipples | -0.036 |
cy | -0.034 |
Token | and |
Token ID | 290 |
Feature activation | -3.71e-02 |
Pos logit contributions | |
udge | +0.037 |
eter | +0.034 |
ipples | +0.033 |
ill | +0.032 |
ased | +0.032 |
Neg logit contributions | |
goodbye | -0.036 |
fs | -0.033 |
Amen | -0.033 |
microphone | -0.029 |
medal | -0.029 |
Token | she |
Token ID | 673 |
Feature activation | +5.11e-03 |
Pos logit contributions | |
goodbye | +0.063 |
Amen | +0.052 |
fs | +0.051 |
saddle | +0.047 |
yes | +0.047 |
Neg logit contributions | |
udge | -0.083 |
eter | -0.076 |
ipples | -0.073 |
ill | -0.071 |
ased | -0.070 |
Token | was |
Token ID | 373 |
Feature activation | -2.28e-02 |
Pos logit contributions | |
udge | +0.010 |
eter | +0.009 |
ipples | +0.009 |
ill | +0.009 |
urious | +0.008 |
Neg logit contributions | |
goodbye | -0.010 |
fs | -0.008 |
Amen | -0.008 |
yes | -0.008 |
saddle | -0.008 |
Token | the |
Token ID | 262 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
udge | +0.089 |
eter | +0.080 |
ipples | +0.078 |
urious | +0.075 |
ill | +0.075 |
Neg logit contributions | |
goodbye | -0.059 |
Amen | -0.045 |
fs | -0.045 |
yes | -0.043 |
saddle | -0.042 |
Token | sky |
Token ID | 6766 |
Feature activation | +5.63e-03 |
Pos logit contributions | |
udge | +0.032 |
ipples | +0.029 |
eter | +0.028 |
urious | +0.027 |
ill | +0.027 |
Neg logit contributions | |
goodbye | -0.024 |
fs | -0.020 |
Amen | -0.019 |
yes | -0.019 |
forehead | -0.018 |
Token | and |
Token ID | 290 |
Feature activation | -1.98e-02 |
Pos logit contributions | |
udge | +0.014 |
eter | +0.012 |
ipples | +0.012 |
urious | +0.012 |
ased | +0.011 |
Neg logit contributions | |
goodbye | -0.009 |
fs | -0.007 |
yes | -0.007 |
Amen | -0.006 |
pron | -0.006 |
Token | heard |
Token ID | 2982 |
Feature activation | +7.41e-03 |
Pos logit contributions | |
goodbye | +0.028 |
fs | +0.020 |
Amen | +0.020 |
yes | +0.019 |
saddle | +0.018 |
Neg logit contributions | |
udge | -0.053 |
eter | -0.048 |
ipples | -0.047 |
ill | -0.045 |
urious | -0.045 |
Token | thunder |
Token ID | 18355 |
Feature activation | +3.53e-02 |
Pos logit contributions | |
udge | +0.018 |
eter | +0.016 |
ipples | +0.016 |
ill | +0.015 |
urious | +0.015 |
Neg logit contributions | |
goodbye | -0.012 |
fs | -0.009 |
Amen | -0.009 |
yes | -0.009 |
saddle | -0.009 |
Token | r |
Token ID | 374 |
Feature activation | +9.90e-02 |
Pos logit contributions | |
udge | +0.079 |
eter | +0.070 |
ipples | +0.069 |
urious | +0.066 |
ill | +0.066 |
Neg logit contributions | |
goodbye | -0.063 |
Amen | -0.049 |
fs | -0.049 |
yes | -0.047 |
forehead | -0.046 |
Token | umbling |
Token ID | 14739 |
Feature activation | +4.47e-02 |
Pos logit contributions | |
udge | +0.121 |
eter | +0.111 |
Like | +0.098 |
cy | +0.091 |
urious | +0.089 |
Neg logit contributions | |
goodbye | -0.201 |
fs | -0.194 |
Amen | -0.186 |
saddle | -0.183 |
on | -0.183 |
Token | . |
Token ID | 13 |
Feature activation | -1.20e-02 |
Pos logit contributions | |
udge | +0.117 |
eter | +0.104 |
ill | +0.100 |
ipples | +0.099 |
urious | +0.098 |
Neg logit contributions | |
goodbye | -0.071 |
helicopter | -0.047 |
Amen | -0.045 |
forehead | -0.044 |
pony | -0.044 |
Token | He |
Token ID | 679 |
Feature activation | +1.11e-02 |
Pos logit contributions | |
goodbye | +0.019 |
Amen | +0.015 |
fs | +0.014 |
| +0.014 |
yes | +0.014 |
Neg logit contributions | |
udge | -0.030 |
eter | -0.026 |
ipples | -0.026 |
ill | -0.025 |
ased | -0.025 |
Token | knew |
Token ID | 2993 |
Feature activation | -8.50e-03 |
Pos logit contributions | |
udge | +0.026 |
eter | +0.023 |
ipples | +0.022 |
ill | +0.021 |
elly | +0.021 |
Neg logit contributions | |
goodbye | -0.019 |
Amen | -0.015 |
fs | -0.015 |
yes | -0.015 |
saddle | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | -2.50e-02 |
Pos logit contributions | |
goodbye | +0.010 |
fs | +0.007 |
Amen | +0.007 |
forehead | +0.006 |
saddle | +0.006 |
Neg logit contributions | |
udge | -0.024 |
eter | -0.022 |
ipples | -0.022 |
ill | -0.021 |
urious | -0.021 |
Token | a |
Token ID | 257 |
Feature activation | -2.38e-03 |
Pos logit contributions | |
udge | +0.067 |
eter | +0.058 |
ipples | +0.058 |
ill | +0.056 |
elly | +0.055 |
Neg logit contributions | |
goodbye | -0.055 |
fs | -0.045 |
Amen | -0.044 |
forehead | -0.042 |
saddle | -0.042 |
Token | little |
Token ID | 1310 |
Feature activation | +1.37e-02 |
Pos logit contributions | |
goodbye | +0.005 |
fs | +0.004 |
Amen | +0.004 |
yes | +0.003 |
forehead | +0.003 |
Neg logit contributions | |
udge | -0.005 |
eter | -0.004 |
ipples | -0.004 |
urious | -0.004 |
elly | -0.004 |
Token | crab |
Token ID | 32202 |
Feature activation | -6.58e-03 |
Pos logit contributions | |
udge | +0.024 |
eter | +0.019 |
ipples | +0.019 |
elly | +0.019 |
cy | +0.018 |
Neg logit contributions | |
goodbye | -0.032 |
fs | -0.028 |
Amen | -0.026 |
yes | -0.025 |
forehead | -0.025 |
Token | named |
Token ID | 3706 |
Feature activation | -1.37e-02 |
Pos logit contributions | |
goodbye | +0.013 |
fs | +0.011 |
yes | +0.010 |
forehead | +0.010 |
Amen | +0.010 |
Neg logit contributions | |
udge | -0.014 |
ipples | -0.011 |
eter | -0.011 |
elly | -0.011 |
urious | -0.011 |
Token | Cr |
Token ID | 3864 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
goodbye | +0.035 |
fs | +0.031 |
Amen | +0.031 |
forehead | +0.030 |
saddle | +0.029 |
Neg logit contributions | |
udge | -0.019 |
eter | -0.016 |
ipples | -0.016 |
ill | -0.015 |
ierce | -0.014 |
Token | ust |
Token ID | 436 |
Feature activation | +9.89e-02 |
Pos logit contributions | |
udge | +0.056 |
eter | +0.050 |
ased | +0.050 |
Esk | +0.047 |
ipples | +0.045 |
Neg logit contributions | |
goodbye | -0.114 |
fs | -0.104 |
pron | -0.103 |
magazine | -0.101 |
medal | -0.099 |
Token | y |
Token ID | 88 |
Feature activation | +5.66e-03 |
Pos logit contributions | |
udge | +0.167 |
eter | +0.140 |
ipples | +0.124 |
ased | +0.122 |
urious | +0.121 |
Neg logit contributions | |
goodbye | -0.236 |
medal | -0.191 |
Amen | -0.185 |
helicopter | -0.185 |
saddle | -0.180 |
Token | . |
Token ID | 13 |
Feature activation | +2.34e-02 |
Pos logit contributions | |
udge | +0.011 |
eter | +0.010 |
ipples | +0.010 |
ill | +0.009 |
elly | +0.009 |
Neg logit contributions | |
goodbye | -0.011 |
fs | -0.010 |
Amen | -0.009 |
yes | -0.009 |
forehead | -0.009 |
Token | Cr |
Token ID | 3864 |
Feature activation | +5.06e-02 |
Pos logit contributions | |
udge | +0.045 |
eter | +0.038 |
ipples | +0.038 |
urious | +0.036 |
ill | +0.035 |
Neg logit contributions | |
goodbye | -0.050 |
Amen | -0.040 |
fs | -0.040 |
yes | -0.040 |
saddle | -0.038 |
Token | ust |
Token ID | 436 |
Feature activation | +9.97e-02 |
Pos logit contributions | |
udge | +0.054 |
Esk | +0.052 |
ased | +0.051 |
eter | +0.048 |
ipples | +0.045 |
Neg logit contributions | |
goodbye | -0.108 |
fs | -0.102 |
magazine | -0.102 |
medal | -0.099 |
on | -0.097 |
Token | y |
Token ID | 88 |
Feature activation | +5.02e-03 |
Pos logit contributions | |
udge | +0.170 |
eter | +0.144 |
ierce | +0.129 |
ipples | +0.129 |
ill | +0.126 |
Neg logit contributions | |
goodbye | -0.231 |
Amen | -0.188 |
medal | -0.186 |
helicopter | -0.180 |
saddle | -0.177 |
Token | inside |
Token ID | 2641 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
udge | +0.010 |
eter | +0.009 |
ipples | +0.009 |
ill | +0.009 |
urious | +0.009 |
Neg logit contributions | |
goodbye | -0.008 |
fs | -0.006 |
Amen | -0.006 |
yes | -0.006 |
pron | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | -1.11e-03 |
Pos logit contributions | |
udge | +0.044 |
eter | +0.038 |
ipples | +0.038 |
ill | +0.037 |
urious | +0.036 |
Neg logit contributions | |
goodbye | -0.032 |
fs | -0.026 |
Amen | -0.026 |
saddle | -0.024 |
yes | -0.024 |
Token | She |
Token ID | 1375 |
Feature activation | +1.24e-02 |
Pos logit contributions | |
goodbye | +0.002 |
Amen | +0.001 |
fs | +0.001 |
yes | +0.001 |
| +0.001 |
Neg logit contributions | |
udge | -0.003 |
ipples | -0.002 |
eter | -0.002 |
ill | -0.002 |
ased | -0.002 |
Token | heard |
Token ID | 2982 |
Feature activation | +6.25e-03 |
Pos logit contributions | |
udge | +0.024 |
eter | +0.021 |
ipples | +0.021 |
ill | +0.020 |
urious | +0.020 |
Neg logit contributions | |
goodbye | -0.026 |
fs | -0.021 |
Amen | -0.021 |
yes | -0.020 |
saddle | -0.020 |
Token | thunder |
Token ID | 18355 |
Feature activation | +3.27e-02 |
Pos logit contributions | |
udge | +0.014 |
eter | +0.012 |
ipples | +0.012 |
ill | +0.012 |
urious | +0.011 |
Neg logit contributions | |
goodbye | -0.011 |
Amen | -0.009 |
fs | -0.009 |
saddle | -0.009 |
pron | -0.009 |
Token | r |
Token ID | 374 |
Feature activation | +9.87e-02 |
Pos logit contributions | |
udge | +0.073 |
eter | +0.065 |
ipples | +0.064 |
ill | +0.061 |
urious | +0.061 |
Neg logit contributions | |
goodbye | -0.059 |
fs | -0.046 |
Amen | -0.046 |
yes | -0.044 |
saddle | -0.043 |
Token | umbling |
Token ID | 14739 |
Feature activation | +4.79e-02 |
Pos logit contributions | |
udge | +0.120 |
eter | +0.112 |
cy | +0.092 |
ierce | +0.091 |
Like | +0.090 |
Neg logit contributions | |
fs | -0.191 |
goodbye | -0.189 |
forehead | -0.179 |
ishly | -0.177 |
on | -0.176 |
Token | and |
Token ID | 290 |
Feature activation | -1.34e-02 |
Pos logit contributions | |
udge | +0.119 |
eter | +0.105 |
ipples | +0.103 |
ill | +0.100 |
urious | +0.100 |
Neg logit contributions | |
goodbye | -0.080 |
forehead | -0.055 |
fs | -0.054 |
Amen | -0.054 |
yes | -0.053 |
Token | felt |
Token ID | 2936 |
Feature activation | -4.03e-03 |
Pos logit contributions | |
goodbye | +0.024 |
yes | +0.018 |
Amen | +0.017 |
fs | +0.017 |
forehead | +0.017 |
Neg logit contributions | |
udge | -0.031 |
eter | -0.028 |
ipples | -0.027 |
elly | -0.026 |
urious | -0.026 |
Token | the |
Token ID | 262 |
Feature activation | +2.04e-03 |
Pos logit contributions | |
goodbye | +0.007 |
Amen | +0.006 |
fs | +0.006 |
yes | +0.006 |
forehead | +0.005 |
Neg logit contributions | |
udge | -0.009 |
eter | -0.008 |
ipples | -0.008 |
ill | -0.007 |
urious | -0.007 |
Token | rain |
Token ID | 6290 |
Feature activation | +2.99e-02 |
Pos logit contributions | |
udge | +0.004 |
ipples | +0.004 |
eter | +0.004 |
urious | +0.003 |
ill | +0.003 |
Neg logit contributions | |
goodbye | -0.004 |
fs | -0.003 |
Amen | -0.003 |
yes | -0.003 |
saddle | -0.003 |
Token | water |
Token ID | 1660 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
goodbye | +0.021 |
Amen | +0.017 |
fs | +0.017 |
yes | +0.016 |
saddle | +0.016 |
Neg logit contributions | |
udge | -0.024 |
eter | -0.021 |
ipples | -0.021 |
ill | -0.020 |
urious | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | +2.03e-02 |
Pos logit contributions | |
goodbye | +0.028 |
Amen | +0.025 |
fs | +0.024 |
pron | +0.022 |
saddle | +0.022 |
Neg logit contributions | |
udge | -0.031 |
ipples | -0.027 |
eter | -0.027 |
ill | -0.026 |
ased | -0.025 |
Token | They |
Token ID | 1119 |
Feature activation | -1.74e-02 |
Pos logit contributions | |
udge | +0.049 |
urious | +0.045 |
eter | +0.044 |
ipples | +0.044 |
rying | +0.043 |
Neg logit contributions | |
goodbye | -0.033 |
| -0.025 |
yes | -0.024 |
fs | -0.023 |
Midnight | -0.023 |
Token | heard |
Token ID | 2982 |
Feature activation | +6.50e-03 |
Pos logit contributions | |
goodbye | +0.032 |
yes | +0.026 |
fs | +0.024 |
saddle | +0.024 |
medal | +0.023 |
Neg logit contributions | |
udge | -0.038 |
eter | -0.035 |
elly | -0.033 |
urious | -0.032 |
ipples | -0.032 |
Token | it |
Token ID | 340 |
Feature activation | +7.25e-03 |
Pos logit contributions | |
udge | +0.015 |
eter | +0.014 |
ipples | +0.013 |
ill | +0.013 |
urious | +0.013 |
Neg logit contributions | |
goodbye | -0.011 |
fs | -0.009 |
Amen | -0.009 |
yes | -0.008 |
saddle | -0.008 |
Token | cre |
Token ID | 1126 |
Feature activation | +9.86e-02 |
Pos logit contributions | |
udge | +0.017 |
eter | +0.015 |
ipples | +0.014 |
elly | +0.014 |
ill | +0.014 |
Neg logit contributions | |
goodbye | -0.013 |
fs | -0.010 |
Amen | -0.010 |
yes | -0.010 |
forehead | -0.009 |
Token | ak |
Token ID | 461 |
Feature activation | +2.50e-03 |
Pos logit contributions | |
eter | +0.162 |
udge | +0.151 |
Esk | +0.149 |
Like | +0.147 |
ierce | +0.138 |
Neg logit contributions | |
on | -0.157 |
goodbye | -0.156 |
fs | -0.145 |
ons | -0.143 |
Amen | -0.142 |
Token | and |
Token ID | 290 |
Feature activation | -1.82e-02 |
Pos logit contributions | |
udge | +0.005 |
eter | +0.005 |
ipples | +0.005 |
ill | +0.004 |
urious | +0.004 |
Neg logit contributions | |
goodbye | -0.005 |
Amen | -0.004 |
fs | -0.004 |
pron | -0.004 |
yes | -0.004 |
Token | gro |
Token ID | 7128 |
Feature activation | +1.66e-02 |
Pos logit contributions | |
goodbye | +0.034 |
Amen | +0.026 |
fs | +0.026 |
yes | +0.026 |
forehead | +0.025 |
Neg logit contributions | |
udge | -0.040 |
eter | -0.036 |
ipples | -0.035 |
urious | -0.033 |
elly | -0.033 |
Token | an |
Token ID | 272 |
Feature activation | -7.04e-03 |
Pos logit contributions | |
udge | +0.031 |
eter | +0.027 |
ipples | +0.026 |
ill | +0.025 |
elly | +0.025 |
Neg logit contributions | |
goodbye | -0.035 |
fs | -0.030 |
Amen | -0.029 |
forehead | -0.028 |
saddle | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | +2.30e-02 |
Pos logit contributions | |
goodbye | +0.013 |
pron | +0.012 |
fs | +0.012 |
Amen | +0.012 |
Midnight | +0.011 |
Neg logit contributions | |
udge | -0.014 |
ipples | -0.012 |
eter | -0.012 |
ill | -0.011 |
ased | -0.011 |
Token | sky |
Token ID | 6766 |
Feature activation | -7.50e-03 |
Pos logit contributions | |
goodbye | +0.032 |
Amen | +0.027 |
fs | +0.027 |
yes | +0.025 |
saddle | +0.025 |
Neg logit contributions | |
udge | -0.035 |
eter | -0.031 |
ipples | -0.031 |
urious | -0.030 |
elly | -0.029 |
Token | was |
Token ID | 373 |
Feature activation | +1.19e-02 |
Pos logit contributions | |
goodbye | +0.014 |
fs | +0.011 |
Amen | +0.011 |
yes | +0.011 |
pron | +0.011 |
Neg logit contributions | |
udge | -0.016 |
eter | -0.014 |
ipples | -0.014 |
ierce | -0.013 |
ill | -0.013 |
Token | gray |
Token ID | 12768 |
Feature activation | +2.87e-02 |
Pos logit contributions | |
udge | +0.022 |
ipples | +0.019 |
eter | +0.019 |
ill | +0.019 |
urious | +0.018 |
Neg logit contributions | |
goodbye | -0.025 |
fs | -0.021 |
Amen | -0.021 |
saddle | -0.020 |
medal | -0.019 |
Token | and |
Token ID | 290 |
Feature activation | -1.93e-02 |
Pos logit contributions | |
udge | +0.064 |
eter | +0.057 |
ipples | +0.056 |
ill | +0.053 |
urious | +0.053 |
Neg logit contributions | |
goodbye | -0.052 |
fs | -0.041 |
Amen | -0.041 |
yes | -0.039 |
saddle | -0.038 |
Token | it |
Token ID | 340 |
Feature activation | +8.14e-03 |
Pos logit contributions | |
goodbye | +0.038 |
fs | +0.031 |
saddle | +0.031 |
Amen | +0.031 |
forehead | +0.030 |
Neg logit contributions | |
udge | -0.036 |
ipples | -0.033 |
eter | -0.032 |
ill | -0.031 |
urious | -0.030 |
Token | r |
Token ID | 374 |
Feature activation | +9.85e-02 |
Pos logit contributions | |
udge | +0.018 |
eter | +0.016 |
ipples | +0.015 |
ill | +0.014 |
urious | +0.014 |
Neg logit contributions | |
goodbye | -0.015 |
fs | -0.012 |
Amen | -0.012 |
yes | -0.012 |
pron | -0.011 |
Token | ained |
Token ID | 1328 |
Feature activation | +1.40e-02 |
Pos logit contributions | |
udge | +0.184 |
eter | +0.165 |
Like | +0.157 |
urious | +0.157 |
roc | +0.155 |
Neg logit contributions | |
goodbye | -0.146 |
fs | -0.142 |
pron | -0.132 |
forehead | -0.127 |
on | -0.122 |
Token | a |
Token ID | 257 |
Feature activation | -3.32e-03 |
Pos logit contributions | |
udge | +0.033 |
eter | +0.029 |
ipples | +0.029 |
ill | +0.028 |
urious | +0.027 |
Neg logit contributions | |
goodbye | -0.024 |
Amen | -0.019 |
fs | -0.018 |
yes | -0.018 |
saddle | -0.017 |
Token | lot |
Token ID | 1256 |
Feature activation | +5.66e-02 |
Pos logit contributions | |
goodbye | +0.006 |
Amen | +0.005 |
fs | +0.005 |
ishly | +0.004 |
chief | +0.004 |
Neg logit contributions | |
udge | -0.007 |
ipples | -0.006 |
eter | -0.006 |
elly | -0.006 |
urious | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | +6.56e-03 |
Pos logit contributions | |
udge | +0.138 |
eter | +0.126 |
urious | +0.123 |
ipples | +0.123 |
roc | +0.118 |
Neg logit contributions | |
goodbye | -0.089 |
fs | -0.067 |
yes | -0.061 |
blanket | -0.059 |
pron | -0.058 |
Token | Anna |
Token ID | 11735 |
Feature activation | +3.23e-03 |
Pos logit contributions | |
udge | +0.015 |
eter | +0.013 |
ipples | +0.013 |
urious | +0.012 |
ill | +0.012 |
Neg logit contributions | |
goodbye | -0.012 |
fs | -0.009 |
Amen | -0.009 |
yes | -0.009 |
forehead | -0.009 |
Token | ground |
Token ID | 2323 |
Feature activation | +3.38e-03 |
Pos logit contributions | |
goodbye | +0.046 |
fs | +0.039 |
Amen | +0.038 |
yes | +0.036 |
Midnight | +0.035 |
Neg logit contributions | |
udge | -0.050 |
ipples | -0.045 |
eter | -0.045 |
urious | -0.042 |
ill | -0.041 |
Token | shake |
Token ID | 13279 |
Feature activation | +2.82e-02 |
Pos logit contributions | |
udge | +0.007 |
eter | +0.006 |
ipples | +0.006 |
ill | +0.006 |
urious | +0.006 |
Neg logit contributions | |
goodbye | -0.006 |
fs | -0.005 |
Amen | -0.005 |
yes | -0.005 |
saddle | -0.005 |
Token | and |
Token ID | 290 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
udge | +0.069 |
eter | +0.061 |
ipples | +0.061 |
urious | +0.059 |
ierce | +0.058 |
Neg logit contributions | |
goodbye | -0.046 |
fs | -0.033 |
yes | -0.032 |
Amen | -0.031 |
forehead | -0.031 |
Token | the |
Token ID | 262 |
Feature activation | -1.62e-02 |
Pos logit contributions | |
goodbye | +0.020 |
fs | +0.015 |
Amen | +0.015 |
yes | +0.015 |
forehead | +0.014 |
Neg logit contributions | |
udge | -0.025 |
eter | -0.022 |
ipples | -0.021 |
urious | -0.021 |
elly | -0.021 |
Token | windows |
Token ID | 9168 |
Feature activation | +1.64e-02 |
Pos logit contributions | |
goodbye | +0.030 |
fs | +0.025 |
Amen | +0.025 |
yes | +0.023 |
Midnight | +0.023 |
Neg logit contributions | |
udge | -0.033 |
ipples | -0.030 |
eter | -0.030 |
urious | -0.028 |
ill | -0.027 |
Token | r |
Token ID | 374 |
Feature activation | +9.84e-02 |
Pos logit contributions | |
udge | +0.037 |
eter | +0.033 |
ipples | +0.032 |
urious | +0.032 |
elly | +0.031 |
Neg logit contributions | |
goodbye | -0.029 |
fs | -0.022 |
yes | -0.021 |
Amen | -0.021 |
forehead | -0.021 |
Token | attle |
Token ID | 1999 |
Feature activation | +2.69e-02 |
Pos logit contributions | |
udge | +0.145 |
eter | +0.123 |
Like | +0.116 |
urious | +0.112 |
ierce | +0.110 |
Neg logit contributions | |
goodbye | -0.192 |
fs | -0.177 |
forehead | -0.174 |
on | -0.172 |
candles | -0.166 |
Token | . |
Token ID | 13 |
Feature activation | +1.90e-02 |
Pos logit contributions | |
udge | +0.062 |
eter | +0.055 |
ipples | +0.054 |
urious | +0.053 |
ill | +0.052 |
Neg logit contributions | |
goodbye | -0.046 |
fs | -0.035 |
yes | -0.034 |
Amen | -0.034 |
forehead | -0.033 |
Token | The |
Token ID | 383 |
Feature activation | -2.18e-02 |
Pos logit contributions | |
udge | +0.043 |
eter | +0.039 |
ipples | +0.038 |
urious | +0.038 |
ased | +0.037 |
Neg logit contributions | |
goodbye | -0.033 |
fs | -0.025 |
| -0.025 |
yes | -0.025 |
Amen | -0.025 |
Token | tower |
Token ID | 10580 |
Feature activation | -1.45e-02 |
Pos logit contributions | |
goodbye | +0.035 |
fs | +0.029 |
Amen | +0.028 |
Midnight | +0.027 |
yes | +0.026 |
Neg logit contributions | |
udge | -0.049 |
ipples | -0.045 |
eter | -0.044 |
urious | -0.042 |
ased | -0.041 |
Token | of |
Token ID | 286 |
Feature activation | +3.06e-02 |
Pos logit contributions | |
goodbye | +0.029 |
fs | +0.023 |
Amen | +0.023 |
yes | +0.022 |
saddle | +0.022 |
Neg logit contributions | |
udge | -0.030 |
eter | -0.026 |
ipples | -0.026 |
ill | -0.025 |
urious | -0.024 |
Token | while |
Token ID | 981 |
Feature activation | +8.48e-03 |
Pos logit contributions | |
goodbye | +0.022 |
fs | +0.017 |
Amen | +0.016 |
yes | +0.016 |
saddle | +0.016 |
Neg logit contributions | |
udge | -0.030 |
eter | -0.027 |
ipples | -0.026 |
urious | -0.025 |
ill | -0.025 |
Token | the |
Token ID | 262 |
Feature activation | -1.53e-02 |
Pos logit contributions | |
udge | +0.019 |
eter | +0.017 |
ipples | +0.016 |
ill | +0.016 |
ased | +0.016 |
Neg logit contributions | |
goodbye | -0.015 |
Amen | -0.012 |
fs | -0.012 |
saddle | -0.012 |
forehead | -0.011 |
Token | lightning |
Token ID | 14357 |
Feature activation | +1.36e-02 |
Pos logit contributions | |
goodbye | +0.027 |
fs | +0.021 |
Amen | +0.021 |
yes | +0.020 |
