Token | in |
Token ID | 287 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
amed | +0.003 |
ished | +0.003 |
aming | +0.002 |
ady | +0.002 |
lling | +0.002 |
Neg logit contributions | |
hill | -0.002 |
mountain | -0.002 |
road | -0.001 |
ladder | -0.001 |
cliff | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
hill | +0.234 |
mountain | +0.216 |
subway | +0.207 |
ladder | +0.206 |
bike | +0.204 |
Neg logit contributions | |
amed | -0.209 |
aming | -0.193 |
ady | -0.189 |
ished | -0.187 |
lling | -0.182 |
Token | tree |
Token ID | 5509 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
hill | +0.187 |
subway | +0.177 |
mountain | +0.175 |
cone | +0.167 |
festival | +0.165 |
Neg logit contributions | |
amed | -0.262 |
aming | -0.242 |
ady | -0.238 |
ished | -0.234 |
ealing | -0.229 |
Token | in |
Token ID | 287 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
hill | +0.329 |
mountain | +0.315 |
subway | +0.304 |
taxi | +0.292 |
road | +0.289 |
Neg logit contributions | |
amed | -0.333 |
ished | -0.314 |
ady | -0.305 |
aming | -0.302 |
lling | -0.290 |
Token | a |
Token ID | 257 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
hill | +0.230 |
mountain | +0.212 |
subway | +0.206 |
bike | +0.203 |
ladder | +0.202 |
Neg logit contributions | |
amed | -0.244 |
aming | -0.226 |
ished | -0.225 |
ady | -0.223 |
lling | -0.219 |
Token | backyard |
Token ID | 24296 |
Feature activation | +1.87e-01 |
Pos logit contributions | |
hill | +0.178 |
subway | +0.175 |
mountain | +0.166 |
cone | +0.164 |
Zig | +0.161 |
Neg logit contributions | |
amed | -0.284 |
aming | -0.263 |
ady | -0.259 |
ealing | -0.255 |
ished | -0.255 |
Token | . |
Token ID | 13 |
Feature activation | +4.05e-02 |
Pos logit contributions | |
hill | +0.411 |
mountain | +0.394 |
subway | +0.387 |
bike | +0.374 |
taxi | +0.373 |
Neg logit contributions | |
amed | -0.365 |
ished | -0.344 |
aming | -0.331 |
ady | -0.327 |
lling | -0.319 |
Token | The |
Token ID | 383 |
Feature activation | -8.10e-03 |
Pos logit contributions | |
hill | +0.095 |
cliff | +0.093 |
mountain | +0.091 |
subway | +0.091 |
festival | +0.088 |
Neg logit contributions | |
amed | -0.061 |
ady | -0.056 |
ished | -0.053 |
aming | -0.052 |
lee | -0.050 |
Token | bird |
Token ID | 6512 |
Feature activation | +2.97e-02 |
Pos logit contributions | |
amed | +0.021 |
aming | +0.019 |
ished | +0.019 |
ady | +0.019 |
ealing | +0.018 |
Neg logit contributions | |
hill | -0.012 |
subway | -0.011 |
mountain | -0.010 |
cone | -0.010 |
festival | -0.010 |
Token | was |
Token ID | 373 |
Feature activation | -1.70e-02 |
Pos logit contributions | |
hill | +0.061 |
mountain | +0.057 |
road | +0.053 |
ladder | +0.052 |
cliff | +0.052 |
Neg logit contributions | |
amed | -0.069 |
ished | -0.066 |
aming | -0.062 |
ady | -0.062 |
lling | -0.062 |
Token | very |
Token ID | 845 |
Feature activation | +5.18e-02 |
Pos logit contributions | |
amed | +0.041 |
ished | +0.039 |
aming | +0.037 |
ady | +0.037 |
lling | +0.037 |
Neg logit contributions | |
hill | -0.033 |
mountain | -0.031 |
road | -0.029 |
ladder | -0.028 |
subway | -0.028 |
Token | tr |
Token ID | 491 |
Feature activation | +4.86e-02 |
Pos logit contributions | |
Zig | +0.084 |
hill | +0.077 |
mountain | +0.073 |
Raven | +0.073 |
subway | +0.072 |
Neg logit contributions | |
amed | -0.134 |
ished | -0.132 |
aming | -0.131 |
ealing | -0.129 |
lling | -0.129 |
Token | amp |
Token ID | 696 |
Feature activation | -5.06e-03 |
Pos logit contributions | |
hill | +0.113 |
mountain | +0.109 |
taxi | +0.108 |
subway | +0.107 |
road | +0.106 |
Neg logit contributions | |
amed | -0.090 |
ished | -0.087 |
aming | -0.083 |
ady | -0.082 |
lling | -0.078 |
Token | oline |
Token ID | 14453 |
Feature activation | +9.46e-02 |
Pos logit contributions | |
amed | +0.009 |
ished | +0.008 |
lling | +0.008 |
aming | +0.008 |
ps | +0.007 |
Neg logit contributions | |
hill | -0.013 |
mountain | -0.012 |
ladder | -0.012 |
taxi | -0.012 |
je | -0.012 |
Token | in |
Token ID | 287 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
hill | +0.189 |
mountain | +0.182 |
road | +0.167 |
cliff | +0.165 |
ladder | +0.165 |
Neg logit contributions | |
amed | -0.213 |
ished | -0.204 |
aming | -0.195 |
lling | -0.193 |
asing | -0.187 |
Token | a |
Token ID | 257 |
Feature activation | +1.20e-01 |
Pos logit contributions | |
hill | +0.228 |
mountain | +0.216 |
volcano | +0.204 |
ladder | +0.204 |
cone | +0.203 |
Neg logit contributions | |
amed | -0.240 |
aming | -0.224 |
lling | -0.221 |
ady | -0.214 |
ished | -0.213 |
Token | backyard |
Token ID | 24296 |
Feature activation | +1.87e-01 |
Pos logit contributions | |
Zig | +0.166 |
hill | +0.147 |
Raven | +0.145 |
mountain | +0.141 |
taxi | +0.136 |
Neg logit contributions | |
amed | -0.295 |
aming | -0.287 |
ady | -0.283 |
lling | -0.267 |
ealing | -0.266 |
Token | . |
Token ID | 13 |
Feature activation | +1.69e-02 |
Pos logit contributions | |
hill | +0.401 |
mountain | +0.390 |
subway | +0.378 |
bike | +0.375 |
cone | +0.374 |
Neg logit contributions | |
amed | -0.353 |
aming | -0.324 |
ished | -0.324 |
lling | -0.320 |
ady | -0.312 |
Token | She |
Token ID | 1375 |
Feature activation | +2.20e-02 |
Pos logit contributions | |
hill | +0.034 |
mountain | +0.032 |
cliff | +0.030 |
taxi | +0.030 |
volcano | +0.029 |
Neg logit contributions | |
amed | -0.039 |
ished | -0.037 |
ady | -0.034 |
aming | -0.034 |
ealing | -0.034 |
Token | couldn |
Token ID | 3521 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
hill | +0.048 |
mountain | +0.045 |
road | +0.042 |
ladder | +0.041 |
cliff | +0.041 |
Neg logit contributions | |
amed | -0.048 |
ished | -0.046 |
aming | -0.043 |
ady | -0.043 |
lling | -0.043 |
Token | 't |
Token ID | 470 |
Feature activation | -3.92e-02 |
Pos logit contributions | |
amed | +0.036 |
ished | +0.034 |
aming | +0.033 |
lling | +0.032 |
ady | +0.032 |
Neg logit contributions | |
hill | -0.019 |
mountain | -0.018 |
road | -0.016 |
taxi | -0.016 |
subway | -0.016 |
Token | believe |
Token ID | 1975 |
Feature activation | -6.65e-02 |
Pos logit contributions | |
amed | +0.083 |
ished | +0.079 |
aming | +0.075 |
lling | +0.075 |
ady | +0.074 |
Neg logit contributions | |
hill | -0.086 |
mountain | -0.081 |
subway | -0.075 |
ladder | -0.075 |
road | -0.075 |
Token | my |
Token ID | 1820 |
Feature activation | -9.35e-02 |
Pos logit contributions | |
hill | +0.184 |
mountain | +0.173 |
road | +0.160 |
subway | +0.157 |
cliff | +0.155 |
Neg logit contributions | |
amed | -0.287 |
ished | -0.273 |
aming | -0.264 |
ady | -0.260 |
lling | -0.257 |
Token | lived |
Token ID | 5615 |
Feature activation | -1.17e-05 |
Pos logit contributions | |
amed | +0.227 |
ished | +0.208 |
aming | +0.207 |
lling | +0.204 |
ady | +0.203 |
Neg logit contributions | |
hill | -0.174 |
mountain | -0.160 |
bike | -0.155 |
je | -0.152 |
taxi | -0.152 |
Token | in |
Token ID | 287 |
Feature activation | +9.41e-02 |
Pos logit contributions | |
amed | +0.000 |
ished | +0.000 |
aming | +0.000 |
lling | +0.000 |
ady | +0.000 |
Neg logit contributions | |
hill | -0.000 |
mountain | -0.000 |
road | -0.000 |
taxi | -0.000 |
ladder | -0.000 |
Token | a |
Token ID | 257 |
Feature activation | +1.03e-01 |
Pos logit contributions | |
hill | +0.213 |
mountain | +0.202 |
ladder | +0.194 |
cone | +0.193 |
cliff | +0.190 |
Neg logit contributions | |
amed | -0.182 |
aming | -0.168 |
ady | -0.164 |
lling | -0.154 |
ished | -0.153 |
Token | small |
Token ID | 1402 |
Feature activation | +4.77e-02 |
Pos logit contributions | |
hill | +0.173 |
cone | +0.172 |
Zig | +0.166 |
mountain | +0.162 |
subway | +0.162 |
Neg logit contributions | |
amed | -0.225 |
ady | -0.213 |
aming | -0.213 |
ealing | -0.202 |
ished | -0.201 |
Token | hut |
Token ID | 40812 |
Feature activation | +1.86e-01 |
Pos logit contributions | |
cone | +0.068 |
Zig | +0.067 |
hill | +0.064 |
subway | +0.063 |
mountain | +0.062 |
Neg logit contributions | |
ady | -0.113 |
amed | -0.113 |
ished | -0.111 |
aming | -0.107 |
arre | -0.103 |
Token | with |
Token ID | 351 |
Feature activation | +2.14e-02 |
Pos logit contributions | |
hill | +0.369 |
mountain | +0.349 |
subway | +0.332 |
road | +0.328 |
cliff | +0.325 |
Neg logit contributions | |
amed | -0.421 |
ished | -0.394 |
aming | -0.381 |
ady | -0.375 |
lling | -0.369 |
Token | his |
Token ID | 465 |
Feature activation | -8.69e-03 |
Pos logit contributions | |
hill | +0.042 |
mountain | +0.039 |
road | +0.036 |
ladder | +0.035 |
cliff | +0.035 |
Neg logit contributions | |
amed | -0.052 |
ished | -0.049 |
aming | -0.047 |
ady | -0.047 |
lling | -0.046 |
Token | mom |
Token ID | 1995 |
Feature activation | -1.97e-02 |
Pos logit contributions | |
amed | +0.018 |
ady | +0.016 |
ished | +0.016 |
aming | +0.015 |
ealing | +0.015 |
Neg logit contributions | |
hill | -0.018 |
mountain | -0.018 |
cliff | -0.017 |
subway | -0.017 |
road | -0.016 |
Token | my |
Token ID | 1820 |
Feature activation | -5.37e-02 |
Pos logit contributions | |
amed | +0.041 |
ished | +0.038 |
lling | +0.036 |
ealing | +0.036 |
asing | +0.036 |
Neg logit contributions | |
hill | -0.045 |
mountain | -0.042 |
road | -0.040 |
taxi | -0.040 |
bike | -0.039 |
Token | and |
Token ID | 290 |
Feature activation | +6.67e-02 |
Pos logit contributions | |
amed | +0.138 |
ished | +0.132 |
aming | +0.125 |
ealing | +0.125 |
lling | +0.124 |
Neg logit contributions | |
hill | -0.098 |
mountain | -0.090 |
road | -0.083 |
taxi | -0.081 |
ladder | -0.081 |
Token | pile |
Token ID | 14540 |
Feature activation | +9.96e-02 |
Pos logit contributions | |
hill | +0.080 |
Zig | +0.079 |
mountain | +0.075 |
road | +0.072 |
Raven | +0.071 |
Neg logit contributions | |
amed | -0.134 |
ished | -0.134 |
ealing | -0.131 |
aming | -0.131 |
lling | -0.130 |
Token | of |
Token ID | 286 |
Feature activation | +8.10e-02 |
Pos logit contributions | |
hill | +0.194 |
mountain | +0.178 |
road | +0.167 |
ladder | +0.166 |
taxi | +0.164 |
Neg logit contributions | |
amed | -0.241 |
ished | -0.232 |
ealing | -0.222 |
ady | -0.219 |
lling | -0.218 |
Token | hay |
Token ID | 27678 |
Feature activation | +1.55e-01 |
Pos logit contributions | |
hill | +0.148 |
mountain | +0.139 |
road | +0.127 |
cliff | +0.126 |
subway | +0.123 |
Neg logit contributions | |
amed | -0.199 |
ished | -0.192 |
ady | -0.183 |
aming | -0.182 |
lling | -0.180 |
Token | in |
Token ID | 287 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
hill | +0.293 |
mountain | +0.283 |
taxi | +0.263 |
subway | +0.263 |
road | +0.262 |
Neg logit contributions | |
amed | -0.358 |
ished | -0.339 |
aming | -0.327 |
lling | -0.322 |
ady | -0.317 |
Token | the |
Token ID | 262 |
Feature activation | +4.40e-02 |
Pos logit contributions | |
hill | +0.201 |
mountain | +0.187 |
road | +0.172 |
subway | +0.170 |
taxi | +0.170 |
Neg logit contributions | |
amed | -0.293 |
ished | -0.279 |
aming | -0.269 |
ady | -0.268 |
lling | -0.266 |
Token | barn |
Token ID | 25203 |
Feature activation | +1.84e-01 |
Pos logit contributions | |
hill | +0.067 |
mountain | +0.063 |
cliff | +0.057 |
subway | +0.056 |
bike | +0.056 |
Neg logit contributions | |
amed | -0.121 |
ished | -0.112 |
ady | -0.111 |
aming | -0.111 |
lling | -0.104 |
Token | . |
Token ID | 13 |
Feature activation | +1.13e-02 |
Pos logit contributions | |
hill | +0.350 |
mountain | +0.323 |
road | +0.305 |
ladder | +0.297 |
cliff | +0.292 |
Neg logit contributions | |
amed | -0.457 |
ished | -0.442 |
aming | -0.416 |
ady | -0.415 |
ealing | -0.414 |
Token | She |
Token ID | 1375 |
Feature activation | +2.09e-02 |
Pos logit contributions | |
hill | +0.024 |
mountain | +0.023 |
festival | +0.022 |
taxi | +0.022 |
cliff | +0.022 |
Neg logit contributions | |
amed | -0.023 |
ished | -0.023 |
ady | -0.021 |
lee | -0.020 |
ealing | -0.020 |
Token | was |
Token ID | 373 |
Feature activation | -2.91e-02 |
Pos logit contributions | |
hill | +0.043 |
mountain | +0.040 |
road | +0.037 |
ladder | +0.037 |
bike | +0.036 |
Neg logit contributions | |
amed | -0.048 |
ished | -0.047 |
lling | -0.045 |
asing | -0.044 |
aming | -0.044 |
Token | very |
Token ID | 845 |
Feature activation | +4.57e-02 |
Pos logit contributions | |
amed | +0.065 |
ished | +0.065 |
asing | +0.061 |
lling | +0.061 |
ealing | +0.060 |
Neg logit contributions | |
hill | -0.058 |
mountain | -0.054 |
ladder | -0.054 |
road | -0.053 |
taxi | -0.052 |
Token | hungry |
Token ID | 14720 |
Feature activation | -2.84e-02 |
Pos logit contributions | |
hill | +0.104 |
ladder | +0.098 |
mountain | +0.097 |
road | +0.094 |
taxi | +0.093 |
Neg logit contributions | |
amed | -0.094 |
ished | -0.090 |
asing | -0.087 |
lling | -0.087 |
eness | -0.085 |
Token | rabbit |
Token ID | 22746 |
Feature activation | +4.62e-02 |
Pos logit contributions | |
hill | +0.050 |
mountain | +0.048 |
subway | +0.046 |
road | +0.046 |
Zig | +0.046 |
Neg logit contributions | |
ished | -0.062 |
amed | -0.060 |
ady | -0.058 |
ealing | -0.057 |
aming | -0.056 |
Token | who |
Token ID | 508 |
Feature activation | +4.25e-02 |
Pos logit contributions | |
hill | +0.095 |
mountain | +0.090 |
ladder | +0.085 |
road | +0.085 |
subway | +0.084 |
Neg logit contributions | |
amed | -0.099 |
ished | -0.095 |
ady | -0.092 |
aming | -0.090 |
lling | -0.089 |
Token | lived |
Token ID | 5615 |
Feature activation | -7.59e-03 |
Pos logit contributions | |
hill | +0.083 |
mountain | +0.078 |
road | +0.073 |
subway | +0.071 |
taxi | +0.071 |
Neg logit contributions | |
amed | -0.102 |
ished | -0.097 |
aming | -0.092 |
ady | -0.092 |
lling | -0.092 |
Token | in |
Token ID | 287 |
Feature activation | +9.69e-02 |
Pos logit contributions | |
amed | +0.021 |
ished | +0.021 |
lling | +0.020 |
ealing | +0.019 |
asing | +0.019 |
Neg logit contributions | |
hill | -0.012 |
mountain | -0.010 |
road | -0.009 |
taxi | -0.009 |
ladder | -0.009 |
Token | a |
Token ID | 257 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
hill | +0.204 |
mountain | +0.191 |
road | +0.179 |
ladder | +0.179 |
subway | +0.179 |
Neg logit contributions | |
amed | -0.212 |
ished | -0.197 |
ady | -0.196 |
aming | -0.194 |
lling | -0.188 |
Token | hut |
Token ID | 40812 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
hill | +0.183 |
mountain | +0.172 |
subway | +0.172 |
ladder | +0.164 |
bike | +0.164 |
Neg logit contributions | |
amed | -0.255 |
ady | -0.245 |
aming | -0.239 |
ished | -0.236 |
ealing | -0.230 |
Token | in |
Token ID | 287 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
hill | +0.375 |
mountain | +0.355 |
subway | +0.341 |
road | +0.335 |
taxi | +0.331 |
Neg logit contributions | |
amed | -0.403 |
ished | -0.381 |
aming | -0.366 |
ady | -0.364 |
lling | -0.356 |
Token | the |
Token ID | 262 |
Feature activation | +4.17e-02 |
Pos logit contributions | |
hill | +0.220 |
mountain | +0.205 |
subway | +0.192 |
road | +0.191 |
taxi | +0.191 |
Neg logit contributions | |
amed | -0.266 |
ished | -0.249 |
ady | -0.245 |
aming | -0.245 |
lling | -0.240 |
Token | woods |
Token ID | 16479 |
Feature activation | +1.52e-01 |
Pos logit contributions | |
hill | +0.050 |
mountain | +0.045 |
subway | +0.041 |
ladder | +0.041 |
bike | +0.040 |
Neg logit contributions | |
amed | -0.129 |
ished | -0.120 |
aming | -0.120 |
ady | -0.119 |
lling | -0.116 |
Token | . |
Token ID | 13 |
Feature activation | +3.20e-02 |
Pos logit contributions | |
hill | +0.292 |
mountain | +0.271 |
road | +0.253 |
ladder | +0.247 |
cliff | +0.246 |
Neg logit contributions | |
amed | -0.373 |
ished | -0.359 |
aming | -0.338 |
ady | -0.336 |
lling | -0.335 |
Token | Every |
Token ID | 3887 |
Feature activation | -1.04e-02 |
Pos logit contributions | |
hill | +0.064 |
mountain | +0.061 |
subway | +0.058 |
taxi | +0.058 |
cliff | +0.058 |
Neg logit contributions | |
amed | -0.067 |
ished | -0.066 |
ady | -0.064 |
ealing | -0.060 |
aming | -0.060 |
Token | , |
Token ID | 11 |
Feature activation | +1.35e-01 |
Pos logit contributions | |
Zig | +0.093 |
hill | +0.092 |
cone | +0.088 |
mountain | +0.084 |
subway | +0.083 |
Neg logit contributions | |
amed | -0.142 |
ished | -0.140 |
ady | -0.137 |
ealing | -0.137 |
aming | -0.136 |
Token | colorful |
Token ID | 20239 |
Feature activation | +1.63e-02 |
Pos logit contributions | |
hill | +0.263 |
mountain | +0.246 |
road | +0.229 |
cliff | +0.229 |
subway | +0.226 |
Neg logit contributions | |
amed | -0.300 |
aming | -0.283 |
ished | -0.279 |
ady | -0.279 |
lling | -0.269 |
Token | grill |
Token ID | 29901 |
Feature activation | +1.03e-01 |
Pos logit contributions | |
Zig | +0.021 |
hill | +0.020 |
cone | +0.019 |
mountain | +0.018 |
Raven | +0.018 |
Neg logit contributions | |
amed | -0.036 |
ady | -0.035 |
ished | -0.035 |
lee | -0.034 |
aming | -0.034 |
Token | in |
Token ID | 287 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
hill | +0.220 |
mountain | +0.210 |
cliff | +0.196 |
road | +0.195 |
ladder | +0.193 |
Neg logit contributions | |
amed | -0.212 |
ished | -0.201 |
aming | -0.193 |
ady | -0.191 |
lling | -0.186 |
Token | the |
Token ID | 262 |
Feature activation | +4.41e-02 |
Pos logit contributions | |
hill | +0.224 |
mountain | +0.210 |
bike | +0.199 |
subway | +0.198 |
cone | +0.197 |
Neg logit contributions | |
amed | -0.263 |
aming | -0.243 |
ady | -0.242 |
lling | -0.238 |
ished | -0.238 |
Token | backyard |
Token ID | 24296 |
Feature activation | +1.82e-01 |
Pos logit contributions | |
hill | +0.061 |
mountain | +0.056 |
ladder | +0.050 |
bike | +0.050 |
cliff | +0.049 |
Neg logit contributions | |
amed | -0.129 |
ished | -0.121 |
aming | -0.120 |
ady | -0.119 |
ealing | -0.115 |
Token | . |
Token ID | 13 |
Feature activation | +3.26e-02 |
Pos logit contributions | |
hill | +0.364 |
mountain | +0.341 |
road | +0.318 |
subway | +0.311 |
ladder | +0.311 |
Neg logit contributions | |
amed | -0.430 |
ished | -0.411 |
aming | -0.389 |
ady | -0.388 |
lling | -0.385 |
Token | The |
Token ID | 383 |
Feature activation | -1.93e-02 |
Pos logit contributions | |
hill | +0.068 |
mountain | +0.065 |
cliff | +0.061 |
subway | +0.060 |
bike | +0.060 |
Neg logit contributions | |
amed | -0.067 |
ady | -0.064 |
ished | -0.063 |
aming | -0.060 |
ealing | -0.058 |
Token | grill |
Token ID | 29901 |
Feature activation | +8.34e-02 |
Pos logit contributions | |
amed | +0.060 |
ished | +0.058 |
asing | +0.056 |
aming | +0.055 |
ady | +0.055 |
Neg logit contributions | |
hill | -0.027 |
mountain | -0.023 |
road | -0.022 |
subway | -0.020 |
ladder | -0.020 |
Token | was |
Token ID | 373 |
Feature activation | -5.45e-04 |
Pos logit contributions | |
hill | +0.171 |
mountain | +0.162 |
road | +0.152 |
subway | +0.149 |
cliff | +0.149 |
Neg logit contributions | |
amed | -0.188 |
ished | -0.179 |
aming | -0.169 |
lling | -0.167 |
ady | -0.166 |
Token | red |
Token ID | 2266 |
Feature activation | -8.53e-03 |
Pos logit contributions | |
amed | +0.001 |
ished | +0.001 |
asing | +0.001 |
lling | +0.001 |
ealing | +0.001 |
Neg logit contributions | |
hill | -0.001 |
mountain | -0.001 |
road | -0.001 |
ladder | -0.001 |
taxi | -0.001 |
Token | Lily |
Token ID | 20037 |
Feature activation | -7.06e-02 |
Pos logit contributions | |
hill | +0.042 |
mountain | +0.040 |
road | +0.036 |
cliff | +0.036 |
subway | +0.036 |
Neg logit contributions | |
amed | -0.054 |
ished | -0.051 |
aming | -0.049 |
ady | -0.048 |
lling | -0.047 |
Token | who |
Token ID | 508 |
Feature activation | +3.79e-02 |
Pos logit contributions | |
amed | +0.174 |
ished | +0.167 |
aming | +0.157 |
lling | +0.156 |
ady | +0.156 |
Neg logit contributions | |
hill | -0.136 |
mountain | -0.126 |
road | -0.118 |
taxi | -0.115 |
subway | -0.114 |
Token | lived |
Token ID | 5615 |
Feature activation | -6.54e-03 |
Pos logit contributions | |
hill | +0.097 |
mountain | +0.093 |
cliff | +0.088 |
road | +0.087 |
ladder | +0.087 |
Neg logit contributions | |
amed | -0.069 |
ished | -0.064 |
aming | -0.061 |
ady | -0.059 |
lling | -0.059 |
Token | in |
Token ID | 287 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
amed | +0.016 |
ished | +0.015 |
aming | +0.015 |
ady | +0.015 |
lling | +0.014 |
Neg logit contributions | |
hill | -0.013 |
mountain | -0.012 |
road | -0.011 |
subway | -0.011 |
cliff | -0.011 |
Token | a |
Token ID | 257 |
Feature activation | +1.11e-01 |
Pos logit contributions | |
hill | +0.236 |
mountain | +0.224 |
cliff | +0.208 |
ladder | +0.205 |
road | +0.203 |
Neg logit contributions | |
amed | -0.209 |
aming | -0.195 |
ady | -0.185 |
ished | -0.184 |
lling | -0.179 |
Token | hut |
Token ID | 40812 |
Feature activation | +1.82e-01 |
Pos logit contributions | |
hill | +0.211 |
mountain | +0.201 |
cone | +0.201 |
Zig | +0.199 |
subway | +0.188 |
Neg logit contributions | |
amed | -0.228 |
aming | -0.218 |
ady | -0.212 |
ished | -0.210 |
ealing | -0.204 |
Token | with |
Token ID | 351 |
Feature activation | +2.06e-02 |
Pos logit contributions | |
hill | +0.385 |
mountain | +0.374 |
subway | +0.360 |
bike | +0.345 |
cone | +0.345 |
Neg logit contributions | |
amed | -0.378 |
aming | -0.353 |
ished | -0.348 |
ady | -0.331 |
lling | -0.325 |
Token | her |
Token ID | 607 |
Feature activation | -1.93e-03 |
Pos logit contributions | |
hill | +0.039 |
mountain | +0.036 |
taxi | +0.034 |
road | +0.034 |
ladder | +0.034 |
Neg logit contributions | |
amed | -0.049 |
ished | -0.049 |
lling | -0.047 |
ealing | -0.046 |
asing | -0.046 |
Token | mom |
Token ID | 1995 |
Feature activation | +3.19e-03 |
Pos logit contributions | |
amed | +0.003 |
ady | +0.003 |
ished | +0.003 |
sisters | +0.003 |
aming | +0.003 |
Neg logit contributions | |
hill | -0.004 |
mountain | -0.004 |
cliff | -0.004 |
road | -0.004 |
cone | -0.004 |
Token | and |
Token ID | 290 |
Feature activation | +7.50e-02 |
Pos logit contributions | |
hill | +0.007 |
mountain | +0.007 |
road | +0.006 |
taxi | +0.006 |
bike | +0.006 |
Neg logit contributions | |
amed | -0.007 |
ished | -0.007 |
lling | -0.006 |
ealing | -0.006 |
aming | -0.006 |
Token | dad |
Token ID | 9955 |
Feature activation | -2.90e-02 |
Pos logit contributions | |
hill | +0.159 |
mountain | +0.149 |
road | +0.140 |
ladder | +0.137 |
subway | +0.137 |
Neg logit contributions | |
amed | -0.170 |
ished | -0.163 |
ady | -0.153 |
aming | -0.153 |
lling | -0.152 |
Token | a |
Token ID | 257 |
Feature activation | +8.22e-02 |
Pos logit contributions | |
amed | +0.052 |
ished | +0.051 |
lling | +0.049 |
ealing | +0.049 |
ady | +0.049 |
Neg logit contributions | |
hill | -0.033 |
mountain | -0.030 |
subway | -0.028 |
taxi | -0.028 |
road | -0.028 |
Token | big |
Token ID | 1263 |
Feature activation | +6.14e-02 |
Pos logit contributions | |
Zig | +0.148 |
hill | +0.148 |
mountain | +0.140 |
road | +0.134 |
subway | +0.132 |
Neg logit contributions | |
amed | -0.156 |
ady | -0.145 |
aming | -0.142 |
ealing | -0.140 |
lling | -0.138 |
Token | chest |
Token ID | 7721 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
Zig | +0.090 |
hill | +0.084 |
mountain | +0.080 |
subway | +0.079 |
Raven | +0.078 |
Neg logit contributions | |
amed | -0.136 |
ished | -0.134 |
ady | -0.130 |
aming | -0.130 |
ealing | -0.127 |
Token | in |
Token ID | 287 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
hill | +0.229 |
mountain | +0.214 |
road | +0.198 |
cliff | +0.194 |
subway | +0.194 |
Neg logit contributions | |
amed | -0.310 |
ished | -0.296 |
aming | -0.283 |
ady | -0.280 |
lling | -0.280 |
Token | the |
Token ID | 262 |
Feature activation | +4.49e-02 |
Pos logit contributions | |
hill | +0.209 |
mountain | +0.197 |
road | +0.180 |
subway | +0.179 |
ladder | +0.178 |
Neg logit contributions | |
amed | -0.287 |
ished | -0.269 |
aming | -0.264 |
lling | -0.264 |
ady | -0.260 |
Token | backyard |
Token ID | 24296 |
Feature activation | +1.81e-01 |
Pos logit contributions | |
hill | +0.070 |
mountain | +0.065 |
ladder | +0.063 |
