Token | was |
Token ID | 373 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
ley | +0.003 |
rant | +0.003 |
work | +0.002 |
ky | +0.002 |
ci | +0.002 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
cheating | -0.003 |
unch | -0.003 |
puck | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | +1.45e-03 |
Pos logit contributions | |
ley | +0.030 |
rant | +0.027 |
work | +0.027 |
ky | +0.026 |
ci | +0.025 |
Neg logit contributions | |
bananas | -0.014 |
banana | -0.013 |
knights | -0.011 |
unch | -0.011 |
catching | -0.011 |
Token | little |
Token ID | 1310 |
Feature activation | +2.87e-03 |
Pos logit contributions | |
ley | +0.003 |
rant | +0.003 |
work | +0.002 |
ky | +0.002 |
ci | +0.002 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.002 |
cheating | -0.002 |
unch | -0.002 |
puck | -0.002 |
Token | girl |
Token ID | 2576 |
Feature activation | +3.69e-03 |
Pos logit contributions | |
ley | +0.006 |
rant | +0.006 |
work | +0.006 |
ky | +0.005 |
ci | +0.005 |
Neg logit contributions | |
bananas | -0.004 |
banana | -0.004 |
knights | -0.004 |
puck | -0.004 |
cheating | -0.003 |
Token | called |
Token ID | 1444 |
Feature activation | +2.32e-03 |
Pos logit contributions | |
ley | +0.008 |
rant | +0.008 |
work | +0.007 |
ky | +0.007 |
ka | +0.006 |
Neg logit contributions | |
bananas | -0.005 |
banana | -0.005 |
swinging | -0.005 |
knights | -0.005 |
catching | -0.005 |
Token | Fel |
Token ID | 13937 |
Feature activation | +6.60e-02 |
Pos logit contributions | |
ley | +0.006 |
rant | +0.005 |
work | +0.005 |
ky | +0.005 |
ci | +0.005 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.002 |
cheating | -0.002 |
unch | -0.002 |
knights | -0.002 |
Token | icia |
Token ID | 33577 |
Feature activation | +7.52e-03 |
Pos logit contributions | |
work | +0.098 |
streaming | +0.093 |
ley | +0.093 |
ci | +0.091 |
rant | +0.090 |
Neg logit contributions | |
knights | -0.107 |
cheating | -0.102 |
court | -0.097 |
math | -0.097 |
beetles | -0.095 |
Token | . |
Token ID | 13 |
Feature activation | -6.42e-04 |
Pos logit contributions | |
rant | +0.017 |
ley | +0.017 |
work | +0.015 |
ky | +0.015 |
ka | +0.015 |
Neg logit contributions | |
bananas | -0.009 |
banana | -0.009 |
knights | -0.009 |
he | -0.008 |
stealing | -0.008 |
Token | She |
Token ID | 1375 |
Feature activation | -3.53e-04 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.001 |
Ready | +0.001 |
cheating | +0.001 |
knights | +0.001 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
work | -0.001 |
ky | -0.001 |
ci | -0.001 |
Token | loved |
Token ID | 6151 |
Feature activation | +7.30e-03 |
Pos logit contributions | |
he | +0.001 |
banana | +0.001 |
bananas | +0.000 |
chewing | +0.000 |
catching | +0.000 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
work | -0.001 |
extingu | -0.001 |
ky | -0.001 |
Token | to |
Token ID | 284 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
ley | +0.014 |
rant | +0.012 |
work | +0.012 |
ky | +0.012 |
ci | +0.011 |
Neg logit contributions | |
bananas | -0.013 |
banana | -0.012 |
cheating | -0.012 |
unch | -0.011 |
puck | -0.011 |
Token | cute |
Token ID | 13779 |
Feature activation | +2.09e-02 |
Pos logit contributions | |
ley | +0.002 |
rant | +0.002 |
work | +0.002 |
ky | +0.002 |
ci | +0.001 |
Neg logit contributions | |
bananas | -0.002 |
banana | -0.001 |
cheating | -0.001 |
unch | -0.001 |
puck | -0.001 |
Token | pink |
Token ID | 11398 |
Feature activation | +5.70e-03 |
Pos logit contributions | |
ley | +0.047 |
rant | +0.047 |
work | +0.042 |
ky | +0.041 |
ci | +0.041 |
Neg logit contributions | |
banana | -0.030 |
bananas | -0.029 |
knights | -0.026 |
mango | -0.024 |
puck | -0.024 |
Token | rabbit |
Token ID | 22746 |
Feature activation | +9.34e-03 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.012 |
work | +0.011 |
ky | +0.011 |
sy | +0.011 |
Neg logit contributions | |
banana | -0.008 |
bananas | -0.008 |
puck | -0.007 |
knights | -0.007 |
mango | -0.007 |
Token | called |
Token ID | 1444 |
Feature activation | +3.84e-03 |
Pos logit contributions | |
ley | +0.018 |
rant | +0.018 |
work | +0.016 |
ky | +0.015 |
ka | +0.014 |
Neg logit contributions | |
bananas | -0.016 |
banana | -0.015 |
knights | -0.014 |
he | -0.014 |
unch | -0.014 |
Token | Mrs |
Token ID | 9074 |
Feature activation | +2.72e-02 |
Pos logit contributions | |
ley | +0.008 |
rant | +0.008 |
work | +0.007 |
ky | +0.007 |
ci | +0.007 |
Neg logit contributions | |
bananas | -0.006 |
banana | -0.005 |
cheating | -0.005 |
unch | -0.005 |
puck | -0.005 |
Token | Fl |
Token ID | 1610 |
Feature activation | +5.96e-02 |
Pos logit contributions | |
ley | +0.051 |
rant | +0.048 |
work | +0.045 |
ky | +0.044 |
ci | +0.043 |
Neg logit contributions | |
bananas | -0.048 |
banana | -0.044 |
knights | -0.042 |
cheating | -0.042 |
mango | -0.041 |
Token | uffy |
Token ID | 15352 |
Feature activation | +2.11e-02 |
Pos logit contributions | |
rant | +0.123 |
ley | +0.116 |
work | +0.108 |
ky | +0.105 |
ened | +0.101 |
Neg logit contributions | |
unch | -0.082 |
banana | -0.079 |
knights | -0.079 |
bananas | -0.078 |
puck | -0.076 |
Token | . |
Token ID | 13 |
Feature activation | +3.22e-03 |
Pos logit contributions | |
ley | +0.043 |
rant | +0.042 |
work | +0.038 |
ky | +0.037 |
ka | +0.036 |
Neg logit contributions | |
bananas | -0.034 |
banana | -0.032 |
knights | -0.030 |
he | -0.028 |
cheating | -0.028 |
Token | She |
Token ID | 1375 |
Feature activation | -2.52e-03 |
Pos logit contributions | |
ley | +0.006 |
rant | +0.005 |
work | +0.005 |
ky | +0.005 |
ci | +0.004 |
Neg logit contributions | |
bananas | -0.006 |
banana | -0.006 |
cheating | -0.006 |
unch | -0.005 |
puck | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | +5.87e-03 |
Pos logit contributions | |
he | +0.004 |
banana | +0.003 |
bananas | +0.003 |
catching | +0.003 |
chewing | +0.003 |
Neg logit contributions | |
rant | -0.005 |
ley | -0.005 |
work | -0.005 |
extingu | -0.005 |
ky | -0.005 |
Token | small |
Token ID | 1402 |
Feature activation | +9.97e-04 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.012 |
work | +0.012 |
ky | +0.011 |
ci | +0.011 |
Neg logit contributions | |
bananas | -0.009 |
banana | -0.008 |
cheating | -0.008 |
catching | -0.008 |
knights | -0.008 |
Token | M |
Token ID | 44 |
Feature activation | +2.41e-02 |
Pos logit contributions | |
ley | +0.008 |
rant | +0.007 |
work | +0.007 |
ky | +0.007 |
ci | +0.006 |
Neg logit contributions | |
bananas | -0.004 |
banana | -0.004 |
cheating | -0.004 |
unch | -0.004 |
puck | -0.004 |
Token | um |
Token ID | 388 |
Feature activation | +6.10e-03 |
Pos logit contributions | |
ley | +0.057 |
rant | +0.055 |
ky | +0.052 |
work | +0.050 |
extingu | +0.049 |
Neg logit contributions | |
bananas | -0.032 |
unch | -0.032 |
banana | -0.030 |
mango | -0.027 |
cheating | -0.027 |
Token | and |
Token ID | 290 |
Feature activation | +1.30e-02 |
Pos logit contributions | |
ley | +0.012 |
rant | +0.011 |
work | +0.011 |
ky | +0.010 |
ci | +0.010 |
Neg logit contributions | |
bananas | -0.010 |
banana | -0.009 |
he | -0.009 |
catching | -0.009 |
knights | -0.008 |
Token | Dad |
Token ID | 17415 |
Feature activation | +1.75e-02 |
Pos logit contributions | |
ley | +0.029 |
rant | +0.027 |
work | +0.025 |
ky | +0.025 |
ci | +0.023 |
Neg logit contributions | |
bananas | -0.020 |
banana | -0.018 |
knights | -0.016 |
cheating | -0.016 |
catching | -0.016 |
Token | took |
Token ID | 1718 |
Feature activation | +2.00e-02 |
Pos logit contributions | |
ley | +0.034 |
rant | +0.032 |
work | +0.031 |
ky | +0.030 |
ci | +0.029 |
Neg logit contributions | |
bananas | -0.029 |
banana | -0.026 |
he | -0.026 |
catching | -0.024 |
unch | -0.024 |
Token | Little |
Token ID | 7703 |
Feature activation | +5.90e-02 |
Pos logit contributions | |
ley | +0.049 |
rant | +0.047 |
work | +0.044 |
sy | +0.042 |
ky | +0.042 |
Neg logit contributions | |
bananas | -0.026 |
banana | -0.021 |
mango | -0.019 |
he | -0.019 |
knights | -0.019 |
Token | Jack |
Token ID | 3619 |
Feature activation | +2.02e-02 |
Pos logit contributions | |
rant | +0.119 |
work | +0.109 |
ley | +0.109 |
ilt | +0.102 |
bul | +0.101 |
Neg logit contributions | |
he | -0.087 |
bananas | -0.085 |
oh | -0.075 |
banana | -0.074 |
Tom | -0.073 |
Token | to |
Token ID | 284 |
Feature activation | +1.26e-02 |
Pos logit contributions | |
rant | +0.048 |
ley | +0.047 |
work | +0.042 |
ky | +0.041 |
ka | +0.041 |
Neg logit contributions | |
bananas | -0.024 |
banana | -0.022 |
he | -0.021 |
chewing | -0.020 |
math | -0.019 |
Token | the |
Token ID | 262 |
Feature activation | +9.20e-03 |
Pos logit contributions | |
ley | +0.029 |
rant | +0.027 |
work | +0.026 |
ky | +0.026 |
ci | +0.024 |
Neg logit contributions | |
bananas | -0.017 |
banana | -0.015 |
cheating | -0.015 |
unch | -0.014 |
puck | -0.014 |
Token | park |
Token ID | 3952 |
Feature activation | +1.05e-02 |
Pos logit contributions | |
ley | +0.023 |
rant | +0.021 |
work | +0.020 |
ky | +0.020 |
ci | +0.019 |
Neg logit contributions | |
bananas | -0.012 |
banana | -0.011 |
cheating | -0.010 |
puck | -0.010 |
knights | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | +1.57e-02 |
Pos logit contributions | |
ley | +0.020 |
rant | +0.018 |
work | +0.017 |
ky | +0.016 |
ci | +0.015 |
Neg logit contributions | |
bananas | -0.020 |
banana | -0.018 |
cheating | -0.017 |
unch | -0.017 |
puck | -0.017 |
Token | . |
Token ID | 13 |
Feature activation | +1.27e-02 |
Pos logit contributions | |
bananas | +0.004 |
banana | +0.004 |
catching | +0.004 |
cheating | +0.003 |
knights | +0.003 |
Neg logit contributions | |
ley | -0.005 |
rant | -0.004 |
work | -0.004 |
ky | -0.004 |
ci | -0.004 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +1.99e-03 |
Pos logit contributions | |
ley | +0.032 |
rant | +0.030 |
work | +0.029 |
ky | +0.028 |
ci | +0.027 |
Neg logit contributions | |
bananas | -0.015 |
banana | -0.013 |
unch | -0.013 |
Ready | -0.012 |
cheating | -0.012 |
Token | Mom |
Token ID | 29252 |
Feature activation | +5.25e-03 |
Pos logit contributions | |
ley | +0.005 |
rant | +0.004 |
work | +0.004 |
ky | +0.004 |
ci | +0.004 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.002 |
unch | -0.002 |
cheating | -0.002 |
puck | -0.002 |
Token | my |
Token ID | 1820 |
Feature activation | +6.66e-03 |
Pos logit contributions | |
ley | +0.011 |
rant | +0.010 |
work | +0.010 |
ky | +0.010 |
ci | +0.009 |
Neg logit contributions | |
bananas | -0.008 |
banana | -0.007 |
he | -0.007 |
knights | -0.006 |
mango | -0.006 |
Token | took |
Token ID | 1718 |
Feature activation | +2.20e-02 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.013 |
work | +0.012 |
ky | +0.011 |
sy | +0.011 |
Neg logit contributions | |
bananas | -0.011 |
banana | -0.010 |
he | -0.010 |
catching | -0.009 |
knights | -0.009 |
Token | Little |
Token ID | 7703 |
Feature activation | +5.42e-02 |
Pos logit contributions | |
ley | +0.060 |
rant | +0.056 |
work | +0.054 |
mer | +0.051 |
ky | +0.051 |
Neg logit contributions | |
bananas | -0.024 |
banana | -0.019 |
math | -0.018 |
he | -0.017 |
mango | -0.017 |
Token | Johnny |
Token ID | 15470 |
Feature activation | +1.71e-02 |
Pos logit contributions | |
rant | +0.103 |
work | +0.095 |
ley | +0.091 |
ilt | +0.084 |
attered | +0.082 |
Neg logit contributions | |
he | -0.086 |
bananas | -0.084 |
climbing | -0.075 |
oh | -0.075 |
banana | -0.073 |
Token | 's |
Token ID | 338 |
Feature activation | +1.28e-02 |
Pos logit contributions | |
rant | +0.035 |
ley | +0.034 |
work | +0.030 |
ky | +0.029 |
ka | +0.029 |
Neg logit contributions | |
bananas | -0.025 |
he | -0.023 |
banana | -0.022 |
swinging | -0.021 |
chewing | -0.021 |
Token | hand |
Token ID | 1021 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
rant | +0.030 |
ley | +0.029 |
ky | +0.026 |
work | +0.024 |
sy | +0.024 |
Neg logit contributions | |
bananas | -0.018 |
banana | -0.017 |
he | -0.016 |
math | -0.015 |
Ready | -0.015 |
Token | in |
Token ID | 287 |
Feature activation | +2.88e-02 |
Pos logit contributions | |
ley | +0.040 |
rant | +0.039 |
work | +0.034 |
ky | +0.033 |
ka | +0.033 |
Neg logit contributions | |
bananas | -0.028 |
he | -0.025 |
banana | -0.025 |
chewing | -0.023 |
knights | -0.022 |
Token | hers |
Token ID | 25144 |
Feature activation | +9.42e-04 |
Pos logit contributions | |
ley | +0.078 |
rant | +0.075 |
work | +0.071 |
ky | +0.068 |
sy | +0.065 |
Neg logit contributions | |
bananas | -0.030 |
banana | -0.026 |
he | -0.024 |
cheating | -0.023 |
catching | -0.023 |
Token | was |
Token ID | 373 |
Feature activation | +1.01e-02 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.001 |
cheating | +0.001 |
unch | +0.001 |
puck | +0.001 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
work | -0.000 |
ky | -0.000 |
ci | -0.000 |
Token | a |
Token ID | 257 |
Feature activation | +1.17e-03 |
Pos logit contributions | |
ley | +0.026 |
rant | +0.024 |
work | +0.023 |
ky | +0.023 |
ci | +0.022 |
Neg logit contributions | |
bananas | -0.012 |
banana | -0.011 |
knights | -0.010 |
unch | -0.010 |
catching | -0.010 |
Token | little |
Token ID | 1310 |
Feature activation | +3.89e-03 |
Pos logit contributions | |
ley | +0.002 |
rant | +0.002 |
work | +0.002 |
ky | +0.002 |
ci | +0.002 |
Neg logit contributions | |
bananas | -0.002 |
banana | -0.002 |
cheating | -0.002 |
unch | -0.002 |
puck | -0.002 |
Token | girl |
Token ID | 2576 |
Feature activation | +4.74e-03 |
Pos logit contributions | |
ley | +0.008 |
rant | +0.008 |
work | +0.008 |
ky | +0.007 |
ci | +0.007 |
Neg logit contributions | |
bananas | -0.006 |
banana | -0.006 |
knights | -0.005 |
puck | -0.005 |
cheating | -0.005 |
Token | named |
Token ID | 3706 |
Feature activation | +5.33e-05 |
Pos logit contributions | |
ley | +0.011 |
rant | +0.011 |
work | +0.010 |
ky | +0.009 |
iful | +0.009 |
Neg logit contributions | |
banana | -0.006 |
bananas | -0.006 |
swinging | -0.006 |
knights | -0.006 |
stealing | -0.006 |
Token | Beat |
Token ID | 12568 |
Feature activation | +5.42e-02 |
Pos logit contributions | |
ley | +0.000 |
rant | +0.000 |
work | +0.000 |
ky | +0.000 |
ci | +0.000 |
Neg logit contributions | |
bananas | -0.000 |
banana | -0.000 |
cheating | -0.000 |
unch | -0.000 |
catching | -0.000 |
Token | rice |
Token ID | 20970 |
Feature activation | -5.39e-04 |
Pos logit contributions | |
rant | +0.113 |
ley | +0.106 |
work | +0.104 |
extingu | +0.095 |
streaming | +0.094 |
Neg logit contributions | |
shadow | -0.071 |
bananas | -0.062 |
catching | -0.062 |
cheating | -0.061 |
Ready | -0.060 |
Token | . |
Token ID | 13 |
Feature activation | -7.59e-03 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.001 |
he | +0.001 |
knights | +0.001 |
swinging | +0.001 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
work | -0.001 |
ky | -0.001 |
ka | -0.001 |
Token | She |
Token ID | 1375 |
Feature activation | -1.10e-03 |
Pos logit contributions | |
bananas | +0.010 |
cheating | +0.009 |
Ready | +0.009 |
banana | +0.009 |
knights | +0.009 |
Neg logit contributions | |
ley | -0.017 |
rant | -0.016 |
work | -0.016 |
ky | -0.015 |
ci | -0.014 |
Token | was |
Token ID | 373 |
Feature activation | +1.75e-03 |
Pos logit contributions | |
banana | +0.002 |
he | +0.002 |
bananas | +0.002 |
chewing | +0.001 |
knights | +0.001 |
Neg logit contributions | |
ley | -0.002 |
rant | -0.002 |
work | -0.002 |
extingu | -0.002 |
ky | -0.002 |
Token | so |
Token ID | 523 |
Feature activation | -7.42e-03 |
Pos logit contributions | |
ley | +0.004 |
rant | +0.003 |
work | +0.003 |
ky | +0.003 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
cheating | -0.002 |
knights | -0.002 |
catching | -0.002 |
Token | Annie |
Token ID | 25860 |
Feature activation | +6.60e-03 |
Pos logit contributions | |
ley | +0.020 |
rant | +0.019 |
work | +0.018 |
ky | +0.017 |
ci | +0.016 |
Neg logit contributions | |
bananas | -0.015 |
banana | -0.014 |
cheating | -0.013 |
unch | -0.013 |
puck | -0.013 |
Token | wanted |
Token ID | 2227 |
Feature activation | +1.09e-02 |
Pos logit contributions | |
rant | +0.015 |
work | +0.014 |
extingu | +0.014 |
ley | +0.014 |
ky | +0.013 |
Neg logit contributions | |
he | -0.008 |
banana | -0.007 |
bananas | -0.007 |
catching | -0.007 |
knights | -0.007 |
Token | to |
Token ID | 284 |
Feature activation | -2.26e-03 |
Pos logit contributions | |
ley | +0.020 |
rant | +0.018 |
work | +0.018 |
ky | +0.017 |
ci | +0.016 |
Neg logit contributions | |
bananas | -0.020 |
banana | -0.018 |
unch | -0.018 |
cheating | -0.018 |
puck | -0.017 |
Token | come |
Token ID | 1282 |
Feature activation | +1.82e-03 |
Pos logit contributions | |
bananas | +0.004 |
banana | +0.003 |
cheating | +0.003 |
unch | +0.003 |
puck | +0.003 |
Neg logit contributions | |
ley | -0.005 |
rant | -0.004 |
work | -0.004 |
ky | -0.004 |
ci | -0.004 |
Token | with |
Token ID | 351 |
Feature activation | +1.39e-02 |
Pos logit contributions | |
ley | +0.003 |
rant | +0.003 |
work | +0.003 |
ky | +0.003 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
cheating | -0.003 |
unch | -0.003 |
puck | -0.003 |
Token | Sp |
Token ID | 1338 |
Feature activation | +5.41e-02 |
Pos logit contributions | |
ley | +0.031 |
rant | +0.029 |
work | +0.028 |
ky | +0.027 |
ci | +0.026 |
Neg logit contributions | |
bananas | -0.020 |
banana | -0.018 |
cheating | -0.017 |
knights | -0.017 |
catching | -0.017 |
Token | ide |
Token ID | 485 |
Feature activation | +1.17e-02 |
Pos logit contributions | |
ley | +0.123 |
work | +0.110 |
rant | +0.109 |
ky | +0.105 |
ci | +0.103 |
Neg logit contributions | |
unch | -0.072 |
bananas | -0.071 |
banana | -0.064 |
cheating | -0.064 |
and | -0.060 |
Token | y |
Token ID | 88 |
Feature activation | -2.40e-03 |
Pos logit contributions | |
ley | +0.012 |
rant | +0.011 |
work | +0.010 |
ky | +0.009 |
ci | +0.008 |
Neg logit contributions | |
bananas | -0.030 |
banana | -0.029 |
knights | -0.028 |
cheating | -0.027 |
beetles | -0.027 |
Token | and |
Token ID | 290 |
Feature activation | -1.56e-03 |
Pos logit contributions | |
bananas | +0.004 |
banana | +0.004 |
cheating | +0.004 |
unch | +0.004 |
puck | +0.004 |
Neg logit contributions | |
ley | -0.005 |
rant | -0.004 |
work | -0.004 |
ky | -0.004 |
ci | -0.004 |
Token | explore |
Token ID | 7301 |
Feature activation | -2.24e-03 |
Pos logit contributions | |
bananas | +0.003 |
unch | +0.002 |
banana | +0.002 |
cheating | +0.002 |
puck | +0.002 |
Neg logit contributions | |
ley | -0.003 |
rant | -0.003 |
work | -0.003 |
ky | -0.003 |
ci | -0.003 |
Token | the |
Token ID | 262 |
Feature activation | -6.82e-03 |
Pos logit contributions | |
bananas | +0.004 |
banana | +0.004 |
unch | +0.004 |
puck | +0.003 |
cheating | +0.003 |
Neg logit contributions | |
ley | -0.004 |
rant | -0.004 |
work | -0.004 |
ky | -0.004 |
ci | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | -6.91e-04 |
Pos logit contributions | |
ley | +0.037 |
rant | +0.033 |
work | +0.033 |
ky | +0.032 |
ci | +0.031 |
Neg logit contributions | |
bananas | -0.027 |
unch | -0.025 |
cheating | -0.025 |
banana | -0.024 |
puck | -0.023 |
Token | potato |
Token ID | 21219 |
Feature activation | -1.90e-03 |
Pos logit contributions | |
bananas | +0.001 |
cheating | +0.001 |
banana | +0.001 |
unch | +0.001 |
puck | +0.001 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
work | -0.001 |
ky | -0.001 |
ci | -0.001 |
Token | named |
Token ID | 3706 |
Feature activation | -1.07e-02 |
Pos logit contributions | |
bananas | +0.003 |
banana | +0.003 |
knights | +0.003 |
puck | +0.003 |
cheating | +0.003 |
Neg logit contributions | |
ley | -0.004 |
rant | -0.004 |
work | -0.003 |
ky | -0.003 |
ci | -0.003 |
Token | Mr |
Token ID | 1770 |
Feature activation | +2.60e-02 |
Pos logit contributions | |
bananas | +0.016 |
banana | +0.014 |
puck | +0.013 |
unch | +0.013 |
cheating | +0.013 |
Neg logit contributions | |
ley | -0.024 |
rant | -0.022 |
work | -0.021 |
ky | -0.021 |
ci | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | +5.16e-04 |
Pos logit contributions | |
ley | +0.050 |
rant | +0.047 |
work | +0.043 |
ky | +0.043 |
ci | +0.041 |
Neg logit contributions | |
bananas | -0.044 |
banana | -0.041 |
cheating | -0.040 |
knights | -0.040 |
unch | -0.038 |
Token | Sp |
Token ID | 1338 |
Feature activation | +5.37e-02 |
Pos logit contributions | |
ley | +0.001 |
rant | +0.001 |
work | +0.001 |
ky | +0.001 |
ci | +0.001 |
Neg logit contributions | |
bananas | -0.001 |
banana | -0.001 |
cheating | -0.001 |
unch | -0.001 |
puck | -0.001 |
Token | ud |
Token ID | 463 |
Feature activation | +1.41e-02 |
Pos logit contributions | |
ley | +0.113 |
rant | +0.106 |
work | +0.101 |
ky | +0.100 |
streaming | +0.096 |
Neg logit contributions | |
unch | -0.076 |
bananas | -0.070 |
banana | -0.070 |
and | -0.065 |
cheating | -0.065 |
Token | . |
Token ID | 13 |
Feature activation | -5.77e-03 |
Pos logit contributions | |
rant | +0.026 |
ley | +0.026 |
work | +0.024 |
ky | +0.023 |
extingu | +0.022 |
Neg logit contributions | |
bananas | -0.023 |
banana | -0.022 |
he | -0.021 |
knights | -0.021 |
catching | -0.020 |
Token | They |
Token ID | 1119 |
Feature activation | +6.85e-03 |
Pos logit contributions | |
bananas | +0.010 |
cheating | +0.010 |
banana | +0.009 |
catching | +0.009 |
unch | +0.009 |
Neg logit contributions | |
ley | -0.011 |
rant | -0.010 |
work | -0.010 |
ky | -0.009 |
ci | -0.009 |
Token | had |
Token ID | 550 |
Feature activation | +2.63e-03 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.013 |
work | +0.013 |
ky | +0.012 |
icol | +0.012 |
Neg logit contributions | |
bananas | -0.011 |
banana | -0.011 |
knights | -0.010 |
he | -0.010 |
beetles | -0.010 |
Token | lots |
Token ID | 6041 |
Feature activation | +1.42e-03 |
Pos logit contributions | |
ley | +0.006 |
rant | +0.006 |
work | +0.006 |
ky | +0.005 |
ci | +0.005 |
Neg logit contributions | |
bananas | -0.004 |
banana | -0.003 |
cheating | -0.003 |
puck | -0.003 |
knights | -0.003 |
Token | was |
Token ID | 373 |
Feature activation | +9.85e-03 |
Pos logit contributions | |
ley | +0.000 |
rant | +0.000 |
work | +0.000 |
ky | +0.000 |
ci | +0.000 |
Neg logit contributions | |
bananas | -0.000 |
banana | -0.000 |
cheating | -0.000 |
unch | -0.000 |
puck | -0.000 |
Token | a |
Token ID | 257 |
Feature activation | -1.28e-05 |
Pos logit contributions | |
ley | +0.023 |
rant | +0.021 |
work | +0.020 |
ky | +0.020 |
ci | +0.019 |
Neg logit contributions | |
bananas | -0.014 |
banana | -0.013 |
cheating | -0.012 |
unch | -0.012 |
puck | -0.012 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.20e-02 |
Pos logit contributions | |
bananas | +0.000 |
cheating | +0.000 |
unch | +0.000 |
banana | +0.000 |
catching | +0.000 |
Neg logit contributions | |
ley | -0.000 |
rant | -0.000 |
work | -0.000 |
ky | -0.000 |
ci | -0.000 |
Token | potato |
Token ID | 21219 |
Feature activation | +6.51e-03 |
Pos logit contributions | |
ley | +0.026 |
rant | +0.024 |
work | +0.023 |
ky | +0.023 |
ci | +0.021 |
Neg logit contributions | |
bananas | -0.018 |
banana | -0.017 |
unch | -0.016 |
cheating | -0.016 |
puck | -0.016 |
Token | named |
Token ID | 3706 |
Feature activation | +3.52e-03 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.012 |
ky | +0.011 |
work | +0.011 |
ka | +0.010 |
Neg logit contributions | |
bananas | -0.011 |
banana | -0.011 |
knights | -0.010 |
unch | -0.010 |
catching | -0.010 |
Token | Sp |
Token ID | 1338 |
Feature activation | +5.37e-02 |
Pos logit contributions | |
ley | +0.007 |
rant | +0.007 |
