Token | smiled |
Token ID | 13541 |
Feature activation | +1.17e-02 |
Pos logit contributions | |
happens | +0.005 |
happened | +0.005 |
ield | +0.005 |
Dan | +0.005 |
meant | +0.004 |
Neg logit contributions | |
Brown | -0.005 |
Buzz | -0.005 |
seek | -0.005 |
| | -0.005 |
Later | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | +2.22e-02 |
Pos logit contributions | |
happens | +0.026 |
ield | +0.025 |
happened | +0.024 |
Dan | +0.023 |
oke | +0.022 |
Neg logit contributions | |
Brown | -0.015 |
Buzz | -0.014 |
Later | -0.014 |
Ruby | -0.013 |
seek | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -3.47e-02 |
Pos logit contributions | |
happens | +0.044 |
ield | +0.044 |
happened | +0.043 |
meant | +0.038 |
Dan | +0.038 |
Neg logit contributions | |
Brown | -0.032 |
Later | -0.031 |
Buzz | -0.031 |
Ruby | -0.031 |
Fl | -0.029 |
Token |
|
Token ID | 198 |
Feature activation | -4.66e-02 |
Pos logit contributions | |
Brown | +0.078 |
seek | +0.073 |
throne | +0.073 |
Buzz | +0.072 |
heads | +0.071 |
Neg logit contributions | |
happens | -0.043 |
happened | -0.040 |
Dan | -0.038 |
ield | -0.037 |
It | -0.036 |
Token | " |
Token ID | 1 |
Feature activation | +1.17e-02 |
Pos logit contributions | |
throne | +0.106 |
hind | +0.098 |
sleep | +0.096 |
| | +0.095 |
cushion | +0.094 |
Neg logit contributions | |
happens | -0.054 |
ield | -0.048 |
It | -0.047 |
Dan | -0.044 |
happened | -0.043 |
Token | Looks |
Token ID | 41102 |
Feature activation | +7.93e-02 |
Pos logit contributions | |
happens | +0.021 |
ield | +0.020 |
happened | +0.020 |
Dan | +0.019 |
meant | +0.018 |
Neg logit contributions | |
Brown | -0.020 |
Buzz | -0.018 |
| | -0.018 |
seek | -0.018 |
Ruby | -0.017 |
Token | like |
Token ID | 588 |
Feature activation | +1.92e-02 |
Pos logit contributions | |
ield | +0.144 |
happened | +0.137 |
Dan | +0.134 |
Matt | +0.132 |
David | +0.130 |
Neg logit contributions | |
Buzz | -0.118 |
sleep | -0.115 |
Blue | -0.110 |
Tweet | -0.109 |
cozy | -0.108 |
Token | the |
Token ID | 262 |
Feature activation | +1.54e-02 |
Pos logit contributions | |
happens | +0.040 |
ield | +0.039 |
happened | +0.038 |
Dan | +0.036 |
meant | +0.034 |
Neg logit contributions | |
Brown | -0.026 |
Buzz | -0.025 |
seek | -0.024 |
Ruby | -0.024 |
Fl | -0.023 |
Token | frog |
Token ID | 21264 |
Feature activation | +5.31e-03 |
Pos logit contributions | |
happens | +0.034 |
happened | +0.033 |
ield | +0.032 |
Dan | +0.031 |
meant | +0.030 |
Neg logit contributions | |
Brown | -0.021 |
Buzz | -0.018 |
Later | -0.018 |
| | -0.017 |
heads | -0.017 |
Token | likes |
Token ID | 7832 |
Feature activation | +9.04e-03 |
Pos logit contributions | |
happens | +0.008 |
happened | +0.008 |
ield | +0.008 |
Dan | +0.007 |
meant | +0.007 |
Neg logit contributions | |
Brown | -0.010 |
Buzz | -0.010 |
Later | -0.009 |
Ruby | -0.009 |
| | -0.009 |
Token | you |
Token ID | 345 |
Feature activation | -6.15e-03 |
Pos logit contributions | |
happens | +0.017 |
ield | +0.016 |
happened | +0.016 |
Dan | +0.015 |
meant | +0.014 |
Neg logit contributions | |
Brown | -0.015 |
Buzz | -0.014 |
seek | -0.013 |
Ruby | -0.013 |
| | -0.013 |
Token | laughed |
Token ID | 13818 |
Feature activation | +2.01e-02 |
Pos logit contributions | |
happens | +0.010 |
ield | +0.009 |
happened | +0.009 |
Dan | +0.009 |
meant | +0.008 |
Neg logit contributions | |
Brown | -0.012 |
Buzz | -0.012 |
seek | -0.011 |
Ruby | -0.011 |
| | -0.011 |
Token | and |
Token ID | 290 |
Feature activation | -1.48e-02 |
Pos logit contributions | |
happens | +0.039 |
ield | +0.038 |
happened | +0.037 |
Dan | +0.035 |
David | +0.033 |
Neg logit contributions | |
Brown | -0.031 |
Buzz | -0.029 |
Later | -0.028 |
Ruby | -0.028 |
seek | -0.028 |
Token | said |
Token ID | 531 |
Feature activation | +4.39e-02 |
Pos logit contributions | |
Brown | +0.027 |
Buzz | +0.026 |
seek | +0.025 |
Ruby | +0.025 |
Fl | +0.025 |
Neg logit contributions | |
happens | -0.024 |
ield | -0.024 |
happened | -0.023 |
Dan | -0.021 |
meant | -0.020 |
Token | , |
Token ID | 11 |
Feature activation | -5.05e-03 |
Pos logit contributions | |
ield | +0.078 |
happens | +0.071 |
happened | +0.068 |
Dan | +0.068 |
meant | +0.067 |
Neg logit contributions | |
Buzz | -0.069 |
Brown | -0.068 |
Fl | -0.066 |
Ruby | -0.064 |
Bun | -0.064 |
Token | " |
Token ID | 366 |
Feature activation | +9.35e-03 |
Pos logit contributions | |
Brown | +0.011 |
Buzz | +0.011 |
Later | +0.011 |
Fl | +0.011 |
Ruby | +0.010 |
Neg logit contributions | |
happens | -0.007 |
ield | -0.006 |
happened | -0.006 |
Dan | -0.005 |
meant | -0.005 |
Token | Looks |
Token ID | 41102 |
Feature activation | +7.92e-02 |
Pos logit contributions | |
happens | +0.017 |
happened | +0.016 |
ield | +0.016 |
Dan | +0.015 |
meant | +0.014 |
Neg logit contributions | |
Brown | -0.016 |
Buzz | -0.015 |
lf | -0.014 |
throne | -0.014 |
seek | -0.014 |
Token | like |
Token ID | 588 |
Feature activation | +1.17e-02 |
Pos logit contributions | |
ield | +0.143 |
happened | +0.131 |
Dan | +0.130 |
happens | +0.128 |
David | +0.125 |
Neg logit contributions | |
Buzz | -0.124 |
sleep | -0.122 |
Blue | -0.116 |
Tweet | -0.115 |
Ruby | -0.115 |
Token | that |
Token ID | 326 |
Feature activation | +4.54e-02 |
Pos logit contributions | |
happens | +0.023 |
happened | +0.022 |
ield | +0.021 |
Dan | +0.020 |
meant | +0.019 |
Neg logit contributions | |
Brown | -0.019 |
Buzz | -0.017 |
| | -0.017 |
Later | -0.017 |
Ruby | -0.017 |
Token | apple |
Token ID | 17180 |
Feature activation | +2.42e-02 |
Pos logit contributions | |
happens | +0.081 |
ield | +0.078 |
happened | +0.077 |
Dan | +0.071 |
meant | +0.068 |
Neg logit contributions | |
Brown | -0.078 |
Buzz | -0.074 |
seek | -0.070 |
Ruby | -0.070 |
| | -0.070 |
Token | made |
Token ID | 925 |
Feature activation | +1.21e-02 |
Pos logit contributions | |
happens | +0.039 |
ield | +0.038 |
happened | +0.037 |
Dan | +0.034 |
meant | +0.032 |
Neg logit contributions | |
Brown | -0.045 |
Buzz | -0.043 |
seek | -0.041 |
Ruby | -0.041 |
Fl | -0.040 |
Token | you |
Token ID | 345 |
Feature activation | -2.86e-03 |
Pos logit contributions | |
happens | +0.024 |
ield | +0.023 |
happened | +0.023 |
Dan | +0.021 |
meant | +0.021 |
Neg logit contributions | |
Brown | -0.019 |
Buzz | -0.017 |
seek | -0.017 |
| | -0.016 |
heads | -0.016 |
Token | smiled |
Token ID | 13541 |
Feature activation | +1.41e-02 |
Pos logit contributions | |
happens | +0.010 |
ield | +0.010 |
happened | +0.010 |
Dan | +0.009 |
meant | +0.008 |
Neg logit contributions | |
Brown | -0.010 |
Buzz | -0.010 |
Ruby | -0.009 |
seek | -0.009 |
Fl | -0.009 |
Token | and |
Token ID | 290 |
Feature activation | -1.13e-02 |
Pos logit contributions | |
happens | +0.030 |
ield | +0.029 |
happened | +0.028 |
Dan | +0.027 |
oke | +0.026 |
Neg logit contributions | |
Brown | -0.019 |
Buzz | -0.018 |
Later | -0.018 |
seek | -0.017 |
Ruby | -0.017 |
Token | said |
Token ID | 531 |
Feature activation | +3.41e-02 |
Pos logit contributions | |
Brown | +0.021 |
Buzz | +0.020 |
seek | +0.019 |
Ruby | +0.019 |
Fl | +0.019 |
Neg logit contributions | |
happens | -0.018 |
ield | -0.018 |
happened | -0.017 |
Dan | -0.015 |
fault | -0.015 |
Token | , |
Token ID | 11 |
Feature activation | -6.45e-03 |
Pos logit contributions | |
ield | +0.062 |
happens | +0.061 |
Dan | +0.059 |
meant | +0.058 |
happened | +0.058 |
Neg logit contributions | |
Buzz | -0.046 |
ide | -0.044 |
Brown | -0.044 |
Fl | -0.043 |
Ruby | -0.043 |
Token | " |
Token ID | 366 |
Feature activation | +1.39e-02 |
Pos logit contributions | |
Brown | +0.015 |
Buzz | +0.014 |
Later | +0.014 |
Fl | +0.014 |
seek | +0.014 |
Neg logit contributions | |
happens | -0.008 |
ield | -0.007 |
happened | -0.007 |
Dan | -0.006 |
meant | -0.006 |
Token | Looks |
Token ID | 41102 |
Feature activation | +7.89e-02 |
Pos logit contributions | |
happens | +0.024 |
ield | +0.023 |
happened | +0.023 |
Dan | +0.021 |
meant | +0.020 |
Neg logit contributions | |
Brown | -0.025 |
Buzz | -0.023 |
seek | -0.022 |
Ruby | -0.022 |
| | -0.022 |
Token | like |
Token ID | 588 |
Feature activation | +1.52e-02 |
Pos logit contributions | |
ield | +0.138 |
happened | +0.131 |
happens | +0.126 |
Dan | +0.126 |
David | +0.122 |
Neg logit contributions | |
Buzz | -0.129 |
sleep | -0.123 |
Brown | -0.118 |
Blue | -0.117 |
games | -0.117 |
Token | you |
Token ID | 345 |
Feature activation | +7.72e-03 |
Pos logit contributions | |
happens | +0.029 |
ield | +0.028 |
happened | +0.028 |
Dan | +0.026 |
meant | +0.024 |
Neg logit contributions | |
Brown | -0.025 |
Buzz | -0.023 |
Ruby | -0.022 |
seek | -0.022 |
| | -0.022 |
Token | like |
Token ID | 588 |
Feature activation | +6.31e-03 |
Pos logit contributions | |
happens | +0.013 |
ield | +0.012 |
happened | +0.012 |
Dan | +0.011 |
David | +0.010 |
Neg logit contributions | |
Brown | -0.014 |
Buzz | -0.014 |
seek | -0.013 |
Ruby | -0.013 |
| | -0.013 |
Token | the |
Token ID | 262 |
Feature activation | +1.21e-02 |
Pos logit contributions | |
happens | +0.011 |
ield | +0.011 |
happened | +0.011 |
Dan | +0.010 |
meant | +0.009 |
Neg logit contributions | |
Brown | -0.011 |
Buzz | -0.010 |
Ruby | -0.010 |
seek | -0.010 |
| | -0.010 |
Token | wardrobe |
Token ID | 35593 |
Feature activation | +2.42e-03 |
Pos logit contributions | |
happens | +0.022 |
happened | +0.022 |
ield | +0.021 |
Dan | +0.019 |
meant | +0.019 |
Neg logit contributions | |
Brown | -0.020 |
Buzz | -0.019 |
Later | -0.018 |
Ruby | -0.018 |
Fl | -0.018 |
Token | 's |
Token ID | 338 |
Feature activation | +8.58e-03 |
Pos logit contributions | |
Brown | +0.031 |
Buzz | +0.029 |
seek | +0.028 |
Ruby | +0.028 |
Fl | +0.028 |
Neg logit contributions | |
happens | -0.030 |
ield | -0.028 |
happened | -0.028 |
Dan | -0.025 |
meant | -0.024 |
Token | mom |
Token ID | 1995 |
Feature activation | -9.31e-04 |
Pos logit contributions | |
happens | +0.014 |
ield | +0.013 |
happened | +0.013 |
Dan | +0.012 |
meant | +0.011 |
Neg logit contributions | |
Brown | -0.016 |
Buzz | -0.015 |
seek | -0.015 |
Later | -0.014 |
lf | -0.014 |
Token | my |
Token ID | 1820 |
Feature activation | -1.13e-03 |
Pos logit contributions | |
Brown | +0.001 |
Buzz | +0.001 |
| | +0.001 |
seek | +0.001 |
Ruby | +0.001 |
Neg logit contributions | |
happens | -0.002 |
ield | -0.002 |
happened | -0.002 |
Dan | -0.002 |
meant | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +9.30e-03 |
Pos logit contributions | |
Brown | +0.002 |
Buzz | +0.002 |
Ruby | +0.002 |
heads | +0.002 |
seek | +0.002 |
Neg logit contributions | |
happens | -0.002 |
happened | -0.002 |