cliff | +0.063 |
taxi | +0.062 |
Neg logit contributions | |
amed | -0.116 |
aming | -0.107 |
ady | -0.104 |
ished | -0.103 |
lling | -0.099 |
Token | . |
Token ID | 13 |
Feature activation | +2.83e-02 |
Pos logit contributions | |
hill | +0.332 |
mountain | +0.301 |
road | +0.285 |
ladder | +0.277 |
cliff | +0.269 |
Neg logit contributions | |
amed | -0.467 |
ished | -0.453 |
ealing | -0.431 |
ady | -0.425 |
aming | -0.424 |
Token | It |
Token ID | 632 |
Feature activation | +5.43e-02 |
Pos logit contributions | |
hill | +0.056 |
mountain | +0.054 |
cliff | +0.049 |
subway | +0.048 |
road | +0.048 |
Neg logit contributions | |
amed | -0.067 |
ished | -0.064 |
ady | -0.059 |
aming | -0.059 |
ealing | -0.059 |
Token | was |
Token ID | 373 |
Feature activation | -2.54e-03 |
Pos logit contributions | |
hill | +0.111 |
mountain | +0.104 |
road | +0.098 |
cliff | +0.096 |
subway | +0.096 |
Neg logit contributions | |
amed | -0.126 |
ished | -0.120 |
aming | -0.113 |
ady | -0.112 |
lling | -0.112 |
Token | locked |
Token ID | 8970 |
Feature activation | +2.15e-03 |
Pos logit contributions | |
amed | +0.006 |
ished | +0.005 |
aming | +0.005 |
ady | +0.005 |
lling | +0.005 |
Neg logit contributions | |
hill | -0.005 |
mountain | -0.005 |
cliff | -0.005 |
subway | -0.004 |
road | -0.004 |
Token | tight |
Token ID | 5381 |
Feature activation | -7.30e-03 |
Pos logit contributions | |
hill | +0.004 |
mountain | +0.004 |
road | +0.004 |
ladder | +0.004 |
cliff | +0.004 |
Neg logit contributions | |
amed | -0.005 |
ished | -0.005 |
aming | -0.005 |
ady | -0.005 |
lling | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | +8.69e-02 |
Pos logit contributions | |
hill | +0.015 |
mountain | +0.014 |
road | +0.013 |
ladder | +0.013 |
cliff | +0.013 |
Neg logit contributions | |
amed | -0.017 |
ished | -0.016 |
aming | -0.016 |
ady | -0.015 |
lling | -0.015 |
Token | tiny |
Token ID | 7009 |
Feature activation | +3.12e-02 |
Pos logit contributions | |
hill | +0.187 |
mountain | +0.177 |
Zig | +0.176 |
subway | +0.173 |
cone | +0.173 |
Neg logit contributions | |
ealing | -0.136 |
amed | -0.131 |
aming | -0.131 |
ady | -0.124 |
sure | -0.118 |
Token | bird |
Token ID | 6512 |
Feature activation | +6.68e-02 |
Pos logit contributions | |
hill | +0.049 |
Zig | +0.047 |
subway | +0.045 |
mountain | +0.045 |
road | +0.044 |
Neg logit contributions | |
ished | -0.065 |
ady | -0.063 |
ealing | -0.062 |
amed | -0.060 |
lee | -0.059 |
Token | in |
Token ID | 287 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
hill | +0.146 |
mountain | +0.138 |
subway | +0.132 |
ladder | +0.131 |
road | +0.129 |
Neg logit contributions | |
amed | -0.137 |
ished | -0.131 |
ady | -0.128 |
aming | -0.125 |
asing | -0.123 |
Token | a |
Token ID | 257 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
hill | +0.236 |
mountain | +0.215 |
bike | +0.213 |
ladder | +0.212 |
subway | +0.210 |
Neg logit contributions | |
amed | -0.192 |
aming | -0.182 |
ady | -0.180 |
ps | -0.177 |
lling | -0.175 |
Token | backyard |
Token ID | 24296 |
Feature activation | +1.80e-01 |
Pos logit contributions | |
hill | +0.192 |
subway | +0.185 |
mountain | +0.182 |
cone | +0.178 |
festival | +0.173 |
Neg logit contributions | |
amed | -0.264 |
ady | -0.248 |
aming | -0.246 |
ealing | -0.236 |
ished | -0.233 |
Token | . |
Token ID | 13 |
Feature activation | +3.85e-02 |
Pos logit contributions | |
hill | +0.406 |
mountain | +0.393 |
subway | +0.387 |
bike | +0.378 |
cone | +0.374 |
Neg logit contributions | |
amed | -0.322 |
ished | -0.303 |
ady | -0.290 |
aming | -0.290 |
lling | -0.279 |
Token | The |
Token ID | 383 |
Feature activation | -5.63e-03 |
Pos logit contributions | |
hill | +0.091 |
mountain | +0.086 |
cliff | +0.085 |
subway | +0.085 |
taxi | +0.083 |
Neg logit contributions | |
amed | -0.061 |
ished | -0.056 |
ady | -0.056 |
aming | -0.054 |
teaches | -0.053 |
Token | bird |
Token ID | 6512 |
Feature activation | +3.08e-02 |
Pos logit contributions | |
amed | +0.015 |
aming | +0.014 |
ished | +0.014 |
ady | +0.014 |
lling | +0.014 |
Neg logit contributions | |
hill | -0.007 |
mountain | -0.007 |
subway | -0.007 |
cone | -0.007 |
festival | -0.006 |
Token | 's |
Token ID | 338 |
Feature activation | -2.11e-02 |
Pos logit contributions | |
hill | +0.062 |
mountain | +0.058 |
road | +0.054 |
ladder | +0.053 |
cliff | +0.053 |
Neg logit contributions | |
amed | -0.072 |
ished | -0.069 |
aming | -0.065 |
ady | -0.065 |
lling | -0.065 |
Token | feathers |
Token ID | 27400 |
Feature activation | +3.21e-02 |
Pos logit contributions | |
amed | +0.046 |
ished | +0.043 |
ady | +0.041 |
aming | +0.041 |
ealing | +0.040 |
Neg logit contributions | |
hill | -0.046 |
mountain | -0.043 |
subway | -0.040 |
road | -0.040 |
cliff | -0.040 |
Token | He |
Token ID | 679 |
Feature activation | +2.50e-02 |
Pos logit contributions | |
hill | +0.088 |
subway | +0.086 |
mountain | +0.084 |
cliff | +0.084 |
taxi | +0.082 |
Neg logit contributions | |
amed | -0.065 |
ished | -0.064 |
ady | -0.063 |
aming | -0.061 |
ealing | -0.057 |
Token | lived |
Token ID | 5615 |
Feature activation | -1.80e-02 |
Pos logit contributions | |
hill | +0.061 |
mountain | +0.058 |
ladder | +0.056 |
cliff | +0.055 |
taxi | +0.054 |
Neg logit contributions | |
amed | -0.048 |
ished | -0.043 |
aming | -0.041 |
ady | -0.040 |
lling | -0.040 |
Token | in |
Token ID | 287 |
Feature activation | +9.35e-02 |
Pos logit contributions | |
amed | +0.047 |
ished | +0.045 |
aming | +0.043 |
ady | +0.043 |
lling | +0.043 |
Neg logit contributions | |
hill | -0.031 |
mountain | -0.029 |
road | -0.027 |
subway | -0.026 |
ladder | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
hill | +0.212 |
mountain | +0.198 |
ladder | +0.191 |
cliff | +0.189 |
subway | +0.188 |
Neg logit contributions | |
amed | -0.188 |
aming | -0.173 |
ady | -0.172 |
ished | -0.169 |
lling | -0.165 |
Token | big |
Token ID | 1263 |
Feature activation | +8.69e-02 |
Pos logit contributions | |
hill | +0.194 |
subway | +0.192 |
mountain | +0.186 |
cone | +0.183 |
Zig | +0.182 |
Neg logit contributions | |
amed | -0.212 |
aming | -0.198 |
ady | -0.198 |
ished | -0.190 |
ealing | -0.187 |
Token | barn |
Token ID | 25203 |
Feature activation | +1.80e-01 |
Pos logit contributions | |
subway | +0.112 |
Zig | +0.111 |
hill | +0.110 |
ladder | +0.106 |
taxi | +0.106 |
Neg logit contributions | |
amed | -0.206 |
ady | -0.203 |
ished | -0.203 |
aming | -0.199 |
lee | -0.194 |
Token | with |
Token ID | 351 |
Feature activation | +2.61e-02 |
Pos logit contributions | |
hill | +0.347 |
mountain | +0.327 |
subway | +0.311 |
cliff | +0.304 |
taxi | +0.304 |
Neg logit contributions | |
amed | -0.423 |
ished | -0.393 |
aming | -0.383 |
ady | -0.379 |
lling | -0.365 |
Token | his |
Token ID | 465 |
Feature activation | +6.22e-03 |
Pos logit contributions | |
hill | +0.052 |
mountain | +0.048 |
road | +0.045 |
cliff | +0.045 |
ladder | +0.044 |
Neg logit contributions | |
amed | -0.063 |
ished | -0.059 |
aming | -0.057 |
ady | -0.057 |
lling | -0.055 |
Token | mom |
Token ID | 1995 |
Feature activation | -1.34e-03 |
Pos logit contributions | |
hill | +0.016 |
mountain | +0.015 |
subway | +0.015 |
cliff | +0.014 |
festival | +0.014 |
Neg logit contributions | |
amed | -0.010 |
ady | -0.009 |
ished | -0.008 |
aming | -0.008 |
ealing | -0.008 |
Token | and |
Token ID | 290 |
Feature activation | +7.24e-02 |
Pos logit contributions | |
amed | +0.003 |
ished | +0.002 |
lling | +0.002 |
aming | +0.002 |
ealing | +0.002 |
Neg logit contributions | |
hill | -0.003 |
mountain | -0.003 |
road | -0.003 |
taxi | -0.003 |
ladder | -0.003 |
Token | dad |
Token ID | 9955 |
Feature activation | -3.11e-02 |
Pos logit contributions | |
hill | +0.172 |
mountain | +0.161 |
road | +0.154 |
cliff | +0.152 |
ladder | +0.151 |
Neg logit contributions | |
amed | -0.141 |
ished | -0.129 |
aming | -0.124 |
ady | -0.122 |
lling | -0.119 |
Token | a |
Token ID | 257 |
Feature activation | +8.32e-02 |
Pos logit contributions | |
hill | +0.011 |
mountain | +0.010 |
road | +0.010 |
subway | +0.009 |
taxi | +0.009 |
Neg logit contributions | |
amed | -0.015 |
ished | -0.014 |
aming | -0.013 |
lling | -0.013 |
ady | -0.013 |
Token | small |
Token ID | 1402 |
Feature activation | +2.50e-02 |
Pos logit contributions | |
hill | +0.179 |
mountain | +0.170 |
Zig | +0.164 |
subway | +0.163 |
cone | +0.162 |
Neg logit contributions | |
ealing | -0.136 |
amed | -0.134 |
aming | -0.132 |
ady | -0.128 |
ished | -0.121 |
Token | farm |
Token ID | 5318 |
Feature activation | +1.35e-01 |
Pos logit contributions | |
Zig | +0.040 |
hill | +0.038 |
cone | +0.037 |
mountain | +0.036 |
subway | +0.036 |
Neg logit contributions | |
ady | -0.049 |
ished | -0.048 |
ealing | -0.047 |
amed | -0.047 |
aming | -0.045 |
Token | with |
Token ID | 351 |
Feature activation | +2.52e-02 |
Pos logit contributions | |
hill | +0.298 |
mountain | +0.288 |
cliff | +0.271 |
subway | +0.270 |
ladder | +0.266 |
Neg logit contributions | |
amed | -0.254 |
aming | -0.234 |
ady | -0.230 |
ished | -0.223 |
lling | -0.210 |
Token | a |
Token ID | 257 |
Feature activation | +9.36e-02 |
Pos logit contributions | |
hill | +0.047 |
mountain | +0.044 |
road | +0.041 |
taxi | +0.040 |
subway | +0.040 |
Neg logit contributions | |
amed | -0.063 |
ished | -0.061 |
aming | -0.057 |
lling | -0.057 |
ady | -0.057 |
Token | barn |
Token ID | 25203 |
Feature activation | +1.79e-01 |
Pos logit contributions | |
hill | +0.153 |
Zig | +0.145 |
cone | +0.144 |
mountain | +0.143 |
subway | +0.135 |
Neg logit contributions | |
amed | -0.194 |
ealing | -0.187 |
ady | -0.187 |
aming | -0.178 |
arre | -0.174 |
Token | full |
Token ID | 1336 |
Feature activation | +2.48e-02 |
Pos logit contributions | |
hill | +0.393 |
mountain | +0.373 |
cliff | +0.350 |
subway | +0.347 |
bike | +0.345 |
Neg logit contributions | |
amed | -0.365 |
ished | -0.333 |
ady | -0.325 |
aming | -0.324 |
lling | -0.309 |
Token | of |
Token ID | 286 |
Feature activation | +7.07e-02 |
Pos logit contributions | |
hill | +0.054 |
mountain | +0.050 |
road | +0.047 |
cliff | +0.047 |
subway | +0.046 |
Neg logit contributions | |
amed | -0.053 |
ished | -0.050 |
ady | -0.048 |
aming | -0.048 |
lling | -0.046 |
Token | hay |
Token ID | 27678 |
Feature activation | +1.54e-01 |
Pos logit contributions | |
hill | +0.123 |
mountain | +0.113 |
road | +0.104 |
cliff | +0.102 |
ladder | +0.102 |
Neg logit contributions | |
amed | -0.185 |
ished | -0.177 |
aming | -0.170 |
ady | -0.169 |
lling | -0.168 |
Token | . |
Token ID | 13 |
Feature activation | +3.81e-02 |
Pos logit contributions | |
hill | +0.335 |
mountain | +0.319 |
cliff | +0.301 |
subway | +0.300 |
taxi | +0.299 |
Neg logit contributions | |
amed | -0.310 |
ished | -0.293 |
ady | -0.279 |
aming | -0.277 |
lling | -0.272 |
Token | One |
Token ID | 1881 |
Feature activation | +2.76e-02 |
Pos logit contributions | |
hill | +0.084 |
mountain | +0.082 |
cliff | +0.080 |
subway | +0.076 |
taxi | +0.076 |
Neg logit contributions | |
amed | -0.066 |
ady | -0.063 |
ished | -0.060 |
aming | -0.058 |
teaches | -0.055 |
Token | Pink |
Token ID | 14657 |
Feature activation | -1.04e-01 |
Pos logit contributions | |
hill | +0.072 |
mountain | +0.067 |
subway | +0.061 |
road | +0.061 |
cliff | +0.060 |
Neg logit contributions | |
amed | -0.110 |
ished | -0.106 |
aming | -0.101 |
ady | -0.100 |
lling | -0.099 |
Token | y |
Token ID | 88 |
Feature activation | -1.59e-02 |
Pos logit contributions | |
amed | +0.239 |
ished | +0.230 |
asing | +0.224 |
ealing | +0.223 |
lling | +0.219 |
Neg logit contributions | |
hill | -0.202 |
mountain | -0.186 |
je | -0.179 |
road | -0.177 |
valley | -0.172 |
Token | lived |
Token ID | 5615 |
Feature activation | -1.73e-02 |
Pos logit contributions | |
amed | +0.036 |
ished | +0.034 |
ady | +0.033 |
aming | +0.033 |
lling | +0.032 |
Neg logit contributions | |
hill | -0.033 |
mountain | -0.031 |
road | -0.029 |
cliff | -0.029 |
ladder | -0.029 |
Token | in |
Token ID | 287 |
Feature activation | +9.82e-02 |
Pos logit contributions | |
amed | +0.045 |
ished | +0.043 |
aming | +0.041 |
ady | +0.041 |
lling | +0.040 |
Neg logit contributions | |
hill | -0.031 |
mountain | -0.029 |
road | -0.027 |
subway | -0.026 |
cliff | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
hill | +0.228 |
mountain | +0.212 |
ladder | +0.204 |
cliff | +0.200 |
subway | +0.198 |
Neg logit contributions | |
amed | -0.196 |
aming | -0.183 |
ady | -0.178 |
ished | -0.173 |
lling | -0.171 |
Token | barn |
Token ID | 25203 |
Feature activation | +1.79e-01 |
Pos logit contributions | |
hill | +0.205 |
subway | +0.193 |
mountain | +0.192 |
cone | +0.186 |
Zig | +0.184 |
Neg logit contributions | |
amed | -0.228 |
aming | -0.214 |
ady | -0.212 |
ished | -0.203 |
ealing | -0.202 |
Token | with |
Token ID | 351 |
Feature activation | +2.31e-02 |
Pos logit contributions | |
hill | +0.339 |
mountain | +0.319 |
subway | +0.301 |
taxi | +0.293 |
cliff | +0.293 |
Neg logit contributions | |
amed | -0.427 |
ished | -0.397 |
aming | -0.389 |
ady | -0.382 |
lling | -0.371 |
Token | his |
Token ID | 465 |
Feature activation | +4.39e-03 |
Pos logit contributions | |
hill | +0.047 |
mountain | +0.043 |
road | +0.040 |
cliff | +0.040 |
subway | +0.040 |
Neg logit contributions | |
amed | -0.056 |
ished | -0.052 |
aming | -0.050 |
ady | -0.049 |
lling | -0.048 |
Token | friends |
Token ID | 2460 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
hill | +0.011 |
mountain | +0.010 |
cliff | +0.010 |
subway | +0.010 |
road | +0.010 |
Neg logit contributions | |
amed | -0.008 |
ady | -0.007 |
ished | -0.007 |
aming | -0.006 |
ealing | -0.006 |
Token | , |
Token ID | 11 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
hill | +0.027 |
mountain | +0.025 |
road | +0.023 |
ladder | +0.023 |
cliff | +0.023 |
Neg logit contributions | |
amed | -0.035 |
ished | -0.034 |
aming | -0.032 |
ady | -0.032 |
lling | -0.031 |
Token | Cow |
Token ID | 10417 |
Feature activation | -6.92e-02 |
Pos logit contributions | |
hill | +0.221 |
mountain | +0.213 |
cliff | +0.198 |
ladder | +0.194 |
subway | +0.193 |
Neg logit contributions | |
amed | -0.227 |
aming | -0.212 |
ady | -0.211 |
ished | -0.209 |
lling | -0.204 |
Token | a |
Token ID | 257 |
Feature activation | +8.24e-02 |
Pos logit contributions | |
amed | +0.040 |
ished | +0.039 |
aming | +0.036 |
ady | +0.036 |
lling | +0.036 |
Neg logit contributions | |
hill | -0.032 |
mountain | -0.030 |
road | -0.028 |
ladder | -0.027 |
cliff | -0.027 |
Token | big |
Token ID | 1263 |
Feature activation | +6.11e-02 |
Pos logit contributions | |
Zig | +0.135 |
hill | +0.128 |
mountain | +0.120 |
cone | +0.119 |
Raven | +0.117 |
Neg logit contributions | |
amed | -0.165 |
aming | -0.159 |
ady | -0.157 |
ealing | -0.157 |
lling | -0.154 |
Token | gear |
Token ID | 7733 |
Feature activation | +8.29e-02 |
Pos logit contributions | |
Zig | +0.076 |
hill | +0.072 |
Raven | +0.067 |
mountain | +0.066 |
subway | +0.064 |
Neg logit contributions | |
amed | -0.144 |
ished | -0.141 |
aming | -0.140 |
ady | -0.139 |
ealing | -0.139 |
Token | in |
Token ID | 287 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
hill | +0.151 |
mountain | +0.145 |
valley | +0.137 |
Zig | +0.137 |
cliff | +0.137 |
Neg logit contributions | |
amed | -0.177 |
ished | -0.168 |
lling | -0.166 |
aming | -0.165 |
ady | -0.162 |
Token | the |
Token ID | 262 |
Feature activation | +4.90e-02 |
Pos logit contributions | |
hill | +0.230 |
mountain | +0.218 |
cone | +0.214 |
ladder | +0.211 |
bike | +0.210 |
Neg logit contributions | |
amed | -0.248 |
aming | -0.235 |
lling | -0.232 |
ady | -0.228 |
ished | -0.222 |
Token | barn |
Token ID | 25203 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
hill | +0.067 |
mountain | +0.061 |
ladder | +0.057 |
cliff | +0.056 |
taxi | +0.054 |
Neg logit contributions | |
amed | -0.142 |
aming | -0.132 |
ady | -0.132 |
ished | -0.131 |
lling | -0.126 |
Token | and |
Token ID | 290 |
Feature activation | +9.03e-02 |
Pos logit contributions | |
hill | +0.354 |
mountain | +0.331 |
road | +0.308 |
subway | +0.303 |
cliff | +0.303 |
Neg logit contributions | |
amed | -0.421 |
ished | -0.402 |
aming | -0.381 |
ady | -0.380 |
lling | -0.377 |
Token | asked |
Token ID | 1965 |
Feature activation | -7.75e-02 |
Pos logit contributions | |
hill | +0.192 |
mountain | +0.179 |
cliff | +0.167 |
road | +0.166 |
volcano | +0.165 |
Neg logit contributions | |
amed | -0.195 |
ished | -0.186 |
lling | -0.174 |
aming | -0.170 |
ady | -0.168 |
Token | his |
Token ID | 465 |
Feature activation | -3.73e-02 |
Pos logit contributions | |
amed | +0.228 |
ished | +0.218 |
aming | +0.216 |
ady | +0.214 |
lling | +0.210 |
Neg logit contributions | |
hill | -0.110 |
mountain | -0.101 |
taxi | -0.099 |
subway | -0.095 |
road | -0.094 |
Token | mom |
Token ID | 1995 |
Feature activation | -3.07e-02 |
Pos logit contributions | |
amed | +0.087 |
ished | +0.082 |
aming | +0.079 |
lling | +0.078 |
asing | +0.076 |
Neg logit contributions | |
hill | -0.079 |
mountain | -0.075 |
taxi | -0.070 |
road | -0.070 |
subway | -0.069 |
Token | what |
Token ID | 644 |
Feature activation | +3.34e-02 |
Pos logit contributions | |
amed | +0.084 |
ished | +0.077 |
aming | +0.077 |
lling | +0.076 |
ady | +0.073 |
Neg logit contributions | |
hill | -0.053 |
mountain | -0.049 |
taxi | -0.046 |
ladder | -0.045 |
bike | -0.045 |
Token | decided |
Token ID | 3066 |
Feature activation | -2.58e-02 |
Pos logit contributions | |
amed | +0.182 |
ished | +0.173 |
asing | +0.167 |
aming | +0.163 |
lling | +0.163 |
Neg logit contributions | |
hill | -0.136 |
mountain | -0.126 |
road | -0.118 |
cliff | -0.113 |
subway | -0.111 |
Token | to |
Token ID | 284 |
Feature activation | +7.88e-02 |
Pos logit contributions | |
amed | +0.056 |
ished | +0.055 |
lling | +0.051 |
sure | +0.051 |
convinced | +0.051 |
Neg logit contributions | |
hill | -0.049 |
mountain | -0.045 |
valley | -0.043 |
taxi | -0.042 |
volcano | -0.041 |
Token | explore |
Token ID | 7301 |
Feature activation | +2.98e-02 |
Pos logit contributions | |
hill | +0.138 |
mountain | +0.130 |
subway | +0.119 |
road | +0.118 |
taxi | +0.118 |
Neg logit contributions | |
amed | -0.196 |
ished | -0.190 |
ady | -0.181 |
aming | -0.179 |
lling | -0.178 |
Token | a |
Token ID | 257 |
Feature activation | +9.64e-02 |
Pos logit contributions | |
hill | +0.059 |
mountain | +0.055 |
taxi | +0.053 |
ladder | +0.053 |
road | +0.053 |
Neg logit contributions | |
amed | -0.065 |
ished | -0.064 |
aming | -0.061 |
lling | -0.060 |
ady | -0.059 |
Token | big |
Token ID | 1263 |
Feature activation | +7.55e-02 |
Pos logit contributions | |
Zig | +0.170 |
cone | +0.158 |
stack | +0.158 |
taxi | +0.153 |
Raven | +0.153 |
Neg logit contributions | |
aming | -0.162 |
mistakes | -0.162 |
amed | -0.161 |
lling | -0.159 |
sure | -0.159 |
Token | chamber |
Token ID | 11847 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
Zig | +0.093 |
taxi | +0.082 |
Raven | +0.081 |
subway | +0.080 |
cone | +0.075 |
Neg logit contributions | |
ished | -0.171 |
aming | -0.169 |
ady | -0.167 |
amed | -0.167 |
lling | -0.164 |
Token | in |
Token ID | 287 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
hill | +0.377 |
subway | +0.368 |
taxi | +0.366 |
mountain | +0.360 |
bicycle | +0.357 |
Neg logit contributions | |
amed | -0.330 |
aming | -0.314 |
ished | -0.312 |
asing | -0.297 |
lling | -0.294 |
Token | the |
Token ID | 262 |
Feature activation | +4.65e-02 |
Pos logit contributions | |
hill | +0.216 |
mountain | +0.201 |
subway | +0.196 |
ladder | +0.193 |
taxi | +0.193 |
Neg logit contributions | |
amed | -0.272 |
ished | -0.256 |
aming | -0.255 |
ady | -0.250 |
lling | -0.249 |
Token | forest |
Token ID | 8222 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
hill | +0.059 |
cone | +0.056 |
taxi | +0.056 |
ladder | +0.055 |
bike | +0.055 |
Neg logit contributions | |
amed | -0.129 |
aming | -0.118 |
ady | -0.117 |
ished | -0.114 |
lling | -0.110 |
Token | . |
Token ID | 13 |
Feature activation | +2.89e-02 |
Pos logit contributions | |
hill | +0.323 |
mountain | +0.304 |
subway | +0.284 |
road | +0.283 |
taxi | +0.283 |
Neg logit contributions | |
amed | -0.351 |
ished | -0.335 |
aming | -0.318 |
ady | -0.316 |
lling | -0.314 |
Token | He |
Token ID | 679 |
Feature activation | +2.27e-02 |
Pos logit contributions | |
hill | +0.055 |
mountain | +0.053 |
subway | +0.050 |
cliff | +0.050 |
taxi | +0.049 |
Neg logit contributions | |
amed | -0.067 |
ished | -0.066 |
ady | -0.061 |
ealing | -0.061 |
aming | -0.060 |
Token | a |
Token ID | 257 |
Feature activation | +7.62e-02 |
Pos logit contributions | |
hill | +0.009 |
mountain | +0.008 |
road | +0.008 |
taxi | +0.007 |
cliff | +0.007 |
Neg logit contributions | |
ished | -0.014 |
amed | -0.014 |
asing | -0.013 |
lling | -0.013 |
ealing | -0.013 |
Token | big |
Token ID | 1263 |
Feature activation | +5.36e-02 |
Pos logit contributions | |
hill | +0.150 |
Zig | +0.141 |
mountain | +0.141 |
cone | +0.137 |
subway | +0.137 |
Neg logit contributions | |
amed | -0.145 |
aming | -0.140 |
ealing | -0.140 |
ady | -0.137 |
ished | -0.133 |
Token | tree |
Token ID | 5509 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
Zig | +0.078 |
hill | +0.076 |
cone | +0.072 |
mountain | +0.070 |
road | +0.069 |
Neg logit contributions | |
amed | -0.118 |
ished | -0.118 |
aming | -0.115 |
ealing | -0.113 |
ady | -0.112 |
Token | in |
Token ID | 287 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
hill | +0.288 |
mountain | +0.274 |
subway | +0.263 |
road | +0.253 |
taxi | +0.253 |
Neg logit contributions | |
amed | -0.335 |
ished | -0.317 |
aming | -0.311 |
ady | -0.310 |
lling | -0.297 |
Token | the |
Token ID | 262 |
Feature activation | +4.82e-02 |
Pos logit contributions | |
hill | +0.210 |
mountain | +0.197 |
subway | +0.188 |
taxi | +0.188 |
bike | +0.188 |
Neg logit contributions | |
amed | -0.258 |
aming | -0.242 |
ady | -0.240 |
ished | -0.239 |
lling | -0.233 |
Token | backyard |
Token ID | 24296 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
hill | +0.066 |
mountain | +0.061 |
cone | +0.058 |
Zig | +0.057 |
subway | +0.057 |
Neg logit contributions | |
amed | -0.134 |
aming | -0.125 |
ady | -0.122 |
ished | -0.120 |
lling | -0.113 |
Token | . |
Token ID | 13 |
Feature activation | +4.15e-02 |
Pos logit contributions | |
hill | +0.335 |
mountain | +0.312 |
road | +0.290 |
subway | +0.284 |
cliff | +0.283 |
Neg logit contributions | |
amed | -0.439 |
ished | -0.420 |
aming | -0.399 |
ady | -0.398 |
lling | -0.395 |
Token | Every |
Token ID | 3887 |
Feature activation | -5.35e-03 |
Pos logit contributions | |
hill | +0.084 |
mountain | +0.079 |
cliff | +0.077 |
subway | +0.074 |
taxi | +0.073 |
Neg logit contributions | |
amed | -0.086 |
ished | -0.083 |
ady | -0.082 |
aming | -0.079 |
ealing | -0.074 |
Token | day |
Token ID | 1110 |
Feature activation | +9.60e-02 |
Pos logit contributions | |
amed | +0.013 |
ished | +0.012 |
ady | +0.012 |
aming | +0.012 |
wed | +0.011 |
Neg logit contributions | |
hill | -0.008 |
mountain | -0.008 |
bike | -0.007 |
bicycle | -0.007 |
Zig | -0.007 |
Token | , |
Token ID | 11 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
hill | +0.169 |
mountain | +0.149 |
cliff | +0.144 |
road | +0.142 |
cone | +0.141 |
Neg logit contributions | |
amed | -0.225 |
ished | -0.216 |
ealing | -0.211 |
aming | -0.207 |
lling | -0.196 |
Token | the |
Token ID | 262 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
hill | +0.197 |
cone | +0.182 |
mountain | +0.180 |
road | +0.173 |
cliff | +0.172 |
Neg logit contributions | |
amed | -0.210 |
aming | -0.197 |
ished | -0.197 |
lling | -0.189 |
sure | -0.187 |
Token | She |
Token ID | 1375 |
Feature activation | +3.11e-02 |
Pos logit contributions | |
