Token | still |
Token ID | 991 |
Feature activation | -1.37e-02 |
Pos logit contributions | |
ey | +0.004 |
principal | +0.004 |
hallway | +0.004 |
orchestra | +0.004 |
classroom | +0.004 |
Neg logit contributions | |
together | -0.002 |
icked | -0.002 |
there | -0.002 |
on | -0.002 |
es | -0.002 |
Token | did |
Token ID | 750 |
Feature activation | -4.96e-02 |
Pos logit contributions | |
icked | +0.022 |
together | +0.021 |
there | +0.021 |
on | +0.021 |
es | +0.020 |
Neg logit contributions | |
principal | -0.026 |
hallway | -0.026 |
hotel | -0.025 |
ey | -0.025 |
city | -0.025 |
Token | not |
Token ID | 407 |
Feature activation | +1.23e-02 |
Pos logit contributions | |
icked | +0.071 |
es | +0.056 |
hot | +0.052 |
there | +0.049 |
iled | +0.048 |
Neg logit contributions | |
city | -0.108 |
playground | -0.105 |
hotel | -0.105 |
classroom | -0.103 |
hallway | -0.102 |
Token | understand |
Token ID | 1833 |
Feature activation | -1.38e-02 |
Pos logit contributions | |
hallway | +0.020 |
ey | +0.020 |
principal | +0.019 |
doorway | +0.019 |
hotel | +0.019 |
Neg logit contributions | |
there | -0.024 |
together | -0.023 |
on | -0.023 |
icked | -0.023 |
in | -0.022 |
Token | why |
Token ID | 1521 |
Feature activation | +9.35e-03 |
Pos logit contributions | |
icked | +0.016 |
es | +0.013 |
together | +0.012 |
there | +0.011 |
on | +0.011 |
Neg logit contributions | |
principal | -0.034 |
city | -0.034 |
hotel | -0.034 |
classroom | -0.033 |
priest | -0.033 |
Token | the |
Token ID | 262 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
city | +0.024 |
hotel | +0.023 |
hallway | +0.023 |
principal | +0.023 |
classroom | +0.023 |
Neg logit contributions | |
icked | -0.011 |
es | -0.008 |
together | -0.008 |
on | -0.007 |
there | -0.007 |
Token | earthquake |
Token ID | 16295 |
Feature activation | -5.01e-03 |
Pos logit contributions | |
Ta | +0.281 |
ey | +0.261 |
salon | +0.256 |
principal | +0.254 |
carn | +0.253 |
Neg logit contributions | |
there | -0.383 |
on | -0.361 |
together | -0.356 |
in | -0.355 |
at | -0.343 |
Token | had |
Token ID | 550 |
Feature activation | -9.59e-03 |
Pos logit contributions | |
icked | +0.008 |
together | +0.007 |
es | +0.007 |
there | +0.007 |
on | +0.007 |
Neg logit contributions | |
hallway | -0.010 |
city | -0.010 |
hotel | -0.010 |
principal | -0.010 |
classroom | -0.009 |
Token | happened |
Token ID | 3022 |
Feature activation | -4.53e-02 |
Pos logit contributions | |
icked | +0.012 |
es | +0.012 |
together | +0.011 |
there | +0.011 |
on | +0.011 |
Neg logit contributions | |
city | -0.020 |
hotel | -0.020 |
principal | -0.020 |
hallway | -0.020 |
classroom | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | -1.07e-02 |
Pos logit contributions | |
icked | +0.060 |
es | +0.043 |
bled | +0.034 |
pped | +0.033 |
hot | +0.032 |
Neg logit contributions | |
city | -0.101 |
classroom | -0.101 |
hallway | -0.100 |
neighbourhood | -0.100 |
principal | -0.098 |
Token | It |
Token ID | 632 |
Feature activation | +1.59e-02 |
Pos logit contributions | |
together | +0.012 |
there | +0.011 |
icked | +0.011 |
on | +0.011 |
es | +0.010 |
Neg logit contributions | |
hallway | -0.026 |
ey | -0.026 |
principal | -0.026 |
priest | -0.026 |
orchestra | -0.025 |
Token | helped |
Token ID | 4193 |
Feature activation | -2.57e-02 |
Pos logit contributions | |
hallway | +0.011 |
city | +0.011 |
doorway | +0.011 |
hotel | +0.011 |
classroom | +0.011 |
Neg logit contributions | |
icked | -0.010 |
es | -0.010 |
there | -0.010 |
together | -0.010 |
on | -0.010 |
Token | the |
Token ID | 262 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
icked | +0.035 |
es | +0.029 |
there | +0.028 |
together | +0.027 |
on | +0.026 |
Neg logit contributions | |
priest | -0.055 |
city | -0.054 |
hotel | -0.054 |
hallway | -0.054 |
principal | -0.054 |
Token | old |
Token ID | 1468 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
ey | +0.208 |
principal | +0.206 |
Ta | +0.196 |
orchestra | +0.191 |
actor | +0.187 |
Neg logit contributions | |
there | -0.328 |
on | -0.311 |
together | -0.309 |
as | -0.298 |
in | -0.297 |
Token | lady |
Token ID | 10846 |
Feature activation | -6.25e-03 |
Pos logit contributions | |
ey | +0.108 |
principal | +0.107 |
hallway | +0.104 |
hotel | +0.100 |
orchestra | +0.100 |
Neg logit contributions | |
there | -0.321 |
together | -0.314 |
on | -0.313 |
icked | -0.305 |
es | -0.301 |
Token | fix |
Token ID | 4259 |
Feature activation | -6.08e-03 |
Pos logit contributions | |
icked | +0.010 |
es | +0.009 |
there | +0.009 |
together | +0.009 |
on | +0.009 |
Neg logit contributions | |
hallway | -0.012 |
city | -0.011 |
hotel | -0.011 |
principal | -0.011 |
classroom | -0.011 |
Token | the |
Token ID | 262 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
icked | +0.006 |
together | +0.006 |
there | +0.005 |
es | +0.005 |
on | +0.005 |
Neg logit contributions | |
hallway | -0.015 |
hotel | -0.015 |
city | -0.015 |
principal | -0.015 |
doorway | -0.015 |
Token | fence |
Token ID | 13990 |
Feature activation | -3.74e-02 |
Pos logit contributions | |
principal | +0.226 |
orchestra | +0.213 |
priest | +0.211 |
ey | +0.211 |
Ta | +0.207 |
Neg logit contributions | |
there | -0.417 |
on | -0.405 |
together | -0.399 |
in | -0.381 |
icked | -0.379 |
Token | . |
Token ID | 13 |
Feature activation | +4.21e-03 |
Pos logit contributions | |
icked | +0.043 |
es | +0.035 |
there | +0.031 |
together | +0.029 |
on | +0.029 |
Neg logit contributions | |
hallway | -0.091 |
city | -0.089 |
doorway | -0.088 |
playground | -0.087 |
street | -0.086 |
Token | |
Token ID | 220 |
Feature activation | +9.62e-03 |
Pos logit contributions | |
hallway | +0.007 |
city | +0.007 |
doorway | +0.007 |
hotel | +0.007 |
classroom | +0.007 |
Neg logit contributions | |
icked | -0.007 |
there | -0.007 |
on | -0.007 |
es | -0.007 |
together | -0.007 |
Token |
|
Token ID | 198 |
Feature activation | +2.65e-03 |
Pos logit contributions | |
principal | +0.017 |
hallway | +0.017 |
city | +0.017 |
classroom | +0.017 |
doorway | +0.016 |
Neg logit contributions | |
icked | -0.015 |
on | -0.014 |
there | -0.014 |
es | -0.014 |
together | -0.014 |
Token |
|
Token ID | 198 |
Feature activation | +4.32e-03 |
Pos logit contributions | |
hallway | +0.005 |
city | +0.005 |
principal | +0.005 |
hotel | +0.005 |
classroom | +0.005 |
Neg logit contributions | |
icked | -0.004 |
there | -0.004 |
together | -0.004 |
on | -0.004 |
es | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +4.53e-02 |
Pos logit contributions | |
icked | +0.028 |
together | +0.027 |
es | +0.027 |
there | +0.026 |
on | +0.026 |
Neg logit contributions | |
hallway | -0.040 |
principal | -0.040 |
hotel | -0.039 |
city | -0.039 |
doorway | -0.039 |
Token | bit |
Token ID | 1643 |
Feature activation | -8.29e-03 |
Pos logit contributions | |
ey | +0.094 |
opl | +0.088 |
orchestra | +0.087 |
alphabet | +0.086 |
Ta | +0.085 |
Neg logit contributions | |
there | -0.076 |
together | -0.075 |
on | -0.074 |
it | -0.068 |
, | -0.068 |
Token | of |
Token ID | 286 |
Feature activation | +2.52e-02 |
Pos logit contributions | |
icked | +0.014 |
together | +0.013 |
there | +0.013 |
es | +0.013 |
on | +0.013 |
Neg logit contributions | |
hallway | -0.015 |
principal | -0.015 |
hotel | -0.015 |
city | -0.015 |
ey | -0.015 |
Token | trouble |
Token ID | 5876 |
Feature activation | -3.46e-02 |
Pos logit contributions | |
hallway | +0.043 |
principal | +0.043 |
priest | +0.042 |
salon | +0.041 |
orchestra | +0.041 |
Neg logit contributions | |
together | -0.047 |
there | -0.047 |
on | -0.046 |
it | -0.045 |
as | -0.043 |
Token | fixing |
Token ID | 18682 |
Feature activation | -1.15e-02 |
Pos logit contributions | |
icked | +0.041 |
es | +0.035 |
together | +0.034 |
there | +0.033 |
on | +0.032 |
Neg logit contributions | |
hallway | -0.082 |
hotel | -0.081 |
principal | -0.081 |
city | -0.081 |
classroom | -0.079 |
Token | the |
Token ID | 262 |
Feature activation | +1.76e-01 |
Pos logit contributions | |
icked | +0.014 |
together | +0.013 |
there | +0.013 |
es | +0.013 |
on | +0.013 |
Neg logit contributions | |
hallway | -0.026 |
principal | -0.026 |
hotel | -0.026 |
city | -0.026 |
classroom | -0.025 |
Token | bag |
Token ID | 6131 |
Feature activation | -3.05e-02 |
Pos logit contributions | |
orchestra | +0.286 |
Ta | +0.274 |
priest | +0.272 |
actor | +0.269 |
ey | +0.269 |
Neg logit contributions | |
there | -0.352 |
on | -0.349 |
in | -0.317 |
it | -0.312 |
at | -0.307 |
Token | , |
Token ID | 11 |
Feature activation | -2.47e-02 |
Pos logit contributions | |
icked | +0.032 |
together | +0.027 |
es | +0.027 |
there | +0.026 |
on | +0.025 |
Neg logit contributions | |
hallway | -0.076 |
hotel | -0.074 |
doorway | -0.074 |
principal | -0.074 |
city | -0.074 |
Token | but |
Token ID | 475 |
Feature activation | -6.68e-03 |
Pos logit contributions | |
icked | +0.025 |
together | +0.024 |
there | +0.023 |
on | +0.022 |
es | +0.022 |
Neg logit contributions | |
hallway | -0.061 |
principal | -0.061 |
hotel | -0.060 |
city | -0.060 |
ey | -0.060 |
Token | eventually |
Token ID | 4191 |
Feature activation | -1.63e-02 |
Pos logit contributions | |
icked | +0.006 |
together | +0.005 |
es | +0.005 |
there | +0.005 |
on | +0.005 |
Neg logit contributions | |
hallway | -0.018 |
principal | -0.018 |
hotel | -0.018 |
city | -0.018 |
classroom | -0.017 |
Token | he |
Token ID | 339 |
Feature activation | +8.67e-03 |
Pos logit contributions | |
icked | +0.021 |
es | +0.018 |
together | +0.018 |
iled | +0.016 |
against | +0.016 |
Neg logit contributions | |
hotel | -0.037 |
city | -0.035 |
principal | -0.035 |
museum | -0.034 |
classroom | -0.034 |
Token | she |
Token ID | 673 |
Feature activation | +2.81e-03 |
Pos logit contributions | |
icked | +0.034 |
es | +0.028 |
together | +0.027 |
on | +0.026 |
there | +0.026 |
Neg logit contributions | |
city | -0.053 |
hallway | -0.053 |
hotel | -0.052 |
library | -0.051 |
principal | -0.051 |
Token | always |
Token ID | 1464 |
Feature activation | -6.79e-03 |
Pos logit contributions | |
hallway | +0.005 |
principal | +0.005 |
city | +0.005 |
hotel | +0.005 |
ey | +0.005 |
Neg logit contributions | |
icked | -0.004 |
together | -0.004 |
there | -0.004 |
on | -0.004 |
es | -0.004 |
Token | remembered |
Token ID | 12086 |
Feature activation | -3.48e-02 |
Pos logit contributions | |
icked | +0.010 |
together | +0.010 |
es | +0.010 |
there | +0.010 |
on | +0.009 |
Neg logit contributions | |
hallway | -0.013 |
classroom | -0.013 |
city | -0.013 |
principal | -0.013 |
hotel | -0.012 |
Token | how |
Token ID | 703 |
Feature activation | -6.24e-03 |
Pos logit contributions | |
icked | +0.032 |
es | +0.028 |
together | +0.025 |
on | +0.022 |
there | +0.022 |
Neg logit contributions | |
principal | -0.089 |
city | -0.089 |
hotel | -0.089 |
classroom | -0.088 |
priest | -0.088 |
Token | kind |
Token ID | 1611 |
Feature activation | -4.72e-03 |
Pos logit contributions | |
icked | +0.008 |
es | +0.007 |
together | +0.007 |
on | +0.007 |
against | +0.006 |
Neg logit contributions | |
city | -0.013 |
priest | -0.013 |
classroom | -0.013 |
principal | -0.013 |
hotel | -0.013 |
Token | the |
Token ID | 262 |
Feature activation | +1.75e-01 |
Pos logit contributions | |
icked | +0.006 |
together | +0.005 |
es | +0.005 |
there | +0.004 |
on | +0.004 |
Neg logit contributions | |
city | -0.011 |
priest | -0.011 |
hallway | -0.010 |
hotel | -0.010 |
ey | -0.010 |
Token | bear |
Token ID | 6842 |
Feature activation | +2.34e-03 |
Pos logit contributions | |
salon | +0.329 |
Ta | +0.328 |
principal | +0.321 |
orchestra | +0.318 |
ey | +0.316 |
Neg logit contributions | |
there | -0.322 |
on | -0.303 |
in | -0.289 |
together | -0.279 |
with | -0.269 |
Token | was |
Token ID | 373 |
Feature activation | +1.26e-03 |
Pos logit contributions | |
hallway | +0.005 |
city | +0.004 |
principal | +0.004 |
hotel | +0.004 |
museum | +0.004 |
Neg logit contributions | |
icked | -0.003 |
together | -0.003 |
there | -0.003 |
on | -0.003 |
es | -0.003 |
Token | to |
Token ID | 284 |
Feature activation | +1.27e-02 |
Pos logit contributions | |
principal | +0.003 |
priest | +0.003 |
city | +0.003 |
museum | +0.003 |
classroom | +0.003 |
Neg logit contributions | |
icked | -0.002 |
es | -0.001 |
together | -0.001 |
there | -0.001 |
bed | -0.001 |
Token | share |
Token ID | 2648 |
Feature activation | -3.76e-02 |
Pos logit contributions | |
hallway | +0.022 |
principal | +0.022 |
city | +0.022 |
hotel | +0.022 |
ey | +0.021 |
Neg logit contributions | |
icked | -0.022 |
together | -0.022 |
there | -0.021 |
es | -0.021 |
on | -0.021 |
Token | his |
Token ID | 465 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
icked | +0.036 |
together | +0.030 |
there | +0.029 |
es | +0.029 |
on | +0.027 |
Neg logit contributions | |
principal | -0.095 |
city | -0.095 |
hallway | -0.095 |
hotel | -0.094 |
ey | -0.094 |
Token | they |
Token ID | 484 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
icked | +0.034 |
es | +0.028 |
pped | +0.027 |
there | +0.025 |
bled | +0.021 |
Neg logit contributions | |
city | -0.067 |
doorway | -0.067 |
hotel | -0.067 |
woman | -0.066 |
orchestra | -0.066 |
Token | were |
Token ID | 547 |
Feature activation | -2.94e-03 |
Pos logit contributions | |
city | +0.021 |
hallway | +0.021 |
doorway | +0.021 |
hotel | +0.021 |
orchestra | +0.021 |
Neg logit contributions | |
icked | -0.017 |
es | -0.015 |
together | -0.015 |
there | -0.015 |
on | -0.014 |
Token | able |
Token ID | 1498 |
Feature activation | -3.75e-02 |
Pos logit contributions | |
icked | +0.003 |
es | +0.003 |
against | +0.003 |
there | +0.003 |
bed | +0.003 |
Neg logit contributions | |
priest | -0.007 |
doorway | -0.006 |
orchestra | -0.006 |
city | -0.006 |
hotel | -0.006 |
Token | to |
Token ID | 284 |
Feature activation | +1.24e-02 |
Pos logit contributions | |
icked | +0.041 |
es | +0.036 |
bed | +0.031 |
pped | +0.030 |
against | +0.029 |
Neg logit contributions | |
doorway | -0.086 |
city | -0.086 |
neighbourhood | -0.084 |
hotel | -0.084 |
priest | -0.084 |
Token | replace |
Token ID | 6330 |
Feature activation | -1.86e-02 |
Pos logit contributions | |
hallway | +0.020 |
principal | +0.020 |
hotel | +0.020 |
city | +0.020 |
ey | +0.020 |
Neg logit contributions | |
icked | -0.023 |
together | -0.023 |
there | -0.022 |
on | -0.022 |
es | -0.022 |
Token | the |
Token ID | 262 |
Feature activation | +1.75e-01 |
Pos logit contributions | |
icked | +0.017 |
together | +0.016 |
there | +0.015 |
on | +0.015 |
es | +0.014 |
Neg logit contributions | |
hallway | -0.049 |
principal | -0.048 |
hotel | -0.048 |
city | -0.048 |
ey | -0.047 |
Token | old |
Token ID | 1468 |
Feature activation | +1.62e-01 |
Pos logit contributions | |
priest | +0.282 |
principal | +0.278 |
ey | +0.261 |
courtroom | +0.256 |
Ta | +0.250 |
Neg logit contributions | |
there | -0.380 |
on | -0.369 |
together | -0.367 |
with | -0.345 |
in | -0.343 |
Token | rug |
Token ID | 14477 |
Feature activation | -2.71e-02 |
Pos logit contributions | |
principal | +0.194 |
priest | +0.189 |
ey | +0.188 |
Ta | +0.177 |
courtroom | +0.173 |
Neg logit contributions | |
together | -0.392 |
there | -0.389 |
on | -0.385 |
es | -0.365 |
it | -0.365 |
Token | with |
Token ID | 351 |
Feature activation | +1.45e-03 |
Pos logit contributions | |
icked | +0.022 |
es | +0.015 |
there | +0.014 |
together | +0.012 |
on | +0.012 |
Neg logit contributions | |
hallway | -0.075 |
hotel | -0.074 |
city | -0.074 |
doorway | -0.073 |
classroom | -0.073 |
Token | a |
Token ID | 257 |
Feature activation | +6.57e-02 |
Pos logit contributions | |
hallway | +0.003 |
principal | +0.003 |
hotel | +0.003 |
city | +0.003 |
ey | +0.003 |
Neg logit contributions | |
icked | -0.002 |
together | -0.002 |
there | -0.002 |
on | -0.002 |
es | -0.002 |
Token | new |
Token ID | 649 |
Feature activation | +9.81e-02 |
Pos logit contributions | |
ey | +0.123 |
principal | +0.118 |
Ta | +0.111 |
priest | +0.110 |
orchestra | +0.107 |
Neg logit contributions | |
together | -0.131 |
there | -0.124 |
on | -0.121 |
it | -0.113 |
with | -0.111 |
Token | gave |
Token ID | 2921 |
Feature activation | -2.21e-02 |
Pos logit contributions | |
icked | +0.018 |
together | +0.017 |
there | +0.017 |
es | +0.017 |
on | +0.017 |
Neg logit contributions | |
hallway | -0.022 |
principal | -0.022 |
hotel | -0.022 |
city | -0.021 |
classroom | -0.021 |
Token | each |
Token ID | 1123 |
Feature activation | +6.85e-02 |
Pos logit contributions | |
icked | +0.023 |
together | +0.022 |
there | +0.022 |
es | +0.021 |
on | +0.021 |
Neg logit contributions | |
principal | -0.053 |
hotel | -0.052 |
hallway | -0.052 |
city | -0.052 |
classroom | -0.052 |
Token | other |
Token ID | 584 |
Feature activation | +6.79e-03 |
Pos logit contributions | |
hallway | +0.118 |
ey | +0.114 |
salon | +0.111 |
principal | +0.110 |
former | +0.110 |
Neg logit contributions | |
on | -0.149 |
it | -0.147 |
there | -0.142 |
together | -0.140 |
in | -0.138 |
Token | half |
Token ID | 2063 |
Feature activation | -2.76e-02 |
Pos logit contributions | |
hallway | +0.018 |
principal | +0.018 |
ey | +0.018 |
city | +0.018 |
orchestra | +0.018 |
Neg logit contributions | |
together | -0.006 |
there | -0.006 |
icked | -0.006 |
on | -0.006 |
es | -0.005 |
Token | of |
Token ID | 286 |
Feature activation | +3.88e-02 |
Pos logit contributions | |
icked | +0.027 |
there | +0.025 |
es | +0.025 |
together | +0.025 |
on | +0.024 |
Neg logit contributions | |
hotel | -0.066 |
principal | -0.065 |
priest | -0.065 |
ey | -0.065 |
hallway | -0.064 |
Token | the |
Token ID | 262 |
Feature activation | +1.75e-01 |
Pos logit contributions | |
hallway | +0.109 |
principal | +0.106 |
city | +0.106 |
ey | +0.105 |
doorway | +0.105 |
Neg logit contributions | |
together | -0.031 |
there | -0.030 |
on | -0.028 |
icked | -0.028 |
it | -0.026 |
Token | lemon |
Token ID | 18873 |
Feature activation | +1.78e-02 |
Pos logit contributions | |
orchestra | +0.384 |
hallway | +0.371 |
doorway | +0.363 |
professor | +0.362 |
actor | +0.361 |
Neg logit contributions | |
there | -0.278 |
on | -0.268 |
together | -0.243 |
it | -0.241 |
in | -0.235 |
Token | . |
Token ID | 13 |
Feature activation | -6.62e-03 |
Pos logit contributions | |
hotel | +0.038 |
principal | +0.037 |
hallway | +0.037 |
ey | +0.037 |
babys | +0.037 |
Neg logit contributions | |
icked | -0.022 |
there | -0.021 |
es | -0.021 |
together | -0.021 |
on | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | -1.08e-02 |
Pos logit contributions | |
icked | +0.010 |
es | +0.009 |
on | +0.009 |
there | +0.009 |
together | +0.009 |
Neg logit contributions | |
classroom | -0.013 |
hallway | -0.013 |
city | -0.013 |
principal | -0.013 |
hotel | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
icked | +0.014 |
together | +0.014 |
there | +0.014 |
on | +0.013 |
es | +0.013 |
Neg logit contributions | |
hallway | -0.023 |
principal | -0.023 |
hotel | -0.023 |
city | -0.023 |
classroom | -0.023 |
Token | " |
Token ID | 1 |
Feature activation | -1.17e-03 |
Pos logit contributions | |
icked | +0.016 |
es | +0.015 |
there | +0.015 |
on | +0.014 |
together | +0.014 |
Neg logit contributions | |
hallway | -0.028 |
principal | -0.028 |
classroom | -0.028 |
city | -0.027 |
hotel | -0.027 |
Token | Even |
Token ID | 3412 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
icked | +0.014 |
together | +0.014 |
there | +0.013 |
es | +0.013 |
on | +0.013 |
Neg logit contributions | |
hallway | -0.023 |
principal | -0.023 |
city | -0.023 |
classroom | -0.023 |
ey | -0.022 |
Token | though |
Token ID | 996 |
Feature activation | -1.45e-02 |
Pos logit contributions | |
icked | +0.011 |
es | +0.009 |
together | +0.007 |
there | +0.006 |
on | +0.006 |
Neg logit contributions | |
city | -0.028 |
hallway | -0.028 |
classroom | -0.028 |
hotel | -0.028 |
library | -0.028 |
Token | she |
Token ID | 673 |
Feature activation | +4.44e-02 |
Pos logit contributions | |
icked | +0.020 |
together | +0.016 |
es | +0.015 |
iled | +0.014 |
on | +0.013 |
Neg logit contributions | |
city | -0.034 |
hallway | -0.033 |
classroom | -0.033 |
hotel | -0.032 |
library | -0.032 |
Token | understood |
Token ID | 7247 |
Feature activation | -4.15e-02 |
Pos logit contributions | |
hallway | +0.095 |
principal | +0.095 |
hotel | +0.094 |
city | +0.093 |
ey | +0.093 |
Neg logit contributions | |
icked | -0.060 |
together | -0.059 |
there | -0.058 |
on | -0.057 |
es | -0.056 |
Token | why |
Token ID | 1521 |
Feature activation | +5.47e-03 |
Pos logit contributions | |
icked | +0.054 |
together | +0.040 |
es | +0.040 |
bled | +0.036 |
iled | +0.033 |
Neg logit contributions | |
library | -0.094 |
principal | -0.094 |
city | -0.093 |
museum | -0.093 |
hotel | -0.092 |
Token | the |
Token ID | 262 |
Feature activation | +1.74e-01 |
Pos logit contributions | |
babys | +0.012 |
city | +0.012 |
hallway | +0.012 |
classroom | +0.012 |
principal | +0.012 |
Neg logit contributions | |
icked | -0.008 |
together | -0.006 |
es | -0.006 |
on | -0.005 |
there | -0.005 |
Token | fence |
Token ID | 13990 |
Feature activation | -8.21e-03 |
Pos logit contributions | |
principal | +0.264 |
ey | +0.259 |
priest | +0.248 |
hotel | +0.241 |
hallway | +0.241 |
Neg logit contributions | |
there | -0.371 |
on | -0.361 |
together | -0.357 |
it | -0.337 |
in | -0.333 |
Token | was |
Token ID | 373 |
Feature activation | +2.82e-03 |
Pos logit contributions | |
icked | +0.012 |
together | +0.011 |
there | +0.011 |
es | +0.011 |
on | +0.011 |
Neg logit contributions | |
hallway | -0.017 |
principal | -0.017 |
city | -0.016 |
hotel | -0.016 |
ey | -0.016 |
Token | there |
Token ID | 612 |
Feature activation | -2.59e-02 |
Pos logit contributions | |
hallway | +0.005 |
hotel | +0.005 |
doorway | +0.005 |
babys | +0.005 |
city | +0.005 |
Neg logit contributions | |
icked | -0.005 |
es | -0.005 |
together | -0.004 |
bed | -0.004 |
on | -0.004 |
Token | , |
Token ID | 11 |
Feature activation | -2.87e-02 |
Pos logit contributions | |
icked | +0.033 |
es | +0.026 |
together | +0.024 |
bled | +0.022 |
there | +0.022 |
Neg logit contributions | |
city | -0.061 |
hallway | -0.061 |
hotel | -0.060 |
classroom | -0.060 |
playground | -0.059 |
Token | she |
Token ID | 673 |
Feature activation | -3.70e-04 |
Pos logit contributions | |
icked | +0.039 |
es | +0.031 |
together | +0.028 |
iled | +0.026 |
on | +0.026 |
Neg logit contributions | |
hallway | -0.066 |
city | -0.065 |
classroom | -0.064 |
doorway | -0.064 |
playground | -0.063 |
Token | but |
Token ID | 475 |
Feature activation | -1.01e-02 |
Pos logit contributions | |
icked | +0.025 |
together | +0.023 |
es | +0.023 |
there | +0.022 |
on | +0.022 |
Neg logit contributions | |
hallway | -0.043 |
principal | -0.043 |
city | -0.042 |
hotel | -0.042 |
ey | -0.041 |
Token | nobody |
Token ID | 8168 |
Feature activation | +4.07e-03 |
Pos logit contributions | |
icked | +0.012 |
es | +0.012 |
together | +0.010 |
on | +0.009 |
there | +0.009 |
Neg logit contributions | |
city | -0.023 |
principal | -0.023 |
priest | -0.023 |
hotel | -0.023 |
museum | -0.022 |
Token | wanted |
Token ID | 2227 |
Feature activation | -3.28e-02 |
Pos logit contributions | |
principal | +0.008 |
hotel | +0.008 |
priest | +0.008 |
hallway | +0.008 |
city | +0.007 |
Neg logit contributions | |
icked | -0.006 |
together | -0.006 |
es | -0.006 |
there | -0.006 |
on | -0.006 |
Token | to |
Token ID | 284 |
Feature activation | +6.11e-03 |
Pos logit contributions | |
icked | +0.038 |
es | +0.032 |
together | +0.031 |
bled | +0.029 |
bed | +0.028 |
Neg logit contributions | |
museum | -0.077 |
principal | -0.075 |
city | -0.075 |
hotel | -0.073 |
library | -0.073 |
Token | buy |
Token ID | 2822 |
Feature activation | -3.22e-02 |
Pos logit contributions | |
principal | +0.009 |
hallway | +0.009 |
city | +0.009 |
hotel | +0.009 |
priest | +0.009 |
Neg logit contributions | |
icked | -0.012 |
together | -0.012 |
there | -0.012 |
es | -0.012 |
on | -0.012 |
Token | the |
Token ID | 262 |
Feature activation | +1.74e-01 |
Pos logit contributions | |
icked | +0.029 |
es | +0.026 |
together | +0.025 |
