Token | cooked |
Token ID | 46591 |
Feature activation | +8.87e-01 |
Pos logit contributions | |
cream | +0.167 |
Fin | +0.161 |
Blue | +0.161 |
Jerry | +0.155 |
Red | +0.151 |
Neg logit contributions | |
otion | -0.211 |
upon | -0.200 |
acy | -0.187 |
ids | -0.186 |
ancer | -0.186 |
Token | and |
Token ID | 290 |
Feature activation | -8.57e-01 |
Pos logit contributions | |
cream | +0.045 |
Fin | +0.040 |
Blue | +0.039 |
oc | +0.037 |
Jerry | +0.037 |
Neg logit contributions | |
otion | -0.063 |
upon | -0.056 |
laptop | -0.055 |
Deal | -0.054 |
spine | -0.054 |
Token | s |
Token ID | 264 |
Feature activation | -1.56e+00 |
Pos logit contributions | |
otion | +0.060 |
upon | +0.054 |
laptop | +0.054 |
Deal | +0.052 |
sister | +0.052 |
Neg logit contributions | |
cream | -0.043 |
Fin | -0.039 |
Blue | -0.039 |
Jerry | -0.036 |
Red | -0.035 |
Token | ous |
Token ID | 516 |
Feature activation | -2.68e+00 |
Pos logit contributions | |
otion | +0.087 |
upon | +0.075 |
laptop | +0.074 |
Deal | +0.072 |
headache | +0.071 |
Neg logit contributions | |
cream | -0.099 |
Fin | -0.094 |
Blue | -0.094 |
Jerry | -0.088 |
Tweet | -0.088 |
Token | - |
Token ID | 12 |
Feature activation | +3.36e+00 |
Pos logit contributions | |
otion | +0.175 |
upon | +0.158 |
laptop | +0.157 |
sister | +0.151 |
Deal | +0.150 |
Neg logit contributions | |
cream | -0.146 |
Fin | -0.134 |
Blue | -0.132 |
Jerry | -0.124 |
Red | -0.122 |
Token | v |
Token ID | 85 |
Feature activation | +1.41e+01 |
Pos logit contributions | |
cream | +0.152 |
Fin | +0.151 |
Blue | +0.149 |
Jerry | +0.143 |
Red | +0.138 |
Neg logit contributions | |
otion | -0.214 |
upon | -0.204 |
ids | -0.192 |
acy | -0.191 |
ancer | -0.190 |
Token | ide |
Token ID | 485 |
Feature activation | +2.47e+00 |
Pos logit contributions | |
cream | +0.837 |
Fin | +0.749 |
oc | +0.726 |
aw | +0.725 |
Blue | +0.708 |
Neg logit contributions | |
otion | -0.831 |
spine | -0.760 |
laptop | -0.758 |
Deal | -0.753 |
upon | -0.727 |
Token | and |
Token ID | 290 |
Feature activation | -8.99e-01 |
Pos logit contributions | |
cream | +0.132 |
Fin | +0.118 |
Blue | +0.116 |
oc | +0.108 |
Tweet | +0.106 |
Neg logit contributions | |
otion | -0.167 |
laptop | -0.150 |
Deal | -0.148 |
upon | -0.148 |
spine | -0.144 |
Token | smoked |
Token ID | 21603 |
Feature activation | +3.02e-02 |
Pos logit contributions | |
otion | +0.056 |
upon | +0.050 |
laptop | +0.050 |
Deal | +0.048 |
sister | +0.048 |
Neg logit contributions | |
cream | -0.052 |
Fin | -0.048 |
Blue | -0.047 |
Jerry | -0.044 |
Red | -0.044 |
Token | and |
Token ID | 290 |
Feature activation | -8.79e-01 |
Pos logit contributions | |
cream | +0.002 |
Fin | +0.001 |
Blue | +0.001 |
Red | +0.001 |
oc | +0.001 |
Neg logit contributions | |
otion | -0.002 |
laptop | -0.002 |
upon | -0.002 |
Deal | -0.002 |
spine | -0.002 |
Token | cured |
Token ID | 30703 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
otion | +0.056 |
upon | +0.049 |
laptop | +0.049 |
Deal | +0.048 |
sister | +0.047 |
Neg logit contributions | |
cream | -0.050 |
Fin | -0.046 |
Blue | -0.045 |
Jerry | -0.042 |
Red | -0.042 |
Token | man |
Token ID | 582 |
Feature activation | +5.14e-01 |
Pos logit contributions | |
cream | +0.066 |
Fin | +0.059 |
Blue | +0.058 |
Jerry | +0.054 |
Red | +0.053 |
Neg logit contributions | |
otion | -0.100 |
upon | -0.091 |
laptop | -0.091 |
sister | -0.088 |
Deal | -0.088 |
Token | was |
Token ID | 373 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
Fin | +0.028 |
Blue | +0.026 |
cream | +0.025 |
rob | +0.025 |
Nut | +0.025 |
Neg logit contributions | |
upon | -0.031 |
otion | -0.031 |
sister | -0.030 |
laptop | -0.029 |
brother | -0.028 |
Token | called |
Token ID | 1444 |
Feature activation | -4.16e-01 |
Pos logit contributions | |
otion | +0.010 |
upon | +0.009 |
laptop | +0.009 |
sister | +0.009 |
Deal | +0.009 |
Neg logit contributions | |
cream | -0.010 |
Fin | -0.009 |
Blue | -0.009 |
oc | -0.008 |
Jerry | -0.008 |
Token | a |
Token ID | 257 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
otion | +0.029 |
spine | +0.027 |
Deal | +0.026 |
laptop | +0.026 |
headache | +0.026 |
Neg logit contributions | |
cream | -0.020 |
Blue | -0.017 |
Fin | -0.017 |
sold | -0.016 |
Jerry | -0.016 |
Token | " |
Token ID | 366 |
Feature activation | -4.19e+00 |
Pos logit contributions | |
otion | +0.106 |
upon | +0.093 |
laptop | +0.092 |
Deal | +0.092 |
spine | +0.091 |
Neg logit contributions | |
cream | -0.071 |
Fin | -0.064 |
Blue | -0.064 |
Jerry | -0.059 |
oc | -0.058 |
Token | v |
Token ID | 85 |
Feature activation | +1.38e+01 |
Pos logit contributions | |
otion | +0.308 |
upon | +0.299 |
ancer | +0.283 |
laptop | +0.278 |
acy | +0.268 |
Neg logit contributions | |
cream | -0.160 |
Fin | -0.150 |
Blue | -0.147 |
Jerry | -0.140 |
sav | -0.134 |
Token | endor |
Token ID | 18738 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
cream | +0.768 |
oc | +0.700 |
Fin | +0.688 |
aw | +0.639 |
Bob | +0.632 |
Neg logit contributions | |
otion | -0.870 |
Deal | -0.805 |
spine | -0.801 |
laptop | -0.789 |
upon | -0.781 |
Token | ". |
Token ID | 1911 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
otion | +0.091 |
laptop | +0.083 |
upon | +0.082 |
Deal | +0.080 |
spine | +0.080 |
Neg logit contributions | |
cream | -0.065 |
Fin | -0.058 |
Blue | -0.057 |
oc | -0.053 |
Jerry | -0.053 |
Token | Tim |
Token ID | 5045 |
Feature activation | +3.34e+00 |
Pos logit contributions | |
otion | +0.117 |
upon | +0.106 |
laptop | +0.106 |
Deal | +0.103 |
sister | +0.102 |
Neg logit contributions | |
cream | -0.071 |
Fin | -0.063 |
Blue | -0.062 |
Jerry | -0.057 |
Red | -0.056 |
Token | my |
Token ID | 1820 |
Feature activation | +5.17e+00 |
Pos logit contributions | |
cream | +0.175 |
Fin | +0.163 |
Blue | +0.161 |
Red | +0.147 |
Tweet | +0.146 |
Neg logit contributions | |
otion | -0.220 |
upon | -0.210 |
laptop | -0.202 |
sister | -0.201 |
Deal | -0.193 |
Token | wanted |
Token ID | 2227 |
Feature activation | +7.72e-01 |
Pos logit contributions | |
cream | +0.270 |
Fin | +0.264 |
Blue | +0.253 |
rob | +0.249 |
Red | +0.238 |
Neg logit contributions | |
upon | -0.320 |
sister | -0.313 |
otion | -0.308 |
laptop | -0.303 |
brother | -0.280 |
Token | little |
Token ID | 1310 |
Feature activation | -6.07e+00 |
Pos logit contributions | |
cream | +0.068 |
Blue | +0.062 |
Fin | +0.061 |
Tweet | +0.058 |
Jerry | +0.057 |
Neg logit contributions | |
otion | -0.087 |
upon | -0.076 |
laptop | -0.075 |
Deal | -0.074 |
spine | -0.073 |
Token | boy |
Token ID | 2933 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
otion | +0.361 |
laptop | +0.313 |
Deal | +0.307 |
upon | +0.304 |
spine | +0.297 |
Neg logit contributions | |
cream | -0.374 |
Fin | -0.347 |
Blue | -0.346 |
Tweet | -0.325 |
oc | -0.323 |
Token | smiled |
Token ID | 13541 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
cream | +0.076 |
Fin | +0.073 |
Blue | +0.070 |
Red | +0.066 |
Jerry | +0.065 |
Neg logit contributions | |
otion | -0.099 |
upon | -0.095 |
sister | -0.093 |
laptop | -0.091 |
headache | -0.087 |
Token | mis |
Token ID | 2984 |
Feature activation | -7.53e+00 |
Pos logit contributions | |
otion | +0.091 |
upon | +0.082 |
laptop | +0.081 |
sister | +0.078 |
Deal | +0.078 |
Neg logit contributions | |
cream | -0.079 |
Fin | -0.072 |
Blue | -0.071 |
Jerry | -0.066 |
Red | -0.066 |
Token | chie |
Token ID | 3043 |
Feature activation | +1.92e+00 |
Pos logit contributions | |
otion | +0.378 |
acy | +0.319 |
upon | +0.319 |
ids | +0.313 |
aughter | +0.312 |
Neg logit contributions | |
cream | -0.482 |
Blue | -0.454 |
Fin | -0.446 |
sav | -0.445 |
Jerry | -0.437 |
Token | v |
Token ID | 85 |
Feature activation | +1.37e+01 |
Pos logit contributions | |
Fin | +0.087 |
Blue | +0.086 |
cream | +0.085 |
Red | +0.084 |
Nut | +0.081 |
Neg logit contributions | |
otion | -0.128 |
upon | -0.122 |
sister | -0.116 |
aughter | -0.115 |
acy | -0.112 |
Token | ously |
Token ID | 3481 |
Feature activation | -4.69e+00 |
Pos logit contributions | |
cream | +0.837 |
Fin | +0.751 |
oc | +0.742 |
bo | +0.741 |
aw | +0.727 |
Neg logit contributions | |
otion | -0.756 |
laptop | -0.733 |
Deal | -0.717 |
spine | -0.703 |
headache | -0.688 |
Token | and |
Token ID | 290 |
Feature activation | -9.80e-01 |
Pos logit contributions | |
otion | +0.292 |
upon | +0.262 |
laptop | +0.261 |
sister | +0.250 |
spine | +0.250 |
Neg logit contributions | |
cream | -0.271 |
Fin | -0.249 |
Blue | -0.246 |
Jerry | -0.230 |
Red | -0.229 |
Token | said |
Token ID | 531 |
Feature activation | +2.82e+00 |
Pos logit contributions | |
otion | +0.066 |
upon | +0.063 |
sister | +0.062 |
laptop | +0.059 |
Lee | +0.059 |
Neg logit contributions | |
cream | -0.048 |
Fin | -0.045 |
Blue | -0.045 |
Red | -0.043 |
Jerry | -0.041 |
Token | , |
Token ID | 11 |
Feature activation | +5.23e-01 |
Pos logit contributions | |
cream | +0.158 |
Blue | +0.145 |
Fin | +0.143 |
oc | +0.132 |
Tweet | +0.132 |
Neg logit contributions | |
otion | -0.175 |
laptop | -0.165 |
spine | -0.157 |
Deal | -0.157 |
upon | -0.157 |
Token | " |
Token ID | 366 |
Feature activation | -4.38e+00 |
Pos logit contributions | |
cream | +0.030 |
Blue | +0.028 |
Fin | +0.027 |
He | +0.025 |
Jerry | +0.025 |
Neg logit contributions | |
otion | -0.033 |
laptop | -0.029 |
upon | -0.028 |
headache | -0.028 |
sister | -0.028 |
Token | He |
Token ID | 679 |
Feature activation | +4.33e+00 |
Pos logit contributions | |
cream | +0.001 |
Fin | +0.001 |
Blue | +0.001 |
rob | +0.001 |
Jerry | +0.001 |
Neg logit contributions | |
otion | -0.001 |
upon | -0.001 |
Deal | -0.001 |
sister | -0.001 |
laptop | -0.001 |
Token | made |
Token ID | 925 |
Feature activation | +7.53e+00 |
Pos logit contributions | |
cream | +0.219 |
Fin | +0.200 |
Blue | +0.195 |
Jerry | +0.184 |
Red | +0.179 |
Neg logit contributions | |
otion | -0.280 |
sister | -0.277 |
upon | -0.276 |
laptop | -0.274 |
ids | -0.253 |
Token | v |
Token ID | 410 |
Feature activation | +2.21e+00 |
Pos logit contributions | |
cream | +0.377 |
Fin | +0.350 |
Blue | +0.345 |
Jerry | +0.320 |
Red | +0.310 |
Neg logit contributions | |
otion | -0.526 |
upon | -0.485 |
laptop | -0.477 |
sister | -0.466 |
headache | -0.454 |
Token | room |
Token ID | 3823 |
Feature activation | -3.02e+00 |
Pos logit contributions | |
cream | +0.153 |
Fin | +0.146 |
Blue | +0.145 |
Jerry | +0.136 |
sav | +0.135 |
Neg logit contributions | |
otion | -0.110 |
laptop | -0.094 |
upon | -0.092 |
sister | -0.087 |
ids | -0.087 |
Token | - |
Token ID | 12 |
Feature activation | +2.90e+00 |
Pos logit contributions | |
otion | +0.187 |
laptop | +0.175 |
upon | +0.168 |
sister | +0.167 |
curtain | +0.162 |
Neg logit contributions | |
cream | -0.167 |
Fin | -0.162 |
Blue | -0.157 |
Jerry | -0.154 |
Frank | -0.142 |
Token | v |
Token ID | 85 |
Feature activation | +1.37e+01 |
Pos logit contributions | |
sav | +0.110 |
Fin | +0.103 |
Jerry | +0.102 |
lake | +0.100 |
Frank | +0.097 |
Neg logit contributions | |
otion | -0.193 |
upon | -0.188 |
laptop | -0.183 |
ids | -0.180 |
ges | -0.170 |
Token | room |
Token ID | 3823 |
Feature activation | -3.09e+00 |
Pos logit contributions | |
cream | +0.775 |
Fin | +0.684 |
oc | +0.668 |
Blue | +0.665 |
aw | +0.638 |
Neg logit contributions | |
otion | -0.864 |
Deal | -0.788 |
laptop | -0.786 |
spine | -0.781 |
headache | -0.775 |
Token | noises |
Token ID | 26782 |
Feature activation | +8.82e-01 |
Pos logit contributions | |
otion | +0.199 |
laptop | +0.188 |
upon | +0.179 |
sister | +0.177 |
curtain | +0.173 |
Neg logit contributions | |
cream | -0.166 |
Fin | -0.157 |
Blue | -0.153 |
Jerry | -0.148 |
Red | -0.137 |
Token | and |
Token ID | 290 |
Feature activation | -8.99e-01 |
Pos logit contributions | |
cream | +0.050 |
Fin | +0.046 |
Blue | +0.045 |
Jerry | +0.043 |
Red | +0.042 |
Neg logit contributions | |
otion | -0.057 |
laptop | -0.051 |
upon | -0.050 |
sister | -0.048 |
spine | -0.048 |
Token | drove |
Token ID | 10357 |
Feature activation | +2.23e+00 |
Pos logit contributions | |
otion | +0.057 |
laptop | +0.055 |
sister | +0.054 |
upon | +0.053 |
curtain | +0.051 |
Neg logit contributions | |
cream | -0.049 |
Fin | -0.044 |
Blue | -0.043 |
Jerry | -0.041 |
Red | -0.040 |
Token | it |
Token ID | 340 |
Feature activation | +3.62e+00 |
Pos logit contributions | |
cream | +0.111 |
Fin | +0.100 |
Blue | +0.098 |
Jerry | +0.091 |
Tweet | +0.089 |
Neg logit contributions | |
otion | -0.159 |
upon | -0.147 |
laptop | -0.145 |
sister | -0.140 |
spine | -0.139 |
Token | his |
Token ID | 465 |
Feature activation | +9.24e-01 |
Pos logit contributions | |
cream | +0.101 |
Fin | +0.090 |
Blue | +0.089 |
Jerry | +0.083 |
oc | +0.081 |
Neg logit contributions | |
otion | -0.153 |
upon | -0.140 |
laptop | -0.139 |
sister | -0.135 |
Deal | -0.133 |
Token | mouth |
Token ID | 5422 |
Feature activation | -4.34e+00 |
Pos logit contributions | |
cream | +0.051 |
Fin | +0.046 |
Blue | +0.045 |
oc | +0.043 |
Red | +0.042 |
Neg logit contributions | |
otion | -0.063 |
upon | -0.053 |
Deal | -0.052 |
acy | -0.051 |
laptop | -0.051 |
Token | and |
Token ID | 290 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
otion | +0.269 |
upon | +0.246 |
laptop | +0.246 |
sister | +0.233 |
spine | +0.233 |
Neg logit contributions | |
cream | -0.249 |
Fin | -0.232 |
Blue | -0.228 |
Jerry | -0.217 |
Red | -0.211 |
Token | says |
Token ID | 1139 |
Feature activation | +8.34e-01 |
Pos logit contributions | |
otion | +0.076 |
laptop | +0.070 |
upon | +0.068 |
sister | +0.068 |
Deal | +0.065 |
Neg logit contributions | |
cream | -0.072 |
Fin | -0.067 |
Blue | -0.066 |
Jerry | -0.062 |
Red | -0.061 |
Token | " |
Token ID | 366 |
Feature activation | -4.41e+00 |
Pos logit contributions | |
cream | +0.049 |
Fin | +0.045 |
Blue | +0.045 |
Tweet | +0.042 |
Red | +0.042 |
Neg logit contributions | |
otion | -0.050 |
laptop | -0.045 |
upon | -0.045 |
spine | -0.044 |
Deal | -0.044 |
Token | v |
Token ID | 85 |
Feature activation | +1.37e+01 |
Pos logit contributions | |
otion | +0.295 |
ancer | +0.294 |
upon | +0.292 |
laptop | +0.283 |
acy | +0.280 |
Neg logit contributions | |
cream | -0.176 |
Jerry | -0.158 |
Fin | -0.151 |
Nut | -0.146 |
Frank | -0.143 |
Token | room |
Token ID | 3823 |
Feature activation | -3.25e+00 |
Pos logit contributions | |
cream | +0.733 |
oc | +0.677 |
Fin | +0.669 |
Blue | +0.632 |
aw | +0.619 |
Neg logit contributions | |
otion | -0.875 |
spine | -0.818 |
upon | -0.791 |
Deal | -0.775 |
laptop | -0.766 |
Token | , |
Token ID | 11 |
Feature activation | +3.68e-01 |
Pos logit contributions | |
otion | +0.211 |
laptop | +0.192 |
upon | +0.187 |
sister | +0.184 |
headache | +0.180 |
Neg logit contributions | |
cream | -0.178 |
Fin | -0.163 |
Blue | -0.163 |
Jerry | -0.153 |
Red | -0.149 |
Token | v |
Token ID | 410 |
Feature activation | +2.18e+00 |
Pos logit contributions | |
cream | +0.020 |
Fin | +0.018 |
Blue | +0.018 |
Red | +0.017 |
Tweet | +0.017 |
Neg logit contributions | |
otion | -0.024 |
upon | -0.022 |
laptop | -0.021 |
Deal | -0.021 |
spine | -0.021 |
Token | room |
Token ID | 3823 |
Feature activation | -3.30e+00 |
Pos logit contributions | |
cream | +0.155 |
Fin | +0.146 |
Blue | +0.146 |
Jerry | +0.139 |
Tweet | +0.135 |
Neg logit contributions | |
otion | -0.106 |
laptop | -0.093 |
upon | -0.087 |
sister | -0.086 |
Deal | -0.085 |
Token | ". |
Token ID | 1911 |
Feature activation | -1.88e+00 |
Pos logit contributions | |
otion | +0.205 |
laptop | +0.185 |
upon | +0.181 |
sister | +0.178 |
headache | +0.173 |
Neg logit contributions | |
cream | -0.190 |
Fin | -0.175 |
Blue | -0.174 |
Jerry | -0.164 |
Red | -0.160 |
Token | , |
Token ID | 11 |
Feature activation | +3.78e-01 |
Pos logit contributions | |
otion | +0.130 |
Deal | +0.123 |
headache | +0.116 |
upon | +0.113 |
laptop | +0.112 |
Neg logit contributions | |
cream | -0.079 |
Blue | -0.067 |
sold | -0.067 |
Fin | -0.067 |
sav | -0.066 |
Token | making |
Token ID | 1642 |
Feature activation | -1.91e+00 |
Pos logit contributions | |
cream | +0.021 |
Blue | +0.019 |
Fin | +0.019 |
sav | +0.018 |
sold | +0.018 |
Neg logit contributions | |
otion | -0.024 |
Deal | -0.022 |
acy | -0.021 |
laptop | -0.020 |
upon | -0.020 |
Token | v |
Token ID | 410 |
Feature activation | +2.11e+00 |
Pos logit contributions | |
otion | +0.124 |
upon | +0.112 |
laptop | +0.112 |
sister | +0.108 |
headache | +0.105 |
Neg logit contributions | |
cream | -0.105 |
Fin | -0.098 |
Blue | -0.096 |
Jerry | -0.091 |
Tweet | -0.088 |
Token | room |
Token ID | 3823 |
Feature activation | -3.26e+00 |
Pos logit contributions | |
cream | +0.143 |
Fin | +0.134 |
Blue | +0.132 |
Jerry | +0.125 |
oc | +0.123 |
Neg logit contributions | |
otion | -0.111 |
laptop | -0.097 |
upon | -0.095 |
sister | -0.090 |
Deal | -0.090 |
Token | - |
Token ID | 12 |
Feature activation | +2.85e+00 |
Pos logit contributions | |
otion | +0.212 |
laptop | +0.201 |
upon | +0.193 |
sister | +0.188 |
headache | +0.184 |
Neg logit contributions | |
cream | -0.172 |
Fin | -0.166 |
Blue | -0.161 |
Jerry | -0.158 |
Tweet | -0.144 |
Token | v |
Token ID | 85 |
Feature activation | +1.37e+01 |
Pos logit contributions | |
sav | +0.086 |
Fin | +0.083 |
Jerry | +0.081 |
cream | +0.076 |
Blue | +0.071 |
Neg logit contributions | |
otion | -0.226 |
laptop | -0.219 |
upon | -0.217 |
ids | -0.210 |
911 | -0.206 |
Token | room |
Token ID | 3823 |
Feature activation | -3.29e+00 |
Pos logit contributions | |
cream | +0.702 |
oc | +0.609 |
v | +0.607 |
Fin | +0.596 |
aw | +0.588 |
Neg logit contributions | |
otion | -0.909 |
Deal | -0.852 |
spine | -0.838 |
headache | -0.821 |
upon | -0.811 |
Token | noises |
Token ID | 26782 |
Feature activation | +7.19e-01 |
Pos logit contributions | |
otion | +0.218 |
laptop | +0.206 |
upon | +0.198 |
sister | +0.193 |
headache | +0.190 |
Neg logit contributions | |
cream | -0.172 |
Fin | -0.163 |
Blue | -0.158 |
Jerry | -0.153 |
Tweet | -0.142 |
Token | . |
Token ID | 13 |
Feature activation | -5.76e-02 |
Pos logit contributions | |
cream | +0.041 |
Fin | +0.037 |
Blue | +0.037 |
Jerry | +0.034 |
Red | +0.034 |
Neg logit contributions | |
otion | -0.046 |
laptop | -0.041 |
upon | -0.041 |
sister | -0.039 |
spine | -0.039 |
Token |
|
Token ID | 198 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
otion | +0.004 |
upon | +0.004 |
laptop | +0.004 |
Deal | +0.004 |
sister | +0.004 |
Neg logit contributions | |
cream | -0.002 |
Fin | -0.002 |
Blue | -0.002 |
Jerry | -0.002 |
Red | -0.002 |
Token |
|
Token ID | 198 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
cream | +0.066 |
Fin | +0.054 |
Blue | +0.053 |
sav | +0.049 |
Jerry | +0.049 |
Neg logit contributions | |
otion | -0.098 |
upon | -0.095 |
laptop | -0.087 |
Deal | -0.084 |
sister | -0.083 |
Token | They |
Token ID | 1119 |
Feature activation | -6.80e-01 |
Pos logit contributions | |
cream | +0.078 |
Blue | +0.070 |
Fin | +0.067 |
oc | +0.066 |
He | +0.063 |
Neg logit contributions | |
otion | -0.166 |
Deal | -0.153 |
acy | -0.149 |
laptop | -0.148 |
ancer | -0.147 |
Token | made |
Token ID | 925 |
Feature activation | +6.18e+00 |
Pos logit contributions | |
otion | +0.047 |
upon | +0.043 |
laptop | +0.043 |
sister | +0.042 |
Deal | +0.041 |
Neg logit contributions | |
cream | -0.034 |
Fin | -0.031 |
Blue | -0.030 |
Jerry | -0.028 |
Red | -0.028 |
Token | v |
Token ID | 410 |
Feature activation | +1.70e+00 |
Pos logit contributions | |
cream | +0.300 |
Fin | +0.262 |
Blue | +0.256 |
Red | +0.236 |
Jerry | +0.235 |
Neg logit contributions | |
otion | -0.448 |
spine | -0.398 |
upon | -0.398 |
laptop | -0.396 |
Deal | -0.393 |
Token | room |
Token ID | 3823 |
Feature activation | -3.56e+00 |
Pos logit contributions | |
cream | +0.108 |
Fin | +0.102 |
Blue | +0.101 |
Jerry | +0.096 |
Red | +0.094 |
Neg logit contributions | |
otion | -0.095 |
laptop | -0.083 |
upon | -0.083 |
sister | -0.078 |
Deal | -0.078 |
Token | - |
Token ID | 12 |
Feature activation | +2.73e+00 |
Pos logit contributions | |
otion | +0.238 |
laptop | +0.222 |
upon | +0.216 |
sister | +0.210 |
headache | +0.208 |
Neg logit contributions | |
cream | -0.182 |
Fin | -0.173 |
Blue | -0.168 |
Jerry | -0.162 |
Tweet | -0.151 |
Token | v |
Token ID | 85 |
Feature activation | +1.36e+01 |
Pos logit contributions | |
Jerry | +0.113 |
Fin | +0.111 |
Frank | +0.106 |
Blue | +0.103 |
sav | +0.101 |
Neg logit contributions | |
otion | -0.182 |
laptop | -0.176 |
upon | -0.175 |
ids | -0.168 |
ancer | -0.161 |
Token | room |
Token ID | 3823 |
Feature activation | -3.52e+00 |
Pos logit contributions | |
cream | +0.787 |
Fin | +0.690 |
Blue | +0.671 |
oc | +0.669 |
aw | +0.646 |
Neg logit contributions | |
otion | -0.855 |
Deal | -0.765 |
laptop | -0.756 |
spine | -0.750 |
headache | -0.748 |
Token | noises |
Token ID | 26782 |
Feature activation | +7.74e-01 |
Pos logit contributions | |
otion | +0.238 |
laptop | +0.223 |
upon | +0.217 |
sister | +0.211 |
headache | +0.208 |
Neg logit contributions | |
cream | -0.178 |
Fin | -0.168 |
Blue | -0.164 |
Jerry | -0.156 |
Red | -0.147 |
Token | as |
Token ID | 355 |
Feature activation | -3.09e+00 |
Pos logit contributions | |
cream | +0.039 |
Fin | +0.035 |
Blue | +0.035 |
Red | +0.032 |
Jerry | +0.032 |
Neg logit contributions | |
otion | -0.053 |
upon | -0.048 |
laptop | -0.048 |
Deal | -0.048 |
sister | -0.046 |
Token | they |
Token ID | 484 |
Feature activation | -3.29e-01 |
Pos logit contributions | |
otion | +0.237 |
Deal | +0.220 |
laptop | +0.212 |
spine | +0.211 |
upon | +0.208 |
Neg logit contributions | |
cream | -0.137 |
Fin | -0.120 |
Blue | -0.117 |
Red | -0.108 |
sailed | -0.108 |
Token | moved |
Token ID | 3888 |
Feature activation | +7.70e-01 |
Pos logit contributions | |
otion | +0.024 |
upon | +0.022 |
laptop | +0.022 |
sister | +0.022 |
headache | +0.021 |
Neg logit contributions | |
cream | -0.015 |
Fin | -0.013 |
Blue | -0.013 |
Jerry | -0.012 |
Red | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
otion | +0.125 |
Deal | +0.113 |
laptop | +0.110 |
upon | +0.110 |
headache | +0.109 |
Neg logit contributions | |
cream | -0.097 |
Fin | -0.087 |
Blue | -0.087 |
oc | -0.082 |
sav | -0.081 |
Token | make |
Token ID | 787 |
Feature activation | +3.10e+00 |
Pos logit contributions | |
otion | +0.062 |
laptop | +0.058 |
upon | +0.057 |
sister | +0.056 |
spine | +0.055 |
Neg logit contributions | |
cream | -0.070 |
Fin | -0.064 |
Blue | -0.062 |
Jerry | -0.060 |
Red | -0.058 |
Token | v |
Token ID | 410 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
cream | +0.186 |
Fin | +0.176 |
Blue | +0.173 |
Jerry | +0.167 |
Tweet | +0.158 |
Neg logit contributions | |
otion | -0.182 |
upon | -0.169 |
laptop | -0.164 |
sister | -0.162 |
911 | -0.155 |
Token | room |
Token ID | 3823 |
Feature activation | -3.27e+00 |
Pos logit contributions | |
cream | +0.142 |
Fin | +0.134 |
Blue | +0.132 |
Jerry | +0.126 |
sav | +0.124 |
Neg logit contributions | |
otion | -0.103 |
laptop | -0.088 |
upon | -0.086 |
sister | -0.082 |
Deal | -0.081 |
Token | - |
Token ID | 12 |
Feature activation | +2.78e+00 |
Pos logit contributions | |
otion | +0.208 |
laptop | +0.196 |
upon | +0.188 |
sister | +0.187 |
headache | +0.181 |
Neg logit contributions | |
cream | -0.174 |
Fin | -0.169 |
Blue | -0.165 |