work | +0.006 |
ky | +0.006 |
ci | +0.006 |
Neg logit contributions | |
bananas | -0.006 |
banana | -0.005 |
puck | -0.005 |
unch | -0.005 |
cheating | -0.005 |
Token | ud |
Token ID | 463 |
Feature activation | +1.80e-02 |
Pos logit contributions | |
ley | +0.118 |
rant | +0.106 |
work | +0.105 |
ky | +0.101 |
ci | +0.098 |
Neg logit contributions | |
unch | -0.074 |
bananas | -0.071 |
banana | -0.068 |
knights | -0.062 |
cheating | -0.062 |
Token | . |
Token ID | 13 |
Feature activation | -2.91e-03 |
Pos logit contributions | |
rant | +0.034 |
ley | +0.034 |
work | +0.032 |
ky | +0.029 |
extingu | +0.028 |
Neg logit contributions | |
bananas | -0.030 |
banana | -0.028 |
he | -0.026 |
cheating | -0.026 |
puck | -0.026 |
Token | Sp |
Token ID | 1338 |
Feature activation | +4.22e-02 |
Pos logit contributions | |
bananas | +0.005 |
cheating | +0.004 |
banana | +0.004 |
unch | +0.004 |
puck | +0.004 |
Neg logit contributions | |
ley | -0.006 |
rant | -0.005 |
work | -0.005 |
ky | -0.005 |
ci | -0.005 |
Token | ud |
Token ID | 463 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
ley | +0.086 |
rant | +0.077 |
work | +0.077 |
ky | +0.072 |
ci | +0.069 |
Neg logit contributions | |
unch | -0.066 |
bananas | -0.065 |
banana | -0.062 |
cheating | -0.060 |
puck | -0.058 |
Token | was |
Token ID | 373 |
Feature activation | +9.12e-03 |
Pos logit contributions | |
ley | +0.016 |
rant | +0.016 |
work | +0.015 |
ky | +0.014 |
extingu | +0.013 |
Neg logit contributions | |
bananas | -0.024 |
banana | -0.023 |
he | -0.022 |
puck | -0.022 |
cheating | -0.022 |
Token | there |
Token ID | 612 |
Feature activation | -2.81e-03 |
Pos logit contributions | |
ley | +0.005 |
rant | +0.005 |
work | +0.005 |
ky | +0.004 |
ci | +0.004 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
puck | -0.003 |
cheating | -0.003 |
unch | -0.002 |
Token | was |
Token ID | 373 |
Feature activation | +1.90e-03 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.005 |
cheating | +0.005 |
unch | +0.005 |
puck | +0.005 |
Neg logit contributions | |
ley | -0.005 |
rant | -0.005 |
work | -0.004 |
ky | -0.004 |
ci | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | -1.27e-03 |
Pos logit contributions | |
ley | +0.004 |
rant | +0.004 |
work | +0.004 |
ky | +0.004 |
ci | +0.004 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.002 |
unch | -0.002 |
cheating | -0.002 |
knights | -0.002 |
Token | girl |
Token ID | 2576 |
Feature activation | -1.94e-03 |
Pos logit contributions | |
bananas | +0.002 |
cheating | +0.002 |
unch | +0.002 |
banana | +0.002 |
puck | +0.002 |
Neg logit contributions | |
ley | -0.002 |
rant | -0.002 |
work | -0.002 |
ky | -0.002 |
ci | -0.002 |
Token | named |
Token ID | 3706 |
Feature activation | -4.56e-03 |
Pos logit contributions | |
bananas | +0.003 |
banana | +0.003 |
swinging | +0.003 |
knights | +0.002 |
catching | +0.002 |
Neg logit contributions | |
ley | -0.004 |
rant | -0.004 |
work | -0.004 |
ky | -0.003 |
ka | -0.003 |
Token | Fel |
Token ID | 13937 |
Feature activation | +5.37e-02 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.004 |
cheating | +0.004 |
catching | +0.004 |
unch | +0.004 |
Neg logit contributions | |
ley | -0.012 |
rant | -0.011 |
work | -0.011 |
ky | -0.011 |
ci | -0.010 |
Token | icity |
Token ID | 8467 |
Feature activation | +4.69e-03 |
Pos logit contributions | |
work | +0.082 |
ley | +0.080 |
streaming | +0.076 |
ci | +0.076 |
attered | +0.075 |
Neg logit contributions | |
knights | -0.091 |
cheating | -0.089 |
bananas | -0.089 |
beetles | -0.086 |
banana | -0.085 |
Token | who |
Token ID | 508 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
ley | +0.010 |
rant | +0.010 |
work | +0.009 |
ky | +0.009 |
extingu | +0.009 |
Neg logit contributions | |
bananas | -0.006 |
banana | -0.006 |
knights | -0.006 |
he | -0.005 |
swinging | -0.005 |
Token | lived |
Token ID | 5615 |
Feature activation | -1.60e-02 |
Pos logit contributions | |
bananas | +0.023 |
banana | +0.022 |
he | +0.021 |
cheating | +0.021 |
catching | +0.020 |
Neg logit contributions | |
ley | -0.024 |
rant | -0.022 |
work | -0.021 |
ky | -0.020 |
extingu | -0.018 |
Token | on |
Token ID | 319 |
Feature activation | +1.68e-03 |
Pos logit contributions | |
bananas | +0.038 |
banana | +0.036 |
cheating | +0.035 |
puck | +0.035 |
unch | +0.034 |
Neg logit contributions | |
ley | -0.021 |
rant | -0.018 |
work | -0.017 |
ky | -0.017 |
ci | -0.015 |
Token | a |
Token ID | 257 |
Feature activation | +3.25e-03 |
Pos logit contributions | |
ley | +0.004 |
rant | +0.004 |
work | +0.004 |
ky | +0.004 |
ci | +0.004 |
Neg logit contributions | |
bananas | -0.002 |
banana | -0.002 |
catching | -0.002 |
knights | -0.002 |
cheating | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +1.04e-02 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.000 |
cheating | +0.000 |
puck | +0.000 |
unch | +0.000 |
Neg logit contributions | |
ley | -0.000 |
rant | -0.000 |
work | -0.000 |
ky | -0.000 |
ci | -0.000 |
Token | a |
Token ID | 257 |
Feature activation | +2.12e-03 |
Pos logit contributions | |
ley | +0.024 |
rant | +0.022 |
work | +0.022 |
ky | +0.021 |
ci | +0.020 |
Neg logit contributions | |
bananas | -0.015 |
banana | -0.013 |
cheating | -0.012 |
unch | -0.012 |
knights | -0.012 |
Token | little |
Token ID | 1310 |
Feature activation | +2.55e-03 |
Pos logit contributions | |
ley | +0.004 |
rant | +0.004 |
work | +0.003 |
ky | +0.003 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.004 |
cheating | -0.003 |
banana | -0.003 |
unch | -0.003 |
puck | -0.003 |
Token | spider |
Token ID | 19230 |
Feature activation | +2.61e-02 |
Pos logit contributions | |
ley | +0.006 |
rant | +0.005 |
work | +0.005 |
ky | +0.005 |
ci | +0.005 |
Neg logit contributions | |
bananas | -0.004 |
banana | -0.004 |
knights | -0.003 |
puck | -0.003 |
unch | -0.003 |
Token | named |
Token ID | 3706 |
Feature activation | -1.36e-03 |
Pos logit contributions | |
ley | +0.054 |
rant | +0.052 |
work | +0.047 |
ky | +0.046 |
ka | +0.044 |
Neg logit contributions | |
bananas | -0.041 |
banana | -0.040 |
knights | -0.037 |
he | -0.036 |
catching | -0.036 |
Token | Sp |
Token ID | 1338 |
Feature activation | +5.25e-02 |
Pos logit contributions | |
bananas | +0.002 |
banana | +0.002 |
cheating | +0.002 |
unch | +0.001 |
puck | +0.001 |
Neg logit contributions | |
ley | -0.003 |
rant | -0.003 |
work | -0.003 |
ky | -0.003 |
ci | -0.003 |
Token | ide |
Token ID | 485 |
Feature activation | +8.75e-03 |
Pos logit contributions | |
ley | +0.114 |
rant | +0.102 |
work | +0.101 |
ky | +0.098 |
ci | +0.094 |
Neg logit contributions | |
bananas | -0.075 |
unch | -0.073 |
banana | -0.071 |
cheating | -0.067 |
knights | -0.065 |
Token | y |
Token ID | 88 |
Feature activation | +6.22e-03 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.013 |
ky | +0.012 |
work | +0.011 |
ci | +0.011 |
Neg logit contributions | |
bananas | -0.018 |
banana | -0.017 |
knights | -0.017 |
he | -0.016 |
beetles | -0.016 |
Token | . |
Token ID | 13 |
Feature activation | -3.21e-03 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.013 |
work | +0.011 |
ky | +0.011 |
ci | +0.011 |
Neg logit contributions | |
bananas | -0.010 |
banana | -0.009 |
knights | -0.009 |
he | -0.008 |
unch | -0.008 |
Token | Sp |
Token ID | 1338 |
Feature activation | +4.54e-02 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.005 |
cheating | +0.005 |
unch | +0.005 |
puck | +0.005 |
Neg logit contributions | |
ley | -0.006 |
rant | -0.006 |
work | -0.006 |
ky | -0.006 |
ci | -0.005 |
Token | ide |
Token ID | 485 |
Feature activation | +4.62e-03 |
Pos logit contributions | |
ley | +0.100 |
rant | +0.089 |
work | +0.089 |
ky | +0.086 |
ci | +0.083 |
Neg logit contributions | |
bananas | -0.063 |
unch | -0.062 |
banana | -0.059 |
cheating | -0.056 |
knights | -0.055 |
Token | was |
Token ID | 373 |
Feature activation | +8.95e-03 |
Pos logit contributions | |
bananas | +0.002 |
banana | +0.002 |
cheating | +0.002 |
puck | +0.002 |
unch | +0.002 |
Neg logit contributions | |
ley | -0.002 |
rant | -0.001 |
work | -0.001 |
ky | -0.001 |
ci | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | -1.29e-04 |
Pos logit contributions | |
ley | +0.020 |
rant | +0.018 |
work | +0.018 |
ky | +0.017 |
ci | +0.017 |
Neg logit contributions | |
bananas | -0.013 |
banana | -0.012 |
cheating | -0.011 |
unch | -0.011 |
puck | -0.011 |
Token | little |
Token ID | 1310 |
Feature activation | +3.11e-04 |
Pos logit contributions | |
bananas | +0.000 |
cheating | +0.000 |
unch | +0.000 |
banana | +0.000 |
catching | +0.000 |
Neg logit contributions | |
ley | -0.000 |
rant | -0.000 |
work | -0.000 |
ky | -0.000 |
ci | -0.000 |
Token | spider |
Token ID | 19230 |
Feature activation | +2.21e-02 |
Pos logit contributions | |
ley | +0.001 |
rant | +0.001 |
work | +0.001 |
ky | +0.001 |
ci | +0.001 |
Neg logit contributions | |
bananas | -0.000 |
banana | -0.000 |
knights | -0.000 |
puck | -0.000 |
cheating | -0.000 |
Token | named |
Token ID | 3706 |
Feature activation | -1.27e-03 |
Pos logit contributions | |
ley | +0.044 |
rant | +0.042 |
work | +0.038 |
ky | +0.037 |
ka | +0.035 |
Neg logit contributions | |
bananas | -0.037 |
banana | -0.036 |
knights | -0.033 |
catching | -0.033 |
cheating | -0.032 |
Token | Sp |
Token ID | 1338 |
Feature activation | +5.23e-02 |
Pos logit contributions | |
bananas | +0.002 |
unch | +0.002 |
banana | +0.002 |
cheating | +0.002 |
puck | +0.002 |
Neg logit contributions | |
ley | -0.003 |
rant | -0.003 |
ky | -0.002 |
work | -0.002 |
ci | -0.002 |
Token | ide |
Token ID | 485 |
Feature activation | +6.46e-03 |
Pos logit contributions | |
ley | +0.113 |
rant | +0.102 |
work | +0.101 |
ky | +0.097 |
ci | +0.094 |
Neg logit contributions | |
bananas | -0.075 |
unch | -0.073 |
banana | -0.071 |
cheating | -0.067 |
knights | -0.065 |
Token | y |
Token ID | 88 |
Feature activation | +3.59e-03 |
Pos logit contributions | |
ley | +0.009 |
rant | +0.008 |
ky | +0.007 |
work | +0.007 |
ci | +0.007 |
Neg logit contributions | |
bananas | -0.015 |
banana | -0.014 |
knights | -0.013 |
he | -0.013 |
beetles | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | -3.89e-03 |
Pos logit contributions | |
ley | +0.007 |
rant | +0.007 |
work | +0.006 |
ky | +0.006 |
ci | +0.006 |
Neg logit contributions | |
bananas | -0.006 |
banana | -0.006 |
knights | -0.005 |
unch | -0.005 |
he | -0.005 |
Token | He |
Token ID | 679 |
Feature activation | -4.39e-03 |
Pos logit contributions | |
bananas | +0.006 |
unch | +0.006 |
cheating | +0.006 |
banana | +0.006 |
puck | +0.006 |
Neg logit contributions | |
ley | -0.008 |
rant | -0.007 |
work | -0.007 |
ky | -0.007 |
ci | -0.006 |
Token | lived |
Token ID | 5615 |
Feature activation | -1.84e-02 |
Pos logit contributions | |
bananas | +0.008 |
banana | +0.008 |
knights | +0.007 |
he | +0.007 |
catching | +0.007 |
Neg logit contributions | |
ley | -0.008 |
rant | -0.008 |
work | -0.008 |
ky | -0.007 |
extingu | -0.007 |
Token | there |
Token ID | 612 |
Feature activation | -7.51e-04 |
Pos logit contributions | |
ley | +0.023 |
rant | +0.021 |
work | +0.020 |
ky | +0.020 |
ci | +0.019 |
Neg logit contributions | |
bananas | -0.012 |
banana | -0.011 |
cheating | -0.011 |
unch | -0.010 |
puck | -0.010 |
Token | was |
Token ID | 373 |
Feature activation | +1.02e-02 |
Pos logit contributions | |
bananas | +0.002 |
cheating | +0.002 |
banana | +0.002 |
puck | +0.002 |
knights | +0.002 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
work | -0.001 |
ky | -0.001 |
ci | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | +8.37e-04 |
Pos logit contributions | |
ley | +0.024 |
rant | +0.022 |
work | +0.021 |
ky | +0.020 |
ci | +0.020 |
Neg logit contributions | |
bananas | -0.015 |
banana | -0.013 |
cheating | -0.012 |
unch | -0.012 |
puck | -0.012 |
Token | spider |
Token ID | 19230 |
Feature activation | +2.49e-02 |
Pos logit contributions | |
ley | +0.001 |
rant | +0.001 |
work | +0.001 |
ky | +0.001 |
ci | +0.001 |
Neg logit contributions | |
bananas | -0.002 |
cheating | -0.001 |
unch | -0.001 |
banana | -0.001 |
catching | -0.001 |
Token | named |
Token ID | 3706 |
Feature activation | +7.77e-04 |
Pos logit contributions | |
ley | +0.050 |
rant | +0.048 |
work | +0.043 |
ky | +0.042 |
ka | +0.040 |
Neg logit contributions | |
bananas | -0.041 |
banana | -0.040 |
knights | -0.037 |
catching | -0.036 |
he | -0.036 |
Token | Sp |
Token ID | 1338 |
Feature activation | +5.22e-02 |
Pos logit contributions | |
ley | +0.002 |
rant | +0.002 |
work | +0.002 |
ky | +0.002 |
ci | +0.001 |
Neg logit contributions | |
bananas | -0.001 |
banana | -0.001 |
cheating | -0.001 |
unch | -0.001 |
puck | -0.001 |
Token | ide |
Token ID | 485 |
Feature activation | +9.02e-03 |
Pos logit contributions | |
ley | +0.114 |
rant | +0.102 |
work | +0.101 |
ky | +0.097 |
ci | +0.094 |
Neg logit contributions | |
bananas | -0.075 |
unch | -0.073 |
banana | -0.071 |
cheating | -0.066 |
knights | -0.064 |
Token | y |
Token ID | 88 |
Feature activation | +6.13e-03 |
Pos logit contributions | |
ley | +0.011 |
rant | +0.010 |
ky | +0.009 |
work | +0.009 |
ci | +0.008 |
Neg logit contributions | |
bananas | -0.022 |
banana | -0.021 |
knights | -0.020 |
beetles | -0.019 |
he | -0.019 |
Token | . |
Token ID | 13 |
Feature activation | -3.55e-03 |
Pos logit contributions | |
ley | +0.012 |
rant | +0.012 |
work | +0.011 |
ky | +0.010 |
ci | +0.010 |
Neg logit contributions | |
bananas | -0.010 |
banana | -0.010 |
knights | -0.009 |
unch | -0.009 |
puck | -0.009 |
Token | Sp |
Token ID | 1338 |
Feature activation | +4.53e-02 |
Pos logit contributions | |
bananas | +0.006 |
unch | +0.005 |
banana | +0.005 |
cheating | +0.005 |
puck | +0.005 |
Neg logit contributions | |
ley | -0.007 |
rant | -0.006 |
work | -0.006 |
ky | -0.006 |
ci | -0.006 |
Token | ide |
Token ID | 485 |
Feature activation | +4.53e-03 |
Pos logit contributions | |
ley | +0.100 |
rant | +0.089 |
work | +0.088 |
ky | +0.085 |
ci | +0.083 |
Neg logit contributions | |
bananas | -0.063 |
unch | -0.062 |
banana | -0.059 |
cheating | -0.056 |
knights | -0.055 |
Token | their |
Token ID | 511 |
Feature activation | +1.64e-02 |
Pos logit contributions | |
ley | +0.022 |
rant | +0.021 |
work | +0.020 |
ky | +0.019 |
ci | +0.018 |
Neg logit contributions | |
bananas | -0.018 |
banana | -0.016 |
cheating | -0.016 |
catching | -0.015 |
knights | -0.015 |
Token | friends |
Token ID | 2460 |
Feature activation | +3.56e-03 |
Pos logit contributions | |
ley | +0.037 |
rant | +0.036 |
ky | +0.033 |
work | +0.032 |
sy | +0.030 |
Neg logit contributions | |
bananas | -0.024 |
banana | -0.022 |
cheating | -0.021 |
knights | -0.021 |
puck | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | +9.60e-03 |
Pos logit contributions | |
ley | +0.007 |
rant | +0.007 |
work | +0.006 |
ky | +0.006 |
ci | +0.005 |
Neg logit contributions | |
bananas | -0.006 |
banana | -0.006 |
cheating | -0.005 |
catching | -0.005 |
knights | -0.005 |
Token |
|
Token ID | 198 |
Feature activation | +6.38e-03 |
Pos logit contributions | |
rant | +0.022 |
ley | +0.021 |
work | +0.021 |
ky | +0.020 |
ci | +0.018 |
Neg logit contributions | |
Ready | -0.011 |
bananas | -0.011 |
he | -0.011 |
many | -0.010 |
knights | -0.010 |
Token |
|
Token ID | 198 |
Feature activation | +3.54e-03 |
Pos logit contributions | |
ley | +0.016 |
rant | +0.015 |
work | +0.015 |
ky | +0.015 |
ci | +0.014 |
Neg logit contributions | |
bananas | -0.007 |
banana | -0.006 |
unch | -0.006 |
cheating | -0.006 |
catching | -0.006 |
Token | Fel |
Token ID | 42493 |
Feature activation | +5.13e-02 |
Pos logit contributions | |
ley | +0.008 |
rant | +0.008 |
work | +0.008 |
ky | +0.007 |
ci | +0.007 |
Neg logit contributions | |
bananas | -0.005 |
cheating | -0.004 |
banana | -0.004 |
unch | -0.004 |
puck | -0.004 |
Token | icity |
Token ID | 8467 |
Feature activation | +3.53e-03 |
Pos logit contributions | |
work | +0.078 |
ley | +0.077 |
rant | +0.077 |
ci | +0.073 |
extingu | +0.071 |
Neg logit contributions | |
cheating | -0.089 |
bananas | -0.081 |
azes | -0.081 |
knights | -0.081 |
cheat | -0.080 |
Token | decided |
Token ID | 3066 |
Feature activation | -1.95e-03 |
Pos logit contributions | |
rant | +0.007 |
ley | +0.007 |
work | +0.007 |
extingu | +0.006 |
ky | +0.006 |
Neg logit contributions | |
bananas | -0.005 |
banana | -0.005 |
he | -0.005 |
knights | -0.005 |
chewing | -0.004 |
Token | to |
Token ID | 284 |
Feature activation | +1.06e-02 |
Pos logit contributions | |
bananas | +0.003 |
banana | +0.003 |
cheating | +0.003 |
unch | +0.003 |
puck | +0.003 |
Neg logit contributions | |
ley | -0.004 |
rant | -0.003 |
work | -0.003 |
ky | -0.003 |
ci | -0.003 |
Token | keep |
Token ID | 1394 |
Feature activation | -3.75e-03 |
Pos logit contributions | |
ley | +0.025 |
rant | +0.023 |
ky | +0.022 |
work | +0.022 |
ci | +0.020 |
Neg logit contributions | |
bananas | -0.015 |
banana | -0.014 |
cheating | -0.012 |
cheat | -0.012 |
puck | -0.012 |
Token | the |
Token ID | 262 |
Feature activation | +6.94e-03 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.005 |
unch | +0.005 |
cheating | +0.005 |
puck | +0.005 |
Neg logit contributions | |
ley | -0.009 |
rant | -0.008 |
work | -0.008 |
ky | -0.007 |
ci | -0.007 |
Token | with |
Token ID | 351 |
Feature activation | +6.51e-03 |
Pos logit contributions | |
ley | +0.039 |
rant | +0.035 |
ky | +0.033 |
work | +0.032 |
ian | +0.032 |
Neg logit contributions | |
bananas | -0.017 |
catching | -0.015 |
banana | -0.015 |
cheating | -0.014 |
knights | -0.013 |
Token | new |
Token ID | 649 |
Feature activation | -7.21e-03 |
Pos logit contributions | |
ley | +0.015 |
rant | +0.014 |
work | +0.013 |
ky | +0.013 |
ci | +0.013 |
Neg logit contributions | |
bananas | -0.010 |
banana | -0.008 |
catching | -0.008 |
knights | -0.008 |
mango | -0.008 |
Token | life |
Token ID | 1204 |
Feature activation | -9.88e-03 |
Pos logit contributions | |
bananas | +0.008 |
banana | +0.007 |
knights | +0.007 |
cheating | +0.006 |
catching | +0.006 |
Neg logit contributions | |
ley | -0.019 |
rant | -0.018 |
work | -0.017 |
ky | -0.017 |
ci | -0.016 |
Token | ! |
Token ID | 0 |
Feature activation | +2.20e-02 |
Pos logit contributions | |
bananas | +0.016 |
banana | +0.014 |
he | +0.014 |
catching | +0.013 |
knights | +0.013 |
Neg logit contributions | |
ley | -0.020 |
rant | -0.019 |
ky | -0.017 |
work | -0.017 |
ka | -0.016 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +3.09e-03 |
Pos logit contributions | |
ley | +0.057 |
rant | +0.054 |
work | +0.051 |
ky | +0.051 |
ci | +0.048 |
Neg logit contributions | |
bananas | -0.023 |
unch | -0.020 |
banana | -0.020 |
Ready | -0.020 |
cheating | -0.019 |
Token | Far |
Token ID | 21428 |
Feature activation | +5.09e-02 |
Pos logit contributions | |
ley | +0.008 |
rant | +0.007 |
work | +0.007 |
ky | +0.007 |
ci | +0.006 |
Neg logit contributions | |
bananas | -0.004 |
banana | -0.003 |
unch | -0.003 |
cheating | -0.003 |
Ready | -0.003 |
Token | mer |
Token ID | 647 |
Feature activation | +1.28e-02 |
Pos logit contributions | |
extingu | +0.045 |
work | +0.037 |
bul | +0.036 |
rant | +0.035 |
streaming | +0.034 |
Neg logit contributions | |
scoring | -0.105 |
he | -0.103 |
bananas | -0.102 |
catching | -0.102 |
shadow | -0.101 |
Token | Fred |
Token ID | 8559 |
Feature activation | +1.61e-02 |
Pos logit contributions | |
rant | +0.028 |
ley | +0.027 |
work | +0.024 |
ky | +0.024 |
ci | +0.024 |
Neg logit contributions | |
he | -0.017 |
bananas | -0.017 |
banana | -0.016 |
knights | -0.014 |
puck | -0.014 |
Token | was |
Token ID | 373 |
Feature activation | -4.94e-04 |
Pos logit contributions | |
work | +0.035 |
rant | +0.035 |
ley | +0.034 |
works | +0.032 |
extingu | +0.031 |
Neg logit contributions | |
bananas | -0.022 |
he | -0.022 |
banana | -0.021 |
chewing | -0.019 |
knights | -0.018 |
Token | very |
Token ID | 845 |
Feature activation | -4.22e-03 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.001 |
catching | +0.001 |
cheating | +0.001 |
knights | +0.001 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
work | -0.001 |
ky | -0.001 |
ci | -0.001 |
Token | cheerful |
Token ID | 37999 |
Feature activation | +7.43e-03 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.005 |
unch | +0.005 |
cheating | +0.005 |
catching | +0.005 |
Neg logit contributions | |
ley | -0.010 |
rant | -0.009 |
work | -0.009 |
ky | -0.009 |
ci | -0.008 |
Token | was |
Token ID | 373 |
Feature activation | +5.89e-03 |
Pos logit contributions | |
ley | +0.002 |
rant | +0.001 |
work | +0.001 |
ky | +0.001 |
ci | +0.001 |
Neg logit contributions | |
bananas | -0.002 |
banana | -0.002 |
cheating | -0.002 |
puck | -0.002 |
unch | -0.002 |
Token | a |
Token ID | 257 |
Feature activation | +9.58e-04 |
Pos logit contributions | |
ley | +0.014 |
rant | +0.013 |
work | +0.013 |
ky | +0.012 |
ci | +0.012 |
Neg logit contributions | |
bananas | -0.008 |
banana | -0.007 |
knights | -0.007 |
unch | -0.006 |
cheating | -0.006 |
Token | little |
Token ID | 1310 |
Feature activation | -4.41e-04 |
Pos logit contributions | |
ley | +0.002 |
rant | +0.002 |
work | +0.002 |
ky | +0.002 |
ci | +0.002 |
Neg logit contributions | |
bananas | -0.002 |
banana | -0.001 |
cheating | -0.001 |
puck | -0.001 |
unch | -0.001 |
Token | spider |
Token ID | 19230 |
Feature activation | +2.57e-02 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.001 |
knights | +0.001 |
puck | +0.001 |
cheating | +0.001 |
Neg logit contributions | |
rant | -0.001 |
ley | -0.001 |
work | -0.001 |
ky | -0.001 |
ci | -0.001 |
Token | named |
Token ID | 3706 |
Feature activation | -6.19e-03 |
Pos logit contributions | |
ley | +0.053 |
rant | +0.052 |
work | +0.046 |
ky | +0.045 |
ka | +0.044 |
Neg logit contributions | |
bananas | -0.041 |
banana | -0.040 |
knights | -0.036 |
he | -0.036 |
catching | -0.036 |
Token | Sp |
Token ID | 1338 |
Feature activation | +5.09e-02 |
Pos logit contributions | |
bananas | +0.008 |
banana | +0.008 |
cheating | +0.007 |
unch | +0.007 |
puck | +0.007 |
Neg logit contributions | |
ley | -0.015 |
rant | -0.013 |
work | -0.013 |
ky | -0.013 |
ci | -0.012 |
Token | ide |
Token ID | 485 |
Feature activation | +9.13e-03 |
Pos logit contributions | |
ley | +0.108 |
rant | +0.097 |
work | +0.095 |
ky | +0.092 |
ci | +0.089 |
Neg logit contributions | |
bananas | -0.076 |
unch | -0.074 |
banana | -0.072 |
cheating | -0.067 |
knights | -0.066 |
Token | y |
Token ID | 88 |
Feature activation | +3.19e-03 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.013 |
ky | +0.012 |
work | +0.011 |
icol | +0.011 |
Neg logit contributions | |
bananas | -0.019 |
banana | -0.018 |
knights | -0.017 |
he | -0.017 |
beetles | -0.017 |
Token | . |
Token ID | 13 |
Feature activation | -6.06e-03 |
Pos logit contributions | |
rant | +0.007 |
ley | +0.007 |
work | +0.006 |
ky | +0.006 |
ci | +0.006 |
Neg logit contributions | |
bananas | -0.005 |
banana | -0.005 |
knights | -0.004 |
he | -0.004 |
beetles | -0.004 |
Token | Sp |
Token ID | 1338 |
Feature activation | +4.34e-02 |
Pos logit contributions | |
bananas | +0.010 |
cheating | +0.009 |
banana | +0.009 |
unch | +0.009 |
puck | +0.009 |
Neg logit contributions | |
ley | -0.012 |
rant | -0.011 |
work | -0.011 |
ky | -0.010 |
ci | -0.010 |
Token | ide |
Token ID | 485 |
Feature activation | +5.05e-03 |
Pos logit contributions | |
ley | +0.093 |
rant | +0.083 |
work | +0.082 |
ky | +0.080 |
ci | +0.077 |
Neg logit contributions | |
bananas | -0.064 |
