Token | had |
Token ID | 550 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
permission | +0.080 |
help | +0.069 |
warning | +0.068 |
unknow | +0.067 |
pap | +0.067 |
Neg logit contributions | |
ing | -0.149 |
ane | -0.139 |
he | -0.137 |
ays | -0.127 |
eddy | -0.125 |
Token | drawn |
Token ID | 7428 |
Feature activation | -2.98e-02 |
Pos logit contributions | |
pap | +0.049 |
forgot | +0.046 |
regret | +0.045 |
robbers | +0.044 |
help | +0.044 |
Neg logit contributions | |
ing | -0.128 |
ane | -0.122 |
he | -0.119 |
ays | -0.114 |
ed | -0.110 |
Token | . |
Token ID | 13 |
Feature activation | -3.19e-02 |
Pos logit contributions | |
ing | +0.056 |
ays | +0.055 |
ane | +0.053 |
eddy | +0.052 |
he | +0.051 |
Neg logit contributions | |
never | -0.058 |
not | -0.055 |
forgotten | -0.051 |
permission | -0.051 |
been | -0.051 |
Token | He |
Token ID | 679 |
Feature activation | +6.77e-02 |
Pos logit contributions | |
ing | +0.063 |
ane | +0.056 |
ays | +0.056 |
he | +0.053 |
eddy | +0.053 |
Neg logit contributions | |
never | -0.062 |
forgot | -0.058 |
forgotten | -0.058 |
grown | -0.057 |
not | -0.057 |
Token | would |
Token ID | 561 |
Feature activation | +7.13e-02 |
Pos logit contributions | |
permission | +0.094 |
stolen | +0.088 |
tresp | +0.086 |
robbers | +0.086 |
help | +0.085 |
Neg logit contributions | |
ing | -0.154 |
ane | -0.147 |
ays | -0.137 |
he | -0.136 |
ement | -0.134 |
Token | also |
Token ID | 635 |
Feature activation | +9.73e-02 |
Pos logit contributions | |
stolen | +0.076 |
forgot | +0.075 |
forgotten | +0.074 |
tresp | +0.072 |
wasted | +0.071 |
Neg logit contributions | |
he | -0.173 |
ane | -0.169 |
ing | -0.162 |
ay | -0.157 |
room | -0.150 |
Token | tell |
Token ID | 1560 |
Feature activation | +1.35e-02 |
Pos logit contributions | |
tresp | +0.103 |
stolen | +0.100 |
forgot | +0.098 |
forgotten | +0.098 |
wasted | +0.096 |
Neg logit contributions | |
he | -0.216 |
ane | -0.209 |
ing | -0.197 |
ay | -0.196 |
ite | -0.195 |
Token | her |
Token ID | 607 |
Feature activation | +3.11e-02 |
Pos logit contributions | |
forgot | +0.021 |
forgotten | +0.020 |
never | +0.020 |
wasted | +0.019 |
stolen | +0.019 |
Neg logit contributions | |
ing | -0.031 |
ane | -0.027 |
he | -0.027 |
ays | -0.026 |
ement | -0.026 |
Token | how |
Token ID | 703 |
Feature activation | +3.07e-02 |
Pos logit contributions | |
forgot | +0.046 |
never | +0.046 |
forgotten | +0.045 |
stolen | +0.044 |
wasted | +0.043 |
Neg logit contributions | |
ing | -0.073 |
ane | -0.068 |
ays | -0.064 |
he | -0.063 |
eddy | -0.062 |
Token | beautiful |
Token ID | 4950 |
Feature activation | -8.37e-03 |
Pos logit contributions | |
forgot | +0.046 |
knocked | +0.043 |
never | +0.042 |
stolen | +0.042 |
unknow | +0.042 |
Neg logit contributions | |
ing | -0.073 |
ane | -0.064 |
he | -0.060 |
eddy | -0.059 |
ays | -0.056 |
Token | each |
Token ID | 1123 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
ing | +0.020 |
ane | +0.019 |
ays | +0.019 |
eddy | +0.018 |
he | +0.018 |
Neg logit contributions | |
never | -0.012 |
forgot | -0.012 |
forgotten | -0.011 |
not | -0.011 |
wasted | -0.011 |
Token | the |
Token ID | 262 |
Feature activation | +1.31e-02 |
Pos logit contributions | |
ing | +0.046 |
ane | +0.045 |
ays | +0.045 |
eddy | +0.043 |
he | +0.042 |
Neg logit contributions | |
never | -0.018 |
help | -0.016 |
not | -0.015 |
forgot | -0.015 |
permission | -0.015 |
Token | bear |
Token ID | 6842 |
Feature activation | -4.45e-02 |
Pos logit contributions | |
never | +0.022 |
forgot | +0.021 |
forgotten | +0.021 |
wasted | +0.020 |
not | +0.019 |
Neg logit contributions | |
ing | -0.030 |
ane | -0.027 |
eddy | -0.027 |
he | -0.026 |
ays | -0.026 |
Token | , |
Token ID | 11 |
Feature activation | +2.77e-02 |
Pos logit contributions | |
ing | +0.086 |
ays | +0.084 |
ane | +0.083 |
eddy | +0.079 |
he | +0.076 |
Neg logit contributions | |
never | -0.086 |
forgot | -0.081 |
help | -0.080 |
not | -0.080 |
forgotten | -0.078 |
Token | she |
Token ID | 673 |
Feature activation | +6.99e-02 |
Pos logit contributions | |
forgot | +0.050 |
never | +0.049 |
forgotten | +0.048 |
wasted | +0.047 |
stolen | +0.047 |
Neg logit contributions | |
ing | -0.058 |
ane | -0.051 |
he | -0.049 |
ays | -0.049 |
eddy | -0.047 |
Token | would |
Token ID | 561 |
Feature activation | +7.30e-02 |
Pos logit contributions | |
permission | +0.087 |
stolen | +0.080 |
forgotten | +0.078 |
help | +0.077 |
tresp | +0.076 |
Neg logit contributions | |
ing | -0.178 |
ane | -0.168 |
he | -0.159 |
ays | -0.159 |
eddy | -0.155 |
Token | also |
Token ID | 635 |
Feature activation | +9.64e-02 |
Pos logit contributions | |
forgot | +0.078 |
stolen | +0.077 |
forgotten | +0.076 |
permission | +0.072 |
known | +0.072 |
Neg logit contributions | |
he | -0.190 |
ane | -0.183 |
ing | -0.182 |
ay | -0.171 |
eddy | -0.169 |
Token | feel |
Token ID | 1254 |
Feature activation | +4.81e-02 |
Pos logit contributions | |
forgot | +0.105 |
forgotten | +0.104 |
stolen | +0.103 |
vitamins | +0.103 |
robbers | +0.102 |
Neg logit contributions | |
he | -0.217 |
ane | -0.203 |
ite | -0.199 |
ing | -0.194 |
ay | -0.189 |
Token | the |
Token ID | 262 |
Feature activation | +3.68e-02 |
Pos logit contributions | |
forgot | +0.070 |
robbers | +0.068 |
knocked | +0.066 |
never | +0.065 |
stolen | +0.065 |
Neg logit contributions | |
ing | -0.111 |
he | -0.103 |
ane | -0.096 |
ays | -0.094 |
ement | -0.094 |
Token | number |
Token ID | 1271 |
Feature activation | -3.04e-02 |
Pos logit contributions | |
forgot | +0.052 |
never | +0.050 |
wasted | +0.048 |
forgotten | +0.047 |
knocked | +0.046 |
Neg logit contributions | |
ing | -0.090 |
he | -0.083 |
ane | -0.083 |
eddy | -0.080 |
ay | -0.078 |
Token | on |
Token ID | 319 |
Feature activation | +1.37e-02 |
Pos logit contributions | |
ing | +0.050 |
ays | +0.049 |
ane | +0.047 |
eddy | +0.046 |
he | +0.043 |
Neg logit contributions | |
never | -0.066 |
help | -0.062 |
forgot | -0.062 |
not | -0.061 |
forgotten | -0.061 |
Token | it |
Token ID | 340 |
Feature activation | -6.76e-03 |
Pos logit contributions | |
never | +0.017 |
forgot | +0.017 |
forgotten | +0.016 |
wasted | +0.015 |
stolen | +0.015 |
Neg logit contributions | |
ing | -0.036 |
ane | -0.034 |
ays | -0.033 |
he | -0.032 |
eddy | -0.032 |
Token | was |
Token ID | 373 |
Feature activation | +5.99e-02 |
Pos logit contributions | |
permission | +0.060 |
never | +0.060 |
forgot | +0.056 |
help | +0.056 |
forgotten | +0.056 |
Neg logit contributions | |
ing | -0.141 |
ane | -0.133 |
ays | -0.130 |
he | -0.128 |
ement | -0.123 |
Token | right |
Token ID | 826 |
Feature activation | +3.26e-04 |
Pos logit contributions | |
forgot | +0.070 |
never | +0.069 |
help | +0.065 |
permission | +0.064 |
wished | +0.062 |
Neg logit contributions | |
ing | -0.154 |
ane | -0.144 |
he | -0.143 |
ays | -0.139 |
ement | -0.135 |
Token | . |
Token ID | 13 |
Feature activation | -2.94e-02 |
Pos logit contributions | |
never | +0.001 |
forgotten | +0.001 |
forgot | +0.001 |
not | +0.001 |
permission | +0.001 |
Neg logit contributions | |
ane | -0.001 |
ing | -0.001 |
ays | -0.001 |
he | -0.000 |
eddy | -0.000 |
Token | They |
Token ID | 1119 |
Feature activation | +7.21e-02 |
Pos logit contributions | |
ing | +0.063 |
ane | +0.058 |
ays | +0.056 |
he | +0.054 |
eddy | +0.054 |
Neg logit contributions | |
never | -0.054 |
forgot | -0.052 |
not | -0.051 |
forgotten | -0.051 |
grown | -0.050 |
Token | would |
Token ID | 561 |
Feature activation | +7.09e-02 |
Pos logit contributions | |
vitamins | +0.102 |
permission | +0.101 |
tresp | +0.098 |
supposed | +0.097 |
stolen | +0.097 |
Neg logit contributions | |
he | -0.161 |
ing | -0.158 |
ane | -0.149 |
ite | -0.133 |
ays | -0.129 |
Token | also |
Token ID | 635 |
Feature activation | +9.51e-02 |
Pos logit contributions | |
forgot | +0.074 |
stolen | +0.073 |
forgotten | +0.072 |
wasted | +0.066 |
permission | +0.066 |
Neg logit contributions | |
he | -0.182 |
ane | -0.176 |
ing | -0.173 |
ay | -0.162 |
eddy | -0.160 |
Token | often |
Token ID | 1690 |
Feature activation | +7.54e-02 |
Pos logit contributions | |
vitamins | +0.097 |
forgot | +0.094 |
stolen | +0.092 |
forgotten | +0.092 |
tresp | +0.092 |
Neg logit contributions | |
he | -0.222 |
ing | -0.215 |
ane | -0.214 |
ite | -0.205 |
ay | -0.194 |
Token | stop |
Token ID | 2245 |
Feature activation | +2.13e-02 |
Pos logit contributions | |
forgot | +0.072 |
forgotten | +0.071 |
stolen | +0.071 |
wasted | +0.070 |
never | +0.069 |
Neg logit contributions | |
ing | -0.194 |
ane | -0.187 |
he | -0.183 |
ite | -0.171 |
ays | -0.170 |
Token | to |
Token ID | 284 |
Feature activation | +5.95e-02 |
Pos logit contributions | |
never | +0.030 |
forgot | +0.029 |
forgotten | +0.028 |
wasted | +0.027 |
stolen | +0.026 |
Neg logit contributions | |
ing | -0.054 |
ane | -0.049 |
ays | -0.049 |
eddy | -0.047 |
he | -0.046 |
Token | share |
Token ID | 2648 |
Feature activation | -6.71e-03 |
Pos logit contributions | |
forgot | +0.073 |
forgotten | +0.064 |
unknow | +0.063 |
wasted | +0.062 |
stolen | +0.061 |
Neg logit contributions | |
ing | -0.142 |
he | -0.139 |
ane | -0.134 |
ay | -0.127 |
ays | -0.125 |
Token | a |
Token ID | 257 |
Feature activation | +1.32e-02 |
Pos logit contributions | |
ing | +0.017 |
ane | +0.016 |
ays | +0.016 |
eddy | +0.015 |
he | +0.015 |
Neg logit contributions | |
never | -0.009 |
forgot | -0.009 |
forgotten | -0.009 |
not | -0.009 |
help | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | -3.50e-02 |
Pos logit contributions | |
ing | +0.057 |
ane | +0.054 |
ays | +0.052 |
ement | +0.051 |
eddy | +0.050 |
Neg logit contributions | |
never | -0.049 |
not | -0.049 |
been | -0.045 |
known | -0.044 |
forgot | -0.044 |
Token |
|
Token ID | 198 |
Feature activation | -2.25e-02 |
Pos logit contributions | |
ing | +0.073 |
ays | +0.069 |
ane | +0.068 |
eddy | +0.063 |
ement | +0.061 |
Neg logit contributions | |
never | -0.069 |
not | -0.062 |
known | -0.061 |
help | -0.060 |
forgot | -0.060 |
Token | The |
Token ID | 464 |
Feature activation | +1.09e-02 |
Pos logit contributions | |
ing | +0.045 |
ane | +0.042 |
ays | +0.041 |
eddy | +0.039 |
he | +0.038 |
Neg logit contributions | |
never | -0.043 |
forgot | -0.042 |
forgotten | -0.041 |
not | -0.040 |
wasted | -0.039 |
Token | farmer |
Token ID | 18739 |
Feature activation | +5.45e-03 |
Pos logit contributions | |
forgot | +0.016 |
wasted | +0.016 |
never | +0.015 |
forgotten | +0.015 |
stolen | +0.015 |
Neg logit contributions | |
ing | -0.026 |
ane | -0.023 |
he | -0.022 |
eddy | -0.022 |
over | -0.022 |
Token | would |
Token ID | 561 |
Feature activation | +7.29e-02 |
Pos logit contributions | |
forgotten | +0.008 |
forgot | +0.008 |
never | +0.008 |
permission | +0.008 |
stolen | +0.008 |
Neg logit contributions | |
ing | -0.013 |
ane | -0.012 |
ays | -0.011 |
he | -0.011 |
eddy | -0.011 |
Token | also |
Token ID | 635 |
Feature activation | +9.45e-02 |
Pos logit contributions | |
forgot | +0.080 |
forgotten | +0.075 |
stolen | +0.074 |
wasted | +0.070 |
permission | +0.070 |
Neg logit contributions | |
he | -0.182 |
ing | -0.165 |
ane | -0.161 |
ay | -0.157 |
amb | -0.152 |
Token | scratch |
Token ID | 12692 |
Feature activation | -2.56e-02 |
Pos logit contributions | |
forgot | +0.105 |
forgotten | +0.100 |
setup | +0.100 |
stolen | +0.097 |
tresp | +0.096 |
Neg logit contributions | |
he | -0.210 |
ing | -0.196 |
ane | -0.189 |
ite | -0.186 |
ore | -0.184 |
Token | the |
Token ID | 262 |
Feature activation | +1.88e-02 |
Pos logit contributions | |
ing | +0.056 |
ane | +0.054 |
ays | +0.053 |
eddy | +0.051 |
he | +0.050 |
Neg logit contributions | |
never | -0.043 |
forgot | -0.041 |
forgotten | -0.040 |
not | -0.039 |
wasted | -0.038 |
Token | sheep |
Token ID | 15900 |
Feature activation | -1.94e-02 |
Pos logit contributions | |
never | +0.031 |
forgot | +0.030 |
known | +0.030 |
wasted | +0.029 |
not | +0.028 |
Neg logit contributions | |
ing | -0.043 |
eddy | -0.036 |
ane | -0.035 |
he | -0.034 |
ays | -0.034 |
Token | 's |
Token ID | 338 |
Feature activation | +3.00e-02 |
Pos logit contributions | |
ing | +0.037 |
ane | +0.034 |
ays | +0.034 |
eddy | +0.032 |
he | +0.032 |
Neg logit contributions | |
never | -0.038 |
forgot | -0.037 |
forgotten | -0.036 |
not | -0.035 |
wasted | -0.035 |
Token | heads |
Token ID | 6665 |
Feature activation | -2.02e-02 |
Pos logit contributions | |
forgot | +0.045 |
never | +0.043 |
known | +0.042 |
wasted | +0.041 |
forgotten | +0.040 |
Neg logit contributions | |
ing | -0.071 |
eddy | -0.063 |
ane | -0.062 |
ement | -0.060 |
over | -0.060 |
Token | bad |
Token ID | 2089 |
Feature activation | -5.31e-03 |
Pos logit contributions | |
forgot | +0.069 |
wasted | +0.067 |
never | +0.063 |
permission | +0.062 |
not | +0.062 |
Neg logit contributions | |
ing | -0.137 |
ane | -0.128 |
eddy | -0.121 |
he | -0.119 |
ay | -0.117 |
Token | thing |
Token ID | 1517 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
ing | +0.012 |
ays | +0.011 |
ane | +0.011 |
he | +0.010 |
eddy | +0.010 |
Neg logit contributions | |
never | -0.009 |
forgot | -0.009 |
forgotten | -0.009 |
not | -0.008 |
help | -0.008 |
Token | and |
Token ID | 290 |
Feature activation | +6.64e-02 |
Pos logit contributions | |
ing | +0.032 |
ane | +0.030 |
ays | +0.030 |
eddy | +0.028 |
he | +0.028 |
Neg logit contributions | |
never | -0.029 |
forgot | -0.028 |
forgotten | -0.027 |
not | -0.027 |
help | -0.026 |
Token | that |
Token ID | 326 |
Feature activation | +3.31e-02 |
Pos logit contributions | |
forgot | +0.080 |
forgotten | +0.079 |
stolen | +0.078 |
permission | +0.078 |
wasted | +0.077 |
Neg logit contributions | |
ing | -0.164 |
ane | -0.142 |
ite | -0.130 |
he | -0.128 |
ays | -0.128 |
Token | everything |
Token ID | 2279 |
Feature activation | +2.38e-02 |
Pos logit contributions | |
forgotten | +0.045 |
forgot | +0.044 |
stolen | +0.043 |
permission | +0.043 |
never | +0.043 |
Neg logit contributions | |
ing | -0.082 |
ane | -0.076 |
ays | -0.070 |
he | -0.068 |
ite | -0.067 |
Token | will |
Token ID | 481 |
Feature activation | +9.43e-02 |
Pos logit contributions | |
permission | +0.036 |
forgotten | +0.035 |
known | +0.035 |
stolen | +0.035 |
forgot | +0.034 |
Neg logit contributions | |
ing | -0.056 |
ane | -0.050 |
ays | -0.047 |
he | -0.047 |
ement | -0.044 |
Token | be |
Token ID | 307 |
Feature activation | +6.22e-02 |
Pos logit contributions | |
robbers | +0.098 |
pap | +0.096 |
forgot | +0.092 |
wasted | +0.090 |
forgotten | +0.090 |
Neg logit contributions | |
he | -0.232 |
ing | -0.218 |
ane | -0.203 |
ay | -0.198 |
ore | -0.182 |
Token | okay |
Token ID | 8788 |
Feature activation | -2.16e-02 |
Pos logit contributions | |
forgot | +0.085 |
never | +0.080 |
forgotten | +0.079 |
permission | +0.079 |
unknow | +0.079 |
Neg logit contributions | |
ing | -0.146 |
ane | -0.127 |
ays | -0.124 |
eddy | -0.122 |
he | -0.122 |
Token | ." |
Token ID | 526 |
Feature activation | -4.60e-02 |
Pos logit contributions | |
ing | +0.041 |
ays | +0.040 |
ane | +0.040 |
he | +0.038 |
eddy | +0.037 |
Neg logit contributions | |
never | -0.043 |
not | -0.040 |
forgot | -0.040 |
help | -0.039 |
forgotten | -0.039 |
Token | |
Token ID | 220 |
Feature activation | -3.05e-02 |
Pos logit contributions | |
ing | +0.113 |
ane | +0.109 |
ays | +0.105 |
eddy | +0.102 |
he | +0.101 |
Neg logit contributions | |
never | -0.066 |
forgot | -0.064 |
help | -0.060 |
forgotten | -0.060 |
not | -0.059 |
Token |
|
Token ID | 198 |
Feature activation | -3.82e-02 |
Pos logit contributions | |
ing | +0.061 |
ane | +0.059 |
ays | +0.057 |
ement | +0.054 |
eddy | +0.054 |
Neg logit contributions | |
never | -0.055 |
not | -0.054 |
forgot | -0.052 |
help | -0.050 |
grown | -0.050 |
Token | our |
Token ID | 674 |
Feature activation | +4.20e-02 |
Pos logit contributions | |
forgot | +0.016 |
never | +0.015 |
forgotten | +0.014 |
stolen | +0.014 |
known | +0.014 |
Neg logit contributions | |
ing | -0.029 |
ane | -0.027 |
ays | -0.027 |
he | -0.026 |
eddy | -0.026 |
Token | journey |
Token ID | 7002 |
Feature activation | -5.18e-02 |
Pos logit contributions | |
known | +0.060 |
forgot | +0.059 |
forgotten | +0.055 |
knocked | +0.055 |
wasted | +0.054 |
Neg logit contributions | |
ing | -0.092 |
eddy | -0.087 |
ane | -0.087 |
ement | -0.083 |
he | -0.082 |
Token | , |
Token ID | 11 |
Feature activation | +3.35e-02 |
Pos logit contributions | |
ing | +0.094 |
ays | +0.088 |
ane | +0.087 |
eddy | +0.082 |
he | +0.079 |
Neg logit contributions | |
never | -0.108 |
forgot | -0.103 |
forgotten | -0.100 |
not | -0.100 |
help | -0.098 |
Token | we |
Token ID | 356 |
Feature activation | +6.06e-02 |
Pos logit contributions | |
forgot | +0.051 |
forgotten | +0.050 |
stolen | +0.049 |
wasted | +0.049 |
known | +0.049 |
Neg logit contributions | |
ing | -0.074 |
ane | -0.067 |
ays | -0.065 |
he | -0.065 |
eddy | -0.065 |
Token | must |
Token ID | 1276 |
Feature activation | +6.74e-02 |
Pos logit contributions | |
pap | +0.090 |
stolen | +0.089 |
forgetting | +0.088 |
tresp | +0.087 |
forgotten | +0.087 |
Neg logit contributions | |
he | -0.123 |
ing | -0.123 |
ane | -0.121 |
ite | -0.112 |
ay | -0.107 |
Token | also |
Token ID | 635 |
Feature activation | +9.31e-02 |
Pos logit contributions | |
robbers | +0.061 |
forgot | +0.060 |
forgotten | +0.059 |
knocked | +0.058 |
stolen | +0.058 |
Neg logit contributions | |
he | -0.174 |
ing | -0.154 |
ane | -0.153 |
ay | -0.146 |
room | -0.145 |
Token | demonstrate |
Token ID | 10176 |
Feature activation | +5.42e-03 |
Pos logit contributions | |
robbers | +0.091 |
vitamins | +0.087 |
tresp | +0.084 |
forgot | +0.083 |
setup | +0.082 |
Neg logit contributions | |
he | -0.218 |
ane | -0.191 |
ing | -0.190 |
ite | -0.186 |
ore | -0.186 |
Token | good |
Token ID | 922 |
Feature activation | -8.15e-03 |
Pos logit contributions | |
forgot | +0.008 |
forgotten | +0.007 |
never | +0.007 |
knocked | +0.007 |
known | +0.007 |
Neg logit contributions | |
ing | -0.013 |
ane | -0.012 |
ays | -0.011 |
eddy | -0.011 |
ement | -0.011 |
Token | moral |
Token ID | 6573 |
Feature activation | -2.87e-02 |
Pos logit contributions | |
ing | +0.023 |
ane | +0.020 |
eddy | +0.019 |
ays | +0.019 |
ay | +0.019 |
Neg logit contributions | |
never | -0.009 |
forgot | -0.009 |
forgotten | -0.009 |
knocked | -0.009 |
known | -0.009 |
Token | values |
Token ID | 3815 |
Feature activation | -4.32e-03 |
Pos logit contributions | |
ing | +0.053 |
ane | +0.049 |
ays | +0.048 |
eddy | +0.045 |
he | +0.045 |
Neg logit contributions | |
never | -0.059 |
forgot | -0.057 |
forgotten | -0.056 |
not | -0.055 |
wasted | -0.054 |
Token | and |
Token ID | 290 |
Feature activation | +5.33e-02 |
Pos logit contributions | |
ing | +0.009 |
ane | +0.007 |
ays | +0.007 |
he | +0.007 |
eddy | +0.007 |
Neg logit contributions | |
forgot | -0.008 |
forgotten | -0.008 |
knocked | -0.008 |
stolen | -0.008 |
unknow | -0.008 |
Token | her |
Token ID | 607 |
Feature activation | +1.91e-02 |
Pos logit contributions | |
forgot | +0.040 |
never | +0.038 |
forgotten | +0.037 |
wasted | +0.036 |
stolen | +0.036 |
Neg logit contributions | |
ing | -0.073 |
ane | -0.068 |
ays | -0.066 |
he | -0.065 |
eddy | -0.064 |
Token | friends |
Token ID | 2460 |
Feature activation | -1.91e-02 |
Pos logit contributions | |
never | +0.026 |
forgot | +0.025 |
forgotten | +0.024 |
known | +0.023 |
wasted | +0.023 |
Neg logit contributions | |
ing | -0.049 |
ane | -0.046 |
ays | -0.044 |
eddy | -0.043 |
he | -0.043 |
Token | . |
Token ID | 13 |
Feature activation | -4.02e-02 |
Pos logit contributions | |
ing | +0.034 |
ays | +0.033 |
ane | +0.032 |
eddy | +0.031 |
he | +0.029 |
Neg logit contributions | |
never | -0.041 |
forgot | -0.038 |
help | -0.038 |
not | -0.038 |
forgotten | -0.037 |
Token | She |
Token ID | 1375 |
Feature activation | +6.87e-02 |
Pos logit contributions | |
ing | +0.093 |
ays | +0.084 |
ane | +0.084 |
ooming | +0.078 |
eddy | +0.078 |
Neg logit contributions | |
never | -0.068 |
help | -0.062 |
always | -0.062 |
been | -0.062 |
grown | -0.061 |
Token | would |
Token ID | 561 |
Feature activation | +6.79e-02 |
Pos logit contributions | |
permission | +0.097 |
stolen | +0.096 |
forgotten | +0.090 |
tresp | +0.087 |
robbers | +0.086 |
Neg logit contributions | |
ing | -0.170 |
ane | -0.163 |
ays | -0.152 |
he | -0.149 |
ite | -0.142 |
Token | also |
Token ID | 635 |
Feature activation | +9.29e-02 |
Pos logit contributions | |
stolen | +0.096 |
forgotten | +0.089 |
wasted | +0.085 |
forgot | +0.084 |
tresp | +0.083 |
Neg logit contributions | |
he | -0.159 |
ing | -0.156 |
ane | -0.155 |
ays | -0.137 |
ay | -0.137 |
Token | keep |
Token ID | 1394 |
Feature activation | +2.55e-02 |
Pos logit contributions | |
stolen | +0.124 |
tresp | +0.113 |
forgotten | +0.109 |
wasted | +0.108 |
forgot | +0.103 |
Neg logit contributions | |
he | -0.206 |
ane | -0.204 |
ing | -0.199 |
ite | -0.186 |
ay | -0.183 |
Token | pretend |
Token ID | 16614 |
Feature activation | +6.96e-04 |
Pos logit contributions | |
forgot | +0.035 |
never | +0.033 |
forgotten | +0.032 |
wasted | +0.031 |
stolen | +0.031 |
Neg logit contributions | |
ing | -0.066 |
ane | -0.060 |
ays | -0.060 |
he | -0.058 |
eddy | -0.057 |
Token | dancing |
Token ID | 15360 |
Feature activation | -5.17e-02 |
Pos logit contributions | |
never | +0.001 |
forgot | +0.001 |
forgotten | +0.001 |
wasted | +0.001 |
stolen | +0.001 |
Neg logit contributions | |
ing | -0.001 |
ane | -0.001 |
ays | -0.001 |
eddy | -0.001 |
he | -0.001 |
Token | around |
Token ID | 1088 |
Feature activation | -1.45e-03 |
Pos logit contributions | |
ays | +0.098 |
ing | +0.095 |
ane | +0.094 |
eddy | +0.092 |
he | +0.084 |
Neg logit contributions | |
never | -0.103 |
not | -0.097 |
help | -0.096 |
forgot | -0.095 |
grown | -0.094 |
Token | the |
Token ID | 262 |
Feature activation | +7.37e-03 |
Pos logit contributions | |
ing | +0.004 |
ane | +0.004 |
ays | +0.004 |
eddy | +0.003 |
he | +0.003 |
Neg logit contributions | |
never | -0.002 |
forgot | -0.002 |
not | -0.002 |
help | -0.002 |
grown | -0.002 |
Token | was |
Token ID | 373 |
Feature activation | +6.03e-02 |
Pos logit contributions | |
ing | +0.062 |
ane | +0.057 |
ays | +0.056 |
eddy | +0.054 |
he | +0.054 |
Neg logit contributions | |
never | -0.048 |
forgot | -0.047 |
forgotten | -0.045 |
not | -0.044 |
wasted | -0.044 |
Token | magical |
Token ID | 10883 |
Feature activation | -2.48e-02 |
Pos logit contributions | |
forgot | +0.078 |
robbers | +0.075 |
forget | +0.072 |
fiber | +0.070 |
setup | +0.070 |
Neg logit contributions | |
ing | -0.141 |
he | -0.124 |
ement | -0.121 |
ane | -0.120 |
over | -0.120 |
Token | and |
Token ID | 290 |
Feature activation | +6.78e-02 |
Pos logit contributions | |
ing | +0.040 |
ane | +0.038 |
ays | +0.037 |
he | +0.034 |
eddy | +0.034 |
Neg logit contributions | |
never | -0.056 |
forgot | -0.054 |
forgotten | -0.053 |
not | -0.052 |
help | -0.052 |
Token | hadn |
Token ID | 8020 |
Feature activation | +2.85e-02 |
Pos logit contributions | |
tresp | +0.093 |
permission | +0.092 |
wasted | +0.086 |
robbers | +0.086 |
unknow | +0.082 |
Neg logit contributions | |
ing | -0.151 |
ane | -0.120 |
eddy | -0.119 |
ement | -0.117 |
ed | -0.116 |
Token | 't |
Token ID | 470 |
Feature activation | +6.63e-02 |
Pos logit contributions | |
forgot | +0.046 |
never | +0.044 |
forgotten | +0.043 |
wasted | +0.042 |
wished | +0.042 |
Neg logit contributions | |
ing | -0.063 |
ane | -0.060 |
he | -0.057 |
ays | -0.057 |
