Token | no |
Token ID | 645 |
Feature activation | +1.59e+00 |
Pos logit contributions | |
mess | +0.041 |
wasteful | +0.038 |
overflow | +0.038 |
roughly | +0.036 |
neither | +0.036 |
Neg logit contributions | |
name | -0.046 |
ka | -0.044 |
ance | -0.043 |
dot | -0.042 |
ti | -0.041 |
Token | ! |
Token ID | 0 |
Feature activation | +9.65e-01 |
Pos logit contributions | |
neither | +0.078 |
overflow | +0.078 |
wasteful | +0.078 |
unfortunately | +0.070 |
ruining | +0.070 |
Neg logit contributions | |
ka | -0.042 |
name | -0.041 |
powers | -0.037 |
dot | -0.035 |
aboard | -0.034 |
Token | We |
Token ID | 775 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
mess | +0.046 |
neither | +0.043 |
wasteful | +0.042 |
overflow | +0.041 |
ruining | +0.040 |
Neg logit contributions | |
name | -0.032 |
ka | -0.029 |
ance | -0.029 |
dot | -0.028 |
anced | -0.027 |
Token | destroyed |
Token ID | 6572 |
Feature activation | +2.90e+00 |
Pos logit contributions | |
neither | +0.058 |
mess | +0.057 |
structures | +0.056 |
wasteful | +0.055 |
overflow | +0.054 |
Neg logit contributions | |
name | -0.043 |
tie | -0.040 |
ka | -0.039 |
ocket | -0.036 |
aved | -0.036 |
Token | the |
Token ID | 262 |
Feature activation | +2.13e+00 |
Pos logit contributions | |
overflow | +0.137 |
forgot | +0.129 |
neither | +0.128 |
wasteful | +0.126 |
ivery | +0.122 |
Neg logit contributions | |
aboard | -0.088 |
ka | -0.081 |
name | -0.081 |
ance | -0.070 |
Magic | -0.070 |
Token | apple |
Token ID | 17180 |
Feature activation | +4.79e+00 |
Pos logit contributions | |
neither | +0.100 |
overflow | +0.091 |
wasteful | +0.087 |
ruining | +0.082 |
roughly | +0.082 |
Neg logit contributions | |
name | -0.075 |
orn | -0.069 |
tie | -0.066 |
amulet | -0.062 |
Crown | -0.061 |
Token | !" |
Token ID | 2474 |
Feature activation | +8.88e-01 |
Pos logit contributions | |
wasteful | +0.259 |
overflow | +0.239 |
sudden | +0.237 |
neither | +0.237 |
Sadly | +0.234 |
Neg logit contributions | |
orn | -0.091 |
aboard | -0.081 |
tie | -0.081 |
ed | -0.067 |
- | -0.066 |
Token | Lily |
Token ID | 20037 |
Feature activation | +2.89e+00 |
Pos logit contributions | |
mess | +0.030 |
neither | +0.030 |
wasteful | +0.029 |
overflow | +0.026 |
ruining | +0.026 |
Neg logit contributions | |
name | -0.041 |
ka | -0.038 |
dot | -0.037 |
leaf | -0.037 |
tie | -0.036 |
Token | cried |
Token ID | 16896 |
Feature activation | +2.57e-01 |
Pos logit contributions | |
overflow | +0.163 |
structures | +0.161 |
flood | +0.159 |
wasteful | +0.158 |
earthquake | +0.154 |
Neg logit contributions | |
ed | -0.056 |
name | -0.055 |
hugs | -0.049 |
u | -0.047 |
tie | -0.046 |
Token | . |
Token ID | 13 |
Feature activation | +5.92e-01 |
Pos logit contributions | |
neither | +0.011 |
mess | +0.011 |
wasteful | +0.011 |
ruining | +0.010 |
overflow | +0.010 |
Neg logit contributions | |
ka | -0.009 |
name | -0.009 |
tie | -0.009 |
ance | -0.009 |
dot | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | +2.95e-01 |
Pos logit contributions | |
mess | +0.028 |
neither | +0.027 |
wasteful | +0.027 |
overflow | +0.025 |
ruining | +0.025 |
Neg logit contributions | |
name | -0.020 |
tie | -0.019 |
ka | -0.018 |
dot | -0.018 |
ance | -0.018 |
Token | go |
Token ID | 467 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
neither | +0.062 |
wasteful | +0.061 |
mess | +0.058 |
roughly | +0.054 |
mosquitoes | +0.054 |
Neg logit contributions | |
name | -0.040 |
tie | -0.031 |
ance | -0.030 |
aboard | -0.030 |
ka | -0.030 |
Token | back |
Token ID | 736 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
ka | +0.041 |
tie | +0.040 |
ance | +0.040 |
anced | +0.039 |
dot | +0.039 |
Neg logit contributions | |
mess | -0.041 |
neither | -0.040 |
overflow | -0.035 |
fighting | -0.034 |
wasteful | -0.034 |
Token | to |
Token ID | 284 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
neither | +0.052 |
wasteful | +0.051 |
mess | +0.049 |
overflow | +0.046 |
ruining | +0.045 |
Neg logit contributions | |
aboard | -0.034 |
name | -0.032 |
ka | -0.031 |
ance | -0.030 |
tie | -0.029 |
Token | the |
Token ID | 262 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
neither | +0.060 |
wasteful | +0.057 |
roughly | +0.053 |
mess | +0.053 |
overflow | +0.050 |
Neg logit contributions | |
name | -0.045 |
aboard | -0.044 |
ance | -0.040 |
Magic | -0.039 |
outer | -0.039 |
Token | pool |
Token ID | 5933 |
Feature activation | +3.26e+00 |
Pos logit contributions | |
neither | +0.046 |
wasteful | +0.040 |
mess | +0.038 |
roughly | +0.037 |
ruining | +0.036 |
Neg logit contributions | |
name | -0.044 |
aboard | -0.040 |
ka | -0.040 |
Crown | -0.039 |
tie | -0.039 |
Token | side |
Token ID | 1589 |
Feature activation | +4.32e+00 |
Pos logit contributions | |
wasteful | +0.155 |
neither | +0.149 |
ruining | +0.149 |
mosquitoes | +0.145 |
spitting | +0.139 |
Neg logit contributions | |
aboard | -0.095 |
name | -0.089 |
tie | -0.085 |
ka | -0.082 |
suit | -0.079 |
Token | , |
Token ID | 11 |
Feature activation | +2.21e+00 |
Pos logit contributions | |
wasteful | +0.237 |
neither | +0.219 |
ruining | +0.210 |
mosquitoes | +0.204 |
roughly | +0.195 |
Neg logit contributions | |
name | -0.130 |
aboard | -0.116 |
powers | -0.099 |
outfit | -0.098 |
costume | -0.096 |
Token | Mia |
Token ID | 32189 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
wasteful | +0.078 |
neither | +0.073 |
roughly | +0.071 |
oving | +0.070 |
mess | +0.068 |
Neg logit contributions | |
name | -0.104 |
aboard | -0.086 |
hello | -0.081 |
ance | -0.078 |
hi | -0.078 |
Token | . |
Token ID | 13 |
Feature activation | +3.78e-01 |
Pos logit contributions | |
neither | +0.053 |
wasteful | +0.050 |
mess | +0.048 |
overflow | +0.047 |
ruining | +0.047 |
Neg logit contributions | |
name | -0.044 |
aboard | -0.043 |
ka | -0.043 |
ance | -0.042 |
tie | -0.042 |
Token | Maybe |
Token ID | 6674 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
mess | +0.015 |
neither | +0.015 |
wasteful | +0.014 |
overflow | +0.013 |
ruining | +0.013 |
Neg logit contributions | |
name | -0.015 |
tie | -0.015 |
ka | -0.015 |
ance | -0.015 |
dot | -0.014 |
Token | we |
Token ID | 356 |
Feature activation | +2.58e-01 |
Pos logit contributions | |
mess | +0.040 |
wasteful | +0.039 |
neither | +0.037 |
overflow | +0.035 |
roughly | +0.035 |
Neg logit contributions | |
name | -0.058 |
aboard | -0.058 |
ance | -0.055 |
tie | -0.052 |
dot | -0.052 |
Token | going |
Token ID | 1016 |
Feature activation | -6.96e-01 |
Pos logit contributions | |
neither | +0.036 |
mess | +0.036 |
wasteful | +0.033 |
structures | +0.032 |
overflow | +0.032 |
Neg logit contributions | |
name | -0.038 |
dot | -0.036 |
leaf | -0.036 |
aboard | -0.035 |
tie | -0.034 |
Token | to |
Token ID | 284 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
ka | +0.026 |
ance | +0.025 |
tie | +0.025 |
leaf | +0.025 |
anced | +0.025 |
Neg logit contributions | |
mess | -0.030 |
neither | -0.030 |
wasteful | -0.027 |
fighting | -0.026 |
ruining | -0.026 |
Token | eat |
Token ID | 4483 |
Feature activation | +1.47e+00 |
Pos logit contributions | |
neither | +0.063 |
wasteful | +0.059 |
mess | +0.055 |
roughly | +0.054 |
structures | +0.051 |
Neg logit contributions | |
name | -0.049 |
aboard | -0.042 |
Magic | -0.041 |
leaf | -0.040 |
inspect | -0.040 |
Token | you |
Token ID | 345 |
Feature activation | +2.02e+00 |
Pos logit contributions | |
neither | +0.063 |
mess | +0.058 |
wasteful | +0.057 |
overflow | +0.055 |
forgot | +0.052 |
Neg logit contributions | |
name | -0.053 |
aboard | -0.052 |
tie | -0.051 |
shirt | -0.051 |
ance | -0.051 |
Token | for |
Token ID | 329 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
neither | +0.091 |
wasteful | +0.087 |
mess | +0.081 |
overflow | +0.079 |
Uh | +0.077 |
Neg logit contributions | |
name | -0.069 |
aboard | -0.068 |
tie | -0.067 |
ka | -0.066 |
ance | -0.066 |
Token | dinner |
Token ID | 8073 |
Feature activation | +4.25e+00 |
Pos logit contributions | |
neither | +0.055 |
wasteful | +0.051 |
mess | +0.051 |
overflow | +0.049 |
roughly | +0.046 |
Neg logit contributions | |
powers | -0.035 |
dot | -0.035 |
ance | -0.035 |
name | -0.035 |
aboard | -0.034 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
wasteful | +0.211 |
Uh | +0.199 |
overflow | +0.198 |
neither | +0.198 |
structures | +0.190 |
Neg logit contributions | |
tie | -0.109 |
ti | -0.104 |
aboard | -0.101 |
ka | -0.099 |
name | -0.098 |
Token | Tim |
Token ID | 5045 |
Feature activation | +6.80e-01 |
Pos logit contributions | |
mess | +0.041 |
neither | +0.040 |
wasteful | +0.040 |
structures | +0.036 |
ruining | +0.035 |
Neg logit contributions | |
name | -0.051 |
tie | -0.048 |
shirt | -0.047 |
Aid | -0.046 |
dot | -0.046 |
Token | my |
Token ID | 1820 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
neither | +0.034 |
wasteful | +0.032 |
mess | +0.032 |
overflow | +0.030 |
ruining | +0.029 |
Neg logit contributions | |
ka | -0.021 |
name | -0.021 |
tie | -0.020 |
shirt | -0.019 |
ance | -0.019 |
Token | was |
Token ID | 373 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
neither | +0.067 |
wasteful | +0.067 |
mess | +0.064 |
overflow | +0.061 |
structures | +0.061 |
Neg logit contributions | |
name | -0.044 |
tie | -0.041 |
ka | -0.041 |
shirt | -0.040 |
dot | -0.039 |
Token | so |
Token ID | 523 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
mess | +0.049 |
neither | +0.048 |
wasteful | +0.047 |
overflow | +0.046 |
structures | +0.046 |
Neg logit contributions | |
name | -0.055 |
dot | -0.047 |
tie | -0.047 |
aboard | -0.045 |
leaf | -0.045 |
Token | want |
Token ID | 765 |
Feature activation | +2.12e-01 |
Pos logit contributions | |
neither | +0.003 |
mess | +0.002 |
wasteful | +0.002 |
overflow | +0.002 |
fighting | +0.002 |
Neg logit contributions | |
ka | -0.002 |
tie | -0.002 |
name | -0.002 |
ance | -0.002 |
dot | -0.002 |
Token | to |
Token ID | 284 |
Feature activation | +8.96e-01 |
Pos logit contributions | |
neither | +0.010 |
mess | +0.010 |
wasteful | +0.010 |
overflow | +0.009 |
roughly | +0.009 |
Neg logit contributions | |
ka | -0.006 |
name | -0.006 |
tie | -0.006 |
ance | -0.006 |
dot | -0.006 |
Token | go |
Token ID | 467 |
Feature activation | -6.94e-01 |
Pos logit contributions | |
wasteful | +0.044 |
neither | +0.043 |
mess | +0.042 |
mosquitoes | +0.038 |
roughly | +0.037 |
Neg logit contributions | |
name | -0.029 |
ance | -0.025 |
tie | -0.025 |
aboard | -0.025 |
dot | -0.025 |
Token | in |
Token ID | 287 |
Feature activation | +9.92e-01 |
Pos logit contributions | |
ka | +0.025 |
leaf | +0.024 |
tie | +0.024 |
ance | +0.024 |
shirt | +0.024 |
Neg logit contributions | |
mess | -0.031 |
neither | -0.031 |
wasteful | -0.028 |
fighting | -0.027 |
ruining | -0.027 |
Token | the |
Token ID | 262 |
Feature activation | +3.80e-02 |
Pos logit contributions | |
neither | +0.043 |
wasteful | +0.043 |
mess | +0.040 |
overflow | +0.037 |
roughly | +0.036 |
Neg logit contributions | |
aboard | -0.035 |
name | -0.034 |
ka | -0.033 |
ance | -0.032 |
tie | -0.032 |
Token | cave |
Token ID | 11527 |
Feature activation | +4.18e+00 |
Pos logit contributions | |
neither | +0.002 |
mess | +0.001 |
wasteful | +0.001 |
overflow | +0.001 |
fighting | +0.001 |
Neg logit contributions | |
ka | -0.001 |
tie | -0.001 |
name | -0.001 |
ance | -0.001 |
dot | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
wasteful | +0.194 |
mosquitoes | +0.188 |
spitting | +0.185 |
neither | +0.185 |
ruining | +0.182 |
Neg logit contributions | |
aboard | -0.116 |
tie | -0.115 |
ka | -0.109 |
name | -0.108 |
ti | -0.103 |
Token | It |
Token ID | 632 |
Feature activation | +9.23e-01 |
Pos logit contributions | |
mess | +0.014 |
neither | +0.014 |
wasteful | +0.013 |
overflow | +0.012 |
ruining | +0.012 |
Neg logit contributions | |
name | -0.010 |
ka | -0.010 |
tie | -0.010 |
dot | -0.010 |
ance | -0.010 |
Token | 's |
Token ID | 338 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
neither | +0.038 |
wasteful | +0.037 |
mess | +0.036 |
roughly | +0.034 |
overflow | +0.033 |
Neg logit contributions | |
name | -0.033 |
ka | -0.033 |
aboard | -0.032 |
ance | -0.032 |
tie | -0.031 |
Token | too |
Token ID | 1165 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
neither | +0.043 |
wasteful | +0.038 |
mess | +0.037 |
structures | +0.036 |
overflow | +0.036 |
Neg logit contributions | |
name | -0.048 |
aboard | -0.045 |
tie | -0.044 |
powers | -0.043 |
ka | -0.043 |
Token | ugly |
Token ID | 13400 |
Feature activation | +8.72e-01 |
Pos logit contributions | |
neither | +0.046 |
wasteful | +0.041 |
mess | +0.040 |
ruining | +0.040 |
fighting | +0.039 |
Neg logit contributions | |
ka | -0.041 |
name | -0.041 |
aboard | -0.040 |
tie | -0.039 |
dot | -0.037 |
Token | not |
Token ID | 407 |
Feature activation | +6.16e-01 |
Pos logit contributions | |
wasteful | +0.064 |
structures | +0.062 |
neither | +0.061 |
rats | +0.060 |
mess | +0.058 |
Neg logit contributions | |
name | -0.075 |
tie | -0.059 |
polish | -0.056 |
ad | -0.056 |
Magic | -0.055 |
Token | pull |
Token ID | 2834 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
neither | +0.028 |
mess | +0.026 |
wasteful | +0.026 |
ruining | +0.024 |
rats | +0.024 |
Neg logit contributions | |
name | -0.022 |
tie | -0.021 |
ka | -0.021 |
shirt | -0.020 |
dot | -0.020 |
Token | them |
Token ID | 606 |
Feature activation | +2.56e+00 |
Pos logit contributions | |
mess | +0.042 |
neither | +0.042 |
wasteful | +0.040 |
overflow | +0.040 |
roughly | +0.037 |
Neg logit contributions | |
ka | -0.038 |
tie | -0.035 |
name | -0.035 |
aboard | -0.035 |
ance | -0.034 |
Token | off |
Token ID | 572 |
Feature activation | +2.72e+00 |
Pos logit contributions | |
wasteful | +0.134 |
neither | +0.128 |
mosquitoes | +0.126 |
Uh | +0.125 |
spitting | +0.123 |
Neg logit contributions | |
aboard | -0.064 |
tie | -0.058 |
name | -0.057 |
ance | -0.054 |
ka | -0.051 |
Token | the |
Token ID | 262 |
Feature activation | +7.93e-01 |
Pos logit contributions | |
neither | +0.123 |
wasteful | +0.115 |
mess | +0.111 |
overflow | +0.108 |
ruining | +0.106 |
Neg logit contributions | |
aboard | -0.084 |
ka | -0.082 |
tie | -0.078 |
name | -0.078 |
outer | -0.076 |
Token | stems |
Token ID | 21552 |
Feature activation | +4.14e+00 |
Pos logit contributions | |
neither | +0.032 |
wasteful | +0.029 |
mess | +0.027 |
overflow | +0.026 |
fighting | +0.026 |
Neg logit contributions | |
name | -0.031 |
tie | -0.031 |
ka | -0.031 |
shirt | -0.031 |
dot | -0.030 |
Token | , |
Token ID | 11 |
Feature activation | +2.34e+00 |
Pos logit contributions | |
wasteful | +0.203 |
neither | +0.196 |
ruining | +0.190 |
mosquitoes | +0.186 |
rats | +0.184 |
Neg logit contributions | |
tie | -0.109 |
aboard | -0.099 |
name | -0.097 |
ance | -0.094 |
ka | -0.089 |
Token | and |
Token ID | 290 |
Feature activation | +2.26e+00 |
Pos logit contributions | |
mess | +0.077 |
neither | +0.075 |
anth | +0.074 |
wasteful | +0.073 |
roughly | +0.073 |
Neg logit contributions | |
name | -0.098 |
tie | -0.083 |
aboard | -0.082 |
shirt | -0.081 |
rib | -0.080 |
Token | we |
Token ID | 356 |
Feature activation | +9.12e-01 |
Pos logit contributions | |
neither | +0.083 |
mess | +0.077 |
overflow | +0.075 |
wasteful | +0.073 |
roughly | +0.073 |
Neg logit contributions | |
name | -0.088 |
aboard | -0.079 |
powers | -0.079 |
tie | -0.079 |
leaf | -0.079 |
Token | could |
Token ID | 714 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
neither | +0.041 |
mess | +0.040 |
wasteful | +0.039 |
structures | +0.037 |
mosquitoes | +0.037 |
Neg logit contributions | |
name | -0.030 |
tie | -0.029 |
ka | -0.029 |
ance | -0.027 |
leaf | -0.026 |
Token | not |
Token ID | 407 |
Feature activation | +7.28e-01 |
Pos logit contributions | |
wasteful | +0.073 |
structures | +0.069 |
neither | +0.068 |
rats | +0.066 |
erupted | +0.064 |
Neg logit contributions | |
name | -0.070 |
tie | -0.055 |
polish | -0.051 |
aboard | -0.051 |
inspect | -0.050 |
Token | want |
Token ID | 765 |
Feature activation | +4.66e-01 |
Pos logit contributions | |
neither | +0.019 |
mess | +0.018 |
wasteful | +0.017 |
structures | +0.016 |
overflow | +0.016 |
Neg logit contributions | |
name | -0.015 |
tie | -0.014 |
ka | -0.014 |
dot | -0.014 |
leaf | -0.014 |
Token | to |
Token ID | 284 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
neither | +0.022 |
mess | +0.021 |
wasteful | +0.021 |
overflow | +0.019 |
roughly | +0.019 |
Neg logit contributions | |
ka | -0.015 |
name | -0.015 |
tie | -0.015 |
ance | -0.014 |
dot | -0.014 |
Token | play |
Token ID | 711 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
neither | +0.048 |
wasteful | +0.048 |
mess | +0.046 |
structures | +0.043 |
roughly | +0.042 |
Neg logit contributions | |
name | -0.040 |
tie | -0.035 |
ka | -0.035 |
dot | -0.034 |
ance | -0.033 |
Token | in |
Token ID | 287 |
Feature activation | +2.01e+00 |
Pos logit contributions | |
neither | +0.049 |
wasteful | +0.046 |
mess | +0.045 |
overflow | +0.042 |
erupted | +0.040 |
Neg logit contributions | |
name | -0.038 |
tie | -0.037 |
ka | -0.035 |
aboard | -0.035 |
powers | -0.035 |
Token | the |
Token ID | 262 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
neither | +0.099 |
wasteful | +0.098 |
forgot | +0.089 |
overflow | +0.088 |
erupted | +0.087 |
Neg logit contributions | |
outer | -0.055 |
aboard | -0.053 |
powers | -0.051 |
ance | -0.048 |
Magic | -0.047 |
Token | hut |
Token ID | 40812 |
Feature activation | +4.13e+00 |
Pos logit contributions | |
neither | +0.057 |
wasteful | +0.048 |
fighting | +0.046 |
overflow | +0.045 |
mess | +0.044 |
Neg logit contributions | |
name | -0.044 |
orn | -0.041 |
amulet | -0.040 |
cloak | -0.040 |
tie | -0.040 |
Token | anymore |
Token ID | 7471 |
Feature activation | +2.90e+00 |
Pos logit contributions | |
wasteful | +0.196 |
neither | +0.195 |
overflow | +0.181 |
mosquitoes | +0.181 |
ruining | +0.181 |
Neg logit contributions | |
aboard | -0.106 |
tie | -0.101 |
name | -0.099 |
ti | -0.092 |
amulet | -0.089 |
Token | ," |
Token ID | 553 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
wasteful | +0.150 |
overflow | +0.145 |
neither | +0.141 |
ruining | +0.132 |
mess | +0.130 |
Neg logit contributions | |
name | -0.076 |
aboard | -0.070 |
powers | -0.068 |
ance | -0.067 |
ti | -0.065 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
mess | +0.049 |
neither | +0.047 |
overflow | +0.047 |
wasteful | +0.046 |
ruining | +0.045 |
Neg logit contributions | |
name | -0.042 |
Magic | -0.036 |
dot | -0.036 |
shirt | -0.036 |
ka | -0.036 |
Token | said |
Token ID | 531 |
Feature activation | -2.80e-02 |
Pos logit contributions | |
overflow | +0.084 |
wasteful | +0.084 |
structures | +0.082 |
ruining | +0.081 |
flood | +0.080 |
Neg logit contributions | |
name | -0.051 |
gown | -0.045 |
shirt | -0.044 |
ka | -0.044 |
ed | -0.044 |
Token | . |
Token ID | 13 |
Feature activation | +4.06e-01 |
Pos logit contributions | |
ka | +0.001 |
tie | +0.001 |
name | +0.001 |
ance | +0.001 |
dot | +0.001 |
Neg logit contributions | |
neither | -0.001 |
mess | -0.001 |
wasteful | -0.001 |
overflow | -0.001 |
fighting | -0.001 |
Token | not |
Token ID | 407 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
wasteful | +0.065 |
mess | +0.064 |
neither | +0.063 |
ruining | +0.058 |
overflow | +0.057 |
Neg logit contributions | |
name | -0.054 |
tie | -0.052 |
ka | -0.050 |
ance | -0.049 |
shirt | -0.049 |
Token | to |
Token ID | 284 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
neither | +0.063 |
mess | +0.061 |
wasteful | +0.060 |
roughly | +0.056 |
overflow | +0.056 |
Neg logit contributions | |
name | -0.034 |
tie | -0.031 |
aboard | -0.028 |
ka | -0.028 |
ance | -0.028 |
Token | fall |
Token ID | 2121 |
Feature activation | +1.51e+00 |
Pos logit contributions | |
neither | +0.092 |
wasteful | +0.091 |
mess | +0.090 |
structures | +0.086 |
unfortunately | +0.085 |
Neg logit contributions | |
name | -0.059 |
tie | -0.046 |
Crown | -0.043 |
dot | -0.043 |
ka | -0.042 |
Token | into |
Token ID | 656 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
wasteful | +0.062 |
neither | +0.061 |
mess | +0.061 |
roughly | +0.054 |
ruining | +0.054 |
Neg logit contributions | |
aboard | -0.055 |
ka | -0.055 |
tie | -0.054 |
name | -0.053 |
ance | -0.052 |
Token | the |
Token ID | 262 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
neither | +0.051 |
mess | +0.049 |
wasteful | +0.048 |
overflow | +0.045 |
roughly | +0.045 |
Neg logit contributions | |
name | -0.044 |
ka | -0.043 |
aboard | -0.043 |
ance | -0.042 |
powers | -0.042 |
Token | pond |
Token ID | 16723 |
Feature activation | +4.12e+00 |
Pos logit contributions | |
neither | +0.005 |
mess | +0.005 |
wasteful | +0.005 |
roughly | +0.004 |
overflow | +0.004 |
Neg logit contributions | |
ka | -0.006 |
name | -0.006 |
tie | -0.006 |
dot | -0.006 |
ance | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
Uh | +0.201 |
wasteful | +0.200 |
neither | +0.200 |
ruining | +0.193 |
mess | +0.189 |
Neg logit contributions | |
name | -0.092 |
tie | -0.090 |
aboard | -0.088 |
ance | -0.078 |
gown | -0.077 |
Token | It |
Token ID | 632 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
mess | +0.056 |
neither | +0.053 |
wasteful | +0.050 |
overflow | +0.048 |
ruining | +0.047 |
Neg logit contributions | |
name | -0.037 |
dot | -0.033 |
ance | -0.033 |
ka | -0.033 |
tie | -0.032 |
Token | 's |
Token ID | 338 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
mess | +0.091 |
wasteful | +0.090 |
neither | +0.090 |
roughly | +0.088 |
structures | +0.086 |
Neg logit contributions | |
name | -0.046 |
aboard | -0.039 |
ka | -0.038 |
ance | -0.038 |
shone | -0.037 |
Token | deep |
Token ID | 2769 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
neither | +0.030 |
mess | +0.028 |
structures | +0.026 |
wasteful | +0.025 |
overflow | +0.025 |
Neg logit contributions | |
name | -0.052 |
tie | -0.048 |
ka | -0.047 |
powers | -0.046 |
dot | -0.046 |
Token | and |
Token ID | 290 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
neither | +0.059 |
wasteful | +0.055 |
mess | +0.054 |
overflow | +0.050 |
ruining | +0.050 |
Neg logit contributions | |
ka | -0.047 |
name | -0.044 |
aboard | -0.044 |
tie | -0.044 |
powers | -0.042 |
Token | We |
Token ID | 1135 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
mess | +0.006 |
neither | +0.006 |
wasteful | +0.006 |
overflow | +0.005 |
ruining | +0.005 |
Neg logit contributions | |
ka | -0.007 |
tie | -0.007 |
name | -0.007 |
leaf | -0.007 |
dot | -0.007 |
Token | want |
Token ID | 765 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
ka | +0.010 |
tie | +0.010 |
ance | +0.010 |
name | +0.010 |
dot | +0.010 |
Neg logit contributions | |
neither | -0.010 |
mess | -0.010 |
wasteful | -0.009 |
fighting | -0.008 |
overflow | -0.008 |
Token | to |
Token ID | 284 |
Feature activation | +8.70e-01 |
Pos logit contributions | |
neither | +0.072 |
wasteful | +0.070 |
mess | +0.069 |
overflow | +0.067 |
roughly | +0.066 |
Neg logit contributions | |
name | -0.033 |
leaf | -0.031 |
powers | -0.030 |
aboard | -0.030 |
tie | -0.029 |
Token | eat |
Token ID | 4483 |
Feature activation | +2.07e+00 |
Pos logit contributions | |
neither | +0.042 |
wasteful | +0.041 |
mess | +0.040 |
fighting | +0.036 |
ruining | +0.036 |
Neg logit contributions | |
name | -0.030 |
tie | -0.027 |
ka | -0.026 |
dot | -0.026 |
ance | -0.025 |
Token | the |
Token ID | 262 |
Feature activation | +6.29e-01 |
Pos logit contributions | |
neither | +0.094 |
overflow | +0.091 |
wasteful | +0.089 |
forgot | +0.088 |
mess | +0.087 |
Neg logit contributions | |
name | -0.066 |
aboard | -0.065 |
tie | -0.059 |
ka | -0.059 |
shirt | -0.057 |
Token | banana |
Token ID | 25996 |
Feature activation | +4.09e+00 |
Pos logit contributions | |