unch | -0.062 |
banana | -0.060 |
cheating | -0.056 |
knights | -0.055 |
Token | was |
Token ID | 373 |
Feature activation | +2.65e-03 |
Pos logit contributions | |
bananas | +0.004 |
banana | +0.004 |
cheating | +0.004 |
puck | +0.004 |
unch | +0.004 |
Neg logit contributions | |
ley | -0.003 |
rant | -0.003 |
work | -0.003 |
ky | -0.003 |
ci | -0.002 |
Token | a |
Token ID | 257 |
Feature activation | -1.78e-04 |
Pos logit contributions | |
ley | +0.006 |
rant | +0.005 |
work | +0.005 |
ky | +0.005 |
ci | +0.005 |
Neg logit contributions | |
bananas | -0.004 |
banana | -0.003 |
cheating | -0.003 |
unch | -0.003 |
puck | -0.003 |
Token | little |
Token ID | 1310 |
Feature activation | +6.26e-04 |
Pos logit contributions | |
bananas | +0.000 |
banana | +0.000 |
cheating | +0.000 |
puck | +0.000 |
unch | +0.000 |
Neg logit contributions | |
ley | -0.000 |
rant | -0.000 |
work | -0.000 |
ky | -0.000 |
ci | -0.000 |
Token | spider |
Token ID | 19230 |
Feature activation | +2.55e-02 |
Pos logit contributions | |
rant | +0.001 |
ley | +0.001 |
work | +0.001 |
ky | +0.001 |
sy | +0.001 |
Neg logit contributions | |
bananas | -0.001 |
banana | -0.001 |
knights | -0.001 |
cheating | -0.001 |
puck | -0.001 |
Token | named |
Token ID | 3706 |
Feature activation | -7.38e-03 |
Pos logit contributions | |
rant | +0.054 |
ley | +0.053 |
work | +0.048 |
ka | +0.046 |
ky | +0.046 |
Neg logit contributions | |
bananas | -0.039 |
banana | -0.038 |
he | -0.035 |
knights | -0.034 |
catching | -0.033 |
Token | Sp |
Token ID | 1338 |
Feature activation | +5.06e-02 |
Pos logit contributions | |
bananas | +0.012 |
banana | +0.011 |
cheating | +0.010 |
unch | +0.010 |
puck | +0.010 |
Neg logit contributions | |
ley | -0.016 |
rant | -0.014 |
work | -0.014 |
ky | -0.013 |
ci | -0.013 |
Token | ide |
Token ID | 485 |
Feature activation | +9.85e-03 |
Pos logit contributions | |
ley | +0.112 |
rant | +0.100 |
work | +0.098 |
ky | +0.096 |
ci | +0.092 |
Neg logit contributions | |
bananas | -0.070 |
unch | -0.068 |
banana | -0.067 |
cheating | -0.062 |
knights | -0.060 |
Token | y |
Token ID | 88 |
Feature activation | +3.90e-03 |
Pos logit contributions | |
ley | +0.012 |
rant | +0.011 |
ky | +0.011 |
icol | +0.010 |
work | +0.010 |
Neg logit contributions | |
bananas | -0.023 |
banana | -0.022 |
knights | -0.021 |
he | -0.021 |
beetles | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | -5.82e-03 |
Pos logit contributions | |
rant | +0.009 |
ley | +0.008 |
work | +0.008 |
ky | +0.007 |
ci | +0.007 |
Neg logit contributions | |
bananas | -0.005 |
banana | -0.005 |
knights | -0.005 |
he | -0.005 |
beetles | -0.004 |
Token | Sp |
Token ID | 1338 |
Feature activation | +4.15e-02 |
Pos logit contributions | |
bananas | +0.009 |
banana | +0.008 |
cheating | +0.008 |
unch | +0.008 |
puck | +0.008 |
Neg logit contributions | |
ley | -0.012 |
rant | -0.011 |
work | -0.011 |
ky | -0.011 |
ci | -0.010 |
Token | ide |
Token ID | 485 |
Feature activation | +4.12e-03 |
Pos logit contributions | |
ley | +0.092 |
rant | +0.082 |
work | +0.081 |
ky | +0.080 |
ci | +0.077 |
Neg logit contributions | |
bananas | -0.057 |
unch | -0.056 |
banana | -0.053 |
cheating | -0.049 |
knights | -0.048 |
Token | was |
Token ID | 373 |
Feature activation | +4.84e-03 |
Pos logit contributions | |
ley | +0.003 |
rant | +0.003 |
work | +0.002 |
ky | +0.002 |
ci | +0.002 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
cheating | -0.003 |
unch | -0.003 |
puck | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | +9.37e-04 |
Pos logit contributions | |
ley | +0.012 |
rant | +0.011 |
work | +0.011 |
ky | +0.011 |
ci | +0.010 |
Neg logit contributions | |
bananas | -0.006 |
banana | -0.006 |
knights | -0.005 |
unch | -0.005 |
catching | -0.005 |
Token | little |
Token ID | 1310 |
Feature activation | -4.80e-04 |
Pos logit contributions | |
ley | +0.002 |
rant | +0.002 |
work | +0.002 |
ky | +0.002 |
ci | +0.001 |
Neg logit contributions | |
bananas | -0.002 |
banana | -0.001 |
cheating | -0.001 |
unch | -0.001 |
puck | -0.001 |
Token | girl |
Token ID | 2576 |
Feature activation | +8.12e-04 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.001 |
knights | +0.001 |
puck | +0.001 |
mango | +0.001 |
Neg logit contributions | |
rant | -0.001 |
ley | -0.001 |
work | -0.001 |
ky | -0.001 |
ci | -0.001 |
Token | named |
Token ID | 3706 |
Feature activation | -5.13e-03 |
Pos logit contributions | |
ley | +0.002 |
rant | +0.002 |
work | +0.002 |
works | +0.002 |
ka | +0.002 |
Neg logit contributions | |
banana | -0.001 |
bananas | -0.001 |
swinging | -0.001 |
chewing | -0.001 |
knights | -0.001 |
Token | Fel |
Token ID | 13937 |
Feature activation | +5.05e-02 |
Pos logit contributions | |
bananas | +0.006 |
banana | +0.005 |
cheating | +0.005 |
unch | +0.005 |
Ready | +0.005 |
Neg logit contributions | |
ley | -0.013 |
rant | -0.012 |
work | -0.012 |
ky | -0.011 |
ci | -0.011 |
Token | icity |
Token ID | 8467 |
Feature activation | +8.18e-03 |
Pos logit contributions | |
work | +0.073 |
ley | +0.072 |
ci | +0.069 |
streaming | +0.068 |
rant | +0.066 |
Neg logit contributions | |
knights | -0.088 |
bananas | -0.084 |
cheating | -0.082 |
beetles | -0.081 |
banana | -0.080 |
Token | . |
Token ID | 13 |
Feature activation | -8.46e-03 |
Pos logit contributions | |
rant | +0.018 |
ley | +0.018 |
work | +0.017 |
ky | +0.016 |
extingu | +0.016 |
Neg logit contributions | |
bananas | -0.010 |
banana | -0.010 |
knights | -0.009 |
he | -0.009 |
many | -0.008 |
Token | She |
Token ID | 1375 |
Feature activation | -4.62e-03 |
Pos logit contributions | |
bananas | +0.012 |
banana | +0.011 |
cheating | +0.010 |
unch | +0.010 |
Ready | +0.010 |
Neg logit contributions | |
ley | -0.019 |
rant | -0.018 |
work | -0.018 |
ky | -0.017 |
ci | -0.016 |
Token | was |
Token ID | 373 |
Feature activation | +3.75e-04 |
Pos logit contributions | |
he | +0.006 |
banana | +0.006 |
bananas | +0.006 |
chewing | +0.005 |
many | +0.005 |
Neg logit contributions | |
ley | -0.011 |
rant | -0.010 |
work | -0.010 |
extingu | -0.010 |
ky | -0.009 |
Token | only |
Token ID | 691 |
Feature activation | -1.79e-02 |
Pos logit contributions | |
ley | +0.001 |
rant | +0.001 |
work | +0.001 |
ky | +0.001 |
ci | +0.001 |
Neg logit contributions | |
bananas | -0.001 |
banana | -0.001 |
cheating | -0.000 |
knights | -0.000 |
catching | -0.000 |
Token | was |
Token ID | 373 |
Feature activation | +9.97e-03 |
Pos logit contributions | |
bananas | +0.000 |
banana | +0.000 |
cheating | +0.000 |
puck | +0.000 |
unch | +0.000 |
Neg logit contributions | |
ley | -0.000 |
rant | -0.000 |
work | -0.000 |
ky | -0.000 |
ci | -0.000 |
Token | a |
Token ID | 257 |
Feature activation | -9.62e-05 |
Pos logit contributions | |
ley | +0.023 |
rant | +0.021 |
work | +0.020 |
ky | +0.020 |
ci | +0.019 |
Neg logit contributions | |
bananas | -0.014 |
banana | -0.013 |
cheating | -0.012 |
unch | -0.012 |
puck | -0.012 |
Token | little |
Token ID | 1310 |
Feature activation | +4.20e-04 |
Pos logit contributions | |
bananas | +0.000 |
cheating | +0.000 |
unch | +0.000 |
banana | +0.000 |
puck | +0.000 |
Neg logit contributions | |
ley | -0.000 |
rant | -0.000 |
work | -0.000 |
ky | -0.000 |
ci | -0.000 |
Token | spider |
Token ID | 19230 |
Feature activation | +2.43e-02 |
Pos logit contributions | |
ley | +0.001 |
rant | +0.001 |
work | +0.001 |
ky | +0.001 |
ci | +0.001 |
Neg logit contributions | |
bananas | -0.001 |
banana | -0.001 |
knights | -0.001 |
puck | -0.001 |
cheating | -0.001 |
Token | named |
Token ID | 3706 |
Feature activation | -3.19e-03 |
Pos logit contributions | |
ley | +0.048 |
rant | +0.047 |
work | +0.042 |
ky | +0.041 |
ka | +0.039 |
Neg logit contributions | |
bananas | -0.040 |
banana | -0.039 |
knights | -0.036 |
catching | -0.035 |
cheating | -0.035 |
Token | Sp |
Token ID | 1338 |
Feature activation | +5.02e-02 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.004 |
cheating | +0.004 |
unch | +0.004 |
puck | +0.004 |
Neg logit contributions | |
ley | -0.007 |
rant | -0.007 |
work | -0.006 |
ky | -0.006 |
ci | -0.006 |
Token | ide |
Token ID | 485 |
Feature activation | +7.89e-03 |
Pos logit contributions | |
ley | +0.108 |
rant | +0.097 |
work | +0.095 |
ky | +0.091 |
ci | +0.088 |
Neg logit contributions | |
bananas | -0.074 |
unch | -0.071 |
banana | -0.070 |
cheating | -0.066 |
knights | -0.064 |
Token | y |
Token ID | 88 |
Feature activation | +4.85e-03 |
Pos logit contributions | |
ley | +0.011 |
rant | +0.011 |
ky | +0.010 |
work | +0.009 |
ci | +0.009 |
Neg logit contributions | |
bananas | -0.017 |
banana | -0.017 |
knights | -0.016 |
he | -0.015 |
beetles | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | -4.83e-03 |
Pos logit contributions | |
ley | +0.010 |
rant | +0.009 |
work | +0.009 |
ky | +0.008 |
ci | +0.008 |
Neg logit contributions | |
bananas | -0.008 |
banana | -0.008 |
knights | -0.007 |
unch | -0.007 |
cheating | -0.007 |
Token | Sp |
Token ID | 1338 |
Feature activation | +4.42e-02 |
Pos logit contributions | |
bananas | +0.008 |
unch | +0.007 |
cheating | +0.007 |
banana | +0.007 |
puck | +0.007 |
Neg logit contributions | |
ley | -0.009 |
rant | -0.009 |
work | -0.008 |
ky | -0.008 |
ci | -0.008 |
Token | ide |
Token ID | 485 |
Feature activation | +3.89e-03 |
Pos logit contributions | |
ley | +0.097 |
rant | +0.087 |
work | +0.086 |
ky | +0.083 |
ci | +0.080 |
Neg logit contributions | |
bananas | -0.062 |
unch | -0.060 |
banana | -0.058 |
cheating | -0.056 |
knights | -0.054 |
Token | knew |
Token ID | 2993 |
Feature activation | +6.12e-03 |
Pos logit contributions | |
ley | +0.015 |
rant | +0.014 |
ky | +0.014 |
work | +0.014 |
ka | +0.013 |
Neg logit contributions | |
he | -0.012 |
bananas | -0.012 |
banana | -0.011 |
cheating | -0.011 |
knights | -0.011 |
Token | that |
Token ID | 326 |
Feature activation | -6.98e-03 |
Pos logit contributions | |
ley | +0.015 |
rant | +0.014 |
ky | +0.013 |
work | +0.013 |
ci | +0.012 |
Neg logit contributions | |
bananas | -0.008 |
cheating | -0.007 |
banana | -0.007 |
catching | -0.006 |
knights | -0.006 |
Token | accidents |
Token ID | 17390 |
Feature activation | +1.45e-02 |
Pos logit contributions | |
bananas | +0.009 |
banana | +0.008 |
cheating | +0.008 |
unch | +0.008 |
puck | +0.008 |
Neg logit contributions | |
ley | -0.017 |
rant | -0.015 |
work | -0.015 |
ky | -0.015 |
ci | -0.014 |
Token | happen |
Token ID | 1645 |
Feature activation | -1.43e-02 |
Pos logit contributions | |
ley | +0.036 |
ky | +0.033 |
rant | +0.032 |
ci | +0.031 |
ka | +0.030 |
Neg logit contributions | |
he | -0.018 |
catching | -0.016 |
bananas | -0.016 |
cheating | -0.016 |
anything | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | +7.49e-03 |
Pos logit contributions | |
bananas | +0.020 |
he | +0.019 |
cheating | +0.019 |
catching | +0.018 |
knights | +0.018 |
Neg logit contributions | |
ley | -0.031 |
rant | -0.029 |
work | -0.027 |
ky | -0.027 |
ci | -0.026 |
Token | Fel |
Token ID | 13937 |
Feature activation | +5.01e-02 |
Pos logit contributions | |
ley | +0.015 |
rant | +0.015 |
work | +0.014 |
ky | +0.014 |
ci | +0.013 |
Neg logit contributions | |
bananas | -0.011 |
Ready | -0.011 |
banana | -0.010 |
unch | -0.010 |
knights | -0.010 |
Token | icia |
Token ID | 33577 |
Feature activation | +3.26e-03 |
Pos logit contributions | |
work | +0.068 |
rant | +0.066 |
ley | +0.065 |
bul | +0.064 |
extingu | +0.062 |
Neg logit contributions | |
cheating | -0.093 |
knights | -0.088 |
math | -0.085 |
principal | -0.083 |
beetles | -0.083 |
Token | learned |
Token ID | 4499 |
Feature activation | +7.63e-03 |
Pos logit contributions | |
rant | +0.007 |
ley | +0.007 |
ky | +0.006 |
work | +0.006 |
extingu | +0.006 |
Neg logit contributions | |
bananas | -0.004 |
knights | -0.004 |
he | -0.004 |
banana | -0.004 |
stealing | -0.003 |
Token | to |
Token ID | 284 |
Feature activation | -1.38e-03 |
Pos logit contributions | |
ley | +0.019 |
rant | +0.018 |
ky | +0.017 |
work | +0.017 |
ci | +0.016 |
Neg logit contributions | |
bananas | -0.009 |
he | -0.008 |
cheating | -0.008 |
stealing | -0.008 |
banana | -0.008 |
Token | be |
Token ID | 307 |
Feature activation | -5.20e-03 |
Pos logit contributions | |
bananas | +0.002 |
banana | +0.002 |
cheating | +0.002 |
unch | +0.002 |
puck | +0.002 |
Neg logit contributions | |
ley | -0.003 |
rant | -0.003 |
work | -0.003 |
ky | -0.003 |
ci | -0.003 |
Token | more |
Token ID | 517 |
Feature activation | -2.62e-02 |
Pos logit contributions | |
bananas | +0.006 |
unch | +0.006 |
banana | +0.006 |
puck | +0.006 |
cheating | +0.005 |
Neg logit contributions | |
ley | -0.012 |
rant | -0.011 |
work | -0.011 |
ky | -0.011 |
sy | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +1.86e-02 |
Pos logit contributions | |
ley | +0.039 |
rant | +0.037 |
ky | +0.033 |
work | +0.033 |
ci | +0.032 |
Neg logit contributions | |
bananas | -0.024 |
banana | -0.023 |
cheating | -0.022 |
he | -0.022 |
math | -0.021 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +7.04e-03 |
Pos logit contributions | |
ley | +0.046 |
rant | +0.042 |
work | +0.041 |
ky | +0.040 |
ci | +0.038 |
Neg logit contributions | |
bananas | -0.023 |
banana | -0.020 |
unch | -0.019 |
cheating | -0.019 |
puck | -0.019 |
Token | One |
Token ID | 3198 |
Feature activation | -3.37e-04 |
Pos logit contributions | |
ley | +0.018 |
rant | +0.017 |
work | +0.016 |
ky | +0.016 |
ci | +0.015 |
Neg logit contributions | |
bananas | -0.008 |
banana | -0.007 |
unch | -0.007 |
Ready | -0.007 |
knights | -0.007 |
Token | day |
Token ID | 1110 |
Feature activation | +1.38e-02 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.001 |
unch | +0.001 |
cheating | +0.001 |
puck | +0.001 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
work | -0.001 |
ky | -0.001 |
ci | -0.000 |
Token | , |
Token ID | 11 |
Feature activation | +1.04e-02 |
Pos logit contributions | |
ley | +0.032 |
rant | +0.031 |
work | +0.030 |
ky | +0.029 |
ci | +0.028 |
Neg logit contributions | |
bananas | -0.018 |
banana | -0.016 |
he | -0.016 |
catching | -0.015 |
cheating | -0.015 |
Token | M |
Token ID | 337 |
Feature activation | +5.00e-02 |
Pos logit contributions | |
ley | +0.026 |
rant | +0.024 |
work | +0.023 |
ky | +0.022 |
ci | +0.022 |
Neg logit contributions | |
bananas | -0.013 |
banana | -0.012 |
unch | -0.011 |
he | -0.011 |
catching | -0.011 |
Token | ummy |
Token ID | 13513 |
Feature activation | +1.54e-02 |
Pos logit contributions | |
rant | +0.109 |
ky | +0.109 |
ley | +0.109 |
extingu | +0.098 |
ples | +0.098 |
Neg logit contributions | |
unch | -0.070 |
banana | -0.050 |
azes | -0.049 |
bananas | -0.049 |
ress | -0.043 |
Token | asked |
Token ID | 1965 |
Feature activation | +1.18e-02 |
Pos logit contributions | |
rant | +0.035 |
ley | +0.032 |
ky | +0.031 |
work | +0.031 |
iful | +0.030 |
Neg logit contributions | |
he | -0.021 |
banana | -0.020 |
bananas | -0.019 |
chewing | -0.019 |
stealing | -0.017 |
Token | Jack |
Token ID | 3619 |
Feature activation | +6.98e-03 |
Pos logit contributions | |
ley | +0.027 |
rant | +0.026 |
work | +0.026 |
ky | +0.025 |
ci | +0.025 |
Neg logit contributions | |
bananas | -0.016 |
banana | -0.014 |
he | -0.013 |
unch | -0.013 |
knights | -0.013 |
Token | to |
Token ID | 284 |
Feature activation | +8.53e-03 |
Pos logit contributions | |
ley | +0.016 |
rant | +0.015 |
work | +0.014 |
ky | +0.014 |
ci | +0.013 |
Neg logit contributions | |
bananas | -0.010 |
banana | -0.009 |
cheating | -0.009 |
knights | -0.008 |
unch | -0.008 |
Token | take |
Token ID | 1011 |
Feature activation | +8.67e-03 |
Pos logit contributions | |
ley | +0.021 |
rant | +0.020 |
ky | +0.019 |
work | +0.018 |
ka | +0.017 |
Neg logit contributions | |
bananas | -0.011 |
banana | -0.011 |
he | -0.009 |
catching | -0.009 |
puck | -0.009 |
Token | " |
Token ID | 366 |
Feature activation | +3.86e-03 |
Pos logit contributions | |
bananas | +0.006 |
banana | +0.005 |
cheating | +0.005 |
puck | +0.004 |
unch | +0.004 |
Neg logit contributions | |
ley | -0.014 |
rant | -0.013 |
work | -0.013 |
ky | -0.013 |
ci | -0.012 |
Token | Hello |
Token ID | 15496 |
Feature activation | +1.86e-03 |
Pos logit contributions | |
ley | +0.007 |
rant | +0.006 |
work | +0.006 |
ky | +0.005 |
ci | +0.005 |
Neg logit contributions | |
bananas | -0.007 |
banana | -0.007 |
Ready | -0.006 |
unch | -0.006 |
mango | -0.006 |
Token | , |
Token ID | 11 |
Feature activation | +1.89e-03 |
Pos logit contributions | |
ley | +0.004 |
rant | +0.003 |
work | +0.003 |
ky | +0.003 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
cheating | -0.003 |
puck | -0.003 |
unch | -0.003 |
Token | Mister |
Token ID | 43438 |
Feature activation | +7.83e-03 |
Pos logit contributions | |
ley | +0.004 |
work | +0.003 |
rant | +0.003 |
ky | +0.003 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.003 |
puck | -0.003 |
banana | -0.003 |
cheating | -0.003 |
unch | -0.003 |
Token | P |
Token ID | 350 |
Feature activation | +8.92e-03 |
Pos logit contributions | |
ley | +0.016 |
rant | +0.015 |
work | +0.013 |
sy | +0.013 |
ky | +0.013 |
Neg logit contributions | |
bananas | -0.013 |
banana | -0.012 |
mango | -0.011 |
cheating | -0.011 |
catching | -0.011 |
Token | ige |
Token ID | 10045 |
Feature activation | -6.00e-02 |
Pos logit contributions | |
ley | +0.018 |
rant | +0.017 |
work | +0.016 |
ky | +0.015 |
ci | +0.015 |
Neg logit contributions | |
bananas | -0.014 |
unch | -0.014 |
banana | -0.013 |
cheating | -0.012 |
mango | -0.012 |
Token | on |
Token ID | 261 |
Feature activation | -1.12e-02 |
Pos logit contributions | |
puck | +0.093 |
Ready | +0.086 |
cheating | +0.083 |
bananas | +0.080 |
winning | +0.079 |
Neg logit contributions | |
ley | -0.119 |
work | -0.113 |
ina | -0.106 |
y | -0.106 |
ka | -0.105 |
Token | !" |
Token ID | 2474 |
Feature activation | +9.86e-04 |
Pos logit contributions | |
bananas | +0.018 |
banana | +0.017 |
cheating | +0.015 |
knights | +0.015 |
unch | +0.015 |
Neg logit contributions | |
ley | -0.024 |
rant | -0.022 |
ky | -0.020 |
work | -0.020 |
ci | -0.019 |
Token |
|
Token ID | 198 |
Feature activation | -4.35e-03 |
Pos logit contributions | |
ley | +0.002 |
rant | +0.002 |
work | +0.002 |
ky | +0.002 |
ci | +0.002 |
Neg logit contributions | |
bananas | -0.002 |
banana | -0.001 |
unch | -0.001 |
puck | -0.001 |
knights | -0.001 |
Token |
|
Token ID | 198 |
Feature activation | -7.81e-03 |
Pos logit contributions | |
bananas | +0.007 |
cheating | +0.006 |
banana | +0.006 |
puck | +0.006 |
knights | +0.006 |
Neg logit contributions | |
ley | -0.009 |
rant | -0.008 |
work | -0.008 |
ky | -0.008 |
sy | -0.007 |
Token | The |
Token ID | 464 |
Feature activation | +9.58e-03 |
Pos logit contributions | |
bananas | +0.012 |
cheating | +0.011 |
banana | +0.011 |
unch | +0.010 |
puck | +0.010 |
Neg logit contributions | |
ley | -0.017 |
ky | -0.015 |
rant | -0.015 |
work | -0.015 |
sy | -0.014 |
Token | I |
Token ID | 40 |
Feature activation | +1.36e-05 |
Pos logit contributions | |
bananas | +0.009 |
banana | +0.008 |
cheating | +0.008 |
knights | +0.008 |
puck | +0.008 |
Neg logit contributions | |
ley | -0.011 |
rant | -0.010 |
ky | -0.010 |
work | -0.009 |
ci | -0.009 |
Token | will |
Token ID | 481 |
Feature activation | -7.28e-03 |
Pos logit contributions | |
ley | +0.000 |
rant | +0.000 |
work | +0.000 |
ky | +0.000 |
ci | +0.000 |
Neg logit contributions | |
bananas | -0.000 |
banana | -0.000 |
cheating | -0.000 |
unch | -0.000 |
puck | -0.000 |
Token | buy |
Token ID | 2822 |
Feature activation | -2.55e-02 |
Pos logit contributions | |
bananas | +0.008 |
banana | +0.008 |
catching | +0.007 |
he | +0.007 |
cheat | +0.007 |
Neg logit contributions | |
ley | -0.019 |
rant | -0.018 |
work | -0.017 |
ky | -0.017 |
ci | -0.016 |
Token | more |
Token ID | 517 |
Feature activation | -2.20e-02 |
Pos logit contributions | |
unch | +0.035 |
puck | +0.033 |
cheating | +0.030 |
coward | +0.030 |
bananas | +0.030 |
Neg logit contributions | |
ley | -0.054 |
work | -0.049 |
rant | -0.049 |
ky | -0.048 |
ka | -0.045 |
Token | av |
Token ID | 1196 |
Feature activation | -4.65e-02 |
Pos logit contributions | |
unch | +0.028 |
bananas | +0.028 |
cheating | +0.027 |
puck | +0.027 |
catching | +0.026 |
Neg logit contributions | |
ley | -0.047 |
rant | -0.043 |
work | -0.043 |
ky | -0.042 |
ci | -0.039 |
Token | oc |
Token ID | 420 |
Feature activation | -5.81e-02 |
Pos logit contributions | |
cheating | +0.075 |
winning | +0.071 |
puck | +0.070 |
Ready | +0.069 |
sandbox | +0.069 |
Neg logit contributions | |
ley | -0.095 |
ky | -0.084 |
rant | -0.081 |
work | -0.076 |
ci | -0.074 |
Token | ados |
Token ID | 22484 |
Feature activation | -1.67e-02 |
Pos logit contributions | |
puck | +0.078 |
sandbox | +0.074 |
Ready | +0.074 |
cheating | +0.074 |
banana | +0.072 |
Neg logit contributions | |
ley | -0.129 |
work | -0.117 |
ky | -0.111 |
rant | -0.111 |
ci | -0.105 |
Token | tomorrow |
Token ID | 9439 |
Feature activation | -6.23e-03 |
Pos logit contributions | |
bananas | +0.030 |
banana | +0.028 |
cheating | +0.028 |
unch | +0.028 |
puck | +0.027 |
Neg logit contributions | |
ley | -0.031 |
rant | -0.027 |
work | -0.027 |
ky | -0.026 |
ci | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | -4.85e-03 |
Pos logit contributions | |
bananas | +0.011 |
banana | +0.010 |
cheating | +0.010 |
unch | +0.009 |
puck | +0.009 |
Neg logit contributions | |
ley | -0.012 |
rant | -0.011 |
work | -0.011 |
ky | -0.010 |
ci | -0.010 |
Token | You |
Token ID | 921 |
Feature activation | -1.08e-02 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.007 |
cheating | +0.007 |
unch | +0.007 |
catching | +0.007 |
Neg logit contributions | |
ley | -0.010 |
rant | -0.009 |
work | -0.009 |
ky | -0.009 |
ci | -0.008 |
Token | can |
Token ID | 460 |
Feature activation | -7.54e-03 |
Pos logit contributions | |
bananas | +0.019 |
banana | +0.017 |
cheating | +0.017 |
unch | +0.016 |
puck | +0.016 |
Neg logit contributions | |
ley | -0.021 |
rant | -0.019 |
work | -0.018 |
ky | -0.018 |
ci | -0.017 |
Token | could |
Token ID | 714 |
Feature activation | -2.13e-02 |
Pos logit contributions | |
ley | +0.017 |
rant | +0.015 |
ky | +0.014 |
work | +0.014 |
ci | +0.013 |
Neg logit contributions | |
bananas | -0.018 |
banana | -0.016 |
catching | -0.015 |
unch | -0.015 |
knights | -0.015 |
Token | see |
Token ID | 766 |
Feature activation | -1.95e-02 |
Pos logit contributions | |
bananas | +0.033 |
banana | +0.031 |
cheating | +0.029 |
unch | +0.029 |
puck | +0.028 |
Neg logit contributions | |
ley | -0.046 |
rant | -0.042 |
work | -0.040 |
ky | -0.040 |
ci | -0.037 |
Token | lots |
Token ID | 6041 |
Feature activation | -2.07e-03 |
Pos logit contributions | |
bananas | +0.029 |
banana | +0.027 |
cheating | +0.027 |
unch | +0.027 |
puck | +0.027 |
Neg logit contributions | |
ley | -0.042 |
rant | -0.038 |
work | -0.037 |
ky | -0.036 |
ci | -0.033 |
Token | of |
Token ID | 286 |
Feature activation | -1.25e-02 |
Pos logit contributions | |
bananas | +0.003 |
cheating | +0.003 |
banana | +0.003 |
puck | +0.003 |
unch | +0.002 |
Neg logit contributions | |
ley | -0.005 |
rant | -0.004 |
work | -0.004 |
ky | -0.004 |
ci | -0.004 |
Token | av |
Token ID | 1196 |
Feature activation | -4.13e-02 |
Pos logit contributions | |
bananas | +0.019 |
cheating | +0.018 |
puck | +0.018 |
unch | +0.018 |
banana | +0.018 |
Neg logit contributions | |
ley | -0.025 |
rant | -0.024 |
work | -0.023 |
ky | -0.022 |
ci | -0.020 |
Token | oc |
Token ID | 420 |
