Token | y |
Token ID | 331 |
Feature activation | -9.08e-02 |
Pos logit contributions | |
le | +0.174 |
ling | +0.165 |
les | +0.163 |
behind | +0.160 |
an | +0.154 |
Neg logit contributions | |
Deal | -0.197 |
Not | -0.181 |
They | -0.170 |
Okay | -0.170 |
Something | -0.160 |
Token | ummy |
Token ID | 13513 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
le | +0.200 |
ling | +0.190 |
behind | +0.186 |
inside | +0.185 |
pe | +0.180 |
Neg logit contributions | |
Deal | -0.160 |
Not | -0.128 |
Okay | -0.125 |
Jake | -0.123 |
Thanks | -0.120 |
Token | ," |
Token ID | 553 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
behind | +0.128 |
le | +0.121 |
ling | +0.121 |
an | +0.117 |
b | +0.114 |
Neg logit contributions | |
Deal | -0.238 |
Not | -0.215 |
They | -0.212 |
Okay | -0.202 |
Jake | -0.195 |
Token | Lily |
Token ID | 20037 |
Feature activation | +2.47e-02 |
Pos logit contributions | |
Deal | +0.317 |
Not | +0.244 |
Okay | +0.233 |
Thanks | +0.229 |
Jake | +0.227 |
Neg logit contributions | |
pe | -0.338 |
behind | -0.332 |
le | -0.327 |
inside | -0.325 |
away | -0.316 |
Token | said |
Token ID | 531 |
Feature activation | +5.18e-02 |
Pos logit contributions | |
Deal | +0.064 |
Not | +0.055 |
Thanks | +0.055 |
Okay | +0.055 |
Jake | +0.054 |
Neg logit contributions | |
le | -0.033 |
ling | -0.030 |
behind | -0.030 |
inside | -0.030 |
pe | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | +2.32e-01 |
Pos logit contributions | |
Deal | +0.140 |
Not | +0.125 |
Okay | +0.125 |
They | +0.121 |
Acc | +0.120 |
Neg logit contributions | |
le | -0.059 |
ling | -0.050 |
les | -0.048 |
pe | -0.048 |
behind | -0.048 |
Token |
|
Token ID | 198 |
Feature activation | +1.93e-01 |
Pos logit contributions | |
Deal | +0.508 |
Thanks | +0.424 |
liv | +0.423 |
omo | +0.407 |
Jake | +0.404 |
Neg logit contributions | |
inside | -0.321 |
. | -0.317 |
on | -0.316 |
behind | -0.310 |
in | -0.293 |
Token |
|
Token ID | 198 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
Deal | +0.424 |
Sorry | +0.346 |
Not | +0.342 |
Okay | +0.337 |
Thanks | +0.336 |
Neg logit contributions | |
inside | -0.298 |
le | -0.292 |
behind | -0.289 |
ling | -0.275 |
. | -0.270 |
Token | " |
Token ID | 1 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
Deal | +0.395 |
Not | +0.332 |
Okay | +0.317 |
Acc | +0.305 |
Sorry | +0.304 |
Neg logit contributions | |
le | -0.315 |
ling | -0.302 |
away | -0.297 |
inside | -0.297 |
behind | -0.293 |
Token | OK |
Token ID | 11380 |
Feature activation | +3.30e-03 |
Pos logit contributions | |
Deal | +0.173 |
Tyler | +0.140 |
Jake | +0.133 |
Kelly | +0.129 |
ossible | +0.128 |
Neg logit contributions | |
pe | -0.222 |
le | -0.216 |
ring | -0.213 |
behind | -0.209 |
ling | -0.207 |
Token | , |
Token ID | 11 |
Feature activation | +6.37e-02 |
Pos logit contributions | |
Deal | +0.009 |
Not | +0.008 |
Okay | +0.008 |
Jake | +0.008 |
They | +0.008 |
Neg logit contributions | |
le | -0.004 |
ling | -0.003 |
behind | -0.003 |
pe | -0.003 |
les | -0.003 |
Token | p |
Token ID | 279 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
Deal | +0.122 |
Not | +0.103 |
Okay | +0.101 |
Jake | +0.101 |
Thanks | +0.099 |
Neg logit contributions | |
le | -0.087 |
ling | -0.078 |
behind | -0.077 |
inside | -0.073 |
les | -0.071 |
Token | uddle |
Token ID | 24500 |
Feature activation | -1.58e-01 |
Pos logit contributions | |
le | +0.197 |
behind | +0.185 |
inside | +0.180 |
ling | +0.177 |
pe | +0.171 |
Neg logit contributions | |
Deal | -0.213 |
Not | -0.178 |
Okay | -0.174 |
Jake | -0.170 |
Thanks | -0.166 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
le | +0.188 |
ling | +0.158 |
pe | +0.154 |
egg | +0.147 |
behind | +0.147 |
Neg logit contributions | |
Deal | -0.374 |
They | -0.356 |
Not | -0.346 |
Thanks | -0.345 |
She | -0.341 |
Token | she |
Token ID | 673 |
Feature activation | +6.54e-02 |
Pos logit contributions | |
Deal | +0.333 |
Okay | +0.246 |
Not | +0.243 |
Thanks | +0.240 |
Jake | +0.236 |
Neg logit contributions | |
behind | -0.400 |
le | -0.398 |
inside | -0.390 |
pe | -0.386 |
ling | -0.377 |
Token | says |
Token ID | 1139 |
Feature activation | +4.07e-02 |
Pos logit contributions | |
Deal | +0.149 |
Not | +0.128 |
Okay | +0.124 |
Jake | +0.122 |
Thanks | +0.119 |
Neg logit contributions | |
le | -0.104 |
ling | -0.099 |
behind | -0.097 |
inside | -0.096 |
away | -0.091 |
Token | . |
Token ID | 13 |
Feature activation | +2.31e-01 |
Pos logit contributions | |
Deal | +0.106 |
Not | +0.096 |
Okay | +0.095 |
Thanks | +0.091 |
Jake | +0.090 |
Neg logit contributions | |
le | -0.049 |
ling | -0.044 |
les | -0.044 |
an | -0.041 |
egg | -0.040 |
Token |
|
Token ID | 198 |
Feature activation | +1.97e-01 |
Pos logit contributions | |
Deal | +0.456 |
Thanks | +0.375 |
Enough | +0.357 |
Jake | +0.355 |
Great | +0.354 |
Neg logit contributions | |
behind | -0.378 |
inside | -0.372 |
on | -0.360 |
pe | -0.358 |
. | -0.345 |
Token |
|
Token ID | 198 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
Deal | +0.409 |
Thanks | +0.331 |
Okay | +0.328 |
Sorry | +0.328 |
Not | +0.325 |
Neg logit contributions | |
inside | -0.332 |
behind | -0.329 |
le | -0.327 |
pe | -0.314 |
ling | -0.302 |
Token | Ben |
Token ID | 11696 |
Feature activation | -1.68e-02 |
Pos logit contributions | |
Deal | +0.397 |
Not | +0.332 |
Okay | +0.321 |
Acc | +0.318 |
Winning | +0.318 |
Neg logit contributions | |
behind | -0.299 |
le | -0.298 |
inside | -0.296 |
pe | -0.295 |
away | -0.288 |
Token | smiles |
Token ID | 21845 |
Feature activation | -2.65e-02 |
Pos logit contributions | |
le | +0.025 |
ling | +0.023 |
behind | +0.023 |
inside | +0.022 |
pe | +0.022 |
Neg logit contributions | |
Deal | -0.041 |
Not | -0.035 |
Okay | -0.035 |
Jake | -0.034 |
Thanks | -0.034 |
Token | too |
Token ID | 1165 |
Feature activation | +2.32e-02 |
Pos logit contributions | |
le | +0.024 |
ling | +0.021 |
behind | +0.021 |
inside | +0.019 |
pe | +0.019 |
Neg logit contributions | |
Deal | -0.079 |
Not | -0.071 |
Okay | -0.069 |
Thanks | -0.068 |
Jake | -0.067 |
Token | said |
Token ID | 531 |
Feature activation | +5.40e-02 |
Pos logit contributions | |
Deal | +0.186 |
Not | +0.160 |
Okay | +0.153 |
Jake | +0.150 |
Thanks | +0.148 |
Neg logit contributions | |
le | -0.149 |
ling | -0.145 |
behind | -0.141 |
inside | -0.136 |
away | -0.132 |
Token | she |
Token ID | 673 |
Feature activation | +4.05e-02 |
Pos logit contributions | |
Deal | +0.128 |
Not | +0.112 |
Okay | +0.108 |
Jake | +0.106 |
Thanks | +0.104 |
Neg logit contributions | |
le | -0.083 |
ling | -0.079 |
behind | -0.076 |
inside | -0.073 |
pe | -0.071 |
Token | learned |
Token ID | 4499 |
Feature activation | +2.39e-02 |
Pos logit contributions | |
Deal | +0.081 |
Not | +0.069 |
Okay | +0.067 |
Jake | +0.065 |
Thanks | +0.064 |
Neg logit contributions | |
le | -0.076 |
ling | -0.073 |
behind | -0.072 |
inside | -0.069 |
away | -0.068 |
Token | a |
Token ID | 257 |
Feature activation | +3.12e-02 |
Pos logit contributions | |
Deal | +0.051 |
Not | +0.044 |
Okay | +0.042 |
Jake | +0.041 |
Thanks | +0.040 |
Neg logit contributions | |
le | -0.042 |
ling | -0.041 |
behind | -0.039 |
inside | -0.038 |
pe | -0.036 |
Token | lesson |
Token ID | 11483 |
Feature activation | -6.01e-02 |
Pos logit contributions | |
Deal | +0.074 |
Not | +0.064 |
Okay | +0.062 |
Jake | +0.061 |
Thanks | +0.060 |
Neg logit contributions | |
le | -0.049 |
ling | -0.046 |
behind | -0.045 |
inside | -0.044 |
pe | -0.042 |
Token | . |
Token ID | 13 |
Feature activation | +2.31e-01 |
Pos logit contributions | |
ling | +0.086 |
le | +0.086 |
pe | +0.075 |
les | +0.075 |
away | +0.073 |
Neg logit contributions | |
Deal | -0.136 |
Not | -0.130 |
They | -0.120 |
Kelly | -0.119 |
Tyler | -0.118 |
Token | She |
Token ID | 1375 |
Feature activation | +8.55e-02 |
Pos logit contributions | |
Deal | +0.323 |
Thanks | +0.256 |
Jake | +0.242 |
Enough | +0.237 |
Tyler | +0.233 |
Neg logit contributions | |
behind | -0.507 |
inside | -0.506 |
pe | -0.483 |
. | -0.456 |
below | -0.452 |
Token | said |
Token ID | 531 |
Feature activation | +6.20e-02 |
Pos logit contributions | |
Deal | +0.187 |
Not | +0.162 |
Okay | +0.156 |
Jake | +0.155 |
Thanks | +0.153 |
Neg logit contributions | |
le | -0.143 |
ling | -0.137 |
behind | -0.134 |
inside | -0.130 |
away | -0.126 |
Token | she |
Token ID | 673 |
Feature activation | +4.64e-02 |
Pos logit contributions | |
Deal | +0.116 |
Not | +0.100 |
Okay | +0.094 |
Jake | +0.091 |
Thanks | +0.090 |
Neg logit contributions | |
le | -0.127 |
ling | -0.120 |
behind | -0.117 |
inside | -0.114 |
away | -0.112 |
Token | would |
Token ID | 561 |
Feature activation | +3.51e-02 |
Pos logit contributions | |
Deal | +0.099 |
Not | +0.083 |
Okay | +0.081 |
Jake | +0.080 |
Thanks | +0.079 |
Neg logit contributions | |
le | -0.083 |
ling | -0.079 |
behind | -0.078 |
inside | -0.077 |
pe | -0.074 |
Token | not |
Token ID | 407 |
Feature activation | +3.03e-02 |
Pos logit contributions | |
Deal | +0.067 |
Not | +0.056 |
Okay | +0.054 |
Thanks | +0.052 |
Jake | +0.051 |
Neg logit contributions | |
le | -0.071 |
ling | -0.069 |
behind | -0.067 |
inside | -0.065 |
les | -0.064 |
Token | to |
Token ID | 284 |
Feature activation | +3.07e-02 |
Pos logit contributions | |
Deal | +0.112 |
Not | +0.102 |
Okay | +0.095 |
They | +0.091 |
Thanks | +0.090 |
Neg logit contributions | |
le | -0.101 |
ling | -0.096 |
behind | -0.095 |
inside | -0.089 |
les | -0.089 |
Token | count |
Token ID | 954 |
Feature activation | -8.51e-02 |
Pos logit contributions | |
Deal | +0.061 |
Not | +0.052 |
Okay | +0.049 |
Thanks | +0.048 |
Jake | +0.047 |
Neg logit contributions | |
le | -0.058 |
ling | -0.058 |
behind | -0.055 |
les | -0.054 |
inside | -0.053 |
Token | ?" |
Token ID | 1701 |
Feature activation | +1.76e-01 |
Pos logit contributions | |
le | +0.123 |
ling | +0.122 |
an | +0.113 |
les | +0.106 |
away | +0.098 |
Neg logit contributions | |
Deal | -0.189 |
Not | -0.179 |
They | -0.174 |
Jake | -0.167 |
She | -0.166 |
Token | Ben |
Token ID | 3932 |
Feature activation | -4.50e-03 |
Pos logit contributions | |
Deal | +0.310 |
Not | +0.234 |
Thanks | +0.231 |
Okay | +0.229 |
Jake | +0.224 |
Neg logit contributions | |
le | -0.371 |
behind | -0.361 |
inside | -0.356 |
pe | -0.353 |
ling | -0.348 |
Token | said |
Token ID | 531 |
Feature activation | +5.21e-02 |
Pos logit contributions | |
le | +0.007 |
ling | +0.007 |
behind | +0.007 |
inside | +0.007 |
away | +0.006 |
Neg logit contributions | |
Deal | -0.010 |
Not | -0.009 |
Okay | -0.009 |
Jake | -0.008 |
Thanks | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | +2.30e-01 |
Pos logit contributions | |
Deal | +0.136 |
Not | +0.122 |
Okay | +0.119 |
They | +0.115 |
Acc | +0.115 |
Neg logit contributions | |
le | -0.062 |
ling | -0.055 |
les | -0.055 |
egg | -0.052 |
pe | -0.051 |
Token |
|
Token ID | 198 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
Deal | +0.453 |
Thanks | +0.372 |
Enough | +0.371 |
Johnny | +0.367 |
liv | +0.364 |
Neg logit contributions | |
inside | -0.366 |
behind | -0.366 |
on | -0.361 |
pe | -0.356 |
. | -0.352 |
Token |
|
Token ID | 198 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
Deal | +0.421 |
Not | +0.344 |
Thanks | +0.341 |
Okay | +0.339 |
Sorry | +0.334 |
Neg logit contributions | |
le | -0.318 |
inside | -0.301 |
ling | -0.292 |
behind | -0.292 |
pe | -0.288 |
Token | " |
Token ID | 1 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
Deal | +0.435 |
Not | +0.383 |
Acc | +0.377 |
idents | +0.368 |
Okay | +0.361 |
Neg logit contributions | |
le | -0.283 |
away | -0.267 |
ling | -0.263 |
b | -0.249 |
pe | -0.241 |
Token | Yes |
Token ID | 5297 |
Feature activation | +6.04e-03 |
Pos logit contributions | |
Deal | +0.173 |
Tyler | +0.136 |
Kelly | +0.135 |
Johnny | +0.130 |
Jake | +0.126 |
Neg logit contributions | |
pe | -0.227 |
le | -0.223 |
ring | -0.215 |
inside | -0.212 |
b | -0.210 |
Token | , |
Token ID | 11 |
Feature activation | +5.32e-02 |
Pos logit contributions | |
Deal | +0.018 |
Not | +0.016 |
Okay | +0.016 |
Jake | +0.016 |
Thanks | +0.015 |
Neg logit contributions | |
le | -0.006 |
ling | -0.005 |
behind | -0.005 |
inside | -0.005 |
pe | -0.004 |
Token | and |
Token ID | 290 |
Feature activation | +4.96e-02 |
Pos logit contributions | |
le | +0.081 |
ling | +0.080 |
behind | +0.074 |
pe | +0.072 |
b | +0.066 |
Neg logit contributions | |
Deal | -0.174 |
Not | -0.157 |
Okay | -0.153 |
They | -0.151 |
Thanks | -0.150 |
Token | hugs |
Token ID | 40657 |
Feature activation | -5.09e-02 |
Pos logit contributions | |
Deal | +0.105 |
Not | +0.090 |
Okay | +0.088 |
Thanks | +0.085 |
Something | +0.084 |
Neg logit contributions | |
le | -0.087 |
ling | -0.084 |
behind | -0.079 |
inside | -0.076 |
les | -0.074 |
Token | ," |
Token ID | 553 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
le | +0.063 |
ling | +0.062 |
behind | +0.058 |
pe | +0.057 |
inside | +0.053 |
Neg logit contributions | |
Deal | -0.127 |
Not | -0.115 |
Okay | -0.112 |
Jake | -0.110 |
Thanks | -0.110 |
Token | Tom |
Token ID | 4186 |
Feature activation | -5.53e-03 |
Pos logit contributions | |
Deal | +0.309 |
Not | +0.241 |
Okay | +0.230 |
Thanks | +0.226 |
Jake | +0.226 |
Neg logit contributions | |
le | -0.342 |
behind | -0.338 |
pe | -0.333 |
inside | -0.333 |
ling | -0.330 |
Token | says |
Token ID | 1139 |
Feature activation | +4.32e-02 |
Pos logit contributions | |
le | +0.006 |
ling | +0.006 |
behind | +0.006 |
inside | +0.006 |
pe | +0.005 |
Neg logit contributions | |
Deal | -0.015 |
Not | -0.013 |
Okay | -0.013 |
Jake | -0.013 |
Thanks | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | +2.30e-01 |
Pos logit contributions | |
Deal | +0.121 |
Not | +0.109 |
Okay | +0.108 |
They | +0.104 |
Thanks | +0.104 |
Neg logit contributions | |
le | -0.045 |
pe | -0.038 |
behind | -0.038 |
ling | -0.038 |
les | -0.036 |
Token |
|
Token ID | 198 |
Feature activation | +2.00e-01 |
Pos logit contributions | |
Deal | +0.504 |
Enough | +0.392 |
Thanks | +0.391 |
liv | +0.389 |
Acc | +0.384 |
Neg logit contributions | |
inside | -0.353 |
behind | -0.351 |
on | -0.345 |
. | -0.342 |
pe | -0.327 |
Token |
|
Token ID | 198 |
Feature activation | +1.99e-01 |
Pos logit contributions | |
Deal | +0.421 |
Not | +0.338 |
Okay | +0.332 |
Thanks | +0.331 |
Sorry | +0.330 |
Neg logit contributions | |
inside | -0.338 |
le | -0.334 |
behind | -0.330 |
ling | -0.323 |
pe | -0.312 |
Token | " |
Token ID | 1 |
Feature activation | +1.11e-01 |
Pos logit contributions | |
Deal | +0.461 |
Not | +0.393 |
Okay | +0.373 |
Acc | +0.366 |
idents | +0.364 |
Neg logit contributions | |
le | -0.276 |
ling | -0.276 |
inside | -0.272 |
behind | -0.265 |
away | -0.264 |
Token | How |
Token ID | 2437 |
Feature activation | +5.40e-02 |
Pos logit contributions | |
Deal | +0.181 |
Tyler | +0.145 |
Jake | +0.141 |
Johnny | +0.137 |
ossible | +0.136 |
Neg logit contributions | |
pe | -0.221 |
le | -0.220 |
ring | -0.217 |
behind | -0.214 |
inside | -0.213 |
Token | do |
Token ID | 466 |
Feature activation | -2.34e-02 |
Pos logit contributions | |
Deal | +0.125 |
Thanks | +0.103 |
Not | +0.103 |
Okay | +0.103 |
Jake | +0.102 |
Neg logit contributions | |
le | -0.088 |
inside | -0.082 |
behind | -0.082 |
ling | -0.081 |
pe | -0.078 |
Token | the |
Token ID | 262 |
Feature activation | +1.85e-02 |
Pos logit contributions | |
le | +0.014 |
ling | +0.014 |
behind | +0.013 |
inside | +0.013 |
pe | +0.013 |
Neg logit contributions | |
Deal | -0.016 |
Not | -0.013 |
Jake | -0.013 |
Okay | -0.013 |
Thanks | -0.012 |
Token | fairy |
Token ID | 25607 |
Feature activation | -1.58e-01 |
Pos logit contributions | |
Deal | +0.042 |
Not | +0.035 |
Okay | +0.034 |
Something | +0.034 |
Thanks | +0.034 |
Neg logit contributions | |
le | -0.029 |
ling | -0.026 |
behind | -0.025 |
away | -0.023 |
les | -0.023 |
Token | and |
Token ID | 290 |
Feature activation | +5.87e-02 |
Pos logit contributions | |
le | +0.206 |
b | +0.193 |
ling | +0.193 |
ls | +0.192 |
pe | +0.190 |
Neg logit contributions | |
Deal | -0.328 |
Not | -0.291 |
They | -0.290 |
She | -0.290 |
ossible | -0.282 |
Token | said |
Token ID | 531 |
Feature activation | +5.32e-02 |
Pos logit contributions | |
Deal | +0.121 |
Not | +0.105 |
Okay | +0.103 |
Thanks | +0.102 |
Jake | +0.101 |
Neg logit contributions | |
le | -0.104 |
ling | -0.099 |
behind | -0.093 |
inside | -0.090 |
les | -0.088 |
Token | goodbye |
Token ID | 24829 |
Feature activation | -1.53e-02 |
Pos logit contributions | |
Deal | +0.120 |
Not | +0.108 |
Okay | +0.104 |
Thanks | +0.101 |
Acc | +0.099 |
Neg logit contributions | |
le | -0.088 |
ling | -0.079 |
behind | -0.079 |
pe | -0.073 |
inside | -0.073 |
Token | . |
Token ID | 13 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
le | +0.022 |
ling | +0.020 |
les | +0.020 |
pe | +0.018 |
egg | +0.018 |
Neg logit contributions | |
Deal | -0.034 |
They | -0.032 |
Not | -0.032 |
Okay | -0.031 |
Thanks | -0.030 |
Token | From |
Token ID | 3574 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
Deal | +0.364 |
Thanks | +0.270 |
Jake | +0.269 |
Enough | +0.267 |
Luke | +0.257 |
Neg logit contributions | |
inside | -0.493 |
behind | -0.477 |
pe | -0.458 |
. | -0.439 |
on | -0.437 |
Token | that |
Token ID | 326 |
Feature activation | +6.66e-02 |
Pos logit contributions | |
Deal | +0.205 |
Thanks | +0.176 |
Not | +0.171 |
Jake | +0.170 |
Okay | +0.169 |
Neg logit contributions | |
le | -0.200 |
ling | -0.191 |
les | -0.179 |
pe | -0.175 |
behind | -0.171 |
Token | day |
Token ID | 1110 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
Deal | +0.129 |
Not | +0.116 |
ossible | +0.109 |
Thanks | +0.108 |
Jake | +0.107 |
Neg logit contributions | |
le | -0.124 |
behind | -0.117 |
ling | -0.112 |
pe | -0.110 |
inside | -0.108 |
Token | on |
Token ID | 319 |
Feature activation | -5.29e-03 |
Pos logit contributions | |
le | +0.008 |
ling | +0.006 |
behind | +0.006 |
pe | +0.005 |
inside | +0.005 |
Neg logit contributions | |
Deal | -0.042 |
Not | -0.038 |
Okay | -0.037 |
Jake | -0.037 |
Thanks | -0.036 |
Token | , |
Token ID | 11 |
Feature activation | +8.83e-02 |
Pos logit contributions | |
le | +0.005 |
ling | +0.004 |
behind | +0.004 |
inside | +0.004 |
pe | +0.004 |
Neg logit contributions | |
Deal | -0.016 |
Not | -0.014 |
Okay | -0.014 |
Jake | -0.014 |
Thanks | -0.013 |
Token | did |
Token ID | 750 |
Feature activation | +4.87e-02 |
Pos logit contributions | |
le | +0.136 |
ling | +0.132 |
behind | +0.127 |
an | +0.116 |
island | +0.116 |
Neg logit contributions | |
Deal | -0.189 |
Not | -0.167 |
Okay | -0.157 |
They | -0.156 |
Thanks | -0.155 |
Token | it |
Token ID | 340 |
Feature activation | -7.00e-02 |
Pos logit contributions | |
Deal | +0.099 |
Not | +0.081 |
Okay | +0.080 |
Jake | +0.078 |
Thanks | +0.077 |
Neg logit contributions | |
le | -0.092 |
ling | -0.087 |
behind | -0.087 |
inside | -0.086 |
pe | -0.083 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.98e-01 |
Pos logit contributions | |
le | +0.074 |
ling | +0.071 |
behind | +0.068 |
egg | +0.067 |
pe | +0.065 |
Neg logit contributions | |
Deal | -0.185 |
Not | -0.168 |
Okay | -0.163 |
Jake | -0.159 |
They | -0.156 |
Token | they |
Token ID | 484 |
Feature activation | +6.21e-02 |
Pos logit contributions | |
Deal | +0.343 |
Okay | +0.256 |
Thanks | +0.254 |
Not | +0.254 |
Great | +0.249 |
Neg logit contributions | |
behind | -0.415 |
le | -0.413 |
inside | -0.404 |
pe | -0.402 |
away | -0.390 |
Token | cheered |
Token ID | 34767 |
Feature activation | -2.53e-02 |
Pos logit contributions | |
Deal | +0.139 |
Not | +0.118 |
Okay | +0.114 |
Jake | +0.113 |
Thanks | +0.111 |
Neg logit contributions | |
le | -0.101 |
ling | -0.097 |
behind | -0.095 |
inside | -0.094 |
pe | -0.088 |
Token | . |
Token ID | 13 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
le | +0.022 |
ling | +0.019 |
behind | +0.019 |
inside | +0.018 |
pe | +0.017 |
Neg logit contributions | |
Deal | -0.077 |
Not | -0.068 |
Okay | -0.067 |
Jake | -0.066 |
Thanks | -0.066 |
Token |
|
Token ID | 198 |
Feature activation | +1.96e-01 |
Pos logit contributions | |
Deal | +0.446 |
Enough | +0.333 |
Okay | +0.332 |
Tyler | +0.331 |
Not | +0.329 |
Neg logit contributions | |
inside | -0.394 |
behind | -0.393 |
. | -0.372 |
on | -0.366 |
in | -0.362 |
Token |
|
Token ID | 198 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
Deal | +0.423 |
Not | +0.340 |
Okay | +0.336 |
Sorry | +0.329 |
Acc | +0.319 |
Neg logit contributions | |
behind | -0.309 |
inside | -0.308 |
le | -0.305 |
ling | -0.295 |
pe | -0.289 |
Token | They |
Token ID | 2990 |
Feature activation | +6.86e-02 |
Pos logit contributions | |
Deal | +0.373 |
Not | +0.300 |
Okay | +0.289 |
Acc | +0.283 |
idents | +0.275 |
Neg logit contributions | |
behind | -0.341 |
le | -0.338 |
ling | -0.334 |
inside | -0.332 |
pe | -0.324 |
Token | opened |
Token ID | 4721 |
Feature activation | -4.31e-02 |
Pos logit contributions | |
Deal | +0.139 |
Not | +0.118 |
Jake | +0.117 |
Thanks | +0.116 |
Okay | +0.114 |
Neg logit contributions | |
le | -0.123 |
ling | -0.119 |
behind | -0.113 |
inside | -0.111 |
egg | -0.106 |
Token | the |
Token ID | 262 |
Feature activation | +2.20e-03 |
Pos logit contributions | |
le | +0.082 |
ling | +0.080 |
les | +0.071 |
b | +0.070 |
an | +0.070 |
Neg logit contributions | |
Deal | -0.080 |
They | -0.068 |
Jake | -0.068 |
Jake | -0.068 |
Not | -0.067 |
Token | break |
Token ID | 2270 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
Deal | +0.027 |
Not | +0.024 |
Okay | +0.022 |
Jake | +0.021 |
They | +0.021 |
Neg logit contributions | |
ling | -0.027 |
le | -0.027 |
behind | -0.027 |
inside | -0.025 |
les | -0.025 |
Token | easily |
Token ID | 3538 |
Feature activation | -7.22e-02 |
Pos logit contributions | |
ling | +0.157 |
le | +0.157 |
egg | +0.146 |
an | +0.146 |
pe | +0.146 |
Neg logit contributions | |
Deal | -0.242 |
They | -0.228 |
Not | -0.225 |
Jake | -0.223 |
Jake | -0.218 |
Token | ," |
Token ID | 553 |
Feature activation | +1.65e-01 |
Pos logit contributions | |
ling | +0.113 |
le | +0.110 |
an | +0.109 |
les | +0.101 |
lers | +0.100 |
Neg logit contributions | |
Deal | -0.138 |
They | -0.130 |
Not | -0.126 |
She | -0.126 |
Okay | -0.122 |
Token | Ben |
Token ID | 3932 |
Feature activation | +2.24e-03 |
Pos logit contributions | |
Deal | +0.323 |
Not | +0.256 |
Okay | +0.245 |
Thanks | +0.242 |
Great | +0.236 |
Neg logit contributions | |
behind | -0.309 |
pe | -0.308 |
le | -0.305 |
inside | -0.302 |
away | -0.301 |
Token | says |
Token ID | 1139 |
Feature activation | +3.98e-02 |
Pos logit contributions | |
Deal | +0.005 |
Not | +0.005 |
Okay | +0.005 |
Thanks | +0.005 |
Jake | +0.005 |
Neg logit contributions | |
le | -0.003 |
ling | -0.003 |
behind | -0.003 |
inside | -0.003 |
pe | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
Deal | +0.113 |
Not | +0.101 |
Okay | +0.099 |
They | +0.096 |
Jake | +0.095 |
Neg logit contributions | |
le | -0.042 |
ling | -0.036 |