eddy | -0.056 |
Token | actually |
Token ID | 1682 |
Feature activation | +9.21e-02 |
Pos logit contributions | |
robbers | +0.062 |
pap | +0.060 |
tresp | +0.057 |
confidential | +0.056 |
protein | +0.056 |
Neg logit contributions | |
ing | -0.154 |
ane | -0.135 |
he | -0.134 |
ed | -0.130 |
ment | -0.130 |
Token | gone |
Token ID | 3750 |
Feature activation | +1.75e-03 |
Pos logit contributions | |
pap | +0.088 |
robbers | +0.088 |
protein | +0.086 |
setup | +0.084 |
vitamins | +0.081 |
Neg logit contributions | |
ane | -0.197 |
ing | -0.196 |
he | -0.192 |
ed | -0.184 |
ite | -0.175 |
Token | anywhere |
Token ID | 6609 |
Feature activation | +2.63e-03 |
Pos logit contributions | |
never | +0.004 |
forgot | +0.004 |
forgotten | +0.003 |
not | +0.003 |
wasted | +0.003 |
Neg logit contributions | |
ing | -0.003 |
ane | -0.003 |
ays | -0.003 |
eddy | -0.003 |
he | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | -2.64e-02 |
Pos logit contributions | |
never | +0.006 |
forgot | +0.006 |
forgotten | +0.006 |
known | +0.006 |
wasted | +0.006 |
Neg logit contributions | |
ing | -0.004 |
ane | -0.003 |
ays | -0.003 |
he | -0.003 |
eddy | -0.003 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -7.56e-02 |
Pos logit contributions | |
ing | +0.048 |
ane | +0.044 |
ays | +0.043 |
eddy | +0.041 |
he | +0.041 |
Neg logit contributions | |
never | -0.055 |
forgot | -0.053 |
forgotten | -0.052 |
not | -0.051 |
wasted | -0.050 |
Token | Once |
Token ID | 7454 |
Feature activation | -3.35e-02 |
Pos logit contributions | |
ing | +0.167 |
ays | +0.161 |
ooming | +0.156 |
ane | +0.156 |
ement | +0.155 |
Neg logit contributions | |
been | -0.114 |
never | -0.112 |
had | -0.103 |
have | -0.100 |
grown | -0.099 |
Token |
|
Token ID | 198 |
Feature activation | -2.66e-02 |
Pos logit contributions | |
ing | +0.067 |
ane | +0.064 |
ays | +0.062 |
ement | +0.057 |
eddy | +0.056 |
Neg logit contributions | |
never | -0.067 |
not | -0.063 |
forgot | -0.061 |
grown | -0.059 |
help | -0.059 |
Token | L |
Token ID | 43 |
Feature activation | -1.87e-02 |
Pos logit contributions | |
ing | +0.050 |
ane | +0.047 |
ays | +0.046 |
eddy | +0.043 |
he | +0.042 |
Neg logit contributions | |
never | -0.055 |
forgot | -0.053 |
forgotten | -0.051 |
not | -0.051 |
wasted | -0.050 |
Token | ily |
Token ID | 813 |
Feature activation | +2.68e-02 |
Pos logit contributions | |
ing | +0.042 |
ane | +0.040 |
ays | +0.039 |
eddy | +0.037 |
ement | +0.037 |
Neg logit contributions | |
never | -0.031 |
forgot | -0.029 |
grown | -0.028 |
been | -0.028 |
wasted | -0.028 |
Token | and |
Token ID | 290 |
Feature activation | +1.48e-02 |
Pos logit contributions | |
permission | +0.045 |
forgotten | +0.044 |
known | +0.044 |
stolen | +0.043 |
help | +0.043 |
Neg logit contributions | |
ing | -0.057 |
ane | -0.052 |
ays | -0.051 |
he | -0.047 |
ement | -0.046 |
Token | Ben |
Token ID | 3932 |
Feature activation | +5.39e-02 |
Pos logit contributions | |
never | +0.021 |
forgot | +0.021 |
forgotten | +0.020 |
stolen | +0.019 |
wasted | +0.019 |
Neg logit contributions | |
ing | -0.037 |
ane | -0.034 |
ays | -0.034 |
he | -0.033 |
eddy | -0.032 |
Token | also |
Token ID | 635 |
Feature activation | +9.21e-02 |
Pos logit contributions | |
supposed | +0.070 |
vitamins | +0.070 |
permission | +0.068 |
forgotten | +0.067 |
stolen | +0.067 |
Neg logit contributions | |
he | -0.122 |
ing | -0.118 |
ane | -0.116 |
ays | -0.105 |
ement | -0.104 |
Token | wake |
Token ID | 7765 |
Feature activation | -2.88e-02 |
Pos logit contributions | |
permission | +0.129 |
supposed | +0.117 |
robbers | +0.116 |
setup | +0.115 |
vitamins | +0.114 |
Neg logit contributions | |
aw | -0.178 |
he | -0.173 |
le | -0.152 |
ed | -0.143 |
ing | -0.143 |
Token | up |
Token ID | 510 |
Feature activation | -1.51e-02 |
Pos logit contributions | |
ane | +0.064 |
ite | +0.061 |
umping | +0.054 |
ays | +0.053 |
ase | +0.053 |
Neg logit contributions | |
not | -0.053 |
never | -0.051 |
grown | -0.047 |
forgot | -0.047 |
be | -0.046 |
Token | and |
Token ID | 290 |
Feature activation | +6.54e-02 |
Pos logit contributions | |
ane | +0.032 |
ays | +0.031 |
ing | +0.030 |
eddy | +0.028 |
ite | +0.028 |
Neg logit contributions | |
not | -0.029 |
never | -0.028 |
help | -0.027 |
grown | -0.026 |
be | -0.026 |
Token | look |
Token ID | 804 |
Feature activation | +2.75e-02 |
Pos logit contributions | |
never | +0.073 |
forgotten | +0.073 |
known | +0.072 |
stolen | +0.072 |
permission | +0.069 |
Neg logit contributions | |
ing | -0.168 |
ays | -0.156 |
ane | -0.155 |
he | -0.153 |
eddy | -0.144 |
Token | out |
Token ID | 503 |
Feature activation | +1.29e-02 |
Pos logit contributions | |
forgot | +0.059 |
never | +0.058 |
forgotten | +0.056 |
not | +0.056 |
grown | +0.056 |
Neg logit contributions | |
ane | -0.048 |
ing | -0.046 |
ays | -0.043 |
eddy | -0.043 |
he | -0.041 |
Token | a |
Token ID | 257 |
Feature activation | +2.23e-02 |
Pos logit contributions | |
stolen | +0.084 |
forgotten | +0.082 |
unknow | +0.080 |
known | +0.078 |
permission | +0.077 |
Neg logit contributions | |
ing | -0.151 |
ays | -0.139 |
ane | -0.137 |
he | -0.136 |
ite | -0.127 |
Token | thrill |
Token ID | 28576 |
Feature activation | -3.03e-02 |
Pos logit contributions | |
never | +0.033 |
permission | +0.031 |
forgot | +0.031 |
known | +0.031 |
forgotten | +0.031 |
Neg logit contributions | |
ing | -0.052 |
ane | -0.047 |
he | -0.047 |
eddy | -0.045 |
ay | -0.045 |
Token | of |
Token ID | 286 |
Feature activation | +3.55e-02 |
Pos logit contributions | |
ing | +0.049 |
ane | +0.048 |
ays | +0.046 |
eddy | +0.045 |
he | +0.043 |
Neg logit contributions | |
never | -0.067 |
forgot | -0.064 |
not | -0.063 |
forgotten | -0.062 |
grown | -0.062 |
Token | joy |
Token ID | 8716 |
Feature activation | -2.06e-02 |
Pos logit contributions | |
forgot | +0.059 |
forgotten | +0.055 |
known | +0.052 |
stolen | +0.051 |
unknow | +0.051 |
Neg logit contributions | |
ing | -0.070 |
he | -0.066 |
ays | -0.064 |
ane | -0.064 |
ement | -0.063 |
Token | . |
Token ID | 13 |
Feature activation | -4.07e-02 |
Pos logit contributions | |
ing | +0.034 |
ane | +0.032 |
ays | +0.031 |
eddy | +0.030 |
he | +0.029 |
Neg logit contributions | |
never | -0.046 |
forgot | -0.045 |
forgotten | -0.043 |
not | -0.043 |
help | -0.043 |
Token | They |
Token ID | 1119 |
Feature activation | +9.13e-02 |
Pos logit contributions | |
ing | +0.087 |
ane | +0.085 |
he | +0.082 |
ement | +0.079 |
ays | +0.079 |
Neg logit contributions | |
never | -0.073 |
not | -0.070 |
grown | -0.069 |
forgot | -0.068 |
known | -0.067 |
Token | think |
Token ID | 892 |
Feature activation | -8.96e-05 |
Pos logit contributions | |
permission | +0.136 |
stolen | +0.130 |
supposed | +0.121 |
forgotten | +0.120 |
unknow | +0.117 |
Neg logit contributions | |
he | -0.213 |
ane | -0.194 |
ite | -0.183 |
ing | -0.174 |
ays | -0.165 |
Token | they |
Token ID | 484 |
Feature activation | +6.16e-02 |
Pos logit contributions | |
ing | +0.000 |
ane | +0.000 |
ays | +0.000 |
eddy | +0.000 |
he | +0.000 |
Neg logit contributions | |
never | -0.000 |
forgot | -0.000 |
not | -0.000 |
help | -0.000 |
forgotten | -0.000 |
Token | are |
Token ID | 389 |
Feature activation | +4.46e-02 |
Pos logit contributions | |
permission | +0.107 |
pap | +0.105 |
setup | +0.098 |
known | +0.098 |
stolen | +0.097 |
Neg logit contributions | |
he | -0.126 |
ane | -0.117 |
ing | -0.116 |
ays | -0.107 |
ite | -0.104 |
Token | flying |
Token ID | 7348 |
Feature activation | -4.58e-02 |
Pos logit contributions | |
unknow | +0.062 |
forgot | +0.060 |
forgotten | +0.057 |
stolen | +0.057 |
permission | +0.056 |
Neg logit contributions | |
ing | -0.099 |
ays | -0.086 |
ane | -0.085 |
eddy | -0.081 |
apping | -0.080 |
Token | . |
Token ID | 13 |
Feature activation | -3.79e-02 |
Pos logit contributions | |
ing | +0.088 |
ane | +0.087 |
ays | +0.086 |
eddy | +0.084 |
he | +0.080 |
Neg logit contributions | |
never | -0.092 |
not | -0.086 |
forgot | -0.084 |
forgotten | -0.081 |
help | -0.080 |
Token | world |
Token ID | 995 |
Feature activation | -9.60e-03 |
Pos logit contributions | |
forgot | +0.041 |
never | +0.039 |
wasted | +0.039 |
forgotten | +0.038 |
knocked | +0.037 |
Neg logit contributions | |
ing | -0.067 |
ane | -0.062 |
eddy | -0.057 |
he | -0.057 |
ay | -0.057 |
Token | might |
Token ID | 1244 |
Feature activation | +5.44e-02 |
Pos logit contributions | |
ing | +0.024 |
ane | +0.022 |
ays | +0.021 |
he | +0.021 |
eddy | +0.020 |
Neg logit contributions | |
never | -0.013 |
forgot | -0.013 |
forgotten | -0.013 |
permission | -0.012 |
wasted | -0.012 |
Token | seem |
Token ID | 1283 |
Feature activation | +2.89e-02 |
Pos logit contributions | |
forgot | +0.051 |
wasted | +0.050 |
forgotten | +0.048 |
pap | +0.048 |
stolen | +0.047 |
Neg logit contributions | |
ing | -0.150 |
ane | -0.140 |
he | -0.138 |
ays | -0.127 |
ement | -0.126 |
Token | , |
Token ID | 11 |
Feature activation | +2.70e-02 |
Pos logit contributions | |
never | +0.063 |
forgot | +0.061 |
forgotten | +0.060 |
wasted | +0.059 |
permission | +0.059 |
Neg logit contributions | |
ing | -0.049 |
ane | -0.042 |
ays | -0.039 |
he | -0.038 |
eddy | -0.037 |
Token | everything |
Token ID | 2279 |
Feature activation | +3.17e-02 |
Pos logit contributions | |
forgot | +0.045 |
wasted | +0.044 |
stolen | +0.044 |
forgotten | +0.043 |
knocked | +0.043 |
Neg logit contributions | |
ing | -0.059 |
ane | -0.051 |
ays | -0.048 |
he | -0.047 |
ement | -0.046 |
Token | would |
Token ID | 561 |
Feature activation | +9.11e-02 |
Pos logit contributions | |
permission | +0.044 |
robbers | +0.044 |
known | +0.043 |
stolen | +0.043 |
forgotten | +0.043 |
Neg logit contributions | |
ing | -0.078 |
ane | -0.070 |
he | -0.067 |
ays | -0.065 |
ement | -0.064 |
Token | work |
Token ID | 670 |
Feature activation | -9.05e-04 |
Pos logit contributions | |
robbers | +0.094 |
wasted | +0.093 |
knocked | +0.088 |
pap | +0.088 |
forgot | +0.087 |
Neg logit contributions | |
he | -0.195 |
ing | -0.185 |
ane | -0.168 |
anymore | -0.165 |
time | -0.165 |
Token | out |
Token ID | 503 |
Feature activation | +1.98e-02 |
Pos logit contributions | |
ing | +0.002 |
ane | +0.002 |
ays | +0.002 |
he | +0.002 |
eddy | +0.002 |
Neg logit contributions | |
never | -0.002 |
forgot | -0.002 |
forgotten | -0.002 |
wasted | -0.002 |
stolen | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | -2.59e-02 |
Pos logit contributions | |
forgot | +0.037 |
never | +0.037 |
forgotten | +0.036 |
wasted | +0.035 |
stolen | +0.035 |
Neg logit contributions | |
ing | -0.039 |
ane | -0.035 |
ays | -0.034 |
he | -0.033 |
eddy | -0.032 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -7.74e-02 |
Pos logit contributions | |
ing | +0.052 |
ays | +0.048 |
ane | +0.047 |
eddy | +0.045 |
he | +0.043 |
Neg logit contributions | |
never | -0.050 |
forgot | -0.046 |
not | -0.046 |
known | -0.045 |
help | -0.045 |
Token | Once |
Token ID | 7454 |
Feature activation | -4.06e-02 |
Pos logit contributions | |
ing | +0.168 |
ays | +0.164 |
ane | +0.159 |
ement | +0.158 |
ooming | +0.158 |
Neg logit contributions | |
been | -0.114 |
never | -0.107 |
had | -0.104 |
have | -0.099 |
grown | -0.098 |
Token |
|
Token ID | 198 |
Feature activation | -4.35e-02 |
Pos logit contributions | |
ing | +0.083 |
ane | +0.077 |
ays | +0.076 |
eddy | +0.074 |
he | +0.070 |
Neg logit contributions | |
never | -0.058 |
not | -0.056 |
grown | -0.055 |
forgot | -0.055 |
help | -0.054 |
Token |
|
Token ID | 198 |
Feature activation | -2.56e-02 |
Pos logit contributions | |
ing | +0.102 |
ays | +0.099 |
ane | +0.096 |
ement | +0.091 |
eddy | +0.090 |
Neg logit contributions | |
never | -0.068 |
not | -0.062 |
known | -0.060 |
know | -0.059 |
grown | -0.059 |
Token | Miss |
Token ID | 17140 |
Feature activation | -3.61e-02 |
Pos logit contributions | |
ing | +0.060 |
ays | +0.057 |
ane | +0.057 |
eddy | +0.054 |
ement | +0.053 |
Neg logit contributions | |
never | -0.041 |
forgot | -0.038 |
forgotten | -0.037 |
not | -0.037 |
grown | -0.036 |
Token | Rachel |
Token ID | 15984 |
Feature activation | +2.46e-02 |
Pos logit contributions | |
he | +0.083 |
ing | +0.083 |
ane | +0.083 |
ays | +0.081 |
ement | +0.078 |
Neg logit contributions | |
forgot | -0.051 |
never | -0.048 |
forgotten | -0.045 |
knocked | -0.045 |
been | -0.044 |
Token | would |
Token ID | 561 |
Feature activation | +6.36e-02 |
Pos logit contributions | |
permission | +0.032 |
forgotten | +0.032 |
stolen | +0.031 |
never | +0.031 |
forgot | +0.031 |
Neg logit contributions | |
ing | -0.061 |
ane | -0.057 |
ays | -0.055 |
he | -0.054 |
ement | -0.052 |
Token | also |
Token ID | 635 |
Feature activation | +9.10e-02 |
Pos logit contributions | |
stolen | +0.078 |
forgot | +0.076 |
forgotten | +0.076 |
wasted | +0.073 |
known | +0.070 |
Neg logit contributions | |
ing | -0.151 |
he | -0.149 |
ane | -0.144 |
ement | -0.130 |
ays | -0.129 |
Token | teach |
Token ID | 4545 |
Feature activation | -3.94e-03 |
Pos logit contributions | |
stolen | +0.104 |
forgotten | +0.101 |
tresp | +0.100 |
forgot | +0.098 |
wasted | +0.098 |
Neg logit contributions | |
ing | -0.212 |
he | -0.208 |
ane | -0.207 |
ite | -0.194 |
ays | -0.189 |
Token | Jack |
Token ID | 3619 |
Feature activation | -9.88e-03 |
Pos logit contributions | |
ing | +0.009 |
ane | +0.009 |
ays | +0.009 |
he | +0.008 |
eddy | +0.008 |
Neg logit contributions | |
never | -0.006 |
forgot | -0.006 |
forgotten | -0.006 |
wasted | -0.005 |
stolen | -0.005 |
Token | maths |
Token ID | 47761 |
Feature activation | -4.03e-02 |
Pos logit contributions | |
ing | +0.025 |
ane | +0.023 |
ays | +0.022 |
eddy | +0.021 |
he | +0.021 |
Neg logit contributions | |
never | -0.014 |
forgot | -0.014 |
forgotten | -0.013 |
stolen | -0.013 |
wasted | -0.013 |
Token | and |
Token ID | 290 |
Feature activation | +6.62e-02 |
Pos logit contributions | |
ing | +0.078 |
ane | +0.074 |
ays | +0.072 |
eddy | +0.070 |
he | +0.068 |
Neg logit contributions | |
never | -0.078 |
forgot | -0.075 |
forgotten | -0.073 |
not | -0.072 |
help | -0.071 |
Token | stories |
Token ID | 3923 |
Feature activation | -3.87e-02 |
Pos logit contributions | |
forgotten | +0.096 |
stolen | +0.095 |
wasted | +0.093 |
tresp | +0.090 |
permission | +0.089 |
Neg logit contributions | |
ing | -0.152 |
ane | -0.127 |
ays | -0.122 |
he | -0.117 |
ay | -0.115 |
Token | river |
Token ID | 7850 |
Feature activation | -5.48e-02 |
Pos logit contributions | |
never | +0.011 |
forgot | +0.011 |
forgotten | +0.010 |
not | +0.010 |
wasted | +0.010 |
Neg logit contributions | |
ing | -0.016 |
ane | -0.015 |
ays | -0.014 |
he | -0.014 |
eddy | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -3.09e-02 |
Pos logit contributions | |
ing | +0.087 |
ane | +0.082 |
ays | +0.080 |
eddy | +0.076 |
he | +0.073 |
Neg logit contributions | |
never | -0.125 |
forgot | -0.120 |
forgotten | -0.118 |
not | -0.117 |
help | -0.115 |
Token | The |
Token ID | 383 |
Feature activation | +1.27e-02 |
Pos logit contributions | |
ing | +0.061 |
ane | +0.056 |
ays | +0.054 |
eddy | +0.052 |
he | +0.052 |
Neg logit contributions | |
never | -0.060 |
forgot | -0.058 |
forgotten | -0.056 |
not | -0.056 |
wasted | -0.055 |
Token | boy |
Token ID | 2933 |
Feature activation | +2.35e-02 |
Pos logit contributions | |
never | +0.025 |
forgot | +0.024 |
forgotten | +0.023 |
known | +0.023 |
wasted | +0.023 |
Neg logit contributions | |
ing | -0.025 |
ane | -0.023 |
he | -0.021 |
eddy | -0.021 |
ays | -0.021 |
Token | could |
Token ID | 714 |
Feature activation | +5.74e-02 |
Pos logit contributions | |
permission | +0.034 |
forgotten | +0.033 |
stolen | +0.032 |
help | +0.032 |
known | +0.032 |
Neg logit contributions | |
ing | -0.056 |
ane | -0.051 |
ays | -0.049 |
he | -0.049 |
eddy | -0.048 |
Token | also |
Token ID | 635 |
Feature activation | +9.07e-02 |
Pos logit contributions | |
stolen | +0.055 |
forgotten | +0.054 |
permission | +0.054 |
forgot | +0.053 |
setup | +0.051 |
Neg logit contributions | |
he | -0.157 |
ane | -0.153 |
ing | -0.153 |
ays | -0.144 |
eddy | -0.138 |
Token | see |
Token ID | 766 |
Feature activation | +1.65e-02 |
Pos logit contributions | |
forgot | +0.097 |
stolen | +0.096 |
forgotten | +0.095 |
setup | +0.095 |
wasted | +0.094 |
Neg logit contributions | |
he | -0.216 |
ane | -0.208 |
ing | -0.201 |
ite | -0.195 |
ays | -0.185 |
Token | flowers |
Token ID | 12734 |
Feature activation | -2.00e-02 |
Pos logit contributions | |
forgot | +0.022 |
never | +0.022 |
forgotten | +0.021 |
wasted | +0.020 |
stolen | +0.020 |
Neg logit contributions | |
ing | -0.043 |
ane | -0.039 |
ays | -0.039 |
eddy | -0.037 |
he | -0.037 |
Token | growing |
Token ID | 3957 |
Feature activation | -2.43e-02 |
Pos logit contributions | |
ing | +0.036 |
ane | +0.034 |
ays | +0.034 |
eddy | +0.033 |
he | +0.031 |
Neg logit contributions | |
never | -0.041 |
forgot | -0.039 |
not | -0.038 |
forgotten | -0.038 |
help | -0.038 |
Token | on |
Token ID | 319 |
Feature activation | +1.23e-02 |
Pos logit contributions | |
ane | +0.045 |
ing | +0.043 |
ays | +0.042 |
eddy | +0.041 |
he | +0.040 |
Neg logit contributions | |
never | -0.048 |
not | -0.045 |
forgotten | -0.045 |
grown | -0.044 |
forgot | -0.044 |
Token | the |
Token ID | 262 |
Feature activation | +9.29e-04 |
Pos logit contributions | |
never | +0.015 |
forgot | +0.015 |
forgotten | +0.014 |
wasted | +0.013 |
not | +0.013 |
Neg logit contributions | |
ing | -0.033 |
ane | -0.031 |
ays | -0.030 |
eddy | -0.029 |
he | -0.029 |
Token | worried |
Token ID | 7960 |
Feature activation | -9.97e-03 |
Pos logit contributions | |
never | +0.034 |
forgot | +0.033 |
forgotten | +0.031 |
known | +0.030 |
stolen | +0.030 |
Neg logit contributions | |
ing | -0.076 |
ane | -0.070 |
ays | -0.069 |
he | -0.067 |
eddy | -0.066 |
Token | she |
Token ID | 673 |
Feature activation | +5.70e-02 |
Pos logit contributions | |
ing | +0.024 |
ane | +0.022 |
ays | +0.021 |
he | +0.021 |
eddy | +0.021 |
Neg logit contributions | |
never | -0.015 |
forgot | -0.015 |
forgotten | -0.014 |
known | -0.014 |
wasted | -0.014 |
Token | was |
Token ID | 373 |
Feature activation | +4.15e-02 |
Pos logit contributions | |
never | +0.068 |
permission | +0.068 |
known | +0.067 |
forgotten | +0.066 |
stolen | +0.065 |
Neg logit contributions | |
ing | -0.147 |
ane | -0.136 |
he | -0.131 |
ays | -0.130 |
ement | -0.125 |
Token | , |
Token ID | 11 |
Feature activation | +2.13e-02 |
Pos logit contributions | |
forgot | +0.084 |
never | +0.083 |
forgotten | +0.080 |
wasted | +0.080 |
known | +0.079 |
Neg logit contributions | |
ing | -0.075 |
ane | -0.065 |
he | -0.061 |
ays | -0.061 |
eddy | -0.058 |
Token | things |
Token ID | 1243 |
Feature activation | +2.54e-02 |
Pos logit contributions | |
forgot | +0.037 |
forgotten | +0.036 |
wasted | +0.036 |
stolen | +0.036 |
never | +0.035 |
Neg logit contributions | |
ing | -0.044 |
ane | -0.038 |
he | -0.036 |
ays | -0.036 |
eddy | -0.035 |
Token | would |
Token ID | 561 |
Feature activation | +9.05e-02 |
Pos logit contributions | |
known | +0.032 |
permission | +0.032 |
unknow | +0.031 |
forgotten | +0.031 |
pap | +0.031 |
Neg logit contributions | |
ing | -0.065 |
ane | -0.059 |
he | -0.059 |
ays | -0.054 |
ite | -0.052 |
Token | turn |
Token ID | 1210 |
Feature activation | -6.90e-03 |
Pos logit contributions | |
robbers | +0.089 |
pap | +0.088 |
unknow | +0.087 |
forgot | +0.082 |
setup | +0.082 |
Neg logit contributions | |
he | -0.204 |
ing | -0.185 |
anymore | -0.169 |
ane | -0.169 |
time | -0.167 |
Token | out |
Token ID | 503 |
Feature activation | +8.77e-03 |
Pos logit contributions | |
ing | +0.013 |
ane | +0.013 |
ays | +0.012 |
eddy | +0.012 |
he | +0.012 |
Neg logit contributions | |
never | -0.013 |
forgot | -0.013 |
forgotten | -0.013 |
not | -0.012 |
wasted | -0.012 |
Token | okay |
Token ID | 8788 |
Feature activation | -1.63e-02 |
Pos logit contributions | |
never | +0.017 |
forgot | +0.017 |
forgotten | +0.016 |
not | +0.016 |
wasted | +0.016 |
Neg logit contributions | |
ing | -0.017 |
ane | -0.016 |
ays | -0.015 |
eddy | -0.015 |
he | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | -2.70e-02 |
Pos logit contributions | |
ing | +0.029 |
ane | +0.028 |
ays | +0.028 |
he | +0.026 |
eddy | +0.026 |
Neg logit contributions | |
never | -0.034 |
forgot | -0.033 |
not | -0.032 |
forgotten | -0.032 |
been | -0.031 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -7.77e-02 |
Pos logit contributions | |
ing | +0.056 |
ane | +0.052 |
ays | +0.050 |
eddy | +0.048 |
he | +0.047 |
Neg logit contributions | |
never | -0.051 |
forgot | -0.048 |
not | -0.047 |
forgotten | -0.046 |
been | -0.046 |
Token | forget |
Token ID | 6044 |
Feature activation | +2.26e-04 |
Pos logit contributions | |
permission | +0.107 |
unknow | +0.098 |
pap | +0.097 |
warning | +0.090 |
stolen | +0.090 |
Neg logit contributions | |
he | -0.179 |
ane | -0.159 |
ing | -0.154 |
ite | -0.142 |
ays | -0.140 |
Token | the |
Token ID | 262 |
Feature activation | +2.19e-02 |
Pos logit contributions | |
forgot | +0.000 |
never | +0.000 |
forgotten | +0.000 |
known | +0.000 |
stolen | +0.000 |
Neg logit contributions | |
ing | -0.000 |
ane | -0.000 |
ays | -0.000 |
eddy | -0.000 |
he | -0.000 |
Token | time |
Token ID | 640 |
Feature activation | -3.12e-02 |
Pos logit contributions | |
forgot | +0.035 |
never | +0.033 |
unknow | +0.032 |
known | +0.031 |
forgotten | +0.031 |
Neg logit contributions | |
ing | -0.047 |
ane | -0.046 |
he | -0.044 |
eddy | -0.042 |
ay | -0.042 |
Token | . |
Token ID | 13 |
Feature activation | -2.40e-02 |
Pos logit contributions | |
ing | +0.053 |
ane | +0.051 |
ays | +0.050 |
eddy | +0.048 |
he | +0.048 |
Neg logit contributions | |
never | -0.066 |
forgot | -0.063 |
not | -0.062 |
forgotten | -0.062 |
wasted | -0.061 |
Token | They |
Token ID | 1119 |
Feature activation | +8.48e-02 |
Pos logit contributions | |
ing | +0.056 |
ane | +0.051 |
he | +0.048 |
ays | +0.048 |
eddy | +0.047 |
Neg logit contributions | |
never | -0.042 |
forgot | -0.039 |
not | -0.039 |
forgotten | -0.038 |
wasted | -0.038 |
Token | also |
Token ID | 635 |
Feature activation | +9.05e-02 |
Pos logit contributions | |
permission | +0.117 |
stolen | +0.110 |
unknow | +0.107 |
supposed | +0.105 |
forgotten | +0.103 |
Neg logit contributions | |
he | -0.207 |
ane | -0.184 |
ite | -0.167 |
ing | -0.163 |
ays | -0.154 |
Token | forget |
Token ID | 6044 |
Feature activation | -7.87e-05 |
Pos logit contributions | |
permission | +0.110 |
unknow | +0.104 |
stolen | +0.100 |
supposed | +0.097 |
forgotten | +0.097 |
Neg logit contributions | |
he | -0.205 |
ane | -0.183 |
ite | -0.168 |
ing | -0.166 |
ays | -0.166 |
Token | to |
Token ID | 284 |
Feature activation | +7.51e-02 |
Pos logit contributions | |
ing | +0.000 |
ane | +0.000 |
ays | +0.000 |
eddy | +0.000 |
he | +0.000 |
Neg logit contributions | |
forgot | -0.000 |
never | -0.000 |
forgotten | -0.000 |
known | -0.000 |
stolen | -0.000 |
Token | eat |
Token ID | 4483 |
Feature activation | -2.74e-02 |
Pos logit contributions | |
forgot | +0.094 |
unknow | +0.087 |
wished | +0.087 |
forgotten | +0.085 |
wasted | +0.084 |
Neg logit contributions | |
he | -0.175 |
room | -0.151 |
ing | -0.148 |
over | -0.147 |
ane | -0.146 |
Token | . |
Token ID | 13 |
Feature activation | -2.66e-02 |
Pos logit contributions | |
ing | +0.076 |
ane | +0.075 |
ays | +0.071 |
eddy | +0.070 |
he | +0.068 |
Neg logit contributions | |
never | -0.031 |
not | -0.029 |
help | -0.028 |
grown | -0.028 |