Feature activation | -5.74e-02 |
Pos logit contributions | |
cheating | +0.059 |
puck | +0.056 |
bananas | +0.055 |
Ready | +0.054 |
banana | +0.053 |
Neg logit contributions | |
ley | -0.090 |
ky | -0.082 |
rant | -0.082 |
work | -0.076 |
sy | -0.073 |
Token | ados |
Token ID | 22484 |
Feature activation | -9.39e-03 |
Pos logit contributions | |
puck | +0.078 |
cheating | +0.075 |
Ready | +0.072 |
sandbox | +0.071 |
banana | +0.070 |
Neg logit contributions | |
ley | -0.125 |
work | -0.116 |
rant | -0.112 |
ky | -0.109 |
ka | -0.102 |
Token | that |
Token ID | 326 |
Feature activation | -8.30e-04 |
Pos logit contributions | |
bananas | +0.018 |
banana | +0.017 |
cheating | +0.017 |
puck | +0.017 |
unch | +0.017 |
Neg logit contributions | |
ley | -0.016 |
rant | -0.014 |
work | -0.014 |
ky | -0.013 |
ci | -0.012 |
Token | the |
Token ID | 262 |
Feature activation | +2.06e-03 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.001 |
cheating | +0.001 |
catching | +0.001 |
puck | +0.001 |
Neg logit contributions | |
ley | -0.002 |
rant | -0.001 |
work | -0.001 |
ky | -0.001 |
ci | -0.001 |
Token | supplier |
Token ID | 22693 |
Feature activation | -4.29e-04 |
Pos logit contributions | |
ley | +0.004 |
work | +0.004 |
rant | +0.004 |
ky | +0.004 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.003 |
unch | -0.003 |
cheating | -0.003 |
puck | -0.003 |
banana | -0.003 |
Token | had |
Token ID | 550 |
Feature activation | -1.22e-03 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.001 |
unch | +0.001 |
cheating | +0.001 |
he | +0.001 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
ky | -0.001 |
work | -0.001 |
ci | -0.001 |
Token | s |
Token ID | 82 |
Feature activation | +1.66e-02 |
Pos logit contributions | |
ley | +0.006 |
rant | +0.005 |
work | +0.005 |
ky | +0.005 |
ci | +0.005 |
Neg logit contributions | |
bananas | -0.005 |
banana | -0.004 |
knights | -0.004 |
cheating | -0.004 |
puck | -0.004 |
Token | wrong |
Token ID | 2642 |
Feature activation | +2.12e-02 |
Pos logit contributions | |
ley | +0.040 |
rant | +0.037 |
ky | +0.036 |
work | +0.035 |
extingu | +0.033 |
Neg logit contributions | |
bananas | -0.022 |
cheating | -0.022 |
banana | -0.020 |
he | -0.019 |
stealing | -0.019 |
Token | , |
Token ID | 11 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
ley | +0.048 |
rant | +0.048 |
work | +0.043 |
ky | +0.043 |
sy | +0.039 |
Neg logit contributions | |
cheating | -0.028 |
bananas | -0.028 |
banana | -0.027 |
puck | -0.025 |
he | -0.025 |
Token | rh |
Token ID | 9529 |
Feature activation | -2.09e-02 |
Pos logit contributions | |
ley | +0.025 |
rant | +0.023 |
work | +0.022 |
ky | +0.021 |
ci | +0.020 |
Neg logit contributions | |
bananas | -0.018 |
banana | -0.016 |
cheating | -0.016 |
puck | -0.015 |
unch | -0.015 |
Token | in |
Token ID | 259 |
Feature activation | +1.12e-04 |
Pos logit contributions | |
bananas | +0.037 |
Ready | +0.034 |
puck | +0.033 |
unch | +0.033 |
banana | +0.032 |
Neg logit contributions | |
ley | -0.038 |
ky | -0.033 |
work | -0.033 |
rant | -0.032 |
ka | -0.031 |
Token | oc |
Token ID | 420 |
Feature activation | -5.70e-02 |
Pos logit contributions | |
ley | +0.000 |
rant | +0.000 |
work | +0.000 |
ky | +0.000 |
ci | +0.000 |
Neg logit contributions | |
bananas | -0.000 |
banana | -0.000 |
cheating | -0.000 |
puck | -0.000 |
unch | -0.000 |
Token | eros |
Token ID | 27498 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
bananas | +0.096 |
banana | +0.090 |
unch | +0.089 |
puck | +0.089 |
cheating | +0.089 |
Neg logit contributions | |
ley | -0.112 |
work | -0.100 |
rant | -0.097 |
ky | -0.094 |
ci | -0.089 |
Token | ? |
Token ID | 30 |
Feature activation | +7.68e-03 |
Pos logit contributions | |
ley | +0.023 |
rant | +0.021 |
work | +0.020 |
ky | +0.020 |
ci | +0.018 |
Neg logit contributions | |
bananas | -0.017 |
banana | -0.016 |
cheating | -0.016 |
catching | -0.015 |
stealing | -0.015 |
Token | â |
Token ID | 22940 |
Feature activation | +1.68e-02 |
Pos logit contributions | |
ley | +0.017 |
rant | +0.015 |
work | +0.015 |
ky | +0.014 |
ci | +0.014 |
Neg logit contributions | |
bananas | -0.012 |
banana | -0.010 |
cheating | -0.010 |
unch | -0.010 |
puck | -0.010 |
Token | € |
Token ID | 26391 |
Feature activation | +2.01e-02 |
Pos logit contributions | |
ley | +0.032 |
work | +0.027 |
rant | +0.027 |
extingu | +0.025 |
ky | +0.024 |
Neg logit contributions | |
bananas | -0.028 |
cheating | -0.027 |
banana | -0.027 |
unch | -0.026 |
mango | -0.026 |
Token | said |
Token ID | 531 |
Feature activation | +1.14e-02 |
Pos logit contributions | |
ley | +0.038 |
work | +0.034 |
ky | +0.034 |
rant | +0.033 |
ci | +0.033 |
Neg logit contributions | |
bananas | -0.035 |
cheating | -0.031 |
banana | -0.031 |
knights | -0.030 |
puck | -0.029 |
Token | learn |
Token ID | 2193 |
Feature activation | +1.14e-03 |
Pos logit contributions | |
bananas | +0.013 |
banana | +0.012 |
cheating | +0.012 |
puck | +0.011 |
unch | +0.011 |
Neg logit contributions | |
ley | -0.022 |
rant | -0.020 |
work | -0.019 |
ky | -0.019 |
ci | -0.018 |
Token | more |
Token ID | 517 |
Feature activation | -9.73e-03 |
Pos logit contributions | |
ley | +0.002 |
rant | +0.002 |
work | +0.002 |
ky | +0.002 |
ci | +0.002 |
Neg logit contributions | |
bananas | -0.002 |
banana | -0.002 |
unch | -0.002 |
cheating | -0.002 |
puck | -0.001 |
Token | about |
Token ID | 546 |
Feature activation | -5.61e-03 |
Pos logit contributions | |
bananas | +0.016 |
unch | +0.015 |
puck | +0.015 |
banana | +0.014 |
cheating | +0.014 |
Neg logit contributions | |
ley | -0.018 |
rant | -0.017 |
work | -0.017 |
ky | -0.016 |
ci | -0.015 |
Token | rh |
Token ID | 9529 |
Feature activation | -3.14e-02 |
Pos logit contributions | |
unch | +0.008 |
bananas | +0.008 |
puck | +0.007 |
banana | +0.007 |
Ready | +0.007 |
Neg logit contributions | |
ley | -0.012 |
work | -0.011 |
rant | -0.011 |
ky | -0.010 |
ci | -0.010 |
Token | in |
Token ID | 259 |
Feature activation | -1.18e-02 |
Pos logit contributions | |
bananas | +0.055 |
Ready | +0.054 |
puck | +0.053 |
mango | +0.053 |
banana | +0.053 |
Neg logit contributions | |
ley | -0.057 |
work | -0.051 |
ky | -0.048 |
rant | -0.047 |
ka | -0.045 |
Token | oc |
Token ID | 420 |
Feature activation | -5.60e-02 |
Pos logit contributions | |
bananas | +0.020 |
unch | +0.019 |
puck | +0.019 |
banana | +0.018 |
cheating | +0.018 |
Neg logit contributions | |
ley | -0.022 |
work | -0.021 |
rant | -0.019 |
ky | -0.019 |
ci | -0.018 |
Token | er |
Token ID | 263 |
Feature activation | -9.29e-03 |
Pos logit contributions | |
bananas | +0.094 |
banana | +0.091 |
puck | +0.089 |
cheating | +0.088 |
unch | +0.087 |
Neg logit contributions | |
ley | -0.108 |
work | -0.097 |
rant | -0.094 |
ky | -0.090 |
ci | -0.084 |
Token | oses |
Token ID | 4629 |
Feature activation | -1.62e-02 |
Pos logit contributions | |
bananas | +0.017 |
banana | +0.016 |
cheating | +0.015 |
unch | +0.015 |
puck | +0.015 |
Neg logit contributions | |
ley | -0.017 |
rant | -0.016 |
work | -0.015 |
ky | -0.014 |
ci | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
bananas | +0.029 |
banana | +0.027 |
cheating | +0.026 |
unch | +0.025 |
puck | +0.025 |
Neg logit contributions | |
ley | -0.030 |
rant | -0.028 |
work | -0.026 |
ky | -0.026 |
ci | -0.024 |
Token | respect |
Token ID | 2461 |
Feature activation | -4.00e-03 |
Pos logit contributions | |
unch | +0.019 |
bananas | +0.017 |
banana | +0.016 |
puck | +0.016 |
cheating | +0.016 |
Neg logit contributions | |
ley | -0.026 |
work | -0.025 |
rant | -0.023 |
ky | -0.022 |
ci | -0.022 |
Token | them |
Token ID | 606 |
Feature activation | +1.82e-03 |
Pos logit contributions | |
bananas | +0.006 |
banana | +0.006 |
unch | +0.006 |
cheating | +0.006 |
puck | +0.006 |
Neg logit contributions | |
ley | -0.008 |
rant | -0.008 |
work | -0.007 |
ky | -0.007 |
ci | -0.007 |
Token | , |
Token ID | 11 |
Feature activation | +2.61e-03 |
Pos logit contributions | |
ley | +0.003 |
rant | +0.002 |
work | +0.002 |
ky | +0.002 |
ci | +0.002 |
Neg logit contributions | |
bananas | -0.002 |
banana | -0.002 |
unch | -0.002 |
cheating | -0.002 |
puck | -0.002 |
Token | " |
Token ID | 366 |
Feature activation | +5.43e-03 |
Pos logit contributions | |
ley | +0.007 |
rant | +0.006 |
work | +0.006 |
ky | +0.006 |
ci | +0.006 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.002 |
cheating | -0.002 |
unch | -0.002 |
puck | -0.002 |
Token | I |
Token ID | 40 |
Feature activation | +2.13e-03 |
Pos logit contributions | |
ley | +0.012 |
rant | +0.011 |
work | +0.011 |
ky | +0.010 |
ci | +0.010 |
Neg logit contributions | |
bananas | -0.008 |
banana | -0.007 |
cheating | -0.007 |
unch | -0.007 |
puck | -0.007 |
Token | love |
Token ID | 1842 |
Feature activation | -8.41e-03 |
Pos logit contributions | |
ley | +0.005 |
rant | +0.005 |
work | +0.005 |
ky | +0.005 |
ci | +0.004 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
cheating | -0.003 |
unch | -0.002 |
cheat | -0.002 |
Token | av |
Token ID | 1196 |
Feature activation | -4.72e-02 |
Pos logit contributions | |
Ready | +0.011 |
puck | +0.011 |
unch | +0.011 |
coward | +0.010 |
sandbox | +0.010 |
Neg logit contributions | |
ley | -0.016 |
work | -0.016 |
rant | -0.015 |
ky | -0.015 |
ci | -0.014 |
Token | oc |
Token ID | 420 |
Feature activation | -5.57e-02 |
Pos logit contributions | |
cheating | +0.071 |
bananas | +0.069 |
Ready | +0.069 |
puck | +0.069 |
sandbox | +0.067 |
Neg logit contributions | |
ley | -0.098 |
ky | -0.090 |
rant | -0.087 |
work | -0.081 |
ci | -0.079 |
Token | ados |
Token ID | 22484 |
Feature activation | -1.52e-02 |
Pos logit contributions | |
puck | +0.079 |
Ready | +0.074 |
cheating | +0.073 |
sandbox | +0.073 |
banana | +0.071 |
Neg logit contributions | |
ley | -0.120 |
work | -0.109 |
rant | -0.102 |
ky | -0.102 |
ci | -0.098 |
Token | ! |
Token ID | 0 |
Feature activation | -5.33e-03 |
Pos logit contributions | |
bananas | +0.027 |
banana | +0.025 |
cheating | +0.025 |
unch | +0.024 |
puck | +0.024 |
Neg logit contributions | |
ley | -0.029 |
rant | -0.026 |
work | -0.025 |
ky | -0.024 |
ci | -0.023 |
Token | They |
Token ID | 1119 |
Feature activation | +1.26e-02 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.006 |
cheating | +0.006 |
unch | +0.006 |
catching | +0.006 |
Neg logit contributions | |
ley | -0.012 |
rant | -0.011 |
work | -0.011 |
ky | -0.011 |
ci | -0.010 |
Token | are |
Token ID | 389 |
Feature activation | +1.14e-02 |
Pos logit contributions | |
ley | +0.030 |
rant | +0.027 |
ky | +0.026 |
work | +0.026 |
ci | +0.025 |
Neg logit contributions | |
bananas | -0.017 |
banana | -0.016 |
cheating | -0.014 |
mango | -0.014 |
unch | -0.014 |
Token | so |
Token ID | 523 |
Feature activation | -4.10e-03 |
Pos logit contributions | |
ley | +0.024 |
rant | +0.022 |
work | +0.021 |
ky | +0.021 |
ci | +0.019 |
Neg logit contributions | |
bananas | -0.018 |
banana | -0.016 |
unch | -0.016 |
cheating | -0.016 |
puck | -0.016 |
Token | to |
Token ID | 284 |
Feature activation | +3.73e-03 |
Pos logit contributions | |
ley | +0.004 |
work | +0.003 |
rant | +0.003 |
ky | +0.003 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.004 |
unch | -0.004 |
puck | -0.004 |
banana | -0.004 |
cheating | -0.004 |
Token | him |
Token ID | 683 |
Feature activation | -5.94e-03 |
Pos logit contributions | |
ley | +0.008 |
rant | +0.008 |
work | +0.007 |
ky | +0.007 |
ci | +0.007 |
Neg logit contributions | |
bananas | -0.006 |
banana | -0.005 |
unch | -0.005 |
puck | -0.005 |
cheating | -0.005 |
Token | , |
Token ID | 11 |
Feature activation | -8.02e-04 |
Pos logit contributions | |
bananas | +0.011 |
banana | +0.010 |
puck | +0.010 |
unch | +0.010 |
cheating | +0.010 |
Neg logit contributions | |
ley | -0.011 |
rant | -0.010 |
work | -0.010 |
ky | -0.009 |
ci | -0.009 |
Token | " |
Token ID | 366 |
Feature activation | -1.78e-03 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.001 |
cheating | +0.001 |
unch | +0.001 |
puck | +0.001 |
Neg logit contributions | |
ley | -0.002 |
rant | -0.002 |
work | -0.002 |
ky | -0.002 |
ci | -0.002 |
Token | P |
Token ID | 47 |
Feature activation | -1.95e-02 |
Pos logit contributions | |
bananas | +0.003 |
banana | +0.003 |
cheating | +0.003 |
unch | +0.003 |
puck | +0.003 |
Neg logit contributions | |
ley | -0.003 |
rant | -0.003 |
ky | -0.003 |
work | -0.003 |
ci | -0.003 |
Token | ige |
Token ID | 10045 |
Feature activation | -5.57e-02 |
Pos logit contributions | |
bananas | +0.037 |
banana | +0.034 |
catching | +0.034 |
cheating | +0.034 |
puck | +0.033 |
Neg logit contributions | |
ley | -0.034 |
rant | -0.029 |
work | -0.029 |
ky | -0.028 |
sy | -0.027 |
Token | on |
Token ID | 261 |
Feature activation | -9.17e-03 |
Pos logit contributions | |
puck | +0.096 |
bananas | +0.091 |
Ready | +0.084 |
moderation | +0.083 |
winning | +0.081 |
Neg logit contributions | |
ley | -0.102 |
ina | -0.095 |
work | -0.089 |
le | -0.089 |
ka | -0.087 |
Token | , |
Token ID | 11 |
Feature activation | +3.45e-03 |
Pos logit contributions | |
bananas | +0.016 |
banana | +0.014 |
cheating | +0.014 |
unch | +0.014 |
puck | +0.014 |
Neg logit contributions | |
ley | -0.018 |
rant | -0.016 |
work | -0.016 |
ky | -0.015 |
ci | -0.015 |
Token | why |
Token ID | 1521 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
ley | +0.008 |
rant | +0.007 |
work | +0.007 |
ky | +0.006 |
ci | +0.006 |
Neg logit contributions | |
bananas | -0.005 |
banana | -0.004 |
unch | -0.004 |
cheating | -0.004 |
puck | -0.004 |
Token | are |
Token ID | 389 |
Feature activation | -1.58e-02 |
Pos logit contributions | |
ley | +0.037 |
work | +0.032 |
rant | +0.031 |
ci | +0.030 |
ky | +0.030 |
Neg logit contributions | |
bananas | -0.032 |
unch | -0.029 |
banana | -0.028 |
mango | -0.027 |
Ready | -0.027 |
Token | you |
Token ID | 345 |
Feature activation | +9.86e-03 |
Pos logit contributions | |
bananas | +0.017 |
unch | +0.017 |
banana | +0.016 |
puck | +0.015 |
mango | +0.014 |
Neg logit contributions | |
ley | -0.039 |
work | -0.036 |
rant | -0.035 |
ky | -0.034 |
ci | -0.033 |
Token |
|
Token ID | 198 |
Feature activation | -2.52e-03 |
Pos logit contributions | |
bananas | +0.006 |
cheating | +0.006 |
banana | +0.005 |
puck | +0.005 |
catching | +0.005 |
Neg logit contributions | |
ley | -0.006 |
rant | -0.005 |
ky | -0.005 |
sy | -0.004 |
work | -0.004 |
Token | " |
Token ID | 1 |
Feature activation | +6.70e-03 |
Pos logit contributions | |
bananas | +0.004 |
cheating | +0.004 |
banana | +0.004 |
catching | +0.003 |
puck | +0.003 |
Neg logit contributions | |
ley | -0.005 |
ky | -0.005 |
rant | -0.004 |
sy | -0.004 |
j | -0.004 |
Token | The |
Token ID | 464 |
Feature activation | -4.39e-03 |
Pos logit contributions | |
ley | +0.014 |
rant | +0.013 |
work | +0.012 |
ky | +0.012 |
ci | +0.011 |
Neg logit contributions | |
bananas | -0.011 |
banana | -0.010 |
cheating | -0.010 |
unch | -0.010 |
puck | -0.009 |
Token | rh |
Token ID | 9529 |
Feature activation | -2.95e-02 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.006 |
unch | +0.006 |
cheating | +0.006 |
puck | +0.006 |
Neg logit contributions | |
ley | -0.010 |
rant | -0.009 |
work | -0.009 |
ky | -0.008 |
ci | -0.008 |
Token | in |
Token ID | 259 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
bananas | +0.057 |
Ready | +0.054 |
unch | +0.051 |
swinging | +0.051 |
puck | +0.051 |
Neg logit contributions | |
ley | -0.050 |
work | -0.048 |
rant | -0.039 |
ky | -0.038 |
ka | -0.038 |
Token | oc |
Token ID | 420 |
Feature activation | -5.49e-02 |
Pos logit contributions | |
bananas | +0.016 |
banana | +0.014 |
cheating | +0.013 |
unch | +0.013 |
puck | +0.013 |
Neg logit contributions | |
ley | -0.031 |
rant | -0.029 |
work | -0.028 |
ky | -0.027 |
ci | -0.026 |
Token | eros |
Token ID | 27498 |
Feature activation | -1.29e-02 |
Pos logit contributions | |
bananas | +0.101 |
unch | +0.098 |
cheating | +0.095 |
banana | +0.095 |
puck | +0.094 |
Neg logit contributions | |
ley | -0.100 |
work | -0.091 |
rant | -0.083 |
ky | -0.080 |
ci | -0.078 |
Token | is |
Token ID | 318 |
Feature activation | +5.14e-03 |
Pos logit contributions | |
bananas | +0.021 |
banana | +0.019 |
cheating | +0.018 |
puck | +0.018 |
knights | +0.017 |
Neg logit contributions | |
ley | -0.027 |
rant | -0.025 |
work | -0.024 |
ky | -0.023 |
ci | -0.022 |
Token | in |
Token ID | 287 |
Feature activation | +6.31e-04 |
Pos logit contributions | |
ley | +0.011 |
rant | +0.010 |
work | +0.010 |
ky | +0.010 |
ci | +0.009 |
Neg logit contributions | |
bananas | -0.008 |
banana | -0.007 |
unch | -0.007 |
cheating | -0.007 |
puck | -0.007 |
Token | there |
Token ID | 612 |
Feature activation | -1.75e-02 |
Pos logit contributions | |
ley | +0.001 |
rant | +0.001 |
work | +0.001 |
ky | +0.001 |
ci | +0.001 |
Neg logit contributions | |
bananas | -0.001 |
banana | -0.001 |
cheating | -0.001 |
unch | -0.001 |
puck | -0.001 |
Token | , |
Token ID | 11 |
Feature activation | -3.20e-03 |
Pos logit contributions | |
bananas | +0.029 |
banana | +0.027 |
cheating | +0.025 |
unch | +0.025 |
puck | +0.024 |
Neg logit contributions | |
ley | -0.036 |
rant | -0.033 |
work | -0.031 |
ky | -0.031 |
ci | -0.029 |
Token | , |
Token ID | 11 |
Feature activation | -6.52e-03 |
Pos logit contributions | |
bananas | +0.040 |
banana | +0.037 |
unch | +0.037 |
cheating | +0.036 |
puck | +0.036 |
Neg logit contributions | |
ley | -0.044 |
rant | -0.040 |
work | -0.038 |
ky | -0.038 |
ci | -0.033 |
Token | fourteen |
Token ID | 29167 |
Feature activation | -2.29e-02 |
Pos logit contributions | |
bananas | +0.010 |
cheating | +0.010 |
unch | +0.010 |
banana | +0.010 |
puck | +0.009 |
Neg logit contributions | |
ley | -0.013 |
rant | -0.012 |
work | -0.012 |
ky | -0.011 |
ka | -0.010 |
Token | , |
Token ID | 11 |
Feature activation | -6.09e-03 |
Pos logit contributions | |
unch | +0.044 |
shadow | +0.043 |
Ready | +0.042 |
and | +0.042 |
puck | +0.042 |
Neg logit contributions | |
ley | -0.033 |
rant | -0.029 |
ky | -0.028 |
work | -0.027 |
colored | -0.025 |
Token | fifteen |
Token ID | 17280 |
Feature activation | -4.42e-02 |
Pos logit contributions | |
unch | +0.008 |
bananas | +0.007 |
cheating | +0.007 |
puck | +0.007 |
banana | +0.007 |
Neg logit contributions | |
ley | -0.013 |
rant | -0.013 |
work | -0.012 |
ky | -0.012 |
extingu | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | -8.87e-03 |
Pos logit contributions | |
unch | +0.081 |
Ready | +0.079 |
puck | +0.079 |
bananas | +0.078 |
cheating | +0.077 |
Neg logit contributions | |
ley | -0.069 |
rant | -0.063 |
ky | -0.062 |
work | -0.057 |
ka | -0.053 |
Token | sixteen |
Token ID | 27913 |
Feature activation | -5.44e-02 |
Pos logit contributions | |
bananas | +0.012 |
unch | +0.012 |
cheating | +0.011 |
banana | +0.011 |
puck | +0.011 |
Neg logit contributions | |
ley | -0.019 |
rant | -0.018 |
work | -0.017 |
ky | -0.017 |
extingu | -0.015 |
Token | , |
Token ID | 11 |
Feature activation | -9.93e-03 |
Pos logit contributions | |
puck | +0.100 |
unch | +0.100 |
bananas | +0.099 |
Ready | +0.098 |
cheating | +0.097 |
Neg logit contributions | |
ley | -0.083 |
rant | -0.077 |
ky | -0.075 |
work | -0.070 |
colored | -0.065 |
Token | seventeen |
Token ID | 38741 |
Feature activation | -5.21e-02 |
Pos logit contributions | |
bananas | +0.014 |
unch | +0.014 |
cheating | +0.014 |
banana | +0.013 |
knights | +0.013 |
Neg logit contributions | |
ley | -0.021 |
rant | -0.020 |
work | -0.019 |
ky | -0.019 |
extingu | -0.017 |
Token | , |
Token ID | 11 |
Feature activation | -9.47e-03 |
Pos logit contributions | |
Ready | +0.105 |
shadow | +0.102 |
puck | +0.100 |
unch | +0.100 |
coward | +0.098 |
Neg logit contributions | |
ley | -0.068 |
ky | -0.064 |
rant | -0.064 |
ka | -0.057 |
work | -0.057 |
Token | eighteen |
Token ID | 29095 |
Feature activation | -4.64e-02 |
Pos logit contributions | |
bananas | +0.013 |
unch | +0.012 |
cheating | +0.012 |
banana | +0.011 |
knights | +0.011 |
Neg logit contributions | |
ley | -0.020 |
rant | -0.020 |
work | -0.019 |
ky | -0.018 |
extingu | -0.017 |
Token | , |
Token ID | 11 |
Feature activation | -6.25e-03 |
Pos logit contributions | |
Ready | +0.096 |
puck | +0.095 |
shadow | +0.093 |
unch | +0.092 |
liar | +0.091 |
Neg logit contributions | |
ley | -0.053 |
ky | -0.047 |
rant | -0.046 |
work | -0.045 |
colored | -0.043 |
Token | , |
Token ID | 11 |
Feature activation | -1.25e-02 |
Pos logit contributions | |
bananas | +0.051 |
unch | +0.050 |
Ready | +0.049 |
cheating | +0.048 |
puck | +0.048 |
Neg logit contributions | |
ley | -0.044 |
rant | -0.041 |
ky | -0.039 |
work | -0.038 |
ka | -0.033 |
Token | fourteen |
Token ID | 29167 |
Feature activation | -2.78e-02 |
Pos logit contributions | |
bananas | +0.020 |
unch | +0.019 |
cheating | +0.019 |
banana | +0.018 |
puck | +0.018 |
Neg logit contributions | |
ley | -0.024 |
rant | -0.024 |
work | -0.022 |
ky | -0.022 |
extingu | -0.019 |
Token | , |
Token ID | 11 |
Feature activation | -1.21e-02 |
Pos logit contributions | |
unch | +0.056 |
shadow | +0.055 |
and | +0.053 |
Ready | +0.052 |
ress | +0.052 |
Neg logit contributions | |
ley | -0.036 |
rant | -0.033 |
ky | -0.032 |
work | -0.032 |
colored | -0.031 |
Token | fifteen |
Token ID | 17280 |
Feature activation | -4.75e-02 |
Pos logit contributions | |
unch | +0.016 |
bananas | +0.014 |
cheating | +0.014 |
moderation | +0.014 |
puck | +0.014 |
Neg logit contributions | |
rant | -0.025 |
work | -0.025 |
ley | -0.025 |
ky | -0.024 |
extingu | -0.022 |
Token | , |
Token ID | 11 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
unch | +0.093 |
shadow | +0.090 |
Ready | +0.089 |
puck | +0.085 |
bananas | +0.085 |
Neg logit contributions | |
ley | -0.063 |
rant | -0.060 |
ky | -0.059 |
work | -0.056 |
colored | -0.053 |
Token | sixteen |
Token ID | 27913 |
Feature activation | -5.44e-02 |
Pos logit contributions | |
unch | +0.020 |
bananas | +0.019 |
cheating | +0.019 |
puck | +0.018 |
catching | +0.018 |
Neg logit contributions | |
rant | -0.029 |
ley | -0.028 |
work | -0.028 |
ky | -0.027 |
extingu | -0.024 |
Token | , |
Token ID | 11 |
Feature activation | -1.36e-02 |
Pos logit contributions | |
shadow | +0.113 |
unch | +0.110 |
Ready | +0.108 |
coward | +0.107 |
puck | +0.105 |
Neg logit contributions | |
ley | -0.062 |
ky | -0.061 |
rant | -0.061 |
work | -0.060 |
colored | -0.058 |
Token | seventeen |
Token ID | 38741 |
Feature activation | -5.33e-02 |
Pos logit contributions | |
unch | +0.020 |
bananas | +0.020 |
cheating | +0.019 |
catching | +0.018 |
puck | +0.018 |
Neg logit contributions | |
rant | -0.027 |
ley | -0.027 |
work | -0.026 |
ky | -0.025 |
extingu | -0.022 |
Token | , |
Token ID | 11 |
Feature activation | -1.31e-02 |
Pos logit contributions | |
shadow | +0.113 |
Ready | +0.111 |
unch | +0.106 |
villains | +0.106 |
coward | +0.106 |
Neg logit contributions | |
ky | -0.061 |
ley | -0.059 |
rant | -0.059 |
colored | -0.054 |
work | -0.054 |
Token | eighteen |
Token ID | 29095 |
Feature activation | -4.87e-02 |
Pos logit contributions | |
unch | +0.018 |
bananas | +0.018 |
cheating | +0.017 |
moderation | +0.016 |
puck | +0.016 |
Neg logit contributions | |
rant | -0.027 |
ley | -0.026 |
work | -0.026 |
ky | -0.025 |
extingu | -0.022 |
Token | , |
Token ID | 11 |
Feature activation | -9.98e-03 |
Pos logit contributions | |
shadow | +0.107 |
Ready | +0.105 |
unch | +0.102 |
puck | +0.102 |
coward | +0.101 |
Neg logit contributions | |
ley | -0.043 |
work | -0.042 |
