Token | , |
Token ID | 11 |
Feature activation | +9.54e-02 |
Pos logit contributions | |
whispered | +0.122 |
chy | +0.116 |
l | +0.109 |
woke | +0.107 |
said | +0.106 |
Neg logit contributions | |
unison | -0.174 |
ric | -0.166 |
latable | -0.165 |
Deal | -0.163 |
yeah | -0.160 |
Token | you |
Token ID | 345 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
said | +0.122 |
whispered | +0.120 |
woke | +0.112 |
fell | +0.106 |
man | +0.105 |
Neg logit contributions | |
ric | -0.158 |
latable | -0.156 |
unison | -0.156 |
igsaw | -0.146 |
Deal | -0.144 |
Token | can |
Token ID | 460 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
whispered | +0.123 |
said | +0.114 |
woke | +0.110 |
fell | +0.109 |
chy | +0.108 |
Neg logit contributions | |
unison | -0.188 |
ric | -0.183 |
latable | -0.181 |
Deal | -0.177 |
igsaw | -0.172 |
Token | keep |
Token ID | 1394 |
Feature activation | +1.55e-01 |
Pos logit contributions | |
said | +0.139 |
whispered | +0.138 |
fell | +0.133 |
woke | +0.130 |
chy | +0.122 |
Neg logit contributions | |
unison | -0.186 |
ric | -0.182 |
latable | -0.175 |
Deal | -0.172 |
igsaw | -0.169 |
Token | the |
Token ID | 262 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
whispered | +0.213 |
said | +0.204 |
fell | +0.188 |
woke | +0.187 |
l | +0.184 |
Neg logit contributions | |
unison | -0.229 |
latable | -0.219 |
ric | -0.217 |
Deal | -0.211 |
igsaw | -0.208 |
Token | coins |
Token ID | 10796 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
whispered | +0.154 |
said | +0.141 |
fell | +0.140 |
chy | +0.139 |
meant | +0.137 |
Neg logit contributions | |
unison | -0.154 |
latable | -0.150 |
ric | -0.148 |
Deal | -0.141 |
yeah | -0.137 |
Token | . |
Token ID | 13 |
Feature activation | +3.31e-02 |
Pos logit contributions | |
whispered | +0.211 |
said | +0.190 |
woke | +0.187 |
fell | +0.180 |
chy | +0.179 |
Neg logit contributions | |
unison | -0.425 |
ric | -0.419 |
latable | -0.412 |
Deal | -0.401 |
igsaw | -0.396 |
Token | You |
Token ID | 921 |
Feature activation | +9.19e-02 |
Pos logit contributions | |
whispered | +0.036 |
said | +0.035 |
meant | +0.033 |
woke | +0.033 |
fell | +0.030 |
Neg logit contributions | |
ric | -0.055 |
latable | -0.054 |
unison | -0.052 |
igsaw | -0.051 |
pieces | -0.047 |
Token | can |
Token ID | 460 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
whispered | +0.109 |
said | +0.100 |
l | +0.099 |
chy | +0.098 |
woke | +0.098 |
Neg logit contributions | |
unison | -0.147 |
ric | -0.141 |
latable | -0.139 |
Deal | -0.138 |
igsaw | -0.132 |
Token | buy |
Token ID | 2822 |
Feature activation | +1.61e-01 |
Pos logit contributions | |
whispered | +0.152 |
said | +0.147 |
fell | +0.140 |
woke | +0.138 |
chy | +0.136 |
Neg logit contributions | |
unison | -0.176 |
ric | -0.170 |
latable | -0.169 |
igsaw | -0.165 |
Deal | -0.162 |
Token | something |
Token ID | 1223 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
whispered | +0.245 |
chy | +0.232 |
woke | +0.230 |
fell | +0.229 |
said | +0.226 |
Neg logit contributions | |
unison | -0.204 |
ric | -0.198 |
latable | -0.193 |
Deal | -0.184 |
igsaw | -0.178 |
Token | Everyone |
Token ID | 11075 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
whispered | +0.024 |
said | +0.024 |
woke | +0.021 |
fell | +0.021 |
meant | +0.020 |
Neg logit contributions | |
unison | -0.034 |
ric | -0.034 |
latable | -0.033 |
Deal | -0.033 |
igsaw | -0.032 |
Token | wanted |
Token ID | 2227 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
chy | +0.088 |
whispered | +0.082 |
l | +0.081 |
prick | +0.080 |
low | +0.075 |
Neg logit contributions | |
unison | -0.203 |
ric | -0.196 |
latable | -0.196 |
teammates | -0.188 |
yeah | -0.188 |
Token | to |
Token ID | 284 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
whispered | +0.186 |
chy | +0.181 |
fell | +0.172 |
l | +0.172 |
woke | +0.169 |
Neg logit contributions | |
unison | -0.148 |
latable | -0.144 |
ric | -0.140 |
igsaw | -0.133 |
yeah | -0.132 |
Token | share |
Token ID | 2648 |
Feature activation | +1.65e-01 |
Pos logit contributions | |
said | +0.153 |
whispered | +0.149 |
woke | +0.135 |
spark | +0.133 |
fell | +0.132 |
Neg logit contributions | |
ric | -0.248 |
latable | -0.222 |
unison | -0.217 |
ifying | -0.216 |
Deal | -0.215 |
Token | the |
Token ID | 262 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
whispered | +0.200 |
chy | +0.192 |
woke | +0.184 |
fell | +0.183 |
said | +0.180 |
Neg logit contributions | |
unison | -0.262 |
ric | -0.249 |
latable | -0.245 |
Deal | -0.233 |
igsaw | -0.232 |
Token | treasure |
Token ID | 14068 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
whispered | +0.182 |
fell | +0.172 |
chy | +0.171 |
said | +0.169 |
l | +0.165 |
Neg logit contributions | |
latable | -0.120 |
unison | -0.118 |
igsaw | -0.118 |
yeah | -0.110 |
ric | -0.110 |
Token | together |
Token ID | 1978 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
whispered | +0.228 |
woke | +0.211 |
chy | +0.203 |
said | +0.202 |
fell | +0.200 |
Neg logit contributions | |
unison | -0.402 |
latable | -0.390 |
igsaw | -0.384 |
ric | -0.384 |
Deal | -0.365 |
Token | . |
Token ID | 13 |
Feature activation | +2.88e-02 |
Pos logit contributions | |
whispered | +0.151 |
chy | +0.140 |
woke | +0.140 |
l | +0.137 |
fell | +0.132 |
Neg logit contributions | |
unison | -0.232 |
ric | -0.210 |
latable | -0.209 |
igsaw | -0.208 |
teammates | -0.207 |
Token | Now |
Token ID | 2735 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
whispered | +0.035 |
said | +0.034 |
woke | +0.032 |
fell | +0.031 |
meant | +0.030 |
Neg logit contributions | |
unison | -0.047 |
ric | -0.046 |
latable | -0.045 |
Deal | -0.044 |
igsaw | -0.043 |
Token | , |
Token ID | 11 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
whispered | +0.116 |
said | +0.103 |
chy | +0.102 |
woke | +0.100 |
l | +0.099 |
Neg logit contributions | |
unison | -0.209 |
latable | -0.203 |
ric | -0.202 |
Deal | -0.195 |
igsaw | -0.193 |
Token | the |
Token ID | 262 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
whispered | +0.127 |
chy | +0.116 |
said | +0.114 |
woke | +0.113 |
l | +0.113 |
Neg logit contributions | |
unison | -0.154 |
ric | -0.151 |
latable | -0.145 |
Deal | -0.142 |
igsaw | -0.139 |
Token | They |
Token ID | 1119 |
Feature activation | +8.97e-02 |
Pos logit contributions | |
chy | +0.019 |
whispered | +0.019 |
woke | +0.017 |
meant | +0.017 |
prick | +0.017 |
Neg logit contributions | |
unison | -0.017 |
Deal | -0.016 |
latable | -0.016 |
ric | -0.015 |
yeah | -0.015 |
Token | paid |
Token ID | 3432 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
chy | +0.080 |
l | +0.073 |
prick | +0.072 |
low | +0.068 |
whispered | +0.067 |
Neg logit contributions | |
unison | -0.162 |
latable | -0.145 |
ric | -0.141 |
teammates | -0.140 |
yeah | -0.136 |
Token | him |
Token ID | 683 |
Feature activation | +1.55e-01 |
Pos logit contributions | |
whispered | +0.202 |
woke | +0.192 |
chy | +0.190 |
fell | +0.184 |
said | +0.181 |
Neg logit contributions | |
unison | -0.200 |
ric | -0.189 |
latable | -0.187 |
Deal | -0.184 |
igsaw | -0.176 |
Token | with |
Token ID | 351 |
Feature activation | +1.20e-01 |
Pos logit contributions | |
whispered | +0.200 |
woke | +0.194 |
chy | +0.182 |
fell | +0.182 |
prick | +0.174 |
Neg logit contributions | |
ric | -0.218 |
unison | -0.215 |
latable | -0.214 |
ap | -0.201 |
Deal | -0.195 |
Token | some |
Token ID | 617 |
Feature activation | +1.59e-01 |
Pos logit contributions | |
whispered | +0.176 |
woke | +0.165 |
chy | +0.163 |
said | +0.162 |
fell | +0.161 |
Neg logit contributions | |
unison | -0.164 |
ric | -0.155 |
latable | -0.155 |
Deal | -0.146 |
igsaw | -0.141 |
Token | coins |
Token ID | 10796 |
Feature activation | +2.27e-01 |
Pos logit contributions | |
whispered | +0.208 |
chy | +0.195 |
woke | +0.195 |
fell | +0.185 |
said | +0.179 |
Neg logit contributions | |
unison | -0.234 |
latable | -0.226 |
ric | -0.221 |
Deal | -0.219 |
yeah | -0.209 |
Token | they |
Token ID | 484 |
Feature activation | +1.22e-01 |
Pos logit contributions | |
whispered | +0.207 |
woke | +0.195 |
chy | +0.183 |
said | +0.176 |
fell | +0.174 |
Neg logit contributions | |
unison | -0.408 |
ric | -0.406 |
latable | -0.398 |
igsaw | -0.377 |
ifying | -0.373 |
Token | had |
Token ID | 550 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
aisle | +0.115 |
l | +0.109 |
chy | +0.107 |
iv | +0.106 |
prick | +0.104 |
Neg logit contributions | |
unison | -0.206 |
latable | -0.181 |
ifying | -0.180 |
teammates | -0.171 |
ric | -0.171 |
Token | . |
Token ID | 13 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
chy | +0.117 |
woke | +0.108 |
whispered | +0.108 |
aisle | +0.106 |
l | +0.104 |
Neg logit contributions | |
unison | -0.155 |
ric | -0.148 |
latable | -0.141 |
such | -0.134 |
gonna | -0.132 |
Token |
|
Token ID | 198 |
Feature activation | +1.73e-02 |
Pos logit contributions | |
chy | +0.025 |
whispered | +0.025 |
woke | +0.023 |
fell | +0.023 |
prick | +0.022 |
Neg logit contributions | |
unison | -0.028 |
latable | -0.026 |
Deal | -0.026 |
ric | -0.026 |
yeah | -0.024 |
Token |
|
Token ID | 198 |
Feature activation | +1.50e-02 |
Pos logit contributions | |
chy | +0.026 |
whispered | +0.025 |
meant | +0.023 |
woke | +0.023 |
prick | +0.023 |
Neg logit contributions | |
unison | -0.022 |
ric | -0.021 |
Deal | -0.020 |
latable | -0.020 |
teammates | -0.018 |
Token | hurried |
Token ID | 31662 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
whispered | +0.073 |
chy | +0.070 |
l | +0.066 |
woke | +0.062 |
meant | +0.061 |
Neg logit contributions | |
unison | -0.186 |
latable | -0.178 |
ric | -0.176 |
igsaw | -0.170 |
Deal | -0.170 |
Token | away |
Token ID | 1497 |
Feature activation | +1.42e-01 |
Pos logit contributions | |
whispered | +0.173 |
said | +0.166 |
woke | +0.160 |
fell | +0.155 |
l | +0.148 |
Neg logit contributions | |
unison | -0.308 |
ric | -0.308 |
latable | -0.290 |
Deal | -0.289 |
yeah | -0.273 |
Token | with |
Token ID | 351 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
chy | +0.166 |
whispered | +0.164 |
woke | +0.157 |
meant | +0.151 |
l | +0.148 |
Neg logit contributions | |
unison | -0.210 |
igsaw | -0.200 |
latable | -0.191 |
teammates | -0.190 |
ric | -0.185 |
Token | his |
Token ID | 465 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
whispered | +0.109 |
said | +0.108 |
fell | +0.095 |
woke | +0.093 |
meant | +0.092 |
Neg logit contributions | |
ric | -0.229 |
unison | -0.207 |
latable | -0.207 |
Deal | -0.203 |
igsaw | -0.196 |
Token | borrowed |
Token ID | 22546 |
Feature activation | +1.85e-01 |
Pos logit contributions | |
whispered | +0.144 |
said | +0.135 |
fell | +0.127 |
woke | +0.127 |
chy | +0.124 |
Neg logit contributions | |
unison | -0.210 |
ric | -0.207 |
latable | -0.203 |
Deal | -0.197 |
igsaw | -0.193 |
Token | hook |
Token ID | 8011 |
Feature activation | +2.27e-01 |
Pos logit contributions | |
whispered | +0.195 |
said | +0.183 |
fell | +0.175 |
woke | +0.172 |
l | +0.170 |
Neg logit contributions | |
unison | -0.329 |
ric | -0.328 |
Deal | -0.320 |
latable | -0.316 |
igsaw | -0.306 |
Token | . |
Token ID | 13 |
Feature activation | +2.33e-02 |
Pos logit contributions | |
whispered | +0.215 |
woke | +0.208 |
chy | +0.187 |
meant | +0.186 |
said | +0.183 |
Neg logit contributions | |
unison | -0.386 |
ric | -0.368 |
latable | -0.367 |
igsaw | -0.364 |
teammates | -0.349 |
Token | From |
Token ID | 3574 |
Feature activation | +7.62e-02 |
Pos logit contributions | |
whispered | +0.030 |
said | +0.029 |
woke | +0.028 |
fell | +0.028 |
chy | +0.027 |
Neg logit contributions | |
unison | -0.036 |
ric | -0.035 |
latable | -0.034 |
Deal | -0.033 |
igsaw | -0.032 |
Token | that |
Token ID | 326 |
Feature activation | +9.80e-02 |
Pos logit contributions | |
whispered | +0.104 |
said | +0.102 |
fell | +0.093 |
woke | +0.091 |
l | +0.090 |
Neg logit contributions | |
unison | -0.117 |
ric | -0.115 |
Deal | -0.113 |
latable | -0.111 |
igsaw | -0.106 |
Token | day |
Token ID | 1110 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
whispered | +0.115 |
l | +0.106 |
iv | +0.093 |
meant | +0.092 |
shouted | +0.090 |
Neg logit contributions | |
igsaw | -0.150 |
Deal | -0.139 |
yeah | -0.139 |
ap | -0.139 |
Wed | -0.138 |
Token | forward |
Token ID | 2651 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
whispered | +0.132 |
l | +0.117 |
chy | +0.116 |
said | +0.113 |
prick | +0.108 |
Neg logit contributions | |
Deal | -0.242 |
unison | -0.237 |
ap | -0.232 |
ric | -0.227 |
latable | -0.226 |
Token | should |
Token ID | 815 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
whispered | +0.078 |
aisle | +0.068 |
chy | +0.066 |
woke | +0.065 |
l | +0.065 |
Neg logit contributions | |
unison | -0.190 |
ric | -0.185 |
latable | -0.184 |
Deal | -0.179 |
yeah | -0.173 |
Token | not |
Token ID | 407 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
said | +0.097 |
hurt | +0.085 |
had | +0.079 |
fell | +0.078 |
whispered | +0.071 |
Neg logit contributions | |
ifying | -0.257 |
unison | -0.252 |
igsaw | -0.241 |
latable | -0.239 |
ric | -0.233 |
Token | have |
Token ID | 423 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
whispered | +0.135 |
said | +0.134 |
fell | +0.126 |
woke | +0.125 |
meant | +0.123 |
Neg logit contributions | |
unison | -0.170 |
ric | -0.164 |
latable | -0.162 |
Deal | -0.159 |
igsaw | -0.156 |
Token | taken |
Token ID | 2077 |
Feature activation | +1.71e-01 |
Pos logit contributions | |
whispered | +0.129 |
said | +0.120 |
woke | +0.114 |
fell | +0.113 |
chy | +0.111 |
Neg logit contributions | |
unison | -0.221 |
ric | -0.216 |
latable | -0.215 |
Deal | -0.209 |
igsaw | -0.205 |
Token | your |
Token ID | 534 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
whispered | +0.239 |
said | +0.236 |
l | +0.219 |
fell | +0.218 |
woke | +0.215 |
Neg logit contributions | |
unison | -0.251 |
latable | -0.247 |
ric | -0.243 |
Deal | -0.229 |
igsaw | -0.219 |
Token | sack |
Token ID | 23704 |
Feature activation | +2.26e-01 |
Pos logit contributions | |
whispered | +0.201 |
said | +0.188 |
woke | +0.185 |
fell | +0.181 |
chy | +0.181 |
Neg logit contributions | |
unison | -0.209 |
ric | -0.200 |
latable | -0.200 |
Deal | -0.194 |
igsaw | -0.192 |
Token | . |
Token ID | 13 |
Feature activation | +4.05e-02 |
Pos logit contributions | |
whispered | +0.218 |
woke | +0.189 |
said | +0.183 |
spoke | +0.177 |
meant | +0.172 |
Neg logit contributions | |
unison | -0.399 |
igsaw | -0.389 |
latable | -0.386 |
teammates | -0.376 |
Deal | -0.375 |
Token | I |
Token ID | 314 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
whispered | +0.052 |
said | +0.052 |
woke | +0.050 |
fell | +0.049 |
meant | +0.049 |
Neg logit contributions | |
ric | -0.065 |
unison | -0.064 |
latable | -0.062 |
pieces | -0.056 |
igsaw | -0.055 |
Token | was |
Token ID | 373 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
whispered | +0.100 |
woke | +0.086 |
chy | +0.086 |
said | +0.086 |
aisle | +0.085 |
Neg logit contributions | |
unison | -0.193 |
ric | -0.189 |
latable | -0.187 |
Deal | -0.182 |
yeah | -0.175 |
Token | curious |
Token ID | 11040 |
Feature activation | +1.47e-01 |
Pos logit contributions | |
said | +0.090 |
whispered | +0.085 |
hurt | +0.078 |
sm | +0.070 |
scared | +0.070 |
Neg logit contributions | |
unison | -0.295 |
igsaw | -0.283 |
latable | -0.275 |
ric | -0.266 |
Deal | -0.261 |
Token | . |
Token ID | 13 |
Feature activation | +4.27e-02 |
Pos logit contributions | |
whispered | +0.136 |
chy | +0.115 |
said | +0.112 |
l | +0.110 |
fell | +0.110 |
Neg logit contributions | |
unison | -0.264 |
latable | -0.250 |
Deal | -0.249 |
ric | -0.248 |
yeah | -0.237 |
Token | 'm |
Token ID | 1101 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
aisle | +0.165 |
whispered | +0.157 |
low | +0.148 |
l | +0.147 |
floor | +0.143 |
Neg logit contributions | |
unison | -0.109 |
ric | -0.109 |
Deal | -0.103 |
latable | -0.102 |
yeah | -0.094 |
Token | sorry |
Token ID | 7926 |
Feature activation | +1.29e-01 |
Pos logit contributions | |
whispered | +0.163 |
said | +0.161 |
fell | +0.146 |
chy | +0.140 |
meant | +0.140 |
Neg logit contributions | |
unison | -0.181 |
latable | -0.168 |
igsaw | -0.168 |
ric | -0.166 |
ifying | -0.155 |
Token | I |
Token ID | 314 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
said | +0.127 |
whispered | +0.126 |
fell | +0.119 |
l | +0.116 |
chy | +0.116 |
Neg logit contributions | |
unison | -0.245 |
latable | -0.239 |
ric | -0.235 |
igsaw | -0.230 |
Deal | -0.216 |
Token | dropped |
Token ID | 5710 |
Feature activation | +1.74e-01 |
Pos logit contributions | |
aisle | +0.126 |
low | +0.109 |
floor | +0.100 |
man | +0.096 |
whispered | +0.094 |
Neg logit contributions | |
ric | -0.168 |
unison | -0.159 |
hey | -0.157 |
latable | -0.156 |
disl | -0.154 |
Token | the |
Token ID | 262 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
whispered | +0.235 |
said | +0.215 |
woke | +0.214 |
chy | +0.212 |
fell | +0.210 |
Neg logit contributions | |
unison | -0.249 |
ric | -0.246 |
latable | -0.238 |
Deal | -0.234 |
igsaw | -0.228 |
Token | bag |
Token ID | 6131 |
Feature activation | +2.26e-01 |
Pos logit contributions | |
whispered | +0.193 |
said | +0.176 |
woke | +0.175 |
fell | +0.172 |
chy | +0.171 |
Neg logit contributions | |
ric | -0.143 |
unison | -0.142 |
latable | -0.141 |
Deal | -0.136 |
igsaw | -0.131 |
Token | and |
Token ID | 290 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
whispered | +0.293 |
spoke | +0.275 |
buzz | +0.268 |
woke | +0.261 |
breathed | +0.255 |
Neg logit contributions | |
ap | -0.275 |
ifying | -0.271 |
unison | -0.259 |
such | -0.256 |
our | -0.255 |
Token | the |
Token ID | 262 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
whispered | +0.134 |
said | +0.132 |
fell | +0.127 |
woke | +0.124 |
needed | +0.120 |
Neg logit contributions | |
unison | -0.214 |
latable | -0.207 |
ric | -0.201 |
igsaw | -0.197 |
Deal | -0.196 |
Token | gum |
Token ID | 27142 |
Feature activation | +2.10e-01 |
Pos logit contributions | |
whispered | +0.163 |
said | +0.146 |
woke | +0.145 |
fell | +0.141 |
chy | +0.141 |
Neg logit contributions | |
ric | -0.164 |
unison | -0.164 |
latable | -0.155 |
Deal | -0.155 |
igsaw | -0.146 |
Token | ." |
Token ID | 526 |
Feature activation | +3.41e-02 |
Pos logit contributions | |
whispered | +0.250 |
spoke | +0.230 |
meant | +0.211 |
woke | +0.211 |
prick | +0.207 |
Neg logit contributions | |
teammates | -0.291 |
unison | -0.277 |
ap | -0.273 |
goalie | -0.268 |
our | -0.265 |
Token |
|
Token ID | 198 |
Feature activation | +1.83e-02 |
Pos logit contributions | |
whispered | +0.029 |
said | +0.027 |
woke | +0.025 |
fell | +0.025 |
l | +0.024 |
Neg logit contributions | |
unison | -0.068 |
ric | -0.067 |
latable | -0.066 |
Deal | -0.064 |
igsaw | -0.063 |
Token | okay |
Token ID | 8788 |
Feature activation | +1.11e-01 |
Pos logit contributions | |
whispered | +0.144 |
said | +0.138 |
meant | +0.133 |
chy | +0.132 |
fell | +0.132 |
Neg logit contributions | |
unison | -0.150 |
ric | -0.147 |
latable | -0.145 |
igsaw | -0.137 |
Deal | -0.136 |
Token | if |
Token ID | 611 |
Feature activation | +1.11e-01 |
Pos logit contributions | |
said | +0.064 |
fell | +0.057 |
woke | +0.057 |
whispered | +0.055 |
needed | +0.054 |
Neg logit contributions | |
ric | -0.260 |
unison | -0.257 |
latable | -0.253 |
Deal | -0.240 |
ifying | -0.238 |
Token | you |
Token ID | 345 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
whispered | +0.185 |
said | +0.177 |
woke | +0.172 |
fell | +0.172 |
chy | +0.169 |
Neg logit contributions | |
unison | -0.129 |
ric | -0.126 |
latable | -0.124 |
Deal | -0.118 |
igsaw | -0.114 |
Token | keep |
Token ID | 1394 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
whispered | +0.165 |
said | +0.155 |
woke | +0.152 |
fell | +0.150 |
l | +0.150 |
Neg logit contributions | |
unison | -0.153 |
ric | -0.147 |
latable | -0.145 |
Deal | -0.140 |
igsaw | -0.136 |
Token | the |
Token ID | 262 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
whispered | +0.189 |
said | +0.184 |
fell | +0.161 |
meant | +0.158 |
needed | +0.156 |
Neg logit contributions | |
unison | -0.258 |
latable | -0.252 |
ric | -0.247 |
Deal | -0.238 |
igsaw | -0.229 |
Token | treasure |
Token ID | 14068 |
Feature activation | +2.26e-01 |
Pos logit contributions | |
whispered | +0.183 |
said | +0.172 |
fell | +0.170 |
chy | +0.169 |
l | +0.167 |
Neg logit contributions | |
unison | -0.116 |
latable | -0.112 |
ric | -0.110 |
Deal | -0.108 |
igsaw | -0.108 |
Token | . |
Token ID | 13 |
Feature activation | +3.64e-02 |
Pos logit contributions | |
whispered | +0.244 |
woke | +0.222 |
said | +0.218 |
fell | +0.209 |
l | +0.209 |
Neg logit contributions | |
unison | -0.370 |
igsaw | -0.355 |
latable | -0.352 |
ric | -0.351 |
Deal | -0.344 |
Token | But |
Token ID | 887 |
Feature activation | +7.78e-02 |
Pos logit contributions | |
said | +0.042 |
whispered | +0.041 |
meant | +0.039 |
needed | +0.037 |
woke | +0.036 |
Neg logit contributions | |
latable | -0.057 |
ric | -0.056 |
unison | -0.055 |
igsaw | -0.054 |
pieces | -0.050 |
Token | if |
Token ID | 611 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
whispered | +0.105 |
said | +0.103 |
woke | +0.096 |
fell | +0.096 |
meant | +0.095 |
Neg logit contributions | |
unison | -0.116 |
latable | -0.113 |
ric | -0.111 |
igsaw | -0.106 |
Deal | -0.105 |
Token | you |
Token ID | 345 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
said | +0.165 |
whispered | +0.163 |
fell | +0.157 |
woke | +0.154 |
needed | +0.152 |
Neg logit contributions | |
ric | -0.170 |
unison | -0.166 |
latable | -0.164 |
Deal | -0.153 |
igsaw | -0.148 |
Token | don |
Token ID | 836 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
whispered | +0.161 |
said | +0.152 |
woke | +0.147 |
fell | +0.146 |
chy | +0.145 |
Neg logit contributions | |
unison | -0.163 |
ric | -0.158 |
latable | -0.156 |
Deal | -0.151 |
igsaw | -0.147 |
Token | okay |
Token ID | 8788 |
Feature activation | +1.11e-01 |
Pos logit contributions | |
whispered | +0.144 |
said | +0.138 |
meant | +0.133 |
chy | +0.132 |
fell | +0.132 |
Neg logit contributions | |
unison | -0.150 |
ric | -0.147 |
latable | -0.145 |
igsaw | -0.137 |
Deal | -0.136 |
Token | if |
Token ID | 611 |
Feature activation | +1.11e-01 |
Pos logit contributions | |
said | +0.064 |
fell | +0.057 |
woke | +0.057 |
whispered | +0.055 |
needed | +0.054 |
Neg logit contributions | |
ric | -0.260 |
unison | -0.257 |
latable | -0.253 |
Deal | -0.240 |
ifying | -0.238 |
Token | you |
Token ID | 345 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
whispered | +0.185 |
said | +0.177 |
woke | +0.172 |
fell | +0.172 |
chy | +0.169 |
Neg logit contributions | |
unison | -0.129 |
ric | -0.126 |
latable | -0.124 |
Deal | -0.118 |
igsaw | -0.114 |
Token | keep |
Token ID | 1394 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
whispered | +0.165 |
said | +0.155 |
woke | +0.152 |
fell | +0.150 |
l | +0.150 |
Neg logit contributions | |
unison | -0.153 |
ric | -0.147 |
latable | -0.145 |
Deal | -0.140 |
igsaw | -0.136 |
Token | the |
Token ID | 262 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
whispered | +0.189 |
said | +0.184 |
fell | +0.161 |
meant | +0.158 |
needed | +0.156 |
Neg logit contributions | |
unison | -0.258 |
latable | -0.252 |
ric | -0.247 |
Deal | -0.238 |
igsaw | -0.229 |
Token | treasure |
Token ID | 14068 |
Feature activation | +2.26e-01 |
Pos logit contributions | |
whispered | +0.183 |
said | +0.172 |
fell | +0.170 |
chy | +0.169 |
l | +0.167 |
