Token | not |
Token ID | 407 |
Feature activation | +5.07e-02 |
Pos logit contributions | |
trick | +0.136 |
show | +0.133 |
grown | +0.122 |
style | +0.121 |
design | +0.120 |
Neg logit contributions | |
mie | -0.259 |
elope | -0.252 |
rified | -0.244 |
hew | -0.238 |
etsy | -0.236 |
Token | listen |
Token ID | 6004 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
new | +0.046 |
show | +0.044 |
trick | +0.044 |
matching | +0.042 |
making | +0.041 |
Neg logit contributions | |
mie | -0.110 |
ppy | -0.108 |
etsy | -0.105 |
elope | -0.105 |
jo | -0.104 |
Token | . |
Token ID | 13 |
Feature activation | +2.24e-02 |
Pos logit contributions | |
show | +0.127 |
trick | +0.121 |
test | +0.117 |
magic | +0.107 |
my | +0.106 |
Neg logit contributions | |
mie | -0.221 |
jo | -0.218 |
storms | -0.211 |
elope | -0.210 |
rified | -0.210 |
Token | He |
Token ID | 679 |
Feature activation | -2.13e-01 |
Pos logit contributions | |
show | +0.037 |
new | +0.036 |
trick | +0.036 |
grown | +0.036 |
to | +0.035 |
Neg logit contributions | |
mie | -0.036 |
jo | -0.035 |
elope | -0.035 |
thel | -0.034 |
etsy | -0.034 |
Token | jumped |
Token ID | 11687 |
Feature activation | +1.61e-01 |
Pos logit contributions | |
mie | +0.398 |
rified | +0.393 |
jo | +0.392 |
och | +0.390 |
etsy | +0.390 |
Neg logit contributions | |
show | -0.294 |
trick | -0.289 |
grown | -0.267 |
new | -0.263 |
to | -0.256 |
Token | into |
Token ID | 656 |
Feature activation | +3.08e-01 |
Pos logit contributions | |
show | +0.164 |
trick | +0.160 |
test | +0.152 |
style | +0.145 |
talent | +0.144 |
Neg logit contributions | |
rified | -0.350 |
mie | -0.338 |
ppy | -0.319 |
hew | -0.318 |
storms | -0.317 |
Token | the |
Token ID | 262 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
trick | +0.171 |
show | +0.163 |
test | +0.138 |
job | +0.129 |
new | +0.111 |
Neg logit contributions | |
mie | -0.793 |
rified | -0.786 |
elope | -0.781 |
hew | -0.745 |
Bear | -0.743 |
Token | pit |
Token ID | 6028 |
Feature activation | -1.42e-01 |
Pos logit contributions | |
trick | +0.316 |
show | +0.308 |
test | +0.285 |
style | +0.284 |
for | +0.278 |
Neg logit contributions | |
rified | -0.261 |
mie | -0.258 |
elope | -0.253 |
hew | -0.249 |
etsy | -0.234 |
Token | . |
Token ID | 13 |
Feature activation | +2.34e-02 |
Pos logit contributions | |
rified | +0.311 |
mie | +0.309 |
elope | +0.309 |
jo | +0.302 |
storms | +0.299 |
Neg logit contributions | |
trick | -0.131 |
show | -0.130 |
test | -0.117 |
new | -0.116 |
grown | -0.113 |
Token | He |
Token ID | 679 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
trick | +0.043 |
show | +0.043 |
test | +0.041 |
new | +0.040 |
work | +0.039 |
Neg logit contributions | |
rified | -0.032 |
mie | -0.032 |
etsy | -0.031 |
elope | -0.030 |
hew | -0.030 |
Token | landed |
Token ID | 11406 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
mie | +0.370 |
elope | +0.356 |
rified | +0.356 |
jo | +0.351 |
etsy | +0.351 |
Neg logit contributions | |
show | -0.319 |
trick | -0.315 |
new | -0.297 |
test | -0.290 |
grown | -0.286 |
Token | Ben |
Token ID | 11696 |
Feature activation | -2.52e-01 |
Pos logit contributions | |
rified | +0.132 |
jo | +0.131 |
storms | +0.130 |
mie | +0.125 |
elope | +0.123 |
Neg logit contributions | |
your | -0.100 |
show | -0.098 |
trick | -0.098 |
new | -0.098 |
my | -0.098 |
Token | nods |
Token ID | 34833 |
Feature activation | +1.57e-02 |
Pos logit contributions | |
mie | +0.431 |
pillar | +0.378 |
elope | +0.378 |
rified | +0.377 |
bo | +0.374 |
Neg logit contributions | |
trick | -0.375 |
job | -0.374 |
test | -0.346 |
style | -0.346 |
show | -0.342 |
Token | . |
Token ID | 13 |
Feature activation | +3.08e-02 |
Pos logit contributions | |
trick | +0.019 |
show | +0.019 |
test | +0.019 |
job | +0.018 |
matching | +0.017 |
Neg logit contributions | |
mie | -0.030 |
elope | -0.029 |
jo | -0.029 |
rified | -0.028 |
ppy | -0.028 |
Token | He |
Token ID | 679 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
show | +0.048 |
new | +0.048 |
trick | +0.046 |
grown | +0.046 |
to | +0.045 |
Neg logit contributions | |
mie | -0.053 |
hew | -0.052 |
jo | -0.051 |
rified | -0.051 |
elope | -0.050 |
Token | bites |
Token ID | 26081 |
Feature activation | +2.18e-01 |
Pos logit contributions | |
mie | +0.381 |
pillar | +0.333 |
hew | +0.331 |
elope | +0.328 |
rified | +0.325 |
Neg logit contributions | |
trick | -0.302 |
job | -0.300 |
test | -0.279 |
style | -0.277 |
show | -0.270 |
Token | into |
Token ID | 656 |
Feature activation | +3.08e-01 |
Pos logit contributions | |
show | +0.223 |
trick | +0.219 |
new | +0.195 |
test | +0.194 |
grown | +0.187 |
Neg logit contributions | |
mie | -0.487 |
rified | -0.475 |
elope | -0.468 |
etsy | -0.466 |
jo | -0.463 |
Token | his |
Token ID | 465 |
Feature activation | +9.44e-02 |
Pos logit contributions | |
show | +0.239 |
trick | +0.235 |
new | +0.200 |
test | +0.199 |
grown | +0.188 |
Neg logit contributions | |
mie | -0.758 |
elope | -0.736 |
rified | -0.736 |
etsy | -0.724 |
jo | -0.724 |
Token | half |
Token ID | 2063 |
Feature activation | +1.36e-02 |
Pos logit contributions | |
trick | +0.154 |
test | +0.145 |
making | +0.145 |
show | +0.144 |
grown | +0.141 |
Neg logit contributions | |
mie | -0.111 |
bo | -0.111 |
elope | -0.110 |
hew | -0.104 |
apa | -0.098 |
Token | of |
Token ID | 286 |
Feature activation | +2.62e-01 |
Pos logit contributions | |
show | +0.017 |
trick | +0.017 |
test | +0.017 |
job | +0.015 |
grown | +0.015 |
Neg logit contributions | |
mie | -0.025 |
ppy | -0.024 |
hew | -0.024 |
bo | -0.024 |
etsy | -0.024 |
Token | the |
Token ID | 262 |
Feature activation | +1.82e-01 |
Pos logit contributions | |
trick | +0.240 |
test | +0.236 |
job | +0.229 |
show | +0.213 |
power | +0.206 |
Neg logit contributions | |
mie | -0.519 |
elope | -0.491 |
ety | -0.489 |
Rex | -0.483 |
Max | -0.479 |
Token | sandwich |
Token ID | 20433 |
Feature activation | -1.76e-01 |
Pos logit contributions | |
trick | +0.261 |
show | +0.258 |
test | +0.240 |
style | +0.232 |
making | +0.231 |
Neg logit contributions | |
hew | -0.291 |
mie | -0.291 |
pillar | -0.281 |
rified | -0.276 |
etsy | -0.268 |
Token | He |
Token ID | 1544 |
Feature activation | -1.84e-01 |
Pos logit contributions | |
mie | +0.094 |
etsy | +0.092 |
storms | +0.090 |
rified | +0.089 |
jo | +0.089 |
Neg logit contributions | |
trick | -0.143 |
show | -0.143 |
new | -0.139 |
my | -0.134 |
to | -0.133 |
Token | p |
Token ID | 279 |
Feature activation | +4.40e-02 |
Pos logit contributions | |
mie | +0.331 |
rified | +0.295 |
hew | +0.292 |
ppy | +0.292 |
azz | +0.291 |
Neg logit contributions | |
job | -0.254 |
trick | -0.243 |
style | -0.224 |
test | -0.223 |
show | -0.218 |
Token | ours |
Token ID | 4662 |
Feature activation | +4.51e-02 |
Pos logit contributions | |
show | +0.061 |
trick | +0.059 |
for | +0.057 |
new | +0.056 |
to | +0.056 |
Neg logit contributions | |
elope | -0.081 |
mie | -0.080 |
rified | -0.080 |
Mic | -0.079 |
etsy | -0.079 |
Token | some |
Token ID | 617 |
Feature activation | +1.67e-01 |
Pos logit contributions | |
trick | +0.038 |
show | +0.038 |
test | +0.035 |
grown | +0.031 |
design | +0.030 |
Neg logit contributions | |
rified | -0.104 |
mie | -0.104 |
bo | -0.099 |
hew | -0.098 |
Bear | -0.098 |
Token | water |
Token ID | 1660 |
Feature activation | -8.94e-03 |
Pos logit contributions | |
trick | +0.209 |
meeting | +0.209 |
show | +0.196 |
test | +0.193 |
spelling | +0.191 |
Neg logit contributions | |
Bear | -0.300 |
rified | -0.285 |
bo | -0.282 |
mie | -0.278 |
pillar | -0.270 |
Token | into |
Token ID | 656 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
bo | +0.018 |
rified | +0.018 |
ppy | +0.017 |
mie | +0.017 |
als | +0.017 |
Neg logit contributions | |
show | -0.010 |
trick | -0.009 |
test | -0.009 |
talent | -0.009 |
job | -0.009 |
Token | Lia |
Token ID | 49520 |
Feature activation | -2.89e-01 |
Pos logit contributions | |
show | +0.164 |
trick | +0.159 |
new | +0.132 |
test | +0.121 |
grown | +0.117 |
Neg logit contributions | |
mie | -0.834 |
elope | -0.814 |
rified | -0.814 |
jo | -0.808 |
etsy | -0.808 |
Token | 's |
Token ID | 338 |
Feature activation | +6.67e-02 |
Pos logit contributions | |
mie | +0.540 |
elope | +0.523 |
etsy | +0.509 |
jo | +0.509 |
rified | +0.509 |
Neg logit contributions | |
show | -0.393 |
trick | -0.391 |
new | -0.362 |
test | -0.358 |
grown | -0.350 |
Token | rocket |
Token ID | 10701 |
Feature activation | -2.45e-01 |
Pos logit contributions | |
trick | +0.111 |
show | +0.103 |
grown | +0.100 |
test | +0.099 |
making | +0.096 |
Neg logit contributions | |
mie | -0.097 |
elope | -0.095 |
rified | -0.091 |
bo | -0.090 |
hew | -0.090 |
Token | and |
Token ID | 290 |
Feature activation | +1.54e-01 |
Pos logit contributions | |
elope | +0.528 |
mie | +0.525 |
jo | +0.521 |
rified | +0.505 |
storms | +0.504 |
Neg logit contributions | |
show | -0.231 |
trick | -0.230 |
new | -0.215 |
grown | -0.214 |
test | -0.208 |
Token | closes |
Token ID | 20612 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
trick | +0.182 |
show | +0.167 |
new | +0.160 |
grown | +0.159 |
test | +0.158 |
Neg logit contributions | |
mie | -0.311 |
elope | -0.296 |
rified | -0.290 |
storms | -0.284 |
hew | -0.283 |
Token | . |
Token ID | 13 |
Feature activation | +1.95e-02 |
Pos logit contributions | |
elope | +0.211 |
jo | +0.211 |
storms | +0.209 |
mie | +0.207 |
rified | +0.205 |
Neg logit contributions | |
trick | -0.158 |
test | -0.152 |
show | -0.152 |
new | -0.148 |
math | -0.147 |
Token | They |
Token ID | 1119 |
Feature activation | -2.13e-01 |
Pos logit contributions | |
show | +0.033 |
trick | +0.033 |
new | +0.032 |
grown | +0.031 |
test | +0.030 |
Neg logit contributions | |
mie | -0.031 |
elope | -0.029 |
rified | -0.029 |
jo | -0.029 |
etsy | -0.029 |
Token | watched |
Token ID | 7342 |
Feature activation | +1.64e-01 |
Pos logit contributions | |
mie | +0.345 |
etsy | +0.337 |
rified | +0.334 |
elope | +0.332 |
jo | +0.327 |
Neg logit contributions | |
show | -0.357 |
trick | -0.341 |
grown | -0.326 |
new | -0.326 |
work | -0.315 |
Token | it |
Token ID | 340 |
Feature activation | -8.95e-02 |
Pos logit contributions | |
trick | +0.198 |
test | +0.184 |
style | +0.178 |
show | +0.173 |
balance | +0.169 |
Neg logit contributions | |
rified | -0.309 |
storms | -0.296 |
elope | -0.291 |
mie | -0.289 |
jo | -0.286 |
Token | explode |
Token ID | 22818 |
Feature activation | -5.70e-02 |
Pos logit contributions | |
elope | +0.178 |
jo | +0.169 |
storms | +0.167 |
hel | +0.166 |
hew | +0.156 |
Neg logit contributions | |
trick | -0.079 |
test | -0.079 |
my | -0.079 |
style | -0.078 |
math | -0.077 |
Token | into |
Token ID | 656 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
rified | +0.121 |
mie | +0.120 |
elope | +0.120 |
jo | +0.119 |
storms | +0.118 |
Neg logit contributions | |
trick | -0.055 |
show | -0.055 |
test | -0.054 |
grown | -0.050 |
new | -0.050 |
Token | many |
Token ID | 867 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
trick | +0.177 |
show | +0.170 |
test | +0.144 |
for | +0.121 |
new | +0.121 |
Neg logit contributions | |
elope | -0.785 |
mie | -0.770 |
rified | -0.767 |
jo | -0.753 |
hew | -0.744 |
Token | colors |
Token ID | 7577 |
Feature activation | +4.78e-02 |
Pos logit contributions | |
trick | +0.170 |
show | +0.161 |
test | +0.154 |
for | +0.147 |
job | +0.144 |
Neg logit contributions | |
mie | -0.285 |
elope | -0.279 |
jo | -0.278 |
rified | -0.277 |
hew | -0.271 |
Token | and |
Token ID | 290 |
Feature activation | +1.52e-01 |
Pos logit contributions | |
trick | +0.053 |
show | +0.052 |
test | +0.050 |
my | +0.048 |
new | +0.047 |
Neg logit contributions | |
mie | -0.094 |
storms | -0.094 |
jo | -0.093 |
rified | -0.090 |
hel | -0.090 |
Token | shapes |
Token ID | 15268 |
Feature activation | -6.96e-02 |
Pos logit contributions | |
trick | +0.180 |
show | +0.176 |
test | +0.166 |
new | +0.162 |
my | +0.155 |
Neg logit contributions | |
mie | -0.296 |
elope | -0.295 |
jo | -0.289 |
rified | -0.287 |
storms | -0.284 |
Token | . |
Token ID | 13 |
Feature activation | +2.21e-02 |
Pos logit contributions | |
storms | +0.127 |
mie | +0.120 |
jo | +0.119 |
rified | +0.117 |
hel | +0.117 |
Neg logit contributions | |
test | -0.081 |
my | -0.080 |
trick | -0.080 |
job | -0.078 |
show | -0.078 |
Token | agreed |
Token ID | 4987 |
Feature activation | +6.72e-02 |
Pos logit contributions | |
mie | +0.411 |
elope | +0.401 |
jo | +0.398 |
rified | +0.396 |
etsy | +0.394 |
Neg logit contributions | |
show | -0.480 |
trick | -0.469 |
new | -0.446 |
grown | -0.436 |
test | -0.436 |
Token | . |
Token ID | 13 |
Feature activation | +2.58e-02 |
Pos logit contributions | |
trick | +0.076 |
show | +0.076 |
test | +0.071 |
my | +0.066 |
magic | +0.066 |
Neg logit contributions | |
mie | -0.138 |
etsy | -0.131 |
hew | -0.129 |
elope | -0.129 |
rified | -0.129 |
Token | They |
Token ID | 1119 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
show | +0.042 |
trick | +0.041 |
new | +0.040 |
to | +0.039 |
for | +0.039 |
Neg logit contributions | |
mie | -0.044 |
rified | -0.041 |
jo | -0.041 |
etsy | -0.041 |
hew | -0.040 |
Token | jumped |
Token ID | 11687 |
Feature activation | +1.58e-01 |
Pos logit contributions | |
mie | +0.335 |
rified | +0.314 |
etsy | +0.311 |
elope | +0.310 |
hew | +0.305 |
Neg logit contributions | |
trick | -0.356 |
show | -0.352 |
new | -0.337 |
test | -0.331 |
magic | -0.322 |
Token | back |
Token ID | 736 |
Feature activation | +9.46e-02 |
Pos logit contributions | |
test | +0.199 |
trick | +0.194 |
show | +0.183 |
design | +0.181 |
spelling | +0.176 |
Neg logit contributions | |
rified | -0.288 |
mie | -0.271 |
Hop | -0.265 |
bo | -0.253 |
storms | -0.252 |
Token | into |
Token ID | 656 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
trick | +0.113 |
show | +0.100 |
grown | +0.099 |
test | +0.098 |
design | +0.095 |
Neg logit contributions | |
mie | -0.176 |
elope | -0.174 |
jo | -0.171 |
storms | -0.169 |
x | -0.168 |
Token | the |
Token ID | 262 |
Feature activation | +1.82e-01 |
Pos logit contributions | |
trick | +0.253 |
job | +0.238 |
show | +0.211 |
test | +0.206 |
for | +0.198 |
Neg logit contributions | |
elope | -0.648 |
mie | -0.629 |
Max | -0.623 |
Le | -0.618 |
rified | -0.611 |
Token | pool |
Token ID | 5933 |
Feature activation | -1.10e-01 |
Pos logit contributions | |
trick | +0.234 |
five | +0.220 |
to | +0.213 |
making | +0.210 |
spelling | +0.210 |
Neg logit contributions | |
Rex | -0.290 |
rified | -0.289 |
hew | -0.287 |
mie | -0.283 |
bo | -0.278 |
Token | and |
Token ID | 290 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
rified | +0.230 |
mie | +0.229 |
elope | +0.226 |
storms | +0.221 |
jo | +0.220 |
Neg logit contributions | |
trick | -0.106 |
show | -0.102 |
test | -0.098 |
my | -0.095 |
new | -0.091 |
Token | sw |
Token ID | 1509 |
Feature activation | +4.38e-02 |
Pos logit contributions | |
trick | +0.252 |
show | +0.238 |
test | +0.235 |
new | +0.232 |
job | +0.228 |
Neg logit contributions | |
mie | -0.245 |
rified | -0.229 |
elope | -0.227 |
jo | -0.217 |
etsy | -0.214 |
Token | am |
Token ID | 321 |
Feature activation | +2.23e-02 |
Pos logit contributions | |
show | +0.088 |
trick | +0.087 |
new | +0.083 |
test | +0.083 |
grown | +0.081 |
Neg logit contributions | |
mie | -0.055 |
hew | -0.053 |
elope | -0.052 |
rified | -0.052 |
jo | -0.051 |
Token | took |
Token ID | 1718 |
Feature activation | +2.12e-01 |
Pos logit contributions | |
mie | +0.405 |
elope | +0.401 |
jo | +0.400 |
etsy | +0.398 |
rified | +0.397 |
Neg logit contributions | |
show | -0.338 |
trick | -0.329 |
new | -0.306 |
grown | -0.304 |
test | -0.297 |
Token | another |
Token ID | 1194 |
Feature activation | +9.51e-02 |
Pos logit contributions | |
show | +0.095 |
trick | +0.090 |
new | +0.080 |
test | +0.065 |
for | +0.064 |
Neg logit contributions | |
mie | -0.598 |
elope | -0.595 |
etsy | -0.593 |
jo | -0.591 |
rified | -0.585 |
Token | balloon |
Token ID | 21190 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
trick | +0.128 |
grown | +0.119 |
making | +0.116 |
show | +0.115 |
my | +0.110 |
Neg logit contributions | |
mie | -0.166 |
rified | -0.149 |
pillar | -0.148 |
etsy | -0.147 |
och | -0.144 |
Token | and |
Token ID | 290 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
mie | +0.284 |
rified | +0.278 |
elope | +0.276 |
etsy | +0.274 |
jo | +0.273 |
Neg logit contributions | |
show | -0.085 |
trick | -0.084 |
new | -0.075 |
grown | -0.072 |
test | -0.071 |
Token | blew |
Token ID | 17948 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
trick | +0.230 |
show | +0.216 |
job | +0.214 |
new | +0.210 |
test | +0.209 |
Neg logit contributions | |
mie | -0.259 |
rified | -0.242 |
elope | -0.228 |
hew | -0.225 |
ppy | -0.223 |
Token | into |
Token ID | 656 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
show | +0.152 |
trick | +0.147 |
new | +0.137 |
for | +0.126 |
grown | +0.124 |
Neg logit contributions | |
mie | -0.431 |
elope | -0.425 |
jo | -0.422 |
etsy | -0.421 |
rified | -0.418 |
Token | it |
Token ID | 340 |
Feature activation | -8.78e-02 |
Pos logit contributions | |
show | +0.224 |
trick | +0.214 |
new | +0.198 |
test | +0.178 |
grown | +0.175 |
Neg logit contributions | |
mie | -0.783 |
elope | -0.768 |
etsy | -0.768 |
jo | -0.767 |
rified | -0.767 |
Token | . |
Token ID | 13 |
Feature activation | +2.70e-02 |
Pos logit contributions | |
mie | +0.147 |
rified | +0.147 |
elope | +0.145 |
storms | +0.144 |
ppy | +0.141 |
Neg logit contributions | |
trick | -0.112 |
show | -0.111 |
test | -0.107 |
grown | -0.106 |
my | -0.102 |
Token | It |
Token ID | 632 |
Feature activation | -1.64e-01 |
Pos logit contributions | |
show | +0.054 |
trick | +0.053 |
new | +0.051 |
test | +0.050 |
grown | +0.050 |
Neg logit contributions | |
mie | -0.034 |
rified | -0.032 |
elope | -0.032 |
etsy | -0.032 |
jo | -0.032 |
Token | became |
Token ID | 2627 |
Feature activation | +1.98e-01 |
Pos logit contributions | |
bo | +0.260 |
mie | +0.251 |
rified | +0.247 |
elope | +0.245 |
storms | +0.238 |
Neg logit contributions | |
show | -0.243 |
job | -0.239 |
new | -0.237 |
test | -0.237 |
trick | -0.234 |
Token | big |
Token ID | 1263 |
Feature activation | +1.54e-01 |
Pos logit contributions | |
show | +0.193 |
trick | +0.187 |
test | +0.173 |
grown | +0.147 |
design | +0.144 |
Neg logit contributions | |
mie | -0.449 |
rified | -0.427 |
elope | -0.427 |
etsy | -0.418 |
hew | -0.417 |
Token | and |
Token ID | 290 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
mie | +0.303 |
rified | +0.302 |
och | +0.301 |
etsy | +0.299 |
jo | +0.297 |
Neg logit contributions | |
show | -0.222 |
trick | -0.215 |
new | -0.201 |
grown | -0.200 |
to | -0.194 |
Token | Max |
Token ID | 5436 |
Feature activation | -3.13e-01 |
Pos logit contributions | |
show | +0.274 |
the | +0.262 |
trick | +0.260 |
his | +0.258 |
new | +0.254 |
Neg logit contributions | |
etsy | -0.236 |
rified | -0.231 |
hew | -0.230 |
mie | -0.229 |
pillar | -0.227 |
Token | followed |
Token ID | 3940 |
Feature activation | +1.70e-01 |
Pos logit contributions | |
och | +0.602 |
lor | +0.579 |
jo | +0.576 |
Mic | +0.566 |
pillar | +0.563 |
Neg logit contributions | |
show | -0.462 |
trick | -0.411 |
for | -0.407 |
grown | -0.406 |
the | -0.406 |
Token | Grand |
Token ID | 5675 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
show | +0.169 |
trick | +0.160 |
new | +0.150 |
for | +0.146 |
to | +0.145 |
Neg logit contributions | |
mie | -0.393 |
rified | -0.383 |
hew | -0.379 |
etsy | -0.379 |
elope | -0.377 |
Token | ma |
Token ID | 2611 |
Feature activation | -2.26e-01 |
Pos logit contributions | |
rified | +0.182 |
lves | +0.181 |
etsy | +0.178 |
elope | +0.175 |
mie | +0.174 |
Neg logit contributions | |
show | -0.211 |
trick | -0.208 |
new | -0.197 |
for | -0.194 |
work | -0.193 |
Token | into |
Token ID | 656 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
mie | +0.638 |
rified | +0.635 |
thel | +0.622 |
etsy | +0.620 |
och | +0.619 |
Neg logit contributions | |
show | -0.104 |
to | -0.104 |
for | -0.100 |
trick | -0.092 |
new | -0.086 |
Token | the |
Token ID | 262 |
Feature activation | +1.84e-01 |
Pos logit contributions | |
trick | +0.117 |
show | +0.116 |
test | +0.077 |
new | +0.077 |
grown | +0.070 |
Neg logit contributions | |
mie | -0.869 |
elope | -0.860 |
rified | -0.851 |
jo | -0.842 |
etsy | -0.838 |
Token | living |
Token ID | 2877 |
Feature activation | +8.74e-02 |
Pos logit contributions | |
show | +0.251 |
trick | +0.243 |
new | +0.236 |
test | +0.224 |
grown | +0.224 |
Neg logit contributions | |
mie | -0.357 |
rified | -0.344 |
etsy | -0.343 |
jo | -0.341 |
elope | -0.338 |
Token | room |
Token ID | 2119 |
Feature activation | -8.47e-02 |
Pos logit contributions | |
trick | +0.119 |
show | +0.114 |
test | +0.110 |
work | +0.106 |
five | +0.104 |
Neg logit contributions | |
elope | -0.162 |
mie | -0.157 |
rified | -0.155 |
jo | -0.150 |
hew | -0.145 |
Token | , |
Token ID | 11 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
mie | +0.196 |
elope | +0.196 |
rified | +0.193 |
jo | +0.191 |
storms | +0.190 |
Neg logit contributions | |
trick | -0.068 |
show | -0.066 |
my | -0.061 |
test | -0.060 |
your | -0.057 |
Token | but |
Token ID | 475 |
Feature activation | +1.83e-02 |
Pos logit contributions | |
trick | +0.145 |
show | +0.143 |
test | +0.126 |
new | +0.126 |
grown | +0.119 |
Neg logit contributions | |
mie | -0.362 |
elope | -0.353 |
rified | -0.349 |
etsy | -0.346 |
jo | -0.345 |
Token | garage |
Token ID | 15591 |
Feature activation | -1.82e-01 |
Pos logit contributions | |
trick | +0.230 |
show | +0.228 |
new | +0.207 |
test | +0.202 |
for | +0.200 |
Neg logit contributions | |
mie | -0.354 |
elope | -0.342 |
rified | -0.341 |
etsy | -0.336 |
jo | -0.334 |
Token | . |
Token ID | 13 |
Feature activation | +1.80e-02 |
Pos logit contributions | |
mie | +0.372 |
rified | +0.358 |
jo | +0.358 |
elope | +0.357 |
storms | +0.352 |
Neg logit contributions | |
trick | -0.203 |
show | -0.199 |
new | -0.186 |
test | -0.185 |
my | -0.179 |
Token | She |
Token ID | 1375 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
show | +0.026 |
new | +0.025 |
trick | +0.024 |
grown | +0.024 |
to | +0.024 |
Neg logit contributions | |
mie | -0.033 |
elope | -0.033 |
rified | -0.033 |
jo | -0.032 |
hew | -0.031 |
Token | cut |
Token ID | 2005 |
Feature activation | +1.64e-01 |
Pos logit contributions | |
mie | +0.392 |
elope | +0.354 |
hew | +0.352 |
jo | +0.352 |
etsy | +0.343 |
Neg logit contributions | |
trick | -0.353 |
show | -0.339 |
test | -0.336 |
job | -0.329 |
new | -0.326 |
Token | it |
Token ID | 340 |
Feature activation | -8.99e-02 |
Pos logit contributions | |
show | +0.125 |
trick | +0.123 |
test | +0.104 |
new | +0.101 |
grown | +0.099 |
Neg logit contributions | |
mie | -0.410 |
rified | -0.398 |
elope | -0.392 |
etsy | -0.391 |
jo | -0.390 |
Token | into |
Token ID | 656 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