saddle | +0.019 |
Neg logit contributions | |
udge | -0.035 |
eter | -0.031 |
ipples | -0.031 |
ill | -0.030 |
urious | -0.029 |
Token | and |
Token ID | 290 |
Feature activation | -3.11e-02 |
Pos logit contributions | |
udge | +0.031 |
eter | +0.027 |
ipples | +0.026 |
urious | +0.026 |
ill | +0.025 |
Neg logit contributions | |
goodbye | -0.024 |
fs | -0.019 |
Amen | -0.019 |
yes | -0.018 |
saddle | -0.018 |
Token | thunder |
Token ID | 18355 |
Feature activation | +2.17e-02 |
Pos logit contributions | |
goodbye | +0.055 |
fs | +0.042 |
Amen | +0.041 |
yes | +0.041 |
forehead | +0.040 |
Neg logit contributions | |
udge | -0.071 |
eter | -0.063 |
ipples | -0.063 |
urious | -0.060 |
elly | -0.059 |
Token | r |
Token ID | 374 |
Feature activation | +9.82e-02 |
Pos logit contributions | |
udge | +0.048 |
eter | +0.042 |
ipples | +0.042 |
urious | +0.040 |
ill | +0.040 |
Neg logit contributions | |
goodbye | -0.040 |
fs | -0.031 |
Amen | -0.030 |
yes | -0.030 |
forehead | -0.029 |
Token | umbled |
Token ID | 11137 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
udge | +0.156 |
eter | +0.131 |
Like | +0.129 |
ierce | +0.125 |
urious | +0.124 |
Neg logit contributions | |
goodbye | -0.180 |
fs | -0.164 |
forehead | -0.162 |
chief | -0.157 |
ishly | -0.156 |
Token | outside |
Token ID | 2354 |
Feature activation | +3.34e-02 |
Pos logit contributions | |
udge | +0.027 |
eter | +0.024 |
ipples | +0.023 |
urious | +0.022 |
ill | +0.022 |
Neg logit contributions | |
goodbye | -0.020 |
fs | -0.015 |
Amen | -0.015 |
yes | -0.015 |
forehead | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -9.10e-03 |
Pos logit contributions | |
udge | +0.072 |
eter | +0.064 |
ipples | +0.062 |
ill | +0.061 |
urious | +0.059 |
Neg logit contributions | |
goodbye | -0.062 |
Amen | -0.051 |
fs | -0.050 |
saddle | -0.048 |
yes | -0.047 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -4.12e-03 |
Pos logit contributions | |
goodbye | +0.014 |
Amen | +0.010 |
fs | +0.010 |
| +0.010 |
yes | +0.010 |
Neg logit contributions | |
udge | -0.023 |
ipples | -0.021 |
eter | -0.020 |
ased | -0.019 |
ill | -0.019 |
Token | Once |
Token ID | 7454 |
Feature activation | +3.26e-04 |
Pos logit contributions | |
goodbye | +0.009 |
Amen | +0.007 |
fs | +0.007 |
yes | +0.007 |
saddle | +0.007 |
Neg logit contributions | |
udge | -0.008 |
eter | -0.007 |
ipples | -0.007 |
urious | -0.006 |
ill | -0.006 |
Token | was |
Token ID | 373 |
Feature activation | +3.16e-03 |
Pos logit contributions | |
goodbye | +0.012 |
fs | +0.010 |
Amen | +0.009 |
yes | +0.009 |
saddle | +0.009 |
Neg logit contributions | |
udge | -0.013 |
eter | -0.011 |
ipples | -0.011 |
urious | -0.011 |
elly | -0.011 |
Token | a |
Token ID | 257 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
udge | +0.007 |
eter | +0.006 |
ipples | +0.006 |
ill | +0.006 |
ased | +0.006 |
Neg logit contributions | |
goodbye | -0.006 |
Amen | -0.005 |
fs | -0.005 |
forehead | -0.004 |
ishly | -0.004 |
Token | box |
Token ID | 3091 |
Feature activation | -1.43e-02 |
Pos logit contributions | |
goodbye | +0.023 |
fs | +0.020 |
ishly | +0.018 |
Amen | +0.018 |
chief | +0.017 |
Neg logit contributions | |
udge | -0.026 |
eter | -0.022 |
elly | -0.021 |
ipples | -0.021 |
Like | -0.021 |
Token | of |
Token ID | 286 |
Feature activation | +3.17e-02 |
Pos logit contributions | |
goodbye | +0.030 |
Amen | +0.025 |
fs | +0.025 |
forehead | +0.023 |
saddle | +0.023 |
Neg logit contributions | |
udge | -0.028 |
eter | -0.024 |
ill | -0.023 |
ipples | -0.023 |
elly | -0.022 |
Token | ch |
Token ID | 442 |
Feature activation | +1.76e-02 |
Pos logit contributions | |
udge | +0.065 |
eter | +0.057 |
ipples | +0.055 |
elly | +0.053 |
ased | +0.053 |
Neg logit contributions | |
goodbye | -0.060 |
Amen | -0.049 |
fs | -0.048 |
chief | -0.045 |
yes | -0.045 |
Token | ocol |
Token ID | 4668 |
Feature activation | +9.81e-02 |
Pos logit contributions | |
udge | +0.027 |
urious | +0.026 |
eter | +0.026 |
ipples | +0.025 |
Like | +0.025 |
Neg logit contributions | |
goodbye | -0.037 |
fs | -0.031 |
on | -0.030 |
microphone | -0.030 |
saddle | -0.030 |
Token | ates |
Token ID | 689 |
Feature activation | -3.19e-02 |
Pos logit contributions | |
udge | +0.229 |
eter | +0.205 |
ipples | +0.201 |
urious | +0.196 |
elly | +0.191 |
Neg logit contributions | |
goodbye | -0.156 |
fs | -0.127 |
Amen | -0.122 |
saddle | -0.116 |
yes | -0.114 |
Token | . |
Token ID | 13 |
Feature activation | +1.83e-02 |
Pos logit contributions | |
goodbye | +0.070 |
Amen | +0.063 |
fs | +0.056 |
forehead | +0.055 |
pron | +0.055 |
Neg logit contributions | |
udge | -0.059 |
ill | -0.050 |
ipples | -0.050 |
eter | -0.049 |
elly | -0.046 |
Token | Lily |
Token ID | 20037 |
Feature activation | +6.57e-03 |
Pos logit contributions | |
udge | +0.040 |
eter | +0.036 |
ipples | +0.036 |
urious | +0.034 |
ased | +0.033 |
Neg logit contributions | |
goodbye | -0.034 |
Amen | -0.026 |
yes | -0.026 |
fs | -0.026 |
| -0.024 |
Token | loved |
Token ID | 6151 |
Feature activation | -8.15e-03 |
Pos logit contributions | |
udge | +0.014 |
eter | +0.012 |
ipples | +0.012 |
elly | +0.012 |
urious | +0.012 |
Neg logit contributions | |
goodbye | -0.012 |
yes | -0.010 |
fs | -0.010 |
Amen | -0.010 |
saddle | -0.009 |
Token | ch |
Token ID | 442 |
Feature activation | -5.15e-03 |
Pos logit contributions | |
goodbye | +0.015 |
fs | +0.012 |
Amen | +0.012 |
forehead | +0.012 |
yes | +0.011 |
Neg logit contributions | |
udge | -0.017 |
eter | -0.015 |
ipples | -0.015 |
ill | -0.014 |
ased | -0.014 |
Token | that |
Token ID | 326 |
Feature activation | -2.93e-02 |
Pos logit contributions | |
goodbye | +0.011 |
fs | +0.010 |
Amen | +0.009 |
pron | +0.009 |
forehead | +0.009 |
Neg logit contributions | |
udge | -0.017 |
ipples | -0.014 |
eter | -0.014 |
urious | -0.014 |
ased | -0.014 |
Token | day |
Token ID | 1110 |
Feature activation | -4.24e-02 |
Pos logit contributions | |
goodbye | +0.067 |
fs | +0.062 |
saddle | +0.061 |
forehead | +0.058 |
clipboard | +0.058 |
Neg logit contributions | |
udge | -0.042 |
cy | -0.034 |
elly | -0.033 |
ill | -0.033 |
aught | -0.033 |
Token | on |
Token ID | 319 |
Feature activation | -3.40e-02 |
Pos logit contributions | |
pron | +0.069 |
Amen | +0.067 |
forehead | +0.067 |
goodbye | +0.063 |
blanket | +0.062 |
Neg logit contributions | |
udge | -0.083 |
ierce | -0.073 |
ased | -0.072 |
ook | -0.071 |
imb | -0.070 |
Token | , |
Token ID | 11 |
Feature activation | -2.33e-02 |
Pos logit contributions | |
goodbye | +0.054 |
Amen | +0.050 |
forehead | +0.048 |
fs | +0.047 |
saddle | +0.046 |
Neg logit contributions | |
udge | -0.074 |
eter | -0.065 |
ill | -0.064 |
cy | -0.062 |
elly | -0.062 |
Token | Cr |
Token ID | 3864 |
Feature activation | +5.35e-02 |
Pos logit contributions | |
goodbye | +0.040 |
fs | +0.031 |
Amen | +0.031 |
yes | +0.029 |
forehead | +0.029 |
Neg logit contributions | |
udge | -0.054 |
eter | -0.048 |
ipples | -0.047 |
ill | -0.045 |
urious | -0.045 |
Token | ust |
Token ID | 436 |
Feature activation | +9.78e-02 |
Pos logit contributions | |
udge | +0.061 |
Esk | +0.051 |
ased | +0.049 |
eter | +0.045 |
ipples | +0.044 |
Neg logit contributions | |
goodbye | -0.120 |
medal | -0.111 |
fs | -0.108 |
magazine | -0.104 |
on | -0.102 |
Token | y |
Token ID | 88 |
Feature activation | -1.78e-02 |
Pos logit contributions | |
udge | +0.162 |
eter | +0.133 |
ipples | +0.122 |
ased | +0.116 |
ill | +0.116 |
Neg logit contributions | |
goodbye | -0.224 |
Amen | -0.183 |
medal | -0.178 |
yes | -0.172 |
saddle | -0.171 |
Token | tried |
Token ID | 3088 |
Feature activation | -1.45e-02 |
Pos logit contributions | |
goodbye | +0.036 |
fs | +0.030 |
Amen | +0.029 |
yes | +0.028 |
saddle | +0.028 |
Neg logit contributions | |
udge | -0.035 |
eter | -0.031 |
ipples | -0.031 |
ill | -0.029 |
urious | -0.029 |
Token | his |
Token ID | 465 |
Feature activation | +2.96e-03 |
Pos logit contributions | |
goodbye | +0.024 |
fs | +0.019 |
Amen | +0.018 |
yes | +0.018 |
saddle | +0.017 |
Neg logit contributions | |
udge | -0.034 |
eter | -0.031 |
ipples | -0.030 |
ill | -0.029 |
urious | -0.029 |
Token | best |
Token ID | 1266 |
Feature activation | -6.27e-03 |
Pos logit contributions | |
udge | +0.006 |
eter | +0.005 |
ipples | +0.005 |
urious | +0.005 |
asured | +0.005 |
Neg logit contributions | |
goodbye | -0.006 |
Amen | -0.005 |
fs | -0.005 |
pron | -0.005 |
yes | -0.004 |
Token | to |
Token ID | 284 |
Feature activation | +1.04e-02 |
Pos logit contributions | |
goodbye | +0.011 |
Amen | +0.010 |
fs | +0.009 |
forehead | +0.009 |
pron | +0.009 |
Neg logit contributions | |
udge | -0.014 |
eter | -0.012 |
ipples | -0.012 |
ill | -0.011 |
urious | -0.011 |
Token | saw |
Token ID | 2497 |
Feature activation | +6.04e-03 |
Pos logit contributions | |
udge | +0.020 |
eter | +0.018 |
ipples | +0.018 |
ierce | +0.017 |
elly | +0.017 |
Neg logit contributions | |
goodbye | -0.018 |
yes | -0.014 |
veil | -0.014 |
fs | -0.013 |
on | -0.013 |
Token | all |
Token ID | 477 |
Feature activation | +1.24e-02 |
Pos logit contributions | |
udge | +0.015 |
eter | +0.014 |
ipples | +0.014 |
urious | +0.013 |
ill | +0.013 |
Neg logit contributions | |
goodbye | -0.009 |
fs | -0.007 |
Amen | -0.007 |
yes | -0.006 |
saddle | -0.006 |
Token | the |
Token ID | 262 |
Feature activation | -1.04e-03 |
Pos logit contributions | |
udge | +0.028 |
eter | +0.025 |
ipples | +0.025 |
ill | +0.024 |
urious | +0.023 |
Neg logit contributions | |
goodbye | -0.022 |
fs | -0.017 |
Amen | -0.017 |
yes | -0.017 |
saddle | -0.016 |
Token | tasty |
Token ID | 25103 |
Feature activation | +5.65e-03 |
Pos logit contributions | |
goodbye | +0.002 |
fs | +0.001 |
Amen | +0.001 |
yes | +0.001 |
saddle | +0.001 |
Neg logit contributions | |
udge | -0.003 |
eter | -0.002 |
ipples | -0.002 |
ill | -0.002 |
ierce | -0.002 |
Token | ch |
Token ID | 442 |
Feature activation | +1.60e-02 |
Pos logit contributions | |
udge | +0.015 |
eter | +0.013 |
ipples | +0.013 |
urious | +0.013 |
ill | +0.013 |
Neg logit contributions | |
goodbye | -0.008 |
fs | -0.006 |
saddle | -0.006 |
medal | -0.006 |
pony | -0.006 |
Token | ocol |
Token ID | 4668 |
Feature activation | +9.73e-02 |
Pos logit contributions | |
udge | +0.025 |
eter | +0.024 |
urious | +0.023 |
ipples | +0.023 |
Like | +0.022 |
Neg logit contributions | |
goodbye | -0.035 |
fs | -0.030 |
on | -0.029 |
medal | -0.028 |
saddle | -0.028 |
Token | ates |
Token ID | 689 |
Feature activation | -2.78e-02 |
Pos logit contributions | |
udge | +0.229 |
eter | +0.207 |
ipples | +0.200 |
urious | +0.198 |
elly | +0.193 |
Neg logit contributions | |
goodbye | -0.151 |
fs | -0.124 |
Amen | -0.116 |
yes | -0.110 |
saddle | -0.110 |
Token | inside |
Token ID | 2641 |
Feature activation | +1.29e-02 |
Pos logit contributions | |
goodbye | +0.056 |
Amen | +0.052 |
fs | +0.046 |
saddle | +0.045 |
pron | +0.045 |
Neg logit contributions | |
udge | -0.054 |
ill | -0.048 |
ipples | -0.047 |
eter | -0.047 |
rying | -0.043 |
Token | . |
Token ID | 13 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
udge | +0.028 |
eter | +0.024 |
ipples | +0.024 |
ill | +0.023 |
urious | +0.023 |
Neg logit contributions | |
goodbye | -0.024 |
Amen | -0.020 |
fs | -0.019 |
saddle | -0.018 |
yes | -0.018 |
Token | She |
Token ID | 1375 |
Feature activation | +7.32e-03 |
Pos logit contributions | |
goodbye | +0.022 |
fs | +0.018 |
Amen | +0.017 |
yes | +0.017 |
saddle | +0.016 |
Neg logit contributions | |
udge | -0.029 |
eter | -0.026 |
ipples | -0.025 |
ill | -0.025 |
urious | -0.024 |
Token | couldn |
Token ID | 3521 |
Feature activation | -2.68e-02 |
Pos logit contributions | |
udge | +0.016 |
eter | +0.014 |
ipples | +0.014 |
ill | +0.013 |
urious | +0.013 |
Neg logit contributions | |
goodbye | -0.014 |
fs | -0.011 |
Amen | -0.011 |
yes | -0.011 |
saddle | -0.010 |
Token | as |
Token ID | 355 |
Feature activation | +2.55e-02 |
Pos logit contributions | |
goodbye | +0.018 |
Amen | +0.017 |
pron | +0.016 |
fs | +0.016 |
headset | +0.015 |
Neg logit contributions | |
udge | -0.014 |
eter | -0.013 |
ipples | -0.012 |
ill | -0.011 |
ased | -0.011 |
Token | he |
Token ID | 339 |
Feature activation | +1.97e-02 |
Pos logit contributions | |
udge | +0.066 |
eter | +0.059 |
ipples | +0.058 |
ill | +0.057 |
ierce | +0.057 |
Neg logit contributions | |
goodbye | -0.034 |
Amen | -0.027 |
fs | -0.026 |
forehead | -0.024 |
saddle | -0.024 |
Token | opened |
Token ID | 4721 |
Feature activation | +1.51e-02 |
Pos logit contributions | |
udge | +0.042 |
eter | +0.037 |
ipples | +0.037 |
ill | +0.035 |
urious | +0.035 |
Neg logit contributions | |
goodbye | -0.037 |
fs | -0.029 |
Amen | -0.029 |
yes | -0.028 |
saddle | -0.028 |
Token | the |
Token ID | 262 |
Feature activation | -6.45e-03 |
Pos logit contributions | |
udge | +0.033 |
eter | +0.029 |
ipples | +0.028 |
ill | +0.027 |
ased | +0.027 |
Neg logit contributions | |
goodbye | -0.027 |
Amen | -0.023 |
fs | -0.023 |
forehead | -0.021 |
yes | -0.021 |
Token | ch |
Token ID | 442 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
goodbye | +0.011 |
fs | +0.010 |
Amen | +0.009 |
yes | +0.009 |
Midnight | +0.009 |
Neg logit contributions | |
udge | -0.013 |
eter | -0.012 |
urious | -0.011 |
elly | -0.011 |
ipples | -0.011 |
Token | ocol |
Token ID | 4668 |
Feature activation | +9.69e-02 |
Pos logit contributions | |
udge | +0.024 |
eter | +0.020 |
ipples | +0.020 |
ill | +0.018 |
urious | +0.018 |
Neg logit contributions | |
goodbye | -0.041 |
fs | -0.035 |
Amen | -0.035 |
yes | -0.034 |
saddle | -0.033 |
Token | ates |
Token ID | 689 |
Feature activation | -3.03e-02 |
Pos logit contributions | |
udge | +0.222 |
eter | +0.198 |
ipples | +0.194 |
urious | +0.188 |
ill | +0.185 |
Neg logit contributions | |
goodbye | -0.163 |
fs | -0.130 |
Amen | -0.127 |
saddle | -0.121 |
yes | -0.121 |
Token | and |
Token ID | 290 |
Feature activation | -2.30e-02 |
Pos logit contributions | |
goodbye | +0.063 |
Amen | +0.061 |
fs | +0.053 |
Midnight | +0.053 |
pron | +0.052 |
Neg logit contributions | |
udge | -0.055 |
ill | -0.051 |
eter | -0.047 |
ipples | -0.047 |
rying | -0.044 |
Token | started |
Token ID | 2067 |
Feature activation | +9.54e-03 |
Pos logit contributions | |
goodbye | +0.042 |
fs | +0.034 |
Amen | +0.034 |
yes | +0.032 |
pron | +0.032 |
Neg logit contributions | |
udge | -0.049 |
eter | -0.044 |
ipples | -0.043 |
ill | -0.041 |
ased | -0.041 |
Token | eating |
Token ID | 6600 |
Feature activation | -6.17e-03 |
Pos logit contributions | |
udge | +0.019 |
eter | +0.017 |
ipples | +0.017 |
ased | +0.016 |
ill | +0.016 |
Neg logit contributions | |
goodbye | -0.017 |
fs | -0.016 |
Amen | -0.015 |
yes | -0.014 |
forehead | -0.014 |
Token | them |
Token ID | 606 |
Feature activation | +1.89e-03 |
Pos logit contributions | |
goodbye | +0.011 |
Amen | +0.009 |
fs | +0.009 |
forehead | +0.008 |
pron | +0.008 |
Neg logit contributions | |
udge | -0.014 |
eter | -0.012 |
ipples | -0.012 |
ill | -0.012 |
ased | -0.011 |
Token | sky |
Token ID | 6766 |
Feature activation | +3.68e-03 |
Pos logit contributions | |
udge | +0.001 |
ipples | +0.001 |
eter | +0.001 |
urious | +0.001 |
ill | +0.001 |
Neg logit contributions | |
goodbye | -0.001 |
fs | -0.001 |
Amen | -0.001 |
yes | -0.001 |
pron | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -6.92e-03 |
Pos logit contributions | |
udge | +0.008 |
eter | +0.007 |
ipples | +0.007 |
ill | +0.007 |
urious | +0.007 |
Neg logit contributions | |
goodbye | -0.007 |
fs | -0.005 |
Amen | -0.005 |
yes | -0.005 |
saddle | -0.005 |
Token | She |
Token ID | 1375 |
Feature activation | +8.37e-03 |
Pos logit contributions | |
goodbye | +0.012 |
Amen | +0.009 |
fs | +0.009 |
yes | +0.009 |
| +0.008 |
Neg logit contributions | |
udge | -0.016 |
ipples | -0.014 |
eter | -0.014 |
urious | -0.014 |
ased | -0.014 |
Token | felt |
Token ID | 2936 |
Feature activation | -5.18e-03 |
Pos logit contributions | |
udge | +0.019 |
eter | +0.017 |
ipples | +0.017 |
elly | +0.016 |
ierce | +0.016 |
Neg logit contributions | |
goodbye | -0.015 |
Amen | -0.012 |
yes | -0.011 |
fs | -0.011 |
saddle | -0.011 |
Token | it |
Token ID | 340 |
Feature activation | +3.19e-03 |
Pos logit contributions | |
goodbye | +0.009 |
fs | +0.008 |
Amen | +0.008 |
pron | +0.007 |
saddle | +0.007 |
Neg logit contributions | |
udge | -0.011 |
ipples | -0.010 |
eter | -0.010 |
ill | -0.009 |
urious | -0.009 |
Token | r |
Token ID | 374 |
Feature activation | +9.69e-02 |
Pos logit contributions | |
udge | +0.006 |
ipples | +0.006 |
eter | +0.006 |
ill | +0.005 |
urious | +0.005 |
Neg logit contributions | |
goodbye | -0.006 |
Amen | -0.005 |
fs | -0.005 |
saddle | -0.005 |
yes | -0.005 |
Token | umbling |
Token ID | 14739 |
Feature activation | +4.50e-02 |
Pos logit contributions | |
udge | +0.130 |
Like | +0.125 |
urious | +0.121 |
bet | +0.115 |
eter | +0.115 |
Neg logit contributions | |
fs | -0.162 |
goodbye | -0.160 |
agging | -0.155 |
on | -0.153 |
ishly | -0.152 |
Token | beneath |
Token ID | 11061 |
Feature activation | +1.45e-02 |
Pos logit contributions | |
udge | +0.105 |
eter | +0.092 |
ipples | +0.091 |
urious | +0.087 |
ill | +0.087 |
Neg logit contributions | |
goodbye | -0.080 |
fs | -0.058 |
Amen | -0.058 |
yes | -0.057 |
forehead | -0.057 |
Token | her |
Token ID | 607 |
Feature activation | +3.68e-02 |
Pos logit contributions | |
udge | +0.033 |
eter | +0.030 |
ipples | +0.028 |
ill | +0.028 |
urious | +0.027 |
Neg logit contributions | |
goodbye | -0.024 |
Amen | -0.020 |
fs | -0.020 |
saddle | -0.019 |
pron | -0.019 |
Token | feet |
Token ID | 3625 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
udge | +0.074 |
eter | +0.065 |
ipples | +0.062 |
ill | +0.061 |
urious | +0.058 |
Neg logit contributions | |
goodbye | -0.073 |
Amen | -0.067 |
fs | -0.062 |
pron | -0.060 |
yes | -0.059 |
Token | . |
Token ID | 13 |
Feature activation | -7.84e-03 |
Pos logit contributions | |
goodbye | +0.020 |
Amen | +0.018 |
fs | +0.016 |
pron | +0.016 |
saddle | +0.015 |
Neg logit contributions | |
udge | -0.021 |
ill | -0.018 |
eter | -0.018 |
ipples | -0.018 |
ased | -0.017 |
Token | apped |
Token ID | 6320 |
Feature activation | -1.64e-02 |
Pos logit contributions | |
goodbye | +0.002 |
ishly | +0.002 |
Amen | +0.002 |
yes | +0.002 |
microphone | +0.002 |
Neg logit contributions | |
udge | -0.003 |
urious | -0.003 |
ipples | -0.003 |
Like | -0.003 |
ierce | -0.003 |
Token | and |
Token ID | 290 |
Feature activation | -2.74e-02 |
Pos logit contributions | |
goodbye | +0.031 |
fs | +0.026 |
Amen | +0.026 |
saddle | +0.024 |
pron | +0.024 |
Neg logit contributions | |
udge | -0.034 |
eter | -0.030 |
ipples | -0.030 |
ill | -0.029 |
ased | -0.028 |
Token | cheered |
Token ID | 34767 |
Feature activation | -3.37e-02 |
Pos logit contributions | |
goodbye | +0.047 |
Amen | +0.040 |
fs | +0.040 |
pron | +0.039 |
saddle | +0.038 |
Neg logit contributions | |
udge | -0.059 |
eter | -0.053 |
ipples | -0.053 |
ill | -0.051 |
ased | -0.049 |
Token | . |
Token ID | 13 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
goodbye | +0.059 |
fs | +0.047 |
Amen | +0.046 |
yes | +0.044 |
saddle | +0.044 |
Neg logit contributions | |
udge | -0.077 |
eter | -0.068 |
ipples | -0.067 |
ill | -0.064 |
urious | -0.064 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.42e-03 |
Pos logit contributions | |
goodbye | +0.035 |
fs | +0.027 |
Amen | +0.027 |
yes | +0.026 |
| +0.026 |
Neg logit contributions | |
udge | -0.035 |
eter | -0.030 |
ipples | -0.030 |
ill | -0.028 |
urious | -0.028 |
Token | and |
Token ID | 290 |
Feature activation | -8.17e-02 |
Pos logit contributions | |
udge | +0.003 |
eter | +0.002 |
ipples | +0.002 |
ill | +0.002 |
urious | +0.002 |
Neg logit contributions | |
goodbye | -0.003 |
Amen | -0.003 |
fs | -0.003 |
pron | -0.002 |
yes | -0.002 |
Token | Tim |
Token ID | 5045 |
Feature activation | +4.30e-04 |
Pos logit contributions | |
goodbye | +0.159 |
fs | +0.137 |
Amen | +0.134 |
pron | +0.128 |
saddle | +0.127 |
Neg logit contributions | |
udge | -0.164 |
eter | -0.143 |
ipples | -0.140 |
urious | -0.135 |
ill | -0.134 |
Token | felt |
Token ID | 2936 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
udge | +0.001 |
eter | +0.001 |
ipples | +0.001 |
ill | +0.001 |
urious | +0.001 |
Neg logit contributions | |
goodbye | -0.001 |
fs | -0.001 |
Amen | -0.001 |
yes | -0.001 |
saddle | -0.001 |
Token | very |
Token ID | 845 |
Feature activation | -1.00e-02 |
Pos logit contributions | |
goodbye | +0.018 |
