Token | and |
Token ID | 290 |
Feature activation | +5.48e-02 |
Pos logit contributions | |
crow | +0.908 |
Hey | +0.893 |
Leave | +0.875 |
Stop | +0.870 |
Bad | +0.845 |
Neg logit contributions | |
pron | -0.547 |
favourite | -0.492 |
favorite | -0.475 |
collection | -0.459 |
masterpiece | -0.453 |
Token | makes |
Token ID | 1838 |
Feature activation | +2.09e-01 |
Pos logit contributions | |
crow | +0.004 |
vious | +0.004 |
Leave | +0.004 |
eper | +0.004 |
Hey | +0.004 |
Neg logit contributions | |
pron | -0.004 |
favourite | -0.004 |
masterpiece | -0.004 |
usual | -0.004 |
favorite | -0.004 |
Token | them |
Token ID | 606 |
Feature activation | +2.61e+00 |
Pos logit contributions | |
crow | +0.014 |
Leave | +0.014 |
Bad | +0.013 |
eper | +0.013 |
vious | +0.012 |
Neg logit contributions | |
pron | -0.017 |
favourite | -0.017 |
masterpiece | -0.017 |
favorite | -0.016 |
brand | -0.016 |
Token | cry |
Token ID | 3960 |
Feature activation | +2.90e+00 |
Pos logit contributions | |
crow | +0.222 |
Leave | +0.205 |
Bad | +0.201 |
Stop | +0.201 |
Hey | +0.199 |
Neg logit contributions | |
pron | -0.189 |
favourite | -0.182 |
masterpiece | -0.175 |
favorite | -0.168 |
collection | -0.162 |
Token | . |
Token ID | 13 |
Feature activation | +1.89e-01 |
Pos logit contributions | |
crow | +0.287 |
Leave | +0.279 |
Bad | +0.271 |
Hey | +0.268 |
Stop | +0.264 |
Neg logit contributions | |
pron | -0.144 |
favourite | -0.143 |
favorite | -0.140 |
masterpiece | -0.129 |
daddy | -0.121 |
Token | War |
Token ID | 1810 |
Feature activation | +1.56e+01 |
Pos logit contributions | |
crow | +0.013 |
vious | +0.012 |
eper | +0.012 |
Leave | +0.012 |
Stop | +0.011 |
Neg logit contributions | |
favourite | -0.014 |
favorite | -0.013 |
pron | -0.013 |
Daddy | -0.013 |
Grand | -0.012 |
Token | is |
Token ID | 318 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
crow | +1.312 |
Leave | +1.272 |
Bad | +1.227 |
eper | +1.198 |
hunters | +1.166 |
Neg logit contributions | |
pron | -0.964 |
favourite | -0.922 |
favorite | -0.913 |
masterpiece | -0.846 |
couldn | -0.833 |
Token | not |
Token ID | 407 |
Feature activation | -2.90e+00 |
Pos logit contributions | |
crow | +0.115 |
Leave | +0.108 |
vious | +0.103 |
Stop | +0.103 |
Bad | +0.102 |
Neg logit contributions | |
pron | -0.090 |
favourite | -0.084 |
masterpiece | -0.082 |
favorite | -0.077 |
brand | -0.076 |
Token | fun |
Token ID | 1257 |
Feature activation | -6.78e-01 |
Pos logit contributions | |
pron | +0.194 |
favourite | +0.176 |
masterpiece | +0.170 |
favorite | +0.162 |
brand | +0.159 |
Neg logit contributions | |
crow | -0.265 |
Leave | -0.253 |
Bad | -0.240 |
vious | -0.239 |
Stop | -0.238 |
Token | . |
Token ID | 13 |
Feature activation | +2.03e-01 |
Pos logit contributions | |
pron | +0.045 |
favourite | +0.042 |
masterpiece | +0.041 |
favorite | +0.039 |
collection | +0.038 |
Neg logit contributions | |
crow | -0.059 |
Leave | -0.058 |
Bad | -0.057 |
Stop | -0.054 |
vious | -0.054 |
Token | You |
Token ID | 921 |
Feature activation | +5.37e+00 |
Pos logit contributions | |
crow | +0.011 |
eper | +0.010 |
Leave | +0.010 |
vious | +0.010 |
hunters | +0.009 |
Neg logit contributions | |
favourite | -0.018 |
favorite | -0.017 |
pron | -0.017 |
masterpiece | -0.016 |
Daddy | -0.016 |
Token | and |
Token ID | 290 |
Feature activation | +4.42e-02 |
Pos logit contributions | |
crow | +0.201 |
Bad | +0.199 |
Leave | +0.199 |
eper | +0.185 |
Hey | +0.185 |
Neg logit contributions | |
favourite | -0.113 |
pron | -0.107 |
favorite | -0.105 |
masterpiece | -0.102 |
collection | -0.098 |
Token | scary |
Token ID | 14343 |
Feature activation | +2.70e+00 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.003 |
eper | +0.003 |
Hey | +0.003 |
vious | +0.003 |
Neg logit contributions | |
pron | -0.003 |
favourite | -0.003 |
masterpiece | -0.003 |
favorite | -0.003 |
brand | -0.003 |
Token | and |
Token ID | 290 |
Feature activation | +4.22e-02 |
Pos logit contributions | |
crow | +0.261 |
Leave | +0.259 |
Bad | +0.254 |
Hey | +0.245 |
vious | +0.243 |
Neg logit contributions | |
favourite | -0.124 |
favorite | -0.117 |
pron | -0.114 |
masterpiece | -0.113 |
collection | -0.110 |
Token | sad |
Token ID | 6507 |
Feature activation | +3.73e+00 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.003 |
eper | +0.003 |
Bad | +0.003 |
vious | +0.003 |
Neg logit contributions | |
pron | -0.003 |
favourite | -0.003 |
masterpiece | -0.003 |
favorite | -0.003 |
usual | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
crow | +0.350 |
Leave | +0.346 |
Bad | +0.336 |
vious | +0.328 |
eper | +0.327 |
Neg logit contributions | |
favourite | -0.187 |
favorite | -0.179 |
pron | -0.178 |
masterpiece | -0.172 |
collection | -0.166 |
Token | War |
Token ID | 1810 |
Feature activation | +1.56e+01 |
Pos logit contributions | |
crow | +0.011 |
vious | +0.010 |
Leave | +0.010 |
eper | +0.010 |
Stop | +0.009 |
Neg logit contributions | |
favourite | -0.013 |
favorite | -0.012 |
Daddy | -0.011 |
Grand | -0.011 |
Mam | -0.011 |
Token | hurts |
Token ID | 20406 |
Feature activation | +8.91e-01 |
Pos logit contributions | |
crow | +1.297 |
Leave | +1.270 |
Bad | +1.222 |
eper | +1.181 |
vious | +1.157 |
Neg logit contributions | |
pron | -0.981 |
favourite | -0.944 |
favorite | -0.923 |
masterpiece | -0.862 |
couldn | -0.842 |
Token | people |
Token ID | 661 |
Feature activation | +9.69e+00 |
Pos logit contributions | |
Leave | +0.074 |
vious | +0.071 |
crow | +0.071 |
eper | +0.068 |
Bad | +0.067 |
Neg logit contributions | |
favourite | -0.048 |
pron | -0.048 |
favorite | -0.047 |
usual | -0.044 |
couldn | -0.043 |
Token | and |
Token ID | 290 |
Feature activation | +5.48e-02 |
Pos logit contributions | |
crow | +0.908 |
Hey | +0.893 |
Leave | +0.875 |
Stop | +0.870 |
Bad | +0.845 |
Neg logit contributions | |
pron | -0.547 |
favourite | -0.492 |
favorite | -0.475 |
collection | -0.459 |
masterpiece | -0.453 |
Token | makes |
Token ID | 1838 |
Feature activation | +2.09e-01 |
Pos logit contributions | |
crow | +0.004 |
vious | +0.004 |
Leave | +0.004 |
eper | +0.004 |
Hey | +0.004 |
Neg logit contributions | |
pron | -0.004 |
favourite | -0.004 |
masterpiece | -0.004 |
usual | -0.004 |
favorite | -0.004 |
Token | them |
Token ID | 606 |
Feature activation | +2.61e+00 |
Pos logit contributions | |
crow | +0.014 |
Leave | +0.014 |
Bad | +0.013 |
eper | +0.013 |
vious | +0.012 |
Neg logit contributions | |
pron | -0.017 |
favourite | -0.017 |
masterpiece | -0.017 |
favorite | -0.016 |
brand | -0.016 |
Token | is |
Token ID | 318 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
crow | +1.326 |
Leave | +1.284 |
Bad | +1.246 |
eper | +1.207 |
vious | +1.178 |
Neg logit contributions | |
pron | -0.962 |
favourite | -0.913 |
favorite | -0.889 |
masterpiece | -0.832 |
couldn | -0.811 |
Token | not |
Token ID | 407 |
Feature activation | -2.96e+00 |
Pos logit contributions | |
crow | +0.113 |
Leave | +0.107 |
vious | +0.102 |
Bad | +0.102 |
Stop | +0.102 |
Neg logit contributions | |
pron | -0.081 |
favourite | -0.075 |
masterpiece | -0.073 |
favorite | -0.067 |
brand | -0.067 |
Token | a |
Token ID | 257 |
Feature activation | +2.10e-01 |
Pos logit contributions | |
pron | +0.201 |
favourite | +0.182 |
masterpiece | +0.178 |
favorite | +0.165 |
usual | +0.165 |
Neg logit contributions | |
crow | -0.267 |
Leave | -0.256 |
Bad | -0.246 |
vious | -0.242 |
Stop | -0.242 |
Token | game |
Token ID | 983 |
Feature activation | +2.34e+00 |
Pos logit contributions | |
crow | +0.017 |
Leave | +0.016 |
Bad | +0.016 |
vious | +0.015 |
Stop | +0.015 |
Neg logit contributions | |
pron | -0.016 |
masterpiece | -0.015 |
favourite | -0.015 |
collection | -0.014 |
brand | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
crow | +0.203 |
Leave | +0.200 |
Bad | +0.198 |
vious | +0.188 |
Stop | +0.187 |
Neg logit contributions | |
pron | -0.153 |
favourite | -0.144 |
masterpiece | -0.140 |
collection | -0.132 |
usual | -0.131 |
Token | War |
Token ID | 1810 |
Feature activation | +1.55e+01 |
Pos logit contributions | |
crow | +0.010 |
Leave | +0.009 |
eper | +0.009 |
vious | +0.009 |
Stop | +0.009 |
Neg logit contributions | |
favourite | -0.012 |
favorite | -0.011 |
pron | -0.011 |
masterpiece | -0.010 |
Daddy | -0.010 |
Token | is |
Token ID | 318 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
crow | +1.280 |
Leave | +1.250 |
Bad | +1.210 |
eper | +1.164 |
vious | +1.148 |
Neg logit contributions | |
pron | -1.026 |
favourite | -0.978 |
favorite | -0.941 |
masterpiece | -0.905 |
usual | -0.874 |
Token | bad |
Token ID | 2089 |
Feature activation | +2.18e+00 |
Pos logit contributions | |
crow | +0.116 |
Leave | +0.110 |
Bad | +0.106 |
vious | +0.105 |
Stop | +0.105 |
Neg logit contributions | |
pron | -0.084 |
favourite | -0.077 |
masterpiece | -0.076 |
favorite | -0.070 |
usual | -0.069 |
Token | and |
Token ID | 290 |
Feature activation | +4.42e-02 |
Pos logit contributions | |
crow | +0.201 |
Bad | +0.199 |
Leave | +0.199 |
eper | +0.185 |
Hey | +0.185 |
Neg logit contributions | |
favourite | -0.113 |
pron | -0.107 |
favorite | -0.105 |
masterpiece | -0.102 |
collection | -0.098 |
Token | scary |
Token ID | 14343 |
Feature activation | +2.70e+00 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.003 |
eper | +0.003 |
Hey | +0.003 |
vious | +0.003 |
Neg logit contributions | |
pron | -0.003 |
favourite | -0.003 |
masterpiece | -0.003 |
favorite | -0.003 |
brand | -0.003 |
Token | and |
Token ID | 290 |
Feature activation | +4.22e-02 |
Pos logit contributions | |
crow | +0.261 |
Leave | +0.259 |
Bad | +0.254 |
Hey | +0.245 |
vious | +0.243 |
Neg logit contributions | |
favourite | -0.124 |
favorite | -0.117 |
pron | -0.114 |
masterpiece | -0.113 |
collection | -0.110 |
Token | oys |
Token ID | 19417 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
Leave | +0.531 |
crow | +0.527 |
Bad | +0.522 |
Stop | +0.493 |
vious | +0.488 |
Neg logit contributions | |
pron | -0.450 |
favourite | -0.416 |
masterpiece | -0.403 |
ail | -0.394 |
collection | -0.380 |
Token | , |
Token ID | 11 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
crow | +0.168 |
Leave | +0.164 |
Bad | +0.162 |
Stop | +0.155 |
vious | +0.154 |
Neg logit contributions | |
pron | -0.127 |
favourite | -0.114 |
masterpiece | -0.112 |
collection | -0.103 |
ail | -0.101 |
Token | stop |
Token ID | 2245 |
Feature activation | +3.86e+00 |
Pos logit contributions | |
crow | +0.083 |
Leave | +0.079 |
Bad | +0.078 |
vious | +0.076 |
Stop | +0.075 |
Neg logit contributions | |
pron | -0.107 |
favourite | -0.098 |
masterpiece | -0.097 |
collection | -0.090 |
usual | -0.089 |
Token | it |
Token ID | 340 |
Feature activation | +7.12e-02 |
Pos logit contributions | |
crow | +0.331 |
Leave | +0.313 |
Bad | +0.308 |
vious | +0.305 |
Stop | +0.296 |
Neg logit contributions | |
pron | -0.274 |
masterpiece | -0.249 |
favourite | -0.246 |
collection | -0.221 |
brand | -0.219 |
Token | . |
Token ID | 13 |
Feature activation | +1.42e-01 |
Pos logit contributions | |
crow | +0.006 |
Leave | +0.006 |
Bad | +0.006 |
Stop | +0.006 |
vious | +0.006 |
Neg logit contributions | |
pron | -0.005 |
masterpiece | -0.004 |
favourite | -0.004 |
collection | -0.004 |
usual | -0.004 |
Token | War |
Token ID | 1810 |
Feature activation | +1.55e+01 |
Pos logit contributions | |
Leave | +0.008 |
crow | +0.008 |
Bad | +0.008 |
vious | +0.008 |
Stop | +0.008 |
Neg logit contributions | |
pron | -0.012 |
favourite | -0.012 |
masterpiece | -0.011 |
favorite | -0.011 |
bedroom | -0.011 |
Token | is |
Token ID | 318 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
crow | +1.326 |
Leave | +1.284 |
Bad | +1.246 |
eper | +1.207 |
vious | +1.178 |
Neg logit contributions | |
pron | -0.962 |
favourite | -0.913 |
favorite | -0.889 |
masterpiece | -0.832 |
couldn | -0.811 |
Token | not |
Token ID | 407 |
Feature activation | -2.96e+00 |
Pos logit contributions | |
crow | +0.113 |
Leave | +0.107 |
vious | +0.102 |
Bad | +0.102 |
Stop | +0.102 |
Neg logit contributions | |
pron | -0.081 |
favourite | -0.075 |
masterpiece | -0.073 |
favorite | -0.067 |
brand | -0.067 |
Token | a |
Token ID | 257 |
Feature activation | +2.10e-01 |
Pos logit contributions | |
pron | +0.201 |
favourite | +0.182 |
masterpiece | +0.178 |
favorite | +0.165 |
usual | +0.165 |
Neg logit contributions | |
crow | -0.267 |
Leave | -0.256 |
Bad | -0.246 |
vious | -0.242 |
Stop | -0.242 |
Token | game |
Token ID | 983 |
Feature activation | +2.34e+00 |
Pos logit contributions | |
crow | +0.017 |
Leave | +0.016 |
Bad | +0.016 |
vious | +0.015 |
Stop | +0.015 |
Neg logit contributions | |
pron | -0.016 |
masterpiece | -0.015 |
favourite | -0.015 |
collection | -0.014 |
brand | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
crow | +0.203 |
Leave | +0.200 |
Bad | +0.198 |
vious | +0.188 |
Stop | +0.187 |
Neg logit contributions | |
pron | -0.153 |
favourite | -0.144 |
masterpiece | -0.140 |
collection | -0.132 |
usual | -0.131 |
Token | I |
Token ID | 314 |
Feature activation | -2.43e+00 |
Pos logit contributions | |
Bad | +0.008 |
Leave | +0.008 |
crow | +0.008 |
vious | +0.008 |
Stop | +0.008 |
Neg logit contributions | |
pron | -0.010 |
favourite | -0.010 |
masterpiece | -0.009 |
favorite | -0.009 |
bedroom | -0.009 |
Token | teach |
Token ID | 4545 |
Feature activation | +3.89e+00 |
Pos logit contributions | |
pron | +0.206 |
favourite | +0.190 |
masterpiece | +0.187 |
collection | +0.174 |
favorite | +0.170 |
Neg logit contributions | |
crow | -0.175 |
Leave | -0.169 |
Bad | -0.167 |
Stop | -0.159 |
vious | -0.157 |
Token | about |
Token ID | 546 |
Feature activation | -2.11e+00 |
Pos logit contributions | |
crow | +0.225 |
Leave | +0.216 |
Bad | +0.213 |
vious | +0.200 |
Stop | +0.199 |
Neg logit contributions | |
pron | -0.382 |
favourite | -0.360 |
masterpiece | -0.356 |
collection | -0.331 |
usual | -0.330 |
Token | war |
Token ID | 1175 |
Feature activation | +6.84e+00 |
Pos logit contributions | |
pron | +0.194 |
masterpiece | +0.177 |
favourite | +0.177 |
collection | +0.163 |
ail | +0.160 |
Neg logit contributions | |
crow | -0.138 |
Bad | -0.133 |
Leave | -0.131 |
Stop | -0.126 |
vious | -0.122 |
Token | . |
Token ID | 13 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
Leave | +0.594 |
crow | +0.585 |
Bad | +0.579 |
Stop | +0.546 |
vious | +0.541 |
Neg logit contributions | |
pron | -0.467 |
favourite | -0.450 |
masterpiece | -0.430 |
favorite | -0.404 |
collection | -0.402 |
Token | War |
Token ID | 1810 |
Feature activation | +1.55e+01 |
Pos logit contributions | |
crow | +0.007 |
Leave | +0.007 |
Bad | +0.007 |
vious | +0.007 |
Stop | +0.007 |
Neg logit contributions | |
favourite | -0.008 |
pron | -0.008 |
masterpiece | -0.008 |
favorite | -0.007 |
bedroom | -0.007 |
Token | is |
Token ID | 318 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
crow | +1.209 |
Leave | +1.208 |
Bad | +1.163 |
vious | +1.091 |
eper | +1.074 |
Neg logit contributions | |
pron | -1.141 |
favourite | -1.088 |
masterpiece | -1.044 |
favorite | -1.017 |
collection | -0.965 |
Token | when |
Token ID | 618 |
Feature activation | +6.78e+00 |
Pos logit contributions | |
crow | +0.106 |
Leave | +0.101 |
Bad | +0.097 |
vious | +0.096 |
Stop | +0.096 |
Neg logit contributions | |
pron | -0.083 |
favourite | -0.077 |
masterpiece | -0.076 |
brand | -0.068 |
favorite | -0.068 |
Token | people |
Token ID | 661 |
Feature activation | +9.60e+00 |
Pos logit contributions | |
crow | +0.502 |
Leave | +0.481 |
Bad | +0.477 |
Stop | +0.458 |
vious | +0.456 |
Neg logit contributions | |
pron | -0.562 |
favourite | -0.518 |
masterpiece | -0.515 |
collection | -0.473 |
usual | -0.467 |
Token | fight |
Token ID | 1907 |
Feature activation | +4.75e+00 |
Pos logit contributions | |
crow | +0.744 |
Leave | +0.717 |
Bad | +0.704 |
vious | +0.654 |
eper | +0.649 |
Neg logit contributions | |
pron | -0.725 |
favourite | -0.657 |
masterpiece | -0.650 |
collection | -0.631 |
usual | -0.617 |
Token | . |
Token ID | 13 |
Feature activation | +8.82e-02 |
Pos logit contributions | |
crow | +0.410 |
Leave | +0.408 |
Bad | +0.386 |
vious | +0.376 |
eper | +0.361 |
Neg logit contributions | |
pron | -0.308 |
favourite | -0.290 |
masterpiece | -0.287 |
favorite | -0.270 |
collection | -0.266 |
Token | , |
Token ID | 11 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
pron | +0.100 |
masterpiece | +0.093 |
favourite | +0.090 |
usual | +0.081 |
collection | +0.081 |
Neg logit contributions | |
crow | -0.095 |
Leave | -0.091 |
Bad | -0.090 |
Stop | -0.086 |
vious | -0.085 |
Token | " |
Token ID | 366 |
Feature activation | -1.87e-01 |
Pos logit contributions | |
Leave | +0.086 |
crow | +0.085 |
Bad | +0.084 |
Stop | +0.080 |
ting | +0.077 |
Neg logit contributions | |
pron | -0.086 |
masterpiece | -0.083 |
favourite | -0.082 |
collection | -0.075 |
usual | -0.073 |
Token | Me |
Token ID | 5308 |
Feature activation | +1.56e+00 |
Pos logit contributions | |
pron | +0.018 |
favourite | +0.016 |
masterpiece | +0.016 |
usual | +0.015 |
brand | +0.014 |
Neg logit contributions | |
crow | -0.012 |
ases | -0.011 |
Leave | -0.011 |
vious | -0.011 |
Bad | -0.011 |
Token | too |
Token ID | 1165 |
Feature activation | +5.19e+00 |
Pos logit contributions | |
crow | +0.094 |
Leave | +0.090 |
Bad | +0.089 |
Stop | +0.085 |
vious | +0.083 |
Neg logit contributions | |
pron | -0.155 |
favourite | -0.144 |
masterpiece | -0.143 |
collection | -0.133 |
brand | -0.131 |
Token | . |
Token ID | 13 |
Feature activation | +4.93e-02 |
Pos logit contributions | |
crow | +0.400 |
Leave | +0.384 |
Bad | +0.383 |
Stop | +0.367 |
vious | +0.364 |
Neg logit contributions | |
pron | -0.421 |
favourite | -0.384 |
masterpiece | -0.383 |
collection | -0.349 |
usual | -0.342 |
Token | War |
Token ID | 1810 |
Feature activation | +1.54e+01 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.003 |
Bad | +0.003 |
Stop | +0.003 |
vious | +0.003 |
Neg logit contributions | |
pron | -0.004 |
favourite | -0.004 |
masterpiece | -0.004 |
collection | -0.004 |
brand | -0.004 |
Token | is |
Token ID | 318 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
crow | +1.226 |
Leave | +1.195 |
Bad | +1.160 |
vious | +1.102 |
eper | +1.098 |
Neg logit contributions | |
pron | -1.133 |
favourite | -1.079 |
masterpiece | -1.023 |
favorite | -0.994 |
collection | -0.960 |
Token | bad |
Token ID | 2089 |
Feature activation | +2.07e+00 |
Pos logit contributions | |
crow | +0.093 |
Leave | +0.090 |
Bad | +0.089 |
Stop | +0.086 |
vious | +0.085 |
Neg logit contributions | |
pron | -0.088 |
favourite | -0.079 |
masterpiece | -0.079 |
collection | -0.072 |
usual | -0.070 |
Token | . |
Token ID | 13 |
Feature activation | +4.12e-02 |
Pos logit contributions | |
crow | +0.177 |
Leave | +0.175 |
Bad | +0.174 |
Stop | +0.164 |
vious | +0.161 |
Neg logit contributions | |
pron | -0.137 |
favourite | -0.134 |
masterpiece | -0.126 |
favorite | -0.120 |
collection | -0.119 |
Token | It |
Token ID | 632 |
Feature activation | +7.48e-01 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.003 |
vious | +0.003 |
Bad | +0.003 |
Stop | +0.003 |
Neg logit contributions | |
pron | -0.003 |
favourite | -0.003 |
masterpiece | -0.003 |
favorite | -0.003 |
brand | -0.003 |
Token | makes |
Token ID | 1838 |
Feature activation | +6.67e-02 |
Pos logit contributions | |
crow | +0.060 |
Leave | +0.058 |
Bad | +0.056 |
ases | +0.054 |
Stop | +0.054 |
Neg logit contributions | |
pron | -0.055 |
favourite | -0.052 |
masterpiece | -0.052 |
favorite | -0.048 |
collection | -0.048 |
Token | missile |
Token ID | 10105 |
Feature activation | +2.54e+00 |
Pos logit contributions | |
crow | +0.105 |
Leave | +0.104 |
ases | +0.099 |
vious | +0.099 |
Bad | +0.097 |
Neg logit contributions | |
pron | -0.093 |
masterpiece | -0.089 |
favourite | -0.089 |
favorite | -0.083 |
collection | -0.082 |
Token | was |
Token ID | 373 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
crow | +0.218 |
Leave | +0.217 |
Bad | +0.210 |
Stop | +0.205 |
vious | +0.199 |
Neg logit contributions | |
pron | -0.168 |
favourite | -0.158 |
masterpiece | -0.154 |
collection | -0.147 |
favorite | -0.145 |
Token | for |
Token ID | 329 |
Feature activation | -5.41e+00 |
Pos logit contributions | |
pron | +0.118 |
favourite | +0.110 |
masterpiece | +0.108 |
favorite | +0.100 |
usual | +0.100 |
Neg logit contributions | |
crow | -0.129 |
Leave | -0.123 |
Stop | -0.119 |
Bad | -0.117 |
vious | -0.117 |
Token | war |
Token ID | 1175 |
Feature activation | +6.78e+00 |
Pos logit contributions | |
pron | +0.427 |
favourite | +0.402 |
masterpiece | +0.402 |
usual | +0.376 |
brand | +0.366 |
Neg logit contributions | |
crow | -0.416 |
Leave | -0.386 |
vious | -0.378 |
Bad | -0.374 |
Stop | -0.365 |
Token | . |
Token ID | 13 |
Feature activation | +5.42e-02 |
Pos logit contributions | |
crow | +0.597 |
Leave | +0.594 |
Bad | +0.586 |
Stop | +0.556 |
vious | +0.542 |
Neg logit contributions | |
pron | -0.436 |
favourite | -0.417 |
masterpiece | -0.395 |
favorite | -0.379 |
collection | -0.376 |
Token | War |
Token ID | 1810 |
Feature activation | +1.54e+01 |
Pos logit contributions | |
crow | +0.004 |
Leave | +0.004 |
eper | +0.004 |
vious | +0.004 |
Stop | +0.004 |
Neg logit contributions | |
favourite | -0.004 |
pron | -0.003 |
favorite | -0.003 |
brand | -0.003 |
masterpiece | -0.003 |
Token | is |
Token ID | 318 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
crow | +1.231 |
Leave | +1.216 |
Bad | +1.169 |
eper | +1.117 |
vious | +1.108 |
Neg logit contributions | |
pron | -1.088 |
favourite | -1.038 |
favorite | -0.982 |
masterpiece | -0.976 |
usual | -0.927 |
Token | bad |
Token ID | 2089 |
Feature activation | +2.06e+00 |
Pos logit contributions | |
crow | +0.100 |
Leave | +0.095 |
vious | +0.091 |
Stop | +0.091 |
Bad | +0.090 |
Neg logit contributions | |
pron | -0.081 |
favourite | -0.076 |
masterpiece | -0.075 |
brand | -0.069 |
favorite | -0.069 |
Token | . |
Token ID | 13 |
Feature activation | +4.82e-02 |
Pos logit contributions | |
crow | +0.186 |
Leave | +0.184 |
Bad | +0.183 |
Stop | +0.171 |
eper | +0.169 |
Neg logit contributions | |
pron | -0.118 |
favourite | -0.116 |
masterpiece | -0.108 |
favorite | -0.106 |
collection | -0.101 |
Token | I |
Token ID | 314 |
Feature activation | -2.52e+00 |
Pos logit contributions | |
crow | +0.004 |
eper | +0.004 |
Leave | +0.004 |
vious | +0.004 |
ting | +0.003 |
Neg logit contributions | |
favourite | -0.003 |
favorite | -0.003 |
pron | -0.003 |
Grand | -0.003 |
Daddy | -0.003 |
Token | don |
Token ID | 836 |
Feature activation | +5.04e+00 |
Pos logit contributions | |
pron | +0.182 |
favourite | +0.181 |
masterpiece | +0.168 |
favorite | +0.167 |
collection | +0.158 |
Neg logit contributions | |
crow | -0.193 |
Leave | -0.190 |
Bad | -0.181 |
ting | -0.177 |
Stop | -0.176 |
Token | It |
Token ID | 632 |
Feature activation | +7.37e-01 |
Pos logit contributions | |
Leave | +0.003 |
crow | +0.003 |
Stop | +0.003 |
Bad | +0.003 |
vious | +0.003 |
Neg logit contributions | |
favourite | -0.002 |
pron | -0.002 |
favorite | -0.002 |
Grand | -0.002 |
Daddy | -0.002 |
Token | is |
Token ID | 318 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
crow | +0.059 |
Leave | +0.059 |
Bad | +0.056 |
Stop | +0.055 |
ases | +0.054 |
Neg logit contributions | |
pron | -0.053 |
favourite | -0.050 |
masterpiece | -0.048 |
favorite | -0.046 |
collection | -0.045 |
Token | for |
Token ID | 329 |
Feature activation | -5.42e+00 |
Pos logit contributions | |
crow | +0.091 |
Bad | +0.090 |
Leave | +0.088 |
Stop | +0.084 |
vious | +0.084 |
Neg logit contributions | |
pron | -0.088 |
masterpiece | -0.079 |
favourite | -0.078 |
collection | -0.072 |
ail | -0.071 |
Token | war |
Token ID | 1175 |
Feature activation | +6.78e+00 |
Pos logit contributions | |
pron | +0.449 |
favourite | +0.416 |
masterpiece | +0.413 |
usual | +0.380 |
collection | +0.378 |
Neg logit contributions | |
crow | -0.400 |
Leave | -0.381 |
Bad | -0.375 |
vious | -0.365 |
Stop | -0.362 |
Token | . |
Token ID | 13 |
Feature activation | +2.88e-02 |
Pos logit contributions | |
crow | +0.564 |
Leave | +0.558 |
Bad | +0.546 |
Stop | +0.522 |
vious | +0.517 |
Neg logit contributions | |
pron | -0.483 |
favourite | -0.459 |
masterpiece | -0.437 |
favorite | -0.417 |
collection | -0.413 |
Token | War |
Token ID | 1810 |
Feature activation | +1.54e+01 |
Pos logit contributions | |
crow | +0.002 |
Leave | +0.002 |
vious | +0.002 |
eper | +0.002 |