ield | -0.002 |
Dan | -0.002 |
meant | -0.002 |
Token | " |
Token ID | 366 |
Feature activation | +3.82e-03 |
Pos logit contributions | |
happens | +0.016 |
happened | +0.014 |
ield | +0.014 |
meant | +0.012 |
vious | +0.011 |
Neg logit contributions | |
Brown | -0.018 |
Fl | -0.017 |
Later | -0.017 |
Ruby | -0.016 |
Ell | -0.016 |
Token | Looks |
Token ID | 41102 |
Feature activation | +7.87e-02 |
Pos logit contributions | |
happens | +0.005 |
happened | +0.005 |
ield | +0.005 |
Dan | +0.005 |
meant | +0.005 |
Neg logit contributions | |
Brown | -0.007 |
throne | -0.007 |
lf | -0.007 |
Buzz | -0.007 |
seek | -0.007 |
Token | like |
Token ID | 588 |
Feature activation | +4.93e-03 |
Pos logit contributions | |
ield | +0.139 |
Dan | +0.132 |
happened | +0.128 |
David | +0.126 |
happens | +0.125 |
Neg logit contributions | |
Buzz | -0.123 |
sleep | -0.118 |
Brown | -0.115 |
Blue | -0.115 |
Tweet | -0.115 |
Token | the |
Token ID | 262 |
Feature activation | +1.77e-02 |
Pos logit contributions | |
happens | +0.010 |
ield | +0.010 |
happened | +0.009 |
Dan | +0.009 |
meant | +0.008 |
Neg logit contributions | |
Brown | -0.007 |
Buzz | -0.007 |
seek | -0.007 |
Ruby | -0.006 |
carpet | -0.006 |
Token | wave |
Token ID | 6769 |
Feature activation | +2.86e-02 |
Pos logit contributions | |
ield | +0.038 |
happens | +0.037 |
happened | +0.035 |
oke | +0.033 |
meant | +0.032 |
Neg logit contributions | |
seek | -0.023 |
Buzz | -0.022 |
walk | -0.022 |
carpet | -0.022 |
throne | -0.021 |
Token | destroyed |
Token ID | 6572 |
Feature activation | -1.99e-03 |
Pos logit contributions | |
happens | +0.049 |
ield | +0.048 |
happened | +0.047 |
Dan | +0.044 |
David | +0.042 |
Neg logit contributions | |
Brown | -0.049 |
Buzz | -0.046 |
seek | -0.045 |
Ruby | -0.044 |
Fl | -0.044 |
Token | all |
Token ID | 477 |
Feature activation | +5.67e-04 |
Pos logit contributions | |
Brown | +0.003 |
Buzz | +0.003 |
seek | +0.003 |
Ruby | +0.003 |
Fl | +0.003 |
Neg logit contributions | |
happens | -0.004 |
ield | -0.004 |
happened | -0.004 |
Dan | -0.004 |
meant | -0.003 |
Token | the |
Token ID | 262 |
Feature activation | +4.14e-03 |
Pos logit contributions | |
Brown | +0.022 |
Buzz | +0.021 |
seek | +0.020 |
Ruby | +0.020 |
Fl | +0.020 |
Neg logit contributions | |
happens | -0.023 |
ield | -0.022 |
happened | -0.021 |
Dan | -0.020 |
meant | -0.019 |
Token | owner |
Token ID | 4870 |
Feature activation | -1.33e-03 |
Pos logit contributions | |
happens | +0.007 |
ield | +0.007 |
happened | +0.007 |
Dan | +0.006 |
David | +0.006 |
Neg logit contributions | |
Brown | -0.007 |
Buzz | -0.007 |
seek | -0.007 |
Ruby | -0.007 |
Fl | -0.006 |
Token | smiled |
Token ID | 13541 |
Feature activation | +1.38e-02 |
Pos logit contributions | |
Brown | +0.003 |
Buzz | +0.002 |
heads | +0.002 |
Later | +0.002 |
| | +0.002 |
Neg logit contributions | |
happens | -0.002 |
happened | -0.002 |
ield | -0.002 |
Dan | -0.002 |
meant | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +1.37e-02 |
Pos logit contributions | |
happens | +0.028 |
ield | +0.027 |
happened | +0.027 |
Dan | +0.025 |
oke | +0.024 |
Neg logit contributions | |
Brown | -0.020 |
Buzz | -0.019 |
Later | -0.018 |
seek | -0.018 |
Mid | -0.017 |
Token | " |
Token ID | 366 |
Feature activation | -6.54e-03 |
Pos logit contributions | |
happens | +0.020 |
happened | +0.019 |
ield | +0.019 |
Dan | +0.016 |
meant | +0.016 |
Neg logit contributions | |
Brown | -0.029 |
Buzz | -0.027 |
Later | -0.026 |
Fl | -0.026 |
Ruby | -0.026 |
Token | Looks |
Token ID | 41102 |
Feature activation | +7.82e-02 |
Pos logit contributions | |
Brown | +0.011 |
Buzz | +0.011 |
lf | +0.011 |
seek | +0.011 |
heads | +0.011 |
Neg logit contributions | |
happens | -0.011 |
ield | -0.010 |
happened | -0.010 |
Dan | -0.010 |
meant | -0.009 |
Token | like |
Token ID | 588 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
ield | +0.136 |
happened | +0.127 |
Dan | +0.127 |
happens | +0.125 |
David | +0.121 |
Neg logit contributions | |
Buzz | -0.126 |
sleep | -0.122 |
Tweet | -0.119 |
Blue | -0.119 |
Brown | -0.119 |
Token | you |
Token ID | 345 |
Feature activation | +6.94e-03 |
Pos logit contributions | |
happens | +0.021 |
ield | +0.020 |
happened | +0.020 |
Dan | +0.018 |
meant | +0.018 |
Neg logit contributions | |
Brown | -0.017 |
Buzz | -0.016 |
seek | -0.015 |
Ruby | -0.015 |
| | -0.015 |
Token | two |
Token ID | 734 |
Feature activation | +1.20e-02 |
Pos logit contributions | |
happens | +0.009 |
happened | +0.009 |
ield | +0.009 |
meant | +0.008 |
Dan | +0.008 |
Neg logit contributions | |
Brown | -0.015 |
Buzz | -0.014 |
Ruby | -0.014 |
lf | -0.014 |
Later | -0.014 |
Token | are |
Token ID | 389 |
Feature activation | +2.82e-02 |
Pos logit contributions | |
happens | +0.021 |
ield | +0.020 |
happened | +0.020 |
Dan | +0.018 |
meant | +0.017 |
Neg logit contributions | |
Brown | -0.022 |
Buzz | -0.020 |
Ruby | -0.020 |
Later | -0.019 |
seek | -0.019 |
Token | friends |
Token ID | 2460 |
Feature activation | -1.48e-02 |
Pos logit contributions | |
happens | +0.044 |
happened | +0.042 |
ield | +0.042 |
Dan | +0.038 |
meant | +0.037 |
Neg logit contributions | |
Brown | -0.054 |
Buzz | -0.052 |
Ruby | -0.050 |
Later | -0.050 |
| | -0.050 |
Token | beach |
Token ID | 10481 |
Feature activation | -2.94e-02 |
Pos logit contributions | |
Brown | +0.007 |
Buzz | +0.006 |
seek | +0.006 |
carpet | +0.006 |
walk | +0.006 |
Neg logit contributions | |
happens | -0.013 |
ield | -0.013 |
happened | -0.013 |
Dan | -0.012 |
meant | -0.011 |
Token | to |
Token ID | 284 |
Feature activation | -4.49e-02 |
Pos logit contributions | |
Brown | +0.050 |
Buzz | +0.046 |
seek | +0.045 |
Ruby | +0.044 |
Later | +0.044 |
Neg logit contributions | |
happens | -0.054 |
ield | -0.051 |
happened | -0.051 |
Dan | -0.046 |
meant | -0.044 |
Token | see |
Token ID | 766 |
Feature activation | +1.65e-02 |
Pos logit contributions | |
Brown | +0.066 |
Buzz | +0.061 |
Ruby | +0.058 |
seek | +0.058 |
Fl | +0.058 |
Neg logit contributions | |
happens | -0.092 |
ield | -0.089 |
happened | -0.088 |
Dan | -0.083 |
meant | -0.079 |
Token | what |
Token ID | 644 |
Feature activation | +4.44e-02 |
Pos logit contributions | |
happens | +0.034 |
ield | +0.033 |
happened | +0.033 |
oke | +0.028 |
Dan | +0.028 |
Neg logit contributions | |
Brown | -0.022 |
Buzz | -0.021 |
ide | -0.020 |
blankets | -0.020 |
cushion | -0.020 |
Token | the |
Token ID | 262 |
Feature activation | +2.11e-02 |
Pos logit contributions | |
ield | +0.094 |
fault | +0.088 |
unes | +0.087 |
itted | +0.087 |
ses | +0.084 |
Neg logit contributions | |
Buzz | -0.059 |
lf | -0.058 |
seek | -0.054 |
ide | -0.053 |
Brown | -0.052 |
Token | sound |
Token ID | 2128 |
Feature activation | +7.81e-02 |
Pos logit contributions | |
happens | +0.046 |
ield | +0.046 |
happened | +0.044 |
oke | +0.041 |
Dan | +0.040 |
Neg logit contributions | |
seek | -0.025 |
Buzz | -0.024 |
Brown | -0.024 |
carpet | -0.024 |
walk | -0.023 |
Token | was |
Token ID | 373 |
Feature activation | +4.10e-02 |
Pos logit contributions | |
fault | +0.141 |
ield | +0.140 |
Dan | +0.127 |
happens | +0.126 |
David | +0.125 |
Neg logit contributions | |
lf | -0.117 |
ide | -0.106 |
ie | -0.106 |
walk | -0.103 |
Brown | -0.102 |
Token | . |
Token ID | 13 |
Feature activation | +2.26e-02 |
Pos logit contributions | |
happens | +0.083 |
ield | +0.083 |
Dan | +0.075 |
happened | +0.074 |
Well | +0.072 |
Neg logit contributions | |
Brown | -0.053 |
seek | -0.052 |
carpet | -0.051 |
sleep | -0.050 |
walk | -0.050 |
Token |
|
Token ID | 198 |
Feature activation | -1.93e-02 |
Pos logit contributions | |
happens | +0.050 |
happened | +0.046 |
ield | +0.044 |
fault | +0.041 |
ier | +0.040 |
Neg logit contributions | |
Later | -0.031 |
Buzz | -0.029 |
Brown | -0.029 |
While | -0.029 |
sleep | -0.027 |
Token |
|
Token ID | 198 |
Feature activation | -2.72e-02 |
Pos logit contributions | |
Brown | +0.032 |
Buzz | +0.030 |
Later | +0.029 |
seek | +0.029 |
Ruby | +0.029 |
Neg logit contributions | |
happens | -0.036 |
ield | -0.035 |
happened | -0.034 |
Dan | -0.031 |
meant | -0.030 |
Token | When |
Token ID | 2215 |
Feature activation | +2.53e-02 |
Pos logit contributions | |
Brown | +0.039 |
Buzz | +0.037 |
Later | +0.035 |
seek | +0.035 |
Ruby | +0.034 |
Neg logit contributions | |
happens | -0.057 |
happened | -0.055 |
ield | -0.055 |
Dan | -0.050 |
meant | -0.049 |
Token | sound |
Token ID | 2128 |
Feature activation | +6.56e-02 |
Pos logit contributions | |
happens | +0.023 |
ield | +0.022 |
happened | +0.022 |
Dan | +0.021 |
meant | +0.020 |
Neg logit contributions | |
Brown | -0.016 |
Buzz | -0.014 |
seek | -0.014 |
walk | -0.014 |
carpet | -0.014 |
Token | made |
Token ID | 925 |
Feature activation | +2.85e-02 |
Pos logit contributions | |
ield | +0.112 |
fault | +0.109 |
happens | +0.109 |
Dan | +0.106 |
David | +0.100 |
Neg logit contributions | |
Brown | -0.108 |
lf | -0.103 |
ide | -0.099 |
Buzz | -0.097 |
ie | -0.096 |
Token | him |
Token ID | 683 |
Feature activation | +6.80e-04 |
Pos logit contributions | |
happens | +0.057 |
ield | +0.054 |
happened | +0.052 |
Dan | +0.048 |
oke | +0.048 |
Neg logit contributions | |
Brown | -0.042 |
Buzz | -0.040 |
ide | -0.038 |
Later | -0.038 |
sleep | -0.038 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.45e-02 |
Pos logit contributions | |
happens | +0.001 |
ield | +0.001 |
happened | +0.001 |
Dan | +0.001 |
David | +0.001 |
Neg logit contributions | |
Brown | -0.001 |
Buzz | -0.001 |
seek | -0.001 |
| | -0.001 |
Ruby | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +2.48e-02 |
Pos logit contributions | |
happens | +0.029 |
ield | +0.028 |
happened | +0.027 |
Dan | +0.026 |
David | +0.025 |
Neg logit contributions | |
Brown | -0.020 |
seek | -0.019 |
sleep | -0.019 |
ide | -0.018 |
Buzz | -0.018 |
Token | Suddenly |
Token ID | 24975 |
Feature activation | +7.78e-02 |
Pos logit contributions | |
happens | +0.053 |
happened | +0.048 |
ield | +0.047 |
fault | +0.043 |
oke | +0.041 |
Neg logit contributions | |
Buzz | -0.036 |
Later | -0.036 |
Brown | -0.032 |
ide | -0.031 |
walk | -0.031 |
Token | , |
Token ID | 11 |
Feature activation | +5.07e-02 |
Pos logit contributions | |
ield | +0.157 |
happens | +0.152 |
fault | +0.151 |
oke | +0.143 |
itted | +0.137 |
Neg logit contributions | |
Buzz | -0.100 |
seek | -0.097 |
Bun | -0.097 |
walk | -0.096 |
Brown | -0.094 |
Token | a |
Token ID | 257 |
Feature activation | +1.88e-02 |
Pos logit contributions | |
happens | +0.112 |
ield | +0.110 |
oke | +0.103 |
fault | +0.102 |
happened | +0.096 |