hill | +0.073 |
mountain | +0.069 |
road | +0.063 |
cliff | +0.062 |
subway | +0.062 |
Neg logit contributions | |
amed | -0.106 |
ished | -0.102 |
aming | -0.097 |
ady | -0.097 |
lling | -0.095 |
Token | lived |
Token ID | 5615 |
Feature activation | -1.74e-02 |
Pos logit contributions | |
hill | +0.080 |
mountain | +0.076 |
cliff | +0.071 |
ladder | +0.071 |
road | +0.071 |
Neg logit contributions | |
amed | -0.058 |
ished | -0.053 |
aming | -0.051 |
ady | -0.049 |
lling | -0.049 |
Token | in |
Token ID | 287 |
Feature activation | +9.49e-02 |
Pos logit contributions | |
amed | +0.045 |
ished | +0.042 |
aming | +0.041 |
ady | +0.040 |
lling | +0.039 |
Neg logit contributions | |
hill | -0.032 |
mountain | -0.030 |
road | -0.028 |
subway | -0.028 |
cliff | -0.027 |
Token | a |
Token ID | 257 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
hill | +0.215 |
mountain | +0.203 |
cliff | +0.189 |
ladder | +0.188 |
road | +0.186 |
Neg logit contributions | |
amed | -0.201 |
aming | -0.186 |
ished | -0.180 |
ady | -0.179 |
lling | -0.174 |
Token | small |
Token ID | 1402 |
Feature activation | +4.91e-02 |
Pos logit contributions | |
hill | +0.214 |
cone | +0.209 |
mountain | +0.206 |
Zig | +0.206 |
subway | +0.198 |
Neg logit contributions | |
amed | -0.202 |
aming | -0.192 |
ady | -0.187 |
ished | -0.176 |
ealing | -0.175 |
Token | hut |
Token ID | 40812 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
cone | +0.088 |
Zig | +0.086 |
hill | +0.081 |
subway | +0.080 |
mountain | +0.080 |
Neg logit contributions | |
ady | -0.102 |
amed | -0.100 |
ished | -0.097 |
aming | -0.097 |
arre | -0.092 |
Token | with |
Token ID | 351 |
Feature activation | +2.15e-02 |
Pos logit contributions | |
hill | +0.359 |
mountain | +0.345 |
subway | +0.334 |
taxi | +0.321 |
road | +0.320 |
Neg logit contributions | |
amed | -0.396 |
aming | -0.366 |
ished | -0.364 |
ady | -0.348 |
asing | -0.340 |
Token | her |
Token ID | 607 |
Feature activation | -2.99e-03 |
Pos logit contributions | |
hill | +0.045 |
mountain | +0.042 |
road | +0.039 |
cliff | +0.039 |
ladder | +0.038 |
Neg logit contributions | |
amed | -0.049 |
ished | -0.047 |
aming | -0.045 |
ady | -0.044 |
lling | -0.044 |
Token | mom |
Token ID | 1995 |
Feature activation | -3.39e-03 |
Pos logit contributions | |
amed | +0.005 |
ady | +0.004 |
ished | +0.004 |
aming | +0.004 |
sisters | +0.004 |
Neg logit contributions | |
hill | -0.007 |
mountain | -0.007 |
cliff | -0.007 |
road | -0.006 |
subway | -0.006 |
Token | my |
Token ID | 1820 |
Feature activation | -3.87e-02 |
Pos logit contributions | |
amed | +0.006 |
ished | +0.006 |
lling | +0.006 |
ealing | +0.006 |
asing | +0.006 |
Neg logit contributions | |
hill | -0.009 |
mountain | -0.008 |
road | -0.008 |
taxi | -0.008 |
bike | -0.007 |
Token | and |
Token ID | 290 |
Feature activation | +6.92e-02 |
Pos logit contributions | |
amed | +0.093 |
ished | +0.089 |
aming | +0.083 |
lling | +0.083 |
ady | +0.083 |
Neg logit contributions | |
hill | -0.077 |
mountain | -0.072 |
road | -0.067 |
taxi | -0.065 |
ladder | -0.065 |
Token | a |
Token ID | 257 |
Feature activation | +8.29e-02 |
Pos logit contributions | |
amed | +0.049 |
ished | +0.048 |
lling | +0.046 |
ady | +0.045 |
ealing | +0.045 |
Neg logit contributions | |
hill | -0.034 |
mountain | -0.032 |
road | -0.029 |
ladder | -0.029 |
taxi | -0.029 |
Token | hidden |
Token ID | 7104 |
Feature activation | +3.73e-02 |
Pos logit contributions | |
Zig | +0.157 |
hill | +0.153 |
mountain | +0.147 |
road | +0.141 |
cliff | +0.137 |
Neg logit contributions | |
amed | -0.145 |
ady | -0.136 |
aming | -0.131 |
ealing | -0.130 |
ished | -0.126 |
Token | box |
Token ID | 3091 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
hill | +0.056 |
Zig | +0.054 |
mountain | +0.050 |
Raven | +0.050 |
subway | +0.049 |
Neg logit contributions | |
amed | -0.087 |
ished | -0.087 |
ady | -0.081 |
aming | -0.079 |
ac | -0.075 |
Token | under |
Token ID | 739 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
hill | +0.242 |
mountain | +0.225 |
road | +0.209 |
ladder | +0.204 |
cliff | +0.204 |
Neg logit contributions | |
amed | -0.328 |
ished | -0.315 |
aming | -0.299 |
ady | -0.298 |
lling | -0.296 |
Token | a |
Token ID | 257 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
hill | +0.251 |
mountain | +0.238 |
road | +0.222 |
subway | +0.221 |
volcano | +0.221 |
Neg logit contributions | |
amed | -0.216 |
aming | -0.194 |
ished | -0.186 |
ady | -0.183 |
lling | -0.178 |
Token | tree |
Token ID | 5509 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
Zig | +0.160 |
Raven | +0.146 |
Hipp | +0.140 |
hill | +0.140 |
subway | +0.136 |
Neg logit contributions | |
amed | -0.264 |
ady | -0.259 |
aming | -0.256 |
lling | -0.238 |
ished | -0.237 |
Token | . |
Token ID | 13 |
Feature activation | +3.26e-02 |
Pos logit contributions | |
hill | +0.385 |
mountain | +0.375 |
subway | +0.363 |
taxi | +0.353 |
volcano | +0.348 |
Neg logit contributions | |
amed | -0.360 |
ished | -0.341 |
aming | -0.329 |
ady | -0.328 |
lling | -0.318 |
Token | Sue |
Token ID | 26089 |
Feature activation | -5.29e-02 |
Pos logit contributions | |
hill | +0.069 |
mountain | +0.068 |
subway | +0.063 |
volcano | +0.062 |
festival | +0.061 |
Neg logit contributions | |
amed | -0.069 |
ished | -0.065 |
ealing | -0.060 |
ady | -0.059 |
aming | -0.058 |
Token | was |
Token ID | 373 |
Feature activation | -1.91e-02 |
Pos logit contributions | |
amed | +0.129 |
ished | +0.122 |
lling | +0.119 |
ealing | +0.118 |
aming | +0.118 |
Neg logit contributions | |
hill | -0.102 |
mountain | -0.095 |
road | -0.088 |
ladder | -0.087 |
bike | -0.085 |
Token | very |
Token ID | 845 |
Feature activation | +5.21e-02 |
Pos logit contributions | |
amed | +0.047 |
ished | +0.047 |
lling | +0.044 |
ady | +0.044 |
asing | +0.044 |
Neg logit contributions | |
hill | -0.035 |
mountain | -0.032 |
road | -0.031 |
ladder | -0.030 |
taxi | -0.030 |
Token | excited |
Token ID | 6568 |
Feature activation | -4.95e-02 |
Pos logit contributions | |
hill | +0.115 |
mountain | +0.108 |
ladder | +0.107 |
road | +0.103 |
taxi | +0.102 |
Neg logit contributions | |
amed | -0.108 |
ished | -0.106 |
lling | -0.101 |
asing | -0.100 |
eness | -0.098 |
Token | pile |
Token ID | 14540 |
Feature activation | +9.67e-02 |
Pos logit contributions | |
Zig | +0.077 |
hill | +0.073 |
Raven | +0.069 |
mountain | +0.068 |
Under | +0.068 |
Neg logit contributions | |
ealing | -0.128 |
ished | -0.128 |
amed | -0.126 |
aming | -0.124 |
ady | -0.124 |
Token | of |
Token ID | 286 |
Feature activation | +8.10e-02 |
Pos logit contributions | |
hill | +0.192 |
mountain | +0.177 |
road | +0.166 |
ladder | +0.164 |
taxi | +0.162 |
Neg logit contributions | |
amed | -0.231 |
ished | -0.221 |
ealing | -0.210 |
ady | -0.209 |
lling | -0.208 |
Token | hay |
Token ID | 27678 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
hill | +0.127 |
mountain | +0.117 |
cliff | +0.107 |
road | +0.105 |
taxi | +0.104 |
Neg logit contributions | |
amed | -0.216 |
ished | -0.211 |
ady | -0.201 |
aming | -0.200 |
lling | -0.199 |
Token | near |
Token ID | 1474 |
Feature activation | +6.50e-02 |
Pos logit contributions | |
hill | +0.298 |
mountain | +0.290 |
taxi | +0.272 |
subway | +0.270 |
road | +0.267 |
Neg logit contributions | |
amed | -0.340 |
ished | -0.320 |
aming | -0.309 |
lling | -0.302 |
ady | -0.300 |
Token | the |
Token ID | 262 |
Feature activation | +3.03e-02 |
Pos logit contributions | |
hill | +0.108 |
mountain | +0.098 |
road | +0.093 |
ladder | +0.089 |
taxi | +0.089 |
Neg logit contributions | |
amed | -0.177 |
ished | -0.173 |
asing | -0.163 |
ealing | -0.162 |
ady | -0.162 |
Token | barn |
Token ID | 25203 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
hill | +0.035 |
mountain | +0.033 |
taxi | +0.032 |
Zig | +0.031 |
bicycle | +0.031 |
Neg logit contributions | |
amed | -0.085 |
aming | -0.080 |
ished | -0.079 |
ady | -0.078 |
lling | -0.076 |
Token | . |
Token ID | 13 |
Feature activation | +1.72e-02 |
Pos logit contributions | |
hill | +0.360 |
mountain | +0.339 |
road | +0.315 |
subway | +0.311 |
taxi | +0.310 |
Neg logit contributions | |
amed | -0.410 |
ished | -0.391 |
aming | -0.370 |
ady | -0.368 |
lling | -0.366 |
Token | They |
Token ID | 1119 |
Feature activation | +1.50e-02 |
Pos logit contributions | |
hill | +0.034 |
mountain | +0.034 |
cliff | +0.030 |
bike | +0.030 |
festival | +0.030 |
Neg logit contributions | |
ished | -0.038 |
amed | -0.038 |
ealing | -0.034 |
ady | -0.034 |
lee | -0.033 |
Token | ran |
Token ID | 4966 |
Feature activation | -6.68e-02 |
Pos logit contributions | |
hill | +0.032 |
mountain | +0.030 |
road | +0.028 |
ladder | +0.028 |
taxi | +0.027 |
Neg logit contributions | |
amed | -0.034 |
ished | -0.033 |
aming | -0.031 |
lling | -0.030 |
ady | -0.030 |
Token | to |
Token ID | 284 |
Feature activation | +7.70e-02 |
Pos logit contributions | |
amed | +0.172 |
ished | +0.162 |
lling | +0.150 |
aming | +0.150 |
ealing | +0.143 |
Neg logit contributions | |
hill | -0.117 |
mountain | -0.115 |
chim | -0.104 |
cliff | -0.101 |
volcano | -0.101 |
Token | it |
Token ID | 340 |
Feature activation | +7.01e-03 |
Pos logit contributions | |
hill | +0.139 |
mountain | +0.133 |
road | +0.121 |
subway | +0.120 |
cliff | +0.119 |
Neg logit contributions | |
amed | -0.189 |
ished | -0.183 |
lling | -0.172 |
aming | -0.172 |
ady | -0.171 |
Token | a |
Token ID | 257 |
Feature activation | +7.44e-02 |
Pos logit contributions | |
amed | +0.032 |
ished | +0.030 |
aming | +0.029 |
ady | +0.028 |
lling | +0.027 |
Neg logit contributions | |
hill | -0.037 |
mountain | -0.035 |
road | -0.033 |
cliff | -0.032 |
ladder | -0.032 |
Token | tea |
Token ID | 8887 |
Feature activation | +8.91e-02 |
Pos logit contributions | |
hill | +0.106 |
mountain | +0.097 |
stack | +0.096 |
Zig | +0.093 |
Raven | +0.092 |
Neg logit contributions | |
sure | -0.161 |
ealing | -0.153 |
ok | -0.152 |
ished | -0.151 |
forgiven | -0.146 |
Token | party |
Token ID | 2151 |
Feature activation | +8.54e-02 |
Pos logit contributions | |
hill | +0.147 |
mountain | +0.139 |
cliff | +0.130 |
road | +0.126 |
ladder | +0.125 |
Neg logit contributions | |
amed | -0.225 |
ished | -0.211 |
aming | -0.205 |
ady | -0.201 |
asing | -0.197 |
Token | in |
Token ID | 287 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
hill | +0.171 |
mountain | +0.161 |
cliff | +0.148 |
road | +0.148 |
ladder | +0.146 |
Neg logit contributions | |
amed | -0.203 |
ished | -0.191 |
aming | -0.184 |
ady | -0.181 |
lling | -0.180 |
Token | their |
Token ID | 511 |
Feature activation | +4.76e-02 |
Pos logit contributions | |
hill | +0.181 |
mountain | +0.168 |
road | +0.155 |
ladder | +0.152 |
cliff | +0.152 |
Neg logit contributions | |
amed | -0.266 |
ished | -0.255 |
aming | -0.244 |
ady | -0.243 |
lling | -0.241 |
Token | hut |
Token ID | 40812 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
hill | +0.088 |
mountain | +0.083 |
cone | +0.082 |
cliff | +0.080 |
Raven | +0.078 |
Neg logit contributions | |
amed | -0.105 |
ished | -0.102 |
ady | -0.098 |
aming | -0.094 |
ealing | -0.093 |
Token | . |
Token ID | 13 |
Feature activation | +2.30e-02 |
Pos logit contributions | |
hill | +0.374 |
mountain | +0.357 |
subway | +0.345 |
taxi | +0.336 |
cliff | +0.335 |
Neg logit contributions | |
amed | -0.366 |
ished | -0.346 |
ady | -0.332 |
aming | -0.331 |
lling | -0.321 |
Token |
|
Token ID | 198 |
Feature activation | +7.06e-03 |
Pos logit contributions | |
hill | +0.046 |
mountain | +0.044 |
cliff | +0.040 |
road | +0.039 |
subway | +0.039 |
Neg logit contributions | |
amed | -0.054 |
ished | -0.052 |
ady | -0.049 |
ealing | -0.048 |
aming | -0.048 |
Token |
|
Token ID | 198 |
Feature activation | +7.87e-03 |
Pos logit contributions | |
hill | +0.012 |
mountain | +0.011 |
road | +0.010 |
ladder | +0.010 |
subway | +0.010 |
Neg logit contributions | |
amed | -0.019 |
ished | -0.018 |
aming | -0.017 |
ady | -0.017 |
lling | -0.017 |
Token | They |
Token ID | 2990 |
Feature activation | +1.09e-02 |
Pos logit contributions | |
hill | +0.015 |
mountain | +0.014 |
road | +0.014 |
taxi | +0.013 |
subway | +0.013 |
Neg logit contributions | |
amed | -0.019 |
ished | -0.018 |
lling | -0.017 |
aming | -0.017 |
ady | -0.017 |
Token | set |
Token ID | 900 |
Feature activation | -3.29e-02 |
Pos logit contributions | |
hill | +0.022 |
mountain | +0.021 |
road | +0.020 |
taxi | +0.019 |
subway | +0.019 |
Neg logit contributions | |
amed | -0.025 |
ished | -0.024 |
aming | -0.023 |
ady | -0.023 |
lling | -0.023 |
Token | big |
Token ID | 1263 |
Feature activation | +6.10e-02 |
Pos logit contributions | |
Zig | +0.139 |
hill | +0.136 |
mountain | +0.126 |
road | +0.123 |
Raven | +0.122 |
Neg logit contributions | |
amed | -0.153 |
ealing | -0.153 |
aming | -0.144 |
lling | -0.143 |
ady | -0.143 |
Token | red |
Token ID | 2266 |
Feature activation | +3.40e-02 |
Pos logit contributions | |
Zig | +0.075 |
hill | +0.071 |
Raven | +0.066 |
Under | +0.065 |
mountain | +0.064 |
Neg logit contributions | |
amed | -0.140 |
ealing | -0.140 |
ished | -0.139 |
ady | -0.137 |
aming | -0.136 |
Token | ball |
Token ID | 2613 |
Feature activation | +9.53e-02 |
Pos logit contributions | |
Zig | +0.049 |
Raven | +0.043 |
Sky | +0.042 |
hill | +0.041 |
mountain | +0.039 |
Neg logit contributions | |
amed | -0.072 |
ished | -0.069 |
ady | -0.069 |
aming | -0.067 |
bling | -0.065 |
Token | near |
Token ID | 1474 |
Feature activation | +6.82e-02 |
Pos logit contributions | |
hill | +0.185 |
mountain | +0.178 |
subway | +0.164 |
road | +0.163 |
taxi | +0.162 |
Neg logit contributions | |
amed | -0.218 |
ished | -0.211 |
aming | -0.200 |
lling | -0.196 |
ady | -0.195 |
Token | the |
Token ID | 262 |
Feature activation | +3.11e-02 |
Pos logit contributions | |
hill | +0.126 |
mountain | +0.117 |
road | +0.107 |
ladder | +0.106 |
cliff | +0.106 |
Neg logit contributions | |
amed | -0.173 |
ished | -0.165 |
aming | -0.158 |
ady | -0.157 |
lling | -0.156 |
Token | barn |
Token ID | 25203 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
hill | +0.034 |
mountain | +0.031 |
Zig | +0.031 |
taxi | +0.030 |
bicycle | +0.030 |
Neg logit contributions | |
amed | -0.088 |
aming | -0.082 |
ady | -0.080 |
ished | -0.079 |
lling | -0.079 |
Token | . |
Token ID | 13 |
Feature activation | +2.02e-02 |
Pos logit contributions | |
hill | +0.368 |
mountain | +0.349 |
subway | +0.324 |
road | +0.322 |
taxi | +0.322 |
Neg logit contributions | |
amed | -0.395 |
ished | -0.375 |
aming | -0.355 |
ady | -0.353 |
lling | -0.351 |
Token | He |
Token ID | 679 |
Feature activation | +2.23e-02 |
Pos logit contributions | |
hill | +0.038 |
mountain | +0.036 |
cliff | +0.034 |
subway | +0.033 |
taxi | +0.033 |
Neg logit contributions | |
amed | -0.048 |
ished | -0.046 |
aming | -0.043 |
ealing | -0.043 |
ady | -0.043 |
Token | wanted |
Token ID | 2227 |
Feature activation | -7.06e-02 |
Pos logit contributions | |
hill | +0.050 |
mountain | +0.047 |
road | +0.045 |
ladder | +0.044 |
cliff | +0.044 |
Neg logit contributions | |
amed | -0.047 |
ished | -0.045 |
aming | -0.042 |
ady | -0.042 |
lling | -0.042 |
Token | to |
Token ID | 284 |
Feature activation | +8.39e-02 |
Pos logit contributions | |
amed | +0.183 |
ished | +0.182 |
aming | +0.167 |
lling | +0.167 |
asing | +0.167 |
Neg logit contributions | |
hill | -0.115 |
mountain | -0.108 |
valley | -0.098 |
cliff | -0.097 |
subway | -0.096 |
Token | play |
Token ID | 711 |
Feature activation | -5.78e-03 |
Pos logit contributions | |
hill | +0.144 |
mountain | +0.140 |
subway | +0.130 |
volcano | +0.126 |
chim | +0.125 |
Neg logit contributions | |
amed | -0.204 |
ished | -0.200 |
ady | -0.186 |
aming | -0.185 |
lling | -0.183 |
Token | love |
Token ID | 1842 |
Feature activation | -6.72e-02 |
Pos logit contributions | |
amed | +0.017 |
ished | +0.017 |
aming | +0.016 |
ady | +0.015 |
lling | +0.015 |
Neg logit contributions | |
hill | -0.013 |
mountain | -0.012 |
road | -0.011 |
ladder | -0.011 |
cliff | -0.011 |
Token | it |
Token ID | 340 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
amed | +0.173 |
ished | +0.168 |
aming | +0.160 |
ady | +0.160 |
lling | +0.160 |
Neg logit contributions | |
hill | -0.116 |
mountain | -0.107 |
road | -0.100 |
taxi | -0.098 |
subway | -0.098 |
Token | !" |
Token ID | 2474 |
Feature activation | -5.32e-03 |
Pos logit contributions | |
amed | +0.029 |
ished | +0.027 |
aming | +0.026 |
ady | +0.026 |
lling | +0.026 |
Neg logit contributions | |
hill | -0.022 |
mountain | -0.021 |
road | -0.019 |
ladder | -0.019 |
taxi | -0.018 |
Token |
|
Token ID | 198 |
Feature activation | -8.86e-03 |
Pos logit contributions | |
amed | +0.014 |
ished | +0.013 |
aming | +0.013 |
ady | +0.013 |
lling | +0.013 |
Neg logit contributions | |
hill | -0.009 |
mountain | -0.009 |
road | -0.008 |
ladder | -0.008 |
cliff | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | -4.20e-03 |
Pos logit contributions | |
amed | +0.026 |
ished | +0.025 |
lling | +0.024 |
asing | +0.024 |
aming | +0.023 |
Neg logit contributions | |
hill | -0.013 |
mountain | -0.012 |
road | -0.011 |
taxi | -0.011 |
ladder | -0.011 |
Token | Isa |
Token ID | 39443 |
Feature activation | -2.24e-01 |
Pos logit contributions | |
amed | +0.010 |
ished | +0.010 |
lling | +0.009 |
aming | +0.009 |
asing | +0.009 |
Neg logit contributions | |
hill | -0.008 |
mountain | -0.007 |
road | -0.007 |
taxi | -0.007 |
cliff | -0.007 |
Token | ac |
Token ID | 330 |
Feature activation | -9.56e-02 |
Pos logit contributions | |
lling | +0.316 |
amed | +0.314 |
ished | +0.296 |
aming | +0.277 |
bling | +0.271 |
Neg logit contributions | |
hill | -0.548 |
mountain | -0.547 |
volcano | -0.516 |
pine | -0.512 |
road | -0.509 |
Token | gul |
Token ID | 47161 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
lling | +0.147 |
bling | +0.141 |
amed | +0.141 |
ealing | +0.141 |
aming | +0.139 |
Neg logit contributions | |
hill | -0.212 |
je | -0.207 |
pine | -0.204 |
mountain | -0.202 |
taxi | -0.197 |
Token | ped |
Token ID | 9124 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
asing | +0.233 |
lling | +0.226 |
mindful | +0.219 |
aming | +0.209 |
alright | +0.208 |
Neg logit contributions | |
taxi | -0.322 |
hill | -0.314 |
mountain | -0.313 |
bamboo | -0.312 |
je | -0.311 |
Token | and |
Token ID | 290 |
Feature activation | +4.13e-02 |
Pos logit contributions | |
amed | +0.305 |
ished | +0.300 |
asing | +0.295 |
ady | +0.290 |
lling | +0.285 |
Neg logit contributions | |
hill | -0.193 |
mountain | -0.178 |
taxi | -0.170 |
ladder | -0.165 |
bike | -0.164 |
Token | said |
Token ID | 531 |
Feature activation | -9.49e-02 |
Pos logit contributions | |
hill | +0.091 |
mountain | +0.086 |
ladder | +0.079 |
taxi | +0.078 |
cliff | +0.078 |
Neg logit contributions | |
amed | -0.090 |
ished | -0.086 |
ady | -0.082 |
asing | -0.080 |
aming | -0.080 |
Token | " |
Token ID | 1 |
Feature activation | -2.23e-02 |
Pos logit contributions | |
hill | +0.003 |
mountain | +0.002 |
road | +0.002 |
ladder | +0.002 |
taxi | +0.002 |
Neg logit contributions | |
amed | -0.003 |
ished | -0.003 |
lling | -0.003 |
asing | -0.003 |
ady | -0.003 |
Token | I |
Token ID | 40 |
Feature activation | -2.86e-02 |
Pos logit contributions | |
amed | +0.053 |
ished | +0.048 |
aming | +0.048 |
ady | +0.047 |
asing | +0.046 |
Neg logit contributions | |
hill | -0.045 |
mountain | -0.044 |
road | -0.040 |
cliff | -0.040 |
volcano | -0.039 |
Token | want |
Token ID | 765 |
Feature activation | -5.93e-02 |
Pos logit contributions | |
amed | +0.064 |
ished | +0.059 |
lling | +0.057 |
aming | +0.057 |
ady | +0.055 |
Neg logit contributions | |
hill | -0.062 |
mountain | -0.058 |
road | -0.055 |
ladder | -0.055 |
taxi | -0.054 |
Token | steak |
Token ID | 26320 |
Feature activation | +6.42e-02 |
Pos logit contributions | |
ished | +0.134 |
amed | +0.133 |
ady | +0.123 |
aming | +0.123 |
lling | +0.121 |
Neg logit contributions | |
hill | -0.116 |
mountain | -0.105 |
road | -0.101 |
subway | -0.099 |
taxi | -0.098 |
Token | . |
Token ID | 13 |
Feature activation | +1.81e-02 |
Pos logit contributions | |
hill | +0.126 |
mountain | +0.118 |
road | +0.111 |
subway | +0.110 |
cliff | +0.109 |
Neg logit contributions | |
amed | -0.151 |
ished | -0.145 |
ady | -0.136 |
aming | -0.136 |
ealing | -0.134 |
Token | Ste |
Token ID | 2441 |
Feature activation | -2.22e-01 |
Pos logit contributions | |
hill | +0.033 |
mountain | +0.031 |
cliff | +0.029 |
subway | +0.029 |
road | +0.029 |
Neg logit contributions | |
amed | -0.043 |
ished | -0.043 |
ady | -0.041 |
ealing | -0.040 |
aming | -0.040 |
Token | ak |
Token ID | 461 |
Feature activation | +1.19e-02 |
Pos logit contributions | |
lling | +0.372 |
amed | +0.366 |
ished | +0.337 |
aming | +0.334 |
bled | +0.332 |
Neg logit contributions | |
hill | -0.485 |
mountain | -0.453 |
road | -0.451 |
volcano | -0.447 |
je | -0.438 |
Token | is |
Token ID | 318 |
Feature activation | -3.92e-03 |
Pos logit contributions | |
hill | +0.023 |
mountain | +0.021 |
ladder | +0.019 |
road | +0.019 |
cliff | +0.019 |
Neg logit contributions | |
amed | -0.030 |
ished | -0.028 |
aming | -0.027 |
ady | -0.027 |
lling | -0.027 |
Token | good |
Token ID | 922 |
Feature activation | -3.46e-02 |
Pos logit contributions | |
amed | +0.010 |
ished | +0.009 |
aming | +0.009 |
ady | +0.009 |
lling | +0.009 |
Neg logit contributions | |
hill | -0.007 |
mountain | -0.007 |
road | -0.006 |
subway | -0.006 |
ladder | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | +1.78e-02 |
Pos logit contributions | |
amed | +0.086 |
ished | +0.083 |
aming | +0.079 |
ady | +0.078 |
lling | +0.078 |
Neg logit contributions | |
hill | -0.065 |
mountain | -0.060 |
road | -0.056 |
ladder | -0.055 |
taxi | -0.055 |
Token | It |
Token ID | 632 |
Feature activation | +4.54e-02 |
Pos logit contributions | |
hill | +0.031 |
mountain | +0.029 |
cliff | +0.027 |
road | +0.026 |
subway | +0.026 |
Neg logit contributions | |
amed | -0.045 |
ished | -0.044 |
ady | -0.042 |
aming | -0.042 |
ealing | -0.042 |
Token | when |
Token ID | 618 |
Feature activation | +5.93e-02 |
Pos logit contributions | |
amed | +0.128 |
ished | +0.123 |
aming | +0.116 |
ady | +0.116 |
lling | +0.115 |
Neg logit contributions | |
hill | -0.108 |
mountain | -0.101 |
road | -0.094 |
ladder | -0.092 |
subway | -0.092 |
Token | I |
Token ID | 314 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
hill | +0.085 |
mountain | +0.078 |
taxi | +0.072 |
road | +0.072 |
subway | +0.071 |
Neg logit contributions | |
amed | -0.161 |
ished | -0.157 |
aming | -0.151 |
lling | -0.151 |
ady | -0.149 |
Token | 'm |
Token ID | 1101 |
Feature activation | -3.89e-02 |
Pos logit contributions | |
amed | +0.045 |
ished | +0.042 |
aming | +0.041 |
lling | +0.041 |
ady | +0.040 |
Neg logit contributions | |
hill | -0.031 |
mountain | -0.028 |
road | -0.026 |
ladder | -0.025 |
taxi | -0.025 |
Token | elderly |
Token ID | 13830 |
Feature activation | -5.30e-02 |
Pos logit contributions | |
amed | +0.092 |
ished | +0.089 |
ady | +0.084 |
asing | +0.083 |
lling | +0.083 |
Neg logit contributions | |
hill | -0.075 |
mountain | -0.070 |
taxi | -0.066 |
road | -0.066 |
ladder | -0.066 |
Token | ." |
Token ID | 526 |
Feature activation | -2.07e-02 |
Pos logit contributions | |
amed | +0.124 |
ished | +0.120 |
aming | +0.113 |
ady | +0.112 |
lling | +0.112 |
Neg logit contributions | |
hill | -0.105 |
mountain | -0.098 |
road | -0.092 |
subway | -0.090 |
cliff | -0.090 |
Token | Gr |
Token ID | 1902 |
Feature activation | -2.18e-01 |
Pos logit contributions | |
amed | +0.054 |
ished | +0.052 |
aming | +0.049 |
lling | +0.049 |
ady | +0.049 |
Neg logit contributions | |
hill | -0.037 |
mountain | -0.034 |
road | -0.032 |
ladder | -0.031 |