pe | -0.034 |
behind | -0.033 |
les | -0.033 |
Token |
|
Token ID | 198 |
Feature activation | +1.96e-01 |
Pos logit contributions | |
Deal | +0.468 |
Thanks | +0.390 |
Enough | +0.382 |
liv | +0.381 |
Great | +0.375 |
Neg logit contributions | |
inside | -0.349 |
behind | -0.347 |
on | -0.339 |
. | -0.336 |
pe | -0.335 |
Token |
|
Token ID | 198 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
Deal | +0.427 |
Sorry | +0.350 |
Not | +0.349 |
Thanks | +0.345 |
Okay | +0.345 |
Neg logit contributions | |
le | -0.308 |
inside | -0.303 |
ling | -0.299 |
behind | -0.293 |
pe | -0.282 |
Token | " |
Token ID | 1 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
Deal | +0.417 |
Not | +0.358 |
Acc | +0.350 |
idents | +0.349 |
Okay | +0.341 |
Neg logit contributions | |
le | -0.299 |
ling | -0.295 |
away | -0.282 |
b | -0.272 |
inside | -0.270 |
Token | Thank |
Token ID | 10449 |
Feature activation | +3.04e-02 |
Pos logit contributions | |
Deal | +0.175 |
Tyler | +0.140 |
Kelly | +0.137 |
Jake | +0.135 |
Johnny | +0.133 |
Neg logit contributions | |
pe | -0.216 |
le | -0.215 |
behind | -0.210 |
inside | -0.208 |
away | -0.206 |
Token | you |
Token ID | 345 |
Feature activation | -1.54e-02 |
Pos logit contributions | |
Deal | +0.067 |
Not | +0.057 |
Okay | +0.056 |
Jake | +0.055 |
Thanks | +0.055 |
Neg logit contributions | |
le | -0.052 |
behind | -0.049 |
inside | -0.048 |
ling | -0.047 |
pe | -0.047 |
Token | time |
Token ID | 640 |
Feature activation | +1.18e-02 |
Pos logit contributions | |
le | +0.010 |
ling | +0.010 |
behind | +0.010 |
inside | +0.009 |
ls | +0.009 |
Neg logit contributions | |
Deal | -0.012 |
Not | -0.011 |
Okay | -0.010 |
Thanks | -0.010 |
ossible | -0.010 |
Token | before |
Token ID | 878 |
Feature activation | +4.98e-02 |
Pos logit contributions | |
Deal | +0.027 |
Not | +0.025 |
Okay | +0.024 |
They | +0.024 |
Jake | +0.024 |
Neg logit contributions | |
le | -0.016 |
ling | -0.015 |
an | -0.014 |
egg | -0.014 |
les | -0.013 |
Token | she |
Token ID | 673 |
Feature activation | +2.92e-02 |
Pos logit contributions | |
Deal | +0.103 |
Not | +0.087 |
Okay | +0.085 |
Jake | +0.082 |
Thanks | +0.082 |
Neg logit contributions | |
le | -0.093 |
ling | -0.088 |
behind | -0.086 |
inside | -0.083 |
pe | -0.081 |
Token | said |
Token ID | 531 |
Feature activation | +3.53e-02 |
Pos logit contributions | |
Deal | +0.056 |
Not | +0.048 |
Okay | +0.048 |
Jake | +0.047 |
Thanks | +0.047 |
Neg logit contributions | |
ling | -0.051 |
le | -0.049 |
behind | -0.047 |
inside | -0.046 |
egg | -0.045 |
Token | goodbye |
Token ID | 24829 |
Feature activation | -2.58e-02 |
Pos logit contributions | |
Deal | +0.073 |
Not | +0.064 |
Okay | +0.062 |
Thanks | +0.059 |
Something | +0.058 |
Neg logit contributions | |
le | -0.063 |
ling | -0.061 |
behind | -0.057 |
les | -0.055 |
pe | -0.055 |
Token | . |
Token ID | 13 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
le | +0.035 |
ling | +0.032 |
les | +0.031 |
pe | +0.028 |
behind | +0.028 |
Neg logit contributions | |
Deal | -0.060 |
Not | -0.056 |
They | -0.055 |
Okay | -0.053 |
She | -0.053 |
Token | |
Token ID | 220 |
Feature activation | +1.35e-01 |
Pos logit contributions | |
Deal | +0.408 |
Thanks | +0.327 |
Jake | +0.318 |
Enough | +0.318 |
Yeah | +0.296 |
Neg logit contributions | |
inside | -0.411 |
behind | -0.403 |
pe | -0.388 |
le | -0.371 |
up | -0.366 |
Token |
|
Token ID | 198 |
Feature activation | +1.86e-01 |
Pos logit contributions | |
Deal | +0.252 |
Thanks | +0.204 |
Not | +0.196 |
Jake | +0.195 |
Okay | +0.194 |
Neg logit contributions | |
inside | -0.255 |
le | -0.246 |
behind | -0.245 |
ling | -0.238 |
up | -0.237 |
Token |
|
Token ID | 198 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
Deal | +0.374 |
Not | +0.293 |
Okay | +0.288 |
Sorry | +0.287 |
Acc | +0.273 |
Neg logit contributions | |
le | -0.320 |
inside | -0.311 |
behind | -0.310 |
pe | -0.307 |
ling | -0.306 |
Token | Summer |
Token ID | 33560 |
Feature activation | -5.46e-02 |
Pos logit contributions | |
Deal | +0.330 |
Not | +0.245 |
Sorry | +0.237 |
Okay | +0.236 |
Acc | +0.234 |
Neg logit contributions | |
le | -0.355 |
pe | -0.339 |
ling | -0.338 |
inside | -0.336 |
behind | -0.332 |
Token | eventually |
Token ID | 4191 |
Feature activation | +4.00e-03 |
Pos logit contributions | |
behind | +0.092 |
le | +0.089 |
inside | +0.088 |
ling | +0.086 |
egg | +0.083 |
Neg logit contributions | |
Deal | -0.101 |
Not | -0.092 |
Thanks | -0.090 |
Jake | -0.089 |
They | -0.085 |
Token | smiled |
Token ID | 13541 |
Feature activation | +4.54e-03 |
Pos logit contributions | |
Deal | +0.192 |
Not | +0.166 |
Jake | +0.159 |
Okay | +0.158 |
Thanks | +0.156 |
Neg logit contributions | |
le | -0.140 |
ling | -0.136 |
behind | -0.132 |
inside | -0.128 |
away | -0.122 |
Token | and |
Token ID | 290 |
Feature activation | +6.84e-02 |
Pos logit contributions | |
Deal | +0.014 |
Not | +0.012 |
Okay | +0.012 |
Jake | +0.012 |
Thanks | +0.012 |
Neg logit contributions | |
le | -0.004 |
ling | -0.004 |
behind | -0.004 |
inside | -0.003 |
pe | -0.003 |
Token | hugged |
Token ID | 41130 |
Feature activation | -1.34e-02 |
Pos logit contributions | |
Deal | +0.148 |
Not | +0.127 |
Okay | +0.124 |
Thanks | +0.123 |
Jake | +0.122 |
Neg logit contributions | |
le | -0.118 |
ling | -0.111 |
behind | -0.108 |
inside | -0.105 |
pe | -0.102 |
Token | them |
Token ID | 606 |
Feature activation | -7.97e-02 |
Pos logit contributions | |
le | +0.025 |
ling | +0.023 |
behind | +0.023 |
inside | +0.022 |
pe | +0.022 |
Neg logit contributions | |
Deal | -0.028 |
Not | -0.023 |
Okay | -0.023 |
Jake | -0.023 |
Thanks | -0.022 |
Token | back |
Token ID | 736 |
Feature activation | -4.52e-02 |
Pos logit contributions | |
le | +0.094 |
ling | +0.083 |
pe | +0.080 |
behind | +0.079 |
inside | +0.076 |
Neg logit contributions | |
Deal | -0.217 |
Not | -0.189 |
Okay | -0.187 |
Thanks | -0.181 |
They | -0.181 |
Token | . |
Token ID | 13 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
le | +0.060 |
ling | +0.054 |
les | +0.051 |
pe | +0.046 |
egg | +0.046 |
Neg logit contributions | |
Deal | -0.111 |
They | -0.102 |
Okay | -0.099 |
Not | -0.099 |
She | -0.096 |
Token | She |
Token ID | 1375 |
Feature activation | +8.36e-02 |
Pos logit contributions | |
Deal | +0.454 |
Thanks | +0.354 |
Jake | +0.348 |
Enough | +0.344 |
Great | +0.343 |
Neg logit contributions | |
inside | -0.402 |
behind | -0.381 |
. | -0.360 |
on | -0.355 |
pe | -0.347 |
Token | picked |
Token ID | 6497 |
Feature activation | -4.59e-02 |
Pos logit contributions | |
Deal | +0.179 |
Not | +0.156 |
Okay | +0.151 |
Jake | +0.151 |
Thanks | +0.150 |
Neg logit contributions | |
le | -0.140 |
ling | -0.135 |
behind | -0.131 |
inside | -0.128 |
away | -0.124 |
Token | up |
Token ID | 510 |
Feature activation | -5.79e-02 |
Pos logit contributions | |
le | +0.041 |
ling | +0.038 |
behind | +0.035 |
island | +0.031 |
pe | +0.030 |
Neg logit contributions | |
Deal | -0.125 |
Tyler | -0.115 |
Not | -0.115 |
They | -0.114 |
First | -0.113 |
Token | the |
Token ID | 262 |
Feature activation | -8.43e-04 |
Pos logit contributions | |
le | +0.104 |
ling | +0.101 |
behind | +0.091 |
inside | +0.091 |
les | +0.087 |
Neg logit contributions | |
Deal | -0.114 |
Not | -0.100 |
Jake | -0.097 |
Tyler | -0.097 |
They | -0.095 |
Token | clay |
Token ID | 21558 |
Feature activation | -1.77e-01 |
Pos logit contributions | |
le | +0.001 |
ling | +0.001 |
les | +0.001 |
aches | +0.001 |
ck | +0.001 |
Neg logit contributions | |
Deal | -0.002 |
Okay | -0.002 |
Not | -0.002 |
Something | -0.001 |
Ryan | -0.001 |
Token | to |
Token ID | 284 |
Feature activation | -1.88e-03 |
Pos logit contributions | |
le | +0.009 |
ling | +0.008 |
behind | +0.008 |
les | +0.007 |
an | +0.007 |
Neg logit contributions | |
Deal | -0.011 |
Not | -0.010 |
Okay | -0.010 |
Jake | -0.009 |
Thanks | -0.009 |
Token | fight |
Token ID | 1907 |
Feature activation | -1.11e-01 |
Pos logit contributions | |
le | +0.003 |
behind | +0.003 |
ling | +0.003 |
inside | +0.003 |
pe | +0.003 |
Neg logit contributions | |
Deal | -0.004 |
Not | -0.003 |
Okay | -0.003 |
Jake | -0.003 |
Thanks | -0.003 |
Token | ," |
Token ID | 553 |
Feature activation | +1.58e-01 |
Pos logit contributions | |
le | +0.117 |
ling | +0.110 |
pe | +0.101 |
egg | +0.100 |
behind | +0.096 |
Neg logit contributions | |
Deal | -0.308 |
Not | -0.274 |
Jake | -0.270 |
Okay | -0.270 |
Thanks | -0.260 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.45e-02 |
Pos logit contributions | |
Deal | +0.309 |
Not | +0.245 |
Okay | +0.237 |
Thanks | +0.228 |
Jake | +0.224 |
Neg logit contributions | |
le | -0.306 |
behind | -0.301 |
pe | -0.296 |
ling | -0.293 |
inside | -0.293 |
Token | says |
Token ID | 1139 |
Feature activation | +3.07e-02 |
Pos logit contributions | |
Deal | +0.038 |
Not | +0.033 |
Okay | +0.032 |
Thanks | +0.032 |
Jake | +0.031 |
Neg logit contributions | |
le | -0.019 |
ling | -0.018 |
pe | -0.017 |
behind | -0.017 |
inside | -0.017 |
Token | . |
Token ID | 13 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
Deal | +0.084 |
Not | +0.077 |
Okay | +0.075 |
Thanks | +0.072 |
They | +0.072 |
Neg logit contributions | |
le | -0.034 |
pe | -0.029 |
les | -0.028 |
ling | -0.028 |
behind | -0.027 |
Token |
|
Token ID | 198 |
Feature activation | +1.90e-01 |
Pos logit contributions | |
Deal | +0.460 |
Thanks | +0.370 |
Jake | +0.369 |
Great | +0.356 |
Johnny | +0.354 |
Neg logit contributions | |
behind | -0.355 |
inside | -0.349 |
. | -0.343 |
on | -0.336 |
up | -0.333 |
Token |
|
Token ID | 198 |
Feature activation | +1.88e-01 |
Pos logit contributions | |
Deal | +0.416 |
Sorry | +0.345 |
Thanks | +0.336 |
Okay | +0.334 |
Not | +0.332 |
Neg logit contributions | |
inside | -0.289 |
le | -0.285 |
behind | -0.282 |
ling | -0.271 |
up | -0.267 |
Token | " |
Token ID | 1 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
Deal | +0.403 |
Not | +0.336 |
Acc | +0.335 |
Okay | +0.328 |
Sorry | +0.325 |
Neg logit contributions | |
le | -0.280 |
away | -0.274 |
ling | -0.270 |
inside | -0.265 |
- | -0.260 |
Token | Yes |
Token ID | 5297 |
Feature activation | -4.17e-03 |
Pos logit contributions | |
Deal | +0.167 |
ossible | +0.127 |
Kelly | +0.126 |
Not | +0.124 |
Acc | +0.123 |
Neg logit contributions | |
le | -0.209 |
pe | -0.206 |
ling | -0.205 |
behind | -0.200 |
away | -0.199 |
Token | , |
Token ID | 11 |
Feature activation | +4.81e-02 |
Pos logit contributions | |
le | +0.003 |
behind | +0.002 |
inside | +0.002 |
ling | +0.002 |
away | +0.002 |
Neg logit contributions | |
Deal | -0.013 |
Not | -0.012 |
Thanks | -0.012 |
Jake | -0.012 |
Okay | -0.012 |
Token | hugged |
Token ID | 41130 |
Feature activation | -1.68e-02 |
Pos logit contributions | |
le | +0.014 |
behind | +0.013 |
ling | +0.013 |
inside | +0.013 |
pe | +0.012 |
Neg logit contributions | |
Deal | -0.027 |
Not | -0.024 |
Okay | -0.023 |
Jake | -0.022 |
Thanks | -0.022 |
Token | Spot |
Token ID | 15899 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
le | +0.027 |
ling | +0.025 |
behind | +0.025 |
inside | +0.024 |
away | +0.023 |
Neg logit contributions | |
Deal | -0.040 |
Not | -0.034 |
Okay | -0.033 |
Jake | -0.033 |
Thanks | -0.032 |
Token | and |
Token ID | 290 |
Feature activation | +5.13e-02 |
Pos logit contributions | |
le | +0.118 |
behind | +0.105 |
pe | +0.105 |
ling | +0.103 |
inside | +0.100 |
Neg logit contributions | |
Deal | -0.344 |
Okay | -0.300 |
Not | -0.299 |
They | -0.290 |
Thanks | -0.281 |
Token | thanked |
Token ID | 26280 |
Feature activation | +1.48e-02 |
Pos logit contributions | |
Deal | +0.113 |
Not | +0.096 |
Okay | +0.093 |
Jake | +0.092 |
Thanks | +0.091 |
Neg logit contributions | |
le | -0.088 |
ling | -0.083 |
behind | -0.081 |
inside | -0.080 |
pe | -0.077 |
Token | him |
Token ID | 683 |
Feature activation | -5.22e-02 |
Pos logit contributions | |
Deal | +0.032 |
Not | +0.027 |
Okay | +0.026 |
Thanks | +0.025 |
Jake | +0.025 |
Neg logit contributions | |
le | -0.027 |
behind | -0.025 |
ling | -0.025 |
inside | -0.025 |
away | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
le | +0.060 |
ling | +0.056 |
pe | +0.055 |
egg | +0.055 |
behind | +0.051 |
Neg logit contributions | |
Deal | -0.137 |
Okay | -0.117 |
Not | -0.116 |
They | -0.114 |
Tyler | -0.109 |
Token | They |
Token ID | 1119 |
Feature activation | +7.68e-02 |
Pos logit contributions | |
Deal | +0.398 |
Enough | +0.307 |
Jake | +0.305 |
Thanks | +0.302 |
Luke | +0.300 |
Neg logit contributions | |
inside | -0.463 |
behind | -0.445 |
pe | -0.425 |
up | -0.414 |
. | -0.404 |
Token | were |
Token ID | 547 |
Feature activation | +4.03e-02 |
Pos logit contributions | |
Deal | +0.160 |
Not | +0.135 |
Okay | +0.132 |
Thanks | +0.130 |
Jake | +0.130 |
Neg logit contributions | |
le | -0.138 |
ling | -0.131 |
behind | -0.129 |
inside | -0.126 |
pe | -0.121 |
Token | sorry |
Token ID | 7926 |
Feature activation | -5.69e-03 |
Pos logit contributions | |
Deal | +0.090 |
Not | +0.076 |
Okay | +0.075 |
Jake | +0.073 |
Thanks | +0.073 |
Neg logit contributions | |
le | -0.068 |
ling | -0.064 |
behind | -0.064 |
inside | -0.062 |
pe | -0.060 |
Token | they |
Token ID | 484 |
Feature activation | +3.16e-03 |
Pos logit contributions | |
inside | +0.005 |
le | +0.005 |
ling | +0.005 |
behind | +0.005 |
away | +0.004 |
Neg logit contributions | |
Deal | -0.018 |
Jake | -0.015 |
Thanks | -0.015 |
Luke | -0.015 |
David | -0.015 |
Token | did |
Token ID | 750 |
Feature activation | +1.52e-02 |
Pos logit contributions | |
Deal | +0.007 |
Not | +0.006 |
Okay | +0.005 |
Jake | +0.005 |
Thanks | +0.005 |
Neg logit contributions | |
le | -0.006 |
ling | -0.005 |
behind | -0.005 |
inside | -0.005 |
pe | -0.005 |
Token | much |
Token ID | 881 |
Feature activation | +9.80e-04 |
Pos logit contributions | |
Deal | +0.172 |
Not | +0.151 |
Okay | +0.146 |
Thanks | +0.140 |
Jake | +0.139 |
Neg logit contributions | |
le | -0.129 |
ling | -0.117 |
behind | -0.115 |
inside | -0.111 |
les | -0.110 |
Token | fun |
Token ID | 1257 |
Feature activation | -4.52e-02 |
Pos logit contributions | |
Deal | +0.002 |
Not | +0.002 |
Okay | +0.002 |
They | +0.001 |
Yeah | +0.001 |
Neg logit contributions | |
le | -0.002 |
ling | -0.002 |
an | -0.002 |
les | -0.002 |
behind | -0.002 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.97e-01 |
Pos logit contributions | |
le | +0.055 |
ling | +0.052 |
behind | +0.046 |
pe | +0.045 |
les | +0.043 |
Neg logit contributions | |
Deal | -0.114 |
Not | -0.104 |
Jake | -0.101 |
Okay | -0.100 |
They | -0.099 |
Token | said |
Token ID | 531 |
Feature activation | +6.29e-02 |
Pos logit contributions | |
Deal | +0.360 |
Not | +0.270 |
Thanks | +0.265 |
Okay | +0.265 |
Great | +0.261 |
Neg logit contributions | |
behind | -0.393 |
le | -0.388 |
inside | -0.384 |
pe | -0.380 |
away | -0.369 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.21e-02 |
Pos logit contributions | |
Deal | +0.102 |
Not | +0.081 |
Okay | +0.078 |
Jake | +0.077 |
Thanks | +0.075 |
Neg logit contributions | |
le | -0.143 |
ling | -0.138 |
behind | -0.136 |
inside | -0.134 |
pe | -0.131 |
Token | . |
Token ID | 13 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
Deal | +0.033 |
Not | +0.030 |
Okay | +0.029 |
Jake | +0.029 |
They | +0.028 |
Neg logit contributions | |
le | -0.012 |
ling | -0.011 |
behind | -0.011 |
egg | -0.010 |
inside | -0.010 |
Token | |
Token ID | 220 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
Deal | +0.456 |
Jake | +0.363 |
Thanks | +0.360 |
David | +0.358 |
Enough | +0.357 |
Neg logit contributions | |
inside | -0.373 |
behind | -0.361 |
on | -0.346 |
. | -0.335 |
again | -0.332 |
Token |
|
Token ID | 198 |
Feature activation | +1.93e-01 |
Pos logit contributions | |
Deal | +0.263 |
Thanks | +0.205 |
Not | +0.205 |
Okay | +0.201 |
Jake | +0.199 |
Neg logit contributions | |
inside | -0.276 |
le | -0.270 |
behind | -0.270 |
ling | -0.264 |
up | -0.257 |
Token |
|
Token ID | 198 |
Feature activation | +1.90e-01 |
Pos logit contributions | |
Deal | +0.402 |
Not | +0.325 |
Okay | +0.325 |
Sorry | +0.311 |
Acc | +0.299 |
Neg logit contributions | |
le | -0.331 |
inside | -0.319 |
ling | -0.318 |
behind | -0.311 |
away | -0.302 |
Token | The |
Token ID | 464 |
Feature activation | +9.47e-02 |
Pos logit contributions | |
Deal | +0.365 |
Not | +0.296 |
Okay | +0.293 |
Acc | +0.284 |
Tyler | +0.275 |
Neg logit contributions | |
le | -0.342 |
ling | -0.334 |
away | -0.323 |
inside | -0.314 |
behind | -0.305 |
Token | butterflies |
Token ID | 44571 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
Deal | +0.216 |
Not | +0.181 |
Okay | +0.181 |
Jake | +0.177 |
Thanks | +0.177 |
Neg logit contributions | |
le | -0.159 |
ling | -0.145 |
behind | -0.142 |
inside | -0.135 |
away | -0.133 |
Token | have |
Token ID | 423 |
Feature activation | +6.98e-03 |
Pos logit contributions | |
Deal | +0.057 |
Not | +0.050 |
Okay | +0.049 |
Thanks | +0.045 |
Johnny | +0.045 |
Neg logit contributions | |
le | -0.053 |
ling | -0.053 |
behind | -0.051 |
les | -0.050 |
inside | -0.047 |
Token | it |
Token ID | 340 |
Feature activation | -9.40e-02 |
Pos logit contributions | |
Deal | +0.015 |
Not | +0.014 |
Okay | +0.013 |
Thanks | +0.012 |
Jake | +0.012 |
Neg logit contributions | |
le | -0.012 |
ling | -0.012 |
behind | -0.011 |
inside | -0.010 |
les | -0.010 |
Token | ," |
Token ID | 553 |
Feature activation | +1.67e-01 |
Pos logit contributions | |
le | +0.127 |
ling | +0.116 |
pe | +0.107 |
egg | +0.103 |
les | +0.103 |
Neg logit contributions | |
Deal | -0.224 |
Not | -0.203 |
Okay | -0.198 |
They | -0.191 |
Thanks | -0.186 |
Token | Lily |
Token ID | 20037 |
Feature activation | +2.72e-02 |
Pos logit contributions | |
Deal | +0.315 |
Not | +0.246 |
Okay | +0.237 |
Thanks | +0.234 |
Jake | +0.231 |
Neg logit contributions | |
le | -0.325 |
behind | -0.323 |
pe | -0.320 |
away | -0.317 |
inside | -0.313 |
Token | says |
Token ID | 1139 |
Feature activation | +4.28e-02 |
Pos logit contributions | |
Deal | +0.068 |
Not | +0.059 |
Okay | +0.058 |
Thanks | +0.058 |
Jake | +0.057 |
Neg logit contributions | |
le | -0.039 |
ling | -0.036 |
behind | -0.035 |
inside | -0.035 |
away | -0.034 |
Token | . |
Token ID | 13 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
Deal | +0.115 |
Not | +0.104 |
Okay | +0.102 |
They | +0.098 |
Acc | +0.097 |
Neg logit contributions | |
le | -0.050 |
ling | -0.043 |
island | -0.042 |
pe | -0.042 |
les | -0.042 |
Token |
|
Token ID | 198 |
Feature activation | +1.96e-01 |
Pos logit contributions | |
Deal | +0.481 |
Thanks | +0.405 |
liv | +0.392 |
Jake | +0.389 |
Tyler | +0.383 |
Neg logit contributions | |
inside | -0.352 |
behind | -0.343 |
on | -0.334 |
. | -0.333 |
again | -0.319 |
Token |
|
Token ID | 198 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
Deal | +0.426 |
Sorry | +0.342 |
Thanks | +0.342 |
Not | +0.341 |
Okay | +0.336 |
Neg logit contributions | |
inside | -0.313 |
le | -0.301 |
behind | -0.298 |
ling | -0.289 |
up | -0.284 |
Token | " |
Token ID | 1 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
Deal | +0.426 |
Not | +0.359 |
Okay | +0.346 |
Acc | +0.342 |
Tyler | +0.335 |
Neg logit contributions | |
le | -0.291 |
ling | -0.282 |
inside | -0.279 |
away | -0.279 |
behind | -0.270 |
Token | It |
Token ID | 1026 |
Feature activation | -3.05e-02 |
Pos logit contributions | |
Deal | +0.178 |
Tyler | +0.134 |
Jake | +0.133 |
Kelly | +0.128 |
Not | +0.128 |
Neg logit contributions | |
pe | -0.223 |
le | -0.221 |
behind | -0.221 |
away | -0.216 |
inside | -0.215 |
Token | 's |
Token ID | 338 |
Feature activation | +1.12e-02 |
Pos logit contributions | |
le | +0.046 |
behind | +0.045 |
ling | +0.042 |
inside | +0.041 |
pe | +0.038 |
Neg logit contributions | |
Deal | -0.070 |
Not | -0.064 |
Okay | -0.060 |
They | -0.059 |
Something | -0.057 |
Token | much |
Token ID | 881 |
Feature activation | -2.23e-02 |
Pos logit contributions | |
Deal | +0.108 |
Not | +0.095 |
Okay | +0.091 |
Jake | +0.090 |
Thanks | +0.089 |
Neg logit contributions | |
le | -0.095 |
behind | -0.089 |
ling | -0.086 |
inside | -0.086 |
away | -0.083 |
Token | noise |
Token ID | 7838 |
Feature activation | -1.24e-01 |
Pos logit contributions | |
le | +0.035 |
ling | +0.032 |
behind | +0.032 |
an | +0.031 |
egg | +0.030 |
Neg logit contributions | |
Deal | -0.047 |
Not | -0.043 |
Okay | -0.042 |
They | -0.041 |
Acc | -0.040 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.89e-01 |
Pos logit contributions | |
egg | +0.136 |
le | +0.133 |
pe | +0.127 |
an | +0.124 |
ling | +0.121 |
Neg logit contributions | |
Deal | -0.285 |
Not | -0.280 |
They | -0.273 |
Jake | -0.267 |
Jake | -0.264 |
Token | Mia |
Token ID | 32189 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
Deal | +0.327 |
Not | +0.245 |
Okay | +0.240 |
Thanks | +0.239 |
Jake | +0.233 |
Neg logit contributions | |
le | -0.399 |
behind | -0.392 |
inside | -0.384 |
pe | -0.379 |
ling | -0.378 |
Token | says |
Token ID | 1139 |
Feature activation | +4.32e-02 |
Pos logit contributions | |
Deal | +0.026 |
Not | +0.023 |
Okay | +0.022 |
Jake | +0.022 |
Thanks | +0.021 |
Neg logit contributions | |
le | -0.018 |
ling | -0.017 |
behind | -0.017 |
inside | -0.017 |
away | -0.016 |
Token | . |
Token ID | 13 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
Deal | +0.109 |
Not | +0.099 |
Okay | +0.098 |
They | +0.095 |
Acc | +0.095 |
Neg logit contributions | |
le | -0.053 |
ling | -0.046 |
les | -0.045 |
island | -0.045 |
an | -0.044 |
Token | " |
Token ID | 366 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
Deal | +0.384 |
Thanks | +0.323 |
Johnny | +0.311 |
Jake | +0.304 |
David | +0.299 |
Neg logit contributions | |
inside | -0.432 |
pe | -0.432 |
behind | -0.431 |
on | -0.416 |
. | -0.401 |
Token | You |
Token ID | 1639 |
Feature activation | +6.00e-02 |
Pos logit contributions | |
Deal | +0.188 |
omo | +0.155 |
Tyler | +0.155 |
addy | +0.152 |
Jake | +0.143 |
Neg logit contributions | |
le | -0.205 |
behind | -0.193 |
inside | -0.191 |
ling | -0.190 |
pe | -0.189 |
Token | are |
Token ID | 389 |
Feature activation | +8.85e-03 |
Pos logit contributions | |
Deal | +0.113 |
Not | +0.098 |
Okay | +0.093 |
Something | +0.088 |
Jake | +0.087 |
Neg logit contributions | |
le | -0.118 |
ling | -0.117 |
behind | -0.114 |
inside | -0.108 |
les | -0.105 |
Token | hurting |
Token ID | 21530 |
Feature activation | -7.15e-02 |
Pos logit contributions | |
Deal | +0.021 |
Not | +0.018 |
Okay | +0.018 |
Jake | +0.017 |
Thanks | +0.017 |
Neg logit contributions | |
le | -0.014 |
ling | -0.013 |
behind | -0.013 |
inside | -0.013 |
pe | -0.012 |
Token | my |
Token ID | 616 |
Feature activation | -3.85e-02 |
Pos logit contributions | |
le | +0.119 |
ling | +0.112 |
behind | +0.110 |
inside | +0.109 |
pe | +0.105 |
Neg logit contributions | |
Deal | -0.161 |
Not | -0.137 |
Okay | -0.134 |
Jake | -0.132 |
Thanks | -0.130 |
Token | of |
Token ID | 286 |
Feature activation | +3.47e-02 |