hel | +0.153 |
mie | +0.150 |
x | +0.146 |
hew | +0.146 |
ocular | +0.145 |
Neg logit contributions | |
talent | -0.099 |
grown | -0.092 |
show | -0.092 |
trick | -0.090 |
meeting | -0.089 |
Token | the |
Token ID | 262 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
show | +0.264 |
trick | +0.261 |
test | +0.229 |
job | +0.204 |
new | +0.204 |
Neg logit contributions | |
mie | -0.715 |
elope | -0.688 |
rified | -0.685 |
hew | -0.681 |
jo | -0.677 |
Token | shape |
Token ID | 5485 |
Feature activation | -1.93e-03 |
Pos logit contributions | |
trick | +0.267 |
show | +0.261 |
new | +0.239 |
test | +0.236 |
grown | +0.234 |
Neg logit contributions | |
mie | -0.325 |
rified | -0.312 |
elope | -0.306 |
hew | -0.304 |
etsy | -0.302 |
Token | of |
Token ID | 286 |
Feature activation | +2.59e-01 |
Pos logit contributions | |
mie | +0.004 |
pillar | +0.004 |
och | +0.004 |
rified | +0.004 |
jo | +0.004 |
Neg logit contributions | |
trick | -0.002 |
show | -0.002 |
test | -0.002 |
new | -0.002 |
power | -0.002 |
Token | a |
Token ID | 257 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
trick | +0.210 |
show | +0.193 |
test | +0.181 |
job | +0.166 |
quiz | +0.159 |
Neg logit contributions | |
mie | -0.581 |
elope | -0.553 |
rified | -0.546 |
etsy | -0.543 |
Max | -0.542 |
Token | doll |
Token ID | 3654 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
show | +0.168 |
trick | +0.168 |
new | +0.149 |
test | +0.146 |
grown | +0.142 |
Neg logit contributions | |
mie | -0.387 |
elope | -0.376 |
rified | -0.373 |
etsy | -0.373 |
jo | -0.370 |
Token | and |
Token ID | 290 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
elope | +0.267 |
mie | +0.263 |
hew | +0.255 |
rified | +0.255 |
etsy | +0.254 |
Neg logit contributions | |
trick | -0.140 |
show | -0.133 |
test | -0.130 |
magic | -0.123 |
new | -0.119 |
Token | threw |
Token ID | 9617 |
Feature activation | +1.47e-01 |
Pos logit contributions | |
show | +0.223 |
trick | +0.208 |
new | +0.199 |
to | +0.198 |
grown | +0.196 |
Neg logit contributions | |
mie | -0.279 |
jo | -0.278 |
hew | -0.272 |
rified | -0.272 |
etsy | -0.271 |
Token | some |
Token ID | 617 |
Feature activation | +1.68e-01 |
Pos logit contributions | |
show | +0.095 |
trick | +0.092 |
new | +0.080 |
test | +0.074 |
grown | +0.072 |
Neg logit contributions | |
mie | -0.385 |
elope | -0.375 |
rified | -0.375 |
jo | -0.373 |
etsy | -0.372 |
Token | cr |
Token ID | 1067 |
Feature activation | +5.65e-02 |
Pos logit contributions | |
show | +0.217 |
trick | +0.215 |
test | +0.211 |
spelling | +0.202 |
school | +0.201 |
Neg logit contributions | |
pillar | -0.291 |
rified | -0.289 |
mie | -0.280 |
hew | -0.278 |
bo | -0.276 |
Token | umbs |
Token ID | 18146 |
Feature activation | -1.28e-01 |
Pos logit contributions | |
trick | +0.127 |
show | +0.126 |
new | +0.120 |
for | +0.118 |
to | +0.118 |
Neg logit contributions | |
mie | -0.058 |
elope | -0.054 |
rified | -0.053 |
jo | -0.049 |
etsy | -0.049 |
Token | into |
Token ID | 656 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
rified | +0.282 |
mie | +0.277 |
ppy | +0.277 |
etsy | +0.275 |
bo | +0.272 |
Neg logit contributions | |
trick | -0.113 |
show | -0.111 |
test | -0.105 |
design | -0.100 |
grown | -0.097 |
Token | the |
Token ID | 262 |
Feature activation | +1.85e-01 |
Pos logit contributions | |
show | +0.121 |
trick | +0.116 |
new | +0.090 |
test | +0.077 |
grown | +0.075 |
Neg logit contributions | |
mie | -0.876 |
elope | -0.856 |
rified | -0.856 |
jo | -0.852 |
etsy | -0.851 |
Token | water |
Token ID | 1660 |
Feature activation | -7.69e-03 |
Pos logit contributions | |
show | +0.171 |
trick | +0.169 |
new | +0.150 |
test | +0.146 |
grown | +0.142 |
Neg logit contributions | |
mie | -0.430 |
rified | -0.419 |
elope | -0.418 |
etsy | -0.413 |
jo | -0.411 |
Token | . |
Token ID | 13 |
Feature activation | +2.68e-02 |
Pos logit contributions | |
mie | +0.014 |
rified | +0.014 |
elope | +0.014 |
storms | +0.014 |
etsy | +0.014 |
Neg logit contributions | |
trick | -0.010 |
show | -0.010 |
test | -0.009 |
new | -0.009 |
math | -0.009 |
Token |
|
Token ID | 198 |
Feature activation | -6.80e-02 |
Pos logit contributions | |
show | +0.036 |
to | +0.036 |
the | +0.035 |
new | +0.034 |
trick | +0.034 |
Neg logit contributions | |
mie | -0.055 |
ppa | -0.052 |
och | -0.052 |
jo | -0.050 |
thel | -0.050 |
Token |
|
Token ID | 198 |
Feature activation | -6.76e-02 |
Pos logit contributions | |
mie | +0.136 |
rified | +0.131 |
lves | +0.126 |
storms | +0.125 |
och | +0.123 |
Neg logit contributions | |
show | -0.092 |
trick | -0.091 |
to | -0.090 |
new | -0.088 |
grown | -0.087 |
Token | from |
Token ID | 422 |
Feature activation | +2.31e-01 |
Pos logit contributions | |
trick | +0.018 |
show | +0.018 |
grown | +0.017 |
making | +0.016 |
test | +0.015 |
Neg logit contributions | |
mie | -0.048 |
etsy | -0.047 |
rified | -0.046 |
pillar | -0.046 |
elope | -0.046 |
Token | the |
Token ID | 262 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
show | +0.378 |
trick | +0.372 |
new | +0.353 |
test | +0.344 |
grown | +0.341 |
Neg logit contributions | |
mie | -0.377 |
elope | -0.363 |
rified | -0.361 |
jo | -0.359 |
etsy | -0.358 |
Token | basket |
Token ID | 7988 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
show | +0.225 |
trick | +0.212 |
new | +0.212 |
test | +0.198 |
grown | +0.197 |
Neg logit contributions | |
jo | -0.390 |
mie | -0.380 |
elope | -0.372 |
hew | -0.366 |
rified | -0.366 |
Token | and |
Token ID | 290 |
Feature activation | +1.54e-01 |
Pos logit contributions | |
mie | +0.289 |
rified | +0.288 |
etsy | +0.285 |
elope | +0.282 |
jo | +0.277 |
Neg logit contributions | |
trick | -0.111 |
show | -0.106 |
test | -0.097 |
magic | -0.093 |
grown | -0.093 |
Token | bit |
Token ID | 1643 |
Feature activation | +2.24e-01 |
Pos logit contributions | |
trick | +0.244 |
show | +0.232 |
new | +0.226 |
test | +0.221 |
magic | +0.214 |
Neg logit contributions | |
mie | -0.265 |
rified | -0.246 |
elope | -0.244 |
etsy | -0.239 |
hew | -0.233 |
Token | into |
Token ID | 656 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
show | +0.273 |
trick | +0.263 |
new | +0.257 |
to | +0.251 |
for | +0.248 |
Neg logit contributions | |
mie | -0.461 |
elope | -0.459 |
jo | -0.455 |
etsy | -0.446 |
rified | -0.441 |
Token | it |
Token ID | 340 |
Feature activation | -8.71e-02 |
Pos logit contributions | |
show | +0.202 |
trick | +0.199 |
new | +0.167 |
test | +0.162 |
grown | +0.155 |
Neg logit contributions | |
mie | -0.793 |
rified | -0.773 |
elope | -0.771 |
etsy | -0.765 |
jo | -0.761 |
Token | . |
Token ID | 13 |
Feature activation | +2.68e-02 |
Pos logit contributions | |
mie | +0.153 |
rified | +0.151 |
elope | +0.149 |
etsy | +0.146 |
storms | +0.145 |
Neg logit contributions | |
trick | -0.114 |
show | -0.112 |
test | -0.108 |
talent | -0.103 |
grown | -0.103 |
Token | It |
Token ID | 632 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
show | +0.046 |
new | +0.044 |
trick | +0.044 |
to | +0.044 |
grown | +0.043 |
Neg logit contributions | |
mie | -0.042 |
jo | -0.041 |
elope | -0.040 |
rified | -0.039 |
hew | -0.039 |
Token | was |
Token ID | 373 |
Feature activation | +1.62e-01 |
Pos logit contributions | |
mie | +0.265 |
elope | +0.257 |
rified | +0.254 |
etsy | +0.249 |
jo | +0.247 |
Neg logit contributions | |
show | -0.260 |
trick | -0.258 |
new | -0.243 |
test | -0.240 |
grown | -0.235 |
Token | sweet |
Token ID | 6029 |
Feature activation | +3.06e-02 |
Pos logit contributions | |
test | +0.191 |
show | +0.185 |
trick | +0.184 |
design | +0.167 |
talent | +0.163 |
Neg logit contributions | |
elope | -0.291 |
mie | -0.290 |
etsy | -0.283 |
ppy | -0.282 |
rified | -0.276 |
Token | the |
Token ID | 262 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
show | +0.092 |
trick | +0.088 |
new | +0.067 |
test | +0.058 |
grown | +0.055 |
Neg logit contributions | |
mie | -0.683 |
elope | -0.667 |
rified | -0.667 |
etsy | -0.663 |
jo | -0.662 |
Token | ground |
Token ID | 2323 |
Feature activation | -6.58e-02 |
Pos logit contributions | |
trick | +0.234 |
show | +0.230 |
new | +0.214 |
for | +0.207 |
test | +0.206 |
Neg logit contributions | |
mie | -0.320 |
elope | -0.313 |
rified | -0.311 |
jo | -0.305 |
etsy | -0.298 |
Token | . |
Token ID | 13 |
Feature activation | +2.55e-02 |
Pos logit contributions | |
rified | +0.125 |
mie | +0.123 |
elope | +0.122 |
storms | +0.121 |
jo | +0.121 |
Neg logit contributions | |
show | -0.082 |
trick | -0.079 |
test | -0.076 |
new | -0.074 |
job | -0.073 |
Token | It |
Token ID | 632 |
Feature activation | -1.68e-01 |
Pos logit contributions | |
trick | +0.043 |
show | +0.043 |
new | +0.040 |
test | +0.040 |
for | +0.039 |
Neg logit contributions | |
mie | -0.038 |
rified | -0.038 |
elope | -0.037 |
jo | -0.036 |
etsy | -0.036 |
Token | breaks |
Token ID | 9457 |
Feature activation | +1.81e-01 |
Pos logit contributions | |
bo | +0.296 |
rified | +0.292 |
storms | +0.280 |
mie | +0.275 |
elope | +0.274 |
Neg logit contributions | |
trick | -0.212 |
show | -0.208 |
job | -0.206 |
grown | -0.205 |
style | -0.199 |
Token | into |
Token ID | 656 |
Feature activation | +3.06e-01 |
Pos logit contributions | |
show | +0.157 |
trick | +0.152 |
test | +0.133 |
new | +0.132 |
grown | +0.124 |
Neg logit contributions | |
mie | -0.433 |
rified | -0.424 |
elope | -0.420 |
jo | -0.416 |
etsy | -0.413 |
Token | pieces |
Token ID | 5207 |
Feature activation | -3.97e-02 |
Pos logit contributions | |
show | +0.279 |
trick | +0.274 |
new | +0.251 |
grown | +0.237 |
test | +0.235 |
Neg logit contributions | |
mie | -0.719 |
rified | -0.698 |
elope | -0.697 |
etsy | -0.696 |
jo | -0.693 |
Token | . |
Token ID | 13 |
Feature activation | +2.68e-02 |
Pos logit contributions | |
mie | +0.082 |
rified | +0.079 |
jo | +0.079 |
elope | +0.079 |
hew | +0.078 |
Neg logit contributions | |
show | -0.044 |
trick | -0.044 |
new | -0.040 |
test | -0.040 |
grown | -0.038 |
Token | The |
Token ID | 383 |
Feature activation | +6.21e-02 |
Pos logit contributions | |
show | +0.042 |
trick | +0.042 |
new | +0.039 |
test | +0.039 |
grown | +0.038 |
Neg logit contributions | |
mie | -0.045 |
rified | -0.043 |
elope | -0.043 |
jo | -0.042 |
etsy | -0.042 |
Token | man |
Token ID | 582 |
Feature activation | -2.57e-01 |
Pos logit contributions | |
trick | +0.086 |
show | +0.084 |
new | +0.079 |
my | +0.077 |
for | +0.076 |
Neg logit contributions | |
mie | -0.109 |
rified | -0.107 |
elope | -0.106 |
etsy | -0.103 |
jo | -0.102 |
Token | is |
Token ID | 318 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
mie | +0.462 |
rified | +0.449 |
elope | +0.445 |
hew | +0.436 |
jo | +0.435 |
Neg logit contributions | |
trick | -0.361 |
show | -0.358 |
new | -0.342 |
test | -0.328 |
grown | -0.328 |
Token | tried |
Token ID | 3088 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
mie | +0.397 |
elope | +0.382 |
rified | +0.382 |
etsy | +0.377 |
jo | +0.377 |
Neg logit contributions | |
show | -0.306 |
trick | -0.302 |
new | -0.283 |
test | -0.276 |
grown | -0.272 |
Token | to |
Token ID | 284 |
Feature activation | +2.75e-01 |
Pos logit contributions | |
show | +0.135 |
test | +0.113 |
trick | +0.112 |
job | +0.106 |
design | +0.095 |
Neg logit contributions | |
mie | -0.448 |
elope | -0.421 |
etsy | -0.417 |
hew | -0.417 |
rified | -0.415 |
Token | fit |
Token ID | 4197 |
Feature activation | +1.68e-01 |
Pos logit contributions | |
new | +0.274 |
grown | +0.253 |
job | +0.253 |
for | +0.253 |
math | +0.250 |
Neg logit contributions | |
mie | -0.566 |
elope | -0.544 |
etsy | -0.532 |
ppy | -0.529 |
Max | -0.528 |
Token | the |
Token ID | 262 |
Feature activation | +1.81e-01 |
Pos logit contributions | |
show | +0.164 |
trick | +0.161 |
new | +0.147 |
test | +0.140 |
grown | +0.138 |
Neg logit contributions | |
mie | -0.382 |
elope | -0.373 |
rified | -0.371 |
jo | -0.369 |
etsy | -0.369 |
Token | wire |
Token ID | 6503 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
show | +0.210 |
trick | +0.206 |
test | +0.184 |
new | +0.180 |
grown | +0.179 |
Neg logit contributions | |
mie | -0.361 |
rified | -0.354 |
hew | -0.350 |
elope | -0.348 |
etsy | -0.346 |
Token | into |
Token ID | 656 |
Feature activation | +3.06e-01 |
Pos logit contributions | |
mie | +0.291 |
rified | +0.286 |
elope | +0.284 |
jo | +0.279 |
etsy | +0.279 |
Neg logit contributions | |
show | -0.095 |
trick | -0.091 |
grown | -0.081 |
test | -0.080 |
new | -0.079 |
Token | the |
Token ID | 262 |
Feature activation | +1.81e-01 |
Pos logit contributions | |
show | +0.174 |
trick | +0.167 |
new | +0.134 |
test | +0.133 |
grown | +0.121 |
Neg logit contributions | |
mie | -0.817 |
elope | -0.798 |
rified | -0.797 |
etsy | -0.785 |
jo | -0.785 |
Token | can |
Token ID | 460 |
Feature activation | +7.42e-02 |
Pos logit contributions | |
show | +0.231 |
trick | +0.227 |
new | +0.212 |
test | +0.206 |
grown | +0.203 |
Neg logit contributions | |
mie | -0.364 |
rified | -0.352 |
elope | -0.352 |
jo | -0.350 |
etsy | -0.349 |
Token | . |
Token ID | 13 |
Feature activation | +2.28e-02 |
Pos logit contributions | |
show | +0.073 |
trick | +0.069 |
new | +0.064 |
test | +0.064 |
grown | +0.063 |
Neg logit contributions | |
mie | -0.159 |
elope | -0.157 |
jo | -0.156 |
etsy | -0.154 |
rified | -0.154 |
Token |
|
Token ID | 198 |
Feature activation | -7.08e-02 |
Pos logit contributions | |
show | +0.037 |
trick | +0.036 |
new | +0.035 |
grown | +0.034 |
to | +0.034 |
Neg logit contributions | |
mie | -0.038 |
jo | -0.036 |
elope | -0.036 |
rified | -0.036 |
etsy | -0.035 |
Token |
|
Token ID | 198 |
Feature activation | -7.04e-02 |
Pos logit contributions | |
rified | +0.141 |
mie | +0.140 |
storms | +0.135 |
elope | +0.132 |
thel | +0.131 |
Neg logit contributions | |
new | -0.094 |
show | -0.093 |
trick | -0.092 |
to | -0.091 |
grown | -0.087 |
Token | Max |
Token ID | 5436 |
Feature activation | -3.11e-01 |
Pos logit contributions | |
lor | +0.050 |
hew | +0.047 |
etsy | +0.046 |
jo | +0.046 |
ppa | +0.046 |
Neg logit contributions | |
show | -0.042 |
the | -0.041 |
new | -0.040 |
to | -0.037 |
my | -0.037 |
Token | said |
Token ID | 531 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
mie | +0.512 |
rified | +0.508 |
elope | +0.505 |
hew | +0.502 |
etsy | +0.498 |
Neg logit contributions | |
show | -0.505 |
trick | -0.497 |
new | -0.454 |
for | -0.451 |
to | -0.451 |
Token | . |
Token ID | 13 |
Feature activation | +2.41e-02 |
Pos logit contributions | |
mie | +0.028 |
elope | +0.027 |
rified | +0.027 |
etsy | +0.027 |
jo | +0.027 |
Neg logit contributions | |
show | -0.015 |
trick | -0.015 |
test | -0.013 |
new | -0.013 |
grown | -0.013 |
Token | He |
Token ID | 679 |
Feature activation | -2.13e-01 |
Pos logit contributions | |
show | +0.037 |
new | +0.036 |
to | +0.036 |
trick | +0.035 |
for | +0.034 |
Neg logit contributions | |
mie | -0.043 |
rified | -0.042 |
hew | -0.042 |
elope | -0.041 |
thel | -0.041 |
Token | climbed |
Token ID | 19952 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
mie | +0.390 |
elope | +0.371 |
rified | +0.368 |
etsy | +0.366 |
jo | +0.365 |
Neg logit contributions | |
show | -0.300 |
trick | -0.298 |
new | -0.282 |
test | -0.276 |
grown | -0.266 |
Token | into |
Token ID | 656 |
Feature activation | +3.06e-01 |
Pos logit contributions | |
trick | +0.121 |
show | +0.117 |
test | +0.108 |
grown | +0.107 |
job | +0.096 |
Neg logit contributions | |
mie | -0.316 |
elope | -0.304 |
rified | -0.302 |
jo | -0.301 |
hel | -0.297 |
Token | the |
Token ID | 262 |
Feature activation | +1.84e-01 |
Pos logit contributions | |
trick | +0.151 |
show | +0.143 |
test | +0.113 |
job | +0.105 |
grown | +0.102 |
Neg logit contributions | |
mie | -0.818 |
elope | -0.804 |
rified | -0.793 |
jo | -0.790 |
Max | -0.770 |
Token | box |
Token ID | 3091 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
show | +0.273 |
trick | +0.267 |
new | +0.254 |
test | +0.246 |
grown | +0.243 |
Neg logit contributions | |
mie | -0.334 |
etsy | -0.322 |
rified | -0.321 |
elope | -0.321 |
jo | -0.317 |
Token | and |
Token ID | 290 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
jo | +0.364 |
mie | +0.364 |
rified | +0.361 |
elope | +0.360 |
hel | +0.348 |
Neg logit contributions | |
trick | -0.172 |
show | -0.167 |
test | -0.165 |
my | -0.158 |
your | -0.154 |
Token | pretended |
Token ID | 36766 |
Feature activation | +7.65e-02 |
Pos logit contributions | |
trick | +0.220 |
show | +0.213 |
new | +0.205 |
test | +0.200 |
grown | +0.196 |
Neg logit contributions | |
mie | -0.285 |
elope | -0.269 |
rified | -0.269 |
jo | -0.260 |
etsy | -0.260 |
Token | to |
Token ID | 284 |
Feature activation | +2.79e-01 |
Pos logit contributions | |
show | +0.051 |
trick | +0.050 |
new | +0.043 |
test | +0.040 |
grown | +0.039 |
Neg logit contributions | |
mie | -0.197 |
rified | -0.192 |
elope | -0.192 |
etsy | -0.191 |
jo | -0.191 |
Token | wanted |
Token ID | 2227 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
mie | +0.361 |
rified | +0.349 |
elope | +0.348 |
jo | +0.348 |
etsy | +0.347 |
Neg logit contributions | |
show | -0.298 |
trick | -0.290 |
new | -0.275 |
grown | -0.268 |
test | -0.265 |
Token | to |
Token ID | 284 |
Feature activation | +2.77e-01 |
Pos logit contributions | |
show | +0.089 |
design | +0.084 |
trick | +0.082 |
style | +0.077 |
test | +0.075 |
Neg logit contributions | |
elope | -0.282 |
etsy | -0.277 |
uddy | -0.275 |
mie | -0.275 |
Rex | -0.273 |
Token | change |
Token ID | 1487 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
new | +0.262 |
for | +0.253 |
matching | +0.246 |
magic | +0.245 |
making | +0.235 |
Neg logit contributions | |
elope | -0.562 |
mie | -0.539 |
etsy | -0.536 |
uddy | -0.530 |
Bird | -0.528 |
Token | the |
Token ID | 262 |
Feature activation | +1.85e-01 |
Pos logit contributions | |
show | +0.074 |
trick | +0.073 |
test | +0.063 |
grown | +0.059 |
design | +0.055 |
Neg logit contributions | |
mie | -0.270 |
elope | -0.266 |
rified | -0.265 |
jo | -0.259 |
hew | -0.255 |
Token | park |
Token ID | 3952 |
Feature activation | -7.21e-02 |
Pos logit contributions | |
show | +0.216 |
trick | +0.207 |
five | +0.206 |
making | +0.202 |
for | +0.197 |
Neg logit contributions | |
pillar | -0.340 |
rified | -0.335 |
mie | -0.327 |
elope | -0.316 |
etsy | -0.312 |
Token | into |
Token ID | 656 |
Feature activation | +3.06e-01 |
Pos logit contributions | |
mie | +0.167 |
elope | +0.163 |
rified | +0.162 |
jo | +0.161 |
etsy | +0.160 |
Neg logit contributions | |
show | -0.062 |
trick | -0.061 |
test | -0.053 |
grown | -0.053 |
new | -0.052 |
Token | his |
Token ID | 465 |
Feature activation | +9.73e-02 |
Pos logit contributions | |
trick | +0.150 |
show | +0.147 |
test | +0.108 |
new | +0.103 |
for | +0.095 |
Neg logit contributions | |
mie | -0.822 |
elope | -0.820 |
rified | -0.810 |
jo | -0.789 |
hew | -0.786 |
Token | own |
Token ID | 898 |
Feature activation | +1.75e-01 |
Pos logit contributions | |
trick | +0.106 |
show | +0.102 |
making | +0.096 |
test | +0.093 |
five | +0.092 |
Neg logit contributions | |
elope | -0.204 |
mie | -0.195 |
etsy | -0.187 |
jo | -0.187 |
hew | -0.186 |
Token | home |
Token ID | 1363 |
Feature activation | +3.19e-02 |
Pos logit contributions | |
show | +0.174 |
trick | +0.173 |
making | +0.158 |
test | +0.156 |
five | +0.153 |
Neg logit contributions | |
elope | -0.372 |
mie | -0.369 |
jo | -0.359 |
hew | -0.357 |
rified | -0.355 |
Token | . |
Token ID | 13 |
Feature activation | +1.94e-02 |
Pos logit contributions | |
trick | +0.037 |
show | +0.037 |
test | +0.033 |
new | +0.032 |
grown | +0.032 |
Neg logit contributions | |
mie | -0.064 |
elope | -0.063 |
rified | -0.062 |
jo | -0.061 |
etsy | -0.061 |
Token | He |
Token ID | 679 |
Feature activation | -2.14e-01 |
Pos logit contributions | |
trick | +0.035 |
show | +0.035 |
test | +0.033 |
new | +0.032 |
magic | +0.032 |
Neg logit contributions | |
elope | -0.027 |
rified | -0.027 |
mie | -0.026 |
etsy | -0.026 |
storms | -0.025 |
Token | asleep |
Token ID | 16039 |
Feature activation | +3.78e-02 |
Pos logit contributions | |
trick | +0.137 |
show | +0.133 |
test | +0.124 |
new | +0.121 |
style | +0.115 |
Neg logit contributions | |
mie | -0.200 |
elope | -0.190 |
rified | -0.185 |
etsy | -0.184 |
hew | -0.183 |
Token | , |
Token ID | 11 |
Feature activation | +1.54e-01 |
Pos logit contributions | |
trick | +0.043 |
show | +0.037 |
test | +0.037 |
style | +0.036 |
new | +0.036 |
Neg logit contributions | |
mie | -0.077 |
elope | -0.076 |
rified | -0.075 |
storms | -0.072 |
jo | -0.071 |
Token | Tom |
Token ID | 4186 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
show | +0.180 |
trick | +0.178 |
new | +0.163 |
test | +0.159 |
grown | +0.156 |
Neg logit contributions | |
mie | -0.319 |
elope | -0.310 |
rified | -0.309 |
jo | -0.305 |
etsy | -0.305 |
Token | sne |
Token ID | 10505 |
Feature activation | +8.90e-02 |
Pos logit contributions | |
mie | +0.515 |
rified | +0.499 |
elope | +0.498 |
etsy | +0.495 |
jo | +0.493 |
Neg logit contributions | |
show | -0.338 |
trick | -0.332 |
new | -0.310 |
test | -0.300 |
grown | -0.298 |
Token | aks |
Token ID | 4730 |
Feature activation | -4.83e-02 |
Pos logit contributions | |
trick | +0.153 |
show | +0.153 |
grown | +0.141 |
for | +0.141 |
new | +0.140 |
Neg logit contributions | |
mie | -0.135 |
etsy | -0.130 |
rified | -0.129 |
elope | -0.126 |
storms | -0.123 |
Token | into |
Token ID | 656 |
Feature activation | +3.06e-01 |
Pos logit contributions | |
mie | +0.116 |
rified | +0.114 |
elope | +0.112 |
jo | +0.112 |
etsy | +0.111 |
Neg logit contributions | |
trick | -0.039 |
show | -0.039 |
test | -0.034 |
new | -0.033 |
grown | -0.032 |
Token | their |
Token ID | 511 |
Feature activation | +6.68e-02 |
Pos logit contributions | |
trick | +0.238 |
test | +0.213 |
job | +0.211 |
show | +0.197 |
style | +0.188 |
Neg logit contributions | |
elope | -0.703 |
rified | -0.656 |
Bear | -0.655 |
mie | -0.653 |
jo | -0.651 |
Token | gardens |
Token ID | 25476 |
Feature activation | -7.01e-02 |
Pos logit contributions | |
trick | +0.104 |
show | +0.097 |
test | +0.092 |
five | +0.092 |
grown | +0.089 |
Neg logit contributions | |
elope | -0.112 |
mie | -0.103 |
etsy | -0.098 |
jo | -0.098 |
lves | -0.095 |
Token | . |
Token ID | 13 |
Feature activation | +2.84e-02 |
Pos logit contributions | |
elope | +0.143 |
mie | +0.140 |
jo | +0.136 |
rified | +0.134 |
storms | +0.132 |
Neg logit contributions | |
trick | -0.079 |
test | -0.073 |
show | -0.072 |
magic | -0.066 |
job | -0.066 |
Token | He |
Token ID | 679 |
Feature activation | -2.13e-01 |
Pos logit contributions | |
show | +0.043 |
trick | +0.042 |
new | +0.042 |
grown | +0.042 |
to | +0.041 |
Neg logit contributions | |
mie | -0.051 |
rified | -0.047 |
ppa | -0.047 |
hew | -0.047 |
etsy | -0.047 |
Token | finds |
Token ID | 7228 |
Feature activation | +1.68e-01 |
Pos logit contributions | |
mie | +0.383 |
elope | +0.373 |
rified | +0.355 |
jo | +0.353 |
hew | +0.351 |
Neg logit contributions | |