forgotten | -0.027 |
Token | They |
Token ID | 1119 |
Feature activation | +7.92e-02 |
Pos logit contributions | |
ing | +0.065 |
ane | +0.059 |
he | +0.056 |
ays | +0.055 |
ement | +0.054 |
Neg logit contributions | |
never | -0.043 |
not | -0.041 |
forgot | -0.040 |
grown | -0.040 |
forgotten | -0.039 |
Token | of |
Token ID | 286 |
Feature activation | +9.86e-03 |
Pos logit contributions | |
ing | +0.012 |
ane | +0.010 |
ays | +0.010 |
he | +0.009 |
eddy | +0.009 |
Neg logit contributions | |
never | -0.025 |
forgot | -0.025 |
forgotten | -0.024 |
not | -0.024 |
wasted | -0.024 |
Token | luck |
Token ID | 8458 |
Feature activation | -5.95e-02 |
Pos logit contributions | |
forgot | +0.015 |
setup | +0.014 |
forget | +0.013 |
never | +0.013 |
unknow | +0.013 |
Neg logit contributions | |
ing | -0.023 |
he | -0.020 |
ays | -0.020 |
eddy | -0.019 |
ement | -0.019 |
Token | and |
Token ID | 290 |
Feature activation | +6.34e-02 |
Pos logit contributions | |
ing | +0.101 |
ane | +0.094 |
ays | +0.091 |
eddy | +0.087 |
he | +0.085 |
Neg logit contributions | |
never | -0.130 |
forgot | -0.127 |
forgotten | -0.123 |
not | -0.121 |
wasted | -0.119 |
Token | good |
Token ID | 922 |
Feature activation | -6.01e-03 |
Pos logit contributions | |
wasted | +0.072 |
permission | +0.072 |
forgotten | +0.071 |
knocked | +0.071 |
tresp | +0.070 |
Neg logit contributions | |
ing | -0.155 |
he | -0.137 |
ane | -0.132 |
ays | -0.127 |
ed | -0.126 |
Token | things |
Token ID | 1243 |
Feature activation | -1.05e-03 |
Pos logit contributions | |
ing | +0.014 |
ane | +0.013 |
ays | +0.013 |
eddy | +0.012 |
he | +0.012 |
Neg logit contributions | |
never | -0.009 |
forgot | -0.009 |
forgotten | -0.009 |
wasted | -0.009 |
not | -0.009 |
Token | would |
Token ID | 561 |
Feature activation | +9.05e-02 |
Pos logit contributions | |
ing | +0.002 |
ane | +0.002 |
ays | +0.002 |
he | +0.002 |
eddy | +0.002 |
Neg logit contributions | |
never | -0.002 |
forgot | -0.002 |
forgotten | -0.002 |
known | -0.002 |
wasted | -0.002 |
Token | happen |
Token ID | 1645 |
Feature activation | -2.30e-02 |
Pos logit contributions | |
pap | +0.092 |
knocked | +0.085 |
unknow | +0.083 |
setup | +0.083 |
wasted | +0.078 |
Neg logit contributions | |
he | -0.219 |
ing | -0.200 |
ane | -0.191 |
ay | -0.182 |
anymore | -0.180 |
Token | to |
Token ID | 284 |
Feature activation | +3.01e-02 |
Pos logit contributions | |
ing | +0.043 |
ane | +0.041 |
ays | +0.040 |
eddy | +0.038 |
he | +0.038 |
Neg logit contributions | |
never | -0.047 |
forgot | -0.045 |
not | -0.043 |
forgotten | -0.043 |
help | -0.042 |
Token | her |
Token ID | 607 |
Feature activation | +2.93e-02 |
Pos logit contributions | |
forgot | +0.047 |
forgotten | +0.045 |
stolen | +0.044 |
knocked | +0.044 |
wasted | +0.043 |
Neg logit contributions | |
ing | -0.068 |
he | -0.060 |
ane | -0.059 |
eddy | -0.057 |
ays | -0.057 |
Token | . |
Token ID | 13 |
Feature activation | -2.41e-02 |
Pos logit contributions | |
never | +0.047 |
forgot | +0.046 |
forgotten | +0.044 |
not | +0.043 |
wasted | +0.043 |
Neg logit contributions | |
ing | -0.067 |
ane | -0.062 |
ays | -0.061 |
eddy | -0.059 |
he | -0.059 |
Token |
|
Token ID | 198 |
Feature activation | -3.12e-02 |
Pos logit contributions | |
ing | +0.050 |
ane | +0.045 |
ays | +0.044 |
eddy | +0.042 |
he | +0.041 |
Neg logit contributions | |
never | -0.046 |
forgot | -0.043 |
not | -0.042 |
forgotten | -0.042 |
been | -0.041 |
Token | felt |
Token ID | 2936 |
Feature activation | +5.40e-02 |
Pos logit contributions | |
permission | +0.108 |
vitamins | +0.106 |
supposed | +0.106 |
pap | +0.103 |
setup | +0.098 |
Neg logit contributions | |
he | -0.179 |
ing | -0.175 |
ane | -0.167 |
aw | -0.162 |
ed | -0.157 |
Token | proud |
Token ID | 6613 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
forgot | +0.073 |
never | +0.073 |
setup | +0.071 |
forgotten | +0.069 |
unknow | +0.068 |
Neg logit contributions | |
ing | -0.123 |
he | -0.112 |
ane | -0.109 |
ays | -0.108 |
ement | -0.106 |
Token | and |
Token ID | 290 |
Feature activation | +5.84e-02 |
Pos logit contributions | |
ing | +0.017 |
ane | +0.016 |
ays | +0.015 |
he | +0.015 |
eddy | +0.015 |
Neg logit contributions | |
never | -0.025 |
forgot | -0.024 |
forgotten | -0.024 |
not | -0.024 |
help | -0.023 |
Token | happy |
Token ID | 3772 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
tresp | +0.082 |
pap | +0.082 |
permission | +0.081 |
setup | +0.080 |
robbers | +0.079 |
Neg logit contributions | |
ing | -0.120 |
ed | -0.108 |
he | -0.106 |
ane | -0.102 |
over | -0.100 |
Token | . |
Token ID | 13 |
Feature activation | -3.18e-02 |
Pos logit contributions | |
ing | +0.017 |
ane | +0.016 |
ays | +0.015 |
he | +0.014 |
eddy | +0.014 |
Neg logit contributions | |
never | -0.023 |
forgot | -0.023 |
not | -0.022 |
forgotten | -0.022 |
help | -0.022 |
Token | They |
Token ID | 1119 |
Feature activation | +9.04e-02 |
Pos logit contributions | |
ing | +0.071 |
ane | +0.066 |
he | +0.063 |
ays | +0.063 |
eddy | +0.061 |
Neg logit contributions | |
never | -0.056 |
forgot | -0.053 |
not | -0.052 |
forgotten | -0.051 |
grown | -0.051 |
Token | had |
Token ID | 550 |
Feature activation | +5.10e-02 |
Pos logit contributions | |
permission | +0.146 |
supervision | +0.130 |
pap | +0.124 |
warning | +0.123 |
vitamins | +0.122 |
Neg logit contributions | |
he | -0.196 |
ane | -0.171 |
aw | -0.169 |
ing | -0.166 |
ite | -0.159 |
Token | discovered |
Token ID | 5071 |
Feature activation | +9.16e-03 |
Pos logit contributions | |
pap | +0.057 |
permission | +0.053 |
help | +0.052 |
warning | +0.050 |
weed | +0.049 |
Neg logit contributions | |
aw | -0.110 |
ing | -0.103 |
ed | -0.100 |
ane | -0.095 |
room | -0.093 |
Token | a |
Token ID | 257 |
Feature activation | +3.68e-02 |
Pos logit contributions | |
forgot | +0.012 |
never | +0.012 |
forgotten | +0.011 |
wasted | +0.011 |
stolen | +0.011 |
Neg logit contributions | |
ing | -0.023 |
ane | -0.022 |
ays | -0.021 |
eddy | -0.020 |
he | -0.020 |
Token | new |
Token ID | 649 |
Feature activation | +5.33e-03 |
Pos logit contributions | |
permission | +0.056 |
forgot | +0.053 |
wasted | +0.052 |
never | +0.052 |
unknow | +0.051 |
Neg logit contributions | |
ing | -0.082 |
ane | -0.077 |
eddy | -0.073 |
ay | -0.071 |
he | -0.070 |
Token | place |
Token ID | 1295 |
Feature activation | -3.00e-02 |
Pos logit contributions | |
forgot | +0.008 |
never | +0.008 |
forgotten | +0.008 |
wasted | +0.008 |
permission | +0.007 |
Neg logit contributions | |
ing | -0.012 |
ane | -0.012 |
ays | -0.011 |
eddy | -0.011 |
he | -0.011 |
Token | helpful |
Token ID | 7613 |
Feature activation | -3.58e-02 |
Pos logit contributions | |
fact | +0.071 |
known | +0.070 |
been | +0.069 |
permission | +0.068 |
tresp | +0.066 |
Neg logit contributions | |
ing | -0.156 |
ite | -0.126 |
apping | -0.125 |
ane | -0.125 |
he | -0.122 |
Token | , |
Token ID | 11 |
Feature activation | +3.28e-02 |
Pos logit contributions | |
ing | +0.059 |
ane | +0.054 |
ays | +0.052 |
eddy | +0.049 |
he | +0.049 |
Neg logit contributions | |
never | -0.080 |
forgot | -0.079 |
forgotten | -0.076 |
not | -0.075 |
wasted | -0.075 |
Token | but |
Token ID | 475 |
Feature activation | +4.71e-02 |
Pos logit contributions | |
forgot | +0.061 |
stolen | +0.059 |
knocked | +0.059 |
forgotten | +0.059 |
wasted | +0.059 |
Neg logit contributions | |
ing | -0.060 |
ane | -0.051 |
eddy | -0.048 |
ays | -0.048 |
ay | -0.047 |
Token | they |
Token ID | 484 |
Feature activation | +7.17e-02 |
Pos logit contributions | |
forgot | +0.075 |
forgotten | +0.071 |
stolen | +0.069 |
wasted | +0.068 |
unknow | +0.068 |
Neg logit contributions | |
ing | -0.107 |
ane | -0.092 |
ite | -0.085 |
ore | -0.084 |
eddy | -0.083 |
Token | can |
Token ID | 460 |
Feature activation | +5.93e-02 |
Pos logit contributions | |
forgotten | +0.101 |
stolen | +0.100 |
robbers | +0.098 |
pap | +0.096 |
tresp | +0.095 |
Neg logit contributions | |
ing | -0.150 |
he | -0.150 |
ane | -0.147 |
ite | -0.137 |
ore | -0.126 |
Token | also |
Token ID | 635 |
Feature activation | +9.03e-02 |
Pos logit contributions | |
stolen | +0.058 |
forgotten | +0.057 |
forgot | +0.056 |
permission | +0.053 |
wasted | +0.052 |
Neg logit contributions | |
ing | -0.162 |
he | -0.159 |
ane | -0.156 |
ite | -0.143 |
eddy | -0.143 |
Token | be |
Token ID | 307 |
Feature activation | +4.05e-02 |
Pos logit contributions | |
forgot | +0.091 |
vitamins | +0.091 |
stolen | +0.089 |
tresp | +0.088 |
forgotten | +0.087 |
Neg logit contributions | |
he | -0.203 |
ing | -0.192 |
ane | -0.191 |
ite | -0.187 |
ore | -0.183 |
Token | very |
Token ID | 845 |
Feature activation | +5.86e-02 |
Pos logit contributions | |
forgot | +0.046 |
wait | +0.046 |
cost | +0.045 |
setup | +0.045 |
fact | +0.044 |
Neg logit contributions | |
ing | -0.099 |
over | -0.084 |
ays | -0.079 |
ane | -0.078 |
apping | -0.078 |
Token | dangerous |
Token ID | 4923 |
Feature activation | -4.94e-02 |
Pos logit contributions | |
fact | +0.067 |
permission | +0.066 |
been | +0.066 |
known | +0.065 |
tresp | +0.064 |
Neg logit contributions | |
ing | -0.149 |
apping | -0.119 |
ane | -0.119 |
ite | -0.118 |
he | -0.117 |
Token | . |
Token ID | 13 |
Feature activation | -2.58e-02 |
Pos logit contributions | |
ane | +0.082 |
ing | +0.081 |
ays | +0.079 |
he | +0.075 |
eddy | +0.074 |
Neg logit contributions | |
never | -0.109 |
forgot | -0.101 |
not | -0.101 |
forgotten | -0.101 |
help | -0.101 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -8.02e-02 |
Pos logit contributions | |
ing | +0.046 |
ane | +0.042 |
ays | +0.041 |
eddy | +0.038 |
he | +0.038 |
Neg logit contributions | |
never | -0.056 |
forgot | -0.053 |
forgotten | -0.052 |
not | -0.052 |
known | -0.051 |
Token | not |
Token ID | 407 |
Feature activation | +5.92e-02 |
Pos logit contributions | |
unknow | +0.068 |
forgot | +0.065 |
forgotten | +0.060 |
fact | +0.060 |
pap | +0.059 |
Neg logit contributions | |
ing | -0.111 |
on | -0.083 |
ane | -0.081 |
ays | -0.080 |
ying | -0.078 |
Token | enough |
Token ID | 1576 |
Feature activation | +1.80e-02 |
Pos logit contributions | |
unknow | +0.082 |
forgot | +0.080 |
never | +0.078 |
barely | +0.077 |
fact | +0.076 |
Neg logit contributions | |
ing | -0.137 |
ane | -0.105 |
eddy | -0.104 |
ays | -0.099 |
apping | -0.097 |
Token | . |
Token ID | 13 |
Feature activation | -3.23e-02 |
Pos logit contributions | |
never | +0.040 |
forgot | +0.040 |
known | +0.038 |
forgotten | +0.038 |
unknow | +0.037 |
Neg logit contributions | |
ing | -0.030 |
ane | -0.025 |
ays | -0.024 |
eddy | -0.023 |
he | -0.023 |
Token | We |
Token ID | 775 |
Feature activation | +5.13e-02 |
Pos logit contributions | |
ing | +0.063 |
ane | +0.058 |
ays | +0.056 |
he | +0.054 |
eddy | +0.053 |
Neg logit contributions | |
never | -0.064 |
forgot | -0.062 |
forgotten | -0.060 |
not | -0.060 |
wasted | -0.059 |
Token | must |
Token ID | 1276 |
Feature activation | +6.17e-02 |
Pos logit contributions | |
pap | +0.068 |
grown | +0.068 |
forgetting | +0.068 |
setup | +0.066 |
permission | +0.066 |
Neg logit contributions | |
he | -0.119 |
ane | -0.117 |
ing | -0.113 |
ite | -0.110 |
ay | -0.105 |
Token | also |
Token ID | 635 |
Feature activation | +9.02e-02 |
Pos logit contributions | |
forgotten | +0.059 |
forgot | +0.057 |
pap | +0.057 |
wasted | +0.055 |
knocked | +0.055 |
Neg logit contributions | |
he | -0.156 |
room | -0.141 |
ing | -0.141 |
ane | -0.140 |
ay | -0.136 |
Token | be |
Token ID | 307 |
Feature activation | +3.66e-02 |
Pos logit contributions | |
vitamins | +0.095 |
setup | +0.091 |
forgotten | +0.089 |
protein | +0.087 |
supposed | +0.086 |
Neg logit contributions | |
he | -0.197 |
ite | -0.187 |
ane | -0.184 |
ing | -0.178 |
room | -0.174 |
Token | careful |
Token ID | 8161 |
Feature activation | -3.76e-02 |
Pos logit contributions | |
cost | +0.048 |
hardly | +0.045 |
barely | +0.045 |
fact | +0.044 |
forgot | +0.043 |
Neg logit contributions | |
ing | -0.087 |
over | -0.074 |
aw | -0.072 |
apping | -0.072 |
ays | -0.069 |
Token | and |
Token ID | 290 |
Feature activation | +6.91e-02 |
Pos logit contributions | |
ing | +0.069 |
ane | +0.066 |
ays | +0.066 |
eddy | +0.062 |
he | +0.062 |
Neg logit contributions | |
never | -0.079 |
not | -0.074 |
forgot | -0.073 |
forgotten | -0.071 |
help | -0.071 |
Token | watch |
Token ID | 2342 |
Feature activation | -1.31e-02 |
Pos logit contributions | |
fact | +0.063 |
unknow | +0.059 |
forgotten | +0.058 |
pap | +0.058 |
tresp | +0.057 |
Neg logit contributions | |
ing | -0.181 |
ane | -0.152 |
ite | -0.150 |
he | -0.148 |
apping | -0.146 |
Token | out |
Token ID | 503 |
Feature activation | -5.61e-03 |
Pos logit contributions | |
ing | +0.034 |
ane | +0.033 |
ays | +0.032 |
eddy | +0.031 |
he | +0.031 |
Neg logit contributions | |
never | -0.017 |
not | -0.015 |
forgot | -0.015 |
help | -0.015 |
forgotten | -0.014 |
Token | hug |
Token ID | 16225 |
Feature activation | -2.06e-02 |
Pos logit contributions | |
stolen | +0.079 |
forgotten | +0.078 |
wasted | +0.073 |
unknow | +0.073 |
tricked | +0.070 |
Neg logit contributions | |
ing | -0.155 |
he | -0.147 |
ane | -0.137 |
ay | -0.135 |
over | -0.134 |
Token | her |
Token ID | 607 |
Feature activation | -1.99e-02 |
Pos logit contributions | |
ane | +0.050 |
ays | +0.048 |
ing | +0.048 |
ite | +0.045 |
he | +0.045 |
Neg logit contributions | |
never | -0.032 |
not | -0.029 |
help | -0.029 |
forgot | -0.029 |
grown | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | -3.89e-02 |
Pos logit contributions | |
ane | +0.040 |
ays | +0.040 |
ing | +0.039 |
he | +0.036 |
eddy | +0.036 |
Neg logit contributions | |
not | -0.037 |
never | -0.037 |
help | -0.036 |
permission | -0.035 |
forgot | -0.034 |
Token | He |
Token ID | 679 |
Feature activation | +6.66e-02 |
Pos logit contributions | |
ing | +0.089 |
ane | +0.085 |
ays | +0.079 |
he | +0.075 |
eddy | +0.074 |
Neg logit contributions | |
never | -0.064 |
forgot | -0.063 |
grown | -0.062 |
not | -0.061 |
forgotten | -0.060 |
Token | would |
Token ID | 561 |
Feature activation | +7.39e-02 |
Pos logit contributions | |
permission | +0.098 |
stolen | +0.085 |
ruining | +0.082 |
forgotten | +0.081 |
unknow | +0.080 |
Neg logit contributions | |
ing | -0.149 |
ane | -0.144 |
he | -0.136 |
eddy | -0.132 |
ays | -0.132 |
Token | also |
Token ID | 635 |
Feature activation | +9.02e-02 |
Pos logit contributions | |
stolen | +0.090 |
forgotten | +0.086 |
forgot | +0.085 |
unknow | +0.083 |
wasted | +0.083 |
Neg logit contributions | |
he | -0.173 |
ane | -0.157 |
ing | -0.149 |
ay | -0.147 |
room | -0.146 |
Token | talk |
Token ID | 1561 |
Feature activation | -2.08e-02 |
Pos logit contributions | |
stolen | +0.099 |
forgotten | +0.095 |
forgot | +0.095 |
vitamins | +0.094 |
unknow | +0.092 |
Neg logit contributions | |
he | -0.206 |
ane | -0.193 |
ing | -0.185 |
ite | -0.183 |
ay | -0.180 |
Token | to |
Token ID | 284 |
Feature activation | +1.45e-02 |
Pos logit contributions | |
ays | +0.046 |
ing | +0.046 |
ane | +0.045 |
eddy | +0.043 |
he | +0.041 |
Neg logit contributions | |
never | -0.036 |
not | -0.033 |
forgot | -0.032 |
grown | -0.032 |
help | -0.032 |
Token | her |
Token ID | 607 |
Feature activation | +1.76e-03 |
Pos logit contributions | |
forgot | +0.025 |
unknow | +0.025 |
never | +0.024 |
forgotten | +0.024 |
stolen | +0.024 |
Neg logit contributions | |
ing | -0.029 |
he | -0.026 |
eddy | -0.025 |
ays | -0.025 |
ane | -0.025 |
Token | , |
Token ID | 11 |
Feature activation | +4.04e-02 |
Pos logit contributions | |
never | +0.004 |
forgot | +0.004 |
not | +0.004 |
help | +0.003 |
permission | +0.003 |
Neg logit contributions | |
ing | -0.003 |
ays | -0.003 |
ane | -0.003 |
eddy | -0.003 |
he | -0.003 |
Token | but |
Token ID | 475 |
Feature activation | +4.17e-02 |
Pos logit contributions | |
forgot | +0.072 |
never | +0.071 |
forgotten | +0.070 |
unknow | +0.070 |
permission | +0.069 |
Neg logit contributions | |
ing | -0.075 |
ane | -0.068 |
he | -0.065 |
ays | -0.065 |
eddy | -0.065 |
Token | , |
Token ID | 11 |
Feature activation | -2.98e-02 |
Pos logit contributions | |
ays | +0.097 |
ane | +0.097 |
eddy | +0.093 |
ing | +0.092 |
ement | +0.092 |
Neg logit contributions | |
forgot | -0.088 |
never | -0.087 |
grown | -0.087 |
help | -0.086 |
not | -0.086 |
Token | she |
Token ID | 673 |
Feature activation | +4.26e-02 |
Pos logit contributions | |
ing | +0.073 |
ane | +0.071 |
ays | +0.069 |
eddy | +0.066 |
ement | +0.065 |
Neg logit contributions | |
never | -0.044 |
forgot | -0.043 |
not | -0.041 |
help | -0.040 |
forgotten | -0.040 |
Token | saw |
Token ID | 2497 |
Feature activation | -8.95e-04 |
Pos logit contributions | |
permission | +0.049 |
never | +0.047 |
known | +0.045 |
help | +0.045 |
forgotten | +0.045 |
Neg logit contributions | |
ing | -0.113 |
ays | -0.104 |
ane | -0.104 |
eddy | -0.103 |
he | -0.102 |
Token | a |
Token ID | 257 |
Feature activation | +9.31e-03 |
Pos logit contributions | |
ing | +0.002 |
ane | +0.002 |
ays | +0.002 |
eddy | +0.002 |
he | +0.002 |
Neg logit contributions | |
forgot | -0.001 |
never | -0.001 |
forgotten | -0.001 |
wasted | -0.001 |
stolen | -0.001 |
Token | big |
Token ID | 1263 |
Feature activation | -2.01e-02 |
Pos logit contributions | |
never | +0.012 |
forgot | +0.012 |
permission | +0.011 |
forgotten | +0.011 |
known | +0.011 |
Neg logit contributions | |
ing | -0.024 |
ane | -0.022 |
eddy | -0.022 |
ays | -0.021 |
he | -0.021 |
Token | ig |
Token ID | 45329 |
Feature activation | -1.24e-01 |
Pos logit contributions | |
ing | +0.045 |
ane | +0.042 |
ays | +0.041 |
eddy | +0.040 |
he | +0.039 |
Neg logit contributions | |
never | -0.033 |
forgot | -0.033 |
forgotten | -0.031 |
not | -0.030 |
wasted | -0.030 |
Token | loo |
Token ID | 29680 |
Feature activation | -8.75e-02 |
Pos logit contributions | |
ays | +0.194 |
ing | +0.187 |
ane | +0.166 |
umping | +0.162 |
over | +0.161 |
Neg logit contributions | |
never | -0.254 |
not | -0.234 |
help | -0.230 |
stolen | -0.227 |
forgotten | -0.226 |
Token | and |
Token ID | 290 |
Feature activation | +3.57e-02 |
Pos logit contributions | |
ane | +0.162 |
ing | +0.160 |
ays | +0.158 |
eddy | +0.150 |
ement | +0.147 |
Neg logit contributions | |
never | -0.181 |
not | -0.171 |
knocked | -0.163 |
forgotten | -0.160 |
always | -0.159 |
Token | wanted |
Token ID | 2227 |
Feature activation | -5.08e-03 |
Pos logit contributions | |
permission | +0.055 |
tresp | +0.051 |
known | +0.050 |
least | +0.049 |
unknow | +0.049 |
Neg logit contributions | |
ing | -0.080 |
he | -0.068 |
ed | -0.066 |
ane | -0.065 |
chair | -0.064 |
Token | to |
Token ID | 284 |
Feature activation | +3.69e-02 |
Pos logit contributions | |
ing | +0.009 |
ane | +0.007 |
eddy | +0.007 |
ays | +0.007 |
ore | +0.006 |
Neg logit contributions | |
forgot | -0.011 |
forget | -0.010 |
known | -0.010 |
setup | -0.010 |
fact | -0.010 |
Token | play |
Token ID | 711 |
Feature activation | -5.19e-02 |
Pos logit contributions | |
forgot | +0.050 |
forgotten | +0.049 |
setup | +0.049 |
known | +0.047 |
wasted | +0.046 |
Neg logit contributions | |
he | -0.087 |
ing | -0.084 |
ane | -0.083 |
eddy | -0.080 |
ay | -0.080 |
Token | loud |
Token ID | 7812 |
Feature activation | -3.36e-02 |
Pos logit contributions | |
never | +0.009 |
forgot | +0.009 |
forgotten | +0.009 |
known | +0.009 |
permission | +0.009 |
Neg logit contributions | |
ing | -0.017 |
ane | -0.015 |
eddy | -0.015 |
he | -0.015 |
ays | -0.014 |
Token | noise |
Token ID | 7838 |
Feature activation | -4.80e-02 |
Pos logit contributions | |
ing | +0.091 |
ane | +0.088 |
ays | +0.087 |
eddy | +0.083 |
he | +0.082 |
Neg logit contributions | |
never | -0.039 |
forgot | -0.036 |
help | -0.035 |
not | -0.035 |
grown | -0.034 |
Token | coming |
Token ID | 2406 |
Feature activation | -2.31e-02 |
Pos logit contributions | |
ays | +0.102 |
ane | +0.102 |
ing | +0.095 |
ite | +0.094 |
eddy | +0.094 |
Neg logit contributions | |
help | -0.085 |
never | -0.082 |
knocked | -0.081 |
not | -0.080 |
grown | -0.077 |
Token | from |
Token ID | 422 |
Feature activation | +8.03e-03 |
Pos logit contributions | |
ane | +0.048 |
ays | +0.047 |
ement | +0.046 |
ing | +0.046 |
ase | +0.042 |
Neg logit contributions | |
never | -0.043 |
help | -0.041 |
not | -0.040 |
be | -0.039 |
grown | -0.039 |
Token | an |
Token ID | 281 |
Feature activation | +5.34e-03 |
Pos logit contributions | |
never | +0.011 |
forgot | +0.011 |
forgotten | +0.010 |
wasted | +0.010 |
not | +0.010 |
Neg logit contributions | |
ing | -0.020 |
ane | -0.019 |
ays | -0.018 |
he | -0.018 |
eddy | -0.018 |
Token | ig |
Token ID | 45329 |
Feature activation | -1.23e-01 |
Pos logit contributions | |
forgot | +0.008 |
never | +0.008 |
forgotten | +0.007 |
known | +0.007 |
wasted | +0.007 |
Neg logit contributions | |
ing | -0.013 |
ane | -0.012 |
ays | -0.012 |
he | -0.012 |
eddy | -0.012 |
Token | loo |
Token ID | 29680 |
Feature activation | -7.94e-02 |
Pos logit contributions | |
ing | +0.180 |
ays | +0.172 |
ane | +0.157 |
eddy | +0.156 |
over | +0.142 |
Neg logit contributions | |
never | -0.277 |
forgotten | -0.258 |
stolen | -0.252 |
help | -0.251 |
not | -0.250 |
Token | . |
Token ID | 13 |
Feature activation | -5.08e-02 |
Pos logit contributions | |
ing | +0.141 |
ane | +0.136 |
ays | +0.133 |
eddy | +0.130 |
ement | +0.123 |
Neg logit contributions | |
never | -0.170 |
not | -0.159 |
forgotten | -0.157 |
knocked | -0.155 |
forgot | -0.153 |
Token |
|
Token ID | 198 |
Feature activation | -6.08e-02 |
Pos logit contributions | |
ing | +0.096 |
ane | +0.091 |
ays | +0.091 |
eddy | +0.087 |
he | +0.086 |
Neg logit contributions | |
never | -0.101 |
forgot | -0.095 |
grown | -0.094 |
not | -0.094 |
help | -0.092 |
Token |
|
Token ID | 198 |
Feature activation | -4.39e-02 |
Pos logit contributions | |
ays | +0.135 |
ing | +0.133 |
ane | +0.129 |
ement | +0.122 |
eddy | +0.119 |
Neg logit contributions | |
never | -0.111 |
not | -0.101 |
know | -0.098 |
help | -0.098 |
grown | -0.097 |
Token | C |
Token ID | 34 |
Feature activation | -3.62e-02 |
Pos logit contributions | |
ing | +0.096 |
ays | +0.096 |
ane | +0.093 |
eddy | +0.088 |
ement | +0.087 |
Neg logit contributions | |
never | -0.076 |
grown | -0.070 |
forgot | -0.069 |
not | -0.068 |
help | -0.067 |
Token | ens |
Token ID | 641 |
Feature activation | -4.79e-03 |
Pos logit contributions | |
ane | +0.186 |
ays | +0.178 |
ing | +0.173 |
ement | +0.172 |
he | +0.164 |
Neg logit contributions | |
forgot | -0.118 |
never | -0.105 |
not | -0.099 |
forgotten | -0.099 |
help | -0.099 |
Token | accidentally |
Token ID | 14716 |
Feature activation | +2.60e-02 |
Pos logit contributions | |
ing | +0.013 |
ane | +0.011 |
he | +0.011 |
ays | +0.011 |
eddy | +0.011 |
Neg logit contributions | |
permission | -0.006 |
known | -0.006 |
never | -0.005 |
help | -0.005 |
unknow | -0.005 |
Token | knocked |
Token ID | 13642 |
Feature activation | -3.14e-02 |
Pos logit contributions | |
known | +0.027 |
setup | +0.027 |
permission | +0.026 |
fact | +0.026 |
never | +0.026 |
Neg logit contributions | |
ing | -0.068 |
ays | -0.061 |
ane | -0.060 |
ed | -0.059 |
eddy | -0.058 |
Token | over |
Token ID | 625 |
Feature activation | -9.48e-03 |
Pos logit contributions | |
ane | +0.063 |
ing | +0.063 |
ement | +0.059 |
ays | +0.058 |
eddy | +0.056 |
Neg logit contributions | |
never | -0.059 |
forgotten | -0.054 |
not | -0.054 |
forgot | -0.054 |
been | -0.053 |
Token | a |
Token ID | 257 |
Feature activation | +7.86e-03 |
Pos logit contributions | |