Neg logit contributions | |
unison | -0.116 |
latable | -0.112 |
ric | -0.110 |
Deal | -0.108 |
igsaw | -0.108 |
Token | . |
Token ID | 13 |
Feature activation | +3.64e-02 |
Pos logit contributions | |
whispered | +0.244 |
woke | +0.222 |
said | +0.218 |
fell | +0.209 |
l | +0.209 |
Neg logit contributions | |
unison | -0.370 |
igsaw | -0.355 |
latable | -0.352 |
ric | -0.351 |
Deal | -0.344 |
Token | But |
Token ID | 887 |
Feature activation | +7.78e-02 |
Pos logit contributions | |
said | +0.042 |
whispered | +0.041 |
meant | +0.039 |
needed | +0.037 |
woke | +0.036 |
Neg logit contributions | |
latable | -0.057 |
ric | -0.056 |
unison | -0.055 |
igsaw | -0.054 |
pieces | -0.050 |
Token | if |
Token ID | 611 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
whispered | +0.105 |
said | +0.103 |
woke | +0.096 |
fell | +0.096 |
meant | +0.095 |
Neg logit contributions | |
unison | -0.116 |
latable | -0.113 |
ric | -0.111 |
igsaw | -0.106 |
Deal | -0.105 |
Token | you |
Token ID | 345 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
said | +0.165 |
whispered | +0.163 |
fell | +0.157 |
woke | +0.154 |
needed | +0.152 |
Neg logit contributions | |
ric | -0.170 |
unison | -0.166 |
latable | -0.164 |
Deal | -0.153 |
igsaw | -0.148 |
Token | don |
Token ID | 836 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
whispered | +0.161 |
said | +0.152 |
woke | +0.147 |
fell | +0.146 |
chy | +0.145 |
Neg logit contributions | |
unison | -0.163 |
ric | -0.158 |
latable | -0.156 |
Deal | -0.151 |
igsaw | -0.147 |
Token | They |
Token ID | 1119 |
Feature activation | +8.83e-02 |
Pos logit contributions | |
whispered | +0.015 |
said | +0.015 |
woke | +0.014 |
fell | +0.013 |
meant | +0.013 |
Neg logit contributions | |
unison | -0.020 |
ric | -0.020 |
latable | -0.020 |
igsaw | -0.019 |
Deal | -0.019 |
Token | do |
Token ID | 466 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
whispered | +0.121 |
l | +0.113 |
woke | +0.112 |
said | +0.111 |
chy | +0.110 |
Neg logit contributions | |
unison | -0.120 |
Deal | -0.115 |
latable | -0.115 |
ric | -0.114 |
hey | -0.103 |
Token | not |
Token ID | 407 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
whispered | +0.135 |
said | +0.131 |
fell | +0.125 |
chy | +0.123 |
woke | +0.122 |
Neg logit contributions | |
unison | -0.200 |
latable | -0.199 |
ric | -0.190 |
igsaw | -0.187 |
ifying | -0.183 |
Token | want |
Token ID | 765 |
Feature activation | +1.26e-01 |
Pos logit contributions | |
whispered | +0.159 |
l | +0.158 |
aisle | +0.158 |
woke | +0.149 |
said | +0.149 |
Neg logit contributions | |
unison | -0.100 |
ric | -0.099 |
latable | -0.097 |
Deal | -0.096 |
ap | -0.091 |
Token | a |
Token ID | 257 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
whispered | +0.186 |
woke | +0.171 |
l | +0.171 |
said | +0.169 |
chy | +0.169 |
Neg logit contributions | |
unison | -0.162 |
latable | -0.157 |
ric | -0.156 |
Deal | -0.145 |
igsaw | -0.144 |
Token | haircut |
Token ID | 45548 |
Feature activation | +2.25e-01 |
Pos logit contributions | |
whispered | +0.215 |
meant | +0.204 |
said | +0.203 |
chy | +0.203 |
woke | +0.198 |
Neg logit contributions | |
unison | -0.096 |
ric | -0.094 |
yeah | -0.093 |
latable | -0.092 |
Deal | -0.092 |
Token | . |
Token ID | 13 |
Feature activation | +2.02e-02 |
Pos logit contributions | |
whispered | +0.201 |
woke | +0.189 |
chy | +0.170 |
prick | +0.170 |
l | +0.168 |
Neg logit contributions | |
unison | -0.407 |
ric | -0.388 |
latable | -0.383 |
yeah | -0.376 |
ap | -0.370 |
Token |
|
Token ID | 198 |
Feature activation | +1.80e-02 |
Pos logit contributions | |
whispered | +0.023 |
said | +0.022 |
woke | +0.020 |
fell | +0.020 |
meant | +0.020 |
Neg logit contributions | |
unison | -0.034 |
ric | -0.034 |
latable | -0.033 |
Deal | -0.032 |
igsaw | -0.032 |
Token |
|
Token ID | 198 |
Feature activation | +1.51e-02 |
Pos logit contributions | |
chy | +0.026 |
whispered | +0.025 |
woke | +0.024 |
prick | +0.023 |
meant | +0.023 |
Neg logit contributions | |
unison | -0.024 |
Deal | -0.023 |
ric | -0.021 |
latable | -0.021 |
ap | -0.020 |
Token | They |
Token ID | 2990 |
Feature activation | +9.33e-02 |
Pos logit contributions | |
whispered | +0.017 |
said | +0.016 |
woke | +0.016 |
fell | +0.015 |
chy | +0.015 |
Neg logit contributions | |
unison | -0.025 |
ric | -0.024 |
latable | -0.024 |
Deal | -0.024 |
igsaw | -0.023 |
Token | see |
Token ID | 766 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
whispered | +0.135 |
said | +0.126 |
l | +0.124 |
woke | +0.124 |
chy | +0.124 |
Neg logit contributions | |
unison | -0.124 |
ric | -0.118 |
latable | -0.117 |
Deal | -0.116 |
igsaw | -0.106 |
Token | friends |
Token ID | 2460 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
whispered | +0.132 |
said | +0.128 |
meant | +0.121 |
fell | +0.117 |
low | +0.114 |
Neg logit contributions | |
unison | -0.174 |
ric | -0.173 |
latable | -0.172 |
igsaw | -0.165 |
Deal | -0.163 |
Token | and |
Token ID | 290 |
Feature activation | +9.98e-02 |
Pos logit contributions | |
whispered | +0.121 |
said | +0.106 |
chy | +0.106 |
woke | +0.104 |
l | +0.103 |
Neg logit contributions | |
unison | -0.279 |
ric | -0.264 |
latable | -0.264 |
Deal | -0.252 |
igsaw | -0.249 |
Token | shared |
Token ID | 4888 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
whispered | +0.112 |
l | +0.105 |
chy | +0.105 |
woke | +0.099 |
said | +0.098 |
Neg logit contributions | |
unison | -0.160 |
ric | -0.154 |
latable | -0.150 |
yeah | -0.143 |
Deal | -0.141 |
Token | the |
Token ID | 262 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
whispered | +0.238 |
chy | +0.235 |
woke | +0.226 |
fell | +0.226 |
meant | +0.217 |
Neg logit contributions | |
unison | -0.185 |
ric | -0.167 |
latable | -0.167 |
igsaw | -0.151 |
Deal | -0.149 |
Token | pretty |
Token ID | 2495 |
Feature activation | +1.60e-01 |
Pos logit contributions | |
whispered | +0.128 |
said | +0.116 |
fell | +0.112 |
meant | +0.108 |
woke | +0.107 |
Neg logit contributions | |
latable | -0.163 |
igsaw | -0.158 |
ric | -0.157 |
Deal | -0.151 |
unison | -0.150 |
Token | rocks |
Token ID | 12586 |
Feature activation | +2.25e-01 |
Pos logit contributions | |
whispered | +0.171 |
said | +0.157 |
woke | +0.151 |
fell | +0.150 |
chy | +0.148 |
Neg logit contributions | |
unison | -0.280 |
ric | -0.275 |
latable | -0.272 |
Deal | -0.263 |
igsaw | -0.259 |
Token | from |
Token ID | 422 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
whispered | +0.191 |
said | +0.173 |
woke | +0.163 |
fell | +0.162 |
chy | +0.160 |
Neg logit contributions | |
unison | -0.442 |
ric | -0.433 |
latable | -0.429 |
Deal | -0.419 |
igsaw | -0.409 |
Token | the |
Token ID | 262 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
whispered | +0.171 |
woke | +0.161 |
said | +0.157 |
fell | +0.156 |
chy | +0.155 |
Neg logit contributions | |
ric | -0.175 |
unison | -0.172 |
latable | -0.163 |
Deal | -0.149 |
hey | -0.144 |
Token | mine |
Token ID | 6164 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
whispered | +0.138 |
said | +0.125 |
l | +0.124 |
meant | +0.123 |
fell | +0.120 |
Neg logit contributions | |
latable | -0.147 |
Deal | -0.145 |
ric | -0.143 |
unison | -0.143 |
yeah | -0.137 |
Token | . |
Token ID | 13 |
Feature activation | +2.12e-02 |
Pos logit contributions | |
whispered | +0.155 |
woke | +0.147 |
meant | +0.137 |
chy | +0.135 |
said | +0.133 |
Neg logit contributions | |
unison | -0.323 |
ric | -0.314 |
teammates | -0.305 |
latable | -0.296 |
igsaw | -0.295 |
Token | They |
Token ID | 1119 |
Feature activation | +9.62e-02 |
Pos logit contributions | |
whispered | +0.034 |
chy | +0.032 |
l | +0.031 |
woke | +0.031 |
fell | +0.031 |
Neg logit contributions | |
unison | -0.026 |
ric | -0.025 |
latable | -0.025 |
Deal | -0.023 |
igsaw | -0.023 |
Token | ! |
Token ID | 0 |
Feature activation | +3.59e-02 |
Pos logit contributions | |
whispered | +0.240 |
said | +0.225 |
woke | +0.216 |
fell | +0.215 |
chy | +0.213 |
Neg logit contributions | |
unison | -0.337 |
ric | -0.327 |
latable | -0.325 |
Deal | -0.315 |
igsaw | -0.308 |
Token | You |
Token ID | 921 |
Feature activation | +9.44e-02 |
Pos logit contributions | |
whispered | +0.056 |
said | +0.054 |
woke | +0.053 |
fell | +0.052 |
chy | +0.051 |
Neg logit contributions | |
unison | -0.044 |
ric | -0.043 |
latable | -0.042 |
igsaw | -0.039 |
Deal | -0.039 |
Token | cannot |
Token ID | 2314 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
whispered | +0.115 |
said | +0.106 |
woke | +0.103 |
l | +0.102 |
chy | +0.102 |
Neg logit contributions | |
unison | -0.148 |
ric | -0.145 |
latable | -0.143 |
Deal | -0.139 |
igsaw | -0.134 |
Token | take |
Token ID | 1011 |
Feature activation | +1.67e-01 |
Pos logit contributions | |
whispered | +0.165 |
said | +0.152 |
l | +0.147 |
meant | +0.144 |
chy | +0.144 |
Neg logit contributions | |
unison | -0.207 |
ric | -0.203 |
latable | -0.202 |
Deal | -0.196 |
igsaw | -0.183 |
Token | my |
Token ID | 616 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
whispered | +0.262 |
chy | +0.250 |
said | +0.240 |
woke | +0.236 |
aisle | +0.235 |
Neg logit contributions | |
ric | -0.197 |
unison | -0.195 |
Deal | -0.184 |
latable | -0.179 |
our | -0.175 |
Token | corn |
Token ID | 11676 |
Feature activation | +2.25e-01 |
Pos logit contributions | |
whispered | +0.191 |
said | +0.176 |
woke | +0.172 |
fell | +0.170 |
chy | +0.169 |
Neg logit contributions | |
unison | -0.209 |
ric | -0.208 |
latable | -0.198 |
Deal | -0.197 |
yeah | -0.184 |
Token | !" |
Token ID | 2474 |
Feature activation | +4.48e-02 |
Pos logit contributions | |
whispered | +0.218 |
said | +0.202 |
fell | +0.192 |
woke | +0.192 |
chy | +0.190 |
Neg logit contributions | |
unison | -0.416 |
ric | -0.404 |
latable | -0.403 |
Deal | -0.392 |
igsaw | -0.383 |
Token |
|
Token ID | 198 |
Feature activation | +2.17e-02 |
Pos logit contributions | |
whispered | +0.050 |
chy | +0.048 |
woke | +0.045 |
fell | +0.044 |
said | +0.043 |
Neg logit contributions | |
unison | -0.073 |
Deal | -0.070 |
latable | -0.069 |
ric | -0.069 |
igsaw | -0.068 |
Token |
|
Token ID | 198 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
whispered | +0.021 |
said | +0.019 |
chy | +0.018 |
woke | +0.018 |
fell | +0.018 |
Neg logit contributions | |
unison | -0.040 |
ric | -0.039 |
latable | -0.038 |
Deal | -0.038 |
igsaw | -0.037 |
Token | Tom |
Token ID | 13787 |
Feature activation | +1.42e-01 |
Pos logit contributions | |
whispered | +0.019 |
said | +0.018 |
fell | +0.017 |
l | +0.017 |
woke | +0.017 |
Neg logit contributions | |
ric | -0.035 |
unison | -0.035 |
latable | -0.034 |
Deal | -0.033 |
igsaw | -0.032 |
Token | and |
Token ID | 290 |
Feature activation | +1.03e-01 |
Pos logit contributions | |
whispered | +0.122 |
said | +0.108 |
woke | +0.105 |
l | +0.104 |
chy | +0.103 |
Neg logit contributions | |
unison | -0.271 |
ric | -0.267 |
latable | -0.265 |
Deal | -0.261 |
igsaw | -0.247 |
Token | He |
Token ID | 679 |
Feature activation | +9.21e-02 |
Pos logit contributions | |
whispered | +0.021 |
chy | +0.020 |
said | +0.019 |
fell | +0.019 |
meant | +0.019 |
Neg logit contributions | |
unison | -0.018 |
Deal | -0.017 |
latable | -0.016 |
ric | -0.016 |
igsaw | -0.016 |
Token | takes |
Token ID | 2753 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
whispered | +0.092 |
said | +0.086 |
woke | +0.082 |
fell | +0.081 |
l | +0.079 |
Neg logit contributions | |
unison | -0.168 |
latable | -0.164 |
ric | -0.163 |
Deal | -0.159 |
igsaw | -0.158 |
Token | his |
Token ID | 465 |
Feature activation | +1.28e-01 |
Pos logit contributions | |
whispered | +0.237 |
chy | +0.224 |
woke | +0.214 |
said | +0.214 |
needed | +0.213 |
Neg logit contributions | |
unison | -0.174 |
Deal | -0.159 |
latable | -0.156 |
igsaw | -0.149 |
ric | -0.147 |
Token | tower |
Token ID | 10580 |
Feature activation | +1.99e-01 |
Pos logit contributions | |
whispered | +0.155 |
said | +0.142 |
woke | +0.139 |
chy | +0.136 |
fell | +0.136 |
Neg logit contributions | |
unison | -0.206 |
ric | -0.202 |
latable | -0.197 |
Deal | -0.194 |
igsaw | -0.190 |
Token | of |
Token ID | 286 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
whispered | +0.215 |
woke | +0.197 |
said | +0.188 |
meant | +0.184 |
chy | +0.183 |
Neg logit contributions | |
ric | -0.318 |
unison | -0.317 |
igsaw | -0.316 |
latable | -0.311 |
ap | -0.294 |
Token | coins |
Token ID | 10796 |
Feature activation | +2.24e-01 |
Pos logit contributions | |
whispered | +0.146 |
woke | +0.138 |
chy | +0.137 |
said | +0.134 |
l | +0.133 |
Neg logit contributions | |
unison | -0.170 |
ric | -0.167 |
latable | -0.160 |
Deal | -0.155 |
igsaw | -0.152 |
Token | and |
Token ID | 290 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
whispered | +0.205 |
woke | +0.191 |
chy | +0.184 |
said | +0.179 |
meant | +0.173 |
Neg logit contributions | |
unison | -0.402 |
ric | -0.398 |
latable | -0.390 |
igsaw | -0.376 |
Deal | -0.368 |
Token | throws |
Token ID | 12542 |
Feature activation | +1.62e-01 |
Pos logit contributions | |
whispered | +0.121 |
said | +0.115 |
fell | +0.108 |
woke | +0.106 |
meant | +0.100 |
Neg logit contributions | |
latable | -0.187 |
unison | -0.187 |
ric | -0.184 |
Deal | -0.181 |
igsaw | -0.179 |
Token | it |
Token ID | 340 |
Feature activation | +1.64e-01 |
Pos logit contributions | |
whispered | +0.246 |
chy | +0.230 |
said | +0.225 |
woke | +0.225 |
fell | +0.222 |
Neg logit contributions | |
unison | -0.210 |
ric | -0.200 |
latable | -0.194 |
Deal | -0.185 |
igsaw | -0.183 |
Token | at |
Token ID | 379 |
Feature activation | +1.11e-01 |
Pos logit contributions | |
whispered | +0.251 |
woke | +0.237 |
meant | +0.234 |
aisle | +0.225 |
said | +0.225 |
Neg logit contributions | |
ap | -0.171 |
ric | -0.167 |
teammates | -0.163 |
unison | -0.158 |
our | -0.157 |
Token | the |
Token ID | 262 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
whispered | +0.181 |
chy | +0.170 |
meant | +0.167 |
l | +0.163 |
woke | +0.162 |
Neg logit contributions | |
unison | -0.135 |
ric | -0.119 |
latable | -0.118 |
Deal | -0.114 |
teammates | -0.110 |
Token | much |
Token ID | 881 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
said | +0.071 |
l | +0.070 |
whispered | +0.070 |
fell | +0.066 |
needed | +0.063 |
Neg logit contributions | |
unison | -0.192 |
Deal | -0.186 |
ric | -0.185 |
latable | -0.179 |
igsaw | -0.167 |
Token | he |
Token ID | 339 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
whispered | +0.118 |
said | +0.110 |
woke | +0.103 |
fell | +0.103 |
meant | +0.101 |
Neg logit contributions | |
unison | -0.225 |
ric | -0.221 |
latable | -0.217 |
Deal | -0.213 |
igsaw | -0.207 |
Token | loved |
Token ID | 6151 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
l | +0.098 |
aisle | +0.096 |
whispered | +0.089 |
prick | +0.084 |
low | +0.082 |
Neg logit contributions | |
latable | -0.189 |
unison | -0.189 |
ric | -0.179 |
gonna | -0.170 |
teammates | -0.169 |
Token | his |
Token ID | 465 |
Feature activation | +1.35e-01 |
Pos logit contributions | |
whispered | +0.177 |
woke | +0.162 |
fell | +0.161 |
chy | +0.160 |
said | +0.160 |
Neg logit contributions | |
ric | -0.150 |
unison | -0.149 |
latable | -0.147 |
igsaw | -0.132 |
yeah | -0.127 |
Token | funny |
Token ID | 8258 |
Feature activation | +1.67e-01 |
Pos logit contributions | |
whispered | +0.188 |
said | +0.173 |
woke | +0.172 |
fell | +0.171 |
chy | +0.169 |
Neg logit contributions | |
unison | -0.187 |
latable | -0.186 |
ric | -0.183 |
igsaw | -0.175 |
Deal | -0.170 |
Token | suit |
Token ID | 6050 |
Feature activation | +2.24e-01 |
Pos logit contributions | |
whispered | +0.218 |
said | +0.203 |
chy | +0.196 |
woke | +0.196 |
fell | +0.196 |
Neg logit contributions | |
unison | -0.249 |
ric | -0.244 |
latable | -0.241 |
Deal | -0.230 |
igsaw | -0.226 |
Token | and |
Token ID | 290 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
whispered | +0.242 |
prick | +0.216 |
woke | +0.209 |
buzz | +0.205 |
l | +0.203 |
Neg logit contributions | |
unison | -0.342 |
latable | -0.338 |
ric | -0.336 |
igsaw | -0.333 |
yeah | -0.313 |
Token | how |
Token ID | 703 |
Feature activation | +9.71e-02 |
Pos logit contributions | |
whispered | +0.127 |
said | +0.122 |
woke | +0.114 |
fell | +0.113 |
meant | +0.111 |
Neg logit contributions | |
unison | -0.184 |
latable | -0.179 |
Deal | -0.179 |
ric | -0.179 |
igsaw | -0.173 |
Token | much |
Token ID | 881 |
Feature activation | +1.27e-01 |
Pos logit contributions | |
said | +0.075 |
whispered | +0.074 |
l | +0.070 |
fell | +0.070 |
needed | +0.066 |
Neg logit contributions | |
unison | -0.211 |
Deal | -0.202 |
ric | -0.200 |
latable | -0.200 |
igsaw | -0.190 |
Token | fun |
Token ID | 1257 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
whispered | +0.126 |
said | +0.116 |
woke | +0.111 |
fell | +0.110 |
chy | +0.109 |
Neg logit contributions | |
unison | -0.229 |
ric | -0.224 |
latable | -0.222 |
Deal | -0.216 |
igsaw | -0.212 |
Token | he |
Token ID | 339 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
whispered | +0.115 |
said | +0.113 |
fell | +0.105 |
woke | +0.103 |
meant | +0.101 |
Neg logit contributions | |
unison | -0.303 |
ric | -0.297 |
Deal | -0.297 |
latable | -0.290 |
igsaw | -0.276 |
Token | " |
Token ID | 1 |
Feature activation | +7.03e-02 |
Pos logit contributions | |
whispered | +0.025 |
said | +0.024 |
woke | +0.023 |
fell | +0.022 |
l | +0.022 |
Neg logit contributions | |
ric | -0.030 |
unison | -0.030 |
latable | -0.029 |
Deal | -0.028 |
igsaw | -0.028 |
Token | Let |
Token ID | 5756 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
whispered | +0.095 |
chy | +0.093 |
fell | +0.089 |
l | +0.087 |
said | +0.086 |
Neg logit contributions | |
unison | -0.108 |
ric | -0.100 |
latable | -0.099 |
Deal | -0.095 |
ifying | -0.094 |
Token | 's |
Token ID | 338 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
said | +0.102 |
whispered | +0.100 |
woke | +0.092 |
fell | +0.086 |
meant | +0.085 |
Neg logit contributions | |
ric | -0.250 |
unison | -0.245 |
latable | -0.237 |
igsaw | -0.233 |
Deal | -0.233 |
Token | replace |
Token ID | 6330 |
Feature activation | +1.68e-01 |
Pos logit contributions | |
whispered | +0.112 |
said | +0.110 |
fell | +0.107 |
tree | +0.101 |
woke | +0.101 |
Neg logit contributions | |
igsaw | -0.199 |
unison | -0.193 |
Deal | -0.191 |
latable | -0.188 |
ric | -0.181 |
Token | the |
Token ID | 262 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
whispered | +0.203 |
said | +0.197 |
l | +0.177 |
meant | +0.172 |
man | +0.170 |
Neg logit contributions | |
unison | -0.279 |
Deal | -0.278 |
latable | -0.266 |
igsaw | -0.263 |
ric | -0.260 |
Token | rocks |
Token ID | 12586 |
Feature activation | +2.24e-01 |
Pos logit contributions | |
whispered | +0.157 |
said | +0.146 |
woke | +0.145 |
chy | +0.144 |
fell | +0.138 |
Neg logit contributions | |
ric | -0.158 |
unison | -0.153 |
latable | -0.147 |
yeah | -0.135 |
igsaw | -0.135 |
Token | at |
Token ID | 379 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
whispered | +0.194 |
said | +0.176 |
woke | +0.166 |
fell | +0.165 |
chy | +0.162 |
Neg logit contributions | |
unison | -0.436 |
ric | -0.426 |
latable | -0.423 |
Deal | -0.413 |
igsaw | -0.403 |
Token | the |
Token ID | 262 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
whispered | +0.183 |
l | +0.177 |
chy | +0.175 |
woke | +0.173 |
spoke | +0.166 |
Neg logit contributions | |
ric | -0.153 |
unison | -0.130 |
our | -0.128 |
latable | -0.122 |
least | -0.116 |
Token | bottom |
Token ID | 4220 |
Feature activation | +1.72e-01 |
Pos logit contributions | |
whispered | +0.164 |
said | +0.152 |
woke | +0.145 |
meant | +0.144 |
chy | +0.142 |
Neg logit contributions | |
ric | -0.145 |
unison | -0.142 |
latable | -0.135 |
hey | -0.128 |
yeah | -0.121 |
Token | of |
Token ID | 286 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
whispered | +0.207 |
woke | +0.190 |
chy | +0.188 |
said | +0.184 |
fell | +0.181 |
Neg logit contributions | |
unison | -0.270 |
ric | -0.262 |
latable | -0.258 |
igsaw | -0.250 |
Deal | -0.242 |
Token | the |
Token ID | 262 |
Feature activation | +1.11e-01 |
Pos logit contributions | |
whispered | +0.134 |
said | +0.125 |
woke | +0.122 |
fell | +0.121 |
chy | +0.120 |
Neg logit contributions | |
unison | -0.170 |
ric | -0.166 |
latable | -0.163 |
Deal | -0.157 |
igsaw | -0.153 |
Token | fast |
Token ID | 3049 |
Feature activation | +1.60e-01 |
Pos logit contributions | |
whispered | +0.239 |
woke | +0.223 |
chy | +0.219 |
fell | +0.214 |
said | +0.210 |
Neg logit contributions | |
unison | -0.245 |
ric | -0.242 |
our | -0.229 |
ifying | -0.224 |
latable | -0.223 |
Token | . |
Token ID | 13 |
Feature activation | +2.91e-02 |
Pos logit contributions | |
whispered | +0.160 |
chy | +0.150 |
fell | +0.139 |
said | +0.138 |
woke | +0.138 |
Neg logit contributions | |
unison | -0.281 |
ric | -0.268 |
latable | -0.256 |
Deal | -0.252 |
ifying | -0.248 |
Token | I |
Token ID | 314 |
Feature activation | +9.84e-02 |
Pos logit contributions | |
whispered | +0.039 |
said | +0.037 |
meant | +0.035 |
woke | +0.035 |
fell | +0.034 |
Neg logit contributions | |
ric | -0.042 |
latable | -0.041 |
unison | -0.041 |
igsaw | -0.040 |
yeah | -0.036 |
Token | put |
Token ID | 1234 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
whispered | +0.130 |
said | +0.119 |
woke | +0.118 |
l | +0.118 |
chy | +0.118 |
Neg logit contributions | |
unison | -0.140 |
ric | -0.137 |
latable | -0.133 |
Deal | -0.130 |
yeah | -0.122 |
Token | the |
Token ID | 262 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
whispered | +0.242 |
said | +0.227 |
woke | +0.223 |
fell | +0.220 |
chy | +0.220 |
Neg logit contributions | |
unison | -0.232 |
ric | -0.225 |
latable | -0.222 |
Deal | -0.215 |
igsaw | -0.208 |
Token | rope |
Token ID | 17182 |
Feature activation | +2.23e-01 |
Pos logit contributions | |
whispered | +0.147 |
said | +0.139 |
woke | +0.134 |
fell | +0.134 |
chy | +0.132 |
Neg logit contributions | |
unison | -0.172 |
ric | -0.167 |
latable | -0.166 |
Deal | -0.159 |
igsaw | -0.156 |
Token | on |
Token ID | 319 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
whispered | +0.226 |
said | +0.204 |
woke | +0.201 |
meant | +0.193 |
fell | +0.192 |
Neg logit contributions | |
unison | -0.389 |
ric | -0.387 |
latable | -0.379 |
Deal | -0.364 |
igsaw | -0.363 |
Token | the |
Token ID | 262 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
whispered | +0.234 |
chy | +0.216 |
said | +0.215 |
woke | +0.215 |
fell | +0.214 |
Neg logit contributions | |
unison | -0.181 |
ric | -0.165 |
latable | -0.161 |
Deal | -0.158 |
igsaw | -0.155 |
Token | bike |
Token ID | 7161 |
Feature activation | +2.01e-01 |
Pos logit contributions | |
whispered | +0.122 |
said | +0.110 |
meant | +0.109 |
woke | +0.106 |
fell | +0.103 |
Neg logit contributions | |
unison | -0.183 |
ric | -0.176 |
latable | -0.172 |
Deal | -0.169 |
igsaw | -0.160 |
Token | to |
Token ID | 284 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
whispered | +0.232 |
woke | +0.211 |
meant | +0.209 |
spoke | +0.208 |
prick | +0.208 |
Neg logit contributions | |
ric | -0.311 |
unison | -0.297 |
ap | -0.286 |
ifying | -0.282 |
igsaw | -0.278 |
Token | see |
Token ID | 766 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
whispered | +0.186 |
chy | +0.173 |
said | +0.173 |
woke | +0.171 |
fell | +0.170 |
Neg logit contributions | |
unison | -0.185 |
ric | -0.178 |
latable | -0.173 |
Deal | -0.169 |
igsaw | -0.162 |
Token | bear |
Token ID | 6842 |
Feature activation | +1.74e-01 |
Pos logit contributions | |