Neg logit contributions | |
Buzz | -0.063 |
seek | -0.060 |
walk | -0.058 |
Brown | -0.057 |
Later | -0.056 |
Token | puff |
Token ID | 33679 |
Feature activation | +5.76e-02 |
Pos logit contributions | |
happens | +0.041 |
ield | +0.039 |
happened | +0.038 |
Dan | +0.037 |
meant | +0.035 |
Neg logit contributions | |
Brown | -0.026 |
Buzz | -0.023 |
seek | -0.023 |
Ruby | -0.022 |
lf | -0.022 |
Token | of |
Token ID | 286 |
Feature activation | +2.98e-03 |
Pos logit contributions | |
happens | +0.100 |
fault | +0.096 |
Dan | +0.095 |
ield | +0.091 |
Well | +0.089 |
Neg logit contributions | |
Brown | -0.094 |
ide | -0.087 |
Buzz | -0.087 |
ie | -0.086 |
lf | -0.086 |
Token | smoke |
Token ID | 7523 |
Feature activation | +4.19e-02 |
Pos logit contributions | |
happens | +0.005 |
ield | +0.005 |
happened | +0.005 |
Dan | +0.005 |
oke | +0.004 |
Neg logit contributions | |
Brown | -0.005 |
Buzz | -0.005 |
blankets | -0.004 |
carpet | -0.004 |
seek | -0.004 |
Token | fence |
Token ID | 13990 |
Feature activation | +9.35e-03 |
Pos logit contributions | |
happens | +0.023 |
happened | +0.022 |
ield | +0.022 |
Dan | +0.021 |
David | +0.020 |
Neg logit contributions | |
Brown | -0.013 |
Buzz | -0.012 |
seek | -0.011 |
carpet | -0.011 |
Fl | -0.011 |
Token | to |
Token ID | 284 |
Feature activation | -2.48e-02 |
Pos logit contributions | |
happens | +0.020 |
happened | +0.018 |
ield | +0.018 |
Dan | +0.016 |
meant | +0.016 |
Neg logit contributions | |
Brown | -0.013 |
Buzz | -0.012 |
ide | -0.012 |
seek | -0.012 |
Later | -0.012 |
Token | see |
Token ID | 766 |
Feature activation | +2.55e-02 |
Pos logit contributions | |
Brown | +0.039 |
Buzz | +0.036 |
Ruby | +0.035 |
Later | +0.034 |
Mid | +0.034 |
Neg logit contributions | |
happens | -0.048 |
happened | -0.047 |
ield | -0.047 |
Dan | -0.046 |
David | -0.042 |
Token | what |
Token ID | 644 |
Feature activation | +4.60e-02 |
Pos logit contributions | |
happens | +0.051 |
ield | +0.050 |
happened | +0.049 |
Dan | +0.043 |
oke | +0.043 |
Neg logit contributions | |
Brown | -0.038 |
Buzz | -0.034 |
Fl | -0.034 |
blankets | -0.033 |
carpet | -0.032 |
Token | the |
Token ID | 262 |
Feature activation | +2.14e-02 |
Pos logit contributions | |
ield | +0.091 |
unes | +0.085 |
itted | +0.084 |
fault | +0.082 |
elt | +0.082 |
Neg logit contributions | |
Buzz | -0.064 |
Brown | -0.064 |
lf | -0.063 |
seek | -0.060 |
Bun | -0.060 |
Token | sound |
Token ID | 2128 |
Feature activation | +7.68e-02 |
Pos logit contributions | |
happens | +0.047 |
ield | +0.044 |
happened | +0.043 |
Dan | +0.041 |
David | +0.040 |
Neg logit contributions | |
Brown | -0.028 |
carpet | -0.026 |
Buzz | -0.026 |
seek | -0.026 |
blankets | -0.025 |
Token | was |
Token ID | 373 |
Feature activation | +4.39e-02 |
Pos logit contributions | |
fault | +0.135 |
ield | +0.134 |
Dan | +0.126 |
happens | +0.124 |
David | +0.121 |
Neg logit contributions | |
lf | -0.116 |
Brown | -0.106 |
walk | -0.103 |
ide | -0.103 |
ie | -0.103 |
Token | . |
Token ID | 13 |
Feature activation | +2.50e-02 |
Pos logit contributions | |
happens | +0.089 |
ield | +0.088 |
Dan | +0.082 |
Well | +0.079 |
happened | +0.078 |
Neg logit contributions | |
Brown | -0.055 |
carpet | -0.055 |
seek | -0.053 |
walk | -0.053 |
sleep | -0.052 |
Token |
|
Token ID | 198 |
Feature activation | -1.73e-02 |
Pos logit contributions | |
happens | +0.049 |
happened | +0.044 |
ield | +0.043 |
meant | +0.038 |
oke | +0.038 |
Neg logit contributions | |
Brown | -0.041 |
Buzz | -0.040 |
Later | -0.038 |
Bun | -0.037 |
| | -0.037 |
Token |
|
Token ID | 198 |
Feature activation | -2.65e-02 |
Pos logit contributions | |
Brown | +0.032 |
Buzz | +0.030 |
seek | +0.029 |
Ruby | +0.029 |
| | +0.028 |
Neg logit contributions | |
happens | -0.029 |
ield | -0.028 |
happened | -0.027 |
Dan | -0.025 |
meant | -0.024 |
Token | It |
Token ID | 1026 |
Feature activation | +3.60e-02 |
Pos logit contributions | |
Brown | +0.036 |
| | +0.033 |
throne | +0.032 |
Buzz | +0.032 |
Ruby | +0.032 |
Neg logit contributions | |
happens | -0.054 |
ield | -0.054 |
happened | -0.053 |
Dan | -0.050 |
David | -0.048 |
Token | asked |
Token ID | 1965 |
Feature activation | +8.05e-03 |
Pos logit contributions | |
Brown | +0.004 |
Buzz | +0.004 |
seek | +0.004 |
Ruby | +0.004 |
Later | +0.004 |
Neg logit contributions | |
happens | -0.003 |
ield | -0.003 |
happened | -0.003 |
Dan | -0.003 |
meant | -0.003 |
Token | his |
Token ID | 465 |
Feature activation | +4.51e-03 |
Pos logit contributions | |
ield | +0.021 |
happens | +0.021 |
happened | +0.020 |
fault | +0.020 |
oke | +0.019 |
Neg logit contributions | |
lf | -0.005 |
seek | -0.005 |
sleep | -0.005 |
carpet | -0.005 |
ide | -0.005 |
Token | dad |
Token ID | 9955 |
Feature activation | +1.32e-02 |
Pos logit contributions | |
happens | +0.010 |
ield | +0.009 |
happened | +0.009 |
Dan | +0.008 |
oke | +0.008 |
Neg logit contributions | |
tidy | -0.005 |
walk | -0.005 |
lf | -0.005 |
carpet | -0.005 |
seek | -0.005 |
Token | what |
Token ID | 644 |
Feature activation | +4.88e-02 |
Pos logit contributions | |
happens | +0.024 |
ield | +0.023 |
happened | +0.023 |
Dan | +0.020 |
meant | +0.019 |
Neg logit contributions | |
ide | -0.020 |
Brown | -0.020 |
Buzz | -0.020 |
seek | -0.019 |
ie | -0.019 |
Token | the |
Token ID | 262 |
Feature activation | +1.60e-02 |
Pos logit contributions | |
ield | +0.104 |
unes | +0.099 |
oke | +0.098 |
fault | +0.097 |
itted | +0.097 |
Neg logit contributions | |
lf | -0.061 |
play | -0.057 |
read | -0.057 |
walk | -0.056 |
seek | -0.056 |
Token | thing |
Token ID | 1517 |
Feature activation | +7.64e-02 |
Pos logit contributions | |
happens | +0.032 |
ield | +0.030 |
happened | +0.030 |
Dan | +0.028 |
oke | +0.027 |
Neg logit contributions | |
Brown | -0.024 |
Buzz | -0.022 |
seek | -0.022 |
carpet | -0.021 |
walk | -0.021 |
Token | with |
Token ID | 351 |
Feature activation | -9.95e-04 |
Pos logit contributions | |
ield | +0.141 |
isel | +0.135 |
happens | +0.131 |
itted | +0.127 |
oke | +0.126 |
Neg logit contributions | |
lf | -0.107 |
walk | -0.104 |
seek | -0.102 |
Brown | -0.099 |
carpet | -0.098 |
Token | the |
Token ID | 262 |
Feature activation | +1.53e-02 |
Pos logit contributions | |
Brown | +0.001 |
blankets | +0.001 |
sleep | +0.001 |
carpet | +0.001 |
pillow | +0.001 |
Neg logit contributions | |
happens | -0.002 |
ield | -0.002 |
happened | -0.002 |
oke | -0.002 |
Well | -0.002 |
Token | blue |
Token ID | 4171 |
Feature activation | +1.69e-02 |
Pos logit contributions | |
happens | +0.031 |
ield | +0.029 |
happened | +0.029 |
Dan | +0.027 |
David | +0.026 |
Neg logit contributions | |
Brown | -0.023 |
Buzz | -0.021 |
seek | -0.021 |
carpet | -0.020 |
Ruby | -0.020 |
Token | water |
Token ID | 1660 |
Feature activation | +1.77e-02 |
Pos logit contributions | |
happens | +0.034 |
happened | +0.032 |
ield | +0.031 |
Dan | +0.031 |
oke | +0.031 |
Neg logit contributions | |
carpet | -0.024 |
Brown | -0.023 |
cushion | -0.023 |
pillow | -0.022 |
blankets | -0.022 |
Token | in |
Token ID | 287 |
Feature activation | -8.31e-03 |
Pos logit contributions | |
happens | +0.036 |
ield | +0.034 |
happened | +0.033 |
Dan | +0.032 |
oke | +0.030 |
Neg logit contributions | |
carpet | -0.022 |
Brown | -0.022 |
lf | -0.022 |
walk | -0.022 |
shelf | -0.022 |
Token | this |
Token ID | 428 |
Feature activation | +2.97e-02 |
Pos logit contributions | |
ield | +0.088 |
happens | +0.086 |
fault | +0.083 |
oke | +0.083 |
elt | +0.083 |
Neg logit contributions | |
Brown | -0.055 |
lf | -0.053 |
walk | -0.050 |
seek | -0.050 |
Bun | -0.048 |
Token | happen |
Token ID | 1645 |
Feature activation | +1.61e-02 |
Pos logit contributions | |
happens | +0.050 |
ield | +0.049 |
Dan | +0.047 |
happened | +0.046 |
David | +0.043 |
Neg logit contributions | |
Brown | -0.051 |
Buzz | -0.047 |
seek | -0.047 |
lf | -0.047 |
carpet | -0.046 |
Token | ? |
Token ID | 30 |
Feature activation | +3.27e-02 |
Pos logit contributions | |
happens | +0.031 |
ield | +0.031 |
happened | +0.029 |
oke | +0.027 |
Dan | +0.027 |
Neg logit contributions | |
seek | -0.023 |
Brown | -0.022 |
lf | -0.021 |
sleep | -0.021 |
ide | -0.021 |
Token |
|
Token ID | 198 |
Feature activation | -8.65e-03 |
Pos logit contributions | |
happens | +0.070 |
happened | +0.065 |
ield | +0.061 |
enough | +0.055 |
meant | +0.055 |
Neg logit contributions | |
Brown | -0.049 |
Bun | -0.043 |
| | -0.043 |
Later | -0.042 |
Buzz | -0.042 |
Token |
|
Token ID | 198 |
Feature activation | -2.29e-02 |
Pos logit contributions | |
Brown | +0.013 |
Buzz | +0.012 |
Later | +0.012 |
Fl | +0.011 |
Ruby | +0.011 |
Neg logit contributions | |
happens | -0.018 |
ield | -0.017 |
happened | -0.017 |
meant | -0.015 |
Dan | -0.015 |
Token | Suddenly |
Token ID | 38582 |
Feature activation | +7.62e-02 |
Pos logit contributions | |
Brown | +0.027 |
Buzz | +0.024 |
Later | +0.024 |
seek | +0.023 |
Fl | +0.022 |
Neg logit contributions | |
happens | -0.055 |
happened | -0.053 |
ield | -0.052 |
Dan | -0.049 |
meant | -0.048 |
Token | , |
Token ID | 11 |
Feature activation | +4.49e-02 |
Pos logit contributions | |
ield | +0.155 |
happens | +0.152 |
oke | +0.148 |
fault | +0.146 |
emon | +0.135 |
Neg logit contributions | |
Bun | -0.101 |
Brown | -0.093 |
seek | -0.093 |
walk | -0.091 |
blankets | -0.091 |
Token | his |
Token ID | 465 |
Feature activation | +1.87e-02 |
Pos logit contributions | |
happens | +0.090 |
ield | +0.086 |
oke | +0.081 |
happened | +0.077 |
fault | +0.076 |
Neg logit contributions | |
Brown | -0.068 |
lf | -0.063 |
Bun | -0.063 |
seek | -0.062 |
blankets | -0.062 |
Token | mum |
Token ID | 25682 |
Feature activation | +2.00e-02 |
Pos logit contributions | |
happens | +0.035 |
ield | +0.033 |
happened | +0.033 |
Dan | +0.030 |
meant | +0.029 |
Neg logit contributions | |
Brown | -0.031 |
Buzz | -0.029 |
seek | -0.028 |
Ruby | -0.028 |
Later | -0.027 |
Token | appeared |
Token ID | 4120 |
Feature activation | +1.33e-02 |
Pos logit contributions | |
happens | +0.032 |
ield | +0.030 |
happened | +0.029 |
Dan | +0.027 |
meant | +0.025 |
Neg logit contributions | |
Brown | -0.039 |
Buzz | -0.036 |
Ruby | -0.035 |
seek | -0.035 |
lf | -0.035 |
Token | by |
Token ID | 416 |
Feature activation | +4.75e-03 |
Pos logit contributions | |
happens | +0.028 |
ield | +0.026 |
happened | +0.026 |
Dan | +0.024 |
oke | +0.023 |
Neg logit contributions | |
Brown | -0.019 |
Buzz | -0.018 |
seek | -0.017 |
Ruby | -0.017 |
Later | -0.017 |
Token | shake |
Token ID | 13279 |
Feature activation | +2.38e-02 |
Pos logit contributions | |
Brown | +0.033 |