ing | +0.025 |
ays | +0.023 |
he | +0.022 |
ane | +0.022 |
eddy | +0.022 |
Neg logit contributions | |
forgot | -0.012 |
never | -0.012 |
forgotten | -0.011 |
known | -0.011 |
unknow | -0.011 |
Token | v |
Token ID | 410 |
Feature activation | -1.23e-01 |
Pos logit contributions | |
never | +0.012 |
forgot | +0.011 |
forgotten | +0.011 |
known | +0.011 |
not | +0.010 |
Neg logit contributions | |
ing | -0.019 |
ane | -0.018 |
ays | -0.017 |
eddy | -0.017 |
he | -0.017 |
Token | ase |
Token ID | 589 |
Feature activation | -8.44e-02 |
Pos logit contributions | |
ing | +0.115 |
ays | +0.090 |
ement | +0.089 |
ane | +0.084 |
eddy | +0.060 |
Neg logit contributions | |
never | -0.362 |
not | -0.342 |
help | -0.333 |
forgotten | -0.330 |
forgot | -0.329 |
Token | and |
Token ID | 290 |
Feature activation | +4.07e-02 |
Pos logit contributions | |
ane | +0.162 |
ing | +0.160 |
ement | +0.155 |
ays | +0.154 |
eddy | +0.154 |
Neg logit contributions | |
never | -0.166 |
not | -0.158 |
knocked | -0.149 |
help | -0.146 |
forgotten | -0.146 |
Token | it |
Token ID | 340 |
Feature activation | +1.48e-02 |
Pos logit contributions | |
known | +0.076 |
permission | +0.076 |
strive | +0.070 |
humor | +0.070 |
suppose | +0.069 |
Neg logit contributions | |
ing | -0.075 |
ed | -0.065 |
he | -0.057 |
s | -0.056 |
ays | -0.056 |
Token | shattered |
Token ID | 23273 |
Feature activation | -3.76e-02 |
Pos logit contributions | |
never | +0.018 |
forgot | +0.018 |
permission | +0.018 |
known | +0.017 |
forgotten | +0.017 |
Neg logit contributions | |
ing | -0.039 |
ays | -0.035 |
ane | -0.035 |
he | -0.034 |
eddy | -0.034 |
Token | into |
Token ID | 656 |
Feature activation | +1.86e-02 |
Pos logit contributions | |
ane | +0.073 |
ing | +0.068 |
he | +0.065 |
ement | +0.063 |
ite | +0.062 |
Neg logit contributions | |
never | -0.070 |
forgotten | -0.068 |
wasted | -0.067 |
knocked | -0.067 |
forgot | -0.066 |
Token | to |
Token ID | 284 |
Feature activation | +4.12e-02 |
Pos logit contributions | |
never | +0.010 |
forgot | +0.010 |
forgotten | +0.010 |
not | +0.010 |
help | +0.009 |
Neg logit contributions | |
ing | -0.016 |
ane | -0.015 |
ays | -0.015 |
eddy | -0.014 |
he | -0.014 |
Token | help |
Token ID | 1037 |
Feature activation | -2.17e-02 |
Pos logit contributions | |
forgot | +0.048 |
wished | +0.047 |
forgotten | +0.043 |
wasted | +0.042 |
known | +0.042 |
Neg logit contributions | |
he | -0.103 |
ing | -0.097 |
ay | -0.097 |
room | -0.096 |
eddy | -0.094 |
Token | carry |
Token ID | 3283 |
Feature activation | -3.10e-02 |
Pos logit contributions | |
ing | +0.056 |
ane | +0.051 |
ays | +0.051 |
he | +0.050 |
eddy | +0.050 |
Neg logit contributions | |
never | -0.029 |
forgot | -0.028 |
forgotten | -0.027 |
known | -0.026 |
wasted | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
ing | +0.081 |
ays | +0.075 |
ane | +0.074 |
eddy | +0.072 |
he | +0.072 |
Neg logit contributions | |
forgot | -0.041 |
never | -0.041 |
forgotten | -0.037 |
known | -0.036 |
wasted | -0.035 |
Token | fragile |
Token ID | 21049 |
Feature activation | -1.43e-02 |
Pos logit contributions | |
never | +0.018 |
forgot | +0.018 |
known | +0.017 |
forgotten | +0.017 |
not | +0.016 |
Neg logit contributions | |
ing | -0.027 |
ane | -0.025 |
ays | -0.024 |
eddy | -0.024 |
he | -0.024 |
Token | v |
Token ID | 410 |
Feature activation | -1.23e-01 |
Pos logit contributions | |
ing | +0.033 |
ane | +0.030 |
ays | +0.029 |
eddy | +0.029 |
he | +0.028 |
Neg logit contributions | |
forgot | -0.024 |
never | -0.024 |
known | -0.022 |
forgotten | -0.022 |
wasted | -0.021 |
Token | ase |
Token ID | 589 |
Feature activation | -6.93e-02 |
Pos logit contributions | |
ing | +0.130 |
ays | +0.113 |
ement | +0.108 |
ane | +0.098 |
ogging | +0.072 |
Neg logit contributions | |
never | -0.335 |
not | -0.320 |
help | -0.313 |
always | -0.302 |
been | -0.302 |
Token | to |
Token ID | 284 |
Feature activation | +3.32e-02 |
Pos logit contributions | |
ing | +0.139 |
ays | +0.137 |
ane | +0.135 |
ement | +0.133 |
eddy | +0.126 |
Neg logit contributions | |
never | -0.133 |
not | -0.128 |
help | -0.124 |
forgotten | -0.121 |
permission | -0.118 |
Token | the |
Token ID | 262 |
Feature activation | +2.70e-04 |
Pos logit contributions | |
forgot | +0.053 |
unknow | +0.047 |
forgotten | +0.046 |
wished | +0.046 |
never | +0.046 |
Neg logit contributions | |
he | -0.072 |
ing | -0.066 |
eddy | -0.062 |
room | -0.060 |
ay | -0.059 |
Token | living |
Token ID | 2877 |
Feature activation | -2.54e-02 |
Pos logit contributions | |
forgot | +0.000 |
wasted | +0.000 |
think | +0.000 |
never | +0.000 |
know | +0.000 |
Neg logit contributions | |
ing | -0.001 |
eddy | -0.001 |
he | -0.001 |
ane | -0.001 |
ays | -0.000 |
Token | room |
Token ID | 2119 |
Feature activation | -4.98e-02 |
Pos logit contributions | |
ing | +0.066 |
ane | +0.062 |
ays | +0.061 |
eddy | +0.059 |
he | +0.059 |
Neg logit contributions | |
never | -0.033 |
forgot | -0.032 |
forgotten | -0.030 |
not | -0.029 |
wasted | -0.029 |
Token | her |
Token ID | 607 |
Feature activation | +2.38e-02 |
Pos logit contributions | |
forgot | +0.003 |
never | +0.003 |
forgotten | +0.003 |
wasted | +0.003 |
permission | +0.003 |
Neg logit contributions | |
ing | -0.006 |
ane | -0.006 |
ays | -0.006 |
eddy | -0.006 |
he | -0.006 |
Token | cat |
Token ID | 3797 |
Feature activation | -5.19e-02 |
Pos logit contributions | |
never | +0.032 |
forgot | +0.031 |
forgotten | +0.030 |
not | +0.029 |
wasted | +0.028 |
Neg logit contributions | |
ing | -0.061 |
ane | -0.057 |
ays | -0.056 |
eddy | -0.054 |
he | -0.054 |
Token | knock |
Token ID | 10643 |
Feature activation | -3.18e-02 |
Pos logit contributions | |
ane | +0.127 |
ing | +0.126 |
ays | +0.125 |
eddy | +0.118 |
ement | +0.116 |
Neg logit contributions | |
never | -0.077 |
not | -0.072 |
forgot | -0.070 |
knocked | -0.068 |
help | -0.067 |
Token | over |
Token ID | 625 |
Feature activation | -1.34e-02 |
Pos logit contributions | |
ing | +0.062 |
ane | +0.058 |
ays | +0.056 |
eddy | +0.055 |
he | +0.053 |
Neg logit contributions | |
never | -0.063 |
forgotten | -0.059 |
forgot | -0.058 |
not | -0.058 |
been | -0.057 |
Token | a |
Token ID | 257 |
Feature activation | -1.34e-03 |
Pos logit contributions | |
ing | +0.034 |
ane | +0.033 |
ays | +0.032 |
eddy | +0.030 |
he | +0.029 |
Neg logit contributions | |
never | -0.019 |
help | -0.018 |
forgotten | -0.018 |
forgot | -0.017 |
not | -0.017 |
Token | v |
Token ID | 410 |
Feature activation | -1.23e-01 |
Pos logit contributions | |
ing | +0.003 |
ane | +0.003 |
ays | +0.003 |
eddy | +0.003 |
he | +0.003 |
Neg logit contributions | |
never | -0.002 |
forgot | -0.002 |
forgotten | -0.002 |
not | -0.002 |
permission | -0.002 |
Token | ase |
Token ID | 589 |
Feature activation | -8.51e-02 |
Pos logit contributions | |
ing | +0.124 |
ays | +0.094 |
ement | +0.090 |
ane | +0.087 |
ogging | +0.065 |
Neg logit contributions | |
never | -0.354 |
not | -0.333 |
forgotten | -0.321 |
been | -0.321 |
forgot | -0.321 |
Token | . |
Token ID | 13 |
Feature activation | -5.75e-02 |
Pos logit contributions | |
ing | +0.172 |
ane | +0.164 |
ays | +0.161 |
ement | +0.159 |
eddy | +0.159 |
Neg logit contributions | |
never | -0.157 |
not | -0.151 |
help | -0.140 |
forgotten | -0.140 |
been | -0.140 |
Token | The |
Token ID | 383 |
Feature activation | -3.58e-03 |
Pos logit contributions | |
ing | +0.127 |
ane | +0.120 |
ays | +0.116 |
eddy | +0.109 |
ement | +0.109 |
Neg logit contributions | |
never | -0.102 |
not | -0.098 |
forgot | -0.097 |
grown | -0.095 |
knocked | -0.094 |
Token | v |
Token ID | 410 |
Feature activation | -1.03e-01 |
Pos logit contributions | |
ing | +0.007 |
ane | +0.007 |
eddy | +0.007 |
he | +0.006 |
ays | +0.006 |
Neg logit contributions | |
forgot | -0.007 |
never | -0.006 |
forgotten | -0.006 |
known | -0.006 |
wasted | -0.006 |
Token | ase |
Token ID | 589 |
Feature activation | -1.77e-02 |
Pos logit contributions | |
ing | +0.103 |
ement | +0.084 |
ays | +0.082 |
ane | +0.077 |
ogging | +0.069 |
Neg logit contributions | |
never | -0.296 |
not | -0.280 |
forgot | -0.275 |
been | -0.270 |
wasted | -0.268 |
Token | , |
Token ID | 11 |
Feature activation | -2.62e-02 |
Pos logit contributions | |
ane | +0.102 |
ays | +0.100 |
ing | +0.097 |
eddy | +0.093 |
ement | +0.085 |
Neg logit contributions | |
help | -0.098 |
never | -0.092 |
not | -0.091 |
grown | -0.089 |
be | -0.086 |
Token | Tim |
Token ID | 5045 |
Feature activation | -1.06e-02 |
Pos logit contributions | |
ing | +0.062 |
ane | +0.057 |
ays | +0.057 |
eddy | +0.055 |
he | +0.055 |
Neg logit contributions | |
never | -0.040 |
forgot | -0.039 |
forgotten | -0.038 |
permission | -0.036 |
wasted | -0.036 |
Token | accidentally |
Token ID | 14716 |
Feature activation | +2.77e-02 |
Pos logit contributions | |
ing | +0.026 |
ane | +0.025 |
ays | +0.025 |
eddy | +0.024 |
he | +0.024 |
Neg logit contributions | |
never | -0.015 |
forgot | -0.015 |
forgotten | -0.014 |
not | -0.013 |
wasted | -0.013 |
Token | hit |
Token ID | 2277 |
Feature activation | -2.58e-02 |
Pos logit contributions | |
never | +0.030 |
known | +0.029 |
forgot | +0.028 |
permission | +0.028 |
setup | +0.028 |
Neg logit contributions | |
ing | -0.073 |
ays | -0.067 |
ane | -0.066 |
eddy | -0.064 |
he | -0.062 |
Token | the |
Token ID | 262 |
Feature activation | -1.81e-02 |
Pos logit contributions | |
ing | +0.074 |
ane | +0.073 |
ays | +0.069 |
ement | +0.067 |
eddy | +0.067 |
Neg logit contributions | |
never | -0.027 |
not | -0.024 |
forgotten | -0.023 |
forgot | -0.023 |
been | -0.023 |
Token | v |
Token ID | 410 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
ing | +0.041 |
ane | +0.038 |
he | +0.037 |
eddy | +0.037 |
ays | +0.037 |
Neg logit contributions | |
forgot | -0.029 |
known | -0.029 |
never | -0.028 |
forgotten | -0.028 |
permission | -0.027 |
Token | ase |
Token ID | 589 |
Feature activation | -7.01e-02 |
Pos logit contributions | |
ing | +0.126 |
ement | +0.102 |
ays | +0.097 |
ane | +0.090 |
ogging | +0.079 |
Neg logit contributions | |
never | -0.347 |
not | -0.328 |
knocked | -0.314 |
grown | -0.312 |
been | -0.311 |
Token | . |
Token ID | 13 |
Feature activation | -4.35e-02 |
Pos logit contributions | |
ane | +0.142 |
eddy | +0.138 |
ing | +0.137 |
ays | +0.133 |
ement | +0.128 |
Neg logit contributions | |
never | -0.141 |
not | -0.138 |
help | -0.123 |
be | -0.118 |
forgotten | -0.116 |
Token | The |
Token ID | 383 |
Feature activation | +3.29e-03 |
Pos logit contributions | |
ing | +0.096 |
ane | +0.091 |
ays | +0.089 |
eddy | +0.084 |
he | +0.083 |
Neg logit contributions | |
never | -0.074 |
forgot | -0.072 |
not | -0.070 |
forgotten | -0.069 |
grown | -0.068 |
Token | v |
Token ID | 410 |
Feature activation | -1.02e-01 |
Pos logit contributions | |
forgot | +0.006 |
known | +0.006 |
forgotten | +0.006 |
never | +0.006 |
unknow | +0.006 |
Neg logit contributions | |
ing | -0.006 |
ane | -0.006 |
eddy | -0.006 |
he | -0.005 |
ay | -0.005 |
Token | ase |
Token ID | 589 |
Feature activation | -1.70e-02 |
Pos logit contributions | |
ing | +0.097 |
ement | +0.078 |
ays | +0.075 |
ane | +0.072 |
ogging | +0.063 |
Neg logit contributions | |
never | -0.297 |
not | -0.281 |
forgot | -0.278 |
wasted | -0.272 |
been | -0.271 |
Token | he |
Token ID | 339 |
Feature activation | +4.26e-02 |
Pos logit contributions | |
ing | +0.022 |
ane | +0.020 |
ays | +0.020 |
eddy | +0.019 |
he | +0.019 |
Neg logit contributions | |
never | -0.016 |
forgot | -0.016 |
forgotten | -0.015 |
not | -0.015 |
wasted | -0.015 |
Token | accidentally |
Token ID | 14716 |
Feature activation | +3.11e-02 |
Pos logit contributions | |
never | +0.039 |
forgot | +0.036 |
forgotten | +0.034 |
not | +0.032 |
help | +0.032 |
Neg logit contributions | |
ing | -0.127 |
ane | -0.120 |
ays | -0.119 |
eddy | -0.115 |
he | -0.115 |
Token | knocked |
Token ID | 13642 |
Feature activation | -3.20e-02 |
Pos logit contributions | |
known | +0.038 |
setup | +0.038 |
permission | +0.036 |
fact | +0.036 |
never | +0.035 |
Neg logit contributions | |
ing | -0.074 |
ays | -0.068 |
ane | -0.067 |
ed | -0.064 |
eddy | -0.062 |
Token | over |
Token ID | 625 |
Feature activation | -9.27e-03 |
Pos logit contributions | |
ing | +0.066 |
ane | +0.064 |
ement | +0.060 |
ays | +0.060 |
eddy | +0.058 |
Neg logit contributions | |
never | -0.059 |
not | -0.054 |
forgotten | -0.054 |
been | -0.054 |
help | -0.053 |
Token | a |
Token ID | 257 |
Feature activation | +7.37e-03 |
Pos logit contributions | |
ing | +0.025 |
ane | +0.023 |
ays | +0.023 |
he | +0.023 |
eddy | +0.023 |
Neg logit contributions | |
never | -0.011 |
forgot | -0.011 |
forgotten | -0.010 |
wasted | -0.010 |
not | -0.010 |
Token | v |
Token ID | 410 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
never | +0.011 |
forgot | +0.011 |
forgotten | +0.010 |
known | +0.010 |
permission | +0.010 |
Neg logit contributions | |
ing | -0.018 |
ane | -0.017 |
ays | -0.016 |
eddy | -0.016 |
he | -0.016 |
Token | ase |
Token ID | 589 |
Feature activation | -8.24e-02 |
Pos logit contributions | |
ing | +0.108 |
ays | +0.081 |
ement | +0.075 |
ane | +0.075 |
eddy | +0.053 |
Neg logit contributions | |
never | -0.365 |
not | -0.345 |
help | -0.340 |
forgotten | -0.337 |
forgot | -0.336 |
Token | . |
Token ID | 13 |
Feature activation | -6.25e-02 |
Pos logit contributions | |
ing | +0.154 |
ane | +0.150 |
ement | +0.148 |
eddy | +0.143 |
ays | +0.143 |
Neg logit contributions | |
never | -0.166 |
not | -0.158 |
knocked | -0.150 |
forgotten | -0.148 |
help | -0.148 |
Token | The |
Token ID | 383 |
Feature activation | +7.11e-03 |
Pos logit contributions | |
ing | +0.142 |
ane | +0.134 |
ays | +0.131 |
eddy | +0.124 |
ement | +0.122 |
Neg logit contributions | |
never | -0.105 |
not | -0.102 |
forgot | -0.101 |
help | -0.096 |
forgotten | -0.096 |
Token | v |
Token ID | 410 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
forgot | +0.014 |
known | +0.014 |
forgotten | +0.013 |
never | +0.013 |
wasted | +0.013 |
Neg logit contributions | |
ing | -0.013 |
ane | -0.012 |
ay | -0.012 |
he | -0.012 |
eddy | -0.012 |
Token | ase |
Token ID | 589 |
Feature activation | -1.99e-02 |
Pos logit contributions | |
ing | +0.095 |
ement | +0.075 |
ays | +0.075 |
ane | +0.069 |
ogging | +0.056 |
Neg logit contributions | |
never | -0.297 |
not | -0.279 |
forgot | -0.275 |
been | -0.270 |
grown | -0.268 |
Token | brother |
Token ID | 3956 |
Feature activation | -3.93e-02 |
Pos logit contributions | |
never | +0.003 |
forgot | +0.003 |
forgotten | +0.003 |
not | +0.003 |
wasted | +0.003 |
Neg logit contributions | |
ing | -0.007 |
ane | -0.007 |
ays | -0.007 |
eddy | -0.006 |
he | -0.006 |
Token | accidentally |
Token ID | 14716 |
Feature activation | +3.07e-02 |
Pos logit contributions | |
ing | +0.083 |
ane | +0.073 |
eddy | +0.071 |
ays | +0.070 |
he | +0.067 |
Neg logit contributions | |
never | -0.067 |
permission | -0.066 |
forgotten | -0.065 |
unknow | -0.065 |
forgot | -0.063 |
Token | knocked |
Token ID | 13642 |
Feature activation | -2.54e-02 |
Pos logit contributions | |
known | +0.034 |
fact | +0.033 |
never | +0.032 |
forgot | +0.032 |
setup | +0.032 |
Neg logit contributions | |
ing | -0.078 |
ays | -0.072 |
ane | -0.071 |
eddy | -0.068 |
ed | -0.067 |
Token | over |
Token ID | 625 |
Feature activation | -7.51e-03 |
Pos logit contributions | |
ing | +0.052 |
ane | +0.050 |
ays | +0.047 |
ement | +0.047 |
eddy | +0.046 |
Neg logit contributions | |
never | -0.047 |
forgotten | -0.043 |
not | -0.043 |
forgot | -0.043 |
been | -0.043 |
Token | the |
Token ID | 262 |
Feature activation | -1.17e-02 |
Pos logit contributions | |
ing | +0.021 |
ane | +0.019 |
ays | +0.019 |
eddy | +0.019 |
he | +0.019 |
Neg logit contributions | |
never | -0.009 |
forgot | -0.008 |
forgotten | -0.008 |
not | -0.008 |
wasted | -0.008 |
Token | v |
Token ID | 410 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
ing | +0.027 |
ane | +0.025 |
eddy | +0.025 |
ays | +0.024 |
he | +0.024 |
Neg logit contributions | |
forgot | -0.019 |
never | -0.018 |
known | -0.018 |
forgotten | -0.017 |
not | -0.016 |
Token | ase |
Token ID | 589 |
Feature activation | -7.67e-02 |
Pos logit contributions | |
ing | +0.133 |
ement | +0.102 |
ays | +0.101 |
ane | +0.090 |
ogging | +0.076 |
Neg logit contributions | |
never | -0.335 |
not | -0.315 |
been | -0.303 |
knocked | -0.301 |
grown | -0.301 |
Token | and |
Token ID | 290 |
Feature activation | +3.92e-02 |
Pos logit contributions | |
ane | +0.151 |
ing | +0.147 |
ays | +0.145 |
eddy | +0.144 |
ement | +0.142 |
Neg logit contributions | |
never | -0.148 |
not | -0.143 |
help | -0.131 |
knocked | -0.128 |
forgotten | -0.128 |
Token | it |
Token ID | 340 |
Feature activation | +1.63e-02 |
Pos logit contributions | |
known | +0.070 |
permission | +0.070 |
unknow | +0.066 |
shown | +0.064 |
setup | +0.063 |
Neg logit contributions | |
ing | -0.082 |
ays | -0.069 |
he | -0.067 |
ane | -0.063 |
eddy | -0.063 |
Token | broke |
Token ID | 6265 |
Feature activation | -3.09e-02 |
Pos logit contributions | |
never | +0.020 |
forgot | +0.019 |
permission | +0.019 |
known | +0.018 |
forgotten | +0.018 |
Neg logit contributions | |
ing | -0.043 |
ays | -0.040 |
ane | -0.040 |
he | -0.038 |
eddy | -0.038 |
Token | into |
Token ID | 656 |
Feature activation | +1.57e-02 |
Pos logit contributions | |
ane | +0.060 |
ing | +0.056 |
ement | +0.054 |
he | +0.053 |
ays | +0.053 |
Neg logit contributions | |
never | -0.060 |
not | -0.057 |
forgotten | -0.057 |
wasted | -0.056 |
knocked | -0.055 |
Token | at |
Token ID | 379 |
Feature activation | +9.14e-03 |
Pos logit contributions | |
never | +0.108 |
forgot | +0.106 |
permission | +0.102 |
wished | +0.101 |
unknow | +0.100 |
Neg logit contributions | |
ing | -0.124 |
ays | -0.114 |
ane | -0.112 |
he | -0.108 |
eddy | -0.102 |
Token | the |
Token ID | 262 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
forgot | +0.014 |
stolen | +0.012 |
unknow | +0.012 |
never | +0.012 |
forgotten | +0.012 |
Neg logit contributions | |
ing | -0.022 |
ays | -0.020 |
ane | -0.020 |
he | -0.019 |
eddy | -0.019 |
Token | top |
Token ID | 1353 |
Feature activation | -5.01e-02 |
Pos logit contributions | |
forgot | +0.038 |
wasted | +0.037 |
wished | +0.036 |
never | +0.036 |
stolen | +0.035 |
Neg logit contributions | |
ing | -0.035 |
he | -0.030 |
ane | -0.029 |
ays | -0.027 |
eddy | -0.026 |
Token | of |
Token ID | 286 |
Feature activation | +3.36e-02 |
Pos logit contributions | |
ing | +0.085 |
ane | +0.081 |
eddy | +0.081 |
ays | +0.080 |
ement | +0.071 |
Neg logit contributions | |
never | -0.109 |
forgotten | -0.103 |
forgot | -0.102 |
not | -0.100 |
grown | -0.099 |
Token | the |
Token ID | 262 |
Feature activation | +1.09e-02 |
Pos logit contributions | |
forgot | +0.062 |
forget | +0.051 |
forgotten | +0.051 |
stolen | +0.050 |
wasted | +0.050 |
Neg logit contributions | |
ing | -0.070 |
ays | -0.063 |
he | -0.063 |
ane | -0.061 |
ement | -0.058 |
Token | anth |
Token ID | 26794 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
forgot | +0.017 |
never | +0.017 |
known | +0.017 |
wasted | +0.016 |
stolen | +0.016 |
Neg logit contributions | |
ing | -0.025 |
ane | -0.022 |
he | -0.022 |
ays | -0.021 |
eddy | -0.020 |
Token | ill |
Token ID | 359 |
Feature activation | -2.01e-02 |
Pos logit contributions | |
ing | +0.189 |
ane | +0.171 |
eddy | +0.170 |
ays | +0.168 |
ement | +0.162 |
Neg logit contributions | |
never | -0.273 |
forgot | -0.260 |
forgotten | -0.257 |
knocked | -0.247 |
been | -0.246 |
Token | . |
Token ID | 13 |
Feature activation | -3.83e-02 |
Pos logit contributions | |
ing | +0.034 |
ane | +0.033 |
ays | +0.033 |
eddy | +0.032 |
he | +0.029 |
Neg logit contributions | |
never | -0.044 |
forgot | -0.043 |
forgotten | -0.041 |
not | -0.040 |
been | -0.040 |
Token | But |
Token ID | 887 |
Feature activation | +1.75e-02 |
Pos logit contributions | |
ing | +0.069 |
ane | +0.063 |
ays | +0.061 |
eddy | +0.058 |
he | +0.058 |
Neg logit contributions | |
never | -0.081 |
forgot | -0.079 |
forgotten | -0.077 |
wasted | -0.075 |
not | -0.075 |
Token | the |
Token ID | 262 |
Feature activation | +2.28e-02 |
Pos logit contributions | |
forgot | +0.029 |
forgotten | +0.029 |
wasted | +0.029 |
stolen | +0.028 |
never | +0.028 |
Neg logit contributions | |
ing | -0.038 |
ane | -0.033 |
he | -0.031 |
ays | -0.031 |
ement | -0.030 |
Token | ant |
Token ID | 1885 |
Feature activation | +1.04e-03 |
Pos logit contributions | |
known | +0.037 |
forgot | +0.037 |
wasted | +0.037 |
never | +0.036 |
stolen | +0.035 |
Neg logit contributions | |
ing | -0.051 |
ane | -0.042 |
he | -0.040 |
ays | -0.040 |
ore | -0.039 |
Token | He |
Token ID | 679 |
Feature activation | +6.90e-02 |
Pos logit contributions | |
ing | +0.079 |
ane | +0.078 |
ays | +0.074 |
ite | +0.070 |
eddy | +0.070 |
Neg logit contributions | |
not | -0.062 |
never | -0.061 |
grown | -0.059 |
help | -0.056 |
be | -0.055 |
Token | tries |
Token ID | 8404 |
Feature activation | -1.74e-03 |
Pos logit contributions | |
permission | +0.135 |
known | +0.127 |
forgotten | +0.116 |
unknow | +0.116 |
pap | +0.114 |
Neg logit contributions | |
ing | -0.126 |
ays | -0.124 |
aw | -0.112 |
ane | -0.105 |
ement | -0.104 |
Token | to |
Token ID | 284 |
Feature activation | +5.47e-02 |
Pos logit contributions | |
ing | +0.004 |
ane | +0.004 |
ays | +0.004 |
eddy | +0.003 |
he | +0.003 |
Neg logit contributions | |
never | -0.003 |
forgot | -0.003 |
forgotten | -0.003 |
wasted | -0.003 |
stolen | -0.003 |
Token | take |
Token ID | 1011 |
Feature activation | -5.48e-03 |
Pos logit contributions | |
forgotten | +0.067 |
known | +0.066 |
vitamins | +0.064 |
forgot | +0.064 |
wasted | +0.061 |
Neg logit contributions | |
he | -0.133 |
ing | -0.128 |
eddy | -0.120 |
ingle | -0.118 |
room | -0.115 |
Token | the |
Token ID | 262 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
ing | +0.014 |
ane | +0.014 |
ays | +0.013 |
eddy | +0.013 |
he | +0.013 |
Neg logit contributions | |
never | -0.007 |
forgot | -0.007 |
forgotten | -0.007 |
not | -0.006 |
wasted | -0.006 |
Token | v |
Token ID | 410 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
ing | +0.023 |
ane | +0.022 |
ays | +0.021 |
eddy | +0.021 |
he | +0.021 |
Neg logit contributions | |
forgot | -0.016 |
never | -0.016 |
forgotten | -0.016 |
known | -0.015 |
wasted | -0.015 |
Token | ase |
Token ID | 589 |
Feature activation | -7.88e-02 |
Pos logit contributions | |
ing | +0.133 |
ement | +0.124 |
ays | +0.102 |
ane | +0.101 |
ogging | +0.098 |
Neg logit contributions | |
never | -0.308 |
not | -0.294 |
help | -0.282 |
be | -0.278 |
been | -0.269 |
Token | from |
Token ID | 422 |
Feature activation | +3.98e-03 |
Pos logit contributions | |
ays | +0.170 |
ing | +0.169 |
ane | +0.168 |
ement | +0.167 |
eddy | +0.158 |
Neg logit contributions | |
not | -0.143 |
never | -0.143 |
help | -0.135 |
be | -0.125 |
forgotten | -0.120 |
Token | Anna |
Token ID | 11735 |
Feature activation | -4.49e-02 |
Pos logit contributions | |
never | +0.006 |
forgot | +0.005 |
forgotten | +0.005 |
not | +0.005 |
help | +0.005 |
Neg logit contributions | |
ing | -0.010 |
ane | -0.010 |
ays | -0.009 |
eddy | -0.009 |
he | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | -3.86e-02 |
Pos logit contributions | |
ing | +0.085 |
ane | +0.084 |
ays | +0.081 |
eddy | +0.078 |
ement | +0.077 |
Neg logit contributions | |
never | -0.090 |
not | -0.085 |
forgot | -0.084 |
forgotten | -0.081 |
help | -0.081 |
