Token | the |
Token ID | 262 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
happened | +0.011 |
teacher | +0.010 |
exam | +0.010 |
sister | +0.010 |
problem | +0.010 |
Neg logit contributions | |
From | -0.014 |
Flo | -0.012 |
Always | -0.012 |
They | -0.012 |
Since | -0.012 |
Token | wall |
Token ID | 3355 |
Feature activation | -1.00e+00 |
Pos logit contributions | |
From | +0.044 |
Flo | +0.037 |
Always | +0.037 |
Since | +0.036 |
They | +0.036 |
Neg logit contributions | |
happened | -0.040 |
teacher | -0.037 |
exam | -0.037 |
sister | -0.036 |
problem | -0.036 |
Token | . |
Token ID | 13 |
Feature activation | +2.11e+00 |
Pos logit contributions | |
From | +0.048 |
Flo | +0.042 |
Always | +0.041 |
Since | +0.041 |
Looks | +0.040 |
Neg logit contributions | |
teacher | -0.021 |
happened | -0.021 |
exam | -0.021 |
sister | -0.020 |
problem | -0.019 |
Token | But |
Token ID | 887 |
Feature activation | -3.41e+00 |
Pos logit contributions | |
happened | +0.083 |
teacher | +0.081 |
sister | +0.078 |
exam | +0.077 |
problem | +0.075 |
Neg logit contributions | |
From | -0.063 |
Always | -0.052 |
Since | -0.050 |
Flo | -0.050 |
oc | -0.049 |
Token | , |
Token ID | 11 |
Feature activation | +4.95e+00 |
Pos logit contributions | |
From | +0.109 |
Flo | +0.087 |
Since | +0.085 |
Always | +0.083 |
The | +0.082 |
Neg logit contributions | |
happened | -0.126 |
exam | -0.122 |
teacher | -0.121 |
problem | -0.118 |
classmate | -0.117 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.08e+01 |
Pos logit contributions | |
happened | +0.148 |
exam | +0.142 |
teacher | +0.141 |
problem | +0.140 |
sister | +0.135 |
Neg logit contributions | |
From | -0.191 |
Flo | -0.159 |
Always | -0.159 |
Since | -0.157 |
They | -0.154 |
Token | no |
Token ID | 645 |
Feature activation | +2.66e+00 |
Pos logit contributions | |
principal | +0.433 |
exam | +0.409 |
teacher | +0.406 |
guest | +0.399 |
classmate | +0.397 |
Neg logit contributions | |
From | -0.283 |
Flo | -0.257 |
The | -0.220 |
Tweet | -0.219 |
Just | -0.214 |
Token | ! |
Token ID | 0 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
happened | +0.074 |
teacher | +0.071 |
exam | +0.069 |
sister | +0.068 |
classmate | +0.066 |
Neg logit contributions | |
From | -0.110 |
Flo | -0.094 |
Always | -0.093 |
They | -0.093 |
Since | -0.093 |
Token | The |
Token ID | 383 |
Feature activation | -9.43e+00 |
Pos logit contributions | |
happened | +0.059 |
teacher | +0.058 |
exam | +0.056 |
sister | +0.056 |
problem | +0.055 |
Neg logit contributions | |
From | -0.043 |
Always | -0.035 |
Flo | -0.034 |
Since | -0.034 |
They | -0.033 |
Token | stack |
Token ID | 8931 |
Feature activation | -4.51e+00 |
Pos logit contributions | |
From | +0.325 |
The | +0.275 |
They | +0.270 |
Since | +0.268 |
Even | +0.256 |
Neg logit contributions | |
happened | -0.314 |
exam | -0.279 |
classmate | -0.275 |
teacher | -0.263 |
haircut | -0.262 |
Token | fell |
Token ID | 3214 |
Feature activation | +1.97e-01 |
Pos logit contributions | |
From | +0.170 |
Always | +0.141 |
Flo | +0.141 |
Since | +0.139 |
They | +0.139 |
Neg logit contributions | |
happened | -0.135 |
exam | -0.133 |
teacher | -0.133 |
sister | -0.130 |
classmate | -0.125 |
Token | |
Token ID | 220 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
happened | +0.066 |
teacher | +0.064 |
exam | +0.064 |
sister | +0.063 |
problem | +0.062 |
Neg logit contributions | |
From | -0.038 |
Flo | -0.029 |
Always | -0.029 |
Since | -0.028 |
They | -0.028 |
Token |
|
Token ID | 198 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
happened | +0.085 |
teacher | +0.082 |
sister | +0.079 |
exam | +0.079 |
problem | +0.077 |
Neg logit contributions | |
From | -0.060 |
Always | -0.049 |
Flo | -0.048 |
Since | -0.046 |
oc | -0.045 |
Token |
|
Token ID | 198 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
From | +0.052 |
Always | +0.043 |
Flo | +0.041 |
Since | +0.041 |
oc | +0.039 |
Neg logit contributions | |
happened | -0.075 |
teacher | -0.073 |
sister | -0.070 |
exam | -0.069 |
problem | -0.068 |
Token | " |
Token ID | 1 |
Feature activation | -8.84e-03 |
Pos logit contributions | |
From | +0.048 |
Always | +0.040 |
Flo | +0.038 |
Since | +0.037 |
oc | +0.035 |
Neg logit contributions | |
happened | -0.079 |
teacher | -0.078 |
exam | -0.076 |
sister | -0.075 |
problem | -0.074 |
Token | Uh |
Token ID | 34653 |
Feature activation | +4.11e+00 |
Pos logit contributions | |
From | +0.000 |
Always | +0.000 |
rig | +0.000 |
Since | +0.000 |
Flo | +0.000 |
Neg logit contributions | |
happened | -0.000 |
teacher | -0.000 |
exam | -0.000 |
sister | -0.000 |
classmate | -0.000 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.08e+01 |
Pos logit contributions | |
teacher | +0.126 |
exam | +0.124 |
principal | +0.123 |
classmate | +0.120 |
happened | +0.118 |
Neg logit contributions | |
From | -0.147 |
oc | -0.137 |
Flo | -0.135 |
Tweet | -0.126 |
They | -0.125 |
Token | !" |
Token ID | 2474 |
Feature activation | +4.40e+00 |
Pos logit contributions | |
principal | +0.389 |
recess | +0.382 |
exam | +0.363 |
classmate | +0.360 |
classroom | +0.351 |
Neg logit contributions | |
Flo | -0.298 |
oc | -0.292 |
From | -0.267 |
spe | -0.261 |
Tweet | -0.239 |
Token | said |
Token ID | 531 |
Feature activation | +7.95e+00 |
Pos logit contributions | |
happened | +0.139 |
exam | +0.138 |
teacher | +0.135 |
problem | +0.133 |
sister | +0.132 |
Neg logit contributions | |
From | -0.163 |
Flo | -0.134 |
They | -0.134 |
Since | -0.133 |
Always | -0.130 |
Token | Liz |
Token ID | 28313 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
problem | +0.317 |
exam | +0.315 |
happened | +0.313 |
register | +0.310 |
principal | +0.306 |
Neg logit contributions | |
From | -0.182 |
Tweet | -0.161 |
Ch | -0.158 |
The | -0.158 |
Flo | -0.157 |
Token | . |
Token ID | 13 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
problem | +0.046 |
exam | +0.046 |
happened | +0.046 |
seeker | +0.043 |
recess | +0.043 |
Neg logit contributions | |
From | -0.040 |
Flo | -0.035 |
The | -0.035 |
He | -0.034 |
They | -0.034 |
Token |
|
Token ID | 198 |
Feature activation | -2.05e+00 |
Pos logit contributions | |
happened | +0.073 |
teacher | +0.071 |
exam | +0.069 |
sister | +0.069 |
problem | +0.067 |
Neg logit contributions | |
From | -0.065 |
Always | -0.054 |
Flo | -0.053 |
Since | -0.053 |
They | -0.051 |
Token | � |
Token ID | 95 |
Feature activation | -7.32e+00 |
Pos logit contributions | |
From | +0.111 |
Always | +0.094 |
They | +0.091 |
Since | +0.091 |
Flo | +0.088 |
Neg logit contributions | |
exam | -0.125 |
teacher | -0.124 |
happened | -0.120 |
problem | -0.116 |
sister | -0.115 |
Token | € |
Token ID | 26391 |
Feature activation | +9.60e+00 |
Pos logit contributions | |
From | +0.233 |
Flo | +0.196 |
They | +0.195 |
Always | +0.192 |
Since | +0.191 |
Neg logit contributions | |
exam | -0.257 |
teacher | -0.255 |
happened | -0.252 |
principal | -0.243 |
classmate | -0.232 |
Token | � |
Token ID | 129 |
Feature activation | -8.25e+00 |
Pos logit contributions | |
happened | +0.360 |
teacher | +0.347 |
exam | +0.343 |
sister | +0.337 |
problem | +0.333 |
Neg logit contributions | |
From | -0.307 |
Always | -0.249 |
Flo | -0.249 |
Since | -0.245 |
They | -0.240 |
Token | � |
Token ID | 241 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
From | +0.249 |
Always | +0.227 |
Since | +0.216 |
Looks | +0.215 |
Flo | +0.211 |
Neg logit contributions | |
happened | -0.290 |
teacher | -0.288 |
exam | -0.286 |
problem | -0.273 |
sister | -0.256 |
Token | Uh |
Token ID | 34653 |
Feature activation | +4.08e+00 |
Pos logit contributions | |
happened | +0.063 |
teacher | +0.060 |
exam | +0.058 |
sister | +0.057 |
problem | +0.056 |
Neg logit contributions | |
From | -0.029 |
Always | -0.022 |
Flo | -0.022 |
Since | -0.021 |
oc | -0.020 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.08e+01 |
Pos logit contributions | |
happened | +0.099 |
teacher | +0.095 |
exam | +0.094 |
sister | +0.090 |
problem | +0.089 |
Neg logit contributions | |
From | -0.185 |
Flo | -0.161 |
Always | -0.159 |
They | -0.158 |
Since | -0.157 |
Token | , |
Token ID | 11 |
Feature activation | +4.84e+00 |
Pos logit contributions | |
exam | +0.341 |
principal | +0.340 |
teacher | +0.330 |
classmate | +0.318 |
happened | +0.315 |
Neg logit contributions | |
From | -0.367 |
Flo | -0.338 |
The | -0.300 |
oc | -0.298 |
They | -0.298 |
Token | what |
Token ID | 644 |
Feature activation | +6.78e-01 |
Pos logit contributions | |
teacher | +0.135 |
happened | +0.134 |
exam | +0.130 |
sister | +0.129 |
problem | +0.120 |
Neg logit contributions | |
From | -0.199 |
Always | -0.179 |
Flo | -0.177 |
oc | -0.173 |
rig | -0.172 |
Token | should |
Token ID | 815 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
exam | +0.019 |
classmate | +0.019 |
teacher | +0.018 |
guest | +0.018 |
principal | +0.018 |
Neg logit contributions | |
From | -0.025 |
Always | -0.021 |
They | -0.021 |
The | -0.020 |
There | -0.019 |
Token | I |
Token ID | 314 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
happened | +0.051 |
exam | +0.050 |
classmate | +0.050 |
problem | +0.050 |
teacher | +0.050 |
Neg logit contributions | |
From | -0.031 |
Flo | -0.023 |
Since | -0.022 |
The | -0.022 |
Always | -0.022 |
Token | do |
Token ID | 466 |
Feature activation | +1.97e+00 |
Pos logit contributions | |
From | +0.049 |
Always | +0.042 |
Since | +0.041 |
Flo | +0.040 |
Even | +0.040 |
Neg logit contributions | |
happened | -0.038 |
teacher | -0.036 |
sister | -0.036 |
principal | -0.036 |
classmate | -0.035 |
Token | fox |
Token ID | 21831 |
Feature activation | -9.42e-01 |
Pos logit contributions | |
From | +0.366 |
Always | +0.311 |
Since | +0.310 |
Flo | +0.309 |
They | +0.308 |
Neg logit contributions | |
happened | -0.282 |
teacher | -0.278 |
exam | -0.275 |
problem | -0.270 |
sister | -0.267 |
Token | said |
Token ID | 531 |
Feature activation | +7.88e+00 |
Pos logit contributions | |
From | +0.040 |
Always | +0.036 |
Since | +0.036 |
Flo | +0.035 |
Even | +0.035 |
Neg logit contributions | |
happened | -0.025 |
teacher | -0.023 |
sister | -0.022 |
problem | -0.022 |
exam | -0.020 |
Token | , |
Token ID | 11 |
Feature activation | +4.78e+00 |
Pos logit contributions | |
exam | +0.236 |
principal | +0.214 |
teacher | +0.210 |
classmate | +0.209 |
recess | +0.202 |
Neg logit contributions | |
From | -0.313 |
Flo | -0.274 |
From | -0.262 |
The | -0.255 |
He | -0.253 |
Token | " |
Token ID | 366 |
Feature activation | -3.52e+00 |
Pos logit contributions | |
happened | +0.119 |
teacher | +0.113 |
sister | +0.106 |
problem | +0.101 |
exam | +0.099 |
Neg logit contributions | |
From | -0.206 |
Always | -0.192 |
Flo | -0.183 |
Even | -0.181 |
Since | -0.181 |
Token | Uh |
Token ID | 34653 |
Feature activation | +4.00e+00 |
Pos logit contributions | |
rig | +0.111 |
igs | +0.109 |
From | +0.105 |
deeply | +0.103 |
From | +0.103 |
Neg logit contributions | |
happened | -0.108 |
ha | -0.093 |
teacher | -0.091 |
guy | -0.086 |
arrived | -0.082 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.07e+01 |
Pos logit contributions | |
happened | +0.107 |
teacher | +0.107 |
exam | +0.106 |
principal | +0.102 |
classmate | +0.100 |
Neg logit contributions | |
From | -0.162 |
Flo | -0.149 |
oc | -0.144 |
They | -0.141 |
Tweet | -0.137 |
Token | , |
Token ID | 11 |
Feature activation | +4.82e+00 |
Pos logit contributions | |
principal | +0.363 |
exam | +0.353 |
recess | +0.344 |
teacher | +0.337 |
classmate | +0.336 |
Neg logit contributions | |
Flo | -0.320 |
From | -0.312 |
oc | -0.310 |
Eagles | -0.269 |
Squ | -0.267 |
Token | I |
Token ID | 314 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
happened | +0.138 |
sister | +0.134 |
teacher | +0.133 |
exam | +0.127 |
problem | +0.127 |
Neg logit contributions | |
From | -0.198 |
Always | -0.181 |
Since | -0.172 |
They | -0.172 |
Even | -0.170 |
Token | shouldn |
Token ID | 6584 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
From | +0.052 |
Always | +0.044 |
Flo | +0.044 |
Since | +0.043 |
They | +0.043 |
Neg logit contributions | |
happened | -0.043 |
teacher | -0.041 |
exam | -0.041 |
sister | -0.040 |
problem | -0.039 |
Token | 't |
Token ID | 470 |
Feature activation | -3.63e-02 |
Pos logit contributions | |
happened | +0.004 |
teacher | +0.004 |
exam | +0.004 |
sister | +0.004 |
problem | +0.004 |
Neg logit contributions | |
From | -0.004 |
Flo | -0.004 |
Always | -0.004 |
Since | -0.004 |
They | -0.003 |
Token | stay |
Token ID | 2652 |
Feature activation | -1.92e+00 |
Pos logit contributions | |
From | +0.001 |
Flo | +0.001 |
Always | +0.001 |
They | +0.001 |
Since | +0.001 |
Neg logit contributions | |
happened | -0.001 |
teacher | -0.001 |
exam | -0.001 |
sister | -0.001 |
problem | -0.001 |
Token | |
Token ID | 220 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
happened | +0.091 |
teacher | +0.088 |
exam | +0.086 |
sister | +0.086 |
problem | +0.084 |
Neg logit contributions | |
From | -0.071 |
Always | -0.058 |
Flo | -0.057 |
Since | -0.057 |
They | -0.055 |
Token |
|
Token ID | 198 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
happened | +0.086 |
teacher | +0.082 |
sister | +0.080 |
exam | +0.079 |
problem | +0.078 |
Neg logit contributions | |
From | -0.062 |
Always | -0.050 |
Flo | -0.050 |
Since | -0.048 |
oc | -0.046 |
Token |
|
Token ID | 198 |
Feature activation | -1.75e+00 |
Pos logit contributions | |
From | +0.048 |
Always | +0.039 |
Flo | +0.038 |
Since | +0.037 |
Looks | +0.035 |
Neg logit contributions | |
happened | -0.068 |
teacher | -0.066 |
sister | -0.063 |
exam | -0.063 |
problem | -0.062 |
Token | " |
Token ID | 1 |
Feature activation | -2.50e-03 |
Pos logit contributions | |
From | +0.056 |
Always | +0.052 |
valleys | +0.051 |
Flo | +0.051 |
Eagles | +0.051 |
Neg logit contributions | |
teacher | -0.056 |
happened | -0.055 |
problem | -0.052 |
sister | -0.052 |
exam | -0.051 |
Token | Uh |
Token ID | 34653 |
Feature activation | +4.15e+00 |
Pos logit contributions | |
From | +0.000 |
Flo | +0.000 |
oc | +0.000 |
rig | +0.000 |
From | +0.000 |
Neg logit contributions | |
happened | -0.000 |
teacher | -0.000 |
exam | -0.000 |
sister | -0.000 |
problem | -0.000 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.07e+01 |
Pos logit contributions | |
teacher | +0.127 |
principal | +0.125 |
happened | +0.124 |
classmate | +0.123 |
exam | +0.122 |
Neg logit contributions | |
From | -0.148 |
They | -0.128 |
oc | -0.127 |
Flo | -0.124 |
Tweet | -0.121 |
Token | , |
Token ID | 11 |
Feature activation | +4.83e+00 |
Pos logit contributions | |
principal | +0.403 |
recess | +0.386 |
classmate | +0.368 |
exam | +0.368 |
classroom | +0.364 |
Neg logit contributions | |
Flo | -0.243 |
oc | -0.235 |
spe | -0.232 |
From | -0.223 |
Chick | -0.207 |
Token | we |
Token ID | 356 |
Feature activation | -2.85e-01 |
Pos logit contributions | |
happened | +0.154 |
exam | +0.146 |
teacher | +0.145 |
problem | +0.141 |
sister | +0.138 |
Neg logit contributions | |
From | -0.181 |
Flo | -0.150 |
Since | -0.149 |
Always | -0.147 |
They | -0.147 |
Token | have |
Token ID | 423 |
Feature activation | -3.32e+00 |
Pos logit contributions | |
From | +0.010 |
Always | +0.008 |
Since | +0.008 |
Flo | +0.008 |
They | +0.008 |
Neg logit contributions | |
happened | -0.010 |
teacher | -0.009 |
exam | -0.009 |
sister | -0.009 |
problem | -0.009 |
Token | a |
Token ID | 257 |
Feature activation | +4.34e+00 |
Pos logit contributions | |
From | +0.128 |
Always | +0.117 |
They | +0.111 |
The | +0.110 |
There | +0.110 |
Neg logit contributions | |
teacher | -0.088 |
exam | -0.086 |
classmate | -0.084 |
happened | -0.082 |
sister | -0.081 |
Token | problem |
Token ID | 1917 |
Feature activation | +7.17e+00 |
Pos logit contributions | |
happened | +0.117 |
teacher | +0.108 |
exam | +0.107 |
classmate | +0.106 |
classmates | +0.105 |
Neg logit contributions | |
From | -0.149 |
Even | -0.136 |
fet | -0.132 |
Flo | -0.131 |
oc | -0.131 |
Token | |
Token ID | 220 |
Feature activation | +1.97e+00 |
Pos logit contributions | |
happened | +0.081 |
teacher | +0.080 |
sister | +0.077 |
exam | +0.076 |
problem | +0.073 |
Neg logit contributions | |
From | -0.056 |
Always | -0.047 |
Flo | -0.046 |
oc | -0.045 |
Since | -0.044 |
Token |
|
Token ID | 198 |
Feature activation | -1.99e+00 |
Pos logit contributions | |
happened | +0.075 |
teacher | +0.071 |
sister | +0.068 |
exam | +0.067 |
problem | +0.065 |
Neg logit contributions | |
From | -0.060 |
Always | -0.050 |
Flo | -0.049 |
oc | -0.048 |
Since | -0.047 |
Token |
|
Token ID | 198 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
From | +0.060 |
Always | +0.052 |
Flo | +0.050 |
Looks | +0.049 |
Since | +0.049 |
Neg logit contributions | |
happened | -0.073 |
teacher | -0.071 |
sister | -0.066 |
exam | -0.065 |
problem | -0.064 |
Token | But |
Token ID | 1537 |
Feature activation | -2.90e+00 |
Pos logit contributions | |
oc | +0.040 |
Eagles | +0.040 |
pill | +0.039 |
valleys | +0.038 |
From | +0.037 |
Neg logit contributions | |
teacher | -0.089 |
happened | -0.083 |
exam | -0.079 |
sister | -0.078 |
classmate | -0.078 |
Token | , |
Token ID | 11 |
Feature activation | +4.79e+00 |
Pos logit contributions | |
From | +0.064 |
Since | +0.044 |
Flo | +0.043 |
They | +0.042 |
The | +0.041 |
Neg logit contributions | |
happened | -0.138 |
exam | -0.137 |
teacher | -0.132 |
principal | -0.130 |
problem | -0.130 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.07e+01 |
Pos logit contributions | |
happened | +0.163 |
exam | +0.159 |
teacher | +0.154 |
problem | +0.152 |
classmate | +0.148 |
Neg logit contributions | |
From | -0.167 |
Since | -0.136 |
They | -0.135 |
Always | -0.134 |
Flo | -0.133 |
Token | no |
Token ID | 645 |
Feature activation | +2.59e+00 |
Pos logit contributions | |
principal | +0.418 |
exam | +0.402 |
teacher | +0.391 |
classmate | +0.383 |
recess | +0.361 |
Neg logit contributions | |
From | -0.282 |
Flo | -0.262 |
Whoever | -0.227 |
Even | -0.224 |
Since | -0.216 |
Token | ! |
Token ID | 0 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
happened | +0.069 |
teacher | +0.067 |
exam | +0.065 |
sister | +0.064 |
classmate | +0.062 |
Neg logit contributions | |
From | -0.110 |
Flo | -0.094 |
Always | -0.094 |
They | -0.094 |
Since | -0.093 |
Token | The |
Token ID | 383 |
Feature activation | -9.67e+00 |
Pos logit contributions | |
happened | +0.045 |
teacher | +0.044 |
sister | +0.043 |
exam | +0.043 |
problem | +0.042 |
Neg logit contributions | |
From | -0.034 |
Always | -0.027 |
Flo | -0.027 |
Since | -0.027 |
They | -0.026 |
Token | ink |
Token ID | 16882 |
Feature activation | -2.10e+00 |
Pos logit contributions | |
From | +0.367 |
Always | +0.307 |
Since | +0.303 |
Flo | +0.303 |
They | +0.301 |
Neg logit contributions | |
happened | -0.303 |
exam | -0.288 |
teacher | -0.284 |
sister | -0.277 |
problem | -0.268 |
Token | was |
Token ID | 373 |
Feature activation | +3.98e+00 |
Pos logit contributions | |
From | +0.083 |
Flo | +0.071 |
Always | +0.070 |
They | +0.070 |
Since | +0.069 |
Neg logit contributions | |
happened | -0.058 |
exam | -0.058 |
teacher | -0.057 |
sister | -0.056 |
problem | -0.054 |
Token | |
Token ID | 220 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
happened | +0.076 |
teacher | +0.074 |
sister | +0.071 |
problem | +0.066 |
dad | +0.065 |
Neg logit contributions | |
From | -0.070 |
Always | -0.062 |
Flo | -0.059 |
Since | -0.058 |
Looks | -0.058 |
Token |
|
Token ID | 198 |
Feature activation | -1.87e+00 |
Pos logit contributions | |
happened | +0.083 |
teacher | +0.079 |
sister | +0.076 |
exam | +0.075 |
problem | +0.074 |
Neg logit contributions | |
From | -0.060 |
Always | -0.049 |
Flo | -0.048 |
Since | -0.047 |
oc | -0.045 |
Token |
|
Token ID | 198 |
Feature activation | -1.92e+00 |
Pos logit contributions | |
From | +0.058 |
Always | +0.049 |
Flo | +0.048 |
Since | +0.047 |
Looks | +0.046 |
Neg logit contributions | |
happened | -0.068 |
teacher | -0.065 |
sister | -0.061 |
exam | -0.060 |
problem | -0.060 |
Token | But |
Token ID | 1537 |
Feature activation | -2.88e+00 |
Pos logit contributions | |
pill | +0.057 |
valleys | +0.053 |
ender | +0.051 |
pollen | +0.049 |
Eagles | +0.048 |
Neg logit contributions | |
happened | -0.065 |
teacher | -0.065 |
problem | -0.058 |
school | -0.055 |
sister | -0.055 |
Token | , |
Token ID | 11 |
Feature activation | +4.86e+00 |
Pos logit contributions | |
From | +0.082 |
Always | +0.062 |
Flo | +0.062 |
Since | +0.062 |
They | +0.062 |
Neg logit contributions | |
happened | -0.119 |
exam | -0.116 |
teacher | -0.116 |
problem | -0.112 |
sister | -0.111 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.07e+01 |
Pos logit contributions | |
happened | +0.168 |
teacher | +0.161 |
exam | +0.159 |
sister | +0.157 |
problem | +0.155 |
Neg logit contributions | |
From | -0.171 |
Flo | -0.142 |
Always | -0.141 |
Since | -0.139 |
They | -0.137 |
Token | no |
Token ID | 645 |
Feature activation | +2.58e+00 |
Pos logit contributions | |
exam | +0.407 |
principal | +0.394 |
teacher | +0.382 |
guest | +0.365 |
classmate | +0.362 |
Neg logit contributions | |
From | -0.304 |
Flo | -0.268 |
oc | -0.265 |
Whoever | -0.249 |
He | -0.244 |
Token | ! |
Token ID | 0 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
happened | +0.069 |
teacher | +0.068 |
exam | +0.066 |
sister | +0.064 |
classmate | +0.062 |
Neg logit contributions | |
From | -0.108 |
Flo | -0.093 |
They | -0.092 |
Always | -0.092 |
Since | -0.091 |
Token | Tim |
Token ID | 5045 |
Feature activation | +5.73e-01 |
Pos logit contributions | |
happened | +0.053 |
teacher | +0.052 |
sister | +0.052 |
exam | +0.051 |
problem | +0.050 |
Neg logit contributions | |
From | -0.038 |
Always | -0.032 |
Flo | -0.032 |
Since | -0.031 |
Hop | -0.030 |
Token | my |
Token ID | 1820 |
Feature activation | +3.96e+00 |
Pos logit contributions | |
happened | +0.023 |
teacher | +0.022 |
exam | +0.022 |
guest | +0.021 |
haircut | +0.021 |
Neg logit contributions | |
From | -0.016 |
Always | -0.014 |
They | -0.012 |
oc | -0.012 |
Flo | -0.012 |
Token | accidentally |
Token ID | 14716 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
happened | +0.120 |
exam | +0.119 |
teacher | +0.116 |
sister | +0.112 |
problem | +0.110 |
Neg logit contributions | |
From | -0.151 |
Always | -0.128 |
Flo | -0.126 |
They | -0.124 |
Since | -0.124 |
Token | them |
Token ID | 606 |
Feature activation | +3.09e+00 |
Pos logit contributions | |
From | +0.068 |
Flo | +0.058 |
Always | +0.057 |
Since | +0.057 |
Even | +0.055 |
Neg logit contributions | |
happened | -0.055 |
teacher | -0.053 |
sister | -0.052 |
exam | -0.052 |
problem | -0.051 |
Token | . |
Token ID | 13 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
happened | +0.081 |
teacher | +0.076 |
exam | +0.075 |
sister | +0.074 |
problem | +0.073 |
Neg logit contributions | |
From | -0.135 |
Always | -0.116 |
Flo | -0.116 |
Since | -0.115 |
They | -0.112 |
Token |
|
Token ID | 198 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
happened | +0.089 |
teacher | +0.087 |
sister | +0.084 |
exam | +0.084 |
problem | +0.082 |
Neg logit contributions | |
From | -0.056 |
Flo | -0.045 |
Always | -0.044 |
Since | -0.042 |
oc | -0.040 |
Token |
|
Token ID | 198 |
Feature activation | -2.00e+00 |
Pos logit contributions | |
From | +0.051 |
Flo | +0.040 |
Always | +0.040 |
Since | +0.039 |
They | +0.037 |
Neg logit contributions | |
happened | -0.085 |
teacher | -0.083 |
exam | -0.081 |
sister | -0.080 |
problem | -0.080 |
Token | But |
Token ID | 1537 |
Feature activation | -2.85e+00 |
Pos logit contributions | |
From | +0.059 |
Flo | +0.054 |
Always | +0.052 |
Bug | +0.050 |
oc | +0.048 |
Neg logit contributions | |
happened | -0.074 |
teacher | -0.073 |
sister | -0.067 |
classmate | -0.067 |
problem | -0.067 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.07e+01 |
Pos logit contributions | |
From | +0.070 |
Since | +0.050 |
They | +0.050 |
Always | +0.048 |
From | +0.046 |
Neg logit contributions | |
exam | -0.128 |
happened | -0.127 |