Mic | +0.032 |
Ruby | +0.031 |
Mid | +0.031 |
heads | +0.031 |
Neg logit contributions | |
happens | -0.028 |
happened | -0.026 |
meant | -0.025 |
ield | -0.024 |
Dan | -0.024 |
Token | and |
Token ID | 290 |
Feature activation | -7.06e-03 |
Pos logit contributions | |
happens | +0.047 |
ield | +0.046 |
happened | +0.045 |
Dan | +0.041 |
oke | +0.040 |
Neg logit contributions | |
Brown | -0.036 |
Buzz | -0.033 |
seek | -0.032 |
Later | -0.032 |
Fl | -0.032 |
Token | rum |
Token ID | 7440 |
Feature activation | +2.09e-02 |
Pos logit contributions | |
Brown | +0.011 |
Buzz | +0.011 |
seek | +0.010 |
Ruby | +0.010 |
| | +0.010 |
Neg logit contributions | |
happens | -0.014 |
ield | -0.013 |
happened | -0.013 |
Dan | -0.012 |
meant | -0.011 |
Token | ble |
Token ID | 903 |
Feature activation | +2.21e-02 |
Pos logit contributions | |
happens | +0.036 |
ield | +0.035 |
happened | +0.034 |
Dan | +0.031 |
meant | +0.030 |
Neg logit contributions | |
Brown | -0.037 |
Buzz | -0.035 |
seek | -0.034 |
Ruby | -0.033 |
Later | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | +2.76e-02 |
Pos logit contributions | |
happens | +0.041 |
ield | +0.040 |
happened | +0.039 |
Dan | +0.037 |
David | +0.035 |
Neg logit contributions | |
Brown | -0.036 |
Buzz | -0.033 |
Later | -0.032 |
seek | -0.032 |
Ruby | -0.032 |
Token | Suddenly |
Token ID | 24975 |
Feature activation | +7.54e-02 |
Pos logit contributions | |
happens | +0.053 |
happened | +0.047 |
ield | +0.046 |
oke | +0.044 |
Dan | +0.042 |
Neg logit contributions | |
Brown | -0.045 |
Later | -0.042 |
Bun | -0.041 |
carpet | -0.041 |
Buzz | -0.040 |
Token | , |
Token ID | 11 |
Feature activation | +4.86e-02 |
Pos logit contributions | |
ield | +0.147 |
happens | +0.140 |
oke | +0.136 |
fault | +0.132 |
Well | +0.130 |
Neg logit contributions | |
Bun | -0.104 |
Brown | -0.102 |
blankets | -0.097 |
walk | -0.095 |
Buzz | -0.095 |
Token | the |
Token ID | 262 |
Feature activation | +8.19e-03 |
Pos logit contributions | |
happens | +0.094 |
ield | +0.094 |
oke | +0.090 |
happened | +0.082 |
fault | +0.082 |
Neg logit contributions | |
Brown | -0.072 |
seek | -0.067 |
Bun | -0.066 |
Buzz | -0.066 |
blankets | -0.065 |
Token | ground |
Token ID | 2323 |
Feature activation | +4.11e-02 |
Pos logit contributions | |
happens | +0.018 |
ield | +0.018 |
happened | +0.017 |
Dan | +0.016 |
David | +0.016 |
Neg logit contributions | |
Brown | -0.011 |
Buzz | -0.010 |
seek | -0.009 |
Ruby | -0.009 |
Later | -0.009 |
Token | gave |
Token ID | 2921 |
Feature activation | +3.29e-02 |
Pos logit contributions | |
happens | +0.076 |
ield | +0.075 |
Dan | +0.073 |
fault | +0.071 |
David | +0.071 |
Neg logit contributions | |
Brown | -0.061 |
Later | -0.060 |
lf | -0.059 |
Buzz | -0.058 |
seek | -0.056 |
Token | way |
Token ID | 835 |
Feature activation | +3.17e-02 |
Pos logit contributions | |
happens | +0.067 |
oke | +0.065 |
ield | +0.061 |
Well | +0.061 |
Dan | +0.060 |
Neg logit contributions | |
Brown | -0.046 |
Bun | -0.043 |
Buzz | -0.042 |
blankets | -0.040 |
Blue | -0.040 |
Token | She |
Token ID | 1375 |
Feature activation | +3.86e-03 |
Pos logit contributions | |
happens | +0.039 |
ield | +0.038 |
happened | +0.037 |
Dan | +0.034 |
meant | +0.032 |
Neg logit contributions | |
Brown | -0.042 |
Buzz | -0.040 |
seek | -0.039 |
Ruby | -0.038 |
| | -0.038 |
Token | realized |
Token ID | 6939 |
Feature activation | +3.70e-02 |
Pos logit contributions | |
happens | +0.006 |
happened | +0.006 |
ield | +0.006 |
Dan | +0.005 |
meant | +0.005 |
Neg logit contributions | |
Brown | -0.007 |
Buzz | -0.007 |
seek | -0.007 |
Ruby | -0.007 |
| | -0.007 |
Token | that |
Token ID | 326 |
Feature activation | +4.23e-02 |
Pos logit contributions | |
ield | +0.065 |
happens | +0.065 |
happened | +0.060 |
Dan | +0.056 |
oke | +0.056 |
Neg logit contributions | |
Brown | -0.061 |
Buzz | -0.056 |
lf | -0.056 |
Later | -0.055 |
Fl | -0.055 |
Token | the |
Token ID | 262 |
Feature activation | +1.63e-02 |
Pos logit contributions | |
happens | +0.075 |
ield | +0.075 |
happened | +0.071 |
Dan | +0.067 |
oke | +0.063 |
Neg logit contributions | |
Brown | -0.071 |
Buzz | -0.066 |
seek | -0.066 |
Ruby | -0.064 |
Fl | -0.063 |
Token | strange |
Token ID | 6283 |
Feature activation | +3.88e-02 |
Pos logit contributions | |
happens | +0.033 |
ield | +0.031 |
happened | +0.031 |
Dan | +0.029 |
David | +0.028 |
Neg logit contributions | |
Brown | -0.024 |
seek | -0.023 |
Buzz | -0.022 |
lf | -0.022 |
Ruby | -0.022 |
Token | sound |
Token ID | 2128 |
Feature activation | +7.53e-02 |
Pos logit contributions | |
happens | +0.072 |
ield | +0.069 |
Dan | +0.067 |
happened | +0.067 |
icked | +0.065 |
Neg logit contributions | |
lf | -0.057 |
shelf | -0.057 |
carpet | -0.056 |
seek | -0.056 |
cushion | -0.055 |
Token | was |
Token ID | 373 |
Feature activation | +5.31e-02 |
Pos logit contributions | |
ield | +0.124 |
Dan | +0.121 |
happens | +0.117 |
David | +0.113 |
fault | +0.110 |
Neg logit contributions | |
Brown | -0.124 |
lf | -0.120 |
carpet | -0.116 |
seek | -0.116 |
walk | -0.113 |
Token | coming |
Token ID | 2406 |
Feature activation | +1.20e-02 |
Pos logit contributions | |
ield | +0.098 |
happens | +0.097 |
happened | +0.088 |
Dan | +0.087 |
oke | +0.085 |
Neg logit contributions | |
Brown | -0.085 |
seek | -0.080 |
Buzz | -0.077 |
carpet | -0.075 |
sleep | -0.075 |
Token | from |
Token ID | 422 |
Feature activation | -5.13e-03 |
Pos logit contributions | |
happens | +0.023 |
ield | +0.022 |
happened | +0.022 |
Dan | +0.021 |
David | +0.019 |
Neg logit contributions | |
Brown | -0.019 |
Buzz | -0.017 |
seek | -0.017 |
Ruby | -0.016 |
Later | -0.016 |
Token | the |
Token ID | 262 |
Feature activation | -1.54e-03 |
Pos logit contributions | |
Brown | +0.008 |
Buzz | +0.008 |
Ruby | +0.007 |
seek | +0.007 |
Later | +0.007 |
Neg logit contributions | |
happens | -0.010 |
ield | -0.009 |
happened | -0.009 |
Dan | -0.009 |
meant | -0.008 |
Token | wall |
Token ID | 3355 |
Feature activation | +2.95e-02 |
Pos logit contributions | |
Brown | +0.002 |
Buzz | +0.002 |
Ruby | +0.002 |
| | +0.002 |
Later | +0.002 |
Neg logit contributions | |
happens | -0.003 |
ield | -0.003 |
happened | -0.003 |
Dan | -0.003 |
meant | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | +3.84e-02 |
Pos logit contributions | |
Brown | +0.011 |
Buzz | +0.010 |
seek | +0.010 |
Ruby | +0.009 |
| | +0.009 |
Neg logit contributions | |
happens | -0.010 |
ield | -0.010 |
happened | -0.010 |
Dan | -0.009 |
meant | -0.008 |
Token | But |
Token ID | 887 |
Feature activation | +4.24e-02 |
Pos logit contributions | |
happens | +0.058 |
happened | +0.053 |
fault | +0.044 |
ield | +0.044 |
ier | +0.044 |
Neg logit contributions | |
Later | -0.075 |
Buzz | -0.071 |
Brown | -0.070 |
sleep | -0.069 |
| | -0.068 |
Token | the |
Token ID | 262 |
Feature activation | +1.86e-02 |
Pos logit contributions | |
happens | +0.064 |
ield | +0.059 |
happened | +0.054 |
oke | +0.051 |
fault | +0.049 |
Neg logit contributions | |
seek | -0.083 |
walk | -0.079 |
Later | -0.079 |
sleep | -0.078 |
Buzz | -0.077 |
Token | most |
Token ID | 749 |
Feature activation | +4.00e-02 |
Pos logit contributions | |
happens | +0.038 |
ield | +0.037 |
happened | +0.037 |
Dan | +0.034 |
David | +0.033 |
Neg logit contributions | |
Brown | -0.027 |
Buzz | -0.025 |
seek | -0.024 |
Ruby | -0.024 |
| | -0.023 |
Token | exciting |
Token ID | 7895 |
Feature activation | +2.42e-02 |
Pos logit contributions | |
ield | +0.080 |
happens | +0.080 |
Dan | +0.072 |
happened | +0.072 |
David | +0.069 |
Neg logit contributions | |
Buzz | -0.058 |
Brown | -0.057 |
seek | -0.054 |
lf | -0.052 |
Tweet | -0.051 |
Token | thing |
Token ID | 1517 |
Feature activation | +7.50e-02 |
Pos logit contributions | |
happens | +0.046 |
ield | +0.045 |
happened | +0.042 |
Dan | +0.041 |
oke | +0.041 |
Neg logit contributions | |
seek | -0.036 |
walk | -0.034 |
lf | -0.034 |
Brown | -0.034 |
Buzz | -0.034 |
Token | of |
Token ID | 286 |
Feature activation | +2.07e-03 |
Pos logit contributions | |
ield | +0.134 |
happens | +0.130 |
oke | +0.113 |
itted | +0.113 |
fault | +0.112 |
Neg logit contributions | |
seek | -0.105 |
ide | -0.103 |
lf | -0.102 |
Brown | -0.101 |
ie | -0.101 |
Token | all |
Token ID | 477 |
Feature activation | +3.44e-02 |
Pos logit contributions | |
happens | +0.005 |
ield | +0.004 |
happened | +0.004 |
oke | +0.004 |
ed | +0.004 |
Neg logit contributions | |
Buzz | -0.002 |
seek | -0.002 |
sleep | -0.002 |
Brown | -0.002 |
walk | -0.002 |
Token | was |
Token ID | 373 |
Feature activation | +4.75e-02 |
Pos logit contributions | |
ield | +0.061 |
happens | +0.060 |
happened | +0.056 |
Dan | +0.054 |
Well | +0.050 |
Neg logit contributions | |
Brown | -0.052 |
Buzz | -0.051 |
ide | -0.050 |
lf | -0.050 |
carpet | -0.049 |
Token | what |
Token ID | 644 |
Feature activation | +6.22e-02 |
Pos logit contributions | |
happens | +0.095 |
ield | +0.094 |
happened | +0.081 |
oke | +0.080 |
Well | +0.079 |
Neg logit contributions | |
seek | -0.064 |
Brown | -0.064 |
Buzz | -0.061 |
walk | -0.061 |
blankets | -0.060 |
Token | Joe |
Token ID | 5689 |
Feature activation | +2.74e-02 |
Pos logit contributions | |
ield | +0.106 |
happens | +0.099 |
ays | +0.096 |
oke | +0.095 |
itted | +0.094 |
Neg logit contributions | |
seek | -0.098 |
Buzz | -0.095 |
walk | -0.089 |
lf | -0.087 |
Bee | -0.087 |
Token | didn |
Token ID | 1422 |
Feature activation | -1.52e-03 |
Pos logit contributions | |
Brown | +0.011 |
Buzz | +0.010 |
Ruby | +0.010 |
seek | +0.010 |
Later | +0.010 |
Neg logit contributions | |
happens | -0.010 |
ield | -0.010 |
happened | -0.010 |
Dan | -0.009 |
fault | -0.009 |
Token | 't |
Token ID | 470 |
Feature activation | +5.17e-03 |
Pos logit contributions | |
Brown | +0.002 |
Buzz | +0.002 |
Later | +0.002 |
Ruby | +0.002 |
Fl | +0.002 |
Neg logit contributions | |
happens | -0.003 |
happened | -0.003 |
ield | -0.003 |
Dan | -0.003 |
meant | -0.002 |
Token | know |
Token ID | 760 |
Feature activation | +3.10e-02 |
Pos logit contributions | |
ield | +0.010 |
happens | +0.010 |
happened | +0.010 |
Dan | +0.009 |
David | +0.009 |
Neg logit contributions | |
Buzz | -0.007 |
Brown | -0.007 |
seek | -0.007 |
sleep | -0.007 |
ide | -0.007 |
Token | what |
Token ID | 644 |
Feature activation | +4.10e-02 |
Pos logit contributions | |
happens | +0.062 |
ield | +0.061 |
happened | +0.057 |
Dan | +0.051 |
meant | +0.051 |
Neg logit contributions | |
Brown | -0.045 |
ide | -0.043 |
lf | -0.042 |
Buzz | -0.042 |
Fl | -0.041 |
Token | the |
Token ID | 262 |