Jerry | -0.165 |
Frank | -0.150 |
Token | v |
Token ID | 85 |
Feature activation | +1.36e+01 |
Pos logit contributions | |
sav | +0.089 |
Jerry | +0.082 |
Fin | +0.080 |
cream | +0.076 |
Blue | +0.072 |
Neg logit contributions | |
otion | -0.216 |
upon | -0.208 |
laptop | -0.206 |
ids | -0.198 |
ancer | -0.198 |
Token | room |
Token ID | 3823 |
Feature activation | -3.29e+00 |
Pos logit contributions | |
cream | +0.723 |
oc | +0.624 |
Fin | +0.623 |
v | +0.612 |
Blue | +0.602 |
Neg logit contributions | |
otion | -0.884 |
spine | -0.828 |
Deal | -0.826 |
upon | -0.806 |
laptop | -0.806 |
Token | noises |
Token ID | 26782 |
Feature activation | +6.41e-01 |
Pos logit contributions | |
otion | +0.216 |
laptop | +0.204 |
upon | +0.194 |
sister | +0.192 |
curtain | +0.188 |
Neg logit contributions | |
cream | -0.173 |
Fin | -0.164 |
Blue | -0.160 |
Jerry | -0.157 |
Nut | -0.143 |
Token | . |
Token ID | 13 |
Feature activation | -1.45e-03 |
Pos logit contributions | |
cream | +0.038 |
Fin | +0.035 |
Blue | +0.034 |
Jerry | +0.032 |
Tweet | +0.031 |
Neg logit contributions | |
otion | -0.040 |
laptop | -0.036 |
upon | -0.035 |
sister | -0.034 |
spine | -0.034 |
Token | One |
Token ID | 1881 |
Feature activation | +4.97e+00 |
Pos logit contributions | |
otion | +0.000 |
upon | +0.000 |
laptop | +0.000 |
sister | +0.000 |
Deal | +0.000 |
Neg logit contributions | |
cream | -0.000 |
Fin | -0.000 |
Blue | -0.000 |
aw | -0.000 |
Jerry | -0.000 |
Token | day |
Token ID | 1110 |
Feature activation | -9.98e-01 |
Pos logit contributions | |
cream | +0.249 |
Fin | +0.225 |
Blue | +0.221 |
oc | +0.205 |
Jerry | +0.203 |
Neg logit contributions | |
otion | -0.356 |
laptop | -0.318 |
upon | -0.315 |
Deal | -0.315 |
spine | -0.306 |
Token | behind |
Token ID | 2157 |
Feature activation | -8.35e+00 |
Pos logit contributions | |
otion | +0.088 |
laptop | +0.078 |
upon | +0.077 |
Deal | +0.076 |
headache | +0.076 |
Neg logit contributions | |
cream | -0.073 |
Fin | -0.066 |
Blue | -0.066 |
Red | -0.061 |
oc | -0.061 |
Token | something |
Token ID | 1223 |
Feature activation | -2.99e+00 |
Pos logit contributions | |
otion | +0.650 |
Deal | +0.591 |
laptop | +0.585 |
spine | +0.580 |
headache | +0.568 |
Neg logit contributions | |
cream | -0.352 |
Blue | -0.315 |
Fin | -0.300 |
aw | -0.288 |
sav | -0.279 |
Token | and |
Token ID | 290 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
otion | +0.220 |
laptop | +0.199 |
Deal | +0.197 |
upon | +0.196 |
spine | +0.192 |
Neg logit contributions | |
cream | -0.138 |
Fin | -0.124 |
Blue | -0.124 |
Red | -0.111 |
oc | -0.111 |
Token | say |
Token ID | 910 |
Feature activation | -8.09e-01 |
Pos logit contributions | |
otion | +0.066 |
upon | +0.060 |
laptop | +0.060 |
sister | +0.057 |
Deal | +0.057 |
Neg logit contributions | |
cream | -0.061 |
Fin | -0.056 |
Blue | -0.055 |
Jerry | -0.052 |
Red | -0.051 |
Token | ' |
Token ID | 705 |
Feature activation | -3.78e+00 |
Pos logit contributions | |
otion | +0.048 |
laptop | +0.045 |
upon | +0.044 |
spine | +0.043 |
aughter | +0.041 |
Neg logit contributions | |
cream | -0.046 |
Blue | -0.044 |
Fin | -0.042 |
aw | -0.040 |
Tweet | -0.040 |
Token | v |
Token ID | 85 |
Feature activation | +1.36e+01 |
Pos logit contributions | |
otion | +0.270 |
upon | +0.261 |
ancer | +0.249 |
laptop | +0.241 |
Deal | +0.238 |
Neg logit contributions | |
cream | -0.141 |
Fin | -0.130 |
Blue | -0.125 |
sav | -0.118 |
Jerry | -0.115 |
Token | anish |
Token ID | 7115 |
Feature activation | -7.02e+00 |
Pos logit contributions | |
cream | +0.795 |
oc | +0.750 |
Fin | +0.736 |
bo | +0.711 |
aw | +0.708 |
Neg logit contributions | |
otion | -0.750 |
spine | -0.744 |
upon | -0.685 |
Deal | -0.684 |
laptop | -0.679 |
Token | !' |
Token ID | 13679 |
Feature activation | -2.61e+00 |
Pos logit contributions | |
otion | +0.476 |
spine | +0.436 |
upon | +0.435 |
laptop | +0.429 |
headache | +0.425 |
Neg logit contributions | |
cream | -0.355 |
Blue | -0.327 |
Fin | -0.321 |
oc | -0.301 |
Red | -0.300 |
Token | and |
Token ID | 290 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
otion | +0.189 |
laptop | +0.171 |
upon | +0.167 |
headache | +0.166 |
spine | +0.164 |
Neg logit contributions | |
cream | -0.126 |
Blue | -0.115 |
Fin | -0.113 |
oc | -0.105 |
Jerry | -0.104 |
Token | then |
Token ID | 788 |
Feature activation | +4.13e+00 |
Pos logit contributions | |
otion | +0.078 |
laptop | +0.070 |
upon | +0.070 |
headache | +0.069 |
spine | +0.069 |
Neg logit contributions | |
cream | -0.043 |
Blue | -0.037 |
Fin | -0.037 |
sav | -0.035 |
oc | -0.033 |
Token | come |
Token ID | 1282 |
Feature activation | +6.89e-02 |
Pos logit contributions | |
cream | +0.218 |
Blue | +0.195 |
Fin | +0.190 |
sav | +0.185 |
sailed | +0.182 |
Neg logit contributions | |
otion | -0.273 |
spine | -0.243 |
Deal | -0.243 |
laptop | -0.242 |
upon | -0.236 |
Token | a |
Token ID | 257 |
Feature activation | -1.84e+00 |
Pos logit contributions | |
cream | +0.162 |
Blue | +0.149 |
Fin | +0.149 |
Jerry | +0.138 |
oc | +0.136 |
Neg logit contributions | |
otion | -0.190 |
upon | -0.173 |
laptop | -0.171 |
sister | -0.168 |
headache | -0.160 |
Token | loud |
Token ID | 7812 |
Feature activation | -5.28e+00 |
Pos logit contributions | |
otion | +0.114 |
laptop | +0.108 |
upon | +0.107 |
sister | +0.105 |
headache | +0.103 |
Neg logit contributions | |
cream | -0.106 |
Fin | -0.093 |
Blue | -0.092 |
Jerry | -0.087 |
Tweet | -0.085 |
Token | v |
Token ID | 410 |
Feature activation | +2.03e+00 |
Pos logit contributions | |
otion | +0.316 |
laptop | +0.289 |
upon | +0.279 |
sister | +0.277 |
headache | +0.274 |
Neg logit contributions | |
cream | -0.323 |
Fin | -0.294 |
Blue | -0.290 |
Jerry | -0.275 |
Red | -0.267 |
Token | room |
Token ID | 3823 |
Feature activation | -3.26e+00 |
Pos logit contributions | |
cream | +0.132 |
Fin | +0.123 |
Blue | +0.121 |
Jerry | +0.114 |
Red | +0.113 |
Neg logit contributions | |
otion | -0.113 |
laptop | -0.100 |
upon | -0.098 |
sister | -0.094 |
spine | -0.094 |
Token | - |
Token ID | 12 |
Feature activation | +2.70e+00 |
Pos logit contributions | |
otion | +0.235 |
laptop | +0.214 |
upon | +0.212 |
sister | +0.205 |
headache | +0.204 |
Neg logit contributions | |
cream | -0.158 |
Fin | -0.143 |
Blue | -0.141 |
Jerry | -0.131 |
Red | -0.127 |
Token | v |
Token ID | 85 |
Feature activation | +1.36e+01 |
Pos logit contributions | |
sav | +0.090 |
cream | +0.088 |
Fin | +0.087 |
Jerry | +0.085 |
Blue | +0.084 |
Neg logit contributions | |
otion | -0.202 |
upon | -0.186 |
laptop | -0.186 |
ancer | -0.181 |
911 | -0.179 |
Token | room |
Token ID | 3823 |
Feature activation | -3.27e+00 |
Pos logit contributions | |
cream | +0.634 |
oc | +0.583 |
v | +0.561 |
aw | +0.530 |
Fin | +0.530 |
Neg logit contributions | |
otion | -0.930 |
Deal | -0.906 |
spine | -0.869 |
upon | -0.859 |
headache | -0.837 |
Token | sound |
Token ID | 2128 |
Feature activation | +3.15e+00 |
Pos logit contributions | |
otion | +0.232 |
laptop | +0.213 |
upon | +0.209 |
sister | +0.203 |
headache | +0.201 |
Neg logit contributions | |
cream | -0.162 |
Fin | -0.148 |
Blue | -0.145 |
Jerry | -0.136 |
Tweet | -0.130 |
Token | . |
Token ID | 13 |
Feature activation | -9.55e-02 |
Pos logit contributions | |
cream | +0.163 |
Fin | +0.142 |
Blue | +0.141 |
oc | +0.134 |
Red | +0.131 |
Neg logit contributions | |
otion | -0.216 |
Deal | -0.197 |
upon | -0.195 |
laptop | -0.194 |
sister | -0.187 |
Token | He |
Token ID | 679 |
Feature activation | +4.10e+00 |
Pos logit contributions | |
otion | +0.007 |
laptop | +0.007 |
upon | +0.007 |
sister | +0.007 |
Deal | +0.007 |
Neg logit contributions | |
cream | -0.004 |
Fin | -0.004 |
Blue | -0.003 |
Jerry | -0.003 |
Tweet | -0.003 |
Token | began |
Token ID | 2540 |
Feature activation | -3.30e+00 |
Pos logit contributions | |
cream | +0.218 |
Blue | +0.195 |
Fin | +0.190 |
Jerry | +0.182 |
Nut | +0.178 |
Neg logit contributions | |
sister | -0.267 |
upon | -0.264 |
otion | -0.261 |
laptop | -0.261 |
brother | -0.242 |
Token | . |
Token ID | 13 |
Feature activation | -2.03e-01 |
Pos logit contributions | |
cream | +0.364 |
Fin | +0.338 |
Blue | +0.335 |
Jerry | +0.316 |
Red | +0.312 |
Neg logit contributions | |
otion | -0.395 |
laptop | -0.359 |
upon | -0.358 |
sister | -0.348 |
spine | -0.348 |
Token |
|
Token ID | 198 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
otion | +0.013 |
upon | +0.012 |
laptop | +0.012 |
Deal | +0.012 |
spine | +0.012 |
Neg logit contributions | |
cream | -0.011 |
Fin | -0.010 |
Blue | -0.010 |
Tweet | -0.009 |
Jerry | -0.009 |
Token |
|
Token ID | 198 |
Feature activation | +1.27e+00 |
Pos logit contributions | |
cream | +0.060 |
Fin | +0.051 |
Blue | +0.050 |
Tweet | +0.045 |
Red | +0.045 |
Neg logit contributions | |
otion | -0.094 |
upon | -0.088 |
laptop | -0.084 |
Deal | -0.081 |
Lee | -0.079 |
Token | I |
Token ID | 40 |
Feature activation | +3.94e+00 |
Pos logit contributions | |
cream | +0.060 |
rob | +0.055 |
Fin | +0.049 |
rust | +0.047 |
sailed | +0.046 |
Neg logit contributions | |
otion | -0.083 |
upon | -0.083 |
laptop | -0.076 |
Deal | -0.075 |
911 | -0.075 |
Token | ne |
Token ID | 710 |
Feature activation | +3.62e+00 |
Pos logit contributions | |
cream | +0.246 |
Fin | +0.226 |
Blue | +0.224 |
Jerry | +0.214 |
Red | +0.209 |
Neg logit contributions | |
otion | -0.224 |
upon | -0.209 |
headache | -0.198 |
laptop | -0.197 |
Deal | -0.194 |
Token | v |
Token ID | 85 |
Feature activation | +1.36e+01 |
Pos logit contributions | |
cream | +0.197 |
oc | +0.176 |
Fin | +0.174 |
Blue | +0.172 |
bo | +0.171 |
Neg logit contributions | |
otion | -0.225 |
spine | -0.212 |
laptop | -0.212 |
Deal | -0.205 |
Lee | -0.197 |
Token | itably |
Token ID | 14829 |
Feature activation | -3.07e+00 |
Pos logit contributions | |
cream | +0.753 |
Fin | +0.687 |
oc | +0.666 |
Blue | +0.657 |
aw | +0.636 |
Neg logit contributions | |
otion | -0.862 |
laptop | -0.785 |
spine | -0.776 |
upon | -0.768 |
Deal | -0.762 |
Token | , |
Token ID | 11 |
Feature activation | +2.72e-01 |
Pos logit contributions | |
otion | +0.195 |
upon | +0.170 |
laptop | +0.170 |
Deal | +0.169 |
spine | +0.163 |
Neg logit contributions | |
cream | -0.176 |
Blue | -0.161 |
Fin | -0.160 |
oc | -0.152 |
Jerry | -0.149 |
Token | John |
Token ID | 1757 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
cream | +0.014 |
Fin | +0.012 |
Blue | +0.012 |
He | +0.011 |
Red | +0.011 |
Neg logit contributions | |
otion | -0.019 |
laptop | -0.017 |
spine | -0.017 |
Deal | -0.016 |
headache | -0.016 |
Token | was |
Token ID | 373 |
Feature activation | -4.41e-01 |
Pos logit contributions | |
cream | +0.111 |
Fin | +0.101 |
Blue | +0.099 |
Jerry | +0.092 |
Red | +0.091 |
Neg logit contributions | |
otion | -0.147 |
upon | -0.133 |
laptop | -0.132 |
sister | -0.128 |
Deal | -0.127 |
Token | not |
Token ID | 407 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
otion | +0.028 |
upon | +0.026 |
laptop | +0.025 |
sister | +0.025 |
headache | +0.025 |
Neg logit contributions | |
cream | -0.025 |
Fin | -0.023 |
Blue | -0.022 |
Jerry | -0.021 |
oc | -0.021 |
Token | the |
Token ID | 262 |
Feature activation | -1.71e-01 |
Pos logit contributions | |
otion | +0.081 |
laptop | +0.072 |
upon | +0.071 |
Deal | +0.071 |
spine | +0.071 |
Neg logit contributions | |
cream | -0.054 |
Fin | -0.049 |
Blue | -0.048 |
oc | -0.045 |
Jerry | -0.044 |
Token | little |
Token ID | 1310 |
Feature activation | -6.25e+00 |
Pos logit contributions | |
otion | +0.012 |
upon | +0.011 |
laptop | +0.011 |
spine | +0.011 |
sister | +0.011 |
Neg logit contributions | |
cream | -0.009 |
Blue | -0.007 |
Fin | -0.007 |
Jerry | -0.007 |
oc | -0.007 |
Token | cracks |
Token ID | 23217 |
Feature activation | +2.92e+00 |
Pos logit contributions | |
otion | +0.383 |
upon | +0.372 |
sister | +0.364 |
spine | +0.359 |
laptop | +0.355 |
Neg logit contributions | |
cream | -0.358 |
Blue | -0.312 |
Fin | -0.310 |
Jerry | -0.299 |
Red | -0.292 |
Token | and |
Token ID | 290 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
cream | +0.145 |
Fin | +0.129 |
Blue | +0.127 |
oc | +0.118 |
aw | +0.116 |
Neg logit contributions | |
otion | -0.203 |
Deal | -0.187 |
laptop | -0.183 |
upon | -0.179 |
headache | -0.178 |
Token | cre |
Token ID | 1126 |
Feature activation | +6.29e+00 |
Pos logit contributions | |
otion | +0.072 |
upon | +0.065 |
laptop | +0.064 |
sister | +0.062 |
Deal | +0.062 |
Neg logit contributions | |
cream | -0.062 |
Fin | -0.057 |
Blue | -0.056 |
Jerry | -0.053 |
Red | -0.052 |
Token | v |
Token ID | 85 |
Feature activation | +1.36e+01 |
Pos logit contributions | |
Blue | +0.372 |
cream | +0.361 |
Fin | +0.353 |
Red | +0.342 |
Jerry | +0.332 |
Neg logit contributions | |
otion | -0.386 |
upon | -0.348 |
acy | -0.327 |
cloth | -0.324 |
ids | -0.316 |
Token | ices |
Token ID | 1063 |
Feature activation | -3.47e+00 |
Pos logit contributions | |
cream | +0.870 |
Fin | +0.807 |
oc | +0.767 |
aw | +0.763 |
Blue | +0.758 |
Neg logit contributions | |
otion | -0.736 |
laptop | -0.674 |
Deal | -0.660 |
spine | -0.650 |
upon | -0.639 |
Token | of |
Token ID | 286 |
Feature activation | -3.05e+00 |
Pos logit contributions | |
otion | +0.234 |
laptop | +0.210 |
upon | +0.208 |
Deal | +0.207 |
headache | +0.202 |
Neg logit contributions | |
cream | -0.182 |
Fin | -0.164 |
Blue | -0.162 |
Jerry | -0.149 |
oc | -0.148 |
Token | the |
Token ID | 262 |
Feature activation | -1.62e-01 |
Pos logit contributions | |
otion | +0.219 |
laptop | +0.197 |
upon | +0.196 |
Deal | +0.192 |
spine | +0.190 |
Neg logit contributions | |
cream | -0.149 |
Fin | -0.135 |
Blue | -0.133 |
Jerry | -0.122 |
oc | -0.121 |
Token | rocks |
Token ID | 12586 |
Feature activation | +4.07e+00 |
Pos logit contributions | |
otion | +0.012 |
laptop | +0.011 |
upon | +0.011 |
Deal | +0.011 |
sister | +0.010 |
Neg logit contributions | |
cream | -0.008 |
Fin | -0.007 |
Blue | -0.007 |
Jerry | -0.006 |
Tweet | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
cream | +0.216 |
Fin | +0.195 |
Blue | +0.193 |
Jerry | +0.178 |
Red | +0.177 |
Neg logit contributions | |
otion | -0.271 |
laptop | -0.244 |
upon | -0.242 |
Deal | -0.239 |
sister | -0.234 |
Token | He |
Token ID | 679 |
Feature activation | +3.95e+00 |
Pos logit contributions | |
otion | +0.005 |
upon | +0.004 |
laptop | +0.004 |
Deal | +0.004 |
sister | +0.004 |
Neg logit contributions | |
cream | -0.003 |
Fin | -0.002 |
Blue | -0.002 |
Jerry | -0.002 |
oc | -0.002 |
Token | makes |
Token ID | 1838 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
cream | +0.250 |
Fin | +0.234 |
Blue | +0.231 |
Jerry | +0.220 |
Tweet | +0.215 |
Neg logit contributions | |
otion | -0.223 |
laptop | -0.204 |
upon | -0.202 |
sister | -0.195 |
headache | -0.190 |
Token | v |
Token ID | 410 |
Feature activation | +2.14e+00 |
Pos logit contributions | |
cream | +0.087 |
Fin | +0.077 |
Blue | +0.076 |
Jerry | +0.070 |
Red | +0.070 |
Neg logit contributions | |
otion | -0.125 |
laptop | -0.112 |
upon | -0.112 |
Deal | -0.110 |
spine | -0.109 |
Token | room |
Token ID | 3823 |
Feature activation | -3.35e+00 |
Pos logit contributions | |
cream | +0.143 |
Fin | +0.134 |
Blue | +0.133 |
Jerry | +0.126 |
Red | +0.125 |
Neg logit contributions | |
otion | -0.114 |
laptop | -0.098 |
upon | -0.097 |
Deal | -0.093 |
sister | -0.092 |
Token | - |
Token ID | 12 |
Feature activation | +2.74e+00 |
Pos logit contributions | |
otion | +0.217 |
laptop | +0.200 |
upon | +0.194 |
sister | +0.192 |
headache | +0.188 |
Neg logit contributions | |
cream | -0.178 |
Fin | -0.170 |
Blue | -0.166 |
Jerry | -0.162 |
Red | -0.149 |
Token | v |
Token ID | 85 |
Feature activation | +1.36e+01 |
Pos logit contributions | |
sav | +0.102 |
Fin | +0.101 |
Jerry | +0.100 |
cream | +0.095 |
Blue | +0.093 |
Neg logit contributions | |
otion | -0.194 |
upon | -0.185 |
laptop | -0.176 |
ids | -0.174 |
911 | -0.174 |
Token | room |
Token ID | 3823 |
Feature activation | -3.39e+00 |
Pos logit contributions | |
cream | +0.681 |
oc | +0.595 |
v | +0.585 |
Fin | +0.581 |
aw | +0.574 |
Neg logit contributions | |
otion | -0.899 |
spine | -0.849 |
Deal | -0.844 |
headache | -0.832 |
laptop | -0.828 |
Token | sounds |
Token ID | 5238 |
Feature activation | +5.36e+00 |
Pos logit contributions | |
otion | +0.225 |
laptop | +0.208 |
upon | +0.202 |
sister | +0.198 |
headache | +0.195 |
Neg logit contributions | |
cream | -0.179 |
Fin | -0.168 |
Blue | -0.163 |
Jerry | -0.158 |
Red | -0.147 |
Token | . |
Token ID | 13 |
Feature activation | -8.14e-02 |
Pos logit contributions | |
cream | +0.282 |
Fin | +0.256 |
Blue | +0.254 |
Red | +0.234 |
Jerry | +0.234 |
Neg logit contributions | |
otion | -0.363 |
laptop | -0.326 |
upon | -0.325 |
Deal | -0.318 |
headache | -0.313 |
Token | Kim |
Token ID | 6502 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
otion | +0.006 |
upon | +0.006 |
laptop | +0.006 |
Deal | +0.006 |
sister | +0.006 |
Neg logit contributions | |
cream | -0.004 |
Fin | -0.003 |
Blue | -0.003 |
oc | -0.003 |
Jerry | -0.003 |
Token | finds |
Token ID | 7228 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
otion | +0.097 |
laptop | +0.088 |
upon | +0.086 |
sister | +0.084 |
headache | +0.081 |
Neg logit contributions | |
cream | -0.099 |
Fin | -0.094 |
Blue | -0.092 |
Jerry | -0.087 |
Tweet | -0.085 |
Token | They |
Token ID | 1119 |
Feature activation | +2.61e+00 |
Pos logit contributions | |
otion | +0.011 |
upon | +0.010 |
laptop | +0.010 |
Deal | +0.010 |
headache | +0.010 |
Neg logit contributions | |
cream | -0.005 |
Fin | -0.005 |
Blue | -0.005 |
Jerry | -0.004 |
Red | -0.004 |
Token | called |
Token ID | 1444 |
Feature activation | -6.65e-01 |
Pos logit contributions | |
cream | +0.103 |
Fin | +0.090 |
Blue | +0.089 |
Jerry | +0.079 |
Red | +0.079 |
Neg logit contributions | |
otion | -0.213 |
laptop | -0.195 |
upon | -0.194 |
Deal | -0.191 |
spine | -0.189 |
Token | it |
Token ID | 340 |
Feature activation | +3.18e+00 |
Pos logit contributions | |
otion | +0.057 |
Deal | +0.054 |
spine | +0.054 |
acy | +0.053 |
kin | +0.053 |
Neg logit contributions | |
cream | -0.022 |
sailed | -0.018 |
sold | -0.017 |
that | -0.016 |
each | -0.016 |
Token | a |
Token ID | 257 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
cream | +0.162 |
Fin | +0.143 |
Blue | +0.142 |
Jerry | +0.129 |
Red | +0.129 |
Neg logit contributions | |
otion | -0.227 |
Deal | -0.202 |
laptop | -0.200 |
headache | -0.198 |
upon | -0.197 |
Token | " |
Token ID | 366 |
Feature activation | -4.54e+00 |
Pos logit contributions | |
otion | +0.146 |
Deal | +0.126 |
ids | +0.126 |
spine | +0.122 |
upon | +0.120 |
Neg logit contributions | |
cream | -0.082 |
Fin | -0.077 |
Blue | -0.075 |
sold | -0.069 |
sav | -0.069 |
Token | v |
Token ID | 85 |
Feature activation | +1.36e+01 |
Pos logit contributions | |
otion | +0.347 |
upon | +0.329 |
laptop | +0.314 |
ancer | +0.310 |
spine | +0.303 |
Neg logit contributions | |
cream | -0.181 |
Fin | -0.163 |
Blue | -0.159 |
Jerry | -0.149 |
sav | -0.144 |
Token | ac |
Token ID | 330 |
Feature activation | -4.30e+00 |
Pos logit contributions | |
cream | +0.734 |
oc | +0.675 |
Fin | +0.669 |
Blue | +0.632 |
sailed | +0.602 |
Neg logit contributions | |
otion | -0.889 |
Deal | -0.808 |
upon | -0.794 |
laptop | -0.783 |
spine | -0.779 |
Token | ation |
Token ID | 341 |
Feature activation | +3.55e+00 |
Pos logit contributions | |
otion | +0.268 |
upon | +0.250 |
sister | +0.249 |
laptop | +0.235 |
oken | +0.227 |
Neg logit contributions | |
Blue | -0.221 |
Red | -0.207 |
cream | -0.203 |
Fin | -0.202 |
sav | -0.183 |
Token | ". |
Token ID | 1911 |
Feature activation | -1.90e+00 |
Pos logit contributions | |
cream | +0.180 |
Fin | +0.160 |
Blue | +0.156 |
oc | +0.148 |
sailed | +0.144 |
Neg logit contributions | |
otion | -0.250 |
Deal | -0.222 |
upon | -0.218 |
laptop | -0.218 |
spine | -0.216 |
Token | They |
Token ID | 1119 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
otion | +0.144 |
laptop | +0.131 |
upon | +0.131 |
sister | +0.126 |
Deal | +0.126 |
Neg logit contributions | |
cream | -0.085 |
Fin | -0.077 |
Blue | -0.075 |
Jerry | -0.069 |
Tweet | -0.068 |
Token | packed |
Token ID | 11856 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
cream | +0.101 |
Fin | +0.088 |
Blue | +0.086 |
Jerry | +0.077 |
Red | +0.076 |
Neg logit contributions | |
otion | -0.209 |
laptop | -0.191 |
upon | -0.188 |
Deal | -0.187 |
spine | -0.186 |
Token | more |
Token ID | 517 |
Feature activation | +9.14e+00 |
Pos logit contributions | |
cream | +0.261 |
Fin | +0.235 |
Blue | +0.231 |
Jerry | +0.212 |
Red | +0.210 |
Neg logit contributions | |
otion | -0.415 |
upon | -0.382 |
laptop | -0.379 |
sister | -0.369 |
spine | -0.367 |
Token | time |
Token ID | 640 |
Feature activation | -5.82e+00 |
Pos logit contributions | |
cream | +0.622 |
Fin | +0.582 |
Blue | +0.574 |
oc | +0.550 |
Jerry | +0.547 |
Neg logit contributions | |
otion | -0.481 |
laptop | -0.407 |
Deal | -0.403 |
upon | -0.403 |
sister | -0.384 |
Token | , |
Token ID | 11 |
Feature activation | +3.30e-01 |
Pos logit contributions | |
otion | +0.377 |
Deal | +0.340 |
laptop | +0.337 |
upon | +0.328 |
headache | +0.323 |
Neg logit contributions | |
cream | -0.319 |
Fin | -0.287 |
Blue | -0.285 |
oc | -0.268 |
sav | -0.266 |
Token | and |
Token ID | 290 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
cream | +0.018 |
Fin | +0.016 |
Blue | +0.016 |
Jerry | +0.015 |
Red | +0.015 |
Neg logit contributions | |
otion | -0.022 |
upon | -0.020 |
laptop | -0.020 |
Deal | -0.019 |
sister | -0.019 |
Token | Sil |
Token ID | 4243 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
otion | +0.073 |
upon | +0.071 |
sister | +0.070 |
laptop | +0.066 |
Lee | +0.066 |
Neg logit contributions | |
cream | -0.059 |
Blue | -0.056 |
Fin | -0.056 |
Jerry | -0.054 |
Red | -0.053 |
Token | v |
Token ID | 85 |
Feature activation | +1.35e+01 |
Pos logit contributions | |
cream | +0.103 |
Fin | +0.095 |
Blue | +0.094 |
Jerry | +0.090 |
sav | +0.088 |
Neg logit contributions | |
otion | -0.092 |
upon | -0.081 |
laptop | -0.079 |
ids | -0.074 |
acy | -0.074 |
Token | ie |
Token ID | 494 |
Feature activation | +3.60e+00 |
Pos logit contributions | |
cream | +0.771 |
Fin | +0.704 |
Blue | +0.694 |
Jerry | +0.648 |
oc | +0.645 |
Neg logit contributions | |
otion | -0.861 |
laptop | -0.771 |
upon | -0.768 |
Deal | -0.743 |
spine | -0.738 |
Token | repeated |
Token ID | 5100 |
Feature activation | +9.99e-01 |
Pos logit contributions | |
cream | +0.171 |
Fin | +0.159 |
Blue | +0.157 |
Jerry | +0.149 |
Red | +0.145 |
Neg logit contributions | |
otion | -0.255 |
upon | -0.241 |
sister | -0.236 |
laptop | -0.234 |
ids | -0.225 |
Token | it |
Token ID | 340 |
Feature activation | +3.35e+00 |
Pos logit contributions | |
cream | +0.048 |
Fin | +0.041 |
Blue | +0.041 |
oc | +0.038 |
aw | +0.037 |
Neg logit contributions | |
otion | -0.073 |
laptop | -0.066 |
Deal | -0.065 |
upon | -0.065 |
spine | -0.064 |
Token | . |
Token ID | 13 |
Feature activation | -1.29e-01 |
Pos logit contributions | |
cream | +0.177 |
Fin | +0.161 |
Blue | +0.159 |
Jerry | +0.148 |
Red | +0.146 |
Neg logit contributions | |
otion | -0.226 |
upon | -0.204 |
laptop | -0.203 |
Deal | -0.195 |
