Token | She |
Token ID | 1375 |
Feature activation | +8.16e-02 |
Pos logit contributions | |
unlike | +0.054 |
matter | +0.050 |
superpower | +0.049 |
perspective | +0.048 |
eway | +0.048 |
Neg logit contributions | |
Ben | -0.041 |
Ben | -0.037 |
bat | -0.036 |
says | -0.035 |
next | -0.035 |
Token | cried |
Token ID | 16896 |
Feature activation | -3.64e-02 |
Pos logit contributions | |
unlike | +0.157 |
matter | +0.149 |
superpower | +0.142 |
blonde | +0.139 |
tactic | +0.134 |
Neg logit contributions | |
ila | -0.099 |
line | -0.095 |
Ben | -0.093 |
Ben | -0.092 |
bat | -0.091 |
Token | and |
Token ID | 290 |
Feature activation | -7.73e-02 |
Pos logit contributions | |
Ben | +0.013 |
bat | +0.012 |
Ben | +0.009 |
se | +0.009 |
next | +0.008 |
Neg logit contributions | |
unlike | -0.108 |
superpower | -0.101 |
matter | -0.101 |
perspective | -0.100 |
eway | -0.099 |
Token | threw |
Token ID | 9617 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
ila | +0.062 |
Ben | +0.062 |
Ben | +0.061 |
bat | +0.061 |
se | +0.060 |
Neg logit contributions | |
unlike | -0.187 |
matter | -0.172 |
superpower | -0.171 |
blonde | -0.168 |
perspective | -0.166 |
Token | a |
Token ID | 257 |
Feature activation | +1.99e-03 |
Pos logit contributions | |
Ben | +0.130 |
bat | +0.114 |
each | +0.114 |
put | +0.107 |
se | +0.105 |
Neg logit contributions | |
unlike | -0.346 |
superpower | -0.318 |
perspective | -0.316 |
eway | -0.314 |
former | -0.309 |
Token | tant |
Token ID | 24246 |
Feature activation | +3.10e-01 |
Pos logit contributions | |
unlike | +0.003 |
former | +0.003 |
eway | +0.003 |
matter | +0.003 |
superpower | +0.003 |
Neg logit contributions | |
bat | -0.004 |
se | -0.004 |
Ben | -0.003 |
slide | -0.003 |
Ben | -0.003 |
Token | rum |
Token ID | 6582 |
Feature activation | +2.09e-01 |
Pos logit contributions | |
unlike | +0.277 |
former | +0.236 |
blonde | +0.226 |
matter | +0.224 |
superpower | +0.223 |
Neg logit contributions | |
ila | -0.731 |
se | -0.701 |
Ben | -0.689 |
Ben | -0.689 |
bat | -0.689 |
Token | . |
Token ID | 13 |
Feature activation | +2.87e-02 |
Pos logit contributions | |
unlike | +0.565 |
matter | +0.546 |
superpower | +0.534 |
tactic | +0.532 |
perspective | +0.519 |
Neg logit contributions | |
Ben | -0.128 |
next | -0.112 |
each | -0.107 |
bat | -0.105 |
ila | -0.101 |
Token | Her |
Token ID | 2332 |
Feature activation | +1.39e-01 |
Pos logit contributions | |
unlike | +0.053 |
matter | +0.052 |
perspective | +0.049 |
eway | +0.048 |
superpower | +0.048 |
Neg logit contributions | |
Ben | -0.046 |
Ben | -0.040 |
bat | -0.039 |
next | -0.039 |
says | -0.038 |
Token | mom |
Token ID | 1995 |
Feature activation | -3.33e-02 |
Pos logit contributions | |
unlike | +0.385 |
eway | +0.356 |
perspective | +0.352 |
tactic | +0.351 |
matter | +0.350 |
Neg logit contributions | |
bat | -0.088 |
Ben | -0.086 |
se | -0.083 |
next | -0.077 |
slide | -0.071 |
Token | tried |
Token ID | 3088 |
Feature activation | +2.49e-02 |
Pos logit contributions | |
ila | +0.040 |
Ben | +0.039 |
Ben | +0.039 |
bat | +0.038 |
se | +0.036 |
Neg logit contributions | |
unlike | -0.066 |
matter | -0.061 |
superpower | -0.058 |
blonde | -0.057 |
tactic | -0.057 |
Token | listen |
Token ID | 6004 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
ila | +0.040 |
Ben | +0.040 |
se | +0.040 |
Ben | +0.039 |
bat | +0.038 |
Neg logit contributions | |
matter | -0.075 |
unlike | -0.075 |
superpower | -0.069 |
perspective | -0.068 |
tactic | -0.067 |
Token | . |
Token ID | 13 |
Feature activation | +1.75e-02 |
Pos logit contributions | |
Ben | +0.108 |
bat | +0.104 |
Ben | +0.103 |
ila | +0.103 |
se | +0.101 |
Neg logit contributions | |
unlike | -0.281 |
superpower | -0.257 |
matter | -0.257 |
perspective | -0.252 |
blonde | -0.251 |
Token | She |
Token ID | 1375 |
Feature activation | +7.20e-02 |
Pos logit contributions | |
unlike | +0.031 |
matter | +0.028 |
superpower | +0.028 |
perspective | +0.027 |
tactic | +0.027 |
Neg logit contributions | |
Ben | -0.026 |
Ben | -0.025 |
bat | -0.025 |
ila | -0.025 |
se | -0.024 |
Token | threw |
Token ID | 9617 |
Feature activation | -1.43e-01 |
Pos logit contributions | |
unlike | +0.136 |
matter | +0.127 |
superpower | +0.122 |
blonde | +0.121 |
former | +0.116 |
Neg logit contributions | |
ila | -0.089 |
Ben | -0.087 |
line | -0.086 |
bat | -0.085 |
Ben | -0.085 |
Token | a |
Token ID | 257 |
Feature activation | -4.38e-03 |
Pos logit contributions | |
Ben | +0.110 |
bat | +0.104 |
Ben | +0.101 |
ila | +0.101 |
se | +0.100 |
Neg logit contributions | |
unlike | -0.359 |
superpower | -0.330 |
matter | -0.327 |
perspective | -0.323 |
blonde | -0.321 |
Token | tant |
Token ID | 24246 |
Feature activation | +3.10e-01 |
Pos logit contributions | |
bat | +0.009 |
se | +0.009 |
Ben | +0.009 |
Ben | +0.009 |
ila | +0.008 |
Neg logit contributions | |
unlike | -0.006 |
matter | -0.005 |
superpower | -0.005 |
blonde | -0.005 |
perspective | -0.005 |
Token | rum |
Token ID | 6582 |
Feature activation | +2.08e-01 |
Pos logit contributions | |
unlike | +0.282 |
former | +0.236 |
blonde | +0.228 |
matter | +0.227 |
superpower | +0.222 |
Neg logit contributions | |
ila | -0.738 |
se | -0.700 |
Ben | -0.696 |
Ben | -0.691 |
bat | -0.687 |
Token | in |
Token ID | 287 |
Feature activation | +3.15e-02 |
Pos logit contributions | |
unlike | +0.545 |
superpower | +0.500 |
matter | +0.499 |
perspective | +0.491 |
blonde | +0.491 |
Neg logit contributions | |
bat | -0.125 |
Ben | -0.124 |
Ben | -0.123 |
ila | -0.122 |
se | -0.120 |
Token | the |
Token ID | 262 |
Feature activation | -6.25e-02 |
Pos logit contributions | |
unlike | +0.079 |
matter | +0.074 |
superpower | +0.073 |
blonde | +0.072 |
perspective | +0.072 |
Neg logit contributions | |
Ben | -0.022 |
Ben | -0.022 |
bat | -0.022 |
ila | -0.022 |
se | -0.021 |
Token | store |
Token ID | 3650 |
Feature activation | -3.85e-02 |
Pos logit contributions | |
ila | +0.059 |
Ben | +0.055 |
Ben | +0.055 |
se | +0.051 |
bat | +0.048 |
Neg logit contributions | |
unlike | -0.137 |
matter | -0.132 |
superpower | -0.131 |
tactic | -0.128 |
perspective | -0.125 |
Token | . |
Token ID | 13 |
Feature activation | +1.82e-02 |
Pos logit contributions | |
Ben | +0.033 |
ila | +0.032 |
bat | +0.031 |
se | +0.030 |
Ben | +0.029 |
Neg logit contributions | |
unlike | -0.092 |
matter | -0.083 |
perspective | -0.082 |
blonde | -0.082 |
superpower | -0.082 |
Token | and |
Token ID | 290 |
Feature activation | -7.79e-02 |
Pos logit contributions | |
unlike | +0.113 |
matter | +0.105 |
superpower | +0.105 |
perspective | +0.103 |
blonde | +0.102 |
Neg logit contributions | |
Ben | -0.020 |
bat | -0.019 |
ila | -0.018 |
Ben | -0.018 |
se | -0.018 |
Token | so |
Token ID | 523 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
Ben | +0.115 |
bat | +0.113 |
ila | +0.112 |
Ben | +0.111 |
se | +0.110 |
Neg logit contributions | |
unlike | -0.140 |
superpower | -0.125 |
perspective | -0.123 |
matter | -0.122 |
tactic | -0.121 |
Token | she |
Token ID | 673 |
Feature activation | +1.28e-02 |
Pos logit contributions | |
unlike | +0.248 |
superpower | +0.226 |
matter | +0.226 |
perspective | +0.221 |
tactic | +0.220 |
Neg logit contributions | |
Ben | -0.108 |
bat | -0.105 |
ila | -0.104 |
Ben | -0.104 |
se | -0.102 |
Token | threw |
Token ID | 9617 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
unlike | +0.027 |
matter | +0.025 |
superpower | +0.024 |
blonde | +0.024 |
tactic | +0.023 |
Neg logit contributions | |
Ben | -0.014 |
ila | -0.014 |
Ben | -0.014 |
bat | -0.014 |
se | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | +6.10e-03 |
Pos logit contributions | |
ila | +0.157 |
Ben | +0.156 |
bat | +0.156 |
Ben | +0.155 |
se | +0.153 |
Neg logit contributions | |
unlike | -0.286 |
matter | -0.262 |
superpower | -0.256 |
blonde | -0.254 |
tactic | -0.250 |
Token | tant |
Token ID | 24246 |
Feature activation | +3.10e-01 |
Pos logit contributions | |
unlike | +0.014 |
matter | +0.012 |
superpower | +0.012 |
blonde | +0.012 |
eway | +0.012 |
Neg logit contributions | |
bat | -0.007 |
se | -0.007 |
Ben | -0.006 |
Ben | -0.006 |
ila | -0.006 |
Token | rum |
Token ID | 6582 |
Feature activation | +2.08e-01 |
Pos logit contributions | |
unlike | +0.271 |
former | +0.236 |
matter | +0.230 |
blonde | +0.227 |
superpower | +0.216 |
Neg logit contributions | |
ila | -0.728 |
se | -0.697 |
bat | -0.682 |
Ben | -0.679 |
Ben | -0.673 |
Token | when |
Token ID | 618 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
unlike | +0.537 |
matter | +0.519 |
superpower | +0.507 |
tactic | +0.499 |
blonde | +0.489 |
Neg logit contributions | |
Ben | -0.144 |
bat | -0.128 |
next | -0.127 |
ila | -0.126 |
each | -0.120 |
Token | her |
Token ID | 607 |
Feature activation | +7.26e-02 |
Pos logit contributions | |
unlike | +0.274 |
superpower | +0.250 |
matter | +0.250 |
tactic | +0.245 |
perspective | +0.244 |
Neg logit contributions | |
Ben | -0.115 |
bat | -0.111 |
ila | -0.110 |
Ben | -0.110 |
se | -0.108 |
Token | parents |
Token ID | 3397 |
Feature activation | -1.34e-02 |
Pos logit contributions | |
unlike | +0.179 |
matter | +0.164 |
superpower | +0.161 |
perspective | +0.161 |
tactic | +0.160 |
Neg logit contributions | |
Ben | -0.060 |
bat | -0.059 |
se | -0.058 |
Ben | -0.056 |
ila | -0.056 |
Token | said |
Token ID | 531 |
Feature activation | -5.44e-02 |
Pos logit contributions | |
ila | +0.015 |
Ben | +0.015 |
Ben | +0.014 |
bat | +0.014 |
se | +0.014 |
Neg logit contributions | |
unlike | -0.026 |
matter | -0.025 |
superpower | -0.023 |
blonde | -0.023 |
darkened | -0.023 |
Token | her |
Token ID | 607 |
Feature activation | +6.28e-02 |
Pos logit contributions | |
unlike | +0.038 |
superpower | +0.036 |
perspective | +0.035 |
matter | +0.035 |
eway | +0.034 |
Neg logit contributions | |
Ben | -0.013 |
each | -0.011 |
bat | -0.011 |
put | -0.010 |
and | -0.010 |
Token | feet |
Token ID | 3625 |
Feature activation | -3.40e-02 |
Pos logit contributions | |
unlike | +0.140 |
matter | +0.126 |
eway | +0.125 |
superpower | +0.124 |
tactic | +0.122 |
Neg logit contributions | |
Ben | -0.071 |
bat | -0.068 |
se | -0.064 |
Ben | -0.063 |
next | -0.063 |
Token | and |
Token ID | 290 |
Feature activation | -8.06e-02 |
Pos logit contributions | |
Ben | +0.017 |
ila | +0.016 |
Ben | +0.016 |
bat | +0.016 |
se | +0.015 |
Neg logit contributions | |
unlike | -0.093 |
matter | -0.086 |
superpower | -0.085 |
blonde | -0.084 |
perspective | -0.084 |
Token | threw |
Token ID | 9617 |
Feature activation | -1.43e-01 |
Pos logit contributions | |
Ben | +0.075 |
bat | +0.070 |
se | +0.069 |
Ben | +0.068 |
ila | +0.067 |
Neg logit contributions | |
unlike | -0.192 |
superpower | -0.178 |
perspective | -0.176 |
matter | -0.175 |
eway | -0.172 |
Token | a |
Token ID | 257 |
Feature activation | -1.77e-03 |
Pos logit contributions | |
Ben | +0.125 |
bat | +0.112 |
ila | +0.109 |
se | +0.108 |
Ben | +0.107 |
Neg logit contributions | |
unlike | -0.354 |
superpower | -0.329 |
perspective | -0.322 |
matter | -0.319 |
tactic | -0.317 |
Token | tant |
Token ID | 24246 |
Feature activation | +3.08e-01 |
Pos logit contributions | |
bat | +0.004 |
se | +0.004 |
Ben | +0.004 |
Ben | +0.004 |
ila | +0.004 |
Neg logit contributions | |
unlike | -0.002 |
superpower | -0.001 |
former | -0.001 |
matter | -0.001 |
eway | -0.001 |
Token | rum |
Token ID | 6582 |
Feature activation | +2.04e-01 |
Pos logit contributions | |
unlike | +0.260 |
former | +0.215 |
matter | +0.211 |
blonde | +0.208 |
superpower | +0.205 |
Neg logit contributions | |
ila | -0.743 |
se | -0.705 |
Ben | -0.697 |
Ben | -0.695 |
bat | -0.694 |
Token | . |
Token ID | 13 |
Feature activation | +2.07e-02 |
Pos logit contributions | |
unlike | +0.522 |
matter | +0.495 |
superpower | +0.490 |
tactic | +0.481 |
perspective | +0.474 |
Neg logit contributions | |
Ben | -0.147 |
ila | -0.133 |
bat | -0.132 |
Ben | -0.129 |
next | -0.125 |
Token | She |
Token ID | 1375 |
Feature activation | +7.36e-02 |
Pos logit contributions | |
unlike | +0.038 |
matter | +0.037 |
superpower | +0.035 |
tactic | +0.035 |
perspective | +0.034 |
Neg logit contributions | |
Ben | -0.033 |
Ben | -0.029 |
ila | -0.029 |
bat | -0.028 |
se | -0.027 |
Token | grabbed |
Token ID | 13176 |
Feature activation | -8.05e-02 |
Pos logit contributions | |
unlike | +0.166 |
superpower | +0.152 |
matter | +0.151 |
perspective | +0.149 |
tactic | +0.147 |
Neg logit contributions | |
Ben | -0.074 |
bat | -0.073 |
Ben | -0.072 |
se | -0.072 |
ila | -0.071 |
Token | the |
Token ID | 262 |
Feature activation | -6.11e-02 |
Pos logit contributions | |
Ben | +0.066 |
bat | +0.065 |
Ben | +0.064 |
ila | +0.064 |
se | +0.063 |
Neg logit contributions | |
unlike | -0.195 |
matter | -0.179 |
superpower | -0.178 |
blonde | -0.174 |
perspective | -0.174 |
Token | to |
Token ID | 284 |
Feature activation | +7.30e-02 |
Pos logit contributions | |
unlike | +0.148 |
matter | +0.134 |
superpower | +0.134 |
perspective | +0.131 |
blonde | +0.130 |
Neg logit contributions | |
Ben | -0.069 |
bat | -0.068 |
se | -0.066 |
Ben | -0.065 |
ila | -0.065 |
Token | cry |
Token ID | 3960 |
Feature activation | -3.02e-03 |
Pos logit contributions | |
unlike | +0.146 |
superpower | +0.131 |
matter | +0.130 |
perspective | +0.127 |
blonde | +0.127 |
Neg logit contributions | |
Ben | -0.092 |
bat | -0.091 |
Ben | -0.090 |
ila | -0.089 |
se | -0.089 |
Token | and |
Token ID | 290 |
Feature activation | -8.00e-02 |
Pos logit contributions | |
Ben | +0.002 |
bat | +0.002 |
Ben | +0.002 |
se | +0.002 |
ila | +0.002 |
Neg logit contributions | |
unlike | -0.008 |
superpower | -0.008 |
matter | -0.008 |
blonde | -0.007 |
perspective | -0.007 |
Token | threw |
Token ID | 9617 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
Ben | +0.084 |
bat | +0.078 |
se | +0.075 |
put | +0.075 |
Ben | +0.071 |
Neg logit contributions | |
unlike | -0.190 |
superpower | -0.173 |
perspective | -0.171 |
eway | -0.170 |
blonde | -0.166 |
Token | a |
Token ID | 257 |
Feature activation | +4.52e-04 |
Pos logit contributions | |
Ben | +0.127 |
bat | +0.114 |
Ben | +0.106 |
ila | +0.106 |
se | +0.105 |
Neg logit contributions | |
unlike | -0.356 |
superpower | -0.326 |
perspective | -0.322 |
eway | -0.319 |
former | -0.318 |
Token | tant |
Token ID | 24246 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
unlike | +0.001 |
former | +0.001 |
eway | +0.001 |
meanings | +0.001 |
avail | +0.001 |
Neg logit contributions | |
bat | -0.001 |
se | -0.001 |
slide | -0.001 |
Ben | -0.001 |
dove | -0.001 |
Token | rum |
Token ID | 6582 |
Feature activation | +2.06e-01 |
Pos logit contributions | |
unlike | +0.273 |
former | +0.232 |
blonde | +0.224 |
matter | +0.221 |
superpower | +0.217 |
Neg logit contributions | |
ila | -0.725 |
se | -0.690 |
Ben | -0.683 |
Ben | -0.682 |
bat | -0.682 |
Token | . |
Token ID | 13 |
Feature activation | +2.41e-02 |
Pos logit contributions | |
unlike | +0.535 |
matter | +0.505 |
superpower | +0.501 |
tactic | +0.491 |
perspective | +0.486 |
Neg logit contributions | |
Ben | -0.142 |
bat | -0.128 |
ila | -0.126 |
next | -0.123 |
Ben | -0.122 |
Token | She |
Token ID | 1375 |
Feature activation | +7.96e-02 |
Pos logit contributions | |
unlike | +0.046 |
matter | +0.045 |
eway | +0.044 |
perspective | +0.044 |
tactic | +0.043 |
Neg logit contributions | |
Ben | -0.037 |
bat | -0.030 |
Tom | -0.030 |
next | -0.030 |
Ben | -0.029 |
Token | didn |
Token ID | 1422 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
unlike | +0.163 |
superpower | +0.146 |
matter | +0.145 |
perspective | +0.143 |
blonde | +0.142 |
Neg logit contributions | |
Ben | -0.098 |
bat | -0.097 |
Ben | -0.095 |
se | -0.095 |
ila | -0.094 |
Token | 't |
Token ID | 470 |
Feature activation | +1.12e-02 |
Pos logit contributions | |
unlike | +0.116 |
superpower | +0.088 |
matter | +0.086 |
eway | +0.085 |
Tears | +0.084 |
Neg logit contributions | |
dads | -0.212 |
se | -0.208 |
pige | -0.207 |
Ben | -0.207 |
ila | -0.206 |
Token | . |
Token ID | 13 |
Feature activation | +1.38e-02 |
Pos logit contributions | |
Ben | +0.088 |
bat | +0.084 |
Ben | +0.080 |
se | +0.079 |
ila | +0.079 |
Neg logit contributions | |
unlike | -0.319 |
superpower | -0.294 |
matter | -0.293 |
blonde | -0.288 |
perspective | -0.287 |
Token | He |
Token ID | 679 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
unlike | +0.024 |
matter | +0.022 |
perspective | +0.021 |
superpower | +0.021 |
tactic | +0.020 |
Neg logit contributions | |
Ben | -0.024 |
Ben | -0.022 |
ila | -0.022 |
bat | -0.021 |
says | -0.021 |
Token | kept |
Token ID | 4030 |
Feature activation | -2.73e-02 |
Pos logit contributions | |
unlike | +0.224 |
matter | +0.203 |
superpower | +0.201 |
blonde | +0.196 |
tactic | +0.192 |
Neg logit contributions | |
Ben | -0.131 |
bat | -0.130 |
ila | -0.129 |
Ben | -0.128 |
se | -0.125 |
Token | throwing |
Token ID | 9644 |
Feature activation | -1.12e-01 |
Pos logit contributions | |
Ben | +0.041 |
ila | +0.040 |
bat | +0.038 |
se | +0.038 |
Ben | +0.037 |
Neg logit contributions | |
unlike | -0.047 |
matter | -0.042 |
superpower | -0.042 |
blonde | -0.041 |
Tears | -0.040 |
Token | a |
Token ID | 257 |
Feature activation | -6.53e-03 |
Pos logit contributions | |
Ben | +0.108 |
bat | +0.107 |
ila | +0.106 |
Ben | +0.106 |
se | +0.104 |
Neg logit contributions | |
unlike | -0.254 |
matter | -0.232 |
superpower | -0.231 |
blonde | -0.226 |
perspective | -0.226 |
Token | tant |
Token ID | 24246 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
bat | +0.008 |
se | +0.008 |
Ben | +0.008 |
Ben | +0.008 |
slide | +0.007 |
Neg logit contributions | |
unlike | -0.014 |
matter | -0.012 |
superpower | -0.012 |
blonde | -0.012 |
eway | -0.012 |
Token | rum |
Token ID | 6582 |
Feature activation | +2.04e-01 |
Pos logit contributions | |
unlike | +0.228 |
blonde | +0.170 |
matter | +0.169 |
superpower | +0.169 |
former | +0.160 |
Neg logit contributions | |
ila | -0.768 |
Ben | -0.747 |
se | -0.747 |
Ben | -0.747 |
bat | -0.743 |
Token | . |
Token ID | 13 |
Feature activation | +1.50e-02 |
Pos logit contributions | |
unlike | +0.498 |
matter | +0.455 |
superpower | +0.455 |
perspective | +0.446 |
blonde | +0.445 |
Neg logit contributions | |
Ben | -0.161 |
bat | -0.159 |
Ben | -0.157 |
ila | -0.157 |
se | -0.154 |
Token | His |
Token ID | 2399 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
unlike | +0.027 |
matter | +0.025 |
perspective | +0.024 |
superpower | +0.024 |
tactic | +0.024 |
Neg logit contributions | |
Ben | -0.026 |
Ben | -0.023 |
ila | -0.023 |
bat | -0.022 |
Tom | -0.022 |
Token | mom |
Token ID | 1995 |
Feature activation | -4.13e-02 |
Pos logit contributions | |
unlike | +0.457 |
matter | +0.416 |
eway | +0.411 |
tactic | +0.410 |
perspective | +0.410 |
Neg logit contributions | |
bat | -0.116 |
Ben | -0.116 |
se | -0.103 |
Ben | -0.098 |
next | -0.098 |
Token | was |
Token ID | 373 |
Feature activation | +8.49e-02 |
Pos logit contributions | |
ila | +0.050 |
bat | +0.050 |
Ben | +0.050 |
Ben | +0.049 |
dads | +0.047 |
Neg logit contributions | |
unlike | -0.078 |
matter | -0.072 |
superpower | -0.068 |
blonde | -0.068 |
darkened | -0.066 |
Token | sad |
Token ID | 6507 |
Feature activation | +1.28e-01 |
Pos logit contributions | |
Ben | +0.113 |
Ben | +0.112 |
bat | +0.112 |
ila | +0.112 |
se | +0.110 |
Neg logit contributions | |
unlike | -0.155 |
matter | -0.139 |
superpower | -0.137 |
blonde | -0.134 |
perspective | -0.133 |
Token | . |
Token ID | 13 |
Feature activation | +1.94e-02 |
Pos logit contributions | |
unlike | +0.367 |
superpower | +0.345 |
eway | +0.344 |
blonde | +0.344 |
matter | +0.344 |
Neg logit contributions | |
Ben | -0.070 |
and | -0.058 |
next | -0.058 |
each | -0.055 |
they | -0.054 |
Token | She |
Token ID | 1375 |
Feature activation | +7.39e-02 |
Pos logit contributions | |
unlike | +0.041 |
superpower | +0.038 |
matter | +0.038 |
perspective | +0.037 |
eway | +0.036 |
Neg logit contributions | |
Ben | -0.024 |
Ben | -0.022 |
ila | -0.021 |
bat | -0.021 |
says | -0.021 |
Token | threw |
Token ID | 9617 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
unlike | +0.164 |
eway | +0.153 |
perspective | +0.153 |
superpower | +0.151 |
tactic | +0.147 |
Neg logit contributions | |
Ben | -0.083 |
put | -0.082 |
bat | -0.078 |
Ben | -0.078 |
se | -0.078 |
Token | a |
Token ID | 257 |
Feature activation | -2.61e-03 |
Pos logit contributions | |
Ben | +0.086 |
bat | +0.078 |
ila | +0.076 |
Ben | +0.075 |
se | +0.073 |
Neg logit contributions | |
unlike | -0.394 |
superpower | -0.365 |
matter | -0.360 |
perspective | -0.359 |
blonde | -0.355 |
Token | tant |
Token ID | 24246 |
Feature activation | +3.04e-01 |
Pos logit contributions | |
bat | +0.005 |
se | +0.005 |
Ben | +0.005 |
slide | +0.004 |
Ben | +0.004 |
Neg logit contributions | |
unlike | -0.004 |
former | -0.004 |
eway | -0.003 |
superpower | -0.003 |
perspective | -0.003 |
Token | rum |
Token ID | 6582 |
Feature activation | +2.02e-01 |
Pos logit contributions | |
unlike | +0.268 |
former | +0.237 |
matter | +0.224 |
blonde | +0.219 |
superpower | +0.215 |
Neg logit contributions | |
ila | -0.722 |
se | -0.677 |
Ben | -0.670 |
Ben | -0.667 |
bat | -0.662 |
Token | and |
Token ID | 290 |
Feature activation | -8.23e-02 |
Pos logit contributions | |
unlike | +0.572 |
matter | +0.546 |
superpower | +0.540 |
tactic | +0.532 |
blonde | +0.524 |
Neg logit contributions | |
Ben | -0.090 |
bat | -0.076 |
ila | -0.075 |
Ben | -0.073 |
next | -0.071 |
Token | cried |
Token ID | 16896 |
Feature activation | -4.43e-02 |
Pos logit contributions | |
Ben | +0.098 |
bat | +0.097 |
Ben | +0.096 |
ila | +0.096 |
se | +0.095 |
Neg logit contributions | |
unlike | -0.169 |
superpower | -0.153 |
matter | -0.153 |
perspective | -0.149 |
blonde | -0.148 |
Token | loudly |
Token ID | 23112 |
Feature activation | +7.31e-02 |
Pos logit contributions | |
Ben | +0.034 |
bat | +0.031 |
each | +0.030 |
next | +0.028 |
and | +0.028 |
Neg logit contributions | |
unlike | -0.114 |
matter | -0.107 |
superpower | -0.107 |
tactic | -0.106 |
eway | -0.106 |
Token | . |
Token ID | 13 |
Feature activation | +2.03e-02 |
Pos logit contributions | |
unlike | +0.191 |
matter | +0.182 |
perspective | +0.179 |
tactic | +0.179 |
superpower | +0.177 |
Neg logit contributions | |
Ben | -0.053 |
next | -0.044 |
bat | -0.043 |
se | -0.039 |
each | -0.039 |
Token | and |
Token ID | 290 |
Feature activation | -8.39e-02 |
Pos logit contributions | |
unlike | +0.130 |
matter | +0.118 |
superpower | +0.117 |
tactic | +0.115 |
perspective | +0.115 |
Neg logit contributions | |
Ben | -0.031 |
ila | -0.031 |
se | -0.029 |
bat | -0.029 |
Ben | -0.026 |
Token | started |
Token ID | 2067 |
Feature activation | +6.20e-02 |
Pos logit contributions | |
Ben | +0.103 |
ila | +0.102 |
bat | +0.100 |
Ben | +0.099 |
se | +0.097 |
Neg logit contributions | |
unlike | -0.167 |
matter | -0.153 |
superpower | -0.148 |
blonde | -0.146 |
tactic | -0.144 |
Token | to |
Token ID | 284 |
Feature activation | +6.86e-02 |
Pos logit contributions | |
unlike | +0.143 |
matter | +0.132 |
superpower | +0.131 |
perspective | +0.129 |
blonde | +0.127 |
Neg logit contributions | |
Ben | -0.061 |
bat | -0.061 |
se | -0.058 |
Ben | -0.056 |
ila | -0.055 |
Token | throw |
Token ID | 3714 |
Feature activation | -2.05e-01 |
Pos logit contributions | |
unlike | +0.139 |
superpower | +0.122 |
perspective | +0.119 |
eway | +0.119 |
matter | +0.118 |
Neg logit contributions | |
bat | -0.092 |
Ben | -0.092 |
se | -0.087 |
Ben | -0.086 |
put | -0.085 |
Token | a |
Token ID | 257 |
Feature activation | -3.82e-03 |
Pos logit contributions | |
Ben | +0.173 |
bat | +0.161 |
Ben | +0.156 |
ila | +0.153 |
se | +0.150 |
Neg logit contributions | |
unlike | -0.513 |
superpower | -0.471 |
matter | -0.465 |
perspective | -0.464 |
eway | -0.457 |
Token | tant |
Token ID | 24246 |
Feature activation | +3.04e-01 |
Pos logit contributions | |
bat | +0.005 |
se | +0.005 |
Ben | +0.005 |
Ben | +0.005 |