Feature activation | +2.96e-02 |
Pos logit contributions | |
ield | +0.084 |
happens | +0.080 |
ed | +0.074 |
ays | +0.073 |
oke | +0.072 |
Neg logit contributions | |
Brown | -0.056 |
lf | -0.055 |
Buzz | -0.054 |
seek | -0.053 |
Ruby | -0.052 |
Token | word |
Token ID | 1573 |
Feature activation | +7.49e-02 |
Pos logit contributions | |
happens | +0.063 |
ield | +0.061 |
Dan | +0.060 |
happened | +0.058 |
oke | +0.057 |
Neg logit contributions | |
Brown | -0.039 |
lf | -0.034 |
walk | -0.033 |
cushion | -0.033 |
seek | -0.032 |
Token | " |
Token ID | 366 |
Feature activation | +1.84e-02 |
Pos logit contributions | |
happens | +0.118 |
ield | +0.117 |
Dan | +0.111 |
isel | +0.110 |
fault | +0.105 |
Neg logit contributions | |
Brown | -0.126 |
lf | -0.119 |
cushion | -0.115 |
shelf | -0.114 |
seek | -0.114 |
Token | harm |
Token ID | 29155 |
Feature activation | +3.06e-02 |
Pos logit contributions | |
happens | +0.033 |
ield | +0.032 |
happened | +0.031 |
Dan | +0.029 |
David | +0.028 |
Neg logit contributions | |
Brown | -0.032 |
Buzz | -0.030 |
seek | -0.028 |
| | -0.028 |
Ruby | -0.028 |
Token | less |
Token ID | 1203 |
Feature activation | +2.49e-02 |
Pos logit contributions | |
happens | +0.055 |
happened | +0.051 |
meant | +0.043 |
ield | +0.043 |
Dan | +0.041 |
Neg logit contributions | |
Brown | -0.052 |
ide | -0.050 |
ge | -0.050 |
heads | -0.049 |
ie | -0.048 |
Token | " |
Token ID | 1 |
Feature activation | +2.55e-02 |
Pos logit contributions | |
happens | +0.047 |
happened | +0.042 |
ield | +0.041 |
Dan | +0.040 |
isel | +0.039 |
Neg logit contributions | |
Brown | -0.037 |
Buzz | -0.036 |
ide | -0.035 |
seek | -0.035 |
Fl | -0.034 |
Token | meant |
Token ID | 4001 |
Feature activation | +4.77e-02 |
Pos logit contributions | |
happens | +0.040 |
ield | +0.037 |
isel | +0.034 |
happened | +0.034 |
Dan | +0.033 |
Neg logit contributions | |
Brown | -0.047 |
seek | -0.044 |
Ell | -0.043 |
Hope | -0.043 |
ide | -0.042 |
Token | parked |
Token ID | 19584 |
Feature activation | -2.83e-03 |
Pos logit contributions | |
happens | +0.004 |
ield | +0.004 |
happened | +0.004 |
Dan | +0.004 |
David | +0.003 |
Neg logit contributions | |
Brown | -0.003 |
lf | -0.002 |
Later | -0.002 |
seek | -0.002 |
walk | -0.002 |
Token | by |
Token ID | 416 |
Feature activation | -8.34e-03 |
Pos logit contributions | |
Brown | +0.004 |
Buzz | +0.004 |
seek | +0.004 |
Ruby | +0.004 |
| | +0.004 |
Neg logit contributions | |
happens | -0.006 |
ield | -0.005 |
happened | -0.005 |
Dan | -0.005 |
meant | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | -3.07e-02 |
Pos logit contributions | |
Brown | +0.013 |
Buzz | +0.012 |
seek | +0.012 |
| | +0.012 |
Later | +0.012 |
Neg logit contributions | |
happens | -0.016 |
ield | -0.015 |
happened | -0.015 |
Dan | -0.014 |
meant | -0.013 |
Token | fence |
Token ID | 13990 |
Feature activation | +1.88e-03 |
Pos logit contributions | |
Brown | +0.044 |
Buzz | +0.043 |
| | +0.041 |
Mid | +0.041 |
Ruby | +0.040 |
Neg logit contributions | |
happens | -0.060 |
ield | -0.059 |
happened | -0.058 |
Dan | -0.054 |
David | -0.052 |
Token | . |
Token ID | 13 |
Feature activation | +3.96e-02 |
Pos logit contributions | |
happens | +0.004 |
happened | +0.004 |
ield | +0.004 |
Dan | +0.003 |
David | +0.003 |
Neg logit contributions | |
Brown | -0.003 |
ide | -0.002 |
Buzz | -0.002 |
Later | -0.002 |
seek | -0.002 |
Token | Suddenly |
Token ID | 24975 |
Feature activation | +7.46e-02 |
Pos logit contributions | |
happens | +0.071 |
fault | +0.060 |
ier | +0.060 |
oke | +0.059 |
happened | +0.059 |
Neg logit contributions | |
Brown | -0.069 |
Later | -0.067 |
walk | -0.065 |
Bun | -0.064 |
Blue | -0.064 |
Token | , |
Token ID | 11 |
Feature activation | +4.86e-02 |
Pos logit contributions | |
ield | +0.144 |
happens | +0.142 |
oke | +0.141 |
fault | +0.138 |
Well | +0.133 |
Neg logit contributions | |
Brown | -0.099 |
Bun | -0.098 |
walk | -0.097 |
blankets | -0.094 |
seek | -0.092 |
Token | they |
Token ID | 484 |
Feature activation | +4.66e-04 |
Pos logit contributions | |
happens | +0.104 |
ield | +0.101 |
oke | +0.100 |
fault | +0.095 |
happened | +0.089 |
Neg logit contributions | |
Brown | -0.062 |
walk | -0.059 |
seek | -0.057 |
Bun | -0.057 |
blankets | -0.056 |
Token | heard |
Token ID | 2982 |
Feature activation | +3.97e-02 |
Pos logit contributions | |
happens | +0.001 |
ield | +0.001 |
happened | +0.001 |
fault | +0.001 |
Well | +0.001 |
Neg logit contributions | |
Brown | -0.001 |
Buzz | -0.001 |
walk | -0.001 |
Later | -0.001 |
seek | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | -1.54e-02 |
Pos logit contributions | |
happens | +0.074 |
ield | +0.070 |
oke | +0.067 |
happened | +0.066 |
fault | +0.065 |
Neg logit contributions | |
Brown | -0.060 |
Later | -0.054 |
blankets | -0.053 |
walk | -0.053 |
Buzz | -0.053 |
Token | loud |
Token ID | 7812 |
Feature activation | +1.19e-02 |
Pos logit contributions | |
Brown | +0.027 |
Buzz | +0.025 |
| | +0.025 |
Ruby | +0.024 |
seek | +0.024 |
Neg logit contributions | |
happens | -0.027 |
ield | -0.026 |
happened | -0.026 |
Dan | -0.024 |
meant | -0.023 |
Token | sky |
Token ID | 6766 |
Feature activation | +1.57e-03 |
Pos logit contributions | |
happens | +0.010 |
ield | +0.009 |
happened | +0.009 |
Dan | +0.009 |
David | +0.008 |
Neg logit contributions | |
Brown | -0.007 |
Buzz | -0.007 |
seek | -0.007 |
Ruby | -0.007 |
Fl | -0.007 |
Token | and |
Token ID | 290 |
Feature activation | -1.04e-02 |
Pos logit contributions | |
happens | +0.004 |
ield | +0.003 |
happened | +0.003 |
oke | +0.003 |
Dan | +0.003 |
Neg logit contributions | |
Brown | -0.002 |
walk | -0.002 |
seek | -0.002 |
ide | -0.002 |
sleep | -0.002 |
Token | explained |
Token ID | 4893 |
Feature activation | +3.88e-02 |
Pos logit contributions | |
Brown | +0.020 |
Buzz | +0.019 |
seek | +0.019 |
Ruby | +0.019 |
Later | +0.019 |
Neg logit contributions | |
happens | -0.016 |
ield | -0.016 |
happened | -0.015 |
Dan | -0.014 |
meant | -0.013 |
Token | that |
Token ID | 326 |
Feature activation | +4.31e-02 |
Pos logit contributions | |
happens | +0.069 |
happened | +0.069 |
ield | +0.068 |
meant | +0.063 |
oke | +0.060 |
Neg logit contributions | |
Buzz | -0.053 |
Brown | -0.053 |
Bun | -0.052 |
cushion | -0.051 |
ide | -0.050 |
Token | the |
Token ID | 262 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
ield | +0.076 |
happens | +0.075 |
happened | +0.071 |
oke | +0.066 |
Dan | +0.064 |
Neg logit contributions | |
seek | -0.070 |
Brown | -0.069 |
Buzz | -0.066 |
lf | -0.066 |
Ruby | -0.064 |
Token | sound |
Token ID | 2128 |
Feature activation | +7.45e-02 |
Pos logit contributions | |
happens | +0.024 |
ield | +0.024 |
happened | +0.023 |
Dan | +0.022 |
David | +0.021 |
Neg logit contributions | |
Brown | -0.024 |
Buzz | -0.023 |
seek | -0.023 |
carpet | -0.022 |
Ruby | -0.022 |
Token | was |
Token ID | 373 |
Feature activation | +4.58e-02 |
Pos logit contributions | |
ield | +0.119 |
Dan | +0.117 |
fault | +0.110 |
David | +0.109 |
happens | +0.109 |
Neg logit contributions | |
lf | -0.122 |
Brown | -0.122 |
carpet | -0.116 |
seek | -0.116 |
walk | -0.116 |
Token | coming |
Token ID | 2406 |
Feature activation | +5.87e-03 |
Pos logit contributions | |
ield | +0.088 |
happens | +0.086 |
oke | +0.077 |
happened | +0.077 |
Dan | +0.074 |
Neg logit contributions | |
Brown | -0.068 |
seek | -0.066 |
sleep | -0.064 |
blankets | -0.063 |
carpet | -0.063 |
Token | from |
Token ID | 422 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
happens | +0.011 |
ield | +0.011 |
happened | +0.011 |
Dan | +0.010 |
meant | +0.010 |
Neg logit contributions | |
Brown | -0.009 |
Buzz | -0.008 |
seek | -0.008 |
Later | -0.008 |
Ruby | -0.008 |
Token | an |
Token ID | 281 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
Brown | +0.024 |
Buzz | +0.023 |
Ruby | +0.022 |
seek | +0.022 |
Later | +0.022 |
Neg logit contributions | |
happens | -0.025 |
ield | -0.024 |
happened | -0.024 |
Dan | -0.022 |
meant | -0.021 |
Token | airplane |
Token ID | 19401 |
Feature activation | -4.08e-03 |
Pos logit contributions | |
Brown | +0.018 |
Buzz | +0.017 |
seek | +0.016 |
Ruby | +0.016 |
Later | +0.016 |
Neg logit contributions | |
happens | -0.023 |
ield | -0.022 |
happened | -0.022 |
Dan | -0.020 |
meant | -0.019 |
Token | different |
Token ID | 1180 |
Feature activation | -1.81e-02 |
Pos logit contributions | |
Brown | +0.005 |
Buzz | +0.005 |
seek | +0.005 |
Ruby | +0.005 |
| | +0.005 |
Neg logit contributions | |
happens | -0.006 |
ield | -0.005 |
happened | -0.005 |
Dan | -0.005 |
meant | -0.005 |
Token | ideas |
Token ID | 4213 |
Feature activation | +1.75e-02 |
Pos logit contributions | |
Brown | +0.034 |
Buzz | +0.032 |
seek | +0.031 |
Ruby | +0.031 |
| | +0.030 |
Neg logit contributions | |
happens | -0.030 |
ield | -0.028 |
happened | -0.028 |
Dan | -0.026 |
meant | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | +2.49e-02 |
Pos logit contributions | |
happens | +0.035 |
ield | +0.034 |
happened | +0.032 |
Dan | +0.031 |
Well | +0.029 |
Neg logit contributions | |
Brown | -0.024 |
ide | -0.024 |
lf | -0.023 |
walk | -0.023 |
seek | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | -1.30e-02 |
Pos logit contributions | |
happens | +0.061 |
happened | +0.057 |
ield | +0.055 |
ier | +0.050 |
emon | +0.050 |
Neg logit contributions | |
Later | -0.028 |
While | -0.024 |
Hope | -0.023 |
Ell | -0.022 |
Brown | -0.022 |
Token |
|
Token ID | 198 |
Feature activation | -2.34e-02 |
Pos logit contributions | |
Brown | +0.018 |
Later | +0.017 |
Buzz | +0.017 |
Ruby | +0.016 |
Ell | +0.016 |
Neg logit contributions | |
happens | -0.028 |
ield | -0.027 |
happened | -0.026 |
meant | -0.024 |
Dan | -0.023 |
Token | Suddenly |
Token ID | 38582 |
Feature activation | +7.42e-02 |
Pos logit contributions | |
Later | +0.027 |
Brown | +0.025 |
While | +0.022 |
Ell | +0.022 |
Buzz | +0.022 |
Neg logit contributions | |
happens | -0.058 |
happened | -0.055 |
ield | -0.053 |
fault | -0.051 |
Dan | -0.049 |
Token | , |
Token ID | 11 |
Feature activation | +4.80e-02 |
Pos logit contributions | |
fault | +0.173 |
oke | +0.164 |
ield | +0.162 |
ses | +0.159 |
unes | +0.154 |
Neg logit contributions | |
walk | -0.071 |
seek | -0.068 |
ie | -0.066 |
carefully | -0.064 |
lf | -0.063 |
Token | Jenna |
Token ID | 45377 |
Feature activation | +2.01e-02 |
Pos logit contributions | |
happens | +0.083 |
oke | +0.082 |
fault | +0.082 |
ield | +0.080 |
unes | +0.075 |
Neg logit contributions | |
walk | -0.072 |
seek | -0.071 |
lf | -0.069 |
Later | -0.069 |
read | -0.067 |