whispered | +0.107 |
meant | +0.097 |
fell | +0.097 |
spoke | +0.095 |
said | +0.094 |
Neg logit contributions | |
Deal | -0.090 |
unison | -0.083 |
ric | -0.083 |
latable | -0.080 |
hey | -0.074 |
Token | did |
Token ID | 750 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
aisle | +0.186 |
chy | +0.177 |
meant | +0.163 |
l | +0.162 |
week | +0.159 |
Neg logit contributions | |
unison | -0.275 |
latable | -0.238 |
teammates | -0.233 |
hey | -0.233 |
yeah | -0.229 |
Token | not |
Token ID | 407 |
Feature activation | +9.90e-02 |
Pos logit contributions | |
fell | +0.087 |
said | +0.083 |
whispered | +0.081 |
low | +0.076 |
hurt | +0.075 |
Neg logit contributions | |
igsaw | -0.241 |
ric | -0.234 |
unison | -0.225 |
latable | -0.222 |
ifying | -0.220 |
Token | like |
Token ID | 588 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
whispered | +0.159 |
aisle | +0.152 |
said | +0.150 |
chy | +0.149 |
l | +0.148 |
Neg logit contributions | |
unison | -0.104 |
latable | -0.101 |
Deal | -0.097 |
ric | -0.095 |
hey | -0.092 |
Token | the |
Token ID | 262 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
whispered | +0.165 |
said | +0.148 |
chy | +0.147 |
fell | +0.144 |
meant | +0.142 |
Neg logit contributions | |
ric | -0.161 |
unison | -0.159 |
latable | -0.159 |
igsaw | -0.142 |
hey | -0.142 |
Token | broom |
Token ID | 47085 |
Feature activation | +2.23e-01 |
Pos logit contributions | |
whispered | +0.145 |
fell | +0.132 |
said | +0.128 |
spoke | +0.126 |
woke | +0.125 |
Neg logit contributions | |
ric | -0.149 |
Deal | -0.144 |
latable | -0.143 |
unison | -0.140 |
yeah | -0.138 |
Token | . |
Token ID | 13 |
Feature activation | +2.90e-02 |
Pos logit contributions | |
whispered | +0.228 |
aisle | +0.196 |
mail | +0.190 |
prick | +0.180 |
l | +0.178 |
Neg logit contributions | |
unison | -0.373 |
ric | -0.349 |
latable | -0.342 |
ifying | -0.326 |
ap | -0.321 |
Token | It |
Token ID | 632 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
whispered | +0.041 |
said | +0.038 |
chy | +0.038 |
l | +0.037 |
woke | +0.037 |
Neg logit contributions | |
unison | -0.040 |
ric | -0.039 |
latable | -0.039 |
Deal | -0.039 |
igsaw | -0.036 |
Token | ran |
Token ID | 4966 |
Feature activation | +1.64e-01 |
Pos logit contributions | |
aisle | +0.112 |
whispered | +0.103 |
l | +0.096 |
books | +0.088 |
man | +0.085 |
Neg logit contributions | |
unison | -0.212 |
latable | -0.207 |
ric | -0.196 |
Deal | -0.181 |
ifying | -0.180 |
Token | to |
Token ID | 284 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
whispered | +0.201 |
said | +0.182 |
woke | +0.177 |
chy | +0.176 |
meant | +0.174 |
Neg logit contributions | |
unison | -0.247 |
latable | -0.244 |
ric | -0.233 |
Deal | -0.227 |
hey | -0.221 |
Token | the |
Token ID | 262 |
Feature activation | +1.03e-01 |
Pos logit contributions | |
whispered | +0.140 |
said | +0.133 |
woke | +0.126 |
fell | +0.125 |
l | +0.117 |
Neg logit contributions | |
unison | -0.241 |
ric | -0.235 |
latable | -0.234 |
igsaw | -0.229 |
Deal | -0.224 |
Token | the |
Token ID | 262 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
whispered | +0.164 |
said | +0.155 |
woke | +0.149 |
fell | +0.147 |
l | +0.145 |
Neg logit contributions | |
unison | -0.209 |
ric | -0.207 |
latable | -0.204 |
Deal | -0.198 |
igsaw | -0.192 |
Token | waves |
Token ID | 9813 |
Feature activation | +1.92e-01 |
Pos logit contributions | |
whispered | +0.165 |
said | +0.152 |
fell | +0.150 |
chy | +0.150 |
woke | +0.149 |
Neg logit contributions | |
unison | -0.155 |
latable | -0.147 |
ric | -0.147 |
Deal | -0.139 |
igsaw | -0.137 |
Token | crashing |
Token ID | 21899 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
whispered | +0.235 |
chy | +0.213 |
prick | +0.211 |
meant | +0.210 |
woke | +0.209 |
Neg logit contributions | |
unison | -0.305 |
ric | -0.286 |
latable | -0.276 |
ifying | -0.266 |
ap | -0.266 |
Token | against |
Token ID | 1028 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
whispered | +0.208 |
said | +0.185 |
woke | +0.182 |
meant | +0.182 |
chy | +0.182 |
Neg logit contributions | |
unison | -0.293 |
ric | -0.281 |
latable | -0.274 |
ifying | -0.269 |
ap | -0.253 |
Token | the |
Token ID | 262 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
whispered | +0.182 |
said | +0.173 |
woke | +0.165 |
fell | +0.163 |
l | +0.161 |
Neg logit contributions | |
ric | -0.222 |
unison | -0.218 |
latable | -0.216 |
Deal | -0.213 |
igsaw | -0.203 |
Token | shore |
Token ID | 15191 |
Feature activation | +2.23e-01 |
Pos logit contributions | |
whispered | +0.158 |
said | +0.147 |
woke | +0.143 |
meant | +0.143 |
fell | +0.143 |
Neg logit contributions | |
unison | -0.154 |
ric | -0.148 |
latable | -0.148 |
Deal | -0.138 |
igsaw | -0.136 |
Token | . |
Token ID | 13 |
Feature activation | +2.35e-02 |
Pos logit contributions | |
whispered | +0.176 |
chy | +0.154 |
woke | +0.152 |
meant | +0.146 |
said | +0.145 |
Neg logit contributions | |
unison | -0.453 |
ric | -0.446 |
latable | -0.423 |
igsaw | -0.406 |
ap | -0.402 |
Token |
|
Token ID | 198 |
Feature activation | +2.30e-02 |
Pos logit contributions | |
said | +0.024 |
whispered | +0.024 |
woke | +0.022 |
fell | +0.021 |
meant | +0.021 |
Neg logit contributions | |
unison | -0.041 |
ric | -0.041 |
latable | -0.041 |
Deal | -0.039 |
igsaw | -0.038 |
Token |
|
Token ID | 198 |
Feature activation | +1.96e-02 |
Pos logit contributions | |
whispered | +0.022 |
said | +0.019 |
woke | +0.019 |
chy | +0.019 |
fell | +0.019 |
Neg logit contributions | |
unison | -0.043 |
ric | -0.042 |
latable | -0.042 |
Deal | -0.041 |
igsaw | -0.040 |
Token | " |
Token ID | 1 |
Feature activation | +7.62e-02 |
Pos logit contributions | |
whispered | +0.026 |
said | +0.026 |
woke | +0.024 |
l | +0.023 |
fell | +0.023 |
Neg logit contributions | |
ric | -0.029 |
latable | -0.028 |
unison | -0.028 |
Deal | -0.027 |
igsaw | -0.027 |
Token | Would |
Token ID | 17353 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
whispered | +0.096 |
said | +0.089 |
fell | +0.087 |
woke | +0.086 |
chy | +0.086 |
Neg logit contributions | |
unison | -0.120 |
ric | -0.115 |
latable | -0.114 |
Deal | -0.111 |
igsaw | -0.107 |
Token | would |
Token ID | 561 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
aisle | +0.087 |
low | +0.083 |
whispered | +0.081 |
prick | +0.077 |
l | +0.077 |
Neg logit contributions | |
unison | -0.190 |
latable | -0.187 |
ric | -0.186 |
yeah | -0.171 |
Deal | -0.171 |
Token | let |
Token ID | 1309 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
whispered | +0.163 |
l | +0.157 |
aisle | +0.153 |
meant | +0.152 |
woke | +0.152 |
Neg logit contributions | |
unison | -0.120 |
ric | -0.117 |
Deal | -0.110 |
latable | -0.108 |
yeah | -0.102 |
Token | them |
Token ID | 606 |
Feature activation | +1.67e-01 |
Pos logit contributions | |
whispered | +0.184 |
said | +0.174 |
woke | +0.169 |
fell | +0.168 |
chy | +0.167 |
Neg logit contributions | |
unison | -0.156 |
ric | -0.150 |
latable | -0.149 |
Deal | -0.143 |
igsaw | -0.139 |
Token | keep |
Token ID | 1394 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
whispered | +0.245 |
said | +0.230 |
woke | +0.227 |
chy | +0.223 |
fell | +0.222 |
Neg logit contributions | |
unison | -0.217 |
ric | -0.212 |
latable | -0.209 |
Deal | -0.198 |
igsaw | -0.193 |
Token | their |
Token ID | 511 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
whispered | +0.238 |
said | +0.223 |
woke | +0.223 |
chy | +0.219 |
fell | +0.218 |
Neg logit contributions | |
unison | -0.183 |
ric | -0.178 |
latable | -0.176 |
Deal | -0.167 |
igsaw | -0.161 |
Token | patch |
Token ID | 8529 |
Feature activation | +2.23e-01 |
Pos logit contributions | |
whispered | +0.174 |
said | +0.163 |
woke | +0.160 |
meant | +0.157 |
l | +0.156 |
Neg logit contributions | |
unison | -0.195 |
ric | -0.194 |
latable | -0.186 |
Deal | -0.180 |
igsaw | -0.172 |
Token | if |
Token ID | 611 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
whispered | +0.200 |
said | +0.172 |
woke | +0.170 |
meant | +0.161 |
spoke | +0.156 |
Neg logit contributions | |
unison | -0.406 |
latable | -0.402 |
ric | -0.397 |
igsaw | -0.387 |
Deal | -0.376 |
Token | they |
Token ID | 484 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
whispered | +0.133 |
said | +0.121 |
woke | +0.118 |
chy | +0.117 |
fell | +0.116 |
Neg logit contributions | |
unison | -0.206 |
ric | -0.200 |
latable | -0.198 |
Deal | -0.193 |
igsaw | -0.189 |
Token | gave |
Token ID | 2921 |
Feature activation | +1.48e-01 |
Pos logit contributions | |
aisle | +0.158 |
l | +0.145 |
floor | +0.136 |
whispered | +0.135 |
man | +0.133 |
Neg logit contributions | |
unison | -0.165 |
ric | -0.155 |
latable | -0.147 |
teammates | -0.142 |
spend | -0.141 |
Token | him |
Token ID | 683 |
Feature activation | +1.60e-01 |
Pos logit contributions | |
whispered | +0.242 |
woke | +0.225 |
chy | +0.224 |
said | +0.224 |
fell | +0.219 |
Neg logit contributions | |
unison | -0.168 |
ric | -0.158 |
latable | -0.154 |
Deal | -0.149 |
igsaw | -0.141 |
Token | their |
Token ID | 511 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
whispered | +0.204 |
said | +0.189 |
woke | +0.184 |
fell | +0.181 |
chy | +0.179 |
Neg logit contributions | |
unison | -0.242 |
ric | -0.237 |
latable | -0.234 |
Deal | -0.222 |
igsaw | -0.220 |
Token | couldn |
Token ID | 3521 |
Feature activation | +9.59e-02 |
Pos logit contributions | |
whispered | +0.093 |
aisle | +0.091 |
l | +0.091 |
low | +0.082 |
man | +0.079 |
Neg logit contributions | |
unison | -0.198 |
ric | -0.196 |
latable | -0.195 |
Deal | -0.181 |
hey | -0.178 |
Token | 't |
Token ID | 470 |
Feature activation | +9.91e-02 |
Pos logit contributions | |
fell | +0.105 |
said | +0.100 |
chy | +0.099 |
woke | +0.099 |
l | +0.098 |
Neg logit contributions | |
unison | -0.174 |
ric | -0.162 |
latable | -0.161 |
Deal | -0.154 |
yeah | -0.153 |
Token | play |
Token ID | 711 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
whispered | +0.136 |
said | +0.128 |
woke | +0.125 |
fell | +0.124 |
chy | +0.123 |
Neg logit contributions | |
unison | -0.144 |
ric | -0.140 |
latable | -0.138 |
Deal | -0.133 |
igsaw | -0.130 |
Token | with |
Token ID | 351 |
Feature activation | +1.20e-01 |
Pos logit contributions | |
whispered | +0.174 |
said | +0.160 |
woke | +0.153 |
fell | +0.151 |
chy | +0.149 |
Neg logit contributions | |
unison | -0.268 |
latable | -0.262 |
ric | -0.260 |
Deal | -0.249 |
igsaw | -0.247 |
Token | his |
Token ID | 465 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
whispered | +0.151 |
said | +0.141 |
woke | +0.137 |
fell | +0.135 |
chy | +0.134 |
Neg logit contributions | |
unison | -0.189 |
ric | -0.184 |
latable | -0.182 |
Deal | -0.176 |
igsaw | -0.172 |
Token | boat |
Token ID | 8848 |
Feature activation | +2.23e-01 |
Pos logit contributions | |
whispered | +0.171 |
said | +0.160 |
woke | +0.155 |
fell | +0.154 |
chy | +0.152 |
Neg logit contributions | |
unison | -0.199 |
ric | -0.193 |
latable | -0.191 |
Deal | -0.186 |
igsaw | -0.181 |
Token | without |
Token ID | 1231 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
whispered | +0.336 |
woke | +0.307 |
said | +0.298 |
l | +0.290 |
meant | +0.287 |
Neg logit contributions | |
unison | -0.264 |
latable | -0.256 |
ric | -0.243 |
teammates | -0.237 |
ap | -0.232 |
Token | bubbles |
Token ID | 25037 |
Feature activation | +2.04e-01 |
Pos logit contributions | |
whispered | +0.184 |
l | +0.173 |
woke | +0.172 |
said | +0.171 |
chy | +0.169 |
Neg logit contributions | |
unison | -0.179 |
ric | -0.177 |
latable | -0.170 |
Deal | -0.167 |
igsaw | -0.157 |
Token | . |
Token ID | 13 |
Feature activation | +2.32e-02 |
Pos logit contributions | |
whispered | +0.251 |
woke | +0.216 |
said | +0.211 |
knew | +0.209 |
spoke | +0.204 |
Neg logit contributions | |
unison | -0.282 |
ifying | -0.280 |
latable | -0.278 |
ric | -0.270 |
igsaw | -0.270 |
Token |
|
Token ID | 198 |
Feature activation | +1.51e-02 |
Pos logit contributions | |
whispered | +0.033 |
l | +0.030 |
chy | +0.030 |
said | +0.029 |
meant | +0.029 |
Neg logit contributions | |
ric | -0.031 |
latable | -0.030 |
unison | -0.030 |
Deal | -0.030 |
igsaw | -0.029 |
Token |
|
Token ID | 198 |
Feature activation | +1.19e-02 |
Pos logit contributions | |
whispered | +0.020 |
chy | +0.018 |
meant | +0.017 |
prick | +0.017 |
woke | +0.017 |
Neg logit contributions | |
ric | -0.022 |
unison | -0.021 |
Deal | -0.021 |
latable | -0.021 |
igsaw | -0.020 |
Token | She |
Token ID | 1375 |
Feature activation | +8.08e-02 |
Pos logit contributions | |
whispered | +0.015 |
said | +0.014 |
chy | +0.014 |
woke | +0.014 |
l | +0.014 |
Neg logit contributions | |
unison | -0.019 |
ric | -0.019 |
latable | -0.019 |
Deal | -0.018 |
igsaw | -0.018 |
Token | even |
Token ID | 772 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
l | +0.066 |
whispered | +0.066 |
chy | +0.065 |
aisle | +0.064 |
low | +0.060 |
Neg logit contributions | |
unison | -0.147 |
latable | -0.141 |
ric | -0.138 |
Deal | -0.132 |
yeah | -0.131 |
Token | shared |
Token ID | 4888 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
chy | +0.105 |
whispered | +0.103 |
l | +0.101 |
aisle | +0.090 |
fell | +0.089 |
Neg logit contributions | |
unison | -0.260 |
latable | -0.245 |
ric | -0.240 |
yeah | -0.237 |
Deal | -0.235 |
Token | her |
Token ID | 607 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
whispered | +0.180 |
chy | +0.177 |
woke | +0.169 |
fell | +0.168 |
l | +0.162 |
Neg logit contributions | |
unison | -0.233 |
ric | -0.221 |
latable | -0.219 |
Deal | -0.210 |
igsaw | -0.210 |
Token | delicious |
Token ID | 12625 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
whispered | +0.151 |
said | +0.141 |
woke | +0.138 |
fell | +0.136 |
chy | +0.136 |
Neg logit contributions | |
unison | -0.177 |
ric | -0.172 |
latable | -0.170 |
Deal | -0.165 |
igsaw | -0.161 |
Token | snack |
Token ID | 26906 |
Feature activation | +2.22e-01 |
Pos logit contributions | |
whispered | +0.255 |
chy | +0.239 |
said | +0.236 |
fell | +0.229 |
l | +0.227 |
Neg logit contributions | |
unison | -0.236 |
latable | -0.236 |
ric | -0.231 |
igsaw | -0.222 |
Deal | -0.217 |
Token | with |
Token ID | 351 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
whispered | +0.308 |
woke | +0.296 |
meant | +0.282 |
l | +0.277 |
chy | +0.276 |
Neg logit contributions | |
igsaw | -0.295 |
unison | -0.280 |
latable | -0.276 |
teammates | -0.265 |
ap | -0.249 |
Token | the |
Token ID | 262 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
whispered | +0.139 |
said | +0.127 |
woke | +0.127 |
chy | +0.125 |
fell | +0.124 |
Neg logit contributions | |
unison | -0.202 |
ric | -0.195 |
latable | -0.192 |
Deal | -0.186 |
igsaw | -0.182 |
Token | little |
Token ID | 1310 |
Feature activation | +1.28e-01 |
Pos logit contributions | |
whispered | +0.108 |
said | +0.096 |
woke | +0.094 |
fell | +0.094 |
chy | +0.092 |
Neg logit contributions | |
unison | -0.197 |
latable | -0.193 |
ric | -0.191 |
Deal | -0.186 |
igsaw | -0.185 |
Token | boy |
Token ID | 2933 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
whispered | +0.074 |
said | +0.070 |
man | +0.058 |
fell | +0.056 |
woke | +0.056 |
Neg logit contributions | |
ric | -0.307 |
unison | -0.302 |
Deal | -0.294 |
latable | -0.290 |
ifying | -0.280 |
Token | as |
Token ID | 355 |
Feature activation | +9.11e-02 |
Pos logit contributions | |
whispered | +0.133 |
woke | +0.118 |
chy | +0.115 |
said | +0.113 |
l | +0.112 |
Neg logit contributions | |
unison | -0.328 |
ric | -0.318 |
latable | -0.310 |
igsaw | -0.299 |
Deal | -0.298 |
Token | When |
Token ID | 1649 |
Feature activation | -2.62e-01 |
Pos logit contributions | |
unison | +0.494 |
ric | +0.476 |
latable | +0.467 |
Deal | +0.465 |
igsaw | +0.446 |
Neg logit contributions | |
whispered | -0.382 |
chy | -0.353 |
said | -0.349 |
woke | -0.347 |
fell | -0.344 |
Token | Daisy |
Token ID | 40355 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
unison | +0.451 |
ric | +0.439 |
latable | +0.433 |
Deal | +0.423 |
igsaw | +0.417 |
Neg logit contributions | |
whispered | -0.277 |
said | -0.249 |
chy | -0.244 |
l | -0.241 |
woke | -0.241 |
Token | put |
Token ID | 1234 |
Feature activation | -1.54e-01 |
Pos logit contributions | |
unison | +0.321 |
latable | +0.317 |
ric | +0.307 |
yeah | +0.298 |
gonna | +0.298 |
Neg logit contributions | |
prick | -0.193 |
chy | -0.182 |
l | -0.178 |
low | -0.178 |
whispered | -0.166 |
Token | the |
Token ID | 262 |
Feature activation | -2.10e-01 |
Pos logit contributions | |
unison | +0.248 |
ric | +0.242 |
Deal | +0.238 |
latable | +0.237 |
igsaw | +0.237 |
Neg logit contributions | |
whispered | -0.169 |
woke | -0.159 |
chy | -0.156 |
l | -0.149 |
said | -0.149 |
Token | magn |
Token ID | 7842 |
Feature activation | -1.79e-01 |
Pos logit contributions | |
unison | +0.327 |
ric | +0.317 |
latable | +0.310 |
igsaw | +0.305 |
Deal | +0.297 |
Neg logit contributions | |
whispered | -0.255 |
woke | -0.233 |
said | -0.232 |
fell | -0.231 |
meant | -0.228 |
Token | ifying |
Token ID | 4035 |
Feature activation | -3.70e-01 |
Pos logit contributions | |
unison | +0.069 |
ric | +0.057 |
latable | +0.055 |
Deal | +0.048 |
ifying | +0.048 |
Neg logit contributions | |
whispered | -0.435 |
said | -0.420 |
woke | -0.414 |
fell | -0.409 |
chy | -0.408 |
Token | glass |
Token ID | 5405 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
unison | +0.703 |
latable | +0.692 |
ric | +0.687 |
Deal | +0.680 |
hey | +0.665 |
Neg logit contributions | |
said | -0.324 |
whispered | -0.323 |
woke | -0.310 |
fell | -0.306 |
l | -0.285 |
Token | down |
Token ID | 866 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
igsaw | +0.200 |
unison | +0.190 |
latable | +0.182 |
ifying | +0.180 |
ric | +0.172 |
Neg logit contributions | |
whispered | -0.154 |
knew | -0.139 |
chy | -0.136 |
woke | -0.136 |
meant | -0.135 |
Token | , |
Token ID | 11 |
Feature activation | -1.85e-01 |
Pos logit contributions | |
unison | +0.198 |
igsaw | +0.194 |
yeah | +0.186 |
Deal | +0.182 |
ifying | +0.182 |
Neg logit contributions | |
whispered | -0.190 |
chy | -0.175 |
woke | -0.173 |
meant | -0.165 |
said | -0.159 |
Token | the |
Token ID | 262 |
Feature activation | -1.94e-01 |
Pos logit contributions | |
unison | +0.403 |
ric | +0.399 |
latable | +0.394 |
Deal | +0.384 |
igsaw | +0.373 |
Neg logit contributions | |
whispered | -0.123 |
said | -0.111 |
fell | -0.102 |
woke | -0.102 |
l | -0.097 |
Token | cater |
Token ID | 20825 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
unison | +0.316 |
ric | +0.291 |
igsaw | +0.284 |
Deal | +0.278 |
latable | +0.277 |
Neg logit contributions | |
whispered | -0.223 |
meant | -0.203 |
woke | -0.200 |
fell | -0.195 |
chy | -0.195 |
Token | John |
Token ID | 1757 |
Feature activation | -3.19e-01 |
Pos logit contributions | |
chy | +0.149 |
prick | +0.149 |
l | +0.142 |
aisle | +0.138 |
iv | +0.132 |
Neg logit contributions | |
unison | -0.252 |
ric | -0.216 |
ap | -0.209 |
hey | -0.207 |
latable | -0.197 |
Token | chose |
Token ID | 7690 |
Feature activation | -2.21e-01 |
Pos logit contributions | |
unison | +0.644 |
ric | +0.631 |
latable | +0.616 |
Deal | +0.601 |
igsaw | +0.580 |
Neg logit contributions | |
whispered | -0.243 |
chy | -0.211 |
l | -0.205 |
said | -0.205 |
woke | -0.201 |
Token | some |
Token ID | 617 |
Feature activation | -2.10e-01 |
Pos logit contributions | |
unison | +0.307 |
ric | +0.299 |
latable | +0.293 |
igsaw | +0.275 |
Deal | +0.271 |
Neg logit contributions | |
whispered | -0.303 |
woke | -0.283 |
chy | -0.280 |
fell | -0.278 |
said | -0.276 |
Token | carrots |
Token ID | 34397 |
Feature activation | -1.70e-01 |
Pos logit contributions | |
unison | +0.329 |
latable | +0.315 |
ric | +0.313 |
igsaw | +0.298 |
Deal | +0.291 |
Neg logit contributions | |
whispered | -0.253 |
woke | -0.237 |
fell | -0.230 |
chy | -0.224 |
said | -0.223 |
Token | and |
Token ID | 290 |
Feature activation | -2.25e-01 |
Pos logit contributions | |
ric | +0.277 |
unison | +0.274 |
latable | +0.268 |
ifying | +0.254 |
igsaw | +0.248 |
Neg logit contributions | |
whispered | -0.180 |
woke | -0.167 |
chy | -0.157 |
fell | -0.156 |
l | -0.154 |
Token | cuc |
Token ID | 38421 |
Feature activation | -3.39e-01 |
Pos logit contributions | |
unison | +0.385 |
ric | +0.373 |
latable | +0.369 |
Deal | +0.356 |
igsaw | +0.346 |
Neg logit contributions | |
whispered | -0.233 |
woke | -0.209 |
said | -0.206 |
l | -0.206 |
chy | -0.205 |
Token | umbers |
Token ID | 17024 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
unison | +0.538 |
latable | +0.523 |
ric | +0.497 |
igsaw | +0.476 |
Deal | +0.470 |
Neg logit contributions | |
whispered | -0.425 |
said | -0.414 |
l | -0.406 |
meant | -0.373 |
fell | -0.371 |
Token | to |
Token ID | 284 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
ric | +0.205 |
unison | +0.195 |
latable | +0.186 |
yeah | +0.182 |
ifying | +0.182 |
Neg logit contributions | |
whispered | -0.133 |
woke | -0.120 |
chy | -0.115 |
aisle | -0.111 |
prick | -0.111 |
Token | take |
Token ID | 1011 |
Feature activation | -1.49e-01 |
Pos logit contributions | |
unison | +0.304 |
ric | +0.294 |
latable | +0.293 |
Deal | +0.286 |
igsaw | +0.279 |
Neg logit contributions | |
whispered | -0.234 |
said | -0.224 |
woke | -0.212 |
fell | -0.210 |
chy | -0.208 |
Token | home |
Token ID | 1363 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
unison | +0.190 |
latable | +0.179 |
ric | +0.174 |
igsaw | +0.173 |
Deal | +0.168 |
Neg logit contributions | |
whispered | -0.226 |
woke | -0.207 |
chy | -0.206 |
fell | -0.205 |
said | -0.205 |
Token | . |
Token ID | 13 |
Feature activation | -2.47e-01 |
Pos logit contributions | |
unison | +0.224 |
ric | +0.219 |
latable | +0.216 |
igsaw | +0.210 |
Deal | +0.210 |
Neg logit contributions | |
whispered | -0.104 |
woke | -0.093 |
said | -0.092 |
fell | -0.089 |
chy | -0.087 |
Token | € |
Token ID | 26391 |
Feature activation | -4.93e-01 |
Pos logit contributions | |
whispered | +0.190 |
said | +0.174 |
l | +0.164 |
chy | +0.164 |
woke | +0.162 |
Neg logit contributions | |
unison | -0.228 |
latable | -0.223 |
ric | -0.219 |
igsaw | -0.199 |
Deal | -0.194 |
Token | � |
Token ID | 129 |
Feature activation | -3.90e-01 |
Pos logit contributions | |
unison | +0.845 |
latable | +0.840 |
ric | +0.800 |
Deal | +0.792 |
igsaw | +0.790 |
Neg logit contributions | |
whispered | -0.540 |
said | -0.528 |
fell | -0.510 |
woke | -0.490 |
chy | -0.470 |
Token | � |
Token ID | 241 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
unison | +0.521 |
ric | +0.501 |
latable | +0.498 |
Deal | +0.478 |
igsaw | +0.464 |
Neg logit contributions | |
whispered | -0.576 |
said | -0.546 |
woke | -0.529 |
fell | -0.525 |
chy | -0.522 |
Token | That |
Token ID | 2504 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
unison | +0.635 |
ric | +0.626 |
latable | +0.610 |
Deal | +0.596 |
igsaw | +0.578 |
Neg logit contributions | |
whispered | -0.454 |
said | -0.418 |
woke | -0.407 |
fell | -0.400 |
chy | -0.398 |
Token | â |
Token ID | 22940 |
Feature activation | -2.84e-01 |
Pos logit contributions | |
unison | +0.461 |
ric | +0.455 |
latable | +0.447 |
Deal | +0.431 |
igsaw | +0.426 |
Neg logit contributions | |
whispered | -0.273 |
said | -0.247 |
woke | -0.239 |
fell | -0.236 |
chy | -0.233 |
Token | € |
Token ID | 26391 |
Feature activation | -3.32e-01 |
Pos logit contributions | |
ric | +0.328 |
unison | +0.325 |
latable | +0.316 |
Deal | +0.307 |
igsaw | +0.303 |
Neg logit contributions | |
whispered | -0.469 |
woke | -0.436 |
said | -0.435 |
fell | -0.429 |