neither | +0.025 |
wasteful | +0.023 |
mess | +0.023 |
overflow | +0.021 |
roughly | +0.021 |
Neg logit contributions | |
name | -0.026 |
ka | -0.025 |
tie | -0.025 |
dot | -0.024 |
shirt | -0.024 |
Token | , |
Token ID | 11 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
wasteful | +0.190 |
ruining | +0.184 |
overflow | +0.182 |
sudden | +0.179 |
neither | +0.179 |
Neg logit contributions | |
tie | -0.110 |
name | -0.110 |
aboard | -0.106 |
orn | -0.096 |
powers | -0.095 |
Token | but |
Token ID | 475 |
Feature activation | +1.47e+00 |
Pos logit contributions | |
wasteful | +0.066 |
neither | +0.066 |
mess | +0.066 |
roughly | +0.062 |
overflow | +0.060 |
Neg logit contributions | |
name | -0.080 |
tie | -0.067 |
aboard | -0.063 |
uit | -0.063 |
shirt | -0.063 |
Token | we |
Token ID | 356 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
mess | +0.052 |
overflow | +0.049 |
roughly | +0.049 |
neither | +0.048 |
wasteful | +0.047 |
Neg logit contributions | |
name | -0.062 |
aboard | -0.055 |
Crown | -0.054 |
tie | -0.053 |
Magic | -0.053 |
Token | can |
Token ID | 460 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
neither | +0.059 |
mess | +0.056 |
wasteful | +0.055 |
structures | +0.054 |
mosquitoes | +0.052 |
Neg logit contributions | |
name | -0.043 |
tie | -0.039 |
ka | -0.037 |
leaf | -0.036 |
ocket | -0.036 |
Token | 't |
Token ID | 470 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
wasteful | +0.095 |
mess | +0.088 |
neither | +0.087 |
rats | +0.085 |
mosquitoes | +0.083 |
Neg logit contributions | |
name | -0.080 |
tie | -0.065 |
ad | -0.061 |
Crown | -0.058 |
dot | -0.057 |
Token | won |
Token ID | 1839 |
Feature activation | +7.83e-01 |
Pos logit contributions | |
ka | +0.006 |
tie | +0.006 |
ance | +0.006 |
name | +0.006 |
dot | +0.006 |
Neg logit contributions | |
neither | -0.006 |
mess | -0.006 |
wasteful | -0.006 |
overflow | -0.005 |
fighting | -0.005 |
Token | 't |
Token ID | 470 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
neither | +0.028 |
mess | +0.028 |
wasteful | +0.028 |
overflow | +0.025 |
rats | +0.024 |
Neg logit contributions | |
name | -0.032 |
ka | -0.032 |
tie | -0.031 |
ance | -0.031 |
aboard | -0.031 |
Token | swim |
Token ID | 9422 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
neither | +0.058 |
wasteful | +0.055 |
rats | +0.052 |
mess | +0.051 |
structures | +0.051 |
Neg logit contributions | |
name | -0.049 |
tie | -0.046 |
ka | -0.043 |
leaf | -0.041 |
yo | -0.041 |
Token | in |
Token ID | 287 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
mess | +0.070 |
neither | +0.070 |
wasteful | +0.070 |
ruining | +0.066 |
mosquitoes | +0.064 |
Neg logit contributions | |
tie | -0.070 |
ka | -0.070 |
aboard | -0.067 |
name | -0.067 |
ance | -0.062 |
Token | the |
Token ID | 262 |
Feature activation | -3.12e-03 |
Pos logit contributions | |
neither | +0.090 |
wasteful | +0.086 |
overflow | +0.079 |
keys | +0.078 |
forgot | +0.078 |
Neg logit contributions | |
outer | -0.056 |
aboard | -0.056 |
powers | -0.055 |
ka | -0.054 |
name | -0.053 |
Token | pond |
Token ID | 16723 |
Feature activation | +4.06e+00 |
Pos logit contributions | |
ka | +0.000 |
tie | +0.000 |
name | +0.000 |
ance | +0.000 |
dot | +0.000 |
Neg logit contributions | |
neither | -0.000 |
mess | -0.000 |
wasteful | -0.000 |
overflow | -0.000 |
fighting | -0.000 |
Token | again |
Token ID | 757 |
Feature activation | +2.79e+00 |
Pos logit contributions | |
wasteful | +0.200 |
neither | +0.195 |
ruining | +0.193 |
Uh | +0.189 |
overflow | +0.186 |
Neg logit contributions | |
tie | -0.090 |
aboard | -0.086 |
name | -0.081 |
ti | -0.079 |
ka | -0.079 |
Token | . |
Token ID | 13 |
Feature activation | -2.33e-01 |
Pos logit contributions | |
neither | +0.139 |
wasteful | +0.138 |
overflow | +0.135 |
ruining | +0.129 |
mess | +0.127 |
Neg logit contributions | |
aboard | -0.066 |
name | -0.064 |
tie | -0.062 |
ka | -0.061 |
powers | -0.058 |
Token | I |
Token ID | 314 |
Feature activation | -6.11e-02 |
Pos logit contributions | |
ka | +0.010 |
tie | +0.010 |
ance | +0.009 |
name | +0.009 |
dot | +0.009 |
Neg logit contributions | |
neither | -0.009 |
mess | -0.008 |
wasteful | -0.008 |
fighting | -0.007 |
overflow | -0.007 |
Token | will |
Token ID | 481 |
Feature activation | +1.52e-01 |
Pos logit contributions | |
ka | +0.002 |
tie | +0.002 |
name | +0.002 |
ance | +0.002 |
dot | +0.002 |
Neg logit contributions | |
neither | -0.003 |
mess | -0.003 |
wasteful | -0.003 |
overflow | -0.002 |
fighting | -0.002 |
Token | be |
Token ID | 307 |
Feature activation | +6.86e-01 |
Pos logit contributions | |
neither | +0.006 |
mess | +0.006 |
wasteful | +0.006 |
overflow | +0.005 |
roughly | +0.005 |
Neg logit contributions | |
name | -0.006 |
ka | -0.006 |
tie | -0.006 |
ance | -0.006 |
dot | -0.006 |
Token | want |
Token ID | 765 |
Feature activation | +9.01e-01 |
Pos logit contributions | |
neither | +0.029 |
mess | +0.028 |
wasteful | +0.027 |
overflow | +0.025 |
ruining | +0.025 |
Neg logit contributions | |
name | -0.021 |
tie | -0.020 |
ka | -0.019 |
dot | -0.019 |
leaf | -0.019 |
Token | to |
Token ID | 284 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
mess | +0.047 |
neither | +0.047 |
wasteful | +0.046 |
overflow | +0.044 |
roughly | +0.043 |
Neg logit contributions | |
tie | -0.022 |
name | -0.022 |
ka | -0.022 |
aboard | -0.021 |
leaf | -0.021 |
Token | play |
Token ID | 711 |
Feature activation | +1.58e+00 |
Pos logit contributions | |
wasteful | +0.065 |
neither | +0.064 |
mess | +0.061 |
rats | +0.058 |
mosquitoes | +0.058 |
Neg logit contributions | |
name | -0.044 |
ka | -0.036 |
tie | -0.036 |
aboard | -0.034 |
dot | -0.034 |
Token | in |
Token ID | 287 |
Feature activation | +2.11e+00 |
Pos logit contributions | |
neither | +0.074 |
wasteful | +0.070 |
mess | +0.067 |
overflow | +0.066 |
forgot | +0.065 |
Neg logit contributions | |
name | -0.051 |
tie | -0.047 |
aboard | -0.047 |
ka | -0.045 |
ti | -0.044 |
Token | the |
Token ID | 262 |
Feature activation | +2.50e-01 |
Pos logit contributions | |
neither | +0.106 |
wasteful | +0.106 |
forgot | +0.099 |
overflow | +0.096 |
mess | +0.094 |
Neg logit contributions | |
outer | -0.055 |
aboard | -0.055 |
powers | -0.047 |
Magic | -0.047 |
name | -0.046 |
Token | barrel |
Token ID | 9036 |
Feature activation | +4.02e+00 |
Pos logit contributions | |
neither | +0.011 |
mess | +0.010 |
wasteful | +0.010 |
overflow | +0.009 |
ruining | +0.009 |
Neg logit contributions | |
name | -0.009 |
ka | -0.009 |
tie | -0.009 |
dot | -0.009 |
aboard | -0.009 |
Token | anymore |
Token ID | 7471 |
Feature activation | +2.86e+00 |
Pos logit contributions | |
wasteful | +0.227 |
sudden | +0.215 |
Uh | +0.210 |
neither | +0.209 |
mosquitoes | +0.209 |
Neg logit contributions | |
tie | -0.074 |
aboard | -0.072 |
ane | -0.060 |
name | -0.059 |
ance | -0.058 |
Token | . |
Token ID | 13 |
Feature activation | +3.12e-01 |
Pos logit contributions | |
wasteful | +0.158 |
overflow | +0.155 |
neither | +0.147 |
mess | +0.146 |
ruining | +0.141 |
Neg logit contributions | |
name | -0.064 |
aboard | -0.060 |
powers | -0.055 |
ance | -0.055 |
tie | -0.053 |
Token | It |
Token ID | 632 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
mess | +0.014 |
neither | +0.013 |
wasteful | +0.012 |
overflow | +0.012 |
ruining | +0.011 |
Neg logit contributions | |
name | -0.011 |
ka | -0.011 |
tie | -0.011 |
dot | -0.011 |
ance | -0.011 |
Token | is |
Token ID | 318 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
neither | +0.091 |
wasteful | +0.088 |
roughly | +0.085 |
overflow | +0.084 |
mess | +0.084 |
Neg logit contributions | |
name | -0.033 |
aboard | -0.029 |
ka | -0.028 |
ance | -0.026 |
shone | -0.025 |
Token | too |
Token ID | 1165 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
neither | +0.046 |
mess | +0.040 |
wasteful | +0.039 |
overflow | +0.038 |
structures | +0.037 |
Neg logit contributions | |
name | -0.056 |
tie | -0.051 |
ka | -0.051 |
aboard | -0.050 |
powers | -0.049 |
Token | 't |
Token ID | 470 |
Feature activation | -2.38e-01 |
Pos logit contributions | |
mess | +0.027 |
neither | +0.027 |
wasteful | +0.026 |
overflow | +0.023 |
roughly | +0.022 |
Neg logit contributions | |
name | -0.032 |
tie | -0.032 |
ance | -0.031 |
ka | -0.031 |
dot | -0.031 |
Token | want |
Token ID | 765 |
Feature activation | +3.98e-01 |
Pos logit contributions | |
ka | +0.009 |
tie | +0.009 |
ance | +0.009 |
name | +0.009 |
dot | +0.009 |
Neg logit contributions | |
neither | -0.010 |
mess | -0.010 |
wasteful | -0.009 |
overflow | -0.008 |
fighting | -0.008 |
Token | to |
Token ID | 284 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
neither | +0.019 |
mess | +0.019 |
wasteful | +0.018 |
overflow | +0.017 |
roughly | +0.017 |
Neg logit contributions | |
tie | -0.012 |
ka | -0.012 |
name | -0.012 |
dot | -0.011 |
ance | -0.011 |
Token | eat |
Token ID | 4483 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
wasteful | +0.059 |
neither | +0.059 |
mess | +0.055 |
rats | +0.052 |
structures | +0.052 |
Neg logit contributions | |
name | -0.046 |
tie | -0.040 |
ka | -0.038 |
dot | -0.037 |
Crown | -0.036 |
Token | the |
Token ID | 262 |
Feature activation | +8.33e-01 |
Pos logit contributions | |
neither | +0.079 |
wasteful | +0.075 |
overflow | +0.074 |
forgot | +0.072 |
mess | +0.071 |
Neg logit contributions | |
name | -0.059 |
aboard | -0.057 |
ka | -0.056 |
tie | -0.055 |
ance | -0.054 |
Token | pit |
Token ID | 6028 |
Feature activation | +4.00e+00 |
Pos logit contributions | |
neither | +0.032 |
wasteful | +0.029 |
mess | +0.027 |
roughly | +0.026 |
overflow | +0.026 |
Neg logit contributions | |
name | -0.036 |
ka | -0.034 |
tie | -0.034 |
aboard | -0.033 |
Crown | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | -1.52e-01 |
Pos logit contributions | |
wasteful | +0.186 |
Sadly | +0.183 |
ruining | +0.180 |
sudden | +0.175 |
overflow | +0.175 |
Neg logit contributions | |
aboard | -0.109 |
tie | -0.108 |
ka | -0.106 |
iest | -0.096 |
ed | -0.091 |
Token | I |
Token ID | 314 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
ka | +0.005 |
tie | +0.004 |
name | +0.004 |
ance | +0.004 |
dot | +0.004 |
Neg logit contributions | |
neither | -0.008 |
mess | -0.007 |
wasteful | -0.007 |
overflow | -0.007 |
fighting | -0.007 |
Token | only |
Token ID | 691 |
Feature activation | +2.08e-02 |
Pos logit contributions | |
ka | +0.005 |
tie | +0.005 |
name | +0.005 |
ance | +0.005 |
dot | +0.005 |
Neg logit contributions | |
neither | -0.004 |
mess | -0.004 |
wasteful | -0.004 |
overflow | -0.004 |
fighting | -0.004 |
Token | want |
Token ID | 765 |
Feature activation | +8.16e-01 |
Pos logit contributions | |
neither | +0.001 |
mess | +0.001 |
wasteful | +0.001 |
overflow | +0.001 |
fighting | +0.001 |
Neg logit contributions | |
ka | -0.001 |
tie | -0.001 |
name | -0.001 |
ance | -0.001 |
dot | -0.001 |
Token | to |
Token ID | 284 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
neither | +0.041 |
mess | +0.040 |
wasteful | +0.039 |
overflow | +0.037 |
roughly | +0.036 |
Neg logit contributions | |
name | -0.023 |
tie | -0.022 |
ka | -0.022 |
aboard | -0.022 |
powers | -0.021 |
Token | you |
Token ID | 345 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
mess | +0.067 |
wasteful | +0.066 |
neither | +0.066 |
overflow | +0.064 |
roughly | +0.059 |
Neg logit contributions | |
ance | -0.079 |
name | -0.076 |
ka | -0.075 |
Magic | -0.072 |
aboard | -0.072 |
Token | to |
Token ID | 284 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
mess | +0.042 |
neither | +0.041 |
wasteful | +0.039 |
overflow | +0.036 |
ruining | +0.035 |
Neg logit contributions | |
name | -0.047 |
tie | -0.046 |
ance | -0.046 |
ka | -0.044 |
dot | -0.044 |
Token | clean |
Token ID | 3424 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
neither | +0.055 |
wasteful | +0.052 |
mess | +0.051 |
structures | +0.047 |
rats | +0.047 |
Neg logit contributions | |
name | -0.046 |
tie | -0.041 |
ance | -0.041 |
ka | -0.040 |
shirt | -0.039 |
Token | up |
Token ID | 510 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
neither | +0.059 |
wasteful | +0.055 |
mess | +0.052 |
overflow | +0.052 |
erupted | +0.050 |
Neg logit contributions | |
ka | -0.039 |
name | -0.038 |
aboard | -0.038 |
tie | -0.036 |
shirt | -0.036 |
Token | your |
Token ID | 534 |
Feature activation | +1.27e+00 |
Pos logit contributions | |
neither | +0.082 |
wasteful | +0.076 |
mess | +0.075 |
overflow | +0.073 |
quarrel | +0.070 |
Neg logit contributions | |
aboard | -0.066 |
ance | -0.066 |
ka | -0.064 |
name | -0.064 |
shirt | -0.062 |
Token | room |
Token ID | 2119 |
Feature activation | +3.99e+00 |
Pos logit contributions | |
neither | +0.051 |
wasteful | +0.044 |
unfortunately | +0.038 |
overflow | +0.038 |
mosquitoes | +0.037 |
Neg logit contributions | |
name | -0.057 |
powers | -0.051 |
ka | -0.051 |
gown | -0.048 |
anger | -0.047 |
Token | ! |
Token ID | 0 |
Feature activation | +2.98e-01 |
Pos logit contributions | |
mosquitoes | +0.173 |
neither | +0.172 |
Uh | +0.168 |
wasteful | +0.168 |
snakes | +0.163 |
Neg logit contributions | |
ka | -0.111 |
tie | -0.110 |
powers | -0.109 |
name | -0.109 |
ance | -0.108 |
Token | Now |
Token ID | 2735 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
mess | +0.013 |
neither | +0.013 |
wasteful | +0.012 |
overflow | +0.011 |
ruining | +0.011 |
Neg logit contributions | |
name | -0.011 |
ka | -0.010 |
tie | -0.010 |
ance | -0.010 |
dot | -0.010 |
Token | , |
Token ID | 11 |
Feature activation | +8.32e-01 |
Pos logit contributions | |
wasteful | +0.042 |
neither | +0.042 |
mess | +0.041 |
overflow | +0.038 |
roughly | +0.037 |
Neg logit contributions | |
name | -0.049 |
ance | -0.047 |
aboard | -0.046 |
tie | -0.045 |
ka | -0.044 |
Token | I |
Token ID | 314 |
Feature activation | -8.33e-02 |
Pos logit contributions | |
neither | +0.030 |
mess | +0.030 |
wasteful | +0.029 |
overflow | +0.026 |
roughly | +0.026 |
Neg logit contributions | |
name | -0.036 |
tie | -0.033 |
ance | -0.033 |
aboard | -0.033 |
ka | -0.032 |
Token | have |
Token ID | 423 |
Feature activation | +6.10e-01 |
Pos logit contributions | |
ka | +0.003 |
tie | +0.003 |
name | +0.003 |
ance | +0.003 |
dot | +0.003 |
Neg logit contributions | |
neither | -0.004 |
mess | -0.004 |
wasteful | -0.003 |
overflow | -0.003 |
fighting | -0.003 |
Token | 's |
Token ID | 338 |
Feature activation | +5.65e-01 |
Pos logit contributions | |
wasteful | +0.035 |
neither | +0.035 |
mess | +0.033 |
roughly | +0.032 |
overflow | +0.031 |
Neg logit contributions | |
name | -0.036 |
ance | -0.033 |
ka | -0.033 |
aboard | -0.032 |
tie | -0.032 |
Token | not |
Token ID | 407 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
neither | +0.027 |
mess | +0.026 |
wasteful | +0.026 |
ruining | +0.023 |
roughly | +0.023 |
Neg logit contributions | |
name | -0.019 |
tie | -0.017 |
ka | -0.017 |
ance | -0.017 |
aboard | -0.016 |
Token | play |
Token ID | 711 |
Feature activation | +1.47e+00 |
Pos logit contributions | |
wasteful | +0.061 |
neither | +0.058 |
mess | +0.056 |
floor | +0.053 |
structures | +0.052 |
Neg logit contributions | |
name | -0.041 |
aboard | -0.034 |
tie | -0.033 |
ance | -0.032 |
ka | -0.031 |
Token | in |
Token ID | 287 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
neither | +0.070 |
wasteful | +0.067 |
mess | +0.063 |
overflow | +0.060 |
anth | +0.060 |
Neg logit contributions | |
name | -0.046 |
aboard | -0.043 |
powers | -0.041 |
tie | -0.040 |
ka | -0.039 |
Token | the |
Token ID | 262 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
wasteful | +0.104 |
neither | +0.103 |
forgot | +0.095 |
overflow | +0.093 |
mess | +0.091 |
Neg logit contributions | |
aboard | -0.057 |
outer | -0.053 |
powers | -0.048 |
name | -0.046 |
ance | -0.045 |
Token | forest |
Token ID | 8222 |
Feature activation | +3.97e+00 |
Pos logit contributions | |
neither | +0.046 |
wasteful | +0.039 |
mess | +0.034 |
fighting | +0.034 |
roughly | +0.034 |
Neg logit contributions | |
name | -0.048 |
orn | -0.043 |
cloak | -0.042 |
tie | -0.041 |
amulet | -0.040 |
Token | again |
Token ID | 757 |
Feature activation | +2.99e+00 |
Pos logit contributions | |
wasteful | +0.189 |
neither | +0.185 |
ruining | +0.175 |
overflow | +0.175 |
spitting | +0.175 |
Neg logit contributions | |
aboard | -0.103 |
tie | -0.086 |
ka | -0.085 |
powers | -0.085 |
ti | -0.083 |
Token | ," |
Token ID | 553 |
Feature activation | -5.81e-01 |
Pos logit contributions | |
wasteful | +0.146 |
neither | +0.143 |
overflow | +0.141 |
ruining | +0.135 |
mess | +0.132 |
Neg logit contributions | |
aboard | -0.087 |
name | -0.079 |
ka | -0.075 |
powers | -0.069 |
outer | -0.067 |
Token | L |
Token ID | 406 |
Feature activation | -9.14e-01 |
Pos logit contributions | |
tie | +0.027 |
ka | +0.027 |
ance | +0.026 |
dot | +0.026 |
shirt | +0.026 |
Neg logit contributions | |
neither | -0.021 |
mess | -0.019 |
wasteful | -0.018 |
fighting | -0.016 |
overflow | -0.016 |
Token | ila |
Token ID | 10102 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
tie | +0.037 |
name | +0.036 |
ka | +0.035 |
shirt | +0.035 |
amulet | +0.035 |
Neg logit contributions | |
neither | -0.037 |
mess | -0.035 |
wasteful | -0.033 |
fighting | -0.031 |
mosquitoes | -0.030 |
Token | said |
Token ID | 531 |
Feature activation | -7.00e-01 |
Pos logit contributions | |
mess | +0.047 |
overflow | +0.046 |
wasteful | +0.046 |
neither | +0.045 |
ruining | +0.044 |
Neg logit contributions | |
tie | -0.041 |
name | -0.040 |
ka | -0.039 |
leaf | -0.038 |
shirt | -0.038 |
Token | Let |
Token ID | 5756 |
Feature activation | +9.11e-01 |
Pos logit contributions | |
neither | +0.010 |
mess | +0.010 |
wasteful | +0.009 |
ruining | +0.008 |
overflow | +0.008 |
Neg logit contributions | |
tie | -0.008 |
ka | -0.008 |
name | -0.008 |
dot | -0.008 |
shirt | -0.008 |
Token | 's |
Token ID | 338 |
Feature activation | +8.76e-01 |
Pos logit contributions | |
neither | +0.037 |
wasteful | +0.037 |
mess | +0.035 |
roughly | +0.034 |
ruining | +0.033 |
Neg logit contributions | |
name | -0.035 |
ka | -0.032 |
ance | -0.032 |
aboard | -0.032 |
tie | -0.032 |
Token | go |
Token ID | 467 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
wasteful | +0.042 |
neither | +0.042 |
mess | +0.042 |
mosquitoes | +0.038 |
rats | +0.037 |
Neg logit contributions | |
name | -0.030 |
tie | -0.025 |
aboard | -0.024 |
ka | -0.024 |
ance | -0.024 |
Token | down |
Token ID | 866 |
Feature activation | +1.37e+00 |
Pos logit contributions | |
ka | +0.065 |
anced | +0.064 |
ance | +0.064 |
tie | +0.063 |
ocket | +0.063 |
Neg logit contributions | |
mess | -0.064 |
neither | -0.062 |
fighting | -0.056 |
overflow | -0.055 |
wasteful | -0.051 |
Token | the |
Token ID | 262 |
Feature activation | +1.79e+00 |
Pos logit contributions | |
wasteful | +0.062 |
neither | +0.061 |
mess | +0.060 |
overflow | +0.054 |
ruining | +0.053 |
Neg logit contributions | |
aboard | -0.049 |
ka | -0.043 |
ance | -0.043 |
name | -0.043 |
tie | -0.041 |
Token | pit |
Token ID | 6028 |
Feature activation | +3.96e+00 |
Pos logit contributions | |
neither | +0.077 |
wasteful | +0.074 |
roughly | +0.063 |
unfortunately | +0.062 |
mosquitoes | +0.061 |
Neg logit contributions | |
name | -0.070 |
aboard | -0.064 |
orn | -0.059 |
anger | -0.058 |
amulet | -0.057 |
Token | ," |
Token ID | 553 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
wasteful | +0.202 |
mosquitoes | +0.192 |
ruining | +0.184 |
neither | +0.183 |
spitting | +0.182 |
Neg logit contributions | |
aboard | -0.120 |
tie | -0.098 |
name | -0.095 |
ka | -0.094 |
anger | -0.088 |
Token | Ben |
Token ID | 3932 |
Feature activation | +9.20e-01 |
Pos logit contributions | |
mess | +0.055 |
wasteful | +0.052 |
neither | +0.050 |
ruining | +0.050 |
overflow | +0.050 |
Neg logit contributions | |
name | -0.062 |
ka | -0.060 |
shirt | -0.056 |
dot | -0.054 |
Magic | -0.054 |
Token | said |
Token ID | 531 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
neither | +0.036 |
mess | +0.036 |
wasteful | +0.036 |
ruining | +0.034 |
overflow | +0.033 |
Neg logit contributions | |
name | -0.036 |
ka | -0.035 |
tie | -0.035 |
shirt | -0.034 |
dot | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | +2.94e-01 |
Pos logit contributions | |
ka | +0.005 |
tie | +0.005 |
name | +0.004 |
ance | +0.004 |
dot | +0.004 |
Neg logit contributions | |
neither | -0.005 |
mess | -0.005 |
wasteful | -0.005 |
overflow | -0.004 |
fighting | -0.004 |
Token | " |
Token ID | 366 |
Feature activation | +2.04e-01 |
Pos logit contributions | |
mess | +0.013 |
neither | +0.013 |
wasteful | +0.012 |
ruining | +0.011 |
overflow | +0.011 |
Neg logit contributions | |
name | -0.011 |
ka | -0.010 |
tie | -0.010 |
dot | -0.010 |
shirt | -0.010 |
Token | sit |
Token ID | 1650 |
Feature activation | +9.46e-01 |
Pos logit contributions | |
wasteful | +0.055 |
neither | +0.054 |
mess | +0.051 |
roughly | +0.049 |
structures | +0.048 |
Neg logit contributions | |
name | -0.034 |
dot | -0.029 |
ance | -0.028 |
tie | -0.028 |
ka | -0.028 |
Token | in |
Token ID | 287 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
neither | +0.046 |
mess | +0.045 |
wasteful | +0.044 |
overflow | +0.042 |
structures | +0.042 |
Neg logit contributions | |
ka | -0.028 |
tie | -0.027 |
name | -0.027 |
aboard | -0.026 |
ance | -0.025 |
Token | the |
Token ID | 262 |
Feature activation | -2.04e-01 |
Pos logit contributions | |
wasteful | +0.089 |
neither | +0.088 |
forgot | +0.080 |
overflow | +0.080 |
keys | +0.079 |
Neg logit contributions | |
aboard | -0.052 |
outer | -0.048 |
ka | -0.047 |
name | -0.045 |
ance | -0.043 |
Token | old |
Token ID | 1468 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
ka | +0.008 |
tie | +0.008 |
name | +0.008 |
ance | +0.008 |
dot | +0.008 |
Neg logit contributions | |
mess | -0.008 |
neither | -0.008 |
wasteful | -0.007 |
overflow | -0.007 |
fighting | -0.007 |
Token | arm |
Token ID | 3211 |
Feature activation | +2.46e+00 |
Pos logit contributions | |
neither | +0.062 |
wasteful | +0.058 |
overflow | +0.054 |
ruining | +0.052 |
mess | +0.050 |
Neg logit contributions | |
name | -0.051 |
gown | -0.050 |
ka | -0.049 |
tie | -0.048 |
shirt | -0.048 |
Token | chair |
Token ID | 16337 |
Feature activation | +3.96e+00 |
Pos logit contributions | |
neither | +0.126 |
wasteful | +0.122 |
overflow | +0.115 |
ruining | +0.115 |
mess | +0.112 |
Neg logit contributions | |
tie | -0.066 |
shirt | -0.065 |
ance | -0.064 |
gown | -0.062 |
anger | -0.061 |
Token | . |
Token ID | 13 |
Feature activation | +5.31e-01 |
Pos logit contributions | |
wasteful | +0.198 |
anth | +0.188 |
structures | +0.188 |
neither | +0.188 |
overflow | +0.187 |
Neg logit contributions | |
tie | -0.094 |
aboard | -0.086 |
ance | -0.083 |
gown | -0.079 |
ka | -0.078 |
Token | It |
Token ID | 632 |
Feature activation | +9.89e-01 |
Pos logit contributions | |
mess | +0.021 |
neither | +0.020 |
wasteful | +0.018 |
ruining | +0.017 |
overflow | +0.017 |
Neg logit contributions | |
name | -0.023 |
tie | -0.021 |
ka | -0.021 |
dot | -0.021 |
ance | -0.021 |
Token | is |
Token ID | 318 |
Feature activation | +1.41e+00 |
Pos logit contributions | |
neither | +0.048 |
wasteful | +0.046 |
mess | +0.045 |
roughly | +0.043 |
overflow | +0.043 |
Neg logit contributions | |
name | -0.029 |
ka | -0.028 |
tie | -0.026 |
aboard | -0.026 |
ance | -0.026 |
Token | wet |
Token ID | 9583 |
Feature activation | +2.14e+00 |
Pos logit contributions | |
neither | +0.049 |
mess | +0.042 |
structures | +0.042 |
overflow | +0.042 |
wasteful | +0.042 |
Neg logit contributions | |
name | -0.061 |
ka | -0.056 |
tie | -0.055 |
aboard | -0.055 |
dot | -0.053 |
Token | and |
Token ID | 290 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
neither | +0.110 |