Pos logit contributions | |
Deal | +0.039 |
Not | +0.037 |
Tyler | +0.035 |
Jake | +0.035 |
Jake | +0.035 |
Neg logit contributions | |
le | -0.032 |
ling | -0.031 |
behind | -0.029 |
away | -0.029 |
inside | -0.029 |
Token | that |
Token ID | 326 |
Feature activation | +3.78e-02 |
Pos logit contributions | |
Deal | +0.072 |
Not | +0.063 |
Jake | +0.059 |
Okay | +0.059 |
Thanks | +0.057 |
Neg logit contributions | |
le | -0.063 |
ling | -0.061 |
behind | -0.057 |
inside | -0.057 |
away | -0.055 |
Token | ," |
Token ID | 553 |
Feature activation | +1.63e-01 |
Pos logit contributions | |
Deal | +0.071 |
Not | +0.066 |
Jake | +0.062 |
Okay | +0.061 |
Jake | +0.061 |
Neg logit contributions | |
le | -0.073 |
ling | -0.073 |
les | -0.066 |
an | -0.064 |
away | -0.064 |
Token | Ben |
Token ID | 3932 |
Feature activation | -1.08e-04 |
Pos logit contributions | |
Deal | +0.322 |
Not | +0.251 |
Okay | +0.244 |
Thanks | +0.238 |
Jake | +0.233 |
Neg logit contributions | |
le | -0.310 |
behind | -0.305 |
pe | -0.303 |
inside | -0.299 |
ling | -0.296 |
Token | says |
Token ID | 1139 |
Feature activation | +3.85e-02 |
Pos logit contributions | |
le | +0.000 |
behind | +0.000 |
ling | +0.000 |
inside | +0.000 |
pe | +0.000 |
Neg logit contributions | |
Deal | -0.000 |
Not | -0.000 |
Okay | -0.000 |
Jake | -0.000 |
Thanks | -0.000 |
Token | . |
Token ID | 13 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
Deal | +0.100 |
Not | +0.091 |
Okay | +0.090 |
They | +0.088 |
Acc | +0.086 |
Neg logit contributions | |
le | -0.046 |
les | -0.039 |
ling | -0.038 |
egg | -0.038 |
pe | -0.038 |
Token |
|
Token ID | 198 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
Deal | +0.443 |
Thanks | +0.359 |
Enough | +0.356 |
liv | +0.354 |
Acc | +0.342 |
Neg logit contributions | |
inside | -0.388 |
behind | -0.386 |
on | -0.380 |
pe | -0.372 |
. | -0.366 |
Token |
|
Token ID | 198 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
Deal | +0.409 |
Not | +0.334 |
Sorry | +0.327 |
Okay | +0.327 |
Thanks | +0.326 |
Neg logit contributions | |
le | -0.323 |
inside | -0.313 |
ling | -0.313 |
behind | -0.305 |
pe | -0.298 |
Token | " |
Token ID | 1 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
Deal | +0.418 |
Not | +0.361 |
Acc | +0.348 |
idents | +0.344 |
Okay | +0.342 |
Neg logit contributions | |
le | -0.301 |
ling | -0.295 |
away | -0.282 |
b | -0.277 |
inside | -0.270 |
Token | You |
Token ID | 1639 |
Feature activation | +6.45e-02 |
Pos logit contributions | |
Deal | +0.176 |
Tyler | +0.132 |
Kelly | +0.131 |
Not | +0.131 |
Jake | +0.128 |
Neg logit contributions | |
pe | -0.223 |
le | -0.219 |
behind | -0.215 |
inside | -0.211 |
ling | -0.211 |
Token | are |
Token ID | 389 |
Feature activation | +8.86e-03 |
Pos logit contributions | |
Deal | +0.129 |
Not | +0.108 |
Okay | +0.104 |
Jake | +0.101 |
Thanks | +0.100 |
Neg logit contributions | |
le | -0.125 |
ling | -0.120 |
behind | -0.119 |
inside | -0.116 |
pe | -0.112 |
Token | the |
Token ID | 262 |
Feature activation | +5.56e-02 |
Pos logit contributions | |
le | +0.013 |
ling | +0.012 |
behind | +0.012 |
inside | +0.011 |
pe | +0.011 |
Neg logit contributions | |
Deal | -0.016 |
Not | -0.014 |
Okay | -0.013 |
Jake | -0.013 |
Thanks | -0.013 |
Token | stars |
Token ID | 5788 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
Deal | +0.133 |
Not | +0.113 |
Okay | +0.110 |
Jake | +0.108 |
Thanks | +0.107 |
Neg logit contributions | |
le | -0.086 |
ling | -0.077 |
behind | -0.075 |
inside | -0.072 |
les | -0.070 |
Token | ," |
Token ID | 553 |
Feature activation | +1.52e-01 |
Pos logit contributions | |
le | +0.147 |
behind | +0.138 |
egg | +0.136 |
an | +0.135 |
ling | +0.132 |
Neg logit contributions | |
Deal | -0.263 |
Not | -0.262 |
They | -0.257 |
She | -0.246 |
Jake | -0.239 |
Token | she |
Token ID | 673 |
Feature activation | +6.49e-02 |
Pos logit contributions | |
Deal | +0.290 |
Not | +0.227 |
Okay | +0.221 |
Thanks | +0.215 |
Jake | +0.211 |
Neg logit contributions | |
le | -0.301 |
behind | -0.294 |
inside | -0.289 |
pe | -0.289 |
ling | -0.284 |
Token | says |
Token ID | 1139 |
Feature activation | +3.88e-02 |
Pos logit contributions | |
Deal | +0.148 |
Not | +0.126 |
Okay | +0.123 |
Jake | +0.121 |
Thanks | +0.118 |
Neg logit contributions | |
le | -0.104 |
ling | -0.099 |
behind | -0.098 |
inside | -0.096 |
pe | -0.092 |
Token | . |
Token ID | 13 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
Deal | +0.105 |
Not | +0.094 |
Okay | +0.093 |
They | +0.089 |
Jake | +0.088 |
Neg logit contributions | |
le | -0.045 |
ling | -0.039 |
les | -0.039 |
pe | -0.037 |
an | -0.036 |
Token |
|
Token ID | 198 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
Deal | +0.434 |
Thanks | +0.365 |
Enough | +0.342 |
Great | +0.340 |
Jake | +0.339 |
Neg logit contributions | |
behind | -0.395 |
inside | -0.389 |
pe | -0.368 |
on | -0.368 |
. | -0.358 |
Token |
|
Token ID | 198 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
Deal | +0.396 |
Thanks | +0.315 |
Not | +0.313 |
Okay | +0.313 |
Sorry | +0.306 |
Neg logit contributions | |
inside | -0.339 |
behind | -0.337 |
le | -0.335 |
pe | -0.324 |
ling | -0.316 |
Token | Anna |
Token ID | 31160 |
Feature activation | +2.76e-03 |
Pos logit contributions | |
Deal | +0.401 |
Not | +0.329 |
Okay | +0.323 |
Winning | +0.311 |
Acc | +0.309 |
Neg logit contributions | |
le | -0.300 |
behind | -0.298 |
pe | -0.296 |
inside | -0.293 |
ling | -0.285 |
Token | and |
Token ID | 290 |
Feature activation | +7.32e-02 |
Pos logit contributions | |
Deal | +0.007 |
Not | +0.006 |
Okay | +0.006 |
Jake | +0.006 |
Thanks | +0.005 |
Neg logit contributions | |
le | -0.004 |
ling | -0.004 |
behind | -0.004 |
inside | -0.004 |
away | -0.003 |
Token | Ben |
Token ID | 3932 |
Feature activation | -2.43e-02 |
Pos logit contributions | |
Deal | +0.097 |
Not | +0.072 |
Okay | +0.069 |
Thanks | +0.069 |
Jake | +0.067 |
Neg logit contributions | |
le | -0.189 |
behind | -0.184 |
inside | -0.182 |
ling | -0.179 |
pe | -0.179 |
Token | cro |
Token ID | 6763 |
Feature activation | -8.51e-02 |
Pos logit contributions | |
Deal | +0.115 |
Not | +0.099 |
Okay | +0.096 |
Jake | +0.092 |
Thanks | +0.091 |
Neg logit contributions | |
le | -0.080 |
ling | -0.076 |
behind | -0.072 |
inside | -0.071 |
les | -0.067 |
Token | aks |
Token ID | 4730 |
Feature activation | -1.34e-01 |
Pos logit contributions | |
behind | +0.135 |
le | +0.135 |
ling | +0.129 |
inside | +0.129 |
egg | +0.125 |
Neg logit contributions | |
Deal | -0.190 |
Not | -0.159 |
Jake | -0.154 |
Okay | -0.154 |
Thanks | -0.149 |
Token | ," |
Token ID | 553 |
Feature activation | +1.70e-01 |
Pos logit contributions | |
le | +0.155 |
egg | +0.139 |
b | +0.137 |
ling | +0.136 |
island | +0.136 |
Neg logit contributions | |
Deal | -0.315 |
They | -0.290 |
Not | -0.289 |
Jake | -0.278 |
She | -0.276 |
Token | he |
Token ID | 339 |
Feature activation | +4.36e-02 |
Pos logit contributions | |
Deal | +0.280 |
Not | +0.210 |
Okay | +0.201 |
Thanks | +0.198 |
Jake | +0.195 |
Neg logit contributions | |
pe | -0.376 |
behind | -0.375 |
le | -0.372 |
inside | -0.364 |
ling | -0.352 |
Token | said |
Token ID | 531 |
Feature activation | +4.75e-02 |
Pos logit contributions | |
Deal | +0.098 |
Not | +0.083 |
Okay | +0.081 |
Jake | +0.080 |
Thanks | +0.079 |
Neg logit contributions | |
le | -0.073 |
ling | -0.069 |
behind | -0.068 |
inside | -0.067 |
pe | -0.065 |
Token | . |
Token ID | 13 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
Deal | +0.118 |
Not | +0.106 |
Okay | +0.106 |
They | +0.104 |
Their | +0.102 |
Neg logit contributions | |
le | -0.059 |
ling | -0.054 |
les | -0.053 |
island | -0.052 |
an | -0.051 |
Token |
|
Token ID | 198 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
Deal | +0.494 |
Thanks | +0.385 |
Acc | +0.384 |
Okay | +0.381 |
Enough | +0.376 |
Neg logit contributions | |
behind | -0.364 |
inside | -0.358 |
pe | -0.345 |
on | -0.341 |
in | -0.328 |
Token |
|
Token ID | 198 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
Deal | +0.405 |
Not | +0.323 |
Okay | +0.316 |
Sorry | +0.313 |
Thanks | +0.312 |
Neg logit contributions | |
le | -0.344 |
behind | -0.336 |
inside | -0.334 |
pe | -0.317 |
ling | -0.316 |
Token | S |
Token ID | 50 |
Feature activation | +4.59e-02 |
Pos logit contributions | |
Deal | +0.362 |
Not | +0.297 |
Okay | +0.278 |
Acc | +0.272 |
idents | +0.268 |
Neg logit contributions | |
le | -0.368 |
behind | -0.354 |
ling | -0.353 |
away | -0.347 |
inside | -0.347 |
Token | ara |
Token ID | 3301 |
Feature activation | -8.72e-03 |
Pos logit contributions | |
Deal | +0.063 |
Not | +0.047 |
Okay | +0.045 |
Jake | +0.044 |
Thanks | +0.043 |
Neg logit contributions | |
le | -0.116 |
ling | -0.113 |
behind | -0.112 |
inside | -0.111 |
pe | -0.107 |
Token | made |
Token ID | 925 |
Feature activation | +2.25e-02 |
Pos logit contributions | |
le | +0.013 |
ling | +0.012 |
behind | +0.012 |
inside | +0.011 |
away | +0.011 |
Neg logit contributions | |
Deal | -0.021 |
Not | -0.018 |
Jake | -0.018 |
Okay | -0.017 |
Thanks | -0.017 |
Token | learned |
Token ID | 4499 |
Feature activation | +1.27e-02 |
Pos logit contributions | |
Deal | +0.062 |
Not | +0.052 |
Okay | +0.050 |
Jake | +0.049 |
Thanks | +0.049 |
Neg logit contributions | |
le | -0.058 |
ling | -0.055 |
behind | -0.055 |
inside | -0.053 |
pe | -0.052 |
Token | a |
Token ID | 257 |
Feature activation | +2.40e-02 |
Pos logit contributions | |
Deal | +0.027 |
Not | +0.023 |
Okay | +0.023 |
Jake | +0.022 |
Thanks | +0.022 |
Neg logit contributions | |
le | -0.022 |
ling | -0.021 |
behind | -0.021 |
inside | -0.020 |
pe | -0.019 |
Token | valuable |
Token ID | 8119 |
Feature activation | -9.89e-02 |
Pos logit contributions | |
Deal | +0.060 |
Not | +0.051 |
Okay | +0.050 |
Jake | +0.049 |
Thanks | +0.049 |
Neg logit contributions | |
le | -0.035 |
ling | -0.033 |
behind | -0.032 |
inside | -0.032 |
pe | -0.031 |
Token | lesson |
Token ID | 11483 |
Feature activation | -6.60e-02 |
Pos logit contributions | |
le | +0.197 |
ling | +0.189 |
behind | +0.188 |
inside | +0.176 |
pe | +0.176 |
Neg logit contributions | |
Deal | -0.171 |
Not | -0.156 |
Thanks | -0.142 |
Okay | -0.137 |
Jake | -0.137 |
Token | today |
Token ID | 1909 |
Feature activation | -2.41e-02 |
Pos logit contributions | |
le | +0.089 |
ling | +0.087 |
pe | +0.078 |
les | +0.075 |
behind | +0.074 |
Neg logit contributions | |
Deal | -0.158 |
Not | -0.147 |
They | -0.138 |
Jake | -0.138 |
Okay | -0.137 |
Token | . |
Token ID | 13 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
ling | +0.027 |
le | +0.027 |
pe | +0.023 |
behind | +0.023 |
away | +0.022 |
Neg logit contributions | |
Deal | -0.061 |
Not | -0.058 |
They | -0.055 |
Jake | -0.054 |
Okay | -0.054 |
Token | She |
Token ID | 1375 |
Feature activation | +8.17e-02 |
Pos logit contributions | |
Deal | +0.355 |
Thanks | +0.281 |
Jake | +0.268 |
Enough | +0.264 |
Tyler | +0.260 |
Neg logit contributions | |
inside | -0.461 |
behind | -0.447 |
pe | -0.422 |
. | -0.414 |
up | -0.406 |
Token | said |
Token ID | 531 |
Feature activation | +5.92e-02 |
Pos logit contributions | |
Deal | +0.180 |
Not | +0.156 |
Okay | +0.151 |
Jake | +0.150 |
Thanks | +0.148 |
Neg logit contributions | |
le | -0.134 |
ling | -0.130 |
behind | -0.126 |
inside | -0.122 |
away | -0.119 |
Token | they |
Token ID | 484 |
Feature activation | +3.04e-02 |
Pos logit contributions | |
Deal | +0.101 |
Not | +0.085 |
Okay | +0.080 |
Jake | +0.078 |
Thanks | +0.077 |
Neg logit contributions | |
le | -0.131 |
ling | -0.124 |
behind | -0.122 |
inside | -0.119 |
away | -0.116 |
Token | should |
Token ID | 815 |
Feature activation | +2.86e-02 |
Pos logit contributions | |
Deal | +0.064 |
Not | +0.053 |
Okay | +0.052 |
Jake | +0.051 |
Thanks | +0.050 |
Neg logit contributions | |
le | -0.056 |
ling | -0.053 |
behind | -0.052 |
inside | -0.052 |
pe | -0.050 |
Token | always |
Token ID | 1464 |
Feature activation | +6.21e-03 |
Pos logit contributions | |
Deal | +0.059 |
Not | +0.048 |
Jake | +0.047 |
Okay | +0.047 |
Thanks | +0.045 |
Neg logit contributions | |
le | -0.053 |
inside | -0.050 |
ling | -0.050 |
behind | -0.050 |
away | -0.049 |
Token | 's |
Token ID | 338 |
Feature activation | +3.43e-02 |
Pos logit contributions | |
Deal | +0.144 |
Not | +0.126 |
Okay | +0.123 |
Jake | +0.120 |
Thanks | +0.119 |
Neg logit contributions | |
le | -0.095 |
ling | -0.088 |
behind | -0.088 |
inside | -0.085 |
away | -0.081 |
Token | cool |
Token ID | 3608 |
Feature activation | -8.49e-02 |
Pos logit contributions | |
Deal | +0.075 |
Not | +0.064 |
Okay | +0.061 |
Jake | +0.061 |
Thanks | +0.060 |
Neg logit contributions | |
le | -0.059 |
behind | -0.056 |
ling | -0.056 |
inside | -0.056 |
pe | -0.054 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.93e-01 |
Pos logit contributions | |
le | +0.104 |
ling | +0.099 |
behind | +0.096 |
pe | +0.088 |
inside | +0.086 |
Neg logit contributions | |
Deal | -0.206 |
Not | -0.185 |
Okay | -0.180 |
They | -0.178 |
Tyler | -0.177 |
Token | Ben |
Token ID | 3932 |
Feature activation | -3.63e-03 |
Pos logit contributions | |
Deal | +0.363 |
Not | +0.268 |
Okay | +0.266 |
Jake | +0.260 |
Thanks | +0.254 |
Neg logit contributions | |
behind | -0.386 |
inside | -0.385 |
le | -0.383 |
away | -0.374 |
pe | -0.373 |
Token | says |
Token ID | 1139 |
Feature activation | +4.28e-02 |
Pos logit contributions | |
le | +0.005 |
ling | +0.005 |
behind | +0.005 |
inside | +0.005 |
pe | +0.005 |
Neg logit contributions | |
Deal | -0.009 |
Not | -0.008 |
Okay | -0.008 |
Jake | -0.007 |
Thanks | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | +2.27e-01 |
Pos logit contributions | |
Deal | +0.117 |
Not | +0.105 |
Okay | +0.105 |
Thanks | +0.102 |
Acc | +0.100 |
Neg logit contributions | |
le | -0.046 |
ling | -0.039 |
egg | -0.038 |
les | -0.038 |
behind | -0.038 |
Token | He |
Token ID | 679 |
Feature activation | +6.25e-02 |
Pos logit contributions | |
Deal | +0.451 |
David | +0.366 |
Enough | +0.361 |
liv | +0.360 |
Jake | +0.358 |
Neg logit contributions | |
inside | -0.388 |
on | -0.380 |
behind | -0.369 |
. | -0.369 |
in | -0.355 |
Token | takes |
Token ID | 2753 |
Feature activation | -2.39e-02 |
Pos logit contributions | |
Deal | +0.141 |
Not | +0.119 |
Okay | +0.117 |
Jake | +0.115 |
Thanks | +0.114 |
Neg logit contributions | |
le | -0.105 |
ling | -0.098 |
behind | -0.097 |
inside | -0.097 |
pe | -0.095 |
Token | the |
Token ID | 262 |
Feature activation | +1.83e-04 |
Pos logit contributions | |
ling | +0.044 |
le | +0.043 |
behind | +0.041 |
pe | +0.039 |
inside | +0.039 |
Neg logit contributions | |
Deal | -0.047 |
Not | -0.038 |
Jake | -0.038 |
Tyler | -0.037 |
Jake | -0.037 |
Token | bucket |
Token ID | 19236 |
Feature activation | -2.14e-01 |
Pos logit contributions | |
Deal | +0.000 |
Okay | +0.000 |
Not | +0.000 |
Something | +0.000 |
Jake | +0.000 |
Neg logit contributions | |
le | -0.000 |
ling | -0.000 |
behind | -0.000 |
les | -0.000 |
ck | -0.000 |
Token | from |
Token ID | 422 |
Feature activation | -2.34e-02 |
Pos logit contributions | |
le | +0.321 |
ling | +0.285 |
les | +0.260 |
lers | +0.255 |
aches | +0.250 |
Neg logit contributions | |
Deal | -0.434 |
They | -0.433 |
She | -0.414 |
Jake | -0.403 |
He | -0.401 |
Token | not |
Token ID | 407 |
Feature activation | +1.13e-02 |
Pos logit contributions | |
Deal | +0.091 |
Jake | +0.074 |
Okay | +0.072 |
Great | +0.072 |
Not | +0.071 |
Neg logit contributions | |
behind | -0.078 |
inside | -0.077 |
le | -0.076 |
pe | -0.074 |
away | -0.072 |
Token | find |
Token ID | 1064 |
Feature activation | -7.79e-02 |
Pos logit contributions | |
Deal | +0.024 |
Not | +0.021 |
Okay | +0.020 |
Something | +0.020 |
Jake | +0.019 |
Neg logit contributions | |
ling | -0.019 |
le | -0.018 |
les | -0.018 |
lers | -0.017 |
ls | -0.017 |
Token | his |
Token ID | 465 |
Feature activation | -3.33e-02 |
Pos logit contributions | |
ling | +0.155 |
le | +0.142 |
an | +0.135 |
les | +0.126 |
ls | +0.125 |
Neg logit contributions | |
Deal | -0.132 |
Jake | -0.127 |
Not | -0.123 |
Dad | -0.121 |
Daddy | -0.118 |
Token | famous |
Token ID | 5863 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
le | +0.067 |
ling | +0.062 |
behind | +0.059 |
les | +0.058 |
away | +0.056 |
Neg logit contributions | |
Deal | -0.064 |
Not | -0.054 |
Okay | -0.052 |
Something | -0.049 |
Jake | -0.049 |
Token | las |
Token ID | 39990 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
le | +0.285 |
ling | +0.279 |
behind | +0.275 |
away | +0.260 |
inside | +0.257 |
Neg logit contributions | |
Deal | -0.243 |
Not | -0.203 |
Jake | -0.190 |
Okay | -0.188 |
Jake | -0.184 |
Token | agna |
Token ID | 48669 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
le | +0.274 |
ling | +0.260 |
behind | +0.256 |
inside | +0.255 |
pe | +0.249 |
Neg logit contributions | |
Deal | -0.202 |
Not | -0.164 |
Okay | -0.159 |
Jake | -0.154 |
Thanks | -0.152 |
Token | . |
Token ID | 13 |
Feature activation | +1.92e-01 |
Pos logit contributions | |
ling | +0.367 |
le | +0.361 |
ck | +0.326 |
ls | +0.319 |
pe | +0.316 |
Neg logit contributions | |
Jake | -0.579 |
Deal | -0.578 |
Tyler | -0.556 |
Not | -0.549 |
Jason | -0.546 |
Token |
|
Token ID | 198 |
Feature activation | +1.62e-01 |
Pos logit contributions | |
Deal | +0.255 |
Thanks | +0.184 |
Jake | +0.181 |
Okay | +0.176 |
Tyler | +0.175 |
Neg logit contributions | |
inside | -0.439 |
behind | -0.428 |
pe | -0.424 |
le | -0.417 |
away | -0.402 |
Token |
|
Token ID | 198 |
Feature activation | +1.63e-01 |
Pos logit contributions | |
Deal | +0.291 |
Not | +0.227 |
Okay | +0.222 |
Thanks | +0.213 |
Sorry | +0.211 |
Neg logit contributions | |
le | -0.335 |
inside | -0.327 |
behind | -0.324 |
ling | -0.322 |
pe | -0.314 |
Token | The |
Token ID | 464 |
Feature activation | +7.81e-02 |
Pos logit contributions | |
Deal | +0.290 |
Acc | +0.224 |
Sorry | +0.221 |
Okay | +0.217 |
Not | +0.216 |
Neg logit contributions | |
le | -0.333 |
ling | -0.329 |
inside | -0.313 |
pe | -0.313 |
b | -0.312 |
Token | chef |
Token ID | 21221 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
Deal | +0.159 |
Not | +0.132 |
Okay | +0.123 |
Jake | +0.123 |
Thanks | +0.122 |
Neg logit contributions | |
le | -0.151 |
behind | -0.140 |
ling | -0.138 |
inside | -0.136 |
away | -0.132 |
Token | finally |
Token ID | 3443 |
Feature activation | -1.61e-02 |
Pos logit contributions | |
Jake | +0.001 |
Deal | +0.001 |
Not | +0.001 |
Thanks | +0.001 |
Johnny | +0.001 |
Neg logit contributions | |
ling | -0.001 |
le | -0.001 |
behind | -0.001 |
egg | -0.001 |
lers | -0.001 |
Token | reached |
Token ID | 4251 |
Feature activation | -5.27e-02 |
Pos logit contributions | |
le | +0.029 |
ling | +0.028 |
behind | +0.027 |
egg | +0.027 |
lers | +0.026 |
Neg logit contributions | |
Thanks | -0.027 |
Deal | -0.027 |
Not | -0.026 |
Jake | -0.026 |
David | -0.026 |
Token | his |
Token ID | 465 |
Feature activation | -2.33e-02 |
Pos logit contributions | |
le | +0.097 |
ling | +0.093 |
behind | +0.084 |
les | +0.083 |
pe | +0.079 |
Neg logit contributions | |
Deal | -0.106 |
Not | -0.092 |
Okay | -0.088 |
Jake | -0.088 |
Tyler | -0.087 |
Token | sparkling |
Token ID | 39072 |
Feature activation | -1.71e-01 |
Pos logit contributions | |
le | +0.044 |
ling | +0.041 |
les | +0.038 |
behind | +0.036 |
ck | +0.035 |
Neg logit contributions | |
Deal | -0.049 |
Not | -0.040 |
Okay | -0.040 |
Jake | -0.039 |
Dad | -0.038 |
Token | red |
Token ID | 2266 |
Feature activation | -1.64e-01 |
Pos logit contributions | |
le | +0.275 |
ck | +0.252 |
behind | +0.247 |
ling | +0.245 |
lling | +0.232 |
Neg logit contributions | |
Deal | -0.311 |
omo | -0.277 |
Jake | -0.276 |
Daddy | -0.273 |
addy | -0.273 |
Token | island |
Token ID | 7022 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
le | +0.250 |
ling | +0.247 |
behind | +0.238 |
ck | +0.230 |
aches | +0.215 |
Neg logit contributions | |
Deal | -0.333 |
Not | -0.299 |
Jake | -0.292 |
They | -0.290 |
Johnny | -0.286 |
Token | . |
Token ID | 13 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
le | +0.387 |
les | +0.336 |
ck | +0.333 |
ling | +0.332 |
ls | +0.303 |
Neg logit contributions | |
They | -0.495 |
Deal | -0.494 |
Jake | -0.492 |
He | -0.482 |
Thanks | -0.478 |
Token | He |
Token ID | 679 |
Feature activation | +3.22e-02 |
Pos logit contributions | |
Deal | +0.312 |
Luke | +0.246 |
Jake | +0.229 |
Emily | +0.221 |
Okay | +0.211 |
Neg logit contributions | |
inside | -0.431 |
pe | -0.425 |
behind | -0.403 |
on | -0.399 |
up | -0.393 |
Token | polished |
Token ID | 23895 |
Feature activation | -8.28e-02 |
Pos logit contributions | |
Deal | +0.054 |
Jake | +0.051 |
Thanks | +0.050 |
Not | +0.049 |
Johnny | +0.047 |
Neg logit contributions | |
le | -0.061 |
ling | -0.058 |
behind | -0.057 |
away | -0.054 |
inside | -0.053 |
Token | and |
Token ID | 290 |
Feature activation | +4.72e-02 |
Pos logit contributions | |
le | +0.152 |
ling | +0.146 |
behind | +0.136 |
egg | +0.134 |
an | +0.131 |
Neg logit contributions | |
Deal | -0.147 |
Jake | -0.136 |
Tyler | -0.135 |
Johnny | -0.130 |
Jake | -0.130 |
Token | polished |
Token ID | 23895 |
Feature activation | -8.40e-02 |
Pos logit contributions | |
Jake | +0.083 |
Thanks | +0.082 |
Deal | +0.081 |
Johnny | +0.080 |
Jake | +0.079 |
Neg logit contributions | |
le | -0.084 |
ling | -0.080 |
egg | -0.076 |
behind | -0.075 |
an | -0.074 |
Token | oven |
Token ID | 14361 |
Feature activation | -1.86e-01 |
Pos logit contributions | |
le | +0.006 |
ling | +0.006 |
behind | +0.006 |
inside | +0.006 |
ck | +0.005 |
Neg logit contributions | |
Deal | -0.008 |
Not | -0.006 |
Okay | -0.006 |
Jake | -0.006 |
Something | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | +4.40e-02 |
Pos logit contributions | |
le | +0.259 |
ling | +0.243 |
ck | +0.215 |
b | +0.212 |
les | +0.209 |
Neg logit contributions | |
Deal | -0.439 |
Not | -0.398 |
They | -0.386 |
Jake | -0.384 |
Jake | -0.382 |
Token | taken |
Token ID | 2077 |
Feature activation | -4.55e-02 |
Pos logit contributions | |
Deal | +0.085 |
Not | +0.077 |
Jake | +0.075 |
Okay | +0.074 |
Thanks | +0.073 |
Neg logit contributions | |
le | -0.080 |
ling | -0.076 |
les | -0.069 |
an | -0.068 |
behind | -0.068 |
Token | the |
Token ID | 262 |
Feature activation | +6.84e-04 |
Pos logit contributions | |
ling | +0.080 |
le | +0.078 |
behind | +0.077 |
an | +0.071 |
ring | +0.070 |
Neg logit contributions | |
Deal | -0.090 |
Not | -0.080 |
Thanks | -0.077 |
Jake | -0.076 |
Okay | -0.075 |
Token | las |
Token ID | 39990 |
Feature activation | -1.12e-01 |
Pos logit contributions | |
Deal | +0.001 |
Not | +0.001 |
Jake | +0.001 |
Thanks | +0.001 |
Okay | +0.001 |
Neg logit contributions | |
le | -0.001 |
behind | -0.001 |
ling | -0.001 |
inside | -0.001 |
away | -0.001 |
Token | agna |
Token ID | 48669 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
le | +0.256 |
ling | +0.245 |
behind | +0.241 |
inside | +0.240 |
pe | +0.234 |