ky | -0.042 |
colored | -0.042 |
rant | -0.040 |
Token | didn |
Token ID | 1422 |
Feature activation | +1.13e-02 |
Pos logit contributions | |
bananas | +0.036 |
banana | +0.034 |
cheating | +0.032 |
unch | +0.032 |
puck | +0.032 |
Neg logit contributions | |
ley | -0.032 |
rant | -0.029 |
work | -0.028 |
ky | -0.027 |
ci | -0.025 |
Token | 't |
Token ID | 470 |
Feature activation | -3.33e-03 |
Pos logit contributions | |
ley | +0.026 |
rant | +0.024 |
work | +0.023 |
ky | +0.023 |
ci | +0.022 |
Neg logit contributions | |
bananas | -0.016 |
banana | -0.014 |
cheating | -0.013 |
unch | -0.013 |
puck | -0.013 |
Token | like |
Token ID | 588 |
Feature activation | -3.36e-03 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.005 |
unch | +0.005 |
cheating | +0.005 |
puck | +0.005 |
Neg logit contributions | |
ley | -0.007 |
rant | -0.006 |
ky | -0.006 |
work | -0.006 |
sy | -0.005 |
Token | wet |
Token ID | 9583 |
Feature activation | +6.88e-05 |
Pos logit contributions | |
bananas | +0.005 |
Ready | +0.005 |
unch | +0.005 |
puck | +0.004 |
cheating | +0.004 |
Neg logit contributions | |
ley | -0.007 |
rant | -0.006 |
work | -0.006 |
ky | -0.006 |
streaming | -0.005 |
Token | av |
Token ID | 1196 |
Feature activation | -4.68e-02 |
Pos logit contributions | |
ley | +0.000 |
rant | +0.000 |
work | +0.000 |
ky | +0.000 |
ci | +0.000 |
Neg logit contributions | |
bananas | -0.000 |
unch | -0.000 |
cheating | -0.000 |
banana | -0.000 |
Ready | -0.000 |
Token | oc |
Token ID | 420 |
Feature activation | -5.38e-02 |
Pos logit contributions | |
bananas | +0.064 |
cheating | +0.063 |
Ready | +0.062 |
puck | +0.060 |
knights | +0.059 |
Neg logit contributions | |
ley | -0.102 |
ky | -0.094 |
rant | -0.093 |
work | -0.084 |
sy | -0.084 |
Token | ados |
Token ID | 22484 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
puck | +0.073 |
sandbox | +0.072 |
Ready | +0.070 |
cheating | +0.070 |
banana | +0.066 |
Neg logit contributions | |
ley | -0.111 |
work | -0.102 |
rant | -0.099 |
ky | -0.097 |
ka | -0.093 |
Token | , |
Token ID | 11 |
Feature activation | +1.77e-03 |
Pos logit contributions | |
bananas | +0.027 |
banana | +0.025 |
cheating | +0.024 |
catching | +0.023 |
unch | +0.023 |
Neg logit contributions | |
ley | -0.031 |
rant | -0.029 |
work | -0.027 |
ky | -0.027 |
ci | -0.025 |
Token | so |
Token ID | 523 |
Feature activation | -7.50e-04 |
Pos logit contributions | |
ley | +0.003 |
rant | +0.003 |
ky | +0.003 |
work | +0.003 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
unch | -0.003 |
cheating | -0.003 |
Ready | -0.003 |
Token | he |
Token ID | 339 |
Feature activation | -9.74e-04 |
Pos logit contributions | |
bananas | +0.001 |
cheating | +0.000 |
banana | +0.000 |
Ready | +0.000 |
unch | +0.000 |
Neg logit contributions | |
ley | -0.002 |
rant | -0.002 |
ky | -0.002 |
work | -0.002 |
ci | -0.002 |
Token | decided |
Token ID | 3066 |
Feature activation | +3.76e-03 |
Pos logit contributions | |
bananas | +0.002 |
banana | +0.002 |
puck | +0.002 |
knights | +0.002 |
he | +0.002 |
Neg logit contributions | |
ley | -0.002 |
work | -0.002 |
rant | -0.002 |
extingu | -0.002 |
ky | -0.002 |
Token | , |
Token ID | 11 |
Feature activation | -1.21e-02 |
Pos logit contributions | |
unch | +0.056 |
shadow | +0.055 |
and | +0.053 |
Ready | +0.052 |
ress | +0.052 |
Neg logit contributions | |
ley | -0.036 |
rant | -0.033 |
ky | -0.032 |
work | -0.032 |
colored | -0.031 |
Token | fifteen |
Token ID | 17280 |
Feature activation | -4.75e-02 |
Pos logit contributions | |
unch | +0.016 |
bananas | +0.014 |
cheating | +0.014 |
moderation | +0.014 |
puck | +0.014 |
Neg logit contributions | |
rant | -0.025 |
work | -0.025 |
ley | -0.025 |
ky | -0.024 |
extingu | -0.022 |
Token | , |
Token ID | 11 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
unch | +0.093 |
shadow | +0.090 |
Ready | +0.089 |
puck | +0.085 |
bananas | +0.085 |
Neg logit contributions | |
ley | -0.063 |
rant | -0.060 |
ky | -0.059 |
work | -0.056 |
colored | -0.053 |
Token | sixteen |
Token ID | 27913 |
Feature activation | -5.44e-02 |
Pos logit contributions | |
unch | +0.020 |
bananas | +0.019 |
cheating | +0.019 |
puck | +0.018 |
catching | +0.018 |
Neg logit contributions | |
rant | -0.029 |
ley | -0.028 |
work | -0.028 |
ky | -0.027 |
extingu | -0.024 |
Token | , |
Token ID | 11 |
Feature activation | -1.36e-02 |
Pos logit contributions | |
shadow | +0.113 |
unch | +0.110 |
Ready | +0.108 |
coward | +0.107 |
puck | +0.105 |
Neg logit contributions | |
ley | -0.062 |
ky | -0.061 |
rant | -0.061 |
work | -0.060 |
colored | -0.058 |
Token | seventeen |
Token ID | 38741 |
Feature activation | -5.33e-02 |
Pos logit contributions | |
unch | +0.020 |
bananas | +0.020 |
cheating | +0.019 |
catching | +0.018 |
puck | +0.018 |
Neg logit contributions | |
rant | -0.027 |
ley | -0.027 |
work | -0.026 |
ky | -0.025 |
extingu | -0.022 |
Token | , |
Token ID | 11 |
Feature activation | -1.31e-02 |
Pos logit contributions | |
shadow | +0.113 |
Ready | +0.111 |
unch | +0.106 |
villains | +0.106 |
coward | +0.106 |
Neg logit contributions | |
ky | -0.061 |
ley | -0.059 |
rant | -0.059 |
colored | -0.054 |
work | -0.054 |
Token | eighteen |
Token ID | 29095 |
Feature activation | -4.87e-02 |
Pos logit contributions | |
unch | +0.018 |
bananas | +0.018 |
cheating | +0.017 |
moderation | +0.016 |
puck | +0.016 |
Neg logit contributions | |
rant | -0.027 |
ley | -0.026 |
work | -0.026 |
ky | -0.025 |
extingu | -0.022 |
Token | , |
Token ID | 11 |
Feature activation | -9.98e-03 |
Pos logit contributions | |
shadow | +0.107 |
Ready | +0.105 |
unch | +0.102 |
puck | +0.102 |
coward | +0.101 |
Neg logit contributions | |
ley | -0.043 |
work | -0.042 |
ky | -0.042 |
colored | -0.042 |
rant | -0.040 |
Token | nineteen |
Token ID | 43063 |
Feature activation | -3.34e-02 |
Pos logit contributions | |
unch | +0.014 |
bananas | +0.013 |
cheating | +0.013 |
moderation | +0.012 |
puck | +0.012 |
Neg logit contributions | |
rant | -0.020 |
work | -0.020 |
ley | -0.020 |
ky | -0.019 |
extingu | -0.017 |
Token | , |
Token ID | 11 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
bananas | +0.064 |
unch | +0.063 |
puck | +0.062 |
banana | +0.060 |
Ready | +0.060 |
Neg logit contributions | |
ley | -0.048 |
rant | -0.047 |
work | -0.044 |
ky | -0.042 |
colored | -0.037 |
Token | tree |
Token ID | 5509 |
Feature activation | -8.54e-03 |
Pos logit contributions | |
unch | +0.010 |
cheating | +0.010 |
puck | +0.010 |
shadow | +0.010 |
bananas | +0.009 |
Neg logit contributions | |
ley | -0.014 |
work | -0.012 |
rant | -0.012 |
ky | -0.012 |
sy | -0.011 |
Token | if |
Token ID | 611 |
Feature activation | -2.18e-03 |
Pos logit contributions | |
puck | +0.016 |
cheating | +0.016 |
unch | +0.016 |
bananas | +0.016 |
cheat | +0.015 |
Neg logit contributions | |
ley | -0.014 |
work | -0.013 |
rant | -0.012 |
ci | -0.011 |
sy | -0.010 |
Token | they |
Token ID | 484 |
Feature activation | +4.71e-03 |
Pos logit contributions | |
bananas | +0.003 |
unch | +0.003 |
banana | +0.002 |
puck | +0.002 |
cheating | +0.002 |
Neg logit contributions | |
ley | -0.005 |
work | -0.005 |
rant | -0.005 |
ci | -0.004 |
ky | -0.004 |
Token | were |
Token ID | 547 |
Feature activation | -1.06e-02 |
Pos logit contributions | |
ley | +0.010 |
rant | +0.009 |
work | +0.009 |
ky | +0.008 |
ci | +0.008 |
Neg logit contributions | |
bananas | -0.008 |
banana | -0.007 |
he | -0.007 |
beetles | -0.007 |
knights | -0.007 |
Token | av |
Token ID | 1196 |
Feature activation | -4.27e-02 |
Pos logit contributions | |
bananas | +0.016 |
unch | +0.014 |
banana | +0.014 |
puck | +0.014 |
cheating | +0.014 |
Neg logit contributions | |
ley | -0.022 |
rant | -0.021 |
work | -0.020 |
ky | -0.020 |
sy | -0.018 |
Token | oc |
Token ID | 420 |
Feature activation | -5.28e-02 |
Pos logit contributions | |
bananas | +0.059 |
cheating | +0.059 |
banana | +0.055 |
puck | +0.055 |
Ready | +0.055 |
Neg logit contributions | |
ley | -0.096 |
ky | -0.086 |
rant | -0.084 |
work | -0.079 |
ci | -0.077 |
Token | ados |
Token ID | 22484 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
puck | +0.066 |
banana | +0.064 |
bananas | +0.063 |
cheating | +0.062 |
Ready | +0.062 |
Neg logit contributions | |
ley | -0.122 |
work | -0.110 |
rant | -0.106 |
ky | -0.105 |
ci | -0.100 |
Token | , |
Token ID | 11 |
Feature activation | +2.12e-03 |
Pos logit contributions | |
bananas | +0.021 |
banana | +0.019 |
cheating | +0.018 |
unch | +0.018 |
puck | +0.018 |
Neg logit contributions | |
ley | -0.026 |
rant | -0.024 |
work | -0.023 |
ky | -0.023 |
ci | -0.021 |
Token | and |
Token ID | 290 |
Feature activation | -8.46e-03 |
Pos logit contributions | |
ley | +0.004 |
rant | +0.004 |
work | +0.004 |
ky | +0.004 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
unch | -0.003 |
cheating | -0.003 |
puck | -0.003 |
Token | they |
Token ID | 484 |
Feature activation | -3.84e-04 |
Pos logit contributions | |
bananas | +0.011 |
banana | +0.010 |
unch | +0.010 |
cheating | +0.010 |
puck | +0.010 |
Neg logit contributions | |
ley | -0.019 |
rant | -0.018 |
work | -0.017 |
ky | -0.017 |
ci | -0.016 |
Token | said |
Token ID | 531 |
Feature activation | -6.25e-05 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.001 |
cheating | +0.001 |
knights | +0.001 |
puck | +0.001 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
work | -0.001 |
ky | -0.001 |
ci | -0.001 |
Token | pepper |
Token ID | 13385 |
Feature activation | -2.71e-02 |
Pos logit contributions | |
bananas | +0.005 |
unch | +0.004 |
cheating | +0.004 |
Ready | +0.004 |
banana | +0.004 |
Neg logit contributions | |
ley | -0.009 |
rant | -0.008 |
work | -0.008 |
ky | -0.008 |
ci | -0.007 |
Token | oni |
Token ID | 14651 |
Feature activation | -2.79e-02 |
Pos logit contributions | |
bananas | +0.044 |
banana | +0.040 |
cheating | +0.038 |
unch | +0.037 |
puck | +0.037 |
Neg logit contributions | |
ley | -0.057 |
rant | -0.052 |
work | -0.050 |
ky | -0.049 |
ci | -0.047 |
Token | , |
Token ID | 11 |
Feature activation | -7.92e-03 |
Pos logit contributions | |
bananas | +0.049 |
cheating | +0.046 |
Ready | +0.046 |
unch | +0.046 |
banana | +0.045 |
Neg logit contributions | |
ley | -0.050 |
rant | -0.044 |
work | -0.044 |
ky | -0.042 |
sy | -0.037 |
Token | mushroom |
Token ID | 28520 |
Feature activation | -6.49e-03 |
Pos logit contributions | |
bananas | +0.012 |
cheating | +0.011 |
banana | +0.011 |
unch | +0.011 |
puck | +0.010 |
Neg logit contributions | |
ley | -0.017 |
rant | -0.016 |
work | -0.015 |
ky | -0.015 |
ci | -0.014 |
Token | , |
Token ID | 11 |
Feature activation | -8.07e-03 |
Pos logit contributions | |
bananas | +0.011 |
banana | +0.010 |
cheating | +0.010 |
unch | +0.010 |
puck | +0.010 |
Neg logit contributions | |
ley | -0.013 |
rant | -0.012 |
work | -0.011 |
ky | -0.011 |
ci | -0.010 |
Token | pineapple |
Token ID | 45540 |
Feature activation | -5.28e-02 |
Pos logit contributions | |
bananas | +0.015 |
banana | +0.014 |
cheating | +0.014 |
unch | +0.013 |
puck | +0.013 |
Neg logit contributions | |
ley | -0.015 |
rant | -0.013 |
work | -0.013 |
ky | -0.012 |
ci | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | -1.00e-02 |
Pos logit contributions | |
bananas | +0.094 |
cheating | +0.091 |
Ready | +0.090 |
unch | +0.089 |
shadow | +0.088 |
Neg logit contributions | |
ley | -0.092 |
rant | -0.082 |
work | -0.078 |
ky | -0.077 |
sy | -0.067 |
Token | and |
Token ID | 290 |
Feature activation | -2.48e-02 |
Pos logit contributions | |
bananas | +0.018 |
cheating | +0.017 |
banana | +0.017 |
unch | +0.017 |
puck | +0.016 |
Neg logit contributions | |
ley | -0.018 |
rant | -0.017 |
work | -0.016 |
ky | -0.015 |
ci | -0.014 |
Token | more |
Token ID | 517 |
Feature activation | -1.74e-02 |
Pos logit contributions | |
unch | +0.031 |
bananas | +0.030 |
cheating | +0.029 |
Ready | +0.027 |
banana | +0.027 |
Neg logit contributions | |
ley | -0.056 |
rant | -0.053 |
work | -0.051 |
ky | -0.050 |
ci | -0.046 |
Token | . |
Token ID | 13 |
Feature activation | -2.33e-03 |
Pos logit contributions | |
bananas | +0.031 |
banana | +0.029 |
cheating | +0.029 |
unch | +0.028 |
puck | +0.028 |
Neg logit contributions | |
ley | -0.032 |
rant | -0.029 |
work | -0.028 |
ky | -0.027 |
ci | -0.025 |
Token | She |
Token ID | 1375 |
Feature activation | -6.33e-03 |
Pos logit contributions | |
bananas | +0.004 |
cheating | +0.004 |
puck | +0.004 |
banana | +0.004 |
catching | +0.004 |
Neg logit contributions | |
ley | -0.004 |
rant | -0.004 |
work | -0.003 |
ky | -0.003 |
ci | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | -6.09e-03 |
Pos logit contributions | |
unch | +0.044 |
shadow | +0.043 |
Ready | +0.042 |
and | +0.042 |
puck | +0.042 |
Neg logit contributions | |
ley | -0.033 |
rant | -0.029 |
ky | -0.028 |
work | -0.027 |
colored | -0.025 |
Token | fifteen |
Token ID | 17280 |
Feature activation | -4.42e-02 |
Pos logit contributions | |
unch | +0.008 |
bananas | +0.007 |
cheating | +0.007 |
puck | +0.007 |
banana | +0.007 |
Neg logit contributions | |
ley | -0.013 |
rant | -0.013 |
work | -0.012 |
ky | -0.012 |
extingu | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | -8.87e-03 |
Pos logit contributions | |
unch | +0.081 |
Ready | +0.079 |
puck | +0.079 |
bananas | +0.078 |
cheating | +0.077 |
Neg logit contributions | |
ley | -0.069 |
rant | -0.063 |
ky | -0.062 |
work | -0.057 |
ka | -0.053 |
Token | sixteen |
Token ID | 27913 |
Feature activation | -5.44e-02 |
Pos logit contributions | |
bananas | +0.012 |
unch | +0.012 |
cheating | +0.011 |
banana | +0.011 |
puck | +0.011 |
Neg logit contributions | |
ley | -0.019 |
rant | -0.018 |
work | -0.017 |
ky | -0.017 |
extingu | -0.015 |
Token | , |
Token ID | 11 |
Feature activation | -9.93e-03 |
Pos logit contributions | |
puck | +0.100 |
unch | +0.100 |
bananas | +0.099 |
Ready | +0.098 |
cheating | +0.097 |
Neg logit contributions | |
ley | -0.083 |
rant | -0.077 |
ky | -0.075 |
work | -0.070 |
colored | -0.065 |
Token | seventeen |
Token ID | 38741 |
Feature activation | -5.21e-02 |
Pos logit contributions | |
bananas | +0.014 |
unch | +0.014 |
cheating | +0.014 |
banana | +0.013 |
knights | +0.013 |
Neg logit contributions | |
ley | -0.021 |
rant | -0.020 |
work | -0.019 |
ky | -0.019 |
extingu | -0.017 |
Token | , |
Token ID | 11 |
Feature activation | -9.47e-03 |
Pos logit contributions | |
Ready | +0.105 |
shadow | +0.102 |
puck | +0.100 |
unch | +0.100 |
coward | +0.098 |
Neg logit contributions | |
ley | -0.068 |
ky | -0.064 |
rant | -0.064 |
ka | -0.057 |
work | -0.057 |
Token | eighteen |
Token ID | 29095 |
Feature activation | -4.64e-02 |
Pos logit contributions | |
bananas | +0.013 |
unch | +0.012 |
cheating | +0.012 |
banana | +0.011 |
knights | +0.011 |
Neg logit contributions | |
ley | -0.020 |
rant | -0.020 |
work | -0.019 |
ky | -0.018 |
extingu | -0.017 |
Token | , |
Token ID | 11 |
Feature activation | -6.25e-03 |
Pos logit contributions | |
Ready | +0.096 |
puck | +0.095 |
shadow | +0.093 |
unch | +0.092 |
liar | +0.091 |
Neg logit contributions | |
ley | -0.053 |
ky | -0.047 |
rant | -0.046 |
work | -0.045 |
colored | -0.043 |
Token | nineteen |
Token ID | 43063 |
Feature activation | -2.91e-02 |
Pos logit contributions | |
unch | +0.008 |
bananas | +0.008 |
cheating | +0.008 |
moderation | +0.007 |
puck | +0.007 |
Neg logit contributions | |
ley | -0.013 |
rant | -0.013 |
work | -0.013 |
ky | -0.012 |
extingu | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | -1.36e-02 |
Pos logit contributions | |
bananas | +0.053 |
unch | +0.051 |
puck | +0.050 |
banana | +0.050 |
cheating | +0.049 |
Neg logit contributions | |
ley | -0.048 |
rant | -0.045 |
work | -0.042 |
ky | -0.042 |
ka | -0.036 |
Token | , |
Token ID | 11 |
Feature activation | +1.24e-02 |
Pos logit contributions | |
ley | +0.003 |
rant | +0.002 |
work | +0.002 |
ky | +0.002 |
ci | +0.002 |
Neg logit contributions | |
bananas | -0.002 |
banana | -0.002 |
cheating | -0.002 |
unch | -0.002 |
puck | -0.002 |
Token | " |
Token ID | 366 |
Feature activation | -8.09e-03 |
Pos logit contributions | |
ley | +0.032 |
rant | +0.030 |
work | +0.029 |
ky | +0.028 |
ci | +0.027 |
Neg logit contributions | |
bananas | -0.013 |
banana | -0.011 |
cheating | -0.011 |
unch | -0.011 |
puck | -0.011 |
Token | The |
Token ID | 464 |
Feature activation | +5.22e-03 |
Pos logit contributions | |
bananas | +0.014 |
banana | +0.013 |
cheating | +0.012 |
unch | +0.012 |
knights | +0.012 |
Neg logit contributions | |
ley | -0.016 |
rant | -0.014 |
ky | -0.014 |
work | -0.014 |
ci | -0.013 |
Token | rh |
Token ID | 9529 |
Feature activation | -2.97e-02 |
Pos logit contributions | |
ley | +0.011 |
rant | +0.010 |
work | +0.010 |
ky | +0.009 |
ci | +0.009 |
Neg logit contributions | |
bananas | -0.008 |
banana | -0.008 |
cheating | -0.007 |
unch | -0.007 |
puck | -0.007 |
Token | in |
Token ID | 259 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
bananas | +0.060 |
Ready | +0.056 |
swinging | +0.053 |
catching | +0.053 |
cheating | +0.052 |
Neg logit contributions | |
ley | -0.049 |
work | -0.046 |
ky | -0.039 |
rant | -0.038 |
ka | -0.037 |
Token | oc |
Token ID | 420 |
Feature activation | -5.17e-02 |
Pos logit contributions | |
bananas | +0.013 |
banana | +0.011 |
cheating | +0.011 |
unch | +0.010 |
puck | +0.010 |
Neg logit contributions | |
ley | -0.028 |
rant | -0.026 |
work | -0.025 |
ky | -0.024 |
ci | -0.023 |
Token | eros |
Token ID | 27498 |
Feature activation | -1.15e-02 |
Pos logit contributions | |
bananas | +0.095 |
unch | +0.089 |
banana | +0.088 |
cheating | +0.088 |
puck | +0.087 |
Neg logit contributions | |
ley | -0.096 |
work | -0.086 |
rant | -0.081 |
ky | -0.078 |
ci | -0.074 |
Token | made |
Token ID | 925 |
Feature activation | -6.10e-04 |
Pos logit contributions | |
bananas | +0.018 |
banana | +0.017 |
cheating | +0.016 |
knights | +0.015 |
puck | +0.015 |
Neg logit contributions | |
ley | -0.024 |
rant | -0.023 |
work | -0.021 |
ky | -0.021 |
ci | -0.020 |
Token | a |
Token ID | 257 |
Feature activation | -7.14e-03 |
Pos logit contributions | |
bananas | +0.001 |
unch | +0.001 |
banana | +0.001 |
puck | +0.001 |
cheating | +0.001 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
work | -0.001 |
ky | -0.001 |
sy | -0.001 |
Token | big |
Token ID | 1263 |
Feature activation | -2.41e-03 |
Pos logit contributions | |
bananas | +0.012 |
unch | +0.011 |
cheating | +0.010 |
knights | +0.010 |
banana | +0.010 |
Neg logit contributions | |
ley | -0.014 |
work | -0.013 |
rant | -0.013 |
ky | -0.012 |
sy | -0.011 |
Token | mess |
Token ID | 2085 |
Feature activation | +5.62e-03 |
Pos logit contributions | |
bananas | +0.004 |
banana | +0.003 |
cheating | +0.003 |
unch | +0.003 |
puck | +0.003 |
Neg logit contributions | |
ley | -0.005 |
rant | -0.005 |
work | -0.005 |
ky | -0.005 |
ci | -0.004 |
Token | The |
Token ID | 464 |
Feature activation | -1.57e-04 |
Pos logit contributions | |
ley | +0.003 |
rant | +0.003 |
work | +0.003 |
ky | +0.003 |
ci | +0.002 |
Neg logit contributions | |
bananas | -0.002 |
banana | -0.002 |
cheating | -0.002 |
knights | -0.002 |
unch | -0.002 |
Token | le |
Token ID | 443 |
Feature activation | -2.09e-02 |
Pos logit contributions | |
bananas | +0.000 |
unch | +0.000 |
knights | +0.000 |
puck | +0.000 |
beetles | +0.000 |
Neg logit contributions | |
ley | -0.000 |
rant | -0.000 |
work | -0.000 |
ky | -0.000 |
ci | -0.000 |
Token | opard |
Token ID | 15478 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
bananas | +0.038 |
cheating | +0.037 |
catching | +0.035 |
moderation | +0.035 |
knights | +0.035 |
Neg logit contributions | |
ley | -0.035 |
work | -0.030 |
rant | -0.029 |
ky | -0.029 |
ci | -0.027 |
Token | toy |
Token ID | 13373 |
Feature activation | -8.44e-03 |
Pos logit contributions | |
bananas | +0.022 |
banana | +0.020 |
cheating | +0.019 |
knights | +0.019 |
puck | +0.018 |
Neg logit contributions | |
ley | -0.029 |
rant | -0.028 |
ky | -0.026 |
work | -0.026 |
ci | -0.024 |
Token | costs |
Token ID | 3484 |
Feature activation | -2.69e-02 |
Pos logit contributions | |
bananas | +0.013 |
banana | +0.012 |
cheating | +0.011 |
puck | +0.011 |
unch | +0.011 |
Neg logit contributions | |
ley | -0.018 |
rant | -0.017 |
work | -0.016 |
ky | -0.016 |
ci | -0.015 |
Token | $ |
Token ID | 720 |
Feature activation | -5.15e-02 |
Pos logit contributions | |
bananas | +0.039 |
unch | +0.037 |
banana | +0.037 |
sandbox | +0.035 |
puck | +0.035 |
Neg logit contributions | |
ley | -0.056 |
rant | -0.051 |
ky | -0.050 |
work | -0.049 |
ka | -0.045 |
Token | 10 |
Token ID | 940 |
Feature activation | -1.98e-02 |
Pos logit contributions | |
bananas | +0.070 |
banana | +0.067 |
cheating | +0.063 |
catching | +0.062 |
he | +0.060 |
Neg logit contributions | |
ley | -0.115 |
ky | -0.108 |
rant | -0.105 |
work | -0.099 |
ka | -0.097 |
Token | ." |
Token ID | 526 |
Feature activation | -4.46e-04 |
Pos logit contributions | |
bananas | +0.032 |
banana | +0.030 |
unch | +0.029 |
cheating | +0.029 |
puck | +0.028 |
Neg logit contributions | |
ley | -0.040 |
rant | -0.037 |
work | -0.035 |
ky | -0.035 |
ci | -0.032 |
Token | |
Token ID | 220 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
bananas | +0.001 |
cheating | +0.001 |
banana | +0.001 |
puck | +0.001 |
knights | +0.001 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
work | -0.001 |
ky | -0.001 |
ci | -0.001 |
Token |
|
Token ID | 198 |
Feature activation | +1.47e-03 |
Pos logit contributions | |
bananas | +0.019 |
cheating | +0.017 |
banana | +0.017 |
puck | +0.017 |
beetles | +0.016 |
Neg logit contributions | |
ley | -0.018 |
rant | -0.016 |
work | -0.016 |
ky | -0.014 |
sy | -0.014 |
Token |
|
Token ID | 198 |
Feature activation | -1.26e-03 |
Pos logit contributions | |
ley | +0.003 |
rant | +0.003 |
work | +0.003 |
ky | +0.003 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.002 |
cheating | -0.002 |
banana | -0.002 |
knights | -0.002 |
puck | -0.002 |
Token | that |
Token ID | 326 |
Feature activation | -1.29e-02 |
Pos logit contributions | |
ley | +0.000 |
rant | +0.000 |
work | +0.000 |
ky | +0.000 |
ci | +0.000 |
Neg logit contributions | |
bananas | -0.000 |
banana | -0.000 |
unch | -0.000 |
cheating | -0.000 |
puck | -0.000 |
Token | boss |
Token ID | 6478 |
Feature activation | -9.79e-04 |
Pos logit contributions | |
bananas | +0.017 |
unch | +0.015 |
banana | +0.015 |
cheating | +0.014 |
Ready | +0.014 |
Neg logit contributions | |
ley | -0.029 |
work | -0.027 |
ci | -0.026 |
rant | -0.026 |
ky | -0.025 |
Token | y |
Token ID | 88 |
Feature activation | -8.67e-03 |
Pos logit contributions | |
bananas | +0.002 |
banana | +0.002 |
cheating | +0.002 |
he | +0.002 |
puck | +0.002 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
ky | -0.001 |
work | -0.001 |
extingu | -0.001 |
Token | rh |
Token ID | 9529 |
Feature activation | -2.79e-02 |
Pos logit contributions | |
bananas | +0.015 |
banana | +0.014 |
cheating | +0.013 |
puck | +0.013 |
knights | +0.013 |
Neg logit contributions | |
ley | -0.017 |
rant | -0.016 |
ky | -0.015 |
work | -0.015 |
ci | -0.014 |
Token | in |
Token ID | 259 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
bananas | +0.052 |
Ready | +0.049 |
banana | +0.047 |
puck | +0.047 |
mango | +0.047 |
Neg logit contributions | |
ley | -0.050 |
work | -0.045 |
rant | -0.042 |
ky | -0.042 |
ka | -0.039 |
Token | oc |
Token ID | 420 |
Feature activation | -5.12e-02 |
Pos logit contributions | |
bananas | +0.018 |