wasteful | +0.107 |
anth | +0.098 |
overflow | +0.097 |
ruining | +0.097 |
Neg logit contributions | |
tie | -0.056 |
ka | -0.055 |
aboard | -0.052 |
name | -0.048 |
ance | -0.047 |
Token | children |
Token ID | 1751 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
neither | +0.038 |
wasteful | +0.037 |
overflow | +0.037 |
mess | +0.036 |
roughly | +0.034 |
Neg logit contributions | |
ka | -0.053 |
name | -0.053 |
ance | -0.050 |
aboard | -0.050 |
powers | -0.048 |
Token | ! |
Token ID | 0 |
Feature activation | +3.09e-01 |
Pos logit contributions | |
overflow | +0.092 |
neither | +0.088 |
wasteful | +0.086 |
structures | +0.083 |
anth | +0.083 |
Neg logit contributions | |
ka | -0.053 |
name | -0.053 |
tie | -0.051 |
aboard | -0.049 |
leaf | -0.048 |
Token | You |
Token ID | 921 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
neither | +0.014 |
mess | +0.014 |
wasteful | +0.013 |
overflow | +0.012 |
ruining | +0.012 |
Neg logit contributions | |
ka | -0.011 |
name | -0.011 |
ance | -0.010 |
dot | -0.010 |
tie | -0.010 |
Token | ruined |
Token ID | 20484 |
Feature activation | +2.80e+00 |
Pos logit contributions | |
neither | +0.052 |
mess | +0.051 |
wasteful | +0.047 |
overflow | +0.047 |
structures | +0.046 |
Neg logit contributions | |
name | -0.043 |
leaf | -0.042 |
ka | -0.042 |
ance | -0.041 |
tie | -0.041 |
Token | the |
Token ID | 262 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
overflow | +0.128 |
neither | +0.127 |
forgot | +0.123 |
erupted | +0.122 |
Uh | +0.122 |
Neg logit contributions | |
name | -0.075 |
ka | -0.072 |
aboard | -0.071 |
ance | -0.066 |
Magic | -0.064 |
Token | cake |
Token ID | 12187 |
Feature activation | +3.94e+00 |
Pos logit contributions | |
neither | +0.046 |
wasteful | +0.039 |
roughly | +0.037 |
overflow | +0.036 |
ruining | +0.035 |
Neg logit contributions | |
name | -0.059 |
ka | -0.054 |
tie | -0.052 |
aboard | -0.052 |
Crown | -0.052 |
Token | ! |
Token ID | 0 |
Feature activation | +2.18e-01 |
Pos logit contributions | |
wasteful | +0.213 |
neither | +0.210 |
Uh | +0.199 |
overflow | +0.197 |
mosquitoes | +0.196 |
Neg logit contributions | |
name | -0.069 |
tie | -0.063 |
aboard | -0.063 |
ance | -0.057 |
anger | -0.056 |
Token | You |
Token ID | 921 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
neither | +0.010 |
mess | +0.010 |
wasteful | +0.009 |
overflow | +0.009 |
ruining | +0.008 |
Neg logit contributions | |
ka | -0.008 |
name | -0.007 |
tie | -0.007 |
ance | -0.007 |
dot | -0.007 |
Token | have |
Token ID | 423 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
neither | +0.057 |
mess | +0.055 |
wasteful | +0.052 |
structures | +0.051 |
overflow | +0.051 |
Neg logit contributions | |
name | -0.047 |
leaf | -0.046 |
ka | -0.046 |
tie | -0.046 |
ance | -0.045 |
Token | to |
Token ID | 284 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
neither | +0.052 |
mess | +0.051 |
overflow | +0.050 |
wasteful | +0.048 |
roughly | +0.048 |
Neg logit contributions | |
name | -0.029 |
ka | -0.029 |
tie | -0.028 |
ance | -0.028 |
leaf | -0.028 |
Token | say |
Token ID | 910 |
Feature activation | +9.86e-01 |
Pos logit contributions | |
neither | +0.048 |
mess | +0.042 |
wasteful | +0.042 |
rats | +0.041 |
snakes | +0.040 |
Neg logit contributions | |
name | -0.041 |
ance | -0.037 |
ka | -0.036 |
ad | -0.035 |
leaf | -0.035 |
Token | should |
Token ID | 815 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
mess | +0.042 |
neither | +0.042 |
wasteful | +0.040 |
overflow | +0.038 |
structures | +0.038 |
Neg logit contributions | |
name | -0.027 |
tie | -0.025 |
ance | -0.024 |
ka | -0.024 |
leaf | -0.023 |
Token | not |
Token ID | 407 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
neither | +0.044 |
mess | +0.041 |
wasteful | +0.041 |
rats | +0.038 |
structures | +0.038 |
Neg logit contributions | |
name | -0.044 |
tie | -0.038 |
ance | -0.037 |
ka | -0.037 |
dot | -0.036 |
Token | fight |
Token ID | 1907 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
neither | +0.094 |
wasteful | +0.090 |
mess | +0.085 |
mosquitoes | +0.083 |
structures | +0.082 |
Neg logit contributions | |
name | -0.071 |
tie | -0.058 |
ad | -0.055 |
orn | -0.054 |
shirt | -0.053 |
Token | over |
Token ID | 625 |
Feature activation | +2.27e+00 |
Pos logit contributions | |
overflow | +0.084 |
neither | +0.084 |
wasteful | +0.082 |
mess | +0.082 |
flood | +0.080 |
Neg logit contributions | |
name | -0.058 |
aboard | -0.053 |
ka | -0.052 |
tie | -0.051 |
ance | -0.050 |
Token | the |
Token ID | 262 |
Feature activation | +1.79e+00 |
Pos logit contributions | |
overflow | +0.100 |
neither | +0.098 |
mess | +0.097 |
wasteful | +0.095 |
timer | +0.091 |
Neg logit contributions | |
name | -0.064 |
aboard | -0.063 |
tie | -0.059 |
powers | -0.059 |
Crown | -0.059 |
Token | banana |
Token ID | 25996 |
Feature activation | +3.94e+00 |
Pos logit contributions | |
neither | +0.081 |
wasteful | +0.072 |
overflow | +0.071 |
roughly | +0.068 |
oving | +0.065 |
Neg logit contributions | |
name | -0.067 |
Crown | -0.060 |
orn | -0.057 |
tie | -0.057 |
ka | -0.057 |
Token | . |
Token ID | 13 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
wasteful | +0.191 |
sudden | +0.189 |
overflow | +0.189 |
neither | +0.184 |
Uh | +0.178 |
Neg logit contributions | |
anger | -0.087 |
orn | -0.085 |
ance | -0.085 |
name | -0.085 |
tie | -0.084 |
Token | You |
Token ID | 921 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
mess | +0.008 |
neither | +0.008 |
wasteful | +0.008 |
overflow | +0.007 |
ruining | +0.007 |
Neg logit contributions | |
name | -0.005 |
ka | -0.005 |
tie | -0.005 |
ance | -0.005 |
dot | -0.005 |
Token | should |
Token ID | 815 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
mess | +0.058 |
neither | +0.056 |
wasteful | +0.054 |
overflow | +0.052 |
structures | +0.052 |
Neg logit contributions | |
name | -0.036 |
tie | -0.034 |
ance | -0.032 |
ka | -0.032 |
leaf | -0.032 |
Token | share |
Token ID | 2648 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
neither | +0.048 |
wasteful | +0.044 |
mess | +0.044 |
rats | +0.042 |
overflow | +0.041 |
Neg logit contributions | |
name | -0.045 |
tie | -0.039 |
ance | -0.038 |
ka | -0.038 |
dot | -0.037 |
Token | and |
Token ID | 290 |
Feature activation | +2.01e+00 |
Pos logit contributions | |
neither | +0.101 |
overflow | +0.099 |
flood | +0.094 |
wasteful | +0.094 |
erupted | +0.090 |
Neg logit contributions | |
name | -0.058 |
tie | -0.051 |
aboard | -0.048 |
Crown | -0.047 |
ka | -0.046 |
Token | No |
Token ID | 2949 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
mess | +0.040 |
neither | +0.036 |
wasteful | +0.036 |
structures | +0.034 |
ruining | +0.034 |
Neg logit contributions | |
leaf | -0.033 |
tie | -0.032 |
name | -0.032 |
shirt | -0.032 |
ka | -0.031 |
Token | , |
Token ID | 11 |
Feature activation | +1.60e+00 |
Pos logit contributions | |
wasteful | +0.051 |
neither | +0.049 |
overflow | +0.047 |
mess | +0.047 |
roughly | +0.047 |
Neg logit contributions | |
name | -0.032 |
powers | -0.029 |
ka | -0.029 |
ance | -0.029 |
aboard | -0.028 |
Token | we |
Token ID | 356 |
Feature activation | +7.64e-01 |
Pos logit contributions | |
mess | +0.053 |
roughly | +0.051 |
wasteful | +0.051 |
oving | +0.049 |
structures | +0.048 |
Neg logit contributions | |
name | -0.068 |
aboard | -0.061 |
ocket | -0.061 |
ance | -0.060 |
hello | -0.060 |
Token | cannot |
Token ID | 2314 |
Feature activation | -9.90e-02 |
Pos logit contributions | |
neither | +0.040 |
mess | +0.040 |
wasteful | +0.039 |
structures | +0.037 |
overflow | +0.037 |
Neg logit contributions | |
name | -0.021 |
tie | -0.018 |
ka | -0.017 |
dot | -0.016 |
ance | -0.016 |
Token | go |
Token ID | 467 |
Feature activation | +8.38e-01 |
Pos logit contributions | |
ka | +0.004 |
tie | +0.004 |
name | +0.004 |
ance | +0.004 |
dot | +0.004 |
Neg logit contributions | |
neither | -0.004 |
mess | -0.004 |
wasteful | -0.004 |
overflow | -0.003 |
fighting | -0.003 |
Token | outside |
Token ID | 2354 |
Feature activation | +3.93e+00 |
Pos logit contributions | |
wasteful | +0.043 |
neither | +0.043 |
mess | +0.043 |
mosquitoes | +0.039 |
roughly | +0.039 |
Neg logit contributions | |
aboard | -0.023 |
name | -0.022 |
ka | -0.021 |
tie | -0.021 |
dot | -0.020 |
Token | ," |
Token ID | 553 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
wasteful | +0.206 |
neither | +0.188 |
Uh | +0.178 |
ruining | +0.176 |
mess | +0.176 |
Neg logit contributions | |
aboard | -0.103 |
name | -0.088 |
ti | -0.079 |
ka | -0.078 |
gown | -0.077 |
Token | Mia |
Token ID | 32189 |
Feature activation | +8.79e-01 |
Pos logit contributions | |
mess | +0.037 |
wasteful | +0.035 |
neither | +0.035 |
ruining | +0.033 |
oving | +0.031 |
Neg logit contributions | |
name | -0.055 |
ka | -0.048 |
aboard | -0.048 |
dot | -0.047 |
Magic | -0.047 |
Token | says |
Token ID | 1139 |
Feature activation | -3.67e-01 |
Pos logit contributions | |
neither | +0.037 |
wasteful | +0.035 |
mess | +0.035 |
overflow | +0.034 |
ruining | +0.033 |
Neg logit contributions | |
name | -0.034 |
ka | -0.031 |
ance | -0.031 |
tie | -0.030 |
leaf | -0.029 |
Token | . |
Token ID | 13 |
Feature activation | +5.73e-01 |
Pos logit contributions | |
ka | +0.015 |
tie | +0.015 |
name | +0.014 |
ance | +0.014 |
dot | +0.014 |
Neg logit contributions | |
neither | -0.015 |
mess | -0.014 |
wasteful | -0.013 |
fighting | -0.012 |
overflow | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | +3.77e-01 |
Pos logit contributions | |
mess | +0.027 |
neither | +0.027 |
wasteful | +0.026 |
overflow | +0.024 |
ruining | +0.024 |
Neg logit contributions | |
name | -0.019 |
tie | -0.018 |
dot | -0.017 |
aboard | -0.017 |
ance | -0.017 |
Token | We |
Token ID | 1135 |
Feature activation | +5.74e-02 |
Pos logit contributions | |
neither | +0.015 |
mess | +0.015 |
wasteful | +0.014 |
ruining | +0.013 |
overflow | +0.013 |
Neg logit contributions | |
tie | -0.013 |
ka | -0.013 |
name | -0.013 |
dot | -0.013 |
leaf | -0.012 |
Token | should |
Token ID | 815 |
Feature activation | +5.39e-01 |
Pos logit contributions | |
neither | +0.003 |
mess | +0.002 |
wasteful | +0.002 |
overflow | +0.002 |
ruining | +0.002 |
Neg logit contributions | |
ka | -0.002 |
tie | -0.002 |
name | -0.002 |
ance | -0.002 |
dot | -0.002 |
Token | not |
Token ID | 407 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
neither | +0.021 |
mess | +0.021 |
wasteful | +0.020 |
overflow | +0.018 |
roughly | +0.018 |
Neg logit contributions | |
name | -0.022 |
tie | -0.021 |
ka | -0.020 |
ance | -0.020 |
dot | -0.020 |
Token | bother |
Token ID | 11393 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
neither | +0.048 |
wasteful | +0.047 |
mess | +0.046 |
rats | +0.042 |
structures | +0.042 |
Neg logit contributions | |
name | -0.039 |
tie | -0.035 |
dot | -0.032 |
ance | -0.032 |
ka | -0.031 |
Token | the |
Token ID | 262 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
mess | +0.072 |
wasteful | +0.066 |
neither | +0.064 |
overflow | +0.063 |
roughly | +0.059 |
Neg logit contributions | |
aboard | -0.075 |
name | -0.073 |
ance | -0.070 |
tie | -0.068 |
ka | -0.067 |
Token | frogs |
Token ID | 37475 |
Feature activation | +3.91e+00 |
Pos logit contributions | |
neither | +0.048 |
wasteful | +0.047 |
mess | +0.044 |
roughly | +0.044 |
ruining | +0.042 |
Neg logit contributions | |
name | -0.038 |
ka | -0.033 |
aboard | -0.032 |
tie | -0.032 |
Crown | -0.031 |
Token | . |
Token ID | 13 |
Feature activation | +8.98e-01 |
Pos logit contributions | |
wasteful | +0.176 |
neither | +0.170 |
structures | +0.169 |
Uh | +0.166 |
mouths | +0.165 |
Neg logit contributions | |
aboard | -0.099 |
tie | -0.097 |
name | -0.095 |
ance | -0.088 |
orn | -0.087 |
Token | They |
Token ID | 1119 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
mess | +0.036 |
neither | +0.033 |
wasteful | +0.031 |
ruining | +0.029 |
overflow | +0.029 |
Neg logit contributions | |
name | -0.039 |
tie | -0.036 |
ka | -0.035 |
dot | -0.035 |
ance | -0.035 |
Token | live |
Token ID | 2107 |
Feature activation | +6.87e-01 |
Pos logit contributions | |
neither | +0.058 |
mess | +0.058 |
wasteful | +0.057 |
roughly | +0.053 |
structures | +0.052 |
Neg logit contributions | |
name | -0.044 |
tie | -0.041 |
ance | -0.040 |
ka | -0.039 |
leaf | -0.039 |
Token | here |
Token ID | 994 |
Feature activation | +2.73e+00 |
Pos logit contributions | |
neither | +0.031 |
mess | +0.030 |
wasteful | +0.029 |
overflow | +0.027 |
ruining | +0.026 |
Neg logit contributions | |
name | -0.023 |
ka | -0.023 |
tie | -0.023 |
aboard | -0.023 |
ance | -0.022 |
Token | and |
Token ID | 290 |
Feature activation | +2.24e+00 |
Pos logit contributions | |
mess | +0.114 |
wasteful | +0.113 |
neither | +0.112 |
overflow | +0.107 |
ruining | +0.102 |
Neg logit contributions | |
name | -0.097 |
tie | -0.096 |
aboard | -0.094 |
ka | -0.087 |
ance | -0.086 |
Token | 's |
Token ID | 338 |
Feature activation | +5.26e-01 |
Pos logit contributions | |
wasteful | +0.044 |
neither | +0.043 |
mess | +0.041 |
overflow | +0.041 |
ruining | +0.041 |
Neg logit contributions | |
name | -0.040 |
ance | -0.036 |
ka | -0.036 |
aboard | -0.035 |
tie | -0.035 |
Token | never |
Token ID | 1239 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
neither | +0.025 |
mess | +0.024 |
wasteful | +0.024 |
rats | +0.022 |
roughly | +0.022 |
Neg logit contributions | |
name | -0.018 |
tie | -0.016 |
ka | -0.016 |
ance | -0.016 |
dot | -0.015 |
Token | go |
Token ID | 467 |
Feature activation | +2.52e-01 |
Pos logit contributions | |
neither | +0.052 |
wasteful | +0.050 |
mess | +0.047 |
roughly | +0.045 |
mosquitoes | +0.044 |
Neg logit contributions | |
name | -0.036 |
tie | -0.031 |
ka | -0.031 |
ance | -0.030 |
Crown | -0.030 |
Token | into |
Token ID | 656 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
neither | +0.012 |
mess | +0.012 |
wasteful | +0.012 |
overflow | +0.011 |
roughly | +0.011 |
Neg logit contributions | |
tie | -0.008 |
name | -0.008 |
ka | -0.008 |
aboard | -0.007 |
dot | -0.007 |
Token | a |
Token ID | 257 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
neither | +0.099 |
wasteful | +0.093 |
overflow | +0.088 |
mess | +0.087 |
forgot | +0.084 |
Neg logit contributions | |
aboard | -0.061 |
outer | -0.060 |
powers | -0.054 |
ance | -0.052 |
name | -0.051 |
Token | cave |
Token ID | 11527 |
Feature activation | +3.91e+00 |
Pos logit contributions | |
neither | +0.049 |
wasteful | +0.042 |
mess | +0.041 |
ruining | +0.039 |
overflow | +0.038 |
Neg logit contributions | |
name | -0.040 |
aboard | -0.037 |
ka | -0.037 |
dot | -0.037 |
tie | -0.037 |
Token | again |
Token ID | 757 |
Feature activation | +3.07e+00 |
Pos logit contributions | |
wasteful | +0.175 |
neither | +0.174 |
mouths | +0.168 |
mosquitoes | +0.167 |
structures | +0.164 |
Neg logit contributions | |
aboard | -0.117 |
tie | -0.108 |
ti | -0.105 |
ka | -0.102 |
name | -0.101 |
Token | !" |
Token ID | 2474 |
Feature activation | +8.54e-02 |
Pos logit contributions | |
wasteful | +0.154 |
neither | +0.147 |
overflow | +0.143 |
structures | +0.139 |
ruining | +0.138 |
Neg logit contributions | |
aboard | -0.090 |
name | -0.080 |
ka | -0.071 |
powers | -0.067 |
ti | -0.067 |
Token | L |
Token ID | 406 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
neither | +0.003 |
mess | +0.002 |
wasteful | +0.002 |
overflow | +0.002 |
ruining | +0.002 |
Neg logit contributions | |
ka | -0.004 |
name | -0.004 |
tie | -0.004 |
dot | -0.004 |
ance | -0.004 |
Token | ila |
Token ID | 10102 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
tie | +0.014 |
ka | +0.013 |
name | +0.013 |
dot | +0.013 |
shirt | +0.013 |
Neg logit contributions | |
neither | -0.018 |
mess | -0.017 |
wasteful | -0.016 |
fighting | -0.015 |
overflow | -0.015 |
Token | said |
Token ID | 531 |
Feature activation | -9.58e-01 |
Pos logit contributions | |
wasteful | +0.071 |
overflow | +0.070 |
mess | +0.069 |
neither | +0.069 |
ruining | +0.069 |
Neg logit contributions | |
tie | -0.048 |
name | -0.045 |
ka | -0.044 |
shirt | -0.044 |
leaf | -0.043 |
Token | it |
Token ID | 340 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
ka | +0.070 |
leaf | +0.065 |
tie | +0.065 |
ance | +0.065 |
anger | +0.064 |
Neg logit contributions | |
mess | -0.071 |
neither | -0.071 |
wasteful | -0.060 |
not | -0.059 |
floor | -0.059 |
Token | would |
Token ID | 561 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
leaf | +0.074 |
ka | +0.071 |
tie | +0.068 |
dot | +0.068 |
aboard | +0.067 |
Neg logit contributions | |
neither | -0.062 |
mess | -0.061 |
floor | -0.052 |
wasteful | -0.050 |
overflow | -0.049 |
Token | give |
Token ID | 1577 |
Feature activation | -2.20e-01 |
Pos logit contributions | |
ka | +0.054 |
ance | +0.053 |
leaf | +0.052 |
dot | +0.051 |
aboard | +0.051 |
Neg logit contributions | |
mess | -0.064 |
neither | -0.061 |
overflow | -0.057 |
wasteful | -0.054 |
not | -0.054 |
Token | her |
Token ID | 607 |
Feature activation | -3.31e+00 |
Pos logit contributions | |
ka | +0.008 |
tie | +0.008 |
name | +0.008 |
dot | +0.008 |
ance | +0.008 |
Neg logit contributions | |
neither | -0.010 |
mess | -0.009 |
wasteful | -0.009 |
fighting | -0.008 |
overflow | -0.008 |
Token | a |
Token ID | 257 |
Feature activation | -4.03e+00 |
Pos logit contributions | |
leaf | +0.159 |
orn | +0.159 |
anced | +0.155 |
reetings | +0.153 |
lynn | +0.150 |
Neg logit contributions | |
mess | -0.109 |
neither | -0.098 |
not | -0.092 |
both | -0.088 |
messy | -0.077 |
Token | special |
Token ID | 2041 |
Feature activation | -5.07e+00 |
Pos logit contributions | |
ance | +0.241 |
anced | +0.239 |
ully | +0.207 |
lynn | +0.204 |
umper | +0.204 |
Neg logit contributions | |
mess | -0.135 |
messy | -0.088 |
not | -0.088 |
bit | -0.086 |
fuss | -0.082 |
Token | power |
Token ID | 1176 |
Feature activation | -2.43e+00 |
Pos logit contributions | |
ance | +0.248 |
anced | +0.244 |
umper | +0.241 |
aved | +0.225 |
igh | +0.224 |
Neg logit contributions | |
mess | -0.203 |
floor | -0.151 |
mistake | -0.147 |
game | -0.144 |
noise | -0.143 |
Token | . |
Token ID | 13 |
Feature activation | -1.73e-01 |
Pos logit contributions | |
ka | +0.119 |
anced | +0.113 |
leaf | +0.113 |
dot | +0.112 |
ance | +0.109 |
Neg logit contributions | |
mess | -0.081 |
neither | -0.075 |
floor | -0.065 |
wasteful | -0.061 |
overflow | -0.060 |
Token |
|
Token ID | 198 |
Feature activation | -3.76e-01 |
Pos logit contributions | |
ka | +0.007 |
tie | +0.007 |
ance | +0.007 |
name | +0.006 |
dot | +0.006 |
Neg logit contributions | |
neither | -0.007 |
mess | -0.007 |
wasteful | -0.006 |
fighting | -0.006 |
overflow | -0.006 |
Token |
|
Token ID | 198 |
Feature activation | +6.08e-01 |
Pos logit contributions | |
ka | +0.016 |
tie | +0.015 |
ance | +0.015 |
name | +0.015 |
dot | +0.015 |
Neg logit contributions | |
neither | -0.014 |
mess | -0.014 |
wasteful | -0.013 |
fighting | -0.012 |
overflow | -0.012 |
Token | One |
Token ID | 3198 |
Feature activation | -7.52e-01 |
Pos logit contributions | |
neither | +0.025 |
mess | +0.024 |
wasteful | +0.023 |
ruining | +0.021 |
overflow | +0.021 |
Neg logit contributions | |
tie | -0.025 |
name | -0.024 |
ka | -0.024 |
dot | -0.024 |
shirt | -0.023 |
Token | parents |
Token ID | 3397 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
tie | +0.081 |
ance | +0.081 |
orn | +0.080 |
ka | +0.080 |
anced | +0.079 |
Neg logit contributions | |
neither | -0.034 |
mess | -0.033 |
fighting | -0.027 |
quarrel | -0.025 |
wasteful | -0.024 |
Token | gave |
Token ID | 2921 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
ka | +0.060 |
dot | +0.059 |
tie | +0.058 |
powers | +0.056 |
ance | +0.056 |
Neg logit contributions | |
mess | -0.054 |
neither | -0.053 |
fighting | -0.049 |
wasteful | -0.046 |
ruining | -0.044 |
Token | him |
Token ID | 683 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
tie | +0.054 |
dot | +0.053 |
ka | +0.053 |
anger | +0.053 |
amulet | +0.052 |
Neg logit contributions | |
neither | -0.051 |
mess | -0.050 |
fighting | -0.045 |
wasteful | -0.043 |
mice | -0.041 |
Token | a |
Token ID | 257 |
Feature activation | -2.79e+00 |
Pos logit contributions | |
amulet | +0.104 |
ka | +0.104 |
ocket | +0.103 |
anced | +0.102 |
tie | +0.102 |
Neg logit contributions | |
neither | -0.081 |
mess | -0.071 |
wasteful | -0.068 |
both | -0.062 |
raw | -0.061 |
Token | pe |
Token ID | 613 |
Feature activation | -3.90e+00 |
Pos logit contributions | |
ance | +0.146 |
anced | +0.140 |
ka | +0.129 |
dot | +0.123 |
tie | +0.123 |
Neg logit contributions | |
mess | -0.099 |
neither | -0.078 |
quarrel | -0.077 |
flood | -0.072 |
fight | -0.070 |
Token | b |
Token ID | 65 |
Feature activation | -5.04e+00 |
Pos logit contributions | |
anced | +0.155 |
ance | +0.139 |
Magic | +0.139 |
icol | +0.139 |
leaf | +0.137 |
Neg logit contributions | |
mess | -0.142 |
neither | -0.136 |
not | -0.125 |
wasteful | -0.124 |
pies | -0.113 |
Token | ble |
Token ID | 903 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
leaf | +0.282 |
outer | +0.272 |
Ta | +0.266 |
shirt | +0.258 |
chery | +0.256 |
Neg logit contributions | |
neither | -0.107 |
not | -0.093 |
both | -0.061 |
mess | -0.058 |
pies | -0.056 |
Token | . |
Token ID | 13 |
Feature activation | -1.50e+00 |
Pos logit contributions | |
ka | +0.079 |
dot | +0.075 |
anced | +0.075 |
leaf | +0.075 |
ance | +0.074 |
Neg logit contributions | |
neither | -0.062 |
mess | -0.060 |
wasteful | -0.053 |
fighting | -0.050 |
roughly | -0.050 |
Token | He |
Token ID | 679 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
ka | +0.069 |
ocket | +0.066 |
powers | +0.066 |
orn | +0.065 |
ance | +0.064 |
Neg logit contributions | |
neither | -0.053 |
mess | -0.047 |
wasteful | -0.044 |
fighting | -0.043 |
roughly | -0.042 |
Token | had |
Token ID | 550 |
Feature activation | -1.22e+00 |
Pos logit contributions | |
ka | +0.054 |
ance | +0.054 |
tie | +0.053 |
dot | +0.053 |
leaf | +0.051 |
Neg logit contributions | |
mess | -0.047 |
neither | -0.046 |
wasteful | -0.041 |
fighting | -0.039 |
roughly | -0.038 |
Token | never |
Token ID | 1239 |
Feature activation | -7.10e-01 |
Pos logit contributions | |
ka | +0.050 |
tie | +0.048 |
ance | +0.048 |
anger | +0.047 |
name | +0.047 |
Neg logit contributions | |
mess | -0.045 |
neither | -0.045 |
wasteful | -0.039 |
fighting | -0.038 |
floor | -0.038 |
Token | He |
Token ID | 679 |
Feature activation | -1.70e-01 |
Pos logit contributions | |
mess | +0.025 |
neither | +0.025 |
wasteful | +0.024 |
ruining | +0.022 |
overflow | +0.022 |
Neg logit contributions | |
name | -0.020 |
ka | -0.019 |
tie | -0.019 |
ance | -0.019 |
shirt | -0.019 |
Token | gave |
Token ID | 2921 |
Feature activation | -4.75e-01 |
Pos logit contributions | |
ka | +0.007 |
tie | +0.007 |
name | +0.007 |
ance | +0.007 |
dot | +0.007 |
Neg logit contributions | |
neither | -0.007 |
mess | -0.007 |
wasteful | -0.006 |
overflow | -0.006 |
fighting | -0.006 |
Token | him |
Token ID | 683 |
Feature activation | -2.08e+00 |
Pos logit contributions | |
ka | +0.021 |
tie | +0.020 |
dot | +0.020 |
name | +0.020 |
ance | +0.020 |
Neg logit contributions | |
neither | -0.017 |
mess | -0.017 |
wasteful | -0.015 |
fighting | -0.014 |
ruining | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | -3.07e+00 |
Pos logit contributions | |
ka | +0.096 |
anger | +0.095 |
dot | +0.092 |
anced | +0.092 |
tie | +0.091 |
Neg logit contributions | |
neither | -0.070 |
mess | -0.062 |
wasteful | -0.060 |
mice | -0.057 |
rats | -0.057 |
Token | beautiful |
Token ID | 4950 |
Feature activation | -3.95e+00 |
Pos logit contributions | |
ance | +0.162 |
anced | +0.154 |
ka | +0.143 |
dot | +0.140 |
ancing | +0.131 |
Neg logit contributions | |
mess | -0.111 |
wasteful | -0.082 |
fuss | -0.079 |