Stop | +0.002 |
Neg logit contributions | |
favourite | -0.002 |
pron | -0.002 |
Grand | -0.002 |
favorite | -0.002 |
Daddy | -0.002 |
Token | is |
Token ID | 318 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
Leave | +1.238 |
crow | +1.237 |
Bad | +1.187 |
vious | +1.130 |
eper | +1.120 |
Neg logit contributions | |
pron | -1.030 |
favourite | -1.003 |
favorite | -0.966 |
masterpiece | -0.914 |
usual | -0.899 |
Token | bad |
Token ID | 2089 |
Feature activation | +2.06e+00 |
Pos logit contributions | |
crow | +0.103 |
Leave | +0.098 |
vious | +0.094 |
Stop | +0.094 |
Bad | +0.093 |
Neg logit contributions | |
pron | -0.074 |
favourite | -0.070 |
masterpiece | -0.068 |
favorite | -0.064 |
usual | -0.063 |
Token | ," |
Token ID | 553 |
Feature activation | -4.70e+00 |
Pos logit contributions | |
Leave | +0.178 |
crow | +0.178 |
Bad | +0.176 |
Stop | +0.164 |
vious | +0.163 |
Neg logit contributions | |
pron | -0.124 |
favourite | -0.122 |
favorite | -0.114 |
masterpiece | -0.113 |
collection | -0.108 |
Token | Lily |
Token ID | 20037 |
Feature activation | -1.93e+00 |
Pos logit contributions | |
pron | +0.353 |
favourite | +0.319 |
masterpiece | +0.318 |
collection | +0.287 |
usual | +0.283 |
Neg logit contributions | |
crow | -0.391 |
Leave | -0.377 |
Bad | -0.374 |
Stop | -0.361 |
vious | -0.359 |
Token | said |
Token ID | 531 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
pron | +0.098 |
favourite | +0.085 |
masterpiece | +0.079 |
favorite | +0.073 |
ail | +0.070 |
Neg logit contributions | |
crow | -0.199 |
Leave | -0.196 |
Bad | -0.194 |
ases | -0.188 |
Stop | -0.187 |
Token | birds |
Token ID | 10087 |
Feature activation | +9.41e+00 |
Pos logit contributions | |
crow | +0.179 |
Bad | +0.171 |
Leave | +0.171 |
Stop | +0.165 |
vious | +0.162 |
Neg logit contributions | |
pron | -0.158 |
favourite | -0.143 |
masterpiece | -0.142 |
collection | -0.128 |
brand | -0.126 |
Token | ," |
Token ID | 553 |
Feature activation | -3.61e+00 |
Pos logit contributions | |
crow | +0.734 |
Leave | +0.705 |
Bad | +0.701 |
Stop | +0.674 |
vious | +0.668 |
Neg logit contributions | |
pron | -0.747 |
favourite | -0.681 |
masterpiece | -0.678 |
collection | -0.617 |
usual | -0.605 |
Token | said |
Token ID | 531 |
Feature activation | -3.02e-01 |
Pos logit contributions | |
pron | +0.254 |
favourite | +0.229 |
masterpiece | +0.228 |
collection | +0.205 |
usual | +0.200 |
Neg logit contributions | |
crow | -0.315 |
Leave | -0.304 |
Bad | -0.303 |
Stop | -0.292 |
vious | -0.290 |
Token | the |
Token ID | 262 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
pron | +0.023 |
masterpiece | +0.021 |
favourite | +0.021 |
collection | +0.019 |
brand | +0.019 |
Neg logit contributions | |
crow | -0.025 |
Bad | -0.023 |
Leave | -0.023 |
Stop | -0.022 |
eper | -0.022 |
Token | scare |
Token ID | 19437 |
Feature activation | +4.88e+00 |
Pos logit contributions | |
crow | +0.148 |
Leave | +0.142 |
Bad | +0.141 |
Stop | +0.136 |
vious | +0.134 |
Neg logit contributions | |
pron | -0.178 |
favourite | -0.164 |
masterpiece | -0.163 |
collection | -0.150 |
usual | -0.147 |
Token | crow |
Token ID | 47114 |
Feature activation | +1.51e+01 |
Pos logit contributions | |
Leave | +0.099 |
Bad | +0.096 |
Stop | +0.075 |
ases | +0.069 |
vious | +0.067 |
Neg logit contributions | |
pron | -0.638 |
favourite | -0.606 |
masterpiece | -0.597 |
ail | -0.592 |
usual | -0.579 |
Token | . |
Token ID | 13 |
Feature activation | +5.33e-01 |
Pos logit contributions | |
crow | +1.032 |
Bad | +0.936 |
Leave | +0.934 |
vious | +0.906 |
eper | +0.904 |
Neg logit contributions | |
pron | -1.347 |
masterpiece | -1.239 |
favourite | -1.204 |
collection | -1.114 |
brand | -1.110 |
Token |
|
Token ID | 198 |
Feature activation | -3.26e-02 |
Pos logit contributions | |
crow | +0.042 |
Leave | +0.041 |
Bad | +0.041 |
Stop | +0.039 |
vious | +0.039 |
Neg logit contributions | |
pron | -0.041 |
favourite | -0.038 |
masterpiece | -0.037 |
collection | -0.034 |
brand | -0.033 |
Token |
|
Token ID | 198 |
Feature activation | -7.88e-02 |
Pos logit contributions | |
pron | +0.002 |
favourite | +0.002 |
masterpiece | +0.002 |
ail | +0.002 |
usual | +0.002 |
Neg logit contributions | |
crow | -0.003 |
Leave | -0.003 |
Bad | -0.003 |
vious | -0.003 |
Stop | -0.002 |
Token | " |
Token ID | 1 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
pron | +0.006 |
masterpiece | +0.005 |
favourite | +0.005 |
ail | +0.005 |
addy | +0.005 |
Neg logit contributions | |
vious | -0.006 |
crow | -0.006 |
ases | -0.005 |
ting | -0.005 |
hunters | -0.005 |
Token | Oh |
Token ID | 5812 |
Feature activation | +2.32e+00 |
Pos logit contributions | |
crow | +0.352 |
ases | +0.338 |
vious | +0.333 |
hunters | +0.318 |
eper | +0.315 |
Neg logit contributions | |
pron | -0.535 |
favourite | -0.484 |
masterpiece | -0.453 |
usual | -0.431 |
favorite | -0.425 |
Token | people |
Token ID | 661 |
Feature activation | +9.71e+00 |
Pos logit contributions | |
Bad | +0.324 |
Leave | +0.319 |
crow | +0.314 |
eper | +0.291 |
Hey | +0.288 |
Neg logit contributions | |
favourite | -0.463 |
masterpiece | -0.448 |
pron | -0.442 |
favorite | -0.435 |
usual | -0.431 |
Token | and |
Token ID | 290 |
Feature activation | +6.92e-02 |
Pos logit contributions | |
Leave | +0.789 |
crow | +0.785 |
Bad | +0.761 |
Stop | +0.760 |
Hey | +0.744 |
Neg logit contributions | |
pron | -0.728 |
favourite | -0.675 |
masterpiece | -0.631 |
collection | -0.617 |
favorite | -0.613 |
Token | their |
Token ID | 511 |
Feature activation | -7.23e-01 |
Pos logit contributions | |
crow | +0.005 |
Leave | +0.004 |
Bad | +0.004 |
vious | +0.004 |
Stop | +0.004 |
Neg logit contributions | |
pron | -0.006 |
favourite | -0.006 |
masterpiece | -0.006 |
collection | -0.005 |
usual | -0.005 |
Token | feelings |
Token ID | 7666 |
Feature activation | +3.41e+00 |
Pos logit contributions | |
pron | +0.060 |
favourite | +0.058 |
masterpiece | +0.057 |
collection | +0.053 |
favorite | +0.052 |
Neg logit contributions | |
crow | -0.052 |
Leave | -0.051 |
Bad | -0.050 |
Stop | -0.048 |
vious | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | +2.14e-01 |
Pos logit contributions | |
crow | +0.279 |
Leave | +0.271 |
Bad | +0.269 |
Stop | +0.260 |
vious | +0.255 |
Neg logit contributions | |
pron | -0.254 |
favourite | -0.234 |
masterpiece | -0.230 |
collection | -0.213 |
usual | -0.208 |
Token | Learn |
Token ID | 14365 |
Feature activation | +1.51e+01 |
Pos logit contributions | |
crow | +0.016 |
Leave | +0.016 |
vious | +0.015 |
Stop | +0.015 |
Bad | +0.015 |
Neg logit contributions | |
pron | -0.015 |
favourite | -0.015 |
masterpiece | -0.014 |
onna | -0.013 |
favorite | -0.013 |
Token | to |
Token ID | 284 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
crow | +1.118 |
Bad | +1.061 |
Leave | +1.049 |
Stop | +1.025 |
vious | +1.000 |
Neg logit contributions | |
pron | -1.283 |
masterpiece | -1.159 |
favourite | -1.152 |
collection | -1.057 |
brand | -1.037 |
Token | appreciate |
Token ID | 9144 |
Feature activation | +9.93e-02 |
Pos logit contributions | |
pron | +0.123 |
favourite | +0.113 |
masterpiece | +0.113 |
collection | +0.104 |
usual | +0.103 |
Neg logit contributions | |
crow | -0.107 |
Leave | -0.102 |
Bad | -0.101 |
vious | -0.097 |
Stop | -0.097 |
Token | and |
Token ID | 290 |
Feature activation | +9.73e-02 |
Pos logit contributions | |
crow | +0.007 |
Leave | +0.007 |
Bad | +0.007 |
vious | +0.006 |
Stop | +0.006 |
Neg logit contributions | |
pron | -0.009 |
favourite | -0.008 |
masterpiece | -0.008 |
collection | -0.007 |
usual | -0.007 |
Token | enjoy |
Token ID | 2883 |
Feature activation | -2.58e+00 |
Pos logit contributions | |
crow | +0.007 |
Leave | +0.007 |
Bad | +0.007 |
vious | +0.006 |
Stop | +0.006 |
Neg logit contributions | |
pron | -0.008 |
favourite | -0.008 |
masterpiece | -0.008 |
collection | -0.007 |
usual | -0.007 |
Token | different |
Token ID | 1180 |
Feature activation | +6.05e-01 |
Pos logit contributions | |
pron | +0.223 |
favourite | +0.210 |
masterpiece | +0.207 |
collection | +0.191 |
usual | +0.191 |
Neg logit contributions | |
crow | -0.180 |
Leave | -0.173 |
Bad | -0.168 |
vious | -0.167 |
Stop | -0.162 |
Token | ? |
Token ID | 30 |
Feature activation | -3.44e+00 |
Pos logit contributions | |
Leave | +0.226 |
Bad | +0.219 |
crow | +0.209 |
Stop | +0.202 |
vious | +0.194 |
Neg logit contributions | |
pron | -0.157 |
favourite | -0.147 |
collection | -0.141 |
masterpiece | -0.139 |
usual | -0.132 |
Token | Share |
Token ID | 8734 |
Feature activation | +8.35e+00 |
Pos logit contributions | |
pron | +0.301 |
favourite | +0.295 |
masterpiece | +0.276 |
favorite | +0.265 |
collection | +0.264 |
Neg logit contributions | |
crow | -0.224 |
Leave | -0.216 |
vious | -0.211 |
Bad | -0.207 |
Stop | -0.202 |
Token | ? |
Token ID | 30 |
Feature activation | -3.44e+00 |
Pos logit contributions | |
crow | +0.660 |
Leave | +0.656 |
Bad | +0.625 |
vious | +0.620 |
Stop | +0.603 |
Neg logit contributions | |
pron | -0.645 |
favourite | -0.603 |
masterpiece | -0.583 |
collection | -0.546 |
favorite | -0.541 |
Token | Play |
Token ID | 3811 |
Feature activation | +1.13e+01 |
Pos logit contributions | |
favourite | +0.280 |
pron | +0.279 |
masterpiece | +0.249 |
favorite | +0.247 |
collection | +0.238 |
Neg logit contributions | |
crow | -0.231 |
Leave | -0.223 |
Bad | -0.220 |
vious | -0.219 |
Stop | -0.212 |
Token | ? |
Token ID | 30 |
Feature activation | -3.44e+00 |
Pos logit contributions | |
Leave | +0.900 |
crow | +0.891 |
Bad | +0.864 |
vious | +0.860 |
Stop | +0.844 |
Neg logit contributions | |
pron | -0.880 |
favourite | -0.808 |
masterpiece | -0.774 |
ail | -0.709 |
favorite | -0.709 |
Token | Learn |
Token ID | 14365 |
Feature activation | +1.50e+01 |
Pos logit contributions | |
pron | +0.284 |
favourite | +0.282 |
masterpiece | +0.252 |
favorite | +0.249 |
brand | +0.241 |
Neg logit contributions | |
crow | -0.230 |
Leave | -0.218 |
Bad | -0.217 |
vious | -0.216 |
Stop | -0.210 |
Token | ?" |
Token ID | 1701 |
Feature activation | -6.20e+00 |
Pos logit contributions | |
crow | +1.119 |
Leave | +1.099 |
Bad | +1.076 |
vious | +1.033 |
Stop | +1.030 |
Neg logit contributions | |
pron | -1.230 |
favourite | -1.152 |
masterpiece | -1.134 |
collection | -1.045 |
favorite | -1.031 |
Token | the |
Token ID | 262 |
Feature activation | +1.46e+00 |
Pos logit contributions | |
pron | +0.366 |
favourite | +0.365 |
masterpiece | +0.357 |
usual | +0.330 |
daddy | +0.319 |
Neg logit contributions | |
crow | -0.531 |
Leave | -0.523 |
Bad | -0.520 |
vious | -0.516 |
ases | -0.503 |
Token | wolf |
Token ID | 17481 |
Feature activation | +3.02e+00 |
Pos logit contributions | |
crow | +0.089 |
Leave | +0.085 |
Bad | +0.084 |
Stop | +0.080 |
vious | +0.079 |
Neg logit contributions | |
pron | -0.142 |
favourite | -0.132 |
masterpiece | -0.130 |
collection | -0.121 |
brand | -0.120 |
Token | says |
Token ID | 1139 |
Feature activation | +2.23e+00 |
Pos logit contributions | |
Bad | +0.253 |
Leave | +0.249 |
crow | +0.242 |
Stop | +0.242 |
ases | +0.226 |
Neg logit contributions | |
pron | -0.194 |
favourite | -0.185 |
masterpiece | -0.184 |
collection | -0.170 |
favorite | -0.169 |
Token | . |
Token ID | 13 |
Feature activation | +2.26e-01 |
Pos logit contributions | |
crow | +0.192 |
Bad | +0.188 |
Leave | +0.187 |
Stop | +0.185 |
vious | +0.177 |
Neg logit contributions | |
pron | -0.159 |
favourite | -0.145 |
masterpiece | -0.142 |
collection | -0.128 |
ail | -0.127 |
Token | Join |
Token ID | 18234 |
Feature activation | +6.38e+00 |
Pos logit contributions | |
crow | +0.002 |
Bad | +0.002 |
eper | +0.002 |
vious | +0.002 |
hunters | +0.002 |
Neg logit contributions | |
pron | -0.003 |
favourite | -0.002 |
masterpiece | -0.002 |
usual | -0.002 |
favorite | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | +1.46e+00 |
Pos logit contributions | |
Leave | +0.482 |
crow | +0.471 |
Bad | +0.457 |
vious | +0.449 |
eper | +0.428 |
Neg logit contributions | |
favourite | -0.496 |
pron | -0.491 |
masterpiece | -0.468 |
favorite | -0.445 |
collection | -0.434 |
Token | Little |
Token ID | 7703 |
Feature activation | -2.68e+00 |
Pos logit contributions | |
crow | +0.122 |
Leave | +0.120 |
Bad | +0.118 |
vious | +0.114 |
Stop | +0.113 |
Neg logit contributions | |
pron | -0.106 |
favourite | -0.098 |
masterpiece | -0.096 |
collection | -0.088 |
usual | -0.086 |
Token | League |
Token ID | 4041 |
Feature activation | +2.27e+00 |
Pos logit contributions | |
pron | +0.218 |
favourite | +0.204 |
masterpiece | +0.199 |
collection | +0.185 |
usual | +0.183 |
Neg logit contributions | |
crow | -0.193 |
Leave | -0.191 |
Bad | -0.187 |
vious | -0.184 |
ases | -0.180 |
Token | ! |
Token ID | 0 |
Feature activation | -3.01e+00 |
Pos logit contributions | |
Leave | +0.175 |
crow | +0.174 |
Bad | +0.172 |
Stop | +0.164 |
vious | +0.163 |
Neg logit contributions | |
pron | -0.176 |
favourite | -0.163 |
masterpiece | -0.158 |
collection | -0.150 |
usual | -0.144 |
Token | Learn |
Token ID | 14365 |
Feature activation | +1.50e+01 |
Pos logit contributions | |
pron | +0.225 |
favourite | +0.225 |
Sus | +0.208 |
masterpiece | +0.203 |
Daddy | +0.203 |
Neg logit contributions | |
Bad | -0.197 |
Leave | -0.195 |
crow | -0.192 |
vious | -0.186 |
Stop | -0.180 |
Token | to |
Token ID | 284 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
crow | +1.154 |
Leave | +1.105 |
Bad | +1.101 |
Stop | +1.060 |
vious | +1.046 |
Neg logit contributions | |
pron | -1.217 |
favourite | -1.105 |
masterpiece | -1.103 |
collection | -1.005 |
brand | -0.986 |
Token | play |
Token ID | 711 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
pron | +0.120 |
favourite | +0.110 |
masterpiece | +0.110 |
collection | +0.101 |
usual | +0.099 |
Neg logit contributions | |
crow | -0.117 |
Leave | -0.113 |
Bad | -0.113 |
vious | -0.108 |
Stop | -0.108 |
Token | baseball |
Token ID | 9283 |
Feature activation | -3.27e-02 |
Pos logit contributions | |
Leave | +0.115 |
crow | +0.114 |
vious | +0.110 |
Bad | +0.110 |
Stop | +0.106 |
Neg logit contributions | |
pron | -0.096 |
favourite | -0.093 |
masterpiece | -0.087 |
collection | -0.083 |
favorite | -0.083 |
Token | with |
Token ID | 351 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
pron | +0.002 |
favourite | +0.002 |
masterpiece | +0.002 |
collection | +0.002 |
favorite | +0.002 |
Neg logit contributions | |
Leave | -0.003 |
crow | -0.003 |
Bad | -0.003 |
Stop | -0.003 |
vious | -0.002 |
Token | other |
Token ID | 584 |
Feature activation | +5.41e+00 |
Pos logit contributions | |
crow | +0.104 |
Leave | +0.102 |
Bad | +0.098 |
vious | +0.096 |
Stop | +0.094 |
Neg logit contributions | |
pron | -0.127 |
favourite | -0.120 |
masterpiece | -0.117 |
collection | -0.109 |
usual | -0.108 |
Token | Join |
Token ID | 18234 |
Feature activation | +6.36e+00 |
Pos logit contributions | |
crow | +0.000 |
hunters | +0.000 |
Bad | +0.000 |
eper | +0.000 |
Leave | +0.000 |
Neg logit contributions | |
pron | -0.000 |
favourite | -0.000 |
masterpiece | -0.000 |
usual | -0.000 |
oph | -0.000 |
Token | the |
Token ID | 262 |
Feature activation | +1.44e+00 |
Pos logit contributions | |
Leave | +0.486 |
crow | +0.478 |
Bad | +0.461 |
vious | +0.451 |
eper | +0.429 |
Neg logit contributions | |
pron | -0.491 |
favourite | -0.487 |
masterpiece | -0.468 |
favorite | -0.436 |
collection | -0.431 |
Token | Jazz |
Token ID | 21406 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
crow | +0.131 |
Leave | +0.128 |
Bad | +0.125 |
Stop | +0.121 |
vious | +0.121 |
Neg logit contributions | |
pron | -0.098 |
favourite | -0.088 |
masterpiece | -0.088 |
collection | -0.079 |
usual | -0.077 |
Token | Club |
Token ID | 6289 |
Feature activation | +4.68e+00 |
Pos logit contributions | |
crow | +0.146 |
Leave | +0.140 |
Bad | +0.140 |
Stop | +0.134 |
vious | +0.133 |
Neg logit contributions | |
pron | -0.161 |
favourite | -0.147 |
masterpiece | -0.147 |
collection | -0.134 |
usual | -0.132 |
Token | ! |
Token ID | 0 |
Feature activation | -3.00e+00 |
Pos logit contributions | |
crow | +0.353 |
Leave | +0.340 |
Bad | +0.338 |
Stop | +0.325 |
vious | +0.321 |
Neg logit contributions | |
pron | -0.383 |
favourite | -0.351 |
masterpiece | -0.349 |
collection | -0.320 |
usual | -0.313 |
Token | Learn |
Token ID | 14365 |
Feature activation | +1.50e+01 |
Pos logit contributions | |
favourite | +0.228 |
pron | +0.225 |
masterpiece | +0.205 |
Sus | +0.204 |
favorite | +0.203 |
Neg logit contributions | |
Bad | -0.202 |
Leave | -0.193 |
crow | -0.190 |
vious | -0.178 |
Stop | -0.177 |
Token | to |
Token ID | 284 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
crow | +1.111 |
Leave | +1.061 |
Bad | +1.056 |
Stop | +1.017 |
vious | +1.003 |
Neg logit contributions | |
pron | -1.262 |
favourite | -1.146 |
masterpiece | -1.146 |
collection | -1.045 |
usual | -1.028 |
Token | play |
Token ID | 711 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
pron | +0.125 |
favourite | +0.114 |
masterpiece | +0.114 |
collection | +0.105 |
usual | +0.103 |
Neg logit contributions | |
crow | -0.116 |
Leave | -0.112 |
Bad | -0.111 |
vious | -0.106 |
Stop | -0.106 |
Token | instruments |
Token ID | 12834 |
Feature activation | +3.58e+00 |
Pos logit contributions | |
crow | +0.112 |
Leave | +0.110 |
Bad | +0.106 |
vious | +0.105 |
Stop | +0.101 |
Neg logit contributions | |
pron | -0.100 |
favourite | -0.095 |
masterpiece | -0.091 |
collection | -0.085 |
brand | -0.084 |
Token | and |
Token ID | 290 |
Feature activation | +1.09e-02 |
Pos logit contributions | |
crow | +0.275 |
Leave | +0.266 |
Bad | +0.263 |
Stop | +0.252 |
vious | +0.248 |
Neg logit contributions | |
pron | -0.291 |
favourite | -0.268 |
masterpiece | -0.264 |
collection | -0.244 |
brand | -0.238 |
Token | sing |
Token ID | 1702 |
Feature activation | +1.75e-01 |
Pos logit contributions | |
crow | +0.001 |
Bad | +0.001 |
Leave | +0.001 |
vious | +0.001 |
Stop | +0.001 |
Neg logit contributions | |
pron | -0.001 |
favourite | -0.001 |
masterpiece | -0.001 |
collection | -0.001 |
usual | -0.001 |
Token | Have |
Token ID | 11980 |
Feature activation | +4.34e+00 |
Pos logit contributions | |
pron | +0.012 |
favourite | +0.011 |
masterpiece | +0.010 |
usual | +0.010 |
favorite | +0.009 |
Neg logit contributions | |
crow | -0.009 |
eper | -0.008 |
ases | -0.008 |
vious | -0.008 |
hunters | -0.007 |
Token | fun |
Token ID | 1257 |
Feature activation | -7.86e-01 |
Pos logit contributions | |
Bad | +0.322 |
crow | +0.321 |
Leave | +0.321 |
Stop | +0.319 |
Hey | +0.306 |
Neg logit contributions | |
pron | -0.361 |
masterpiece | -0.327 |
favourite | -0.302 |
collection | -0.295 |
ail | -0.290 |
Token | , |
Token ID | 11 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
pron | +0.065 |
favourite | +0.060 |
masterpiece | +0.059 |
collection | +0.055 |
brand | +0.053 |
Neg logit contributions | |
crow | -0.059 |
Leave | -0.057 |
Bad | -0.056 |
Stop | -0.054 |
vious | -0.054 |
Token | Lily |
Token ID | 20037 |
Feature activation | -1.85e+00 |
Pos logit contributions | |
crow | +0.098 |
Bad | +0.095 |
Leave | +0.095 |
Stop | +0.091 |
vious | +0.090 |
Neg logit contributions | |
pron | -0.087 |
masterpiece | -0.078 |
favourite | -0.078 |
collection | -0.070 |
usual | -0.069 |
Token | . |
Token ID | 13 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
pron | +0.146 |
favourite | +0.132 |
masterpiece | +0.131 |
collection | +0.121 |
usual | +0.117 |
Neg logit contributions | |
crow | -0.147 |
Leave | -0.143 |
Bad | -0.140 |
Stop | -0.136 |
vious | -0.134 |
Token | Learn |
Token ID | 14365 |
Feature activation | +1.50e+01 |
Pos logit contributions | |
crow | +0.006 |
Leave | +0.006 |
Bad | +0.006 |
vious | +0.006 |
eper | +0.006 |
Neg logit contributions | |
pron | -0.009 |
favourite | -0.009 |
masterpiece | -0.008 |
bedroom | -0.008 |
favorite | -0.008 |
Token | a |
Token ID | 257 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
crow | +1.110 |
Bad | +1.052 |
Leave | +1.028 |
Stop | +1.013 |
ases | +1.003 |
Neg logit contributions | |
pron | -1.270 |
masterpiece | -1.138 |
favourite | -1.114 |
collection | -1.022 |
brand | -1.013 |
Token | lot |
Token ID | 1256 |
Feature activation | -3.82e+00 |
Pos logit contributions | |
crow | +0.015 |
Bad | +0.014 |
Stop | +0.014 |
Leave | +0.014 |
hunters | +0.013 |
Neg logit contributions | |
pron | -0.013 |
masterpiece | -0.011 |
favourite | -0.011 |
ail | -0.010 |
usual | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
pron | +0.322 |
favourite | +0.296 |
masterpiece | +0.292 |
collection | +0.270 |
usual | +0.269 |
Neg logit contributions | |
crow | -0.274 |
Leave | -0.273 |
Bad | -0.268 |
Stop | -0.255 |
vious | -0.250 |
Token | We |
Token ID | 775 |
Feature activation | -6.33e-01 |
Pos logit contributions | |
crow | +0.008 |
Bad | +0.007 |
Leave | +0.007 |
vious | +0.007 |
eper | +0.007 |
Neg logit contributions | |
pron | -0.010 |
favourite | -0.010 |
masterpiece | -0.009 |
bedroom | -0.009 |
favorite | -0.008 |
Token | are |
Token ID | 389 |
Feature activation | +3.99e+00 |
Pos logit contributions | |
pron | +0.050 |
favourite | +0.046 |
masterpiece | +0.045 |
collection | +0.043 |
favorite | +0.041 |
Neg logit contributions | |
crow | -0.049 |
Leave | -0.049 |
Bad | -0.048 |
Stop | -0.045 |
vious | -0.045 |
Token | any |
Token ID | 597 |
Feature activation | +4.37e-02 |
Pos logit contributions | |
crow | +0.396 |
Leave | +0.378 |
Bad | +0.375 |
Stop | +0.360 |
vious | +0.357 |
Neg logit contributions | |
pron | -0.449 |
favourite | -0.413 |
masterpiece | -0.411 |
collection | -0.377 |
usual | -0.372 |
Token | situation |
Token ID | 3074 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.003 |
Bad | +0.003 |
Stop | +0.003 |
vious | +0.003 |
Neg logit contributions | |
pron | -0.004 |
favourite | -0.003 |
masterpiece | -0.003 |
collection | -0.003 |
usual | -0.003 |
Token | too |
Token ID | 1165 |
Feature activation | +5.21e+00 |
Pos logit contributions | |
crow | +0.153 |
Leave | +0.150 |
Bad | +0.148 |
Stop | +0.142 |
vious | +0.140 |
Neg logit contributions | |
pron | -0.147 |
favourite | -0.137 |
masterpiece | -0.135 |
collection | -0.124 |
usual | -0.123 |
Token | seriously |
Token ID | 6411 |
Feature activation | +2.37e+00 |
Pos logit contributions | |
crow | +0.442 |
Leave | +0.427 |
Bad | +0.424 |
Stop | +0.409 |
vious | +0.405 |
Neg logit contributions | |
pron | -0.379 |
favourite | -0.342 |
masterpiece | -0.341 |
collection | -0.308 |
usual | -0.302 |
Token | . |
Token ID | 13 |
Feature activation | +3.35e-02 |
Pos logit contributions | |
crow | +0.182 |
Leave | +0.175 |
Bad | +0.174 |
Stop | +0.167 |
vious | +0.164 |
Neg logit contributions | |
pron | -0.190 |
favourite | -0.173 |
masterpiece | -0.173 |
collection | -0.159 |
brand | -0.156 |
Token | Learn |
Token ID | 14365 |
Feature activation | +1.49e+01 |
Pos logit contributions | |
crow | +0.002 |
Leave | +0.002 |
Bad | +0.002 |
vious | +0.002 |
Stop | +0.002 |
Neg logit contributions | |
pron | -0.003 |
favourite | -0.003 |
masterpiece | -0.002 |
collection | -0.002 |
favorite | -0.002 |
Token | to |
Token ID | 284 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
crow | +1.253 |
Leave | +1.207 |
Bad | +1.202 |
Stop | +1.160 |
vious | +1.148 |
Neg logit contributions | |
pron | -1.095 |
favourite | -0.988 |
masterpiece | -0.985 |
collection | -0.888 |
usual | -0.869 |
Token | laugh |
Token ID | 6487 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
pron | +0.136 |
favourite | +0.124 |
masterpiece | +0.124 |
collection | +0.113 |
usual | +0.111 |
Neg logit contributions | |
crow | -0.125 |
Leave | -0.120 |
Bad | -0.119 |
Stop | -0.115 |
vious | -0.113 |
Token | at |
Token ID | 379 |
Feature activation | +1.60e-01 |
Pos logit contributions | |
crow | +0.118 |
Leave | +0.116 |
Bad | +0.114 |
Stop | +0.108 |
vious | +0.106 |
Neg logit contributions | |
pron | -0.116 |
favourite | -0.109 |
masterpiece | -0.105 |
collection | -0.096 |
favorite | -0.096 |
Token | yourself |
Token ID | 3511 |
Feature activation | +5.78e+00 |
Pos logit contributions | |
crow | +0.011 |
Leave | +0.010 |
Bad | +0.010 |
vious | +0.010 |
Stop | +0.010 |
Neg logit contributions | |
pron | -0.014 |
favourite | -0.013 |
masterpiece | -0.013 |
usual | -0.012 |
collection | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | -8.33e-02 |
Pos logit contributions | |
crow | +0.443 |
Leave | +0.439 |
Bad | +0.432 |
Stop | +0.414 |
vious | +0.403 |