there | +0.023 |
on | +0.023 |
Neg logit contributions | |
city | -0.083 |
principal | -0.083 |
hotel | -0.082 |
museum | -0.082 |
priest | -0.081 |
Token | earth |
Token ID | 4534 |
Feature activation | -2.19e-02 |
Pos logit contributions | |
principal | +0.256 |
salon | +0.240 |
babys | +0.232 |
carn | +0.222 |
hallway | +0.221 |
Neg logit contributions | |
there | -0.408 |
together | -0.389 |
on | -0.388 |
in | -0.379 |
it | -0.378 |
Token | . |
Token ID | 13 |
Feature activation | -9.25e-03 |
Pos logit contributions | |
icked | +0.021 |
es | +0.017 |
together | +0.016 |
there | +0.015 |
on | +0.015 |
Neg logit contributions | |
city | -0.057 |
hotel | -0.057 |
hallway | -0.056 |
principal | -0.056 |
classroom | -0.055 |
Token | Finally |
Token ID | 9461 |
Feature activation | -7.25e-03 |
Pos logit contributions | |
icked | +0.013 |
there | +0.012 |
together | +0.012 |
on | +0.012 |
es | +0.012 |
Neg logit contributions | |
principal | -0.020 |
hallway | -0.019 |
city | -0.019 |
hotel | -0.019 |
classroom | -0.019 |
Token | , |
Token ID | 11 |
Feature activation | -3.30e-02 |
Pos logit contributions | |
icked | +0.012 |
es | +0.011 |
uts | +0.011 |
run | +0.010 |
ppers | +0.010 |
Neg logit contributions | |
hotel | -0.012 |
babys | -0.012 |
city | -0.011 |
museum | -0.011 |
audience | -0.011 |
Token | a |
Token ID | 257 |
Feature activation | +5.54e-02 |
Pos logit contributions | |
icked | +0.045 |
es | +0.040 |
together | +0.033 |
there | +0.031 |
iled | +0.030 |
Neg logit contributions | |
city | -0.074 |
hotel | -0.072 |
principal | -0.072 |
hallway | -0.072 |
museum | -0.072 |
Token | was |
Token ID | 373 |
Feature activation | +3.23e-03 |
Pos logit contributions | |
icked | +0.001 |
together | +0.001 |
there | +0.001 |
on | +0.001 |
es | +0.001 |
Neg logit contributions | |
principal | -0.002 |
hallway | -0.002 |
hotel | -0.002 |
city | -0.002 |
classroom | -0.002 |
Token | glad |
Token ID | 9675 |
Feature activation | -2.31e-02 |
Pos logit contributions | |
principal | +0.005 |
priest | +0.005 |
hotel | +0.005 |
classroom | +0.005 |
babys | +0.005 |
Neg logit contributions | |
es | -0.006 |
icked | -0.006 |
together | -0.005 |
instead | -0.005 |
bed | -0.005 |
Token | to |
Token ID | 284 |
Feature activation | -3.98e-04 |
Pos logit contributions | |
icked | +0.030 |
es | +0.026 |
pped | +0.023 |
bed | +0.021 |
together | +0.021 |
Neg logit contributions | |
hotel | -0.053 |
principal | -0.052 |
priest | -0.052 |
city | -0.051 |
museum | -0.051 |
Token | have |
Token ID | 423 |
Feature activation | -1.45e-02 |
Pos logit contributions | |
icked | +0.001 |
together | +0.001 |
there | +0.001 |
on | +0.001 |
es | +0.001 |
Neg logit contributions | |
hallway | -0.001 |
principal | -0.001 |
hotel | -0.001 |
ey | -0.001 |
city | -0.001 |
Token | fixed |
Token ID | 5969 |
Feature activation | -1.61e-02 |
Pos logit contributions | |
icked | +0.021 |
es | +0.019 |
against | +0.017 |
on | +0.017 |
bed | +0.017 |
Neg logit contributions | |
city | -0.027 |
audience | -0.027 |
classroom | -0.027 |
principal | -0.027 |
playground | -0.027 |
Token | the |
Token ID | 262 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
icked | +0.019 |
together | +0.016 |
there | +0.016 |
es | +0.015 |
on | +0.015 |
Neg logit contributions | |
principal | -0.039 |
city | -0.039 |
hotel | -0.038 |
hallway | -0.038 |
classroom | -0.038 |
Token | stage |
Token ID | 3800 |
Feature activation | -2.99e-02 |
Pos logit contributions | |
ey | +0.211 |
principal | +0.204 |
priest | +0.201 |
Ta | +0.201 |
ostr | +0.199 |
Neg logit contributions | |
there | -0.431 |
on | -0.426 |
in | -0.394 |
together | -0.393 |
it | -0.386 |
Token | and |
Token ID | 290 |
Feature activation | -4.77e-03 |
Pos logit contributions | |
icked | +0.024 |
together | +0.020 |
there | +0.020 |
es | +0.019 |
on | +0.018 |
Neg logit contributions | |
hallway | -0.082 |
principal | -0.081 |
hotel | -0.081 |
city | -0.081 |
classroom | -0.080 |
Token | then |
Token ID | 788 |
Feature activation | -2.88e-02 |
Pos logit contributions | |
together | +0.005 |
icked | +0.005 |
there | +0.005 |
on | +0.005 |
es | +0.005 |
Neg logit contributions | |
hallway | -0.011 |
principal | -0.011 |
hotel | -0.011 |
city | -0.011 |
ey | -0.011 |
Token | he |
Token ID | 339 |
Feature activation | -4.75e-04 |
Pos logit contributions | |
icked | +0.041 |
es | +0.038 |
together | +0.036 |
there | +0.035 |
on | +0.034 |
Neg logit contributions | |
principal | -0.059 |
hallway | -0.059 |
city | -0.059 |
hotel | -0.059 |
classroom | -0.058 |
Token | left |
Token ID | 1364 |
Feature activation | -3.45e-02 |
Pos logit contributions | |
icked | +0.001 |
together | +0.001 |
there | +0.001 |
on | +0.001 |
es | +0.001 |
Neg logit contributions | |
hallway | -0.001 |
principal | -0.001 |
hotel | -0.001 |
city | -0.001 |
ey | -0.001 |
Token | that |
Token ID | 326 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
icked | +0.035 |
es | +0.033 |
together | +0.026 |
iled | +0.026 |
against | +0.024 |
Neg logit contributions | |
museum | -0.086 |
hotel | -0.084 |
city | -0.083 |
principal | -0.083 |
priest | -0.082 |
Token | it |
Token ID | 340 |
Feature activation | +6.62e-03 |
Pos logit contributions | |
icked | +0.015 |
es | +0.012 |
together | +0.011 |
there | +0.010 |
on | +0.010 |
Neg logit contributions | |
city | -0.033 |
hallway | -0.032 |
hotel | -0.032 |
museum | -0.032 |
principal | -0.032 |
Token | had |
Token ID | 550 |
Feature activation | -2.49e-02 |
Pos logit contributions | |
hallway | +0.012 |
city | +0.012 |
hotel | +0.012 |
principal | +0.012 |
doorway | +0.012 |
Neg logit contributions | |
icked | -0.011 |
together | -0.010 |
there | -0.010 |
on | -0.010 |
es | -0.009 |
Token | helped |
Token ID | 4193 |
Feature activation | -2.67e-02 |
Pos logit contributions | |
icked | +0.039 |
es | +0.038 |
together | +0.038 |
there | +0.037 |
on | +0.037 |
Neg logit contributions | |
hallway | -0.046 |
principal | -0.046 |
hotel | -0.045 |
city | -0.045 |
doorway | -0.045 |
Token | fix |
Token ID | 4259 |
Feature activation | -2.23e-02 |
Pos logit contributions | |
icked | +0.031 |
es | +0.026 |
together | +0.025 |
there | +0.025 |
on | +0.023 |
Neg logit contributions | |
principal | -0.063 |
city | -0.063 |
hallway | -0.062 |
museum | -0.062 |
hotel | -0.062 |
Token | the |
Token ID | 262 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
icked | +0.025 |
together | +0.023 |
there | +0.022 |
es | +0.021 |
on | +0.021 |
Neg logit contributions | |
hallway | -0.053 |
principal | -0.053 |
hotel | -0.053 |
city | -0.053 |
ey | -0.052 |
Token | earth |
Token ID | 4534 |
Feature activation | -3.12e-03 |
Pos logit contributions | |
principal | +0.236 |
salon | +0.221 |
babys | +0.218 |
Ta | +0.218 |
priest | +0.215 |
Neg logit contributions | |
there | -0.404 |
on | -0.393 |
together | -0.386 |
in | -0.383 |
it | -0.372 |
Token | . |
Token ID | 13 |
Feature activation | -9.52e-03 |
Pos logit contributions | |
icked | +0.003 |
es | +0.002 |
together | +0.002 |
there | +0.002 |
on | +0.002 |
Neg logit contributions | |
hallway | -0.008 |
city | -0.008 |
hotel | -0.008 |
classroom | -0.008 |
principal | -0.008 |
Token | He |
Token ID | 679 |
Feature activation | +3.78e-03 |
Pos logit contributions | |
icked | +0.014 |
there | +0.014 |
together | +0.014 |
on | +0.013 |
es | +0.013 |
Neg logit contributions | |
hallway | -0.019 |
principal | -0.019 |
city | -0.019 |
hotel | -0.019 |
ey | -0.019 |
Token | sang |
Token ID | 25889 |
Feature activation | -5.19e-02 |
Pos logit contributions | |
hallway | +0.008 |
principal | +0.008 |
hotel | +0.008 |
city | +0.008 |
ey | +0.008 |
Neg logit contributions | |
icked | -0.006 |
together | -0.006 |
there | -0.005 |
on | -0.005 |
es | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | +3.09e-02 |
Pos logit contributions | |
icked | +0.066 |
es | +0.053 |
there | +0.053 |
together | +0.049 |
on | +0.048 |
Neg logit contributions | |
orchestra | -0.114 |
city | -0.111 |
hallway | -0.110 |
audience | -0.110 |
principal | -0.110 |
Token | , |
Token ID | 11 |
Feature activation | -4.03e-02 |
Pos logit contributions | |
icked | +0.050 |
es | +0.039 |
there | +0.037 |
together | +0.035 |
on | +0.034 |
Neg logit contributions | |
museum | -0.135 |
city | -0.135 |
hallway | -0.134 |
hotel | -0.134 |
principal | -0.131 |
Token | they |
Token ID | 484 |
Feature activation | -5.68e-03 |
Pos logit contributions | |
icked | +0.040 |
es | +0.031 |
together | +0.030 |
there | +0.030 |
on | +0.029 |
Neg logit contributions | |
city | -0.104 |
hallway | -0.103 |
principal | -0.103 |
hotel | -0.102 |
museum | -0.101 |
Token | finally |
Token ID | 3443 |
Feature activation | -2.92e-02 |
Pos logit contributions | |
icked | +0.007 |
together | +0.007 |
there | +0.007 |
on | +0.007 |
es | +0.007 |
Neg logit contributions | |
hallway | -0.012 |
principal | -0.012 |
ey | -0.012 |
hotel | -0.012 |
city | -0.012 |
Token | discovered |
Token ID | 5071 |
Feature activation | -3.98e-02 |
Pos logit contributions | |
icked | +0.040 |
together | +0.040 |
there | +0.040 |
on | +0.039 |
es | +0.038 |
Neg logit contributions | |
hallway | -0.062 |
principal | -0.061 |
ey | -0.061 |
hotel | -0.061 |
city | -0.060 |
Token | why |
Token ID | 1521 |
Feature activation | -3.82e-03 |
Pos logit contributions | |
icked | +0.037 |
together | +0.034 |
there | +0.033 |
es | +0.032 |
on | +0.032 |
Neg logit contributions | |
hotel | -0.100 |
city | -0.100 |
hallway | -0.100 |
principal | -0.100 |
doorway | -0.098 |
Token | the |
Token ID | 262 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
icked | +0.004 |
together | +0.003 |
es | +0.003 |
on | +0.003 |
there | +0.003 |
Neg logit contributions | |
city | -0.009 |
hotel | -0.009 |
principal | -0.009 |
hallway | -0.009 |
museum | -0.009 |
Token | fox |
Token ID | 21831 |
Feature activation | -9.51e-03 |
Pos logit contributions | |
principal | +0.295 |
ey | +0.294 |
Ta | +0.286 |
hallway | +0.285 |
orchestra | +0.285 |
Neg logit contributions | |
on | -0.319 |
there | -0.316 |
it | -0.292 |
in | -0.291 |
as | -0.281 |
Token | was |
Token ID | 373 |
Feature activation | +5.58e-03 |
Pos logit contributions | |
icked | +0.014 |
together | +0.014 |
there | +0.013 |
on | +0.013 |
es | +0.013 |
Neg logit contributions | |
hallway | -0.019 |
principal | -0.019 |
city | -0.019 |
hotel | -0.018 |
classroom | -0.018 |
Token | different |
Token ID | 1180 |
Feature activation | +3.24e-02 |
Pos logit contributions | |
hallway | +0.011 |
principal | +0.011 |
hotel | +0.011 |
city | +0.011 |
classroom | +0.011 |
Neg logit contributions | |
icked | -0.008 |
together | -0.008 |
es | -0.008 |
there | -0.008 |
on | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | -6.80e-03 |
Pos logit contributions | |
hallway | +0.080 |
principal | +0.080 |
hotel | +0.079 |
city | +0.079 |
ey | +0.079 |
Neg logit contributions | |
icked | -0.033 |
together | -0.032 |
there | -0.031 |
on | -0.030 |
es | -0.029 |
Token | It |
Token ID | 632 |
Feature activation | +6.85e-03 |
Pos logit contributions | |
icked | +0.008 |
together | +0.008 |
there | +0.008 |
on | +0.008 |
es | +0.008 |
Neg logit contributions | |
hallway | -0.015 |
principal | -0.015 |
hotel | -0.015 |
city | -0.015 |
ey | -0.015 |
Token | happy |
Token ID | 3772 |
Feature activation | -2.96e-02 |
Pos logit contributions | |
hallway | +0.018 |
principal | +0.018 |
hotel | +0.018 |
ey | +0.018 |
city | +0.018 |
Neg logit contributions | |
together | -0.017 |
there | -0.017 |
icked | -0.016 |
on | -0.016 |
es | -0.015 |
Token | to |
Token ID | 284 |
Feature activation | +2.80e-03 |
Pos logit contributions | |
icked | +0.045 |
es | +0.033 |
bled | +0.032 |
pped | +0.030 |
hot | +0.028 |
Neg logit contributions | |
playground | -0.063 |
classroom | -0.063 |
principal | -0.061 |
city | -0.061 |
hotel | -0.061 |
Token | finally |
Token ID | 3443 |
Feature activation | -1.34e-02 |
Pos logit contributions | |
hallway | +0.005 |
principal | +0.005 |
hotel | +0.005 |
city | +0.005 |
ey | +0.005 |
Neg logit contributions | |
icked | -0.005 |
together | -0.005 |
there | -0.005 |
on | -0.004 |
es | -0.004 |
Token | understand |
Token ID | 1833 |
Feature activation | -1.81e-02 |
Pos logit contributions | |
icked | +0.024 |
es | +0.023 |
together | +0.023 |
there | +0.022 |
on | +0.021 |
Neg logit contributions | |
hallway | -0.023 |
doorway | -0.022 |
classroom | -0.022 |
playground | -0.022 |
city | -0.022 |
Token | why |
Token ID | 1521 |
Feature activation | +6.97e-03 |
Pos logit contributions | |
icked | +0.019 |
es | +0.016 |
together | +0.015 |
there | +0.014 |
on | +0.013 |
Neg logit contributions | |
principal | -0.046 |
hotel | -0.045 |
city | -0.045 |
classroom | -0.044 |
hallway | -0.044 |
Token | the |
Token ID | 262 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
city | +0.016 |
classroom | +0.016 |
principal | +0.016 |
babys | +0.016 |
hallway | +0.016 |
Neg logit contributions | |
icked | -0.009 |
es | -0.007 |
together | -0.007 |
on | -0.006 |
against | -0.006 |
Token | screw |
Token ID | 9580 |
Feature activation | +7.65e-03 |
Pos logit contributions | |
actor | +0.278 |
orchestra | +0.278 |
ey | +0.275 |
Ta | +0.270 |
principal | +0.268 |
Neg logit contributions | |
there | -0.360 |
on | -0.352 |
it | -0.339 |
in | -0.320 |
together | -0.318 |
Token | was |
Token ID | 373 |
Feature activation | -6.16e-04 |
Pos logit contributions | |
hallway | +0.015 |
principal | +0.015 |
city | +0.015 |
hotel | +0.015 |
doorway | +0.015 |
Neg logit contributions | |
icked | -0.011 |
together | -0.011 |
there | -0.010 |
es | -0.010 |
on | -0.010 |
Token | hidden |
Token ID | 7104 |
Feature activation | -3.05e-02 |
Pos logit contributions | |
icked | +0.001 |
es | +0.001 |
together | +0.001 |
bed | +0.001 |
instead | +0.001 |
Neg logit contributions | |
doorway | -0.001 |
hallway | -0.001 |
babys | -0.001 |
city | -0.001 |
classroom | -0.001 |
Token | in |
Token ID | 287 |
Feature activation | +1.50e-02 |
Pos logit contributions | |
icked | +0.034 |
es | +0.025 |
together | +0.024 |
there | +0.021 |
able | +0.019 |
Neg logit contributions | |
hotel | -0.073 |
hallway | -0.072 |
city | -0.072 |
doorway | -0.071 |
principal | -0.071 |
Token | the |
Token ID | 262 |
Feature activation | +1.35e-01 |
Pos logit contributions | |
principal | +0.039 |
hallway | +0.039 |
hotel | +0.039 |
ey | +0.039 |
city | +0.039 |
Neg logit contributions | |
together | -0.013 |
there | -0.013 |
icked | -0.012 |
on | -0.012 |
es | -0.011 |
Token | neighbor |
Token ID | 4780 |
Feature activation | -1.51e-02 |
Pos logit contributions | |
ey | +0.174 |
Ta | +0.166 |
hallway | +0.165 |
doorway | +0.158 |
orchestra | +0.154 |
Neg logit contributions | |
together | -0.176 |
there | -0.173 |
on | -0.172 |
in | -0.165 |
at | -0.161 |
Token | to |
Token ID | 284 |
Feature activation | -1.78e-03 |
Pos logit contributions | |
icked | +0.013 |
together | +0.010 |
there | +0.010 |
es | +0.010 |
on | +0.010 |
Neg logit contributions | |
principal | -0.041 |
city | -0.040 |
hotel | -0.040 |
classroom | -0.040 |
museum | -0.040 |
Token | help |
Token ID | 1037 |
Feature activation | -2.76e-02 |
Pos logit contributions | |
icked | +0.003 |
together | +0.003 |
there | +0.003 |
on | +0.003 |
es | +0.003 |
Neg logit contributions | |
hallway | -0.003 |
principal | -0.003 |
hotel | -0.003 |
city | -0.003 |
ey | -0.003 |
Token | them |
Token ID | 606 |
Feature activation | +2.41e-03 |
Pos logit contributions | |
icked | +0.044 |
es | +0.036 |
there | +0.034 |
together | +0.034 |
on | +0.033 |
Neg logit contributions | |
babys | -0.055 |
hallway | -0.055 |
principal | -0.055 |
city | -0.055 |
hotel | -0.054 |
Token | repair |
Token ID | 9185 |
Feature activation | -3.24e-02 |
Pos logit contributions | |
city | +0.005 |
classroom | +0.005 |
hallway | +0.005 |
museum | +0.005 |
playground | +0.005 |
Neg logit contributions | |
icked | -0.004 |
es | -0.003 |
bled | -0.003 |
there | -0.003 |
pped | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
icked | +0.038 |
es | +0.031 |
there | +0.030 |
together | +0.030 |
on | +0.028 |
Neg logit contributions | |
city | -0.077 |
hotel | -0.077 |
hallway | -0.077 |
principal | -0.076 |
museum | -0.075 |
Token | tractor |
Token ID | 38278 |
Feature activation | -3.24e-02 |
Pos logit contributions | |
principal | +0.271 |
hallway | +0.268 |
ey | +0.265 |
hotel | +0.259 |
orchestra | +0.258 |
Neg logit contributions | |
there | -0.332 |
together | -0.331 |
on | -0.328 |
icked | -0.324 |
es | -0.310 |
Token | . |
Token ID | 13 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
icked | +0.033 |
es | +0.027 |
there | +0.026 |
together | +0.025 |
on | +0.023 |
Neg logit contributions | |
hallway | -0.082 |
city | -0.081 |
principal | -0.079 |
hotel | -0.079 |
street | -0.078 |
Token | |
Token ID | 220 |
Feature activation | +1.05e-02 |
Pos logit contributions | |
icked | +0.013 |
together | +0.013 |
there | +0.013 |
on | +0.012 |
es | +0.012 |
Neg logit contributions | |
hallway | -0.026 |
principal | -0.026 |
hotel | -0.026 |
city | -0.026 |
ey | -0.026 |
Token |
|
Token ID | 198 |
Feature activation | -2.78e-03 |
Pos logit contributions | |
principal | +0.020 |
hallway | +0.020 |
city | +0.020 |
classroom | +0.020 |
hotel | +0.020 |
Neg logit contributions | |
icked | -0.016 |
on | -0.015 |
together | -0.014 |
there | -0.014 |
es | -0.014 |
Token |
|
Token ID | 198 |
Feature activation | -3.94e-03 |
Pos logit contributions | |
icked | +0.004 |
together | +0.003 |
there | +0.003 |
on | +0.003 |
es | +0.003 |
Neg logit contributions | |
hallway | -0.006 |
principal | -0.006 |
hotel | -0.006 |
city | -0.006 |
ey | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | +2.78e-03 |
Pos logit contributions | |
icked | +0.037 |
together | +0.032 |
there | +0.030 |
on | +0.029 |
es | +0.029 |
Neg logit contributions | |
principal | -0.066 |
city | -0.065 |
hallway | -0.065 |
babys | -0.065 |
classroom | -0.064 |
Token | The |
Token ID | 383 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
principal | +0.006 |
hallway | +0.006 |
city | +0.006 |
classroom | +0.006 |
ey | +0.006 |
Neg logit contributions | |
icked | -0.004 |
together | -0.004 |
there | -0.004 |
on | -0.004 |
es | -0.004 |
Token | fox |
Token ID | 21831 |
Feature activation | -1.31e-02 |
Pos logit contributions | |
hallway | +0.253 |
ey | +0.252 |
principal | +0.248 |
tant | +0.246 |
orchestra | +0.246 |
Neg logit contributions | |
there | -0.258 |
on | -0.249 |
in | -0.226 |
it | -0.223 |
with | -0.216 |
Token | explained |
Token ID | 4893 |
Feature activation | -3.85e-02 |
Pos logit contributions | |
icked | +0.020 |
together | +0.019 |
there | +0.019 |
on | +0.019 |
es | +0.018 |
Neg logit contributions | |
principal | -0.025 |
hallway | -0.025 |
city | -0.025 |
hotel | -0.025 |
priest | -0.024 |
Token | why |
Token ID | 1521 |
Feature activation | -6.09e-03 |
Pos logit contributions | |
icked | +0.035 |
es | +0.029 |
together | +0.027 |
against | +0.025 |
bed | +0.025 |
Neg logit contributions | |
principal | -0.099 |
museum | -0.097 |
city | -0.096 |
priest | -0.096 |
ostr | -0.095 |
Token | the |
Token ID | 262 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
icked | +0.007 |
together | +0.006 |
es | +0.006 |
on | +0.005 |
against | +0.005 |
Neg logit contributions | |
principal | -0.015 |
museum | -0.014 |
classroom | -0.014 |
city | -0.014 |
hotel | -0.014 |
Token | bird |
Token ID | 6512 |
Feature activation | -6.49e-03 |
Pos logit contributions | |
ey | +0.359 |
principal | +0.345 |
hallway | +0.343 |
igsaw | +0.337 |
orchestra | +0.332 |
Neg logit contributions | |
there | -0.297 |
on | -0.281 |
in | -0.264 |
it | -0.251 |
together | -0.246 |
Token | bit |
Token ID | 1643 |
Feature activation | -2.33e-02 |
Pos logit contributions | |
icked | +0.010 |
together | +0.009 |
there | +0.009 |
on | +0.009 |
es | +0.009 |
Neg logit contributions | |
hallway | -0.013 |
principal | -0.013 |
city | -0.013 |
hotel | -0.013 |
ey | -0.012 |
Token | him |
Token ID | 683 |
Feature activation | +9.32e-03 |
Pos logit contributions | |
icked | +0.025 |
together | +0.024 |
there | +0.023 |
es | +0.023 |
on | +0.022 |
Neg logit contributions | |
principal | -0.055 |
hallway | -0.055 |
ey | -0.055 |
hotel | -0.055 |
city | -0.054 |
Token | - |
Token ID | 532 |
Feature activation | -2.21e-02 |
Pos logit contributions | |
city | +0.020 |
museum | +0.020 |
principal | +0.020 |
classroom | +0.020 |
library | +0.020 |
Neg logit contributions | |
icked | -0.012 |
together | -0.009 |
there | -0.009 |
es | -0.009 |
bled | -0.008 |
Token | it |
Token ID | 340 |
Feature activation | +2.74e-03 |
Pos logit contributions | |
icked | +0.021 |
together | +0.019 |
es | +0.018 |
there | +0.017 |
on | +0.016 |
Neg logit contributions | |
principal | -0.056 |
city | -0.056 |
hallway | -0.055 |
hotel | -0.055 |
priest | -0.055 |
Token | and |
Token ID | 290 |
Feature activation | -2.50e-02 |
Pos logit contributions | |
icked | +0.031 |
es | +0.024 |
there | +0.021 |
instead | +0.019 |
together | +0.019 |
Neg logit contributions | |
city | -0.077 |
hallway | -0.076 |
classroom | -0.074 |
neighbourhood | -0.074 |
hotel | -0.074 |
Token | worked |
Token ID | 3111 |
Feature activation | -5.36e-02 |
Pos logit contributions | |
icked | +0.036 |
together | +0.035 |
there | +0.035 |
on | +0.034 |
es | +0.034 |
Neg logit contributions | |
hallway | -0.051 |
principal | -0.051 |
hotel | -0.050 |
city | -0.050 |
ey | -0.049 |
Token | hard |
Token ID | 1327 |
Feature activation | -1.08e-02 |
Pos logit contributions | |
icked | +0.056 |
es | +0.039 |
pped | +0.033 |
bed | +0.029 |
there | +0.028 |
Neg logit contributions | |
hallway | -0.120 |
city | -0.118 |
museum | -0.118 |
tant | -0.117 |
hotel | -0.116 |
Token | to |
Token ID | 284 |
Feature activation | +6.59e-03 |
Pos logit contributions | |
icked | +0.013 |
es | +0.009 |
bled | +0.009 |
pped | +0.008 |
there | +0.008 |
Neg logit contributions | |
hallway | -0.024 |
city | -0.024 |
museum | -0.024 |
street | -0.023 |
hotel | -0.023 |
Token | restore |
Token ID | 11169 |
Feature activation | -2.55e-02 |
Pos logit contributions | |
hallway | +0.012 |
principal | +0.012 |
ey | +0.011 |
hotel | +0.011 |
city | +0.011 |
Neg logit contributions | |
together | -0.011 |
there | -0.011 |
icked | -0.011 |
on | -0.011 |
es | -0.010 |
Token | the |
Token ID | 262 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
icked | +0.026 |
together | +0.024 |
there | +0.024 |
es | +0.023 |
on | +0.023 |
Neg logit contributions | |
hallway | -0.063 |
hotel | -0.062 |
city | -0.062 |
principal | -0.062 |
classroom | -0.061 |
Token | mountain |
Token ID | 8598 |
Feature activation | -2.71e-02 |
Pos logit contributions | |
Ta | +0.261 |
principal | +0.256 |
babys | +0.251 |
orchestra | +0.240 |
salon | +0.240 |
Neg logit contributions | |
there | -0.391 |
on | -0.373 |
in | -0.353 |
it | -0.343 |
together | -0.340 |
Token | . |
Token ID | 13 |
Feature activation | -5.78e-03 |
Pos logit contributions | |
icked | +0.025 |
there | +0.020 |
es | +0.019 |
together | +0.019 |
on | +0.018 |
Neg logit contributions | |
hallway | -0.071 |
city | -0.071 |
hotel | -0.070 |
principal | -0.069 |
doorway | -0.069 |
Token | With |
Token ID | 2080 |
Feature activation | -7.19e-03 |
Pos logit contributions | |
icked | +0.007 |
together | +0.007 |
there | +0.006 |
on | +0.006 |
es | +0.006 |
Neg logit contributions | |
hallway | -0.013 |
principal | -0.013 |
ey | -0.013 |
hotel | -0.013 |
city | -0.013 |
Token | help |
Token ID | 1037 |
Feature activation | -2.64e-02 |
Pos logit contributions | |
icked | +0.007 |
together | +0.006 |
there | +0.006 |
on | +0.006 |
es | +0.006 |
Neg logit contributions | |
hallway | -0.019 |
principal | -0.019 |
hotel | -0.018 |
city | -0.018 |
ey | -0.018 |
Token | from |
Token ID | 422 |
Feature activation | +1.34e-02 |
Pos logit contributions | |
icked | +0.025 |
es | +0.018 |
there | +0.018 |
together | +0.017 |
on | +0.016 |
Neg logit contributions | |
hallway | -0.067 |
city | -0.067 |
hotel | -0.066 |
principal | -0.065 |
babys | -0.065 |
Token | , |
Token ID | 11 |