trick | -0.302 |
show | -0.279 |
job | -0.274 |
new | -0.274 |
test | -0.266 |
Token | were |
Token ID | 547 |
Feature activation | +8.96e-02 |
Pos logit contributions | |
jo | +0.303 |
mie | +0.298 |
elope | +0.295 |
storms | +0.293 |
ksh | +0.291 |
Neg logit contributions | |
show | -0.330 |
trick | -0.314 |
grown | -0.304 |
to | -0.303 |
making | -0.299 |
Token | lost |
Token ID | 2626 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
show | +0.118 |
trick | +0.117 |
design | +0.112 |
test | +0.110 |
job | +0.106 |
Neg logit contributions | |
mie | -0.169 |
elope | -0.157 |
ppy | -0.156 |
rified | -0.155 |
Rex | -0.150 |
Token | . |
Token ID | 13 |
Feature activation | +2.33e-02 |
Pos logit contributions | |
show | +0.085 |
trick | +0.083 |
new | +0.075 |
test | +0.070 |
grown | +0.069 |
Neg logit contributions | |
mie | -0.246 |
elope | -0.240 |
rified | -0.239 |
etsy | -0.239 |
jo | -0.239 |
Token | They |
Token ID | 1119 |
Feature activation | -2.14e-01 |
Pos logit contributions | |
new | +0.029 |
to | +0.029 |
the | +0.029 |
show | +0.028 |
grown | +0.028 |
Neg logit contributions | |
mie | -0.050 |
thel | -0.048 |
ppa | -0.047 |
jo | -0.045 |
pillar | -0.044 |
Token | ran |
Token ID | 4966 |
Feature activation | +1.03e-01 |
Pos logit contributions | |
mie | +0.357 |
jo | +0.348 |
lor | +0.348 |
etsy | +0.347 |
hew | +0.345 |
Neg logit contributions | |
show | -0.362 |
trick | -0.338 |
grown | -0.328 |
making | -0.321 |
the | -0.321 |
Token | into |
Token ID | 656 |
Feature activation | +3.06e-01 |
Pos logit contributions | |
show | +0.075 |
trick | +0.073 |
new | +0.063 |
test | +0.060 |
grown | +0.059 |
Neg logit contributions | |
mie | -0.259 |
rified | -0.252 |
elope | -0.252 |
etsy | -0.250 |
jo | -0.250 |
Token | a |
Token ID | 257 |
Feature activation | +1.70e-01 |
Pos logit contributions | |
show | +0.110 |
trick | +0.107 |
new | +0.074 |
test | +0.070 |
grown | +0.063 |
Neg logit contributions | |
mie | -0.877 |
elope | -0.861 |
rified | -0.858 |
jo | -0.850 |
etsy | -0.846 |
Token | dead |
Token ID | 2636 |
Feature activation | +9.35e-02 |
Pos logit contributions | |
show | +0.154 |
trick | +0.150 |
new | +0.140 |
grown | +0.131 |
test | +0.130 |
Neg logit contributions | |
mie | -0.404 |
rified | -0.394 |
etsy | -0.390 |
elope | -0.389 |
jo | -0.388 |
Token | end |
Token ID | 886 |
Feature activation | +9.56e-02 |
Pos logit contributions | |
trick | +0.130 |
show | +0.130 |
new | +0.119 |
test | +0.118 |
making | +0.117 |
Neg logit contributions | |
elope | -0.160 |
mie | -0.160 |
jo | -0.156 |
rified | -0.155 |
pillar | -0.153 |
Token | . |
Token ID | 13 |
Feature activation | +2.44e-02 |
Pos logit contributions | |
show | +0.086 |
trick | +0.084 |
new | +0.077 |
for | +0.071 |
test | +0.071 |
Neg logit contributions | |
mie | -0.230 |
rified | -0.224 |
etsy | -0.222 |
elope | -0.220 |
hew | -0.218 |
Token | They |
Token ID | 1119 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
show | +0.033 |
new | +0.033 |
to | +0.032 |
trick | +0.032 |
grown | +0.031 |
Neg logit contributions | |
mie | -0.050 |
thel | -0.047 |
ppa | -0.046 |
jo | -0.045 |
rified | -0.045 |
Token | juice |
Token ID | 13135 |
Feature activation | -6.35e-02 |
Pos logit contributions | |
trick | +0.239 |
meeting | +0.215 |
test | +0.214 |
show | +0.210 |
spelling | +0.209 |
Neg logit contributions | |
als | -0.247 |
mie | -0.244 |
pillar | -0.244 |
Bear | -0.240 |
rified | -0.239 |
Token | from |
Token ID | 422 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
bo | +0.123 |
ppy | +0.119 |
rified | +0.119 |
als | +0.118 |
mie | +0.118 |
Neg logit contributions | |
trick | -0.066 |
test | -0.066 |
show | -0.066 |
talent | -0.062 |
job | -0.062 |
Token | a |
Token ID | 257 |
Feature activation | +1.68e-01 |
Pos logit contributions | |
show | +0.194 |
trick | +0.187 |
new | +0.171 |
test | +0.159 |
grown | +0.155 |
Neg logit contributions | |
mie | -0.555 |
elope | -0.540 |
jo | -0.540 |
etsy | -0.539 |
rified | -0.538 |
Token | plastic |
Token ID | 7309 |
Feature activation | -3.66e-03 |
Pos logit contributions | |
trick | +0.214 |
show | +0.210 |
work | +0.192 |
test | +0.191 |
making | +0.188 |
Neg logit contributions | |
mie | -0.308 |
etsy | -0.305 |
och | -0.303 |
pillar | -0.302 |
elope | -0.293 |
Token | pitcher |
Token ID | 18086 |
Feature activation | -6.63e-02 |
Pos logit contributions | |
pillar | +0.006 |
etsy | +0.006 |
ley | +0.006 |
och | +0.006 |
rified | +0.006 |
Neg logit contributions | |
show | -0.005 |
trick | -0.005 |
new | -0.005 |
my | -0.005 |
test | -0.004 |
Token | into |
Token ID | 656 |
Feature activation | +3.06e-01 |
Pos logit contributions | |
och | +0.121 |
ppy | +0.121 |
bo | +0.121 |
mie | +0.119 |
rified | +0.119 |
Neg logit contributions | |
trick | -0.079 |
test | -0.076 |
show | -0.076 |
new | -0.074 |
job | -0.071 |
Token | a |
Token ID | 257 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
trick | +0.277 |
show | +0.259 |
test | +0.235 |
job | +0.225 |
new | +0.214 |
Neg logit contributions | |
mie | -0.691 |
elope | -0.676 |
rified | -0.676 |
Max | -0.656 |
Bear | -0.653 |
Token | plastic |
Token ID | 7309 |
Feature activation | -2.61e-03 |
Pos logit contributions | |
trick | +0.220 |
show | +0.209 |
work | +0.198 |
test | +0.192 |
power | +0.191 |
Neg logit contributions | |
mie | -0.294 |
pillar | -0.294 |
och | -0.292 |
etsy | -0.291 |
elope | -0.287 |
Token | cup |
Token ID | 6508 |
Feature activation | -3.34e-02 |
Pos logit contributions | |
etsy | +0.004 |
pillar | +0.004 |
ley | +0.004 |
och | +0.004 |
elope | +0.004 |
Neg logit contributions | |
trick | -0.004 |
show | -0.003 |
my | -0.003 |
new | -0.003 |
test | -0.003 |
Token | and |
Token ID | 290 |
Feature activation | +1.52e-01 |
Pos logit contributions | |
och | +0.060 |
rified | +0.058 |
pillar | +0.057 |
mie | +0.057 |
storms | +0.057 |
Neg logit contributions | |
trick | -0.044 |
new | -0.041 |
show | -0.041 |
test | -0.040 |
job | -0.039 |
Token | gives |
Token ID | 3607 |
Feature activation | +2.26e-01 |
Pos logit contributions | |
trick | +0.201 |
job | +0.171 |
chance | +0.167 |
test | +0.166 |
new | +0.165 |
Neg logit contributions | |
mie | -0.279 |
elope | -0.254 |
rified | -0.249 |
hew | -0.249 |
ksh | -0.248 |
Token | 's |
Token ID | 338 |
Feature activation | +6.22e-02 |
Pos logit contributions | |
mie | +0.396 |
rified | +0.395 |
etsy | +0.391 |
elope | +0.386 |
och | +0.384 |
Neg logit contributions | |
show | -0.280 |
trick | -0.265 |
grown | -0.253 |
new | -0.249 |
to | -0.247 |
Token | big |
Token ID | 1263 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
show | +0.077 |
trick | +0.075 |
new | +0.070 |
test | +0.068 |
grown | +0.067 |
Neg logit contributions | |
mie | -0.126 |
rified | -0.122 |
elope | -0.122 |
jo | -0.121 |
etsy | -0.121 |
Token | brother |
Token ID | 3956 |
Feature activation | -2.35e-01 |
Pos logit contributions | |
trick | +0.258 |
show | +0.252 |
grown | +0.236 |
new | +0.234 |
test | +0.233 |
Neg logit contributions | |
mie | -0.221 |
elope | -0.213 |
jo | -0.210 |
rified | -0.209 |
storms | -0.203 |
Token | Tom |
Token ID | 4186 |
Feature activation | -2.61e-01 |
Pos logit contributions | |
hew | +0.433 |
mie | +0.433 |
rified | +0.427 |
etsy | +0.427 |
elope | +0.425 |
Neg logit contributions | |
show | -0.330 |
trick | -0.313 |
to | -0.304 |
grown | -0.303 |
new | -0.297 |
Token | came |
Token ID | 1625 |
Feature activation | +1.36e-01 |
Pos logit contributions | |
mie | +0.487 |
rified | +0.477 |
elope | +0.475 |
etsy | +0.474 |
jo | +0.466 |
Neg logit contributions | |
show | -0.360 |
trick | -0.349 |
new | -0.327 |
grown | -0.322 |
for | -0.320 |
Token | into |
Token ID | 656 |
Feature activation | +3.06e-01 |
Pos logit contributions | |
trick | +0.115 |
show | +0.115 |
test | +0.105 |
magic | +0.103 |
job | +0.101 |
Neg logit contributions | |
mie | -0.300 |
elope | -0.299 |
jo | -0.298 |
ppy | -0.296 |
bo | -0.292 |
Token | her |
Token ID | 607 |
Feature activation | +8.16e-02 |
Pos logit contributions | |
trick | +0.225 |
show | +0.210 |
test | +0.191 |
job | +0.179 |
new | +0.176 |
Neg logit contributions | |
elope | -0.709 |
Max | -0.708 |
mie | -0.704 |
Bear | -0.684 |
Mic | -0.680 |
Token | room |
Token ID | 2119 |
Feature activation | -8.49e-02 |
Pos logit contributions | |
trick | +0.119 |
show | +0.107 |
grown | +0.105 |
five | +0.103 |
test | +0.102 |
Neg logit contributions | |
elope | -0.147 |
mie | -0.136 |
jo | -0.132 |
etsy | -0.125 |
uffy | -0.122 |
Token | and |
Token ID | 290 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
elope | +0.140 |
storms | +0.138 |
jo | +0.138 |
rified | +0.138 |
mie | +0.137 |
Neg logit contributions | |
trick | -0.118 |
show | -0.109 |
my | -0.109 |
new | -0.108 |
magic | -0.107 |
Token | saw |
Token ID | 2497 |
Feature activation | +1.74e-01 |
Pos logit contributions | |
show | +0.235 |
trick | +0.229 |
new | +0.217 |
test | +0.211 |
grown | +0.210 |
Neg logit contributions | |
mie | -0.260 |
rified | -0.252 |
elope | -0.251 |
etsy | -0.249 |
jo | -0.248 |
Token | her |
Token ID | 607 |
Feature activation | +8.24e-02 |
Pos logit contributions | |
show | +0.168 |
trick | +0.164 |
new | +0.151 |
test | +0.142 |
grown | +0.141 |
Neg logit contributions | |
mie | -0.404 |
elope | -0.394 |
rified | -0.392 |
etsy | -0.390 |
jo | -0.390 |
Token | and |
Token ID | 290 |
Feature activation | +1.48e-01 |
Pos logit contributions | |
mie | +0.322 |
rified | +0.317 |
elope | +0.313 |
jo | +0.310 |
etsy | +0.308 |
Neg logit contributions | |
show | -0.151 |
trick | -0.151 |
new | -0.137 |
test | -0.134 |
grown | -0.130 |
Token | was |
Token ID | 373 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
trick | +0.224 |
show | +0.220 |
test | +0.208 |
new | +0.207 |
style | +0.197 |
Neg logit contributions | |
mie | -0.253 |
rified | -0.244 |
elope | -0.243 |
etsy | -0.235 |
hew | -0.233 |
Token | able |
Token ID | 1498 |
Feature activation | +1.20e-01 |
Pos logit contributions | |
trick | +0.227 |
test | +0.221 |
show | +0.220 |
style | +0.206 |
design | +0.204 |
Neg logit contributions | |
mie | -0.257 |
elope | -0.255 |
rified | -0.237 |
etsy | -0.226 |
lves | -0.222 |
Token | to |
Token ID | 284 |
Feature activation | +2.76e-01 |
Pos logit contributions | |
show | +0.064 |
trick | +0.058 |
test | +0.052 |
design | +0.042 |
style | +0.040 |
Neg logit contributions | |
mie | -0.313 |
etsy | -0.305 |
elope | -0.302 |
rified | -0.302 |
hew | -0.299 |
Token | get |
Token ID | 651 |
Feature activation | +2.09e-01 |
Pos logit contributions | |
new | +0.280 |
red | +0.261 |
style | +0.258 |
talent | +0.257 |
magic | +0.253 |
Neg logit contributions | |
elope | -0.574 |
mie | -0.555 |
etsy | -0.543 |
ppy | -0.532 |
hew | -0.527 |
Token | into |
Token ID | 656 |
Feature activation | +3.06e-01 |
Pos logit contributions | |
trick | +0.218 |
show | +0.207 |
test | +0.193 |
style | +0.187 |
new | +0.180 |
Neg logit contributions | |
mie | -0.450 |
rified | -0.433 |
elope | -0.424 |
etsy | -0.424 |
hew | -0.423 |
Token | the |
Token ID | 262 |
Feature activation | +1.82e-01 |
Pos logit contributions | |
trick | +0.320 |
show | +0.317 |
test | +0.294 |
style | +0.272 |
grown | +0.269 |
Neg logit contributions | |
mie | -0.644 |
elope | -0.636 |
rified | -0.619 |
hew | -0.597 |
etsy | -0.591 |
Token | garden |
Token ID | 11376 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
show | +0.251 |
trick | +0.244 |
new | +0.239 |
grown | +0.226 |
test | +0.224 |
Neg logit contributions | |
mie | -0.360 |
jo | -0.348 |
etsy | -0.341 |
rified | -0.337 |
elope | -0.333 |
Token | . |
Token ID | 13 |
Feature activation | +1.70e-02 |
Pos logit contributions | |
mie | +0.284 |
elope | +0.276 |
rified | +0.276 |
jo | +0.272 |
etsy | +0.272 |
Neg logit contributions | |
show | -0.122 |
trick | -0.121 |
new | -0.109 |
test | -0.106 |
grown | -0.104 |
Token | He |
Token ID | 679 |
Feature activation | -2.17e-01 |
Pos logit contributions | |
new | +0.026 |
show | +0.025 |
to | +0.025 |
for | +0.024 |
grown | +0.023 |
Neg logit contributions | |
mie | -0.031 |
la | -0.030 |
jo | -0.030 |
elope | -0.030 |
thel | -0.030 |
Token | got |
Token ID | 1392 |
Feature activation | +2.33e-01 |
Pos logit contributions | |
mie | +0.377 |
elope | +0.366 |
rified | +0.359 |
etsy | +0.356 |
jo | +0.354 |
Neg logit contributions | |
show | -0.326 |
trick | -0.324 |
new | -0.305 |
test | -0.303 |
grown | -0.291 |
Token | angry |
Token ID | 7954 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
trick | +0.248 |
show | +0.246 |
test | +0.228 |
new | +0.218 |
work | +0.214 |
Neg logit contributions | |
mie | -0.324 |
elope | -0.316 |
etsy | -0.310 |
rified | -0.304 |
jo | -0.301 |
Token | that |
Token ID | 326 |
Feature activation | +4.27e-02 |
Pos logit contributions | |
show | +0.101 |
trick | +0.097 |
for | +0.092 |
new | +0.091 |
to | +0.089 |
Neg logit contributions | |
mie | -0.309 |
elope | -0.305 |
etsy | -0.305 |
jo | -0.299 |
hew | -0.298 |
Token | he |
Token ID | 339 |
Feature activation | -1.96e-01 |
Pos logit contributions | |
show | +0.068 |
new | +0.064 |
the | +0.063 |
trick | +0.063 |
to | +0.062 |
Neg logit contributions | |
hew | -0.077 |
mie | -0.076 |
etsy | -0.076 |
elope | -0.075 |
rified | -0.072 |
Token | had |
Token ID | 550 |
Feature activation | +7.14e-02 |
Pos logit contributions | |
mie | +0.331 |
elope | +0.323 |
rified | +0.320 |
jo | +0.319 |
etsy | +0.319 |
Neg logit contributions | |
show | -0.310 |
trick | -0.301 |
new | -0.279 |
grown | -0.278 |
test | -0.275 |
Token | come |
Token ID | 1282 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
show | +0.107 |
trick | +0.105 |
test | +0.102 |
job | +0.100 |
new | +0.099 |
Neg logit contributions | |
mie | -0.117 |
etsy | -0.108 |
rified | -0.105 |
elope | -0.105 |
hew | -0.104 |
Token | into |
Token ID | 656 |
Feature activation | +3.06e-01 |
Pos logit contributions | |
show | +0.061 |
trick | +0.060 |
test | +0.050 |
new | +0.047 |
grown | +0.046 |
Neg logit contributions | |
mie | -0.265 |
etsy | -0.258 |
elope | -0.257 |
rified | -0.254 |
jo | -0.253 |
Token | her |
Token ID | 607 |
Feature activation | +8.26e-02 |
Pos logit contributions | |
show | +0.184 |
trick | +0.183 |
test | +0.154 |
new | +0.144 |
grown | +0.134 |
Neg logit contributions | |
mie | -0.771 |
elope | -0.765 |
rified | -0.755 |
jo | -0.746 |
etsy | -0.734 |
Token | garden |
Token ID | 11376 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
trick | +0.111 |
show | +0.106 |
grown | +0.098 |
test | +0.097 |
five | +0.097 |
Neg logit contributions | |
elope | -0.155 |
mie | -0.153 |
etsy | -0.145 |
rified | -0.144 |
jo | -0.144 |
Token | . |
Token ID | 13 |
Feature activation | +1.77e-02 |
Pos logit contributions | |
mie | +0.275 |
elope | +0.269 |
rified | +0.267 |
etsy | +0.262 |
jo | +0.261 |
Neg logit contributions | |
show | -0.125 |
trick | -0.125 |
test | -0.112 |
new | -0.111 |
grown | -0.109 |
Token | She |
Token ID | 1375 |
Feature activation | -2.28e-01 |
Pos logit contributions | |
to | +0.026 |
show | +0.026 |
the | +0.025 |
new | +0.025 |
my | +0.025 |
Neg logit contributions | |
ppy | -0.032 |
la | -0.032 |
mie | -0.032 |
ppa | -0.032 |
lor | -0.032 |
Token | decided |
Token ID | 3066 |
Feature activation | +4.98e-02 |
Pos logit contributions | |
mie | +0.410 |
rified | +0.397 |
elope | +0.395 |
jo | +0.393 |
etsy | +0.392 |
Neg logit contributions | |
show | -0.333 |
trick | -0.326 |
new | -0.306 |
grown | -0.299 |
test | -0.296 |
Token | her |
Token ID | 607 |
Feature activation | -5.16e-02 |
Pos logit contributions | |
trick | +0.115 |
new | +0.105 |
show | +0.103 |
for | +0.096 |
magic | +0.095 |
Neg logit contributions | |
mie | -0.338 |
elope | -0.335 |
jo | -0.320 |
etsy | -0.319 |
rified | -0.312 |
Token | and |
Token ID | 290 |
Feature activation | +2.34e-02 |
Pos logit contributions | |
mie | +0.100 |
elope | +0.098 |
jo | +0.094 |
etsy | +0.093 |
rified | +0.091 |
Neg logit contributions | |
trick | -0.064 |
show | -0.060 |
grown | -0.058 |
test | -0.056 |
magic | -0.055 |
Token | comfort |
Token ID | 4467 |
Feature activation | -2.11e-02 |
Pos logit contributions | |
trick | +0.034 |
show | +0.032 |
new | +0.031 |
test | +0.031 |
grown | +0.030 |
Neg logit contributions | |
mie | -0.042 |
elope | -0.040 |
jo | -0.039 |
etsy | -0.038 |
rified | -0.038 |
Token | ed |
Token ID | 276 |
Feature activation | -9.90e-03 |
Pos logit contributions | |
elope | +0.054 |
mie | +0.053 |
hew | +0.053 |
Mic | +0.053 |
lves | +0.052 |
Neg logit contributions | |
show | -0.014 |
trick | -0.013 |
to | -0.013 |
for | -0.012 |
new | -0.012 |
Token | Pen |
Token ID | 7507 |
Feature activation | -2.84e-01 |
Pos logit contributions | |
mie | +0.023 |
rified | +0.023 |
elope | +0.023 |
jo | +0.023 |
etsy | +0.023 |
Neg logit contributions | |
show | -0.009 |
trick | -0.009 |
new | -0.008 |
test | -0.007 |
grown | -0.007 |
Token | elope |
Token ID | 47329 |
Feature activation | -4.87e-01 |
Pos logit contributions | |
Mic | +0.391 |
och | +0.383 |
lves | +0.376 |
storms | +0.371 |
azz | +0.371 |
Neg logit contributions | |
to | -0.538 |
the | -0.530 |
for | -0.527 |
show | -0.526 |
new | -0.513 |
Token | . |
Token ID | 13 |
Feature activation | -8.77e-02 |
Pos logit contributions | |
mie | +0.955 |
rified | +0.925 |
elope | +0.924 |
jo | +0.915 |
etsy | +0.915 |
Neg logit contributions | |
show | -0.624 |
trick | -0.614 |
new | -0.573 |
test | -0.554 |
grown | -0.549 |
Token | " |
Token ID | 366 |
Feature activation | -1.73e-01 |
Pos logit contributions | |
x | +0.170 |
rified | +0.168 |
lves | +0.167 |
hew | +0.164 |
ppy | +0.164 |
Neg logit contributions | |
my | -0.114 |
the | -0.112 |
to | -0.111 |
your | -0.108 |
show | -0.105 |
Token | I |
Token ID | 40 |
Feature activation | -1.92e-01 |
Pos logit contributions | |
uddy | +0.274 |
isc | +0.268 |
jo | +0.267 |
storms | +0.258 |
ongo | +0.251 |
Neg logit contributions | |
five | -0.267 |
show | -0.266 |
trick | -0.261 |
my | -0.261 |
your | -0.250 |
Token | told |
Token ID | 1297 |
Feature activation | +2.94e-02 |
Pos logit contributions | |
mie | +0.323 |
elope | +0.309 |
etsy | +0.308 |
rified | +0.304 |
jo | +0.304 |
Neg logit contributions | |
trick | -0.294 |
show | -0.293 |
new | -0.279 |
test | -0.271 |
grown | -0.266 |
Token | you |
Token ID | 345 |
Feature activation | -2.23e-01 |
Pos logit contributions | |
show | +0.041 |
my | +0.039 |
trick | +0.039 |
new | +0.038 |
your | +0.037 |
Neg logit contributions | |
mie | -0.056 |
jo | -0.055 |
och | -0.055 |
etsy | -0.054 |
rified | -0.054 |
Token |
|
Token ID | 198 |
Feature activation | -1.87e-01 |
Pos logit contributions | |
rified | +0.255 |
hew | +0.247 |
mie | +0.246 |
lves | +0.244 |
storms | +0.244 |
Neg logit contributions | |
show | -0.205 |
the | -0.199 |
to | -0.198 |
my | -0.197 |
his | -0.193 |
Token |
|
Token ID | 198 |
Feature activation | -1.83e-01 |
Pos logit contributions | |
rified | +0.478 |
storms | +0.456 |
mie | +0.440 |
lves | +0.438 |
ishly | +0.438 |
Neg logit contributions | |
the | -0.181 |
new | -0.177 |
his | -0.174 |
your | -0.165 |
my | -0.162 |
Token | " |
Token ID | 1 |
Feature activation | -1.72e-01 |
Pos logit contributions | |
rified | +0.341 |
storms | +0.334 |
mie | +0.309 |
ksh | +0.298 |
ley | +0.296 |
Neg logit contributions | |
my | -0.272 |
five | -0.270 |
your | -0.270 |
grown | -0.270 |
show | -0.268 |
Token | Who |
Token ID | 8241 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
mie | +0.279 |
jo | +0.279 |
hew | +0.267 |
rified | +0.264 |
isc | +0.256 |
Neg logit contributions | |
trick | -0.287 |
show | -0.286 |
five | -0.280 |
test | -0.271 |
grown | -0.266 |
Token | is |
Token ID | 318 |
Feature activation | +4.37e-02 |
Pos logit contributions | |
mie | +0.245 |
elope | +0.237 |
rified | +0.235 |
jo | +0.234 |
hew | +0.232 |
Neg logit contributions | |
show | -0.202 |
trick | -0.200 |
new | -0.190 |
test | -0.181 |
grown | -0.181 |
Token | Lincoln |
Token ID | 12406 |
Feature activation | -4.80e-01 |
Pos logit contributions | |
show | +0.043 |
trick | +0.043 |
new | +0.038 |
test | +0.037 |
grown | +0.037 |
Neg logit contributions | |
mie | -0.098 |
elope | -0.096 |
rified | -0.095 |
jo | -0.095 |
etsy | -0.094 |
Token | , |
Token ID | 11 |
Feature activation | +5.63e-02 |
Pos logit contributions | |
mie | +1.098 |
jo | +1.056 |
elope | +1.036 |
rified | +1.021 |
hew | +1.014 |
Neg logit contributions | |
trick | -0.446 |
show | -0.433 |
new | -0.422 |
grown | -0.395 |
magic | -0.390 |
Token | Ben |
Token ID | 3932 |
Feature activation | -3.65e-01 |
Pos logit contributions | |
show | +0.078 |
trick | +0.073 |
new | +0.072 |
for | +0.070 |
my | +0.070 |
Neg logit contributions | |
elope | -0.110 |
etsy | -0.109 |
mie | -0.109 |
rified | -0.108 |
jo | -0.107 |
Token | ? |
Token ID | 30 |
Feature activation | -4.11e-02 |
Pos logit contributions | |
mie | +1.011 |
rified | +0.991 |
elope | +0.988 |
etsy | +0.984 |
jo | +0.980 |
Neg logit contributions | |
show | -0.175 |
trick | -0.167 |
new | -0.134 |
test | -0.121 |
grown | -0.117 |
Token | Do |
Token ID | 2141 |
Feature activation | -9.35e-02 |
Pos logit contributions | |
mie | +0.070 |
etsy | +0.070 |
hew | +0.069 |
elope | +0.068 |
ppa | +0.067 |
Neg logit contributions | |
show | -0.064 |
my | -0.062 |
to | -0.060 |
your | -0.060 |
test | -0.059 |
Token | you |
Token ID | 345 |
Feature activation | -2.16e-01 |
Pos logit contributions | |
rified | +0.207 |
etsy | +0.204 |
mie | +0.198 |
elope | +0.197 |
jo | +0.194 |
Neg logit contributions | |
show | -0.116 |
my | -0.113 |
your | -0.111 |
trick | -0.110 |
new | -0.107 |
Token |
|
Token ID | 198 |
Feature activation | -1.73e-01 |
Pos logit contributions | |
rified | +0.404 |
storms | +0.385 |
mie | +0.375 |
ley | +0.373 |
lves | +0.367 |
Neg logit contributions | |
new | -0.190 |
show | -0.180 |
the | -0.179 |
grown | -0.175 |
trick | -0.173 |
Token | Once |
Token ID | 7454 |
Feature activation | -6.73e-02 |
Pos logit contributions | |
rified | +0.334 |
mie | +0.322 |
storms | +0.321 |
ley | +0.308 |
isc | +0.302 |
Neg logit contributions | |
show | -0.235 |
trick | -0.232 |
grown | -0.231 |
five | -0.228 |
new | -0.219 |
Token | home |
Token ID | 1363 |
Feature activation | -6.51e-02 |
Pos logit contributions | |
Mic | +0.165 |
mie | +0.158 |
ollo | +0.157 |
thel | +0.157 |
och | +0.156 |
Neg logit contributions | |
the | -0.062 |
for | -0.061 |
new | -0.061 |
to | -0.060 |
show | -0.056 |
Token | , |
Token ID | 11 |
Feature activation | +5.30e-02 |
Pos logit contributions | |
mie | +0.164 |
elope | +0.161 |
etsy | +0.158 |
rified | +0.157 |
jo | +0.157 |
Neg logit contributions | |
trick | -0.047 |
show | -0.045 |
test | -0.037 |
new | -0.037 |
magic | -0.034 |
Token | Pen |
Token ID | 7507 |
Feature activation | -2.72e-01 |
Pos logit contributions | |
show | +0.057 |
new | +0.054 |
the | +0.053 |
trick | +0.051 |
for | +0.050 |
Neg logit contributions | |
mie | -0.117 |
hew | -0.117 |
jo | -0.117 |
rified | -0.113 |
och | -0.113 |
Token | elope |
Token ID | 47329 |