teacher | -0.123 |
problem | -0.121 |
principal | -0.118 |
Token | dear |
Token ID | 13674 |
Feature activation | +9.44e-01 |
Pos logit contributions | |
exam | +0.415 |
principal | +0.402 |
teacher | +0.382 |
guest | +0.360 |
classmate | +0.354 |
Neg logit contributions | |
From | -0.277 |
Whoever | -0.242 |
Flo | -0.232 |
From | -0.213 |
They | -0.213 |
Token | , |
Token ID | 11 |
Feature activation | +4.74e+00 |
Pos logit contributions | |
exam | +0.029 |
happened | +0.026 |
problem | +0.025 |
haircut | +0.025 |
principal | +0.025 |
Neg logit contributions | |
From | -0.029 |
They | -0.027 |
They | -0.026 |
Since | -0.024 |
He | -0.024 |
Token | the |
Token ID | 262 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
happened | +0.150 |
teacher | +0.141 |
sister | +0.141 |
exam | +0.134 |
problem | +0.131 |
Neg logit contributions | |
From | -0.185 |
Flo | -0.171 |
Always | -0.161 |
oc | -0.160 |
Bug | -0.158 |
Token | stream |
Token ID | 4269 |
Feature activation | -7.32e+00 |
Pos logit contributions | |
From | +0.042 |
Flo | +0.035 |
Always | +0.035 |
They | +0.034 |
Since | +0.033 |
Neg logit contributions | |
happened | -0.048 |
exam | -0.045 |
teacher | -0.044 |
classmate | -0.043 |
sister | -0.043 |
Token | was |
Token ID | 373 |
Feature activation | +4.05e+00 |
Pos logit contributions | |
From | +0.261 |
Flo | +0.218 |
They | +0.216 |
Always | +0.215 |
Since | +0.214 |
Neg logit contributions | |
exam | -0.231 |
teacher | -0.222 |
happened | -0.222 |
sister | -0.219 |
problem | -0.209 |
Token | and |
Token ID | 290 |
Feature activation | -7.88e+00 |
Pos logit contributions | |
From | +0.122 |
Flo | +0.103 |
Always | +0.101 |
Since | +0.099 |
They | +0.097 |
Neg logit contributions | |
happened | -0.125 |
teacher | -0.121 |
exam | -0.119 |
sister | -0.117 |
problem | -0.115 |
Token | said |
Token ID | 531 |
Feature activation | +7.85e+00 |
Pos logit contributions | |
From | +0.305 |
They | +0.259 |
Always | +0.255 |
Since | +0.252 |
Flo | +0.249 |
Neg logit contributions | |
exam | -0.237 |
teacher | -0.229 |
happened | -0.226 |
problem | -0.223 |
sister | -0.218 |
Token | , |
Token ID | 11 |
Feature activation | +4.82e+00 |
Pos logit contributions | |
exam | +0.224 |
classmate | +0.216 |
teacher | +0.210 |
guest | +0.210 |
seeker | +0.205 |
Neg logit contributions | |
From | -0.292 |
They | -0.249 |
From | -0.247 |
They | -0.244 |
Flo | -0.241 |
Token | " |
Token ID | 366 |
Feature activation | -3.59e+00 |
Pos logit contributions | |
happened | +0.138 |
teacher | +0.128 |
sister | +0.121 |
exam | +0.119 |
problem | +0.119 |
Neg logit contributions | |
From | -0.200 |
Always | -0.181 |
Flo | -0.178 |
Since | -0.172 |
oc | -0.170 |
Token | Uh |
Token ID | 34653 |
Feature activation | +4.11e+00 |
Pos logit contributions | |
From | +0.098 |
rig | +0.097 |
Flo | +0.091 |
ender | +0.090 |
grate | +0.090 |
Neg logit contributions | |
happened | -0.141 |
teacher | -0.115 |
problem | -0.114 |
exam | -0.113 |
ha | -0.108 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.07e+01 |
Pos logit contributions | |
teacher | +0.125 |
exam | +0.120 |
happened | +0.120 |
classmate | +0.119 |
principal | +0.119 |
Neg logit contributions | |
From | -0.149 |
They | -0.135 |
oc | -0.127 |
Flo | -0.124 |
Always | -0.122 |
Token | , |
Token ID | 11 |
Feature activation | +4.91e+00 |
Pos logit contributions | |
principal | +0.374 |
exam | +0.363 |
teacher | +0.350 |
classmate | +0.346 |
recess | +0.345 |
Neg logit contributions | |
From | -0.296 |
oc | -0.272 |
Flo | -0.259 |
They | -0.253 |
He | -0.252 |
Token | you |
Token ID | 345 |
Feature activation | +2.81e+00 |
Pos logit contributions | |
happened | +0.166 |
teacher | +0.155 |
sister | +0.154 |
exam | +0.152 |
problem | +0.149 |
Neg logit contributions | |
From | -0.177 |
Always | -0.157 |
Flo | -0.154 |
Since | -0.148 |
oc | -0.147 |
Token | boys |
Token ID | 6510 |
Feature activation | +3.41e+00 |
Pos logit contributions | |
happened | +0.099 |
exam | +0.096 |
teacher | +0.095 |
classmate | +0.090 |
problem | +0.089 |
Neg logit contributions | |
From | -0.093 |
Since | -0.075 |
They | -0.075 |
Always | -0.075 |
Even | -0.074 |
Token | made |
Token ID | 925 |
Feature activation | -4.50e-02 |
Pos logit contributions | |
teacher | +0.103 |
exam | +0.103 |
classmate | +0.101 |
happened | +0.101 |
principal | +0.095 |
Neg logit contributions | |
From | -0.111 |
They | -0.100 |
Always | -0.099 |
Even | -0.093 |
oc | -0.093 |
Token | a |
Token ID | 257 |
Feature activation | +4.22e+00 |
Pos logit contributions | |
From | +0.002 |
They | +0.001 |
Always | +0.001 |
The | +0.001 |
Even | +0.001 |
Neg logit contributions | |
happened | -0.002 |
teacher | -0.002 |
exam | -0.002 |
sister | -0.002 |
classmate | -0.002 |
Token | on |
Token ID | 319 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
happened | +0.056 |
teacher | +0.054 |
exam | +0.053 |
sister | +0.052 |
problem | +0.051 |
Neg logit contributions | |
From | -0.065 |
Flo | -0.055 |
Always | -0.055 |
Since | -0.054 |
They | -0.054 |
Token | a |
Token ID | 257 |
Feature activation | +4.13e+00 |
Pos logit contributions | |
From | +0.059 |
Always | +0.051 |
Flo | +0.050 |
Since | +0.050 |
They | +0.049 |
Neg logit contributions | |
happened | -0.043 |
teacher | -0.041 |
exam | -0.040 |
sister | -0.040 |
problem | -0.039 |
Token | plate |
Token ID | 7480 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
happened | +0.129 |
exam | +0.117 |
teacher | +0.117 |
sister | +0.116 |
problem | +0.112 |
Neg logit contributions | |
From | -0.159 |
Flo | -0.137 |
They | -0.133 |
Always | -0.131 |
Since | -0.131 |
Token | . |
Token ID | 13 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
happened | +0.035 |
exam | +0.033 |
sister | +0.032 |
teacher | +0.032 |
problem | +0.032 |
Neg logit contributions | |
From | -0.045 |
They | -0.039 |
Always | -0.038 |
Flo | -0.038 |
Since | -0.038 |
Token | But |
Token ID | 887 |
Feature activation | -3.62e+00 |
Pos logit contributions | |
happened | +0.066 |
teacher | +0.064 |
sister | +0.061 |
exam | +0.058 |
classmate | +0.055 |
Neg logit contributions | |
From | -0.063 |
Always | -0.058 |
Flo | -0.055 |
cking | -0.055 |
Since | -0.054 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.07e+01 |
Pos logit contributions | |
From | +0.119 |
Flo | +0.095 |
Since | +0.094 |
Always | +0.093 |
They | +0.092 |
Neg logit contributions | |
happened | -0.133 |
exam | -0.129 |
teacher | -0.129 |
sister | -0.123 |
problem | -0.123 |
Token | no |
Token ID | 645 |
Feature activation | +2.56e+00 |
Pos logit contributions | |
principal | +0.427 |
exam | +0.426 |
teacher | +0.410 |
classmate | +0.398 |
guest | +0.389 |
Neg logit contributions | |
From | -0.270 |
Flo | -0.247 |
oc | -0.228 |
Whoever | -0.221 |
He | -0.209 |
Token | ! |
Token ID | 0 |
Feature activation | +1.27e+00 |
Pos logit contributions | |
happened | +0.072 |
teacher | +0.072 |
exam | +0.070 |
sister | +0.068 |
classmate | +0.066 |
Neg logit contributions | |
From | -0.104 |
Flo | -0.089 |
They | -0.088 |
Always | -0.087 |
Since | -0.087 |
Token | The |
Token ID | 383 |
Feature activation | -9.60e+00 |
Pos logit contributions | |
happened | +0.047 |
teacher | +0.045 |
sister | +0.045 |
exam | +0.043 |
problem | +0.042 |
Neg logit contributions | |
From | -0.040 |
Always | -0.034 |
Flo | -0.034 |
Since | -0.033 |
They | -0.032 |
Token | toast |
Token ID | 27805 |
Feature activation | +3.13e-01 |
Pos logit contributions | |
From | +0.356 |
They | +0.310 |
The | +0.304 |
Always | +0.294 |
Since | +0.291 |
Neg logit contributions | |
happened | -0.279 |
exam | -0.272 |
classmate | -0.267 |
sister | -0.261 |
teacher | -0.260 |
Token | was |
Token ID | 373 |
Feature activation | +3.96e+00 |
Pos logit contributions | |
happened | +0.009 |
exam | +0.009 |
teacher | +0.009 |
sister | +0.009 |
problem | +0.009 |
Neg logit contributions | |
From | -0.012 |
Flo | -0.010 |
Always | -0.010 |
Since | -0.010 |
They | -0.010 |
Token | in |
Token ID | 287 |
Feature activation | -3.55e+00 |
Pos logit contributions | |
From | +0.149 |
Flo | +0.128 |
They | +0.125 |
Since | +0.121 |
Always | +0.120 |
Neg logit contributions | |
happened | -0.134 |
exam | -0.128 |
teacher | -0.127 |
sister | -0.125 |
problem | -0.121 |
Token | too |
Token ID | 1165 |
Feature activation | +3.69e+00 |
Pos logit contributions | |
From | +0.158 |
Always | +0.137 |
Flo | +0.136 |
Since | +0.135 |
They | +0.133 |
Neg logit contributions | |
happened | -0.089 |
teacher | -0.085 |
exam | -0.083 |
sister | -0.082 |
problem | -0.081 |
Token | . |
Token ID | 13 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
happened | +0.073 |
teacher | +0.069 |
exam | +0.067 |
sister | +0.065 |
problem | +0.064 |
Neg logit contributions | |
From | -0.183 |
Always | -0.161 |
Flo | -0.160 |
Since | -0.160 |
They | -0.157 |
Token | But |
Token ID | 887 |
Feature activation | -3.73e+00 |
Pos logit contributions | |
teacher | +0.057 |
happened | +0.056 |
sister | +0.056 |
exam | +0.051 |
classmate | +0.050 |
Neg logit contributions | |
From | -0.050 |
Always | -0.043 |
Since | -0.041 |
cking | -0.040 |
Looks | -0.039 |
Token | , |
Token ID | 11 |
Feature activation | +4.74e+00 |
Pos logit contributions | |
From | +0.151 |
Always | +0.129 |
Flo | +0.128 |
Since | +0.127 |
They | +0.126 |
Neg logit contributions | |
happened | -0.110 |
teacher | -0.105 |
exam | -0.103 |
sister | -0.102 |
problem | -0.100 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.06e+01 |
Pos logit contributions | |
happened | +0.151 |
sister | +0.144 |
teacher | +0.143 |
exam | +0.141 |
problem | +0.135 |
Neg logit contributions | |
From | -0.183 |
Always | -0.154 |
Since | -0.153 |
Flo | -0.152 |
They | -0.148 |
Token | no |
Token ID | 645 |
Feature activation | +2.46e+00 |
Pos logit contributions | |
principal | +0.387 |
exam | +0.375 |
teacher | +0.364 |
classmate | +0.351 |
guest | +0.331 |
Neg logit contributions | |
From | -0.325 |
Flo | -0.303 |
Tweet | -0.267 |
Whoever | -0.261 |
Squ | -0.260 |
Token | ! |
Token ID | 0 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
happened | +0.062 |
teacher | +0.060 |
exam | +0.059 |
sister | +0.057 |
classmate | +0.055 |
Neg logit contributions | |
From | -0.107 |
Flo | -0.093 |
They | -0.092 |
Always | -0.092 |
Since | -0.091 |
Token | The |
Token ID | 383 |
Feature activation | -9.75e+00 |
Pos logit contributions | |
sister | +0.037 |
happened | +0.036 |
teacher | +0.036 |
exam | +0.033 |
problem | +0.033 |
Neg logit contributions | |
From | -0.035 |
Always | -0.030 |
Since | -0.029 |
cking | -0.028 |
oc | -0.027 |
Token | box |
Token ID | 3091 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
From | +0.423 |
Since | +0.355 |
They | +0.354 |
Always | +0.353 |
Flo | +0.343 |
Neg logit contributions | |
happened | -0.258 |
exam | -0.243 |
teacher | -0.226 |
sister | -0.223 |
classmate | -0.222 |
Token | was |
Token ID | 373 |
Feature activation | +3.85e+00 |
Pos logit contributions | |
happened | +0.036 |
teacher | +0.034 |
exam | +0.034 |
sister | +0.033 |
problem | +0.033 |
Neg logit contributions | |
From | -0.048 |
Flo | -0.040 |
Always | -0.040 |
Since | -0.040 |
They | -0.039 |
Token | of |
Token ID | 286 |
Feature activation | -4.24e+00 |
Pos logit contributions | |
From | +0.039 |
Flo | +0.034 |
Always | +0.034 |
Since | +0.033 |
Bug | +0.032 |
Neg logit contributions | |
happened | -0.031 |
teacher | -0.030 |
problem | -0.028 |
sister | -0.028 |
exam | -0.028 |
Token | yogurt |
Token ID | 32132 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
From | +0.156 |
Always | +0.132 |
Flo | +0.128 |
Since | +0.127 |
They | +0.127 |
Neg logit contributions | |
happened | -0.136 |
teacher | -0.130 |
exam | -0.127 |
sister | -0.125 |
problem | -0.121 |
Token | ! |
Token ID | 0 |
Feature activation | +9.31e-01 |
Pos logit contributions | |
happened | +0.054 |
exam | +0.053 |
teacher | +0.053 |
sister | +0.053 |
problem | +0.050 |
Neg logit contributions | |
From | -0.065 |
They | -0.057 |
Always | -0.055 |
Flo | -0.055 |
Since | -0.053 |
Token | But |
Token ID | 887 |
Feature activation | -3.81e+00 |
Pos logit contributions | |
happened | +0.030 |
teacher | +0.028 |
sister | +0.028 |
exam | +0.026 |
problem | +0.025 |
Neg logit contributions | |
From | -0.033 |
Flo | -0.030 |
Always | -0.030 |
cking | -0.029 |
Hop | -0.029 |
Token | , |
Token ID | 11 |
Feature activation | +4.58e+00 |
Pos logit contributions | |
From | +0.133 |
Flo | +0.112 |
Always | +0.111 |
Since | +0.109 |
They | +0.107 |
Neg logit contributions | |
happened | -0.134 |
teacher | -0.127 |
sister | -0.125 |
exam | -0.124 |
problem | -0.123 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.06e+01 |
Pos logit contributions | |
happened | +0.145 |
sister | +0.139 |
teacher | +0.138 |
exam | +0.132 |
problem | +0.130 |
Neg logit contributions | |
From | -0.178 |
Flo | -0.155 |
Always | -0.152 |
Since | -0.152 |
pill | -0.148 |
Token | no |
Token ID | 645 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
exam | +0.393 |
principal | +0.380 |
teacher | +0.377 |
classmate | +0.361 |
guest | +0.351 |
Neg logit contributions | |
From | -0.307 |
Flo | -0.266 |
oc | -0.264 |
Whoever | -0.249 |
From | -0.244 |
Token | ! |
Token ID | 0 |
Feature activation | +9.77e-01 |
Pos logit contributions | |
happened | +0.066 |
teacher | +0.065 |
exam | +0.064 |
sister | +0.061 |
classmate | +0.060 |
Neg logit contributions | |
From | -0.099 |
Flo | -0.085 |
They | -0.085 |
Always | -0.084 |
Since | -0.084 |
Token | The |
Token ID | 383 |
Feature activation | -9.80e+00 |
Pos logit contributions | |
happened | +0.033 |
sister | +0.032 |
teacher | +0.032 |
exam | +0.030 |
problem | +0.029 |
Neg logit contributions | |
From | -0.033 |
Always | -0.028 |
cking | -0.028 |
Flo | -0.028 |
rig | -0.028 |
Token | yogurt |
Token ID | 32132 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
From | +0.374 |
There | +0.304 |
Since | +0.304 |
They | +0.303 |
This | +0.302 |
Neg logit contributions | |
happened | -0.288 |
exam | -0.259 |
classmate | -0.248 |
sister | -0.244 |
teacher | -0.238 |
Token | was |
Token ID | 373 |
Feature activation | +3.82e+00 |
Pos logit contributions | |
exam | +0.054 |
teacher | +0.053 |
happened | +0.053 |
sister | +0.052 |
classmate | +0.050 |
Neg logit contributions | |
From | -0.061 |
Flo | -0.051 |
Always | -0.051 |
They | -0.050 |
Since | -0.049 |
Token | took |
Token ID | 1718 |
Feature activation | -3.65e+00 |
Pos logit contributions | |
From | +0.305 |
Since | +0.266 |
Always | +0.264 |
Flo | +0.258 |
They | +0.258 |
Neg logit contributions | |
happened | -0.242 |
teacher | -0.232 |
sister | -0.223 |
problem | -0.210 |
exam | -0.209 |
Token | a |
Token ID | 257 |
Feature activation | +4.14e+00 |
Pos logit contributions | |
From | +0.141 |
Flo | +0.120 |
Always | +0.118 |
Since | +0.117 |
They | +0.116 |
Neg logit contributions | |
happened | -0.112 |
teacher | -0.108 |
exam | -0.107 |
sister | -0.105 |
problem | -0.103 |
Token | big |
Token ID | 1263 |
Feature activation | -2.13e+00 |
Pos logit contributions | |
happened | +0.147 |
teacher | +0.144 |
exam | +0.143 |
sister | +0.140 |
principal | +0.135 |
Neg logit contributions | |
From | -0.134 |
Flo | -0.115 |
oc | -0.108 |
They | -0.106 |
Since | -0.106 |
Token | bite |
Token ID | 13197 |
Feature activation | +1.00e-01 |
Pos logit contributions | |
From | +0.080 |
Flo | +0.067 |
Always | +0.067 |
Since | +0.066 |
They | +0.065 |
Neg logit contributions | |
happened | -0.068 |
teacher | -0.065 |
exam | -0.064 |
sister | -0.063 |
problem | -0.062 |
Token | and |
Token ID | 290 |
Feature activation | -7.86e+00 |
Pos logit contributions | |
happened | +0.004 |
teacher | +0.004 |
exam | +0.004 |
sister | +0.003 |
problem | +0.003 |
Neg logit contributions | |
From | -0.003 |
Always | -0.003 |
Flo | -0.003 |
Since | -0.003 |
Even | -0.003 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.06e+01 |
Pos logit contributions | |
From | +0.288 |
Always | +0.243 |
Flo | +0.241 |
Since | +0.239 |
They | +0.236 |
Neg logit contributions | |
happened | -0.260 |
teacher | -0.246 |
exam | -0.241 |
sister | -0.239 |
problem | -0.234 |
Token | , |
Token ID | 11 |
Feature activation | +4.75e+00 |
Pos logit contributions | |
exam | +0.375 |
principal | +0.366 |
teacher | +0.357 |
classmate | +0.340 |
happened | +0.333 |
Neg logit contributions | |
From | -0.326 |
Flo | -0.294 |
Always | -0.259 |
They | -0.257 |
From | -0.257 |
Token | it |
Token ID | 340 |
Feature activation | +1.78e+00 |
Pos logit contributions | |
teacher | +0.128 |
happened | +0.127 |
sister | +0.125 |
exam | +0.121 |
problem | +0.112 |
Neg logit contributions | |
From | -0.197 |
Always | -0.177 |
rig | -0.177 |
Bug | -0.171 |
Hop | -0.171 |
Token | tasted |
Token ID | 29187 |
Feature activation | +2.22e+00 |
Pos logit contributions | |
happened | +0.055 |
problem | +0.049 |
teacher | +0.049 |
sister | +0.048 |
exam | +0.047 |
Neg logit contributions | |
From | -0.072 |
Since | -0.061 |
Always | -0.060 |
They | -0.060 |
Flo | -0.059 |
Token | so |
Token ID | 523 |
Feature activation | +5.61e+00 |
Pos logit contributions | |
happened | +0.071 |
teacher | +0.066 |
exam | +0.065 |
problem | +0.065 |
sister | +0.064 |
Neg logit contributions | |
From | -0.084 |
Always | -0.070 |
Since | -0.070 |
Flo | -0.070 |
They | -0.067 |
Token | good |
Token ID | 922 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
happened | +0.151 |
problem | +0.129 |
teacher | +0.123 |
sister | +0.117 |
exam | +0.112 |
Neg logit contributions | |
From | -0.237 |
Always | -0.222 |
oc | -0.220 |
Since | -0.213 |
rig | -0.213 |
Token | l |
Token ID | 300 |
Feature activation | -2.79e+00 |
Pos logit contributions | |
From | +0.279 |
Since | +0.241 |
Flo | +0.238 |
Always | +0.234 |
They | +0.233 |
Neg logit contributions | |
happened | -0.216 |
teacher | -0.199 |
sister | -0.194 |
exam | -0.186 |
problem | -0.182 |
Token | icked |
Token ID | 9484 |
Feature activation | -2.31e+00 |
Pos logit contributions | |
From | +0.101 |
Always | +0.085 |
Flo | +0.083 |
They | +0.083 |
Since | +0.082 |
Neg logit contributions | |
happened | -0.092 |
exam | -0.090 |
teacher | -0.089 |
problem | -0.086 |
sister | -0.085 |
Token | the |
Token ID | 262 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
From | +0.102 |
Always | +0.095 |
Flo | +0.090 |
Since | +0.090 |
pill | +0.090 |
Neg logit contributions | |
sister | -0.056 |
happened | -0.055 |
teacher | -0.055 |
exam | -0.048 |
problem | -0.047 |
Token | cube |
Token ID | 23441 |
Feature activation | -3.84e-01 |
Pos logit contributions | |
From | +0.043 |
Always | +0.034 |
Since | +0.034 |
They | +0.034 |
oc | +0.034 |
Neg logit contributions | |
happened | -0.040 |
exam | -0.038 |
teacher | -0.038 |
problem | -0.035 |
sister | -0.035 |
Token | and |
Token ID | 290 |
Feature activation | -7.63e+00 |
Pos logit contributions | |
From | +0.014 |
Since | +0.012 |
Always | +0.012 |
Flo | +0.011 |
Hop | +0.011 |
Neg logit contributions | |
sister | -0.012 |
happened | -0.012 |
teacher | -0.012 |
problem | -0.012 |
exam | -0.012 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.06e+01 |
Pos logit contributions | |
From | +0.286 |
Always | +0.245 |
Flo | +0.243 |
Since | +0.241 |
They | +0.236 |
Neg logit contributions | |
happened | -0.251 |
teacher | -0.235 |
sister | -0.230 |
exam | -0.222 |
problem | -0.219 |
Token | ! |
Token ID | 0 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
exam | +0.384 |
principal | +0.367 |
teacher | +0.362 |
classmate | +0.344 |
happened | +0.331 |
Neg logit contributions | |
From | -0.321 |
Flo | -0.290 |
oc | -0.271 |
They | -0.261 |
Always | -0.260 |
Token | What |
Token ID | 1867 |
Feature activation | -2.87e-01 |
Pos logit contributions | |
teacher | +0.054 |
happened | +0.054 |
sister | +0.052 |
classmate | +0.048 |
exam | +0.048 |
Neg logit contributions | |
From | -0.044 |
cking | -0.039 |
Always | -0.039 |
Flo | -0.037 |
rig | -0.037 |
Token | a |
Token ID | 257 |
Feature activation | +4.26e+00 |
Pos logit contributions | |
From | +0.010 |
They | +0.009 |
Always | +0.008 |
The | +0.008 |
oc | +0.008 |
Neg logit contributions | |
exam | -0.010 |
babys | -0.009 |
teacher | -0.009 |
principal | -0.009 |
classmate | -0.009 |
Token | taste |
Token ID | 6938 |
Feature activation | -7.68e-01 |
Pos logit contributions | |
happened | +0.136 |
exam | +0.124 |
teacher | +0.121 |
sister | +0.117 |
principal | +0.114 |
Neg logit contributions | |
From | -0.159 |
oc | -0.132 |
Flo | -0.128 |
Always | -0.127 |
They | -0.126 |
Token | ! |
Token ID | 0 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
From | +0.030 |
Always | +0.025 |
They | +0.025 |
Flo | +0.025 |
Since | +0.025 |
Neg logit contributions | |
teacher | -0.021 |
sister | -0.021 |
exam | -0.021 |
happened | -0.020 |
classmate | -0.020 |
Token | 's |
Token ID | 338 |
Feature activation | +4.03e+00 |
Pos logit contributions | |
happened | +0.083 |
teacher | +0.081 |
sister | +0.081 |
exam | +0.080 |
problem | +0.077 |
Neg logit contributions | |
From | -0.124 |
Since | -0.105 |
Always | -0.105 |
Flo | -0.102 |
Even | -0.102 |
Token | hair |
Token ID | 4190 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
happened | +0.101 |
exam | +0.093 |
teacher | +0.091 |
problem | +0.087 |
sister | +0.086 |
Neg logit contributions | |
From | -0.179 |
Flo | -0.159 |
Always | -0.155 |
They | -0.153 |
Since | -0.152 |
Token | . |
Token ID | 13 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
From | +0.047 |
Always | +0.041 |
Flo | +0.040 |
Since | +0.040 |
They | +0.039 |
Neg logit contributions | |
happened | -0.031 |
teacher | -0.029 |
exam | -0.029 |
sister | -0.028 |
problem | -0.028 |
Token | But |
Token ID | 887 |
Feature activation | -3.79e+00 |
Pos logit contributions | |
teacher | +0.060 |
happened | +0.060 |
sister | +0.059 |
exam | +0.057 |
problem | +0.056 |
Neg logit contributions | |
From | -0.044 |
Flo | -0.038 |
Always | -0.037 |
Since | -0.036 |
oc | -0.035 |
Token | , |
Token ID | 11 |
Feature activation | +4.67e+00 |
Pos logit contributions | |
From | +0.137 |
Always | +0.112 |
Since | +0.111 |
Flo | +0.111 |
They | +0.109 |
Neg logit contributions | |
happened | -0.127 |
teacher | -0.122 |
exam | -0.121 |
sister | -0.117 |
problem | -0.116 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.06e+01 |
Pos logit contributions | |
happened | +0.150 |
teacher | +0.142 |
exam | +0.142 |
sister | +0.137 |
problem | +0.137 |
Neg logit contributions | |
From | -0.174 |
Always | -0.145 |
Since | -0.144 |
Flo | -0.143 |
They | -0.142 |
Token | no |
Token ID | 645 |
Feature activation | +2.51e+00 |
Pos logit contributions | |
principal | +0.393 |
exam | +0.375 |
teacher | +0.370 |
classmate | +0.360 |
guest | +0.356 |
Neg logit contributions | |
From | -0.309 |
Flo | -0.271 |
Tweet | -0.244 |
Even | -0.237 |
Always | -0.236 |
Token | ! |
Token ID | 0 |
Feature activation | +8.80e-01 |
Pos logit contributions | |
teacher | +0.071 |
happened | +0.070 |
exam | +0.069 |
classmate | +0.069 |
principal | +0.067 |
Neg logit contributions | |
From | -0.099 |
They | -0.084 |
Flo | -0.083 |
Always | -0.082 |
Since | -0.082 |
Token | The |
Token ID | 383 |
Feature activation | -9.96e+00 |
Pos logit contributions | |
happened | +0.031 |
teacher | +0.031 |
sister | +0.030 |
exam | +0.030 |
problem | +0.029 |
Neg logit contributions | |
From | -0.029 |
Always | -0.024 |
Flo | -0.024 |
Since | -0.024 |
They | -0.023 |
Token | shampoo |
Token ID | 47259 |
Feature activation | -3.57e+00 |
Pos logit contributions | |
From | +0.376 |
Always | +0.324 |
Flo | +0.315 |
Since | +0.314 |
They | +0.311 |
Neg logit contributions | |
happened | -0.306 |
exam | -0.269 |
classmate | -0.259 |
problem | -0.247 |
principal | -0.245 |
Token | got |
Token ID | 1392 |
Feature activation | +3.80e+00 |
Pos logit contributions | |
From | +0.132 |
Flo | +0.113 |
Always | +0.112 |
They | +0.111 |
Since | +0.109 |
Neg logit contributions | |
happened | -0.105 |
exam | -0.105 |
teacher | -0.103 |
sister | -0.100 |
problem | -0.100 |
Token | pretty |
Token ID | 2495 |
Feature activation | -4.48e+00 |