l | -0.423 |
Token | ™ |
Token ID | 8151 |
Feature activation | -2.99e-01 |
Pos logit contributions | |
unison | +0.492 |
latable | +0.476 |
ric | +0.471 |
Deal | +0.457 |
igsaw | +0.440 |
Neg logit contributions | |
whispered | -0.440 |
said | -0.421 |
fell | -0.410 |
woke | -0.406 |
l | -0.396 |
Token | s |
Token ID | 82 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
ric | +0.501 |
unison | +0.450 |
igsaw | +0.444 |
latable | +0.441 |
Deal | +0.430 |
Neg logit contributions | |
whispered | -0.363 |
prick | -0.323 |
woke | -0.305 |
chy | -0.295 |
said | -0.290 |
Token | right |
Token ID | 826 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
unison | +0.381 |
latable | +0.367 |
ric | +0.365 |
Deal | +0.357 |
igsaw | +0.350 |
Neg logit contributions | |
whispered | -0.285 |
said | -0.271 |
woke | -0.260 |
fell | -0.258 |
chy | -0.256 |
Token | , |
Token ID | 11 |
Feature activation | -2.28e-01 |
Pos logit contributions | |
ric | +0.307 |
unison | +0.304 |
latable | +0.292 |
Deal | +0.286 |
igsaw | +0.278 |
Neg logit contributions | |
whispered | -0.224 |
woke | -0.201 |
said | -0.200 |
chy | -0.198 |
fell | -0.196 |
Token | Emma |
Token ID | 18966 |
Feature activation | -2.03e-01 |
Pos logit contributions | |
unison | +0.331 |
ric | +0.321 |
latable | +0.318 |
Deal | +0.307 |
igsaw | +0.298 |
Neg logit contributions | |
whispered | -0.308 |
said | -0.288 |
woke | -0.280 |
fell | -0.278 |
chy | -0.277 |
Token | € |
Token ID | 26391 |
Feature activation | -4.93e-01 |
Pos logit contributions | |
l | +0.113 |
chy | +0.112 |
whispered | +0.112 |
aisle | +0.110 |
woke | +0.106 |
Neg logit contributions | |
unison | -0.111 |
latable | -0.104 |
yeah | -0.102 |
ric | -0.101 |
hey | -0.098 |
Token | � |
Token ID | 129 |
Feature activation | -3.90e-01 |
Pos logit contributions | |
unison | +0.845 |
latable | +0.840 |
ric | +0.800 |
Deal | +0.792 |
igsaw | +0.790 |
Neg logit contributions | |
whispered | -0.540 |
said | -0.528 |
fell | -0.510 |
woke | -0.490 |
chy | -0.470 |
Token | � |
Token ID | 241 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
unison | +0.521 |
ric | +0.501 |
latable | +0.498 |
Deal | +0.478 |
igsaw | +0.464 |
Neg logit contributions | |
whispered | -0.576 |
said | -0.546 |
woke | -0.529 |
fell | -0.525 |
chy | -0.522 |
Token | That |
Token ID | 2504 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
unison | +0.635 |
ric | +0.626 |
latable | +0.610 |
Deal | +0.596 |
igsaw | +0.578 |
Neg logit contributions | |
whispered | -0.454 |
said | -0.418 |
woke | -0.407 |
fell | -0.400 |
chy | -0.398 |
Token | â |
Token ID | 22940 |
Feature activation | -2.84e-01 |
Pos logit contributions | |
unison | +0.461 |
ric | +0.455 |
latable | +0.447 |
Deal | +0.431 |
igsaw | +0.426 |
Neg logit contributions | |
whispered | -0.273 |
said | -0.247 |
woke | -0.239 |
fell | -0.236 |
chy | -0.233 |
Token | € |
Token ID | 26391 |
Feature activation | -3.32e-01 |
Pos logit contributions | |
ric | +0.328 |
unison | +0.325 |
latable | +0.316 |
Deal | +0.307 |
igsaw | +0.303 |
Neg logit contributions | |
whispered | -0.469 |
woke | -0.436 |
said | -0.435 |
fell | -0.429 |
l | -0.423 |
Token | ™ |
Token ID | 8151 |
Feature activation | -2.99e-01 |
Pos logit contributions | |
unison | +0.492 |
latable | +0.476 |
ric | +0.471 |
Deal | +0.457 |
igsaw | +0.440 |
Neg logit contributions | |
whispered | -0.440 |
said | -0.421 |
fell | -0.410 |
woke | -0.406 |
l | -0.396 |
Token | s |
Token ID | 82 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
ric | +0.501 |
unison | +0.450 |
igsaw | +0.444 |
latable | +0.441 |
Deal | +0.430 |
Neg logit contributions | |
whispered | -0.363 |
prick | -0.323 |
woke | -0.305 |
chy | -0.295 |
said | -0.290 |
Token | the |
Token ID | 262 |
Feature activation | -2.43e-01 |
Pos logit contributions | |
unison | +0.381 |
latable | +0.367 |
ric | +0.365 |
Deal | +0.357 |
igsaw | +0.350 |
Neg logit contributions | |
whispered | -0.285 |
said | -0.271 |
woke | -0.260 |
fell | -0.258 |
chy | -0.256 |
Token | gas |
Token ID | 3623 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
ric | +0.399 |
unison | +0.396 |
Deal | +0.378 |
latable | +0.376 |
igsaw | +0.366 |
Neg logit contributions | |
whispered | -0.280 |
said | -0.249 |
woke | -0.248 |
fell | -0.247 |
chy | -0.240 |
Token | from |
Token ID | 422 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
ric | +0.250 |
unison | +0.245 |
latable | +0.235 |
Deal | +0.218 |
ifying | +0.216 |
Neg logit contributions | |
whispered | -0.165 |
woke | -0.144 |
said | -0.143 |
chy | -0.142 |
fell | -0.141 |
Token | those |
Token ID | 883 |
Feature activation | -3.54e-01 |
Pos logit contributions | |
whispered | +0.024 |
said | +0.022 |
woke | +0.021 |
fell | +0.021 |
l | +0.020 |
Neg logit contributions | |
unison | -0.056 |
ric | -0.054 |
latable | -0.054 |
Deal | -0.053 |
igsaw | -0.051 |
Token | in |
Token ID | 287 |
Feature activation | -3.09e-01 |
Pos logit contributions | |
unison | +0.578 |
ric | +0.567 |
latable | +0.549 |
Deal | +0.537 |
igsaw | +0.524 |
Neg logit contributions | |
whispered | -0.412 |
said | -0.375 |
woke | -0.374 |
fell | -0.371 |
chy | -0.367 |
Token | need |
Token ID | 761 |
Feature activation | -2.59e-01 |
Pos logit contributions | |
unison | +0.468 |
ric | +0.455 |
latable | +0.450 |
Deal | +0.435 |
igsaw | +0.424 |
Neg logit contributions | |
whispered | -0.401 |
said | -0.375 |
woke | -0.364 |
fell | -0.361 |
chy | -0.358 |
Token | â |
Token ID | 22940 |
Feature activation | -2.85e-01 |
Pos logit contributions | |
unison | +0.362 |
ric | +0.347 |
latable | +0.336 |
Deal | +0.333 |
igsaw | +0.320 |
Neg logit contributions | |
whispered | -0.359 |
woke | -0.330 |
chy | -0.330 |
said | -0.329 |
fell | -0.326 |
Token | € |
Token ID | 26391 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
ric | +0.358 |
unison | +0.354 |
latable | +0.345 |
Deal | +0.339 |
igsaw | +0.329 |
Neg logit contributions | |
whispered | -0.438 |
said | -0.410 |
fell | -0.410 |
woke | -0.407 |
l | -0.405 |
Token | � |
Token ID | 447 |
Feature activation | -3.31e-01 |
Pos logit contributions | |
unison | +0.491 |
latable | +0.475 |
ric | +0.475 |
Deal | +0.459 |
igsaw | +0.446 |
Neg logit contributions | |
whispered | -0.432 |
said | -0.409 |
fell | -0.395 |
woke | -0.395 |
chy | -0.385 |
Token | � |
Token ID | 251 |
Feature activation | -2.87e-01 |
Pos logit contributions | |
unison | +0.347 |
ric | +0.335 |
Deal | +0.328 |
latable | +0.303 |
igsaw | +0.300 |
Neg logit contributions | |
whispered | -0.580 |
woke | -0.545 |
meant | -0.544 |
prick | -0.539 |
chy | -0.538 |
Token | even |
Token ID | 10197 |
Feature activation | -1.72e-01 |
Pos logit contributions | |
unison | +0.454 |
Deal | +0.430 |
ric | +0.426 |
yeah | +0.409 |
latable | +0.397 |
Neg logit contributions | |
chy | -0.329 |
whispered | -0.304 |
prick | -0.300 |
woke | -0.296 |
l | -0.286 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.65e-01 |
Pos logit contributions | |
unison | +0.288 |
ric | +0.276 |
latable | +0.274 |
Deal | +0.269 |
igsaw | +0.265 |
Neg logit contributions | |
whispered | -0.188 |
said | -0.171 |
woke | -0.171 |
fell | -0.170 |
chy | -0.167 |
Token | One |
Token ID | 3198 |
Feature activation | -2.17e-01 |
Pos logit contributions | |
ric | +0.388 |
unison | +0.383 |
latable | +0.362 |
Deal | +0.358 |
igsaw | +0.357 |
Neg logit contributions | |
whispered | -0.343 |
chy | -0.322 |
prick | -0.307 |
said | -0.300 |
fell | -0.300 |
Token | day |
Token ID | 1110 |
Feature activation | -1.70e-01 |
Pos logit contributions | |
ric | +0.456 |
unison | +0.455 |
latable | +0.442 |
igsaw | +0.438 |
Deal | +0.426 |
Neg logit contributions | |
whispered | -0.144 |
woke | -0.112 |
chy | -0.111 |
said | -0.110 |
fell | -0.106 |
Token | by |
Token ID | 416 |
Feature activation | -3.25e-01 |
Pos logit contributions | |
whispered | +0.026 |
chy | +0.024 |
fell | +0.023 |
meant | +0.023 |
prick | +0.023 |
Neg logit contributions | |
unison | -0.032 |
ric | -0.030 |
latable | -0.030 |
Deal | -0.029 |
igsaw | -0.028 |
Token | one |
Token ID | 530 |
Feature activation | -2.84e-01 |
Pos logit contributions | |
unison | +0.474 |
latable | +0.470 |
ric | +0.455 |
Deal | +0.450 |
igsaw | +0.439 |
Neg logit contributions | |
whispered | -0.433 |
said | -0.413 |
woke | -0.397 |
fell | -0.393 |
chy | -0.387 |
Token | � |
Token ID | 6184 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
unison | +0.463 |
ric | +0.438 |
latable | +0.414 |
Deal | +0.410 |
igsaw | +0.409 |
Neg logit contributions | |
whispered | -0.335 |
woke | -0.301 |
chy | -0.299 |
said | -0.298 |
fell | -0.298 |
Token | � |
Token ID | 95 |
Feature activation | -2.42e-01 |
Pos logit contributions | |
unison | +0.344 |
ric | +0.337 |
latable | +0.326 |
Deal | +0.320 |
igsaw | +0.301 |
Neg logit contributions | |
whispered | -0.387 |
said | -0.365 |
woke | -0.351 |
l | -0.351 |
chy | -0.350 |
Token | € |
Token ID | 26391 |
Feature activation | -3.25e-01 |
Pos logit contributions | |
unison | +0.305 |
ric | +0.294 |
latable | +0.292 |
Deal | +0.280 |
igsaw | +0.272 |
Neg logit contributions | |
whispered | -0.377 |
said | -0.357 |
woke | -0.349 |
fell | -0.347 |
chy | -0.343 |
Token | � |
Token ID | 447 |
Feature activation | -3.31e-01 |
Pos logit contributions | |
unison | +0.622 |
latable | +0.609 |
ric | +0.603 |
Deal | +0.588 |
igsaw | +0.577 |
Neg logit contributions | |
whispered | -0.292 |
said | -0.271 |
fell | -0.257 |
woke | -0.257 |
l | -0.246 |
Token | � |
Token ID | 250 |
Feature activation | -2.47e-01 |
Pos logit contributions | |
unison | +0.292 |
ric | +0.278 |
Deal | +0.263 |
latable | +0.262 |
igsaw | +0.246 |
Neg logit contributions | |
whispered | -0.644 |
said | -0.608 |
chy | -0.606 |
woke | -0.604 |
meant | -0.602 |
Token | 1 |
Token ID | 352 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
unison | +0.365 |
ric | +0.354 |
Deal | +0.343 |
latable | +0.342 |
igsaw | +0.330 |
Neg logit contributions | |
whispered | -0.332 |
said | -0.311 |
woke | -0.305 |
fell | -0.303 |
chy | -0.300 |
Token | , |
Token ID | 11 |
Feature activation | -2.04e-01 |
Pos logit contributions | |
unison | +0.226 |
ric | +0.220 |
Deal | +0.208 |
latable | +0.207 |
yeah | +0.204 |
Neg logit contributions | |
whispered | -0.159 |
chy | -0.152 |
woke | -0.146 |
prick | -0.143 |
fell | -0.142 |
Token | 2 |
Token ID | 362 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
unison | +0.281 |
ric | +0.272 |
latable | +0.268 |
Deal | +0.260 |
igsaw | +0.252 |
Neg logit contributions | |
whispered | -0.291 |
said | -0.273 |
woke | -0.266 |
fell | -0.265 |
chy | -0.264 |
Token | , |
Token ID | 11 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
unison | +0.205 |
ric | +0.197 |
latable | +0.190 |
Deal | +0.188 |
igsaw | +0.185 |
Neg logit contributions | |
whispered | -0.146 |
chy | -0.139 |
woke | -0.135 |
fell | -0.132 |
said | -0.127 |
Token | ™ |
Token ID | 8151 |
Feature activation | -4.76e-01 |
Pos logit contributions | |
whispered | +0.156 |
said | +0.148 |
woke | +0.141 |
fell | +0.139 |
chy | +0.138 |
Neg logit contributions | |
unison | -0.204 |
ric | -0.199 |
latable | -0.196 |
Deal | -0.190 |
igsaw | -0.185 |
Token | s |
Token ID | 82 |
Feature activation | -3.49e-01 |
Pos logit contributions | |
ric | +0.889 |
unison | +0.850 |
latable | +0.831 |
Deal | +0.808 |
igsaw | +0.802 |
Neg logit contributions | |
whispered | -0.445 |
prick | -0.376 |
said | -0.374 |
chy | -0.371 |
woke | -0.369 |
Token | ok |
Token ID | 12876 |
Feature activation | -2.78e-01 |
Pos logit contributions | |
unison | +0.557 |
ric | +0.553 |
latable | +0.531 |
Deal | +0.517 |
igsaw | +0.498 |
Neg logit contributions | |
whispered | -0.416 |
said | -0.377 |
woke | -0.371 |
fell | -0.367 |
l | -0.367 |
Token | , |
Token ID | 11 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
unison | +0.583 |
ric | +0.571 |
latable | +0.568 |
Deal | +0.554 |
igsaw | +0.544 |
Neg logit contributions | |
whispered | -0.200 |
said | -0.178 |
woke | -0.168 |
fell | -0.165 |
chy | -0.162 |
Token | â |
Token ID | 22940 |
Feature activation | -2.84e-01 |
Pos logit contributions | |
unison | +0.428 |
ric | +0.417 |
latable | +0.412 |
Deal | +0.399 |
igsaw | +0.389 |
Neg logit contributions | |
whispered | -0.368 |
said | -0.345 |
woke | -0.335 |
fell | -0.332 |
chy | -0.329 |
Token | € |
Token ID | 26391 |
Feature activation | -3.29e-01 |
Pos logit contributions | |
unison | +0.331 |
ric | +0.331 |
Deal | +0.317 |
latable | +0.314 |
igsaw | +0.294 |
Neg logit contributions | |
whispered | -0.454 |
woke | -0.424 |
fell | -0.417 |
said | -0.417 |
prick | -0.415 |
Token | said |
Token ID | 531 |
Feature activation | -2.23e-01 |
Pos logit contributions | |
unison | +0.537 |
latable | +0.524 |
ric | +0.513 |
Deal | +0.506 |
igsaw | +0.495 |
Neg logit contributions | |
whispered | -0.391 |
said | -0.373 |
fell | -0.363 |
woke | -0.357 |
l | -0.345 |
Token | her |
Token ID | 607 |
Feature activation | -2.23e-01 |
Pos logit contributions | |
unison | +0.386 |
ric | +0.375 |
latable | +0.371 |
Deal | +0.366 |
igsaw | +0.353 |
Neg logit contributions | |
whispered | -0.236 |
said | -0.213 |
woke | -0.207 |
chy | -0.205 |
fell | -0.203 |
Token | mum |
Token ID | 25682 |
Feature activation | -1.70e-01 |
Pos logit contributions | |
unison | +0.287 |
ric | +0.283 |
Deal | +0.273 |
latable | +0.269 |
igsaw | +0.261 |
Neg logit contributions | |
whispered | -0.332 |
said | -0.307 |
woke | -0.300 |
chy | -0.297 |
fell | -0.295 |
Token | . |
Token ID | 13 |
Feature activation | -2.67e-01 |
Pos logit contributions | |
ric | +0.278 |
unison | +0.265 |
yeah | +0.248 |
hey | +0.241 |
ifying | +0.239 |
Neg logit contributions | |
prick | -0.167 |
whispered | -0.155 |
aisle | -0.152 |
low | -0.150 |
mail | -0.138 |
Token | � |
Token ID | 6184 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
unison | +0.420 |
ric | +0.404 |
Deal | +0.399 |
latable | +0.396 |
igsaw | +0.381 |
Neg logit contributions | |
whispered | -0.328 |
chy | -0.310 |
l | -0.299 |
fell | -0.299 |
woke | -0.298 |
Token | arm |
Token ID | 3211 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
unison | +0.379 |
ric | +0.366 |
latable | +0.350 |
igsaw | +0.338 |
Deal | +0.332 |
Neg logit contributions | |
whispered | -0.279 |
said | -0.248 |
fell | -0.247 |
chy | -0.247 |
woke | -0.247 |
Token | chair |
Token ID | 16337 |
Feature activation | -2.41e-01 |
Pos logit contributions | |
ric | +0.290 |
unison | +0.266 |
latable | +0.265 |
ifying | +0.257 |
igsaw | +0.251 |
Neg logit contributions | |
whispered | -0.188 |
prick | -0.155 |
woke | -0.152 |
said | -0.150 |
low | -0.148 |
Token | . |
Token ID | 13 |
Feature activation | -2.85e-01 |
Pos logit contributions | |
ric | +0.380 |
unison | +0.367 |
latable | +0.347 |
ifying | +0.337 |
ap | +0.333 |
Neg logit contributions | |
whispered | -0.256 |
prick | -0.232 |
l | -0.227 |
meant | -0.213 |
said | -0.211 |
Token |
|
Token ID | 198 |
Feature activation | -2.78e-01 |
Pos logit contributions | |
unison | +0.449 |
ric | +0.434 |
latable | +0.424 |
Deal | +0.421 |
igsaw | +0.405 |
Neg logit contributions | |
whispered | -0.351 |
said | -0.322 |
chy | -0.321 |
fell | -0.317 |
woke | -0.316 |
Token |
|
Token ID | 198 |
Feature activation | -2.74e-01 |
Pos logit contributions | |
unison | +0.463 |
ric | +0.445 |
Deal | +0.432 |
latable | +0.427 |
igsaw | +0.414 |
Neg logit contributions | |
whispered | -0.317 |
chy | -0.301 |
fell | -0.287 |
woke | -0.283 |
meant | -0.283 |
Token | THE |
Token ID | 10970 |
Feature activation | -3.29e-01 |
Pos logit contributions | |
unison | +0.437 |
ric | +0.435 |
latable | +0.429 |
Deal | +0.411 |
igsaw | +0.403 |
Neg logit contributions | |
whispered | -0.328 |
said | -0.308 |
woke | -0.297 |
fell | -0.292 |
l | -0.289 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.56e-01 |
Pos logit contributions | |
ric | +0.531 |
unison | +0.525 |
latable | +0.495 |
Deal | +0.487 |
yeah | +0.483 |
Neg logit contributions | |
whispered | -0.369 |
meant | -0.336 |
l | -0.334 |
prick | -0.330 |
chy | -0.327 |
Token | Once |
Token ID | 7454 |
Feature activation | -2.23e-01 |
Pos logit contributions | |
unison | +0.348 |
ric | +0.334 |
latable | +0.308 |
Deal | +0.303 |
igsaw | +0.301 |
Neg logit contributions | |
whispered | -0.363 |
chy | -0.350 |
prick | -0.340 |
said | -0.317 |
l | -0.312 |
Token | upon |
Token ID | 2402 |
Feature activation | -2.79e-01 |
Pos logit contributions | |
unison | +0.239 |
ric | +0.236 |
latable | +0.225 |
igsaw | +0.221 |
Deal | +0.218 |
Neg logit contributions | |
whispered | -0.376 |
said | -0.349 |
woke | -0.346 |
chy | -0.345 |
fell | -0.341 |
Token | a |
Token ID | 257 |
Feature activation | -1.89e-01 |
Pos logit contributions | |
ric | +0.520 |
unison | +0.496 |
latable | +0.495 |
igsaw | +0.488 |
Deal | +0.488 |
Neg logit contributions | |
said | -0.295 |
whispered | -0.294 |
woke | -0.284 |
fell | -0.281 |
chy | -0.257 |
Token | time |
Token ID | 640 |
Feature activation | -1.71e-01 |
Pos logit contributions | |
ric | +0.398 |
unison | +0.388 |
latable | +0.376 |
Deal | +0.371 |
igsaw | +0.365 |
Neg logit contributions | |
whispered | -0.135 |
woke | -0.131 |
said | -0.129 |
fell | -0.125 |
chy | -0.121 |
Token | � |
Token ID | 129 |
Feature activation | -4.32e-01 |
Pos logit contributions | |
whispered | +0.146 |
woke | +0.135 |
fell | +0.129 |
chy | +0.129 |
said | +0.129 |
Neg logit contributions | |
unison | -0.239 |
ric | -0.225 |
latable | -0.224 |
Deal | -0.222 |
igsaw | -0.219 |
Token | � |
Token ID | 241 |
Feature activation | -4.07e-01 |
Pos logit contributions | |
unison | +0.657 |
latable | +0.613 |
Deal | +0.603 |
ric | +0.598 |
igsaw | +0.590 |
Neg logit contributions | |
whispered | -0.575 |
said | -0.557 |
fell | -0.534 |
woke | -0.531 |
l | -0.526 |
Token | T |
Token ID | 51 |
Feature activation | -2.60e-01 |
Pos logit contributions | |
unison | +0.706 |
ric | +0.686 |
latable | +0.683 |
Deal | +0.663 |
igsaw | +0.650 |
Neg logit contributions | |
whispered | -0.437 |
said | -0.406 |
woke | -0.390 |
fell | -0.386 |
chy | -0.381 |
Token | IM |
Token ID | 3955 |
Feature activation | -1.86e-01 |
Pos logit contributions | |
unison | +0.372 |
ric | +0.361 |
latable | +0.358 |
Deal | +0.345 |
igsaw | +0.336 |
Neg logit contributions | |
whispered | -0.360 |
said | -0.338 |
woke | -0.329 |
fell | -0.326 |
chy | -0.323 |
Token | â |
Token ID | 22940 |
Feature activation | -2.82e-01 |
Pos logit contributions | |
unison | +0.253 |
ric | +0.246 |
latable | +0.242 |
Deal | +0.232 |
igsaw | +0.228 |
Neg logit contributions | |
whispered | -0.268 |
woke | -0.246 |
said | -0.245 |
fell | -0.243 |
chy | -0.243 |
Token | € |
Token ID | 26391 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
ric | +0.310 |
unison | +0.306 |
Deal | +0.294 |
latable | +0.291 |
igsaw | +0.278 |
Neg logit contributions | |
whispered | -0.486 |
said | -0.451 |
woke | -0.450 |
fell | -0.446 |
l | -0.446 |
Token | . |
Token ID | 13 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
unison | +0.530 |
latable | +0.519 |
ric | +0.507 |
Deal | +0.498 |
igsaw | +0.485 |
Neg logit contributions | |
whispered | -0.391 |
said | -0.374 |
fell | -0.364 |
woke | -0.359 |
l | -0.347 |
Token | It |
Token ID | 632 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
unison | +0.388 |
ric | +0.375 |
latable | +0.374 |
Deal | +0.356 |
igsaw | +0.350 |
Neg logit contributions | |
whispered | -0.404 |
said | -0.383 |
woke | -0.370 |
fell | -0.366 |
chy | -0.361 |
Token | looked |
Token ID | 3114 |
Feature activation | -1.62e-01 |
Pos logit contributions | |
ric | +0.333 |
unison | +0.318 |
latable | +0.310 |
Deal | +0.296 |
yeah | +0.293 |
Neg logit contributions | |
whispered | -0.206 |
aisle | -0.185 |
l | -0.181 |
prick | -0.178 |
woke | -0.174 |
Token | nice |
Token ID | 3621 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
unison | +0.230 |
latable | +0.223 |
ifying | +0.211 |
Deal | +0.211 |
gonna | +0.209 |
Neg logit contributions | |
whispered | -0.212 |
aisle | -0.196 |
woke | -0.193 |
fell | -0.190 |
said | -0.186 |
Token | and |
Token ID | 290 |
Feature activation | -1.76e-01 |
Pos logit contributions | |
unison | +0.225 |
latable | +0.214 |
ric | +0.212 |
igsaw | +0.202 |
ifying | +0.201 |
Neg logit contributions | |
whispered | -0.148 |
chy | -0.138 |
fell | -0.135 |
l | -0.133 |
woke | -0.132 |
Token | nos |
Token ID | 43630 |
Feature activation | -3.47e-01 |
Pos logit contributions | |
whispered | +0.147 |
said | +0.137 |
fell | +0.135 |
woke | +0.133 |
chy | +0.132 |
Neg logit contributions | |
unison | -0.139 |
ric | -0.138 |
latable | -0.137 |
Deal | -0.127 |
igsaw | -0.127 |
Token | y |
Token ID | 88 |
Feature activation | -3.10e-01 |
Pos logit contributions | |
ric | +0.630 |
unison | +0.611 |
latable | +0.587 |
igsaw | +0.574 |
Deal | +0.566 |
Neg logit contributions | |
whispered | -0.331 |
chy | -0.297 |
fell | -0.296 |
said | -0.282 |
woke | -0.282 |
Token | . |
Token ID | 13 |
Feature activation | -3.33e-01 |
Pos logit contributions | |
unison | +0.599 |
ric | +0.589 |
latable | +0.564 |
Deal | +0.550 |
igsaw | +0.544 |
Neg logit contributions | |
whispered | -0.254 |
said | -0.211 |
fell | -0.211 |
chy | -0.208 |
woke | -0.207 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -3.00e-01 |
Pos logit contributions | |
unison | +0.506 |
ric | +0.490 |
Deal | +0.474 |
latable | +0.473 |
igsaw | +0.455 |
Neg logit contributions | |
whispered | -0.426 |
chy | -0.395 |
said | -0.389 |
woke | -0.388 |
fell | -0.387 |
Token | Once |
Token ID | 7454 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
ric | +0.397 |
unison | +0.370 |
igsaw | +0.354 |
Deal | +0.351 |
latable | +0.343 |
Neg logit contributions | |
whispered | -0.429 |
chy | -0.409 |
prick | -0.396 |
meant | -0.368 |
said | -0.365 |
Token | upon |
Token ID | 2402 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
unison | +0.305 |
ric | +0.302 |
latable | +0.289 |
Deal | +0.282 |
igsaw | +0.275 |
Neg logit contributions | |
whispered | -0.447 |
said | -0.419 |
woke | -0.412 |
chy | -0.407 |
fell | -0.407 |
Token | a |
Token ID | 257 |
Feature activation | -2.33e-01 |
Pos logit contributions | |
ric | +0.628 |
unison | +0.609 |
latable | +0.605 |
Deal | +0.599 |
igsaw | +0.596 |
Neg logit contributions | |
whispered | -0.333 |
said | -0.329 |
fell | -0.320 |
woke | -0.319 |
chy | -0.285 |
Token | time |
Token ID | 640 |
Feature activation | -2.09e-01 |
Pos logit contributions | |
ric | +0.459 |
unison | +0.458 |
latable | +0.442 |
Deal | +0.433 |
igsaw | +0.428 |
Neg logit contributions | |
whispered | -0.191 |
said | -0.177 |
woke | -0.173 |
fell | -0.170 |
chy | -0.167 |
Token | , |
Token ID | 11 |
Feature activation | -2.14e-01 |
Pos logit contributions | |
unison | +0.452 |
ric | +0.435 |
latable | +0.433 |
Deal | +0.420 |
igsaw | +0.408 |
Neg logit contributions | |
whispered | -0.146 |
said | -0.139 |
fell | -0.132 |
woke | -0.129 |
l | -0.120 |
Token | a |
Token ID | 257 |
Feature activation | -2.09e-01 |
Pos logit contributions | |
unison | +0.341 |
ric | +0.335 |
latable | +0.329 |
Deal | +0.318 |
igsaw | +0.311 |
Neg logit contributions | |
whispered | -0.259 |
said | -0.237 |
woke | -0.233 |
fell | -0.227 |
chy | -0.227 |
Token | little |
Token ID | 1310 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
unison | +0.339 |
latable | +0.329 |
ric | +0.321 |
Deal | +0.311 |
igsaw | +0.308 |
Neg logit contributions | |
whispered | -0.248 |
said | -0.223 |
meant | -0.220 |
woke | -0.215 |
chy | -0.214 |
Token | € |
Token ID | 26391 |
Feature activation | -4.93e-01 |
Pos logit contributions | |
whispered | +0.193 |
woke | +0.183 |