fs | +0.017 |
Amen | +0.016 |
pron | +0.016 |
saddle | +0.016 |
Neg logit contributions | |
udge | -0.023 |
ipples | -0.021 |
eter | -0.020 |
ill | -0.019 |
ased | -0.018 |
Token | happy |
Token ID | 3772 |
Feature activation | -4.35e-03 |
Pos logit contributions | |
goodbye | +0.015 |
fs | +0.012 |
Amen | +0.012 |
yes | +0.011 |
saddle | +0.011 |
Neg logit contributions | |
udge | -0.025 |
eter | -0.022 |
ipples | -0.022 |
ill | -0.021 |
urious | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | -1.39e-02 |
Pos logit contributions | |
goodbye | +0.008 |
fs | +0.006 |
Amen | +0.006 |
saddle | +0.006 |
pron | +0.006 |
Neg logit contributions | |
udge | -0.010 |
eter | -0.009 |
ipples | -0.008 |
ill | -0.008 |
elly | -0.008 |
Token | And |
Token ID | 1870 |
Feature activation | -2.81e-02 |
Pos logit contributions | |
goodbye | +0.024 |
fs | +0.020 |
Amen | +0.019 |
yes | +0.019 |
saddle | +0.018 |
Neg logit contributions | |
udge | -0.021 |
eter | -0.018 |
ipples | -0.018 |
ill | -0.017 |
urious | -0.017 |
Token | so |
Token ID | 523 |
Feature activation | -4.95e-02 |
Pos logit contributions | |
goodbye | +0.043 |
Amen | +0.038 |
fs | +0.038 |
pron | +0.036 |
saddle | +0.035 |
Neg logit contributions | |
udge | -0.064 |
eter | -0.059 |
ipples | -0.056 |
ill | -0.055 |
ased | -0.053 |
Token | , |
Token ID | 11 |
Feature activation | -2.81e-02 |
Pos logit contributions | |
goodbye | +0.075 |
fs | +0.057 |
Amen | +0.056 |
yes | +0.053 |
saddle | +0.052 |
Neg logit contributions | |
udge | -0.124 |
eter | -0.112 |
ipples | -0.110 |
ill | -0.106 |
urious | -0.105 |
Token | Mom |
Token ID | 11254 |
Feature activation | +1.33e-02 |
Pos logit contributions | |
goodbye | +0.053 |
fs | +0.044 |
Amen | +0.043 |
saddle | +0.041 |
yes | +0.040 |
Neg logit contributions | |
udge | -0.058 |
eter | -0.052 |
ipples | -0.051 |
ill | -0.048 |
urious | -0.047 |
Token | my |
Token ID | 1820 |
Feature activation | -6.96e-03 |
Pos logit contributions | |
udge | +0.023 |
eter | +0.020 |
ipples | +0.019 |
ill | +0.018 |
urious | +0.018 |
Neg logit contributions | |
goodbye | -0.031 |
fs | -0.026 |
Amen | -0.026 |
yes | -0.025 |
saddle | -0.025 |
Token | and |
Token ID | 290 |
Feature activation | -7.82e-02 |
Pos logit contributions | |
goodbye | +0.013 |
fs | +0.011 |
Amen | +0.011 |
yes | +0.010 |
forehead | +0.010 |
Neg logit contributions | |
udge | -0.014 |
eter | -0.013 |
ipples | -0.013 |
ill | -0.012 |
ased | -0.012 |
Token | Baby |
Token ID | 14801 |
Feature activation | -1.60e-02 |
Pos logit contributions | |
goodbye | +0.165 |
fs | +0.138 |
Amen | +0.135 |
yes | +0.130 |
saddle | +0.129 |
Neg logit contributions | |
udge | -0.148 |
eter | -0.130 |
ipples | -0.127 |
ill | -0.120 |
urious | -0.120 |
Token | continued |
Token ID | 3767 |
Feature activation | +1.63e-03 |
Pos logit contributions | |
goodbye | +0.030 |
Amen | +0.024 |
fs | +0.024 |
yes | +0.023 |
saddle | +0.023 |
Neg logit contributions | |
udge | -0.035 |
eter | -0.030 |
ipples | -0.030 |
urious | -0.029 |
ill | -0.028 |
Token | their |
Token ID | 511 |
Feature activation | +1.45e-02 |
Pos logit contributions | |
udge | +0.004 |
eter | +0.003 |
ipples | +0.003 |
ased | +0.003 |
ierce | +0.003 |
Neg logit contributions | |
goodbye | -0.003 |
fs | -0.002 |
Amen | -0.002 |
forehead | -0.002 |
pron | -0.002 |
Token | walk |
Token ID | 2513 |
Feature activation | +8.70e-03 |
Pos logit contributions | |
udge | +0.031 |
eter | +0.028 |
ipples | +0.028 |
urious | +0.027 |
ierce | +0.026 |
Neg logit contributions | |
goodbye | -0.025 |
Amen | -0.021 |
fs | -0.021 |
pron | -0.020 |
yes | -0.019 |
Token | down |
Token ID | 866 |
Feature activation | +3.85e-02 |
Pos logit contributions | |
udge | +0.021 |
eter | +0.019 |
ipples | +0.019 |
urious | +0.018 |
ill | +0.018 |
Neg logit contributions | |
goodbye | -0.014 |
fs | -0.010 |
Amen | -0.010 |
yes | -0.010 |
saddle | -0.010 |
Token | that |
Token ID | 326 |
Feature activation | -6.31e-03 |
Pos logit contributions | |
goodbye | +0.019 |
Amen | +0.018 |
saddle | +0.017 |
fs | +0.017 |
forehead | +0.017 |
Neg logit contributions | |
udge | -0.025 |
eter | -0.023 |
ill | -0.023 |
ipples | -0.022 |
ierce | -0.022 |
Token | special |
Token ID | 2041 |
Feature activation | +3.82e-03 |
Pos logit contributions | |
goodbye | +0.011 |
fs | +0.010 |
Amen | +0.009 |
saddle | +0.009 |
forehead | +0.009 |
Neg logit contributions | |
udge | -0.013 |
eter | -0.012 |
ipples | -0.012 |
ill | -0.011 |
urious | -0.011 |
Token | day |
Token ID | 1110 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
udge | +0.007 |
eter | +0.006 |
ipples | +0.006 |
urious | +0.006 |
elly | +0.006 |
Neg logit contributions | |
goodbye | -0.007 |
fs | -0.007 |
Amen | -0.006 |
forehead | -0.006 |
yes | -0.006 |
Token | when |
Token ID | 618 |
Feature activation | +2.04e-02 |
Pos logit contributions | |
goodbye | +0.033 |
Amen | +0.030 |
fs | +0.029 |
saddle | +0.028 |
forehead | +0.028 |
Neg logit contributions | |
udge | -0.033 |
eter | -0.029 |
ill | -0.028 |
ipples | -0.027 |
ased | -0.026 |
Token | mom |
Token ID | 1995 |
Feature activation | +4.45e-03 |
Pos logit contributions | |
udge | +0.054 |
eter | +0.049 |
ipples | +0.048 |
ill | +0.046 |
urious | +0.046 |
Neg logit contributions | |
goodbye | -0.028 |
fs | -0.021 |
Amen | -0.021 |
yes | -0.019 |
saddle | -0.019 |
Token | and |
Token ID | 290 |
Feature activation | -7.82e-02 |
Pos logit contributions | |
udge | +0.010 |
eter | +0.009 |
ipples | +0.008 |
ill | +0.008 |
urious | +0.008 |
Neg logit contributions | |
goodbye | -0.008 |
fs | -0.007 |
Amen | -0.007 |
yes | -0.006 |
saddle | -0.006 |
Token | dad |
Token ID | 9955 |
Feature activation | -1.89e-02 |
Pos logit contributions | |
goodbye | +0.141 |
fs | +0.124 |
Amen | +0.122 |
yes | +0.109 |
forehead | +0.109 |
Neg logit contributions | |
udge | -0.165 |
eter | -0.149 |
ill | -0.141 |
ipples | -0.140 |
ased | -0.138 |
Token | said |
Token ID | 531 |
Feature activation | -3.62e-02 |
Pos logit contributions | |
goodbye | +0.035 |
fs | +0.030 |
Amen | +0.029 |
forehead | +0.027 |
saddle | +0.027 |
Neg logit contributions | |
udge | -0.039 |
eter | -0.034 |
ipples | -0.034 |
ill | -0.033 |
ased | -0.032 |
Token | " |
Token ID | 366 |
Feature activation | +3.96e-02 |
Pos logit contributions | |
fs | +0.046 |
saddle | +0.046 |
Amen | +0.044 |
pron | +0.043 |
forehead | +0.042 |
Neg logit contributions | |
udge | -0.076 |
eter | -0.074 |
ill | -0.070 |
ipples | -0.068 |
ased | -0.067 |
Token | I |
Token ID | 40 |
Feature activation | +1.01e-02 |
Pos logit contributions | |
udge | +0.074 |
ipples | +0.062 |
eter | +0.060 |
urious | +0.059 |
ierce | +0.058 |
Neg logit contributions | |
goodbye | -0.081 |
fs | -0.069 |
Amen | -0.068 |
yes | -0.068 |
pron | -0.067 |
Token | do |
Token ID | 466 |
Feature activation | -3.85e-02 |
Pos logit contributions | |
udge | +0.019 |
eter | +0.016 |
ipples | +0.016 |
urious | +0.015 |
ill | +0.015 |
Neg logit contributions | |
goodbye | -0.022 |
Amen | -0.018 |
fs | -0.018 |
yes | -0.018 |
saddle | -0.017 |
Token | was |
Token ID | 373 |
Feature activation | -1.72e-02 |
Pos logit contributions | |
goodbye | +0.003 |
fs | +0.003 |
Amen | +0.003 |
saddle | +0.003 |
yes | +0.003 |
Neg logit contributions | |
udge | -0.003 |
eter | -0.003 |
ipples | -0.003 |
ill | -0.003 |
urious | -0.003 |
Token | so |
Token ID | 523 |
Feature activation | -2.56e-02 |
Pos logit contributions | |
goodbye | +0.028 |
fs | +0.027 |
saddle | +0.023 |
Amen | +0.023 |
medal | +0.023 |
Neg logit contributions | |
udge | -0.038 |
ipples | -0.034 |
eter | -0.034 |
ill | -0.033 |
ased | -0.032 |
Token | glad |
Token ID | 9675 |
Feature activation | -8.01e-03 |
Pos logit contributions | |
goodbye | +0.038 |
fs | +0.030 |
Amen | +0.029 |
saddle | +0.027 |
yes | +0.027 |
Neg logit contributions | |
udge | -0.065 |
eter | -0.058 |
ipples | -0.058 |
ill | -0.056 |
urious | -0.055 |
Token | that |
Token ID | 326 |
Feature activation | -2.12e-02 |
Pos logit contributions | |
goodbye | +0.011 |
fs | +0.009 |
Amen | +0.009 |
saddle | +0.008 |
pron | +0.008 |
Neg logit contributions | |
udge | -0.020 |
eter | -0.019 |
ipples | -0.018 |
ill | -0.018 |
urious | -0.017 |
Token | Mom |
Token ID | 11254 |
Feature activation | +2.01e-03 |
Pos logit contributions | |
goodbye | +0.021 |
Amen | +0.012 |
fs | +0.012 |
yes | +0.012 |
forehead | +0.010 |
Neg logit contributions | |
udge | -0.066 |
eter | -0.060 |
ipples | -0.060 |
urious | -0.058 |
elly | -0.057 |
Token | and |
Token ID | 290 |
Feature activation | -7.80e-02 |
Pos logit contributions | |
udge | +0.004 |
eter | +0.004 |
ill | +0.003 |
ipples | +0.003 |
ased | +0.003 |
Neg logit contributions | |
goodbye | -0.004 |
fs | -0.003 |
Amen | -0.003 |
saddle | -0.003 |
yes | -0.003 |
Token | Dad |
Token ID | 17415 |
Feature activation | -1.70e-02 |
Pos logit contributions | |
goodbye | +0.176 |
fs | +0.154 |
Amen | +0.151 |
saddle | +0.145 |
yes | +0.142 |
Neg logit contributions | |
udge | -0.135 |
eter | -0.116 |
ipples | -0.114 |
ill | -0.110 |
urious | -0.107 |
Token | had |
Token ID | 550 |
Feature activation | -2.11e-02 |
Pos logit contributions | |
goodbye | +0.033 |
Amen | +0.028 |
fs | +0.028 |
saddle | +0.026 |
pron | +0.026 |
Neg logit contributions | |
udge | -0.034 |
eter | -0.030 |
ipples | -0.030 |
ill | -0.029 |
ased | -0.028 |
Token | not |
Token ID | 407 |
Feature activation | +2.49e-02 |
Pos logit contributions | |
goodbye | +0.041 |
Amen | +0.034 |
fs | +0.034 |
saddle | +0.032 |
forehead | +0.031 |
Neg logit contributions | |
udge | -0.043 |
eter | -0.039 |
ipples | -0.038 |
ill | -0.036 |
ased | -0.036 |
Token | been |
Token ID | 587 |
Feature activation | +2.42e-02 |
Pos logit contributions | |
udge | +0.050 |
eter | +0.043 |
ipples | +0.042 |
urious | +0.041 |
elly | +0.040 |
Neg logit contributions | |
goodbye | -0.052 |
yes | -0.041 |
Amen | -0.040 |
fs | -0.040 |
forehead | -0.039 |
Token | scared |
Token ID | 12008 |
Feature activation | +1.74e-03 |
Pos logit contributions | |
udge | +0.061 |
eter | +0.055 |
ipples | +0.054 |
urious | +0.053 |
ill | +0.052 |
Neg logit contributions | |
goodbye | -0.037 |
yes | -0.026 |
fs | -0.026 |
Amen | -0.026 |
pron | -0.025 |
Token | ." |
Token ID | 526 |
Feature activation | -1.15e-02 |
Pos logit contributions | |
udge | +0.080 |
ipples | +0.069 |
eter | +0.069 |
urious | +0.067 |
ased | +0.065 |
Neg logit contributions | |
goodbye | -0.057 |
Amen | -0.045 |
yes | -0.043 |
fs | -0.043 |
forehead | -0.040 |
Token |
|
Token ID | 198 |
Feature activation | -6.63e-03 |
Pos logit contributions | |
goodbye | +0.020 |
fs | +0.016 |
Amen | +0.016 |
yes | +0.015 |
saddle | +0.015 |
Neg logit contributions | |
udge | -0.026 |
eter | -0.023 |
ipples | -0.023 |
urious | -0.022 |
ill | -0.022 |
Token |
|
Token ID | 198 |
Feature activation | -4.52e-03 |
Pos logit contributions | |
goodbye | +0.012 |
fs | +0.009 |
Amen | +0.009 |
yes | +0.009 |
saddle | +0.009 |
Neg logit contributions | |
udge | -0.015 |
eter | -0.013 |
ipples | -0.013 |
ill | -0.012 |
urious | -0.012 |
Token | So |
Token ID | 2396 |
Feature activation | -2.84e-02 |
Pos logit contributions | |
goodbye | +0.009 |
fs | +0.008 |
Amen | +0.008 |
yes | +0.008 |
saddle | +0.008 |
Neg logit contributions | |
udge | -0.008 |
eter | -0.007 |
ipples | -0.007 |
ased | -0.007 |
ill | -0.007 |
Token | Mom |
Token ID | 11254 |
Feature activation | +1.06e-02 |
Pos logit contributions | |
goodbye | +0.050 |
Amen | +0.042 |
fs | +0.041 |
forehead | +0.039 |
pron | +0.038 |
Neg logit contributions | |
udge | -0.063 |
eter | -0.057 |
ipples | -0.055 |
ill | -0.053 |
urious | -0.052 |
Token | and |
Token ID | 290 |
Feature activation | -7.76e-02 |
Pos logit contributions | |
udge | +0.021 |
eter | +0.019 |
ipples | +0.018 |
ill | +0.018 |
ased | +0.018 |
Neg logit contributions | |
goodbye | -0.021 |
fs | -0.018 |
Amen | -0.018 |
yes | -0.016 |
medal | -0.016 |
Token | Sammy |
Token ID | 44067 |
Feature activation | -1.04e-02 |
Pos logit contributions | |
goodbye | +0.176 |
fs | +0.146 |
Amen | +0.146 |
yes | +0.141 |
saddle | +0.139 |
Neg logit contributions | |
udge | -0.138 |
eter | -0.119 |
ipples | -0.116 |
ill | -0.110 |
urious | -0.109 |
Token | went |
Token ID | 1816 |
Feature activation | +9.30e-03 |
Pos logit contributions | |
goodbye | +0.021 |
Amen | +0.017 |
fs | +0.017 |
yes | +0.016 |
forehead | +0.016 |
Neg logit contributions | |
udge | -0.021 |
eter | -0.019 |
ipples | -0.018 |
ill | -0.018 |
urious | -0.017 |
Token | back |
Token ID | 736 |
Feature activation | +4.08e-02 |
Pos logit contributions | |
udge | +0.018 |
ipples | +0.015 |
eter | +0.015 |
ill | +0.015 |
ased | +0.015 |
Neg logit contributions | |
goodbye | -0.019 |
fs | -0.016 |
Amen | -0.016 |
pron | -0.016 |
yes | -0.015 |
Token | inside |
Token ID | 2641 |
Feature activation | +1.91e-02 |
Pos logit contributions | |
udge | +0.109 |
eter | +0.102 |
ipples | +0.100 |
roc | +0.098 |
urious | +0.097 |
Neg logit contributions | |
goodbye | -0.058 |
blanket | -0.039 |
| -0.037 |
bed | -0.035 |
saddle | -0.034 |
Token | and |
Token ID | 290 |
Feature activation | -3.15e-02 |
Pos logit contributions | |
udge | +0.047 |
eter | +0.043 |
ipples | +0.042 |
urious | +0.041 |
elly | +0.040 |
Neg logit contributions | |
goodbye | -0.030 |
fs | -0.022 |
yes | -0.022 |
Amen | -0.021 |
forehead | -0.020 |
Token | And |
Token ID | 1870 |
Feature activation | -2.72e-02 |
Pos logit contributions | |
goodbye | +0.031 |
fs | +0.025 |
Amen | +0.025 |
yes | +0.024 |
forehead | +0.024 |
Neg logit contributions | |
udge | -0.030 |
eter | -0.026 |
ipples | -0.025 |
ill | -0.024 |
urious | -0.024 |
Token | so |
Token ID | 523 |
Feature activation | -5.19e-02 |
Pos logit contributions | |
goodbye | +0.043 |
Amen | +0.037 |
fs | +0.036 |
pron | +0.035 |
saddle | +0.035 |
Neg logit contributions | |
udge | -0.063 |
eter | -0.057 |
ill | -0.054 |
ipples | -0.054 |
ased | -0.052 |
Token | , |
Token ID | 11 |
Feature activation | -3.00e-02 |
Pos logit contributions | |
goodbye | +0.077 |
Amen | +0.059 |
saddle | +0.058 |
fs | +0.058 |
forehead | +0.055 |
Neg logit contributions | |
udge | -0.130 |
eter | -0.117 |
ipples | -0.113 |
ill | -0.111 |
urious | -0.110 |
Token | Mom |
Token ID | 11254 |
Feature activation | +1.50e-02 |
Pos logit contributions | |
goodbye | +0.053 |
Amen | +0.044 |
fs | +0.043 |
saddle | +0.042 |
forehead | +0.041 |
Neg logit contributions | |
udge | -0.066 |
eter | -0.060 |
ipples | -0.056 |
ill | -0.055 |
cy | -0.054 |
Token | my |
Token ID | 1820 |
Feature activation | -4.88e-03 |
Pos logit contributions | |
udge | +0.024 |
eter | +0.021 |
ipples | +0.020 |
ill | +0.020 |
ased | +0.019 |
Neg logit contributions | |
goodbye | -0.036 |
fs | -0.031 |
Amen | -0.030 |
yes | -0.029 |
saddle | -0.029 |
Token | and |
Token ID | 290 |
Feature activation | -7.73e-02 |
Pos logit contributions | |
goodbye | +0.009 |
fs | +0.008 |
Amen | +0.008 |
forehead | +0.007 |
yes | +0.007 |
Neg logit contributions | |
udge | -0.010 |
eter | -0.009 |
ipples | -0.009 |
ill | -0.008 |
ased | -0.008 |
Token | Courtney |
Token ID | 36677 |
Feature activation | -6.08e-03 |
Pos logit contributions | |
goodbye | +0.149 |
fs | +0.118 |
Amen | +0.118 |
yes | +0.114 |
forehead | +0.111 |
Neg logit contributions | |
udge | -0.164 |
eter | -0.144 |
ipples | -0.143 |
ill | -0.135 |
urious | -0.135 |
Token | gathered |
Token ID | 9272 |
Feature activation | +1.61e-02 |
Pos logit contributions | |
goodbye | +0.012 |
fs | +0.011 |
Amen | +0.010 |
pron | +0.010 |
magazine | +0.010 |
Neg logit contributions | |
udge | -0.011 |
eter | -0.010 |
ill | -0.010 |
ipples | -0.010 |
ased | -0.010 |
Token | up |
Token ID | 510 |
Feature activation | +2.36e-02 |
Pos logit contributions | |
udge | +0.038 |
ipples | +0.034 |
eter | +0.034 |
ill | +0.033 |
urious | +0.033 |
Neg logit contributions | |
goodbye | -0.027 |
fs | -0.020 |
yes | -0.019 |
Amen | -0.019 |
saddle | -0.018 |
Token | their |
Token ID | 511 |
Feature activation | +2.03e-02 |
Pos logit contributions | |
udge | +0.059 |
ipples | +0.053 |
eter | +0.052 |
urious | +0.051 |
elly | +0.050 |
Neg logit contributions | |
goodbye | -0.038 |
yes | -0.027 |
fs | -0.027 |
Amen | -0.025 |
medal | -0.024 |
Token | things |
Token ID | 1243 |
Feature activation | +8.43e-03 |
Pos logit contributions | |
udge | +0.047 |
ipples | +0.043 |
eter | +0.042 |
ill | +0.039 |
elly | +0.039 |
Neg logit contributions | |
goodbye | -0.036 |
fs | -0.027 |
yes | -0.027 |
magazine | -0.026 |
microphone | -0.025 |
Token | tired |
Token ID | 10032 |
Feature activation | +3.11e-02 |
Pos logit contributions | |
udge | +0.008 |
ipples | +0.008 |
eter | +0.008 |
ill | +0.007 |
ased | +0.007 |
Neg logit contributions | |
goodbye | -0.006 |
fs | -0.006 |
Amen | -0.005 |
saddle | -0.005 |
clipboard | -0.005 |
Token | but |
Token ID | 475 |
Feature activation | -7.17e-03 |
Pos logit contributions | |
udge | +0.081 |
urious | +0.072 |
eter | +0.072 |
ipples | +0.071 |
roc | +0.069 |
Neg logit contributions | |
goodbye | -0.048 |
yes | -0.033 |
fs | -0.032 |
blanket | -0.031 |
forehead | -0.031 |
Token | happy |
Token ID | 3772 |
Feature activation | -7.92e-03 |
Pos logit contributions | |
goodbye | +0.012 |
fs | +0.010 |
Amen | +0.010 |
saddle | +0.009 |
pron | +0.009 |
Neg logit contributions | |
udge | -0.016 |
eter | -0.015 |
ipples | -0.014 |
ill | -0.014 |
urious | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
goodbye | +0.013 |
fs | +0.010 |
Amen | +0.010 |
yes | +0.010 |
saddle | +0.009 |
Neg logit contributions | |
udge | -0.019 |
eter | -0.017 |
ipples | -0.017 |
ill | -0.016 |
urious | -0.016 |
Token | Lucy |
Token ID | 22162 |
Feature activation | +8.26e-04 |
Pos logit contributions | |
goodbye | +0.022 |
Amen | +0.017 |
fs | +0.016 |
| +0.016 |
yes | +0.016 |
Neg logit contributions | |
udge | -0.030 |
ipples | -0.027 |
eter | -0.026 |
ased | -0.025 |
ill | -0.025 |
Token | and |
Token ID | 290 |
Feature activation | -7.72e-02 |
Pos logit contributions | |
udge | +0.001 |
eter | +0.001 |
ipples | +0.001 |
ill | +0.001 |
urious | +0.001 |
Neg logit contributions | |
goodbye | -0.002 |
fs | -0.002 |
Amen | -0.001 |
saddle | -0.001 |
magazine | -0.001 |
Token | her |
Token ID | 607 |
Feature activation | +2.96e-03 |
Pos logit contributions | |
goodbye | +0.133 |
fs | +0.117 |
Amen | +0.108 |
pron | +0.108 |
yes | +0.108 |
Neg logit contributions | |
udge | -0.167 |
eter | -0.149 |
ipples | -0.140 |
urious | -0.140 |
ill | -0.139 |
Token | mom |
Token ID | 1995 |
Feature activation | +4.29e-03 |
Pos logit contributions | |
udge | +0.006 |
eter | +0.006 |
ipples | +0.006 |
urious | +0.005 |
ill | +0.005 |
Neg logit contributions | |
goodbye | -0.005 |
Amen | -0.005 |
fs | -0.004 |
yes | -0.004 |
forehead | -0.004 |
Token | went |
Token ID | 1816 |
Feature activation | +1.61e-03 |
Pos logit contributions | |
udge | +0.008 |
eter | +0.007 |
ill | +0.007 |
ipples | +0.007 |
cy | +0.007 |
Neg logit contributions | |
goodbye | -0.008 |
fs | -0.007 |
Amen | -0.007 |
pron | -0.007 |
medal | -0.007 |
Token | home |
Token ID | 1363 |
Feature activation | +2.69e-02 |
Pos logit contributions | |
udge | +0.003 |
ipples | +0.002 |
eter | +0.002 |
ook | +0.002 |
rying | +0.002 |
Neg logit contributions | |
Amen | -0.003 |
fs | -0.003 |
goodbye | -0.003 |
pron | -0.003 |
yes | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | +1.38e-02 |
Pos logit contributions | |
udge | +0.062 |
eter | +0.056 |
ipples | +0.054 |
ill | +0.053 |
ased | +0.051 |
Neg logit contributions | |
goodbye | -0.044 |
fs | -0.036 |
Amen | -0.036 |
saddle | -0.033 |
pron | -0.033 |
Token | We |
Token ID | 775 |
Feature activation | -1.17e-02 |
Pos logit contributions | |
udge | +0.070 |
ipples | +0.063 |
urious | +0.062 |
eter | +0.060 |
ill | +0.059 |
Neg logit contributions | |
goodbye | -0.058 |
Amen | -0.049 |
yes | -0.043 |
Midnight | -0.042 |
| -0.042 |
Token | love |
Token ID | 1842 |
Feature activation | -1.35e-02 |
Pos logit contributions | |
goodbye | +0.026 |
fs | +0.022 |
Amen | +0.022 |
forehead | +0.021 |
saddle | +0.021 |
Neg logit contributions | |
udge | -0.021 |
ipples | -0.018 |
eter | -0.018 |
ill | -0.017 |
ased | -0.016 |
Token | you |
Token ID | 345 |
Feature activation | +1.40e-03 |
Pos logit contributions | |
goodbye | +0.024 |
fs | +0.020 |
Amen | +0.019 |
forehead | +0.019 |
saddle | +0.019 |
Neg logit contributions | |
udge | -0.028 |
eter | -0.026 |
ipples | -0.025 |
urious | -0.024 |
ill | -0.024 |
Token | ." |
Token ID | 526 |