neither | -0.078 |
fight | -0.076 |
Token | golden |
Token ID | 10861 |
Feature activation | -4.98e+00 |
Pos logit contributions | |
ance | +0.177 |
anced | +0.164 |
aved | +0.153 |
leaf | +0.149 |
dot | +0.147 |
Neg logit contributions | |
mess | -0.175 |
floor | -0.136 |
mistake | -0.131 |
table | -0.130 |
pie | -0.127 |
Token | sword |
Token ID | 8429 |
Feature activation | -1.99e+00 |
Pos logit contributions | |
ance | +0.220 |
anced | +0.210 |
leaf | +0.196 |
dot | +0.190 |
aved | +0.189 |
Neg logit contributions | |
mess | -0.217 |
floor | -0.176 |
mistake | -0.162 |
table | -0.159 |
pie | -0.158 |
Token | . |
Token ID | 13 |
Feature activation | -1.23e-01 |
Pos logit contributions | |
ka | +0.096 |
leaf | +0.092 |
dot | +0.091 |
ance | +0.089 |
anced | +0.089 |
Neg logit contributions | |
neither | -0.063 |
mess | -0.061 |
fighting | -0.051 |
roughly | -0.050 |
not | -0.049 |
Token |
|
Token ID | 198 |
Feature activation | -1.95e-01 |
Pos logit contributions | |
ka | +0.005 |
tie | +0.005 |
name | +0.005 |
ance | +0.004 |
dot | +0.004 |
Neg logit contributions | |
neither | -0.005 |
mess | -0.005 |
wasteful | -0.005 |
overflow | -0.004 |
fighting | -0.004 |
Token |
|
Token ID | 198 |
Feature activation | +8.53e-01 |
Pos logit contributions | |
ka | +0.008 |
tie | +0.008 |
name | +0.007 |
ance | +0.007 |
dot | +0.007 |
Neg logit contributions | |
neither | -0.008 |
mess | -0.008 |
wasteful | -0.007 |
overflow | -0.007 |
fighting | -0.007 |
Token | The |
Token ID | 464 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
neither | +0.034 |
mess | +0.033 |
wasteful | +0.030 |
ruining | +0.028 |
fighting | +0.028 |
Neg logit contributions | |
shirt | -0.035 |
tie | -0.035 |
leaf | -0.035 |
ka | -0.034 |
yo | -0.034 |
Token | and |
Token ID | 290 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
anced | +0.070 |
leaf | +0.069 |
tie | +0.069 |
ka | +0.068 |
ance | +0.068 |
Neg logit contributions | |
mess | -0.051 |
neither | -0.043 |
structures | -0.041 |
floor | -0.040 |
snakes | -0.040 |
Token | always |
Token ID | 1464 |
Feature activation | -1.88e+00 |
Pos logit contributions | |
ka | +0.075 |
aboard | +0.075 |
anger | +0.074 |
tie | +0.072 |
ance | +0.072 |
Neg logit contributions | |
neither | -0.048 |
mess | -0.047 |
not | -0.043 |
wasteful | -0.041 |
floor | -0.039 |
Token | wore |
Token ID | 12408 |
Feature activation | -2.76e+00 |
Pos logit contributions | |
anger | +0.091 |
leaf | +0.087 |
ance | +0.085 |
tie | +0.084 |
aboard | +0.083 |
Neg logit contributions | |
neither | -0.064 |
mess | -0.061 |
fighting | -0.056 |
ruining | -0.055 |
forgot | -0.055 |
Token | a |
Token ID | 257 |
Feature activation | -2.88e+00 |
Pos logit contributions | |
anced | +0.151 |
ane | +0.146 |
ance | +0.144 |
oured | +0.142 |
anger | +0.142 |
Neg logit contributions | |
mess | -0.074 |
not | -0.064 |
both | -0.062 |
mice | -0.056 |
structures | -0.053 |
Token | spark |
Token ID | 9009 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
ance | +0.179 |
anced | +0.170 |
leaf | +0.148 |
aboard | +0.145 |
ancing | +0.141 |
Neg logit contributions | |
mess | -0.103 |
floor | -0.069 |
messy | -0.068 |
not | -0.065 |
roughly | -0.061 |
Token | ly |
Token ID | 306 |
Feature activation | -4.98e+00 |
Pos logit contributions | |
anced | +0.102 |
ance | +0.096 |
leaf | +0.094 |
dot | +0.091 |
aboard | +0.090 |
Neg logit contributions | |
mess | -0.088 |
neither | -0.080 |
floor | -0.071 |
anymore | -0.071 |
wasteful | -0.069 |
Token | dress |
Token ID | 6576 |
Feature activation | -2.31e+00 |
Pos logit contributions | |
ance | +0.240 |
leaf | +0.228 |
anced | +0.226 |
iddled | +0.204 |
year | +0.194 |
Neg logit contributions | |
mess | -0.206 |
floor | -0.191 |
bit | -0.158 |
table | -0.154 |
while | -0.153 |
Token | . |
Token ID | 13 |
Feature activation | -6.66e-01 |
Pos logit contributions | |
leaf | +0.121 |
anced | +0.120 |
ka | +0.116 |
dot | +0.115 |
amulet | +0.114 |
Neg logit contributions | |
neither | -0.064 |
mess | -0.061 |
not | -0.052 |
floor | -0.052 |
anymore | -0.051 |
Token | One |
Token ID | 1881 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
ka | +0.033 |
leaf | +0.032 |
tie | +0.032 |
ance | +0.031 |
aboard | +0.031 |
Neg logit contributions | |
neither | -0.021 |
mess | -0.020 |
wasteful | -0.018 |
fighting | -0.017 |
roughly | -0.017 |
Token | day |
Token ID | 1110 |
Feature activation | -6.14e-01 |
Pos logit contributions | |
anced | +0.083 |
ance | +0.081 |
ka | +0.080 |
leaf | +0.079 |
tie | +0.079 |
Neg logit contributions | |
mess | -0.037 |
neither | -0.035 |
wasteful | -0.028 |
floor | -0.028 |
roughly | -0.027 |
Token | , |
Token ID | 11 |
Feature activation | +2.39e-01 |
Pos logit contributions | |
tie | +0.030 |
leaf | +0.030 |
ka | +0.030 |
ance | +0.029 |
shirt | +0.029 |
Neg logit contributions | |
neither | -0.020 |
mess | -0.018 |
wasteful | -0.016 |
fighting | -0.016 |
roughly | -0.015 |
Token | Pixie |
Token ID | 46687 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
neither | +0.005 |
mess | +0.005 |
wasteful | +0.005 |
overflow | +0.004 |
ruining | +0.004 |
Neg logit contributions | |
name | -0.005 |
ka | -0.005 |
tie | -0.005 |
ance | -0.005 |
dot | -0.005 |
Token | gave |
Token ID | 2921 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
ka | +0.059 |
tie | +0.058 |
dot | +0.057 |
leaf | +0.056 |
shirt | +0.056 |
Neg logit contributions | |
mess | -0.051 |
neither | -0.051 |
wasteful | -0.044 |
fighting | -0.041 |
roughly | -0.041 |
Token | her |
Token ID | 607 |
Feature activation | -2.54e+00 |
Pos logit contributions | |
dot | +0.073 |
ka | +0.072 |
tie | +0.071 |
shirt | +0.070 |
anger | +0.069 |
Neg logit contributions | |
neither | -0.053 |
mess | -0.046 |
Amy | -0.043 |
fighting | -0.041 |
mice | -0.040 |
Token | a |
Token ID | 257 |
Feature activation | -3.32e+00 |
Pos logit contributions | |
anced | +0.119 |
tie | +0.117 |
dot | +0.116 |
orn | +0.114 |
ka | +0.114 |
Neg logit contributions | |
mess | -0.085 |
neither | -0.081 |
not | -0.071 |
both | -0.066 |
floor | -0.064 |
Token | spark |
Token ID | 9009 |
Feature activation | -2.92e+00 |
Pos logit contributions | |
ance | +0.191 |
anced | +0.189 |
dot | +0.166 |
ka | +0.165 |
ancing | +0.162 |
Neg logit contributions | |
mess | -0.106 |
not | -0.071 |
bit | -0.070 |
neither | -0.069 |
million | -0.066 |
Token | ly |
Token ID | 306 |
Feature activation | -4.96e+00 |
Pos logit contributions | |
dot | +0.130 |
anced | +0.128 |
lynn | +0.123 |
leaf | +0.123 |
Crown | +0.123 |
Neg logit contributions | |
mess | -0.099 |
neither | -0.091 |
anymore | -0.073 |
floor | -0.072 |
worse | -0.072 |
Token | wand |
Token ID | 11569 |
Feature activation | -2.42e+00 |
Pos logit contributions | |
ance | +0.206 |
anced | +0.200 |
leaf | +0.188 |
lynn | +0.185 |
ully | +0.181 |
Neg logit contributions | |
mess | -0.208 |
floor | -0.171 |
bit | -0.164 |
longer | -0.153 |
noise | -0.152 |
Token | , |
Token ID | 11 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
ka | +0.121 |
dot | +0.118 |
ance | +0.115 |
leaf | +0.115 |
yo | +0.115 |
Neg logit contributions | |
neither | -0.074 |
mess | -0.068 |
fighting | -0.057 |
roughly | -0.056 |
wasteful | -0.054 |
Token | which |
Token ID | 543 |
Feature activation | -9.37e-01 |
Pos logit contributions | |
ka | +0.062 |
ance | +0.060 |
dot | +0.058 |
anger | +0.057 |
tie | +0.057 |
Neg logit contributions | |
neither | -0.042 |
mess | -0.042 |
wasteful | -0.036 |
fighting | -0.034 |
ruining | -0.033 |
Token | she |
Token ID | 673 |
Feature activation | -1.85e+00 |
Pos logit contributions | |
ka | +0.041 |
dot | +0.039 |
tie | +0.039 |
shirt | +0.038 |
ance | +0.038 |
Neg logit contributions | |
neither | -0.036 |
mess | -0.035 |
wasteful | -0.031 |
fighting | -0.029 |
overflow | -0.029 |
Token | could |
Token ID | 714 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
dot | +0.080 |
ka | +0.080 |
ance | +0.079 |
anger | +0.078 |
aboard | +0.078 |
Neg logit contributions | |
mess | -0.063 |
neither | -0.062 |
forgot | -0.055 |
wasteful | -0.053 |
roughly | -0.051 |
Token | bedroom |
Token ID | 14043 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
leaf | +0.174 |
ance | +0.158 |
reetings | +0.155 |
tie | +0.154 |
aboard | +0.150 |
Neg logit contributions | |
mess | -0.110 |
floor | -0.101 |
messy | -0.085 |
beds | -0.083 |
room | -0.080 |
Token | and |
Token ID | 290 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
neither | +0.050 |
wasteful | +0.049 |
mess | +0.047 |
overflow | +0.044 |
mosquitoes | +0.044 |
Neg logit contributions | |
name | -0.038 |
tie | -0.037 |
ance | -0.036 |
ka | -0.036 |
gown | -0.035 |
Token | put |
Token ID | 1234 |
Feature activation | -3.33e+00 |
Pos logit contributions | |
ka | +0.040 |
ance | +0.038 |
tie | +0.038 |
anced | +0.037 |
dot | +0.037 |
Neg logit contributions | |
neither | -0.044 |
mess | -0.044 |
wasteful | -0.039 |
fighting | -0.037 |
ruining | -0.037 |
Token | on |
Token ID | 319 |
Feature activation | -2.56e+00 |
Pos logit contributions | |
reetings | +0.207 |
leaf | +0.202 |
anced | +0.197 |
found | +0.197 |
chery | +0.197 |
Neg logit contributions | |
both | -0.065 |
mess | -0.056 |
neither | -0.053 |
Lily | -0.052 |
they | -0.046 |
Token | the |
Token ID | 262 |
Feature activation | -4.36e+00 |
Pos logit contributions | |
reetings | +0.151 |
leaf | +0.145 |
orn | +0.144 |
ane | +0.143 |
anced | +0.141 |
Neg logit contributions | |
mess | -0.063 |
both | -0.059 |
not | -0.052 |
floor | -0.049 |
neither | -0.046 |
Token | pretty |
Token ID | 2495 |
Feature activation | -4.93e+00 |
Pos logit contributions | |
anced | +0.217 |
ully | +0.216 |
enna | +0.214 |
igh | +0.212 |
year | +0.208 |
Neg logit contributions | |
mess | -0.157 |
floor | -0.157 |
noise | -0.120 |
ground | -0.119 |
floors | -0.115 |
Token | buckle |
Token ID | 44983 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
leaf | +0.249 |
year | +0.246 |
ance | +0.232 |
anced | +0.230 |
uttered | +0.226 |
Neg logit contributions | |
mess | -0.186 |
floor | -0.175 |
table | -0.132 |
floors | -0.130 |
windows | -0.121 |
Token | . |
Token ID | 13 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
ka | +0.043 |
leaf | +0.041 |
dot | +0.041 |
name | +0.040 |
ance | +0.040 |
Neg logit contributions | |
neither | -0.041 |
mess | -0.039 |
wasteful | -0.035 |
roughly | -0.034 |
fighting | -0.034 |
Token | |
Token ID | 220 |
Feature activation | -7.86e-01 |
Pos logit contributions | |
ka | +0.048 |
leaf | +0.046 |
tie | +0.046 |
ance | +0.046 |
aboard | +0.045 |
Neg logit contributions | |
neither | -0.033 |
mess | -0.032 |
wasteful | -0.028 |
fighting | -0.027 |
roughly | -0.027 |
Token |
|
Token ID | 198 |
Feature activation | -2.98e-01 |
Pos logit contributions | |
ka | +0.035 |
tie | +0.034 |
dot | +0.033 |
leaf | +0.033 |
ance | +0.033 |
Neg logit contributions | |
neither | -0.028 |
mess | -0.027 |
wasteful | -0.025 |
fighting | -0.023 |
roughly | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | +3.38e-01 |
Pos logit contributions | |
ka | +0.012 |
tie | +0.012 |
ance | +0.012 |
name | +0.012 |
dot | +0.011 |
Neg logit contributions | |
neither | -0.012 |
mess | -0.011 |
wasteful | -0.011 |
fighting | -0.010 |
overflow | -0.010 |
Token | green |
Token ID | 4077 |
Feature activation | -1.98e+00 |
Pos logit contributions | |
ance | +0.068 |
anced | +0.065 |
ka | +0.063 |
tie | +0.060 |
aboard | +0.060 |
Neg logit contributions | |
mess | -0.061 |
neither | -0.055 |
wasteful | -0.052 |
floor | -0.050 |
ruining | -0.049 |
Token | dress |
Token ID | 6576 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
leaf | +0.082 |
anced | +0.081 |
ance | +0.080 |
dot | +0.078 |
tie | +0.076 |
Neg logit contributions | |
mess | -0.080 |
floor | -0.072 |
neither | -0.068 |
table | -0.061 |
overflow | -0.061 |
Token | with |
Token ID | 351 |
Feature activation | -2.35e+00 |
Pos logit contributions | |
ka | +0.052 |
leaf | +0.052 |
dot | +0.051 |
anced | +0.051 |
ance | +0.050 |
Neg logit contributions | |
neither | -0.035 |
mess | -0.035 |
wasteful | -0.029 |
floor | -0.028 |
overflow | -0.028 |
Token | a |
Token ID | 257 |
Feature activation | -1.98e+00 |
Pos logit contributions | |
int | +0.098 |
orn | +0.096 |
tie | +0.096 |
anger | +0.096 |
aboard | +0.095 |
Neg logit contributions | |
mess | -0.083 |
neither | -0.079 |
floor | -0.073 |
mice | -0.072 |
rats | -0.071 |
Token | spark |
Token ID | 9009 |
Feature activation | -2.47e+00 |
Pos logit contributions | |
ance | +0.084 |
anced | +0.079 |
aboard | +0.074 |
leaf | +0.074 |
ka | +0.072 |
Neg logit contributions | |
mess | -0.089 |
floor | -0.074 |
neither | -0.073 |
wasteful | -0.067 |
fighting | -0.067 |
Token | ly |
Token ID | 306 |
Feature activation | -4.91e+00 |
Pos logit contributions | |
anced | +0.111 |
ance | +0.108 |
leaf | +0.107 |
lynn | +0.103 |
reetings | +0.102 |
Neg logit contributions | |
mess | -0.093 |
neither | -0.077 |
floor | -0.076 |
anymore | -0.068 |
fighting | -0.065 |
Token | belt |
Token ID | 10999 |
Feature activation | -2.60e+00 |
Pos logit contributions | |
ance | +0.218 |
leaf | +0.205 |
anced | +0.198 |
year | +0.189 |
found | +0.186 |
Neg logit contributions | |
mess | -0.212 |
floor | -0.199 |
table | -0.161 |
noise | -0.152 |
bit | -0.143 |
Token | . |
Token ID | 13 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
ka | +0.124 |
leaf | +0.120 |
anced | +0.120 |
Aid | +0.119 |
ance | +0.116 |
Neg logit contributions | |
mess | -0.081 |
neither | -0.081 |
floor | -0.074 |
fighting | -0.070 |
not | -0.068 |
Token | She |
Token ID | 1375 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
ka | +0.072 |
leaf | +0.072 |
tie | +0.071 |
powers | +0.070 |
amulet | +0.069 |
Neg logit contributions | |
mess | -0.045 |
neither | -0.044 |
floor | -0.037 |
wasteful | -0.035 |
roughly | -0.035 |
Token | couldn |
Token ID | 3521 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
ance | +0.061 |
aboard | +0.061 |
leaf | +0.061 |
tie | +0.060 |
anger | +0.060 |
Neg logit contributions | |
neither | -0.050 |
mess | -0.050 |
forgot | -0.042 |
floor | -0.041 |
wasteful | -0.041 |
Token | 't |
Token ID | 470 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
ka | +0.066 |
tie | +0.065 |
shirt | +0.063 |
leaf | +0.063 |
name | +0.062 |
Neg logit contributions | |
neither | -0.026 |
mess | -0.024 |
wasteful | -0.019 |
fighting | -0.018 |
overflow | -0.018 |
Token | she |
Token ID | 673 |
Feature activation | -1.13e+00 |
Pos logit contributions | |
neither | +0.005 |
mess | +0.005 |
wasteful | +0.005 |
overflow | +0.004 |
ruining | +0.004 |
Neg logit contributions | |
ka | -0.005 |
name | -0.005 |
tie | -0.005 |
ance | -0.005 |
dot | -0.005 |
Token | liked |
Token ID | 8288 |
Feature activation | +2.57e-02 |
Pos logit contributions | |
ka | +0.042 |
tie | +0.041 |
ance | +0.040 |
dot | +0.040 |
aboard | +0.040 |
Neg logit contributions | |
mess | -0.047 |
neither | -0.047 |
wasteful | -0.043 |
roughly | -0.040 |
fighting | -0.040 |
Token | to |
Token ID | 284 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
neither | +0.001 |
mess | +0.001 |
wasteful | +0.001 |
overflow | +0.001 |
fighting | +0.001 |
Neg logit contributions | |
ka | -0.001 |
tie | -0.001 |
name | -0.001 |
ance | -0.001 |
dot | -0.001 |
Token | wear |
Token ID | 5806 |
Feature activation | -2.98e+00 |
Pos logit contributions | |
ka | +0.056 |
tie | +0.053 |
ance | +0.052 |
aboard | +0.052 |
leaf | +0.051 |
Neg logit contributions | |
mess | -0.053 |
neither | -0.050 |
overflow | -0.045 |
wasteful | -0.045 |
quarrel | -0.045 |
Token | her |
Token ID | 607 |
Feature activation | -3.50e+00 |
Pos logit contributions | |
anced | +0.167 |
leaf | +0.163 |
ance | +0.160 |
reetings | +0.158 |
int | +0.157 |
Neg logit contributions | |
mess | -0.078 |
both | -0.063 |
not | -0.058 |
they | -0.058 |
floor | -0.056 |
Token | special |
Token ID | 2041 |
Feature activation | -4.90e+00 |
Pos logit contributions | |
ance | +0.187 |
year | +0.183 |
leaf | +0.180 |
anced | +0.178 |
reetings | +0.176 |
Neg logit contributions | |
mess | -0.105 |
not | -0.089 |
floor | -0.085 |
neither | -0.083 |
messy | -0.078 |
Token | dress |
Token ID | 6576 |
Feature activation | -1.93e+00 |
Pos logit contributions | |
ance | +0.223 |
year | +0.214 |
anced | +0.213 |
leaf | +0.212 |
aved | +0.207 |
Neg logit contributions | |
mess | -0.196 |
floor | -0.167 |
quarrel | -0.147 |
table | -0.145 |
fight | -0.137 |
Token | . |
Token ID | 13 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
leaf | +0.097 |
dot | +0.093 |
ka | +0.092 |
anced | +0.089 |
aboard | +0.088 |
Neg logit contributions | |
neither | -0.061 |
mess | -0.058 |
floor | -0.049 |
roughly | -0.046 |
not | -0.046 |
Token | It |
Token ID | 632 |
Feature activation | -7.05e-01 |
Pos logit contributions | |
ka | +0.048 |
tie | +0.046 |
leaf | +0.046 |
ance | +0.045 |
aboard | +0.045 |
Neg logit contributions | |
neither | -0.033 |
mess | -0.030 |
wasteful | -0.028 |
fighting | -0.027 |
roughly | -0.026 |
Token | was |
Token ID | 373 |
Feature activation | -1.75e+00 |
Pos logit contributions | |
ka | +0.029 |
tie | +0.029 |
leaf | +0.028 |
dot | +0.028 |
ance | +0.028 |
Neg logit contributions | |
neither | -0.028 |
mess | -0.027 |
wasteful | -0.024 |
overflow | -0.023 |
fighting | -0.022 |
Token | purple |
Token ID | 14032 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
ance | +0.080 |
leaf | +0.080 |
ka | +0.078 |
anger | +0.078 |
tie | +0.077 |
Neg logit contributions | |
mess | -0.061 |
neither | -0.060 |
wasteful | -0.053 |
not | -0.053 |
ruining | -0.050 |
Token | unique |
Token ID | 3748 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
ance | +0.034 |
tie | +0.034 |
ka | +0.033 |
anced | +0.033 |
aboard | +0.032 |
Neg logit contributions | |
mess | -0.025 |
neither | -0.024 |
wasteful | -0.023 |
fighting | -0.020 |
overflow | -0.020 |
Token | girl |
Token ID | 2576 |
Feature activation | -9.31e-01 |
Pos logit contributions | |
anced | +0.064 |
ance | +0.063 |
ka | +0.061 |
tie | +0.061 |
dot | +0.060 |
Neg logit contributions | |
mess | -0.045 |
neither | -0.039 |
wasteful | -0.036 |
floor | -0.034 |
fighting | -0.034 |
Token | with |
Token ID | 351 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
ka | +0.045 |
anced | +0.045 |
ance | +0.044 |
tie | +0.043 |
leaf | +0.043 |
Neg logit contributions | |
neither | -0.030 |
mess | -0.028 |
wasteful | -0.026 |
fighting | -0.025 |
roughly | -0.024 |
Token | a |
Token ID | 257 |
Feature activation | -9.62e-01 |
Pos logit contributions | |
aboard | +0.077 |
ka | +0.073 |
tie | +0.072 |
anger | +0.071 |
int | +0.071 |
Neg logit contributions | |
neither | -0.052 |
mess | -0.051 |
wasteful | -0.047 |
fighting | -0.045 |
mice | -0.045 |
Token | spark |
Token ID | 9009 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
ance | +0.043 |
ka | +0.041 |
anced | +0.041 |
aboard | +0.040 |
tie | +0.040 |
Neg logit contributions | |
mess | -0.038 |
neither | -0.035 |
wasteful | -0.032 |
fighting | -0.030 |
overflow | -0.029 |
Token | ly |
Token ID | 306 |
Feature activation | -4.90e+00 |
Pos logit contributions | |
anced | +0.089 |
ance | +0.087 |
dot | +0.087 |
leaf | +0.086 |
lynn | +0.084 |
Neg logit contributions | |
mess | -0.083 |
neither | -0.077 |
floor | -0.066 |
fighting | -0.064 |
wasteful | -0.064 |
Token | smile |
Token ID | 8212 |
Feature activation | -2.80e+00 |
Pos logit contributions | |
leaf | +0.217 |
ance | +0.217 |
anced | +0.198 |
iddled | +0.193 |
ancing | +0.190 |
Neg logit contributions | |
mess | -0.209 |
floor | -0.179 |
noise | -0.148 |
hole | -0.146 |
table | -0.143 |
Token | and |
Token ID | 290 |
Feature activation | -1.67e+00 |
Pos logit contributions | |
leaf | +0.138 |
ance | +0.121 |
lynn | +0.121 |
dot | +0.120 |
amulet | +0.120 |
Neg logit contributions | |
neither | -0.089 |
mess | -0.087 |
fighting | -0.078 |
floor | -0.074 |
anymore | -0.071 |
Token | bright |
Token ID | 6016 |
Feature activation | -3.10e+00 |
Pos logit contributions | |
ka | +0.071 |
anger | +0.071 |
tie | +0.069 |
aboard | +0.069 |
orn | +0.069 |
Neg logit contributions | |
neither | -0.061 |
mess | -0.061 |
not | -0.052 |
wasteful | -0.051 |
fighting | -0.051 |
Token | eyes |
Token ID | 2951 |
Feature activation | -3.05e+00 |
Pos logit contributions | |
ance | +0.161 |
leaf | +0.156 |
chery | +0.155 |
cher | +0.155 |
lynn | +0.152 |
Neg logit contributions | |
mess | -0.103 |
floor | -0.081 |
noise | -0.069 |
fighting | -0.067 |
faces | -0.066 |
Token | that |
Token ID | 326 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
lynn | +0.182 |
leaf | +0.180 |
amulet | +0.178 |
yo | +0.178 |
ance | +0.177 |
Neg logit contributions | |
mess | -0.061 |
neither | -0.054 |
fighting | -0.053 |
anymore | -0.047 |
floor | -0.041 |
Token | yellow |
Token ID | 7872 |
Feature activation | -2.89e+00 |
Pos logit contributions | |
ance | +0.068 |
anced | +0.066 |
ka | +0.061 |
tie | +0.060 |
leaf | +0.060 |
Neg logit contributions | |
mess | -0.054 |
neither | -0.047 |
wasteful | -0.045 |
floor | -0.042 |
fighting | -0.041 |
Token | dress |
Token ID | 6576 |
Feature activation | -8.72e-01 |
Pos logit contributions | |
anced | +0.113 |
leaf | +0.112 |
ance | +0.110 |
iddled | +0.102 |
aved | +0.101 |
Neg logit contributions | |
mess | -0.130 |
floor | -0.116 |
neither | -0.111 |
table | -0.105 |
longer | -0.100 |
Token | with |
Token ID | 351 |
Feature activation | -2.52e+00 |
Pos logit contributions | |
ka | +0.041 |
leaf | +0.040 |
anced | +0.040 |
dot | +0.040 |
ance | +0.040 |
Neg logit contributions | |
neither | -0.028 |
mess | -0.027 |
wasteful | -0.023 |
overflow | -0.022 |
floor | -0.022 |
Token | a |
Token ID | 257 |
Feature activation | -2.21e+00 |
Pos logit contributions | |
int | +0.115 |
anger | +0.110 |
Aid | +0.110 |
lynn | +0.109 |
aboard | +0.109 |
Neg logit contributions | |
mess | -0.081 |
neither | -0.077 |
floor | -0.072 |
mice | -0.072 |
both | -0.070 |
Token | spark |
Token ID | 9009 |
Feature activation | -1.96e+00 |
Pos logit contributions | |
ance | +0.096 |
anced | +0.091 |
leaf | +0.086 |
aboard | +0.084 |
ka | +0.082 |
Neg logit contributions | |
mess | -0.097 |
floor | -0.081 |
neither | -0.079 |
wasteful | -0.074 |
fighting | -0.073 |
Token | ly |
Token ID | 306 |
Feature activation | -4.89e+00 |
Pos logit contributions | |
anced | +0.085 |
ance | +0.082 |
leaf | +0.082 |
dot | +0.080 |
lynn | +0.077 |
Neg logit contributions | |
mess | -0.076 |
neither | -0.068 |
floor | -0.063 |
fighting | -0.058 |
wasteful | -0.057 |
Token | cape |
Token ID | 36023 |
Feature activation | -2.03e+00 |
Pos logit contributions | |
ance | +0.214 |
leaf | +0.209 |
anced | +0.202 |
year | +0.189 |
aved | +0.185 |
Neg logit contributions | |
mess | -0.212 |
floor | -0.203 |
table | -0.164 |
noise | -0.156 |
ground | -0.144 |
Token | . |
Token ID | 13 |
Feature activation | +2.98e-01 |
Pos logit contributions | |
ka | +0.099 |
leaf | +0.096 |
Aid | +0.096 |
dot | +0.093 |
shirt | +0.092 |
Neg logit contributions | |
neither | -0.065 |
mess | -0.062 |
fighting | -0.056 |
floor | -0.053 |
ruining | -0.053 |
Token |
|
Token ID | 198 |
Feature activation | -1.94e-01 |
Pos logit contributions | |
neither | +0.012 |
mess | +0.012 |
wasteful | +0.012 |
overflow | +0.011 |
ruining | +0.011 |
Neg logit contributions | |
name | -0.012 |
ka | -0.011 |
tie | -0.011 |
dot | -0.011 |
shirt | -0.011 |
Token |
|
Token ID | 198 |
Feature activation | +2.53e-01 |
Pos logit contributions | |
ka | +0.008 |
tie | +0.008 |
name | +0.007 |
ance | +0.007 |
dot | +0.007 |
Neg logit contributions | |
neither | -0.008 |
mess | -0.008 |
wasteful | -0.007 |
fighting | -0.006 |
overflow | -0.006 |
Token | One |
Token ID | 3198 |
Feature activation | -9.23e-01 |
Pos logit contributions | |
neither | +0.009 |
mess | +0.009 |
wasteful | +0.008 |
overflow | +0.008 |
ruining | +0.008 |