Neg logit contributions | |
Deal | -0.181 |
Not | -0.144 |
Okay | -0.139 |
Jake | -0.135 |
Thanks | -0.133 |
Token | . |
Token ID | 13 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
le | +0.387 |
ck | +0.342 |
ling | +0.335 |
b | +0.327 |
les | +0.317 |
Neg logit contributions | |
Deal | -0.510 |
They | -0.497 |
Jake | -0.490 |
Dad | -0.480 |
Not | -0.477 |
Token |
|
Token ID | 198 |
Feature activation | +1.65e-01 |
Pos logit contributions | |
Deal | +0.313 |
Thanks | +0.254 |
Okay | +0.244 |
Jake | +0.241 |
omo | +0.241 |
Neg logit contributions | |
inside | -0.377 |
behind | -0.363 |
pe | -0.351 |
le | -0.347 |
up | -0.333 |
Token |
|
Token ID | 198 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
Deal | +0.319 |
Not | +0.252 |
Okay | +0.248 |
Sorry | +0.243 |
Thanks | +0.236 |
Neg logit contributions | |
inside | -0.311 |
le | -0.311 |
behind | -0.305 |
ling | -0.303 |
pe | -0.297 |
Token | But |
Token ID | 1537 |
Feature activation | +9.13e-02 |
Pos logit contributions | |
Deal | +0.326 |
Acc | +0.266 |
Sorry | +0.256 |
Okay | +0.253 |
Not | +0.250 |
Neg logit contributions | |
ling | -0.294 |
le | -0.292 |
pe | -0.280 |
b | -0.277 |
inside | -0.271 |
Token | the |
Token ID | 262 |
Feature activation | +4.79e-02 |
Pos logit contributions | |
Deal | +0.184 |
Not | +0.166 |
Okay | +0.157 |
Tyler | +0.156 |
Jake | +0.155 |
Neg logit contributions | |
le | -0.165 |
ling | -0.157 |
les | -0.153 |
an | -0.139 |
behind | -0.139 |
Token | want |
Token ID | 765 |
Feature activation | -3.70e-02 |
Pos logit contributions | |
Deal | +0.006 |
Not | +0.005 |
Okay | +0.005 |
They | +0.004 |
Something | +0.004 |
Neg logit contributions | |
ling | -0.011 |
le | -0.011 |
les | -0.010 |
ls | -0.010 |
behind | -0.010 |
Token | anyone |
Token ID | 2687 |
Feature activation | -7.87e-02 |
Pos logit contributions | |
ling | +0.080 |
le | +0.078 |
les | +0.070 |
an | +0.069 |
b | +0.065 |
Neg logit contributions | |
Deal | -0.065 |
Not | -0.060 |
Okay | -0.055 |
Jake | -0.054 |
They | -0.053 |
Token | to |
Token ID | 284 |
Feature activation | -3.41e-02 |
Pos logit contributions | |
ling | +0.172 |
le | +0.164 |
les | +0.161 |
an | +0.158 |
lling | +0.157 |
Neg logit contributions | |
Not | -0.101 |
Okay | -0.090 |
I | -0.087 |
We | -0.084 |
Deal | -0.084 |
Token | steal |
Token ID | 8711 |
Feature activation | -1.53e-01 |
Pos logit contributions | |
ling | +0.075 |
les | +0.069 |
led | +0.069 |
le | +0.067 |
lers | +0.066 |
Neg logit contributions | |
Deal | -0.054 |
Not | -0.051 |
Okay | -0.050 |
She | -0.048 |
They | -0.046 |
Token | my |
Token ID | 616 |
Feature activation | -7.20e-02 |
Pos logit contributions | |
ling | +0.271 |
le | +0.244 |
an | +0.236 |
b | +0.235 |
les | +0.229 |
Neg logit contributions | |
Deal | -0.277 |
They | -0.250 |
Okay | -0.243 |
Dad | -0.240 |
Not | -0.238 |
Token | treasure |
Token ID | 14068 |
Feature activation | -2.66e-01 |
Pos logit contributions | |
le | +0.120 |
ling | +0.113 |
les | +0.098 |
lling | +0.093 |
ck | +0.090 |
Neg logit contributions | |
Deal | -0.156 |
Okay | -0.126 |
Not | -0.126 |
Something | -0.123 |
ossible | -0.122 |
Token | ." |
Token ID | 526 |
Feature activation | +1.86e-01 |
Pos logit contributions | |
ling | +0.482 |
le | +0.430 |
ck | +0.415 |
b | +0.413 |
lling | +0.413 |
Neg logit contributions | |
Deal | -0.421 |
They | -0.398 |
Not | -0.379 |
She | -0.371 |
Max | -0.370 |
Token |
|
Token ID | 198 |
Feature activation | +1.72e-01 |
Pos logit contributions | |
Deal | +0.344 |
Thanks | +0.267 |
Not | +0.266 |
Okay | +0.264 |
Jake | +0.260 |
Neg logit contributions | |
le | -0.377 |
behind | -0.370 |
inside | -0.370 |
pe | -0.360 |
ling | -0.355 |
Token |
|
Token ID | 198 |
Feature activation | +1.74e-01 |
Pos logit contributions | |
Deal | +0.303 |
Not | +0.239 |
Okay | +0.231 |
Thanks | +0.224 |
Jake | +0.215 |
Neg logit contributions | |
le | -0.371 |
behind | -0.356 |
inside | -0.355 |
ling | -0.355 |
pe | -0.345 |
Token | The |
Token ID | 464 |
Feature activation | +8.72e-02 |
Pos logit contributions | |
Deal | +0.328 |
Not | +0.256 |
Okay | +0.248 |
Acc | +0.231 |
Tyler | +0.226 |
Neg logit contributions | |
le | -0.363 |
ling | -0.350 |
away | -0.343 |
pe | -0.343 |
behind | -0.339 |
Token | little |
Token ID | 1310 |
Feature activation | -4.70e-03 |
Pos logit contributions | |
Deal | +0.156 |
Not | +0.124 |
Thanks | +0.116 |
Okay | +0.116 |
Great | +0.113 |
Neg logit contributions | |
le | -0.179 |
ling | -0.165 |
lers | -0.158 |
behind | -0.158 |
away | -0.156 |
Token | a |
Token ID | 257 |
Feature activation | -2.37e-02 |
Pos logit contributions | |
ling | +0.197 |
le | +0.180 |
an | +0.165 |
b | +0.165 |
les | +0.163 |
Neg logit contributions | |
Deal | -0.130 |
Okay | -0.111 |
Dad | -0.108 |
Not | -0.107 |
Something | -0.107 |
Token | way |
Token ID | 835 |
Feature activation | -1.12e-01 |
Pos logit contributions | |
le | +0.045 |
ling | +0.041 |
les | +0.040 |
led | +0.039 |
behind | +0.037 |
Neg logit contributions | |
Deal | -0.047 |
Okay | -0.045 |
Thanks | -0.042 |
Something | -0.042 |
Not | -0.041 |
Token | to |
Token ID | 284 |
Feature activation | -2.99e-02 |
Pos logit contributions | |
le | +0.180 |
ling | +0.166 |
les | +0.153 |
pe | +0.151 |
behind | +0.148 |
Neg logit contributions | |
Deal | -0.249 |
Not | -0.221 |
Okay | -0.216 |
They | -0.206 |
Jake | -0.205 |
Token | explore |
Token ID | 7301 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
ling | +0.064 |
les | +0.059 |
le | +0.059 |
lers | +0.059 |
ls | +0.057 |
Neg logit contributions | |
Deal | -0.050 |
Not | -0.044 |
Something | -0.044 |
Okay | -0.042 |
Thanks | -0.041 |
Token | my |
Token ID | 616 |
Feature activation | -5.79e-02 |
Pos logit contributions | |
ling | +0.231 |
le | +0.226 |
an | +0.222 |
b | +0.215 |
ring | +0.213 |
Neg logit contributions | |
Deal | -0.265 |
Not | -0.242 |
Jake | -0.235 |
Jake | -0.235 |
They | -0.233 |
Token | island |
Token ID | 7022 |
Feature activation | -2.66e-01 |
Pos logit contributions | |
le | +0.105 |
ling | +0.089 |
les | +0.088 |
ier | +0.079 |
b | +0.078 |
Neg logit contributions | |
Deal | -0.117 |
Dad | -0.102 |
Okay | -0.100 |
Thanks | -0.100 |
ossible | -0.099 |
Token | - |
Token ID | 532 |
Feature activation | +5.48e-02 |
Pos logit contributions | |
le | +0.396 |
ling | +0.356 |
les | +0.350 |
ck | +0.337 |
b | +0.314 |
Neg logit contributions | |
Deal | -0.533 |
They | -0.504 |
Not | -0.495 |
Okay | -0.472 |
She | -0.469 |
Token | even |
Token ID | 772 |
Feature activation | -5.61e-02 |
Pos logit contributions | |
Deal | +0.109 |
Not | +0.100 |
Okay | +0.096 |
Something | +0.090 |
Thanks | +0.087 |
Neg logit contributions | |
le | -0.106 |
ling | -0.102 |
les | -0.095 |
lers | -0.088 |
ls | -0.087 |
Token | with |
Token ID | 351 |
Feature activation | -1.17e-02 |
Pos logit contributions | |
le | +0.106 |
ling | +0.103 |
an | +0.100 |
les | +0.099 |
lers | +0.089 |
Neg logit contributions | |
Deal | -0.106 |
Not | -0.092 |
Okay | -0.091 |
How | -0.084 |
Something | -0.084 |
Token | the |
Token ID | 262 |
Feature activation | +1.01e-02 |
Pos logit contributions | |
le | +0.023 |
ling | +0.023 |
les | +0.020 |
behind | +0.020 |
an | +0.020 |
Neg logit contributions | |
Deal | -0.022 |
Not | -0.020 |
Okay | -0.018 |
Something | -0.018 |
Thanks | -0.017 |
Token | storm |
Token ID | 6388 |
Feature activation | -1.73e-01 |
Pos logit contributions | |
Deal | +0.022 |
Okay | +0.019 |
Thanks | +0.018 |
Not | +0.018 |
Something | +0.017 |
Neg logit contributions | |
le | -0.018 |
ling | -0.015 |
les | -0.015 |
behind | -0.014 |
ls | -0.013 |
Token | your |
Token ID | 534 |
Feature activation | -6.24e-02 |
Pos logit contributions | |
le | +0.121 |
ling | +0.115 |
behind | +0.112 |
inside | +0.109 |
pe | +0.106 |
Neg logit contributions | |
Deal | -0.167 |
Not | -0.143 |
Okay | -0.139 |
Jake | -0.137 |
Thanks | -0.135 |
Token | treasure |
Token ID | 14068 |
Feature activation | -2.52e-01 |
Pos logit contributions | |
le | +0.111 |
ling | +0.105 |
behind | +0.093 |
les | +0.092 |
inside | +0.088 |
Neg logit contributions | |
Deal | -0.143 |
Not | -0.122 |
Okay | -0.119 |
Something | -0.111 |
Thanks | -0.108 |
Token | or |
Token ID | 393 |
Feature activation | -3.73e-02 |
Pos logit contributions | |
ling | +0.408 |
le | +0.399 |
lling | +0.339 |
les | +0.320 |
ck | +0.317 |
Neg logit contributions | |
Deal | -0.540 |
Not | -0.485 |
They | -0.464 |
Okay | -0.455 |
Something | -0.439 |
Token | walk |
Token ID | 2513 |
Feature activation | -1.43e-01 |
Pos logit contributions | |
le | +0.078 |
ling | +0.075 |
behind | +0.073 |
inside | +0.072 |
pe | +0.070 |
Neg logit contributions | |
Deal | -0.069 |
Not | -0.057 |
Okay | -0.055 |
Jake | -0.053 |
Thanks | -0.053 |
Token | the |
Token ID | 262 |
Feature activation | -2.00e-02 |
Pos logit contributions | |
le | +0.195 |
ling | +0.190 |
egg | +0.185 |
pe | +0.174 |
les | +0.172 |
Neg logit contributions | |
They | -0.286 |
Deal | -0.286 |
Not | -0.269 |
Jake | -0.262 |
Max | -0.260 |
Token | plank |
Token ID | 39599 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
le | +0.033 |
ling | +0.030 |
les | +0.026 |
led | +0.025 |
behind | +0.024 |
Neg logit contributions | |
Deal | -0.048 |
Not | -0.039 |
Luke | -0.039 |
Jake | -0.039 |
Max | -0.039 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
egg | +0.395 |
ling | +0.391 |
pe | +0.390 |
le | +0.374 |
lers | +0.344 |
Neg logit contributions | |
Deal | -0.481 |
They | -0.451 |
Not | -0.447 |
Jake | -0.431 |
ossible | -0.416 |
Token |
|
Token ID | 198 |
Feature activation | +1.71e-01 |
Pos logit contributions | |
Deal | +0.308 |
Okay | +0.236 |
Not | +0.234 |
Jake | +0.224 |
Thanks | +0.224 |
Neg logit contributions | |
le | -0.389 |
behind | -0.380 |
inside | -0.374 |
ling | -0.371 |
away | -0.370 |
Token |
|
Token ID | 198 |
Feature activation | +1.72e-01 |
Pos logit contributions | |
Deal | +0.337 |
Not | +0.265 |
Okay | +0.261 |
Acc | +0.248 |
Sorry | +0.243 |
Neg logit contributions | |
le | -0.345 |
ling | -0.328 |
inside | -0.326 |
behind | -0.324 |
pe | -0.311 |
Token | L |
Token ID | 43 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
Deal | +0.325 |
Acc | +0.257 |
Okay | +0.251 |
Not | +0.250 |
Kelly | +0.226 |
Neg logit contributions | |
le | -0.370 |
ling | -0.359 |
away | -0.342 |
pe | -0.329 |
behind | -0.325 |
Token | ily |
Token ID | 813 |
Feature activation | -4.93e-03 |
Pos logit contributions | |
Deal | +0.096 |
Not | +0.079 |
Jake | +0.076 |
Okay | +0.075 |
Thanks | +0.074 |
Neg logit contributions | |
le | -0.093 |
ling | -0.091 |
behind | -0.090 |
inside | -0.088 |
pe | -0.087 |
Token | had |
Token ID | 550 |
Feature activation | +7.30e-03 |
Pos logit contributions | |
le | +0.009 |
ling | +0.008 |
egg | +0.008 |
away | +0.008 |
crown | +0.008 |
Neg logit contributions | |
Not | -0.006 |
They | -0.005 |
Thanks | -0.005 |
Jake | -0.005 |
Jake | -0.005 |
Token | come |
Token ID | 1282 |
Feature activation | -9.00e-02 |
Pos logit contributions | |
Not | +0.011 |
Thanks | +0.011 |
Their | +0.011 |
David | +0.011 |
They | +0.011 |
Neg logit contributions | |
le | -0.016 |
ling | -0.015 |
les | -0.013 |
ie | -0.013 |
ls | -0.012 |
Token | to |
Token ID | 284 |
Feature activation | -3.32e-02 |
Pos logit contributions | |
le | +0.167 |
les | +0.154 |
egg | +0.148 |
ling | +0.142 |
aches | +0.139 |
Neg logit contributions | |
Deal | -0.157 |
They | -0.154 |
Not | -0.147 |
Okay | -0.145 |
Johnny | -0.137 |
Token | save |
Token ID | 3613 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
ling | +0.079 |
le | +0.075 |
les | +0.074 |
lers | +0.072 |
ls | +0.071 |
Neg logit contributions | |
Deal | -0.049 |
Not | -0.044 |
Okay | -0.043 |
Something | -0.041 |
Thanks | -0.041 |
Token | the |
Token ID | 262 |
Feature activation | -1.06e-02 |
Pos logit contributions | |
ling | +0.249 |
le | +0.243 |
ring | +0.227 |
an | +0.226 |
egg | +0.225 |
Neg logit contributions | |
They | -0.203 |
Sam | -0.199 |
Jake | -0.196 |
Max | -0.192 |
Kelly | -0.188 |
Token | island |
Token ID | 7022 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
le | +0.017 |
les | +0.014 |
ling | +0.014 |
aches | +0.014 |
lers | +0.014 |
Neg logit contributions | |
Deal | -0.018 |
Dad | -0.018 |
Thanks | -0.018 |
They | -0.018 |
Not | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | +1.86e-01 |
Pos logit contributions | |
le | +0.422 |
les | +0.366 |
ling | +0.357 |
egg | +0.335 |
ck | +0.332 |
Neg logit contributions | |
They | -0.524 |
Not | -0.451 |
Their | -0.426 |
He | -0.423 |
Dad | -0.416 |
Token | The |
Token ID | 383 |
Feature activation | +7.95e-02 |
Pos logit contributions | |
Deal | +0.306 |
Okay | +0.222 |
Thanks | +0.214 |
Acc | +0.197 |
Brother | +0.194 |
Neg logit contributions | |
behind | -0.436 |
inside | -0.415 |
around | -0.394 |
pe | -0.394 |
tree | -0.390 |
Token | people |
Token ID | 661 |
Feature activation | -1.02e-01 |
Pos logit contributions | |
Deal | +0.157 |
Not | +0.141 |
Thanks | +0.141 |
They | +0.139 |
ossible | +0.135 |
Neg logit contributions | |
le | -0.140 |
behind | -0.115 |
ling | -0.111 |
les | -0.107 |
inside | -0.106 |
Token | on |
Token ID | 319 |
Feature activation | -2.96e-02 |
Pos logit contributions | |
le | +0.201 |
egg | +0.195 |
lands | +0.190 |
an | +0.189 |
potion | +0.183 |
Neg logit contributions | |
They | -0.134 |
Their | -0.133 |
Not | -0.121 |
Thanks | -0.120 |
Yeah | -0.115 |
Token | the |
Token ID | 262 |
Feature activation | +3.99e-02 |
Pos logit contributions | |
le | +0.048 |
ling | +0.046 |
inside | +0.043 |
behind | +0.042 |
pe | +0.041 |
Neg logit contributions | |
Deal | -0.066 |
Not | -0.059 |
Jake | -0.056 |
Okay | -0.055 |
Thanks | -0.054 |
Token | could |
Token ID | 714 |
Feature activation | +1.45e-02 |
Pos logit contributions | |
le | +0.049 |
ling | +0.049 |
behind | +0.046 |
egg | +0.045 |
inside | +0.044 |
Neg logit contributions | |
Deal | -0.038 |
Not | -0.036 |
Thanks | -0.032 |
Jake | -0.032 |
Okay | -0.031 |
Token | not |
Token ID | 407 |
Feature activation | -4.25e-03 |
Pos logit contributions | |
Deal | +0.029 |
Not | +0.026 |
Okay | +0.024 |
Thanks | +0.023 |
Jake | +0.023 |
Neg logit contributions | |
le | -0.027 |
ling | -0.026 |
behind | -0.025 |
les | -0.024 |
inside | -0.023 |
Token | repair |
Token ID | 9185 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
le | +0.007 |
ling | +0.007 |
behind | +0.006 |
les | +0.006 |
ls | +0.006 |
Neg logit contributions | |
Deal | -0.009 |
Not | -0.009 |
Okay | -0.008 |
They | -0.008 |
Something | -0.008 |
Token | the |
Token ID | 262 |
Feature activation | -2.18e-02 |
Pos logit contributions | |
le | +0.264 |
ling | +0.256 |
an | +0.239 |
fish | +0.232 |
worm | +0.232 |
Neg logit contributions | |
Deal | -0.269 |
Tyler | -0.252 |
Not | -0.251 |
Jake | -0.246 |
Jake | -0.242 |
Token | o |
Token ID | 267 |
Feature activation | -1.65e-01 |
Pos logit contributions | |
le | +0.036 |
ling | +0.030 |
les | +0.029 |
led | +0.029 |
ck | +0.028 |
Neg logit contributions | |
Deal | -0.043 |
Not | -0.038 |
They | -0.038 |
Okay | -0.037 |
Good | -0.037 |
Token | asis |
Token ID | 17765 |
Feature activation | -2.62e-01 |
Pos logit contributions | |
le | +0.175 |
behind | +0.161 |
inside | +0.157 |
ling | +0.153 |
away | +0.152 |
Neg logit contributions | |
Deal | -0.450 |
Not | -0.406 |
Okay | -0.386 |
Jake | -0.372 |
omo | -0.371 |
Token | . |
Token ID | 13 |
Feature activation | +1.88e-01 |
Pos logit contributions | |
le | +0.339 |
ling | +0.290 |
ck | +0.277 |
egg | +0.273 |
pe | +0.269 |
Neg logit contributions | |
Deal | -0.600 |
Not | -0.594 |
They | -0.564 |
Jake | -0.557 |
Tyler | -0.549 |
Token | He |
Token ID | 679 |
Feature activation | +2.55e-02 |
Pos logit contributions | |
Deal | +0.368 |
Jake | +0.278 |
Luke | +0.278 |
Thanks | +0.276 |
Okay | +0.270 |
Neg logit contributions | |
inside | -0.351 |
behind | -0.334 |
up | -0.330 |
pe | -0.329 |
le | -0.325 |
Token | was |
Token ID | 373 |
Feature activation | +1.28e-02 |
Pos logit contributions | |
Deal | +0.038 |
Jake | +0.037 |
Not | +0.036 |
Brother | +0.035 |
Thanks | +0.034 |
Neg logit contributions | |
le | -0.050 |
behind | -0.049 |
ling | -0.047 |
inside | -0.046 |
ls | -0.046 |
Token | so |
Token ID | 523 |
Feature activation | +5.70e-02 |
Pos logit contributions | |
Deal | +0.028 |
Not | +0.025 |
Okay | +0.024 |
Thanks | +0.023 |
Jake | +0.022 |
Neg logit contributions | |
le | -0.023 |
ling | -0.020 |
les | -0.020 |
behind | -0.019 |
ls | -0.019 |
Token | sad |
Token ID | 6507 |
Feature activation | -5.17e-02 |
Pos logit contributions | |
Deal | +0.131 |
Not | +0.114 |
Okay | +0.109 |
Thanks | +0.106 |
Jake | +0.105 |
Neg logit contributions | |
le | -0.097 |
behind | -0.087 |
ling | -0.086 |
inside | -0.084 |
les | -0.082 |
Token | keep |
Token ID | 1394 |
Feature activation | -4.62e-02 |
Pos logit contributions | |
le | +0.029 |
ling | +0.029 |
behind | +0.027 |
inside | +0.026 |
les | +0.025 |
Neg logit contributions | |
Deal | -0.033 |
Not | -0.029 |
Okay | -0.028 |
Jake | -0.027 |
Thanks | -0.027 |
Token | it |
Token ID | 340 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
le | +0.088 |
ling | +0.081 |
behind | +0.071 |
les | +0.070 |
an | +0.069 |
Neg logit contributions | |
Deal | -0.086 |
Not | -0.084 |
Tyler | -0.076 |
Ryan | -0.074 |
Johnny | -0.074 |
Token | safe |
Token ID | 3338 |
Feature activation | -1.24e-01 |
Pos logit contributions | |
le | +0.190 |
les | +0.184 |
ling | +0.179 |
lers | +0.162 |
pe | +0.159 |
Neg logit contributions | |
Deal | -0.248 |
Not | -0.218 |
Okay | -0.211 |
Thanks | -0.203 |
Johnny | -0.202 |
Token | on |
Token ID | 319 |
Feature activation | -6.53e-02 |
Pos logit contributions | |
le | +0.176 |
ling | +0.165 |
les | +0.164 |
pe | +0.158 |
ck | +0.156 |
Neg logit contributions | |
Deal | -0.269 |
They | -0.244 |
Not | -0.239 |
Okay | -0.236 |
She | -0.230 |
Token | my |
Token ID | 616 |
Feature activation | -4.98e-02 |
Pos logit contributions | |
le | +0.140 |
ling | +0.139 |
behind | +0.125 |
les | +0.123 |
inside | +0.122 |
Neg logit contributions | |
Deal | -0.108 |
Not | -0.098 |
Jake | -0.086 |
Okay | -0.084 |
Tyler | -0.083 |
Token | island |
Token ID | 7022 |
Feature activation | -2.61e-01 |
Pos logit contributions | |
le | +0.086 |
ling | +0.078 |
les | +0.070 |
inside | +0.070 |
behind | +0.068 |
Neg logit contributions | |
Deal | -0.105 |
Not | -0.092 |
Thanks | -0.089 |
ossible | -0.088 |
Jake | -0.086 |
Token | of |
Token ID | 286 |
Feature activation | -3.89e-02 |
Pos logit contributions | |
le | +0.393 |
ling | +0.374 |
les | +0.348 |
ck | +0.324 |
ls | +0.312 |
Neg logit contributions | |
Deal | -0.524 |
They | -0.498 |
Not | -0.494 |
Jake | -0.459 |
ossible | -0.453 |
Token | seas |
Token ID | 21547 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
le | +0.083 |
ling | +0.083 |
les | +0.070 |
led | +0.069 |
ier | +0.067 |
Neg logit contributions | |
Not | -0.059 |
Deal | -0.057 |
Dad | -0.052 |
Thanks | -0.052 |
Jake | -0.051 |
Token | hell |
Token ID | 12758 |
Feature activation | -1.89e-01 |
Pos logit contributions | |
inside | +0.303 |
behind | +0.301 |
pe | +0.297 |
ling | +0.283 |
egg | +0.283 |
Neg logit contributions | |
Deal | -0.454 |
Not | -0.450 |
ossible | -0.404 |
addy | -0.402 |
omo | -0.400 |
Token | s |
Token ID | 82 |
Feature activation | -3.70e-02 |
Pos logit contributions | |
le | +0.308 |
pe | +0.294 |
ling | +0.292 |
behind | +0.291 |
inside | +0.276 |
Neg logit contributions | |
Not | -0.343 |
Deal | -0.342 |
They | -0.323 |
She | -0.314 |
Sorry | -0.293 |
Token | ." |
Token ID | 526 |
Feature activation | +1.84e-01 |
Pos logit contributions | |
le | +0.061 |
ling | +0.057 |
ck | +0.053 |
les | +0.052 |
b | +0.049 |
Neg logit contributions | |
Deal | -0.069 |
They | -0.065 |
Not | -0.065 |
Have | -0.061 |
ossible | -0.060 |
Token | she |
Token ID | 673 |
Feature activation | +1.07e-03 |
Pos logit contributions | |
le | +0.059 |
ling | +0.056 |
behind | +0.056 |
inside | +0.055 |
pe | +0.054 |
Neg logit contributions | |
Deal | -0.047 |
Not | -0.038 |
Okay | -0.037 |
Jake | -0.036 |
Thanks | -0.035 |
Token | had |
Token ID | 550 |
Feature activation | +1.39e-02 |
Pos logit contributions | |
Deal | +0.002 |
Not | +0.002 |
Okay | +0.002 |
Jake | +0.002 |
Thanks | +0.002 |
Neg logit contributions | |
le | -0.002 |
ling | -0.002 |
behind | -0.002 |
pe | -0.002 |
inside | -0.002 |
Token | never |
Token ID | 1239 |
Feature activation | -4.29e-02 |
Pos logit contributions | |
Deal | +0.025 |
Not | +0.022 |
Okay | +0.021 |
Thanks | +0.021 |
Jake | +0.020 |
Neg logit contributions | |
le | -0.029 |
ling | -0.028 |
les | -0.026 |
behind | -0.026 |
inside | -0.025 |
Token | seen |
Token ID | 1775 |
Feature activation | -8.46e-02 |
Pos logit contributions | |
le | +0.087 |
ling | +0.087 |
les | +0.078 |
an | +0.075 |
egg | +0.075 |
Neg logit contributions | |
Deal | -0.069 |
Thanks | -0.066 |
Not | -0.064 |
Okay | -0.063 |
Something | -0.063 |
Token | the |
Token ID | 262 |
Feature activation | -1.81e-02 |
Pos logit contributions | |
ling | +0.185 |
le | +0.183 |
behind | +0.166 |
away | +0.163 |
inside | +0.163 |
Neg logit contributions | |
Deal | -0.138 |
Not | -0.118 |
Jake | -0.113 |
Okay | -0.112 |
Jake | -0.112 |
Token | pit |
Token ID | 6028 |
Feature activation | -2.60e-01 |
Pos logit contributions | |
le | +0.034 |
ling | +0.031 |
row | +0.029 |
les | +0.029 |
lers | +0.028 |
Neg logit contributions | |
Deal | -0.034 |
Okay | -0.031 |
Not | -0.030 |
Thanks | -0.030 |
Yeah | -0.029 |
Token | . |
Token ID | 13 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
le | +0.349 |
lers | +0.327 |
ling | +0.322 |
b | +0.321 |
ck | +0.319 |
Neg logit contributions | |
Not | -0.514 |
Deal | -0.508 |
She | -0.507 |
He | -0.498 |
They | -0.484 |
Token | She |
Token ID | 1375 |
Feature activation | +4.29e-02 |
Pos logit contributions | |
Deal | +0.342 |
omo | +0.304 |
Johnny | +0.304 |
addy | +0.299 |
Jake | +0.297 |
Neg logit contributions | |
behind | -0.348 |
inside | -0.343 |
. | -0.301 |
on | -0.300 |
pe | -0.297 |
Token | wished |
Token ID | 16555 |
Feature activation | -2.86e-02 |
Pos logit contributions | |
Deal | +0.076 |
Not | +0.066 |
Thanks | +0.062 |
Okay | +0.062 |
Jake | +0.062 |
Neg logit contributions | |
le | -0.086 |
ling | -0.085 |
behind | -0.079 |
inside | -0.078 |
les | -0.076 |
Token | she |
Token ID | 673 |
Feature activation | +3.17e-04 |
Pos logit contributions | |
le | +0.062 |
ling | +0.059 |
behind | +0.058 |
inside | +0.058 |
pe | +0.056 |
Neg logit contributions | |
Deal | -0.050 |
Not | -0.041 |
Okay | -0.039 |
Jake | -0.038 |
Thanks | -0.038 |
Token | had |
Token ID | 550 |
Feature activation | +1.34e-02 |
Pos logit contributions | |
Deal | +0.001 |
Not | +0.001 |
Okay | +0.000 |
Jake | +0.000 |
Thanks | +0.000 |
Neg logit contributions | |
le | -0.001 |
ling | -0.001 |
behind | -0.001 |
pe | -0.001 |
inside | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | -2.61e-03 |
Pos logit contributions | |
le | +0.032 |