Feature activation | -5.04e-03 |
Pos logit contributions | |
principal | +0.005 |
hotel | +0.005 |
babys | +0.005 |
city | +0.005 |
hallway | +0.005 |
Neg logit contributions | |
icked | -0.001 |
together | -0.001 |
es | -0.001 |
there | -0.001 |
on | -0.001 |
Token | can |
Token ID | 460 |
Feature activation | -3.00e-02 |
Pos logit contributions | |
there | +0.007 |
on | +0.007 |
icked | +0.006 |
together | +0.006 |
it | +0.006 |
Neg logit contributions | |
hallway | -0.011 |
principal | -0.011 |
ey | -0.011 |
orchestra | -0.011 |
doorway | -0.011 |
Token | you |
Token ID | 345 |
Feature activation | +1.31e-02 |
Pos logit contributions | |
icked | +0.040 |
together | +0.038 |
es | +0.036 |
on | +0.034 |
against | +0.031 |
Neg logit contributions | |
babys | -0.066 |
principal | -0.065 |
hallway | -0.065 |
hotel | -0.064 |
city | -0.064 |
Token | please |
Token ID | 3387 |
Feature activation | -2.73e-02 |
Pos logit contributions | |
hotel | +0.020 |
babys | +0.020 |
principal | +0.020 |
hallway | +0.019 |
priest | +0.019 |
Neg logit contributions | |
icked | -0.026 |
es | -0.025 |
together | -0.025 |
there | -0.023 |
on | -0.023 |
Token | replace |
Token ID | 6330 |
Feature activation | -1.67e-02 |
Pos logit contributions | |
icked | +0.054 |
es | +0.054 |
together | +0.052 |
there | +0.048 |
on | +0.048 |
Neg logit contributions | |
hotel | -0.042 |
hallway | -0.041 |
babys | -0.041 |
classroom | -0.041 |
principal | -0.041 |
Token | the |
Token ID | 262 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
icked | +0.025 |
together | +0.022 |
es | +0.022 |
on | +0.021 |
there | +0.021 |
Neg logit contributions | |
city | -0.034 |
hotel | -0.034 |
principal | -0.033 |
ey | -0.033 |
classroom | -0.033 |
Token | rock |
Token ID | 3881 |
Feature activation | -2.76e-02 |
Pos logit contributions | |
principal | +0.301 |
Ta | +0.293 |
orchestra | +0.285 |
courtroom | +0.283 |
igsaw | +0.279 |
Neg logit contributions | |
there | -0.352 |
on | -0.331 |
in | -0.309 |
it | -0.303 |
at | -0.293 |
Token | you |
Token ID | 345 |
Feature activation | +1.72e-02 |
Pos logit contributions | |
icked | +0.034 |
es | +0.029 |
together | +0.028 |
there | +0.024 |
it | +0.023 |
Neg logit contributions | |
hotel | -0.063 |
classroom | -0.062 |
principal | -0.062 |
museum | -0.062 |
hallway | -0.062 |
Token | moved |
Token ID | 3888 |
Feature activation | -3.57e-02 |
Pos logit contributions | |
hotel | +0.038 |
museum | +0.037 |
library | +0.037 |
classroom | +0.037 |
city | +0.037 |
Neg logit contributions | |
icked | -0.023 |
es | -0.020 |
together | -0.019 |
against | -0.016 |
able | -0.015 |
Token | from |
Token ID | 422 |
Feature activation | +3.78e-03 |
Pos logit contributions | |
icked | +0.049 |
es | +0.044 |
together | +0.036 |
able | +0.034 |
they | +0.031 |
Neg logit contributions | |
classroom | -0.079 |
hotel | -0.079 |
city | -0.077 |
hallway | -0.077 |
museum | -0.077 |
Token | the |
Token ID | 262 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
city | +0.008 |
hotel | +0.008 |
hallway | +0.008 |
principal | +0.008 |
museum | +0.008 |
Neg logit contributions | |
icked | -0.005 |
es | -0.005 |
together | -0.005 |
there | -0.004 |
on | -0.004 |
Token | the |
Token ID | 262 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
icked | +0.040 |
es | +0.034 |
together | +0.029 |
on | +0.027 |
there | +0.027 |
Neg logit contributions | |
city | -0.080 |
hotel | -0.080 |
hallway | -0.079 |
principal | -0.079 |
babys | -0.079 |
Token | boy |
Token ID | 2933 |
Feature activation | -5.36e-04 |
Pos logit contributions | |
orchestra | +0.281 |
ey | +0.281 |
hallway | +0.267 |
priest | +0.266 |
principal | +0.265 |
Neg logit contributions | |
there | -0.264 |
on | -0.262 |
together | -0.255 |
in | -0.243 |
instead | -0.240 |
Token | finally |
Token ID | 3443 |
Feature activation | -2.30e-02 |
Pos logit contributions | |
icked | +0.001 |
together | +0.001 |
there | +0.001 |
on | +0.001 |
es | +0.001 |
Neg logit contributions | |
hallway | -0.001 |
principal | -0.001 |
ey | -0.001 |
hotel | -0.001 |
city | -0.001 |
Token | understood |
Token ID | 7247 |
Feature activation | -3.63e-02 |
Pos logit contributions | |
icked | +0.034 |
there | +0.033 |
together | +0.033 |
on | +0.032 |
es | +0.032 |
Neg logit contributions | |
hallway | -0.047 |
principal | -0.046 |
ey | -0.046 |
hotel | -0.046 |
city | -0.046 |
Token | why |
Token ID | 1521 |
Feature activation | +9.68e-04 |
Pos logit contributions | |
icked | +0.039 |
es | +0.032 |
together | +0.031 |
there | +0.028 |
on | +0.027 |
Neg logit contributions | |
principal | -0.090 |
hotel | -0.088 |
city | -0.088 |
museum | -0.087 |
priest | -0.087 |
Token | the |
Token ID | 262 |
Feature activation | +1.72e-01 |
Pos logit contributions | |
principal | +0.002 |
city | +0.002 |
hotel | +0.002 |
babys | +0.002 |
hallway | +0.002 |
Neg logit contributions | |
icked | -0.001 |
together | -0.001 |
es | -0.001 |
on | -0.001 |
there | -0.001 |
Token | chip |
Token ID | 11594 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
orchestra | +0.349 |
ey | +0.343 |
principal | +0.335 |
hallway | +0.332 |
Ta | +0.327 |
Neg logit contributions | |
on | -0.265 |
there | -0.261 |
together | -0.256 |
instead | -0.242 |
it | -0.242 |
Token | m |
Token ID | 76 |
Feature activation | +1.22e-01 |
Pos logit contributions | |
hallway | +0.203 |
principal | +0.199 |
hotel | +0.198 |
orchestra | +0.198 |
ey | +0.197 |
Neg logit contributions | |
together | -0.208 |
there | -0.204 |
icked | -0.204 |
on | -0.203 |
es | -0.201 |
Token | unk |
Token ID | 2954 |
Feature activation | +8.13e-03 |
Pos logit contributions | |
hallway | +0.202 |
principal | +0.200 |
hotel | +0.197 |
city | +0.196 |
ey | +0.195 |
Neg logit contributions | |
icked | -0.220 |
together | -0.219 |
there | -0.216 |
on | -0.211 |
es | -0.209 |
Token | had |
Token ID | 550 |
Feature activation | -1.18e-02 |
Pos logit contributions | |
principal | +0.016 |
hallway | +0.016 |
city | +0.016 |
hotel | +0.016 |
priest | +0.016 |
Neg logit contributions | |
icked | -0.012 |
together | -0.011 |
there | -0.011 |
es | -0.011 |
on | -0.010 |
Token | been |
Token ID | 587 |
Feature activation | +4.30e-03 |
Pos logit contributions | |
icked | +0.016 |
es | +0.015 |
together | +0.015 |
there | +0.014 |
on | +0.014 |
Neg logit contributions | |
principal | -0.025 |
city | -0.025 |
hallway | -0.025 |
hotel | -0.024 |
classroom | -0.024 |
Token | boy |
Token ID | 2933 |
Feature activation | -2.54e-03 |
Pos logit contributions | |
priest | +0.201 |
orchestra | +0.200 |
ey | +0.199 |
salon | +0.193 |
doorway | +0.191 |
Neg logit contributions | |
there | -0.307 |
on | -0.297 |
in | -0.276 |
together | -0.268 |
at | -0.265 |
Token | worked |
Token ID | 3111 |
Feature activation | -4.85e-02 |
Pos logit contributions | |
icked | +0.004 |
there | +0.004 |
together | +0.004 |
es | +0.004 |
on | +0.004 |
Neg logit contributions | |
hallway | -0.005 |
principal | -0.005 |
city | -0.005 |
hotel | -0.005 |
classroom | -0.005 |
Token | for |
Token ID | 329 |
Feature activation | +3.64e-03 |
Pos logit contributions | |
icked | +0.089 |
es | +0.079 |
there | +0.072 |
pped | +0.066 |
iled | +0.066 |
Neg logit contributions | |
hallway | -0.079 |
classroom | -0.075 |
principal | -0.075 |
museum | -0.075 |
doorway | -0.074 |
Token | hours |
Token ID | 2250 |
Feature activation | -3.97e-02 |
Pos logit contributions | |
hallway | +0.009 |
principal | +0.009 |
hotel | +0.009 |
ey | +0.009 |
city | +0.009 |
Neg logit contributions | |
together | -0.004 |
icked | -0.004 |
there | -0.003 |
on | -0.003 |
es | -0.003 |
Token | putting |
Token ID | 5137 |
Feature activation | -2.19e-02 |
Pos logit contributions | |
icked | +0.051 |
es | +0.035 |
there | +0.033 |
together | +0.030 |
bled | +0.029 |
Neg logit contributions | |
hallway | -0.091 |
city | -0.089 |
principal | -0.088 |
playground | -0.087 |
classroom | -0.086 |
Token | the |
Token ID | 262 |
Feature activation | +1.72e-01 |
Pos logit contributions | |
icked | +0.028 |
es | +0.025 |
together | +0.024 |
there | +0.024 |
on | +0.022 |
Neg logit contributions | |
hallway | -0.048 |
city | -0.048 |
hotel | -0.048 |
ey | -0.048 |
principal | -0.047 |
Token | pieces |
Token ID | 5207 |
Feature activation | -2.83e-02 |
Pos logit contributions | |
priest | +0.287 |
salon | +0.277 |
principal | +0.276 |
orchestra | +0.266 |
babys | +0.264 |
Neg logit contributions | |
on | -0.346 |
there | -0.337 |
together | -0.323 |
in | -0.314 |
it | -0.310 |
Token | of |
Token ID | 286 |
Feature activation | +8.55e-02 |
Pos logit contributions | |
icked | +0.027 |
es | +0.023 |
there | +0.018 |
bed | +0.017 |
together | +0.016 |
Neg logit contributions | |
hallway | -0.069 |
hotel | -0.068 |
city | -0.067 |
doorway | -0.067 |
ey | -0.066 |
Token | gear |
Token ID | 7733 |
Feature activation | -5.23e-02 |
Pos logit contributions | |
hallway | +0.129 |
principal | +0.127 |
hotel | +0.125 |
city | +0.125 |
ey | +0.124 |
Neg logit contributions | |
icked | -0.170 |
together | -0.167 |
there | -0.165 |
on | -0.163 |
es | -0.162 |
Token | together |
Token ID | 1978 |
Feature activation | -3.51e-02 |
Pos logit contributions | |
icked | +0.046 |
es | +0.033 |
there | +0.032 |
together | +0.026 |
pped | +0.025 |
Neg logit contributions | |
hallway | -0.134 |
hotel | -0.132 |
city | -0.131 |
museum | -0.130 |
classroom | -0.130 |
Token | . |
Token ID | 13 |
Feature activation | -8.21e-03 |
Pos logit contributions | |
icked | +0.052 |
es | +0.039 |
pped | +0.036 |
bled | +0.034 |
there | +0.034 |
Neg logit contributions | |
city | -0.070 |
hallway | -0.070 |
playground | -0.069 |
museum | -0.068 |
hotel | -0.068 |
Token | she |
Token ID | 673 |
Feature activation | -2.15e-03 |
Pos logit contributions | |
icked | +0.026 |
es | +0.022 |
together | +0.021 |
against | +0.020 |
on | +0.019 |
Neg logit contributions | |
principal | -0.076 |
priest | -0.074 |
babys | -0.074 |
hotel | -0.074 |
salon | -0.073 |
Token | would |
Token ID | 561 |
Feature activation | -3.35e-02 |
Pos logit contributions | |
icked | +0.003 |
together | +0.003 |
there | +0.003 |
on | +0.003 |
es | +0.003 |
Neg logit contributions | |
hallway | -0.004 |
principal | -0.004 |
hotel | -0.004 |
ey | -0.004 |
city | -0.004 |
Token | try |
Token ID | 1949 |
Feature activation | -4.81e-02 |
Pos logit contributions | |
icked | +0.060 |
together | +0.058 |
es | +0.058 |
there | +0.058 |
on | +0.056 |
Neg logit contributions | |
principal | -0.057 |
hallway | -0.057 |
city | -0.056 |
hotel | -0.055 |
classroom | -0.055 |
Token | to |
Token ID | 284 |
Feature activation | +1.56e-03 |
Pos logit contributions | |
icked | +0.047 |
es | +0.037 |
there | +0.034 |
together | +0.034 |
on | +0.029 |
Neg logit contributions | |
principal | -0.124 |
classroom | -0.123 |
city | -0.121 |
babys | -0.120 |
hotel | -0.120 |
Token | fix |
Token ID | 4259 |
Feature activation | -3.13e-02 |
Pos logit contributions | |
hallway | +0.003 |
hotel | +0.003 |
principal | +0.003 |
doorway | +0.003 |
city | +0.003 |
Neg logit contributions | |
together | -0.002 |
there | -0.002 |
on | -0.002 |
icked | -0.002 |
in | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | +1.72e-01 |
Pos logit contributions | |
icked | +0.028 |
together | +0.027 |
there | +0.026 |
on | +0.025 |
es | +0.025 |
Neg logit contributions | |
principal | -0.081 |
hallway | -0.081 |
hotel | -0.080 |
city | -0.080 |
ey | -0.079 |
Token | train |
Token ID | 4512 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
principal | +0.282 |
Ta | +0.281 |
priest | +0.273 |
orchestra | +0.272 |
hallway | +0.270 |
Neg logit contributions | |
there | -0.352 |
on | -0.342 |
together | -0.333 |
it | -0.317 |
in | -0.314 |
Token | . |
Token ID | 13 |
Feature activation | -2.70e-03 |
Pos logit contributions | |
icked | +0.011 |
there | +0.009 |
together | +0.009 |
es | +0.009 |
on | +0.008 |
Neg logit contributions | |
hallway | -0.030 |
principal | -0.030 |
city | -0.030 |
hotel | -0.030 |
classroom | -0.030 |
Token | She |
Token ID | 1375 |
Feature activation | -2.54e-03 |
Pos logit contributions | |
icked | +0.004 |
es | +0.004 |
together | +0.004 |
on | +0.004 |
there | +0.003 |
Neg logit contributions | |
principal | -0.006 |
hallway | -0.006 |
classroom | -0.006 |
city | -0.006 |
ey | -0.005 |
Token | said |
Token ID | 531 |
Feature activation | -3.16e-02 |
Pos logit contributions | |
icked | +0.003 |
together | +0.003 |
there | +0.003 |
on | +0.003 |
es | +0.003 |
Neg logit contributions | |
hallway | -0.006 |
ey | -0.006 |
city | -0.006 |
hotel | -0.006 |
orchestra | -0.006 |
Token | they |
Token ID | 484 |
Feature activation | +1.17e-04 |
Pos logit contributions | |
icked | +0.030 |
es | +0.026 |
together | +0.025 |
against | +0.023 |
on | +0.022 |
Neg logit contributions | |
principal | -0.080 |
priest | -0.078 |
salon | -0.078 |
babys | -0.078 |
hotel | -0.077 |
Token | person |
Token ID | 1048 |
Feature activation | -6.45e-03 |
Pos logit contributions | |
ey | +0.128 |
hallway | +0.125 |
principal | +0.123 |
city | +0.120 |
hotel | +0.120 |
Neg logit contributions | |
there | -0.308 |
together | -0.304 |
on | -0.304 |
icked | -0.297 |
es | -0.297 |
Token | who |
Token ID | 508 |
Feature activation | +5.80e-03 |
Pos logit contributions | |
icked | +0.006 |
together | +0.005 |
es | +0.005 |
there | +0.005 |
on | +0.005 |
Neg logit contributions | |
principal | -0.016 |
hallway | -0.016 |
hotel | -0.016 |
city | -0.016 |
classroom | -0.016 |
Token | must |
Token ID | 1276 |
Feature activation | -2.69e-03 |
Pos logit contributions | |
hallway | +0.012 |
principal | +0.012 |
hotel | +0.012 |
city | +0.012 |
ey | +0.012 |
Neg logit contributions | |
icked | -0.008 |
together | -0.008 |
there | -0.008 |
on | -0.008 |
es | -0.007 |
Token | have |
Token ID | 423 |
Feature activation | +2.65e-05 |
Pos logit contributions | |
icked | +0.004 |
together | +0.004 |
there | +0.004 |
on | +0.004 |
es | +0.004 |
Neg logit contributions | |
hallway | -0.005 |
principal | -0.005 |
hotel | -0.005 |
city | -0.005 |
ey | -0.005 |
Token | owned |
Token ID | 6898 |
Feature activation | -7.44e-03 |
Pos logit contributions | |
principal | +0.000 |
hotel | +0.000 |
city | +0.000 |
priest | +0.000 |
hallway | +0.000 |
Neg logit contributions | |
icked | -0.000 |
es | -0.000 |
together | -0.000 |
on | -0.000 |
there | -0.000 |
Token | the |
Token ID | 262 |
Feature activation | +1.72e-01 |
Pos logit contributions | |
icked | +0.007 |
together | +0.006 |
es | +0.006 |
there | +0.006 |
on | +0.005 |
Neg logit contributions | |
city | -0.019 |
hallway | -0.019 |
hotel | -0.019 |
principal | -0.019 |
doorway | -0.019 |
Token | mug |
Token ID | 25152 |
Feature activation | -2.10e-02 |
Pos logit contributions | |
ey | +0.282 |
principal | +0.269 |
priest | +0.266 |
hallway | +0.256 |
orchestra | +0.254 |
Neg logit contributions | |
there | -0.348 |
on | -0.346 |
together | -0.335 |
in | -0.319 |
it | -0.315 |
Token | . |
Token ID | 13 |
Feature activation | -4.92e-03 |
Pos logit contributions | |
icked | +0.020 |
es | +0.016 |
there | +0.015 |
together | +0.015 |
on | +0.014 |
Neg logit contributions | |
hallway | -0.054 |
principal | -0.054 |
hotel | -0.053 |
city | -0.053 |
classroom | -0.053 |
Token | She |
Token ID | 1375 |
Feature activation | +6.37e-03 |
Pos logit contributions | |
icked | +0.006 |
together | +0.006 |
there | +0.006 |
on | +0.006 |
es | +0.006 |
Neg logit contributions | |
hallway | -0.011 |
principal | -0.011 |
hotel | -0.011 |
city | -0.011 |
ey | -0.011 |
Token | thought |
Token ID | 1807 |
Feature activation | -2.20e-02 |
Pos logit contributions | |
hallway | +0.014 |
ey | +0.014 |
principal | +0.014 |
hotel | +0.014 |
city | +0.014 |
Neg logit contributions | |
icked | -0.008 |
together | -0.008 |
there | -0.008 |
on | -0.008 |
es | -0.008 |
Token | that |
Token ID | 326 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
icked | +0.020 |
es | +0.016 |
together | +0.015 |
there | +0.013 |
on | +0.012 |
Neg logit contributions | |
hotel | -0.057 |
principal | -0.057 |
city | -0.056 |
hallway | -0.056 |
orchestra | -0.056 |
Token | the |
Token ID | 262 |
Feature activation | +8.48e-02 |
Pos logit contributions | |
together | +0.023 |
there | +0.022 |
icked | +0.022 |
on | +0.021 |
es | +0.020 |
Neg logit contributions | |
principal | -0.070 |
hallway | -0.069 |
hotel | -0.068 |
ey | -0.068 |
city | -0.068 |
Token | rough |
Token ID | 5210 |
Feature activation | +6.64e-02 |
Pos logit contributions | |
principal | +0.117 |
priest | +0.113 |
Ta | +0.109 |
babys | +0.108 |
child | +0.104 |
Neg logit contributions | |
on | -0.189 |
in | -0.184 |
there | -0.183 |
with | -0.171 |
together | -0.169 |
Token | and |
Token ID | 290 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
principal | +0.079 |
priest | +0.076 |
Ta | +0.074 |
orchestra | +0.071 |
babys | +0.071 |
Neg logit contributions | |
there | -0.153 |
on | -0.153 |
together | -0.150 |
in | -0.145 |
it | -0.145 |
Token | bump |
Token ID | 13852 |
Feature activation | -5.21e-02 |
Pos logit contributions | |
icked | +0.018 |
together | +0.018 |
there | +0.018 |
on | +0.018 |
es | +0.017 |
Neg logit contributions | |
hallway | -0.021 |
principal | -0.021 |
hotel | -0.021 |
city | -0.021 |
ey | -0.021 |
Token | y |
Token ID | 88 |
Feature activation | +1.35e-03 |
Pos logit contributions | |
icked | +0.063 |
together | +0.052 |
es | +0.051 |
there | +0.050 |
able | +0.047 |
Neg logit contributions | |
hallway | -0.113 |
tant | -0.109 |
street | -0.107 |
city | -0.107 |
hotel | -0.106 |
Token | terrain |
Token ID | 15510 |
Feature activation | -9.93e-02 |
Pos logit contributions | |
hallway | +0.003 |
principal | +0.003 |
hotel | +0.003 |
city | +0.003 |
ey | +0.003 |
Neg logit contributions | |
icked | -0.001 |
together | -0.001 |
there | -0.001 |
es | -0.001 |
on | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -1.84e-02 |
Pos logit contributions | |
icked | +0.103 |
es | +0.084 |
there | +0.078 |
together | +0.077 |
on | +0.067 |
Neg logit contributions | |
hallway | -0.250 |
city | -0.242 |
hotel | -0.239 |
doorway | -0.238 |
street | -0.235 |
Token |
|
Token ID | 198 |
Feature activation | -9.62e-03 |
Pos logit contributions | |
icked | +0.024 |
together | +0.023 |
there | +0.023 |
on | +0.022 |
es | +0.022 |
Neg logit contributions | |
hallway | -0.040 |
principal | -0.040 |
hotel | -0.040 |
city | -0.040 |
ey | -0.039 |
Token |
|
Token ID | 198 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
together | +0.011 |
there | +0.011 |
icked | +0.010 |
on | +0.010 |
es | +0.010 |
Neg logit contributions | |
hallway | -0.023 |
principal | -0.023 |
hotel | -0.023 |
ey | -0.023 |
orchestra | -0.023 |
Token | Finally |
Token ID | 11158 |
Feature activation | -2.00e-02 |
Pos logit contributions | |
there | +0.018 |
together | +0.018 |
icked | +0.016 |
on | +0.016 |
es | +0.016 |
Neg logit contributions | |
hotel | -0.027 |
ey | -0.027 |
principal | -0.027 |
hallway | -0.027 |
orchestra | -0.026 |
Token | , |
Token ID | 11 |
Feature activation | -3.20e-02 |
Pos logit contributions | |
icked | +0.026 |
es | +0.021 |
hot | +0.020 |
iled | +0.020 |
together | +0.020 |
Neg logit contributions | |
hotel | -0.045 |
city | -0.043 |
hallway | -0.042 |
classroom | -0.041 |
hall | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | -1.58e-03 |
Pos logit contributions | |
icked | +0.043 |
together | +0.037 |
there | +0.037 |
es | +0.035 |
able | +0.029 |
Neg logit contributions | |
hallway | -0.126 |
city | -0.126 |
principal | -0.125 |
salon | -0.125 |
bus | -0.124 |
Token | They |
Token ID | 1119 |
Feature activation | +1.40e-02 |
Pos logit contributions | |
icked | +0.002 |
together | +0.002 |
there | +0.002 |
es | +0.002 |
on | +0.002 |
Neg logit contributions | |
hallway | -0.003 |
principal | -0.003 |
hotel | -0.003 |
city | -0.003 |
classroom | -0.003 |
Token | wear |
Token ID | 5806 |
Feature activation | -4.67e-02 |
Pos logit contributions | |
doorway | +0.034 |
principal | +0.034 |
hallway | +0.033 |
hotel | +0.033 |
babys | +0.033 |
Neg logit contributions | |
on | -0.017 |
there | -0.017 |
in | -0.016 |
together | -0.016 |
as | -0.016 |
Token | helmets |
Token ID | 28359 |
Feature activation | -7.82e-02 |
Pos logit contributions | |
icked | +0.081 |
together | +0.076 |
es | +0.076 |
there | +0.075 |
on | +0.072 |
Neg logit contributions | |
city | -0.082 |
hallway | -0.082 |
hotel | -0.080 |
principal | -0.080 |
classroom | -0.080 |
Token | and |
Token ID | 290 |
Feature activation | -1.08e-02 |
Pos logit contributions | |
icked | +0.090 |
es | +0.078 |
there | +0.074 |
able | +0.066 |
together | +0.065 |
Neg logit contributions | |
city | -0.178 |
hallway | -0.177 |
bus | -0.175 |
street | -0.174 |
salon | -0.173 |
Token | pads |
Token ID | 21226 |
Feature activation | -9.90e-02 |
Pos logit contributions | |
on | +0.018 |
together | +0.018 |
there | +0.018 |
icked | +0.018 |
as | +0.017 |
Neg logit contributions | |
principal | -0.020 |
hotel | -0.020 |
doorway | -0.020 |
hallway | -0.019 |
ey | -0.019 |
Token | to |
Token ID | 284 |
Feature activation | -1.11e-03 |
Pos logit contributions | |
icked | +0.099 |
there | +0.086 |
es | +0.084 |
able | +0.077 |
together | +0.074 |
Neg logit contributions | |
bus | -0.229 |
city | -0.227 |
salon | -0.226 |
principal | -0.224 |
priest | -0.222 |
Token | be |
Token ID | 307 |
Feature activation | -7.77e-03 |
Pos logit contributions | |
on | +0.002 |
together | +0.002 |
there | +0.002 |
it | +0.002 |
in | +0.002 |
Neg logit contributions | |
ey | -0.002 |
doorway | -0.002 |
hotel | -0.002 |
hallway | -0.002 |
principal | -0.002 |
Token | safe |
Token ID | 3338 |
Feature activation | -3.30e-02 |
Pos logit contributions | |
icked | +0.014 |
es | +0.012 |
together | +0.012 |
there | +0.012 |
instead | +0.012 |
Neg logit contributions | |
city | -0.013 |
classroom | -0.013 |
hallway | -0.013 |
principal | -0.013 |
orchestra | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
icked | +0.045 |
es | +0.032 |
pped | +0.031 |
bled | +0.028 |
able | +0.028 |
Neg logit contributions | |
city | -0.077 |
hotel | -0.074 |
street | -0.073 |
classroom | -0.073 |
hallway | -0.073 |
Token |
|
Token ID | 198 |
Feature activation | -1.10e-02 |
Pos logit contributions | |
icked | +0.020 |
together | +0.019 |
there | +0.019 |
es | +0.018 |
on | +0.018 |
Neg logit contributions | |
hallway | -0.033 |
principal | -0.032 |
hotel | -0.032 |
city | -0.032 |
classroom | -0.032 |
Token | sk |
Token ID | 1341 |
Feature activation | +2.14e-02 |
Pos logit contributions | |
icked | +0.080 |
es | +0.074 |
there | +0.073 |
together | +0.072 |
on | +0.069 |
Neg logit contributions | |
hallway | -0.095 |
city | -0.095 |
hotel | -0.095 |
ey | -0.094 |
classroom | -0.094 |
Token | ates |
Token ID | 689 |
Feature activation | -7.12e-02 |
Pos logit contributions | |
principal | +0.038 |
ey | +0.037 |
classroom | +0.036 |
hallway | +0.035 |
city | +0.035 |
Neg logit contributions | |
there | -0.036 |
together | -0.034 |
on | -0.034 |
icked | -0.033 |
it | -0.033 |
Token | and |
Token ID | 290 |
Feature activation | -1.79e-02 |
Pos logit contributions | |
icked | +0.092 |
able | +0.077 |
es | +0.070 |
pped | +0.069 |
there | +0.068 |
Neg logit contributions | |
tant | -0.138 |
city | -0.138 |
hotel | -0.138 |
museum | -0.135 |
classroom | -0.135 |
Token | helmets |
Token ID | 28359 |
Feature activation | -7.56e-02 |
Pos logit contributions | |
icked | +0.031 |
together | +0.031 |
there | +0.030 |
on | +0.030 |
es | +0.030 |
Neg logit contributions | |
hallway | -0.031 |
principal | -0.031 |
hotel | -0.030 |
city | -0.030 |
ey | -0.030 |
Token | and |
Token ID | 290 |
Feature activation | -1.48e-02 |
Pos logit contributions | |
icked | +0.077 |
es | +0.067 |
there | +0.063 |
together | +0.057 |
able | +0.056 |
Neg logit contributions | |
city | -0.177 |
hallway | -0.176 |
bus | -0.176 |
street | -0.176 |
hotel | -0.176 |
Token | pads |
Token ID | 21226 |
Feature activation | -9.63e-02 |
Pos logit contributions | |
icked | +0.026 |
together | +0.026 |
there | +0.025 |
on | +0.025 |
es | +0.025 |
Neg logit contributions | |