Neg logit contributions | |
pron | -0.458 |
favourite | -0.425 |
masterpiece | -0.416 |
collection | -0.387 |
usual | -0.378 |
Token | ockey |
Token ID | 8337 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
pron | +0.023 |
favourite | +0.021 |
masterpiece | +0.021 |
collection | +0.020 |
ail | +0.018 |
Neg logit contributions | |
Leave | -0.026 |
crow | -0.026 |
Bad | -0.026 |
Stop | -0.025 |
vious | -0.024 |
Token | Camp |
Token ID | 5425 |
Feature activation | +4.59e+00 |
Pos logit contributions | |
pron | +0.008 |
favourite | +0.008 |
masterpiece | +0.007 |
collection | +0.007 |
usual | +0.007 |
Neg logit contributions | |
Leave | -0.008 |
crow | -0.008 |
Bad | -0.007 |
Stop | -0.007 |
vious | -0.007 |
Token | for |
Token ID | 329 |
Feature activation | -5.42e+00 |
Pos logit contributions | |
crow | +0.340 |
Leave | +0.336 |
Bad | +0.331 |
Stop | +0.314 |
vious | +0.312 |
Neg logit contributions | |
pron | -0.378 |
favourite | -0.350 |
masterpiece | -0.344 |
collection | -0.321 |
usual | -0.314 |
Token | Kids |
Token ID | 17476 |
Feature activation | +1.78e+00 |
Pos logit contributions | |
pron | +0.462 |
favourite | +0.430 |
masterpiece | +0.424 |
collection | +0.394 |
usual | +0.391 |
Neg logit contributions | |
crow | -0.392 |
Leave | -0.374 |
Bad | -0.366 |
vious | -0.356 |
Stop | -0.353 |
Token | . |
Token ID | 13 |
Feature activation | +3.40e-02 |
Pos logit contributions | |
Leave | +0.137 |
crow | +0.137 |
Bad | +0.136 |
Stop | +0.129 |
vious | +0.127 |
Neg logit contributions | |
pron | -0.138 |
favourite | -0.130 |
masterpiece | -0.126 |
collection | -0.119 |
usual | -0.118 |
Token | Learn |
Token ID | 14365 |
Feature activation | +1.49e+01 |
Pos logit contributions | |
Bad | +0.002 |
Leave | +0.002 |
crow | +0.002 |
vious | +0.002 |
Stop | +0.002 |
Neg logit contributions | |
favourite | -0.003 |
pron | -0.002 |
Sus | -0.002 |
masterpiece | -0.002 |
favorite | -0.002 |
Token | from |
Token ID | 422 |
Feature activation | +3.75e+00 |
Pos logit contributions | |
crow | +1.198 |
Leave | +1.161 |
Bad | +1.151 |
Stop | +1.107 |
vious | +1.098 |
Neg logit contributions | |
pron | -1.142 |
favourite | -1.048 |
masterpiece | -1.038 |
collection | -0.950 |
usual | -0.930 |
Token | the |
Token ID | 262 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
crow | +0.277 |
Leave | +0.267 |
Bad | +0.264 |
Stop | +0.254 |
vious | +0.252 |
Neg logit contributions | |
pron | -0.314 |
favourite | -0.289 |
masterpiece | -0.288 |
collection | -0.264 |
usual | -0.260 |
Token | Best |
Token ID | 6705 |
Feature activation | +2.89e+00 |
Pos logit contributions | |
crow | +0.111 |
Leave | +0.108 |
vious | +0.104 |
Bad | +0.104 |
ases | +0.102 |
Neg logit contributions | |
pron | -0.091 |
favourite | -0.085 |
masterpiece | -0.083 |
collection | -0.079 |
usual | -0.078 |
Token | . |
Token ID | 13 |
Feature activation | +3.16e-02 |
Pos logit contributions | |
Leave | +0.209 |
crow | +0.208 |
Bad | +0.204 |
Stop | +0.191 |
vious | +0.188 |
Neg logit contributions | |
pron | -0.240 |
favourite | -0.231 |
masterpiece | -0.223 |
collection | -0.215 |
usual | -0.208 |
Token | Sign |
Token ID | 5865 |
Feature activation | +6.60e+00 |
Pos logit contributions | |
Bad | +0.002 |
Leave | +0.002 |
crow | +0.002 |
vious | +0.002 |
Stop | +0.002 |
Neg logit contributions | |
favourite | -0.002 |
pron | -0.002 |
masterpiece | -0.002 |
Sus | -0.002 |
mummy | -0.002 |
Token | for |
Token ID | 329 |
Feature activation | -4.92e+00 |
Pos logit contributions | |
crow | +0.070 |
Leave | +0.066 |
Bad | +0.066 |
Stop | +0.064 |
vious | +0.063 |
Neg logit contributions | |
pron | -0.075 |
masterpiece | -0.069 |
favourite | -0.069 |
collection | -0.063 |
brand | -0.061 |
Token | help |
Token ID | 1037 |
Feature activation | +2.76e+00 |
Pos logit contributions | |
pron | +0.419 |
favourite | +0.389 |
masterpiece | +0.385 |
collection | +0.355 |
usual | +0.353 |
Neg logit contributions | |
crow | -0.352 |
Leave | -0.339 |
Bad | -0.336 |
Stop | -0.325 |
vious | -0.322 |
Token | . |
Token ID | 13 |
Feature activation | +4.37e-01 |
Pos logit contributions | |
crow | +0.219 |
Leave | +0.214 |
Bad | +0.213 |
Stop | +0.205 |
vious | +0.203 |
Neg logit contributions | |
pron | -0.214 |
favourite | -0.195 |
masterpiece | -0.193 |
collection | -0.177 |
usual | -0.175 |
Token | The |
Token ID | 383 |
Feature activation | +2.43e-01 |
Pos logit contributions | |
crow | +0.034 |
Leave | +0.033 |
Bad | +0.033 |
Stop | +0.032 |
vious | +0.032 |
Neg logit contributions | |
pron | -0.034 |
favourite | -0.031 |
masterpiece | -0.031 |
collection | -0.028 |
brand | -0.028 |
Token | scare |
Token ID | 19437 |
Feature activation | +4.53e+00 |
Pos logit contributions | |
Leave | +0.016 |
Bad | +0.015 |
ases | +0.015 |
crow | +0.015 |
vious | +0.015 |
Neg logit contributions | |
pron | -0.021 |
favourite | -0.020 |
masterpiece | -0.020 |
usual | -0.018 |
collection | -0.018 |
Token | crow |
Token ID | 47114 |
Feature activation | +1.49e+01 |
Pos logit contributions | |
Leave | +0.063 |
Bad | +0.060 |
Stop | +0.044 |
ases | +0.043 |
vious | +0.036 |
Neg logit contributions | |
pron | -0.607 |
favourite | -0.588 |
masterpiece | -0.577 |
ail | -0.568 |
usual | -0.561 |
Token | said |
Token ID | 531 |
Feature activation | -9.94e-01 |
Pos logit contributions | |
crow | +1.276 |
Bad | +1.208 |
Leave | +1.200 |
vious | +1.173 |
Stop | +1.155 |
Neg logit contributions | |
pron | -1.065 |
masterpiece | -0.949 |
favourite | -0.929 |
collection | -0.835 |
brand | -0.831 |
Token | " |
Token ID | 366 |
Feature activation | +4.48e-02 |
Pos logit contributions | |
pron | +0.072 |
masterpiece | +0.065 |
favourite | +0.064 |
collection | +0.058 |
usual | +0.057 |
Neg logit contributions | |
crow | -0.085 |
Leave | -0.082 |
Bad | -0.082 |
Stop | -0.079 |
vious | -0.078 |
Token | Come |
Token ID | 16773 |
Feature activation | +9.70e+00 |
Pos logit contributions | |
crow | +0.003 |
ases | +0.003 |
Leave | +0.003 |
Bad | +0.003 |
eper | +0.003 |
Neg logit contributions | |
pron | -0.004 |
favourite | -0.004 |
masterpiece | -0.004 |
usual | -0.003 |
collection | -0.003 |
Token | with |
Token ID | 351 |
Feature activation | +1.59e+00 |
Pos logit contributions | |
crow | +0.775 |
vious | +0.750 |
Bad | +0.749 |
Leave | +0.744 |
Stop | +0.729 |
Neg logit contributions | |
pron | -0.745 |
favourite | -0.686 |
masterpiece | -0.656 |
collection | -0.603 |
brand | -0.599 |
Token | me |
Token ID | 502 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
crow | +0.108 |
Leave | +0.107 |
Bad | +0.104 |
vious | +0.100 |
Stop | +0.099 |
Neg logit contributions | |
pron | -0.140 |
favourite | -0.128 |
masterpiece | -0.127 |
collection | -0.117 |
usual | -0.116 |
Token | . |
Token ID | 13 |
Feature activation | +5.33e-01 |
Pos logit contributions | |
crow | +0.130 |
Bad | +0.125 |
Leave | +0.125 |
Stop | +0.120 |
vious | +0.116 |
Neg logit contributions | |
pron | -0.118 |
masterpiece | -0.102 |
favourite | -0.102 |
collection | -0.094 |
brand | -0.089 |
Token | They |
Token ID | 1119 |
Feature activation | +2.86e-01 |
Pos logit contributions | |
crow | +0.035 |
Bad | +0.033 |
Leave | +0.033 |
Stop | +0.032 |
vious | +0.030 |
Neg logit contributions | |
pron | -0.051 |
masterpiece | -0.046 |
favourite | -0.045 |
collection | -0.042 |
usual | -0.042 |
Token | thanked |
Token ID | 26280 |
Feature activation | +3.90e-02 |
Pos logit contributions | |
crow | +0.021 |
Bad | +0.020 |
Leave | +0.020 |
vious | +0.020 |
Stop | +0.019 |
Neg logit contributions | |
pron | -0.025 |
masterpiece | -0.022 |
favourite | -0.022 |
brand | -0.020 |
usual | -0.020 |
Token | the |
Token ID | 262 |
Feature activation | +1.63e+00 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.003 |
Bad | +0.003 |
Stop | +0.003 |
vious | +0.002 |
Neg logit contributions | |
pron | -0.003 |
favourite | -0.003 |
masterpiece | -0.003 |
collection | -0.003 |
brand | -0.003 |
Token | scare |
Token ID | 19437 |
Feature activation | +4.53e+00 |
Pos logit contributions | |
crow | +0.126 |
Bad | +0.121 |
Leave | +0.121 |
Stop | +0.116 |
vious | +0.114 |
Neg logit contributions | |
pron | -0.131 |
favourite | -0.119 |
masterpiece | -0.119 |
collection | -0.109 |
brand | -0.106 |
Token | crow |
Token ID | 47114 |
Feature activation | +1.49e+01 |
Pos logit contributions | |
Leave | +0.315 |
Bad | +0.309 |
Stop | +0.289 |
crow | +0.286 |
vious | +0.275 |
Neg logit contributions | |
pron | -0.376 |
favourite | -0.350 |
masterpiece | -0.349 |
ail | -0.331 |
usual | -0.322 |
Token | and |
Token ID | 290 |
Feature activation | +1.67e-01 |
Pos logit contributions | |
crow | +1.010 |
Leave | +0.927 |
Bad | +0.926 |
vious | +0.896 |
eper | +0.885 |
Neg logit contributions | |
pron | -1.321 |
masterpiece | -1.202 |
favourite | -1.174 |
brand | -1.081 |
collection | -1.079 |
Token | hugged |
Token ID | 41130 |
Feature activation | -1.00e+00 |
Pos logit contributions | |
crow | +0.013 |
Bad | +0.012 |
Leave | +0.012 |
Stop | +0.011 |
eper | +0.011 |
Neg logit contributions | |
pron | -0.014 |
masterpiece | -0.013 |
favourite | -0.013 |
ail | -0.011 |
collection | -0.011 |
Token | him |
Token ID | 683 |
Feature activation | +7.70e+00 |
Pos logit contributions | |
pron | +0.086 |
favourite | +0.081 |
masterpiece | +0.080 |
usual | +0.073 |
collection | +0.073 |
Neg logit contributions | |
crow | -0.070 |
Leave | -0.068 |
Bad | -0.067 |
vious | -0.065 |
Stop | -0.064 |
Token | . |
Token ID | 13 |
Feature activation | +2.30e-01 |
Pos logit contributions | |
crow | +0.611 |
Bad | +0.585 |
Leave | +0.583 |
Stop | +0.555 |
vious | +0.554 |
Neg logit contributions | |
pron | -0.599 |
masterpiece | -0.544 |
favourite | -0.537 |
collection | -0.490 |
brand | -0.481 |
Token | They |
Token ID | 1119 |
Feature activation | +1.14e-02 |
Pos logit contributions | |
crow | +0.016 |
Leave | +0.016 |
Bad | +0.016 |
Stop | +0.015 |
vious | +0.014 |
Neg logit contributions | |
pron | -0.020 |
masterpiece | -0.019 |
favourite | -0.018 |
collection | -0.017 |
usual | -0.017 |
Token | . |
Token ID | 13 |
Feature activation | +2.58e-01 |
Pos logit contributions | |
crow | +0.014 |
Leave | +0.014 |
Bad | +0.013 |
vious | +0.013 |
Stop | +0.013 |
Neg logit contributions | |
pron | -0.017 |
masterpiece | -0.015 |
favourite | -0.014 |
collection | -0.013 |
brand | -0.013 |
Token | Suddenly |
Token ID | 24975 |
Feature activation | +2.42e-01 |
Pos logit contributions | |
crow | +0.016 |
Leave | +0.016 |
Bad | +0.016 |
Stop | +0.016 |
Hey | +0.015 |
Neg logit contributions | |
pron | -0.024 |
masterpiece | -0.022 |
collection | -0.021 |
favourite | -0.020 |
ail | -0.020 |
Token | , |
Token ID | 11 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
crow | +0.016 |
Bad | +0.015 |
Leave | +0.015 |
Stop | +0.015 |
Hey | +0.014 |
Neg logit contributions | |
pron | -0.022 |
masterpiece | -0.021 |
favourite | -0.020 |
collection | -0.018 |
brand | -0.018 |
Token | the |
Token ID | 262 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
Bad | +0.078 |
crow | +0.075 |
Stop | +0.074 |
Leave | +0.072 |
Hey | +0.071 |
Neg logit contributions | |
pron | -0.133 |
masterpiece | -0.129 |
collection | -0.121 |
favourite | -0.120 |
ail | -0.115 |
Token | scare |
Token ID | 19437 |
Feature activation | +4.45e+00 |
Pos logit contributions | |
crow | +0.117 |
Leave | +0.117 |
Bad | +0.114 |
vious | +0.112 |
ases | +0.112 |
Neg logit contributions | |
pron | -0.117 |
favourite | -0.107 |
masterpiece | -0.106 |
usual | -0.097 |
collection | -0.096 |
Token | crow |
Token ID | 47114 |
Feature activation | +1.48e+01 |
Pos logit contributions | |
Leave | +0.091 |
Bad | +0.086 |
ases | +0.083 |
vious | +0.069 |
Stop | +0.063 |
Neg logit contributions | |
pron | -0.566 |
favourite | -0.534 |
masterpiece | -0.526 |
ail | -0.522 |
usual | -0.515 |
Token | moved |
Token ID | 3888 |
Feature activation | +1.78e+00 |
Pos logit contributions | |
crow | +1.201 |
Leave | +1.162 |
Bad | +1.160 |
Stop | +1.123 |
vious | +1.115 |
Neg logit contributions | |
pron | -1.129 |
favourite | -0.996 |
masterpiece | -0.995 |
collection | -0.905 |
brand | -0.874 |
Token | and |
Token ID | 290 |
Feature activation | +1.61e-01 |
Pos logit contributions | |
Leave | +0.138 |
crow | +0.137 |
Bad | +0.136 |
ases | +0.131 |
vious | +0.130 |
Neg logit contributions | |
pron | -0.124 |
favourite | -0.123 |
masterpiece | -0.119 |
usual | -0.110 |
favorite | -0.109 |
Token | waved |
Token ID | 26834 |
Feature activation | +5.41e-01 |
Pos logit contributions | |
Bad | +0.011 |
crow | +0.011 |
Leave | +0.011 |
Stop | +0.010 |
Hey | +0.010 |
Neg logit contributions | |
pron | -0.014 |
masterpiece | -0.013 |
favourite | -0.013 |
collection | -0.012 |
ail | -0.012 |
Token | its |
Token ID | 663 |
Feature activation | +5.50e-01 |
Pos logit contributions | |
crow | +0.043 |
Leave | +0.041 |
Bad | +0.041 |
Stop | +0.040 |
vious | +0.040 |
Neg logit contributions | |
pron | -0.041 |
favourite | -0.039 |
masterpiece | -0.037 |
favorite | -0.034 |
collection | -0.034 |
Token | arms |
Token ID | 5101 |
Feature activation | -3.08e+00 |
Pos logit contributions | |
crow | +0.044 |
Leave | +0.042 |
Bad | +0.042 |
Stop | +0.040 |
vious | +0.040 |
Neg logit contributions | |
pron | -0.043 |
masterpiece | -0.039 |
favourite | -0.039 |
collection | -0.035 |
brand | -0.035 |
Token | stop |
Token ID | 2245 |
Feature activation | +3.97e+00 |
Pos logit contributions | |
Leave | +0.096 |
crow | +0.095 |
vious | +0.091 |
ases | +0.090 |
Bad | +0.088 |
Neg logit contributions | |
pron | -0.096 |
favourite | -0.090 |
masterpiece | -0.086 |
favorite | -0.083 |
collection | -0.080 |
Token | it |
Token ID | 340 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
crow | +0.338 |
Leave | +0.330 |
vious | +0.319 |
Bad | +0.315 |
Stop | +0.309 |
Neg logit contributions | |
pron | -0.272 |
masterpiece | -0.253 |
favourite | -0.244 |
collection | -0.223 |
brand | -0.219 |
Token | !" |
Token ID | 2474 |
Feature activation | -3.14e+00 |
Pos logit contributions | |
crow | +0.015 |
Leave | +0.015 |
Bad | +0.014 |
Stop | +0.014 |
vious | +0.013 |
Neg logit contributions | |
pron | -0.014 |
masterpiece | -0.012 |
favourite | -0.012 |
collection | -0.012 |
ail | -0.011 |
Token | the |
Token ID | 262 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
pron | +0.230 |
favourite | +0.213 |
masterpiece | +0.209 |
usual | +0.189 |
brand | +0.188 |
Neg logit contributions | |
crow | -0.256 |
Leave | -0.248 |
Bad | -0.245 |
Stop | -0.237 |
vious | -0.235 |
Token | scare |
Token ID | 19437 |
Feature activation | +4.42e+00 |
Pos logit contributions | |
Leave | +0.100 |
vious | +0.099 |
Bad | +0.097 |
crow | +0.094 |
ases | +0.092 |
Neg logit contributions | |
pron | -0.119 |
favourite | -0.112 |
masterpiece | -0.111 |
ail | -0.106 |
porch | -0.105 |
Token | crow |
Token ID | 47114 |
Feature activation | +1.48e+01 |
Pos logit contributions | |
Leave | +0.097 |
Bad | +0.096 |
Learn | +0.065 |
risks | +0.060 |
ases | +0.059 |
Neg logit contributions | |
ail | -0.509 |
pron | -0.484 |
yt | -0.471 |
Sus | -0.467 |
iana | -0.466 |
Token | said |
Token ID | 531 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
crow | +1.367 |
Leave | +1.358 |
Bad | +1.331 |
Stop | +1.292 |
vious | +1.269 |
Neg logit contributions | |
pron | -0.900 |
favourite | -0.815 |
masterpiece | -0.798 |
collection | -0.714 |
favorite | -0.707 |
Token | . |
Token ID | 13 |
Feature activation | +2.56e-01 |
Pos logit contributions | |
pron | +0.072 |
favourite | +0.066 |
masterpiece | +0.065 |
collection | +0.059 |
brand | +0.058 |
Neg logit contributions | |
crow | -0.088 |
Bad | -0.086 |
Leave | -0.086 |
Stop | -0.085 |
vious | -0.081 |
Token | " |
Token ID | 366 |
Feature activation | +3.00e-02 |
Pos logit contributions | |
crow | +0.024 |
Leave | +0.023 |
Bad | +0.023 |
vious | +0.023 |
Stop | +0.022 |
Neg logit contributions | |
pron | -0.015 |
favourite | -0.013 |
masterpiece | -0.013 |
brand | -0.012 |
favorite | -0.012 |
Token | You |
Token ID | 1639 |
Feature activation | +6.86e+00 |
Pos logit contributions | |
crow | +0.002 |
hunters | +0.002 |
eper | +0.002 |
thieves | +0.002 |
risks | +0.002 |
Neg logit contributions | |
pron | -0.002 |
favourite | -0.002 |
masterpiece | -0.002 |
hardly | -0.002 |
favorite | -0.002 |
Token | are |
Token ID | 389 |
Feature activation | +4.12e+00 |
Pos logit contributions | |
crow | +0.518 |
Leave | +0.504 |
Bad | +0.482 |
Stop | +0.463 |
ases | +0.442 |
Neg logit contributions | |
pron | -0.546 |
favourite | -0.514 |
masterpiece | -0.503 |
collection | -0.487 |
favorite | -0.469 |
Token | a |
Token ID | 257 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
crow | +0.532 |
Bad | +0.499 |
Leave | +0.494 |
vious | +0.487 |
Stop | +0.471 |
Neg logit contributions | |
pron | -0.653 |
masterpiece | -0.589 |
favourite | -0.561 |
collection | -0.528 |
brand | -0.519 |
Token | cup |
Token ID | 6508 |
Feature activation | -5.08e+00 |
Pos logit contributions | |
crow | +0.008 |
Bad | +0.008 |
Leave | +0.007 |
Stop | +0.007 |
vious | +0.007 |
Neg logit contributions | |
pron | -0.006 |
masterpiece | -0.006 |
favourite | -0.006 |
ail | -0.006 |
usual | -0.005 |
Token | of |
Token ID | 286 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
pron | +0.515 |
masterpiece | +0.497 |
favourite | +0.483 |
usual | +0.452 |
collection | +0.448 |
Neg logit contributions | |
crow | -0.282 |
vious | -0.236 |
Leave | -0.230 |
Bad | -0.217 |
ases | -0.213 |
Token | warm |
Token ID | 5814 |
Feature activation | -4.28e+00 |
Pos logit contributions | |
crow | +0.206 |
Bad | +0.194 |
Leave | +0.190 |
Stop | +0.188 |
vious | +0.186 |
Neg logit contributions | |
pron | -0.174 |
masterpiece | -0.156 |
favourite | -0.149 |
ail | -0.143 |
collection | -0.139 |
Token | , |
Token ID | 11 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
pron | +0.332 |
masterpiece | +0.301 |
favourite | +0.293 |
usual | +0.271 |
collection | +0.268 |
Neg logit contributions | |
crow | -0.347 |
Leave | -0.316 |
Bad | -0.307 |
Stop | -0.307 |
vious | -0.306 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.67e+01 |
Pos logit contributions | |
Leave | +0.086 |
crow | +0.085 |
Bad | +0.084 |
vious | +0.080 |
hunters | +0.079 |
Neg logit contributions | |
pron | -0.080 |
masterpiece | -0.076 |
ail | -0.073 |
favourite | -0.071 |
collection | -0.070 |
Token | made |
Token ID | 925 |
Feature activation | -9.23e-01 |
Pos logit contributions | |
pron | +1.240 |
masterpiece | +1.148 |
favourite | +1.129 |
collection | +1.038 |
brand | +1.018 |
Neg logit contributions | |
crow | -1.372 |
Leave | -1.317 |
Bad | -1.315 |
Stop | -1.271 |
vious | -1.232 |
Token | soup |
Token ID | 17141 |
Feature activation | -4.06e+00 |
Pos logit contributions | |
pron | +0.096 |
favourite | +0.089 |
masterpiece | +0.089 |
collection | +0.083 |
usual | +0.082 |
Neg logit contributions | |
crow | -0.050 |
Leave | -0.047 |
Bad | -0.047 |
Stop | -0.044 |
vious | -0.043 |
Token | and |
Token ID | 290 |
Feature activation | -8.93e-02 |
Pos logit contributions | |
pron | +0.351 |
masterpiece | +0.323 |
favourite | +0.311 |
usual | +0.295 |
complexion | +0.281 |
Neg logit contributions | |
crow | -0.256 |
hunters | -0.231 |
ases | -0.229 |
Leave | -0.226 |
eper | -0.226 |
Token | a |
Token ID | 257 |
Feature activation | +9.54e-02 |
Pos logit contributions | |
pron | +0.007 |
masterpiece | +0.006 |
favourite | +0.006 |
ail | +0.006 |
collection | +0.005 |
Neg logit contributions | |
crow | -0.007 |
Bad | -0.007 |
Leave | -0.007 |
hunters | -0.006 |
Stop | -0.006 |
Token | cozy |
Token ID | 37438 |
Feature activation | -6.21e+00 |
Pos logit contributions | |
Bad | +0.008 |
crow | +0.008 |
Leave | +0.008 |
Stop | +0.007 |
hunters | +0.007 |
Neg logit contributions | |
pron | -0.005 |
favourite | -0.005 |
ail | -0.005 |
masterpiece | -0.005 |
ilda | -0.005 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.67e-01 |
Pos logit contributions | |
pron | +0.415 |
masterpiece | +0.382 |
favourite | +0.377 |
collection | +0.347 |
usual | +0.339 |
Neg logit contributions | |
crow | -0.307 |
Bad | -0.303 |
Leave | -0.299 |
Stop | -0.289 |
vious | -0.283 |
Token | her |
Token ID | 607 |
Feature activation | -2.77e+00 |
Pos logit contributions | |
crow | +0.030 |
Leave | +0.029 |
Bad | +0.028 |
Stop | +0.027 |
vious | +0.027 |
Neg logit contributions | |
pron | -0.028 |
favourite | -0.026 |
masterpiece | -0.026 |
collection | -0.023 |
brand | -0.023 |
Token | mother |
Token ID | 2802 |
Feature activation | -6.70e+00 |
Pos logit contributions | |
pron | +0.194 |
onna | +0.149 |
masterpiece | +0.145 |
cakes | +0.130 |
ail | +0.129 |
Neg logit contributions | |
hunters | -0.226 |
Bad | -0.219 |
eper | -0.217 |
crow | -0.216 |
Leave | -0.204 |
Token | holding |
Token ID | 4769 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
pron | +0.520 |
favourite | +0.473 |
masterpiece | +0.472 |
collection | +0.428 |
usual | +0.420 |
Neg logit contributions | |
crow | -0.536 |
Leave | -0.515 |
Bad | -0.513 |
Stop | -0.494 |
vious | -0.489 |
Token | a |
Token ID | 257 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
pron | +0.088 |
favourite | +0.083 |
masterpiece | +0.080 |
collection | +0.073 |
brand | +0.072 |
Neg logit contributions | |
crow | -0.100 |
Leave | -0.098 |
Bad | -0.097 |
Stop | -0.094 |
vious | -0.093 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.67e+01 |
Pos logit contributions | |
Bad | +0.010 |
crow | +0.010 |
Leave | +0.009 |
Stop | +0.009 |
Hey | +0.009 |
Neg logit contributions | |
pron | -0.006 |
masterpiece | -0.006 |
favourite | -0.006 |
ail | -0.006 |
usual | -0.006 |
Token | - |
Token ID | 12 |
Feature activation | -2.96e-01 |
Pos logit contributions | |
pron | +1.316 |
favourite | +1.192 |
masterpiece | +1.185 |
collection | +1.076 |
brand | +1.057 |
Neg logit contributions | |
crow | -1.324 |
Leave | -1.277 |
Bad | -1.269 |
Stop | -1.220 |
vious | -1.210 |
Token | b |
Token ID | 65 |
Feature activation | +4.26e+00 |
Pos logit contributions | |
pron | +0.023 |
favourite | +0.021 |
masterpiece | +0.020 |
favorite | +0.018 |
usual | +0.018 |
Neg logit contributions | |
Leave | -0.023 |
Bad | -0.023 |
crow | -0.023 |
ases | -0.022 |
eper | -0.021 |
Token | aked |
Token ID | 4335 |
Feature activation | +7.87e+00 |
Pos logit contributions | |
crow | +0.344 |
Leave | +0.327 |
Bad | +0.326 |
vious | +0.313 |
eper | +0.312 |
Neg logit contributions | |
pron | -0.320 |
masterpiece | -0.294 |
favourite | -0.292 |
collection | -0.265 |
usual | -0.263 |
Token | muff |
Token ID | 27563 |
Feature activation | -3.66e+00 |
Pos logit contributions | |
crow | +0.625 |
Bad | +0.616 |
Leave | +0.608 |
Stop | +0.587 |
vious | +0.573 |
Neg logit contributions | |
pron | -0.609 |
favourite | -0.559 |
masterpiece | -0.558 |
collection | -0.505 |
brand | -0.494 |
Token | in |
Token ID | 259 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
pron | +0.397 |
favourite | +0.368 |
masterpiece | +0.367 |
collection | +0.346 |
ail | +0.340 |
Neg logit contributions | |
crow | -0.182 |
Leave | -0.179 |
Bad | -0.175 |
Stop | -0.164 |
vious | -0.154 |
Token | went |
Token ID | 1816 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
crow | +0.105 |
Leave | +0.101 |
Bad | +0.100 |
Stop | +0.097 |
vious | +0.097 |
Neg logit contributions | |
pron | -0.082 |
favourite | -0.074 |
masterpiece | -0.073 |
collection | -0.065 |
brand | -0.065 |
Token | home |
Token ID | 1363 |
Feature activation | -5.20e-01 |
Pos logit contributions | |
pron | +0.098 |
favourite | +0.090 |
masterpiece | +0.089 |
collection | +0.081 |
usual | +0.080 |
Neg logit contributions | |
crow | -0.094 |
Leave | -0.090 |
Bad | -0.089 |
Stop | -0.085 |
vious | -0.085 |
Token | and |
Token ID | 290 |
Feature activation | -7.52e-02 |
Pos logit contributions | |
pron | +0.047 |
favourite | +0.044 |
masterpiece | +0.044 |
collection | +0.040 |
usual | +0.039 |
Neg logit contributions | |
crow | -0.032 |
Leave | -0.030 |
Bad | -0.030 |
eper | -0.028 |
vious | -0.028 |
Token | enjoyed |
Token ID | 8359 |
Feature activation | -4.56e+00 |
Pos logit contributions | |
pron | +0.006 |
masterpiece | +0.005 |
favourite | +0.005 |
favorite | +0.005 |
ail | +0.005 |
Neg logit contributions | |
crow | -0.006 |
Bad | -0.005 |
Leave | -0.005 |
eper | -0.005 |
Stop | -0.005 |
Token | his |
Token ID | 465 |
Feature activation | +2.78e+00 |
Pos logit contributions | |
pron | +0.450 |
masterpiece | +0.414 |
favourite | +0.409 |
ail | +0.391 |
ouse | +0.376 |
Neg logit contributions | |