taxi | -0.031 |
Token | anny |
Token ID | 7737 |
Feature activation | -7.47e-02 |
Pos logit contributions | |
amed | +0.391 |
ished | +0.388 |
ealing | +0.386 |
convinced | +0.374 |
lling | +0.373 |
Neg logit contributions | |
hill | -0.427 |
volcano | -0.400 |
road | -0.396 |
mountain | -0.392 |
chim | -0.388 |
Token | laughed |
Token ID | 13818 |
Feature activation | -1.50e-01 |
Pos logit contributions | |
amed | +0.151 |
ished | +0.142 |
ady | +0.137 |
aming | +0.137 |
lling | +0.135 |
Neg logit contributions | |
hill | -0.168 |
mountain | -0.158 |
taxi | -0.152 |
ladder | -0.147 |
subway | -0.147 |
Token | and |
Token ID | 290 |
Feature activation | +3.97e-02 |
Pos logit contributions | |
amed | +0.420 |
ished | +0.417 |
ady | +0.411 |
asing | +0.411 |
lling | +0.401 |
Neg logit contributions | |
hill | -0.201 |
taxi | -0.183 |
mountain | -0.179 |
bike | -0.176 |
bicycle | -0.168 |
Token | said |
Token ID | 531 |
Feature activation | -9.55e-02 |
Pos logit contributions | |
hill | +0.092 |
mountain | +0.086 |
road | +0.081 |
subway | +0.081 |
taxi | +0.081 |
Neg logit contributions | |
amed | -0.081 |
ished | -0.077 |
ady | -0.074 |
aming | -0.073 |
asing | -0.071 |
Token | , |
Token ID | 11 |
Feature activation | +5.67e-02 |
Pos logit contributions | |
amed | +0.294 |
ished | +0.283 |
ady | +0.279 |
aming | +0.278 |
lling | +0.275 |
Neg logit contributions | |
hill | -0.120 |
mountain | -0.109 |
taxi | -0.100 |
ladder | -0.099 |
bike | -0.095 |
Token | the |
Token ID | 262 |
Feature activation | -3.08e-02 |
Pos logit contributions | |
ished | +0.103 |
amed | +0.102 |
ady | +0.096 |
asing | +0.095 |
wed | +0.094 |
Neg logit contributions | |
hill | -0.082 |
mountain | -0.078 |
cliff | -0.072 |
taxi | -0.071 |
road | -0.071 |
Token | best |
Token ID | 1266 |
Feature activation | -1.72e-02 |
Pos logit contributions | |
amed | +0.049 |
ished | +0.044 |
aming | +0.042 |
ady | +0.042 |
lling | +0.041 |
Neg logit contributions | |
hill | -0.083 |
mountain | -0.079 |
road | -0.076 |
ladder | -0.075 |
cliff | -0.075 |
Token | at |
Token ID | 379 |
Feature activation | +6.65e-02 |
Pos logit contributions | |
amed | +0.035 |
ished | +0.035 |
ady | +0.034 |
aming | +0.033 |
asing | +0.033 |
Neg logit contributions | |
hill | -0.036 |
mountain | -0.033 |
road | -0.032 |
cliff | -0.031 |
subway | -0.031 |
Token | sliding |
Token ID | 22292 |
Feature activation | -2.44e-02 |
Pos logit contributions | |
hill | +0.134 |
mountain | +0.128 |
subway | +0.121 |
road | +0.121 |
taxi | +0.121 |
Neg logit contributions | |
amed | -0.143 |
ished | -0.135 |
lling | -0.130 |
aming | -0.128 |
asing | -0.126 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.39e-02 |
Pos logit contributions | |
amed | +0.056 |
ished | +0.055 |
aming | +0.050 |
lling | +0.050 |
ealing | +0.049 |
Neg logit contributions | |
hill | -0.046 |
mountain | -0.045 |
subway | -0.041 |
road | -0.041 |
taxi | -0.041 |
Token | Gr |
Token ID | 1902 |
Feature activation | -2.16e-01 |
Pos logit contributions | |
amed | +0.034 |
ished | +0.033 |
aming | +0.031 |
ady | +0.031 |
lling | +0.031 |
Neg logit contributions | |
hill | -0.026 |
mountain | -0.025 |
road | -0.023 |
cliff | -0.022 |
ladder | -0.022 |
Token | anny |
Token ID | 7737 |
Feature activation | -7.51e-02 |
Pos logit contributions | |
ished | +0.389 |
ealing | +0.388 |
amed | +0.387 |
ok | +0.373 |
bled | +0.368 |
Neg logit contributions | |
hill | -0.410 |
road | -0.385 |
volcano | -0.382 |
mountain | -0.375 |
chim | -0.367 |
Token | smiled |
Token ID | 13541 |
Feature activation | -1.50e-01 |
Pos logit contributions | |
amed | +0.153 |
ished | +0.145 |
ady | +0.141 |
aming | +0.139 |
lling | +0.137 |
Neg logit contributions | |
hill | -0.164 |
mountain | -0.154 |
taxi | -0.149 |
subway | -0.144 |
ladder | -0.143 |
Token | and |
Token ID | 290 |
Feature activation | +3.75e-02 |
Pos logit contributions | |
amed | +0.418 |
ished | +0.417 |
asing | +0.412 |
lling | +0.400 |
wed | +0.399 |
Neg logit contributions | |
hill | -0.214 |
mountain | -0.195 |
taxi | -0.192 |
bike | -0.186 |
road | -0.178 |
Token | said |
Token ID | 531 |
Feature activation | -9.49e-02 |
Pos logit contributions | |
hill | +0.081 |
mountain | +0.076 |
taxi | +0.072 |
subway | +0.071 |
bike | +0.071 |
Neg logit contributions | |
amed | -0.080 |
ished | -0.077 |
ady | -0.074 |
aming | -0.072 |
asing | -0.072 |
Token | , |
Token ID | 11 |
Feature activation | +5.65e-02 |
Pos logit contributions | |
amed | +0.292 |
ished | +0.282 |
ady | +0.278 |
aming | +0.276 |
lling | +0.273 |
Neg logit contributions | |
hill | -0.118 |
mountain | -0.107 |
taxi | -0.100 |
ladder | -0.098 |
subway | -0.094 |
Token | the |
Token ID | 262 |
Feature activation | -3.17e-02 |
Pos logit contributions | |
hill | +0.122 |
mountain | +0.115 |
road | +0.108 |
cliff | +0.106 |
ladder | +0.106 |
Neg logit contributions | |
amed | -0.116 |
ished | -0.110 |
aming | -0.103 |
lling | -0.103 |
ady | -0.102 |
Token | owner |
Token ID | 4870 |
Feature activation | -7.85e-04 |
Pos logit contributions | |
amed | +0.106 |
ished | +0.102 |
aming | +0.099 |
ady | +0.099 |
lling | +0.098 |
Neg logit contributions | |
hill | -0.032 |
mountain | -0.029 |
road | -0.024 |
cliff | -0.023 |
ladder | -0.023 |
Token | smiled |
Token ID | 13541 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
amed | +0.002 |
ished | +0.001 |
ady | +0.001 |
lling | +0.001 |
aming | +0.001 |
Neg logit contributions | |
hill | -0.002 |
mountain | -0.002 |
taxi | -0.002 |
road | -0.002 |
ladder | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +1.06e-03 |
Pos logit contributions | |
amed | +0.390 |
ished | +0.378 |
asing | +0.376 |
ady | +0.372 |
lling | +0.368 |
Neg logit contributions | |
hill | -0.200 |
mountain | -0.184 |
taxi | -0.177 |
bike | -0.173 |
bicycle | -0.166 |
Token | " |
Token ID | 366 |
Feature activation | -1.45e-02 |
Pos logit contributions | |
hill | +0.002 |
mountain | +0.002 |
road | +0.002 |
ladder | +0.002 |
taxi | +0.002 |
Neg logit contributions | |
amed | -0.003 |
ished | -0.003 |
aming | -0.003 |
lling | -0.003 |
ady | -0.003 |
Token | Looks |
Token ID | 41102 |
Feature activation | -2.14e-01 |
Pos logit contributions | |
amed | +0.033 |
ished | +0.033 |
ady | +0.030 |
ealing | +0.030 |
aming | +0.030 |
Neg logit contributions | |
hill | -0.029 |
mountain | -0.028 |
road | -0.026 |
subway | -0.025 |
taxi | -0.025 |
Token | like |
Token ID | 588 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
amed | +0.610 |
aming | +0.578 |
bled | +0.568 |
ished | +0.568 |
ady | +0.567 |
Neg logit contributions | |
hill | -0.274 |
road | -0.258 |
mountain | -0.244 |
bike | -0.238 |
saf | -0.233 |
Token | you |
Token ID | 345 |
Feature activation | -6.56e-02 |
Pos logit contributions | |
amed | +0.025 |
ished | +0.024 |
asing | +0.023 |
lling | +0.023 |
aming | +0.022 |
Neg logit contributions | |
hill | -0.021 |
mountain | -0.020 |
cliff | -0.018 |
ladder | -0.018 |
road | -0.018 |
Token | two |
Token ID | 734 |
Feature activation | -8.42e-03 |
Pos logit contributions | |
amed | +0.151 |
ished | +0.146 |
aming | +0.138 |
ady | +0.137 |
lling | +0.136 |
Neg logit contributions | |
hill | -0.135 |
mountain | -0.125 |
road | -0.119 |
taxi | -0.116 |
ladder | -0.114 |
Token | are |
Token ID | 389 |
Feature activation | -7.19e-04 |
Pos logit contributions | |
amed | +0.021 |
ished | +0.020 |
aming | +0.019 |
ady | +0.018 |
lling | +0.018 |
Neg logit contributions | |
hill | -0.016 |
mountain | -0.015 |
road | -0.014 |
taxi | -0.014 |
ladder | -0.014 |
Token | friends |
Token ID | 2460 |
Feature activation | +2.76e-03 |
Pos logit contributions | |
amed | +0.002 |
ished | +0.002 |
aming | +0.002 |
lling | +0.002 |
ady | +0.002 |
Neg logit contributions | |
hill | -0.001 |
mountain | -0.001 |
road | -0.001 |
taxi | -0.001 |
ladder | -0.001 |
Token | bit |
Token ID | 1643 |
Feature activation | +1.59e-02 |
Pos logit contributions | |
hill | +0.028 |
mountain | +0.026 |
subway | +0.024 |
road | +0.024 |
taxi | +0.023 |
Neg logit contributions | |
amed | -0.042 |
ished | -0.041 |
ady | -0.038 |
aming | -0.038 |
lling | -0.038 |
Token | of |
Token ID | 286 |
Feature activation | +4.40e-02 |
Pos logit contributions | |
hill | +0.029 |
mountain | +0.027 |
ladder | +0.025 |
road | +0.025 |
cliff | +0.025 |
Neg logit contributions | |
amed | -0.041 |
ished | -0.037 |
aming | -0.037 |
ady | -0.036 |
lling | -0.036 |
Token | magic |
Token ID | 5536 |
Feature activation | +1.04e-02 |
Pos logit contributions | |
hill | +0.099 |
mountain | +0.093 |
road | +0.088 |
ladder | +0.086 |
cliff | +0.086 |
Neg logit contributions | |
amed | -0.095 |
ished | -0.090 |
aming | -0.084 |
ady | -0.084 |
lling | -0.083 |
Token | . |
Token ID | 13 |
Feature activation | +9.81e-03 |
Pos logit contributions | |
hill | +0.020 |
mountain | +0.019 |
road | +0.017 |
subway | +0.017 |
taxi | +0.017 |
Neg logit contributions | |
amed | -0.025 |
ished | -0.024 |
ady | -0.023 |
aming | -0.023 |
lling | -0.023 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -3.41e-02 |
Pos logit contributions | |
hill | +0.017 |
mountain | +0.016 |
road | +0.015 |
subway | +0.015 |
ladder | +0.015 |
Neg logit contributions | |
amed | -0.025 |
ished | -0.024 |
aming | -0.023 |
ady | -0.023 |
lling | -0.023 |
Token | Isa |
Token ID | 39443 |
Feature activation | -2.14e-01 |
Pos logit contributions | |
amed | +0.088 |
ished | +0.084 |
aming | +0.081 |
ady | +0.080 |
lling | +0.079 |
Neg logit contributions | |
hill | -0.059 |
mountain | -0.055 |
road | -0.051 |
taxi | -0.050 |
subway | -0.050 |
Token | ac |
Token ID | 330 |
Feature activation | -8.56e-02 |
Pos logit contributions | |
ished | +0.340 |
amed | +0.339 |
lling | +0.337 |
asing | +0.318 |
bled | +0.310 |
Neg logit contributions | |
hill | -0.497 |
mountain | -0.489 |
road | -0.468 |
saf | -0.468 |
pine | -0.468 |
Token | was |
Token ID | 373 |
Feature activation | -2.56e-02 |
Pos logit contributions | |
amed | +0.158 |
asing | +0.153 |
ealing | +0.148 |
ished | +0.147 |
aming | +0.147 |
Neg logit contributions | |
hill | -0.184 |
mountain | -0.179 |
je | -0.173 |
pine | -0.172 |
taxi | -0.167 |
Token | preparing |
Token ID | 10629 |
Feature activation | -4.03e-02 |
Pos logit contributions | |
ished | +0.065 |
amed | +0.064 |
asing | +0.063 |
lling | +0.062 |
ealing | +0.061 |
Neg logit contributions | |
hill | -0.041 |
mountain | -0.038 |
taxi | -0.038 |
ladder | -0.036 |
road | -0.035 |
Token | breakfast |
Token ID | 12607 |
Feature activation | +6.76e-02 |
Pos logit contributions | |
amed | +0.097 |
ished | +0.093 |
aming | +0.089 |
ady | +0.088 |
lling | +0.087 |
Neg logit contributions | |
hill | -0.078 |
mountain | -0.072 |
road | -0.067 |
cliff | -0.066 |
taxi | -0.066 |
Token | in |
Token ID | 287 |
Feature activation | +8.02e-02 |
Pos logit contributions | |
hill | +0.126 |
mountain | +0.117 |
road | +0.109 |
ladder | +0.106 |
subway | +0.106 |
Neg logit contributions | |
amed | -0.169 |
ished | -0.162 |
aming | -0.154 |
ady | -0.153 |
lling | -0.152 |
Token | mas |
Token ID | 12422 |
Feature activation | -6.68e-02 |
Pos logit contributions | |
amed | +0.182 |
ished | +0.176 |
aming | +0.166 |
asing | +0.166 |
ady | +0.166 |
Neg logit contributions | |
hill | -0.144 |
mountain | -0.134 |
road | -0.126 |
taxi | -0.124 |
subway | -0.122 |
Token | ala |
Token ID | 6081 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
amed | +0.120 |
ished | +0.115 |
lling | +0.106 |
aming | +0.104 |
ady | +0.103 |
Neg logit contributions | |
hill | -0.167 |
mountain | -0.159 |
road | -0.152 |
taxi | -0.150 |
cliff | -0.149 |
Token | and |
Token ID | 290 |
Feature activation | +5.47e-02 |
Pos logit contributions | |
amed | +0.257 |
ished | +0.244 |
aming | +0.231 |
ady | +0.229 |
lling | +0.228 |
Neg logit contributions | |
hill | -0.237 |
mountain | -0.223 |
road | -0.208 |
taxi | -0.205 |
subway | -0.205 |
Token | r |
Token ID | 374 |
Feature activation | -8.86e-02 |
Pos logit contributions | |
hill | +0.109 |
mountain | +0.103 |
road | +0.096 |
ladder | +0.094 |
cliff | +0.093 |
Neg logit contributions | |
amed | -0.129 |
ished | -0.124 |
aming | -0.117 |
ady | -0.116 |
lling | -0.115 |
Token | as |
Token ID | 292 |
Feature activation | -4.05e-02 |
Pos logit contributions | |
ealing | +0.106 |
amed | +0.105 |
eyebrow | +0.102 |
lling | +0.101 |
ished | +0.100 |
Neg logit contributions | |
hill | -0.232 |
mountain | -0.221 |
road | -0.214 |
valley | -0.212 |
taxi | -0.210 |
Token | el |
Token ID | 1288 |
Feature activation | -2.13e-01 |
Pos logit contributions | |
amed | +0.099 |
ished | +0.095 |
aming | +0.090 |
ady | +0.090 |
lling | +0.089 |
Neg logit contributions | |
hill | -0.077 |
mountain | -0.072 |
road | -0.066 |
ladder | -0.065 |
subway | -0.065 |
Token | h |
Token ID | 289 |
Feature activation | -5.39e-02 |
Pos logit contributions | |
amed | +0.215 |
ished | +0.215 |
ealing | +0.193 |
oking | +0.191 |
asing | +0.186 |
Neg logit contributions | |
hill | -0.560 |
mountain | -0.549 |
road | -0.547 |
rh | -0.533 |
cliff | -0.530 |
Token | an |
Token ID | 272 |
Feature activation | -3.67e-02 |
Pos logit contributions | |
amed | +0.106 |
lling | +0.095 |
ished | +0.095 |
aming | +0.092 |
ady | +0.091 |
Neg logit contributions | |
hill | -0.117 |
mountain | -0.115 |
taxi | -0.109 |
subway | -0.108 |
ladder | -0.108 |
Token | out |
Token ID | 448 |
Feature activation | -2.31e-02 |
Pos logit contributions | |
amed | +0.086 |
ished | +0.082 |
aming | +0.077 |
ady | +0.077 |
lling | +0.077 |
Neg logit contributions | |
hill | -0.074 |
mountain | -0.069 |
road | -0.065 |
ladder | -0.063 |
subway | -0.063 |
Token | and |
Token ID | 290 |
Feature activation | +5.78e-02 |
Pos logit contributions | |
amed | +0.056 |
ished | +0.053 |
aming | +0.050 |
ady | +0.050 |
lling | +0.050 |
Neg logit contributions | |
hill | -0.045 |
mountain | -0.042 |
road | -0.039 |
ladder | -0.038 |
subway | -0.038 |
Token | b |
Token ID | 275 |
Feature activation | -8.13e-02 |
Pos logit contributions | |
hill | +0.114 |
mountain | +0.107 |
road | +0.099 |
subway | +0.097 |
ladder | +0.097 |
Neg logit contributions | |
amed | -0.137 |
ished | -0.131 |
aming | -0.124 |
ady | -0.124 |
lling | -0.123 |
Token | smiled |
Token ID | 13541 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
amed | +0.124 |
ished | +0.116 |
ady | +0.113 |
ealing | +0.113 |
aming | +0.111 |
Neg logit contributions | |
hill | -0.141 |
mountain | -0.131 |
taxi | -0.122 |
ladder | -0.120 |
bike | -0.120 |
Token | . |
Token ID | 13 |
Feature activation | -1.20e-02 |
Pos logit contributions | |
amed | +0.412 |
ished | +0.407 |
asing | +0.393 |
wed | +0.388 |
ady | +0.387 |
Neg logit contributions | |
hill | -0.206 |
mountain | -0.190 |
taxi | -0.177 |
bike | -0.176 |
road | -0.167 |
Token |
|
Token ID | 198 |
Feature activation | -1.04e-02 |
Pos logit contributions | |
amed | +0.033 |
ished | +0.032 |
lling | +0.031 |
aming | +0.030 |
ady | +0.030 |
Neg logit contributions | |
hill | -0.019 |
mountain | -0.018 |
road | -0.016 |
ladder | -0.016 |
taxi | -0.016 |
Token |
|
Token ID | 198 |
Feature activation | -5.68e-03 |
Pos logit contributions | |
ished | +0.032 |
amed | +0.032 |
asing | +0.030 |
lling | +0.030 |
aming | +0.029 |
Neg logit contributions | |
hill | -0.014 |
mountain | -0.012 |
road | -0.011 |
ladder | -0.011 |
cliff | -0.010 |
Token | " |
Token ID | 1 |
Feature activation | -2.52e-02 |
Pos logit contributions | |
ished | +0.015 |
amed | +0.014 |
ps | +0.013 |
asing | +0.013 |
lling | +0.013 |
Neg logit contributions | |
hill | -0.010 |
mountain | -0.009 |
road | -0.009 |
ladder | -0.008 |
taxi | -0.008 |
Token | Looks |
Token ID | 41102 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
amed | +0.066 |
ished | +0.064 |
ady | +0.061 |
aming | +0.060 |
lling | +0.060 |
Neg logit contributions | |
hill | -0.044 |
mountain | -0.041 |
road | -0.038 |
subway | -0.037 |
taxi | -0.037 |
Token | like |
Token ID | 588 |
Feature activation | -1.30e-02 |
Pos logit contributions | |
amed | +0.617 |
aming | +0.591 |
bled | +0.589 |
lling | +0.581 |
ated | +0.568 |
Neg logit contributions | |
hill | -0.238 |
road | -0.227 |
pine | -0.218 |
bike | -0.217 |
mountain | -0.215 |
Token | the |
Token ID | 262 |
Feature activation | -2.05e-02 |
Pos logit contributions | |
amed | +0.033 |
ished | +0.032 |
asing | +0.030 |
lling | +0.030 |
aming | +0.030 |
Neg logit contributions | |
hill | -0.023 |
mountain | -0.022 |
road | -0.020 |
ladder | -0.020 |
taxi | -0.019 |
Token | frog |
Token ID | 21264 |
Feature activation | +1.32e-02 |
Pos logit contributions | |
amed | +0.059 |
ished | +0.057 |
lling | +0.055 |
ady | +0.055 |
aming | +0.054 |
Neg logit contributions | |
hill | -0.028 |
mountain | -0.026 |
road | -0.023 |
subway | -0.023 |
ladder | -0.023 |
Token | likes |
Token ID | 7832 |
Feature activation | -3.94e-02 |
Pos logit contributions | |
hill | +0.026 |
mountain | +0.024 |
road | +0.022 |
subway | +0.022 |
cliff | +0.022 |
Neg logit contributions | |
amed | -0.031 |
ished | -0.030 |
ady | -0.028 |
aming | -0.028 |
lling | -0.028 |
Token | you |
Token ID | 345 |
Feature activation | -5.64e-02 |
Pos logit contributions | |
amed | +0.097 |
ished | +0.094 |
ady | +0.089 |
lling | +0.089 |
asing | +0.089 |
Neg logit contributions | |
hill | -0.070 |
mountain | -0.065 |
ladder | -0.060 |
road | -0.060 |
subway | -0.060 |
Token | smiled |
Token ID | 13541 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
amed | +0.071 |
ady | +0.069 |
ished | +0.068 |
ealing | +0.066 |
asing | +0.066 |
Neg logit contributions | |
hill | -0.071 |
mountain | -0.067 |
taxi | -0.063 |
ladder | -0.061 |
volcano | -0.061 |
Token | and |
Token ID | 290 |
Feature activation | +4.01e-02 |
Pos logit contributions | |
amed | +0.411 |
ished | +0.411 |
asing | +0.410 |
wed | +0.395 |
lling | +0.394 |
Neg logit contributions | |
hill | -0.188 |
mountain | -0.170 |
taxi | -0.165 |
bike | -0.163 |
bicycle | -0.155 |
Token | said |
Token ID | 531 |
Feature activation | -9.05e-02 |
Pos logit contributions | |
hill | +0.089 |
mountain | +0.084 |
taxi | +0.079 |
ladder | +0.079 |
subway | +0.078 |
Neg logit contributions | |
amed | -0.083 |
ished | -0.080 |
ady | -0.077 |
lling | -0.075 |
asing | -0.075 |
Token | , |
Token ID | 11 |
Feature activation | +5.83e-02 |
Pos logit contributions | |
amed | +0.277 |
ished | +0.267 |
ady | +0.263 |
aming | +0.260 |
lling | +0.260 |
Neg logit contributions | |
hill | -0.114 |
mountain | -0.104 |
ladder | -0.096 |
taxi | -0.096 |
subway | -0.091 |
Token | " |
Token ID | 366 |
Feature activation | -2.13e-02 |
Pos logit contributions | |
hill | +0.104 |
mountain | +0.097 |
road | +0.089 |
ladder | +0.088 |
taxi | +0.087 |
Neg logit contributions | |
amed | -0.150 |
ished | -0.144 |
aming | -0.137 |
ady | -0.137 |
lling | -0.136 |
Token | Looks |
Token ID | 41102 |
Feature activation | -2.11e-01 |
Pos logit contributions | |
amed | +0.051 |
ished | +0.050 |
ady | +0.047 |
aming | +0.046 |
ealing | +0.046 |
Neg logit contributions | |
hill | -0.041 |
mountain | -0.039 |
road | -0.036 |
subway | -0.035 |
taxi | -0.035 |
Token | like |
Token ID | 588 |
Feature activation | -1.13e-02 |
Pos logit contributions | |
amed | +0.621 |
aming | +0.583 |
bled | +0.572 |
lling | +0.569 |
ady | +0.567 |
Neg logit contributions | |
hill | -0.256 |
mountain | -0.235 |
road | -0.234 |
pine | -0.219 |
bike | -0.214 |
Token | you |
Token ID | 345 |
Feature activation | -6.46e-02 |
Pos logit contributions | |
amed | +0.027 |
ished | +0.026 |
lling | +0.025 |
aming | +0.025 |
asing | +0.025 |
Neg logit contributions | |
hill | -0.021 |
mountain | -0.020 |
road | -0.018 |
cliff | -0.018 |
ladder | -0.018 |
Token | like |
Token ID | 588 |
Feature activation | -7.66e-03 |
Pos logit contributions | |
amed | +0.162 |
ished | +0.156 |
aming | +0.148 |
lling | +0.146 |
ady | +0.146 |
Neg logit contributions | |
hill | -0.121 |
mountain | -0.113 |
road | -0.104 |
cliff | -0.101 |
taxi | -0.101 |
Token | the |
Token ID | 262 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
amed | +0.017 |
ished | +0.016 |
lling | +0.016 |
aming | +0.016 |
ady | +0.015 |
Neg logit contributions | |
hill | -0.015 |
mountain | -0.014 |
road | -0.013 |
cliff | -0.013 |
ladder | -0.013 |
Token | wardrobe |
Token ID | 35593 |
Feature activation | +7.90e-02 |
Pos logit contributions | |
amed | +0.033 |
ished | +0.031 |
lling | +0.030 |
aming | +0.030 |
ady | +0.030 |
Neg logit contributions | |
hill | -0.027 |
mountain | -0.026 |
road | -0.024 |
cliff | -0.024 |
ladder | -0.024 |
Token | hard |
Token ID | 1327 |
Feature activation | -1.47e-02 |
Pos logit contributions | |
hill | +0.103 |
mountain | +0.098 |
ladder | +0.093 |
cliff | +0.093 |
taxi | +0.092 |
Neg logit contributions | |
amed | -0.117 |
ished | -0.109 |
ady | -0.104 |
aming | -0.103 |
lling | -0.100 |
Token | on |
Token ID | 319 |
Feature activation | +7.21e-02 |
Pos logit contributions | |
amed | +0.038 |
ished | +0.036 |
ady | +0.035 |
aming | +0.035 |
ealing | +0.033 |
Neg logit contributions | |
hill | -0.027 |
mountain | -0.025 |
road | -0.023 |
ladder | -0.023 |
cliff | -0.023 |
Token | yourself |
Token ID | 3511 |
Feature activation | -8.77e-02 |
Pos logit contributions | |
hill | +0.131 |
mountain | +0.122 |
road | +0.113 |
ladder | +0.110 |
subway | +0.110 |
Neg logit contributions | |
amed | -0.183 |
ished | -0.176 |
aming | -0.167 |
ady | -0.166 |
lling | -0.166 |
Token | !' |
Token ID | 13679 |
Feature activation | -4.03e-04 |
Pos logit contributions | |
amed | +0.218 |
ished | +0.209 |
aming | +0.198 |
ady | +0.198 |
lling | +0.196 |
Neg logit contributions | |
hill | -0.164 |
mountain | -0.153 |
road | -0.142 |
ladder | -0.139 |
cliff | -0.138 |
Token | said |
Token ID | 531 |
Feature activation | -8.63e-02 |
Pos logit contributions | |
amed | +0.001 |
ished | +0.001 |
lling | +0.001 |
aming | +0.001 |
ady | +0.001 |
Neg logit contributions | |
hill | -0.001 |
mountain | -0.000 |
ladder | -0.000 |
taxi | -0.000 |
bike | -0.000 |
Token | Gr |
Token ID | 1902 |
Feature activation | -2.11e-01 |
Pos logit contributions | |
amed | +0.209 |
ished | +0.200 |
aming | +0.189 |
ady | +0.188 |
lling | +0.188 |
Neg logit contributions | |
hill | -0.168 |
mountain | -0.157 |
road | -0.146 |
cliff | -0.142 |
ladder | -0.142 |
Token | anny |
Token ID | 7737 |
Feature activation | -7.53e-02 |
Pos logit contributions | |
amed | +0.422 |
ealing | +0.418 |
bled | +0.403 |
ished | +0.402 |
aming | +0.396 |
Neg logit contributions | |
hill | -0.354 |
mountain | -0.327 |
volcano | -0.325 |
road | -0.323 |
museum | -0.315 |
Token | . |
Token ID | 13 |
Feature activation | -7.51e-03 |
Pos logit contributions | |
amed | +0.199 |
ished | +0.189 |
aming | +0.185 |
ady | +0.184 |
lling | +0.179 |
Neg logit contributions | |
hill | -0.126 |
mountain | -0.119 |
taxi | -0.110 |
ladder | -0.109 |
bike | -0.105 |
Token |
|
Token ID | 198 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
amed | +0.020 |
ished | +0.019 |
aming | +0.018 |
lling | +0.018 |
ady | +0.018 |
Neg logit contributions | |
hill | -0.013 |
mountain | -0.012 |
road | -0.011 |
ladder | -0.011 |
taxi | -0.011 |
Token |
|
Token ID | 198 |
Feature activation | -8.01e-03 |
Pos logit contributions | |
amed | +0.034 |
ished | +0.034 |
lling | +0.032 |
asing | +0.032 |
aming | +0.031 |
Neg logit contributions | |
hill | -0.016 |
mountain | -0.014 |
road | -0.013 |
ladder | -0.013 |
taxi | -0.013 |
Token | L |
Token ID | 43 |
Feature activation | -1.65e-01 |
Pos logit contributions | |
ished | +0.020 |
amed | +0.019 |
ealing | +0.018 |
ady | +0.018 |
lling | +0.018 |
Neg logit contributions | |
hill | -0.014 |
mountain | -0.013 |
road | -0.012 |
taxi | -0.012 |
bike | -0.012 |
Token | we |
Token ID | 356 |
Feature activation | -4.11e-02 |
Pos logit contributions | |