hallway | -0.026 |
principal | -0.025 |
hotel | -0.025 |
city | -0.025 |
ey | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | -2.47e-02 |
Pos logit contributions | |
icked | +0.104 |
there | +0.089 |
es | +0.089 |
able | +0.080 |
together | +0.078 |
Neg logit contributions | |
salon | -0.213 |
city | -0.212 |
principal | -0.211 |
bus | -0.211 |
hotel | -0.208 |
Token | They |
Token ID | 1119 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
icked | +0.035 |
there | +0.032 |
together | +0.032 |
on | +0.032 |
es | +0.032 |
Neg logit contributions | |
principal | -0.051 |
hallway | -0.050 |
city | -0.050 |
classroom | -0.049 |
orchestra | -0.049 |
Token | use |
Token ID | 779 |
Feature activation | -4.16e-02 |
Pos logit contributions | |
on | +0.014 |
together | +0.014 |
there | +0.014 |
icked | +0.014 |
instead | +0.013 |
Neg logit contributions | |
doorway | -0.022 |
hallway | -0.022 |
principal | -0.022 |
ey | -0.022 |
classroom | -0.021 |
Token | a |
Token ID | 257 |
Feature activation | +5.40e-02 |
Pos logit contributions | |
icked | +0.053 |
es | +0.048 |
there | +0.048 |
together | +0.047 |
on | +0.045 |
Neg logit contributions | |
hallway | -0.091 |
hotel | -0.091 |
city | -0.090 |
principal | -0.090 |
ey | -0.089 |
Token | puck |
Token ID | 20695 |
Feature activation | -6.72e-02 |
Pos logit contributions | |
principal | +0.093 |
hallway | +0.089 |
priest | +0.088 |
ey | +0.088 |
city | +0.088 |
Neg logit contributions | |
on | -0.101 |
together | -0.100 |
there | -0.099 |
it | -0.091 |
icked | -0.091 |
Token | but |
Token ID | 475 |
Feature activation | -9.82e-03 |
Pos logit contributions | |
icked | +0.036 |
together | +0.032 |
es | +0.032 |
there | +0.032 |
on | +0.031 |
Neg logit contributions | |
hallway | -0.076 |
principal | -0.075 |
city | -0.074 |
hotel | -0.074 |
classroom | -0.073 |
Token | then |
Token ID | 788 |
Feature activation | -2.50e-02 |
Pos logit contributions | |
icked | +0.010 |
es | +0.008 |
together | +0.007 |
there | +0.007 |
on | +0.007 |
Neg logit contributions | |
hallway | -0.025 |
city | -0.025 |
principal | -0.025 |
hotel | -0.024 |
classroom | -0.024 |
Token | the |
Token ID | 262 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
icked | +0.026 |
together | +0.022 |
es | +0.020 |
on | +0.020 |
against | +0.019 |
Neg logit contributions | |
principal | -0.060 |
hotel | -0.060 |
city | -0.060 |
hallway | -0.059 |
priest | -0.059 |
Token | vegetables |
Token ID | 13701 |
Feature activation | -2.73e-02 |
Pos logit contributions | |
principal | +0.221 |
orchestra | +0.218 |
salon | +0.217 |
hotel | +0.217 |
ey | +0.215 |
Neg logit contributions | |
there | -0.254 |
on | -0.253 |
in | -0.242 |
together | -0.241 |
it | -0.229 |
Token | g |
Token ID | 308 |
Feature activation | +4.81e-02 |
Pos logit contributions | |
icked | +0.041 |
together | +0.040 |
es | +0.038 |
there | +0.038 |
on | +0.036 |
Neg logit contributions | |
hallway | -0.052 |
hotel | -0.051 |
principal | -0.051 |
city | -0.051 |
doorway | -0.050 |
Token | anged |
Token ID | 5102 |
Feature activation | -9.58e-02 |
Pos logit contributions | |
ey | +0.084 |
hotel | +0.082 |
hallway | +0.082 |
city | +0.081 |
principal | +0.081 |
Neg logit contributions | |
there | -0.084 |
together | -0.081 |
on | -0.081 |
icked | -0.078 |
es | -0.076 |
Token | up |
Token ID | 510 |
Feature activation | -4.88e-02 |
Pos logit contributions | |
icked | +0.077 |
es | +0.054 |
there | +0.053 |
able | +0.051 |
together | +0.048 |
Neg logit contributions | |
hallway | -0.238 |
tant | -0.233 |
city | -0.229 |
doorway | -0.228 |
principal | -0.227 |
Token | and |
Token ID | 290 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
icked | +0.062 |
es | +0.054 |
there | +0.043 |
able | +0.041 |
together | +0.039 |
Neg logit contributions | |
hallway | -0.112 |
city | -0.111 |
doorway | -0.108 |
playground | -0.108 |
hotel | -0.107 |
Token | blocked |
Token ID | 10226 |
Feature activation | -6.18e-02 |
Pos logit contributions | |
icked | +0.018 |
es | +0.017 |
together | +0.015 |
there | +0.015 |
on | +0.014 |
Neg logit contributions | |
hallway | -0.028 |
doorway | -0.028 |
principal | -0.027 |
city | -0.027 |
priest | -0.027 |
Token | it |
Token ID | 340 |
Feature activation | -1.88e-02 |
Pos logit contributions | |
icked | +0.081 |
es | +0.070 |
there | +0.060 |
able | +0.057 |
pped | +0.057 |
Neg logit contributions | |
doorway | -0.131 |
city | -0.131 |
hallway | -0.130 |
hotel | -0.128 |
classroom | -0.128 |
Token | . |
Token ID | 13 |
Feature activation | -1.35e-02 |
Pos logit contributions | |
icked | +0.023 |
es | +0.018 |
there | +0.017 |
together | +0.016 |
bled | +0.015 |
Neg logit contributions | |
hallway | -0.044 |
doorway | -0.042 |
playground | -0.042 |
classroom | -0.041 |
street | -0.041 |
Token | unicorn |
Token ID | 44986 |
Feature activation | -2.38e-02 |
Pos logit contributions | |
ey | +0.198 |
principal | +0.178 |
courtroom | +0.178 |
hallway | +0.169 |
hotel | +0.167 |
Neg logit contributions | |
on | -0.325 |
in | -0.305 |
there | -0.304 |
it | -0.298 |
as | -0.298 |
Token | . |
Token ID | 13 |
Feature activation | -7.12e-03 |
Pos logit contributions | |
icked | +0.023 |
there | +0.019 |
together | +0.019 |
es | +0.017 |
on | +0.016 |
Neg logit contributions | |
hallway | -0.059 |
hotel | -0.058 |
classroom | -0.058 |
principal | -0.058 |
city | -0.058 |
Token | She |
Token ID | 1375 |
Feature activation | -9.83e-04 |
Pos logit contributions | |
icked | +0.012 |
es | +0.011 |
together | +0.011 |
on | +0.010 |
able | +0.010 |
Neg logit contributions | |
principal | -0.013 |
classroom | -0.013 |
hallway | -0.012 |
babys | -0.012 |
library | -0.012 |
Token | cut |
Token ID | 2005 |
Feature activation | -5.43e-02 |
Pos logit contributions | |
icked | +0.001 |
together | +0.001 |
there | +0.001 |
es | +0.001 |
on | +0.001 |
Neg logit contributions | |
hallway | -0.002 |
city | -0.002 |
principal | -0.002 |
hotel | -0.002 |
ey | -0.002 |
Token | a |
Token ID | 257 |
Feature activation | +5.93e-02 |
Pos logit contributions | |
icked | +0.050 |
together | +0.045 |
there | +0.045 |
able | +0.040 |
es | +0.039 |
Neg logit contributions | |
hallway | -0.130 |
principal | -0.128 |
ey | -0.127 |
doorway | -0.127 |
babys | -0.125 |
Token | hole |
Token ID | 7604 |
Feature activation | -9.52e-02 |
Pos logit contributions | |
principal | +0.099 |
hotel | +0.092 |
city | +0.091 |
mall | +0.089 |
courtroom | +0.089 |
Neg logit contributions | |
on | -0.123 |
together | -0.115 |
it | -0.114 |
there | -0.114 |
in | -0.111 |
Token | in |
Token ID | 287 |
Feature activation | +2.19e-03 |
Pos logit contributions | |
icked | +0.096 |
es | +0.084 |
Remy | +0.079 |
dd | +0.071 |
Buddy | +0.069 |
Neg logit contributions | |
hallway | -0.212 |
doorway | -0.210 |
mirror | -0.202 |
ey | -0.199 |
salon | -0.198 |
Token | the |
Token ID | 262 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
hallway | +0.006 |
principal | +0.006 |
hotel | +0.005 |
city | +0.005 |
ey | +0.005 |
Neg logit contributions | |
icked | -0.002 |
together | -0.002 |
there | -0.002 |
es | -0.002 |
on | -0.002 |
Token | middle |
Token ID | 3504 |
Feature activation | -3.40e-03 |
Pos logit contributions | |
principal | +0.168 |
priest | +0.161 |
orchestra | +0.156 |
courtroom | +0.154 |
pastor | +0.154 |
Neg logit contributions | |
on | -0.250 |
there | -0.234 |
in | -0.232 |
it | -0.224 |
together | -0.221 |
Token | for |
Token ID | 329 |
Feature activation | +2.37e-03 |
Pos logit contributions | |
icked | +0.004 |
together | +0.003 |
there | +0.003 |
es | +0.003 |
on | +0.003 |
Neg logit contributions | |
hallway | -0.008 |
city | -0.008 |
principal | -0.008 |
ey | -0.008 |
hotel | -0.008 |
Token | the |
Token ID | 262 |
Feature activation | +1.39e-01 |
Pos logit contributions | |
hallway | +0.006 |
principal | +0.006 |
hotel | +0.006 |
city | +0.006 |
ey | +0.006 |
Neg logit contributions | |
icked | -0.002 |
together | -0.002 |
there | -0.002 |
on | -0.002 |
es | -0.002 |
Token | garden |
Token ID | 11376 |
Feature activation | -6.73e-02 |
Pos logit contributions | |
principal | +0.074 |
hallway | +0.068 |
orchestra | +0.067 |
ey | +0.066 |
babys | +0.066 |
Neg logit contributions | |
there | -0.139 |
together | -0.138 |
on | -0.137 |
in | -0.131 |
icked | -0.131 |
Token | . |
Token ID | 13 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
icked | +0.068 |
es | +0.054 |
together | +0.052 |
there | +0.050 |
on | +0.050 |
Neg logit contributions | |
hallway | -0.172 |
city | -0.169 |
doorway | -0.168 |
hotel | -0.168 |
playground | -0.167 |
Token | She |
Token ID | 1375 |
Feature activation | +8.60e-03 |
Pos logit contributions | |
icked | +0.019 |
together | +0.019 |
there | +0.019 |
on | +0.018 |
es | +0.018 |
Neg logit contributions | |
hallway | -0.033 |
principal | -0.033 |
hotel | -0.032 |
city | -0.032 |
ey | -0.032 |
Token | dug |
Token ID | 18735 |
Feature activation | -6.48e-02 |
Pos logit contributions | |
principal | +0.017 |
hallway | +0.017 |
hotel | +0.017 |
city | +0.017 |
ey | +0.017 |
Neg logit contributions | |
icked | -0.013 |
together | -0.013 |
there | -0.012 |
on | -0.012 |
es | -0.012 |
Token | a |
Token ID | 257 |
Feature activation | +4.74e-02 |
Pos logit contributions | |
icked | +0.076 |
es | +0.067 |
together | +0.065 |
there | +0.064 |
on | +0.058 |
Neg logit contributions | |
hallway | -0.149 |
hotel | -0.147 |
ey | -0.146 |
doorway | -0.145 |
city | -0.145 |
Token | hole |
Token ID | 7604 |
Feature activation | -9.51e-02 |
Pos logit contributions | |
principal | +0.097 |
child | +0.086 |
babys | +0.085 |
alphabet | +0.085 |
orchestra | +0.083 |
Neg logit contributions | |
on | -0.083 |
together | -0.083 |
there | -0.082 |
in | -0.079 |
it | -0.079 |
Token | and |
Token ID | 290 |
Feature activation | -1.91e-02 |
Pos logit contributions | |
icked | +0.117 |
es | +0.097 |
pped | +0.082 |
Remy | +0.077 |
hot | +0.076 |
Neg logit contributions | |
doorway | -0.211 |
hallway | -0.210 |
playground | -0.194 |
city | -0.191 |
ey | -0.190 |
Token | planted |
Token ID | 18969 |
Feature activation | -5.25e-02 |
Pos logit contributions | |
icked | +0.028 |
together | +0.026 |
es | +0.026 |
on | +0.025 |
there | +0.025 |
Neg logit contributions | |
hallway | -0.039 |
doorway | -0.038 |
hotel | -0.037 |
city | -0.037 |
ey | -0.037 |
Token | the |
Token ID | 262 |
Feature activation | +1.29e-01 |
Pos logit contributions | |
icked | +0.060 |
es | +0.052 |
together | +0.049 |
able | +0.045 |
there | +0.044 |
Neg logit contributions | |
ey | -0.118 |
doorway | -0.118 |
city | -0.117 |
hallway | -0.117 |
hotel | -0.117 |
Token | sun |
Token ID | 4252 |
Feature activation | -1.45e-02 |
Pos logit contributions | |
principal | +0.228 |
orchestra | +0.206 |
salon | +0.205 |
courtroom | +0.200 |
carn | +0.200 |
Neg logit contributions | |
there | -0.254 |
on | -0.250 |
in | -0.239 |
together | -0.232 |
it | -0.227 |
Token | flower |
Token ID | 25547 |
Feature activation | -2.86e-02 |
Pos logit contributions | |
icked | +0.027 |
es | +0.024 |
together | +0.023 |
able | +0.023 |
there | +0.022 |
Neg logit contributions | |
hallway | -0.023 |
city | -0.022 |
ostr | -0.022 |
doorway | -0.022 |
hotel | -0.021 |
Token | They |
Token ID | 1119 |
Feature activation | -9.19e-03 |
Pos logit contributions | |
icked | +0.022 |
together | +0.021 |
there | +0.021 |
on | +0.020 |
es | +0.020 |
Neg logit contributions | |
hallway | -0.041 |
principal | -0.041 |
hotel | -0.040 |
city | -0.040 |
ey | -0.040 |
Token | wear |
Token ID | 5806 |
Feature activation | -5.08e-02 |
Pos logit contributions | |
there | +0.014 |
together | +0.013 |
on | +0.013 |
in | +0.012 |
instead | +0.012 |
Neg logit contributions | |
principal | -0.019 |
doorway | -0.019 |
ey | -0.019 |
hallway | -0.019 |
babys | -0.019 |
Token | colorful |
Token ID | 20239 |
Feature activation | +9.23e-03 |
Pos logit contributions | |
together | +0.090 |
icked | +0.089 |
there | +0.089 |
on | +0.087 |
es | +0.085 |
Neg logit contributions | |
hallway | -0.087 |
principal | -0.087 |
hotel | -0.085 |
city | -0.085 |
ey | -0.084 |
Token | clothes |
Token ID | 8242 |
Feature activation | -7.95e-02 |
Pos logit contributions | |
principal | +0.014 |
hallway | +0.013 |
doorway | +0.013 |
hotel | +0.013 |
priest | +0.013 |
Neg logit contributions | |
together | -0.019 |
on | -0.019 |
there | -0.019 |
icked | -0.018 |
es | -0.018 |
Token | and |
Token ID | 290 |
Feature activation | -1.06e-02 |
Pos logit contributions | |
icked | +0.112 |
es | +0.090 |
able | +0.081 |
there | +0.081 |
together | +0.079 |
Neg logit contributions | |
city | -0.167 |
hotel | -0.166 |
hallway | -0.163 |
bus | -0.163 |
classroom | -0.163 |
Token | hats |
Token ID | 23910 |
Feature activation | -9.51e-02 |
Pos logit contributions | |
together | +0.019 |
there | +0.019 |
on | +0.019 |
icked | +0.019 |
es | +0.018 |
Neg logit contributions | |
principal | -0.018 |
hallway | -0.018 |
hotel | -0.018 |
doorway | -0.018 |
ey | -0.017 |
Token | and |
Token ID | 290 |
Feature activation | -2.27e-03 |
Pos logit contributions | |
icked | +0.137 |
es | +0.107 |
able | +0.104 |
there | +0.102 |
bled | +0.101 |
Neg logit contributions | |
city | -0.188 |
bus | -0.186 |
hotel | -0.185 |
classroom | -0.182 |
priest | -0.182 |
Token | carry |
Token ID | 3283 |
Feature activation | -3.83e-02 |
Pos logit contributions | |
on | +0.004 |
there | +0.004 |
together | +0.004 |
icked | +0.004 |
es | +0.004 |
Neg logit contributions | |
principal | -0.004 |
doorway | -0.004 |
hallway | -0.004 |
hotel | -0.004 |
ey | -0.004 |
Token | flags |
Token ID | 9701 |
Feature activation | -7.59e-02 |
Pos logit contributions | |
icked | +0.063 |
together | +0.062 |
there | +0.062 |
on | +0.060 |
es | +0.060 |
Neg logit contributions | |
hallway | -0.070 |
principal | -0.070 |
hotel | -0.069 |
city | -0.068 |
ey | -0.068 |
Token | and |
Token ID | 290 |
Feature activation | -5.88e-03 |
Pos logit contributions | |
icked | +0.071 |
es | +0.055 |
there | +0.054 |
together | +0.051 |
on | +0.047 |
Neg logit contributions | |
city | -0.196 |
hotel | -0.195 |
hallway | -0.193 |
classroom | -0.192 |
priest | -0.191 |
Token | balloons |
Token ID | 35485 |
Feature activation | -7.03e-02 |
Pos logit contributions | |
on | +0.011 |
together | +0.011 |
there | +0.011 |
icked | +0.010 |
in | +0.010 |
Neg logit contributions | |
principal | -0.010 |
doorway | -0.010 |
hallway | -0.010 |
hotel | -0.009 |
ey | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | -1.05e-02 |
Pos logit contributions | |
icked | +0.054 |
together | +0.046 |
es | +0.046 |
there | +0.045 |
against | +0.038 |
Neg logit contributions | |
hallway | -0.137 |
hotel | -0.135 |
bus | -0.135 |
salon | -0.135 |
babys | -0.134 |
Token | She |
Token ID | 1375 |
Feature activation | +1.26e-02 |
Pos logit contributions | |
icked | +0.016 |
es | +0.015 |
together | +0.015 |
there | +0.015 |
on | +0.015 |
Neg logit contributions | |
hallway | -0.021 |
city | -0.021 |
principal | -0.021 |
classroom | -0.020 |
hotel | -0.020 |
Token | se |
Token ID | 384 |
Feature activation | +3.49e-02 |
Pos logit contributions | |
hallway | +0.025 |
principal | +0.025 |
hotel | +0.025 |
city | +0.025 |
ey | +0.024 |
Neg logit contributions | |
icked | -0.019 |
together | -0.018 |
there | -0.018 |
on | -0.018 |
es | -0.017 |
Token | wed |
Token ID | 19103 |
Feature activation | -6.05e-02 |
Pos logit contributions | |
hallway | +0.072 |
principal | +0.071 |
hotel | +0.070 |
city | +0.069 |
doorway | +0.069 |
Neg logit contributions | |
icked | -0.054 |
together | -0.052 |
es | -0.049 |
there | -0.049 |
on | -0.049 |
Token | a |
Token ID | 257 |
Feature activation | +5.97e-02 |
Pos logit contributions | |
icked | +0.052 |
there | +0.043 |
es | +0.042 |
together | +0.040 |
able | +0.038 |
Neg logit contributions | |
ey | -0.151 |
hallway | -0.150 |
tant | -0.149 |
city | -0.149 |
principal | -0.149 |
Token | hole |
Token ID | 7604 |
Feature activation | -9.48e-02 |
Pos logit contributions | |
principal | +0.109 |
hallway | +0.101 |
city | +0.100 |
orchestra | +0.099 |
priest | +0.099 |
Neg logit contributions | |
on | -0.117 |
together | -0.111 |
there | -0.111 |
it | -0.109 |
in | -0.108 |
Token | in |
Token ID | 287 |
Feature activation | +1.33e-03 |
Pos logit contributions | |
es | +0.097 |
icked | +0.095 |
Remy | +0.074 |
dd | +0.069 |
rums | +0.068 |
Neg logit contributions | |
hallway | -0.209 |
doorway | -0.207 |
bus | -0.204 |
elevator | -0.195 |
city | -0.194 |
Token | the |
Token ID | 262 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
hallway | +0.004 |
principal | +0.004 |
hotel | +0.003 |
city | +0.003 |
ey | +0.003 |
Neg logit contributions | |
icked | -0.001 |
together | -0.001 |
there | -0.001 |
on | -0.001 |
es | -0.001 |
Token | sail |
Token ID | 14936 |
Feature activation | -4.69e-02 |
Pos logit contributions | |
principal | +0.184 |
orchestra | +0.176 |
ey | +0.175 |
priest | +0.172 |
classroom | +0.171 |
Neg logit contributions | |
on | -0.236 |
there | -0.235 |
in | -0.221 |
it | -0.220 |
together | -0.215 |
Token | . |
Token ID | 13 |
Feature activation | -1.64e-02 |
Pos logit contributions | |
icked | +0.043 |
together | +0.037 |
there | +0.037 |
es | +0.036 |
on | +0.034 |
Neg logit contributions | |
hallway | -0.123 |
city | -0.120 |
principal | -0.120 |
hotel | -0.120 |
classroom | -0.118 |
Token | Now |
Token ID | 2735 |
Feature activation | -2.20e-02 |
Pos logit contributions | |
icked | +0.024 |
together | +0.023 |
es | +0.023 |
there | +0.023 |
on | +0.022 |
Neg logit contributions | |
hallway | -0.034 |
principal | -0.034 |
city | -0.033 |
classroom | -0.033 |
hotel | -0.033 |
Token | X |
Token ID | 1395 |
Feature activation | -4.06e-02 |
Pos logit contributions | |
Ta | +0.153 |
ey | +0.146 |
orchestra | +0.145 |
opl | +0.141 |
child | +0.131 |
Neg logit contributions | |
on | -0.265 |
there | -0.260 |
in | -0.255 |
it | -0.239 |
with | -0.236 |
Token | . |
Token ID | 13 |
Feature activation | -5.95e-03 |
Pos logit contributions | |
icked | +0.044 |
together | +0.040 |
there | +0.038 |
es | +0.035 |
on | +0.034 |
Neg logit contributions | |
hotel | -0.095 |
hallway | -0.095 |
priest | -0.095 |
principal | -0.095 |
salon | -0.094 |
Token | He |
Token ID | 679 |
Feature activation | -1.92e-03 |
Pos logit contributions | |
icked | +0.008 |
together | +0.008 |
es | +0.008 |
on | +0.008 |
there | +0.008 |
Neg logit contributions | |
principal | -0.013 |
hallway | -0.013 |
city | -0.012 |
hotel | -0.012 |
ey | -0.012 |
Token | digs |
Token ID | 46561 |
Feature activation | -6.05e-02 |
Pos logit contributions | |
icked | +0.003 |
there | +0.003 |
on | +0.003 |
es | +0.003 |
together | +0.003 |
Neg logit contributions | |
hallway | -0.003 |
orchestra | -0.003 |
babys | -0.003 |
city | -0.003 |
principal | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
icked | +0.067 |
es | +0.055 |
together | +0.055 |
there | +0.053 |
able | +0.052 |
Neg logit contributions | |
hotel | -0.142 |
hallway | -0.142 |
city | -0.140 |
ey | -0.138 |
museum | -0.138 |
Token | hole |
Token ID | 7604 |
Feature activation | -9.45e-02 |
Pos logit contributions | |
babys | +0.091 |
orchestra | +0.090 |
pastor | +0.090 |
principal | +0.090 |
salon | +0.087 |
Neg logit contributions | |
on | -0.091 |
there | -0.089 |
it | -0.084 |
with | -0.083 |
in | -0.083 |
Token | and |
Token ID | 290 |
Feature activation | -1.29e-02 |
Pos logit contributions | |
icked | +0.111 |
es | +0.097 |
able | +0.081 |
pped | +0.081 |
dd | +0.080 |
Neg logit contributions | |
hallway | -0.200 |
doorway | -0.196 |
bus | -0.184 |
hotel | -0.182 |
city | -0.180 |
Token | pulls |
Token ID | 16194 |
Feature activation | -4.95e-02 |
Pos logit contributions | |
icked | +0.023 |
together | +0.023 |
there | +0.023 |
on | +0.022 |
es | +0.022 |
Neg logit contributions | |
hallway | -0.022 |
principal | -0.022 |
hotel | -0.021 |
city | -0.021 |
ey | -0.021 |
Token | out |
Token ID | 503 |
Feature activation | -4.96e-02 |
Pos logit contributions | |
icked | +0.049 |
together | +0.040 |
es | +0.039 |
there | +0.038 |
able | +0.035 |
Neg logit contributions | |
hallway | -0.120 |
hotel | -0.119 |
city | -0.119 |
principal | -0.119 |
ey | -0.116 |
Token | a |
Token ID | 257 |
Feature activation | +5.77e-02 |
Pos logit contributions | |
icked | +0.059 |
together | +0.052 |
es | +0.049 |
bed | +0.048 |
able | +0.047 |
Neg logit contributions | |
hotel | -0.108 |
hallway | -0.107 |
doorway | -0.105 |
city | -0.105 |
museum | -0.102 |
Token | carrot |
Token ID | 40562 |
Feature activation | -4.40e-02 |
Pos logit contributions | |
principal | +0.095 |
orchestra | +0.094 |
hallway | +0.090 |
babys | +0.088 |
salon | +0.088 |
Neg logit contributions | |
on | -0.114 |
there | -0.113 |
it | -0.108 |
together | -0.107 |
in | -0.104 |
Token | tree |
Token ID | 5509 |
Feature activation | -4.93e-02 |
Pos logit contributions | |
principal | +0.158 |
actor | +0.146 |
Ta | +0.144 |
orchestra | +0.144 |
carn | +0.142 |
Neg logit contributions | |
on | -0.188 |
there | -0.186 |
in | -0.174 |
it | -0.166 |
together | -0.164 |
Token | , |
Token ID | 11 |
Feature activation | -3.18e-02 |
Pos logit contributions | |
icked | +0.049 |
together | +0.043 |
es | +0.041 |
on | +0.038 |
able | +0.038 |
Neg logit contributions | |
hallway | -0.125 |
city | -0.124 |
hotel | -0.123 |
classroom | -0.121 |
salon | -0.120 |
Token | they |
Token ID | 484 |
Feature activation | +1.13e-02 |
Pos logit contributions | |
icked | +0.021 |
es | +0.016 |
together | +0.016 |
on | +0.012 |
able | +0.011 |
Neg logit contributions | |
hallway | -0.092 |
city | -0.092 |
hotel | -0.091 |
principal | -0.090 |
classroom | -0.090 |
Token | dug |
Token ID | 18735 |
Feature activation | -6.26e-02 |
Pos logit contributions | |
hallway | +0.023 |
principal | +0.023 |
ey | +0.023 |
hotel | +0.023 |
orchestra | +0.023 |
Neg logit contributions | |
there | -0.016 |
icked | -0.016 |
together | -0.016 |
on | -0.016 |
es | -0.015 |
Token | a |
Token ID | 257 |
Feature activation | +4.32e-02 |
Pos logit contributions | |
icked | +0.073 |
es | +0.064 |
there | +0.059 |
together | +0.059 |
able | +0.055 |
Neg logit contributions | |
hallway | -0.143 |
hotel | -0.142 |
city | -0.141 |
playground | -0.138 |
doorway | -0.138 |
Token | hole |
Token ID | 7604 |
Feature activation | -9.39e-02 |
Pos logit contributions | |
principal | +0.087 |
child | +0.081 |
orchestra | +0.080 |
babys | +0.079 |
Ta | +0.079 |
Neg logit contributions | |
there | -0.075 |
on | -0.075 |
together | -0.075 |
in | -0.071 |
it | -0.071 |
Token | and |
Token ID | 290 |
Feature activation | -1.52e-02 |
Pos logit contributions | |
es | +0.106 |
icked | +0.100 |
Remy | +0.082 |
rums | +0.075 |
pped | +0.075 |
Neg logit contributions | |
hallway | -0.205 |
doorway | -0.204 |
playground | -0.193 |
street | -0.191 |
bus | -0.187 |
Token | found |
Token ID | 1043 |
Feature activation | -4.10e-02 |
Pos logit contributions | |
icked | +0.022 |
together | +0.021 |
es | +0.021 |
on | +0.020 |
there | +0.020 |
Neg logit contributions | |
hallway | -0.031 |
hotel | -0.030 |
doorway | -0.030 |
principal | -0.030 |
city | -0.030 |
Token | a |
Token ID | 257 |
Feature activation | +4.97e-02 |
Pos logit contributions | |
icked | +0.040 |
es | +0.040 |
together | +0.040 |
on | +0.038 |
there | +0.037 |
Neg logit contributions | |
doorway | -0.097 |
hotel | -0.094 |
hallway | -0.094 |
principal | -0.093 |
city | -0.093 |
Token | shiny |
Token ID | 22441 |
Feature activation | +4.71e-02 |
Pos logit contributions | |
principal | +0.083 |
orchestra | +0.081 |
Ta | +0.080 |
ey | +0.079 |
child | +0.078 |
Neg logit contributions | |
there | -0.099 |
on | -0.096 |
together | -0.093 |
it | -0.090 |
in | -0.089 |
Token | box |
Token ID | 3091 |
Feature activation | -6.45e-02 |
Pos logit contributions | |
principal | +0.064 |
hallway | +0.064 |
hotel | +0.062 |
city | +0.061 |
orchestra | +0.061 |
Neg logit contributions | |
there | -0.099 |
together | -0.099 |
icked | -0.099 |
on | -0.098 |
es | -0.095 |
Token | moved |
Token ID | 3888 |
Feature activation | -6.82e-02 |
Pos logit contributions | |