Neg logit contributions | |
ka | -0.011 |
tie | -0.011 |
name | -0.011 |
dot | -0.011 |
shirt | -0.011 |
Token | feeling |
Token ID | 4203 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
neither | +0.013 |
mess | +0.013 |
wasteful | +0.013 |
overflow | +0.011 |
mosquitoes | +0.011 |
Neg logit contributions | |
ka | -0.013 |
name | -0.013 |
tie | -0.012 |
ance | -0.012 |
dot | -0.012 |
Token | very |
Token ID | 845 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
ka | +0.070 |
anger | +0.069 |
tie | +0.069 |
anced | +0.069 |
ance | +0.068 |
Neg logit contributions | |
neither | -0.034 |
mess | -0.033 |
wasteful | -0.031 |
roughly | -0.026 |
overflow | -0.025 |
Token | proud |
Token ID | 6613 |
Feature activation | -6.75e-01 |
Pos logit contributions | |
ka | +0.008 |
tie | +0.008 |
name | +0.008 |
ance | +0.008 |
dot | +0.007 |
Neg logit contributions | |
neither | -0.006 |
mess | -0.006 |
wasteful | -0.006 |
overflow | -0.005 |
fighting | -0.005 |
Token | wearing |
Token ID | 5762 |
Feature activation | -2.41e+00 |
Pos logit contributions | |
ka | +0.029 |
tie | +0.028 |
dot | +0.028 |
name | +0.028 |
ance | +0.028 |
Neg logit contributions | |
neither | -0.025 |
mess | -0.025 |
wasteful | -0.022 |
fighting | -0.021 |
roughly | -0.021 |
Token | her |
Token ID | 607 |
Feature activation | -3.41e+00 |
Pos logit contributions | |
anced | +0.124 |
ka | +0.122 |
leaf | +0.121 |
dot | +0.121 |
ance | +0.121 |
Neg logit contributions | |
mess | -0.072 |
not | -0.059 |
roughly | -0.057 |
neither | -0.057 |
floor | -0.055 |
Token | special |
Token ID | 2041 |
Feature activation | -4.89e+00 |
Pos logit contributions | |
ance | +0.173 |
anced | +0.164 |
leaf | +0.155 |
dot | +0.151 |
aved | +0.149 |
Neg logit contributions | |
mess | -0.125 |
not | -0.100 |
floor | -0.092 |
quarrel | -0.090 |
neither | -0.089 |
Token | suit |
Token ID | 6050 |
Feature activation | -1.67e+00 |
Pos logit contributions | |
ance | +0.211 |
anced | +0.191 |
leaf | +0.182 |
aved | +0.173 |
tie | +0.170 |
Neg logit contributions | |
mess | -0.234 |
floor | -0.185 |
mistake | -0.185 |
fuss | -0.176 |
quarrel | -0.174 |
Token | . |
Token ID | 13 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
ka | +0.077 |
leaf | +0.076 |
dot | +0.075 |
ance | +0.072 |
name | +0.072 |
Neg logit contributions | |
neither | -0.056 |
mess | -0.055 |
wasteful | -0.046 |
roughly | -0.045 |
fighting | -0.045 |
Token | She |
Token ID | 1375 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
ka | +0.068 |
tie | +0.065 |
leaf | +0.063 |
powers | +0.062 |
dot | +0.062 |
Neg logit contributions | |
neither | -0.040 |
mess | -0.035 |
wasteful | -0.033 |
roughly | -0.032 |
fighting | -0.031 |
Token | will |
Token ID | 481 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
ka | +0.059 |
tie | +0.057 |
dot | +0.057 |
leaf | +0.056 |
anger | +0.056 |
Neg logit contributions | |
mess | -0.044 |
neither | -0.042 |
wasteful | -0.037 |
forgot | -0.036 |
roughly | -0.034 |
Token | never |
Token ID | 1239 |
Feature activation | +5.53e-01 |
Pos logit contributions | |
ka | +0.092 |
shirt | +0.089 |
dot | +0.088 |
leaf | +0.087 |
tie | +0.087 |
Neg logit contributions | |
neither | -0.037 |
mess | -0.035 |
not | -0.030 |
overflow | -0.027 |
wasteful | -0.026 |
Token | to |
Token ID | 284 |
Feature activation | -1.49e+00 |
Pos logit contributions | |
neither | +0.004 |
mess | +0.004 |
wasteful | +0.003 |
overflow | +0.003 |
ruining | +0.003 |
Neg logit contributions | |
ka | -0.003 |
tie | -0.003 |
name | -0.003 |
dot | -0.003 |
ance | -0.003 |
Token | her |
Token ID | 607 |
Feature activation | -3.54e-01 |
Pos logit contributions | |
tie | +0.073 |
ka | +0.072 |
ance | +0.071 |
shirt | +0.069 |
anger | +0.069 |
Neg logit contributions | |
mess | -0.049 |
neither | -0.048 |
fighting | -0.046 |
wasteful | -0.043 |
ruining | -0.041 |
Token | and |
Token ID | 290 |
Feature activation | -8.15e-03 |
Pos logit contributions | |
ka | +0.016 |
tie | +0.015 |
ance | +0.015 |
dot | +0.015 |
leaf | +0.015 |
Neg logit contributions | |
mess | -0.013 |
neither | -0.013 |
wasteful | -0.012 |
fighting | -0.011 |
ruining | -0.010 |
Token | put |
Token ID | 1234 |
Feature activation | -2.19e+00 |
Pos logit contributions | |
ka | +0.000 |
tie | +0.000 |
name | +0.000 |
ance | +0.000 |
dot | +0.000 |
Neg logit contributions | |
neither | -0.000 |
mess | -0.000 |
wasteful | -0.000 |
overflow | -0.000 |
fighting | -0.000 |
Token | a |
Token ID | 257 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
dot | +0.116 |
leaf | +0.113 |
amulet | +0.113 |
tie | +0.112 |
ka | +0.109 |
Neg logit contributions | |
neither | -0.063 |
mess | -0.055 |
wasteful | -0.050 |
fighting | -0.048 |
both | -0.046 |
Token | comforting |
Token ID | 34228 |
Feature activation | -4.89e+00 |
Pos logit contributions | |
ance | +0.112 |
anced | +0.107 |
tie | +0.096 |
dot | +0.096 |
ka | +0.092 |
Neg logit contributions | |
mess | -0.099 |
wasteful | -0.077 |
neither | -0.074 |
floor | -0.073 |
overflow | -0.073 |
Token | hand |
Token ID | 1021 |
Feature activation | -2.45e+00 |
Pos logit contributions | |
ance | +0.225 |
anced | +0.212 |
ane | +0.200 |
amb | +0.190 |
tie | +0.189 |
Neg logit contributions | |
mess | -0.204 |
floor | -0.169 |
noise | -0.159 |
table | -0.151 |
fuss | -0.147 |
Token | on |
Token ID | 319 |
Feature activation | -6.16e-01 |
Pos logit contributions | |
ka | +0.114 |
Crown | +0.112 |
leaf | +0.110 |
ance | +0.109 |
amulet | +0.109 |
Neg logit contributions | |
neither | -0.082 |
mess | -0.072 |
fighting | -0.066 |
roughly | -0.065 |
wasteful | -0.063 |
Token | her |
Token ID | 607 |
Feature activation | -2.06e+00 |
Pos logit contributions | |
ka | +0.027 |
tie | +0.026 |
ance | +0.026 |
dot | +0.026 |
shirt | +0.025 |
Neg logit contributions | |
neither | -0.023 |
mess | -0.022 |
wasteful | -0.020 |
fighting | -0.019 |
ruining | -0.019 |
Token | shoulder |
Token ID | 8163 |
Feature activation | -3.21e-01 |
Pos logit contributions | |
orn | +0.106 |
leaf | +0.105 |
anced | +0.105 |
ka | +0.104 |
ance | +0.104 |
Neg logit contributions | |
mess | -0.068 |
floor | -0.060 |
neither | -0.050 |
mistake | -0.048 |
not | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
ka | +0.013 |
tie | +0.012 |
name | +0.012 |
ance | +0.012 |
dot | +0.012 |
Neg logit contributions | |
neither | -0.013 |
mess | -0.013 |
wasteful | -0.012 |
overflow | -0.011 |
fighting | -0.011 |
Token | had |
Token ID | 550 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
dot | +0.077 |
ka | +0.076 |
ance | +0.076 |
tie | +0.075 |
leaf | +0.074 |
Neg logit contributions | |
mess | -0.057 |
neither | -0.056 |
fighting | -0.050 |
wasteful | -0.048 |
ruining | -0.044 |
Token | just |
Token ID | 655 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
ka | +0.052 |
tie | +0.051 |
ance | +0.051 |
dot | +0.050 |
anger | +0.050 |
Neg logit contributions | |
mess | -0.052 |
neither | -0.051 |
wasteful | -0.046 |
fighting | -0.045 |
ruining | -0.043 |
Token | given |
Token ID | 1813 |
Feature activation | -1.50e+00 |
Pos logit contributions | |
anger | +0.100 |
ance | +0.092 |
lor | +0.092 |
orn | +0.091 |
name | +0.090 |
Neg logit contributions | |
mess | -0.084 |
neither | -0.084 |
fighting | -0.080 |
wasteful | -0.075 |
ruining | -0.072 |
Token | him |
Token ID | 683 |
Feature activation | -2.40e+00 |
Pos logit contributions | |
tie | +0.062 |
ka | +0.061 |
dot | +0.061 |
anger | +0.061 |
anced | +0.061 |
Neg logit contributions | |
neither | -0.056 |
mess | -0.056 |
wasteful | -0.049 |
fighting | -0.048 |
pies | -0.044 |
Token | a |
Token ID | 257 |
Feature activation | -3.27e+00 |
Pos logit contributions | |
ka | +0.108 |
anced | +0.106 |
orn | +0.103 |
dot | +0.103 |
anger | +0.103 |
Neg logit contributions | |
neither | -0.085 |
mess | -0.075 |
wasteful | -0.073 |
not | -0.067 |
roughly | -0.066 |
Token | shiny |
Token ID | 22441 |
Feature activation | -4.84e+00 |
Pos logit contributions | |
ance | +0.178 |
anced | +0.174 |
ka | +0.153 |
ed | +0.149 |
dot | +0.148 |
Neg logit contributions | |
mess | -0.119 |
quarrel | -0.091 |
fight | -0.087 |
neither | -0.086 |
mistake | -0.085 |
Token | new |
Token ID | 649 |
Feature activation | -4.08e+00 |
Pos logit contributions | |
ance | +0.231 |
anced | +0.227 |
leaf | +0.218 |
dot | +0.204 |
lor | +0.202 |
Neg logit contributions | |
mess | -0.203 |
floor | -0.153 |
mistake | -0.152 |
longer | -0.137 |
table | -0.137 |
Token | pistol |
Token ID | 16790 |
Feature activation | -1.98e+00 |
Pos logit contributions | |
ance | +0.211 |
anced | +0.199 |
leaf | +0.181 |
year | +0.181 |
aved | +0.180 |
Neg logit contributions | |
mess | -0.155 |
floor | -0.120 |
mistake | -0.120 |
quarrel | -0.115 |
fight | -0.111 |
Token | . |
Token ID | 13 |
Feature activation | -4.64e-01 |
Pos logit contributions | |
ka | +0.094 |
dot | +0.090 |
leaf | +0.089 |
anced | +0.089 |
ance | +0.087 |
Neg logit contributions | |
neither | -0.067 |
mess | -0.061 |
wasteful | -0.054 |
fighting | -0.054 |
roughly | -0.053 |
Token | Joey |
Token ID | 26154 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
ka | +0.018 |
tie | +0.017 |
ance | +0.017 |
dot | +0.017 |
leaf | +0.017 |
Neg logit contributions | |
neither | -0.019 |
mess | -0.019 |
wasteful | -0.018 |
fighting | -0.017 |
overflow | -0.016 |
Token | had |
Token ID | 550 |
Feature activation | -2.18e+00 |
Pos logit contributions | |
ance | +0.045 |
ka | +0.044 |
tie | +0.044 |
leaf | +0.044 |
dot | +0.044 |
Neg logit contributions | |
mess | -0.037 |
neither | -0.036 |
wasteful | -0.032 |
fighting | -0.031 |
roughly | -0.030 |
Token | she |
Token ID | 673 |
Feature activation | -8.72e-01 |
Pos logit contributions | |
leaf | +0.056 |
tie | +0.054 |
ka | +0.054 |
dot | +0.053 |
orn | +0.052 |
Neg logit contributions | |
neither | -0.048 |
mess | -0.046 |
wasteful | -0.039 |
fighting | -0.039 |
roughly | -0.037 |
Token | put |
Token ID | 1234 |
Feature activation | -3.33e+00 |
Pos logit contributions | |
leaf | +0.036 |
tie | +0.035 |
ka | +0.035 |
ance | +0.035 |
name | +0.034 |
Neg logit contributions | |
mess | -0.034 |
neither | -0.034 |
wasteful | -0.029 |
fighting | -0.028 |
floor | -0.028 |
Token | on |
Token ID | 319 |
Feature activation | -1.88e+00 |
Pos logit contributions | |
reetings | +0.215 |
leaf | +0.214 |
found | +0.206 |
chery | +0.201 |
int | +0.198 |
Neg logit contributions | |
both | -0.065 |
neither | -0.055 |
Lily | -0.048 |
they | -0.047 |
mess | -0.045 |
Token | her |
Token ID | 607 |
Feature activation | -3.61e+00 |
Pos logit contributions | |
reetings | +0.099 |
orn | +0.098 |
leaf | +0.098 |
anced | +0.095 |
tie | +0.094 |
Neg logit contributions | |
mess | -0.058 |
neither | -0.049 |
floor | -0.047 |
both | -0.047 |
not | -0.046 |
Token | spark |
Token ID | 9009 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
leaf | +0.177 |
ance | +0.171 |
orn | +0.170 |
year | +0.167 |
aved | +0.165 |
Neg logit contributions | |
mess | -0.130 |
floor | -0.115 |
not | -0.107 |
messy | -0.098 |
neither | -0.096 |
Token | ly |
Token ID | 306 |
Feature activation | -4.82e+00 |
Pos logit contributions | |
leaf | +0.130 |
anced | +0.118 |
Magic | +0.117 |
dot | +0.117 |
name | +0.115 |
Neg logit contributions | |
mess | -0.102 |
floor | -0.087 |
neither | -0.083 |
anymore | -0.073 |
roughly | -0.068 |
Token | shoes |
Token ID | 10012 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
leaf | +0.258 |
ance | +0.234 |
year | +0.230 |
anced | +0.223 |
aboard | +0.221 |
Neg logit contributions | |
mess | -0.170 |
floor | -0.159 |
table | -0.121 |
messy | -0.112 |
floors | -0.107 |
Token | . |
Token ID | 13 |
Feature activation | -1.91e+00 |
Pos logit contributions | |
leaf | +0.084 |
ka | +0.081 |
name | +0.081 |
anced | +0.080 |
dot | +0.080 |
Neg logit contributions | |
neither | -0.055 |
mess | -0.051 |
roughly | -0.043 |
fighting | -0.042 |
floor | -0.042 |
Token |
|
Token ID | 198 |
Feature activation | -6.11e-01 |
Pos logit contributions | |
ka | +0.100 |
leaf | +0.099 |
powers | +0.098 |
tie | +0.098 |
amulet | +0.096 |
Neg logit contributions | |
mess | -0.051 |
neither | -0.050 |
floor | -0.044 |
roughly | -0.042 |
fighting | -0.041 |
Token |
|
Token ID | 198 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
ka | +0.026 |
tie | +0.026 |
ance | +0.025 |
dot | +0.025 |
anced | +0.025 |
Neg logit contributions | |
neither | -0.022 |
mess | -0.022 |
wasteful | -0.020 |
roughly | -0.018 |
fighting | -0.018 |
Token | She |
Token ID | 3347 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
ka | +0.010 |
tie | +0.010 |
name | +0.010 |
ance | +0.010 |
dot | +0.010 |
Neg logit contributions | |
neither | -0.006 |
mess | -0.006 |
wasteful | -0.006 |
overflow | -0.005 |
fighting | -0.005 |
Token | mom |
Token ID | 1995 |
Feature activation | +5.22e-01 |
Pos logit contributions | |
leaf | +0.103 |
tie | +0.098 |
aboard | +0.098 |
ka | +0.097 |
amulet | +0.097 |
Neg logit contributions | |
neither | -0.033 |
mess | -0.032 |
fighting | -0.021 |
not | -0.021 |
floor | -0.020 |
Token | my |
Token ID | 1820 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
neither | +0.023 |
mess | +0.022 |
wasteful | +0.022 |
overflow | +0.020 |
ruining | +0.020 |
Neg logit contributions | |
ka | -0.018 |
name | -0.018 |
tie | -0.018 |
aboard | -0.017 |
ance | -0.017 |
Token | gave |
Token ID | 2921 |
Feature activation | -1.58e-01 |
Pos logit contributions | |
neither | +0.006 |
mess | +0.006 |
wasteful | +0.005 |
overflow | +0.005 |
ruining | +0.005 |
Neg logit contributions | |
ka | -0.006 |
tie | -0.006 |
name | -0.006 |
ance | -0.005 |
dot | -0.005 |
Token | her |
Token ID | 607 |
Feature activation | -2.48e+00 |
Pos logit contributions | |
ka | +0.005 |
tie | +0.005 |
name | +0.005 |
ance | +0.005 |
dot | +0.005 |
Neg logit contributions | |
neither | -0.007 |
mess | -0.007 |
wasteful | -0.007 |
fighting | -0.006 |
overflow | -0.006 |
Token | a |
Token ID | 257 |
Feature activation | -3.96e+00 |
Pos logit contributions | |
anced | +0.126 |
reetings | +0.126 |
orn | +0.122 |
amulet | +0.121 |
ance | +0.120 |
Neg logit contributions | |
neither | -0.070 |
mess | -0.067 |
wasteful | -0.056 |
not | -0.054 |
both | -0.054 |
Token | cool |
Token ID | 3608 |
Feature activation | -4.82e+00 |
Pos logit contributions | |
ance | +0.247 |
anced | +0.242 |
ancing | +0.206 |
ed | +0.203 |
aved | +0.202 |
Neg logit contributions | |
mess | -0.129 |
messy | -0.088 |
table | -0.083 |
quarrel | -0.081 |
fight | -0.078 |
Token | mask |
Token ID | 9335 |
Feature activation | -2.15e+00 |
Pos logit contributions | |
ance | +0.234 |
reetings | +0.232 |
anced | +0.230 |
year | +0.226 |
lor | +0.224 |
Neg logit contributions | |
mess | -0.195 |
table | -0.142 |
floor | -0.142 |
bit | -0.127 |
mistake | -0.122 |
Token | to |
Token ID | 284 |
Feature activation | -2.14e+00 |
Pos logit contributions | |
ka | +0.100 |
dot | +0.094 |
ance | +0.092 |
anced | +0.092 |
amulet | +0.092 |
Neg logit contributions | |
neither | -0.073 |
mess | -0.069 |
roughly | -0.060 |
fighting | -0.058 |
wasteful | -0.057 |
Token | wear |
Token ID | 5806 |
Feature activation | -2.60e+00 |
Pos logit contributions | |
ka | +0.084 |
aboard | +0.083 |
anced | +0.083 |
dot | +0.081 |
ance | +0.081 |
Neg logit contributions | |
mess | -0.087 |
neither | -0.079 |
wasteful | -0.075 |
not | -0.075 |
overflow | -0.074 |
Token | . |
Token ID | 13 |
Feature activation | +3.25e-01 |
Pos logit contributions | |
leaf | +0.138 |
anced | +0.138 |
dot | +0.135 |
amulet | +0.133 |
int | +0.132 |
Neg logit contributions | |
mess | -0.068 |
not | -0.059 |
neither | -0.058 |
both | -0.053 |
floor | -0.053 |
Token | It |
Token ID | 632 |
Feature activation | -6.10e-01 |
Pos logit contributions | |
neither | +0.012 |
mess | +0.011 |
wasteful | +0.011 |
ruining | +0.010 |
overflow | +0.010 |
Neg logit contributions | |
name | -0.014 |
ka | -0.014 |
tie | -0.014 |
dot | -0.013 |
shirt | -0.013 |
Token | she |
Token ID | 673 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
mess | +0.030 |
neither | +0.029 |
wasteful | +0.028 |
overflow | +0.025 |
ruining | +0.025 |
Neg logit contributions | |
name | -0.033 |
ka | -0.033 |
tie | -0.032 |
dot | -0.031 |
shirt | -0.031 |
Token | found |
Token ID | 1043 |
Feature activation | -1.13e+00 |
Pos logit contributions | |
ka | +0.012 |
tie | +0.011 |
name | +0.011 |
ance | +0.011 |
leaf | +0.011 |
Neg logit contributions | |
neither | -0.011 |
mess | -0.011 |
wasteful | -0.010 |
fighting | -0.009 |
overflow | -0.009 |
Token | a |
Token ID | 257 |
Feature activation | -3.03e-01 |
Pos logit contributions | |
ka | +0.046 |
tie | +0.044 |
anced | +0.044 |
aboard | +0.044 |
dot | +0.043 |
Neg logit contributions | |
neither | -0.045 |
mess | -0.041 |
wasteful | -0.040 |
rats | -0.039 |
mice | -0.037 |
Token | shiny |
Token ID | 22441 |
Feature activation | -3.16e+00 |
Pos logit contributions | |
ka | +0.012 |
ance | +0.012 |
tie | +0.012 |
dot | +0.012 |
name | +0.011 |
Neg logit contributions | |
mess | -0.012 |
neither | -0.012 |
wasteful | -0.011 |
overflow | -0.010 |
fighting | -0.010 |
Token | pe |
Token ID | 613 |
Feature activation | -3.85e+00 |
Pos logit contributions | |
ance | +0.139 |
anced | +0.136 |
leaf | +0.132 |
dot | +0.124 |
found | +0.124 |
Neg logit contributions | |
mess | -0.145 |
floor | -0.116 |
neither | -0.109 |
wasteful | -0.104 |
fighting | -0.103 |
Token | b |
Token ID | 65 |
Feature activation | -4.76e+00 |
Pos logit contributions | |
anced | +0.164 |
leaf | +0.149 |
ance | +0.143 |
Music | +0.135 |
farewell | +0.134 |
Neg logit contributions | |
mess | -0.151 |
neither | -0.127 |
not | -0.126 |
wasteful | -0.121 |
floor | -0.117 |
Token | ble |
Token ID | 903 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
leaf | +0.286 |
outer | +0.263 |
shirt | +0.262 |
chery | +0.260 |
farewell | +0.259 |
Neg logit contributions | |
neither | -0.101 |
not | -0.088 |
mess | -0.053 |
fighting | -0.048 |
both | -0.045 |
Token | and |
Token ID | 290 |
Feature activation | -1.00e+00 |
Pos logit contributions | |
ka | +0.046 |
anced | +0.044 |
ance | +0.044 |
name | +0.043 |
leaf | +0.043 |
Neg logit contributions | |
neither | -0.045 |
mess | -0.044 |
wasteful | -0.039 |
fighting | -0.039 |
roughly | -0.038 |
Token | picked |
Token ID | 6497 |
Feature activation | -3.14e+00 |
Pos logit contributions | |
ance | +0.043 |
ka | +0.041 |
tie | +0.041 |
aboard | +0.040 |
anced | +0.040 |
Neg logit contributions | |
mess | -0.039 |
neither | -0.038 |
wasteful | -0.036 |
roughly | -0.032 |
fighting | -0.031 |
Token | it |
Token ID | 340 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
anced | +0.197 |
chery | +0.193 |
ka | +0.191 |
uster | +0.190 |
leaf | +0.190 |
Neg logit contributions | |
neither | -0.048 |
mess | -0.045 |
not | -0.035 |
fighting | -0.029 |
both | -0.027 |
Token | up |
Token ID | 510 |
Feature activation | -4.16e-01 |
Pos logit contributions | |
leaf | +0.047 |
ka | +0.046 |
ance | +0.046 |
anced | +0.045 |
dot | +0.045 |
Neg logit contributions | |
neither | -0.043 |
mess | -0.042 |
fighting | -0.035 |
roughly | -0.035 |
wasteful | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | -5.17e-01 |
Pos logit contributions | |
ka | +0.168 |
anced | +0.156 |
leaf | +0.155 |
dot | +0.155 |
amulet | +0.153 |
Neg logit contributions | |
neither | -0.085 |
mess | -0.071 |
not | -0.068 |
they | -0.061 |
floor | -0.060 |
Token | He |
Token ID | 679 |
Feature activation | -2.20e+00 |
Pos logit contributions | |
ka | +0.021 |
tie | +0.020 |
ance | +0.020 |
leaf | +0.020 |
dot | +0.020 |
Neg logit contributions | |
neither | -0.020 |
mess | -0.020 |
wasteful | -0.019 |
overflow | -0.017 |
fighting | -0.017 |
Token | put |
Token ID | 1234 |
Feature activation | -3.35e+00 |
Pos logit contributions | |
ance | +0.105 |
leaf | +0.103 |
tie | +0.102 |
dot | +0.100 |
anger | +0.100 |
Neg logit contributions | |
mess | -0.071 |
neither | -0.063 |
forgot | -0.060 |
floor | -0.056 |
stomp | -0.052 |
Token | on |
Token ID | 319 |
Feature activation | -2.50e+00 |
Pos logit contributions | |
reetings | +0.176 |
found | +0.175 |
leaf | +0.175 |
lor | +0.169 |
dot | +0.166 |
Neg logit contributions | |
neither | -0.086 |
both | -0.086 |
mess | -0.074 |
they | -0.073 |
fighting | -0.064 |
Token | a |
Token ID | 257 |
Feature activation | -3.81e+00 |
Pos logit contributions | |
reetings | +0.141 |
leaf | +0.139 |
orn | +0.138 |
ocket | +0.137 |
anced | +0.133 |
Neg logit contributions | |
mess | -0.066 |
floor | -0.055 |
both | -0.054 |
neither | -0.052 |
not | -0.052 |
Token | blue |
Token ID | 4171 |
Feature activation | -4.76e+00 |
Pos logit contributions | |
ance | +0.205 |
anced | +0.198 |
aboard | +0.164 |
reetings | +0.161 |
ancing | +0.158 |
Neg logit contributions | |
mess | -0.155 |
floor | -0.129 |
messy | -0.115 |
table | -0.113 |
timer | -0.109 |
Token | shirt |
Token ID | 10147 |
Feature activation | -2.46e+00 |
Pos logit contributions | |
reetings | +0.246 |
leaf | +0.241 |
ance | +0.238 |
anced | +0.235 |
year | +0.234 |
Neg logit contributions | |
floor | -0.138 |
mess | -0.136 |
table | -0.127 |
longer | -0.106 |
both | -0.102 |
Token | , |
Token ID | 11 |
Feature activation | -2.76e+00 |
Pos logit contributions | |
leaf | +0.142 |
ka | +0.138 |
anced | +0.134 |
dot | +0.133 |
ance | +0.131 |
Neg logit contributions | |
neither | -0.053 |
mess | -0.048 |
floor | -0.046 |
not | -0.041 |
roughly | -0.037 |
Token | blue |
Token ID | 4171 |
Feature activation | -3.81e+00 |
Pos logit contributions | |
orn | +0.130 |
ance | +0.129 |
anger | +0.128 |
ane | +0.125 |
reetings | +0.123 |
Neg logit contributions | |
mess | -0.082 |
neither | -0.082 |
floor | -0.078 |
not | -0.076 |
they | -0.074 |
Token | pants |
Token ID | 12581 |
Feature activation | -2.43e+00 |
Pos logit contributions | |
leaf | +0.194 |
year | +0.192 |
reetings | +0.188 |
ocket | +0.180 |
amulet | +0.176 |
Neg logit contributions | |
floor | -0.102 |
mess | -0.097 |
table | -0.087 |
mouths | -0.086 |
cups | -0.085 |
Token | , |
Token ID | 11 |
Feature activation | -2.52e+00 |
Pos logit contributions | |
leaf | +0.137 |
ka | +0.132 |
amulet | +0.132 |
name | +0.131 |
orn | +0.129 |
Neg logit contributions | |
neither | -0.057 |
mess | -0.052 |
floor | -0.045 |
fighting | -0.044 |
they | -0.039 |
Token | She |
Token ID | 1375 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
ka | +0.057 |
tie | +0.054 |
orn | +0.052 |
leaf | +0.052 |
ocket | +0.052 |
Neg logit contributions | |
neither | -0.035 |
mess | -0.032 |
wasteful | -0.030 |
roughly | -0.029 |
floor | -0.029 |
Token | picked |
Token ID | 6497 |
Feature activation | -2.55e+00 |
Pos logit contributions | |
leaf | +0.073 |
tie | +0.073 |
ka | +0.072 |
dot | +0.071 |
ocket | +0.070 |
Neg logit contributions | |
mess | -0.055 |
neither | -0.054 |
forgot | -0.047 |
quarrel | -0.045 |
fighting | -0.045 |
Token | up |
Token ID | 510 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
och | +0.121 |
ka | +0.121 |
aved | +0.120 |
ocket | +0.120 |
uster | +0.118 |
Neg logit contributions | |
neither | -0.069 |
mess | -0.065 |
mosquitoes | -0.064 |
snakes | -0.058 |
mice | -0.058 |
Token | a |
Token ID | 257 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
tie | +0.062 |
ka | +0.060 |
leaf | +0.060 |
dot | +0.059 |
anced | +0.059 |
Neg logit contributions | |
neither | -0.052 |
mess | -0.051 |
rats | -0.047 |
mosquitoes | -0.045 |
mice | -0.045 |
Token | shiny |
Token ID | 22441 |
Feature activation | -4.10e+00 |
Pos logit contributions | |
ance | +0.084 |
anced | +0.083 |
dot | +0.075 |
ka | +0.073 |
leaf | +0.072 |
Neg logit contributions | |
mess | -0.071 |
neither | -0.055 |
floor | -0.054 |
wasteful | -0.053 |
quarrel | -0.051 |
Token | golden |
Token ID | 10861 |
Feature activation | -4.75e+00 |
Pos logit contributions | |
anced | +0.183 |
ance | +0.181 |