ling | +0.030 |
behind | +0.029 |
les | +0.027 |
inside | +0.027 |
Neg logit contributions | |
Deal | -0.044 |
Not | -0.039 |
Okay | -0.038 |
Jake | -0.037 |
Tyler | -0.037 |
Token | place |
Token ID | 1295 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
le | +0.005 |
behind | +0.004 |
ling | +0.004 |
les | +0.004 |
inside | +0.004 |
Neg logit contributions | |
Deal | -0.005 |
Not | -0.005 |
Okay | -0.004 |
Thanks | -0.004 |
Winning | -0.004 |
Token | I |
Token ID | 314 |
Feature activation | -6.58e-03 |
Pos logit contributions | |
le | +0.246 |
ling | +0.234 |
pe | +0.214 |
les | +0.211 |
behind | +0.209 |
Neg logit contributions | |
Deal | -0.414 |
Not | -0.377 |
Okay | -0.369 |
Jake | -0.358 |
Jake | -0.357 |
Token | buried |
Token ID | 11694 |
Feature activation | -1.53e-01 |
Pos logit contributions | |
ling | +0.013 |
le | +0.013 |
behind | +0.012 |
egg | +0.012 |
inside | +0.012 |
Neg logit contributions | |
Deal | -0.013 |
Not | -0.010 |
Okay | -0.010 |
Thanks | -0.010 |
Something | -0.010 |
Token | my |
Token ID | 616 |
Feature activation | -7.46e-02 |
Pos logit contributions | |
ling | +0.184 |
le | +0.181 |
les | +0.164 |
an | +0.155 |
ck | +0.154 |
Neg logit contributions | |
Deal | -0.334 |
Jake | -0.304 |
Okay | -0.293 |
Jake | -0.291 |
Dad | -0.289 |
Token | treasure |
Token ID | 14068 |
Feature activation | -2.60e-01 |
Pos logit contributions | |
le | +0.128 |
ling | +0.118 |
les | +0.110 |
lling | +0.097 |
ck | +0.096 |
Neg logit contributions | |
Deal | -0.151 |
Okay | -0.128 |
Not | -0.128 |
Something | -0.126 |
Dad | -0.123 |
Token | ! |
Token ID | 0 |
Feature activation | +1.70e-01 |
Pos logit contributions | |
ling | +0.371 |
le | +0.362 |
ck | +0.317 |
les | +0.316 |
b | +0.289 |
Neg logit contributions | |
Deal | -0.558 |
Okay | -0.487 |
Not | -0.482 |
Jake | -0.463 |
Have | -0.454 |
Token | â |
Token ID | 22940 |
Feature activation | +3.56e-02 |
Pos logit contributions | |
Deal | +0.335 |
Luke | +0.292 |
Jake | +0.291 |
Enough | +0.288 |
Thanks | +0.284 |
Neg logit contributions | |
inside | -0.242 |
on | -0.224 |
behind | -0.217 |
pe | -0.211 |
in | -0.208 |
Token | € |
Token ID | 26391 |
Feature activation | +7.10e-02 |
Pos logit contributions | |
Deal | +0.051 |
Not | +0.039 |
Okay | +0.038 |
Jake | +0.036 |
Thanks | +0.034 |
Neg logit contributions | |
le | -0.090 |
ling | -0.086 |
behind | -0.085 |
inside | -0.084 |
away | -0.082 |
Token | |
Token ID | 220 |
Feature activation | +9.88e-02 |
Pos logit contributions | |
Deal | +0.146 |
Thanks | +0.111 |
Great | +0.105 |
Okay | +0.105 |
Not | +0.105 |
Neg logit contributions | |
inside | -0.129 |
behind | -0.125 |
le | -0.122 |
- | -0.120 |
ling | -0.120 |
Token |
|
Token ID | 198 |
Feature activation | +1.57e-01 |
Pos logit contributions | |
Deal | +0.168 |
Thanks | +0.133 |
Not | +0.129 |
Jake | +0.127 |
Okay | +0.125 |
Neg logit contributions | |
inside | -0.200 |
le | -0.196 |
behind | -0.196 |
ling | -0.191 |
up | -0.190 |
Token | you |
Token ID | 345 |
Feature activation | -3.21e-02 |
Pos logit contributions | |
Deal | +0.010 |
Not | +0.009 |
Okay | +0.009 |
Thanks | +0.008 |
Jake | +0.008 |
Neg logit contributions | |
le | -0.008 |
behind | -0.008 |
inside | -0.007 |
ling | -0.007 |
pe | -0.007 |
Token | ! |
Token ID | 0 |
Feature activation | +1.58e-01 |
Pos logit contributions | |
le | +0.030 |
behind | +0.027 |
inside | +0.027 |
ling | +0.026 |
away | +0.025 |
Neg logit contributions | |
Deal | -0.094 |
Not | -0.085 |
Thanks | -0.083 |
Jake | -0.083 |
Okay | -0.082 |
Token | Where |
Token ID | 6350 |
Feature activation | -6.72e-03 |
Pos logit contributions | |
Deal | +0.240 |
Jake | +0.218 |
David | +0.210 |
Luke | +0.206 |
Enough | +0.202 |
Neg logit contributions | |
inside | -0.296 |
behind | -0.274 |
in | -0.272 |
on | -0.271 |
pe | -0.268 |
Token | is |
Token ID | 318 |
Feature activation | -3.05e-02 |
Pos logit contributions | |
le | +0.014 |
ling | +0.014 |
behind | +0.013 |
inside | +0.013 |
away | +0.013 |
Neg logit contributions | |
Deal | -0.012 |
Not | -0.010 |
Okay | -0.010 |
Jake | -0.010 |
Something | -0.009 |
Token | my |
Token ID | 616 |
Feature activation | -4.34e-02 |
Pos logit contributions | |
le | +0.058 |
ling | +0.056 |
les | +0.051 |
behind | +0.051 |
inside | +0.049 |
Neg logit contributions | |
Deal | -0.061 |
Not | -0.054 |
Okay | -0.051 |
Jake | -0.049 |
Thanks | -0.049 |
Token | island |
Token ID | 7022 |
Feature activation | -2.60e-01 |
Pos logit contributions | |
le | +0.081 |
ling | +0.075 |
les | +0.071 |
lling | +0.063 |
led | +0.061 |
Neg logit contributions | |
Deal | -0.088 |
Dad | -0.078 |
Not | -0.078 |
Daddy | -0.076 |
Okay | -0.076 |
Token | ?" |
Token ID | 1701 |
Feature activation | +1.55e-01 |
Pos logit contributions | |
le | +0.529 |
ling | +0.501 |
les | +0.477 |
lers | +0.460 |
behind | +0.443 |
Neg logit contributions | |
Deal | -0.443 |
Not | -0.378 |
Okay | -0.366 |
Something | -0.332 |
ossible | -0.332 |
Token |
|
Token ID | 198 |
Feature activation | +1.67e-01 |
Pos logit contributions | |
Deal | +0.284 |
Jake | +0.214 |
Okay | +0.212 |
Not | +0.210 |
Tyler | +0.205 |
Neg logit contributions | |
behind | -0.316 |
inside | -0.315 |
le | -0.312 |
pe | -0.309 |
ling | -0.298 |
Token |
|
Token ID | 198 |
Feature activation | +1.70e-01 |
Pos logit contributions | |
Deal | +0.318 |
Not | +0.252 |
Okay | +0.244 |
Tyler | +0.230 |
Thanks | +0.229 |
Neg logit contributions | |
inside | -0.321 |
le | -0.317 |
behind | -0.316 |
pe | -0.310 |
ling | -0.303 |
Token | The |
Token ID | 464 |
Feature activation | +8.25e-02 |
Pos logit contributions | |
Deal | +0.311 |
Tyler | +0.237 |
Not | +0.236 |
Okay | +0.232 |
omo | +0.216 |
Neg logit contributions | |
pe | -0.339 |
le | -0.338 |
inside | -0.335 |
behind | -0.333 |
ling | -0.326 |
Token | fairy |
Token ID | 25607 |
Feature activation | -1.10e-01 |
Pos logit contributions | |
Deal | +0.185 |
Not | +0.153 |
Thanks | +0.151 |
Okay | +0.147 |
Emily | +0.145 |
Neg logit contributions | |
le | -0.139 |
ling | -0.127 |
les | -0.120 |
behind | -0.114 |
lers | -0.109 |
Token | he |
Token ID | 339 |
Feature activation | -4.26e-03 |
Pos logit contributions | |
Deal | +0.042 |
Not | +0.037 |
Okay | +0.035 |
Jake | +0.034 |
Thanks | +0.033 |
Neg logit contributions | |
le | -0.053 |
ling | -0.050 |
behind | -0.049 |
inside | -0.047 |
les | -0.047 |
Token | could |
Token ID | 714 |
Feature activation | +2.37e-02 |
Pos logit contributions | |
le | +0.008 |
ling | +0.008 |
behind | +0.008 |
inside | +0.008 |
away | +0.008 |
Neg logit contributions | |
Deal | -0.008 |
Not | -0.007 |
Okay | -0.007 |
Jake | -0.006 |
Thanks | -0.006 |
Token | help |
Token ID | 1037 |
Feature activation | -4.04e-02 |
Pos logit contributions | |
Deal | +0.041 |
Not | +0.036 |
Okay | +0.034 |
Thanks | +0.033 |
They | +0.032 |
Neg logit contributions | |
le | -0.050 |
ling | -0.050 |
les | -0.047 |
behind | -0.046 |
lers | -0.044 |
Token | fill |
Token ID | 6070 |
Feature activation | -1.55e-01 |
Pos logit contributions | |
le | +0.080 |
ling | +0.079 |
les | +0.077 |
ls | +0.075 |
an | +0.074 |
Neg logit contributions | |
Deal | -0.063 |
Not | -0.059 |
They | -0.057 |
Jake | -0.057 |
Johnny | -0.057 |
Token | the |
Token ID | 262 |
Feature activation | -2.36e-02 |
Pos logit contributions | |
le | +0.278 |
ling | +0.273 |
les | +0.246 |
behind | +0.239 |
an | +0.238 |
Neg logit contributions | |
Deal | -0.255 |
They | -0.243 |
Jake | -0.242 |
Jake | -0.236 |
Tyler | -0.229 |
Token | pit |
Token ID | 6028 |
Feature activation | -2.60e-01 |
Pos logit contributions | |
le | +0.045 |
ling | +0.036 |
les | +0.035 |
ck | +0.034 |
behind | +0.033 |
Neg logit contributions | |
Deal | -0.048 |
They | -0.044 |
Not | -0.043 |
Something | -0.042 |
Okay | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
le | +0.371 |
les | +0.345 |
ling | +0.313 |
ck | +0.296 |
n | +0.285 |
Neg logit contributions | |
They | -0.538 |
Deal | -0.517 |
She | -0.500 |
He | -0.499 |
Together | -0.471 |
Token | The |
Token ID | 383 |
Feature activation | +8.71e-02 |
Pos logit contributions | |
Deal | +0.294 |
Okay | +0.209 |
Jake | +0.203 |
Thanks | +0.196 |
David | +0.194 |
Neg logit contributions | |
behind | -0.458 |
inside | -0.455 |
pe | -0.453 |
away | -0.452 |
le | -0.446 |
Token | boy |
Token ID | 2933 |
Feature activation | -8.88e-02 |
Pos logit contributions | |
Deal | +0.140 |
Not | +0.120 |
Thanks | +0.114 |
Okay | +0.113 |
Jake | +0.112 |
Neg logit contributions | |
le | -0.198 |
ling | -0.178 |
lers | -0.171 |
away | -0.171 |
behind | -0.169 |
Token | said |
Token ID | 531 |
Feature activation | +3.51e-02 |
Pos logit contributions | |
le | +0.160 |
behind | +0.153 |
ling | +0.151 |
inside | +0.147 |
les | +0.142 |
Neg logit contributions | |
Deal | -0.172 |
Not | -0.153 |
Thanks | -0.147 |
Okay | -0.145 |
Jake | -0.145 |
Token | yes |
Token ID | 3763 |
Feature activation | -1.63e-02 |
Pos logit contributions | |
Deal | +0.055 |
Not | +0.049 |
Okay | +0.045 |
Thanks | +0.044 |
Jake | +0.041 |
Neg logit contributions | |
le | -0.083 |
ling | -0.078 |
les | -0.076 |
behind | -0.075 |
an | -0.072 |
Token | could |
Token ID | 714 |
Feature activation | +1.19e-02 |
Pos logit contributions | |
le | +0.328 |
ling | +0.318 |
behind | +0.310 |
egg | +0.293 |
pe | +0.288 |
Neg logit contributions | |
Deal | -0.355 |
Not | -0.348 |
Something | -0.323 |
Jake | -0.323 |
They | -0.320 |
Token | help |
Token ID | 1037 |
Feature activation | -4.83e-02 |
Pos logit contributions | |
Deal | +0.023 |
Not | +0.020 |
Okay | +0.019 |
Thanks | +0.019 |
Something | +0.018 |
Neg logit contributions | |
le | -0.022 |
ling | -0.022 |
behind | -0.021 |
les | -0.020 |
inside | -0.020 |
Token | him |
Token ID | 683 |
Feature activation | -7.26e-02 |
Pos logit contributions | |
ling | +0.079 |
le | +0.078 |
ls | +0.077 |
les | +0.075 |
behind | +0.072 |
Neg logit contributions | |
Deal | -0.088 |
Not | -0.087 |
Jake | -0.084 |
Johnny | -0.082 |
Tyler | -0.082 |
Token | build |
Token ID | 1382 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
le | +0.116 |
les | +0.112 |
ling | +0.112 |
ls | +0.111 |
behind | +0.104 |
Neg logit contributions | |
Deal | -0.132 |
Not | -0.131 |
Something | -0.123 |
They | -0.122 |
Okay | -0.117 |
Token | an |
Token ID | 281 |
Feature activation | -4.27e-02 |
Pos logit contributions | |
ling | +0.217 |
le | +0.216 |
an | +0.199 |
les | +0.192 |
egg | +0.191 |
Neg logit contributions | |
Deal | -0.185 |
Jake | -0.182 |
Jake | -0.176 |
They | -0.175 |
Something | -0.174 |
Token | island |
Token ID | 7022 |
Feature activation | -2.59e-01 |
Pos logit contributions | |
ls | +0.067 |
ling | +0.067 |
le | +0.064 |
les | +0.062 |
away | +0.062 |
Neg logit contributions | |
Deal | -0.085 |
Not | -0.073 |
Johnny | -0.072 |
Thanks | -0.070 |
Jake | -0.069 |
Token | just |
Token ID | 655 |
Feature activation | -2.14e-02 |
Pos logit contributions | |
le | +0.361 |
ling | +0.334 |
les | +0.326 |
ck | +0.304 |
led | +0.304 |
Neg logit contributions | |
They | -0.498 |
Not | -0.473 |
Deal | -0.469 |
Jake | -0.456 |
She | -0.444 |
Token | like |
Token ID | 588 |
Feature activation | -1.39e-02 |
Pos logit contributions | |
le | +0.041 |
ling | +0.040 |
ls | +0.039 |
les | +0.039 |
egg | +0.038 |
Neg logit contributions | |
Deal | -0.036 |
Not | -0.035 |
Something | -0.031 |
They | -0.031 |
Thanks | -0.031 |
Token | this |
Token ID | 428 |
Feature activation | -5.73e-02 |
Pos logit contributions | |
le | +0.024 |
ling | +0.023 |
behind | +0.020 |
les | +0.019 |
away | +0.019 |
Neg logit contributions | |
Deal | -0.029 |
Not | -0.027 |
Jake | -0.026 |
Johnny | -0.025 |
Tyler | -0.025 |
Token | ! |
Token ID | 0 |
Feature activation | +1.92e-01 |
Pos logit contributions | |
le | +0.070 |
ling | +0.068 |
behind | +0.061 |
ls | +0.060 |
ck | +0.058 |
Neg logit contributions | |
Deal | -0.123 |
Not | -0.121 |
They | -0.119 |
Jake | -0.117 |
Tyler | -0.113 |
Token |
|
Token ID | 198 |
Feature activation | +1.64e-01 |
Pos logit contributions | |
Deal | +0.394 |
Luke | +0.327 |
Jake | +0.303 |
Enough | +0.299 |
Thanks | +0.295 |
Neg logit contributions | |
inside | -0.336 |
pe | -0.327 |
on | -0.313 |
b | -0.313 |
behind | -0.303 |
Token | she |
Token ID | 673 |
Feature activation | +1.57e-02 |
Pos logit contributions | |
Deal | +0.042 |
Not | +0.035 |
Okay | +0.033 |
Jake | +0.033 |
Thanks | +0.032 |
Neg logit contributions | |
le | -0.050 |
ling | -0.048 |
behind | -0.047 |
inside | -0.046 |
pe | -0.045 |
Token | could |
Token ID | 714 |
Feature activation | +2.50e-02 |
Pos logit contributions | |
Deal | +0.026 |
Not | +0.023 |
Okay | +0.021 |
Jake | +0.021 |
Something | +0.020 |
Neg logit contributions | |
ling | -0.034 |
le | -0.033 |
behind | -0.032 |
away | -0.031 |
inside | -0.031 |
Token | go |
Token ID | 467 |
Feature activation | -7.79e-02 |
Pos logit contributions | |
Deal | +0.046 |
Not | +0.040 |
Okay | +0.038 |
Something | +0.036 |
They | +0.036 |
Neg logit contributions | |
ling | -0.052 |
le | -0.050 |
les | -0.047 |
behind | -0.046 |
ls | -0.045 |
Token | get |
Token ID | 651 |
Feature activation | -6.60e-02 |
Pos logit contributions | |
ling | +0.112 |
le | +0.106 |
les | +0.098 |
egg | +0.096 |
an | +0.096 |
Neg logit contributions | |
Deal | -0.166 |
Johnny | -0.152 |
Okay | -0.146 |
Not | -0.146 |
They | -0.145 |
Token | an |
Token ID | 281 |
Feature activation | -4.56e-02 |
Pos logit contributions | |
ling | +0.131 |
le | +0.127 |
les | +0.114 |
an | +0.114 |
behind | +0.107 |
Neg logit contributions | |
Deal | -0.111 |
Dad | -0.105 |
Johnny | -0.101 |
Daddy | -0.101 |
Jake | -0.100 |
Token | X |
Token ID | 1395 |
Feature activation | -2.58e-01 |
Pos logit contributions | |
ling | +0.083 |
le | +0.073 |
ls | +0.071 |
les | +0.070 |
away | +0.068 |
Neg logit contributions | |
Deal | -0.091 |
Not | -0.081 |
First | -0.074 |
Something | -0.073 |
Jake | -0.073 |
Token | - |
Token ID | 12 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
ling | +0.298 |
lling | +0.244 |
bank | +0.234 |
les | +0.233 |
le | +0.228 |
Neg logit contributions | |
Deal | -0.559 |
Jake | -0.529 |
Johnny | -0.518 |
Jake | -0.514 |
Not | -0.511 |
Token | ray |
Token ID | 2433 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
ling | +0.233 |
le | +0.226 |
behind | +0.216 |
les | +0.197 |
inside | +0.195 |
Neg logit contributions | |
Deal | -0.241 |
Not | -0.232 |
Jake | -0.219 |
Johnny | -0.214 |
Thanks | -0.210 |
Token | . |
Token ID | 13 |
Feature activation | +2.00e-01 |
Pos logit contributions | |
ling | +0.193 |
le | +0.165 |
an | +0.154 |
aches | +0.150 |
ck | +0.150 |
Neg logit contributions | |
Deal | -0.299 |
Jake | -0.297 |
Jake | -0.281 |
Not | -0.281 |
They | -0.272 |
Token |
|
Token ID | 198 |
Feature activation | +1.60e-01 |
Pos logit contributions | |
Deal | +0.414 |
Tyler | +0.314 |
Luke | +0.311 |
Jake | +0.308 |
Chris | +0.308 |
Neg logit contributions | |
inside | -0.377 |
up | -0.359 |
pe | -0.355 |
behind | -0.344 |
. | -0.337 |
Token |
|
Token ID | 198 |
Feature activation | +1.61e-01 |
Pos logit contributions | |
Deal | +0.320 |
Not | +0.247 |
Okay | +0.246 |
Thanks | +0.231 |
Acc | +0.228 |
Neg logit contributions | |
le | -0.324 |
inside | -0.310 |
behind | -0.305 |
ling | -0.301 |
pe | -0.295 |
Token | was |
Token ID | 373 |
Feature activation | -4.08e-03 |
Pos logit contributions | |
le | +0.147 |
ling | +0.140 |
behind | +0.133 |
pe | +0.131 |
egg | +0.129 |
Neg logit contributions | |
Deal | -0.174 |
Not | -0.157 |
Okay | -0.148 |
Thanks | -0.148 |
Jake | -0.146 |
Token | time |
Token ID | 640 |
Feature activation | -1.23e-02 |
Pos logit contributions | |
le | +0.006 |
ling | +0.006 |
behind | +0.006 |
pe | +0.006 |
les | +0.006 |
Neg logit contributions | |
Deal | -0.009 |
Not | -0.008 |
Okay | -0.008 |
Thanks | -0.008 |
Jake | -0.007 |
Token | to |
Token ID | 284 |
Feature activation | -1.36e-02 |
Pos logit contributions | |
le | +0.022 |
ling | +0.020 |
les | +0.018 |
pe | +0.018 |
behind | +0.018 |
Neg logit contributions | |
Deal | -0.027 |
Not | -0.023 |
Okay | -0.023 |
Thanks | -0.022 |
Jake | -0.021 |
Token | leave |
Token ID | 2666 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
ling | +0.025 |
le | +0.025 |
behind | +0.024 |
les | +0.023 |
inside | +0.023 |
Neg logit contributions | |
Deal | -0.028 |
Not | -0.023 |
Okay | -0.022 |
Thanks | -0.022 |
Jake | -0.021 |
Token | the |
Token ID | 262 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
ling | +0.172 |
le | +0.164 |
les | +0.152 |
b | +0.148 |
an | +0.144 |
Neg logit contributions | |
Deal | -0.245 |
Kelly | -0.231 |
They | -0.228 |
She | -0.226 |
Something | -0.221 |
Token | island |
Token ID | 7022 |
Feature activation | -2.58e-01 |
Pos logit contributions | |
le | +0.016 |
les | +0.014 |
ling | +0.013 |
lers | +0.012 |
ier | +0.011 |
Neg logit contributions | |
Thanks | -0.021 |
Deal | -0.021 |
Not | -0.020 |
Bye | -0.020 |
Have | -0.020 |
Token | and |
Token ID | 290 |
Feature activation | +2.86e-02 |
Pos logit contributions | |
le | +0.382 |
les | +0.343 |
ling | +0.320 |
lers | +0.307 |
ck | +0.303 |
Neg logit contributions | |
They | -0.475 |
She | -0.456 |
Deal | -0.442 |
ossible | -0.433 |
Thanks | -0.429 |
Token | go |
Token ID | 467 |
Feature activation | -6.30e-02 |
Pos logit contributions | |
Deal | +0.060 |
Not | +0.052 |
Okay | +0.051 |
Thanks | +0.049 |
Jake | +0.048 |
Neg logit contributions | |
le | -0.052 |
ling | -0.049 |
behind | -0.046 |
les | -0.045 |
inside | -0.044 |
Token | back |
Token ID | 736 |
Feature activation | -6.11e-02 |
Pos logit contributions | |
le | +0.081 |
ling | +0.071 |
egg | +0.064 |
les | +0.061 |
lling | +0.055 |
Neg logit contributions | |
Deal | -0.151 |
They | -0.142 |
Okay | -0.137 |
ossible | -0.135 |
Not | -0.134 |
Token | home |
Token ID | 1363 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
le | +0.095 |
ling | +0.089 |
les | +0.082 |
egg | +0.080 |
an | +0.077 |
Neg logit contributions | |
They | -0.125 |
Deal | -0.120 |
She | -0.117 |
Jake | -0.115 |
Okay | -0.114 |
Token | . |
Token ID | 13 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
le | +0.187 |
les | +0.166 |
ling | +0.160 |
egg | +0.147 |
an | +0.144 |
Neg logit contributions | |
Deal | -0.265 |
They | -0.252 |
Okay | -0.250 |
Not | -0.248 |
She | -0.240 |
Token | You |
Token ID | 1639 |
Feature activation | +3.50e-02 |
Pos logit contributions | |
Deal | +0.111 |
Not | +0.082 |
Okay | +0.078 |
Acc | +0.074 |
Jake | +0.072 |
Neg logit contributions | |
le | -0.194 |
ling | -0.184 |
behind | -0.183 |
pe | -0.182 |
inside | -0.180 |
Token | shall |
Token ID | 2236 |
Feature activation | -6.20e-02 |
Pos logit contributions | |
Deal | +0.064 |
Not | +0.057 |
Okay | +0.052 |
Something | +0.050 |
Thanks | +0.049 |
Neg logit contributions | |
le | -0.071 |
ling | -0.070 |
behind | -0.067 |
les | -0.063 |
inside | -0.063 |
Token | not |
Token ID | 407 |
Feature activation | -2.46e-03 |
Pos logit contributions | |
le | +0.119 |
ling | +0.113 |
behind | +0.111 |
inside | +0.109 |
pe | +0.106 |
Neg logit contributions | |
Deal | -0.125 |
Not | -0.105 |
Okay | -0.101 |
Jake | -0.099 |
Thanks | -0.098 |
Token | take |
Token ID | 1011 |
Feature activation | -6.76e-02 |
Pos logit contributions | |
ling | +0.004 |
le | +0.004 |
lers | +0.004 |
les | +0.004 |
led | +0.004 |
Neg logit contributions | |
Not | -0.005 |
Deal | -0.004 |
Okay | -0.004 |
They | -0.004 |
Something | -0.004 |
Token | my |
Token ID | 616 |
Feature activation | -5.15e-02 |
Pos logit contributions | |
ling | +0.144 |
le | +0.139 |
an | +0.125 |
behind | +0.124 |
fish | +0.123 |
Neg logit contributions | |
Deal | -0.111 |
Not | -0.108 |
Max | -0.092 |
Sam | -0.088 |
What | -0.087 |
Token | treasure |
Token ID | 14068 |
Feature activation | -2.58e-01 |
Pos logit contributions | |
le | +0.078 |
ling | +0.071 |
les | +0.065 |
b | +0.059 |
er | +0.059 |
Neg logit contributions | |
Deal | -0.116 |
Okay | -0.099 |
Not | -0.098 |
Daddy | -0.095 |
Sorry | -0.095 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
ling | +0.379 |
le | +0.365 |
les | +0.339 |
ck | +0.333 |
b | +0.332 |
Neg logit contributions | |
Deal | -0.501 |
Not | -0.434 |
ossible | -0.406 |
Okay | -0.406 |
They | -0.402 |
Token | Tom |
Token ID | 4186 |
Feature activation | -3.11e-02 |
Pos logit contributions | |
Deal | +0.275 |
Okay | +0.194 |
Not | +0.186 |
Thanks | +0.186 |
Jake | +0.183 |
Neg logit contributions | |
behind | -0.408 |
inside | -0.399 |
le | -0.391 |
pe | -0.386 |
away | -0.383 |
Token | was |
Token ID | 373 |
Feature activation | +2.30e-02 |
Pos logit contributions | |
le | +0.051 |
ling | +0.049 |
behind | +0.048 |
inside | +0.048 |
egg | +0.046 |
Neg logit contributions | |
Deal | -0.064 |
Not | -0.057 |
Jake | -0.055 |
Thanks | -0.053 |
Okay | -0.053 |
Token | scared |
Token ID | 12008 |
Feature activation | -4.54e-02 |
Pos logit contributions | |
Deal | +0.052 |
Not | +0.046 |
Okay | +0.044 |
Something | +0.042 |
Thanks | +0.042 |
Neg logit contributions | |
le | -0.039 |
les | -0.035 |
ling | -0.035 |
egg | -0.031 |
ls | -0.031 |
Token | and |
Token ID | 290 |
Feature activation | +5.20e-02 |
Pos logit contributions | |
le | +0.057 |
ling | +0.052 |
pe | +0.050 |
egg | +0.049 |
behind | +0.049 |
Neg logit contributions | |
Deal | -0.114 |
Not | -0.103 |
Okay | -0.098 |
Jake | -0.097 |
Tyler | -0.096 |
Token | to |
Token ID | 284 |
Feature activation | -3.51e-02 |
Pos logit contributions | |
ling | +0.202 |
le | +0.191 |
les | +0.167 |
ls | +0.162 |
behind | +0.162 |
Neg logit contributions | |
Deal | -0.210 |
Not | -0.194 |
Something | -0.190 |
ossible | -0.187 |
They | -0.183 |
Token | help |
Token ID | 1037 |
Feature activation | -5.12e-02 |
Pos logit contributions | |
ling | +0.069 |
le | +0.068 |
les | +0.063 |
b | +0.060 |
behind | +0.059 |
Neg logit contributions | |
Deal | -0.068 |
Not | -0.061 |
Okay | -0.057 |
Something | -0.055 |
They | -0.055 |
Token | her |
Token ID | 607 |
Feature activation | -5.73e-03 |
Pos logit contributions | |
ling | +0.085 |
le | +0.082 |
les | +0.080 |
b | +0.078 |
an | +0.077 |
Neg logit contributions | |
Deal | -0.096 |
Not | -0.090 |
They | -0.089 |
Kelly | -0.086 |
Jake | -0.084 |
Token | explore |
Token ID | 7301 |
Feature activation | -1.52e-01 |
Pos logit contributions | |
le | +0.009 |
les | +0.008 |
ling | +0.008 |
ls | +0.008 |
b | +0.007 |
Neg logit contributions | |
Deal | -0.012 |
Not | -0.011 |
Something | -0.011 |
They | -0.011 |
Okay | -0.011 |
Token | the |
Token ID | 262 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
ling | +0.278 |
le | +0.274 |
an | +0.269 |
b | +0.258 |
les | +0.254 |
Neg logit contributions | |
They | -0.239 |
Deal | -0.230 |
Something | -0.225 |
Not | -0.213 |
Kelly | -0.212 |
Token | island |
Token ID | 7022 |
Feature activation | -2.58e-01 |
Pos logit contributions | |
le | +0.024 |
les | +0.020 |
ier | +0.020 |
aches | +0.019 |
lers | +0.019 |
Neg logit contributions | |
They | -0.028 |
Thanks | -0.028 |
Good | -0.028 |
Not | -0.027 |