icked | +0.008 |
together | +0.008 |
there | +0.008 |
on | +0.008 |
es | +0.008 |
Neg logit contributions | |
principal | -0.012 |
hallway | -0.012 |
ey | -0.012 |
hotel | -0.012 |
city | -0.012 |
Token | his |
Token ID | 465 |
Feature activation | +7.59e-02 |
Pos logit contributions | |
icked | +0.080 |
es | +0.073 |
together | +0.061 |
there | +0.058 |
on | +0.053 |
Neg logit contributions | |
hallway | -0.156 |
city | -0.152 |
principal | -0.152 |
orchestra | -0.152 |
ey | -0.151 |
Token | feet |
Token ID | 3625 |
Feature activation | -8.10e-02 |
Pos logit contributions | |
principal | +0.134 |
Ta | +0.130 |
hallway | +0.130 |
orchestra | +0.128 |
city | +0.128 |
Neg logit contributions | |
on | -0.138 |
there | -0.132 |
in | -0.130 |
together | -0.128 |
as | -0.123 |
Token | back |
Token ID | 736 |
Feature activation | -1.88e-02 |
Pos logit contributions | |
icked | +0.080 |
es | +0.070 |
there | +0.057 |
together | +0.055 |
it | +0.051 |
Neg logit contributions | |
hallway | -0.199 |
doorway | -0.192 |
playground | -0.191 |
bus | -0.191 |
city | -0.189 |
Token | and |
Token ID | 290 |
Feature activation | -1.47e-02 |
Pos logit contributions | |
icked | +0.017 |
es | +0.014 |
there | +0.012 |
together | +0.012 |
on | +0.010 |
Neg logit contributions | |
hallway | -0.050 |
city | -0.049 |
classroom | -0.049 |
hotel | -0.048 |
doorway | -0.048 |
Token | forth |
Token ID | 6071 |
Feature activation | -9.37e-02 |
Pos logit contributions | |
icked | +0.019 |
together | +0.017 |
es | +0.017 |
there | +0.017 |
on | +0.016 |
Neg logit contributions | |
hallway | -0.033 |
principal | -0.032 |
hotel | -0.032 |
city | -0.032 |
doorway | -0.031 |
Token | to |
Token ID | 284 |
Feature activation | -6.65e-03 |
Pos logit contributions | |
icked | +0.101 |
es | +0.078 |
there | +0.069 |
able | +0.060 |
together | +0.059 |
Neg logit contributions | |
hallway | -0.223 |
playground | -0.218 |
classroom | -0.214 |
doorway | -0.214 |
principal | -0.214 |
Token | practice |
Token ID | 3357 |
Feature activation | -6.82e-02 |
Pos logit contributions | |
there | +0.006 |
together | +0.006 |
on | +0.006 |
it | +0.006 |
in | +0.006 |
Neg logit contributions | |
hallway | -0.017 |
ey | -0.017 |
babys | -0.017 |
hotel | -0.016 |
principal | -0.016 |
Token | standing |
Token ID | 5055 |
Feature activation | -4.07e-02 |
Pos logit contributions | |
icked | +0.073 |
there | +0.059 |
es | +0.057 |
together | +0.057 |
on | +0.053 |
Neg logit contributions | |
hallway | -0.171 |
principal | -0.169 |
city | -0.166 |
classroom | -0.166 |
hotel | -0.165 |
Token | on |
Token ID | 319 |
Feature activation | -7.36e-03 |
Pos logit contributions | |
icked | +0.045 |
es | +0.037 |
together | +0.035 |
there | +0.034 |
on | +0.031 |
Neg logit contributions | |
hallway | -0.099 |
city | -0.099 |
principal | -0.098 |
classroom | -0.097 |
hotel | -0.097 |
Token | one |
Token ID | 530 |
Feature activation | +2.50e-02 |
Pos logit contributions | |
icked | +0.008 |
together | +0.007 |
there | +0.007 |
es | +0.007 |
on | +0.006 |
Neg logit contributions | |
hallway | -0.018 |
principal | -0.018 |
city | -0.018 |
hotel | -0.018 |
classroom | -0.018 |
Token | he |
Token ID | 339 |
Feature activation | +6.48e-03 |
Pos logit contributions | |
icked | +0.025 |
es | +0.021 |
together | +0.021 |
on | +0.020 |
there | +0.019 |
Neg logit contributions | |
city | -0.062 |
hallway | -0.062 |
priest | -0.062 |
hotel | -0.062 |
principal | -0.061 |
Token | crawled |
Token ID | 45668 |
Feature activation | -6.86e-02 |
Pos logit contributions | |
hallway | +0.014 |
principal | +0.014 |
hotel | +0.014 |
city | +0.014 |
priest | +0.013 |
Neg logit contributions | |
icked | -0.009 |
together | -0.008 |
there | -0.008 |
on | -0.008 |
es | -0.008 |
Token | through |
Token ID | 832 |
Feature activation | -2.65e-02 |
Pos logit contributions | |
icked | +0.083 |
es | +0.079 |
together | +0.066 |
able | +0.059 |
there | +0.059 |
Neg logit contributions | |
hallway | -0.156 |
doorway | -0.155 |
playground | -0.147 |
hotel | -0.147 |
bus | -0.144 |
Token | a |
Token ID | 257 |
Feature activation | +3.91e-02 |
Pos logit contributions | |
icked | +0.025 |
together | +0.022 |
es | +0.020 |
there | +0.020 |
on | +0.020 |
Neg logit contributions | |
hallway | -0.070 |
city | -0.069 |
hotel | -0.069 |
principal | -0.068 |
ey | -0.067 |
Token | small |
Token ID | 1402 |
Feature activation | +5.13e-02 |
Pos logit contributions | |
principal | +0.074 |
babys | +0.071 |
Ta | +0.069 |
priest | +0.067 |
actor | +0.067 |
Neg logit contributions | |
there | -0.071 |
on | -0.067 |
it | -0.066 |
together | -0.064 |
in | -0.063 |
Token | hole |
Token ID | 7604 |
Feature activation | -9.34e-02 |
Pos logit contributions | |
principal | +0.087 |
Ta | +0.082 |
priest | +0.080 |
babys | +0.078 |
salon | +0.075 |
Neg logit contributions | |
there | -0.103 |
together | -0.099 |
on | -0.098 |
in | -0.094 |
it | -0.093 |
Token | . |
Token ID | 13 |
Feature activation | -1.35e-02 |
Pos logit contributions | |
icked | +0.101 |
es | +0.088 |
Remy | +0.074 |
together | +0.073 |
okay | +0.071 |
Neg logit contributions | |
hallway | -0.212 |
doorway | -0.210 |
playground | -0.197 |
priest | -0.196 |
bus | -0.195 |
Token |
|
Token ID | 198 |
Feature activation | -5.89e-03 |
Pos logit contributions | |
icked | +0.018 |
together | +0.017 |
there | +0.017 |
on | +0.017 |
es | +0.016 |
Neg logit contributions | |
hallway | -0.029 |
principal | -0.029 |
hotel | -0.029 |
city | -0.029 |
ey | -0.029 |
Token |
|
Token ID | 198 |
Feature activation | -9.16e-03 |
Pos logit contributions | |
together | +0.007 |
icked | +0.007 |
there | +0.007 |
on | +0.007 |
es | +0.006 |
Neg logit contributions | |
hallway | -0.014 |
principal | -0.014 |
hotel | -0.013 |
city | -0.013 |
ey | -0.013 |
Token | On |
Token ID | 2202 |
Feature activation | -1.85e-02 |
Pos logit contributions | |
there | +0.012 |
together | +0.012 |
icked | +0.012 |
on | +0.011 |
es | +0.011 |
Neg logit contributions | |
principal | -0.020 |
hotel | -0.020 |
hallway | -0.020 |
ey | -0.020 |
city | -0.020 |
Token | the |
Token ID | 262 |
Feature activation | +9.00e-02 |
Pos logit contributions | |
icked | +0.019 |
together | +0.018 |
es | +0.017 |
there | +0.017 |
on | +0.016 |
Neg logit contributions | |
hallway | -0.047 |
hotel | -0.046 |
city | -0.046 |
principal | -0.046 |
doorway | -0.045 |
Token | . |
Token ID | 13 |
Feature activation | -3.11e-03 |
Pos logit contributions | |
es | +0.058 |
icked | +0.057 |
there | +0.050 |
together | +0.049 |
able | +0.048 |
Neg logit contributions | |
hallway | -0.137 |
doorway | -0.135 |
hotel | -0.135 |
ey | -0.132 |
bus | -0.131 |
Token | He |
Token ID | 679 |
Feature activation | -4.21e-03 |
Pos logit contributions | |
es | +0.005 |
icked | +0.005 |
together | +0.005 |
able | +0.005 |
on | +0.005 |
Neg logit contributions | |
ey | -0.005 |
hotel | -0.005 |
principal | -0.005 |
museum | -0.005 |
hallway | -0.005 |
Token | digs |
Token ID | 46561 |
Feature activation | -6.49e-02 |
Pos logit contributions | |
icked | +0.007 |
together | +0.007 |
there | +0.007 |
on | +0.007 |
es | +0.007 |
Neg logit contributions | |
hallway | -0.008 |
principal | -0.008 |
orchestra | -0.008 |
city | -0.008 |
doorway | -0.007 |
Token | a |
Token ID | 257 |
Feature activation | +5.02e-02 |
Pos logit contributions | |
icked | +0.075 |
es | +0.065 |
able | +0.062 |
together | +0.056 |
iled | +0.053 |
Neg logit contributions | |
hotel | -0.142 |
hallway | -0.140 |
city | -0.138 |
ey | -0.138 |
museum | -0.136 |
Token | big |
Token ID | 1263 |
Feature activation | +5.25e-02 |
Pos logit contributions | |
principal | +0.093 |
orchestra | +0.089 |
child | +0.088 |
salon | +0.086 |
babys | +0.085 |
Neg logit contributions | |
there | -0.095 |
together | -0.094 |
on | -0.094 |
in | -0.089 |
it | -0.088 |
Token | hole |
Token ID | 7604 |
Feature activation | -9.34e-02 |
Pos logit contributions | |
principal | +0.077 |
hallway | +0.073 |
priest | +0.072 |
orchestra | +0.072 |
babys | +0.071 |
Neg logit contributions | |
there | -0.110 |
together | -0.110 |
on | -0.109 |
icked | -0.106 |
it | -0.102 |
Token | in |
Token ID | 287 |
Feature activation | -4.27e-03 |
Pos logit contributions | |
icked | +0.102 |
es | +0.094 |
Remy | +0.084 |
mmm | +0.083 |
Buddy | +0.083 |
Neg logit contributions | |
hallway | -0.196 |
doorway | -0.193 |
priest | -0.185 |
hotel | -0.185 |
bus | -0.183 |
Token | the |
Token ID | 262 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
icked | +0.005 |
es | +0.004 |
together | +0.004 |
on | +0.004 |
there | +0.004 |
Neg logit contributions | |
hallway | -0.010 |
hotel | -0.010 |
city | -0.010 |
doorway | -0.010 |
classroom | -0.010 |
Token | sand |
Token ID | 6450 |
Feature activation | -3.02e-02 |
Pos logit contributions | |
orchestra | +0.178 |
principal | +0.172 |
child | +0.167 |
priest | +0.165 |
salon | +0.163 |
Neg logit contributions | |
on | -0.214 |
there | -0.214 |
in | -0.205 |
it | -0.196 |
together | -0.196 |
Token | . |
Token ID | 13 |
Feature activation | -7.56e-03 |
Pos logit contributions | |
icked | +0.029 |
together | +0.024 |
there | +0.023 |
es | +0.023 |
on | +0.022 |
Neg logit contributions | |
hallway | -0.077 |
hotel | -0.076 |
city | -0.075 |
principal | -0.075 |
playground | -0.074 |
Token | He |
Token ID | 679 |
Feature activation | -2.28e-03 |
Pos logit contributions | |
together | +0.012 |
es | +0.012 |
icked | +0.012 |
on | +0.011 |
able | +0.011 |
Neg logit contributions | |
principal | -0.014 |
hallway | -0.013 |
hotel | -0.013 |
ey | -0.013 |
museum | -0.013 |
Token | again |
Token ID | 757 |
Feature activation | -2.07e-02 |
Pos logit contributions | |
icked | +0.001 |
es | +0.001 |
iled | +0.000 |
bled | +0.000 |
together | +0.000 |
Neg logit contributions | |
city | -0.001 |
playground | -0.001 |
hallway | -0.001 |
hotel | -0.001 |
classroom | -0.001 |
Token | so |
Token ID | 523 |
Feature activation | -2.22e-02 |
Pos logit contributions | |
icked | +0.023 |
es | +0.017 |
together | +0.016 |
there | +0.015 |
on | +0.014 |
Neg logit contributions | |
city | -0.052 |
classroom | -0.051 |
hallway | -0.051 |
playground | -0.051 |
principal | -0.050 |
Token | she |
Token ID | 673 |
Feature activation | +1.41e-02 |
Pos logit contributions | |
icked | +0.028 |
es | +0.027 |
together | +0.023 |
on | +0.022 |
there | +0.021 |
Neg logit contributions | |
priest | -0.048 |
city | -0.048 |
principal | -0.048 |
hallway | -0.047 |
classroom | -0.047 |
Token | began |
Token ID | 2540 |
Feature activation | -5.71e-02 |
Pos logit contributions | |
hallway | +0.029 |
principal | +0.029 |
priest | +0.029 |
hotel | +0.029 |
city | +0.028 |
Neg logit contributions | |
icked | -0.019 |
together | -0.019 |
es | -0.018 |
there | -0.018 |
on | -0.017 |
Token | to |
Token ID | 284 |
Feature activation | -6.56e-03 |
Pos logit contributions | |
icked | +0.041 |
es | +0.034 |
together | +0.034 |
there | +0.031 |
on | +0.028 |
Neg logit contributions | |
hallway | -0.160 |
principal | -0.158 |
city | -0.157 |
classroom | -0.157 |
doorway | -0.156 |
Token | flap |
Token ID | 37699 |
Feature activation | -9.33e-02 |
Pos logit contributions | |
there | +0.012 |
on | +0.011 |
together | +0.011 |
it | +0.011 |
in | +0.011 |
Neg logit contributions | |
principal | -0.012 |
hotel | -0.011 |
hallway | -0.011 |
doorway | -0.011 |
orchestra | -0.011 |
Token | her |
Token ID | 607 |
Feature activation | +7.55e-02 |
Pos logit contributions | |
icked | +0.110 |
es | +0.095 |
together | +0.083 |
there | +0.083 |
able | +0.076 |
Neg logit contributions | |
hallway | -0.210 |
ey | -0.209 |
doorway | -0.208 |
orchestra | -0.208 |
city | -0.207 |
Token | wings |
Token ID | 12098 |
Feature activation | -7.70e-02 |
Pos logit contributions | |
principal | +0.165 |
hotel | +0.160 |
city | +0.156 |
museum | +0.156 |
hallway | +0.156 |
Neg logit contributions | |
on | -0.111 |
there | -0.107 |
icked | -0.105 |
together | -0.103 |
in | -0.102 |
Token | very |
Token ID | 845 |
Feature activation | +2.85e-03 |
Pos logit contributions | |
icked | +0.095 |
es | +0.079 |
there | +0.067 |
together | +0.064 |
bled | +0.058 |
Neg logit contributions | |
hallway | -0.173 |
doorway | -0.169 |
orchestra | -0.167 |
playground | -0.166 |
city | -0.164 |
Token | quickly |
Token ID | 2952 |
Feature activation | -2.73e-02 |
Pos logit contributions | |
priest | +0.005 |
hallway | +0.005 |
city | +0.005 |
hotel | +0.005 |
orchestra | +0.005 |
Neg logit contributions | |
icked | -0.005 |
es | -0.005 |
there | -0.004 |
together | -0.004 |
on | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -1.46e-02 |
Pos logit contributions | |
icked | +0.028 |
es | +0.025 |
together | +0.023 |
there | +0.023 |
on | +0.019 |
Neg logit contributions | |
hallway | -0.066 |
orchestra | -0.064 |
city | -0.064 |
playground | -0.064 |
doorway | -0.064 |
Token | ! |
Token ID | 0 |
Feature activation | +3.53e-03 |
Pos logit contributions | |
icked | +0.058 |
es | +0.046 |
together | +0.045 |
bled | +0.041 |
there | +0.040 |
Neg logit contributions | |
hallway | -0.106 |
classroom | -0.106 |
city | -0.106 |
principal | -0.106 |
babys | -0.105 |
Token | She |
Token ID | 1375 |
Feature activation | +9.67e-03 |
Pos logit contributions | |
hallway | +0.008 |
principal | +0.008 |
classroom | +0.008 |
babys | +0.008 |
hotel | +0.008 |
Neg logit contributions | |
together | -0.005 |
icked | -0.005 |
on | -0.004 |
es | -0.004 |
there | -0.004 |
Token | has |
Token ID | 468 |
Feature activation | -1.15e-02 |
Pos logit contributions | |
hallway | +0.018 |
principal | +0.018 |
city | +0.017 |
hotel | +0.017 |
ey | +0.017 |
Neg logit contributions | |
icked | -0.016 |
together | -0.016 |
there | -0.016 |
on | -0.016 |
es | -0.015 |
Token | an |
Token ID | 281 |
Feature activation | +9.02e-02 |
Pos logit contributions | |
icked | +0.013 |
there | +0.013 |
together | +0.013 |
on | +0.012 |
es | +0.012 |
Neg logit contributions | |
hallway | -0.027 |
principal | -0.027 |
hotel | -0.027 |
city | -0.027 |
ey | -0.027 |
Token | open |
Token ID | 1280 |
Feature activation | +2.67e-02 |
Pos logit contributions | |
hallway | +0.117 |
principal | +0.116 |
hotel | +0.113 |
city | +0.112 |
priest | +0.108 |
Neg logit contributions | |
icked | -0.201 |
there | -0.197 |
on | -0.196 |
together | -0.191 |
as | -0.188 |
Token | hole |
Token ID | 7604 |
Feature activation | -9.31e-02 |
Pos logit contributions | |
principal | +0.032 |
hallway | +0.032 |
city | +0.030 |
orchestra | +0.030 |
hotel | +0.030 |
Neg logit contributions | |
there | -0.061 |
on | -0.061 |
together | -0.060 |
icked | -0.060 |
it | -0.057 |
Token | in |
Token ID | 287 |
Feature activation | -9.22e-04 |
Pos logit contributions | |
icked | +0.089 |
es | +0.085 |
together | +0.065 |
bed | +0.061 |
able | +0.060 |
Neg logit contributions | |
doorway | -0.218 |
hallway | -0.216 |
bus | -0.214 |
classroom | -0.213 |
hotel | -0.212 |
Token | her |
Token ID | 607 |
Feature activation | +9.59e-02 |
Pos logit contributions | |
icked | +0.001 |
together | +0.001 |
es | +0.001 |
on | +0.001 |
there | +0.001 |
Neg logit contributions | |
hallway | -0.002 |
hotel | -0.002 |
city | -0.002 |
doorway | -0.002 |
classroom | -0.002 |
Token | skin |
Token ID | 4168 |
Feature activation | -4.92e-02 |
Pos logit contributions | |
principal | +0.124 |
orchestra | +0.120 |
hallway | +0.119 |
priest | +0.118 |
hotel | +0.116 |
Neg logit contributions | |
on | -0.214 |
there | -0.213 |
together | -0.207 |
icked | -0.205 |
in | -0.202 |
Token | ! |
Token ID | 0 |
Feature activation | -9.69e-03 |
Pos logit contributions | |
icked | +0.065 |
together | +0.053 |
es | +0.050 |
able | +0.048 |
bled | +0.048 |
Neg logit contributions | |
hallway | -0.107 |
salon | -0.106 |
street | -0.106 |
playground | -0.105 |
classroom | -0.104 |
Token | Can |
Token ID | 1680 |
Feature activation | -4.16e-02 |
Pos logit contributions | |
together | +0.014 |
icked | +0.014 |
es | +0.013 |
on | +0.012 |
there | +0.012 |
Neg logit contributions | |
hallway | -0.020 |
principal | -0.020 |
classroom | -0.020 |
babys | -0.020 |
ey | -0.019 |
Token | and |
Token ID | 290 |
Feature activation | +2.60e-02 |
Pos logit contributions | |
hallway | +0.021 |
principal | +0.021 |
hotel | +0.021 |
city | +0.020 |
classroom | +0.020 |
Neg logit contributions | |
icked | -0.016 |
together | -0.015 |
there | -0.015 |
on | -0.015 |
es | -0.015 |
Token | his |
Token ID | 465 |
Feature activation | +8.18e-02 |
Pos logit contributions | |
hallway | +0.068 |
principal | +0.068 |
ey | +0.067 |
hotel | +0.067 |
city | +0.067 |
Neg logit contributions | |
together | -0.023 |
icked | -0.023 |
there | -0.023 |
on | -0.021 |
es | -0.020 |
Token | friends |
Token ID | 2460 |
Feature activation | -2.74e-02 |
Pos logit contributions | |
ey | +0.163 |
hallway | +0.159 |
orchestra | +0.157 |
Ta | +0.154 |
principal | +0.154 |
Neg logit contributions | |
on | -0.133 |
there | -0.127 |
in | -0.125 |
together | -0.124 |
icked | -0.116 |
Token | dug |
Token ID | 18735 |
Feature activation | -6.15e-02 |
Pos logit contributions | |
icked | +0.039 |
together | +0.039 |
there | +0.038 |
on | +0.037 |
es | +0.037 |
Neg logit contributions | |
hallway | -0.056 |
principal | -0.055 |
hotel | -0.055 |
city | -0.055 |
ey | -0.054 |
Token | a |
Token ID | 257 |
Feature activation | +5.01e-02 |
Pos logit contributions | |
icked | +0.075 |
es | +0.065 |
there | +0.058 |
together | +0.055 |
able | +0.054 |
Neg logit contributions | |
hotel | -0.138 |
hallway | -0.138 |
city | -0.138 |
doorway | -0.134 |
playground | -0.133 |
Token | hole |
Token ID | 7604 |
Feature activation | -9.31e-02 |
Pos logit contributions | |
principal | +0.099 |
orchestra | +0.096 |
Ta | +0.095 |
actor | +0.093 |
alphabet | +0.092 |
Neg logit contributions | |
together | -0.089 |
there | -0.087 |
on | -0.086 |
it | -0.083 |
in | -0.082 |
Token | and |
Token ID | 290 |
Feature activation | -1.94e-02 |
Pos logit contributions | |
icked | +0.106 |
es | +0.096 |
pped | +0.074 |
Remy | +0.069 |
dd | +0.065 |
Neg logit contributions | |
hallway | -0.205 |
doorway | -0.204 |
playground | -0.192 |
city | -0.191 |
hotel | -0.191 |
Token | put |
Token ID | 1234 |
Feature activation | -3.53e-02 |
Pos logit contributions | |
icked | +0.028 |
es | +0.026 |
on | +0.025 |
together | +0.025 |
there | +0.025 |
Neg logit contributions | |
hallway | -0.040 |
hotel | -0.038 |
doorway | -0.038 |
city | -0.038 |
principal | -0.038 |
Token | the |
Token ID | 262 |
Feature activation | +1.26e-01 |
Pos logit contributions | |
icked | +0.034 |
es | +0.033 |
together | +0.028 |
there | +0.027 |
able | +0.025 |
Neg logit contributions | |
hotel | -0.084 |
city | -0.084 |
hallway | -0.084 |
doorway | -0.083 |
museum | -0.081 |
Token | bottle |
Token ID | 9294 |
Feature activation | -4.76e-02 |
Pos logit contributions | |
principal | +0.211 |
orchestra | +0.207 |
ey | +0.196 |
priest | +0.196 |
classroom | +0.195 |
Neg logit contributions | |
there | -0.250 |
on | -0.245 |
in | -0.230 |
together | -0.228 |
it | -0.223 |
Token | in |
Token ID | 287 |
Feature activation | -2.73e-03 |
Pos logit contributions | |
icked | +0.063 |
es | +0.062 |
mmm | +0.054 |
spending | +0.052 |
Remy | +0.050 |
Neg logit contributions | |
doorway | -0.086 |
hallway | -0.085 |
hotel | -0.081 |
distance | -0.079 |
tant | -0.079 |
Token | had |
Token ID | 550 |
Feature activation | -2.28e-02 |
Pos logit contributions | |
icked | +0.007 |
together | +0.007 |
es | +0.007 |
there | +0.006 |
on | +0.006 |
Neg logit contributions | |
hotel | -0.013 |
city | -0.012 |
hallway | -0.012 |
principal | -0.012 |
priest | -0.012 |
Token | stripes |
Token ID | 28806 |
Feature activation | -6.63e-02 |
Pos logit contributions | |
icked | +0.034 |
together | +0.034 |
there | +0.033 |
on | +0.033 |
es | +0.032 |
Neg logit contributions | |
hallway | -0.046 |
principal | -0.045 |
hotel | -0.044 |
city | -0.044 |
ey | -0.044 |
Token | and |
Token ID | 290 |
Feature activation | -4.77e-04 |
Pos logit contributions | |
icked | +0.087 |
oh | +0.076 |
es | +0.075 |
able | +0.073 |
together | +0.073 |
Neg logit contributions | |
bus | -0.114 |
hallway | -0.114 |
playground | -0.113 |
city | -0.112 |
beach | -0.112 |
Token | spots |
Token ID | 10222 |
Feature activation | -7.63e-02 |
Pos logit contributions | |
icked | +0.001 |
together | +0.001 |
there | +0.001 |
es | +0.001 |
on | +0.001 |
Neg logit contributions | |
hallway | -0.001 |
principal | -0.001 |
hotel | -0.001 |
city | -0.001 |
ey | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | -1.08e-02 |
Pos logit contributions | |
icked | +0.092 |
es | +0.078 |
oh | +0.072 |
bed | +0.070 |
able | +0.069 |
Neg logit contributions | |
hallway | -0.144 |
bus | -0.143 |
city | -0.142 |
park | -0.141 |
playground | -0.141 |
Token | fins |
Token ID | 42080 |
Feature activation | -9.30e-02 |
Pos logit contributions | |
icked | +0.019 |
together | +0.019 |
there | +0.018 |
on | +0.018 |
es | +0.018 |
Neg logit contributions | |
hallway | -0.019 |
principal | -0.019 |
hotel | -0.018 |
city | -0.018 |
ey | -0.018 |
Token | and |
Token ID | 290 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
icked | +0.113 |
es | +0.100 |
bled | +0.093 |
bed | +0.091 |
able | +0.091 |
Neg logit contributions | |
bus | -0.170 |
hotel | -0.166 |
mirror | -0.165 |
priest | -0.165 |
playground | -0.165 |
Token | tails |
Token ID | 30514 |
Feature activation | -7.01e-02 |
Pos logit contributions | |
icked | +0.019 |
together | +0.019 |
there | +0.019 |
on | +0.018 |
es | +0.018 |
Neg logit contributions | |
hallway | -0.019 |
principal | -0.019 |
hotel | -0.018 |
city | -0.018 |
ey | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | -1.11e-02 |
Pos logit contributions | |
icked | +0.083 |
es | +0.069 |
together | +0.069 |
there | +0.062 |
bled | +0.062 |
Neg logit contributions | |
playground | -0.140 |
hallway | -0.139 |
city | -0.137 |
park | -0.135 |
beach | -0.135 |
Token | He |
Token ID | 679 |
Feature activation | +6.46e-03 |
Pos logit contributions | |
icked | +0.018 |
es | +0.017 |
together | +0.017 |
on | +0.017 |
there | +0.016 |
Neg logit contributions | |
principal | -0.021 |
classroom | -0.020 |
ey | -0.020 |
museum | -0.020 |
city | -0.020 |
Token | was |
Token ID | 373 |
Feature activation | -1.12e-02 |
Pos logit contributions | |
hallway | +0.013 |
principal | +0.013 |
city | +0.013 |
hotel | +0.013 |
ey | +0.013 |
Neg logit contributions | |
icked | -0.009 |
together | -0.009 |
there | -0.009 |
on | -0.009 |
es | -0.009 |
Token | it |
Token ID | 340 |
Feature activation | -2.45e-02 |
Pos logit contributions | |
icked | +0.062 |
together | +0.046 |
es | +0.043 |
able | +0.040 |
okay | +0.033 |
Neg logit contributions | |
hotel | -0.164 |
city | -0.160 |
tant | -0.160 |
museum | -0.160 |
ey | -0.159 |
Token | . |
Token ID | 13 |
Feature activation | -1.15e-02 |
Pos logit contributions | |
icked | +0.030 |
es | +0.027 |
apy | +0.025 |
iled | +0.022 |
mmm | +0.021 |
Neg logit contributions | |
hallway | -0.053 |
doorway | -0.051 |
hall | -0.050 |
classroom | -0.049 |
museum | -0.049 |
Token | He |
Token ID | 679 |
Feature activation | +6.21e-03 |
Pos logit contributions | |
icked | +0.020 |
es | +0.019 |
together | +0.018 |
on | +0.018 |
there | +0.017 |
Neg logit contributions | |
hallway | -0.021 |
doorway | -0.020 |
city | -0.020 |
street | -0.020 |
hotel | -0.020 |
Token | dug |
Token ID | 18735 |
Feature activation | -6.32e-02 |
Pos logit contributions | |
hallway | +0.013 |
principal | +0.013 |
hotel | +0.013 |
city | +0.013 |
ey | +0.013 |
Neg logit contributions | |
icked | -0.008 |
together | -0.008 |
there | -0.008 |
on | -0.008 |
es | -0.008 |
Token | a |
Token ID | 257 |
Feature activation | +5.13e-02 |
Pos logit contributions | |
icked | +0.070 |
es | +0.061 |
together | +0.057 |
there | +0.054 |
able | +0.050 |
Neg logit contributions | |
hallway | -0.147 |
hotel | -0.145 |
ey | -0.143 |
city | -0.143 |
doorway | -0.143 |
Token | hole |
Token ID | 7604 |
Feature activation | -9.29e-02 |
Pos logit contributions | |
principal | +0.101 |
pastor | +0.098 |
priest | +0.097 |
orchestra | +0.094 |
child | +0.094 |