Token | had |
Token ID | 550 |
Feature activation | +5.53e-02 |
Pos logit contributions | |
happens | +0.034 |
ield | +0.033 |
happened | +0.031 |
Dan | +0.029 |
Well | +0.028 |
Neg logit contributions | |
Brown | -0.035 |
Buzz | -0.033 |
lf | -0.033 |
seek | -0.032 |
Ruby | -0.032 |
Token | an |
Token ID | 281 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
ield | +0.085 |
oke | +0.084 |
happens | +0.084 |
Well | +0.075 |
ement | +0.071 |
Neg logit contributions | |
Brown | -0.096 |
seek | -0.092 |
Buzz | -0.092 |
unw | -0.091 |
read | -0.090 |
Token | idea |
Token ID | 2126 |
Feature activation | +3.09e-02 |
Pos logit contributions | |
happens | +0.020 |
happened | +0.018 |
Dan | +0.018 |
ield | +0.017 |
oke | +0.017 |
Neg logit contributions | |
sleep | -0.019 |
carpet | -0.019 |
lf | -0.019 |
Brown | -0.018 |
hind | -0.018 |
Token | 's |
Token ID | 338 |
Feature activation | +1.14e-02 |
Pos logit contributions | |
Brown | +0.033 |
Buzz | +0.031 |
seek | +0.030 |
Ruby | +0.030 |
| | +0.030 |
Neg logit contributions | |
happens | -0.033 |
ield | -0.031 |
happened | -0.031 |
Dan | -0.029 |
meant | -0.027 |
Token | eyes |
Token ID | 2951 |
Feature activation | +2.26e-02 |
Pos logit contributions | |
happens | +0.021 |
ield | +0.020 |
happened | +0.019 |
Dan | +0.018 |
meant | +0.017 |
Neg logit contributions | |
Brown | -0.020 |
Buzz | -0.019 |
seek | -0.018 |
Ruby | -0.018 |
Fl | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | +2.30e-02 |
Pos logit contributions | |
happens | +0.046 |
ield | +0.044 |
happened | +0.043 |
Dan | +0.041 |
oke | +0.039 |
Neg logit contributions | |
Brown | -0.032 |
walk | -0.030 |
Buzz | -0.030 |
ide | -0.029 |
Later | -0.029 |
Token |
|
Token ID | 198 |
Feature activation | -2.66e-02 |
Pos logit contributions | |
happens | +0.048 |
ield | +0.044 |
happened | +0.044 |
meant | +0.038 |
fault | +0.037 |
Neg logit contributions | |
Brown | -0.034 |
Later | -0.033 |
Buzz | -0.031 |
| | -0.031 |
Fl | -0.031 |
Token |
|
Token ID | 198 |
Feature activation | -3.14e-02 |
Pos logit contributions | |
Brown | +0.052 |
Buzz | +0.049 |
seek | +0.048 |
Mid | +0.047 |
Ruby | +0.047 |
Neg logit contributions | |
happens | -0.040 |
happened | -0.039 |
ield | -0.038 |
Dan | -0.036 |
David | -0.033 |
Token | Suddenly |
Token ID | 38582 |
Feature activation | +7.42e-02 |
Pos logit contributions | |
Brown | +0.038 |
throne | +0.037 |
| | +0.036 |
Buzz | +0.036 |
Ruby | +0.036 |
Neg logit contributions | |
happens | -0.067 |
ield | -0.066 |
happened | -0.065 |
Dan | -0.063 |
David | -0.060 |
Token | , |
Token ID | 11 |
Feature activation | +4.58e-02 |
Pos logit contributions | |
ield | +0.141 |
happens | +0.137 |
fault | +0.130 |
oke | +0.130 |
happened | +0.120 |
Neg logit contributions | |
Brown | -0.109 |
Bun | -0.106 |
Buzz | -0.105 |
seek | -0.104 |
walk | -0.101 |
Token | the |
Token ID | 262 |
Feature activation | +1.88e-02 |
Pos logit contributions | |
happens | +0.092 |
ield | +0.090 |
happened | +0.082 |
oke | +0.081 |
fault | +0.076 |
Neg logit contributions | |
Brown | -0.068 |
Buzz | -0.064 |
seek | -0.063 |
Fl | -0.060 |
Later | -0.060 |
Token | cat |
Token ID | 3797 |
Feature activation | +2.90e-02 |
Pos logit contributions | |
happens | +0.037 |
ield | +0.036 |
happened | +0.035 |
Dan | +0.033 |
David | +0.031 |
Neg logit contributions | |
Brown | -0.029 |
Buzz | -0.027 |
Ruby | -0.026 |
seek | -0.026 |
Fl | -0.026 |
Token | jumped |
Token ID | 11687 |
Feature activation | +1.70e-02 |
Pos logit contributions | |
happens | +0.046 |
ield | +0.044 |
happened | +0.043 |
Dan | +0.040 |
David | +0.037 |
Neg logit contributions | |
Brown | -0.055 |
Buzz | -0.053 |
Ruby | -0.051 |
seek | -0.051 |
Later | -0.050 |
Token | back |
Token ID | 736 |
Feature activation | -1.37e-02 |
Pos logit contributions | |
happens | +0.034 |
happened | +0.032 |
ield | +0.032 |
Dan | +0.029 |
oke | +0.028 |
Neg logit contributions | |
Brown | -0.026 |
Buzz | -0.025 |
Ruby | -0.024 |
Later | -0.024 |
seek | -0.024 |
Token | told |
Token ID | 1297 |
Feature activation | -4.43e-05 |
Pos logit contributions | |
happens | +0.001 |
ield | +0.001 |
happened | +0.001 |
Dan | +0.000 |
fault | +0.000 |
Neg logit contributions | |
Brown | -0.001 |
Buzz | -0.001 |
seek | -0.001 |
Ruby | -0.001 |
Later | -0.001 |
Token | her |
Token ID | 607 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
Brown | +0.000 |
seek | +0.000 |
Buzz | +0.000 |
| | +0.000 |
Later | +0.000 |
Neg logit contributions | |
happens | -0.000 |
happened | -0.000 |
ield | -0.000 |
Dan | -0.000 |
meant | -0.000 |
Token | , |
Token ID | 11 |
Feature activation | -1.95e-03 |
Pos logit contributions | |
happens | +0.019 |
ield | +0.019 |
happened | +0.018 |
Dan | +0.017 |
meant | +0.016 |
Neg logit contributions | |
Brown | -0.021 |
Buzz | -0.020 |
seek | -0.019 |
Ruby | -0.019 |
lf | -0.019 |
Token | " |
Token ID | 366 |
Feature activation | +1.60e-02 |
Pos logit contributions | |
Brown | +0.004 |
Buzz | +0.004 |
Later | +0.004 |
Fl | +0.004 |
seek | +0.004 |
Neg logit contributions | |
happens | -0.003 |
happened | -0.002 |
ield | -0.002 |
Dan | -0.002 |
meant | -0.002 |
Token | The |
Token ID | 464 |
Feature activation | +2.22e-02 |
Pos logit contributions | |
happens | +0.030 |
ield | +0.028 |
happened | +0.028 |
Dan | +0.026 |
David | +0.025 |
Neg logit contributions | |
Brown | -0.027 |
Buzz | -0.026 |
Ruby | -0.024 |
Fl | -0.024 |
Later | -0.024 |
Token | word |
Token ID | 1573 |
Feature activation | +7.41e-02 |
Pos logit contributions | |
happens | +0.041 |
ield | +0.039 |
happened | +0.039 |
Dan | +0.037 |
David | +0.034 |
Neg logit contributions | |
Brown | -0.037 |
Buzz | -0.035 |
seek | -0.033 |
Ruby | -0.033 |
Fl | -0.033 |
Token | is |
Token ID | 318 |
Feature activation | +6.32e-02 |
Pos logit contributions | |
happened | +0.093 |
happens | +0.093 |
Dan | +0.092 |
David | +0.088 |
ield | +0.086 |
Neg logit contributions | |
Brown | -0.146 |
shelf | -0.132 |
Fl | -0.131 |
cushion | -0.129 |
Hope | -0.129 |
Token | kind |
Token ID | 1611 |
Feature activation | +3.63e-02 |
Pos logit contributions | |
happens | +0.112 |
ield | +0.111 |
happened | +0.106 |
Dan | +0.101 |
ement | +0.100 |
Neg logit contributions | |
Brown | -0.097 |
seek | -0.087 |
Hope | -0.085 |
Buzz | -0.084 |
Fl | -0.084 |
Token | . |
Token ID | 13 |
Feature activation | +5.09e-02 |
Pos logit contributions | |
happens | +0.067 |
happened | +0.066 |
ield | +0.063 |
Dan | +0.062 |
David | +0.059 |
Neg logit contributions | |
Brown | -0.054 |
Buzz | -0.051 |
ide | -0.050 |
seek | -0.050 |
sleep | -0.049 |
Token | Kind |
Token ID | 14927 |
Feature activation | +4.76e-02 |
Pos logit contributions | |
happened | +0.072 |
happens | +0.069 |
ield | +0.058 |
Dan | +0.054 |
fault | +0.053 |
Neg logit contributions | |
Brown | -0.113 |
Buzz | -0.104 |
Fl | -0.102 |
Ruby | -0.101 |
Blue | -0.100 |
Token | ness |
Token ID | 1108 |
Feature activation | +4.25e-02 |
Pos logit contributions | |
happens | +0.072 |
happened | +0.071 |
Dan | +0.069 |
David | +0.064 |
ield | +0.063 |
Neg logit contributions | |
Brown | -0.087 |
Buzz | -0.079 |
sleep | -0.077 |
carpet | -0.077 |
walk | -0.076 |
Token | Lily |
Token ID | 20037 |
Feature activation | -5.36e-03 |
Pos logit contributions | |
happens | +0.012 |
happened | +0.012 |
ield | +0.011 |
meant | +0.010 |
Dan | +0.010 |
Neg logit contributions | |
Brown | -0.011 |
heads | -0.011 |
Buzz | -0.010 |
hind | -0.010 |
Later | -0.010 |
Token | knew |
Token ID | 2993 |
Feature activation | +2.61e-02 |
Pos logit contributions | |
Brown | +0.009 |
Buzz | +0.008 |
lf | +0.008 |
seek | +0.008 |
Ruby | +0.008 |
Neg logit contributions | |
ield | -0.010 |
happens | -0.010 |
happened | -0.009 |
Dan | -0.009 |
David | -0.008 |
Token | that |
Token ID | 326 |
Feature activation | +3.90e-02 |
Pos logit contributions | |
ield | +0.052 |
happens | +0.052 |
happened | +0.047 |
oke | +0.047 |
Dan | +0.044 |
Neg logit contributions | |
Brown | -0.035 |
seek | -0.032 |
Bun | -0.032 |
lf | -0.032 |
Later | -0.031 |
Token | even |
Token ID | 772 |
Feature activation | +1.83e-02 |
Pos logit contributions | |
ield | +0.080 |
happens | +0.078 |
happened | +0.072 |
Dan | +0.071 |
oke | +0.071 |
Neg logit contributions | |
Brown | -0.052 |
seek | -0.052 |
lf | -0.049 |
carpet | -0.047 |
sleep | -0.047 |
Token | the |
Token ID | 262 |
Feature activation | +1.51e-02 |
Pos logit contributions | |
happens | +0.037 |
ield | +0.036 |
happened | +0.035 |
Dan | +0.034 |
David | +0.032 |
Neg logit contributions | |
Brown | -0.027 |
Buzz | -0.025 |
seek | -0.024 |
Ruby | -0.024 |
Fl | -0.023 |
Token | sill |
Token ID | 49276 |
Feature activation | +7.41e-02 |
Pos logit contributions | |
happens | +0.031 |
ield | +0.030 |
happened | +0.029 |
Dan | +0.028 |
David | +0.026 |
Neg logit contributions | |
Brown | -0.022 |
seek | -0.020 |
Buzz | -0.020 |
carpet | -0.020 |
shelf | -0.019 |
Token | iest |
Token ID | 6386 |
Feature activation | +2.13e-02 |
Pos logit contributions | |
happens | +0.115 |
ield | +0.113 |
Dan | +0.106 |
gonna | +0.101 |
Where | +0.100 |
Neg logit contributions | |
lf | -0.132 |
ie | -0.127 |
ide | -0.122 |
carpet | -0.114 |
Brown | -0.113 |
Token | things |
Token ID | 1243 |
Feature activation | +4.21e-02 |
Pos logit contributions | |
happens | +0.047 |
ield | +0.045 |
happened | +0.043 |
Dan | +0.042 |
David | +0.039 |
Neg logit contributions | |
Brown | -0.027 |
seek | -0.026 |
carpet | -0.025 |
lf | -0.025 |
Buzz | -0.024 |
Token | could |
Token ID | 714 |
Feature activation | -4.62e-03 |
Pos logit contributions | |
ield | +0.075 |
happens | +0.073 |
Dan | +0.069 |
happened | +0.066 |
David | +0.063 |
Neg logit contributions | |
Brown | -0.066 |
walk | -0.062 |
seek | -0.062 |
Buzz | -0.062 |
Ruby | -0.061 |
Token | be |
Token ID | 307 |
Feature activation | +2.53e-02 |
Pos logit contributions | |
Brown | +0.007 |
Buzz | +0.007 |
Ruby | +0.006 |
Fl | +0.006 |
Later | +0.006 |
Neg logit contributions | |
happens | -0.009 |
ield | -0.009 |
happened | -0.009 |
Dan | -0.008 |
meant | -0.008 |
Token | turned |
Token ID | 2900 |
Feature activation | +3.51e-02 |
Pos logit contributions | |
ield | +0.047 |
happens | +0.046 |
Dan | +0.042 |
happened | +0.041 |
Well | +0.039 |
Neg logit contributions | |
Brown | -0.039 |
seek | -0.038 |
Buzz | -0.037 |
carpet | -0.035 |
Hope | -0.035 |
Token | who |
Token ID | 508 |
Feature activation | +5.52e-03 |
Pos logit contributions | |
Brown | +0.067 |
hind | +0.064 |
Elsa | +0.064 |
heads | +0.063 |
Hope | +0.063 |
Neg logit contributions | |
ield | -0.038 |
happened | -0.037 |