Token |
|
Token ID | 198 |
Feature activation | -4.13e-02 |
Pos logit contributions | |
ing | +0.092 |
ane | +0.090 |
ays | +0.085 |
eddy | +0.084 |
ite | +0.082 |
Neg logit contributions | |
not | -0.068 |
never | -0.066 |
grown | -0.060 |
help | -0.059 |
forgot | -0.058 |
Token | walk |
Token ID | 2513 |
Feature activation | -5.38e-02 |
Pos logit contributions | |
ing | +0.066 |
ane | +0.063 |
he | +0.063 |
ays | +0.059 |
eddy | +0.059 |
Neg logit contributions | |
forgot | -0.045 |
known | -0.045 |
permission | -0.045 |
forgotten | -0.045 |
wasted | -0.045 |
Token | and |
Token ID | 290 |
Feature activation | +4.33e-02 |
Pos logit contributions | |
ays | +0.119 |
ane | +0.110 |
eddy | +0.110 |
ement | +0.106 |
ay | +0.099 |
Neg logit contributions | |
never | -0.105 |
always | -0.093 |
not | -0.091 |
be | -0.089 |
been | -0.087 |
Token | found |
Token ID | 1043 |
Feature activation | -9.41e-04 |
Pos logit contributions | |
permission | +0.057 |
wasted | +0.052 |
tresp | +0.052 |
help | +0.051 |
need | +0.050 |
Neg logit contributions | |
ing | -0.109 |
he | -0.096 |
ane | -0.095 |
ays | -0.090 |
ed | -0.088 |
Token | a |
Token ID | 257 |
Feature activation | +2.97e-02 |
Pos logit contributions | |
ing | +0.003 |
ane | +0.002 |
ays | +0.002 |
eddy | +0.002 |
he | +0.002 |
Neg logit contributions | |
forgot | -0.001 |
never | -0.001 |
forgotten | -0.001 |
wasted | -0.001 |
setup | -0.001 |
Token | big |
Token ID | 1263 |
Feature activation | -8.08e-03 |
Pos logit contributions | |
permission | +0.044 |
never | +0.044 |
forgot | +0.042 |
known | +0.041 |
wasted | +0.041 |
Neg logit contributions | |
ing | -0.069 |
eddy | -0.063 |
ane | -0.062 |
he | -0.060 |
ay | -0.059 |
Token | ig |
Token ID | 45329 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
ing | +0.017 |
ane | +0.016 |
eddy | +0.015 |
ays | +0.015 |
he | +0.015 |
Neg logit contributions | |
never | -0.014 |
forgot | -0.014 |
known | -0.014 |
forgotten | -0.013 |
permission | -0.013 |
Token | loo |
Token ID | 29680 |
Feature activation | -8.76e-02 |
Pos logit contributions | |
ays | +0.197 |
ing | +0.187 |
ane | +0.170 |
umping | +0.168 |
eddy | +0.167 |
Neg logit contributions | |
never | -0.236 |
forgotten | -0.212 |
help | -0.211 |
not | -0.211 |
stolen | -0.210 |
Token | . |
Token ID | 13 |
Feature activation | -3.37e-02 |
Pos logit contributions | |
ing | +0.153 |
ane | +0.150 |
ays | +0.149 |
eddy | +0.139 |
ement | +0.134 |
Neg logit contributions | |
never | -0.187 |
not | -0.175 |
forgotten | -0.174 |
knocked | -0.171 |
stolen | -0.169 |
Token | He |
Token ID | 679 |
Feature activation | +5.10e-02 |
Pos logit contributions | |
ing | +0.060 |
ane | +0.056 |
ays | +0.054 |
eddy | +0.051 |
he | +0.051 |
Neg logit contributions | |
never | -0.072 |
forgot | -0.070 |
forgotten | -0.068 |
not | -0.067 |
help | -0.067 |
Token | thought |
Token ID | 1807 |
Feature activation | +1.52e-02 |
Pos logit contributions | |
permission | +0.076 |
punish | +0.069 |
known | +0.069 |
supposed | +0.068 |
help | +0.068 |
Neg logit contributions | |
ing | -0.114 |
ane | -0.099 |
he | -0.097 |
eddy | -0.096 |
ed | -0.095 |
Token | it |
Token ID | 340 |
Feature activation | +1.28e-02 |
Pos logit contributions | |
never | +0.023 |
forgot | +0.023 |
forgotten | +0.022 |
wasted | +0.021 |
known | +0.021 |
Neg logit contributions | |
ing | -0.036 |
ane | -0.033 |
ays | -0.033 |
eddy | -0.032 |
he | -0.031 |
Token | Anna |
Token ID | 31160 |
Feature activation | +2.69e-02 |
Pos logit contributions | |
ane | +0.094 |
ing | +0.093 |
ays | +0.092 |
eddy | +0.085 |
ement | +0.084 |
Neg logit contributions | |
never | -0.080 |
forgot | -0.077 |
not | -0.074 |
grown | -0.073 |
be | -0.071 |
Token | wants |
Token ID | 3382 |
Feature activation | +8.59e-04 |
Pos logit contributions | |
never | +0.052 |
forgot | +0.051 |
forgotten | +0.050 |
known | +0.049 |
stolen | +0.049 |
Neg logit contributions | |
ing | -0.052 |
ane | -0.047 |
ays | -0.047 |
he | -0.044 |
eddy | -0.043 |
Token | to |
Token ID | 284 |
Feature activation | +4.27e-02 |
Pos logit contributions | |
forgot | +0.002 |
known | +0.002 |
forget | +0.002 |
unknow | +0.002 |
setup | +0.002 |
Neg logit contributions | |
ing | -0.001 |
eddy | -0.001 |
ays | -0.001 |
ane | -0.001 |
he | -0.001 |
Token | touch |
Token ID | 3638 |
Feature activation | -3.90e-02 |
Pos logit contributions | |
setup | +0.065 |
forgotten | +0.062 |
unknow | +0.061 |
wasted | +0.061 |
known | +0.061 |
Neg logit contributions | |
he | -0.084 |
room | -0.080 |
ay | -0.075 |
ingle | -0.075 |
ane | -0.074 |
Token | the |
Token ID | 262 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
ing | +0.112 |
ane | +0.108 |
ays | +0.104 |
eddy | +0.103 |
he | +0.102 |
Neg logit contributions | |
never | -0.040 |
not | -0.036 |
help | -0.036 |
forgot | -0.035 |
forgotten | -0.034 |
Token | v |
Token ID | 410 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
ing | +0.026 |
ane | +0.024 |
ays | +0.024 |
he | +0.023 |
eddy | +0.023 |
Neg logit contributions | |
forgot | -0.019 |
never | -0.019 |
forgotten | -0.018 |
known | -0.018 |
unknow | -0.018 |
Token | ase |
Token ID | 589 |
Feature activation | -7.67e-02 |
Pos logit contributions | |
ing | +0.137 |
ement | +0.120 |
ays | +0.101 |
ane | +0.100 |
over | +0.094 |
Neg logit contributions | |
never | -0.293 |
not | -0.280 |
help | -0.266 |
be | -0.265 |
been | -0.261 |
Token | . |
Token ID | 13 |
Feature activation | -5.10e-02 |
Pos logit contributions | |
ing | +0.157 |
ane | +0.155 |
ays | +0.152 |
eddy | +0.148 |
acked | +0.136 |
Neg logit contributions | |
not | -0.148 |
never | -0.145 |
help | -0.140 |
be | -0.133 |
permission | -0.131 |
Token | She |
Token ID | 1375 |
Feature activation | +5.42e-02 |
Pos logit contributions | |
ane | +0.111 |
ing | +0.110 |
eddy | +0.107 |
ays | +0.103 |
he | +0.102 |
Neg logit contributions | |
never | -0.085 |
not | -0.084 |
help | -0.080 |
forgot | -0.080 |
think | -0.080 |
Token | climbs |
Token ID | 34332 |
Feature activation | -3.30e-02 |
Pos logit contributions | |
known | +0.090 |
forgotten | +0.085 |
never | +0.083 |
permission | +0.082 |
supposed | +0.082 |
Neg logit contributions | |
ing | -0.116 |
ays | -0.113 |
ane | -0.102 |
ement | -0.100 |
aw | -0.096 |
Token | on |
Token ID | 319 |
Feature activation | -1.57e-02 |
Pos logit contributions | |
ane | +0.059 |
ing | +0.059 |
eddy | +0.055 |
ays | +0.054 |
ement | +0.054 |
Neg logit contributions | |
never | -0.069 |
not | -0.063 |
knocked | -0.063 |
help | -0.062 |
forgotten | -0.062 |
Token | dared |
Token ID | 28765 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
pap | +0.007 |
never | +0.007 |
permission | +0.007 |
known | +0.007 |
vitamins | +0.007 |
Neg logit contributions | |
ing | -0.019 |
he | -0.017 |
ane | -0.017 |
ays | -0.016 |
ement | -0.016 |
Token | her |
Token ID | 607 |
Feature activation | +1.54e-02 |
Pos logit contributions | |
ing | +0.025 |
ane | +0.022 |
ays | +0.021 |
eddy | +0.021 |
he | +0.021 |
Neg logit contributions | |
forgot | -0.015 |
never | -0.015 |
wasted | -0.014 |
forgotten | -0.014 |
stolen | -0.013 |
Token | to |
Token ID | 284 |
Feature activation | +5.27e-02 |
Pos logit contributions | |
never | +0.024 |
forgot | +0.023 |
forgotten | +0.022 |
not | +0.022 |
help | +0.022 |
Neg logit contributions | |
ing | -0.036 |
ane | -0.035 |
ays | -0.034 |
eddy | -0.033 |
he | -0.033 |
Token | touch |
Token ID | 3638 |
Feature activation | -2.67e-02 |
Pos logit contributions | |
forgot | +0.065 |
wasted | +0.064 |
never | +0.063 |
known | +0.062 |
wished | +0.062 |
Neg logit contributions | |
ing | -0.128 |
ane | -0.115 |
he | -0.115 |
ays | -0.109 |
ite | -0.107 |
Token | the |
Token ID | 262 |
Feature activation | -2.65e-03 |
Pos logit contributions | |
ing | +0.076 |
ane | +0.074 |
ays | +0.072 |
eddy | +0.071 |
he | +0.070 |
Neg logit contributions | |
never | -0.028 |
help | -0.024 |
not | -0.024 |
forgot | -0.024 |
forgotten | -0.024 |
Token | ic |
Token ID | 14158 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
ing | +0.007 |
ane | +0.006 |
ays | +0.006 |
he | +0.006 |
eddy | +0.006 |
Neg logit contributions | |
never | -0.003 |
forgot | -0.003 |
forgotten | -0.003 |
wasted | -0.003 |
known | -0.003 |
Token | icle |
Token ID | 1548 |
Feature activation | -6.96e-02 |
Pos logit contributions | |
ane | +0.199 |
over | +0.192 |
ing | +0.190 |
he | +0.188 |
apping | +0.187 |
Neg logit contributions | |
never | -0.216 |
not | -0.211 |
help | -0.207 |
be | -0.205 |
permission | -0.204 |
Token | , |
Token ID | 11 |
Feature activation | +1.96e-02 |
Pos logit contributions | |
ane | +0.141 |
ays | +0.139 |
eddy | +0.136 |
ing | +0.128 |
ement | +0.127 |
Neg logit contributions | |
not | -0.136 |
never | -0.131 |
help | -0.123 |
be | -0.121 |
always | -0.113 |
Token | so |
Token ID | 523 |
Feature activation | +7.07e-03 |
Pos logit contributions | |
known | +0.043 |
forgot | +0.043 |
wasted | +0.042 |
forgotten | +0.042 |
unknow | +0.042 |
Neg logit contributions | |
ing | -0.030 |
ays | -0.025 |
ane | -0.023 |
he | -0.022 |
ement | -0.021 |
Token | she |
Token ID | 673 |
Feature activation | +5.73e-02 |
Pos logit contributions | |
never | +0.011 |
forgot | +0.011 |
forgotten | +0.011 |
permission | +0.011 |
unknow | +0.011 |
Neg logit contributions | |
ing | -0.016 |
ays | -0.014 |
ane | -0.014 |
he | -0.014 |
eddy | -0.013 |
Token | did |
Token ID | 750 |
Feature activation | +1.59e-02 |
Pos logit contributions | |
permission | +0.071 |
fact | +0.067 |
known | +0.067 |
need | +0.065 |
supposed | +0.064 |
Neg logit contributions | |
ing | -0.146 |
ays | -0.125 |
ane | -0.125 |
ed | -0.124 |
he | -0.124 |
Token | he |
Token ID | 339 |
Feature activation | +5.05e-02 |
Pos logit contributions | |
unknow | +0.061 |
shown | +0.061 |
known | +0.060 |
tresp | +0.060 |
forgot | +0.060 |
Neg logit contributions | |
ing | -0.071 |
he | -0.059 |
ane | -0.059 |
ays | -0.058 |
eddy | -0.057 |
Token | accidentally |
Token ID | 14716 |
Feature activation | +3.48e-02 |
Pos logit contributions | |
permission | +0.063 |
help | +0.060 |
known | +0.059 |
setup | +0.058 |
unknow | +0.057 |
Neg logit contributions | |
ing | -0.130 |
he | -0.117 |
ane | -0.116 |
ays | -0.116 |
eddy | -0.113 |
Token | knocked |
Token ID | 13642 |
Feature activation | -2.97e-02 |
Pos logit contributions | |
known | +0.038 |
fact | +0.035 |
robbers | +0.034 |
setup | +0.034 |
never | +0.034 |
Neg logit contributions | |
ing | -0.091 |
ays | -0.084 |
eddy | -0.078 |
ane | -0.078 |
he | -0.077 |
Token | over |
Token ID | 625 |
Feature activation | -8.39e-03 |
Pos logit contributions | |
ane | +0.068 |
ing | +0.066 |
ays | +0.063 |
ement | +0.062 |
eddy | +0.062 |
Neg logit contributions | |
never | -0.049 |
forgotten | -0.044 |
not | -0.044 |
been | -0.043 |
help | -0.042 |
Token | a |
Token ID | 257 |
Feature activation | +6.47e-03 |
Pos logit contributions | |
ing | +0.023 |
ane | +0.022 |
ays | +0.022 |
eddy | +0.021 |
he | +0.021 |
Neg logit contributions | |
never | -0.010 |
forgot | -0.009 |
forgotten | -0.009 |
not | -0.009 |
help | -0.009 |
Token | v |
Token ID | 410 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
never | +0.010 |
forgot | +0.009 |
forgotten | +0.009 |
known | +0.009 |
permission | +0.009 |
Neg logit contributions | |
ing | -0.016 |
ane | -0.014 |
ays | -0.014 |
he | -0.014 |
eddy | -0.014 |
Token | ase |
Token ID | 589 |
Feature activation | -7.93e-02 |
Pos logit contributions | |
ing | +0.115 |
ane | +0.092 |
ays | +0.092 |
ement | +0.089 |
ogging | +0.068 |
Neg logit contributions | |
never | -0.335 |
not | -0.317 |
help | -0.316 |
grown | -0.309 |
forgotten | -0.309 |
Token | . |
Token ID | 13 |
Feature activation | -3.82e-02 |
Pos logit contributions | |
ane | +0.148 |
ing | +0.146 |
eddy | +0.139 |
ays | +0.138 |
ement | +0.137 |
Neg logit contributions | |
never | -0.161 |
not | -0.152 |
forgotten | -0.145 |
help | -0.144 |
knocked | -0.144 |
Token | The |
Token ID | 383 |
Feature activation | +4.85e-03 |
Pos logit contributions | |
ing | +0.075 |
ane | +0.070 |
ays | +0.068 |
eddy | +0.065 |
he | +0.064 |
Neg logit contributions | |
never | -0.074 |
forgot | -0.072 |
forgotten | -0.070 |
not | -0.069 |
help | -0.068 |
Token | v |
Token ID | 410 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
forgot | +0.009 |
known | +0.009 |
never | +0.008 |
forgotten | +0.008 |
unknow | +0.008 |
Neg logit contributions | |
ing | -0.010 |
he | -0.009 |
eddy | -0.009 |
ane | -0.009 |
ays | -0.009 |
Token | ase |
Token ID | 589 |
Feature activation | -1.54e-02 |
Pos logit contributions | |
ing | +0.095 |
ane | +0.081 |
ement | +0.080 |
ays | +0.079 |
ogging | +0.066 |
Neg logit contributions | |
never | -0.280 |
not | -0.267 |
forgot | -0.265 |
grown | -0.260 |
wasted | -0.258 |
Token | , |
Token ID | 11 |
Feature activation | -1.85e-02 |
Pos logit contributions | |
ays | +0.104 |
ing | +0.104 |
eddy | +0.103 |
ane | +0.099 |
he | +0.092 |
Neg logit contributions | |
not | -0.086 |
never | -0.084 |
help | -0.077 |
be | -0.076 |
been | -0.075 |
Token | there |
Token ID | 612 |
Feature activation | -3.46e-02 |
Pos logit contributions | |
ane | +0.044 |
ing | +0.044 |
ays | +0.043 |
eddy | +0.042 |
ement | +0.042 |
Neg logit contributions | |
never | -0.030 |
not | -0.029 |
also | -0.027 |
forgot | -0.027 |
been | -0.026 |
Token | was |
Token ID | 373 |
Feature activation | +3.38e-04 |
Pos logit contributions | |
ing | +0.094 |
ane | +0.091 |
ays | +0.090 |
eddy | +0.086 |
ement | +0.083 |
Neg logit contributions | |
never | -0.044 |
forgot | -0.039 |
not | -0.038 |
been | -0.037 |
knocked | -0.036 |
Token | a |
Token ID | 257 |
Feature activation | -1.53e-02 |
Pos logit contributions | |
never | +0.000 |
not | +0.000 |
forgot | +0.000 |
forgotten | +0.000 |
been | +0.000 |
Neg logit contributions | |
ing | -0.001 |
ane | -0.001 |
ays | -0.001 |
eddy | -0.001 |
ement | -0.001 |
Token | big |
Token ID | 1263 |
Feature activation | -2.00e-02 |
Pos logit contributions | |
ing | +0.040 |
ane | +0.038 |
ays | +0.036 |
eddy | +0.036 |
he | +0.036 |
Neg logit contributions | |
never | -0.019 |
forgot | -0.018 |
forgotten | -0.018 |
permission | -0.017 |
wasted | -0.017 |
Token | v |
Token ID | 410 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
ing | +0.042 |
ane | +0.040 |
ays | +0.038 |
eddy | +0.038 |
he | +0.037 |
Neg logit contributions | |
never | -0.035 |
forgot | -0.035 |
forgotten | -0.034 |
known | -0.033 |
wasted | -0.032 |
Token | ase |
Token ID | 589 |
Feature activation | -8.20e-02 |
Pos logit contributions | |
ing | +0.113 |
ays | +0.111 |
ement | +0.098 |
ane | +0.080 |
ogging | +0.067 |
Neg logit contributions | |
never | -0.350 |
not | -0.329 |
forgot | -0.314 |
always | -0.313 |
forgotten | -0.310 |
Token | in |
Token ID | 287 |
Feature activation | +1.01e-02 |
Pos logit contributions | |
ays | +0.166 |
ing | +0.161 |
ement | +0.156 |
ane | +0.154 |
eddy | +0.145 |
Neg logit contributions | |
never | -0.166 |
not | -0.152 |
always | -0.146 |
been | -0.138 |
help | -0.136 |
Token | a |
Token ID | 257 |
Feature activation | +4.19e-03 |
Pos logit contributions | |
forgot | +0.018 |
forgotten | +0.015 |
stolen | +0.015 |
permission | +0.015 |
wasted | +0.014 |
Neg logit contributions | |
ing | -0.021 |
he | -0.021 |
ays | -0.020 |
ane | -0.020 |
eddy | -0.019 |
Token | small |
Token ID | 1402 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
forgot | +0.006 |
never | +0.006 |
permission | +0.006 |
forgotten | +0.006 |
known | +0.006 |
Neg logit contributions | |
ing | -0.010 |
ane | -0.009 |
eddy | -0.009 |
he | -0.009 |
ays | -0.009 |
Token | house |
Token ID | 2156 |
Feature activation | -7.22e-02 |
Pos logit contributions | |
ing | +0.030 |
ane | +0.028 |
ays | +0.027 |
eddy | +0.026 |
he | +0.026 |
Neg logit contributions | |
never | -0.025 |
forgot | -0.025 |
forgotten | -0.024 |
not | -0.024 |
known | -0.023 |
Token | her |
Token ID | 607 |
Feature activation | +4.50e-02 |
Pos logit contributions | |
unknow | +0.054 |
tresp | +0.051 |
forgot | +0.050 |
wasted | +0.050 |
forgotten | +0.050 |
Neg logit contributions | |
ing | -0.054 |
ays | -0.047 |
ane | -0.045 |
eddy | -0.044 |
he | -0.044 |
Token | best |
Token ID | 1266 |
Feature activation | +5.33e-03 |
Pos logit contributions | |
never | +0.094 |
known | +0.094 |
forgot | +0.093 |
unknow | +0.093 |
forgotten | +0.091 |
Neg logit contributions | |
ing | -0.070 |
ane | -0.068 |
eddy | -0.067 |
ays | -0.064 |
ay | -0.062 |
Token | toy |
Token ID | 13373 |
Feature activation | -2.25e-02 |
Pos logit contributions | |
never | +0.008 |
forgot | +0.008 |
forgotten | +0.007 |
not | +0.007 |
wasted | +0.007 |
Neg logit contributions | |
ing | -0.013 |
ane | -0.012 |
ays | -0.012 |
eddy | -0.011 |
he | -0.011 |
Token | was |
Token ID | 373 |
Feature activation | +3.22e-02 |
Pos logit contributions | |
ing | +0.050 |
ane | +0.047 |
ays | +0.046 |
eddy | +0.044 |
he | +0.043 |
Neg logit contributions | |
never | -0.037 |
forgot | -0.037 |
forgotten | -0.035 |
wasted | -0.034 |
not | -0.034 |
Token | a |
Token ID | 257 |
Feature activation | +2.14e-02 |
Pos logit contributions | |
forgot | +0.051 |
unknow | +0.051 |
think | +0.050 |
wait | +0.050 |
forget | +0.049 |
Neg logit contributions | |
ing | -0.068 |
eddy | -0.060 |
ays | -0.058 |
ane | -0.058 |
he | -0.054 |
Token | crocod |
Token ID | 37565 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
permission | +0.035 |
known | +0.033 |
never | +0.032 |
unknow | +0.032 |
wasted | +0.031 |
Neg logit contributions | |
ing | -0.045 |
eddy | -0.044 |
ane | -0.042 |
he | -0.040 |
ay | -0.040 |
Token | ile |
Token ID | 576 |
Feature activation | -8.73e-02 |
Pos logit contributions | |
ays | +0.256 |
ing | +0.252 |
ane | +0.233 |
ement | +0.232 |
eddy | +0.224 |
Neg logit contributions | |
never | -0.186 |
help | -0.186 |
always | -0.170 |
tresp | -0.169 |
been | -0.167 |
Token | . |
Token ID | 13 |
Feature activation | -4.36e-02 |
Pos logit contributions | |
ing | +0.154 |
ane | +0.147 |
ays | +0.146 |
ement | +0.135 |
he | +0.135 |
Neg logit contributions | |
never | -0.188 |
not | -0.174 |
forgot | -0.174 |
help | -0.170 |
forgotten | -0.170 |
Token | It |
Token ID | 632 |
Feature activation | +2.76e-02 |
Pos logit contributions | |
ing | +0.118 |
ane | +0.110 |
ays | +0.102 |
ement | +0.099 |
he | +0.098 |
Neg logit contributions | |
grown | -0.053 |
help | -0.053 |
never | -0.052 |
not | -0.051 |
be | -0.051 |
Token | was |
Token ID | 373 |
Feature activation | +4.38e-02 |
Pos logit contributions | |
permission | +0.062 |
setup | +0.053 |
supervision | +0.052 |
unknow | +0.051 |
forgotten | +0.051 |
Neg logit contributions | |
ing | -0.048 |
ane | -0.044 |
ays | -0.044 |
he | -0.042 |
eddy | -0.042 |
Token | green |
Token ID | 4077 |
Feature activation | -6.89e-02 |
Pos logit contributions | |
forgot | +0.076 |
permission | +0.076 |
unknow | +0.075 |
wait | +0.073 |
forget | +0.070 |
Neg logit contributions | |
ing | -0.083 |
eddy | -0.074 |
ays | -0.073 |
ay | -0.068 |
he | -0.066 |
Token | his |
Token ID | 465 |
Feature activation | +5.80e-03 |
Pos logit contributions | |
forgot | +0.008 |
never | +0.008 |
forgotten | +0.007 |
known | +0.007 |
stolen | +0.007 |
Neg logit contributions | |
ing | -0.013 |
he | -0.012 |
eddy | -0.012 |
ays | -0.012 |
ane | -0.012 |
Token | room |
Token ID | 2119 |
Feature activation | -6.39e-02 |
Pos logit contributions | |
never | +0.009 |
forgot | +0.009 |
known | +0.009 |
forgotten | +0.009 |
permission | +0.008 |
Neg logit contributions | |
ing | -0.013 |
eddy | -0.012 |
ane | -0.012 |
ays | -0.012 |
he | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | +4.37e-02 |
Pos logit contributions | |
ing | +0.103 |
ane | +0.096 |
ays | +0.093 |
eddy | +0.088 |
he | +0.085 |
Neg logit contributions | |
never | -0.146 |
forgot | -0.141 |
forgotten | -0.139 |
not | -0.138 |
help | -0.135 |
Token | grabbed |
Token ID | 13176 |
Feature activation | +8.70e-03 |
Pos logit contributions | |
permission | +0.065 |
vitamins | +0.059 |
tresp | +0.058 |
fact | +0.057 |
known | +0.057 |
Neg logit contributions | |
ing | -0.093 |
ed | -0.079 |
he | -0.078 |
s | -0.077 |
ays | -0.076 |
Token | the |
Token ID | 262 |
Feature activation | -3.02e-03 |
Pos logit contributions | |
never | +0.013 |
forgot | +0.013 |
forgotten | +0.012 |
known | +0.012 |
not | +0.011 |
Neg logit contributions | |
ing | -0.021 |
ays | -0.020 |
ane | -0.019 |
he | -0.019 |
eddy | -0.019 |
Token | v |
Token ID | 410 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
ing | +0.007 |
ane | +0.006 |
ays | +0.006 |
he | +0.006 |
eddy | +0.006 |
Neg logit contributions | |
forgot | -0.005 |
never | -0.005 |
forgotten | -0.005 |
known | -0.005 |
wasted | -0.004 |
Token | ase |
Token ID | 589 |
Feature activation | -6.79e-02 |
Pos logit contributions | |
ing | +0.129 |
ement | +0.109 |
ays | +0.106 |
ane | +0.095 |
ogging | +0.076 |
Neg logit contributions | |
never | -0.318 |
not | -0.296 |
been | -0.289 |
help | -0.285 |
grown | -0.284 |
Token | . |
Token ID | 13 |
Feature activation | -3.56e-02 |
Pos logit contributions | |
ing | +0.121 |
ane | +0.120 |
ays | +0.119 |
ement | +0.110 |
eddy | +0.107 |
Neg logit contributions | |
never | -0.141 |
not | -0.135 |
forgotten | -0.132 |
help | -0.130 |
permission | -0.126 |
Token | He |
Token ID | 679 |
Feature activation | +4.84e-02 |
Pos logit contributions | |
ing | +0.067 |
ane | +0.061 |
ays | +0.061 |
eddy | +0.058 |
he | +0.057 |
Neg logit contributions | |
never | -0.072 |
forgot | -0.070 |
forgotten | -0.067 |
known | -0.067 |
permission | -0.066 |
Token | was |
Token ID | 373 |
Feature activation | +5.02e-02 |
Pos logit contributions | |
permission | +0.074 |
known | +0.064 |
help | +0.062 |
fact | +0.061 |
supervision | +0.061 |
Neg logit contributions | |
ing | -0.112 |
he | -0.097 |
eddy | -0.097 |
ays | -0.094 |
ed | -0.093 |
Token | going |
Token ID | 1016 |
Feature activation | -1.10e-02 |
Pos logit contributions | |
forgot | +0.084 |
permission | +0.078 |
forget | +0.078 |
fiber | +0.076 |
wait | +0.075 |
Neg logit contributions | |
ing | -0.100 |
he | -0.092 |
eddy | -0.087 |
ays | -0.086 |
apping | -0.085 |
Token | . |
Token ID | 13 |
Feature activation | -3.90e-02 |
Pos logit contributions | |
ane | +0.115 |
ing | +0.114 |
eddy | +0.110 |
ays | +0.110 |
he | +0.100 |
Neg logit contributions | |
never | -0.132 |
not | -0.126 |
grown | -0.120 |
help | -0.120 |
forgot | -0.118 |
Token | They |
Token ID | 1119 |
Feature activation | +6.40e-02 |
Pos logit contributions | |
ing | +0.089 |
ane | +0.084 |
ays | +0.078 |
eddy | +0.077 |
he | +0.076 |
Neg logit contributions | |
never | -0.064 |
forgot | -0.063 |
not | -0.061 |
grown | -0.060 |
forgotten | -0.060 |
Token | see |
Token ID | 766 |
Feature activation | -5.43e-04 |
Pos logit contributions | |
permission | +0.099 |
setup | +0.096 |
unknow | +0.094 |
warning | +0.091 |
pap | +0.091 |
Neg logit contributions | |
he | -0.140 |
ane | -0.127 |
ing | -0.126 |
room | -0.118 |
ite | -0.118 |
Token | a |
Token ID | 257 |
Feature activation | +9.23e-03 |
Pos logit contributions | |
ing | +0.002 |
ane | +0.001 |
ays | +0.001 |
eddy | +0.001 |
he | +0.001 |
Neg logit contributions | |
never | -0.001 |
forgot | -0.001 |
forgotten | -0.001 |
not | -0.001 |
wasted | -0.001 |
Token | big |
Token ID | 1263 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
never | +0.012 |
forgot | +0.012 |
permission | +0.011 |
forgotten | +0.011 |
known | +0.011 |
Neg logit contributions | |
ing | -0.024 |