fell | +0.168 |
chy | +0.167 |
said | +0.167 |
Neg logit contributions | |
unison | -0.315 |
latable | -0.291 |
ifying | -0.289 |
ric | -0.288 |
igsaw | -0.286 |
Token | � |
Token ID | 129 |
Feature activation | -3.90e-01 |
Pos logit contributions | |
unison | +0.845 |
latable | +0.840 |
ric | +0.800 |
Deal | +0.792 |
igsaw | +0.790 |
Neg logit contributions | |
whispered | -0.540 |
said | -0.528 |
fell | -0.510 |
woke | -0.490 |
chy | -0.470 |
Token | � |
Token ID | 241 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
unison | +0.521 |
ric | +0.501 |
latable | +0.498 |
Deal | +0.478 |
igsaw | +0.464 |
Neg logit contributions | |
whispered | -0.576 |
said | -0.546 |
woke | -0.529 |
fell | -0.525 |
chy | -0.522 |
Token | Let |
Token ID | 5756 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
unison | +0.635 |
ric | +0.626 |
latable | +0.610 |
Deal | +0.596 |
igsaw | +0.578 |
Neg logit contributions | |
whispered | -0.454 |
said | -0.418 |
woke | -0.407 |
fell | -0.400 |
chy | -0.398 |
Token | â |
Token ID | 22940 |
Feature activation | -2.80e-01 |
Pos logit contributions | |
unison | +0.468 |
ric | +0.465 |
latable | +0.455 |
Deal | +0.445 |
igsaw | +0.436 |
Neg logit contributions | |
whispered | -0.265 |
said | -0.260 |
woke | -0.241 |
fell | -0.235 |
meant | -0.232 |
Token | € |
Token ID | 26391 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
unison | +0.355 |
ric | +0.341 |
latable | +0.338 |
Deal | +0.324 |
igsaw | +0.314 |
Neg logit contributions | |
whispered | -0.434 |
said | -0.413 |
woke | -0.401 |
fell | -0.398 |
chy | -0.395 |
Token | ™ |
Token ID | 8151 |
Feature activation | -2.95e-01 |
Pos logit contributions | |
unison | +0.528 |
latable | +0.519 |
Deal | +0.496 |
ric | +0.489 |
igsaw | +0.478 |
Neg logit contributions | |
whispered | -0.396 |
said | -0.396 |
fell | -0.393 |
woke | -0.371 |
l | -0.362 |
Token | s |
Token ID | 82 |
Feature activation | -2.36e-01 |
Pos logit contributions | |
ric | +0.522 |
unison | +0.481 |
igsaw | +0.464 |
latable | +0.464 |
Deal | +0.463 |
Neg logit contributions | |
whispered | -0.318 |
prick | -0.288 |
chy | -0.271 |
woke | -0.269 |
spark | -0.256 |
Token | explore |
Token ID | 7301 |
Feature activation | -1.71e-01 |
Pos logit contributions | |
unison | +0.384 |
latable | +0.381 |
Deal | +0.367 |
igsaw | +0.361 |
ric | +0.360 |
Neg logit contributions | |
said | -0.295 |
whispered | -0.291 |
fell | -0.279 |
woke | -0.266 |
meant | -0.263 |
Token | and |
Token ID | 290 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
unison | +0.272 |
ric | +0.264 |
latable | +0.262 |
Deal | +0.252 |
igsaw | +0.247 |
Neg logit contributions | |
whispered | -0.209 |
said | -0.193 |
woke | -0.189 |
fell | -0.187 |
chy | -0.186 |
Token | have |
Token ID | 423 |
Feature activation | -1.93e-01 |
Pos logit contributions | |
unison | +0.341 |
ric | +0.332 |
latable | +0.331 |
Deal | +0.324 |
igsaw | +0.313 |
Neg logit contributions | |
whispered | -0.242 |
said | -0.232 |
fell | -0.217 |
woke | -0.214 |
meant | -0.205 |
Token | € |
Token ID | 26391 |
Feature activation | -4.93e-01 |
Pos logit contributions | |
whispered | +0.140 |
chy | +0.129 |
l | +0.123 |
said | +0.121 |
prick | +0.118 |
Neg logit contributions | |
unison | -0.231 |
latable | -0.222 |
Deal | -0.215 |
ric | -0.214 |
igsaw | -0.209 |
Token | � |
Token ID | 129 |
Feature activation | -3.90e-01 |
Pos logit contributions | |
unison | +0.845 |
latable | +0.840 |
ric | +0.800 |
Deal | +0.792 |
igsaw | +0.790 |
Neg logit contributions | |
whispered | -0.540 |
said | -0.528 |
fell | -0.510 |
woke | -0.490 |
chy | -0.470 |
Token | � |
Token ID | 241 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
unison | +0.521 |
ric | +0.501 |
latable | +0.498 |
Deal | +0.478 |
igsaw | +0.464 |
Neg logit contributions | |
whispered | -0.576 |
said | -0.546 |
woke | -0.529 |
fell | -0.525 |
chy | -0.522 |
Token | Let |
Token ID | 5756 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
unison | +0.635 |
ric | +0.626 |
latable | +0.610 |
Deal | +0.596 |
igsaw | +0.578 |
Neg logit contributions | |
whispered | -0.454 |
said | -0.418 |
woke | -0.407 |
fell | -0.400 |
chy | -0.398 |
Token | â |
Token ID | 22940 |
Feature activation | -2.80e-01 |
Pos logit contributions | |
unison | +0.468 |
ric | +0.465 |
latable | +0.455 |
Deal | +0.445 |
igsaw | +0.436 |
Neg logit contributions | |
whispered | -0.265 |
said | -0.260 |
woke | -0.241 |
fell | -0.235 |
meant | -0.232 |
Token | € |
Token ID | 26391 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
unison | +0.355 |
ric | +0.341 |
latable | +0.338 |
Deal | +0.324 |
igsaw | +0.314 |
Neg logit contributions | |
whispered | -0.434 |
said | -0.413 |
woke | -0.401 |
fell | -0.398 |
chy | -0.395 |
Token | ™ |
Token ID | 8151 |
Feature activation | -2.95e-01 |
Pos logit contributions | |
unison | +0.528 |
latable | +0.519 |
Deal | +0.496 |
ric | +0.489 |
igsaw | +0.478 |
Neg logit contributions | |
whispered | -0.396 |
said | -0.396 |
fell | -0.393 |
woke | -0.371 |
l | -0.362 |
Token | s |
Token ID | 82 |
Feature activation | -2.36e-01 |
Pos logit contributions | |
ric | +0.522 |
unison | +0.481 |
igsaw | +0.464 |
latable | +0.464 |
Deal | +0.463 |
Neg logit contributions | |
whispered | -0.318 |
prick | -0.288 |
chy | -0.271 |
woke | -0.269 |
spark | -0.256 |
Token | not |
Token ID | 407 |
Feature activation | -2.44e-01 |
Pos logit contributions | |
unison | +0.384 |
latable | +0.381 |
Deal | +0.367 |
igsaw | +0.361 |
ric | +0.360 |
Neg logit contributions | |
said | -0.295 |
whispered | -0.291 |
fell | -0.279 |
woke | -0.266 |
meant | -0.263 |
Token | think |
Token ID | 892 |
Feature activation | -2.01e-01 |
Pos logit contributions | |
unison | +0.379 |
latable | +0.367 |
ric | +0.365 |
Deal | +0.355 |
igsaw | +0.345 |
Neg logit contributions | |
whispered | -0.312 |
said | -0.295 |
woke | -0.283 |
fell | -0.282 |
meant | -0.277 |
Token | about |
Token ID | 546 |
Feature activation | -1.94e-01 |
Pos logit contributions | |
unison | +0.309 |
ric | +0.301 |
latable | +0.297 |
Deal | +0.280 |
igsaw | +0.279 |
Neg logit contributions | |
whispered | -0.248 |
woke | -0.223 |
chy | -0.223 |
said | -0.220 |
fell | -0.216 |
Token | € |
Token ID | 26391 |
Feature activation | -4.93e-01 |
Pos logit contributions | |
l | +0.114 |
whispered | +0.111 |
chy | +0.109 |
low | +0.107 |
prick | +0.106 |
Neg logit contributions | |
latable | -0.205 |
unison | -0.203 |
ric | -0.195 |
ap | -0.185 |
hey | -0.184 |
Token | � |
Token ID | 129 |
Feature activation | -3.90e-01 |
Pos logit contributions | |
unison | +0.845 |
latable | +0.840 |
ric | +0.800 |
Deal | +0.792 |
igsaw | +0.790 |
Neg logit contributions | |
whispered | -0.540 |
said | -0.528 |
fell | -0.510 |
woke | -0.490 |
chy | -0.470 |
Token | � |
Token ID | 241 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
unison | +0.521 |
ric | +0.501 |
latable | +0.498 |
Deal | +0.478 |
igsaw | +0.464 |
Neg logit contributions | |
whispered | -0.576 |
said | -0.546 |
woke | -0.529 |
fell | -0.525 |
chy | -0.522 |
Token | Let |
Token ID | 5756 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
unison | +0.635 |
ric | +0.626 |
latable | +0.610 |
Deal | +0.596 |
igsaw | +0.578 |
Neg logit contributions | |
whispered | -0.454 |
said | -0.418 |
woke | -0.407 |
fell | -0.400 |
chy | -0.398 |
Token | â |
Token ID | 22940 |
Feature activation | -2.80e-01 |
Pos logit contributions | |
unison | +0.468 |
ric | +0.465 |
latable | +0.455 |
Deal | +0.445 |
igsaw | +0.436 |
Neg logit contributions | |
whispered | -0.265 |
said | -0.260 |
woke | -0.241 |
fell | -0.235 |
meant | -0.232 |
Token | € |
Token ID | 26391 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
unison | +0.355 |
ric | +0.341 |
latable | +0.338 |
Deal | +0.324 |
igsaw | +0.314 |
Neg logit contributions | |
whispered | -0.434 |
said | -0.413 |
woke | -0.401 |
fell | -0.398 |
chy | -0.395 |
Token | ™ |
Token ID | 8151 |
Feature activation | -2.95e-01 |
Pos logit contributions | |
unison | +0.528 |
latable | +0.519 |
Deal | +0.496 |
ric | +0.489 |
igsaw | +0.478 |
Neg logit contributions | |
whispered | -0.396 |
said | -0.396 |
fell | -0.393 |
woke | -0.371 |
l | -0.362 |
Token | s |
Token ID | 82 |
Feature activation | -2.36e-01 |
Pos logit contributions | |
ric | +0.522 |
unison | +0.481 |
igsaw | +0.464 |
latable | +0.464 |
Deal | +0.463 |
Neg logit contributions | |
whispered | -0.318 |
prick | -0.288 |
chy | -0.271 |
woke | -0.269 |
spark | -0.256 |
Token | play |
Token ID | 711 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
unison | +0.384 |
latable | +0.381 |
Deal | +0.367 |
igsaw | +0.361 |
ric | +0.360 |
Neg logit contributions | |
said | -0.295 |
whispered | -0.291 |
fell | -0.279 |
woke | -0.266 |
meant | -0.263 |
Token | a |
Token ID | 257 |
Feature activation | -2.18e-01 |
Pos logit contributions | |
unison | +0.316 |
ric | +0.315 |
latable | +0.308 |
Deal | +0.301 |
igsaw | +0.295 |
Neg logit contributions | |
whispered | -0.216 |
said | -0.206 |
woke | -0.195 |
fell | -0.193 |
l | -0.189 |
Token | joke |
Token ID | 9707 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
unison | +0.354 |
latable | +0.341 |
ric | +0.341 |
Deal | +0.327 |
igsaw | +0.323 |
Neg logit contributions | |
whispered | -0.260 |
said | -0.240 |
woke | -0.232 |
fell | -0.230 |
meant | -0.228 |
Token | € |
Token ID | 26391 |
Feature activation | -4.93e-01 |
Pos logit contributions | |
prick | +0.122 |
chy | +0.110 |
iv | +0.108 |
looking | +0.106 |
whispered | +0.105 |
Neg logit contributions | |
ric | -0.187 |
unison | -0.179 |
teammates | -0.164 |
latable | -0.163 |
ap | -0.159 |
Token | � |
Token ID | 129 |
Feature activation | -3.90e-01 |
Pos logit contributions | |
unison | +0.845 |
latable | +0.840 |
ric | +0.800 |
Deal | +0.792 |
igsaw | +0.790 |
Neg logit contributions | |
whispered | -0.540 |
said | -0.528 |
fell | -0.510 |
woke | -0.490 |
chy | -0.470 |
Token | � |
Token ID | 241 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
unison | +0.521 |
ric | +0.501 |
latable | +0.498 |
Deal | +0.478 |
igsaw | +0.464 |
Neg logit contributions | |
whispered | -0.576 |
said | -0.546 |
woke | -0.529 |
fell | -0.525 |
chy | -0.522 |
Token | Let |
Token ID | 5756 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
unison | +0.635 |
ric | +0.626 |
latable | +0.610 |
Deal | +0.596 |
igsaw | +0.578 |
Neg logit contributions | |
whispered | -0.454 |
said | -0.418 |
woke | -0.407 |
fell | -0.400 |
chy | -0.398 |
Token | â |
Token ID | 22940 |
Feature activation | -2.80e-01 |
Pos logit contributions | |
unison | +0.468 |
ric | +0.465 |
latable | +0.455 |
Deal | +0.445 |
igsaw | +0.436 |
Neg logit contributions | |
whispered | -0.265 |
said | -0.260 |
woke | -0.241 |
fell | -0.235 |
meant | -0.232 |
Token | € |
Token ID | 26391 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
unison | +0.355 |
ric | +0.341 |
latable | +0.338 |
Deal | +0.324 |
igsaw | +0.314 |
Neg logit contributions | |
whispered | -0.434 |
said | -0.413 |
woke | -0.401 |
fell | -0.398 |
chy | -0.395 |
Token | ™ |
Token ID | 8151 |
Feature activation | -2.95e-01 |
Pos logit contributions | |
unison | +0.528 |
latable | +0.519 |
Deal | +0.496 |
ric | +0.489 |
igsaw | +0.478 |
Neg logit contributions | |
whispered | -0.396 |
said | -0.396 |
fell | -0.393 |
woke | -0.371 |
l | -0.362 |
Token | s |
Token ID | 82 |
Feature activation | -2.36e-01 |
Pos logit contributions | |
ric | +0.522 |
unison | +0.481 |
igsaw | +0.464 |
latable | +0.464 |
Deal | +0.463 |
Neg logit contributions | |
whispered | -0.318 |
prick | -0.288 |
chy | -0.271 |
woke | -0.269 |
spark | -0.256 |
Token | have |
Token ID | 423 |
Feature activation | -2.22e-01 |
Pos logit contributions | |
unison | +0.384 |
latable | +0.381 |
Deal | +0.367 |
igsaw | +0.361 |
ric | +0.360 |
Neg logit contributions | |
said | -0.295 |
whispered | -0.291 |
fell | -0.279 |
woke | -0.266 |
meant | -0.263 |
Token | a |
Token ID | 257 |
Feature activation | -2.14e-01 |
Pos logit contributions | |
ric | +0.349 |
unison | +0.343 |
latable | +0.334 |
Deal | +0.326 |
igsaw | +0.313 |
Neg logit contributions | |
whispered | -0.292 |
said | -0.285 |
woke | -0.262 |
fell | -0.261 |
meant | -0.254 |
Token | stretching |
Token ID | 20880 |
Feature activation | -1.56e-01 |
Pos logit contributions | |
unison | +0.348 |
ric | +0.343 |
latable | +0.336 |
Deal | +0.327 |
igsaw | +0.317 |
Neg logit contributions | |
whispered | -0.252 |
said | -0.236 |
woke | -0.228 |
fell | -0.225 |
chy | -0.223 |
Token | € |
Token ID | 26391 |
Feature activation | -4.93e-01 |
Pos logit contributions | |
whispered | +0.107 |
said | +0.100 |
woke | +0.096 |
fell | +0.095 |
chy | +0.094 |
Neg logit contributions | |
unison | -0.155 |
ric | -0.151 |
latable | -0.150 |
Deal | -0.145 |
igsaw | -0.142 |
Token | � |
Token ID | 129 |
Feature activation | -3.90e-01 |
Pos logit contributions | |
unison | +0.845 |
latable | +0.840 |
ric | +0.800 |
Deal | +0.792 |
igsaw | +0.790 |
Neg logit contributions | |
whispered | -0.540 |
said | -0.528 |
fell | -0.510 |
woke | -0.490 |
chy | -0.470 |
Token | � |
Token ID | 241 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
unison | +0.521 |
ric | +0.501 |
latable | +0.498 |
Deal | +0.478 |
igsaw | +0.464 |
Neg logit contributions | |
whispered | -0.576 |
said | -0.546 |
woke | -0.529 |
fell | -0.525 |
chy | -0.522 |
Token | please |
Token ID | 29688 |
Feature activation | -2.28e-01 |
Pos logit contributions | |
unison | +0.635 |
ric | +0.626 |
latable | +0.610 |
Deal | +0.596 |
igsaw | +0.578 |
Neg logit contributions | |
whispered | -0.454 |
said | -0.418 |
woke | -0.407 |
fell | -0.400 |
chy | -0.398 |
Token | â |
Token ID | 22940 |
Feature activation | -2.79e-01 |
Pos logit contributions | |
ric | +0.322 |
unison | +0.299 |
Deal | +0.286 |
yeah | +0.284 |
ap | +0.278 |
Neg logit contributions | |
whispered | -0.308 |
chy | -0.301 |
woke | -0.291 |
aisle | -0.288 |
prick | -0.283 |
Token | € |
Token ID | 26391 |
Feature activation | -3.26e-01 |
Pos logit contributions | |
ric | +0.330 |
unison | +0.325 |
latable | +0.314 |
Deal | +0.310 |
igsaw | +0.296 |
Neg logit contributions | |
whispered | -0.456 |
woke | -0.424 |
said | -0.422 |
fell | -0.419 |
l | -0.410 |
Token | and |
Token ID | 290 |
Feature activation | -2.24e-01 |
Pos logit contributions | |
unison | +0.508 |
latable | +0.492 |
ric | +0.489 |
Deal | +0.474 |
igsaw | +0.460 |
Neg logit contributions | |
whispered | -0.409 |
said | -0.389 |
fell | -0.375 |
woke | -0.374 |
l | -0.365 |
Token | � |
Token ID | 6184 |
Feature activation | -2.20e-01 |
Pos logit contributions | |
unison | +0.344 |
ric | +0.335 |
latable | +0.332 |
Deal | +0.321 |
igsaw | +0.313 |
Neg logit contributions | |
whispered | -0.284 |
said | -0.265 |
woke | -0.257 |
fell | -0.255 |
chy | -0.253 |
Token | � |
Token ID | 95 |
Feature activation | -2.13e-01 |
Pos logit contributions | |
unison | +0.278 |
ric | +0.270 |
latable | +0.262 |
Deal | +0.255 |
igsaw | +0.245 |
Neg logit contributions | |
whispered | -0.331 |
said | -0.321 |
l | -0.308 |
fell | -0.307 |
woke | -0.303 |
Token | € |
Token ID | 26391 |
Feature activation | -2.84e-01 |
Pos logit contributions | |
unison | +0.271 |
ric | +0.261 |
latable | +0.255 |
Deal | +0.247 |
igsaw | +0.237 |
Neg logit contributions | |
whispered | -0.325 |
said | -0.311 |
woke | -0.300 |
chy | -0.297 |
fell | -0.297 |
Token | � |
Token ID | 129 |
Feature activation | -2.36e-01 |
Pos logit contributions | |
unison | +0.451 |
latable | +0.439 |
ric | +0.433 |
Deal | +0.421 |
igsaw | +0.411 |
Neg logit contributions | |
whispered | -0.349 |
said | -0.333 |
fell | -0.324 |
woke | -0.321 |
l | -0.313 |
Token | , |
Token ID | 11 |
Feature activation | -3.57e-01 |
Pos logit contributions | |
whispered | +0.139 |
said | +0.130 |
woke | +0.128 |
fell | +0.127 |
chy | +0.126 |
Neg logit contributions | |
unison | -0.123 |
ric | -0.118 |
latable | -0.117 |
Deal | -0.112 |
igsaw | -0.109 |
Token | the |
Token ID | 262 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
unison | +0.580 |
ric | +0.568 |
latable | +0.552 |
Deal | +0.542 |
igsaw | +0.520 |
Neg logit contributions | |
whispered | -0.408 |
chy | -0.367 |
said | -0.365 |
woke | -0.364 |
l | -0.360 |
Token | little |
Token ID | 1310 |
Feature activation | -2.94e-01 |
Pos logit contributions | |
unison | +0.556 |
ric | +0.542 |
latable | +0.522 |
Deal | +0.515 |
igsaw | +0.504 |
Neg logit contributions | |
whispered | -0.365 |
said | -0.328 |
woke | -0.323 |
fell | -0.321 |
meant | -0.317 |
Token | girl |
Token ID | 2576 |
Feature activation | -2.42e-01 |
Pos logit contributions | |
unison | +0.565 |
ric | +0.553 |
latable | +0.548 |
Deal | +0.533 |
igsaw | +0.524 |
Neg logit contributions | |
whispered | -0.259 |
said | -0.234 |
woke | -0.225 |
fell | -0.221 |
chy | -0.219 |
Token | â |
Token ID | 22940 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
unison | +0.444 |
ric | +0.431 |
ap | +0.398 |
latable | +0.394 |
yeah | +0.386 |
Neg logit contributions | |
prick | -0.182 |
chy | -0.180 |
l | -0.175 |
whispered | -0.170 |
aisle | -0.169 |
Token | € |
Token ID | 26391 |
Feature activation | -3.24e-01 |
Pos logit contributions | |
ric | +0.340 |
unison | +0.337 |
latable | +0.318 |
Deal | +0.307 |
igsaw | +0.304 |
Neg logit contributions | |
whispered | -0.448 |
l | -0.434 |
prick | -0.427 |
fell | -0.420 |
woke | -0.417 |
Token | ™ |
Token ID | 8151 |
Feature activation | -2.85e-01 |
Pos logit contributions | |
unison | +0.610 |
latable | +0.606 |
Deal | +0.593 |
ric | +0.582 |
igsaw | +0.573 |
Neg logit contributions | |
whispered | -0.301 |
said | -0.288 |
fell | -0.279 |
woke | -0.269 |
l | -0.254 |
Token | s |
Token ID | 82 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
ric | +0.458 |
unison | +0.411 |
ifying | +0.392 |
yeah | +0.382 |
Deal | +0.380 |
Neg logit contributions | |
whispered | -0.353 |
prick | -0.346 |
chy | -0.321 |
spark | -0.310 |
meant | -0.306 |
Token | l |
Token ID | 300 |
Feature activation | -2.16e-01 |
Pos logit contributions | |
unison | +0.345 |
ric | +0.335 |
latable | +0.326 |
Deal | +0.316 |
igsaw | +0.307 |
Neg logit contributions | |
whispered | -0.323 |
said | -0.298 |
woke | -0.292 |
chy | -0.290 |
fell | -0.289 |
Token | imes |
Token ID | 999 |
Feature activation | -1.85e-01 |
Pos logit contributions | |
unison | +0.294 |
Deal | +0.280 |
ric | +0.276 |
latable | +0.274 |
igsaw | +0.266 |
Neg logit contributions | |
whispered | -0.317 |
said | -0.297 |
woke | -0.297 |
meant | -0.287 |
fell | -0.286 |
Token | were |
Token ID | 547 |
Feature activation | -1.99e-01 |
Pos logit contributions | |
ifying | +0.243 |
unison | +0.236 |
ric | +0.224 |
ap | +0.223 |
such | +0.222 |
Neg logit contributions | |
whispered | -0.196 |
prick | -0.188 |
low | -0.187 |
aisle | -0.180 |
woke | -0.179 |
Token | � |
Token ID | 129 |
Feature activation | -4.32e-01 |
Pos logit contributions | |
whispered | +0.230 |
said | +0.225 |
fell | +0.218 |
chy | +0.217 |
woke | +0.214 |
Neg logit contributions | |
ric | -0.205 |
unison | -0.205 |
Deal | -0.189 |
latable | -0.183 |
yeah | -0.173 |
Token | � |
Token ID | 241 |
Feature activation | -4.07e-01 |
Pos logit contributions | |
unison | +0.657 |
latable | +0.613 |
Deal | +0.603 |
ric | +0.598 |
igsaw | +0.590 |
Neg logit contributions | |
whispered | -0.575 |
said | -0.557 |
fell | -0.534 |
woke | -0.531 |
l | -0.526 |
Token | Please |
Token ID | 5492 |
Feature activation | -2.93e-01 |
Pos logit contributions | |
unison | +0.706 |
ric | +0.686 |
latable | +0.683 |
Deal | +0.663 |
igsaw | +0.650 |
Neg logit contributions | |
whispered | -0.437 |
said | -0.406 |
woke | -0.390 |
fell | -0.386 |
chy | -0.381 |
Token | don |
Token ID | 836 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
unison | +0.444 |
ric | +0.433 |
latable | +0.422 |
Deal | +0.420 |
yeah | +0.394 |
Neg logit contributions | |
whispered | -0.374 |
said | -0.340 |
chy | -0.337 |
woke | -0.337 |
l | -0.333 |
Token | â |
Token ID | 22940 |
Feature activation | -2.74e-01 |
Pos logit contributions | |
unison | +0.491 |
ric | +0.474 |
latable | +0.467 |
yeah | +0.438 |
igsaw | +0.437 |
Neg logit contributions | |
said | -0.286 |
whispered | -0.280 |
fell | -0.276 |
l | -0.273 |
meant | -0.266 |
Token | € |
Token ID | 26391 |
Feature activation | -3.24e-01 |
Pos logit contributions | |
unison | +0.346 |
ric | +0.334 |
latable | +0.331 |
Deal | +0.319 |
igsaw | +0.308 |
Neg logit contributions | |
whispered | -0.426 |
said | -0.402 |
woke | -0.393 |
fell | -0.390 |
chy | -0.387 |
Token | ™ |
Token ID | 8151 |
Feature activation | -2.88e-01 |
Pos logit contributions | |
unison | +0.505 |
latable | +0.491 |
ric | +0.482 |
Deal | +0.466 |
igsaw | +0.458 |
Neg logit contributions | |
whispered | -0.410 |
said | -0.397 |
fell | -0.384 |
woke | -0.376 |
l | -0.366 |
Token | t |
Token ID | 83 |
Feature activation | -2.36e-01 |
Pos logit contributions | |
ric | +0.498 |
unison | +0.459 |
Deal | +0.457 |
ifying | +0.449 |
igsaw | +0.448 |
Neg logit contributions | |
whispered | -0.318 |
prick | -0.287 |
woke | -0.266 |
spark | -0.265 |
buzz | -0.259 |
Token | eat |
Token ID | 4483 |
Feature activation | -1.56e-01 |
Pos logit contributions | |
ric | +0.333 |
unison | +0.329 |
latable | +0.311 |
Deal | +0.308 |
igsaw | +0.301 |
Neg logit contributions | |
whispered | -0.321 |
said | -0.290 |
woke | -0.285 |
chy | -0.285 |
l | -0.283 |
Token | me |
Token ID | 502 |
Feature activation | -1.67e-01 |
Pos logit contributions | |
unison | +0.198 |
ric | +0.193 |
latable | +0.191 |
Deal | +0.183 |
igsaw | +0.177 |
Neg logit contributions | |
whispered | -0.238 |
said | -0.220 |
woke | -0.219 |
fell | -0.216 |
chy | -0.216 |
Token | , |
Token ID | 11 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
ric | +0.261 |
unison | +0.260 |
Deal | +0.250 |
latable | +0.250 |
igsaw | +0.240 |
Neg logit contributions | |
whispered | -0.191 |
woke | -0.175 |
l | -0.174 |
chy | -0.169 |
said | -0.168 |
Token | very |
Token ID | 845 |
Feature activation | -3.60e-01 |
Pos logit contributions | |
whispered | +0.077 |
l | +0.064 |
said | +0.061 |
shouted | +0.058 |
cried | +0.057 |
Neg logit contributions | |
latable | -0.166 |
unison | -0.160 |
ric | -0.160 |
igsaw | -0.156 |
ap | -0.151 |
Token | much |
Token ID | 881 |
Feature activation | -3.05e-01 |
Pos logit contributions | |
unison | +0.539 |
Deal | +0.536 |
latable | +0.509 |
gonna | +0.505 |
ric | +0.503 |
Neg logit contributions | |
whispered | -0.384 |
woke | -0.375 |
chy | -0.358 |
meant | -0.344 |
aisle | -0.330 |
Token | . |
Token ID | 13 |
Feature activation | -3.18e-01 |
Pos logit contributions | |
unison | +0.475 |
igsaw | +0.461 |
latable | +0.457 |
ric | +0.433 |
Deal | +0.432 |
Neg logit contributions | |
whispered | -0.324 |
woke | -0.309 |
chy | -0.292 |
fell | -0.281 |
l | -0.274 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.90e-01 |
Pos logit contributions | |
unison | +0.498 |
ric | +0.482 |
latable | +0.474 |
Deal | +0.466 |
igsaw | +0.450 |
Neg logit contributions | |
whispered | -0.397 |
said | -0.367 |
chy | -0.360 |
woke | -0.360 |
fell | -0.358 |
Token | Once |
Token ID | 7454 |
Feature activation | -2.66e-01 |
Pos logit contributions | |
ric | +0.377 |
unison | +0.354 |
Deal | +0.335 |
igsaw | +0.331 |
latable | +0.331 |
Neg logit contributions | |
whispered | -0.420 |
chy | -0.405 |
prick | -0.391 |
meant | -0.361 |
wil | -0.354 |
Token | upon |
Token ID | 2402 |
Feature activation | -3.23e-01 |
Pos logit contributions | |
ric | +0.282 |
unison | +0.278 |
latable | +0.261 |
Deal | +0.259 |
igsaw | +0.254 |
Neg logit contributions | |
whispered | -0.454 |
said | -0.421 |