ila | +0.005 |
Neg logit contributions | |
unlike | -0.008 |
former | -0.007 |
superpower | -0.007 |
matter | -0.007 |
blonde | -0.007 |
Token | rum |
Token ID | 6582 |
Feature activation | +2.02e-01 |
Pos logit contributions | |
unlike | +0.253 |
blonde | +0.198 |
superpower | +0.195 |
matter | +0.195 |
former | +0.191 |
Neg logit contributions | |
ila | -0.736 |
Ben | -0.711 |
Ben | -0.710 |
se | -0.710 |
bat | -0.707 |
Token | . |
Token ID | 13 |
Feature activation | +2.02e-02 |
Pos logit contributions | |
unlike | +0.511 |
matter | +0.476 |
superpower | +0.474 |
tactic | +0.462 |
perspective | +0.461 |
Neg logit contributions | |
Ben | -0.151 |
bat | -0.142 |
ila | -0.140 |
Ben | -0.138 |
se | -0.135 |
Token | Suddenly |
Token ID | 24975 |
Feature activation | +1.90e-01 |
Pos logit contributions | |
unlike | +0.039 |
matter | +0.037 |
eway | +0.036 |
superpower | +0.035 |
perspective | +0.035 |
Neg logit contributions | |
Ben | -0.032 |
Ben | -0.027 |
bat | -0.027 |
next | -0.026 |
says | -0.026 |
Token | , |
Token ID | 11 |
Feature activation | +1.02e-02 |
Pos logit contributions | |
unlike | +0.471 |
matter | +0.432 |
superpower | +0.431 |
tactic | +0.426 |
perspective | +0.425 |
Neg logit contributions | |
Ben | -0.159 |
bat | -0.147 |
Ben | -0.146 |
ila | -0.145 |
se | -0.141 |
Token | she |
Token ID | 673 |
Feature activation | +6.13e-03 |
Pos logit contributions | |
unlike | +0.027 |
perspective | +0.026 |
tactic | +0.026 |
superpower | +0.026 |
eway | +0.026 |
Neg logit contributions | |
Ben | -0.009 |
they | -0.007 |
bat | -0.007 |
ila | -0.006 |
Ben | -0.006 |
Token | love |
Token ID | 1842 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
Ben | +0.006 |
each | +0.005 |
bat | +0.005 |
se | +0.005 |
ila | +0.005 |
Neg logit contributions | |
unlike | -0.013 |
superpower | -0.012 |
eway | -0.012 |
darkened | -0.012 |
tactic | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | -7.41e-02 |
Pos logit contributions | |
Ben | +0.010 |
bat | +0.010 |
ila | +0.010 |
Ben | +0.010 |
se | +0.009 |
Neg logit contributions | |
unlike | -0.038 |
superpower | -0.035 |
matter | -0.035 |
tactic | -0.035 |
blonde | -0.035 |
Token | honesty |
Token ID | 24345 |
Feature activation | +1.73e-02 |
Pos logit contributions | |
Ben | +0.078 |
bat | +0.078 |
se | +0.076 |
Ben | +0.075 |
ila | +0.074 |
Neg logit contributions | |
unlike | -0.162 |
superpower | -0.147 |
matter | -0.146 |
perspective | -0.143 |
blonde | -0.143 |
Token | , |
Token ID | 11 |
Feature activation | +1.88e-02 |
Pos logit contributions | |
unlike | +0.043 |
superpower | +0.039 |
matter | +0.039 |
perspective | +0.039 |
blonde | +0.039 |
Neg logit contributions | |
Ben | -0.013 |
bat | -0.013 |
Ben | -0.012 |
ila | -0.012 |
se | -0.012 |
Token | no |
Token ID | 645 |
Feature activation | +3.29e-02 |
Pos logit contributions | |
unlike | +0.039 |
matter | +0.035 |
superpower | +0.035 |
blonde | +0.034 |
perspective | +0.034 |
Neg logit contributions | |
Ben | -0.022 |
ila | -0.022 |
Ben | -0.022 |
bat | -0.021 |
se | -0.021 |
Token | matter |
Token ID | 2300 |
Feature activation | +3.03e-01 |
Pos logit contributions | |
unlike | +0.028 |
eway | +0.024 |
Terr | +0.021 |
superpower | +0.018 |
inct | +0.018 |
Neg logit contributions | |
Ben | -0.083 |
each | -0.081 |
put | -0.081 |
se | -0.081 |
bat | -0.081 |
Token | what |
Token ID | 644 |
Feature activation | +8.20e-02 |
Pos logit contributions | |
unlike | +0.638 |
superpower | +0.576 |
blonde | +0.567 |
Terr | +0.557 |
darkened | +0.556 |
Neg logit contributions | |
Ben | -0.369 |
ila | -0.342 |
each | -0.337 |
bat | -0.330 |
we | -0.326 |
Token | . |
Token ID | 13 |
Feature activation | +2.83e-02 |
Pos logit contributions | |
unlike | +0.194 |
eway | +0.179 |
tactic | +0.178 |
superpower | +0.176 |
perspective | +0.176 |
Neg logit contributions | |
Ben | -0.076 |
bat | -0.071 |
se | -0.068 |
ila | -0.067 |
Ben | -0.067 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +9.09e-03 |
Pos logit contributions | |
unlike | +0.064 |
superpower | +0.058 |
matter | +0.057 |
perspective | +0.056 |
blonde | +0.056 |
Neg logit contributions | |
se | -0.027 |
bat | -0.027 |
ila | -0.027 |
Ben | -0.026 |
Ben | -0.026 |
Token | Jack |
Token ID | 14295 |
Feature activation | +2.23e-02 |
Pos logit contributions | |
unlike | +0.020 |
superpower | +0.018 |
matter | +0.018 |
blonde | +0.018 |
perspective | +0.018 |
Neg logit contributions | |
bat | -0.009 |
se | -0.009 |
ila | -0.009 |
Ben | -0.009 |
Ben | -0.009 |
Token | was |
Token ID | 373 |
Feature activation | +9.31e-02 |
Pos logit contributions | |
unlike | +0.047 |
superpower | +0.043 |
matter | +0.042 |
blonde | +0.042 |
perspective | +0.040 |
Neg logit contributions | |
ila | -0.023 |
Ben | -0.022 |
se | -0.021 |
dads | -0.021 |
bat | -0.021 |
Token | smart |
Token ID | 4451 |
Feature activation | -1.23e-02 |
Pos logit contributions | |
Ben | +0.120 |
ila | +0.119 |
bat | +0.115 |
se | +0.114 |
Ben | +0.110 |
Neg logit contributions | |
unlike | -0.147 |
matter | -0.140 |
superpower | -0.130 |
blonde | -0.129 |
meanings | -0.121 |
Token | . |
Token ID | 13 |
Feature activation | +2.57e-02 |
Pos logit contributions | |
ila | +0.011 |
Ben | +0.011 |
bat | +0.011 |
se | +0.011 |
Ben | +0.011 |
Neg logit contributions | |
unlike | -0.029 |
matter | -0.026 |
superpower | -0.026 |
perspective | -0.026 |
blonde | -0.025 |
Token | It |
Token ID | 632 |
Feature activation | +6.45e-02 |
Pos logit contributions | |
unlike | +0.042 |
superpower | +0.036 |
matter | +0.036 |
blonde | +0.036 |
Tears | +0.034 |
Neg logit contributions | |
se | -0.042 |
bat | -0.040 |
ila | -0.039 |
Ben | -0.039 |
dove | -0.038 |
Token | does |
Token ID | 857 |
Feature activation | +5.19e-02 |
Pos logit contributions | |
unlike | +0.144 |
matter | +0.132 |
superpower | +0.129 |
tactic | +0.124 |
Terr | +0.122 |
Neg logit contributions | |
Ben | -0.065 |
ila | -0.063 |
Ben | -0.062 |
bat | -0.062 |
line | -0.059 |
Token | not |
Token ID | 407 |
Feature activation | -3.39e-02 |
Pos logit contributions | |
unlike | +0.106 |
matter | +0.099 |
superpower | +0.096 |
blonde | +0.092 |
tactic | +0.092 |
Neg logit contributions | |
Ben | -0.061 |
Ben | -0.061 |
se | -0.060 |
ila | -0.060 |
bat | -0.059 |
Token | matter |
Token ID | 2300 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
ila | +0.039 |
Ben | +0.038 |
Ben | +0.038 |
se | +0.037 |
bat | +0.037 |
Neg logit contributions | |
unlike | -0.069 |
matter | -0.068 |
superpower | -0.063 |
blonde | -0.061 |
perspective | -0.060 |
Token | if |
Token ID | 611 |
Feature activation | +8.38e-02 |
Pos logit contributions | |
unlike | +0.565 |
matter | +0.522 |
superpower | +0.510 |
perspective | +0.502 |
tactic | +0.492 |
Neg logit contributions | |
bat | -0.405 |
Ben | -0.403 |
se | -0.402 |
Ben | -0.395 |
ila | -0.394 |
Token | you |
Token ID | 345 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
unlike | +0.207 |
matter | +0.200 |
superpower | +0.191 |
blonde | +0.187 |
perspective | +0.187 |
Neg logit contributions | |
se | -0.056 |
Ben | -0.055 |
bat | -0.054 |
ila | -0.052 |
line | -0.045 |
Token | have |
Token ID | 423 |
Feature activation | -1.34e-01 |
Pos logit contributions | |
ila | +0.129 |
Ben | +0.118 |
se | +0.115 |
Ben | +0.112 |
pige | +0.112 |
Neg logit contributions | |
matter | -0.237 |
unlike | -0.224 |
superpower | -0.219 |
blonde | -0.214 |
tactic | -0.210 |
Token | long |
Token ID | 890 |
Feature activation | -7.10e-03 |
Pos logit contributions | |
Ben | +0.153 |
bat | +0.149 |
Ben | +0.146 |
se | +0.145 |
ila | +0.145 |
Neg logit contributions | |
unlike | -0.289 |
superpower | -0.259 |
matter | -0.254 |
eway | -0.253 |
perspective | -0.252 |
Token | hair |
Token ID | 4190 |
Feature activation | +7.59e-02 |
Pos logit contributions | |
Ben | +0.007 |
Ben | +0.007 |
bat | +0.007 |
ila | +0.007 |
se | +0.007 |
Neg logit contributions | |
unlike | -0.015 |
matter | -0.014 |
superpower | -0.014 |
blonde | -0.014 |
perspective | -0.014 |
Token | kept |
Token ID | 4030 |
Feature activation | -3.01e-02 |
Pos logit contributions | |
unlike | +0.146 |
matter | +0.133 |
superpower | +0.131 |
blonde | +0.128 |
perspective | +0.127 |
Neg logit contributions | |
Ben | -0.088 |
ila | -0.087 |
bat | -0.087 |
Ben | -0.086 |
se | -0.085 |
Token | crying |
Token ID | 13774 |
Feature activation | +5.60e-02 |
Pos logit contributions | |
ila | +0.044 |
Ben | +0.044 |
Ben | +0.044 |
bat | +0.044 |
se | +0.043 |
Neg logit contributions | |
unlike | -0.053 |
matter | -0.047 |
superpower | -0.047 |
blonde | -0.046 |
perspective | -0.045 |
Token | and |
Token ID | 290 |
Feature activation | -8.55e-02 |
Pos logit contributions | |
unlike | +0.159 |
superpower | +0.149 |
matter | +0.148 |
eway | +0.146 |
perspective | +0.146 |
Neg logit contributions | |
Ben | -0.025 |
bat | -0.022 |
Ben | -0.021 |
ila | -0.021 |
se | -0.020 |
Token | throwing |
Token ID | 9644 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
Ben | +0.113 |
bat | +0.108 |
se | +0.105 |
Ben | +0.105 |
ila | +0.104 |
Neg logit contributions | |
unlike | -0.172 |
superpower | -0.156 |
matter | -0.152 |
perspective | -0.151 |
eway | -0.150 |
Token | a |
Token ID | 257 |
Feature activation | -4.28e-03 |
Pos logit contributions | |
Ben | +0.099 |
bat | +0.091 |
Ben | +0.085 |
each | +0.084 |
se | +0.081 |
Neg logit contributions | |
unlike | -0.291 |
superpower | -0.268 |
perspective | -0.267 |
eway | -0.265 |
matter | -0.260 |
Token | tant |
Token ID | 24246 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
bat | +0.008 |
se | +0.007 |
Ben | +0.007 |
Ben | +0.007 |
slide | +0.007 |
Neg logit contributions | |
unlike | -0.007 |
eway | -0.006 |
former | -0.006 |
superpower | -0.006 |
matter | -0.006 |
Token | rum |
Token ID | 6582 |
Feature activation | +2.00e-01 |
Pos logit contributions | |
unlike | +0.266 |
former | +0.221 |
blonde | +0.213 |
matter | +0.211 |
superpower | +0.210 |
Neg logit contributions | |
ila | -0.722 |
Ben | -0.684 |
se | -0.682 |
Ben | -0.679 |
bat | -0.677 |
Token | . |
Token ID | 13 |
Feature activation | +1.79e-02 |
Pos logit contributions | |
unlike | +0.529 |
matter | +0.509 |
superpower | +0.502 |
tactic | +0.496 |
perspective | +0.485 |
Neg logit contributions | |
Ben | -0.132 |
next | -0.116 |
each | -0.110 |
ila | -0.110 |
bat | -0.108 |
Token |
|
Token ID | 198 |
Feature activation | -2.40e-02 |
Pos logit contributions | |
unlike | +0.044 |
matter | +0.043 |
superpower | +0.042 |
perspective | +0.041 |
eway | +0.041 |
Neg logit contributions | |
Ben | -0.019 |
Ben | -0.015 |
next | -0.014 |
Tom | -0.014 |
bat | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -2.34e-02 |
Pos logit contributions | |
Ben | +0.017 |
Ben | +0.017 |
ila | +0.017 |
bat | +0.016 |
se | +0.015 |
Neg logit contributions | |
unlike | -0.061 |
matter | -0.057 |
superpower | -0.057 |
blonde | -0.055 |
perspective | -0.055 |
Token | Finally |
Token ID | 11158 |
Feature activation | +1.86e-01 |
Pos logit contributions | |
Ben | +0.035 |
Ben | +0.034 |
bat | +0.034 |
ila | +0.034 |
se | +0.032 |
Neg logit contributions | |
unlike | -0.042 |
superpower | -0.038 |
matter | -0.037 |
perspective | -0.036 |
tactic | -0.036 |
Token | Billy |
Token ID | 49640 |
Feature activation | +5.67e-02 |
Pos logit contributions | |
unlike | +0.017 |
superpower | +0.015 |
matter | +0.015 |
perspective | +0.015 |
blonde | +0.015 |
Neg logit contributions | |
Ben | -0.008 |
bat | -0.008 |
ila | -0.008 |
Ben | -0.008 |
se | -0.008 |
Token | was |
Token ID | 373 |
Feature activation | +8.96e-02 |
Pos logit contributions | |
unlike | +0.116 |
blonde | +0.105 |
superpower | +0.105 |
matter | +0.104 |
perspective | +0.100 |
Neg logit contributions | |
ila | -0.065 |
Ben | -0.061 |
se | -0.061 |
bat | -0.061 |
Ben | -0.059 |
Token | sad |
Token ID | 6507 |
Feature activation | +1.36e-01 |
Pos logit contributions | |
unlike | +0.190 |
matter | +0.174 |
superpower | +0.173 |
perspective | +0.170 |
tactic | +0.169 |
Neg logit contributions | |
Ben | -0.101 |
bat | -0.099 |
ila | -0.096 |
Ben | -0.096 |
se | -0.096 |
Token | . |
Token ID | 13 |
Feature activation | +2.52e-02 |
Pos logit contributions | |
unlike | +0.355 |
matter | +0.333 |
superpower | +0.331 |
tactic | +0.330 |
eway | +0.327 |
Neg logit contributions | |
bat | -0.088 |
Ben | -0.087 |
se | -0.077 |
each | -0.075 |
next | -0.074 |
Token | No |
Token ID | 1400 |
Feature activation | +8.54e-02 |
Pos logit contributions | |
unlike | +0.044 |
superpower | +0.043 |
perspective | +0.042 |
matter | +0.042 |
former | +0.039 |
Neg logit contributions | |
Ben | -0.038 |
bat | -0.036 |
Tom | -0.034 |
next | -0.032 |
each | -0.032 |
Token | matter |
Token ID | 2300 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
unlike | +0.169 |
superpower | +0.151 |
eway | +0.148 |
Terr | +0.145 |
blonde | +0.144 |
Neg logit contributions | |
Ben | -0.111 |
bat | -0.110 |
se | -0.109 |
next | -0.105 |
Ben | -0.104 |
Token | how |
Token ID | 703 |
Feature activation | +4.58e-02 |
Pos logit contributions | |
unlike | +0.652 |
superpower | +0.603 |
blonde | +0.600 |
Terr | +0.593 |
former | +0.581 |
Neg logit contributions | |
Ben | -0.349 |
next | -0.316 |
we | -0.315 |
each | -0.315 |
ila | -0.313 |
Token | hard |
Token ID | 1327 |
Feature activation | +4.53e-02 |
Pos logit contributions | |
unlike | +0.090 |
matter | +0.081 |
superpower | +0.081 |
perspective | +0.078 |
blonde | +0.078 |
Neg logit contributions | |
Ben | -0.059 |
bat | -0.058 |
ila | -0.057 |
Ben | -0.057 |
se | -0.057 |
Token | he |
Token ID | 339 |
Feature activation | +1.70e-02 |
Pos logit contributions | |
unlike | +0.097 |
matter | +0.088 |
superpower | +0.087 |
blonde | +0.085 |
perspective | +0.084 |
Neg logit contributions | |
ila | -0.050 |
Ben | -0.049 |
Ben | -0.049 |
bat | -0.049 |
se | -0.048 |
Token | tried |
Token ID | 3088 |
Feature activation | +2.45e-02 |
Pos logit contributions | |
unlike | +0.031 |
matter | +0.028 |
superpower | +0.028 |
blonde | +0.027 |
tactic | +0.026 |
Neg logit contributions | |
ila | -0.024 |
Ben | -0.024 |
Ben | -0.024 |
bat | -0.023 |
se | -0.023 |
Token | , |
Token ID | 11 |
Feature activation | +1.67e-02 |
Pos logit contributions | |
unlike | +0.062 |
superpower | +0.058 |
eway | +0.058 |
tactic | +0.058 |
matter | +0.058 |
Neg logit contributions | |
Ben | -0.018 |
bat | -0.017 |
next | -0.015 |
and | -0.015 |
each | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | +2.54e-02 |
Pos logit contributions | |
Ben | +0.041 |
bat | +0.040 |
ila | +0.040 |
Ben | +0.040 |
se | +0.039 |
Neg logit contributions | |
unlike | -0.071 |
matter | -0.064 |
superpower | -0.064 |
perspective | -0.062 |
blonde | -0.062 |
Token | She |
Token ID | 1375 |
Feature activation | +8.21e-02 |
Pos logit contributions | |
unlike | +0.048 |
matter | +0.044 |
superpower | +0.043 |
perspective | +0.042 |
tactic | +0.042 |
Neg logit contributions | |
Ben | -0.036 |
ila | -0.034 |
Ben | -0.033 |
bat | -0.033 |
says | -0.032 |
Token | knew |
Token ID | 2993 |
Feature activation | +4.00e-02 |
Pos logit contributions | |
unlike | +0.148 |
superpower | +0.132 |
matter | +0.131 |
former | +0.123 |
rarely | +0.121 |
Neg logit contributions | |
ila | -0.116 |
Ben | -0.113 |
Ben | -0.113 |
line | -0.110 |
bat | -0.108 |
Token | that |
Token ID | 326 |
Feature activation | -1.02e-01 |
Pos logit contributions | |
unlike | +0.106 |
matter | +0.098 |
superpower | +0.098 |
perspective | +0.096 |
blonde | +0.096 |
Neg logit contributions | |
Ben | -0.024 |
bat | -0.024 |
Ben | -0.023 |
ila | -0.023 |
se | -0.022 |
Token | no |
Token ID | 645 |
Feature activation | +3.18e-02 |
Pos logit contributions | |
ila | +0.111 |
Ben | +0.111 |
Ben | +0.111 |
bat | +0.110 |
se | +0.108 |
Neg logit contributions | |
unlike | -0.220 |
matter | -0.199 |
superpower | -0.197 |
perspective | -0.193 |
blonde | -0.192 |
Token | matter |
Token ID | 2300 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
unlike | +0.051 |
Terr | +0.045 |
eway | +0.045 |
superpower | +0.043 |
blonde | +0.043 |
Neg logit contributions | |
next | -0.053 |
Ben | -0.052 |
put | -0.051 |
each | -0.051 |
bat | -0.050 |
Token | how |
Token ID | 703 |
Feature activation | +4.57e-02 |
Pos logit contributions | |
unlike | +0.607 |
superpower | +0.548 |
blonde | +0.541 |
matter | +0.529 |
tactic | +0.527 |
Neg logit contributions | |
Ben | -0.383 |
ila | -0.365 |
bat | -0.357 |
se | -0.348 |
next | -0.348 |
Token | bitter |
Token ID | 12922 |
Feature activation | +7.26e-02 |
Pos logit contributions | |
unlike | +0.090 |
matter | +0.081 |
superpower | +0.081 |
tactic | +0.079 |
blonde | +0.079 |
Neg logit contributions | |
Ben | -0.059 |
bat | -0.057 |
se | -0.057 |
ila | -0.057 |
Ben | -0.056 |
Token | life |
Token ID | 1204 |
Feature activation | +5.37e-03 |
Pos logit contributions | |
unlike | +0.167 |
superpower | +0.155 |
matter | +0.152 |
perspective | +0.152 |
blonde | +0.151 |
Neg logit contributions | |
Ben | -0.068 |
bat | -0.065 |
ila | -0.063 |
se | -0.063 |
Ben | -0.063 |
Token | got |
Token ID | 1392 |
Feature activation | -5.82e-02 |
Pos logit contributions | |
unlike | +0.011 |
matter | +0.010 |
superpower | +0.010 |
blonde | +0.009 |
tactic | +0.009 |
Neg logit contributions | |
Ben | -0.006 |
Ben | -0.006 |
ila | -0.006 |
bat | -0.006 |
se | -0.006 |
Token | , |
Token ID | 11 |
Feature activation | +1.77e-02 |
Pos logit contributions | |
Ben | +0.051 |
bat | +0.049 |
se | +0.049 |
ila | +0.046 |
next | +0.046 |
Neg logit contributions | |
unlike | -0.137 |
superpower | -0.127 |
perspective | -0.125 |
tactic | -0.125 |
eway | -0.125 |
Token | ami |
Token ID | 6277 |
Feature activation | -1.86e-02 |
Pos logit contributions | |
Ben | +0.067 |
bat | +0.066 |
Ben | +0.064 |
ila | +0.064 |
se | +0.062 |
Neg logit contributions | |
unlike | -0.075 |
matter | -0.068 |
superpower | -0.067 |
perspective | -0.067 |
blonde | -0.065 |
Token | figure |
Token ID | 3785 |
Feature activation | -2.06e-02 |
Pos logit contributions | |
Ben | +0.016 |
se | +0.016 |
bat | +0.015 |
ila | +0.015 |
Ben | +0.015 |
Neg logit contributions | |
unlike | -0.044 |
matter | -0.041 |
superpower | -0.040 |
blonde | -0.040 |
tactic | -0.040 |
Token | . |
Token ID | 13 |
Feature activation | +2.77e-02 |
Pos logit contributions | |
Ben | +0.018 |
Ben | +0.018 |
bat | +0.018 |
ila | +0.017 |
se | +0.017 |
Neg logit contributions | |
unlike | -0.049 |
superpower | -0.044 |
matter | -0.044 |
blonde | -0.043 |
perspective | -0.043 |
Token | But |
Token ID | 887 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
unlike | +0.056 |
matter | +0.054 |
superpower | +0.054 |
perspective | +0.052 |
avail | +0.051 |
Neg logit contributions | |
Ben | -0.034 |
put | -0.030 |
says | -0.030 |
goes | -0.030 |
next | -0.030 |
Token | no |
Token ID | 645 |
Feature activation | +3.35e-02 |
Pos logit contributions | |
unlike | +0.283 |
superpower | +0.275 |
perspective | +0.272 |
eway | +0.272 |
blonde | +0.270 |
Neg logit contributions | |
next | -0.079 |
Ben | -0.078 |
put | -0.074 |
they | -0.069 |
we | -0.066 |
Token | matter |
Token ID | 2300 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
unlike | +0.053 |
Terr | +0.049 |
eway | +0.048 |
superpower | +0.045 |
Tears | +0.044 |
Neg logit contributions | |
Ben | -0.058 |
put | -0.058 |
se | -0.056 |
next | -0.056 |
ila | -0.055 |
Token | how |
Token ID | 703 |
Feature activation | +4.49e-02 |
Pos logit contributions | |
unlike | +0.624 |
superpower | +0.571 |
blonde | +0.556 |
tactic | +0.552 |
Terr | +0.546 |
Neg logit contributions | |
Ben | -0.376 |
ila | -0.353 |
bat | -0.339 |
next | -0.338 |
each | -0.337 |
Token | hard |
Token ID | 1327 |
Feature activation | +4.52e-02 |
Pos logit contributions | |
unlike | +0.086 |
matter | +0.077 |
superpower | +0.077 |
perspective | +0.075 |
blonde | +0.074 |
Neg logit contributions | |
Ben | -0.060 |
bat | -0.059 |
Ben | -0.059 |
ila | -0.058 |
se | -0.058 |
Token | she |
Token ID | 673 |
Feature activation | +1.34e-02 |
Pos logit contributions | |
unlike | +0.107 |
superpower | +0.099 |
perspective | +0.098 |
tactic | +0.097 |
matter | +0.097 |
Neg logit contributions | |
Ben | -0.040 |
bat | -0.037 |
ila | -0.036 |
se | -0.036 |
Ben | -0.036 |
Token | tried |
Token ID | 3088 |
Feature activation | +2.49e-02 |
Pos logit contributions | |
unlike | +0.024 |
matter | +0.023 |
superpower | +0.022 |
tactic | +0.021 |
blonde | +0.020 |
Neg logit contributions | |
ila | -0.018 |
Ben | -0.018 |
Ben | -0.018 |
dads | -0.018 |
pige | -0.018 |
Token | , |
Token ID | 11 |
Feature activation | +1.83e-02 |
Pos logit contributions | |
unlike | +0.064 |
superpower | +0.059 |
matter | +0.059 |
eway | +0.059 |
tactic | +0.059 |
Neg logit contributions | |
Ben | -0.018 |
bat | -0.016 |
next | -0.015 |
ila | -0.015 |
se | -0.014 |
Token | remember |
Token ID | 3505 |
Feature activation | -2.05e-02 |
Pos logit contributions | |
Ben | +0.000 |
ila | +0.000 |
bat | +0.000 |
se | +0.000 |
dads | +0.000 |
Neg logit contributions | |
unlike | -0.001 |
matter | -0.001 |
blonde | -0.001 |
superpower | -0.001 |
alas | -0.001 |
Token | , |
Token ID | 11 |
Feature activation | +1.66e-02 |
Pos logit contributions | |
Ben | +0.019 |
bat | +0.019 |
ila | +0.019 |
Ben | +0.019 |
se | +0.018 |
Neg logit contributions | |
unlike | -0.047 |
matter | -0.043 |
superpower | -0.043 |
perspective | -0.042 |
blonde | -0.042 |
Token | it |
Token ID | 340 |
Feature activation | -1.63e-02 |
Pos logit contributions | |
unlike | +0.034 |
matter | +0.032 |
superpower | +0.030 |
blonde | +0.030 |
former | +0.029 |
Neg logit contributions | |
ila | -0.020 |
Ben | -0.020 |
se | -0.019 |
bat | -0.018 |
Ben | -0.017 |
Token | doesn |
Token ID | 1595 |
Feature activation | +4.49e-02 |
Pos logit contributions | |
Ben | +0.017 |
ila | +0.016 |
bat | +0.016 |
Ben | +0.016 |
se | +0.016 |
Neg logit contributions | |
unlike | -0.036 |
matter | -0.033 |
superpower | -0.033 |
tactic | -0.032 |
perspective | -0.031 |
Token | 't |
Token ID | 470 |
Feature activation | +1.37e-02 |
Pos logit contributions | |
unlike | +0.061 |
matter | +0.051 |
superpower | +0.050 |
tactic | +0.046 |
blonde | +0.045 |
Neg logit contributions | |
Ben | -0.085 |
ila | -0.085 |
Ben | -0.083 |
dads | -0.083 |
se | -0.083 |
Token | matter |
Token ID | 2300 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
unlike | +0.019 |
superpower | +0.016 |
matter | +0.016 |
perspective | +0.015 |
blonde | +0.015 |
Neg logit contributions | |
Ben | -0.026 |
bat | -0.026 |
ila | -0.025 |
Ben | -0.025 |
se | -0.025 |
Token | if |
Token ID | 611 |
Feature activation | +8.45e-02 |
Pos logit contributions | |
unlike | +0.551 |
matter | +0.495 |
superpower | +0.490 |
perspective | +0.479 |
blonde | +0.473 |
Neg logit contributions | |
Ben | -0.422 |
bat | -0.422 |
Ben | -0.420 |
ila | -0.417 |
se | -0.415 |
Token | your |
Token ID | 534 |
Feature activation | +8.34e-03 |
Pos logit contributions | |
unlike | +0.210 |
matter | +0.197 |
superpower | +0.193 |
perspective | +0.189 |
blonde | +0.189 |
Neg logit contributions | |
Ben | -0.061 |
bat | -0.060 |
ila | -0.060 |
se | -0.059 |
Ben | -0.058 |
Token | painting |
Token ID | 12036 |
Feature activation | +3.23e-02 |
Pos logit contributions | |
unlike | +0.016 |
matter | +0.014 |
superpower | +0.014 |
perspective | +0.013 |
blonde | +0.013 |
Neg logit contributions | |
Ben | -0.011 |
bat | -0.011 |
ila | -0.011 |
Ben | -0.011 |
se | -0.011 |
Token | is |
Token ID | 318 |
Feature activation | -4.39e-03 |
Pos logit contributions | |
unlike | +0.070 |
matter | +0.065 |
superpower | +0.064 |
perspective | +0.062 |
tactic | +0.062 |
Neg logit contributions | |
Ben | -0.034 |
Ben | -0.033 |
bat | -0.033 |
ila | -0.033 |
se | -0.032 |
Token | not |
Token ID | 407 |
Feature activation | -3.05e-02 |
Pos logit contributions | |
Ben | +0.005 |
bat | +0.005 |
se | +0.005 |
ila | +0.005 |
Ben | +0.005 |
Neg logit contributions | |
unlike | -0.009 |
superpower | -0.008 |
matter | -0.008 |
tactic | -0.008 |
perspective | -0.008 |
Token | She |