Leave | -0.252 |
crow | -0.250 |
Bad | -0.248 |
Stop | -0.221 |
hunters | -0.221 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.67e+01 |
Pos logit contributions | |
crow | +0.229 |
Bad | +0.224 |
Leave | +0.221 |
vious | +0.211 |
Stop | +0.209 |
Neg logit contributions | |
pron | -0.186 |
masterpiece | -0.161 |
favourite | -0.156 |
ail | -0.152 |
ilda | -0.151 |
Token | caught |
Token ID | 4978 |
Feature activation | +3.54e+00 |
Pos logit contributions | |
pron | +1.133 |
masterpiece | +1.024 |
favourite | +1.023 |
collection | +0.916 |
brand | +0.906 |
Neg logit contributions | |
crow | -1.480 |
Leave | -1.420 |
Bad | -1.415 |
Stop | -1.370 |
vious | -1.352 |
Token | fish |
Token ID | 5916 |
Feature activation | +5.57e+00 |
Pos logit contributions | |
crow | +0.289 |
Leave | +0.279 |
Bad | +0.277 |
Stop | +0.268 |
vious | +0.265 |
Neg logit contributions | |
pron | -0.270 |
favourite | -0.247 |
masterpiece | -0.245 |
collection | -0.223 |
usual | -0.218 |
Token | . |
Token ID | 13 |
Feature activation | +4.78e-02 |
Pos logit contributions | |
crow | +0.447 |
Leave | +0.429 |
Bad | +0.427 |
Stop | +0.411 |
vious | +0.408 |
Neg logit contributions | |
pron | -0.431 |
favourite | -0.391 |
masterpiece | -0.390 |
collection | -0.353 |
usual | -0.347 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -8.70e-01 |
Pos logit contributions | |
crow | +0.004 |
Leave | +0.003 |
Bad | +0.003 |
Stop | +0.003 |
vious | +0.003 |
Neg logit contributions | |
pron | -0.004 |
masterpiece | -0.004 |
favourite | -0.004 |
collection | -0.003 |
usual | -0.003 |
Token | Once |
Token ID | 7454 |
Feature activation | +2.14e+00 |
Pos logit contributions | |
pron | +0.060 |
favourite | +0.056 |
masterpiece | +0.053 |
usual | +0.051 |
ail | +0.050 |
Neg logit contributions | |
Leave | -0.072 |
Bad | -0.071 |
crow | -0.070 |
vious | -0.068 |
Stop | -0.066 |
Token | Tim |
Token ID | 5045 |
Feature activation | -2.04e-01 |
Pos logit contributions | |
pron | +0.132 |
masterpiece | +0.122 |
favourite | +0.120 |
collection | +0.111 |
brand | +0.108 |
Neg logit contributions | |
crow | -0.102 |
Bad | -0.096 |
Leave | -0.096 |
Stop | -0.093 |
vious | -0.090 |
Token | my |
Token ID | 1820 |
Feature activation | +3.23e+00 |
Pos logit contributions | |
pron | +0.018 |
favourite | +0.017 |
masterpiece | +0.017 |
collection | +0.016 |
usual | +0.016 |
Neg logit contributions | |
crow | -0.014 |
Leave | -0.013 |
Bad | -0.013 |
Stop | -0.013 |
vious | -0.012 |
Token | a |
Token ID | 257 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
crow | +0.232 |
Bad | +0.220 |
Leave | +0.219 |
vious | +0.214 |
Stop | +0.209 |
Neg logit contributions | |
pron | -0.273 |
masterpiece | -0.250 |
favourite | -0.244 |
collection | -0.227 |
brand | -0.223 |
Token | warm |
Token ID | 5814 |
Feature activation | -4.27e+00 |
Pos logit contributions | |
crow | +0.008 |
Bad | +0.008 |
vious | +0.008 |
Stop | +0.007 |
bully | +0.007 |
Neg logit contributions | |
masterpiece | -0.008 |
favourite | -0.007 |
yt | -0.007 |
pron | -0.007 |
ail | -0.007 |
Token | , |
Token ID | 11 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
pron | +0.304 |
masterpiece | +0.300 |
favourite | +0.295 |
usual | +0.272 |
collection | +0.271 |
Neg logit contributions | |
crow | -0.335 |
Leave | -0.304 |
Bad | -0.303 |
vious | -0.295 |
Stop | -0.289 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.67e+01 |
Pos logit contributions | |
Leave | +0.097 |
Bad | +0.095 |
crow | +0.092 |
vious | +0.091 |
Stop | +0.088 |
Neg logit contributions | |
pron | -0.071 |
ail | -0.071 |
masterpiece | -0.069 |
collection | -0.065 |
favourite | -0.065 |
Token | baked |
Token ID | 22979 |
Feature activation | -7.49e+00 |
Pos logit contributions | |
pron | +1.006 |
masterpiece | +0.905 |
favourite | +0.904 |
collection | +0.798 |
usual | +0.778 |
Neg logit contributions | |
crow | -1.608 |
Leave | -1.557 |
Bad | -1.556 |
Stop | -1.509 |
vious | -1.492 |
Token | cookie |
Token ID | 19751 |
Feature activation | -3.60e+00 |
Pos logit contributions | |
pron | +0.520 |
favourite | +0.468 |
masterpiece | +0.468 |
collection | +0.419 |
usual | +0.411 |
Neg logit contributions | |
crow | -0.661 |
Leave | -0.634 |
Bad | -0.632 |
Stop | -0.612 |
vious | -0.607 |
Token | . |
Token ID | 13 |
Feature activation | +3.79e-02 |
Pos logit contributions | |
pron | +0.284 |
masterpiece | +0.262 |
favourite | +0.261 |
usual | +0.237 |
brand | +0.236 |
Neg logit contributions | |
crow | -0.275 |
Bad | -0.252 |
Leave | -0.249 |
vious | -0.242 |
Stop | -0.241 |
Token |
|
Token ID | 198 |
Feature activation | -4.82e-01 |
Pos logit contributions | |
crow | +0.002 |
Bad | +0.002 |
Leave | +0.002 |
Stop | +0.002 |
Hey | +0.002 |
Neg logit contributions | |
pron | -0.003 |
masterpiece | -0.003 |
favourite | -0.003 |
collection | -0.003 |
usual | -0.003 |
Token |
|
Token ID | 198 |
Feature activation | -4.67e-01 |
Pos logit contributions | |
pron | +0.045 |
masterpiece | +0.044 |
favourite | +0.043 |
collection | +0.041 |
usual | +0.039 |
Neg logit contributions | |
crow | -0.029 |
Bad | -0.027 |
Leave | -0.027 |
Stop | -0.026 |
Hey | -0.025 |
Token | the |
Token ID | 262 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
crow | +0.096 |
Leave | +0.092 |
Bad | +0.092 |
Stop | +0.088 |
vious | +0.088 |
Neg logit contributions | |
pron | -0.100 |
favourite | -0.093 |
masterpiece | -0.092 |
collection | -0.084 |
usual | -0.083 |
Token | church |
Token ID | 4928 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
crow | +0.107 |
Leave | +0.103 |
Bad | +0.102 |
Stop | +0.098 |
vious | +0.098 |
Neg logit contributions | |
pron | -0.102 |
favourite | -0.093 |
masterpiece | -0.092 |
collection | -0.083 |
brand | -0.082 |
Token | and |
Token ID | 290 |
Feature activation | -5.77e-02 |
Pos logit contributions | |
pron | +0.104 |
favourite | +0.093 |
masterpiece | +0.093 |
brand | +0.085 |
collection | +0.084 |
Neg logit contributions | |
crow | -0.090 |
Leave | -0.083 |
Bad | -0.083 |
vious | -0.080 |
Stop | -0.078 |
Token | noticed |
Token ID | 6810 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
pron | +0.004 |
masterpiece | +0.004 |
favourite | +0.004 |
collection | +0.003 |
brand | +0.003 |
Neg logit contributions | |
crow | -0.005 |
Bad | -0.005 |
Leave | -0.005 |
Stop | -0.004 |
vious | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
pron | +0.097 |
favourite | +0.089 |
masterpiece | +0.088 |
collection | +0.080 |
usual | +0.079 |
Neg logit contributions | |
crow | -0.100 |
Leave | -0.097 |
Bad | -0.096 |
Stop | -0.092 |
vious | -0.091 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.67e+01 |
Pos logit contributions | |
crow | +0.011 |
Bad | +0.011 |
Leave | +0.010 |
Stop | +0.010 |
Hey | +0.010 |
Neg logit contributions | |
pron | -0.009 |
favourite | -0.009 |
masterpiece | -0.008 |
ail | -0.008 |
collection | -0.008 |
Token | dug |
Token ID | 18735 |
Feature activation | -2.05e+00 |
Pos logit contributions | |
pron | +1.327 |
favourite | +1.210 |
masterpiece | +1.206 |
collection | +1.100 |
usual | +1.079 |
Neg logit contributions | |
crow | -1.302 |
Leave | -1.251 |
Bad | -1.245 |
Stop | -1.198 |
vious | -1.183 |
Token | hole |
Token ID | 7604 |
Feature activation | +6.25e-01 |
Pos logit contributions | |
pron | +0.158 |
favourite | +0.145 |
masterpiece | +0.145 |
collection | +0.132 |
usual | +0.130 |
Neg logit contributions | |
crow | -0.162 |
Bad | -0.158 |
Leave | -0.158 |
Stop | -0.151 |
vious | -0.148 |
Token | in |
Token ID | 287 |
Feature activation | -2.08e+00 |
Pos logit contributions | |
crow | +0.045 |
Leave | +0.043 |
Bad | +0.042 |
vious | +0.041 |
Stop | +0.040 |
Neg logit contributions | |
pron | -0.053 |
favourite | -0.049 |
masterpiece | -0.048 |
collection | -0.044 |
brand | -0.043 |
Token | the |
Token ID | 262 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
pron | +0.167 |
masterpiece | +0.156 |
favourite | +0.155 |
collection | +0.142 |
usual | +0.141 |
Neg logit contributions | |
crow | -0.157 |
Leave | -0.151 |
Bad | -0.149 |
vious | -0.145 |
Stop | -0.144 |
Token | ground |
Token ID | 2323 |
Feature activation | -8.37e-01 |
Pos logit contributions | |
crow | +0.109 |
Leave | +0.105 |
Bad | +0.105 |
Stop | +0.101 |
vious | +0.099 |
Neg logit contributions | |
pron | -0.101 |
favourite | -0.092 |
masterpiece | -0.091 |
collection | -0.083 |
brand | -0.081 |
Token | sky |
Token ID | 6766 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
crow | +0.115 |
Leave | +0.111 |
Bad | +0.110 |
Stop | +0.106 |
vious | +0.105 |
Neg logit contributions | |
pron | -0.099 |
favourite | -0.089 |
masterpiece | -0.089 |
collection | -0.080 |
usual | -0.079 |
Token | . |
Token ID | 13 |
Feature activation | +9.97e-02 |
Pos logit contributions | |
crow | +0.097 |
Leave | +0.092 |
Bad | +0.092 |
vious | +0.088 |
Stop | +0.088 |
Neg logit contributions | |
pron | -0.114 |
masterpiece | -0.103 |
favourite | -0.103 |
collection | -0.095 |
brand | -0.093 |
Token | The |
Token ID | 383 |
Feature activation | -5.58e-02 |
Pos logit contributions | |
crow | +0.008 |
Bad | +0.007 |
Leave | +0.007 |
Stop | +0.007 |
vious | +0.007 |
Neg logit contributions | |
pron | -0.008 |
favourite | -0.007 |
masterpiece | -0.007 |
usual | -0.006 |
collection | -0.006 |
Token | grass |
Token ID | 8701 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
pron | +0.004 |
masterpiece | +0.004 |
favourite | +0.004 |
collection | +0.004 |
brand | +0.003 |
Neg logit contributions | |
crow | -0.004 |
Bad | -0.004 |
Leave | -0.004 |
Stop | -0.004 |
vious | -0.004 |
Token | was |
Token ID | 373 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
crow | +0.009 |
Leave | +0.008 |
Bad | +0.008 |
Stop | +0.008 |
Hey | +0.008 |
Neg logit contributions | |
pron | -0.008 |
masterpiece | -0.007 |
favourite | -0.007 |
brand | -0.007 |
collection | -0.006 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.67e+01 |
Pos logit contributions | |
pron | +0.111 |
masterpiece | +0.102 |
ail | +0.098 |
collection | +0.097 |
favourite | +0.095 |
Neg logit contributions | |
crow | -0.117 |
Leave | -0.116 |
Bad | -0.116 |
vious | -0.109 |
Stop | -0.107 |
Token | cut |
Token ID | 2005 |
Feature activation | -3.71e+00 |
Pos logit contributions | |
pron | +1.831 |
favourite | +1.723 |
masterpiece | +1.710 |
collection | +1.604 |
usual | +1.580 |
Neg logit contributions | |
crow | -0.799 |
Leave | -0.754 |
Bad | -0.749 |
Stop | -0.699 |
vious | -0.687 |
Token | , |
Token ID | 11 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
pron | +0.305 |
favourite | +0.283 |
masterpiece | +0.276 |
collection | +0.255 |
usual | +0.254 |
Neg logit contributions | |
crow | -0.272 |
Leave | -0.271 |
Bad | -0.270 |
Stop | -0.259 |
vious | -0.251 |
Token | and |
Token ID | 290 |
Feature activation | -5.38e-02 |
Pos logit contributions | |
crow | +0.089 |
Leave | +0.086 |
Bad | +0.085 |
Stop | +0.082 |
vious | +0.080 |
Neg logit contributions | |
pron | -0.095 |
masterpiece | -0.086 |
favourite | -0.086 |
collection | -0.079 |
usual | -0.077 |
Token | a |
Token ID | 257 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
pron | +0.005 |
masterpiece | +0.004 |
favourite | +0.004 |
ail | +0.004 |
collection | +0.004 |
Neg logit contributions | |
crow | -0.004 |
Bad | -0.004 |
Leave | -0.004 |
Stop | -0.003 |
vious | -0.003 |
Token | gentle |
Token ID | 10296 |
Feature activation | +3.46e+00 |
Pos logit contributions | |
crow | +0.011 |
Bad | +0.011 |
Leave | +0.010 |
Stop | +0.010 |
Hey | +0.010 |
Neg logit contributions | |
pron | -0.009 |
favourite | -0.008 |
masterpiece | -0.008 |
ail | -0.008 |
collection | -0.007 |
Token | to |
Token ID | 284 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
crow | +0.149 |
Leave | +0.144 |
Bad | +0.142 |
Stop | +0.136 |
vious | +0.136 |
Neg logit contributions | |
pron | -0.151 |
masterpiece | -0.135 |
favourite | -0.134 |
collection | -0.122 |
ail | -0.120 |
Token | play |
Token ID | 711 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
pron | +0.142 |
favourite | +0.129 |
masterpiece | +0.127 |
ail | +0.118 |
collection | +0.115 |
Neg logit contributions | |
crow | -0.110 |
Leave | -0.106 |
Bad | -0.104 |
Stop | -0.099 |
ases | -0.098 |
Token | and |
Token ID | 290 |
Feature activation | -3.05e-02 |
Pos logit contributions | |
crow | +0.101 |
Leave | +0.097 |
Bad | +0.097 |
Stop | +0.093 |
vious | +0.092 |
Neg logit contributions | |
pron | -0.107 |
favourite | -0.097 |
masterpiece | -0.097 |
collection | -0.089 |
usual | -0.087 |
Token | explore |
Token ID | 7301 |
Feature activation | -6.59e+00 |
Pos logit contributions | |
pron | +0.003 |
masterpiece | +0.002 |
favourite | +0.002 |
ail | +0.002 |
brand | +0.002 |
Neg logit contributions | |
crow | -0.002 |
Leave | -0.002 |
Bad | -0.002 |
Stop | -0.002 |
hunters | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | +1.35e+00 |
Pos logit contributions | |
pron | +0.557 |
favourite | +0.508 |
masterpiece | +0.507 |
collection | +0.464 |
usual | +0.456 |
Neg logit contributions | |
crow | -0.484 |
Leave | -0.464 |
Bad | -0.461 |
Stop | -0.442 |
vious | -0.437 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.66e+01 |
Pos logit contributions | |
crow | +0.119 |
Leave | +0.115 |
Stop | +0.114 |
Bad | +0.114 |
vious | +0.109 |
Neg logit contributions | |
pron | -0.094 |
masterpiece | -0.078 |
favourite | -0.077 |
collection | -0.073 |
ail | -0.072 |
Token | organized |
Token ID | 8389 |
Feature activation | -3.13e+00 |
Pos logit contributions | |
pron | +1.257 |
favourite | +1.137 |
masterpiece | +1.133 |
collection | +1.028 |
brand | +1.016 |
Neg logit contributions | |
crow | -1.375 |
Leave | -1.303 |
Bad | -1.294 |
Stop | -1.257 |
vious | -1.221 |
Token | lawn |
Token ID | 21133 |
Feature activation | -5.80e+00 |
Pos logit contributions | |
pron | +0.209 |
favourite | +0.177 |
masterpiece | +0.174 |
brand | +0.163 |
onna | +0.161 |
Neg logit contributions | |
crow | -0.291 |
Stop | -0.264 |
Leave | -0.264 |
Bad | -0.260 |
vious | -0.254 |
Token | . |
Token ID | 13 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
pron | +0.498 |
masterpiece | +0.446 |
favourite | +0.445 |
brand | +0.401 |
collection | +0.400 |
Neg logit contributions | |
crow | -0.418 |
Leave | -0.392 |
Bad | -0.388 |
vious | -0.376 |
Stop | -0.369 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -7.78e-01 |
Pos logit contributions | |
crow | +0.008 |
Leave | +0.007 |
Bad | +0.007 |
Stop | +0.007 |
vious | +0.007 |
Neg logit contributions | |
pron | -0.009 |
masterpiece | -0.008 |
favourite | -0.008 |
collection | -0.007 |
usual | -0.007 |
Token | Once |
Token ID | 7454 |
Feature activation | +2.21e+00 |
Pos logit contributions | |
pron | +0.065 |
favourite | +0.061 |
masterpiece | +0.059 |
usual | +0.055 |
collection | +0.055 |
Neg logit contributions | |
Leave | -0.054 |
crow | -0.054 |
Bad | -0.054 |
vious | -0.051 |
Stop | -0.050 |
Token | . |
Token ID | 13 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
pron | +0.313 |
favourite | +0.295 |
masterpiece | +0.282 |
usual | +0.265 |
ail | +0.264 |
Neg logit contributions | |
crow | -0.192 |
vious | -0.164 |
ting | -0.163 |
eper | -0.159 |
Stop | -0.158 |
Token | The |
Token ID | 383 |
Feature activation | -2.27e-02 |
Pos logit contributions | |
crow | +0.008 |
Stop | +0.008 |
Leave | +0.008 |
vious | +0.008 |
ases | +0.008 |
Neg logit contributions | |
pron | -0.010 |
masterpiece | -0.009 |
favourite | -0.008 |
collection | -0.008 |
ail | -0.008 |
Token | sweet |
Token ID | 6029 |
Feature activation | -5.62e+00 |
Pos logit contributions | |
pron | +0.002 |
onna | +0.001 |
favourite | +0.001 |
masterpiece | +0.001 |
brand | +0.001 |
Neg logit contributions | |
crow | -0.002 |
Bad | -0.002 |
Stop | -0.002 |
eper | -0.002 |
vious | -0.002 |
Token | smell |
Token ID | 8508 |
Feature activation | -5.01e+00 |
Pos logit contributions | |
pron | +0.473 |
favourite | +0.430 |
masterpiece | +0.422 |
usual | +0.405 |
brand | +0.397 |
Neg logit contributions | |
crow | -0.411 |
vious | -0.373 |
Stop | -0.363 |
Leave | -0.361 |
Bad | -0.352 |
Token | of |
Token ID | 286 |
Feature activation | +2.50e+00 |
Pos logit contributions | |
pron | +0.440 |
masterpiece | +0.416 |
favourite | +0.415 |
collection | +0.403 |
icol | +0.394 |
Neg logit contributions | |
crow | -0.326 |
Leave | -0.301 |
Stop | -0.298 |
vious | -0.298 |
Bad | -0.294 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.66e+01 |
Pos logit contributions | |
crow | +0.151 |
Bad | +0.145 |
Leave | +0.140 |
Stop | +0.138 |
vious | +0.132 |
Neg logit contributions | |
pron | -0.234 |
favourite | -0.213 |
ail | -0.213 |
ouse | -0.212 |
masterpiece | -0.209 |
Token | baked |
Token ID | 22979 |
Feature activation | -7.41e+00 |
Pos logit contributions | |
pron | +0.811 |
favourite | +0.716 |
masterpiece | +0.706 |
collection | +0.629 |
brand | +0.617 |
Neg logit contributions | |
crow | -1.755 |
Bad | -1.692 |
Stop | -1.682 |
Leave | -1.676 |
vious | -1.633 |
Token | port |
Token ID | 2493 |
Feature activation | -3.20e+00 |
Pos logit contributions | |
pron | +0.541 |
favourite | +0.503 |
masterpiece | +0.487 |
usual | +0.454 |
collection | +0.453 |
Neg logit contributions | |
crow | -0.595 |
Leave | -0.534 |
Stop | -0.534 |
Bad | -0.531 |
vious | -0.523 |
Token | filled |
Token ID | 5901 |
Feature activation | -4.53e+00 |
Pos logit contributions | |
pron | +0.288 |
favourite | +0.260 |
masterpiece | +0.252 |
usual | +0.236 |
complexion | +0.235 |
Neg logit contributions | |
crow | -0.209 |
Leave | -0.183 |
vious | -0.181 |
Bad | -0.179 |
Stop | -0.176 |
Token | the |
Token ID | 262 |
Feature activation | +1.39e+00 |
Pos logit contributions | |
pron | +0.420 |
favourite | +0.381 |
masterpiece | +0.381 |
collection | +0.363 |
brand | +0.356 |
Neg logit contributions | |
crow | -0.281 |
Bad | -0.267 |
Stop | -0.259 |
Leave | -0.253 |
Hey | -0.245 |
Token | house |
Token ID | 2156 |
Feature activation | -1.56e-01 |
Pos logit contributions | |
crow | +0.125 |
Bad | +0.119 |
Leave | +0.118 |
Stop | +0.117 |
vious | +0.116 |
Neg logit contributions | |
pron | -0.093 |
favourite | -0.081 |
masterpiece | -0.080 |
collection | -0.073 |
usual | -0.073 |
Token | so |
Token ID | 523 |
Feature activation | -7.45e-01 |
Pos logit contributions | |
pron | +0.121 |
masterpiece | +0.108 |
favourite | +0.108 |
collection | +0.101 |
usual | +0.099 |
Neg logit contributions | |
crow | -0.111 |
Leave | -0.107 |
Bad | -0.107 |
vious | -0.102 |
eper | -0.100 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
pron | +0.061 |
masterpiece | +0.055 |
favourite | +0.054 |
collection | +0.050 |
usual | +0.049 |
Neg logit contributions | |
crow | -0.056 |
Leave | -0.055 |
Bad | -0.055 |
Stop | -0.052 |
vious | -0.052 |
Token | to |
Token ID | 284 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
crow | +0.155 |
Leave | +0.149 |
Bad | +0.149 |
Stop | +0.143 |
vious | +0.142 |
Neg logit contributions | |
pron | -0.140 |
favourite | -0.126 |
masterpiece | -0.125 |
collection | -0.113 |
brand | -0.111 |
Token | see |
Token ID | 766 |
Feature activation | -3.89e-02 |
Pos logit contributions | |
pron | +0.153 |
favourite | +0.141 |
masterpiece | +0.138 |
brand | +0.130 |
collection | +0.129 |
Neg logit contributions | |
crow | -0.094 |
Leave | -0.091 |
Bad | -0.088 |
Stop | -0.085 |
vious | -0.083 |
Token | a |
Token ID | 257 |
Feature activation | +2.14e-01 |
Pos logit contributions | |
pron | +0.003 |
favourite | +0.003 |
masterpiece | +0.003 |
collection | +0.003 |
brand | +0.003 |
Neg logit contributions | |
crow | -0.003 |
Leave | -0.003 |
Bad | -0.003 |
Stop | -0.003 |
vious | -0.003 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.66e+01 |
Pos logit contributions | |
Bad | +0.017 |
crow | +0.017 |
bully | +0.016 |
Stop | +0.016 |
Leave | +0.016 |
Neg logit contributions | |
favourite | -0.013 |
ail | -0.012 |
pron | -0.012 |
usual | -0.012 |
masterpiece | -0.012 |
Token | baked |
Token ID | 22979 |
Feature activation | -7.38e+00 |
Pos logit contributions | |
pron | +1.155 |
favourite | +1.039 |
masterpiece | +1.035 |
collection | +0.928 |
usual | +0.908 |
Neg logit contributions | |
crow | -1.459 |
Leave | -1.408 |
Bad | -1.402 |
Stop | -1.356 |
vious | -1.342 |
Token | bean |
Token ID | 26394 |
Feature activation | -5.64e+00 |
Pos logit contributions | |
pron | +0.443 |
favourite | +0.413 |
masterpiece | +0.398 |
usual | +0.359 |
collection | +0.355 |
Neg logit contributions | |
crow | -0.711 |
Bad | -0.663 |
Leave | -0.662 |
Stop | -0.649 |
vious | -0.646 |
Token | . |
Token ID | 13 |
Feature activation | +1.58e-01 |
Pos logit contributions | |
pron | +0.473 |
favourite | +0.432 |
masterpiece | +0.419 |
usual | +0.395 |
brand | +0.384 |
Neg logit contributions | |
crow | -0.411 |
Leave | -0.381 |
vious | -0.373 |
Bad | -0.373 |
ases | -0.360 |
Token | After |
Token ID | 2293 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
crow | +0.010 |
Bad | +0.010 |
Leave | +0.010 |
Stop | +0.009 |
vious | +0.009 |
Neg logit contributions | |
pron | -0.015 |
masterpiece | -0.014 |
favourite | -0.013 |
collection | -0.013 |
usual | -0.013 |
Token | that |
Token ID | 326 |
Feature activation | -2.74e+00 |
Pos logit contributions | |
pron | +0.107 |
favourite | +0.096 |
masterpiece | +0.095 |
ouse | +0.088 |
ail | +0.087 |
Neg logit contributions | |
Bad | -0.088 |
crow | -0.087 |
Leave | -0.086 |
hunters | -0.085 |
Stop | -0.080 |
Token | mom |
Token ID | 1995 |
Feature activation | -5.19e+00 |
Pos logit contributions | |
Bad | +0.249 |
crow | +0.245 |
hunters | +0.240 |
eper | +0.238 |
Leave | +0.235 |
Neg logit contributions | |
pron | -0.183 |
masterpiece | -0.152 |
yt | -0.143 |
oph | -0.142 |
favourite | -0.142 |
Token | my |
Token ID | 1820 |
Feature activation | +3.34e+00 |
Pos logit contributions | |
pron | +0.488 |
favourite | +0.465 |
masterpiece | +0.457 |
collection | +0.423 |
usual | +0.420 |
Neg logit contributions | |
crow | -0.323 |
Leave | -0.309 |
Bad | -0.308 |
Stop | -0.294 |
vious | -0.289 |
Token | surprised |
Token ID | 6655 |
Feature activation | +2.94e+00 |
Pos logit contributions | |
crow | +0.257 |
Bad | +0.246 |
Leave | +0.244 |
eper | +0.238 |
Stop | +0.234 |
Neg logit contributions | |
pron | -0.258 |
favourite | -0.249 |
masterpiece | -0.240 |
usual | -0.222 |
favorite | -0.214 |
Token | him |
Token ID | 683 |
Feature activation | +7.62e+00 |
Pos logit contributions | |
crow | +0.135 |
Leave | +0.129 |
Bad | +0.128 |
vious | +0.116 |
Stop | +0.115 |
Neg logit contributions | |
pron | -0.327 |
masterpiece | -0.306 |
favourite | -0.306 |
collection | -0.287 |
usual | -0.285 |
Token | with |
Token ID | 351 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
crow | +0.528 |
Leave | +0.518 |
Bad | +0.511 |
vious | +0.483 |
Stop | +0.476 |
Neg logit contributions | |
pron | -0.659 |
favourite | -0.601 |
masterpiece | -0.599 |
collection | -0.550 |
usual | -0.542 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.65e+01 |
Pos logit contributions | |
crow | +0.115 |
Leave | +0.112 |
Bad | +0.112 |
Stop | +0.106 |
vious | +0.105 |
Neg logit contributions | |
pron | -0.130 |
masterpiece | -0.119 |
favourite | -0.118 |
collection | -0.108 |
ail | -0.105 |
Token | baked |
Token ID | 22979 |
Feature activation | -7.33e+00 |
Pos logit contributions | |
pron | +1.101 |
favourite | +0.975 |
masterpiece | +0.961 |
collection | +0.851 |
usual | +0.834 |
Neg logit contributions | |
crow | -1.528 |
Leave | -1.469 |
Bad | -1.451 |
vious | -1.406 |
Stop | -1.403 |
Token | cookies |
Token ID | 14746 |
Feature activation | -1.94e+00 |
Pos logit contributions | |
pron | +0.494 |
favourite | +0.448 |
masterpiece | +0.436 |
brand | +0.393 |
collection | +0.393 |
Neg logit contributions | |
crow | -0.663 |
Leave | -0.658 |
Bad | -0.644 |
Stop | -0.625 |
vious | -0.619 |
Token | . |
Token ID | 13 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
pron | +0.175 |
masterpiece | +0.164 |
favourite | +0.163 |
collection | +0.148 |
brand | +0.146 |
Neg logit contributions | |
crow | -0.126 |
Bad | -0.116 |
Leave | -0.116 |
vious | -0.113 |
Stop | -0.110 |
Token | Tim |
Token ID | 5045 |
Feature activation | -6.23e-02 |
Pos logit contributions | |
Leave | +0.011 |
crow | +0.011 |
Stop | +0.010 |
Bad | +0.010 |
vious | +0.010 |
Neg logit contributions | |
pron | -0.019 |
masterpiece | -0.018 |
favourite | -0.017 |
collection | -0.017 |
usual | -0.016 |
Token | my |
Token ID | 1820 |
Feature activation | +3.36e+00 |
Pos logit contributions | |
pron | +0.006 |