leaf | +0.179 |
dot | +0.164 |
aved | +0.164 |
Neg logit contributions | |
mess | -0.181 |
floor | -0.146 |
mistake | -0.126 |
table | -0.124 |
tower | -0.120 |
Token | sword |
Token ID | 8429 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
ance | +0.220 |
anced | +0.214 |
leaf | +0.203 |
aved | +0.194 |
iddled | +0.188 |
Neg logit contributions | |
mess | -0.204 |
floor | -0.173 |
table | -0.150 |
tower | -0.145 |
pie | -0.144 |
Token | and |
Token ID | 290 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
ka | +0.072 |
leaf | +0.070 |
dot | +0.069 |
shirt | +0.067 |
Aid | +0.067 |
Neg logit contributions | |
neither | -0.042 |
mess | -0.040 |
fighting | -0.034 |
floor | -0.033 |
roughly | -0.032 |
Token | went |
Token ID | 1816 |
Feature activation | -3.36e+00 |
Pos logit contributions | |
ka | +0.049 |
tie | +0.047 |
dot | +0.047 |
leaf | +0.046 |
ance | +0.046 |
Neg logit contributions | |
neither | -0.043 |
mess | -0.042 |
wasteful | -0.037 |
roughly | -0.036 |
fighting | -0.035 |
Token | out |
Token ID | 503 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
orn | +0.189 |
ocket | +0.186 |
anced | +0.185 |
igh | +0.184 |
ka | +0.181 |
Neg logit contributions | |
not | -0.079 |
mess | -0.072 |
after | -0.068 |
they | -0.065 |
ruining | -0.060 |
Token | into |
Token ID | 656 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
tie | +0.107 |
ka | +0.106 |
yo | +0.102 |
orn | +0.102 |
Aid | +0.101 |
Neg logit contributions | |
mess | -0.058 |
neither | -0.058 |
fighting | -0.055 |
snakes | -0.050 |
ruining | -0.049 |
Token | fairy |
Token ID | 25607 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
tie | +0.055 |
shirt | +0.052 |
anced | +0.051 |
dot | +0.049 |
ance | +0.049 |
Neg logit contributions | |
mess | -0.067 |
neither | -0.059 |
wasteful | -0.057 |
fighting | -0.056 |
floor | -0.054 |
Token | gave |
Token ID | 2921 |
Feature activation | -3.85e-01 |
Pos logit contributions | |
ka | +0.007 |
tie | +0.007 |
name | +0.007 |
ance | +0.007 |
dot | +0.007 |
Neg logit contributions | |
neither | -0.008 |
mess | -0.008 |
wasteful | -0.007 |
fighting | -0.007 |
overflow | -0.007 |
Token | Lily |
Token ID | 20037 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
tie | +0.019 |
ka | +0.019 |
dot | +0.019 |
name | +0.019 |
leaf | +0.019 |
Neg logit contributions | |
neither | -0.012 |
mess | -0.011 |
wasteful | -0.010 |
fighting | -0.009 |
overflow | -0.009 |
Token | a |
Token ID | 257 |
Feature activation | -3.36e+00 |
Pos logit contributions | |
anger | +0.096 |
ka | +0.095 |
anced | +0.091 |
leaf | +0.091 |
dot | +0.089 |
Neg logit contributions | |
neither | -0.070 |
mess | -0.065 |
both | -0.057 |
not | -0.057 |
mice | -0.056 |
Token | spark |
Token ID | 9009 |
Feature activation | -2.88e+00 |
Pos logit contributions | |
ance | +0.206 |
anced | +0.199 |
dot | +0.177 |
ka | +0.176 |
lynn | +0.173 |
Neg logit contributions | |
mess | -0.097 |
not | -0.060 |
neither | -0.058 |
overflow | -0.057 |
wasteful | -0.057 |
Token | ly |
Token ID | 306 |
Feature activation | -4.74e+00 |
Pos logit contributions | |
anced | +0.143 |
dot | +0.136 |
ance | +0.134 |
aved | +0.133 |
lynn | +0.133 |
Neg logit contributions | |
mess | -0.089 |
neither | -0.075 |
floor | -0.062 |
longer | -0.060 |
worse | -0.060 |
Token | wand |
Token ID | 11569 |
Feature activation | -2.55e+00 |
Pos logit contributions | |
ance | +0.195 |
anced | +0.182 |
leaf | +0.176 |
lynn | +0.174 |
aden | +0.167 |
Neg logit contributions | |
mess | -0.205 |
floor | -0.172 |
bit | -0.160 |
noise | -0.150 |
longer | -0.149 |
Token | and |
Token ID | 290 |
Feature activation | -4.24e-01 |
Pos logit contributions | |
ka | +0.144 |
dot | +0.136 |
ance | +0.135 |
leaf | +0.135 |
yo | +0.135 |
Neg logit contributions | |
neither | -0.064 |
mess | -0.057 |
not | -0.048 |
fighting | -0.047 |
roughly | -0.046 |
Token | said |
Token ID | 531 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
ka | +0.019 |
tie | +0.019 |
name | +0.018 |
dot | +0.018 |
ance | +0.018 |
Neg logit contributions | |
neither | -0.015 |
mess | -0.015 |
wasteful | -0.013 |
fighting | -0.012 |
overflow | -0.012 |
Token | to |
Token ID | 284 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
leaf | +0.077 |
tie | +0.076 |
ka | +0.076 |
shirt | +0.076 |
Aid | +0.074 |
Neg logit contributions | |
neither | -0.038 |
mess | -0.031 |
wasteful | -0.027 |
fighting | -0.027 |
not | -0.026 |
Token | make |
Token ID | 787 |
Feature activation | -2.96e-01 |
Pos logit contributions | |
ka | +0.007 |
tie | +0.007 |
name | +0.007 |
dot | +0.007 |
ance | +0.007 |
Neg logit contributions | |
neither | -0.005 |
mess | -0.005 |
wasteful | -0.004 |
overflow | -0.004 |
fighting | -0.004 |
Token | her |
Token ID | 607 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
ka | +0.018 |
tie | +0.018 |
ance | +0.018 |
leaf | +0.018 |
name | +0.018 |
Neg logit contributions | |
neither | -0.013 |
mess | -0.013 |
wasteful | -0.012 |
fighting | -0.011 |
roughly | -0.010 |
Token | mum |
Token ID | 25682 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
tie | +0.066 |
ka | +0.066 |
ance | +0.065 |
dot | +0.064 |
aboard | +0.064 |
Neg logit contributions | |
mess | -0.033 |
neither | -0.031 |
fighting | -0.025 |
wasteful | -0.025 |
mosquitoes | -0.024 |
Token | put |
Token ID | 1234 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
ka | +0.041 |
tie | +0.040 |
dot | +0.040 |
ance | +0.040 |
anced | +0.039 |
Neg logit contributions | |
neither | -0.043 |
mess | -0.042 |
wasteful | -0.039 |
fighting | -0.036 |
overflow | -0.035 |
Token | on |
Token ID | 319 |
Feature activation | -2.73e-01 |
Pos logit contributions | |
dot | +0.059 |
leaf | +0.059 |
ance | +0.059 |
anced | +0.058 |
amulet | +0.058 |
Neg logit contributions | |
neither | -0.061 |
mess | -0.057 |
wasteful | -0.053 |
raw | -0.051 |
fighting | -0.050 |
Token | her |
Token ID | 607 |
Feature activation | -3.07e+00 |
Pos logit contributions | |
tie | +0.010 |
ka | +0.010 |
name | +0.010 |
ance | +0.010 |
dot | +0.010 |
Neg logit contributions | |
neither | -0.011 |
mess | -0.011 |
wasteful | -0.011 |
fighting | -0.010 |
overflow | -0.010 |
Token | special |
Token ID | 2041 |
Feature activation | -4.73e+00 |
Pos logit contributions | |
ance | +0.146 |
anced | +0.139 |
ka | +0.135 |
tie | +0.134 |
leaf | +0.131 |
Neg logit contributions | |
mess | -0.105 |
neither | -0.093 |
floor | -0.091 |
quarrel | -0.086 |
raw | -0.086 |
Token | clothes |
Token ID | 8242 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
ance | +0.218 |
anced | +0.208 |
aved | +0.199 |
tie | +0.197 |
leaf | +0.195 |
Neg logit contributions | |
mess | -0.183 |
floor | -0.160 |
quarrel | -0.150 |
table | -0.146 |
fight | -0.144 |
Token | and |
Token ID | 290 |
Feature activation | -7.61e-01 |
Pos logit contributions | |
ka | +0.056 |
leaf | +0.055 |
dot | +0.054 |
name | +0.053 |
ance | +0.052 |
Neg logit contributions | |
neither | -0.040 |
mess | -0.037 |
wasteful | -0.032 |
fighting | -0.032 |
roughly | -0.031 |
Token | said |
Token ID | 531 |
Feature activation | -8.22e-01 |
Pos logit contributions | |
ka | +0.032 |
tie | +0.032 |
ance | +0.031 |
dot | +0.031 |
leaf | +0.031 |
Neg logit contributions | |
neither | -0.029 |
mess | -0.028 |
wasteful | -0.026 |
overflow | -0.024 |
fighting | -0.023 |
Token | to |
Token ID | 284 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
tie | +0.040 |
ka | +0.040 |
leaf | +0.039 |
shirt | +0.038 |
name | +0.038 |
Neg logit contributions | |
neither | -0.026 |
mess | -0.024 |
wasteful | -0.022 |
fighting | -0.021 |
overflow | -0.020 |
Token | her |
Token ID | 607 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
neither | +0.007 |
mess | +0.007 |
wasteful | +0.007 |
overflow | +0.006 |
fighting | +0.006 |
Neg logit contributions | |
ka | -0.005 |
name | -0.005 |
tie | -0.005 |
ance | -0.005 |
dot | -0.005 |
Token | At |
Token ID | 2953 |
Feature activation | +5.29e-01 |
Pos logit contributions | |
neither | +0.027 |
mess | +0.026 |
wasteful | +0.026 |
ruining | +0.024 |
overflow | +0.023 |
Neg logit contributions | |
name | -0.019 |
tie | -0.018 |
ka | -0.018 |
dot | -0.018 |
shirt | -0.017 |
Token | first |
Token ID | 717 |
Feature activation | +1.57e+00 |
Pos logit contributions | |
neither | +0.020 |
mess | +0.019 |
wasteful | +0.019 |
overflow | +0.017 |
roughly | +0.017 |
Neg logit contributions | |
aboard | -0.021 |
ka | -0.021 |
tie | -0.021 |
ance | -0.021 |
name | -0.021 |
Token | , |
Token ID | 11 |
Feature activation | +1.44e+00 |
Pos logit contributions | |
wasteful | +0.078 |
mess | +0.073 |
neither | +0.073 |
overflow | +0.073 |
ruining | +0.072 |
Neg logit contributions | |
name | -0.041 |
aboard | -0.041 |
ance | -0.039 |
powers | -0.038 |
orn | -0.038 |
Token | none |
Token ID | 4844 |
Feature activation | +4.56e-01 |
Pos logit contributions | |
mess | +0.074 |
wasteful | +0.071 |
overflow | +0.066 |
neither | +0.065 |
ruining | +0.064 |
Neg logit contributions | |
name | -0.040 |
tie | -0.038 |
aboard | -0.038 |
shone | -0.034 |
Magic | -0.034 |
Token | of |
Token ID | 286 |
Feature activation | +5.01e-02 |
Pos logit contributions | |
neither | +0.022 |
mess | +0.022 |
wasteful | +0.021 |
overflow | +0.020 |
ruining | +0.020 |
Neg logit contributions | |
tie | -0.014 |
ka | -0.013 |
name | -0.013 |
ance | -0.013 |
aboard | -0.013 |
Token | the |
Token ID | 262 |
Feature activation | +2.85e+00 |
Pos logit contributions | |
neither | +0.002 |
mess | +0.002 |
wasteful | +0.002 |
overflow | +0.002 |
ruining | +0.002 |
Neg logit contributions | |
ka | -0.002 |
tie | -0.002 |
name | -0.002 |
ance | -0.002 |
dot | -0.002 |
Token | other |
Token ID | 584 |
Feature activation | +2.34e+00 |
Pos logit contributions | |
neither | +0.134 |
wasteful | +0.127 |
iders | +0.119 |
roughly | +0.119 |
oving | +0.118 |
Neg logit contributions | |
name | -0.099 |
compass | -0.078 |
suitcase | -0.074 |
armour | -0.073 |
amulet | -0.071 |
Token | children |
Token ID | 1751 |
Feature activation | +3.51e-01 |
Pos logit contributions | |
wasteful | +0.102 |
neither | +0.094 |
ruining | +0.094 |
roughly | +0.090 |
overflow | +0.090 |
Neg logit contributions | |
name | -0.079 |
aboard | -0.074 |
shone | -0.072 |
ed | -0.070 |
anced | -0.068 |
Token | wanted |
Token ID | 2227 |
Feature activation | +2.20e-01 |
Pos logit contributions | |
neither | +0.014 |
mess | +0.014 |
wasteful | +0.013 |
overflow | +0.012 |
ruining | +0.012 |
Neg logit contributions | |
tie | -0.014 |
ka | -0.014 |
name | -0.013 |
ance | -0.013 |
dot | -0.013 |
Token | anything |
Token ID | 1997 |
Feature activation | +6.45e-01 |
Pos logit contributions | |
neither | +0.011 |
mess | +0.011 |
wasteful | +0.010 |
overflow | +0.009 |
ruining | +0.009 |
Neg logit contributions | |
ka | -0.007 |
name | -0.007 |
tie | -0.007 |
ance | -0.006 |
dot | -0.006 |
Token | to |
Token ID | 284 |
Feature activation | +4.23e-01 |
Pos logit contributions | |
neither | +0.026 |
mess | +0.025 |
wasteful | +0.025 |
overflow | +0.023 |
rats | +0.022 |
Neg logit contributions | |
tie | -0.024 |
name | -0.024 |
aboard | -0.023 |
ka | -0.023 |
dot | -0.023 |
Token | want |
Token ID | 765 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
ka | +0.014 |
tie | +0.013 |
ance | +0.013 |
dot | +0.013 |
leaf | +0.013 |
Neg logit contributions | |
neither | -0.013 |
mess | -0.012 |
wasteful | -0.011 |
fighting | -0.010 |
overflow | -0.010 |
Token | to |
Token ID | 284 |
Feature activation | +7.13e-01 |
Pos logit contributions | |
neither | +0.065 |
mess | +0.063 |
wasteful | +0.063 |
overflow | +0.061 |
roughly | +0.060 |
Neg logit contributions | |
name | -0.041 |
ka | -0.040 |
aboard | -0.040 |
tie | -0.039 |
powers | -0.038 |
Token | build |
Token ID | 1382 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
neither | +0.032 |
wasteful | +0.030 |
mess | +0.030 |
mosquitoes | +0.028 |
rats | +0.028 |
Neg logit contributions | |
name | -0.026 |
ka | -0.024 |
tie | -0.024 |
dot | -0.023 |
ance | -0.023 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
neither | +0.048 |
wasteful | +0.045 |
mess | +0.045 |
overflow | +0.043 |
roughly | +0.041 |
Neg logit contributions | |
name | -0.032 |
ka | -0.032 |
aboard | -0.031 |
ance | -0.030 |
tie | -0.030 |
Token | wide |
Token ID | 3094 |
Feature activation | +2.06e+00 |
Pos logit contributions | |
neither | +0.107 |
wasteful | +0.097 |
ruining | +0.089 |
mosquitoes | +0.087 |
wasting | +0.086 |
Neg logit contributions | |
name | -0.097 |
Crown | -0.080 |
yo | -0.078 |
aboard | -0.077 |
ka | -0.077 |
Token | tower |
Token ID | 10580 |
Feature activation | +3.08e+00 |
Pos logit contributions | |
neither | +0.079 |
wasteful | +0.076 |
mosquitoes | +0.065 |
ruining | +0.065 |
wasting | +0.063 |
Neg logit contributions | |
name | -0.094 |
ka | -0.089 |
aboard | -0.084 |
tie | -0.083 |
Crown | -0.083 |
Token | , |
Token ID | 11 |
Feature activation | +1.71e+00 |
Pos logit contributions | |
wasteful | +0.146 |
mosquitoes | +0.135 |
neither | +0.135 |
sudden | +0.134 |
overflow | +0.130 |
Neg logit contributions | |
aboard | -0.097 |
tie | -0.085 |
name | -0.085 |
ka | -0.079 |
ance | -0.076 |
Token | like |
Token ID | 588 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
mess | +0.059 |
neither | +0.056 |
wasteful | +0.056 |
roughly | +0.051 |
overflow | +0.051 |
Neg logit contributions | |
name | -0.077 |
tie | -0.068 |
shirt | -0.066 |
aboard | -0.064 |
Magic | -0.064 |
Token | a |
Token ID | 257 |
Feature activation | +2.39e+00 |
Pos logit contributions | |
neither | +0.042 |
mess | +0.041 |
wasteful | +0.040 |
overflow | +0.037 |
roughly | +0.036 |
Neg logit contributions | |
name | -0.042 |
ka | -0.042 |
dot | -0.041 |
tie | -0.041 |
aboard | -0.041 |
Token | castle |
Token ID | 16669 |
Feature activation | +2.98e+00 |
Pos logit contributions | |
neither | +0.096 |
wasteful | +0.088 |
ruining | +0.077 |
overflow | +0.077 |
wasting | +0.074 |
Neg logit contributions | |
name | -0.104 |
Crown | -0.090 |
tie | -0.087 |
ti | -0.087 |
ka | -0.086 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
wasteful | +0.139 |
neither | +0.137 |
overflow | +0.131 |
sudden | +0.128 |
mosquitoes | +0.127 |
Neg logit contributions | |
name | -0.090 |
aboard | -0.085 |
tie | -0.083 |
ti | -0.079 |
ka | -0.078 |
Token | . |
Token ID | 13 |
Feature activation | -5.91e-01 |
Pos logit contributions | |
neither | +0.034 |
mess | +0.032 |
wasteful | +0.032 |
overflow | +0.030 |
ruining | +0.029 |
Neg logit contributions | |
name | -0.033 |
ka | -0.032 |
tie | -0.032 |
ance | -0.031 |
aboard | -0.031 |
Token | Now |
Token ID | 2735 |
Feature activation | +8.76e-01 |
Pos logit contributions | |
ka | +0.023 |
tie | +0.022 |
dot | +0.021 |
ance | +0.021 |
leaf | +0.021 |
Neg logit contributions | |
neither | -0.025 |
mess | -0.023 |
wasteful | -0.023 |
fighting | -0.021 |
overflow | -0.020 |
Token | we |
Token ID | 356 |
Feature activation | +2.47e-01 |
Pos logit contributions | |
neither | +0.033 |
wasteful | +0.033 |
mess | +0.032 |
roughly | +0.029 |
overflow | +0.029 |
Neg logit contributions | |
name | -0.036 |
ance | -0.034 |
aboard | -0.033 |
ka | -0.033 |
tie | -0.033 |
Token | can |
Token ID | 460 |
Feature activation | +8.17e-01 |
Pos logit contributions | |
neither | +0.012 |
mess | +0.012 |
wasteful | +0.012 |
overflow | +0.011 |
ruining | +0.011 |
Neg logit contributions | |
name | -0.007 |
ka | -0.007 |
tie | -0.007 |
ance | -0.007 |
dot | -0.007 |
Token | enjoy |
Token ID | 2883 |
Feature activation | +9.99e-01 |
Pos logit contributions | |
neither | +0.034 |
wasteful | +0.034 |
mess | +0.033 |
rats | +0.030 |
roughly | +0.029 |
Neg logit contributions | |
name | -0.033 |
tie | -0.029 |
ance | -0.029 |
ka | -0.029 |
dot | -0.028 |
Token | cocoa |
Token ID | 35845 |
Feature activation | +2.94e+00 |
Pos logit contributions | |
neither | +0.043 |
mess | +0.040 |
wasteful | +0.040 |
overflow | +0.037 |
roughly | +0.036 |
Neg logit contributions | |
name | -0.036 |
aboard | -0.035 |
dot | -0.034 |
tie | -0.034 |
ka | -0.034 |
Token | and |
Token ID | 290 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
wasteful | +0.122 |
neither | +0.111 |
mess | +0.110 |
Uh | +0.106 |
overflow | +0.103 |
Neg logit contributions | |
aboard | -0.111 |
ance | -0.107 |
tie | -0.104 |
name | -0.103 |
ka | -0.101 |
Token | stay |
Token ID | 2652 |
Feature activation | +3.70e-01 |
Pos logit contributions | |
neither | +0.075 |
mess | +0.070 |
wasteful | +0.069 |
roughly | +0.066 |
overflow | +0.063 |
Neg logit contributions | |
name | -0.064 |
leaf | -0.055 |
tie | -0.055 |
dot | -0.054 |
powers | -0.054 |
Token | healthy |
Token ID | 5448 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
neither | +0.014 |
mess | +0.014 |
wasteful | +0.013 |
overflow | +0.012 |
rats | +0.012 |
Neg logit contributions | |
name | -0.015 |
ka | -0.015 |
tie | -0.015 |
dot | -0.014 |
aboard | -0.014 |
Token | too |
Token ID | 1165 |
Feature activation | +3.26e+00 |
Pos logit contributions | |
wasteful | +0.102 |
neither | +0.101 |
mess | +0.097 |
overflow | +0.095 |
erupted | +0.088 |
Neg logit contributions | |
aboard | -0.088 |
tie | -0.087 |
name | -0.085 |
ka | -0.084 |
ance | -0.082 |
Token | !" |
Token ID | 2474 |
Feature activation | +4.79e-01 |
Pos logit contributions | |
wasteful | +0.167 |
mess | +0.156 |
Uh | +0.153 |
neither | +0.153 |
mosquitoes | +0.151 |
Neg logit contributions | |
name | -0.086 |
ka | -0.079 |
aboard | -0.078 |
tie | -0.075 |
ti | -0.073 |
Token | very |
Token ID | 845 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
neither | +0.034 |
mess | +0.033 |
wasteful | +0.031 |
structures | +0.030 |
overflow | +0.030 |
Neg logit contributions | |
name | -0.031 |
ka | -0.030 |
tie | -0.030 |
dot | -0.030 |
leaf | -0.029 |
Token | grateful |
Token ID | 14066 |
Feature activation | -5.79e-01 |
Pos logit contributions | |
neither | +0.054 |
mess | +0.050 |
structures | +0.050 |
snakes | +0.048 |
mosquitoes | +0.047 |
Neg logit contributions | |
name | -0.041 |
ka | -0.036 |
leaf | -0.036 |
tie | -0.034 |
dot | -0.034 |
Token | . |
Token ID | 13 |
Feature activation | +2.17e+00 |
Pos logit contributions | |
ka | +0.024 |
dot | +0.023 |
tie | +0.023 |
ance | +0.023 |
leaf | +0.023 |
Neg logit contributions | |
neither | -0.023 |
mess | -0.022 |
wasteful | -0.020 |
fighting | -0.019 |
roughly | -0.019 |
Token | |
Token ID | 220 |
Feature activation | +4.77e-01 |
Pos logit contributions | |
wasteful | +0.127 |
neither | +0.126 |
ruining | +0.122 |
mosquitoes | +0.121 |
overflow | +0.120 |
Neg logit contributions | |
name | -0.052 |
tie | -0.039 |
Magic | -0.036 |
aboard | -0.035 |
shirt | -0.034 |
Token |
|
Token ID | 198 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
neither | +0.021 |
mess | +0.021 |
wasteful | +0.020 |
overflow | +0.018 |
ruining | +0.018 |
Neg logit contributions | |
name | -0.017 |
tie | -0.016 |
ka | -0.016 |
ance | -0.016 |
dot | -0.016 |
Token |
|
Token ID | 198 |
Feature activation | +2.38e+00 |
Pos logit contributions | |
neither | +0.106 |
wasteful | +0.102 |
mess | +0.098 |
ruining | +0.096 |
rats | +0.092 |
Neg logit contributions | |
name | -0.047 |
tie | -0.041 |
shirt | -0.040 |
aboard | -0.040 |
dot | -0.039 |
Token | Mrs |
Token ID | 27034 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
neither | +0.115 |
snakes | +0.094 |
drown | +0.091 |
wasteful | +0.090 |
mouths | +0.089 |
Neg logit contributions | |
shirt | -0.087 |
name | -0.081 |
tie | -0.081 |
yo | -0.080 |
new | -0.076 |
Token | . |
Token ID | 13 |
Feature activation | +9.53e-01 |
Pos logit contributions | |
neither | +0.006 |
mess | +0.006 |
wasteful | +0.006 |
overflow | +0.005 |
ruining | +0.005 |
Neg logit contributions | |
ka | -0.005 |
tie | -0.005 |
name | -0.005 |
ance | -0.005 |
dot | -0.005 |
Token | Johnson |
Token ID | 5030 |
Feature activation | +8.35e-01 |
Pos logit contributions | |
neither | +0.032 |
mess | +0.031 |
wasteful | +0.030 |
overflow | +0.027 |
mosquitoes | +0.026 |
Neg logit contributions | |
name | -0.045 |
ka | -0.042 |
tie | -0.042 |
aboard | -0.041 |
anced | -0.041 |
Token | asked |
Token ID | 1965 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
neither | +0.035 |
mess | +0.034 |
wasteful | +0.033 |
overflow | +0.031 |
ruining | +0.031 |
Neg logit contributions | |
ka | -0.030 |
tie | -0.030 |
name | -0.030 |
aboard | -0.028 |
dot | -0.028 |
Token | Lily |
Token ID | 20037 |
Feature activation | -5.34e-01 |
Pos logit contributions | |
neither | +0.050 |
mess | +0.049 |
overflow | +0.047 |
wasteful | +0.046 |
ruining | +0.044 |
Neg logit contributions | |
name | -0.042 |
ance | -0.039 |
ka | -0.038 |
tie | -0.038 |
aboard | -0.038 |
Token | " |
Token ID | 366 |
Feature activation | -3.68e-01 |
Pos logit contributions | |
neither | +0.043 |
mess | +0.043 |
wasteful | +0.041 |
overflow | +0.038 |
ruining | +0.037 |
Neg logit contributions | |
name | -0.045 |
anced | -0.040 |
dot | -0.040 |
tie | -0.039 |
ka | -0.039 |
Token | Yes |
Token ID | 5297 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
ka | +0.016 |
ance | +0.015 |
name | +0.015 |
tie | +0.015 |
aboard | +0.015 |
Neg logit contributions | |
neither | -0.013 |
mess | -0.013 |
wasteful | -0.012 |
roughly | -0.011 |
overflow | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | +1.39e+00 |
Pos logit contributions | |
wasteful | +0.098 |
mess | +0.095 |
ruining | +0.091 |
overflow | +0.091 |
structures | +0.091 |
Neg logit contributions | |
name | -0.042 |
ka | -0.042 |
anger | -0.036 |
aboard | -0.035 |
tie | -0.034 |
Token | it |
Token ID | 340 |
Feature activation | -5.03e-02 |
Pos logit contributions | |
mess | +0.047 |
wasteful | +0.045 |
neither | +0.044 |
roughly | +0.041 |
overflow | +0.041 |
Neg logit contributions | |
name | -0.062 |
tie | -0.058 |
aboard | -0.057 |
yo | -0.056 |
ance | -0.056 |
Token | is |
Token ID | 318 |
Feature activation | +5.50e-01 |
Pos logit contributions | |
ka | +0.002 |
tie | +0.002 |
name | +0.002 |
ance | +0.002 |
dot | +0.002 |
Neg logit contributions | |
neither | -0.002 |
mess | -0.002 |
wasteful | -0.002 |
overflow | -0.002 |
fighting | -0.002 |
Token | safe |
Token ID | 3338 |
Feature activation | +2.48e+00 |
Pos logit contributions | |
neither | +0.020 |
mess | +0.019 |
wasteful | +0.019 |
overflow | +0.018 |
fighting | +0.017 |
Neg logit contributions | |
name | -0.024 |
ka | -0.023 |
tie | -0.023 |
aboard | -0.022 |
dot | -0.022 |
Token | to |
Token ID | 284 |
Feature activation | +8.98e-01 |
Pos logit contributions | |
neither | +0.108 |
wasteful | +0.106 |
overflow | +0.101 |
mess | +0.100 |
ruining | +0.100 |
Neg logit contributions | |
aboard | -0.082 |
tie | -0.081 |
name | -0.080 |
ka | -0.076 |
powers | -0.076 |
Token | eat |
Token ID | 4483 |
Feature activation | +2.88e+00 |
Pos logit contributions | |
neither | +0.036 |
wasteful | +0.032 |
mess | +0.032 |
structures | +0.029 |
fighting | +0.029 |
Neg logit contributions | |
name | -0.038 |
tie | -0.035 |
dot | -0.033 |
ka | -0.033 |
Crown | -0.033 |
Token | ," |
Token ID | 553 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
neither | +0.135 |
overflow | +0.123 |
wasteful | +0.118 |
Uh | +0.114 |
ruining | +0.114 |
Neg logit contributions | |
name | -0.083 |
powers | -0.083 |
tie | -0.082 |
aboard | -0.081 |
ka | -0.081 |
Token | she |
Token ID | 673 |
Feature activation | +4.30e-01 |
Pos logit contributions | |
mess | +0.063 |
wasteful | +0.062 |
overflow | +0.060 |
neither | +0.059 |
ruining | +0.059 |
Neg logit contributions | |
name | -0.045 |
ka | -0.038 |
shirt | -0.037 |
shone | -0.037 |
tie | -0.037 |
Token | said |
Token ID | 531 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
neither | +0.016 |
mess | +0.015 |
wasteful | +0.014 |
overflow | +0.014 |
ruining | +0.014 |
Neg logit contributions | |
name | -0.019 |
ka | -0.018 |