Token ID | 1375 |
Feature activation | +8.04e-02 |
Pos logit contributions | |
unlike | +0.059 |
matter | +0.055 |
superpower | +0.054 |
perspective | +0.053 |
tactic | +0.053 |
Neg logit contributions | |
Ben | -0.025 |
Ben | -0.022 |
bat | -0.022 |
ila | -0.021 |
se | -0.019 |
Token | was |
Token ID | 373 |
Feature activation | +9.09e-02 |
Pos logit contributions | |
unlike | +0.148 |
superpower | +0.134 |
matter | +0.134 |
former | +0.126 |
blonde | +0.125 |
Neg logit contributions | |
ila | -0.109 |
Ben | -0.107 |
Ben | -0.106 |
bat | -0.102 |
line | -0.102 |
Token | reminded |
Token ID | 14516 |
Feature activation | +9.00e-02 |
Pos logit contributions | |
unlike | +0.179 |
superpower | +0.157 |
matter | +0.156 |
blonde | +0.153 |
perspective | +0.149 |
Neg logit contributions | |
Ben | -0.120 |
ila | -0.117 |
Ben | -0.116 |
bat | -0.114 |
se | -0.113 |
Token | that |
Token ID | 326 |
Feature activation | -1.04e-01 |
Pos logit contributions | |
unlike | +0.215 |
matter | +0.196 |
superpower | +0.195 |
tactic | +0.189 |
blonde | +0.188 |
Neg logit contributions | |
bat | -0.076 |
Ben | -0.076 |
Ben | -0.074 |
se | -0.073 |
ila | -0.073 |
Token | no |
Token ID | 645 |
Feature activation | +3.20e-02 |
Pos logit contributions | |
ila | +0.119 |
Ben | +0.117 |
bat | +0.110 |
se | +0.108 |
Ben | +0.107 |
Neg logit contributions | |
unlike | -0.221 |
matter | -0.200 |
superpower | -0.195 |
perspective | -0.191 |
blonde | -0.188 |
Token | matter |
Token ID | 2300 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
unlike | +0.053 |
Terr | +0.047 |
eway | +0.047 |
superpower | +0.045 |
blonde | +0.045 |
Neg logit contributions | |
Ben | -0.053 |
next | -0.052 |
bat | -0.051 |
se | -0.051 |
put | -0.050 |
Token | what |
Token ID | 644 |
Feature activation | +8.14e-02 |
Pos logit contributions | |
unlike | +0.634 |
superpower | +0.575 |
blonde | +0.572 |
Terr | +0.561 |
darkened | +0.556 |
Neg logit contributions | |
Ben | -0.365 |
ila | -0.339 |
each | -0.336 |
next | -0.329 |
we | -0.328 |
Token | , |
Token ID | 11 |
Feature activation | +1.68e-02 |
Pos logit contributions | |
unlike | +0.172 |
eway | +0.157 |
tactic | +0.156 |
perspective | +0.156 |
former | +0.154 |
Neg logit contributions | |
Ben | -0.097 |
bat | -0.091 |
we | -0.087 |
se | -0.087 |
next | -0.086 |
Token | she |
Token ID | 673 |
Feature activation | +1.10e-02 |
Pos logit contributions | |
unlike | +0.039 |
superpower | +0.036 |
matter | +0.036 |
tactic | +0.036 |
eway | +0.035 |
Neg logit contributions | |
Ben | -0.015 |
bat | -0.014 |
ila | -0.014 |
Ben | -0.014 |
se | -0.014 |
Token | had |
Token ID | 550 |
Feature activation | -3.26e-02 |
Pos logit contributions | |
unlike | +0.021 |
matter | +0.019 |
superpower | +0.018 |
blonde | +0.017 |
perspective | +0.017 |
Neg logit contributions | |
Ben | -0.015 |
Ben | -0.015 |
ila | -0.015 |
bat | -0.015 |
se | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | +3.95e-03 |
Pos logit contributions | |
Ben | +0.042 |
bat | +0.041 |
Ben | +0.040 |
ila | +0.040 |
se | +0.040 |
Neg logit contributions | |
unlike | -0.065 |
superpower | -0.058 |
matter | -0.058 |
perspective | -0.057 |
tactic | -0.057 |
Token | always |
Token ID | 1464 |
Feature activation | +1.18e-02 |
Pos logit contributions | |
unlike | +0.113 |
matter | +0.110 |
superpower | +0.108 |
blonde | +0.105 |
perspective | +0.102 |
Neg logit contributions | |
ila | -0.053 |
Ben | -0.050 |
Ben | -0.048 |
se | -0.048 |
pige | -0.048 |
Token | love |
Token ID | 1842 |
Feature activation | -1.55e-02 |
Pos logit contributions | |
matter | +0.023 |
unlike | +0.023 |
superpower | +0.022 |
blonde | +0.021 |
former | +0.020 |
Neg logit contributions | |
ila | -0.014 |
Ben | -0.013 |
pige | -0.013 |
Ben | -0.013 |
se | -0.012 |
Token | you |
Token ID | 345 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
Ben | +0.015 |
ila | +0.014 |
bat | +0.014 |
se | +0.013 |
line | +0.012 |
Neg logit contributions | |
unlike | -0.035 |
matter | -0.032 |
superpower | -0.032 |
blonde | -0.031 |
perspective | -0.031 |
Token | , |
Token ID | 11 |
Feature activation | +1.08e-02 |
Pos logit contributions | |
ila | +0.117 |
se | +0.116 |
Ben | +0.112 |
bat | +0.108 |
Ben | +0.107 |
Neg logit contributions | |
unlike | -0.269 |
matter | -0.249 |
superpower | -0.248 |
blonde | -0.241 |
perspective | -0.239 |
Token | no |
Token ID | 645 |
Feature activation | +3.19e-02 |
Pos logit contributions | |
unlike | +0.020 |
matter | +0.019 |
blonde | +0.017 |
superpower | +0.017 |
alas | +0.016 |
Neg logit contributions | |
Ben | -0.014 |
ila | -0.014 |
se | -0.014 |
bat | -0.013 |
ars | -0.013 |
Token | matter |
Token ID | 2300 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
unlike | +0.049 |
eway | +0.041 |
Terr | +0.041 |
superpower | +0.040 |
blonde | +0.039 |
Neg logit contributions | |
Ben | -0.059 |
bat | -0.058 |
se | -0.056 |
next | -0.055 |
put | -0.055 |
Token | what |
Token ID | 644 |
Feature activation | +8.00e-02 |
Pos logit contributions | |
unlike | +0.646 |
superpower | +0.578 |
blonde | +0.568 |
matter | +0.567 |
tactic | +0.559 |
Neg logit contributions | |
Ben | -0.355 |
ila | -0.336 |
bat | -0.333 |
Ben | -0.326 |
se | -0.318 |
Token | ." |
Token ID | 526 |
Feature activation | -5.10e-02 |
Pos logit contributions | |
unlike | +0.188 |
matter | +0.173 |
superpower | +0.173 |
blonde | +0.169 |
perspective | +0.168 |
Neg logit contributions | |
Ben | -0.068 |
bat | -0.068 |
se | -0.068 |
ila | -0.067 |
Ben | -0.067 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +8.01e-03 |
Pos logit contributions | |
se | +0.064 |
ila | +0.061 |
bat | +0.060 |
wave | +0.059 |
dove | +0.058 |
Neg logit contributions | |
unlike | -0.096 |
Tears | -0.089 |
blonde | -0.088 |
superpower | -0.086 |
Terr | -0.086 |
Token | L |
Token ID | 43 |
Feature activation | -4.55e-02 |
Pos logit contributions | |
unlike | +0.013 |
superpower | +0.012 |
blonde | +0.012 |
Tears | +0.012 |
perspective | +0.011 |
Neg logit contributions | |
se | -0.011 |
dads | -0.011 |
bat | -0.011 |
ila | -0.011 |
Ben | -0.011 |
Token | ily |
Token ID | 813 |
Feature activation | -1.02e-01 |
Pos logit contributions | |
bat | +0.054 |
se | +0.054 |
wave | +0.052 |
dads | +0.049 |
pige | +0.049 |
Neg logit contributions | |
unlike | -0.082 |
matter | -0.081 |
tactic | -0.080 |
blonde | -0.076 |
eway | -0.076 |
Token | and |
Token ID | 290 |
Feature activation | -7.52e-02 |
Pos logit contributions | |
unlike | +0.056 |
matter | +0.052 |
superpower | +0.052 |
perspective | +0.051 |
tactic | +0.051 |
Neg logit contributions | |
Ben | -0.012 |
bat | -0.011 |
se | -0.011 |
Ben | -0.011 |
ila | -0.011 |
Token | felt |
Token ID | 2936 |
Feature activation | +1.26e-02 |
Pos logit contributions | |
ila | +0.081 |
Ben | +0.079 |
bat | +0.074 |
Ben | +0.072 |
se | +0.066 |
Neg logit contributions | |
unlike | -0.168 |
blonde | -0.147 |
matter | -0.147 |
superpower | -0.144 |
Tears | -0.142 |
Token | content |
Token ID | 2695 |
Feature activation | +6.68e-02 |
Pos logit contributions | |
unlike | +0.027 |
matter | +0.024 |
superpower | +0.023 |
blonde | +0.023 |
tactic | +0.023 |
Neg logit contributions | |
Ben | -0.014 |
ila | -0.014 |
Ben | -0.014 |
bat | -0.014 |
se | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | +2.99e-02 |
Pos logit contributions | |
unlike | +0.156 |
superpower | +0.140 |
matter | +0.139 |
perspective | +0.138 |
blonde | +0.137 |
Neg logit contributions | |
ila | -0.062 |
Ben | -0.062 |
Ben | -0.062 |
bat | -0.060 |
se | -0.058 |
Token | No |
Token ID | 1400 |
Feature activation | +8.70e-02 |
Pos logit contributions | |
unlike | +0.047 |
matter | +0.043 |
superpower | +0.041 |
perspective | +0.040 |
tactic | +0.040 |
Neg logit contributions | |
Ben | -0.053 |
bat | -0.049 |
Ben | -0.049 |
ila | -0.048 |
se | -0.047 |
Token | matter |
Token ID | 2300 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
unlike | +0.090 |
eway | +0.071 |
superpower | +0.070 |
blonde | +0.066 |
former | +0.064 |
Neg logit contributions | |
Ben | -0.198 |
se | -0.195 |
bat | -0.193 |
Ben | -0.192 |
ila | -0.192 |
Token | what |
Token ID | 644 |
Feature activation | +8.39e-02 |
Pos logit contributions | |
unlike | +0.657 |
blonde | +0.616 |
superpower | +0.610 |
Terr | +0.597 |
darkened | +0.588 |
Neg logit contributions | |
Ben | -0.360 |
each | -0.318 |
we | -0.317 |
next | -0.314 |
ila | -0.310 |
Token | had |
Token ID | 550 |
Feature activation | -3.13e-02 |
Pos logit contributions | |
unlike | +0.167 |
superpower | +0.149 |
matter | +0.148 |
perspective | +0.147 |
tactic | +0.147 |
Neg logit contributions | |
Ben | -0.111 |
bat | -0.107 |
se | -0.106 |
Ben | -0.105 |
ila | -0.104 |
Token | happened |
Token ID | 3022 |
Feature activation | +7.92e-02 |
Pos logit contributions | |
bat | +0.034 |
Ben | +0.033 |
ila | +0.033 |
Ben | +0.033 |
se | +0.032 |
Neg logit contributions | |
unlike | -0.068 |
matter | -0.062 |
superpower | -0.061 |
blonde | -0.059 |
tactic | -0.059 |
Token | before |
Token ID | 878 |
Feature activation | +7.24e-02 |
Pos logit contributions | |
unlike | +0.183 |
superpower | +0.166 |
matter | +0.166 |
perspective | +0.163 |
blonde | +0.162 |
Neg logit contributions | |
Ben | -0.073 |
bat | -0.072 |
ila | -0.072 |
Ben | -0.072 |
se | -0.070 |
Token | he |
Token ID | 339 |
Feature activation | +1.76e-02 |
Pos logit contributions | |
unlike | +0.162 |
matter | +0.146 |
superpower | +0.146 |
blonde | +0.143 |
perspective | +0.142 |
Neg logit contributions | |
ila | -0.073 |
Ben | -0.073 |
Ben | -0.071 |
bat | -0.070 |
se | -0.068 |
Token | again |
Token ID | 757 |
Feature activation | +7.57e-02 |
Pos logit contributions | |
Ben | +0.015 |
ila | +0.015 |
bat | +0.015 |
Ben | +0.015 |
se | +0.014 |
Neg logit contributions | |
unlike | -0.039 |
superpower | -0.035 |
perspective | -0.035 |
matter | -0.035 |
tactic | -0.034 |
Token | and |
Token ID | 290 |
Feature activation | -7.53e-02 |
Pos logit contributions | |
unlike | +0.160 |
superpower | +0.147 |
perspective | +0.146 |
matter | +0.141 |
former | +0.140 |
Neg logit contributions | |
Ben | -0.083 |
ila | -0.082 |
se | -0.076 |
Ben | -0.075 |
bat | -0.075 |
Token | again |
Token ID | 757 |
Feature activation | +7.61e-02 |
Pos logit contributions | |
ila | +0.092 |
Ben | +0.088 |
pige | +0.083 |
bat | +0.081 |
irts | +0.080 |
Neg logit contributions | |
unlike | -0.138 |
matter | -0.132 |
Tears | -0.130 |
alas | -0.126 |
terribly | -0.123 |
Token | . |
Token ID | 13 |
Feature activation | +2.90e-02 |
Pos logit contributions | |
unlike | +0.155 |
superpower | +0.142 |
perspective | +0.140 |
matter | +0.138 |
blonde | +0.133 |
Neg logit contributions | |
ila | -0.088 |
Ben | -0.088 |
bat | -0.081 |
se | -0.080 |
pige | -0.079 |
Token | No |
Token ID | 1400 |
Feature activation | +8.61e-02 |
Pos logit contributions | |
unlike | +0.062 |
matter | +0.057 |
superpower | +0.056 |
perspective | +0.055 |
blonde | +0.055 |
Neg logit contributions | |
Ben | -0.033 |
bat | -0.032 |
Ben | -0.031 |
ila | -0.031 |
se | -0.030 |
Token | matter |
Token ID | 2300 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
unlike | +0.127 |
superpower | +0.107 |
blonde | +0.103 |
perspective | +0.103 |
eway | +0.103 |
Neg logit contributions | |
Ben | -0.157 |
bat | -0.153 |
se | -0.153 |
ila | -0.153 |
Ben | -0.152 |
Token | how |
Token ID | 703 |
Feature activation | +4.58e-02 |
Pos logit contributions | |
unlike | +0.614 |
blonde | +0.552 |
superpower | +0.551 |
matter | +0.534 |
tactic | +0.533 |
Neg logit contributions | |
Ben | -0.384 |
bat | -0.363 |
ila | -0.358 |
each | -0.348 |
se | -0.346 |
Token | many |
Token ID | 867 |
Feature activation | -1.66e-02 |
Pos logit contributions | |
unlike | +0.081 |
matter | +0.073 |
superpower | +0.073 |
perspective | +0.071 |
blonde | +0.069 |
Neg logit contributions | |
Ben | -0.066 |
bat | -0.065 |
ila | -0.064 |
Ben | -0.064 |
se | -0.063 |
Token | times |
Token ID | 1661 |
Feature activation | +2.89e-02 |
Pos logit contributions | |
Ben | +0.023 |
ila | +0.022 |
Ben | +0.021 |
bat | +0.021 |
se | +0.021 |
Neg logit contributions | |
unlike | -0.032 |
matter | -0.030 |
superpower | -0.028 |
blonde | -0.028 |
tactic | -0.027 |
Token | they |
Token ID | 484 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
unlike | +0.067 |
matter | +0.063 |
superpower | +0.062 |
eway | +0.060 |
tactic | +0.060 |
Neg logit contributions | |
Ben | -0.026 |
ila | -0.023 |
bat | -0.022 |
se | -0.022 |
Ben | -0.021 |
Token | came |
Token ID | 1625 |
Feature activation | +1.64e-02 |
Pos logit contributions | |
ila | +0.132 |
Ben | +0.130 |
Ben | +0.129 |
bat | +0.125 |
se | +0.124 |
Neg logit contributions | |
unlike | -0.189 |
matter | -0.179 |
superpower | -0.173 |
perspective | -0.167 |
blonde | -0.166 |
Token | forget |
Token ID | 6044 |
Feature activation | -7.78e-02 |
Pos logit contributions | |
Ben | +0.034 |
ila | +0.033 |
Ben | +0.033 |
bat | +0.032 |
se | +0.031 |
Neg logit contributions | |
unlike | -0.099 |
matter | -0.092 |
superpower | -0.091 |
perspective | -0.089 |
tactic | -0.089 |
Token | to |
Token ID | 284 |
Feature activation | +7.78e-02 |
Pos logit contributions | |
Ben | +0.055 |
bat | +0.050 |
ila | +0.047 |
se | +0.047 |
Ben | +0.045 |
Neg logit contributions | |
unlike | -0.202 |
superpower | -0.186 |
blonde | -0.186 |
matter | -0.186 |
tactic | -0.182 |
Token | dream |
Token ID | 4320 |
Feature activation | +9.41e-03 |
Pos logit contributions | |
unlike | +0.175 |
matter | +0.164 |
superpower | +0.162 |
perspective | +0.158 |
blonde | +0.157 |
Neg logit contributions | |
Ben | -0.075 |
ila | -0.075 |
Ben | -0.074 |
bat | -0.071 |
se | -0.070 |
Token | , |
Token ID | 11 |
Feature activation | +1.71e-02 |
Pos logit contributions | |
unlike | +0.018 |
matter | +0.016 |
superpower | +0.016 |
perspective | +0.015 |
eway | +0.015 |
Neg logit contributions | |
Ben | -0.013 |
bat | -0.012 |
ila | -0.012 |
Ben | -0.012 |
se | -0.012 |
Token | no |
Token ID | 645 |
Feature activation | +3.26e-02 |
Pos logit contributions | |
unlike | +0.036 |
matter | +0.032 |
superpower | +0.032 |
perspective | +0.031 |
blonde | +0.031 |
Neg logit contributions | |
Ben | -0.020 |
ila | -0.020 |
Ben | -0.019 |
bat | -0.019 |
se | -0.019 |
Token | matter |
Token ID | 2300 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
unlike | +0.036 |
Terr | +0.029 |
blonde | +0.028 |
eway | +0.027 |
superpower | +0.027 |
Neg logit contributions | |
Ben | -0.073 |
se | -0.073 |
next | -0.072 |
bat | -0.071 |
wave | -0.070 |
Token | how |
Token ID | 703 |
Feature activation | +4.62e-02 |
Pos logit contributions | |
unlike | +0.631 |
blonde | +0.579 |
superpower | +0.570 |
tactic | +0.555 |
former | +0.554 |
Neg logit contributions | |
Ben | -0.369 |
ila | -0.345 |
bat | -0.336 |
next | -0.332 |
each | -0.328 |
Token | far |
Token ID | 1290 |
Feature activation | -6.02e-02 |
Pos logit contributions | |
unlike | +0.094 |
matter | +0.085 |
superpower | +0.085 |
blonde | +0.083 |
tactic | +0.083 |
Neg logit contributions | |
Ben | -0.056 |
bat | -0.055 |
ila | -0.055 |
se | -0.054 |
Ben | -0.054 |
Token | away |
Token ID | 1497 |
Feature activation | +1.89e-01 |
Pos logit contributions | |
Ben | +0.063 |
bat | +0.061 |
ila | +0.061 |
se | +0.060 |
Ben | +0.059 |
Neg logit contributions | |
unlike | -0.131 |
matter | -0.123 |
superpower | -0.118 |
tactic | -0.114 |
blonde | -0.113 |
Token | they |
Token ID | 484 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
unlike | +0.432 |
superpower | +0.394 |
matter | +0.394 |
blonde | +0.386 |
perspective | +0.386 |
Neg logit contributions | |
Ben | -0.183 |
bat | -0.178 |
ila | -0.175 |
Ben | -0.173 |
se | -0.172 |
Token | seemed |
Token ID | 3947 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
Ben | +0.137 |
ila | +0.137 |
Ben | +0.137 |
bat | +0.136 |
se | +0.133 |
Neg logit contributions | |
unlike | -0.186 |
matter | -0.170 |
superpower | -0.166 |
perspective | -0.161 |
tactic | -0.159 |
Token | down |
Token ID | 866 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
Ben | +0.109 |
bat | +0.104 |
ila | +0.099 |
se | +0.098 |
Ben | +0.097 |
Neg logit contributions | |
unlike | -0.387 |
matter | -0.360 |
superpower | -0.359 |
blonde | -0.353 |
tactic | -0.352 |
Token | the |
Token ID | 262 |
Feature activation | -6.88e-02 |
Pos logit contributions | |
unlike | +0.308 |
matter | +0.286 |
superpower | +0.286 |
blonde | +0.281 |
perspective | +0.279 |
Neg logit contributions | |
Ben | -0.056 |
bat | -0.053 |
ila | -0.052 |
Ben | -0.050 |
se | -0.050 |
Token | tree |
Token ID | 5509 |
Feature activation | -4.01e-02 |
Pos logit contributions | |
ila | +0.091 |
Ben | +0.089 |
Ben | +0.088 |
bat | +0.085 |
se | +0.085 |
Neg logit contributions | |
unlike | -0.133 |
matter | -0.120 |
superpower | -0.120 |
perspective | -0.117 |
blonde | -0.116 |
Token | and |
Token ID | 290 |
Feature activation | -9.08e-02 |
Pos logit contributions | |
Ben | +0.036 |
ila | +0.035 |
Ben | +0.035 |
bat | +0.034 |
se | +0.034 |
Neg logit contributions | |
unlike | -0.094 |
superpower | -0.086 |
matter | -0.086 |
perspective | -0.085 |
blonde | -0.084 |
Token | scam |
Token ID | 19126 |
Feature activation | -8.89e-02 |
Pos logit contributions | |
ila | +0.084 |
Ben | +0.083 |
Ben | +0.081 |
bat | +0.080 |
se | +0.079 |
Neg logit contributions | |
unlike | -0.209 |
matter | -0.192 |
superpower | -0.190 |
perspective | -0.185 |
blonde | -0.185 |
Token | per |
Token ID | 525 |
Feature activation | -2.57e-01 |
Pos logit contributions | |
Ben | +0.146 |
se | +0.146 |
bat | +0.143 |
Ben | +0.141 |
says | +0.140 |
Neg logit contributions | |
unlike | -0.137 |
eway | -0.120 |
superpower | -0.120 |
blonde | -0.119 |
perspective | -0.116 |
Token | away |
Token ID | 1497 |
Feature activation | +1.59e-01 |
Pos logit contributions | |
Ben | +0.332 |
bat | +0.328 |
ila | +0.326 |
Ben | +0.325 |
se | +0.322 |
Neg logit contributions | |
unlike | -0.499 |
matter | -0.447 |
superpower | -0.446 |
perspective | -0.434 |
blonde | -0.433 |
Token | . |
Token ID | 13 |
Feature activation | +1.28e-02 |
Pos logit contributions | |
unlike | +0.389 |
matter | +0.369 |
superpower | +0.362 |
blonde | +0.356 |
tactic | +0.354 |
Neg logit contributions | |
Ben | -0.124 |
bat | -0.123 |
se | -0.117 |
next | -0.113 |
ila | -0.110 |
Token |
|
Token ID | 198 |
Feature activation | -3.22e-02 |
Pos logit contributions | |
unlike | +0.028 |
matter | +0.026 |
superpower | +0.025 |
perspective | +0.025 |
tactic | +0.025 |
Neg logit contributions | |
Ben | -0.015 |
Ben | -0.014 |
bat | -0.014 |
ila | -0.013 |
se | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -3.15e-02 |
Pos logit contributions | |
Ben | +0.030 |
bat | +0.030 |
se | +0.030 |
ila | +0.029 |
Ben | +0.028 |
Neg logit contributions | |
unlike | -0.073 |
superpower | -0.067 |
matter | -0.065 |
perspective | -0.065 |
blonde | -0.065 |
Token | The |
Token ID | 464 |
Feature activation | -4.24e-02 |
Pos logit contributions | |
Ben | +0.038 |
se | +0.037 |
bat | +0.037 |
ila | +0.036 |
Ben | +0.035 |
Neg logit contributions | |
unlike | -0.064 |
superpower | -0.058 |
matter | -0.057 |
perspective | -0.056 |
blonde | -0.056 |
Token | and |
Token ID | 290 |
Feature activation | -9.44e-02 |
Pos logit contributions | |
unlike | +0.376 |
superpower | +0.354 |
matter | +0.351 |
tactic | +0.348 |
perspective | +0.344 |
Neg logit contributions | |
se | -0.087 |
Ben | -0.085 |
bat | -0.083 |
next | -0.077 |
ila | -0.071 |
Token | fierce |
Token ID | 14800 |
Feature activation | +2.93e-02 |
Pos logit contributions | |
Ben | +0.122 |
se | +0.121 |
bat | +0.118 |
ila | +0.113 |
put | +0.113 |
Neg logit contributions | |
unlike | -0.182 |
superpower | -0.170 |
perspective | -0.167 |
tactic | -0.166 |
blonde | -0.161 |
Token | winds |
Token ID | 13520 |
Feature activation | -7.40e-02 |
Pos logit contributions | |
unlike | +0.062 |
superpower | +0.056 |
matter | +0.056 |
tactic | +0.056 |
perspective | +0.055 |
Neg logit contributions | |
Ben | -0.034 |
se | -0.032 |
bat | -0.032 |
ila | -0.031 |
Ben | -0.030 |
Token | started |
Token ID | 2067 |
Feature activation | +4.12e-02 |
Pos logit contributions | |
Ben | +0.077 |
Ben | +0.076 |
ila | +0.075 |
bat | +0.074 |
se | +0.072 |
Neg logit contributions | |
unlike | -0.163 |
matter | -0.147 |
superpower | -0.145 |
perspective | -0.144 |
blonde | -0.141 |
Token | to |
Token ID | 284 |
Feature activation | +4.30e-02 |
Pos logit contributions | |
unlike | +0.095 |
tactic | +0.090 |
superpower | +0.090 |
matter | +0.089 |
blonde | +0.088 |
Neg logit contributions | |
se | -0.037 |
bat | -0.036 |
Ben | -0.035 |
next | -0.034 |
put | -0.032 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.56e-01 |
Pos logit contributions | |
unlike | +0.092 |
superpower | +0.083 |
matter | +0.082 |
blonde | +0.081 |
perspective | +0.081 |
Neg logit contributions | |
bat | -0.047 |
Ben | -0.047 |
se | -0.046 |
Ben | -0.046 |
ila | -0.046 |
Token | . |
Token ID | 13 |
Feature activation | +1.49e-03 |
Pos logit contributions | |
Ben | +0.261 |
bat | +0.260 |
se | +0.252 |
Ben | +0.241 |
ila | +0.239 |
Neg logit contributions | |
unlike | -0.571 |
superpower | -0.525 |
matter | -0.515 |
perspective | -0.513 |
tactic | -0.511 |
Token | Tim |
Token ID | 5045 |
Feature activation | -3.60e-02 |
Pos logit contributions | |
unlike | +0.003 |
matter | +0.003 |
superpower | +0.003 |
perspective | +0.003 |
tactic | +0.003 |
Neg logit contributions | |
Ben | -0.002 |
Ben | -0.002 |
bat | -0.002 |
ila | -0.002 |
se | -0.002 |
Token | my |
Token ID | 1820 |
Feature activation | +6.26e-02 |
Pos logit contributions | |
Ben | +0.040 |
ila | +0.040 |
bat | +0.039 |
Ben | +0.039 |
se | +0.039 |
Neg logit contributions | |
unlike | -0.076 |
matter | -0.069 |
superpower | -0.069 |
blonde | -0.067 |
perspective | -0.067 |
Token | was |
Token ID | 373 |
Feature activation | +6.89e-02 |
Pos logit contributions | |
unlike | +0.124 |
matter | +0.111 |
superpower | +0.110 |
blonde | +0.108 |
perspective | +0.107 |
Neg logit contributions | |
ila | -0.078 |
Ben | -0.078 |
Ben | -0.076 |
bat | -0.075 |
se | -0.072 |
Token | scared |
Token ID | 12008 |
Feature activation | +9.00e-02 |
Pos logit contributions | |
unlike | +0.134 |
matter | +0.128 |
superpower | +0.126 |
perspective | +0.124 |
tactic | +0.123 |
Neg logit contributions | |
bat | -0.086 |
Ben | -0.086 |
se | -0.083 |
put | -0.083 |
next | -0.081 |
Token | sleeping |
Token ID | 11029 |
Feature activation | -4.76e-02 |
Pos logit contributions | |
unlike | +0.000 |
matter | +0.000 |
superpower | +0.000 |
perspective | +0.000 |
blonde | +0.000 |
Neg logit contributions | |
Ben | -0.000 |
bat | -0.000 |
ila | -0.000 |
Ben | -0.000 |
se | -0.000 |
Token | , |
Token ID | 11 |
Feature activation | -2.59e-03 |
Pos logit contributions | |
Ben | +0.040 |
ila | +0.039 |
bat | +0.039 |
Ben | +0.039 |
se | +0.039 |
Neg logit contributions | |
unlike | -0.113 |
superpower | -0.104 |
matter | -0.103 |
perspective | -0.101 |
blonde | -0.101 |
Token | okay |
Token ID | 8788 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
Ben | +0.003 |
ila | +0.003 |
bat | +0.003 |
Ben | +0.003 |
se | +0.003 |
Neg logit contributions | |
unlike | -0.006 |
matter | -0.005 |
superpower | -0.005 |
blonde | -0.005 |
perspective | -0.005 |
Token | ? |
Token ID | 30 |
Feature activation | +8.90e-03 |
Pos logit contributions | |
unlike | +0.340 |
blonde | +0.315 |
superpower | +0.313 |
perspective | +0.312 |
tactic | +0.309 |
Neg logit contributions | |
Ben | -0.097 |
bat | -0.089 |
next | -0.084 |
we | -0.082 |
put | -0.077 |
Token | Whis |
Token ID | 28424 |
Feature activation | -6.22e-02 |
Pos logit contributions | |
unlike | +0.019 |
matter | +0.017 |
perspective | +0.017 |
superpower | +0.017 |
tactic | +0.017 |
Neg logit contributions | |
Ben | -0.011 |
bat | -0.010 |
Ben | -0.010 |
ila | -0.010 |
says | -0.010 |
Token | per |
Token ID | 525 |
Feature activation | -2.56e-01 |
Pos logit contributions | |
bat | +0.084 |
se | +0.080 |
Ben | +0.073 |
Ben | +0.073 |
goes | +0.073 |