sister | -0.195 |
Token |
|
Token ID | 198 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
otion | +0.008 |
upon | +0.007 |
laptop | +0.007 |
Lee | +0.007 |
ids | +0.007 |
Neg logit contributions | |
cream | -0.007 |
Fin | -0.007 |
Blue | -0.007 |
Jerry | -0.006 |
oc | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
otion | +0.143 |
laptop | +0.128 |
upon | +0.128 |
spine | +0.122 |
sister | +0.121 |
Neg logit contributions | |
cream | -0.127 |
Fin | -0.120 |
Blue | -0.118 |
Jerry | -0.112 |
Tweet | -0.109 |
Token | It |
Token ID | 632 |
Feature activation | +2.97e+00 |
Pos logit contributions | |
otion | +0.015 |
upon | +0.014 |
laptop | +0.014 |
Deal | +0.013 |
sister | +0.013 |
Neg logit contributions | |
cream | -0.008 |
Fin | -0.007 |
Blue | -0.007 |
Jerry | -0.006 |
Red | -0.006 |
Token | had |
Token ID | 550 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
cream | +0.170 |
Fin | +0.170 |
Blue | +0.157 |
Jerry | +0.156 |
Nut | +0.152 |
Neg logit contributions | |
sister | -0.158 |
laptop | -0.158 |
otion | -0.157 |
upon | -0.155 |
kin | -0.151 |
Token | some |
Token ID | 617 |
Feature activation | +2.39e+00 |
Pos logit contributions | |
cream | +0.062 |
Fin | +0.060 |
Blue | +0.058 |
oc | +0.057 |
Jerry | +0.056 |
Neg logit contributions | |
otion | -0.055 |
upon | -0.054 |
laptop | -0.052 |
sister | -0.051 |
ids | -0.050 |
Token | car |
Token ID | 1097 |
Feature activation | +8.22e-01 |
Pos logit contributions | |
cream | +0.122 |
Fin | +0.112 |
Blue | +0.110 |
Jerry | +0.103 |
Red | +0.100 |
Neg logit contributions | |
otion | -0.163 |
laptop | -0.150 |
upon | -0.150 |
sister | -0.144 |
headache | -0.143 |
Token | v |
Token ID | 85 |
Feature activation | +1.35e+01 |
Pos logit contributions | |
cream | +0.040 |
Fin | +0.038 |
Blue | +0.037 |
Jerry | +0.035 |
Tweet | +0.034 |
Neg logit contributions | |
otion | -0.057 |
laptop | -0.053 |
upon | -0.052 |
sister | -0.050 |
spine | -0.050 |
Token | ings |
Token ID | 654 |
Feature activation | -6.53e+00 |
Pos logit contributions | |
cream | +0.858 |
Fin | +0.774 |
oc | +0.757 |
Blue | +0.750 |
He | +0.730 |
Neg logit contributions | |
otion | -0.735 |
Deal | -0.671 |
laptop | -0.668 |
spine | -0.663 |
headache | -0.649 |
Token | and |
Token ID | 290 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
otion | +0.432 |
upon | +0.387 |
laptop | +0.386 |
Deal | +0.371 |
sister | +0.370 |
Neg logit contributions | |
cream | -0.354 |
Fin | -0.322 |
Blue | -0.318 |
Jerry | -0.296 |
Red | -0.293 |
Token | letters |
Token ID | 7475 |
Feature activation | +1.65e+00 |
Pos logit contributions | |
otion | +0.084 |
laptop | +0.078 |
upon | +0.077 |
sister | +0.075 |
headache | +0.073 |
Neg logit contributions | |
cream | -0.073 |
Fin | -0.069 |
Blue | -0.068 |
Jerry | -0.064 |
Red | -0.062 |
Token | on |
Token ID | 319 |
Feature activation | -2.49e+00 |
Pos logit contributions | |
cream | +0.092 |
Fin | +0.085 |
Blue | +0.084 |
Jerry | +0.078 |
Tweet | +0.077 |
Neg logit contributions | |
otion | -0.107 |
upon | -0.096 |
laptop | -0.095 |
sister | -0.091 |
spine | -0.090 |
Token | it |
Token ID | 340 |
Feature activation | +3.26e+00 |
Pos logit contributions | |
otion | +0.183 |
laptop | +0.168 |
upon | +0.167 |
sister | +0.161 |
headache | +0.159 |
Neg logit contributions | |
cream | -0.116 |
Fin | -0.103 |
Blue | -0.101 |
Jerry | -0.094 |
Tweet | -0.092 |
Token | vehicle |
Token ID | 4038 |
Feature activation | +2.89e+00 |
Pos logit contributions | |
cream | +0.094 |
Blue | +0.089 |
Fin | +0.088 |
Jerry | +0.082 |
Tweet | +0.081 |
Neg logit contributions | |
otion | -0.105 |
laptop | -0.101 |
upon | -0.098 |
sister | -0.097 |
curtain | -0.094 |
Token | and |
Token ID | 290 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
Fin | +0.169 |
cream | +0.168 |
Blue | +0.163 |
Jerry | +0.155 |
Tweet | +0.152 |
Neg logit contributions | |
otion | -0.172 |
laptop | -0.159 |
upon | -0.157 |
spine | -0.153 |
Lee | -0.145 |
Token | happily |
Token ID | 18177 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
otion | +0.084 |
upon | +0.078 |
laptop | +0.077 |
sister | +0.075 |
spine | +0.074 |
Neg logit contributions | |
cream | -0.057 |
Fin | -0.054 |
Blue | -0.053 |
Jerry | -0.049 |
Red | -0.048 |
Token | exclaimed |
Token ID | 33552 |
Feature activation | +1.80e+00 |
Pos logit contributions | |
cream | +0.005 |
Blue | +0.005 |
Fin | +0.005 |
Red | +0.005 |
Jerry | +0.005 |
Neg logit contributions | |
otion | -0.008 |
upon | -0.007 |
laptop | -0.007 |
sister | -0.007 |
spine | -0.007 |
Token | " |
Token ID | 366 |
Feature activation | -4.46e+00 |
Pos logit contributions | |
cream | +0.083 |
He | +0.071 |
Fin | +0.068 |
Blue | +0.068 |
They | +0.066 |
Neg logit contributions | |
otion | -0.129 |
laptop | -0.125 |
spine | -0.122 |
Deal | -0.118 |
headache | -0.117 |
Token | v |
Token ID | 85 |
Feature activation | +1.35e+01 |
Pos logit contributions | |
upon | +0.302 |
ancer | +0.295 |
otion | +0.289 |
Deal | +0.283 |
acy | +0.276 |
Neg logit contributions | |
cream | -0.159 |
Fin | -0.153 |
Blue | -0.150 |
bank | -0.148 |
Jerry | -0.146 |
Token | room |
Token ID | 3823 |
Feature activation | -3.31e+00 |
Pos logit contributions | |
cream | +0.751 |
Fin | +0.676 |
oc | +0.657 |
Blue | +0.643 |
Jerry | +0.617 |
Neg logit contributions | |
otion | -0.871 |
laptop | -0.784 |
spine | -0.782 |
upon | -0.777 |
Deal | -0.768 |
Token | v |
Token ID | 410 |
Feature activation | +1.98e+00 |
Pos logit contributions | |
otion | +0.218 |
laptop | +0.197 |
upon | +0.195 |
sister | +0.193 |
headache | +0.188 |
Neg logit contributions | |
cream | -0.172 |
Blue | -0.161 |
Fin | -0.160 |
Jerry | -0.149 |
Red | -0.148 |
Token | room |
Token ID | 3823 |
Feature activation | -3.33e+00 |
Pos logit contributions | |
cream | +0.123 |
Fin | +0.115 |
Blue | +0.114 |
Jerry | +0.107 |
Red | +0.106 |
Neg logit contributions | |
otion | -0.114 |
laptop | -0.100 |
upon | -0.100 |
Deal | -0.096 |
sister | -0.095 |
Token | !" |
Token ID | 2474 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
otion | +0.231 |
laptop | +0.211 |
upon | +0.207 |
sister | +0.203 |
headache | +0.200 |
Neg logit contributions | |
cream | -0.163 |
Fin | -0.151 |
Blue | -0.150 |
Jerry | -0.141 |
Red | -0.137 |
Token |
|
Token ID | 198 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
cream | +0.074 |
Fin | +0.057 |
Blue | +0.056 |
Jerry | +0.051 |
He | +0.051 |
Neg logit contributions | |
otion | -0.180 |
laptop | -0.162 |
upon | -0.161 |
Deal | -0.158 |
sister | -0.158 |
Token | thing |
Token ID | 1517 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
cream | +0.268 |
Fin | +0.238 |
Blue | +0.235 |
oc | +0.219 |
Red | +0.217 |
Neg logit contributions | |
otion | -0.367 |
upon | -0.330 |
laptop | -0.326 |
Deal | -0.325 |
spine | -0.315 |
Token | . |
Token ID | 13 |
Feature activation | -2.20e-01 |
Pos logit contributions | |
cream | +0.052 |
Blue | +0.046 |
Fin | +0.045 |
Jerry | +0.042 |
Red | +0.041 |
Neg logit contributions | |
otion | -0.075 |
Deal | -0.068 |
laptop | -0.067 |
upon | -0.067 |
headache | -0.066 |
Token | She |
Token ID | 1375 |
Feature activation | +9.30e-02 |
Pos logit contributions | |
otion | +0.016 |
laptop | +0.014 |
upon | +0.014 |
headache | +0.014 |
sister | +0.013 |
Neg logit contributions | |
cream | -0.011 |
Blue | -0.010 |
Fin | -0.010 |
Red | -0.009 |
Jerry | -0.009 |
Token | shouted |
Token ID | 17293 |
Feature activation | -6.87e-01 |
Pos logit contributions | |
cream | +0.005 |
Fin | +0.005 |
Blue | +0.005 |
Jerry | +0.005 |
Nut | +0.004 |
Neg logit contributions | |
sister | -0.005 |
upon | -0.005 |
otion | -0.005 |
laptop | -0.005 |
ids | -0.005 |
Token | " |
Token ID | 366 |
Feature activation | -4.58e+00 |
Pos logit contributions | |
otion | +0.053 |
laptop | +0.049 |
Deal | +0.049 |
ancer | +0.048 |
spine | +0.047 |
Neg logit contributions | |
that | -0.024 |
Blue | -0.023 |
The | -0.023 |
cream | -0.023 |
They | -0.023 |
Token | v |
Token ID | 85 |
Feature activation | +1.35e+01 |
Pos logit contributions | |
otion | +0.303 |
upon | +0.295 |
Deal | +0.284 |
ancer | +0.283 |
laptop | +0.277 |
Neg logit contributions | |
cream | -0.209 |
Fin | -0.187 |
Blue | -0.176 |
oc | -0.176 |
Jerry | -0.167 |
Token | room |
Token ID | 3823 |
Feature activation | -3.35e+00 |
Pos logit contributions | |
cream | +0.719 |
oc | +0.660 |
Fin | +0.652 |
Blue | +0.607 |
aw | +0.606 |
Neg logit contributions | |
otion | -0.857 |
spine | -0.820 |
Deal | -0.786 |
laptop | -0.785 |
upon | -0.785 |
Token | v |
Token ID | 410 |
Feature activation | +1.93e+00 |
Pos logit contributions | |
otion | +0.230 |
laptop | +0.207 |
upon | +0.207 |
sister | +0.200 |
Deal | +0.199 |
Neg logit contributions | |
cream | -0.173 |
Fin | -0.156 |
Blue | -0.155 |
Jerry | -0.143 |
Red | -0.142 |
Token | room |
Token ID | 3823 |
Feature activation | -3.36e+00 |
Pos logit contributions | |
cream | +0.124 |
Fin | +0.115 |
Blue | +0.114 |
Jerry | +0.107 |
Red | +0.106 |
Neg logit contributions | |
otion | -0.109 |
laptop | -0.095 |
upon | -0.095 |
Deal | -0.091 |
sister | -0.090 |
Token | " |
Token ID | 1 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
otion | +0.239 |
laptop | +0.216 |
upon | +0.214 |
sister | +0.209 |
headache | +0.207 |
Neg logit contributions | |
cream | -0.164 |
Fin | -0.148 |
Blue | -0.147 |
Jerry | -0.136 |
Red | -0.133 |
Token | and |
Token ID | 290 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
cream | +0.063 |
Blue | +0.056 |
Fin | +0.055 |
Red | +0.052 |
Jerry | +0.052 |
Neg logit contributions | |
otion | -0.087 |
Deal | -0.078 |
laptop | -0.078 |
sister | -0.077 |
upon | -0.077 |
Token | The |
Token ID | 383 |
Feature activation | +9.77e-01 |
Pos logit contributions | |
otion | +0.009 |
laptop | +0.008 |
upon | +0.008 |
Deal | +0.008 |
sister | +0.008 |
Neg logit contributions | |
cream | -0.006 |
Fin | -0.005 |
Blue | -0.005 |
Jerry | -0.005 |
Tweet | -0.005 |
Token | mighty |
Token ID | 18680 |
Feature activation | +2.26e+00 |
Pos logit contributions | |
cream | +0.061 |
Blue | +0.056 |
Fin | +0.056 |
Red | +0.053 |
Tweet | +0.052 |
Neg logit contributions | |
otion | -0.058 |
upon | -0.050 |
laptop | -0.050 |
Deal | -0.049 |
spine | -0.048 |
Token | ambulance |
Token ID | 22536 |
Feature activation | -3.59e+00 |
Pos logit contributions | |
cream | +0.131 |
Blue | +0.120 |
Fin | +0.120 |
oc | +0.114 |
Red | +0.112 |
Neg logit contributions | |
otion | -0.138 |
Deal | -0.124 |
upon | -0.123 |
acy | -0.122 |
laptop | -0.120 |
Token | went |
Token ID | 1816 |
Feature activation | +6.00e-01 |
Pos logit contributions | |
otion | +0.237 |
laptop | +0.221 |
sister | +0.219 |
upon | +0.217 |
ids | +0.207 |
Neg logit contributions | |
cream | -0.175 |
Fin | -0.174 |
Blue | -0.164 |
Jerry | -0.159 |
Red | -0.154 |
Token | " |
Token ID | 366 |
Feature activation | -4.57e+00 |
Pos logit contributions | |
cream | +0.032 |
Fin | +0.029 |
Blue | +0.029 |
Jerry | +0.027 |
oc | +0.027 |
Neg logit contributions | |
otion | -0.040 |
upon | -0.036 |
laptop | -0.036 |
sister | -0.035 |
Deal | -0.035 |
Token | v |
Token ID | 85 |
Feature activation | +1.35e+01 |
Pos logit contributions | |
otion | +0.329 |
upon | +0.318 |
ancer | +0.298 |
laptop | +0.290 |
Deal | +0.287 |
Neg logit contributions | |
cream | -0.182 |
Fin | -0.173 |
Blue | -0.163 |
Jerry | -0.158 |
sav | -0.149 |
Token | room |
Token ID | 3823 |
Feature activation | -3.37e+00 |
Pos logit contributions | |
cream | +0.742 |
oc | +0.697 |
Fin | +0.692 |
Blue | +0.655 |
aw | +0.655 |
Neg logit contributions | |
otion | -0.818 |
spine | -0.767 |
upon | -0.766 |
laptop | -0.759 |
Deal | -0.755 |
Token | , |
Token ID | 11 |
Feature activation | +2.39e-01 |
Pos logit contributions | |
otion | +0.229 |
laptop | +0.206 |
upon | +0.205 |
sister | +0.199 |
headache | +0.197 |
Neg logit contributions | |
cream | -0.176 |
Fin | -0.160 |
Blue | -0.158 |
Jerry | -0.148 |
Red | -0.145 |
Token | v |
Token ID | 410 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
cream | +0.013 |
Blue | +0.012 |
Fin | +0.012 |
oc | +0.011 |
Red | +0.011 |
Neg logit contributions | |
otion | -0.016 |
laptop | -0.014 |
upon | -0.014 |
Deal | -0.013 |
spine | -0.013 |
Token | room |
Token ID | 3823 |
Feature activation | -3.39e+00 |
Pos logit contributions | |
cream | +0.129 |
Fin | +0.120 |
Blue | +0.119 |
Jerry | +0.112 |
Red | +0.111 |
Neg logit contributions | |
otion | -0.107 |
laptop | -0.093 |
upon | -0.093 |
Deal | -0.088 |
sister | -0.088 |
Token | " |
Token ID | 1 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
otion | +0.228 |
laptop | +0.204 |
upon | +0.202 |
sister | +0.197 |
headache | +0.195 |
Neg logit contributions | |
cream | -0.179 |
Fin | -0.163 |
Blue | -0.161 |
Jerry | -0.150 |
Red | -0.148 |
Token | . |
Token ID | 13 |
Feature activation | -1.96e-01 |
Pos logit contributions | |
otion | +0.820 |
Deal | +0.750 |
laptop | +0.744 |
upon | +0.726 |
aughter | +0.724 |
Neg logit contributions | |
cream | -0.497 |
Fin | -0.431 |
Blue | -0.426 |
He | -0.409 |
sav | -0.404 |
Token |
|
Token ID | 198 |
Feature activation | +1.27e+00 |
Pos logit contributions | |
otion | +0.011 |
upon | +0.010 |
laptop | +0.010 |
Lee | +0.010 |
Deal | +0.010 |
Neg logit contributions | |
cream | -0.012 |
Fin | -0.011 |
Blue | -0.011 |
Jerry | -0.010 |
Tweet | -0.010 |
Token |
|
Token ID | 198 |
Feature activation | +1.27e+00 |
Pos logit contributions | |
cream | +0.070 |
Fin | +0.059 |
Blue | +0.056 |
sav | +0.055 |
sold | +0.054 |
Neg logit contributions | |
otion | -0.082 |
upon | -0.081 |
laptop | -0.072 |
Deal | -0.071 |
ids | -0.069 |
Token | " |
Token ID | 1 |
Feature activation | +1.39e+00 |
Pos logit contributions | |
cream | +0.080 |
rust | +0.072 |
tast | +0.069 |
rob | +0.069 |
sav | +0.067 |
Neg logit contributions | |
upon | -0.061 |
laptop | -0.056 |
911 | -0.054 |
otion | -0.054 |
stool | -0.053 |
Token | Sil |
Token ID | 15086 |
Feature activation | +5.23e+00 |
Pos logit contributions | |
cream | +0.064 |
oc | +0.054 |
bank | +0.049 |
aw | +0.049 |
Fin | +0.049 |
Neg logit contributions | |
otion | -0.097 |
upon | -0.092 |
Deal | -0.091 |
ancer | -0.090 |
Lee | -0.086 |
Token | v |
Token ID | 85 |
Feature activation | +1.35e+01 |
Pos logit contributions | |
cream | +0.365 |
Fin | +0.339 |
Blue | +0.336 |
Jerry | +0.318 |
Red | +0.316 |
Neg logit contributions | |
otion | -0.268 |
upon | -0.231 |
laptop | -0.231 |
Deal | -0.220 |
sister | -0.219 |
Token | ie |
Token ID | 494 |
Feature activation | +3.57e+00 |
Pos logit contributions | |
cream | +0.697 |
Fin | +0.642 |
oc | +0.618 |
Blue | +0.615 |
Jerry | +0.579 |
Neg logit contributions | |
otion | -0.907 |
laptop | -0.823 |
upon | -0.819 |
Deal | -0.811 |
spine | -0.808 |
Token | ," |
Token ID | 553 |
Feature activation | +2.69e+00 |
Pos logit contributions | |
cream | +0.208 |
Fin | +0.189 |
Blue | +0.187 |
oc | +0.177 |
Jerry | +0.174 |
Neg logit contributions | |
otion | -0.222 |
laptop | -0.197 |
upon | -0.195 |
Deal | -0.192 |
spine | -0.187 |
Token | he |
Token ID | 339 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
Fin | +0.146 |
cream | +0.144 |
Blue | +0.136 |
Jerry | +0.133 |
Red | +0.131 |
Neg logit contributions | |
otion | -0.145 |
sister | -0.142 |
Lee | -0.138 |
upon | -0.138 |
laptop | -0.137 |
Token | called |
Token ID | 1444 |
Feature activation | -6.75e-01 |
Pos logit contributions | |
cream | +0.122 |
Fin | +0.106 |
Blue | +0.105 |
Red | +0.100 |
Jerry | +0.096 |
Neg logit contributions | |
otion | -0.180 |
upon | -0.172 |
sister | -0.167 |
laptop | -0.165 |
headache | -0.160 |
Token | to |
Token ID | 284 |
Feature activation | +8.33e-01 |
Pos logit contributions | |
otion | +0.051 |
Deal | +0.050 |
acy | +0.048 |
ancer | +0.047 |
laptop | +0.046 |
Neg logit contributions | |
cream | -0.023 |
Blue | -0.023 |
sailed | -0.022 |
He | -0.021 |
penter | -0.020 |
Token | the |
Token ID | 262 |
Feature activation | -5.75e-01 |
Pos logit contributions | |
otion | +0.057 |
laptop | +0.051 |
Deal | +0.050 |
spine | +0.050 |
upon | +0.050 |
Neg logit contributions | |
cream | -0.034 |
Blue | -0.029 |
Fin | -0.029 |
Jerry | -0.027 |
Red | -0.027 |
Token | best |
Token ID | 1266 |
Feature activation | +5.11e+00 |
Pos logit contributions | |
otion | +0.044 |
upon | +0.038 |
ids | +0.038 |
laptop | +0.038 |
Deal | +0.037 |
Neg logit contributions | |
cream | -0.025 |
Blue | -0.023 |
Fin | -0.023 |
sav | -0.022 |
sold | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | -3.89e-01 |
Pos logit contributions | |
cream | +0.237 |
Fin | +0.198 |
sold | +0.197 |
sav | +0.192 |
This | +0.188 |
Neg logit contributions | |
otion | -0.371 |
Deal | -0.354 |
ids | -0.329 |
ancer | -0.329 |
spine | -0.329 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.90e+00 |
Pos logit contributions | |
otion | +0.027 |
upon | +0.025 |
Deal | +0.024 |
Lee | +0.023 |
sister | +0.023 |
Neg logit contributions | |
cream | -0.019 |
Fin | -0.017 |
Blue | -0.017 |
Red | -0.016 |
rob | -0.016 |
Token | Once |
Token ID | 7454 |
Feature activation | +6.42e-01 |
Pos logit contributions | |
otion | +0.232 |
upon | +0.214 |
laptop | +0.209 |
Deal | +0.203 |
sister | +0.203 |
Neg logit contributions | |
cream | -0.119 |
Fin | -0.103 |
Blue | -0.101 |
Jerry | -0.091 |
Red | -0.090 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
Fin | +0.044 |
The | +0.044 |
He | +0.043 |
One | +0.043 |
Bob | +0.043 |
Neg logit contributions | |
laptop | -0.027 |
spine | -0.024 |
Deal | -0.024 |
otion | -0.023 |
scalp | -0.023 |
Token | a |
Token ID | 257 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
otion | +1.311 |
laptop | +1.181 |
upon | +1.168 |
Deal | +1.138 |
spine | +1.137 |
Neg logit contributions | |
cream | -0.983 |
Fin | -0.890 |
Blue | -0.885 |
Jerry | -0.819 |
Red | -0.808 |
Token | time |
Token ID | 640 |
Feature activation | -6.10e+00 |
Pos logit contributions | |
otion | +0.100 |
laptop | +0.085 |
upon | +0.084 |
sister | +0.080 |
Deal | +0.080 |
Neg logit contributions | |
cream | -0.141 |
Fin | -0.134 |
Blue | -0.133 |
Jerry | -0.127 |
oc | -0.125 |
Token | , |
Token ID | 11 |
Feature activation | -6.61e-03 |
Pos logit contributions | |
laptop | +0.502 |
spine | +0.483 |
forehead | +0.477 |
Deal | +0.474 |
scalp | +0.470 |
Neg logit contributions | |
Fin | -0.196 |
Bob | -0.191 |
Blue | -0.175 |
Jerry | -0.169 |
One | -0.165 |
Token | in |
Token ID | 287 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
laptop | +0.001 |
spine | +0.000 |
Deal | +0.000 |
otion | +0.000 |
headache | +0.000 |
Neg logit contributions | |
Blue | -0.000 |
Fin | -0.000 |
The | -0.000 |
Tweet | -0.000 |
It | -0.000 |
Token | a |
Token ID | 257 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
otion | +0.102 |
laptop | +0.092 |
upon | +0.090 |
spine | +0.090 |
Deal | +0.089 |
Neg logit contributions | |
cream | -0.067 |
Fin | -0.061 |
Blue | -0.061 |
Jerry | -0.056 |
Tweet | -0.055 |
Token | all |
Token ID | 477 |
Feature activation | +3.46e+00 |
Pos logit contributions | |
otion | +0.220 |
laptop | +0.198 |
upon | +0.197 |
Deal | +0.191 |
sister | +0.189 |
Neg logit contributions | |
cream | -0.170 |
Fin | -0.154 |
Blue | -0.153 |
Jerry | -0.141 |
Red | -0.140 |
Token | day |
Token ID | 1110 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
cream | +0.236 |
Fin | +0.215 |
Blue | +0.210 |
Nut | +0.208 |
Jerry | +0.206 |
Neg logit contributions | |
otion | -0.177 |
upon | -0.172 |
sister | -0.162 |
curtain | -0.153 |
laptop | -0.152 |
Token | . |
Token ID | 13 |
Feature activation | -2.66e-01 |
Pos logit contributions | |
otion | +0.086 |
upon | +0.076 |
laptop | +0.075 |
Deal | +0.075 |
sister | +0.073 |
Neg logit contributions | |
cream | -0.067 |
Blue | -0.062 |
Fin | -0.062 |
Jerry | -0.057 |
Red | -0.057 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.93e+00 |
Pos logit contributions | |
otion | +0.018 |
upon | +0.017 |
Deal | +0.016 |
Lee | +0.015 |
laptop | +0.015 |
Neg logit contributions | |
cream | -0.014 |
Fin | -0.012 |
Blue | -0.012 |
Red | -0.011 |
oc | -0.011 |
Token | Once |
Token ID | 7454 |
Feature activation | +6.52e-01 |
Pos logit contributions | |
otion | +0.235 |
upon | +0.217 |
laptop | +0.213 |
Deal | +0.207 |
sister | +0.207 |
Neg logit contributions | |
cream | -0.119 |
Fin | -0.103 |
Blue | -0.101 |
Jerry | -0.091 |
Red | -0.090 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
Fin | +0.045 |
The | +0.045 |
One | +0.044 |
He | +0.044 |
Blue | +0.044 |
Neg logit contributions | |
laptop | -0.026 |
otion | -0.024 |
Deal | -0.024 |
spine | -0.024 |
scalp | -0.023 |
Token | a |
Token ID | 257 |
Feature activation | -2.00e+00 |
Pos logit contributions | |
otion | +1.241 |
upon | +1.126 |
laptop | +1.107 |
sister | +1.067 |
Deal | +1.064 |
Neg logit contributions | |
cream | -1.046 |
Fin | -0.950 |
Blue | -0.934 |
Jerry | -0.872 |
oc | -0.864 |
Token | time |
Token ID | 640 |
Feature activation | -6.07e+00 |
Pos logit contributions | |
otion | +0.104 |
laptop | +0.088 |
upon | +0.086 |
sister | +0.084 |
spine | +0.083 |
Neg logit contributions | |
cream | -0.134 |
Fin | -0.129 |
Blue | -0.128 |
Jerry | -0.122 |
oc | -0.120 |
Token | , |
Token ID | 11 |
Feature activation | +5.14e-02 |
Pos logit contributions | |
laptop | +0.469 |
otion | +0.451 |
spine | +0.449 |
forehead | +0.446 |
headache | +0.436 |
Neg logit contributions | |
Fin | -0.220 |
Bob | -0.208 |
Blue | -0.198 |
Jerry | -0.190 |
One | -0.189 |
Token | there |
Token ID | 612 |
Feature activation | -2.00e+00 |
Pos logit contributions | |
Blue | +0.002 |
Fin | +0.002 |
cream | +0.002 |
The | +0.002 |
Tweet | +0.002 |
Neg logit contributions | |
laptop | -0.004 |
otion | -0.004 |
spine | -0.003 |
headache | -0.003 |
Deal | -0.003 |
Token | was |
Token ID | 373 |
Feature activation | -6.46e-01 |
Pos logit contributions | |
otion | +0.175 |
laptop | +0.163 |
Deal | +0.160 |
upon | +0.159 |
spine | +0.159 |
Neg logit contributions | |
cream | -0.065 |
Fin | -0.056 |
Blue | -0.055 |
Jerry | -0.048 |
oc | -0.048 |
Token | had |
Token ID | 550 |
Feature activation | +7.11e-01 |
Pos logit contributions | |
otion | +0.104 |
upon | +0.100 |
sister | +0.094 |
spine | +0.093 |
laptop | +0.092 |
Neg logit contributions | |
cream | -0.080 |
Fin | -0.070 |
Blue | -0.068 |
oc | -0.067 |
Red | -0.066 |
Token | fun |
Token ID | 1257 |
Feature activation | +8.65e+00 |
Pos logit contributions | |
oc | +0.051 |
cream | +0.050 |
rob | +0.044 |
penter | +0.044 |
bike | +0.042 |
Neg logit contributions | |
upon | -0.031 |
bed | -0.029 |
sister | -0.028 |
eyes | -0.028 |
toes | -0.026 |
Token | . |
Token ID | 13 |
Feature activation | -5.17e-01 |
Pos logit contributions | |
cream | +0.528 |
Fin | +0.475 |
Blue | +0.469 |
Jerry | +0.450 |
Red | +0.442 |
Neg logit contributions | |
otion | -0.532 |
upon | -0.484 |
laptop | -0.458 |
sister | -0.457 |
spine | -0.445 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.93e+00 |
Pos logit contributions | |
otion | +0.031 |
upon | +0.030 |
Deal | +0.028 |
sister | +0.028 |
laptop | +0.028 |
Neg logit contributions | |
cream | -0.030 |
Fin | -0.027 |
Blue | -0.025 |
rob | -0.025 |
Jerry | -0.025 |
Token | Once |
Token ID | 7454 |
Feature activation | +7.16e-01 |
Pos logit contributions | |
otion | +0.239 |
upon | +0.219 |
laptop | +0.219 |