Feature activation | -3.77e-03 |
Pos logit contributions | |
eter | +0.002 |
udge | +0.002 |
ipples | +0.002 |
cy | +0.002 |
ill | +0.002 |
Neg logit contributions | |
fs | -0.003 |
goodbye | -0.003 |
chief | -0.003 |
saddle | -0.003 |
veil | -0.002 |
Token | Lily |
Token ID | 20037 |
Feature activation | +4.74e-03 |
Pos logit contributions | |
goodbye | +0.009 |
fs | +0.008 |
saddle | +0.008 |
pron | +0.008 |
Amen | +0.008 |
Neg logit contributions | |
udge | -0.006 |
eter | -0.005 |
elly | -0.005 |
ill | -0.005 |
ipples | -0.005 |
Token | and |
Token ID | 290 |
Feature activation | -7.71e-02 |
Pos logit contributions | |
udge | +0.008 |
eter | +0.007 |
ill | +0.007 |
ipples | +0.007 |
urious | +0.006 |
Neg logit contributions | |
goodbye | -0.010 |
fs | -0.010 |
Amen | -0.009 |
pron | -0.009 |
yes | -0.009 |
Token | Max |
Token ID | 5436 |
Feature activation | -1.13e-02 |
Pos logit contributions | |
goodbye | +0.145 |
pron | +0.140 |
fs | +0.133 |
Amen | +0.131 |
forehead | +0.126 |
Neg logit contributions | |
udge | -0.124 |
isk | -0.114 |
eter | -0.113 |
roc | -0.110 |
ierce | -0.109 |
Token | feel |
Token ID | 1254 |
Feature activation | -3.90e-03 |
Pos logit contributions | |
fs | +0.019 |
goodbye | +0.019 |
pron | +0.018 |
Amen | +0.017 |
saddle | +0.015 |
Neg logit contributions | |
udge | -0.022 |
ipples | -0.021 |
eter | -0.020 |
ill | -0.020 |
cy | -0.020 |
Token | warm |
Token ID | 5814 |
Feature activation | +2.41e-02 |
Pos logit contributions | |
goodbye | +0.006 |
fs | +0.006 |
pron | +0.005 |
saddle | +0.005 |
chief | +0.005 |
Neg logit contributions | |
udge | -0.008 |
ipples | -0.008 |
eter | -0.008 |
elly | -0.007 |
ased | -0.007 |
Token | and |
Token ID | 290 |
Feature activation | -3.85e-02 |
Pos logit contributions | |
udge | +0.039 |
eter | +0.036 |
ipples | +0.035 |
ill | +0.033 |
cy | +0.032 |
Neg logit contributions | |
fs | -0.050 |
goodbye | -0.049 |
pron | -0.046 |
Amen | -0.044 |
medal | -0.043 |
Token | cozy |
Token ID | 37438 |
Feature activation | +3.00e-02 |
Pos logit contributions | |
goodbye | +0.065 |
fs | +0.062 |
pron | +0.055 |
saddle | +0.054 |
Amen | +0.053 |
Neg logit contributions | |
udge | -0.077 |
ipples | -0.071 |
eter | -0.070 |
cy | -0.065 |
urious | -0.063 |
Token | were |
Token ID | 547 |
Feature activation | -3.16e-02 |
Pos logit contributions | |
goodbye | +0.053 |
fs | +0.046 |
Amen | +0.045 |
saddle | +0.042 |
forehead | +0.042 |
Neg logit contributions | |
udge | -0.052 |
ill | -0.047 |
ipples | -0.046 |
eter | -0.046 |
cy | -0.045 |
Token | so |
Token ID | 523 |
Feature activation | -2.35e-02 |
Pos logit contributions | |
goodbye | +0.051 |
fs | +0.049 |
Amen | +0.044 |
saddle | +0.043 |
medal | +0.042 |
Neg logit contributions | |
udge | -0.066 |
eter | -0.063 |
ipples | -0.062 |
ill | -0.059 |
ased | -0.058 |
Token | delicious |
Token ID | 12625 |
Feature activation | -8.80e-03 |
Pos logit contributions | |
goodbye | +0.037 |
fs | +0.029 |
Amen | +0.028 |
saddle | +0.027 |
yes | +0.026 |
Neg logit contributions | |
udge | -0.057 |
eter | -0.051 |
ipples | -0.051 |
ill | -0.049 |
urious | -0.049 |
Token | ! |
Token ID | 0 |
Feature activation | -1.05e-02 |
Pos logit contributions | |
goodbye | +0.016 |
fs | +0.013 |
Amen | +0.013 |
yes | +0.012 |
saddle | +0.012 |
Neg logit contributions | |
udge | -0.019 |
eter | -0.017 |
ipples | -0.017 |
ill | -0.016 |
urious | -0.016 |
Token | Lucy |
Token ID | 22162 |
Feature activation | +6.16e-03 |
Pos logit contributions | |
goodbye | +0.021 |
fs | +0.017 |
Amen | +0.017 |
yes | +0.016 |
saddle | +0.016 |
Neg logit contributions | |
udge | -0.021 |
eter | -0.019 |
ipples | -0.018 |
ill | -0.017 |
urious | -0.017 |
Token | and |
Token ID | 290 |
Feature activation | -7.71e-02 |
Pos logit contributions | |
udge | +0.011 |
eter | +0.011 |
ipples | +0.010 |
ill | +0.010 |
urious | +0.009 |
Neg logit contributions | |
goodbye | -0.012 |
fs | -0.011 |
Amen | -0.011 |
saddle | -0.010 |
yes | -0.010 |
Token | her |
Token ID | 607 |
Feature activation | +7.07e-03 |
Pos logit contributions | |
goodbye | +0.148 |
fs | +0.142 |
pron | +0.133 |
Amen | +0.132 |
saddle | +0.132 |
Neg logit contributions | |
udge | -0.138 |
eter | -0.117 |
ill | -0.113 |
urious | -0.111 |
cy | -0.111 |
Token | friends |
Token ID | 2460 |
Feature activation | -2.64e-02 |
Pos logit contributions | |
udge | +0.013 |
eter | +0.012 |
ipples | +0.012 |
urious | +0.011 |
ill | +0.011 |
Neg logit contributions | |
goodbye | -0.014 |
fs | -0.012 |
Amen | -0.012 |
saddle | -0.011 |
chief | -0.011 |
Token | ate |
Token ID | 15063 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
goodbye | +0.052 |
fs | +0.051 |
Amen | +0.049 |
pron | +0.046 |
magazine | +0.045 |
Neg logit contributions | |
udge | -0.042 |
ill | -0.039 |
cy | -0.039 |
eter | -0.037 |
ipples | -0.037 |
Token | the |
Token ID | 262 |
Feature activation | -1.58e-02 |
Pos logit contributions | |
goodbye | +0.022 |
Amen | +0.020 |
fs | +0.020 |
saddle | +0.019 |
headset | +0.018 |
Neg logit contributions | |
udge | -0.026 |
eter | -0.024 |
ipples | -0.022 |
ill | -0.022 |
cy | -0.022 |
Token | sund |
Token ID | 37437 |
Feature activation | +2.39e-02 |
Pos logit contributions | |
goodbye | +0.025 |
Amen | +0.024 |
fs | +0.023 |
chief | +0.022 |
saddle | +0.022 |
Neg logit contributions | |
udge | -0.029 |
cy | -0.028 |
eter | -0.028 |
ipples | -0.027 |
urious | -0.027 |
Token | find |
Token ID | 1064 |
Feature activation | +6.03e-03 |
Pos logit contributions | |
udge | +0.042 |
eter | +0.037 |
ipples | +0.036 |
urious | +0.034 |
ased | +0.034 |
Neg logit contributions | |
goodbye | -0.039 |
Amen | -0.032 |
fs | -0.031 |
yes | -0.031 |
forehead | -0.030 |
Token | my |
Token ID | 616 |
Feature activation | -3.96e-03 |
Pos logit contributions | |
udge | +0.014 |
eter | +0.012 |
ipples | +0.012 |
ill | +0.011 |
urious | +0.011 |
Neg logit contributions | |
goodbye | -0.010 |
fs | -0.008 |
Amen | -0.008 |
yes | -0.008 |
pron | -0.008 |
Token | home |
Token ID | 1363 |
Feature activation | +1.13e-02 |
Pos logit contributions | |
goodbye | +0.007 |
fs | +0.006 |
Amen | +0.006 |
pron | +0.006 |
yes | +0.006 |
Neg logit contributions | |
udge | -0.008 |
eter | -0.008 |
ipples | -0.007 |
urious | -0.007 |
elly | -0.007 |
Token | ." |
Token ID | 526 |
Feature activation | -8.01e-03 |
Pos logit contributions | |
udge | +0.023 |
eter | +0.020 |
ipples | +0.020 |
ill | +0.020 |
elly | +0.019 |
Neg logit contributions | |
goodbye | -0.022 |
Amen | -0.018 |
fs | -0.018 |
saddle | -0.017 |
yes | -0.017 |
Token | Mia |
Token ID | 32189 |
Feature activation | +8.36e-03 |
Pos logit contributions | |
goodbye | +0.020 |
fs | +0.016 |
Amen | +0.016 |
yes | +0.016 |
| +0.016 |
Neg logit contributions | |
udge | -0.013 |
eter | -0.011 |
ipples | -0.011 |
urious | -0.010 |
ased | -0.010 |
Token | and |
Token ID | 290 |
Feature activation | -7.70e-02 |
Pos logit contributions | |
udge | +0.018 |
eter | +0.016 |
ill | +0.016 |
ipples | +0.016 |
urious | +0.016 |
Neg logit contributions | |
goodbye | -0.015 |
fs | -0.012 |
Amen | -0.012 |
pron | -0.011 |
saddle | -0.011 |
Token | Max |
Token ID | 5436 |
Feature activation | -1.04e-02 |
Pos logit contributions | |
goodbye | +0.143 |
pron | +0.136 |
fs | +0.136 |
Amen | +0.133 |
saddle | +0.126 |
Neg logit contributions | |
udge | -0.128 |
eter | -0.116 |
roc | -0.115 |
cy | -0.113 |
urious | -0.113 |
Token | decided |
Token ID | 3066 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
goodbye | +0.020 |
Amen | +0.017 |
fs | +0.017 |
pron | +0.017 |
medal | +0.016 |
Neg logit contributions | |
udge | -0.021 |
eter | -0.018 |
ipples | -0.018 |
ill | -0.017 |
urious | -0.017 |
Token | to |
Token ID | 284 |
Feature activation | -3.28e-03 |
Pos logit contributions | |
Amen | +0.027 |
pron | +0.027 |
fs | +0.026 |
goodbye | +0.025 |
magazine | +0.024 |
Neg logit contributions | |
udge | -0.019 |
eter | -0.018 |
ill | -0.017 |
rying | -0.017 |
ook | -0.016 |
Token | help |
Token ID | 1037 |
Feature activation | -1.91e-02 |
Pos logit contributions | |
goodbye | +0.007 |
fs | +0.005 |
Amen | +0.005 |
yes | +0.005 |
pron | +0.005 |
Neg logit contributions | |
udge | -0.007 |
eter | -0.006 |
ipples | -0.006 |
ill | -0.005 |
urious | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -3.99e-02 |
Pos logit contributions | |
goodbye | +0.031 |
fs | +0.023 |
Amen | +0.023 |
yes | +0.022 |
forehead | +0.021 |
Neg logit contributions | |
udge | -0.046 |
eter | -0.042 |
ipples | -0.042 |
urious | -0.040 |
elly | -0.040 |
Token | Everybody |
Token ID | 23080 |
Feature activation | -1.29e-04 |
Pos logit contributions | |
goodbye | +0.014 |
fs | +0.011 |
Amen | +0.011 |
yes | +0.011 |
saddle | +0.011 |
Neg logit contributions | |
udge | -0.015 |
eter | -0.013 |
ipples | -0.013 |
ill | -0.012 |
urious | -0.012 |
Token | cheered |
Token ID | 34767 |
Feature activation | -3.39e-02 |
Pos logit contributions | |
goodbye | +0.000 |
fs | +0.000 |
Amen | +0.000 |
saddle | +0.000 |
forehead | +0.000 |
Neg logit contributions | |
udge | -0.000 |
eter | -0.000 |
ipples | -0.000 |
ill | -0.000 |
ased | -0.000 |
Token | ! |
Token ID | 0 |
Feature activation | -9.95e-03 |
Pos logit contributions | |
goodbye | +0.063 |
fs | +0.054 |
Amen | +0.053 |
saddle | +0.051 |
pron | +0.049 |
Neg logit contributions | |
udge | -0.070 |
eter | -0.063 |
ipples | -0.060 |
ill | -0.059 |
ased | -0.058 |
Token | Mom |
Token ID | 11254 |
Feature activation | +1.58e-02 |
Pos logit contributions | |
goodbye | +0.016 |
fs | +0.012 |
Amen | +0.012 |
saddle | +0.011 |
yes | +0.011 |
Neg logit contributions | |
udge | -0.024 |
eter | -0.022 |
ipples | -0.022 |
ill | -0.021 |
urious | -0.021 |
Token | my |
Token ID | 1820 |
Feature activation | -1.67e-03 |
Pos logit contributions | |
udge | +0.024 |
eter | +0.020 |
ill | +0.020 |
ipples | +0.019 |
ased | +0.018 |
Neg logit contributions | |
goodbye | -0.040 |
fs | -0.035 |
Amen | -0.034 |
yes | -0.033 |
saddle | -0.032 |
Token | and |
Token ID | 290 |
Feature activation | -7.70e-02 |
Pos logit contributions | |
goodbye | +0.003 |
fs | +0.003 |
Amen | +0.003 |
saddle | +0.003 |
magazine | +0.003 |
Neg logit contributions | |
udge | -0.003 |
eter | -0.003 |
ill | -0.003 |
ipples | -0.003 |
ased | -0.003 |
Token | Daddy |
Token ID | 26926 |
Feature activation | -1.57e-02 |
Pos logit contributions | |
goodbye | +0.167 |
fs | +0.158 |
saddle | +0.149 |
Amen | +0.147 |
medal | +0.138 |
Neg logit contributions | |
udge | -0.127 |
eter | -0.110 |
ill | -0.108 |
ipples | -0.104 |
ased | -0.103 |
Token | were |
Token ID | 547 |
Feature activation | -3.03e-02 |
Pos logit contributions | |
goodbye | +0.028 |
fs | +0.024 |
Amen | +0.024 |
pron | +0.022 |
magazine | +0.022 |
Neg logit contributions | |
udge | -0.033 |
ipples | -0.029 |
eter | -0.029 |
ill | -0.029 |
ased | -0.028 |
Token | married |
Token ID | 6405 |
Feature activation | -2.38e-02 |
Pos logit contributions | |
goodbye | +0.049 |
fs | +0.045 |
saddle | +0.041 |
Amen | +0.041 |
Midnight | +0.039 |
Neg logit contributions | |
udge | -0.066 |
ipples | -0.060 |
eter | -0.060 |
ill | -0.059 |
ased | -0.056 |
Token | now |
Token ID | 783 |
Feature activation | -1.79e-02 |
Pos logit contributions | |
goodbye | +0.045 |
Amen | +0.040 |
fs | +0.039 |
saddle | +0.037 |
chief | +0.036 |
Neg logit contributions | |
udge | -0.047 |
eter | -0.043 |
ipples | -0.040 |
ill | -0.040 |
ased | -0.038 |
Token | and |
Token ID | 290 |
Feature activation | -4.78e-02 |
Pos logit contributions | |
goodbye | +0.034 |
fs | +0.031 |
Amen | +0.031 |
saddle | +0.029 |
pron | +0.029 |
Neg logit contributions | |
udge | -0.034 |
eter | -0.030 |
ipples | -0.028 |
ased | -0.027 |
ill | -0.027 |
Token | " |
Token ID | 1 |
Feature activation | +4.77e-02 |
Pos logit contributions | |
udge | +0.023 |
eter | +0.020 |
ipples | +0.020 |
ill | +0.019 |
urious | +0.019 |
Neg logit contributions | |
goodbye | -0.020 |
fs | -0.016 |
Amen | -0.016 |
yes | -0.015 |
forehead | -0.015 |
Token | Oh |
Token ID | 5812 |
Feature activation | -9.10e-03 |
Pos logit contributions | |
udge | +0.089 |
eter | +0.079 |
ipples | +0.075 |
ierce | +0.072 |
urious | +0.072 |
Neg logit contributions | |
goodbye | -0.099 |
fs | -0.086 |
Amen | -0.085 |
pron | -0.084 |
yes | -0.084 |
Token | no |
Token ID | 645 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
goodbye | +0.015 |
fs | +0.014 |
saddle | +0.014 |
Amen | +0.013 |
clipboard | +0.013 |
Neg logit contributions | |
udge | -0.018 |
eter | -0.017 |
ill | -0.016 |
ipples | -0.015 |
Esk | -0.015 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.87e-02 |
Pos logit contributions | |
udge | +0.028 |
eter | +0.025 |
ipples | +0.025 |
urious | +0.024 |
ill | +0.024 |
Neg logit contributions | |
goodbye | -0.019 |
Amen | -0.014 |
fs | -0.014 |
yes | -0.014 |
forehead | -0.013 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.51e-02 |
Pos logit contributions | |
goodbye | +0.043 |
fs | +0.037 |
saddle | +0.037 |
Amen | +0.037 |
pron | +0.037 |
Neg logit contributions | |
udge | -0.030 |
eter | -0.025 |
ipples | -0.024 |
ill | -0.023 |
ased | -0.023 |
Token | and |
Token ID | 290 |
Feature activation | -7.69e-02 |
Pos logit contributions | |
udge | +0.036 |
eter | +0.032 |
ill | +0.031 |
ipples | +0.031 |
urious | +0.030 |
Neg logit contributions | |
goodbye | -0.025 |
Amen | -0.020 |
fs | -0.019 |
saddle | -0.018 |
pron | -0.018 |
Token | Max |
Token ID | 5436 |
Feature activation | -2.49e-03 |
Pos logit contributions | |
goodbye | +0.122 |
Amen | +0.098 |
fs | +0.098 |
pron | +0.093 |
saddle | +0.092 |
Neg logit contributions | |
udge | -0.182 |
eter | -0.163 |
ill | -0.155 |
ipples | -0.154 |
ierce | -0.153 |
Token | said |
Token ID | 531 |
Feature activation | -1.44e-02 |
Pos logit contributions | |
goodbye | +0.004 |
Amen | +0.003 |
fs | +0.003 |
pron | +0.003 |
saddle | +0.003 |
Neg logit contributions | |
udge | -0.006 |
eter | -0.005 |
ipples | -0.005 |
ill | -0.005 |
ased | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | +1.97e-02 |
Pos logit contributions | |
goodbye | +0.029 |
fs | +0.028 |
pron | +0.028 |
saddle | +0.028 |
Amen | +0.028 |
Neg logit contributions | |
udge | -0.025 |
eter | -0.021 |
ill | -0.019 |
ipples | -0.019 |
aught | -0.019 |
Token | " |
Token ID | 366 |
Feature activation | +4.94e-02 |
Pos logit contributions | |
udge | +0.042 |
eter | +0.037 |
ipples | +0.036 |
ill | +0.035 |
elly | +0.034 |
Neg logit contributions | |
goodbye | -0.037 |
fs | -0.030 |
Amen | -0.029 |
saddle | -0.029 |
medal | -0.028 |
Token | Now |
Token ID | 3844 |
Feature activation | -3.42e-02 |
Pos logit contributions | |
udge | +0.098 |
eter | +0.093 |
rying | +0.086 |
ierce | +0.086 |
ook | +0.086 |
Neg logit contributions | |
goodbye | -0.080 |
fs | -0.076 |
Amen | -0.073 |
Midnight | -0.070 |
magazine | -0.069 |
Token | him |
Token ID | 683 |
Feature activation | +1.86e-02 |
Pos logit contributions | |
goodbye | +0.032 |
fs | +0.029 |
pron | +0.027 |
Amen | +0.027 |
saddle | +0.027 |
Neg logit contributions | |
udge | -0.036 |
eter | -0.033 |
ipples | -0.032 |
ill | -0.031 |
ierce | -0.030 |
Token | sleep |
Token ID | 3993 |
Feature activation | -8.86e-04 |
Pos logit contributions | |
udge | +0.038 |
eter | +0.034 |
ipples | +0.034 |
ill | +0.032 |
urious | +0.032 |
Neg logit contributions | |
goodbye | -0.034 |
fs | -0.029 |
Amen | -0.028 |
pron | -0.027 |
yes | -0.026 |
Token | ." |
Token ID | 526 |
Feature activation | -1.23e-02 |
Pos logit contributions | |
goodbye | +0.002 |
fs | +0.002 |
chief | +0.001 |
Amen | +0.001 |
saddle | +0.001 |
Neg logit contributions | |
udge | -0.002 |
eter | -0.001 |
ipples | -0.001 |
ill | -0.001 |
elly | -0.001 |
Token | So |
Token ID | 1406 |
Feature activation | -2.09e-02 |
Pos logit contributions | |
goodbye | +0.026 |
| +0.022 |
yes | +0.020 |
fs | +0.019 |
blanket | +0.019 |
Neg logit contributions | |
udge | -0.024 |
eter | -0.020 |
ipples | -0.020 |
ased | -0.019 |
asured | -0.019 |
Token | Mia |
Token ID | 32189 |
Feature activation | +9.48e-03 |
Pos logit contributions | |
goodbye | +0.036 |
fs | +0.027 |
Amen | +0.027 |
yes | +0.026 |
forehead | +0.025 |
Neg logit contributions | |
udge | -0.049 |
eter | -0.043 |
ipples | -0.043 |
urious | -0.041 |
ill | -0.041 |
Token | and |
Token ID | 290 |
Feature activation | -7.67e-02 |
Pos logit contributions | |
udge | +0.017 |
ipples | +0.016 |
eter | +0.016 |
ill | +0.016 |
urious | +0.015 |
Neg logit contributions | |
goodbye | -0.018 |
fs | -0.016 |
Amen | -0.016 |
magazine | -0.015 |
microphone | -0.015 |
Token | her |
Token ID | 607 |
Feature activation | +9.84e-03 |
Pos logit contributions | |
fs | +0.142 |
goodbye | +0.140 |
pron | +0.130 |
Amen | +0.128 |
saddle | +0.125 |
Neg logit contributions | |
udge | -0.141 |
eter | -0.125 |
ipples | -0.124 |
ill | -0.121 |
urious | -0.121 |
Token | mom |
Token ID | 1995 |
Feature activation | +5.15e-03 |
Pos logit contributions | |
udge | +0.020 |
eter | +0.019 |
ipples | +0.018 |
urious | +0.018 |
ill | +0.018 |
Neg logit contributions | |
goodbye | -0.017 |
fs | -0.015 |
Amen | -0.015 |
saddle | -0.014 |
chief | -0.014 |
Token | sat |
Token ID | 3332 |
Feature activation | +4.04e-02 |
Pos logit contributions | |
udge | +0.010 |
ipples | +0.009 |
eter | +0.009 |
ill | +0.009 |
ased | +0.008 |
Neg logit contributions | |
goodbye | -0.010 |
fs | -0.009 |
Amen | -0.008 |
pron | -0.008 |
yes | -0.008 |
Token | on |
Token ID | 319 |
Feature activation | -5.20e-03 |
Pos logit contributions | |
udge | +0.091 |
eter | +0.081 |
ipples | +0.081 |
ased | +0.077 |
urious | +0.076 |
Neg logit contributions | |
goodbye | -0.070 |
fs | -0.058 |
Amen | -0.055 |
yes | -0.053 |
pron | -0.053 |
Token | the |
Token ID | 262 |
Feature activation | -1.01e-02 |
Pos logit contributions | |
goodbye | +0.010 |
fs | +0.010 |
Amen | +0.009 |
pron | +0.009 |
forehead | +0.009 |
Neg logit contributions | |
udge | -0.009 |
eter | -0.008 |
ill | -0.008 |
ipples | -0.007 |
cy | -0.007 |
Token | ." |
Token ID | 526 |
Feature activation | -7.07e-03 |
Pos logit contributions | |
goodbye | +0.026 |
fs | +0.021 |
Amen | +0.021 |
yes | +0.020 |
saddle | +0.019 |
Neg logit contributions | |
udge | -0.031 |
eter | -0.028 |
ipples | -0.027 |
ill | -0.026 |
urious | -0.026 |
Token | And |
Token ID | 843 |
Feature activation | -2.28e-02 |
Pos logit contributions | |
goodbye | +0.013 |
fs | +0.011 |
Amen | +0.011 |
yes | +0.011 |
saddle | +0.010 |
Neg logit contributions | |
udge | -0.015 |
eter | -0.013 |
ipples | -0.013 |
ill | -0.012 |
elly | -0.012 |
Token | so |
Token ID | 523 |
Feature activation | -4.14e-02 |
Pos logit contributions | |
goodbye | +0.041 |
Amen | +0.038 |
fs | +0.037 |
pron | +0.036 |
saddle | +0.034 |
Neg logit contributions | |
udge | -0.045 |
eter | -0.041 |
ill | -0.038 |
ipples | -0.038 |
ased | -0.037 |
Token | , |
Token ID | 11 |
Feature activation | -2.14e-02 |
Pos logit contributions | |
goodbye | +0.060 |
Amen | +0.047 |
fs | +0.046 |
saddle | +0.044 |
forehead | +0.043 |
Neg logit contributions | |
udge | -0.103 |
eter | -0.094 |
ipples | -0.092 |
ill | -0.090 |
urious | -0.088 |
Token | Lily |
Token ID | 20037 |
Feature activation | +3.67e-03 |
Pos logit contributions | |
goodbye | +0.037 |
Amen | +0.035 |
fs | +0.033 |
saddle | +0.032 |
forehead | +0.032 |
Neg logit contributions | |
udge | -0.042 |
eter | -0.040 |
ill | -0.037 |
ipples | -0.037 |
cy | -0.035 |
Token | and |
Token ID | 290 |
Feature activation | -7.66e-02 |
Pos logit contributions | |
udge | +0.007 |
eter | +0.007 |
ipples | +0.007 |
ill | +0.006 |
ased | +0.006 |
Neg logit contributions | |
goodbye | -0.007 |
Amen | -0.006 |
fs | -0.006 |
yes | -0.005 |
pron | -0.005 |
Token | her |
Token ID | 607 |
Feature activation | +4.51e-03 |
Pos logit contributions | |
fs | +0.153 |
goodbye | +0.143 |
pron | +0.133 |
saddle | +0.131 |
Amen | +0.130 |
Neg logit contributions | |
udge | -0.123 |
eter | -0.116 |
cy | -0.109 |
urious | -0.106 |
roc | -0.105 |
Token | mom |
Token ID | 1995 |
Feature activation | +6.88e-03 |
Pos logit contributions | |
udge | +0.009 |
eter | +0.009 |
ipples | +0.008 |
ill | +0.008 |
urious | +0.008 |
Neg logit contributions | |
goodbye | -0.007 |
fs | -0.007 |
Amen | -0.006 |
saddle | -0.006 |
chief | -0.006 |
Token | spent |
Token ID | 3377 |
Feature activation | -9.62e-03 |
Pos logit contributions | |
udge | +0.014 |
eter | +0.013 |
ipples | +0.012 |
ill | +0.012 |
urious | +0.012 |
Neg logit contributions | |
goodbye | -0.013 |
fs | -0.011 |
Amen | -0.011 |
yes | -0.010 |