Neg logit contributions | |
unlike | -0.115 |
superpower | -0.107 |
perspective | -0.106 |
blonde | -0.103 |
matter | -0.102 |
Token | means |
Token ID | 1724 |
Feature activation | +3.35e-02 |
Pos logit contributions | |
Ben | +0.259 |
ila | +0.254 |
bat | +0.249 |
Ben | +0.247 |
se | +0.244 |
Neg logit contributions | |
unlike | -0.572 |
matter | -0.518 |
superpower | -0.518 |
perspective | -0.509 |
blonde | -0.507 |
Token | to |
Token ID | 284 |
Feature activation | +4.73e-02 |
Pos logit contributions | |
unlike | +0.075 |
matter | +0.068 |
superpower | +0.067 |
tactic | +0.067 |
perspective | +0.067 |
Neg logit contributions | |
Ben | -0.035 |
bat | -0.034 |
ila | -0.032 |
se | -0.032 |
next | -0.031 |
Token | talk |
Token ID | 1561 |
Feature activation | -4.75e-02 |
Pos logit contributions | |
unlike | +0.104 |
matter | +0.101 |
superpower | +0.097 |
blonde | +0.094 |
perspective | +0.093 |
Neg logit contributions | |
ila | -0.044 |
Ben | -0.043 |
se | -0.042 |
Ben | -0.041 |
bat | -0.041 |
Token | very |
Token ID | 845 |
Feature activation | +1.03e-01 |
Pos logit contributions | |
Ben | +0.037 |
next | +0.034 |
bat | +0.033 |
and | +0.033 |
goes | +0.032 |
Neg logit contributions | |
unlike | -0.118 |
matter | -0.110 |
blonde | -0.110 |
perspective | -0.110 |
tactic | -0.110 |
Token | softly |
Token ID | 26625 |
Feature activation | +1.02e-02 |
Pos logit contributions | |
unlike | +0.164 |
matter | +0.144 |
superpower | +0.143 |
blonde | +0.138 |
perspective | +0.137 |
Neg logit contributions | |
Ben | -0.169 |
bat | -0.168 |
Ben | -0.167 |
ila | -0.167 |
se | -0.166 |
Token | excited |
Token ID | 6568 |
Feature activation | -4.96e-03 |
Pos logit contributions | |
unlike | +0.027 |
matter | +0.024 |
superpower | +0.024 |
perspective | +0.023 |
tactic | +0.023 |
Neg logit contributions | |
Ben | -0.016 |
bat | -0.016 |
ila | -0.016 |
Ben | -0.016 |
se | -0.015 |
Token | to |
Token ID | 284 |
Feature activation | +4.75e-02 |
Pos logit contributions | |
Ben | +0.005 |
ila | +0.005 |
Ben | +0.005 |
bat | +0.005 |
se | +0.005 |
Neg logit contributions | |
unlike | -0.011 |
matter | -0.010 |
superpower | -0.010 |
perspective | -0.010 |
blonde | -0.010 |
Token | feel |
Token ID | 1254 |
Feature activation | -1.70e-02 |
Pos logit contributions | |
unlike | +0.106 |
superpower | +0.101 |
matter | +0.101 |
blonde | +0.097 |
tactic | +0.097 |
Neg logit contributions | |
ila | -0.044 |
Ben | -0.044 |
Ben | -0.039 |
pige | -0.038 |
se | -0.037 |
Token | the |
Token ID | 262 |
Feature activation | -7.71e-02 |
Pos logit contributions | |
Ben | +0.016 |
Ben | +0.016 |
ila | +0.016 |
bat | +0.016 |
se | +0.015 |
Neg logit contributions | |
unlike | -0.039 |
matter | -0.035 |
superpower | -0.035 |
blonde | -0.034 |
perspective | -0.034 |
Token | wind |
Token ID | 2344 |
Feature activation | -8.55e-02 |
Pos logit contributions | |
bat | +0.089 |
se | +0.086 |
Ben | +0.085 |
Ben | +0.082 |
ila | +0.079 |
Neg logit contributions | |
unlike | -0.167 |
superpower | -0.148 |
eway | -0.146 |
tactic | -0.146 |
matter | -0.145 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
Ben | +0.102 |
ila | +0.100 |
Ben | +0.097 |
bat | +0.089 |
dads | +0.086 |
Neg logit contributions | |
unlike | -0.182 |
matter | -0.160 |
superpower | -0.157 |
perspective | -0.155 |
blonde | -0.153 |
Token | through |
Token ID | 832 |
Feature activation | +7.59e-02 |
Pos logit contributions | |
Ben | +0.269 |
bat | +0.265 |
ila | +0.264 |
Ben | +0.264 |
se | +0.259 |
Neg logit contributions | |
unlike | -0.554 |
matter | -0.503 |
superpower | -0.501 |
perspective | -0.489 |
blonde | -0.489 |
Token | their |
Token ID | 511 |
Feature activation | -3.49e-02 |
Pos logit contributions | |
unlike | +0.212 |
superpower | +0.196 |
matter | +0.196 |
tactic | +0.192 |
blonde | +0.192 |
Neg logit contributions | |
Ben | -0.036 |
bat | -0.034 |
Ben | -0.032 |
ila | -0.032 |
se | -0.032 |
Token | hair |
Token ID | 4190 |
Feature activation | +4.73e-02 |
Pos logit contributions | |
ila | +0.050 |
Ben | +0.048 |
Ben | +0.048 |
bat | +0.046 |
se | +0.046 |
Neg logit contributions | |
unlike | -0.063 |
superpower | -0.058 |
matter | -0.057 |
blonde | -0.056 |
perspective | -0.055 |
Token | and |
Token ID | 290 |
Feature activation | -1.07e-01 |
Pos logit contributions | |
unlike | +0.112 |
matter | +0.102 |
superpower | +0.101 |
blonde | +0.100 |
perspective | +0.100 |
Neg logit contributions | |
ila | -0.041 |
Ben | -0.041 |
Ben | -0.039 |
bat | -0.038 |
se | -0.037 |
Token | spread |
Token ID | 4104 |
Feature activation | -1.54e-01 |
Pos logit contributions | |
ila | +0.096 |
Ben | +0.094 |
Ben | +0.093 |
bat | +0.091 |
se | +0.089 |
Neg logit contributions | |
unlike | -0.250 |
matter | -0.229 |
superpower | -0.227 |
blonde | -0.224 |
perspective | -0.222 |
Token | , |
Token ID | 11 |
Feature activation | +1.44e-02 |
Pos logit contributions | |
unlike | +0.276 |
superpower | +0.251 |
matter | +0.251 |
perspective | +0.245 |
tactic | +0.245 |
Neg logit contributions | |
Ben | -0.129 |
bat | -0.125 |
Ben | -0.124 |
ila | -0.123 |
se | -0.123 |
Token | the |
Token ID | 262 |
Feature activation | -5.33e-02 |
Pos logit contributions | |
unlike | +0.033 |
superpower | +0.030 |
matter | +0.030 |
perspective | +0.030 |
tactic | +0.030 |
Neg logit contributions | |
Ben | -0.014 |
bat | -0.013 |
ila | -0.013 |
Ben | -0.013 |
se | -0.013 |
Token | wind |
Token ID | 2344 |
Feature activation | -8.18e-02 |
Pos logit contributions | |
Ben | +0.065 |
bat | +0.064 |
Ben | +0.064 |
se | +0.064 |
ila | +0.063 |
Neg logit contributions | |
unlike | -0.107 |
matter | -0.096 |
superpower | -0.096 |
perspective | -0.094 |
blonde | -0.093 |
Token | started |
Token ID | 2067 |
Feature activation | +4.34e-02 |
Pos logit contributions | |
Ben | +0.100 |
Ben | +0.099 |
ila | +0.096 |
bat | +0.095 |
se | +0.088 |
Neg logit contributions | |
unlike | -0.167 |
matter | -0.149 |
superpower | -0.146 |
perspective | -0.144 |
darkened | -0.144 |
Token | to |
Token ID | 284 |
Feature activation | +4.66e-02 |
Pos logit contributions | |
unlike | +0.099 |
superpower | +0.094 |
tactic | +0.094 |
matter | +0.093 |
blonde | +0.092 |
Neg logit contributions | |
se | -0.041 |
bat | -0.040 |
next | -0.038 |
Ben | -0.037 |
put | -0.036 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.52e-01 |
Pos logit contributions | |
unlike | +0.104 |
matter | +0.100 |
superpower | +0.094 |
perspective | +0.094 |
darkened | +0.092 |
Neg logit contributions | |
ila | -0.047 |
Ben | -0.046 |
Ben | -0.045 |
bat | -0.041 |
dads | -0.039 |
Token | and |
Token ID | 290 |
Feature activation | -8.99e-02 |
Pos logit contributions | |
bat | +0.208 |
Ben | +0.207 |
se | +0.200 |
Ben | +0.187 |
next | +0.183 |
Neg logit contributions | |
unlike | -0.614 |
superpower | -0.571 |
matter | -0.555 |
perspective | -0.554 |
blonde | -0.554 |
Token | it |
Token ID | 340 |
Feature activation | -3.68e-02 |
Pos logit contributions | |
Ben | +0.062 |
bat | +0.060 |
se | +0.058 |
Ben | +0.055 |
ila | +0.055 |
Neg logit contributions | |
unlike | -0.231 |
superpower | -0.214 |
matter | -0.212 |
perspective | -0.211 |
blonde | -0.211 |
Token | made |
Token ID | 925 |
Feature activation | -9.30e-02 |
Pos logit contributions | |
Ben | +0.046 |
bat | +0.045 |
Ben | +0.045 |
ila | +0.045 |
se | +0.044 |
Neg logit contributions | |
unlike | -0.073 |
matter | -0.066 |
superpower | -0.066 |
perspective | -0.064 |
blonde | -0.064 |
Token | Lily |
Token ID | 20037 |
Feature activation | -5.45e-02 |
Pos logit contributions | |
Ben | +0.061 |
bat | +0.051 |
se | +0.048 |
Ben | +0.048 |
each | +0.047 |
Neg logit contributions | |
unlike | -0.252 |
superpower | -0.236 |
tactic | -0.230 |
perspective | -0.230 |
blonde | -0.227 |
Token | sh |
Token ID | 427 |
Feature activation | -6.68e-02 |
Pos logit contributions | |
bat | +0.059 |
Ben | +0.057 |
se | +0.055 |
ila | +0.053 |
put | +0.052 |
Neg logit contributions | |
unlike | -0.120 |
matter | -0.109 |
superpower | -0.109 |
perspective | -0.107 |
blonde | -0.105 |
Token | , |
Token ID | 11 |
Feature activation | +6.51e-04 |
Pos logit contributions | |
next | +0.100 |
Ben | +0.098 |
bat | +0.092 |
ila | +0.088 |
se | +0.087 |
Neg logit contributions | |
unlike | -0.317 |
matter | -0.302 |
blonde | -0.289 |
superpower | -0.287 |
meanings | -0.286 |
Token | the |
Token ID | 262 |
Feature activation | -7.32e-02 |
Pos logit contributions | |
unlike | +0.002 |
tactic | +0.001 |
superpower | +0.001 |
matter | +0.001 |
perspective | +0.001 |
Neg logit contributions | |
Ben | -0.001 |
they | -0.001 |
bat | -0.001 |
Ben | -0.001 |
ila | -0.001 |
Token | wind |
Token ID | 2344 |
Feature activation | -8.00e-02 |
Pos logit contributions | |
Ben | +0.088 |
bat | +0.087 |
Ben | +0.086 |
ila | +0.085 |
se | +0.085 |
Neg logit contributions | |
unlike | -0.150 |
matter | -0.135 |
superpower | -0.134 |
tactic | -0.131 |
perspective | -0.131 |
Token | started |
Token ID | 2067 |
Feature activation | +4.26e-02 |
Pos logit contributions | |
Ben | +0.101 |
Ben | +0.098 |
ila | +0.098 |
bat | +0.096 |
se | +0.090 |
Neg logit contributions | |
unlike | -0.161 |
matter | -0.144 |
superpower | -0.142 |
perspective | -0.138 |
darkened | -0.137 |
Token | to |
Token ID | 284 |
Feature activation | +4.75e-02 |
Pos logit contributions | |
unlike | +0.094 |
matter | +0.087 |
superpower | +0.087 |
tactic | +0.085 |
blonde | +0.085 |
Neg logit contributions | |
bat | -0.043 |
se | -0.043 |
Ben | -0.043 |
Ben | -0.040 |
ila | -0.040 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.51e-01 |
Pos logit contributions | |
unlike | +0.101 |
matter | +0.099 |
darkened | +0.091 |
superpower | +0.091 |
perspective | +0.090 |
Neg logit contributions | |
ila | -0.053 |
Ben | -0.050 |
Ben | -0.049 |
pige | -0.046 |
dads | -0.045 |
Token | and |
Token ID | 290 |
Feature activation | -9.70e-02 |
Pos logit contributions | |
bat | +0.190 |
Ben | +0.189 |
se | +0.183 |
Ben | +0.167 |
next | +0.164 |
Neg logit contributions | |
unlike | -0.629 |
superpower | -0.589 |
perspective | -0.577 |
matter | -0.573 |
tactic | -0.571 |
Token | the |
Token ID | 262 |
Feature activation | -7.55e-02 |
Pos logit contributions | |
Ben | +0.082 |
bat | +0.081 |
se | +0.078 |
Ben | +0.078 |
ila | +0.077 |
Neg logit contributions | |
unlike | -0.233 |
superpower | -0.214 |
matter | -0.213 |
perspective | -0.210 |
blonde | -0.209 |
Token | trees |
Token ID | 7150 |
Feature activation | +8.10e-03 |
Pos logit contributions | |
Ben | +0.078 |
ila | +0.077 |
Ben | +0.076 |
bat | +0.075 |
se | +0.073 |
Neg logit contributions | |
unlike | -0.166 |
matter | -0.152 |
superpower | -0.151 |
perspective | -0.148 |
blonde | -0.147 |
Token | began |
Token ID | 2540 |
Feature activation | +7.69e-02 |
Pos logit contributions | |
unlike | +0.017 |
matter | +0.015 |
superpower | +0.015 |
blonde | +0.014 |
perspective | +0.014 |
Neg logit contributions | |
Ben | -0.010 |
Ben | -0.010 |
ila | -0.010 |
bat | -0.009 |
se | -0.009 |
Token | to |
Token ID | 284 |
Feature activation | +4.97e-02 |
Pos logit contributions | |
unlike | +0.182 |
matter | +0.172 |
superpower | +0.170 |
tactic | +0.169 |
eway | +0.166 |
Neg logit contributions | |
se | -0.069 |
bat | -0.067 |
Ben | -0.063 |
next | -0.061 |
put | -0.058 |
Token | his |
Token ID | 465 |
Feature activation | +7.78e-02 |
Pos logit contributions | |
unlike | +0.108 |
matter | +0.103 |
superpower | +0.099 |
former | +0.096 |
blonde | +0.096 |
Neg logit contributions | |
Ben | -0.043 |
ila | -0.043 |
se | -0.039 |
bat | -0.039 |
goes | -0.039 |
Token | trunk |
Token ID | 21427 |
Feature activation | +5.41e-02 |
Pos logit contributions | |
unlike | +0.168 |
matter | +0.150 |
superpower | +0.147 |
perspective | +0.147 |
blonde | +0.146 |
Neg logit contributions | |
Ben | -0.090 |
bat | -0.089 |
se | -0.084 |
Ben | -0.083 |
ila | -0.083 |
Token | and |
Token ID | 290 |
Feature activation | -9.75e-02 |
Pos logit contributions | |
unlike | +0.143 |
matter | +0.131 |
superpower | +0.131 |
perspective | +0.129 |
tactic | +0.129 |
Neg logit contributions | |
Ben | -0.032 |
ila | -0.031 |
Ben | -0.031 |
bat | -0.029 |
se | -0.028 |
Token | started |
Token ID | 2067 |
Feature activation | +4.73e-02 |
Pos logit contributions | |
ila | +0.087 |
Ben | +0.086 |
bat | +0.082 |
Ben | +0.081 |
se | +0.079 |
Neg logit contributions | |
unlike | -0.227 |
matter | -0.209 |
superpower | -0.207 |
blonde | -0.202 |
tactic | -0.200 |
Token | to |
Token ID | 284 |
Feature activation | +5.01e-02 |
Pos logit contributions | |
unlike | +0.104 |
matter | +0.094 |
superpower | +0.094 |
perspective | +0.092 |
blonde | +0.091 |
Neg logit contributions | |
Ben | -0.050 |
bat | -0.050 |
se | -0.048 |
ila | -0.048 |
Ben | -0.048 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.50e-01 |
Pos logit contributions | |
unlike | +0.116 |
superpower | +0.105 |
matter | +0.105 |
perspective | +0.103 |
tactic | +0.102 |
Neg logit contributions | |
bat | -0.048 |
Ben | -0.047 |
se | -0.046 |
Ben | -0.045 |
ila | -0.044 |
Token | . |
Token ID | 13 |
Feature activation | +4.63e-03 |
Pos logit contributions | |
ila | +0.233 |
Ben | +0.232 |
Ben | +0.231 |
bat | +0.230 |
se | +0.225 |
Neg logit contributions | |
unlike | -0.569 |
matter | -0.521 |
superpower | -0.518 |
perspective | -0.505 |
blonde | -0.505 |
Token | A |
Token ID | 317 |
Feature activation | +7.63e-02 |
Pos logit contributions | |
unlike | +0.008 |
matter | +0.007 |
superpower | +0.007 |
perspective | +0.007 |
tactic | +0.007 |
Neg logit contributions | |
Ben | -0.007 |
bat | -0.007 |
Ben | -0.007 |
ila | -0.007 |
se | -0.007 |
Token | bunch |
Token ID | 7684 |
Feature activation | +3.01e-02 |
Pos logit contributions | |
unlike | +0.170 |
eway | +0.152 |
tactic | +0.151 |
superpower | +0.150 |
matter | +0.150 |
Neg logit contributions | |
se | -0.085 |
bat | -0.085 |
Ben | -0.082 |
Ben | -0.082 |
slide | -0.077 |
Token | of |
Token ID | 286 |
Feature activation | -7.97e-04 |
Pos logit contributions | |
unlike | +0.064 |
superpower | +0.055 |
matter | +0.054 |
darkened | +0.053 |
eway | +0.053 |
Neg logit contributions | |
Ben | -0.032 |
Ben | -0.032 |
bat | -0.031 |
ila | -0.031 |
dads | -0.031 |
Token | delicious |
Token ID | 12625 |
Feature activation | +6.47e-02 |
Pos logit contributions | |
Ben | +0.001 |
bat | +0.001 |
se | +0.001 |
ila | +0.001 |
Ben | +0.001 |
Neg logit contributions | |
unlike | -0.002 |
superpower | -0.002 |
matter | -0.002 |
perspective | -0.002 |
tactic | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | -7.25e-02 |
Pos logit contributions | |
unlike | +0.098 |
superpower | +0.089 |
blonde | +0.088 |
matter | +0.088 |
perspective | +0.087 |
Neg logit contributions | |
Ben | -0.047 |
bat | -0.044 |
se | -0.044 |
Ben | -0.041 |
ila | -0.041 |
Token | other |
Token ID | 584 |
Feature activation | +3.01e-02 |
Pos logit contributions | |
bat | +0.090 |
Ben | +0.090 |
se | +0.088 |
Ben | +0.087 |
ila | +0.085 |
Neg logit contributions | |
unlike | -0.147 |
superpower | -0.131 |
matter | -0.128 |
blonde | -0.127 |
perspective | -0.127 |
Token | balls |
Token ID | 11333 |
Feature activation | -4.08e-02 |
Pos logit contributions | |
unlike | +0.064 |
matter | +0.061 |
blonde | +0.058 |
superpower | +0.057 |
perspective | +0.057 |
Neg logit contributions | |
ila | -0.032 |
se | -0.031 |
Ben | -0.031 |
bat | -0.030 |
Ben | -0.029 |
Token | . |
Token ID | 13 |
Feature activation | +1.03e-03 |
Pos logit contributions | |
Ben | +0.042 |
ila | +0.041 |
Ben | +0.038 |
bat | +0.037 |
se | +0.037 |
Neg logit contributions | |
unlike | -0.092 |
matter | -0.081 |
superpower | -0.080 |
perspective | -0.080 |
tactic | -0.079 |
Token | Cas |
Token ID | 11294 |
Feature activation | -5.04e-02 |
Pos logit contributions | |
unlike | +0.002 |
matter | +0.002 |
superpower | +0.002 |
perspective | +0.002 |
tactic | +0.002 |
Neg logit contributions | |
Ben | -0.001 |
Ben | -0.001 |
ila | -0.001 |
bat | -0.001 |
says | -0.001 |
Token | per |
Token ID | 525 |
Feature activation | -2.50e-01 |
Pos logit contributions | |
bat | +0.048 |
se | +0.046 |
Ben | +0.045 |
Ben | +0.045 |
ila | +0.045 |
Neg logit contributions | |
unlike | -0.116 |
superpower | -0.105 |
matter | -0.105 |
perspective | -0.103 |
blonde | -0.101 |
Token | started |
Token ID | 2067 |
Feature activation | +4.82e-02 |
Pos logit contributions | |
Ben | +0.297 |
bat | +0.295 |
Ben | +0.292 |
ila | +0.290 |
se | +0.281 |
Neg logit contributions | |
unlike | -0.505 |
matter | -0.454 |
superpower | -0.451 |
blonde | -0.440 |
tactic | -0.438 |
Token | to |
Token ID | 284 |
Feature activation | +5.20e-02 |
Pos logit contributions | |
unlike | +0.117 |
matter | +0.109 |
perspective | +0.106 |
eway | +0.106 |
superpower | +0.106 |
Neg logit contributions | |
bat | -0.043 |
se | -0.042 |
Ben | -0.041 |
put | -0.039 |
next | -0.036 |
Token | display |
Token ID | 3359 |
Feature activation | -5.48e-02 |
Pos logit contributions | |
unlike | +0.120 |
superpower | +0.105 |
eway | +0.104 |
perspective | +0.103 |
blonde | +0.102 |
Neg logit contributions | |
bat | -0.058 |
se | -0.054 |
Ben | -0.054 |
put | -0.054 |
say | -0.052 |
Token | them |
Token ID | 606 |
Feature activation | -5.67e-03 |
Pos logit contributions | |
Ben | +0.038 |
bat | +0.031 |
se | +0.029 |
each | +0.029 |
ila | +0.028 |
Neg logit contributions | |
unlike | -0.148 |
perspective | -0.137 |
matter | -0.136 |
superpower | -0.136 |
tactic | -0.134 |
Token | in |
Token ID | 287 |
Feature activation | +1.36e-02 |
Pos logit contributions | |
Ben | +0.005 |
bat | +0.005 |
next | +0.005 |
each | +0.004 |
se | +0.004 |
Neg logit contributions | |
unlike | -0.014 |
superpower | -0.013 |
matter | -0.013 |
blonde | -0.013 |
perspective | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | +7.71e-03 |
Pos logit contributions | |
Ben | +0.004 |
bat | +0.004 |
se | +0.004 |
ila | +0.004 |
Ben | +0.004 |
Neg logit contributions | |
unlike | -0.016 |
matter | -0.015 |
superpower | -0.015 |
blonde | -0.014 |
perspective | -0.014 |
Token | |
Token ID | 220 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
unlike | +0.015 |
matter | +0.014 |
superpower | +0.014 |
perspective | +0.014 |
tactic | +0.013 |
Neg logit contributions | |
Ben | -0.011 |
Ben | -0.010 |
bat | -0.010 |
ila | -0.010 |
se | -0.009 |
Token |
|
Token ID | 198 |
Feature activation | -3.39e-02 |
Pos logit contributions | |
unlike | +0.245 |
superpower | +0.223 |
matter | +0.220 |
perspective | +0.218 |
tactic | +0.218 |
Neg logit contributions | |
bat | -0.095 |
se | -0.095 |
Ben | -0.092 |
Ben | -0.092 |
ila | -0.091 |
Token |
|
Token ID | 198 |
Feature activation | -3.18e-02 |
Pos logit contributions | |
Ben | +0.028 |
Ben | +0.028 |
bat | +0.027 |
ila | +0.027 |
se | +0.027 |
Neg logit contributions | |
unlike | -0.082 |
matter | -0.075 |
superpower | -0.075 |
blonde | -0.073 |
perspective | -0.073 |
Token | Har |
Token ID | 13587 |
Feature activation | -8.90e-02 |
Pos logit contributions | |
Ben | +0.041 |
Ben | +0.041 |
bat | +0.040 |
ila | +0.040 |
se | +0.039 |
Neg logit contributions | |
unlike | -0.063 |
matter | -0.056 |
superpower | -0.056 |
blonde | -0.054 |
perspective | -0.054 |
Token | per |
Token ID | 525 |
Feature activation | -2.50e-01 |
Pos logit contributions | |
bat | +0.131 |
Ben | +0.128 |
Ben | +0.126 |
se | +0.126 |
ila | +0.123 |
Neg logit contributions | |
unlike | -0.159 |
matter | -0.142 |
superpower | -0.141 |
perspective | -0.139 |
tactic | -0.137 |
Token | 's |
Token ID | 338 |
Feature activation | -4.11e-02 |
Pos logit contributions | |
Ben | +0.318 |
bat | +0.315 |
Ben | +0.313 |
ila | +0.311 |
se | +0.303 |
Neg logit contributions | |
unlike | -0.486 |
matter | -0.435 |
superpower | -0.434 |
blonde | -0.423 |
perspective | -0.421 |
Token | mom |
Token ID | 1995 |
Feature activation | -4.97e-02 |
Pos logit contributions | |
Ben | +0.043 |
bat | +0.042 |
se | +0.042 |
Ben | +0.039 |
next | +0.039 |
Neg logit contributions | |
unlike | -0.094 |
eway | -0.085 |
matter | -0.085 |
tactic | -0.083 |
superpower | -0.082 |
Token | put |
Token ID | 1234 |
Feature activation | -2.21e-01 |
Pos logit contributions | |
bat | +0.058 |
Ben | +0.058 |
ila | +0.057 |
Ben | +0.057 |
se | +0.054 |
Neg logit contributions | |
unlike | -0.101 |
matter | -0.092 |
superpower | -0.090 |
blonde | -0.089 |
perspective | -0.087 |
Token | the |
Token ID | 262 |
Feature activation | -6.88e-02 |
Pos logit contributions | |
Ben | +0.173 |
Ben | +0.173 |
ila | +0.173 |
bat | +0.172 |
se | +0.166 |
Neg logit contributions | |
unlike | -0.538 |
matter | -0.493 |
superpower | -0.493 |
blonde | -0.482 |
tactic | -0.478 |
Token | bowls |
Token ID | 34796 |
Feature activation | -3.14e-02 |
Pos logit contributions | |
se | +0.083 |
bat | +0.083 |
Ben | +0.081 |
Ben | +0.078 |
ila | +0.077 |
Neg logit contributions | |
unlike | -0.142 |
perspective | -0.128 |
superpower | -0.127 |
eway | -0.127 |
matter | -0.127 |
Token |
|
Token ID | 198 |
Feature activation | -3.87e-02 |
Pos logit contributions | |
bat | +0.040 |
se | +0.040 |
Ben | +0.039 |
goes | +0.039 |
ila | +0.039 |
Neg logit contributions | |
unlike | -0.082 |
superpower | -0.075 |
eway | -0.074 |
perspective | -0.074 |
tactic | -0.074 |
Token | The |
Token ID | 464 |
Feature activation | -4.36e-02 |
Pos logit contributions | |
Ben | +0.046 |
ila | +0.045 |
bat | +0.045 |
se | +0.044 |
goes | +0.043 |
Neg logit contributions | |
unlike | -0.077 |
superpower | -0.070 |
matter | -0.069 |
perspective | -0.069 |
blonde | -0.068 |
Token | fan |
Token ID | 4336 |
Feature activation | +5.40e-02 |
Pos logit contributions | |
Ben | +0.048 |
bat | +0.047 |
Ben | +0.046 |
ila | +0.046 |
se | +0.046 |
Neg logit contributions | |
unlike | -0.094 |
matter | -0.085 |
superpower | -0.085 |
perspective | -0.083 |
blonde | -0.083 |
Token | started |
Token ID | 2067 |
Feature activation | +4.96e-02 |
Pos logit contributions | |
unlike | +0.113 |
matter | +0.103 |
superpower | +0.102 |
perspective | +0.100 |
blonde | +0.100 |
Neg logit contributions | |
Ben | -0.061 |
bat | -0.060 |
ila | -0.060 |
Ben | -0.060 |
se | -0.059 |
Token | to |
Token ID | 284 |
Feature activation | +5.34e-02 |
Pos logit contributions | |
unlike | +0.117 |
matter | +0.110 |
superpower | +0.108 |
perspective | +0.107 |
tactic | +0.107 |
Neg logit contributions | |
se | -0.045 |
bat | -0.045 |
Ben | -0.043 |
next | -0.040 |
put | -0.039 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
unlike | +0.116 |
matter | +0.107 |
superpower | +0.105 |
perspective | +0.103 |
blonde | +0.103 |
Neg logit contributions | |
Ben | -0.056 |
ila | -0.056 |
Ben | -0.055 |
bat | -0.054 |
se | -0.053 |
Token | gently |
Token ID | 15165 |
Feature activation | -1.95e-02 |
Pos logit contributions | |
Ben | +0.198 |
bat | +0.193 |
Ben | +0.186 |
se | +0.186 |
ila | +0.183 |
Neg logit contributions | |
unlike | -0.617 |
superpower | -0.568 |
matter | -0.564 |
perspective | -0.556 |
blonde | -0.554 |
Token | and |
Token ID | 290 |
Feature activation | -9.68e-02 |
Pos logit contributions | |
ila | +0.016 |
Ben | +0.016 |
Ben | +0.016 |
bat | +0.015 |
se | +0.015 |
Neg logit contributions | |
unlike | -0.047 |
matter | -0.043 |
superpower | -0.043 |
blonde | -0.042 |
tactic | -0.042 |
Token | soon |
Token ID | 2582 |
Feature activation | +2.47e-04 |
Pos logit contributions | |
Ben | +0.088 |
bat | +0.087 |
ila | +0.086 |
Ben | +0.086 |
se | +0.084 |
Neg logit contributions | |
unlike | -0.224 |
matter | -0.205 |
superpower | -0.205 |
blonde | -0.200 |
perspective | -0.200 |
Token | Molly |
Token ID | 30236 |
Feature activation | -1.83e-02 |
Pos logit contributions | |
unlike | +0.001 |
matter | +0.001 |
superpower | +0.001 |
blonde | +0.001 |
perspective | +0.001 |
Neg logit contributions | |
bat | -0.000 |
Ben | -0.000 |
ila | -0.000 |
se | -0.000 |
Ben | -0.000 |
Token | felt |
Token ID | 2936 |
Feature activation | -6.28e-03 |
Pos logit contributions | |
ila | +0.021 |
Ben | +0.020 |
Ben | +0.020 |
bat | +0.020 |
se | +0.019 |
Neg logit contributions | |
unlike | -0.038 |