masterpiece | +0.005 |
favourite | +0.005 |
collection | +0.005 |
brand | +0.005 |
Neg logit contributions | |
crow | -0.004 |
Leave | -0.004 |
Bad | -0.004 |
Stop | -0.004 |
vious | -0.004 |
Token | it |
Token ID | 340 |
Feature activation | +2.73e-01 |
Pos logit contributions | |
pron | +0.202 |
masterpiece | +0.182 |
favourite | +0.179 |
collection | +0.168 |
ail | +0.167 |
Neg logit contributions | |
crow | -0.186 |
Leave | -0.180 |
Bad | -0.178 |
Stop | -0.173 |
vious | -0.164 |
Token | when |
Token ID | 618 |
Feature activation | +6.98e+00 |
Pos logit contributions | |
crow | +0.017 |
Bad | +0.016 |
Leave | +0.016 |
vious | +0.016 |
hunters | +0.015 |
Neg logit contributions | |
pron | -0.025 |
masterpiece | -0.023 |
favourite | -0.021 |
ail | -0.021 |
complexion | -0.020 |
Token | the |
Token ID | 262 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
crow | +0.481 |
Leave | +0.471 |
Bad | +0.469 |
Stop | +0.442 |
vious | +0.418 |
Neg logit contributions | |
pron | -0.607 |
masterpiece | -0.555 |
favourite | -0.542 |
ail | -0.511 |
collection | -0.507 |
Token | snow |
Token ID | 6729 |
Feature activation | +7.32e-01 |
Pos logit contributions | |
crow | +0.127 |
Leave | +0.120 |
Bad | +0.120 |
Stop | +0.116 |
vious | +0.114 |
Neg logit contributions | |
pron | -0.114 |
masterpiece | -0.102 |
favourite | -0.102 |
collection | -0.094 |
brand | -0.091 |
Token | was |
Token ID | 373 |
Feature activation | -1.38e+00 |
Pos logit contributions | |
crow | +0.063 |
Bad | +0.058 |
Leave | +0.058 |
Stop | +0.056 |
eper | +0.055 |
Neg logit contributions | |
pron | -0.053 |
masterpiece | -0.049 |
favourite | -0.048 |
brand | -0.044 |
collection | -0.043 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.65e+01 |
Pos logit contributions | |
pron | +0.103 |
masterpiece | +0.096 |
favourite | +0.094 |
collection | +0.092 |
ail | +0.087 |
Neg logit contributions | |
crow | -0.107 |
Bad | -0.104 |
Leave | -0.104 |
vious | -0.100 |
Stop | -0.098 |
Token | fallen |
Token ID | 9292 |
Feature activation | -5.86e-01 |
Pos logit contributions | |
pron | +1.284 |
masterpiece | +1.171 |
favourite | +1.158 |
collection | +1.064 |
brand | +1.044 |
Neg logit contributions | |
crow | -1.325 |
Leave | -1.271 |
Bad | -1.265 |
Stop | -1.219 |
vious | -1.204 |
Token | . |
Token ID | 13 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
pron | +0.052 |
favourite | +0.047 |
masterpiece | +0.047 |
collection | +0.043 |
usual | +0.042 |
Neg logit contributions | |
crow | -0.041 |
Leave | -0.039 |
Bad | -0.039 |
Stop | -0.037 |
vious | -0.037 |
Token | Today |
Token ID | 6288 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
crow | +0.013 |
Leave | +0.013 |
Bad | +0.013 |
Stop | +0.012 |
vious | +0.012 |
Neg logit contributions | |
pron | -0.015 |
masterpiece | -0.014 |
favourite | -0.014 |
collection | -0.012 |
usual | -0.012 |
Token | , |
Token ID | 11 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
pron | +0.149 |
masterpiece | +0.132 |
favourite | +0.129 |
brand | +0.119 |
collection | +0.117 |
Neg logit contributions | |
crow | -0.136 |
Bad | -0.124 |
Leave | -0.121 |
Stop | -0.119 |
Hey | -0.117 |
Token | he |
Token ID | 339 |
Feature activation | +6.30e-01 |
Pos logit contributions | |
crow | +0.096 |
Bad | +0.093 |
Leave | +0.092 |
Stop | +0.089 |
vious | +0.086 |
Neg logit contributions | |
pron | -0.100 |
masterpiece | -0.091 |
favourite | -0.090 |
collection | -0.083 |
brand | -0.081 |
Token | the |
Token ID | 262 |
Feature activation | +1.47e+00 |
Pos logit contributions | |
pron | +0.514 |
masterpiece | +0.465 |
favourite | +0.464 |
collection | +0.429 |
ail | +0.421 |
Neg logit contributions | |
crow | -0.435 |
Leave | -0.416 |
Bad | -0.414 |
Stop | -0.400 |
vious | -0.398 |
Token | sweet |
Token ID | 6029 |
Feature activation | -5.52e+00 |
Pos logit contributions | |
crow | +0.128 |
Bad | +0.127 |
vious | +0.125 |
Stop | +0.124 |
Leave | +0.121 |
Neg logit contributions | |
pron | -0.088 |
favourite | -0.077 |
ail | -0.075 |
masterpiece | -0.074 |
collection | -0.072 |
Token | aroma |
Token ID | 31242 |
Feature activation | -2.02e+00 |
Pos logit contributions | |
pron | +0.377 |
favourite | +0.338 |
masterpiece | +0.331 |
collection | +0.299 |
usual | +0.297 |
Neg logit contributions | |
crow | -0.495 |
Leave | -0.471 |
Bad | -0.469 |
Stop | -0.460 |
vious | -0.458 |
Token | of |
Token ID | 286 |
Feature activation | +2.60e+00 |
Pos logit contributions | |
pron | +0.196 |
favourite | +0.187 |
masterpiece | +0.181 |
collection | +0.174 |
ail | +0.172 |
Neg logit contributions | |
crow | -0.115 |
Leave | -0.107 |
vious | -0.103 |
Bad | -0.102 |
Stop | -0.102 |
Token | a |
Token ID | 257 |
Feature activation | +3.01e-01 |
Pos logit contributions | |
Bad | +0.150 |
crow | +0.149 |
Stop | +0.141 |
Leave | +0.141 |
vious | +0.133 |
Neg logit contributions | |
pron | -0.249 |
ail | -0.231 |
favourite | -0.229 |
ouse | -0.228 |
masterpiece | -0.218 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.65e+01 |
Pos logit contributions | |
Bad | +0.026 |
crow | +0.025 |
Leave | +0.025 |
Stop | +0.024 |
Hey | +0.023 |
Neg logit contributions | |
pron | -0.018 |
favourite | -0.018 |
ail | -0.016 |
masterpiece | -0.016 |
usual | -0.016 |
Token | baked |
Token ID | 22979 |
Feature activation | -7.29e+00 |
Pos logit contributions | |
pron | +0.972 |
favourite | +0.885 |
masterpiece | +0.869 |
collection | +0.770 |
usual | +0.765 |
Neg logit contributions | |
crow | -1.602 |
Bad | -1.530 |
Leave | -1.528 |
Stop | -1.500 |
vious | -1.484 |
Token | cake |
Token ID | 12187 |
Feature activation | -6.66e+00 |
Pos logit contributions | |
pron | +0.466 |
favourite | +0.447 |
masterpiece | +0.416 |
usual | +0.392 |
collection | +0.383 |
Neg logit contributions | |
crow | -0.658 |
Leave | -0.594 |
Bad | -0.593 |
vious | -0.588 |
Stop | -0.585 |
Token | . |
Token ID | 13 |
Feature activation | +2.48e-01 |
Pos logit contributions | |
pron | +0.653 |
favourite | +0.632 |
usual | +0.582 |
masterpiece | +0.581 |
aj | +0.570 |
Neg logit contributions | |
crow | -0.303 |
ting | -0.292 |
ters | -0.261 |
vious | -0.259 |
eper | -0.256 |
Token |
|
Token ID | 198 |
Feature activation | -2.58e-01 |
Pos logit contributions | |
crow | +0.014 |
Stop | +0.013 |
Leave | +0.012 |
Bad | +0.012 |
Hey | +0.012 |
Neg logit contributions | |
pron | -0.025 |
masterpiece | -0.024 |
favourite | -0.023 |
collection | -0.023 |
usual | -0.022 |
Token |
|
Token ID | 198 |
Feature activation | -2.60e-01 |
Pos logit contributions | |
pron | +0.023 |
masterpiece | +0.021 |
favourite | +0.021 |
collection | +0.020 |
usual | +0.019 |
Neg logit contributions | |
crow | -0.018 |
Leave | -0.017 |
Bad | -0.017 |
Stop | -0.016 |
Hey | -0.015 |
Token | he |
Token ID | 339 |
Feature activation | +6.89e-01 |
Pos logit contributions | |
crow | +0.103 |
Leave | +0.099 |
Bad | +0.097 |
vious | +0.095 |
Stop | +0.095 |
Neg logit contributions | |
pron | -0.102 |
favourite | -0.094 |
masterpiece | -0.091 |
usual | -0.084 |
collection | -0.084 |
Token | saw |
Token ID | 2497 |
Feature activation | +5.50e-01 |
Pos logit contributions | |
crow | +0.048 |
Leave | +0.047 |
Bad | +0.047 |
Stop | +0.046 |
Hey | +0.046 |
Neg logit contributions | |
pron | -0.058 |
favourite | -0.054 |
masterpiece | -0.053 |
usual | -0.048 |
brand | -0.048 |
Token | a |
Token ID | 257 |
Feature activation | +2.88e-01 |
Pos logit contributions | |
crow | +0.043 |
Leave | +0.041 |
Bad | +0.041 |
Stop | +0.039 |
vious | +0.039 |
Neg logit contributions | |
pron | -0.045 |
favourite | -0.041 |
masterpiece | -0.041 |
collection | -0.037 |
usual | -0.036 |
Token | basket |
Token ID | 7988 |
Feature activation | -7.03e+00 |
Pos logit contributions | |
Bad | +0.023 |
crow | +0.023 |
Leave | +0.022 |
Stop | +0.022 |
Hey | +0.021 |
Neg logit contributions | |
pron | -0.020 |
favourite | -0.019 |
masterpiece | -0.019 |
ail | -0.018 |
usual | -0.017 |
Token | of |
Token ID | 286 |
Feature activation | +2.60e+00 |
Pos logit contributions | |
pron | +0.576 |
favourite | +0.530 |
masterpiece | +0.527 |
collection | +0.479 |
usual | +0.471 |
Neg logit contributions | |
crow | -0.531 |
Leave | -0.510 |
Bad | -0.504 |
vious | -0.483 |
Stop | -0.483 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.65e+01 |
Pos logit contributions | |
crow | +0.245 |
Leave | +0.237 |
Bad | +0.236 |
Stop | +0.231 |
vious | +0.230 |
Neg logit contributions | |
pron | -0.163 |
masterpiece | -0.144 |
favourite | -0.142 |
collection | -0.127 |
ail | -0.126 |
Token | baked |
Token ID | 22979 |
Feature activation | -7.29e+00 |
Pos logit contributions | |
pron | +1.021 |
masterpiece | +0.922 |
favourite | +0.917 |
brand | +0.813 |
collection | +0.807 |
Neg logit contributions | |
crow | -1.565 |
Leave | -1.500 |
Bad | -1.497 |
Stop | -1.470 |
vious | -1.441 |
Token | bread |
Token ID | 8509 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
pron | +0.562 |
masterpiece | +0.514 |
favourite | +0.514 |
brand | +0.464 |
usual | +0.460 |
Neg logit contributions | |
crow | -0.587 |
Leave | -0.536 |
Bad | -0.535 |
vious | -0.528 |
Stop | -0.525 |
Token | . |
Token ID | 13 |
Feature activation | +2.45e-01 |
Pos logit contributions | |
crow | +0.069 |
Bad | +0.061 |
Leave | +0.061 |
vious | +0.061 |
hunters | +0.060 |
Neg logit contributions | |
pron | -0.092 |
masterpiece | -0.083 |
favourite | -0.083 |
brand | -0.076 |
usual | -0.076 |
Token | It |
Token ID | 632 |
Feature activation | +9.42e-01 |
Pos logit contributions | |
crow | +0.013 |
Leave | +0.013 |
Bad | +0.012 |
Stop | +0.012 |
vious | +0.012 |
Neg logit contributions | |
pron | -0.026 |
masterpiece | -0.024 |
favourite | -0.024 |
collection | -0.022 |
usual | -0.022 |
Token | smelled |
Token ID | 37609 |
Feature activation | -5.96e+00 |
Pos logit contributions | |
Hey | +0.058 |
crow | +0.057 |
Bad | +0.054 |
Stop | +0.053 |
Leave | +0.052 |
Neg logit contributions | |
pron | -0.086 |
masterpiece | -0.077 |
favourite | -0.075 |
aj | -0.074 |
complexion | -0.074 |
Token | crown |
Token ID | 12389 |
Feature activation | -4.77e+00 |
Pos logit contributions | |
pron | +0.129 |
favourite | +0.101 |
addy | +0.100 |
icol | +0.100 |
usual | +0.099 |
Neg logit contributions | |
crow | -0.120 |
hunters | -0.114 |
vious | -0.112 |
Leave | -0.112 |
eper | -0.110 |
Token | and |
Token ID | 290 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
pron | +0.420 |
favourite | +0.365 |
masterpiece | +0.365 |
bedroom | +0.351 |
icol | +0.345 |
Neg logit contributions | |
vious | -0.277 |
crow | -0.275 |
Leave | -0.261 |
ases | -0.252 |
eper | -0.252 |
Token | playing |
Token ID | 2712 |
Feature activation | +9.73e-01 |
Pos logit contributions | |
Bad | +0.008 |
crow | +0.008 |
Leave | +0.008 |
Stop | +0.007 |
hunters | +0.007 |
Neg logit contributions | |
pron | -0.008 |
favourite | -0.007 |
masterpiece | -0.007 |
ail | -0.007 |
brand | -0.006 |
Token | in |
Token ID | 287 |
Feature activation | -1.93e+00 |
Pos logit contributions | |
crow | +0.070 |
Leave | +0.067 |
Bad | +0.066 |
Stop | +0.064 |
vious | +0.063 |
Neg logit contributions | |
pron | -0.084 |
favourite | -0.077 |
masterpiece | -0.077 |
collection | -0.070 |
usual | -0.069 |
Token | the |
Token ID | 262 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
pron | +0.187 |
favourite | +0.172 |
masterpiece | +0.167 |
ail | +0.159 |
collection | +0.157 |
Neg logit contributions | |
crow | -0.116 |
Bad | -0.113 |
Leave | -0.112 |
Stop | -0.108 |
ases | -0.102 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.65e+01 |
Pos logit contributions | |
crow | +0.126 |
Leave | +0.121 |
Bad | +0.121 |
Stop | +0.119 |
vious | +0.114 |
Neg logit contributions | |
pron | -0.108 |
masterpiece | -0.096 |
favourite | -0.096 |
collection | -0.088 |
ail | -0.087 |
Token | cut |
Token ID | 2005 |
Feature activation | -3.54e+00 |
Pos logit contributions | |
pron | +1.090 |
favourite | +0.998 |
masterpiece | +0.984 |
collection | +0.886 |
usual | +0.876 |
Neg logit contributions | |
crow | -1.471 |
Bad | -1.414 |
Leave | -1.410 |
Stop | -1.384 |
vious | -1.344 |
Token | grass |
Token ID | 8701 |
Feature activation | +2.80e-01 |
Pos logit contributions | |
pron | +0.329 |
favourite | +0.303 |
masterpiece | +0.303 |
collection | +0.280 |
usual | +0.275 |
Neg logit contributions | |
crow | -0.229 |
Leave | -0.220 |
Bad | -0.218 |
Stop | -0.208 |
vious | -0.205 |
Token | . |
Token ID | 13 |
Feature activation | +2.56e-01 |
Pos logit contributions | |
crow | +0.020 |
Leave | +0.019 |
Bad | +0.019 |
vious | +0.018 |
Stop | +0.018 |
Neg logit contributions | |
pron | -0.024 |
favourite | -0.022 |
masterpiece | -0.022 |
brand | -0.020 |
collection | -0.019 |
Token | It |
Token ID | 632 |
Feature activation | +9.58e-01 |
Pos logit contributions | |
Stop | +0.011 |
Fighting | +0.010 |
If | +0.010 |
crow | +0.010 |
Leave | +0.010 |
Neg logit contributions | |
pron | -0.027 |
masterpiece | -0.025 |
collection | -0.024 |
usual | -0.024 |
complexion | -0.023 |
Token | looked |
Token ID | 3114 |
Feature activation | -4.59e+00 |
Pos logit contributions | |
crow | +0.067 |
Bad | +0.065 |
Stop | +0.064 |
vious | +0.064 |
Leave | +0.064 |
Neg logit contributions | |
pron | -0.083 |
favourite | -0.071 |
masterpiece | -0.070 |
usual | -0.067 |
complexion | -0.066 |
Token | tidy |
Token ID | 43044 |
Feature activation | -3.38e+00 |
Pos logit contributions | |
pron | +0.096 |
favourite | +0.087 |
masterpiece | +0.086 |
collection | +0.076 |
usual | +0.076 |
Neg logit contributions | |
crow | -0.121 |
Leave | -0.116 |
Bad | -0.116 |
Stop | -0.112 |
vious | -0.111 |
Token | , |
Token ID | 11 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
pron | +0.298 |
masterpiece | +0.276 |
favourite | +0.270 |
collection | +0.247 |
ail | +0.246 |
Neg logit contributions | |
crow | -0.225 |
Bad | -0.215 |
Leave | -0.214 |
Stop | -0.202 |
vious | -0.201 |
Token | with |
Token ID | 351 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
crow | +0.107 |
Leave | +0.106 |
Bad | +0.105 |
Stop | +0.100 |
vious | +0.098 |
Neg logit contributions | |
pron | -0.101 |
masterpiece | -0.092 |
favourite | -0.090 |
collection | -0.084 |
ail | -0.083 |
Token | his |
Token ID | 465 |
Feature activation | +2.99e+00 |
Pos logit contributions | |
crow | +0.114 |
Leave | +0.112 |
Bad | +0.110 |
Stop | +0.106 |
vious | +0.105 |
Neg logit contributions | |
pron | -0.135 |
masterpiece | -0.123 |
favourite | -0.117 |
ail | -0.115 |
collection | -0.111 |
Token | clothes |
Token ID | 8242 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
crow | +0.238 |
Bad | +0.237 |
Leave | +0.231 |
Stop | +0.219 |
hunters | +0.219 |
Neg logit contributions | |
pron | -0.204 |
masterpiece | -0.180 |
onna | -0.179 |
oph | -0.172 |
yt | -0.163 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.65e+01 |
Pos logit contributions | |
pron | +0.220 |
masterpiece | +0.206 |
favourite | +0.197 |
onna | +0.181 |
oph | +0.178 |
Neg logit contributions | |
crow | -0.137 |
Leave | -0.134 |
vious | -0.129 |
Bad | -0.125 |
ting | -0.124 |
Token | cleaned |
Token ID | 20750 |
Feature activation | -2.83e+00 |
Pos logit contributions | |
pron | +1.185 |
masterpiece | +1.165 |
onna | +1.071 |
favourite | +1.056 |
yt | +1.041 |
Neg logit contributions | |
crow | -1.272 |
Leave | -1.218 |
Stop | -1.207 |
Bad | -1.206 |
Hey | -1.177 |
Token | . |
Token ID | 13 |
Feature activation | +2.68e-01 |
Pos logit contributions | |
pron | +0.245 |
masterpiece | +0.221 |
favourite | +0.217 |
collection | +0.199 |
usual | +0.196 |
Neg logit contributions | |
crow | -0.202 |
Leave | -0.189 |
Bad | -0.189 |
Stop | -0.182 |
vious | -0.179 |
Token | Suddenly |
Token ID | 24975 |
Feature activation | +2.71e-01 |
Pos logit contributions | |
crow | +0.018 |
Leave | +0.017 |
Bad | +0.017 |
Stop | +0.016 |
vious | +0.016 |
Neg logit contributions | |
pron | -0.025 |
favourite | -0.023 |
masterpiece | -0.023 |
collection | -0.021 |
usual | -0.021 |
Token | , |
Token ID | 11 |
Feature activation | +1.35e+00 |
Pos logit contributions | |
crow | +0.024 |
Leave | +0.023 |
Bad | +0.023 |
vious | +0.022 |
Stop | +0.022 |
Neg logit contributions | |
pron | -0.019 |
favourite | -0.017 |
masterpiece | -0.017 |
collection | -0.015 |
usual | -0.015 |
Token | he |
Token ID | 339 |
Feature activation | +7.47e-01 |
Pos logit contributions | |
crow | +0.104 |
Leave | +0.100 |
Bad | +0.099 |
Stop | +0.096 |
vious | +0.094 |
Neg logit contributions | |
pron | -0.109 |
favourite | -0.099 |
masterpiece | -0.099 |
collection | -0.090 |
brand | -0.088 |
Token |
|
Token ID | 198 |
Feature activation | -2.60e-01 |
Pos logit contributions | |
pron | +0.274 |
masterpiece | +0.253 |
favourite | +0.253 |
collection | +0.235 |
usual | +0.230 |
Neg logit contributions | |
crow | -0.188 |
Leave | -0.177 |
Bad | -0.175 |
Stop | -0.168 |
vious | -0.163 |
Token | Andy |
Token ID | 35314 |
Feature activation | +2.53e+00 |
Pos logit contributions | |
pron | +0.025 |
masterpiece | +0.023 |
favourite | +0.022 |
ail | +0.021 |
brand | +0.021 |
Neg logit contributions | |
crow | -0.015 |
Bad | -0.015 |
vious | -0.015 |
Leave | -0.015 |
ases | -0.014 |
Token | looked |
Token ID | 3114 |
Feature activation | -4.59e+00 |
Pos logit contributions | |
crow | +0.202 |
Bad | +0.196 |
Leave | +0.194 |
Stop | +0.188 |
vious | +0.188 |
Neg logit contributions | |
pron | -0.191 |
favourite | -0.170 |
masterpiece | -0.163 |
usual | -0.158 |
collection | -0.152 |
Token | at |
Token ID | 379 |
Feature activation | +3.84e-01 |
Pos logit contributions | |
pron | +0.426 |
masterpiece | +0.374 |
favourite | +0.366 |
usual | +0.359 |
ail | +0.354 |
Neg logit contributions | |
crow | -0.290 |
vious | -0.285 |
Leave | -0.282 |
Bad | -0.282 |
Stop | -0.270 |
Token | the |
Token ID | 262 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
crow | +0.028 |
Leave | +0.027 |
Bad | +0.027 |
Stop | +0.026 |
vious | +0.025 |
Neg logit contributions | |
pron | -0.033 |
favourite | -0.030 |
masterpiece | -0.030 |
collection | -0.028 |
usual | -0.027 |
Token | freshly |
Token ID | 29026 |
Feature activation | -1.65e+01 |
Pos logit contributions | |
crow | +0.139 |
Bad | +0.137 |
Stop | +0.134 |
Leave | +0.133 |
vious | +0.131 |
Neg logit contributions | |
pron | -0.091 |
favourite | -0.084 |
masterpiece | -0.079 |
collection | -0.077 |
ail | -0.074 |
Token | printed |
Token ID | 10398 |
Feature activation | -3.71e+00 |
Pos logit contributions | |
pron | +0.901 |
favourite | +0.790 |
masterpiece | +0.773 |
collection | +0.678 |
usual | +0.667 |
Neg logit contributions | |
crow | -1.685 |
Bad | -1.636 |
Leave | -1.633 |
Stop | -1.595 |
vious | -1.571 |
Token | picture |
Token ID | 4286 |
Feature activation | -5.48e+00 |
Pos logit contributions | |
pron | +0.250 |
favourite | +0.232 |
masterpiece | +0.220 |
usual | +0.205 |
collection | +0.202 |
Neg logit contributions | |
crow | -0.327 |
Bad | -0.304 |
Leave | -0.304 |
Stop | -0.296 |
vious | -0.294 |
Token | and |
Token ID | 290 |
Feature activation | +1.61e-01 |
Pos logit contributions | |
pron | +0.496 |
favourite | +0.456 |
masterpiece | +0.446 |
collection | +0.413 |
usual | +0.413 |
Neg logit contributions | |
crow | -0.359 |
Leave | -0.327 |
vious | -0.321 |
Bad | -0.321 |
Stop | -0.313 |
Token | smiled |
Token ID | 13541 |
Feature activation | -9.51e-01 |
Pos logit contributions | |
crow | +0.012 |
Bad | +0.012 |
Leave | +0.012 |
Stop | +0.012 |
Hey | +0.011 |
Neg logit contributions | |
pron | -0.013 |
favourite | -0.011 |
masterpiece | -0.011 |
collection | -0.010 |
usual | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | +2.80e-01 |
Pos logit contributions | |
pron | +0.085 |
masterpiece | +0.076 |
favourite | +0.075 |
collection | +0.070 |
usual | +0.069 |
Neg logit contributions | |
crow | -0.065 |
Leave | -0.062 |
Bad | -0.062 |
vious | -0.059 |
Stop | -0.058 |
Token | mom |
Token ID | 1995 |
Feature activation | -5.35e+00 |
Pos logit contributions | |
crow | +0.264 |
eper | +0.257 |
Bad | +0.255 |
bully | +0.250 |
Stop | +0.249 |
Neg logit contributions | |
pron | -0.243 |
onna | -0.234 |
masterpiece | -0.233 |
icol | -0.216 |
yt | -0.215 |
Token | put |
Token ID | 1234 |
Feature activation | -5.58e+00 |
Pos logit contributions | |
favourite | +0.431 |
masterpiece | +0.429 |
pron | +0.428 |
usual | +0.398 |
brand | +0.389 |
Neg logit contributions | |
crow | -0.359 |
Stop | -0.346 |
vious | -0.346 |
Bad | -0.345 |
Hey | -0.340 |
Token | on |
Token ID | 319 |
Feature activation | -1.49e+00 |
Pos logit contributions | |
pron | +0.463 |
masterpiece | +0.433 |
favourite | +0.426 |
collection | +0.394 |
brand | +0.385 |
Neg logit contributions | |
crow | -0.416 |
Bad | -0.395 |
Leave | -0.391 |
Stop | -0.380 |
vious | -0.372 |
Token | her |
Token ID | 607 |
Feature activation | -2.79e+00 |
Pos logit contributions | |
pron | +0.131 |
masterpiece | +0.121 |
favourite | +0.120 |
collection | +0.116 |
ail | +0.116 |
Neg logit contributions | |
crow | -0.093 |
Bad | -0.092 |
Stop | -0.090 |
Leave | -0.087 |
Hey | -0.083 |
Token | a |
Token ID | 257 |
Feature activation | +9.40e-02 |
Pos logit contributions | |
yt | +0.161 |
masterpiece | +0.161 |
onna | +0.160 |
pron | +0.160 |
oph | +0.154 |
Neg logit contributions | |
crow | -0.220 |
vious | -0.215 |
Bad | -0.212 |
Leave | -0.208 |
Stop | -0.198 |
Token | pron |
Token ID | 31186 |
Feature activation | -1.45e+01 |
Pos logit contributions | |
crow | +0.008 |
Hey | +0.007 |
Stop | +0.007 |
Bad | +0.007 |
mean | +0.007 |
Neg logit contributions | |
ert | -0.004 |
icol | -0.004 |
ail | -0.004 |
yt | -0.004 |
onna | -0.004 |
Token | and |
Token ID | 290 |
Feature activation | -8.30e-02 |
Pos logit contributions | |
masterpiece | +1.311 |
pron | +1.248 |
favourite | +1.241 |
brand | +1.160 |
usual | +1.152 |
Neg logit contributions | |
crow | -0.911 |
vious | -0.753 |
Bad | -0.748 |
ting | -0.737 |
eper | -0.737 |
Token | said |
Token ID | 531 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
masterpiece | +0.007 |
pron | +0.006 |
ail | +0.006 |
onna | +0.006 |
favourite | +0.006 |
Neg logit contributions | |
Bad | -0.005 |
Stop | -0.005 |
Hey | -0.005 |
crow | -0.005 |
Leave | -0.005 |
Token | , |
Token ID | 11 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
pron | +0.112 |
masterpiece | +0.111 |
favourite | +0.105 |
usual | +0.098 |
collection | +0.098 |
Neg logit contributions | |
crow | -0.077 |
Bad | -0.071 |
Leave | -0.071 |
Stop | -0.070 |
Stop | -0.066 |
Token | " |
Token ID | 366 |
Feature activation | -1.89e-01 |
Pos logit contributions | |
Leave | +0.068 |
Bad | +0.067 |
crow | +0.067 |
Stop | +0.065 |
ting | +0.064 |
Neg logit contributions | |
masterpiece | -0.095 |
pron | -0.095 |
favourite | -0.091 |
collection | -0.089 |
ail | -0.084 |
Token | Let |
Token ID | 5756 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
pron | +0.018 |
favourite | +0.015 |
masterpiece | +0.015 |
usual | +0.014 |
favorite | +0.014 |
Neg logit contributions | |
crow | -0.013 |
Bad | -0.012 |
ases | -0.012 |
Leave | -0.012 |
eper | -0.012 |
Token | Dad |
Token ID | 17415 |
Feature activation | -5.66e+00 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.003 |
Bad | +0.003 |
Stop | +0.003 |
vious | +0.003 |
Neg logit contributions | |
pron | -0.003 |
masterpiece | -0.003 |
favourite | -0.003 |
collection | -0.003 |
ail | -0.003 |
Token | put |
Token ID | 1234 |
Feature activation | -5.57e+00 |
Pos logit contributions | |
pron | +0.426 |
masterpiece | +0.416 |
favourite | +0.407 |
usual | +0.389 |
collection | +0.366 |
Neg logit contributions | |
eper | -0.374 |
crow | -0.374 |
vious | -0.362 |
Bad | -0.357 |
Hey | -0.356 |
Token | on |
Token ID | 319 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
pron | +0.490 |
masterpiece | +0.459 |
favourite | +0.445 |
collection | +0.419 |
ail | +0.412 |
Neg logit contributions | |
crow | -0.380 |
Bad | -0.358 |
Leave | -0.350 |
Stop | -0.345 |
vious | -0.337 |
Token | his |
Token ID | 465 |
Feature activation | +2.77e+00 |
Pos logit contributions | |
pron | +0.141 |
masterpiece | +0.128 |
favourite | +0.123 |
ail | +0.123 |
collection | +0.120 |
Neg logit contributions | |
crow | -0.083 |
Bad | -0.082 |
Leave | -0.079 |
Stop | -0.078 |
Hey | -0.075 |
Token | a |
Token ID | 257 |
Feature activation | +9.76e-02 |
Pos logit contributions | |
Leave | +0.225 |
Bad | +0.222 |
crow | +0.221 |
vious | +0.216 |
Stop | +0.209 |
Neg logit contributions | |
pron | -0.161 |
masterpiece | -0.154 |
onna | -0.150 |
yt | -0.146 |
ail | -0.146 |
Token | pron |