Pos logit contributions | |
happened | +0.180 |
principal | +0.170 |
grade | +0.154 |
classmate | +0.154 |
exam | +0.152 |
Neg logit contributions | |
From | -0.211 |
Always | -0.200 |
refreshing | -0.190 |
Flo | -0.189 |
On | -0.189 |
Token | dress |
Token ID | 6576 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
Flo | +0.111 |
From | +0.108 |
Always | +0.104 |
bur | +0.102 |
which | +0.099 |
Neg logit contributions | |
principal | -0.149 |
happened | -0.145 |
liar | -0.134 |
classmate | -0.130 |
register | -0.130 |
Token | . |
Token ID | 13 |
Feature activation | +2.17e+00 |
Pos logit contributions | |
From | +0.049 |
They | +0.044 |
Always | +0.043 |
Flo | +0.042 |
On | +0.041 |
Neg logit contributions | |
happened | -0.049 |
sister | -0.048 |
problem | -0.046 |
teacher | -0.046 |
exam | -0.046 |
Token | But |
Token ID | 887 |
Feature activation | -3.55e+00 |
Pos logit contributions | |
happened | +0.093 |
teacher | +0.090 |
exam | +0.088 |
sister | +0.087 |
problem | +0.084 |
Neg logit contributions | |
From | -0.058 |
Always | -0.046 |
Flo | -0.045 |
Since | -0.044 |
They | -0.043 |
Token | , |
Token ID | 11 |
Feature activation | +4.89e+00 |
Pos logit contributions | |
From | +0.120 |
Flo | +0.096 |
Since | +0.095 |
Always | +0.093 |
The | +0.090 |
Neg logit contributions | |
happened | -0.124 |
teacher | -0.123 |
exam | -0.122 |
classmate | -0.119 |
problem | -0.118 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.06e+01 |
Pos logit contributions | |
happened | +0.150 |
problem | +0.147 |
teacher | +0.146 |
exam | +0.146 |
sister | +0.141 |
Neg logit contributions | |
From | -0.181 |
Flo | -0.149 |
Always | -0.149 |
Since | -0.148 |
The | -0.145 |
Token | no |
Token ID | 645 |
Feature activation | +2.62e+00 |
Pos logit contributions | |
principal | +0.424 |
exam | +0.397 |
teacher | +0.393 |
classmate | +0.387 |
guest | +0.376 |
Neg logit contributions | |
From | -0.275 |
Flo | -0.261 |
Tweet | -0.219 |
oc | -0.216 |
The | -0.209 |
Token | ! |
Token ID | 0 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
teacher | +0.079 |
happened | +0.077 |
classmate | +0.076 |
sister | +0.075 |
principal | +0.073 |
Neg logit contributions | |
From | -0.099 |
Flo | -0.084 |
Always | -0.082 |
oc | -0.082 |
They | -0.081 |
Token | Her |
Token ID | 2332 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
happened | +0.060 |
teacher | +0.058 |
exam | +0.058 |
sister | +0.057 |
problem | +0.056 |
Neg logit contributions | |
From | -0.039 |
Always | -0.031 |
Flo | -0.030 |
Since | -0.030 |
They | -0.029 |
Token | dress |
Token ID | 6576 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
From | +0.067 |
Flo | +0.056 |
From | +0.055 |
The | +0.054 |
grace | +0.054 |
Neg logit contributions | |
happened | -0.060 |
principal | -0.058 |
exam | -0.057 |
register | -0.055 |
liar | -0.054 |
Token | was |
Token ID | 373 |
Feature activation | +4.06e+00 |
Pos logit contributions | |
From | +0.048 |
Flo | +0.040 |
Always | +0.038 |
They | +0.037 |
The | +0.037 |
Neg logit contributions | |
exam | -0.053 |
teacher | -0.052 |
sister | -0.052 |
principal | -0.051 |
classmate | -0.051 |
Token | the |
Token ID | 262 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
From | +0.157 |
Always | +0.140 |
rig | +0.137 |
Flo | +0.136 |
Since | +0.135 |
Neg logit contributions | |
happened | -0.086 |
teacher | -0.082 |
sister | -0.076 |
exam | -0.075 |
problem | -0.075 |
Token | distance |
Token ID | 5253 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
From | +0.041 |
Always | +0.033 |
Flo | +0.033 |
Since | +0.032 |
They | +0.032 |
Neg logit contributions | |
happened | -0.049 |
teacher | -0.047 |
exam | -0.046 |
sister | -0.046 |
problem | -0.045 |
Token | . |
Token ID | 13 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
happened | +0.038 |
teacher | +0.036 |
exam | +0.035 |
sister | +0.034 |
problem | +0.034 |
Neg logit contributions | |
From | -0.076 |
Always | -0.067 |
Flo | -0.066 |
Since | -0.066 |
Even | -0.065 |
Token | " |
Token ID | 366 |
Feature activation | -3.47e+00 |
Pos logit contributions | |
happened | +0.068 |
teacher | +0.063 |
sister | +0.061 |
problem | +0.061 |
exam | +0.058 |
Neg logit contributions | |
From | -0.060 |
rig | -0.053 |
Always | -0.052 |
Looks | -0.051 |
Since | -0.051 |
Token | Uh |
Token ID | 34653 |
Feature activation | +4.02e+00 |
Pos logit contributions | |
igs | +0.113 |
rig | +0.109 |
From | +0.102 |
pill | +0.101 |
Using | +0.099 |
Neg logit contributions | |
happened | -0.103 |
ha | -0.088 |
teacher | -0.086 |
lawyer | -0.081 |
exam | -0.080 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.06e+01 |
Pos logit contributions | |
teacher | +0.123 |
principal | +0.119 |
exam | +0.118 |
classmate | +0.116 |
sister | +0.111 |
Neg logit contributions | |
From | -0.153 |
Flo | -0.137 |
They | -0.132 |
Tweet | -0.131 |
oc | -0.130 |
Token | , |
Token ID | 11 |
Feature activation | +4.87e+00 |
Pos logit contributions | |
principal | +0.389 |
exam | +0.356 |
classmate | +0.352 |
recess | +0.351 |
teacher | +0.349 |
Neg logit contributions | |
Flo | -0.328 |
From | -0.318 |
oc | -0.299 |
Tweet | -0.281 |
Bug | -0.278 |
Token | that |
Token ID | 326 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
happened | +0.207 |
teacher | +0.200 |
exam | +0.199 |
sister | +0.195 |
problem | +0.193 |
Neg logit contributions | |
From | -0.132 |
Flo | -0.102 |
Always | -0.102 |
Since | -0.100 |
They | -0.098 |
Token | cloud |
Token ID | 6279 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
From | +0.041 |
Flo | +0.034 |
They | +0.033 |
Always | +0.033 |
Since | +0.033 |
Neg logit contributions | |
teacher | -0.039 |
exam | -0.038 |
happened | -0.037 |
sister | -0.037 |
classmate | -0.037 |
Token | looks |
Token ID | 3073 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
From | +0.092 |
Always | +0.078 |
Flo | +0.078 |
Since | +0.077 |
They | +0.076 |
Neg logit contributions | |
happened | -0.066 |
teacher | -0.063 |
exam | -0.062 |
sister | -0.061 |
problem | -0.060 |
Token | scary |
Token ID | 14343 |
Feature activation | +1.93e-01 |
Pos logit contributions | |
happened | +0.068 |
teacher | +0.066 |
exam | +0.065 |
sister | +0.064 |
problem | +0.062 |
Neg logit contributions | |
From | -0.075 |
Flo | -0.063 |
Always | -0.063 |
Since | -0.062 |
They | -0.061 |
Token | tigers |
Token ID | 43114 |
Feature activation | -2.90e+00 |
Pos logit contributions | |
teacher | +0.140 |
sister | +0.139 |
happened | +0.132 |
problem | +0.125 |
brother | +0.121 |
Neg logit contributions | |
From | -0.201 |
Always | -0.184 |
Even | -0.178 |
rig | -0.177 |
inning | -0.176 |
Token | , |
Token ID | 11 |
Feature activation | +4.99e+00 |
Pos logit contributions | |
From | +0.156 |
Always | +0.141 |
Since | +0.139 |
Looks | +0.138 |
Flo | +0.137 |
Neg logit contributions | |
happened | -0.048 |
teacher | -0.044 |
sister | -0.041 |
problem | -0.039 |
exam | -0.036 |
Token | and |
Token ID | 290 |
Feature activation | -7.56e+00 |
Pos logit contributions | |
happened | +0.137 |
sister | +0.137 |
teacher | +0.135 |
problem | +0.128 |
brother | +0.124 |
Neg logit contributions | |
From | -0.197 |
rig | -0.175 |
Always | -0.173 |
Since | -0.172 |
izing | -0.171 |
Token | bears |
Token ID | 13062 |
Feature activation | +7.49e-01 |
Pos logit contributions | |
From | +0.289 |
Always | +0.245 |
Since | +0.244 |
Even | +0.237 |
Flo | +0.236 |
Neg logit contributions | |
sister | -0.235 |
happened | -0.235 |
teacher | -0.232 |
exam | -0.219 |
problem | -0.216 |
Token | , |
Token ID | 11 |
Feature activation | +4.98e+00 |
Pos logit contributions | |
happened | +0.020 |
exam | +0.019 |
teacher | +0.019 |
sister | +0.018 |
problem | +0.018 |
Neg logit contributions | |
From | -0.032 |
Always | -0.027 |
They | -0.027 |
Flo | -0.027 |
Since | -0.027 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.06e+01 |
Pos logit contributions | |
teacher | +0.156 |
sister | +0.155 |
happened | +0.154 |
exam | +0.141 |
brother | +0.141 |
Neg logit contributions | |
From | -0.189 |
Even | -0.163 |
Since | -0.162 |
Always | -0.157 |
Looks | -0.157 |
Token | my |
Token ID | 616 |
Feature activation | +2.66e+00 |
Pos logit contributions | |
exam | +0.384 |
happened | +0.337 |
teacher | +0.336 |
principal | +0.333 |
classmate | +0.327 |
Neg logit contributions | |
From | -0.324 |
oc | -0.303 |
Flo | -0.300 |
They | -0.291 |
They | -0.276 |
Token | ! |
Token ID | 0 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
happened | +0.081 |
exam | +0.077 |
teacher | +0.072 |
classmate | +0.068 |
problem | +0.068 |
Neg logit contributions | |
From | -0.093 |
oc | -0.086 |
Flo | -0.084 |
They | -0.082 |
rig | -0.080 |
Token | They |
Token ID | 1119 |
Feature activation | -8.61e+00 |
Pos logit contributions | |
teacher | +0.063 |
sister | +0.062 |
happened | +0.059 |
principal | +0.057 |
exam | +0.057 |
Neg logit contributions | |
From | -0.039 |
Always | -0.032 |
cking | -0.031 |
ized | -0.031 |
izing | -0.030 |
Token | all |
Token ID | 477 |
Feature activation | -4.52e+00 |
Pos logit contributions | |
From | +0.302 |
Always | +0.249 |
Flo | +0.249 |
Since | +0.245 |
They | +0.245 |
Neg logit contributions | |
happened | -0.294 |
exam | -0.284 |
teacher | -0.284 |
sister | -0.274 |
problem | -0.273 |
Token | played |
Token ID | 2826 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
From | +0.166 |
Flo | +0.139 |
Always | +0.139 |
Since | +0.138 |
They | +0.133 |
Neg logit contributions | |
happened | -0.152 |
teacher | -0.141 |
sister | -0.138 |
exam | -0.138 |
problem | -0.135 |
Token | his |
Token ID | 465 |
Feature activation | +4.52e+00 |
Pos logit contributions | |
happened | +0.038 |
teacher | +0.036 |
exam | +0.036 |
problem | +0.035 |
sister | +0.034 |
Neg logit contributions | |
From | -0.042 |
Always | -0.035 |
They | -0.035 |
Flo | -0.034 |
Since | -0.034 |
Token | guitar |
Token ID | 10047 |
Feature activation | -4.44e+00 |
Pos logit contributions | |
happened | +0.099 |
exam | +0.082 |
principal | +0.072 |
teacher | +0.069 |
problem | +0.066 |
Neg logit contributions | |
From | -0.198 |
rig | -0.181 |
oc | -0.180 |
Flo | -0.177 |
Always | -0.176 |
Token | . |
Token ID | 13 |
Feature activation | +1.98e+00 |
Pos logit contributions | |
From | +0.167 |
They | +0.145 |
Always | +0.144 |
Flo | +0.140 |
Since | +0.137 |
Neg logit contributions | |
happened | -0.132 |
exam | -0.121 |
teacher | -0.120 |
sister | -0.117 |
problem | -0.116 |
Token | Luke |
Token ID | 11336 |
Feature activation | +4.39e+00 |
Pos logit contributions | |
happened | +0.068 |
exam | +0.066 |
problem | +0.063 |
teacher | +0.062 |
sister | +0.061 |
Neg logit contributions | |
From | -0.071 |
They | -0.058 |
Since | -0.057 |
Always | -0.057 |
Flo | -0.057 |
Token | said |
Token ID | 531 |
Feature activation | +7.82e+00 |
Pos logit contributions | |
happened | +0.121 |
teacher | +0.116 |
exam | +0.111 |
sister | +0.111 |
problem | +0.108 |
Neg logit contributions | |
From | -0.184 |
Always | -0.158 |
Flo | -0.158 |
Since | -0.156 |
They | -0.154 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.06e+01 |
Pos logit contributions | |
exam | +0.304 |
problem | +0.289 |
teacher | +0.278 |
sister | +0.276 |
happened | +0.276 |
Neg logit contributions | |
From | -0.240 |
They | -0.179 |
Flo | -0.179 |
The | -0.177 |
Since | -0.174 |
Token | yes |
Token ID | 3763 |
Feature activation | +1.57e+00 |
Pos logit contributions | |
exam | +0.396 |
principal | +0.357 |
happened | +0.352 |
teacher | +0.336 |
recess | +0.332 |
Neg logit contributions | |
From | -0.304 |
Whoever | -0.273 |
They | -0.265 |
He | -0.263 |
Flo | -0.242 |
Token | ! |
Token ID | 0 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
happened | +0.041 |
exam | +0.041 |
sister | +0.040 |
teacher | +0.039 |
classmate | +0.038 |
Neg logit contributions | |
From | -0.065 |
They | -0.057 |
Flo | -0.054 |
oc | -0.054 |
He | -0.054 |
Token | Ben |
Token ID | 3932 |
Feature activation | +2.73e+00 |
Pos logit contributions | |
happened | +0.047 |
teacher | +0.047 |
sister | +0.046 |
exam | +0.045 |
problem | +0.044 |
Neg logit contributions | |
From | -0.032 |
Flo | -0.026 |
Always | -0.026 |
Since | -0.025 |
Even | -0.024 |
Token | got |
Token ID | 1392 |
Feature activation | +3.89e+00 |
Pos logit contributions | |
happened | +0.084 |
teacher | +0.081 |
sister | +0.078 |
exam | +0.076 |
problem | +0.075 |
Neg logit contributions | |
From | -0.105 |
Flo | -0.091 |
Always | -0.090 |
Since | -0.089 |
They | -0.087 |
Token | out |
Token ID | 503 |
Feature activation | +4.87e+00 |
Pos logit contributions | |
happened | +0.141 |
exam | +0.139 |
problem | +0.132 |
teacher | +0.131 |
sister | +0.127 |
Neg logit contributions | |
From | -0.126 |
They | -0.105 |
Always | -0.103 |
From | -0.100 |
fet | -0.097 |
Token | Tim |
Token ID | 5045 |
Feature activation | +5.24e-01 |
Pos logit contributions | |
happened | +0.153 |
teacher | +0.147 |
exam | +0.145 |
sister | +0.143 |
problem | +0.140 |
Neg logit contributions | |
From | -0.174 |
Always | -0.146 |
Flo | -0.145 |
Since | -0.144 |
They | -0.142 |
Token | my |
Token ID | 1820 |
Feature activation | +3.98e+00 |
Pos logit contributions | |
happened | +0.021 |
teacher | +0.019 |
exam | +0.019 |
classmate | +0.018 |
problem | +0.018 |
Neg logit contributions | |
From | -0.015 |
Always | -0.012 |
They | -0.012 |
Since | -0.012 |
Flo | -0.012 |
Token | , |
Token ID | 11 |
Feature activation | +4.80e+00 |
Pos logit contributions | |
happened | +0.086 |
exam | +0.077 |
teacher | +0.074 |
problem | +0.067 |
classmate | +0.067 |
Neg logit contributions | |
From | -0.188 |
They | -0.162 |
Since | -0.160 |
Always | -0.160 |
From | -0.160 |
Token | " |
Token ID | 366 |
Feature activation | -3.69e+00 |
Pos logit contributions | |
happened | +0.145 |
teacher | +0.137 |
sister | +0.132 |
exam | +0.130 |
problem | +0.126 |
Neg logit contributions | |
From | -0.190 |
Always | -0.166 |
Flo | -0.166 |
Since | -0.160 |
oc | -0.158 |
Token | Uh |
Token ID | 34653 |
Feature activation | +4.14e+00 |
Pos logit contributions | |
rig | +0.108 |
From | +0.102 |
pering | +0.098 |
valleys | +0.097 |
igs | +0.094 |
Neg logit contributions | |
happened | -0.130 |
teacher | -0.117 |
ha | -0.107 |
exam | -0.107 |
classmate | -0.106 |
Token | oh |
Token ID | 11752 |
Feature activation | +1.05e+01 |
Pos logit contributions | |
teacher | +0.121 |
principal | +0.117 |
happened | +0.117 |
exam | +0.116 |
classmate | +0.115 |
Neg logit contributions | |
From | -0.154 |
They | -0.139 |
Flo | -0.132 |
oc | -0.131 |
Tweet | -0.128 |
Token | , |
Token ID | 11 |
Feature activation | +4.87e+00 |
Pos logit contributions | |
principal | +0.397 |
exam | +0.372 |
classmate | +0.358 |
teacher | +0.357 |
recess | +0.355 |
Neg logit contributions | |
From | -0.261 |
Flo | -0.246 |
Whoever | -0.232 |
run | -0.226 |
They | -0.220 |
Token | the |
Token ID | 262 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
happened | +0.166 |
teacher | +0.159 |
exam | +0.157 |
sister | +0.156 |
problem | +0.151 |
Neg logit contributions | |
From | -0.175 |
Always | -0.148 |
Flo | -0.146 |
Since | -0.144 |
They | -0.142 |
Token | roof |
Token ID | 9753 |
Feature activation | -5.23e+00 |
Pos logit contributions | |
From | +0.049 |
Flo | +0.041 |
Since | +0.040 |
Always | +0.040 |
They | +0.040 |
Neg logit contributions | |
happened | -0.051 |
exam | -0.048 |
teacher | -0.048 |
sister | -0.046 |
classmate | -0.046 |
Token | has |
Token ID | 468 |
Feature activation | -2.35e+00 |
Pos logit contributions | |
From | +0.214 |
Always | +0.182 |
Flo | +0.182 |
Since | +0.180 |
They | +0.177 |
Neg logit contributions | |
happened | -0.151 |
teacher | -0.144 |
exam | -0.142 |
sister | -0.139 |
problem | -0.138 |
Token | a |
Token ID | 257 |
Feature activation | +4.08e+00 |
Pos logit contributions | |
From | +0.092 |
Flo | +0.079 |
Always | +0.078 |
Since | +0.077 |
They | +0.077 |
Neg logit contributions | |
happened | -0.067 |
teacher | -0.066 |
exam | -0.066 |
sister | -0.064 |
problem | -0.062 |
Token | learned |
Token ID | 4499 |
Feature activation | +7.60e-01 |
Pos logit contributions | |
From | +0.375 |
Flo | +0.323 |
Always | +0.322 |
Since | +0.317 |
Even | +0.305 |
Neg logit contributions | |
happened | -0.275 |
teacher | -0.271 |
exam | -0.257 |
sister | -0.257 |
problem | -0.243 |
Token | a |
Token ID | 257 |
Feature activation | +3.65e+00 |
Pos logit contributions | |
happened | +0.022 |
teacher | +0.021 |
exam | +0.021 |
sister | +0.020 |
problem | +0.020 |
Neg logit contributions | |
From | -0.031 |
Flo | -0.027 |
Always | -0.027 |
Since | -0.026 |
oc | -0.026 |
Token | hard |
Token ID | 1327 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
happened | +0.102 |
teacher | +0.097 |
exam | +0.096 |
sister | +0.094 |
problem | +0.092 |
Neg logit contributions | |
From | -0.152 |
Flo | -0.130 |
Always | -0.130 |
Since | -0.129 |
They | -0.127 |
Token | lesson |
Token ID | 11483 |
Feature activation | +2.40e+00 |
Pos logit contributions | |
From | +0.047 |
They | +0.040 |
Flo | +0.040 |
Always | +0.040 |
Since | +0.040 |
Neg logit contributions | |
happened | -0.023 |
teacher | -0.022 |
sister | -0.022 |
exam | -0.021 |
classmate | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | +8.62e-01 |
Pos logit contributions | |
happened | +0.066 |
teacher | +0.063 |
exam | +0.062 |
sister | +0.060 |
problem | +0.059 |
Neg logit contributions | |
From | -0.101 |
Flo | -0.087 |
Always | -0.087 |
Since | -0.086 |
They | -0.084 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.38e+01 |
Pos logit contributions | |
teacher | +0.034 |
sister | +0.033 |
happened | +0.032 |
dad | +0.031 |
cousin | +0.030 |
Neg logit contributions | |
From | -0.021 |
cking | -0.021 |
ized | -0.020 |
inning | -0.020 |
Always | -0.019 |
Token | things |
Token ID | 1243 |
Feature activation | +1.31e-02 |
Pos logit contributions | |
From | +0.592 |
Flo | +0.508 |
Since | +0.507 |
Always | +0.500 |
They | +0.491 |
Neg logit contributions | |
happened | -0.390 |
exam | -0.367 |
teacher | -0.363 |
problem | -0.358 |
sister | -0.352 |
Token | that |
Token ID | 326 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
happened | +0.000 |
teacher | +0.000 |
exam | +0.000 |
problem | +0.000 |
sister | +0.000 |
Neg logit contributions | |
From | -0.001 |
Flo | -0.000 |
Since | -0.000 |
Always | -0.000 |
oc | -0.000 |
Token | are |
Token ID | 389 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
From | +0.058 |
Flo | +0.050 |
Since | +0.049 |
Always | +0.049 |
They | +0.048 |
Neg logit contributions | |
happened | -0.042 |
teacher | -0.040 |
exam | -0.039 |
sister | -0.038 |
problem | -0.037 |
Token | not |
Token ID | 407 |
Feature activation | -5.03e-01 |
Pos logit contributions | |
happened | +0.060 |
teacher | +0.057 |
problem | +0.056 |
exam | +0.054 |
sister | +0.054 |
Neg logit contributions | |
From | -0.087 |
Flo | -0.076 |
oc | -0.075 |
Always | -0.074 |
Since | -0.074 |
Token | yours |
Token ID | 12431 |
Feature activation | +3.09e+00 |
Pos logit contributions | |
From | +0.021 |
Flo | +0.018 |
Since | +0.018 |
Always | +0.018 |
They | +0.017 |
Neg logit contributions | |
happened | -0.014 |
teacher | -0.013 |
exam | -0.013 |
sister | -0.013 |
problem | -0.013 |
Token | be |
Token ID | 307 |
Feature activation | +6.58e-01 |
Pos logit contributions | |
happened | +0.138 |
teacher | +0.131 |
exam | +0.130 |
sister | +0.127 |
problem | +0.125 |
Neg logit contributions | |
From | -0.175 |
Flo | -0.148 |
Always | -0.148 |
Since | -0.146 |
They | -0.144 |
Token | healthy |
Token ID | 5448 |
Feature activation | -2.16e+00 |
Pos logit contributions | |
happened | +0.024 |
exam | +0.023 |
problem | +0.023 |
teacher | +0.021 |
principal | +0.020 |
Neg logit contributions | |
From | -0.021 |
They | -0.019 |
Always | -0.018 |
Flo | -0.017 |
The | -0.017 |
Token | and |
Token ID | 290 |
Feature activation | -8.07e+00 |
Pos logit contributions | |
From | +0.057 |
Always | +0.049 |
They | +0.049 |
oc | +0.046 |
Flo | +0.045 |
Neg logit contributions | |
happened | -0.090 |
exam | -0.084 |
teacher | -0.084 |
problem | -0.081 |
classmate | -0.081 |
Token | happy |
Token ID | 3772 |
Feature activation | +5.34e-01 |
Pos logit contributions | |
From | +0.308 |
Flo | +0.262 |
Always | +0.259 |
Since | +0.259 |
They | +0.250 |
Neg logit contributions | |
happened | -0.254 |
teacher | -0.247 |
exam | -0.242 |
sister | -0.240 |
problem | -0.233 |
Token | . |
Token ID | 13 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
happened | +0.018 |
classmate | +0.017 |
exam | +0.017 |
teacher | +0.017 |
problem | +0.016 |
Neg logit contributions | |
From | -0.017 |
They | -0.015 |
Always | -0.015 |
Flo | -0.014 |
Since | -0.014 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.38e+01 |
Pos logit contributions | |
sister | +0.039 |
teacher | +0.039 |
exam | +0.037 |
happened | +0.035 |
brother | +0.035 |
Neg logit contributions | |
From | -0.037 |
ized | -0.036 |
cking | -0.036 |
rig | -0.034 |
valleys | -0.033 |
Token | your |
Token ID | 534 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
From | +0.505 |
Always | +0.438 |
Flo | +0.434 |
They | +0.423 |
Since | +0.408 |
Neg logit contributions | |
happened | -0.424 |
teacher | -0.413 |
exam | -0.396 |
sister | -0.388 |
classmate | -0.383 |
Token | pill |
Token ID | 9582 |
Feature activation | -2.71e+00 |
Pos logit contributions | |
happened | +0.037 |
teacher | +0.035 |
classmate | +0.034 |
exam | +0.033 |
principal | +0.032 |
Neg logit contributions | |
From | -0.034 |
Flo | -0.030 |
oc | -0.029 |
Always | -0.029 |
Squ | -0.028 |
Token | is |
Token ID | 318 |
Feature activation | +3.25e+00 |
Pos logit contributions | |
From | +0.095 |
Flo | +0.079 |
Always | +0.079 |
Since | +0.078 |
They | +0.077 |
Neg logit contributions | |
happened | -0.092 |
teacher | -0.089 |
exam | -0.088 |
sister | -0.086 |
problem | -0.085 |
Token | good |
Token ID | 922 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
happened | +0.102 |
teacher | +0.098 |
exam | +0.097 |
sister | +0.094 |
problem | +0.092 |
Neg logit contributions | |
From | -0.122 |
Flo | -0.102 |
Always | -0.102 |
They | -0.101 |
Since | -0.100 |
Token | for |
Token ID | 329 |
Feature activation | -8.24e-01 |
Pos logit contributions | |
happened | +0.066 |
classmate | +0.066 |
exam | +0.064 |
teacher | +0.063 |
principal | +0.061 |
Neg logit contributions | |
From | -0.046 |
Always | -0.044 |
They | -0.043 |
They | -0.041 |
Now | -0.038 |
Token | share |
Token ID | 2648 |
Feature activation | -2.16e+00 |
Pos logit contributions | |
happened | +0.145 |
teacher | +0.135 |
exam | +0.134 |
sister | +0.133 |
problem | +0.132 |
Neg logit contributions | |
From | -0.160 |
Flo | -0.134 |
Always | -0.133 |
They | -0.132 |
Since | -0.131 |
Token | with |
Token ID | 351 |
Feature activation | -6.08e+00 |
Pos logit contributions | |
From | +0.076 |
They | +0.062 |
Flo | +0.062 |
Always | +0.062 |
Since | +0.061 |
Neg logit contributions | |
happened | -0.075 |
exam | -0.071 |
teacher | -0.070 |
sister | -0.069 |
problem | -0.068 |
Token | your |
Token ID | 534 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
From | +0.215 |
Flo | +0.179 |
Always | +0.178 |
They | +0.176 |
Since | +0.173 |
Neg logit contributions | |
happened | -0.207 |
exam | -0.195 |
teacher | -0.194 |
problem | -0.190 |
sister | -0.188 |
Token | sister |
Token ID | 6621 |
Feature activation | +7.36e+00 |
Pos logit contributions | |
happened | +0.027 |
exam | +0.022 |
classmate | +0.021 |
teacher | +0.020 |
problem | +0.020 |
Neg logit contributions | |
From | -0.039 |
Flo | -0.034 |
Always | -0.034 |
oc | -0.034 |
From | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
happened | +0.234 |
teacher | +0.215 |
exam | +0.215 |
problem | +0.202 |
classmate | +0.200 |
Neg logit contributions | |
From | -0.266 |
They | -0.235 |
Always | -0.234 |
oc | -0.231 |
Flo | -0.229 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.38e+01 |
Pos logit contributions | |
teacher | +0.041 |
sister | +0.041 |
exam | +0.038 |
cousin | +0.037 |
brother | +0.037 |
Neg logit contributions | |
From | -0.035 |