superpower | -0.034 |
blonde | -0.033 |
matter | -0.033 |
darkened | -0.032 |
Token | the |
Token ID | 262 |
Feature activation | -7.18e-02 |
Pos logit contributions | |
Ben | +0.131 |
ila | +0.115 |
bat | +0.112 |
Ben | +0.110 |
each | +0.109 |
Neg logit contributions | |
unlike | -0.299 |
superpower | -0.275 |
matter | -0.271 |
tactic | -0.269 |
perspective | -0.265 |
Token | rare |
Token ID | 4071 |
Feature activation | +4.63e-02 |
Pos logit contributions | |
next | +0.089 |
Ben | +0.084 |
se | +0.083 |
bat | +0.082 |
slide | +0.082 |
Neg logit contributions | |
unlike | -0.152 |
tactic | -0.141 |
eway | -0.135 |
superpower | -0.135 |
rum | -0.132 |
Token | rac |
Token ID | 3444 |
Feature activation | -7.66e-02 |
Pos logit contributions | |
unlike | +0.095 |
superpower | +0.085 |
tactic | +0.084 |
alas | +0.083 |
matter | +0.082 |
Neg logit contributions | |
next | -0.055 |
slide | -0.055 |
Ben | -0.055 |
se | -0.054 |
bat | -0.054 |
Token | coon |
Token ID | 20912 |
Feature activation | +1.23e-02 |
Pos logit contributions | |
Ben | +0.135 |
bat | +0.135 |
se | +0.132 |
Ben | +0.129 |
says | +0.129 |
Neg logit contributions | |
unlike | -0.110 |
eway | -0.100 |
superpower | -0.096 |
perspective | -0.094 |
matter | -0.093 |
Token | scam |
Token ID | 19126 |
Feature activation | -8.36e-02 |
Pos logit contributions | |
unlike | +0.028 |
matter | +0.025 |
superpower | +0.025 |
perspective | +0.025 |
blonde | +0.024 |
Neg logit contributions | |
Ben | -0.012 |
ila | -0.012 |
Ben | -0.012 |
bat | -0.012 |
se | -0.012 |
Token | per |
Token ID | 525 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
se | +0.131 |
Ben | +0.131 |
says | +0.126 |
bat | +0.126 |
dads | +0.125 |
Neg logit contributions | |
unlike | -0.126 |
eway | -0.124 |
perspective | -0.115 |
terribly | -0.114 |
Tears | -0.113 |
Token | away |
Token ID | 1497 |
Feature activation | +1.64e-01 |
Pos logit contributions | |
Ben | +0.256 |
Ben | +0.253 |
bat | +0.252 |
ila | +0.252 |
se | +0.247 |
Neg logit contributions | |
unlike | -0.547 |
matter | -0.496 |
superpower | -0.495 |
perspective | -0.487 |
blonde | -0.482 |
Token | . |
Token ID | 13 |
Feature activation | +8.64e-03 |
Pos logit contributions | |
unlike | +0.425 |
superpower | +0.399 |
matter | +0.399 |
blonde | +0.390 |
tactic | +0.388 |
Neg logit contributions | |
Ben | -0.108 |
bat | -0.102 |
next | -0.098 |
se | -0.097 |
ila | -0.093 |
Token |
|
Token ID | 198 |
Feature activation | -3.29e-02 |
Pos logit contributions | |
unlike | +0.019 |
matter | +0.019 |
superpower | +0.018 |
perspective | +0.018 |
tactic | +0.018 |
Neg logit contributions | |
Ben | -0.010 |
Ben | -0.009 |
ila | -0.008 |
bat | -0.008 |
says | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | -3.25e-02 |
Pos logit contributions | |
se | +0.035 |
bat | +0.034 |
Ben | +0.033 |
ila | +0.033 |
Ben | +0.032 |
Neg logit contributions | |
unlike | -0.069 |
superpower | -0.063 |
eway | -0.063 |
Tears | -0.061 |
blonde | -0.061 |
Token | Luckily |
Token ID | 42332 |
Feature activation | +1.74e-02 |
Pos logit contributions | |
Ben | +0.037 |
se | +0.036 |
bat | +0.036 |
ila | +0.034 |
says | +0.034 |
Neg logit contributions | |
unlike | -0.067 |
superpower | -0.061 |
eway | -0.060 |
blonde | -0.060 |
perspective | -0.059 |
Token | kids |
Token ID | 3988 |
Feature activation | -7.52e-02 |
Pos logit contributions | |
Ben | +0.078 |
bat | +0.077 |
ila | +0.077 |
Ben | +0.077 |
se | +0.075 |
Neg logit contributions | |
unlike | -0.147 |
matter | -0.134 |
superpower | -0.133 |
perspective | -0.130 |
blonde | -0.130 |
Token | would |
Token ID | 561 |
Feature activation | +8.12e-02 |
Pos logit contributions | |
Ben | +0.087 |
Ben | +0.087 |
ila | +0.086 |
bat | +0.085 |
se | +0.082 |
Neg logit contributions | |
unlike | -0.156 |
superpower | -0.139 |
matter | -0.139 |
perspective | -0.137 |
blonde | -0.136 |
Token | join |
Token ID | 4654 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
unlike | +0.182 |
matter | +0.168 |
superpower | +0.166 |
perspective | +0.163 |
blonde | +0.161 |
Neg logit contributions | |
ila | -0.079 |
Ben | -0.079 |
Ben | -0.078 |
bat | -0.074 |
se | -0.073 |
Token | together |
Token ID | 1978 |
Feature activation | +6.42e-02 |
Pos logit contributions | |
bat | +0.086 |
Ben | +0.085 |
Ben | +0.085 |
ila | +0.084 |
se | +0.084 |
Neg logit contributions | |
unlike | -0.255 |
matter | -0.234 |
superpower | -0.234 |
perspective | -0.229 |
blonde | -0.229 |
Token | to |
Token ID | 284 |
Feature activation | +5.24e-02 |
Pos logit contributions | |
unlike | +0.157 |
superpower | +0.144 |
matter | +0.144 |
perspective | +0.141 |
blonde | +0.140 |
Neg logit contributions | |
ila | -0.050 |
Ben | -0.050 |
Ben | -0.049 |
bat | -0.049 |
se | -0.048 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
unlike | +0.118 |
matter | +0.112 |
superpower | +0.110 |
perspective | +0.108 |
blonde | +0.107 |
Neg logit contributions | |
ila | -0.048 |
Ben | -0.048 |
Ben | -0.046 |
se | -0.045 |
bat | -0.044 |
Token | bubbles |
Token ID | 25037 |
Feature activation | -2.04e-02 |
Pos logit contributions | |
ila | +0.264 |
Ben | +0.264 |
Ben | +0.262 |
bat | +0.258 |
se | +0.255 |
Neg logit contributions | |
unlike | -0.531 |
matter | -0.483 |
superpower | -0.482 |
tactic | -0.469 |
blonde | -0.469 |
Token | and |
Token ID | 290 |
Feature activation | -9.38e-02 |
Pos logit contributions | |
ila | +0.022 |
Ben | +0.022 |
Ben | +0.021 |
dads | +0.019 |
se | +0.019 |
Neg logit contributions | |
unlike | -0.044 |
superpower | -0.039 |
matter | -0.038 |
eway | -0.038 |
tactic | -0.038 |
Token | have |
Token ID | 423 |
Feature activation | -1.54e-01 |
Pos logit contributions | |
ila | +0.090 |
Ben | +0.088 |
Ben | +0.084 |
bat | +0.082 |
se | +0.081 |
Neg logit contributions | |
unlike | -0.213 |
matter | -0.197 |
superpower | -0.195 |
blonde | -0.190 |
perspective | -0.188 |
Token | a |
Token ID | 257 |
Feature activation | -1.19e-02 |
Pos logit contributions | |
Ben | +0.183 |
bat | +0.181 |
se | +0.177 |
Ben | +0.177 |
ila | +0.175 |
Neg logit contributions | |
unlike | -0.317 |
superpower | -0.285 |
matter | -0.284 |
perspective | -0.278 |
tactic | -0.277 |
Token | good |
Token ID | 922 |
Feature activation | -2.44e-02 |
Pos logit contributions | |
bat | +0.017 |
se | +0.016 |
Ben | +0.016 |
Ben | +0.016 |
slide | +0.015 |
Neg logit contributions | |
unlike | -0.023 |
eway | -0.020 |
matter | -0.019 |
tactic | -0.019 |
perspective | -0.019 |
Token | Soon |
Token ID | 15894 |
Feature activation | +5.89e-02 |
Pos logit contributions | |
unlike | +0.026 |
matter | +0.024 |
superpower | +0.024 |
perspective | +0.023 |
tactic | +0.023 |
Neg logit contributions | |
Ben | -0.019 |
Ben | -0.018 |
bat | -0.017 |
ila | -0.017 |
se | -0.017 |
Token | the |
Token ID | 262 |
Feature activation | -5.89e-02 |
Pos logit contributions | |
unlike | +0.135 |
matter | +0.123 |
superpower | +0.122 |
blonde | +0.119 |
perspective | +0.119 |
Neg logit contributions | |
bat | -0.055 |
Ben | -0.054 |
ila | -0.054 |
se | -0.053 |
Ben | -0.053 |
Token | wind |
Token ID | 2344 |
Feature activation | -7.92e-02 |
Pos logit contributions | |
ila | +0.064 |
Ben | +0.062 |
Ben | +0.062 |
bat | +0.059 |
says | +0.055 |
Neg logit contributions | |
unlike | -0.126 |
matter | -0.116 |
superpower | -0.115 |
perspective | -0.113 |
blonde | -0.110 |
Token | started |
Token ID | 2067 |
Feature activation | +4.98e-02 |
Pos logit contributions | |
Ben | +0.094 |
Ben | +0.093 |
ila | +0.092 |
bat | +0.090 |
se | +0.086 |
Neg logit contributions | |
unlike | -0.164 |
matter | -0.148 |
superpower | -0.145 |
perspective | -0.141 |
darkened | -0.140 |
Token | to |
Token ID | 284 |
Feature activation | +5.10e-02 |
Pos logit contributions | |
unlike | +0.111 |
superpower | +0.104 |
matter | +0.103 |
tactic | +0.102 |
blonde | +0.101 |
Neg logit contributions | |
se | -0.049 |
bat | -0.049 |
Ben | -0.048 |
next | -0.044 |
ila | -0.043 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
unlike | +0.111 |
matter | +0.109 |
darkened | +0.101 |
superpower | +0.099 |
perspective | +0.098 |
Neg logit contributions | |
ila | -0.057 |
Ben | -0.054 |
Ben | -0.052 |
pige | -0.048 |
dads | -0.048 |
Token | and |
Token ID | 290 |
Feature activation | -9.08e-02 |
Pos logit contributions | |
Ben | +0.218 |
bat | +0.215 |
se | +0.209 |
Ben | +0.207 |
ila | +0.204 |
Neg logit contributions | |
unlike | -0.589 |
superpower | -0.541 |
matter | -0.537 |
perspective | -0.529 |
blonde | -0.528 |
Token | the |
Token ID | 262 |
Feature activation | -6.67e-02 |
Pos logit contributions | |
ila | +0.075 |
Ben | +0.074 |
Ben | +0.070 |
bat | +0.069 |
se | +0.067 |
Neg logit contributions | |
unlike | -0.220 |
matter | -0.203 |
superpower | -0.200 |
perspective | -0.194 |
blonde | -0.194 |
Token | waves |
Token ID | 9813 |
Feature activation | -1.70e-01 |
Pos logit contributions | |
ila | +0.070 |
Ben | +0.068 |
Ben | +0.068 |
bat | +0.064 |
se | +0.061 |
Neg logit contributions | |
unlike | -0.146 |
matter | -0.135 |
superpower | -0.133 |
perspective | -0.131 |
blonde | -0.129 |
Token | started |
Token ID | 2067 |
Feature activation | +4.78e-02 |
Pos logit contributions | |
Ben | +0.204 |
Ben | +0.203 |
ila | +0.202 |
bat | +0.197 |
se | +0.191 |
Neg logit contributions | |
unlike | -0.346 |
matter | -0.312 |
superpower | -0.306 |
blonde | -0.299 |
perspective | -0.299 |
Token | to |
Token ID | 284 |
Feature activation | +5.05e-02 |
Pos logit contributions | |
unlike | +0.104 |
matter | +0.095 |
superpower | +0.095 |
perspective | +0.092 |
tactic | +0.092 |
Neg logit contributions | |
Ben | -0.050 |
bat | -0.050 |
se | -0.050 |
ila | -0.048 |
Ben | -0.048 |
Token | fast |
Token ID | 3049 |
Feature activation | -9.27e-02 |
Pos logit contributions | |
unlike | +0.191 |
matter | +0.176 |
superpower | +0.173 |
perspective | +0.171 |
blonde | +0.169 |
Neg logit contributions | |
ila | -0.083 |
Ben | -0.083 |
bat | -0.082 |
se | -0.081 |
Ben | -0.081 |
Token | it |
Token ID | 340 |
Feature activation | -3.19e-02 |
Pos logit contributions | |
Ben | +0.069 |
ila | +0.068 |
bat | +0.065 |
se | +0.064 |
Ben | +0.064 |
Neg logit contributions | |
unlike | -0.233 |
superpower | -0.210 |
matter | -0.209 |
perspective | -0.208 |
blonde | -0.203 |
Token | made |
Token ID | 925 |
Feature activation | -8.85e-02 |
Pos logit contributions | |
Ben | +0.040 |
Ben | +0.039 |
ila | +0.039 |
bat | +0.039 |
se | +0.038 |
Neg logit contributions | |
unlike | -0.064 |
matter | -0.057 |
superpower | -0.057 |
perspective | -0.055 |
blonde | -0.055 |
Token | the |
Token ID | 262 |
Feature activation | -7.13e-02 |
Pos logit contributions | |
Ben | +0.093 |
bat | +0.089 |
ila | +0.088 |
se | +0.087 |
dads | +0.082 |
Neg logit contributions | |
unlike | -0.194 |
matter | -0.178 |
superpower | -0.171 |
eway | -0.169 |
perspective | -0.166 |
Token | wind |
Token ID | 2344 |
Feature activation | -8.09e-02 |
Pos logit contributions | |
ila | +0.081 |
Ben | +0.081 |
Ben | +0.081 |
bat | +0.079 |
se | +0.077 |
Neg logit contributions | |
unlike | -0.148 |
matter | -0.136 |
superpower | -0.134 |
perspective | -0.132 |
blonde | -0.130 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.48e-01 |
Pos logit contributions | |
Ben | +0.093 |
Ben | +0.091 |
ila | +0.090 |
bat | +0.081 |
se | +0.080 |
Neg logit contributions | |
unlike | -0.175 |
matter | -0.156 |
superpower | -0.153 |
perspective | -0.151 |
tactic | -0.147 |
Token | . |
Token ID | 13 |
Feature activation | +1.06e-02 |
Pos logit contributions | |
Ben | +0.271 |
ila | +0.270 |
Ben | +0.270 |
bat | +0.263 |
se | +0.259 |
Neg logit contributions | |
unlike | -0.531 |
matter | -0.484 |
superpower | -0.476 |
perspective | -0.466 |
tactic | -0.462 |
Token |
|
Token ID | 198 |
Feature activation | -3.51e-02 |
Pos logit contributions | |
unlike | +0.020 |
matter | +0.018 |
superpower | +0.018 |
perspective | +0.017 |
tactic | +0.017 |
Neg logit contributions | |
Ben | -0.015 |
bat | -0.014 |
Ben | -0.014 |
ila | -0.014 |
se | -0.014 |
Token |
|
Token ID | 198 |
Feature activation | -3.43e-02 |
Pos logit contributions | |
bat | +0.031 |
Ben | +0.031 |
se | +0.031 |
ila | +0.030 |
Ben | +0.030 |
Neg logit contributions | |
unlike | -0.081 |
superpower | -0.074 |
matter | -0.073 |
perspective | -0.073 |
tactic | -0.072 |
Token | The |
Token ID | 464 |
Feature activation | -4.27e-02 |
Pos logit contributions | |
Ben | +0.051 |
bat | +0.050 |
ila | +0.050 |
se | +0.050 |
Ben | +0.049 |
Neg logit contributions | |
unlike | -0.060 |
matter | -0.054 |
superpower | -0.053 |
perspective | -0.052 |
blonde | -0.052 |
Token | car |
Token ID | 1097 |
Feature activation | +1.44e-02 |
Pos logit contributions | |
ila | +0.054 |
Ben | +0.053 |
Ben | +0.053 |
says | +0.048 |
se | +0.048 |
Neg logit contributions | |
unlike | -0.080 |
matter | -0.078 |
superpower | -0.076 |
perspective | -0.072 |
blonde | -0.072 |
Token | , |
Token ID | 11 |
Feature activation | -1.44e-03 |
Pos logit contributions | |
Ben | +0.113 |
Ben | +0.085 |
they | +0.084 |
ila | +0.081 |
bat | +0.081 |
Neg logit contributions | |
unlike | -0.340 |
perspective | -0.318 |
tactic | -0.317 |
superpower | -0.313 |
eway | -0.310 |
Token | the |
Token ID | 262 |
Feature activation | -7.39e-02 |
Pos logit contributions | |
Ben | +0.001 |
bat | +0.001 |
Ben | +0.001 |
ila | +0.001 |
se | +0.001 |
Neg logit contributions | |
unlike | -0.004 |
superpower | -0.003 |
perspective | -0.003 |
tactic | -0.003 |
matter | -0.003 |
Token | wind |
Token ID | 2344 |
Feature activation | -8.36e-02 |
Pos logit contributions | |
Ben | +0.076 |
ila | +0.074 |
bat | +0.074 |
Ben | +0.074 |
se | +0.072 |
Neg logit contributions | |
unlike | -0.163 |
matter | -0.148 |
superpower | -0.148 |
blonde | -0.145 |
perspective | -0.145 |
Token | started |
Token ID | 2067 |
Feature activation | +4.56e-02 |
Pos logit contributions | |
Ben | +0.095 |
Ben | +0.093 |
ila | +0.093 |
bat | +0.092 |
se | +0.088 |
Neg logit contributions | |
unlike | -0.176 |
matter | -0.159 |
superpower | -0.157 |
perspective | -0.153 |
blonde | -0.152 |
Token | to |
Token ID | 284 |
Feature activation | +4.90e-02 |
Pos logit contributions | |
unlike | +0.106 |
superpower | +0.101 |
tactic | +0.100 |
matter | +0.100 |
eway | +0.100 |
Neg logit contributions | |
se | -0.042 |
bat | -0.042 |
Ben | -0.038 |
next | -0.037 |
put | -0.035 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.48e-01 |
Pos logit contributions | |
unlike | +0.107 |
matter | +0.106 |
darkened | +0.097 |
superpower | +0.096 |
tactic | +0.094 |
Neg logit contributions | |
ila | -0.050 |
Ben | -0.049 |
Ben | -0.048 |
pige | -0.046 |
dads | -0.045 |
Token | harder |
Token ID | 7069 |
Feature activation | +3.74e-02 |
Pos logit contributions | |
Ben | +0.219 |
bat | +0.217 |
se | +0.208 |
Ben | +0.206 |
ila | +0.200 |
Neg logit contributions | |
unlike | -0.588 |
superpower | -0.547 |
perspective | -0.534 |
matter | -0.534 |
blonde | -0.528 |
Token | and |
Token ID | 290 |
Feature activation | -9.79e-02 |
Pos logit contributions | |
unlike | +0.093 |
matter | +0.086 |
perspective | +0.084 |
superpower | +0.084 |
tactic | +0.083 |
Neg logit contributions | |
ila | -0.029 |
Ben | -0.029 |
Ben | -0.027 |
bat | -0.026 |
se | -0.024 |
Token | harder |
Token ID | 7069 |
Feature activation | +3.81e-02 |
Pos logit contributions | |
Ben | +0.114 |
bat | +0.111 |
se | +0.109 |
Ben | +0.109 |
ila | +0.109 |
Neg logit contributions | |
unlike | -0.206 |
superpower | -0.187 |
matter | -0.184 |
perspective | -0.183 |
blonde | -0.183 |
Token | . |
Token ID | 13 |
Feature activation | +3.97e-03 |
Pos logit contributions | |
unlike | +0.088 |
matter | +0.080 |
superpower | +0.080 |
perspective | +0.079 |
blonde | +0.078 |
Neg logit contributions | |
Ben | -0.035 |
ila | -0.035 |
Ben | -0.035 |
bat | -0.034 |
se | -0.033 |
Token | Suddenly |
Token ID | 24975 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
unlike | +0.009 |
matter | +0.009 |
perspective | +0.008 |
superpower | +0.008 |
tactic | +0.008 |
Neg logit contributions | |
Ben | -0.005 |
Ben | -0.004 |
bat | -0.004 |
ila | -0.004 |
Tom | -0.004 |
Token | , |
Token ID | 11 |
Feature activation | +1.05e-03 |
Pos logit contributions | |
unlike | +0.522 |
superpower | +0.485 |
matter | +0.485 |
blonde | +0.475 |
perspective | +0.475 |
Neg logit contributions | |
Ben | -0.076 |
bat | -0.073 |
Ben | -0.071 |
ila | -0.071 |
se | -0.069 |
Token | the |
Token ID | 262 |
Feature activation | -6.36e-02 |
Pos logit contributions | |
unlike | +0.002 |
matter | +0.002 |
blonde | +0.002 |
superpower | +0.002 |
perspective | +0.002 |
Neg logit contributions | |
ila | -0.001 |
Ben | -0.001 |
bat | -0.001 |
Ben | -0.001 |
se | -0.001 |
Token | wind |
Token ID | 2344 |
Feature activation | -8.07e-02 |
Pos logit contributions | |
bat | +0.071 |
Ben | +0.071 |
se | +0.071 |
Ben | +0.068 |
next | +0.068 |
Neg logit contributions | |
unlike | -0.136 |
tactic | -0.123 |
superpower | -0.122 |
eway | -0.120 |
matter | -0.120 |
Token | started |
Token ID | 2067 |
Feature activation | +4.93e-02 |
Pos logit contributions | |
Ben | +0.101 |
Ben | +0.099 |
ila | +0.099 |
bat | +0.097 |
se | +0.093 |
Neg logit contributions | |
unlike | -0.163 |
matter | -0.144 |
superpower | -0.143 |
perspective | -0.141 |
blonde | -0.139 |
Token | to |
Token ID | 284 |
Feature activation | +5.10e-02 |
Pos logit contributions | |
unlike | +0.113 |
superpower | +0.108 |
matter | +0.107 |
tactic | +0.107 |
blonde | +0.105 |
Neg logit contributions | |
se | -0.045 |
bat | -0.043 |
Ben | -0.042 |
next | -0.041 |
each | -0.039 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.48e-01 |
Pos logit contributions | |
unlike | +0.109 |
matter | +0.102 |
superpower | +0.098 |
perspective | +0.096 |
tactic | +0.095 |
Neg logit contributions | |
Ben | -0.057 |
ila | -0.057 |
Ben | -0.056 |
bat | -0.053 |
se | -0.051 |
Token | , |
Token ID | 11 |
Feature activation | +6.82e-04 |
Pos logit contributions | |
bat | +0.219 |
Ben | +0.218 |
se | +0.215 |
next | +0.199 |
Ben | +0.197 |
Neg logit contributions | |
unlike | -0.580 |
superpower | -0.545 |
perspective | -0.527 |
tactic | -0.527 |
matter | -0.527 |
Token | and |
Token ID | 290 |
Feature activation | -9.31e-02 |
Pos logit contributions | |
unlike | +0.002 |
matter | +0.002 |
superpower | +0.002 |
blonde | +0.001 |
perspective | +0.001 |
Neg logit contributions | |
ila | -0.001 |
Ben | -0.001 |
Ben | -0.001 |
bat | -0.001 |
se | -0.000 |
Token | it |
Token ID | 340 |
Feature activation | -3.45e-02 |
Pos logit contributions | |
Ben | +0.057 |
bat | +0.054 |
se | +0.052 |
ila | +0.052 |
Ben | +0.052 |
Neg logit contributions | |
unlike | -0.246 |
superpower | -0.228 |
matter | -0.227 |
perspective | -0.224 |
tactic | -0.223 |
Token | started |
Token ID | 2067 |
Feature activation | +4.96e-02 |
Pos logit contributions | |
Ben | +0.043 |
ila | +0.043 |
Ben | +0.043 |
bat | +0.042 |
se | +0.041 |
Neg logit contributions | |
unlike | -0.069 |
matter | -0.062 |
superpower | -0.061 |
blonde | -0.059 |
perspective | -0.059 |
Token | to |
Token ID | 284 |
Feature activation | +5.20e-02 |
Pos logit contributions | |
unlike | +0.120 |
tactic | +0.116 |
matter | +0.115 |
superpower | +0.114 |
eway | +0.112 |
Neg logit contributions | |
se | -0.041 |
bat | -0.040 |
next | -0.037 |
Ben | -0.036 |
put | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | +1.42e-03 |
Pos logit contributions | |
unlike | +0.019 |
superpower | +0.018 |
matter | +0.017 |
tactic | +0.017 |
blonde | +0.017 |
Neg logit contributions | |
bat | -0.010 |
se | -0.010 |
Ben | -0.010 |
Ben | -0.009 |
slide | -0.009 |
Token | The |
Token ID | 383 |
Feature activation | -2.35e-02 |
Pos logit contributions | |
unlike | +0.003 |
matter | +0.003 |
superpower | +0.003 |
perspective | +0.003 |
tactic | +0.003 |
Neg logit contributions | |
Ben | -0.002 |
bat | -0.001 |
Ben | -0.001 |
ila | -0.001 |
se | -0.001 |
Token | wind |
Token ID | 2344 |
Feature activation | -7.82e-02 |
Pos logit contributions | |
Ben | +0.025 |
se | +0.025 |
bat | +0.025 |
Ben | +0.024 |
next | +0.024 |
Neg logit contributions | |
unlike | -0.052 |
superpower | -0.047 |
tactic | -0.046 |
matter | -0.046 |
eway | -0.046 |
Token | started |
Token ID | 2067 |
Feature activation | +4.82e-02 |
Pos logit contributions | |
Ben | +0.096 |
Ben | +0.094 |
bat | +0.092 |
ila | +0.092 |
se | +0.089 |
Neg logit contributions | |
unlike | -0.159 |
matter | -0.143 |
superpower | -0.141 |
perspective | -0.136 |
blonde | -0.136 |
Token | to |
Token ID | 284 |
Feature activation | +5.14e-02 |
Pos logit contributions | |
unlike | +0.114 |
superpower | +0.110 |
matter | +0.109 |
tactic | +0.109 |
blonde | +0.108 |
Neg logit contributions | |
se | -0.041 |
bat | -0.039 |
next | -0.037 |
Ben | -0.036 |
put | -0.035 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.47e-01 |
Pos logit contributions | |
unlike | +0.104 |
matter | +0.098 |
superpower | +0.092 |
perspective | +0.091 |
tactic | +0.091 |
Neg logit contributions | |
ila | -0.063 |
Ben | -0.062 |
Ben | -0.062 |
bat | -0.057 |
dads | -0.056 |
Token | hard |
Token ID | 1327 |
Feature activation | +1.61e-02 |
Pos logit contributions | |
bat | +0.178 |
Ben | +0.175 |
se | +0.169 |
next | +0.156 |
Ben | +0.155 |
Neg logit contributions | |
unlike | -0.627 |
superpower | -0.590 |
perspective | -0.577 |
matter | -0.571 |
blonde | -0.570 |
Token | and |
Token ID | 290 |
Feature activation | -9.90e-02 |
Pos logit contributions | |
unlike | +0.043 |
superpower | +0.040 |
matter | +0.039 |
blonde | +0.039 |
perspective | +0.039 |
Neg logit contributions | |
Ben | -0.010 |
bat | -0.009 |
se | -0.009 |
Ben | -0.008 |
ila | -0.008 |
Token | it |
Token ID | 340 |
Feature activation | -3.70e-02 |
Pos logit contributions | |
Ben | +0.077 |
bat | +0.074 |
se | +0.071 |
Ben | +0.069 |
ila | +0.068 |
Neg logit contributions | |
unlike | -0.247 |
superpower | -0.229 |
perspective | -0.226 |
matter | -0.226 |
tactic | -0.224 |
Token | began |
Token ID | 2540 |
Feature activation | +8.10e-02 |
Pos logit contributions | |
bat | +0.042 |
Ben | +0.042 |
se | +0.042 |
Ben | +0.041 |
ila | +0.041 |
Neg logit contributions | |
unlike | -0.077 |
superpower | -0.070 |
matter | -0.069 |
perspective | -0.069 |
blonde | -0.068 |
Token | to |
Token ID | 284 |
Feature activation | +5.15e-02 |
Pos logit contributions | |
unlike | +0.200 |
matter | +0.191 |
tactic | +0.189 |
superpower | +0.189 |
blonde | +0.187 |
Neg logit contributions | |
se | -0.065 |
bat | -0.063 |
next | -0.058 |
Ben | -0.056 |
put | -0.053 |
Token | , |
Token ID | 11 |
Feature activation | -5.53e-04 |
Pos logit contributions | |
Ben | +0.076 |
bat | +0.073 |
se | +0.072 |
ila | +0.072 |
Ben | +0.071 |
Neg logit contributions | |
unlike | -0.188 |
matter | -0.175 |
superpower | -0.172 |
perspective | -0.170 |
tactic | -0.169 |
Token | " |
Token ID | 366 |
Feature activation | +6.06e-02 |
Pos logit contributions | |
Ben | +0.001 |
ila | +0.001 |
bat | +0.001 |
Ben | +0.001 |
se | +0.001 |
Neg logit contributions | |
unlike | -0.001 |
superpower | -0.001 |
matter | -0.001 |
blonde | -0.001 |
perspective | -0.001 |
Token | You |
Token ID | 1639 |
Feature activation | -3.22e-02 |
Pos logit contributions | |
unlike | +0.120 |
matter | +0.110 |
superpower | +0.108 |
perspective | +0.106 |
blonde | +0.105 |
Neg logit contributions | |
Ben | -0.076 |
bat | -0.075 |
ila | -0.074 |
se | -0.074 |
Ben | -0.071 |
Token | have |
Token ID | 423 |
Feature activation | -1.50e-01 |
Pos logit contributions | |
ila | +0.036 |
Ben | +0.036 |
Ben | +0.036 |
bat | +0.035 |
se | +0.035 |
Neg logit contributions | |
unlike | -0.067 |
matter | -0.062 |
superpower | -0.061 |
blonde | -0.059 |
perspective | -0.059 |
Token | to |
Token ID | 284 |
Feature activation | +5.39e-02 |
Pos logit contributions | |
ila | +0.174 |
Ben | +0.170 |
Ben | +0.170 |
bat | +0.166 |
se | +0.163 |
Neg logit contributions | |
unlike | -0.313 |
matter | -0.284 |
superpower | -0.284 |
blonde | -0.275 |