woke | -0.417 |
chy | -0.413 |
meant | -0.411 |
Token | a |
Token ID | 257 |
Feature activation | -2.28e-01 |
Pos logit contributions | |
ric | +0.642 |
unison | +0.615 |
latable | +0.609 |
Deal | +0.607 |
igsaw | +0.607 |
Neg logit contributions | |
said | -0.314 |
whispered | -0.313 |
fell | -0.300 |
woke | -0.300 |
low | -0.265 |
Token | time |
Token ID | 640 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
unison | +0.444 |
ric | +0.442 |
latable | +0.428 |
Deal | +0.419 |
igsaw | +0.414 |
Neg logit contributions | |
whispered | -0.192 |
said | -0.178 |
woke | -0.173 |
fell | -0.170 |
chy | -0.167 |
Token | , |
Token ID | 11 |
Feature activation | -2.10e-01 |
Pos logit contributions | |
unison | +0.428 |
ric | +0.414 |
latable | +0.412 |
Deal | +0.399 |
igsaw | +0.390 |
Neg logit contributions | |
whispered | -0.159 |
said | -0.149 |
fell | -0.142 |
woke | -0.140 |
l | -0.133 |
Token | there |
Token ID | 612 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
unison | +0.333 |
ric | +0.328 |
latable | +0.322 |
Deal | +0.312 |
igsaw | +0.304 |
Neg logit contributions | |
whispered | -0.253 |
said | -0.231 |
woke | -0.227 |
chy | -0.221 |
l | -0.220 |
Token | was |
Token ID | 373 |
Feature activation | -1.87e-01 |
Pos logit contributions | |
ric | +0.340 |
unison | +0.328 |
latable | +0.326 |
yeah | +0.302 |
ap | +0.302 |
Neg logit contributions | |
whispered | -0.187 |
chy | -0.160 |
l | -0.155 |
prick | -0.147 |
aisle | -0.137 |
Token | € |
Token ID | 26391 |
Feature activation | -4.93e-01 |
Pos logit contributions | |
whispered | +0.028 |
said | +0.026 |
fell | +0.025 |
woke | +0.025 |
chy | +0.025 |
Neg logit contributions | |
unison | -0.037 |
ric | -0.037 |
latable | -0.036 |
Deal | -0.035 |
igsaw | -0.034 |
Token | � |
Token ID | 129 |
Feature activation | -3.90e-01 |
Pos logit contributions | |
unison | +0.845 |
latable | +0.840 |
ric | +0.800 |
Deal | +0.792 |
igsaw | +0.790 |
Neg logit contributions | |
whispered | -0.540 |
said | -0.528 |
fell | -0.510 |
woke | -0.490 |
chy | -0.470 |
Token | � |
Token ID | 241 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
unison | +0.521 |
ric | +0.501 |
latable | +0.498 |
Deal | +0.478 |
igsaw | +0.464 |
Neg logit contributions | |
whispered | -0.576 |
said | -0.546 |
woke | -0.529 |
fell | -0.525 |
chy | -0.522 |
Token | Who |
Token ID | 8241 |
Feature activation | -2.59e-01 |
Pos logit contributions | |
unison | +0.635 |
ric | +0.626 |
latable | +0.610 |
Deal | +0.596 |
igsaw | +0.578 |
Neg logit contributions | |
whispered | -0.454 |
said | -0.418 |
woke | -0.407 |
fell | -0.400 |
chy | -0.398 |
Token | â |
Token ID | 22940 |
Feature activation | -2.73e-01 |
Pos logit contributions | |
unison | +0.445 |
ric | +0.431 |
latable | +0.425 |
Deal | +0.411 |
igsaw | +0.402 |
Neg logit contributions | |
whispered | -0.272 |
said | -0.244 |
chy | -0.244 |
l | -0.240 |
woke | -0.239 |
Token | € |
Token ID | 26391 |
Feature activation | -3.22e-01 |
Pos logit contributions | |
unison | +0.313 |
ric | +0.312 |
latable | +0.300 |
Deal | +0.289 |
igsaw | +0.282 |
Neg logit contributions | |
whispered | -0.452 |
said | -0.426 |
woke | -0.419 |
fell | -0.417 |
chy | -0.414 |
Token | ™ |
Token ID | 8151 |
Feature activation | -2.92e-01 |
Pos logit contributions | |
unison | +0.502 |
latable | +0.488 |
ric | +0.482 |
Deal | +0.467 |
igsaw | +0.451 |
Neg logit contributions | |
whispered | -0.408 |
said | -0.389 |
fell | -0.377 |
woke | -0.376 |
l | -0.363 |
Token | s |
Token ID | 82 |
Feature activation | -2.33e-01 |
Pos logit contributions | |
ric | +0.509 |
unison | +0.444 |
igsaw | +0.437 |
ifying | +0.436 |
Deal | +0.436 |
Neg logit contributions | |
whispered | -0.332 |
prick | -0.304 |
chy | -0.280 |
woke | -0.278 |
spark | -0.271 |
Token | there |
Token ID | 612 |
Feature activation | -2.16e-01 |
Pos logit contributions | |
unison | +0.320 |
latable | +0.307 |
ric | +0.304 |
Deal | +0.297 |
igsaw | +0.285 |
Neg logit contributions | |
whispered | -0.339 |
said | -0.326 |
woke | -0.314 |
fell | -0.310 |
meant | -0.307 |
Token | ? |
Token ID | 30 |
Feature activation | -2.67e-01 |
Pos logit contributions | |
ric | +0.253 |
unison | +0.231 |
latable | +0.223 |
Deal | +0.214 |
yeah | +0.212 |
Neg logit contributions | |
whispered | -0.350 |
chy | -0.320 |
prick | -0.313 |
l | -0.312 |
fell | -0.309 |
Token | â |
Token ID | 22940 |
Feature activation | -2.17e-01 |
Pos logit contributions | |
unison | +0.359 |
ric | +0.348 |
latable | +0.343 |
Deal | +0.332 |
igsaw | +0.322 |
Neg logit contributions | |
whispered | -0.389 |
said | -0.367 |
woke | -0.358 |
fell | -0.355 |
chy | -0.354 |
Token | � |
Token ID | 129 |
Feature activation | -4.32e-01 |
Pos logit contributions | |
whispered | +0.034 |
chy | +0.033 |
l | +0.032 |
woke | +0.031 |
prick | +0.031 |
Neg logit contributions | |
unison | -0.017 |
ric | -0.016 |
latable | -0.016 |
Deal | -0.016 |
gonna | -0.014 |
Token | � |
Token ID | 241 |
Feature activation | -4.07e-01 |
Pos logit contributions | |
unison | +0.657 |
latable | +0.613 |
Deal | +0.603 |
ric | +0.598 |
igsaw | +0.590 |
Neg logit contributions | |
whispered | -0.575 |
said | -0.557 |
fell | -0.534 |
woke | -0.531 |
l | -0.526 |
Token | Why |
Token ID | 5195 |
Feature activation | -3.13e-01 |
Pos logit contributions | |
unison | +0.706 |
ric | +0.686 |
latable | +0.683 |
Deal | +0.663 |
igsaw | +0.650 |
Neg logit contributions | |
whispered | -0.437 |
said | -0.406 |
woke | -0.390 |
fell | -0.386 |
chy | -0.381 |
Token | don |
Token ID | 836 |
Feature activation | -2.67e-01 |
Pos logit contributions | |
unison | +0.487 |
latable | +0.470 |
ric | +0.470 |
Deal | +0.453 |
igsaw | +0.444 |
Neg logit contributions | |
whispered | -0.395 |
said | -0.375 |
woke | -0.361 |
fell | -0.357 |
chy | -0.353 |
Token | â |
Token ID | 22940 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
unison | +0.573 |
latable | +0.538 |
ric | +0.535 |
Deal | +0.516 |
Majesty | +0.509 |
Neg logit contributions | |
said | -0.228 |
fell | -0.222 |
l | -0.216 |
meant | -0.214 |
whispered | -0.203 |
Token | € |
Token ID | 26391 |
Feature activation | -3.21e-01 |
Pos logit contributions | |
unison | +0.337 |
ric | +0.329 |
latable | +0.324 |
Deal | +0.313 |
igsaw | +0.301 |
Neg logit contributions | |
whispered | -0.421 |
said | -0.395 |
woke | -0.388 |
fell | -0.386 |
l | -0.384 |
Token | ™ |
Token ID | 8151 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
unison | +0.494 |
latable | +0.478 |
ric | +0.470 |
Deal | +0.456 |
igsaw | +0.446 |
Neg logit contributions | |
whispered | -0.415 |
said | -0.399 |
fell | -0.386 |
woke | -0.384 |
l | -0.369 |
Token | t |
Token ID | 83 |
Feature activation | -2.33e-01 |
Pos logit contributions | |
ric | +0.477 |
ifying | +0.426 |
unison | +0.424 |
Deal | +0.420 |
yeah | +0.420 |
Neg logit contributions | |
whispered | -0.322 |
prick | -0.305 |
spark | -0.272 |
buzz | -0.268 |
chy | -0.263 |
Token | you |
Token ID | 345 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
ric | +0.337 |
unison | +0.334 |
latable | +0.320 |
igsaw | +0.308 |
Deal | +0.308 |
Neg logit contributions | |
whispered | -0.311 |
said | -0.283 |
woke | -0.278 |
chy | -0.277 |
fell | -0.276 |
Token | come |
Token ID | 1282 |
Feature activation | -1.59e-01 |
Pos logit contributions | |
ric | +0.279 |
unison | +0.274 |
Deal | +0.258 |
yeah | +0.248 |
latable | +0.246 |
Neg logit contributions | |
whispered | -0.281 |
aisle | -0.277 |
l | -0.266 |
chy | -0.258 |
prick | -0.257 |
Token | with |
Token ID | 351 |
Feature activation | -1.85e-01 |
Pos logit contributions | |
unison | +0.280 |
latable | +0.271 |
ric | +0.270 |
Deal | +0.261 |
igsaw | +0.258 |
Neg logit contributions | |
whispered | -0.172 |
said | -0.160 |
woke | -0.153 |
fell | -0.151 |
chy | -0.148 |
Token | but |
Token ID | 475 |
Feature activation | +5.44e-02 |
Pos logit contributions | |
said | +0.064 |
whispered | +0.064 |
meant | +0.058 |
man | +0.057 |
woke | +0.054 |
Neg logit contributions | |
ric | -0.111 |
latable | -0.107 |
unison | -0.106 |
igsaw | -0.106 |
ifying | -0.102 |
Token | you |
Token ID | 345 |
Feature activation | +7.83e-02 |
Pos logit contributions | |
said | +0.077 |
whispered | +0.077 |
meant | +0.071 |
fell | +0.069 |
knew | +0.068 |
Neg logit contributions | |
unison | -0.082 |
latable | -0.079 |
ric | -0.076 |
igsaw | -0.076 |
Deal | -0.073 |
Token | have |
Token ID | 423 |
Feature activation | +8.65e-02 |
Pos logit contributions | |
whispered | +0.086 |
said | +0.081 |
woke | +0.077 |
fell | +0.076 |
meant | +0.075 |
Neg logit contributions | |
unison | -0.137 |
latable | -0.134 |
ric | -0.133 |
Deal | -0.129 |
igsaw | -0.128 |
Token | to |
Token ID | 284 |
Feature activation | +9.51e-02 |
Pos logit contributions | |
whispered | +0.099 |
said | +0.095 |
fell | +0.088 |
woke | +0.088 |
meant | +0.086 |
Neg logit contributions | |
unison | -0.150 |
latable | -0.146 |
ric | -0.145 |
Deal | -0.141 |
igsaw | -0.138 |
Token | be |
Token ID | 307 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
whispered | +0.135 |
said | +0.128 |
woke | +0.122 |
fell | +0.121 |
chy | +0.120 |
Neg logit contributions | |
unison | -0.133 |
ric | -0.132 |
latable | -0.130 |
igsaw | -0.124 |
Deal | -0.123 |
Token | careful |
Token ID | 8161 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
whispered | +0.124 |
said | +0.116 |
fell | +0.107 |
meant | +0.106 |
woke | +0.105 |
Neg logit contributions | |
unison | -0.192 |
ric | -0.190 |
latable | -0.187 |
igsaw | -0.181 |
Deal | -0.178 |
Token | with |
Token ID | 351 |
Feature activation | +9.90e-02 |
Pos logit contributions | |
whispered | +0.065 |
said | +0.060 |
fell | +0.056 |
chy | +0.055 |
meant | +0.052 |
Neg logit contributions | |
unison | -0.247 |
latable | -0.240 |
ric | -0.240 |
Deal | -0.236 |
igsaw | -0.232 |
Token | other |
Token ID | 584 |
Feature activation | +1.22e-01 |
Pos logit contributions | |
whispered | +0.154 |
woke | +0.145 |
said | +0.144 |
fell | +0.143 |
chy | +0.141 |
Neg logit contributions | |
unison | -0.123 |
ric | -0.121 |
latable | -0.116 |
Deal | -0.111 |
igsaw | -0.108 |
Token | people |
Token ID | 661 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
whispered | +0.165 |
said | +0.154 |
fell | +0.153 |
chy | +0.153 |
woke | +0.153 |
Neg logit contributions | |
unison | -0.178 |
ric | -0.167 |
latable | -0.165 |
Deal | -0.163 |
igsaw | -0.156 |
Token | 's |
Token ID | 338 |
Feature activation | +9.53e-02 |
Pos logit contributions | |
whispered | +0.135 |
woke | +0.125 |
fell | +0.119 |
chy | +0.117 |
said | +0.116 |
Neg logit contributions | |
unison | -0.275 |
ric | -0.270 |
latable | -0.260 |
Deal | -0.255 |
igsaw | -0.252 |
Token | things |
Token ID | 1243 |
Feature activation | +1.40e-01 |
Pos logit contributions | |
whispered | +0.141 |
fell | +0.138 |
woke | +0.136 |
chy | +0.136 |
said | +0.132 |
Neg logit contributions | |
unison | -0.119 |
ric | -0.112 |
latable | -0.104 |
Deal | -0.102 |
igsaw | -0.098 |
Token | old |
Token ID | 1468 |
Feature activation | +1.42e-01 |
Pos logit contributions | |
whispered | +0.100 |
woke | +0.094 |
meant | +0.091 |
chy | +0.090 |
fell | +0.089 |
Neg logit contributions | |
igsaw | -0.136 |
latable | -0.118 |
teammates | -0.115 |
unison | -0.112 |
ocol | -0.112 |
Token | . |
Token ID | 13 |
Feature activation | +1.71e-02 |
Pos logit contributions | |
whispered | +0.128 |
chy | +0.122 |
fell | +0.116 |
woke | +0.116 |
l | +0.116 |
Neg logit contributions | |
unison | -0.261 |
ric | -0.247 |
latable | -0.243 |
yeah | -0.232 |
Deal | -0.230 |
Token | Every |
Token ID | 3887 |
Feature activation | +7.77e-02 |
Pos logit contributions | |
whispered | +0.025 |
l | +0.023 |
fell | +0.022 |
chy | +0.022 |
said | +0.022 |
Neg logit contributions | |
unison | -0.023 |
ric | -0.023 |
latable | -0.023 |
Deal | -0.023 |
igsaw | -0.022 |
Token | morning |
Token ID | 3329 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
whispered | +0.075 |
l | +0.069 |
said | +0.062 |
shouted | +0.060 |
spoke | +0.055 |
Neg logit contributions | |
latable | -0.139 |
unison | -0.138 |
igsaw | -0.134 |
ric | -0.129 |
ap | -0.128 |
Token | , |
Token ID | 11 |
Feature activation | +9.20e-02 |
Pos logit contributions | |
whispered | +0.155 |
chy | +0.147 |
meant | +0.135 |
l | +0.134 |
prick | +0.129 |
Neg logit contributions | |
unison | -0.194 |
yeah | -0.190 |
ric | -0.188 |
latable | -0.187 |
ap | -0.186 |
Token | she |
Token ID | 673 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
whispered | +0.062 |
said | +0.055 |
woke | +0.052 |
fell | +0.050 |
chy | +0.050 |
Neg logit contributions | |
unison | -0.197 |
ric | -0.193 |
latable | -0.192 |
Deal | -0.187 |
igsaw | -0.184 |
Token | would |
Token ID | 561 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
l | +0.080 |
whispered | +0.079 |
aisle | +0.072 |
prick | +0.071 |
meant | +0.069 |
Neg logit contributions | |
latable | -0.172 |
ric | -0.170 |
unison | -0.169 |
ap | -0.158 |
hey | -0.156 |
Token | pray |
Token ID | 12472 |
Feature activation | +1.72e-01 |
Pos logit contributions | |
whispered | +0.163 |
said | +0.152 |
l | +0.152 |
meant | +0.147 |
chy | +0.145 |
Neg logit contributions | |
unison | -0.149 |
ric | -0.144 |
latable | -0.144 |
Deal | -0.137 |
igsaw | -0.127 |
Token | . |
Token ID | 13 |
Feature activation | +2.19e-02 |
Pos logit contributions | |
whispered | +0.256 |
l | +0.248 |
fell | +0.240 |
aisle | +0.233 |
said | +0.232 |
Neg logit contributions | |
ric | -0.205 |
unison | -0.202 |
ap | -0.192 |
Deal | -0.187 |
latable | -0.176 |
Token | She |
Token ID | 1375 |
Feature activation | +9.12e-02 |
Pos logit contributions | |
whispered | +0.028 |
l | +0.026 |
said | +0.026 |
fell | +0.025 |
chy | +0.025 |
Neg logit contributions | |
ric | -0.034 |
unison | -0.033 |
Deal | -0.033 |
latable | -0.032 |
igsaw | -0.031 |
Token | would |
Token ID | 561 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
l | +0.093 |
aisle | +0.093 |
prick | +0.083 |
low | +0.078 |
ho | +0.078 |
Neg logit contributions | |
unison | -0.138 |
ric | -0.137 |
latable | -0.122 |
ap | -0.121 |
Deal | -0.119 |
Token | were |
Token ID | 547 |
Feature activation | +9.41e-02 |
Pos logit contributions | |
l | +0.082 |
chy | +0.081 |
dies | +0.076 |
whispered | +0.074 |
man | +0.072 |
Neg logit contributions | |
unison | -0.124 |
latable | -0.113 |
ric | -0.107 |
pieces | -0.102 |
hey | -0.101 |
Token | bright |
Token ID | 6016 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
whispered | +0.120 |
woke | +0.119 |
l | +0.116 |
chy | +0.112 |
fell | +0.109 |
Neg logit contributions | |
unison | -0.138 |
latable | -0.135 |
ric | -0.134 |
Deal | -0.127 |
gonna | -0.122 |
Token | and |
Token ID | 290 |
Feature activation | +9.49e-02 |
Pos logit contributions | |
whispered | +0.081 |
said | +0.075 |
l | +0.065 |
meant | +0.063 |
fell | +0.063 |
Neg logit contributions | |
ric | -0.295 |
unison | -0.291 |
Deal | -0.290 |
latable | -0.280 |
igsaw | -0.275 |
Token | colorful |
Token ID | 20239 |
Feature activation | +1.47e-01 |
Pos logit contributions | |
whispered | +0.098 |
woke | +0.089 |
l | +0.085 |
chy | +0.085 |
said | +0.083 |
Neg logit contributions | |
unison | -0.166 |
latable | -0.160 |
ric | -0.158 |
Deal | -0.153 |
igsaw | -0.151 |
Token | , |
Token ID | 11 |
Feature activation | +8.13e-02 |
Pos logit contributions | |
whispered | +0.115 |
said | +0.101 |
woke | +0.098 |
fell | +0.095 |
chy | +0.094 |
Neg logit contributions | |
unison | -0.299 |
ric | -0.291 |
latable | -0.290 |
Deal | -0.281 |
igsaw | -0.278 |
Token | and |
Token ID | 290 |
Feature activation | +9.06e-02 |
Pos logit contributions | |
whispered | +0.090 |
woke | +0.083 |
l | +0.082 |
chy | +0.081 |
said | +0.079 |
Neg logit contributions | |
unison | -0.136 |
latable | -0.132 |
ric | -0.127 |
Deal | -0.121 |
igsaw | -0.121 |
Token | smelled |
Token ID | 37609 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
whispered | +0.129 |
woke | +0.120 |
chy | +0.119 |
said | +0.118 |
l | +0.118 |
Neg logit contributions | |
unison | -0.124 |
latable | -0.118 |
ric | -0.116 |
Deal | -0.111 |
igsaw | -0.109 |
Token | so |
Token ID | 523 |
Feature activation | +8.61e-02 |
Pos logit contributions | |
whispered | +0.166 |
said | +0.153 |
woke | +0.149 |
fell | +0.148 |
chy | +0.146 |
Neg logit contributions | |
unison | -0.212 |
ric | -0.205 |
latable | -0.205 |
Deal | -0.196 |
igsaw | -0.192 |
Token | good |
Token ID | 922 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
whispered | +0.101 |
woke | +0.091 |
aisle | +0.085 |
said | +0.084 |
chy | +0.084 |
Neg logit contributions | |
unison | -0.139 |
latable | -0.135 |
igsaw | -0.132 |
Deal | -0.127 |
yeah | -0.126 |
Token | ! |
Token ID | 0 |
Feature activation | +1.68e-02 |
Pos logit contributions | |
whispered | +0.097 |
said | +0.089 |
woke | +0.079 |
fell | +0.079 |
meant | +0.077 |
Neg logit contributions | |
unison | -0.315 |
ric | -0.311 |
latable | -0.306 |
Deal | -0.302 |
igsaw | -0.293 |
Token | When |
Token ID | 1649 |
Feature activation | +6.72e-02 |
Pos logit contributions | |
whispered | +0.022 |
said | +0.021 |
woke | +0.020 |
fell | +0.020 |
meant | +0.019 |
Neg logit contributions | |
unison | -0.025 |
ric | -0.025 |
latable | -0.024 |
Deal | -0.023 |
igsaw | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | -5.09e-02 |
Pos logit contributions | |
unison | +0.059 |
ric | +0.058 |
latable | +0.054 |
Deal | +0.053 |
igsaw | +0.047 |
Neg logit contributions | |
chy | -0.071 |
whispered | -0.070 |
prick | -0.065 |
meant | -0.065 |
woke | -0.064 |
Token | They |
Token ID | 2990 |
Feature activation | +2.65e-02 |
Pos logit contributions | |
unison | +0.074 |
latable | +0.070 |
ric | +0.070 |
Deal | +0.068 |
igsaw | +0.065 |
Neg logit contributions | |
whispered | -0.070 |
said | -0.066 |
woke | -0.064 |
chy | -0.064 |
fell | -0.064 |
Token | waited |
Token ID | 13488 |
Feature activation | +8.38e-02 |
Pos logit contributions | |
l | +0.033 |
chy | +0.031 |
prick | +0.030 |
iv | +0.029 |
dies | +0.028 |
Neg logit contributions | |
unison | -0.041 |
ric | -0.034 |
hey | -0.034 |
ap | -0.032 |
pieces | -0.032 |
Token | for |
Token ID | 329 |
Feature activation | +6.64e-02 |
Pos logit contributions | |
whispered | +0.116 |
woke | +0.107 |
meant | +0.106 |
said | +0.106 |
l | +0.106 |
Neg logit contributions | |
unison | -0.118 |
latable | -0.111 |
ric | -0.108 |
Deal | -0.107 |
igsaw | -0.101 |
Token | their |
Token ID | 511 |
Feature activation | +6.54e-02 |
Pos logit contributions | |
whispered | +0.084 |
said | +0.080 |
woke | +0.075 |
fell | +0.075 |
meant | +0.073 |
Neg logit contributions | |
ric | -0.103 |
unison | -0.101 |
latable | -0.100 |
Deal | -0.097 |
igsaw | -0.095 |
Token | mom |
Token ID | 1995 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
whispered | +0.101 |
woke | +0.094 |
meant | +0.093 |
l | +0.093 |
chy | +0.092 |
Neg logit contributions | |
unison | -0.086 |
latable | -0.077 |
ric | -0.075 |
Deal | -0.075 |
igsaw | -0.074 |
Token | to |
Token ID | 284 |
Feature activation | +7.44e-02 |
Pos logit contributions | |
whispered | +0.144 |
prick | +0.137 |
fell | +0.136 |
l | +0.135 |
woke | +0.133 |
Neg logit contributions | |
unison | -0.138 |
ric | -0.137 |
latable | -0.130 |
pieces | -0.127 |
igsaw | -0.126 |
Token | come |
Token ID | 1282 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
whispered | +0.103 |
said | +0.097 |
woke | +0.094 |
fell | +0.093 |
chy | +0.092 |
Neg logit contributions | |
unison | -0.106 |
ric | -0.104 |
latable | -0.103 |
Deal | -0.099 |
igsaw | -0.097 |
Token | home |
Token ID | 1363 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
whispered | +0.096 |
said | +0.088 |
fell | +0.082 |
woke | +0.082 |
chy | +0.077 |
Neg logit contributions | |
unison | -0.216 |
ric | -0.214 |
latable | -0.213 |
Deal | -0.207 |
igsaw | -0.203 |
Token | from |
Token ID | 422 |
Feature activation | +5.83e-02 |
Pos logit contributions | |
whispered | +0.139 |
woke | +0.135 |
l | +0.134 |
chy | +0.132 |
meant | +0.124 |
Neg logit contributions | |
unison | -0.196 |
ap | -0.179 |
ric | -0.178 |
igsaw | -0.177 |
latable | -0.171 |
Token | work |
Token ID | 670 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
whispered | +0.094 |
l | +0.090 |
chy | +0.089 |
woke | +0.089 |
said | +0.087 |
Neg logit contributions | |
unison | -0.070 |
ric | -0.063 |
latable | -0.062 |
igsaw | -0.062 |
Deal | -0.058 |
Token | hungry |
Token ID | 14720 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
whispered | +0.052 |
aisle | +0.050 |
woke | +0.048 |
books | +0.048 |
meant | +0.046 |
Neg logit contributions | |
Deal | -0.057 |
ric | -0.056 |
hey | -0.055 |
igsaw | -0.054 |
gonna | -0.054 |
Token | , |
Token ID | 11 |
Feature activation | +4.73e-02 |
Pos logit contributions | |
whispered | +0.123 |
said | +0.111 |
woke | +0.109 |
fell | +0.107 |
l | +0.107 |
Neg logit contributions | |
unison | -0.172 |
latable | -0.169 |
ric | -0.168 |
Deal | -0.162 |
igsaw | -0.160 |
Token | but |
Token ID | 475 |
Feature activation | +3.86e-02 |
Pos logit contributions | |
whispered | +0.058 |
said | +0.054 |
woke | +0.053 |
fell | +0.052 |
chy | +0.051 |
Neg logit contributions | |
unison | -0.076 |
ric | -0.074 |
latable | -0.073 |
Deal | -0.070 |
igsaw | -0.069 |
Token | I |
Token ID | 314 |
Feature activation | +5.58e-02 |
Pos logit contributions | |
whispered | +0.063 |
said | +0.060 |
woke | +0.059 |
fell | +0.058 |
chy | +0.058 |
Neg logit contributions | |
unison | -0.045 |
ric | -0.044 |
latable | -0.043 |
Deal | -0.041 |
igsaw | -0.040 |
Token | don |
Token ID | 836 |
Feature activation | +6.08e-02 |
Pos logit contributions | |
whispered | +0.068 |
said | +0.063 |
woke | +0.061 |
l | +0.060 |
fell | +0.060 |
Neg logit contributions | |
unison | -0.087 |
ric | -0.086 |
latable | -0.085 |
Deal | -0.082 |
igsaw | -0.080 |
Token | 't |
Token ID | 470 |
Feature activation | +5.91e-02 |
Pos logit contributions | |
whispered | +0.082 |
said | +0.078 |
fell | +0.076 |
woke | +0.075 |
l | +0.074 |
Neg logit contributions | |
unison | -0.090 |
ric | -0.087 |
latable | -0.086 |
Deal | -0.082 |
igsaw | -0.081 |
Token | have |
Token ID | 423 |
Feature activation | +7.81e-02 |
Pos logit contributions | |
whispered | +0.093 |
l | +0.088 |
aisle | +0.088 |
said | +0.085 |
woke | +0.085 |
Neg logit contributions | |
ric | -0.063 |
unison | -0.062 |
Deal | -0.061 |
latable | -0.059 |
igsaw | -0.058 |
Token | anything |
Token ID | 1997 |
Feature activation | +9.03e-02 |
Pos logit contributions | |
whispered | +0.133 |
said | +0.126 |
l | +0.124 |
chy | +0.123 |
woke | +0.122 |
Neg logit contributions | |
unison | -0.082 |
ric | -0.080 |
latable | -0.077 |
Deal | -0.074 |
igsaw | -0.072 |
Token | to |
Token ID | 284 |
Feature activation | +7.82e-02 |
Pos logit contributions | |
whispered | +0.122 |
woke | +0.113 |
aisle | +0.113 |
l | +0.109 |
said | +0.107 |
Neg logit contributions | |
Deal | -0.106 |
gonna | -0.105 |
latable | -0.104 |
yeah | -0.103 |
igsaw | -0.102 |
Token | eat |
Token ID | 4483 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
whispered | +0.109 |
l | +0.100 |
chy | +0.099 |
said | +0.099 |
woke | +0.098 |
Neg logit contributions | |
unison | -0.107 |
latable | -0.103 |
ric | -0.103 |
Deal | -0.098 |
igsaw | -0.093 |
Token | ." |
Token ID | 526 |
Feature activation | -1.21e-02 |
Pos logit contributions | |
whispered | +0.181 |
spoke | +0.161 |
woke | +0.160 |
said | +0.158 |
knew | +0.155 |
Neg logit contributions | |
unison | -0.176 |
latable | -0.176 |
ric | -0.174 |
our | -0.173 |
Deal | -0.173 |
Token | bird |
Token ID | 6512 |
Feature activation | +8.40e-02 |
Pos logit contributions | |