banana | +0.016 |
cheating | +0.016 |
unch | +0.015 |
puck | +0.015 |
Neg logit contributions | |
ley | -0.025 |
rant | -0.023 |
work | -0.022 |
ky | -0.021 |
ci | -0.020 |
Token | eros |
Token ID | 27498 |
Feature activation | -5.51e-03 |
Pos logit contributions | |
bananas | +0.089 |
banana | +0.083 |
cheating | +0.081 |
puck | +0.081 |
unch | +0.080 |
Neg logit contributions | |
ley | -0.100 |
work | -0.088 |
rant | -0.088 |
ky | -0.084 |
ci | -0.078 |
Token | . |
Token ID | 13 |
Feature activation | +3.83e-03 |
Pos logit contributions | |
bananas | +0.010 |
banana | +0.009 |
cheating | +0.009 |
knights | +0.008 |
catching | +0.008 |
Neg logit contributions | |
ley | -0.010 |
rant | -0.010 |
ky | -0.009 |
work | -0.009 |
ci | -0.008 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -3.88e-03 |
Pos logit contributions | |
ley | +0.008 |
work | +0.007 |
rant | +0.007 |
ky | +0.006 |
ci | +0.006 |
Neg logit contributions | |
bananas | -0.006 |
cheating | -0.006 |
banana | -0.006 |
puck | -0.006 |
knights | -0.005 |
Token | D |
Token ID | 35 |
Feature activation | +1.41e-02 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.005 |
cheating | +0.004 |
unch | +0.004 |
puck | +0.004 |
Neg logit contributions | |
ley | -0.009 |
rant | -0.008 |
work | -0.008 |
ky | -0.008 |
ci | -0.008 |
Token | ee |
Token ID | 1453 |
Feature activation | +1.06e-02 |
Pos logit contributions | |
ley | +0.026 |
rant | +0.024 |
work | +0.024 |
ky | +0.023 |
ci | +0.022 |
Neg logit contributions | |
bananas | -0.024 |
banana | -0.023 |
unch | -0.023 |
puck | -0.021 |
mango | -0.021 |
Token | and |
Token ID | 290 |
Feature activation | -1.93e-02 |
Pos logit contributions | |
unch | +0.015 |
cheating | +0.014 |
bananas | +0.014 |
Ready | +0.014 |
shadow | +0.014 |
Neg logit contributions | |
ley | -0.010 |
rant | -0.009 |
ky | -0.009 |
work | -0.008 |
ka | -0.007 |
Token | pige |
Token ID | 32372 |
Feature activation | -4.00e-02 |
Pos logit contributions | |
unch | +0.029 |
shadow | +0.024 |
cheating | +0.024 |
ress | +0.024 |
bananas | +0.023 |
Neg logit contributions | |
ley | -0.036 |
rant | -0.036 |
work | -0.035 |
ky | -0.035 |
colored | -0.033 |
Token | ons |
Token ID | 684 |
Feature activation | -3.26e-03 |
Pos logit contributions | |
puck | +0.053 |
Ready | +0.049 |
bananas | +0.049 |
cheating | +0.049 |
oh | +0.047 |
Neg logit contributions | |
ley | -0.085 |
rant | -0.083 |
ky | -0.082 |
work | -0.076 |
sy | -0.072 |
Token | . |
Token ID | 13 |
Feature activation | -1.52e-03 |
Pos logit contributions | |
bananas | +0.006 |
banana | +0.005 |
cheating | +0.005 |
knights | +0.005 |
puck | +0.005 |
Neg logit contributions | |
ley | -0.006 |
rant | -0.006 |
work | -0.005 |
ky | -0.005 |
ci | -0.005 |
Token | P |
Token ID | 350 |
Feature activation | +5.53e-03 |
Pos logit contributions | |
bananas | +0.003 |
banana | +0.002 |
cheating | +0.002 |
unch | +0.002 |
puck | +0.002 |
Neg logit contributions | |
ley | -0.003 |
rant | -0.003 |
work | -0.002 |
ky | -0.002 |
ci | -0.002 |
Token | ige |
Token ID | 10045 |
Feature activation | -5.11e-02 |
Pos logit contributions | |
ley | +0.011 |
rant | +0.010 |
work | +0.010 |
ky | +0.009 |
ci | +0.009 |
Neg logit contributions | |
bananas | -0.009 |
banana | -0.009 |
cheating | -0.008 |
puck | -0.008 |
unch | -0.008 |
Token | ons |
Token ID | 684 |
Feature activation | -8.30e-04 |
Pos logit contributions | |
puck | +0.082 |
cheating | +0.081 |
bananas | +0.081 |
unch | +0.081 |
Ready | +0.081 |
Neg logit contributions | |
ley | -0.098 |
rant | -0.089 |
ky | -0.084 |
work | -0.083 |
colored | -0.078 |
Token | are |
Token ID | 389 |
Feature activation | -3.53e-04 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.001 |
catching | +0.001 |
he | +0.001 |
cheating | +0.001 |
Neg logit contributions | |
ley | -0.002 |
rant | -0.002 |
ky | -0.002 |
ci | -0.002 |
work | -0.002 |
Token | birds |
Token ID | 10087 |
Feature activation | +4.90e-03 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.001 |
cheating | +0.001 |
unch | +0.001 |
puck | +0.000 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
work | -0.001 |
ky | -0.001 |
ci | -0.001 |
Token | that |
Token ID | 326 |
Feature activation | -4.43e-03 |
Pos logit contributions | |
ley | +0.011 |
rant | +0.010 |
work | +0.009 |
ky | +0.009 |
ci | +0.009 |
Neg logit contributions | |
bananas | -0.007 |
banana | -0.007 |
unch | -0.006 |
catching | -0.006 |
swinging | -0.006 |
Token | have |
Token ID | 423 |
Feature activation | -2.01e-02 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.006 |
cheating | +0.005 |
catching | +0.005 |
puck | +0.005 |
Neg logit contributions | |
ley | -0.010 |
rant | -0.009 |
work | -0.009 |
ky | -0.009 |
ci | -0.008 |
Token | see |
Token ID | 766 |
Feature activation | -2.23e-02 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.007 |
cheating | +0.006 |
unch | +0.006 |
puck | +0.006 |
Neg logit contributions | |
ley | -0.008 |
rant | -0.008 |
work | -0.007 |
ky | -0.007 |
ci | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
bananas | +0.041 |
banana | +0.038 |
cheating | +0.037 |
unch | +0.037 |
puck | +0.036 |
Neg logit contributions | |
ley | -0.041 |
rant | -0.038 |
work | -0.036 |
ky | -0.035 |
ci | -0.032 |
Token | They |
Token ID | 1119 |
Feature activation | +4.09e-03 |
Pos logit contributions | |
ley | +0.026 |
rant | +0.024 |
ky | +0.022 |
work | +0.022 |
ci | +0.022 |
Neg logit contributions | |
bananas | -0.018 |
Ready | -0.017 |
banana | -0.016 |
knights | -0.016 |
he | -0.016 |
Token | see |
Token ID | 766 |
Feature activation | -2.28e-02 |
Pos logit contributions | |
ley | +0.010 |
rant | +0.009 |
work | +0.009 |
ky | +0.008 |
ci | +0.008 |
Neg logit contributions | |
bananas | -0.006 |
banana | -0.005 |
cheating | -0.005 |
unch | -0.005 |
puck | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -3.69e-03 |
Pos logit contributions | |
bananas | +0.034 |
unch | +0.032 |
banana | +0.032 |
cheating | +0.031 |
puck | +0.031 |
Neg logit contributions | |
ley | -0.049 |
rant | -0.045 |
work | -0.044 |
ky | -0.042 |
ci | -0.039 |
Token | pump |
Token ID | 8901 |
Feature activation | -5.11e-02 |
Pos logit contributions | |
bananas | +0.006 |
banana | +0.006 |
cheating | +0.005 |
unch | +0.005 |
puck | +0.005 |
Neg logit contributions | |
ley | -0.008 |
rant | -0.007 |
work | -0.007 |
ky | -0.007 |
ci | -0.006 |
Token | kins |
Token ID | 5331 |
Feature activation | -4.06e-03 |
Pos logit contributions | |
bananas | +0.085 |
banana | +0.078 |
cheating | +0.078 |
Ready | +0.077 |
puck | +0.077 |
Neg logit contributions | |
ley | -0.102 |
rant | -0.093 |
work | -0.092 |
ky | -0.086 |
ci | -0.083 |
Token | . |
Token ID | 13 |
Feature activation | +6.74e-03 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.007 |
cheating | +0.007 |
unch | +0.006 |
puck | +0.006 |
Neg logit contributions | |
ley | -0.008 |
rant | -0.007 |
work | -0.007 |
ky | -0.007 |
ci | -0.006 |
Token | They |
Token ID | 1119 |
Feature activation | -1.46e-04 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.012 |
work | +0.011 |
ky | +0.011 |
ci | +0.010 |
Neg logit contributions | |
bananas | -0.013 |
banana | -0.012 |
Ready | -0.011 |
knights | -0.011 |
puck | -0.011 |
Token | see |
Token ID | 766 |
Feature activation | -2.14e-02 |
Pos logit contributions | |
bananas | +0.000 |
banana | +0.000 |
cheating | +0.000 |
unch | +0.000 |
puck | +0.000 |
Neg logit contributions | |
ley | -0.000 |
rant | -0.000 |
work | -0.000 |
ky | -0.000 |
ci | -0.000 |
Token | the |
Token ID | 262 |
Feature activation | -1.87e-03 |
Pos logit contributions | |
bananas | +0.031 |
unch | +0.029 |
cheating | +0.029 |
banana | +0.029 |
puck | +0.028 |
Neg logit contributions | |
ley | -0.046 |
rant | -0.043 |
work | -0.042 |
ky | -0.040 |
ci | -0.037 |
Token | and |
Token ID | 290 |
Feature activation | +8.11e-03 |
Pos logit contributions | |
bananas | +0.024 |
banana | +0.022 |
cheating | +0.022 |
unch | +0.021 |
puck | +0.021 |
Neg logit contributions | |
ley | -0.026 |
rant | -0.024 |
work | -0.023 |
ky | -0.022 |
ci | -0.021 |
Token | they |
Token ID | 484 |
Feature activation | +6.89e-03 |
Pos logit contributions | |
ley | +0.016 |
rant | +0.015 |
work | +0.014 |
ky | +0.014 |
ci | +0.013 |
Neg logit contributions | |
bananas | -0.014 |
banana | -0.012 |
cheating | -0.012 |
unch | -0.012 |
puck | -0.012 |
Token | both |
Token ID | 1111 |
Feature activation | +1.14e-02 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.013 |
work | +0.012 |
ky | +0.012 |
extingu | +0.011 |
Neg logit contributions | |
bananas | -0.011 |
banana | -0.011 |
puck | -0.010 |
he | -0.010 |
catching | -0.010 |
Token | agreed |
Token ID | 4987 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
ley | +0.020 |
rant | +0.019 |
work | +0.018 |
ky | +0.018 |
ci | +0.016 |
Neg logit contributions | |
bananas | -0.021 |
banana | -0.020 |
puck | -0.019 |
knights | -0.018 |
mango | -0.018 |
Token | that |
Token ID | 326 |
Feature activation | -4.23e-03 |
Pos logit contributions | |
ley | +0.025 |
rant | +0.023 |
work | +0.022 |
ky | +0.022 |
ci | +0.020 |
Neg logit contributions | |
bananas | -0.018 |
banana | -0.016 |
cheating | -0.016 |
unch | -0.015 |
catching | -0.015 |
Token | their |
Token ID | 511 |
Feature activation | +3.64e-02 |
Pos logit contributions | |
bananas | +0.006 |
unch | +0.006 |
banana | +0.006 |
puck | +0.006 |
Ready | +0.006 |
Neg logit contributions | |
ley | -0.009 |
work | -0.008 |
ci | -0.008 |
rant | -0.007 |
ky | -0.007 |
Token | sandwiches |
Token ID | 29777 |
Feature activation | +1.72e-03 |
Pos logit contributions | |
ley | +0.084 |
rant | +0.083 |
ky | +0.074 |
work | +0.074 |
any | +0.068 |
Neg logit contributions | |
bananas | -0.053 |
banana | -0.048 |
knights | -0.047 |
cheating | -0.046 |
puck | -0.045 |
Token | were |
Token ID | 547 |
Feature activation | +7.90e-03 |
Pos logit contributions | |
ley | +0.004 |
rant | +0.003 |
ky | +0.003 |
work | +0.003 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
he | -0.002 |
catching | -0.002 |
puck | -0.002 |
Token | very |
Token ID | 845 |
Feature activation | +2.66e-03 |
Pos logit contributions | |
ley | +0.019 |
rant | +0.017 |
work | +0.016 |
ky | +0.016 |
ci | +0.015 |
Neg logit contributions | |
bananas | -0.012 |
banana | -0.010 |
catching | -0.010 |
cheating | -0.010 |
he | -0.009 |
Token | tasty |
Token ID | 25103 |
Feature activation | -1.99e-03 |
Pos logit contributions | |
ley | +0.007 |
rant | +0.006 |
work | +0.006 |
ky | +0.006 |
ci | +0.006 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
unch | -0.003 |
puck | -0.003 |
cheating | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | +9.01e-03 |
Pos logit contributions | |
bananas | +0.003 |
banana | +0.003 |
cheating | +0.003 |
catching | +0.003 |
puck | +0.003 |
Neg logit contributions | |
ley | -0.004 |
rant | -0.004 |
work | -0.004 |
ky | -0.003 |
ci | -0.003 |
Token | were |
Token ID | 547 |
Feature activation | +1.14e-02 |
Pos logit contributions | |
ley | +0.041 |
rant | +0.039 |
work | +0.037 |
extingu | +0.037 |
ky | +0.037 |
Neg logit contributions | |
bananas | -0.025 |
he | -0.024 |
banana | -0.022 |
knights | -0.021 |
beetles | -0.021 |
Token | playing |
Token ID | 2712 |
Feature activation | +6.34e-03 |
Pos logit contributions | |
ley | +0.026 |
rant | +0.024 |
work | +0.023 |
ky | +0.023 |
ci | +0.022 |
Neg logit contributions | |
bananas | -0.016 |
banana | -0.014 |
cheating | -0.014 |
unch | -0.014 |
knights | -0.013 |
Token | at |
Token ID | 379 |
Feature activation | +7.80e-03 |
Pos logit contributions | |
ley | +0.010 |
rant | +0.009 |
work | +0.009 |
ky | +0.009 |
ci | +0.008 |
Neg logit contributions | |
bananas | -0.013 |
banana | -0.012 |
unch | -0.012 |
cheating | -0.012 |
puck | -0.011 |
Token | Sally |
Token ID | 25737 |
Feature activation | +7.66e-03 |
Pos logit contributions | |
ley | +0.016 |
work | +0.015 |
rant | +0.014 |
ky | +0.014 |
ci | +0.013 |
Neg logit contributions | |
bananas | -0.013 |
unch | -0.012 |
banana | -0.011 |
cheating | -0.011 |
puck | -0.011 |
Token | â |
Token ID | 22940 |
Feature activation | +2.85e-02 |
Pos logit contributions | |
ley | +0.016 |
rant | +0.014 |
work | +0.014 |
ky | +0.013 |
ci | +0.013 |
Neg logit contributions | |
bananas | -0.013 |
banana | -0.012 |
cheating | -0.011 |
unch | -0.011 |
puck | -0.011 |
Token | € |
Token ID | 26391 |
Feature activation | +3.84e-02 |
Pos logit contributions | |
ley | +0.057 |
work | +0.051 |
rant | +0.048 |
extingu | +0.045 |
bul | +0.044 |
Neg logit contributions | |
bananas | -0.045 |
unch | -0.044 |
cheating | -0.043 |
catching | -0.042 |
banana | -0.042 |
Token | ™ |
Token ID | 8151 |
Feature activation | +3.04e-04 |
Pos logit contributions | |
ley | +0.076 |
rant | +0.068 |
work | +0.067 |
ky | +0.067 |
ci | +0.065 |
Neg logit contributions | |
bananas | -0.064 |
banana | -0.058 |
cheating | -0.056 |
knights | -0.054 |
puck | -0.053 |
Token | s |
Token ID | 82 |
Feature activation | +2.15e-03 |
Pos logit contributions | |
ley | +0.001 |
rant | +0.001 |
work | +0.001 |
ky | +0.000 |
ci | +0.000 |
Neg logit contributions | |
bananas | -0.001 |
cheating | -0.001 |
banana | -0.001 |
puck | -0.000 |
knights | -0.000 |
Token | house |
Token ID | 2156 |
Feature activation | +2.12e-03 |
Pos logit contributions | |
ley | +0.005 |
rant | +0.004 |
work | +0.004 |
ky | +0.004 |
ci | +0.004 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
cheating | -0.003 |
puck | -0.003 |
unch | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | +1.77e-03 |
Pos logit contributions | |
ley | +0.004 |
rant | +0.004 |
work | +0.003 |
ky | +0.003 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.004 |
banana | -0.004 |
cheating | -0.003 |
knights | -0.003 |
puck | -0.003 |
Token | Sally |
Token ID | 25737 |
Feature activation | -6.25e-03 |
Pos logit contributions | |
ley | +0.004 |
rant | +0.003 |
work | +0.003 |
ky | +0.003 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
cheating | -0.003 |
knights | -0.003 |
unch | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | +1.03e-02 |
Pos logit contributions | |
ley | +0.007 |
rant | +0.007 |
work | +0.006 |
ky | +0.006 |
ci | +0.006 |
Neg logit contributions | |
bananas | -0.004 |
banana | -0.004 |
cheating | -0.004 |
unch | -0.004 |
puck | -0.004 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +4.87e-04 |
Pos logit contributions | |
ley | +0.023 |
rant | +0.021 |
work | +0.021 |
ky | +0.020 |
ci | +0.019 |
Neg logit contributions | |
bananas | -0.015 |
banana | -0.013 |
cheating | -0.013 |
unch | -0.012 |
puck | -0.012 |
Token | One |
Token ID | 3198 |
Feature activation | -3.39e-03 |
Pos logit contributions | |
ley | +0.001 |
rant | +0.001 |
work | +0.001 |
ky | +0.001 |
ci | +0.001 |
Neg logit contributions | |
bananas | -0.001 |
banana | -0.001 |
cheating | -0.001 |
unch | -0.001 |
puck | -0.001 |
Token | morning |
Token ID | 3329 |
Feature activation | +2.63e-02 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.006 |
cheating | +0.006 |
unch | +0.006 |
puck | +0.006 |
Neg logit contributions | |
ley | -0.006 |
rant | -0.005 |
work | -0.005 |
ky | -0.005 |
ci | -0.005 |
Token | , |
Token ID | 11 |
Feature activation | +7.63e-03 |
Pos logit contributions | |
ley | +0.064 |
rant | +0.060 |
work | +0.058 |
ky | +0.056 |
ci | +0.054 |
Neg logit contributions | |
bananas | -0.034 |
banana | -0.030 |
cheating | -0.028 |
unch | -0.028 |
puck | -0.028 |
Token | M |
Token ID | 337 |
Feature activation | +4.65e-02 |
Pos logit contributions | |
ley | +0.018 |
rant | +0.016 |
work | +0.016 |
ky | +0.015 |
ci | +0.015 |
Neg logit contributions | |
bananas | -0.011 |
banana | -0.010 |
cheating | -0.009 |
unch | -0.009 |
puck | -0.009 |
Token | ummy |
Token ID | 13513 |
Feature activation | +1.46e-02 |
Pos logit contributions | |
ley | +0.094 |
rant | +0.094 |
ky | +0.088 |
work | +0.082 |
ples | +0.081 |
Neg logit contributions | |
unch | -0.073 |
bananas | -0.062 |
banana | -0.061 |
puck | -0.053 |
knights | -0.053 |
Token | and |
Token ID | 290 |
Feature activation | +7.43e-03 |
Pos logit contributions | |
rant | +0.033 |
ley | +0.031 |
ky | +0.030 |
work | +0.030 |
iful | +0.029 |
Neg logit contributions | |
he | -0.020 |
banana | -0.019 |
bananas | -0.018 |
chewing | -0.017 |
knights | -0.015 |
Token | Josh |
Token ID | 8518 |
Feature activation | +2.70e-02 |
Pos logit contributions | |
ley | +0.016 |
rant | +0.015 |
work | +0.015 |
ky | +0.014 |
ci | +0.014 |
Neg logit contributions | |
bananas | -0.011 |
he | -0.010 |
banana | -0.010 |
mango | -0.010 |
chewing | -0.009 |
Token | went |
Token ID | 1816 |
Feature activation | +1.02e-02 |
Pos logit contributions | |
ley | +0.065 |
rant | +0.062 |
work | +0.061 |
ky | +0.061 |
ci | +0.060 |
Neg logit contributions | |
bananas | -0.030 |
he | -0.029 |
banana | -0.028 |
knights | -0.024 |
chewing | -0.024 |
Token | to |
Token ID | 284 |
Feature activation | +2.06e-02 |
Pos logit contributions | |
ley | +0.023 |
rant | +0.022 |
work | +0.021 |
ky | +0.020 |
ci | +0.019 |
Neg logit contributions | |
bananas | -0.015 |
banana | -0.013 |
knights | -0.013 |
catching | -0.012 |
he | -0.012 |
Token | wouldn |
Token ID | 3636 |
Feature activation | +4.23e-03 |
Pos logit contributions | |
ley | +0.025 |
rant | +0.023 |
work | +0.022 |
ky | +0.022 |
extingu | +0.020 |
Neg logit contributions | |
bananas | -0.025 |
banana | -0.023 |
cheating | -0.023 |
he | -0.022 |
catching | -0.022 |
Token | 't |
Token ID | 470 |
Feature activation | +5.51e-03 |
Pos logit contributions | |
ley | +0.010 |
rant | +0.010 |
work | +0.009 |
ky | +0.009 |
ci | +0.009 |
Neg logit contributions | |
bananas | -0.005 |
banana | -0.005 |
unch | -0.005 |
cheating | -0.004 |
catching | -0.004 |
Token | forget |
Token ID | 6044 |
Feature activation | +2.73e-03 |
Pos logit contributions | |
ley | +0.012 |
rant | +0.012 |
work | +0.011 |
ky | +0.011 |
ci | +0.010 |
Neg logit contributions | |
bananas | -0.008 |
banana | -0.007 |
cheating | -0.007 |
unch | -0.007 |
puck | -0.007 |
Token | it |
Token ID | 340 |
Feature activation | +6.18e-03 |
Pos logit contributions | |
ley | +0.006 |
rant | +0.006 |
work | +0.006 |
ky | +0.006 |
ci | +0.005 |
Neg logit contributions | |
bananas | -0.004 |
banana | -0.003 |
cheating | -0.003 |
unch | -0.003 |
puck | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | +8.35e-03 |
Pos logit contributions | |
ley | +0.012 |
rant | +0.011 |
work | +0.011 |
ky | +0.010 |
ci | +0.010 |
Neg logit contributions | |
bananas | -0.011 |
banana | -0.010 |
cheating | -0.010 |
catching | -0.009 |
knights | -0.009 |
Token | Jay |
Token ID | 9180 |
Feature activation | +3.87e-02 |
Pos logit contributions | |
ley | +0.018 |
rant | +0.016 |
work | +0.016 |
ky | +0.015 |
ci | +0.014 |
Neg logit contributions | |
bananas | -0.014 |
banana | -0.012 |
cheating | -0.012 |
puck | -0.011 |
unch | -0.011 |
Token | den |
Token ID | 6559 |
Feature activation | +1.41e-02 |
Pos logit contributions | |
extingu | +0.050 |
rant | +0.047 |
icol | +0.043 |
attered | +0.043 |
ley | +0.042 |
Neg logit contributions | |
he | -0.073 |
cheating | -0.065 |
and | -0.065 |
knights | -0.063 |
shadow | -0.063 |
Token | j |
Token ID | 474 |
Feature activation | +2.61e-03 |
Pos logit contributions | |
rant | +0.028 |
ley | +0.028 |
ky | +0.025 |
work | +0.025 |
extingu | +0.025 |
Neg logit contributions | |
bananas | -0.021 |
he | -0.021 |
banana | -0.020 |
knights | -0.019 |
puck | -0.018 |
Token | ogged |
Token ID | 42545 |
Feature activation | +3.07e-03 |
Pos logit contributions | |
ley | +0.006 |
rant | +0.006 |
work | +0.006 |
extingu | +0.005 |
ky | +0.005 |
Neg logit contributions | |
unch | -0.003 |
bananas | -0.003 |
banana | -0.003 |
knights | -0.003 |
puck | -0.003 |
Token | his |
Token ID | 465 |
Feature activation | +9.95e-03 |
Pos logit contributions | |
ley | +0.006 |
rant | +0.006 |
work | +0.005 |
ky | +0.005 |
ci | +0.005 |
Neg logit contributions | |
bananas | -0.005 |
banana | -0.004 |
he | -0.004 |
knights | -0.004 |
puck | -0.004 |
Token | way |
Token ID | 835 |
Feature activation | +7.47e-03 |
Pos logit contributions | |
ley | +0.021 |
rant | +0.020 |
work | +0.019 |
ky | +0.018 |
ci | +0.017 |
Neg logit contributions | |
bananas | -0.016 |
banana | -0.014 |
cheating | -0.014 |
puck | -0.013 |
unch | -0.013 |
Token | friends |
Token ID | 2460 |
Feature activation | -4.35e-03 |
Pos logit contributions | |
bananas | +0.005 |
cheating | +0.004 |
knights | +0.004 |
banana | +0.004 |
unch | +0.003 |
Neg logit contributions | |
ley | -0.013 |
rant | -0.012 |
ky | -0.011 |
work | -0.011 |
ci | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | -4.00e-03 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.007 |
cheating | +0.006 |
unch | +0.006 |
puck | +0.006 |
Neg logit contributions | |
ley | -0.009 |
rant | -0.008 |
work | -0.008 |
ky | -0.008 |
ci | -0.007 |
Token | Can |
Token ID | 1680 |
Feature activation | -9.77e-03 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.004 |
cheating | +0.004 |
unch | +0.004 |
puck | +0.004 |
Neg logit contributions | |
ley | -0.010 |
rant | -0.009 |
work | -0.009 |
ky | -0.009 |
ci | -0.008 |
Token | we |
Token ID | 356 |
Feature activation | +2.85e-03 |
Pos logit contributions | |
bananas | +0.013 |
banana | +0.011 |
unch | +0.011 |
puck | +0.011 |
cheating | +0.011 |
Neg logit contributions | |
ley | -0.022 |
work | -0.021 |
rant | -0.020 |
ci | -0.020 |
ky | -0.019 |
Token | make |
Token ID | 787 |
Feature activation | +8.39e-03 |
Pos logit contributions | |
ley | +0.007 |
rant | +0.007 |
ky | +0.006 |
work | +0.006 |
any | +0.006 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
cheat | -0.003 |
puck | -0.003 |
cheating | -0.003 |
Token | something |
Token ID | 1223 |
Feature activation | +3.51e-02 |
Pos logit contributions | |
ley | +0.016 |
work | +0.016 |
rant | +0.014 |
ky | +0.014 |
ci | +0.014 |
Neg logit contributions | |
bananas | -0.012 |
unch | -0.012 |
banana | -0.012 |
puck | -0.012 |
cheating | -0.011 |
Token | else |
Token ID | 2073 |
Feature activation | +1.76e-02 |
Pos logit contributions | |
ley | +0.079 |
rant | +0.072 |
work | +0.070 |
ky | +0.069 |
ci | +0.065 |
Neg logit contributions | |
bananas | -0.052 |
banana | -0.047 |
cheating | -0.045 |
unch | -0.044 |
puck | -0.043 |
Token | with |
Token ID | 351 |
Feature activation | +1.52e-02 |
Pos logit contributions | |
work | +0.030 |
ley | +0.030 |
rant | +0.028 |
ci | +0.026 |
ky | +0.025 |
Neg logit contributions | |
bananas | -0.031 |
unch | -0.029 |
banana | -0.028 |
Ready | -0.028 |
puck | -0.028 |
Token | the |
Token ID | 262 |
Feature activation | -6.01e-03 |
Pos logit contributions | |
ley | +0.037 |
rant | +0.034 |
work | +0.033 |
ky | +0.032 |
ci | +0.031 |
Neg logit contributions | |
bananas | -0.020 |
banana | -0.018 |
cheating | -0.017 |
puck | -0.016 |
knights | -0.016 |
Token | blocks |
Token ID | 7021 |
Feature activation | +1.52e-02 |
Pos logit contributions | |
bananas | +0.009 |
banana | +0.008 |
cheating | +0.008 |
puck | +0.007 |
unch | +0.007 |
Neg logit contributions | |
ley | -0.013 |
rant | -0.012 |
work | -0.012 |
ky | -0.012 |
ci | -0.011 |
Token | ?" |
Token ID | 1701 |
Feature activation | +7.80e-03 |
Pos logit contributions | |
ley | +0.032 |
rant | +0.028 |
ky | +0.027 |
work | +0.027 |
ci | +0.025 |
Neg logit contributions | |
bananas | -0.025 |
banana | -0.023 |
cheating | -0.023 |
unch | -0.022 |
puck | -0.022 |
Token | more |
Token ID | 517 |
Feature activation | -2.00e-02 |
Pos logit contributions | |