Token ID | 31186 |
Feature activation | -1.45e+01 |
Pos logit contributions | |
Bad | +0.007 |
Leave | +0.007 |
Hey | +0.007 |
crow | +0.007 |
Stop | +0.007 |
Neg logit contributions | |
ail | -0.004 |
oph | -0.004 |
icol | -0.004 |
onna | -0.004 |
yt | -0.004 |
Token | and |
Token ID | 290 |
Feature activation | -8.94e-02 |
Pos logit contributions | |
pron | +1.266 |
masterpiece | +1.223 |
favourite | +1.192 |
usual | +1.106 |
brand | +1.095 |
Neg logit contributions | |
crow | -0.916 |
vious | -0.784 |
eper | -0.761 |
ases | -0.759 |
Leave | -0.753 |
Token | seemed |
Token ID | 3947 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
pron | +0.007 |
masterpiece | +0.007 |
ail | +0.006 |
collection | +0.006 |
favourite | +0.006 |
Neg logit contributions | |
Bad | -0.006 |
crow | -0.006 |
Hey | -0.006 |
Stop | -0.006 |
Leave | -0.005 |
Token | really |
Token ID | 1107 |
Feature activation | -2.76e+00 |
Pos logit contributions | |
pron | +0.099 |
masterpiece | +0.090 |
ail | +0.087 |
favourite | +0.084 |
collection | +0.083 |
Neg logit contributions | |
crow | -0.092 |
vious | -0.091 |
Leave | -0.090 |
Bad | -0.089 |
Stop | -0.085 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
pron | +0.195 |
masterpiece | +0.190 |
favourite | +0.182 |
porch | +0.176 |
ail | +0.175 |
Neg logit contributions | |
vious | -0.200 |
crow | -0.199 |
Bad | -0.199 |
Leave | -0.194 |
Stop | -0.188 |
Token | . |
Token ID | 13 |
Feature activation | +4.38e-02 |
Pos logit contributions | |
crow | +0.113 |
Leave | +0.107 |
Bad | +0.106 |
Stop | +0.104 |
vious | +0.102 |
Neg logit contributions | |
pron | -0.160 |
masterpiece | -0.143 |
favourite | -0.142 |
ail | -0.134 |
collection | -0.130 |
Token | and |
Token ID | 290 |
Feature activation | -8.40e-02 |
Pos logit contributions | |
pron | +0.151 |
masterpiece | +0.140 |
favourite | +0.138 |
ail | +0.129 |
usual | +0.128 |
Neg logit contributions | |
crow | -0.092 |
vious | -0.081 |
Stop | -0.080 |
eper | -0.079 |
Bad | -0.079 |
Token | put |
Token ID | 1234 |
Feature activation | -5.56e+00 |
Pos logit contributions | |
pron | +0.006 |
masterpiece | +0.005 |
favourite | +0.005 |
ail | +0.005 |
brand | +0.005 |
Neg logit contributions | |
crow | -0.006 |
Bad | -0.006 |
Stop | -0.006 |
Leave | -0.006 |
eper | -0.006 |
Token | on |
Token ID | 319 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
pron | +0.512 |
masterpiece | +0.474 |
favourite | +0.469 |
collection | +0.437 |
usual | +0.430 |
Neg logit contributions | |
crow | -0.359 |
Bad | -0.333 |
Leave | -0.327 |
Stop | -0.324 |
vious | -0.313 |
Token | an |
Token ID | 281 |
Feature activation | -2.69e+00 |
Pos logit contributions | |
pron | +0.136 |
masterpiece | +0.122 |
favourite | +0.120 |
yt | +0.116 |
ail | +0.115 |
Neg logit contributions | |
crow | -0.085 |
Bad | -0.084 |
Leave | -0.080 |
Stop | -0.080 |
ases | -0.077 |
Token | a |
Token ID | 257 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
pron | +0.185 |
masterpiece | +0.162 |
ail | +0.162 |
favourite | +0.157 |
yt | +0.156 |
Neg logit contributions | |
crow | -0.218 |
vious | -0.213 |
Leave | -0.203 |
eper | -0.202 |
Bad | -0.202 |
Token | pron |
Token ID | 31186 |
Feature activation | -1.45e+01 |
Pos logit contributions | |
crow | +0.010 |
Bad | +0.010 |
Stop | +0.010 |
Hey | +0.010 |
Leave | +0.009 |
Neg logit contributions | |
ail | -0.004 |
masterpiece | -0.004 |
usual | -0.004 |
ilda | -0.004 |
liv | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | +4.12e-02 |
Pos logit contributions | |
pron | +1.194 |
masterpiece | +1.147 |
favourite | +1.126 |
brand | +1.065 |
usual | +1.031 |
Neg logit contributions | |
crow | -1.029 |
vious | -0.875 |
eper | -0.861 |
ases | -0.854 |
Leave | -0.840 |
Token | His |
Token ID | 2399 |
Feature activation | -4.70e-02 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.002 |
Bad | +0.002 |
Stop | +0.002 |
vious | +0.002 |
Neg logit contributions | |
pron | -0.004 |
favourite | -0.004 |
masterpiece | -0.004 |
collection | -0.003 |
usual | -0.003 |
Token | siblings |
Token ID | 20569 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
pron | +0.003 |
yt | +0.002 |
onna | +0.002 |
masterpiece | +0.002 |
oph | +0.002 |
Neg logit contributions | |
Bad | -0.004 |
crow | -0.004 |
eper | -0.004 |
hunters | -0.004 |
Leave | -0.004 |
Token | watched |
Token ID | 7342 |
Feature activation | +3.65e-01 |
Pos logit contributions | |
crow | +0.151 |
Leave | +0.148 |
vious | +0.144 |
Stop | +0.142 |
Bad | +0.141 |
Neg logit contributions | |
pron | -0.141 |
favourite | -0.129 |
masterpiece | -0.128 |
brand | -0.116 |
usual | -0.115 |
Token | him |
Token ID | 683 |
Feature activation | +7.48e+00 |
Pos logit contributions | |
crow | +0.025 |
Leave | +0.024 |
Bad | +0.024 |
Stop | +0.023 |
vious | +0.023 |
Neg logit contributions | |
pron | -0.033 |
favourite | -0.030 |
masterpiece | -0.030 |
collection | -0.028 |
brand | -0.027 |
Token | put |
Token ID | 1234 |
Feature activation | -5.57e+00 |
Pos logit contributions | |
crow | +0.278 |
Bad | +0.275 |
Leave | +0.267 |
Stop | +0.260 |
Hey | +0.257 |
Neg logit contributions | |
pron | -0.210 |
masterpiece | -0.200 |
favourite | -0.192 |
collection | -0.172 |
ail | -0.171 |
Token | on |
Token ID | 319 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
pron | +0.491 |
favourite | +0.452 |
masterpiece | +0.445 |
collection | +0.413 |
brand | +0.406 |
Neg logit contributions | |
crow | -0.393 |
Leave | -0.380 |
Bad | -0.375 |
Stop | -0.359 |
vious | -0.356 |
Token | his |
Token ID | 465 |
Feature activation | +2.77e+00 |
Pos logit contributions | |
pron | +0.157 |
masterpiece | +0.146 |
favourite | +0.145 |
ail | +0.136 |
collection | +0.135 |
Neg logit contributions | |
crow | -0.072 |
Leave | -0.068 |
Bad | -0.067 |
Stop | -0.064 |
vious | -0.063 |
Token | special |
Token ID | 2041 |
Feature activation | -6.23e+00 |
Pos logit contributions | |
crow | +0.201 |
Leave | +0.200 |
Bad | +0.195 |
hunters | +0.194 |
brake | +0.191 |
Neg logit contributions | |
pron | -0.183 |
oph | -0.181 |
onna | -0.177 |
masterpiece | -0.167 |
yt | -0.161 |
Token | a |
Token ID | 257 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
pron | +0.423 |
masterpiece | +0.380 |
oph | +0.379 |
onna | +0.373 |
favourite | +0.371 |
Neg logit contributions | |
crow | -0.492 |
Leave | -0.455 |
Bad | -0.448 |
hunters | -0.429 |
Stop | -0.427 |
Token | pron |
Token ID | 31186 |
Feature activation | -1.45e+01 |
Pos logit contributions | |
crow | +0.008 |
Bad | +0.008 |
bully | +0.007 |
Leave | +0.007 |
brake | +0.007 |
Neg logit contributions | |
masterpiece | -0.006 |
favourite | -0.006 |
ail | -0.006 |
oph | -0.006 |
liv | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | -7.80e-02 |
Pos logit contributions | |
masterpiece | +1.222 |
pron | +1.209 |
favourite | +1.123 |
brand | +1.073 |
ail | +1.045 |
Neg logit contributions | |
crow | -0.997 |
eper | -0.840 |
Bad | -0.839 |
hunters | -0.838 |
Leave | -0.833 |
Token | decided |
Token ID | 3066 |
Feature activation | +3.93e+00 |
Pos logit contributions | |
masterpiece | +0.006 |
pron | +0.006 |
ail | +0.006 |
onna | +0.006 |
favourite | +0.006 |
Neg logit contributions | |
Bad | -0.005 |
Leave | -0.005 |
crow | -0.005 |
Stop | -0.005 |
Hey | -0.005 |
Token | to |
Token ID | 284 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
Leave | +0.305 |
crow | +0.302 |
Bad | +0.298 |
Stop | +0.293 |
hunters | +0.278 |
Neg logit contributions | |
pron | -0.302 |
masterpiece | -0.277 |
favourite | -0.262 |
ail | -0.247 |
collection | -0.243 |
Token | travel |
Token ID | 3067 |
Feature activation | -2.11e+00 |
Pos logit contributions | |
pron | +0.123 |
favourite | +0.116 |
masterpiece | +0.114 |
brand | +0.104 |
ail | +0.103 |
Neg logit contributions | |
crow | -0.131 |
Leave | -0.126 |
Bad | -0.123 |
Stop | -0.121 |
vious | -0.118 |
Token | . |
Token ID | 13 |
Feature activation | +4.05e-02 |
Pos logit contributions | |
pron | +0.131 |
favourite | +0.116 |
masterpiece | +0.115 |
collection | +0.102 |
usual | +0.100 |
Neg logit contributions | |
crow | -0.203 |
Leave | -0.197 |
Bad | -0.196 |
Stop | -0.190 |
vious | -0.188 |
Token | , |
Token ID | 11 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
pron | +0.130 |
favourite | +0.114 |
masterpiece | +0.113 |
brand | +0.104 |
favorite | +0.101 |
Neg logit contributions | |
crow | -0.130 |
Leave | -0.118 |
Bad | -0.115 |
Stop | -0.114 |
vious | -0.113 |
Token | the |
Token ID | 262 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
crow | +0.082 |
Bad | +0.078 |
Leave | +0.077 |
Stop | +0.074 |
vious | +0.071 |
Neg logit contributions | |
pron | -0.095 |
masterpiece | -0.086 |
favourite | -0.083 |
brand | -0.077 |
collection | -0.077 |
Token | baby |
Token ID | 5156 |
Feature activation | -8.00e-01 |
Pos logit contributions | |
crow | +0.098 |
Bad | +0.095 |
Leave | +0.092 |
vious | +0.089 |
Stop | +0.088 |
Neg logit contributions | |
pron | -0.103 |
masterpiece | -0.089 |
favourite | -0.088 |
brand | -0.084 |
ail | -0.083 |
Token | wasn |
Token ID | 2492 |
Feature activation | -6.67e-01 |
Pos logit contributions | |
pron | +0.063 |
favourite | +0.057 |
masterpiece | +0.057 |
collection | +0.052 |
brand | +0.051 |
Neg logit contributions | |
crow | -0.064 |
Leave | -0.061 |
Bad | -0.061 |
Stop | -0.058 |
vious | -0.058 |
Token | 't |
Token ID | 470 |
Feature activation | +8.58e-01 |
Pos logit contributions | |
pron | +0.054 |
favourite | +0.049 |
masterpiece | +0.049 |
collection | +0.045 |
brand | +0.044 |
Neg logit contributions | |
crow | -0.051 |
Leave | -0.050 |
Bad | -0.049 |
Stop | -0.047 |
vious | -0.047 |
Token | scared |
Token ID | 12008 |
Feature activation | +8.22e+00 |
Pos logit contributions | |
crow | +0.061 |
Leave | +0.058 |
Bad | +0.058 |
Stop | +0.056 |
vious | +0.054 |
Neg logit contributions | |
pron | -0.075 |
favourite | -0.069 |
masterpiece | -0.068 |
collection | -0.063 |
usual | -0.062 |
Token | anymore |
Token ID | 7471 |
Feature activation | +3.11e+00 |
Pos logit contributions | |
Bad | +0.626 |
crow | +0.614 |
Leave | +0.602 |
Stop | +0.569 |
eper | +0.564 |
Neg logit contributions | |
pron | -0.610 |
favourite | -0.603 |
masterpiece | -0.592 |
usual | -0.553 |
favorite | -0.543 |
Token | . |
Token ID | 13 |
Feature activation | +4.38e-02 |
Pos logit contributions | |
crow | +0.233 |
Leave | +0.223 |
Bad | +0.222 |
Stop | +0.212 |
vious | +0.211 |
Neg logit contributions | |
pron | -0.259 |
favourite | -0.233 |
masterpiece | -0.233 |
collection | -0.212 |
brand | -0.209 |
Token | She |
Token ID | 1375 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.003 |
Bad | +0.003 |
Stop | +0.003 |
vious | +0.003 |
Neg logit contributions | |
pron | -0.004 |
masterpiece | -0.004 |
favourite | -0.003 |
collection | -0.003 |
usual | -0.003 |
Token | now |
Token ID | 783 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
pron | +0.095 |
masterpiece | +0.082 |
favourite | +0.082 |
usual | +0.076 |
brand | +0.076 |
Neg logit contributions | |
crow | -0.092 |
Bad | -0.087 |
Leave | -0.086 |
vious | -0.085 |
Stop | -0.084 |
Token | knew |
Token ID | 2993 |
Feature activation | +2.03e-02 |
Pos logit contributions | |
pron | +0.154 |
favourite | +0.139 |
masterpiece | +0.136 |
brand | +0.124 |
usual | +0.123 |
Neg logit contributions | |
crow | -0.163 |
Leave | -0.154 |
Bad | -0.152 |
vious | -0.150 |
Stop | -0.147 |
Token | , |
Token ID | 11 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
pron | +0.242 |
masterpiece | +0.216 |
favourite | +0.214 |
collection | +0.192 |
brand | +0.191 |
Neg logit contributions | |
crow | -0.213 |
Bad | -0.197 |
Leave | -0.196 |
Stop | -0.191 |
eper | -0.188 |
Token | he |
Token ID | 339 |
Feature activation | +4.96e-01 |
Pos logit contributions | |
crow | +0.084 |
Bad | +0.081 |
Leave | +0.080 |
Stop | +0.077 |
vious | +0.075 |
Neg logit contributions | |
pron | -0.092 |
masterpiece | -0.084 |
favourite | -0.083 |
collection | -0.076 |
ail | -0.074 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.70e-01 |
Pos logit contributions | |
crow | +0.039 |
Bad | +0.038 |
Leave | +0.038 |
Stop | +0.036 |
eper | +0.036 |
Neg logit contributions | |
pron | -0.039 |
masterpiece | -0.034 |
favourite | -0.034 |
usual | -0.030 |
collection | -0.030 |
Token | two |
Token ID | 734 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
crow | +0.031 |
Leave | +0.030 |
Bad | +0.029 |
Stop | +0.028 |
vious | +0.028 |
Neg logit contributions | |
pron | -0.027 |
favourite | -0.025 |
masterpiece | -0.025 |
brand | -0.023 |
collection | -0.023 |
Token | other |
Token ID | 584 |
Feature activation | +5.31e+00 |
Pos logit contributions | |
crow | +0.116 |
Leave | +0.111 |
Bad | +0.111 |
Stop | +0.107 |
vious | +0.105 |
Neg logit contributions | |
pron | -0.120 |
favourite | -0.111 |
masterpiece | -0.110 |
collection | -0.101 |
usual | -0.099 |
Token | mice |
Token ID | 10693 |
Feature activation | +9.24e+00 |
Pos logit contributions | |
crow | +0.275 |
Leave | +0.261 |
Bad | +0.258 |
Stop | +0.243 |
vious | +0.237 |
Neg logit contributions | |
pron | -0.559 |
favourite | -0.530 |
masterpiece | -0.525 |
collection | -0.493 |
usual | -0.486 |
Token | there |
Token ID | 612 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
crow | +0.802 |
Bad | +0.774 |
Leave | +0.773 |
Stop | +0.750 |
vious | +0.737 |
Neg logit contributions | |
pron | -0.639 |
favourite | -0.589 |
masterpiece | -0.588 |
collection | -0.538 |
brand | -0.518 |
Token | . |
Token ID | 13 |
Feature activation | +2.76e-02 |
Pos logit contributions | |
crow | +0.154 |
Leave | +0.147 |
Bad | +0.147 |
Stop | +0.141 |
vious | +0.139 |
Neg logit contributions | |
pron | -0.162 |
favourite | -0.147 |
masterpiece | -0.147 |
collection | -0.133 |
brand | -0.130 |
Token | " |
Token ID | 366 |
Feature activation | -1.96e-01 |
Pos logit contributions | |
crow | +0.002 |
Leave | +0.002 |
Bad | +0.002 |
Stop | +0.002 |
vious | +0.002 |
Neg logit contributions | |
pron | -0.002 |
favourite | -0.002 |
masterpiece | -0.002 |
collection | -0.002 |
brand | -0.002 |
Token | These |
Token ID | 4711 |
Feature activation | +2.76e+00 |
Pos logit contributions | |
pron | +0.016 |
favourite | +0.014 |
masterpiece | +0.014 |
oph | +0.013 |
Sus | +0.013 |
Neg logit contributions | |
crow | -0.013 |
ases | -0.012 |
vious | -0.012 |
risks | -0.012 |
hunters | -0.012 |
Token | are |
Token ID | 389 |
Feature activation | +3.89e+00 |
Pos logit contributions | |
crow | +0.179 |
Leave | +0.175 |
Bad | +0.169 |
Stop | +0.159 |
vious | +0.156 |
Neg logit contributions | |
favourite | -0.238 |
pron | -0.231 |
masterpiece | -0.224 |
collection | -0.220 |
favorite | -0.217 |
Token | 's |
Token ID | 338 |
Feature activation | +3.72e+00 |
Pos logit contributions | |
pron | +0.015 |
masterpiece | +0.014 |
favourite | +0.014 |
brand | +0.012 |
collection | +0.012 |
Neg logit contributions | |
crow | -0.015 |
Leave | -0.015 |
Bad | -0.015 |
vious | -0.014 |
Stop | -0.014 |
Token | mom |
Token ID | 1995 |
Feature activation | -5.31e+00 |
Pos logit contributions | |
crow | +0.334 |
Bad | +0.330 |
Leave | +0.325 |
hunters | +0.319 |
eper | +0.313 |
Neg logit contributions | |
pron | -0.221 |
masterpiece | -0.183 |
onna | -0.178 |
yt | -0.163 |
favourite | -0.160 |
Token | asked |
Token ID | 1965 |
Feature activation | +3.72e-01 |
Pos logit contributions | |
pron | +0.408 |
masterpiece | +0.393 |
favourite | +0.388 |
brand | +0.346 |
usual | +0.343 |
Neg logit contributions | |
crow | -0.401 |
eper | -0.376 |
Bad | -0.376 |
Leave | -0.375 |
vious | -0.371 |
Token | , |
Token ID | 11 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
crow | +0.017 |
Leave | +0.017 |
Bad | +0.017 |
vious | +0.016 |
Stop | +0.015 |
Neg logit contributions | |
pron | -0.041 |
favourite | -0.039 |
masterpiece | -0.038 |
usual | -0.036 |
collection | -0.036 |
Token | " |
Token ID | 366 |
Feature activation | -1.83e-01 |
Pos logit contributions | |
crow | +0.097 |
Bad | +0.094 |
Leave | +0.094 |
Stop | +0.090 |
vious | +0.088 |
Neg logit contributions | |
pron | -0.081 |
masterpiece | -0.075 |
favourite | -0.074 |
collection | -0.067 |
usual | -0.065 |
Token | Did |
Token ID | 11633 |
Feature activation | +8.42e+00 |
Pos logit contributions | |
pron | +0.019 |
favourite | +0.017 |
masterpiece | +0.016 |
usual | +0.015 |
brand | +0.015 |
Neg logit contributions | |
crow | -0.011 |
ases | -0.010 |
vious | -0.010 |
Leave | -0.009 |
Bad | -0.009 |
Token | you |
Token ID | 345 |
Feature activation | +3.31e+00 |
Pos logit contributions | |
crow | +0.469 |
Leave | +0.443 |
Bad | +0.438 |
Stop | +0.414 |
vious | +0.411 |
Neg logit contributions | |
pron | -0.857 |
favourite | -0.804 |
masterpiece | -0.796 |
collection | -0.747 |
usual | -0.738 |
Token | let |
Token ID | 1309 |
Feature activation | +3.65e-01 |
Pos logit contributions | |
crow | +0.244 |
Leave | +0.235 |
Bad | +0.233 |
vious | +0.225 |
Stop | +0.223 |
Neg logit contributions | |
pron | -0.276 |
masterpiece | -0.254 |
favourite | -0.253 |
collection | -0.230 |
brand | -0.227 |
Token | Bob |
Token ID | 5811 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
crow | +0.025 |
Leave | +0.025 |
Bad | +0.024 |
vious | +0.023 |
Stop | +0.023 |
Neg logit contributions | |
pron | -0.032 |
favourite | -0.030 |
masterpiece | -0.029 |
collection | -0.027 |
usual | -0.027 |
Token | go |
Token ID | 467 |
Feature activation | -2.19e-01 |
Pos logit contributions | |
crow | +0.083 |
Leave | +0.080 |
Bad | +0.079 |
Stop | +0.076 |
vious | +0.076 |
Neg logit contributions | |
pron | -0.080 |
masterpiece | -0.073 |
favourite | -0.073 |
collection | -0.066 |
brand | -0.065 |
Token | ?" |
Token ID | 1701 |
Feature activation | -6.35e+00 |
Pos logit contributions | |
pron | +0.019 |
masterpiece | +0.017 |
favourite | +0.016 |
collection | +0.015 |
brand | +0.014 |
Neg logit contributions | |
crow | -0.016 |
Leave | -0.015 |
vious | -0.015 |
Bad | -0.015 |
ases | -0.014 |
Token | be |
Token ID | 307 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
pron | +0.122 |
favourite | +0.111 |
masterpiece | +0.111 |
collection | +0.101 |
usual | +0.099 |
Neg logit contributions | |
crow | -0.128 |
Leave | -0.123 |
Bad | -0.122 |
Stop | -0.118 |
vious | -0.117 |
Token | mail |
Token ID | 6920 |
Feature activation | +4.21e+00 |
Pos logit contributions | |
pron | +0.130 |
favourite | +0.129 |
masterpiece | +0.123 |
favorite | +0.121 |
collection | +0.116 |
Neg logit contributions | |
crow | -0.125 |
Leave | -0.124 |
Stop | -0.119 |
Bad | -0.118 |
vious | -0.114 |
Token | men |
Token ID | 3653 |
Feature activation | +6.54e+00 |
Pos logit contributions | |
Leave | +0.338 |
crow | +0.335 |
Bad | +0.330 |
vious | +0.320 |
Stop | +0.319 |
Neg logit contributions | |
pron | -0.320 |
favourite | -0.303 |
masterpiece | -0.293 |
favorite | -0.274 |
collection | -0.271 |
Token | and |
Token ID | 290 |
Feature activation | -3.51e-02 |
Pos logit contributions | |
crow | +0.515 |
Leave | +0.500 |
Bad | +0.496 |
Stop | +0.478 |
vious | +0.469 |
Neg logit contributions | |
pron | -0.515 |
favourite | -0.472 |
masterpiece | -0.464 |
collection | -0.431 |
usual | -0.427 |
Token | deliver |
Token ID | 5203 |
Feature activation | +3.02e+00 |
Pos logit contributions | |
pron | +0.003 |
favourite | +0.003 |
masterpiece | +0.003 |
collection | +0.002 |
ail | +0.002 |
Neg logit contributions | |
crow | -0.003 |
Leave | -0.003 |
Bad | -0.003 |
Stop | -0.002 |
vious | -0.002 |
Token | letters |
Token ID | 7475 |
Feature activation | +7.74e+00 |
Pos logit contributions | |
crow | +0.227 |
Leave | +0.218 |
Bad | +0.217 |
Stop | +0.209 |
vious | +0.206 |
Neg logit contributions | |
pron | -0.248 |
favourite | -0.227 |
masterpiece | -0.226 |
collection | -0.207 |
usual | -0.203 |
Token | to |
Token ID | 284 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
crow | +0.649 |
Bad | +0.609 |
Leave | +0.609 |
vious | +0.588 |
Stop | +0.586 |
Neg logit contributions | |
pron | -0.572 |
masterpiece | -0.514 |
favourite | -0.504 |
collection | -0.454 |
brand | -0.449 |
Token | their |
Token ID | 511 |
Feature activation | -8.65e-01 |
Pos logit contributions | |
pron | +0.131 |
favourite | +0.119 |
masterpiece | +0.119 |
collection | +0.108 |
brand | +0.106 |
Neg logit contributions | |
crow | -0.124 |
Leave | -0.118 |
Bad | -0.118 |
Stop | -0.114 |
vious | -0.112 |
Token | mom |
Token ID | 1995 |
Feature activation | -5.31e+00 |
Pos logit contributions | |
pron | +0.059 |
masterpiece | +0.050 |
favourite | +0.049 |
brand | +0.044 |
ail | +0.044 |
Neg logit contributions | |
crow | -0.078 |
Bad | -0.075 |
Leave | -0.073 |
Stop | -0.072 |
eper | -0.071 |
Token | and |
Token ID | 290 |
Feature activation | -5.15e-02 |
Pos logit contributions | |
pron | +0.446 |
masterpiece | +0.408 |
favourite | +0.408 |
collection | +0.371 |
brand | +0.366 |
Neg logit contributions | |
crow | -0.392 |
Leave | -0.372 |
Bad | -0.371 |
Stop | -0.356 |
vious | -0.355 |
Token | dad |
Token ID | 9955 |
Feature activation | -5.57e+00 |
Pos logit contributions | |
pron | +0.003 |
masterpiece | +0.002 |
favourite | +0.002 |
ail | +0.002 |
brand | +0.002 |
Neg logit contributions | |
crow | -0.005 |
Bad | -0.005 |
Leave | -0.005 |
Stop | -0.005 |
eper | -0.005 |
Token | he |
Token ID | 339 |
Feature activation | +5.33e-01 |
Pos logit contributions | |
pron | +0.130 |
favourite | +0.119 |
masterpiece | +0.111 |
usual | +0.110 |
favorite | +0.105 |
Neg logit contributions | |
crow | -0.144 |
Leave | -0.139 |
vious | -0.136 |
ases | -0.135 |
Bad | -0.134 |
Token | did |
Token ID | 750 |
Feature activation | +2.44e+00 |
Pos logit contributions | |
crow | +0.043 |
Bad | +0.043 |
Leave | +0.041 |
Stop | +0.040 |
eper | +0.040 |
Neg logit contributions | |
pron | -0.040 |
masterpiece | -0.038 |
favourite | -0.037 |
brand | -0.034 |
collection | -0.033 |
Token | not |
Token ID | 407 |
Feature activation | -3.02e+00 |
Pos logit contributions | |
crow | +0.272 |
Leave | +0.258 |
Bad | +0.256 |
Stop | +0.252 |
vious | +0.252 |
Neg logit contributions | |
pron | -0.119 |
favourite | -0.103 |
masterpiece | -0.101 |
collection | -0.086 |
usual | -0.086 |
Token | recognize |
Token ID | 7564 |
Feature activation | -2.48e-01 |
Pos logit contributions | |
pron | +0.210 |
masterpiece | +0.192 |
favourite | +0.186 |
collection | +0.184 |
usual | +0.178 |
Neg logit contributions | |
crow | -0.254 |
Bad | -0.254 |
Leave | -0.248 |
Stop | -0.236 |
Hey | -0.227 |
Token | the |
Token ID | 262 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
pron | +0.017 |
favourite | +0.017 |
masterpiece | +0.016 |
usual | +0.016 |
favorite | +0.015 |
Neg logit contributions | |
crow | -0.020 |
Leave | -0.020 |
vious | -0.019 |
Bad | -0.019 |
Stop | -0.019 |
Token | danger |
Token ID | 3514 |
Feature activation | +9.00e+00 |
Pos logit contributions | |
crow | +0.100 |
Leave | +0.098 |
Bad | +0.094 |
Stop | +0.092 |
vious | +0.091 |
Neg logit contributions | |
pron | -0.109 |
favourite | -0.102 |
masterpiece | -0.100 |
favorite | -0.093 |
collection | -0.093 |
Token | . |
Token ID | 13 |
Feature activation | +8.90e-02 |
Pos logit contributions | |
Bad | +0.805 |
Leave | +0.799 |
crow | +0.772 |
Stop | +0.769 |
vious | +0.767 |
Neg logit contributions | |
pron | -0.534 |
favourite | -0.491 |
usual | -0.479 |
masterpiece | -0.474 |
collection | -0.452 |
Token | There |
Token ID | 1318 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
crow | +0.007 |
vious | +0.007 |
Leave | +0.007 |
Bad | +0.007 |
Stop | +0.006 |
Neg logit contributions | |
pron | -0.007 |
favourite | -0.006 |
favorite | -0.006 |
masterpiece | -0.006 |
brand | -0.006 |
Token | was |
Token ID | 373 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
crow | +0.101 |
Leave | +0.099 |
Bad | +0.098 |
Stop | +0.095 |
vious | +0.095 |
Neg logit contributions | |
pron | -0.070 |
favourite | -0.064 |
masterpiece | -0.063 |
collection | -0.057 |
usual | -0.056 |
Token | a |
Token ID | 257 |
Feature activation | +1.49e-01 |
Pos logit contributions | |
pron | +0.117 |
favourite | +0.108 |
masterpiece | +0.106 |
usual | +0.097 |
collection | +0.096 |
Neg logit contributions | |
crow | -0.123 |
Leave | -0.118 |
Bad | -0.116 |
Stop | -0.113 |
vious | -0.112 |
Token | big |
Token ID | 1263 |
Feature activation | -5.86e+00 |
Pos logit contributions | |
crow | +0.013 |
Bad | +0.012 |
Leave | +0.012 |
Stop | +0.012 |
vious | +0.011 |
Neg logit contributions | |
pron | -0.011 |
favourite | -0.010 |
masterpiece | -0.010 |
collection | -0.009 |
ail | -0.009 |