ized | -0.034 |
cking | -0.034 |
refreshing | -0.031 |
izing | -0.031 |
Token | her |
Token ID | 607 |
Feature activation | +6.07e+00 |
Pos logit contributions | |
From | +0.509 |
Always | +0.435 |
Flo | +0.430 |
They | +0.424 |
Since | +0.415 |
Neg logit contributions | |
happened | -0.430 |
teacher | -0.412 |
exam | -0.401 |
sister | -0.393 |
problem | -0.384 |
Token | dolls |
Token ID | 36062 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
happened | +0.165 |
teacher | +0.149 |
classmate | +0.146 |
principal | +0.141 |
exam | +0.140 |
Neg logit contributions | |
From | -0.228 |
Always | -0.201 |
Flo | -0.198 |
oc | -0.190 |
They | -0.188 |
Token | was |
Token ID | 373 |
Feature activation | +3.72e+00 |
Pos logit contributions | |
From | +0.051 |
Always | +0.043 |
Flo | +0.043 |
They | +0.043 |
Since | +0.042 |
Neg logit contributions | |
happened | -0.042 |
teacher | -0.040 |
exam | -0.039 |
sister | -0.039 |
problem | -0.038 |
Token | very |
Token ID | 845 |
Feature activation | +6.68e+00 |
Pos logit contributions | |
happened | +0.117 |
teacher | +0.114 |
exam | +0.114 |
sister | +0.111 |
classmate | +0.109 |
Neg logit contributions | |
From | -0.133 |
Always | -0.113 |
They | -0.111 |
Flo | -0.110 |
Since | -0.106 |
Token | bad |
Token ID | 2089 |
Feature activation | +3.42e+00 |
Pos logit contributions | |
happened | +0.209 |
exam | +0.196 |
teacher | +0.196 |
sister | +0.189 |
problem | +0.184 |
Neg logit contributions | |
From | -0.258 |
Flo | -0.215 |
They | -0.214 |
Always | -0.211 |
Since | -0.209 |
Token | that |
Token ID | 326 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
teacher | +0.141 |
exam | +0.141 |
happened | +0.139 |
sister | +0.138 |
problem | +0.128 |
Neg logit contributions | |
From | -0.186 |
Always | -0.161 |
Flo | -0.160 |
oc | -0.159 |
Since | -0.159 |
Token | 's |
Token ID | 338 |
Feature activation | +3.99e+00 |
Pos logit contributions | |
From | +0.065 |
Since | +0.056 |
Flo | +0.055 |
Always | +0.055 |
Even | +0.053 |
Neg logit contributions | |
happened | -0.041 |
teacher | -0.037 |
sister | -0.037 |
problem | -0.037 |
exam | -0.036 |
Token | not |
Token ID | 407 |
Feature activation | -3.24e-01 |
Pos logit contributions | |
happened | +0.128 |
teacher | +0.126 |
exam | +0.123 |
sister | +0.121 |
classmate | +0.120 |
Neg logit contributions | |
From | -0.145 |
Always | -0.119 |
Flo | -0.118 |
They | -0.118 |
Since | -0.117 |
Token | nice |
Token ID | 3621 |
Feature activation | +3.28e-01 |
Pos logit contributions | |
From | +0.013 |
Always | +0.011 |
They | +0.011 |
Flo | +0.011 |
Since | +0.010 |
Neg logit contributions | |
happened | -0.009 |
teacher | -0.009 |
exam | -0.009 |
sister | -0.009 |
classmate | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | +1.39e+00 |
Pos logit contributions | |
happened | +0.010 |
teacher | +0.010 |
exam | +0.010 |
classmate | +0.009 |
principal | +0.009 |
Neg logit contributions | |
From | -0.010 |
They | -0.010 |
Always | -0.009 |
oc | -0.009 |
They | -0.009 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.37e+01 |
Pos logit contributions | |
sister | +0.050 |
teacher | +0.046 |
brother | +0.043 |
exam | +0.043 |
cousin | +0.042 |
Neg logit contributions | |
From | -0.049 |
cking | -0.043 |
rig | -0.042 |
Together | -0.042 |
Always | -0.041 |
Token | things |
Token ID | 1243 |
Feature activation | +1.27e-01 |
Pos logit contributions | |
From | +0.574 |
Since | +0.488 |
Always | +0.486 |
Flo | +0.486 |
They | +0.477 |
Neg logit contributions | |
happened | -0.387 |
teacher | -0.366 |
exam | -0.365 |
sister | -0.357 |
problem | -0.354 |
Token | that |
Token ID | 326 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
happened | +0.004 |
teacher | +0.003 |
exam | +0.003 |
sister | +0.003 |
problem | +0.003 |
Neg logit contributions | |
From | -0.005 |
Flo | -0.004 |
Always | -0.004 |
Since | -0.004 |
They | -0.004 |
Token | aren |
Token ID | 3588 |
Feature activation | -7.25e-01 |
Pos logit contributions | |
From | +0.055 |
Always | +0.047 |
Flo | +0.047 |
Since | +0.047 |
They | +0.046 |
Neg logit contributions | |
happened | -0.039 |
teacher | -0.037 |
exam | -0.037 |
sister | -0.036 |
problem | -0.035 |
Token | 't |
Token ID | 470 |
Feature activation | -1.72e-02 |
Pos logit contributions | |
oc | +0.033 |
From | +0.033 |
valleys | +0.030 |
Since | +0.030 |
Flo | +0.029 |
Neg logit contributions | |
happened | -0.016 |
exam | -0.015 |
teacher | -0.014 |
problem | -0.014 |
sister | -0.012 |
Token | yours |
Token ID | 12431 |
Feature activation | +3.24e+00 |
Pos logit contributions | |
From | +0.001 |
Always | +0.001 |
Flo | +0.001 |
Since | +0.001 |
They | +0.001 |
Neg logit contributions | |
happened | -0.000 |
teacher | -0.000 |
exam | -0.000 |
sister | -0.000 |
problem | -0.000 |
Token | 's |
Token ID | 338 |
Feature activation | +3.94e+00 |
Pos logit contributions | |
classmate | +0.024 |
exam | +0.024 |
seeker | +0.023 |
problem | +0.023 |
principal | +0.023 |
Neg logit contributions | |
From | -0.013 |
They | -0.011 |
Always | -0.011 |
The | -0.010 |
Sun | -0.010 |
Token | best |
Token ID | 1266 |
Feature activation | +5.31e+00 |
Pos logit contributions | |
classmate | +0.115 |
teacher | +0.110 |
principal | +0.107 |
happened | +0.107 |
sister | +0.105 |
Neg logit contributions | |
From | -0.122 |
Always | -0.114 |
They | -0.106 |
From | -0.096 |
Now | -0.095 |
Token | for |
Token ID | 329 |
Feature activation | -7.55e-01 |
Pos logit contributions | |
principal | +0.202 |
classmate | +0.199 |
happened | +0.190 |
register | +0.183 |
babys | +0.182 |
Neg logit contributions | |
From | -0.096 |
They | -0.091 |
They | -0.088 |
Always | -0.086 |
But | -0.080 |
Token | you |
Token ID | 345 |
Feature activation | +2.72e+00 |
Pos logit contributions | |
From | +0.023 |
They | +0.019 |
Flo | +0.019 |
Always | +0.019 |
Since | +0.018 |
Neg logit contributions | |
happened | -0.029 |
teacher | -0.027 |
exam | -0.027 |
problem | -0.027 |
sister | -0.027 |
Token | . |
Token ID | 13 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
classmate | +0.094 |
exam | +0.094 |
teacher | +0.092 |
happened | +0.091 |
principal | +0.090 |
Neg logit contributions | |
From | -0.082 |
They | -0.072 |
Always | -0.070 |
From | -0.064 |
They | -0.063 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.37e+01 |
Pos logit contributions | |
sister | +0.036 |
teacher | +0.035 |
exam | +0.032 |
brother | +0.031 |
ha | +0.031 |
Neg logit contributions | |
ized | -0.042 |
cking | -0.041 |
From | -0.038 |
izing | -0.038 |
valleys | -0.037 |
Token | a |
Token ID | 257 |
Feature activation | +3.78e+00 |
Pos logit contributions | |
From | +0.522 |
Always | +0.449 |
Flo | +0.444 |
They | +0.434 |
Since | +0.429 |
Neg logit contributions | |
happened | -0.412 |
teacher | -0.401 |
exam | -0.389 |
sister | -0.380 |
problem | -0.370 |
Token | bath |
Token ID | 7837 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
happened | +0.123 |
teacher | +0.116 |
exam | +0.114 |
sister | +0.111 |
classmate | +0.109 |
Neg logit contributions | |
From | -0.132 |
Flo | -0.115 |
oc | -0.109 |
Squ | -0.109 |
They | -0.108 |
Token | is |
Token ID | 318 |
Feature activation | +3.24e+00 |
Pos logit contributions | |
From | +0.047 |
They | +0.041 |
Always | +0.041 |
Flo | +0.040 |
Since | +0.039 |
Neg logit contributions | |
happened | -0.041 |
exam | -0.038 |
teacher | -0.038 |
problem | -0.037 |
sister | -0.036 |
Token | good |
Token ID | 922 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
happened | +0.101 |
teacher | +0.097 |
exam | +0.096 |
sister | +0.094 |
problem | +0.092 |
Neg logit contributions | |
From | -0.123 |
Flo | -0.103 |
Always | -0.103 |
Since | -0.101 |
They | -0.101 |
Token | for |
Token ID | 329 |
Feature activation | -7.79e-01 |
Pos logit contributions | |
happened | +0.058 |
classmate | +0.056 |
exam | +0.055 |
teacher | +0.055 |
principal | +0.054 |
Neg logit contributions | |
From | -0.056 |
Always | -0.053 |
They | -0.050 |
They | -0.048 |
Now | -0.046 |
Token | happier |
Token ID | 23030 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
From | +0.131 |
Flo | +0.115 |
Since | +0.112 |
Always | +0.111 |
Tweet | +0.111 |
Neg logit contributions | |
happened | -0.088 |
sister | -0.078 |
teacher | -0.077 |
problem | -0.077 |
exam | -0.074 |
Token | . |
Token ID | 13 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
From | +0.004 |
Flo | +0.004 |
Always | +0.004 |
Since | +0.004 |
Even | +0.004 |
Neg logit contributions | |
happened | -0.003 |
teacher | -0.003 |
sister | -0.003 |
exam | -0.003 |
problem | -0.003 |
Token | Everyone |
Token ID | 11075 |
Feature activation | -5.74e+00 |
Pos logit contributions | |
sister | +0.044 |
teacher | +0.043 |
happened | +0.041 |
brother | +0.040 |
dad | +0.040 |
Neg logit contributions | |
cking | -0.044 |
From | -0.042 |
Tweet | -0.042 |
Flo | -0.040 |
Hop | -0.039 |
Token | agreed |
Token ID | 4987 |
Feature activation | +7.55e+00 |
Pos logit contributions | |
From | +0.227 |
Flo | +0.198 |
Always | +0.193 |
Since | +0.192 |
They | +0.187 |
Neg logit contributions | |
happened | -0.175 |
teacher | -0.163 |
sister | -0.159 |
exam | -0.158 |
problem | -0.153 |
Token | . |
Token ID | 13 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
happened | +0.238 |
teacher | +0.226 |
exam | +0.211 |
sister | +0.210 |
problem | +0.209 |
Neg logit contributions | |
From | -0.285 |
Flo | -0.261 |
Bug | -0.255 |
Tweet | -0.254 |
Always | -0.252 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.37e+01 |
Pos logit contributions | |
sister | +0.041 |
teacher | +0.039 |
dad | +0.038 |
brother | +0.038 |
mom | +0.037 |
Neg logit contributions | |
cking | -0.052 |
Tweet | -0.051 |
cod | -0.049 |
From | -0.048 |
Flo | -0.048 |
Token | the |
Token ID | 262 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
From | +0.591 |
Since | +0.508 |
Flo | +0.504 |
Always | +0.494 |
They | +0.488 |
Neg logit contributions | |
happened | -0.390 |
problem | -0.347 |
sister | -0.346 |
exam | -0.345 |
teacher | -0.342 |
Token | medicine |
Token ID | 9007 |
Feature activation | -1.85e+00 |
Pos logit contributions | |
From | +0.061 |
Always | +0.053 |
Flo | +0.053 |
Since | +0.052 |
They | +0.051 |
Neg logit contributions | |
happened | -0.041 |
teacher | -0.039 |
exam | -0.039 |
sister | -0.038 |
problem | -0.037 |
Token | felt |
Token ID | 2936 |
Feature activation | -4.55e-01 |
Pos logit contributions | |
From | +0.077 |
Always | +0.065 |
Flo | +0.065 |
Since | +0.065 |
They | +0.064 |
Neg logit contributions | |
happened | -0.052 |
teacher | -0.050 |
exam | -0.049 |
sister | -0.048 |
problem | -0.047 |
Token | nice |
Token ID | 3621 |
Feature activation | +3.03e-01 |
Pos logit contributions | |
From | +0.018 |
Flo | +0.016 |
oc | +0.016 |
Always | +0.016 |
Bug | +0.015 |
Neg logit contributions | |
happened | -0.014 |
teacher | -0.012 |
sister | -0.012 |
problem | -0.012 |
exam | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | -7.88e+00 |
Pos logit contributions | |
happened | +0.011 |
teacher | +0.010 |
exam | +0.010 |
sister | +0.010 |
problem | +0.010 |
Neg logit contributions | |
From | -0.010 |
Flo | -0.009 |
Always | -0.008 |
Since | -0.008 |
They | -0.008 |
Token | , |
Token ID | 11 |
Feature activation | +4.57e+00 |
Pos logit contributions | |
happened | +0.021 |
teacher | +0.020 |
exam | +0.019 |
problem | +0.018 |
sister | +0.018 |
Neg logit contributions | |
From | -0.022 |
Flo | -0.019 |
Always | -0.019 |
Since | -0.018 |
Bug | -0.018 |
Token | and |
Token ID | 290 |
Feature activation | -7.96e+00 |
Pos logit contributions | |
happened | +0.122 |
teacher | +0.110 |
dad | +0.108 |
sister | +0.105 |
exam | +0.101 |
Neg logit contributions | |
oc | -0.197 |
From | -0.190 |
Flo | -0.184 |
Bug | -0.181 |
Ho | -0.175 |
Token | she |
Token ID | 673 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
From | +0.316 |
Flo | +0.278 |
Always | +0.274 |
oc | +0.268 |
Since | +0.268 |
Neg logit contributions | |
happened | -0.237 |
teacher | -0.224 |
sister | -0.218 |
exam | -0.216 |
problem | -0.207 |
Token | smiled |
Token ID | 13541 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
From | +0.005 |
Always | +0.004 |
Flo | +0.004 |
oc | +0.004 |
Bug | +0.004 |
Neg logit contributions | |
happened | -0.003 |
teacher | -0.003 |
sister | -0.003 |
problem | -0.003 |
classmate | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
From | +0.007 |
Always | +0.006 |
Flo | +0.006 |
Since | +0.006 |
They | +0.006 |
Neg logit contributions | |
happened | -0.005 |
teacher | -0.005 |
exam | -0.005 |
sister | -0.005 |
problem | -0.005 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.37e+01 |
Pos logit contributions | |
teacher | +0.051 |
happened | +0.050 |
sister | +0.048 |
exam | +0.045 |
classmate | +0.045 |
Neg logit contributions | |
From | -0.042 |
cking | -0.038 |
oc | -0.038 |
Flo | -0.038 |
Always | -0.038 |
Token | a |
Token ID | 257 |
Feature activation | +3.79e+00 |
Pos logit contributions | |
From | +0.602 |
Flo | +0.517 |
Since | +0.515 |
Always | +0.511 |
They | +0.504 |
Neg logit contributions | |
happened | -0.367 |
teacher | -0.332 |
exam | -0.332 |
sister | -0.324 |
problem | -0.322 |
Token | shower |
Token ID | 14643 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
happened | +0.124 |
exam | +0.120 |
teacher | +0.119 |
sister | +0.115 |
principal | +0.111 |
Neg logit contributions | |
From | -0.136 |
Flo | -0.114 |
They | -0.112 |
Since | -0.110 |
Even | -0.109 |
Token | was |
Token ID | 373 |
Feature activation | +3.83e+00 |
Pos logit contributions | |
From | +0.068 |
Always | +0.059 |
Flo | +0.059 |
Since | +0.058 |
They | +0.058 |
Neg logit contributions | |
happened | -0.041 |
teacher | -0.039 |
exam | -0.039 |
sister | -0.038 |
problem | -0.037 |
Token | the |
Token ID | 262 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
happened | +0.121 |
teacher | +0.119 |
exam | +0.118 |
sister | +0.116 |
problem | +0.114 |
Neg logit contributions | |
From | -0.142 |
Always | -0.118 |
Flo | -0.118 |
Since | -0.117 |
They | -0.116 |
Token | best |
Token ID | 1266 |
Feature activation | +5.16e+00 |
Pos logit contributions | |
From | +0.062 |
Always | +0.053 |
Flo | +0.053 |
Since | +0.053 |
They | +0.052 |
Neg logit contributions | |
happened | -0.042 |
teacher | -0.039 |
exam | -0.039 |
sister | -0.038 |
problem | -0.037 |
Token | oral |
Token ID | 6864 |
Feature activation | -7.02e-02 |
Pos logit contributions | |
happened | +0.053 |
teacher | +0.052 |
exam | +0.052 |
problem | +0.051 |
sister | +0.050 |
Neg logit contributions | |
From | -0.077 |
Flo | -0.067 |
Always | -0.066 |
Since | -0.065 |
They | -0.065 |
Token | of |
Token ID | 286 |
Feature activation | -4.26e+00 |
Pos logit contributions | |
From | +0.003 |
Flo | +0.002 |
Always | +0.002 |
Since | +0.002 |
Bug | +0.002 |
Neg logit contributions | |
happened | -0.002 |
teacher | -0.002 |
exam | -0.002 |
problem | -0.002 |
sister | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
From | +0.171 |
Flo | +0.157 |
They | +0.151 |
Always | +0.147 |
Looks | +0.147 |
Neg logit contributions | |
happened | -0.118 |
teacher | -0.112 |
exam | -0.112 |
sister | -0.111 |
problem | -0.105 |
Token | story |
Token ID | 1621 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
From | +0.092 |
Flo | +0.083 |
They | +0.082 |
Since | +0.082 |
Always | +0.082 |
Neg logit contributions | |
happened | -0.020 |
exam | -0.019 |
teacher | -0.019 |
problem | -0.019 |
sister | -0.017 |
Token | : |
Token ID | 25 |
Feature activation | +5.75e+00 |
Pos logit contributions | |
happened | +0.025 |
exam | +0.024 |
teacher | +0.023 |
sister | +0.022 |
problem | +0.022 |
Neg logit contributions | |
From | -0.059 |
Since | -0.051 |
They | -0.051 |
Always | -0.051 |
Flo | -0.051 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.37e+01 |
Pos logit contributions | |
sister | +0.185 |
teacher | +0.169 |
brother | +0.167 |
exam | +0.153 |
problem | +0.152 |
Neg logit contributions | |
Twe | -0.187 |
ized | -0.185 |
From | -0.185 |
refreshing | -0.185 |
rig | -0.185 |
Token | regular |
Token ID | 3218 |
Feature activation | -2.30e+00 |
Pos logit contributions | |
From | +0.594 |
Since | +0.518 |
Flo | +0.505 |
They | +0.493 |
Always | +0.492 |
Neg logit contributions | |
happened | -0.377 |
exam | -0.351 |
problem | -0.350 |
sister | -0.342 |
teacher | -0.338 |
Token | n |
Token ID | 299 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
From | +0.098 |
Since | +0.084 |
Flo | +0.083 |
Always | +0.083 |
They | +0.082 |
Neg logit contributions | |
happened | -0.065 |
exam | -0.061 |
teacher | -0.060 |
problem | -0.059 |
sister | -0.059 |
Token | aps |
Token ID | 1686 |
Feature activation | -7.98e+00 |
Pos logit contributions | |
From | +0.098 |
Bug | +0.084 |
Tweet | +0.083 |
Flo | +0.083 |
Looks | +0.083 |
Neg logit contributions | |
sister | -0.055 |
teacher | -0.054 |
happened | -0.052 |
exam | -0.050 |
problem | -0.050 |
Token | can |
Token ID | 460 |
Feature activation | +2.98e+00 |
Pos logit contributions | |
From | +0.337 |
Flo | +0.292 |
Since | +0.287 |
Always | +0.287 |
Even | +0.279 |
Neg logit contributions | |
happened | -0.222 |
teacher | -0.207 |
sister | -0.203 |
exam | -0.200 |
problem | -0.197 |
Token | help |
Token ID | 1037 |
Feature activation | +3.39e+00 |
Pos logit contributions | |
teacher | +0.067 |
happened | +0.066 |
problem | +0.064 |
exam | +0.064 |
sister | +0.063 |
Neg logit contributions | |
From | -0.139 |
Since | -0.123 |
Looks | -0.118 |
Always | -0.118 |
oc | -0.116 |
Token | bird |
Token ID | 6512 |
Feature activation | -8.32e-01 |
Pos logit contributions | |
From | +0.073 |
Flo | +0.064 |
Always | +0.063 |
Since | +0.063 |
They | +0.062 |
Neg logit contributions | |
happened | -0.037 |
teacher | -0.035 |
exam | -0.034 |
sister | -0.033 |
problem | -0.033 |
Token | for |
Token ID | 329 |
Feature activation | -8.77e-01 |
Pos logit contributions | |
From | +0.038 |
Always | +0.033 |
Since | +0.033 |
Flo | +0.033 |
Even | +0.033 |
Neg logit contributions | |
happened | -0.019 |
teacher | -0.019 |
sister | -0.019 |
exam | -0.018 |
problem | -0.017 |
Token | the |
Token ID | 262 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
From | +0.035 |
Flo | +0.031 |
Always | +0.030 |
Since | +0.030 |
oc | +0.029 |
Neg logit contributions | |
happened | -0.026 |
teacher | -0.025 |
exam | -0.024 |
sister | -0.024 |
problem | -0.023 |
Token | advice |
Token ID | 5608 |
Feature activation | +1.54e-01 |
Pos logit contributions | |
From | +0.067 |
Always | +0.057 |
Flo | +0.057 |
Since | +0.057 |
They | +0.056 |
Neg logit contributions | |
happened | -0.039 |
teacher | -0.037 |
exam | -0.036 |
sister | -0.035 |
problem | -0.034 |
Token | . |
Token ID | 13 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
happened | +0.004 |
teacher | +0.003 |
exam | +0.003 |
sister | +0.003 |
problem | +0.003 |
Neg logit contributions | |
From | -0.007 |
Flo | -0.006 |
Always | -0.006 |
Since | -0.006 |
Even | -0.006 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.36e+01 |
Pos logit contributions | |
teacher | +0.052 |
sister | +0.052 |
brother | +0.048 |
classmate | +0.047 |
dad | +0.047 |
Neg logit contributions | |
cking | -0.036 |
From | -0.032 |
inning | -0.031 |
Always | -0.030 |
Flo | -0.030 |
Token | the |
Token ID | 262 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
From | +0.599 |
Flo | +0.520 |
Since | +0.519 |
Always | +0.508 |
They | +0.501 |
Neg logit contributions | |
happened | -0.375 |
problem | -0.337 |
teacher | -0.332 |
exam | -0.331 |
sister | -0.328 |
Token | camera |
Token ID | 4676 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
From | +0.061 |
Always | +0.051 |
They | +0.050 |
Since | +0.050 |
oc | +0.050 |
Neg logit contributions | |
happened | -0.038 |
teacher | -0.038 |
exam | -0.038 |
sister | -0.037 |
classmate | -0.035 |
Token | , |
Token ID | 11 |
Feature activation | +4.65e+00 |
Pos logit contributions | |
From | +0.059 |
Flo | +0.051 |
Always | +0.051 |
They | +0.051 |
Since | +0.050 |
Neg logit contributions | |
happened | -0.035 |
teacher | -0.033 |
exam | -0.033 |
sister | -0.032 |
problem | -0.031 |
Token | he |
Token ID | 339 |
Feature activation | -3.06e-01 |
Pos logit contributions | |
sister | +0.127 |
teacher | +0.117 |
brother | +0.115 |
happened | +0.112 |
dad | +0.108 |
Neg logit contributions | |
From | -0.197 |
Flo | -0.194 |
Tweet | -0.190 |
Bug | -0.189 |
Twe | -0.186 |
Token | and |
Token ID | 290 |
Feature activation | -7.98e+00 |
Pos logit contributions | |
From | +0.012 |
Flo | +0.010 |
Always | +0.010 |
Since | +0.010 |
Bug | +0.010 |
Neg logit contributions | |
happened | -0.009 |
teacher | -0.009 |
sister | -0.009 |
problem | -0.008 |
classmate | -0.008 |
Token | what |
Token ID | 644 |
Feature activation | +6.46e-01 |
Pos logit contributions | |
happened | +0.036 |
teacher | +0.034 |
exam | +0.034 |
sister | +0.033 |
problem | +0.032 |
Neg logit contributions | |
From | -0.048 |
Flo | -0.040 |
Always | -0.040 |
Since | -0.040 |
They | -0.039 |
Token | it |
Token ID | 340 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
exam | +0.021 |
sister | +0.020 |
teacher | +0.020 |
happened | +0.020 |
problem | +0.020 |
Neg logit contributions | |
From | -0.023 |
Always | -0.019 |
They | -0.018 |
Flo | -0.018 |
The | -0.018 |
Token | was |
Token ID | 373 |
Feature activation | +3.87e+00 |
Pos logit contributions | |
happened | +0.047 |
teacher | +0.041 |
exam | +0.040 |
problem | +0.039 |
sister | +0.039 |
Neg logit contributions | |
From | -0.070 |
Flo | -0.060 |
Since | -0.060 |
Always | -0.059 |
They | -0.058 |
Token | like |
Token ID | 588 |
Feature activation | -4.64e+00 |
Pos logit contributions | |
teacher | +0.121 |
sister | +0.118 |
happened | +0.118 |
exam | +0.118 |
classmate | +0.114 |
Neg logit contributions | |
From | -0.137 |
Always | -0.121 |
They | -0.115 |
Flo | -0.114 |
Since | -0.113 |
Token | . |
Token ID | 13 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
From | +0.194 |
Always | +0.166 |
Flo | +0.166 |
Since | +0.164 |
They | +0.162 |
Neg logit contributions | |
happened | -0.127 |
teacher | -0.121 |
exam | -0.120 |
sister | -0.118 |
problem | -0.115 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.36e+01 |
Pos logit contributions | |
happened | +0.056 |
teacher | +0.056 |
sister | +0.054 |
exam | +0.053 |
classmate | +0.051 |
Neg logit contributions | |
From | -0.046 |
cking | -0.042 |
Flo | -0.041 |
oc | -0.038 |
Tweet | -0.038 |
Token | a |
Token ID | 257 |
Feature activation | +3.86e+00 |
Pos logit contributions | |
From | +0.580 |
Flo | +0.498 |
Since | +0.495 |
Always | +0.490 |
They | +0.488 |
Neg logit contributions | |
happened | -0.387 |
exam | -0.353 |
teacher | -0.352 |
problem | -0.347 |
sister | -0.343 |
Token | leap |
Token ID | 16470 |
Feature activation | -3.90e+00 |
Pos logit contributions | |
happened | +0.125 |
teacher | +0.120 |
exam | +0.119 |
sister | +0.116 |
problem | +0.113 |
Neg logit contributions | |
From | -0.143 |
Flo | -0.120 |
Always | -0.118 |
Since | -0.117 |
They | -0.116 |
Token | of |
Token ID | 286 |
Feature activation | -4.15e+00 |
Pos logit contributions | |
From | +0.186 |
Always | +0.163 |
Flo | +0.162 |
Since | +0.161 |
They | +0.159 |
Neg logit contributions | |
happened | -0.085 |
teacher | -0.079 |
exam | -0.078 |
sister | -0.075 |
problem | -0.074 |
Token | faith |
Token ID | 4562 |
Feature activation | -2.40e+00 |
Pos logit contributions | |
From | +0.158 |
Flo | +0.135 |
Always | +0.134 |
Since | +0.134 |
They | +0.133 |
Neg logit contributions | |
happened | -0.135 |
teacher | -0.127 |
exam | -0.126 |
sister | -0.122 |
problem | -0.122 |
Token | , |
Token ID | 11 |
Feature activation | +4.84e+00 |
Pos logit contributions | |
From | +0.112 |
Flo | +0.098 |
Always | +0.097 |
Since | +0.097 |
Even | +0.095 |
Neg logit contributions | |
happened | -0.057 |
teacher | -0.051 |
problem | -0.050 |
exam | -0.050 |
sister | -0.049 |
Token | the |
Token ID | 262 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
From | +0.046 |
Flo | +0.040 |
Since | +0.039 |
They | +0.038 |
Even | +0.038 |