Neg logit contributions | |
on | -0.091 |
there | -0.089 |
with | -0.086 |
in | -0.086 |
together | -0.085 |
Token | and |
Token ID | 290 |
Feature activation | -2.18e-02 |
Pos logit contributions | |
icked | +0.109 |
es | +0.099 |
rums | +0.081 |
OK | +0.076 |
dd | +0.075 |
Neg logit contributions | |
hallway | -0.206 |
doorway | -0.205 |
hall | -0.186 |
playground | -0.184 |
ey | -0.183 |
Token | put |
Token ID | 1234 |
Feature activation | -3.57e-02 |
Pos logit contributions | |
icked | +0.032 |
together | +0.030 |
es | +0.030 |
on | +0.029 |
there | +0.029 |
Neg logit contributions | |
hallway | -0.045 |
hotel | -0.042 |
doorway | -0.042 |
playground | -0.042 |
principal | -0.042 |
Token | the |
Token ID | 262 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
icked | +0.025 |
es | +0.022 |
together | +0.019 |
able | +0.015 |
there | +0.014 |
Neg logit contributions | |
hallway | -0.095 |
doorway | -0.094 |
ey | -0.094 |
museum | -0.094 |
hotel | -0.094 |
Token | screw |
Token ID | 9580 |
Feature activation | -3.63e-02 |
Pos logit contributions | |
priest | +0.235 |
orchestra | +0.230 |
principal | +0.226 |
hotel | +0.222 |
pastor | +0.219 |
Neg logit contributions | |
on | -0.293 |
there | -0.292 |
it | -0.271 |
in | -0.270 |
together | -0.268 |
Token | inside |
Token ID | 2641 |
Feature activation | -5.86e-02 |
Pos logit contributions | |
icked | +0.030 |
together | +0.024 |
es | +0.023 |
spending | +0.023 |
there | +0.021 |
Neg logit contributions | |
hallway | -0.088 |
doorway | -0.084 |
hall | -0.083 |
bus | -0.083 |
desk | -0.082 |
Token | . |
Token ID | 13 |
Feature activation | -5.49e-03 |
Pos logit contributions | |
icked | +0.051 |
there | +0.043 |
es | +0.042 |
together | +0.041 |
on | +0.041 |
Neg logit contributions | |
hallway | -0.114 |
hotel | -0.113 |
city | -0.113 |
classroom | -0.111 |
playground | -0.111 |
Token | Jack |
Token ID | 3619 |
Feature activation | -1.87e-03 |
Pos logit contributions | |
icked | +0.008 |
es | +0.008 |
on | +0.007 |
together | +0.007 |
there | +0.007 |
Neg logit contributions | |
hallway | -0.011 |
city | -0.011 |
classroom | -0.011 |
playground | -0.011 |
doorway | -0.011 |
Token | dug |
Token ID | 18735 |
Feature activation | -6.02e-02 |
Pos logit contributions | |
icked | +0.003 |
together | +0.003 |
there | +0.003 |
es | +0.003 |
on | +0.003 |
Neg logit contributions | |
hallway | -0.003 |
city | -0.003 |
hotel | -0.003 |
principal | -0.003 |
classroom | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
icked | +0.074 |
es | +0.066 |
together | +0.063 |
there | +0.060 |
on | +0.056 |
Neg logit contributions | |
hallway | -0.137 |
hotel | -0.135 |
ey | -0.134 |
city | -0.133 |
playground | -0.132 |
Token | big |
Token ID | 1263 |
Feature activation | +4.28e-02 |
Pos logit contributions | |
principal | +0.098 |
child | +0.092 |
orchestra | +0.090 |
opl | +0.089 |
alphabet | +0.088 |
Neg logit contributions | |
on | -0.089 |
there | -0.088 |
together | -0.086 |
in | -0.084 |
it | -0.080 |
Token | hole |
Token ID | 7604 |
Feature activation | -9.29e-02 |
Pos logit contributions | |
principal | +0.063 |
priest | +0.059 |
hallway | +0.058 |
babys | +0.058 |
orchestra | +0.058 |
Neg logit contributions | |
there | -0.090 |
together | -0.089 |
on | -0.089 |
icked | -0.085 |
it | -0.084 |
Token | and |
Token ID | 290 |
Feature activation | -1.63e-02 |
Pos logit contributions | |
icked | +0.110 |
es | +0.101 |
rums | +0.077 |
Remy | +0.075 |
OK | +0.073 |
Neg logit contributions | |
hallway | -0.213 |
doorway | -0.210 |
playground | -0.199 |
bus | -0.192 |
hotel | -0.192 |
Token | his |
Token ID | 465 |
Feature activation | +9.08e-02 |
Pos logit contributions | |
icked | +0.024 |
es | +0.023 |
together | +0.023 |
on | +0.022 |
there | +0.022 |
Neg logit contributions | |
hallway | -0.033 |
doorway | -0.032 |
hotel | -0.032 |
principal | -0.031 |
playground | -0.031 |
Token | dad |
Token ID | 9955 |
Feature activation | -2.66e-02 |
Pos logit contributions | |
orchestra | +0.160 |
Ta | +0.159 |
principal | +0.159 |
hallway | +0.157 |
salon | +0.156 |
Neg logit contributions | |
there | -0.158 |
on | -0.158 |
in | -0.150 |
together | -0.145 |
it | -0.145 |
Token | built |
Token ID | 3170 |
Feature activation | -5.07e-02 |
Pos logit contributions | |
icked | +0.042 |
together | +0.039 |
there | +0.038 |
es | +0.037 |
on | +0.037 |
Neg logit contributions | |
hallway | -0.050 |
doorway | -0.048 |
playground | -0.047 |
city | -0.047 |
hotel | -0.047 |
Token | a |
Token ID | 257 |
Feature activation | +5.29e-02 |
Pos logit contributions | |
icked | +0.059 |
es | +0.052 |
together | +0.050 |
there | +0.047 |
on | +0.045 |
Neg logit contributions | |
hallway | -0.120 |
hotel | -0.119 |
city | -0.118 |
playground | -0.118 |
doorway | -0.117 |
Token | on |
Token ID | 319 |
Feature activation | -3.97e-02 |
Pos logit contributions | |
icked | +0.027 |
es | +0.025 |
able | +0.025 |
instead | +0.018 |
iled | +0.017 |
Neg logit contributions | |
hotel | -0.094 |
city | -0.089 |
hallway | -0.089 |
mall | -0.089 |
mirror | -0.088 |
Token | their |
Token ID | 511 |
Feature activation | +9.37e-02 |
Pos logit contributions | |
icked | +0.049 |
together | +0.048 |
there | +0.047 |
on | +0.046 |
es | +0.045 |
Neg logit contributions | |
hallway | -0.090 |
principal | -0.089 |
hotel | -0.088 |
city | -0.088 |
ey | -0.088 |
Token | swim |
Token ID | 9422 |
Feature activation | -2.65e-02 |
Pos logit contributions | |
hallway | +0.138 |
principal | +0.137 |
hotel | +0.135 |
city | +0.134 |
ey | +0.133 |
Neg logit contributions | |
icked | -0.189 |
together | -0.186 |
there | -0.184 |
es | -0.181 |
on | -0.180 |
Token | suits |
Token ID | 14803 |
Feature activation | -7.18e-02 |
Pos logit contributions | |
icked | +0.058 |
together | +0.055 |
there | +0.055 |
es | +0.055 |
on | +0.053 |
Neg logit contributions | |
hallway | -0.035 |
hotel | -0.035 |
city | -0.034 |
principal | -0.033 |
classroom | -0.033 |
Token | and |
Token ID | 290 |
Feature activation | -2.28e-02 |
Pos logit contributions | |
icked | +0.106 |
es | +0.088 |
able | +0.078 |
there | +0.076 |
together | +0.076 |
Neg logit contributions | |
bus | -0.146 |
hallway | -0.145 |
hotel | -0.145 |
city | -0.143 |
street | -0.142 |
Token | hats |
Token ID | 23910 |
Feature activation | -9.28e-02 |
Pos logit contributions | |
icked | +0.043 |
together | +0.041 |
there | +0.041 |
es | +0.040 |
on | +0.039 |
Neg logit contributions | |
hallway | -0.038 |
hotel | -0.037 |
principal | -0.037 |
city | -0.036 |
classroom | -0.036 |
Token | . |
Token ID | 13 |
Feature activation | -8.74e-03 |
Pos logit contributions | |
icked | +0.122 |
es | +0.101 |
able | +0.089 |
Remy | +0.088 |
there | +0.086 |
Neg logit contributions | |
bus | -0.194 |
hotel | -0.187 |
hallway | -0.185 |
babys | -0.184 |
playground | -0.182 |
Token | They |
Token ID | 1119 |
Feature activation | -2.68e-03 |
Pos logit contributions | |
icked | +0.015 |
es | +0.014 |
together | +0.014 |
able | +0.013 |
there | +0.013 |
Neg logit contributions | |
hotel | -0.016 |
hallway | -0.016 |
city | -0.016 |
priest | -0.015 |
babys | -0.015 |
Token | get |
Token ID | 651 |
Feature activation | -2.66e-02 |
Pos logit contributions | |
there | +0.004 |
on | +0.004 |
together | +0.004 |
icked | +0.004 |
in | +0.004 |
Neg logit contributions | |
principal | -0.005 |
ey | -0.005 |
doorway | -0.005 |
hallway | -0.005 |
orchestra | -0.005 |
Token | in |
Token ID | 287 |
Feature activation | -9.19e-03 |
Pos logit contributions | |
icked | +0.033 |
es | +0.032 |
together | +0.027 |
instead | +0.026 |
there | +0.026 |
Neg logit contributions | |
hotel | -0.060 |
babys | -0.058 |
priest | -0.057 |
hallway | -0.057 |
city | -0.056 |
Token | the |
Token ID | 262 |
Feature activation | +9.38e-02 |
Pos logit contributions | |
icked | +0.009 |
together | +0.008 |
es | +0.008 |
there | +0.008 |
on | +0.008 |
Neg logit contributions | |
hallway | -0.024 |
hotel | -0.023 |
principal | -0.023 |
city | -0.023 |
classroom | -0.023 |
Token | The |
Token ID | 464 |
Feature activation | +1.29e-01 |
Pos logit contributions | |
together | +0.004 |
there | +0.004 |
es | +0.004 |
icked | +0.003 |
on | +0.003 |
Neg logit contributions | |
hotel | -0.006 |
ey | -0.005 |
hallway | -0.005 |
principal | -0.005 |
city | -0.005 |
Token | girl |
Token ID | 2576 |
Feature activation | -1.62e-03 |
Pos logit contributions | |
opl | +0.194 |
ey | +0.180 |
principal | +0.173 |
hallway | +0.165 |
terminal | +0.161 |
Neg logit contributions | |
there | -0.290 |
on | -0.280 |
in | -0.276 |
as | -0.255 |
together | -0.255 |
Token | nodded |
Token ID | 14464 |
Feature activation | -4.71e-02 |
Pos logit contributions | |
together | +0.002 |
icked | +0.002 |
there | +0.002 |
on | +0.002 |
es | +0.002 |
Neg logit contributions | |
hallway | -0.003 |
principal | -0.003 |
hotel | -0.003 |
city | -0.003 |
ey | -0.003 |
Token | and |
Token ID | 290 |
Feature activation | -1.50e-02 |
Pos logit contributions | |
icked | +0.067 |
es | +0.059 |
there | +0.053 |
together | +0.050 |
able | +0.049 |
Neg logit contributions | |
babys | -0.095 |
mirror | -0.093 |
orchestra | -0.092 |
woman | -0.091 |
priest | -0.091 |
Token | thanked |
Token ID | 26280 |
Feature activation | -3.07e-02 |
Pos logit contributions | |
icked | +0.021 |
there | +0.019 |
together | +0.019 |
es | +0.019 |
on | +0.019 |
Neg logit contributions | |
hallway | -0.031 |
principal | -0.030 |
hotel | -0.030 |
doorway | -0.030 |
city | -0.030 |
Token | the |
Token ID | 262 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
icked | +0.027 |
together | +0.026 |
there | +0.025 |
on | +0.024 |
es | +0.024 |
Neg logit contributions | |
hallway | -0.080 |
principal | -0.080 |
hotel | -0.079 |
city | -0.079 |
ey | -0.078 |
Token | old |
Token ID | 1468 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
opl | +0.184 |
principal | +0.170 |
carn | +0.166 |
ey | +0.161 |
Ta | +0.160 |
Neg logit contributions | |
there | -0.316 |
on | -0.301 |
in | -0.297 |
it | -0.292 |
with | -0.271 |
Token | lady |
Token ID | 10846 |
Feature activation | -8.47e-03 |
Pos logit contributions | |
Ta | +0.122 |
hallway | +0.120 |
ey | +0.119 |
principal | +0.116 |
hotel | +0.115 |
Neg logit contributions | |
there | -0.298 |
on | -0.291 |
it | -0.289 |
together | -0.282 |
in | -0.280 |
Token | for |
Token ID | 329 |
Feature activation | +7.93e-03 |
Pos logit contributions | |
icked | +0.007 |
there | +0.006 |
together | +0.006 |
es | +0.006 |
on | +0.006 |
Neg logit contributions | |
principal | -0.022 |
hallway | -0.022 |
hotel | -0.021 |
city | -0.021 |
classroom | -0.021 |
Token | her |
Token ID | 607 |
Feature activation | +6.21e-02 |
Pos logit contributions | |
principal | +0.017 |
hallway | +0.017 |
doorway | +0.017 |
orchestra | +0.017 |
city | +0.017 |
Neg logit contributions | |
icked | -0.010 |
es | -0.010 |
together | -0.009 |
there | -0.009 |
on | -0.009 |
Token | advice |
Token ID | 5608 |
Feature activation | -4.10e-02 |
Pos logit contributions | |
hallway | +0.108 |
principal | +0.108 |
hotel | +0.106 |
city | +0.106 |
ey | +0.105 |
Neg logit contributions | |
together | -0.109 |
icked | -0.108 |
there | -0.107 |
on | -0.106 |
es | -0.102 |
Token | him |
Token ID | 683 |
Feature activation | -3.64e-03 |
Pos logit contributions | |
together | +0.012 |
icked | +0.011 |
there | +0.011 |
on | +0.011 |
es | +0.010 |
Neg logit contributions | |
hallway | -0.039 |
principal | -0.039 |
ey | -0.038 |
hotel | -0.038 |
city | -0.038 |
Token | gently |
Token ID | 15165 |
Feature activation | -2.73e-02 |
Pos logit contributions | |
icked | +0.003 |
there | +0.003 |
together | +0.003 |
es | +0.003 |
on | +0.003 |
Neg logit contributions | |
principal | -0.009 |
hotel | -0.009 |
city | -0.009 |
classroom | -0.009 |
hallway | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | -1.07e-02 |
Pos logit contributions | |
icked | +0.026 |
together | +0.023 |
there | +0.023 |
es | +0.022 |
on | +0.021 |
Neg logit contributions | |
hallway | -0.069 |
principal | -0.069 |
hotel | -0.069 |
city | -0.068 |
classroom | -0.068 |
Token | He |
Token ID | 679 |
Feature activation | +4.82e-04 |
Pos logit contributions | |
icked | +0.013 |
together | +0.013 |
there | +0.013 |
on | +0.013 |
es | +0.012 |
Neg logit contributions | |
hallway | -0.024 |
principal | -0.024 |
city | -0.023 |
hotel | -0.023 |
classroom | -0.023 |
Token | said |
Token ID | 531 |
Feature activation | -2.52e-02 |
Pos logit contributions | |
hallway | +0.001 |
ey | +0.001 |
principal | +0.001 |
city | +0.001 |
hotel | +0.001 |
Neg logit contributions | |
icked | -0.001 |
together | -0.001 |
there | -0.001 |
on | -0.001 |
es | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
icked | +0.025 |
es | +0.022 |
together | +0.022 |
against | +0.020 |
there | +0.019 |
Neg logit contributions | |
priest | -0.062 |
principal | -0.060 |
hotel | -0.060 |
babys | -0.060 |
ostr | -0.059 |
Token | elephant |
Token ID | 20950 |
Feature activation | -2.41e-02 |
Pos logit contributions | |
salon | +0.197 |
ey | +0.193 |
hallway | +0.192 |
doorway | +0.190 |
Ta | +0.190 |
Neg logit contributions | |
there | -0.268 |
on | -0.267 |
together | -0.251 |
in | -0.249 |
it | -0.241 |
Token | 's |
Token ID | 338 |
Feature activation | +5.84e-02 |
Pos logit contributions | |
icked | +0.036 |
together | +0.035 |
there | +0.035 |
on | +0.033 |
es | +0.033 |
Neg logit contributions | |
principal | -0.046 |
hotel | -0.046 |
hallway | -0.045 |
city | -0.045 |
priest | -0.045 |
Token | name |
Token ID | 1438 |
Feature activation | -4.07e-02 |
Pos logit contributions | |
hallway | +0.084 |
principal | +0.082 |
ey | +0.080 |
orchestra | +0.080 |
doorway | +0.079 |
Neg logit contributions | |
there | -0.119 |
on | -0.119 |
icked | -0.119 |
together | -0.118 |
es | -0.113 |
Token | was |
Token ID | 373 |
Feature activation | -2.57e-03 |
Pos logit contributions | |
icked | +0.065 |
together | +0.061 |
there | +0.059 |
able | +0.057 |
bed | +0.055 |
Neg logit contributions | |
priest | -0.069 |
hotel | -0.068 |
woman | -0.068 |
city | -0.068 |
library | -0.067 |
Token | R |
Token ID | 371 |
Feature activation | +9.83e-02 |
Pos logit contributions | |
icked | +0.003 |
together | +0.003 |
there | +0.003 |
es | +0.003 |
on | +0.003 |
Neg logit contributions | |
principal | -0.005 |
hotel | -0.005 |
city | -0.005 |
hallway | -0.005 |
classroom | -0.005 |
Token |
|
Token ID | 198 |
Feature activation | -8.15e-03 |
Pos logit contributions | |
together | +0.017 |
there | +0.017 |
on | +0.016 |
icked | +0.015 |
it | +0.015 |
Neg logit contributions | |
principal | -0.034 |
ey | -0.034 |
orchestra | -0.034 |
hallway | -0.034 |
priest | -0.034 |
Token |
|
Token ID | 198 |
Feature activation | -9.15e-03 |
Pos logit contributions | |
icked | +0.010 |
together | +0.010 |
there | +0.009 |
on | +0.009 |
es | +0.009 |
Neg logit contributions | |
hallway | -0.019 |
principal | -0.019 |
hotel | -0.018 |
ey | -0.018 |
city | -0.018 |
Token | One |
Token ID | 3198 |
Feature activation | +2.63e-02 |
Pos logit contributions | |
there | +0.013 |
together | +0.013 |
on | +0.013 |
instead | +0.012 |
es | +0.012 |
Neg logit contributions | |
ey | -0.019 |
principal | -0.019 |
doorway | -0.019 |
orchestra | -0.019 |
babys | -0.018 |
Token | day |
Token ID | 1110 |
Feature activation | -3.21e-02 |
Pos logit contributions | |
principal | +0.035 |
hallway | +0.034 |
ey | +0.034 |
hotel | +0.033 |
city | +0.033 |
Neg logit contributions | |
together | -0.057 |
there | -0.057 |
icked | -0.057 |
on | -0.056 |
es | -0.055 |
Token | , |
Token ID | 11 |
Feature activation | -3.48e-02 |
Pos logit contributions | |
icked | +0.045 |
able | +0.040 |
bled | +0.036 |
together | +0.036 |
es | +0.036 |
Neg logit contributions | |
hotel | -0.063 |
babys | -0.061 |
museum | -0.060 |
city | -0.059 |
ostr | -0.059 |
Token | the |
Token ID | 262 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
icked | +0.046 |
es | +0.040 |
together | +0.036 |
able | +0.034 |
on | +0.031 |
Neg logit contributions | |
hotel | -0.078 |
city | -0.078 |
museum | -0.077 |
hallway | -0.076 |
babys | -0.075 |
Token | tiger |
Token ID | 26241 |
Feature activation | -1.30e-02 |
Pos logit contributions | |
ey | +0.240 |
salon | +0.232 |
principal | +0.232 |
babys | +0.222 |
hallway | +0.216 |
Neg logit contributions | |
on | -0.279 |
there | -0.270 |
in | -0.262 |
together | -0.250 |
with | -0.247 |
Token | heard |
Token ID | 2982 |
Feature activation | -4.62e-02 |
Pos logit contributions | |
icked | +0.021 |
together | +0.020 |
there | +0.020 |
on | +0.019 |
es | +0.019 |
Neg logit contributions | |
hallway | -0.024 |
hotel | -0.024 |
city | -0.024 |
principal | -0.024 |
classroom | -0.023 |
Token | a |
Token ID | 257 |
Feature activation | +3.27e-02 |
Pos logit contributions | |
icked | +0.039 |
together | +0.038 |
there | +0.037 |
on | +0.035 |
es | +0.035 |
Neg logit contributions | |
hallway | -0.122 |
principal | -0.122 |
hotel | -0.121 |
city | -0.120 |
ey | -0.120 |
Token | knock |
Token ID | 10643 |
Feature activation | -5.70e-02 |
Pos logit contributions | |
principal | +0.066 |
ey | +0.064 |
salon | +0.063 |
classroom | +0.062 |
hallway | +0.062 |
Neg logit contributions | |
on | -0.049 |
together | -0.048 |
there | -0.048 |
it | -0.047 |
in | -0.046 |
Token | on |
Token ID | 319 |
Feature activation | -2.54e-02 |
Pos logit contributions | |
icked | +0.029 |
together | +0.019 |
es | +0.016 |
able | +0.016 |
there | +0.014 |
Neg logit contributions | |
principal | -0.165 |
hallway | -0.164 |
city | -0.163 |
priest | -0.162 |
bus | -0.162 |
Token | was |
Token ID | 373 |
Feature activation | +4.92e-03 |
Pos logit contributions | |
principal | +0.025 |
hallway | +0.025 |
hotel | +0.025 |
city | +0.025 |
classroom | +0.024 |
Neg logit contributions | |
icked | -0.018 |
together | -0.018 |
there | -0.018 |
on | -0.017 |
es | -0.017 |
Token | really |
Token ID | 1107 |
Feature activation | +8.48e-03 |
Pos logit contributions | |
principal | +0.009 |
babys | +0.009 |
classroom | +0.009 |
hotel | +0.009 |
city | +0.009 |
Neg logit contributions | |
icked | -0.008 |
es | -0.008 |
together | -0.007 |
there | -0.007 |
on | -0.007 |
Token | nice |
Token ID | 3621 |
Feature activation | -7.01e-03 |
Pos logit contributions | |
hallway | +0.013 |
principal | +0.013 |
hotel | +0.012 |
city | +0.012 |
ey | +0.012 |
Neg logit contributions | |
icked | -0.017 |
together | -0.017 |
there | -0.016 |
on | -0.016 |
es | -0.016 |
Token | and |
Token ID | 290 |
Feature activation | -7.63e-03 |
Pos logit contributions | |
icked | +0.008 |
together | +0.006 |
es | +0.006 |
bled | +0.006 |
there | +0.006 |
Neg logit contributions | |
classroom | -0.017 |
principal | -0.017 |
hotel | -0.017 |
city | -0.017 |
hallway | -0.017 |
Token | that |
Token ID | 326 |
Feature activation | -1.69e-02 |
Pos logit contributions | |
icked | +0.011 |
es | +0.010 |
together | +0.010 |
on | +0.010 |
there | +0.010 |
Neg logit contributions | |
principal | -0.015 |
hallway | -0.015 |
city | -0.015 |
hotel | -0.015 |
classroom | -0.015 |
Token | the |
Token ID | 262 |
Feature activation | +1.48e-01 |
Pos logit contributions | |
icked | +0.020 |
together | +0.017 |
es | +0.016 |
there | +0.015 |
on | +0.015 |
Neg logit contributions | |
hotel | -0.039 |
hallway | -0.039 |
principal | -0.038 |
library | -0.038 |
city | -0.038 |
Token | cat |
Token ID | 3797 |
Feature activation | -1.47e-02 |
Pos logit contributions | |
ey | +0.266 |
priest | +0.265 |
orchestra | +0.258 |
hotel | +0.256 |
principal | +0.255 |
Neg logit contributions | |
there | -0.277 |
on | -0.271 |
in | -0.240 |
together | -0.238 |
with | -0.236 |
Token | should |
Token ID | 815 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
icked | +0.023 |
together | +0.021 |
there | +0.021 |
es | +0.020 |
on | +0.020 |
Neg logit contributions | |
principal | -0.028 |
hallway | -0.028 |
hotel | -0.027 |
city | -0.027 |
babys | -0.027 |
Token | give |
Token ID | 1577 |
Feature activation | -2.70e-02 |
Pos logit contributions | |
there | +0.015 |
icked | +0.015 |
together | +0.015 |
on | +0.015 |
it | +0.014 |
Neg logit contributions | |
hallway | -0.020 |
principal | -0.020 |
ey | -0.020 |
hotel | -0.020 |
doorway | -0.020 |
Token | them |
Token ID | 606 |
Feature activation | -3.23e-03 |
Pos logit contributions | |
icked | +0.031 |
together | +0.028 |
there | +0.028 |
es | +0.027 |
on | +0.026 |
Neg logit contributions | |
principal | -0.062 |
hotel | -0.061 |
hallway | -0.061 |
babys | -0.061 |
city | -0.060 |
Token | a |
Token ID | 257 |
Feature activation | +5.97e-02 |
Pos logit contributions | |
icked | +0.004 |
together | +0.003 |
es | +0.003 |
there | +0.003 |
on | +0.003 |
Neg logit contributions | |
hotel | -0.007 |
principal | -0.007 |
hallway | -0.007 |
city | -0.007 |
classroom | -0.007 |
Token | excited |
Token ID | 6568 |
Feature activation | -4.24e-02 |
Pos logit contributions | |
hotel | +0.011 |
priest | +0.011 |
hallway | +0.011 |
city | +0.011 |
classroom | +0.011 |
Neg logit contributions | |
icked | -0.014 |
es | -0.014 |
together | -0.013 |
on | -0.013 |
there | -0.013 |
Token | ! |
Token ID | 0 |
Feature activation | +6.11e-03 |
Pos logit contributions | |
icked | +0.063 |
es | +0.052 |
bled | +0.045 |
pped | +0.042 |
against | +0.041 |
Neg logit contributions | |
playground | -0.092 |
bus | -0.089 |
hotel | -0.089 |
classroom | -0.089 |
hallway | -0.089 |
Token | She |
Token ID | 1375 |
Feature activation | +1.24e-02 |
Pos logit contributions | |
hallway | +0.013 |
principal | +0.013 |
city | +0.013 |
hotel | +0.013 |
classroom | +0.013 |
Neg logit contributions | |
icked | -0.008 |
together | -0.008 |
there | -0.008 |
es | -0.008 |
on | -0.008 |
Token | skipped |
Token ID | 26684 |
Feature activation | -5.57e-02 |
Pos logit contributions | |
principal | +0.027 |
hallway | +0.027 |
hotel | +0.026 |
city | +0.026 |
ey | +0.026 |
Neg logit contributions | |
icked | -0.017 |
together | -0.016 |
there | -0.016 |
on | -0.016 |
es | -0.015 |
Token | along |
Token ID | 1863 |
Feature activation | -2.40e-02 |
Pos logit contributions | |
icked | +0.052 |
es | +0.047 |
together | +0.044 |
there | +0.044 |
on | +0.040 |
Neg logit contributions | |
hallway | -0.145 |
hotel | -0.142 |
city | -0.141 |
doorway | -0.141 |
classroom | -0.141 |
Token | the |
Token ID | 262 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
icked | +0.018 |
together | +0.018 |
there | +0.017 |
on | +0.016 |
es | +0.016 |
Neg logit contributions | |
hallway | -0.066 |
principal | -0.065 |
hotel | -0.065 |
city | -0.065 |
ey | -0.064 |
Token | sidewalk |
Token ID | 24058 |
Feature activation | -5.65e-02 |
Pos logit contributions | |
ey | +0.148 |
principal | +0.140 |
Ta | +0.133 |
babys | +0.131 |
tant | +0.130 |
Neg logit contributions | |
on | -0.246 |
there | -0.243 |
together | -0.232 |
it | -0.228 |
in | -0.227 |
Token | to |
Token ID | 284 |
Feature activation | +5.25e-03 |
Pos logit contributions | |
icked | +0.046 |
there | +0.041 |
together | +0.041 |
es | +0.039 |
on | +0.039 |
Neg logit contributions | |
hallway | -0.152 |
principal | -0.150 |
city | -0.149 |
hotel | -0.149 |
classroom | -0.148 |
Token | the |
Token ID | 262 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
principal | +0.014 |
hallway | +0.014 |
ey | +0.013 |
babys | +0.013 |
Ta | +0.013 |
Neg logit contributions | |
there | -0.005 |
on | -0.005 |
it | -0.005 |
together | -0.004 |
as | -0.004 |
Token | ice |
Token ID | 4771 |
Feature activation | +1.02e-02 |
Pos logit contributions | |
ey | +0.141 |
tant | +0.132 |
Ta | +0.125 |
principal | +0.121 |
orchestra | +0.119 |
Neg logit contributions | |
on | -0.249 |
there | -0.241 |
in | -0.234 |
together | -0.231 |
at | -0.227 |
Token | cream |
Token ID | 8566 |
Feature activation | -9.66e-03 |
Pos logit contributions | |
principal | +0.017 |
hallway | +0.017 |
hotel | +0.017 |
city | +0.017 |
ey | +0.017 |
Neg logit contributions | |
together | -0.019 |
icked | -0.019 |
there | -0.018 |
on | -0.018 |
es | -0.018 |
Token | hug |
Token ID | 16225 |
Feature activation | -5.33e-02 |
Pos logit contributions | |