bananas | +0.004 |
unch | +0.004 |
banana | +0.003 |
puck | +0.003 |
mango | +0.003 |
Neg logit contributions | |
ley | -0.007 |
rant | -0.006 |
ky | -0.006 |
work | -0.006 |
sy | -0.006 |
Token | careful |
Token ID | 8161 |
Feature activation | -2.17e-03 |
Pos logit contributions | |
bananas | +0.017 |
banana | +0.014 |
unch | +0.013 |
cheating | +0.013 |
puck | +0.013 |
Neg logit contributions | |
ley | -0.056 |
rant | -0.053 |
work | -0.051 |
ky | -0.051 |
ci | -0.048 |
Token | . |
Token ID | 13 |
Feature activation | +4.40e-03 |
Pos logit contributions | |
bananas | +0.004 |
banana | +0.004 |
cheating | +0.004 |
unch | +0.003 |
puck | +0.003 |
Neg logit contributions | |
ley | -0.004 |
rant | -0.004 |
work | -0.004 |
ky | -0.003 |
ci | -0.003 |
Token | This |
Token ID | 770 |
Feature activation | -5.53e-03 |
Pos logit contributions | |
ley | +0.009 |
rant | +0.009 |
work | +0.008 |
ky | +0.008 |
ci | +0.008 |
Neg logit contributions | |
bananas | -0.007 |
banana | -0.006 |
cheating | -0.006 |
puck | -0.006 |
catching | -0.006 |
Token | calendar |
Token ID | 11845 |
Feature activation | -3.61e-03 |
Pos logit contributions | |
bananas | +0.009 |
banana | +0.008 |
cheating | +0.008 |
unch | +0.008 |
puck | +0.007 |
Neg logit contributions | |
ley | -0.012 |
rant | -0.011 |
work | -0.010 |
ky | -0.010 |
ci | -0.010 |
Token | is |
Token ID | 318 |
Feature activation | +1.49e-02 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.004 |
cheating | +0.004 |
knights | +0.004 |
catching | +0.004 |
Neg logit contributions | |
ley | -0.009 |
rant | -0.008 |
ky | -0.008 |
work | -0.007 |
ci | -0.007 |
Token | very |
Token ID | 845 |
Feature activation | +3.80e-03 |
Pos logit contributions | |
ley | +0.039 |
rant | +0.036 |
work | +0.035 |
ky | +0.035 |
ci | +0.034 |
Neg logit contributions | |
bananas | -0.016 |
banana | -0.014 |
cheating | -0.014 |
catching | -0.013 |
puck | -0.013 |
Token | important |
Token ID | 1593 |
Feature activation | -9.77e-03 |
Pos logit contributions | |
ley | +0.009 |
rant | +0.008 |
work | +0.008 |
ky | +0.007 |
ci | +0.007 |
Neg logit contributions | |
bananas | -0.006 |
banana | -0.005 |
cheating | -0.005 |
catching | -0.005 |
knights | -0.005 |
Token | to |
Token ID | 284 |
Feature activation | -5.40e-03 |
Pos logit contributions | |
bananas | +0.018 |
banana | +0.016 |
cheating | +0.016 |
knights | +0.016 |
puck | +0.015 |
Neg logit contributions | |
ley | -0.018 |
rant | -0.017 |
work | -0.016 |
ky | -0.016 |
ci | -0.015 |
Token | me |
Token ID | 502 |
Feature activation | +2.38e-03 |
Pos logit contributions | |
bananas | +0.009 |
banana | +0.008 |
unch | +0.008 |
cheating | +0.008 |
puck | +0.008 |
Neg logit contributions | |
ley | -0.011 |
rant | -0.010 |
work | -0.010 |
ky | -0.010 |
ci | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | +1.98e-03 |
Pos logit contributions | |
ley | +0.004 |
rant | +0.004 |
work | +0.003 |
ky | +0.003 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.005 |
banana | -0.004 |
cheating | -0.004 |
puck | -0.004 |
unch | -0.004 |
Token | cards |
Token ID | 4116 |
Feature activation | -1.65e-02 |
Pos logit contributions | |
bananas | +0.022 |
banana | +0.020 |
cheating | +0.019 |
knights | +0.019 |
catching | +0.018 |
Neg logit contributions | |
ley | -0.033 |
rant | -0.030 |
ky | -0.028 |
work | -0.028 |
ci | -0.027 |
Token | at |
Token ID | 379 |
Feature activation | -1.57e-02 |
Pos logit contributions | |
bananas | +0.027 |
banana | +0.025 |
cheating | +0.024 |
catching | +0.024 |
he | +0.024 |
Neg logit contributions | |
ley | -0.034 |
rant | -0.031 |
ky | -0.030 |
work | -0.029 |
ci | -0.028 |
Token | the |
Token ID | 262 |
Feature activation | -5.87e-03 |
Pos logit contributions | |
bananas | +0.022 |
banana | +0.020 |
cheating | +0.020 |
puck | +0.019 |
catching | +0.019 |
Neg logit contributions | |
ley | -0.036 |
rant | -0.033 |
work | -0.031 |
ky | -0.031 |
ci | -0.030 |
Token | end |
Token ID | 886 |
Feature activation | +7.99e-03 |
Pos logit contributions | |
bananas | +0.009 |
unch | +0.009 |
banana | +0.008 |
cheating | +0.008 |
puck | +0.008 |
Neg logit contributions | |
ley | -0.012 |
rant | -0.011 |
work | -0.011 |
ky | -0.011 |
ci | -0.010 |
Token | would |
Token ID | 561 |
Feature activation | -4.23e-03 |
Pos logit contributions | |
ley | +0.018 |
rant | +0.017 |
work | +0.016 |
ky | +0.016 |
ci | +0.015 |
Neg logit contributions | |
bananas | -0.012 |
banana | -0.011 |
cheating | -0.010 |
knights | -0.010 |
unch | -0.010 |
Token | win |
Token ID | 1592 |
Feature activation | +5.92e-03 |
Pos logit contributions | |
bananas | +0.006 |
banana | +0.006 |
cheating | +0.005 |
catching | +0.005 |
puck | +0.005 |
Neg logit contributions | |
ley | -0.010 |
rant | -0.009 |
work | -0.009 |
ky | -0.008 |
ci | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | -1.05e-03 |
Pos logit contributions | |
ley | +0.011 |
rant | +0.010 |
work | +0.009 |
ky | +0.009 |
ci | +0.009 |
Neg logit contributions | |
bananas | -0.011 |
banana | -0.010 |
cheating | -0.010 |
puck | -0.010 |
unch | -0.010 |
Token |
|
Token ID | 198 |
Feature activation | -3.33e-04 |
Pos logit contributions | |
bananas | +0.002 |
banana | +0.002 |
cheating | +0.002 |
catching | +0.002 |
puck | +0.002 |
Neg logit contributions | |
ley | -0.002 |
rant | -0.002 |
work | -0.002 |
ky | -0.002 |
ci | -0.002 |
Token |
|
Token ID | 198 |
Feature activation | -4.11e-03 |
Pos logit contributions | |
bananas | +0.001 |
cheating | +0.001 |
banana | +0.001 |
knights | +0.000 |
puck | +0.000 |
Neg logit contributions | |
ley | -0.001 |
rant | -0.001 |
work | -0.001 |
ky | -0.001 |
sy | -0.001 |
Token | Tom |
Token ID | 13787 |
Feature activation | -1.36e-03 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.006 |
cheating | +0.006 |
knights | +0.006 |
mango | +0.006 |
Neg logit contributions | |
ley | -0.008 |
rant | -0.007 |
ky | -0.007 |
sy | -0.007 |
work | -0.007 |
Token | flipped |
Token ID | 26157 |
Feature activation | +8.12e-03 |
Pos logit contributions | |
bananas | +0.002 |
banana | +0.002 |
cheating | +0.002 |
puck | +0.002 |
knights | +0.002 |
Neg logit contributions | |
ley | -0.003 |
rant | -0.003 |
work | -0.002 |
extingu | -0.002 |
ky | -0.002 |
Token | wheels |
Token ID | 13666 |
Feature activation | +1.04e-03 |
Pos logit contributions | |
bananas | +0.009 |
banana | +0.008 |
puck | +0.008 |
cheating | +0.008 |
knights | +0.007 |
Neg logit contributions | |
ley | -0.013 |
rant | -0.012 |
work | -0.011 |
ky | -0.011 |
ci | -0.011 |
Token | were |
Token ID | 547 |
Feature activation | +2.52e-03 |
Pos logit contributions | |
ley | +0.002 |
rant | +0.002 |
ky | +0.002 |
work | +0.002 |
sy | +0.002 |
Neg logit contributions | |
bananas | -0.002 |
he | -0.002 |
banana | -0.002 |
cheating | -0.002 |
puck | -0.001 |
Token | thick |
Token ID | 6546 |
Feature activation | +6.73e-03 |
Pos logit contributions | |
ley | +0.006 |
rant | +0.005 |
work | +0.005 |
ci | +0.005 |
ky | +0.005 |
Neg logit contributions | |
bananas | -0.004 |
cheating | -0.003 |
banana | -0.003 |
catching | -0.003 |
winning | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | -1.26e-03 |
Pos logit contributions | |
ley | +0.015 |
rant | +0.014 |
work | +0.013 |
ci | +0.013 |
sy | +0.012 |
Neg logit contributions | |
bananas | -0.010 |
banana | -0.009 |
he | -0.009 |
knights | -0.009 |
cheating | -0.008 |
Token | He |
Token ID | 679 |
Feature activation | -4.42e-03 |
Pos logit contributions | |
bananas | +0.002 |
cheating | +0.002 |
banana | +0.002 |
unch | +0.002 |
catching | +0.002 |
Neg logit contributions | |
ley | -0.003 |
rant | -0.003 |
work | -0.002 |
ky | -0.002 |
ci | -0.002 |
Token | thought |
Token ID | 1807 |
Feature activation | +1.03e-02 |
Pos logit contributions | |
bananas | +0.008 |
banana | +0.008 |
puck | +0.008 |
cheating | +0.008 |
knights | +0.007 |
Neg logit contributions | |
ley | -0.008 |
rant | -0.007 |
work | -0.007 |
ky | -0.007 |
ci | -0.006 |
Token | to |
Token ID | 284 |
Feature activation | +1.02e-02 |
Pos logit contributions | |
ley | +0.025 |
rant | +0.024 |
work | +0.023 |
ky | +0.022 |
ci | +0.021 |
Neg logit contributions | |
bananas | -0.012 |
banana | -0.011 |
cheating | -0.010 |
unch | -0.010 |
puck | -0.010 |
Token | himself |
Token ID | 2241 |
Feature activation | -4.14e-03 |
Pos logit contributions | |
ley | +0.029 |
rant | +0.028 |
work | +0.027 |
ky | +0.026 |
ci | +0.025 |
Neg logit contributions | |
bananas | -0.008 |
unch | -0.007 |
banana | -0.006 |
cheating | -0.006 |
knights | -0.006 |
Token | , |
Token ID | 11 |
Feature activation | +9.94e-05 |
Pos logit contributions | |
bananas | +0.008 |
banana | +0.008 |
puck | +0.008 |
unch | +0.008 |
moderation | +0.007 |
Neg logit contributions | |
ley | -0.007 |
work | -0.006 |
rant | -0.006 |
ky | -0.006 |
ci | -0.005 |
Token | " |
Token ID | 366 |
Feature activation | +6.18e-03 |
Pos logit contributions | |
ley | +0.000 |
rant | +0.000 |
ky | +0.000 |
work | +0.000 |
ci | +0.000 |
Neg logit contributions | |
bananas | -0.000 |
banana | -0.000 |
unch | -0.000 |
cheating | -0.000 |
puck | -0.000 |
Token | My |
Token ID | 3666 |
Feature activation | -4.20e-03 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.012 |
work | +0.011 |
ky | +0.011 |
ci | +0.010 |
Neg logit contributions | |
bananas | -0.010 |
banana | -0.009 |
cheating | -0.009 |
puck | -0.009 |
unch | -0.008 |
Token | time |
Token ID | 640 |
Feature activation | +1.70e-02 |
Pos logit contributions | |
bananas | +0.002 |
banana | +0.002 |
cheating | +0.002 |
knights | +0.002 |
catching | +0.002 |
Neg logit contributions | |
ley | -0.002 |
rant | -0.002 |
work | -0.002 |
ky | -0.002 |
ci | -0.002 |
Token | , |
Token ID | 11 |
Feature activation | +9.70e-03 |
Pos logit contributions | |
ley | +0.039 |
rant | +0.038 |
ky | +0.035 |
work | +0.035 |
ci | +0.034 |
Neg logit contributions | |
bananas | -0.022 |
he | -0.020 |
knights | -0.020 |
cheating | -0.020 |
banana | -0.019 |
Token | Peter |
Token ID | 5613 |
Feature activation | +3.24e-03 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.011 |
work | +0.010 |
ky | +0.010 |
ci | +0.009 |
Neg logit contributions | |
bananas | -0.023 |
banana | -0.022 |
cheating | -0.021 |
puck | -0.021 |
knights | -0.021 |
Token | and |
Token ID | 290 |
Feature activation | -6.84e-03 |
Pos logit contributions | |
rant | +0.007 |
ley | +0.006 |
work | +0.006 |
extingu | +0.006 |
ky | +0.006 |
Neg logit contributions | |
he | -0.005 |
bananas | -0.004 |
banana | -0.004 |
knights | -0.004 |
puck | -0.004 |
Token | his |
Token ID | 465 |
Feature activation | +9.31e-03 |
Pos logit contributions | |
bananas | +0.006 |
banana | +0.005 |
cheating | +0.005 |
knights | +0.004 |
puck | +0.004 |
Neg logit contributions | |
ley | -0.019 |
rant | -0.018 |
work | -0.018 |
ky | -0.017 |
ci | -0.017 |
Token | bicycle |
Token ID | 17026 |
Feature activation | +9.13e-03 |
Pos logit contributions | |
ley | +0.021 |
rant | +0.020 |
work | +0.019 |
ky | +0.019 |
sy | +0.017 |
Neg logit contributions | |
bananas | -0.013 |
banana | -0.012 |
cheating | -0.011 |
puck | -0.011 |
knights | -0.011 |
Token | appeared |
Token ID | 4120 |
Feature activation | -5.20e-03 |
Pos logit contributions | |
ley | +0.016 |
rant | +0.015 |
work | +0.014 |
ky | +0.013 |
ci | +0.012 |
Neg logit contributions | |
bananas | -0.017 |
banana | -0.016 |
cheating | -0.016 |
puck | -0.016 |
knights | -0.015 |
Token | again |
Token ID | 757 |
Feature activation | +5.26e-03 |
Pos logit contributions | |
bananas | +0.009 |
banana | +0.008 |
cheating | +0.008 |
knights | +0.007 |
puck | +0.007 |
Neg logit contributions | |
ley | -0.011 |
rant | -0.010 |
work | -0.009 |
ky | -0.009 |
ci | -0.009 |
Token | like |
Token ID | 588 |
Feature activation | +1.66e-02 |
Pos logit contributions | |
ley | +0.010 |
rant | +0.010 |
work | +0.009 |
ky | +0.009 |
ci | +0.008 |
Neg logit contributions | |
bananas | -0.009 |
banana | -0.008 |
cheating | -0.008 |
catching | -0.008 |
unch | -0.008 |
Token | nothing |
Token ID | 2147 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
ley | +0.039 |
rant | +0.036 |
work | +0.035 |
ky | +0.034 |
ci | +0.032 |
Neg logit contributions | |
bananas | -0.022 |
banana | -0.020 |
cheating | -0.019 |
unch | -0.019 |
puck | -0.018 |
Token | had |
Token ID | 550 |
Feature activation | -3.54e-03 |
Pos logit contributions | |
ley | +0.030 |
rant | +0.027 |
work | +0.026 |
ky | +0.026 |
ci | +0.025 |
Neg logit contributions | |
bananas | -0.023 |
he | -0.022 |
cheating | -0.021 |
banana | -0.021 |
catching | -0.021 |
Token | They |
Token ID | 1119 |
Feature activation | +3.96e-03 |
Pos logit contributions | |
ley | +0.004 |
rant | +0.004 |
work | +0.004 |
ky | +0.004 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.005 |
banana | -0.004 |
cheating | -0.004 |
unch | -0.004 |
puck | -0.004 |
Token | decided |
Token ID | 3066 |
Feature activation | -3.32e-04 |
Pos logit contributions | |
ley | +0.008 |
rant | +0.008 |
work | +0.008 |
ky | +0.007 |
extingu | +0.007 |
Neg logit contributions | |
bananas | -0.006 |
banana | -0.006 |
he | -0.006 |
puck | -0.005 |
knights | -0.005 |
Token | to |
Token ID | 284 |
Feature activation | +8.59e-03 |
Pos logit contributions | |
bananas | +0.001 |
banana | +0.001 |
unch | +0.001 |
cheating | +0.001 |
puck | +0.001 |
Neg logit contributions | |
ley | -0.001 |
work | -0.000 |
rant | -0.000 |
ky | -0.000 |
ci | -0.000 |
Token | take |
Token ID | 1011 |
Feature activation | +1.40e-03 |
Pos logit contributions | |
ley | +0.021 |
rant | +0.020 |
ky | +0.019 |
work | +0.018 |
ka | +0.017 |
Neg logit contributions | |
bananas | -0.011 |
banana | -0.010 |
puck | -0.009 |
cheat | -0.009 |
cheating | -0.009 |
Token | a |
Token ID | 257 |
Feature activation | +1.64e-02 |
Pos logit contributions | |
ley | +0.003 |
rant | +0.003 |
work | +0.003 |
ky | +0.003 |
ci | +0.002 |
Neg logit contributions | |
bananas | -0.002 |
banana | -0.002 |
cheating | -0.002 |
puck | -0.002 |
unch | -0.002 |
Token | nap |
Token ID | 25422 |
Feature activation | +7.85e-03 |
Pos logit contributions | |
ley | +0.035 |
rant | +0.032 |
work | +0.031 |
ky | +0.031 |
ci | +0.029 |
Neg logit contributions | |
bananas | -0.025 |
banana | -0.023 |
cheating | -0.022 |
unch | -0.022 |
puck | -0.021 |
Token | under |
Token ID | 739 |
Feature activation | +1.38e-02 |
Pos logit contributions | |
ley | +0.016 |
rant | +0.015 |
work | +0.014 |
ky | +0.014 |
ka | +0.013 |
Neg logit contributions | |
bananas | -0.012 |
banana | -0.011 |
knights | -0.011 |
puck | -0.011 |
cheating | -0.011 |
Token | a |
Token ID | 257 |
Feature activation | -7.69e-03 |
Pos logit contributions | |
ley | +0.032 |
rant | +0.029 |
work | +0.028 |
ky | +0.028 |
ci | +0.026 |
Neg logit contributions | |
bananas | -0.020 |
banana | -0.018 |
cheating | -0.017 |
unch | -0.017 |
puck | -0.017 |
Token | big |
Token ID | 1263 |
Feature activation | +1.67e-03 |
Pos logit contributions | |
bananas | +0.012 |
cheating | +0.011 |
unch | +0.010 |
banana | +0.010 |
knights | +0.010 |
Neg logit contributions | |
ley | -0.016 |
rant | -0.015 |
work | -0.014 |
ky | -0.014 |
ci | -0.013 |
Token | tree |
Token ID | 5509 |
Feature activation | +7.33e-03 |
Pos logit contributions | |
ley | +0.004 |
rant | +0.004 |
work | +0.004 |
ky | +0.004 |
ci | +0.004 |
Neg logit contributions | |
bananas | -0.002 |
banana | -0.002 |
cheating | -0.002 |
unch | -0.002 |
puck | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +6.80e-03 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.011 |
work | +0.011 |
ky | +0.011 |
ci | +0.010 |
Neg logit contributions | |
bananas | -0.014 |
banana | -0.013 |
cheating | -0.013 |
unch | -0.013 |
puck | -0.013 |
Token | sprink |
Token ID | 27562 |
Feature activation | -8.02e-03 |
Pos logit contributions | |
ley | +0.017 |
rant | +0.015 |
work | +0.014 |
ky | +0.014 |
ci | +0.013 |
Neg logit contributions | |
bananas | -0.014 |
banana | -0.013 |
cheating | -0.012 |
puck | -0.012 |
catching | -0.012 |
Token | les |
Token ID | 829 |
Feature activation | +7.41e-04 |
Pos logit contributions | |
bananas | +0.015 |
banana | +0.014 |
unch | +0.013 |
puck | +0.013 |
he | +0.012 |
Neg logit contributions | |
ley | -0.013 |
work | -0.012 |
ci | -0.012 |
extingu | -0.012 |
rant | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +3.02e-03 |
Pos logit contributions | |
ley | +0.002 |
rant | +0.001 |
ky | +0.001 |
ci | +0.001 |
work | +0.001 |
Neg logit contributions | |
bananas | -0.001 |
banana | -0.001 |
knights | -0.001 |
he | -0.001 |
catching | -0.001 |
Token | Lily |
Token ID | 20037 |
Feature activation | -9.46e-03 |
Pos logit contributions | |
ley | +0.008 |
rant | +0.007 |
work | +0.007 |
ky | +0.007 |
ci | +0.007 |
Neg logit contributions | |
bananas | -0.003 |
Ready | -0.003 |
banana | -0.003 |
knights | -0.003 |
puck | -0.003 |
Token | was |
Token ID | 373 |
Feature activation | +1.84e-03 |
Pos logit contributions | |
bananas | +0.015 |
banana | +0.014 |
catching | +0.013 |
puck | +0.013 |
knights | +0.013 |
Neg logit contributions | |
ley | -0.020 |
rant | -0.019 |
work | -0.017 |
ky | -0.017 |
extingu | -0.016 |
Token | so |
Token ID | 523 |
Feature activation | -1.01e-02 |
Pos logit contributions | |
ley | +0.004 |
rant | +0.003 |
work | +0.003 |
ky | +0.003 |
ci | +0.003 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
cheating | -0.003 |
puck | -0.003 |
unch | -0.003 |
Token | happy |
Token ID | 3772 |
Feature activation | -8.34e-03 |
Pos logit contributions | |
bananas | +0.010 |
banana | +0.009 |
cheating | +0.008 |
puck | +0.008 |
unch | +0.008 |
Neg logit contributions | |
ley | -0.027 |
rant | -0.026 |
work | -0.025 |
ky | -0.024 |
ci | -0.023 |
Token | to |
Token ID | 284 |
Feature activation | +1.71e-02 |
Pos logit contributions | |
bananas | +0.016 |
banana | +0.014 |
cheating | +0.014 |
puck | +0.014 |
unch | +0.014 |
Neg logit contributions | |
ley | -0.015 |
rant | -0.013 |
work | -0.013 |
ky | -0.013 |
ci | -0.012 |
Token | have |
Token ID | 423 |
Feature activation | +8.66e-03 |
Pos logit contributions | |
ley | +0.048 |
rant | +0.043 |
ky | +0.043 |
work | +0.042 |
ci | +0.040 |
Neg logit contributions | |
bananas | -0.016 |
banana | -0.015 |
catching | -0.013 |
cheating | -0.012 |
unch | -0.012 |
Token | made |
Token ID | 925 |
Feature activation | +1.28e-02 |
Pos logit contributions | |
ley | +0.017 |
rant | +0.016 |
work | +0.015 |
ky | +0.015 |
ci | +0.014 |
Neg logit contributions | |
bananas | -0.015 |
banana | -0.014 |
cheating | -0.013 |
catching | -0.013 |
knights | -0.013 |
Token | a |
Token ID | 257 |
Feature activation | +1.59e-04 |
Pos logit contributions | |
ley | +0.023 |
work | +0.022 |
rant | +0.021 |
ky | +0.020 |
ci | +0.019 |
Neg logit contributions | |
bananas | -0.021 |
puck | -0.020 |
cheating | -0.020 |
unch | -0.019 |
banana | -0.019 |
Token | at |
Token ID | 379 |
Feature activation | -7.91e-03 |
Pos logit contributions | |
bananas | +0.000 |
banana | +0.000 |
knights | +0.000 |
he | +0.000 |
many | +0.000 |
Neg logit contributions | |
ley | -0.000 |
work | -0.000 |
rant | -0.000 |
ci | -0.000 |
ky | -0.000 |
Token | Tom |
Token ID | 4186 |
Feature activation | -3.59e-03 |
Pos logit contributions | |
bananas | +0.010 |
banana | +0.009 |
cheating | +0.009 |
catching | +0.008 |
puck | +0.008 |
Neg logit contributions | |
ley | -0.019 |
rant | -0.018 |
work | -0.017 |
ky | -0.017 |
ci | -0.016 |
Token | and |
Token ID | 290 |
Feature activation | +1.77e-04 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.007 |
cheating | +0.007 |
unch | +0.006 |
puck | +0.006 |
Neg logit contributions | |
ley | -0.006 |
rant | -0.005 |
work | -0.005 |
ky | -0.005 |
ci | -0.005 |
Token | said |
Token ID | 531 |
Feature activation | -1.44e-02 |
Pos logit contributions | |
ley | +0.000 |
rant | +0.000 |
work | +0.000 |
ky | +0.000 |
ci | +0.000 |
Neg logit contributions | |
bananas | -0.000 |
banana | -0.000 |
cheating | -0.000 |
puck | -0.000 |
catching | -0.000 |
Token | , |
Token ID | 11 |
Feature activation | +2.91e-03 |
Pos logit contributions | |
bananas | +0.029 |
puck | +0.027 |
banana | +0.027 |
cheating | +0.026 |
Ready | +0.026 |
Neg logit contributions | |
ley | -0.023 |
rant | -0.020 |
work | -0.020 |
ky | -0.018 |
sy | -0.017 |
Token | " |
Token ID | 366 |
Feature activation | -1.03e-02 |
Pos logit contributions | |
ley | +0.007 |
rant | +0.007 |
work | +0.007 |
ky | +0.006 |
ci | +0.006 |
Neg logit contributions | |
bananas | -0.004 |
banana | -0.003 |
cheating | -0.003 |
puck | -0.003 |
unch | -0.003 |
Token | Hello |
Token ID | 15496 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
bananas | +0.021 |
banana | +0.019 |
unch | +0.019 |
cheating | +0.019 |
catching | +0.018 |
Neg logit contributions | |
ley | -0.017 |
ky | -0.015 |
work | -0.014 |
rant | -0.014 |
ci | -0.014 |
Token | , |
Token ID | 11 |
Feature activation | -7.72e-03 |
Pos logit contributions | |
bananas | +0.024 |
banana | +0.022 |
puck | +0.022 |
unch | +0.022 |
cheating | +0.021 |
Neg logit contributions | |
ley | -0.024 |
rant | -0.022 |
work | -0.021 |
ky | -0.020 |
ci | -0.018 |
Token | Tom |
Token ID | 4186 |
Feature activation | -1.32e-02 |
Pos logit contributions | |
bananas | +0.012 |
unch | +0.011 |
puck | +0.011 |
ulk | +0.011 |
banana | +0.011 |
Neg logit contributions | |
ley | -0.014 |
rant | -0.013 |
work | -0.012 |
ky | -0.012 |
extingu | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | -1.29e-02 |
Pos logit contributions | |
bananas | +0.023 |
banana | +0.022 |
cheating | +0.021 |
knights | +0.020 |
unch | +0.020 |
Neg logit contributions | |
ley | -0.025 |
rant | -0.023 |
work | -0.023 |
ky | -0.021 |
ci | -0.020 |
Token | I |
Token ID | 314 |
Feature activation | -3.87e-04 |
Pos logit contributions | |
bananas | +0.018 |
unch | +0.016 |
banana | +0.016 |
cheating | +0.015 |
puck | +0.015 |
Neg logit contributions | |
ley | -0.030 |
work | -0.028 |
rant | -0.027 |
ci | -0.025 |
ky | -0.025 |
Token |
|
Token ID | 198 |
Feature activation | -1.63e-03 |
Pos logit contributions | |
ley | +0.001 |
rant | +0.001 |
work | +0.001 |
ky | +0.001 |
ci | +0.001 |
Neg logit contributions | |
bananas | -0.001 |
banana | -0.001 |
cheating | -0.001 |
knights | -0.000 |
puck | -0.000 |
Token |
|
Token ID | 198 |
Feature activation | -4.45e-03 |
Pos logit contributions | |
bananas | +0.003 |
cheating | +0.002 |
banana | +0.002 |
knights | +0.002 |
puck | +0.002 |
Neg logit contributions | |
ley | -0.003 |
rant | -0.003 |
work | -0.003 |
ky | -0.003 |
sy | -0.003 |
Token | Mom |
Token ID | 29252 |
Feature activation | -1.79e-03 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.007 |
cheating | +0.007 |
knights | +0.006 |
puck | +0.006 |
Neg logit contributions | |
ley | -0.009 |
rant | -0.008 |
ky | -0.008 |
work | -0.008 |
sy | -0.007 |
Token | my |
Token ID | 1820 |
Feature activation | -1.20e-03 |
Pos logit contributions | |
bananas | +0.002 |
banana | +0.002 |
unch | +0.002 |
cheating | +0.002 |
chewing | +0.002 |
Neg logit contributions | |
ley | -0.004 |
rant | -0.004 |
work | -0.004 |
ky | -0.003 |
ci | -0.003 |
Token | then |
Token ID | 788 |
Feature activation | -1.97e-03 |
Pos logit contributions | |
bananas | +0.002 |
banana | +0.002 |
he | +0.002 |
catching | +0.002 |
chewing | +0.002 |
Neg logit contributions | |
ley | -0.002 |
rant | -0.002 |
work | -0.002 |
ky | -0.002 |
ci | -0.002 |
Token | gave |
Token ID | 2921 |
Feature activation | -2.98e-03 |
Pos logit contributions | |
bananas | +0.004 |
banana | +0.004 |