Neg logit contributions | |
happened | -0.037 |
teacher | -0.034 |
sister | -0.033 |
exam | -0.033 |
problem | -0.033 |
Token | bear |
Token ID | 6842 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
From | +0.051 |
Always | +0.043 |
Flo | +0.043 |
Since | +0.043 |
They | +0.042 |
Neg logit contributions | |
happened | -0.038 |
teacher | -0.036 |
exam | -0.036 |
sister | -0.035 |
problem | -0.034 |
Token | was |
Token ID | 373 |
Feature activation | +4.09e+00 |
Pos logit contributions | |
happened | +0.038 |
teacher | +0.033 |
sister | +0.033 |
problem | +0.032 |
exam | +0.032 |
Neg logit contributions | |
From | -0.054 |
Flo | -0.046 |
Since | -0.046 |
They | -0.046 |
Always | -0.045 |
Token | right |
Token ID | 826 |
Feature activation | +2.40e+00 |
Pos logit contributions | |
happened | +0.121 |
teacher | +0.112 |
exam | +0.110 |
sister | +0.110 |
problem | +0.110 |
Neg logit contributions | |
From | -0.167 |
Flo | -0.142 |
Since | -0.141 |
Always | -0.140 |
oc | -0.139 |
Token | . |
Token ID | 13 |
Feature activation | +1.57e+00 |
Pos logit contributions | |
happened | +0.070 |
exam | +0.068 |
teacher | +0.067 |
sister | +0.065 |
problem | +0.062 |
Neg logit contributions | |
From | -0.095 |
Always | -0.081 |
They | -0.080 |
Since | -0.080 |
Flo | -0.079 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.36e+01 |
Pos logit contributions | |
teacher | +0.051 |
sister | +0.050 |
brother | +0.046 |
principal | +0.046 |
dad | +0.046 |
Neg logit contributions | |
cking | -0.054 |
inning | -0.048 |
From | -0.047 |
Looks | -0.045 |
bing | -0.044 |
Token | the |
Token ID | 262 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
From | +0.607 |
Since | +0.532 |
Flo | +0.520 |
They | +0.519 |
Always | +0.514 |
Neg logit contributions | |
happened | -0.371 |
problem | -0.337 |
exam | -0.324 |
sister | -0.320 |
teacher | -0.315 |
Token | gold |
Token ID | 3869 |
Feature activation | -4.50e+00 |
Pos logit contributions | |
From | +0.047 |
oc | +0.040 |
Flo | +0.040 |
Always | +0.039 |
Squ | +0.038 |
Neg logit contributions | |
happened | -0.043 |
teacher | -0.042 |
exam | -0.042 |
sister | -0.040 |
classmate | -0.040 |
Token | would |
Token ID | 561 |
Feature activation | +4.45e+00 |
Pos logit contributions | |
From | +0.184 |
Flo | +0.157 |
Always | +0.156 |
Since | +0.155 |
They | +0.153 |
Neg logit contributions | |
happened | -0.128 |
teacher | -0.122 |
exam | -0.121 |
sister | -0.118 |
problem | -0.116 |
Token | have |
Token ID | 423 |
Feature activation | -3.37e+00 |
Pos logit contributions | |
teacher | +0.115 |
happened | +0.109 |
exam | +0.104 |
problem | +0.100 |
sister | +0.099 |
Neg logit contributions | |
From | -0.192 |
pill | -0.175 |
ksh | -0.175 |
Looks | -0.172 |
Since | -0.171 |
Token | been |
Token ID | 587 |
Feature activation | +2.92e+00 |
Pos logit contributions | |
From | +0.138 |
Flo | +0.117 |
Always | +0.117 |
Since | +0.116 |
They | +0.114 |
Neg logit contributions | |
happened | -0.097 |
teacher | -0.092 |
exam | -0.091 |
sister | -0.089 |
problem | -0.088 |
Token | no |
Token ID | 645 |
Feature activation | +2.49e+00 |
Pos logit contributions | |
happened | +0.055 |
teacher | +0.052 |
exam | +0.052 |
sister | +0.051 |
problem | +0.050 |
Neg logit contributions | |
From | -0.066 |
Flo | -0.055 |
Always | -0.055 |
Since | -0.055 |
They | -0.054 |
Token | one |
Token ID | 530 |
Feature activation | +2.48e+00 |
Pos logit contributions | |
happened | +0.070 |
teacher | +0.070 |
classmate | +0.068 |
principal | +0.068 |
exam | +0.067 |
Neg logit contributions | |
From | -0.091 |
Always | -0.079 |
Flo | -0.077 |
They | -0.075 |
Since | -0.073 |
Token | is |
Token ID | 318 |
Feature activation | +3.22e+00 |
Pos logit contributions | |
teacher | +0.078 |
exam | +0.076 |
happened | +0.075 |
sister | +0.075 |
classmate | +0.074 |
Neg logit contributions | |
From | -0.085 |
Always | -0.074 |
They | -0.070 |
Flo | -0.070 |
oc | -0.069 |
Token | watching |
Token ID | 4964 |
Feature activation | -2.47e+00 |
Pos logit contributions | |
happened | +0.098 |
teacher | +0.093 |
exam | +0.092 |
sister | +0.090 |
problem | +0.089 |
Neg logit contributions | |
From | -0.126 |
Always | -0.107 |
Flo | -0.107 |
Since | -0.106 |
They | -0.104 |
Token | . |
Token ID | 13 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
From | +0.100 |
Flo | +0.086 |
Always | +0.085 |
Since | +0.084 |
They | +0.084 |
Neg logit contributions | |
happened | -0.070 |
teacher | -0.066 |
exam | -0.066 |
sister | -0.064 |
problem | -0.063 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.36e+01 |
Pos logit contributions | |
teacher | +0.043 |
sister | +0.043 |
cousin | +0.042 |
principal | +0.039 |
brother | +0.039 |
Neg logit contributions | |
cking | -0.050 |
ized | -0.047 |
valleys | -0.045 |
refreshing | -0.044 |
winding | -0.043 |
Token | something |
Token ID | 1223 |
Feature activation | +4.17e+00 |
Pos logit contributions | |
From | +0.594 |
Flo | +0.510 |
Since | +0.509 |
Always | +0.506 |
They | +0.499 |
Neg logit contributions | |
happened | -0.363 |
teacher | -0.337 |
exam | -0.337 |
sister | -0.330 |
problem | -0.328 |
Token | that |
Token ID | 326 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
happened | +0.128 |
teacher | +0.122 |
exam | +0.121 |
sister | +0.119 |
problem | +0.117 |
Neg logit contributions | |
From | -0.162 |
Always | -0.136 |
Flo | -0.136 |
Since | -0.135 |
They | -0.133 |
Token | doesn |
Token ID | 1595 |
Feature activation | -3.68e-01 |
Pos logit contributions | |
From | +0.050 |
Flo | +0.043 |
Always | +0.042 |
Since | +0.042 |
They | +0.041 |
Neg logit contributions | |
happened | -0.041 |
teacher | -0.039 |
exam | -0.039 |
sister | -0.038 |
problem | -0.038 |
Token | 't |
Token ID | 470 |
Feature activation | -2.66e-01 |
Pos logit contributions | |
From | +0.015 |
Flo | +0.013 |
Always | +0.013 |
oc | +0.013 |
Looks | +0.013 |
Neg logit contributions | |
happened | -0.010 |
exam | -0.010 |
teacher | -0.010 |
problem | -0.009 |
sister | -0.009 |
Token | belong |
Token ID | 5594 |
Feature activation | +9.71e-01 |
Pos logit contributions | |
From | +0.012 |
Flo | +0.010 |
Since | +0.010 |
Always | +0.010 |
oc | +0.010 |
Neg logit contributions | |
happened | -0.007 |
teacher | -0.007 |
exam | -0.007 |
sister | -0.006 |
problem | -0.006 |
Token | learned |
Token ID | 4499 |
Feature activation | +7.95e-01 |
Pos logit contributions | |
exam | +0.012 |
happened | +0.012 |
teacher | +0.012 |
sister | +0.012 |
problem | +0.012 |
Neg logit contributions | |
From | -0.016 |
Always | -0.013 |
Since | -0.013 |
They | -0.013 |
Flo | -0.013 |
Token | an |
Token ID | 281 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
happened | +0.024 |
sister | +0.023 |
teacher | +0.023 |
exam | +0.023 |
problem | +0.022 |
Neg logit contributions | |
From | -0.031 |
Always | -0.026 |
Since | -0.026 |
They | -0.025 |
Flo | -0.025 |
Token | important |
Token ID | 1593 |
Feature activation | -8.06e-01 |
Pos logit contributions | |
happened | +0.051 |
teacher | +0.049 |
exam | +0.048 |
sister | +0.047 |
problem | +0.046 |
Neg logit contributions | |
From | -0.070 |
Always | -0.059 |
Flo | -0.059 |
Since | -0.058 |
They | -0.057 |
Token | lesson |
Token ID | 11483 |
Feature activation | +2.63e+00 |
Pos logit contributions | |
From | +0.035 |
Even | +0.029 |
Since | +0.029 |
Always | +0.029 |
oc | +0.029 |
Neg logit contributions | |
happened | -0.020 |
sister | -0.019 |
teacher | -0.019 |
exam | -0.018 |
classmate | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
happened | +0.076 |
teacher | +0.074 |
sister | +0.073 |
exam | +0.073 |
problem | +0.070 |
Neg logit contributions | |
From | -0.105 |
Always | -0.089 |
Since | -0.088 |
They | -0.087 |
Flo | -0.087 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.36e+01 |
Pos logit contributions | |
teacher | +0.059 |
happened | +0.057 |
sister | +0.056 |
exam | +0.054 |
classmate | +0.053 |
Neg logit contributions | |
From | -0.045 |
cking | -0.044 |
Flo | -0.041 |
Tweet | -0.041 |
Ho | -0.040 |
Token | things |
Token ID | 1243 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
From | +0.551 |
Flo | +0.470 |
Always | +0.465 |
Since | +0.465 |
They | +0.458 |
Neg logit contributions | |
happened | -0.400 |
teacher | -0.379 |
exam | -0.377 |
sister | -0.367 |
problem | -0.364 |
Token | without |
Token ID | 1231 |
Feature activation | -1.56e+00 |
Pos logit contributions | |
happened | +0.005 |
teacher | +0.005 |
exam | +0.005 |
problem | +0.004 |
sister | +0.004 |
Neg logit contributions | |
From | -0.008 |
Flo | -0.007 |
oc | -0.006 |
Since | -0.006 |
Always | -0.006 |
Token | asking |
Token ID | 4737 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
From | +0.066 |
Flo | +0.057 |
Since | +0.056 |
Always | +0.056 |
They | +0.055 |
Neg logit contributions | |
happened | -0.044 |
teacher | -0.041 |
exam | -0.040 |
sister | -0.040 |
problem | -0.039 |
Token | was |
Token ID | 373 |
Feature activation | +3.78e+00 |
Pos logit contributions | |
happened | +0.042 |
teacher | +0.041 |
exam | +0.040 |
sister | +0.039 |
problem | +0.038 |
Neg logit contributions | |
From | -0.061 |
Flo | -0.052 |
Always | -0.052 |
Since | -0.052 |
They | -0.051 |
Token | wrong |
Token ID | 2642 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
happened | +0.111 |
teacher | +0.106 |
exam | +0.105 |
sister | +0.103 |
problem | +0.100 |
Neg logit contributions | |
From | -0.150 |
Always | -0.127 |
Flo | -0.126 |
Since | -0.125 |
They | -0.124 |
Token | and |
Token ID | 290 |
Feature activation | -7.78e+00 |
Pos logit contributions | |
teacher | +0.026 |
happened | +0.026 |
sister | +0.025 |
exam | +0.025 |
problem | +0.024 |
Neg logit contributions | |
From | -0.035 |
Flo | -0.030 |
Since | -0.030 |
Always | -0.029 |
Even | -0.029 |
Token | cake |
Token ID | 12187 |
Feature activation | +1.68e+00 |
Pos logit contributions | |
From | +0.316 |
Flo | +0.272 |
Since | +0.268 |
Always | +0.268 |
They | +0.261 |
Neg logit contributions | |
happened | -0.225 |
teacher | -0.215 |
exam | -0.210 |
sister | -0.208 |
problem | -0.203 |
Token | with |
Token ID | 351 |
Feature activation | -5.62e+00 |
Pos logit contributions | |
happened | +0.054 |
teacher | +0.052 |
exam | +0.051 |
sister | +0.051 |
problem | +0.049 |
Neg logit contributions | |
From | -0.063 |
Flo | -0.052 |
Always | -0.052 |
Since | -0.052 |
They | -0.050 |
Token | them |
Token ID | 606 |
Feature activation | +2.85e+00 |
Pos logit contributions | |
From | +0.233 |
Flo | +0.200 |
Since | +0.199 |
Always | +0.198 |
They | +0.192 |
Neg logit contributions | |
happened | -0.159 |
teacher | -0.152 |
sister | -0.148 |
exam | -0.147 |
problem | -0.143 |
Token | . |
Token ID | 13 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
happened | +0.073 |
teacher | +0.070 |
sister | +0.069 |
exam | +0.068 |
problem | +0.066 |
Neg logit contributions | |
From | -0.124 |
Flo | -0.107 |
Always | -0.107 |
Since | -0.106 |
Even | -0.104 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.36e+01 |
Pos logit contributions | |
teacher | +0.048 |
sister | +0.047 |
cousin | +0.046 |
birthday | +0.045 |
mom | +0.043 |
Neg logit contributions | |
cking | -0.064 |
ized | -0.058 |
winding | -0.054 |
valleys | -0.054 |
bing | -0.053 |
Token | things |
Token ID | 1243 |
Feature activation | +3.17e-01 |
Pos logit contributions | |
From | +0.580 |
Flo | +0.496 |
Since | +0.495 |
Always | +0.492 |
They | +0.483 |
Neg logit contributions | |
happened | -0.373 |
teacher | -0.350 |
exam | -0.348 |
sister | -0.342 |
problem | -0.338 |
Token | that |
Token ID | 326 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
happened | +0.010 |
teacher | +0.009 |
problem | +0.009 |
exam | +0.009 |
sister | +0.009 |
Neg logit contributions | |
From | -0.013 |
Flo | -0.011 |
Since | -0.011 |
oc | -0.011 |
Always | -0.011 |
Token | do |
Token ID | 466 |
Feature activation | +1.82e+00 |
Pos logit contributions | |
From | +0.053 |
Always | +0.045 |
Flo | +0.045 |
Since | +0.044 |
They | +0.044 |
Neg logit contributions | |
happened | -0.039 |
teacher | -0.038 |
exam | -0.037 |
sister | -0.036 |
problem | -0.036 |
Token | not |
Token ID | 407 |
Feature activation | -3.22e-01 |
Pos logit contributions | |
happened | +0.054 |
teacher | +0.051 |
problem | +0.050 |
sister | +0.049 |
exam | +0.048 |
Neg logit contributions | |
From | -0.074 |
Flo | -0.066 |
Bug | -0.063 |
Always | -0.063 |
Since | -0.063 |
Token | belong |
Token ID | 5594 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
From | +0.013 |
Flo | +0.011 |
Since | +0.011 |
Always | +0.011 |
They | +0.010 |
Neg logit contributions | |
happened | -0.010 |
teacher | -0.010 |
exam | -0.009 |
sister | -0.009 |
problem | -0.009 |
Token | always |
Token ID | 1464 |
Feature activation | -5.45e+00 |
Pos logit contributions | |
happened | +0.063 |
teacher | +0.061 |
exam | +0.058 |
problem | +0.056 |
sister | +0.052 |
Neg logit contributions | |
From | -0.121 |
pill | -0.109 |
ang | -0.109 |
Looks | -0.109 |
ksh | -0.108 |
Token | have |
Token ID | 423 |
Feature activation | -3.59e+00 |
Pos logit contributions | |
From | +0.235 |
Since | +0.204 |
Flo | +0.198 |
Always | +0.195 |
pill | +0.193 |
Neg logit contributions | |
happened | -0.150 |
teacher | -0.142 |
exam | -0.142 |
sister | -0.138 |
problem | -0.137 |
Token | its |
Token ID | 663 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
From | +0.143 |
Always | +0.122 |
Flo | +0.122 |
Since | +0.119 |
They | +0.119 |
Neg logit contributions | |
happened | -0.103 |
teacher | -0.101 |
exam | -0.100 |
sister | -0.097 |
problem | -0.095 |
Token | rewards |
Token ID | 11530 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
From | +0.058 |
Always | +0.049 |
Flo | +0.049 |
Since | +0.048 |
They | +0.048 |
Neg logit contributions | |
happened | -0.042 |
teacher | -0.040 |
exam | -0.039 |
sister | -0.038 |
problem | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
From | +0.045 |
Flo | +0.038 |
Always | +0.038 |
Since | +0.038 |
They | +0.037 |
Neg logit contributions | |
happened | -0.030 |
teacher | -0.029 |
exam | -0.028 |
sister | -0.028 |
problem | -0.027 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.36e+01 |
Pos logit contributions | |
sister | +0.036 |
teacher | +0.034 |
ha | +0.031 |
brother | +0.030 |
mom | +0.027 |
Neg logit contributions | |
cking | -0.060 |
ized | -0.059 |
valleys | -0.052 |
izing | -0.051 |
inning | -0.051 |
Token | a |
Token ID | 257 |
Feature activation | +3.86e+00 |
Pos logit contributions | |
From | +0.600 |
Since | +0.525 |
Flo | +0.507 |
They | +0.506 |
Always | +0.501 |
Neg logit contributions | |
happened | -0.370 |
problem | -0.339 |
exam | -0.335 |
sister | -0.327 |
teacher | -0.323 |
Token | break |
Token ID | 2270 |
Feature activation | -3.18e+00 |
Pos logit contributions | |
happened | +0.115 |
teacher | +0.109 |
exam | +0.108 |
sister | +0.106 |
problem | +0.103 |
Neg logit contributions | |
From | -0.154 |
Flo | -0.130 |
Always | -0.130 |
Since | -0.129 |
They | -0.127 |
Token | is |
Token ID | 318 |
Feature activation | +3.26e+00 |
Pos logit contributions | |
From | +0.143 |
Since | +0.123 |
Always | +0.123 |
Flo | +0.122 |
Looks | +0.119 |
Neg logit contributions | |
happened | -0.079 |
teacher | -0.074 |
sister | -0.073 |
exam | -0.073 |
problem | -0.073 |
Token | important |
Token ID | 1593 |
Feature activation | -7.77e-01 |
Pos logit contributions | |
happened | +0.098 |
teacher | +0.092 |
exam | +0.091 |
sister | +0.089 |
problem | +0.088 |
Neg logit contributions | |
From | -0.130 |
Always | -0.110 |
Flo | -0.110 |
Since | -0.109 |
They | -0.107 |
Token | , |
Token ID | 11 |
Feature activation | +4.66e+00 |
Pos logit contributions | |
From | +0.033 |
Always | +0.028 |
Flo | +0.028 |
Since | +0.028 |
They | +0.027 |
Neg logit contributions | |
happened | -0.020 |
teacher | -0.020 |
exam | -0.019 |
sister | -0.019 |
problem | -0.018 |
Token | when |
Token ID | 618 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
From | +0.141 |
Flo | +0.122 |
Since | +0.120 |
Always | +0.117 |
Even | +0.116 |
Neg logit contributions | |
happened | -0.098 |
teacher | -0.093 |
exam | -0.092 |
sister | -0.092 |
problem | -0.091 |
Token | we |
Token ID | 356 |
Feature activation | -5.23e-01 |
Pos logit contributions | |
happened | +0.057 |
problem | +0.053 |
teacher | +0.053 |
exam | +0.052 |
sister | +0.052 |
Neg logit contributions | |
From | -0.072 |
Flo | -0.062 |
Since | -0.062 |
Always | -0.061 |
oc | -0.060 |
Token | are |
Token ID | 389 |
Feature activation | +2.22e+00 |
Pos logit contributions | |
From | +0.023 |
Since | +0.020 |
Flo | +0.020 |
Looks | +0.019 |
Always | +0.019 |
Neg logit contributions | |
happened | -0.014 |
exam | -0.013 |
teacher | -0.013 |
sister | -0.013 |
problem | -0.013 |
Token | busy |
Token ID | 8179 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
happened | +0.068 |
teacher | +0.065 |
exam | +0.064 |
sister | +0.063 |
problem | +0.062 |
Neg logit contributions | |
From | -0.087 |
Flo | -0.073 |
Always | -0.073 |
Since | -0.073 |
They | -0.072 |
Token | . |
Token ID | 13 |
Feature activation | +1.57e+00 |
Pos logit contributions | |
happened | +0.006 |
teacher | +0.006 |
exam | +0.006 |
sister | +0.006 |
problem | +0.006 |
Neg logit contributions | |
From | -0.009 |
Always | -0.008 |
Flo | -0.008 |
Since | -0.008 |
They | -0.008 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.36e+01 |
Pos logit contributions | |
sister | +0.042 |
teacher | +0.042 |
cousin | +0.039 |
brother | +0.039 |
principal | +0.037 |
Neg logit contributions | |
cking | -0.061 |
ized | -0.059 |
refreshing | -0.053 |
izing | -0.052 |
winding | -0.052 |
Token | a |
Token ID | 257 |
Feature activation | +3.93e+00 |
Pos logit contributions | |
From | +0.590 |
Since | +0.511 |
Flo | +0.504 |
Always | +0.491 |
They | +0.489 |
Neg logit contributions | |
happened | -0.379 |
problem | -0.350 |
exam | -0.347 |
sister | -0.340 |
teacher | -0.336 |
Token | break |
Token ID | 2270 |
Feature activation | -3.29e+00 |
Pos logit contributions | |
happened | +0.115 |
teacher | +0.111 |
exam | +0.110 |
sister | +0.107 |
problem | +0.107 |
Neg logit contributions | |
From | -0.158 |
Always | -0.135 |
Flo | -0.134 |
Since | -0.133 |
They | -0.131 |
Token | will |
Token ID | 481 |
Feature activation | +2.62e+00 |
Pos logit contributions | |
From | +0.160 |
Always | +0.140 |
Since | +0.139 |
Flo | +0.139 |
Even | +0.135 |
Neg logit contributions | |
happened | -0.071 |
teacher | -0.065 |
sister | -0.065 |
exam | -0.064 |
problem | -0.063 |
Token | help |
Token ID | 1037 |
Feature activation | +3.45e+00 |
Pos logit contributions | |
happened | +0.072 |
teacher | +0.069 |
exam | +0.066 |
problem | +0.065 |
sister | +0.063 |
Neg logit contributions | |
From | -0.109 |
Flo | -0.095 |
Looks | -0.095 |
pill | -0.093 |
Since | -0.093 |
Token | us |
Token ID | 514 |
Feature activation | +4.12e+00 |
Pos logit contributions | |
happened | +0.097 |
teacher | +0.095 |
exam | +0.093 |
sister | +0.093 |
problem | +0.090 |
Neg logit contributions | |
From | -0.143 |
Since | -0.123 |
Flo | -0.121 |
Always | -0.121 |
Looks | -0.119 |
Token | you |
Token ID | 345 |
Feature activation | +2.83e+00 |
Pos logit contributions | |
happened | +0.146 |
exam | +0.143 |
teacher | +0.143 |
sister | +0.140 |
problem | +0.132 |
Neg logit contributions | |
From | -0.196 |
Always | -0.169 |
Flo | -0.169 |
Since | -0.167 |
oc | -0.167 |
Token | broke |
Token ID | 6265 |
Feature activation | +8.60e-01 |
Pos logit contributions | |
happened | +0.095 |
teacher | +0.092 |
exam | +0.091 |
sister | +0.088 |
problem | +0.087 |
Neg logit contributions | |
From | -0.101 |
Always | -0.084 |
Flo | -0.083 |
Since | -0.082 |
They | -0.081 |
Token | the |
Token ID | 262 |
Feature activation | -1.38e+00 |
Pos logit contributions | |
happened | +0.032 |
teacher | +0.031 |
exam | +0.030 |
sister | +0.029 |
classmate | +0.029 |
Neg logit contributions | |
From | -0.027 |
Flo | -0.023 |
They | -0.022 |
Always | -0.022 |
Since | -0.022 |
Token | law |
Token ID | 1099 |
Feature activation | -2.80e+00 |
Pos logit contributions | |
From | +0.048 |
Flo | +0.043 |
Always | +0.042 |
The | +0.040 |
Since | +0.039 |
Neg logit contributions | |
happened | -0.043 |
teacher | -0.038 |
exam | -0.038 |
classmate | -0.038 |
sister | -0.038 |
Token | . |
Token ID | 13 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
From | +0.100 |
They | +0.084 |
Always | +0.084 |
Since | +0.083 |
Flo | +0.082 |
Neg logit contributions | |
happened | -0.092 |
teacher | -0.084 |
exam | -0.084 |
sister | -0.083 |
problem | -0.082 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.36e+01 |
Pos logit contributions | |
sister | +0.067 |
teacher | +0.065 |
exam | +0.062 |
happened | +0.060 |
brother | +0.059 |
Neg logit contributions | |
From | -0.059 |
cking | -0.051 |
Always | -0.051 |
Flo | -0.051 |
They | -0.050 |
Token | things |
Token ID | 1243 |
Feature activation | +3.80e-01 |
Pos logit contributions | |
From | +0.510 |
Always | +0.434 |
Flo | +0.432 |
They | +0.420 |
Since | +0.419 |
Neg logit contributions | |
happened | -0.417 |
teacher | -0.405 |
exam | -0.394 |
sister | -0.386 |
classmate | -0.373 |
Token | without |
Token ID | 1231 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
happened | +0.011 |
teacher | +0.011 |
exam | +0.010 |
sister | +0.010 |
problem | +0.010 |
Neg logit contributions | |
From | -0.015 |
Always | -0.013 |
Flo | -0.013 |
Since | -0.013 |
They | -0.013 |
Token | asking |
Token ID | 4737 |
Feature activation | +1.58e+00 |
Pos logit contributions | |
From | +0.053 |
Always | +0.045 |
Flo | +0.045 |
They | +0.044 |
Since | +0.043 |
Neg logit contributions | |
happened | -0.048 |
teacher | -0.047 |
exam | -0.046 |
sister | -0.045 |
classmate | -0.044 |
Token | is |
Token ID | 318 |
Feature activation | +3.25e+00 |
Pos logit contributions | |
happened | +0.052 |
exam | +0.047 |
teacher | +0.047 |
sister | +0.046 |
problem | +0.046 |
Neg logit contributions | |
From | -0.058 |
They | -0.048 |
Always | -0.048 |
Flo | -0.047 |
Since | -0.047 |
Token | wrong |
Token ID | 2642 |
Feature activation | +5.97e+00 |
Pos logit contributions | |
happened | +0.099 |
teacher | +0.093 |
classmate | +0.092 |
exam | +0.091 |
sister | +0.088 |
Neg logit contributions | |
From | -0.115 |
Always | -0.097 |
They | -0.096 |
Flo | -0.096 |
Since | -0.093 |
Token | she |
Token ID | 673 |
Feature activation | +3.61e-01 |
Pos logit contributions | |
happened | +0.167 |
exam | +0.162 |
teacher | +0.161 |
sister | +0.159 |
problem | +0.157 |
Neg logit contributions | |
From | -0.223 |
Always | -0.189 |
Since | -0.186 |
They | -0.185 |
Flo | -0.184 |
Token | had |
Token ID | 550 |
Feature activation | +4.27e-01 |
Pos logit contributions | |
happened | +0.012 |
exam | +0.012 |
teacher | +0.011 |
sister | +0.011 |
problem | +0.011 |
Neg logit contributions | |
From | -0.013 |
Always | -0.011 |
Since | -0.010 |
They | -0.010 |
Flo | -0.010 |
Token | to |
Token ID | 284 |
Feature activation | +4.72e+00 |
Pos logit contributions | |
exam | +0.013 |
sister | +0.013 |
teacher | +0.012 |
problem | +0.012 |
happened | +0.012 |
Neg logit contributions | |
From | -0.016 |
Always | -0.014 |
They | -0.013 |
The | -0.013 |
Since | -0.013 |
Token | hurry |
Token ID | 23290 |
Feature activation | +7.92e-01 |
Pos logit contributions | |
happened | +0.139 |
teacher | +0.133 |
exam | +0.132 |
sister | +0.128 |
problem | +0.126 |
Neg logit contributions | |
From | -0.188 |
Flo | -0.160 |
Always | -0.160 |
Since | -0.158 |
They | -0.156 |
Token | . |
Token ID | 13 |
Feature activation | +1.62e+00 |
Pos logit contributions | |
happened | +0.024 |
teacher | +0.023 |
sister | +0.023 |
exam | +0.022 |
problem | +0.022 |
Neg logit contributions | |
From | -0.030 |
They | -0.026 |
Always | -0.026 |
Since | -0.025 |
Flo | -0.025 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.36e+01 |
Pos logit contributions | |
teacher | +0.060 |
happened | +0.058 |
sister | +0.057 |
exam | +0.055 |
classmate | +0.054 |
Neg logit contributions | |
From | -0.049 |
cking | -0.042 |
Always | -0.041 |
Flo | -0.040 |
Squ | -0.039 |