perspective | -0.274 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.47e-01 |
Pos logit contributions | |
matter | +0.109 |
superpower | +0.102 |
unlike | +0.098 |
former | +0.097 |
perspective | +0.095 |
Neg logit contributions | |
ila | -0.066 |
pige | -0.063 |
says | -0.057 |
Ben | -0.057 |
Ben | -0.056 |
Token | on |
Token ID | 319 |
Feature activation | +8.99e-03 |
Pos logit contributions | |
ila | +0.252 |
Ben | +0.244 |
Ben | +0.237 |
se | +0.234 |
bat | +0.229 |
Neg logit contributions | |
unlike | -0.539 |
matter | -0.498 |
superpower | -0.494 |
tactic | -0.475 |
perspective | -0.474 |
Token | the |
Token ID | 262 |
Feature activation | -7.78e-02 |
Pos logit contributions | |
unlike | +0.022 |
matter | +0.021 |
superpower | +0.021 |
perspective | +0.020 |
blonde | +0.020 |
Neg logit contributions | |
Ben | -0.007 |
ila | -0.007 |
bat | -0.007 |
Ben | -0.006 |
se | -0.006 |
Token | sail |
Token ID | 14936 |
Feature activation | -1.53e-01 |
Pos logit contributions | |
ila | +0.109 |
Ben | +0.097 |
Ben | +0.093 |
says | +0.087 |
dads | +0.083 |
Neg logit contributions | |
unlike | -0.150 |
matter | -0.141 |
superpower | -0.139 |
perspective | -0.136 |
former | -0.132 |
Token | and |
Token ID | 290 |
Feature activation | -9.95e-02 |
Pos logit contributions | |
ila | +0.163 |
Ben | +0.162 |
Ben | +0.158 |
se | +0.156 |
bat | +0.151 |
Neg logit contributions | |
unlike | -0.327 |
matter | -0.300 |
superpower | -0.295 |
perspective | -0.294 |
tactic | -0.289 |
Token | steer |
Token ID | 27401 |
Feature activation | -1.64e-01 |
Pos logit contributions | |
ila | +0.103 |
Ben | +0.094 |
Ben | +0.085 |
pige | +0.080 |
se | +0.080 |
Neg logit contributions | |
unlike | -0.223 |
matter | -0.211 |
superpower | -0.205 |
perspective | -0.196 |
former | -0.196 |
Token |
|
Token ID | 198 |
Feature activation | -3.92e-02 |
Pos logit contributions | |
se | +0.043 |
bat | +0.042 |
Ben | +0.041 |
ila | +0.040 |
goes | +0.040 |
Neg logit contributions | |
unlike | -0.082 |
superpower | -0.076 |
perspective | -0.075 |
eway | -0.075 |
tactic | -0.073 |
Token | The |
Token ID | 464 |
Feature activation | -4.37e-02 |
Pos logit contributions | |
Ben | +0.050 |
se | +0.048 |
bat | +0.048 |
ila | +0.047 |
says | +0.047 |
Neg logit contributions | |
unlike | -0.076 |
perspective | -0.069 |
superpower | -0.068 |
matter | -0.068 |
eway | -0.067 |
Token | wind |
Token ID | 2344 |
Feature activation | -8.01e-02 |
Pos logit contributions | |
Ben | +0.064 |
bat | +0.063 |
ila | +0.063 |
Ben | +0.063 |
se | +0.062 |
Neg logit contributions | |
unlike | -0.078 |
matter | -0.069 |
superpower | -0.069 |
perspective | -0.067 |
blonde | -0.067 |
Token | began |
Token ID | 2540 |
Feature activation | +8.39e-02 |
Pos logit contributions | |
Ben | +0.100 |
Ben | +0.098 |
ila | +0.096 |
bat | +0.095 |
se | +0.089 |
Neg logit contributions | |
unlike | -0.161 |
matter | -0.145 |
superpower | -0.143 |
perspective | -0.138 |
darkened | -0.137 |
Token | to |
Token ID | 284 |
Feature activation | +5.52e-02 |
Pos logit contributions | |
unlike | +0.188 |
matter | +0.172 |
superpower | +0.172 |
perspective | +0.168 |
tactic | +0.168 |
Neg logit contributions | |
Ben | -0.083 |
bat | -0.082 |
se | -0.080 |
ila | -0.079 |
Ben | -0.078 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.47e-01 |
Pos logit contributions | |
matter | +0.118 |
unlike | +0.118 |
superpower | +0.109 |
darkened | +0.107 |
perspective | +0.107 |
Neg logit contributions | |
ila | -0.061 |
Ben | -0.058 |
Ben | -0.057 |
pige | -0.053 |
dads | -0.051 |
Token | harder |
Token ID | 7069 |
Feature activation | +4.08e-02 |
Pos logit contributions | |
Ben | +0.192 |
bat | +0.186 |
se | +0.180 |
Ben | +0.175 |
ila | +0.167 |
Neg logit contributions | |
unlike | -0.612 |
superpower | -0.567 |
matter | -0.556 |
perspective | -0.555 |
tactic | -0.549 |
Token | and |
Token ID | 290 |
Feature activation | -9.75e-02 |
Pos logit contributions | |
unlike | +0.106 |
matter | +0.098 |
superpower | +0.097 |
perspective | +0.096 |
blonde | +0.095 |
Neg logit contributions | |
Ben | -0.026 |
ila | -0.026 |
Ben | -0.026 |
bat | -0.026 |
se | -0.024 |
Token | the |
Token ID | 262 |
Feature activation | -7.17e-02 |
Pos logit contributions | |
Ben | +0.126 |
bat | +0.124 |
Ben | +0.123 |
ila | +0.123 |
se | +0.122 |
Neg logit contributions | |
unlike | -0.191 |
superpower | -0.171 |
matter | -0.170 |
perspective | -0.166 |
blonde | -0.166 |
Token | child |
Token ID | 1200 |
Feature activation | +5.19e-02 |
Pos logit contributions | |
Ben | +0.096 |
ila | +0.096 |
Ben | +0.095 |
bat | +0.095 |
se | +0.092 |
Neg logit contributions | |
unlike | -0.135 |
matter | -0.122 |
superpower | -0.121 |
perspective | -0.118 |
blonde | -0.118 |
Token | laughed |
Token ID | 13818 |
Feature activation | -5.32e-02 |
Pos logit contributions | |
unlike | +0.104 |
matter | +0.094 |
superpower | +0.093 |
blonde | +0.091 |
perspective | +0.090 |
Neg logit contributions | |
ila | -0.062 |
Ben | -0.062 |
Ben | -0.062 |
bat | -0.061 |
se | -0.058 |
Token | sounds |
Token ID | 5238 |
Feature activation | +2.62e-02 |
Pos logit contributions | |
unlike | +0.062 |
superpower | +0.055 |
matter | +0.055 |
perspective | +0.054 |
blonde | +0.053 |
Neg logit contributions | |
Ben | -0.045 |
bat | -0.045 |
Ben | -0.044 |
ila | -0.044 |
se | -0.044 |
Token | . |
Token ID | 13 |
Feature activation | +8.04e-03 |
Pos logit contributions | |
unlike | +0.063 |
matter | +0.056 |
superpower | +0.056 |
blonde | +0.055 |
perspective | +0.055 |
Neg logit contributions | |
ila | -0.022 |
Ben | -0.022 |
Ben | -0.022 |
bat | -0.022 |
se | -0.021 |
Token | He |
Token ID | 679 |
Feature activation | +9.58e-02 |
Pos logit contributions | |
unlike | +0.014 |
blonde | +0.012 |
superpower | +0.012 |
matter | +0.012 |
Tears | +0.012 |
Neg logit contributions | |
se | -0.012 |
bat | -0.011 |
ila | -0.011 |
Ben | -0.011 |
next | -0.011 |
Token | likes |
Token ID | 7832 |
Feature activation | -1.64e-02 |
Pos logit contributions | |
unlike | +0.192 |
matter | +0.177 |
superpower | +0.177 |
blonde | +0.174 |
tactic | +0.160 |
Neg logit contributions | |
ila | -0.104 |
Ben | -0.104 |
line | -0.103 |
bat | -0.100 |
next | -0.097 |
Token | to |
Token ID | 284 |
Feature activation | +4.90e-02 |
Pos logit contributions | |
ila | +0.017 |
Ben | +0.017 |
bat | +0.017 |
Ben | +0.017 |
se | +0.016 |
Neg logit contributions | |
unlike | -0.035 |
matter | -0.032 |
superpower | -0.032 |
perspective | -0.031 |
blonde | -0.031 |
Token | blow |
Token ID | 6611 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
unlike | +0.105 |
matter | +0.102 |
superpower | +0.099 |
blonde | +0.097 |
tactic | +0.095 |
Neg logit contributions | |
Ben | -0.048 |
ila | -0.048 |
Ben | -0.047 |
pige | -0.044 |
next | -0.043 |
Token | the |
Token ID | 262 |
Feature activation | -8.29e-02 |
Pos logit contributions | |
ila | +0.238 |
Ben | +0.236 |
bat | +0.226 |
Ben | +0.225 |
se | +0.225 |
Neg logit contributions | |
unlike | -0.546 |
matter | -0.502 |
superpower | -0.495 |
tactic | -0.486 |
blonde | -0.482 |
Token | trumpet |
Token ID | 40334 |
Feature activation | +8.28e-02 |
Pos logit contributions | |
Ben | +0.135 |
bat | +0.135 |
se | +0.133 |
Ben | +0.132 |
ila | +0.132 |
Neg logit contributions | |
unlike | -0.135 |
matter | -0.117 |
superpower | -0.117 |
perspective | -0.114 |
blonde | -0.113 |
Token | in |
Token ID | 287 |
Feature activation | +8.76e-03 |
Pos logit contributions | |
unlike | +0.194 |
matter | +0.177 |
superpower | +0.176 |
tactic | +0.172 |
blonde | +0.170 |
Neg logit contributions | |
ila | -0.074 |
Ben | -0.073 |
Ben | -0.073 |
bat | -0.070 |
se | -0.069 |
Token | the |
Token ID | 262 |
Feature activation | -8.25e-02 |
Pos logit contributions | |
unlike | +0.020 |
matter | +0.018 |
superpower | +0.018 |
blonde | +0.018 |
perspective | +0.018 |
Neg logit contributions | |
ila | -0.008 |
Ben | -0.008 |
bat | -0.008 |
Ben | -0.008 |
se | -0.008 |
Token | morning |
Token ID | 3329 |
Feature activation | +6.03e-02 |
Pos logit contributions | |
ila | +0.097 |
Ben | +0.093 |
Ben | +0.092 |
se | +0.088 |
bat | +0.087 |
Neg logit contributions | |
unlike | -0.163 |
superpower | -0.153 |
matter | -0.152 |
blonde | -0.151 |
tactic | -0.145 |
Token | one |
Token ID | 530 |
Feature activation | +2.42e-02 |
Pos logit contributions | |
unlike | +0.024 |
Terr | +0.022 |
eway | +0.022 |
tactic | +0.021 |
superpower | +0.021 |
Neg logit contributions | |
Ben | -0.011 |
bat | -0.010 |
next | -0.010 |
se | -0.010 |
put | -0.010 |
Token | could |
Token ID | 714 |
Feature activation | +3.70e-02 |
Pos logit contributions | |
unlike | +0.050 |
matter | +0.046 |
superpower | +0.045 |
perspective | +0.045 |
blonde | +0.045 |
Neg logit contributions | |
Ben | -0.028 |
bat | -0.028 |
ila | -0.028 |
se | -0.028 |
Ben | -0.027 |
Token | open |
Token ID | 1280 |
Feature activation | -6.53e-02 |
Pos logit contributions | |
unlike | +0.090 |
matter | +0.083 |
superpower | +0.083 |
perspective | +0.081 |
blonde | +0.081 |
Neg logit contributions | |
bat | -0.030 |
Ben | -0.029 |
ila | -0.029 |
se | -0.029 |
Ben | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | +2.13e-04 |
Pos logit contributions | |
Ben | +0.046 |
bat | +0.041 |
next | +0.040 |
se | +0.039 |
put | +0.039 |
Neg logit contributions | |
unlike | -0.167 |
superpower | -0.156 |
tactic | -0.153 |
matter | -0.152 |
blonde | -0.152 |
Token | No |
Token ID | 1400 |
Feature activation | +5.97e-02 |
Pos logit contributions | |
unlike | +0.000 |
matter | +0.000 |
perspective | +0.000 |
superpower | +0.000 |
tactic | +0.000 |
Neg logit contributions | |
Ben | -0.000 |
bat | -0.000 |
Ben | -0.000 |
ila | -0.000 |
put | -0.000 |
Token | matter |
Token ID | 2300 |
Feature activation | +2.83e-01 |
Pos logit contributions | |
unlike | +0.134 |
Terr | +0.125 |
eway | +0.124 |
former | +0.121 |
Tears | +0.118 |
Neg logit contributions | |
put | -0.063 |
Ben | -0.063 |
next | -0.060 |
bat | -0.060 |
se | -0.059 |
Token | how |
Token ID | 703 |
Feature activation | +2.77e-02 |
Pos logit contributions | |
unlike | +0.595 |
superpower | +0.537 |
blonde | +0.535 |
matter | +0.523 |
tactic | +0.520 |
Neg logit contributions | |
Ben | -0.339 |
ila | -0.322 |
bat | -0.318 |
next | -0.304 |
each | -0.303 |
Token | hard |
Token ID | 1327 |
Feature activation | +1.99e-02 |
Pos logit contributions | |
unlike | +0.058 |
matter | +0.053 |
superpower | +0.053 |
perspective | +0.052 |
blonde | +0.051 |
Neg logit contributions | |
Ben | -0.031 |
Ben | -0.031 |
bat | -0.031 |
ila | -0.031 |
se | -0.030 |
Token | they |
Token ID | 484 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
unlike | +0.042 |
matter | +0.039 |
superpower | +0.038 |
blonde | +0.037 |
perspective | +0.036 |
Neg logit contributions | |
ila | -0.023 |
Ben | -0.023 |
se | -0.022 |
bat | -0.021 |
Ben | -0.021 |
Token | tried |
Token ID | 3088 |
Feature activation | +3.77e-03 |
Pos logit contributions | |
Ben | +0.154 |
bat | +0.152 |
ila | +0.151 |
Ben | +0.151 |
se | +0.150 |
Neg logit contributions | |
unlike | -0.241 |
superpower | -0.216 |
matter | -0.215 |
perspective | -0.210 |
blonde | -0.210 |
Token | they |
Token ID | 484 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
unlike | +0.009 |
superpower | +0.009 |
matter | +0.009 |
blonde | +0.009 |
tactic | +0.009 |
Neg logit contributions | |
Ben | -0.003 |
bat | -0.003 |
each | -0.002 |
next | -0.002 |
they | -0.002 |
Token | tree |
Token ID | 5509 |
Feature activation | -2.64e-02 |
Pos logit contributions | |
ila | +0.059 |
Ben | +0.054 |
Ben | +0.052 |
dads | +0.051 |
next | +0.049 |
Neg logit contributions | |
unlike | -0.080 |
matter | -0.074 |
perspective | -0.073 |
superpower | -0.072 |
blonde | -0.071 |
Token | . |
Token ID | 13 |
Feature activation | +1.34e-02 |
Pos logit contributions | |
Ben | +0.031 |
ila | +0.031 |
Ben | +0.030 |
dads | +0.029 |
se | +0.027 |
Neg logit contributions | |
unlike | -0.055 |
perspective | -0.050 |
matter | -0.049 |
tactic | -0.048 |
Tears | -0.047 |
Token |
|
Token ID | 198 |
Feature activation | -2.91e-02 |
Pos logit contributions | |
unlike | +0.026 |
matter | +0.024 |
superpower | +0.024 |
perspective | +0.023 |
blonde | +0.023 |
Neg logit contributions | |
Ben | -0.018 |
bat | -0.017 |
Ben | -0.017 |
ila | -0.017 |
se | -0.016 |
Token |
|
Token ID | 198 |
Feature activation | -2.87e-02 |
Pos logit contributions | |
se | +0.033 |
bat | +0.032 |
Ben | +0.031 |
ila | +0.031 |
Ben | +0.030 |
Neg logit contributions | |
unlike | -0.058 |
tactic | -0.053 |
eway | -0.053 |
perspective | -0.052 |
superpower | -0.052 |
Token | " |
Token ID | 1 |
Feature activation | -7.34e-02 |
Pos logit contributions | |
Ben | +0.047 |
se | +0.047 |
ila | +0.046 |
bat | +0.046 |
Ben | +0.045 |
Neg logit contributions | |
unlike | -0.046 |
matter | -0.040 |
superpower | -0.039 |
eway | -0.039 |
blonde | -0.038 |
Token | Wow |
Token ID | 22017 |
Feature activation | +1.80e-01 |
Pos logit contributions | |
Ben | +0.105 |
Ben | +0.103 |
bat | +0.102 |
se | +0.099 |
ila | +0.099 |
Neg logit contributions | |
unlike | -0.137 |
superpower | -0.124 |
matter | -0.121 |
tactic | -0.120 |
perspective | -0.119 |
Token | , |
Token ID | 11 |
Feature activation | +7.27e-03 |
Pos logit contributions | |
unlike | +0.390 |
matter | +0.353 |
superpower | +0.351 |
perspective | +0.345 |
blonde | +0.343 |
Neg logit contributions | |
bat | -0.187 |
se | -0.187 |
Ben | -0.185 |
ila | -0.184 |
Ben | -0.177 |
Token | this |
Token ID | 428 |
Feature activation | -6.79e-02 |
Pos logit contributions | |
unlike | +0.016 |
superpower | +0.015 |
perspective | +0.015 |
blonde | +0.014 |
matter | +0.014 |
Neg logit contributions | |
Ben | -0.008 |
bat | -0.008 |
ila | -0.007 |
Ben | -0.007 |
se | -0.007 |
Token | is |
Token ID | 318 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
Ben | +0.065 |
Ben | +0.065 |
ila | +0.064 |
bat | +0.064 |
se | +0.063 |
Neg logit contributions | |
unlike | -0.154 |
matter | -0.141 |
superpower | -0.141 |
perspective | -0.137 |
tactic | -0.136 |
Token | a |
Token ID | 257 |
Feature activation | -3.53e-03 |
Pos logit contributions | |
Ben | +0.017 |
bat | +0.015 |
next | +0.014 |
put | +0.014 |
we | +0.014 |
Neg logit contributions | |
unlike | -0.034 |
perspective | -0.034 |
superpower | -0.033 |
tactic | -0.033 |
blonde | -0.033 |
Token | nice |
Token ID | 3621 |
Feature activation | +1.98e-02 |
Pos logit contributions | |
bat | +0.004 |
se | +0.004 |
Ben | +0.004 |
Ben | +0.004 |
slide | +0.004 |
Neg logit contributions | |
unlike | -0.008 |
tactic | -0.007 |
blonde | -0.007 |
eway | -0.007 |
superpower | -0.007 |
Token | him |
Token ID | 683 |
Feature activation | +4.24e-02 |
Pos logit contributions | |
Ben | +0.023 |
bat | +0.022 |
Ben | +0.022 |
ila | +0.022 |
se | +0.022 |
Neg logit contributions | |
unlike | -0.033 |
superpower | -0.030 |
matter | -0.029 |
blonde | -0.029 |
perspective | -0.029 |
Token | out |
Token ID | 503 |
Feature activation | +1.03e-01 |
Pos logit contributions | |
unlike | +0.100 |
matter | +0.092 |
superpower | +0.092 |
perspective | +0.089 |
blonde | +0.089 |
Neg logit contributions | |
Ben | -0.037 |
Ben | -0.036 |
ila | -0.036 |
bat | -0.036 |
se | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | +1.08e-02 |
Pos logit contributions | |
unlike | +0.253 |
matter | +0.232 |
superpower | +0.232 |
blonde | +0.227 |
perspective | +0.227 |
Neg logit contributions | |
Ben | -0.081 |
bat | -0.079 |
ila | -0.077 |
se | -0.077 |
Ben | -0.076 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
unlike | +0.023 |
superpower | +0.021 |
perspective | +0.020 |
matter | +0.020 |
Tears | +0.020 |
Neg logit contributions | |
se | -0.012 |
bat | -0.012 |
ila | -0.012 |
Ben | -0.011 |
Ben | -0.011 |
Token | Once |
Token ID | 7454 |
Feature activation | +9.64e-02 |
Pos logit contributions | |
se | +0.012 |
bat | +0.012 |
ila | +0.012 |
Ben | +0.012 |
Ben | +0.012 |
Neg logit contributions | |
unlike | -0.024 |
superpower | -0.022 |
eway | -0.021 |
perspective | -0.021 |
matter | -0.021 |
Token | upon |
Token ID | 2402 |
Feature activation | +1.92e-01 |
Pos logit contributions | |
unlike | +0.111 |
matter | +0.091 |
superpower | +0.091 |
perspective | +0.086 |
blonde | +0.086 |
Neg logit contributions | |
Ben | -0.201 |
bat | -0.199 |
ila | -0.199 |
Ben | -0.199 |
se | -0.197 |
Token | a |
Token ID | 257 |
Feature activation | -7.63e-03 |
Pos logit contributions | |
unlike | +0.441 |
matter | +0.403 |
superpower | +0.401 |
blonde | +0.393 |
perspective | +0.392 |
Neg logit contributions | |
Ben | -0.176 |
bat | -0.173 |
ila | -0.173 |
Ben | -0.172 |
se | -0.169 |
Token | time |
Token ID | 640 |
Feature activation | +7.29e-02 |
Pos logit contributions | |
se | +0.014 |
ila | +0.014 |
Ben | +0.014 |
bat | +0.014 |
line | +0.013 |
Neg logit contributions | |
unlike | -0.010 |
matter | -0.009 |
blonde | -0.009 |
superpower | -0.009 |
perspective | -0.009 |
Token | there |
Token ID | 612 |
Feature activation | +6.02e-02 |
Pos logit contributions | |
unlike | +0.177 |
matter | +0.164 |
superpower | +0.163 |
perspective | +0.160 |
blonde | +0.159 |
Neg logit contributions | |
Ben | -0.061 |
ila | -0.058 |
bat | -0.058 |
Ben | -0.057 |
se | -0.055 |
Token | was |
Token ID | 373 |
Feature activation | +7.45e-02 |
Pos logit contributions | |
unlike | +0.135 |
superpower | +0.126 |
former | +0.125 |
matter | +0.123 |
perspective | +0.122 |
Neg logit contributions | |
Ben | -0.066 |
ila | -0.062 |
Ben | -0.061 |
bat | -0.061 |
se | -0.058 |
Token | a |
Token ID | 257 |
Feature activation | -6.68e-03 |
Pos logit contributions | |
unlike | +0.164 |
matter | +0.147 |
blonde | +0.146 |
superpower | +0.146 |
perspective | +0.141 |
Neg logit contributions | |
ila | -0.076 |
bat | -0.074 |
Ben | -0.074 |
se | -0.074 |
Ben | -0.073 |
Token | friends |
Token ID | 2460 |
Feature activation | +2.77e-02 |
Pos logit contributions | |
unlike | +0.170 |
superpower | +0.153 |
matter | +0.152 |
perspective | +0.149 |
blonde | +0.148 |
Neg logit contributions | |
Ben | -0.093 |
Ben | -0.092 |
ila | -0.092 |
bat | -0.092 |
se | -0.090 |
Token | forever |
Token ID | 8097 |
Feature activation | +1.67e-01 |
Pos logit contributions | |
unlike | +0.064 |
superpower | +0.059 |
matter | +0.059 |
perspective | +0.057 |
blonde | +0.057 |
Neg logit contributions | |
Ben | -0.025 |
bat | -0.025 |
ila | -0.025 |
Ben | -0.025 |
se | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | +2.32e-02 |
Pos logit contributions | |
unlike | +0.357 |
perspective | +0.325 |
superpower | +0.318 |
matter | +0.318 |
tactic | +0.303 |
Neg logit contributions | |
ila | -0.194 |
Ben | -0.190 |
Ben | -0.181 |
bat | -0.171 |
se | -0.170 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +5.10e-03 |
Pos logit contributions | |
unlike | +0.046 |
Tears | +0.041 |
superpower | +0.041 |
perspective | +0.040 |
matter | +0.039 |
Neg logit contributions | |
se | -0.029 |
ila | -0.028 |
bat | -0.027 |
Ben | -0.027 |
dove | -0.026 |
Token | Once |
Token ID | 7454 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
unlike | +0.010 |
eway | +0.009 |
Tears | +0.009 |
superpower | +0.009 |
tactic | +0.009 |
Neg logit contributions | |
se | -0.006 |
dads | -0.006 |
ila | -0.006 |
bat | -0.006 |
Ben | -0.006 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.08e-01 |
Pos logit contributions | |
unlike | +0.127 |
matter | +0.105 |
superpower | +0.104 |
perspective | +0.101 |
tactic | +0.099 |
Neg logit contributions | |
bat | -0.233 |
Ben | -0.231 |
Ben | -0.231 |
se | -0.230 |
ila | -0.229 |
Token | a |
Token ID | 257 |
Feature activation | +2.33e-03 |
Pos logit contributions | |
unlike | +0.453 |
matter | +0.413 |
superpower | +0.408 |
blonde | +0.404 |
tactic | +0.398 |
Neg logit contributions | |
ila | -0.211 |
Ben | -0.206 |
Ben | -0.203 |
se | -0.203 |
bat | -0.202 |
Token | time |
Token ID | 640 |
Feature activation | +8.68e-02 |
Pos logit contributions | |
unlike | +0.003 |
matter | +0.003 |
blonde | +0.003 |
superpower | +0.003 |
perspective | +0.002 |
Neg logit contributions | |
se | -0.004 |
ila | -0.004 |
Ben | -0.004 |
line | -0.004 |
bat | -0.004 |
Token | , |
Token ID | 11 |
Feature activation | +1.36e-02 |
Pos logit contributions | |
unlike | +0.197 |
superpower | +0.179 |
matter | +0.178 |
blonde | +0.175 |
perspective | +0.174 |
Neg logit contributions | |
bat | -0.082 |
Ben | -0.081 |
Ben | -0.081 |
se | -0.081 |
ila | -0.081 |
Token | there |
Token ID | 612 |
Feature activation | +7.23e-02 |
Pos logit contributions | |
unlike | +0.033 |
superpower | +0.031 |
perspective | +0.030 |
former | +0.030 |
matter | +0.030 |
Neg logit contributions | |
Ben | -0.012 |
Ben | -0.011 |
bat | -0.011 |
ila | -0.011 |
next | -0.011 |
Token | was |
Token ID | 373 |
Feature activation | +8.89e-02 |
Pos logit contributions | |
unlike | +0.157 |
superpower | +0.144 |
matter | +0.141 |
former | +0.141 |
blonde | +0.140 |
Neg logit contributions | |
Ben | -0.084 |
bat | -0.079 |
Ben | -0.078 |
ila | -0.077 |
se | -0.075 |
Token | cake |
Token ID | 12187 |
Feature activation | +1.32e-02 |
Pos logit contributions | |
se | +0.069 |
Ben | +0.067 |
bat | +0.065 |
slide | +0.064 |
next | +0.063 |
Neg logit contributions | |
unlike | -0.150 |
eway | -0.139 |
superpower | -0.134 |
perspective | -0.133 |
matter | -0.133 |
Token | . |
Token ID | 13 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
unlike | +0.029 |
matter | +0.026 |
superpower | +0.026 |
blonde | +0.025 |
perspective | +0.025 |
Neg logit contributions | |
Ben | -0.013 |
ila | -0.013 |
bat | -0.013 |
Ben | -0.012 |
se | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | -3.09e-02 |
Pos logit contributions | |
unlike | +0.025 |
matter | +0.023 |
superpower | +0.023 |
perspective | +0.022 |
tactic | +0.022 |
Neg logit contributions | |
Ben | -0.010 |
Ben | -0.010 |
bat | -0.010 |
ila | -0.010 |
se | -0.009 |
Token |
|
Token ID | 198 |
Feature activation | -3.07e-02 |
Pos logit contributions | |
se | +0.046 |
wave | +0.044 |
dads | +0.043 |
pige | +0.043 |
bat | +0.043 |
Neg logit contributions | |
unlike | -0.045 |
eway | -0.044 |
Tears | -0.041 |
Terr | -0.040 |
Where | -0.040 |
Token | " |
Token ID | 1 |
Feature activation | -7.87e-02 |
Pos logit contributions | |
se | +0.040 |
wave | +0.039 |
waves | +0.039 |
dove | +0.038 |
pige | +0.038 |
Neg logit contributions | |
unlike | -0.045 |
Andy | -0.044 |
Fear | -0.043 |
total | -0.043 |
slightly | -0.043 |
Token | Oh |
Token ID | 5812 |
Feature activation | +2.03e-01 |
Pos logit contributions | |
Ben | +0.144 |
bat | +0.143 |
ila | +0.140 |
goes | +0.135 |
says | +0.134 |
Neg logit contributions | |
unlike | -0.102 |
matter | -0.095 |
blonde | -0.089 |
superpower | -0.087 |
perspective | -0.083 |
Token | no |
Token ID | 645 |
Feature activation | +2.57e-02 |
Pos logit contributions | |
unlike | +0.483 |
superpower | +0.440 |
perspective | +0.436 |
tactic | +0.434 |
matter | +0.433 |
Neg logit contributions | |
Ben | -0.208 |
ila | -0.188 |
bat | -0.182 |
Ben | -0.180 |
se | -0.178 |
Token | !" |
Token ID | 2474 |
Feature activation | -6.65e-02 |
Pos logit contributions | |
unlike | +0.061 |
superpower | +0.056 |
matter | +0.055 |
blonde | +0.054 |
tactic | +0.054 |
Neg logit contributions | |
Ben | -0.023 |
bat | -0.022 |
ila | -0.022 |
Ben | -0.022 |
se | -0.022 |
Token | Lily |
Token ID | 20037 |
Feature activation | -4.54e-02 |
Pos logit contributions | |
bat | +0.046 |
se | +0.045 |
Ben | +0.045 |
ila | +0.044 |
slide | +0.044 |
Neg logit contributions | |
unlike | -0.147 |
blonde | -0.139 |
matter | -0.136 |
Tears | -0.135 |
Nobody | -0.131 |
Token | and |
Token ID | 290 |
Feature activation | -9.21e-02 |
Pos logit contributions | |
ila | +0.035 |
Ben | +0.035 |
se | +0.034 |
bat | +0.033 |
line | +0.032 |
Neg logit contributions | |
unlike | -0.101 |
tactic | -0.097 |
Tears | -0.097 |
Terr | -0.094 |
blonde | -0.092 |
Token | Ben |
Token ID | 3932 |
Feature activation | -2.07e-01 |
Pos logit contributions | |
ars | +0.100 |
ites | +0.096 |
is | +0.089 |
aks | +0.086 |
pige | +0.085 |