Bye | -0.027 |
Token | . |
Token ID | 13 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
le | +0.382 |
les | +0.328 |
ling | +0.325 |
ck | +0.313 |
lers | +0.298 |
Neg logit contributions | |
They | -0.540 |
Not | -0.486 |
She | -0.484 |
Deal | -0.481 |
Something | -0.465 |
Token | The |
Token ID | 383 |
Feature activation | +8.50e-02 |
Pos logit contributions | |
Deal | +0.283 |
Jake | +0.220 |
David | +0.207 |
Johnny | +0.205 |
Tyler | +0.204 |
Neg logit contributions | |
inside | -0.479 |
behind | -0.469 |
away | -0.461 |
le | -0.454 |
up | -0.449 |
Token | curious |
Token ID | 11040 |
Feature activation | -6.21e-02 |
Pos logit contributions | |
Deal | +0.135 |
Thanks | +0.116 |
Not | +0.115 |
They | +0.109 |
Dad | +0.109 |
Neg logit contributions | |
le | -0.169 |
lers | -0.153 |
away | -0.153 |
b | -0.151 |
ling | -0.151 |
Token | girl |
Token ID | 2576 |
Feature activation | -6.35e-02 |
Pos logit contributions | |
b | +0.124 |
cking | +0.116 |
lers | +0.116 |
away | +0.114 |
ck | +0.113 |
Neg logit contributions | |
They | -0.082 |
Dad | -0.081 |
She | -0.079 |
Kelly | -0.079 |
Deal | -0.079 |
Token | asked |
Token ID | 1965 |
Feature activation | +2.81e-02 |
Pos logit contributions | |
le | +0.117 |
ling | +0.114 |
behind | +0.112 |
les | +0.104 |
away | +0.104 |
Neg logit contributions | |
Deal | -0.099 |
Not | -0.098 |
Thanks | -0.094 |
They | -0.094 |
Yeah | -0.091 |
Token | had |
Token ID | 550 |
Feature activation | +1.63e-02 |
Pos logit contributions | |
le | +0.314 |
egg | +0.279 |
ling | +0.263 |
aches | +0.259 |
behind | +0.258 |
Neg logit contributions | |
Thanks | -0.284 |
Not | -0.280 |
Deal | -0.271 |
They | -0.270 |
Jake | -0.269 |
Token | seen |
Token ID | 1775 |
Feature activation | -7.46e-02 |
Pos logit contributions | |
Deal | +0.031 |
Not | +0.030 |
Thanks | +0.028 |
Okay | +0.027 |
Jake | +0.027 |
Neg logit contributions | |
le | -0.030 |
ling | -0.029 |
behind | -0.027 |
les | -0.025 |
inside | -0.025 |
Token | her |
Token ID | 607 |
Feature activation | -2.81e-02 |
Pos logit contributions | |
le | +0.136 |
ling | +0.136 |
an | +0.120 |
ls | +0.118 |
b | +0.116 |
Neg logit contributions | |
Not | -0.127 |
Deal | -0.126 |
Jake | -0.122 |
Jake | -0.121 |
Kelly | -0.119 |
Token | lose |
Token ID | 4425 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
le | +0.048 |
ling | +0.044 |
les | +0.043 |
ls | +0.038 |
behind | +0.038 |
Neg logit contributions | |
Deal | -0.058 |
Not | -0.052 |
Okay | -0.051 |
Thanks | -0.048 |
Something | -0.048 |
Token | her |
Token ID | 607 |
Feature activation | -3.97e-02 |
Pos logit contributions | |
le | +0.159 |
ling | +0.145 |
an | +0.128 |
les | +0.128 |
ck | +0.121 |
Neg logit contributions | |
Deal | -0.260 |
Not | -0.248 |
Something | -0.241 |
They | -0.238 |
Tyler | -0.237 |
Token | anchor |
Token ID | 18021 |
Feature activation | -2.57e-01 |
Pos logit contributions | |
le | +0.069 |
ling | +0.063 |
les | +0.062 |
ls | +0.057 |
behind | +0.055 |
Neg logit contributions | |
Deal | -0.080 |
Not | -0.071 |
Okay | -0.068 |
Thanks | -0.068 |
Dad | -0.066 |
Token | and |
Token ID | 290 |
Feature activation | +3.40e-02 |
Pos logit contributions | |
le | +0.359 |
les | +0.320 |
ls | +0.315 |
ling | +0.292 |
pe | +0.286 |
Neg logit contributions | |
Not | -0.490 |
Deal | -0.472 |
Jake | -0.457 |
They | -0.453 |
She | -0.446 |
Token | it |
Token ID | 340 |
Feature activation | -8.63e-02 |
Pos logit contributions | |
Deal | +0.063 |
Not | +0.059 |
Thanks | +0.056 |
Okay | +0.056 |
Jake | +0.055 |
Neg logit contributions | |
le | -0.064 |
ling | -0.059 |
les | -0.053 |
behind | -0.053 |
an | -0.051 |
Token | sailed |
Token ID | 33835 |
Feature activation | -1.03e-01 |
Pos logit contributions | |
le | +0.165 |
egg | +0.150 |
behind | +0.148 |
ling | +0.147 |
inside | +0.142 |
Neg logit contributions | |
Not | -0.125 |
Thanks | -0.121 |
Jake | -0.119 |
Deal | -0.118 |
Jake | -0.114 |
Token | over |
Token ID | 625 |
Feature activation | -6.26e-02 |
Pos logit contributions | |
le | +0.086 |
egg | +0.080 |
ling | +0.068 |
lers | +0.066 |
pe | +0.065 |
Neg logit contributions | |
Deal | -0.275 |
Thanks | -0.271 |
Jake | -0.257 |
They | -0.252 |
Not | -0.251 |
Token | to |
Token ID | 284 |
Feature activation | -2.38e-02 |
Pos logit contributions | |
le | +0.108 |
ling | +0.105 |
les | +0.094 |
an | +0.093 |
b | +0.088 |
Neg logit contributions | |
Deal | -0.129 |
Not | -0.114 |
Jake | -0.113 |
They | -0.111 |
Okay | -0.110 |
Token | never |
Token ID | 1239 |
Feature activation | -3.38e-02 |
Pos logit contributions | |
Deal | +0.054 |
Not | +0.050 |
Thanks | +0.047 |
Okay | +0.046 |
David | +0.045 |
Neg logit contributions | |
le | -0.058 |
ling | -0.054 |
behind | -0.050 |
les | -0.050 |
inside | -0.048 |
Token | seen |
Token ID | 1775 |
Feature activation | -7.90e-02 |
Pos logit contributions | |
le | +0.065 |
ling | +0.063 |
ls | +0.061 |
les | +0.060 |
lers | +0.059 |
Neg logit contributions | |
Deal | -0.054 |
Not | -0.054 |
Thanks | -0.052 |
David | -0.051 |
Something | -0.050 |
Token | such |
Token ID | 884 |
Feature activation | -1.04e-02 |
Pos logit contributions | |
ling | +0.169 |
le | +0.161 |
ls | +0.157 |
an | +0.153 |
lers | +0.144 |
Neg logit contributions | |
Not | -0.110 |
Jake | -0.109 |
Kelly | -0.100 |
Sam | -0.100 |
How | -0.099 |
Token | a |
Token ID | 257 |
Feature activation | -2.99e-02 |
Pos logit contributions | |
le | +0.022 |
ling | +0.020 |
les | +0.020 |
behind | +0.019 |
ls | +0.019 |
Neg logit contributions | |
Deal | -0.017 |
Not | -0.016 |
Something | -0.015 |
Okay | -0.015 |
Thanks | -0.014 |
Token | small |
Token ID | 1402 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
le | +0.056 |
les | +0.047 |
behind | +0.047 |
ling | +0.047 |
led | +0.046 |
Neg logit contributions | |
Thanks | -0.052 |
Deal | -0.052 |
David | -0.049 |
Okay | -0.048 |
Not | -0.048 |
Token | island |
Token ID | 7022 |
Feature activation | -2.57e-01 |
Pos logit contributions | |
le | +0.180 |
aches | +0.172 |
ck | +0.169 |
ls | +0.166 |
b | +0.162 |
Neg logit contributions | |
They | -0.193 |
Dad | -0.189 |
Daddy | -0.187 |
Ryan | -0.185 |
Deal | -0.185 |
Token | before |
Token ID | 878 |
Feature activation | +2.92e-03 |
Pos logit contributions | |
ls | +0.447 |
les | +0.441 |
lers | +0.440 |
le | +0.431 |
ck | +0.412 |
Neg logit contributions | |
� | -0.389 |
Thanks | -0.374 |
Not | -0.372 |
Have | -0.364 |
Jake | -0.357 |
Token | . |
Token ID | 13 |
Feature activation | +1.87e-01 |
Pos logit contributions | |
Not | +0.007 |
Deal | +0.007 |
They | +0.007 |
Tyler | +0.006 |
Jake | +0.006 |
Neg logit contributions | |
le | -0.004 |
ling | -0.004 |
an | -0.003 |
les | -0.003 |
away | -0.003 |
Token | But |
Token ID | 887 |
Feature activation | +8.46e-02 |
Pos logit contributions | |
Deal | +0.237 |
Jake | +0.164 |
Okay | +0.159 |
Thanks | +0.157 |
Not | +0.146 |
Neg logit contributions | |
le | -0.481 |
behind | -0.476 |
inside | -0.472 |
pe | -0.459 |
ling | -0.455 |
Token | they |
Token ID | 484 |
Feature activation | +2.48e-02 |
Pos logit contributions | |
Deal | +0.155 |
Not | +0.148 |
Thanks | +0.136 |
Okay | +0.135 |
They | +0.134 |
Neg logit contributions | |
le | -0.161 |
les | -0.152 |
ling | -0.151 |
an | -0.136 |
behind | -0.133 |
Token | soon |
Token ID | 2582 |
Feature activation | -6.56e-03 |
Pos logit contributions | |
Deal | +0.038 |
Not | +0.035 |
Thanks | +0.033 |
Okay | +0.032 |
Something | +0.031 |
Neg logit contributions | |
le | -0.053 |
ling | -0.050 |
les | -0.049 |
behind | -0.047 |
egg | -0.046 |
Token | book |
Token ID | 1492 |
Feature activation | -1.62e-01 |
Pos logit contributions | |
Deal | +0.037 |
Okay | +0.031 |
Not | +0.031 |
Something | +0.030 |
Jake | +0.029 |
Neg logit contributions | |
le | -0.029 |
ling | -0.026 |
behind | -0.023 |
les | -0.023 |
ls | -0.022 |
Token | , |
Token ID | 11 |
Feature activation | +5.50e-02 |
Pos logit contributions | |
le | +0.269 |
ling | +0.256 |
egg | +0.228 |
an | +0.220 |
pe | +0.220 |
Neg logit contributions | |
Deal | -0.338 |
Not | -0.317 |
They | -0.303 |
She | -0.292 |
Jake | -0.287 |
Token | and |
Token ID | 290 |
Feature activation | +4.25e-02 |
Pos logit contributions | |
Deal | +0.131 |
Not | +0.118 |
Okay | +0.112 |
Jake | +0.111 |
They | +0.110 |
Neg logit contributions | |
le | -0.084 |
ling | -0.075 |
les | -0.067 |
inside | -0.065 |
behind | -0.065 |
Token | the |
Token ID | 262 |
Feature activation | +2.68e-02 |
Pos logit contributions | |
Deal | +0.092 |
Not | +0.081 |
Okay | +0.077 |
Jake | +0.076 |
They | +0.075 |
Neg logit contributions | |
le | -0.074 |
ling | -0.069 |
behind | -0.062 |
inside | -0.061 |
les | -0.060 |
Token | hat |
Token ID | 6877 |
Feature activation | -1.89e-01 |
Pos logit contributions | |
Deal | +0.061 |
Okay | +0.051 |
Not | +0.051 |
Something | +0.050 |
Jake | +0.049 |
Neg logit contributions | |
le | -0.046 |
ling | -0.041 |
behind | -0.038 |
les | -0.037 |
inside | -0.036 |
Token | . |
Token ID | 13 |
Feature activation | +1.92e-01 |
Pos logit contributions | |
le | +0.252 |
ling | +0.224 |
egg | +0.209 |
behind | +0.194 |
les | +0.190 |
Neg logit contributions | |
They | -0.423 |
Deal | -0.420 |
Not | -0.416 |
She | -0.404 |
He | -0.386 |
Token | He |
Token ID | 679 |
Feature activation | +3.69e-02 |
Pos logit contributions | |
Deal | +0.288 |
Jake | +0.209 |
Thanks | +0.209 |
liv | +0.199 |
omo | +0.196 |
Neg logit contributions | |
inside | -0.443 |
behind | -0.439 |
pe | -0.419 |
on | -0.411 |
le | -0.398 |
Token | does |
Token ID | 857 |
Feature activation | +1.80e-02 |
Pos logit contributions | |
Deal | +0.083 |
Not | +0.070 |
Okay | +0.068 |
Jake | +0.067 |
Thanks | +0.067 |
Neg logit contributions | |
le | -0.062 |
ling | -0.059 |
behind | -0.058 |
inside | -0.057 |
pe | -0.056 |
Token | not |
Token ID | 407 |
Feature activation | +2.14e-02 |
Pos logit contributions | |
Deal | +0.028 |
Not | +0.023 |
Okay | +0.021 |
Thanks | +0.020 |
Jake | +0.020 |
Neg logit contributions | |
le | -0.043 |
ling | -0.041 |
behind | -0.040 |
inside | -0.039 |
les | -0.039 |
Token | see |
Token ID | 766 |
Feature activation | -3.96e-02 |
Pos logit contributions | |
Deal | +0.032 |
Not | +0.027 |
Okay | +0.025 |
Thanks | +0.025 |
Something | +0.024 |
Neg logit contributions | |
ling | -0.048 |
le | -0.045 |
lers | -0.044 |
ls | -0.044 |
egg | -0.044 |
Token | the |
Token ID | 262 |
Feature activation | -3.47e-03 |
Pos logit contributions | |
le | +0.076 |
ling | +0.075 |
pe | +0.066 |
an | +0.065 |
inside | +0.065 |
Neg logit contributions | |
Deal | -0.069 |
Not | -0.065 |
Jake | -0.060 |
They | -0.059 |
Tyler | -0.058 |
Token | fun |
Token ID | 1257 |
Feature activation | -8.36e-02 |
Pos logit contributions | |
Deal | +0.017 |
Acc | +0.015 |
Not | +0.015 |
Winning | +0.014 |
Okay | +0.014 |
Neg logit contributions | |
le | -0.015 |
ling | -0.014 |
behind | -0.013 |
les | -0.013 |
ls | -0.013 |
Token | she |
Token ID | 673 |
Feature activation | -2.00e-02 |
Pos logit contributions | |
le | +0.154 |
ling | +0.150 |
behind | +0.143 |
away | +0.139 |
pe | +0.136 |
Neg logit contributions | |
Deal | -0.145 |
Not | -0.137 |
Jake | -0.131 |
Kelly | -0.128 |
Tyler | -0.128 |
Token | had |
Token ID | 550 |
Feature activation | +4.81e-03 |
Pos logit contributions | |
ling | +0.042 |
egg | +0.041 |
ls | +0.039 |
rier | +0.039 |
le | +0.039 |
Neg logit contributions | |
Deal | -0.025 |
Thanks | -0.024 |
Not | -0.024 |
Jake | -0.024 |
Max | -0.022 |
Token | that |
Token ID | 326 |
Feature activation | +1.35e-02 |
Pos logit contributions | |
Thanks | +0.009 |
Not | +0.008 |
Jake | +0.008 |
Deal | +0.008 |
Okay | +0.008 |
Neg logit contributions | |
ling | -0.008 |
egg | -0.008 |
an | -0.008 |
le | -0.007 |
rier | -0.007 |
Token | night |
Token ID | 1755 |
Feature activation | -5.59e-02 |
Pos logit contributions | |
Deal | +0.025 |
Not | +0.024 |
Okay | +0.023 |
Thanks | +0.022 |
Something | +0.022 |
Neg logit contributions | |
le | -0.026 |
ling | -0.024 |
les | -0.023 |
ls | -0.022 |
behind | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | +1.96e-01 |
Pos logit contributions | |
le | +0.065 |
ling | +0.064 |
pe | +0.062 |
an | +0.062 |
egg | +0.062 |
Neg logit contributions | |
Not | -0.125 |
Deal | -0.125 |
Jake | -0.119 |
Kelly | -0.118 |
Jake | -0.116 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +4.79e-02 |
Pos logit contributions | |
Deal | +0.428 |
Tyler | +0.365 |
Jake | +0.365 |
Yeah | +0.358 |
David | +0.353 |
Neg logit contributions | |
inside | -0.277 |
. | -0.253 |
tree | -0.246 |
behind | -0.244 |
again | -0.235 |
Token | Once |
Token ID | 7454 |
Feature activation | +6.50e-02 |
Pos logit contributions | |
Deal | +0.068 |
Not | +0.053 |
Okay | +0.051 |
Jake | +0.050 |
Thanks | +0.048 |
Neg logit contributions | |
le | -0.118 |
ling | -0.114 |
behind | -0.112 |
inside | -0.109 |
pe | -0.108 |
Token | upon |
Token ID | 2402 |
Feature activation | +9.93e-02 |
Pos logit contributions | |
Deal | +0.120 |
Not | +0.098 |
Okay | +0.095 |
Jake | +0.093 |
Thanks | +0.091 |
Neg logit contributions | |
le | -0.135 |
ling | -0.128 |
behind | -0.127 |
inside | -0.125 |
pe | -0.122 |
Token | a |
Token ID | 257 |
Feature activation | +4.44e-02 |
Pos logit contributions | |
Deal | +0.250 |
Okay | +0.206 |
Not | +0.205 |
Jake | +0.202 |
Thanks | +0.200 |
Neg logit contributions | |
le | -0.144 |
behind | -0.138 |
inside | -0.137 |
pe | -0.130 |
ling | -0.129 |
Token | time |
Token ID | 640 |
Feature activation | +5.55e-03 |
Pos logit contributions | |
Deal | +0.122 |
Not | +0.107 |
Okay | +0.106 |
Jake | +0.105 |
Thanks | +0.104 |
Neg logit contributions | |
le | -0.052 |
ling | -0.047 |
behind | -0.046 |
inside | -0.043 |
pe | -0.042 |
Token | will |
Token ID | 481 |
Feature activation | -3.47e-02 |
Pos logit contributions | |
behind | +0.158 |
ling | +0.153 |
le | +0.152 |
egg | +0.148 |
inside | +0.144 |
Neg logit contributions | |
Deal | -0.148 |
Not | -0.139 |
They | -0.125 |
Okay | -0.123 |
Something | -0.120 |
Token | come |
Token ID | 1282 |
Feature activation | -8.66e-02 |
Pos logit contributions | |
ling | +0.069 |
le | +0.069 |
lers | +0.066 |
les | +0.065 |
behind | +0.064 |
Neg logit contributions | |
Deal | -0.064 |
Not | -0.056 |
Okay | -0.052 |
Thanks | -0.051 |
They | -0.049 |
Token | true |
Token ID | 2081 |
Feature activation | -7.62e-02 |
Pos logit contributions | |
le | +0.159 |
ling | +0.154 |
les | +0.150 |
lers | +0.140 |
egg | +0.139 |
Neg logit contributions | |
Deal | -0.173 |
Okay | -0.148 |
Not | -0.144 |
They | -0.141 |
Thanks | -0.132 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.71e-01 |
Pos logit contributions | |
ling | +0.121 |
le | +0.106 |
b | +0.106 |
pe | +0.101 |
les | +0.100 |
Neg logit contributions | |
Deal | -0.163 |
Not | -0.149 |
Okay | -0.146 |
They | -0.141 |
Jake | -0.136 |
Token |
|
Token ID | 198 |
Feature activation | +1.67e-01 |
Pos logit contributions | |
Deal | +0.307 |
Jake | +0.232 |
Okay | +0.227 |
Not | +0.225 |
Johnny | +0.220 |
Neg logit contributions | |
le | -0.360 |
inside | -0.359 |
behind | -0.355 |
pe | -0.347 |
away | -0.344 |
Token |
|
Token ID | 198 |
Feature activation | +1.68e-01 |
Pos logit contributions | |
Deal | +0.314 |
Not | +0.247 |
Okay | +0.238 |
Jake | +0.230 |
Thanks | +0.229 |
Neg logit contributions | |
le | -0.342 |
inside | -0.337 |
behind | -0.329 |
ling | -0.327 |
pe | -0.317 |
Token | But |
Token ID | 1537 |
Feature activation | +9.47e-02 |
Pos logit contributions | |
Deal | +0.290 |
Not | +0.219 |
Okay | +0.206 |
Acc | +0.203 |
Jake | +0.197 |
Neg logit contributions | |
le | -0.383 |
inside | -0.367 |
ling | -0.364 |
away | -0.358 |
behind | -0.358 |
Token | Wendy |
Token ID | 28314 |
Feature activation | -4.87e-02 |
Pos logit contributions | |
Deal | +0.150 |
Not | +0.119 |
Okay | +0.115 |
Thanks | +0.111 |
Jake | +0.110 |
Neg logit contributions | |
le | -0.221 |
ling | -0.212 |
behind | -0.207 |
inside | -0.205 |
pe | -0.202 |
Token | said |
Token ID | 531 |
Feature activation | +3.85e-02 |
Pos logit contributions | |
le | +0.083 |
ling | +0.081 |
behind | +0.079 |
inside | +0.077 |
away | +0.074 |
Neg logit contributions | |
Deal | -0.096 |
Not | -0.086 |
Jake | -0.082 |
Thanks | -0.080 |
Okay | -0.079 |
Token | , |
Token ID | 11 |
Feature activation | +8.66e-02 |
Pos logit contributions | |
Deal | +0.100 |
Not | +0.091 |
Okay | +0.087 |
Thanks | +0.085 |
Jake | +0.084 |
Neg logit contributions | |
le | -0.049 |
ling | -0.044 |
behind | -0.043 |
les | -0.041 |
pe | -0.040 |
Token | " |
Token ID | 366 |
Feature activation | +8.51e-02 |
Pos logit contributions | |
Deal | +0.121 |
Not | +0.088 |
Okay | +0.086 |
Jake | +0.084 |
Thanks | +0.081 |
Neg logit contributions | |
le | -0.218 |
ling | -0.211 |
behind | -0.211 |
inside | -0.209 |
pe | -0.204 |
Token | know |
Token ID | 760 |
Feature activation | -3.14e-02 |
Pos logit contributions | |
Deal | +0.013 |
Not | +0.010 |
Okay | +0.010 |
Jake | +0.010 |
Thanks | +0.010 |
Neg logit contributions | |
le | -0.014 |
ling | -0.014 |
behind | -0.014 |
inside | -0.013 |
pe | -0.013 |
Token | , |
Token ID | 11 |
Feature activation | +5.15e-02 |
Pos logit contributions | |
ling | +0.047 |
le | +0.043 |
away | +0.040 |
an | +0.038 |
les | +0.037 |
Neg logit contributions | |
Deal | -0.069 |
Not | -0.065 |
Jake | -0.064 |
Okay | -0.062 |
They | -0.061 |
Token | it |
Token ID | 340 |
Feature activation | -8.73e-02 |
Pos logit contributions | |
Deal | +0.107 |
Not | +0.088 |
Okay | +0.086 |
Jake | +0.085 |
Thanks | +0.085 |
Neg logit contributions | |
le | -0.096 |
behind | -0.091 |
ling | -0.090 |
inside | -0.089 |
pe | -0.087 |
Token | looks |
Token ID | 3073 |
Feature activation | -8.93e-02 |
Pos logit contributions | |
behind | +0.171 |
ling | +0.165 |
le | +0.165 |
inside | +0.156 |
away | +0.153 |
Neg logit contributions | |
Deal | -0.135 |
Not | -0.134 |
They | -0.119 |
Something | -0.115 |
Okay | -0.114 |
Token | hard |
Token ID | 1327 |
Feature activation | -9.29e-02 |
Pos logit contributions | |
le | +0.168 |
ling | +0.165 |
les | +0.160 |
an | +0.159 |
egg | +0.150 |
Neg logit contributions | |
Deal | -0.156 |
Not | -0.134 |
Okay | -0.127 |
They | -0.123 |
Something | -0.120 |
Token | ," |
Token ID | 553 |
Feature activation | +1.42e-01 |
Pos logit contributions | |
le | +0.109 |
ling | +0.104 |
pe | +0.101 |
an | +0.099 |
behind | +0.093 |
Neg logit contributions | |
Deal | -0.222 |
They | -0.202 |
Not | -0.201 |
She | -0.190 |
First | -0.190 |
Token | Lily |
Token ID | 20037 |
Feature activation | -5.24e-04 |
Pos logit contributions | |
Deal | +0.232 |
Not | +0.176 |
Thanks | +0.175 |
Okay | +0.171 |
Jake | +0.159 |
Neg logit contributions | |
pe | -0.322 |
behind | -0.321 |
le | -0.319 |
inside | -0.314 |
away | -0.301 |
Token | says |
Token ID | 1139 |
Feature activation | +2.28e-02 |
Pos logit contributions | |
le | +0.001 |
ling | +0.001 |
behind | +0.001 |
inside | +0.001 |
pe | +0.001 |
Neg logit contributions | |
Deal | -0.001 |
Not | -0.001 |
Okay | -0.001 |
Jake | -0.001 |
Thanks | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +2.06e-01 |
Pos logit contributions | |
Deal | +0.053 |
They | +0.047 |
Not | +0.046 |
She | +0.046 |
Okay | +0.046 |
Neg logit contributions | |
le | -0.032 |
island | -0.031 |
asis | -0.030 |
ling | -0.029 |
pe | -0.029 |
Token | " |
Token ID | 366 |
Feature activation | +9.10e-02 |
Pos logit contributions | |
Deal | +0.411 |
Thanks | +0.333 |
Jake | +0.315 |
Okay | +0.314 |
Johnny | +0.311 |
Neg logit contributions | |
behind | -0.396 |
inside | -0.394 |
pe | -0.371 |
on | -0.360 |
le | -0.355 |
Token | But |
Token ID | 1537 |
Feature activation | +8.58e-02 |
Pos logit contributions | |
Deal | +0.128 |
Not | +0.093 |
Okay | +0.090 |
Jake | +0.086 |
Thanks | +0.085 |
Neg logit contributions | |
le | -0.232 |
ling | -0.223 |
behind | -0.222 |
inside | -0.220 |
pe | -0.215 |
Token | made |
Token ID | 925 |
Feature activation | -2.84e-02 |
Pos logit contributions | |
le | +0.262 |
les | +0.242 |
ling | +0.228 |
an | +0.224 |
lers | +0.221 |
Neg logit contributions | |
She | -0.390 |
Deal | -0.382 |
Thanks | -0.377 |
Jake | -0.374 |
They | -0.374 |
Token | of |
Token ID | 286 |
Feature activation | -3.48e-02 |
Pos logit contributions | |
ling | +0.047 |
le | +0.046 |
behind | +0.042 |
pe | +0.042 |
an | +0.041 |
Neg logit contributions | |
Deal | -0.059 |
Not | -0.053 |
Jake | -0.051 |
Jake | -0.049 |
Okay | -0.049 |
Token | past |
Token ID | 1613 |
Feature activation | -9.39e-02 |
Pos logit contributions | |
le | +0.058 |
ling | +0.052 |
behind | +0.048 |
les | +0.047 |
an | +0.046 |
Neg logit contributions | |
Deal | -0.075 |
Not | -0.066 |
Something | -0.066 |
Jake | -0.064 |
Okay | -0.064 |
Token | el |
Token ID | 417 |
Feature activation | -1.53e-01 |
Pos logit contributions | |
ling | +0.122 |
le | +0.121 |
b | +0.121 |
an | +0.120 |
ring | +0.119 |
Neg logit contributions | |
They | -0.182 |
Jake | -0.180 |
She | -0.174 |
Not | -0.169 |
Jake | -0.167 |
Token | colors |
Token ID | 7577 |
Feature activation | -1.55e-01 |
Pos logit contributions | |
le | +0.163 |
behind | +0.156 |
away | +0.147 |
ls | +0.144 |
aches | +0.143 |
Neg logit contributions | |
Deal | -0.328 |
They | -0.322 |
Not | -0.303 |
She | -0.297 |
He | -0.294 |
Token | . |
Token ID | 13 |
Feature activation | +1.96e-01 |
Pos logit contributions | |
le | +0.200 |
an | +0.185 |
ling | +0.183 |
away | +0.164 |
egg | +0.162 |
Neg logit contributions | |
They | -0.325 |
She | -0.315 |
Not | -0.300 |
Jake | -0.290 |
Deal | -0.288 |
Token |
|
Token ID | 198 |
Feature activation | +1.62e-01 |
Pos logit contributions | |
Deal | +0.294 |
Thanks | +0.270 |
Johnny | +0.265 |
omo | +0.254 |
David | +0.247 |
Neg logit contributions | |
inside | -0.411 |
behind | -0.407 |
le | -0.391 |
pe | -0.386 |
up | -0.379 |
Token |
|
Token ID | 198 |
Feature activation | +1.63e-01 |
Pos logit contributions | |
Deal | +0.291 |
Okay | +0.227 |
Thanks | +0.227 |
Acc | +0.224 |
Not | +0.224 |
Neg logit contributions | |
le | -0.333 |
inside | -0.320 |
behind | -0.319 |
ling | -0.312 |
pe | -0.306 |
Token | When |
Token ID | 2215 |
Feature activation | +6.53e-02 |
Pos logit contributions | |
Deal | +0.287 |
Acc | +0.238 |
Sorry | +0.228 |
Okay | +0.227 |
Thanks | +0.224 |
Neg logit contributions | |
le | -0.343 |
behind | -0.320 |
ling | -0.318 |
inside | -0.318 |
pe | -0.308 |
Token | Ashley |
Token ID | 20899 |
Feature activation | -4.42e-02 |
Pos logit contributions | |
Deal | +0.137 |
Not | +0.111 |
Okay | +0.110 |
Thanks | +0.109 |
Jake | +0.107 |
Neg logit contributions | |
le | -0.120 |
behind | -0.115 |
inside | -0.115 |
ling | -0.111 |
away | -0.110 |
Token | tried |
Token ID | 3088 |
Feature activation | -2.91e-03 |
Pos logit contributions | |
ling | +0.081 |
le | +0.074 |
aches | +0.072 |
behind | +0.072 |
b | +0.071 |
Neg logit contributions | |
Deal | -0.074 |
Not | -0.071 |
Jake | -0.069 |
Something | -0.068 |
Tyler | -0.065 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +4.00e-02 |