Token | of |
Token ID | 286 |
Feature activation | +2.53e+00 |
Pos logit contributions | |
pron | +0.357 |
favourite | +0.330 |
masterpiece | +0.324 |
brand | +0.297 |
collection | +0.297 |
Neg logit contributions | |
crow | -0.300 |
Bad | -0.296 |
Leave | -0.294 |
Stop | -0.276 |
eper | -0.269 |
Token | hot |
Token ID | 3024 |
Feature activation | -2.41e+00 |
Pos logit contributions | |
crow | +0.228 |
Leave | +0.220 |
Bad | +0.220 |
Stop | +0.213 |
vious | +0.211 |
Neg logit contributions | |
pron | -0.172 |
favourite | -0.153 |
masterpiece | -0.153 |
collection | -0.137 |
usual | -0.134 |
Token | pepper |
Token ID | 13385 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
pron | +0.162 |
favourite | +0.146 |
masterpiece | +0.144 |
collection | +0.128 |
brand | +0.125 |
Neg logit contributions | |
crow | -0.218 |
Bad | -0.213 |
Leave | -0.212 |
Stop | -0.201 |
vious | -0.201 |
Token | to |
Token ID | 284 |
Feature activation | -1.50e+00 |
Pos logit contributions | |
pron | +0.080 |
masterpiece | +0.072 |
favourite | +0.072 |
brand | +0.064 |
collection | +0.064 |
Neg logit contributions | |
crow | -0.087 |
Bad | -0.082 |
Leave | -0.082 |
Stop | -0.080 |
vious | -0.079 |
Token | his |
Token ID | 465 |
Feature activation | +2.91e+00 |
Pos logit contributions | |
pron | +0.124 |
favourite | +0.115 |
masterpiece | +0.114 |
collection | +0.105 |
ail | +0.103 |
Neg logit contributions | |
crow | -0.110 |
Bad | -0.105 |
Leave | -0.105 |
Stop | -0.101 |
vious | -0.098 |
Token | food |
Token ID | 2057 |
Feature activation | +5.17e+00 |
Pos logit contributions | |
crow | +0.232 |
Bad | +0.229 |
Leave | +0.223 |
Stop | +0.213 |
ases | +0.211 |
Neg logit contributions | |
pron | -0.213 |
masterpiece | -0.190 |
favourite | -0.183 |
ail | -0.174 |
brand | -0.169 |
Token | , |
Token ID | 11 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
crow | +0.352 |
Bad | +0.317 |
Leave | +0.309 |
ases | +0.306 |
vious | +0.304 |
Neg logit contributions | |
pron | -0.445 |
masterpiece | -0.414 |
favourite | -0.409 |
brand | -0.370 |
ail | -0.366 |
Token | and |
Token ID | 290 |
Feature activation | +7.44e-02 |
Pos logit contributions | |
crow | +0.091 |
Leave | +0.090 |
Bad | +0.090 |
Stop | +0.085 |
vious | +0.084 |
Neg logit contributions | |
pron | -0.102 |
masterpiece | -0.095 |
favourite | -0.095 |
ail | -0.088 |
collection | -0.087 |
Token | it |
Token ID | 340 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
crow | +0.006 |
Bad | +0.006 |
Leave | +0.005 |
Stop | +0.005 |
vious | +0.005 |
Neg logit contributions | |
pron | -0.006 |
favourite | -0.005 |
masterpiece | -0.005 |
ail | -0.005 |
brand | -0.005 |
Token | became |
Token ID | 2627 |
Feature activation | -2.61e-01 |
Pos logit contributions | |
crow | +0.010 |
Bad | +0.010 |
Stop | +0.010 |
vious | +0.010 |
Leave | +0.009 |
Neg logit contributions | |
pron | -0.011 |
favourite | -0.010 |
masterpiece | -0.010 |
complexion | -0.009 |
brand | -0.009 |
Token | very |
Token ID | 845 |
Feature activation | +8.88e-01 |
Pos logit contributions | |
pron | +0.019 |
masterpiece | +0.018 |
favourite | +0.017 |
ail | +0.016 |
collection | +0.016 |
Neg logit contributions | |
crow | -0.021 |
Bad | -0.021 |
Leave | -0.020 |
Stop | -0.020 |
vious | -0.019 |
Token | jump |
Token ID | 4391 |
Feature activation | +7.04e-01 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.003 |
Bad | +0.003 |
Stop | +0.003 |
vious | +0.002 |
Neg logit contributions | |
pron | -0.003 |
masterpiece | -0.003 |
favourite | -0.003 |
brand | -0.002 |
collection | -0.002 |
Token | and |
Token ID | 290 |
Feature activation | +5.77e-02 |
Pos logit contributions | |
crow | +0.056 |
Leave | +0.054 |
Bad | +0.054 |
Stop | +0.051 |
vious | +0.051 |
Neg logit contributions | |
pron | -0.055 |
favourite | -0.051 |
masterpiece | -0.050 |
collection | -0.046 |
usual | -0.045 |
Token | slide |
Token ID | 10649 |
Feature activation | -3.23e+00 |
Pos logit contributions | |
crow | +0.005 |
Leave | +0.005 |
Bad | +0.005 |
Stop | +0.004 |
vious | +0.004 |
Neg logit contributions | |
pron | -0.004 |
favourite | -0.004 |
masterpiece | -0.004 |
collection | -0.003 |
brand | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | +1.89e-01 |
Pos logit contributions | |
pron | +0.265 |
favourite | +0.241 |
masterpiece | +0.241 |
collection | +0.219 |
brand | +0.215 |
Neg logit contributions | |
crow | -0.247 |
Leave | -0.237 |
Bad | -0.235 |
Stop | -0.226 |
vious | -0.223 |
Token | They |
Token ID | 1119 |
Feature activation | -2.22e-02 |
Pos logit contributions | |
crow | +0.013 |
Leave | +0.012 |
Bad | +0.012 |
Stop | +0.012 |
Hey | +0.011 |
Neg logit contributions | |
pron | -0.018 |
masterpiece | -0.016 |
favourite | -0.015 |
collection | -0.015 |
brand | -0.015 |
Token | saw |
Token ID | 2497 |
Feature activation | +5.13e-01 |
Pos logit contributions | |
pron | +0.002 |
masterpiece | +0.002 |
favourite | +0.002 |
brand | +0.002 |
collection | +0.001 |
Neg logit contributions | |
crow | -0.002 |
Leave | -0.002 |
Bad | -0.002 |
Stop | -0.002 |
Hey | -0.002 |
Token | a |
Token ID | 257 |
Feature activation | +2.41e-01 |
Pos logit contributions | |
crow | +0.044 |
Leave | +0.043 |
Bad | +0.042 |
Stop | +0.041 |
vious | +0.040 |
Neg logit contributions | |
pron | -0.034 |
favourite | -0.034 |
masterpiece | -0.031 |
favorite | -0.030 |
usual | -0.029 |
Token | big |
Token ID | 1263 |
Feature activation | -5.75e+00 |
Pos logit contributions | |
Bad | +0.021 |
crow | +0.021 |
Stop | +0.020 |
Leave | +0.020 |
Hey | +0.019 |
Neg logit contributions | |
pron | -0.015 |
favourite | -0.014 |
masterpiece | -0.014 |
ail | -0.013 |
collection | -0.012 |
Token | hill |
Token ID | 12788 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
pron | +0.426 |
favourite | +0.387 |
masterpiece | +0.384 |
collection | +0.345 |
usual | +0.342 |
Neg logit contributions | |
crow | -0.480 |
Leave | -0.461 |
Bad | -0.461 |
Stop | -0.444 |
vious | -0.437 |
Token | with |
Token ID | 351 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
pron | +0.093 |
favourite | +0.085 |
masterpiece | +0.085 |
collection | +0.078 |
usual | +0.076 |
Neg logit contributions | |
crow | -0.084 |
Leave | -0.081 |
Bad | -0.080 |
Stop | -0.077 |
vious | -0.076 |
Token | a |
Token ID | 257 |
Feature activation | +2.52e-01 |
Pos logit contributions | |
crow | +0.151 |
Leave | +0.146 |
Bad | +0.145 |
Stop | +0.141 |
vious | +0.140 |
Neg logit contributions | |
pron | -0.091 |
favourite | -0.085 |
masterpiece | -0.082 |
collection | -0.073 |
usual | -0.072 |
Token | and |
Token ID | 290 |
Feature activation | -8.56e-02 |
Pos logit contributions | |
crow | +0.669 |
Bad | +0.628 |
Leave | +0.628 |
vious | +0.609 |
Stop | +0.604 |
Neg logit contributions | |
pron | -0.452 |
favourite | -0.425 |
masterpiece | -0.404 |
favorite | -0.389 |
brand | -0.370 |
Token | cl |
Token ID | 537 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
pron | +0.007 |
masterpiece | +0.007 |
favourite | +0.007 |
collection | +0.006 |
usual | +0.006 |
Neg logit contributions | |
Leave | -0.006 |
Bad | -0.006 |
crow | -0.006 |
Stop | -0.006 |
vious | -0.006 |
Token | apped |
Token ID | 6320 |
Feature activation | +5.85e-01 |
Pos logit contributions | |
crow | +0.122 |
Leave | +0.121 |
Bad | +0.120 |
Stop | +0.118 |
hunters | +0.115 |
Neg logit contributions | |
pron | -0.074 |
favourite | -0.069 |
masterpiece | -0.065 |
collection | -0.060 |
ouse | -0.058 |
Token | . |
Token ID | 13 |
Feature activation | +3.40e-02 |
Pos logit contributions | |
crow | +0.052 |
Bad | +0.050 |
Leave | +0.050 |
Stop | +0.048 |
vious | +0.047 |
Neg logit contributions | |
pron | -0.041 |
favourite | -0.038 |
masterpiece | -0.036 |
favorite | -0.034 |
collection | -0.033 |
Token | " |
Token ID | 366 |
Feature activation | -1.95e-01 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.003 |
Bad | +0.002 |
Stop | +0.002 |
vious | +0.002 |
Neg logit contributions | |
pron | -0.003 |
favourite | -0.003 |
masterpiece | -0.003 |
collection | -0.002 |
usual | -0.002 |
Token | You |
Token ID | 1639 |
Feature activation | +6.63e+00 |
Pos logit contributions | |
pron | +0.018 |
favourite | +0.017 |
masterpiece | +0.016 |
favorite | +0.015 |
usual | +0.015 |
Neg logit contributions | |
crow | -0.012 |
eper | -0.011 |
thieves | -0.010 |
vious | -0.010 |
hunters | -0.010 |
Token | 're |
Token ID | 821 |
Feature activation | +7.59e-02 |
Pos logit contributions | |
crow | +0.483 |
Leave | +0.463 |
Bad | +0.459 |
Stop | +0.440 |
vious | +0.434 |
Neg logit contributions | |
pron | -0.566 |
favourite | -0.521 |
masterpiece | -0.518 |
collection | -0.476 |
brand | -0.467 |
Token | awesome |
Token ID | 7427 |
Feature activation | -2.91e+00 |
Pos logit contributions | |
crow | +0.006 |
Leave | +0.006 |
Bad | +0.005 |
Stop | +0.005 |
vious | +0.005 |
Neg logit contributions | |
pron | -0.006 |
favourite | -0.006 |
masterpiece | -0.006 |
brand | -0.005 |
favorite | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | +3.16e-02 |
Pos logit contributions | |
pron | +0.202 |
favourite | +0.185 |
masterpiece | +0.182 |
collection | +0.168 |
favorite | +0.162 |
Neg logit contributions | |
crow | -0.255 |
Leave | -0.250 |
Bad | -0.243 |
Stop | -0.235 |
vious | -0.234 |
Token | I |
Token ID | 314 |
Feature activation | -2.52e+00 |
Pos logit contributions | |
crow | +0.002 |
Leave | +0.002 |
Bad | +0.002 |
vious | +0.002 |
Stop | +0.002 |
Neg logit contributions | |
pron | -0.003 |
favourite | -0.003 |
masterpiece | -0.002 |
favorite | -0.002 |
bedroom | -0.002 |
Token | 'm |
Token ID | 1101 |
Feature activation | -4.81e-01 |
Pos logit contributions | |
pron | +0.204 |
favourite | +0.186 |
masterpiece | +0.186 |
collection | +0.169 |
brand | +0.166 |
Neg logit contributions | |
crow | -0.194 |
Leave | -0.186 |
Bad | -0.186 |
Stop | -0.179 |
vious | -0.178 |
Token | away |
Token ID | 1497 |
Feature activation | +7.52e+00 |
Pos logit contributions | |
pron | +0.058 |
favourite | +0.058 |
masterpiece | +0.055 |
favorite | +0.051 |
usual | +0.051 |
Neg logit contributions | |
crow | -0.062 |
Bad | -0.062 |
Leave | -0.061 |
vious | -0.059 |
Stop | -0.058 |
Token | . |
Token ID | 13 |
Feature activation | +1.96e-01 |
Pos logit contributions | |
crow | +0.540 |
Leave | +0.511 |
Bad | +0.509 |
Stop | +0.479 |
vious | +0.475 |
Neg logit contributions | |
pron | -0.674 |
masterpiece | -0.598 |
favourite | -0.595 |
collection | -0.542 |
brand | -0.537 |
Token | They |
Token ID | 1119 |
Feature activation | -7.24e-03 |
Pos logit contributions | |
crow | +0.011 |
Leave | +0.010 |
Stop | +0.010 |
Bad | +0.010 |
vious | +0.010 |
Neg logit contributions | |
pron | -0.021 |
masterpiece | -0.019 |
favourite | -0.018 |
collection | -0.017 |
usual | -0.017 |
Token | were |
Token ID | 547 |
Feature activation | +2.71e-02 |
Pos logit contributions | |
pron | +0.001 |
masterpiece | +0.001 |
favourite | +0.001 |
brand | +0.000 |
collection | +0.000 |
Neg logit contributions | |
crow | -0.001 |
Bad | -0.001 |
Leave | -0.001 |
Stop | -0.001 |
vious | -0.001 |
Token | so |
Token ID | 523 |
Feature activation | -6.90e-01 |
Pos logit contributions | |
crow | +0.002 |
Leave | +0.002 |
Bad | +0.002 |
Stop | +0.002 |
vious | +0.002 |
Neg logit contributions | |
pron | -0.002 |
favourite | -0.002 |
masterpiece | -0.002 |
collection | -0.002 |
brand | -0.002 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.92e+00 |
Pos logit contributions | |
pron | +0.049 |
favourite | +0.045 |
masterpiece | +0.044 |
collection | +0.040 |
brand | +0.040 |
Neg logit contributions | |
crow | -0.060 |
Leave | -0.058 |
Bad | -0.057 |
Stop | -0.056 |
vious | -0.055 |
Token | and |
Token ID | 290 |
Feature activation | +8.74e-02 |
Pos logit contributions | |
crow | +0.152 |
Leave | +0.147 |
Bad | +0.147 |
Stop | +0.141 |
vious | +0.138 |
Neg logit contributions | |
pron | -0.157 |
masterpiece | -0.137 |
favourite | -0.135 |
collection | -0.123 |
brand | -0.121 |
Token | proud |
Token ID | 6613 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
crow | +0.006 |
Bad | +0.006 |
Leave | +0.006 |
vious | +0.006 |
Stop | +0.006 |
Neg logit contributions | |
pron | -0.008 |
masterpiece | -0.007 |
favourite | -0.007 |
collection | -0.006 |
ail | -0.006 |
Token | of |
Token ID | 286 |
Feature activation | +2.58e+00 |
Pos logit contributions | |
pron | +0.148 |
masterpiece | +0.136 |
favourite | +0.136 |
collection | +0.125 |
usual | +0.123 |
Neg logit contributions | |
crow | -0.107 |
Leave | -0.102 |
Bad | -0.101 |
Stop | -0.096 |
vious | -0.096 |
Token | themselves |
Token ID | 2405 |
Feature activation | +3.37e+00 |
Pos logit contributions | |
crow | +0.206 |
Leave | +0.198 |
Bad | +0.196 |
vious | +0.190 |
Stop | +0.190 |
Neg logit contributions | |
pron | -0.201 |
favourite | -0.186 |
masterpiece | -0.185 |
collection | -0.168 |
usual | -0.166 |
Token | for |
Token ID | 329 |
Feature activation | -5.24e+00 |
Pos logit contributions | |
crow | +0.262 |
Leave | +0.252 |
Bad | +0.251 |
Stop | +0.240 |
vious | +0.239 |
Neg logit contributions | |
pron | -0.270 |
favourite | -0.244 |
masterpiece | -0.244 |
collection | -0.221 |
usual | -0.218 |
Token | boy |
Token ID | 2933 |
Feature activation | +4.85e+00 |
Pos logit contributions | |
pron | +0.214 |
favourite | +0.207 |
masterpiece | +0.198 |
collection | +0.194 |
favorite | +0.193 |
Neg logit contributions | |
crow | -0.179 |
Leave | -0.178 |
Bad | -0.171 |
ases | -0.167 |
Stop | -0.164 |
Token | was |
Token ID | 373 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
crow | +0.406 |
Leave | +0.390 |
Bad | +0.388 |
Stop | +0.376 |
vious | +0.374 |
Neg logit contributions | |
pron | -0.363 |
masterpiece | -0.323 |
favourite | -0.322 |
collection | -0.289 |
brand | -0.288 |
Token | very |
Token ID | 845 |
Feature activation | +7.48e-01 |
Pos logit contributions | |
pron | +0.135 |
masterpiece | +0.123 |
favourite | +0.123 |
collection | +0.113 |
usual | +0.111 |
Neg logit contributions | |
crow | -0.114 |
Leave | -0.109 |
Bad | -0.108 |
Stop | -0.103 |
vious | -0.103 |
Token | surprised |
Token ID | 6655 |
Feature activation | +2.80e+00 |
Pos logit contributions | |
crow | +0.053 |
Leave | +0.051 |
Bad | +0.050 |
Stop | +0.048 |
vious | +0.047 |
Neg logit contributions | |
pron | -0.066 |
favourite | -0.061 |
masterpiece | -0.060 |
collection | -0.055 |
usual | -0.055 |
Token | . |
Token ID | 13 |
Feature activation | +4.78e-02 |
Pos logit contributions | |
crow | +0.246 |
Leave | +0.243 |
Bad | +0.242 |
Stop | +0.228 |
vious | +0.224 |
Neg logit contributions | |
pron | -0.188 |
favourite | -0.180 |
masterpiece | -0.173 |
favorite | -0.160 |
usual | -0.160 |
Token | He |
Token ID | 679 |
Feature activation | +3.19e+00 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.002 |
Bad | +0.002 |
Stop | +0.002 |
vious | +0.002 |
Neg logit contributions | |
pron | -0.005 |
favourite | -0.005 |
masterpiece | -0.005 |
collection | -0.004 |
brand | -0.004 |
Token | had |
Token ID | 550 |
Feature activation | -5.24e+00 |
Pos logit contributions | |
crow | +0.235 |
Bad | +0.232 |
Leave | +0.230 |
Stop | +0.230 |
vious | +0.220 |
Neg logit contributions | |
pron | -0.262 |
masterpiece | -0.235 |
favourite | -0.232 |
collection | -0.216 |
ail | -0.212 |
Token | never |
Token ID | 1239 |
Feature activation | -7.07e-01 |
Pos logit contributions | |
pron | +0.447 |
favourite | +0.392 |
masterpiece | +0.390 |
ail | +0.362 |
collection | +0.358 |
Neg logit contributions | |
crow | -0.382 |
Bad | -0.378 |
Leave | -0.377 |
Stop | -0.366 |
vious | -0.365 |
Token | seen |
Token ID | 1775 |
Feature activation | +3.14e+00 |
Pos logit contributions | |
pron | +0.068 |
favourite | +0.063 |
masterpiece | +0.062 |
collection | +0.058 |
ail | +0.057 |
Neg logit contributions | |
crow | -0.044 |
Leave | -0.042 |
Bad | -0.041 |
vious | -0.041 |
Stop | -0.040 |
Token | a |
Token ID | 257 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
crow | +0.229 |
Leave | +0.221 |
Bad | +0.219 |
Stop | +0.210 |
vious | +0.207 |
Neg logit contributions | |
pron | -0.265 |
favourite | -0.245 |
masterpiece | -0.243 |
collection | -0.222 |
usual | -0.220 |
Token | fake |
Token ID | 8390 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
crow | +0.010 |
Bad | +0.010 |
Stop | +0.010 |
Leave | +0.010 |
eper | +0.009 |
Neg logit contributions | |
pron | -0.009 |
favourite | -0.008 |
masterpiece | -0.008 |
ail | -0.008 |
collection | -0.007 |
Token | kne |
Token ID | 24813 |
Feature activation | +7.42e-01 |
Pos logit contributions | |
pron | +0.397 |
favourite | +0.370 |
masterpiece | +0.361 |
collection | +0.327 |
usual | +0.327 |
Neg logit contributions | |
crow | -0.446 |
Leave | -0.426 |
Bad | -0.424 |
Stop | -0.411 |
vious | -0.409 |
Token | eled |
Token ID | 18449 |
Feature activation | +2.08e-01 |
Pos logit contributions | |
Bad | +0.058 |
crow | +0.057 |
Leave | +0.056 |
hunters | +0.055 |
Stop | +0.055 |
Neg logit contributions | |
pron | -0.052 |
masterpiece | -0.045 |
ail | -0.043 |
favourite | -0.042 |
complexion | -0.040 |
Token | down |
Token ID | 866 |
Feature activation | +3.73e-01 |
Pos logit contributions | |
crow | +0.014 |
Leave | +0.014 |
Bad | +0.014 |
Stop | +0.013 |
vious | +0.013 |
Neg logit contributions | |
pron | -0.019 |
favourite | -0.017 |
masterpiece | -0.017 |
collection | -0.016 |
usual | -0.015 |
Token | and |
Token ID | 290 |
Feature activation | -8.90e-02 |
Pos logit contributions | |
crow | +0.025 |
Leave | +0.024 |
Bad | +0.024 |
Stop | +0.023 |
vious | +0.023 |
Neg logit contributions | |
pron | -0.033 |
favourite | -0.031 |
masterpiece | -0.031 |
collection | -0.029 |
usual | -0.028 |
Token | told |
Token ID | 1297 |
Feature activation | -2.63e+00 |
Pos logit contributions | |
pron | +0.007 |
favourite | +0.006 |
masterpiece | +0.006 |
collection | +0.006 |
usual | +0.006 |
Neg logit contributions | |
Bad | -0.007 |
crow | -0.007 |
Leave | -0.007 |
Stop | -0.007 |
Hey | -0.007 |
Token | L |
Token ID | 406 |
Feature activation | -7.99e-01 |
Pos logit contributions | |
pron | +0.201 |
favourite | +0.177 |
masterpiece | +0.176 |
collection | +0.162 |
brand | +0.156 |
Neg logit contributions | |
crow | -0.221 |
Leave | -0.210 |
Bad | -0.208 |
Stop | -0.207 |
vious | -0.202 |
Token | ola |
Token ID | 5708 |
Feature activation | -5.23e+00 |
Pos logit contributions | |
pron | +0.051 |
favourite | +0.046 |
masterpiece | +0.045 |
collection | +0.041 |
brand | +0.040 |
Neg logit contributions | |
crow | -0.076 |
Leave | -0.072 |
vious | -0.071 |
Bad | -0.071 |
Stop | -0.071 |
Token | to |
Token ID | 284 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
pron | +0.414 |
masterpiece | +0.372 |
favourite | +0.366 |
collection | +0.330 |
brand | +0.328 |
Neg logit contributions | |
crow | -0.421 |
Bad | -0.389 |
Leave | -0.388 |
vious | -0.382 |
Stop | -0.375 |
Token | be |
Token ID | 307 |
Feature activation | -1.73e+00 |
Pos logit contributions | |
pron | +0.135 |
favourite | +0.125 |
masterpiece | +0.124 |
collection | +0.112 |
brand | +0.112 |
Neg logit contributions | |
crow | -0.125 |
Leave | -0.120 |
Bad | -0.119 |
Stop | -0.116 |
vious | -0.112 |
Token | careful |
Token ID | 8161 |
Feature activation | +6.74e+00 |
Pos logit contributions | |
pron | +0.142 |
favourite | +0.133 |
masterpiece | +0.129 |
brand | +0.119 |
favorite | +0.118 |
Neg logit contributions | |
crow | -0.128 |
Bad | -0.122 |
Leave | -0.121 |
Stop | -0.119 |
vious | -0.114 |
Token | . |
Token ID | 13 |
Feature activation | +3.63e-02 |
Pos logit contributions | |
crow | +0.515 |
Leave | +0.495 |
Bad | +0.492 |
Stop | +0.473 |
vious | +0.468 |
Neg logit contributions | |
pron | -0.548 |
favourite | -0.501 |
masterpiece | -0.499 |
collection | -0.455 |
usual | -0.447 |
Token | find |
Token ID | 1064 |
Feature activation | -2.54e+00 |
Pos logit contributions | |
pron | +0.210 |
favourite | +0.190 |
masterpiece | +0.189 |
collection | +0.171 |
usual | +0.168 |
Neg logit contributions | |
crow | -0.255 |
Bad | -0.247 |
Leave | -0.246 |
Stop | -0.238 |
vious | -0.234 |
Token | Lily |
Token ID | 20037 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
pron | +0.175 |
favourite | +0.170 |
masterpiece | +0.159 |
favorite | +0.152 |
usual | +0.151 |
Neg logit contributions | |
crow | -0.212 |
Leave | -0.208 |
Bad | -0.204 |
Stop | -0.199 |
vious | -0.199 |
Token | , |
Token ID | 11 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
pron | +0.150 |
masterpiece | +0.135 |
favourite | +0.133 |
collection | +0.120 |
brand | +0.119 |
Neg logit contributions | |
crow | -0.138 |
Leave | -0.132 |
Bad | -0.131 |
vious | -0.126 |
Stop | -0.124 |
Token | so |
Token ID | 523 |
Feature activation | -7.39e-01 |
Pos logit contributions | |
crow | +0.097 |
Leave | +0.094 |
Bad | +0.093 |
Stop | +0.089 |
vious | +0.088 |
Neg logit contributions | |
pron | -0.097 |
masterpiece | -0.089 |
favourite | -0.088 |
collection | -0.081 |
brand | -0.078 |
Token | he |
Token ID | 339 |
Feature activation | +6.28e-01 |
Pos logit contributions | |
pron | +0.064 |
masterpiece | +0.061 |
favourite | +0.058 |
collection | +0.054 |
ail | +0.054 |
Neg logit contributions | |
Bad | -0.050 |
crow | -0.050 |
Leave | -0.050 |
Stop | -0.046 |
vious | -0.044 |
Token | went |
Token ID | 1816 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
crow | +0.050 |
Leave | +0.047 |
Bad | +0.047 |
Stop | +0.045 |
eper | +0.045 |
Neg logit contributions | |
pron | -0.050 |
masterpiece | -0.047 |
favourite | -0.045 |
brand | -0.041 |
collection | -0.041 |
Token | to |
Token ID | 284 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
pron | +0.076 |
masterpiece | +0.067 |
favourite | +0.067 |
collection | +0.059 |
brand | +0.058 |
Neg logit contributions | |
crow | -0.100 |
Leave | -0.096 |
Bad | -0.095 |
Stop | -0.092 |
vious | -0.091 |
Token | get |
Token ID | 651 |
Feature activation | +4.14e-01 |
Pos logit contributions | |
pron | +0.127 |
masterpiece | +0.114 |
favourite | +0.114 |
collection | +0.104 |
brand | +0.102 |
Neg logit contributions | |
crow | -0.115 |
Leave | -0.111 |
Bad | -0.110 |
Stop | -0.106 |
vious | -0.102 |
Token | help |
Token ID | 1037 |
Feature activation | +2.44e+00 |
Pos logit contributions | |
crow | +0.031 |
Leave | +0.029 |
Bad | +0.029 |
Stop | +0.028 |
vious | +0.027 |
Neg logit contributions | |
pron | -0.035 |
masterpiece | -0.032 |
favourite | -0.032 |
collection | -0.029 |
brand | -0.029 |
Token | . |
Token ID | 13 |
Feature activation | +1.81e-01 |
Pos logit contributions | |
Bad | +0.207 |
crow | +0.205 |
Leave | +0.204 |
Stop | +0.200 |
vious | +0.197 |
Neg logit contributions | |
pron | -0.170 |
favourite | -0.156 |
masterpiece | -0.155 |
collection | -0.143 |
usual | -0.142 |
Token | He |
Token ID | 679 |
Feature activation | +3.30e+00 |
Pos logit contributions | |
crow | +0.012 |
Leave | +0.011 |
Bad | +0.011 |
Stop | +0.010 |
Hey | +0.010 |
Neg logit contributions | |
pron | -0.017 |
masterpiece | -0.016 |
favourite | -0.015 |
collection | -0.015 |
usual | -0.014 |
Token | so |
Token ID | 523 |
Feature activation | -7.42e-01 |
Pos logit contributions | |
pron | +0.144 |
masterpiece | +0.127 |
ail | +0.125 |
favourite | +0.124 |
collection | +0.124 |
Neg logit contributions | |
vious | -0.080 |
crow | -0.079 |
Leave | -0.077 |
Bad | -0.077 |
eper | -0.075 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
pron | +0.070 |
masterpiece | +0.066 |
ail | +0.062 |
icol | +0.061 |
favourite | +0.061 |
Neg logit contributions | |
vious | -0.038 |
Bad | -0.037 |
Leave | -0.037 |
crow | -0.036 |
ases | -0.035 |
Token | and |
Token ID | 290 |
Feature activation | +2.73e-02 |
Pos logit contributions | |
crow | +0.130 |
Leave | +0.124 |
Bad | +0.122 |
Stop | +0.120 |
vious | +0.118 |
Neg logit contributions | |
pron | -0.168 |
masterpiece | -0.150 |
favourite | -0.148 |
collection | -0.136 |
brand | -0.134 |
Token | excited |
Token ID | 6568 |
Feature activation | +2.81e+00 |
Pos logit contributions | |
Leave | +0.002 |
Bad | +0.002 |
crow | +0.002 |
ting | +0.002 |
vious | +0.001 |
Neg logit contributions | |
pron | -0.002 |
masterpiece | -0.002 |
favourite | -0.002 |
ool | -0.002 |
ail | -0.002 |
Token | ! |
Token ID | 0 |
Feature activation | -3.00e+00 |
Pos logit contributions | |
crow | +0.196 |
Stop | +0.184 |
vious | +0.183 |
Leave | +0.180 |
Bad | +0.177 |
Neg logit contributions | |
pron | -0.255 |
masterpiece | -0.224 |
favourite | -0.218 |
collection | -0.205 |
brand | -0.200 |
Token | |
Token ID | 220 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