happens | -0.034 |
Dan | -0.033 |
: | -0.032 |
Token | liked |
Token ID | 8288 |
Feature activation | -6.08e-03 |
Pos logit contributions | |
happened | +0.008 |
happens | +0.008 |
ield | +0.008 |
meant | +0.007 |
It | +0.007 |
Neg logit contributions | |
Brown | -0.011 |
lf | -0.010 |
heads | -0.010 |
Later | -0.010 |
hind | -0.010 |
Token | to |
Token ID | 284 |
Feature activation | -5.59e-02 |
Pos logit contributions | |
Brown | +0.015 |
| | +0.014 |
Buzz | +0.014 |
lf | +0.014 |
Ruby | +0.014 |
Neg logit contributions | |
happens | -0.007 |
happened | -0.007 |
ield | -0.006 |
Dan | -0.005 |
meant | -0.005 |
Token | play |
Token ID | 711 |
Feature activation | -5.95e-02 |
Pos logit contributions | |
Brown | +0.089 |
ge | +0.081 |
heads | +0.078 |
Ell | +0.078 |
Later | +0.076 |
Neg logit contributions | |
happens | -0.106 |
happened | -0.104 |
ield | -0.102 |
It | -0.100 |
Dan | -0.099 |
Token | hide |
Token ID | 7808 |
Feature activation | -5.20e-02 |
Pos logit contributions | |
Brown | +0.126 |
ge | +0.124 |
heads | +0.117 |
| | +0.114 |
ppa | +0.111 |
Neg logit contributions | |
happens | -0.077 |
It | -0.075 |
happened | -0.074 |
ield | -0.071 |
at | -0.063 |
Token | and |
Token ID | 290 |
Feature activation | -1.03e-01 |
Pos logit contributions | |
Brown | +0.127 |
Hope | +0.115 |
Ell | +0.115 |
| | +0.114 |
Ruby | +0.114 |
Neg logit contributions | |
ed | -0.052 |
happens | -0.050 |
enough | -0.047 |
ield | -0.046 |
aw | -0.046 |
Token | seek |
Token ID | 5380 |
Feature activation | -5.59e-02 |
Pos logit contributions | |
ge | +0.127 |
ide | +0.126 |
Hope | +0.122 |
ppa | +0.121 |
Brown | +0.121 |
Neg logit contributions | |
happens | -0.185 |
happened | -0.178 |
J | -0.177 |
Dan | -0.176 |
Joe | -0.173 |
Token | . |
Token ID | 13 |
Feature activation | -3.93e-03 |
Pos logit contributions | |
Brown | +0.133 |
Mic | +0.124 |
| | +0.123 |
Hope | +0.122 |
Buzz | +0.119 |
Neg logit contributions | |
happens | -0.057 |
happened | -0.056 |
enough | -0.054 |
ield | -0.054 |
meant | -0.050 |
Token | One |
Token ID | 1881 |
Feature activation | +1.10e-02 |
Pos logit contributions | |
Brown | +0.009 |
heads | +0.009 |
ge | +0.009 |
Buzz | +0.009 |
lf | +0.009 |
Neg logit contributions | |
It | -0.004 |
ield | -0.004 |
happens | -0.004 |
Dan | -0.004 |
happened | -0.004 |
Token | day |
Token ID | 1110 |
Feature activation | +2.78e-02 |
Pos logit contributions | |
happened | +0.018 |
happens | +0.018 |
ield | +0.016 |
Dan | +0.016 |
meant | +0.015 |
Neg logit contributions | |
Brown | -0.021 |
Buzz | -0.019 |
heads | -0.019 |
Elsa | -0.019 |
Hope | -0.019 |
Token | , |
Token ID | 11 |
Feature activation | +1.40e-02 |
Pos logit contributions | |
happens | +0.045 |
ield | +0.045 |
happened | +0.040 |
Dan | +0.040 |
meant | +0.038 |
Neg logit contributions | |
Brown | -0.050 |
Buzz | -0.048 |
seek | -0.047 |
Fl | -0.047 |
Ruby | -0.046 |
Token | She |
Token ID | 1375 |
Feature activation | -2.39e-02 |
Pos logit contributions | |
Brown | +0.025 |
Buzz | +0.023 |
heads | +0.022 |
lf | +0.022 |
ge | +0.022 |
Neg logit contributions | |
happens | -0.012 |
happened | -0.011 |
It | -0.011 |
ield | -0.011 |
Dan | -0.010 |
Token | loved |
Token ID | 6151 |
Feature activation | -2.42e-02 |
Pos logit contributions | |
Brown | +0.048 |
Buzz | +0.047 |
ge | +0.047 |
heads | +0.047 |
ide | +0.047 |
Neg logit contributions | |
happened | -0.037 |
happens | -0.035 |
ield | -0.030 |
meant | -0.029 |
It | -0.027 |
Token | to |
Token ID | 284 |
Feature activation | -5.66e-02 |
Pos logit contributions | |
Brown | +0.057 |
Buzz | +0.056 |
Ruby | +0.055 |
| | +0.055 |
Fl | +0.054 |
Neg logit contributions | |
happened | -0.028 |
happens | -0.027 |
ield | -0.025 |
Dan | -0.023 |
meant | -0.021 |
Token | play |
Token ID | 711 |
Feature activation | -5.24e-02 |
Pos logit contributions | |
Brown | +0.071 |
Buzz | +0.063 |
Fl | +0.061 |
Ruby | +0.061 |
Later | +0.060 |
Neg logit contributions | |
happens | -0.128 |
happened | -0.128 |
ield | -0.124 |
Dan | -0.121 |
It | -0.117 |
Token | hide |
Token ID | 7808 |
Feature activation | -5.18e-02 |
Pos logit contributions | |
ge | +0.101 |
Brown | +0.097 |
| | +0.093 |
charged | +0.093 |
Hope | +0.090 |
Neg logit contributions | |
happened | -0.077 |
happens | -0.074 |
ield | -0.073 |
It | -0.072 |
at | -0.063 |
Token | and |
Token ID | 290 |
Feature activation | -1.02e-01 |
Pos logit contributions | |
Brown | +0.133 |
Hope | +0.122 |
| | +0.121 |
Buzz | +0.121 |
Ell | +0.120 |
Neg logit contributions | |
ed | -0.044 |
happens | -0.040 |
aw | -0.038 |
ield | -0.037 |
enough | -0.036 |
Token | seek |
Token ID | 5380 |
Feature activation | -5.69e-02 |
Pos logit contributions | |
Hope | +0.119 |
Brown | +0.116 |
ide | +0.113 |
ppa | +0.108 |
ind | +0.106 |
Neg logit contributions | |
happens | -0.195 |
happened | -0.194 |
meant | -0.184 |
Dan | -0.183 |
J | -0.180 |
Token | with |
Token ID | 351 |
Feature activation | -6.00e-02 |
Pos logit contributions | |
Brown | +0.119 |
| | +0.110 |
Buzz | +0.110 |
Mic | +0.109 |
Hope | +0.107 |
Neg logit contributions | |
happened | -0.077 |
happens | -0.076 |
enough | -0.072 |
ield | -0.072 |
meant | -0.071 |
Token | her |
Token ID | 607 |
Feature activation | -3.08e-02 |
Pos logit contributions | |
Brown | +0.113 |
| | +0.108 |
heads | +0.107 |
ge | +0.105 |
Buzz | +0.104 |
Neg logit contributions | |
happens | -0.093 |
Dan | -0.093 |
happened | -0.090 |
David | -0.086 |
Bill | -0.084 |
Token | friends |
Token ID | 2460 |
Feature activation | -4.38e-02 |
Pos logit contributions | |
Brown | +0.057 |
Buzz | +0.055 |
| | +0.055 |
Later | +0.054 |
heads | +0.054 |
Neg logit contributions | |
happened | -0.051 |
happens | -0.049 |
Dan | -0.045 |
ield | -0.044 |
meant | -0.043 |
Token | . |
Token ID | 13 |
Feature activation | -1.37e-02 |
Pos logit contributions | |
Brown | +0.086 |
Buzz | +0.082 |
| | +0.079 |
seek | +0.077 |
Later | +0.077 |
Neg logit contributions | |
happens | -0.066 |
ield | -0.065 |
happened | -0.064 |
Dan | -0.059 |
meant | -0.056 |
Token | Ben |
Token ID | 3932 |
Feature activation | +6.01e-03 |
Pos logit contributions | |
Brown | +0.020 |
Buzz | +0.019 |
Ruby | +0.017 |
Mid | +0.017 |
walk | +0.017 |
Neg logit contributions | |
happens | -0.040 |
ield | -0.040 |
happened | -0.037 |
oke | -0.036 |
Dan | -0.035 |
Token | liked |
Token ID | 8288 |
Feature activation | -1.52e-02 |
Pos logit contributions | |
happens | +0.010 |
ield | +0.009 |
happened | +0.009 |
Dan | +0.008 |
meant | +0.008 |
Neg logit contributions | |
Brown | -0.011 |
Buzz | -0.011 |
seek | -0.010 |
Ruby | -0.010 |
Later | -0.010 |
Token | to |
Token ID | 284 |
Feature activation | -5.67e-02 |
Pos logit contributions | |
Brown | +0.038 |
Buzz | +0.036 |
Ruby | +0.036 |
| | +0.036 |
Fl | +0.035 |
Neg logit contributions | |
happened | -0.016 |
happens | -0.016 |
ield | -0.015 |
Dan | -0.014 |
David | -0.012 |
Token | play |
Token ID | 711 |
Feature activation | -5.99e-02 |
Pos logit contributions | |
Brown | +0.063 |
Buzz | +0.054 |
Later | +0.054 |
Fl | +0.053 |
Ruby | +0.053 |
Neg logit contributions | |
happens | -0.135 |
happened | -0.133 |
ield | -0.132 |
Dan | -0.130 |
David | -0.124 |
Token | hide |
Token ID | 7808 |
Feature activation | -5.46e-02 |
Pos logit contributions | |
Brown | +0.130 |
| | +0.120 |
heads | +0.119 |
ge | +0.118 |
Hope | +0.115 |
Neg logit contributions | |
happens | -0.080 |
happened | -0.078 |
ield | -0.075 |
It | -0.074 |
at | -0.067 |
Token | and |
Token ID | 290 |
Feature activation | -1.02e-01 |
Pos logit contributions | |
Brown | +0.135 |
| | +0.126 |
Ell | +0.125 |
Hope | +0.124 |
Ruby | +0.122 |
Neg logit contributions | |
ed | -0.052 |
happens | -0.050 |
ield | -0.048 |
aw | -0.046 |
enough | -0.045 |
Token | seek |
Token ID | 5380 |
Feature activation | -6.12e-02 |
Pos logit contributions | |
Brown | +0.120 |
ind | +0.115 |
Hope | +0.113 |
ppa | +0.112 |
ivering | +0.107 |
Neg logit contributions | |
Dan | -0.198 |
happened | -0.197 |
happens | -0.194 |
J | -0.188 |
meant | -0.187 |
Token | in |
Token ID | 287 |
Feature activation | -4.76e-02 |
Pos logit contributions | |
Brown | +0.141 |
Mic | +0.130 |
| | +0.129 |
Buzz | +0.128 |
Hope | +0.125 |
Neg logit contributions | |
happens | -0.073 |
happened | -0.071 |
enough | -0.067 |
ield | -0.067 |
meant | -0.065 |
Token | the |
Token ID | 262 |
Feature activation | -3.29e-02 |
Pos logit contributions | |
Brown | +0.112 |
Buzz | +0.108 |
heads | +0.107 |
Tweet | +0.105 |
Mic | +0.103 |
Neg logit contributions | |
Dan | -0.055 |
David | -0.049 |
It | -0.046 |
Matt | -0.046 |
Bill | -0.045 |
Token | house |
Token ID | 2156 |
Feature activation | -4.80e-02 |
Pos logit contributions | |
Brown | +0.055 |
ge | +0.053 |
heads | +0.050 |
Buzz | +0.049 |
ide | +0.048 |
Neg logit contributions | |
happens | -0.062 |
happened | -0.058 |
ield | -0.056 |
Dan | -0.055 |
It | -0.053 |
Token | . |
Token ID | 13 |
Feature activation | -2.00e-03 |
Pos logit contributions | |
Brown | +0.105 |
Buzz | +0.098 |
| | +0.094 |
Mic | +0.094 |
Mid | +0.093 |
Neg logit contributions | |
happens | -0.065 |
ield | -0.063 |
happened | -0.061 |
Dan | -0.056 |
meant | -0.055 |
Token | She |
Token ID | 1375 |
Feature activation | -2.42e-02 |
Pos logit contributions | |
Brown | +0.030 |
Buzz | +0.028 |
heads | +0.028 |
lf | +0.028 |
ge | +0.027 |
Neg logit contributions | |
happens | -0.014 |
happened | -0.013 |
It | -0.013 |
ield | -0.013 |
Dan | -0.012 |
Token | loved |
Token ID | 6151 |
Feature activation | -2.74e-02 |
Pos logit contributions | |
Brown | +0.049 |
Buzz | +0.048 |
ge | +0.048 |
ide | +0.047 |
heads | +0.047 |
Neg logit contributions | |
happened | -0.037 |
happens | -0.035 |
ield | -0.030 |
meant | -0.029 |
It | -0.027 |
Token | to |
Token ID | 284 |
Feature activation | -5.60e-02 |
Pos logit contributions | |
Brown | +0.064 |
Buzz | +0.063 |
Ruby | +0.061 |
| | +0.061 |
Fl | +0.060 |
Neg logit contributions | |
happened | -0.032 |
happens | -0.031 |
ield | -0.029 |
Dan | -0.027 |
meant | -0.024 |
Token | play |
Token ID | 711 |
Feature activation | -5.22e-02 |
Pos logit contributions | |
Brown | +0.069 |
Buzz | +0.061 |
Later | +0.060 |
Ruby | +0.060 |
Fl | +0.060 |
Neg logit contributions | |
happens | -0.127 |
happened | -0.126 |
ield | -0.123 |
Dan | -0.121 |
David | -0.116 |
Token | hide |
Token ID | 7808 |
Feature activation | -5.31e-02 |
Pos logit contributions | |
ge | +0.101 |
Brown | +0.098 |
| | +0.095 |
charged | +0.094 |
Hope | +0.091 |
Neg logit contributions | |
happened | -0.073 |
happens | -0.071 |
ield | -0.071 |
It | -0.070 |
at | -0.062 |