ane | -0.022 |
eddy | -0.021 |
ays | -0.021 |
he | -0.021 |
Token | v |
Token ID | 410 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
ing | +0.032 |
ane | +0.030 |
ays | +0.029 |
eddy | +0.028 |
he | +0.028 |
Neg logit contributions | |
never | -0.024 |
forgot | -0.023 |
forgotten | -0.022 |
known | -0.022 |
not | -0.021 |
Token | ase |
Token ID | 589 |
Feature activation | -7.38e-02 |
Pos logit contributions | |
ing | +0.120 |
ays | +0.102 |
ement | +0.100 |
ane | +0.084 |
ogging | +0.071 |
Neg logit contributions | |
never | -0.331 |
not | -0.313 |
help | -0.298 |
forgot | -0.296 |
always | -0.295 |
Token | on |
Token ID | 319 |
Feature activation | -5.13e-03 |
Pos logit contributions | |
eddy | +0.132 |
ing | +0.128 |
ays | +0.128 |
ane | +0.128 |
ement | +0.123 |
Neg logit contributions | |
never | -0.156 |
not | -0.152 |
also | -0.138 |
always | -0.137 |
been | -0.135 |
Token | the |
Token ID | 262 |
Feature activation | -1.66e-02 |
Pos logit contributions | |
ing | +0.014 |
ane | +0.013 |
ays | +0.013 |
he | +0.013 |
eddy | +0.012 |
Neg logit contributions | |
forgot | -0.006 |
never | -0.006 |
forgotten | -0.006 |
stolen | -0.005 |
known | -0.005 |
Token | table |
Token ID | 3084 |
Feature activation | -6.92e-02 |
Pos logit contributions | |
ing | +0.039 |
ane | +0.038 |
he | +0.037 |
eddy | +0.037 |
ays | +0.036 |
Neg logit contributions | |
forgot | -0.025 |
known | -0.023 |
forgotten | -0.023 |
never | -0.023 |
wasted | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | -5.11e-02 |
Pos logit contributions | |
ing | +0.122 |
ane | +0.114 |
eddy | +0.112 |
ays | +0.109 |
ement | +0.100 |
Neg logit contributions | |
never | -0.151 |
not | -0.144 |
forgot | -0.138 |
grown | -0.137 |
forgotten | -0.137 |
Token | carrying |
Token ID | 6872 |
Feature activation | -1.08e-02 |
Pos logit contributions | |
tresp | +0.056 |
permission | +0.054 |
unknow | +0.054 |
least | +0.053 |
punish | +0.052 |
Neg logit contributions | |
ing | -0.107 |
ed | -0.084 |
he | -0.083 |
eddy | -0.079 |
ore | -0.079 |
Token | it |
Token ID | 340 |
Feature activation | -1.82e-02 |
Pos logit contributions | |
ing | +0.025 |
ays | +0.021 |
ane | +0.021 |
ay | +0.020 |
he | +0.020 |
Neg logit contributions | |
forgot | -0.018 |
never | -0.017 |
forgotten | -0.016 |
forget | -0.016 |
known | -0.016 |
Token | back |
Token ID | 736 |
Feature activation | -1.88e-03 |
Pos logit contributions | |
ing | +0.032 |
ane | +0.029 |
ays | +0.028 |
he | +0.027 |
eddy | +0.027 |
Neg logit contributions | |
never | -0.038 |
forgot | -0.038 |
forgotten | -0.037 |
known | -0.036 |
wasted | -0.036 |
Token | to |
Token ID | 284 |
Feature activation | +3.34e-02 |
Pos logit contributions | |
ing | +0.004 |
ane | +0.004 |
ays | +0.004 |
eddy | +0.004 |
he | +0.004 |
Neg logit contributions | |
never | -0.003 |
forgot | -0.003 |
forgotten | -0.003 |
not | -0.003 |
help | -0.003 |
Token | the |
Token ID | 262 |
Feature activation | +2.20e-02 |
Pos logit contributions | |
forgot | +0.061 |
forgotten | +0.053 |
stolen | +0.052 |
wasted | +0.052 |
knocked | +0.052 |
Neg logit contributions | |
ing | -0.062 |
he | -0.059 |
room | -0.054 |
ore | -0.053 |
ay | -0.053 |
Token | anth |
Token ID | 26794 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
forgot | +0.037 |
never | +0.035 |
wasted | +0.034 |
knocked | +0.033 |
forgotten | +0.033 |
Neg logit contributions | |
ing | -0.050 |
ane | -0.044 |
eddy | -0.043 |
he | -0.041 |
ays | -0.040 |
Token | ill |
Token ID | 359 |
Feature activation | -3.53e-02 |
Pos logit contributions | |
ing | +0.185 |
ane | +0.164 |
eddy | +0.164 |
ement | +0.163 |
ays | +0.160 |
Neg logit contributions | |
never | -0.271 |
forgot | -0.255 |
forgotten | -0.248 |
been | -0.242 |
not | -0.242 |
Token | . |
Token ID | 13 |
Feature activation | -5.45e-02 |
Pos logit contributions | |
ing | +0.061 |
ane | +0.058 |
ays | +0.058 |
eddy | +0.056 |
ement | +0.054 |
Neg logit contributions | |
never | -0.077 |
forgot | -0.072 |
grown | -0.070 |
not | -0.069 |
help | -0.069 |
Token |
|
Token ID | 198 |
Feature activation | -6.13e-02 |
Pos logit contributions | |
ing | +0.135 |
ays | +0.123 |
ane | +0.122 |
eddy | +0.118 |
he | +0.114 |
Neg logit contributions | |
never | -0.085 |
help | -0.077 |
not | -0.077 |
always | -0.075 |
grown | -0.075 |
Token |
|
Token ID | 198 |
Feature activation | -4.25e-02 |
Pos logit contributions | |
ing | +0.152 |
ays | +0.152 |
ane | +0.142 |
ement | +0.138 |
ipping | +0.133 |
Neg logit contributions | |
never | -0.097 |
not | -0.090 |
know | -0.082 |
help | -0.081 |
also | -0.081 |
Token | One |
Token ID | 3198 |
Feature activation | +8.94e-03 |
Pos logit contributions | |
ing | +0.105 |
ays | +0.100 |
ane | +0.098 |
eddy | +0.094 |
ement | +0.091 |
Neg logit contributions | |
never | -0.066 |
not | -0.059 |
forgot | -0.058 |
forgotten | -0.057 |
grown | -0.057 |
Token | with |
Token ID | 351 |
Feature activation | +2.70e-03 |
Pos logit contributions | |
ane | +0.093 |
ing | +0.090 |
ays | +0.086 |
eddy | +0.082 |
umbled | +0.081 |
Neg logit contributions | |
help | -0.087 |
grown | -0.081 |
not | -0.081 |
never | -0.079 |
put | -0.078 |
Token | her |
Token ID | 607 |
Feature activation | +1.45e-02 |
Pos logit contributions | |
never | +0.004 |
forgot | +0.003 |
forgotten | +0.003 |
wasted | +0.003 |
known | +0.003 |
Neg logit contributions | |
ing | -0.007 |
ane | -0.007 |
ays | -0.007 |
eddy | -0.006 |
he | -0.006 |
Token | mom |
Token ID | 1995 |
Feature activation | -3.12e-02 |
Pos logit contributions | |
never | +0.020 |
forgot | +0.020 |
forgotten | +0.019 |
not | +0.018 |
wasted | +0.018 |
Neg logit contributions | |
ing | -0.036 |
ane | -0.034 |
ays | -0.033 |
eddy | -0.032 |
he | -0.032 |
Token | 's |
Token ID | 338 |
Feature activation | -6.93e-03 |
Pos logit contributions | |
ane | +0.067 |
ing | +0.065 |
ays | +0.064 |
ement | +0.063 |
umbled | +0.060 |
Neg logit contributions | |
never | -0.052 |
always | -0.051 |
not | -0.050 |
be | -0.049 |
been | -0.049 |
Token | large |
Token ID | 1588 |
Feature activation | -5.95e-03 |
Pos logit contributions | |
ing | +0.018 |
ays | +0.017 |
ane | +0.017 |
eddy | +0.016 |
he | +0.016 |
Neg logit contributions | |
never | -0.009 |
help | -0.009 |
grown | -0.009 |
forgot | -0.009 |
not | -0.008 |
Token | v |
Token ID | 410 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
ing | +0.014 |
ane | +0.013 |
ays | +0.013 |
eddy | +0.012 |
he | +0.012 |
Neg logit contributions | |
never | -0.009 |
forgot | -0.009 |
forgotten | -0.009 |
not | -0.008 |
known | -0.008 |
Token | ase |
Token ID | 589 |
Feature activation | -7.11e-02 |
Pos logit contributions | |
ing | +0.135 |
ement | +0.118 |
ays | +0.116 |
ane | +0.092 |
rily | +0.091 |
Neg logit contributions | |
never | -0.301 |
not | -0.288 |
been | -0.272 |
help | -0.271 |
always | -0.269 |
Token | . |
Token ID | 13 |
Feature activation | -4.14e-02 |
Pos logit contributions | |
ing | +0.143 |
ays | +0.140 |
ane | +0.138 |
eddy | +0.132 |
ement | +0.130 |
Neg logit contributions | |
never | -0.138 |
not | -0.133 |
help | -0.129 |
always | -0.125 |
forgot | -0.119 |
Token | It |
Token ID | 632 |
Feature activation | +3.30e-02 |
Pos logit contributions | |
ing | +0.092 |
ane | +0.086 |
ays | +0.081 |
he | +0.080 |
eddy | +0.077 |
Neg logit contributions | |
never | -0.071 |
not | -0.068 |
grown | -0.068 |
help | -0.067 |
forgot | -0.067 |
Token | was |
Token ID | 373 |
Feature activation | +3.98e-02 |
Pos logit contributions | |
permission | +0.068 |
setup | +0.059 |
unknow | +0.057 |
forgotten | +0.057 |
supervision | +0.056 |
Neg logit contributions | |
ing | -0.062 |
ane | -0.059 |
eddy | -0.056 |
ays | -0.056 |
he | -0.056 |
Token | blue |
Token ID | 4171 |
Feature activation | -6.59e-02 |
Pos logit contributions | |
forgot | +0.065 |
permission | +0.064 |
unknow | +0.064 |
wasted | +0.059 |
regret | +0.058 |
Neg logit contributions | |
ing | -0.083 |
eddy | -0.076 |
ays | -0.075 |
ay | -0.072 |
he | -0.072 |
Token | , |
Token ID | 11 |
Feature activation | +3.14e-02 |
Pos logit contributions | |
ing | +0.017 |
ane | +0.016 |
ays | +0.015 |
he | +0.015 |
eddy | +0.015 |
Neg logit contributions | |
never | -0.015 |
forgot | -0.014 |
forgotten | -0.014 |
wasted | -0.013 |
known | -0.013 |
Token | even |
Token ID | 772 |
Feature activation | +4.79e-02 |
Pos logit contributions | |
forgot | +0.063 |
known | +0.062 |
wasted | +0.061 |
stolen | +0.061 |
forgotten | +0.061 |
Neg logit contributions | |
ing | -0.054 |
ays | -0.045 |
ane | -0.045 |
he | -0.044 |
eddy | -0.042 |
Token | though |
Token ID | 996 |
Feature activation | +9.08e-03 |
Pos logit contributions | |
never | +0.102 |
forgot | +0.100 |
forgotten | +0.097 |
wasted | +0.095 |
stolen | +0.095 |
Neg logit contributions | |
ing | -0.085 |
ane | -0.072 |
ays | -0.071 |
he | -0.069 |
eddy | -0.068 |
Token | she |
Token ID | 673 |
Feature activation | +6.26e-02 |
Pos logit contributions | |
forgot | +0.014 |
never | +0.014 |
forgotten | +0.013 |
stolen | +0.013 |
wasted | +0.013 |
Neg logit contributions | |
ing | -0.020 |
ane | -0.018 |
ays | -0.018 |
he | -0.017 |
eddy | -0.017 |
Token | didn |
Token ID | 1422 |
Feature activation | -9.22e-03 |
Pos logit contributions | |
permission | +0.084 |
robbers | +0.080 |
known | +0.080 |
setup | +0.078 |
help | +0.077 |
Neg logit contributions | |
ing | -0.145 |
ane | -0.136 |
he | -0.133 |
ays | -0.126 |
ement | -0.125 |
Token | 't |
Token ID | 470 |
Feature activation | +4.27e-02 |
Pos logit contributions | |
ing | +0.025 |
ays | +0.023 |
ane | +0.023 |
eddy | +0.023 |
ement | +0.022 |
Neg logit contributions | |
never | -0.012 |
not | -0.011 |
help | -0.010 |
forgotten | -0.010 |
been | -0.009 |
Token | want |
Token ID | 765 |
Feature activation | -2.27e-03 |
Pos logit contributions | |
unknow | +0.039 |
stolen | +0.039 |
forgot | +0.038 |
known | +0.038 |
setup | +0.037 |
Neg logit contributions | |
ing | -0.117 |
he | -0.114 |
ane | -0.111 |
eddy | -0.102 |
ite | -0.101 |
Token | to |
Token ID | 284 |
Feature activation | +4.93e-02 |
Pos logit contributions | |
ing | +0.005 |
ane | +0.004 |
ays | +0.004 |
eddy | +0.004 |
he | +0.004 |
Neg logit contributions | |
forgot | -0.004 |
never | -0.004 |
wasted | -0.004 |
known | -0.004 |
forgotten | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -3.82e-02 |
Pos logit contributions | |
forgot | +0.049 |
never | +0.049 |
forgotten | +0.045 |
stolen | +0.044 |
wasted | +0.044 |
Neg logit contributions | |
ing | -0.138 |
ane | -0.131 |
he | -0.128 |
ays | -0.127 |
eddy | -0.125 |
Token | |
Token ID | 220 |
Feature activation | -3.60e-02 |
Pos logit contributions | |
ing | +0.081 |
ane | +0.077 |
ays | +0.074 |
eddy | +0.072 |
he | +0.070 |
Neg logit contributions | |
never | -0.069 |
forgot | -0.066 |
forgotten | -0.064 |
not | -0.064 |
grown | -0.062 |
Token |
|
Token ID | 198 |
Feature activation | -4.40e-02 |
Pos logit contributions | |
ing | +0.075 |
ane | +0.073 |
ays | +0.070 |
eddy | +0.069 |
ement | +0.067 |
Neg logit contributions | |
not | -0.062 |
never | -0.062 |
forgot | -0.056 |
been | -0.055 |
grown | -0.055 |
Token | , |
Token ID | 11 |
Feature activation | +1.54e-02 |
Pos logit contributions | |
ing | +0.069 |
ane | +0.065 |
ays | +0.064 |
eddy | +0.060 |
he | +0.059 |
Neg logit contributions | |
never | -0.095 |
forgot | -0.093 |
forgotten | -0.090 |
not | -0.089 |
help | -0.088 |
Token | she |
Token ID | 673 |
Feature activation | +6.55e-02 |
Pos logit contributions | |
never | +0.026 |
forgot | +0.026 |
forgotten | +0.025 |
wasted | +0.024 |
stolen | +0.024 |
Neg logit contributions | |
ing | -0.034 |
ane | -0.031 |
ays | -0.030 |
he | -0.029 |
eddy | -0.029 |
Token | opened |
Token ID | 4721 |
Feature activation | +1.01e-02 |
Pos logit contributions | |
permission | +0.095 |
help | +0.086 |
tresp | +0.085 |
supposed | +0.084 |
vitamins | +0.083 |
Neg logit contributions | |
ing | -0.156 |
ane | -0.136 |
ays | -0.135 |
he | -0.131 |
ed | -0.130 |
Token | the |
Token ID | 262 |
Feature activation | +9.84e-03 |
Pos logit contributions | |
never | +0.013 |
forgot | +0.013 |
forgotten | +0.012 |
wasted | +0.011 |
not | +0.011 |
Neg logit contributions | |
ing | -0.026 |
ane | -0.025 |
ays | -0.025 |
eddy | -0.024 |
he | -0.024 |
Token | box |
Token ID | 3091 |
Feature activation | -5.17e-02 |
Pos logit contributions | |
never | +0.018 |
forgot | +0.018 |
known | +0.017 |
not | +0.017 |
forgotten | +0.016 |
Neg logit contributions | |
ing | -0.021 |
ane | -0.019 |
ays | -0.018 |
he | -0.018 |
eddy | -0.018 |
Token | and |
Token ID | 290 |
Feature activation | +6.13e-02 |
Pos logit contributions | |
ing | +0.090 |
ane | +0.088 |
ays | +0.085 |
eddy | +0.083 |
he | +0.076 |
Neg logit contributions | |
never | -0.109 |
not | -0.102 |
forgotten | -0.102 |
forgot | -0.101 |
help | -0.098 |
Token | put |
Token ID | 1234 |
Feature activation | -7.38e-03 |
Pos logit contributions | |
tresp | +0.074 |
permission | +0.072 |
never | +0.072 |
fact | +0.071 |
supposed | +0.071 |
Neg logit contributions | |
ing | -0.161 |
ays | -0.140 |
he | -0.139 |
ed | -0.138 |
ane | -0.137 |
Token | a |
Token ID | 257 |
Feature activation | +3.53e-02 |
Pos logit contributions | |
ing | +0.018 |
ays | +0.017 |
ane | +0.017 |
he | +0.016 |
eddy | +0.016 |
Neg logit contributions | |
forgot | -0.011 |
never | -0.011 |
forgotten | -0.010 |
wasted | -0.010 |
not | -0.009 |
Token | mint |
Token ID | 21043 |
Feature activation | -6.03e-02 |
Pos logit contributions | |
never | +0.053 |
forgot | +0.049 |
known | +0.049 |
permission | +0.048 |
forgotten | +0.047 |
Neg logit contributions | |
ing | -0.084 |
eddy | -0.077 |
he | -0.077 |
ane | -0.075 |
ay | -0.074 |
Token | in |
Token ID | 287 |
Feature activation | +1.74e-02 |
Pos logit contributions | |
ing | +0.118 |
ane | +0.117 |
ays | +0.112 |
he | +0.106 |
ement | +0.105 |
Neg logit contributions | |
never | -0.110 |
forgot | -0.106 |
help | -0.104 |
forgotten | -0.103 |
grown | -0.101 |
Token | her |
Token ID | 607 |
Feature activation | +1.37e-02 |
Pos logit contributions | |
forgot | +0.026 |
never | +0.024 |
wasted | +0.023 |
forgotten | +0.023 |
stolen | +0.022 |
Neg logit contributions | |
ing | -0.041 |
he | -0.039 |
ays | -0.039 |
eddy | -0.038 |
ane | -0.038 |
Token | yellow |
Token ID | 7872 |
Feature activation | -3.04e-02 |
Pos logit contributions | |
never | +0.087 |
forgotten | +0.086 |
stolen | +0.084 |
permission | +0.083 |
forgot | +0.083 |
Neg logit contributions | |
ing | -0.101 |
ays | -0.096 |
ane | -0.090 |
ay | -0.084 |
eddy | -0.084 |
Token | . |
Token ID | 13 |
Feature activation | -4.29e-02 |
Pos logit contributions | |
ing | +0.058 |
ane | +0.056 |
he | +0.054 |
eddy | +0.053 |
ays | +0.051 |
Neg logit contributions | |
help | -0.057 |
never | -0.057 |
be | -0.057 |
not | -0.055 |
grown | -0.055 |
Token | Ben |
Token ID | 3932 |
Feature activation | +3.28e-02 |
Pos logit contributions | |
ing | +0.099 |
ane | +0.098 |
ement | +0.092 |
he | +0.092 |
eddy | +0.089 |
Neg logit contributions | |
not | -0.067 |
never | -0.066 |
grown | -0.064 |
help | -0.063 |
always | -0.058 |
Token | picks |
Token ID | 11103 |
Feature activation | -1.91e-02 |
Pos logit contributions | |
forgotten | +0.051 |
never | +0.051 |
stolen | +0.050 |
known | +0.050 |
permission | +0.050 |
Neg logit contributions | |
ing | -0.075 |
ays | -0.072 |
ane | -0.069 |
he | -0.064 |
eddy | -0.063 |
Token | blue |
Token ID | 4171 |
Feature activation | -3.58e-02 |
Pos logit contributions | |
ane | +0.045 |
ing | +0.043 |
he | +0.042 |
eddy | +0.042 |
ays | +0.042 |
Neg logit contributions | |
never | -0.027 |
not | -0.026 |
grown | -0.025 |
help | -0.024 |
be | -0.024 |
Token | and |
Token ID | 290 |
Feature activation | +4.93e-02 |
Pos logit contributions | |
ing | +0.071 |
he | +0.069 |
ane | +0.069 |
eddy | +0.066 |
over | +0.066 |
Neg logit contributions | |
be | -0.062 |
help | -0.062 |
grown | -0.056 |
never | -0.055 |
permission | -0.055 |
Token | green |
Token ID | 4077 |
Feature activation | -3.76e-02 |
Pos logit contributions | |
never | +0.089 |
forgotten | +0.085 |
forgot | +0.083 |
stolen | +0.082 |
permission | +0.082 |
Neg logit contributions | |
ing | -0.106 |
ays | -0.100 |
ane | -0.095 |
eddy | -0.090 |
ay | -0.086 |
Token | and |
Token ID | 290 |
Feature activation | +4.82e-02 |
Pos logit contributions | |
ane | +0.079 |
ing | +0.077 |
he | +0.076 |
eddy | +0.075 |
ays | +0.070 |
Neg logit contributions | |
help | -0.062 |
be | -0.062 |
never | -0.061 |
not | -0.060 |
grown | -0.059 |
Token | red |
Token ID | 2266 |
Feature activation | -3.76e-02 |
Pos logit contributions | |
never | +0.084 |
forgotten | +0.081 |
permission | +0.080 |
forgot | +0.079 |
stolen | +0.079 |
Neg logit contributions | |
ing | -0.104 |
ays | -0.098 |
ane | -0.093 |
eddy | -0.089 |
he | -0.085 |
Token | . |
Token ID | 13 |
Feature activation | -4.23e-02 |
Pos logit contributions | |
ane | +0.074 |
ing | +0.072 |
he | +0.071 |
eddy | +0.068 |
ays | +0.065 |
Neg logit contributions | |
help | -0.068 |
be | -0.067 |
never | -0.066 |
not | -0.064 |
grown | -0.064 |
Token | They |
Token ID | 1119 |
Feature activation | +7.45e-02 |
Pos logit contributions | |
ing | +0.098 |
ane | +0.096 |
ement | +0.089 |
ays | +0.088 |
he | +0.088 |
Neg logit contributions | |
not | -0.068 |
never | -0.067 |
grown | -0.066 |
help | -0.062 |
forgot | -0.061 |
Token | . |
Token ID | 13 |
Feature activation | -2.57e-02 |
Pos logit contributions | |
ing | +0.053 |
ane | +0.049 |
ays | +0.048 |
he | +0.046 |
eddy | +0.046 |
Neg logit contributions | |
never | -0.048 |
forgot | -0.048 |
forgotten | -0.045 |
wasted | -0.045 |
known | -0.045 |
Token | It |
Token ID | 632 |
Feature activation | +4.39e-02 |
Pos logit contributions | |
ing | +0.052 |
ane | +0.048 |
ays | +0.046 |
eddy | +0.045 |
he | +0.045 |
Neg logit contributions | |
never | -0.049 |
forgot | -0.047 |
not | -0.046 |
forgotten | -0.045 |
grown | -0.045 |
Token | protected |
Token ID | 6861 |
Feature activation | -2.82e-02 |
Pos logit contributions | |
permission | +0.073 |
robbers | +0.065 |
supposed | +0.064 |
tresp | +0.063 |
supervision | +0.063 |
Neg logit contributions | |
ing | -0.091 |
he | -0.082 |
ane | -0.080 |
ed | -0.076 |
ays | -0.076 |
Token | the |
Token ID | 262 |
Feature activation | +2.15e-02 |
Pos logit contributions | |
ing | +0.071 |
ane | +0.066 |
ays | +0.065 |
eddy | +0.063 |
he | +0.063 |
Neg logit contributions | |
never | -0.039 |
forgot | -0.038 |
forgotten | -0.036 |
wasted | -0.035 |
not | -0.035 |
Token | animals |
Token ID | 4695 |
Feature activation | -2.79e-02 |
Pos logit contributions | |
forgot | +0.037 |
never | +0.036 |
known | +0.035 |
forgotten | +0.035 |
wasted | +0.034 |
Neg logit contributions | |
ing | -0.047 |
ane | -0.042 |
eddy | -0.041 |
he | -0.040 |
ays | -0.040 |
Token | and |
Token ID | 290 |
Feature activation | +6.34e-02 |
Pos logit contributions | |
ing | +0.047 |
ane | +0.043 |
ays | +0.042 |
eddy | +0.040 |
he | +0.039 |
Neg logit contributions | |
never | -0.061 |
forgot | -0.060 |
forgotten | -0.058 |
not | -0.057 |
wasted | -0.057 |
Token | they |
Token ID | 484 |
Feature activation | +6.69e-02 |
Pos logit contributions | |
supposed | +0.074 |
unknow | +0.073 |
forgotten | +0.072 |
tresp | +0.072 |
permission | +0.072 |
Neg logit contributions | |
ing | -0.143 |
ed | -0.120 |
ment | -0.114 |
ane | -0.112 |
es | -0.111 |
Token | were |
Token ID | 547 |
Feature activation | +5.38e-02 |
Pos logit contributions | |
permission | +0.081 |
vitamins | +0.080 |
fact | +0.079 |
stolen | +0.079 |
setup | +0.078 |
Neg logit contributions | |
ing | -0.153 |
he | -0.149 |
ane | -0.147 |
ite | -0.135 |
ays | -0.131 |
Token | never |
Token ID | 1239 |
Feature activation | +7.46e-02 |
Pos logit contributions | |
forgot | +0.069 |
setup | +0.067 |
cost | +0.066 |
permission | +0.065 |
forget | +0.065 |
Neg logit contributions | |
ing | -0.125 |
ane | -0.108 |
ays | -0.108 |
he | -0.108 |
over | -0.107 |
Token | scared |
Token ID | 12008 |
Feature activation | -2.80e-02 |
Pos logit contributions | |
never | +0.095 |
help | +0.088 |
fact | +0.086 |
pap | +0.085 |
wait | +0.084 |
Neg logit contributions | |
ing | -0.181 |
ane | -0.150 |
he | -0.148 |
umping | -0.147 |
ays | -0.147 |
Token | . |
Token ID | 13 |
Feature activation | -3.07e-02 |
Pos logit contributions | |
ane | +0.043 |
ays | +0.042 |
ing | +0.040 |
eddy | +0.039 |
he | +0.039 |
Neg logit contributions | |
never | -0.065 |
forgotten | -0.062 |
grown | -0.061 |
not | -0.061 |
forgot | -0.060 |
Token | One |
Token ID | 3198 |
Feature activation | +8.52e-03 |
Pos logit contributions | |
ing | +0.087 |
ays | +0.084 |
ane | +0.083 |
eddy | +0.079 |
he | +0.076 |
Neg logit contributions | |
never | -0.056 |
forgot | -0.050 |
grown | -0.049 |
not | -0.049 |
forgotten | -0.049 |
Token | day |
Token ID | 1110 |
Feature activation | -3.63e-02 |
Pos logit contributions | |
never | +0.011 |
forgot | +0.011 |
grown | +0.011 |
help | +0.011 |
not | +0.011 |
Neg logit contributions | |
ays | -0.022 |
ing | -0.022 |
ane | -0.021 |
ement | -0.021 |
eddy | -0.020 |
Token | , |
Token ID | 11 |
Feature activation | -1.44e-02 |
Pos logit contributions | |
ays | +0.070 |
ane | +0.070 |
ing | +0.067 |
eddy | +0.066 |
he | +0.063 |
Neg logit contributions | |
never | -0.068 |
grown | -0.068 |
forgot | -0.068 |
help | -0.067 |
knocked | -0.067 |
Token | Sus |
Token ID | 8932 |
Feature activation | -1.97e-02 |
Pos logit contributions | |
ing | +0.035 |
ane | +0.033 |
ays | +0.033 |
eddy | +0.031 |
he | +0.031 |
Neg logit contributions | |
never | -0.021 |
forgot | -0.021 |
forgotten | -0.020 |
not | -0.020 |
help | -0.019 |
Token | ie |
Token ID | 494 |
Feature activation | +1.01e-02 |
Pos logit contributions | |
ing | +0.031 |
ane | +0.031 |
ays | +0.030 |
ement | +0.028 |
ay | +0.027 |
Neg logit contributions | |
never | -0.043 |
forgot | -0.043 |
forgotten | -0.040 |
been | -0.040 |
wasted | -0.040 |
Token | was |
Token ID | 373 |
Feature activation | +4.40e-02 |
Pos logit contributions | |
never | +0.011 |
forgot | +0.010 |
forgotten | +0.010 |
permission | +0.010 |
help | +0.010 |
Neg logit contributions | |
ing | -0.028 |
ane | -0.026 |
ays | -0.026 |
eddy | -0.025 |
he | -0.025 |
Token | looking |
Token ID | 2045 |
Feature activation | +1.45e-02 |
Pos logit contributions | |
forgot | +0.071 |
fiber | +0.071 |
vitamins | +0.068 |
forget | +0.067 |
permission | +0.067 |
Neg logit contributions | |
ing | -0.095 |
he | -0.078 |
ane | -0.078 |
apping | -0.077 |
ays | -0.075 |
Token | for |
Token ID | 329 |
Feature activation | +1.66e-02 |
Pos logit contributions | |
never | +0.020 |
forgot | +0.019 |
help | +0.019 |
grown | +0.019 |
permission | +0.019 |
Neg logit contributions | |
ane | -0.035 |
ing | -0.035 |
ays | -0.034 |
eddy | -0.033 |
he | -0.032 |
Token | something |
Token ID | 1223 |
Feature activation | +2.15e-04 |
Pos logit contributions | |
forgot | +0.031 |
never | +0.029 |
wasted | +0.028 |
setup | +0.028 |
forget | +0.027 |
Neg logit contributions | |
ing | -0.030 |
eddy | -0.025 |
ement | -0.025 |
ays | -0.024 |
he | -0.024 |
Token | special |
Token ID | 2041 |
Feature activation | -3.21e-02 |
Pos logit contributions | |
forgot | +0.000 |
never | +0.000 |
wasted | +0.000 |
known | +0.000 |
setup | +0.000 |
Neg logit contributions | |
ing | -0.000 |
ane | -0.000 |
ays | -0.000 |
eddy | -0.000 |
he | -0.000 |
Token | . |
Token ID | 13 |
Feature activation | -2.92e-02 |
Pos logit contributions | |
ing | +0.061 |
ane | +0.057 |
ays | +0.056 |