pron | -0.010 |
Token | the |
Token ID | 262 |
Feature activation | +9.62e-03 |
Pos logit contributions | |
goodbye | +0.016 |
Amen | +0.013 |
fs | +0.013 |
yes | +0.012 |
forehead | +0.012 |
Neg logit contributions | |
udge | -0.022 |
eter | -0.020 |
ipples | -0.020 |
ill | -0.019 |
urious | -0.019 |
Token | day |
Token ID | 1110 |
Feature activation | +5.77e-03 |
Pos logit contributions | |
udge | +0.020 |
ipples | +0.018 |
eter | +0.017 |
urious | +0.017 |
ill | +0.017 |
Neg logit contributions | |
goodbye | -0.017 |
Amen | -0.014 |
fs | -0.014 |
pron | -0.013 |
saddle | -0.013 |
Token | It |
Token ID | 1026 |
Feature activation | +1.03e-02 |
Pos logit contributions | |
goodbye | +0.016 |
fs | +0.014 |
Amen | +0.013 |
yes | +0.013 |
pron | +0.013 |
Neg logit contributions | |
udge | -0.016 |
eter | -0.014 |
ipples | -0.014 |
ill | -0.013 |
ased | -0.013 |
Token | was |
Token ID | 373 |
Feature activation | -4.26e-03 |
Pos logit contributions | |
udge | +0.022 |
eter | +0.019 |
ipples | +0.019 |
elly | +0.018 |
ill | +0.018 |
Neg logit contributions | |
goodbye | -0.020 |
fs | -0.015 |
yes | -0.015 |
Amen | -0.015 |
saddle | -0.014 |
Token | great |
Token ID | 1049 |
Feature activation | -1.48e-02 |
Pos logit contributions | |
goodbye | +0.007 |
fs | +0.006 |
Amen | +0.006 |
yes | +0.005 |
saddle | +0.005 |
Neg logit contributions | |
udge | -0.010 |
eter | -0.009 |
ipples | -0.009 |
ill | -0.008 |
urious | -0.008 |
Token | ! |
Token ID | 0 |
Feature activation | +6.18e-03 |
Pos logit contributions | |
goodbye | +0.027 |
Amen | +0.022 |
fs | +0.022 |
saddle | +0.021 |
forehead | +0.021 |
Neg logit contributions | |
udge | -0.032 |
eter | -0.029 |
ipples | -0.028 |
ill | -0.027 |
urious | -0.026 |
Token | Linda |
Token ID | 21776 |
Feature activation | -8.75e-03 |
Pos logit contributions | |
udge | +0.013 |
ipples | +0.011 |
eter | +0.011 |
urious | +0.011 |
ased | +0.010 |
Neg logit contributions | |
goodbye | -0.012 |
fs | -0.009 |
Amen | -0.009 |
| -0.009 |
yes | -0.009 |
Token | and |
Token ID | 290 |
Feature activation | -7.66e-02 |
Pos logit contributions | |
goodbye | +0.017 |
Amen | +0.015 |
fs | +0.015 |
saddle | +0.013 |
forehead | +0.013 |
Neg logit contributions | |
udge | -0.016 |
eter | -0.015 |
ill | -0.014 |
ipples | -0.014 |
urious | -0.014 |
Token | her |
Token ID | 607 |
Feature activation | +2.10e-03 |
Pos logit contributions | |
goodbye | +0.133 |
Amen | +0.113 |
fs | +0.112 |
saddle | +0.105 |
pron | +0.102 |
Neg logit contributions | |
udge | -0.173 |
eter | -0.149 |
ill | -0.146 |
ipples | -0.145 |
urious | -0.142 |
Token | family |
Token ID | 1641 |
Feature activation | -2.43e-02 |
Pos logit contributions | |
udge | +0.004 |
eter | +0.004 |
urious | +0.004 |
ipples | +0.003 |
ill | +0.003 |
Neg logit contributions | |
goodbye | -0.004 |
Amen | -0.004 |
fs | -0.004 |
saddle | -0.003 |
Midnight | -0.003 |
Token | smiled |
Token ID | 13541 |
Feature activation | +5.68e-03 |
Pos logit contributions | |
goodbye | +0.050 |
fs | +0.046 |
Amen | +0.046 |
Midnight | +0.043 |
pron | +0.042 |
Neg logit contributions | |
udge | -0.041 |
ipples | -0.036 |
eter | -0.036 |
ill | -0.035 |
ased | -0.035 |
Token | , |
Token ID | 11 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
udge | +0.012 |
eter | +0.011 |
ipples | +0.011 |
ill | +0.010 |
urious | +0.010 |
Neg logit contributions | |
goodbye | -0.010 |
fs | -0.009 |
Amen | -0.008 |
pron | -0.008 |
saddle | -0.008 |
Token | and |
Token ID | 290 |
Feature activation | -5.71e-02 |
Pos logit contributions | |
udge | +0.022 |
eter | +0.021 |
ipples | +0.019 |
ill | +0.019 |
ased | +0.018 |
Neg logit contributions | |
goodbye | -0.021 |
fs | -0.020 |
Amen | -0.019 |
pron | -0.017 |
medal | -0.017 |
Token | like |
Token ID | 588 |
Feature activation | +4.39e-03 |
Pos logit contributions | |
udge | +0.010 |
eter | +0.009 |
ipples | +0.009 |
urious | +0.008 |
ill | +0.008 |
Neg logit contributions | |
goodbye | -0.012 |
fs | -0.009 |
yes | -0.009 |
Amen | -0.009 |
forehead | -0.009 |
Token | fun |
Token ID | 1257 |
Feature activation | -2.31e-02 |
Pos logit contributions | |
udge | +0.009 |
eter | +0.008 |
ipples | +0.008 |
ill | +0.007 |
urious | +0.007 |
Neg logit contributions | |
goodbye | -0.008 |
Amen | -0.007 |
fs | -0.007 |
saddle | -0.006 |
forehead | -0.006 |
Token | ." |
Token ID | 526 |
Feature activation | -2.53e-02 |
Pos logit contributions | |
goodbye | +0.040 |
fs | +0.031 |
Amen | +0.030 |
yes | +0.029 |
forehead | +0.028 |
Neg logit contributions | |
udge | -0.054 |
eter | -0.048 |
ipples | -0.047 |
urious | -0.045 |
ill | -0.045 |
Token | And |
Token ID | 843 |
Feature activation | -2.74e-02 |
Pos logit contributions | |
goodbye | +0.044 |
fs | +0.039 |
medal | +0.039 |
Amen | +0.039 |
saddle | +0.038 |
Neg logit contributions | |
udge | -0.052 |
eter | -0.048 |
ill | -0.046 |
ipples | -0.045 |
elly | -0.044 |
Token | Lily |
Token ID | 20037 |
Feature activation | -5.50e-03 |
Pos logit contributions | |
goodbye | +0.043 |
Amen | +0.038 |
fs | +0.036 |
saddle | +0.035 |
pron | +0.035 |
Neg logit contributions | |
udge | -0.063 |
eter | -0.057 |
ill | -0.055 |
ipples | -0.054 |
ased | -0.052 |
Token | and |
Token ID | 290 |
Feature activation | -7.64e-02 |
Pos logit contributions | |
goodbye | +0.010 |
Amen | +0.009 |
fs | +0.009 |
saddle | +0.008 |
yes | +0.008 |
Neg logit contributions | |
udge | -0.011 |
eter | -0.010 |
ipples | -0.010 |
ill | -0.010 |
ased | -0.009 |
Token | Tim |
Token ID | 5045 |
Feature activation | -2.70e-06 |
Pos logit contributions | |
goodbye | +0.138 |
fs | +0.121 |
Amen | +0.118 |
saddle | +0.111 |
pron | +0.111 |
Neg logit contributions | |
udge | -0.161 |
eter | -0.143 |
ipples | -0.137 |
ill | -0.135 |
urious | -0.132 |
Token | my |
Token ID | 1820 |
Feature activation | -7.35e-03 |
Token | smiled |
Token ID | 13541 |
Feature activation | +5.94e-03 |
Pos logit contributions | |
goodbye | +0.014 |
fs | +0.011 |
Amen | +0.011 |
yes | +0.011 |
saddle | +0.011 |
Neg logit contributions | |
udge | -0.016 |
eter | -0.014 |
ipples | -0.013 |
ill | -0.013 |
urious | -0.013 |
Token | , |
Token ID | 11 |
Feature activation | +1.05e-02 |
Pos logit contributions | |
udge | +0.013 |
eter | +0.011 |
ipples | +0.011 |
ill | +0.011 |
urious | +0.010 |
Neg logit contributions | |
goodbye | -0.011 |
fs | -0.009 |
Amen | -0.009 |
pron | -0.009 |
saddle | -0.008 |
Token | happy |
Token ID | 3772 |
Feature activation | -1.04e-02 |
Pos logit contributions | |
udge | +0.019 |
eter | +0.018 |
ipples | +0.018 |
ill | +0.017 |
asured | +0.016 |
Neg logit contributions | |
goodbye | -0.019 |
fs | -0.018 |
Amen | -0.018 |
medal | -0.016 |
clipboard | -0.016 |
Token | embarrassed |
Token ID | 21100 |
Feature activation | +1.40e-02 |
Pos logit contributions | |
goodbye | +0.015 |
fs | +0.013 |
Amen | +0.012 |
saddle | +0.012 |
pron | +0.012 |
Neg logit contributions | |
udge | -0.021 |
ipples | -0.018 |
eter | -0.018 |
elly | -0.017 |
urious | -0.017 |
Token | , |
Token ID | 11 |
Feature activation | +1.71e-02 |
Pos logit contributions | |
udge | +0.031 |
eter | +0.028 |
ipples | +0.028 |
urious | +0.026 |
ased | +0.026 |
Neg logit contributions | |
goodbye | -0.026 |
fs | -0.020 |
yes | -0.020 |
Amen | -0.019 |
forehead | -0.019 |
Token | but |
Token ID | 475 |
Feature activation | -1.75e-02 |
Pos logit contributions | |
udge | +0.025 |
eter | +0.021 |
ill | +0.021 |
ipples | +0.020 |
elly | +0.019 |
Neg logit contributions | |
goodbye | -0.038 |
Amen | -0.037 |
fs | -0.035 |
saddle | -0.034 |
pron | -0.033 |
Token | her |
Token ID | 607 |
Feature activation | -3.67e-03 |
Pos logit contributions | |
goodbye | +0.024 |
fs | +0.020 |
Amen | +0.020 |
pron | +0.019 |
saddle | +0.019 |
Neg logit contributions | |
udge | -0.044 |
eter | -0.039 |
ipples | -0.039 |
urious | -0.039 |
ill | -0.038 |
Token | mom |
Token ID | 1995 |
Feature activation | -4.78e-03 |
Pos logit contributions | |
goodbye | +0.006 |
Amen | +0.006 |
fs | +0.006 |
ner | +0.005 |
yes | +0.005 |
Neg logit contributions | |
udge | -0.008 |
eter | -0.007 |
urious | -0.007 |
ipples | -0.007 |
elly | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | -7.64e-02 |
Pos logit contributions | |
goodbye | +0.010 |
Amen | +0.008 |
fs | +0.008 |
medal | +0.008 |
saddle | +0.007 |
Neg logit contributions | |
udge | -0.009 |
ill | -0.008 |
eter | -0.008 |
ipples | -0.008 |
ased | -0.008 |
Token | dad |
Token ID | 9955 |
Feature activation | -2.38e-02 |
Pos logit contributions | |
goodbye | +0.156 |
fs | +0.147 |
Amen | +0.139 |
saddle | +0.131 |
medal | +0.127 |
Neg logit contributions | |
udge | -0.130 |
urious | -0.112 |
ased | -0.111 |
eter | -0.110 |
ill | -0.107 |
Token | just |
Token ID | 655 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
goodbye | +0.046 |
Amen | +0.041 |
fs | +0.040 |
saddle | +0.038 |
forehead | +0.036 |
Neg logit contributions | |
udge | -0.046 |
ill | -0.041 |
ipples | -0.039 |
eter | -0.039 |
ased | -0.038 |
Token | laughed |
Token ID | 13818 |
Feature activation | -1.31e-02 |
Pos logit contributions | |
udge | +0.042 |
eter | +0.037 |
ipples | +0.037 |
ierce | +0.035 |
elly | +0.035 |
Neg logit contributions | |
goodbye | -0.025 |
yes | -0.018 |
fs | -0.017 |
Amen | -0.017 |
forehead | -0.017 |
Token | and |
Token ID | 290 |
Feature activation | -4.02e-02 |
Pos logit contributions | |
goodbye | +0.027 |
fs | +0.024 |
Amen | +0.024 |
saddle | +0.024 |
pron | +0.023 |
Neg logit contributions | |
udge | -0.023 |
eter | -0.020 |
ipples | -0.019 |
ill | -0.019 |
elly | -0.019 |
Token | said |
Token ID | 531 |
Feature activation | -2.67e-02 |
Pos logit contributions | |
goodbye | +0.068 |
fs | +0.060 |
Amen | +0.060 |
saddle | +0.055 |
pron | +0.054 |
Neg logit contributions | |
udge | -0.086 |
eter | -0.078 |
ipples | -0.077 |
ill | -0.074 |
ased | -0.073 |
Token | . |
Token ID | 13 |
Feature activation | +2.87e-02 |
Pos logit contributions | |
goodbye | +0.018 |
fs | +0.016 |
Amen | +0.015 |
pron | +0.014 |
saddle | +0.014 |
Neg logit contributions | |
udge | -0.021 |
eter | -0.019 |
ill | -0.018 |
ipples | -0.018 |
ased | -0.017 |
Token | Good |
Token ID | 4599 |
Feature activation | +3.94e-02 |
Pos logit contributions | |
udge | +0.061 |
eter | +0.054 |
ipples | +0.053 |
ill | +0.051 |
urious | +0.050 |
Neg logit contributions | |
goodbye | -0.055 |
Amen | -0.043 |
fs | -0.043 |
yes | -0.042 |
saddle | -0.041 |
Token | night |
Token ID | 1755 |
Feature activation | -6.20e-03 |
Pos logit contributions | |
udge | +0.104 |
eter | +0.093 |
ipples | +0.092 |
ill | +0.090 |
ierce | +0.088 |
Neg logit contributions | |
goodbye | -0.059 |
Amen | -0.043 |
forehead | -0.039 |
fs | -0.038 |
yes | -0.038 |
Token | , |
Token ID | 11 |
Feature activation | +1.40e-02 |
Pos logit contributions | |
goodbye | +0.009 |
fs | +0.007 |
Amen | +0.007 |
saddle | +0.007 |
yes | +0.007 |
Neg logit contributions | |
udge | -0.015 |
eter | -0.014 |
ipples | -0.014 |
ill | -0.013 |
elly | -0.013 |
Token | mom |
Token ID | 1995 |
Feature activation | +1.19e-02 |
Pos logit contributions | |
udge | +0.027 |
eter | +0.024 |
ipples | +0.023 |
urious | +0.022 |
ased | +0.022 |
Neg logit contributions | |
goodbye | -0.031 |
yes | -0.025 |
Amen | -0.024 |
fs | -0.024 |
medal | -0.023 |
Token | and |
Token ID | 290 |
Feature activation | -7.60e-02 |
Pos logit contributions | |
udge | +0.025 |
eter | +0.022 |
ipples | +0.021 |
ill | +0.021 |
ased | +0.020 |
Neg logit contributions | |
goodbye | -0.023 |
fs | -0.019 |
Amen | -0.018 |
yes | -0.018 |
saddle | -0.018 |
Token | dad |
Token ID | 9955 |
Feature activation | -1.59e-02 |
Pos logit contributions | |
fs | +0.150 |
goodbye | +0.150 |
Amen | +0.133 |
pron | +0.129 |
forehead | +0.127 |
Neg logit contributions | |
udge | -0.130 |
eter | -0.113 |
ased | -0.106 |
ipples | -0.106 |
ierce | -0.104 |
Token | . |
Token ID | 13 |
Feature activation | +3.13e-02 |
Pos logit contributions | |
goodbye | +0.029 |
fs | +0.026 |
Amen | +0.024 |
saddle | +0.023 |
pron | +0.023 |
Neg logit contributions | |
udge | -0.032 |
eter | -0.028 |
ipples | -0.028 |
ill | -0.028 |
ased | -0.027 |
Token | Good |
Token ID | 4599 |
Feature activation | +3.79e-02 |
Pos logit contributions | |
udge | +0.064 |
eter | +0.055 |
ipples | +0.055 |
urious | +0.052 |
ill | +0.052 |
Neg logit contributions | |
goodbye | -0.064 |
Amen | -0.052 |
fs | -0.050 |
yes | -0.049 |
forehead | -0.048 |
Token | night |
Token ID | 1755 |
Feature activation | -5.34e-03 |
Pos logit contributions | |
udge | +0.099 |
eter | +0.088 |
ipples | +0.087 |
ill | +0.085 |
ierce | +0.083 |
Neg logit contributions | |
goodbye | -0.058 |
Amen | -0.042 |
fs | -0.039 |
forehead | -0.039 |
yes | -0.038 |
Token | , |
Token ID | 11 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
goodbye | +0.008 |
fs | +0.007 |
Amen | +0.006 |
saddle | +0.006 |
yes | +0.006 |
Neg logit contributions | |
udge | -0.013 |
eter | -0.012 |
ipples | -0.012 |
ill | -0.011 |
ased | -0.011 |
Token | playing |
Token ID | 2712 |
Feature activation | -2.65e-02 |
Pos logit contributions | |
udge | +0.036 |
eter | +0.032 |
ipples | +0.032 |
ill | +0.031 |
urious | +0.031 |
Neg logit contributions | |
goodbye | -0.023 |
fs | -0.016 |
Midnight | -0.016 |
Amen | -0.016 |
microphone | -0.016 |
Token | with |
Token ID | 351 |
Feature activation | +1.15e-03 |
Pos logit contributions | |
goodbye | +0.048 |
fs | +0.038 |
Amen | +0.037 |
yes | +0.036 |
magazine | +0.035 |
Neg logit contributions | |
udge | -0.059 |
eter | -0.053 |
ipples | -0.053 |
urious | -0.050 |
ill | -0.050 |
Token | that |
Token ID | 326 |
Feature activation | -6.65e-03 |
Pos logit contributions | |
udge | +0.003 |
eter | +0.002 |
ipples | +0.002 |
ill | +0.002 |
urious | +0.002 |
Neg logit contributions | |
goodbye | -0.002 |
fs | -0.001 |
Amen | -0.001 |
yes | -0.001 |
forehead | -0.001 |
Token | ." |
Token ID | 526 |
Feature activation | -1.80e-02 |
Pos logit contributions | |
goodbye | +0.013 |
fs | +0.010 |
magazine | +0.010 |
yes | +0.010 |
helicopter | +0.010 |
Neg logit contributions | |
udge | -0.014 |
eter | -0.012 |
ipples | -0.012 |
urious | -0.012 |
ased | -0.011 |
Token | Mia |
Token ID | 32189 |
Feature activation | +2.38e-03 |
Pos logit contributions | |
goodbye | +0.040 |
| +0.034 |
fs | +0.031 |
yes | +0.031 |
Midnight | +0.030 |
Neg logit contributions | |
udge | -0.033 |
ipples | -0.028 |
eter | -0.028 |
rying | -0.027 |
ased | -0.027 |
Token | and |
Token ID | 290 |
Feature activation | -7.60e-02 |
Pos logit contributions | |
udge | +0.005 |
eter | +0.004 |
ipples | +0.004 |
ill | +0.004 |
urious | +0.004 |
Neg logit contributions | |
goodbye | -0.005 |
fs | -0.004 |
Amen | -0.004 |
saddle | -0.004 |
pron | -0.004 |
Token | Tom |
Token ID | 4186 |
Feature activation | -1.11e-02 |
Pos logit contributions | |
goodbye | +0.180 |
Amen | +0.164 |
fs | +0.159 |
pron | +0.154 |
saddle | +0.151 |
Neg logit contributions | |
udge | -0.119 |
eter | -0.100 |
ipples | -0.096 |
ill | -0.095 |
urious | -0.093 |
Token | looked |
Token ID | 3114 |
Feature activation | +1.58e-02 |
Pos logit contributions | |
goodbye | +0.022 |
Amen | +0.018 |
fs | +0.018 |
saddle | +0.017 |
pron | +0.017 |
Neg logit contributions | |
udge | -0.022 |
eter | -0.020 |
ipples | -0.020 |
ill | -0.019 |
urious | -0.019 |
Token | at |
Token ID | 379 |
Feature activation | +2.82e-02 |
Pos logit contributions | |
udge | +0.030 |
eter | +0.026 |
ipples | +0.025 |
urious | +0.024 |
elly | +0.024 |
Neg logit contributions | |
goodbye | -0.033 |
fs | -0.028 |
Amen | -0.028 |
yes | -0.026 |
saddle | -0.026 |
Token | each |
Token ID | 1123 |
Feature activation | -2.14e-02 |
Pos logit contributions | |
udge | +0.057 |
eter | +0.051 |
ipples | +0.050 |
elly | +0.047 |
urious | +0.047 |
Neg logit contributions | |
goodbye | -0.058 |
fs | -0.045 |
yes | -0.044 |
Amen | -0.044 |
pony | -0.042 |
Token | other |
Token ID | 584 |
Feature activation | +4.21e-03 |
Pos logit contributions | |
goodbye | +0.032 |
Amen | +0.024 |
fs | +0.023 |
yes | +0.022 |
saddle | +0.022 |
Neg logit contributions | |
udge | -0.055 |
eter | -0.049 |
ipples | -0.049 |
ill | -0.048 |
urious | -0.047 |
Token | was |
Token ID | 373 |
Feature activation | -1.53e-02 |
Pos logit contributions | |
udge | +0.007 |
ipples | +0.006 |
eter | +0.006 |
ill | +0.006 |
ased | +0.006 |
Neg logit contributions | |
goodbye | -0.007 |
fs | -0.006 |
Amen | -0.006 |
saddle | -0.005 |
forehead | -0.005 |
Token | so |
Token ID | 523 |
Feature activation | -1.67e-02 |
Pos logit contributions | |
goodbye | +0.023 |
fs | +0.019 |
Amen | +0.018 |
saddle | +0.017 |
forehead | +0.017 |
Neg logit contributions | |
udge | -0.037 |
eter | -0.034 |
ipples | -0.034 |
ill | -0.032 |
ased | -0.032 |
Token | happy |
Token ID | 3772 |
Feature activation | -2.58e-03 |
Pos logit contributions | |
goodbye | +0.025 |
Amen | +0.018 |
fs | +0.018 |
yes | +0.017 |
forehead | +0.016 |
Neg logit contributions | |
udge | -0.043 |
eter | -0.039 |
ipples | -0.038 |
urious | -0.037 |
ill | -0.037 |
Token | that |
Token ID | 326 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
goodbye | +0.004 |
fs | +0.003 |
Amen | +0.003 |
yes | +0.003 |
forehead | +0.003 |
Neg logit contributions | |
udge | -0.006 |
eter | -0.006 |
ipples | -0.006 |
urious | -0.005 |
ill | -0.005 |
Token | Sarah |
Token ID | 10490 |
Feature activation | +2.44e-03 |
Pos logit contributions | |
goodbye | +0.024 |
Amen | +0.019 |
fs | +0.018 |
saddle | +0.018 |
forehead | +0.017 |
Neg logit contributions | |
udge | -0.038 |
eter | -0.035 |
ipples | -0.034 |
ill | -0.033 |
urious | -0.032 |
Token | and |
Token ID | 290 |
Feature activation | -7.59e-02 |
Pos logit contributions | |
udge | +0.005 |
eter | +0.004 |
ipples | +0.004 |
ill | +0.004 |
ased | +0.004 |
Neg logit contributions | |
goodbye | -0.005 |
Amen | -0.004 |
fs | -0.004 |
yes | -0.004 |
saddle | -0.004 |
Token | her |
Token ID | 607 |
Feature activation | -6.00e-05 |
Pos logit contributions | |
goodbye | +0.128 |
fs | +0.117 |
Amen | +0.110 |
saddle | +0.105 |
pron | +0.104 |
Neg logit contributions | |
udge | -0.165 |
eter | -0.147 |
ipples | -0.141 |
ill | -0.140 |
ased | -0.137 |
Token | dad |
Token ID | 9955 |
Feature activation | -1.55e-02 |
Pos logit contributions | |
goodbye | +0.000 |
Amen | +0.000 |
fs | +0.000 |
saddle | +0.000 |
Midnight | +0.000 |
Neg logit contributions | |
udge | -0.000 |
eter | -0.000 |
ipples | -0.000 |
ill | -0.000 |
urious | -0.000 |
Token | had |
Token ID | 550 |
Feature activation | -1.96e-02 |
Pos logit contributions | |
goodbye | +0.032 |
fs | +0.028 |
Amen | +0.028 |
pron | +0.026 |
saddle | +0.026 |
Neg logit contributions | |
udge | -0.028 |
eter | -0.025 |
ipples | -0.025 |
ill | -0.024 |
ased | -0.023 |
Token | saved |
Token ID | 7448 |
Feature activation | +1.22e-02 |
Pos logit contributions | |
goodbye | +0.037 |
Amen | +0.031 |
fs | +0.031 |
forehead | +0.028 |
saddle | +0.028 |
Neg logit contributions | |
udge | -0.041 |
eter | -0.037 |
ipples | -0.036 |
ill | -0.035 |
ased | -0.034 |
Token | the |
Token ID | 262 |
Feature activation | -1.34e-02 |
Pos logit contributions | |
udge | +0.032 |
ipples | +0.030 |
elly | +0.029 |
urious | +0.029 |
eter | +0.028 |
Neg logit contributions | |
goodbye | -0.016 |
magazine | -0.011 |
fs | -0.011 |
yes | -0.011 |
| -0.010 |
Token | rock |
Token ID | 3881 |
Feature activation | +8.14e-03 |
Pos logit contributions | |
udge | +0.005 |
ill | +0.004 |
ipples | +0.004 |
eter | +0.004 |
elly | +0.004 |
Neg logit contributions | |
goodbye | -0.005 |
Amen | -0.004 |
fs | -0.004 |
yes | -0.004 |
ishly | -0.003 |
Token | and |
Token ID | 290 |
Feature activation | -1.57e-02 |
Pos logit contributions | |
udge | +0.015 |
ill | +0.013 |
ipples | +0.013 |
eter | +0.013 |
elly | +0.012 |
Neg logit contributions | |
goodbye | -0.017 |
Amen | -0.015 |
fs | -0.015 |
pron | -0.014 |
Midnight | -0.014 |
Token | his |
Token ID | 465 |
Feature activation | +1.29e-02 |
Pos logit contributions | |
goodbye | +0.026 |
Amen | +0.021 |
fs | +0.021 |
yes | +0.020 |
pron | +0.019 |
Neg logit contributions | |
udge | -0.036 |
eter | -0.032 |
ipples | -0.032 |
ill | -0.031 |
ased | -0.030 |
Token | axe |
Token ID | 23272 |
Feature activation | -9.48e-03 |
Pos logit contributions | |
udge | +0.027 |
eter | +0.024 |
ipples | +0.024 |
urious | +0.022 |
ased | +0.022 |
Neg logit contributions | |
goodbye | -0.020 |
Amen | -0.020 |
fs | -0.019 |
Midnight | -0.016 |
ishly | -0.016 |
Token | flew |
Token ID | 13112 |
Feature activation | +1.74e-02 |