tie | -0.018 |
leaf | -0.018 |
dot | -0.017 |
Token | was |
Token ID | 373 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
ka | +0.039 |
dot | +0.038 |
tie | +0.037 |
ance | +0.037 |
name | +0.037 |
Neg logit contributions | |
neither | -0.038 |
mess | -0.037 |
wasteful | -0.033 |
fighting | -0.032 |
roughly | -0.031 |
Token | no |
Token ID | 645 |
Feature activation | +4.73e-01 |
Pos logit contributions | |
ka | +0.006 |
tie | +0.006 |
name | +0.006 |
ance | +0.006 |
dot | +0.006 |
Neg logit contributions | |
neither | -0.005 |
mess | -0.005 |
wasteful | -0.004 |
fighting | -0.004 |
overflow | -0.004 |
Token | longer |
Token ID | 2392 |
Feature activation | +5.52e-01 |
Pos logit contributions | |
neither | +0.010 |
mess | +0.009 |
wasteful | +0.009 |
overflow | +0.008 |
structures | +0.007 |
Neg logit contributions | |
ka | -0.028 |
tie | -0.027 |
name | -0.026 |
dot | -0.026 |
leaf | -0.026 |
Token | mad |
Token ID | 8805 |
Feature activation | +7.44e-01 |
Pos logit contributions | |
neither | +0.017 |
mess | +0.016 |
wasteful | +0.015 |
overflow | +0.014 |
roughly | +0.013 |
Neg logit contributions | |
ka | -0.027 |
name | -0.026 |
tie | -0.026 |
dot | -0.025 |
ance | -0.025 |
Token | at |
Token ID | 379 |
Feature activation | +6.66e-01 |
Pos logit contributions | |
mess | +0.031 |
neither | +0.031 |
wasteful | +0.031 |
overflow | +0.028 |
ruining | +0.027 |
Neg logit contributions | |
name | -0.027 |
tie | -0.026 |
ka | -0.026 |
ance | -0.026 |
aboard | -0.025 |
Token | her |
Token ID | 607 |
Feature activation | +5.59e-01 |
Pos logit contributions | |
neither | +0.028 |
mess | +0.027 |
wasteful | +0.027 |
overflow | +0.025 |
roughly | +0.024 |
Neg logit contributions | |
name | -0.024 |
ka | -0.024 |
ance | -0.024 |
tie | -0.023 |
aboard | -0.023 |
Token | ! |
Token ID | 0 |
Feature activation | +3.20e-01 |
Pos logit contributions | |
neither | +0.024 |
mess | +0.023 |
wasteful | +0.022 |
overflow | +0.021 |
ruining | +0.020 |
Neg logit contributions | |
name | -0.020 |
tie | -0.020 |
ka | -0.019 |
ance | -0.019 |
dot | -0.019 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.88e+00 |
Pos logit contributions | |
mess | +0.014 |
neither | +0.014 |
wasteful | +0.013 |
overflow | +0.013 |
ruining | +0.012 |
Neg logit contributions | |
name | -0.011 |
tie | -0.011 |
ka | -0.011 |
ance | -0.011 |
dot | -0.011 |
Token | Once |
Token ID | 7454 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
ka | +0.078 |
tie | +0.077 |
leaf | +0.077 |
dot | +0.076 |
shirt | +0.075 |
Neg logit contributions | |
neither | -0.067 |
mess | -0.066 |
fighting | -0.061 |
roughly | -0.056 |
Amy | -0.056 |
Token | upon |
Token ID | 2402 |
Feature activation | -7.77e-01 |
Pos logit contributions | |
neither | +0.005 |
mess | +0.005 |
wasteful | +0.005 |
overflow | +0.005 |
ruining | +0.005 |
Neg logit contributions | |
ka | -0.003 |
tie | -0.003 |
name | -0.003 |
ance | -0.003 |
dot | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
ka | +0.029 |
ance | +0.028 |
tie | +0.028 |
anced | +0.027 |
aboard | +0.026 |
Neg logit contributions | |
neither | -0.033 |
mess | -0.033 |
wasteful | -0.030 |
fighting | -0.029 |
roughly | -0.028 |
Token | with |
Token ID | 351 |
Feature activation | +2.80e-01 |
Pos logit contributions | |
neither | +0.030 |
wasteful | +0.030 |
mess | +0.029 |
overflow | +0.027 |
ruining | +0.026 |
Neg logit contributions | |
name | -0.022 |
ka | -0.022 |
tie | -0.022 |
aboard | -0.022 |
ance | -0.022 |
Token | their |
Token ID | 511 |
Feature activation | -2.22e-01 |
Pos logit contributions | |
neither | +0.008 |
mess | +0.008 |
wasteful | +0.008 |
overflow | +0.007 |
roughly | +0.007 |
Neg logit contributions | |
name | -0.014 |
ka | -0.013 |
tie | -0.013 |
ance | -0.013 |
dot | -0.013 |
Token | work |
Token ID | 670 |
Feature activation | +7.68e-03 |
Pos logit contributions | |
ka | +0.009 |
tie | +0.009 |
ance | +0.008 |
name | +0.008 |
dot | +0.008 |
Neg logit contributions | |
mess | -0.009 |
neither | -0.009 |
wasteful | -0.008 |
fighting | -0.008 |
overflow | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | +1.32e-03 |
Pos logit contributions | |
neither | +0.000 |
mess | +0.000 |
wasteful | +0.000 |
overflow | +0.000 |
fighting | +0.000 |
Neg logit contributions | |
ka | -0.000 |
tie | -0.000 |
name | -0.000 |
ance | -0.000 |
dot | -0.000 |
Token | The |
Token ID | 383 |
Feature activation | +1.29e-02 |
Pos logit contributions | |
neither | +0.000 |
mess | +0.000 |
wasteful | +0.000 |
overflow | +0.000 |
ruining | +0.000 |
Neg logit contributions | |
ka | -0.000 |
tie | -0.000 |
name | -0.000 |
ance | -0.000 |
dot | -0.000 |
Token | palace |
Token ID | 20562 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
neither | +0.000 |
mess | +0.000 |
wasteful | +0.000 |
overflow | +0.000 |
fighting | +0.000 |
Neg logit contributions | |
ka | -0.001 |
tie | -0.001 |
name | -0.001 |
ance | -0.001 |
dot | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
neither | +0.061 |
wasteful | +0.061 |
mess | +0.059 |
ruining | +0.057 |
overflow | +0.056 |
Neg logit contributions | |
name | -0.037 |
ka | -0.036 |
tie | -0.036 |
shone | -0.035 |
ance | -0.035 |
Token | neat |
Token ID | 15049 |
Feature activation | -2.17e-01 |
Pos logit contributions | |
neither | +0.007 |
mess | +0.006 |
wasteful | +0.006 |
overflow | +0.005 |
fighting | +0.005 |
Neg logit contributions | |
ka | -0.008 |
tie | -0.008 |
name | -0.008 |
dot | -0.008 |
ance | -0.008 |
Token | and |
Token ID | 290 |
Feature activation | +4.21e-01 |
Pos logit contributions | |
ka | +0.009 |
tie | +0.009 |
name | +0.009 |
ance | +0.009 |
dot | +0.009 |
Neg logit contributions | |
neither | -0.008 |
mess | -0.008 |
wasteful | -0.007 |
overflow | -0.007 |
fighting | -0.007 |
Token | tidy |
Token ID | 43044 |
Feature activation | +7.01e-01 |
Pos logit contributions | |
neither | +0.014 |
mess | +0.014 |
wasteful | +0.013 |
overflow | +0.012 |
fighting | +0.012 |
Neg logit contributions | |
ka | -0.019 |
tie | -0.018 |
name | -0.018 |
ance | -0.018 |
dot | -0.018 |
Token | once |
Token ID | 1752 |
Feature activation | +7.34e-01 |
Pos logit contributions | |
neither | +0.030 |
wasteful | +0.028 |
mess | +0.028 |
overflow | +0.025 |
mosquitoes | +0.025 |
Neg logit contributions | |
ka | -0.026 |
tie | -0.025 |
name | -0.025 |
ance | -0.024 |
aboard | -0.024 |
Token | Tom |
Token ID | 13787 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
mess | +0.087 |
neither | +0.085 |
quarrel | +0.076 |
wasteful | +0.076 |
anymore | +0.075 |
Neg logit contributions | |
name | -0.056 |
da | -0.056 |
dot | -0.055 |
ka | -0.055 |
orn | -0.054 |
Token | and |
Token ID | 290 |
Feature activation | -2.39e-01 |
Pos logit contributions | |
neither | +0.062 |
wasteful | +0.058 |
overflow | +0.054 |
mess | +0.053 |
unfortunately | +0.052 |
Neg logit contributions | |
name | -0.041 |
ka | -0.040 |
ance | -0.038 |
tie | -0.037 |
aboard | -0.037 |
Token | Mia |
Token ID | 32189 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
ka | +0.015 |
tie | +0.015 |
ance | +0.014 |
name | +0.014 |
dot | +0.014 |
Neg logit contributions | |
neither | -0.004 |
mess | -0.004 |
wasteful | -0.004 |
overflow | -0.003 |
fighting | -0.003 |
Token | are |
Token ID | 389 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
neither | +0.052 |
wasteful | +0.051 |
mess | +0.049 |
structures | +0.045 |
rats | +0.044 |
Neg logit contributions | |
name | -0.046 |
tie | -0.041 |
ance | -0.041 |
ka | -0.041 |
dot | -0.039 |
Token | shocked |
Token ID | 11472 |
Feature activation | +7.52e-01 |
Pos logit contributions | |
wasteful | +0.069 |
neither | +0.068 |
timer | +0.064 |
mess | +0.064 |
overflow | +0.064 |
Neg logit contributions | |
name | -0.079 |
tie | -0.071 |
aboard | -0.066 |
ocket | -0.066 |
ka | -0.066 |
Token | . |
Token ID | 13 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
neither | +0.037 |
wasteful | +0.036 |
mess | +0.035 |
overflow | +0.033 |
ruining | +0.032 |
Neg logit contributions | |
name | -0.022 |
ka | -0.021 |
tie | -0.021 |
aboard | -0.020 |
dot | -0.020 |
Token | They |
Token ID | 1119 |
Feature activation | +5.39e-01 |
Pos logit contributions | |
mess | +0.050 |
wasteful | +0.048 |
neither | +0.048 |
ruining | +0.044 |
overflow | +0.042 |
Neg logit contributions | |
name | -0.058 |
dot | -0.049 |
ka | -0.048 |
anced | -0.048 |
ance | -0.047 |
Token | drop |
Token ID | 4268 |
Feature activation | +5.90e-01 |
Pos logit contributions | |
neither | +0.021 |
mess | +0.019 |
wasteful | +0.019 |
overflow | +0.017 |
rats | +0.017 |
Neg logit contributions | |
name | -0.022 |
ka | -0.021 |
tie | -0.021 |
leaf | -0.020 |
dot | -0.020 |
Token | the |
Token ID | 262 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
mess | +0.022 |
neither | +0.022 |
wasteful | +0.021 |
ruining | +0.019 |
overflow | +0.019 |
Neg logit contributions | |
name | -0.023 |
ka | -0.023 |
tie | -0.023 |
aboard | -0.023 |
ance | -0.022 |
Token | bow |
Token ID | 9563 |
Feature activation | +1.58e-01 |
Pos logit contributions | |
ance | +0.038 |
dot | +0.037 |
shirt | +0.036 |
anced | +0.036 |
ka | +0.035 |
Neg logit contributions | |
mess | -0.059 |
neither | -0.053 |
fighting | -0.052 |
wasteful | -0.049 |
floor | -0.049 |
Token | and |
Token ID | 290 |
Feature activation | +9.34e-01 |
Pos logit contributions | |
neither | +0.007 |
mess | +0.007 |
wasteful | +0.006 |
overflow | +0.006 |
ruining | +0.006 |
Neg logit contributions | |
tie | -0.006 |
ka | -0.006 |
name | -0.006 |
ance | -0.006 |
dot | -0.005 |
Token |
|
Token ID | 198 |
Feature activation | +5.70e-01 |
Pos logit contributions | |
ka | +0.002 |
tie | +0.002 |
name | +0.002 |
ance | +0.002 |
dot | +0.002 |
Neg logit contributions | |
neither | -0.002 |
mess | -0.002 |
wasteful | -0.002 |
overflow | -0.002 |
fighting | -0.002 |
Token |
|
Token ID | 198 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
neither | +0.027 |
mess | +0.026 |
wasteful | +0.026 |
ruining | +0.024 |
overflow | +0.024 |
Neg logit contributions | |
name | -0.019 |
tie | -0.018 |
ka | -0.018 |
dot | -0.017 |
shirt | -0.017 |
Token | Suddenly |
Token ID | 38582 |
Feature activation | +8.96e-01 |
Pos logit contributions | |
neither | +0.070 |
wasteful | +0.063 |
mess | +0.062 |
mouths | +0.057 |
overflow | +0.056 |
Neg logit contributions | |
shirt | -0.065 |
tie | -0.064 |
name | -0.062 |
dot | -0.062 |
leaf | -0.062 |
Token | , |
Token ID | 11 |
Feature activation | +1.58e+00 |
Pos logit contributions | |
mess | +0.040 |
wasteful | +0.039 |
neither | +0.039 |
overflow | +0.036 |
ruining | +0.035 |
Neg logit contributions | |
ka | -0.030 |
name | -0.030 |
ance | -0.029 |
aboard | -0.029 |
tie | -0.028 |
Token | the |
Token ID | 262 |
Feature activation | -6.67e-01 |
Pos logit contributions | |
mess | +0.074 |
wasteful | +0.073 |
neither | +0.069 |
overflow | +0.067 |
anth | +0.066 |
Neg logit contributions | |
name | -0.048 |
ka | -0.046 |
shone | -0.043 |
ance | -0.043 |
aboard | -0.043 |
Token | paper |
Token ID | 3348 |
Feature activation | +2.97e-01 |
Pos logit contributions | |
ka | +0.027 |
ance | +0.027 |
tie | +0.027 |
dot | +0.027 |
anced | +0.026 |
Neg logit contributions | |
mess | -0.027 |
neither | -0.026 |
wasteful | -0.023 |
fighting | -0.023 |
floor | -0.022 |
Token | boat |
Token ID | 8848 |
Feature activation | +3.80e-01 |
Pos logit contributions | |
neither | +0.013 |
mess | +0.012 |
wasteful | +0.012 |
overflow | +0.011 |
ruining | +0.011 |
Neg logit contributions | |
ka | -0.011 |
tie | -0.011 |
name | -0.011 |
ance | -0.010 |
dot | -0.010 |
Token | started |
Token ID | 2067 |
Feature activation | -1.88e-01 |
Pos logit contributions | |
neither | +0.016 |
mess | +0.016 |
wasteful | +0.015 |
overflow | +0.014 |
ruining | +0.014 |
Neg logit contributions | |
ka | -0.014 |
tie | -0.014 |
name | -0.014 |
aboard | -0.013 |
ance | -0.013 |
Token | to |
Token ID | 284 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
ka | +0.006 |
tie | +0.006 |
name | +0.006 |
ance | +0.006 |
dot | +0.006 |
Neg logit contributions | |
neither | -0.009 |
mess | -0.009 |
wasteful | -0.008 |
fighting | -0.008 |
overflow | -0.008 |
Token | fill |
Token ID | 6070 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
neither | +0.052 |
wasteful | +0.050 |
mess | +0.046 |
roughly | +0.045 |
rats | +0.045 |
Neg logit contributions | |
name | -0.039 |
aboard | -0.038 |
ka | -0.037 |
shone | -0.036 |
outer | -0.035 |
Token | with |
Token ID | 351 |
Feature activation | -1.55e-01 |
Pos logit contributions | |
neither | +0.054 |
wasteful | +0.052 |
mess | +0.049 |
ruining | +0.047 |
overflow | +0.047 |
Neg logit contributions | |
name | -0.032 |
ka | -0.032 |
aboard | -0.030 |
tie | -0.029 |
ance | -0.028 |
Token | , |
Token ID | 11 |
Feature activation | +2.14e+00 |
Pos logit contributions | |
neither | +0.042 |
overflow | +0.040 |
wasteful | +0.040 |
mess | +0.039 |
flood | +0.036 |
Neg logit contributions | |
name | -0.036 |
ka | -0.036 |
powers | -0.035 |
ance | -0.034 |
tie | -0.033 |
Token | sweet |
Token ID | 6029 |
Feature activation | +7.51e-01 |
Pos logit contributions | |
neither | +0.080 |
overflow | +0.077 |
erupted | +0.076 |
wasteful | +0.073 |
roughly | +0.073 |
Neg logit contributions | |
name | -0.087 |
Magic | -0.073 |
rib | -0.071 |
ka | -0.071 |
ti | -0.069 |
Token | ie |
Token ID | 494 |
Feature activation | +7.78e-01 |
Pos logit contributions | |
neither | +0.033 |
wasteful | +0.030 |
mess | +0.029 |
ruining | +0.029 |
overflow | +0.028 |
Neg logit contributions | |
ka | -0.028 |
name | -0.027 |
ance | -0.026 |
tie | -0.025 |
shirt | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
neither | +0.031 |
mess | +0.030 |
wasteful | +0.029 |
overflow | +0.028 |
ruining | +0.027 |
Neg logit contributions | |
ka | -0.030 |
name | -0.030 |
tie | -0.029 |
ance | -0.028 |
aboard | -0.028 |
Token | Painting |
Token ID | 48892 |
Feature activation | -3.00e-01 |
Pos logit contributions | |
ka | +0.072 |
tie | +0.071 |
aboard | +0.068 |
leaf | +0.067 |
ocket | +0.067 |
Neg logit contributions | |
neither | -0.051 |
wasteful | -0.046 |
mess | -0.041 |
fighting | -0.040 |
roughly | -0.039 |
Token | is |
Token ID | 318 |
Feature activation | +5.56e-01 |
Pos logit contributions | |
ka | +0.011 |
tie | +0.011 |
name | +0.010 |
dot | +0.010 |
ance | +0.010 |
Neg logit contributions | |
neither | -0.013 |
mess | -0.013 |
wasteful | -0.012 |
overflow | -0.011 |
fighting | -0.011 |
Token | fun |
Token ID | 1257 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
neither | +0.019 |
mess | +0.019 |
wasteful | +0.017 |
overflow | +0.017 |
fighting | +0.016 |
Neg logit contributions | |
name | -0.024 |
tie | -0.023 |
ka | -0.023 |
ance | -0.023 |
aboard | -0.023 |
Token | , |
Token ID | 11 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
neither | +0.061 |
wasteful | +0.057 |
overflow | +0.054 |
mess | +0.053 |
ruining | +0.051 |
Neg logit contributions | |
name | -0.041 |
powers | -0.041 |
ka | -0.039 |
tie | -0.039 |
aboard | -0.038 |
Token | but |
Token ID | 475 |
Feature activation | +4.90e-01 |
Pos logit contributions | |
neither | +0.040 |
wasteful | +0.038 |
mess | +0.037 |
overflow | +0.034 |
roughly | +0.033 |
Neg logit contributions | |
name | -0.061 |
tie | -0.057 |
ka | -0.055 |
shirt | -0.055 |
aboard | -0.053 |
Token | it |
Token ID | 340 |
Feature activation | +9.98e-01 |
Pos logit contributions | |
neither | +0.015 |
mess | +0.015 |
wasteful | +0.013 |
overflow | +0.012 |
roughly | +0.012 |
Neg logit contributions | |
name | -0.023 |
tie | -0.023 |
ka | -0.023 |
ance | -0.023 |
aboard | -0.022 |
Token | can |
Token ID | 460 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
neither | +0.042 |
mess | +0.040 |
wasteful | +0.040 |
overflow | +0.037 |
roughly | +0.037 |
Neg logit contributions | |
name | -0.035 |
ka | -0.035 |
ance | -0.034 |
tie | -0.033 |
aboard | -0.032 |
Token | saw |
Token ID | 2497 |
Feature activation | -7.97e-01 |
Pos logit contributions | |
mess | +0.026 |
neither | +0.026 |
wasteful | +0.025 |
overflow | +0.023 |
ruining | +0.022 |
Neg logit contributions | |
tie | -0.022 |
ka | -0.021 |
name | -0.021 |
aboard | -0.020 |
dot | -0.020 |
Token | that |
Token ID | 326 |
Feature activation | +2.87e-01 |
Pos logit contributions | |
ka | +0.031 |
tie | +0.030 |
shirt | +0.030 |
dot | +0.030 |
leaf | +0.029 |
Neg logit contributions | |
neither | -0.034 |
mess | -0.031 |
wasteful | -0.030 |
fighting | -0.029 |
rats | -0.027 |
Token | it |
Token ID | 340 |
Feature activation | +3.63e-01 |
Pos logit contributions | |
mess | +0.011 |
neither | +0.011 |
wasteful | +0.011 |
overflow | +0.010 |
ruining | +0.010 |
Neg logit contributions | |
name | -0.011 |
tie | -0.011 |
ka | -0.011 |
aboard | -0.011 |
ance | -0.011 |
Token | was |
Token ID | 373 |
Feature activation | -3.40e-01 |
Pos logit contributions | |
neither | +0.017 |
mess | +0.016 |
wasteful | +0.016 |
overflow | +0.014 |
roughly | +0.014 |
Neg logit contributions | |
ka | -0.012 |
tie | -0.012 |
name | -0.012 |
aboard | -0.011 |
ance | -0.011 |
Token | a |
Token ID | 257 |
Feature activation | -2.81e-01 |
Pos logit contributions | |
ka | +0.012 |
tie | +0.012 |
ance | +0.012 |
dot | +0.012 |
shirt | +0.012 |
Neg logit contributions | |
neither | -0.015 |
mess | -0.015 |
wasteful | -0.014 |
fighting | -0.013 |
ruining | -0.013 |
Token | whale |
Token ID | 22206 |
Feature activation | -4.43e-01 |
Pos logit contributions | |
ka | +0.011 |
tie | +0.011 |
ance | +0.011 |
name | +0.011 |
dot | +0.011 |
Neg logit contributions | |
neither | -0.011 |
mess | -0.011 |
wasteful | -0.010 |
overflow | -0.009 |
fighting | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | -8.46e-01 |
Pos logit contributions | |
ka | +0.017 |
tie | +0.017 |
ance | +0.017 |
name | +0.016 |
dot | +0.016 |
Neg logit contributions | |
neither | -0.018 |
mess | -0.018 |
wasteful | -0.016 |
fighting | -0.016 |
ruining | -0.015 |
Token |
|
Token ID | 198 |
Feature activation | -9.21e-01 |
Pos logit contributions | |
ka | +0.037 |
tie | +0.036 |
ance | +0.036 |
leaf | +0.036 |
powers | +0.035 |
Neg logit contributions | |
neither | -0.031 |
mess | -0.029 |
wasteful | -0.027 |
fighting | -0.026 |
overflow | -0.025 |
Token |
|
Token ID | 198 |
Feature activation | -8.16e-01 |
Pos logit contributions | |
ka | +0.043 |
ance | +0.042 |
tie | +0.042 |
anced | +0.041 |
dot | +0.041 |
Neg logit contributions | |
neither | -0.030 |
mess | -0.030 |
wasteful | -0.026 |
roughly | -0.025 |
fighting | -0.025 |
Token | The |
Token ID | 464 |
Feature activation | +4.14e-01 |
Pos logit contributions | |
ka | +0.036 |
ance | +0.036 |
anced | +0.035 |
tie | +0.035 |
leaf | +0.035 |
Neg logit contributions | |
mess | -0.028 |
neither | -0.028 |
wasteful | -0.026 |
roughly | -0.024 |
overflow | -0.023 |
Token | whale |
Token ID | 22206 |
Feature activation | +9.57e-01 |
Pos logit contributions | |
neither | +0.017 |
mess | +0.016 |
wasteful | +0.016 |
roughly | +0.015 |
ruining | +0.014 |
Neg logit contributions | |
name | -0.016 |
ka | -0.016 |
aboard | -0.016 |
tie | -0.015 |
leaf | -0.015 |
Token | saw |
Token ID | 2497 |
Feature activation | -7.11e-01 |
Pos logit contributions | |
neither | +0.000 |
mess | +0.000 |
wasteful | +0.000 |
overflow | +0.000 |
fighting | +0.000 |
Neg logit contributions | |
ka | -0.000 |
tie | -0.000 |
name | -0.000 |
ance | -0.000 |
dot | -0.000 |
Token | that |
Token ID | 326 |
Feature activation | +9.25e-02 |
Pos logit contributions | |
ka | +0.032 |
tie | +0.031 |
dot | +0.030 |
name | +0.030 |
shirt | +0.030 |
Neg logit contributions | |
neither | -0.026 |
mess | -0.025 |
wasteful | -0.023 |
fighting | -0.021 |
ruining | -0.021 |
Token | it |
Token ID | 340 |
Feature activation | -8.10e-02 |
Pos logit contributions | |
neither | +0.004 |
mess | +0.004 |
wasteful | +0.003 |
overflow | +0.003 |
ruining | +0.003 |
Neg logit contributions | |
ka | -0.004 |
tie | -0.004 |
name | -0.004 |
ance | -0.004 |
dot | -0.004 |
Token | was |
Token ID | 373 |
Feature activation | -4.11e-03 |
Pos logit contributions | |
ka | +0.003 |
tie | +0.003 |
name | +0.003 |
ance | +0.003 |
dot | +0.003 |
Neg logit contributions | |
neither | -0.003 |
mess | -0.003 |
wasteful | -0.003 |
overflow | -0.003 |
fighting | -0.003 |
Token | the |
Token ID | 262 |
Feature activation | +8.28e-02 |
Pos logit contributions | |
ka | +0.000 |
tie | +0.000 |
name | +0.000 |
ance | +0.000 |
dot | +0.000 |
Neg logit contributions | |
neither | -0.000 |
mess | -0.000 |
wasteful | -0.000 |
overflow | -0.000 |
fighting | -0.000 |
Token | perfect |
Token ID | 2818 |
Feature activation | -6.03e-01 |
Pos logit contributions | |
neither | +0.004 |
mess | +0.003 |
wasteful | +0.003 |
overflow | +0.003 |
fighting | +0.003 |
Neg logit contributions | |
ka | -0.003 |
name | -0.003 |
tie | -0.003 |
ance | -0.003 |
dot | -0.003 |
Token | size |
Token ID | 2546 |
Feature activation | -2.02e+00 |
Pos logit contributions | |
ka | +0.025 |
tie | +0.025 |
ance | +0.025 |
dot | +0.024 |
leaf | +0.024 |
Neg logit contributions | |
mess | -0.024 |
neither | -0.023 |
wasteful | -0.020 |
fighting | -0.019 |
overflow | -0.019 |
Token | . |
Token ID | 13 |
Feature activation | +6.43e-02 |
Pos logit contributions | |
ka | +0.090 |
name | +0.087 |
tie | +0.087 |
dot | +0.086 |
Aid | +0.085 |
Neg logit contributions | |
mess | -0.072 |
neither | -0.071 |
wasteful | -0.061 |
floor | -0.059 |
fighting | -0.058 |
Token | Sm |
Token ID | 2439 |
Feature activation | +2.69e-01 |
Pos logit contributions | |
neither | +0.003 |
mess | +0.003 |
wasteful | +0.003 |
overflow | +0.003 |
ruining | +0.002 |
Neg logit contributions | |
ka | -0.002 |
tie | -0.002 |
name | -0.002 |
ance | -0.002 |
dot | -0.002 |
Token | iling |
Token ID | 4386 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
neither | +0.008 |
mess | +0.008 |
wasteful | +0.007 |
overflow | +0.006 |
ruining | +0.006 |
Neg logit contributions | |
ka | -0.014 |
tie | -0.014 |
ance | -0.013 |
dot | -0.013 |
name | -0.013 |
Token | , |
Token ID | 11 |
Feature activation | +8.67e-01 |
Pos logit contributions | |
ka | +0.007 |
tie | +0.007 |
name | +0.007 |
dot | +0.007 |
ance | +0.007 |
Neg logit contributions | |
neither | -0.006 |
mess | -0.006 |
wasteful | -0.006 |
fighting | -0.005 |
overflow | -0.005 |
Token | gathered |
Token ID | 9272 |
Feature activation | -9.57e-01 |
Pos logit contributions | |
ka | +0.029 |
ance | +0.028 |
tie | +0.028 |
name | +0.028 |
dot | +0.028 |
Neg logit contributions | |
mess | -0.027 |
neither | -0.027 |
fighting | -0.025 |
wasteful | -0.025 |
ruining | -0.023 |
Token | around |
Token ID | 1088 |
Feature activation | +7.97e-02 |
Pos logit contributions | |
ka | +0.037 |
ance | +0.037 |
tie | +0.037 |
dot | +0.037 |
name | +0.036 |
Neg logit contributions | |
neither | -0.039 |
mess | -0.038 |
wasteful | -0.035 |
fighting | -0.033 |
roughly | -0.032 |
Token | a |
Token ID | 257 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
neither | +0.003 |
mess | +0.003 |
wasteful | +0.003 |
overflow | +0.003 |
ruining | +0.003 |
Neg logit contributions | |
ka | -0.003 |
tie | -0.003 |
name | -0.003 |
dot | -0.003 |
ance | -0.003 |
Token | funny |
Token ID | 8258 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
ka | +0.005 |
tie | +0.005 |
name | +0.005 |
ance | +0.005 |
dot | +0.005 |
Neg logit contributions | |
neither | -0.005 |
mess | -0.005 |
wasteful | -0.005 |
overflow | -0.004 |
fighting | -0.004 |
Token | - |
Token ID | 12 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
ance | +0.066 |
anced | +0.063 |
tie | +0.061 |
ka | +0.060 |
leaf | +0.059 |
Neg logit contributions | |
mess | -0.075 |
neither | -0.063 |
floor | -0.060 |
fighting | -0.060 |
quarrel | -0.059 |
Token | looking |
Token ID | 11534 |
Feature activation | -8.39e-01 |
Pos logit contributions | |
anced | +0.089 |
ance | +0.083 |