eddy | +0.053 |
he | +0.053 |
Neg logit contributions | |
never | -0.064 |
forgot | -0.062 |
forgotten | -0.061 |
not | -0.060 |
help | -0.059 |
Token | twins |
Token ID | 20345 |
Feature activation | -4.49e-02 |
Pos logit contributions | |
never | +0.004 |
forgot | +0.004 |
forgotten | +0.004 |
not | +0.004 |
wasted | +0.004 |
Neg logit contributions | |
ing | -0.007 |
ane | -0.007 |
ays | -0.007 |
eddy | -0.006 |
he | -0.006 |
Token | who |
Token ID | 508 |
Feature activation | +2.84e-02 |
Pos logit contributions | |
ing | +0.064 |
ane | +0.056 |
ays | +0.053 |
eddy | +0.052 |
he | +0.052 |
Neg logit contributions | |
forgot | -0.108 |
never | -0.107 |
forgotten | -0.105 |
known | -0.103 |
stolen | -0.103 |
Token | liked |
Token ID | 8288 |
Feature activation | +1.49e-02 |
Pos logit contributions | |
permission | +0.042 |
setup | +0.040 |
unknow | +0.039 |
forgotten | +0.037 |
stolen | +0.037 |
Neg logit contributions | |
ing | -0.065 |
he | -0.065 |
ane | -0.061 |
eddy | -0.058 |
ays | -0.056 |
Token | to |
Token ID | 284 |
Feature activation | +3.68e-02 |
Pos logit contributions | |
forgot | +0.029 |
never | +0.027 |
forgotten | +0.026 |
unknow | +0.026 |
known | +0.025 |
Neg logit contributions | |
ing | -0.030 |
ane | -0.027 |
eddy | -0.027 |
ays | -0.026 |
he | -0.024 |
Token | play |
Token ID | 711 |
Feature activation | -3.52e-02 |
Pos logit contributions | |
never | +0.042 |
forgot | +0.041 |
forgotten | +0.040 |
wasted | +0.038 |
stolen | +0.037 |
Neg logit contributions | |
ing | -0.098 |
ane | -0.094 |
ays | -0.091 |
he | -0.090 |
eddy | -0.089 |
Token | in |
Token ID | 287 |
Feature activation | -2.20e-03 |
Pos logit contributions | |
ane | +0.074 |
ing | +0.070 |
ays | +0.068 |
eddy | +0.064 |
umbled | +0.063 |
Neg logit contributions | |
help | -0.068 |
not | -0.064 |
never | -0.063 |
grown | -0.063 |
put | -0.061 |
Token | the |
Token ID | 262 |
Feature activation | +2.77e-03 |
Pos logit contributions | |
ing | +0.006 |
ane | +0.005 |
ays | +0.005 |
he | +0.005 |
eddy | +0.005 |
Neg logit contributions | |
forgot | -0.003 |
never | -0.003 |
forgotten | -0.003 |
wasted | -0.003 |
known | -0.003 |
Token | garden |
Token ID | 11376 |
Feature activation | -6.87e-02 |
Pos logit contributions | |
never | +0.004 |
forgot | +0.004 |
forgotten | +0.004 |
wasted | +0.004 |
not | +0.004 |
Neg logit contributions | |
ing | -0.007 |
ane | -0.006 |
ays | -0.006 |
eddy | -0.006 |
he | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | -4.35e-02 |
Pos logit contributions | |
ane | +0.125 |
ing | +0.124 |
ays | +0.120 |
eddy | +0.116 |
he | +0.108 |
Neg logit contributions | |
never | -0.142 |
not | -0.136 |
help | -0.132 |
always | -0.129 |
grown | -0.128 |
Token | They |
Token ID | 1119 |
Feature activation | +6.34e-02 |
Pos logit contributions | |
ing | +0.105 |
ane | +0.099 |
eddy | +0.094 |
ays | +0.092 |
he | +0.092 |
Neg logit contributions | |
never | -0.065 |
forgot | -0.064 |
not | -0.062 |
forgotten | -0.061 |
grown | -0.061 |
Token | had |
Token ID | 550 |
Feature activation | +2.48e-02 |
Pos logit contributions | |
permission | +0.109 |
warning | +0.106 |
tresp | +0.100 |
mistake | +0.099 |
setup | +0.099 |
Neg logit contributions | |
he | -0.125 |
ing | -0.118 |
ane | -0.113 |
ite | -0.110 |
y | -0.107 |
Token | near |
Token ID | 1474 |
Feature activation | -9.79e-03 |
Pos logit contributions | |
ing | +0.117 |
ane | +0.110 |
ays | +0.107 |
eddy | +0.101 |
he | +0.098 |
Neg logit contributions | |
never | -0.170 |
forgot | -0.161 |
not | -0.161 |
forgotten | -0.160 |
help | -0.157 |
Token | her |
Token ID | 607 |
Feature activation | +1.63e-02 |
Pos logit contributions | |
ing | +0.023 |
ays | +0.021 |
he | +0.021 |
eddy | +0.021 |
ane | +0.021 |
Neg logit contributions | |
forgot | -0.015 |
never | -0.014 |
forgotten | -0.013 |
known | -0.013 |
unknow | -0.013 |
Token | house |
Token ID | 2156 |
Feature activation | -6.05e-02 |
Pos logit contributions | |
never | +0.029 |
forgot | +0.029 |
forgotten | +0.028 |
known | +0.027 |
wasted | +0.027 |
Neg logit contributions | |
ing | -0.034 |
ane | -0.032 |
ays | -0.031 |
eddy | -0.030 |
he | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | -5.79e-02 |
Pos logit contributions | |
ing | +0.100 |
ane | +0.094 |
ays | +0.089 |
eddy | +0.087 |
he | +0.082 |
Neg logit contributions | |
never | -0.138 |
forgot | -0.131 |
not | -0.131 |
forgotten | -0.130 |
help | -0.127 |
Token | She |
Token ID | 1375 |
Feature activation | +4.68e-02 |
Pos logit contributions | |
ing | +0.133 |
ane | +0.126 |
ays | +0.119 |
ement | +0.116 |
eddy | +0.113 |
Neg logit contributions | |
grown | -0.094 |
never | -0.092 |
help | -0.092 |
not | -0.092 |
forgot | -0.090 |
Token | thought |
Token ID | 1807 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
permission | +0.087 |
supervision | +0.082 |
backup | +0.080 |
tresp | +0.078 |
payoff | +0.077 |
Neg logit contributions | |
ing | -0.085 |
ed | -0.078 |
s | -0.073 |
eddy | -0.073 |
ays | -0.071 |
Token | it |
Token ID | 340 |
Feature activation | +1.40e-02 |
Pos logit contributions | |
never | +0.022 |
forgot | +0.021 |
forgotten | +0.020 |
wasted | +0.019 |
known | +0.019 |
Neg logit contributions | |
ing | -0.034 |
ane | -0.032 |
ays | -0.031 |
eddy | -0.030 |
he | -0.030 |
Token | would |
Token ID | 561 |
Feature activation | +6.39e-02 |
Pos logit contributions | |
never | +0.017 |
forgot | +0.016 |
forgotten | +0.015 |
not | +0.015 |
help | +0.014 |
Neg logit contributions | |
ing | -0.038 |
ane | -0.037 |
ays | -0.036 |
eddy | -0.035 |
he | -0.034 |
Token | be |
Token ID | 307 |
Feature activation | +4.55e-02 |
Pos logit contributions | |
unknow | +0.090 |
robbers | +0.088 |
tresp | +0.083 |
wasted | +0.082 |
supervision | +0.080 |
Neg logit contributions | |
he | -0.111 |
time | -0.107 |
room | -0.104 |
anymore | -0.103 |
with | -0.099 |
Token | fun |
Token ID | 1257 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
unknow | +0.058 |
forgot | +0.050 |
pap | +0.050 |
never | +0.049 |
permission | +0.049 |
Neg logit contributions | |
ing | -0.102 |
ays | -0.091 |
he | -0.090 |
eddy | -0.089 |
ement | -0.089 |
Token | to |
Token ID | 284 |
Feature activation | +5.15e-02 |
Pos logit contributions | |
ing | +0.035 |
ane | +0.033 |
ays | +0.032 |
eddy | +0.031 |
he | +0.031 |
Neg logit contributions | |
never | -0.023 |
forgot | -0.023 |
forgotten | -0.022 |
not | -0.021 |
help | -0.021 |
Token | Sara |
Token ID | 24799 |
Feature activation | +1.43e-02 |
Pos logit contributions | |
ing | +0.130 |
ane | +0.123 |
ement | +0.115 |
ays | +0.115 |
he | +0.113 |
Neg logit contributions | |
never | -0.081 |
forgot | -0.079 |
not | -0.079 |
grown | -0.074 |
wasted | -0.074 |
Token | and |
Token ID | 290 |
Feature activation | +2.66e-03 |
Pos logit contributions | |
forgotten | +0.020 |
permission | +0.020 |
stolen | +0.020 |
known | +0.020 |
never | +0.019 |
Neg logit contributions | |
ing | -0.034 |
ane | -0.031 |
ays | -0.030 |
eddy | -0.030 |
he | -0.030 |
Token | the |
Token ID | 262 |
Feature activation | -1.66e-03 |
Pos logit contributions | |
never | +0.003 |
forgot | +0.003 |
forgotten | +0.003 |
not | +0.003 |
wasted | +0.003 |
Neg logit contributions | |
ing | -0.007 |
ane | -0.007 |
ays | -0.007 |
eddy | -0.006 |
he | -0.006 |
Token | bear |
Token ID | 6842 |
Feature activation | +2.59e-03 |
Pos logit contributions | |
ing | +0.004 |
ane | +0.004 |
eddy | +0.003 |
he | +0.003 |
ays | +0.003 |
Neg logit contributions | |
never | -0.003 |
forgot | -0.003 |
forgotten | -0.003 |
wasted | -0.002 |
permission | -0.002 |
Token | said |
Token ID | 531 |
Feature activation | -8.49e-03 |
Pos logit contributions | |
never | +0.003 |
forgot | +0.003 |
forgotten | +0.003 |
not | +0.003 |
stolen | +0.003 |
Neg logit contributions | |
ing | -0.007 |
ane | -0.007 |
ays | -0.006 |
he | -0.006 |
eddy | -0.006 |
Token | , |
Token ID | 11 |
Feature activation | -6.88e-03 |
Pos logit contributions | |
ays | +0.017 |
ane | +0.017 |
he | +0.016 |
ing | +0.016 |
eddy | +0.016 |
Neg logit contributions | |
never | -0.016 |
help | -0.016 |
not | -0.016 |
forgot | -0.015 |
grown | -0.015 |
Token | " |
Token ID | 366 |
Feature activation | -2.67e-02 |
Pos logit contributions | |
ing | +0.018 |
ays | +0.018 |
ane | +0.017 |
he | +0.017 |
eddy | +0.016 |
Neg logit contributions | |
never | -0.009 |
not | -0.009 |
forgot | -0.008 |
help | -0.008 |
grown | -0.008 |
Token | Okay |
Token ID | 16454 |
Feature activation | -2.61e-02 |
Pos logit contributions | |
ing | +0.067 |
ane | +0.064 |
ays | +0.063 |
he | +0.059 |
eddy | +0.058 |
Neg logit contributions | |
never | -0.040 |
forgot | -0.037 |
forgotten | -0.034 |
wished | -0.034 |
help | -0.034 |
Token | , |
Token ID | 11 |
Feature activation | +1.00e-02 |
Pos logit contributions | |
ane | +0.053 |
ays | +0.052 |
ing | +0.052 |
he | +0.049 |
eddy | +0.047 |
Neg logit contributions | |
never | -0.051 |
not | -0.050 |
help | -0.049 |
be | -0.048 |
forgot | -0.048 |
Token | mom |
Token ID | 1995 |
Feature activation | -3.92e-02 |
Pos logit contributions | |
never | +0.013 |
forgot | +0.013 |
forgotten | +0.013 |
wasted | +0.013 |
stolen | +0.013 |
Neg logit contributions | |
ing | -0.025 |
ane | -0.023 |
eddy | -0.023 |
ays | -0.023 |
he | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
ays | +0.073 |
ane | +0.073 |
ing | +0.072 |
he | +0.069 |
ement | +0.069 |
Neg logit contributions | |
never | -0.079 |
forgot | -0.075 |
not | -0.074 |
help | -0.074 |
be | -0.072 |
Token | upon |
Token ID | 2402 |
Feature activation | -6.02e-03 |
Pos logit contributions | |
ing | +0.075 |
ays | +0.072 |
ane | +0.070 |
eddy | +0.070 |
ement | +0.066 |
Neg logit contributions | |
never | -0.049 |
forgot | -0.047 |
not | -0.047 |
forgotten | -0.043 |
help | -0.043 |
Token | a |
Token ID | 257 |
Feature activation | -3.42e-02 |
Pos logit contributions | |
ing | +0.017 |
ane | +0.017 |
ays | +0.016 |
eddy | +0.016 |
he | +0.016 |
Neg logit contributions | |
never | -0.006 |
forgot | -0.006 |
forgotten | -0.005 |
not | -0.005 |
help | -0.005 |
Token | time |
Token ID | 640 |
Feature activation | -4.39e-02 |
Pos logit contributions | |
ing | +0.094 |
ane | +0.090 |
ays | +0.089 |
eddy | +0.084 |
he | +0.081 |
Neg logit contributions | |
never | -0.044 |
forgot | -0.041 |
not | -0.039 |
forgotten | -0.038 |
grown | -0.036 |
Token | , |
Token ID | 11 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
ays | +0.096 |
eddy | +0.093 |
ing | +0.092 |
ane | +0.090 |
he | +0.084 |
Neg logit contributions | |
not | -0.079 |
never | -0.077 |
help | -0.071 |
be | -0.070 |
forgot | -0.068 |
Token | there |
Token ID | 612 |
Feature activation | -2.51e-02 |
Pos logit contributions | |
ane | +0.030 |
ays | +0.030 |
ing | +0.030 |
ement | +0.029 |
eddy | +0.029 |
Neg logit contributions | |
never | -0.020 |
not | -0.020 |
also | -0.019 |
always | -0.017 |
knocked | -0.017 |
Token | was |
Token ID | 373 |
Feature activation | +7.52e-03 |
Pos logit contributions | |
ing | +0.068 |
ane | +0.066 |
ays | +0.065 |
eddy | +0.062 |
ement | +0.060 |
Neg logit contributions | |
never | -0.032 |
forgot | -0.029 |
not | -0.028 |
been | -0.028 |
grown | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | -7.21e-03 |
Pos logit contributions | |
never | +0.007 |
forgot | +0.006 |
not | +0.006 |
forgotten | +0.006 |
been | +0.006 |
Neg logit contributions | |
ing | -0.023 |
ane | -0.022 |
ays | -0.021 |
eddy | -0.020 |
he | -0.020 |
Token | little |
Token ID | 1310 |
Feature activation | -1.70e-02 |
Pos logit contributions | |
ing | +0.018 |
ane | +0.017 |
eddy | +0.017 |
he | +0.017 |
ays | +0.017 |
Neg logit contributions | |
never | -0.009 |
forgot | -0.009 |
forgotten | -0.009 |
permission | -0.009 |
known | -0.008 |
Token | girl |
Token ID | 2576 |
Feature activation | -4.59e-02 |
Pos logit contributions | |
ing | +0.037 |
ane | +0.035 |
ays | +0.035 |
he | +0.033 |
eddy | +0.032 |
Neg logit contributions | |
never | -0.028 |
forgot | -0.027 |
forgotten | -0.026 |
help | -0.026 |
grown | -0.026 |
Token | named |
Token ID | 3706 |
Feature activation | -6.93e-03 |
Pos logit contributions | |
ing | +0.125 |
ane | +0.120 |
ays | +0.118 |
he | +0.113 |
eddy | +0.112 |
Neg logit contributions | |
never | -0.057 |
forgot | -0.052 |
not | -0.049 |
forgotten | -0.048 |
help | -0.048 |
Token | Lily |
Token ID | 20037 |
Feature activation | -4.11e-02 |
Pos logit contributions | |
ing | +0.018 |
ane | +0.017 |
ays | +0.016 |
eddy | +0.016 |
he | +0.016 |
Neg logit contributions | |
never | -0.010 |
forgot | -0.009 |
forgotten | -0.009 |
not | -0.009 |
wasted | -0.008 |
Token | her |
Token ID | 607 |
Feature activation | +2.27e-02 |
Pos logit contributions | |
forgot | +0.013 |
never | +0.012 |
forgotten | +0.012 |
wasted | +0.012 |
stolen | +0.012 |
Neg logit contributions | |
ing | -0.021 |
ane | -0.019 |
ays | -0.019 |
he | -0.019 |
ement | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | -3.43e-02 |
Pos logit contributions | |
never | +0.045 |
forgot | +0.044 |
forgotten | +0.043 |
not | +0.042 |
help | +0.042 |
Neg logit contributions | |
ing | -0.044 |
ane | -0.041 |
ays | -0.040 |
eddy | -0.038 |
he | -0.038 |
Token | But |
Token ID | 887 |
Feature activation | +1.92e-02 |
Pos logit contributions | |
ing | +0.070 |
ane | +0.066 |
eddy | +0.065 |
ays | +0.064 |
he | +0.063 |
Neg logit contributions | |
never | -0.065 |
not | -0.061 |
forgot | -0.060 |
grown | -0.059 |
help | -0.057 |
Token | Mom |
Token ID | 11254 |
Feature activation | -1.21e-02 |
Pos logit contributions | |
never | +0.030 |
forgot | +0.030 |
forgotten | +0.029 |
wasted | +0.028 |
stolen | +0.028 |
Neg logit contributions | |
ing | -0.044 |
ane | -0.041 |
ays | -0.040 |
he | -0.038 |
eddy | -0.038 |
Token | my |
Token ID | 1820 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
ing | +0.024 |
ane | +0.023 |
ays | +0.022 |
eddy | +0.021 |
he | +0.021 |
Neg logit contributions | |
never | -0.023 |
forgot | -0.022 |
forgotten | -0.021 |
not | -0.021 |
wasted | -0.021 |
Token | said |
Token ID | 531 |
Feature activation | +3.71e-03 |
Pos logit contributions | |
permission | +0.013 |
never | +0.013 |
forgotten | +0.013 |
wasted | +0.012 |
known | +0.012 |
Neg logit contributions | |
ing | -0.031 |
ane | -0.028 |
he | -0.027 |
ays | -0.026 |
ement | -0.026 |
Token | it |
Token ID | 340 |
Feature activation | +8.88e-03 |
Pos logit contributions | |
never | +0.008 |
forgot | +0.008 |
forgotten | +0.008 |
wasted | +0.008 |
known | +0.008 |
Neg logit contributions | |
ing | -0.006 |
ane | -0.006 |
ays | -0.005 |
eddy | -0.005 |
he | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | +5.31e-02 |
Pos logit contributions | |
forgot | +0.014 |
permission | +0.014 |
never | +0.014 |
forgotten | +0.013 |
known | +0.013 |
Neg logit contributions | |
ing | -0.020 |
ane | -0.018 |
ays | -0.017 |
eddy | -0.017 |
he | -0.017 |
Token | alright |
Token ID | 23036 |
Feature activation | -2.85e-02 |
Pos logit contributions | |
unknow | +0.082 |
forgot | +0.080 |
wished | +0.079 |
wasted | +0.074 |
forget | +0.072 |
Neg logit contributions | |
ing | -0.099 |
ay | -0.083 |
ays | -0.081 |
over | -0.079 |
ement | -0.078 |
Token | . |
Token ID | 13 |
Feature activation | -2.87e-02 |
Pos logit contributions | |
ing | +0.049 |
ane | +0.049 |
ays | +0.048 |
he | +0.046 |
eddy | +0.044 |
Neg logit contributions | |
never | -0.062 |
not | -0.059 |
forgot | -0.059 |
forgotten | -0.058 |
help | -0.057 |
Token |
|
Token ID | 198 |
Feature activation | -3.85e-02 |
Pos logit contributions | |
ing | +0.059 |
ane | +0.056 |
eddy | +0.054 |
ays | +0.053 |
he | +0.052 |
Neg logit contributions | |
never | -0.053 |
forgot | -0.050 |
not | -0.050 |
grown | -0.048 |
known | -0.048 |
Token | bunny |
Token ID | 44915 |
Feature activation | +2.43e-02 |
Pos logit contributions | |
never | +0.028 |
forgot | +0.027 |
forgotten | +0.026 |
wasted | +0.026 |
known | +0.025 |
Neg logit contributions | |
ing | -0.034 |
ane | -0.031 |
eddy | -0.030 |
ays | -0.030 |
he | -0.030 |
Token | became |
Token ID | 2627 |
Feature activation | +5.16e-02 |
Pos logit contributions | |
forgotten | +0.033 |
permission | +0.033 |
never | +0.032 |
stolen | +0.032 |
wasted | +0.032 |
Neg logit contributions | |
ing | -0.057 |
ane | -0.052 |
he | -0.049 |
ays | -0.049 |
ite | -0.047 |
Token | very |
Token ID | 845 |
Feature activation | +4.71e-02 |
Pos logit contributions | |
forgot | +0.077 |
never | +0.077 |
forgotten | +0.074 |
permission | +0.072 |
unknow | +0.072 |
Neg logit contributions | |
ing | -0.121 |
ays | -0.107 |
ane | -0.106 |
eddy | -0.105 |
he | -0.101 |
Token | good |
Token ID | 922 |
Feature activation | -5.65e-03 |
Pos logit contributions | |
never | +0.058 |
permission | +0.057 |
forgotten | +0.055 |
unknow | +0.055 |
been | +0.054 |
Neg logit contributions | |
ing | -0.120 |
ane | -0.109 |
eddy | -0.106 |
ays | -0.106 |
he | -0.103 |
Token | friends |
Token ID | 2460 |
Feature activation | -3.51e-02 |
Pos logit contributions | |
ane | +0.015 |
ing | +0.015 |
ays | +0.015 |
he | +0.015 |
ement | +0.014 |
Neg logit contributions | |
never | -0.006 |
forgot | -0.006 |
forgotten | -0.005 |
not | -0.005 |
help | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | -2.71e-02 |
Pos logit contributions | |
ays | +0.068 |
ane | +0.066 |
ing | +0.066 |
eddy | +0.063 |
he | +0.060 |
Neg logit contributions | |
never | -0.073 |
not | -0.064 |
forgotten | -0.063 |
forgot | -0.062 |
wished | -0.062 |
Token | They |
Token ID | 1119 |
Feature activation | +7.63e-02 |
Pos logit contributions | |
ing | +0.056 |
ane | +0.052 |
ays | +0.050 |
eddy | +0.049 |
he | +0.049 |
Neg logit contributions | |
never | -0.051 |
forgot | -0.049 |
forgotten | -0.047 |
not | -0.047 |
help | -0.046 |
Token | played |
Token ID | 2826 |
Feature activation | -2.57e-02 |
Pos logit contributions | |
warning | +0.117 |
permission | +0.116 |
mistake | +0.115 |
unknow | +0.113 |
payoff | +0.110 |
Neg logit contributions | |
ing | -0.147 |
he | -0.143 |
ane | -0.142 |
ite | -0.133 |
hat | -0.125 |
Token | together |
Token ID | 1978 |
Feature activation | -2.18e-03 |
Pos logit contributions | |
ane | +0.048 |
ing | +0.047 |
ays | +0.046 |
eddy | +0.044 |
ite | +0.039 |
Neg logit contributions | |
never | -0.055 |
not | -0.053 |
help | -0.053 |
always | -0.050 |
grown | -0.050 |
Token | everyday |
Token ID | 10908 |
Feature activation | -2.79e-03 |
Pos logit contributions | |
ing | +0.005 |
ane | +0.005 |
ays | +0.004 |
eddy | +0.004 |
he | +0.004 |
Neg logit contributions | |
never | -0.004 |
forgot | -0.004 |
not | -0.004 |
forgotten | -0.003 |
help | -0.003 |
Token | and |
Token ID | 290 |
Feature activation | +4.15e-02 |
Pos logit contributions | |
ing | +0.005 |
ane | +0.004 |
ays | +0.004 |
eddy | +0.004 |
he | +0.004 |
Neg logit contributions | |
never | -0.006 |
forgot | -0.006 |
forgotten | -0.006 |
not | -0.006 |
help | -0.006 |
Token | a |
Token ID | 257 |
Feature activation | -4.10e-02 |
Pos logit contributions | |
ing | +0.046 |
ane | +0.044 |
ays | +0.043 |
eddy | +0.041 |
ement | +0.041 |
Neg logit contributions | |
never | -0.014 |
not | -0.013 |
forgot | -0.013 |
forgotten | -0.012 |
been | -0.012 |
Token | time |
Token ID | 640 |
Feature activation | -4.86e-02 |
Pos logit contributions | |
ing | +0.115 |
ane | +0.110 |
ays | +0.110 |
eddy | +0.104 |
ement | +0.099 |
Neg logit contributions | |
never | -0.050 |
forgot | -0.046 |
not | -0.044 |
forgotten | -0.043 |
been | -0.041 |
Token | , |
Token ID | 11 |
Feature activation | -2.17e-02 |
Pos logit contributions | |
ays | +0.108 |
eddy | +0.106 |
ing | +0.105 |
ane | +0.100 |
acked | +0.097 |
Neg logit contributions | |
not | -0.084 |
never | -0.080 |
help | -0.075 |
be | -0.073 |
grown | -0.071 |
Token | there |
Token ID | 612 |
Feature activation | -3.64e-02 |
Pos logit contributions | |
ane | +0.055 |
ing | +0.055 |
ays | +0.054 |
ement | +0.053 |
eddy | +0.052 |
Neg logit contributions | |
never | -0.032 |
not | -0.032 |
also | -0.029 |
been | -0.028 |
knocked | -0.028 |
Token | was |
Token ID | 373 |
Feature activation | -6.01e-04 |
Pos logit contributions | |
ing | +0.102 |
ane | +0.099 |
ays | +0.099 |
eddy | +0.094 |
ement | +0.091 |
Neg logit contributions | |
never | -0.043 |
not | -0.037 |
forgot | -0.037 |
been | -0.036 |
be | -0.035 |
Token | a |
Token ID | 257 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
ing | +0.002 |
ane | +0.002 |
ays | +0.002 |
ement | +0.002 |
eddy | +0.002 |
Neg logit contributions | |
never | -0.001 |
not | -0.000 |
been | -0.000 |
forgotten | -0.000 |
forgot | -0.000 |
Token | boy |
Token ID | 2933 |
Feature activation | -6.21e-02 |
Pos logit contributions | |
ing | +0.038 |
ane | +0.035 |
ays | +0.034 |
eddy | +0.034 |
he | +0.034 |
Neg logit contributions | |
never | -0.018 |
forgot | -0.017 |
forgotten | -0.016 |
permission | -0.015 |
wasted | -0.015 |
Token | named |
Token ID | 3706 |
Feature activation | -8.16e-03 |
Pos logit contributions | |
ing | +0.177 |
ane | +0.171 |
ays | +0.170 |
ement | +0.163 |
he | +0.160 |
Neg logit contributions | |
never | -0.072 |
forgot | -0.061 |
help | -0.061 |
not | -0.061 |
knocked | -0.057 |
Token | Tim |
Token ID | 5045 |
Feature activation | -5.07e-02 |
Pos logit contributions | |
ing | +0.019 |
ane | +0.018 |
ays | +0.018 |
eddy | +0.017 |
he | +0.017 |
Neg logit contributions | |
never | -0.013 |
forgot | -0.012 |
forgotten | -0.012 |
not | -0.012 |
help | -0.011 |
Token | my |
Token ID | 1820 |
Feature activation | -5.35e-02 |
Pos logit contributions | |
ing | +0.117 |
ane | +0.111 |
ays | +0.110 |
ement | +0.105 |
eddy | +0.099 |
Neg logit contributions | |
never | -0.086 |
forgot | -0.080 |
been | -0.077 |
not | -0.077 |
forgotten | -0.076 |
Token | who |
Token ID | 508 |
Feature activation | -5.02e-04 |
Pos logit contributions | |
ing | +0.090 |
ane | +0.084 |
ays | +0.084 |
ement | +0.077 |
he | +0.075 |
Neg logit contributions | |
never | -0.122 |
forgot | -0.115 |
not | -0.113 |
help | -0.112 |
forgotten | -0.111 |
Token | end |
Token ID | 886 |
Feature activation | -2.92e-02 |
Pos logit contributions | |
never | +0.031 |
forgot | +0.029 |
forgotten | +0.029 |
known | +0.029 |
wasted | +0.028 |
Neg logit contributions | |
ing | -0.046 |
he | -0.041 |
ane | -0.041 |
ement | -0.039 |
eddy | -0.039 |
Token | . |
Token ID | 13 |
Feature activation | -3.03e-02 |
Pos logit contributions | |
ays | +0.057 |
ane | +0.056 |
ing | +0.055 |
he | +0.050 |
ement | +0.048 |
Neg logit contributions | |
never | -0.058 |
help | -0.056 |
not | -0.055 |
forgot | -0.053 |
forgotten | -0.052 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -7.89e-02 |
Pos logit contributions | |
ing | +0.061 |
ane | +0.056 |
ays | +0.055 |
ement | +0.051 |
eddy | +0.051 |
Neg logit contributions | |
never | -0.059 |
forgot | -0.055 |
not | -0.054 |
been | -0.053 |
known | -0.053 |
Token | Once |
Token ID | 7454 |
Feature activation | -4.02e-02 |
Pos logit contributions | |
ing | +0.201 |
ays | +0.194 |
ane | +0.192 |
ement | +0.189 |
ooming | +0.188 |
Neg logit contributions | |
never | -0.098 |
been | -0.097 |
have | -0.088 |
grown | -0.086 |
help | -0.083 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.18e-02 |
Pos logit contributions | |
ing | +0.093 |
ays | +0.089 |
eddy | +0.087 |
ane | +0.087 |
umping | +0.083 |
Neg logit contributions | |
never | -0.060 |
forgot | -0.059 |
not | -0.058 |
grown | -0.053 |
help | -0.053 |
Token | a |
Token ID | 257 |
Feature activation | -3.62e-02 |
Pos logit contributions | |
ing | +0.035 |
ane | +0.033 |