principal | +0.042 |
ey | +0.041 |
city | +0.041 |
courtroom | +0.039 |
orchestra | +0.039 |
Neg logit contributions | |
on | -0.056 |
there | -0.054 |
together | -0.052 |
in | -0.050 |
with | -0.049 |
Token | and |
Token ID | 290 |
Feature activation | -2.52e-02 |
Pos logit contributions | |
icked | +0.059 |
there | +0.048 |
together | +0.045 |
es | +0.043 |
able | +0.040 |
Neg logit contributions | |
principal | -0.125 |
babys | -0.123 |
hallway | -0.122 |
hotel | -0.121 |
bus | -0.121 |
Token | tucked |
Token ID | 29779 |
Feature activation | -4.57e-02 |
Pos logit contributions | |
icked | +0.034 |
together | +0.033 |
there | +0.032 |
es | +0.032 |
on | +0.031 |
Neg logit contributions | |
hallway | -0.053 |
principal | -0.053 |
hotel | -0.052 |
city | -0.052 |
priest | -0.051 |
Token | him |
Token ID | 683 |
Feature activation | +1.77e-04 |
Pos logit contributions | |
icked | +0.045 |
es | +0.041 |
together | +0.038 |
there | +0.035 |
on | +0.034 |
Neg logit contributions | |
principal | -0.112 |
hotel | -0.109 |
hallway | -0.109 |
doorway | -0.109 |
priest | -0.108 |
Token | in |
Token ID | 287 |
Feature activation | -9.28e-04 |
Pos logit contributions | |
hallway | +0.000 |
hotel | +0.000 |
classroom | +0.000 |
doorway | +0.000 |
babys | +0.000 |
Neg logit contributions | |
icked | -0.000 |
es | -0.000 |
together | -0.000 |
it | -0.000 |
there | -0.000 |
Token | his |
Token ID | 465 |
Feature activation | +8.48e-02 |
Pos logit contributions | |
icked | +0.001 |
together | +0.001 |
es | +0.001 |
there | +0.001 |
on | +0.001 |
Neg logit contributions | |
hallway | -0.002 |
hotel | -0.002 |
principal | -0.002 |
city | -0.002 |
classroom | -0.002 |
Token | crib |
Token ID | 48083 |
Feature activation | -5.55e-02 |
Pos logit contributions | |
principal | +0.110 |
orchestra | +0.106 |
ey | +0.106 |
priest | +0.104 |
Ta | +0.103 |
Neg logit contributions | |
on | -0.189 |
there | -0.188 |
together | -0.186 |
in | -0.179 |
icked | -0.178 |
Token | . |
Token ID | 13 |
Feature activation | -7.94e-03 |
Pos logit contributions | |
icked | +0.048 |
es | +0.039 |
there | +0.039 |
together | +0.039 |
on | +0.036 |
Neg logit contributions | |
hallway | -0.150 |
hotel | -0.146 |
principal | -0.146 |
doorway | -0.145 |
classroom | -0.144 |
Token | She |
Token ID | 1375 |
Feature activation | -4.63e-04 |
Pos logit contributions | |
icked | +0.011 |
together | +0.010 |
es | +0.010 |
there | +0.010 |
on | +0.010 |
Neg logit contributions | |
hallway | -0.017 |
principal | -0.017 |
city | -0.017 |
classroom | -0.017 |
hotel | -0.017 |
Token | turned |
Token ID | 2900 |
Feature activation | -4.78e-02 |
Pos logit contributions | |
icked | +0.001 |
together | +0.001 |
there | +0.001 |
on | +0.001 |
es | +0.001 |
Neg logit contributions | |
hallway | -0.001 |
principal | -0.001 |
city | -0.001 |
ey | -0.001 |
hotel | -0.001 |
Token | off |
Token ID | 572 |
Feature activation | -4.69e-02 |
Pos logit contributions | |
icked | +0.044 |
es | +0.037 |
together | +0.032 |
there | +0.030 |
able | +0.028 |
Neg logit contributions | |
hotel | -0.122 |
hallway | -0.121 |
tant | -0.121 |
doorway | -0.120 |
babys | -0.120 |
Token | . |
Token ID | 13 |
Feature activation | -9.45e-03 |
Pos logit contributions | |
icked | +0.086 |
bled | +0.063 |
pped | +0.062 |
es | +0.061 |
ions | +0.060 |
Neg logit contributions | |
classroom | -0.104 |
orchestra | -0.102 |
woman | -0.102 |
city | -0.101 |
mirror | -0.101 |
Token | |
Token ID | 220 |
Feature activation | +1.26e-03 |
Pos logit contributions | |
icked | +0.014 |
together | +0.014 |
es | +0.014 |
on | +0.013 |
there | +0.013 |
Neg logit contributions | |
hallway | -0.019 |
principal | -0.019 |
classroom | -0.019 |
city | -0.019 |
doorway | -0.018 |
Token |
|
Token ID | 198 |
Feature activation | -8.53e-03 |
Pos logit contributions | |
principal | +0.002 |
hallway | +0.002 |
classroom | +0.002 |
city | +0.002 |
ey | +0.002 |
Neg logit contributions | |
icked | -0.002 |
there | -0.002 |
on | -0.002 |
together | -0.002 |
es | -0.002 |
Token |
|
Token ID | 198 |
Feature activation | -9.09e-03 |
Pos logit contributions | |
icked | +0.012 |
together | +0.012 |
there | +0.011 |
es | +0.011 |
on | +0.011 |
Neg logit contributions | |
hallway | -0.018 |
principal | -0.017 |
city | -0.017 |
hotel | -0.017 |
classroom | -0.017 |
Token | The |
Token ID | 464 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
icked | +0.014 |
together | +0.014 |
there | +0.013 |
on | +0.013 |
es | +0.013 |
Neg logit contributions | |
hallway | -0.018 |
principal | -0.018 |
city | -0.018 |
hotel | -0.018 |
ey | -0.017 |
Token | two |
Token ID | 734 |
Feature activation | +6.30e-02 |
Pos logit contributions | |
principal | +0.240 |
salon | +0.238 |
ey | +0.235 |
courtroom | +0.233 |
priest | +0.230 |
Neg logit contributions | |
there | -0.288 |
on | -0.272 |
in | -0.260 |
with | -0.243 |
as | -0.237 |
Token | of |
Token ID | 286 |
Feature activation | +5.30e-02 |
Pos logit contributions | |
hallway | +0.110 |
principal | +0.110 |
hotel | +0.107 |
city | +0.106 |
ey | +0.106 |
Neg logit contributions | |
icked | -0.111 |
together | -0.110 |
there | -0.109 |
on | -0.107 |
es | -0.106 |
Token | them |
Token ID | 606 |
Feature activation | +5.83e-03 |
Pos logit contributions | |
hallway | +0.144 |
principal | +0.143 |
hotel | +0.140 |
city | +0.139 |
doorway | +0.139 |
Neg logit contributions | |
icked | -0.045 |
there | -0.045 |
together | -0.044 |
on | -0.042 |
es | -0.039 |
Token | were |
Token ID | 547 |
Feature activation | -8.46e-04 |
Pos logit contributions | |
hallway | +0.012 |
principal | +0.011 |
hotel | +0.011 |
city | +0.011 |
classroom | +0.011 |
Neg logit contributions | |
icked | -0.009 |
together | -0.008 |
there | -0.008 |
es | -0.008 |
on | -0.008 |
Token | suddenly |
Token ID | 6451 |
Feature activation | -2.64e-02 |
Pos logit contributions | |
icked | +0.001 |
es | +0.001 |
together | +0.001 |
there | +0.001 |
on | +0.001 |
Neg logit contributions | |
city | -0.002 |
principal | -0.002 |
classroom | -0.002 |
hallway | -0.002 |
hotel | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
icked | +0.039 |
together | +0.035 |
es | +0.034 |
there | +0.034 |
on | +0.033 |
Neg logit contributions | |
hallway | -0.055 |
city | -0.054 |
classroom | -0.054 |
hotel | -0.053 |
principal | -0.053 |
Token |
|
Token ID | 198 |
Feature activation | -5.97e-03 |
Pos logit contributions | |
icked | +0.003 |
together | +0.003 |
there | +0.003 |
on | +0.003 |
es | +0.003 |
Neg logit contributions | |
hallway | -0.006 |
principal | -0.006 |
hotel | -0.006 |
city | -0.006 |
ey | -0.006 |
Token | Her |
Token ID | 9360 |
Feature activation | +6.82e-02 |
Pos logit contributions | |
icked | +0.009 |
together | +0.008 |
there | +0.008 |
on | +0.008 |
es | +0.008 |
Neg logit contributions | |
hallway | -0.012 |
principal | -0.012 |
city | -0.012 |
hotel | -0.012 |
classroom | -0.012 |
Token | mom |
Token ID | 1995 |
Feature activation | -8.99e-03 |
Pos logit contributions | |
hallway | +0.132 |
orchestra | +0.128 |
ey | +0.126 |
city | +0.126 |
hotel | +0.125 |
Neg logit contributions | |
on | -0.115 |
there | -0.115 |
in | -0.109 |
together | -0.106 |
at | -0.103 |
Token | my |
Token ID | 1820 |
Feature activation | -1.67e-02 |
Pos logit contributions | |
icked | +0.015 |
together | +0.015 |
there | +0.014 |
es | +0.014 |
on | +0.014 |
Neg logit contributions | |
hallway | -0.016 |
principal | -0.016 |
doorway | -0.015 |
city | -0.015 |
babys | -0.015 |
Token | saw |
Token ID | 2497 |
Feature activation | -3.74e-02 |
Pos logit contributions | |
icked | +0.027 |
together | +0.025 |
there | +0.024 |
es | +0.024 |
on | +0.024 |
Neg logit contributions | |
principal | -0.031 |
hallway | -0.031 |
babys | -0.031 |
doorway | -0.030 |
city | -0.030 |
Token | her |
Token ID | 607 |
Feature activation | +4.67e-02 |
Pos logit contributions | |
icked | +0.046 |
together | +0.041 |
es | +0.041 |
against | +0.039 |
on | +0.037 |
Neg logit contributions | |
ey | -0.083 |
principal | -0.078 |
mirror | -0.078 |
babys | -0.077 |
classroom | -0.076 |
Token | ankle |
Token ID | 19847 |
Feature activation | -6.01e-02 |
Pos logit contributions | |
hallway | +0.094 |
principal | +0.093 |
hotel | +0.092 |
city | +0.092 |
ey | +0.091 |
Neg logit contributions | |
icked | -0.069 |
together | -0.069 |
there | -0.069 |
on | -0.068 |
es | -0.066 |
Token | and |
Token ID | 290 |
Feature activation | -1.53e-02 |
Pos logit contributions | |
icked | +0.081 |
together | +0.069 |
es | +0.066 |
there | +0.065 |
instead | +0.056 |
Neg logit contributions | |
hallway | -0.136 |
salon | -0.128 |
doorway | -0.128 |
street | -0.127 |
playground | -0.126 |
Token | said |
Token ID | 531 |
Feature activation | -3.66e-02 |
Pos logit contributions | |
icked | +0.025 |
together | +0.023 |
es | +0.023 |
on | +0.023 |
there | +0.022 |
Neg logit contributions | |
babys | -0.027 |
principal | -0.027 |
doorway | -0.027 |
hallway | -0.027 |
priest | -0.026 |
Token | , |
Token ID | 11 |
Feature activation | -3.66e-02 |
Pos logit contributions | |
icked | +0.040 |
together | +0.036 |
against | +0.034 |
able | +0.034 |
pped | +0.033 |
Neg logit contributions | |
babys | -0.081 |
principal | -0.077 |
priest | -0.076 |
mirror | -0.074 |
woman | -0.073 |
Token | " |
Token ID | 366 |
Feature activation | +8.45e-03 |
Pos logit contributions | |
icked | +0.060 |
es | +0.054 |
together | +0.050 |
there | +0.049 |
pped | +0.048 |
Neg logit contributions | |
babys | -0.068 |
hallway | -0.065 |
woman | -0.065 |
principal | -0.064 |
doorway | -0.063 |
Token | to |
Token ID | 284 |
Feature activation | +8.51e-03 |
Pos logit contributions | |
icked | +0.027 |
es | +0.025 |
bled | +0.022 |
ged | +0.021 |
pped | +0.020 |
Neg logit contributions | |
museum | -0.053 |
hotel | -0.052 |
beach | -0.051 |
city | -0.051 |
audience | -0.051 |
Token | fly |
Token ID | 6129 |
Feature activation | -5.51e-02 |
Pos logit contributions | |
city | +0.012 |
playground | +0.012 |
priest | +0.012 |
classroom | +0.012 |
hotel | +0.012 |
Neg logit contributions | |
icked | -0.018 |
es | -0.017 |
together | -0.017 |
there | -0.016 |
on | -0.015 |
Token | ." |
Token ID | 526 |
Feature activation | -3.57e-03 |
Pos logit contributions | |
icked | +0.069 |
es | +0.049 |
able | +0.049 |
bled | +0.044 |
pped | +0.042 |
Neg logit contributions | |
city | -0.125 |
hotel | -0.124 |
classroom | -0.122 |
museum | -0.121 |
zoo | -0.120 |
Token | He |
Token ID | 679 |
Feature activation | +7.48e-03 |
Pos logit contributions | |
icked | +0.005 |
es | +0.005 |
there | +0.005 |
on | +0.005 |
together | +0.005 |
Neg logit contributions | |
principal | -0.007 |
hallway | -0.007 |
classroom | -0.007 |
city | -0.007 |
ey | -0.007 |
Token | took |
Token ID | 1718 |
Feature activation | -2.23e-02 |
Pos logit contributions | |
hallway | +0.015 |
principal | +0.015 |
hotel | +0.014 |
city | +0.014 |
priest | +0.014 |
Neg logit contributions | |
icked | -0.011 |
together | -0.010 |
there | -0.010 |
on | -0.010 |
es | -0.010 |
Token | a |
Token ID | 257 |
Feature activation | +5.74e-02 |
Pos logit contributions | |
icked | +0.020 |
together | +0.019 |
there | +0.019 |
on | +0.018 |
es | +0.018 |
Neg logit contributions | |
hallway | -0.059 |
principal | -0.058 |
hotel | -0.058 |
city | -0.058 |
ey | -0.057 |
Token | sip |
Token ID | 31145 |
Feature activation | -3.48e-02 |
Pos logit contributions | |
principal | +0.105 |
city | +0.097 |
hallway | +0.096 |
hotel | +0.096 |
orchestra | +0.095 |
Neg logit contributions | |
on | -0.107 |
there | -0.106 |
together | -0.104 |
it | -0.098 |
in | -0.097 |
Token | of |
Token ID | 286 |
Feature activation | +1.58e-02 |
Pos logit contributions | |
icked | +0.052 |
es | +0.047 |
there | +0.046 |
able | +0.046 |
together | +0.043 |
Neg logit contributions | |
ey | -0.065 |
mirror | -0.063 |
tant | -0.063 |
hotel | -0.062 |
hallway | -0.061 |
Token | the |
Token ID | 262 |
Feature activation | +1.40e-01 |
Pos logit contributions | |
hallway | +0.041 |
principal | +0.041 |
city | +0.041 |
hotel | +0.041 |
classroom | +0.040 |
Neg logit contributions | |
icked | -0.014 |
together | -0.014 |
there | -0.013 |
on | -0.013 |
es | -0.012 |
Token | soup |
Token ID | 17141 |
Feature activation | -3.18e-02 |
Pos logit contributions | |
principal | +0.264 |
hallway | +0.254 |
city | +0.250 |
orchestra | +0.249 |
classroom | +0.248 |
Neg logit contributions | |
there | -0.239 |
on | -0.237 |
together | -0.236 |
icked | -0.231 |
es | -0.222 |
Token | . |
Token ID | 13 |
Feature activation | -4.98e-04 |
Pos logit contributions | |
icked | +0.042 |
there | +0.037 |
es | +0.034 |
pped | +0.033 |
together | +0.031 |
Neg logit contributions | |
hallway | -0.067 |
doorway | -0.065 |
hotel | -0.065 |
ey | -0.064 |
playground | -0.064 |
Token | 't |
Token ID | 470 |
Feature activation | +2.90e-03 |
Pos logit contributions | |
icked | +0.027 |
together | +0.025 |
there | +0.025 |
es | +0.024 |
on | +0.023 |
Neg logit contributions | |
hallway | -0.040 |
principal | -0.040 |
city | -0.039 |
hotel | -0.039 |
babys | -0.039 |
Token | because |
Token ID | 780 |
Feature activation | -4.52e-03 |
Pos logit contributions | |
hallway | +0.005 |
principal | +0.005 |
hotel | +0.005 |
city | +0.005 |
classroom | +0.005 |
Neg logit contributions | |
icked | -0.005 |
together | -0.005 |
there | -0.005 |
es | -0.005 |
on | -0.005 |
Token | they |
Token ID | 484 |
Feature activation | +1.97e-02 |
Pos logit contributions | |
icked | +0.003 |
together | +0.003 |
es | +0.003 |
there | +0.002 |
on | +0.002 |
Neg logit contributions | |
hallway | -0.013 |
city | -0.013 |
hotel | -0.013 |
principal | -0.013 |
classroom | -0.013 |
Token | already |
Token ID | 1541 |
Feature activation | -3.40e-03 |
Pos logit contributions | |
hallway | +0.040 |
principal | +0.039 |
hotel | +0.039 |
city | +0.039 |
classroom | +0.038 |
Neg logit contributions | |
icked | -0.029 |
together | -0.028 |
there | -0.027 |
es | -0.027 |
on | -0.027 |
Token | had |
Token ID | 550 |
Feature activation | -1.61e-02 |
Pos logit contributions | |
icked | +0.005 |
together | +0.005 |
es | +0.005 |
there | +0.005 |
on | +0.005 |
Neg logit contributions | |
hallway | -0.007 |
hotel | -0.006 |
principal | -0.006 |
city | -0.006 |
classroom | -0.006 |
Token | a |
Token ID | 257 |
Feature activation | +7.37e-02 |
Pos logit contributions | |
icked | +0.020 |
es | +0.019 |
together | +0.018 |
there | +0.017 |
on | +0.017 |
Neg logit contributions | |
hotel | -0.036 |
principal | -0.035 |
city | -0.035 |
hallway | -0.035 |
classroom | -0.035 |
Token | cat |
Token ID | 3797 |
Feature activation | -1.45e-02 |
Pos logit contributions | |
ey | +0.136 |
Ta | +0.133 |
principal | +0.124 |
hallway | +0.119 |
orchestra | +0.117 |
Neg logit contributions | |
there | -0.142 |
on | -0.139 |
it | -0.135 |
together | -0.134 |
with | -0.126 |
Token | . |
Token ID | 13 |
Feature activation | -8.60e-03 |
Pos logit contributions | |
icked | +0.016 |
together | +0.014 |
es | +0.013 |
there | +0.013 |
on | +0.012 |
Neg logit contributions | |
principal | -0.035 |
hallway | -0.035 |
hotel | -0.035 |
city | -0.034 |
classroom | -0.034 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.51e-02 |
Pos logit contributions | |
icked | +0.012 |
es | +0.011 |
there | +0.011 |
on | +0.011 |
together | +0.011 |
Neg logit contributions | |
hallway | -0.018 |
principal | -0.018 |
city | -0.018 |
classroom | -0.018 |
hotel | -0.018 |
Token | was |
Token ID | 373 |
Feature activation | +4.32e-03 |
Pos logit contributions | |
hallway | +0.032 |
principal | +0.031 |
hotel | +0.031 |
city | +0.031 |
ey | +0.031 |
Neg logit contributions | |
icked | -0.021 |
together | -0.021 |
there | -0.021 |
on | -0.020 |
es | -0.020 |
Token | sad |
Token ID | 6507 |
Feature activation | -1.75e-02 |
Pos logit contributions | |
hallway | +0.006 |
principal | +0.006 |
hotel | +0.006 |
city | +0.006 |
ey | +0.006 |
Neg logit contributions | |
icked | -0.009 |
together | -0.009 |
there | -0.008 |
on | -0.008 |
es | -0.008 |
Token | not |
Token ID | 407 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
icked | +0.015 |
es | +0.012 |
together | +0.011 |
there | +0.010 |
on | +0.010 |
Neg logit contributions | |
principal | -0.055 |
hotel | -0.054 |
city | -0.054 |
classroom | -0.053 |
hallway | -0.053 |
Token | to |
Token ID | 284 |
Feature activation | +5.27e-03 |
Pos logit contributions | |
classroom | +0.026 |
street | +0.026 |
hallway | +0.026 |
salon | +0.026 |
hotel | +0.025 |
Neg logit contributions | |
icked | -0.011 |
es | -0.010 |
together | -0.007 |
bled | -0.007 |
bed | -0.007 |
Token | spin |
Token ID | 7906 |
Feature activation | -4.58e-02 |
Pos logit contributions | |
hallway | +0.009 |
principal | +0.009 |
hotel | +0.009 |
city | +0.009 |
classroom | +0.009 |
Neg logit contributions | |
icked | -0.009 |
together | -0.009 |
there | -0.009 |
es | -0.009 |
on | -0.009 |
Token | with |
Token ID | 351 |
Feature activation | -6.67e-03 |
Pos logit contributions | |
icked | +0.054 |
es | +0.040 |
there | +0.037 |
able | +0.036 |
together | +0.034 |
Neg logit contributions | |
classroom | -0.107 |
tant | -0.107 |
city | -0.107 |
hotel | -0.107 |
hallway | -0.107 |
Token | candles |
Token ID | 32268 |
Feature activation | -5.47e-02 |
Pos logit contributions | |
together | +0.008 |
icked | +0.008 |
there | +0.008 |
on | +0.008 |
es | +0.008 |
Neg logit contributions | |
hallway | -0.015 |
principal | -0.015 |
hotel | -0.015 |
city | -0.015 |
ey | -0.015 |
Token | or |
Token ID | 393 |
Feature activation | -3.14e-03 |
Pos logit contributions | |
icked | +0.064 |
es | +0.048 |
there | +0.045 |
bed | +0.044 |
able | +0.043 |
Neg logit contributions | |
hotel | -0.121 |
city | -0.117 |
salon | -0.116 |
classroom | -0.115 |
airport | -0.114 |
Token | other |
Token ID | 584 |
Feature activation | +7.44e-02 |
Pos logit contributions | |
together | +0.005 |
there | +0.005 |
on | +0.005 |
icked | +0.005 |
es | +0.005 |
Neg logit contributions | |
hallway | -0.006 |
principal | -0.006 |
hotel | -0.006 |
doorway | -0.006 |
ey | -0.006 |
Token | things |
Token ID | 1243 |
Feature activation | -2.00e-02 |
Pos logit contributions | |
principal | +0.117 |
hallway | +0.116 |
doorway | +0.112 |
ey | +0.111 |
salon | +0.109 |
Neg logit contributions | |
together | -0.150 |
there | -0.149 |
on | -0.147 |
es | -0.140 |
icked | -0.139 |
Token | that |
Token ID | 326 |
Feature activation | +1.83e-02 |
Pos logit contributions | |
icked | +0.023 |
es | +0.019 |
bed | +0.016 |
bled | +0.014 |
there | +0.014 |
Neg logit contributions | |
salon | -0.044 |
street | -0.043 |
hotel | -0.043 |
city | -0.042 |
bus | -0.042 |
Token | could |
Token ID | 714 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
hallway | +0.040 |
principal | +0.040 |
hotel | +0.039 |
city | +0.039 |
ey | +0.039 |
Neg logit contributions | |
together | -0.024 |
icked | -0.024 |
there | -0.023 |
on | -0.023 |
es | -0.022 |
Token | break |
Token ID | 2270 |
Feature activation | -3.99e-02 |
Pos logit contributions | |
together | +0.020 |
icked | +0.020 |
there | +0.020 |
on | +0.019 |
es | +0.019 |
Neg logit contributions | |
hallway | -0.024 |
principal | -0.024 |
hotel | -0.024 |
ey | -0.024 |
city | -0.024 |
Token | and |
Token ID | 290 |
Feature activation | -5.27e-03 |
Pos logit contributions | |
icked | +0.079 |
ppers | +0.068 |
bled | +0.067 |
pped | +0.066 |
fed | +0.065 |
Neg logit contributions | |
bus | -0.068 |
orchestra | -0.068 |
classroom | -0.066 |
mirror | -0.066 |
woman | -0.066 |
Token | says |
Token ID | 1139 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
icked | +0.011 |
together | +0.011 |
there | +0.011 |
es | +0.010 |
on | +0.010 |
Neg logit contributions | |
hallway | -0.008 |
principal | -0.007 |
hotel | -0.007 |
ey | -0.007 |
city | -0.007 |
Token | , |
Token ID | 11 |
Feature activation | -2.30e-02 |
Pos logit contributions | |
pped | +0.020 |
able | +0.019 |
bed | +0.018 |
against | +0.018 |
bing | +0.017 |
Neg logit contributions | |
priest | -0.033 |
mirror | -0.033 |
principal | -0.032 |
hotel | -0.032 |
museum | -0.032 |
Token | " |
Token ID | 366 |
Feature activation | +1.29e-02 |
Pos logit contributions | |
icked | +0.037 |
es | +0.034 |
together | +0.032 |
able | +0.031 |
there | +0.031 |
Neg logit contributions | |
principal | -0.042 |
hallway | -0.042 |
babys | -0.041 |
city | -0.040 |
bus | -0.040 |
Token | Ha |
Token ID | 23303 |
Feature activation | +1.01e-02 |
Pos logit contributions | |
hallway | +0.025 |
hotel | +0.025 |
principal | +0.025 |
city | +0.024 |
doorway | +0.024 |
Neg logit contributions | |
together | -0.020 |
icked | -0.020 |
there | -0.020 |
on | -0.020 |
es | -0.019 |
Token | ha |
Token ID | 387 |
Feature activation | -1.21e-02 |
Pos logit contributions | |
audience | +0.016 |
hotel | +0.015 |
city | +0.015 |
principal | +0.015 |
classroom | +0.015 |
Neg logit contributions | |
icked | -0.019 |
together | -0.018 |
es | -0.018 |
there | -0.017 |
against | -0.016 |
Token | , |
Token ID | 11 |
Feature activation | -2.58e-03 |
Pos logit contributions | |
icked | +0.014 |
together | +0.013 |
there | +0.013 |
on | +0.012 |
es | +0.012 |
Neg logit contributions | |
principal | -0.028 |
salon | -0.027 |
hotel | -0.027 |
classroom | -0.027 |
babys | -0.027 |
Token | now |
Token ID | 783 |
Feature activation | -2.37e-02 |
Pos logit contributions | |
icked | +0.004 |
together | +0.004 |
there | +0.003 |
es | +0.003 |
on | +0.003 |
Neg logit contributions | |
principal | -0.005 |
hallway | -0.005 |
hotel | -0.005 |
city | -0.005 |
ey | -0.005 |
Token | it |
Token ID | 340 |
Feature activation | -2.17e-03 |
Pos logit contributions | |
icked | +0.038 |
bled | +0.033 |
es | +0.030 |
bed | +0.030 |
together | +0.029 |
Neg logit contributions | |
classroom | -0.044 |
hotel | -0.044 |
library | -0.043 |
museum | -0.043 |
teacher | -0.042 |
Token | is |
Token ID | 318 |
Feature activation | -4.11e-03 |
Pos logit contributions | |
icked | +0.004 |
together | +0.004 |
there | +0.004 |
on | +0.004 |
es | +0.003 |
Neg logit contributions | |
hallway | -0.004 |
principal | -0.004 |
hotel | -0.004 |
city | -0.004 |
classroom | -0.004 |
Token | mine |
Token ID | 6164 |
Feature activation | -5.43e-02 |
Pos logit contributions | |
es | +0.008 |
icked | +0.008 |
together | +0.007 |
bed | +0.007 |
iled | +0.007 |
Neg logit contributions | |
classroom | -0.006 |
hotel | -0.006 |
playground | -0.006 |
city | -0.006 |
principal | -0.006 |
Token | L |
Token ID | 406 |
Feature activation | +8.10e-02 |
Pos logit contributions | |
icked | +0.024 |
together | +0.024 |
es | +0.023 |
on | +0.022 |
there | +0.021 |
Neg logit contributions | |
city | -0.042 |
museum | -0.041 |
hotel | -0.041 |
principal | -0.041 |
ostr | -0.041 |
Token | izz |
Token ID | 6457 |
Feature activation | +2.53e-02 |
Pos logit contributions | |
hallway | +0.132 |
city | +0.129 |
principal | +0.129 |
ey | +0.127 |
orchestra | +0.127 |
Neg logit contributions | |
icked | -0.154 |
there | -0.152 |
together | -0.150 |
on | -0.148 |
es | -0.148 |
Token | ie |
Token ID | 494 |
Feature activation | -9.54e-03 |
Pos logit contributions | |
hallway | +0.048 |
principal | +0.047 |
hotel | +0.047 |
city | +0.047 |
priest | +0.046 |
Neg logit contributions | |
icked | -0.039 |
together | -0.039 |
there | -0.038 |
on | -0.037 |
es | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | +7.30e-03 |
Pos logit contributions | |
icked | +0.009 |
together | +0.008 |
there | +0.008 |
es | +0.007 |
on | +0.007 |
Neg logit contributions | |
hallway | -0.024 |
city | -0.024 |
hotel | -0.024 |
principal | -0.024 |
priest | -0.024 |
Token | She |
Token ID | 1375 |
Feature activation | +2.54e-02 |
Pos logit contributions | |
hallway | +0.015 |
principal | +0.015 |
hotel | +0.015 |
city | +0.015 |
ey | +0.015 |
Neg logit contributions | |
icked | -0.010 |
together | -0.010 |
there | -0.010 |
es | -0.010 |
on | -0.010 |
Token | was |
Token ID | 373 |
Feature activation | +2.14e-03 |
Pos logit contributions | |
hallway | +0.044 |
hotel | +0.043 |
city | +0.043 |
principal | +0.042 |
priest | +0.042 |
Neg logit contributions | |
icked | -0.044 |