hill | +0.044 |
mountain | +0.042 |
road | +0.039 |
subway | +0.037 |
cliff | +0.037 |
Neg logit contributions | |
amed | -0.067 |
ished | -0.066 |
asing | -0.065 |
ealing | -0.064 |
aming | -0.062 |
Token | do |
Token ID | 466 |
Feature activation | +2.40e-03 |
Pos logit contributions | |
amed | +0.085 |
ished | +0.084 |
ealing | +0.079 |
ady | +0.079 |
aming | +0.078 |
Neg logit contributions | |
hill | -0.080 |
mountain | -0.077 |
subway | -0.074 |
road | -0.071 |
ladder | -0.071 |
Token | it |
Token ID | 340 |
Feature activation | +2.53e-03 |
Pos logit contributions | |
hill | +0.004 |
mountain | +0.004 |
road | +0.004 |
ladder | +0.004 |
cliff | +0.004 |
Neg logit contributions | |
amed | -0.006 |
ished | -0.006 |
ealing | -0.005 |
asing | -0.005 |
lling | -0.005 |
Token | again |
Token ID | 757 |
Feature activation | +2.82e-02 |
Pos logit contributions | |
hill | +0.005 |
mountain | +0.005 |
road | +0.005 |
ladder | +0.005 |
subway | +0.005 |
Neg logit contributions | |
amed | -0.006 |
ished | -0.005 |
aming | -0.005 |
ady | -0.005 |
lling | -0.005 |
Token | ?" |
Token ID | 1701 |
Feature activation | -1.17e-02 |
Pos logit contributions | |
hill | +0.050 |
mountain | +0.046 |
road | +0.043 |
ladder | +0.042 |
subway | +0.042 |
Neg logit contributions | |
amed | -0.073 |
ished | -0.070 |
aming | -0.067 |
ady | -0.067 |
lling | -0.066 |
Token | Gr |
Token ID | 1902 |
Feature activation | -2.10e-01 |
Pos logit contributions | |
amed | +0.026 |
ished | +0.025 |
ady | +0.023 |
ealing | +0.023 |
aming | +0.023 |
Neg logit contributions | |
hill | -0.024 |
mountain | -0.023 |
cliff | -0.021 |
road | -0.021 |
subway | -0.021 |
Token | anny |
Token ID | 7737 |
Feature activation | -7.07e-02 |
Pos logit contributions | |
amed | +0.375 |
ished | +0.375 |
ealing | +0.358 |
lling | +0.350 |
bled | +0.350 |
Neg logit contributions | |
hill | -0.422 |
volcano | -0.399 |
mountain | -0.392 |
road | -0.391 |
chim | -0.381 |
Token | said |
Token ID | 531 |
Feature activation | -9.83e-02 |
Pos logit contributions | |
amed | +0.151 |
ady | +0.139 |
aming | +0.139 |
ished | +0.138 |
lling | +0.133 |
Neg logit contributions | |
hill | -0.148 |
mountain | -0.139 |
taxi | -0.134 |
ladder | -0.130 |
bike | -0.129 |
Token | , |
Token ID | 11 |
Feature activation | +5.17e-02 |
Pos logit contributions | |
amed | +0.313 |
ady | +0.304 |
ished | +0.303 |
aming | +0.294 |
lling | +0.294 |
Neg logit contributions | |
hill | -0.106 |
mountain | -0.093 |
taxi | -0.092 |
ladder | -0.089 |
bike | -0.086 |
Token | " |
Token ID | 366 |
Feature activation | -1.45e-02 |
Pos logit contributions | |
hill | +0.085 |
mountain | +0.078 |
road | +0.072 |
taxi | +0.072 |
ladder | +0.071 |
Neg logit contributions | |
amed | -0.139 |
ished | -0.137 |
ady | -0.129 |
aming | -0.128 |
lling | -0.128 |
Token | Of |
Token ID | 5189 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
amed | +0.041 |
ished | +0.040 |
ady | +0.038 |
aming | +0.038 |
lling | +0.038 |
Neg logit contributions | |
hill | -0.022 |
mountain | -0.020 |
road | -0.018 |
taxi | -0.018 |
ladder | -0.018 |
Token | to |
Token ID | 284 |
Feature activation | +8.33e-02 |
Pos logit contributions | |
amed | +0.096 |
ished | +0.093 |
aming | +0.089 |
ady | +0.089 |
lling | +0.088 |
Neg logit contributions | |
hill | -0.046 |
mountain | -0.042 |
road | -0.038 |
ladder | -0.037 |
cliff | -0.037 |
Token | buzz |
Token ID | 14713 |
Feature activation | -9.32e-02 |
Pos logit contributions | |
hill | +0.151 |
mountain | +0.140 |
road | +0.130 |
ladder | +0.127 |
subway | +0.127 |
Neg logit contributions | |
amed | -0.213 |
ished | -0.204 |
aming | -0.194 |
ady | -0.193 |
lling | -0.192 |
Token | around |
Token ID | 1088 |
Feature activation | +8.31e-03 |
Pos logit contributions | |
amed | +0.229 |
ished | +0.218 |
ady | +0.209 |
aming | +0.208 |
asing | +0.206 |
Neg logit contributions | |
hill | -0.175 |
mountain | -0.163 |
road | -0.151 |
cliff | -0.148 |
taxi | -0.148 |
Token | him |
Token ID | 683 |
Feature activation | -3.73e-02 |
Pos logit contributions | |
hill | +0.016 |
mountain | +0.015 |
road | +0.014 |
ladder | +0.013 |
cliff | +0.013 |
Neg logit contributions | |
amed | -0.020 |
ished | -0.020 |
aming | -0.019 |
lling | -0.019 |
ady | -0.019 |
Token | . |
Token ID | 13 |
Feature activation | +6.49e-04 |
Pos logit contributions | |
amed | +0.091 |
ished | +0.087 |
aming | +0.082 |
ady | +0.082 |
lling | +0.081 |
Neg logit contributions | |
hill | -0.072 |
mountain | -0.067 |
road | -0.062 |
ladder | -0.061 |
cliff | -0.061 |
Token | Th |
Token ID | 536 |
Feature activation | -2.09e-01 |
Pos logit contributions | |
hill | +0.001 |
mountain | +0.001 |
road | +0.001 |
ladder | +0.001 |
taxi | +0.001 |
Neg logit contributions | |
amed | -0.002 |
ished | -0.002 |
lling | -0.002 |
aming | -0.002 |
asing | -0.002 |
Token | ump |
Token ID | 931 |
Feature activation | -1.31e-01 |
Pos logit contributions | |
amed | +0.284 |
ished | +0.263 |
bled | +0.252 |
lling | +0.248 |
ealing | +0.245 |
Neg logit contributions | |
hill | -0.531 |
road | -0.487 |
mountain | -0.482 |
je | -0.473 |
cliff | -0.459 |
Token | was |
Token ID | 373 |
Feature activation | -3.37e-02 |
Pos logit contributions | |
amed | +0.232 |
ished | +0.214 |
ady | +0.212 |
aming | +0.208 |
asing | +0.198 |
Neg logit contributions | |
hill | -0.315 |
mountain | -0.295 |
taxi | -0.291 |
ladder | -0.283 |
subway | -0.282 |
Token | very |
Token ID | 845 |
Feature activation | +4.47e-02 |
Pos logit contributions | |
ished | +0.081 |
asing | +0.078 |
amed | +0.077 |
ealing | +0.075 |
ins | +0.073 |
Neg logit contributions | |
hill | -0.059 |
mountain | -0.056 |
taxi | -0.055 |
road | -0.054 |
ladder | -0.052 |
Token | scared |
Token ID | 12008 |
Feature activation | -8.26e-02 |
Pos logit contributions | |
hill | +0.095 |
mountain | +0.090 |
ladder | +0.085 |
road | +0.084 |
cliff | +0.084 |
Neg logit contributions | |
amed | -0.100 |
ished | -0.097 |
asing | -0.091 |
aming | -0.089 |
ady | -0.089 |
Token | , |
Token ID | 11 |
Feature activation | +8.55e-02 |
Pos logit contributions | |
amed | +0.221 |
ished | +0.209 |
aming | +0.199 |
ady | +0.199 |
lling | +0.195 |
Neg logit contributions | |
hill | -0.144 |
mountain | -0.132 |
road | -0.123 |
taxi | -0.120 |
ladder | -0.119 |
Token | 's |
Token ID | 338 |
Feature activation | -5.14e-02 |
Pos logit contributions | |
amed | +0.369 |
ished | +0.358 |
aming | +0.342 |
wed | +0.336 |
asing | +0.336 |
Neg logit contributions | |
hill | -0.206 |
mountain | -0.185 |
ladder | -0.176 |
bike | -0.168 |
road | -0.168 |
Token | mom |
Token ID | 1995 |
Feature activation | -5.33e-02 |
Pos logit contributions | |
amed | +0.138 |
ished | +0.133 |
lling | +0.128 |
ady | +0.127 |
asing | +0.127 |
Neg logit contributions | |
hill | -0.089 |
mountain | -0.081 |
ladder | -0.076 |
road | -0.075 |
taxi | -0.075 |
Token | my |
Token ID | 1820 |
Feature activation | -8.16e-02 |
Pos logit contributions | |
amed | +0.105 |
ished | +0.099 |
lling | +0.091 |
ealing | +0.091 |
aming | +0.091 |
Neg logit contributions | |
hill | -0.123 |
mountain | -0.113 |
taxi | -0.112 |
ladder | -0.111 |
bike | -0.109 |
Token | . |
Token ID | 13 |
Feature activation | -9.69e-04 |
Pos logit contributions | |
amed | +0.233 |
ished | +0.221 |
aming | +0.212 |
asing | +0.211 |
ady | +0.211 |
Neg logit contributions | |
hill | -0.116 |
mountain | -0.104 |
taxi | -0.104 |
ladder | -0.102 |
bike | -0.094 |
Token | " |
Token ID | 366 |
Feature activation | -1.60e-02 |
Pos logit contributions | |
amed | +0.002 |
ished | +0.002 |
aming | +0.002 |
ady | +0.002 |
lling | +0.002 |
Neg logit contributions | |
hill | -0.002 |
mountain | -0.002 |
road | -0.002 |
ladder | -0.001 |
cliff | -0.001 |
Token | Looks |
Token ID | 41102 |
Feature activation | -2.09e-01 |
Pos logit contributions | |
amed | +0.038 |
ished | +0.038 |
ealing | +0.035 |
ady | +0.035 |
aming | +0.034 |
Neg logit contributions | |
hill | -0.031 |
mountain | -0.028 |
road | -0.027 |
cliff | -0.026 |
subway | -0.026 |
Token | like |
Token ID | 588 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
amed | +0.558 |
aming | +0.526 |
ished | +0.516 |
lling | +0.514 |
ady | +0.513 |
Neg logit contributions | |
hill | -0.306 |
road | -0.289 |
mountain | -0.277 |
bike | -0.267 |
ladder | -0.266 |
Token | the |
Token ID | 262 |
Feature activation | -1.98e-02 |
Pos logit contributions | |
amed | +0.029 |
ished | +0.028 |
aming | +0.027 |
lling | +0.027 |
asing | +0.027 |
Neg logit contributions | |
hill | -0.020 |
mountain | -0.018 |
road | -0.017 |
taxi | -0.017 |
subway | -0.016 |
Token | wave |
Token ID | 6769 |
Feature activation | -4.45e-02 |
Pos logit contributions | |
amed | +0.057 |
ished | +0.055 |
aming | +0.052 |
ady | +0.052 |
lling | +0.052 |
Neg logit contributions | |
hill | -0.030 |
mountain | -0.027 |
road | -0.025 |
ladder | -0.024 |
cliff | -0.024 |
Token | destroyed |
Token ID | 6572 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
amed | +0.105 |
ished | +0.101 |
lling | +0.095 |
aming | +0.095 |
ady | +0.095 |
Neg logit contributions | |
hill | -0.086 |
mountain | -0.080 |
road | -0.076 |
subway | -0.074 |
cliff | -0.074 |
Token | all |
Token ID | 477 |
Feature activation | +2.16e-02 |
Pos logit contributions | |
amed | +0.035 |
ished | +0.034 |
lling | +0.032 |
aming | +0.032 |
ady | +0.031 |
Neg logit contributions | |
hill | -0.029 |
mountain | -0.027 |
road | -0.025 |
ladder | -0.025 |
subway | -0.025 |
Token | table |
Token ID | 3084 |
Feature activation | +8.56e-02 |
Pos logit contributions | |
hill | +0.020 |
mountain | +0.019 |
cliff | +0.018 |
taxi | +0.017 |
road | +0.017 |
Neg logit contributions | |
amed | -0.028 |
ished | -0.027 |
lling | -0.026 |
aming | -0.026 |
ady | -0.026 |
Token | . |
Token ID | 13 |
Feature activation | -1.31e-02 |
Pos logit contributions | |
hill | +0.157 |
mountain | +0.144 |
road | +0.136 |
ladder | +0.133 |
taxi | +0.129 |
Neg logit contributions | |
amed | -0.221 |
ished | -0.211 |
aming | -0.201 |
ady | -0.200 |
ealing | -0.199 |
Token |
|
Token ID | 198 |
Feature activation | -1.51e-02 |
Pos logit contributions | |
amed | +0.033 |
ished | +0.032 |
ady | +0.031 |
aming | +0.030 |
ealing | +0.030 |
Neg logit contributions | |
hill | -0.024 |
mountain | -0.022 |
cliff | -0.020 |
subway | -0.020 |
road | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
amed | +0.043 |
ished | +0.042 |
lling | +0.040 |
asing | +0.040 |
aming | +0.040 |
Neg logit contributions | |
hill | -0.023 |
mountain | -0.021 |
road | -0.019 |
ladder | -0.019 |
taxi | -0.018 |
Token | " |
Token ID | 1 |
Feature activation | -2.63e-02 |
Pos logit contributions | |
amed | +0.025 |
ished | +0.025 |
lling | +0.023 |
aming | +0.023 |
ady | +0.023 |
Neg logit contributions | |
hill | -0.025 |
mountain | -0.023 |
road | -0.022 |
ladder | -0.021 |
taxi | -0.021 |
Token | Looks |
Token ID | 41102 |
Feature activation | -2.08e-01 |
Pos logit contributions | |
amed | +0.064 |
ished | +0.061 |
aming | +0.058 |
ady | +0.058 |
lling | +0.058 |
Neg logit contributions | |
hill | -0.052 |
mountain | -0.048 |
road | -0.045 |
ladder | -0.044 |
cliff | -0.044 |
Token | like |
Token ID | 588 |
Feature activation | -1.12e-02 |
Pos logit contributions | |
amed | +0.594 |
aming | +0.565 |
bled | +0.561 |
lling | +0.561 |
ished | +0.543 |
Neg logit contributions | |
hill | -0.243 |
road | -0.241 |
mountain | -0.223 |
ladder | -0.220 |
bike | -0.219 |
Token | mom |
Token ID | 1995 |
Feature activation | -2.64e-02 |
Pos logit contributions | |
amed | +0.026 |
ished | +0.026 |
aming | +0.024 |
asing | +0.024 |
lling | +0.024 |
Neg logit contributions | |
hill | -0.022 |
mountain | -0.020 |
road | -0.019 |
cliff | -0.018 |
taxi | -0.018 |
Token | has |
Token ID | 468 |
Feature activation | -3.53e-03 |
Pos logit contributions | |
amed | +0.068 |
ished | +0.065 |
lling | +0.062 |
aming | +0.060 |
ealing | +0.060 |
Neg logit contributions | |
hill | -0.047 |
mountain | -0.044 |
taxi | -0.041 |
road | -0.041 |
ladder | -0.041 |
Token | some |
Token ID | 617 |
Feature activation | +3.46e-02 |
Pos logit contributions | |
amed | +0.009 |
ished | +0.008 |
lling | +0.008 |
aming | +0.008 |
ady | +0.008 |
Neg logit contributions | |
hill | -0.007 |
mountain | -0.006 |
road | -0.006 |
subway | -0.006 |
taxi | -0.006 |
Token | mail |
Token ID | 6920 |
Feature activation | +2.71e-02 |
Pos logit contributions | |
hill | +0.063 |
mountain | +0.059 |
road | +0.054 |
cliff | +0.053 |
ladder | +0.053 |
Neg logit contributions | |
amed | -0.088 |
ished | -0.085 |
aming | -0.080 |
ady | -0.080 |
lling | -0.079 |
Token | their |
Token ID | 511 |
Feature activation | -2.95e-02 |
Pos logit contributions | |
asing | +0.188 |
amed | +0.186 |
ady | +0.183 |
ished | +0.179 |
arre | +0.172 |
Neg logit contributions | |
hill | -0.297 |
mountain | -0.291 |
taxi | -0.284 |
plane | -0.271 |
ladder | -0.270 |
Token | cups |
Token ID | 14180 |
Feature activation | +9.07e-03 |
Pos logit contributions | |
amed | +0.067 |
ished | +0.064 |
ady | +0.061 |
aming | +0.061 |
lling | +0.060 |
Neg logit contributions | |
hill | -0.060 |
mountain | -0.057 |
road | -0.053 |
cliff | -0.052 |
subway | -0.052 |
Token | and |
Token ID | 290 |
Feature activation | +6.07e-02 |
Pos logit contributions | |
hill | +0.017 |
mountain | +0.015 |
road | +0.014 |
ladder | +0.014 |
taxi | +0.014 |
Neg logit contributions | |
amed | -0.023 |
ished | -0.022 |
ealing | -0.021 |
lling | -0.021 |
ady | -0.021 |
Token | say |
Token ID | 910 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
hill | +0.122 |
mountain | +0.114 |
road | +0.107 |
ladder | +0.105 |
subway | +0.104 |
Neg logit contributions | |
amed | -0.143 |
ished | -0.138 |
aming | -0.130 |
ady | -0.130 |
lling | -0.128 |
Token | " |
Token ID | 366 |
Feature activation | -1.04e-02 |
Pos logit contributions | |
amed | +0.348 |
ished | +0.335 |
aming | +0.319 |
ady | +0.318 |
lling | +0.316 |
Neg logit contributions | |
hill | -0.220 |
mountain | -0.204 |
road | -0.187 |
taxi | -0.185 |
subway | -0.185 |
Token | che |
Token ID | 2395 |
Feature activation | -2.08e-01 |
Pos logit contributions | |
amed | +0.024 |
ished | +0.023 |
aming | +0.022 |
ady | +0.022 |
lling | +0.022 |
Neg logit contributions | |
hill | -0.021 |
mountain | -0.020 |
road | -0.018 |
subway | -0.018 |
ladder | -0.018 |
Token | ers |
Token ID | 364 |
Feature activation | -5.22e-02 |
Pos logit contributions | |
amed | +0.362 |
ished | +0.359 |
lling | +0.347 |
ady | +0.346 |
aming | +0.324 |
Neg logit contributions | |
hill | -0.455 |
taxi | -0.443 |
mountain | -0.435 |
subway | -0.421 |
bike | -0.414 |
Token | ". |
Token ID | 1911 |
Feature activation | -3.86e-03 |
Pos logit contributions | |
amed | +0.121 |
ished | +0.115 |
aming | +0.109 |
ady | +0.109 |
lling | +0.108 |
Neg logit contributions | |
hill | -0.106 |
mountain | -0.099 |
road | -0.093 |
ladder | -0.091 |
cliff | -0.091 |
Token |
|
Token ID | 198 |
Feature activation | -5.57e-03 |
Pos logit contributions | |
amed | +0.010 |
ished | +0.009 |
aming | +0.009 |
ady | +0.009 |
lling | +0.009 |
Neg logit contributions | |
hill | -0.007 |
mountain | -0.007 |
road | -0.006 |
ladder | -0.006 |
cliff | -0.006 |
Token |
|
Token ID | 198 |
Feature activation | -1.25e-03 |
Pos logit contributions | |
amed | +0.016 |
ished | +0.016 |
lling | +0.015 |
asing | +0.015 |
aming | +0.015 |
Neg logit contributions | |
hill | -0.008 |
mountain | -0.007 |
road | -0.007 |
taxi | -0.006 |
ladder | -0.006 |
Token | They |
Token ID | 2990 |
Feature activation | +6.00e-03 |
Pos logit contributions | |
amed | +0.003 |
ished | +0.003 |
lling | +0.003 |
aming | +0.003 |
ady | +0.003 |
Neg logit contributions | |
hill | -0.003 |
mountain | -0.002 |
road | -0.002 |
taxi | -0.002 |
subway | -0.002 |
Token | against |
Token ID | 1028 |
Feature activation | +4.10e-03 |
Pos logit contributions | |
amed | +0.130 |
ished | +0.124 |
aming | +0.118 |
ady | +0.117 |
lling | +0.117 |
Neg logit contributions | |
hill | -0.098 |
mountain | -0.091 |
road | -0.085 |
ladder | -0.083 |
subway | -0.082 |
Token | being |
Token ID | 852 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
hill | +0.007 |
mountain | +0.007 |
road | +0.006 |
subway | +0.006 |
cliff | +0.006 |
Neg logit contributions | |
amed | -0.010 |
ished | -0.010 |
ady | -0.009 |
lling | -0.009 |
aming | -0.009 |
Token | too |
Token ID | 1165 |
Feature activation | +1.17e-02 |
Pos logit contributions | |
amed | +0.280 |
ished | +0.271 |
aming | +0.254 |
lling | +0.252 |
ady | +0.252 |
Neg logit contributions | |
hill | -0.221 |
mountain | -0.208 |
road | -0.197 |
taxi | -0.193 |
ladder | -0.190 |
Token | curious |
Token ID | 11040 |
Feature activation | -6.76e-02 |
Pos logit contributions | |
hill | +0.022 |
mountain | +0.020 |
road | +0.019 |
ladder | +0.019 |
taxi | +0.018 |
Neg logit contributions | |
amed | -0.030 |
ished | -0.028 |
aming | -0.027 |
ady | -0.026 |
lling | -0.026 |
Token | . |
Token ID | 13 |
Feature activation | +3.43e-03 |
Pos logit contributions | |
amed | +0.178 |
ished | +0.170 |
aming | +0.163 |
ady | +0.162 |
lling | +0.160 |
Neg logit contributions | |
hill | -0.118 |
mountain | -0.109 |
road | -0.101 |
ladder | -0.098 |
taxi | -0.098 |
Token | N |
Token ID | 399 |
Feature activation | -2.07e-01 |
Pos logit contributions | |
hill | +0.007 |
mountain | +0.006 |
road | +0.006 |
subway | +0.006 |
cliff | +0.006 |
Neg logit contributions | |
amed | -0.008 |
ished | -0.008 |
ady | -0.007 |
aming | -0.007 |
ealing | -0.007 |
Token | ont |
Token ID | 756 |
Feature activation | -7.01e-02 |
Pos logit contributions | |
amed | +0.383 |
ished | +0.354 |
lling | +0.339 |
aming | +0.327 |
ady | +0.312 |
Neg logit contributions | |
hill | -0.513 |
mountain | -0.491 |
road | -0.461 |
cliff | -0.460 |
taxi | -0.458 |
Token | he |
Token ID | 258 |
Feature activation | -4.72e-03 |
Pos logit contributions | |
amed | +0.131 |
aming | +0.125 |
lling | +0.120 |
ished | +0.117 |
ady | +0.113 |
Neg logit contributions | |
hill | -0.156 |
mountain | -0.152 |
je | -0.143 |
ladder | -0.143 |
bike | -0.141 |
Token | less |
Token ID | 1203 |
Feature activation | -8.45e-02 |
Pos logit contributions | |
amed | +0.011 |
ished | +0.010 |
aming | +0.010 |
ady | +0.010 |
lling | +0.010 |
Neg logit contributions | |
hill | -0.010 |
mountain | -0.009 |
road | -0.008 |
ladder | -0.008 |
subway | -0.008 |
Token | , |
Token ID | 11 |
Feature activation | +7.62e-02 |
Pos logit contributions | |
amed | +0.185 |
ished | +0.172 |
aming | +0.167 |
lling | +0.167 |
ady | +0.163 |
Neg logit contributions | |
hill | -0.172 |
mountain | -0.162 |
road | -0.153 |
subway | -0.152 |
taxi | -0.151 |
Token | it |
Token ID | 340 |
Feature activation | -8.39e-03 |
Pos logit contributions | |
hill | +0.153 |
mountain | +0.143 |
road | +0.133 |
ladder | +0.131 |
taxi | +0.131 |
Neg logit contributions | |
amed | -0.180 |
ished | -0.172 |
aming | -0.162 |
ady | -0.162 |
lling | -0.161 |
Token | the |
Token ID | 262 |
Feature activation | -1.72e-02 |
Pos logit contributions | |
amed | +0.042 |
ished | +0.040 |
aming | +0.038 |
ady | +0.038 |
asing | +0.038 |
Neg logit contributions | |
hill | -0.027 |
mountain | -0.025 |
road | -0.023 |
taxi | -0.023 |
subway | -0.023 |
Token | other |
Token ID | 584 |
Feature activation | -4.77e-03 |
Pos logit contributions | |
amed | +0.047 |
ished | +0.046 |
aming | +0.043 |
ady | +0.043 |
asing | +0.042 |
Neg logit contributions | |
hill | -0.029 |
mountain | -0.026 |
road | -0.025 |
subway | -0.024 |
ladder | -0.024 |
Token | children |
Token ID | 1751 |
Feature activation | -1.69e-02 |
Pos logit contributions | |
amed | +0.012 |
ished | +0.011 |
aming | +0.011 |
ady | +0.010 |
lling | +0.010 |
Neg logit contributions | |
hill | -0.009 |
mountain | -0.008 |
road | -0.008 |
subway | -0.008 |
taxi | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | +4.36e-03 |
Pos logit contributions | |
amed | +0.041 |
ished | +0.039 |
aming | +0.037 |
ady | +0.036 |
lling | +0.036 |
Neg logit contributions | |
hill | -0.033 |
mountain | -0.031 |
road | -0.029 |
ladder | -0.028 |
taxi | -0.028 |
Token | But |
Token ID | 887 |
Feature activation | -2.02e-03 |
Pos logit contributions | |
hill | +0.008 |
mountain | +0.007 |
road | +0.007 |
ladder | +0.006 |
volcano | +0.006 |
Neg logit contributions | |
amed | -0.012 |
ished | -0.011 |
asing | -0.011 |
aming | -0.011 |
lling | -0.010 |
Token | Gr |
Token ID | 1902 |
Feature activation | -2.07e-01 |
Pos logit contributions | |
amed | +0.005 |
ished | +0.005 |
lling | +0.004 |
aming | +0.004 |
ealing | +0.004 |
Neg logit contributions | |
hill | -0.004 |
mountain | -0.004 |
road | -0.003 |
ladder | -0.003 |
subway | -0.003 |
Token | umpy |
Token ID | 32152 |
Feature activation | -6.84e-02 |
Pos logit contributions | |
bled | +0.307 |
amed | +0.307 |
ished | +0.284 |
ps | +0.278 |
lling | +0.277 |
Neg logit contributions | |
hill | -0.441 |
road | -0.432 |
volcano | -0.394 |
mountain | -0.391 |
museum | -0.391 |
Token | Joe |
Token ID | 5689 |
Feature activation | -8.57e-02 |
Pos logit contributions | |
amed | +0.161 |
ished | +0.152 |
ady | +0.146 |
aming | +0.146 |
asing | +0.141 |
Neg logit contributions | |
hill | -0.139 |
mountain | -0.130 |
road | -0.122 |
subway | -0.120 |
taxi | -0.118 |
Token | was |
Token ID | 373 |
Feature activation | -4.38e-02 |
Pos logit contributions | |
amed | +0.203 |
ished | +0.188 |
aming | +0.183 |
ady | +0.182 |
asing | +0.176 |
Neg logit contributions | |
hill | -0.172 |
mountain | -0.159 |
road | -0.146 |
taxi | -0.144 |
ladder | -0.144 |
Token | already |
Token ID | 1541 |
Feature activation | -5.05e-03 |
Pos logit contributions | |
ished | +0.112 |
amed | +0.106 |
wed | +0.103 |
asing | +0.103 |
ady | +0.101 |
Neg logit contributions | |
hill | -0.069 |
taxi | -0.064 |
mountain | -0.064 |
cliff | -0.062 |
bench | -0.062 |
Token | there |
Token ID | 612 |
Feature activation | +5.70e-02 |
Pos logit contributions | |
ished | +0.012 |
amed | +0.012 |
asing | +0.011 |
ealing | +0.011 |
wed | +0.011 |
Neg logit contributions | |
hill | -0.009 |
mountain | -0.008 |
taxi | -0.008 |
road | -0.008 |
cliff | -0.007 |
Token | It |
Token ID | 632 |
Feature activation | +4.76e-02 |
Pos logit contributions | |
hill | +0.052 |
mountain | +0.048 |
cliff | +0.043 |
road | +0.043 |
subway | +0.042 |
Neg logit contributions | |
amed | -0.082 |
ished | -0.080 |
ady | -0.076 |
asing | -0.076 |
ealing | -0.075 |
Token | 's |
Token ID | 338 |
Feature activation | -1.99e-02 |
Pos logit contributions | |
hill | +0.109 |
mountain | +0.103 |
ladder | +0.096 |
road | +0.096 |
taxi | +0.094 |
Neg logit contributions | |
amed | -0.099 |
ished | -0.092 |
aming | -0.090 |
ady | -0.090 |
ealing | -0.086 |
Token | an |
Token ID | 281 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
amed | +0.047 |
ished | +0.045 |
aming | +0.042 |
ady | +0.042 |
lling | +0.042 |
Neg logit contributions | |
hill | -0.040 |
mountain | -0.038 |
road | -0.036 |
ladder | -0.035 |
taxi | -0.035 |
Token | emergency |
Token ID | 6334 |
Feature activation | +7.29e-02 |
Pos logit contributions | |
hill | +0.033 |
mountain | +0.031 |
road | +0.029 |
taxi | +0.029 |
subway | +0.028 |
Neg logit contributions | |
amed | -0.038 |
ished | -0.037 |
ealing | -0.035 |
asing | -0.034 |
aming | -0.034 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.06e-02 |
Pos logit contributions | |
hill | +0.146 |
mountain | +0.139 |
cliff | +0.131 |
valley | +0.127 |
road | +0.127 |
Neg logit contributions | |
amed | -0.156 |
ished | -0.149 |
ady | -0.140 |
aming | -0.139 |
lling | -0.136 |
Token | Che |
Token ID | 2580 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
amed | +0.022 |
ished | +0.021 |
ady | +0.019 |
aming | +0.019 |
asing | +0.019 |
Neg logit contributions | |
hill | -0.023 |
mountain | -0.022 |