Deal | +0.212 |
sister | +0.212 |
Neg logit contributions | |
cream | -0.113 |
Fin | -0.099 |
Blue | -0.097 |
Jerry | -0.087 |
Red | -0.086 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
The | +0.052 |
Fin | +0.051 |
that | +0.051 |
He | +0.051 |
One | +0.050 |
Neg logit contributions | |
laptop | -0.026 |
Deal | -0.024 |
scalp | -0.023 |
spine | -0.023 |
otion | -0.023 |
Token | a |
Token ID | 257 |
Feature activation | -1.96e+00 |
Pos logit contributions | |
otion | +1.305 |
laptop | +1.174 |
upon | +1.164 |
Deal | +1.132 |
spine | +1.129 |
Neg logit contributions | |
cream | -0.986 |
Fin | -0.893 |
Blue | -0.887 |
Jerry | -0.821 |
Red | -0.812 |
Token | time |
Token ID | 640 |
Feature activation | -6.02e+00 |
Pos logit contributions | |
otion | +0.101 |
laptop | +0.086 |
upon | +0.084 |
sister | +0.082 |
spine | +0.081 |
Neg logit contributions | |
cream | -0.133 |
Fin | -0.128 |
Blue | -0.127 |
Jerry | -0.121 |
oc | -0.120 |
Token | , |
Token ID | 11 |
Feature activation | +1.81e-02 |
Pos logit contributions | |
laptop | +0.481 |
forehead | +0.462 |
spine | +0.460 |
scalp | +0.456 |
Deal | +0.454 |
Neg logit contributions | |
Fin | -0.201 |
Bob | -0.192 |
Blue | -0.184 |
One | -0.175 |
Jerry | -0.172 |
Token | there |
Token ID | 612 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
Blue | +0.001 |
Fin | +0.001 |
The | +0.001 |
cream | +0.001 |
Tweet | +0.001 |
Neg logit contributions | |
laptop | -0.001 |
otion | -0.001 |
spine | -0.001 |
Deal | -0.001 |
headache | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | -6.35e-01 |
Pos logit contributions | |
otion | +0.167 |
laptop | +0.154 |
upon | +0.153 |
Deal | +0.150 |
spine | +0.150 |
Neg logit contributions | |
cream | -0.069 |
Fin | -0.060 |
Blue | -0.059 |
Jerry | -0.052 |
oc | -0.051 |
Token | the |
Token ID | 262 |
Feature activation | -6.28e-01 |
Pos logit contributions | |
cream | +0.223 |
Blue | +0.202 |
Fin | +0.201 |
Red | +0.188 |
Jerry | +0.187 |
Neg logit contributions | |
otion | -0.250 |
upon | -0.221 |
laptop | -0.220 |
spine | -0.213 |
Deal | -0.212 |
Token | doctor |
Token ID | 6253 |
Feature activation | -1.95e+00 |
Pos logit contributions | |
otion | +0.042 |
upon | +0.037 |
spine | +0.036 |
ids | +0.036 |
laptop | +0.036 |
Neg logit contributions | |
cream | -0.033 |
Blue | -0.031 |
Red | -0.029 |
Fin | -0.029 |
He | -0.029 |
Token | . |
Token ID | 13 |
Feature activation | -3.59e-01 |
Pos logit contributions | |
otion | +0.130 |
upon | +0.114 |
laptop | +0.114 |
Deal | +0.112 |
headache | +0.111 |
Neg logit contributions | |
cream | -0.107 |
Blue | -0.094 |
Fin | -0.094 |
oc | -0.091 |
Jerry | -0.087 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.86e+00 |
Pos logit contributions | |
otion | +0.025 |
upon | +0.024 |
laptop | +0.023 |
sister | +0.023 |
Deal | +0.023 |
Neg logit contributions | |
cream | -0.017 |
Fin | -0.016 |
Blue | -0.015 |
Jerry | -0.015 |
Tweet | -0.014 |
Token | Once |
Token ID | 7454 |
Feature activation | +6.91e-01 |
Pos logit contributions | |
otion | +0.240 |
laptop | +0.221 |
upon | +0.220 |
Deal | +0.215 |
spine | +0.214 |
Neg logit contributions | |
cream | -0.103 |
Fin | -0.089 |
Blue | -0.088 |
Jerry | -0.079 |
Red | -0.077 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
The | +0.053 |
He | +0.053 |
that | +0.053 |
One | +0.052 |
cream | +0.051 |
Neg logit contributions | |
laptop | -0.021 |
otion | -0.020 |
Deal | -0.020 |
scalp | -0.019 |
spine | -0.018 |
Token | a |
Token ID | 257 |
Feature activation | -1.94e+00 |
Pos logit contributions | |
otion | +1.310 |
laptop | +1.177 |
upon | +1.167 |
Deal | +1.137 |
spine | +1.135 |
Neg logit contributions | |
cream | -0.984 |
Fin | -0.887 |
Blue | -0.884 |
Jerry | -0.818 |
Red | -0.809 |
Token | time |
Token ID | 640 |
Feature activation | -6.07e+00 |
Pos logit contributions | |
otion | +0.112 |
laptop | +0.094 |
ancer | +0.092 |
ids | +0.091 |
sister | +0.091 |
Neg logit contributions | |
cream | -0.117 |
Fin | -0.116 |
Blue | -0.115 |
Jerry | -0.111 |
oc | -0.109 |
Token | , |
Token ID | 11 |
Feature activation | +2.46e-02 |
Pos logit contributions | |
laptop | +0.496 |
spine | +0.484 |
scalp | +0.481 |
forehead | +0.476 |
Deal | +0.474 |
Neg logit contributions | |
Bob | -0.185 |
Fin | -0.180 |
Blue | -0.175 |
Jerry | -0.164 |
oc | -0.164 |
Token | there |
Token ID | 612 |
Feature activation | -1.96e+00 |
Pos logit contributions | |
Blue | +0.001 |
cream | +0.001 |
Fin | +0.001 |
He | +0.001 |
The | +0.001 |
Neg logit contributions | |
laptop | -0.002 |
otion | -0.002 |
spine | -0.002 |
Deal | -0.002 |
headache | -0.002 |
Token | was |
Token ID | 373 |
Feature activation | -6.36e-01 |
Pos logit contributions | |
otion | +0.166 |
laptop | +0.152 |
upon | +0.152 |
Deal | +0.148 |
spine | +0.148 |
Neg logit contributions | |
cream | -0.071 |
Fin | -0.061 |
Blue | -0.060 |
Jerry | -0.053 |
Red | -0.052 |
Token | love |
Token ID | 1842 |
Feature activation | -7.88e-01 |
Pos logit contributions | |
cream | +0.124 |
Fin | +0.114 |
Blue | +0.112 |
Jerry | +0.106 |
Red | +0.105 |
Neg logit contributions | |
otion | -0.119 |
upon | -0.105 |
laptop | -0.105 |
Deal | -0.101 |
sister | -0.101 |
Token | Mom |
Token ID | 11254 |
Feature activation | -3.00e+00 |
Pos logit contributions | |
otion | +0.047 |
upon | +0.041 |
laptop | +0.041 |
Deal | +0.040 |
spine | +0.040 |
Neg logit contributions | |
cream | -0.049 |
Fin | -0.044 |
Blue | -0.044 |
Jerry | -0.042 |
Red | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | -6.28e-01 |
Pos logit contributions | |
otion | +0.191 |
upon | +0.175 |
laptop | +0.175 |
sister | +0.171 |
headache | +0.164 |
Neg logit contributions | |
cream | -0.164 |
Fin | -0.159 |
Blue | -0.153 |
Jerry | -0.144 |
Tweet | -0.144 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.97e+00 |
Pos logit contributions | |
otion | +0.051 |
laptop | +0.046 |
upon | +0.046 |
spine | +0.045 |
sister | +0.045 |
Neg logit contributions | |
cream | -0.025 |
Fin | -0.022 |
Blue | -0.022 |
Jerry | -0.020 |
Red | -0.019 |
Token | Once |
Token ID | 7454 |
Feature activation | +7.05e-01 |
Pos logit contributions | |
otion | +0.207 |
spine | +0.198 |
laptop | +0.194 |
headache | +0.189 |
Deal | +0.187 |
Neg logit contributions | |
cream | -0.140 |
Blue | -0.131 |
Fin | -0.129 |
This | -0.121 |
Red | -0.119 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
that | +0.050 |
The | +0.049 |
He | +0.048 |
One | +0.048 |
| +0.048 |
Neg logit contributions | |
laptop | -0.025 |
scalp | -0.025 |
Deal | -0.024 |
spine | -0.024 |
forehead | -0.024 |
Token | a |
Token ID | 257 |
Feature activation | -1.96e+00 |
Pos logit contributions | |
otion | +1.331 |
laptop | +1.198 |
upon | +1.180 |
spine | +1.161 |
Deal | +1.159 |
Neg logit contributions | |
cream | -0.962 |
Blue | -0.869 |
Fin | -0.867 |
Jerry | -0.799 |
Red | -0.793 |
Token | time |
Token ID | 640 |
Feature activation | -6.02e+00 |
Pos logit contributions | |
otion | +0.107 |
laptop | +0.090 |
upon | +0.087 |
sister | +0.086 |
ancer | +0.086 |
Neg logit contributions | |
cream | -0.126 |
Fin | -0.122 |
Blue | -0.121 |
Jerry | -0.116 |
oc | -0.115 |
Token | , |
Token ID | 11 |
Feature activation | +2.08e-02 |
Pos logit contributions | |
laptop | +0.488 |
forehead | +0.482 |
spine | +0.481 |
scalp | +0.480 |
headache | +0.470 |
Neg logit contributions | |
Bob | -0.166 |
that | -0.156 |
Fin | -0.156 |
One | -0.151 |
Blue | -0.148 |
Token | there |
Token ID | 612 |
Feature activation | -1.94e+00 |
Pos logit contributions | |
Blue | +0.001 |
The | +0.001 |
It | +0.001 |
Fin | +0.001 |
He | +0.001 |
Neg logit contributions | |
laptop | -0.002 |
spine | -0.001 |
Deal | -0.001 |
headache | -0.001 |
otion | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | -5.37e-01 |
Pos logit contributions | |
otion | +0.171 |
laptop | +0.160 |
Deal | +0.158 |
spine | +0.157 |
upon | +0.155 |
Neg logit contributions | |
cream | -0.059 |
Fin | -0.051 |
Blue | -0.050 |
oc | -0.044 |
Jerry | -0.043 |
Token | day |
Token ID | 1110 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
cream | +0.189 |
Fin | +0.172 |
Blue | +0.170 |
Jerry | +0.158 |
Red | +0.156 |
Neg logit contributions | |
otion | -0.231 |
upon | -0.208 |
laptop | -0.206 |
sister | -0.199 |
Deal | -0.198 |
Token | long |
Token ID | 890 |
Feature activation | -8.85e-01 |
Pos logit contributions | |
otion | +0.082 |
upon | +0.073 |
laptop | +0.073 |
Deal | +0.071 |
sister | +0.070 |
Neg logit contributions | |
cream | -0.071 |
Fin | -0.065 |
Blue | -0.065 |
Jerry | -0.061 |
Red | -0.060 |
Token | . |
Token ID | 13 |
Feature activation | -3.27e-01 |
Pos logit contributions | |
otion | +0.068 |
Deal | +0.064 |
upon | +0.061 |
laptop | +0.061 |
headache | +0.060 |
Neg logit contributions | |
cream | -0.037 |
Blue | -0.033 |
Fin | -0.032 |
sav | -0.031 |
aw | -0.030 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.89e+00 |
Pos logit contributions | |
otion | +0.023 |
upon | +0.021 |
Deal | +0.020 |
sister | +0.020 |
Lee | +0.020 |
Neg logit contributions | |
cream | -0.016 |
Fin | -0.014 |
Blue | -0.014 |
Red | -0.014 |
oc | -0.013 |
Token | Once |
Token ID | 7454 |
Feature activation | +6.69e-01 |
Pos logit contributions | |
otion | +0.235 |
upon | +0.216 |
laptop | +0.215 |
Deal | +0.209 |
sister | +0.208 |
Neg logit contributions | |
cream | -0.113 |
Fin | -0.098 |
Blue | -0.097 |
Jerry | -0.087 |
Red | -0.085 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
The | +0.048 |
that | +0.048 |
Fin | +0.047 |
He | +0.047 |
One | +0.047 |
Neg logit contributions | |
laptop | -0.025 |
Deal | -0.022 |
spine | -0.022 |
otion | -0.022 |
scalp | -0.021 |
Token | a |
Token ID | 257 |
Feature activation | -1.90e+00 |
Pos logit contributions | |
otion | +1.264 |
upon | +1.140 |
laptop | +1.131 |
Deal | +1.089 |
sister | +1.088 |
Neg logit contributions | |
cream | -1.023 |
Fin | -0.928 |
Blue | -0.915 |
Jerry | -0.852 |
Red | -0.842 |
Token | time |
Token ID | 640 |
Feature activation | -6.05e+00 |
Pos logit contributions | |
otion | +0.099 |
laptop | +0.084 |
upon | +0.081 |
sister | +0.080 |
spine | +0.079 |
Neg logit contributions | |
cream | -0.128 |
Fin | -0.124 |
Blue | -0.122 |
Jerry | -0.117 |
oc | -0.115 |
Token | , |
Token ID | 11 |
Feature activation | +6.58e-02 |
Pos logit contributions | |
laptop | +0.483 |
forehead | +0.461 |
spine | +0.460 |
scalp | +0.458 |
otion | +0.453 |
Neg logit contributions | |
Fin | -0.202 |
Bob | -0.197 |
Blue | -0.185 |
Jerry | -0.178 |
oc | -0.172 |
Token | in |
Token ID | 287 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
Blue | +0.003 |
Fin | +0.003 |
cream | +0.003 |
The | +0.003 |
Tweet | +0.003 |
Neg logit contributions | |
laptop | -0.005 |
otion | -0.005 |
spine | -0.004 |
Deal | -0.004 |
headache | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | -1.88e+00 |
Pos logit contributions | |
otion | +0.098 |
laptop | +0.089 |
upon | +0.087 |
spine | +0.086 |
Deal | +0.086 |
Neg logit contributions | |
cream | -0.062 |
Blue | -0.058 |
Fin | -0.058 |
Jerry | -0.053 |
Tweet | -0.052 |
Token | happy |
Token ID | 3772 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
otion | +0.206 |
upon | +0.188 |
sister | +0.184 |
laptop | +0.183 |
Deal | +0.177 |
Neg logit contributions | |
cream | -0.197 |
Fin | -0.181 |
Blue | -0.177 |
oc | -0.171 |
Red | -0.166 |
Token | friends |
Token ID | 2460 |
Feature activation | +3.34e+00 |
Pos logit contributions | |
otion | +0.071 |
upon | +0.066 |
sister | +0.063 |
laptop | +0.063 |
headache | +0.061 |
Neg logit contributions | |
cream | -0.064 |
Fin | -0.059 |
Blue | -0.059 |
Red | -0.055 |
Jerry | -0.055 |
Token | . |
Token ID | 13 |
Feature activation | -4.35e-01 |
Pos logit contributions | |
cream | +0.208 |
Fin | +0.197 |
Blue | +0.192 |
Tweet | +0.185 |
Red | +0.182 |
Neg logit contributions | |
otion | -0.191 |
upon | -0.172 |
sister | -0.164 |
laptop | -0.162 |
Lee | -0.160 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.92e+00 |
Pos logit contributions | |
otion | +0.026 |
Deal | +0.025 |
Lee | +0.024 |
upon | +0.024 |
laptop | +0.024 |
Neg logit contributions | |
cream | -0.024 |
Fin | -0.021 |
rob | -0.020 |
sailed | -0.020 |
v | -0.019 |
Token | Once |
Token ID | 7454 |
Feature activation | +6.95e-01 |
Pos logit contributions | |
otion | +0.233 |
upon | +0.214 |
laptop | +0.212 |
Deal | +0.206 |
sister | +0.205 |
Neg logit contributions | |
cream | -0.120 |
Fin | -0.105 |
Blue | -0.103 |
Jerry | -0.093 |
Red | -0.091 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
Fin | +0.050 |
He | +0.049 |
cream | +0.048 |
The | +0.048 |
Blue | +0.048 |
Neg logit contributions | |
laptop | -0.027 |
otion | -0.026 |
Deal | -0.024 |
spine | -0.024 |
scalp | -0.023 |
Token | a |
Token ID | 257 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
otion | +1.257 |
upon | +1.137 |
laptop | +1.124 |
Deal | +1.081 |
sister | +1.081 |
Neg logit contributions | |
cream | -1.029 |
Fin | -0.932 |
Blue | -0.919 |
Jerry | -0.856 |
Red | -0.846 |
Token | time |
Token ID | 640 |
Feature activation | -6.02e+00 |
Pos logit contributions | |
otion | +0.103 |
laptop | +0.087 |
upon | +0.085 |
sister | +0.083 |
Deal | +0.082 |
Neg logit contributions | |
cream | -0.137 |
Fin | -0.131 |
Blue | -0.130 |
Jerry | -0.124 |
oc | -0.122 |
Token | , |
Token ID | 11 |
Feature activation | +5.12e-02 |
Pos logit contributions | |
laptop | +0.443 |
otion | +0.426 |
spine | +0.422 |
headache | +0.421 |
Deal | +0.419 |
Neg logit contributions | |
Fin | -0.234 |
Blue | -0.224 |
Bob | -0.218 |
Jerry | -0.209 |
Tweet | -0.205 |
Token | there |
Token ID | 612 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
Blue | +0.002 |
Fin | +0.002 |
cream | +0.002 |
The | +0.002 |
Tweet | +0.002 |
Neg logit contributions | |
laptop | -0.003 |
otion | -0.003 |
spine | -0.003 |
Deal | -0.003 |
headache | -0.003 |
Token | was |
Token ID | 373 |
Feature activation | -5.81e-01 |
Pos logit contributions | |
otion | +0.163 |
upon | +0.150 |
laptop | +0.149 |
sister | +0.145 |
Deal | +0.144 |
Neg logit contributions | |
cream | -0.074 |
Fin | -0.064 |
Blue | -0.063 |
Jerry | -0.057 |
Red | -0.056 |
Token | the |
Token ID | 262 |
Feature activation | -4.69e-01 |
Pos logit contributions | |
otion | +0.121 |
laptop | +0.110 |
sister | +0.110 |
upon | +0.108 |
Lee | +0.102 |
Neg logit contributions | |
cream | -0.140 |
Fin | -0.134 |
Blue | -0.132 |
Tweet | -0.124 |
Red | -0.123 |
Token | store |
Token ID | 3650 |
Feature activation | -9.84e-01 |
Pos logit contributions | |
otion | +0.028 |
laptop | +0.027 |
upon | +0.026 |
headache | +0.025 |
sister | +0.025 |
Neg logit contributions | |
cream | -0.029 |
Fin | -0.026 |
Blue | -0.024 |
Tweet | -0.024 |
oc | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | -3.51e-01 |
Pos logit contributions | |
otion | +0.058 |
upon | +0.054 |
laptop | +0.053 |
sister | +0.051 |
Lee | +0.050 |
Neg logit contributions | |
cream | -0.059 |
Fin | -0.055 |
Blue | -0.054 |
Tweet | -0.051 |
Red | -0.051 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.94e+00 |
Pos logit contributions | |
otion | +0.023 |
upon | +0.022 |
Deal | +0.021 |
Lee | +0.020 |
laptop | +0.020 |
Neg logit contributions | |
cream | -0.018 |
Fin | -0.016 |
Blue | -0.015 |
rob | -0.015 |
Red | -0.015 |
Token | Once |
Token ID | 7454 |
Feature activation | +7.33e-01 |
Pos logit contributions | |
otion | +0.233 |
upon | +0.217 |
laptop | +0.211 |
Deal | +0.205 |
sister | +0.204 |
Neg logit contributions | |
cream | -0.123 |
Fin | -0.106 |
Blue | -0.103 |
Jerry | -0.093 |
oc | -0.092 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
Fin | +0.053 |
He | +0.052 |
The | +0.052 |
Blue | +0.051 |
One | +0.051 |
Neg logit contributions | |
laptop | -0.028 |
otion | -0.026 |
Deal | -0.025 |
spine | -0.024 |
scalp | -0.024 |
Token | a |
Token ID | 257 |
Feature activation | -1.96e+00 |
Pos logit contributions | |
otion | +1.264 |
upon | +1.140 |
laptop | +1.132 |
Deal | +1.090 |
sister | +1.088 |
Neg logit contributions | |
cream | -1.021 |
Fin | -0.927 |
Blue | -0.914 |
Jerry | -0.851 |
Red | -0.841 |
Token | time |
Token ID | 640 |
Feature activation | -6.02e+00 |
Pos logit contributions | |
otion | +0.103 |
laptop | +0.088 |
upon | +0.085 |
sister | +0.083 |
spine | +0.082 |
Neg logit contributions | |
cream | -0.131 |
Fin | -0.126 |
Blue | -0.125 |
Jerry | -0.120 |
oc | -0.118 |
Token | , |
Token ID | 11 |
Feature activation | +5.28e-02 |
Pos logit contributions | |
laptop | +0.469 |
otion | +0.445 |
Deal | +0.444 |
spine | +0.443 |
forehead | +0.443 |
Neg logit contributions | |
Fin | -0.211 |
Bob | -0.203 |
Blue | -0.198 |
Jerry | -0.188 |
One | -0.183 |
Token | in |
Token ID | 287 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
Blue | +0.002 |
Fin | +0.002 |
cream | +0.002 |
The | +0.002 |
Tweet | +0.002 |
Neg logit contributions | |
laptop | -0.004 |
otion | -0.004 |
spine | -0.004 |
headache | -0.004 |
Deal | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | -1.98e+00 |
Pos logit contributions | |
otion | +0.091 |
upon | +0.082 |
laptop | +0.082 |
Deal | +0.079 |
sister | +0.079 |
Neg logit contributions | |
cream | -0.072 |
Fin | -0.065 |
Blue | -0.064 |
Jerry | -0.060 |
Red | -0.059 |
Token | good |
Token ID | 922 |
Feature activation | +2.14e+00 |
Pos logit contributions | |
cream | +0.231 |
Blue | +0.202 |
Fin | +0.201 |
oc | +0.192 |
Red | +0.189 |
Neg logit contributions | |
otion | -0.229 |
upon | -0.211 |
sister | -0.207 |
laptop | -0.201 |
Deal | -0.197 |
Token | friends |
Token ID | 2460 |
Feature activation | +3.34e+00 |
Pos logit contributions | |
cream | +0.107 |
Fin | +0.096 |
Blue | +0.093 |
Jerry | +0.088 |
aw | +0.087 |
Neg logit contributions | |
otion | -0.152 |
laptop | -0.137 |
upon | -0.136 |
Deal | -0.135 |
spine | -0.132 |
Token | . |
Token ID | 13 |
Feature activation | -2.87e-01 |
Pos logit contributions | |
cream | +0.215 |
Fin | +0.198 |
Blue | +0.192 |
Tweet | +0.183 |
Red | +0.182 |
Neg logit contributions | |
otion | -0.188 |
upon | -0.168 |
sister | -0.161 |
laptop | -0.159 |
Lee | -0.158 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.94e+00 |
Pos logit contributions | |
otion | +0.017 |
upon | +0.016 |
Lee | +0.015 |
sister | +0.015 |
Deal | +0.015 |
Neg logit contributions | |
cream | -0.017 |
Fin | -0.014 |
Red | -0.014 |
Blue | -0.014 |
rob | -0.014 |
Token | Once |
Token ID | 7454 |
Feature activation | +6.53e-01 |
Pos logit contributions | |
otion | +0.239 |
upon | +0.220 |
laptop | +0.217 |
Deal | +0.211 |
sister | +0.210 |
Neg logit contributions | |
cream | -0.116 |
Fin | -0.100 |
Blue | -0.098 |
Jerry | -0.088 |
Red | -0.087 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
Fin | +0.043 |
The | +0.043 |
He | +0.042 |
Blue | +0.042 |
Bob | +0.042 |
Neg logit contributions | |
laptop | -0.028 |
Deal | -0.026 |
otion | -0.026 |
spine | -0.025 |
scalp | -0.025 |
Token | a |
Token ID | 257 |
Feature activation | -1.94e+00 |
Pos logit contributions | |
otion | +1.263 |
upon | +1.139 |
laptop | +1.130 |
Deal | +1.088 |
sister | +1.087 |
Neg logit contributions | |
cream | -1.023 |
Fin | -0.928 |
Blue | -0.915 |
Jerry | -0.852 |
Red | -0.842 |
Token | time |
Token ID | 640 |
Feature activation | -6.00e+00 |
Pos logit contributions | |
otion | +0.102 |
laptop | +0.086 |
upon | +0.083 |
sister | +0.082 |
spine | +0.081 |
Neg logit contributions | |
cream | -0.130 |
Fin | -0.125 |
Blue | -0.124 |
Jerry | -0.118 |
oc | -0.117 |
Token | , |
Token ID | 11 |
Feature activation | +7.10e-02 |
Pos logit contributions | |
laptop | +0.464 |
spine | +0.441 |
otion | +0.438 |
Deal | +0.438 |
headache | +0.436 |
Neg logit contributions | |
Fin | -0.215 |
Bob | -0.204 |
Blue | -0.198 |
Jerry | -0.185 |
oc | -0.185 |
Token | there |
Token ID | 612 |
Feature activation | -1.95e+00 |
Pos logit contributions | |
Blue | +0.003 |
Fin | +0.003 |
cream | +0.003 |
The | +0.003 |
He | +0.003 |
Neg logit contributions | |
laptop | -0.005 |
otion | -0.005 |
spine | -0.005 |
headache | -0.005 |
Deal | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | -5.65e-01 |
Pos logit contributions | |
otion | +0.167 |
laptop | +0.155 |
upon | +0.153 |
Deal | +0.152 |
spine | +0.151 |
Neg logit contributions | |
cream | -0.066 |
Fin | -0.057 |
Blue | -0.056 |
Jerry | -0.049 |
oc | -0.049 |
Token | the |
Token ID | 262 |
Feature activation | -6.19e-01 |
Pos logit contributions | |
otion | +0.130 |
sister | +0.115 |
laptop | +0.115 |
upon | +0.115 |
headache | +0.109 |
Neg logit contributions | |
cream | -0.138 |
Fin | -0.130 |
Blue | -0.129 |
Tweet | -0.121 |
Jerry | -0.121 |
Token | theater |
Token ID | 13766 |
Feature activation | -6.61e+00 |
Pos logit contributions | |
otion | +0.034 |
laptop | +0.032 |
upon | +0.031 |
headache | +0.031 |
curtain | +0.030 |
Neg logit contributions | |
cream | -0.041 |
Fin | -0.037 |
Blue | -0.036 |
Jerry | -0.035 |
Tweet | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | -3.79e-01 |
Pos logit contributions | |
otion | +0.415 |
laptop | +0.378 |
upon | +0.376 |
sister | +0.360 |
spine | +0.356 |
Neg logit contributions | |
cream | -0.375 |
Fin | -0.351 |
Blue | -0.346 |
Jerry | -0.324 |
Tweet | -0.323 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.98e+00 |
Pos logit contributions | |
otion | +0.027 |
upon | +0.025 |
sister | +0.024 |
Deal | +0.024 |
laptop | +0.023 |
Neg logit contributions | |
cream | -0.018 |
Fin | -0.017 |
Blue | -0.016 |
Jerry | -0.015 |
Tweet | -0.015 |
Token | Once |
Token ID | 7454 |
Feature activation | +6.92e-01 |
Pos logit contributions | |
otion | +0.246 |
upon | +0.226 |
laptop | +0.225 |
Deal | +0.219 |
sister | +0.218 |
Neg logit contributions | |
cream | -0.113 |
Fin | -0.098 |
Blue | -0.096 |
Jerry | -0.086 |
Red | -0.085 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
The | +0.051 |
He | +0.051 |
that | +0.051 |
Fin | +0.051 |
One | +0.050 |
Neg logit contributions | |
laptop | -0.024 |
Deal | -0.022 |
otion | -0.021 |
scalp | -0.021 |
spine | -0.020 |
Token | a |
Token ID | 257 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
otion | +1.270 |
upon | +1.143 |
laptop | +1.137 |
Deal | +1.095 |
sister | +1.093 |
Neg logit contributions | |
cream | -1.016 |
Fin | -0.922 |
Blue | -0.911 |
Jerry | -0.847 |
Red | -0.837 |
Token | time |
Token ID | 640 |
Feature activation | -6.06e+00 |
Pos logit contributions | |
otion | +0.113 |
laptop | +0.096 |
ancer | +0.092 |
sister | +0.091 |
upon | +0.091 |
Neg logit contributions | |
cream | -0.125 |
Fin | -0.123 |
Blue | -0.122 |
Jerry | -0.117 |
oc | -0.115 |
Token | , |
Token ID | 11 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
laptop | +0.484 |
forehead | +0.458 |
scalp | +0.457 |
spine | +0.455 |
Deal | +0.453 |
Neg logit contributions | |
Fin | -0.202 |
Bob | -0.194 |
Blue | -0.182 |
Jerry | -0.177 |
oc | -0.170 |
Token | in |
Token ID | 287 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
Blue | +0.001 |
Fin | +0.001 |
cream | +0.001 |
The | +0.001 |
He | +0.001 |
Neg logit contributions | |
laptop | -0.001 |
otion | -0.001 |
spine | -0.001 |
Deal | -0.001 |
headache | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | -2.02e+00 |
Pos logit contributions | |
otion | +0.102 |
laptop | +0.092 |
upon | +0.090 |
spine | +0.089 |
Deal | +0.089 |
Neg logit contributions | |