Feature activation | -4.76e-01 |
Pos logit contributions | |
och | +0.265 |
azz | +0.225 |
Mic | +0.219 |
emic | +0.213 |
storms | +0.210 |
Neg logit contributions | |
show | -0.656 |
the | -0.639 |
to | -0.637 |
trick | -0.635 |
for | -0.633 |
Token | tucked |
Token ID | 29779 |
Feature activation | +6.69e-02 |
Pos logit contributions | |
mie | +0.807 |
elope | +0.775 |
rified | +0.773 |
jo | +0.764 |
etsy | +0.763 |
Neg logit contributions | |
show | -0.736 |
trick | -0.731 |
new | -0.688 |
test | -0.672 |
grown | -0.663 |
Token | the |
Token ID | 262 |
Feature activation | +8.88e-02 |
Pos logit contributions | |
show | +0.050 |
trick | +0.050 |
test | +0.042 |
new | +0.041 |
grown | +0.040 |
Neg logit contributions | |
mie | -0.169 |
rified | -0.164 |
elope | -0.163 |
etsy | -0.162 |
jo | -0.160 |
Token | cloth |
Token ID | 16270 |
Feature activation | -2.25e-01 |
Pos logit contributions | |
trick | +0.109 |
show | +0.109 |
new | +0.098 |
making | +0.097 |
grown | +0.097 |
Neg logit contributions | |
mie | -0.174 |
rified | -0.168 |
etsy | -0.167 |
elope | -0.164 |
jo | -0.161 |
Token | bird |
Token ID | 6512 |
Feature activation | -3.12e-01 |
Pos logit contributions | |
mie | +0.507 |
etsy | +0.489 |
rified | +0.489 |
elope | +0.482 |
jo | +0.481 |
Neg logit contributions | |
trick | -0.224 |
show | -0.219 |
test | -0.198 |
grown | -0.191 |
new | -0.189 |
Token | snug |
Token ID | 47032 |
Feature activation | -3.12e-02 |
Pos logit contributions | |
mie | +0.731 |
rified | +0.717 |
elope | +0.711 |
jo | +0.706 |
etsy | +0.706 |
Neg logit contributions | |
show | -0.281 |
trick | -0.274 |
new | -0.250 |
test | -0.233 |
for | -0.232 |
Token | this |
Token ID | 428 |
Feature activation | -9.06e-03 |
Pos logit contributions | |
show | +0.059 |
trick | +0.059 |
new | +0.050 |
test | +0.049 |
grown | +0.046 |
Neg logit contributions | |
mie | -0.211 |
elope | -0.206 |
rified | -0.206 |
jo | -0.204 |
etsy | -0.202 |
Token | special |
Token ID | 2041 |
Feature activation | +4.50e-02 |
Pos logit contributions | |
mie | +0.017 |
elope | +0.017 |
jo | +0.017 |
rified | +0.016 |
storms | +0.016 |
Neg logit contributions | |
trick | -0.009 |
five | -0.009 |
show | -0.008 |
test | -0.008 |
my | -0.008 |
Token | lamp |
Token ID | 20450 |
Feature activation | -2.86e-01 |
Pos logit contributions | |
trick | +0.041 |
my | +0.040 |
show | +0.039 |
your | +0.038 |
five | +0.038 |
Neg logit contributions | |
elope | -0.090 |
jo | -0.089 |
mie | -0.088 |
pillar | -0.086 |
rified | -0.085 |
Token | that |
Token ID | 326 |
Feature activation | -5.90e-02 |
Pos logit contributions | |
mie | +0.613 |
elope | +0.601 |
jo | +0.595 |
rified | +0.584 |
etsy | +0.575 |
Neg logit contributions | |
trick | -0.300 |
show | -0.298 |
new | -0.271 |
test | -0.265 |
grown | -0.265 |
Token | Pen |
Token ID | 7507 |
Feature activation | -2.71e-01 |
Pos logit contributions | |
mie | +0.116 |
rified | +0.112 |
elope | +0.112 |
etsy | +0.111 |
jo | +0.111 |
Neg logit contributions | |
show | -0.075 |
trick | -0.074 |
new | -0.069 |
test | -0.067 |
grown | -0.066 |
Token | elope |
Token ID | 47329 |
Feature activation | -4.76e-01 |
Pos logit contributions | |
Mic | +0.281 |
rified | +0.281 |
och | +0.271 |
etsy | +0.271 |
mie | +0.265 |
Neg logit contributions | |
show | -0.601 |
trick | -0.591 |
for | -0.561 |
grown | -0.558 |
to | -0.557 |
Token | had |
Token ID | 550 |
Feature activation | -2.65e-02 |
Pos logit contributions | |
mie | +0.772 |
rified | +0.753 |
etsy | +0.743 |
elope | +0.742 |
jo | +0.733 |
Neg logit contributions | |
show | -0.772 |
trick | -0.754 |
new | -0.715 |
test | -0.699 |
grown | -0.697 |
Token | ! |
Token ID | 0 |
Feature activation | -7.88e-02 |
Pos logit contributions | |
mie | +0.044 |
elope | +0.041 |
jo | +0.041 |
hew | +0.041 |
la | +0.040 |
Neg logit contributions | |
show | -0.039 |
trick | -0.039 |
test | -0.037 |
new | -0.036 |
job | -0.035 |
Token | |
Token ID | 220 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
ppa | +0.166 |
lves | +0.163 |
mie | +0.161 |
rified | +0.160 |
ppy | +0.160 |
Neg logit contributions | |
my | -0.094 |
the | -0.093 |
show | -0.089 |
to | -0.087 |
your | -0.087 |
Token |
|
Token ID | 198 |
Feature activation | -1.58e-01 |
Pos logit contributions | |
rified | +0.639 |
mie | +0.610 |
lves | +0.604 |
ley | +0.590 |
thel | +0.590 |
Neg logit contributions | |
new | -0.279 |
to | -0.276 |
the | -0.270 |
for | -0.247 |
your | -0.236 |
Token |
|
Token ID | 198 |
Feature activation | -1.57e-01 |
Pos logit contributions | |
rified | +0.380 |
storms | +0.358 |
mie | +0.351 |
lves | +0.351 |
ppy | +0.346 |
Neg logit contributions | |
new | -0.167 |
the | -0.163 |
show | -0.162 |
grown | -0.155 |
trick | -0.153 |
Token | ! |
Token ID | 0 |
Feature activation | -7.88e-02 |
Pos logit contributions | |
mie | +0.044 |
elope | +0.041 |
jo | +0.041 |
hew | +0.041 |
la | +0.040 |
Neg logit contributions | |
show | -0.039 |
trick | -0.039 |
test | -0.037 |
new | -0.036 |
job | -0.035 |
Token | |
Token ID | 220 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
ppa | +0.166 |
lves | +0.163 |
mie | +0.161 |
rified | +0.160 |
ppy | +0.160 |
Neg logit contributions | |
my | -0.094 |
the | -0.093 |
show | -0.089 |
to | -0.087 |
your | -0.087 |
Token |
|
Token ID | 198 |
Feature activation | -1.58e-01 |
Pos logit contributions | |
rified | +0.639 |
mie | +0.610 |
lves | +0.604 |
ley | +0.590 |
thel | +0.590 |
Neg logit contributions | |
new | -0.279 |
to | -0.276 |
the | -0.270 |
for | -0.247 |
your | -0.236 |
Token |
|
Token ID | 198 |
Feature activation | -1.57e-01 |
Pos logit contributions | |
rified | +0.380 |
storms | +0.358 |
mie | +0.351 |
lves | +0.351 |
ppy | +0.346 |
Neg logit contributions | |
new | -0.167 |
the | -0.163 |
show | -0.162 |
grown | -0.155 |
trick | -0.153 |
Token | Pen |
Token ID | 25553 |
Feature activation | -2.68e-01 |
Pos logit contributions | |
rified | +0.288 |
mie | +0.277 |
storms | +0.274 |
isc | +0.259 |
lves | +0.258 |
Neg logit contributions | |
show | -0.235 |
trick | -0.230 |
five | -0.225 |
grown | -0.224 |
new | -0.218 |
Token | elope |
Token ID | 47329 |
Feature activation | -4.66e-01 |
Pos logit contributions | |
och | +0.161 |
Mic | +0.153 |
mie | +0.149 |
rified | +0.141 |
etsy | +0.139 |
Neg logit contributions | |
show | -0.717 |
trick | -0.698 |
grown | -0.681 |
the | -0.677 |
five | -0.675 |
Token | used |
Token ID | 973 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
mie | +0.797 |
elope | +0.766 |
jo | +0.751 |
rified | +0.746 |
hew | +0.737 |
Neg logit contributions | |
trick | -0.718 |
show | -0.706 |
new | -0.674 |
test | -0.658 |
magic | -0.638 |
Token | her |
Token ID | 607 |
Feature activation | +9.91e-04 |
Pos logit contributions | |
trick | +0.095 |
show | +0.094 |
test | +0.088 |
job | +0.081 |
quiz | +0.076 |
Neg logit contributions | |
mie | -0.246 |
la | -0.234 |
etsy | -0.234 |
elope | -0.234 |
jo | -0.233 |
Token | lamp |
Token ID | 20450 |
Feature activation | -2.72e-01 |
Pos logit contributions | |
trick | +0.001 |
show | +0.001 |
grown | +0.001 |
making | +0.001 |
sport | +0.001 |
Neg logit contributions | |
jo | -0.002 |
elope | -0.002 |
mie | -0.002 |
apa | -0.002 |
hew | -0.002 |
Token | every |
Token ID | 790 |
Feature activation | +9.30e-02 |
Pos logit contributions | |
mie | +0.716 |
elope | +0.706 |
jo | +0.705 |
etsy | +0.692 |
rified | +0.688 |
Neg logit contributions | |
show | -0.147 |
trick | -0.138 |
grown | -0.113 |
test | -0.108 |
my | -0.105 |
Token | night |
Token ID | 1755 |
Feature activation | -6.92e-02 |
Pos logit contributions | |
ags | +0.117 |
my | +0.115 |
your | +0.113 |
trick | +0.112 |
matching | +0.110 |
Neg logit contributions | |
elope | -0.156 |
jo | -0.152 |
pillar | -0.144 |
mie | -0.140 |
bo | -0.138 |
Token | " |
Token ID | 366 |
Feature activation | -2.14e-01 |
Pos logit contributions | |
rified | +0.150 |
mie | +0.149 |
storms | +0.142 |
ksh | +0.139 |
etsy | +0.139 |
Neg logit contributions | |
show | -0.096 |
trick | -0.094 |
new | -0.092 |
to | -0.090 |
the | -0.087 |
Token | Don |
Token ID | 3987 |
Feature activation | -1.93e-01 |
Pos logit contributions | |
jo | +0.308 |
mie | +0.305 |
rified | +0.304 |
storms | +0.300 |
isc | +0.298 |
Neg logit contributions | |
show | -0.375 |
trick | -0.372 |
five | -0.355 |
grown | -0.348 |
test | -0.347 |
Token | 't |
Token ID | 470 |
Feature activation | -6.06e-02 |
Pos logit contributions | |
mie | +0.398 |
elope | +0.394 |
rified | +0.391 |
lves | +0.382 |
jo | +0.381 |
Neg logit contributions | |
show | -0.243 |
trick | -0.231 |
new | -0.219 |
to | -0.218 |
my | -0.214 |
Token | worry |
Token ID | 5490 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
mie | +0.132 |
elope | +0.128 |
rified | +0.128 |
etsy | +0.127 |
jo | +0.127 |
Neg logit contributions | |
show | -0.066 |
trick | -0.065 |
new | -0.060 |
test | -0.058 |
grown | -0.057 |
Token | , |
Token ID | 11 |
Feature activation | +3.67e-02 |
Pos logit contributions | |
mie | +0.342 |
rified | +0.336 |
elope | +0.334 |
och | +0.331 |
etsy | +0.331 |
Neg logit contributions | |
show | -0.053 |
trick | -0.053 |
new | -0.050 |
for | -0.047 |
to | -0.042 |
Token | Harry |
Token ID | 5850 |
Feature activation | -4.61e-01 |
Pos logit contributions | |
show | +0.052 |
trick | +0.050 |
new | +0.048 |
grown | +0.046 |
test | +0.045 |
Neg logit contributions | |
mie | -0.070 |
jo | -0.068 |
elope | -0.068 |
rified | -0.068 |
etsy | -0.067 |
Token | . |
Token ID | 13 |
Feature activation | -9.42e-02 |
Pos logit contributions | |
rified | +1.140 |
mie | +1.133 |
etsy | +1.115 |
Mic | +1.112 |
och | +1.108 |
Neg logit contributions | |
show | -0.366 |
trick | -0.339 |
for | -0.333 |
to | -0.318 |
new | -0.310 |
Token | I |
Token ID | 314 |
Feature activation | -2.65e-01 |
Pos logit contributions | |
mie | +0.193 |
ppa | +0.191 |
thel | +0.191 |
hel | +0.189 |
lves | +0.188 |
Neg logit contributions | |
my | -0.119 |
show | -0.115 |
your | -0.107 |
to | -0.106 |
help | -0.104 |
Token | love |
Token ID | 1842 |
Feature activation | +8.00e-02 |
Pos logit contributions | |
mie | +0.419 |
rified | +0.407 |
elope | +0.403 |
jo | +0.399 |
etsy | +0.396 |
Neg logit contributions | |
show | -0.446 |
trick | -0.439 |
new | -0.412 |
test | -0.406 |
grown | -0.402 |
Token | being |
Token ID | 852 |
Feature activation | -5.78e-03 |
Pos logit contributions | |
trick | +0.108 |
show | +0.106 |
test | +0.099 |
grown | +0.097 |
new | +0.096 |
Neg logit contributions | |
mie | -0.143 |
elope | -0.137 |
jo | -0.136 |
storms | -0.136 |
hew | -0.133 |
Token | covered |
Token ID | 5017 |
Feature activation | +7.41e-02 |
Pos logit contributions | |
mie | +0.013 |
Mic | +0.012 |
bo | +0.012 |
pillar | +0.012 |
rified | +0.012 |
Neg logit contributions | |
show | -0.005 |
trick | -0.005 |
test | -0.005 |
work | -0.005 |
balance | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | +2.73e-02 |
Pos logit contributions | |
mie | +0.321 |
rified | +0.312 |
etsy | +0.312 |
jo | +0.311 |
elope | +0.310 |
Neg logit contributions | |
show | -0.210 |
trick | -0.203 |
new | -0.195 |
grown | -0.187 |
for | -0.185 |
Token | a |
Token ID | 257 |
Feature activation | +5.24e-02 |
Pos logit contributions | |
show | +0.016 |
trick | +0.015 |
test | +0.012 |
new | +0.012 |
grown | +0.011 |
Neg logit contributions | |
mie | -0.072 |
elope | -0.071 |
rified | -0.071 |
etsy | -0.070 |
jo | -0.070 |
Token | little |
Token ID | 1310 |
Feature activation | -6.87e-02 |
Pos logit contributions | |
show | +0.050 |
trick | +0.048 |
new | +0.046 |
test | +0.042 |
grown | +0.042 |
Neg logit contributions | |
mie | -0.123 |
rified | -0.121 |
jo | -0.120 |
hew | -0.119 |
etsy | -0.119 |
Token | boy |
Token ID | 2933 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
mie | +0.105 |
hew | +0.102 |
jo | +0.102 |
rified | +0.101 |
etsy | +0.101 |
Neg logit contributions | |
show | -0.120 |
trick | -0.117 |
new | -0.114 |
test | -0.110 |
magic | -0.108 |
Token | called |
Token ID | 1444 |
Feature activation | -4.61e-02 |
Pos logit contributions | |
mie | +0.708 |
rified | +0.695 |
etsy | +0.684 |
elope | +0.680 |
hew | +0.675 |
Neg logit contributions | |
show | -0.561 |
trick | -0.543 |
new | -0.509 |
test | -0.496 |
for | -0.495 |
Token | Steven |
Token ID | 8239 |
Feature activation | -4.58e-01 |
Pos logit contributions | |
mie | +0.074 |
hew | +0.072 |
jo | +0.070 |
elope | +0.070 |
rified | +0.070 |
Neg logit contributions | |
show | -0.078 |
trick | -0.075 |
new | -0.073 |
to | -0.073 |
for | -0.073 |
Token | . |
Token ID | 13 |
Feature activation | -9.08e-02 |
Pos logit contributions | |
rified | +1.183 |
hew | +1.169 |
etsy | +1.165 |
mie | +1.144 |
Mic | +1.142 |
Neg logit contributions | |
to | -0.311 |
for | -0.307 |
show | -0.306 |
the | -0.289 |
new | -0.287 |
Token | Steven |
Token ID | 8239 |
Feature activation | -4.54e-01 |
Pos logit contributions | |
etsy | +0.136 |
mie | +0.134 |
lves | +0.127 |
rified | +0.123 |
Hop | +0.121 |
Neg logit contributions | |
show | -0.170 |
new | -0.161 |
to | -0.161 |
the | -0.159 |
his | -0.158 |
Token | was |
Token ID | 373 |
Feature activation | +5.43e-02 |
Pos logit contributions | |
rified | +0.878 |
och | +0.862 |
etsy | +0.851 |
mie | +0.837 |
Mic | +0.830 |
Neg logit contributions | |
show | -0.612 |
trick | -0.571 |
new | -0.563 |
to | -0.554 |
grown | -0.553 |
Token | only |
Token ID | 691 |
Feature activation | -1.60e-02 |
Pos logit contributions | |
trick | +0.058 |
show | +0.057 |
test | +0.050 |
new | +0.049 |
work | +0.048 |
Neg logit contributions | |
mie | -0.118 |
elope | -0.115 |
rified | -0.112 |
jo | -0.110 |
etsy | -0.109 |
Token | three |
Token ID | 1115 |
Feature activation | +9.81e-03 |
Pos logit contributions | |
mie | +0.033 |
elope | +0.032 |
rified | +0.032 |
jo | +0.031 |
etsy | +0.031 |
Neg logit contributions | |
trick | -0.018 |
show | -0.018 |
grown | -0.015 |
test | -0.015 |
new | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -6.68e-02 |
Pos logit contributions | |
mie | +0.370 |
elope | +0.369 |
jo | +0.354 |
rified | +0.350 |
storms | +0.348 |
Neg logit contributions | |
trick | -0.264 |
show | -0.244 |
test | -0.235 |
my | -0.229 |
new | -0.219 |
Token | The |
Token ID | 383 |
Feature activation | -9.33e-03 |
Pos logit contributions | |
mie | +0.109 |
rified | +0.105 |
jo | +0.102 |
hew | +0.102 |
ppy | +0.101 |
Neg logit contributions | |
show | -0.110 |
new | -0.104 |
trick | -0.102 |
grown | -0.102 |
to | -0.101 |
Token | next |
Token ID | 1306 |
Feature activation | +1.83e-02 |
Pos logit contributions | |
mie | +0.020 |
jo | +0.020 |
rified | +0.020 |
etsy | +0.020 |
elope | +0.020 |
Neg logit contributions | |
show | -0.011 |
trick | -0.010 |
new | -0.010 |
grown | -0.009 |
magic | -0.009 |
Token | day |
Token ID | 1110 |
Feature activation | -5.44e-02 |
Pos logit contributions | |
show | +0.019 |
trick | +0.019 |
making | +0.017 |
new | +0.017 |
grown | +0.016 |
Neg logit contributions | |
elope | -0.038 |
mie | -0.037 |
jo | -0.037 |
etsy | -0.036 |
rified | -0.036 |
Token | Pen |
Token ID | 7507 |
Feature activation | -2.53e-01 |
Pos logit contributions | |
mie | +0.147 |
rified | +0.143 |
elope | +0.143 |
jo | +0.142 |
etsy | +0.142 |
Neg logit contributions | |
show | -0.030 |
trick | -0.029 |
new | -0.024 |
test | -0.022 |
grown | -0.022 |
Token | elope |
Token ID | 47329 |
Feature activation | -4.58e-01 |
Pos logit contributions | |
och | +0.238 |
Mic | +0.209 |
azz | +0.191 |
aden | +0.190 |
rified | +0.182 |
Neg logit contributions | |
show | -0.616 |
for | -0.612 |
to | -0.611 |
trick | -0.606 |
five | -0.606 |
Token | went |
Token ID | 1816 |
Feature activation | +6.70e-02 |
Pos logit contributions | |
mie | +0.751 |
elope | +0.719 |
rified | +0.712 |
jo | +0.707 |
etsy | +0.701 |
Neg logit contributions | |
trick | -0.734 |
show | -0.731 |
new | -0.689 |
test | -0.674 |
grown | -0.661 |
Token | outside |
Token ID | 2354 |
Feature activation | +2.18e-02 |
Pos logit contributions | |
trick | +0.067 |
color | +0.060 |
job | +0.057 |
test | +0.057 |
power | +0.057 |
Neg logit contributions | |
mie | -0.140 |
jo | -0.130 |
och | -0.130 |
elope | -0.128 |
etsy | -0.127 |
Token | to |
Token ID | 284 |
Feature activation | +2.11e-01 |
Pos logit contributions | |
trick | +0.022 |
test | +0.019 |
show | +0.018 |
magic | +0.018 |
job | +0.017 |
Neg logit contributions | |
elope | -0.047 |
mie | -0.047 |
jo | -0.046 |
storms | -0.044 |
rified | -0.044 |
Token | pick |
Token ID | 2298 |
Feature activation | +9.59e-02 |
Pos logit contributions | |
trick | +0.209 |
new | +0.188 |
skills | +0.185 |
for | +0.184 |
five | +0.183 |
Neg logit contributions | |
elope | -0.436 |
mie | -0.434 |
jo | -0.392 |
etsy | -0.391 |
Bear | -0.388 |
Token | more |
Token ID | 517 |
Feature activation | +1.40e-01 |
Pos logit contributions | |
show | +0.056 |
trick | +0.053 |
new | +0.049 |
for | +0.041 |
grown | +0.040 |
Neg logit contributions | |
mie | -0.256 |
elope | -0.252 |
jo | -0.252 |
etsy | -0.251 |
rified | -0.250 |
Token | ". |
Token ID | 1911 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
mie | +0.146 |
jo | +0.145 |
elope | +0.144 |
storms | +0.143 |
hew | +0.138 |
Neg logit contributions | |
trick | -0.086 |
show | -0.084 |
test | -0.077 |
new | -0.073 |
five | -0.071 |
Token | Everyone |
Token ID | 11075 |
Feature activation | -2.86e-01 |
Pos logit contributions | |
rified | +0.254 |
mie | +0.252 |
lves | +0.252 |
hew | +0.248 |
ppy | +0.247 |
Neg logit contributions | |
show | -0.181 |
the | -0.175 |
to | -0.173 |
his | -0.170 |
your | -0.164 |
Token | laughed |
Token ID | 13818 |
Feature activation | +1.65e-02 |
Pos logit contributions | |
mie | +0.464 |
jo | +0.448 |
rified | +0.445 |
elope | +0.442 |
hew | +0.419 |
Neg logit contributions | |
trick | -0.454 |
show | -0.444 |
new | -0.429 |
test | -0.416 |
magic | -0.408 |
Token | . |
Token ID | 13 |
Feature activation | -8.47e-02 |
Pos logit contributions | |
show | +0.013 |
trick | +0.013 |
test | +0.011 |
new | +0.011 |
grown | +0.011 |
Neg logit contributions | |
mie | -0.040 |
rified | -0.039 |
jo | -0.039 |
elope | -0.039 |
etsy | -0.039 |
Token | Mrs |
Token ID | 9074 |
Feature activation | -2.19e-01 |
Pos logit contributions | |
mie | +0.165 |
ppa | +0.164 |
lves | +0.159 |
rified | +0.158 |
thel | +0.156 |
Neg logit contributions | |
show | -0.113 |
the | -0.108 |
to | -0.107 |
new | -0.107 |
five | -0.103 |
Token | Jones |
Token ID | 5437 |
Feature activation | -4.57e-01 |
Pos logit contributions | |
Mic | +0.610 |
lves | +0.607 |
etsy | +0.601 |
hew | +0.589 |
oped | +0.587 |
Neg logit contributions | |
for | -0.132 |
to | -0.114 |
the | -0.108 |
new | -0.107 |
show | -0.104 |
Token | said |
Token ID | 531 |
Feature activation | -1.04e-01 |
Pos logit contributions | |
mie | +0.886 |
rified | +0.874 |
etsy | +0.867 |
elope | +0.855 |
hew | +0.850 |
Neg logit contributions | |
show | -0.602 |
trick | -0.579 |
new | -0.533 |
for | -0.523 |
grown | -0.522 |
Token | , |
Token ID | 11 |
Feature activation | +5.90e-02 |
Pos logit contributions | |
mie | +0.246 |
rified | +0.239 |
etsy | +0.237 |
elope | +0.236 |
jo | +0.235 |
Neg logit contributions | |
show | -0.090 |
trick | -0.089 |
test | -0.078 |
new | -0.078 |
work | -0.073 |
Token | " |
Token ID | 366 |
Feature activation | -1.61e-01 |
Pos logit contributions | |
show | +0.094 |
trick | +0.093 |
new | +0.088 |
grown | +0.086 |
test | +0.085 |
Neg logit contributions | |
mie | -0.098 |
rified | -0.094 |
elope | -0.094 |
etsy | -0.093 |
jo | -0.093 |
Token | Well |
Token ID | 5779 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
elope | +0.227 |
mie | +0.223 |
apa | +0.222 |
rified | +0.218 |
uddy | +0.218 |
Neg logit contributions | |
trick | -0.276 |
show | -0.275 |
five | -0.270 |
to | -0.260 |
color | -0.257 |
Token | Jenny |
Token ID | 28437 |
Feature activation | -3.35e-01 |
Pos logit contributions | |
mie | +0.460 |
hew | +0.455 |
rified | +0.440 |
jo | +0.433 |
elope | +0.432 |
Neg logit contributions | |
trick | -0.164 |
show | -0.147 |
new | -0.135 |
test | -0.128 |
magic | -0.123 |
Token | humble |
Token ID | 19695 |
Feature activation | -3.78e-03 |
Pos logit contributions | |
show | +0.055 |
trick | +0.051 |
new | +0.050 |
grown | +0.046 |
test | +0.046 |
Neg logit contributions | |
mie | -0.148 |
rified | -0.147 |
jo | -0.145 |
hew | -0.144 |
etsy | -0.143 |
Token | lady |
Token ID | 10846 |
Feature activation | -3.01e-01 |
Pos logit contributions | |
mie | +0.007 |
elope | +0.007 |
rified | +0.007 |
etsy | +0.007 |
jo | +0.007 |
Neg logit contributions | |
show | -0.005 |
trick | -0.005 |
new | -0.005 |
test | -0.005 |
grown | -0.005 |
Token | called |
Token ID | 1444 |
Feature activation | -4.26e-02 |
Pos logit contributions | |
rified | +0.593 |
hew | +0.587 |
etsy | +0.577 |
mie | +0.575 |
elope | +0.554 |
Neg logit contributions | |
show | -0.405 |
trick | -0.372 |
to | -0.370 |
for | -0.365 |
new | -0.356 |
Token | Mrs |
Token ID | 9074 |
Feature activation | -2.19e-01 |
Pos logit contributions | |
hew | +0.072 |
mie | +0.071 |
rified | +0.069 |
jo | +0.069 |
lves | +0.067 |
Neg logit contributions | |
show | -0.066 |
the | -0.065 |
to | -0.064 |
new | -0.063 |
five | -0.063 |
Token | . |
Token ID | 13 |
Feature activation | -8.35e-02 |
Pos logit contributions | |
rified | +0.569 |
hew | +0.567 |
lves | +0.561 |
jo | +0.559 |
oped | +0.557 |
Neg logit contributions | |
for | -0.175 |
to | -0.145 |
the | -0.141 |
new | -0.135 |
five | -0.116 |
Token | Jones |
Token ID | 5437 |
Feature activation | -4.57e-01 |
Pos logit contributions | |
hew | +0.127 |
lves | +0.116 |
rified | +0.116 |
mie | +0.113 |
jo | +0.111 |
Neg logit contributions | |
the | -0.146 |
new | -0.144 |
to | -0.143 |
show | -0.141 |
five | -0.141 |
Token | . |
Token ID | 13 |
Feature activation | -7.88e-02 |
Pos logit contributions | |
rified | +1.002 |
mie | +0.972 |
hew | +0.965 |
etsy | +0.957 |
Mic | +0.941 |
Neg logit contributions | |
show | -0.507 |
trick | -0.469 |
new | -0.464 |
for | -0.447 |
to | -0.446 |
Token | One |
Token ID | 1881 |
Feature activation | -9.58e-02 |
Pos logit contributions | |
hew | +0.130 |
rified | +0.126 |
lves | +0.124 |
mie | +0.124 |
ppy | +0.122 |
Neg logit contributions | |
show | -0.128 |
new | -0.125 |
to | -0.122 |
the | -0.122 |
for | -0.119 |
Token | day |
Token ID | 1110 |
Feature activation | -6.63e-02 |
Pos logit contributions | |
mie | +0.237 |
rified | +0.231 |
elope | +0.230 |
etsy | +0.229 |
jo | +0.228 |
Neg logit contributions | |
show | -0.076 |
trick | -0.073 |
new | -0.066 |
test | -0.062 |
grown | -0.061 |
Token | she |
Token ID | 673 |
Feature activation | -2.87e-01 |
Pos logit contributions | |
mie | +0.114 |
rified | +0.109 |
etsy | +0.109 |
elope | +0.108 |
jo | +0.106 |
Neg logit contributions | |
trick | -0.102 |
show | -0.100 |
new | -0.093 |
test | -0.092 |
magic | -0.089 |
Token | wanted |
Token ID | 2227 |
Feature activation | +3.71e-02 |
Pos logit contributions | |
mie | +0.459 |
elope | +0.439 |
jo | +0.434 |
rified | +0.433 |