subway | -0.021 |
taxi | -0.020 |
road | -0.020 |
Token | wy |
Token ID | 21768 |
Feature activation | -6.82e-02 |
Pos logit contributions | |
amed | +0.370 |
ished | +0.363 |
ady | +0.344 |
lee | +0.340 |
ealing | +0.336 |
Neg logit contributions | |
hill | -0.490 |
mountain | -0.454 |
cliff | -0.417 |
volcano | -0.413 |
road | -0.413 |
Token | wanted |
Token ID | 2227 |
Feature activation | -7.34e-02 |
Pos logit contributions | |
amed | +0.161 |
ished | +0.155 |
aming | +0.149 |
lling | +0.147 |
ady | +0.146 |
Neg logit contributions | |
hill | -0.136 |
mountain | -0.125 |
road | -0.118 |
bike | -0.115 |
ladder | -0.115 |
Token | to |
Token ID | 284 |
Feature activation | +8.26e-02 |
Pos logit contributions | |
amed | +0.195 |
ished | +0.193 |
aming | +0.179 |
lling | +0.177 |
ady | +0.176 |
Neg logit contributions | |
hill | -0.118 |
mountain | -0.108 |
cliff | -0.101 |
road | -0.096 |
valley | -0.095 |
Token | help |
Token ID | 1037 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
hill | +0.185 |
mountain | +0.177 |
subway | +0.167 |
cliff | +0.163 |
volcano | +0.162 |
Neg logit contributions | |
amed | -0.158 |
ished | -0.148 |
ady | -0.139 |
aming | -0.139 |
ps | -0.138 |
Token | , |
Token ID | 11 |
Feature activation | +9.31e-02 |
Pos logit contributions | |
amed | +0.036 |
ished | +0.035 |
lling | +0.033 |
aming | +0.033 |
ady | +0.033 |
Neg logit contributions | |
hill | -0.025 |
mountain | -0.023 |
road | -0.021 |
subway | -0.021 |
cliff | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | +9.99e-03 |
Pos logit contributions | |
hill | +0.104 |
mountain | +0.096 |
road | +0.090 |
ladder | +0.089 |
taxi | +0.088 |
Neg logit contributions | |
amed | -0.131 |
ished | -0.125 |
aming | -0.119 |
ady | -0.118 |
lling | -0.117 |
Token | |
Token ID | 220 |
Feature activation | -5.68e-02 |
Pos logit contributions | |
hill | +0.017 |
mountain | +0.015 |
road | +0.014 |
ladder | +0.014 |
taxi | +0.014 |
Neg logit contributions | |
amed | -0.027 |
ished | -0.026 |
aming | -0.025 |
lling | -0.025 |
ady | -0.025 |
Token |
|
Token ID | 198 |
Feature activation | +3.90e-03 |
Pos logit contributions | |
ished | +0.174 |
amed | +0.174 |
lling | +0.173 |
aming | +0.166 |
asing | +0.164 |
Neg logit contributions | |
hill | -0.063 |
ladder | -0.053 |
mountain | -0.052 |
road | -0.051 |
taxi | -0.048 |
Token |
|
Token ID | 198 |
Feature activation | +6.77e-03 |
Pos logit contributions | |
hill | +0.006 |
mountain | +0.005 |
road | +0.005 |
taxi | +0.005 |
subway | +0.005 |
Neg logit contributions | |
amed | -0.011 |
ished | -0.011 |
aming | -0.010 |
lling | -0.010 |
asing | -0.010 |
Token | When |
Token ID | 2215 |
Feature activation | +7.52e-02 |
Pos logit contributions | |
hill | +0.013 |
mountain | +0.012 |
road | +0.011 |
subway | +0.011 |
taxi | +0.011 |
Neg logit contributions | |
amed | -0.016 |
ished | -0.016 |
ady | -0.015 |
ealing | -0.015 |
asing | -0.015 |
Token | Gr |
Token ID | 1902 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
hill | +0.142 |
mountain | +0.131 |
road | +0.125 |
cliff | +0.122 |
ladder | +0.121 |
Neg logit contributions | |
amed | -0.181 |
ished | -0.171 |
lling | -0.166 |
aming | -0.165 |
ealing | -0.160 |
Token | acy |
Token ID | 1590 |
Feature activation | -8.09e-02 |
Pos logit contributions | |
amed | +0.363 |
ished | +0.344 |
bled | +0.339 |
ealing | +0.333 |
ok | +0.319 |
Neg logit contributions | |
hill | -0.432 |
mountain | -0.415 |
library | -0.408 |
road | -0.396 |
museum | -0.393 |
Token | got |
Token ID | 1392 |
Feature activation | -4.13e-02 |
Pos logit contributions | |
amed | +0.186 |
ished | +0.177 |
aming | +0.168 |
ady | +0.167 |
lling | +0.166 |
Neg logit contributions | |
hill | -0.167 |
mountain | -0.157 |
road | -0.146 |
cliff | -0.143 |
subway | -0.143 |
Token | to |
Token ID | 284 |
Feature activation | +7.53e-02 |
Pos logit contributions | |
amed | +0.101 |
ished | +0.095 |
aming | +0.091 |
ady | +0.088 |
lling | +0.087 |
Neg logit contributions | |
hill | -0.082 |
mountain | -0.077 |
road | -0.072 |
subway | -0.071 |
cliff | -0.069 |
Token | the |
Token ID | 262 |
Feature activation | -2.11e-02 |
Pos logit contributions | |
hill | +0.148 |
mountain | +0.135 |
ladder | +0.127 |
road | +0.127 |
taxi | +0.127 |
Neg logit contributions | |
amed | -0.163 |
ished | -0.154 |
lling | -0.154 |
aming | -0.152 |
ady | -0.152 |
Token | house |
Token ID | 2156 |
Feature activation | +8.73e-02 |
Pos logit contributions | |
amed | +0.055 |
ished | +0.050 |
aming | +0.050 |
ealing | +0.049 |
lling | +0.049 |
Neg logit contributions | |
hill | -0.029 |
Zig | -0.027 |
sled | -0.025 |
mountain | -0.025 |
subway | -0.025 |
Token | so |
Token ID | 523 |
Feature activation | +5.18e-02 |
Pos logit contributions | |
ished | +0.103 |
asing | +0.098 |
ealing | +0.097 |
wed | +0.096 |
oking | +0.096 |
Neg logit contributions | |
taxi | -0.047 |
hill | -0.047 |
mountain | -0.043 |
ladder | -0.043 |
road | -0.042 |
Token | happy |
Token ID | 3772 |
Feature activation | -6.13e-02 |
Pos logit contributions | |
hill | +0.091 |
taxi | +0.086 |
mountain | +0.084 |
ladder | +0.083 |
bike | +0.080 |
Neg logit contributions | |
ished | -0.125 |
amed | -0.116 |
lling | -0.116 |
ady | -0.115 |
asing | -0.114 |
Token | that |
Token ID | 326 |
Feature activation | +3.02e-02 |
Pos logit contributions | |
amed | +0.128 |
ished | +0.126 |
lling | +0.119 |
asing | +0.118 |
aming | +0.118 |
Neg logit contributions | |
hill | -0.132 |
mountain | -0.124 |
road | -0.119 |
subway | -0.117 |
cliff | -0.117 |
Token | she |
Token ID | 673 |
Feature activation | -7.47e-03 |
Pos logit contributions | |
hill | +0.037 |
mountain | +0.033 |
road | +0.030 |
cliff | +0.028 |
ladder | +0.028 |
Neg logit contributions | |
amed | -0.095 |
ished | -0.091 |
aming | -0.088 |
ady | -0.087 |
lling | -0.087 |
Token | gave |
Token ID | 2921 |
Feature activation | -1.09e-01 |
Pos logit contributions | |
amed | +0.015 |
ished | +0.014 |
ady | +0.014 |
ealing | +0.013 |
asing | +0.013 |
Neg logit contributions | |
hill | -0.017 |
mountain | -0.016 |
road | -0.015 |
taxi | -0.015 |
bike | -0.015 |
Token | Gr |
Token ID | 1902 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
amed | +0.297 |
ished | +0.283 |
ady | +0.279 |
aming | +0.274 |
ealing | +0.270 |
Neg logit contributions | |
hill | -0.178 |
mountain | -0.163 |
taxi | -0.149 |
road | -0.149 |
subway | -0.148 |
Token | anny |
Token ID | 7737 |
Feature activation | -6.08e-02 |
Pos logit contributions | |
amed | +0.404 |
ealing | +0.393 |
bled | +0.391 |
lling | +0.388 |
ished | +0.377 |
Neg logit contributions | |
hill | -0.368 |
road | -0.348 |
mountain | -0.327 |
je | -0.325 |
volcano | -0.319 |
Token | a |
Token ID | 257 |
Feature activation | +5.29e-02 |
Pos logit contributions | |
amed | +0.152 |
ished | +0.146 |
lling | +0.140 |
aming | +0.139 |
ady | +0.139 |
Neg logit contributions | |
hill | -0.107 |
mountain | -0.100 |
taxi | -0.095 |
bike | -0.092 |
subway | -0.092 |
Token | big |
Token ID | 1263 |
Feature activation | +3.03e-02 |
Pos logit contributions | |
hill | +0.108 |
mountain | +0.102 |
cliff | +0.097 |
road | +0.095 |
valley | +0.093 |
Neg logit contributions | |
amed | -0.119 |
ished | -0.107 |
ady | -0.104 |
aming | -0.102 |
ealing | -0.100 |
Token | hug |
Token ID | 16225 |
Feature activation | -5.44e-02 |
Pos logit contributions | |
hill | +0.069 |
mountain | +0.066 |
road | +0.063 |
cliff | +0.063 |
subway | +0.061 |
Neg logit contributions | |
amed | -0.059 |
ished | -0.057 |
ady | -0.052 |
aming | -0.051 |
ealing | -0.051 |
Token | . |
Token ID | 13 |
Feature activation | +5.13e-03 |
Pos logit contributions | |
amed | +0.138 |
ished | +0.133 |
ady | +0.128 |
aming | +0.127 |
ealing | +0.126 |
Neg logit contributions | |
hill | -0.097 |
mountain | -0.089 |
road | -0.084 |
taxi | -0.083 |
subway | -0.082 |
Token | " |
Token ID | 366 |
Feature activation | -1.72e-02 |
Pos logit contributions | |
amed | +0.001 |
ished | +0.001 |
lling | +0.001 |
aming | +0.001 |
ady | +0.001 |
Neg logit contributions | |
hill | -0.000 |
mountain | -0.000 |
road | -0.000 |
ladder | -0.000 |
taxi | -0.000 |
Token | Thanks |
Token ID | 9690 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
amed | +0.041 |
ished | +0.039 |
aming | +0.037 |
ady | +0.037 |
lling | +0.037 |
Neg logit contributions | |
hill | -0.034 |
mountain | -0.032 |
road | -0.030 |
ladder | -0.029 |
cliff | -0.029 |
Token | for |
Token ID | 329 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
amed | +0.262 |
ished | +0.250 |
aming | +0.232 |
ady | +0.231 |
lling | +0.229 |
Neg logit contributions | |
hill | -0.290 |
mountain | -0.274 |
road | -0.258 |
cliff | -0.254 |
subway | -0.253 |
Token | helping |
Token ID | 5742 |
Feature activation | -2.74e-02 |
Pos logit contributions | |
amed | +0.034 |
ished | +0.033 |
ady | +0.031 |
ealing | +0.030 |
aming | +0.030 |
Neg logit contributions | |
hill | -0.031 |
mountain | -0.029 |
road | -0.027 |
ladder | -0.027 |
volcano | -0.027 |
Token | me |
Token ID | 502 |
Feature activation | -6.24e-02 |
Pos logit contributions | |
amed | +0.074 |
ished | +0.069 |
aming | +0.066 |
ady | +0.066 |
lling | +0.065 |
Neg logit contributions | |
hill | -0.048 |
mountain | -0.044 |
road | -0.042 |
volcano | -0.039 |
ladder | -0.039 |
Token | , |
Token ID | 11 |
Feature activation | +8.97e-02 |
Pos logit contributions | |
amed | +0.165 |
ished | +0.156 |
aming | +0.150 |
ady | +0.149 |
lling | +0.148 |
Neg logit contributions | |
hill | -0.111 |
mountain | -0.103 |
road | -0.095 |
ladder | -0.094 |
cliff | -0.092 |
Token | Sammy |
Token ID | 44067 |
Feature activation | -9.35e-02 |
Pos logit contributions | |
hill | +0.176 |
mountain | +0.163 |
road | +0.152 |
bike | +0.148 |
ladder | +0.148 |
Neg logit contributions | |
amed | -0.223 |
ished | -0.213 |
aming | -0.202 |
ady | -0.202 |
lling | -0.197 |
Token | ." |
Token ID | 526 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
amed | +0.245 |
ished | +0.231 |
aming | +0.224 |
ady | +0.221 |
lling | +0.218 |
Neg logit contributions | |
hill | -0.169 |
mountain | -0.157 |
road | -0.142 |
ladder | -0.141 |
cliff | -0.137 |
Token |
|
Token ID | 198 |
Feature activation | -1.23e-02 |
Pos logit contributions | |
amed | +0.050 |
ished | +0.050 |
lling | +0.048 |
aming | +0.048 |
ady | +0.047 |
Neg logit contributions | |
hill | -0.023 |
mountain | -0.021 |
ladder | -0.019 |
road | -0.018 |
bike | -0.018 |
Token |
|
Token ID | 198 |
Feature activation | -7.54e-03 |
Pos logit contributions | |
ished | +0.037 |
amed | +0.036 |
lling | +0.035 |
asing | +0.034 |
ealing | +0.034 |
Neg logit contributions | |
hill | -0.016 |
mountain | -0.014 |
road | -0.013 |
ladder | -0.013 |
je | -0.012 |
Token | " |
Token ID | 1 |
Feature activation | -2.71e-02 |
Pos logit contributions | |
ished | +0.019 |
amed | +0.018 |
ealing | +0.017 |
lling | +0.017 |
ps | +0.017 |
Neg logit contributions | |
hill | -0.014 |
mountain | -0.013 |
road | -0.012 |
bike | -0.012 |
je | -0.011 |
Token | to |
Token ID | 284 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
amed | +0.117 |
ished | +0.114 |
lee | +0.111 |
ins | +0.110 |
aming | +0.108 |
Neg logit contributions | |
hill | -0.124 |
mountain | -0.115 |
cliff | -0.110 |
ladder | -0.106 |
valley | -0.101 |
Token | play |
Token ID | 711 |
Feature activation | +2.27e-03 |
Pos logit contributions | |
hill | +0.178 |
mountain | +0.167 |
cliff | +0.154 |
volcano | +0.151 |
ladder | +0.148 |
Neg logit contributions | |
amed | -0.244 |
ished | -0.234 |
ady | -0.232 |
aming | -0.229 |
wed | -0.223 |
Token | outside |
Token ID | 2354 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
hill | +0.005 |
mountain | +0.005 |
cliff | +0.004 |
ladder | +0.004 |
road | +0.004 |
Neg logit contributions | |
amed | -0.005 |
ished | -0.004 |
aming | -0.004 |
asing | -0.004 |
lling | -0.004 |
Token | in |
Token ID | 287 |
Feature activation | +9.39e-02 |
Pos logit contributions | |
hill | +0.241 |
mountain | +0.237 |
subway | +0.232 |
taxi | +0.225 |
cliff | +0.225 |
Neg logit contributions | |
amed | -0.200 |
aming | -0.190 |
ished | -0.187 |
lling | -0.175 |
ady | -0.170 |
Token | the |
Token ID | 262 |
Feature activation | +3.90e-02 |
Pos logit contributions | |
hill | +0.189 |
mountain | +0.183 |
subway | +0.172 |
cone | +0.171 |
road | +0.171 |
Neg logit contributions | |
amed | -0.195 |
aming | -0.188 |
ished | -0.177 |
ady | -0.177 |
lling | -0.173 |
Token | garden |
Token ID | 11376 |
Feature activation | +1.49e-01 |
Pos logit contributions | |
hill | +0.063 |
mountain | +0.060 |
ladder | +0.056 |
cliff | +0.054 |
subway | +0.053 |
Neg logit contributions | |
amed | -0.101 |
aming | -0.096 |
ished | -0.093 |
ady | -0.092 |
ealing | -0.087 |
Token | . |
Token ID | 13 |
Feature activation | +3.78e-02 |
Pos logit contributions | |
hill | +0.290 |
mountain | +0.271 |
road | +0.253 |
subway | +0.248 |
taxi | +0.247 |
Neg logit contributions | |
amed | -0.359 |
ished | -0.344 |
aming | -0.326 |
ady | -0.325 |
lling | -0.322 |
Token | One |
Token ID | 1881 |
Feature activation | +2.95e-02 |
Pos logit contributions | |
hill | +0.066 |
mountain | +0.063 |
cliff | +0.058 |
subway | +0.056 |
taxi | +0.056 |
Neg logit contributions | |
amed | -0.094 |
ished | -0.092 |
ady | -0.088 |
ealing | -0.085 |
aming | -0.085 |
Token | day |
Token ID | 1110 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
hill | +0.044 |
mountain | +0.042 |
cliff | +0.037 |
ladder | +0.037 |
bike | +0.036 |
Neg logit contributions | |
amed | -0.081 |
aming | -0.077 |
ished | -0.077 |
ady | -0.076 |
lling | -0.073 |
Token | , |
Token ID | 11 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
hill | +0.232 |
mountain | +0.221 |
road | +0.204 |
cliff | +0.203 |
volcano | +0.194 |
Neg logit contributions | |
amed | -0.228 |
ished | -0.219 |
ealing | -0.210 |
aming | -0.206 |
ady | -0.202 |
Token | she |
Token ID | 673 |
Feature activation | +1.68e-02 |
Pos logit contributions | |
hill | +0.247 |
mountain | +0.244 |
cliff | +0.225 |
volcano | +0.222 |
road | +0.217 |
Neg logit contributions | |
amed | -0.223 |
ished | -0.204 |
bled | -0.204 |
aming | -0.202 |
convinced | -0.189 |
Token | lot |
Token ID | 1256 |
Feature activation | +5.70e-02 |
Pos logit contributions | |
hill | +0.095 |
mountain | +0.092 |
cone | +0.080 |
road | +0.080 |
Zig | +0.079 |
Neg logit contributions | |
amed | -0.166 |
ealing | -0.162 |
ady | -0.158 |
arre | -0.154 |
aming | -0.153 |
Token | of |
Token ID | 286 |
Feature activation | +6.49e-02 |
Pos logit contributions | |
hill | +0.113 |
mountain | +0.105 |
road | +0.099 |
ladder | +0.098 |
taxi | +0.097 |
Neg logit contributions | |
amed | -0.135 |
ished | -0.129 |
lling | -0.123 |
ady | -0.122 |
aming | -0.122 |
Token | fun |
Token ID | 1257 |
Feature activation | +5.26e-03 |
Pos logit contributions | |
hill | +0.097 |
mountain | +0.087 |
road | +0.084 |
subway | +0.081 |
ladder | +0.081 |
Neg logit contributions | |
amed | -0.186 |
ished | -0.181 |
lling | -0.173 |
aming | -0.172 |
ady | -0.170 |
Token | at |
Token ID | 379 |
Feature activation | +6.86e-02 |
Pos logit contributions | |
hill | +0.011 |
mountain | +0.010 |
road | +0.010 |
taxi | +0.009 |
subway | +0.009 |
Neg logit contributions | |
amed | -0.012 |
ished | -0.011 |
aming | -0.011 |
lling | -0.011 |
ady | -0.011 |
Token | the |
Token ID | 262 |
Feature activation | -4.04e-03 |
Pos logit contributions | |
hill | +0.128 |
mountain | +0.119 |
taxi | +0.114 |
bike | +0.113 |
road | +0.113 |
Neg logit contributions | |
amed | -0.159 |
ished | -0.147 |
lling | -0.146 |
aming | -0.144 |
ps | -0.138 |
Token | mountain |
Token ID | 8598 |
Feature activation | +1.27e-01 |
Pos logit contributions | |
amed | +0.010 |
aming | +0.010 |
lling | +0.009 |
ished | +0.009 |
ealing | +0.009 |
Neg logit contributions | |
hill | -0.005 |
bicycle | -0.005 |
sled | -0.005 |
mountain | -0.005 |
cone | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | -7.44e-03 |
Pos logit contributions | |
hill | +0.270 |
mountain | +0.255 |
subway | +0.248 |
bicycle | +0.245 |
taxi | +0.243 |
Neg logit contributions | |
amed | -0.255 |
ished | -0.245 |
lling | -0.235 |
ady | -0.230 |
aming | -0.229 |
Token | They |
Token ID | 1119 |
Feature activation | +1.61e-03 |
Pos logit contributions | |
amed | +0.018 |
ished | +0.018 |
aming | +0.017 |
ady | +0.017 |
ealing | +0.017 |
Neg logit contributions | |
hill | -0.014 |
mountain | -0.013 |
road | -0.012 |
cliff | -0.012 |
taxi | -0.012 |
Token | ate |
Token ID | 15063 |
Feature activation | -4.76e-02 |
Pos logit contributions | |
hill | +0.004 |
mountain | +0.003 |
road | +0.003 |
ladder | +0.003 |
subway | +0.003 |
Neg logit contributions | |
amed | -0.003 |
ished | -0.003 |
aming | -0.003 |
lling | -0.003 |
ady | -0.003 |
Token | their |
Token ID | 511 |
Feature activation | -2.18e-02 |
Pos logit contributions | |
amed | +0.105 |
ished | +0.100 |
aming | +0.094 |
lling | +0.093 |
ady | +0.091 |
Neg logit contributions | |
hill | -0.101 |
mountain | -0.096 |
road | -0.090 |
subway | -0.088 |
cliff | -0.087 |
Token | snacks |
Token ID | 28487 |
Feature activation | +6.23e-02 |
Pos logit contributions | |
amed | +0.054 |
ished | +0.051 |
aming | +0.048 |
lling | +0.047 |
ady | +0.047 |
Neg logit contributions | |
hill | -0.040 |
mountain | -0.038 |
road | -0.035 |
bicycle | -0.034 |
subway | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | +9.86e-03 |
Pos logit contributions | |
hill | +0.125 |
mountain | +0.115 |
road | +0.107 |
ladder | +0.105 |
cliff | +0.103 |
Neg logit contributions | |
amed | -0.170 |
ished | -0.162 |
aming | -0.153 |
ady | -0.153 |
ealing | -0.152 |
Token | |
Token ID | 220 |
Feature activation | -6.23e-02 |
Pos logit contributions | |
hill | +0.019 |
mountain | +0.018 |
road | +0.017 |
subway | +0.017 |
cliff | +0.016 |
Neg logit contributions | |
amed | -0.024 |
ished | -0.023 |
ady | -0.022 |
aming | -0.021 |
ealing | -0.021 |
Token |
|
Token ID | 198 |
Feature activation | -3.26e-03 |
Pos logit contributions | |
amed | +0.194 |
ished | +0.192 |
lling | +0.190 |
aming | +0.185 |
asing | +0.181 |
Neg logit contributions | |
hill | -0.070 |
ladder | -0.059 |
mountain | -0.057 |
road | -0.056 |
Zig | -0.052 |
Token |
|
Token ID | 198 |
Feature activation | -6.52e-05 |
Pos logit contributions | |
ished | +0.010 |
amed | +0.009 |
lling | +0.009 |
asing | +0.009 |
aming | +0.009 |
Neg logit contributions | |
hill | -0.005 |
mountain | -0.004 |
road | -0.004 |
ladder | -0.004 |
cliff | -0.004 |
Token | Next |
Token ID | 10019 |
Feature activation | -3.13e-03 |
Pos logit contributions | |
amed | +0.000 |
ished | +0.000 |
lling | +0.000 |
asing | +0.000 |
ady | +0.000 |
Neg logit contributions | |
hill | -0.000 |
mountain | -0.000 |
road | -0.000 |
taxi | -0.000 |
cliff | -0.000 |
Token | , |
Token ID | 11 |
Feature activation | +9.22e-02 |
Pos logit contributions | |
amed | +0.008 |
ished | +0.007 |
aming | +0.007 |
ady | +0.007 |
lling | +0.007 |
Neg logit contributions | |
hill | -0.006 |
mountain | -0.006 |
road | -0.005 |
ladder | -0.005 |
bike | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -3.10e-02 |
Pos logit contributions | |
hill | +0.199 |
road | +0.184 |
mountain | +0.184 |
bicycle | +0.182 |
bike | +0.181 |
Neg logit contributions | |
amed | -0.178 |
lling | -0.165 |
ished | -0.160 |
aming | -0.160 |
sure | -0.159 |
Token | mouse |
Token ID | 10211 |
Feature activation | +4.48e-02 |
Pos logit contributions | |
amed | +0.085 |
ished | +0.081 |
ady | +0.078 |
aming | +0.078 |
lling | +0.077 |
Neg logit contributions | |
hill | -0.046 |
mountain | -0.042 |
cliff | -0.040 |
subway | -0.039 |
taxi | -0.039 |
Token | decided |
Token ID | 3066 |
Feature activation | -3.54e-02 |
Pos logit contributions | |
hill | +0.094 |
mountain | +0.089 |
road | +0.083 |
ladder | +0.081 |
taxi | +0.081 |
Neg logit contributions | |
amed | -0.101 |
ished | -0.096 |
aming | -0.091 |
ady | -0.090 |
ealing | -0.090 |
Token | to |
Token ID | 284 |
Feature activation | +7.86e-02 |
Pos logit contributions | |
amed | +0.099 |
ished | +0.095 |
lling | +0.092 |
aming | +0.091 |
ealing | +0.090 |
Neg logit contributions | |
hill | -0.054 |
mountain | -0.049 |
road | -0.044 |
cliff | -0.043 |
subway | -0.042 |
Token | peek |
Token ID | 27185 |
Feature activation | -7.97e-02 |
Pos logit contributions | |
hill | +0.150 |
mountain | +0.141 |
subway | +0.130 |
road | +0.130 |
cliff | +0.128 |
Neg logit contributions | |
amed | -0.183 |
ished | -0.177 |
ady | -0.169 |
aming | -0.167 |
lling | -0.166 |
Token | never |
Token ID | 1239 |
Feature activation | -3.69e-02 |
Pos logit contributions | |
amed | +0.066 |
ished | +0.063 |
aming | +0.060 |
lling | +0.060 |
ealing | +0.059 |
Neg logit contributions | |
hill | -0.069 |
mountain | -0.065 |
ladder | -0.060 |
taxi | -0.060 |
road | -0.060 |
Token | let |
Token ID | 1309 |
Feature activation | -5.36e-02 |
Pos logit contributions | |
amed | +0.094 |
ished | +0.092 |
ady | +0.088 |
ealing | +0.087 |
aming | +0.087 |
Neg logit contributions | |
hill | -0.064 |
mountain | -0.059 |
road | -0.056 |
bike | -0.054 |
taxi | -0.054 |
Token | her |
Token ID | 607 |
Feature activation | -5.37e-02 |
Pos logit contributions | |
amed | +0.126 |
ished | +0.121 |
aming | +0.114 |
ady | +0.114 |
lling | +0.113 |
Neg logit contributions | |
hill | -0.108 |
mountain | -0.101 |
road | -0.094 |
ladder | -0.092 |
subway | -0.092 |
Token | enter |
Token ID | 3802 |
Feature activation | -6.59e-03 |
Pos logit contributions | |
amed | +0.134 |
ished | +0.129 |
aming | +0.122 |
ady | +0.122 |
lling | +0.121 |
Neg logit contributions | |
hill | -0.101 |
mountain | -0.094 |
road | -0.087 |
cliff | -0.085 |
ladder | -0.085 |
Token | her |
Token ID | 607 |
Feature activation | -2.22e-02 |
Pos logit contributions | |
amed | +0.014 |
ished | +0.013 |
aming | +0.013 |
ady | +0.013 |
lling | +0.013 |
Neg logit contributions | |
hill | -0.014 |
mountain | -0.013 |
cliff | -0.012 |
road | -0.012 |
ladder | -0.012 |
Token | room |
Token ID | 2119 |
Feature activation | +9.96e-02 |
Pos logit contributions | |
amed | +0.045 |
ished | +0.044 |
ady | +0.043 |
aming | +0.041 |
asing | +0.039 |
Neg logit contributions | |
hill | -0.046 |
mountain | -0.044 |
cliff | -0.042 |
cone | -0.041 |
Zig | -0.040 |
Token | , |
Token ID | 11 |
Feature activation | +9.33e-02 |
Pos logit contributions | |
hill | +0.195 |
mountain | +0.183 |
road | +0.170 |
cliff | +0.168 |
subway | +0.168 |
Neg logit contributions | |
amed | -0.235 |
ished | -0.226 |
aming | -0.214 |
ady | -0.212 |
lling | -0.211 |
Token | so |
Token ID | 523 |
Feature activation | +6.99e-02 |
Pos logit contributions | |
hill | +0.180 |
mountain | +0.169 |
road | +0.156 |
cliff | +0.155 |
subway | +0.153 |
Neg logit contributions | |
amed | -0.222 |
ished | -0.212 |
aming | -0.204 |
lling | -0.202 |
ady | -0.200 |
Token | she |
Token ID | 673 |
Feature activation | +7.97e-04 |
Pos logit contributions | |
hill | +0.119 |
mountain | +0.109 |
road | +0.102 |
cliff | +0.097 |
subway | +0.097 |
Neg logit contributions | |
amed | -0.185 |
ished | -0.177 |
aming | -0.171 |
lling | -0.170 |
ady | -0.166 |
Token | was |
Token ID | 373 |
Feature activation | -2.78e-02 |
Pos logit contributions | |
hill | +0.002 |
mountain | +0.002 |
road | +0.002 |
cliff | +0.002 |
subway | +0.002 |
Neg logit contributions | |
amed | -0.002 |
ished | -0.002 |
aming | -0.001 |
ady | -0.001 |
lling | -0.001 |
Token | really |
Token ID | 1107 |
Feature activation | +2.40e-02 |
Pos logit contributions | |
ished | +0.067 |
amed | +0.066 |
asing | +0.063 |
ealing | +0.062 |
lling | +0.062 |
Neg logit contributions | |
hill | -0.049 |
taxi | -0.046 |
mountain | -0.046 |
ladder | -0.045 |
road | -0.044 |
Token | plant |
Token ID | 4618 |
Feature activation | +2.70e-02 |
Pos logit contributions | |