whispered | +0.028 |
said | +0.026 |
woke | +0.025 |
fell | +0.024 |
chy | +0.024 |
Neg logit contributions | |
unison | -0.065 |
ric | -0.063 |
latable | -0.063 |
Deal | -0.061 |
igsaw | -0.060 |
Token | . |
Token ID | 13 |
Feature activation | -6.16e-02 |
Pos logit contributions | |
whispered | +0.061 |
said | +0.056 |
woke | +0.052 |
fell | +0.052 |
l | +0.050 |
Neg logit contributions | |
unison | -0.180 |
latable | -0.176 |
Deal | -0.175 |
igsaw | -0.172 |
ric | -0.171 |
Token | The |
Token ID | 383 |
Feature activation | -5.79e-03 |
Pos logit contributions | |
latable | +0.072 |
unison | +0.071 |
igsaw | +0.071 |
Deal | +0.067 |
ric | +0.067 |
Neg logit contributions | |
whispered | -0.091 |
chy | -0.085 |
l | -0.085 |
meant | -0.082 |
fell | -0.081 |
Token | bird |
Token ID | 6512 |
Feature activation | +7.76e-02 |
Pos logit contributions | |
unison | +0.008 |
ric | +0.008 |
latable | +0.008 |
yeah | +0.007 |
hey | +0.007 |
Neg logit contributions | |
whispered | -0.007 |
chy | -0.007 |
fell | -0.006 |
aisle | -0.006 |
meant | -0.006 |
Token | said |
Token ID | 531 |
Feature activation | +2.39e-02 |
Pos logit contributions | |
aisle | +0.075 |
books | +0.062 |
book | +0.059 |
chy | +0.057 |
post | +0.056 |
Neg logit contributions | |
ric | -0.133 |
unison | -0.121 |
gonna | -0.112 |
how | -0.112 |
yeah | -0.112 |
Token | , |
Token ID | 11 |
Feature activation | +7.47e-03 |
Pos logit contributions | |
whispered | +0.024 |
said | +0.022 |
chy | +0.021 |
woke | +0.021 |
meant | +0.020 |
Neg logit contributions | |
unison | -0.042 |
ric | -0.041 |
latable | -0.041 |
Deal | -0.040 |
igsaw | -0.039 |
Token | " |
Token ID | 366 |
Feature activation | -1.35e-03 |
Pos logit contributions | |
whispered | +0.011 |
chy | +0.011 |
aisle | +0.011 |
prick | +0.011 |
meant | +0.010 |
Neg logit contributions | |
ric | -0.009 |
unison | -0.008 |
Deal | -0.008 |
gonna | -0.008 |
latable | -0.007 |
Token | I |
Token ID | 40 |
Feature activation | +9.74e-03 |
Pos logit contributions | |
unison | +0.002 |
ric | +0.002 |
hey | +0.002 |
Deal | +0.002 |
yeah | +0.002 |
Neg logit contributions | |
chy | -0.002 |
whispered | -0.002 |
fell | -0.002 |
said | -0.002 |
aisle | -0.002 |
Token | can |
Token ID | 460 |
Feature activation | +3.60e-02 |
Pos logit contributions | |
whispered | +0.013 |
l | +0.012 |
said | +0.012 |
chy | +0.012 |
aisle | +0.012 |
Neg logit contributions | |
unison | -0.013 |
ric | -0.013 |
latable | -0.013 |
Deal | -0.012 |
igsaw | -0.012 |
Token | 't |
Token ID | 470 |
Feature activation | +2.55e-02 |
Pos logit contributions | |
whispered | +0.054 |
said | +0.051 |
woke | +0.050 |
l | +0.049 |
fell | +0.049 |
Neg logit contributions | |
unison | -0.046 |
ric | -0.045 |
latable | -0.044 |
Deal | -0.043 |
igsaw | -0.040 |
Token | find |
Token ID | 1064 |
Feature activation | +8.26e-02 |
Pos logit contributions | |
whispered | +0.038 |
said | +0.035 |
chy | +0.035 |
l | +0.035 |
meant | +0.034 |
Neg logit contributions | |
unison | -0.031 |
latable | -0.030 |
ric | -0.029 |
Deal | -0.027 |
hey | -0.027 |
Token | ens |
Token ID | 641 |
Feature activation | +8.97e-02 |
Pos logit contributions | |
whispered | +0.046 |
low | +0.041 |
woke | +0.038 |
l | +0.038 |
fell | +0.038 |
Neg logit contributions | |
latable | -0.088 |
unison | -0.084 |
Deal | -0.083 |
igsaw | -0.082 |
ric | -0.081 |
Token | . |
Token ID | 13 |
Feature activation | -6.09e-02 |
Pos logit contributions | |
whispered | +0.109 |
l | +0.106 |
woke | +0.103 |
prick | +0.102 |
low | +0.099 |
Neg logit contributions | |
unison | -0.118 |
pieces | -0.112 |
latable | -0.109 |
ap | -0.107 |
teammates | -0.107 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -3.97e-02 |
Pos logit contributions | |
unison | +0.096 |
ric | +0.095 |
latable | +0.093 |
Deal | +0.090 |
igsaw | +0.088 |
Neg logit contributions | |
whispered | -0.074 |
said | -0.068 |
woke | -0.066 |
chy | -0.066 |
fell | -0.065 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.86e-02 |
Pos logit contributions | |
unison | +0.060 |
ric | +0.058 |
latable | +0.058 |
Deal | +0.056 |
igsaw | +0.054 |
Neg logit contributions | |
whispered | -0.049 |
chy | -0.044 |
prick | -0.042 |
said | -0.042 |
fell | -0.041 |
Token | upon |
Token ID | 2402 |
Feature activation | -5.57e-02 |
Pos logit contributions | |
unison | +0.023 |
ric | +0.022 |
latable | +0.022 |
Deal | +0.021 |
igsaw | +0.020 |
Neg logit contributions | |
whispered | -0.029 |
said | -0.027 |
woke | -0.027 |
chy | -0.026 |
fell | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
ric | +0.113 |
igsaw | +0.112 |
Deal | +0.108 |
unison | +0.108 |
latable | +0.108 |
Neg logit contributions | |
woke | -0.053 |
said | -0.051 |
fell | -0.050 |
whispered | -0.049 |
low | -0.044 |
Token | time |
Token ID | 640 |
Feature activation | +3.25e-02 |
Pos logit contributions | |
whispered | +0.012 |
said | +0.011 |
woke | +0.011 |
fell | +0.011 |
chy | +0.010 |
Neg logit contributions | |
ric | -0.034 |
unison | -0.033 |
latable | -0.032 |
Deal | -0.032 |
igsaw | -0.031 |
Token | , |
Token ID | 11 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
said | +0.017 |
fell | +0.017 |
woke | +0.016 |
whispered | +0.016 |
l | +0.013 |
Neg logit contributions | |
unison | -0.078 |
ric | -0.075 |
latable | -0.075 |
Deal | -0.071 |
ifying | -0.070 |
Token | there |
Token ID | 612 |
Feature activation | +3.54e-02 |
Pos logit contributions | |
whispered | +0.012 |
said | +0.011 |
woke | +0.011 |
fell | +0.011 |
chy | +0.010 |
Neg logit contributions | |
unison | -0.021 |
ric | -0.021 |
latable | -0.021 |
Deal | -0.020 |
igsaw | -0.020 |
Token | was |
Token ID | 373 |
Feature activation | +2.66e-02 |
Pos logit contributions | |
whispered | +0.024 |
chy | +0.018 |
l | +0.018 |
said | +0.018 |
fell | +0.016 |
Neg logit contributions | |
unison | -0.075 |
ric | -0.075 |
latable | -0.074 |
Deal | -0.070 |
yeah | -0.069 |
Token | a |
Token ID | 257 |
Feature activation | +2.29e-02 |
Pos logit contributions | |
said | +0.025 |
whispered | +0.024 |
fell | +0.022 |
woke | +0.022 |
meant | +0.021 |
Neg logit contributions | |
ric | -0.054 |
unison | -0.050 |
yeah | -0.049 |
Deal | -0.049 |
latable | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
whispered | +0.031 |
woke | +0.028 |
fell | +0.027 |
knew | +0.026 |
chy | +0.026 |
Neg logit contributions | |
unison | -0.060 |
ric | -0.058 |
latable | -0.056 |
ap | -0.054 |
ifying | -0.054 |
Token | They |
Token ID | 1119 |
Feature activation | -4.86e-02 |
Pos logit contributions | |
unison | +0.173 |
ric | +0.165 |
latable | +0.162 |
Deal | +0.161 |
igsaw | +0.153 |
Neg logit contributions | |
whispered | -0.156 |
chy | -0.144 |
said | -0.144 |
woke | -0.144 |
fell | -0.143 |
Token | put |
Token ID | 1234 |
Feature activation | +2.47e-02 |
Pos logit contributions | |
unison | +0.085 |
ric | +0.078 |
latable | +0.071 |
teammates | +0.069 |
hey | +0.068 |
Neg logit contributions | |
l | -0.045 |
prick | -0.043 |
aisle | -0.042 |
chy | -0.042 |
whispered | -0.041 |
Token | Lily |
Token ID | 20037 |
Feature activation | +4.89e-03 |
Pos logit contributions | |
whispered | +0.036 |
woke | +0.035 |
chy | +0.034 |
fell | +0.033 |
said | +0.033 |
Neg logit contributions | |
unison | -0.033 |
ric | -0.031 |
latable | -0.031 |
Deal | -0.028 |
ifying | -0.028 |
Token | 's |
Token ID | 338 |
Feature activation | -1.68e-02 |
Pos logit contributions | |
whispered | +0.006 |
l | +0.005 |
meant | +0.005 |
woke | +0.005 |
said | +0.005 |
Neg logit contributions | |
ric | -0.008 |
unison | -0.008 |
latable | -0.007 |
ifying | -0.007 |
igsaw | -0.007 |
Token | cow |
Token ID | 9875 |
Feature activation | +4.72e-02 |
Pos logit contributions | |
unison | +0.025 |
ric | +0.025 |
latable | +0.024 |
Deal | +0.023 |
ifying | +0.022 |
Neg logit contributions | |
whispered | -0.021 |
woke | -0.020 |
said | -0.020 |
fell | -0.019 |
meant | -0.019 |
Token | in |
Token ID | 287 |
Feature activation | -2.06e-02 |
Pos logit contributions | |
whispered | +0.059 |
said | +0.055 |
woke | +0.054 |
fell | +0.053 |
chy | +0.053 |
Neg logit contributions | |
unison | -0.072 |
ric | -0.070 |
latable | -0.069 |
Deal | -0.065 |
igsaw | -0.065 |
Token | the |
Token ID | 262 |
Feature activation | -3.22e-02 |
Pos logit contributions | |
unison | +0.028 |
ric | +0.025 |
latable | +0.025 |
Deal | +0.024 |
igsaw | +0.024 |
Neg logit contributions | |
whispered | -0.031 |
chy | -0.029 |
woke | -0.029 |
fell | -0.029 |
said | -0.029 |
Token | barn |
Token ID | 25203 |
Feature activation | +5.69e-02 |
Pos logit contributions | |
unison | +0.049 |
ric | +0.048 |
latable | +0.047 |
Deal | +0.046 |
igsaw | +0.043 |
Neg logit contributions | |
whispered | -0.043 |
said | -0.040 |
fell | -0.038 |
meant | -0.038 |
l | -0.038 |
Token | . |
Token ID | 13 |
Feature activation | -1.02e-01 |
Pos logit contributions | |
whispered | +0.064 |
chy | +0.064 |
woke | +0.062 |
meant | +0.060 |
prick | +0.060 |
Neg logit contributions | |
ric | -0.085 |
unison | -0.085 |
ap | -0.082 |
latable | -0.075 |
teammates | -0.071 |
Token | They |
Token ID | 1119 |
Feature activation | -3.21e-02 |
Pos logit contributions | |
unison | +0.144 |
latable | +0.134 |
ric | +0.134 |
Deal | +0.133 |
igsaw | +0.125 |
Neg logit contributions | |
whispered | -0.144 |
chy | -0.136 |
woke | -0.134 |
fell | -0.132 |
said | -0.131 |
Token | feels |
Token ID | 5300 |
Feature activation | +2.02e-02 |
Pos logit contributions | |
igsaw | +0.018 |
unison | +0.018 |
ric | +0.017 |
ifying | +0.017 |
pieces | +0.017 |
Neg logit contributions | |
said | -0.007 |
whispered | -0.007 |
woke | -0.006 |
l | -0.006 |
fell | -0.006 |
Token | bad |
Token ID | 2089 |
Feature activation | +2.53e-02 |
Pos logit contributions | |
whispered | +0.021 |
said | +0.021 |
fell | +0.019 |
meant | +0.019 |
chy | +0.019 |
Neg logit contributions | |
unison | -0.037 |
ric | -0.037 |
latable | -0.036 |
igsaw | -0.036 |
Deal | -0.034 |
Token | for |
Token ID | 329 |
Feature activation | +2.65e-02 |
Pos logit contributions | |
whispered | +0.023 |
aisle | +0.020 |
chy | +0.020 |
l | +0.020 |
low | +0.019 |
Neg logit contributions | |
unison | -0.044 |
ric | -0.044 |
Deal | -0.043 |
latable | -0.042 |
yeah | -0.041 |
Token | pin |
Token ID | 6757 |
Feature activation | +7.99e-02 |
Pos logit contributions | |
whispered | +0.036 |
said | +0.034 |
woke | +0.032 |
fell | +0.032 |
chy | +0.032 |
Neg logit contributions | |
unison | -0.039 |
ric | -0.038 |
latable | -0.038 |
Deal | -0.036 |
igsaw | -0.036 |
Token | ching |
Token ID | 10813 |
Feature activation | +6.62e-02 |
Pos logit contributions | |
whispered | +0.096 |
aisle | +0.092 |
woke | +0.089 |
said | +0.085 |
low | +0.082 |
Neg logit contributions | |
ric | -0.114 |
latable | -0.111 |
unison | -0.109 |
Deal | -0.107 |
igsaw | -0.106 |
Token | Lily |
Token ID | 20037 |
Feature activation | +4.67e-02 |
Pos logit contributions | |
whispered | +0.085 |
woke | +0.077 |
chy | +0.075 |
fell | +0.074 |
said | +0.074 |
Neg logit contributions | |
unison | -0.097 |
ric | -0.093 |
Deal | -0.093 |
latable | -0.091 |
igsaw | -0.088 |
Token | . |
Token ID | 13 |
Feature activation | -6.80e-02 |
Pos logit contributions | |
whispered | +0.056 |
woke | +0.053 |
aisle | +0.053 |
spoke | +0.051 |
flowed | +0.050 |
Neg logit contributions | |
ap | -0.061 |
unison | -0.060 |
ric | -0.059 |
pieces | -0.056 |
latable | -0.055 |
Token | He |
Token ID | 679 |
Feature activation | +4.43e-03 |
Pos logit contributions | |
unison | +0.100 |
Deal | +0.100 |
latable | +0.096 |
ric | +0.095 |
igsaw | +0.092 |
Neg logit contributions | |
whispered | -0.088 |
woke | -0.079 |
chy | -0.079 |
said | -0.077 |
fell | -0.077 |
Token | sees |
Token ID | 7224 |
Feature activation | +3.63e-02 |
Pos logit contributions | |
whispered | +0.004 |
said | +0.004 |
woke | +0.004 |
fell | +0.004 |
chy | +0.004 |
Neg logit contributions | |
igsaw | -0.009 |
unison | -0.009 |
ric | -0.009 |
latable | -0.009 |
ifying | -0.008 |
Token | a |
Token ID | 257 |
Feature activation | +1.88e-02 |
Pos logit contributions | |
whispered | +0.038 |
said | +0.037 |
l | +0.034 |
woke | +0.034 |
fell | +0.034 |
Neg logit contributions | |
unison | -0.066 |
ric | -0.064 |
latable | -0.064 |
Deal | -0.062 |
igsaw | -0.060 |
Token | red |
Token ID | 2266 |
Feature activation | +6.49e-02 |
Pos logit contributions | |
whispered | +0.020 |
said | +0.018 |
chy | +0.017 |
meant | +0.017 |
fell | +0.017 |
Neg logit contributions | |
unison | -0.032 |
ric | -0.032 |
latable | -0.030 |
Deal | -0.030 |
igsaw | -0.029 |
Token | lucky |
Token ID | 9670 |
Feature activation | -8.34e-02 |
Pos logit contributions | |
unison | +0.189 |
ric | +0.186 |
latable | +0.183 |
Deal | +0.179 |
igsaw | +0.174 |
Neg logit contributions | |
whispered | -0.109 |
said | -0.103 |
woke | -0.096 |
fell | -0.096 |
meant | -0.094 |
Token | car |
Token ID | 1097 |
Feature activation | -2.23e-02 |
Pos logit contributions | |
unison | +0.133 |
latable | +0.124 |
ric | +0.123 |
igsaw | +0.121 |
yeah | +0.115 |
Neg logit contributions | |
whispered | -0.100 |
l | -0.092 |
chy | -0.090 |
said | -0.088 |
fell | -0.087 |
Token | and |
Token ID | 290 |
Feature activation | -1.00e-01 |
Pos logit contributions | |
unison | +0.036 |
ric | +0.035 |
latable | +0.034 |
yeah | +0.033 |
hey | +0.032 |
Neg logit contributions | |
whispered | -0.022 |
prick | -0.021 |
low | -0.019 |
aisle | -0.019 |
l | -0.019 |
Token | was |
Token ID | 373 |
Feature activation | -9.53e-02 |
Pos logit contributions | |
unison | +0.175 |
ric | +0.170 |
latable | +0.168 |
Deal | +0.161 |
igsaw | +0.160 |
Neg logit contributions | |
whispered | -0.101 |
said | -0.090 |
chy | -0.090 |
l | -0.090 |
woke | -0.089 |
Token | so |
Token ID | 523 |
Feature activation | -1.12e-01 |
Pos logit contributions | |
unison | +0.152 |
ric | +0.149 |
latable | +0.145 |
Deal | +0.140 |
igsaw | +0.136 |
Neg logit contributions | |
whispered | -0.114 |
said | -0.105 |
woke | -0.104 |
l | -0.102 |
fell | -0.102 |
Token | thankful |
Token ID | 25535 |
Feature activation | -8.86e-02 |
Pos logit contributions | |
unison | +0.171 |
ric | +0.169 |
latable | +0.161 |
Deal | +0.160 |
igsaw | +0.157 |
Neg logit contributions | |
whispered | -0.126 |
woke | -0.116 |
chy | -0.115 |
aisle | -0.113 |
l | -0.109 |
Token | for |
Token ID | 329 |
Feature activation | -6.60e-02 |
Pos logit contributions | |
unison | +0.135 |
ric | +0.130 |
latable | +0.128 |
Deal | +0.124 |
igsaw | +0.121 |
Neg logit contributions | |
whispered | -0.112 |
woke | -0.102 |
said | -0.101 |
chy | -0.101 |
fell | -0.100 |
Token | the |
Token ID | 262 |
Feature activation | -9.28e-02 |
Pos logit contributions | |
unison | +0.093 |
ric | +0.084 |
Deal | +0.081 |
latable | +0.080 |
yeah | +0.078 |
Neg logit contributions | |
whispered | -0.095 |
chy | -0.092 |
woke | -0.091 |
l | -0.090 |
fell | -0.090 |
Token | adventure |
Token ID | 8855 |
Feature activation | -6.67e-03 |
Pos logit contributions | |
unison | +0.146 |
ric | +0.136 |
latable | +0.135 |
Deal | +0.132 |
igsaw | +0.131 |
Neg logit contributions | |
whispered | -0.117 |
said | -0.105 |
fell | -0.105 |
woke | -0.104 |
chy | -0.103 |
Token | it |
Token ID | 340 |
Feature activation | -4.12e-02 |
Pos logit contributions | |
unison | +0.011 |
igsaw | +0.010 |
latable | +0.010 |
pieces | +0.009 |
ric | +0.009 |
Neg logit contributions | |
whispered | -0.007 |
prick | -0.007 |
chy | -0.007 |
woke | -0.006 |
l | -0.006 |
Token | was |
Token ID | 373 |
Feature activation | -7.85e-02 |
Pos logit contributions | |
unison | +0.063 |
such | +0.062 |
how | +0.059 |
gonna | +0.056 |
our | +0.056 |
Neg logit contributions | |
prick | -0.035 |
aisle | -0.035 |
ho | -0.033 |
l | -0.033 |
whispered | -0.031 |
Token | over |
Token ID | 625 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
unison | +0.219 |
ric | +0.215 |
latable | +0.211 |
Deal | +0.205 |
igsaw | +0.200 |
Neg logit contributions | |
whispered | -0.144 |
said | -0.133 |
woke | -0.130 |
chy | -0.129 |
fell | -0.129 |
Token | to |
Token ID | 284 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
unison | +0.247 |
ric | +0.239 |
latable | +0.237 |
Deal | +0.232 |
igsaw | +0.227 |
Neg logit contributions | |
whispered | -0.142 |
said | -0.129 |
chy | -0.127 |
woke | -0.126 |
fell | -0.125 |
Token | the |
Token ID | 262 |
Feature activation | -1.79e-01 |
Pos logit contributions | |
unison | +0.197 |
ric | +0.188 |
latable | +0.183 |
Deal | +0.180 |
igsaw | +0.174 |
Neg logit contributions | |
whispered | -0.223 |
chy | -0.213 |
fell | -0.208 |
l | -0.206 |
said | -0.206 |
Token | kitten |
Token ID | 42143 |
Feature activation | -1.10e-01 |
Pos logit contributions | |
unison | +0.306 |
ric | +0.289 |
latable | +0.289 |
igsaw | +0.279 |
Deal | +0.272 |
Neg logit contributions | |
whispered | -0.195 |
said | -0.174 |
meant | -0.174 |
fell | -0.173 |
chy | -0.172 |
Token | . |
Token ID | 13 |
Feature activation | -2.22e-01 |
Pos logit contributions | |
unison | +0.194 |
ric | +0.191 |
latable | +0.184 |
igsaw | +0.171 |
ap | +0.169 |
Neg logit contributions | |
whispered | -0.101 |
l | -0.095 |
prick | -0.093 |
meant | -0.093 |
chy | -0.092 |
Token | He |
Token ID | 679 |
Feature activation | -1.59e-01 |
Pos logit contributions | |
unison | +0.322 |
ric | +0.309 |
latable | +0.306 |
igsaw | +0.299 |
Deal | +0.299 |
Neg logit contributions | |
whispered | -0.290 |
chy | -0.282 |
fell | -0.267 |
l | -0.264 |
woke | -0.264 |
Token | was |
Token ID | 373 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
unison | +0.294 |
ric | +0.273 |
latable | +0.272 |
yeah | +0.255 |
gonna | +0.254 |
Neg logit contributions | |
chy | -0.112 |
aisle | -0.111 |
l | -0.101 |
prick | -0.101 |
low | -0.093 |
Token | carrying |
Token ID | 6872 |
Feature activation | -7.77e-02 |
Pos logit contributions | |
unison | +0.219 |
ric | +0.212 |
latable | +0.211 |
Deal | +0.202 |
igsaw | +0.193 |
Neg logit contributions | |
whispered | -0.166 |
woke | -0.156 |
l | -0.154 |
chy | -0.153 |
said | -0.152 |
Token | a |
Token ID | 257 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
unison | +0.090 |
latable | +0.086 |
igsaw | +0.081 |
ric | +0.080 |
Deal | +0.078 |
Neg logit contributions | |
whispered | -0.125 |
woke | -0.120 |
chy | -0.119 |
fell | -0.116 |
said | -0.114 |
Token | warm |
Token ID | 5814 |
Feature activation | -7.70e-02 |
Pos logit contributions | |
unison | +0.228 |
latable | +0.219 |
ric | +0.215 |
igsaw | +0.213 |
Deal | +0.208 |
Neg logit contributions | |
whispered | -0.142 |
said | -0.128 |
woke | -0.125 |
meant | -0.125 |
chy | -0.124 |
Token | blanket |
Token ID | 18447 |
Feature activation | -6.04e-02 |
Pos logit contributions | |
ric | +0.145 |
unison | +0.145 |
Deal | +0.143 |
latable | +0.137 |
hey | +0.133 |
Neg logit contributions | |
whispered | -0.073 |
woke | -0.069 |
said | -0.069 |
fell | -0.066 |
chy | -0.062 |
Token | spaghetti |
Token ID | 48241 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
unison | +0.256 |
latable | +0.242 |
ric | +0.239 |
Deal | +0.227 |
igsaw | +0.225 |
Neg logit contributions | |
whispered | -0.279 |
chy | -0.262 |
l | -0.259 |
woke | -0.257 |
fell | -0.256 |
Token | for |
Token ID | 329 |
Feature activation | -1.54e-01 |
Pos logit contributions | |
unison | +0.218 |
latable | +0.210 |
ric | +0.207 |
igsaw | +0.201 |
Deal | +0.201 |
Neg logit contributions | |
whispered | -0.155 |
woke | -0.139 |
l | -0.138 |
said | -0.136 |
chy | -0.135 |
Token | dinner |
Token ID | 8073 |
Feature activation | -9.83e-02 |
Pos logit contributions | |
unison | +0.200 |
ric | +0.193 |
latable | +0.191 |
Deal | +0.183 |
igsaw | +0.178 |
Neg logit contributions | |
whispered | -0.235 |
said | -0.223 |
woke | -0.217 |
fell | -0.215 |
chy | -0.214 |
Token | , |
Token ID | 11 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
unison | +0.182 |
latable | +0.173 |
igsaw | +0.167 |
Deal | +0.165 |
ric | +0.164 |
Neg logit contributions | |
whispered | -0.089 |
l | -0.081 |
woke | -0.079 |
chy | -0.075 |
fell | -0.075 |
Token | which |
Token ID | 543 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
unison | +0.354 |
ric | +0.350 |
latable | +0.346 |
Deal | +0.335 |
igsaw | +0.334 |
Neg logit contributions | |
whispered | -0.157 |
said | -0.149 |
woke | -0.133 |
fell | -0.132 |
meant | -0.128 |
Token | was |
Token ID | 373 |
Feature activation | -1.54e-01 |
Pos logit contributions | |
unison | +0.232 |
ric | +0.224 |
latable | +0.218 |
Deal | +0.214 |
igsaw | +0.214 |
Neg logit contributions | |
whispered | -0.141 |
l | -0.129 |
woke | -0.117 |
chy | -0.116 |
low | -0.116 |
Token | Lily |
Token ID | 20037 |
Feature activation | -1.33e-01 |
Pos logit contributions | |
unison | +0.227 |
latable | +0.221 |
ric | +0.220 |
Deal | +0.215 |
gonna | +0.202 |
Neg logit contributions | |
whispered | -0.191 |
woke | -0.181 |
l | -0.177 |
fell | -0.172 |
aisle | -0.165 |
Token | 's |
Token ID | 338 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
unison | +0.213 |
ric | +0.205 |
latable | +0.199 |
yeah | +0.193 |
ap | +0.192 |
Neg logit contributions | |
whispered | -0.132 |
l | -0.131 |
aisle | -0.125 |
woke | -0.123 |
prick | -0.120 |
Token | favorite |
Token ID | 4004 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
unison | +0.215 |
latable | +0.207 |
ric | +0.206 |
Deal | +0.204 |
igsaw | +0.194 |
Neg logit contributions | |
whispered | -0.210 |
woke | -0.191 |
said | -0.191 |
l | -0.190 |
fell | -0.189 |
Token | . |
Token ID | 13 |
Feature activation | -2.23e-01 |
Pos logit contributions | |
unison | +0.213 |
latable | +0.204 |
ric | +0.204 |
igsaw | +0.196 |
Deal | +0.194 |
Neg logit contributions | |
whispered | -0.110 |
l | -0.095 |
fell | -0.093 |
said | -0.093 |
woke | -0.090 |
Token | Lily |
Token ID | 20037 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
unison | +0.379 |
ric | +0.369 |
latable | +0.365 |
Deal | +0.356 |
igsaw | +0.347 |
Neg logit contributions | |
whispered | -0.247 |
said | -0.229 |
woke | -0.221 |
fell | -0.219 |
chy | -0.217 |
Token | crown |
Token ID | 12389 |
Feature activation | -8.37e-02 |
Pos logit contributions | |
unison | +0.184 |
ric | +0.174 |
latable | +0.169 |
Deal | +0.165 |
igsaw | +0.164 |
Neg logit contributions | |
whispered | -0.236 |
said | -0.222 |
woke | -0.219 |
chy | -0.218 |
fell | -0.218 |
Token | and |
Token ID | 290 |
Feature activation | -1.47e-01 |
Pos logit contributions | |
unison | +0.137 |
ric | +0.137 |
igsaw | +0.133 |
latable | +0.131 |
yeah | +0.127 |
Neg logit contributions | |
whispered | -0.086 |
woke | -0.079 |
chy | -0.076 |
said | -0.075 |
fell | -0.075 |
Token | make |
Token ID | 787 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
latable | +0.243 |
unison | +0.241 |
ric | +0.241 |
Deal | +0.232 |
igsaw | +0.228 |
Neg logit contributions | |
whispered | -0.183 |
said | -0.177 |
woke | -0.163 |
fell | -0.162 |
meant | -0.155 |
Token | them |
Token ID | 606 |
Feature activation | -8.42e-02 |
Pos logit contributions | |
unison | +0.188 |
latable | +0.184 |
ric | +0.183 |
Deal | +0.181 |
igsaw | +0.174 |
Neg logit contributions | |
whispered | -0.133 |
said | -0.129 |
l | -0.122 |
woke | -0.118 |
fell | -0.117 |
Token | jealous |
Token ID | 19354 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
unison | +0.158 |
latable | +0.156 |
Deal | +0.153 |
ric | +0.151 |
igsaw | +0.145 |
Neg logit contributions | |
whispered | -0.085 |
said | -0.080 |
l | -0.076 |
fell | -0.075 |
chy | -0.073 |
Token | . |
Token ID | 13 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
unison | +0.240 |
ric | +0.229 |
latable | +0.224 |
Deal | +0.221 |
igsaw | +0.218 |
Neg logit contributions | |