etsy | +0.430 |
Neg logit contributions | |
trick | -0.470 |
show | -0.470 |
new | -0.444 |
test | -0.434 |
grown | -0.425 |
Token | special |
Token ID | 2041 |
Feature activation | +4.98e-02 |
Pos logit contributions | |
trick | +0.057 |
show | +0.055 |
new | +0.048 |
my | +0.048 |
test | +0.047 |
Neg logit contributions | |
mie | -0.156 |
elope | -0.155 |
etsy | -0.150 |
jo | -0.150 |
rified | -0.147 |
Token | birthday |
Token ID | 10955 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
my | +0.044 |
trick | +0.043 |
your | +0.042 |
ags | +0.042 |
grown | +0.040 |
Neg logit contributions | |
elope | -0.100 |
jo | -0.099 |
mie | -0.098 |
etsy | -0.094 |
thel | -0.093 |
Token | present |
Token ID | 1944 |
Feature activation | -1.49e-01 |
Pos logit contributions | |
mie | +0.279 |
elope | +0.272 |
jo | +0.271 |
etsy | +0.266 |
rified | +0.265 |
Neg logit contributions | |
trick | -0.109 |
show | -0.107 |
new | -0.097 |
test | -0.095 |
grown | -0.094 |
Token | from |
Token ID | 422 |
Feature activation | +1.42e-01 |
Pos logit contributions | |
mie | +0.349 |
elope | +0.340 |
jo | +0.338 |
rified | +0.337 |
etsy | +0.336 |
Neg logit contributions | |
show | -0.130 |
trick | -0.130 |
new | -0.116 |
test | -0.112 |
grown | -0.111 |
Token | Mr |
Token ID | 1770 |
Feature activation | -1.95e-01 |
Pos logit contributions | |
show | +0.122 |
trick | +0.121 |
new | +0.107 |
test | +0.103 |
grown | +0.101 |
Neg logit contributions | |
mie | -0.339 |
elope | -0.330 |
rified | -0.329 |
jo | -0.326 |
etsy | -0.326 |
Token | Jones |
Token ID | 5437 |
Feature activation | -4.56e-01 |
Pos logit contributions | |
rified | +0.499 |
mie | +0.486 |
lves | +0.485 |
och | +0.482 |
hew | +0.479 |
Neg logit contributions | |
for | -0.149 |
new | -0.132 |
to | -0.127 |
five | -0.120 |
show | -0.118 |
Token | and |
Token ID | 290 |
Feature activation | +6.15e-02 |
Pos logit contributions | |
mie | +1.118 |
rified | +1.117 |
etsy | +1.098 |
hew | +1.087 |
elope | +1.082 |
Neg logit contributions | |
show | -0.367 |
for | -0.338 |
trick | -0.328 |
to | -0.317 |
new | -0.306 |
Token | the |
Token ID | 262 |
Feature activation | +9.21e-02 |
Pos logit contributions | |
trick | +0.099 |
show | +0.096 |
new | +0.090 |
test | +0.090 |
grown | +0.086 |
Neg logit contributions | |
mie | -0.102 |
elope | -0.097 |
rified | -0.096 |
jo | -0.095 |
etsy | -0.094 |
Token | family |
Token ID | 1641 |
Feature activation | -3.17e-01 |
Pos logit contributions | |
show | +0.119 |
trick | +0.114 |
new | +0.110 |
test | +0.104 |
grown | +0.104 |
Neg logit contributions | |
mie | -0.185 |
jo | -0.180 |
rified | -0.180 |
elope | -0.180 |
etsy | -0.180 |
Token | decided |
Token ID | 3066 |
Feature activation | -3.33e-02 |
Pos logit contributions | |
mie | +0.708 |
elope | +0.685 |
jo | +0.682 |
rified | +0.679 |
etsy | +0.677 |
Neg logit contributions | |
trick | -0.317 |
show | -0.313 |
new | -0.284 |
test | -0.275 |
grown | -0.269 |
Token | to |
Token ID | 284 |
Feature activation | +1.98e-01 |
Pos logit contributions | |
mie | +0.089 |
etsy | +0.084 |
jo | +0.084 |
elope | +0.084 |
rified | +0.083 |
Neg logit contributions | |
trick | -0.017 |
show | -0.016 |
test | -0.014 |
new | -0.012 |
magic | -0.012 |
Token | little |
Token ID | 1310 |
Feature activation | -6.87e-02 |
Pos logit contributions | |
show | +0.050 |
trick | +0.048 |
new | +0.046 |
test | +0.042 |
grown | +0.042 |
Neg logit contributions | |
mie | -0.123 |
rified | -0.121 |
jo | -0.120 |
hew | -0.119 |
etsy | -0.119 |
Token | boy |
Token ID | 2933 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
mie | +0.105 |
hew | +0.102 |
jo | +0.102 |
rified | +0.101 |
etsy | +0.101 |
Neg logit contributions | |
show | -0.120 |
trick | -0.117 |
new | -0.114 |
test | -0.110 |
magic | -0.108 |
Token | called |
Token ID | 1444 |
Feature activation | -4.61e-02 |
Pos logit contributions | |
mie | +0.708 |
rified | +0.695 |
etsy | +0.684 |
elope | +0.680 |
hew | +0.675 |
Neg logit contributions | |
show | -0.561 |
trick | -0.543 |
new | -0.509 |
test | -0.496 |
for | -0.495 |
Token | Steven |
Token ID | 8239 |
Feature activation | -4.58e-01 |
Pos logit contributions | |
mie | +0.074 |
hew | +0.072 |
jo | +0.070 |
elope | +0.070 |
rified | +0.070 |
Neg logit contributions | |
show | -0.078 |
trick | -0.075 |
new | -0.073 |
to | -0.073 |
for | -0.073 |
Token | . |
Token ID | 13 |
Feature activation | -9.08e-02 |
Pos logit contributions | |
rified | +1.183 |
hew | +1.169 |
etsy | +1.165 |
mie | +1.144 |
Mic | +1.142 |
Neg logit contributions | |
to | -0.311 |
for | -0.307 |
show | -0.306 |
the | -0.289 |
new | -0.287 |
Token | Steven |
Token ID | 8239 |
Feature activation | -4.54e-01 |
Pos logit contributions | |
etsy | +0.136 |
mie | +0.134 |
lves | +0.127 |
rified | +0.123 |
Hop | +0.121 |
Neg logit contributions | |
show | -0.170 |
new | -0.161 |
to | -0.161 |
the | -0.159 |
his | -0.158 |
Token | was |
Token ID | 373 |
Feature activation | +5.43e-02 |
Pos logit contributions | |
rified | +0.878 |
och | +0.862 |
etsy | +0.851 |
mie | +0.837 |
Mic | +0.830 |
Neg logit contributions | |
show | -0.612 |
trick | -0.571 |
new | -0.563 |
to | -0.554 |
grown | -0.553 |
Token | only |
Token ID | 691 |
Feature activation | -1.60e-02 |
Pos logit contributions | |
trick | +0.058 |
show | +0.057 |
test | +0.050 |
new | +0.049 |
work | +0.048 |
Neg logit contributions | |
mie | -0.118 |
elope | -0.115 |
rified | -0.112 |
jo | -0.110 |
etsy | -0.109 |
Token | three |
Token ID | 1115 |
Feature activation | +9.81e-03 |
Pos logit contributions | |
mie | +0.033 |
elope | +0.032 |
rified | +0.032 |
jo | +0.031 |
etsy | +0.031 |
Neg logit contributions | |
trick | -0.018 |
show | -0.018 |
grown | -0.015 |
test | -0.015 |
new | -0.014 |
Token | , |
Token ID | 11 |
Feature activation | +5.85e-02 |
Pos logit contributions | |
trick | +0.018 |
show | +0.018 |
test | +0.017 |
new | +0.016 |
magic | +0.016 |
Neg logit contributions | |
mie | -0.014 |
jo | -0.013 |
elope | -0.013 |
rified | -0.013 |
hew | -0.012 |
Token | but |
Token ID | 475 |
Feature activation | -7.16e-02 |
Pos logit contributions | |
show | +0.064 |
trick | +0.064 |
test | +0.057 |
new | +0.057 |
grown | +0.054 |
Neg logit contributions | |
mie | -0.125 |
elope | -0.122 |
rified | -0.121 |
jo | -0.120 |
etsy | -0.119 |
Token | line |
Token ID | 1627 |
Feature activation | -6.79e-02 |
Pos logit contributions | |
show | +0.125 |
trick | +0.121 |
new | +0.118 |
test | +0.114 |
grown | +0.112 |
Neg logit contributions | |
mie | -0.155 |
etsy | -0.150 |
rified | -0.148 |
jo | -0.148 |
elope | -0.147 |
Token | . |
Token ID | 13 |
Feature activation | -7.41e-02 |
Pos logit contributions | |
mie | +0.148 |
jo | +0.145 |
rified | +0.144 |
elope | +0.143 |
hew | +0.141 |
Neg logit contributions | |
trick | -0.069 |
show | -0.067 |
new | -0.061 |
test | -0.060 |
magic | -0.059 |
Token |
|
Token ID | 198 |
Feature activation | -1.64e-01 |
Pos logit contributions | |
hew | +0.144 |
ppy | +0.143 |
rified | +0.141 |
mie | +0.141 |
ppa | +0.139 |
Neg logit contributions | |
the | -0.095 |
to | -0.095 |
new | -0.091 |
show | -0.090 |
grown | -0.084 |
Token |
|
Token ID | 198 |
Feature activation | -1.62e-01 |
Pos logit contributions | |
rified | +0.373 |
storms | +0.358 |
mie | +0.349 |
ksh | +0.346 |
ppy | +0.342 |
Neg logit contributions | |
the | -0.179 |
show | -0.175 |
new | -0.175 |
to | -0.173 |
trick | -0.165 |
Token | G |
Token ID | 38 |
Feature activation | -2.44e-01 |
Pos logit contributions | |
rified | +0.319 |
storms | +0.317 |
mie | +0.300 |
jo | +0.299 |
elope | +0.296 |
Neg logit contributions | |
show | -0.205 |
grown | -0.200 |
five | -0.199 |
my | -0.199 |
your | -0.195 |
Token | avin |
Token ID | 20637 |
Feature activation | -4.54e-01 |
Pos logit contributions | |
storms | +0.366 |
rified | +0.365 |
ksh | +0.354 |
Mic | +0.347 |
girls | +0.323 |
Neg logit contributions | |
work | -0.397 |
for | -0.392 |
grown | -0.391 |
to | -0.383 |
new | -0.376 |
Token | stopped |
Token ID | 5025 |
Feature activation | +8.54e-02 |
Pos logit contributions | |
rified | +0.858 |
etsy | +0.856 |
och | +0.850 |
Mic | +0.843 |
mie | +0.831 |
Neg logit contributions | |
show | -0.619 |
trick | -0.580 |
grown | -0.552 |
to | -0.547 |
the | -0.536 |
Token | and |
Token ID | 290 |
Feature activation | +6.43e-02 |
Pos logit contributions | |
trick | +0.082 |
show | +0.077 |
test | +0.070 |
job | +0.069 |
magic | +0.068 |
Neg logit contributions | |
mie | -0.189 |
elope | -0.185 |
jo | -0.182 |
rified | -0.177 |
ollo | -0.174 |
Token | the |
Token ID | 262 |
Feature activation | +9.62e-02 |
Pos logit contributions | |
show | +0.101 |
trick | +0.094 |
new | +0.092 |
grown | +0.090 |
to | +0.089 |
Neg logit contributions | |
etsy | -0.109 |
mie | -0.108 |
jo | -0.108 |
hew | -0.107 |
elope | -0.107 |
Token | woman |
Token ID | 2415 |
Feature activation | -3.68e-01 |
Pos logit contributions | |
show | +0.181 |
trick | +0.177 |
new | +0.172 |
test | +0.168 |
grown | +0.167 |
Neg logit contributions | |
mie | -0.140 |
rified | -0.134 |
etsy | -0.133 |
elope | -0.131 |
hew | -0.130 |
Token | said |
Token ID | 531 |
Feature activation | -9.19e-02 |
Pos logit contributions | |
etsy | +0.637 |
mie | +0.632 |
rified | +0.628 |
Mic | +0.622 |
hew | +0.620 |
Neg logit contributions | |
show | -0.566 |
trick | -0.527 |
new | -0.513 |
grown | -0.501 |
to | -0.495 |
Token | To |
Token ID | 1675 |
Feature activation | +4.86e-02 |
Pos logit contributions | |
ppy | +0.141 |
mie | +0.137 |
rified | +0.136 |
elope | +0.135 |
pillar | +0.134 |
Neg logit contributions | |
to | -0.114 |
show | -0.113 |
new | -0.111 |
grown | -0.110 |
the | -0.107 |
Token | encourage |
Token ID | 7898 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
trick | +0.052 |
new | +0.049 |
show | +0.044 |
red | +0.044 |
magic | +0.044 |
Neg logit contributions | |
mie | -0.103 |
elope | -0.099 |
jo | -0.099 |
hew | -0.096 |
thel | -0.095 |
Token | her |
Token ID | 607 |
Feature activation | -8.48e-03 |
Pos logit contributions | |
trick | +0.126 |
show | +0.125 |
new | +0.115 |
test | +0.111 |
grown | +0.110 |
Neg logit contributions | |
mie | -0.224 |
elope | -0.215 |
rified | -0.215 |
jo | -0.214 |
etsy | -0.213 |
Token | , |
Token ID | 11 |
Feature activation | +5.89e-02 |
Pos logit contributions | |
mie | +0.020 |
elope | +0.019 |
jo | +0.019 |
etsy | +0.019 |
rified | +0.018 |
Neg logit contributions | |
trick | -0.007 |
show | -0.007 |
new | -0.006 |
five | -0.006 |
grown | -0.006 |
Token | Mrs |
Token ID | 9074 |
Feature activation | -2.11e-01 |
Pos logit contributions | |
trick | +0.077 |
show | +0.077 |
new | +0.070 |
test | +0.069 |
grown | +0.067 |
Neg logit contributions | |
mie | -0.114 |
rified | -0.110 |
elope | -0.109 |
etsy | -0.108 |
jo | -0.108 |
Token | Jones |
Token ID | 5437 |
Feature activation | -4.54e-01 |
Pos logit contributions | |
rified | +0.605 |
Hop | +0.597 |
Mic | +0.594 |
lves | +0.591 |
etsy | +0.584 |
Neg logit contributions | |
for | -0.126 |
to | -0.108 |
new | -0.100 |
show | -0.089 |
the | -0.087 |
Token | said |
Token ID | 531 |
Feature activation | -9.98e-02 |
Pos logit contributions | |
mie | +0.900 |
rified | +0.894 |
etsy | +0.880 |
hew | +0.880 |
elope | +0.878 |
Neg logit contributions | |
show | -0.578 |
trick | -0.542 |
for | -0.514 |
to | -0.511 |
new | -0.507 |
Token | , |
Token ID | 11 |
Feature activation | +6.56e-02 |
Pos logit contributions | |
mie | +0.236 |
etsy | +0.230 |
rified | +0.230 |
elope | +0.226 |
jo | +0.224 |
Neg logit contributions | |
trick | -0.085 |
show | -0.085 |
test | -0.074 |
new | -0.071 |
work | -0.069 |
Token | " |
Token ID | 366 |
Feature activation | -1.53e-01 |
Pos logit contributions | |
show | +0.092 |
trick | +0.090 |
new | +0.085 |
test | +0.083 |
grown | +0.082 |
Neg logit contributions | |
mie | -0.121 |
elope | -0.117 |
rified | -0.117 |
jo | -0.116 |
etsy | -0.115 |
Token | It |
Token ID | 1026 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
elope | +0.211 |
uddy | +0.199 |
rified | +0.198 |
apa | +0.198 |
storms | +0.195 |
Neg logit contributions | |
five | -0.260 |
to | -0.256 |
show | -0.252 |
trick | -0.251 |
grown | -0.250 |
Token | 's |
Token ID | 338 |
Feature activation | -1.62e-02 |
Pos logit contributions | |
mie | +0.397 |
rified | +0.384 |
etsy | +0.381 |
elope | +0.379 |
jo | +0.377 |
Neg logit contributions | |
show | -0.460 |
trick | -0.451 |
new | -0.434 |
test | -0.420 |
grown | -0.419 |
Token | and |
Token ID | 290 |
Feature activation | +4.48e-02 |
Pos logit contributions | |
show | +0.015 |
trick | +0.014 |
new | +0.013 |
test | +0.013 |
magic | +0.012 |
Neg logit contributions | |
mie | -0.022 |
storms | -0.021 |
rified | -0.021 |
etsy | -0.020 |
hew | -0.020 |
Token | could |
Token ID | 714 |
Feature activation | -6.17e-02 |
Pos logit contributions | |
trick | +0.065 |
show | +0.064 |
test | +0.059 |
new | +0.059 |
five | +0.057 |
Neg logit contributions | |
mie | -0.078 |
elope | -0.075 |
rified | -0.074 |
jo | -0.073 |
etsy | -0.073 |
Token | hurt |
Token ID | 5938 |
Feature activation | +8.08e-02 |
Pos logit contributions | |
mie | +0.142 |
elope | +0.140 |
etsy | +0.137 |
jo | +0.136 |
rified | +0.135 |
Neg logit contributions | |
show | -0.052 |
trick | -0.052 |
new | -0.049 |
grown | -0.045 |
magic | -0.044 |
Token | Mrs |
Token ID | 9074 |
Feature activation | -2.19e-01 |
Pos logit contributions | |
show | +0.083 |
trick | +0.081 |
new | +0.080 |
your | +0.079 |
my | +0.079 |
Neg logit contributions | |
mie | -0.183 |
rified | -0.181 |
etsy | -0.180 |
jo | -0.179 |
storms | -0.175 |
Token | . |
Token ID | 13 |
Feature activation | -7.49e-02 |
Pos logit contributions | |
oped | +0.624 |
lves | +0.620 |
hew | +0.611 |
rified | +0.605 |
emic | +0.603 |
Neg logit contributions | |
for | -0.141 |
to | -0.102 |
new | -0.094 |
the | -0.085 |
five | -0.074 |
Token | Jones |
Token ID | 5437 |
Feature activation | -4.52e-01 |
Pos logit contributions | |
hew | +0.120 |
jo | +0.118 |
rified | +0.117 |
wl | +0.116 |
emic | +0.115 |
Neg logit contributions | |
my | -0.118 |
show | -0.118 |
new | -0.118 |
grown | -0.117 |
the | -0.117 |
Token | ' |
Token ID | 6 |
Feature activation | -3.34e-02 |
Pos logit contributions | |
mie | +0.780 |
elope | +0.751 |
rified | +0.751 |
jo | +0.741 |
etsy | +0.741 |
Neg logit contributions | |
show | -0.685 |
trick | -0.677 |
new | -0.637 |
test | -0.622 |
grown | -0.616 |
Token | body |
Token ID | 1767 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
mie | +0.071 |
elope | +0.069 |
rified | +0.069 |
jo | +0.068 |
etsy | +0.068 |
Neg logit contributions | |
show | -0.037 |
trick | -0.037 |
new | -0.034 |
test | -0.033 |
grown | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | -6.92e-02 |
Pos logit contributions | |
mie | +0.267 |
rified | +0.256 |
elope | +0.255 |
etsy | +0.255 |
jo | +0.255 |
Neg logit contributions | |
show | -0.144 |
trick | -0.142 |
test | -0.129 |
new | -0.127 |
five | -0.121 |
Token | So |
Token ID | 1406 |
Feature activation | -1.65e-01 |
Pos logit contributions | |
rified | +0.135 |
wl | +0.132 |
x | +0.132 |
ppy | +0.132 |
lves | +0.130 |
Neg logit contributions | |
my | -0.091 |
new | -0.090 |
to | -0.089 |
the | -0.088 |
show | -0.086 |
Token | , |
Token ID | 11 |
Feature activation | +6.75e-02 |
Pos logit contributions | |
mie | +0.334 |
elope | +0.323 |
rified | +0.322 |
etsy | +0.320 |
jo | +0.318 |
Neg logit contributions | |
show | -0.201 |
trick | -0.201 |
new | -0.184 |
test | -0.180 |
grown | -0.176 |
Token | you |
Token ID | 345 |
Feature activation | -2.43e-01 |
Pos logit contributions | |
show | +0.126 |
new | +0.125 |
trick | +0.120 |
my | +0.114 |
your | +0.113 |
Neg logit contributions | |
etsy | -0.309 |
rified | -0.304 |
och | -0.303 |
jo | -0.301 |
elope | -0.300 |
Token | dirty |
Token ID | 11841 |
Feature activation | +2.05e-02 |
Pos logit contributions | |
mie | +0.424 |
bo | +0.408 |
hew | +0.383 |
ppy | +0.377 |
elope | +0.371 |
Neg logit contributions | |
trick | -0.319 |
show | -0.314 |
matching | -0.296 |
test | -0.295 |
grown | -0.294 |
Token | ." |
Token ID | 526 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
show | +0.022 |
trick | +0.022 |
test | +0.020 |
grown | +0.020 |
new | +0.020 |
Neg logit contributions | |
mie | -0.044 |
rified | -0.041 |
jo | -0.040 |
hew | -0.040 |
storms | -0.040 |
Token | She |
Token ID | 1375 |
Feature activation | -3.26e-01 |
Pos logit contributions | |
rified | +0.265 |
lor | +0.260 |
emic | +0.255 |
hew | +0.253 |
x | +0.253 |
Neg logit contributions | |
show | -0.181 |
the | -0.180 |
to | -0.178 |
his | -0.174 |
your | -0.168 |
Token | took |
Token ID | 1718 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
mie | +0.554 |
elope | +0.514 |
rified | +0.504 |
jo | +0.498 |
hew | +0.498 |
Neg logit contributions | |
trick | -0.503 |
show | -0.481 |
new | -0.478 |
test | -0.467 |
job | -0.458 |
Token | Harry |
Token ID | 5850 |
Feature activation | -4.51e-01 |
Pos logit contributions | |
show | +0.058 |
trick | +0.051 |
new | +0.050 |
the | +0.047 |
to | +0.045 |
Neg logit contributions | |
elope | -0.333 |
jo | -0.331 |
mie | -0.328 |
etsy | -0.324 |
hew | -0.324 |
Token | from |
Token ID | 422 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
mie | +1.077 |
ppy | +1.021 |
etsy | +1.019 |
elope | +1.014 |
bo | +1.003 |
Neg logit contributions | |
show | -0.376 |
trick | -0.373 |
test | -0.322 |
grown | -0.321 |
new | -0.313 |
Token | Tom |
Token ID | 4186 |
Feature activation | -3.57e-01 |
Pos logit contributions | |
show | +0.192 |
trick | +0.185 |
new | +0.174 |
test | +0.168 |
for | +0.167 |
Neg logit contributions | |
mie | -0.264 |
etsy | -0.255 |
rified | -0.253 |
hew | -0.252 |
elope | -0.251 |
Token | and |
Token ID | 290 |
Feature activation | +5.72e-02 |
Pos logit contributions | |
mie | +0.911 |
rified | +0.893 |
elope | +0.890 |
etsy | +0.885 |
jo | +0.881 |
Neg logit contributions | |
show | -0.251 |
trick | -0.241 |
new | -0.215 |
for | -0.203 |
test | -0.197 |
Token | rubbed |
Token ID | 31862 |
Feature activation | +3.67e-02 |
Pos logit contributions | |
show | +0.082 |
trick | +0.080 |
new | +0.075 |
test | +0.073 |
grown | +0.073 |
Neg logit contributions | |
mie | -0.104 |
rified | -0.102 |
elope | -0.101 |
jo | -0.101 |
etsy | -0.100 |
Token | his |
Token ID | 465 |
Feature activation | +1.36e-02 |
Pos logit contributions | |
show | +0.013 |
trick | +0.012 |
new | +0.011 |
the | +0.010 |
for | +0.010 |
Neg logit contributions | |
mie | -0.106 |
elope | -0.105 |
etsy | -0.105 |
hew | -0.104 |
jo | -0.104 |
Token | . |
Token ID | 13 |
Feature activation | -1.02e-01 |
Pos logit contributions | |
mie | +0.009 |
rified | +0.009 |
jo | +0.009 |
elope | +0.009 |
hew | +0.009 |
Neg logit contributions | |
show | -0.003 |
trick | -0.003 |
test | -0.003 |
new | -0.003 |
grown | -0.003 |
Token | The |
Token ID | 383 |
Feature activation | -4.25e-02 |
Pos logit contributions | |
mie | +0.183 |
rified | +0.181 |
lves | +0.177 |
hew | +0.176 |
elope | +0.175 |
Neg logit contributions | |
show | -0.146 |
to | -0.141 |
my | -0.137 |
the | -0.137 |
new | -0.134 |
Token | end |
Token ID | 886 |
Feature activation | -1.53e-02 |
Pos logit contributions | |
mie | +0.080 |
elope | +0.077 |
rified | +0.077 |
jo | +0.076 |
etsy | +0.076 |
Neg logit contributions | |
show | -0.057 |
trick | -0.057 |
new | -0.053 |
test | -0.051 |
grown | -0.051 |
Token | ! |
Token ID | 0 |
Feature activation | -8.51e-02 |
Pos logit contributions | |
mie | +0.035 |
rified | +0.034 |
elope | +0.034 |
etsy | +0.033 |
jo | +0.033 |
Neg logit contributions | |
show | -0.015 |
trick | -0.015 |
new | -0.014 |
test | -0.013 |
grown | -0.013 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.66e-01 |
Pos logit contributions | |
mie | +0.193 |
lves | +0.192 |
storms | +0.191 |
rified | +0.190 |
ppa | +0.188 |
Neg logit contributions | |
the | -0.085 |
my | -0.081 |
to | -0.081 |
show | -0.078 |
your | -0.075 |
Token | Harry |
Token ID | 18308 |
Feature activation | -4.51e-01 |
Pos logit contributions | |
ppa | +0.359 |
lves | +0.354 |
etsy | +0.350 |
mie | +0.350 |
ppy | +0.342 |
Neg logit contributions | |
the | -0.201 |
show | -0.193 |
to | -0.190 |
new | -0.185 |
trick | -0.171 |
Token | had |
Token ID | 550 |
Feature activation | -3.05e-02 |
Pos logit contributions | |
mie | +1.013 |
rified | +0.993 |
elope | +0.986 |
etsy | +0.984 |
jo | +0.975 |
Neg logit contributions | |
show | -0.448 |
trick | -0.432 |
new | -0.398 |
grown | -0.379 |
test | -0.377 |
Token | a |
Token ID | 257 |
Feature activation | +7.64e-02 |
Pos logit contributions | |
mie | +0.081 |
jo | +0.078 |
elope | +0.078 |
rified | +0.078 |
hew | +0.078 |
Neg logit contributions | |
trick | -0.019 |
show | -0.018 |
test | -0.015 |
new | -0.014 |
grown | -0.012 |
Token | new |
Token ID | 649 |
Feature activation | +4.20e-02 |
Pos logit contributions | |
show | +0.059 |
trick | +0.058 |
new | +0.052 |
test | +0.049 |
grown | +0.047 |
Neg logit contributions | |
mie | -0.189 |
rified | -0.185 |
elope | -0.185 |
jo | -0.183 |
etsy | -0.183 |
Token | model |
Token ID | 2746 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
trick | +0.052 |
show | +0.049 |
grown | +0.048 |
making | +0.047 |
my | +0.045 |
Neg logit contributions | |
mie | -0.078 |
elope | -0.077 |
jo | -0.074 |
etsy | -0.073 |
pillar | -0.071 |
Token | . |
Token ID | 13 |
Feature activation | -7.56e-02 |
Pos logit contributions | |
mie | +0.280 |
och | +0.268 |
jo | +0.266 |
rified | +0.266 |
pillar | +0.263 |
Neg logit contributions | |
trick | -0.158 |
grown | -0.150 |
show | -0.148 |
test | -0.140 |
work | -0.140 |
Token | to |
Token ID | 284 |
Feature activation | +1.79e-01 |
Pos logit contributions | |
show | +0.016 |
trick | +0.016 |
test | +0.012 |
grown | +0.012 |
new | +0.011 |
Neg logit contributions | |
mie | -0.074 |
etsy | -0.071 |
storms | -0.071 |
jo | -0.071 |
elope | -0.070 |
Token | talk |
Token ID | 1561 |
Feature activation | -7.94e-04 |
Pos logit contributions | |
new | +0.155 |
trick | +0.150 |
show | +0.144 |
magic | +0.140 |
for | +0.139 |
Neg logit contributions | |
mie | -0.405 |
elope | -0.386 |
etsy | -0.383 |
jo | -0.379 |
bo | -0.377 |
Token | to |
Token ID | 284 |
Feature activation | +1.82e-01 |
Pos logit contributions | |
mie | +0.002 |
jo | +0.002 |
storms | +0.002 |
rified | +0.002 |
etsy | +0.002 |
Neg logit contributions | |
show | -0.001 |
trick | -0.001 |
test | -0.001 |
design | -0.000 |
matching | -0.000 |
Token | Mr |
Token ID | 1770 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
show | +0.185 |
trick | +0.176 |
new | +0.165 |
test | +0.156 |
for | +0.154 |
Neg logit contributions | |
mie | -0.417 |
rified | -0.409 |
etsy | -0.407 |
elope | -0.402 |
hew | -0.401 |
Token | . |
Token ID | 13 |
Feature activation | -7.03e-02 |
Pos logit contributions | |
rified | +0.619 |
oped | +0.599 |
lves | +0.597 |
etsy | +0.588 |
hew | +0.585 |
Neg logit contributions | |
to | -0.076 |
for | -0.073 |
new | -0.065 |
, | -0.051 |
the | -0.047 |
Token | Jones |
Token ID | 5437 |