amed | +0.047 |
ished | +0.047 |
ady | +0.044 |
mistakes | +0.044 |
ps | +0.044 |
Neg logit contributions | |
hill | -0.035 |
mountain | -0.033 |
subway | -0.031 |
festival | -0.031 |
road | -0.031 |
Token | ! |
Token ID | 0 |
Feature activation | +1.21e-02 |
Pos logit contributions | |
hill | +0.053 |
mountain | +0.049 |
road | +0.046 |
ladder | +0.045 |
cliff | +0.045 |
Neg logit contributions | |
amed | -0.065 |
ished | -0.062 |
aming | -0.059 |
ady | -0.059 |
lling | -0.058 |
Token | You |
Token ID | 921 |
Feature activation | -9.94e-02 |
Pos logit contributions | |
hill | +0.027 |
mountain | +0.025 |
road | +0.024 |
cliff | +0.023 |
subway | +0.023 |
Neg logit contributions | |
amed | -0.026 |
ished | -0.025 |
ady | -0.024 |
aming | -0.024 |
lling | -0.023 |
Token | are |
Token ID | 389 |
Feature activation | -2.30e-02 |
Pos logit contributions | |
amed | +0.205 |
ished | +0.203 |
asing | +0.185 |
ady | +0.182 |
lling | +0.181 |
Neg logit contributions | |
hill | -0.220 |
mountain | -0.210 |
road | -0.197 |
subway | -0.195 |
cliff | -0.192 |
Token | hurting |
Token ID | 21530 |
Feature activation | -9.82e-02 |
Pos logit contributions | |
amed | +0.050 |
ished | +0.048 |
aming | +0.045 |
ady | +0.045 |
lling | +0.044 |
Neg logit contributions | |
hill | -0.050 |
mountain | -0.047 |
road | -0.045 |
ladder | -0.044 |
cliff | -0.044 |
Token | it |
Token ID | 340 |
Feature activation | -5.31e-03 |
Pos logit contributions | |
amed | +0.232 |
ished | +0.222 |
aming | +0.211 |
ady | +0.209 |
lling | +0.209 |
Neg logit contributions | |
hill | -0.196 |
mountain | -0.185 |
road | -0.171 |
subway | -0.169 |
ladder | -0.169 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.70e-03 |
Pos logit contributions | |
amed | +0.013 |
ished | +0.013 |
aming | +0.012 |
ady | +0.012 |
ealing | +0.012 |
Neg logit contributions | |
hill | -0.010 |
mountain | -0.009 |
road | -0.009 |
ladder | -0.008 |
taxi | -0.008 |
Token | Sara |
Token ID | 24799 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
hill | +0.004 |
mountain | +0.003 |
road | +0.003 |
subway | +0.003 |
bike | +0.003 |
Neg logit contributions | |
amed | -0.004 |
ished | -0.003 |
ady | -0.003 |
aming | -0.003 |
ealing | -0.003 |
Token | said |
Token ID | 531 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
amed | +0.326 |
ished | +0.312 |
aming | +0.295 |
lling | +0.291 |
ady | +0.290 |
Neg logit contributions | |
hill | -0.279 |
mountain | -0.250 |
road | -0.245 |
ladder | -0.240 |
bike | -0.234 |
Token | . |
Token ID | 13 |
Feature activation | -1.95e-02 |
Pos logit contributions | |
amed | +0.305 |
ished | +0.285 |
ady | +0.280 |
aming | +0.279 |
lling | +0.275 |
Neg logit contributions | |
hill | -0.127 |
ladder | -0.116 |
bike | -0.109 |
taxi | -0.109 |
mountain | -0.107 |
Token |
|
Token ID | 198 |
Feature activation | -2.45e-02 |
Pos logit contributions | |
amed | +0.050 |
ished | +0.047 |
ady | +0.045 |
aming | +0.045 |
lling | +0.044 |
Neg logit contributions | |
hill | -0.035 |
mountain | -0.033 |
road | -0.030 |
subway | -0.030 |
cliff | -0.029 |
Token | sky |
Token ID | 6766 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
amed | +0.030 |
aming | +0.028 |
lling | +0.027 |
ealing | +0.027 |
ished | +0.027 |
Neg logit contributions | |
hill | -0.021 |
mountain | -0.020 |
cone | -0.018 |
volcano | -0.018 |
saf | -0.018 |
Token | and |
Token ID | 290 |
Feature activation | +7.29e-02 |
Pos logit contributions | |
hill | +0.219 |
mountain | +0.205 |
road | +0.193 |
subway | +0.190 |
cliff | +0.190 |
Neg logit contributions | |
amed | -0.233 |
ished | -0.221 |
aming | -0.211 |
ady | -0.208 |
lling | -0.208 |
Token | sighed |
Token ID | 21893 |
Feature activation | -1.09e-01 |
Pos logit contributions | |
hill | +0.154 |
mountain | +0.144 |
road | +0.137 |
ladder | +0.135 |
taxi | +0.134 |
Neg logit contributions | |
amed | -0.164 |
ished | -0.154 |
lling | -0.148 |
aming | -0.147 |
ady | -0.142 |
Token | . |
Token ID | 13 |
Feature activation | +4.80e-03 |
Pos logit contributions | |
amed | +0.259 |
ished | +0.247 |
aming | +0.234 |
ady | +0.233 |
lling | +0.232 |
Neg logit contributions | |
hill | -0.217 |
mountain | -0.202 |
road | -0.189 |
ladder | -0.185 |
subway | -0.185 |
Token | A |
Token ID | 317 |
Feature activation | +4.85e-02 |
Pos logit contributions | |
hill | +0.010 |
mountain | +0.009 |
road | +0.008 |
cliff | +0.008 |
taxi | +0.008 |
Neg logit contributions | |
amed | -0.011 |
ished | -0.011 |
ady | -0.010 |
aming | -0.010 |
ealing | -0.010 |
Token | glass |
Token ID | 5405 |
Feature activation | +3.54e-02 |
Pos logit contributions | |
hill | +0.090 |
mountain | +0.085 |
cone | +0.082 |
Zig | +0.081 |
road | +0.080 |
Neg logit contributions | |
amed | -0.098 |
ealing | -0.091 |
aming | -0.088 |
ady | -0.086 |
lling | -0.085 |
Token | suddenly |
Token ID | 6451 |
Feature activation | +5.63e-02 |
Pos logit contributions | |
hill | +0.067 |
mountain | +0.063 |
road | +0.060 |
taxi | +0.058 |
subway | +0.058 |
Neg logit contributions | |
amed | -0.084 |
ished | -0.078 |
aming | -0.075 |
lling | -0.075 |
ady | -0.073 |
Token | appeared |
Token ID | 4120 |
Feature activation | -3.19e-02 |
Pos logit contributions | |
hill | +0.119 |
mountain | +0.112 |
road | +0.105 |
ladder | +0.102 |
cliff | +0.102 |
Neg logit contributions | |
amed | -0.127 |
ished | -0.121 |
aming | -0.114 |
lling | -0.113 |
ady | -0.112 |
Token | on |
Token ID | 319 |
Feature activation | +1.03e-01 |
Pos logit contributions | |
amed | +0.081 |
ished | +0.078 |
ady | +0.074 |
lling | +0.073 |
aming | +0.073 |
Neg logit contributions | |
hill | -0.058 |
mountain | -0.053 |
road | -0.049 |
ladder | -0.048 |
taxi | -0.048 |
Token | the |
Token ID | 262 |
Feature activation | +1.57e-02 |
Pos logit contributions | |
hill | +0.225 |
mountain | +0.214 |
ladder | +0.205 |
bicycle | +0.203 |
bike | +0.201 |
Neg logit contributions | |
amed | -0.194 |
aming | -0.188 |
lling | -0.178 |
ps | -0.172 |
ished | -0.170 |
Token | ground |
Token ID | 2323 |
Feature activation | +9.60e-02 |
Pos logit contributions | |
hill | +0.029 |
mountain | +0.028 |
volcano | +0.027 |
taxi | +0.026 |
ladder | +0.026 |
Neg logit contributions | |
amed | -0.034 |
aming | -0.032 |
lling | -0.031 |
ished | -0.030 |
ady | -0.029 |
Token | smiled |
Token ID | 13541 |
Feature activation | -1.36e-01 |
Pos logit contributions | |
hill | +0.022 |
mountain | +0.021 |
road | +0.019 |
ladder | +0.019 |
cliff | +0.019 |
Neg logit contributions | |
amed | -0.022 |
ished | -0.021 |
ady | -0.020 |
aming | -0.020 |
ealing | -0.020 |
Token | and |
Token ID | 290 |
Feature activation | +4.45e-02 |
Pos logit contributions | |
amed | +0.338 |
ished | +0.326 |
ady | +0.310 |
lling | +0.310 |
asing | +0.307 |
Neg logit contributions | |
hill | -0.253 |
mountain | -0.238 |
road | -0.219 |
taxi | -0.217 |
bike | -0.215 |
Token | started |
Token ID | 2067 |
Feature activation | -3.85e-02 |
Pos logit contributions | |
hill | +0.096 |
mountain | +0.090 |
road | +0.084 |
ladder | +0.083 |
taxi | +0.083 |
Neg logit contributions | |
amed | -0.096 |
ished | -0.092 |
ady | -0.089 |
asing | -0.087 |
aming | -0.086 |
Token | to |
Token ID | 284 |
Feature activation | +8.14e-02 |
Pos logit contributions | |
amed | +0.111 |
ished | +0.108 |
aming | +0.103 |
ady | +0.102 |
lling | +0.102 |
Neg logit contributions | |
hill | -0.057 |
mountain | -0.052 |
road | -0.047 |
subway | -0.045 |
cliff | -0.045 |
Token | eat |
Token ID | 4483 |
Feature activation | -2.38e-02 |
Pos logit contributions | |
hill | +0.126 |
mountain | +0.116 |
road | +0.106 |
subway | +0.105 |
taxi | +0.104 |
Neg logit contributions | |
amed | -0.226 |
ished | -0.219 |
aming | -0.209 |
ady | -0.209 |
lling | -0.208 |
Token | it |
Token ID | 340 |
Feature activation | +1.36e-02 |
Pos logit contributions | |
amed | +0.064 |
ished | +0.061 |
aming | +0.058 |
ady | +0.058 |
lling | +0.058 |
Neg logit contributions | |
hill | -0.040 |
mountain | -0.037 |
road | -0.034 |
subway | -0.033 |
ladder | -0.033 |
Token | all |
Token ID | 477 |
Feature activation | +1.80e-02 |
Pos logit contributions | |
hill | +0.027 |
mountain | +0.025 |
road | +0.024 |
ladder | +0.023 |
cliff | +0.023 |
Neg logit contributions | |
amed | -0.032 |
ished | -0.031 |
aming | -0.029 |
ady | -0.029 |
lling | -0.029 |
Token | up |
Token ID | 510 |
Feature activation | +2.70e-02 |
Pos logit contributions | |
hill | +0.035 |
mountain | +0.033 |
road | +0.030 |
ladder | +0.030 |
cliff | +0.029 |
Neg logit contributions | |
amed | -0.044 |
ished | -0.042 |
ady | -0.040 |
aming | -0.040 |
ealing | -0.040 |
Token | . |
Token ID | 13 |
Feature activation | +4.42e-03 |
Pos logit contributions | |
hill | +0.059 |
mountain | +0.056 |
subway | +0.054 |
taxi | +0.053 |
road | +0.053 |
Neg logit contributions | |
amed | -0.055 |
ished | -0.053 |
lling | -0.051 |
aming | -0.051 |
ady | -0.049 |
Token | Sam |
Token ID | 3409 |
Feature activation | -9.41e-02 |
Pos logit contributions | |
hill | +0.007 |
mountain | +0.007 |
road | +0.006 |
ladder | +0.006 |
cliff | +0.006 |
Neg logit contributions | |
amed | -0.012 |
ished | -0.012 |
lling | -0.012 |
aming | -0.011 |
ady | -0.011 |
Token | was |
Token ID | 373 |
Feature activation | -3.64e-02 |
Pos logit contributions | |
amed | +0.225 |
ished | +0.212 |
ady | +0.210 |
ealing | +0.210 |
asing | +0.210 |
Neg logit contributions | |
hill | -0.181 |
mountain | -0.170 |
pine | -0.151 |
ladder | -0.151 |
road | -0.150 |
Token | to |
Token ID | 284 |
Feature activation | +7.81e-02 |
Pos logit contributions | |
hill | +0.109 |
mountain | +0.101 |
road | +0.093 |
ladder | +0.090 |
cliff | +0.090 |
Neg logit contributions | |
amed | -0.177 |
ished | -0.171 |
aming | -0.163 |
ady | -0.163 |
lling | -0.161 |
Token | always |
Token ID | 1464 |
Feature activation | -5.77e-02 |
Pos logit contributions | |
hill | +0.155 |
mountain | +0.144 |
road | +0.135 |
ladder | +0.132 |
cliff | +0.132 |
Neg logit contributions | |
amed | -0.187 |
ished | -0.179 |
aming | -0.169 |
ady | -0.169 |
lling | -0.168 |
Token | lead |
Token ID | 1085 |
Feature activation | -8.66e-02 |
Pos logit contributions | |
amed | +0.142 |
ished | +0.136 |
aming | +0.128 |
ady | +0.128 |
lling | +0.127 |
Neg logit contributions | |
hill | -0.112 |
mountain | -0.105 |
road | -0.097 |
ladder | -0.095 |
cliff | -0.094 |
Token | the |
Token ID | 262 |
Feature activation | -3.93e-02 |
Pos logit contributions | |
amed | +0.214 |
ished | +0.207 |
lling | +0.196 |
aming | +0.195 |
ady | +0.195 |
Neg logit contributions | |
hill | -0.162 |
mountain | -0.150 |
road | -0.139 |
ladder | -0.137 |
subway | -0.137 |
Token | way |
Token ID | 835 |
Feature activation | +3.56e-02 |
Pos logit contributions | |
amed | +0.088 |
ished | +0.085 |
lling | +0.084 |
aming | +0.081 |
ady | +0.081 |
Neg logit contributions | |
hill | -0.073 |
mountain | -0.067 |
taxi | -0.067 |
road | -0.065 |
subway | -0.065 |
Token | . |
Token ID | 13 |
Feature activation | +1.01e-02 |
Pos logit contributions | |
hill | +0.058 |
mountain | +0.054 |
road | +0.049 |
ladder | +0.048 |
cliff | +0.048 |
Neg logit contributions | |
amed | -0.097 |
ished | -0.093 |
aming | -0.089 |
ady | -0.089 |
lling | -0.088 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -3.50e-02 |
Pos logit contributions | |
hill | +0.018 |
mountain | +0.017 |
road | +0.016 |
subway | +0.016 |
taxi | +0.015 |
Neg logit contributions | |
amed | -0.026 |
ished | -0.025 |
ady | -0.023 |
aming | -0.023 |
lling | -0.023 |
Token | George |
Token ID | 20191 |
Feature activation | -2.05e-02 |
Pos logit contributions | |
amed | +0.090 |
ished | +0.085 |
aming | +0.082 |
ady | +0.082 |
lling | +0.081 |
Neg logit contributions | |
hill | -0.061 |
mountain | -0.057 |
road | -0.053 |
taxi | -0.052 |
subway | -0.052 |
Token | was |
Token ID | 373 |
Feature activation | -1.41e-02 |
Pos logit contributions | |
amed | +0.050 |
ished | +0.047 |
lling | +0.045 |
asing | +0.045 |
aming | +0.045 |
Neg logit contributions | |
hill | -0.040 |
mountain | -0.037 |
road | -0.034 |
taxi | -0.033 |
subway | -0.033 |
Token | a |
Token ID | 257 |
Feature activation | +6.66e-02 |
Pos logit contributions | |
ished | +0.035 |
amed | +0.034 |
asing | +0.033 |
lling | +0.033 |
ealing | +0.033 |
Neg logit contributions | |
hill | -0.025 |
mountain | -0.023 |
taxi | -0.022 |
road | -0.022 |
ladder | -0.022 |
Token | little |
Token ID | 1310 |
Feature activation | +2.03e-02 |
Pos logit contributions | |
hill | +0.152 |
mountain | +0.141 |
subway | +0.139 |
cone | +0.139 |
Zig | +0.138 |
Neg logit contributions | |
amed | -0.114 |
aming | -0.107 |
ealing | -0.105 |
ady | -0.104 |
ished | -0.104 |
Token | Amy |
Token ID | 14235 |
Feature activation | -5.57e-02 |
Pos logit contributions | |
amed | +0.261 |
ished | +0.250 |
ady | +0.239 |
aming | +0.239 |
lling | +0.237 |
Neg logit contributions | |
hill | -0.197 |
mountain | -0.183 |
road | -0.171 |
taxi | -0.170 |
subway | -0.169 |
Token | 's |
Token ID | 338 |
Feature activation | -3.90e-02 |
Pos logit contributions | |
amed | +0.143 |
ished | +0.138 |
aming | +0.131 |
ady | +0.130 |
lling | +0.130 |
Neg logit contributions | |
hill | -0.099 |
mountain | -0.092 |
road | -0.085 |
ladder | -0.083 |
taxi | -0.083 |
Token | hand |
Token ID | 1021 |
Feature activation | -1.68e-02 |
Pos logit contributions | |
amed | +0.092 |
ished | +0.089 |
ady | +0.085 |
aming | +0.084 |
lling | +0.083 |
Neg logit contributions | |
hill | -0.072 |
mountain | -0.069 |
road | -0.063 |
cliff | -0.062 |
subway | -0.062 |
Token | and |
Token ID | 290 |
Feature activation | +6.71e-02 |
Pos logit contributions | |
amed | +0.041 |
ished | +0.039 |
aming | +0.037 |
lling | +0.037 |
ady | +0.037 |
Neg logit contributions | |
hill | -0.032 |
mountain | -0.030 |
road | -0.028 |
taxi | -0.028 |
ladder | -0.028 |
Token | the |
Token ID | 262 |
Feature activation | -2.19e-02 |
Pos logit contributions | |
hill | +0.138 |
mountain | +0.129 |
road | +0.122 |
cliff | +0.118 |
subway | +0.118 |
Neg logit contributions | |
amed | -0.154 |
ished | -0.147 |
aming | -0.139 |
lling | -0.138 |
ady | -0.137 |
Token | two |
Token ID | 734 |
Feature activation | +3.47e-03 |
Pos logit contributions | |
amed | +0.056 |
ished | +0.053 |
aming | +0.051 |
ady | +0.051 |
lling | +0.051 |
Neg logit contributions | |
hill | -0.037 |
mountain | -0.034 |
road | -0.032 |
subway | -0.030 |
cone | -0.030 |
Token | of |
Token ID | 286 |
Feature activation | +5.08e-02 |
Pos logit contributions | |
hill | +0.007 |
mountain | +0.007 |
taxi | +0.006 |
ladder | +0.006 |
road | +0.006 |
Neg logit contributions | |
amed | -0.008 |
ished | -0.008 |
ady | -0.007 |
aming | -0.007 |
lling | -0.007 |
Token | them |
Token ID | 606 |
Feature activation | -2.79e-02 |
Pos logit contributions | |
hill | +0.112 |
mountain | +0.106 |
taxi | +0.100 |
road | +0.099 |
subway | +0.098 |
Neg logit contributions | |
amed | -0.109 |
ished | -0.105 |
ady | -0.099 |
ealing | -0.098 |
asing | -0.097 |
Token | played |
Token ID | 2826 |
Feature activation | -4.74e-02 |
Pos logit contributions | |
amed | +0.065 |
ished | +0.062 |
ady | +0.059 |
aming | +0.059 |
lling | +0.058 |
Neg logit contributions | |
hill | -0.057 |
mountain | -0.052 |
road | -0.049 |
taxi | -0.049 |
subway | -0.049 |
Token | together |
Token ID | 1978 |
Feature activation | +5.43e-02 |
Pos logit contributions | |
amed | +0.122 |
ished | +0.119 |
ady | +0.113 |
aming | +0.112 |
lling | +0.111 |
Neg logit contributions | |
hill | -0.085 |
mountain | -0.077 |
road | -0.072 |
ladder | -0.070 |
subway | -0.070 |
Token | in |
Token ID | 287 |
Feature activation | +7.41e-02 |
Pos logit contributions | |
hill | +0.107 |
mountain | +0.100 |
road | +0.093 |
subway | +0.091 |
ladder | +0.091 |
Neg logit contributions | |
amed | -0.129 |
ished | -0.124 |
aming | -0.117 |
ady | -0.117 |
lling | -0.117 |
Token | to |
Token ID | 284 |
Feature activation | +6.92e-02 |
Pos logit contributions | |
amed | +0.215 |
ished | +0.210 |
lling | +0.189 |
aming | +0.186 |
ealing | +0.185 |
Neg logit contributions | |
hill | -0.172 |
mountain | -0.166 |
chim | -0.151 |
cliff | -0.150 |
volcano | -0.149 |
Token | Anna |
Token ID | 11735 |
Feature activation | -8.91e-02 |
Pos logit contributions | |
hill | +0.140 |
mountain | +0.133 |
subway | +0.124 |
cliff | +0.121 |
road | +0.121 |
Neg logit contributions | |
amed | -0.143 |
ished | -0.141 |
lling | -0.134 |
ps | -0.133 |
ady | -0.131 |
Token | and |
Token ID | 290 |
Feature activation | +5.10e-02 |
Pos logit contributions | |
amed | +0.212 |
ished | +0.199 |
aming | +0.190 |
lling | +0.187 |
ady | +0.186 |
Neg logit contributions | |
hill | -0.182 |
mountain | -0.169 |
road | -0.159 |
ladder | -0.155 |
taxi | -0.155 |
Token | jumps |
Token ID | 18045 |
Feature activation | -9.43e-02 |
Pos logit contributions | |
hill | +0.137 |
mountain | +0.129 |
road | +0.125 |
subway | +0.122 |
cliff | +0.121 |
Neg logit contributions | |
ished | -0.081 |
amed | -0.078 |
ealing | -0.070 |
ady | -0.069 |
ps | -0.069 |
Token | on |
Token ID | 319 |
Feature activation | +7.00e-02 |
Pos logit contributions | |
amed | +0.241 |
ished | +0.229 |
ady | +0.221 |
aming | +0.221 |
asing | +0.219 |
Neg logit contributions | |
hill | -0.172 |
mountain | -0.157 |
road | -0.148 |
ladder | -0.145 |
taxi | -0.145 |
Token | her |
Token ID | 607 |
Feature activation | -3.08e-02 |
Pos logit contributions | |
hill | +0.136 |
mountain | +0.127 |
road | +0.119 |
cliff | +0.119 |
subway | +0.119 |
Neg logit contributions | |
amed | -0.148 |
ished | -0.147 |
lling | -0.142 |
ps | -0.141 |
ady | -0.137 |
Token | . |
Token ID | 13 |
Feature activation | -1.30e-02 |
Pos logit contributions | |
amed | +0.073 |
ished | +0.070 |
aming | +0.066 |
ady | +0.066 |
lling | +0.065 |
Neg logit contributions | |
hill | -0.061 |
mountain | -0.058 |
road | -0.054 |
ladder | -0.053 |
cliff | -0.052 |
Token | He |
Token ID | 679 |
Feature activation | +6.44e-03 |
Pos logit contributions | |
amed | +0.030 |
ished | +0.029 |
ady | +0.027 |
ealing | +0.027 |
aming | +0.026 |
Neg logit contributions | |
hill | -0.026 |
mountain | -0.025 |
subway | -0.023 |
cliff | -0.023 |
road | -0.022 |
Token | tries |
Token ID | 8404 |
Feature activation | -6.18e-02 |
Pos logit contributions | |
hill | +0.015 |
mountain | +0.014 |
road | +0.014 |
subway | +0.013 |
ladder | +0.013 |
Neg logit contributions | |
amed | -0.013 |
ished | -0.012 |
aming | -0.011 |
ady | -0.011 |
lling | -0.011 |
Token | to |
Token ID | 284 |
Feature activation | +6.17e-02 |
Pos logit contributions | |
amed | +0.193 |
ished | +0.187 |
aming | +0.179 |
ady | +0.179 |
lling | +0.178 |
Neg logit contributions | |
hill | -0.077 |
mountain | -0.070 |
road | -0.062 |
subway | -0.059 |
cliff | -0.059 |
Token | bite |
Token ID | 13197 |
Feature activation | -3.55e-02 |
Pos logit contributions | |
hill | +0.112 |
mountain | +0.106 |
road | +0.096 |
subway | +0.096 |
cliff | +0.095 |
Neg logit contributions | |
amed | -0.155 |
ished | -0.149 |
lling | -0.141 |
aming | -0.141 |
ady | -0.141 |
Token | she |
Token ID | 673 |
Feature activation | -8.51e-03 |
Pos logit contributions | |
hill | +0.129 |
mountain | +0.118 |
road | +0.112 |
subway | +0.107 |
taxi | +0.104 |
Neg logit contributions | |
amed | -0.200 |
ished | -0.198 |
lling | -0.189 |
ealing | -0.187 |
aming | -0.187 |
Token | might |
Token ID | 1244 |
Feature activation | +5.92e-03 |
Pos logit contributions | |
amed | +0.019 |
ished | +0.018 |
ady | +0.017 |
aming | +0.017 |
lling | +0.017 |
Neg logit contributions | |
hill | -0.018 |
mountain | -0.017 |
road | -0.016 |
subway | -0.016 |
taxi | -0.016 |
Token | get |
Token ID | 651 |
Feature activation | -5.81e-02 |
Pos logit contributions | |
hill | +0.012 |
mountain | +0.011 |
subway | +0.011 |
taxi | +0.010 |
road | +0.010 |
Neg logit contributions | |
amed | -0.013 |
ished | -0.013 |
lling | -0.012 |
ady | -0.012 |
aming | -0.012 |
Token | angry |
Token ID | 7954 |
Feature activation | -8.49e-02 |
Pos logit contributions | |
amed | +0.134 |
ished | +0.128 |
aming | +0.121 |
ady | +0.120 |
lling | +0.120 |
Neg logit contributions | |
hill | -0.121 |
mountain | -0.113 |
road | -0.106 |
subway | -0.104 |
ladder | -0.103 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.57e-02 |
Pos logit contributions | |
amed | +0.202 |
ished | +0.194 |
aming | +0.183 |
ady | +0.183 |
lling | +0.182 |
Neg logit contributions | |
hill | -0.168 |
mountain | -0.157 |
road | -0.147 |
ladder | -0.144 |
cliff | -0.143 |
Token |
|
Token ID | 198 |
Feature activation | -2.28e-02 |
Pos logit contributions | |
amed | +0.045 |
ished | +0.043 |
lling | +0.041 |
aming | +0.041 |
ady | +0.041 |
Neg logit contributions | |
hill | -0.025 |
mountain | -0.022 |
road | -0.021 |
ladder | -0.020 |
cliff | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
amed | +0.066 |
ished | +0.065 |
lling | +0.063 |
aming | +0.061 |
asing | +0.061 |
Neg logit contributions | |
hill | -0.033 |
mountain | -0.030 |
road | -0.028 |
ladder | -0.027 |
cliff | -0.026 |
Token | They |
Token ID | 2990 |
Feature activation | -4.49e-03 |
Pos logit contributions | |
ished | +0.038 |
amed | +0.038 |
lling | +0.035 |
asing | +0.034 |
ady | +0.034 |
Neg logit contributions | |
hill | -0.037 |
mountain | -0.034 |
road | -0.033 |
bike | -0.032 |
cliff | -0.032 |
Token | watched |
Token ID | 7342 |
Feature activation | -8.55e-02 |
Pos logit contributions | |
amed | +0.010 |
ished | +0.010 |
aming | +0.009 |
ady | +0.009 |
lling | +0.009 |
Neg logit contributions | |
hill | -0.009 |
mountain | -0.009 |
road | -0.008 |
cliff | -0.008 |
taxi | -0.008 |
Token | the |
Token ID | 262 |
Feature activation | -2.42e-02 |
Pos logit contributions | |
amed | +0.206 |
ished | +0.199 |
lling | +0.189 |
aming | +0.188 |
ealing | +0.185 |
Neg logit contributions | |
hill | -0.164 |
mountain | -0.152 |
road | -0.141 |
subway | -0.138 |
cliff | -0.137 |
Token | queen |
Token ID | 16599 |
Feature activation | +2.16e-02 |
Pos logit contributions | |
amed | +0.058 |
ished | +0.054 |
lling | +0.053 |
aming | +0.052 |
ealing | +0.051 |
Neg logit contributions | |
hill | -0.044 |
mountain | -0.040 |
road | -0.039 |
subway | -0.039 |
bike | -0.038 |
Token | . |
Token ID | 13 |
Feature activation | -8.06e-03 |
Pos logit contributions | |
amed | +0.026 |
ished | +0.025 |
aming | +0.023 |
ady | +0.023 |
lling | +0.023 |
Neg logit contributions | |
hill | -0.021 |
mountain | -0.019 |
road | -0.018 |
ladder | -0.017 |
taxi | -0.017 |
Token | The |
Token ID | 383 |
Feature activation | -2.74e-02 |
Pos logit contributions | |
amed | +0.021 |
ished | +0.020 |
aming | +0.019 |
ady | +0.019 |
lling | +0.019 |
Neg logit contributions | |
hill | -0.015 |
mountain | -0.014 |
road | -0.013 |
subway | -0.012 |
cliff | -0.012 |
Token | apple |
Token ID | 17180 |
Feature activation | +7.79e-02 |
Pos logit contributions | |
amed | +0.077 |
ished | +0.073 |
aming | +0.070 |
ady | +0.070 |
lling | +0.069 |
Neg logit contributions | |
hill | -0.042 |
mountain | -0.039 |
subway | -0.035 |
road | -0.035 |
cliff | -0.034 |
Token | was |
Token ID | 373 |
Feature activation | -2.69e-02 |
Pos logit contributions | |
hill | +0.161 |
mountain | +0.151 |
road | +0.143 |
subway | +0.142 |
taxi | +0.140 |
Neg logit contributions | |
amed | -0.176 |
ished | -0.168 |
aming | -0.158 |
lling | -0.157 |
ady | -0.155 |
Token | not |
Token ID | 407 |
Feature activation | -4.97e-02 |
Pos logit contributions | |
amed | +0.069 |
ished | +0.067 |
asing | +0.066 |
ady | +0.064 |
ealing | +0.064 |
Neg logit contributions | |
hill | -0.045 |
mountain | -0.042 |
ladder | -0.040 |
road | -0.039 |
taxi | -0.039 |
Token | y |
Token ID | 331 |
Feature activation | -2.58e-02 |
Pos logit contributions | |
amed | +0.120 |
ished | +0.114 |
aming | +0.108 |
ady | +0.108 |
asing | +0.107 |