whispered | -0.106 |
chy | -0.091 |
woke | -0.090 |
fell | -0.089 |
said | -0.088 |
Token | He |
Token ID | 679 |
Feature activation | -1.42e-01 |
Pos logit contributions | |
unison | +0.374 |
ric | +0.365 |
latable | +0.362 |
Deal | +0.352 |
igsaw | +0.344 |
Neg logit contributions | |
whispered | -0.222 |
said | -0.205 |
woke | -0.197 |
fell | -0.195 |
chy | -0.193 |
Token | put |
Token ID | 1234 |
Feature activation | -6.76e-02 |
Pos logit contributions | |
unison | +0.266 |
latable | +0.243 |
ric | +0.241 |
teammates | +0.236 |
yeah | +0.228 |
Neg logit contributions | |
aisle | -0.099 |
l | -0.094 |
chy | -0.092 |
prick | -0.089 |
low | -0.088 |
Token | on |
Token ID | 319 |
Feature activation | -8.63e-02 |
Pos logit contributions | |
unison | +0.090 |
ric | +0.088 |
latable | +0.084 |
Deal | +0.083 |
igsaw | +0.081 |
Neg logit contributions | |
whispered | -0.094 |
chy | -0.091 |
woke | -0.088 |
fell | -0.085 |
meant | -0.084 |
Token | his |
Token ID | 465 |
Feature activation | -1.09e-01 |
Pos logit contributions | |
unison | +0.134 |
ric | +0.130 |
latable | +0.129 |
Deal | +0.125 |
igsaw | +0.122 |
Neg logit contributions | |
whispered | -0.109 |
said | -0.102 |
woke | -0.099 |
fell | -0.098 |
chy | -0.097 |
Token | best |
Token ID | 1266 |
Feature activation | -6.74e-02 |
Pos logit contributions | |
unison | +0.173 |
ric | +0.168 |
latable | +0.166 |
Deal | +0.162 |
igsaw | +0.158 |
Neg logit contributions | |
whispered | -0.131 |
said | -0.122 |
woke | -0.118 |
fell | -0.117 |
chy | -0.117 |
Token | � |
Token ID | 241 |
Feature activation | -4.61e-01 |
Pos logit contributions | |
chy | +0.114 |
aisle | +0.110 |
prick | +0.104 |
whispered | +0.103 |
l | +0.101 |
Neg logit contributions | |
unison | -0.189 |
ifying | -0.167 |
ric | -0.166 |
ap | -0.165 |
teammates | -0.161 |
Token | That |
Token ID | 2504 |
Feature activation | -2.92e-01 |
Pos logit contributions | |
unison | +0.800 |
latable | +0.782 |
igsaw | +0.761 |
Deal | +0.736 |
ric | +0.717 |
Neg logit contributions | |
said | -0.507 |
whispered | -0.496 |
meant | -0.470 |
fell | -0.446 |
l | -0.445 |
Token | â |
Token ID | 22940 |
Feature activation | -3.06e-01 |
Pos logit contributions | |
unison | +0.523 |
ric | +0.515 |
latable | +0.507 |
Deal | +0.491 |
igsaw | +0.483 |
Neg logit contributions | |
whispered | -0.293 |
said | -0.264 |
woke | -0.258 |
fell | -0.254 |
chy | -0.252 |
Token | € |
Token ID | 26391 |
Feature activation | -3.49e-01 |
Pos logit contributions | |
unison | +0.375 |
ric | +0.373 |
latable | +0.361 |
Deal | +0.350 |
igsaw | +0.343 |
Neg logit contributions | |
whispered | -0.484 |
woke | -0.452 |
said | -0.447 |
fell | -0.446 |
l | -0.440 |
Token | ™ |
Token ID | 8151 |
Feature activation | -3.13e-01 |
Pos logit contributions | |
unison | +0.525 |
latable | +0.511 |
ric | +0.502 |
Deal | +0.491 |
igsaw | +0.474 |
Neg logit contributions | |
whispered | -0.455 |
said | -0.438 |
fell | -0.425 |
woke | -0.419 |
l | -0.408 |
Token | s |
Token ID | 82 |
Feature activation | -2.50e-01 |
Pos logit contributions | |
ric | +0.523 |
unison | +0.469 |
igsaw | +0.457 |
yeah | +0.454 |
latable | +0.454 |
Neg logit contributions | |
whispered | -0.375 |
prick | -0.347 |
woke | -0.320 |
chy | -0.317 |
buzz | -0.308 |
Token | why |
Token ID | 1521 |
Feature activation | -2.53e-01 |
Pos logit contributions | |
unison | +0.413 |
latable | +0.400 |
ric | +0.397 |
Deal | +0.389 |
igsaw | +0.382 |
Neg logit contributions | |
whispered | -0.289 |
said | -0.280 |
woke | -0.262 |
fell | -0.261 |
meant | -0.261 |
Token | it |
Token ID | 340 |
Feature activation | -1.79e-01 |
Pos logit contributions | |
unison | +0.363 |
ric | +0.363 |
latable | +0.347 |
Deal | +0.338 |
igsaw | +0.329 |
Neg logit contributions | |
whispered | -0.338 |
said | -0.308 |
woke | -0.307 |
chy | -0.304 |
fell | -0.301 |
Token | â |
Token ID | 22940 |
Feature activation | -2.42e-01 |
Pos logit contributions | |
ric | +0.317 |
latable | +0.294 |
unison | +0.290 |
Deal | +0.284 |
igsaw | +0.278 |
Neg logit contributions | |
whispered | -0.178 |
aisle | -0.149 |
woke | -0.147 |
said | -0.144 |
l | -0.143 |
Token | € |
Token ID | 26391 |
Feature activation | -2.87e-01 |
Pos logit contributions | |
ric | +0.269 |
unison | +0.265 |
latable | +0.257 |
Deal | +0.248 |
igsaw | +0.241 |
Neg logit contributions | |
whispered | -0.408 |
woke | -0.380 |
said | -0.377 |
fell | -0.377 |
l | -0.375 |
Token | ™ |
Token ID | 8151 |
Feature activation | -2.54e-01 |
Pos logit contributions | |
unison | +0.400 |
latable | +0.387 |
ric | +0.382 |
Deal | +0.372 |
igsaw | +0.359 |
Neg logit contributions | |
whispered | -0.408 |
said | -0.393 |
fell | -0.383 |
woke | -0.379 |
l | -0.369 |
Token | ily |
Token ID | 813 |
Feature activation | -1.78e-01 |
Pos logit contributions | |
unison | +0.467 |
ric | +0.458 |
latable | +0.449 |
Deal | +0.444 |
igsaw | +0.438 |
Neg logit contributions | |
whispered | -0.212 |
said | -0.190 |
fell | -0.184 |
woke | -0.182 |
chy | -0.179 |
Token | was |
Token ID | 373 |
Feature activation | -1.88e-01 |
Pos logit contributions | |
unison | +0.311 |
latable | +0.294 |
ric | +0.290 |
ap | +0.271 |
Deal | +0.271 |
Neg logit contributions | |
l | -0.155 |
prick | -0.150 |
low | -0.145 |
whispered | -0.140 |
chy | -0.138 |
Token | so |
Token ID | 523 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
unison | +0.313 |
ric | +0.305 |
latable | +0.302 |
Deal | +0.293 |
igsaw | +0.286 |
Neg logit contributions | |
whispered | -0.218 |
said | -0.202 |
woke | -0.196 |
fell | -0.193 |
chy | -0.191 |
Token | happy |
Token ID | 3772 |
Feature activation | -1.70e-01 |
Pos logit contributions | |
Deal | +0.273 |
unison | +0.260 |
ap | +0.258 |
latable | +0.245 |
igsaw | +0.245 |
Neg logit contributions | |
aisle | -0.244 |
woke | -0.238 |
whispered | -0.235 |
iv | -0.234 |
l | -0.233 |
Token | that |
Token ID | 326 |
Feature activation | -1.77e-01 |
Pos logit contributions | |
unison | +0.324 |
ric | +0.318 |
latable | +0.316 |
Deal | +0.312 |
igsaw | +0.298 |
Neg logit contributions | |
whispered | -0.163 |
said | -0.157 |
woke | -0.144 |
fell | -0.142 |
meant | -0.136 |
Token | she |
Token ID | 673 |
Feature activation | -1.70e-01 |
Pos logit contributions | |
unison | +0.252 |
Wed | +0.250 |
igsaw | +0.247 |
yeah | +0.244 |
Deal | +0.242 |
Neg logit contributions | |
whispered | -0.203 |
l | -0.196 |
chy | -0.180 |
woke | -0.174 |
prick | -0.173 |
Token | made |
Token ID | 925 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
unison | +0.311 |
latable | +0.299 |
ric | +0.289 |
gonna | +0.286 |
yeah | +0.283 |
Neg logit contributions | |
l | -0.130 |
whispered | -0.124 |
aisle | -0.121 |
chy | -0.115 |
prick | -0.112 |
Token | an |
Token ID | 281 |
Feature activation | -1.60e-01 |
Pos logit contributions | |
unison | +0.242 |
ric | +0.236 |
Deal | +0.232 |
latable | +0.231 |
ifying | +0.224 |
Neg logit contributions | |
whispered | -0.149 |
said | -0.146 |
fell | -0.130 |
meant | -0.129 |
l | -0.129 |
Token | original |
Token ID | 2656 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
unison | +0.260 |
ric | +0.260 |
latable | +0.247 |
Deal | +0.245 |
igsaw | +0.237 |
Neg logit contributions | |
whispered | -0.191 |
said | -0.179 |
chy | -0.169 |
meant | -0.169 |
woke | -0.167 |
Token | cookie |
Token ID | 19751 |
Feature activation | -7.05e-02 |
Pos logit contributions | |
unison | +0.175 |
latable | +0.167 |
ric | +0.167 |
igsaw | +0.160 |
Deal | +0.158 |
Neg logit contributions | |
whispered | -0.146 |
said | -0.134 |
fell | -0.131 |
woke | -0.131 |
l | -0.131 |
Token | all |
Token ID | 477 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
unison | +0.121 |
latable | +0.114 |
ric | +0.111 |
igsaw | +0.111 |
yeah | +0.107 |
Neg logit contributions | |
whispered | -0.073 |
woke | -0.067 |
l | -0.065 |
fell | -0.062 |
said | -0.062 |
Token | little |
Token ID | 1310 |
Feature activation | -3.23e-01 |
Pos logit contributions | |
unison | +0.550 |
ric | +0.534 |
latable | +0.519 |
Deal | +0.510 |
igsaw | +0.489 |
Neg logit contributions | |
whispered | -0.447 |
said | -0.410 |
woke | -0.406 |
fell | -0.401 |
l | -0.397 |
Token | dog |
Token ID | 3290 |
Feature activation | -2.35e-01 |
Pos logit contributions | |
unison | +0.554 |
ric | +0.537 |
latable | +0.529 |
Deal | +0.518 |
igsaw | +0.506 |
Neg logit contributions | |
whispered | -0.350 |
said | -0.320 |
woke | -0.312 |
fell | -0.309 |
chy | -0.309 |
Token | . |
Token ID | 13 |
Feature activation | -3.00e-01 |
Pos logit contributions | |
unison | +0.396 |
ric | +0.370 |
latable | +0.355 |
ap | +0.354 |
Deal | +0.343 |
Neg logit contributions | |
prick | -0.228 |
whispered | -0.214 |
l | -0.214 |
aisle | -0.208 |
chy | -0.208 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.82e-01 |
Pos logit contributions | |
unison | +0.481 |
ric | +0.466 |
latable | +0.460 |
Deal | +0.450 |
igsaw | +0.436 |
Neg logit contributions | |
whispered | -0.360 |
said | -0.333 |
woke | -0.325 |
chy | -0.323 |
fell | -0.323 |
Token | Once |
Token ID | 7454 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
ric | +0.377 |
unison | +0.371 |
Deal | +0.345 |
igsaw | +0.339 |
latable | +0.338 |
Neg logit contributions | |
whispered | -0.394 |
chy | -0.380 |
prick | -0.367 |
meant | -0.338 |
said | -0.336 |
Token | upon |
Token ID | 2402 |
Feature activation | -3.12e-01 |
Pos logit contributions | |
unison | +0.298 |
ric | +0.291 |
latable | +0.283 |
Deal | +0.274 |
igsaw | +0.266 |
Neg logit contributions | |
whispered | -0.415 |
said | -0.391 |
woke | -0.383 |
fell | -0.379 |
chy | -0.378 |
Token | a |
Token ID | 257 |
Feature activation | -2.19e-01 |
Pos logit contributions | |
ric | +0.600 |
unison | +0.580 |
latable | +0.577 |
igsaw | +0.577 |
Deal | +0.573 |
Neg logit contributions | |
whispered | -0.311 |
said | -0.307 |
woke | -0.299 |
fell | -0.298 |
chy | -0.269 |
Token | time |
Token ID | 640 |
Feature activation | -1.99e-01 |
Pos logit contributions | |
ric | +0.441 |
unison | +0.439 |
latable | +0.423 |
Deal | +0.413 |
igsaw | +0.411 |
Neg logit contributions | |
whispered | -0.170 |
said | -0.159 |
woke | -0.156 |
fell | -0.153 |
chy | -0.151 |
Token | , |
Token ID | 11 |
Feature activation | -2.03e-01 |
Pos logit contributions | |
unison | +0.431 |
ric | +0.414 |
latable | +0.414 |
Deal | +0.400 |
igsaw | +0.389 |
Neg logit contributions | |
whispered | -0.137 |
said | -0.131 |
fell | -0.125 |
woke | -0.123 |
l | -0.115 |
Token | in |
Token ID | 287 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
unison | +0.331 |
ric | +0.325 |
latable | +0.319 |
Deal | +0.310 |
igsaw | +0.303 |
Neg logit contributions | |
whispered | -0.239 |
said | -0.220 |
woke | -0.214 |
fell | -0.210 |
chy | -0.209 |
Token | a |
Token ID | 257 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
ric | +0.413 |
Deal | +0.370 |
pieces | +0.366 |
unison | +0.357 |
igsaw | +0.351 |
Neg logit contributions | |
whispered | -0.148 |
prick | -0.148 |
said | -0.146 |
woke | -0.139 |
spark | -0.137 |
Token | playful |
Token ID | 34264 |
Feature activation | -3.38e-01 |
Pos logit contributions | |
whispered | +0.172 |
said | +0.152 |
woke | +0.152 |
chy | +0.151 |
meant | +0.150 |
Neg logit contributions | |
unison | -0.302 |
latable | -0.294 |
ric | -0.293 |
ap | -0.284 |
igsaw | -0.283 |
Token | adventure |
Token ID | 8855 |
Feature activation | -2.15e-01 |
Pos logit contributions | |
unison | +0.593 |
ric | +0.587 |
latable | +0.582 |
Deal | +0.575 |
igsaw | +0.555 |
Neg logit contributions | |
whispered | -0.353 |
said | -0.339 |
woke | -0.323 |
fell | -0.312 |
chy | -0.306 |
Token | . |
Token ID | 13 |
Feature activation | -3.00e-01 |
Pos logit contributions | |
unison | +0.372 |
latable | +0.355 |
ric | +0.354 |
igsaw | +0.342 |
Deal | +0.326 |
Neg logit contributions | |
whispered | -0.220 |
woke | -0.192 |
chy | -0.191 |
said | -0.188 |
fell | -0.185 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.86e-01 |
Pos logit contributions | |
unison | +0.498 |
ric | +0.485 |
latable | +0.480 |
Deal | +0.466 |
igsaw | +0.455 |
Neg logit contributions | |
whispered | -0.345 |
said | -0.320 |
woke | -0.309 |
fell | -0.307 |
chy | -0.305 |
Token | Once |
Token ID | 7454 |
Feature activation | -2.59e-01 |
Pos logit contributions | |
ric | +0.383 |
unison | +0.364 |
Deal | +0.342 |
igsaw | +0.342 |
latable | +0.336 |
Neg logit contributions | |
whispered | -0.401 |
chy | -0.384 |
prick | -0.373 |
meant | -0.349 |
said | -0.342 |
Token | upon |
Token ID | 2402 |
Feature activation | -3.17e-01 |
Pos logit contributions | |
ric | +0.277 |
unison | +0.274 |
latable | +0.257 |
Deal | +0.254 |
igsaw | +0.250 |
Neg logit contributions | |
whispered | -0.441 |
said | -0.411 |
woke | -0.407 |
chy | -0.402 |
meant | -0.401 |
Token | a |
Token ID | 257 |
Feature activation | -2.25e-01 |
Pos logit contributions | |
ric | +0.617 |
latable | +0.594 |
unison | +0.593 |
igsaw | +0.592 |
Deal | +0.592 |
Neg logit contributions | |
whispered | -0.312 |
said | -0.312 |
woke | -0.296 |
fell | -0.296 |
low | -0.264 |
Token | time |
Token ID | 640 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
unison | +0.441 |
ric | +0.440 |
latable | +0.426 |
Deal | +0.416 |
igsaw | +0.412 |
Neg logit contributions | |
whispered | -0.185 |
said | -0.172 |
woke | -0.168 |
fell | -0.165 |
chy | -0.163 |
Token | , |
Token ID | 11 |
Feature activation | -2.08e-01 |
Pos logit contributions | |
unison | +0.427 |
latable | +0.412 |
ric | +0.411 |
Deal | +0.398 |
igsaw | +0.389 |
Neg logit contributions | |
whispered | -0.148 |
said | -0.141 |
fell | -0.134 |
woke | -0.132 |
l | -0.125 |
Token | there |
Token ID | 612 |
Feature activation | -1.87e-01 |
Pos logit contributions | |
unison | +0.326 |
ric | +0.323 |
latable | +0.314 |
Deal | +0.305 |
igsaw | +0.297 |
Neg logit contributions | |
whispered | -0.254 |
said | -0.232 |
woke | -0.228 |
chy | -0.222 |
fell | -0.222 |
Token | was |
Token ID | 373 |
Feature activation | -1.83e-01 |
Pos logit contributions | |
ric | +0.350 |
unison | +0.342 |
latable | +0.334 |
yeah | +0.312 |
ap | +0.307 |
Neg logit contributions | |
whispered | -0.169 |
chy | -0.144 |
l | -0.135 |
prick | -0.132 |
aisle | -0.126 |
Token | ever |
Token ID | 1683 |
Feature activation | -2.93e-01 |
Pos logit contributions | |
whispered | +0.190 |
fell | +0.165 |
said | +0.164 |
woke | +0.162 |
chy | +0.162 |
Neg logit contributions | |
unison | -0.303 |
latable | -0.299 |
ric | -0.299 |
igsaw | -0.290 |
yeah | -0.277 |
Token | after |
Token ID | 706 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
unison | +0.514 |
ric | +0.502 |
latable | +0.494 |
Deal | +0.483 |
igsaw | +0.471 |
Neg logit contributions | |
whispered | -0.305 |
said | -0.280 |
woke | -0.270 |
chy | -0.267 |
fell | -0.266 |
Token | . |
Token ID | 13 |
Feature activation | -3.06e-01 |
Pos logit contributions | |
unison | +0.421 |
ric | +0.413 |
latable | +0.400 |
Deal | +0.391 |
igsaw | +0.388 |
Neg logit contributions | |
whispered | -0.306 |
woke | -0.275 |
said | -0.273 |
chy | -0.273 |
l | -0.266 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
unison | +0.503 |
ric | +0.488 |
latable | +0.480 |
Deal | +0.473 |
igsaw | +0.458 |
Neg logit contributions | |
whispered | -0.356 |
said | -0.326 |
chy | -0.320 |
woke | -0.319 |
fell | -0.317 |
Token | Once |
Token ID | 7454 |
Feature activation | -2.58e-01 |
Pos logit contributions | |
ric | +0.379 |
unison | +0.375 |
igsaw | +0.350 |
Deal | +0.350 |
latable | +0.347 |
Neg logit contributions | |
whispered | -0.395 |
chy | -0.380 |
prick | -0.367 |
meant | -0.340 |
knew | -0.336 |
Token | upon |
Token ID | 2402 |
Feature activation | -3.14e-01 |
Pos logit contributions | |
ric | +0.280 |
unison | +0.278 |
latable | +0.262 |
Deal | +0.258 |
igsaw | +0.253 |
Neg logit contributions | |
whispered | -0.436 |
said | -0.406 |
woke | -0.403 |
chy | -0.398 |
fell | -0.396 |
Token | a |
Token ID | 257 |
Feature activation | -2.24e-01 |
Pos logit contributions | |
ric | +0.622 |
unison | +0.593 |
Deal | +0.592 |
igsaw | +0.592 |
latable | +0.592 |
Neg logit contributions | |
said | -0.303 |
whispered | -0.302 |
woke | -0.288 |
fell | -0.286 |
l | -0.256 |
Token | time |
Token ID | 640 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
ric | +0.447 |
unison | +0.445 |
latable | +0.429 |
Deal | +0.419 |
igsaw | +0.416 |
Neg logit contributions | |
whispered | -0.179 |
said | -0.166 |
woke | -0.164 |
fell | -0.160 |
chy | -0.158 |
Token | , |
Token ID | 11 |
Feature activation | -2.08e-01 |
Pos logit contributions | |
unison | +0.428 |
ric | +0.414 |
latable | +0.413 |
Deal | +0.399 |
igsaw | +0.389 |
Neg logit contributions | |
whispered | -0.146 |
said | -0.139 |
fell | -0.133 |
woke | -0.130 |
l | -0.123 |
Token | there |
Token ID | 612 |
Feature activation | -1.86e-01 |
Pos logit contributions | |
unison | +0.332 |
ric | +0.327 |
latable | +0.320 |
Deal | +0.310 |
igsaw | +0.302 |
Neg logit contributions | |
whispered | -0.249 |
said | -0.228 |
woke | -0.223 |
chy | -0.218 |
fell | -0.218 |
Token | was |
Token ID | 373 |
Feature activation | -1.82e-01 |
Pos logit contributions | |
ric | +0.342 |
unison | +0.335 |
latable | +0.330 |
yeah | +0.306 |
ap | +0.301 |
Neg logit contributions | |
whispered | -0.173 |
chy | -0.148 |
l | -0.140 |
prick | -0.135 |
said | -0.124 |
Token | flew |
Token ID | 13112 |
Feature activation | -2.74e-01 |
Pos logit contributions | |
whispered | +0.142 |
l | +0.135 |
aisle | +0.130 |
meant | +0.126 |
chy | +0.125 |
Neg logit contributions | |
unison | -0.254 |
ric | -0.251 |
latable | -0.230 |
ap | -0.219 |
ifying | -0.212 |
Token | away |
Token ID | 1497 |
Feature activation | -2.47e-01 |
Pos logit contributions | |
unison | +0.430 |
ric | +0.418 |
latable | +0.408 |
Deal | +0.390 |
igsaw | +0.382 |
Neg logit contributions | |
whispered | -0.311 |
woke | -0.284 |
meant | -0.282 |
chy | -0.280 |
said | -0.279 |
Token | . |
Token ID | 13 |
Feature activation | -3.07e-01 |
Pos logit contributions | |
unison | +0.377 |
ric | +0.373 |
latable | +0.343 |
igsaw | +0.341 |
ap | +0.335 |
Neg logit contributions | |
whispered | -0.262 |
woke | -0.244 |
aisle | -0.242 |
meant | -0.242 |
mail | -0.229 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.84e-01 |
Pos logit contributions | |
unison | +0.459 |
ric | +0.436 |
Deal | +0.432 |
latable | +0.424 |
igsaw | +0.412 |
Neg logit contributions | |
whispered | -0.398 |
chy | -0.373 |
l | -0.363 |
woke | -0.361 |
fell | -0.360 |
Token | Once |
Token ID | 7454 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
ric | +0.365 |
unison | +0.358 |
igsaw | +0.337 |
Deal | +0.334 |
latable | +0.325 |
Neg logit contributions | |
whispered | -0.405 |
chy | -0.395 |
prick | -0.382 |
meant | -0.349 |
knew | -0.348 |
Token | upon |
Token ID | 2402 |
Feature activation | -3.14e-01 |
Pos logit contributions | |
ric | +0.274 |
unison | +0.271 |
latable | +0.254 |
Deal | +0.251 |
igsaw | +0.247 |
Neg logit contributions | |
whispered | -0.434 |
said | -0.403 |
woke | -0.402 |
chy | -0.397 |
fell | -0.394 |
Token | a |
Token ID | 257 |
Feature activation | -2.24e-01 |
Pos logit contributions | |
ric | +0.597 |
unison | +0.572 |
latable | +0.571 |
Deal | +0.568 |
igsaw | +0.562 |
Neg logit contributions | |
whispered | -0.321 |
said | -0.316 |
woke | -0.303 |
fell | -0.302 |
meant | -0.275 |
Token | time |
Token ID | 640 |
Feature activation | -2.01e-01 |
Pos logit contributions | |
ric | +0.447 |
unison | +0.444 |
latable | +0.430 |
Deal | +0.419 |
igsaw | +0.416 |
Neg logit contributions | |
whispered | -0.177 |
said | -0.165 |
woke | -0.162 |
fell | -0.159 |
chy | -0.157 |
Token | , |
Token ID | 11 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
unison | +0.435 |
latable | +0.419 |
ric | +0.418 |
Deal | +0.404 |
igsaw | +0.393 |
Neg logit contributions | |
whispered | -0.138 |
said | -0.133 |
fell | -0.126 |
woke | -0.123 |
l | -0.117 |
Token | there |
Token ID | 612 |
Feature activation | -1.86e-01 |
Pos logit contributions | |
unison | +0.324 |
ric | +0.320 |
latable | +0.311 |
Deal | +0.301 |
igsaw | +0.294 |
Neg logit contributions | |
whispered | -0.251 |
said | -0.229 |
woke | -0.227 |
chy | -0.220 |
l | -0.219 |
Token | was |
Token ID | 373 |
Feature activation | -1.83e-01 |
Pos logit contributions | |
ric | +0.346 |
unison | +0.337 |
latable | +0.329 |
yeah | +0.306 |
ap | +0.302 |
Neg logit contributions | |
whispered | -0.172 |
chy | -0.147 |
l | -0.139 |
prick | -0.136 |
aisle | -0.124 |
Token | best |
Token ID | 1266 |
Feature activation | -3.09e-01 |
Pos logit contributions | |
aisle | +0.097 |
l | +0.097 |
whispered | +0.095 |
low | +0.085 |
prick | +0.083 |
Neg logit contributions | |
latable | -0.118 |
ric | -0.114 |
unison | -0.112 |
Wed | -0.104 |
disl | -0.102 |
Token | friends |
Token ID | 2460 |
Feature activation | -2.73e-01 |
Pos logit contributions | |
unison | +0.515 |
latable | +0.514 |
ric | +0.512 |
Deal | +0.492 |
igsaw | +0.486 |
Neg logit contributions | |
whispered | -0.356 |
said | -0.350 |
woke | -0.322 |
fell | -0.321 |
l | -0.311 |
Token | . |
Token ID | 13 |
Feature activation | -3.08e-01 |
Pos logit contributions | |
unison | +0.479 |
ric | +0.468 |
latable | +0.449 |
Deal | +0.436 |
yeah | +0.431 |
Neg logit contributions | |
whispered | -0.262 |
chy | -0.238 |
l | -0.234 |
prick | -0.230 |
woke | -0.228 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.88e-01 |
Pos logit contributions | |
unison | +0.507 |
ric | +0.494 |
latable | +0.490 |
Deal | +0.475 |
igsaw | +0.464 |
Neg logit contributions | |
whispered | -0.356 |
said | -0.332 |
woke | -0.320 |
fell | -0.316 |
chy | -0.313 |
Token | Once |
Token ID | 7454 |
Feature activation | -2.62e-01 |
Pos logit contributions | |
ric | +0.393 |
unison | +0.385 |
Deal | +0.359 |
latable | +0.359 |
igsaw | +0.355 |
Neg logit contributions | |
whispered | -0.397 |
chy | -0.372 |
prick | -0.361 |
meant | -0.341 |
said | -0.338 |
Token | upon |
Token ID | 2402 |
Feature activation | -3.17e-01 |
Pos logit contributions | |
ric | +0.281 |
unison | +0.279 |
latable | +0.262 |
Deal | +0.259 |
igsaw | +0.255 |
Neg logit contributions | |
whispered | -0.445 |
said | -0.414 |
woke | -0.409 |
chy | -0.404 |
meant | -0.403 |
Token | a |
Token ID | 257 |
Feature activation | -2.26e-01 |
Pos logit contributions | |
ric | +0.612 |
unison | +0.589 |
latable | +0.587 |
igsaw | +0.585 |
Deal | +0.581 |
Neg logit contributions | |
whispered | -0.318 |
said | -0.316 |
fell | -0.303 |
woke | -0.302 |
low | -0.273 |
Token | time |
Token ID | 640 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
unison | +0.440 |
ric | +0.438 |
latable | +0.425 |
Deal | +0.415 |
igsaw | +0.411 |
Neg logit contributions | |
whispered | -0.190 |
said | -0.176 |
woke | -0.171 |
fell | -0.168 |
chy | -0.165 |
Token | , |
Token ID | 11 |
Feature activation | -2.08e-01 |
Pos logit contributions | |
unison | +0.426 |
ric | +0.411 |
latable | +0.411 |
Deal | +0.397 |
igsaw | +0.387 |
Neg logit contributions | |
whispered | -0.150 |
said | -0.142 |
fell | -0.135 |
woke | -0.133 |
l | -0.126 |
Token | there |
Token ID | 612 |
Feature activation | -1.87e-01 |
Pos logit contributions | |
unison | +0.333 |
ric | +0.329 |
latable | +0.321 |
Deal | +0.312 |
igsaw | +0.304 |
Neg logit contributions | |
whispered | -0.247 |
said | -0.226 |
woke | -0.221 |
fell | -0.216 |
chy | -0.215 |
Token | was |
Token ID | 373 |
Feature activation | -1.83e-01 |
Pos logit contributions | |
ric | +0.356 |
unison | +0.347 |
latable | +0.339 |
yeah | +0.319 |
ap | +0.313 |
Neg logit contributions | |
whispered | -0.163 |
chy | -0.137 |
l | -0.129 |
prick | -0.125 |
aisle | -0.118 |