Token | a |
Token ID | 257 |
Feature activation | +4.00e+00 |
Pos logit contributions | |
From | +0.565 |
Flo | +0.481 |
Always | +0.481 |
Since | +0.478 |
They | +0.471 |
Neg logit contributions | |
happened | -0.382 |
teacher | -0.362 |
exam | -0.359 |
sister | -0.350 |
problem | -0.345 |
Token | deep |
Token ID | 2769 |
Feature activation | -3.23e+00 |
Pos logit contributions | |
happened | +0.137 |
teacher | +0.130 |
exam | +0.128 |
sister | +0.127 |
problem | +0.122 |
Neg logit contributions | |
From | -0.138 |
Flo | -0.116 |
Since | -0.112 |
They | -0.112 |
Always | -0.111 |
Token | breath |
Token ID | 8033 |
Feature activation | -2.64e+00 |
Pos logit contributions | |
From | +0.132 |
Always | +0.112 |
Flo | +0.111 |
Since | +0.111 |
They | +0.109 |
Neg logit contributions | |
happened | -0.094 |
teacher | -0.089 |
exam | -0.089 |
sister | -0.087 |
problem | -0.086 |
Token | , |
Token ID | 11 |
Feature activation | +4.82e+00 |
Pos logit contributions | |
From | +0.115 |
Always | +0.100 |
Flo | +0.100 |
Since | +0.099 |
They | +0.096 |
Neg logit contributions | |
happened | -0.070 |
teacher | -0.064 |
exam | -0.062 |
problem | -0.061 |
sister | -0.060 |
Token | she |
Token ID | 673 |
Feature activation | +3.71e-01 |
Pos logit contributions | |
happened | +0.117 |
teacher | +0.113 |
sister | +0.109 |
exam | +0.106 |
problem | +0.094 |
Neg logit contributions | |
From | -0.220 |
oc | -0.197 |
Always | -0.194 |
Flo | -0.194 |
Bug | -0.192 |
Token | with |
Token ID | 351 |
Feature activation | -5.61e+00 |
Pos logit contributions | |
From | +0.084 |
Always | +0.072 |
Since | +0.072 |
Flo | +0.071 |
oc | +0.071 |
Neg logit contributions | |
happened | -0.049 |
teacher | -0.048 |
exam | -0.048 |
sister | -0.045 |
problem | -0.045 |
Token | all |
Token ID | 477 |
Feature activation | -4.63e+00 |
Pos logit contributions | |
From | +0.249 |
Always | +0.215 |
Flo | +0.215 |
Since | +0.213 |
They | +0.210 |
Neg logit contributions | |
happened | -0.142 |
teacher | -0.135 |
exam | -0.133 |
sister | -0.130 |
problem | -0.126 |
Token | her |
Token ID | 607 |
Feature activation | +6.43e+00 |
Pos logit contributions | |
From | +0.204 |
Flo | +0.180 |
Always | +0.176 |
oc | +0.174 |
Since | +0.173 |
Neg logit contributions | |
happened | -0.118 |
teacher | -0.113 |
exam | -0.111 |
sister | -0.111 |
problem | -0.104 |
Token | might |
Token ID | 1244 |
Feature activation | +3.00e+00 |
Pos logit contributions | |
happened | +0.175 |
exam | +0.168 |
teacher | +0.165 |
sister | +0.160 |
problem | +0.157 |
Neg logit contributions | |
From | -0.264 |
Flo | -0.228 |
Always | -0.227 |
They | -0.222 |
Since | -0.221 |
Token | . |
Token ID | 13 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
happened | +0.075 |
teacher | +0.071 |
exam | +0.069 |
sister | +0.066 |
problem | +0.066 |
Neg logit contributions | |
From | -0.135 |
Always | -0.118 |
Flo | -0.116 |
Since | -0.115 |
They | -0.113 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.35e+01 |
Pos logit contributions | |
happened | +0.075 |
teacher | +0.072 |
exam | +0.070 |
sister | +0.068 |
problem | +0.065 |
Neg logit contributions | |
From | -0.055 |
Always | -0.046 |
oc | -0.044 |
Flo | -0.043 |
Since | -0.043 |
Token | a |
Token ID | 257 |
Feature activation | +4.08e+00 |
Pos logit contributions | |
From | +0.580 |
Since | +0.495 |
Flo | +0.493 |
Always | +0.492 |
They | +0.484 |
Neg logit contributions | |
happened | -0.373 |
exam | -0.349 |
teacher | -0.348 |
sister | -0.339 |
problem | -0.337 |
Token | step |
Token ID | 2239 |
Feature activation | -2.36e+00 |
Pos logit contributions | |
happened | +0.141 |
teacher | +0.134 |
exam | +0.133 |
sister | +0.131 |
problem | +0.126 |
Neg logit contributions | |
From | -0.142 |
Flo | -0.119 |
Since | -0.115 |
They | -0.114 |
Always | -0.113 |
Token | was |
Token ID | 373 |
Feature activation | +4.03e+00 |
Pos logit contributions | |
From | +0.099 |
Flo | +0.085 |
Always | +0.085 |
Since | +0.084 |
They | +0.083 |
Neg logit contributions | |
happened | -0.065 |
teacher | -0.061 |
exam | -0.061 |
sister | -0.059 |
problem | -0.058 |
Token | a |
Token ID | 257 |
Feature activation | +4.09e+00 |
Pos logit contributions | |
happened | +0.128 |
teacher | +0.125 |
exam | +0.124 |
sister | +0.123 |
problem | +0.120 |
Neg logit contributions | |
From | -0.148 |
Flo | -0.124 |
Always | -0.123 |
Since | -0.122 |
They | -0.122 |
Token | big |
Token ID | 1263 |
Feature activation | -2.15e+00 |
Pos logit contributions | |
happened | +0.146 |
exam | +0.135 |
teacher | +0.132 |
sister | +0.131 |
classmate | +0.125 |
Neg logit contributions | |
From | -0.136 |
Flo | -0.113 |
Even | -0.109 |
Since | -0.109 |
They | -0.108 |
Token | with |
Token ID | 351 |
Feature activation | -5.59e+00 |
Pos logit contributions | |
sister | +0.205 |
teacher | +0.204 |
happened | +0.199 |
exam | +0.198 |
problem | +0.191 |
Neg logit contributions | |
From | -0.236 |
Even | -0.206 |
Flo | -0.202 |
Since | -0.201 |
Bug | -0.201 |
Token | his |
Token ID | 465 |
Feature activation | +4.58e+00 |
Pos logit contributions | |
From | +0.234 |
Flo | +0.204 |
Always | +0.202 |
Since | +0.201 |
They | +0.197 |
Neg logit contributions | |
happened | -0.159 |
teacher | -0.148 |
sister | -0.143 |
exam | -0.142 |
problem | -0.141 |
Token | fierce |
Token ID | 14800 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
happened | +0.136 |
exam | +0.132 |
teacher | +0.127 |
principal | +0.125 |
classmate | +0.118 |
Neg logit contributions | |
From | -0.163 |
oc | -0.141 |
Flo | -0.140 |
Always | -0.139 |
Squ | -0.135 |
Token | eyes |
Token ID | 2951 |
Feature activation | +3.28e-01 |
Pos logit contributions | |
From | +0.040 |
oc | +0.037 |
Flo | +0.035 |
Squ | +0.034 |
Tweet | +0.034 |
Neg logit contributions | |
happened | -0.038 |
exam | -0.036 |
principal | -0.036 |
teacher | -0.036 |
classmate | -0.034 |
Token | . |
Token ID | 13 |
Feature activation | +1.71e+00 |
Pos logit contributions | |
happened | +0.009 |
teacher | +0.008 |
problem | +0.008 |
exam | +0.008 |
, | +0.008 |
Neg logit contributions | |
From | -0.014 |
Flo | -0.012 |
Bug | -0.012 |
Since | -0.012 |
Looks | -0.012 |
Token | Taking |
Token ID | 20879 |
Feature activation | -1.35e+01 |
Pos logit contributions | |
teacher | +0.063 |
sister | +0.059 |
happened | +0.059 |
exam | +0.056 |
principal | +0.056 |
Neg logit contributions | |
From | -0.053 |
cking | -0.049 |
Always | -0.046 |
Looks | -0.045 |
inning | -0.044 |
Token | a |
Token ID | 257 |
Feature activation | +4.10e+00 |
Pos logit contributions | |
From | +0.593 |
Since | +0.511 |
Flo | +0.505 |
Always | +0.504 |
They | +0.496 |
Neg logit contributions | |
happened | -0.363 |
teacher | -0.333 |
exam | -0.330 |
sister | -0.326 |
problem | -0.323 |
Token | brave |
Token ID | 14802 |
Feature activation | +2.69e-01 |
Pos logit contributions | |
happened | +0.127 |
teacher | +0.121 |
exam | +0.121 |
sister | +0.117 |
problem | +0.113 |
Neg logit contributions | |
From | -0.156 |
Flo | -0.133 |
Always | -0.129 |
Since | -0.128 |
They | -0.128 |
Token | step |
Token ID | 2239 |
Feature activation | -2.43e+00 |
Pos logit contributions | |
happened | +0.007 |
teacher | +0.007 |
exam | +0.007 |
sister | +0.007 |
problem | +0.006 |
Neg logit contributions | |
From | -0.011 |
Flo | -0.010 |
Always | -0.010 |
Since | -0.009 |
They | -0.009 |
Token | forward |
Token ID | 2651 |
Feature activation | +3.86e+00 |
Pos logit contributions | |
From | +0.112 |
Always | +0.097 |
Since | +0.096 |
Flo | +0.096 |
Even | +0.094 |
Neg logit contributions | |
happened | -0.058 |
teacher | -0.055 |
exam | -0.054 |
sister | -0.053 |
problem | -0.052 |
Token | , |
Token ID | 11 |
Feature activation | +4.78e+00 |
Pos logit contributions | |
happened | +0.107 |
teacher | +0.103 |
exam | +0.101 |
sister | +0.099 |
problem | +0.097 |
Neg logit contributions | |
From | -0.157 |
Always | -0.135 |
Flo | -0.135 |
They | -0.133 |
Since | -0.131 |
Token | hoops |
Token ID | 46730 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
From | +0.119 |
Always | +0.101 |
Since | +0.100 |
Flo | +0.099 |
Even | +0.097 |
Neg logit contributions | |
happened | -0.088 |
teacher | -0.086 |
sister | -0.083 |
exam | -0.083 |
problem | -0.081 |
Token | . |
Token ID | 13 |
Feature activation | +1.57e+00 |
Pos logit contributions | |
From | +0.101 |
Always | +0.087 |
Flo | +0.087 |
Since | +0.087 |
They | +0.085 |
Neg logit contributions | |
happened | -0.057 |
teacher | -0.054 |
exam | -0.053 |
sister | -0.052 |
problem | -0.051 |
Token | Tim |
Token ID | 5045 |
Feature activation | +4.41e-01 |
Pos logit contributions | |
happened | +0.051 |
teacher | +0.049 |
Lee | +0.045 |
sister | +0.045 |
cousin | +0.045 |
Neg logit contributions | |
cking | -0.048 |
From | -0.047 |
Always | -0.045 |
inning | -0.045 |
Looks | -0.042 |
Token | my |
Token ID | 1820 |
Feature activation | +3.90e+00 |
Pos logit contributions | |
happened | +0.013 |
teacher | +0.013 |
sister | +0.013 |
exam | +0.012 |
problem | +0.012 |
Neg logit contributions | |
From | -0.017 |
Flo | -0.015 |
Since | -0.015 |
Always | -0.015 |
They | -0.014 |
Token | was |
Token ID | 373 |
Feature activation | +3.81e+00 |
Pos logit contributions | |
happened | +0.122 |
teacher | +0.117 |
exam | +0.116 |
sister | +0.113 |
problem | +0.111 |
Neg logit contributions | |
From | -0.148 |
Always | -0.125 |
Flo | -0.124 |
Since | -0.123 |
They | -0.122 |
Token | so |
Token ID | 523 |
Feature activation | +5.40e+00 |
Pos logit contributions | |
happened | +0.141 |
exam | +0.137 |
teacher | +0.132 |
problem | +0.131 |
sister | +0.127 |
Neg logit contributions | |
From | -0.127 |
Since | -0.103 |
They | -0.101 |
Always | -0.101 |
Flo | -0.099 |
Token | proud |
Token ID | 6613 |
Feature activation | -6.28e-01 |
Pos logit contributions | |
happened | +0.169 |
teacher | +0.162 |
exam | +0.160 |
sister | +0.157 |
problem | +0.154 |
Neg logit contributions | |
From | -0.207 |
Always | -0.175 |
Flo | -0.175 |
Since | -0.173 |
They | -0.170 |
Token | of |
Token ID | 286 |
Feature activation | -4.25e+00 |
Pos logit contributions | |
From | +0.023 |
Flo | +0.019 |
Always | +0.019 |
Since | +0.019 |
Even | +0.018 |
Neg logit contributions | |
happened | -0.021 |
teacher | -0.021 |
exam | -0.020 |
sister | -0.020 |
problem | -0.019 |
Token | Pepper |
Token ID | 24346 |
Feature activation | -2.03e+00 |
Pos logit contributions | |
From | +0.172 |
ksh | +0.168 |
rig | +0.164 |
plets | +0.163 |
oc | +0.161 |
Neg logit contributions | |
teacher | -0.107 |
happened | -0.103 |
sister | -0.101 |
exam | -0.099 |
school | -0.098 |
Token | and |
Token ID | 290 |
Feature activation | -7.91e+00 |
Pos logit contributions | |
From | +0.084 |
Since | +0.072 |
Always | +0.072 |
Even | +0.072 |
Flo | +0.072 |
Neg logit contributions | |
happened | -0.058 |
teacher | -0.056 |
sister | -0.055 |
exam | -0.053 |
problem | -0.051 |
Token | gave |
Token ID | 2921 |
Feature activation | -2.18e+00 |
Pos logit contributions | |
From | +0.293 |
Always | +0.238 |
They | +0.238 |
Since | +0.237 |
Flo | +0.236 |
Neg logit contributions | |
happened | -0.258 |
exam | -0.256 |
teacher | -0.246 |
problem | -0.246 |
sister | -0.236 |
Token | said |
Token ID | 531 |
Feature activation | +7.68e+00 |
Pos logit contributions | |
From | +0.300 |
Always | +0.248 |
Flo | +0.247 |
Since | +0.247 |
They | +0.245 |
Neg logit contributions | |
happened | -0.244 |
exam | -0.235 |
teacher | -0.233 |
problem | -0.227 |
sister | -0.226 |
Token | , |
Token ID | 11 |
Feature activation | +4.67e+00 |
Pos logit contributions | |
happened | +0.205 |
exam | +0.198 |
cousin | +0.181 |
problem | +0.181 |
principal | +0.181 |
Neg logit contributions | |
From | -0.319 |
They | -0.270 |
From | -0.265 |
He | -0.263 |
The | -0.262 |
Token | " |
Token ID | 366 |
Feature activation | -3.75e+00 |
Pos logit contributions | |
teacher | +0.114 |
happened | +0.110 |
sister | +0.103 |
exam | +0.096 |
dad | +0.096 |
Neg logit contributions | |
From | -0.198 |
Always | -0.189 |
Flo | -0.183 |
pill | -0.181 |
oc | -0.179 |
Token | Thanks |
Token ID | 9690 |
Feature activation | -7.51e-01 |
Pos logit contributions | |
rig | +0.099 |
From | +0.098 |
fet | +0.097 |
pering | +0.095 |
Flo | +0.094 |
Neg logit contributions | |
teacher | -0.127 |
happened | -0.124 |
ha | -0.115 |
classmate | -0.114 |
exam | -0.114 |
Token | , |
Token ID | 11 |
Feature activation | +4.62e+00 |
Pos logit contributions | |
From | +0.028 |
Flo | +0.025 |
They | +0.025 |
From | +0.023 |
They | +0.023 |
Neg logit contributions | |
happened | -0.024 |
classmate | -0.019 |
problem | -0.019 |
principal | -0.018 |
exam | -0.018 |
Token | Mom |
Token ID | 11254 |
Feature activation | +7.25e+00 |
Pos logit contributions | |
teacher | +0.078 |
sister | +0.073 |
exam | +0.072 |
happened | +0.070 |
problem | +0.064 |
Neg logit contributions | |
From | -0.243 |
Always | -0.221 |
Flo | -0.220 |
Since | -0.216 |
They | -0.215 |
Token | . |
Token ID | 13 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
happened | +0.258 |
exam | +0.246 |
teacher | +0.245 |
problem | +0.235 |
classmate | +0.234 |
Neg logit contributions | |
From | -0.230 |
They | -0.197 |
Always | -0.196 |
oc | -0.187 |
Since | -0.185 |
Token | Can |
Token ID | 1680 |
Feature activation | +2.94e+00 |
Pos logit contributions | |
sister | +0.043 |
teacher | +0.043 |
happened | +0.039 |
exam | +0.039 |
brother | +0.037 |
Neg logit contributions | |
From | -0.056 |
cking | -0.054 |
ized | -0.051 |
Always | -0.050 |
Flo | -0.050 |
Token | I |
Token ID | 314 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
teacher | +0.071 |
sister | +0.067 |
exam | +0.067 |
ha | +0.061 |
dad | +0.059 |
Neg logit contributions | |
rig | -0.127 |
oc | -0.125 |
From | -0.124 |
valleys | -0.122 |
inning | -0.118 |
Token | please |
Token ID | 3387 |
Feature activation | +5.26e+00 |
Pos logit contributions | |
From | +0.065 |
Flo | +0.055 |
Always | +0.055 |
Since | +0.055 |
They | +0.054 |
Neg logit contributions | |
happened | -0.048 |
teacher | -0.047 |
exam | -0.046 |
sister | -0.045 |
problem | -0.044 |
Token | have |
Token ID | 423 |
Feature activation | -3.64e+00 |
Pos logit contributions | |
teacher | +0.147 |
exam | +0.145 |
happened | +0.145 |
sister | +0.141 |
problem | +0.133 |
Neg logit contributions | |
From | -0.218 |
Flo | -0.187 |
Since | -0.186 |
Always | -0.184 |
They | -0.180 |
Token | so |
Token ID | 523 |
Feature activation | +5.41e+00 |
Pos logit contributions | |
happened | +0.138 |
teacher | +0.136 |
sister | +0.131 |
exam | +0.126 |
problem | +0.122 |
Neg logit contributions | |
From | -0.176 |
Flo | -0.155 |
Always | -0.154 |
Bug | -0.150 |
oc | -0.149 |
Token | Tom |
Token ID | 4186 |
Feature activation | +2.23e+00 |
Pos logit contributions | |
exam | +0.198 |
happened | +0.192 |
teacher | +0.187 |
problem | +0.186 |
sister | +0.184 |
Neg logit contributions | |
From | -0.176 |
They | -0.143 |
Since | -0.137 |
Flo | -0.134 |
Always | -0.131 |
Token | put |
Token ID | 1234 |
Feature activation | -7.81e+00 |
Pos logit contributions | |
happened | +0.075 |
teacher | +0.073 |
exam | +0.071 |
sister | +0.071 |
problem | +0.069 |
Neg logit contributions | |
From | -0.079 |
Flo | -0.066 |
Always | -0.066 |
Since | -0.065 |
They | -0.064 |
Token | the |
Token ID | 262 |
Feature activation | -1.50e+00 |
Pos logit contributions | |
From | +0.313 |
Flo | +0.266 |
Always | +0.264 |
Since | +0.263 |
They | +0.259 |
Neg logit contributions | |
happened | -0.234 |
teacher | -0.224 |
exam | -0.221 |
sister | -0.219 |
problem | -0.213 |
Token | string |
Token ID | 4731 |
Feature activation | -7.73e+00 |
Pos logit contributions | |
From | +0.040 |
Flo | +0.032 |
Always | +0.031 |
Since | +0.031 |
They | +0.030 |
Neg logit contributions | |
happened | -0.064 |
teacher | -0.062 |
exam | -0.062 |
sister | -0.061 |
problem | -0.061 |
Token | back |
Token ID | 736 |
Feature activation | +5.69e+00 |
Pos logit contributions | |
From | +0.386 |
Since | +0.338 |
Always | +0.337 |
Flo | +0.335 |
oc | +0.332 |
Neg logit contributions | |
happened | -0.152 |
teacher | -0.145 |
sister | -0.142 |
exam | -0.141 |
problem | -0.132 |
Token | into |
Token ID | 656 |
Feature activation | +6.52e-01 |
Pos logit contributions | |
happened | +0.173 |
exam | +0.165 |
teacher | +0.165 |
sister | +0.161 |
problem | +0.157 |
Neg logit contributions | |
From | -0.223 |
Flo | -0.189 |
Since | -0.188 |
Always | -0.187 |
They | -0.183 |
Token | his |
Token ID | 465 |
Feature activation | +4.46e+00 |
Pos logit contributions | |
happened | +0.018 |
teacher | +0.018 |
sister | +0.018 |
exam | +0.017 |
problem | +0.017 |
Neg logit contributions | |
From | -0.027 |
Flo | -0.023 |
Always | -0.023 |
Since | -0.023 |
They | -0.022 |
Token | toy |
Token ID | 13373 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
happened | +0.132 |
teacher | +0.125 |
exam | +0.124 |
sister | +0.120 |
problem | +0.119 |
Neg logit contributions | |
From | -0.178 |
Always | -0.151 |
Flo | -0.151 |
Since | -0.149 |
They | -0.147 |
Token | box |
Token ID | 3091 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
happened | +0.042 |
teacher | +0.042 |
exam | +0.041 |
sister | +0.040 |
problem | +0.039 |
Neg logit contributions | |
From | -0.070 |
Flo | -0.060 |
Since | -0.060 |
Always | -0.059 |
Even | -0.058 |
Token | . |
Token ID | 13 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
happened | +0.038 |
teacher | +0.037 |
sister | +0.036 |
exam | +0.036 |
problem | +0.035 |
Neg logit contributions | |
From | -0.048 |
Flo | -0.041 |
Since | -0.040 |
Always | -0.040 |
Even | -0.039 |
Token | know |
Token ID | 760 |
Feature activation | +5.35e+00 |
Pos logit contributions | |
From | +0.015 |
They | +0.014 |
Flo | +0.013 |
Always | +0.013 |
Since | +0.012 |
Neg logit contributions | |
happened | -0.014 |
problem | -0.013 |
exam | -0.013 |
classmate | -0.013 |
teacher | -0.013 |
Token | that |
Token ID | 326 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
happened | +0.191 |
exam | +0.185 |
teacher | +0.185 |
sister | +0.181 |
problem | +0.178 |
Neg logit contributions | |
From | -0.177 |
Flo | -0.145 |
They | -0.143 |
Since | -0.143 |
Always | -0.143 |
Token | the |
Token ID | 262 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
From | +0.043 |
They | +0.036 |
Since | +0.035 |
Always | +0.035 |
Flo | +0.034 |
Neg logit contributions | |
happened | -0.042 |
exam | -0.042 |
teacher | -0.041 |
sister | -0.040 |
classmate | -0.039 |
Token | park |
Token ID | 3952 |
Feature activation | -8.08e-01 |
Pos logit contributions | |
From | +0.051 |
Always | +0.042 |
Since | +0.042 |
Flo | +0.042 |
They | +0.042 |
Neg logit contributions | |
happened | -0.053 |
exam | -0.051 |
teacher | -0.051 |
sister | -0.050 |
problem | -0.048 |
Token | was |
Token ID | 373 |
Feature activation | +3.89e+00 |
Pos logit contributions | |
From | +0.030 |
They | +0.026 |
Always | +0.025 |
Flo | +0.025 |
Since | +0.025 |
Neg logit contributions | |
happened | -0.024 |
exam | -0.024 |
teacher | -0.023 |
sister | -0.023 |
problem | -0.022 |
Token | very |
Token ID | 845 |
Feature activation | +6.72e+00 |
Pos logit contributions | |
happened | +0.122 |
exam | +0.120 |
teacher | +0.120 |
sister | +0.116 |
classmate | +0.112 |
Neg logit contributions | |
From | -0.141 |
Always | -0.118 |
They | -0.118 |
Flo | -0.117 |
Since | -0.116 |
Token | dry |
Token ID | 5894 |
Feature activation | -5.36e+00 |
Pos logit contributions | |
happened | +0.204 |
teacher | +0.196 |
exam | +0.194 |
sister | +0.190 |
problem | +0.185 |
Neg logit contributions | |
From | -0.262 |
Flo | -0.221 |
Always | -0.220 |
Since | -0.219 |
They | -0.217 |
Token | and |
Token ID | 290 |
Feature activation | -8.27e+00 |
Pos logit contributions | |
From | +0.204 |
Flo | +0.171 |
Always | +0.170 |
Since | +0.169 |
Even | +0.165 |
Neg logit contributions | |
happened | -0.170 |
teacher | -0.163 |
exam | -0.160 |
sister | -0.158 |
problem | -0.156 |
Token | had |
Token ID | 550 |
Feature activation | +3.65e-01 |
Pos logit contributions | |
From | +0.332 |
Flo | +0.289 |
Always | +0.285 |
Since | +0.281 |
oc | +0.276 |
Neg logit contributions | |
happened | -0.261 |
teacher | -0.238 |
problem | -0.237 |
sister | -0.237 |
exam | -0.231 |
Token | many |
Token ID | 867 |
Feature activation | -4.57e+00 |
Pos logit contributions | |
happened | +0.011 |
exam | +0.010 |
teacher | +0.010 |
sister | +0.010 |
problem | +0.010 |
Neg logit contributions | |
From | -0.014 |
Flo | -0.012 |
Always | -0.012 |
They | -0.012 |
Since | -0.012 |
Token | trees |
Token ID | 7150 |
Feature activation | -3.08e+00 |
Pos logit contributions | |
From | +0.177 |
Always | +0.149 |
They | +0.149 |
Flo | +0.147 |
Since | +0.147 |
Neg logit contributions | |
happened | -0.134 |
exam | -0.131 |
teacher | -0.130 |
sister | -0.126 |
classmate | -0.121 |
Token | house |
Token ID | 2156 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
happened | +0.221 |
exam | +0.199 |
principal | +0.189 |
teacher | +0.188 |
grade | +0.186 |
Neg logit contributions | |
From | -0.209 |
Flo | -0.188 |
They | -0.185 |
The | -0.184 |
Always | -0.180 |
Token | . |
Token ID | 13 |
Feature activation | +2.30e+00 |
Pos logit contributions | |
happened | +0.052 |
teacher | +0.049 |
exam | +0.049 |
sister | +0.048 |
problem | +0.047 |
Neg logit contributions | |
From | -0.082 |
Always | -0.070 |
They | -0.070 |
Since | -0.070 |
Flo | -0.069 |
Token | She |
Token ID | 1375 |
Feature activation | -7.73e+00 |
Pos logit contributions | |
happened | +0.097 |
teacher | +0.095 |
exam | +0.093 |
sister | +0.092 |
problem | +0.090 |
Neg logit contributions | |
From | -0.062 |
Always | -0.049 |
Flo | -0.048 |
Since | -0.048 |
They | -0.047 |
Token | didn |
Token ID | 1422 |
Feature activation | +2.37e+00 |
Pos logit contributions | |
From | +0.291 |
Flo | +0.240 |
Always | +0.239 |
They | +0.236 |
Since | +0.232 |
Neg logit contributions | |
exam | -0.248 |
happened | -0.240 |
teacher | -0.239 |
problem | -0.226 |
sister | -0.224 |
Token | 't |
Token ID | 470 |
Feature activation | -1.08e-03 |
Pos logit contributions | |
happened | +0.070 |
teacher | +0.067 |
exam | +0.066 |
problem | +0.062 |
sister | +0.061 |
Neg logit contributions | |
From | -0.091 |
Always | -0.081 |
Flo | -0.080 |
oc | -0.079 |
Since | -0.077 |
Token | know |
Token ID | 760 |
Feature activation | +5.62e+00 |
Pos logit contributions | |
From | +0.000 |
Always | +0.000 |
They | +0.000 |
Flo | +0.000 |
Since | +0.000 |
Neg logit contributions | |
happened | -0.000 |
teacher | -0.000 |
exam | -0.000 |
problem | -0.000 |
sister | -0.000 |
Token | what |
Token ID | 644 |
Feature activation | +8.09e-01 |
Pos logit contributions | |
happened | +0.192 |
teacher | +0.187 |
exam | +0.184 |
sister | +0.181 |
problem | +0.179 |
Neg logit contributions | |
From | -0.194 |
Flo | -0.160 |
Always | -0.159 |
Since | -0.159 |
They | -0.158 |
Token | to |
Token ID | 284 |
Feature activation | +4.85e+00 |
Pos logit contributions | |
sister | +0.022 |
exam | +0.021 |
classmate | +0.021 |
problem | +0.021 |
teacher | +0.021 |
Neg logit contributions | |
From | -0.031 |
The | -0.028 |
Always | -0.027 |
Flo | -0.027 |
Bug | -0.027 |
Token | do |
Token ID | 466 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
classmate | +0.177 |
principal | +0.176 |
sister | +0.169 |
servant | +0.167 |
problem | +0.163 |
Neg logit contributions | |
From | -0.134 |
They | -0.121 |
The | -0.112 |
Always | -0.108 |
This | -0.107 |
Token | and |
Token ID | 290 |
Feature activation | -7.60e+00 |
Pos logit contributions | |
sister | +0.076 |
teacher | +0.072 |
classmate | +0.072 |
exam | +0.071 |
problem | +0.070 |
Neg logit contributions | |
From | -0.064 |
Always | -0.055 |
They | -0.053 |
Since | -0.050 |
Now | -0.050 |
Token | started |
Token ID | 2067 |
Feature activation | +6.14e+00 |
Pos logit contributions | |
From | +0.255 |
They | +0.205 |
From | +0.202 |
Always | +0.201 |
Since | +0.199 |
Neg logit contributions | |
exam | -0.248 |
classmate | -0.242 |
teacher | -0.242 |
guest | -0.238 |
problem | -0.236 |
Token | then |
Token ID | 788 |
Feature activation | +6.94e-01 |