powers | +0.077 |
name | +0.077 |
amulet | +0.077 |
Neg logit contributions | |
neither | -0.075 |
roughly | -0.073 |
mess | -0.071 |
wasteful | -0.070 |
fighting | -0.065 |
Token | machine |
Token ID | 4572 |
Feature activation | -7.57e-01 |
Pos logit contributions | |
ka | +0.034 |
ance | +0.034 |
tie | +0.034 |
dot | +0.033 |
anced | +0.033 |
Neg logit contributions | |
mess | -0.033 |
neither | -0.031 |
wasteful | -0.029 |
fighting | -0.027 |
overflow | -0.027 |
Token | . |
Token ID | 13 |
Feature activation | +9.20e-01 |
Pos logit contributions | |
ka | +0.031 |
tie | +0.030 |
ance | +0.030 |
name | +0.030 |
dot | +0.030 |
Neg logit contributions | |
neither | -0.029 |
mess | -0.029 |
wasteful | -0.026 |
fighting | -0.025 |
ruining | -0.024 |
Token |
|
Token ID | 198 |
Feature activation | +6.51e-01 |
Pos logit contributions | |
mess | +0.047 |
neither | +0.047 |
wasteful | +0.046 |
ruining | +0.043 |
overflow | +0.043 |
Neg logit contributions | |
name | -0.028 |
tie | -0.025 |
aboard | -0.024 |
dot | -0.024 |
shirt | -0.024 |
Token |
|
Token ID | 198 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
neither | +0.031 |
mess | +0.030 |
wasteful | +0.030 |
ruining | +0.027 |
overflow | +0.027 |
Neg logit contributions | |
name | -0.022 |
tie | -0.021 |
ka | -0.021 |
aboard | -0.020 |
dot | -0.020 |
Token | John |
Token ID | 7554 |
Feature activation | +4.16e-01 |
Pos logit contributions | |
neither | +0.067 |
mess | +0.066 |
wasteful | +0.063 |
ruining | +0.059 |
snakes | +0.058 |
Neg logit contributions | |
name | -0.052 |
shirt | -0.051 |
tie | -0.051 |
dot | -0.050 |
leaf | -0.050 |
Token | path |
Token ID | 3108 |
Feature activation | -5.83e-01 |
Pos logit contributions | |
ka | +0.001 |
tie | +0.001 |
name | +0.001 |
ance | +0.001 |
dot | +0.001 |
Neg logit contributions | |
neither | -0.001 |
mess | -0.001 |
wasteful | -0.001 |
overflow | -0.001 |
fighting | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +7.99e-01 |
Pos logit contributions | |
ka | +0.024 |
tie | +0.024 |
name | +0.023 |
dot | +0.023 |
ance | +0.023 |
Neg logit contributions | |
neither | -0.023 |
mess | -0.022 |
wasteful | -0.020 |
fighting | -0.019 |
overflow | -0.019 |
Token | Suddenly |
Token ID | 24975 |
Feature activation | +6.09e-01 |
Pos logit contributions | |
mess | +0.031 |
neither | +0.028 |
wasteful | +0.028 |
ruining | +0.025 |
overflow | +0.025 |
Neg logit contributions | |
name | -0.036 |
tie | -0.034 |
ka | -0.033 |
dot | -0.033 |
anced | -0.033 |
Token | , |
Token ID | 11 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
mess | +0.026 |
neither | +0.024 |
wasteful | +0.024 |
overflow | +0.022 |
ruining | +0.022 |
Neg logit contributions | |
ka | -0.023 |
name | -0.022 |
ance | -0.022 |
tie | -0.022 |
aboard | -0.022 |
Token | the |
Token ID | 262 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
mess | +0.047 |
wasteful | +0.042 |
neither | +0.039 |
overflow | +0.038 |
ruining | +0.038 |
Neg logit contributions | |
aboard | -0.044 |
name | -0.044 |
ka | -0.044 |
tie | -0.043 |
ance | -0.042 |
Token | path |
Token ID | 3108 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
ka | +0.006 |
tie | +0.006 |
name | +0.006 |
dot | +0.006 |
ance | +0.006 |
Neg logit contributions | |
neither | -0.006 |
mess | -0.006 |
wasteful | -0.005 |
overflow | -0.005 |
fighting | -0.005 |
Token | said |
Token ID | 531 |
Feature activation | -9.36e-01 |
Pos logit contributions | |
ka | +0.008 |
tie | +0.007 |
name | +0.007 |
dot | +0.007 |
ance | +0.007 |
Neg logit contributions | |
neither | -0.009 |
mess | -0.009 |
wasteful | -0.008 |
fighting | -0.007 |
overflow | -0.007 |
Token | hello |
Token ID | 23748 |
Feature activation | -9.78e-01 |
Pos logit contributions | |
ka | +0.045 |
leaf | +0.044 |
tie | +0.043 |
shirt | +0.043 |
dot | +0.043 |
Neg logit contributions | |
neither | -0.033 |
mess | -0.028 |
wasteful | -0.027 |
fighting | -0.026 |
overflow | -0.024 |
Token | back |
Token ID | 736 |
Feature activation | -8.88e-01 |
Pos logit contributions | |
ka | +0.043 |
leaf | +0.043 |
dot | +0.041 |
shirt | +0.041 |
orn | +0.041 |
Neg logit contributions | |
neither | -0.036 |
mess | -0.034 |
fighting | -0.030 |
wasteful | -0.029 |
overflow | -0.028 |
Token | ! |
Token ID | 0 |
Feature activation | +5.39e-01 |
Pos logit contributions | |
ka | +0.040 |
leaf | +0.039 |
dot | +0.039 |
shirt | +0.039 |
tie | +0.038 |
Neg logit contributions | |
neither | -0.032 |
mess | -0.030 |
wasteful | -0.027 |
fighting | -0.027 |
roughly | -0.026 |
Token | They |
Token ID | 1119 |
Feature activation | -6.52e-02 |
Pos logit contributions | |
mess | +0.023 |
neither | +0.022 |
wasteful | +0.021 |
ruining | +0.019 |
overflow | +0.019 |
Neg logit contributions | |
name | -0.021 |
ka | -0.021 |
tie | -0.020 |
dot | -0.020 |
ance | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | -1.94e+00 |
Pos logit contributions | |
ka | +0.032 |
leaf | +0.032 |
dot | +0.031 |
ance | +0.030 |
aboard | +0.030 |
Neg logit contributions | |
neither | -0.027 |
mess | -0.026 |
wasteful | -0.023 |
fighting | -0.023 |
roughly | -0.022 |
Token | She |
Token ID | 1375 |
Feature activation | -3.10e-02 |
Pos logit contributions | |
leaf | +0.102 |
ka | +0.101 |
tie | +0.097 |
ocket | +0.097 |
shirt | +0.093 |
Neg logit contributions | |
neither | -0.056 |
mess | -0.050 |
structures | -0.047 |
fighting | -0.046 |
floor | -0.045 |
Token | was |
Token ID | 373 |
Feature activation | +3.83e-01 |
Pos logit contributions | |
ka | +0.001 |
tie | +0.001 |
name | +0.001 |
ance | +0.001 |
dot | +0.001 |
Neg logit contributions | |
neither | -0.001 |
mess | -0.001 |
wasteful | -0.001 |
overflow | -0.001 |
fighting | -0.001 |
Token | having |
Token ID | 1719 |
Feature activation | -2.09e-01 |
Pos logit contributions | |
neither | +0.014 |
mess | +0.014 |
wasteful | +0.013 |
overflow | +0.012 |
roughly | +0.012 |
Neg logit contributions | |
name | -0.016 |
ka | -0.016 |
tie | -0.015 |
dot | -0.015 |
leaf | -0.015 |
Token | so |
Token ID | 523 |
Feature activation | +1.47e-01 |
Pos logit contributions | |
ka | +0.009 |
tie | +0.009 |
name | +0.009 |
ance | +0.008 |
dot | +0.008 |
Neg logit contributions | |
neither | -0.008 |
mess | -0.008 |
wasteful | -0.007 |
fighting | -0.007 |
ruining | -0.007 |
Token | much |
Token ID | 881 |
Feature activation | -9.52e-01 |
Pos logit contributions | |
neither | +0.005 |
mess | +0.005 |
wasteful | +0.005 |
overflow | +0.004 |
ruining | +0.004 |
Neg logit contributions | |
ka | -0.007 |
tie | -0.007 |
name | -0.006 |
dot | -0.006 |
ance | -0.006 |
Token | fun |
Token ID | 1257 |
Feature activation | +1.01e-02 |
Pos logit contributions | |
ka | +0.059 |
ance | +0.058 |
tie | +0.057 |
anced | +0.057 |
ocket | +0.056 |
Neg logit contributions | |
mess | -0.021 |
neither | -0.017 |
wasteful | -0.014 |
fighting | -0.014 |
roughly | -0.013 |
Token | that |
Token ID | 326 |
Feature activation | -3.57e-01 |
Pos logit contributions | |
neither | +0.000 |
mess | +0.000 |
wasteful | +0.000 |
overflow | +0.000 |
fighting | +0.000 |
Neg logit contributions | |
ka | -0.000 |
tie | -0.000 |
name | -0.000 |
ance | -0.000 |
dot | -0.000 |
Token | she |
Token ID | 673 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
ka | +0.017 |
tie | +0.016 |
leaf | +0.016 |
name | +0.016 |
ance | +0.016 |
Neg logit contributions | |
neither | -0.012 |
mess | -0.012 |
wasteful | -0.011 |
fighting | -0.010 |
roughly | -0.010 |
Token | started |
Token ID | 2067 |
Feature activation | +2.17e-01 |
Pos logit contributions | |
neither | +0.048 |
wasteful | +0.046 |
mess | +0.046 |
structures | +0.044 |
overflow | +0.042 |
Neg logit contributions | |
ka | -0.052 |
tie | -0.049 |
name | -0.049 |
shirt | -0.048 |
u | -0.047 |
Token | to |
Token ID | 284 |
Feature activation | +5.51e-01 |
Pos logit contributions | |
neither | +0.010 |
mess | +0.010 |
wasteful | +0.010 |
overflow | +0.009 |
roughly | +0.009 |
Neg logit contributions | |
ka | -0.007 |
name | -0.007 |
tie | -0.007 |
dot | -0.007 |
ance | -0.006 |
Token | showed |
Token ID | 3751 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
ka | +0.021 |
ance | +0.021 |
tie | +0.021 |
name | +0.021 |
dot | +0.021 |
Neg logit contributions | |
neither | -0.026 |
mess | -0.026 |
wasteful | -0.024 |
fighting | -0.022 |
roughly | -0.022 |
Token | her |
Token ID | 607 |
Feature activation | -1.99e+00 |
Pos logit contributions | |
dot | +0.073 |
tie | +0.073 |
ocket | +0.071 |
leaf | +0.070 |
name | +0.070 |
Neg logit contributions | |
neither | -0.055 |
mess | -0.051 |
mice | -0.047 |
rats | -0.045 |
roughly | -0.044 |
Token | a |
Token ID | 257 |
Feature activation | -2.47e+00 |
Pos logit contributions | |
ka | +0.096 |
leaf | +0.093 |
dot | +0.092 |
tie | +0.091 |
orn | +0.090 |
Neg logit contributions | |
neither | -0.067 |
mess | -0.063 |
fighting | -0.054 |
not | -0.052 |
wasteful | -0.050 |
Token | beautiful |
Token ID | 4950 |
Feature activation | -3.79e+00 |
Pos logit contributions | |
ance | +0.119 |
anced | +0.110 |
ka | +0.102 |
dot | +0.101 |
tie | +0.101 |
Neg logit contributions | |
mess | -0.102 |
floor | -0.082 |
neither | -0.077 |
quarrel | -0.075 |
wasteful | -0.072 |
Token | dress |
Token ID | 6576 |
Feature activation | -2.34e+00 |
Pos logit contributions | |
ance | +0.188 |
anced | +0.174 |
aved | +0.165 |
leaf | +0.160 |
int | +0.153 |
Neg logit contributions | |
mess | -0.171 |
floor | -0.137 |
mistake | -0.121 |
table | -0.119 |
noise | -0.117 |
Token | with |
Token ID | 351 |
Feature activation | -2.65e+00 |
Pos logit contributions | |
ka | +0.128 |
leaf | +0.126 |
dot | +0.122 |
Aid | +0.121 |
ance | +0.119 |
Neg logit contributions | |
neither | -0.062 |
mess | -0.061 |
floor | -0.052 |
not | -0.047 |
fighting | -0.045 |
Token | colorful |
Token ID | 20239 |
Feature activation | -3.30e+00 |
Pos logit contributions | |
int | +0.115 |
anger | +0.111 |
orn | +0.109 |
Aid | +0.109 |
ance | +0.109 |
Neg logit contributions | |
mess | -0.093 |
neither | -0.092 |
floor | -0.081 |
fighting | -0.080 |
mice | -0.079 |
Token | flowers |
Token ID | 12734 |
Feature activation | -1.78e+00 |
Pos logit contributions | |
ance | +0.142 |
leaf | +0.132 |
anced | +0.131 |
int | +0.131 |
Aid | +0.130 |
Neg logit contributions | |
mess | -0.132 |
floor | -0.118 |
structures | -0.097 |
fighting | -0.095 |
table | -0.094 |
Token | on |
Token ID | 319 |
Feature activation | -8.56e-01 |
Pos logit contributions | |
ka | +0.087 |
leaf | +0.082 |
shirt | +0.082 |
ance | +0.082 |
Aid | +0.081 |
Neg logit contributions | |
neither | -0.055 |
mess | -0.054 |
fighting | -0.048 |
ruining | -0.044 |
floor | -0.043 |
Token | it |
Token ID | 340 |
Feature activation | -8.02e-01 |
Pos logit contributions | |
ka | +0.038 |
ance | +0.038 |
leaf | +0.038 |
dot | +0.037 |
tie | +0.037 |
Neg logit contributions | |
neither | -0.029 |
mess | -0.029 |
wasteful | -0.025 |
roughly | -0.024 |
fighting | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
ka | +0.034 |
dot | +0.032 |
tie | +0.032 |
leaf | +0.032 |
ance | +0.032 |
Neg logit contributions | |
neither | -0.031 |
mess | -0.030 |
wasteful | -0.027 |
fighting | -0.026 |
roughly | -0.025 |
Token | neck |
Token ID | 7393 |
Feature activation | -9.17e-01 |
Pos logit contributions | |
neither | +0.018 |
mess | +0.017 |
wasteful | +0.017 |
overflow | +0.015 |
ruining | +0.015 |
Neg logit contributions | |
name | -0.014 |
tie | -0.014 |
ka | -0.014 |
dot | -0.013 |
ance | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | -1.70e+00 |
Pos logit contributions | |
ka | +0.040 |
leaf | +0.038 |
dot | +0.038 |
ance | +0.037 |
name | +0.037 |
Neg logit contributions | |
neither | -0.034 |
mess | -0.033 |
wasteful | -0.029 |
roughly | -0.028 |
fighting | -0.027 |
Token | |
Token ID | 220 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
ka | +0.082 |
leaf | +0.080 |
tie | +0.079 |
orn | +0.077 |
ocket | +0.077 |
Neg logit contributions | |
neither | -0.056 |
mess | -0.050 |
floor | -0.044 |
fighting | -0.044 |
wasteful | -0.043 |
Token |
|
Token ID | 198 |
Feature activation | -7.57e-01 |
Pos logit contributions | |
ka | +0.052 |
tie | +0.049 |
leaf | +0.049 |
dot | +0.049 |
powers | +0.047 |
Neg logit contributions | |
neither | -0.037 |
mess | -0.034 |
wasteful | -0.031 |
fighting | -0.030 |
roughly | -0.029 |
Token |
|
Token ID | 198 |
Feature activation | -7.20e-01 |
Pos logit contributions | |
ka | +0.034 |
tie | +0.033 |
ance | +0.032 |
anced | +0.032 |
dot | +0.032 |
Neg logit contributions | |
neither | -0.026 |
mess | -0.025 |
wasteful | -0.022 |
fighting | -0.021 |
roughly | -0.021 |
Token | T |
Token ID | 51 |
Feature activation | -3.15e+00 |
Pos logit contributions | |
ka | +0.032 |
tie | +0.031 |
anced | +0.030 |
leaf | +0.030 |
ance | +0.030 |
Neg logit contributions | |
mess | -0.026 |
neither | -0.025 |
wasteful | -0.023 |
roughly | -0.022 |
overflow | -0.021 |
Token | illy |
Token ID | 6548 |
Feature activation | -4.21e-01 |
Pos logit contributions | |
leaf | +0.136 |
tie | +0.130 |
hy | +0.127 |
chery | +0.126 |
dot | +0.124 |
Neg logit contributions | |
mess | -0.119 |
neither | -0.089 |
floor | -0.086 |
anth | -0.085 |
wasteful | -0.085 |
Token | spent |
Token ID | 3377 |
Feature activation | +5.95e-01 |
Pos logit contributions | |
ka | +0.016 |
tie | +0.016 |
name | +0.016 |
dot | +0.016 |
ance | +0.016 |
Neg logit contributions | |
neither | -0.017 |
mess | -0.017 |
wasteful | -0.016 |
fighting | -0.015 |
overflow | -0.014 |
Token | every |
Token ID | 790 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
neither | +0.025 |
mess | +0.024 |
wasteful | +0.024 |
overflow | +0.022 |
rats | +0.020 |
Neg logit contributions | |
ka | -0.022 |
tie | -0.022 |
name | -0.022 |
dot | -0.021 |
ance | -0.021 |
Token | day |
Token ID | 1110 |
Feature activation | +4.83e-01 |
Pos logit contributions | |
ance | +0.056 |
anced | +0.054 |
ka | +0.054 |
name | +0.053 |
leaf | +0.052 |
Neg logit contributions | |
neither | -0.041 |
mess | -0.040 |
wasteful | -0.034 |
roughly | -0.034 |
fighting | -0.033 |
Token | playing |
Token ID | 2712 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
neither | +0.021 |
mess | +0.020 |
wasteful | +0.020 |
overflow | +0.018 |
rats | +0.018 |
Neg logit contributions | |
ka | -0.017 |
tie | -0.017 |
dot | -0.017 |
name | -0.016 |
ance | -0.016 |
Token | light |
Token ID | 1657 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
ka | +0.023 |
ance | +0.023 |
tie | +0.022 |
dot | +0.022 |
anced | +0.022 |
Neg logit contributions | |
mess | -0.029 |
neither | -0.029 |
wasteful | -0.027 |
fighting | -0.025 |
ruining | -0.025 |
Token | , |
Token ID | 11 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
ance | +0.081 |
ka | +0.081 |
dot | +0.081 |
leaf | +0.081 |
anced | +0.079 |
Neg logit contributions | |
mess | -0.079 |
neither | -0.075 |
fighting | -0.065 |
wasteful | -0.064 |
floor | -0.063 |
Token | like |
Token ID | 588 |
Feature activation | -1.77e+00 |
Pos logit contributions | |
ka | +0.043 |
ance | +0.043 |
tie | +0.041 |
dot | +0.041 |
anced | +0.041 |
Neg logit contributions | |
neither | -0.040 |
mess | -0.040 |
wasteful | -0.036 |
fighting | -0.034 |
ruining | -0.034 |
Token | a |
Token ID | 257 |
Feature activation | -8.98e-02 |
Pos logit contributions | |
ka | +0.072 |
tie | +0.070 |
shirt | +0.070 |
dot | +0.069 |
yo | +0.067 |
Neg logit contributions | |
neither | -0.069 |
mess | -0.066 |
fighting | -0.060 |
wasteful | -0.060 |
snakes | -0.059 |
Token | spark |
Token ID | 9009 |
Feature activation | -2.14e+00 |
Pos logit contributions | |
ka | +0.003 |
tie | +0.003 |
name | +0.003 |
ance | +0.003 |
dot | +0.003 |
Neg logit contributions | |
neither | -0.004 |
mess | -0.004 |
wasteful | -0.004 |
overflow | -0.004 |
fighting | -0.004 |
Token | ly |
Token ID | 306 |
Feature activation | -2.71e+00 |
Pos logit contributions | |
anced | +0.088 |
ance | +0.087 |
leaf | +0.087 |
dot | +0.086 |
lynn | +0.082 |
Neg logit contributions | |
mess | -0.087 |
neither | -0.075 |
anymore | -0.068 |
fighting | -0.066 |
floor | -0.066 |
Token | star |
Token ID | 3491 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
ance | +0.107 |
leaf | +0.100 |
anced | +0.099 |
dot | +0.095 |
iddled | +0.093 |
Neg logit contributions | |
mess | -0.121 |
floor | -0.105 |
neither | -0.098 |
snakes | -0.094 |
rats | -0.091 |
Token | . |
Token ID | 13 |
Feature activation | +8.72e-01 |
Pos logit contributions | |
leaf | +0.072 |
ka | +0.070 |
ance | +0.070 |
dot | +0.070 |
shirt | +0.069 |
Neg logit contributions | |
mess | -0.062 |
neither | -0.060 |
fighting | -0.052 |
wasteful | -0.050 |
ruining | -0.050 |
Token |
|
Token ID | 198 |
Feature activation | +7.78e-02 |
Pos logit contributions | |
neither | +0.043 |
mess | +0.042 |
wasteful | +0.042 |
overflow | +0.039 |
ruining | +0.038 |
Neg logit contributions | |
name | -0.028 |
aboard | -0.026 |
leaf | -0.026 |
tie | -0.026 |
dot | -0.025 |
Token |
|
Token ID | 198 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
neither | +0.003 |
mess | +0.003 |
wasteful | +0.003 |
overflow | +0.003 |
ruining | +0.003 |
Neg logit contributions | |
ka | -0.003 |
tie | -0.003 |
name | -0.003 |
dot | -0.003 |
ance | -0.003 |
Token | John |
Token ID | 7554 |
Feature activation | -7.98e-01 |
Pos logit contributions | |
neither | +0.058 |
mess | +0.054 |
wasteful | +0.052 |
overflow | +0.049 |
snakes | +0.047 |
Neg logit contributions | |
leaf | -0.044 |
tie | -0.044 |
shirt | -0.043 |
yo | -0.042 |
dot | -0.042 |
Token | , |
Token ID | 11 |
Feature activation | +8.27e-01 |
Pos logit contributions | |
ka | +0.010 |
tie | +0.010 |
name | +0.010 |
ance | +0.010 |
leaf | +0.010 |
Neg logit contributions | |
neither | -0.008 |
mess | -0.008 |
wasteful | -0.007 |
fighting | -0.007 |
overflow | -0.007 |
Token | Tim |
Token ID | 5045 |
Feature activation | +6.09e-01 |
Pos logit contributions | |
mess | +0.026 |
wasteful | +0.025 |
neither | +0.023 |
ruining | +0.020 |
overflow | +0.020 |
Neg logit contributions | |
name | -0.040 |
tie | -0.040 |
ka | -0.039 |
aboard | -0.038 |
dot | -0.038 |
Token | my |
Token ID | 1820 |
Feature activation | +1.19e+00 |
Pos logit contributions | |
neither | +0.030 |
wasteful | +0.028 |
mess | +0.028 |
overflow | +0.026 |
ruining | +0.026 |
Neg logit contributions | |
ka | -0.019 |
name | -0.018 |
tie | -0.018 |
ance | -0.018 |
orn | -0.017 |
Token | found |
Token ID | 1043 |
Feature activation | -8.34e-01 |
Pos logit contributions | |
neither | +0.055 |
wasteful | +0.053 |
mess | +0.049 |
ruining | +0.049 |
overflow | +0.047 |
Neg logit contributions | |
name | -0.040 |
ka | -0.038 |
tie | -0.038 |
aboard | -0.037 |
dot | -0.036 |
Token | a |
Token ID | 257 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
ka | +0.032 |
tie | +0.031 |
name | +0.031 |
dot | +0.031 |
anced | +0.031 |
Neg logit contributions | |
neither | -0.035 |
mess | -0.033 |
wasteful | -0.032 |
fighting | -0.029 |
rats | -0.029 |
Token | shiny |
Token ID | 22441 |
Feature activation | -3.08e+00 |
Pos logit contributions | |
ance | +0.064 |
anced | +0.062 |
ka | +0.059 |
dot | +0.058 |
leaf | +0.058 |
Neg logit contributions | |
mess | -0.055 |
neither | -0.048 |
wasteful | -0.045 |
fighting | -0.043 |
overflow | -0.042 |
Token | rock |
Token ID | 3881 |
Feature activation | -9.34e-01 |
Pos logit contributions | |
anced | +0.137 |
leaf | +0.137 |
ance | +0.136 |
dot | +0.127 |
found | +0.125 |
Neg logit contributions | |
mess | -0.137 |
floor | -0.107 |
neither | -0.105 |
wasteful | -0.098 |
fighting | -0.097 |
Token | on |
Token ID | 319 |
Feature activation | +3.74e-01 |
Pos logit contributions | |
ka | +0.040 |
name | +0.038 |
ance | +0.038 |
leaf | +0.038 |
anced | +0.038 |
Neg logit contributions | |
neither | -0.037 |
mess | -0.036 |
wasteful | -0.032 |
fighting | -0.032 |
ruining | -0.031 |
Token | the |
Token ID | 262 |
Feature activation | -2.30e-01 |
Pos logit contributions | |
neither | +0.015 |
mess | +0.015 |
wasteful | +0.015 |
overflow | +0.013 |
ruining | +0.013 |
Neg logit contributions | |
name | -0.014 |
ka | -0.014 |
tie | -0.014 |
ance | -0.013 |
aboard | -0.013 |
Token | ground |
Token ID | 2323 |
Feature activation | +3.28e-01 |
Pos logit contributions | |
ka | +0.013 |
tie | +0.013 |
name | +0.012 |
ance | +0.012 |
dot | +0.012 |
Neg logit contributions | |
mess | -0.006 |
neither | -0.006 |
wasteful | -0.005 |
overflow | -0.004 |
ruining | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -2.40e-01 |
Pos logit contributions | |
neither | +0.015 |
mess | +0.014 |
wasteful | +0.014 |
overflow | +0.013 |
ruining | +0.013 |
Neg logit contributions | |
tie | -0.011 |
ka | -0.011 |
name | -0.011 |
ance | -0.011 |
dot | -0.011 |
Token | ™ |
Token ID | 8151 |
Feature activation | +2.98e-01 |
Pos logit contributions | |
ka | +0.025 |
tie | +0.025 |
name | +0.024 |
leaf | +0.024 |
ance | +0.024 |
Neg logit contributions | |
neither | -0.028 |
mess | -0.027 |
wasteful | -0.025 |
fighting | -0.024 |
roughly | -0.024 |
Token | s |
Token ID | 82 |
Feature activation | -1.78e+00 |
Pos logit contributions | |
mess | +0.013 |
neither | +0.013 |
wasteful | +0.012 |
overflow | +0.011 |
ruining | +0.011 |
Neg logit contributions | |
ka | -0.011 |
tie | -0.010 |
name | -0.010 |
ance | -0.010 |
dot | -0.010 |
Token | new |
Token ID | 649 |
Feature activation | -1.98e+00 |
Pos logit contributions | |
ka | +0.082 |
dot | +0.078 |
orn | +0.077 |
tie | +0.076 |
shirt | +0.075 |
Neg logit contributions | |
neither | -0.065 |
mess | -0.064 |
fighting | -0.058 |
not | -0.054 |
ruining | -0.052 |
Token | skill |
Token ID | 5032 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
ance | +0.083 |
ka | +0.083 |
ocket | +0.083 |
tie | +0.081 |
anced | +0.081 |
Neg logit contributions | |
mess | -0.081 |
neither | -0.070 |
fighting | -0.064 |
quarrel | -0.064 |
mistake | -0.062 |
Token | . |
Token ID | 13 |
Feature activation | -1.80e+00 |
Pos logit contributions | |
ka | +0.072 |
name | +0.070 |
dot | +0.069 |
Aid | +0.067 |
tie | +0.067 |
Neg logit contributions | |
neither | -0.058 |
mess | -0.053 |
fighting | -0.047 |
wasteful | -0.047 |
roughly | -0.045 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.78e+00 |
Pos logit contributions | |
ka | +0.091 |
tie | +0.090 |
ocket | +0.086 |
powers | +0.085 |
leaf | +0.085 |
Neg logit contributions | |
neither | -0.054 |
mess | -0.045 |
fighting | -0.042 |
wasteful | -0.042 |
Amy | -0.039 |
Token | One |
Token ID | 3198 |
Feature activation | -1.00e+00 |
Pos logit contributions | |
tie | +0.129 |
ocket | +0.128 |
leaf | +0.128 |
ka | +0.127 |
dot | +0.126 |
Neg logit contributions | |
mess | -0.080 |
neither | -0.078 |
fighting | -0.076 |
Amy | -0.069 |
mice | -0.067 |
Token | day |
Token ID | 1110 |
Feature activation | -8.17e-01 |
Pos logit contributions | |
tie | +0.048 |
ka | +0.047 |
ance | +0.046 |
anced | +0.046 |
leaf | +0.046 |
Neg logit contributions | |
neither | -0.033 |
mess | -0.032 |
wasteful | -0.028 |
fighting | -0.026 |
roughly | -0.026 |
Token | , |
Token ID | 11 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
tie | +0.040 |
leaf | +0.039 |
ka | +0.039 |
dot | +0.038 |
ance | +0.038 |
Neg logit contributions | |
neither | -0.026 |
mess | -0.024 |
wasteful | -0.021 |
fighting | -0.021 |
roughly | -0.020 |
Token | a |
Token ID | 257 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
ance | +0.046 |
leaf | +0.046 |
ka | +0.045 |
ocket | +0.045 |
tie | +0.044 |
Neg logit contributions | |
neither | -0.044 |
mess | -0.038 |
fighting | -0.036 |
wasteful | -0.036 |
roughly | -0.035 |
Token | 3 |
Token ID | 513 |
Feature activation | +4.43e-01 |
Pos logit contributions | |
ance | +0.073 |
anced | +0.072 |
tie | +0.069 |
ka | +0.068 |
aboard | +0.067 |
Neg logit contributions | |
mess | -0.032 |
neither | -0.031 |
wasteful | -0.028 |
fighting | -0.024 |
roughly | -0.023 |