cream | -0.066 |
Fin | -0.061 |
Blue | -0.061 |
Jerry | -0.056 |
Red | -0.055 |
Token | the |
Token ID | 262 |
Feature activation | -4.33e-01 |
Pos logit contributions | |
otion | +0.039 |
upon | +0.037 |
laptop | +0.036 |
sister | +0.035 |
Deal | +0.035 |
Neg logit contributions | |
cream | -0.030 |
Fin | -0.027 |
Blue | -0.027 |
Tweet | -0.025 |
oc | -0.025 |
Token | best |
Token ID | 1266 |
Feature activation | +5.09e+00 |
Pos logit contributions | |
otion | +0.026 |
laptop | +0.024 |
upon | +0.024 |
Deal | +0.023 |
sister | +0.023 |
Neg logit contributions | |
cream | -0.026 |
Fin | -0.024 |
Blue | -0.023 |
Jerry | -0.022 |
oc | -0.022 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
cream | +0.306 |
Fin | +0.270 |
Blue | +0.269 |
oc | +0.259 |
sav | +0.256 |
Neg logit contributions | |
otion | -0.310 |
upon | -0.271 |
laptop | -0.271 |
Deal | -0.270 |
spine | -0.264 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.88e+00 |
Pos logit contributions | |
Fin | +0.109 |
cream | +0.108 |
Blue | +0.105 |
Jerry | +0.099 |
Red | +0.098 |
Neg logit contributions | |
otion | -0.111 |
upon | -0.107 |
sister | -0.106 |
laptop | -0.101 |
headache | -0.098 |
Token | Once |
Token ID | 7454 |
Feature activation | +7.08e-01 |
Pos logit contributions | |
otion | +0.227 |
upon | +0.212 |
laptop | +0.206 |
Deal | +0.200 |
sister | +0.200 |
Neg logit contributions | |
cream | -0.120 |
Fin | -0.105 |
Blue | -0.102 |
Jerry | -0.092 |
Tweet | -0.092 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
He | +0.055 |
cream | +0.054 |
The | +0.054 |
Fin | +0.053 |
that | +0.053 |
Neg logit contributions | |
laptop | -0.022 |
otion | -0.021 |
Deal | -0.020 |
spine | -0.019 |
scalp | -0.018 |
Token | a |
Token ID | 257 |
Feature activation | -1.95e+00 |
Pos logit contributions | |
otion | +1.279 |
upon | +1.149 |
laptop | +1.147 |
Deal | +1.105 |
sister | +1.102 |
Neg logit contributions | |
cream | -1.008 |
Fin | -0.914 |
Blue | -0.904 |
Jerry | -0.840 |
Red | -0.830 |
Token | time |
Token ID | 640 |
Feature activation | -6.01e+00 |
Pos logit contributions | |
otion | +0.109 |
laptop | +0.092 |
ancer | +0.088 |
sister | +0.088 |
upon | +0.087 |
Neg logit contributions | |
cream | -0.122 |
Fin | -0.119 |
Blue | -0.118 |
Jerry | -0.114 |
oc | -0.112 |
Token | , |
Token ID | 11 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
laptop | +0.486 |
spine | +0.463 |
scalp | +0.458 |
forehead | +0.457 |
otion | +0.451 |
Neg logit contributions | |
Fin | -0.184 |
Bob | -0.182 |
Blue | -0.172 |
Jerry | -0.170 |
One | -0.167 |
Token | there |
Token ID | 612 |
Feature activation | -1.98e+00 |
Pos logit contributions | |
Blue | +0.005 |
Fin | +0.005 |
cream | +0.005 |
He | +0.005 |
The | +0.005 |
Neg logit contributions | |
laptop | -0.008 |
otion | -0.008 |
spine | -0.008 |
headache | -0.007 |
Deal | -0.007 |
Token | was |
Token ID | 373 |
Feature activation | -5.84e-01 |
Pos logit contributions | |
otion | +0.171 |
laptop | +0.159 |
upon | +0.156 |
Deal | +0.155 |
spine | +0.154 |
Neg logit contributions | |
cream | -0.066 |
Fin | -0.057 |
Blue | -0.056 |
Jerry | -0.050 |
oc | -0.049 |
Token | day |
Token ID | 1110 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
cream | +0.267 |
Fin | +0.245 |
Blue | +0.244 |
Jerry | +0.225 |
Red | +0.223 |
Neg logit contributions | |
otion | -0.355 |
upon | -0.324 |
laptop | -0.321 |
sister | -0.311 |
headache | -0.309 |
Token | ever |
Token ID | 1683 |
Feature activation | +2.35e-02 |
Pos logit contributions | |
otion | +0.098 |
laptop | +0.090 |
Deal | +0.089 |
upon | +0.089 |
headache | +0.088 |
Neg logit contributions | |
cream | -0.057 |
Blue | -0.049 |
Fin | -0.049 |
Jerry | -0.045 |
oc | -0.044 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
cream | +0.001 |
Fin | +0.001 |
Blue | +0.001 |
Red | +0.001 |
Jerry | +0.001 |
Neg logit contributions | |
otion | -0.002 |
upon | -0.001 |
laptop | -0.001 |
sister | -0.001 |
spine | -0.001 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.99e+00 |
Pos logit contributions | |
cream | +0.061 |
Blue | +0.052 |
Fin | +0.047 |
Jerry | +0.044 |
Red | +0.043 |
Neg logit contributions | |
otion | -0.166 |
laptop | -0.150 |
spine | -0.148 |
upon | -0.148 |
Deal | -0.147 |
Token | Once |
Token ID | 7454 |
Feature activation | +6.62e-01 |
Pos logit contributions | |
otion | +0.242 |
upon | +0.224 |
laptop | +0.221 |
Deal | +0.215 |
sister | +0.214 |
Neg logit contributions | |
cream | -0.119 |
Fin | -0.103 |
Blue | -0.101 |
Jerry | -0.090 |
Red | -0.089 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
He | +0.048 |
The | +0.048 |
that | +0.048 |
One | +0.048 |
Fin | +0.046 |
Neg logit contributions | |
laptop | -0.023 |
otion | -0.021 |
Deal | -0.021 |
scalp | -0.020 |
spine | -0.020 |
Token | a |
Token ID | 257 |
Feature activation | -1.99e+00 |
Pos logit contributions | |
otion | +1.297 |
laptop | +1.164 |
upon | +1.159 |
Deal | +1.123 |
spine | +1.119 |
Neg logit contributions | |
cream | -0.992 |
Fin | -0.897 |
Blue | -0.891 |
Jerry | -0.826 |
Red | -0.816 |
Token | time |
Token ID | 640 |
Feature activation | -6.03e+00 |
Pos logit contributions | |
otion | +0.116 |
laptop | +0.097 |
ancer | +0.096 |
ids | +0.094 |
sister | +0.093 |
Neg logit contributions | |
cream | -0.119 |
Fin | -0.118 |
Blue | -0.117 |
Jerry | -0.114 |
oc | -0.111 |
Token | , |
Token ID | 11 |
Feature activation | +2.61e-02 |
Pos logit contributions | |
laptop | +0.497 |
spine | +0.475 |
forehead | +0.469 |
scalp | +0.468 |
otion | +0.460 |
Neg logit contributions | |
Fin | -0.173 |
Bob | -0.172 |
Blue | -0.161 |
Jerry | -0.158 |
that | -0.158 |
Token | there |
Token ID | 612 |
Feature activation | -2.00e+00 |
Pos logit contributions | |
Blue | +0.001 |
Fin | +0.001 |
The | +0.001 |
It | +0.001 |
cream | +0.001 |
Neg logit contributions | |
laptop | -0.002 |
spine | -0.002 |
otion | -0.002 |
headache | -0.002 |
Deal | -0.002 |
Token | was |
Token ID | 373 |
Feature activation | -5.97e-01 |
Pos logit contributions | |
otion | +0.173 |
laptop | +0.161 |
upon | +0.158 |
Deal | +0.157 |
spine | +0.157 |
Neg logit contributions | |
cream | -0.066 |
Fin | -0.057 |
Blue | -0.056 |
Jerry | -0.049 |
oc | -0.049 |
Token | the |
Token ID | 262 |
Feature activation | -3.68e-01 |
Pos logit contributions | |
cream | +0.024 |
Fin | +0.021 |
Blue | +0.021 |
Jerry | +0.019 |
oc | +0.019 |
Neg logit contributions | |
otion | -0.032 |
laptop | -0.028 |
upon | -0.028 |
Deal | -0.028 |
spine | -0.027 |
Token | water |
Token ID | 1660 |
Feature activation | +7.22e-01 |
Pos logit contributions | |
otion | +0.026 |
upon | +0.024 |
laptop | +0.024 |
sister | +0.023 |
spine | +0.023 |
Neg logit contributions | |
cream | -0.018 |
Fin | -0.016 |
Blue | -0.016 |
Jerry | -0.015 |
Red | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -2.03e-01 |
Pos logit contributions | |
cream | +0.038 |
Fin | +0.034 |
Blue | +0.034 |
Jerry | +0.031 |
oc | +0.031 |
Neg logit contributions | |
otion | -0.049 |
laptop | -0.044 |
upon | -0.044 |
Deal | -0.043 |
sister | -0.042 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.87e+00 |
Pos logit contributions | |
otion | +0.012 |
Deal | +0.012 |
Lee | +0.011 |
upon | +0.011 |
headache | +0.011 |
Neg logit contributions | |
cream | -0.010 |
Fin | -0.010 |
v | -0.010 |
Blue | -0.009 |
aw | -0.009 |
Token | Once |
Token ID | 7454 |
Feature activation | +7.63e-01 |
Pos logit contributions | |
otion | +0.232 |
upon | +0.213 |
laptop | +0.211 |
Deal | +0.205 |
sister | +0.205 |
Neg logit contributions | |
cream | -0.114 |
Fin | -0.100 |
Blue | -0.098 |
Jerry | -0.088 |
Red | -0.087 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
He | +0.056 |
cream | +0.056 |
Fin | +0.055 |
The | +0.055 |
Bob | +0.055 |
Neg logit contributions | |
laptop | -0.026 |
otion | -0.025 |
Deal | -0.024 |
scalp | -0.023 |
spine | -0.023 |
Token | a |
Token ID | 257 |
Feature activation | -1.90e+00 |
Pos logit contributions | |
otion | +1.284 |
laptop | +1.152 |
upon | +1.152 |
Deal | +1.110 |
sister | +1.106 |
Neg logit contributions | |
cream | -1.004 |
Fin | -0.910 |
Blue | -0.900 |
Jerry | -0.836 |
Red | -0.826 |
Token | time |
Token ID | 640 |
Feature activation | -5.96e+00 |
Pos logit contributions | |
otion | +0.104 |
laptop | +0.088 |
upon | +0.084 |
sister | +0.084 |
ancer | +0.084 |
Neg logit contributions | |
cream | -0.121 |
Fin | -0.119 |
Blue | -0.118 |
Jerry | -0.113 |
oc | -0.111 |
Token | , |
Token ID | 11 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
laptop | +0.494 |
scalp | +0.477 |
forehead | +0.472 |
Deal | +0.468 |
spine | +0.466 |
Neg logit contributions | |
Bob | -0.181 |
Fin | -0.179 |
Blue | -0.163 |
Jerry | -0.154 |
that | -0.152 |
Token | a |
Token ID | 257 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
Blue | +0.004 |
Fin | +0.004 |
The | +0.004 |
It | +0.004 |
Tweet | +0.004 |
Neg logit contributions | |
laptop | -0.008 |
spine | -0.008 |
Deal | -0.008 |
otion | -0.008 |
headache | -0.008 |
Token | little |
Token ID | 1310 |
Feature activation | -6.43e+00 |
Pos logit contributions | |
otion | +0.110 |
laptop | +0.094 |
upon | +0.093 |
Deal | +0.091 |
spine | +0.090 |
Neg logit contributions | |
cream | -0.118 |
Fin | -0.113 |
Blue | -0.112 |
Red | -0.105 |
Jerry | -0.104 |
Token | the |
Token ID | 262 |
Feature activation | -6.10e-01 |
Pos logit contributions | |
otion | +0.361 |
laptop | +0.319 |
upon | +0.319 |
Deal | +0.316 |
sister | +0.308 |
Neg logit contributions | |
cream | -0.258 |
Blue | -0.232 |
Fin | -0.230 |
Red | -0.215 |
Jerry | -0.211 |
Token | others |
Token ID | 1854 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
otion | +0.038 |
upon | +0.034 |
laptop | +0.033 |
Deal | +0.032 |
sister | +0.032 |
Neg logit contributions | |
cream | -0.036 |
Fin | -0.033 |
Blue | -0.033 |
Red | -0.030 |
Jerry | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | -3.62e-01 |
Pos logit contributions | |
cream | +0.047 |
Blue | +0.040 |
Fin | +0.038 |
oc | +0.037 |
sold | +0.036 |
Neg logit contributions | |
otion | -0.082 |
Deal | -0.077 |
headache | -0.074 |
laptop | -0.072 |
upon | -0.072 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.94e+00 |
Pos logit contributions | |
otion | +0.027 |
upon | +0.025 |
laptop | +0.024 |
Deal | +0.024 |
sister | +0.024 |
Neg logit contributions | |
cream | -0.017 |
Fin | -0.015 |
Blue | -0.014 |
Jerry | -0.013 |
Tweet | -0.013 |
Token | Once |
Token ID | 7454 |
Feature activation | +6.76e-01 |
Pos logit contributions | |
otion | +0.236 |
upon | +0.219 |
laptop | +0.214 |
Deal | +0.208 |
sister | +0.207 |
Neg logit contributions | |
cream | -0.119 |
Fin | -0.103 |
Blue | -0.101 |
Jerry | -0.090 |
Tweet | -0.090 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
The | +0.051 |
He | +0.051 |
that | +0.050 |
One | +0.050 |
Fin | +0.050 |
Neg logit contributions | |
laptop | -0.022 |
Deal | -0.020 |
otion | -0.020 |
spine | -0.019 |
scalp | -0.019 |
Token | a |
Token ID | 257 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
otion | +1.272 |
upon | +1.145 |
laptop | +1.139 |
Deal | +1.097 |
sister | +1.095 |
Neg logit contributions | |
cream | -1.014 |
Fin | -0.920 |
Blue | -0.908 |
Jerry | -0.845 |
Red | -0.835 |
Token | time |
Token ID | 640 |
Feature activation | -6.15e+00 |
Pos logit contributions | |
otion | +0.105 |
laptop | +0.089 |
upon | +0.087 |
sister | +0.085 |
spine | +0.084 |
Neg logit contributions | |
cream | -0.134 |
Fin | -0.130 |
Blue | -0.128 |
Jerry | -0.123 |
oc | -0.121 |
Token | , |
Token ID | 11 |
Feature activation | +1.85e-02 |
Pos logit contributions | |
laptop | +0.486 |
spine | +0.469 |
forehead | +0.462 |
scalp | +0.461 |
Deal | +0.459 |
Neg logit contributions | |
Fin | -0.207 |
Bob | -0.206 |
Blue | -0.199 |
Jerry | -0.189 |
oc | -0.189 |
Token | there |
Token ID | 612 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
Blue | +0.001 |
Fin | +0.001 |
cream | +0.001 |
The | +0.001 |
Tweet | +0.001 |
Neg logit contributions | |
laptop | -0.001 |
otion | -0.001 |
spine | -0.001 |
headache | -0.001 |
Deal | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | -6.84e-01 |
Pos logit contributions | |
otion | +0.176 |
laptop | +0.164 |
upon | +0.161 |
Deal | +0.160 |
spine | +0.159 |
Neg logit contributions | |
cream | -0.068 |
Fin | -0.059 |
Blue | -0.058 |
Jerry | -0.051 |
oc | -0.050 |
Token | were |
Token ID | 547 |
Feature activation | -3.33e+00 |
Pos logit contributions | |
cream | +0.089 |
Fin | +0.079 |
Blue | +0.078 |
Jerry | +0.071 |
Red | +0.070 |
Neg logit contributions | |
otion | -0.160 |
laptop | -0.146 |
upon | -0.146 |
Deal | -0.141 |
sister | -0.141 |
Token | happy |
Token ID | 3772 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
otion | +0.224 |
laptop | +0.202 |
upon | +0.197 |
spine | +0.193 |
Deal | +0.193 |
Neg logit contributions | |
cream | -0.177 |
Fin | -0.164 |
Blue | -0.162 |
Jerry | -0.151 |
Tweet | -0.148 |
Token | . |
Token ID | 13 |
Feature activation | -4.41e-01 |
Pos logit contributions | |
otion | +0.069 |
upon | +0.064 |
sister | +0.061 |
laptop | +0.059 |
spine | +0.058 |
Neg logit contributions | |
cream | -0.063 |
Blue | -0.058 |
Fin | -0.058 |
Red | -0.055 |
Jerry | -0.054 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.97e+00 |
Pos logit contributions | |
otion | +0.024 |
upon | +0.023 |
Deal | +0.023 |
sister | +0.022 |
Lee | +0.022 |
Neg logit contributions | |
cream | -0.027 |
rob | -0.023 |
Fin | -0.022 |
Red | -0.022 |
Blue | -0.022 |
Token | Once |
Token ID | 7454 |
Feature activation | +6.82e-01 |
Pos logit contributions | |
otion | +0.239 |
upon | +0.220 |
laptop | +0.217 |
Deal | +0.211 |
sister | +0.210 |
Neg logit contributions | |
cream | -0.119 |
Fin | -0.104 |
Blue | -0.102 |
Jerry | -0.092 |
Red | -0.090 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
Fin | +0.047 |
He | +0.047 |
The | +0.046 |
cream | +0.046 |
Bob | +0.045 |
Neg logit contributions | |
laptop | -0.027 |
otion | -0.025 |
Deal | -0.025 |
scalp | -0.024 |
spine | -0.023 |
Token | a |
Token ID | 257 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
otion | +1.269 |
upon | +1.143 |
laptop | +1.137 |
Deal | +1.095 |
sister | +1.093 |
Neg logit contributions | |
cream | -1.015 |
Fin | -0.921 |
Blue | -0.909 |
Jerry | -0.846 |
Red | -0.836 |
Token | time |
Token ID | 640 |
Feature activation | -5.98e+00 |
Pos logit contributions | |
otion | +0.106 |
laptop | +0.090 |
upon | +0.086 |
sister | +0.085 |
Deal | +0.085 |
Neg logit contributions | |
cream | -0.133 |
Fin | -0.129 |
Blue | -0.127 |
Jerry | -0.122 |
oc | -0.120 |
Token | , |
Token ID | 11 |
Feature activation | +4.55e-02 |
Pos logit contributions | |
laptop | +0.458 |
Deal | +0.443 |
headache | +0.442 |
otion | +0.439 |
spine | +0.439 |
Neg logit contributions | |
Fin | -0.206 |
Bob | -0.196 |
Blue | -0.192 |
Jerry | -0.178 |
One | -0.175 |
Token | there |
Token ID | 612 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
Blue | +0.002 |
Fin | +0.002 |
cream | +0.002 |
The | +0.002 |
Tweet | +0.002 |
Neg logit contributions | |
laptop | -0.003 |
otion | -0.003 |
spine | -0.003 |
Deal | -0.003 |
headache | -0.003 |
Token | was |
Token ID | 373 |
Feature activation | -5.44e-01 |
Pos logit contributions | |
otion | +0.169 |
laptop | +0.156 |
upon | +0.154 |
Deal | +0.153 |
spine | +0.152 |
Neg logit contributions | |
cream | -0.067 |
Fin | -0.058 |
Blue | -0.057 |
Jerry | -0.050 |
oc | -0.049 |
Token | , |
Token ID | 11 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
otion | +0.104 |
laptop | +0.090 |
spine | +0.087 |
headache | +0.086 |
upon | +0.083 |
Neg logit contributions | |
cream | -0.073 |
Blue | -0.066 |
oc | -0.066 |
Fin | -0.066 |
They | -0.061 |
Token | bee |
Token ID | 20697 |
Feature activation | -4.41e+00 |
Pos logit contributions | |
cream | +0.006 |
Blue | +0.006 |
sav | +0.006 |
oc | +0.005 |
Tweet | +0.005 |
Neg logit contributions | |
otion | -0.006 |
spine | -0.005 |
laptop | -0.005 |
ancer | -0.005 |
ids | -0.005 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
otion | +0.314 |
upon | +0.284 |
laptop | +0.283 |
sister | +0.274 |
Deal | +0.273 |
Neg logit contributions | |
cream | -0.217 |
Fin | -0.195 |
Blue | -0.193 |
Jerry | -0.178 |
Red | -0.176 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.92e+00 |
Pos logit contributions | |
cream | +0.085 |
Blue | +0.073 |
Fin | +0.069 |
sav | +0.068 |
oc | +0.066 |
Neg logit contributions | |
otion | -0.139 |
laptop | -0.125 |
Deal | -0.122 |
upon | -0.122 |
spine | -0.121 |
Token | Once |
Token ID | 7454 |
Feature activation | +7.01e-01 |
Pos logit contributions | |
otion | +0.239 |
upon | +0.220 |
laptop | +0.219 |
Deal | +0.213 |
sister | +0.212 |
Neg logit contributions | |
cream | -0.112 |
Fin | -0.098 |
Blue | -0.096 |
Jerry | -0.086 |
Red | -0.085 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
The | +0.052 |
that | +0.052 |
He | +0.052 |
One | +0.051 |
cream | +0.050 |
Neg logit contributions | |
laptop | -0.023 |
otion | -0.022 |
Deal | -0.021 |
spine | -0.021 |
scalp | -0.020 |
Token | a |
Token ID | 257 |
Feature activation | -1.91e+00 |
Pos logit contributions | |
otion | +1.275 |
upon | +1.146 |
laptop | +1.143 |
Deal | +1.101 |
sister | +1.098 |
Neg logit contributions | |
cream | -1.010 |
Fin | -0.916 |
Blue | -0.906 |
Jerry | -0.842 |
Red | -0.832 |
Token | time |
Token ID | 640 |
Feature activation | -6.00e+00 |
Pos logit contributions | |
otion | +0.107 |
laptop | +0.091 |
ancer | +0.088 |
sister | +0.086 |
upon | +0.086 |
Neg logit contributions | |
cream | -0.118 |
Fin | -0.117 |
Blue | -0.115 |
Jerry | -0.112 |
oc | -0.109 |
Token | , |
Token ID | 11 |
Feature activation | +7.55e-02 |
Pos logit contributions | |
laptop | +0.490 |
spine | +0.473 |
scalp | +0.459 |
forehead | +0.457 |
Deal | +0.456 |
Neg logit contributions | |
Fin | -0.183 |
Bob | -0.178 |
Blue | -0.171 |
One | -0.164 |
oc | -0.163 |
Token | there |
Token ID | 612 |
Feature activation | -1.91e+00 |
Pos logit contributions | |
Blue | +0.004 |
Fin | +0.003 |
cream | +0.003 |
The | +0.003 |
Tweet | +0.003 |
Neg logit contributions | |
laptop | -0.005 |
otion | -0.005 |
spine | -0.005 |
Deal | -0.005 |
headache | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | -5.71e-01 |
Pos logit contributions | |
otion | +0.162 |
laptop | +0.149 |
upon | +0.148 |
Deal | +0.145 |
spine | +0.144 |
Neg logit contributions | |
cream | -0.068 |
Fin | -0.059 |
Blue | -0.058 |
Jerry | -0.051 |
Red | -0.050 |
Token | anxious |
Token ID | 18116 |
Feature activation | -5.47e+00 |
Pos logit contributions | |
otion | +0.087 |
laptop | +0.077 |
Deal | +0.076 |
upon | +0.074 |
spine | +0.072 |
Neg logit contributions | |
cream | -0.071 |
Fin | -0.064 |
Blue | -0.063 |
Jerry | -0.060 |
sold | -0.059 |
Token | anymore |
Token ID | 7471 |
Feature activation | +1.00e-01 |
Pos logit contributions | |
otion | +0.358 |
laptop | +0.347 |
Deal | +0.345 |
kin | +0.317 |
ids | +0.316 |
Neg logit contributions | |
cream | -0.272 |
Fin | -0.241 |
Blue | -0.237 |
He | -0.232 |
sold | -0.228 |
Token | . |
Token ID | 13 |
Feature activation | -2.09e-01 |
Pos logit contributions | |
cream | +0.005 |
Fin | +0.005 |
Blue | +0.004 |
oc | +0.004 |
Jerry | +0.004 |
Neg logit contributions | |
otion | -0.007 |
laptop | -0.006 |
upon | -0.006 |
Deal | -0.006 |
spine | -0.006 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
otion | +0.012 |
upon | +0.012 |
Deal | +0.011 |
Lee | +0.011 |
sister | +0.011 |
Neg logit contributions | |
cream | -0.011 |
rob | -0.010 |
Fin | -0.010 |
Blue | -0.010 |
Red | -0.010 |
Token | Once |
Token ID | 7454 |
Feature activation | +7.37e-01 |
Pos logit contributions | |
otion | +0.236 |
laptop | +0.217 |
upon | +0.216 |
Deal | +0.210 |
spine | +0.210 |
Neg logit contributions | |
cream | -0.102 |
Fin | -0.088 |
Blue | -0.087 |
Jerry | -0.078 |
Red | -0.076 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
Fin | +0.056 |
He | +0.056 |
The | +0.054 |
cream | +0.054 |
Bob | +0.054 |
Neg logit contributions | |
laptop | -0.025 |
otion | -0.022 |
Deal | -0.022 |
scalp | -0.022 |
spine | -0.021 |
Token | a |
Token ID | 257 |
Feature activation | -1.90e+00 |
Pos logit contributions | |
otion | +1.291 |
laptop | +1.159 |
upon | +1.155 |
Deal | +1.117 |
sister | +1.112 |
Neg logit contributions | |
cream | -0.997 |
Fin | -0.903 |
Blue | -0.895 |
Jerry | -0.830 |
Red | -0.820 |
Token | time |
Token ID | 640 |
Feature activation | -5.99e+00 |
Pos logit contributions | |
otion | +0.100 |
laptop | +0.084 |
upon | +0.082 |
sister | +0.080 |
spine | +0.079 |
Neg logit contributions | |
cream | -0.128 |
Fin | -0.123 |
Blue | -0.122 |
Jerry | -0.117 |
oc | -0.115 |
Token | , |
Token ID | 11 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
laptop | +0.488 |
forehead | +0.470 |
spine | +0.465 |
Deal | +0.462 |
scalp | +0.460 |
Neg logit contributions | |
Fin | -0.191 |
Bob | -0.186 |
Blue | -0.168 |
Jerry | -0.162 |
One | -0.158 |
Token | in |
Token ID | 287 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
Fin | +0.006 |
Blue | +0.006 |
cream | +0.006 |
Tweet | +0.006 |
The | +0.006 |
Neg logit contributions | |
laptop | -0.009 |
otion | -0.009 |
spine | -0.009 |
Deal | -0.009 |
headache | -0.009 |
Token | a |
Token ID | 257 |
Feature activation | -1.91e+00 |
Pos logit contributions | |
otion | +0.085 |
upon | +0.077 |
laptop | +0.076 |
Deal | +0.073 |
sister | +0.073 |
Neg logit contributions | |
cream | -0.072 |
Fin | -0.065 |
Blue | -0.064 |
Jerry | -0.059 |
Red | -0.059 |
Token | ay |
Token ID | 323 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
otion | +0.290 |
upon | +0.257 |
laptop | +0.254 |
Deal | +0.243 |
sister | +0.242 |
Neg logit contributions | |
cream | -0.296 |
Fin | -0.278 |
Blue | -0.272 |
Jerry | -0.256 |
Red | -0.253 |
Token | on |
Token ID | 261 |
Feature activation | -6.86e-01 |
Pos logit contributions | |
otion | +0.082 |
upon | +0.074 |
laptop | +0.074 |
Deal | +0.072 |
spine | +0.072 |
Neg logit contributions | |
cream | -0.061 |
Fin | -0.054 |
Blue | -0.054 |
Jerry | -0.050 |
oc | -0.050 |
Token | . |
Token ID | 13 |
Feature activation | -2.95e-01 |
Pos logit contributions | |
otion | +0.048 |
upon | +0.043 |
laptop | +0.043 |
Deal | +0.042 |
headache | +0.041 |
Neg logit contributions | |
cream | -0.035 |
Blue | -0.031 |
Fin | -0.031 |
oc | -0.029 |
Jerry | -0.029 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.79e+00 |
Pos logit contributions | |
otion | +0.019 |
upon | +0.018 |
sister | +0.017 |
Lee | +0.017 |
Deal | +0.016 |
Neg logit contributions | |
cream | -0.015 |
Fin | -0.014 |
Blue | -0.014 |
Red | -0.013 |
rob | -0.013 |
Token | Once |
Token ID | 7454 |
Feature activation | +7.22e-01 |
Pos logit contributions | |
otion | +0.221 |
upon | +0.205 |
laptop | +0.199 |
Deal | +0.193 |
sister | +0.193 |
Neg logit contributions | |
cream | -0.117 |
Fin | -0.101 |
Blue | -0.099 |
Jerry | -0.089 |
Red | -0.089 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
Fin | +0.055 |
cream | +0.054 |
Blue | +0.054 |
The | +0.054 |
He | +0.054 |
Neg logit contributions | |
laptop | -0.024 |
otion | -0.022 |
Deal | -0.022 |
spine | -0.020 |
scalp | -0.020 |
Token | a |
Token ID | 257 |
Feature activation | -1.91e+00 |
Pos logit contributions | |
otion | +1.258 |
upon | +1.137 |
laptop | +1.125 |
Deal | +1.083 |
sister | +1.082 |