together | -0.043 |
there | -0.041 |
es | -0.041 |
on | -0.040 |
Token | sitting |
Token ID | 5586 |
Feature activation | -3.64e-02 |
Pos logit contributions | |
hallway | +0.004 |
principal | +0.004 |
hotel | +0.004 |
city | +0.004 |
ey | +0.004 |
Neg logit contributions | |
icked | -0.003 |
together | -0.003 |
there | -0.003 |
es | -0.003 |
on | -0.003 |
Token | on |
Token ID | 319 |
Feature activation | +8.18e-03 |
Pos logit contributions | |
icked | +0.025 |
es | +0.019 |
together | +0.015 |
it | +0.013 |
able | +0.012 |
Neg logit contributions | |
city | -0.101 |
hallway | -0.100 |
classroom | -0.099 |
hotel | -0.099 |
doorway | -0.098 |
Token | a |
Token ID | 257 |
Feature activation | +3.82e-02 |
Pos logit contributions | |
principal | +0.025 |
hallway | +0.024 |
orchestra | +0.024 |
ey | +0.024 |
priest | +0.024 |
Neg logit contributions | |
there | -0.005 |
on | -0.005 |
it | -0.004 |
together | -0.004 |
in | -0.004 |
Token | branch |
Token ID | 8478 |
Feature activation | -5.53e-02 |
Pos logit contributions | |
principal | +0.067 |
ey | +0.063 |
alphabet | +0.061 |
Ta | +0.060 |
babys | +0.060 |
Neg logit contributions | |
there | -0.070 |
on | -0.070 |
in | -0.069 |
it | -0.068 |
that | -0.067 |
Token | high |
Token ID | 1029 |
Feature activation | -1.57e-02 |
Pos logit contributions | |
icked | +0.049 |
es | +0.042 |
together | +0.041 |
there | +0.037 |
it | +0.033 |
Neg logit contributions | |
hallway | -0.149 |
doorway | -0.144 |
city | -0.144 |
classroom | -0.142 |
hotel | -0.142 |
Token | boy |
Token ID | 2933 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
principal | +0.055 |
hallway | +0.052 |
ey | +0.051 |
doorway | +0.050 |
alphabet | +0.049 |
Neg logit contributions | |
there | -0.099 |
on | -0.097 |
together | -0.096 |
in | -0.094 |
it | -0.091 |
Token | named |
Token ID | 3706 |
Feature activation | -4.13e-04 |
Pos logit contributions | |
icked | +0.027 |
together | +0.025 |
es | +0.024 |
there | +0.024 |
on | +0.023 |
Neg logit contributions | |
hallway | -0.028 |
city | -0.028 |
principal | -0.028 |
classroom | -0.028 |
hotel | -0.027 |
Token | Tim |
Token ID | 5045 |
Feature activation | +2.33e-02 |
Pos logit contributions | |
icked | +0.000 |
es | +0.000 |
together | +0.000 |
on | +0.000 |
against | +0.000 |
Neg logit contributions | |
museum | -0.001 |
hotel | -0.001 |
classroom | -0.001 |
principal | -0.001 |
hallway | -0.001 |
Token | my |
Token ID | 1820 |
Feature activation | -1.61e-02 |
Pos logit contributions | |
hallway | +0.035 |
principal | +0.035 |
hotel | +0.034 |
city | +0.034 |
ey | +0.034 |
Neg logit contributions | |
icked | -0.046 |
together | -0.046 |
there | -0.046 |
on | -0.045 |
es | -0.044 |
Token | . |
Token ID | 13 |
Feature activation | +1.04e-02 |
Pos logit contributions | |
icked | +0.011 |
together | +0.011 |
there | +0.010 |
es | +0.010 |
on | +0.010 |
Neg logit contributions | |
hallway | -0.045 |
principal | -0.045 |
city | -0.044 |
hotel | -0.044 |
ey | -0.044 |
Token | Tim |
Token ID | 5045 |
Feature activation | +3.48e-02 |
Pos logit contributions | |
city | +0.024 |
hallway | +0.024 |
classroom | +0.024 |
museum | +0.024 |
playground | +0.023 |
Neg logit contributions | |
es | -0.014 |
icked | -0.014 |
together | -0.012 |
able | -0.012 |
on | -0.011 |
Token | my |
Token ID | 1820 |
Feature activation | +1.60e-03 |
Pos logit contributions | |
principal | +0.052 |
hallway | +0.052 |
hotel | +0.051 |
city | +0.050 |
orchestra | +0.050 |
Neg logit contributions | |
icked | -0.069 |
there | -0.069 |
together | -0.069 |
on | -0.068 |
es | -0.066 |
Token | had |
Token ID | 550 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
hallway | +0.003 |
principal | +0.003 |
hotel | +0.003 |
city | +0.003 |
ey | +0.003 |
Neg logit contributions | |
icked | -0.003 |
together | -0.003 |
there | -0.003 |
on | -0.003 |
es | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | +4.19e-02 |
Pos logit contributions | |
there | +0.009 |
together | +0.009 |
icked | +0.009 |
on | +0.009 |
es | +0.008 |
Neg logit contributions | |
hallway | -0.032 |
principal | -0.032 |
hotel | -0.031 |
ey | -0.031 |
city | -0.031 |
Token | hammer |
Token ID | 15554 |
Feature activation | -5.46e-02 |
Pos logit contributions | |
orchestra | +0.081 |
opl | +0.079 |
Ta | +0.079 |
ey | +0.079 |
principal | +0.077 |
Neg logit contributions | |
on | -0.071 |
there | -0.069 |
in | -0.067 |
together | -0.065 |
to | -0.063 |
Token | that |
Token ID | 326 |
Feature activation | -3.17e-03 |
Pos logit contributions | |
icked | +0.036 |
together | +0.033 |
there | +0.031 |
es | +0.030 |
on | +0.027 |
Neg logit contributions | |
hallway | -0.155 |
principal | -0.154 |
city | -0.152 |
hotel | -0.152 |
classroom | -0.151 |
Token | ." |
Token ID | 526 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
icked | +0.008 |
es | +0.007 |
able | +0.006 |
pped | +0.006 |
there | +0.006 |
Neg logit contributions | |
city | -0.014 |
hotel | -0.013 |
universe | -0.013 |
hallway | -0.013 |
playground | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -4.34e-03 |
Pos logit contributions | |
icked | +0.017 |
together | +0.016 |
es | +0.015 |
there | +0.015 |
on | +0.015 |
Neg logit contributions | |
principal | -0.021 |
hallway | -0.021 |
city | -0.021 |
classroom | -0.021 |
ey | -0.021 |
Token |
|
Token ID | 198 |
Feature activation | -4.05e-03 |
Pos logit contributions | |
icked | +0.006 |
together | +0.006 |
there | +0.006 |
es | +0.006 |
on | +0.006 |
Neg logit contributions | |
hallway | -0.009 |
principal | -0.009 |
city | -0.009 |
hotel | -0.009 |
classroom | -0.009 |
Token | Tom |
Token ID | 13787 |
Feature activation | +1.29e-02 |
Pos logit contributions | |
icked | +0.007 |
together | +0.007 |
there | +0.007 |
on | +0.007 |
es | +0.007 |
Neg logit contributions | |
hallway | -0.007 |
principal | -0.007 |
city | -0.007 |
classroom | -0.007 |
hotel | -0.007 |
Token | was |
Token ID | 373 |
Feature activation | +1.13e-03 |
Pos logit contributions | |
hallway | +0.026 |
principal | +0.026 |
city | +0.026 |
classroom | +0.026 |
hotel | +0.026 |
Neg logit contributions | |
icked | -0.018 |
together | -0.018 |
there | -0.018 |
on | -0.017 |
es | -0.017 |
Token | so |
Token ID | 523 |
Feature activation | +5.45e-03 |
Pos logit contributions | |
principal | +0.002 |
hallway | +0.002 |
classroom | +0.002 |
city | +0.002 |
orchestra | +0.002 |
Neg logit contributions | |
icked | -0.002 |
es | -0.002 |
together | -0.002 |
there | -0.002 |
on | -0.002 |
Token | fascinated |
Token ID | 30103 |
Feature activation | -3.26e-02 |
Pos logit contributions | |
hallway | +0.009 |
city | +0.009 |
principal | +0.009 |
priest | +0.009 |
hotel | +0.009 |
Neg logit contributions | |
icked | -0.010 |
es | -0.010 |
together | -0.010 |
there | -0.010 |
on | -0.009 |
Token | by |
Token ID | 416 |
Feature activation | -1.31e-04 |
Pos logit contributions | |
icked | +0.048 |
es | +0.038 |
able | +0.032 |
bled | +0.032 |
together | +0.031 |
Neg logit contributions | |
city | -0.071 |
museum | -0.069 |
hotel | -0.068 |
classroom | -0.068 |
principal | -0.067 |
Token | this |
Token ID | 428 |
Feature activation | +7.16e-02 |
Pos logit contributions | |
there | +0.000 |
together | +0.000 |
icked | +0.000 |
on | +0.000 |
es | +0.000 |
Neg logit contributions | |
hallway | -0.000 |
principal | -0.000 |
hotel | -0.000 |
city | -0.000 |
ey | -0.000 |
Token | answer |
Token ID | 3280 |
Feature activation | -4.22e-02 |
Pos logit contributions | |
principal | +0.171 |
hallway | +0.168 |
ey | +0.165 |
hotel | +0.164 |
city | +0.163 |
Neg logit contributions | |
together | -0.083 |
there | -0.081 |
on | -0.080 |
icked | -0.078 |
es | -0.075 |
Token | . |
Token ID | 13 |
Feature activation | -3.38e-03 |
Pos logit contributions | |
icked | +0.048 |
together | +0.036 |
es | +0.035 |
there | +0.034 |
able | +0.030 |
Neg logit contributions | |
city | -0.102 |
principal | -0.101 |
hotel | -0.100 |
hallway | -0.100 |
classroom | -0.100 |
Token | farmer |
Token ID | 18739 |
Feature activation | -3.63e-02 |
Pos logit contributions | |
ey | +0.052 |
opl | +0.050 |
Ta | +0.050 |
alphabet | +0.049 |
principal | +0.049 |
Neg logit contributions | |
there | -0.061 |
on | -0.057 |
it | -0.056 |
in | -0.054 |
together | -0.054 |
Token | appeared |
Token ID | 4120 |
Feature activation | -6.25e-02 |
Pos logit contributions | |
icked | +0.058 |
together | +0.056 |
es | +0.052 |
there | +0.051 |
on | +0.050 |
Neg logit contributions | |
hallway | -0.068 |
city | -0.067 |
principal | -0.066 |
hotel | -0.066 |
doorway | -0.066 |
Token | from |
Token ID | 422 |
Feature activation | +7.87e-03 |
Pos logit contributions | |
icked | +0.079 |
es | +0.063 |
neither | +0.059 |
there | +0.056 |
together | +0.055 |
Neg logit contributions | |
doorway | -0.131 |
hallway | -0.129 |
city | -0.127 |
ey | -0.122 |
classroom | -0.122 |
Token | behind |
Token ID | 2157 |
Feature activation | -1.62e-02 |
Pos logit contributions | |
hallway | +0.020 |
principal | +0.020 |
city | +0.020 |
hotel | +0.020 |
doorway | +0.020 |
Neg logit contributions | |
icked | -0.008 |
together | -0.007 |
es | -0.007 |
there | -0.007 |
on | -0.006 |
Token | a |
Token ID | 257 |
Feature activation | +5.61e-02 |
Pos logit contributions | |
icked | +0.021 |
together | +0.018 |
es | +0.017 |
there | +0.016 |
able | +0.016 |
Neg logit contributions | |
hallway | -0.037 |
city | -0.037 |
doorway | -0.036 |
hotel | -0.036 |
principal | -0.035 |
Token | tree |
Token ID | 5509 |
Feature activation | -4.50e-02 |
Pos logit contributions | |
Ta | +0.092 |
principal | +0.091 |
alphabet | +0.088 |
orchestra | +0.087 |
former | +0.086 |
Neg logit contributions | |
there | -0.111 |
it | -0.105 |
on | -0.105 |
that | -0.098 |
in | -0.097 |
Token | . |
Token ID | 13 |
Feature activation | -9.66e-03 |
Pos logit contributions | |
icked | +0.043 |
together | +0.036 |
es | +0.034 |
there | +0.032 |
on | +0.031 |
Neg logit contributions | |
hallway | -0.120 |
city | -0.117 |
doorway | -0.116 |
hotel | -0.115 |
classroom | -0.114 |
Token | He |
Token ID | 679 |
Feature activation | +5.56e-03 |
Pos logit contributions | |
icked | +0.013 |
together | +0.013 |
there | +0.013 |
es | +0.012 |
on | +0.012 |
Neg logit contributions | |
hallway | -0.021 |
principal | -0.020 |
city | -0.020 |
hotel | -0.020 |
doorway | -0.020 |
Token | tapped |
Token ID | 22499 |
Feature activation | -6.24e-02 |
Pos logit contributions | |
hallway | +0.011 |
principal | +0.011 |
hotel | +0.011 |
city | +0.011 |
ey | +0.011 |
Neg logit contributions | |
icked | -0.008 |
together | -0.008 |
there | -0.008 |
on | -0.008 |
es | -0.008 |
Token | his |
Token ID | 465 |
Feature activation | +9.09e-02 |
Pos logit contributions | |
icked | +0.071 |
together | +0.063 |
es | +0.059 |
there | +0.059 |
able | +0.058 |
Neg logit contributions | |
hallway | -0.140 |
city | -0.139 |
ey | -0.137 |
principal | -0.137 |
tant | -0.137 |
Token | foot |
Token ID | 2366 |
Feature activation | -6.17e-02 |
Pos logit contributions | |
principal | +0.133 |
salon | +0.128 |
hallway | +0.127 |
hotel | +0.127 |
city | +0.125 |
Neg logit contributions | |
on | -0.193 |
there | -0.185 |
icked | -0.183 |
in | -0.183 |
together | -0.180 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.01e-02 |
Pos logit contributions | |
together | +0.003 |
there | +0.003 |
icked | +0.003 |
on | +0.003 |
es | +0.003 |
Neg logit contributions | |
hallway | -0.007 |
hotel | -0.007 |
ey | -0.007 |
city | -0.007 |
priest | -0.007 |
Token | did |
Token ID | 750 |
Feature activation | -4.53e-02 |
Pos logit contributions | |
hallway | +0.022 |
hotel | +0.022 |
city | +0.022 |
principal | +0.022 |
ey | +0.022 |
Neg logit contributions | |
together | -0.013 |
icked | -0.013 |
there | -0.013 |
on | -0.013 |
es | -0.013 |
Token | not |
Token ID | 407 |
Feature activation | +1.54e-02 |
Pos logit contributions | |
icked | +0.057 |
es | +0.049 |
together | +0.043 |
there | +0.043 |
on | +0.042 |
Neg logit contributions | |
principal | -0.102 |
classroom | -0.100 |
babys | -0.098 |
playground | -0.098 |
hallway | -0.098 |
Token | like |
Token ID | 588 |
Feature activation | +4.69e-03 |
Pos logit contributions | |
principal | +0.026 |
hallway | +0.026 |
hotel | +0.025 |
city | +0.025 |
orchestra | +0.025 |
Neg logit contributions | |
icked | -0.028 |
together | -0.027 |
there | -0.027 |
es | -0.026 |
on | -0.026 |
Token | this |
Token ID | 428 |
Feature activation | +5.84e-02 |
Pos logit contributions | |
hallway | +0.013 |
hotel | +0.013 |
ey | +0.013 |
doorway | +0.013 |
principal | +0.013 |
Neg logit contributions | |
there | -0.003 |
together | -0.003 |
icked | -0.003 |
on | -0.003 |
es | -0.003 |
Token | idea |
Token ID | 2126 |
Feature activation | -4.16e-02 |
Pos logit contributions | |
hallway | +0.125 |
principal | +0.124 |
hotel | +0.122 |
city | +0.122 |
classroom | +0.122 |
Neg logit contributions | |
icked | -0.082 |
together | -0.076 |
es | -0.075 |
there | -0.074 |
on | -0.071 |
Token | . |
Token ID | 13 |
Feature activation | -1.38e-03 |
Pos logit contributions | |
icked | +0.062 |
es | +0.048 |
bled | +0.048 |
bed | +0.044 |
together | +0.044 |
Neg logit contributions | |
principal | -0.088 |
classroom | -0.086 |
babys | -0.084 |
city | -0.083 |
museum | -0.083 |
Token | She |
Token ID | 1375 |
Feature activation | +1.47e-02 |
Pos logit contributions | |
icked | +0.002 |
there | +0.002 |
together | +0.002 |
on | +0.002 |
es | +0.002 |
Neg logit contributions | |
principal | -0.003 |
hallway | -0.003 |
city | -0.003 |
classroom | -0.003 |
ey | -0.003 |
Token | said |
Token ID | 531 |
Feature activation | -2.49e-02 |
Pos logit contributions | |
hallway | +0.033 |
ey | +0.033 |
principal | +0.033 |
hotel | +0.032 |
city | +0.032 |
Neg logit contributions | |
icked | -0.019 |
together | -0.019 |
there | -0.018 |
on | -0.018 |
es | -0.018 |
Token | , |
Token ID | 11 |
Feature activation | -2.82e-02 |
Pos logit contributions | |
icked | +0.027 |
es | +0.024 |
pped | +0.024 |
able | +0.024 |
together | +0.023 |
Neg logit contributions | |
babys | -0.051 |
principal | -0.051 |
teacher | -0.050 |
salon | -0.049 |
priest | -0.049 |
Token | " |
Token ID | 366 |
Feature activation | +6.91e-03 |
Pos logit contributions | |
icked | +0.046 |
es | +0.042 |
together | +0.040 |
there | +0.040 |
able | +0.037 |
Neg logit contributions | |
principal | -0.052 |
babys | -0.051 |
hallway | -0.050 |
woman | -0.049 |
bus | -0.048 |
Token | smiled |
Token ID | 13541 |
Feature activation | -3.41e-02 |
Pos logit contributions | |
principal | +0.023 |
ey | +0.023 |
hallway | +0.023 |
city | +0.022 |
classroom | +0.022 |
Neg logit contributions | |
there | -0.014 |
together | -0.014 |
on | -0.014 |
icked | -0.014 |
es | -0.013 |
Token | and |
Token ID | 290 |
Feature activation | -1.01e-02 |
Pos logit contributions | |
icked | +0.044 |
es | +0.037 |
there | +0.032 |
pped | +0.031 |
together | +0.030 |
Neg logit contributions | |
city | -0.075 |
orchestra | -0.074 |
doorway | -0.074 |
hotel | -0.074 |
classroom | -0.073 |
Token | sn |
Token ID | 3013 |
Feature activation | -3.58e-03 |
Pos logit contributions | |
icked | +0.017 |
together | +0.016 |
es | +0.016 |
there | +0.016 |
on | +0.015 |
Neg logit contributions | |
hallway | -0.018 |
hotel | -0.018 |
doorway | -0.018 |
principal | -0.018 |
orchestra | -0.018 |
Token | ugg |
Token ID | 6837 |
Feature activation | -7.01e-03 |
Pos logit contributions | |
there | +0.008 |
es | +0.008 |
icked | +0.008 |
on | +0.008 |
together | +0.008 |
Neg logit contributions | |
doorway | -0.003 |
hotel | -0.003 |
salon | -0.003 |
hallway | -0.003 |
neighbourhood | -0.003 |
Token | led |
Token ID | 992 |
Feature activation | -6.28e-02 |
Pos logit contributions | |
icked | +0.008 |
together | +0.008 |
there | +0.007 |
able | +0.007 |
instead | +0.007 |
Neg logit contributions | |
hallway | -0.016 |
doorway | -0.016 |
hotel | -0.015 |
orchestra | -0.015 |
ey | -0.015 |
Token | up |
Token ID | 510 |
Feature activation | -4.13e-02 |
Pos logit contributions | |
icked | +0.075 |
es | +0.063 |
there | +0.051 |
together | +0.047 |
able | +0.044 |
Neg logit contributions | |
hallway | -0.148 |
doorway | -0.145 |
hotel | -0.143 |
city | -0.140 |
orchestra | -0.140 |
Token | close |
Token ID | 1969 |
Feature activation | -2.91e-02 |
Pos logit contributions | |
icked | +0.035 |
es | +0.029 |
there | +0.021 |
together | +0.021 |
it | +0.017 |
Neg logit contributions | |
hallway | -0.114 |
doorway | -0.111 |
hotel | -0.110 |
city | -0.110 |
classroom | -0.109 |
Token | to |
Token ID | 284 |
Feature activation | -8.43e-03 |
Pos logit contributions | |
icked | +0.030 |
es | +0.027 |
together | +0.021 |
there | +0.020 |
on | +0.018 |
Neg logit contributions | |
doorway | -0.073 |
hallway | -0.072 |
hotel | -0.071 |
city | -0.071 |
priest | -0.069 |
Token | her |
Token ID | 607 |
Feature activation | +7.05e-02 |
Pos logit contributions | |
icked | +0.008 |
together | +0.008 |
there | +0.008 |
on | +0.008 |
es | +0.008 |
Neg logit contributions | |
hallway | -0.021 |
principal | -0.021 |
hotel | -0.021 |
city | -0.021 |
ey | -0.021 |
Token | mom |
Token ID | 1995 |
Feature activation | +4.64e-03 |
Pos logit contributions | |
principal | +0.120 |
hallway | +0.118 |
ey | +0.116 |
orchestra | +0.114 |
hotel | +0.114 |
Neg logit contributions | |
together | -0.128 |
there | -0.127 |
on | -0.127 |
icked | -0.122 |
es | -0.119 |
Token | to |
Token ID | 284 |
Feature activation | -1.07e-03 |
Pos logit contributions | |
hallway | +0.011 |
city | +0.011 |
classroom | +0.011 |
principal | +0.011 |
hotel | +0.011 |
Neg logit contributions | |
icked | -0.005 |
together | -0.004 |
es | -0.004 |
there | -0.004 |
on | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -9.39e-03 |
Pos logit contributions | |
icked | +0.067 |
es | +0.056 |
bed | +0.052 |
pped | +0.051 |
bled | +0.049 |
Neg logit contributions | |
hotel | -0.096 |
museum | -0.096 |
hallway | -0.094 |
beach | -0.094 |
playground | -0.094 |
Token | He |
Token ID | 679 |
Feature activation | +6.64e-03 |
Pos logit contributions | |
icked | +0.014 |
on | +0.014 |
together | +0.014 |
there | +0.014 |
es | +0.013 |
Neg logit contributions | |
hallway | -0.018 |
city | -0.018 |
doorway | -0.018 |
ey | -0.018 |
museum | -0.018 |
Token | looked |
Token ID | 3114 |
Feature activation | -5.14e-02 |
Pos logit contributions | |
hallway | +0.014 |
principal | +0.014 |
ey | +0.014 |
hotel | +0.014 |
city | +0.014 |
Neg logit contributions | |
icked | -0.009 |
together | -0.009 |
there | -0.009 |
on | -0.009 |
es | -0.009 |
Token | at |
Token ID | 379 |
Feature activation | -1.94e-02 |
Pos logit contributions | |
icked | +0.054 |
es | +0.052 |
together | +0.046 |
there | +0.042 |
on | +0.039 |
Neg logit contributions | |
hotel | -0.126 |
city | -0.125 |
hallway | -0.123 |
classroom | -0.121 |
orchestra | -0.121 |
Token | the |
Token ID | 262 |
Feature activation | +1.29e-01 |
Pos logit contributions | |
icked | +0.018 |
together | +0.017 |
es | +0.017 |
there | +0.016 |
on | +0.015 |
Neg logit contributions | |
hotel | -0.049 |
hallway | -0.049 |
city | -0.049 |
principal | -0.048 |
doorway | -0.047 |
Token | bottles |
Token ID | 14096 |
Feature activation | -3.35e-02 |
Pos logit contributions | |
principal | +0.163 |
ey | +0.156 |
priest | +0.152 |
hallway | +0.151 |
classroom | +0.150 |
Neg logit contributions | |
on | -0.301 |
there | -0.297 |
together | -0.292 |
icked | -0.281 |
es | -0.281 |
Token | and |
Token ID | 290 |
Feature activation | -9.00e-03 |
Pos logit contributions | |
icked | +0.042 |
es | +0.037 |
together | +0.034 |
there | +0.032 |
on | +0.030 |
Neg logit contributions | |
hallway | -0.077 |
hotel | -0.075 |
doorway | -0.075 |
playground | -0.075 |
city | -0.075 |
Token | read |
Token ID | 1100 |
Feature activation | -6.04e-02 |
Pos logit contributions | |
icked | +0.015 |
together | +0.014 |
there | +0.014 |
on | +0.014 |
es | +0.014 |
Neg logit contributions | |
hallway | -0.017 |
principal | -0.017 |
city | -0.017 |
hotel | -0.017 |
ey | -0.016 |
Token | the |
Token ID | 262 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
icked | +0.072 |
able | +0.056 |
es | +0.054 |
together | +0.054 |
there | +0.052 |
Neg logit contributions | |
hotel | -0.137 |
city | -0.135 |
museum | -0.135 |
hallway | -0.132 |
library | -0.131 |
Token | directions |
Token ID | 11678 |
Feature activation | -1.21e-02 |
Pos logit contributions | |
principal | +0.172 |
ey | +0.166 |
priest | +0.162 |
hallway | +0.161 |
classroom | +0.161 |
Neg logit contributions | |
on | -0.299 |
there | -0.296 |
together | -0.296 |
icked | -0.286 |
es | -0.283 |
Token | . |
Token ID | 13 |
Feature activation | -6.68e-03 |
Pos logit contributions | |
icked | +0.013 |
es | +0.011 |
together | +0.011 |
there | +0.010 |
able | +0.009 |
Neg logit contributions | |
hallway | -0.030 |
hotel | -0.029 |
city | -0.029 |
doorway | -0.029 |
classroom | -0.028 |
Token | and |
Token ID | 290 |
Feature activation | -8.99e-03 |
Pos logit contributions | |
icked | +0.049 |
es | +0.040 |
together | +0.035 |
there | +0.033 |
on | +0.030 |
Neg logit contributions | |
hallway | -0.109 |
principal | -0.109 |
doorway | -0.109 |
salon | -0.109 |
priest | -0.108 |
Token | put |
Token ID | 1234 |
Feature activation | -2.95e-02 |
Pos logit contributions | |
icked | +0.013 |
together | +0.013 |
there | +0.012 |
es | +0.012 |
on | +0.012 |
Neg logit contributions | |
hallway | -0.018 |
principal | -0.018 |
hotel | -0.018 |
city | -0.018 |
doorway | -0.018 |
Token | it |
Token ID | 340 |
Feature activation | -1.52e-02 |
Pos logit contributions | |
icked | +0.025 |
es | +0.022 |
together | +0.022 |
there | +0.020 |
on | +0.018 |
Neg logit contributions | |
hotel | -0.075 |
hallway | -0.075 |
city | -0.075 |
principal | -0.075 |
doorway | -0.074 |
Token | into |
Token ID | 656 |
Feature activation | -1.64e-02 |
Pos logit contributions | |
icked | +0.014 |
es | +0.014 |
it | +0.011 |
together | +0.010 |
Remy | +0.010 |
Neg logit contributions | |
hallway | -0.035 |
doorway | -0.035 |
hotel | -0.034 |
tour | -0.034 |
bus | -0.034 |
Token | her |
Token ID | 607 |
Feature activation | +7.94e-02 |
Pos logit contributions | |
together | +0.014 |
there | +0.014 |
icked | +0.014 |
on | +0.013 |
es | +0.013 |
Neg logit contributions | |
hallway | -0.043 |
principal | -0.043 |
hotel | -0.042 |
city | -0.042 |
ey | -0.042 |
Token | pocket |
Token ID | 10000 |
Feature activation | -4.91e-02 |
Pos logit contributions | |
principal | +0.098 |
hallway | +0.097 |
orchestra | +0.094 |
ey | +0.094 |
city | +0.092 |
Neg logit contributions | |
there | -0.180 |
together | -0.180 |
on | -0.178 |
icked | -0.176 |
es | -0.171 |
Token | . |
Token ID | 13 |
Feature activation | -5.31e-03 |
Pos logit contributions | |
icked | +0.043 |
together | +0.036 |
es | +0.036 |
there | +0.035 |
on | +0.033 |
Neg logit contributions | |
hallway | -0.129 |
hotel | -0.126 |
doorway | -0.126 |
principal | -0.126 |
city | -0.125 |
Token | � |
Token ID | 6184 |
Feature activation | +5.28e-02 |
Pos logit contributions | |
icked | +0.007 |
together | +0.007 |
there | +0.006 |
on | +0.006 |
es | +0.006 |
Neg logit contributions | |
hallway | -0.012 |
principal | -0.012 |
city | -0.012 |
hotel | -0.012 |
ey | -0.011 |
Token | � |
Token ID | 95 |
Feature activation | +8.49e-02 |
Pos logit contributions | |
principal | +0.088 |
hotel | +0.087 |
city | +0.086 |
museum | +0.086 |
hallway | +0.085 |
Neg logit contributions | |
together | -0.095 |
icked | -0.095 |
there | -0.094 |
on | -0.093 |
es | -0.092 |
Token | € |
Token ID | 26391 |
Feature activation | +3.71e-02 |
Pos logit contributions | |
hallway | +0.143 |
principal | +0.142 |
city | +0.138 |
hotel | +0.138 |
ey | +0.137 |
Neg logit contributions | |
icked | -0.155 |
together | -0.153 |
there | -0.152 |
on | -0.149 |
es | -0.147 |
Token | � |
Token ID | 129 |
Feature activation | +5.54e-02 |
Pos logit contributions | |
city | +0.071 |
hallway | +0.070 |
classroom | +0.070 |
principal | +0.068 |
playground | +0.068 |
Neg logit contributions | |
icked | -0.061 |
together | -0.056 |
on | -0.054 |
against | -0.053 |
there | -0.051 |