ays | +0.033 |
eddy | +0.031 |
he | +0.031 |
Neg logit contributions | |
never | -0.012 |
forgot | -0.011 |
not | -0.010 |
forgotten | -0.010 |
help | -0.010 |
Token | time |
Token ID | 640 |
Feature activation | -4.95e-02 |
Pos logit contributions | |
ing | +0.099 |
ane | +0.094 |
ays | +0.093 |
eddy | +0.089 |
he | +0.085 |
Neg logit contributions | |
never | -0.047 |
forgot | -0.043 |
not | -0.041 |
forgotten | -0.040 |
help | -0.039 |
Token | , |
Token ID | 11 |
Feature activation | -2.09e-02 |
Pos logit contributions | |
ays | +0.111 |
eddy | +0.108 |
ing | +0.106 |
ane | +0.103 |
he | +0.097 |
Neg logit contributions | |
not | -0.087 |
never | -0.084 |
help | -0.080 |
be | -0.077 |
grown | -0.075 |
Token | there |
Token ID | 612 |
Feature activation | -3.27e-02 |
Pos logit contributions | |
ane | +0.054 |
ays | +0.053 |
ing | +0.052 |
eddy | +0.051 |
ement | +0.051 |
Neg logit contributions | |
never | -0.031 |
not | -0.030 |
also | -0.029 |
been | -0.026 |
be | -0.026 |
Token | was |
Token ID | 373 |
Feature activation | +1.07e-03 |
Pos logit contributions | |
ing | +0.091 |
ane | +0.089 |
ays | +0.088 |
eddy | +0.084 |
ement | +0.081 |
Neg logit contributions | |
never | -0.040 |
forgot | -0.035 |
not | -0.034 |
been | -0.034 |
grown | -0.032 |
Token | a |
Token ID | 257 |
Feature activation | -1.06e-02 |
Pos logit contributions | |
never | +0.001 |
not | +0.001 |
forgot | +0.001 |
been | +0.001 |
forgotten | +0.001 |
Neg logit contributions | |
ing | -0.003 |
ane | -0.003 |
ays | -0.003 |
ement | -0.003 |
eddy | -0.003 |
Token | happy |
Token ID | 3772 |
Feature activation | -8.08e-03 |
Pos logit contributions | |
robbers | +0.071 |
forgot | +0.070 |
wished | +0.070 |
tresp | +0.069 |
never | +0.069 |
Neg logit contributions | |
ing | -0.132 |
he | -0.122 |
ays | -0.118 |
ane | -0.116 |
ement | -0.113 |
Token | and |
Token ID | 290 |
Feature activation | +5.23e-02 |
Pos logit contributions | |
ing | +0.016 |
ane | +0.015 |
ays | +0.015 |
he | +0.014 |
eddy | +0.014 |
Neg logit contributions | |
never | -0.016 |
forgot | -0.015 |
not | -0.015 |
forgotten | -0.015 |
help | -0.015 |
Token | went |
Token ID | 1816 |
Feature activation | -2.82e-03 |
Pos logit contributions | |
tresp | +0.081 |
permission | +0.080 |
robbers | +0.077 |
setup | +0.075 |
wasted | +0.074 |
Neg logit contributions | |
ing | -0.120 |
ed | -0.101 |
he | -0.099 |
ane | -0.095 |
ays | -0.094 |
Token | back |
Token ID | 736 |
Feature activation | +5.97e-03 |
Pos logit contributions | |
ing | +0.006 |
ane | +0.005 |
ays | +0.005 |
eddy | +0.005 |
he | +0.005 |
Neg logit contributions | |
never | -0.005 |
forgot | -0.005 |
forgotten | -0.005 |
not | -0.005 |
wasted | -0.005 |
Token | home |
Token ID | 1363 |
Feature activation | -2.03e-02 |
Pos logit contributions | |
never | +0.011 |
forgot | +0.011 |
forgotten | +0.010 |
not | +0.010 |
wasted | +0.010 |
Neg logit contributions | |
ing | -0.012 |
ane | -0.011 |
ays | -0.011 |
eddy | -0.011 |
he | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | -2.86e-02 |
Pos logit contributions | |
ing | +0.027 |
ane | +0.021 |
he | +0.021 |
eddy | +0.021 |
ays | +0.020 |
Neg logit contributions | |
forgot | -0.051 |
never | -0.050 |
known | -0.049 |
wasted | -0.049 |
stolen | -0.049 |
Token | She |
Token ID | 1375 |
Feature activation | +7.26e-02 |
Pos logit contributions | |
ing | +0.050 |
ane | +0.046 |
ays | +0.045 |
eddy | +0.042 |
he | +0.042 |
Neg logit contributions | |
never | -0.062 |
forgot | -0.060 |
forgotten | -0.058 |
not | -0.057 |
wasted | -0.057 |
Token | was |
Token ID | 373 |
Feature activation | +5.26e-02 |
Pos logit contributions | |
permission | +0.124 |
robbers | +0.118 |
tresp | +0.109 |
ruining | +0.108 |
setup | +0.108 |
Neg logit contributions | |
ing | -0.140 |
he | -0.131 |
eddy | -0.129 |
ane | -0.124 |
ays | -0.121 |
Token | excited |
Token ID | 6568 |
Feature activation | -8.94e-03 |
Pos logit contributions | |
robbers | +0.082 |
forgot | +0.076 |
fiber | +0.076 |
vitamins | +0.075 |
permission | +0.074 |
Neg logit contributions | |
ing | -0.111 |
he | -0.106 |
ane | -0.100 |
ays | -0.096 |
eddy | -0.094 |
Token | to |
Token ID | 284 |
Feature activation | +5.43e-02 |
Pos logit contributions | |
ing | +0.020 |
ane | +0.019 |
ays | +0.019 |
eddy | +0.018 |
he | +0.018 |
Neg logit contributions | |
never | -0.015 |
forgot | -0.014 |
forgotten | -0.014 |
not | -0.013 |
help | -0.013 |
Token | show |
Token ID | 905 |
Feature activation | +6.21e-03 |
Pos logit contributions | |
wasted | +0.074 |
setup | +0.071 |
forgot | +0.069 |
unknow | +0.067 |
stolen | +0.067 |
Neg logit contributions | |
he | -0.113 |
ay | -0.111 |
ane | -0.107 |
ing | -0.107 |
ingle | -0.100 |
Token | their |
Token ID | 511 |
Feature activation | +2.43e-02 |
Pos logit contributions | |
ing | +0.014 |
ays | +0.013 |
ane | +0.013 |
eddy | +0.013 |
he | +0.013 |
Neg logit contributions | |
forgot | -0.008 |
never | -0.007 |
forgotten | -0.007 |
known | -0.007 |
unknow | -0.007 |
Token | coats |
Token ID | 30720 |
Feature activation | -7.05e-02 |
Pos logit contributions | |
forgot | +0.038 |
never | +0.037 |
known | +0.036 |
forgotten | +0.035 |
wasted | +0.034 |
Neg logit contributions | |
ing | -0.052 |
ane | -0.050 |
ays | -0.048 |
eddy | -0.048 |
ement | -0.047 |
Token | and |
Token ID | 290 |
Feature activation | +5.42e-02 |
Pos logit contributions | |
ing | +0.143 |
ays | +0.138 |
ane | +0.135 |
ement | +0.127 |
eddy | +0.126 |
Neg logit contributions | |
never | -0.134 |
not | -0.126 |
always | -0.115 |
forgotten | -0.115 |
help | -0.115 |
Token | shoes |
Token ID | 10012 |
Feature activation | -4.71e-02 |
Pos logit contributions | |
never | +0.082 |
regret | +0.081 |
known | +0.081 |
unknow | +0.080 |
setup | +0.079 |
Neg logit contributions | |
ing | -0.113 |
ane | -0.101 |
ays | -0.100 |
eddy | -0.098 |
ite | -0.098 |
Token | and |
Token ID | 290 |
Feature activation | +5.28e-02 |
Pos logit contributions | |
ing | +0.090 |
ays | +0.087 |
ane | +0.084 |
eddy | +0.080 |
ement | +0.079 |
Neg logit contributions | |
never | -0.095 |
not | -0.088 |
help | -0.087 |
forgot | -0.084 |
forgotten | -0.084 |
Token | walked |
Token ID | 6807 |
Feature activation | -2.64e-02 |
Pos logit contributions | |
unknow | +0.077 |
never | +0.076 |
setup | +0.076 |
regret | +0.076 |
known | +0.076 |
Neg logit contributions | |
ing | -0.108 |
ane | -0.098 |
ed | -0.094 |
he | -0.094 |
ite | -0.094 |
Token | to |
Token ID | 284 |
Feature activation | +2.42e-02 |
Pos logit contributions | |
ane | +0.064 |
ays | +0.063 |
ing | +0.063 |
ement | +0.060 |
eddy | +0.057 |
Neg logit contributions | |
never | -0.041 |
help | -0.037 |
not | -0.037 |
forgotten | -0.036 |
knocked | -0.035 |
Token | the |
Token ID | 262 |
Feature activation | +1.96e-02 |
Pos logit contributions | |
forgot | +0.038 |
never | +0.036 |
forgotten | +0.034 |
stolen | +0.033 |
unknow | +0.033 |
Neg logit contributions | |
ing | -0.053 |
he | -0.052 |
eddy | -0.050 |
ane | -0.049 |
ays | -0.049 |
Token | mailbox |
Token ID | 37282 |
Feature activation | -5.90e-02 |
Pos logit contributions | |
forgot | +0.040 |
known | +0.038 |
never | +0.037 |
forgotten | +0.037 |
wasted | +0.036 |
Neg logit contributions | |
ing | -0.035 |
ane | -0.032 |
eddy | -0.031 |
he | -0.031 |
ays | -0.029 |
Token | . |
Token ID | 13 |
Feature activation | -3.56e-02 |
Pos logit contributions | |
ing | +0.103 |
ane | +0.096 |
ays | +0.094 |
eddy | +0.088 |
ement | +0.085 |
Neg logit contributions | |
never | -0.129 |
not | -0.121 |
forgot | -0.120 |
forgotten | -0.120 |
help | -0.118 |
Token | Lily |
Token ID | 20037 |
Feature activation | +3.01e-02 |
Pos logit contributions | |
ing | +0.090 |
ane | +0.087 |
ays | +0.079 |
ement | +0.078 |
he | +0.076 |
Neg logit contributions | |
never | -0.051 |
not | -0.050 |
grown | -0.050 |
forgot | -0.049 |
forgotten | -0.047 |
Token | throw |
Token ID | 3714 |
Feature activation | -2.94e-02 |
Pos logit contributions | |
forgotten | +0.083 |
setup | +0.083 |
unknow | +0.081 |
forgot | +0.080 |
pap | +0.080 |
Neg logit contributions | |
ing | -0.155 |
he | -0.155 |
ane | -0.153 |
ite | -0.152 |
ore | -0.141 |
Token | round |
Token ID | 2835 |
Feature activation | -6.14e-02 |
Pos logit contributions | |
ing | +0.071 |
ane | +0.066 |
ays | +0.065 |
eddy | +0.062 |
he | +0.062 |
Neg logit contributions | |
never | -0.044 |
forgot | -0.043 |
forgotten | -0.041 |
not | -0.039 |
wasted | -0.039 |
Token | and |
Token ID | 290 |
Feature activation | +4.03e-02 |
Pos logit contributions | |
ane | +0.125 |
he | +0.122 |
ing | +0.121 |
ays | +0.115 |
eddy | +0.115 |
Neg logit contributions | |
never | -0.104 |
permission | -0.099 |
forgot | -0.099 |
help | -0.099 |
grown | -0.097 |
Token | smooth |
Token ID | 7209 |
Feature activation | -5.15e-02 |
Pos logit contributions | |
supposed | +0.054 |
permission | +0.052 |
known | +0.052 |
never | +0.051 |
forgotten | +0.051 |
Neg logit contributions | |
ing | -0.092 |
ite | -0.084 |
s | -0.078 |
ays | -0.075 |
ane | -0.074 |
Token | pe |
Token ID | 613 |
Feature activation | -3.86e-02 |
Pos logit contributions | |
ane | +0.100 |
ing | +0.099 |
he | +0.096 |
ays | +0.092 |
eddy | +0.090 |
Neg logit contributions | |
never | -0.092 |
forgot | -0.090 |
permission | -0.088 |
help | -0.087 |
not | -0.085 |
Token | b |
Token ID | 65 |
Feature activation | -7.32e-02 |
Pos logit contributions | |
ing | +0.081 |
ane | +0.075 |
ays | +0.073 |
eddy | +0.070 |
he | +0.070 |
Neg logit contributions | |
never | -0.070 |
forgot | -0.068 |
forgotten | -0.066 |
not | -0.064 |
wasted | -0.064 |
Token | bles |
Token ID | 7689 |
Feature activation | -6.85e-02 |
Pos logit contributions | |
he | +0.139 |
ane | +0.132 |
ement | +0.132 |
ing | +0.129 |
eddy | +0.128 |
Neg logit contributions | |
not | -0.128 |
permission | -0.126 |
never | -0.124 |
help | -0.120 |
be | -0.117 |
Token | . |
Token ID | 13 |
Feature activation | -1.68e-02 |
Pos logit contributions | |
ing | +0.116 |
ane | +0.114 |
eddy | +0.111 |
ays | +0.110 |
ement | +0.110 |
Neg logit contributions | |
never | -0.142 |
not | -0.133 |
help | -0.133 |
forgot | -0.132 |
permission | -0.131 |
Token | They |
Token ID | 1119 |
Feature activation | +6.85e-02 |
Pos logit contributions | |
ing | +0.037 |
ane | +0.034 |
he | +0.033 |
ays | +0.032 |
ement | +0.031 |
Neg logit contributions | |
never | -0.030 |
not | -0.029 |
forgot | -0.027 |
help | -0.027 |
wished | -0.027 |
Token | are |
Token ID | 389 |
Feature activation | +5.95e-02 |
Pos logit contributions | |
setup | +0.113 |
permission | +0.113 |
pap | +0.112 |
robbers | +0.110 |
unknow | +0.107 |
Neg logit contributions | |
he | -0.130 |
ing | -0.126 |
ane | -0.123 |
ite | -0.119 |
s | -0.113 |
Token | safer |
Token ID | 14178 |
Feature activation | -1.53e-02 |
Pos logit contributions | |
unknow | +0.083 |
overs | +0.071 |
forgot | +0.071 |
wait | +0.070 |
payoff | +0.070 |
Neg logit contributions | |
ing | -0.123 |
ays | -0.109 |
at | -0.108 |
eddy | -0.107 |
on | -0.106 |
Token | weren |
Token ID | 6304 |
Feature activation | +1.32e-02 |
Pos logit contributions | |
permission | +0.060 |
pap | +0.056 |
stolen | +0.055 |
help | +0.055 |
tresp | +0.054 |
Neg logit contributions | |
ing | -0.127 |
he | -0.114 |
ane | -0.113 |
ays | -0.107 |
ed | -0.104 |
Token | 't |
Token ID | 470 |
Feature activation | +5.47e-02 |
Pos logit contributions | |
never | +0.019 |
forgot | +0.018 |
forgotten | +0.017 |
not | +0.017 |
wasted | +0.017 |
Neg logit contributions | |
ing | -0.033 |
ane | -0.031 |
ays | -0.030 |
eddy | -0.029 |
he | -0.029 |
Token | embarrassed |
Token ID | 21100 |
Feature activation | -2.42e-02 |
Pos logit contributions | |
help | +0.078 |
never | +0.077 |
permission | +0.074 |
protect | +0.073 |
unknow | +0.073 |
Neg logit contributions | |
ing | -0.125 |
ays | -0.104 |
he | -0.104 |
ane | -0.104 |
umping | -0.100 |
Token | anymore |
Token ID | 7471 |
Feature activation | -4.13e-03 |
Pos logit contributions | |
ane | +0.033 |
ays | +0.031 |
ing | +0.031 |
eddy | +0.030 |
ite | +0.029 |
Neg logit contributions | |
never | -0.061 |
forgot | -0.059 |
forgotten | -0.059 |
not | -0.058 |
grown | -0.058 |
Token | . |
Token ID | 13 |
Feature activation | -2.48e-02 |
Pos logit contributions | |
ing | +0.007 |
ane | +0.006 |
ays | +0.006 |
eddy | +0.006 |
he | +0.006 |
Neg logit contributions | |
never | -0.010 |
forgot | -0.009 |
forgotten | -0.009 |
not | -0.009 |
help | -0.009 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -8.24e-02 |
Pos logit contributions | |
ing | +0.046 |
ane | +0.042 |
ays | +0.041 |
eddy | +0.040 |
he | +0.039 |
Neg logit contributions | |
never | -0.051 |
forgot | -0.050 |
forgotten | -0.048 |
not | -0.048 |
wasted | -0.047 |
Token | Once |
Token ID | 7454 |
Feature activation | -4.30e-02 |
Pos logit contributions | |
ing | +0.180 |
ays | +0.176 |
ane | +0.175 |
ement | +0.173 |
ooming | +0.171 |
Neg logit contributions | |
been | -0.122 |
never | -0.115 |
had | -0.112 |
grown | -0.107 |
have | -0.106 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.06e-02 |
Pos logit contributions | |
ing | +0.100 |
ays | +0.095 |
eddy | +0.094 |
ane | +0.093 |
umping | +0.089 |
Neg logit contributions | |
never | -0.064 |
forgot | -0.063 |
not | -0.062 |
help | -0.057 |
forgotten | -0.057 |
Token | a |
Token ID | 257 |
Feature activation | -4.18e-02 |
Pos logit contributions | |
ing | +0.031 |
ane | +0.030 |
ays | +0.030 |
eddy | +0.028 |
he | +0.028 |
Neg logit contributions | |
never | -0.010 |
forgot | -0.010 |
not | -0.009 |
forgotten | -0.009 |
help | -0.009 |
Token | time |
Token ID | 640 |
Feature activation | -4.95e-02 |
Pos logit contributions | |
ing | +0.113 |
ane | +0.108 |
ays | +0.107 |
eddy | +0.102 |
he | +0.098 |
Neg logit contributions | |
never | -0.054 |
forgot | -0.051 |
not | -0.048 |
forgotten | -0.048 |
help | -0.045 |
Token | , |
Token ID | 11 |
Feature activation | -2.01e-02 |
Pos logit contributions | |
ays | +0.107 |
eddy | +0.106 |
ing | +0.103 |
ane | +0.100 |
he | +0.095 |
Neg logit contributions | |
not | -0.090 |
never | -0.088 |
help | -0.080 |
be | -0.079 |
knocked | -0.078 |
Token | lived |
Token ID | 5615 |
Feature activation | -9.28e-03 |
Pos logit contributions | |
ing | +0.073 |
eddy | +0.070 |
ays | +0.069 |
ane | +0.068 |
he | +0.064 |
Neg logit contributions | |
never | -0.085 |
forgot | -0.083 |
not | -0.080 |
forgotten | -0.077 |
grown | -0.077 |
Token | a |
Token ID | 257 |
Feature activation | +1.17e-02 |
Pos logit contributions | |
ing | +0.027 |
ane | +0.025 |
ays | +0.025 |
eddy | +0.024 |
he | +0.024 |
Neg logit contributions | |
never | -0.010 |
forgot | -0.009 |
forgotten | -0.009 |
not | -0.009 |
help | -0.008 |
Token | tiny |
Token ID | 7009 |
Feature activation | -1.91e-02 |
Pos logit contributions | |
permission | +0.017 |
never | +0.016 |
unknow | +0.016 |
been | +0.016 |
forgotten | +0.016 |
Neg logit contributions | |
ing | -0.027 |
ane | -0.025 |
eddy | -0.025 |
he | -0.024 |
ite | -0.023 |
Token | , |
Token ID | 11 |
Feature activation | +1.54e-02 |
Pos logit contributions | |
ing | +0.045 |
ane | +0.043 |
ays | +0.042 |
eddy | +0.040 |
he | +0.040 |
Neg logit contributions | |
never | -0.029 |
forgot | -0.028 |
forgotten | -0.027 |
not | -0.026 |
wasted | -0.026 |
Token | friendly |
Token ID | 8030 |
Feature activation | -1.69e-02 |
Pos logit contributions | |
never | +0.020 |
forgot | +0.019 |
forgotten | +0.018 |
permission | +0.018 |
known | +0.018 |
Neg logit contributions | |
ing | -0.040 |
ane | -0.038 |
ays | -0.036 |
he | -0.036 |
eddy | -0.036 |
Token | bug |
Token ID | 5434 |
Feature activation | -6.88e-02 |
Pos logit contributions | |
ing | +0.041 |
ane | +0.039 |
ays | +0.039 |
eddy | +0.037 |
he | +0.037 |
Neg logit contributions | |
never | -0.024 |
forgot | -0.023 |
forgotten | -0.022 |
not | -0.022 |
help | -0.022 |
Token | named |
Token ID | 3706 |
Feature activation | -3.75e-03 |
Pos logit contributions | |
ing | +0.143 |
ane | +0.135 |
ays | +0.134 |
he | +0.128 |
ement | +0.126 |
Neg logit contributions | |
never | -0.130 |
forgot | -0.121 |
not | -0.118 |
forgotten | -0.114 |
help | -0.114 |
Token | Hug |
Token ID | 12905 |
Feature activation | -5.12e-02 |
Pos logit contributions | |
ing | +0.009 |
ane | +0.008 |
ays | +0.008 |
eddy | +0.008 |
he | +0.008 |
Neg logit contributions | |
never | -0.006 |
forgot | -0.006 |
forgotten | -0.005 |
wasted | -0.005 |
known | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | -5.63e-02 |
Pos logit contributions | |
ing | +0.097 |
ane | +0.091 |
ays | +0.090 |
he | +0.085 |
eddy | +0.084 |
Neg logit contributions | |
never | -0.102 |
forgot | -0.099 |
forgotten | -0.095 |
not | -0.095 |
wasted | -0.092 |
Token | Hug |
Token ID | 12905 |
Feature activation | -2.69e-02 |
Pos logit contributions | |
ing | +0.122 |
ane | +0.113 |
ays | +0.113 |
ement | +0.104 |
eddy | +0.103 |
Neg logit contributions | |
never | -0.099 |
not | -0.094 |
help | -0.094 |
grown | -0.091 |
always | -0.091 |
Token | was |
Token ID | 373 |
Feature activation | +4.23e-02 |
Pos logit contributions | |
ing | +0.065 |
ane | +0.061 |
ays | +0.059 |
eddy | +0.057 |
he | +0.057 |
Neg logit contributions | |
never | -0.039 |
forgot | -0.038 |
forgotten | -0.037 |
permission | -0.036 |
wasted | -0.036 |
Token | 'm |
Token ID | 1101 |
Feature activation | +3.54e-02 |
Pos logit contributions | |
permission | +0.045 |
setup | +0.044 |
never | +0.043 |
forgotten | +0.043 |
unknow | +0.042 |
Neg logit contributions | |
ing | -0.090 |
he | -0.085 |
ane | -0.083 |
eddy | -0.082 |
ays | -0.079 |
Token | sorry |
Token ID | 7926 |
Feature activation | -3.33e-02 |
Pos logit contributions | |
forgot | +0.053 |
setup | +0.049 |
forgotten | +0.049 |
unknow | +0.048 |
permission | +0.047 |
Neg logit contributions | |
ing | -0.081 |
he | -0.074 |
ane | -0.073 |
eddy | -0.073 |
ays | -0.072 |
Token | I |
Token ID | 314 |
Feature activation | +4.72e-02 |
Pos logit contributions | |
ane | +0.084 |
he | +0.083 |
ays | +0.082 |
ing | +0.080 |
ement | +0.078 |
Neg logit contributions | |
not | -0.048 |
never | -0.046 |
help | -0.046 |
be | -0.043 |
grown | -0.041 |
Token | dropped |
Token ID | 5710 |
Feature activation | -2.32e-02 |
Pos logit contributions | |
permission | +0.055 |
never | +0.055 |
setup | +0.053 |
pap | +0.052 |
forgotten | +0.052 |
Neg logit contributions | |
ing | -0.120 |
he | -0.112 |
ane | -0.112 |
ays | -0.109 |
eddy | -0.108 |
Token | the |
Token ID | 262 |
Feature activation | -8.66e-03 |
Pos logit contributions | |
ing | +0.064 |
ane | +0.061 |
ays | +0.060 |
eddy | +0.058 |
he | +0.058 |
Neg logit contributions | |
never | -0.026 |
forgot | -0.025 |
forgotten | -0.024 |
not | -0.023 |
help | -0.022 |
Token | bag |
Token ID | 6131 |
Feature activation | -7.63e-02 |
Pos logit contributions | |
ing | +0.018 |
ane | +0.017 |
ays | +0.016 |
eddy | +0.016 |
he | +0.016 |
Neg logit contributions | |
never | -0.016 |
forgot | -0.016 |
forgotten | -0.015 |
known | -0.015 |
not | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | +4.63e-02 |
Pos logit contributions | |
ing | +0.147 |
ays | +0.142 |
ane | +0.142 |
ement | +0.138 |
eddy | +0.130 |
Neg logit contributions | |
never | -0.153 |
not | -0.142 |
help | -0.137 |
forgotten | -0.133 |
forgot | -0.132 |
Token | the |
Token ID | 262 |
Feature activation | +9.59e-03 |
Pos logit contributions | |
known | +0.071 |
supposed | +0.069 |
tresp | +0.068 |
unknow | +0.067 |
forgotten | +0.067 |
Neg logit contributions | |
ing | -0.088 |
ite | -0.085 |
eddy | -0.081 |
he | -0.079 |
ane | -0.078 |
Token | gum |
Token ID | 27142 |
Feature activation | -7.04e-02 |
Pos logit contributions | |
never | +0.017 |
forgot | +0.017 |
known | +0.016 |
forgotten | +0.016 |
unknow | +0.015 |
Neg logit contributions | |
ing | -0.020 |
ane | -0.018 |
eddy | -0.018 |
he | -0.017 |
ays | -0.017 |
Token | ." |
Token ID | 526 |
Feature activation | -5.37e-02 |
Pos logit contributions | |
ing | +0.121 |
ane | +0.114 |
ays | +0.110 |
ement | +0.106 |
eddy | +0.105 |
Neg logit contributions | |
never | -0.153 |
forgot | -0.146 |
not | -0.143 |
forgotten | -0.141 |
help | -0.139 |
Token |
|
Token ID | 198 |
Feature activation | -4.24e-02 |
Pos logit contributions | |
ane | +0.136 |
ing | +0.134 |
ays | +0.129 |
ement | +0.124 |
eddy | +0.118 |
Neg logit contributions | |
help | -0.079 |
not | -0.077 |
grown | -0.075 |
never | -0.072 |
be | -0.070 |
Token | want |
Token ID | 765 |
Feature activation | +3.55e-03 |
Pos logit contributions | |
pap | +0.059 |
unknow | +0.057 |
forgot | +0.054 |
setup | +0.054 |
forgotten | +0.054 |
Neg logit contributions | |
ing | -0.119 |
he | -0.116 |
ane | -0.112 |
eddy | -0.106 |
ite | -0.104 |
Token | to |
Token ID | 284 |
Feature activation | +4.43e-02 |
Pos logit contributions | |
forgot | +0.006 |
never | +0.006 |
forgotten | +0.006 |
known | +0.006 |
unknow | +0.006 |
Neg logit contributions | |
ing | -0.007 |
ane | -0.007 |
ays | -0.007 |
eddy | -0.006 |
he | -0.006 |
Token | come |
Token ID | 1282 |
Feature activation | -1.86e-02 |
Pos logit contributions | |
unknow | +0.043 |
forgot | +0.043 |
forgotten | +0.043 |
never | +0.041 |
setup | +0.041 |
Neg logit contributions | |
ing | -0.117 |
he | -0.113 |
ane | -0.111 |
eddy | -0.111 |
ays | -0.106 |
Token | to |
Token ID | 284 |
Feature activation | +2.57e-02 |
Pos logit contributions | |
ement | +0.046 |
ing | +0.044 |
ane | +0.042 |
ays | +0.042 |
he | +0.040 |
Neg logit contributions | |
never | -0.030 |
not | -0.030 |
be | -0.028 |
help | -0.028 |
find | -0.027 |
Token | your |
Token ID | 534 |
Feature activation | +1.70e-02 |
Pos logit contributions | |
forgot | +0.037 |
never | +0.036 |
forgotten | +0.035 |
unknow | +0.034 |
known | +0.033 |
Neg logit contributions | |
ing | -0.057 |
eddy | -0.055 |
he | -0.054 |
ane | -0.054 |
ays | -0.053 |
Token | party |
Token ID | 2151 |
Feature activation | -7.32e-02 |
Pos logit contributions | |
never | +0.024 |
forgot | +0.023 |
forgotten | +0.022 |
wasted | +0.021 |
not | +0.021 |
Neg logit contributions | |
ing | -0.042 |
ane | -0.040 |
ays | -0.039 |
eddy | -0.038 |
he | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | -1.05e-02 |
Pos logit contributions | |
ing | +0.151 |
ays | +0.144 |
ane | +0.143 |
ement | +0.134 |
eddy | +0.134 |
Neg logit contributions | |
never | -0.140 |
not | -0.134 |
forgot | -0.130 |
forgotten | -0.123 |
help | -0.123 |
Token | Goodbye |
Token ID | 44442 |
Feature activation | -5.19e-02 |
Pos logit contributions | |
ing | +0.025 |
ane | +0.023 |
ays | +0.022 |
he | +0.022 |
eddy | +0.021 |
Neg logit contributions | |
never | -0.017 |
forgot | -0.016 |
not | -0.016 |
forgotten | -0.015 |
help | -0.015 |
Token | , |
Token ID | 11 |
Feature activation | +2.67e-02 |
Pos logit contributions | |
ing | +0.096 |
ane | +0.096 |
ays | +0.096 |
he | +0.095 |
ement | +0.089 |
Neg logit contributions | |
never | -0.102 |
forgot | -0.100 |
not | -0.100 |
help | -0.098 |
grown | -0.095 |
Token | Lily |
Token ID | 20037 |
Feature activation | -2.54e-02 |
Pos logit contributions | |
never | +0.046 |
forgotten | +0.045 |
known | +0.045 |
wasted | +0.044 |
stolen | +0.044 |
Neg logit contributions | |
ing | -0.052 |
ane | -0.048 |
ay | -0.046 |
eddy | -0.046 |
ays | -0.046 |
Token | ." |
Token ID | 526 |
Feature activation | -4.88e-02 |
Pos logit contributions | |
ing | +0.050 |
ays | +0.048 |
ane | +0.047 |
ement | +0.045 |
eddy | +0.044 |
Neg logit contributions | |
never | -0.051 |
forgot | -0.047 |
not | -0.046 |
forgotten | -0.044 |
always | -0.044 |