Feature activation | -4.50e-01 |
Pos logit contributions | |
jo | +0.118 |
hew | +0.116 |
wl | +0.113 |
emic | +0.112 |
pillar | +0.110 |
Neg logit contributions | |
my | -0.115 |
show | -0.113 |
the | -0.112 |
your | -0.109 |
new | -0.109 |
Token | . |
Token ID | 13 |
Feature activation | -6.84e-02 |
Pos logit contributions | |
mie | +0.970 |
rified | +0.945 |
elope | +0.942 |
etsy | +0.934 |
jo | +0.932 |
Neg logit contributions | |
show | -0.490 |
trick | -0.480 |
new | -0.442 |
test | -0.425 |
grown | -0.420 |
Token |
|
Token ID | 198 |
Feature activation | -1.60e-01 |
Pos logit contributions | |
hew | +0.125 |
rified | +0.124 |
x | +0.124 |
lves | +0.123 |
ppy | +0.123 |
Neg logit contributions | |
my | -0.102 |
your | -0.097 |
the | -0.097 |
to | -0.096 |
new | -0.096 |
Token |
|
Token ID | 198 |
Feature activation | -1.59e-01 |
Pos logit contributions | |
rified | +0.392 |
storms | +0.361 |
mie | +0.359 |
ley | +0.350 |
ppy | +0.347 |
Neg logit contributions | |
new | -0.175 |
to | -0.166 |
the | -0.162 |
show | -0.160 |
big | -0.159 |
Token | Mr |
Token ID | 5246 |
Feature activation | -2.04e-01 |
Pos logit contributions | |
rified | +0.327 |
storms | +0.318 |
mie | +0.299 |
jo | +0.293 |
elope | +0.293 |
Neg logit contributions | |
my | -0.207 |
your | -0.205 |
show | -0.203 |
five | -0.201 |
grown | -0.198 |
Token | . |
Token ID | 13 |
Feature activation | -6.42e-02 |
Pos logit contributions | |
rified | +0.589 |
etsy | +0.585 |
hew | +0.572 |
Mic | +0.571 |
lves | +0.570 |
Neg logit contributions | |
to | -0.096 |
new | -0.096 |
for | -0.090 |
the | -0.082 |
show | -0.078 |
Token | explore |
Token ID | 7301 |
Feature activation | -7.53e-03 |
Pos logit contributions | |
for | +0.163 |
new | +0.161 |
trick | +0.150 |
making | +0.150 |
red | +0.148 |
Neg logit contributions | |
elope | -0.341 |
mie | -0.339 |
jo | -0.322 |
etsy | -0.313 |
thel | -0.312 |
Token | the |
Token ID | 262 |
Feature activation | +7.15e-02 |
Pos logit contributions | |
mie | +0.017 |
elope | +0.016 |
jo | +0.016 |
rified | +0.016 |
storms | +0.016 |
Neg logit contributions | |
trick | -0.007 |
show | -0.006 |
test | -0.006 |
job | -0.006 |
grown | -0.005 |
Token | field |
Token ID | 2214 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
show | +0.095 |
trick | +0.093 |
new | +0.088 |
test | +0.085 |
grown | +0.084 |
Neg logit contributions | |
mie | -0.139 |
rified | -0.135 |
elope | -0.134 |
etsy | -0.134 |
jo | -0.134 |
Token | , |
Token ID | 11 |
Feature activation | +4.86e-02 |
Pos logit contributions | |
mie | +0.510 |
elope | +0.503 |
rified | +0.497 |
jo | +0.491 |
etsy | +0.487 |
Neg logit contributions | |
trick | -0.242 |
show | -0.237 |
test | -0.214 |
new | -0.212 |
grown | -0.209 |
Token | but |
Token ID | 475 |
Feature activation | -7.88e-02 |
Pos logit contributions | |
show | +0.051 |
trick | +0.050 |
new | +0.045 |
test | +0.044 |
grown | +0.043 |
Neg logit contributions | |
mie | -0.107 |
elope | -0.104 |
rified | -0.103 |
etsy | -0.102 |
jo | -0.102 |
Token | Harry |
Token ID | 5850 |
Feature activation | -4.49e-01 |
Pos logit contributions | |
mie | +0.158 |
jo | +0.155 |
rified | +0.153 |
hew | +0.152 |
etsy | +0.152 |
Neg logit contributions | |
show | -0.100 |
trick | -0.095 |
new | -0.091 |
for | -0.089 |
to | -0.088 |
Token | warned |
Token ID | 7728 |
Feature activation | +1.53e-02 |
Pos logit contributions | |
och | +0.807 |
rified | +0.800 |
Mic | +0.785 |
etsy | +0.784 |
jo | +0.780 |
Neg logit contributions | |
show | -0.640 |
trick | -0.621 |
to | -0.589 |
new | -0.587 |
grown | -0.580 |
Token | him |
Token ID | 683 |
Feature activation | -1.69e-01 |
Pos logit contributions | |
show | +0.021 |
trick | +0.020 |
new | +0.019 |
for | +0.019 |
to | +0.019 |
Neg logit contributions | |
mie | -0.030 |
etsy | -0.029 |
elope | -0.029 |
rified | -0.029 |
jo | -0.028 |
Token | not |
Token ID | 407 |
Feature activation | -4.31e-02 |
Pos logit contributions | |
mie | +0.414 |
elope | +0.404 |
rified | +0.402 |
jo | +0.400 |
etsy | +0.399 |
Neg logit contributions | |
show | -0.133 |
trick | -0.129 |
new | -0.112 |
test | -0.110 |
grown | -0.105 |
Token | to |
Token ID | 284 |
Feature activation | +1.89e-01 |
Pos logit contributions | |
mie | +0.111 |
elope | +0.110 |
jo | +0.109 |
etsy | +0.107 |
hew | +0.106 |
Neg logit contributions | |
show | -0.023 |
trick | -0.020 |
test | -0.016 |
matching | -0.015 |
design | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | -7.27e-02 |
Pos logit contributions | |
show | +0.181 |
trick | +0.179 |
new | +0.171 |
magic | +0.159 |
grown | +0.156 |
Neg logit contributions | |
mie | -0.415 |
elope | -0.407 |
etsy | -0.399 |
rified | -0.395 |
jo | -0.395 |
Token | excited |
Token ID | 6568 |
Feature activation | +4.07e-02 |
Pos logit contributions | |
trick | +0.097 |
show | +0.095 |
test | +0.086 |
new | +0.085 |
for | +0.082 |
Neg logit contributions | |
mie | -0.145 |
elope | -0.141 |
rified | -0.139 |
etsy | -0.137 |
jo | -0.137 |
Token | . |
Token ID | 13 |
Feature activation | -5.37e-02 |
Pos logit contributions | |
show | +0.036 |
trick | +0.035 |
new | +0.031 |
test | +0.030 |
grown | +0.030 |
Neg logit contributions | |
mie | -0.096 |
rified | -0.094 |
elope | -0.094 |
jo | -0.093 |
etsy | -0.093 |
Token |
|
Token ID | 198 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
pper | +0.126 |
ppy | +0.123 |
ppers | +0.123 |
ppa | +0.123 |
x | +0.120 |
Neg logit contributions | |
to | -0.061 |
the | -0.060 |
new | -0.052 |
my | -0.050 |
show | -0.047 |
Token |
|
Token ID | 198 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
ksh | +0.377 |
rified | +0.365 |
storms | +0.361 |
ley | +0.352 |
x | +0.345 |
Neg logit contributions | |
the | -0.121 |
new | -0.120 |
to | -0.113 |
big | -0.111 |
grown | -0.105 |
Token | Pen |
Token ID | 25553 |
Feature activation | -2.51e-01 |
Pos logit contributions | |
storms | +0.286 |
rified | +0.278 |
ksh | +0.276 |
ley | +0.272 |
isc | +0.267 |
Neg logit contributions | |
grown | -0.179 |
five | -0.170 |
show | -0.165 |
the | -0.164 |
my | -0.163 |
Token | elope |
Token ID | 47329 |
Feature activation | -4.49e-01 |
Pos logit contributions | |
och | +0.181 |
Mic | +0.160 |
mie | +0.150 |
rified | +0.148 |
lves | +0.146 |
Neg logit contributions | |
show | -0.649 |
trick | -0.628 |
grown | -0.625 |
the | -0.624 |
to | -0.616 |
Token | started |
Token ID | 2067 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
mie | +0.784 |
rified | +0.762 |
elope | +0.755 |
etsy | +0.753 |
jo | +0.748 |
Neg logit contributions | |
show | -0.672 |
trick | -0.658 |
new | -0.620 |
grown | -0.604 |
test | -0.602 |
Token | to |
Token ID | 284 |
Feature activation | +2.14e-01 |
Pos logit contributions | |
trick | +0.084 |
show | +0.082 |
my | +0.082 |
job | +0.077 |
test | +0.074 |
Neg logit contributions | |
elope | -0.264 |
jo | -0.260 |
mie | -0.257 |
storms | -0.254 |
Jean | -0.246 |
Token | call |
Token ID | 869 |
Feature activation | -2.47e-02 |
Pos logit contributions | |
trick | +0.237 |
new | +0.229 |
show | +0.226 |
for | +0.218 |
my | +0.209 |
Neg logit contributions | |
elope | -0.439 |
mie | -0.435 |
jo | -0.423 |
rified | -0.415 |
etsy | -0.403 |
Token | her |
Token ID | 607 |
Feature activation | +1.88e-02 |
Pos logit contributions | |
rified | +0.059 |
mie | +0.059 |
elope | +0.058 |
jo | +0.057 |
etsy | +0.056 |
Neg logit contributions | |
trick | -0.018 |
show | -0.017 |
grown | -0.015 |
test | -0.015 |
matching | -0.014 |
Token | mom |
Token ID | 1995 |
Feature activation | -3.10e-01 |
Pos logit contributions | |
show | +0.024 |
trick | +0.024 |
new | +0.022 |
test | +0.021 |
grown | +0.021 |
Neg logit contributions | |
mie | -0.037 |
elope | -0.036 |
rified | -0.036 |
jo | -0.035 |
etsy | -0.035 |
Token | bear |
Token ID | 6842 |
Feature activation | -2.58e-01 |
Pos logit contributions | |
rified | +0.067 |
mie | +0.065 |
elope | +0.063 |
pillar | +0.063 |
jo | +0.063 |
Neg logit contributions | |
trick | -0.053 |
show | -0.051 |
new | -0.051 |
grown | -0.051 |
test | -0.049 |
Token | . |
Token ID | 13 |
Feature activation | +1.53e-03 |
Pos logit contributions | |
mie | +0.637 |
hew | +0.621 |
rified | +0.619 |
etsy | +0.616 |
elope | +0.614 |
Neg logit contributions | |
show | -0.218 |
trick | -0.208 |
for | -0.189 |
to | -0.184 |
new | -0.177 |
Token | " |
Token ID | 366 |
Feature activation | -8.87e-02 |
Pos logit contributions | |
show | +0.002 |
trick | +0.002 |
new | +0.002 |
test | +0.002 |
grown | +0.002 |
Neg logit contributions | |
mie | -0.003 |
elope | -0.002 |
rified | -0.002 |
etsy | -0.002 |
jo | -0.002 |
Token | I |
Token ID | 40 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
mie | +0.138 |
hew | +0.138 |
elope | +0.134 |
rified | +0.133 |
jo | +0.133 |
Neg logit contributions | |
show | -0.154 |
trick | -0.151 |
grown | -0.141 |
five | -0.141 |
test | -0.138 |
Token | want |
Token ID | 765 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
mie | +0.210 |
elope | +0.204 |
etsy | +0.199 |
jo | +0.198 |
rified | +0.197 |
Neg logit contributions | |
trick | -0.157 |
show | -0.155 |
new | -0.154 |
grown | -0.144 |
my | -0.143 |
Token | to |
Token ID | 284 |
Feature activation | +2.58e-01 |
Pos logit contributions | |
trick | +0.065 |
show | +0.062 |
test | +0.060 |
job | +0.057 |
balance | +0.052 |
Neg logit contributions | |
mie | -0.247 |
Bear | -0.242 |
storms | -0.242 |
jo | -0.238 |
etsy | -0.236 |
Token | eat |
Token ID | 4483 |
Feature activation | +1.63e-01 |
Pos logit contributions | |
new | +0.240 |
for | +0.221 |
making | +0.204 |
trick | +0.201 |
red | +0.197 |
Neg logit contributions | |
elope | -0.535 |
mie | -0.520 |
bo | -0.511 |
etsy | -0.505 |
ppy | -0.503 |
Token | some |
Token ID | 617 |
Feature activation | +1.49e-01 |
Pos logit contributions | |
trick | +0.178 |
show | +0.173 |
test | +0.165 |
grown | +0.158 |
job | +0.155 |
Neg logit contributions | |
mie | -0.333 |
rified | -0.314 |
hew | -0.312 |
elope | -0.310 |
Hop | -0.308 |
Token | honey |
Token ID | 12498 |
Feature activation | -1.61e-01 |
Pos logit contributions | |
trick | +0.178 |
grown | +0.168 |
making | +0.168 |
test | +0.168 |
design | +0.159 |
Neg logit contributions | |
pillar | -0.252 |
mie | -0.245 |
Bear | -0.241 |
Hop | -0.238 |
rified | -0.234 |
Token | , |
Token ID | 11 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
mie | +0.366 |
elope | +0.352 |
etsy | +0.348 |
jo | +0.348 |
ppy | +0.346 |
Neg logit contributions | |
show | -0.146 |
trick | -0.145 |
new | -0.133 |
grown | -0.133 |
test | -0.130 |
Token | but |
Token ID | 475 |
Feature activation | -7.60e-05 |
Pos logit contributions | |
trick | +0.168 |
show | +0.166 |
test | +0.154 |
new | +0.152 |
grown | +0.149 |
Neg logit contributions | |
mie | -0.268 |
rified | -0.260 |
elope | -0.259 |
etsy | -0.257 |
jo | -0.255 |
Token | seems |
Token ID | 2331 |
Feature activation | +3.19e-02 |
Pos logit contributions | |
five | +0.001 |
trick | +0.001 |
show | +0.001 |
order | +0.001 |
style | +0.001 |
Neg logit contributions | |
mie | -0.002 |
hew | -0.001 |
pillar | -0.001 |
rified | -0.001 |
elope | -0.001 |
Token | fun |
Token ID | 1257 |
Feature activation | +3.06e-02 |
Pos logit contributions | |
show | +0.032 |
trick | +0.031 |
test | +0.029 |
balance | +0.027 |
skills | +0.026 |
Neg logit contributions | |
mie | -0.064 |
storms | -0.062 |
lves | -0.060 |
etsy | -0.060 |
rified | -0.059 |
Token | , |
Token ID | 11 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
trick | +0.026 |
show | +0.025 |
five | +0.023 |
test | +0.022 |
grown | +0.021 |
Neg logit contributions | |
mie | -0.069 |
storms | -0.066 |
jo | -0.065 |
rified | -0.064 |
hew | -0.064 |
Token | it |
Token ID | 340 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
trick | +0.179 |
show | +0.176 |
test | +0.166 |
new | +0.158 |
five | +0.156 |
Neg logit contributions | |
mie | -0.209 |
elope | -0.207 |
rified | -0.204 |
etsy | -0.198 |
hew | -0.198 |
Token | might |
Token ID | 1244 |
Feature activation | -1.73e-02 |
Pos logit contributions | |
mie | +0.196 |
etsy | +0.194 |
rified | +0.190 |
jo | +0.189 |
elope | +0.188 |
Neg logit contributions | |
show | -0.200 |
trick | -0.193 |
new | -0.186 |
test | -0.180 |
grown | -0.180 |
Token | be |
Token ID | 307 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
elope | +0.034 |
mie | +0.033 |
bo | +0.032 |
storms | +0.032 |
ppy | +0.032 |
Neg logit contributions | |
new | -0.016 |
red | -0.015 |
show | -0.015 |
five | -0.015 |
skills | -0.015 |
Token | harsh |
Token ID | 11859 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
show | +0.145 |
trick | +0.136 |
test | +0.136 |
design | +0.118 |
quiz | +0.116 |
Neg logit contributions | |
mie | -0.231 |
elope | -0.219 |
storms | -0.213 |
hew | -0.211 |
lves | -0.211 |
Token | and |
Token ID | 290 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
show | +0.101 |
trick | +0.098 |
test | +0.086 |
new | +0.081 |
grown | +0.081 |
Neg logit contributions | |
mie | -0.276 |
rified | -0.271 |
storms | -0.269 |
jo | -0.267 |
elope | -0.264 |
Token | it |
Token ID | 340 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
trick | +0.145 |
show | +0.141 |
test | +0.126 |
five | +0.125 |
new | +0.124 |
Neg logit contributions | |
mie | -0.235 |
elope | -0.227 |
storms | -0.225 |
rified | -0.224 |
etsy | -0.220 |
Token | might |
Token ID | 1244 |
Feature activation | -1.53e-02 |
Pos logit contributions | |
mie | +0.192 |
etsy | +0.185 |
rified | +0.185 |
elope | +0.184 |
jo | +0.183 |
Neg logit contributions | |
show | -0.196 |
trick | -0.192 |
new | -0.183 |
test | -0.178 |
grown | -0.177 |
Token | not |
Token ID | 407 |
Feature activation | +2.50e-02 |
Pos logit contributions | |
elope | +0.031 |
mie | +0.030 |
bo | +0.029 |
storms | +0.029 |
etsy | +0.029 |
Neg logit contributions | |
new | -0.014 |
show | -0.014 |
trick | -0.013 |
red | -0.013 |
style | -0.013 |
Token | The |
Token ID | 464 |
Feature activation | +4.77e-02 |
Pos logit contributions | |
mie | +0.104 |
elope | +0.099 |
rified | +0.099 |
jo | +0.097 |
etsy | +0.097 |
Neg logit contributions | |
show | -0.150 |
trick | -0.149 |
new | -0.142 |
test | -0.139 |
grown | -0.138 |
Token | old |
Token ID | 1468 |
Feature activation | +4.63e-02 |
Pos logit contributions | |
trick | +0.049 |
show | +0.043 |
my | +0.042 |
your | +0.042 |
balance | +0.041 |
Neg logit contributions | |
mie | -0.100 |
thel | -0.093 |
elope | -0.093 |
rified | -0.091 |
keepers | -0.091 |
Token | man |
Token ID | 582 |
Feature activation | -2.64e-01 |
Pos logit contributions | |
trick | +0.086 |
show | +0.083 |
test | +0.079 |
my | +0.078 |
magic | +0.078 |
Neg logit contributions | |
mie | -0.064 |
elope | -0.060 |
thel | -0.058 |
rified | -0.057 |
jo | -0.057 |
Token | gave |
Token ID | 2921 |
Feature activation | +2.17e-01 |
Pos logit contributions | |
mie | +0.470 |
elope | +0.449 |
rified | +0.444 |
jo | +0.440 |
etsy | +0.439 |
Neg logit contributions | |
trick | -0.385 |
show | -0.385 |
new | -0.363 |
test | -0.355 |
grown | -0.346 |
Token | Sally |
Token ID | 25737 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
show | +0.271 |
new | +0.264 |
trick | +0.259 |
for | +0.253 |
the | +0.249 |
Neg logit contributions | |
jo | -0.444 |
elope | -0.435 |
storms | -0.431 |
mie | -0.425 |
uddy | -0.420 |
Token | a |
Token ID | 257 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
mie | +0.689 |
elope | +0.663 |
etsy | +0.661 |
rified | +0.661 |
jo | +0.655 |
Neg logit contributions | |
show | -0.144 |
trick | -0.141 |
test | -0.110 |
new | -0.100 |
grown | -0.099 |
Token | m |
Token ID | 285 |
Feature activation | +1.65e-02 |
Pos logit contributions | |
making | +0.161 |
trick | +0.152 |
ags | +0.150 |
adding | +0.147 |
stacking | +0.145 |
Neg logit contributions | |
etsy | -0.341 |
pillar | -0.322 |
mie | -0.322 |
apa | -0.317 |
och | -0.314 |
Token | op |
Token ID | 404 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
show | +0.032 |
trick | +0.031 |
test | +0.031 |
job | +0.030 |
for | +0.030 |
Neg logit contributions | |
storms | -0.022 |
elope | -0.021 |
Mic | -0.021 |
rified | -0.021 |
mie | -0.020 |
Token | and |
Token ID | 290 |
Feature activation | +1.47e-01 |
Pos logit contributions | |
mie | +0.270 |
rified | +0.259 |
jo | +0.258 |
etsy | +0.258 |
elope | +0.255 |
Neg logit contributions | |
trick | -0.100 |
show | -0.098 |
test | -0.086 |
grown | -0.085 |
new | -0.084 |
Token | a |
Token ID | 257 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
trick | +0.214 |
show | +0.213 |
new | +0.198 |
test | +0.197 |
grown | +0.191 |
Neg logit contributions | |
mie | -0.262 |
elope | -0.250 |
rified | -0.249 |
etsy | -0.247 |
jo | -0.245 |
Token | bucket |
Token ID | 19236 |
Feature activation | -1.29e-01 |
Pos logit contributions | |
trick | +0.181 |
show | +0.168 |
my | +0.166 |
making | +0.162 |
your | +0.161 |
Neg logit contributions | |
etsy | -0.327 |
mie | -0.323 |
pillar | -0.312 |
och | -0.308 |
apa | -0.301 |
Token | appeared |
Token ID | 4120 |
Feature activation | +7.25e-02 |
Pos logit contributions | |
show | +0.088 |
test | +0.087 |
trick | +0.087 |
quiz | +0.082 |
work | +0.081 |
Neg logit contributions | |
mie | -0.098 |
hew | -0.088 |
Max | -0.088 |
elope | -0.087 |
etsy | -0.086 |
Token | . |
Token ID | 13 |
Feature activation | +3.49e-03 |
Pos logit contributions | |
show | +0.079 |
trick | +0.077 |
test | +0.072 |
grown | +0.066 |
design | +0.065 |
Neg logit contributions | |
mie | -0.154 |
rified | -0.150 |
elope | -0.148 |
etsy | -0.145 |
jo | -0.144 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -8.10e-02 |
Pos logit contributions | |
new | +0.005 |
show | +0.005 |
to | +0.005 |
for | +0.005 |
the | +0.005 |
Neg logit contributions | |
rified | -0.006 |
mie | -0.006 |
ppy | -0.006 |
thel | -0.006 |
ppa | -0.006 |
Token | Once |
Token ID | 7454 |
Feature activation | +2.05e-02 |
Pos logit contributions | |
ppa | +0.185 |
lves | +0.182 |
etsy | +0.179 |
mie | +0.177 |
hew | +0.176 |
Neg logit contributions | |
the | -0.094 |
to | -0.092 |
show | -0.090 |
new | -0.088 |
big | -0.081 |
Token | upon |
Token ID | 2402 |
Feature activation | +6.44e-02 |
Pos logit contributions | |
new | +0.025 |
to | +0.023 |
for | +0.023 |
show | +0.023 |
the | +0.023 |
Neg logit contributions | |
rified | -0.044 |
elope | -0.044 |
thel | -0.044 |
mie | -0.043 |
Mic | -0.043 |
Token | a |
Token ID | 257 |
Feature activation | +1.63e-01 |
Pos logit contributions | |
show | +0.012 |
trick | +0.010 |
new | +0.009 |
grown | +0.005 |
to | +0.004 |
Neg logit contributions | |
mie | -0.200 |
etsy | -0.198 |
rified | -0.198 |
och | -0.197 |
jo | -0.195 |
Token | time |
Token ID | 640 |
Feature activation | +1.22e-01 |
Pos logit contributions | |
show | +0.050 |
trick | +0.049 |
new | +0.040 |
grown | +0.027 |
test | +0.027 |
Neg logit contributions | |
mie | -0.484 |
rified | -0.481 |
och | -0.469 |
etsy | -0.466 |
elope | -0.464 |
Token | , |
Token ID | 11 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
show | +0.034 |
trick | +0.030 |
new | +0.029 |
for | +0.023 |
to | +0.021 |
Neg logit contributions | |
mie | -0.360 |
rified | -0.354 |
elope | -0.353 |
jo | -0.351 |
etsy | -0.349 |
Token | there |
Token ID | 612 |
Feature activation | -5.72e-02 |
Pos logit contributions | |
show | +0.168 |
trick | +0.163 |
new | +0.158 |
for | +0.149 |
test | +0.148 |
Neg logit contributions | |
mie | -0.297 |
rified | -0.288 |
jo | -0.288 |
elope | -0.287 |
etsy | -0.286 |
Token | was |
Token ID | 373 |
Feature activation | +1.48e-01 |
Pos logit contributions | |
mie | +0.120 |
etsy | +0.119 |
ppa | +0.118 |
och | +0.117 |
rified | +0.117 |
Neg logit contributions | |
show | -0.067 |
trick | -0.064 |
new | -0.062 |
the | -0.060 |
grown | -0.059 |
Token | a |
Token ID | 257 |
Feature activation | +1.65e-01 |
Pos logit contributions | |
show | +0.052 |
trick | +0.048 |
new | +0.041 |
grown | +0.031 |
test | +0.030 |
Neg logit contributions | |
mie | -0.431 |
rified | -0.423 |
elope | -0.422 |
jo | -0.421 |
etsy | -0.420 |
Token | bird |
Token ID | 6512 |
Feature activation | -2.39e-01 |
Pos logit contributions | |
show | +0.073 |
trick | +0.073 |
new | +0.067 |
test | +0.067 |
for | +0.066 |
Neg logit contributions | |
mie | -0.070 |
elope | -0.069 |
rified | -0.068 |
etsy | -0.067 |
hew | -0.066 |
Token | wanted |
Token ID | 2227 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
mie | +0.386 |
elope | +0.374 |
jo | +0.366 |
etsy | +0.357 |
rified | +0.357 |
Neg logit contributions | |
trick | -0.389 |
show | -0.385 |
test | -0.364 |
new | -0.359 |
grown | -0.350 |
Token | to |
Token ID | 284 |
Feature activation | +2.60e-01 |
Pos logit contributions | |
show | +0.095 |
test | +0.094 |
job | +0.091 |
trick | +0.091 |
design | +0.086 |
Neg logit contributions | |
ety | -0.201 |
elope | -0.197 |
Le | -0.196 |
Rex | -0.195 |
mie | -0.195 |
Token | un |
Token ID | 555 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
new | +0.274 |
for | +0.267 |
matching | +0.260 |
math | +0.258 |
making | +0.255 |
Neg logit contributions | |
elope | -0.464 |
Bird | -0.460 |
etsy | -0.449 |
bo | -0.445 |
Mic | -0.444 |
Token | pack |
Token ID | 8002 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
show | +0.174 |
trick | +0.172 |
new | +0.160 |
test | +0.155 |
grown | +0.153 |
Neg logit contributions | |
mie | -0.262 |
rified | -0.255 |
elope | -0.254 |
jo | -0.252 |
etsy | -0.251 |
Token | the |
Token ID | 262 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
trick | +0.118 |
test | +0.105 |
show | +0.099 |
design | +0.099 |
quiz | +0.094 |
Neg logit contributions | |
mie | -0.222 |
hew | -0.208 |
rified | -0.204 |
la | -0.203 |
thel | -0.191 |
Token | bag |
Token ID | 6131 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
trick | +0.269 |
show | +0.267 |
test | +0.243 |
making | +0.242 |
grown | +0.239 |
Neg logit contributions | |
mie | -0.266 |
etsy | -0.249 |
rified | -0.248 |
hew | -0.244 |
elope | -0.240 |
Token | . |
Token ID | 13 |
Feature activation | +4.42e-03 |
Pos logit contributions | |
mie | +0.255 |
rified | +0.242 |
hew | +0.241 |
etsy | +0.239 |
hel | +0.238 |
Neg logit contributions | |
trick | -0.156 |
show | -0.149 |
test | -0.146 |
grown | -0.141 |
magic | -0.136 |
Token | The |
Token ID | 383 |
Feature activation | +4.71e-02 |
Pos logit contributions | |
show | +0.006 |
new | +0.006 |
trick | +0.006 |
to | +0.006 |
the | +0.006 |
Neg logit contributions | |
mie | -0.008 |
rified | -0.008 |
etsy | -0.008 |
jo | -0.008 |
ppa | -0.008 |
Token | bag |
Token ID | 6131 |
Feature activation | -1.31e-01 |
Pos logit contributions | |
show | +0.079 |
trick | +0.079 |
new | +0.074 |
test | +0.073 |
grown | +0.072 |
Neg logit contributions | |
mie | -0.073 |
elope | -0.070 |
rified | -0.070 |
etsy | -0.069 |
jo | -0.068 |
Token | had |
Token ID | 550 |
Feature activation | +5.88e-02 |
Pos logit contributions | |
mie | +0.213 |
elope | +0.203 |
rified | +0.203 |
etsy | +0.200 |
jo | +0.200 |
Neg logit contributions | |
trick | -0.207 |
show | -0.207 |
test | -0.194 |
new | -0.193 |