Neg logit contributions | |
blonde | -0.172 |
Tears | -0.163 |
avail | -0.162 |
matter | -0.154 |
disgust | -0.150 |
Token | ." |
Token ID | 526 |
Feature activation | -5.19e-02 |
Pos logit contributions | |
Ben | +0.024 |
bat | +0.023 |
Ben | +0.023 |
ila | +0.023 |
se | +0.023 |
Neg logit contributions | |
unlike | -0.053 |
matter | -0.049 |
superpower | -0.048 |
perspective | -0.047 |
blonde | -0.046 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +6.24e-03 |
Pos logit contributions | |
Ben | +0.044 |
Ben | +0.036 |
bat | +0.035 |
says | +0.032 |
they | +0.031 |
Neg logit contributions | |
unlike | -0.137 |
matter | -0.130 |
perspective | -0.126 |
superpower | -0.125 |
eway | -0.122 |
Token | Once |
Token ID | 7454 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
unlike | +0.012 |
superpower | +0.011 |
matter | +0.011 |
blonde | +0.011 |
perspective | +0.011 |
Neg logit contributions | |
bat | -0.008 |
se | -0.008 |
Ben | -0.008 |
ila | -0.008 |
Ben | -0.008 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.07e-01 |
Pos logit contributions | |
unlike | +0.148 |
superpower | +0.125 |
matter | +0.125 |
blonde | +0.119 |
perspective | +0.119 |
Neg logit contributions | |
Ben | -0.220 |
ila | -0.217 |
bat | -0.215 |
Ben | -0.215 |
se | -0.212 |
Token | a |
Token ID | 257 |
Feature activation | +2.04e-03 |
Pos logit contributions | |
unlike | +0.489 |
superpower | +0.447 |
matter | +0.447 |
perspective | +0.438 |
blonde | +0.435 |
Neg logit contributions | |
Ben | -0.184 |
bat | -0.180 |
Ben | -0.178 |
ila | -0.177 |
se | -0.174 |
Token | time |
Token ID | 640 |
Feature activation | +8.58e-02 |
Pos logit contributions | |
unlike | +0.003 |
matter | +0.003 |
superpower | +0.003 |
blonde | +0.003 |
perspective | +0.003 |
Neg logit contributions | |
Ben | -0.003 |
ila | -0.003 |
se | -0.003 |
bat | -0.003 |
Ben | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | +1.40e-02 |
Pos logit contributions | |
unlike | +0.208 |
matter | +0.192 |
superpower | +0.191 |
perspective | +0.187 |
blonde | +0.186 |
Neg logit contributions | |
Ben | -0.073 |
ila | -0.069 |
bat | -0.068 |
Ben | -0.068 |
se | -0.066 |
Token | there |
Token ID | 612 |
Feature activation | +7.29e-02 |
Pos logit contributions | |
unlike | +0.035 |
former | +0.034 |
superpower | +0.034 |
perspective | +0.033 |
matter | +0.033 |
Neg logit contributions | |
Ben | -0.012 |
ila | -0.010 |
Ben | -0.010 |
next | -0.010 |
they | -0.009 |
Token | was |
Token ID | 373 |
Feature activation | +8.77e-02 |
Pos logit contributions | |
unlike | +0.171 |
former | +0.163 |
superpower | +0.161 |
perspective | +0.157 |
matter | +0.155 |
Neg logit contributions | |
Ben | -0.075 |
Ben | -0.067 |
bat | -0.066 |
ila | -0.065 |
put | -0.064 |
Token | a |
Token ID | 257 |
Feature activation | +2.75e-03 |
Pos logit contributions | |
unlike | +0.199 |
matter | +0.180 |
superpower | +0.179 |
blonde | +0.177 |
perspective | +0.175 |
Neg logit contributions | |
ila | -0.083 |
Ben | -0.083 |
bat | -0.083 |
Ben | -0.082 |
se | -0.081 |
Token | small |
Token ID | 1402 |
Feature activation | +3.80e-02 |
Pos logit contributions | |
unlike | +0.004 |
matter | +0.004 |
superpower | +0.004 |
blonde | +0.004 |
perspective | +0.004 |
Neg logit contributions | |
ila | -0.004 |
Ben | -0.004 |
goes | -0.004 |
says | -0.004 |
irts | -0.004 |
Token | old |
Token ID | 1468 |
Feature activation | -5.37e-02 |
Pos logit contributions | |
unlike | +0.116 |
blonde | +0.104 |
superpower | +0.103 |
perspective | +0.100 |
former | +0.097 |
Neg logit contributions | |
Ben | -0.103 |
bat | -0.094 |
ila | -0.092 |
Ben | -0.089 |
throw | -0.087 |
Token | . |
Token ID | 13 |
Feature activation | +1.19e-02 |
Pos logit contributions | |
Ben | +0.019 |
and | +0.019 |
each | +0.017 |
next | +0.013 |
bat | +0.012 |
Neg logit contributions | |
unlike | -0.164 |
eway | -0.159 |
superpower | -0.153 |
matter | -0.149 |
former | -0.149 |
Token | |
Token ID | 220 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
unlike | +0.024 |
superpower | +0.022 |
matter | +0.021 |
perspective | +0.021 |
tactic | +0.021 |
Neg logit contributions | |
Ben | -0.015 |
Ben | -0.014 |
bat | -0.014 |
ila | -0.014 |
next | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -2.84e-02 |
Pos logit contributions | |
unlike | +0.291 |
matter | +0.268 |
superpower | +0.268 |
perspective | +0.262 |
blonde | +0.261 |
Neg logit contributions | |
Ben | -0.094 |
ila | -0.090 |
bat | -0.090 |
Ben | -0.090 |
se | -0.086 |
Token | One |
Token ID | 3198 |
Feature activation | +4.87e-02 |
Pos logit contributions | |
Ben | +0.023 |
bat | +0.022 |
Ben | +0.022 |
ila | +0.022 |
se | +0.022 |
Neg logit contributions | |
unlike | -0.069 |
matter | -0.063 |
superpower | -0.063 |
perspective | -0.062 |
blonde | -0.062 |
Token | day |
Token ID | 1110 |
Feature activation | +2.79e-02 |
Pos logit contributions | |
unlike | +0.087 |
matter | +0.081 |
superpower | +0.076 |
blonde | +0.073 |
tactic | +0.073 |
Neg logit contributions | |
Ben | -0.067 |
ila | -0.066 |
se | -0.066 |
Ben | -0.066 |
bat | -0.063 |
Token | , |
Token ID | 11 |
Feature activation | +7.17e-03 |
Pos logit contributions | |
unlike | +0.079 |
former | +0.077 |
superpower | +0.074 |
perspective | +0.074 |
tactic | +0.073 |
Neg logit contributions | |
Ben | -0.016 |
they | -0.013 |
Ben | -0.011 |
bat | -0.011 |
next | -0.011 |
Token | Sarah |
Token ID | 10490 |
Feature activation | +1.03e-01 |
Pos logit contributions | |
unlike | +0.014 |
former | +0.014 |
superpower | +0.014 |
perspective | +0.013 |
eway | +0.013 |
Neg logit contributions | |
Ben | -0.010 |
they | -0.010 |
next | -0.009 |
put | -0.009 |
Tom | -0.009 |
Token | went |
Token ID | 1816 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
unlike | +0.184 |
blonde | +0.167 |
matter | +0.164 |
shouldn | +0.157 |
darkened | +0.156 |
Neg logit contributions | |
dads | -0.134 |
ila | -0.130 |
line | -0.126 |
bat | -0.117 |
Ben | -0.116 |
Token | to |
Token ID | 284 |
Feature activation | +6.84e-02 |
Pos logit contributions | |
Ben | +0.188 |
ila | +0.184 |
Ben | +0.170 |
se | +0.166 |
pige | +0.165 |
Neg logit contributions | |
unlike | -0.210 |
superpower | -0.192 |
tactic | -0.175 |
eway | -0.173 |
matter | -0.172 |
Token | the |
Token ID | 262 |
Feature activation | -6.38e-02 |
Pos logit contributions | |
unlike | +0.159 |
matter | +0.155 |
superpower | +0.150 |
blonde | +0.148 |
tactic | +0.147 |
Neg logit contributions | |
ila | -0.055 |
Ben | -0.052 |
se | -0.049 |
bat | -0.047 |
line | -0.045 |
Token | , |
Token ID | 11 |
Feature activation | +8.04e-05 |
Pos logit contributions | |
unlike | +0.164 |
matter | +0.151 |
superpower | +0.150 |
perspective | +0.146 |
blonde | +0.146 |
Neg logit contributions | |
Ben | -0.062 |
bat | -0.062 |
ila | -0.062 |
Ben | -0.062 |
se | -0.061 |
Token | Leo |
Token ID | 19632 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
unlike | +0.000 |
matter | +0.000 |
superpower | +0.000 |
tactic | +0.000 |
blonde | +0.000 |
Neg logit contributions | |
Ben | -0.000 |
ila | -0.000 |
bat | -0.000 |
se | -0.000 |
Ben | -0.000 |
Token | ." |
Token ID | 526 |
Feature activation | -6.39e-02 |
Pos logit contributions | |
Ben | +0.132 |
ila | +0.130 |
bat | +0.129 |
Ben | +0.129 |
se | +0.128 |
Neg logit contributions | |
unlike | -0.316 |
matter | -0.291 |
superpower | -0.289 |
perspective | -0.278 |
blonde | -0.277 |
Token |
|
Token ID | 198 |
Feature activation | -3.32e-02 |
Pos logit contributions | |
bat | +0.060 |
se | +0.060 |
ila | +0.059 |
Ben | +0.058 |
Ben | +0.056 |
Neg logit contributions | |
unlike | -0.144 |
superpower | -0.130 |
tactic | -0.128 |
blonde | -0.127 |
perspective | -0.127 |
Token |
|
Token ID | 198 |
Feature activation | -3.29e-02 |
Pos logit contributions | |
se | +0.043 |
bat | +0.042 |
wave | +0.039 |
goes | +0.039 |
put | +0.039 |
Neg logit contributions | |
unlike | -0.060 |
eway | -0.056 |
Tears | -0.055 |
superpower | -0.053 |
Terr | -0.053 |
Token | But |
Token ID | 1537 |
Feature activation | +7.72e-02 |
Pos logit contributions | |
se | +0.035 |
bat | +0.034 |
ila | +0.032 |
Ben | +0.032 |
put | +0.031 |
Neg logit contributions | |
unlike | -0.069 |
superpower | -0.063 |
eway | -0.063 |
perspective | -0.062 |
tactic | -0.061 |
Token | Leo |
Token ID | 19632 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
unlike | +0.196 |
superpower | +0.182 |
matter | +0.182 |
perspective | +0.180 |
blonde | +0.179 |
Neg logit contributions | |
Ben | -0.060 |
ila | -0.054 |
bat | -0.053 |
Ben | -0.053 |
se | -0.050 |
Token | was |
Token ID | 373 |
Feature activation | +7.77e-02 |
Pos logit contributions | |
bat | +0.175 |
Ben | +0.175 |
ila | +0.175 |
Ben | +0.173 |
se | +0.172 |
Neg logit contributions | |
unlike | -0.272 |
matter | -0.244 |
superpower | -0.243 |
blonde | -0.236 |
tactic | -0.232 |
Token | angry |
Token ID | 7954 |
Feature activation | +4.07e-02 |
Pos logit contributions | |
unlike | +0.169 |
matter | +0.153 |
superpower | +0.152 |
blonde | +0.149 |
perspective | +0.148 |
Neg logit contributions | |
Ben | -0.084 |
bat | -0.083 |
Ben | -0.082 |
ila | -0.082 |
se | -0.081 |
Token | . |
Token ID | 13 |
Feature activation | +9.96e-03 |
Pos logit contributions | |
unlike | +0.092 |
matter | +0.083 |
superpower | +0.081 |
tactic | +0.080 |
perspective | +0.080 |
Neg logit contributions | |
bat | -0.039 |
Ben | -0.039 |
se | -0.038 |
ila | -0.038 |
Ben | -0.036 |
Token | He |
Token ID | 679 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
matter | +0.018 |
unlike | +0.018 |
perspective | +0.017 |
superpower | +0.016 |
former | +0.015 |
Neg logit contributions | |
Ben | -0.017 |
ila | -0.015 |
Ben | -0.015 |
says | -0.014 |
bat | -0.014 |
Token | who |
Token ID | 508 |
Feature activation | -7.95e-03 |
Pos logit contributions | |
ila | +0.010 |
Ben | +0.009 |
Ben | +0.009 |
bat | +0.009 |
se | +0.009 |
Neg logit contributions | |
unlike | -0.021 |
superpower | -0.019 |
perspective | -0.018 |
matter | -0.018 |
blonde | -0.018 |
Token | liked |
Token ID | 8288 |
Feature activation | +5.06e-02 |
Pos logit contributions | |
Ben | +0.011 |
ila | +0.011 |
bat | +0.011 |
Ben | +0.011 |
se | +0.011 |
Neg logit contributions | |
unlike | -0.015 |
matter | -0.013 |
superpower | -0.013 |
blonde | -0.013 |
tactic | -0.012 |
Token | to |
Token ID | 284 |
Feature activation | +7.51e-02 |
Pos logit contributions | |
unlike | +0.110 |
matter | +0.099 |
superpower | +0.099 |
perspective | +0.097 |
blonde | +0.097 |
Neg logit contributions | |
Ben | -0.054 |
bat | -0.053 |
ila | -0.053 |
Ben | -0.053 |
se | -0.052 |
Token | play |
Token ID | 711 |
Feature activation | -9.03e-02 |
Pos logit contributions | |
unlike | +0.184 |
matter | +0.171 |
superpower | +0.171 |
blonde | +0.168 |
tactic | +0.165 |
Neg logit contributions | |
ila | -0.057 |
Ben | -0.056 |
Ben | -0.056 |
bat | -0.053 |
se | -0.052 |
Token | with |
Token ID | 351 |
Feature activation | -9.81e-03 |
Pos logit contributions | |
Ben | +0.087 |
bat | +0.085 |
Ben | +0.084 |
ila | +0.084 |
se | +0.083 |
Neg logit contributions | |
unlike | -0.207 |
matter | -0.189 |
superpower | -0.189 |
perspective | -0.184 |
blonde | -0.184 |
Token | fire |
Token ID | 2046 |
Feature activation | +9.37e-02 |
Pos logit contributions | |
Ben | +0.007 |
bat | +0.006 |
se | +0.006 |
each | +0.006 |
Ben | +0.006 |
Neg logit contributions | |
unlike | -0.027 |
matter | -0.025 |
superpower | -0.025 |
perspective | -0.024 |
eway | -0.024 |
Token | work |
Token ID | 1818 |
Feature activation | +6.80e-02 |
Pos logit contributions | |
unlike | +0.237 |
matter | +0.220 |
blonde | +0.219 |
superpower | +0.217 |
tactic | +0.213 |
Neg logit contributions | |
Ben | -0.070 |
bat | -0.068 |
se | -0.065 |
next | -0.063 |
each | -0.061 |
Token | . |
Token ID | 13 |
Feature activation | +2.20e-02 |
Pos logit contributions | |
unlike | +0.164 |
matter | +0.150 |
superpower | +0.150 |
perspective | +0.146 |
blonde | +0.146 |
Neg logit contributions | |
Ben | -0.056 |
ila | -0.055 |
Ben | -0.055 |
bat | -0.055 |
se | -0.054 |
Token | They |
Token ID | 1119 |
Feature activation | -2.90e-02 |
Pos logit contributions | |
unlike | +0.031 |
matter | +0.027 |
superpower | +0.027 |
blonde | +0.026 |
tactic | +0.026 |
Neg logit contributions | |
ila | -0.039 |
se | -0.039 |
bat | -0.039 |
Ben | -0.039 |
Ben | -0.038 |
Token | found |
Token ID | 1043 |
Feature activation | -1.12e-02 |
Pos logit contributions | |
Ben | +0.032 |
ila | +0.032 |
Ben | +0.032 |
bat | +0.031 |
se | +0.030 |
Neg logit contributions | |
unlike | -0.061 |
matter | -0.056 |
superpower | -0.055 |
blonde | -0.054 |
perspective | -0.053 |
Token | some |
Token ID | 617 |
Feature activation | -1.64e-02 |
Pos logit contributions | |
Ben | +0.009 |
se | +0.009 |
bat | +0.009 |
ila | +0.008 |
each | +0.008 |
Neg logit contributions | |
unlike | -0.027 |
matter | -0.026 |
superpower | -0.026 |
perspective | -0.025 |
blonde | -0.025 |
Token | other |
Token ID | 584 |
Feature activation | +4.67e-02 |
Pos logit contributions | |
Ben | +0.163 |
bat | +0.161 |
ila | +0.159 |
Ben | +0.159 |
se | +0.157 |
Neg logit contributions | |
unlike | -0.452 |
matter | -0.417 |
superpower | -0.414 |
perspective | -0.406 |
blonde | -0.404 |
Token | . |
Token ID | 13 |
Feature activation | +2.62e-02 |
Pos logit contributions | |
unlike | +0.102 |
superpower | +0.094 |
matter | +0.093 |
tactic | +0.092 |
perspective | +0.092 |
Neg logit contributions | |
ila | -0.050 |
Ben | -0.047 |
se | -0.046 |
Ben | -0.044 |
bat | -0.044 |
Token | They |
Token ID | 1119 |
Feature activation | -3.21e-02 |
Pos logit contributions | |
unlike | +0.061 |
matter | +0.057 |
superpower | +0.057 |
tactic | +0.055 |
perspective | +0.055 |
Neg logit contributions | |
Ben | -0.025 |
ila | -0.024 |
Ben | -0.023 |
bat | -0.023 |
se | -0.022 |
Token | love |
Token ID | 1842 |
Feature activation | -1.60e-02 |
Pos logit contributions | |
ila | +0.033 |
Ben | +0.031 |
se | +0.030 |
Ben | +0.030 |
pige | +0.030 |
Neg logit contributions | |
matter | -0.065 |
unlike | -0.063 |
superpower | -0.061 |
blonde | -0.060 |
perspective | -0.059 |
Token | their |
Token ID | 511 |
Feature activation | -4.93e-03 |
Pos logit contributions | |
ila | +0.013 |
Ben | +0.013 |
bat | +0.012 |
se | +0.012 |
Ben | +0.011 |
Neg logit contributions | |
unlike | -0.039 |
matter | -0.036 |
superpower | -0.035 |
perspective | -0.035 |
blonde | -0.034 |
Token | globe |
Token ID | 13342 |
Feature activation | +4.77e-02 |
Pos logit contributions | |
Ben | +0.006 |
bat | +0.006 |
ila | +0.006 |
Ben | +0.006 |
se | +0.006 |
Neg logit contributions | |
unlike | -0.010 |
matter | -0.009 |
superpower | -0.009 |
blonde | -0.009 |
perspective | -0.009 |
Token | and |
Token ID | 290 |
Feature activation | -8.20e-02 |
Pos logit contributions | |
unlike | +0.114 |
matter | +0.102 |
superpower | +0.102 |
perspective | +0.100 |
tactic | +0.099 |
Neg logit contributions | |
Ben | -0.043 |
ila | -0.042 |
Ben | -0.040 |
bat | -0.039 |
se | -0.038 |
Token | their |
Token ID | 511 |
Feature activation | -5.21e-03 |
Pos logit contributions | |
ila | +0.087 |
Ben | +0.086 |
se | +0.080 |
Ben | +0.079 |
bat | +0.078 |
Neg logit contributions | |
unlike | -0.179 |
matter | -0.165 |
superpower | -0.164 |
blonde | -0.159 |
perspective | -0.158 |
Token | planet |
Token ID | 5440 |
Feature activation | +7.95e-05 |
Pos logit contributions | |
Ben | +0.006 |
bat | +0.006 |
Ben | +0.006 |
ila | +0.006 |
se | +0.006 |
Neg logit contributions | |
unlike | -0.010 |
matter | -0.009 |
superpower | -0.009 |
blonde | -0.009 |
perspective | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | +2.64e-02 |
Pos logit contributions | |
unlike | +0.000 |
superpower | +0.000 |
matter | +0.000 |
perspective | +0.000 |
tactic | +0.000 |
Neg logit contributions | |
Ben | -0.000 |
ila | -0.000 |
Ben | -0.000 |
bat | -0.000 |
se | -0.000 |
Token | They |
Token ID | 1119 |
Feature activation | -3.17e-02 |
Pos logit contributions | |
unlike | +0.062 |
matter | +0.057 |
superpower | +0.056 |
perspective | +0.055 |
tactic | +0.055 |
Neg logit contributions | |
Ben | -0.025 |
bat | -0.024 |
Ben | -0.024 |
ila | -0.024 |
se | -0.023 |
Token | They |
Token ID | 1119 |
Feature activation | -5.51e-02 |
Pos logit contributions | |
unlike | +0.016 |
superpower | +0.014 |
matter | +0.014 |
blonde | +0.014 |
perspective | +0.013 |
Neg logit contributions | |
se | -0.013 |
ila | -0.012 |
bat | -0.012 |
Ben | -0.012 |
Ben | -0.012 |
Token | must |
Token ID | 1276 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
ila | +0.060 |
Ben | +0.060 |
Ben | +0.056 |
se | +0.052 |
bat | +0.051 |
Neg logit contributions | |
unlike | -0.112 |
matter | -0.109 |
superpower | -0.104 |
blonde | -0.100 |
former | -0.099 |
Token | be |
Token ID | 307 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
matter | +0.312 |
unlike | +0.301 |
former | +0.298 |
superpower | +0.288 |
eway | +0.277 |
Neg logit contributions | |
Ben | -0.161 |
se | -0.159 |
ila | -0.158 |
Ben | -0.154 |
pige | -0.153 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.70e-02 |
Pos logit contributions | |
Ben | +0.152 |
bat | +0.150 |
ila | +0.148 |
Ben | +0.148 |
se | +0.147 |
Neg logit contributions | |
unlike | -0.255 |
superpower | -0.230 |
matter | -0.230 |
perspective | -0.225 |
blonde | -0.223 |
Token | ." |
Token ID | 526 |
Feature activation | -6.44e-02 |
Pos logit contributions | |
unlike | +0.040 |
superpower | +0.035 |
matter | +0.035 |
perspective | +0.035 |
tactic | +0.034 |
Neg logit contributions | |
Ben | -0.016 |
ila | -0.016 |
bat | -0.015 |
se | -0.015 |
Ben | -0.015 |
Token |
|
Token ID | 198 |
Feature activation | -3.84e-02 |
Pos logit contributions | |
bat | +0.062 |
se | +0.062 |
ila | +0.060 |
Ben | +0.059 |
Ben | +0.058 |
Neg logit contributions | |
unlike | -0.143 |
superpower | -0.129 |
matter | -0.129 |
blonde | -0.128 |
tactic | -0.127 |
Token |
|
Token ID | 198 |
Feature activation | -3.77e-02 |
Pos logit contributions | |
se | +0.050 |
bat | +0.049 |
goes | +0.048 |
blow | +0.048 |
wave | +0.047 |
Neg logit contributions | |
unlike | -0.064 |
eway | -0.060 |
Tears | -0.059 |
tactic | -0.058 |
perspective | -0.058 |
Token | They |
Token ID | 2990 |
Feature activation | -1.00e-01 |
Pos logit contributions | |
se | +0.039 |
Ben | +0.038 |
bat | +0.038 |
goes | +0.036 |
waves | +0.035 |
Neg logit contributions | |
unlike | -0.076 |
matter | -0.072 |
perspective | -0.071 |
eway | -0.070 |
superpower | -0.070 |
Token | walk |
Token ID | 2513 |
Feature activation | -1.29e-01 |
Pos logit contributions | |
ila | +0.103 |
pige | +0.099 |
Ben | +0.096 |
we | +0.094 |
next | +0.094 |
Neg logit contributions | |
matter | -0.190 |
blonde | -0.184 |
unlike | -0.181 |
superpower | -0.181 |
darkened | -0.174 |
Token | closer |
Token ID | 5699 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
Ben | +0.158 |
ila | +0.157 |
se | +0.157 |
bat | +0.156 |
Ben | +0.155 |
Neg logit contributions | |
unlike | -0.251 |
superpower | -0.226 |
matter | -0.225 |
tactic | -0.224 |
perspective | -0.222 |
Token | to |
Token ID | 284 |
Feature activation | +5.61e-02 |
Pos logit contributions | |
unlike | +0.279 |
superpower | +0.251 |
perspective | +0.248 |
tactic | +0.246 |
matter | +0.246 |
Neg logit contributions | |
bat | -0.145 |
Ben | -0.145 |
ila | -0.141 |
se | -0.140 |
Ben | -0.137 |
Token |
|
Token ID | 198 |
Feature activation | -1.79e-02 |
Pos logit contributions | |
unlike | +0.297 |
superpower | +0.271 |
matter | +0.269 |
blonde | +0.265 |
perspective | +0.264 |
Neg logit contributions | |
bat | -0.106 |
Ben | -0.105 |
se | -0.104 |
ila | -0.104 |
Ben | -0.104 |
Token |
|
Token ID | 198 |
Feature activation | -1.75e-02 |
Pos logit contributions | |
se | +0.023 |
bat | +0.023 |
goes | +0.022 |
wave | +0.021 |
dove | +0.021 |
Neg logit contributions | |
unlike | -0.032 |
Terr | -0.030 |
eway | -0.030 |
Tears | -0.029 |
tactic | -0.028 |
Token | The |
Token ID | 464 |
Feature activation | -2.56e-02 |
Pos logit contributions | |
Ben | +0.022 |
says | +0.020 |
bat | +0.020 |
se | +0.020 |
ila | +0.020 |
Neg logit contributions | |
unlike | -0.035 |
eway | -0.030 |
matter | -0.030 |
superpower | -0.030 |
perspective | -0.030 |
Token | girl |
Token ID | 2576 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
Ben | +0.031 |
bat | +0.031 |
ila | +0.031 |
Ben | +0.031 |
se | +0.031 |
Neg logit contributions | |
unlike | -0.052 |
superpower | -0.047 |
matter | -0.047 |
perspective | -0.046 |
blonde | -0.046 |
Token | then |
Token ID | 788 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
unlike | +0.249 |
superpower | +0.218 |
matter | +0.212 |
blonde | +0.211 |
Tears | +0.208 |
Neg logit contributions | |
ila | -0.179 |
Ben | -0.177 |
Ben | -0.172 |
bat | -0.169 |
dads | -0.165 |
Token | gave |
Token ID | 2921 |
Feature activation | -8.25e-02 |
Pos logit contributions | |
ila | +0.119 |
Ben | +0.119 |
bat | +0.113 |
Ben | +0.111 |
line | +0.109 |
Neg logit contributions | |
unlike | -0.222 |
matter | -0.200 |
blonde | -0.196 |
superpower | -0.195 |
darkened | -0.195 |
Token | him |
Token ID | 683 |
Feature activation | +5.80e-02 |
Pos logit contributions | |
bat | +0.089 |
Ben | +0.087 |
ila | +0.086 |
se | +0.085 |
Ben | +0.084 |
Neg logit contributions | |
unlike | -0.176 |
matter | -0.162 |
superpower | -0.159 |
perspective | -0.156 |
blonde | -0.156 |
Token | a |
Token ID | 257 |
Feature activation | +1.10e-03 |
Pos logit contributions | |
unlike | +0.118 |
superpower | +0.110 |
matter | +0.108 |
blonde | +0.107 |
Tears | +0.103 |
Neg logit contributions | |
Ben | -0.064 |
ila | -0.063 |
se | -0.061 |
bat | -0.059 |
line | -0.058 |
Token | hug |
Token ID | 16225 |
Feature activation | -3.14e-02 |
Pos logit contributions | |
unlike | +0.003 |
matter | +0.003 |
perspective | +0.003 |
eway | +0.002 |
superpower | +0.002 |
Neg logit contributions | |
bat | -0.001 |
se | -0.001 |
Ben | -0.001 |
Ben | -0.001 |
ila | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | -7.77e-02 |
Pos logit contributions | |
Ben | +0.015 |
bat | +0.015 |
ila | +0.014 |
Ben | +0.014 |
se | +0.014 |
Neg logit contributions | |
unlike | -0.086 |
matter | -0.080 |
superpower | -0.080 |
perspective | -0.078 |
blonde | -0.078 |
Token | said |
Token ID | 531 |
Feature activation | -5.77e-02 |
Pos logit contributions | |
Ben | +0.068 |
ila | +0.067 |
bat | +0.063 |
Ben | +0.059 |
line | +0.057 |
Neg logit contributions | |
unlike | -0.183 |
matter | -0.167 |
superpower | -0.165 |
blonde | -0.162 |
tactic | -0.159 |
Token | a |
Token ID | 257 |
Feature activation | -4.30e-03 |
Pos logit contributions | |
unlike | +0.258 |
matter | +0.238 |
superpower | +0.238 |
perspective | +0.233 |
tactic | +0.233 |
Neg logit contributions | |
Ben | -0.049 |
bat | -0.048 |
Ben | -0.045 |
se | -0.045 |
ila | -0.044 |
Token | field |
Token ID | 2214 |
Feature activation | -4.16e-02 |
Pos logit contributions | |
ila | +0.005 |
Ben | +0.005 |
Ben | +0.005 |
bat | +0.005 |
se | +0.005 |
Neg logit contributions | |
unlike | -0.009 |
superpower | -0.008 |
matter | -0.008 |
perspective | -0.008 |
blonde | -0.008 |
Token | of |
Token ID | 286 |
Feature activation | +9.81e-03 |
Pos logit contributions | |
Ben | +0.044 |
ila | +0.044 |
Ben | +0.043 |
se | +0.041 |
bat | +0.041 |
Neg logit contributions | |
unlike | -0.091 |
perspective | -0.082 |
eway | -0.080 |
tactic | -0.080 |
superpower | -0.080 |
Token | grass |
Token ID | 8701 |
Feature activation | -3.42e-02 |
Pos logit contributions | |
unlike | +0.020 |
matter | +0.018 |
blonde | +0.018 |
superpower | +0.018 |
perspective | +0.018 |
Neg logit contributions | |
ila | -0.012 |
Ben | -0.011 |
Ben | -0.011 |
bat | -0.011 |
se | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
Ben | +0.027 |
bat | +0.027 |
ila | +0.027 |
Ben | +0.027 |
se | +0.026 |
Neg logit contributions | |
unlike | -0.083 |
superpower | -0.076 |
matter | -0.076 |
perspective | -0.075 |
blonde | -0.074 |
Token | The |
Token ID | 383 |
Feature activation | -7.81e-03 |
Pos logit contributions | |