Pos logit contributions | |
goodbye | +0.019 |
Amen | +0.017 |
fs | +0.015 |
pron | +0.015 |
magazine | +0.015 |
Neg logit contributions | |
udge | -0.018 |
ipples | -0.016 |
eter | -0.016 |
ased | -0.015 |
ill | -0.015 |
Token | out |
Token ID | 503 |
Feature activation | +7.04e-02 |
Pos logit contributions | |
udge | +0.035 |
eter | +0.032 |
ipples | +0.031 |
ill | +0.031 |
ased | +0.030 |
Neg logit contributions | |
goodbye | -0.031 |
Amen | -0.028 |
fs | -0.028 |
pron | -0.026 |
clipboard | -0.025 |
Token | of |
Token ID | 286 |
Feature activation | +3.23e-02 |
Pos logit contributions | |
udge | +0.148 |
eter | +0.130 |
ipples | +0.129 |
ill | +0.125 |
ased | +0.121 |
Neg logit contributions | |
goodbye | -0.134 |
fs | -0.111 |
Amen | -0.110 |
pron | -0.104 |
yes | -0.103 |
Token | his |
Token ID | 465 |
Feature activation | +4.09e-02 |
Pos logit contributions | |
udge | +0.074 |
eter | +0.065 |
ipples | +0.065 |
ill | +0.063 |
ased | +0.062 |
Neg logit contributions | |
goodbye | -0.056 |
fs | -0.044 |
Amen | -0.043 |
yes | -0.042 |
forehead | -0.040 |
Token | hand |
Token ID | 1021 |
Feature activation | -5.83e-03 |
Pos logit contributions | |
udge | +0.086 |
eter | +0.078 |
ipples | +0.075 |
urious | +0.074 |
ased | +0.071 |
Neg logit contributions | |
goodbye | -0.071 |
fs | -0.065 |
Amen | -0.063 |
yes | -0.060 |
Midnight | -0.055 |
Token | . |
Token ID | 13 |
Feature activation | -3.33e-03 |
Pos logit contributions | |
goodbye | +0.012 |
Amen | +0.012 |
fs | +0.011 |
pron | +0.011 |
medal | +0.011 |
Neg logit contributions | |
udge | -0.010 |
eter | -0.009 |
ill | -0.008 |
ipples | -0.008 |
ased | -0.008 |
Token | It |
Token ID | 632 |
Feature activation | -7.93e-03 |
Pos logit contributions | |
goodbye | +0.006 |
fs | +0.005 |
Amen | +0.005 |
yes | +0.005 |
saddle | +0.005 |
Neg logit contributions | |
udge | -0.007 |
eter | -0.006 |
ipples | -0.006 |
ill | -0.006 |
urious | -0.006 |
Token | woods |
Token ID | 16479 |
Feature activation | +7.12e-03 |
Pos logit contributions | |
goodbye | +0.011 |
saddle | +0.008 |
Amen | +0.008 |
fs | +0.008 |
yes | +0.008 |
Neg logit contributions | |
udge | -0.014 |
eter | -0.012 |
ipples | -0.012 |
ill | -0.012 |
ased | -0.011 |
Token | he |
Token ID | 339 |
Feature activation | +3.20e-02 |
Pos logit contributions | |
udge | +0.021 |
eter | +0.019 |
ipples | +0.019 |
urious | +0.018 |
elly | +0.018 |
Neg logit contributions | |
goodbye | -0.008 |
yes | -0.005 |
fs | -0.005 |
Amen | -0.005 |
forehead | -0.004 |
Token | stopped |
Token ID | 5025 |
Feature activation | +3.14e-02 |
Pos logit contributions | |
udge | +0.076 |
eter | +0.066 |
elly | +0.064 |
ipples | +0.063 |
roc | +0.059 |
Neg logit contributions | |
goodbye | -0.058 |
yes | -0.048 |
| -0.047 |
replied | -0.040 |
mask | -0.039 |
Token | and |
Token ID | 290 |
Feature activation | -1.90e-02 |
Pos logit contributions | |
udge | +0.089 |
eter | +0.082 |
ipples | +0.079 |
elly | +0.076 |
urious | +0.076 |
Neg logit contributions | |
goodbye | -0.039 |
yes | -0.026 |
| -0.024 |
Amen | -0.023 |
pony | -0.022 |
Token | gazed |
Token ID | 50255 |
Feature activation | +2.42e-02 |
Pos logit contributions | |
goodbye | +0.035 |
yes | +0.027 |
Amen | +0.026 |
fs | +0.025 |
forehead | +0.025 |
Neg logit contributions | |
udge | -0.044 |
eter | -0.039 |
ipples | -0.038 |
elly | -0.037 |
urious | -0.036 |
Token | out |
Token ID | 503 |
Feature activation | +6.06e-02 |
Pos logit contributions | |
udge | +0.052 |
eter | +0.045 |
ipples | +0.044 |
urious | +0.042 |
ill | +0.042 |
Neg logit contributions | |
goodbye | -0.047 |
fs | -0.036 |
Amen | -0.036 |
yes | -0.036 |
saddle | -0.034 |
Token | over |
Token ID | 625 |
Feature activation | +4.11e-02 |
Pos logit contributions | |
udge | +0.144 |
eter | +0.129 |
ipples | +0.127 |
urious | +0.122 |
ill | +0.121 |
Neg logit contributions | |
goodbye | -0.104 |
Amen | -0.077 |
fs | -0.077 |
yes | -0.075 |
saddle | -0.072 |
Token | a |
Token ID | 257 |
Feature activation | -4.65e-03 |
Pos logit contributions | |
udge | +0.112 |
ipples | +0.100 |
eter | +0.100 |
ierce | +0.097 |
elly | +0.095 |
Neg logit contributions | |
goodbye | -0.061 |
yes | -0.039 |
Amen | -0.036 |
| -0.036 |
blanket | -0.035 |
Token | large |
Token ID | 1588 |
Feature activation | +4.94e-03 |
Pos logit contributions | |
goodbye | +0.009 |
fs | +0.007 |
Amen | +0.007 |
yes | +0.007 |
forehead | +0.007 |
Neg logit contributions | |
udge | -0.010 |
eter | -0.008 |
ipples | -0.008 |
urious | -0.008 |
ill | -0.008 |
Token | pool |
Token ID | 5933 |
Feature activation | -2.40e-02 |
Pos logit contributions | |
udge | +0.011 |
eter | +0.009 |
ipples | +0.009 |
ill | +0.009 |
urious | +0.009 |
Neg logit contributions | |
goodbye | -0.009 |
fs | -0.007 |
Amen | -0.007 |
yes | -0.007 |
forehead | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | +1.43e-03 |
Pos logit contributions | |
goodbye | +0.048 |
Amen | +0.041 |
fs | +0.039 |
forehead | +0.038 |
saddle | +0.037 |
Neg logit contributions | |
udge | -0.049 |
ipples | -0.042 |
eter | -0.041 |
ill | -0.041 |
elly | -0.039 |
Token | , |
Token ID | 11 |
Feature activation | +6.71e-03 |
Pos logit contributions | |
goodbye | +0.036 |
Amen | +0.031 |
fs | +0.029 |
forehead | +0.029 |
saddle | +0.028 |
Neg logit contributions | |
udge | -0.052 |
eter | -0.045 |
ipples | -0.044 |
ill | -0.044 |
elly | -0.044 |
Token | something |
Token ID | 1223 |
Feature activation | +1.31e-02 |
Pos logit contributions | |
udge | +0.014 |
eter | +0.012 |
ipples | +0.012 |
ill | +0.011 |
ierce | +0.011 |
Neg logit contributions | |
goodbye | -0.013 |
Amen | -0.011 |
fs | -0.010 |
saddle | -0.010 |
forehead | -0.010 |
Token | slow |
Token ID | 3105 |
Feature activation | -7.60e-03 |
Pos logit contributions | |
udge | +0.029 |
eter | +0.025 |
ipples | +0.025 |
urious | +0.024 |
ill | +0.024 |
Neg logit contributions | |
goodbye | -0.024 |
Amen | -0.019 |
fs | -0.019 |
yes | -0.018 |
saddle | -0.018 |
Token | and |
Token ID | 290 |
Feature activation | -2.62e-02 |
Pos logit contributions | |
goodbye | +0.016 |
Amen | +0.014 |
fs | +0.013 |
pron | +0.013 |
yes | +0.013 |
Neg logit contributions | |
udge | -0.014 |
eter | -0.012 |
ipples | -0.012 |
urious | -0.012 |
ased | -0.012 |
Token | scary |
Token ID | 14343 |
Feature activation | +5.97e-03 |
Pos logit contributions | |
goodbye | +0.051 |
Amen | +0.042 |
fs | +0.042 |
forehead | +0.039 |
saddle | +0.039 |
Neg logit contributions | |
udge | -0.051 |
eter | -0.046 |
ipples | -0.045 |
urious | -0.044 |
ased | -0.043 |
Token | came |
Token ID | 1625 |
Feature activation | +6.31e-02 |
Pos logit contributions | |
udge | +0.013 |
eter | +0.012 |
ipples | +0.011 |
ill | +0.011 |
urious | +0.011 |
Neg logit contributions | |
goodbye | -0.012 |
fs | -0.009 |
yes | -0.009 |
Amen | -0.009 |
microphone | -0.008 |
Token | to |
Token ID | 284 |
Feature activation | +1.86e-02 |
Pos logit contributions | |
udge | +0.158 |
eter | +0.141 |
ipples | +0.139 |
ased | +0.135 |
urious | +0.132 |
Neg logit contributions | |
goodbye | -0.096 |
fs | -0.066 |
yes | -0.065 |
| -0.064 |
Amen | -0.063 |
Token | the |
Token ID | 262 |
Feature activation | -1.15e-02 |
Pos logit contributions | |
udge | +0.048 |
eter | +0.043 |
elly | +0.043 |
ipples | +0.042 |
ased | +0.041 |
Neg logit contributions | |
goodbye | -0.029 |
Amen | -0.021 |
yes | -0.021 |
helicopter | -0.020 |
blanket | -0.020 |
Token | house |
Token ID | 2156 |
Feature activation | +1.24e-02 |
Pos logit contributions | |
goodbye | +0.020 |
Amen | +0.015 |
fs | +0.015 |
yes | +0.014 |
saddle | +0.014 |
Neg logit contributions | |
udge | -0.027 |
eter | -0.024 |
ipples | -0.023 |
ill | -0.023 |
ased | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | -4.46e-03 |
Pos logit contributions | |
udge | +0.030 |
eter | +0.026 |
ipples | +0.026 |
urious | +0.025 |
elly | +0.024 |
Neg logit contributions | |
goodbye | -0.021 |
fs | -0.016 |
Amen | -0.015 |
yes | -0.015 |
forehead | -0.015 |
Token | It |
Token ID | 632 |
Feature activation | -4.00e-03 |
Pos logit contributions | |
goodbye | +0.008 |
fs | +0.007 |
Amen | +0.006 |
yes | +0.006 |
saddle | +0.006 |
Neg logit contributions | |
udge | -0.010 |
eter | -0.009 |
ipples | -0.009 |
ill | -0.008 |
urious | -0.008 |
Token | , |
Token ID | 11 |
Feature activation | -1.95e-02 |
Pos logit contributions | |
goodbye | +0.041 |
fs | +0.034 |
Amen | +0.033 |
pron | +0.031 |
forehead | +0.031 |
Neg logit contributions | |
udge | -0.064 |
eter | -0.059 |
ipples | -0.057 |
ill | -0.056 |
ased | -0.055 |
Token | an |
Token ID | 281 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
goodbye | +0.033 |
fs | +0.028 |
Amen | +0.027 |
pron | +0.026 |
saddle | +0.026 |
Neg logit contributions | |
udge | -0.043 |
eter | -0.039 |
ipples | -0.038 |
ill | -0.037 |
ased | -0.036 |
Token | even |
Token ID | 772 |
Feature activation | +3.10e-02 |
Pos logit contributions | |
udge | +0.043 |
eter | +0.039 |
ipples | +0.038 |
ill | +0.036 |
ased | +0.036 |
Neg logit contributions | |
goodbye | -0.034 |
Amen | -0.027 |
forehead | -0.025 |
fs | -0.025 |
yes | -0.025 |
Token | bigger |
Token ID | 5749 |
Feature activation | +1.46e-02 |
Pos logit contributions | |
udge | +0.069 |
eter | +0.062 |
ipples | +0.061 |
urious | +0.059 |
ased | +0.058 |
Neg logit contributions | |
goodbye | -0.054 |
fs | -0.043 |
Amen | -0.043 |
yes | -0.040 |
saddle | -0.040 |
Token | darkness |
Token ID | 11854 |
Feature activation | +1.18e-02 |
Pos logit contributions | |
udge | +0.033 |
eter | +0.029 |
ipples | +0.029 |
ill | +0.028 |
urious | +0.027 |
Neg logit contributions | |
goodbye | -0.027 |
Amen | -0.021 |
fs | -0.020 |
yes | -0.020 |
saddle | -0.019 |
Token | came |
Token ID | 1625 |
Feature activation | +6.07e-02 |
Pos logit contributions | |
udge | +0.025 |
eter | +0.022 |
ipples | +0.022 |
ill | +0.021 |
urious | +0.020 |
Neg logit contributions | |
goodbye | -0.022 |
Amen | -0.018 |
fs | -0.018 |
yes | -0.017 |
saddle | -0.017 |
Token | and |
Token ID | 290 |
Feature activation | -1.63e-02 |
Pos logit contributions | |
udge | +0.138 |
eter | +0.121 |
ipples | +0.120 |
urious | +0.114 |
ill | +0.113 |
Neg logit contributions | |
goodbye | -0.109 |
fs | -0.083 |
Amen | -0.083 |
yes | -0.081 |
saddle | -0.078 |
Token | covered |
Token ID | 5017 |
Feature activation | +2.33e-02 |
Pos logit contributions | |
goodbye | +0.029 |
fs | +0.023 |
Amen | +0.023 |
yes | +0.022 |
saddle | +0.021 |
Neg logit contributions | |
udge | -0.037 |
eter | -0.033 |
ipples | -0.032 |
ill | -0.031 |
urious | -0.031 |
Token | the |
Token ID | 262 |
Feature activation | -6.89e-03 |
Pos logit contributions | |
udge | +0.053 |
eter | +0.048 |
ipples | +0.047 |
urious | +0.045 |
ill | +0.044 |
Neg logit contributions | |
goodbye | -0.041 |
fs | -0.032 |
Amen | -0.031 |
yes | -0.030 |
forehead | -0.029 |
Token | city |
Token ID | 1748 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
goodbye | +0.013 |
Amen | +0.010 |
fs | +0.010 |
forehead | +0.009 |
yes | +0.009 |
Neg logit contributions | |
udge | -0.015 |
eter | -0.013 |
ipples | -0.013 |
ill | -0.013 |
ased | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | -6.59e-03 |
Pos logit contributions | |
udge | +0.022 |
eter | +0.020 |
ipples | +0.019 |
ill | +0.019 |
urious | +0.018 |
Neg logit contributions | |
goodbye | -0.020 |
fs | -0.016 |
Amen | -0.016 |
yes | -0.016 |
saddle | -0.015 |
Token | led |
Token ID | 992 |
Feature activation | -3.86e-04 |
Pos logit contributions | |
udge | +0.036 |
eter | +0.032 |
ipples | +0.031 |
ill | +0.031 |
ierce | +0.030 |
Neg logit contributions | |
goodbye | -0.023 |
fs | -0.017 |
forehead | -0.016 |
saddle | -0.015 |
yes | -0.015 |
Token | around |
Token ID | 1088 |
Feature activation | +2.33e-03 |
Pos logit contributions | |
goodbye | +0.001 |
fs | +0.001 |
Amen | +0.001 |
yes | +0.001 |
saddle | +0.001 |
Neg logit contributions | |
udge | -0.001 |
eter | -0.001 |
ipples | -0.001 |
urious | -0.001 |
ill | -0.001 |
Token | , |
Token ID | 11 |
Feature activation | +1.34e-02 |
Pos logit contributions | |
udge | +0.006 |
ipples | +0.006 |
eter | +0.006 |
urious | +0.005 |
elly | +0.005 |
Neg logit contributions | |
goodbye | -0.004 |
yes | -0.002 |
Amen | -0.002 |
forehead | -0.002 |
fs | -0.002 |
Token | pe |
Token ID | 613 |
Feature activation | +4.55e-02 |
Pos logit contributions | |
udge | +0.029 |
eter | +0.026 |
ipples | +0.025 |
ill | +0.024 |
ased | +0.024 |
Neg logit contributions | |
goodbye | -0.024 |
fs | -0.020 |
Amen | -0.020 |
yes | -0.018 |
saddle | -0.018 |
Token | eking |
Token ID | 18754 |
Feature activation | +2.76e-02 |
Pos logit contributions | |
udge | +0.093 |
eter | +0.083 |
ipples | +0.079 |
elly | +0.076 |
ill | +0.076 |
Neg logit contributions | |
goodbye | -0.090 |
fs | -0.074 |
Amen | -0.071 |
yes | -0.071 |
saddle | -0.069 |
Token | out |
Token ID | 503 |
Feature activation | +6.72e-02 |
Pos logit contributions | |
udge | +0.064 |
eter | +0.057 |
ipples | +0.056 |
urious | +0.054 |
ill | +0.053 |
Neg logit contributions | |
goodbye | -0.049 |
fs | -0.037 |
yes | -0.036 |
Amen | -0.036 |
saddle | -0.034 |
Token | from |
Token ID | 422 |
Feature activation | +3.29e-02 |
Pos logit contributions | |
udge | +0.153 |
eter | +0.134 |
ipples | +0.134 |
urious | +0.129 |
elly | +0.126 |
Neg logit contributions | |
goodbye | -0.125 |
yes | -0.093 |
fs | -0.090 |
Amen | -0.088 |
forehead | -0.087 |
Token | her |
Token ID | 607 |
Feature activation | +3.44e-02 |
Pos logit contributions | |
udge | +0.072 |
eter | +0.066 |
ipples | +0.065 |
urious | +0.061 |
elly | +0.061 |
Neg logit contributions | |
goodbye | -0.062 |
Amen | -0.046 |
blanket | -0.045 |
veil | -0.045 |
saddle | -0.045 |
Token | cozy |
Token ID | 37438 |
Feature activation | +1.76e-02 |
Pos logit contributions | |
udge | +0.080 |
eter | +0.073 |
ipples | +0.070 |
ill | +0.068 |
urious | +0.068 |
Neg logit contributions | |
goodbye | -0.055 |
Amen | -0.048 |
fs | -0.046 |
yes | -0.042 |
pron | -0.041 |
Token | blanket |
Token ID | 18447 |
Feature activation | -3.02e-02 |
Pos logit contributions | |
udge | +0.044 |
eter | +0.040 |
ipples | +0.039 |
ill | +0.038 |
ased | +0.037 |
Neg logit contributions | |
goodbye | -0.026 |
Amen | -0.021 |
fs | -0.021 |
yes | -0.020 |
pron | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
Amen | +0.069 |
goodbye | +0.065 |
fs | +0.060 |
headset | +0.057 |
pron | +0.057 |
Neg logit contributions | |
udge | -0.047 |
ill | -0.043 |
eter | -0.042 |
ipples | -0.038 |
elly | -0.037 |
Token | same |
Token ID | 976 |
Feature activation | +8.13e-04 |
Pos logit contributions | |
goodbye | +0.024 |
fs | +0.021 |
Amen | +0.019 |
forehead | +0.018 |
pron | +0.018 |
Neg logit contributions | |
udge | -0.026 |
eter | -0.024 |
ipples | -0.023 |
ierce | -0.022 |
ill | -0.022 |
Token | as |
Token ID | 355 |
Feature activation | +3.80e-02 |
Pos logit contributions | |
udge | +0.002 |
eter | +0.002 |
ipples | +0.002 |
urious | +0.002 |
ill | +0.002 |
Neg logit contributions | |
goodbye | -0.001 |
fs | -0.001 |
Amen | -0.001 |
yes | -0.001 |
saddle | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
udge | +0.097 |
eter | +0.087 |
ipples | +0.086 |
urious | +0.083 |
ill | +0.082 |
Neg logit contributions | |
goodbye | -0.057 |
fs | -0.042 |
yes | -0.041 |
Amen | -0.041 |
saddle | -0.038 |
Token | one |
Token ID | 530 |
Feature activation | +1.32e-02 |
Pos logit contributions | |
goodbye | +0.021 |
fs | +0.019 |
Amen | +0.018 |
Midnight | +0.018 |
pron | +0.017 |
Neg logit contributions | |
udge | -0.029 |
ipples | -0.026 |
eter | -0.025 |
ased | -0.025 |
ierce | -0.024 |
Token | on |
Token ID | 319 |
Feature activation | +2.13e-02 |
Pos logit contributions | |
udge | +0.033 |
eter | +0.029 |
ipples | +0.029 |
ill | +0.028 |
urious | +0.027 |
Neg logit contributions | |
goodbye | -0.021 |
Amen | -0.016 |
fs | -0.016 |
forehead | -0.015 |
saddle | -0.014 |
Token | his |
Token ID | 465 |
Feature activation | +2.72e-02 |
Pos logit contributions | |
udge | +0.047 |
eter | +0.041 |
ipples | +0.040 |
ill | +0.039 |
urious | +0.038 |
Neg logit contributions | |
goodbye | -0.039 |
Amen | -0.032 |
fs | -0.031 |
forehead | -0.030 |
yes | -0.030 |
Token | je |
Token ID | 11223 |
Feature activation | +8.84e-03 |
Pos logit contributions | |
udge | +0.060 |
eter | +0.054 |
ipples | +0.051 |
urious | +0.050 |
ill | +0.049 |
Neg logit contributions | |
goodbye | -0.044 |
Amen | -0.041 |
fs | -0.040 |
pron | -0.037 |
Midnight | -0.037 |
Token | ep |
Token ID | 538 |
Feature activation | -7.61e-03 |
Pos logit contributions | |
udge | +0.017 |
eter | +0.015 |
ipples | +0.014 |
urious | +0.014 |
ill | +0.013 |
Neg logit contributions | |
goodbye | -0.019 |
Amen | -0.016 |
fs | -0.015 |
yes | -0.015 |
forehead | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | +1.64e-02 |
Pos logit contributions | |
goodbye | +0.014 |
Amen | +0.011 |
fs | +0.011 |
pron | +0.011 |
yes | +0.011 |
Neg logit contributions | |
udge | -0.016 |
eter | -0.015 |
ipples | -0.014 |
ill | -0.014 |
ased | -0.014 |
Token | They |
Token ID | 1119 |
Feature activation | -2.26e-02 |
Pos logit contributions | |
udge | +0.034 |
eter | +0.029 |
ipples | +0.029 |
urious | +0.028 |
rying | +0.027 |
Neg logit contributions | |
goodbye | -0.033 |
yes | -0.026 |
fs | -0.026 |
Amen | -0.026 |
| -0.025 |
Token | had |
Token ID | 550 |
Feature activation | -1.48e-02 |
Pos logit contributions | |
goodbye | +0.045 |
fs | +0.037 |
Amen | +0.037 |
forehead | +0.035 |
pron | +0.035 |
Neg logit contributions | |
udge | -0.045 |
ipples | -0.039 |
eter | -0.039 |
ill | -0.038 |
ased | -0.037 |
Token | was |
Token ID | 373 |
Feature activation | -8.49e-03 |
Pos logit contributions | |
udge | +0.022 |
eter | +0.019 |
ipples | +0.019 |
elly | +0.018 |
ill | +0.018 |
Neg logit contributions | |
goodbye | -0.017 |
yes | -0.013 |
fs | -0.013 |
saddle | -0.013 |
Amen | -0.012 |
Token | surrounded |
Token ID | 11191 |
Feature activation | +1.96e-02 |
Pos logit contributions | |
goodbye | +0.014 |
fs | +0.010 |
Amen | +0.010 |
yes | +0.010 |
saddle | +0.010 |
Neg logit contributions | |
udge | -0.021 |
eter | -0.018 |
ipples | -0.018 |
urious | -0.017 |
ill | -0.017 |
Token | by |
Token ID | 416 |
Feature activation | +3.04e-02 |
Pos logit contributions | |
udge | +0.037 |
ipples | +0.033 |
eter | +0.032 |
ierce | +0.030 |
urious | +0.030 |
Neg logit contributions | |
goodbye | -0.042 |
yes | -0.032 |
Amen | -0.032 |
fs | -0.032 |
magazine | -0.032 |
Token | a |
Token ID | 257 |
Feature activation | -1.62e-02 |
Pos logit contributions | |
udge | +0.074 |
eter | +0.066 |
ipples | +0.065 |
elly | +0.063 |
ill | +0.062 |
Neg logit contributions | |
goodbye | -0.051 |
yes | -0.037 |
fs | -0.037 |
medal | -0.035 |
blanket | -0.034 |
Token | jungle |
Token ID | 20712 |
Feature activation | -2.67e-03 |
Pos logit contributions | |
goodbye | +0.032 |
fs | +0.026 |
Amen | +0.025 |
yes | +0.024 |
forehead | +0.024 |
Neg logit contributions | |
udge | -0.034 |
eter | -0.030 |
ipples | -0.029 |
urious | -0.028 |
ill | -0.027 |
Token | of |
Token ID | 286 |
Feature activation | +2.43e-02 |
Pos logit contributions | |
goodbye | +0.005 |
fs | +0.004 |
Amen | +0.004 |
yes | +0.004 |
forehead | +0.004 |
Neg logit contributions | |
udge | -0.006 |
eter | -0.005 |
ipples | -0.005 |
ill | -0.005 |
urious | -0.005 |
Token | sweets |
Token ID | 42402 |
Feature activation | -1.05e-02 |
Pos logit contributions | |
udge | +0.055 |
eter | +0.049 |
ipples | +0.048 |
ill | +0.046 |
urious | +0.046 |
Neg logit contributions | |
goodbye | -0.044 |
fs | -0.034 |
Amen | -0.034 |
yes | -0.033 |
saddle | -0.032 |
Token | and |
Token ID | 290 |
Feature activation | -1.32e-02 |
Pos logit contributions | |
goodbye | +0.021 |
Amen | +0.018 |
fs | +0.016 |
forehead | +0.016 |
yes | +0.016 |
Neg logit contributions | |
udge | -0.021 |
ipples | -0.019 |
eter | -0.019 |
ill | -0.018 |
ased | -0.017 |
Token | treats |
Token ID | 18432 |
Feature activation | -5.57e-03 |
Pos logit contributions | |
goodbye | +0.024 |
fs | +0.019 |
Amen | +0.019 |
yes | +0.018 |
saddle | +0.017 |
Neg logit contributions | |
udge | -0.029 |
eter | -0.026 |
ipples | -0.026 |
ill | -0.025 |
urious | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | -8.33e-04 |
Pos logit contributions | |
goodbye | +0.011 |
Amen | +0.010 |
fs | +0.009 |
forehead | +0.009 |
pron | +0.009 |
Neg logit contributions | |
udge | -0.011 |
ipples | -0.010 |
eter | -0.010 |
ill | -0.009 |
ased | -0.009 |
Token | Tom |
Token ID | 4186 |
Feature activation | +2.71e-03 |