cheating | +0.003 |
puck | +0.003 |
he | +0.003 |
Neg logit contributions | |
ley | -0.004 |
rant | -0.003 |
work | -0.003 |
ky | -0.003 |
ci | -0.003 |
Token | herself |
Token ID | 5223 |
Feature activation | -2.27e-03 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.004 |
cheating | +0.004 |
catching | +0.004 |
puck | +0.004 |
Neg logit contributions | |
ley | -0.006 |
rant | -0.006 |
work | -0.005 |
ky | -0.005 |
ci | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | +2.15e-03 |
Pos logit contributions | |
bananas | +0.003 |
banana | +0.003 |
puck | +0.002 |
knights | +0.002 |
unch | +0.002 |
Neg logit contributions | |
ley | -0.005 |
rant | -0.005 |
work | -0.005 |
ky | -0.005 |
ci | -0.005 |
Token | scoop |
Token ID | 23529 |
Feature activation | -2.79e-02 |
Pos logit contributions | |
ley | +0.005 |
rant | +0.004 |
work | +0.004 |
ky | +0.004 |
ci | +0.004 |
Neg logit contributions | |
bananas | -0.003 |
banana | -0.003 |
puck | -0.003 |
unch | -0.003 |
cheating | -0.003 |
Token | of |
Token ID | 286 |
Feature activation | +4.43e-03 |
Pos logit contributions | |
bananas | +0.034 |
banana | +0.031 |
unch | +0.028 |
puck | +0.027 |
he | +0.026 |
Neg logit contributions | |
ley | -0.069 |
rant | -0.064 |
work | -0.062 |
ky | -0.062 |
ci | -0.060 |
Token | yogurt |
Token ID | 32132 |
Feature activation | -9.31e-03 |
Pos logit contributions | |
ley | +0.011 |
rant | +0.010 |
work | +0.010 |
ky | +0.010 |
ci | +0.009 |
Neg logit contributions | |
bananas | -0.005 |
banana | -0.005 |
cheating | -0.004 |
puck | -0.004 |
unch | -0.004 |
Token | his |
Token ID | 465 |
Feature activation | +2.80e-03 |
Pos logit contributions | |
ley | +0.009 |
rant | +0.009 |
work | +0.009 |
ky | +0.008 |
ci | +0.008 |
Neg logit contributions | |
bananas | -0.004 |
banana | -0.004 |
cheating | -0.003 |
unch | -0.003 |
puck | -0.003 |
Token | toy |
Token ID | 13373 |
Feature activation | -3.05e-03 |
Pos logit contributions | |
ley | +0.006 |
rant | +0.006 |
ky | +0.006 |
work | +0.005 |
sy | +0.005 |
Neg logit contributions | |
bananas | -0.004 |
banana | -0.004 |
puck | -0.004 |
cheating | -0.004 |
knights | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -3.51e-03 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.004 |
knights | +0.004 |
puck | +0.004 |
cheating | +0.004 |
Neg logit contributions | |
ley | -0.007 |
rant | -0.006 |
ky | -0.006 |
work | -0.006 |
ci | -0.005 |
Token | She |
Token ID | 1375 |
Feature activation | -1.82e-05 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.005 |
cheating | +0.005 |
puck | +0.004 |
knights | +0.004 |
Neg logit contributions | |
ley | -0.008 |
rant | -0.007 |
work | -0.007 |
ky | -0.007 |
ci | -0.006 |
Token | asked |
Token ID | 1965 |
Feature activation | -1.67e-02 |
Pos logit contributions | |
bananas | +0.000 |
banana | +0.000 |
knights | +0.000 |
he | +0.000 |
puck | +0.000 |
Neg logit contributions | |
rant | -0.000 |
work | -0.000 |
ley | -0.000 |
ky | -0.000 |
ifying | -0.000 |
Token | her |
Token ID | 607 |
Feature activation | -6.33e-03 |
Pos logit contributions | |
unch | +0.031 |
puck | +0.030 |
bananas | +0.030 |
banana | +0.029 |
sandbox | +0.029 |
Neg logit contributions | |
ley | -0.030 |
rant | -0.025 |
ky | -0.024 |
work | -0.024 |
ci | -0.021 |
Token | family |
Token ID | 1641 |
Feature activation | -2.12e-02 |
Pos logit contributions | |
bananas | +0.011 |
unch | +0.010 |
banana | +0.010 |
puck | +0.010 |
cheating | +0.009 |
Neg logit contributions | |
ley | -0.013 |
work | -0.011 |
rant | -0.011 |
ky | -0.010 |
ci | -0.010 |
Token | to |
Token ID | 284 |
Feature activation | -4.49e-03 |
Pos logit contributions | |
bananas | +0.037 |
unch | +0.036 |
puck | +0.035 |
banana | +0.034 |
cheating | +0.034 |
Neg logit contributions | |
ley | -0.040 |
work | -0.036 |
rant | -0.035 |
ky | -0.032 |
ci | -0.031 |
Token | help |
Token ID | 1037 |
Feature activation | -5.12e-03 |
Pos logit contributions | |
bananas | +0.006 |
banana | +0.006 |
cheating | +0.005 |
catching | +0.005 |
he | +0.005 |
Neg logit contributions | |
ley | -0.010 |
rant | -0.010 |
ky | -0.009 |
work | -0.009 |
ci | -0.009 |
Token | too |
Token ID | 1165 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.006 |
cheating | +0.006 |
unch | +0.006 |
catching | +0.006 |
Neg logit contributions | |
ley | -0.012 |
rant | -0.011 |
work | -0.011 |
ky | -0.011 |
ci | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | +1.68e-03 |
Pos logit contributions | |
bananas | +0.040 |
banana | +0.038 |
puck | +0.038 |
cheating | +0.037 |
mango | +0.037 |
Neg logit contributions | |
ley | -0.022 |
rant | -0.018 |
work | -0.018 |
ky | -0.017 |
extingu | -0.014 |
Token | rabbit |
Token ID | 22746 |
Feature activation | -1.12e-02 |
Pos logit contributions | |
ley | +0.016 |
rant | +0.015 |
work | +0.014 |
ky | +0.014 |
ci | +0.013 |
Neg logit contributions | |
bananas | -0.015 |
banana | -0.014 |
cheating | -0.013 |
knights | -0.013 |
puck | -0.013 |
Token | was |
Token ID | 373 |
Feature activation | -3.89e-03 |
Pos logit contributions | |
bananas | +0.022 |
banana | +0.020 |
cheating | +0.020 |
unch | +0.020 |
puck | +0.019 |
Neg logit contributions | |
ley | -0.020 |
rant | -0.018 |
work | -0.017 |
ky | -0.016 |
ci | -0.015 |
Token | happy |
Token ID | 3772 |
Feature activation | -6.92e-03 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.005 |
unch | +0.005 |
cheating | +0.005 |
puck | +0.005 |
Neg logit contributions | |
ley | -0.009 |
rant | -0.008 |
work | -0.008 |
ky | -0.008 |
sy | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | -2.60e-03 |
Pos logit contributions | |
bananas | +0.014 |
banana | +0.013 |
cheating | +0.013 |
puck | +0.013 |
unch | +0.013 |
Neg logit contributions | |
ley | -0.011 |
rant | -0.010 |
work | -0.010 |
ky | -0.009 |
ci | -0.008 |
Token | It |
Token ID | 632 |
Feature activation | +1.68e-02 |
Pos logit contributions | |
bananas | +0.004 |
cheating | +0.004 |
banana | +0.004 |
unch | +0.004 |
catching | +0.004 |
Neg logit contributions | |
ley | -0.005 |
rant | -0.005 |
work | -0.004 |
ky | -0.004 |
extingu | -0.004 |
Token | hopped |
Token ID | 45230 |
Feature activation | -3.71e-03 |
Pos logit contributions | |
ley | +0.033 |
rant | +0.030 |
work | +0.030 |
ky | +0.029 |
ci | +0.028 |
Neg logit contributions | |
bananas | -0.030 |
banana | -0.028 |
puck | -0.025 |
knights | -0.025 |
he | -0.025 |
Token | over |
Token ID | 625 |
Feature activation | +4.97e-03 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.007 |
knights | +0.006 |
catching | +0.006 |
puck | +0.006 |
Neg logit contributions | |
ley | -0.007 |
rant | -0.006 |
work | -0.006 |
ci | -0.006 |
ky | -0.006 |
Token | to |
Token ID | 284 |
Feature activation | -7.20e-03 |
Pos logit contributions | |
ley | +0.010 |
rant | +0.010 |
work | +0.009 |
ky | +0.009 |
ci | +0.008 |
Neg logit contributions | |
bananas | -0.008 |
cheating | -0.007 |
unch | -0.007 |
banana | -0.007 |
puck | -0.007 |
Token | the |
Token ID | 262 |
Feature activation | +1.33e-03 |
Pos logit contributions | |
bananas | +0.011 |
unch | +0.010 |
cheating | +0.010 |
banana | +0.010 |
puck | +0.010 |
Neg logit contributions | |
ley | -0.016 |
rant | -0.014 |
work | -0.014 |
ky | -0.013 |
ci | -0.012 |
Token | seeds |
Token ID | 11904 |
Feature activation | -6.98e-03 |
Pos logit contributions | |
ley | +0.003 |
rant | +0.002 |
work | +0.002 |
ky | +0.002 |
ci | +0.002 |
Neg logit contributions | |
bananas | -0.002 |
unch | -0.002 |
banana | -0.002 |
cheating | -0.002 |
puck | -0.002 |
Token | and |
Token ID | 290 |
Feature activation | +3.99e-04 |
Pos logit contributions | |
bananas | +0.013 |
banana | +0.012 |
unch | +0.012 |
cheating | +0.012 |
puck | +0.012 |
Neg logit contributions | |
ley | -0.012 |
rant | -0.011 |
work | -0.011 |
ky | -0.010 |
ci | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | -5.14e-03 |
Pos logit contributions | |
bananas | +0.017 |
banana | +0.016 |
unch | +0.015 |
mango | +0.015 |
puck | +0.015 |
Neg logit contributions | |
ley | -0.015 |
rant | -0.013 |
work | -0.013 |
ky | -0.012 |
ci | -0.012 |
Token | |
Token ID | 220 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
bananas | +0.008 |
cheating | +0.008 |
cheat | +0.008 |
banana | +0.008 |
puck | +0.008 |
Neg logit contributions | |
ley | -0.010 |
rant | -0.009 |
work | -0.008 |
ky | -0.008 |
any | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | -6.07e-03 |
Pos logit contributions | |
bananas | +0.031 |
banana | +0.028 |
cheating | +0.028 |
puck | +0.027 |
mango | +0.026 |
Neg logit contributions | |
ley | -0.031 |
rant | -0.028 |
work | -0.027 |
sy | -0.026 |
ky | -0.026 |
Token |
|
Token ID | 198 |
Feature activation | -7.15e-03 |
Pos logit contributions | |
bananas | +0.011 |
cheating | +0.010 |
banana | +0.009 |
puck | +0.009 |
knights | +0.009 |
Neg logit contributions | |
ley | -0.011 |
rant | -0.011 |
ky | -0.010 |
sy | -0.010 |
work | -0.010 |
Token | The |
Token ID | 464 |
Feature activation | +1.25e-03 |
Pos logit contributions | |
bananas | +0.011 |
cheating | +0.011 |
banana | +0.010 |
knights | +0.010 |
puck | +0.010 |
Neg logit contributions | |
ley | -0.014 |
rant | -0.013 |
sy | -0.013 |
ky | -0.013 |
work | -0.012 |
Token | two |
Token ID | 734 |
Feature activation | -3.16e-02 |
Pos logit contributions | |
ley | +0.003 |
rant | +0.003 |
work | +0.003 |
ky | +0.003 |
ci | +0.002 |
Neg logit contributions | |
bananas | -0.002 |
banana | -0.002 |
cheating | -0.002 |
puck | -0.001 |
knights | -0.001 |
Token | friends |
Token ID | 2460 |
Feature activation | -7.01e-03 |
Pos logit contributions | |
bananas | +0.036 |
banana | +0.032 |
cheating | +0.031 |
unch | +0.030 |
puck | +0.030 |
Neg logit contributions | |
ley | -0.079 |
rant | -0.071 |
work | -0.070 |
ky | -0.069 |
ci | -0.065 |
Token | marched |
Token ID | 23558 |
Feature activation | +9.93e-03 |
Pos logit contributions | |
bananas | +0.012 |
banana | +0.011 |
he | +0.011 |
knights | +0.011 |
puck | +0.011 |
Neg logit contributions | |
ley | -0.013 |
rant | -0.013 |
work | -0.012 |
ky | -0.011 |
ci | -0.011 |
Token | around |
Token ID | 1088 |
Feature activation | +6.91e-03 |
Pos logit contributions | |
ley | +0.019 |
rant | +0.018 |
work | +0.016 |
ky | +0.016 |
ci | +0.015 |
Neg logit contributions | |
bananas | -0.017 |
puck | -0.016 |
knights | -0.016 |
he | -0.015 |
banana | -0.015 |
Token | the |
Token ID | 262 |
Feature activation | -4.17e-03 |
Pos logit contributions | |
ley | +0.016 |
rant | +0.015 |
work | +0.014 |
ky | +0.014 |
ci | +0.013 |
Neg logit contributions | |
bananas | -0.010 |
banana | -0.009 |
cheating | -0.008 |
unch | -0.008 |
puck | -0.008 |
Token | garden |
Token ID | 11376 |
Feature activation | -5.06e-03 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.007 |
cheating | +0.006 |
unch | +0.006 |
puck | +0.006 |
Neg logit contributions | |
ley | -0.008 |
rant | -0.007 |
work | -0.007 |
ky | -0.007 |
ci | -0.007 |
Token | are |
Token ID | 389 |
Feature activation | -1.99e-02 |
Pos logit contributions | |
bananas | +0.002 |
unch | +0.002 |
mango | +0.002 |
puck | +0.002 |
banana | +0.001 |
Neg logit contributions | |
work | -0.002 |
ley | -0.002 |
rant | -0.001 |
printed | -0.001 |
ci | -0.001 |
Token | you |
Token ID | 345 |
Feature activation | -3.99e-03 |
Pos logit contributions | |
bananas | +0.025 |
unch | +0.025 |
banana | +0.025 |
mango | +0.023 |
puck | +0.022 |
Neg logit contributions | |
ley | -0.044 |
work | -0.041 |
rant | -0.040 |
ky | -0.038 |
ci | -0.037 |
Token | doing |
Token ID | 1804 |
Feature activation | +1.23e-03 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.004 |
puck | +0.004 |
mango | +0.004 |
unch | +0.004 |
Neg logit contributions | |
ley | -0.009 |
rant | -0.009 |
work | -0.009 |
ky | -0.009 |
ci | -0.008 |
Token | ? |
Token ID | 30 |
Feature activation | -1.38e-02 |
Pos logit contributions | |
ley | +0.003 |
rant | +0.003 |
ky | +0.002 |
work | +0.002 |
sy | +0.002 |
Neg logit contributions | |
cheating | -0.002 |
bananas | -0.002 |
catching | -0.002 |
banana | -0.002 |
stealing | -0.002 |
Token | Why |
Token ID | 4162 |
Feature activation | +1.23e-02 |
Pos logit contributions | |
bananas | +0.019 |
banana | +0.017 |
cheating | +0.016 |
unch | +0.016 |
puck | +0.015 |
Neg logit contributions | |
ley | -0.032 |
rant | -0.030 |
work | -0.029 |
ky | -0.028 |
ci | -0.027 |
Token | are |
Token ID | 389 |
Feature activation | -2.95e-02 |
Pos logit contributions | |
ley | +0.024 |
work | +0.023 |
rant | +0.022 |
ky | +0.020 |
ci | +0.020 |
Neg logit contributions | |
bananas | -0.020 |
unch | -0.018 |
banana | -0.018 |
mango | -0.018 |
puck | -0.017 |
Token | you |
Token ID | 345 |
Feature activation | -3.81e-03 |
Pos logit contributions | |
unch | +0.040 |
bananas | +0.037 |
banana | +0.036 |
puck | +0.035 |
mango | +0.035 |
Neg logit contributions | |
ley | -0.062 |
work | -0.059 |
rant | -0.055 |
ky | -0.053 |
ci | -0.052 |
Token | fighting |
Token ID | 4330 |
Feature activation | +8.83e-03 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.005 |
mango | +0.005 |
puck | +0.004 |
unch | +0.004 |
Neg logit contributions | |
ley | -0.009 |
rant | -0.008 |
work | -0.008 |
ky | -0.008 |
ci | -0.007 |
Token | ?" |
Token ID | 1701 |
Feature activation | +1.66e-02 |
Pos logit contributions | |
ley | +0.016 |
rant | +0.014 |
ky | +0.013 |
work | +0.013 |
ci | +0.013 |
Neg logit contributions | |
bananas | -0.017 |
cheating | -0.016 |
banana | -0.016 |
catching | -0.015 |
puck | -0.015 |
Token | Mom |
Token ID | 11254 |
Feature activation | +2.63e-02 |
Pos logit contributions | |
ley | +0.029 |
rant | +0.025 |
work | +0.025 |
ky | +0.024 |
ci | +0.022 |
Neg logit contributions | |
bananas | -0.032 |
banana | -0.029 |
puck | -0.029 |
unch | -0.028 |
cheating | -0.028 |
Token | asks |
Token ID | 7893 |
Feature activation | -7.63e-03 |
Pos logit contributions | |
ley | +0.061 |
work | +0.056 |
rant | +0.056 |
ky | +0.052 |
ci | +0.052 |
Neg logit contributions | |
bananas | -0.035 |
cheating | -0.034 |
banana | -0.031 |
knights | -0.029 |
Ready | -0.029 |
Token | was |
Token ID | 373 |
Feature activation | -2.05e-03 |
Pos logit contributions | |
bananas | +0.013 |
banana | +0.012 |
cheating | +0.011 |
puck | +0.011 |
unch | +0.011 |
Neg logit contributions | |
ley | -0.012 |
rant | -0.011 |
work | -0.011 |
ky | -0.010 |
ci | -0.009 |
Token | amazed |
Token ID | 24562 |
Feature activation | -5.09e-03 |
Pos logit contributions | |
bananas | +0.004 |
banana | +0.003 |
unch | +0.003 |
cheating | +0.003 |
puck | +0.003 |
Neg logit contributions | |
ley | -0.004 |
rant | -0.004 |
work | -0.003 |
ky | -0.003 |
ci | -0.003 |
Token | by |
Token ID | 416 |
Feature activation | +1.67e-02 |
Pos logit contributions | |
bananas | +0.007 |
he | +0.007 |
banana | +0.006 |
cheating | +0.006 |
catching | +0.006 |
Neg logit contributions | |
ley | -0.012 |
rant | -0.011 |
work | -0.010 |
sy | -0.010 |
ci | -0.010 |
Token | how |
Token ID | 703 |
Feature activation | +1.81e-02 |
Pos logit contributions | |
ley | +0.032 |
work | +0.030 |
rant | +0.029 |
ky | +0.027 |
ci | +0.026 |
Neg logit contributions | |
bananas | -0.026 |
unch | -0.026 |
puck | -0.025 |
banana | -0.025 |
Ready | -0.024 |
Token | it |
Token ID | 340 |
Feature activation | +4.58e-03 |
Pos logit contributions | |
ley | +0.037 |
rant | +0.034 |
work | +0.033 |
ky | +0.032 |
ci | +0.031 |
Neg logit contributions | |
bananas | -0.028 |
banana | -0.026 |
cheating | -0.026 |
unch | -0.025 |
puck | -0.025 |
Token | landed |
Token ID | 11406 |
Feature activation | -3.18e-02 |
Pos logit contributions | |
ley | +0.008 |
rant | +0.008 |
work | +0.008 |
ky | +0.007 |
extingu | +0.007 |
Neg logit contributions | |
bananas | -0.008 |
banana | -0.008 |
he | -0.007 |
puck | -0.007 |
catching | -0.007 |
Token | so |
Token ID | 523 |
Feature activation | -3.32e-03 |
Pos logit contributions | |
bananas | +0.063 |
banana | +0.059 |
cheating | +0.057 |
unch | +0.056 |
puck | +0.056 |
Neg logit contributions | |
ley | -0.054 |
rant | -0.048 |
work | -0.046 |
ky | -0.045 |
ci | -0.042 |
Token | smoothly |
Token ID | 21461 |
Feature activation | +7.25e-03 |
Pos logit contributions | |
bananas | +0.005 |
banana | +0.005 |
cheating | +0.005 |
unch | +0.005 |
puck | +0.005 |
Neg logit contributions | |
ley | -0.007 |
rant | -0.006 |
work | -0.006 |
ky | -0.006 |
ci | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | -3.10e-03 |
Pos logit contributions | |
ley | +0.016 |
rant | +0.015 |
work | +0.014 |
ky | +0.013 |
ci | +0.013 |
Neg logit contributions | |
bananas | -0.011 |
cheating | -0.010 |
banana | -0.010 |
he | -0.009 |
catching | -0.009 |
Token | From |
Token ID | 3574 |
Feature activation | +5.60e-03 |
Pos logit contributions | |
bananas | +0.005 |
cheating | +0.005 |
banana | +0.005 |
cheat | +0.005 |
liar | +0.005 |
Neg logit contributions | |
ley | -0.006 |
rant | -0.005 |
ky | -0.005 |
work | -0.005 |
any | -0.004 |
Token | that |
Token ID | 326 |
Feature activation | -9.96e-03 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.012 |
work | +0.011 |
extingu | +0.011 |
ky | +0.011 |
Neg logit contributions | |
bananas | -0.008 |
he | -0.007 |
catching | -0.007 |
banana | -0.007 |
cheating | -0.007 |
Token | and |
Token ID | 290 |
Feature activation | +1.05e-04 |
Pos logit contributions | |
bananas | +0.015 |
banana | +0.014 |
cheating | +0.013 |
unch | +0.013 |
puck | +0.013 |
Neg logit contributions | |
ley | -0.014 |
rant | -0.013 |
work | -0.012 |
ky | -0.012 |
ci | -0.011 |
Token | looked |
Token ID | 3114 |
Feature activation | -3.94e-03 |
Pos logit contributions | |
ley | +0.000 |
rant | +0.000 |
work | +0.000 |
ky | +0.000 |
ci | +0.000 |
Neg logit contributions | |
bananas | -0.000 |
banana | -0.000 |
cheating | -0.000 |
puck | -0.000 |
he | -0.000 |
Token | around |
Token ID | 1088 |
Feature activation | +6.47e-03 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.006 |
he | +0.006 |
knights | +0.006 |
catching | +0.006 |
Neg logit contributions | |
ley | -0.008 |
work | -0.007 |
rant | -0.007 |
ci | -0.007 |
ky | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | -2.16e-03 |
Pos logit contributions | |
ley | +0.013 |
rant | +0.012 |
work | +0.011 |
ky | +0.011 |
ci | +0.011 |
Neg logit contributions | |
bananas | -0.011 |
banana | -0.010 |
cheating | -0.010 |
catching | -0.010 |
knights | -0.010 |
Token | There |
Token ID | 1318 |
Feature activation | -1.54e-02 |
Pos logit contributions | |
bananas | +0.003 |
banana | +0.003 |
cheating | +0.003 |
unch | +0.003 |
puck | +0.003 |
Neg logit contributions | |
ley | -0.004 |
rant | -0.004 |
work | -0.004 |
ky | -0.004 |
ci | -0.004 |
Token | were |
Token ID | 547 |
Feature activation | -2.98e-02 |
Pos logit contributions | |
bananas | +0.031 |
banana | +0.028 |
puck | +0.028 |
cheating | +0.028 |
unch | +0.027 |
Neg logit contributions | |
ley | -0.027 |
rant | -0.023 |
work | -0.023 |
ky | -0.021 |
ci | -0.021 |
Token | lots |
Token ID | 6041 |
Feature activation | -8.55e-03 |
Pos logit contributions | |
bananas | +0.037 |
unch | +0.034 |
banana | +0.034 |
cheating | +0.033 |
puck | +0.031 |
Neg logit contributions | |
ley | -0.069 |
rant | -0.064 |
work | -0.063 |
ky | -0.062 |
ci | -0.058 |
Token | of |
Token ID | 286 |
Feature activation | -1.66e-02 |
Pos logit contributions | |
bananas | +0.011 |
banana | +0.010 |
cheating | +0.010 |
unch | +0.010 |
puck | +0.010 |
Neg logit contributions | |
ley | -0.020 |
rant | -0.019 |
work | -0.018 |
ky | -0.018 |
ci | -0.017 |
Token | buildings |
Token ID | 6832 |
Feature activation | -1.35e-02 |
Pos logit contributions | |
bananas | +0.023 |
banana | +0.020 |
cheating | +0.019 |
knights | +0.018 |
puck | +0.018 |
Neg logit contributions | |
ley | -0.040 |
rant | -0.037 |
work | -0.036 |
ky | -0.034 |
ci | -0.033 |
Token | , |
Token ID | 11 |
Feature activation | +3.65e-03 |
Pos logit contributions | |
bananas | +0.025 |
banana | +0.023 |
cheating | +0.023 |
unch | +0.022 |
catching | +0.022 |
Neg logit contributions | |
ley | -0.025 |
rant | -0.022 |
work | -0.021 |
ky | -0.021 |
ci | -0.019 |
Token | people |
Token ID | 661 |
Feature activation | -8.37e-03 |
Pos logit contributions | |
ley | +0.008 |
rant | +0.007 |
work | +0.007 |
ky | +0.007 |
ci | +0.006 |
Neg logit contributions | |
bananas | -0.005 |
unch | -0.005 |
banana | -0.005 |
cheating | -0.005 |
puck | -0.005 |
Token | He |
Token ID | 679 |
Feature activation | -1.81e-03 |
Pos logit contributions | |
bananas | +0.012 |
cheating | +0.012 |
beetles | +0.012 |
unch | +0.012 |
knights | +0.012 |
Neg logit contributions | |
ley | -0.012 |
rant | -0.011 |
ky | -0.011 |
work | -0.011 |
j | -0.010 |
Token | scoop |
Token ID | 23529 |
Feature activation | -2.24e-02 |
Pos logit contributions | |
bananas | +0.003 |
banana | +0.003 |
puck | +0.003 |
catching | +0.003 |
cheating | +0.003 |
Neg logit contributions | |
ley | -0.003 |
rant | -0.003 |
work | -0.003 |
ky | -0.003 |
extingu | -0.003 |
Token | ed |
Token ID | 276 |
Feature activation | -3.74e-03 |
Pos logit contributions | |
bananas | +0.055 |
cheating | +0.053 |
unch | +0.052 |
banana | +0.052 |
catching | +0.051 |
Neg logit contributions | |
ley | -0.027 |
rant | -0.023 |
work | -0.020 |
ky | -0.020 |
sy | -0.017 |
Token | up |
Token ID | 510 |
Feature activation | +1.01e-03 |
Pos logit contributions | |
bananas | +0.007 |
banana | +0.007 |
cheating | +0.007 |
unch | +0.006 |
puck | +0.006 |
Neg logit contributions | |
ley | -0.007 |
rant | -0.006 |
work | -0.006 |
ky | -0.005 |
ci | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -2.66e-03 |
Pos logit contributions | |
ley | +0.002 |
rant | +0.002 |
work | +0.002 |
ky | +0.002 |
ci | +0.002 |
Neg logit contributions | |
bananas | -0.001 |
banana | -0.001 |
unch | -0.001 |
cheating | -0.001 |
puck | -0.001 |
Token | golf |
Token ID | 13126 |
Feature activation | -3.46e-02 |
Pos logit contributions | |
bananas | +0.004 |
banana | +0.003 |
cheating | +0.003 |
unch | +0.003 |
puck | +0.003 |
Neg logit contributions | |
ley | -0.006 |
rant | -0.006 |
work | -0.006 |
ky | -0.005 |
ci | -0.005 |
Token | clubs |
Token ID | 9784 |
Feature activation | -1.87e-02 |
Pos logit contributions | |
bananas | +0.051 |
unch | +0.047 |
cheating | +0.047 |
banana | +0.046 |
moderation | +0.044 |
Neg logit contributions | |
ley | -0.073 |
work | -0.066 |
rant | -0.066 |
ky | -0.064 |
ci | -0.060 |
Token | and |
Token ID | 290 |
Feature activation | -1.36e-03 |
Pos logit contributions | |
bananas | +0.035 |
banana | +0.032 |
puck | +0.031 |
unch | +0.030 |
cheating | +0.030 |
Neg logit contributions | |
ley | -0.035 |
rant | -0.031 |
work | -0.030 |
ky | -0.029 |
ci | -0.027 |
Token | balls |
Token ID | 11333 |
Feature activation | -1.36e-02 |
Pos logit contributions | |
bananas | +0.002 |
banana | +0.002 |
unch | +0.002 |
cheating | +0.002 |
puck | +0.002 |
Neg logit contributions | |
ley | -0.003 |
rant | -0.002 |
work | -0.002 |
ky | -0.002 |
ci | -0.002 |
Token | , |
Token ID | 11 |
Feature activation | -3.26e-03 |
Pos logit contributions | |
bananas | +0.024 |
banana | +0.022 |
puck | +0.022 |
unch | +0.021 |
cheating | +0.021 |
Neg logit contributions | |
ley | -0.026 |
rant | -0.024 |
work | -0.022 |
ky | -0.022 |
ci | -0.021 |
Token | and |
Token ID | 290 |
Feature activation | -1.73e-02 |
Pos logit contributions | |
bananas | +0.006 |
unch | +0.006 |
banana | +0.005 |
cheating | +0.005 |
mango | +0.005 |
Neg logit contributions | |
ley | -0.006 |
rant | -0.005 |
ky | -0.005 |
work | -0.005 |
ci | -0.005 |