Pos logit contributions | |
Johnny | +0.331 |
David | +0.320 |
Deal | +0.319 |
Kim | +0.315 |
Jake | +0.308 |
Neg logit contributions | |
around | -0.347 |
inside | -0.346 |
up | -0.338 |
le | -0.332 |
les | -0.326 |
Token | Once |
Token ID | 7454 |
Feature activation | +6.20e-02 |
Pos logit contributions | |
Deal | +0.066 |
Not | +0.053 |
Okay | +0.051 |
Jake | +0.050 |
Thanks | +0.049 |
Neg logit contributions | |
le | -0.090 |
ling | -0.086 |
behind | -0.086 |
inside | -0.085 |
pe | -0.082 |
Token | upon |
Token ID | 2402 |
Feature activation | +9.69e-02 |
Pos logit contributions | |
Deal | +0.124 |
Not | +0.101 |
Okay | +0.099 |
Thanks | +0.098 |
Jake | +0.097 |
Neg logit contributions | |
le | -0.119 |
behind | -0.114 |
ling | -0.114 |
inside | -0.114 |
pe | -0.108 |
Token | a |
Token ID | 257 |
Feature activation | +3.27e-02 |
Pos logit contributions | |
Deal | +0.251 |
Okay | +0.206 |
Not | +0.205 |
Jake | +0.204 |
Tyler | +0.201 |
Neg logit contributions | |
le | -0.134 |
inside | -0.129 |
behind | -0.129 |
pe | -0.122 |
ling | -0.119 |
Token | time |
Token ID | 640 |
Feature activation | -4.08e-03 |
Pos logit contributions | |
Deal | +0.092 |
Okay | +0.081 |
Not | +0.080 |
Jake | +0.080 |
Thanks | +0.078 |
Neg logit contributions | |
le | -0.036 |
ling | -0.032 |
behind | -0.032 |
inside | -0.030 |
pe | -0.028 |
Token | there |
Token ID | 612 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
le | +0.003 |
behind | +0.003 |
ling | +0.003 |
inside | +0.003 |
pe | +0.002 |
Neg logit contributions | |
Deal | -0.013 |
Not | -0.011 |
Okay | -0.011 |
Thanks | -0.011 |
Jake | -0.011 |
Token | was |
Token ID | 373 |
Feature activation | -6.53e-04 |
Pos logit contributions | |
le | +0.019 |
ling | +0.018 |
behind | +0.017 |
inside | +0.017 |
away | +0.016 |
Neg logit contributions | |
Deal | -0.030 |
Not | -0.026 |
Jake | -0.025 |
Okay | -0.025 |
Thanks | -0.025 |
Token | a |
Token ID | 257 |
Feature activation | +1.02e-02 |
Pos logit contributions | |
inside | +0.001 |
le | +0.001 |
behind | +0.001 |
pe | +0.001 |
up | +0.001 |
Neg logit contributions | |
Deal | -0.002 |
Okay | -0.002 |
Not | -0.001 |
Sorry | -0.001 |
Jake | -0.001 |
Token | little |
Token ID | 1310 |
Feature activation | -4.70e-02 |
Pos logit contributions | |
Deal | +0.026 |
Not | +0.023 |
Okay | +0.022 |
Jake | +0.022 |
Thanks | +0.022 |
Neg logit contributions | |
le | -0.014 |
ling | -0.013 |
behind | -0.013 |
inside | -0.013 |
pe | -0.012 |
Token | girl |
Token ID | 2576 |
Feature activation | -9.43e-02 |
Pos logit contributions | |
le | +0.084 |
ling | +0.078 |
behind | +0.078 |
inside | +0.074 |
away | +0.074 |
Neg logit contributions | |
Deal | -0.094 |
Not | -0.081 |
Okay | -0.079 |
Jake | -0.079 |
Johnny | -0.075 |
Token | called |
Token ID | 1444 |
Feature activation | -2.27e-02 |
Pos logit contributions | |
le | +0.201 |
ling | +0.195 |
les | +0.187 |
away | +0.181 |
behind | +0.179 |
Neg logit contributions | |
Deal | -0.146 |
Not | -0.128 |
Jake | -0.124 |
Okay | -0.121 |
They | -0.119 |
Token | â |
Token ID | 22940 |
Feature activation | +4.41e-02 |
Pos logit contributions | |
Deal | +0.261 |
Jake | +0.202 |
Thanks | +0.191 |
Luke | +0.183 |
Not | +0.180 |
Neg logit contributions | |
inside | -0.451 |
behind | -0.432 |
pe | -0.423 |
le | -0.420 |
on | -0.414 |
Token | € |
Token ID | 26391 |
Feature activation | +7.74e-02 |
Pos logit contributions | |
Deal | +0.059 |
Not | +0.044 |
Okay | +0.042 |
Jake | +0.040 |
Thanks | +0.038 |
Neg logit contributions | |
le | -0.115 |
ling | -0.111 |
behind | -0.110 |
inside | -0.109 |
away | -0.106 |
Token |
|
Token ID | 198 |
Feature activation | +1.63e-01 |
Pos logit contributions | |
Deal | +0.159 |
Okay | +0.122 |
Thanks | +0.121 |
Brother | +0.115 |
Acc | +0.114 |
Neg logit contributions | |
inside | -0.140 |
behind | -0.136 |
le | -0.133 |
pe | -0.132 |
- | -0.127 |
Token |
|
Token ID | 198 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
Deal | +0.316 |
Not | +0.251 |
Sorry | +0.246 |
Okay | +0.245 |
Acc | +0.236 |
Neg logit contributions | |
inside | -0.300 |
le | -0.300 |
behind | -0.291 |
ling | -0.290 |
pe | -0.284 |
Token | Tim |
Token ID | 14967 |
Feature activation | -3.76e-02 |
Pos logit contributions | |
Deal | +0.330 |
Acc | +0.263 |
Not | +0.259 |
Sorry | +0.257 |
Okay | +0.254 |
Neg logit contributions | |
le | -0.300 |
ling | -0.293 |
inside | -0.281 |
pe | -0.276 |
behind | -0.275 |
Token | smiled |
Token ID | 13541 |
Feature activation | -1.58e-02 |
Pos logit contributions | |
le | +0.068 |
ling | +0.065 |
behind | +0.064 |
inside | +0.063 |
pe | +0.061 |
Neg logit contributions | |
Deal | -0.078 |
Not | -0.066 |
Okay | -0.064 |
Jake | -0.063 |
Thanks | -0.062 |
Token | in |
Token ID | 287 |
Feature activation | +2.07e-02 |
Pos logit contributions | |
le | +0.014 |
ling | +0.013 |
behind | +0.013 |
inside | +0.012 |
pe | +0.011 |
Neg logit contributions | |
Deal | -0.048 |
Not | -0.042 |
Okay | -0.042 |
Jake | -0.041 |
Thanks | -0.041 |
Token | delight |
Token ID | 10974 |
Feature activation | -8.83e-02 |
Pos logit contributions | |
Deal | +0.046 |
Jake | +0.041 |
Not | +0.041 |
Acc | +0.040 |
Okay | +0.039 |
Neg logit contributions | |
behind | -0.031 |
inside | -0.030 |
pe | -0.030 |
away | -0.030 |
le | -0.030 |
Token | and |
Token ID | 290 |
Feature activation | +6.01e-02 |
Pos logit contributions | |
le | +0.086 |
ling | +0.075 |
behind | +0.074 |
inside | +0.071 |
pe | +0.069 |
Neg logit contributions | |
Deal | -0.249 |
Not | -0.223 |
Jake | -0.217 |
Okay | -0.214 |
Thanks | -0.211 |
Token | proudly |
Token ID | 23017 |
Feature activation | -1.37e-02 |
Pos logit contributions | |
Deal | +0.135 |
Not | +0.116 |
Okay | +0.113 |
Jake | +0.112 |
Thanks | +0.112 |
Neg logit contributions | |
le | -0.099 |
ling | -0.093 |
behind | -0.091 |
inside | -0.088 |
away | -0.085 |
Token | said |
Token ID | 531 |
Feature activation | +3.53e-02 |
Pos logit contributions | |
le | +0.025 |
ling | +0.024 |
behind | +0.024 |
inside | +0.022 |
egg | +0.022 |
Neg logit contributions | |
Deal | -0.025 |
Thanks | -0.023 |
Jake | -0.023 |
Not | -0.022 |
Tyler | -0.022 |
Token | nice |
Token ID | 3621 |
Feature activation | -6.52e-02 |
Pos logit contributions | |
Deal | +0.008 |
Not | +0.007 |
Okay | +0.007 |
Thanks | +0.006 |
Kelly | +0.006 |
Neg logit contributions | |
le | -0.007 |
ling | -0.007 |
les | -0.007 |
ls | -0.006 |
an | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | +1.24e-02 |
Pos logit contributions | |
le | +0.095 |
ling | +0.090 |
behind | +0.082 |
pe | +0.081 |
ck | +0.078 |
Neg logit contributions | |
Deal | -0.148 |
Not | -0.138 |
They | -0.132 |
Okay | -0.132 |
She | -0.125 |
Token | honest |
Token ID | 5508 |
Feature activation | -8.70e-02 |
Pos logit contributions | |
Deal | +0.024 |
Not | +0.021 |
Okay | +0.020 |
Jake | +0.019 |
Thanks | +0.019 |
Neg logit contributions | |
le | -0.024 |
ling | -0.023 |
behind | -0.022 |
inside | -0.021 |
les | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
le | +0.101 |
ling | +0.095 |
behind | +0.090 |
pe | +0.088 |
egg | +0.080 |
Neg logit contributions | |
Deal | -0.228 |
Not | -0.208 |
Okay | -0.201 |
They | -0.199 |
Jake | -0.193 |
Token | He |
Token ID | 679 |
Feature activation | +2.28e-02 |
Pos logit contributions | |
Deal | +0.205 |
Jake | +0.183 |
Luke | +0.177 |
Thanks | +0.176 |
David | +0.174 |
Neg logit contributions | |
inside | -0.387 |
le | -0.376 |
behind | -0.373 |
pe | -0.367 |
up | -0.359 |
Token | will |
Token ID | 481 |
Feature activation | -1.19e-02 |
Pos logit contributions | |
Deal | +0.041 |
Not | +0.038 |
Okay | +0.034 |
They | +0.032 |
Something | +0.032 |
Neg logit contributions | |
le | -0.043 |
behind | -0.043 |
ling | -0.043 |
away | -0.040 |
inside | -0.040 |
Token | take |
Token ID | 1011 |
Feature activation | -4.83e-02 |
Pos logit contributions | |
le | +0.024 |
ling | +0.023 |
behind | +0.023 |
inside | +0.022 |
les | +0.021 |
Neg logit contributions | |
Deal | -0.023 |
Not | -0.019 |
Okay | -0.019 |
Thanks | -0.018 |
Jake | -0.017 |
Token | good |
Token ID | 922 |
Feature activation | -7.20e-02 |
Pos logit contributions | |
ling | +0.090 |
le | +0.088 |
behind | +0.087 |
inside | +0.081 |
pe | +0.081 |
Neg logit contributions | |
Deal | -0.093 |
Not | -0.079 |
Jake | -0.076 |
Okay | -0.074 |
Thanks | -0.073 |
Token | care |
Token ID | 1337 |
Feature activation | -8.98e-02 |
Pos logit contributions | |
ling | +0.147 |
behind | +0.146 |
le | +0.144 |
inside | +0.134 |
pe | +0.134 |
Neg logit contributions | |
Deal | -0.128 |
Not | -0.103 |
Okay | -0.100 |
Thanks | -0.095 |
Now | -0.094 |
Token | of |
Token ID | 286 |
Feature activation | -3.62e-02 |
Pos logit contributions | |
le | +0.134 |
pe | +0.127 |
ling | +0.122 |
behind | +0.121 |
egg | +0.114 |
Neg logit contributions | |
Deal | -0.204 |
Not | -0.185 |
Okay | -0.174 |
They | -0.173 |
Now | -0.170 |
Token | us |
Token ID | 514 |
Feature activation | -1.02e-01 |
Pos logit contributions | |
le | +0.069 |
ling | +0.067 |
behind | +0.064 |
inside | +0.062 |
les | +0.062 |
Neg logit contributions | |
Deal | -0.069 |
Not | -0.061 |
Jake | -0.057 |
Okay | -0.057 |
Jake | -0.055 |
Token | had |
Token ID | 550 |
Feature activation | +2.84e-02 |
Pos logit contributions | |
Deal | +0.047 |
Jake | +0.042 |
Not | +0.042 |
Thanks | +0.041 |
Something | +0.039 |
Neg logit contributions | |
le | -0.040 |
ling | -0.038 |
behind | -0.036 |
inside | -0.035 |
away | -0.035 |
Token | seen |
Token ID | 1775 |
Feature activation | -6.47e-02 |
Pos logit contributions | |
Deal | +0.059 |
Not | +0.052 |
Okay | +0.049 |
Jake | +0.049 |
Thanks | +0.049 |
Neg logit contributions | |
le | -0.053 |
ling | -0.050 |
behind | -0.048 |
inside | -0.046 |
les | -0.045 |
Token | his |
Token ID | 465 |
Feature activation | -2.81e-02 |
Pos logit contributions | |
le | +0.130 |
ling | +0.128 |
les | +0.112 |
behind | +0.111 |
an | +0.110 |
Neg logit contributions | |
Deal | -0.107 |
Not | -0.106 |
Jake | -0.102 |
Something | -0.100 |
Tyler | -0.099 |
Token | dad |
Token ID | 9955 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
le | +0.052 |
ling | +0.048 |
les | +0.045 |
behind | +0.044 |
inside | +0.043 |
Neg logit contributions | |
Deal | -0.060 |
Not | -0.052 |
Okay | -0.051 |
Something | -0.048 |
Thanks | -0.048 |
Token | use |
Token ID | 779 |
Feature activation | -6.04e-02 |
Pos logit contributions | |
le | +0.201 |
les | +0.195 |
ls | +0.191 |
ling | +0.184 |
an | +0.184 |
Neg logit contributions | |
Deal | -0.263 |
Not | -0.243 |
Jake | -0.232 |
Johnny | -0.232 |
Tyler | -0.231 |
Token | the |
Token ID | 262 |
Feature activation | -1.30e-02 |
Pos logit contributions | |
le | +0.098 |
ling | +0.093 |
ls | +0.084 |
an | +0.084 |
les | +0.081 |
Neg logit contributions | |
Deal | -0.120 |
Not | -0.115 |
Tyler | -0.111 |
Jake | -0.111 |
Ryan | -0.109 |
Token | same |
Token ID | 976 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
le | +0.025 |
ling | +0.023 |
les | +0.021 |
behind | +0.020 |
ls | +0.019 |
Neg logit contributions | |
Deal | -0.028 |
Not | -0.025 |
Something | -0.024 |
Okay | -0.024 |
Thanks | -0.023 |
Token | pick |
Token ID | 2298 |
Feature activation | -1.55e-01 |
Pos logit contributions | |
le | +0.167 |
ling | +0.160 |
ls | +0.159 |
behind | +0.148 |
les | +0.144 |
Neg logit contributions | |
Deal | -0.197 |
Something | -0.178 |
Not | -0.177 |
Okay | -0.164 |
Winning | -0.161 |
Token | before |
Token ID | 878 |
Feature activation | -1.60e-04 |
Pos logit contributions | |
ck | +0.192 |
le | +0.179 |
ling | +0.174 |
cking | +0.170 |
pe | +0.161 |
Neg logit contributions | |
She | -0.304 |
Not | -0.302 |
Deal | -0.297 |
They | -0.296 |
� | -0.290 |
Token | . |
Token ID | 13 |
Feature activation | +1.88e-01 |
Pos logit contributions | |
le | +0.000 |
ling | +0.000 |
les | +0.000 |
behind | +0.000 |
away | +0.000 |
Neg logit contributions | |
Deal | -0.000 |
Not | -0.000 |
Okay | -0.000 |
Jake | -0.000 |
Tyler | -0.000 |
Token |
|
Token ID | 198 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
Deal | +0.306 |
Jake | +0.226 |
Okay | +0.220 |
Luke | +0.218 |
Not | +0.212 |
Neg logit contributions | |
inside | -0.428 |
pe | -0.419 |
behind | -0.419 |
ling | -0.402 |
le | -0.402 |
Token | way |
Token ID | 835 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
Deal | +0.017 |
They | +0.015 |
Not | +0.015 |
Thanks | +0.015 |
Max | +0.014 |
Neg logit contributions | |
le | -0.017 |
ling | -0.015 |
les | -0.014 |
inside | -0.013 |
away | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | +1.87e-01 |
Pos logit contributions | |
le | +0.128 |
ling | +0.110 |
pe | +0.106 |
egg | +0.100 |
les | +0.100 |
Neg logit contributions | |
Deal | -0.244 |
They | -0.235 |
Not | -0.231 |
She | -0.224 |
Tyler | -0.221 |
Token | They |
Token ID | 1119 |
Feature activation | +3.43e-02 |
Pos logit contributions | |
Deal | +0.292 |
Jake | +0.243 |
Thanks | +0.239 |
Tyler | +0.224 |
Johnny | +0.224 |
Neg logit contributions | |
behind | -0.422 |
inside | -0.420 |
pe | -0.390 |
up | -0.387 |
le | -0.375 |
Token | see |
Token ID | 766 |
Feature activation | -3.10e-02 |
Pos logit contributions | |
Deal | +0.068 |
Not | +0.056 |
Okay | +0.055 |
Jake | +0.054 |
Thanks | +0.053 |
Neg logit contributions | |
le | -0.066 |
ling | -0.063 |
behind | -0.063 |
inside | -0.062 |
pe | -0.060 |
Token | cars |
Token ID | 5006 |
Feature activation | -1.69e-01 |
Pos logit contributions | |
le | +0.058 |
ling | +0.054 |
away | +0.049 |
behind | +0.049 |
les | +0.048 |
Neg logit contributions | |
Deal | -0.059 |
Not | -0.055 |
They | -0.053 |
Something | -0.053 |
Okay | -0.051 |
Token | and |
Token ID | 290 |
Feature activation | +2.04e-02 |
Pos logit contributions | |
le | +0.260 |
egg | +0.230 |
ling | +0.229 |
an | +0.200 |
les | +0.197 |
Neg logit contributions | |
They | -0.356 |
Not | -0.349 |
Deal | -0.339 |
She | -0.329 |
Tyler | -0.322 |
Token | buses |
Token ID | 16893 |
Feature activation | -1.60e-01 |
Pos logit contributions | |
Deal | +0.046 |
Not | +0.040 |
Jake | +0.039 |
Okay | +0.038 |
Thanks | +0.037 |
Neg logit contributions | |
le | -0.033 |
ling | -0.031 |
behind | -0.030 |
inside | -0.029 |
away | -0.029 |
Token | and |
Token ID | 290 |
Feature activation | +2.18e-02 |
Pos logit contributions | |
le | +0.231 |
egg | +0.228 |
ling | +0.205 |
an | +0.204 |
pearl | +0.186 |
Neg logit contributions | |
They | -0.345 |
She | -0.311 |
Not | -0.311 |
He | -0.291 |
Tyler | -0.290 |
Token | bikes |
Token ID | 16715 |
Feature activation | -1.49e-01 |
Pos logit contributions | |
Deal | +0.048 |
Not | +0.042 |
Okay | +0.040 |
Jake | +0.040 |
Something | +0.039 |
Neg logit contributions | |
le | -0.037 |
ling | -0.034 |
behind | -0.032 |
inside | -0.032 |
away | -0.031 |
Token | . |
Token ID | 13 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
egg | +0.222 |
le | +0.215 |
an | +0.202 |
ling | +0.199 |
les | +0.183 |
Neg logit contributions | |
They | -0.317 |
She | -0.289 |
Not | -0.277 |
He | -0.264 |
Tyler | -0.262 |
Token | They |
Token ID | 1119 |
Feature activation | +3.99e-02 |
Pos logit contributions | |
Deal | +0.311 |
Thanks | +0.252 |
Jake | +0.250 |
David | +0.236 |
Tyler | +0.235 |
Neg logit contributions | |
behind | -0.432 |
inside | -0.427 |
pe | -0.403 |
up | -0.395 |
le | -0.387 |
Token | jar |
Token ID | 17379 |
Feature activation | -2.13e-01 |
Pos logit contributions | |
le | +0.002 |
ling | +0.002 |
behind | +0.002 |
les | +0.002 |
inside | +0.002 |
Neg logit contributions | |
Deal | -0.003 |
Not | -0.002 |
Okay | -0.002 |
Thanks | -0.002 |
Something | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +1.97e-01 |
Pos logit contributions | |
le | +0.286 |
ling | +0.259 |
pe | +0.245 |
ck | +0.239 |
les | +0.238 |
Neg logit contributions | |
Deal | -0.467 |
She | -0.450 |
Not | -0.437 |
They | -0.433 |
Jake | -0.421 |
Token |
|
Token ID | 198 |
Feature activation | +1.65e-01 |
Pos logit contributions | |
Johnny | +0.342 |
Deal | +0.342 |
Jake | +0.309 |
Tyler | +0.305 |
David | +0.297 |
Neg logit contributions | |
inside | -0.377 |
up | -0.355 |
behind | -0.354 |
around | -0.332 |
away | -0.328 |
Token |
|
Token ID | 198 |
Feature activation | +1.65e-01 |
Pos logit contributions | |
Deal | +0.311 |
Not | +0.248 |
Okay | +0.246 |
Jake | +0.236 |
Thanks | +0.234 |
Neg logit contributions | |
le | -0.331 |
inside | -0.324 |
behind | -0.322 |
ling | -0.316 |
pe | -0.305 |
Token | When |
Token ID | 2215 |
Feature activation | +6.47e-02 |
Pos logit contributions | |
Deal | +0.253 |
Acc | +0.196 |
Okay | +0.191 |
Not | +0.190 |
Tyler | +0.186 |
Neg logit contributions | |
le | -0.386 |
inside | -0.369 |
behind | -0.369 |
ling | -0.366 |
away | -0.358 |
Token | Eva |
Token ID | 32355 |
Feature activation | -3.06e-02 |
Pos logit contributions | |
Deal | +0.122 |
Not | +0.100 |
Okay | +0.097 |
Jake | +0.095 |
Thanks | +0.094 |
Neg logit contributions | |
le | -0.132 |
ling | -0.125 |
behind | -0.125 |
inside | -0.123 |
pe | -0.120 |
Token | 's |
Token ID | 338 |
Feature activation | +3.72e-02 |
Pos logit contributions | |
ling | +0.060 |
le | +0.056 |
behind | +0.054 |
b | +0.052 |
an | +0.052 |
Neg logit contributions | |
Deal | -0.049 |
Not | -0.047 |
Jake | -0.045 |
Now | -0.043 |
Tyler | -0.043 |
Token | mom |
Token ID | 1995 |
Feature activation | -5.60e-02 |
Pos logit contributions | |
Deal | +0.062 |
Not | +0.050 |
Okay | +0.048 |
Thanks | +0.045 |
Jake | +0.045 |
Neg logit contributions | |
le | -0.086 |
ling | -0.081 |
behind | -0.079 |
inside | -0.076 |
les | -0.076 |
Token | saw |
Token ID | 2497 |
Feature activation | -2.36e-02 |
Pos logit contributions | |
le | +0.097 |
ling | +0.097 |
behind | +0.085 |
an | +0.084 |
b | +0.083 |
Neg logit contributions | |
Deal | -0.109 |
Jake | -0.104 |
Tyler | -0.098 |
Johnny | -0.098 |
Jake | -0.097 |
Token | the |
Token ID | 262 |
Feature activation | +2.47e-03 |
Pos logit contributions | |
ling | +0.044 |
le | +0.044 |
b | +0.041 |
an | +0.041 |
behind | +0.041 |
Neg logit contributions | |
Deal | -0.045 |
Not | -0.039 |
Luke | -0.036 |
Jake | -0.036 |
Kelly | -0.036 |
Token | jar |
Token ID | 17379 |
Feature activation | -2.15e-01 |
Pos logit contributions | |
Deal | +0.005 |
Not | +0.005 |
Okay | +0.005 |
Thanks | +0.004 |
Something | +0.004 |
Neg logit contributions | |
le | -0.004 |
ling | -0.004 |
behind | -0.004 |
away | -0.003 |
les | -0.003 |
Token | her |
Token ID | 607 |
Feature activation | -1.30e-02 |
Pos logit contributions | |
le | +0.082 |
ling | +0.079 |
les | +0.073 |
behind | +0.071 |
inside | +0.071 |
Neg logit contributions | |
Deal | -0.091 |
Not | -0.079 |
Okay | -0.075 |
Jake | -0.074 |
Tyler | -0.072 |
Token | sadness |
Token ID | 25303 |
Feature activation | -1.54e-01 |
Pos logit contributions | |
le | +0.023 |
ling | +0.022 |
les | +0.021 |
ls | +0.020 |
b | +0.020 |
Neg logit contributions | |
Deal | -0.028 |
Not | -0.024 |
Okay | -0.023 |
Acc | -0.022 |
Jake | -0.022 |
Token | and |
Token ID | 290 |
Feature activation | +3.12e-02 |
Pos logit contributions | |
le | +0.178 |
pe | +0.162 |
ling | +0.161 |
an | +0.149 |
les | +0.149 |
Neg logit contributions | |
Deal | -0.364 |
Not | -0.343 |
They | -0.327 |
Okay | -0.323 |
Jake | -0.319 |
Token | tears |
Token ID | 10953 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
Deal | +0.062 |
Not | +0.054 |
Okay | +0.051 |
Thanks | +0.050 |
Jake | +0.050 |
Neg logit contributions | |
le | -0.058 |
ling | -0.055 |
behind | -0.053 |
inside | -0.051 |
les | -0.051 |
Token | . |
Token ID | 13 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
an | +0.128 |
egg | +0.126 |
pe | +0.125 |
island | +0.121 |
cage | +0.121 |
Neg logit contributions | |
Deal | -0.255 |
Not | -0.249 |
Thanks | -0.247 |
They | -0.247 |
Jake | -0.247 |
Token | Billy |
Token ID | 15890 |
Feature activation | -4.65e-02 |
Pos logit contributions | |
Deal | +0.266 |
Jake | +0.235 |
Luke | +0.217 |
Kids | +0.213 |
Tyler | +0.211 |
Neg logit contributions | |
inside | -0.406 |
behind | -0.394 |
up | -0.385 |
away | -0.382 |
er | -0.373 |
Token | wanted |
Token ID | 2227 |
Feature activation | -1.20e-02 |
Pos logit contributions | |
le | +0.085 |
behind | +0.083 |
inside | +0.079 |
ling | +0.079 |
ls | +0.076 |
Neg logit contributions | |
Deal | -0.081 |
Not | -0.074 |
Jake | -0.071 |
Thanks | -0.070 |
Brother | -0.068 |
Token | to |
Token ID | 284 |
Feature activation | -1.32e-02 |
Pos logit contributions | |
le | +0.021 |
ling | +0.021 |
behind | +0.019 |
les | +0.018 |
inside | +0.018 |
Neg logit contributions | |
Deal | -0.025 |
Not | -0.023 |
Okay | -0.022 |
Jake | -0.021 |
Thanks | -0.021 |
Token | cry |
Token ID | 3960 |
Feature activation | -6.87e-02 |
Pos logit contributions | |
le | +0.025 |
ling | +0.024 |
behind | +0.023 |
inside | +0.023 |
les | +0.022 |
Neg logit contributions | |
Deal | -0.027 |
Not | -0.023 |
Okay | -0.022 |
Thanks | -0.021 |
Jake | -0.021 |
Token | , |
Token ID | 11 |
Feature activation | +6.20e-02 |
Pos logit contributions | |
le | +0.080 |
ling | +0.074 |
pe | +0.072 |
behind | +0.070 |
les | +0.069 |
Neg logit contributions | |
Deal | -0.179 |
Not | -0.162 |
Okay | -0.158 |
They | -0.154 |
Jake | -0.152 |
Token | but |
Token ID | 475 |
Feature activation | +6.71e-02 |
Pos logit contributions | |
Deal | +0.146 |
Not | +0.126 |
Okay | +0.123 |
Jake | +0.120 |
Thanks | +0.119 |
Neg logit contributions | |
le | -0.097 |
ling | -0.091 |
behind | -0.089 |
inside | -0.088 |
pe | -0.084 |
Token | was |
Token ID | 373 |
Feature activation | +2.02e-02 |
Pos logit contributions | |
Deal | +0.048 |
Not | +0.042 |
Okay | +0.041 |
Jake | +0.041 |
Thanks | +0.040 |
Neg logit contributions | |
le | -0.031 |
ling | -0.029 |
behind | -0.029 |
inside | -0.028 |
pe | -0.027 |
Token | a |
Token ID | 257 |
Feature activation | +2.54e-02 |
Pos logit contributions | |
Deal | +0.052 |
Okay | +0.043 |
Not | +0.042 |
Jake | +0.041 |
Sorry | +0.040 |
Neg logit contributions | |
inside | -0.027 |
behind | -0.027 |
le | -0.027 |
pe | -0.026 |
ling | -0.024 |
Token | little |
Token ID | 1310 |
Feature activation | -2.68e-02 |
Pos logit contributions | |
Deal | +0.058 |
Not | +0.050 |
Okay | +0.049 |
Jake | +0.049 |
Thanks | +0.048 |
Neg logit contributions | |
le | -0.042 |
ling | -0.038 |
behind | -0.037 |
inside | -0.036 |
les | -0.036 |
Token | bunny |
Token ID | 44915 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
le | +0.053 |
ling | +0.050 |
behind | +0.049 |
les | +0.048 |
aches | +0.047 |
Neg logit contributions | |
Deal | -0.046 |
Okay | -0.038 |
Not | -0.038 |
Jake | -0.037 |
Acc | -0.036 |
Token | named |
Token ID | 3706 |
Feature activation | -9.14e-03 |
Pos logit contributions | |
le | +0.265 |
ling | +0.247 |
les | +0.247 |
away | +0.242 |
behind | +0.240 |
Neg logit contributions | |
Not | -0.174 |
Deal | -0.173 |
They | -0.173 |
She | -0.160 |
Jake | -0.157 |
Token | Bob |
Token ID | 5811 |
Feature activation | -8.36e-02 |
Pos logit contributions | |
le | +0.021 |
inside | +0.020 |