grown | -0.187 |
Token | and |
Token ID | 290 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
mie | +0.196 |
rified | +0.190 |
elope | +0.190 |
jo | +0.188 |
etsy | +0.188 |
Neg logit contributions | |
show | -0.064 |
trick | -0.064 |
test | -0.055 |
new | -0.055 |
grown | -0.053 |
Token | gave |
Token ID | 2921 |
Feature activation | +2.05e-01 |
Pos logit contributions | |
trick | +0.224 |
show | +0.213 |
test | +0.212 |
job | +0.208 |
new | +0.208 |
Neg logit contributions | |
mie | -0.220 |
elope | -0.204 |
rified | -0.201 |
etsy | -0.194 |
thel | -0.192 |
Token | them |
Token ID | 606 |
Feature activation | -7.98e-02 |
Pos logit contributions | |
the | +0.179 |
new | +0.175 |
show | +0.170 |
to | +0.164 |
for | +0.161 |
Neg logit contributions | |
etsy | -0.528 |
elope | -0.526 |
och | -0.523 |
ongo | -0.517 |
storms | -0.513 |
Token | some |
Token ID | 617 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
mie | +0.209 |
rified | +0.203 |
elope | +0.202 |
etsy | +0.201 |
jo | +0.200 |
Neg logit contributions | |
show | -0.049 |
trick | -0.048 |
new | -0.039 |
test | -0.039 |
grown | -0.037 |
Token | cookies |
Token ID | 14746 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
trick | +0.207 |
test | +0.199 |
show | +0.192 |
making | +0.187 |
spelling | +0.187 |
Neg logit contributions | |
mie | -0.265 |
etsy | -0.259 |
apa | -0.251 |
pillar | -0.245 |
hew | -0.243 |
Token | . |
Token ID | 13 |
Feature activation | +1.09e-02 |
Pos logit contributions | |
mie | +0.215 |
etsy | +0.205 |
elope | +0.203 |
rified | +0.201 |
jo | +0.200 |
Neg logit contributions | |
trick | -0.118 |
show | -0.117 |
test | -0.111 |
new | -0.108 |
grown | -0.107 |
Token | They |
Token ID | 1119 |
Feature activation | -2.25e-01 |
Pos logit contributions | |
show | +0.020 |
trick | +0.020 |
new | +0.019 |
test | +0.019 |
grown | +0.018 |
Neg logit contributions | |
mie | -0.015 |
rified | -0.015 |
elope | -0.015 |
jo | -0.014 |
etsy | -0.014 |
Token | ate |
Token ID | 15063 |
Feature activation | +2.18e-01 |
Pos logit contributions | |
mie | +0.397 |
elope | +0.375 |
rified | +0.372 |
etsy | +0.372 |
jo | +0.363 |
Neg logit contributions | |
trick | -0.332 |
show | -0.323 |
new | -0.311 |
test | -0.306 |
magic | -0.295 |
Token | and |
Token ID | 290 |
Feature activation | +1.39e-01 |
Pos logit contributions | |
trick | +0.293 |
test | +0.285 |
show | +0.284 |
job | +0.258 |
style | +0.255 |
Neg logit contributions | |
mie | -0.396 |
rified | -0.374 |
etsy | -0.368 |
hew | -0.358 |
elope | -0.355 |
Token | talked |
Token ID | 6619 |
Feature activation | +8.42e-02 |
Pos logit contributions | |
trick | +0.208 |
show | +0.208 |
new | +0.196 |
test | +0.193 |
grown | +0.187 |
Neg logit contributions | |
mie | -0.243 |
elope | -0.234 |
rified | -0.231 |
etsy | -0.230 |
jo | -0.227 |
Token | and |
Token ID | 290 |
Feature activation | +1.40e-01 |
Pos logit contributions | |
test | +0.111 |
show | +0.107 |
trick | +0.106 |
design | +0.102 |
my | +0.100 |
Neg logit contributions | |
mie | -0.142 |
jo | -0.140 |
storms | -0.136 |
hel | -0.136 |
rified | -0.136 |
Token | What |
Token ID | 2061 |
Feature activation | -6.54e-02 |
Pos logit contributions | |
mie | +0.155 |
jo | +0.152 |
rified | +0.151 |
elope | +0.150 |
etsy | +0.149 |
Neg logit contributions | |
trick | -0.199 |
show | -0.199 |
grown | -0.195 |
making | -0.190 |
test | -0.185 |
Token | have |
Token ID | 423 |
Feature activation | +5.75e-02 |
Pos logit contributions | |
mie | +0.107 |
hew | +0.101 |
rified | +0.101 |
elope | +0.100 |
pillar | +0.100 |
Neg logit contributions | |
show | -0.095 |
trick | -0.094 |
new | -0.084 |
work | -0.084 |
five | -0.084 |
Token | you |
Token ID | 345 |
Feature activation | -1.67e-01 |
Pos logit contributions | |
trick | +0.110 |
show | +0.106 |
answer | +0.099 |
design | +0.099 |
test | +0.098 |
Neg logit contributions | |
mie | -0.074 |
hew | -0.069 |
storms | -0.065 |
rified | -0.065 |
la | -0.064 |
Token | done |
Token ID | 1760 |
Feature activation | +1.48e-01 |
Pos logit contributions | |
mie | +0.298 |
hew | +0.288 |
ppy | +0.268 |
ppa | +0.258 |
bo | +0.256 |
Neg logit contributions | |
trick | -0.231 |
new | -0.230 |
skills | -0.210 |
math | -0.206 |
show | -0.200 |
Token | ? |
Token ID | 30 |
Feature activation | +8.93e-03 |
Pos logit contributions | |
trick | +0.157 |
show | +0.147 |
new | +0.146 |
chance | +0.146 |
power | +0.140 |
Neg logit contributions | |
mie | -0.282 |
jo | -0.264 |
storms | -0.263 |
hel | -0.263 |
hew | -0.259 |
Token | This |
Token ID | 770 |
Feature activation | +4.82e-02 |
Pos logit contributions | |
show | +0.017 |
trick | +0.016 |
test | +0.016 |
new | +0.016 |
grown | +0.015 |
Neg logit contributions | |
mie | -0.012 |
elope | -0.012 |
rified | -0.012 |
jo | -0.012 |
etsy | -0.012 |
Token | is |
Token ID | 318 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
show | +0.066 |
trick | +0.063 |
for | +0.058 |
help | +0.058 |
work | +0.058 |
Neg logit contributions | |
mie | -0.080 |
elope | -0.077 |
bo | -0.077 |
hew | -0.077 |
jo | -0.076 |
Token | not |
Token ID | 407 |
Feature activation | +1.40e-02 |
Pos logit contributions | |
trick | +0.106 |
show | +0.104 |
design | +0.095 |
skills | +0.094 |
test | +0.092 |
Neg logit contributions | |
mie | -0.212 |
ppy | -0.204 |
bo | -0.203 |
Max | -0.203 |
mi | -0.203 |
Token | fun |
Token ID | 1257 |
Feature activation | +2.31e-02 |
Pos logit contributions | |
show | +0.012 |
trick | +0.011 |
test | +0.010 |
design | +0.010 |
skills | +0.009 |
Neg logit contributions | |
mie | -0.030 |
hew | -0.028 |
ppy | -0.028 |
ety | -0.028 |
ppa | -0.027 |
Token | . |
Token ID | 13 |
Feature activation | -1.18e-02 |
Pos logit contributions | |
show | +0.030 |
new | +0.030 |
trick | +0.029 |
magic | +0.029 |
grown | +0.029 |
Neg logit contributions | |
jo | -0.038 |
mie | -0.038 |
hew | -0.037 |
storms | -0.036 |
uddy | -0.035 |
Token | This |
Token ID | 770 |
Feature activation | +5.02e-02 |
Pos logit contributions | |
mie | +0.018 |
elope | +0.017 |
rified | +0.017 |
etsy | +0.017 |
jo | +0.017 |
Neg logit contributions | |
show | -0.020 |
my | -0.019 |
trick | -0.019 |
your | -0.019 |
test | -0.019 |
Token | in |
Token ID | 287 |
Feature activation | +2.04e-01 |
Pos logit contributions | |
trick | +0.063 |
show | +0.059 |
test | +0.056 |
new | +0.053 |
job | +0.052 |
Neg logit contributions | |
mie | -0.169 |
rified | -0.167 |
elope | -0.167 |
storms | -0.162 |
jo | -0.161 |
Token | the |
Token ID | 262 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
trick | +0.144 |
show | +0.131 |
test | +0.121 |
job | +0.115 |
grown | +0.115 |
Neg logit contributions | |
mie | -0.489 |
rified | -0.486 |
elope | -0.482 |
jo | -0.470 |
pillar | -0.455 |
Token | garden |
Token ID | 11376 |
Feature activation | -1.70e-01 |
Pos logit contributions | |
show | +0.145 |
trick | +0.139 |
new | +0.133 |
grown | +0.125 |
test | +0.125 |
Neg logit contributions | |
mie | -0.292 |
rified | -0.284 |
jo | -0.284 |
elope | -0.284 |
etsy | -0.282 |
Token | with |
Token ID | 351 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
mie | +0.382 |
elope | +0.375 |
rified | +0.373 |
jo | +0.366 |
etsy | +0.365 |
Neg logit contributions | |
trick | -0.160 |
show | -0.160 |
test | -0.143 |
new | -0.140 |
grown | -0.134 |
Token | her |
Token ID | 607 |
Feature activation | +3.92e-02 |
Pos logit contributions | |
show | +0.108 |
trick | +0.102 |
new | +0.091 |
grown | +0.080 |
test | +0.079 |
Neg logit contributions | |
mie | -0.462 |
elope | -0.453 |
jo | -0.451 |
rified | -0.451 |
etsy | -0.450 |
Token | . |
Token ID | 13 |
Feature activation | -2.77e-02 |
Pos logit contributions | |
show | +0.040 |
trick | +0.039 |
test | +0.035 |
new | +0.035 |
grown | +0.034 |
Neg logit contributions | |
mie | -0.087 |
elope | -0.085 |
etsy | -0.084 |
rified | -0.084 |
jo | -0.083 |
Token | From |
Token ID | 3574 |
Feature activation | +1.93e-03 |
Pos logit contributions | |
ppy | +0.052 |
mie | +0.052 |
ppa | +0.051 |
rified | +0.051 |
hew | +0.051 |
Neg logit contributions | |
show | -0.037 |
to | -0.037 |
the | -0.036 |
new | -0.036 |
his | -0.036 |
Token | that |
Token ID | 326 |
Feature activation | -2.29e-04 |
Pos logit contributions | |
show | +0.002 |
trick | +0.002 |
new | +0.002 |
test | +0.002 |
grown | +0.002 |
Neg logit contributions | |
mie | -0.004 |
elope | -0.004 |
rified | -0.004 |
etsy | -0.004 |
jo | -0.004 |
Token | day |
Token ID | 1110 |
Feature activation | -2.11e-02 |
Pos logit contributions | |
elope | +0.001 |
mie | +0.001 |
jo | +0.000 |
rified | +0.000 |
etsy | +0.000 |
Neg logit contributions | |
trick | -0.000 |
show | -0.000 |
new | -0.000 |
test | -0.000 |
grown | -0.000 |
Token | the |
Token ID | 262 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
mie | +0.049 |
elope | +0.047 |
rified | +0.047 |
jo | +0.047 |
etsy | +0.047 |
Neg logit contributions | |
show | -0.020 |
trick | -0.020 |
new | -0.018 |
test | -0.017 |
grown | -0.017 |
Token | little |
Token ID | 1310 |
Feature activation | -1.97e-03 |
Pos logit contributions | |
show | +0.185 |
trick | +0.181 |
new | +0.171 |
grown | +0.164 |
test | +0.164 |
Neg logit contributions | |
mie | -0.276 |
jo | -0.269 |
rified | -0.268 |
elope | -0.268 |
etsy | -0.263 |
Token | good |
Token ID | 922 |
Feature activation | +1.29e-01 |
Pos logit contributions | |
trick | +0.097 |
show | +0.095 |
test | +0.081 |
making | +0.080 |
for | +0.078 |
Neg logit contributions | |
mie | -0.332 |
elope | -0.320 |
etsy | -0.316 |
rified | -0.311 |
hew | -0.308 |
Token | helper |
Token ID | 31904 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
show | +0.170 |
trick | +0.168 |
new | +0.156 |
test | +0.152 |
grown | +0.150 |
Neg logit contributions | |
mie | -0.250 |
elope | -0.241 |
rified | -0.241 |
etsy | -0.238 |
jo | -0.238 |
Token | ." |
Token ID | 526 |
Feature activation | -6.31e-02 |
Pos logit contributions | |
mie | +0.358 |
rified | +0.348 |
elope | +0.348 |
etsy | +0.345 |
jo | +0.345 |
Neg logit contributions | |
show | -0.132 |
trick | -0.129 |
new | -0.116 |
test | -0.111 |
grown | -0.109 |
Token | And |
Token ID | 843 |
Feature activation | -4.92e-02 |
Pos logit contributions | |
rified | +0.119 |
jo | +0.117 |
elope | +0.115 |
wl | +0.114 |
thel | +0.114 |
Neg logit contributions | |
show | -0.089 |
new | -0.088 |
for | -0.087 |
to | -0.084 |
the | -0.083 |
Token | from |
Token ID | 422 |
Feature activation | +1.98e-01 |
Pos logit contributions | |
mie | +0.102 |
rified | +0.100 |
elope | +0.099 |
jo | +0.099 |
etsy | +0.098 |
Neg logit contributions | |
show | -0.059 |
trick | -0.057 |
new | -0.053 |
for | -0.051 |
grown | -0.051 |
Token | then |
Token ID | 788 |
Feature activation | -4.50e-02 |
Pos logit contributions | |
trick | +0.267 |
test | +0.240 |
show | +0.238 |
grown | +0.233 |
new | +0.231 |
Neg logit contributions | |
mie | -0.367 |
elope | -0.346 |
rified | -0.345 |
thel | -0.336 |
jo | -0.332 |
Token | on |
Token ID | 319 |
Feature activation | +2.10e-01 |
Pos logit contributions | |
mie | +0.104 |
elope | +0.099 |
rified | +0.098 |
och | +0.098 |
hew | +0.098 |
Neg logit contributions | |
trick | -0.045 |
show | -0.037 |
new | -0.037 |
power | -0.035 |
test | -0.034 |
Token | , |
Token ID | 11 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
trick | +0.211 |
show | +0.178 |
test | +0.169 |
power | +0.167 |
answer | +0.159 |
Neg logit contributions | |
mie | -0.485 |
elope | -0.469 |
etsy | -0.465 |
jo | -0.453 |
rified | -0.443 |
Token | Lily |
Token ID | 20037 |
Feature activation | -2.40e-01 |
Pos logit contributions | |
trick | +0.176 |
show | +0.170 |
test | +0.159 |
new | +0.156 |
grown | +0.154 |
Neg logit contributions | |
mie | -0.219 |
etsy | -0.209 |
elope | -0.209 |
rified | -0.205 |
jo | -0.201 |
Token | always |
Token ID | 1464 |
Feature activation | +8.12e-02 |
Pos logit contributions | |
mie | +0.402 |
elope | +0.381 |
jo | +0.373 |
hew | +0.364 |
ppy | +0.359 |
Neg logit contributions | |
trick | -0.381 |
show | -0.371 |
new | -0.359 |
test | -0.358 |
my | -0.340 |
Token | helped |
Token ID | 4193 |
Feature activation | +1.92e-01 |
Pos logit contributions | |
trick | +0.149 |
new | +0.141 |
show | +0.139 |
test | +0.138 |
math | +0.135 |
Neg logit contributions | |
mie | -0.116 |
elope | -0.108 |
bo | -0.101 |
ppy | -0.101 |
thel | -0.096 |
Token | whole |
Token ID | 2187 |
Feature activation | +7.79e-02 |
Pos logit contributions | |
show | +0.144 |
trick | +0.143 |
new | +0.130 |
test | +0.128 |
grown | +0.126 |
Neg logit contributions | |
mie | -0.226 |
rified | -0.220 |
elope | -0.219 |
jo | -0.216 |
etsy | -0.214 |
Token | bedroom |
Token ID | 14043 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
trick | +0.114 |
show | +0.112 |
test | +0.103 |
making | +0.100 |
help | +0.099 |
Neg logit contributions | |
mie | -0.134 |
elope | -0.132 |
jo | -0.129 |
rified | -0.128 |
pillar | -0.127 |
Token | ! |
Token ID | 0 |
Feature activation | -5.25e-02 |
Pos logit contributions | |
elope | +0.329 |
storms | +0.321 |
rified | +0.318 |
jo | +0.313 |
mie | +0.313 |
Neg logit contributions | |
trick | -0.255 |
show | -0.220 |
test | -0.218 |
my | -0.216 |
your | -0.207 |
Token | And |
Token ID | 843 |
Feature activation | -8.01e-02 |
Pos logit contributions | |
mie | +0.114 |
ppa | +0.110 |
ppy | +0.109 |
hew | +0.107 |
rified | +0.106 |
Neg logit contributions | |
show | -0.063 |
the | -0.059 |
new | -0.058 |
grown | -0.057 |
to | -0.057 |
Token | on |
Token ID | 319 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
mie | +0.152 |
elope | +0.150 |
rified | +0.146 |
hew | +0.143 |
etsy | +0.139 |
Neg logit contributions | |
trick | -0.102 |
show | -0.093 |
test | -0.093 |
new | -0.085 |
grown | -0.083 |
Token | this |
Token ID | 428 |
Feature activation | +3.61e-02 |
Pos logit contributions | |
trick | +0.148 |
show | +0.137 |
test | +0.121 |
style | +0.103 |
new | +0.102 |
Neg logit contributions | |
elope | -0.447 |
mie | -0.440 |
rified | -0.421 |
jo | -0.420 |
hew | -0.418 |
Token | enormous |
Token ID | 9812 |
Feature activation | +7.25e-02 |
Pos logit contributions | |
trick | +0.036 |
ags | +0.036 |
show | +0.034 |
skills | +0.034 |
your | +0.033 |
Neg logit contributions | |
elope | -0.072 |
jo | -0.071 |
abel | -0.065 |
hew | -0.064 |
mie | -0.064 |
Token | bed |
Token ID | 3996 |
Feature activation | -1.47e-01 |
Pos logit contributions | |
trick | +0.095 |
show | +0.091 |
your | +0.089 |
making | +0.089 |
new | +0.087 |
Neg logit contributions | |
elope | -0.120 |
jo | -0.118 |
pillar | -0.112 |
rified | -0.111 |
mie | -0.110 |
Token | slept |
Token ID | 21256 |
Feature activation | +3.45e-02 |
Pos logit contributions | |
mie | +0.259 |
elope | +0.255 |
rified | +0.253 |
jo | +0.252 |
etsy | +0.248 |
Neg logit contributions | |
show | -0.211 |
trick | -0.210 |
new | -0.196 |
test | -0.194 |
work | -0.186 |
Token | the |
Token ID | 262 |
Feature activation | +1.22e-01 |
Pos logit contributions | |
test | +0.035 |
show | +0.034 |
trick | +0.034 |
turn | +0.031 |
style | +0.030 |
Neg logit contributions | |
rified | -0.071 |
elope | -0.070 |
mie | -0.068 |
lor | -0.066 |
thel | -0.066 |
Token | most |
Token ID | 749 |
Feature activation | +5.70e-02 |
Pos logit contributions | |
trick | +0.125 |
show | +0.123 |
new | +0.111 |
test | +0.109 |
grown | +0.106 |
Neg logit contributions | |
mie | -0.255 |
rified | -0.254 |
elope | -0.251 |
jo | -0.246 |
hew | -0.243 |
Token | at |
Token ID | 379 |
Feature activation | +2.31e-01 |
Pos logit contributions | |
trick | +0.151 |
show | +0.140 |
test | +0.123 |
magic | +0.111 |
grown | +0.110 |
Neg logit contributions | |
mie | -0.399 |
rified | -0.379 |
elope | -0.378 |
etsy | -0.378 |
hew | -0.373 |
Token | the |
Token ID | 262 |
Feature activation | +1.55e-01 |
Pos logit contributions | |
trick | +0.117 |
show | +0.111 |
test | +0.085 |
grown | +0.084 |
new | +0.081 |
Neg logit contributions | |
mie | -0.624 |
elope | -0.622 |
rified | -0.607 |
jo | -0.602 |
etsy | -0.597 |
Token | magic |
Token ID | 5536 |
Feature activation | +6.96e-02 |
Pos logit contributions | |
trick | +0.210 |
show | +0.206 |
test | +0.191 |
making | +0.189 |
for | +0.186 |
Neg logit contributions | |
mie | -0.277 |
rified | -0.269 |
elope | -0.261 |
hew | -0.261 |
pillar | -0.252 |
Token | island |
Token ID | 7022 |
Feature activation | -2.25e-01 |
Pos logit contributions | |
trick | +0.094 |
show | +0.091 |
test | +0.087 |
my | +0.086 |
five | +0.086 |
Neg logit contributions | |
mie | -0.118 |
elope | -0.115 |
pillar | -0.114 |
rified | -0.114 |
jo | -0.113 |
Token | , |
Token ID | 11 |
Feature activation | +1.26e-01 |
Pos logit contributions | |
mie | +0.552 |
elope | +0.545 |
rified | +0.536 |
jo | +0.528 |
hew | +0.524 |
Neg logit contributions | |
trick | -0.170 |
show | -0.163 |
test | -0.144 |
grown | -0.126 |
magic | -0.124 |
Token | Marina |
Token ID | 35321 |
Feature activation | -2.80e-01 |
Pos logit contributions | |
show | +0.247 |
new | +0.240 |
trick | +0.235 |
for | +0.230 |
grown | +0.229 |
Neg logit contributions | |
mie | -0.166 |
etsy | -0.157 |
rified | -0.157 |
jo | -0.156 |
Mic | -0.156 |
Token | wanted |
Token ID | 2227 |
Feature activation | +9.81e-02 |
Pos logit contributions | |
mie | +0.489 |
rified | +0.480 |
etsy | +0.479 |
elope | +0.474 |
jo | +0.467 |
Neg logit contributions | |
show | -0.423 |
trick | -0.404 |
new | -0.377 |
grown | -0.374 |
test | -0.371 |
Token | to |
Token ID | 284 |
Feature activation | +2.54e-01 |
Pos logit contributions | |
show | +0.054 |
trick | +0.052 |
test | +0.048 |
job | +0.047 |
design | +0.039 |
Neg logit contributions | |
mie | -0.246 |
la | -0.241 |
elope | -0.236 |
etsy | -0.234 |
jo | -0.232 |
Token | explore |
Token ID | 7301 |
Feature activation | +7.46e-02 |
Pos logit contributions | |
math | +0.282 |
new | +0.274 |
for | +0.264 |
five | +0.254 |
red | +0.253 |
Neg logit contributions | |
mie | -0.467 |
elope | -0.456 |
Hop | -0.443 |
bo | -0.423 |
Flash | -0.417 |
Token | , |
Token ID | 11 |
Feature activation | +1.27e-01 |
Pos logit contributions | |
trick | +0.069 |
job | +0.067 |
test | +0.066 |
show | +0.060 |
design | +0.056 |
Neg logit contributions | |
mie | -0.166 |
elope | -0.155 |
rified | -0.154 |
Hop | -0.151 |
la | -0.150 |
Token | but |
Token ID | 475 |
Feature activation | -7.18e-03 |
Pos logit contributions | |
trick | +0.151 |
show | +0.143 |
test | +0.136 |
job | +0.129 |
new | +0.127 |
Neg logit contributions | |
mie | -0.257 |
elope | -0.248 |
rified | -0.247 |
jo | -0.241 |
etsy | -0.240 |
Token | she |
Token ID | 673 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
trick | +0.273 |
show | +0.263 |
test | +0.258 |
job | +0.246 |
work | +0.244 |
Neg logit contributions | |
rified | -0.181 |
mie | -0.178 |
elope | -0.171 |
etsy | -0.170 |
pillar | -0.163 |
Token | decided |
Token ID | 3066 |
Feature activation | +3.91e-02 |
Pos logit contributions | |
rified | +0.398 |
mie | +0.398 |
elope | +0.391 |
etsy | +0.387 |
jo | +0.386 |
Neg logit contributions | |
show | -0.292 |
trick | -0.281 |
new | -0.262 |
grown | -0.259 |
to | -0.254 |
Token | to |
Token ID | 284 |
Feature activation | +2.68e-01 |
Pos logit contributions | |
trick | +0.019 |
show | +0.019 |
test | +0.015 |
job | +0.014 |
design | +0.013 |
Neg logit contributions | |
mie | -0.106 |
etsy | -0.101 |
elope | -0.098 |
jo | -0.098 |
rified | -0.097 |
Token | share |
Token ID | 2648 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
new | +0.219 |
trick | +0.214 |
for | +0.196 |
show | +0.195 |
magic | +0.188 |
Neg logit contributions | |
mie | -0.616 |
elope | -0.597 |
etsy | -0.595 |
jo | -0.581 |
rified | -0.575 |
Token | her |
Token ID | 607 |
Feature activation | +7.26e-02 |
Pos logit contributions | |
show | +0.127 |
trick | +0.126 |
new | +0.113 |
test | +0.113 |
grown | +0.111 |
Neg logit contributions | |
mie | -0.265 |
elope | -0.256 |
rified | -0.254 |
jo | -0.252 |
etsy | -0.252 |
Token | bear |
Token ID | 6842 |
Feature activation | -2.47e-01 |
Pos logit contributions | |
trick | +0.077 |
making | +0.073 |
grown | +0.071 |
show | +0.070 |
test | +0.070 |
Neg logit contributions | |
mie | -0.142 |
elope | -0.134 |
jo | -0.131 |
etsy | -0.128 |
Brown | -0.124 |
Token | with |
Token ID | 351 |
Feature activation | +2.08e-01 |
Pos logit contributions | |
mie | +0.587 |
elope | +0.572 |
jo | +0.562 |
etsy | +0.560 |
rified | +0.559 |
Neg logit contributions | |
show | -0.203 |
trick | -0.202 |
new | -0.181 |
grown | -0.179 |
test | -0.177 |
Token | her |
Token ID | 607 |
Feature activation | +7.35e-02 |
Pos logit contributions | |
show | +0.171 |
trick | +0.159 |
new | +0.152 |
for | +0.135 |
the | +0.133 |
Neg logit contributions | |
mie | -0.517 |
jo | -0.510 |
elope | -0.508 |
rified | -0.507 |
hew | -0.505 |
Token | little |
Token ID | 1310 |
Feature activation | +3.02e-02 |
Pos logit contributions | |
trick | +0.085 |
show | +0.082 |
test | +0.077 |
grown | +0.077 |
five | +0.073 |
Neg logit contributions | |
mie | -0.150 |
elope | -0.144 |
rified | -0.142 |
jo | -0.142 |
etsy | -0.140 |
Token | brother |
Token ID | 3956 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
show | +0.040 |
trick | +0.039 |
new | +0.037 |
test | +0.036 |
for | +0.035 |
Neg logit contributions | |
mie | -0.058 |
rified | -0.057 |
elope | -0.056 |
etsy | -0.056 |
jo | -0.056 |
Token | . |
Token ID | 13 |
Feature activation | +9.55e-03 |
Pos logit contributions | |
etsy | +0.602 |
mie | +0.601 |
rified | +0.599 |
elope | +0.599 |
hew | +0.595 |
Neg logit contributions | |
show | -0.194 |
for | -0.179 |
trick | -0.175 |
new | -0.167 |
to | -0.165 |
Token | it |
Token ID | 340 |
Feature activation | -9.40e-02 |
Pos logit contributions | |
trick | +0.067 |
show | +0.062 |
job | +0.057 |
test | +0.057 |
power | +0.055 |
Neg logit contributions | |
mie | -0.098 |
jo | -0.098 |
elope | -0.094 |
hew | -0.092 |
abel | -0.092 |
Token | must |
Token ID | 1276 |
Feature activation | +8.53e-02 |
Pos logit contributions | |
mie | +0.172 |
etsy | +0.169 |
rified | +0.167 |
elope | +0.165 |
hew | +0.164 |
Neg logit contributions | |
show | -0.138 |
trick | -0.135 |
new | -0.130 |
magic | -0.122 |
test | -0.121 |
Token | be |
Token ID | 307 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
trick | +0.080 |
five | +0.075 |
new | +0.075 |
show | +0.074 |
red | +0.072 |
Neg logit contributions | |
elope | -0.176 |
mie | -0.171 |
hew | -0.169 |
jo | -0.165 |
etsy | -0.160 |
Token | her |
Token ID | 607 |
Feature activation | +7.87e-02 |
Pos logit contributions | |
trick | +0.139 |
show | +0.139 |
test | +0.135 |
balance | +0.113 |
job | +0.109 |
Neg logit contributions | |
mie | -0.309 |
elope | -0.302 |
storms | -0.285 |
ppy | -0.280 |
etsy | -0.278 |
Token | secret |
Token ID | 3200 |
Feature activation | +7.48e-02 |
Pos logit contributions | |
trick | +0.074 |
show | +0.070 |
grown | +0.064 |
my | +0.063 |
making | +0.063 |
Neg logit contributions | |
elope | -0.173 |
mie | -0.165 |
rified | -0.161 |
jo | -0.160 |
etsy | -0.157 |
Token | room |
Token ID | 2119 |