Pos logit contributions | |
From | +0.084 |
Always | +0.065 |
Since | +0.064 |
Flo | +0.064 |
They | +0.063 |
Neg logit contributions | |
happened | -0.122 |
teacher | -0.119 |
exam | -0.119 |
sister | -0.115 |
problem | -0.115 |
Token | , |
Token ID | 11 |
Feature activation | +4.83e+00 |
Pos logit contributions | |
exam | +0.022 |
teacher | +0.021 |
principal | +0.021 |
guest | +0.020 |
classmate | +0.020 |
Neg logit contributions | |
From | -0.026 |
They | -0.022 |
oc | -0.021 |
From | -0.021 |
The | -0.020 |
Token | Tim |
Token ID | 5045 |
Feature activation | +4.35e-01 |
Pos logit contributions | |
happened | +0.177 |
exam | +0.176 |
teacher | +0.176 |
problem | +0.170 |
sister | +0.169 |
Neg logit contributions | |
From | -0.155 |
Always | -0.124 |
They | -0.123 |
Since | -0.123 |
Flo | -0.121 |
Token | my |
Token ID | 1820 |
Feature activation | +3.99e+00 |
Pos logit contributions | |
exam | +0.018 |
happened | +0.018 |
teacher | +0.018 |
haircut | +0.018 |
guest | +0.017 |
Neg logit contributions | |
From | -0.010 |
Always | -0.009 |
They | -0.008 |
He | -0.008 |
The | -0.007 |
Token | 's |
Token ID | 338 |
Feature activation | +4.10e+00 |
Pos logit contributions | |
exam | +0.146 |
teacher | +0.130 |
guest | +0.130 |
happened | +0.130 |
haircut | +0.129 |
Neg logit contributions | |
From | -0.129 |
They | -0.109 |
Always | -0.105 |
He | -0.102 |
From | -0.101 |
Token | other |
Token ID | 584 |
Feature activation | +3.86e+00 |
Pos logit contributions | |
happened | +0.097 |
exam | +0.081 |
register | +0.070 |
grade | +0.069 |
principal | +0.069 |
Neg logit contributions | |
From | -0.160 |
Squ | -0.148 |
He | -0.143 |
The | -0.142 |
Tweet | -0.141 |
Token | friend |
Token ID | 1545 |
Feature activation | +5.69e+00 |
Pos logit contributions | |
exam | +0.086 |
happened | +0.080 |
grade | +0.072 |
haircut | +0.072 |
teacher | +0.071 |
Neg logit contributions | |
From | -0.162 |
oc | -0.151 |
They | -0.150 |
They | -0.147 |
From | -0.144 |
Token | Sally |
Token ID | 25737 |
Feature activation | +2.17e+00 |
Pos logit contributions | |
exam | +0.219 |
haircut | +0.203 |
grade | +0.198 |
guest | +0.195 |
hairc | +0.193 |
Neg logit contributions | |
From | -0.142 |
From | -0.130 |
They | -0.126 |
They | -0.126 |
He | -0.120 |
Token | gave |
Token ID | 2921 |
Feature activation | -2.24e+00 |
Pos logit contributions | |
exam | +0.067 |
haircut | +0.064 |
teacher | +0.063 |
guest | +0.058 |
problem | +0.058 |
Neg logit contributions | |
From | -0.076 |
Always | -0.068 |
They | -0.067 |
fet | -0.065 |
From | -0.064 |
Token | him |
Token ID | 683 |
Feature activation | +6.48e+00 |
Pos logit contributions | |
From | +0.082 |
Always | +0.070 |
Flo | +0.069 |
They | +0.068 |
Since | +0.067 |
Neg logit contributions | |
happened | -0.073 |
teacher | -0.071 |
exam | -0.070 |
problem | -0.068 |
sister | -0.068 |
Token | a |
Token ID | 257 |
Feature activation | +4.04e+00 |
Pos logit contributions | |
happened | +0.199 |
teacher | +0.187 |
exam | +0.185 |
problem | +0.181 |
sister | +0.178 |
Neg logit contributions | |
From | -0.252 |
Always | -0.213 |
Flo | -0.211 |
Since | -0.209 |
They | -0.208 |
Token |
|
Token ID | 198 |
Feature activation | -2.65e+00 |
Pos logit contributions | |
teacher | +0.101 |
happened | +0.100 |
sister | +0.097 |
dad | +0.095 |
daughter | +0.088 |
Neg logit contributions | |
From | -0.134 |
cking | -0.132 |
Looks | -0.131 |
Flo | -0.126 |
inning | -0.125 |
Token |
|
Token ID | 198 |
Feature activation | -2.68e+00 |
Pos logit contributions | |
Looks | +0.086 |
From | +0.086 |
ksh | +0.085 |
Always | +0.083 |
Eagles | +0.082 |
Neg logit contributions | |
teacher | -0.080 |
happened | -0.077 |
exam | -0.068 |
dad | -0.067 |
sister | -0.067 |
Token | And |
Token ID | 1870 |
Feature activation | -6.46e+00 |
Pos logit contributions | |
ksh | +0.065 |
Eagles | +0.054 |
rant | +0.054 |
plets | +0.054 |
spe | +0.052 |
Neg logit contributions | |
teacher | -0.112 |
ha | -0.097 |
principal | -0.094 |
classmate | -0.093 |
happened | -0.093 |
Token | that |
Token ID | 326 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
From | +0.244 |
Flo | +0.216 |
Always | +0.215 |
Bug | +0.215 |
Even | +0.212 |
Neg logit contributions | |
happened | -0.192 |
teacher | -0.190 |
sister | -0.184 |
exam | -0.175 |
problem | -0.171 |
Token | was |
Token ID | 373 |
Feature activation | +3.76e+00 |
Pos logit contributions | |
From | +0.060 |
Flo | +0.052 |
Always | +0.051 |
Since | +0.051 |
They | +0.050 |
Neg logit contributions | |
happened | -0.040 |
sister | -0.036 |
teacher | -0.036 |
problem | -0.036 |
exam | -0.035 |
Token | how |
Token ID | 703 |
Feature activation | +9.37e-01 |
Pos logit contributions | |
happened | +0.117 |
teacher | +0.109 |
exam | +0.107 |
sister | +0.106 |
problem | +0.105 |
Neg logit contributions | |
From | -0.145 |
Flo | -0.123 |
Always | -0.123 |
Since | -0.121 |
They | -0.120 |
Token | the |
Token ID | 262 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
happened | +0.030 |
teacher | +0.028 |
sister | +0.028 |
exam | +0.027 |
problem | +0.027 |
Neg logit contributions | |
From | -0.036 |
Always | -0.031 |
Flo | -0.030 |
Since | -0.030 |
They | -0.030 |
Token | little |
Token ID | 1310 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
From | +0.055 |
Since | +0.045 |
oc | +0.044 |
Always | +0.044 |
The | +0.043 |
Neg logit contributions | |
happened | -0.047 |
exam | -0.043 |
teacher | -0.042 |
principal | -0.041 |
classmate | -0.040 |
Token | girl |
Token ID | 2576 |
Feature activation | +6.20e+00 |
Pos logit contributions | |
From | +0.055 |
Always | +0.048 |
Since | +0.046 |
The | +0.045 |
oc | +0.045 |
Neg logit contributions | |
happened | -0.038 |
exam | -0.035 |
teacher | -0.034 |
gh | -0.033 |
principal | -0.033 |
Token | learned |
Token ID | 4499 |
Feature activation | +8.52e-01 |
Pos logit contributions | |
happened | +0.181 |
teacher | +0.163 |
sister | +0.162 |
exam | +0.148 |
problem | +0.143 |
Neg logit contributions | |
From | -0.261 |
Flo | -0.233 |
Ho | -0.228 |
They | -0.227 |
Since | -0.226 |
Token | her |
Token ID | 607 |
Feature activation | +6.13e+00 |
Pos logit contributions | |
happened | +0.023 |
teacher | +0.022 |
exam | +0.022 |
sister | +0.021 |
problem | +0.021 |
Neg logit contributions | |
From | -0.036 |
Flo | -0.031 |
Always | -0.031 |
Since | -0.030 |
oc | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | +9.64e-01 |
Pos logit contributions | |
From | +0.315 |
Since | +0.273 |
Looks | +0.273 |
Even | +0.273 |
Grab | +0.261 |
Neg logit contributions | |
teacher | -0.159 |
sister | -0.155 |
problem | -0.153 |
exam | -0.152 |
happened | -0.150 |
Token | He |
Token ID | 679 |
Feature activation | -7.67e+00 |
Pos logit contributions | |
sister | +0.032 |
teacher | +0.032 |
brother | +0.030 |
Lee | +0.029 |
principal | +0.029 |
Neg logit contributions | |
From | -0.030 |
ized | -0.030 |
cking | -0.029 |
Always | -0.027 |
rig | -0.027 |
Token | will |
Token ID | 481 |
Feature activation | +2.56e+00 |
Pos logit contributions | |
From | +0.355 |
Since | +0.319 |
Even | +0.303 |
Always | +0.300 |
They | +0.299 |
Neg logit contributions | |
happened | -0.200 |
problem | -0.184 |
teacher | -0.182 |
sister | -0.180 |
exam | -0.162 |
Token | be |
Token ID | 307 |
Feature activation | +5.38e-01 |
Pos logit contributions | |
teacher | +0.066 |
exam | +0.060 |
happened | +0.059 |
sister | +0.059 |
problem | +0.056 |
Neg logit contributions | |
From | -0.115 |
Since | -0.102 |
pill | -0.102 |
Always | -0.102 |
Looks | -0.100 |
Token | angry |
Token ID | 7954 |
Feature activation | +3.91e+00 |
Pos logit contributions | |
happened | +0.014 |
teacher | +0.014 |
sister | +0.014 |
problem | +0.013 |
exam | +0.013 |
Neg logit contributions | |
From | -0.023 |
Since | -0.020 |
Always | -0.020 |
Flo | -0.020 |
Even | -0.019 |
Token | if |
Token ID | 611 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
happened | +0.131 |
teacher | +0.125 |
exam | +0.125 |
sister | +0.121 |
problem | +0.120 |
Neg logit contributions | |
From | -0.140 |
Always | -0.117 |
Flo | -0.117 |
They | -0.116 |
Since | -0.115 |
Token | we |
Token ID | 356 |
Feature activation | -5.59e-01 |
Pos logit contributions | |
happened | +0.038 |
teacher | +0.036 |
exam | +0.036 |
sister | +0.035 |
problem | +0.035 |
Neg logit contributions | |
From | -0.036 |
Flo | -0.029 |
Always | -0.029 |
Since | -0.029 |
They | -0.029 |
Token | mess |
Token ID | 2085 |
Feature activation | +6.89e-02 |
Pos logit contributions | |
From | +0.023 |
Since | +0.019 |
Always | +0.019 |
Flo | +0.019 |
Even | +0.018 |
Neg logit contributions | |
happened | -0.017 |
teacher | -0.016 |
exam | -0.016 |
sister | -0.016 |
problem | -0.015 |
Token | it |
Token ID | 340 |
Feature activation | +1.51e+00 |
Pos logit contributions | |
happened | +0.002 |
teacher | +0.002 |
exam | +0.002 |
problem | +0.002 |
sister | +0.002 |
Neg logit contributions | |
From | -0.002 |
They | -0.002 |
Flo | -0.002 |
Always | -0.002 |
Since | -0.002 |
Token | up |
Token ID | 510 |
Feature activation | +2.03e+00 |
Pos logit contributions | |
happened | +0.047 |
teacher | +0.046 |
exam | +0.045 |
sister | +0.044 |
problem | +0.043 |
Neg logit contributions | |
From | -0.057 |
Flo | -0.049 |
Always | -0.048 |
They | -0.047 |
Since | -0.047 |
Token | ." |
Token ID | 526 |
Feature activation | +2.77e+00 |
Pos logit contributions | |
happened | +0.051 |
teacher | +0.049 |
sister | +0.048 |
problem | +0.048 |
exam | +0.047 |
Neg logit contributions | |
From | -0.090 |
Since | -0.077 |
Always | -0.077 |
Even | -0.077 |
Flo | -0.077 |
Token |
|
Token ID | 198 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
happened | +0.157 |
teacher | +0.156 |
sister | +0.151 |
exam | +0.149 |
problem | +0.143 |
Neg logit contributions | |
From | -0.115 |
Always | -0.095 |
Flo | -0.094 |
Since | -0.091 |
Tweet | -0.087 |
Token |
|
Token ID | 198 |
Feature activation | -2.32e+00 |
Pos logit contributions | |
From | +0.062 |
Always | +0.051 |
Flo | +0.050 |
Since | +0.049 |
Looks | +0.047 |
Neg logit contributions | |
happened | -0.093 |
teacher | -0.093 |
sister | -0.089 |
exam | -0.088 |
problem | -0.086 |
Token | M |
Token ID | 44 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
From | +0.064 |
Always | +0.054 |
Flo | +0.053 |
Since | +0.050 |
Looks | +0.049 |
Neg logit contributions | |
teacher | -0.093 |
happened | -0.091 |
sister | -0.088 |
exam | -0.087 |
problem | -0.085 |
Token | ama |
Token ID | 1689 |
Feature activation | +2.45e+00 |
Pos logit contributions | |
happened | +0.060 |
exam | +0.055 |
teacher | +0.054 |
sister | +0.054 |
classmate | +0.053 |
Neg logit contributions | |
From | -0.076 |
They | -0.063 |
Since | -0.062 |
Flo | -0.061 |
oc | -0.061 |
Token | smiled |
Token ID | 13541 |
Feature activation | -1.36e-01 |
Pos logit contributions | |
happened | +0.068 |
teacher | +0.065 |
sister | +0.063 |
exam | +0.061 |
problem | +0.060 |
Neg logit contributions | |
From | -0.103 |
Always | -0.090 |
Flo | -0.089 |
Since | -0.088 |
Even | -0.086 |
Token | . |
Token ID | 13 |
Feature activation | +1.56e+00 |
Pos logit contributions | |
From | +0.004 |
Always | +0.003 |
Flo | +0.003 |
Since | +0.003 |
They | +0.003 |
Neg logit contributions | |
happened | -0.006 |
exam | -0.005 |
teacher | -0.005 |
sister | -0.005 |
problem | -0.005 |
Token | " |
Token ID | 366 |
Feature activation | -3.65e+00 |
Pos logit contributions | |
happened | +0.051 |
teacher | +0.050 |
sister | +0.047 |
exam | +0.045 |
problem | +0.044 |
Neg logit contributions | |
From | -0.055 |
Always | -0.050 |
Flo | -0.049 |
Tweet | -0.049 |
Hop | -0.047 |
Token | That |
Token ID | 2504 |
Feature activation | -3.97e+00 |
Pos logit contributions | |
oc | +0.092 |
bing | +0.091 |
athed | +0.091 |
Flo | +0.091 |
From | +0.091 |
Neg logit contributions | |
happened | -0.128 |
teacher | -0.127 |
exam | -0.119 |
ha | -0.116 |
haircut | -0.116 |
Token | 's |
Token ID | 338 |
Feature activation | +3.99e+00 |
Pos logit contributions | |
From | +0.141 |
Always | +0.118 |
They | +0.117 |
Flo | +0.115 |
Since | +0.114 |
Neg logit contributions | |
happened | -0.127 |
teacher | -0.124 |
exam | -0.123 |
sister | -0.118 |
classmate | -0.116 |
Token | so |
Token ID | 523 |
Feature activation | +5.43e+00 |
Pos logit contributions | |
happened | +0.105 |
exam | +0.099 |
teacher | +0.099 |
problem | +0.097 |
sister | +0.096 |
Neg logit contributions | |
From | -0.172 |
Flo | -0.150 |
Always | -0.150 |
Since | -0.148 |
oc | -0.145 |
Token | lovely |
Token ID | 14081 |
Feature activation | -2.63e+00 |
Pos logit contributions | |
happened | +0.136 |
teacher | +0.132 |
exam | +0.129 |
problem | +0.129 |
sister | +0.126 |
Neg logit contributions | |
From | -0.239 |
Always | -0.214 |
Since | -0.214 |
Flo | -0.210 |
Even | -0.201 |
Token | tracks |
Token ID | 8339 |
Feature activation | -6.23e+00 |
Pos logit contributions | |
From | +0.045 |
Flo | +0.037 |
Always | +0.036 |
Since | +0.036 |
They | +0.036 |
Neg logit contributions | |
happened | -0.050 |
teacher | -0.047 |
exam | -0.047 |
sister | -0.047 |
problem | -0.046 |
Token | . |
Token ID | 13 |
Feature activation | +1.35e+00 |
Pos logit contributions | |
From | +0.296 |
Flo | +0.261 |
Since | +0.257 |
Always | +0.256 |
Even | +0.251 |
Neg logit contributions | |
happened | -0.142 |
teacher | -0.132 |
exam | -0.129 |
sister | -0.124 |
problem | -0.123 |
Token | The |
Token ID | 383 |
Feature activation | -9.80e+00 |
Pos logit contributions | |
teacher | +0.053 |
sister | +0.052 |
happened | +0.052 |
classmate | +0.047 |
brother | +0.047 |
Neg logit contributions | |
From | -0.039 |
Always | -0.036 |
cking | -0.034 |
Flo | -0.033 |
inning | -0.033 |
Token | train |
Token ID | 4512 |
Feature activation | -3.36e+00 |
Pos logit contributions | |
From | +0.381 |
Flo | +0.322 |
Since | +0.319 |
Always | +0.319 |
They | +0.316 |
Neg logit contributions | |
happened | -0.300 |
exam | -0.282 |
teacher | -0.275 |
sister | -0.266 |
problem | -0.259 |
Token | had |
Token ID | 550 |
Feature activation | +4.12e-01 |
Pos logit contributions | |
From | +0.138 |
Always | +0.120 |
Flo | +0.119 |
Since | +0.118 |
They | +0.115 |
Neg logit contributions | |
happened | -0.098 |
teacher | -0.093 |
sister | -0.088 |
problem | -0.087 |
exam | -0.085 |
Token | stopped |
Token ID | 5025 |
Feature activation | +3.51e+00 |
Pos logit contributions | |
happened | +0.012 |
exam | +0.012 |
teacher | +0.012 |
sister | +0.011 |
problem | +0.011 |
Neg logit contributions | |
From | -0.016 |
Flo | -0.014 |
Always | -0.014 |
They | -0.014 |
Since | -0.013 |
Token | to |
Token ID | 284 |
Feature activation | +4.61e+00 |
Pos logit contributions | |
happened | +0.105 |
sister | +0.097 |
problem | +0.096 |
teacher | +0.091 |
brother | +0.090 |
Neg logit contributions | |
From | -0.151 |
Since | -0.132 |
Even | -0.130 |
Together | -0.126 |
Flo | -0.125 |
Token | keep |
Token ID | 1394 |
Feature activation | -4.56e+00 |
Pos logit contributions | |
happened | +0.123 |
teacher | +0.116 |
exam | +0.112 |
sister | +0.109 |
problem | +0.100 |
Neg logit contributions | |
From | -0.202 |
Since | -0.178 |
Always | -0.178 |
pill | -0.177 |
Even | -0.172 |
Token | the |
Token ID | 262 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
From | +0.192 |
Since | +0.165 |
Always | +0.162 |
Even | +0.160 |
Flo | +0.157 |
Neg logit contributions | |
happened | -0.134 |
sister | -0.123 |
teacher | -0.122 |
problem | -0.120 |
exam | -0.120 |
Token | kitten |
Token ID | 42143 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
From | +0.057 |
Always | +0.048 |
Flo | +0.048 |
Since | +0.047 |
They | +0.047 |
Neg logit contributions | |
happened | -0.041 |
teacher | -0.040 |
exam | -0.039 |
sister | -0.038 |
problem | -0.038 |
Token | safe |
Token ID | 3338 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
sister | +0.005 |
happened | +0.005 |
exam | +0.005 |
teacher | +0.005 |
problem | +0.005 |
Neg logit contributions | |
From | -0.007 |
Since | -0.006 |
Even | -0.006 |
Always | -0.006 |
Flo | -0.006 |
Token | They |
Token ID | 1119 |
Feature activation | -8.81e+00 |
Pos logit contributions | |
happened | +0.072 |
teacher | +0.067 |
sister | +0.065 |
problem | +0.061 |
dad | +0.060 |
Neg logit contributions | |
From | -0.058 |
Always | -0.051 |
aling | -0.049 |
Ho | -0.047 |
oc | -0.047 |
Token | made |
Token ID | 925 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
From | +0.347 |
Flo | +0.304 |
Since | +0.300 |
Always | +0.300 |
inning | +0.293 |
Neg logit contributions | |
sister | -0.265 |
teacher | -0.263 |
happened | -0.261 |
exam | -0.232 |
dad | -0.227 |
Token | a |
Token ID | 257 |
Feature activation | +4.00e+00 |
Pos logit contributions | |
From | +0.008 |
Always | +0.007 |
Since | +0.007 |
Flo | +0.007 |
Even | +0.006 |
Neg logit contributions | |
happened | -0.005 |
teacher | -0.005 |
sister | -0.004 |
exam | -0.004 |
problem | -0.004 |
Token | face |
Token ID | 1986 |
Feature activation | -9.18e-01 |
Pos logit contributions | |
happened | +0.132 |
exam | +0.124 |
teacher | +0.121 |
sister | +0.118 |
problem | +0.115 |
Neg logit contributions | |
From | -0.148 |
Flo | -0.124 |
They | -0.123 |
Since | -0.120 |
Always | -0.119 |
Token | for |
Token ID | 329 |
Feature activation | -7.33e-01 |
Pos logit contributions | |
From | +0.033 |
Always | +0.027 |
Since | +0.027 |
Flo | +0.027 |
Bug | +0.026 |
Neg logit contributions | |
happened | -0.032 |
teacher | -0.031 |
sister | -0.030 |
exam | -0.030 |
problem | -0.029 |
Token | the |
Token ID | 262 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
From | +0.027 |
Always | +0.022 |
Flo | +0.022 |
Since | +0.022 |
They | +0.022 |
Neg logit contributions | |
happened | -0.024 |
teacher | -0.023 |
exam | -0.023 |
sister | -0.023 |
problem | -0.022 |
Token | snow |
Token ID | 6729 |
Feature activation | -3.32e+00 |
Pos logit contributions | |
From | +0.052 |
Always | +0.044 |
Flo | +0.044 |
They | +0.043 |
Since | +0.043 |
Neg logit contributions | |
happened | -0.044 |
exam | -0.041 |
teacher | -0.041 |
sister | -0.040 |
problem | -0.039 |
Token | man |
Token ID | 805 |
Feature activation | +5.69e+00 |
Pos logit contributions | |
From | +0.126 |
Flo | +0.111 |
They | +0.111 |
Always | +0.107 |
oc | +0.106 |
Neg logit contributions | |
happened | -0.097 |
exam | -0.095 |
teacher | -0.093 |
sister | -0.090 |
problem | -0.089 |
Token | . |
Token ID | 13 |
Feature activation | +1.80e+00 |
Pos logit contributions | |
happened | +0.159 |
exam | +0.155 |
teacher | +0.152 |
problem | +0.146 |
sister | +0.145 |
Neg logit contributions | |
From | -0.226 |
They | -0.197 |
Flo | -0.196 |
Always | -0.194 |
oc | -0.192 |
Token |
|
Token ID | 198 |
Feature activation | -2.10e+00 |
Pos logit contributions | |
happened | +0.064 |
teacher | +0.059 |
sister | +0.056 |
mom | +0.054 |
dad | +0.053 |
Neg logit contributions | |
From | -0.060 |
Always | -0.056 |
aling | -0.053 |
cking | -0.051 |
Ho | -0.051 |
Token |
|
Token ID | 198 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
From | +0.071 |
Always | +0.067 |
pill | +0.066 |
ksh | +0.064 |
Looks | +0.064 |
Neg logit contributions | |
happened | -0.067 |
teacher | -0.062 |
school | -0.054 |
sister | -0.053 |
problem | -0.053 |
Token | . |
Token ID | 13 |
Feature activation | +9.87e-01 |
Pos logit contributions | |
happened | +0.088 |
exam | +0.087 |
teacher | +0.084 |
problem | +0.082 |
classmate | +0.081 |
Neg logit contributions | |
From | -0.066 |
They | -0.058 |
Always | -0.057 |
He | -0.053 |
Flo | -0.053 |
Token | He |
Token ID | 679 |
Feature activation | -7.83e+00 |
Pos logit contributions | |
teacher | +0.043 |
happened | +0.041 |
sister | +0.041 |
classmate | +0.038 |
problem | +0.038 |
Neg logit contributions | |
From | -0.025 |
Always | -0.021 |
Flo | -0.020 |
rig | -0.020 |
Hop | -0.019 |
Token | said |
Token ID | 531 |
Feature activation | +7.19e+00 |
Pos logit contributions | |
From | +0.339 |
Flo | +0.294 |
Always | +0.291 |
Since | +0.291 |
They | +0.285 |
Neg logit contributions | |
happened | -0.210 |
teacher | -0.196 |
sister | -0.188 |
exam | -0.187 |
problem | -0.184 |
Token | he |
Token ID | 339 |
Feature activation | -7.35e-01 |
Pos logit contributions | |
exam | +0.273 |
happened | +0.268 |
teacher | +0.265 |
problem | +0.261 |
classmate | +0.260 |
Neg logit contributions | |
From | -0.213 |
They | -0.167 |
Flo | -0.166 |
Always | -0.163 |
Since | -0.162 |
Token | hated |
Token ID | 16563 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
From | +0.028 |
Always | +0.024 |
Flo | +0.023 |
They | +0.023 |
Since | +0.023 |
Neg logit contributions | |
happened | -0.022 |
exam | -0.022 |
teacher | -0.022 |
sister | -0.021 |
problem | -0.021 |
Token | the |
Token ID | 262 |
Feature activation | -1.93e+00 |
Pos logit contributions | |
From | +0.044 |
Always | +0.038 |
Flo | +0.038 |
Since | +0.037 |
They | +0.037 |
Neg logit contributions | |
happened | -0.034 |
teacher | -0.033 |
exam | -0.032 |
sister | -0.032 |
problem | -0.031 |
Token | blanket |
Token ID | 18447 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
From | +0.072 |
Always | +0.067 |
Flo | +0.062 |
Since | +0.061 |
They | +0.060 |
Neg logit contributions | |
happened | -0.050 |
classmate | -0.043 |
exam | -0.042 |
teacher | -0.042 |
sister | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | +8.92e-01 |
Pos logit contributions | |
From | +0.055 |
Always | +0.047 |
Flo | +0.047 |
They | +0.046 |
Since | +0.046 |
Neg logit contributions | |
happened | -0.038 |
teacher | -0.036 |
exam | -0.036 |
sister | -0.034 |
problem | -0.034 |
Token | He |
Token ID | 679 |
Feature activation | -7.85e+00 |
Pos logit contributions | |
teacher | +0.037 |
happened | +0.036 |
sister | +0.036 |
classmate | +0.033 |
problem | +0.033 |
Neg logit contributions | |
From | -0.023 |
Always | -0.020 |
rig | -0.020 |
Flo | -0.019 |
cking | -0.019 |
Token | said |
Token ID | 531 |
Feature activation | +7.13e+00 |
Pos logit contributions | |
From | +0.342 |
Flo | +0.298 |
Always | +0.295 |
Since | +0.295 |
They | +0.289 |
Neg logit contributions | |
happened | -0.207 |
teacher | -0.193 |
sister | -0.185 |
exam | -0.184 |
problem | -0.181 |
Token | he |
Token ID | 339 |
Feature activation | -7.72e-01 |
Pos logit contributions | |
exam | +0.273 |
happened | +0.269 |
teacher | +0.266 |
problem | +0.261 |
classmate | +0.258 |
Neg logit contributions | |
From | -0.213 |
Flo | -0.166 |
They | -0.164 |
Since | -0.163 |
Always | -0.162 |
Token | book |
Token ID | 1492 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
From | +0.077 |
Always | +0.067 |
Flo | +0.066 |
They | +0.066 |
Since | +0.065 |
Neg logit contributions | |
happened | -0.054 |
teacher | -0.049 |
exam | -0.048 |
sister | -0.048 |
classmate | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | +7.16e-01 |
Pos logit contributions | |
happened | +0.047 |
exam | +0.045 |
sister | +0.044 |
teacher | +0.044 |
problem | +0.043 |
Neg logit contributions | |
From | -0.074 |
They | -0.063 |
Always | -0.063 |
Flo | -0.063 |
Since | -0.062 |
Token | They |
Token ID | 1119 |
Feature activation | -9.31e+00 |
Pos logit contributions | |
happened | +0.032 |
teacher | +0.032 |
sister | +0.029 |
dad | +0.029 |
cousin | +0.028 |
Neg logit contributions | |
From | -0.017 |
cking | -0.014 |
Always | -0.014 |
ized | -0.014 |
oc | -0.014 |
Token | smiled |
Token ID | 13541 |
Feature activation | -4.90e-01 |
Pos logit contributions | |
From | +0.362 |
Always | +0.305 |
Flo | +0.304 |
Since | +0.302 |
They | +0.299 |
Neg logit contributions | |
happened | -0.284 |
teacher | -0.271 |
exam | -0.270 |
sister | -0.263 |
problem | -0.260 |
Token | and |
Token ID | 290 |
Feature activation | -8.54e+00 |
Pos logit contributions | |
From | +0.013 |
They | +0.010 |
Always | +0.010 |
Since | +0.010 |
Flo | +0.010 |
Neg logit contributions | |
happened | -0.021 |
exam | -0.021 |
teacher | -0.020 |
problem | -0.020 |
sister | -0.020 |
Token | gigg |
Token ID | 30442 |
Feature activation | -5.80e+00 |
Pos logit contributions | |
From | +0.278 |
They | +0.228 |
Always | +0.220 |
Flo | +0.218 |
Since | +0.216 |