Neg logit contributions | |
hill | -0.097 |
mountain | -0.091 |
road | -0.085 |
ladder | -0.084 |
taxi | -0.082 |
Token | ummy |
Token ID | 13513 |
Feature activation | -1.10e-02 |
Pos logit contributions | |
amed | +0.051 |
ished | +0.046 |
lling | +0.044 |
aming | +0.044 |
ady | +0.044 |
Neg logit contributions | |
hill | -0.062 |
mountain | -0.058 |
road | -0.056 |
ladder | -0.055 |
cliff | -0.054 |
Token | . |
Token ID | 13 |
Feature activation | -2.77e-03 |
Pos logit contributions | |
amed | +0.026 |
ished | +0.025 |
aming | +0.024 |
ady | +0.024 |
lling | +0.024 |
Neg logit contributions | |
hill | -0.022 |
mountain | -0.020 |
road | -0.019 |
ladder | -0.019 |
subway | -0.019 |
Token | It |
Token ID | 632 |
Feature activation | +3.53e-02 |
Pos logit contributions | |
amed | +0.007 |
ished | +0.007 |
lling | +0.007 |
aming | +0.007 |
ady | +0.007 |
Neg logit contributions | |
hill | -0.005 |
mountain | -0.004 |
road | -0.004 |
ladder | -0.004 |
cliff | -0.004 |
Token | was |
Token ID | 373 |
Feature activation | -2.52e-02 |
Pos logit contributions | |
hill | +0.074 |
mountain | +0.069 |
road | +0.065 |
subway | +0.064 |
cliff | +0.063 |
Neg logit contributions | |
amed | -0.080 |
ished | -0.077 |
aming | -0.072 |
lling | -0.072 |
ady | -0.072 |
Token | bitter |
Token ID | 12922 |
Feature activation | -5.80e-02 |
Pos logit contributions | |
amed | +0.064 |
ished | +0.062 |
asing | +0.059 |
ady | +0.058 |
lling | +0.058 |
Neg logit contributions | |
hill | -0.045 |
mountain | -0.042 |
ladder | -0.039 |
road | -0.039 |
taxi | -0.038 |
Token | Her |
Token ID | 2332 |
Feature activation | -5.70e-02 |
Pos logit contributions | |
hill | +0.012 |
mountain | +0.011 |
road | +0.010 |
ladder | +0.010 |
subway | +0.010 |
Neg logit contributions | |
amed | -0.018 |
ished | -0.017 |
lling | -0.016 |
aming | -0.016 |
ady | -0.016 |
Token | tooth |
Token ID | 16162 |
Feature activation | -1.65e-03 |
Pos logit contributions | |
amed | +0.133 |
ished | +0.127 |
aming | +0.120 |
ady | +0.120 |
lling | +0.120 |
Neg logit contributions | |
hill | -0.117 |
mountain | -0.109 |
road | -0.102 |
ladder | -0.100 |
subway | -0.100 |
Token | felt |
Token ID | 2936 |
Feature activation | -9.37e-02 |
Pos logit contributions | |
amed | +0.004 |
ished | +0.004 |
aming | +0.004 |
lling | +0.004 |
ady | +0.004 |
Neg logit contributions | |
hill | -0.003 |
mountain | -0.003 |
ladder | -0.003 |
road | -0.003 |
bike | -0.003 |
Token | dull |
Token ID | 19222 |
Feature activation | -5.99e-02 |
Pos logit contributions | |
amed | +0.242 |
ished | +0.236 |
asing | +0.228 |
ady | +0.224 |
wed | +0.219 |
Neg logit contributions | |
hill | -0.145 |
mountain | -0.136 |
road | -0.127 |
volcano | -0.127 |
cliff | -0.126 |
Token | and |
Token ID | 290 |
Feature activation | +7.73e-02 |
Pos logit contributions | |
amed | +0.153 |
ished | +0.144 |
ady | +0.138 |
aming | +0.138 |
ealing | +0.137 |
Neg logit contributions | |
hill | -0.110 |
mountain | -0.101 |
road | -0.094 |
ladder | -0.091 |
bike | -0.091 |
Token | w |
Token ID | 266 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
hill | +0.137 |
mountain | +0.127 |
road | +0.116 |
ladder | +0.116 |
cliff | +0.114 |
Neg logit contributions | |
amed | -0.201 |
ished | -0.193 |
ady | -0.186 |
aming | -0.182 |
asing | -0.181 |
Token | obb |
Token ID | 21963 |
Feature activation | -6.92e-02 |
Pos logit contributions | |
amed | +0.155 |
lling | +0.138 |
ady | +0.134 |
aming | +0.129 |
ished | +0.128 |
Neg logit contributions | |
hill | -0.247 |
mountain | -0.237 |
cliff | -0.233 |
volcano | -0.230 |
subway | -0.225 |
Token | ly |
Token ID | 306 |
Feature activation | +1.92e-03 |
Pos logit contributions | |
amed | +0.169 |
ady | +0.153 |
aming | +0.151 |
asing | +0.143 |
ished | +0.139 |
Neg logit contributions | |
hill | -0.117 |
mountain | -0.109 |
road | -0.105 |
je | -0.103 |
bike | -0.103 |
Token | . |
Token ID | 13 |
Feature activation | +3.47e-03 |
Pos logit contributions | |
hill | +0.004 |
mountain | +0.003 |
road | +0.003 |
ladder | +0.003 |
cliff | +0.003 |
Neg logit contributions | |
amed | -0.005 |
ished | -0.004 |
aming | -0.004 |
ady | -0.004 |
lling | -0.004 |
Token | She |
Token ID | 1375 |
Feature activation | +8.74e-03 |
Pos logit contributions | |
hill | +0.006 |
mountain | +0.005 |
road | +0.005 |
ladder | +0.005 |
subway | +0.005 |
Neg logit contributions | |
amed | -0.009 |
ished | -0.009 |
lling | -0.009 |
aming | -0.009 |
asing | -0.009 |
Token | quickly |
Token ID | 2952 |
Feature activation | -2.03e-02 |
Pos logit contributions | |
hill | +0.018 |
mountain | +0.017 |
road | +0.016 |
taxi | +0.016 |
ladder | +0.016 |
Neg logit contributions | |
amed | -0.019 |
ished | -0.019 |
aming | -0.018 |
ady | -0.018 |
asing | -0.017 |
Token | in |
Token ID | 287 |
Feature activation | +9.49e-02 |
Pos logit contributions | |
hill | +0.142 |
mountain | +0.133 |
road | +0.124 |
subway | +0.122 |
taxi | +0.122 |
Neg logit contributions | |
amed | -0.162 |
ished | -0.155 |
aming | -0.147 |
lling | -0.147 |
ady | -0.147 |
Token | a |
Token ID | 257 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
hill | +0.165 |
mountain | +0.152 |
road | +0.141 |
ladder | +0.137 |
subway | +0.137 |
Neg logit contributions | |
amed | -0.250 |
ished | -0.240 |
aming | -0.228 |
ady | -0.228 |
lling | -0.227 |
Token | distant |
Token ID | 12899 |
Feature activation | +6.07e-02 |
Pos logit contributions | |
hill | +0.159 |
mountain | +0.144 |
subway | +0.141 |
Zig | +0.141 |
ladder | +0.137 |
Neg logit contributions | |
amed | -0.275 |
ady | -0.259 |
aming | -0.256 |
ished | -0.256 |
lling | -0.254 |
Token | land |
Token ID | 1956 |
Feature activation | +6.62e-02 |
Pos logit contributions | |
hill | +0.063 |
subway | +0.057 |
cone | +0.057 |
taxi | +0.055 |
ladder | +0.055 |
Neg logit contributions | |
amed | -0.175 |
ady | -0.167 |
ished | -0.166 |
lling | -0.164 |
aming | -0.163 |
Token | . |
Token ID | 13 |
Feature activation | +1.18e-02 |
Pos logit contributions | |
hill | +0.133 |
mountain | +0.125 |
road | +0.116 |
subway | +0.114 |
ladder | +0.114 |
Neg logit contributions | |
amed | -0.154 |
ished | -0.147 |
aming | -0.140 |
ady | -0.139 |
lling | -0.138 |
Token | |
Token ID | 220 |
Feature activation | -5.96e-02 |
Pos logit contributions | |
hill | +0.024 |
mountain | +0.023 |
cliff | +0.021 |
subway | +0.021 |
road | +0.021 |
Neg logit contributions | |
amed | -0.026 |
ished | -0.025 |
ady | -0.024 |
aming | -0.023 |
ealing | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | -4.24e-03 |
Pos logit contributions | |
amed | +0.179 |
ished | +0.176 |
lling | +0.173 |
aming | +0.169 |
asing | +0.166 |
Neg logit contributions | |
hill | -0.077 |
mountain | -0.067 |
ladder | -0.065 |
road | -0.063 |
taxi | -0.060 |
Token |
|
Token ID | 198 |
Feature activation | -1.60e-03 |
Pos logit contributions | |
amed | +0.011 |
ished | +0.011 |
aming | +0.010 |
lling | +0.010 |
ady | +0.010 |
Neg logit contributions | |
hill | -0.007 |
mountain | -0.007 |
road | -0.006 |
ladder | -0.006 |
taxi | -0.006 |
Token | Exc |
Token ID | 40127 |
Feature activation | -4.70e-02 |
Pos logit contributions | |
amed | +0.004 |
ished | +0.004 |
aming | +0.003 |
ady | +0.003 |
lling | +0.003 |
Neg logit contributions | |
hill | -0.003 |
mountain | -0.003 |
road | -0.003 |
ladder | -0.003 |
cliff | -0.003 |
Token | ited |
Token ID | 863 |
Feature activation | -4.65e-02 |
Pos logit contributions | |
amed | +0.070 |
ished | +0.065 |
aming | +0.060 |
lling | +0.059 |
ady | +0.058 |
Neg logit contributions | |
hill | -0.135 |
mountain | -0.129 |
road | -0.124 |
ladder | -0.123 |
cliff | -0.122 |
Token | to |
Token ID | 284 |
Feature activation | +8.96e-02 |
Pos logit contributions | |
amed | +0.096 |
ished | +0.093 |
aming | +0.086 |
lling | +0.086 |
ady | +0.085 |
Neg logit contributions | |
hill | -0.107 |
mountain | -0.101 |
road | -0.095 |
ladder | -0.093 |
cliff | -0.093 |
Token | â |
Token ID | 22940 |
Feature activation | -1.61e-01 |
Pos logit contributions | |
hill | +0.021 |
mountain | +0.020 |
road | +0.019 |
cliff | +0.018 |
subway | +0.018 |
Neg logit contributions | |
amed | -0.027 |
ished | -0.026 |
ady | -0.025 |
aming | -0.025 |
lling | -0.024 |
Token | € |
Token ID | 26391 |
Feature activation | -9.40e-02 |
Pos logit contributions | |
amed | +0.346 |
ished | +0.341 |
ady | +0.316 |
lling | +0.311 |
asing | +0.305 |
Neg logit contributions | |
hill | -0.349 |
mountain | -0.321 |
road | -0.308 |
subway | -0.307 |
taxi | -0.299 |
Token | The |
Token ID | 383 |
Feature activation | -3.18e-02 |
Pos logit contributions | |
amed | +0.231 |
ished | +0.220 |
aming | +0.212 |
ealing | +0.212 |
ady | +0.212 |
Neg logit contributions | |
hill | -0.172 |
mountain | -0.162 |
road | -0.150 |
cliff | -0.148 |
taxi | -0.146 |
Token | mom |
Token ID | 1995 |
Feature activation | -3.35e-02 |
Pos logit contributions | |
amed | +0.075 |
ished | +0.070 |
aming | +0.068 |
ady | +0.067 |
lling | +0.066 |
Neg logit contributions | |
hill | -0.062 |
mountain | -0.058 |
road | -0.054 |
cliff | -0.054 |
ladder | -0.052 |
Token | my |
Token ID | 1820 |
Feature activation | -6.83e-02 |
Pos logit contributions | |
amed | +0.057 |
ished | +0.052 |
aming | +0.050 |
lling | +0.049 |
ady | +0.048 |
Neg logit contributions | |
hill | -0.088 |
mountain | -0.085 |
taxi | -0.080 |
ladder | -0.080 |
road | -0.079 |
Token | â |
Token ID | 22940 |
Feature activation | -1.57e-01 |
Pos logit contributions | |
amed | +0.145 |
ished | +0.136 |
aming | +0.130 |
lling | +0.129 |
ady | +0.128 |
Neg logit contributions | |
hill | -0.152 |
mountain | -0.145 |
ladder | -0.133 |
road | -0.133 |
taxi | -0.133 |
Token | € |
Token ID | 26391 |
Feature activation | -9.26e-02 |
Pos logit contributions | |
amed | +0.336 |
ished | +0.328 |
ady | +0.306 |
lling | +0.304 |
aming | +0.299 |
Neg logit contributions | |
hill | -0.341 |
mountain | -0.313 |
road | -0.302 |
subway | -0.302 |
taxi | -0.296 |
Token | ™ |
Token ID | 8151 |
Feature activation | -8.50e-03 |
Pos logit contributions | |
amed | +0.222 |
ished | +0.211 |
ealing | +0.204 |
ady | +0.204 |
aming | +0.204 |
Neg logit contributions | |
hill | -0.173 |
mountain | -0.164 |
road | -0.152 |
cliff | -0.150 |
taxi | -0.148 |
Token | s |
Token ID | 82 |
Feature activation | -3.93e-02 |
Pos logit contributions | |
amed | +0.020 |
ished | +0.019 |
aming | +0.018 |
lling | +0.018 |
ady | +0.018 |
Neg logit contributions | |
hill | -0.017 |
mountain | -0.016 |
road | -0.015 |
ladder | -0.014 |
taxi | -0.014 |
Token | sadness |
Token ID | 25303 |
Feature activation | -6.94e-02 |
Pos logit contributions | |
amed | +0.093 |
ished | +0.089 |
aming | +0.084 |
ady | +0.083 |
lling | +0.083 |
Neg logit contributions | |
hill | -0.079 |
mountain | -0.074 |
road | -0.069 |
ladder | -0.067 |
subway | -0.067 |
Token | disappeared |
Token ID | 12120 |
Feature activation | -1.83e-02 |
Pos logit contributions | |
amed | +0.172 |
ished | +0.165 |
aming | +0.157 |
ady | +0.157 |
lling | +0.154 |
Neg logit contributions | |
hill | -0.132 |
mountain | -0.122 |
road | -0.113 |
ladder | -0.112 |
subway | -0.111 |
Token | it |
Token ID | 340 |
Feature activation | +1.73e-02 |
Pos logit contributions | |
hill | +0.021 |
mountain | +0.020 |
road | +0.019 |
cliff | +0.018 |
subway | +0.018 |
Neg logit contributions | |
ished | -0.028 |
amed | -0.028 |
lling | -0.026 |
aming | -0.026 |
ady | -0.026 |
Token | together |
Token ID | 1978 |
Feature activation | +5.99e-02 |
Pos logit contributions | |
hill | +0.036 |
mountain | +0.034 |
road | +0.032 |
cliff | +0.031 |
subway | +0.031 |
Neg logit contributions | |
amed | -0.039 |
ished | -0.038 |
aming | -0.035 |
ady | -0.035 |
lling | -0.035 |
Token | ." |
Token ID | 526 |
Feature activation | -7.32e-03 |
Pos logit contributions | |
hill | +0.114 |
mountain | +0.105 |
road | +0.097 |
ladder | +0.096 |
cliff | +0.095 |
Neg logit contributions | |
amed | -0.150 |
ished | -0.143 |
aming | -0.135 |
ady | -0.135 |
lling | -0.134 |
Token |
|
Token ID | 198 |
Feature activation | +6.90e-04 |
Pos logit contributions | |
amed | +0.016 |
ished | +0.016 |
ady | +0.015 |
aming | +0.014 |
ealing | +0.014 |
Neg logit contributions | |
hill | -0.015 |
mountain | -0.015 |
cliff | -0.014 |
subway | -0.014 |
road | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | +4.10e-03 |
Pos logit contributions | |
hill | +0.001 |
mountain | +0.001 |
road | +0.001 |
taxi | +0.001 |
ladder | +0.001 |
Neg logit contributions | |
amed | -0.002 |
ished | -0.002 |
lling | -0.002 |
aming | -0.002 |
asing | -0.002 |
Token | L |
Token ID | 43 |
Feature activation | -1.59e-01 |
Pos logit contributions | |
hill | +0.010 |
mountain | +0.009 |
road | +0.009 |
ladder | +0.008 |
cliff | +0.008 |
Neg logit contributions | |
amed | -0.008 |
ished | -0.008 |
aming | -0.007 |
ady | -0.007 |
lling | -0.007 |
Token | ily |
Token ID | 813 |
Feature activation | -9.66e-02 |
Pos logit contributions | |
amed | +0.231 |
ished | +0.213 |
ealing | +0.212 |
lling | +0.211 |
aming | +0.196 |
Neg logit contributions | |
hill | -0.404 |
mountain | -0.387 |
taxi | -0.382 |
je | -0.378 |
library | -0.376 |
Token | opened |
Token ID | 4721 |
Feature activation | -2.49e-02 |
Pos logit contributions | |
amed | +0.224 |
lling | +0.215 |
aming | +0.215 |
ady | +0.213 |
ealing | +0.212 |
Neg logit contributions | |
hill | -0.193 |
mountain | -0.178 |
ladder | -0.168 |
bike | -0.160 |
road | -0.158 |
Token | the |
Token ID | 262 |
Feature activation | -6.24e-04 |
Pos logit contributions | |
amed | +0.063 |
ished | +0.060 |
aming | +0.057 |
ady | +0.056 |
lling | +0.056 |
Neg logit contributions | |
hill | -0.046 |
mountain | -0.042 |
road | -0.040 |
subway | -0.039 |
cliff | -0.038 |
Token | package |
Token ID | 5301 |
Feature activation | +9.23e-02 |
Pos logit contributions | |
amed | +0.002 |
ished | +0.001 |
ady | +0.001 |
aming | +0.001 |
ealing | +0.001 |
Neg logit contributions | |
hill | -0.001 |
mountain | -0.001 |
Zig | -0.001 |
road | -0.001 |
saf | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | +7.53e-02 |
Pos logit contributions | |
hill | +0.182 |
mountain | +0.170 |
road | +0.159 |
ladder | +0.156 |
subway | +0.155 |
Neg logit contributions | |
amed | -0.220 |
ished | -0.211 |
aming | -0.200 |
ady | -0.199 |
lling | -0.198 |
Token | . |
Token ID | 13 |
Feature activation | +3.80e-03 |
Pos logit contributions | |
amed | +0.228 |
ished | +0.218 |
aming | +0.208 |
lling | +0.206 |
ady | +0.206 |
Neg logit contributions | |
hill | -0.161 |
mountain | -0.149 |
road | -0.138 |
ladder | -0.135 |
taxi | -0.135 |
Token |
|
Token ID | 198 |
Feature activation | -1.52e-03 |
Pos logit contributions | |
hill | +0.007 |
mountain | +0.007 |
road | +0.006 |
cliff | +0.006 |
subway | +0.006 |
Neg logit contributions | |
amed | -0.009 |
ished | -0.009 |
ady | -0.008 |
aming | -0.008 |
ealing | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | +1.03e-03 |
Pos logit contributions | |
amed | +0.004 |
ished | +0.004 |
lling | +0.004 |
aming | +0.004 |
asing | +0.004 |
Neg logit contributions | |
hill | -0.002 |
mountain | -0.002 |
road | -0.002 |
taxi | -0.002 |
ladder | -0.002 |
Token | His |
Token ID | 6653 |
Feature activation | -3.79e-02 |
Pos logit contributions | |
hill | +0.002 |
mountain | +0.002 |
road | +0.002 |
taxi | +0.002 |
subway | +0.002 |
Neg logit contributions | |
amed | -0.003 |
ished | -0.003 |
lling | -0.002 |
aming | -0.002 |
ady | -0.002 |
Token | mom |
Token ID | 1995 |
Feature activation | -3.46e-02 |
Pos logit contributions | |
amed | +0.092 |
ished | +0.089 |
ady | +0.084 |
aming | +0.083 |
ealing | +0.082 |
Neg logit contributions | |
hill | -0.072 |
mountain | -0.068 |
road | -0.062 |
cliff | -0.062 |
subway | -0.061 |
Token | smiled |
Token ID | 13541 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
amed | +0.077 |
ished | +0.075 |
ealing | +0.071 |
aming | +0.071 |
lling | +0.071 |
Neg logit contributions | |
hill | -0.068 |
mountain | -0.064 |
taxi | -0.062 |
subway | -0.060 |
je | -0.059 |
Token | . |
Token ID | 13 |
Feature activation | +1.44e-03 |
Pos logit contributions | |
ished | +0.404 |
amed | +0.400 |
asing | +0.392 |
wed | +0.386 |
lling | +0.382 |
Neg logit contributions | |
hill | -0.203 |
mountain | -0.188 |
taxi | -0.185 |
bike | -0.179 |
road | -0.171 |
Token | � |
Token ID | 6184 |
Feature activation | -9.70e-02 |
Pos logit contributions | |
hill | +0.003 |
mountain | +0.002 |
road | +0.002 |
ladder | +0.002 |
taxi | +0.002 |
Neg logit contributions | |
amed | -0.004 |
ished | -0.004 |
aming | -0.003 |
lling | -0.003 |
ady | -0.003 |
Token | � |
Token ID | 95 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
amed | +0.204 |
ished | +0.197 |
ealing | +0.186 |
aming | +0.185 |
ady | +0.182 |
Neg logit contributions | |
hill | -0.217 |
mountain | -0.206 |
road | -0.193 |
cliff | -0.188 |
ladder | -0.188 |
Token | € |
Token ID | 26391 |
Feature activation | -9.66e-02 |
Pos logit contributions | |
amed | +0.272 |
ished | +0.264 |
lling | +0.247 |
ady | +0.240 |
aming | +0.239 |
Neg logit contributions | |
hill | -0.270 |
mountain | -0.251 |
road | -0.244 |
subway | -0.242 |
taxi | -0.234 |
Token | � |
Token ID | 129 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
amed | +0.231 |
ished | +0.221 |
ady | +0.211 |
aming | +0.211 |
ealing | +0.210 |
Neg logit contributions | |
hill | -0.186 |
mountain | -0.175 |
road | -0.163 |
cliff | -0.160 |
taxi | -0.159 |
Token | . |
Token ID | 13 |
Feature activation | +8.74e-03 |
Pos logit contributions | |
hill | +0.105 |
mountain | +0.099 |
subway | +0.095 |
road | +0.093 |
taxi | +0.092 |
Neg logit contributions | |
amed | -0.094 |
ished | -0.089 |
lling | -0.087 |
aming | -0.085 |
ady | -0.082 |
Token | He |
Token ID | 679 |
Feature activation | +1.17e-02 |
Pos logit contributions | |
hill | +0.013 |
mountain | +0.012 |
road | +0.011 |
ladder | +0.010 |
volcano | +0.010 |
Neg logit contributions | |
amed | -0.026 |
ished | -0.024 |
lling | -0.024 |
aming | -0.024 |
ady | -0.024 |
Token | and |
Token ID | 290 |
Feature activation | +5.76e-02 |
Pos logit contributions | |
hill | +0.026 |
mountain | +0.025 |
road | +0.023 |
cliff | +0.023 |
ladder | +0.023 |
Neg logit contributions | |
amed | -0.025 |
ished | -0.024 |
aming | -0.023 |
ady | -0.022 |
lling | -0.022 |
Token | the |
Token ID | 262 |
Feature activation | -3.22e-02 |
Pos logit contributions | |
hill | +0.110 |
mountain | +0.103 |
road | +0.096 |
subway | +0.096 |
ladder | +0.095 |
Neg logit contributions | |
amed | -0.140 |
ished | -0.135 |
lling | -0.128 |
aming | -0.128 |
ady | -0.125 |
Token | monster |
Token ID | 9234 |
Feature activation | +2.97e-02 |
Pos logit contributions | |
amed | +0.089 |
ished | +0.084 |
lling | +0.083 |
aming | +0.082 |
ady | +0.080 |
Neg logit contributions | |
hill | -0.045 |
mountain | -0.041 |
road | -0.038 |
taxi | -0.038 |
ladder | -0.038 |
Token | smiled |
Token ID | 13541 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
hill | +0.064 |
mountain | +0.060 |
road | +0.056 |
cliff | +0.056 |
ladder | +0.056 |
Neg logit contributions | |
amed | -0.065 |
ished | -0.062 |
lling | -0.057 |
aming | -0.057 |
ady | -0.056 |
Token | at |
Token ID | 379 |
Feature activation | +3.95e-02 |
Pos logit contributions | |
amed | +0.330 |
ished | +0.320 |
ady | +0.306 |
lling | +0.306 |
asing | +0.302 |
Neg logit contributions | |
hill | -0.239 |
mountain | -0.224 |
road | -0.207 |
taxi | -0.202 |
bike | -0.202 |
Token | each |
Token ID | 1123 |
Feature activation | -4.80e-02 |
Pos logit contributions | |
hill | +0.067 |
mountain | +0.062 |
volcano | +0.056 |
road | +0.056 |
festival | +0.054 |
Neg logit contributions | |
amed | -0.108 |
ished | -0.102 |
ady | -0.101 |
asing | -0.098 |
ealing | -0.096 |
Token | other |
Token ID | 584 |
Feature activation | -2.89e-02 |
Pos logit contributions | |
amed | +0.119 |
ished | +0.112 |
aming | +0.107 |
ealing | +0.104 |
ady | +0.104 |
Neg logit contributions | |
hill | -0.096 |
mountain | -0.088 |
road | -0.084 |
taxi | -0.081 |
subway | -0.081 |
Token | and |
Token ID | 290 |
Feature activation | +6.40e-02 |
Pos logit contributions | |
amed | +0.070 |
ished | +0.066 |
aming | +0.063 |
ady | +0.063 |
lling | +0.062 |
Neg logit contributions | |
hill | -0.057 |
mountain | -0.053 |
road | -0.050 |
subway | -0.048 |
taxi | -0.048 |
Token | laughed |
Token ID | 13818 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
hill | +0.138 |
mountain | +0.130 |
road | +0.122 |
ladder | +0.119 |
subway | +0.119 |
Neg logit contributions | |
amed | -0.142 |
ished | -0.135 |
aming | -0.127 |
ady | -0.127 |
lling | -0.126 |
Token | Ryan |
Token ID | 6047 |
Feature activation | -1.09e-01 |
Pos logit contributions | |
amed | +0.040 |
ished | +0.038 |
aming | +0.037 |
lling | +0.036 |
ady | +0.036 |
Neg logit contributions | |
hill | -0.027 |
mountain | -0.025 |
road | -0.023 |
ladder | -0.023 |
cliff | -0.023 |
Token | opening |
Token ID | 4756 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
amed | +0.238 |
ished | +0.228 |
aming | +0.214 |
ady | +0.213 |
lling | +0.210 |
Neg logit contributions | |
hill | -0.236 |
mountain | -0.222 |
road | -0.207 |
subway | -0.204 |
taxi | -0.203 |
Token | the |
Token ID | 262 |
Feature activation | +7.27e-04 |
Pos logit contributions | |
amed | +0.039 |
ished | +0.037 |
aming | +0.035 |
ady | +0.035 |
lling | +0.035 |
Neg logit contributions | |
hill | -0.026 |
mountain | -0.024 |
road | -0.023 |
cliff | -0.022 |
ladder | -0.022 |
Token | box |
Token ID | 3091 |
Feature activation | +1.03e-01 |
Pos logit contributions | |
hill | +0.001 |
mountain | +0.001 |
taxi | +0.001 |
Zig | +0.001 |
road | +0.001 |
Neg logit contributions | |
amed | -0.002 |
ished | -0.002 |
ady | -0.002 |
aming | -0.002 |
lling | -0.002 |
Token | and |
Token ID | 290 |
Feature activation | +6.78e-02 |
Pos logit contributions | |
hill | +0.197 |
mountain | +0.184 |
road | +0.171 |
ladder | +0.167 |
taxi | +0.166 |
Neg logit contributions | |
amed | -0.253 |
ished | -0.243 |
aming | -0.230 |
ady | -0.229 |
lling | -0.228 |
Token | being |
Token ID | 852 |
Feature activation | -1.23e-01 |
Pos logit contributions | |
hill | +0.142 |
mountain | +0.134 |
road | +0.125 |
ladder | +0.122 |
cliff | +0.122 |
Neg logit contributions | |
amed | -0.154 |
ished | -0.147 |
aming | -0.139 |
ady | -0.138 |
lling | -0.137 |
Token | surprised |
Token ID | 6655 |
Feature activation | -1.00e-01 |
Pos logit contributions | |
ished | +0.259 |
ealing | +0.245 |
amed | +0.242 |
oking | +0.240 |
lling | +0.240 |
Neg logit contributions | |
hill | -0.237 |
mountain | -0.231 |
taxi | -0.225 |
road | -0.216 |
bike | -0.215 |
Token | by |
Token ID | 416 |
Feature activation | +1.39e-03 |
Pos logit contributions | |
amed | +0.272 |
ished | +0.258 |
aming | +0.248 |
ady | +0.248 |
lling | +0.244 |
Neg logit contributions | |
hill | -0.169 |
mountain | -0.156 |
road | -0.144 |
taxi | -0.139 |
subway | -0.138 |
Token | the |
Token ID | 262 |
Feature activation | -2.40e-02 |
Pos logit contributions | |
hill | +0.003 |
mountain | +0.002 |
road | +0.002 |
subway | +0.002 |
taxi | +0.002 |
Neg logit contributions | |
amed | -0.004 |
ished | -0.003 |
aming | -0.003 |
ady | -0.003 |
ealing | -0.003 |
Token | beautiful |
Token ID | 4950 |
Feature activation | -1.57e-02 |
Pos logit contributions | |
amed | +0.057 |
ished | +0.054 |
lling | +0.053 |
aming | +0.053 |
ady | +0.052 |
Neg logit contributions | |
hill | -0.042 |
mountain | -0.039 |
ladder | -0.036 |
road | -0.036 |
taxi | -0.036 |
Token | marble |
Token ID | 30623 |
Feature activation | +3.59e-03 |
Pos logit contributions | |
amed | +0.040 |
ished | +0.039 |
lling | +0.037 |
ady | +0.037 |
aming | +0.036 |
Neg logit contributions | |
hill | -0.025 |
mountain | -0.023 |
subway | -0.021 |
ladder | -0.021 |
taxi | -0.021 |