Feature activation | -8.96e-02 |
Pos logit contributions | |
trick | +0.072 |
show | +0.071 |
five | +0.068 |
test | +0.067 |
my | +0.067 |
Neg logit contributions | |
elope | -0.156 |
mie | -0.150 |
jo | -0.149 |
storms | -0.143 |
etsy | -0.139 |
Token | . |
Token ID | 13 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
mie | +0.190 |
elope | +0.180 |
jo | +0.179 |
etsy | +0.176 |
storms | +0.176 |
Neg logit contributions | |
trick | -0.094 |
show | -0.090 |
test | -0.083 |
grown | -0.079 |
five | -0.079 |
Token | She |
Token ID | 1375 |
Feature activation | -2.32e-01 |
Pos logit contributions | |
show | +0.020 |
to | +0.020 |
new | +0.020 |
grown | +0.019 |
for | +0.018 |
Neg logit contributions | |
mie | -0.018 |
ppa | -0.018 |
ppy | -0.018 |
pper | -0.018 |
rified | -0.018 |
Token | was |
Token ID | 373 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
mie | +0.362 |
elope | +0.345 |
jo | +0.339 |
hew | +0.331 |
rified | +0.331 |
Neg logit contributions | |
trick | -0.389 |
show | -0.383 |
new | -0.369 |
test | -0.364 |
job | -0.352 |
Token | ready |
Token ID | 3492 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
trick | +0.255 |
test | +0.224 |
show | +0.219 |
skills | +0.215 |
work | +0.211 |
Neg logit contributions | |
mie | -0.251 |
elope | -0.240 |
la | -0.220 |
rified | -0.216 |
abel | -0.209 |
Token | to |
Token ID | 284 |
Feature activation | +2.75e-01 |
Pos logit contributions | |
trick | +0.111 |
show | +0.098 |
test | +0.096 |
match | +0.088 |
style | +0.081 |
Neg logit contributions | |
mie | -0.351 |
jo | -0.343 |
la | -0.339 |
hew | -0.339 |
elope | -0.337 |
Token | You |
Token ID | 1639 |
Feature activation | -1.84e-01 |
Pos logit contributions | |
apa | +0.172 |
uddy | +0.170 |
elope | +0.168 |
mie | +0.164 |
rified | +0.163 |
Neg logit contributions | |
color | -0.185 |
five | -0.185 |
show | -0.183 |
trick | -0.180 |
my | -0.178 |
Token | 're |
Token ID | 821 |
Feature activation | +4.42e-02 |
Pos logit contributions | |
mie | +0.320 |
rified | +0.304 |
hew | +0.300 |
elope | +0.299 |
etsy | +0.297 |
Neg logit contributions | |
trick | -0.277 |
show | -0.271 |
new | -0.262 |
test | -0.247 |
for | -0.245 |
Token | such |
Token ID | 884 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
trick | +0.045 |
show | +0.037 |
test | +0.036 |
balance | +0.035 |
new | +0.035 |
Neg logit contributions | |
mie | -0.096 |
elope | -0.090 |
thel | -0.088 |
rified | -0.086 |
hew | -0.085 |
Token | a |
Token ID | 257 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
trick | +0.130 |
show | +0.126 |
test | +0.105 |
new | +0.099 |
balance | +0.090 |
Neg logit contributions | |
mie | -0.481 |
elope | -0.470 |
rified | -0.467 |
etsy | -0.461 |
thel | -0.461 |
Token | good |
Token ID | 922 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
trick | +0.115 |
show | +0.108 |
making | +0.104 |
for | +0.103 |
to | +0.096 |
Neg logit contributions | |
mie | -0.287 |
elope | -0.276 |
thel | -0.270 |
etsy | -0.267 |
pillar | -0.264 |
Token | artist |
Token ID | 6802 |
Feature activation | -1.64e-01 |
Pos logit contributions | |
trick | +0.158 |
show | +0.155 |
new | +0.148 |
my | +0.145 |
making | +0.143 |
Neg logit contributions | |
mie | -0.236 |
elope | -0.222 |
etsy | -0.221 |
jo | -0.218 |
thel | -0.216 |
Token | , |
Token ID | 11 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
mie | +0.385 |
elope | +0.373 |
jo | +0.372 |
rified | +0.370 |
etsy | +0.367 |
Neg logit contributions | |
trick | -0.144 |
show | -0.144 |
new | -0.131 |
test | -0.125 |
grown | -0.124 |
Token | Lily |
Token ID | 20037 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
show | +0.123 |
trick | +0.120 |
new | +0.112 |
test | +0.106 |
grown | +0.106 |
Neg logit contributions | |
mie | -0.248 |
rified | -0.242 |
elope | -0.242 |
jo | -0.240 |
etsy | -0.238 |
Token | . |
Token ID | 13 |
Feature activation | -1.77e-02 |
Pos logit contributions | |
mie | +0.574 |
rified | +0.568 |
etsy | +0.563 |
elope | +0.560 |
jo | +0.554 |
Neg logit contributions | |
show | -0.225 |
trick | -0.217 |
new | -0.199 |
for | -0.192 |
grown | -0.184 |
Token | Maybe |
Token ID | 6674 |
Feature activation | -1.83e-01 |
Pos logit contributions | |
mie | +0.031 |
jo | +0.029 |
rified | +0.029 |
elope | +0.029 |
ppa | +0.028 |
Neg logit contributions | |
show | -0.028 |
my | -0.027 |
your | -0.026 |
trick | -0.025 |
for | -0.025 |
Token | one |
Token ID | 530 |
Feature activation | -1.66e-02 |
Pos logit contributions | |
mie | +0.250 |
hew | +0.221 |
elope | +0.216 |
la | +0.214 |
och | +0.211 |
Neg logit contributions | |
trick | -0.312 |
show | -0.283 |
test | -0.274 |
job | -0.272 |
skills | -0.272 |
Token | bright |
Token ID | 6016 |
Feature activation | +6.79e-02 |
Pos logit contributions | |
show | +0.110 |
trick | +0.108 |
new | +0.096 |
test | +0.091 |
grown | +0.090 |
Neg logit contributions | |
mie | -0.321 |
rified | -0.312 |
elope | -0.312 |
etsy | -0.310 |
jo | -0.309 |
Token | flash |
Token ID | 7644 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
math | +0.071 |
trick | +0.070 |
new | +0.070 |
show | +0.069 |
matching | +0.069 |
Neg logit contributions | |
jo | -0.126 |
elope | -0.124 |
rified | -0.122 |
bo | -0.119 |
pillar | -0.116 |
Token | . |
Token ID | 13 |
Feature activation | -1.23e-02 |
Pos logit contributions | |
show | +0.017 |
trick | +0.017 |
test | +0.017 |
new | +0.016 |
grown | +0.016 |
Neg logit contributions | |
jo | -0.026 |
rified | -0.026 |
storms | -0.025 |
bo | -0.025 |
elope | -0.025 |
Token | It |
Token ID | 632 |
Feature activation | -2.01e-01 |
Pos logit contributions | |
mie | +0.022 |
jo | +0.022 |
elope | +0.021 |
rified | +0.021 |
thel | +0.021 |
Neg logit contributions | |
show | -0.018 |
new | -0.018 |
trick | -0.018 |
grown | -0.017 |
the | -0.017 |
Token | hurt |
Token ID | 5938 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
elope | +0.338 |
rified | +0.337 |
mie | +0.336 |
jo | +0.328 |
bo | +0.323 |
Neg logit contributions | |
show | -0.275 |
trick | -0.273 |
new | -0.271 |
test | -0.267 |
job | -0.262 |
Token | Lily |
Token ID | 20037 |
Feature activation | -2.39e-01 |
Pos logit contributions | |
new | +0.074 |
the | +0.070 |
trick | +0.067 |
show | +0.066 |
grown | +0.065 |
Neg logit contributions | |
mie | -0.375 |
jo | -0.368 |
elope | -0.366 |
etsy | -0.366 |
thel | -0.364 |
Token | and |
Token ID | 290 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
mie | +0.586 |
rified | +0.576 |
etsy | +0.574 |
elope | +0.571 |
jo | +0.564 |
Neg logit contributions | |
show | -0.193 |
trick | -0.189 |
new | -0.165 |
for | -0.161 |
to | -0.159 |
Token | Ben |
Token ID | 3932 |
Feature activation | -3.07e-01 |
Pos logit contributions | |
show | +0.192 |
trick | +0.185 |
new | +0.180 |
grown | +0.176 |
for | +0.175 |
Neg logit contributions | |
mie | -0.195 |
etsy | -0.194 |
jo | -0.193 |
elope | -0.191 |
hew | -0.185 |
Token | very |
Token ID | 845 |
Feature activation | +1.47e-01 |
Pos logit contributions | |
mie | +0.679 |
etsy | +0.664 |
rified | +0.658 |
elope | +0.656 |
jo | +0.653 |
Neg logit contributions | |
show | -0.327 |
trick | -0.316 |
for | -0.288 |
new | -0.288 |
to | -0.284 |
Token | much |
Token ID | 881 |
Feature activation | +4.28e-02 |
Pos logit contributions | |
show | +0.123 |
new | +0.118 |
big | +0.115 |
good | +0.115 |
trick | +0.115 |
Neg logit contributions | |
etsy | -0.375 |
mie | -0.372 |
Mic | -0.362 |
rified | -0.361 |
lor | -0.359 |
Token | . |
Token ID | 13 |
Feature activation | -8.73e-03 |
Pos logit contributions | |
show | +0.045 |
trick | +0.044 |
new | +0.040 |
test | +0.039 |
grown | +0.038 |
Neg logit contributions | |
mie | -0.093 |
rified | -0.091 |
elope | -0.090 |
jo | -0.089 |
etsy | -0.089 |
Token | little |
Token ID | 1310 |
Feature activation | -7.33e-02 |
Pos logit contributions | |
show | +0.045 |
trick | +0.043 |
new | +0.040 |
test | +0.038 |
grown | +0.038 |
Neg logit contributions | |
mie | -0.103 |
rified | -0.101 |
jo | -0.100 |
elope | -0.100 |
etsy | -0.100 |
Token | girl |
Token ID | 2576 |
Feature activation | -3.32e-01 |
Pos logit contributions | |
mie | +0.111 |
jo | +0.107 |
rified | +0.107 |
hew | +0.107 |
etsy | +0.107 |
Neg logit contributions | |
show | -0.129 |
trick | -0.125 |
new | -0.122 |
test | -0.118 |
magic | -0.116 |
Token | named |
Token ID | 3706 |
Feature activation | -9.64e-02 |
Pos logit contributions | |
mie | +0.621 |
rified | +0.616 |
etsy | +0.602 |
hew | +0.598 |
elope | +0.595 |
Neg logit contributions | |
show | -0.460 |
trick | -0.440 |
new | -0.418 |
for | -0.405 |
to | -0.402 |
Token | Lily |
Token ID | 20037 |
Feature activation | -3.04e-01 |
Pos logit contributions | |
lor | +0.206 |
jo | +0.206 |
storms | +0.203 |
thel | +0.203 |
rified | +0.202 |
Neg logit contributions | |
the | -0.115 |
show | -0.111 |
to | -0.106 |
my | -0.106 |
for | -0.106 |
Token | . |
Token ID | 13 |
Feature activation | -8.74e-02 |
Pos logit contributions | |
rified | +0.858 |
etsy | +0.827 |
uddy | +0.818 |
storms | +0.813 |
Mic | +0.809 |
Neg logit contributions | |
to | -0.180 |
for | -0.163 |
show | -0.157 |
the | -0.156 |
new | -0.136 |
Token | Lily |
Token ID | 20037 |
Feature activation | -2.97e-01 |
Pos logit contributions | |
thel | +0.160 |
mie | +0.156 |
ppa | +0.156 |
rified | +0.155 |
lor | +0.153 |
Neg logit contributions | |
new | -0.127 |
my | -0.125 |
blue | -0.124 |
red | -0.123 |
the | -0.121 |
Token | loved |
Token ID | 6151 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
rified | +0.645 |
och | +0.620 |
etsy | +0.613 |
lor | +0.605 |
uddy | +0.601 |
Neg logit contributions | |
show | -0.345 |
to | -0.310 |
grown | -0.307 |
new | -0.307 |
trick | -0.305 |
Token | to |
Token ID | 284 |
Feature activation | +1.85e-01 |
Pos logit contributions | |
trick | +0.082 |
show | +0.076 |
match | +0.070 |
five | +0.066 |
test | +0.064 |
Neg logit contributions | |
storms | -0.275 |
mie | -0.267 |
jo | -0.260 |
Mic | -0.259 |
pillar | -0.252 |
Token | explore |
Token ID | 7301 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
for | +0.211 |
new | +0.209 |
my | +0.208 |
making | +0.198 |
example | +0.196 |
Neg logit contributions | |
jo | -0.324 |
elope | -0.315 |
ongo | -0.302 |
mie | -0.302 |
bo | -0.295 |
Token | . |
Token ID | 13 |
Feature activation | -7.82e-02 |
Pos logit contributions | |
trick | +0.011 |
job | +0.010 |
show | +0.009 |
test | +0.009 |
balance | +0.009 |
Neg logit contributions | |
mie | -0.025 |
jo | -0.024 |
hel | -0.024 |
elope | -0.024 |
storms | -0.023 |
Token | One |
Token ID | 1881 |
Feature activation | -9.20e-02 |
Pos logit contributions | |
ppa | +0.194 |
rified | +0.192 |
thel | +0.190 |
mie | +0.188 |
dd | +0.186 |
Neg logit contributions | |
new | -0.070 |
show | -0.066 |
to | -0.064 |
for | -0.063 |
the | -0.063 |
Token | down |
Token ID | 866 |
Feature activation | +4.21e-02 |
Pos logit contributions | |
show | +0.030 |
new | +0.029 |
trick | +0.028 |
for | +0.028 |
to | +0.027 |
Neg logit contributions | |
mie | -0.110 |
jo | -0.108 |
elope | -0.107 |
etsy | -0.107 |
rified | -0.107 |
Token | on |
Token ID | 319 |
Feature activation | +1.88e-01 |
Pos logit contributions | |
trick | +0.059 |
test | +0.054 |
show | +0.053 |
design | +0.048 |
job | +0.047 |
Neg logit contributions | |
rified | -0.074 |
mie | -0.074 |
storms | -0.070 |
hel | -0.070 |
elope | -0.069 |
Token | . |
Token ID | 13 |
Feature activation | -4.08e-02 |
Pos logit contributions | |
trick | +0.136 |
show | +0.135 |
test | +0.115 |
grown | +0.109 |
new | +0.109 |
Neg logit contributions | |
mie | -0.463 |
elope | -0.457 |
rified | -0.453 |
jo | -0.449 |
etsy | -0.441 |
Token |
|
Token ID | 198 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
mie | +0.088 |
ppa | +0.087 |
thel | +0.087 |
rified | +0.085 |
ppy | +0.084 |
Neg logit contributions | |
to | -0.050 |
new | -0.047 |
show | -0.047 |
the | -0.047 |
his | -0.045 |
Token |
|
Token ID | 198 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
rified | +0.301 |
storms | +0.286 |
mie | +0.284 |
ksh | +0.275 |
thel | +0.270 |
Neg logit contributions | |
new | -0.137 |
to | -0.135 |
show | -0.130 |
trick | -0.129 |
big | -0.126 |
Token | Sam |
Token ID | 16305 |
Feature activation | -3.12e-01 |
Pos logit contributions | |
mie | +0.231 |
rified | +0.230 |
storms | +0.220 |
elope | +0.216 |
jo | +0.211 |
Neg logit contributions | |
trick | -0.185 |
show | -0.185 |
new | -0.177 |
grown | -0.176 |
to | -0.172 |
Token | my |
Token ID | 1820 |
Feature activation | -2.17e-01 |
Pos logit contributions | |
rified | +0.389 |
och | +0.375 |
etsy | +0.373 |
storms | +0.365 |
lves | +0.345 |
Neg logit contributions | |
show | -0.616 |
trick | -0.595 |
new | -0.591 |
to | -0.585 |
for | -0.584 |
Token | and |
Token ID | 290 |
Feature activation | +9.56e-02 |
Pos logit contributions | |
rified | +0.386 |
mie | +0.383 |
etsy | +0.379 |
elope | +0.374 |
och | +0.373 |
Neg logit contributions | |
show | -0.317 |
trick | -0.304 |
new | -0.290 |
grown | -0.286 |
to | -0.280 |
Token | his |
Token ID | 465 |
Feature activation | +4.53e-02 |
Pos logit contributions | |
show | +0.054 |
his | +0.054 |
the | +0.052 |
trick | +0.043 |
to | +0.043 |
Neg logit contributions | |
rified | -0.264 |
hew | -0.262 |
och | -0.261 |
pillar | -0.261 |
lor | -0.261 |
Token | mom |
Token ID | 1995 |
Feature activation | -2.99e-01 |
Pos logit contributions | |
show | +0.057 |
trick | +0.055 |
new | +0.054 |
test | +0.050 |
grown | +0.049 |
Neg logit contributions | |
mie | -0.094 |
rified | -0.092 |
etsy | -0.090 |
elope | -0.090 |
jo | -0.090 |
Token | decided |
Token ID | 3066 |
Feature activation | -3.10e-03 |
Pos logit contributions | |
rified | +0.382 |
storms | +0.376 |
etsy | +0.370 |
elope | +0.369 |
lor | +0.369 |
Neg logit contributions | |
show | -0.600 |
for | -0.577 |
to | -0.564 |
the | -0.553 |
work | -0.553 |
Token | I |
Token ID | 40 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
mie | +0.139 |
rified | +0.137 |
Mic | +0.135 |
hew | +0.132 |
ppy | +0.131 |
Neg logit contributions | |
to | -0.170 |
trick | -0.165 |
show | -0.165 |
five | -0.163 |
grown | -0.161 |
Token | know |
Token ID | 760 |
Feature activation | -1.91e-03 |
Pos logit contributions | |
mie | +0.287 |
elope | +0.276 |
rified | +0.275 |
etsy | +0.274 |
jo | +0.273 |
Neg logit contributions | |
show | -0.242 |
trick | -0.240 |
new | -0.227 |
test | -0.220 |
grown | -0.218 |
Token | it |
Token ID | 340 |
Feature activation | -1.66e-01 |
Pos logit contributions | |
mie | +0.004 |
elope | +0.004 |
rified | +0.004 |
jo | +0.004 |
etsy | +0.004 |
Neg logit contributions | |
show | -0.002 |
trick | -0.002 |
new | -0.002 |
test | -0.002 |
grown | -0.002 |
Token | â |
Token ID | 22940 |
Feature activation | -6.93e-02 |
Pos logit contributions | |
mie | +0.287 |
elope | +0.277 |
rified | +0.276 |
jo | +0.273 |
etsy | +0.273 |
Neg logit contributions | |
show | -0.250 |
trick | -0.247 |
new | -0.233 |
test | -0.227 |
grown | -0.225 |
Token | € |
Token ID | 26391 |
Feature activation | -1.94e-01 |
Pos logit contributions | |
mie | +0.088 |
rified | +0.084 |
elope | +0.084 |
jo | +0.082 |
etsy | +0.082 |
Neg logit contributions | |
show | -0.137 |
trick | -0.136 |
test | -0.128 |
new | -0.127 |
grown | -0.126 |
Token | ™ |
Token ID | 8151 |
Feature activation | -3.54e-01 |
Pos logit contributions | |
rified | +0.239 |
ley | +0.219 |
Mic | +0.214 |
mie | +0.210 |
hetically | +0.208 |
Neg logit contributions | |
to | -0.399 |
show | -0.394 |
trick | -0.375 |
test | -0.371 |
grown | -0.369 |
Token | s |
Token ID | 82 |
Feature activation | -3.12e-03 |
Pos logit contributions | |
elope | +0.539 |
mie | +0.516 |
Mic | +0.503 |
lves | +0.499 |
rified | +0.498 |
Neg logit contributions | |
to | -0.593 |
show | -0.582 |
for | -0.576 |
trick | -0.563 |
new | -0.560 |
Token | around |
Token ID | 1088 |
Feature activation | +5.22e-02 |
Pos logit contributions | |
mie | +0.006 |
elope | +0.006 |
hew | +0.006 |
jo | +0.006 |
rified | +0.006 |
Neg logit contributions | |
trick | -0.004 |
show | -0.004 |
test | -0.003 |
new | -0.003 |
grown | -0.003 |
Token | here |
Token ID | 994 |
Feature activation | +9.06e-04 |
Pos logit contributions | |
trick | +0.054 |
show | +0.051 |
new | +0.047 |
test | +0.046 |
grown | +0.045 |
Neg logit contributions | |
mie | -0.112 |
rified | -0.109 |
elope | -0.107 |
jo | -0.107 |
hew | -0.106 |
Token | somewhere |
Token ID | 7382 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
trick | +0.001 |
show | +0.001 |
new | +0.001 |
test | +0.001 |
grown | +0.001 |
Neg logit contributions | |
mie | -0.002 |
elope | -0.002 |
rified | -0.002 |
jo | -0.002 |
hew | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | -4.87e-02 |
Pos logit contributions | |
mie | +0.025 |
elope | +0.024 |
hew | +0.024 |
jo | +0.024 |
etsy | +0.023 |
Neg logit contributions | |
trick | -0.019 |
show | -0.019 |
test | -0.017 |
five | -0.017 |
job | -0.017 |
Token | " |
Token ID | 366 |
Feature activation | -8.17e-02 |
Pos logit contributions | |
trick | +0.194 |
show | +0.185 |
test | +0.178 |
job | +0.173 |
new | +0.172 |
Neg logit contributions | |
mie | -0.265 |
rified | -0.255 |
etsy | -0.250 |
elope | -0.249 |
storms | -0.246 |
Token | OK |
Token ID | 11380 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
jo | +0.134 |
uddy | +0.125 |
mie | +0.124 |
elope | +0.123 |
hew | +0.120 |
Neg logit contributions | |
show | -0.138 |
trick | -0.136 |
new | -0.133 |
good | -0.128 |
my | -0.127 |
Token | , |
Token ID | 11 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
mie | +0.256 |
elope | +0.245 |
rified | +0.244 |
hew | +0.244 |
etsy | +0.243 |
Neg logit contributions | |
trick | -0.070 |
show | -0.067 |
test | -0.057 |
new | -0.057 |
grown | -0.056 |
Token | Dr |
Token ID | 1583 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
show | +0.165 |
trick | +0.156 |
new | +0.147 |
test | +0.141 |
for | +0.140 |
Neg logit contributions | |
mie | -0.306 |
jo | -0.303 |
elope | -0.302 |
rified | -0.301 |
etsy | -0.299 |
Token | . |
Token ID | 13 |
Feature activation | +1.29e-02 |
Pos logit contributions | |
rified | +0.313 |
mie | +0.313 |
etsy | +0.306 |
lves | +0.304 |
jo | +0.303 |
Neg logit contributions | |
trick | -0.084 |
show | -0.082 |
for | -0.079 |
five | -0.073 |
to | -0.070 |
Token | Lee |
Token ID | 5741 |
Feature activation | -3.31e-01 |
Pos logit contributions | |
show | +0.023 |
my | +0.023 |
trick | +0.022 |
to | +0.022 |
for | +0.022 |
Neg logit contributions | |
lves | -0.019 |
etsy | -0.019 |
mie | -0.018 |
Hop | -0.018 |
rified | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | +1.39e-02 |
Pos logit contributions | |
mie | +0.740 |
elope | +0.718 |
rified | +0.715 |
jo | +0.711 |
etsy | +0.709 |
Neg logit contributions | |
show | -0.331 |
trick | -0.327 |
new | -0.297 |
test | -0.286 |
grown | -0.283 |
Token | Thank |
Token ID | 6952 |
Feature activation | +2.31e-03 |
Pos logit contributions | |
my | +0.017 |
show | +0.015 |
your | +0.015 |
to | +0.014 |
for | +0.014 |
Neg logit contributions | |
lves | -0.031 |
ppa | -0.030 |
mie | -0.030 |
onda | -0.029 |
ollo | -0.029 |
Token | you |
Token ID | 345 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
show | +0.003 |
trick | +0.003 |
my | +0.003 |
your | +0.003 |
help | +0.003 |
Neg logit contributions | |
jo | -0.005 |
mie | -0.004 |
rified | -0.004 |
elope | -0.004 |
etsy | -0.004 |
Token | for |
Token ID | 329 |
Feature activation | +2.44e-01 |
Pos logit contributions | |
mie | +0.365 |
elope | +0.357 |
rified | +0.356 |
etsy | +0.354 |
jo | +0.354 |
Neg logit contributions | |
show | -0.085 |
trick | -0.084 |
new | -0.069 |
test | -0.067 |
grown | -0.064 |
Token | helping |
Token ID | 5742 |
Feature activation | +1.75e-01 |
Pos logit contributions | |
chance | +0.269 |
new | +0.236 |
trick | +0.232 |
blue | +0.227 |
red | +0.226 |
Neg logit contributions | |
ppa | -0.443 |
elope | -0.426 |
Max | -0.425 |
Mic | -0.422 |
Rex | -0.422 |
Token | the |
Token ID | 262 |
Feature activation | +8.96e-02 |
Pos logit contributions | |
trick | +0.105 |
test | +0.105 |
grown | +0.103 |
power | +0.102 |
show | +0.101 |
Neg logit contributions | |
Max | -0.200 |
elope | -0.200 |
mie | -0.197 |
Hop | -0.192 |
rified | -0.191 |
Token | game |
Token ID | 983 |
Feature activation | -8.24e-02 |
Pos logit contributions | |
show | +0.092 |
trick | +0.090 |
new | +0.082 |
test | +0.080 |
grown | +0.079 |
Neg logit contributions | |
mie | -0.196 |
elope | -0.190 |
rified | -0.190 |
etsy | -0.188 |
jo | -0.188 |
Token | and |
Token ID | 290 |
Feature activation | +6.19e-02 |
Pos logit contributions | |
mie | +0.185 |
elope | +0.183 |
jo | +0.183 |
rified | +0.180 |
storms | +0.178 |
Neg logit contributions | |
show | -0.070 |
trick | -0.070 |
new | -0.063 |
grown | -0.063 |
test | -0.062 |
Token | the |
Token ID | 262 |
Feature activation | +9.16e-02 |
Pos logit contributions | |
show | +0.096 |
trick | +0.092 |
new | +0.087 |
for | +0.086 |
grown | +0.085 |
Neg logit contributions | |
mie | -0.107 |
rified | -0.106 |
etsy | -0.104 |
elope | -0.104 |
jo | -0.104 |
Token | little |
Token ID | 1310 |
Feature activation | -4.62e-02 |
Pos logit contributions | |
show | +0.112 |
trick | +0.110 |
new | +0.104 |
test | +0.099 |
grown | +0.098 |
Neg logit contributions | |
mie | -0.188 |
elope | -0.182 |
rified | -0.182 |
jo | -0.181 |
etsy | -0.181 |
Token | mouse |
Token ID | 10211 |
Feature activation | -3.48e-01 |
Pos logit contributions | |
mie | +0.071 |
etsy | +0.069 |
jo | +0.069 |
rified | +0.069 |
elope | +0.069 |
Neg logit contributions | |
show | -0.081 |
trick | -0.079 |
new | -0.077 |
test | -0.074 |
magic | -0.072 |
Token | was |
Token ID | 373 |
Feature activation | +7.72e-02 |
Pos logit contributions | |
mie | +0.596 |
elope | +0.573 |
rified | +0.563 |
jo | +0.561 |
etsy | +0.556 |
Neg logit contributions | |
trick | -0.518 |
show | -0.517 |
new | -0.490 |
test | -0.483 |
my | -0.474 |
Token | very |
Token ID | 845 |
Feature activation | +9.86e-02 |
Pos logit contributions | |
trick | +0.084 |
show | +0.083 |
test | +0.077 |
new | +0.071 |
grown | +0.070 |
Neg logit contributions | |
mie | -0.167 |
elope | -0.160 |
jo | -0.155 |
etsy | -0.153 |
rified | -0.152 |
Token | fast |
Token ID | 3049 |
Feature activation | +7.31e-03 |
Pos logit contributions | |
trick | +0.103 |
show | +0.100 |
test | +0.091 |
new | +0.086 |
magic | +0.085 |
Neg logit contributions | |
mie | -0.217 |
elope | -0.208 |
etsy | -0.207 |
jo | -0.204 |
hew | -0.202 |
Token | . |
Token ID | 13 |
Feature activation | -6.16e-02 |
Pos logit contributions | |
trick | +0.009 |
show | +0.008 |
test | +0.008 |
new | +0.008 |
grown | +0.008 |
Neg logit contributions | |
mie | -0.015 |
rified | -0.014 |
storms | -0.014 |
jo | -0.014 |
elope | -0.014 |
Token | The |
Token ID | 383 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
mie | +0.119 |
ppa | +0.112 |
thel | +0.111 |
rified | +0.111 |
elope | +0.110 |
Neg logit contributions | |
show | -0.086 |
new | -0.081 |
to | -0.079 |
trick | -0.078 |
the | -0.078 |