pron | +0.308 |
masterpiece | +0.284 |
favourite | +0.259 |
collection | +0.251 |
brand | +0.250 |
Neg logit contributions | |
crow | -0.192 |
Leave | -0.164 |
Stop | -0.163 |
Bad | -0.157 |
vious | -0.156 |
Token |
|
Token ID | 198 |
Feature activation | -3.48e-01 |
Pos logit contributions | |
pron | +0.111 |
masterpiece | +0.103 |
favourite | +0.100 |
collection | +0.095 |
brand | +0.092 |
Neg logit contributions | |
crow | -0.079 |
Leave | -0.073 |
Bad | -0.073 |
Stop | -0.071 |
ases | -0.069 |
Token |
|
Token ID | 198 |
Feature activation | -3.55e-01 |
Pos logit contributions | |
pron | +0.032 |
masterpiece | +0.030 |
favourite | +0.029 |
collection | +0.028 |
brand | +0.027 |
Neg logit contributions | |
crow | -0.023 |
Leave | -0.022 |
Bad | -0.022 |
Stop | -0.021 |
Hey | -0.020 |
Token | The |
Token ID | 464 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
pron | +0.030 |
favourite | +0.027 |
masterpiece | +0.027 |
usual | +0.024 |
collection | +0.024 |
Neg logit contributions | |
crow | -0.026 |
Bad | -0.025 |
Leave | -0.024 |
vious | -0.024 |
ases | -0.024 |
Token | bee |
Token ID | 20697 |
Feature activation | +8.90e-01 |
Pos logit contributions | |
crow | +0.090 |
Bad | +0.087 |
Leave | +0.086 |
Stop | +0.083 |
vious | +0.083 |
Neg logit contributions | |
pron | -0.107 |
masterpiece | -0.100 |
favourite | -0.100 |
brand | -0.094 |
collection | -0.093 |
Token | cut |
Token ID | 2005 |
Feature activation | -3.62e+00 |
Pos logit contributions | |
crow | +0.070 |
Bad | +0.065 |
Leave | +0.065 |
vious | +0.064 |
eper | +0.063 |
Neg logit contributions | |
pron | -0.069 |
favourite | -0.063 |
masterpiece | -0.060 |
usual | -0.058 |
brand | -0.057 |
Token | sand |
Token ID | 6450 |
Feature activation | -3.05e+00 |
Pos logit contributions | |
crow | +0.101 |
Leave | +0.096 |
Bad | +0.096 |
Stop | +0.094 |
vious | +0.091 |
Neg logit contributions | |
pron | -0.105 |
masterpiece | -0.093 |
favourite | -0.093 |
collection | -0.085 |
ail | -0.084 |
Token | . |
Token ID | 13 |
Feature activation | +7.44e-02 |
Pos logit contributions | |
pron | +0.261 |
masterpiece | +0.236 |
favourite | +0.233 |
brand | +0.215 |
usual | +0.209 |
Neg logit contributions | |
crow | -0.222 |
Leave | -0.206 |
Bad | -0.206 |
vious | -0.201 |
ases | -0.201 |
Token | They |
Token ID | 1119 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
crow | +0.006 |
Leave | +0.005 |
Bad | +0.005 |
Stop | +0.005 |
vious | +0.005 |
Neg logit contributions | |
pron | -0.006 |
masterpiece | -0.006 |
favourite | -0.006 |
collection | -0.005 |
ail | -0.005 |
Token | make |
Token ID | 787 |
Feature activation | -1.70e+00 |
Pos logit contributions | |
pron | +0.012 |
masterpiece | +0.011 |
favourite | +0.011 |
brand | +0.010 |
ail | +0.010 |
Neg logit contributions | |
crow | -0.009 |
Leave | -0.009 |
Bad | -0.009 |
Stop | -0.009 |
vious | -0.008 |
Token | a |
Token ID | 257 |
Feature activation | +1.28e-01 |
Pos logit contributions | |
pron | +0.107 |
favourite | +0.099 |
masterpiece | +0.097 |
collection | +0.086 |
favorite | +0.085 |
Neg logit contributions | |
crow | -0.161 |
Leave | -0.155 |
Bad | -0.154 |
Stop | -0.149 |
vious | -0.147 |
Token | big |
Token ID | 1263 |
Feature activation | -5.87e+00 |
Pos logit contributions | |
Bad | +0.011 |
crow | +0.010 |
Leave | +0.010 |
Stop | +0.010 |
hunters | +0.010 |
Neg logit contributions | |
pron | -0.008 |
favourite | -0.007 |
masterpiece | -0.007 |
ail | -0.007 |
ilda | -0.007 |
Token | pile |
Token ID | 14540 |
Feature activation | -4.39e+00 |
Pos logit contributions | |
pron | +0.392 |
favourite | +0.338 |
masterpiece | +0.328 |
usual | +0.305 |
ail | +0.302 |
Neg logit contributions | |
crow | -0.525 |
Bad | -0.509 |
Leave | -0.503 |
Stop | -0.488 |
vious | -0.485 |
Token | of |
Token ID | 286 |
Feature activation | +2.45e+00 |
Pos logit contributions | |
pron | +0.422 |
favourite | +0.387 |
masterpiece | +0.386 |
collection | +0.354 |
brand | +0.354 |
Neg logit contributions | |
crow | -0.274 |
Leave | -0.256 |
Bad | -0.251 |
vious | -0.243 |
ases | -0.239 |
Token | sand |
Token ID | 6450 |
Feature activation | -3.04e+00 |
Pos logit contributions | |
crow | +0.221 |
Leave | +0.215 |
Bad | +0.213 |
Stop | +0.206 |
vious | +0.204 |
Neg logit contributions | |
pron | -0.164 |
favourite | -0.149 |
masterpiece | -0.148 |
collection | -0.133 |
brand | -0.130 |
Token | and |
Token ID | 290 |
Feature activation | -3.76e-02 |
Pos logit contributions | |
pron | +0.248 |
favourite | +0.214 |
masterpiece | +0.213 |
brand | +0.196 |
usual | +0.193 |
Neg logit contributions | |
crow | -0.231 |
ases | -0.222 |
eper | -0.218 |
hunters | -0.212 |
Leave | -0.211 |
Token | say |
Token ID | 910 |
Feature activation | +2.19e+00 |
Pos logit contributions | |
pron | +0.003 |
masterpiece | +0.003 |
favourite | +0.002 |
ail | +0.002 |
brand | +0.002 |
Neg logit contributions | |
Leave | -0.003 |
Bad | -0.003 |
crow | -0.003 |
Stop | -0.003 |
vious | -0.003 |
Token | box |
Token ID | 3091 |
Feature activation | -7.75e-01 |
Pos logit contributions | |
pron | +0.135 |
favourite | +0.123 |
masterpiece | +0.123 |
collection | +0.111 |
brand | +0.109 |
Neg logit contributions | |
crow | -0.141 |
Leave | -0.136 |
Bad | -0.135 |
Stop | -0.130 |
vious | -0.129 |
Token | opened |
Token ID | 4721 |
Feature activation | -1.92e+00 |
Pos logit contributions | |
pron | +0.062 |
favourite | +0.055 |
masterpiece | +0.053 |
usual | +0.049 |
favorite | +0.049 |
Neg logit contributions | |
crow | -0.062 |
Leave | -0.059 |
Bad | -0.055 |
vious | -0.055 |
Stop | -0.054 |
Token | up |
Token ID | 510 |
Feature activation | -8.90e-01 |
Pos logit contributions | |
pron | +0.176 |
favourite | +0.157 |
masterpiece | +0.155 |
usual | +0.144 |
collection | +0.144 |
Neg logit contributions | |
crow | -0.128 |
Leave | -0.121 |
Bad | -0.119 |
Stop | -0.116 |
vious | -0.115 |
Token | with |
Token ID | 351 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
pron | +0.088 |
favourite | +0.076 |
masterpiece | +0.076 |
collection | +0.071 |
ail | +0.069 |
Neg logit contributions | |
crow | -0.057 |
Leave | -0.053 |
Bad | -0.051 |
Stop | -0.051 |
vious | -0.051 |
Token | candy |
Token ID | 18550 |
Feature activation | -2.07e-01 |
Pos logit contributions | |
crow | +0.107 |
Leave | +0.103 |
Bad | +0.103 |
Stop | +0.100 |
vious | +0.099 |
Neg logit contributions | |
pron | -0.136 |
masterpiece | -0.118 |
favourite | -0.117 |
ail | -0.111 |
collection | -0.110 |
Token | inside |
Token ID | 2641 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
pron | +0.018 |
favourite | +0.017 |
masterpiece | +0.017 |
usual | +0.015 |
favorite | +0.015 |
Neg logit contributions | |
crow | -0.014 |
Leave | -0.013 |
Bad | -0.013 |
Stop | -0.013 |
vious | -0.013 |
Token | ! |
Token ID | 0 |
Feature activation | -2.98e+00 |
Pos logit contributions | |
pron | +0.102 |
favourite | +0.095 |
masterpiece | +0.091 |
favorite | +0.086 |
collection | +0.084 |
Neg logit contributions | |
crow | -0.071 |
Leave | -0.066 |
Bad | -0.065 |
Stop | -0.064 |
vious | -0.063 |
Token |
|
Token ID | 198 |
Feature activation | -3.41e-01 |
Pos logit contributions | |
pron | +0.292 |
favourite | +0.265 |
masterpiece | +0.261 |
collection | +0.251 |
usual | +0.247 |
Neg logit contributions | |
crow | -0.181 |
Stop | -0.172 |
Leave | -0.166 |
Bad | -0.160 |
Hey | -0.158 |
Token |
|
Token ID | 198 |
Feature activation | -3.35e-01 |
Pos logit contributions | |
pron | +0.032 |
favourite | +0.029 |
masterpiece | +0.029 |
collection | +0.028 |
usual | +0.027 |
Neg logit contributions | |
crow | -0.022 |
Leave | -0.021 |
Stop | -0.020 |
Bad | -0.020 |
Hey | -0.019 |
Token | Sarah |
Token ID | 29284 |
Feature activation | -2.42e+00 |
Pos logit contributions | |
pron | +0.026 |
favourite | +0.024 |
masterpiece | +0.024 |
collection | +0.022 |
brand | +0.021 |
Neg logit contributions | |
crow | -0.027 |
Leave | -0.026 |
Bad | -0.025 |
Stop | -0.024 |
vious | -0.024 |
Token | was |
Token ID | 373 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
pron | +0.193 |
favourite | +0.165 |
masterpiece | +0.162 |
usual | +0.159 |
favorite | +0.148 |
Neg logit contributions | |
crow | -0.181 |
Leave | -0.177 |
Bad | -0.177 |
Stop | -0.171 |
eper | -0.170 |
Token | Mom |
Token ID | 11254 |
Feature activation | -5.52e+00 |
Pos logit contributions | |
pron | +0.367 |
masterpiece | +0.344 |
favourite | +0.321 |
usual | +0.304 |
collection | +0.303 |
Neg logit contributions | |
crow | -0.381 |
Bad | -0.366 |
Leave | -0.357 |
Stop | -0.354 |
Hey | -0.347 |
Token | said |
Token ID | 531 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
pron | +0.396 |
favourite | +0.376 |
masterpiece | +0.367 |
usual | +0.345 |
brand | +0.336 |
Neg logit contributions | |
crow | -0.466 |
vious | -0.426 |
Bad | -0.422 |
eper | -0.417 |
Stop | -0.412 |
Token | , |
Token ID | 11 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
pron | +0.113 |
masterpiece | +0.106 |
favourite | +0.103 |
collection | +0.096 |
brand | +0.095 |
Neg logit contributions | |
crow | -0.080 |
Bad | -0.073 |
Leave | -0.073 |
vious | -0.070 |
ases | -0.069 |
Token | " |
Token ID | 366 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
Leave | +0.081 |
crow | +0.081 |
Bad | +0.080 |
Stop | +0.074 |
vious | +0.073 |
Neg logit contributions | |
masterpiece | -0.087 |
pron | -0.087 |
favourite | -0.087 |
collection | -0.082 |
usual | -0.080 |
Token | Your |
Token ID | 7120 |
Feature activation | +4.08e+00 |
Pos logit contributions | |
pron | +0.016 |
favourite | +0.014 |
masterpiece | +0.014 |
usual | +0.013 |
favorite | +0.013 |
Neg logit contributions | |
crow | -0.012 |
Bad | -0.011 |
Leave | -0.011 |
vious | -0.011 |
ases | -0.011 |
Token | favorite |
Token ID | 4004 |
Feature activation | -1.10e+01 |
Pos logit contributions | |
crow | +0.380 |
Leave | +0.379 |
Bad | +0.373 |
vious | +0.368 |
Stop | +0.363 |
Neg logit contributions | |
pron | -0.227 |
masterpiece | -0.211 |
onna | -0.187 |
collection | -0.175 |
favourite | -0.171 |
Token | color |
Token ID | 3124 |
Feature activation | -2.24e+00 |
Pos logit contributions | |
pron | +0.869 |
masterpiece | +0.826 |
favourite | +0.788 |
usual | +0.720 |
brand | +0.716 |
Neg logit contributions | |
crow | -0.863 |
Leave | -0.798 |
Bad | -0.797 |
Stop | -0.775 |
vious | -0.765 |
Token | is |
Token ID | 318 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
pron | +0.227 |
masterpiece | +0.217 |
favourite | +0.206 |
icol | +0.201 |
usual | +0.194 |
Neg logit contributions | |
crow | -0.115 |
vious | -0.098 |
ases | -0.095 |
Bad | -0.092 |
Leave | -0.091 |
Token | blue |
Token ID | 4171 |
Feature activation | -5.03e+00 |
Pos logit contributions | |
crow | +0.068 |
Bad | +0.067 |
hunters | +0.064 |
vious | +0.064 |
Leave | +0.063 |
Neg logit contributions | |
pron | -0.098 |
masterpiece | -0.096 |
ail | -0.092 |
favourite | -0.089 |
complexion | -0.088 |
Token | , |
Token ID | 11 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
pron | +0.449 |
masterpiece | +0.417 |
favourite | +0.412 |
collection | +0.379 |
usual | +0.376 |
Neg logit contributions | |
crow | -0.345 |
Leave | -0.325 |
Bad | -0.322 |
vious | -0.313 |
Stop | -0.311 |
Token | right |
Token ID | 826 |
Feature activation | +2.27e+00 |
Pos logit contributions | |
crow | +0.088 |
Bad | +0.087 |
Leave | +0.087 |
Stop | +0.084 |
vious | +0.082 |
Neg logit contributions | |
pron | -0.083 |
masterpiece | -0.080 |
favourite | -0.076 |
ail | -0.074 |
collection | -0.074 |
Token | a |
Token ID | 257 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
pron | +0.142 |
masterpiece | +0.124 |
ouse | +0.121 |
ail | +0.121 |
collection | +0.118 |
Neg logit contributions | |
Bad | -0.068 |
Leave | -0.066 |
Stop | -0.064 |
hunters | -0.063 |
Hey | -0.062 |
Token | trail |
Token ID | 8025 |
Feature activation | -2.82e+00 |
Pos logit contributions | |
Leave | +0.014 |
crow | +0.014 |
Bad | +0.014 |
hunters | +0.013 |
bully | +0.013 |
Neg logit contributions | |
pron | -0.012 |
masterpiece | -0.011 |
ail | -0.010 |
favourite | -0.010 |
ouse | -0.010 |
Token | ! |
Token ID | 0 |
Feature activation | -3.02e+00 |
Pos logit contributions | |
pron | +0.239 |
masterpiece | +0.224 |
favourite | +0.219 |
brand | +0.198 |
collection | +0.196 |
Neg logit contributions | |
crow | -0.202 |
Leave | -0.187 |
vious | -0.185 |
Bad | -0.176 |
Stop | -0.174 |
Token |
|
Token ID | 198 |
Feature activation | -3.63e-01 |
Pos logit contributions | |
pron | +0.299 |
masterpiece | +0.289 |
collection | +0.270 |
favourite | +0.264 |
usual | +0.258 |
Neg logit contributions | |
crow | -0.162 |
Stop | -0.161 |
Leave | -0.160 |
Hey | -0.149 |
Bad | -0.147 |
Token |
|
Token ID | 198 |
Feature activation | -3.65e-01 |
Pos logit contributions | |
pron | +0.037 |
masterpiece | +0.035 |
favourite | +0.033 |
collection | +0.033 |
brand | +0.031 |
Neg logit contributions | |
crow | -0.020 |
Stop | -0.019 |
Leave | -0.019 |
Bad | -0.018 |
Hey | -0.017 |
Token | Em |
Token ID | 10161 |
Feature activation | -9.22e+00 |
Pos logit contributions | |
pron | +0.029 |
masterpiece | +0.028 |
favourite | +0.027 |
collection | +0.026 |
usual | +0.025 |
Neg logit contributions | |
Leave | -0.027 |
crow | -0.026 |
Bad | -0.026 |
Stop | -0.026 |
Hey | -0.025 |
Token | ma |
Token ID | 2611 |
Feature activation | -6.66e-01 |
Pos logit contributions | |
pron | +0.748 |
favourite | +0.714 |
masterpiece | +0.690 |
usual | +0.639 |
brand | +0.638 |
Neg logit contributions | |
crow | -0.679 |
Bad | -0.658 |
Leave | -0.655 |
Stop | -0.644 |
vious | -0.623 |
Token | got |
Token ID | 1392 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
pron | +0.059 |
masterpiece | +0.053 |
brand | +0.050 |
favourite | +0.050 |
usual | +0.048 |
Neg logit contributions | |
crow | -0.048 |
eper | -0.044 |
vious | -0.042 |
Leave | -0.041 |
Hey | -0.041 |
Token | ready |
Token ID | 3492 |
Feature activation | -6.77e-01 |
Pos logit contributions | |
pron | +0.017 |
masterpiece | +0.016 |
favourite | +0.015 |
collection | +0.014 |
brand | +0.014 |
Neg logit contributions | |
crow | -0.016 |
Leave | -0.016 |
Bad | -0.015 |
vious | -0.015 |
Stop | -0.015 |
Token | to |
Token ID | 284 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
pron | +0.063 |
masterpiece | +0.060 |
favourite | +0.058 |
collection | +0.056 |
oph | +0.054 |
Neg logit contributions | |
ting | -0.034 |
crow | -0.031 |
Leave | -0.031 |
vious | -0.030 |
smart | -0.029 |
Token | go |
Token ID | 467 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
pron | +0.138 |
favourite | +0.125 |
masterpiece | +0.124 |
brand | +0.113 |
favorite | +0.111 |
Neg logit contributions | |
crow | -0.100 |
Leave | -0.098 |
Stop | -0.092 |
Bad | -0.091 |
ases | -0.090 |
Token | many |
Token ID | 867 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
pron | +0.481 |
masterpiece | +0.440 |
ail | +0.410 |
collection | +0.395 |
favourite | +0.391 |
Neg logit contributions | |
crow | -0.321 |
Leave | -0.320 |
Bad | -0.317 |
Stop | -0.308 |
hunters | -0.304 |
Token | games |
Token ID | 1830 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
pron | +0.153 |
favourite | +0.142 |
masterpiece | +0.139 |
collection | +0.127 |
favorite | +0.124 |
Neg logit contributions | |
crow | -0.178 |
Leave | -0.172 |
Bad | -0.171 |
Stop | -0.165 |
vious | -0.162 |
Token | , |
Token ID | 11 |
Feature activation | +1.27e+00 |
Pos logit contributions | |
crow | +0.131 |
Leave | +0.120 |
ases | +0.120 |
Bad | +0.118 |
hunters | +0.115 |
Neg logit contributions | |
pron | -0.179 |
masterpiece | -0.164 |
favourite | -0.157 |
brand | -0.146 |
ail | -0.144 |
Token | but |
Token ID | 475 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
crow | +0.096 |
Leave | +0.095 |
Bad | +0.093 |
Stop | +0.090 |
vious | +0.089 |
Neg logit contributions | |
pron | -0.102 |
masterpiece | -0.094 |
favourite | -0.090 |
collection | -0.085 |
ail | -0.084 |
Token | their |
Token ID | 511 |
Feature activation | -7.33e-01 |
Pos logit contributions | |
pron | +0.118 |
masterpiece | +0.109 |
favourite | +0.097 |
ail | +0.097 |
brand | +0.095 |
Neg logit contributions | |
crow | -0.139 |
Bad | -0.138 |
Leave | -0.135 |
Stop | -0.130 |
vious | -0.127 |
Token | favorite |
Token ID | 4004 |
Feature activation | -1.09e+01 |
Pos logit contributions | |
pron | +0.034 |
masterpiece | +0.025 |
ail | +0.021 |
brand | +0.020 |
onna | +0.020 |
Neg logit contributions | |
crow | -0.080 |
Bad | -0.080 |
Leave | -0.078 |
Stop | -0.077 |
vious | -0.076 |
Token | was |
Token ID | 373 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
pron | +0.793 |
masterpiece | +0.711 |
favourite | +0.697 |
brand | +0.632 |
collection | +0.621 |
Neg logit contributions | |
crow | -0.938 |
Bad | -0.888 |
Leave | -0.875 |
Stop | -0.859 |
vious | -0.840 |
Token | pretend |
Token ID | 16614 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
pron | +0.123 |
masterpiece | +0.115 |
favourite | +0.110 |
ail | +0.105 |
collection | +0.104 |
Neg logit contributions | |
crow | -0.094 |
Bad | -0.090 |
Leave | -0.088 |
vious | -0.086 |
hunters | -0.084 |
Token | . |
Token ID | 13 |
Feature activation | +2.14e-01 |
Pos logit contributions | |
crow | +0.081 |
Bad | +0.076 |
Leave | +0.075 |
Stop | +0.074 |
ases | +0.074 |
Neg logit contributions | |
pron | -0.075 |
masterpiece | -0.066 |
favourite | -0.065 |
ail | -0.060 |
brand | -0.059 |
Token | They |
Token ID | 1119 |
Feature activation | +2.94e-02 |
Pos logit contributions | |
crow | +0.017 |
Leave | +0.016 |
Bad | +0.016 |
Stop | +0.015 |
vious | +0.015 |
Neg logit contributions | |
pron | -0.017 |
masterpiece | -0.016 |
favourite | -0.015 |
collection | -0.014 |
brand | -0.014 |
Token | pretended |
Token ID | 36766 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
crow | +0.002 |
Leave | +0.002 |
vious | +0.002 |
Bad | +0.002 |
Stop | +0.002 |
Neg logit contributions | |
pron | -0.002 |
masterpiece | -0.002 |
favourite | -0.002 |
brand | -0.002 |
ail | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +2.30e-01 |
Pos logit contributions | |
pron | +0.572 |
favourite | +0.525 |
masterpiece | +0.514 |
ail | +0.480 |
brand | +0.474 |
Neg logit contributions | |
crow | -0.453 |
Bad | -0.403 |
vious | -0.403 |
Leave | -0.402 |
Stop | -0.398 |
Token | |
Token ID | 220 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
crow | +0.017 |
Leave | +0.016 |
Bad | +0.016 |
Stop | +0.015 |
vious | +0.015 |
Neg logit contributions | |
pron | -0.020 |
masterpiece | -0.018 |
favourite | -0.018 |
collection | -0.016 |
usual | -0.016 |
Token |
|
Token ID | 198 |
Feature activation | -3.18e-01 |
Pos logit contributions | |
pron | +0.102 |
masterpiece | +0.094 |
favourite | +0.093 |
collection | +0.087 |
usual | +0.085 |
Neg logit contributions | |
crow | -0.079 |
Leave | -0.074 |
Bad | -0.074 |
Stop | -0.071 |
vious | -0.070 |
Token |
|
Token ID | 198 |
Feature activation | -3.37e-01 |
Pos logit contributions | |
pron | +0.027 |
masterpiece | +0.025 |
favourite | +0.025 |
collection | +0.023 |
brand | +0.022 |
Neg logit contributions | |
crow | -0.023 |
Leave | -0.022 |
Bad | -0.022 |
Stop | -0.021 |
vious | -0.021 |
Token | S |
Token ID | 50 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
pron | +0.031 |
favourite | +0.028 |
masterpiece | +0.028 |
collection | +0.026 |
usual | +0.025 |
Neg logit contributions | |
crow | -0.023 |
Leave | -0.022 |
Bad | -0.022 |
vious | -0.021 |
Stop | -0.021 |
Token | ally |
Token ID | 453 |
Feature activation | -8.74e+00 |
Pos logit contributions | |
crow | +0.223 |
Bad | +0.213 |
Leave | +0.211 |
vious | +0.211 |
Stop | +0.210 |
Neg logit contributions | |
pron | -0.105 |
favourite | -0.093 |
masterpiece | -0.089 |
brand | -0.078 |
collection | -0.076 |
Token | couldn |
Token ID | 3521 |
Feature activation | -6.77e+00 |
Pos logit contributions | |
pron | +0.686 |
favourite | +0.614 |
masterpiece | +0.603 |
brand | +0.575 |
usual | +0.571 |
Neg logit contributions | |
crow | -0.678 |
vious | -0.639 |
Bad | -0.635 |
Leave | -0.634 |
Stop | -0.619 |
Token | 't |
Token ID | 470 |
Feature activation | +9.58e-01 |
Pos logit contributions | |
pron | +0.588 |
favourite | +0.539 |
masterpiece | +0.534 |
collection | +0.495 |
brand | +0.485 |
Neg logit contributions | |
crow | -0.489 |
Leave | -0.467 |
Bad | -0.458 |
Stop | -0.447 |
vious | -0.442 |
Token | wait |
Token ID | 4043 |
Feature activation | -5.13e-01 |
Pos logit contributions | |
crow | +0.072 |
Leave | +0.069 |
Bad | +0.067 |
vious | +0.066 |
Stop | +0.066 |
Neg logit contributions | |
pron | -0.079 |
favourite | -0.072 |
masterpiece | -0.071 |
brand | -0.066 |
collection | -0.065 |
Token | to |
Token ID | 284 |
Feature activation | -1.56e+00 |
Pos logit contributions | |
pron | +0.044 |
masterpiece | +0.038 |
favourite | +0.037 |
ail | +0.036 |
ouse | +0.036 |
Neg logit contributions | |
Leave | -0.033 |
vious | -0.031 |
ases | -0.031 |
crow | -0.030 |
Stop | -0.030 |
Token | give |
Token ID | 1577 |
Feature activation | +2.03e+00 |
Pos logit contributions | |
pron | +0.133 |
favourite | +0.123 |
masterpiece | +0.117 |
ail | +0.110 |
favorite | +0.109 |
Neg logit contributions | |
crow | -0.109 |
Leave | -0.106 |
Bad | -0.104 |
Stop | -0.102 |
ases | -0.099 |
Token | than |
Token ID | 621 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
Leave | +0.430 |
crow | +0.424 |
Bad | +0.409 |
vious | +0.406 |
Stop | +0.404 |
Neg logit contributions | |
favourite | -0.449 |
pron | -0.446 |
favorite | -0.418 |
masterpiece | -0.413 |
collection | -0.412 |
Token | having |
Token ID | 1719 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
crow | +0.127 |
Leave | +0.121 |
Bad | +0.118 |
Stop | +0.114 |
vious | +0.114 |
Neg logit contributions | |
pron | -0.145 |
favourite | -0.134 |
masterpiece | -0.133 |
usual | -0.125 |
collection | -0.123 |
Token | a |
Token ID | 257 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
pron | +0.093 |
favourite | +0.091 |
masterpiece | +0.087 |
collection | +0.081 |
favorite | +0.080 |
Neg logit contributions | |
crow | -0.115 |
Leave | -0.110 |
Bad | -0.108 |
Stop | -0.105 |
vious | -0.104 |
Token | l |
Token ID | 300 |
Feature activation | -2.68e+00 |
Pos logit contributions | |
crow | +0.009 |
Leave | +0.009 |
Bad | +0.009 |
Stop | +0.008 |
vious | +0.008 |
Neg logit contributions | |
pron | -0.008 |
favourite | -0.007 |
masterpiece | -0.007 |
collection | -0.006 |
usual | -0.006 |
Token | oll |
Token ID | 692 |
Feature activation | -9.71e+00 |
Pos logit contributions | |
pron | +0.112 |
favourite | +0.094 |
masterpiece | +0.080 |
collection | +0.078 |
ail | +0.073 |
Neg logit contributions | |
Bad | -0.312 |
Leave | -0.310 |
Stop | -0.305 |
crow | -0.301 |
Hey | -0.290 |
Token | ipop |
Token ID | 42800 |
Feature activation | -8.91e+00 |
Pos logit contributions | |
pron | +0.384 |
favourite | +0.314 |
masterpiece | +0.310 |
collection | +0.250 |
usual | +0.237 |
Neg logit contributions | |
crow | -1.151 |
Leave | -1.123 |
Bad | -1.118 |
Stop | -1.092 |
vious | -1.083 |
Token | . |
Token ID | 13 |
Feature activation | +3.52e-02 |
Pos logit contributions | |
pron | +0.712 |
favourite | +0.651 |
masterpiece | +0.646 |
collection | +0.592 |
usual | +0.580 |
Neg logit contributions | |
crow | -0.691 |
Leave | -0.667 |
Bad | -0.662 |
Stop | -0.639 |
vious | -0.628 |
Token | He |
Token ID | 679 |
Feature activation | +3.16e+00 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.003 |
Bad | +0.003 |
Stop | +0.003 |
vious | +0.003 |
Neg logit contributions | |
pron | -0.003 |
favourite | -0.002 |
masterpiece | -0.002 |
favorite | -0.002 |
onna | -0.002 |
Token | learned |
Token ID | 4499 |
Feature activation | +4.83e+00 |
Pos logit contributions | |
crow | +0.230 |
Bad | +0.223 |
Leave | +0.222 |
vious | +0.212 |
Stop | +0.212 |
Neg logit contributions | |
pron | -0.269 |
masterpiece | -0.245 |
favourite | -0.244 |
collection | -0.222 |
usual | -0.221 |
Token | that |
Token ID | 326 |
Feature activation | -2.85e+00 |
Pos logit contributions | |
crow | +0.407 |
Bad | +0.393 |
Leave | +0.390 |
Stop | +0.377 |
vious | +0.373 |
Neg logit contributions | |
pron | -0.361 |
masterpiece | -0.324 |
favourite | -0.318 |
collection | -0.291 |
usual | -0.281 |
Token | he |
Token ID | 339 |
Feature activation | +4.99e-01 |
Pos logit contributions | |
pron | +0.227 |
favourite | +0.211 |
masterpiece | +0.208 |
collection | +0.193 |
usual | +0.190 |
Neg logit contributions | |
crow | -0.221 |
Leave | -0.213 |
Bad | -0.210 |
Stop | -0.204 |
vious | -0.203 |