Neg logit contributions | |
cream | -1.026 |
Fin | -0.931 |
Blue | -0.918 |
Jerry | -0.855 |
Red | -0.845 |
Token | time |
Token ID | 640 |
Feature activation | -6.03e+00 |
Pos logit contributions | |
otion | +0.100 |
laptop | +0.085 |
upon | +0.082 |
sister | +0.081 |
spine | +0.079 |
Neg logit contributions | |
cream | -0.127 |
Fin | -0.123 |
Blue | -0.122 |
Jerry | -0.117 |
oc | -0.115 |
Token | , |
Token ID | 11 |
Feature activation | +6.03e-02 |
Pos logit contributions | |
laptop | +0.463 |
otion | +0.440 |
spine | +0.438 |
Deal | +0.436 |
headache | +0.435 |
Neg logit contributions | |
Fin | -0.223 |
Bob | -0.211 |
Blue | -0.208 |
oc | -0.199 |
Jerry | -0.198 |
Token | there |
Token ID | 612 |
Feature activation | -1.95e+00 |
Pos logit contributions | |
Blue | +0.003 |
cream | +0.003 |
Fin | +0.003 |
Tweet | +0.003 |
He | +0.003 |
Neg logit contributions | |
laptop | -0.004 |
otion | -0.004 |
spine | -0.004 |
headache | -0.004 |
Deal | -0.004 |
Token | was |
Token ID | 373 |
Feature activation | -5.39e-01 |
Pos logit contributions | |
otion | +0.164 |
upon | +0.150 |
laptop | +0.150 |
Deal | +0.146 |
sister | +0.145 |
Neg logit contributions | |
cream | -0.071 |
Fin | -0.061 |
Blue | -0.060 |
Jerry | -0.054 |
Red | -0.053 |
Token | healthy |
Token ID | 5448 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
otion | +0.092 |
upon | +0.085 |
sister | +0.085 |
laptop | +0.083 |
headache | +0.082 |
Neg logit contributions | |
cream | -0.069 |
Fin | -0.065 |
Blue | -0.064 |
Jerry | -0.060 |
Red | -0.059 |
Token | again |
Token ID | 757 |
Feature activation | +3.15e+00 |
Pos logit contributions | |
cream | +0.109 |
Fin | +0.098 |
Blue | +0.097 |
Jerry | +0.090 |
oc | +0.089 |
Neg logit contributions | |
otion | -0.143 |
upon | -0.128 |
laptop | -0.128 |
Deal | -0.124 |
spine | -0.122 |
Token | . |
Token ID | 13 |
Feature activation | -1.72e-01 |
Pos logit contributions | |
cream | +0.171 |
Fin | +0.155 |
Blue | +0.153 |
Jerry | +0.143 |
Red | +0.141 |
Neg logit contributions | |
otion | -0.209 |
upon | -0.188 |
laptop | -0.187 |
Deal | -0.180 |
sister | -0.180 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.75e+00 |
Pos logit contributions | |
otion | +0.013 |
upon | +0.012 |
Deal | +0.011 |
sister | +0.011 |
Lee | +0.011 |
Neg logit contributions | |
cream | -0.008 |
Fin | -0.007 |
Blue | -0.007 |
Red | -0.007 |
Tweet | -0.006 |
Token | Once |
Token ID | 7454 |
Feature activation | +7.14e-01 |
Pos logit contributions | |
otion | +0.226 |
upon | +0.207 |
laptop | +0.206 |
Deal | +0.200 |
sister | +0.200 |
Neg logit contributions | |
cream | -0.106 |
Fin | -0.092 |
Blue | -0.091 |
Jerry | -0.081 |
Red | -0.080 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
He | +0.054 |
The | +0.053 |
that | +0.053 |
Fin | +0.053 |
One | +0.052 |
Neg logit contributions | |
laptop | -0.025 |
Deal | -0.022 |
otion | -0.021 |
scalp | -0.021 |
spine | -0.021 |
Token | a |
Token ID | 257 |
Feature activation | -1.91e+00 |
Pos logit contributions | |
otion | +1.293 |
laptop | +1.162 |
upon | +1.157 |
Deal | +1.120 |
spine | +1.115 |
Neg logit contributions | |
cream | -0.994 |
Fin | -0.900 |
Blue | -0.893 |
Jerry | -0.827 |
Red | -0.818 |
Token | time |
Token ID | 640 |
Feature activation | -6.05e+00 |
Pos logit contributions | |
otion | +0.102 |
laptop | +0.086 |
upon | +0.083 |
sister | +0.082 |
spine | +0.081 |
Neg logit contributions | |
cream | -0.126 |
Fin | -0.122 |
Blue | -0.121 |
Jerry | -0.115 |
oc | -0.114 |
Token | , |
Token ID | 11 |
Feature activation | +8.17e-02 |
Pos logit contributions | |
laptop | +0.490 |
forehead | +0.462 |
scalp | +0.462 |
spine | +0.460 |
Deal | +0.458 |
Neg logit contributions | |
Fin | -0.203 |
Bob | -0.201 |
Blue | -0.188 |
Jerry | -0.181 |
oc | -0.171 |
Token | there |
Token ID | 612 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
Blue | +0.003 |
Fin | +0.003 |
The | +0.003 |
He | +0.003 |
cream | +0.003 |
Neg logit contributions | |
laptop | -0.006 |
spine | -0.006 |
Deal | -0.006 |
otion | -0.006 |
headache | -0.006 |
Token | was |
Token ID | 373 |
Feature activation | -6.12e-01 |
Pos logit contributions | |
otion | +0.175 |
laptop | +0.166 |
Deal | +0.163 |
spine | +0.162 |
headache | +0.159 |
Neg logit contributions | |
cream | -0.058 |
Fin | -0.050 |
Blue | -0.048 |
oc | -0.043 |
Jerry | -0.042 |
Token | y |
Token ID | 331 |
Feature activation | +4.00e-01 |
Pos logit contributions | |
cream | +0.007 |
Fin | +0.006 |
Blue | +0.006 |
Jerry | +0.005 |
Red | +0.005 |
Neg logit contributions | |
otion | -0.010 |
laptop | -0.009 |
upon | -0.009 |
Deal | -0.009 |
spine | -0.009 |
Token | ummy |
Token ID | 13513 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
cream | +0.031 |
Blue | +0.029 |
Fin | +0.029 |
Tweet | +0.028 |
Jerry | +0.027 |
Neg logit contributions | |
otion | -0.017 |
upon | -0.015 |
laptop | -0.014 |
Deal | -0.014 |
spine | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | -2.87e-01 |
Pos logit contributions | |
cream | +0.121 |
Fin | +0.113 |
Blue | +0.112 |
Jerry | +0.104 |
Tweet | +0.104 |
Neg logit contributions | |
otion | -0.135 |
upon | -0.122 |
laptop | -0.122 |
sister | -0.118 |
spine | -0.116 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.87e+00 |
Pos logit contributions | |
otion | +0.021 |
upon | +0.019 |
laptop | +0.018 |
Deal | +0.018 |
sister | +0.018 |
Neg logit contributions | |
cream | -0.014 |
Fin | -0.013 |
Blue | -0.012 |
oc | -0.011 |
Jerry | -0.011 |
Token | Once |
Token ID | 7454 |
Feature activation | +6.80e-01 |
Pos logit contributions | |
otion | +0.238 |
upon | +0.219 |
laptop | +0.218 |
Deal | +0.212 |
sister | +0.212 |
Neg logit contributions | |
cream | -0.108 |
Fin | -0.093 |
Blue | -0.092 |
Jerry | -0.082 |
Red | -0.081 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.90e+01 |
Pos logit contributions | |
He | +0.049 |
that | +0.049 |
The | +0.049 |
One | +0.048 |
Fin | +0.048 |
Neg logit contributions | |
laptop | -0.024 |
Deal | -0.022 |
otion | -0.022 |
scalp | -0.021 |
spine | -0.021 |
Token | a |
Token ID | 257 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
otion | +1.311 |
laptop | +1.179 |
upon | +1.166 |
Deal | +1.138 |
spine | +1.137 |
Neg logit contributions | |
cream | -0.978 |
Fin | -0.884 |
Blue | -0.880 |
Jerry | -0.813 |
Red | -0.804 |
Token | time |
Token ID | 640 |
Feature activation | -6.03e+00 |
Pos logit contributions | |
otion | +0.104 |
laptop | +0.088 |
upon | +0.084 |
ancer | +0.084 |
sister | +0.084 |
Neg logit contributions | |
cream | -0.120 |
Fin | -0.118 |
Blue | -0.116 |
Jerry | -0.112 |
oc | -0.110 |
Token | , |
Token ID | 11 |
Feature activation | +6.73e-02 |
Pos logit contributions | |
laptop | +0.474 |
forehead | +0.450 |
otion | +0.450 |
spine | +0.450 |
scalp | +0.444 |
Neg logit contributions | |
Fin | -0.201 |
Bob | -0.198 |
Blue | -0.188 |
oc | -0.184 |
Jerry | -0.181 |
Token | there |
Token ID | 612 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
Blue | +0.003 |
Fin | +0.003 |
The | +0.003 |
cream | +0.003 |
He | +0.003 |
Neg logit contributions | |
laptop | -0.005 |
otion | -0.005 |
spine | -0.005 |
headache | -0.005 |
Deal | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | -5.35e-01 |
Pos logit contributions | |
otion | +0.170 |
laptop | +0.158 |
upon | +0.155 |
Deal | +0.154 |
spine | +0.154 |
Neg logit contributions | |
cream | -0.066 |
Fin | -0.056 |
Blue | -0.055 |
Jerry | -0.049 |
oc | -0.048 |
Token | little |
Token ID | 1310 |
Feature activation | -5.74e+00 |
Pos logit contributions | |
cream | +0.048 |
Fin | +0.044 |
oc | +0.043 |
sav | +0.042 |
Blue | +0.042 |
Neg logit contributions | |
otion | -0.077 |
Deal | -0.072 |
upon | -0.067 |
laptop | -0.066 |
ids | -0.065 |
Token | boy |
Token ID | 2933 |
Feature activation | +2.02e+00 |
Pos logit contributions | |
otion | +0.368 |
upon | +0.323 |
laptop | +0.323 |
Deal | +0.317 |
headache | +0.307 |
Neg logit contributions | |
cream | -0.326 |
Fin | -0.298 |
Blue | -0.296 |
Red | -0.274 |
oc | -0.273 |
Token | . |
Token ID | 13 |
Feature activation | +4.43e-01 |
Pos logit contributions | |
cream | +0.116 |
Fin | +0.109 |
Blue | +0.107 |
Jerry | +0.102 |
Red | +0.099 |
Neg logit contributions | |
otion | -0.124 |
upon | -0.113 |
laptop | -0.112 |
sister | -0.110 |
spine | -0.106 |
Token | He |
Token ID | 679 |
Feature activation | +4.57e+00 |
Pos logit contributions | |
cream | +0.017 |
Fin | +0.015 |
Blue | +0.015 |
Tweet | +0.013 |
oc | +0.013 |
Neg logit contributions | |
otion | -0.036 |
upon | -0.034 |
laptop | -0.033 |
Deal | -0.033 |
sister | -0.033 |
Token | was |
Token ID | 373 |
Feature activation | +3.12e-02 |
Pos logit contributions | |
cream | +0.197 |
Fin | +0.175 |
Blue | +0.171 |
Jerry | +0.158 |
Red | +0.151 |
Neg logit contributions | |
otion | -0.344 |
upon | -0.330 |
laptop | -0.320 |
sister | -0.315 |
headache | -0.309 |
Token | always |
Token ID | 1464 |
Feature activation | +9.30e+00 |
Pos logit contributions | |
cream | +0.002 |
Fin | +0.002 |
Blue | +0.002 |
Jerry | +0.001 |
oc | +0.001 |
Neg logit contributions | |
otion | -0.002 |
upon | -0.002 |
laptop | -0.002 |
sister | -0.002 |
Deal | -0.002 |
Token | looking |
Token ID | 2045 |
Feature activation | +3.54e-01 |
Pos logit contributions | |
cream | +0.549 |
Fin | +0.497 |
Blue | +0.493 |
Jerry | +0.473 |
sav | +0.469 |
Neg logit contributions | |
otion | -0.568 |
laptop | -0.489 |
Deal | -0.481 |
headache | -0.470 |
upon | -0.469 |
Token | around |
Token ID | 1088 |
Feature activation | +3.68e+00 |
Pos logit contributions | |
cream | +0.019 |
aw | +0.016 |
oc | +0.016 |
He | +0.016 |
Fin | +0.016 |
Neg logit contributions | |
otion | -0.024 |
kin | -0.022 |
laptop | -0.021 |
Deal | -0.020 |
ids | -0.020 |
Token | and |
Token ID | 290 |
Feature activation | -7.51e-01 |
Pos logit contributions | |
cream | +0.124 |
He | +0.123 |
sav | +0.118 |
The | +0.115 |
This | +0.113 |
Neg logit contributions | |
otion | -0.306 |
ids | -0.265 |
kin | -0.263 |
laptop | -0.259 |
Deal | -0.258 |
Token | noticed |
Token ID | 6810 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
otion | +0.044 |
upon | +0.040 |
laptop | +0.039 |
sister | +0.037 |
spine | +0.037 |
Neg logit contributions | |
cream | -0.047 |
Fin | -0.043 |
Blue | -0.043 |
Jerry | -0.040 |
Red | -0.040 |
Token | everything |
Token ID | 2279 |
Feature activation | +2.87e+00 |
Pos logit contributions | |
cream | +0.095 |
Fin | +0.090 |
Blue | +0.089 |
Jerry | +0.082 |
oc | +0.082 |
Neg logit contributions | |
otion | -0.124 |
upon | -0.114 |
laptop | -0.111 |
sister | -0.108 |
headache | -0.106 |
Token | had |
Token ID | 550 |
Feature activation | +8.59e-02 |
Pos logit contributions | |
otion | +0.242 |
upon | +0.220 |
laptop | +0.220 |
sister | +0.213 |
Deal | +0.213 |
Neg logit contributions | |
cream | -0.140 |
Fin | -0.124 |
Blue | -0.122 |
Jerry | -0.112 |
Red | -0.110 |
Token | her |
Token ID | 607 |
Feature activation | -4.36e+00 |
Pos logit contributions | |
cream | +0.004 |
Fin | +0.004 |
Blue | +0.004 |
oc | +0.004 |
Red | +0.004 |
Neg logit contributions | |
otion | -0.006 |
upon | -0.005 |
laptop | -0.005 |
sister | -0.005 |
headache | -0.005 |
Token | own |
Token ID | 898 |
Feature activation | +6.69e-01 |
Pos logit contributions | |
otion | +0.309 |
upon | +0.251 |
Deal | +0.250 |
ancer | +0.247 |
spine | +0.245 |
Neg logit contributions | |
cream | -0.231 |
Blue | -0.203 |
Fin | -0.201 |
sav | -0.194 |
Jerry | -0.189 |
Token | stick |
Token ID | 4859 |
Feature activation | +2.88e+00 |
Pos logit contributions | |
cream | +0.035 |
Blue | +0.030 |
Fin | +0.030 |
sav | +0.028 |
sold | +0.028 |
Neg logit contributions | |
otion | -0.048 |
Deal | -0.041 |
ancer | -0.039 |
upon | -0.039 |
spine | -0.039 |
Token | , |
Token ID | 11 |
Feature activation | +5.10e-01 |
Pos logit contributions | |
cream | +0.148 |
Fin | +0.132 |
Blue | +0.131 |
Jerry | +0.121 |
Red | +0.120 |
Neg logit contributions | |
otion | -0.201 |
laptop | -0.178 |
upon | -0.177 |
Deal | -0.177 |
headache | -0.172 |
Token | which |
Token ID | 543 |
Feature activation | +5.61e+00 |
Pos logit contributions | |
cream | +0.027 |
Blue | +0.024 |
Fin | +0.024 |
Jerry | +0.022 |
sav | +0.022 |
Neg logit contributions | |
otion | -0.035 |
Deal | -0.031 |
laptop | -0.031 |
upon | -0.030 |
spine | -0.030 |
Token | was |
Token ID | 373 |
Feature activation | -2.16e-01 |
Pos logit contributions | |
cream | +0.231 |
Fin | +0.203 |
Blue | +0.200 |
Jerry | +0.181 |
Red | +0.178 |
Neg logit contributions | |
otion | -0.444 |
upon | -0.406 |
laptop | -0.405 |
sister | -0.392 |
Deal | -0.392 |
Token | very |
Token ID | 845 |
Feature activation | +6.65e-01 |
Pos logit contributions | |
otion | +0.015 |
upon | +0.013 |
laptop | +0.013 |
sister | +0.013 |
headache | +0.013 |
Neg logit contributions | |
cream | -0.011 |
Fin | -0.010 |
Blue | -0.010 |
oc | -0.009 |
Jerry | -0.009 |
Token | hard |
Token ID | 1327 |
Feature activation | -3.17e-01 |
Pos logit contributions | |
cream | +0.034 |
Fin | +0.030 |
Blue | +0.030 |
oc | +0.029 |
rob | +0.028 |
Neg logit contributions | |
otion | -0.045 |
upon | -0.044 |
laptop | -0.041 |
sister | -0.041 |
headache | -0.040 |
Token | . |
Token ID | 13 |
Feature activation | +1.84e-01 |
Pos logit contributions | |
otion | +0.022 |
laptop | +0.020 |
upon | +0.020 |
Deal | +0.019 |
sister | +0.019 |
Neg logit contributions | |
cream | -0.016 |
Fin | -0.014 |
Blue | -0.014 |
Jerry | -0.013 |
oc | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
cream | +0.009 |
Blue | +0.008 |
Fin | +0.008 |
Jerry | +0.008 |
Red | +0.008 |
Neg logit contributions | |
otion | -0.013 |
laptop | -0.012 |
upon | -0.011 |
spine | -0.011 |
headache | -0.011 |
Token | L |
Token ID | 43 |
Feature activation | -8.29e-01 |
Pos logit contributions | |
cream | +0.083 |
rob | +0.068 |
sailed | +0.065 |
Fin | +0.064 |
aw | +0.062 |
Neg logit contributions | |
otion | -0.088 |
upon | -0.086 |
Deal | -0.084 |
laptop | -0.080 |
acy | -0.078 |
Token | ily |
Token ID | 813 |
Feature activation | -4.65e+00 |
Pos logit contributions | |
spine | +0.045 |
headache | +0.044 |
laptop | +0.043 |
otion | +0.042 |
Deal | +0.042 |
Neg logit contributions | |
cream | -0.052 |
Blue | -0.048 |
Fin | -0.047 |
Red | -0.046 |
aw | -0.046 |
Token | was |
Token ID | 373 |
Feature activation | -1.65e-01 |
Pos logit contributions | |
sister | +0.279 |
upon | +0.268 |
laptop | +0.261 |
otion | +0.259 |
brother | +0.253 |
Neg logit contributions | |
cream | -0.252 |
Fin | -0.249 |
Blue | -0.231 |
Red | -0.224 |
rob | -0.217 |
Token | so |
Token ID | 523 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
otion | +0.011 |
upon | +0.010 |
sister | +0.010 |
laptop | +0.010 |
Deal | +0.010 |
Neg logit contributions | |
cream | -0.009 |
Fin | -0.008 |
Blue | -0.008 |
oc | -0.007 |
Red | -0.007 |
Token | happy |
Token ID | 3772 |
Feature activation | -5.82e-01 |
Pos logit contributions | |
cream | +0.148 |
Fin | +0.135 |
Blue | +0.133 |
Jerry | +0.125 |
Red | +0.123 |
Neg logit contributions | |
otion | -0.162 |
upon | -0.145 |
laptop | -0.144 |
Deal | -0.139 |
sister | -0.138 |
Token | that |
Token ID | 326 |
Feature activation | +9.57e+00 |
Pos logit contributions | |
otion | +0.044 |
laptop | +0.040 |
upon | +0.040 |
Deal | +0.039 |
sister | +0.039 |
Neg logit contributions | |
cream | -0.026 |
Fin | -0.023 |
Blue | -0.023 |
Jerry | -0.021 |
oc | -0.021 |
Token | she |
Token ID | 673 |
Feature activation | -3.21e+00 |
Pos logit contributions | |
cream | +0.520 |
Blue | +0.510 |
Fin | +0.497 |
Jerry | +0.479 |
This | +0.468 |
Neg logit contributions | |
otion | -0.588 |
laptop | -0.546 |
spine | -0.540 |
Deal | -0.532 |
upon | -0.511 |
Token | gave |
Token ID | 2921 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
otion | +0.181 |
sister | +0.180 |
upon | +0.179 |
kin | +0.167 |
laptop | +0.164 |
Neg logit contributions | |
cream | -0.170 |
Fin | -0.162 |
Blue | -0.153 |
One | -0.147 |
Jerry | -0.147 |
Token | her |
Token ID | 607 |
Feature activation | -3.86e+00 |
Pos logit contributions | |
cream | +0.114 |
Fin | +0.104 |
Blue | +0.104 |
Jerry | +0.097 |
Red | +0.096 |
Neg logit contributions | |
otion | -0.110 |
laptop | -0.098 |
upon | -0.096 |
spine | -0.094 |
Deal | -0.093 |
Token | little |
Token ID | 1310 |
Feature activation | -6.00e+00 |
Pos logit contributions | |
otion | +0.203 |
Deal | +0.161 |
spine | +0.160 |
laptop | +0.143 |
acy | +0.139 |
Neg logit contributions | |
cream | -0.272 |
oc | -0.245 |
Blue | -0.245 |
Fin | -0.244 |
The | -0.241 |
Token | brother |
Token ID | 3956 |
Feature activation | -6.91e+00 |
Pos logit contributions | |
otion | +0.247 |
Deal | +0.160 |
spine | +0.145 |
acy | +0.142 |
laptop | +0.141 |
Neg logit contributions | |
cream | -0.495 |
Fin | -0.467 |
oc | -0.465 |
Blue | -0.461 |
Tweet | -0.452 |
Token | Thank |
Token ID | 10449 |
Feature activation | +6.30e+00 |
Pos logit contributions | |
otion | +0.280 |
upon | +0.278 |
ancer | +0.274 |
acy | +0.273 |
Deal | +0.267 |
Neg logit contributions | |
cream | -0.177 |
bank | -0.128 |
Fin | -0.125 |
oc | -0.122 |
penter | -0.119 |
Token | you |
Token ID | 345 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
cream | +0.337 |
Fin | +0.304 |
Blue | +0.301 |
Jerry | +0.280 |
Red | +0.277 |
Neg logit contributions | |
otion | -0.427 |
upon | -0.383 |
laptop | -0.381 |
Deal | -0.369 |
sister | -0.367 |
Token | , |
Token ID | 11 |
Feature activation | +4.86e-01 |
Pos logit contributions | |
otion | +0.069 |
laptop | +0.061 |
spine | +0.059 |
upon | +0.058 |
headache | +0.058 |
Neg logit contributions | |
cream | -0.054 |
Fin | -0.051 |
Blue | -0.050 |
oc | -0.048 |
Jerry | -0.046 |
Token | f |
Token ID | 277 |
Feature activation | -1.70e+00 |
Pos logit contributions | |
cream | +0.025 |
Blue | +0.025 |
Fin | +0.024 |
sav | +0.023 |
Tweet | +0.023 |
Neg logit contributions | |
otion | -0.030 |
laptop | -0.028 |
spine | -0.028 |
ids | -0.027 |
upon | -0.026 |
Token | auc |
Token ID | 14272 |
Feature activation | -4.58e+00 |
Pos logit contributions | |
laptop | +0.085 |
otion | +0.083 |
sister | +0.080 |
upon | +0.079 |
spine | +0.079 |
Neg logit contributions | |
cream | -0.116 |
Fin | -0.105 |
Blue | -0.103 |
aw | -0.101 |
Tweet | -0.099 |
Token | et |
Token ID | 316 |
Feature activation | +8.62e+00 |
Pos logit contributions | |
otion | +0.355 |
laptop | +0.323 |
upon | +0.321 |
sister | +0.320 |
headache | +0.313 |
Neg logit contributions | |
cream | -0.186 |
Blue | -0.170 |
Fin | -0.167 |
Jerry | -0.155 |
Red | -0.153 |
Token | , |
Token ID | 11 |
Feature activation | +4.66e-01 |
Pos logit contributions | |
cream | +0.401 |
Fin | +0.345 |
Blue | +0.332 |
oc | +0.322 |
aw | +0.306 |
Neg logit contributions | |
otion | -0.647 |
Deal | -0.588 |
laptop | -0.588 |
spine | -0.573 |
upon | -0.570 |
Token | for |
Token ID | 329 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
cream | +0.024 |
Fin | +0.022 |
Blue | +0.022 |
sav | +0.021 |
Tweet | +0.020 |
Neg logit contributions | |
otion | -0.031 |
laptop | -0.029 |
spine | -0.028 |
Deal | -0.027 |
upon | -0.026 |
Token | helping |
Token ID | 5742 |
Feature activation | -8.52e-02 |
Pos logit contributions | |
otion | +0.099 |
upon | +0.087 |
laptop | +0.087 |
sister | +0.084 |
Deal | +0.084 |
Neg logit contributions | |
cream | -0.094 |
Fin | -0.086 |
Blue | -0.085 |
Jerry | -0.080 |
Red | -0.079 |
Token | me |
Token ID | 502 |
Feature activation | +2.58e+00 |
Pos logit contributions | |
otion | +0.006 |
upon | +0.006 |
laptop | +0.006 |
sister | +0.006 |
headache | +0.006 |
Neg logit contributions | |
cream | -0.004 |
Fin | -0.003 |
Blue | -0.003 |
Jerry | -0.003 |
Red | -0.003 |
Token | grow |
Token ID | 1663 |
Feature activation | +2.79e+00 |
Pos logit contributions | |
cream | +0.141 |
Fin | +0.127 |
Blue | +0.126 |
oc | +0.116 |
Jerry | +0.116 |
Neg logit contributions | |
otion | -0.170 |
laptop | -0.152 |
upon | -0.151 |
Deal | -0.147 |
spine | -0.145 |
Token | friends |
Token ID | 2460 |
Feature activation | +3.50e+00 |
Pos logit contributions | |
cream | +0.113 |
Fin | +0.102 |
Blue | +0.099 |
Jerry | +0.093 |
oc | +0.093 |
Neg logit contributions | |
otion | -0.151 |
upon | -0.135 |
laptop | -0.134 |
Deal | -0.131 |
spine | -0.131 |
Token | and |
Token ID | 290 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
cream | +0.212 |
Fin | +0.194 |
Blue | +0.189 |
Jerry | +0.178 |
Tweet | +0.178 |
Neg logit contributions | |
otion | -0.211 |
upon | -0.191 |
laptop | -0.186 |
sister | -0.184 |
spine | -0.179 |
Token | had |
Token ID | 550 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
otion | +0.098 |
sister | +0.095 |
upon | +0.092 |
laptop | +0.092 |
headache | +0.089 |
Neg logit contributions | |
cream | -0.054 |
Fin | -0.049 |
Blue | -0.047 |
Red | -0.045 |
Jerry | -0.043 |
Token | lots |
Token ID | 6041 |
Feature activation | +5.06e-02 |
Pos logit contributions | |
cream | +0.071 |
oc | +0.068 |
Fin | +0.066 |
rob | +0.065 |
bike | +0.064 |
Neg logit contributions | |
upon | -0.047 |
sister | -0.047 |
bed | -0.045 |
brother | -0.042 |
toes | -0.040 |
Token | of |
Token ID | 286 |
Feature activation | -3.21e+00 |
Pos logit contributions | |
cream | +0.003 |
Fin | +0.003 |
Blue | +0.003 |
Jerry | +0.003 |
Red | +0.003 |
Neg logit contributions | |
otion | -0.003 |
upon | -0.002 |
laptop | -0.002 |
sister | -0.002 |
headache | -0.002 |
Token | fun |
Token ID | 1257 |
Feature activation | +8.62e+00 |
Pos logit contributions | |
otion | +0.288 |
upon | +0.273 |
laptop | +0.272 |
sister | +0.267 |
Deal | +0.267 |
Neg logit contributions | |
cream | -0.099 |
Fin | -0.084 |
Blue | -0.081 |
Tweet | -0.070 |
Jerry | -0.069 |
Token | together |
Token ID | 1978 |
Feature activation | +4.98e+00 |
Pos logit contributions | |
cream | +0.496 |
Blue | +0.472 |
Fin | +0.469 |
Tweet | +0.438 |
Red | +0.435 |
Neg logit contributions | |
otion | -0.540 |
laptop | -0.494 |
upon | -0.494 |
sister | -0.482 |
headache | -0.456 |
Token | . |
Token ID | 13 |
Feature activation | -1.82e-01 |
Pos logit contributions | |
cream | +0.310 |
Fin | +0.285 |
Blue | +0.280 |
Red | +0.266 |
Jerry | +0.264 |
Neg logit contributions | |
otion | -0.293 |
upon | -0.265 |
laptop | -0.258 |
sister | -0.252 |
spine | -0.246 |
Token | And |
Token ID | 843 |
Feature activation | -2.87e+00 |
Pos logit contributions | |
otion | +0.014 |
upon | +0.013 |
laptop | +0.012 |
sister | +0.012 |
Deal | +0.012 |
Neg logit contributions | |
cream | -0.008 |
Fin | -0.007 |
Blue | -0.007 |
Jerry | -0.007 |
Red | -0.006 |
Token | they |
Token ID | 484 |
Feature activation | -5.46e-01 |
Pos logit contributions | |
otion | +0.188 |
upon | +0.171 |
laptop | +0.166 |
sister | +0.165 |
headache | +0.159 |
Neg logit contributions | |
cream | -0.160 |
Fin | -0.143 |
Blue | -0.141 |
Red | -0.133 |
Jerry | -0.132 |
Token | lived |
Token ID | 5615 |
Feature activation | -1.50e+00 |
Pos logit contributions | |
otion | +0.034 |
upon | +0.032 |
sister | +0.031 |
kin | +0.030 |
acy | +0.029 |
Neg logit contributions | |
cream | -0.031 |
Fin | -0.026 |
Blue | -0.026 |
Red | -0.026 |
Jerry | -0.025 |
Token | was |
Token ID | 373 |
Feature activation | -3.45e-01 |
Pos logit contributions | |
otion | +0.421 |
upon | +0.411 |
laptop | +0.394 |
sister | +0.387 |
spine | +0.377 |