Neg logit contributions | |
exam | -0.316 |
happened | -0.303 |
problem | -0.302 |
teacher | -0.295 |
sister | -0.286 |
Token | led |
Token ID | 992 |
Feature activation | -1.93e+00 |
Pos logit contributions | |
From | +0.162 |
Always | +0.127 |
Flo | +0.125 |
They | +0.125 |
Since | +0.124 |
Neg logit contributions | |
happened | -0.241 |
teacher | -0.236 |
exam | -0.234 |
sister | -0.229 |
problem | -0.225 |
Token | . |
Token ID | 13 |
Feature activation | +8.35e-01 |
Pos logit contributions | |
From | +0.085 |
Always | +0.073 |
Flo | +0.073 |
Since | +0.072 |
They | +0.071 |
Neg logit contributions | |
happened | -0.049 |
teacher | -0.047 |
exam | -0.046 |
sister | -0.045 |
problem | -0.044 |
Token |
|
Token ID | 198 |
Feature activation | -2.93e+00 |
Pos logit contributions | |
happened | +0.035 |
teacher | +0.035 |
sister | +0.033 |
classmate | +0.031 |
problem | +0.031 |
Neg logit contributions | |
From | -0.022 |
Always | -0.019 |
Flo | -0.018 |
oc | -0.018 |
ized | -0.017 |
Token |
|
Token ID | 198 |
Feature activation | -2.97e+00 |
Pos logit contributions | |
From | +0.088 |
Always | +0.081 |
Flo | +0.077 |
Looks | +0.076 |
pill | +0.073 |
Neg logit contributions | |
happened | -0.106 |
teacher | -0.104 |
sister | -0.097 |
problem | -0.094 |
exam | -0.093 |
Token | But |
Token ID | 1537 |
Feature activation | -3.49e+00 |
Pos logit contributions | |
ksh | +0.087 |
pill | +0.086 |
Always | +0.084 |
Eagles | +0.084 |
rig | +0.082 |
Neg logit contributions | |
teacher | -0.099 |
happened | -0.091 |
school | -0.087 |
ha | -0.086 |
dad | -0.085 |
Token | and |
Token ID | 290 |
Feature activation | -7.87e+00 |
Pos logit contributions | |
happened | +0.064 |
teacher | +0.063 |
exam | +0.060 |
sister | +0.060 |
problem | +0.059 |
Neg logit contributions | |
From | -0.071 |
Flo | -0.061 |
Always | -0.060 |
Since | -0.059 |
Even | -0.058 |
Token | she |
Token ID | 673 |
Feature activation | +6.04e-02 |
Pos logit contributions | |
From | +0.328 |
Flo | +0.283 |
Always | +0.282 |
Since | +0.277 |
They | +0.274 |
Neg logit contributions | |
happened | -0.224 |
teacher | -0.210 |
exam | -0.203 |
sister | -0.203 |
problem | -0.197 |
Token | quickly |
Token ID | 2952 |
Feature activation | +7.62e-01 |
Pos logit contributions | |
happened | +0.002 |
teacher | +0.002 |
sister | +0.001 |
problem | +0.001 |
classmate | +0.001 |
Neg logit contributions | |
From | -0.003 |
Flo | -0.002 |
Always | -0.002 |
Since | -0.002 |
Bug | -0.002 |
Token | thanked |
Token ID | 26280 |
Feature activation | +4.00e-02 |
Pos logit contributions | |
happened | +0.024 |
teacher | +0.022 |
sister | +0.021 |
problem | +0.021 |
exam | +0.021 |
Neg logit contributions | |
From | -0.030 |
Always | -0.026 |
Flo | -0.025 |
Since | -0.025 |
Bug | -0.024 |
Token | the |
Token ID | 262 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
teacher | +0.001 |
happened | +0.001 |
exam | +0.001 |
sister | +0.001 |
school | +0.001 |
Neg logit contributions | |
From | -0.002 |
Flo | -0.002 |
Always | -0.002 |
Even | -0.002 |
oc | -0.002 |
Token | old |
Token ID | 1468 |
Feature activation | -4.04e+00 |
Pos logit contributions | |
From | +0.057 |
Flo | +0.048 |
Always | +0.048 |
Since | +0.047 |
They | +0.047 |
Neg logit contributions | |
happened | -0.045 |
exam | -0.042 |
teacher | -0.041 |
sister | -0.041 |
problem | -0.040 |
Token | lady |
Token ID | 10846 |
Feature activation | +3.91e+00 |
Pos logit contributions | |
From | +0.161 |
Since | +0.135 |
Always | +0.135 |
The | +0.134 |
Flo | +0.134 |
Neg logit contributions | |
happened | -0.108 |
exam | -0.101 |
sister | -0.098 |
problem | -0.093 |
classmate | -0.090 |
Token | . |
Token ID | 13 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
happened | +0.094 |
teacher | +0.094 |
sister | +0.092 |
exam | +0.091 |
problem | +0.087 |
Neg logit contributions | |
From | -0.175 |
Always | -0.151 |
Flo | -0.150 |
Since | -0.149 |
Even | -0.149 |
Token | But |
Token ID | 887 |
Feature activation | -3.85e+00 |
Pos logit contributions | |
teacher | +0.053 |
happened | +0.051 |
sister | +0.047 |
dad | +0.047 |
classmate | +0.047 |
Neg logit contributions | |
From | -0.047 |
cking | -0.047 |
Flo | -0.044 |
inning | -0.043 |
Always | -0.043 |
Token | just |
Token ID | 655 |
Feature activation | -2.53e+00 |
Pos logit contributions | |
From | +0.151 |
Flo | +0.132 |
Always | +0.130 |
They | +0.129 |
Since | +0.127 |
Neg logit contributions | |
happened | -0.116 |
teacher | -0.111 |
sister | -0.107 |
exam | -0.107 |
problem | -0.105 |
Token | as |
Token ID | 355 |
Feature activation | -6.13e-01 |
Pos logit contributions | |
From | +0.086 |
Flo | +0.072 |
Always | +0.071 |
Since | +0.071 |
They | +0.070 |
Neg logit contributions | |
happened | -0.092 |
teacher | -0.086 |
exam | -0.086 |
sister | -0.086 |
problem | -0.084 |
Token | would |
Token ID | 561 |
Feature activation | +4.35e+00 |
Pos logit contributions | |
From | +0.001 |
Flo | +0.001 |
Always | +0.001 |
Since | +0.001 |
They | +0.001 |
Neg logit contributions | |
happened | -0.001 |
teacher | -0.001 |
exam | -0.001 |
sister | -0.001 |
problem | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | -7.81e+00 |
Pos logit contributions | |
teacher | +0.116 |
exam | +0.108 |
sister | +0.102 |
happened | +0.097 |
haircut | +0.094 |
Neg logit contributions | |
From | -0.187 |
pill | -0.178 |
Looks | -0.172 |
valleys | -0.167 |
Flo | -0.167 |
Token | they |
Token ID | 484 |
Feature activation | -7.40e-01 |
Pos logit contributions | |
From | +0.304 |
Always | +0.257 |
Flo | +0.257 |
Since | +0.256 |
They | +0.251 |
Neg logit contributions | |
happened | -0.240 |
teacher | -0.237 |
sister | -0.230 |
exam | -0.228 |
problem | -0.213 |
Token | went |
Token ID | 1816 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
From | +0.029 |
Since | +0.024 |
Flo | +0.024 |
Always | +0.024 |
Even | +0.023 |
Neg logit contributions | |
happened | -0.023 |
teacher | -0.023 |
exam | -0.022 |
sister | -0.022 |
problem | -0.020 |
Token | home |
Token ID | 1363 |
Feature activation | +8.42e+00 |
Pos logit contributions | |
happened | +0.035 |
teacher | +0.034 |
exam | +0.033 |
sister | +0.031 |
problem | +0.031 |
Neg logit contributions | |
From | -0.115 |
Always | -0.101 |
Since | -0.101 |
Flo | -0.101 |
Even | -0.098 |
Token | with |
Token ID | 351 |
Feature activation | -5.63e+00 |
Pos logit contributions | |
happened | +0.253 |
teacher | +0.240 |
exam | +0.238 |
sister | +0.232 |
problem | +0.229 |
Neg logit contributions | |
From | -0.333 |
Always | -0.282 |
Flo | -0.281 |
Since | -0.279 |
They | -0.276 |
Token | the |
Token ID | 262 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
From | +0.219 |
Always | +0.186 |
Flo | +0.185 |
Since | +0.183 |
They | +0.182 |
Neg logit contributions | |
happened | -0.169 |
exam | -0.162 |
teacher | -0.161 |
sister | -0.155 |
problem | -0.154 |
Token | new |
Token ID | 649 |
Feature activation | -1.73e+00 |
Pos logit contributions | |
From | +0.059 |
Always | +0.051 |
Flo | +0.050 |
Since | +0.049 |
They | +0.048 |
Neg logit contributions | |
happened | -0.047 |
teacher | -0.043 |
sister | -0.043 |
exam | -0.042 |
classmate | -0.041 |
Token | pony |
Token ID | 26902 |
Feature activation | -5.37e-01 |
Pos logit contributions | |
From | +0.066 |
Always | +0.058 |
Flo | +0.057 |
Since | +0.055 |
They | +0.055 |
Neg logit contributions | |
happened | -0.053 |
teacher | -0.050 |
exam | -0.049 |
sister | -0.049 |
principal | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
From | +0.023 |
Flo | +0.020 |
Always | +0.020 |
Since | +0.020 |
They | +0.019 |
Neg logit contributions | |
happened | -0.014 |
teacher | -0.014 |
exam | -0.014 |
sister | -0.013 |
problem | -0.013 |
Token | |
Token ID | 220 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
teacher | +0.047 |
sister | +0.045 |
happened | +0.043 |
dad | +0.043 |
home | +0.041 |
Neg logit contributions | |
From | -0.050 |
cking | -0.048 |
inning | -0.047 |
Always | -0.046 |
izing | -0.046 |
Token | many |
Token ID | 867 |
Feature activation | -4.52e+00 |
Pos logit contributions | |
From | +0.113 |
Flo | +0.093 |
They | +0.093 |
Always | +0.092 |
Since | +0.091 |
Neg logit contributions | |
happened | -0.114 |
exam | -0.109 |
teacher | -0.109 |
sister | -0.106 |
problem | -0.104 |
Token | things |
Token ID | 1243 |
Feature activation | +3.68e-01 |
Pos logit contributions | |
From | +0.174 |
Always | +0.146 |
Flo | +0.146 |
Since | +0.145 |
They | +0.144 |
Neg logit contributions | |
happened | -0.140 |
teacher | -0.134 |
exam | -0.133 |
sister | -0.130 |
problem | -0.127 |
Token | with |
Token ID | 351 |
Feature activation | -5.65e+00 |
Pos logit contributions | |
happened | +0.011 |
teacher | +0.011 |
exam | +0.010 |
problem | +0.010 |
sister | +0.010 |
Neg logit contributions | |
From | -0.015 |
Since | -0.013 |
Flo | -0.012 |
Always | -0.012 |
Even | -0.012 |
Token | blocks |
Token ID | 7021 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
From | +0.215 |
Flo | +0.180 |
Always | +0.180 |
Since | +0.178 |
They | +0.177 |
Neg logit contributions | |
happened | -0.178 |
teacher | -0.170 |
exam | -0.169 |
sister | -0.165 |
problem | -0.163 |
Token | . |
Token ID | 13 |
Feature activation | +2.11e+00 |
Pos logit contributions | |
From | +0.077 |
Always | +0.066 |
Flo | +0.066 |
Since | +0.065 |
They | +0.064 |
Neg logit contributions | |
happened | -0.049 |
teacher | -0.047 |
exam | -0.046 |
sister | -0.045 |
problem | -0.044 |
Token | They |
Token ID | 1119 |
Feature activation | -8.73e+00 |
Pos logit contributions | |
sister | +0.078 |
happened | +0.078 |
teacher | +0.078 |
problem | +0.073 |
exam | +0.073 |
Neg logit contributions | |
From | -0.068 |
Always | -0.055 |
Flo | -0.054 |
oc | -0.053 |
Since | -0.052 |
Token | could |
Token ID | 714 |
Feature activation | +2.93e+00 |
Pos logit contributions | |
From | +0.313 |
Flo | +0.260 |
Always | +0.260 |
Since | +0.257 |
They | +0.253 |
Neg logit contributions | |
happened | -0.295 |
teacher | -0.283 |
exam | -0.280 |
sister | -0.275 |
problem | -0.271 |
Token | build |
Token ID | 1382 |
Feature activation | -3.22e+00 |
Pos logit contributions | |
teacher | +0.088 |
sister | +0.083 |
exam | +0.082 |
happened | +0.078 |
problem | +0.077 |
Neg logit contributions | |
From | -0.123 |
Always | -0.106 |
Since | -0.104 |
pill | -0.103 |
Flo | -0.103 |
Token | houses |
Token ID | 7777 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
From | +0.116 |
Flo | +0.097 |
Always | +0.096 |
Since | +0.095 |
They | +0.095 |
Neg logit contributions | |
happened | -0.107 |
teacher | -0.103 |
exam | -0.102 |
sister | -0.100 |
problem | -0.098 |
Token | , |
Token ID | 11 |
Feature activation | +4.76e+00 |
Pos logit contributions | |
From | +0.098 |
Always | +0.086 |
Since | +0.085 |
Flo | +0.084 |
Bug | +0.082 |
Neg logit contributions | |
teacher | -0.050 |
happened | -0.050 |
sister | -0.047 |
problem | -0.047 |
exam | -0.046 |
Token | towers |
Token ID | 18028 |
Feature activation | -4.68e-02 |
Pos logit contributions | |
teacher | +0.121 |
exam | +0.116 |
happened | +0.115 |
sister | +0.114 |
problem | +0.108 |
Neg logit contributions | |
From | -0.213 |
Since | -0.189 |
Flo | -0.183 |
Always | -0.183 |
rig | -0.180 |
Token | sitting |
Token ID | 5586 |
Feature activation | -5.39e+00 |
Pos logit contributions | |
happened | +0.111 |
teacher | +0.106 |
exam | +0.106 |
sister | +0.102 |
problem | +0.100 |
Neg logit contributions | |
From | -0.147 |
Always | -0.126 |
Flo | -0.126 |
They | -0.125 |
Since | -0.123 |
Token | on |
Token ID | 319 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
From | +0.226 |
Flo | +0.194 |
Since | +0.191 |
Always | +0.190 |
Even | +0.188 |
Neg logit contributions | |
happened | -0.148 |
teacher | -0.145 |
sister | -0.141 |
exam | -0.140 |
problem | -0.138 |
Token | a |
Token ID | 257 |
Feature activation | +3.79e+00 |
Pos logit contributions | |
From | +0.070 |
Always | +0.061 |
Since | +0.060 |
Flo | +0.060 |
Even | +0.059 |
Neg logit contributions | |
happened | -0.042 |
teacher | -0.041 |
exam | -0.039 |
sister | -0.038 |
problem | -0.036 |
Token | bench |
Token ID | 7624 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
happened | +0.129 |
exam | +0.119 |
teacher | +0.118 |
sister | +0.118 |
problem | +0.116 |
Neg logit contributions | |
From | -0.136 |
Flo | -0.115 |
They | -0.112 |
Since | -0.110 |
Always | -0.109 |
Token | . |
Token ID | 13 |
Feature activation | +8.44e-01 |
Pos logit contributions | |
From | +0.069 |
Flo | +0.060 |
Always | +0.060 |
Since | +0.059 |
Even | +0.058 |
Neg logit contributions | |
happened | -0.032 |
teacher | -0.032 |
exam | -0.029 |
problem | -0.029 |
sister | -0.028 |
Token | They |
Token ID | 1119 |
Feature activation | -9.02e+00 |
Pos logit contributions | |
happened | +0.037 |
teacher | +0.036 |
sister | +0.035 |
exam | +0.035 |
problem | +0.034 |
Neg logit contributions | |
From | -0.022 |
Always | -0.017 |
Flo | -0.017 |
Since | -0.017 |
They | -0.016 |
Token | hugged |
Token ID | 41130 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
From | +0.420 |
Flo | +0.375 |
Always | +0.369 |
Since | +0.366 |
inning | +0.366 |
Neg logit contributions | |
happened | -0.212 |
teacher | -0.206 |
sister | -0.193 |
dad | -0.174 |
problem | -0.170 |
Token | her |
Token ID | 607 |
Feature activation | +5.95e+00 |
Pos logit contributions | |
From | +0.072 |
Always | +0.063 |
Flo | +0.062 |
Since | +0.062 |
Even | +0.060 |
Neg logit contributions | |
happened | -0.037 |
teacher | -0.036 |
exam | -0.035 |
sister | -0.035 |
problem | -0.033 |
Token | and |
Token ID | 290 |
Feature activation | -8.44e+00 |
Pos logit contributions | |
happened | +0.244 |
teacher | +0.237 |
exam | +0.234 |
sister | +0.232 |
problem | +0.228 |
Neg logit contributions | |
From | -0.172 |
Always | -0.136 |
Flo | -0.135 |
Since | -0.133 |
They | -0.130 |
Token | told |
Token ID | 1297 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
From | +0.335 |
Flo | +0.286 |
Always | +0.284 |
Since | +0.281 |
They | +0.275 |
Neg logit contributions | |
happened | -0.257 |
teacher | -0.243 |
exam | -0.237 |
sister | -0.237 |
problem | -0.230 |
Token | her |
Token ID | 607 |
Feature activation | +5.97e+00 |
Pos logit contributions | |
happened | +0.054 |
classmate | +0.052 |
problem | +0.052 |
exam | +0.052 |
teacher | +0.049 |
Neg logit contributions | |
From | -0.059 |
They | -0.052 |
Flo | -0.048 |
The | -0.046 |
Since | -0.046 |
Token | . |
Token ID | 13 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
happened | +0.031 |
teacher | +0.029 |
exam | +0.027 |
sister | +0.027 |
problem | +0.027 |
Neg logit contributions | |
From | -0.052 |
Always | -0.046 |
Since | -0.045 |
Flo | -0.045 |
Even | -0.044 |
Token | They |
Token ID | 1119 |
Feature activation | -9.25e+00 |
Pos logit contributions | |
teacher | +0.040 |
happened | +0.038 |
dad | +0.038 |
sister | +0.037 |
mom | +0.036 |
Neg logit contributions | |
cking | -0.033 |
From | -0.032 |
Always | -0.032 |
inning | -0.031 |
ized | -0.030 |
Token | made |
Token ID | 925 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
From | +0.388 |
Always | +0.335 |
Flo | +0.332 |
Since | +0.330 |
Even | +0.316 |
Neg logit contributions | |
happened | -0.269 |
teacher | -0.257 |
sister | -0.246 |
exam | -0.231 |
problem | -0.224 |
Token | crane |
Token ID | 41175 |
Feature activation | -5.66e+00 |
Pos logit contributions | |
From | +0.006 |
Flo | +0.005 |
Always | +0.005 |
Since | +0.005 |
They | +0.005 |
Neg logit contributions | |
happened | -0.005 |
teacher | -0.004 |
exam | -0.004 |
sister | -0.004 |
problem | -0.004 |
Token | noises |
Token ID | 26782 |
Feature activation | -2.29e+00 |
Pos logit contributions | |
From | +0.223 |
Always | +0.191 |
Since | +0.188 |
Flo | +0.185 |
Even | +0.178 |
Neg logit contributions | |
happened | -0.174 |
teacher | -0.165 |
sister | -0.164 |
problem | -0.163 |
exam | -0.159 |
Token | and |
Token ID | 290 |
Feature activation | -8.42e+00 |
Pos logit contributions | |
From | +0.078 |
Always | +0.066 |
Flo | +0.065 |
Since | +0.064 |
Bug | +0.063 |
Neg logit contributions | |
happened | -0.083 |
teacher | -0.081 |
exam | -0.077 |
sister | -0.076 |
problem | -0.076 |
Token | laughed |
Token ID | 13818 |
Feature activation | -9.38e-01 |
Pos logit contributions | |
From | +0.285 |
Flo | +0.233 |
Always | +0.232 |
Since | +0.230 |
They | +0.230 |
Neg logit contributions | |
happened | -0.295 |
exam | -0.285 |
teacher | -0.285 |
sister | -0.276 |
problem | -0.275 |
Token | . |
Token ID | 13 |
Feature activation | +1.00e+00 |
Pos logit contributions | |
From | +0.040 |
Always | +0.035 |
Flo | +0.034 |
Since | +0.034 |
Even | +0.033 |
Neg logit contributions | |
happened | -0.025 |
teacher | -0.024 |
exam | -0.023 |
sister | -0.023 |
problem | -0.022 |
Token | They |
Token ID | 1119 |
Feature activation | -9.21e+00 |
Pos logit contributions | |
teacher | +0.040 |
happened | +0.039 |
sister | +0.038 |
dad | +0.036 |
school | +0.035 |
Neg logit contributions | |
From | -0.027 |
Always | -0.025 |
oc | -0.024 |
cking | -0.023 |
rig | -0.023 |
Token | were |
Token ID | 547 |
Feature activation | +2.07e+00 |
Pos logit contributions | |
From | +0.358 |
Flo | +0.304 |
Always | +0.304 |
Since | +0.300 |
They | +0.292 |
Neg logit contributions | |
happened | -0.286 |
teacher | -0.273 |
sister | -0.263 |
exam | -0.262 |
problem | -0.255 |
Token | happy |
Token ID | 3772 |
Feature activation | +2.43e-02 |
Pos logit contributions | |
happened | +0.070 |
exam | +0.068 |
teacher | +0.066 |
problem | +0.064 |
sister | +0.063 |
Neg logit contributions | |
From | -0.074 |
They | -0.062 |
Flo | -0.061 |
Always | -0.060 |
Since | -0.060 |
Token | wouldn |
Token ID | 3636 |
Feature activation | +9.34e-01 |
Pos logit contributions | |
happened | +0.045 |
teacher | +0.041 |
exam | +0.039 |
sister | +0.039 |
problem | +0.039 |
Neg logit contributions | |
From | -0.068 |
Flo | -0.058 |
Since | -0.058 |
Always | -0.057 |
They | -0.056 |
Token | 't |
Token ID | 470 |
Feature activation | -2.30e-01 |
Pos logit contributions | |
happened | +0.029 |
exam | +0.027 |
teacher | +0.027 |
sister | +0.026 |
problem | +0.025 |
Neg logit contributions | |
From | -0.036 |
Always | -0.031 |
Flo | -0.030 |
Since | -0.030 |
Looks | -0.030 |
Token | be |
Token ID | 307 |
Feature activation | +7.21e-01 |
Pos logit contributions | |
From | +0.009 |
Always | +0.008 |
Flo | +0.008 |
They | +0.008 |
Since | +0.008 |
Neg logit contributions | |
happened | -0.007 |
teacher | -0.006 |
exam | -0.006 |
sister | -0.006 |
problem | -0.006 |
Token | raining |
Token ID | 43079 |
Feature activation | -5.70e+00 |
Pos logit contributions | |
teacher | +0.024 |
exam | +0.024 |
happened | +0.022 |
classmate | +0.022 |
principal | +0.022 |
Neg logit contributions | |
From | -0.024 |
Flo | -0.021 |
Always | -0.021 |
They | -0.020 |
Bug | -0.019 |
Token | anymore |
Token ID | 7471 |
Feature activation | +3.47e+00 |
Pos logit contributions | |
From | +0.221 |
They | +0.194 |
Always | +0.191 |
Flo | +0.190 |
Since | +0.184 |
Neg logit contributions | |
teacher | -0.162 |
happened | -0.162 |
sister | -0.159 |
exam | -0.154 |
classmate | -0.153 |
Token | and |
Token ID | 290 |
Feature activation | -8.04e+00 |
Pos logit contributions | |
happened | +0.095 |
exam | +0.095 |
teacher | +0.095 |
sister | +0.091 |
problem | +0.090 |
Neg logit contributions | |
From | -0.137 |
Always | -0.119 |
They | -0.119 |
Flo | -0.116 |
Since | -0.115 |
Token | he |
Token ID | 339 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
From | +0.291 |
Flo | +0.245 |
Always | +0.241 |
Since | +0.240 |
They | +0.234 |
Neg logit contributions | |
happened | -0.273 |
teacher | -0.259 |
exam | -0.257 |
sister | -0.253 |
problem | -0.249 |
Token | could |
Token ID | 714 |
Feature activation | +2.94e+00 |
Pos logit contributions | |
From | +0.016 |
Flo | +0.013 |
Since | +0.013 |
Always | +0.013 |
They | +0.013 |
Neg logit contributions | |
happened | -0.012 |
teacher | -0.011 |
sister | -0.011 |
exam | -0.010 |
problem | -0.010 |
Token | go |
Token ID | 467 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
teacher | +0.083 |
exam | +0.081 |
sister | +0.076 |
happened | +0.074 |
problem | +0.070 |
Neg logit contributions | |
From | -0.126 |
pill | -0.111 |
Ho | -0.109 |
Looks | -0.109 |
oc | -0.108 |
Token | to |
Token ID | 284 |
Feature activation | +4.47e+00 |
Pos logit contributions | |
happened | +0.025 |
exam | +0.024 |
teacher | +0.024 |
sister | +0.023 |
problem | +0.023 |
Neg logit contributions | |
From | -0.049 |
Since | -0.042 |
Flo | -0.042 |
Always | -0.041 |
Even | -0.040 |
Token | the |
Token ID | 262 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
happened | +0.142 |
teacher | +0.135 |
exam | +0.133 |
sister | +0.131 |
problem | +0.129 |
Neg logit contributions | |
From | -0.169 |
Always | -0.142 |
Flo | -0.142 |
Since | -0.140 |
They | -0.139 |
Token | said |
Token ID | 531 |
Feature activation | +7.90e+00 |
Pos logit contributions | |
happened | +0.006 |
exam | +0.006 |
teacher | +0.006 |
problem | +0.006 |
sister | +0.005 |
Neg logit contributions | |
From | -0.007 |
Flo | -0.006 |
They | -0.006 |
Always | -0.006 |
Since | -0.006 |
Token | as |
Token ID | 355 |
Feature activation | -3.63e-01 |
Pos logit contributions | |
happened | +0.230 |
exam | +0.218 |
teacher | +0.216 |
problem | +0.212 |
sister | +0.208 |
Neg logit contributions | |
From | -0.318 |
Flo | -0.268 |
They | -0.268 |
Always | -0.267 |
Since | -0.266 |
Token | she |
Token ID | 673 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
From | +0.015 |
Always | +0.014 |
Flo | +0.013 |
Since | +0.013 |
Even | +0.013 |
Neg logit contributions | |
happened | -0.010 |
teacher | -0.010 |
sister | -0.009 |
exam | -0.009 |
problem | -0.009 |
Token | jumped |
Token ID | 11687 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
happened | +0.004 |
teacher | +0.004 |
sister | +0.004 |
exam | +0.004 |
problem | +0.004 |
Neg logit contributions | |
From | -0.006 |
Always | -0.005 |
Since | -0.005 |
Flo | -0.005 |
They | -0.005 |
Token | up |
Token ID | 510 |
Feature activation | +2.17e+00 |
Pos logit contributions | |
From | +0.060 |
Always | +0.052 |
Since | +0.051 |
Even | +0.051 |
Flo | +0.051 |
Neg logit contributions | |
happened | -0.039 |
teacher | -0.038 |
exam | -0.037 |
sister | -0.036 |
problem | -0.036 |
Token | and |
Token ID | 290 |
Feature activation | -7.94e+00 |
Pos logit contributions | |
happened | +0.070 |
teacher | +0.068 |
exam | +0.066 |
sister | +0.065 |
problem | +0.064 |
Neg logit contributions | |
From | -0.082 |
Always | -0.069 |
Flo | -0.069 |
Since | -0.068 |
Even | -0.068 |
Token | down |
Token ID | 866 |
Feature activation | +2.58e+00 |
Pos logit contributions | |
From | +0.323 |
Always | +0.276 |
Flo | +0.274 |
Since | +0.273 |
They | +0.267 |
Neg logit contributions | |
happened | -0.233 |
teacher | -0.222 |
sister | -0.217 |
exam | -0.216 |
problem | -0.208 |
Token | . |
Token ID | 13 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
happened | +0.070 |
teacher | +0.067 |
exam | +0.065 |
sister | +0.064 |
problem | +0.063 |
Neg logit contributions | |
From | -0.110 |
Always | -0.095 |
Flo | -0.095 |
Since | -0.094 |
Even | -0.092 |
Token |
|
Token ID | 198 |
Feature activation | -2.07e+00 |
Pos logit contributions | |
teacher | +0.065 |
happened | +0.064 |
sister | +0.060 |
exam | +0.060 |
dad | +0.057 |
Neg logit contributions | |
From | -0.066 |
Flo | -0.058 |
Always | -0.058 |
Tweet | -0.055 |
Hop | -0.054 |
Token |
|
Token ID | 198 |
Feature activation | -2.11e+00 |
Pos logit contributions | |
From | +0.063 |
Always | +0.057 |
Looks | +0.056 |
refreshing | +0.055 |
Flo | +0.054 |
Neg logit contributions | |
teacher | -0.072 |
happened | -0.071 |
exam | -0.064 |
sister | -0.063 |
dad | -0.062 |
Token | Her |
Token ID | 9360 |
Feature activation | -3.75e-01 |
Pos logit contributions | |
From | +0.050 |
Flo | +0.047 |
Always | +0.045 |
pill | +0.045 |
refreshing | +0.045 |
Neg logit contributions | |
teacher | -0.082 |
happened | -0.079 |
exam | -0.076 |
classmate | -0.073 |
problem | -0.072 |