Pos logit contributions | |
goodbye | +0.002 |
fs | +0.001 |
yes | +0.001 |
Amen | +0.001 |
| +0.001 |
Neg logit contributions | |
udge | -0.002 |
ipples | -0.002 |
eter | -0.002 |
urious | -0.002 |
ill | -0.002 |
Token | you |
Token ID | 345 |
Feature activation | +1.38e-02 |
Pos logit contributions | |
goodbye | +0.036 |
pron | +0.032 |
fs | +0.031 |
saddle | +0.030 |
forehead | +0.030 |
Neg logit contributions | |
udge | -0.033 |
eter | -0.029 |
ipples | -0.027 |
ased | -0.027 |
urious | -0.027 |
Token | , |
Token ID | 11 |
Feature activation | +8.56e-03 |
Pos logit contributions | |
udge | +0.022 |
eter | +0.021 |
ill | +0.020 |
ipples | +0.019 |
elly | +0.018 |
Neg logit contributions | |
goodbye | -0.027 |
fs | -0.027 |
pron | -0.026 |
saddle | -0.026 |
chief | -0.025 |
Token | Sara |
Token ID | 24799 |
Feature activation | +1.14e-02 |
Pos logit contributions | |
udge | +0.016 |
eter | +0.014 |
ipples | +0.013 |
ill | +0.013 |
ierce | +0.013 |
Neg logit contributions | |
goodbye | -0.017 |
fs | -0.015 |
Amen | -0.015 |
pron | -0.014 |
forehead | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | -5.66e-02 |
Pos logit contributions | |
udge | +0.020 |
eter | +0.017 |
ipples | +0.017 |
ill | +0.017 |
elly | +0.017 |
Neg logit contributions | |
goodbye | -0.025 |
fs | -0.022 |
Amen | -0.021 |
yes | -0.020 |
pron | -0.020 |
Token | Ben |
Token ID | 3932 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
goodbye | +0.130 |
fs | +0.123 |
Amen | +0.117 |
pron | +0.112 |
saddle | +0.110 |
Neg logit contributions | |
udge | -0.085 |
eter | -0.074 |
ipples | -0.071 |
ill | -0.068 |
ased | -0.067 |
Token | , |
Token ID | 11 |
Feature activation | +3.59e-02 |
Pos logit contributions | |
goodbye | +0.031 |
fs | +0.030 |
pron | +0.027 |
chief | +0.027 |
Amen | +0.027 |
Neg logit contributions | |
udge | -0.025 |
ipples | -0.022 |
eter | -0.022 |
ill | -0.021 |
elly | -0.020 |
Token | for |
Token ID | 329 |
Feature activation | -1.91e-02 |
Pos logit contributions | |
udge | +0.066 |
eter | +0.058 |
ill | +0.055 |
ipples | +0.055 |
ierce | +0.054 |
Neg logit contributions | |
goodbye | -0.068 |
fs | -0.063 |
Amen | -0.060 |
pron | -0.058 |
forehead | -0.058 |
Token | finding |
Token ID | 4917 |
Feature activation | -2.97e-03 |
Pos logit contributions | |
goodbye | +0.038 |
fs | +0.034 |
saddle | +0.032 |
yes | +0.031 |
pron | +0.030 |
Neg logit contributions | |
udge | -0.036 |
ipples | -0.030 |
eter | -0.029 |
aught | -0.029 |
ased | -0.028 |
Token | him |
Token ID | 683 |
Feature activation | +7.95e-03 |
Pos logit contributions | |
goodbye | +0.005 |
fs | +0.004 |
Amen | +0.004 |
pron | +0.004 |
forehead | +0.004 |
Neg logit contributions | |
udge | -0.006 |
eter | -0.006 |
ipples | -0.005 |
ill | -0.005 |
ased | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | +2.40e-02 |
Pos logit contributions | |
udge | +0.015 |
eter | +0.014 |
ipples | +0.013 |
ill | +0.013 |
urious | +0.012 |
Neg logit contributions | |
goodbye | -0.015 |
Amen | -0.014 |
fs | -0.014 |
pron | -0.013 |
saddle | -0.013 |
Token | You |
Token ID | 921 |
Feature activation | +8.03e-03 |
Pos logit contributions | |
udge | +0.047 |
eter | +0.042 |
ipples | +0.040 |
ill | +0.039 |
elly | +0.038 |
Neg logit contributions | |
goodbye | -0.049 |
fs | -0.040 |
Amen | -0.038 |
yes | -0.038 |
pron | -0.038 |
Token | meals |
Token ID | 13840 |
Feature activation | -1.52e-02 |
Pos logit contributions | |
udge | +0.016 |
eter | +0.015 |
ipples | +0.013 |
urious | +0.013 |
ill | +0.013 |
Neg logit contributions | |
goodbye | -0.013 |
Amen | -0.011 |
fs | -0.011 |
saddle | -0.010 |
headset | -0.009 |
Token | and |
Token ID | 290 |
Feature activation | -3.04e-02 |
Pos logit contributions | |
goodbye | +0.030 |
Amen | +0.029 |
saddle | +0.025 |
headset | +0.025 |
fs | +0.025 |
Neg logit contributions | |
udge | -0.029 |
eter | -0.026 |
ill | -0.025 |
ipples | -0.025 |
elly | -0.023 |
Token | tried |
Token ID | 3088 |
Feature activation | -1.29e-02 |
Pos logit contributions | |
goodbye | +0.055 |
fs | +0.045 |
Amen | +0.045 |
yes | +0.042 |
saddle | +0.042 |
Neg logit contributions | |
udge | -0.066 |
eter | -0.059 |
ipples | -0.058 |
ill | -0.055 |
urious | -0.055 |
Token | her |
Token ID | 607 |
Feature activation | +8.86e-03 |
Pos logit contributions | |
goodbye | +0.022 |
fs | +0.017 |
Amen | +0.017 |
yes | +0.016 |
saddle | +0.016 |
Neg logit contributions | |
udge | -0.030 |
eter | -0.027 |
ipples | -0.027 |
ill | -0.025 |
urious | -0.025 |
Token | best |
Token ID | 1266 |
Feature activation | -9.68e-03 |
Pos logit contributions | |
udge | +0.018 |
eter | +0.017 |
urious | +0.016 |
ipples | +0.016 |
cy | +0.015 |
Neg logit contributions | |
goodbye | -0.016 |
Amen | -0.014 |
fs | -0.014 |
headset | -0.013 |
pron | -0.013 |
Token | to |
Token ID | 284 |
Feature activation | +1.59e-02 |
Pos logit contributions | |
goodbye | +0.017 |
Amen | +0.014 |
fs | +0.014 |
yes | +0.013 |
forehead | +0.013 |
Neg logit contributions | |
udge | -0.022 |
eter | -0.019 |
ipples | -0.019 |
ill | -0.018 |
urious | -0.018 |
Token | not |
Token ID | 407 |
Feature activation | +3.00e-02 |
Pos logit contributions | |
udge | +0.035 |
eter | +0.030 |
ipples | +0.029 |
elly | +0.029 |
urious | +0.028 |
Neg logit contributions | |
goodbye | -0.030 |
Amen | -0.024 |
yes | -0.024 |
fs | -0.023 |
forehead | -0.022 |
Token | waste |
Token ID | 7030 |
Feature activation | +1.02e-02 |
Pos logit contributions | |
udge | +0.077 |
ipples | +0.066 |
eter | +0.066 |
urious | +0.066 |
ased | +0.065 |
Neg logit contributions | |
goodbye | -0.046 |
yes | -0.031 |
medal | -0.031 |
Midnight | -0.030 |
pony | -0.030 |
Token | any |
Token ID | 597 |
Feature activation | +2.78e-02 |
Pos logit contributions | |
udge | +0.023 |
eter | +0.021 |
ipples | +0.021 |
urious | +0.020 |
ill | +0.020 |
Neg logit contributions | |
goodbye | -0.018 |
fs | -0.014 |
Amen | -0.014 |
yes | -0.013 |
saddle | -0.013 |
Token | food |
Token ID | 2057 |
Feature activation | +3.38e-03 |
Pos logit contributions | |
udge | +0.064 |
eter | +0.055 |
ipples | +0.055 |
ill | +0.054 |
ierce | +0.053 |
Neg logit contributions | |
goodbye | -0.049 |
fs | -0.037 |
yes | -0.036 |
medal | -0.035 |
magazine | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | -1.45e-02 |
Pos logit contributions | |
udge | +0.007 |
eter | +0.006 |
ipples | +0.006 |
ill | +0.006 |
elly | +0.006 |
Neg logit contributions | |
goodbye | -0.006 |
Amen | -0.005 |
fs | -0.005 |
yes | -0.005 |
forehead | -0.005 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.12e-02 |
Pos logit contributions | |
udge | +0.003 |
eter | +0.002 |
ipples | +0.002 |
ill | +0.002 |
urious | +0.002 |
Neg logit contributions | |
goodbye | -0.006 |
Amen | -0.005 |
fs | -0.005 |
saddle | -0.005 |
forehead | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | -1.38e-02 |
Pos logit contributions | |
udge | +0.046 |
ipples | +0.040 |
eter | +0.039 |
ill | +0.039 |
urious | +0.038 |
Neg logit contributions | |
goodbye | -0.038 |
fs | -0.031 |
Amen | -0.030 |
saddle | -0.029 |
ishly | -0.029 |
Token | time |
Token ID | 640 |
Feature activation | -1.87e-02 |
Pos logit contributions | |
goodbye | +0.023 |
yes | +0.018 |
Amen | +0.018 |
fs | +0.017 |
pron | +0.017 |
Neg logit contributions | |
udge | -0.033 |
eter | -0.030 |
ipples | -0.029 |
ased | -0.028 |
ill | -0.027 |
Token | , |
Token ID | 11 |
Feature activation | +1.73e-02 |
Pos logit contributions | |
goodbye | +0.020 |
yes | +0.013 |
fs | +0.012 |
Amen | +0.011 |
| +0.011 |
Neg logit contributions | |
udge | -0.056 |
eter | -0.052 |
ipples | -0.052 |
urious | -0.050 |
ased | -0.049 |
Token | there |
Token ID | 612 |
Feature activation | -7.95e-03 |
Pos logit contributions | |
udge | +0.043 |
eter | +0.040 |
ipples | +0.038 |
urious | +0.036 |
ased | +0.036 |
Neg logit contributions | |
goodbye | -0.027 |
yes | -0.021 |
Amen | -0.021 |
fs | -0.020 |
| -0.019 |
Token | was |
Token ID | 373 |
Feature activation | +2.93e-02 |
Pos logit contributions | |
goodbye | +0.010 |
| +0.009 |
on | +0.008 |
yes | +0.007 |
blanket | +0.007 |
Neg logit contributions | |
udge | -0.021 |
eter | -0.020 |
ipples | -0.020 |
rying | -0.020 |
ased | -0.019 |
Token | a |
Token ID | 257 |
Feature activation | -3.15e-03 |
Pos logit contributions | |
udge | +0.066 |
eter | +0.059 |
ipples | +0.058 |
ill | +0.056 |
urious | +0.055 |
Neg logit contributions | |
goodbye | -0.051 |
fs | -0.041 |
Amen | -0.040 |
forehead | -0.038 |
saddle | -0.038 |
Token | small |
Token ID | 1402 |
Feature activation | +1.79e-02 |
Pos logit contributions | |
goodbye | +0.006 |
fs | +0.005 |
Amen | +0.005 |
yes | +0.005 |
saddle | +0.004 |
Neg logit contributions | |
udge | -0.007 |
eter | -0.006 |
ipples | -0.006 |
urious | -0.005 |
ill | -0.005 |
Token | girl |
Token ID | 2576 |
Feature activation | -6.59e-03 |
Pos logit contributions | |
udge | +0.036 |
eter | +0.032 |
ipples | +0.031 |
ill | +0.029 |
elly | +0.029 |
Neg logit contributions | |
goodbye | -0.036 |
fs | -0.030 |
Amen | -0.029 |
yes | -0.028 |
forehead | -0.027 |
Token | who |
Token ID | 508 |
Feature activation | +2.58e-03 |
Pos logit contributions | |
goodbye | +0.010 |
Amen | +0.008 |
fs | +0.008 |
yes | +0.007 |
saddle | +0.007 |
Neg logit contributions | |
udge | -0.016 |
eter | -0.015 |
ipples | -0.015 |
ill | -0.014 |
urious | -0.014 |
Token | wanted |
Token ID | 2227 |
Feature activation | -2.48e-03 |
Pos logit contributions | |
udge | +0.006 |
eter | +0.005 |
ipples | +0.005 |
ill | +0.005 |
ased | +0.005 |
Neg logit contributions | |
goodbye | -0.005 |
Amen | -0.004 |
fs | -0.004 |
yes | -0.004 |
pron | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | -1.01e-02 |
Pos logit contributions | |
goodbye | +0.001 |
fs | +0.001 |
Amen | +0.001 |
forehead | +0.001 |
pron | +0.001 |
Neg logit contributions | |
udge | -0.001 |
eter | -0.001 |
ipples | -0.001 |
ill | -0.001 |
ased | -0.001 |
Token | beautiful |
Token ID | 4950 |
Feature activation | +4.62e-04 |
Pos logit contributions | |
goodbye | +0.019 |
fs | +0.017 |
Amen | +0.016 |
ishly | +0.015 |
Midnight | +0.015 |
Neg logit contributions | |
udge | -0.020 |
eter | -0.018 |
ipples | -0.017 |
urious | -0.016 |
ill | -0.016 |
Token | golden |
Token ID | 10861 |
Feature activation | +4.35e-03 |
Pos logit contributions | |
udge | +0.001 |
eter | +0.001 |
ipples | +0.001 |
ill | +0.001 |
cy | +0.001 |
Neg logit contributions | |
goodbye | -0.001 |
fs | -0.001 |
Amen | -0.001 |
Midnight | -0.001 |
forehead | -0.001 |
Token | trophy |
Token ID | 16383 |
Feature activation | -2.85e-02 |
Pos logit contributions | |
udge | +0.011 |
eter | +0.010 |
ipples | +0.009 |
ill | +0.009 |
elly | +0.009 |
Neg logit contributions | |
goodbye | -0.006 |
fs | -0.005 |
Amen | -0.005 |
Midnight | -0.005 |
ishly | -0.004 |
Token | , |
Token ID | 11 |
Feature activation | +2.55e-02 |
Pos logit contributions | |
goodbye | +0.061 |
Amen | +0.056 |
fs | +0.052 |
forehead | +0.050 |
Midnight | +0.049 |
Neg logit contributions | |
udge | -0.051 |
eter | -0.042 |
ill | -0.042 |
ipples | -0.041 |
elly | -0.039 |
Token | shining |
Token ID | 22751 |
Feature activation | +1.03e-02 |
Pos logit contributions | |
udge | +0.049 |
eter | +0.045 |
ipples | +0.043 |
ill | +0.041 |
ased | +0.040 |
Neg logit contributions | |
goodbye | -0.049 |
Amen | -0.043 |
fs | -0.043 |
forehead | -0.040 |
pron | -0.039 |
Token | in |
Token ID | 287 |
Feature activation | +3.15e-02 |
Pos logit contributions | |
udge | +0.023 |
eter | +0.020 |
ipples | +0.020 |
ill | +0.019 |
ased | +0.019 |
Neg logit contributions | |
goodbye | -0.018 |
Amen | -0.015 |
fs | -0.015 |
forehead | -0.014 |
yes | -0.014 |
Token | the |
Token ID | 262 |
Feature activation | +1.30e-02 |
Pos logit contributions | |
udge | +0.080 |
eter | +0.072 |
ipples | +0.071 |
elly | +0.069 |
urious | +0.068 |
Neg logit contributions | |
goodbye | -0.048 |
Amen | -0.034 |
yes | -0.034 |
fs | -0.033 |
saddle | -0.032 |
Token | sunlight |
Token ID | 19606 |
Feature activation | +1.14e-02 |
Pos logit contributions | |
udge | +0.031 |
eter | +0.028 |
ipples | +0.028 |
urious | +0.027 |
ill | +0.027 |
Neg logit contributions | |
goodbye | -0.021 |
fs | -0.017 |
Amen | -0.016 |
yes | -0.015 |
saddle | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | -7.00e-03 |
Pos logit contributions | |
udge | +0.023 |
ipples | +0.020 |
eter | +0.020 |
ill | +0.020 |
elly | +0.018 |
Neg logit contributions | |
goodbye | -0.023 |
Amen | -0.020 |
fs | -0.019 |
Midnight | -0.018 |
yes | -0.018 |
Token | The |
Token ID | 383 |
Feature activation | -1.58e-02 |
Pos logit contributions | |
goodbye | +0.012 |
fs | +0.010 |
Amen | +0.010 |
pron | +0.009 |
yes | +0.009 |
Neg logit contributions | |
udge | -0.016 |
eter | -0.014 |
ipples | -0.014 |
ill | -0.013 |
urious | -0.013 |
Token | a |
Token ID | 257 |
Feature activation | -1.10e-02 |
Pos logit contributions | |
goodbye | +0.037 |
fs | +0.029 |
Amen | +0.028 |
saddle | +0.027 |
yes | +0.027 |
Neg logit contributions | |
udge | -0.053 |
eter | -0.047 |
ipples | -0.047 |
ill | -0.045 |
urious | -0.044 |
Token | good |
Token ID | 922 |
Feature activation | +1.43e-02 |
Pos logit contributions | |
goodbye | +0.019 |
fs | +0.015 |
Amen | +0.014 |
yes | +0.014 |
forehead | +0.014 |
Neg logit contributions | |
udge | -0.025 |
eter | -0.023 |
ipples | -0.022 |
ill | -0.021 |
urious | -0.021 |
Token | friend |
Token ID | 1545 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
udge | +0.032 |
ipples | +0.030 |
eter | +0.029 |
urious | +0.029 |
ill | +0.029 |
Neg logit contributions | |
goodbye | -0.025 |
Amen | -0.019 |
pony | -0.018 |
medal | -0.018 |
yes | -0.018 |
Token | ." |
Token ID | 526 |
Feature activation | -2.02e-02 |
Pos logit contributions | |
goodbye | +0.028 |
fs | +0.026 |
forehead | +0.024 |
chief | +0.024 |
pron | +0.024 |
Neg logit contributions | |
udge | -0.023 |
eter | -0.021 |
ill | -0.020 |
elly | -0.019 |
ipples | -0.019 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
goodbye | +0.043 |
fs | +0.036 |
Amen | +0.036 |
yes | +0.035 |
saddle | +0.034 |
Neg logit contributions | |
udge | -0.038 |
eter | -0.033 |
ipples | -0.032 |
ill | -0.031 |
urious | -0.030 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.60e-04 |
Pos logit contributions | |
goodbye | +0.020 |
fs | +0.016 |
Amen | +0.016 |
yes | +0.015 |
saddle | +0.015 |
Neg logit contributions | |
udge | -0.026 |
eter | -0.023 |
ipples | -0.022 |
urious | -0.021 |
ill | -0.021 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.18e-02 |
Pos logit contributions | |
goodbye | +0.000 |
Amen | +0.000 |
fs | +0.000 |
saddle | +0.000 |
forehead | +0.000 |
Neg logit contributions | |
udge | -0.000 |
eter | -0.000 |
ipples | -0.000 |
ill | -0.000 |
urious | -0.000 |
Token | a |
Token ID | 257 |
Feature activation | -1.18e-02 |
Pos logit contributions | |
udge | +0.046 |
ipples | +0.040 |
eter | +0.039 |
ill | +0.039 |
urious | +0.039 |
Neg logit contributions | |
goodbye | -0.040 |
fs | -0.033 |
Amen | -0.032 |
saddle | -0.031 |
ishly | -0.031 |
Token | time |
Token ID | 640 |
Feature activation | -2.09e-02 |
Pos logit contributions | |
goodbye | +0.022 |
Amen | +0.017 |
yes | +0.016 |
fs | +0.016 |
pron | +0.016 |
Neg logit contributions | |
udge | -0.026 |
eter | -0.024 |
ipples | -0.023 |
ased | -0.022 |
ill | -0.022 |
Token | , |
Token ID | 11 |
Feature activation | +1.82e-02 |
Pos logit contributions | |
goodbye | +0.025 |
fs | +0.016 |
yes | +0.016 |
Amen | +0.016 |
forehead | +0.013 |
Neg logit contributions | |
udge | -0.060 |
eter | -0.056 |
ipples | -0.055 |
urious | -0.053 |
ased | -0.052 |
Token | there |
Token ID | 612 |
Feature activation | -6.41e-03 |
Pos logit contributions | |
udge | +0.041 |
eter | +0.037 |
ipples | +0.036 |
ill | +0.034 |
urious | +0.034 |
Neg logit contributions | |
goodbye | -0.033 |
Amen | -0.026 |
fs | -0.025 |
yes | -0.025 |
pron | -0.023 |
Token | of |
Token ID | 286 |
Feature activation | +2.32e-03 |
Pos logit contributions | |
goodbye | +0.030 |
fs | +0.023 |
Amen | +0.023 |
yes | +0.022 |
saddle | +0.022 |
Neg logit contributions | |
udge | -0.041 |
eter | -0.036 |
ipples | -0.036 |
ill | -0.034 |
urious | -0.034 |
Token | chocolate |
Token ID | 11311 |
Feature activation | -1.54e-02 |
Pos logit contributions | |
udge | +0.006 |
eter | +0.005 |
ipples | +0.005 |
ill | +0.005 |
urious | +0.005 |
Neg logit contributions | |
goodbye | -0.004 |
fs | -0.003 |
Amen | -0.003 |
yes | -0.003 |
saddle | -0.003 |
Token | for |
Token ID | 329 |
Feature activation | -1.41e-03 |
Pos logit contributions | |
goodbye | +0.026 |
Amen | +0.020 |
fs | +0.020 |
yes | +0.019 |
saddle | +0.019 |
Neg logit contributions | |
udge | -0.036 |
eter | -0.032 |
ipples | -0.031 |
ill | -0.030 |
urious | -0.030 |
Token | you |
Token ID | 345 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
goodbye | +0.002 |
Amen | +0.002 |
fs | +0.002 |
yes | +0.002 |
saddle | +0.002 |
Neg logit contributions | |
udge | -0.003 |
eter | -0.003 |
ipples | -0.003 |
ill | -0.003 |
urious | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | +3.13e-02 |
Pos logit contributions | |
udge | +0.038 |
eter | +0.034 |
ipples | +0.033 |
ill | +0.033 |
ierce | +0.031 |
Neg logit contributions | |
goodbye | -0.036 |
Amen | -0.032 |
fs | -0.031 |
saddle | -0.029 |
pron | -0.029 |
Token | little |
Token ID | 1310 |
Feature activation | -9.83e-03 |
Pos logit contributions | |
udge | +0.061 |
eter | +0.052 |
ipples | +0.052 |
urious | +0.049 |
ill | +0.048 |
Neg logit contributions | |
goodbye | -0.067 |
fs | -0.054 |
Amen | -0.054 |
yes | -0.054 |
saddle | -0.051 |
Token | guy |
Token ID | 3516 |
Feature activation | -1.04e-02 |
Pos logit contributions | |
goodbye | +0.020 |
Amen | +0.016 |
fs | +0.016 |
yes | +0.016 |
saddle | +0.015 |
Neg logit contributions | |
udge | -0.020 |
eter | -0.017 |
ipples | -0.017 |
urious | -0.016 |
ill | -0.016 |
Token | ." |
Token ID | 526 |
Feature activation | -9.82e-03 |
Pos logit contributions | |
goodbye | +0.020 |
fs | +0.016 |
Amen | +0.016 |
yes | +0.015 |
forehead | +0.015 |
Neg logit contributions | |
udge | -0.022 |
eter | -0.020 |
ipples | -0.019 |
ill | -0.018 |
ased | -0.018 |
Token | Tim |
Token ID | 5045 |
Feature activation | +2.74e-02 |
Pos logit contributions | |
goodbye | +0.021 |
| +0.017 |
Amen | +0.016 |
yes | +0.016 |
fs | +0.016 |
Neg logit contributions | |
udge | -0.020 |
ipples | -0.017 |
eter | -0.017 |
urious | -0.016 |
elly | -0.016 |
Token | my |
Token ID | 1820 |
Feature activation | +4.49e-03 |
Pos logit contributions | |
udge | +0.052 |
eter | +0.044 |
ipples | +0.044 |
ill | +0.041 |
urious | +0.040 |
Neg logit contributions | |
goodbye | -0.059 |
Amen | -0.048 |
fs | -0.048 |
yes | -0.046 |
forehead | -0.046 |
Token | was |
Token ID | 373 |
Feature activation | -1.81e-02 |
Pos logit contributions | |
udge | +0.008 |
eter | +0.007 |
ipples | +0.007 |
urious | +0.007 |
ill | +0.007 |
Neg logit contributions | |
goodbye | -0.010 |
fs | -0.008 |
Amen | -0.008 |
saddle | -0.008 |
yes | -0.008 |
Token | " |
Token ID | 1 |
Feature activation | +4.50e-02 |
Pos logit contributions | |
udge | +0.011 |
eter | +0.009 |
ipples | +0.009 |
ill | +0.009 |
urious | +0.009 |
Neg logit contributions | |
goodbye | -0.009 |
fs | -0.007 |
Amen | -0.007 |
yes | -0.007 |
forehead | -0.007 |
Token | Good |
Token ID | 10248 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
udge | +0.088 |
eter | +0.076 |
ipples | +0.074 |
urious | +0.072 |
ased | +0.070 |
Neg logit contributions | |
goodbye | -0.092 |
Amen | -0.078 |
fs | -0.077 |
yes | -0.077 |
pron | -0.074 |
Token | fight |
Token ID | 1907 |
Feature activation | -3.75e-02 |
Pos logit contributions | |
udge | +0.037 |
eter | +0.032 |
ipples | +0.032 |
ill | +0.031 |
urious | +0.031 |
Neg logit contributions | |
goodbye | -0.029 |
Amen | -0.023 |
fs | -0.022 |
yes | -0.022 |
forehead | -0.021 |
Token | ," |
Token ID | 553 |
Feature activation | -2.59e-02 |
Pos logit contributions | |
goodbye | +0.068 |
fs | +0.056 |
Amen | +0.055 |
pron | +0.054 |
saddle | +0.053 |
Neg logit contributions | |
udge | -0.081 |
eter | -0.072 |
ipples | -0.069 |
ill | -0.068 |
ased | -0.066 |
Token | Sam |
Token ID | 3409 |
Feature activation | +6.18e-04 |
Pos logit contributions | |
goodbye | +0.053 |
fs | +0.046 |
pron | +0.044 |
Amen | +0.044 |
saddle | +0.044 |
Neg logit contributions | |
udge | -0.049 |
eter | -0.042 |
ill | -0.042 |
ipples | -0.042 |
ased | -0.039 |
Token | said |
Token ID | 531 |
Feature activation | -1.50e-02 |