Neg logit contributions | |
cream | -0.314 |
Fin | -0.287 |
Blue | -0.273 |
Jerry | -0.262 |
rob | -0.257 |
Token | scared |
Token ID | 12008 |
Feature activation | -2.33e+00 |
Pos logit contributions | |
otion | +0.025 |
laptop | +0.023 |
upon | +0.023 |
Deal | +0.022 |
spine | +0.022 |
Neg logit contributions | |
cream | -0.016 |
Fin | -0.014 |
Blue | -0.014 |
Jerry | -0.013 |
Red | -0.013 |
Token | and |
Token ID | 290 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
otion | +0.161 |
laptop | +0.154 |
Deal | +0.147 |
kin | +0.143 |
ids | +0.140 |
Neg logit contributions | |
cream | -0.107 |
Fin | -0.102 |
Blue | -0.100 |
He | -0.099 |
aw | -0.096 |
Token | wanted |
Token ID | 2227 |
Feature activation | +5.61e-01 |
Pos logit contributions | |
otion | +0.081 |
laptop | +0.076 |
Deal | +0.074 |
ancer | +0.072 |
spine | +0.071 |
Neg logit contributions | |
cream | -0.047 |
Blue | -0.043 |
Fin | -0.042 |
sav | -0.041 |
He | -0.040 |
Token | to |
Token ID | 284 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
cream | +0.027 |
Fin | +0.024 |
Blue | +0.024 |
Jerry | +0.022 |
Tweet | +0.021 |
Neg logit contributions | |
otion | -0.041 |
laptop | -0.037 |
upon | -0.037 |
Deal | -0.036 |
spine | -0.036 |
Token | go |
Token ID | 467 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
cream | +0.060 |
Fin | +0.057 |
Blue | +0.056 |
Jerry | +0.052 |
Tweet | +0.052 |
Neg logit contributions | |
otion | -0.071 |
laptop | -0.064 |
upon | -0.063 |
Deal | -0.061 |
spine | -0.060 |
Token | back |
Token ID | 736 |
Feature activation | +4.06e+00 |
Pos logit contributions | |
otion | +0.085 |
upon | +0.077 |
laptop | +0.074 |
spine | +0.072 |
sister | +0.072 |
Neg logit contributions | |
cream | -0.090 |
Blue | -0.082 |
Fin | -0.082 |
Jerry | -0.077 |
Red | -0.076 |
Token | home |
Token ID | 1363 |
Feature activation | +2.12e+00 |
Pos logit contributions | |
cream | +0.238 |
Fin | +0.218 |
Blue | +0.215 |
Jerry | +0.201 |
Red | +0.199 |
Neg logit contributions | |
otion | -0.253 |
laptop | -0.225 |
upon | -0.223 |
Deal | -0.216 |
headache | -0.214 |
Token | . |
Token ID | 13 |
Feature activation | +8.12e-02 |
Pos logit contributions | |
cream | +0.091 |
Fin | +0.079 |
Blue | +0.078 |
He | +0.075 |
sav | +0.075 |
Neg logit contributions | |
otion | -0.161 |
Deal | -0.146 |
laptop | -0.139 |
headache | -0.138 |
ancer | -0.135 |
Token | |
Token ID | 220 |
Feature activation | -2.43e-01 |
Pos logit contributions | |
cream | +0.004 |
Blue | +0.003 |
Fin | +0.003 |
Jerry | +0.003 |
Red | +0.003 |
Neg logit contributions | |
otion | -0.006 |
laptop | -0.006 |
upon | -0.006 |
headache | -0.005 |
spine | -0.005 |
Token |
|
Token ID | 198 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
otion | +0.022 |
laptop | +0.021 |
spine | +0.021 |
headache | +0.020 |
sister | +0.020 |
Neg logit contributions | |
cream | -0.007 |
Blue | -0.006 |
Fin | -0.006 |
Jerry | -0.005 |
Red | -0.005 |
Token | and |
Token ID | 290 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
cream | +0.013 |
Fin | +0.012 |
Blue | +0.012 |
sav | +0.011 |
Jerry | +0.011 |
Neg logit contributions | |
otion | -0.016 |
laptop | -0.014 |
upon | -0.014 |
Deal | -0.013 |
spine | -0.013 |
Token | so |
Token ID | 523 |
Feature activation | +2.21e+00 |
Pos logit contributions | |
otion | +0.087 |
upon | +0.079 |
laptop | +0.077 |
sister | +0.076 |
Deal | +0.074 |
Neg logit contributions | |
cream | -0.076 |
Fin | -0.070 |
Blue | -0.069 |
Jerry | -0.065 |
Red | -0.064 |
Token | the |
Token ID | 262 |
Feature activation | -3.07e-01 |
Pos logit contributions | |
cream | +0.132 |
Fin | +0.122 |
Blue | +0.120 |
Jerry | +0.112 |
oc | +0.112 |
Neg logit contributions | |
otion | -0.134 |
laptop | -0.119 |
upon | -0.117 |
Deal | -0.114 |
spine | -0.113 |
Token | old |
Token ID | 1468 |
Feature activation | +2.97e+00 |
Pos logit contributions | |
otion | +0.019 |
upon | +0.017 |
sister | +0.017 |
laptop | +0.017 |
Deal | +0.016 |
Neg logit contributions | |
cream | -0.018 |
Fin | -0.017 |
Blue | -0.016 |
oc | -0.016 |
Tweet | -0.016 |
Token | man |
Token ID | 582 |
Feature activation | +7.64e-02 |
Pos logit contributions | |
cream | +0.168 |
Fin | +0.155 |
Blue | +0.152 |
Red | +0.143 |
Tweet | +0.142 |
Neg logit contributions | |
otion | -0.190 |
upon | -0.175 |
laptop | -0.169 |
sister | -0.169 |
Lee | -0.164 |
Token | took |
Token ID | 1718 |
Feature activation | +7.15e-01 |
Pos logit contributions | |
cream | +0.003 |
Fin | +0.003 |
Blue | +0.003 |
Jerry | +0.003 |
Red | +0.003 |
Neg logit contributions | |
otion | -0.006 |
upon | -0.005 |
sister | -0.005 |
laptop | -0.005 |
headache | -0.005 |
Token | her |
Token ID | 607 |
Feature activation | -4.11e+00 |
Pos logit contributions | |
cream | +0.044 |
Fin | +0.040 |
Blue | +0.040 |
Jerry | +0.038 |
Red | +0.037 |
Neg logit contributions | |
otion | -0.042 |
upon | -0.037 |
laptop | -0.037 |
Deal | -0.036 |
sister | -0.036 |
Token | to |
Token ID | 284 |
Feature activation | +8.87e-01 |
Pos logit contributions | |
otion | +0.319 |
Deal | +0.275 |
laptop | +0.269 |
Lee | +0.263 |
headache | +0.261 |
Neg logit contributions | |
cream | -0.187 |
Blue | -0.155 |
oc | -0.152 |
Fin | -0.151 |
This | -0.147 |
Token | a |
Token ID | 257 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
cream | +0.036 |
Fin | +0.033 |
Blue | +0.033 |
sav | +0.033 |
The | +0.031 |
Neg logit contributions | |
otion | -0.070 |
laptop | -0.060 |
spine | -0.060 |
upon | -0.060 |
Deal | -0.059 |
Token | room |
Token ID | 2119 |
Feature activation | -1.99e+00 |
Pos logit contributions | |
otion | +0.116 |
upon | +0.116 |
laptop | +0.115 |
sister | +0.111 |
headache | +0.110 |
Neg logit contributions | |
cream | -0.103 |
Fin | -0.089 |
Blue | -0.087 |
Jerry | -0.083 |
Tweet | -0.082 |
Token | in |
Token ID | 287 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
otion | +0.149 |
Deal | +0.138 |
laptop | +0.130 |
headache | +0.130 |
sister | +0.129 |
Neg logit contributions | |
cream | -0.089 |
sold | -0.073 |
sav | -0.072 |
Fin | -0.072 |
Blue | -0.071 |
Token | boy |
Token ID | 2933 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
otion | +0.035 |
ids | +0.032 |
Deal | +0.032 |
acy | +0.032 |
spine | +0.031 |
Neg logit contributions | |
cream | -0.018 |
Blue | -0.015 |
Red | -0.015 |
Fin | -0.015 |
He | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | -5.43e-01 |
Pos logit contributions | |
cream | +0.048 |
Blue | +0.042 |
Fin | +0.041 |
oc | +0.041 |
He | +0.038 |
Neg logit contributions | |
otion | -0.076 |
upon | -0.069 |
Deal | -0.069 |
laptop | -0.068 |
headache | -0.068 |
Token | The |
Token ID | 383 |
Feature activation | +9.85e-01 |
Pos logit contributions | |
otion | +0.050 |
laptop | +0.046 |
upon | +0.046 |
spine | +0.045 |
Deal | +0.045 |
Neg logit contributions | |
cream | -0.015 |
Blue | -0.013 |
Fin | -0.013 |
Jerry | -0.011 |
Red | -0.011 |
Token | boy |
Token ID | 2933 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
cream | +0.041 |
Fin | +0.037 |
Blue | +0.036 |
Red | +0.034 |
He | +0.034 |
Neg logit contributions | |
otion | -0.077 |
ids | -0.070 |
spine | -0.068 |
acy | -0.067 |
Deal | -0.067 |
Token | had |
Token ID | 550 |
Feature activation | +8.42e-01 |
Pos logit contributions | |
cream | +0.057 |
Fin | +0.056 |
Blue | +0.052 |
Jerry | +0.049 |
Red | +0.049 |
Neg logit contributions | |
otion | -0.066 |
upon | -0.061 |
sister | -0.061 |
laptop | -0.060 |
Lee | -0.057 |
Token | a |
Token ID | 257 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
cream | +0.054 |
Blue | +0.049 |
Fin | +0.049 |
oc | +0.049 |
Red | +0.046 |
Neg logit contributions | |
otion | -0.045 |
upon | -0.044 |
laptop | -0.043 |
sister | -0.042 |
headache | -0.042 |
Token | red |
Token ID | 2266 |
Feature activation | +5.27e+00 |
Pos logit contributions | |
otion | +0.129 |
upon | +0.107 |
acy | +0.107 |
ancer | +0.106 |
laptop | +0.106 |
Neg logit contributions | |
cream | -0.097 |
Fin | -0.094 |
Blue | -0.094 |
Jerry | -0.089 |
Red | -0.086 |
Token | mark |
Token ID | 1317 |
Feature activation | +3.83e+00 |
Pos logit contributions | |
cream | +0.212 |
Blue | +0.182 |
Fin | +0.178 |
oc | +0.177 |
He | +0.171 |
Neg logit contributions | |
otion | -0.425 |
acy | -0.386 |
Deal | -0.374 |
ancer | -0.365 |
upon | -0.361 |
Token | on |
Token ID | 319 |
Feature activation | -2.66e+00 |
Pos logit contributions | |
cream | +0.201 |
Blue | +0.173 |
Fin | +0.170 |
He | +0.163 |
aw | +0.161 |
Neg logit contributions | |
otion | -0.258 |
Deal | -0.240 |
upon | -0.233 |
laptop | -0.232 |
spine | -0.227 |
Token | his |
Token ID | 465 |
Feature activation | +5.55e-01 |
Pos logit contributions | |
otion | +0.226 |
upon | +0.205 |
laptop | +0.202 |
ancer | +0.200 |
Deal | +0.200 |
Neg logit contributions | |
cream | -0.096 |
Fin | -0.085 |
Blue | -0.084 |
Jerry | -0.077 |
Red | -0.074 |
Token | eye |
Token ID | 4151 |
Feature activation | -3.71e+00 |
Pos logit contributions | |
cream | +0.038 |
Blue | +0.034 |
Fin | +0.033 |
Red | +0.033 |
oc | +0.033 |
Neg logit contributions | |
otion | -0.031 |
upon | -0.026 |
ancer | -0.024 |
Deal | -0.024 |
acy | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | -4.73e-02 |
Pos logit contributions | |
cream | +0.086 |
Fin | +0.077 |
Blue | +0.076 |
Jerry | +0.071 |
oc | +0.070 |
Neg logit contributions | |
otion | -0.115 |
laptop | -0.102 |
upon | -0.102 |
Deal | -0.101 |
headache | -0.099 |
Token | As |
Token ID | 1081 |
Feature activation | -4.21e+00 |
Pos logit contributions | |
otion | +0.004 |
upon | +0.003 |
laptop | +0.003 |
headache | +0.003 |
spine | +0.003 |
Neg logit contributions | |
cream | -0.002 |
Blue | -0.002 |
Fin | -0.002 |
Red | -0.002 |
Jerry | -0.002 |
Token | they |
Token ID | 484 |
Feature activation | -3.16e-01 |
Pos logit contributions | |
otion | +0.320 |
Deal | +0.289 |
laptop | +0.287 |
spine | +0.282 |
upon | +0.278 |
Neg logit contributions | |
cream | -0.189 |
Fin | -0.176 |
Blue | -0.170 |
Jerry | -0.156 |
Tweet | -0.153 |
Token | arrived |
Token ID | 5284 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
otion | +0.023 |
upon | +0.021 |
laptop | +0.020 |
sister | +0.020 |
headache | +0.020 |
Neg logit contributions | |
cream | -0.015 |
Fin | -0.014 |
Blue | -0.014 |
Jerry | -0.013 |
Red | -0.012 |
Token | , |
Token ID | 11 |
Feature activation | +5.40e-01 |
Pos logit contributions | |
otion | +0.093 |
laptop | +0.087 |
Deal | +0.087 |
headache | +0.083 |
acy | +0.080 |
Neg logit contributions | |
cream | -0.065 |
Fin | -0.064 |
aw | -0.062 |
Blue | -0.059 |
Jerry | -0.057 |
Token | they |
Token ID | 484 |
Feature activation | -3.58e-01 |
Pos logit contributions | |
cream | +0.019 |
Fin | +0.018 |
Blue | +0.017 |
sold | +0.016 |
sailed | +0.016 |
Neg logit contributions | |
otion | -0.044 |
Deal | -0.043 |
spine | -0.042 |
laptop | -0.041 |
arl | -0.040 |
Token | found |
Token ID | 1043 |
Feature activation | +4.70e+00 |
Pos logit contributions | |
otion | +0.028 |
laptop | +0.025 |
Deal | +0.025 |
spine | +0.025 |
headache | +0.024 |
Neg logit contributions | |
cream | -0.015 |
Fin | -0.013 |
Blue | -0.013 |
Jerry | -0.012 |
Tweet | -0.012 |
Token | it |
Token ID | 340 |
Feature activation | +3.44e+00 |
Pos logit contributions | |
cream | +0.248 |
Fin | +0.224 |
Blue | +0.222 |
Jerry | +0.206 |
Red | +0.203 |
Neg logit contributions | |
otion | -0.321 |
laptop | -0.288 |
upon | -0.288 |
Deal | -0.278 |
spine | -0.277 |
Token | empty |
Token ID | 6565 |
Feature activation | +5.46e-01 |
Pos logit contributions | |
cream | +0.171 |
Fin | +0.155 |
Blue | +0.153 |
Jerry | +0.140 |
Tweet | +0.139 |
Neg logit contributions | |
otion | -0.241 |
Deal | -0.221 |
laptop | -0.218 |
headache | -0.215 |
spine | -0.211 |
Token | . |
Token ID | 13 |
Feature activation | +7.87e-02 |
Pos logit contributions | |
cream | +0.025 |
aw | +0.020 |
Fin | +0.020 |
sav | +0.020 |
Blue | +0.020 |
Neg logit contributions | |
otion | -0.041 |
Deal | -0.039 |
headache | -0.036 |
ancer | -0.036 |
acy | -0.035 |
Token |
|
Token ID | 198 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
cream | +0.003 |
Blue | +0.003 |
Fin | +0.003 |
Jerry | +0.003 |
Red | +0.003 |
Neg logit contributions | |
otion | -0.006 |
laptop | -0.006 |
upon | -0.006 |
spine | -0.006 |
headache | -0.006 |
Token | in |
Token ID | 287 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
cream | +0.506 |
Fin | +0.440 |
Blue | +0.430 |
aw | +0.402 |
sailed | +0.401 |
Neg logit contributions | |
otion | -0.649 |
Deal | -0.624 |
laptop | -0.594 |
headache | -0.579 |
sister | -0.574 |
Token | the |
Token ID | 262 |
Feature activation | -1.58e-01 |
Pos logit contributions | |
otion | +0.077 |
laptop | +0.069 |
upon | +0.068 |
Deal | +0.067 |
spine | +0.067 |
Neg logit contributions | |
cream | -0.055 |
Fin | -0.050 |
Blue | -0.049 |
Jerry | -0.045 |
oc | -0.045 |
Token | bushes |
Token ID | 37413 |
Feature activation | -2.25e+00 |
Pos logit contributions | |
upon | +0.008 |
spine | +0.008 |
laptop | +0.008 |
curtain | +0.008 |
sister | +0.008 |
Neg logit contributions | |
cream | -0.010 |
Fin | -0.009 |
Jerry | -0.009 |
rob | -0.009 |
oc | -0.009 |
Token | with |
Token ID | 351 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
otion | +0.151 |
laptop | +0.136 |
upon | +0.136 |
Deal | +0.132 |
sister | +0.131 |
Neg logit contributions | |
cream | -0.119 |
Fin | -0.108 |
Blue | -0.107 |
Jerry | -0.099 |
Red | -0.098 |
Token | a |
Token ID | 257 |
Feature activation | -1.79e+00 |
Pos logit contributions | |
cream | +0.110 |
Fin | +0.100 |
Blue | +0.099 |
Jerry | +0.093 |
Red | +0.092 |
Neg logit contributions | |
otion | -0.124 |
upon | -0.113 |
laptop | -0.111 |
sister | -0.108 |
headache | -0.107 |
Token | cheerful |
Token ID | 37999 |
Feature activation | +2.29e+00 |
Pos logit contributions | |
otion | +0.120 |
laptop | +0.107 |
upon | +0.107 |
Deal | +0.104 |
sister | +0.103 |
Neg logit contributions | |
cream | -0.095 |
Fin | -0.086 |
Blue | -0.086 |
Jerry | -0.079 |
Red | -0.078 |
Token | chicken |
Token ID | 9015 |
Feature activation | +6.46e-01 |
Pos logit contributions | |
cream | +0.125 |
Fin | +0.116 |
Blue | +0.115 |
oc | +0.107 |
Tweet | +0.107 |
Neg logit contributions | |
otion | -0.150 |
laptop | -0.132 |
Deal | -0.130 |
upon | -0.129 |
spine | -0.124 |
Token | sitting |
Token ID | 5586 |
Feature activation | -2.75e+00 |
Pos logit contributions | |
cream | +0.033 |
Fin | +0.029 |
Blue | +0.029 |
oc | +0.027 |
Tweet | +0.026 |
Neg logit contributions | |
otion | -0.045 |
Deal | -0.041 |
laptop | -0.040 |
upon | -0.040 |
headache | -0.039 |
Token | on |
Token ID | 319 |
Feature activation | -2.46e+00 |
Pos logit contributions | |
otion | +0.175 |
laptop | +0.159 |
Deal | +0.153 |
headache | +0.152 |
upon | +0.151 |
Neg logit contributions | |
cream | -0.159 |
Fin | -0.143 |
Blue | -0.141 |
Jerry | -0.131 |
oc | -0.131 |
Token | it |
Token ID | 340 |
Feature activation | +3.31e+00 |
Pos logit contributions | |
otion | +0.162 |
upon | +0.146 |
laptop | +0.146 |
headache | +0.140 |
sister | +0.138 |
Neg logit contributions | |
cream | -0.133 |
Fin | -0.121 |
Blue | -0.118 |
Jerry | -0.112 |
Red | -0.110 |
Token | . |
Token ID | 13 |
Feature activation | -1.03e-01 |
Pos logit contributions | |
cream | +0.180 |
Fin | +0.164 |
Blue | +0.161 |
Jerry | +0.151 |
Red | +0.149 |
Neg logit contributions | |
otion | -0.218 |
upon | -0.196 |
laptop | -0.195 |
sister | -0.187 |
spine | -0.187 |
Token | little |
Token ID | 1310 |
Feature activation | -6.44e+00 |
Pos logit contributions | |
otion | +0.128 |
upon | +0.110 |
Deal | +0.109 |
laptop | +0.109 |
ids | +0.107 |
Neg logit contributions | |
cream | -0.095 |
Blue | -0.092 |
Fin | -0.092 |
oc | -0.085 |
Red | -0.085 |
Token | clumsy |
Token ID | 39120 |
Feature activation | -3.88e-02 |
Pos logit contributions | |
otion | +0.378 |
upon | +0.330 |
laptop | +0.327 |
Deal | +0.325 |
spine | +0.312 |
Neg logit contributions | |
cream | -0.398 |
Fin | -0.366 |
Blue | -0.365 |
oc | -0.343 |
Red | -0.341 |
Token | and |
Token ID | 290 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
otion | +0.003 |
Deal | +0.003 |
upon | +0.003 |
laptop | +0.002 |
acy | +0.002 |
Neg logit contributions | |
cream | -0.002 |
oc | -0.002 |
Fin | -0.002 |
Blue | -0.002 |
aw | -0.002 |
Token | she |
Token ID | 673 |
Feature activation | -3.46e+00 |
Pos logit contributions | |
otion | +0.094 |
laptop | +0.085 |
Deal | +0.084 |
spine | +0.083 |
upon | +0.082 |
Neg logit contributions | |
cream | -0.065 |
Blue | -0.059 |
Fin | -0.058 |
Jerry | -0.054 |
Tweet | -0.054 |
Token | accidentally |
Token ID | 14716 |
Feature activation | -5.30e+00 |
Pos logit contributions | |
otion | +0.193 |
upon | +0.184 |
sister | +0.182 |
laptop | +0.179 |
headache | +0.171 |
Neg logit contributions | |
cream | -0.213 |
Fin | -0.197 |
Blue | -0.192 |
Jerry | -0.182 |
rob | -0.180 |
Token | dropped |
Token ID | 5710 |
Feature activation | -5.25e+00 |
Pos logit contributions | |
otion | +0.375 |
Deal | +0.355 |
ancer | +0.342 |
spine | +0.336 |
upon | +0.333 |
Neg logit contributions | |
cream | -0.287 |
Blue | -0.231 |
sailed | -0.229 |
Fin | -0.228 |
sold | -0.226 |
Token | the |
Token ID | 262 |
Feature activation | -2.94e-01 |
Pos logit contributions | |
otion | +0.413 |
laptop | +0.378 |
Deal | +0.374 |
spine | +0.372 |
ancer | +0.369 |
Neg logit contributions | |
cream | -0.228 |
Blue | -0.192 |
Fin | -0.187 |
Red | -0.172 |
This | -0.171 |
Token | box |
Token ID | 3091 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
otion | +0.022 |
Deal | +0.021 |
acy | +0.020 |
ancer | +0.019 |
ids | +0.019 |
Neg logit contributions | |
cream | -0.012 |
Blue | -0.011 |
Red | -0.010 |
sold | -0.010 |
oc | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | -2.98e-01 |
Pos logit contributions | |
otion | +0.072 |
Deal | +0.066 |
laptop | +0.065 |
headache | +0.064 |
upon | +0.064 |
Neg logit contributions | |
cream | -0.051 |
Blue | -0.044 |
Fin | -0.043 |
Jerry | -0.041 |
Red | -0.040 |
Token | Her |
Token ID | 2332 |
Feature activation | -3.89e+00 |
Pos logit contributions | |
otion | +0.023 |
ancer | +0.021 |
upon | +0.021 |
Deal | +0.020 |
kin | +0.020 |
Neg logit contributions | |
cream | -0.012 |
Blue | -0.012 |
Red | -0.011 |
sav | -0.011 |
This | -0.010 |
Token | mom |
Token ID | 1995 |
Feature activation | -3.66e+00 |
Pos logit contributions | |
otion | +0.276 |
Deal | +0.235 |
ancer | +0.231 |
acy | +0.223 |
911 | +0.202 |
Neg logit contributions | |
cream | -0.186 |
feathers | -0.172 |
nut | -0.172 |
Tweet | -0.170 |
Blue | -0.168 |
Token | . |
Token ID | 13 |
Feature activation | +4.55e-01 |
Pos logit contributions | |
cream | +0.375 |
Fin | +0.340 |
Blue | +0.335 |
Red | +0.312 |
oc | +0.310 |
Neg logit contributions | |
otion | -0.456 |
laptop | -0.404 |
Deal | -0.393 |
upon | -0.388 |
spine | -0.388 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.94e+00 |
Pos logit contributions | |
cream | +0.025 |
Fin | +0.022 |
Blue | +0.021 |
Tweet | +0.021 |
Jerry | +0.020 |
Neg logit contributions | |
otion | -0.030 |
Deal | -0.028 |
upon | -0.027 |
laptop | -0.027 |
sister | -0.026 |
Token | Once |
Token ID | 7454 |
Feature activation | +1.51e+00 |
Pos logit contributions | |
otion | +0.139 |
laptop | +0.128 |
spine | +0.125 |
headache | +0.123 |
Deal | +0.123 |
Neg logit contributions | |
cream | -0.094 |
Fin | -0.085 |
Blue | -0.084 |
Jerry | -0.078 |
oc | -0.077 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.82e+01 |
Pos logit contributions | |
He | +0.117 |
Fin | +0.117 |
cream | +0.116 |
that | +0.115 |
The | +0.115 |
Neg logit contributions | |
laptop | -0.047 |
Deal | -0.043 |
otion | -0.039 |
scalp | -0.039 |
spine | -0.036 |
Token | a |
Token ID | 257 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
otion | +1.242 |
laptop | +1.115 |
upon | +1.110 |
Deal | +1.074 |
sister | +1.070 |
Neg logit contributions | |
cream | -0.952 |
Fin | -0.862 |
Blue | -0.855 |
Jerry | -0.792 |
Red | -0.784 |
Token | time |
Token ID | 640 |
Feature activation | -5.45e+00 |
Pos logit contributions | |
otion | +0.073 |
laptop | +0.061 |
upon | +0.059 |
sister | +0.058 |
spine | +0.057 |
Neg logit contributions | |
cream | -0.088 |
Fin | -0.087 |
Blue | -0.086 |
Jerry | -0.082 |
oc | -0.081 |
Token | , |
Token ID | 11 |
Feature activation | +7.36e-01 |
Pos logit contributions | |
laptop | +0.456 |
Deal | +0.433 |
forehead | +0.430 |
scalp | +0.429 |
spine | +0.427 |
Neg logit contributions | |
Fin | -0.165 |
Bob | -0.161 |
Blue | -0.152 |
that | -0.146 |
Tim | -0.139 |
Token | there |
Token ID | 612 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
Blue | +0.035 |
Fin | +0.034 |
cream | +0.033 |
Tweet | +0.032 |
He | +0.031 |
Neg logit contributions | |
laptop | -0.051 |
otion | -0.050 |
headache | -0.048 |
Deal | -0.048 |
spine | -0.048 |
Token | was |
Token ID | 373 |
Feature activation | +5.29e-02 |
Pos logit contributions | |
otion | +0.118 |
laptop | +0.109 |
upon | +0.108 |
Deal | +0.107 |
spine | +0.107 |
Neg logit contributions | |
cream | -0.045 |
Fin | -0.038 |
Blue | -0.038 |
Jerry | -0.033 |
oc | -0.033 |
Token | a |
Token ID | 257 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
cream | +0.003 |
Fin | +0.002 |
Blue | +0.002 |
Jerry | +0.002 |
Red | +0.002 |
Neg logit contributions | |
otion | -0.004 |
laptop | -0.003 |
upon | -0.003 |
spine | -0.003 |
Deal | -0.003 |
Token | j |
Token ID | 474 |
Feature activation | +1.71e+00 |
Pos logit contributions | |
otion | +0.081 |
upon | +0.070 |
laptop | +0.070 |
Deal | +0.068 |
spine | +0.067 |
Neg logit contributions | |
cream | -0.092 |
Fin | -0.086 |
Blue | -0.086 |
Jerry | -0.080 |
Red | -0.080 |
Token | is |
Token ID | 318 |
Feature activation | -5.39e-01 |
Pos logit contributions | |
cream | +0.015 |
Blue | +0.013 |
Fin | +0.013 |
oc | +0.012 |
Red | +0.011 |
Neg logit contributions | |
otion | -0.023 |
Deal | -0.021 |
laptop | -0.021 |
upon | -0.021 |
spine | -0.020 |
Token | ugly |
Token ID | 13400 |
Feature activation | +1.82e+00 |
Pos logit contributions | |
otion | +0.037 |
upon | +0.034 |
laptop | +0.034 |
Deal | +0.033 |
sister | +0.032 |
Neg logit contributions | |
cream | -0.028 |
Fin | -0.025 |
Blue | -0.025 |
Jerry | -0.023 |
Red | -0.023 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
cream | +0.080 |
Blue | +0.067 |
Fin | +0.067 |
oc | +0.062 |
Red | +0.062 |
Neg logit contributions | |
otion | -0.140 |
Deal | -0.127 |
laptop | -0.126 |
upon | -0.125 |
spine | -0.122 |
Token | Tim |
Token ID | 5045 |
Feature activation | +3.00e+00 |
Pos logit contributions | |
cream | +0.099 |
Fin | +0.092 |
Blue | +0.090 |
Jerry | +0.082 |
oc | +0.082 |
Neg logit contributions | |
otion | -0.139 |
upon | -0.127 |
laptop | -0.126 |
sister | -0.122 |
spine | -0.121 |
Token | my |
Token ID | 1820 |
Feature activation | +5.01e+00 |
Pos logit contributions | |
cream | +0.159 |
Fin | +0.148 |
Blue | +0.145 |
Red | +0.133 |
Jerry | +0.132 |
Neg logit contributions | |
otion | -0.197 |
upon | -0.186 |
sister | -0.180 |
laptop | -0.179 |
Deal | -0.174 |
Token | didn |
Token ID | 1422 |
Feature activation | -7.50e+00 |
Pos logit contributions | |
cream | +0.284 |
Fin | +0.274 |
Blue | +0.261 |
rob | +0.252 |
Red | +0.252 |
Neg logit contributions | |
sister | -0.296 |
upon | -0.285 |
otion | -0.285 |