behind | +0.020 |
ling | +0.020 |
pe | +0.019 |
Neg logit contributions | |
Deal | -0.015 |
Not | -0.011 |
Okay | -0.011 |
Jake | -0.011 |
Thanks | -0.011 |
Token | o |
Token ID | 78 |
Feature activation | -9.03e-02 |
Pos logit contributions | |
le | +0.100 |
behind | +0.091 |
ling | +0.089 |
les | +0.085 |
inside | +0.084 |
Neg logit contributions | |
Deal | -0.206 |
Not | -0.189 |
They | -0.184 |
Okay | -0.184 |
Jake | -0.180 |
Token | . |
Token ID | 13 |
Feature activation | +1.88e-01 |
Pos logit contributions | |
le | +0.096 |
behind | +0.083 |
ling | +0.081 |
inside | +0.078 |
pe | +0.076 |
Neg logit contributions | |
Deal | -0.232 |
Not | -0.220 |
They | -0.214 |
Jake | -0.209 |
Okay | -0.209 |
Token | Bob |
Token ID | 5811 |
Feature activation | -3.50e-02 |
Pos logit contributions | |
Deal | +0.401 |
Jake | +0.342 |
Thanks | +0.329 |
Johnny | +0.322 |
Luke | +0.317 |
Neg logit contributions | |
behind | -0.310 |
pe | -0.309 |
inside | -0.300 |
le | -0.299 |
ling | -0.291 |
Token | o |
Token ID | 78 |
Feature activation | -5.59e-02 |
Pos logit contributions | |
le | +0.081 |
behind | +0.077 |
ling | +0.077 |
inside | +0.076 |
away | +0.074 |
Neg logit contributions | |
Deal | -0.055 |
Not | -0.044 |
Okay | -0.042 |
Jake | -0.041 |
Thanks | -0.040 |
Token | loved |
Token ID | 6151 |
Feature activation | -6.95e-03 |
Pos logit contributions | |
le | +0.109 |
behind | +0.103 |
ling | +0.103 |
inside | +0.101 |
away | +0.097 |
Neg logit contributions | |
Deal | -0.106 |
Not | -0.091 |
Okay | -0.086 |
Jake | -0.085 |
Thanks | -0.085 |
Token | and |
Token ID | 290 |
Feature activation | +2.45e-02 |
Pos logit contributions | |
le | +0.079 |
ling | +0.074 |
pe | +0.069 |
behind | +0.069 |
inside | +0.065 |
Neg logit contributions | |
Deal | -0.218 |
Not | -0.199 |
Okay | -0.193 |
Jake | -0.189 |
Thanks | -0.188 |
Token | learned |
Token ID | 4499 |
Feature activation | -2.37e-03 |
Pos logit contributions | |
Deal | +0.053 |
Not | +0.043 |
Okay | +0.042 |
Jake | +0.042 |
Thanks | +0.040 |
Neg logit contributions | |
le | -0.044 |
inside | -0.042 |
ling | -0.042 |
behind | -0.041 |
pe | -0.040 |
Token | many |
Token ID | 867 |
Feature activation | -5.18e-02 |
Pos logit contributions | |
le | +0.004 |
ling | +0.003 |
behind | +0.003 |
inside | +0.003 |
pe | +0.003 |
Neg logit contributions | |
Deal | -0.006 |
Not | -0.005 |
Okay | -0.005 |
Jake | -0.005 |
Thanks | -0.005 |
Token | new |
Token ID | 649 |
Feature activation | -6.83e-02 |
Pos logit contributions | |
ling | +0.087 |
le | +0.087 |
hatch | +0.087 |
bank | +0.086 |
egg | +0.086 |
Neg logit contributions | |
Deal | -0.083 |
Not | -0.082 |
Winning | -0.076 |
Jake | -0.074 |
Math | -0.073 |
Token | things |
Token ID | 1243 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
le | +0.094 |
ling | +0.089 |
behind | +0.087 |
inside | +0.079 |
away | +0.079 |
Neg logit contributions | |
Deal | -0.164 |
Not | -0.152 |
Jake | -0.142 |
Okay | -0.141 |
They | -0.141 |
Token | together |
Token ID | 1978 |
Feature activation | -4.46e-02 |
Pos logit contributions | |
le | +0.186 |
ling | +0.183 |
egg | +0.174 |
an | +0.163 |
pe | +0.161 |
Neg logit contributions | |
Deal | -0.313 |
Not | -0.301 |
They | -0.287 |
Okay | -0.280 |
She | -0.275 |
Token | . |
Token ID | 13 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
ling | +0.055 |
le | +0.054 |
egg | +0.051 |
an | +0.049 |
les | +0.048 |
Neg logit contributions | |
Deal | -0.103 |
Not | -0.103 |
They | -0.099 |
Thanks | -0.093 |
Okay | -0.092 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +5.05e-02 |
Pos logit contributions | |
Deal | +0.480 |
Jake | +0.379 |
Tyler | +0.369 |
David | +0.368 |
Yeah | +0.366 |
Neg logit contributions | |
inside | -0.287 |
on | -0.270 |
. | -0.256 |
behind | -0.249 |
in | -0.245 |
Token | Once |
Token ID | 7454 |
Feature activation | +6.78e-02 |
Pos logit contributions | |
Deal | +0.065 |
Not | +0.050 |
Jake | +0.047 |
Okay | +0.047 |
Thanks | +0.044 |
Neg logit contributions | |
le | -0.131 |
ling | -0.127 |
behind | -0.123 |
inside | -0.121 |
pe | -0.121 |
Token | upon |
Token ID | 2402 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
Deal | +0.132 |
Not | +0.108 |
Okay | +0.105 |
Jake | +0.103 |
Thanks | +0.103 |
Neg logit contributions | |
le | -0.133 |
ling | -0.128 |
behind | -0.127 |
inside | -0.126 |
pe | -0.121 |
Token | a |
Token ID | 257 |
Feature activation | +3.87e-02 |
Pos logit contributions | |
Deal | +0.257 |
Okay | +0.210 |
Not | +0.210 |
Jake | +0.207 |
Tyler | +0.205 |
Neg logit contributions | |
le | -0.148 |
behind | -0.141 |
inside | -0.140 |
ling | -0.133 |
pe | -0.132 |
Token | race |
Token ID | 3234 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
Deal | +0.007 |
Not | +0.006 |
Okay | +0.005 |
Thanks | +0.005 |
Jake | +0.005 |
Neg logit contributions | |
le | -0.007 |
ling | -0.006 |
behind | -0.006 |
inside | -0.006 |
away | -0.006 |
Token | too |
Token ID | 1165 |
Feature activation | -2.28e-02 |
Pos logit contributions | |
le | +0.188 |
ling | +0.175 |
egg | +0.153 |
les | +0.147 |
pe | +0.147 |
Neg logit contributions | |
Deal | -0.471 |
Not | -0.435 |
Jake | -0.423 |
Okay | -0.420 |
Tyler | -0.418 |
Token | . |
Token ID | 13 |
Feature activation | +1.87e-01 |
Pos logit contributions | |
le | +0.028 |
ling | +0.027 |
les | +0.025 |
an | +0.022 |
egg | +0.022 |
Neg logit contributions | |
Deal | -0.059 |
Not | -0.055 |
Okay | -0.053 |
They | -0.053 |
Tyler | -0.051 |
Token | His |
Token ID | 2399 |
Feature activation | +4.61e-02 |
Pos logit contributions | |
Deal | +0.329 |
Okay | +0.242 |
Luke | +0.242 |
David | +0.237 |
Jake | +0.236 |
Neg logit contributions | |
up | -0.412 |
inside | -0.406 |
behind | -0.401 |
pe | -0.397 |
around | -0.385 |
Token | mom |
Token ID | 1995 |
Feature activation | -5.36e-02 |
Pos logit contributions | |
Deal | +0.094 |
Not | +0.078 |
Okay | +0.075 |
Jake | +0.075 |
Thanks | +0.074 |
Neg logit contributions | |
behind | -0.082 |
inside | -0.082 |
le | -0.081 |
ling | -0.080 |
pe | -0.079 |
Token | signed |
Token ID | 4488 |
Feature activation | -6.67e-02 |
Pos logit contributions | |
le | +0.098 |
ling | +0.092 |
behind | +0.091 |
away | +0.087 |
egg | +0.085 |
Neg logit contributions | |
Deal | -0.096 |
Jake | -0.091 |
Johnny | -0.088 |
Not | -0.087 |
Tyler | -0.086 |
Token | him |
Token ID | 683 |
Feature activation | -5.63e-02 |
Pos logit contributions | |
le | +0.116 |
ling | +0.107 |
behind | +0.100 |
les | +0.100 |
an | +0.099 |
Neg logit contributions | |
Deal | -0.131 |
Not | -0.120 |
Jake | -0.117 |
Tyler | -0.116 |
Jake | -0.116 |
Token | up |
Token ID | 510 |
Feature activation | -7.77e-02 |
Pos logit contributions | |
le | +0.082 |
ling | +0.074 |
les | +0.070 |
pe | +0.068 |
egg | +0.067 |
Neg logit contributions | |
Deal | -0.132 |
Not | -0.122 |
Jake | -0.113 |
Okay | -0.113 |
Tyler | -0.112 |
Token | and |
Token ID | 290 |
Feature activation | +3.49e-02 |
Pos logit contributions | |
le | +0.094 |
ling | +0.085 |
les | +0.079 |
inside | +0.077 |
behind | +0.075 |
Neg logit contributions | |
Deal | -0.207 |
Not | -0.184 |
Thanks | -0.179 |
Okay | -0.178 |
Jake | -0.178 |
Token | he |
Token ID | 339 |
Feature activation | +8.09e-03 |
Pos logit contributions | |
Deal | +0.067 |
Jake | +0.060 |
Tyler | +0.058 |
Okay | +0.058 |
Not | +0.058 |
Neg logit contributions | |
le | -0.067 |
ling | -0.063 |
egg | -0.058 |
les | -0.058 |
behind | -0.057 |
Token | was |
Token ID | 373 |
Feature activation | +8.22e-03 |
Pos logit contributions | |
Deal | +0.016 |
Jake | +0.014 |
Not | +0.014 |
Tyler | +0.014 |
Thanks | +0.014 |
Neg logit contributions | |
le | -0.015 |
ling | -0.014 |
behind | -0.013 |
egg | -0.013 |
away | -0.013 |
Token | saw |
Token ID | 2497 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
Deal | +0.045 |
Not | +0.040 |
Jake | +0.040 |
Thanks | +0.039 |
Okay | +0.039 |
Neg logit contributions | |
le | -0.033 |
ling | -0.030 |
behind | -0.030 |
inside | -0.029 |
away | -0.028 |
Token | a |
Token ID | 257 |
Feature activation | -3.65e-03 |
Pos logit contributions | |
le | +0.021 |
inside | +0.021 |
behind | +0.021 |
pe | +0.020 |
ling | +0.020 |
Neg logit contributions | |
Deal | -0.035 |
Not | -0.029 |
Okay | -0.029 |
Jake | -0.028 |
Thanks | -0.028 |
Token | small |
Token ID | 1402 |
Feature activation | -9.61e-02 |
Pos logit contributions | |
le | +0.006 |
ling | +0.006 |
behind | +0.005 |
les | +0.005 |
away | +0.005 |
Neg logit contributions | |
Deal | -0.008 |
Not | -0.007 |
Jake | -0.007 |
Okay | -0.007 |
Johnny | -0.007 |
Token | , |
Token ID | 11 |
Feature activation | +2.84e-02 |
Pos logit contributions | |
le | +0.161 |
ling | +0.149 |
led | +0.140 |
ck | +0.140 |
les | +0.139 |
Neg logit contributions | |
Deal | -0.175 |
Not | -0.165 |
They | -0.165 |
omo | -0.160 |
Jake | -0.159 |
Token | round |
Token ID | 2835 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
Deal | +0.060 |
Not | +0.054 |
Something | +0.051 |
Okay | +0.051 |
Jake | +0.051 |
Neg logit contributions | |
le | -0.049 |
les | -0.043 |
ling | -0.043 |
away | -0.039 |
inside | -0.038 |
Token | drain |
Token ID | 14782 |
Feature activation | -2.00e-01 |
Pos logit contributions | |
le | +0.187 |
ling | +0.175 |
behind | +0.161 |
b | +0.156 |
away | +0.155 |
Neg logit contributions | |
Deal | -0.268 |
Not | -0.250 |
They | -0.242 |
omo | -0.237 |
Jake | -0.235 |
Token | in |
Token ID | 287 |
Feature activation | -1.48e-02 |
Pos logit contributions | |
le | +0.224 |
les | +0.169 |
ling | +0.155 |
egg | +0.146 |
away | +0.139 |
Neg logit contributions | |
They | -0.474 |
Not | -0.450 |
Deal | -0.450 |
She | -0.443 |
Why | -0.441 |
Token | the |
Token ID | 262 |
Feature activation | +5.94e-03 |
Pos logit contributions | |
le | +0.028 |
ling | +0.027 |
les | +0.025 |
behind | +0.024 |
inside | +0.024 |
Neg logit contributions | |
Deal | -0.029 |
Jake | -0.025 |
Not | -0.025 |
They | -0.024 |
Okay | -0.024 |
Token | ground |
Token ID | 2323 |
Feature activation | -1.65e-01 |
Pos logit contributions | |
Deal | +0.015 |
Not | +0.013 |
Thanks | +0.013 |
Okay | +0.013 |
Jake | +0.013 |
Neg logit contributions | |
le | -0.008 |
ling | -0.007 |
behind | -0.007 |
inside | -0.007 |
les | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | +1.90e-01 |
Pos logit contributions | |
le | +0.246 |
les | +0.206 |
ling | +0.201 |
egg | +0.193 |
ls | +0.192 |
Neg logit contributions | |
They | -0.343 |
She | -0.335 |
Not | -0.333 |
Deal | -0.314 |
He | -0.312 |
Token | He |
Token ID | 679 |
Feature activation | +2.99e-02 |
Pos logit contributions | |
Deal | +0.337 |
Luke | +0.232 |
Okay | +0.231 |
Jake | +0.229 |
Thanks | +0.226 |
Neg logit contributions | |
inside | -0.427 |
behind | -0.424 |
pe | -0.423 |
up | -0.399 |
tree | -0.391 |
Token | I |
Token ID | 40 |
Feature activation | +4.79e-02 |
Pos logit contributions | |
Deal | +0.157 |
ossible | +0.108 |
Tyler | +0.108 |
Their | +0.105 |
Okay | +0.101 |
Neg logit contributions | |
le | -0.166 |
ling | -0.161 |
pe | -0.160 |
b | -0.156 |
ring | -0.151 |
Token | wish |
Token ID | 4601 |
Feature activation | -7.69e-02 |
Pos logit contributions | |
Deal | +0.092 |
Not | +0.076 |
Okay | +0.074 |
Jake | +0.072 |
Thanks | +0.071 |
Neg logit contributions | |
le | -0.096 |
ling | -0.092 |
behind | -0.092 |
inside | -0.089 |
pe | -0.087 |
Token | I |
Token ID | 314 |
Feature activation | +5.96e-03 |
Pos logit contributions | |
le | +0.205 |
ling | +0.198 |
behind | +0.193 |
inside | +0.191 |
away | +0.188 |
Neg logit contributions | |
Deal | -0.098 |
Not | -0.074 |
Okay | -0.069 |
Jake | -0.066 |
Thanks | -0.064 |
Token | had |
Token ID | 550 |
Feature activation | +1.41e-03 |
Pos logit contributions | |
Deal | +0.010 |
Not | +0.009 |
Okay | +0.009 |
Thanks | +0.009 |
Jake | +0.009 |
Neg logit contributions | |
ling | -0.012 |
le | -0.012 |
behind | -0.011 |
les | -0.011 |
away | -0.011 |
Token | a |
Token ID | 257 |
Feature activation | -1.51e-02 |
Pos logit contributions | |
Deal | +0.002 |
Not | +0.002 |
Thanks | +0.002 |
Okay | +0.002 |
Kelly | +0.002 |
Neg logit contributions | |
ling | -0.003 |
le | -0.003 |
les | -0.003 |
an | -0.002 |
behind | -0.002 |
Token | cane |
Token ID | 33009 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
le | +0.026 |
ling | +0.024 |
behind | +0.024 |
inside | +0.022 |
les | +0.022 |
Neg logit contributions | |
Deal | -0.034 |
Not | -0.030 |
Okay | -0.029 |
Thanks | -0.028 |
Jake | -0.027 |
Token | like |
Token ID | 588 |
Feature activation | -2.43e-02 |
Pos logit contributions | |
le | +0.357 |
ling | +0.356 |
les | +0.319 |
an | +0.301 |
lling | +0.298 |
Neg logit contributions | |
Deal | -0.397 |
Not | -0.363 |
Okay | -0.355 |
Bye | -0.321 |
She | -0.319 |
Token | yours |
Token ID | 12431 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
le | +0.048 |
ling | +0.046 |
behind | +0.043 |
inside | +0.042 |
away | +0.041 |
Neg logit contributions | |
Deal | -0.048 |
Not | -0.041 |
Okay | -0.039 |
Thanks | -0.038 |
Jake | -0.038 |
Token | ." |
Token ID | 526 |
Feature activation | +1.82e-01 |
Pos logit contributions | |
ling | +0.172 |
le | +0.169 |
les | +0.148 |
egg | +0.140 |
lers | +0.137 |
Neg logit contributions | |
Deal | -0.286 |
Not | -0.259 |
Okay | -0.257 |
They | -0.241 |
She | -0.238 |
Token | Tim |
Token ID | 5045 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
Deal | +0.389 |
Jake | +0.285 |
Luke | +0.284 |
Okay | +0.272 |
Not | +0.272 |
Neg logit contributions | |
inside | -0.349 |
up | -0.329 |
behind | -0.326 |
pe | -0.322 |
ling | -0.309 |
Token | my |
Token ID | 1820 |
Feature activation | -7.00e-03 |
Pos logit contributions | |
le | +0.025 |
ling | +0.024 |
behind | +0.023 |
inside | +0.023 |
pe | +0.022 |
Neg logit contributions | |
Deal | -0.030 |
Not | -0.026 |
Okay | -0.025 |
Jake | -0.024 |
Thanks | -0.024 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
le | +0.003 |
ling | +0.003 |
behind | +0.003 |
inside | +0.003 |
pe | +0.002 |
Neg logit contributions | |
Deal | -0.010 |
Not | -0.009 |
Okay | -0.008 |
Jake | -0.008 |
Thanks | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | +1.76e-01 |
Pos logit contributions | |
Deal | +0.288 |
Jake | +0.206 |
Okay | +0.200 |
Not | +0.199 |
Luke | +0.197 |
Neg logit contributions | |
inside | -0.408 |
behind | -0.406 |
away | -0.395 |
le | -0.389 |
pe | -0.388 |
Token |
|
Token ID | 198 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
Deal | +0.338 |
Not | +0.267 |
Okay | +0.256 |
Sorry | +0.242 |
Acc | +0.240 |
Neg logit contributions | |
le | -0.349 |
ling | -0.343 |
inside | -0.342 |
behind | -0.336 |
les | -0.319 |
Token | They |
Token ID | 2990 |
Feature activation | +3.56e-02 |
Pos logit contributions | |
Deal | +0.291 |
Not | +0.221 |
Okay | +0.209 |
Acc | +0.203 |
Something | +0.192 |
Neg logit contributions | |
le | -0.404 |
ling | -0.403 |
inside | -0.389 |
behind | -0.385 |
away | -0.379 |
Token | played |
Token ID | 2826 |
Feature activation | -4.22e-02 |
Pos logit contributions | |
Deal | +0.063 |
Jake | +0.053 |
Not | +0.052 |
Thanks | +0.052 |
Okay | +0.051 |
Neg logit contributions | |
le | -0.073 |
ling | -0.070 |
behind | -0.068 |
inside | -0.066 |
egg | -0.065 |
Token | golf |
Token ID | 13126 |
Feature activation | -1.64e-01 |
Pos logit contributions | |
le | +0.059 |
ling | +0.054 |
egg | +0.049 |
pe | +0.048 |
les | +0.048 |
Neg logit contributions | |
Deal | -0.104 |
Not | -0.093 |
Thanks | -0.090 |
Jake | -0.090 |
Okay | -0.089 |
Token | for |
Token ID | 329 |
Feature activation | +2.23e-03 |
Pos logit contributions | |
le | +0.217 |
egg | +0.206 |
ling | +0.201 |
inside | +0.190 |
pe | +0.188 |
Neg logit contributions | |
Deal | -0.373 |
Jake | -0.346 |
Not | -0.345 |
Thanks | -0.337 |
Luke | -0.336 |
Token | a |
Token ID | 257 |
Feature activation | +6.88e-03 |
Pos logit contributions | |
Deal | +0.005 |
Not | +0.004 |
Okay | +0.004 |
Jake | +0.004 |
Thanks | +0.004 |
Neg logit contributions | |
le | -0.004 |
ling | -0.004 |
behind | -0.003 |
inside | -0.003 |
pe | -0.003 |
Token | long |
Token ID | 890 |
Feature activation | -7.87e-02 |
Pos logit contributions | |
Deal | +0.015 |
Not | +0.013 |
Jake | +0.013 |
Thanks | +0.012 |
Okay | +0.012 |
Neg logit contributions | |
le | -0.011 |
ling | -0.011 |
behind | -0.011 |
inside | -0.010 |
pe | -0.010 |
Token | time |
Token ID | 640 |
Feature activation | -2.46e-02 |
Pos logit contributions | |
le | +0.119 |
ling | +0.113 |
behind | +0.112 |
inside | +0.111 |
pe | +0.106 |
Neg logit contributions | |
Deal | -0.181 |
Not | -0.160 |
Jake | -0.153 |
Thanks | -0.152 |
Okay | -0.149 |
Token | . |
Token ID | 13 |
Feature activation | +1.88e-01 |
Pos logit contributions | |
le | +0.036 |
ling | +0.034 |
egg | +0.032 |
les | +0.031 |
an | +0.031 |
Neg logit contributions | |
Deal | -0.055 |
Not | -0.054 |
They | -0.050 |
Okay | -0.050 |
Jake | -0.050 |
Token | ball |
Token ID | 2613 |
Feature activation | -1.89e-01 |
Pos logit contributions | |
le | +0.006 |
ling | +0.005 |
behind | +0.005 |
les | +0.005 |
inside | +0.005 |
Neg logit contributions | |
Deal | -0.008 |
Not | -0.007 |
Okay | -0.007 |
Jake | -0.007 |
Something | -0.007 |
Token | , |
Token ID | 11 |
Feature activation | +3.40e-02 |
Pos logit contributions | |
le | +0.275 |
ling | +0.265 |
les | +0.216 |
aches | +0.213 |
b | +0.213 |
Neg logit contributions | |
They | -0.410 |
Not | -0.405 |
Deal | -0.402 |
She | -0.397 |
Jake | -0.375 |
Token | and |
Token ID | 290 |
Feature activation | +2.85e-02 |
Pos logit contributions | |
Deal | +0.077 |
Not | +0.071 |
They | +0.067 |
Something | +0.067 |
Okay | +0.065 |
Neg logit contributions | |
le | -0.054 |
ling | -0.053 |
les | -0.045 |
away | -0.043 |
an | -0.043 |
Token | a |
Token ID | 257 |
Feature activation | +6.11e-03 |
Pos logit contributions | |
Deal | +0.056 |
Not | +0.049 |
Okay | +0.047 |
Jake | +0.046 |
Something | +0.045 |
Neg logit contributions | |
le | -0.055 |
ling | -0.053 |
behind | -0.048 |
les | -0.048 |
away | -0.047 |
Token | big |
Token ID | 1263 |
Feature activation | -5.96e-02 |
Pos logit contributions | |
Deal | +0.014 |
Okay | +0.012 |
Not | +0.012 |
Jake | +0.012 |
Something | +0.012 |
Neg logit contributions | |
le | -0.010 |
ling | -0.009 |
behind | -0.009 |
les | -0.008 |
inside | -0.008 |
Token | iron |
Token ID | 6953 |
Feature activation | -2.15e-01 |
Pos logit contributions | |
le | +0.097 |
ling | +0.090 |
behind | +0.081 |
les | +0.079 |
inside | +0.076 |
Neg logit contributions | |
Deal | -0.139 |
Okay | -0.119 |
Not | -0.119 |
Something | -0.113 |
Jake | -0.113 |
Token | . |
Token ID | 13 |
Feature activation | +1.90e-01 |
Pos logit contributions | |
le | +0.255 |
ling | +0.237 |
behind | +0.198 |
ck | +0.190 |
les | +0.187 |
Neg logit contributions | |
Deal | -0.501 |
They | -0.482 |
Tyler | -0.471 |
Not | -0.470 |
She | -0.461 |
Token | The |
Token ID | 383 |
Feature activation | +6.87e-02 |
Pos logit contributions | |
Deal | +0.283 |
Thanks | +0.216 |
Jake | +0.215 |
Tyler | +0.205 |
David | +0.201 |
Neg logit contributions | |
inside | -0.434 |
behind | -0.428 |
on | -0.402 |
le | -0.394 |
in | -0.394 |
Token | iron |
Token ID | 6953 |
Feature activation | -1.82e-01 |
Pos logit contributions | |
Deal | +0.170 |
Not | +0.147 |
Okay | +0.144 |
Jake | +0.141 |
Thanks | +0.140 |
Neg logit contributions | |
le | -0.104 |
ling | -0.095 |
behind | -0.092 |
inside | -0.089 |
pe | -0.085 |
Token | was |
Token ID | 373 |
Feature activation | +2.48e-03 |
Pos logit contributions | |
le | +0.326 |
ling | +0.299 |
behind | +0.290 |
away | +0.274 |
inside | +0.273 |
Neg logit contributions | |
Deal | -0.324 |
Not | -0.318 |
Jake | -0.315 |
They | -0.305 |
Tyler | -0.303 |
Token | heavy |
Token ID | 4334 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
Deal | +0.005 |
Not | +0.005 |
Okay | +0.005 |
Thanks | +0.005 |
Tyler | +0.005 |
Neg logit contributions | |
le | -0.004 |
ling | -0.004 |
les | -0.004 |
an | -0.003 |
ls | -0.003 |
Token | liked |
Token ID | 8288 |
Feature activation | +6.89e-03 |
Pos logit contributions | |
le | +0.077 |
ling | +0.071 |
inside | +0.071 |
behind | +0.071 |
pe | +0.067 |
Neg logit contributions | |
Deal | -0.156 |
Not | -0.134 |
Okay | -0.132 |
Jake | -0.129 |
Thanks | -0.128 |
Token | to |
Token ID | 284 |
Feature activation | +2.75e-03 |
Pos logit contributions | |
Deal | +0.018 |
Not | +0.015 |
Okay | +0.015 |
Jake | +0.015 |
Thanks | +0.015 |
Neg logit contributions | |
le | -0.009 |
behind | -0.009 |
inside | -0.009 |
ling | -0.008 |
pe | -0.008 |
Token | play |
Token ID | 711 |
Feature activation | -6.97e-02 |
Pos logit contributions | |
Deal | +0.007 |
Okay | +0.006 |
Not | +0.005 |
Thanks | +0.005 |
Jake | +0.005 |
Neg logit contributions | |
le | -0.004 |
pe | -0.004 |
behind | -0.004 |
inside | -0.004 |
ring | -0.004 |
Token | with |
Token ID | 351 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
le | +0.078 |
ling | +0.072 |
behind | +0.066 |
egg | +0.063 |
pe | +0.063 |
Neg logit contributions | |
Deal | -0.184 |
Not | -0.164 |
Jake | -0.161 |
Thanks | -0.161 |
They | -0.158 |
Token | his |
Token ID | 465 |
Feature activation | -2.09e-02 |
Pos logit contributions | |
le | +0.024 |
behind | +0.023 |
ling | +0.023 |
inside | +0.022 |
pe | +0.022 |
Neg logit contributions | |
Deal | -0.031 |
Not | -0.026 |
Okay | -0.026 |
Jake | -0.025 |
Thanks | -0.025 |
Token | robot |
Token ID | 9379 |
Feature activation | -1.89e-01 |
Pos logit contributions | |
le | +0.027 |
behind | +0.026 |
ling | +0.026 |
inside | +0.025 |
pe | +0.024 |
Neg logit contributions | |
Deal | -0.054 |
Not | -0.047 |
Okay | -0.046 |
Jake | -0.045 |
Thanks | -0.045 |
Token | . |
Token ID | 13 |
Feature activation | +1.79e-01 |
Pos logit contributions | |
le | +0.207 |
ling | +0.186 |
behind | +0.156 |
pe | +0.156 |
les | +0.156 |
Neg logit contributions | |
Deal | -0.494 |
Not | -0.460 |
They | -0.451 |
Their | -0.430 |
Jake | -0.430 |
Token | It |
Token ID | 632 |
Feature activation | -5.14e-02 |
Pos logit contributions | |
Deal | +0.325 |
Kelly | +0.274 |
liv | +0.253 |
Chris | +0.252 |
David | +0.249 |
Neg logit contributions | |
le | -0.365 |
. | -0.358 |
pe | -0.353 |
inside | -0.353 |
rock | -0.352 |
Token | was |
Token ID | 373 |
Feature activation | -1.72e-03 |
Pos logit contributions | |
le | +0.080 |
ling | +0.075 |
behind | +0.073 |
inside | +0.072 |
away | +0.069 |
Neg logit contributions | |
Deal | -0.120 |
Not | -0.104 |
Okay | -0.101 |
Jake | -0.101 |
Thanks | -0.100 |
Token | red |
Token ID | 2266 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
behind | +0.002 |
inside | +0.002 |
pe | +0.002 |
le | +0.002 |
ling | +0.002 |
Neg logit contributions | |
Deal | -0.005 |
Okay | -0.004 |
Not | -0.004 |
Jake | -0.004 |
Sorry | -0.004 |
Token | and |
Token ID | 290 |
Feature activation | +4.71e-03 |
Pos logit contributions | |
le | +0.145 |
ling | +0.134 |
b | +0.120 |
behind | +0.120 |
away | +0.117 |
Neg logit contributions | |
Deal | -0.277 |
Not | -0.257 |
They | -0.251 |
She | -0.238 |
Okay | -0.236 |