Pos logit contributions | |
udge | +0.001 |
eter | +0.001 |
ill | +0.001 |
ipples | +0.001 |
ased | +0.001 |
Neg logit contributions | |
goodbye | -0.001 |
Amen | -0.001 |
fs | -0.001 |
pron | -0.001 |
saddle | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +1.82e-02 |
Pos logit contributions | |
goodbye | +0.030 |
fs | +0.030 |
Amen | +0.029 |
pron | +0.029 |
saddle | +0.028 |
Neg logit contributions | |
udge | -0.027 |
eter | -0.023 |
ill | -0.022 |
ipples | -0.020 |
ased | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | +3.64e-03 |
Pos logit contributions | |
udge | +0.038 |
eter | +0.033 |
ipples | +0.032 |
ill | +0.031 |
ierce | +0.030 |
Neg logit contributions | |
goodbye | -0.035 |
fs | -0.029 |
Amen | -0.028 |
saddle | -0.027 |
pron | -0.027 |
Token |
|
Token ID | 198 |
Feature activation | +5.42e-03 |
Pos logit contributions | |
udge | +0.007 |
ipples | +0.006 |
eter | +0.006 |
ill | +0.006 |
urious | +0.006 |
Neg logit contributions | |
goodbye | -0.007 |
fs | -0.006 |
Amen | -0.006 |
yes | -0.006 |
forehead | -0.006 |
Token | Peter |
Token ID | 19727 |
Feature activation | -3.99e-03 |
Pos logit contributions | |
udge | +0.011 |
eter | +0.009 |
ipples | +0.009 |
ill | +0.009 |
urious | +0.009 |
Neg logit contributions | |
goodbye | -0.011 |
fs | -0.009 |
Amen | -0.009 |
yes | -0.009 |
forehead | -0.009 |
Token | smiled |
Token ID | 13541 |
Feature activation | +1.08e-02 |
Pos logit contributions | |
goodbye | +0.007 |
Amen | +0.006 |
fs | +0.006 |
saddle | +0.005 |
pron | +0.005 |
Neg logit contributions | |
udge | -0.009 |
eter | -0.008 |
ipples | -0.008 |
ill | -0.007 |
ased | -0.007 |
Token | waste |
Token ID | 7030 |
Feature activation | -2.48e-04 |
Pos logit contributions | |
udge | +0.029 |
eter | +0.026 |
ipples | +0.026 |
ill | +0.025 |
urious | +0.025 |
Neg logit contributions | |
goodbye | -0.016 |
fs | -0.011 |
yes | -0.011 |
Amen | -0.010 |
helicopter | -0.010 |
Token | in |
Token ID | 287 |
Feature activation | +1.75e-02 |
Pos logit contributions | |
goodbye | +0.000 |
fs | +0.000 |
Amen | +0.000 |
yes | +0.000 |
saddle | +0.000 |
Neg logit contributions | |
udge | -0.001 |
eter | -0.001 |
ipples | -0.000 |
ill | -0.000 |
urious | -0.000 |
Token | the |
Token ID | 262 |
Feature activation | +6.40e-03 |
Pos logit contributions | |
udge | +0.042 |
eter | +0.038 |
ipples | +0.037 |
ill | +0.036 |
urious | +0.036 |
Neg logit contributions | |
goodbye | -0.028 |
fs | -0.022 |
Amen | -0.021 |
yes | -0.020 |
saddle | -0.020 |
Token | water |
Token ID | 1660 |
Feature activation | -4.21e-03 |
Pos logit contributions | |
udge | +0.013 |
ipples | +0.012 |
eter | +0.011 |
urious | +0.011 |
elly | +0.011 |
Neg logit contributions | |
goodbye | -0.013 |
fs | -0.011 |
Amen | -0.010 |
yes | -0.010 |
saddle | -0.010 |
Token | ! |
Token ID | 0 |
Feature activation | +9.95e-03 |
Pos logit contributions | |
goodbye | +0.008 |
Amen | +0.006 |
fs | +0.006 |
yes | +0.006 |
saddle | +0.006 |
Neg logit contributions | |
udge | -0.009 |
ipples | -0.008 |
eter | -0.008 |
ill | -0.008 |
elly | -0.008 |
Token | We |
Token ID | 775 |
Feature activation | -1.32e-02 |
Pos logit contributions | |
udge | +0.021 |
ipples | +0.018 |
eter | +0.018 |
urious | +0.017 |
ill | +0.017 |
Neg logit contributions | |
goodbye | -0.020 |
Amen | -0.016 |
yes | -0.015 |
fs | -0.015 |
Midnight | -0.015 |
Token | need |
Token ID | 761 |
Feature activation | +1.39e-03 |
Pos logit contributions | |
goodbye | +0.027 |
fs | +0.022 |
Amen | +0.022 |
yes | +0.021 |
saddle | +0.021 |
Neg logit contributions | |
udge | -0.026 |
eter | -0.023 |
ipples | -0.022 |
ill | -0.021 |
urious | -0.021 |
Token | to |
Token ID | 284 |
Feature activation | +3.04e-04 |
Pos logit contributions | |
udge | +0.003 |
eter | +0.003 |
ipples | +0.003 |
ased | +0.002 |
ill | +0.002 |
Neg logit contributions | |
goodbye | -0.002 |
fs | -0.002 |
Amen | -0.002 |
pron | -0.002 |
yes | -0.002 |
Token | be |
Token ID | 307 |
Feature activation | +4.37e-03 |
Pos logit contributions | |
udge | +0.001 |
ipples | +0.000 |
eter | +0.000 |
ill | +0.000 |
urious | +0.000 |
Neg logit contributions | |
goodbye | -0.001 |
fs | -0.001 |
Amen | -0.001 |
yes | -0.001 |
pron | -0.001 |
Token | careful |
Token ID | 8161 |
Feature activation | +1.23e-02 |
Pos logit contributions | |
udge | +0.010 |
eter | +0.009 |
ipples | +0.008 |
ill | +0.008 |
urious | +0.008 |
Neg logit contributions | |
goodbye | -0.008 |
fs | -0.006 |
Amen | -0.006 |
yes | -0.006 |
saddle | -0.006 |
Token | !" |
Token ID | 2474 |
Feature activation | -2.92e-02 |
Pos logit contributions | |
udge | +0.024 |
ipples | +0.022 |
eter | +0.021 |
ill | +0.020 |
ased | +0.020 |
Neg logit contributions | |
goodbye | -0.025 |
fs | -0.020 |
Amen | -0.020 |
yes | -0.020 |
pron | -0.019 |
Token |
|
Token ID | 198 |
Feature activation | -1.22e-02 |
Pos logit contributions | |
udge | +0.079 |
eter | +0.072 |
ipples | +0.072 |
urious | +0.069 |
ased | +0.069 |
Neg logit contributions | |
goodbye | -0.046 |
| -0.037 |
fs | -0.035 |
yes | -0.033 |
Midnight | -0.033 |
Token |
|
Token ID | 198 |
Feature activation | -6.49e-03 |
Pos logit contributions | |
goodbye | +0.024 |
Amen | +0.019 |
fs | +0.019 |
yes | +0.018 |
forehead | +0.018 |
Neg logit contributions | |
udge | -0.026 |
eter | -0.022 |
ipples | -0.022 |
ill | -0.021 |
urious | -0.021 |
Token | Her |
Token ID | 9360 |
Feature activation | +1.06e-02 |
Pos logit contributions | |
goodbye | +0.014 |
fs | +0.011 |
Amen | +0.011 |
yes | +0.011 |
forehead | +0.011 |
Neg logit contributions | |
udge | -0.012 |
eter | -0.011 |
ipples | -0.011 |
ill | -0.010 |
urious | -0.010 |
Token | mom |
Token ID | 1995 |
Feature activation | +5.57e-04 |
Pos logit contributions | |
udge | +0.024 |
eter | +0.022 |
ill | +0.021 |
ipples | +0.021 |
urious | +0.020 |
Neg logit contributions | |
goodbye | -0.017 |
Amen | -0.015 |
fs | -0.015 |
saddle | -0.014 |
ishly | -0.013 |
Token | smiled |
Token ID | 13541 |
Feature activation | +1.13e-02 |
Pos logit contributions | |
udge | +0.001 |
eter | +0.001 |
ill | +0.001 |
ipples | +0.001 |
ased | +0.001 |
Neg logit contributions | |
goodbye | -0.001 |
fs | -0.001 |
Amen | -0.001 |
saddle | -0.001 |
forehead | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | -3.99e-02 |
Pos logit contributions | |
udge | +0.023 |
eter | +0.021 |
ill | +0.020 |
ipples | +0.020 |
ased | +0.019 |
Neg logit contributions | |
goodbye | -0.021 |
fs | -0.019 |
Amen | -0.018 |
saddle | -0.017 |
pron | -0.017 |
Token | said |
Token ID | 531 |
Feature activation | -2.48e-02 |
Pos logit contributions | |
goodbye | +0.057 |
Amen | +0.048 |
fs | +0.047 |
saddle | +0.042 |
magazine | +0.041 |
Neg logit contributions | |
udge | -0.098 |
eter | -0.089 |
ipples | -0.086 |
ill | -0.085 |
ased | -0.083 |
Token | , |
Token ID | 11 |
Feature activation | +1.01e-02 |
Pos logit contributions | |
saddle | +0.034 |
goodbye | +0.034 |
Amen | +0.033 |
fs | +0.033 |
forehead | +0.031 |
Neg logit contributions | |
udge | -0.054 |
eter | -0.050 |
ill | -0.048 |
ased | -0.046 |
elly | -0.046 |
Token | � |
Token ID | 6184 |
Feature activation | +6.53e-03 |
Pos logit contributions | |
udge | +0.018 |
eter | +0.016 |
ill | +0.016 |
ierce | +0.015 |
ipples | +0.015 |
Neg logit contributions | |
goodbye | -0.019 |
fs | -0.017 |
saddle | -0.016 |
medal | -0.016 |
Amen | -0.016 |
Token | � |
Token ID | 95 |
Feature activation | +7.55e-03 |
Pos logit contributions | |
udge | +0.013 |
ipples | +0.011 |
eter | +0.011 |
ill | +0.010 |
urious | +0.010 |
Neg logit contributions | |
goodbye | -0.014 |
fs | -0.011 |
Amen | -0.011 |
yes | -0.011 |
pron | -0.011 |
Token | € |
Token ID | 26391 |
Feature activation | -2.42e-02 |
Pos logit contributions | |
udge | +0.015 |
ipples | +0.014 |
eter | +0.014 |
ill | +0.013 |
urious | +0.013 |
Neg logit contributions | |
goodbye | -0.015 |
Amen | -0.012 |
fs | -0.012 |
pron | -0.012 |
yes | -0.011 |
Token | " |
Token ID | 366 |
Feature activation | +4.38e-02 |
Pos logit contributions | |
goodbye | +0.012 |
Amen | +0.010 |
fs | +0.009 |
yes | +0.009 |
forehead | +0.009 |
Neg logit contributions | |
udge | -0.016 |
eter | -0.014 |
ipples | -0.014 |
ill | -0.013 |
ased | -0.013 |
Token | The |
Token ID | 464 |
Feature activation | -1.50e-02 |
Pos logit contributions | |
udge | +0.080 |
ipples | +0.066 |
eter | +0.066 |
urious | +0.064 |
elly | +0.063 |
Neg logit contributions | |
goodbye | -0.090 |
fs | -0.078 |
yes | -0.077 |
Amen | -0.076 |
saddle | -0.075 |
Token | Thunder |
Token ID | 9850 |
Feature activation | +2.07e-02 |
Pos logit contributions | |
goodbye | +0.028 |
fs | +0.023 |
Amen | +0.021 |
pron | +0.021 |
yes | +0.021 |
Neg logit contributions | |
udge | -0.031 |
ipples | -0.028 |
eter | -0.028 |
urious | -0.026 |
ill | -0.025 |
Token | Escape |
Token ID | 14473 |
Feature activation | +3.44e-02 |
Pos logit contributions | |
udge | +0.042 |
eter | +0.037 |
ipples | +0.037 |
ill | +0.035 |
elly | +0.035 |
Neg logit contributions | |
goodbye | -0.040 |
fs | -0.033 |
Amen | -0.032 |
yes | -0.031 |
saddle | -0.031 |
Token | " |
Token ID | 1 |
Feature activation | +4.58e-02 |
Pos logit contributions | |
udge | +0.071 |
eter | +0.063 |
ipples | +0.062 |
ill | +0.059 |
urious | +0.059 |
Neg logit contributions | |
goodbye | -0.066 |
fs | -0.054 |
Amen | -0.053 |
yes | -0.051 |
saddle | -0.050 |
Token | and |
Token ID | 290 |
Feature activation | -4.23e-02 |
Pos logit contributions | |
udge | +0.105 |
eter | +0.092 |
ipples | +0.092 |
urious | +0.088 |
ill | +0.087 |
Neg logit contributions | |
goodbye | -0.079 |
fs | -0.062 |
Amen | -0.060 |
yes | -0.059 |
saddle | -0.057 |
Token | made |
Token ID | 925 |
Feature activation | -2.63e-02 |
Pos logit contributions | |
goodbye | +0.078 |
fs | +0.064 |
Amen | +0.063 |
yes | +0.059 |
pron | +0.059 |
Neg logit contributions | |
udge | -0.090 |
eter | -0.080 |
ipples | -0.080 |
ill | -0.076 |
ased | -0.075 |
Token | it |
Token ID | 340 |
Feature activation | -2.02e-03 |
Pos logit contributions | |
goodbye | +0.047 |
fs | +0.036 |
Amen | +0.035 |
yes | +0.034 |
forehead | +0.033 |
Neg logit contributions | |
udge | -0.059 |
eter | -0.053 |
ipples | -0.052 |
ill | -0.050 |
elly | -0.050 |
Token | their |
Token ID | 511 |
Feature activation | +1.91e-03 |
Pos logit contributions | |
goodbye | +0.004 |
fs | +0.003 |
Amen | +0.003 |
yes | +0.003 |
pron | +0.003 |
Neg logit contributions | |
udge | -0.005 |
eter | -0.004 |
ipples | -0.004 |
ill | -0.004 |
urious | -0.004 |
Token | mission |
Token ID | 4365 |
Feature activation | -1.91e-02 |
Pos logit contributions | |
udge | +0.004 |
eter | +0.004 |
ipples | +0.004 |
urious | +0.003 |
ill | +0.003 |
Neg logit contributions | |
goodbye | -0.003 |
fs | -0.003 |
Amen | -0.003 |
yes | -0.003 |
pron | -0.003 |
Token | to |
Token ID | 284 |
Feature activation | -9.07e-04 |
Pos logit contributions | |
goodbye | +0.036 |
Amen | +0.031 |
fs | +0.030 |
pron | +0.028 |
saddle | +0.028 |
Neg logit contributions | |
udge | -0.039 |
eter | -0.035 |
ipples | -0.034 |
ill | -0.033 |
ased | -0.032 |
Token | his |
Token ID | 465 |
Feature activation | -7.11e-03 |
Pos logit contributions | |
goodbye | +0.032 |
Amen | +0.027 |
fs | +0.026 |
forehead | +0.025 |
saddle | +0.025 |
Neg logit contributions | |
udge | -0.039 |
eter | -0.034 |
ipples | -0.033 |
ill | -0.033 |
ased | -0.032 |
Token | symbols |
Token ID | 14354 |
Feature activation | -1.61e-02 |
Pos logit contributions | |
goodbye | +0.013 |
Amen | +0.012 |
fs | +0.011 |
ishly | +0.011 |
saddle | +0.010 |
Neg logit contributions | |
udge | -0.015 |
eter | -0.013 |
ipples | -0.012 |
ill | -0.012 |
urious | -0.012 |
Token | back |
Token ID | 736 |
Feature activation | +3.62e-02 |
Pos logit contributions | |
goodbye | +0.032 |
Amen | +0.031 |
fs | +0.029 |
ishly | +0.028 |
chief | +0.028 |
Neg logit contributions | |
udge | -0.028 |
ill | -0.026 |
eter | -0.025 |
ipples | -0.025 |
ased | -0.023 |
Token | . |
Token ID | 13 |
Feature activation | +2.63e-02 |
Pos logit contributions | |
udge | +0.076 |
eter | +0.067 |
ipples | +0.065 |
ill | +0.064 |
urious | +0.062 |
Neg logit contributions | |
goodbye | -0.069 |
Amen | -0.057 |
fs | -0.057 |
yes | -0.053 |
pron | -0.053 |
Token | Lily |
Token ID | 20037 |
Feature activation | -8.57e-04 |
Pos logit contributions | |
udge | +0.067 |
ipples | +0.062 |
eter | +0.061 |
rying | +0.060 |
urious | +0.058 |
Neg logit contributions | |
goodbye | -0.039 |
| -0.031 |
Amen | -0.029 |
yes | -0.028 |
helmet | -0.027 |
Token | and |
Token ID | 290 |
Feature activation | -5.72e-02 |
Pos logit contributions | |
goodbye | +0.002 |
fs | +0.002 |
Amen | +0.002 |
pron | +0.002 |
yes | +0.002 |
Neg logit contributions | |
udge | -0.001 |
eter | -0.001 |
ipples | -0.001 |
urious | -0.001 |
ill | -0.001 |
Token | Ben |
Token ID | 3932 |
Feature activation | -1.64e-02 |
Pos logit contributions | |
goodbye | +0.122 |
Amen | +0.095 |
fs | +0.095 |
yes | +0.094 |
saddle | +0.092 |
Neg logit contributions | |
udge | -0.113 |
eter | -0.098 |
ipples | -0.097 |
urious | -0.092 |
elly | -0.091 |
Token | are |
Token ID | 389 |
Feature activation | -2.29e-02 |
Pos logit contributions | |
goodbye | +0.034 |
fs | +0.032 |
pron | +0.032 |
Amen | +0.030 |
saddle | +0.029 |
Neg logit contributions | |
udge | -0.027 |
ipples | -0.024 |
urious | -0.023 |
ill | -0.022 |
eter | -0.022 |
Token | sad |
Token ID | 6507 |
Feature activation | +2.07e-02 |
Pos logit contributions | |
goodbye | +0.037 |
fs | +0.034 |
pron | +0.029 |
saddle | +0.029 |
chief | +0.029 |
Neg logit contributions | |
udge | -0.052 |
ipples | -0.048 |
eter | -0.046 |
asured | -0.045 |
Esk | -0.044 |
Token | too |
Token ID | 1165 |
Feature activation | +1.86e-02 |
Pos logit contributions | |
udge | +0.050 |
eter | +0.045 |
ipples | +0.044 |
urious | +0.042 |
ill | +0.042 |
Neg logit contributions | |
goodbye | -0.034 |
Amen | -0.026 |
fs | -0.025 |
yes | -0.025 |
forehead | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | +2.73e-02 |
Pos logit contributions | |
udge | +0.037 |
eter | +0.033 |
ipples | +0.032 |
urious | +0.031 |
elly | +0.030 |
Neg logit contributions | |
goodbye | -0.037 |
fs | -0.031 |
Amen | -0.030 |
pron | -0.030 |
saddle | -0.029 |
Token | how |
Token ID | 703 |
Feature activation | +3.47e-02 |
Pos logit contributions | |
udge | +0.024 |
ipples | +0.021 |
eter | +0.020 |
ill | +0.020 |
ierce | +0.019 |
Neg logit contributions | |
goodbye | -0.023 |
fs | -0.020 |
pron | -0.020 |
Amen | -0.019 |
forehead | -0.019 |
Token | everyone |
Token ID | 2506 |
Feature activation | +2.22e-02 |
Pos logit contributions | |
udge | +0.076 |
eter | +0.070 |
ipples | +0.068 |
urious | +0.067 |
ierce | +0.064 |
Neg logit contributions | |
goodbye | -0.055 |
fs | -0.049 |
Amen | -0.046 |
pron | -0.046 |
forehead | -0.045 |
Token | was |
Token ID | 373 |
Feature activation | -8.51e-04 |
Pos logit contributions | |
udge | +0.049 |
eter | +0.044 |
ipples | +0.042 |
urious | +0.041 |
ill | +0.041 |
Neg logit contributions | |
goodbye | -0.041 |
Amen | -0.032 |
fs | -0.032 |
yes | -0.031 |
medal | -0.030 |
Token | able |
Token ID | 1498 |
Feature activation | -7.09e-03 |
Pos logit contributions | |
goodbye | +0.002 |
fs | +0.001 |
Amen | +0.001 |
pron | +0.001 |
saddle | +0.001 |
Neg logit contributions | |
udge | -0.002 |
ipples | -0.002 |
eter | -0.001 |
ill | -0.001 |
aught | -0.001 |
Token | to |
Token ID | 284 |
Feature activation | +7.67e-03 |
Pos logit contributions | |
goodbye | +0.014 |
fs | +0.013 |
Amen | +0.013 |
pron | +0.012 |
forehead | +0.012 |
Neg logit contributions | |
udge | -0.013 |
eter | -0.012 |
ipples | -0.011 |
ill | -0.011 |
urious | -0.011 |
Token | make |
Token ID | 787 |
Feature activation | -3.50e-02 |
Pos logit contributions | |
udge | +0.015 |
ipples | +0.013 |
eter | +0.012 |
ill | +0.012 |
ierce | +0.011 |
Neg logit contributions | |
goodbye | -0.016 |
fs | -0.014 |
Amen | -0.013 |
pron | -0.013 |
forehead | -0.012 |
Token | nice |
Token ID | 3621 |
Feature activation | +6.66e-03 |
Pos logit contributions | |
goodbye | +0.059 |
fs | +0.046 |
Amen | +0.046 |
yes | +0.044 |
saddle | +0.043 |
Neg logit contributions | |
udge | -0.082 |
eter | -0.073 |
ipples | -0.072 |
ill | -0.069 |
urious | -0.068 |
Token | sounds |
Token ID | 5238 |
Feature activation | +1.33e-02 |
Pos logit contributions | |
udge | +0.015 |
eter | +0.013 |
ipples | +0.013 |
urious | +0.013 |
ill | +0.013 |
Neg logit contributions | |
goodbye | -0.012 |
Amen | -0.009 |
fs | -0.009 |
yes | -0.009 |
forehead | -0.009 |
Token | together |
Token ID | 1978 |
Feature activation | +3.70e-03 |
Pos logit contributions | |
udge | +0.025 |
eter | +0.022 |
ipples | +0.021 |
ill | +0.021 |
elly | +0.020 |
Neg logit contributions | |
goodbye | -0.027 |
Amen | -0.024 |
fs | -0.024 |
forehead | -0.022 |
pron | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | -1.74e-02 |
Pos logit contributions | |
udge | +0.007 |
eter | +0.007 |
ipples | +0.007 |
ill | +0.006 |
urious | +0.006 |
Neg logit contributions | |
goodbye | -0.007 |
Amen | -0.006 |
fs | -0.006 |
pron | -0.006 |
saddle | -0.006 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -9.28e-04 |
Pos logit contributions | |
goodbye | +0.030 |
fs | +0.023 |
Amen | +0.023 |
yes | +0.022 |
saddle | +0.022 |
Neg logit contributions | |
udge | -0.040 |
eter | -0.036 |
ipples | -0.035 |
ill | -0.034 |
urious | -0.034 |
Token | than |
Token ID | 621 |
Feature activation | +3.82e-02 |
Pos logit contributions | |
udge | +0.095 |
eter | +0.085 |
ipples | +0.084 |
ill | +0.081 |
urious | +0.080 |
Neg logit contributions | |
goodbye | -0.077 |
fs | -0.064 |
Amen | -0.064 |
pron | -0.060 |
yes | -0.059 |
Token | ever |
Token ID | 1683 |
Feature activation | +3.47e-02 |
Pos logit contributions | |
udge | +0.069 |
eter | +0.060 |
ipples | +0.059 |
ill | +0.057 |
ierce | +0.054 |
Neg logit contributions | |
goodbye | -0.082 |
Amen | -0.072 |
fs | -0.070 |
saddle | -0.067 |
pron | -0.066 |
Token | before |
Token ID | 878 |
Feature activation | +1.54e-02 |
Pos logit contributions | |
udge | +0.080 |
eter | +0.069 |
ipples | +0.068 |
urious | +0.066 |
ierce | +0.065 |
Neg logit contributions | |
goodbye | -0.064 |
yes | -0.047 |
fs | -0.047 |
Amen | -0.046 |
forehead | -0.045 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e-02 |
Pos logit contributions | |
udge | +0.029 |
eter | +0.025 |
ipples | +0.025 |
ill | +0.024 |
urious | +0.023 |
Neg logit contributions | |
goodbye | -0.032 |
Amen | -0.027 |
fs | -0.027 |
forehead | -0.026 |
pron | -0.026 |
Token | From |
Token ID | 3574 |
Feature activation | -1.12e-02 |
Pos logit contributions | |
goodbye | +0.019 |
| +0.014 |
Amen | +0.014 |
fs | +0.013 |
yes | +0.013 |
Neg logit contributions | |
udge | -0.024 |
ipples | -0.022 |
eter | -0.021 |
ased | -0.021 |
urious | -0.020 |
Token | then |
Token ID | 788 |
Feature activation | -3.80e-02 |
Pos logit contributions | |
goodbye | +0.016 |
fs | +0.013 |
Amen | +0.013 |
pron | +0.012 |
forehead | +0.012 |
Neg logit contributions | |
udge | -0.028 |
ipples | -0.025 |
eter | -0.025 |
urious | -0.024 |
ased | -0.023 |
Token | on |
Token ID | 319 |
Feature activation | -3.42e-02 |
Pos logit contributions | |
goodbye | +0.049 |
Amen | +0.038 |
forehead | +0.036 |
pron | +0.035 |
fs | +0.034 |
Neg logit contributions | |
udge | -0.101 |
eter | -0.088 |
ipples | -0.088 |
ased | -0.087 |
elly | -0.086 |
Token | , |
Token ID | 11 |
Feature activation | -2.93e-02 |
Pos logit contributions | |
goodbye | +0.056 |
Amen | +0.053 |
forehead | +0.049 |
saddle | +0.048 |
fs | +0.046 |
Neg logit contributions | |
udge | -0.071 |
eter | -0.063 |
ill | -0.061 |
ipples | -0.061 |
ierce | -0.060 |
Token | they |
Token ID | 484 |
Feature activation | -2.13e-02 |
Pos logit contributions | |
goodbye | +0.036 |
fs | +0.025 |
Amen | +0.024 |
yes | +0.023 |
saddle | +0.022 |
Neg logit contributions | |
udge | -0.083 |
eter | -0.075 |
ipples | -0.074 |
urious | -0.072 |
ill | -0.072 |
Token | were |
Token ID | 547 |
Feature activation | -2.65e-02 |
Pos logit contributions | |
goodbye | +0.040 |
fs | +0.032 |
Amen | +0.032 |
yes | +0.031 |
saddle | +0.030 |
Neg logit contributions | |
udge | -0.046 |
eter | -0.040 |
ipples | -0.040 |
ill | -0.038 |
urious | -0.037 |
Token | the |
Token ID | 262 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
goodbye | +0.049 |
fs | +0.043 |
Amen | +0.041 |
forehead | +0.039 |
yes | +0.038 |
Neg logit contributions | |
udge | -0.054 |
eter | -0.050 |
ipples | -0.050 |
ill | -0.046 |
ased | -0.046 |