laptop | -0.267 |
brother | -0.262 |
Token | 't |
Token ID | 470 |
Feature activation | -5.46e+00 |
Pos logit contributions | |
otion | +0.433 |
upon | +0.394 |
spine | +0.392 |
laptop | +0.390 |
sister | +0.379 |
Neg logit contributions | |
cream | -0.460 |
Blue | -0.423 |
Fin | -0.419 |
Red | -0.396 |
Jerry | -0.394 |
Token | like |
Token ID | 588 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
otion | +0.328 |
laptop | +0.296 |
upon | +0.291 |
Deal | +0.282 |
spine | +0.280 |
Neg logit contributions | |
cream | -0.327 |
Fin | -0.305 |
Blue | -0.303 |
Tweet | -0.282 |
Jerry | -0.281 |
Token | that |
Token ID | 326 |
Feature activation | +9.23e+00 |
Pos logit contributions | |
cream | +0.044 |
Fin | +0.039 |
Blue | +0.038 |
Jerry | +0.034 |
Red | +0.034 |
Neg logit contributions | |
otion | -0.086 |
upon | -0.079 |
laptop | -0.079 |
Deal | -0.076 |
spine | -0.076 |
Token | and |
Token ID | 290 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
cream | +0.437 |
Fin | +0.384 |
oc | +0.384 |
Blue | +0.378 |
Jerry | +0.358 |
Neg logit contributions | |
otion | -0.652 |
upon | -0.614 |
Deal | -0.610 |
acy | -0.587 |
ancer | -0.586 |
Token | said |
Token ID | 531 |
Feature activation | +2.61e+00 |
Pos logit contributions | |
otion | +0.090 |
upon | +0.082 |
sister | +0.081 |
laptop | +0.080 |
headache | +0.078 |
Neg logit contributions | |
cream | -0.065 |
Fin | -0.061 |
Blue | -0.060 |
oc | -0.056 |
Red | -0.055 |
Token | a |
Token ID | 257 |
Feature activation | -1.84e+00 |
Pos logit contributions | |
cream | +0.065 |
Fin | +0.062 |
Blue | +0.061 |
oc | +0.059 |
Tweet | +0.057 |
Neg logit contributions | |
otion | -0.065 |
upon | -0.064 |
laptop | -0.061 |
sister | -0.059 |
headache | -0.058 |
Token | big |
Token ID | 1263 |
Feature activation | +9.67e-01 |
Pos logit contributions | |
otion | +0.156 |
ids | +0.132 |
ancer | +0.131 |
acy | +0.130 |
spine | +0.128 |
Neg logit contributions | |
cream | -0.065 |
Blue | -0.064 |
Fin | -0.062 |
sold | -0.061 |
sav | -0.060 |
Token | smile |
Token ID | 8212 |
Feature activation | -2.51e+00 |
Pos logit contributions | |
sav | +0.047 |
oc | +0.046 |
sold | +0.044 |
cream | +0.044 |
nut | +0.044 |
Neg logit contributions | |
otion | -0.066 |
ancer | -0.059 |
acy | -0.057 |
ids | -0.055 |
upon | -0.055 |
Token | on |
Token ID | 319 |
Feature activation | -2.53e+00 |
Pos logit contributions | |
otion | +0.183 |
Deal | +0.178 |
laptop | +0.162 |
upon | +0.161 |
spine | +0.159 |
Neg logit contributions | |
cream | -0.125 |
oc | -0.096 |
sold | -0.095 |
sailed | -0.095 |
bo | -0.093 |
Token | her |
Token ID | 607 |
Feature activation | -4.12e+00 |
Pos logit contributions | |
otion | +0.188 |
laptop | +0.171 |
upon | +0.170 |
Deal | +0.165 |
sister | +0.165 |
Neg logit contributions | |
cream | -0.119 |
Fin | -0.107 |
Blue | -0.105 |
Jerry | -0.096 |
Red | -0.095 |
Token | face |
Token ID | 1986 |
Feature activation | -4.17e+00 |
Pos logit contributions | |
otion | +0.292 |
ancer | +0.245 |
upon | +0.237 |
acy | +0.222 |
ids | +0.221 |
Neg logit contributions | |
cream | -0.234 |
He | -0.201 |
sav | -0.191 |
bo | -0.191 |
oc | -0.184 |
Token | . |
Token ID | 13 |
Feature activation | -1.64e-01 |
Pos logit contributions | |
otion | +0.294 |
laptop | +0.264 |
upon | +0.263 |
Deal | +0.261 |
headache | +0.256 |
Neg logit contributions | |
cream | -0.212 |
Fin | -0.182 |
Blue | -0.181 |
oc | -0.170 |
Jerry | -0.168 |
Token | She |
Token ID | 1375 |
Feature activation | +3.99e-02 |
Pos logit contributions | |
otion | +0.011 |
laptop | +0.009 |
upon | +0.009 |
Deal | +0.009 |
spine | +0.009 |
Neg logit contributions | |
cream | -0.009 |
Fin | -0.008 |
Blue | -0.008 |
Jerry | -0.008 |
Red | -0.008 |
Token | said |
Token ID | 531 |
Feature activation | +2.65e+00 |
Pos logit contributions | |
cream | +0.002 |
Fin | +0.002 |
Blue | +0.002 |
Jerry | +0.002 |
Red | +0.002 |
Neg logit contributions | |
otion | -0.003 |
upon | -0.003 |
laptop | -0.003 |
sister | -0.003 |
headache | -0.002 |
Token | : |
Token ID | 25 |
Feature activation | -5.49e-01 |
Pos logit contributions | |
cream | +0.144 |
Blue | +0.126 |
Fin | +0.125 |
Jerry | +0.115 |
He | +0.115 |
Neg logit contributions | |
otion | -0.173 |
laptop | -0.163 |
Deal | -0.160 |
spine | -0.160 |
upon | -0.157 |
Token | " |
Token ID | 366 |
Feature activation | -4.46e+00 |
Pos logit contributions | |
otion | +0.036 |
laptop | +0.033 |
spine | +0.033 |
headache | +0.032 |
sister | +0.032 |
Neg logit contributions | |
cream | -0.028 |
Blue | -0.027 |
Fin | -0.026 |
He | -0.025 |
Jerry | -0.024 |
Token | makes |
Token ID | 1838 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
cream | +0.243 |
Fin | +0.230 |
Blue | +0.225 |
Jerry | +0.216 |
Red | +0.211 |
Neg logit contributions | |
otion | -0.206 |
laptop | -0.191 |
upon | -0.189 |
sister | -0.185 |
headache | -0.177 |
Token | funny |
Token ID | 8258 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
cream | +0.081 |
Fin | +0.074 |
Blue | +0.073 |
Jerry | +0.068 |
Red | +0.067 |
Neg logit contributions | |
otion | -0.104 |
upon | -0.095 |
laptop | -0.094 |
sister | -0.091 |
headache | -0.090 |
Token | noises |
Token ID | 26782 |
Feature activation | +3.23e-01 |
Pos logit contributions | |
cream | +0.097 |
Fin | +0.089 |
Blue | +0.088 |
Jerry | +0.082 |
Red | +0.082 |
Neg logit contributions | |
otion | -0.110 |
upon | -0.100 |
laptop | -0.099 |
sister | -0.095 |
headache | -0.094 |
Token | . |
Token ID | 13 |
Feature activation | -4.71e-01 |
Pos logit contributions | |
cream | +0.018 |
Fin | +0.017 |
Blue | +0.016 |
Jerry | +0.015 |
Red | +0.015 |
Neg logit contributions | |
otion | -0.021 |
upon | -0.019 |
laptop | -0.019 |
sister | -0.018 |
headache | -0.018 |
Token | L |
Token ID | 406 |
Feature activation | -6.85e+00 |
Pos logit contributions | |
otion | +0.035 |
upon | +0.032 |
laptop | +0.032 |
Deal | +0.031 |
sister | +0.031 |
Neg logit contributions | |
cream | -0.022 |
Fin | -0.019 |
Blue | -0.019 |
Jerry | -0.018 |
Red | -0.017 |
Token | ila |
Token ID | 10102 |
Feature activation | -8.33e+00 |
Pos logit contributions | |
spine | +0.209 |
headache | +0.204 |
otion | +0.203 |
laptop | +0.201 |
Deal | +0.194 |
Neg logit contributions | |
cream | -0.595 |
Fin | -0.559 |
aw | -0.541 |
Blue | -0.540 |
Jerry | -0.534 |
Token | and |
Token ID | 290 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
otion | +0.482 |
upon | +0.444 |
laptop | +0.440 |
sister | +0.428 |
Deal | +0.407 |
Neg logit contributions | |
cream | -0.507 |
Fin | -0.479 |
Blue | -0.462 |
Jerry | -0.440 |
Red | -0.434 |
Token | Ben |
Token ID | 3932 |
Feature activation | -6.56e+00 |
Pos logit contributions | |
otion | +0.074 |
laptop | +0.060 |
upon | +0.058 |
spine | +0.057 |
headache | +0.056 |
Neg logit contributions | |
cream | -0.100 |
sav | -0.095 |
Fin | -0.095 |
aw | -0.091 |
Blue | -0.091 |
Token | laugh |
Token ID | 6487 |
Feature activation | -4.33e+00 |
Pos logit contributions | |
otion | +0.310 |
upon | +0.267 |
laptop | +0.266 |
Deal | +0.251 |
sister | +0.251 |
Neg logit contributions | |
cream | -0.481 |
Fin | -0.447 |
Blue | -0.442 |
Jerry | -0.422 |
Red | -0.418 |
Token | and |
Token ID | 290 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
otion | +0.281 |
upon | +0.253 |
laptop | +0.250 |
sister | +0.242 |
Deal | +0.242 |
Neg logit contributions | |
cream | -0.243 |
Fin | -0.221 |
Blue | -0.216 |
Jerry | -0.204 |
Red | -0.199 |
Token | talk |
Token ID | 1561 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
otion | +0.077 |
laptop | +0.069 |
upon | +0.068 |
sister | +0.067 |
Deal | +0.065 |
Neg logit contributions | |
cream | -0.100 |
Fin | -0.092 |
Blue | -0.090 |
Jerry | -0.086 |
Red | -0.086 |
Token | fire |
Token ID | 2046 |
Feature activation | -8.67e-01 |
Pos logit contributions | |
cream | +0.083 |
Fin | +0.073 |
Blue | +0.072 |
oc | +0.070 |
Bob | +0.065 |
Neg logit contributions | |
otion | -0.112 |
laptop | -0.104 |
spine | -0.103 |
Deal | -0.103 |
upon | -0.101 |
Token | man |
Token ID | 805 |
Feature activation | +5.88e+00 |
Pos logit contributions | |
otion | +0.072 |
upon | +0.066 |
laptop | +0.066 |
Deal | +0.064 |
spine | +0.064 |
Neg logit contributions | |
cream | -0.033 |
Fin | -0.028 |
Blue | -0.028 |
Jerry | -0.025 |
oc | -0.024 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.92e+00 |
Pos logit contributions | |
cream | +0.295 |
Fin | +0.268 |
Blue | +0.265 |
Jerry | +0.246 |
Red | +0.243 |
Neg logit contributions | |
otion | -0.409 |
upon | -0.373 |
laptop | -0.369 |
sister | -0.364 |
Deal | -0.354 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.86e+00 |
Pos logit contributions | |
cream | +0.080 |
Blue | +0.069 |
Fin | +0.067 |
Jerry | +0.063 |
Red | +0.060 |
Neg logit contributions | |
otion | -0.155 |
laptop | -0.142 |
upon | -0.139 |
Deal | -0.137 |
headache | -0.136 |
Token | Once |
Token ID | 7454 |
Feature activation | +7.53e-01 |
Pos logit contributions | |
otion | +0.232 |
upon | +0.215 |
laptop | +0.212 |
Deal | +0.206 |
sister | +0.205 |
Neg logit contributions | |
cream | -0.112 |
Fin | -0.098 |
Blue | -0.096 |
Jerry | -0.086 |
Red | -0.085 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.89e+01 |
Pos logit contributions | |
He | +0.055 |
The | +0.054 |
that | +0.054 |
cream | +0.053 |
Fin | +0.052 |
Neg logit contributions | |
laptop | -0.027 |
otion | -0.025 |
Deal | -0.024 |
spine | -0.024 |
scalp | -0.023 |
Token | a |
Token ID | 257 |
Feature activation | -1.92e+00 |
Pos logit contributions | |
otion | +1.274 |
upon | +1.144 |
laptop | +1.142 |
Deal | +1.100 |
sister | +1.097 |
Neg logit contributions | |
cream | -1.006 |
Fin | -0.912 |
Blue | -0.902 |
Jerry | -0.838 |
Red | -0.828 |
Token | time |
Token ID | 640 |
Feature activation | -6.01e+00 |
Pos logit contributions | |
otion | +0.107 |
laptop | +0.090 |
ancer | +0.087 |
upon | +0.086 |
sister | +0.086 |
Neg logit contributions | |
cream | -0.121 |
Fin | -0.119 |
Blue | -0.118 |
Jerry | -0.114 |
oc | -0.112 |
Token | , |
Token ID | 11 |
Feature activation | +6.68e-02 |
Pos logit contributions | |
laptop | +0.496 |
spine | +0.476 |
scalp | +0.467 |
forehead | +0.462 |
Deal | +0.459 |
Neg logit contributions | |
Bob | -0.178 |
Fin | -0.177 |
Blue | -0.168 |
Jerry | -0.161 |
that | -0.160 |
Token | there |
Token ID | 612 |
Feature activation | -1.96e+00 |
Pos logit contributions | |
Blue | +0.003 |
Fin | +0.003 |
The | +0.003 |
cream | +0.003 |
It | +0.003 |
Neg logit contributions | |
laptop | -0.005 |
spine | -0.005 |
otion | -0.005 |
headache | -0.005 |
Deal | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | -5.52e-01 |
Pos logit contributions | |
otion | +0.160 |
upon | +0.148 |
laptop | +0.145 |
sister | +0.142 |
headache | +0.140 |
Neg logit contributions | |
cream | -0.078 |
Fin | -0.067 |
Blue | -0.066 |
Jerry | -0.059 |
Red | -0.059 |
Token | wonderful |
Token ID | 7932 |
Feature activation | +6.60e+00 |
Pos logit contributions | |
cream | +0.033 |
Fin | +0.029 |
Blue | +0.029 |
Red | +0.027 |
Tweet | +0.027 |
Neg logit contributions | |
otion | -0.046 |
upon | -0.045 |
sister | -0.044 |
laptop | -0.044 |
headache | -0.042 |
Token | adventures |
Token ID | 17545 |
Feature activation | +9.48e+00 |
Pos logit contributions | |
cream | +0.307 |
Blue | +0.284 |
Fin | +0.279 |
Red | +0.259 |
Tweet | +0.256 |
Neg logit contributions | |
otion | -0.473 |
upon | -0.449 |
laptop | -0.441 |
sister | -0.437 |
headache | -0.431 |
Token | ! |
Token ID | 0 |
Feature activation | +8.92e-01 |
Pos logit contributions | |
cream | +0.500 |
Fin | +0.466 |
Blue | +0.462 |
Red | +0.431 |
Tweet | +0.430 |
Neg logit contributions | |
otion | -0.622 |
upon | -0.571 |
laptop | -0.567 |
sister | -0.540 |
spine | -0.537 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.68e+00 |
Pos logit contributions | |
cream | +0.037 |
Fin | +0.032 |
Blue | +0.031 |
Jerry | +0.028 |
Red | +0.028 |
Neg logit contributions | |
otion | -0.071 |
upon | -0.065 |
laptop | -0.065 |
Deal | -0.063 |
sister | -0.063 |
Token | Once |
Token ID | 7454 |
Feature activation | +9.32e-01 |
Pos logit contributions | |
otion | +0.205 |
laptop | +0.188 |
upon | +0.183 |
spine | +0.182 |
Deal | +0.181 |
Neg logit contributions | |
cream | -0.117 |
Fin | -0.105 |
Blue | -0.104 |
Jerry | -0.095 |
Red | -0.094 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.87e+01 |
Pos logit contributions | |
He | +0.073 |
The | +0.072 |
that | +0.072 |
Fin | +0.071 |
Blue | +0.070 |
Neg logit contributions | |
laptop | -0.029 |
Deal | -0.028 |
otion | -0.025 |
scalp | -0.025 |
spine | -0.023 |
Token | a |
Token ID | 257 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
otion | +1.233 |
upon | +1.115 |
laptop | +1.102 |
Deal | +1.060 |
sister | +1.059 |
Neg logit contributions | |
cream | -1.019 |
Fin | -0.927 |
Blue | -0.912 |
Jerry | -0.851 |
oc | -0.841 |
Token | time |
Token ID | 640 |
Feature activation | -5.94e+00 |
Pos logit contributions | |
otion | +0.097 |
laptop | +0.082 |
upon | +0.079 |
sister | +0.078 |
spine | +0.077 |
Neg logit contributions | |
cream | -0.125 |
Fin | -0.121 |
Blue | -0.119 |
Jerry | -0.114 |
oc | -0.113 |
Token | there |
Token ID | 612 |
Feature activation | -1.84e+00 |
Pos logit contributions | |
laptop | +0.487 |
Deal | +0.473 |
forehead | +0.468 |
scalp | +0.467 |
headache | +0.460 |
Neg logit contributions | |
Fin | -0.185 |
Bob | -0.183 |
Blue | -0.168 |
that | -0.164 |
Jerry | -0.159 |
Token | was |
Token ID | 373 |
Feature activation | -4.15e-01 |
Pos logit contributions | |
otion | +0.149 |
upon | +0.137 |
laptop | +0.133 |
sister | +0.131 |
headache | +0.129 |
Neg logit contributions | |
cream | -0.075 |
Fin | -0.065 |
Blue | -0.064 |
Jerry | -0.058 |
Red | -0.057 |
Token | a |
Token ID | 257 |
Feature activation | -1.87e+00 |
Pos logit contributions | |
otion | +0.030 |
laptop | +0.028 |
spine | +0.027 |
upon | +0.027 |
Deal | +0.027 |
Neg logit contributions | |
cream | -0.020 |
Blue | -0.018 |
Fin | -0.018 |
Jerry | -0.016 |
Red | -0.016 |
Token | all |
Token ID | 477 |
Feature activation | +3.62e+00 |
Pos logit contributions | |
otion | +0.030 |
upon | +0.027 |
laptop | +0.026 |
sister | +0.026 |
spine | +0.026 |
Neg logit contributions | |
cream | -0.026 |
Fin | -0.023 |
Blue | -0.023 |
Jerry | -0.021 |
Red | -0.021 |
Token | day |
Token ID | 1110 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
cream | +0.243 |
Blue | +0.224 |
Fin | +0.218 |
Red | +0.215 |
Nut | +0.213 |
Neg logit contributions | |
otion | -0.192 |
upon | -0.192 |
sister | -0.165 |
spine | -0.158 |
curtain | -0.157 |
Token | . |
Token ID | 13 |
Feature activation | -1.79e-01 |
Pos logit contributions | |
otion | +0.063 |
upon | +0.058 |
laptop | +0.057 |
spine | +0.056 |
sister | +0.054 |
Neg logit contributions | |
cream | -0.068 |
Fin | -0.061 |
Blue | -0.060 |
oc | -0.057 |
Red | -0.056 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.74e+00 |
Pos logit contributions | |
otion | +0.011 |
upon | +0.010 |
sister | +0.009 |
Deal | +0.009 |
Lee | +0.008 |
Neg logit contributions | |
cream | -0.010 |
Fin | -0.009 |
Blue | -0.009 |
rob | -0.009 |
bo | -0.009 |
Token | Once |
Token ID | 7454 |
Feature activation | +7.81e-01 |
Pos logit contributions | |
otion | +0.216 |
upon | +0.201 |
laptop | +0.195 |
Deal | +0.189 |
sister | +0.189 |
Neg logit contributions | |
cream | -0.116 |
Fin | -0.100 |
Blue | -0.098 |
Jerry | -0.087 |
Red | -0.087 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.89e+01 |
Pos logit contributions | |
Fin | +0.057 |
The | +0.056 |
He | +0.056 |
cream | +0.056 |
Blue | +0.056 |
Neg logit contributions | |
laptop | -0.027 |
otion | -0.026 |
Deal | -0.025 |
scalp | -0.024 |
spine | -0.023 |
Token | a |
Token ID | 257 |
Feature activation | -1.83e+00 |
Pos logit contributions | |
otion | +1.232 |
upon | +1.117 |
laptop | +1.099 |
sister | +1.059 |
Deal | +1.058 |
Neg logit contributions | |
cream | -1.032 |
Fin | -0.939 |
Blue | -0.923 |
Jerry | -0.861 |
oc | -0.853 |
Token | time |
Token ID | 640 |
Feature activation | -5.92e+00 |
Pos logit contributions | |
otion | +0.097 |
laptop | +0.083 |
upon | +0.080 |
sister | +0.079 |
ancer | +0.078 |
Neg logit contributions | |
cream | -0.121 |
Fin | -0.117 |
Blue | -0.116 |
Jerry | -0.111 |
oc | -0.109 |
Token | , |
Token ID | 11 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
laptop | +0.458 |
otion | +0.437 |
acy | +0.428 |
spine | +0.426 |
forehead | +0.423 |
Neg logit contributions | |
Fin | -0.214 |
Bob | -0.206 |
Blue | -0.198 |
Jerry | -0.194 |
One | -0.185 |
Token | there |
Token ID | 612 |
Feature activation | -1.87e+00 |
Pos logit contributions | |
Blue | +0.008 |
Fin | +0.007 |
cream | +0.007 |
The | +0.007 |
He | +0.007 |
Neg logit contributions | |
laptop | -0.013 |
otion | -0.013 |
spine | -0.012 |
headache | -0.012 |
Deal | -0.012 |
Token | was |
Token ID | 373 |
Feature activation | -4.73e-01 |
Pos logit contributions | |
otion | +0.163 |
laptop | +0.152 |
Deal | +0.149 |
upon | +0.149 |
spine | +0.148 |
Neg logit contributions | |
cream | -0.061 |
Fin | -0.053 |
Blue | -0.051 |
Jerry | -0.045 |
oc | -0.045 |
Token | free |
Token ID | 1479 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
cream | +0.182 |
Blue | +0.174 |
Fin | +0.174 |
Jerry | +0.164 |
Red | +0.162 |
Neg logit contributions | |
otion | -0.219 |
upon | -0.211 |
sister | -0.202 |
laptop | -0.202 |
Deal | -0.194 |
Token | bread |
Token ID | 8509 |
Feature activation | +3.79e+00 |
Pos logit contributions | |
cream | +0.103 |
Fin | +0.096 |
Blue | +0.094 |
Jerry | +0.089 |
Red | +0.087 |
Neg logit contributions | |
otion | -0.129 |
upon | -0.119 |
laptop | -0.117 |
sister | -0.114 |
headache | -0.112 |
Token | !" |
Token ID | 2474 |
Feature activation | +2.22e+00 |
Pos logit contributions | |
cream | +0.201 |
Fin | +0.183 |
Blue | +0.182 |
Jerry | +0.170 |
Red | +0.169 |
Neg logit contributions | |
otion | -0.252 |
upon | -0.229 |
laptop | -0.227 |
sister | -0.220 |
Deal | -0.217 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.41e+00 |
Pos logit contributions | |
cream | +0.110 |
Blue | +0.097 |
Fin | +0.093 |
Jerry | +0.089 |
sav | +0.087 |
Neg logit contributions | |
otion | -0.162 |
upon | -0.145 |
laptop | -0.144 |
Deal | -0.141 |
spine | -0.139 |
Token | Once |
Token ID | 7454 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
otion | +0.191 |
upon | +0.177 |
laptop | +0.173 |
Deal | +0.168 |
sister | +0.167 |
Neg logit contributions | |
cream | -0.099 |
Fin | -0.087 |
Blue | -0.085 |
Jerry | -0.076 |
Tweet | -0.076 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.85e+01 |
Pos logit contributions | |
He | +0.094 |
The | +0.093 |
One | +0.091 |
Fin | +0.091 |
cream | +0.091 |
Neg logit contributions | |
laptop | -0.036 |
otion | -0.033 |
Deal | -0.029 |
spine | -0.029 |
scalp | -0.028 |
Token | a |
Token ID | 257 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
otion | +1.175 |
upon | +1.076 |
laptop | +1.046 |
sister | +1.012 |
Deal | +1.006 |
Neg logit contributions | |
cream | -1.042 |
Fin | -0.954 |
Blue | -0.931 |
Jerry | -0.872 |
oc | -0.872 |
Token | time |
Token ID | 640 |
Feature activation | -5.66e+00 |
Pos logit contributions | |
otion | +0.084 |
laptop | +0.071 |
upon | +0.068 |
ancer | +0.068 |
ids | +0.067 |
Neg logit contributions | |
cream | -0.096 |
Fin | -0.094 |
Blue | -0.093 |
Jerry | -0.090 |
oc | -0.088 |
Token | , |
Token ID | 11 |
Feature activation | +4.63e-01 |
Pos logit contributions | |
laptop | +0.433 |
otion | +0.412 |
spine | +0.407 |
scalp | +0.402 |
ids | +0.400 |
Neg logit contributions | |
Fin | -0.200 |
Blue | -0.198 |
Bob | -0.193 |
Jerry | -0.188 |
oc | -0.183 |
Token | there |
Token ID | 612 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
Blue | +0.022 |
Fin | +0.021 |
cream | +0.021 |
The | +0.020 |
He | +0.020 |
Neg logit contributions | |
laptop | -0.031 |
otion | -0.030 |
spine | -0.029 |
headache | -0.029 |
Deal | -0.029 |
Token | was |
Token ID | 373 |
Feature activation | -1.52e-01 |
Pos logit contributions | |
otion | +0.130 |
upon | +0.120 |
laptop | +0.118 |
sister | +0.115 |
Deal | +0.114 |
Neg logit contributions | |
cream | -0.061 |
Fin | -0.053 |
Blue | -0.052 |
Jerry | -0.047 |
Red | -0.046 |
Token | clever |
Token ID | 14169 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
cream | +0.039 |
Fin | +0.033 |
oc | +0.033 |
sav | +0.033 |
Blue | +0.033 |
Neg logit contributions | |
otion | -0.051 |
ids | -0.042 |
upon | -0.042 |
ancer | -0.042 |
Deal | -0.041 |
Token | plan |
Token ID | 1410 |
Feature activation | -4.51e+00 |
Pos logit contributions | |
cream | +0.120 |
Fin | +0.109 |
oc | +0.107 |
Blue | +0.105 |
sav | +0.104 |
Neg logit contributions | |
otion | -0.142 |
Deal | -0.119 |
upon | -0.117 |
laptop | -0.114 |
ids | -0.114 |
Token | . |
Token ID | 13 |
Feature activation | -8.40e-02 |
Pos logit contributions | |
otion | +0.310 |
laptop | +0.277 |
upon | +0.276 |
Deal | +0.271 |
headache | +0.267 |
Neg logit contributions | |
cream | -0.233 |
Fin | -0.209 |
Blue | -0.207 |
Jerry | -0.192 |
oc | -0.190 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.91e+00 |
Pos logit contributions | |
otion | +0.006 |
upon | +0.006 |
laptop | +0.006 |
Deal | +0.006 |
sister | +0.006 |
Neg logit contributions | |
cream | -0.004 |
Fin | -0.003 |
Blue | -0.003 |
Jerry | -0.003 |
Tweet | -0.003 |
Token | Once |
Token ID | 7454 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
otion | +0.154 |
laptop | +0.142 |
upon | +0.139 |
spine | +0.137 |
Deal | +0.137 |
Neg logit contributions | |
cream | -0.076 |
Fin | -0.067 |
Blue | -0.066 |
Jerry | -0.060 |
oc | -0.059 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.81e+01 |
Pos logit contributions | |
He | +0.134 |
Fin | +0.133 |
Blue | +0.131 |
The | +0.131 |
cream | +0.130 |
Neg logit contributions | |
laptop | -0.046 |
otion | -0.042 |
Deal | -0.039 |
scalp | -0.034 |
spine | -0.034 |
Token | a |
Token ID | 257 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
otion | +1.057 |
upon | +0.997 |
laptop | +0.932 |
sister | +0.914 |
Deal | +0.899 |
Neg logit contributions | |
cream | -1.089 |
Fin | -0.995 |
Blue | -0.961 |
oc | -0.932 |
Jerry | -0.909 |
Token | time |
Token ID | 640 |
Feature activation | -5.30e+00 |
Pos logit contributions | |
otion | +0.063 |
laptop | +0.053 |
upon | +0.051 |
spine | +0.049 |
sister | +0.049 |
Neg logit contributions | |
cream | -0.081 |
Fin | -0.079 |
Blue | -0.078 |
Jerry | -0.075 |
oc | -0.074 |
Token | there |
Token ID | 612 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
laptop | +0.403 |
otion | +0.382 |
headache | +0.375 |
spine | +0.370 |
Deal | +0.368 |
Neg logit contributions | |
Fin | -0.207 |
Blue | -0.203 |
Bob | -0.195 |
Jerry | -0.185 |
Tweet | -0.181 |
Token | was |
Token ID | 373 |
Feature activation | +1.58e-01 |
Pos logit contributions | |
otion | +0.117 |
laptop | +0.110 |
Deal | +0.109 |
spine | +0.107 |
headache | +0.106 |
Neg logit contributions | |
cream | -0.040 |
Fin | -0.035 |
Blue | -0.034 |
oc | -0.030 |
Jerry | -0.029 |
Token | a |
Token ID | 257 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
cream | +0.009 |
Fin | +0.008 |
Blue | +0.008 |
Jerry | +0.007 |
oc | +0.007 |
Neg logit contributions | |
otion | -0.011 |
upon | -0.009 |
laptop | -0.009 |
Deal | -0.009 |
sister | -0.009 |