unlike | +0.038 |
superpower | +0.036 |
matter | +0.035 |
perspective | +0.035 |
tactic | +0.035 |
Neg logit contributions | |
Ben | -0.025 |
bat | -0.023 |
Ben | -0.022 |
put | -0.021 |
Tom | -0.021 |
Token | bright |
Token ID | 6016 |
Feature activation | +3.89e-02 |
Pos logit contributions | |
ila | +0.013 |
Ben | +0.011 |
gives | +0.010 |
Ben | +0.010 |
says | +0.010 |
Neg logit contributions | |
unlike | -0.013 |
matter | -0.012 |
perspective | -0.011 |
blonde | -0.011 |
darkened | -0.011 |
Token | green |
Token ID | 4077 |
Feature activation | -4.19e-03 |
Pos logit contributions | |
unlike | +0.081 |
matter | +0.074 |
perspective | +0.071 |
blonde | +0.070 |
heart | +0.069 |
Neg logit contributions | |
ila | -0.048 |
Ben | -0.046 |
Ben | -0.044 |
bat | -0.040 |
se | -0.039 |
Token | grass |
Token ID | 8701 |
Feature activation | -3.37e-02 |
Pos logit contributions | |
ila | +0.005 |
Ben | +0.005 |
Ben | +0.004 |
bat | +0.004 |
se | +0.004 |
Neg logit contributions | |
unlike | -0.009 |
matter | -0.008 |
perspective | -0.008 |
superpower | -0.008 |
blonde | -0.008 |
Token | was |
Token ID | 373 |
Feature activation | +7.97e-02 |
Pos logit contributions | |
Ben | +0.042 |
bat | +0.041 |
ila | +0.041 |
Ben | +0.041 |
se | +0.040 |
Neg logit contributions | |
unlike | -0.067 |
matter | -0.062 |
superpower | -0.061 |
perspective | -0.060 |
blonde | -0.059 |
Token | very |
Token ID | 845 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
unlike | +0.170 |
superpower | +0.164 |
perspective | +0.163 |
tactic | +0.163 |
matter | +0.161 |
Neg logit contributions | |
bat | -0.086 |
Ben | -0.085 |
se | -0.082 |
next | -0.080 |
put | -0.079 |
Token | anyway |
Token ID | 6949 |
Feature activation | +1.59e-01 |
Pos logit contributions | |
ila | +0.011 |
Ben | +0.011 |
Ben | +0.011 |
se | +0.010 |
bat | +0.010 |
Neg logit contributions | |
unlike | -0.023 |
superpower | -0.021 |
matter | -0.020 |
tactic | -0.020 |
blonde | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | +2.83e-02 |
Pos logit contributions | |
unlike | +0.340 |
matter | +0.306 |
eway | +0.297 |
superpower | +0.297 |
tactic | +0.294 |
Neg logit contributions | |
Ben | -0.185 |
bat | -0.173 |
Ben | -0.169 |
se | -0.168 |
ila | -0.166 |
Token |
|
Token ID | 198 |
Feature activation | -1.72e-02 |
Pos logit contributions | |
unlike | +0.053 |
matter | +0.048 |
superpower | +0.047 |
perspective | +0.046 |
blonde | +0.045 |
Neg logit contributions | |
Ben | -0.039 |
bat | -0.039 |
ila | -0.038 |
Ben | -0.038 |
se | -0.038 |
Token |
|
Token ID | 198 |
Feature activation | -1.70e-02 |
Pos logit contributions | |
Ben | +0.020 |
se | +0.020 |
bat | +0.019 |
goes | +0.019 |
wave | +0.019 |
Neg logit contributions | |
unlike | -0.033 |
eway | -0.031 |
superpower | -0.030 |
Tears | -0.030 |
tactic | -0.029 |
Token | After |
Token ID | 3260 |
Feature activation | +1.28e-01 |
Pos logit contributions | |
Ben | +0.025 |
bat | +0.023 |
se | +0.023 |
ila | +0.022 |
says | +0.022 |
Neg logit contributions | |
unlike | -0.030 |
eway | -0.027 |
superpower | -0.026 |
perspective | -0.026 |
tactic | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | +1.51e-03 |
Pos logit contributions | |
unlike | +0.330 |
superpower | +0.304 |
matter | +0.302 |
blonde | +0.297 |
perspective | +0.295 |
Neg logit contributions | |
Ben | -0.082 |
ila | -0.081 |
Ben | -0.080 |
bat | -0.080 |
se | -0.077 |
Token | little |
Token ID | 1310 |
Feature activation | +6.75e-02 |
Pos logit contributions | |
unlike | +0.002 |
adoes | +0.002 |
eway | +0.002 |
blonde | +0.002 |
perspective | +0.002 |
Neg logit contributions | |
bat | -0.003 |
se | -0.003 |
Ben | -0.003 |
Ben | -0.002 |
next | -0.002 |
Token | while |
Token ID | 981 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
unlike | +0.127 |
matter | +0.114 |
superpower | +0.114 |
perspective | +0.110 |
blonde | +0.110 |
Neg logit contributions | |
Ben | -0.091 |
ila | -0.090 |
bat | -0.090 |
Ben | -0.090 |
se | -0.088 |
Token | , |
Token ID | 11 |
Feature activation | +1.77e-02 |
Pos logit contributions | |
unlike | +0.412 |
matter | +0.375 |
superpower | +0.374 |
blonde | +0.365 |
perspective | +0.365 |
Neg logit contributions | |
Ben | -0.161 |
bat | -0.159 |
Ben | -0.159 |
ila | -0.155 |
se | -0.153 |
Token | Mam |
Token ID | 29926 |
Feature activation | -3.65e-02 |
Pos logit contributions | |
unlike | +0.047 |
superpower | +0.043 |
matter | +0.043 |
perspective | +0.042 |
blonde | +0.042 |
Neg logit contributions | |
Ben | -0.011 |
bat | -0.010 |
ila | -0.010 |
se | -0.010 |
Ben | -0.010 |
Token | ma |
Token ID | 2611 |
Feature activation | +2.35e-02 |
Pos logit contributions | |
bat | +0.054 |
Ben | +0.053 |
pige | +0.051 |
Ben | +0.051 |
ashes | +0.048 |
Neg logit contributions | |
matter | -0.058 |
unlike | -0.055 |
heart | -0.054 |
exactly | -0.052 |
tactic | -0.052 |
Token | good |
Token ID | 922 |
Feature activation | -1.45e-02 |
Pos logit contributions | |
unlike | +0.315 |
superpower | +0.287 |
perspective | +0.286 |
matter | +0.284 |
blonde | +0.282 |
Neg logit contributions | |
Ben | -0.181 |
bat | -0.175 |
se | -0.173 |
ila | -0.173 |
Ben | -0.170 |
Token | and |
Token ID | 290 |
Feature activation | -8.58e-02 |
Pos logit contributions | |
Ben | +0.011 |
bat | +0.010 |
ila | +0.010 |
Ben | +0.010 |
se | +0.010 |
Neg logit contributions | |
unlike | -0.037 |
superpower | -0.034 |
matter | -0.034 |
blonde | -0.033 |
perspective | -0.033 |
Token | make |
Token ID | 787 |
Feature activation | -9.67e-02 |
Pos logit contributions | |
ila | +0.104 |
Ben | +0.104 |
Ben | +0.103 |
bat | +0.102 |
se | +0.099 |
Neg logit contributions | |
unlike | -0.174 |
matter | -0.160 |
superpower | -0.156 |
blonde | -0.152 |
tactic | -0.151 |
Token | someone |
Token ID | 2130 |
Feature activation | +6.51e-02 |
Pos logit contributions | |
bat | +0.114 |
Ben | +0.113 |
Ben | +0.113 |
ila | +0.113 |
se | +0.111 |
Neg logit contributions | |
unlike | -0.198 |
matter | -0.180 |
superpower | -0.178 |
perspective | -0.173 |
blonde | -0.173 |
Token | happy |
Token ID | 3772 |
Feature activation | +2.40e-02 |
Pos logit contributions | |
unlike | +0.138 |
matter | +0.125 |
superpower | +0.124 |
blonde | +0.121 |
perspective | +0.121 |
Neg logit contributions | |
Ben | -0.073 |
Ben | -0.071 |
ila | -0.071 |
bat | -0.071 |
se | -0.069 |
Token | . |
Token ID | 13 |
Feature activation | +1.70e-02 |
Pos logit contributions | |
unlike | +0.065 |
blonde | +0.061 |
superpower | +0.060 |
perspective | +0.060 |
matter | +0.060 |
Neg logit contributions | |
Ben | -0.015 |
next | -0.013 |
bat | -0.012 |
says | -0.012 |
se | -0.011 |
Token | The |
Token ID | 383 |
Feature activation | -1.00e-02 |
Pos logit contributions | |
unlike | +0.031 |
Tears | +0.028 |
superpower | +0.027 |
Terr | +0.027 |
blonde | +0.026 |
Neg logit contributions | |
se | -0.024 |
ila | -0.023 |
bat | -0.023 |
wave | -0.022 |
Ben | -0.022 |
Token | end |
Token ID | 886 |
Feature activation | -2.04e-02 |
Pos logit contributions | |
ila | +0.011 |
Ben | +0.010 |
Ben | +0.010 |
bat | +0.010 |
se | +0.010 |
Neg logit contributions | |
unlike | -0.021 |
matter | -0.021 |
superpower | -0.020 |
blonde | -0.019 |
perspective | -0.019 |
Token | . |
Token ID | 13 |
Feature activation | +1.80e-02 |
Pos logit contributions | |
Ben | +0.012 |
bat | +0.011 |
Ben | +0.011 |
ila | +0.011 |
next | +0.011 |
Neg logit contributions | |
unlike | -0.056 |
blonde | -0.052 |
perspective | -0.051 |
matter | -0.051 |
superpower | -0.051 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -3.60e-03 |
Pos logit contributions | |
unlike | +0.036 |
Tears | +0.031 |
matter | +0.031 |
superpower | +0.031 |
Terr | +0.031 |
Neg logit contributions | |
se | -0.022 |
ila | -0.022 |
dads | -0.021 |
bat | -0.021 |
wave | -0.020 |
Token | Once |
Token ID | 7454 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
se | +0.004 |
dads | +0.004 |
ila | +0.004 |
bat | +0.004 |
Ben | +0.004 |
Neg logit contributions | |
unlike | -0.007 |
superpower | -0.007 |
Tears | -0.006 |
eway | -0.006 |
tactic | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | -9.41e-02 |
Pos logit contributions | |
unlike | +0.009 |
superpower | +0.008 |
matter | +0.008 |
blonde | +0.008 |
tactic | +0.008 |
Neg logit contributions | |
bat | -0.004 |
Ben | -0.004 |
se | -0.004 |
ila | -0.004 |
Ben | -0.003 |
Token | says |
Token ID | 1139 |
Feature activation | -1.29e-01 |
Pos logit contributions | |
Ben | +0.087 |
ila | +0.086 |
bat | +0.083 |
se | +0.075 |
pige | +0.070 |
Neg logit contributions | |
unlike | -0.209 |
matter | -0.197 |
superpower | -0.193 |
blonde | -0.190 |
tactic | -0.188 |
Token | hello |
Token ID | 23748 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
bat | +0.122 |
Ben | +0.121 |
Ben | +0.120 |
ila | +0.119 |
se | +0.117 |
Neg logit contributions | |
unlike | -0.293 |
superpower | -0.267 |
matter | -0.265 |
blonde | -0.262 |
perspective | -0.258 |
Token | . |
Token ID | 13 |
Feature activation | +1.52e-02 |
Pos logit contributions | |
unlike | +0.280 |
matter | +0.264 |
superpower | +0.260 |
perspective | +0.256 |
tactic | +0.256 |
Neg logit contributions | |
Ben | -0.064 |
next | -0.055 |
each | -0.050 |
bat | -0.049 |
and | -0.049 |
Token | Anna |
Token ID | 11735 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
unlike | +0.030 |
matter | +0.027 |
superpower | +0.027 |
perspective | +0.027 |
tactic | +0.026 |
Neg logit contributions | |
Ben | -0.021 |
Ben | -0.019 |
bat | -0.019 |
ila | -0.019 |
se | -0.019 |
Token | says |
Token ID | 1139 |
Feature activation | -1.28e-01 |
Pos logit contributions | |
ila | +0.097 |
line | +0.091 |
se | +0.090 |
Ben | +0.089 |
bat | +0.085 |
Neg logit contributions | |
unlike | -0.228 |
superpower | -0.223 |
Terr | -0.211 |
blonde | -0.210 |
Tears | -0.206 |
Token | , |
Token ID | 11 |
Feature activation | -1.04e-04 |
Pos logit contributions | |
bat | +0.158 |
Ben | +0.155 |
ila | +0.148 |
se | +0.144 |
Ben | +0.139 |
Neg logit contributions | |
unlike | -0.250 |
superpower | -0.226 |
blonde | -0.226 |
matter | -0.219 |
tactic | -0.212 |
Token | " |
Token ID | 366 |
Feature activation | +6.71e-02 |
Pos logit contributions | |
ila | +0.000 |
bat | +0.000 |
Ben | +0.000 |
se | +0.000 |
Ben | +0.000 |
Neg logit contributions | |
unlike | -0.000 |
superpower | -0.000 |
blonde | -0.000 |
matter | -0.000 |
Tears | -0.000 |
Token | Hi |
Token ID | 17250 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
unlike | +0.100 |
matter | +0.093 |
blonde | +0.089 |
perspective | +0.088 |
superpower | +0.086 |
Neg logit contributions | |
bat | -0.110 |
Ben | -0.108 |
ila | -0.106 |
se | -0.104 |
put | -0.102 |
Token | , |
Token ID | 11 |
Feature activation | +7.45e-04 |
Pos logit contributions | |
unlike | +0.286 |
matter | +0.264 |
superpower | +0.261 |
blonde | +0.257 |
perspective | +0.253 |
Neg logit contributions | |
bat | -0.134 |
se | -0.131 |
ila | -0.131 |
Ben | -0.130 |
Ben | -0.124 |
Token | dad |
Token ID | 9955 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
unlike | +0.002 |
matter | +0.001 |
superpower | +0.001 |
blonde | +0.001 |
tactic | +0.001 |
Neg logit contributions | |
bat | -0.001 |
Ben | -0.001 |
ila | -0.001 |
se | -0.001 |
irts | -0.001 |
Token | Mom |
Token ID | 29252 |
Feature activation | -3.88e-02 |
Pos logit contributions | |
unlike | +0.165 |
superpower | +0.157 |
tactic | +0.152 |
matter | +0.151 |
perspective | +0.150 |
Neg logit contributions | |
Ben | -0.065 |
Ben | -0.059 |
ila | -0.058 |
bat | -0.058 |
se | -0.057 |
Token | my |
Token ID | 1820 |
Feature activation | +7.72e-02 |
Pos logit contributions | |
Ben | +0.048 |
ila | +0.048 |
bat | +0.048 |
Ben | +0.047 |
se | +0.047 |
Neg logit contributions | |
unlike | -0.076 |
matter | -0.070 |
superpower | -0.068 |
blonde | -0.067 |
perspective | -0.067 |
Token | , |
Token ID | 11 |
Feature activation | +5.37e-03 |
Pos logit contributions | |
unlike | +0.161 |
matter | +0.147 |
superpower | +0.143 |
blonde | +0.143 |
perspective | +0.138 |
Neg logit contributions | |
ila | -0.088 |
Ben | -0.086 |
bat | -0.084 |
se | -0.083 |
Ben | -0.081 |
Token | can |
Token ID | 460 |
Feature activation | +4.91e-02 |
Pos logit contributions | |
unlike | +0.011 |
matter | +0.010 |
superpower | +0.010 |
blonde | +0.010 |
perspective | +0.010 |
Neg logit contributions | |
ila | -0.006 |
Ben | -0.006 |
bat | -0.006 |
Ben | -0.006 |
se | -0.006 |
Token | I |
Token ID | 314 |
Feature activation | -1.52e-01 |
Pos logit contributions | |
unlike | +0.112 |
matter | +0.105 |
blonde | +0.101 |
superpower | +0.099 |
alas | +0.098 |
Neg logit contributions | |
Ben | -0.045 |
ila | -0.044 |
bat | -0.043 |
se | -0.040 |
dads | -0.038 |
Token | have |
Token ID | 423 |
Feature activation | -1.42e-01 |
Pos logit contributions | |
ila | +0.139 |
Ben | +0.138 |
Ben | +0.134 |
se | +0.125 |
bat | +0.122 |
Neg logit contributions | |
unlike | -0.352 |
matter | -0.329 |
superpower | -0.325 |
blonde | -0.315 |
tactic | -0.310 |
Token | this |
Token ID | 428 |
Feature activation | -7.01e-02 |
Pos logit contributions | |
Ben | +0.107 |
bat | +0.104 |
se | +0.101 |
Ben | +0.100 |
ila | +0.100 |
Neg logit contributions | |
unlike | -0.355 |
superpower | -0.326 |
matter | -0.326 |
perspective | -0.321 |
tactic | -0.320 |
Token | toy |
Token ID | 13373 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
bat | +0.083 |
Ben | +0.081 |
slide | +0.077 |
se | +0.075 |
next | +0.075 |
Neg logit contributions | |
unlike | -0.148 |
perspective | -0.137 |
blonde | -0.136 |
matter | -0.133 |
eway | -0.133 |
Token | please |
Token ID | 3387 |
Feature activation | +6.21e-02 |
Pos logit contributions | |
unlike | +0.262 |
matter | +0.243 |
blonde | +0.242 |
perspective | +0.241 |
superpower | +0.241 |
Neg logit contributions | |
Ben | -0.098 |
bat | -0.097 |
se | -0.092 |
Ben | -0.088 |
ila | -0.087 |
Token | ?" |
Token ID | 1701 |
Feature activation | -3.32e-02 |
Pos logit contributions | |
unlike | +0.109 |
matter | +0.100 |
superpower | +0.097 |
tactic | +0.090 |
blonde | +0.090 |
Neg logit contributions | |
ila | -0.091 |
Ben | -0.089 |
Ben | -0.083 |
se | -0.082 |
dads | -0.080 |
Token | Mia |
Token ID | 32189 |
Feature activation | -7.99e-02 |
Pos logit contributions | |
Ben | +0.031 |
ila | +0.029 |
bat | +0.029 |
Ben | +0.029 |
se | +0.028 |
Neg logit contributions | |
unlike | -0.079 |
superpower | -0.072 |
matter | -0.072 |
perspective | -0.071 |
blonde | -0.070 |
Token | and |
Token ID | 290 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
ila | +0.007 |
Ben | +0.007 |
se | +0.007 |
bat | +0.007 |
Ben | +0.007 |
Neg logit contributions | |
unlike | -0.018 |
matter | -0.016 |
superpower | -0.016 |
perspective | -0.015 |
blonde | -0.015 |
Token | not |
Token ID | 407 |
Feature activation | -5.98e-02 |
Pos logit contributions | |
ila | +0.090 |
Ben | +0.089 |
bat | +0.084 |
Ben | +0.083 |
se | +0.081 |
Neg logit contributions | |
unlike | -0.250 |
matter | -0.230 |
superpower | -0.228 |
blonde | -0.220 |
perspective | -0.219 |
Token | bother |
Token ID | 11393 |
Feature activation | -8.79e-02 |
Pos logit contributions | |
Ben | +0.064 |
bat | +0.063 |
se | +0.061 |
Ben | +0.061 |
ila | +0.060 |
Neg logit contributions | |
unlike | -0.131 |
superpower | -0.119 |
matter | -0.117 |
blonde | -0.116 |
tactic | -0.116 |
Token | them |
Token ID | 606 |
Feature activation | -8.60e-03 |
Pos logit contributions | |
Ben | +0.075 |
bat | +0.072 |
ila | +0.071 |
Ben | +0.071 |
se | +0.069 |
Neg logit contributions | |
unlike | -0.212 |
superpower | -0.194 |
matter | -0.193 |
blonde | -0.190 |
perspective | -0.190 |
Token | when |
Token ID | 618 |
Feature activation | +9.83e-02 |
Pos logit contributions | |
Ben | +0.008 |
ila | +0.008 |
bat | +0.008 |
Ben | +0.008 |
se | +0.008 |
Neg logit contributions | |
unlike | -0.020 |
matter | -0.018 |
superpower | -0.018 |
perspective | -0.018 |
blonde | -0.017 |
Token | they |
Token ID | 484 |
Feature activation | -1.29e-01 |
Pos logit contributions | |
unlike | +0.272 |
matter | +0.253 |
superpower | +0.250 |
perspective | +0.245 |
blonde | +0.245 |
Neg logit contributions | |
Ben | -0.042 |
bat | -0.041 |
ila | -0.041 |
se | -0.040 |
Ben | -0.038 |
Token | are |
Token ID | 389 |
Feature activation | -8.67e-02 |
Pos logit contributions | |
ila | +0.128 |
Ben | +0.127 |
Ben | +0.127 |
bat | +0.122 |
se | +0.120 |
Neg logit contributions | |
unlike | -0.275 |
matter | -0.262 |
superpower | -0.254 |
perspective | -0.249 |
tactic | -0.245 |
Token | working |
Token ID | 1762 |
Feature activation | -3.93e-02 |
Pos logit contributions | |
Ben | +0.110 |
Ben | +0.109 |
bat | +0.108 |
ila | +0.108 |
se | +0.107 |
Neg logit contributions | |
unlike | -0.171 |
matter | -0.153 |
superpower | -0.153 |
blonde | -0.149 |
perspective | -0.148 |
Token | . |
Token ID | 13 |
Feature activation | +1.97e-03 |
Pos logit contributions | |
Ben | +0.038 |
ila | +0.037 |
Ben | +0.037 |
bat | +0.037 |
se | +0.036 |
Neg logit contributions | |
unlike | -0.089 |
matter | -0.081 |
superpower | -0.081 |
perspective | -0.079 |
blonde | -0.079 |
Token | They |
Token ID | 1119 |
Feature activation | -5.70e-02 |
Pos logit contributions | |
unlike | +0.005 |
matter | +0.004 |
superpower | +0.004 |
perspective | +0.004 |
tactic | +0.004 |
Neg logit contributions | |
Ben | -0.002 |
Ben | -0.002 |
ila | -0.002 |
bat | -0.002 |
se | -0.002 |
Token | learned |
Token ID | 4499 |
Feature activation | -7.02e-02 |
Pos logit contributions | |
ila | +0.059 |
Ben | +0.058 |
Ben | +0.058 |
bat | +0.055 |
se | +0.054 |
Neg logit contributions | |
unlike | -0.117 |
matter | -0.113 |
superpower | -0.110 |
perspective | -0.106 |
blonde | -0.106 |
Token | . |
Token ID | 13 |
Feature activation | +2.98e-02 |
Pos logit contributions | |
bat | +0.055 |
Ben | +0.053 |
se | +0.052 |
ila | +0.052 |
dads | +0.051 |
Neg logit contributions | |
unlike | -0.113 |
matter | -0.098 |
superpower | -0.097 |
blonde | -0.097 |
slightly | -0.096 |
Token | They |
Token ID | 1119 |
Feature activation | -2.84e-02 |
Pos logit contributions | |
unlike | +0.053 |
blonde | +0.048 |
Tears | +0.047 |
superpower | +0.046 |
Nobody | +0.046 |
Neg logit contributions | |
se | -0.043 |
ila | -0.040 |
bat | -0.040 |
dads | -0.039 |
next | -0.038 |
Token | g |
Token ID | 308 |
Feature activation | +5.22e-02 |
Pos logit contributions | |
ila | +0.034 |
Ben | +0.032 |
pige | +0.031 |
dads | +0.030 |
next | +0.030 |
Neg logit contributions | |
unlike | -0.051 |
matter | -0.050 |
superpower | -0.048 |
blonde | -0.048 |
perspective | -0.046 |
Token | iggle |
Token ID | 24082 |
Feature activation | -6.93e-02 |
Pos logit contributions | |
unlike | +0.091 |
superpower | +0.078 |
blonde | +0.077 |
perspective | +0.076 |
Tears | +0.075 |
Neg logit contributions | |
Ben | -0.079 |
se | -0.078 |
ila | -0.077 |
bat | -0.076 |
next | -0.074 |
Token | and |
Token ID | 290 |
Feature activation | -8.03e-02 |
Pos logit contributions | |
ila | +0.049 |
Ben | +0.046 |
se | +0.044 |
bat | +0.042 |
Ben | +0.040 |
Neg logit contributions | |
unlike | -0.175 |
superpower | -0.160 |
matter | -0.158 |
perspective | -0.157 |
blonde | -0.156 |
Token | have |
Token ID | 423 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
ila | +0.110 |
Ben | +0.102 |
iza | +0.098 |
ily | +0.097 |
pige | +0.096 |
Neg logit contributions | |
unlike | -0.133 |
matter | -0.131 |
blonde | -0.128 |
superpower | -0.125 |
fascination | -0.119 |
Token | fun |
Token ID | 1257 |
Feature activation | +1.92e-02 |
Pos logit contributions | |
ila | +0.178 |
Ben | +0.173 |
bat | +0.169 |
Ben | +0.168 |
se | +0.168 |
Neg logit contributions | |
unlike | -0.240 |
matter | -0.218 |
blonde | -0.215 |
superpower | -0.214 |
perspective | -0.208 |
Token | . |
Token ID | 13 |
Feature activation | +3.03e-02 |
Pos logit contributions | |
unlike | +0.044 |
superpower | +0.040 |
perspective | +0.040 |
matter | +0.040 |
blonde | +0.039 |
Neg logit contributions | |
Ben | -0.018 |
Ben | -0.018 |
ila | -0.018 |
bat | -0.017 |
se | -0.017 |
Token |
|
Token ID | 198 |
Feature activation | -1.62e-02 |
Pos logit contributions | |
unlike | +0.062 |
blonde | +0.056 |
Tears | +0.055 |
superpower | +0.055 |
matter | +0.054 |
Neg logit contributions | |
se | -0.036 |
ila | -0.034 |
bat | -0.033 |
Ben | -0.031 |
slide | -0.031 |
Token |
|
Token ID | 198 |
Feature activation | -1.58e-02 |
Pos logit contributions | |
se | +0.024 |
bat | +0.022 |
wave | +0.022 |
dads | +0.022 |
line | +0.022 |
Neg logit contributions | |
unlike | -0.025 |
eway | -0.023 |
Tears | -0.023 |
Finally | -0.022 |
Before | -0.022 |
Token | The |
Token ID | 464 |
Feature activation | -2.10e-02 |
Pos logit contributions | |
se | +0.024 |
bat | +0.023 |
wave | +0.022 |
goes | +0.022 |
dads | +0.022 |
Neg logit contributions | |
unlike | -0.023 |
slightly | -0.021 |
darkened | -0.021 |
total | -0.021 |
alas | -0.021 |
Token | go |
Token ID | 467 |
Feature activation | -1.94e-01 |
Pos logit contributions | |
matter | +0.105 |
unlike | +0.104 |
superpower | +0.098 |
former | +0.095 |
Frankie | +0.094 |
Neg logit contributions | |
ila | -0.057 |
Ben | -0.053 |
line | -0.048 |
Ben | -0.047 |
goes | -0.046 |
Token | on |
Token ID | 319 |
Feature activation | +7.81e-03 |
Pos logit contributions | |
Ben | +0.296 |
ila | +0.279 |
Ben | +0.266 |
Tom | +0.247 |
pige | +0.245 |
Neg logit contributions | |
unlike | -0.318 |
superpower | -0.293 |
matter | -0.287 |
eway | -0.274 |
tactic | -0.271 |
Token | your |
Token ID | 534 |
Feature activation | -1.35e-02 |
Pos logit contributions | |
unlike | +0.016 |
superpower | +0.015 |
matter | +0.014 |
perspective | +0.014 |
former | +0.014 |
Neg logit contributions | |
ila | -0.009 |
Ben | -0.009 |
se | -0.008 |
line | -0.008 |
Ben | -0.008 |
Token | ship |
Token ID | 4074 |
Feature activation | -5.64e-02 |
Pos logit contributions | |
ila | +0.017 |
Ben | +0.016 |
Ben | +0.015 |
line | +0.013 |
se | +0.013 |
Neg logit contributions | |
unlike | -0.027 |
superpower | -0.026 |
matter | -0.026 |
tactic | -0.025 |
former | -0.025 |
Token | ?" |
Token ID | 1701 |
Feature activation | -4.23e-02 |
Pos logit contributions | |
ila | +0.064 |
Ben | +0.063 |
Ben | +0.061 |
se | +0.060 |
bat | +0.059 |
Neg logit contributions | |
unlike | -0.118 |
matter | -0.108 |
superpower | -0.107 |
tactic | -0.103 |
perspective | -0.103 |
Token | Tom |
Token ID | 4186 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
Ben | +0.049 |
bat | +0.049 |
ila | +0.049 |
Ben | +0.049 |
se | +0.048 |
Neg logit contributions | |
unlike | -0.088 |
matter | -0.079 |
superpower | -0.079 |
blonde | -0.077 |
perspective | -0.077 |
Token | agreed |
Token ID | 4987 |
Feature activation | -2.45e-02 |
Pos logit contributions | |
Ben | +0.149 |
bat | +0.147 |
ila | +0.145 |
Ben | +0.145 |
se | +0.142 |
Neg logit contributions | |
unlike | -0.419 |
matter | -0.384 |
superpower | -0.383 |
blonde | -0.374 |
perspective | -0.374 |
Token | and |
Token ID | 290 |
Feature activation | -9.75e-02 |
Pos logit contributions | |
ila | +0.026 |
Ben | +0.024 |
pige | +0.022 |
se | +0.021 |
goes | +0.021 |
Neg logit contributions | |
unlike | -0.054 |
tactic | -0.049 |
matter | -0.049 |
superpower | -0.048 |
Tears | -0.047 |
Token | they |
Token ID | 484 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
Ben | +0.094 |
ila | +0.093 |
pige | +0.080 |
irts | +0.078 |
bat | +0.076 |
Neg logit contributions | |
unlike | -0.211 |
Frankie | -0.193 |
matter | -0.190 |
Mill | -0.187 |
superpower | -0.186 |
Token | went |
Token ID | 1816 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
ila | +0.153 |
Ben | +0.144 |
Ben | +0.137 |
pige | +0.136 |
bat | +0.127 |
Neg logit contributions | |
unlike | -0.229 |
matter | -0.222 |
superpower | -0.216 |
darkened | -0.212 |
blonde | -0.210 |
Token | on |
Token ID | 319 |
Feature activation | +8.06e-03 |
Pos logit contributions | |
Ben | +0.204 |
ila | +0.192 |
Ben | +0.184 |
pige | +0.183 |
se | +0.170 |
Neg logit contributions | |
unlike | -0.203 |
eway | -0.196 |
superpower | -0.193 |
tactic | -0.184 |
terribly | -0.181 |