Token | and |
Token ID | 290 |
Feature activation | -1.02e-01 |
Pos logit contributions | |
Brown | +0.135 |
Hope | +0.124 |
Buzz | +0.123 |
| | +0.123 |
Ell | +0.122 |
Neg logit contributions | |
ed | -0.047 |
happens | -0.042 |
aw | -0.039 |
enough | -0.039 |
ield | -0.038 |
Token | seek |
Token ID | 5380 |
Feature activation | -5.58e-02 |
Pos logit contributions | |
Hope | +0.118 |
Brown | +0.116 |
ide | +0.109 |
ppa | +0.108 |
ind | +0.107 |
Neg logit contributions | |
happened | -0.196 |
happens | -0.193 |
meant | -0.188 |
Dan | -0.185 |
J | -0.181 |
Token | with |
Token ID | 351 |
Feature activation | -6.18e-02 |
Pos logit contributions | |
Brown | +0.114 |
Buzz | +0.106 |
| | +0.105 |
Mic | +0.104 |
Hope | +0.102 |
Neg logit contributions | |
happened | -0.078 |
happens | -0.076 |
ield | -0.074 |
enough | -0.073 |
meant | -0.073 |
Token | her |
Token ID | 607 |
Feature activation | -3.11e-02 |
Pos logit contributions | |
Brown | +0.116 |
| | +0.111 |
heads | +0.109 |
Buzz | +0.108 |
ge | +0.106 |
Neg logit contributions | |
happens | -0.096 |
Dan | -0.096 |
happened | -0.094 |
David | -0.089 |
Bill | -0.087 |
Token | friends |
Token ID | 2460 |
Feature activation | -4.19e-02 |
Pos logit contributions | |
Brown | +0.056 |
Buzz | +0.054 |
| | +0.054 |
Later | +0.054 |
heads | +0.052 |
Neg logit contributions | |
happened | -0.053 |
happens | -0.050 |
Dan | -0.047 |
ield | -0.046 |
meant | -0.044 |
Token | . |
Token ID | 13 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
Brown | +0.082 |
Buzz | +0.078 |
| | +0.075 |
seek | +0.074 |
Ruby | +0.073 |
Neg logit contributions | |
happens | -0.064 |
ield | -0.063 |
happened | -0.062 |
Dan | -0.057 |
meant | -0.054 |
Token | They |
Token ID | 1119 |
Feature activation | -3.52e-02 |
Pos logit contributions | |
happens | +0.037 |
ield | +0.035 |
happened | +0.035 |
Dan | +0.031 |
meant | +0.030 |
Neg logit contributions | |
Brown | -0.037 |
Buzz | -0.034 |
Later | -0.033 |
Fl | -0.033 |
Ruby | -0.033 |
Token | decide |
Token ID | 5409 |
Feature activation | +3.92e-03 |
Pos logit contributions | |
Brown | +0.068 |
heads | +0.063 |
lf | +0.062 |
Later | +0.062 |
Buzz | +0.061 |
Neg logit contributions | |
happens | -0.057 |
Dan | -0.054 |
happened | -0.053 |
ield | -0.053 |
meant | -0.049 |
Token | to |
Token ID | 284 |
Feature activation | -6.80e-02 |
Pos logit contributions | |
happens | +0.004 |
happened | +0.004 |
ield | +0.004 |
Dan | +0.004 |
meant | +0.003 |
Neg logit contributions | |
Buzz | -0.009 |
Brown | -0.009 |
| | -0.009 |
Ruby | -0.009 |
ide | -0.009 |
Token | play |
Token ID | 711 |
Feature activation | -3.46e-02 |
Pos logit contributions | |
Brown | +0.099 |
Later | +0.098 |
Ruby | +0.096 |
Mid | +0.095 |
heads | +0.094 |
Neg logit contributions | |
ield | -0.128 |
Dan | -0.121 |
happens | -0.118 |
Bill | -0.118 |
happened | -0.115 |
Token | hide |
Token ID | 7808 |
Feature activation | -3.26e-02 |
Pos logit contributions | |
Brown | +0.057 |
Buzz | +0.051 |
Later | +0.049 |
lf | +0.049 |
Hope | +0.048 |
Neg logit contributions | |
happens | -0.066 |
happened | -0.064 |
ield | -0.063 |
Dan | -0.059 |
meant | -0.055 |
Token | and |
Token ID | 290 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
Brown | +0.063 |
Buzz | +0.058 |
Ruby | +0.056 |
| | +0.056 |
Hope | +0.056 |
Neg logit contributions | |
happens | -0.051 |
ield | -0.048 |
happened | -0.048 |
Dan | -0.047 |
ed | -0.043 |
Token | seek |
Token ID | 5380 |
Feature activation | -4.95e-02 |
Pos logit contributions | |
ide | +0.104 |
Brown | +0.101 |
ind | +0.096 |
Hope | +0.090 |
Mic | +0.085 |
Neg logit contributions | |
happens | -0.228 |
Dan | -0.225 |
happened | -0.222 |
Bill | -0.203 |
ield | -0.202 |
Token | again |
Token ID | 757 |
Feature activation | -4.14e-02 |
Pos logit contributions | |
Brown | +0.090 |
Buzz | +0.081 |
Mic | +0.080 |
Hope | +0.078 |
| | +0.078 |
Neg logit contributions | |
happens | -0.086 |
happened | -0.082 |
ield | -0.079 |
Dan | -0.075 |
meant | -0.072 |
Token | . |
Token ID | 13 |
Feature activation | +2.36e-02 |
Pos logit contributions | |
Brown | +0.082 |
Buzz | +0.076 |
Mid | +0.075 |
| | +0.074 |
Ruby | +0.074 |
Neg logit contributions | |
happens | -0.065 |
happened | -0.061 |
Dan | -0.060 |
ield | -0.058 |
meant | -0.053 |
Token | But |
Token ID | 887 |
Feature activation | +1.56e-02 |
Pos logit contributions | |
happens | +0.040 |
ield | +0.039 |
happened | +0.039 |
Dan | +0.034 |
meant | +0.033 |
Neg logit contributions | |
Brown | -0.043 |
Buzz | -0.040 |
| | -0.039 |
Later | -0.039 |
Ruby | -0.039 |
Token | this |
Token ID | 428 |
Feature activation | +3.71e-02 |
Pos logit contributions | |
happens | +0.029 |
ield | +0.029 |
happened | +0.028 |
Dan | +0.025 |
meant | +0.024 |
Neg logit contributions | |
Brown | -0.026 |
Buzz | -0.024 |
seek | -0.024 |
Fl | -0.023 |
Ruby | -0.023 |
Token | She |
Token ID | 1375 |
Feature activation | -1.85e-02 |
Pos logit contributions | |
Brown | +0.008 |
ge | +0.007 |
Buzz | +0.007 |
heads | +0.007 |
lf | +0.007 |
Neg logit contributions | |
happens | -0.005 |
It | -0.005 |
happened | -0.005 |
ield | -0.005 |
There | -0.005 |
Token | loved |
Token ID | 6151 |
Feature activation | -2.32e-02 |
Pos logit contributions | |
Brown | +0.038 |
Buzz | +0.036 |
ge | +0.036 |
heads | +0.035 |
ide | +0.035 |
Neg logit contributions | |
happened | -0.028 |
happens | -0.027 |
ield | -0.024 |
meant | -0.023 |
Dan | -0.021 |
Token | to |
Token ID | 284 |
Feature activation | -5.87e-02 |
Pos logit contributions | |
Brown | +0.057 |
Buzz | +0.055 |
Ruby | +0.054 |
| | +0.054 |
Fl | +0.053 |
Neg logit contributions | |
happened | -0.026 |
happens | -0.025 |
ield | -0.022 |
Dan | -0.020 |
meant | -0.019 |
Token | play |
Token ID | 711 |
Feature activation | -4.92e-02 |
Pos logit contributions | |
Brown | +0.080 |
Buzz | +0.072 |
Fl | +0.072 |
Ruby | +0.071 |
ge | +0.070 |
Neg logit contributions | |
happens | -0.127 |
happened | -0.127 |
ield | -0.122 |
Dan | -0.119 |
It | -0.117 |
Token | hide |
Token ID | 7808 |
Feature activation | -4.97e-02 |
Pos logit contributions | |
ge | +0.100 |
Brown | +0.096 |
charged | +0.092 |
| | +0.090 |
Buzz | +0.089 |
Neg logit contributions | |
happened | -0.069 |
happens | -0.067 |
It | -0.067 |
ield | -0.064 |
at | -0.057 |
Token | and |
Token ID | 290 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
Brown | +0.120 |
Hope | +0.110 |
Buzz | +0.109 |
| | +0.109 |
Ell | +0.108 |
Neg logit contributions | |
ed | -0.050 |
happens | -0.049 |
ield | -0.045 |
enough | -0.044 |
Dan | -0.043 |
Token | seek |
Token ID | 5380 |
Feature activation | -5.81e-02 |
Pos logit contributions | |
Hope | +0.119 |
Brown | +0.118 |
ide | +0.115 |
ppa | +0.111 |
ind | +0.110 |
Neg logit contributions | |
happened | -0.189 |
happens | -0.189 |
Dan | -0.177 |
meant | -0.176 |
It | -0.172 |
Token | with |
Token ID | 351 |
Feature activation | -6.07e-02 |
Pos logit contributions | |
Brown | +0.124 |
Mic | +0.114 |
Buzz | +0.113 |
| | +0.113 |
Hope | +0.111 |
Neg logit contributions | |
happens | -0.077 |
happened | -0.077 |
enough | -0.072 |
ield | -0.071 |
meant | -0.070 |
Token | her |
Token ID | 607 |
Feature activation | -2.90e-02 |
Pos logit contributions | |
Brown | +0.114 |
ge | +0.110 |
heads | +0.108 |
| | +0.107 |
Buzz | +0.102 |
Neg logit contributions | |
happens | -0.096 |
Dan | -0.095 |
happened | -0.091 |
David | -0.086 |
It | -0.086 |
Token | toys |
Token ID | 14958 |
Feature activation | -6.08e-02 |
Pos logit contributions | |
Brown | +0.053 |
Buzz | +0.052 |
Later | +0.052 |
| | +0.051 |
heads | +0.050 |
Neg logit contributions | |
happened | -0.048 |
happens | -0.046 |
Dan | -0.042 |
ield | -0.042 |
meant | -0.040 |
Token | . |
Token ID | 13 |
Feature activation | -1.15e-02 |
Pos logit contributions | |
Brown | +0.127 |
Buzz | +0.120 |
| | +0.116 |
Ruby | +0.114 |
Later | +0.113 |
Neg logit contributions | |
happens | -0.089 |
happened | -0.086 |
ield | -0.084 |
meant | -0.077 |
Dan | -0.075 |
Token | Betsy |
Token ID | 45775 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
happens | +0.015 |
ield | +0.015 |
happened | +0.015 |
It | +0.014 |
meant | +0.014 |
Neg logit contributions | |
Brown | -0.009 |
lf | -0.009 |
heads | -0.009 |
seek | -0.008 |
Buzz | -0.008 |
Token | liked |
Token ID | 8288 |
Feature activation | -9.17e-03 |
Pos logit contributions | |
Brown | +0.035 |
heads | +0.035 |
hind | +0.034 |
Buzz | +0.033 |
Later | +0.033 |
Neg logit contributions | |
happened | -0.024 |
happens | -0.023 |
meant | -0.021 |
ield | -0.020 |
Dan | -0.019 |
Token | to |
Token ID | 284 |
Feature activation | -5.72e-02 |
Pos logit contributions | |
Brown | +0.022 |
| | +0.021 |
Buzz | +0.021 |
Ruby | +0.021 |
heads | +0.021 |
Neg logit contributions | |
happens | -0.010 |
happened | -0.010 |
ield | -0.009 |
Dan | -0.008 |
meant | -0.007 |
Token | play |
Token ID | 711 |
Feature activation | -5.62e-02 |
Pos logit contributions | |
Brown | +0.086 |
ge | +0.078 |
Ell | +0.075 |
Later | +0.074 |
| | +0.073 |
Neg logit contributions | |
happens | -0.112 |
ield | -0.109 |
happened | -0.109 |
Dan | -0.105 |
It | -0.105 |
Token | hide |
Token ID | 7808 |
Feature activation | -5.17e-02 |
Pos logit contributions | |
ge | +0.119 |
Brown | +0.116 |
heads | +0.109 |
| | +0.107 |
ppa | +0.106 |
Neg logit contributions | |
happened | -0.074 |
happens | -0.074 |
It | -0.072 |
ield | -0.067 |
at | -0.060 |
Token | and |
Token ID | 290 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
Brown | +0.127 |
| | +0.115 |
Hope | +0.115 |
Ell | +0.115 |
Elsa | +0.112 |
Neg logit contributions | |
ed | -0.050 |
happens | -0.047 |
aw | -0.044 |
enough | -0.043 |
ield | -0.041 |
Token | seek |
Token ID | 5380 |
Feature activation | -5.59e-02 |
Pos logit contributions | |
Brown | +0.122 |
Hope | +0.122 |
ge | +0.118 |
ide | +0.118 |
ppa | +0.117 |
Neg logit contributions | |
happens | -0.185 |
happened | -0.176 |
Dan | -0.175 |
It | -0.170 |
J | -0.168 |
Token | with |
Token ID | 351 |
Feature activation | -5.78e-02 |
Pos logit contributions | |
Brown | +0.124 |
Mic | +0.115 |
| | +0.113 |
Hope | +0.112 |
Buzz | +0.109 |
Neg logit contributions | |
happened | -0.070 |
happens | -0.069 |
enough | -0.064 |
meant | -0.062 |
ield | -0.061 |
Token | her |
Token ID | 607 |
Feature activation | -2.75e-02 |
Pos logit contributions | |
Brown | +0.111 |
heads | +0.108 |
ge | +0.108 |
| | +0.105 |
Later | +0.100 |
Neg logit contributions | |
Dan | -0.088 |
happens | -0.085 |
happened | -0.081 |
It | -0.081 |
Bill | -0.080 |
Token | friends |
Token ID | 2460 |
Feature activation | -4.50e-02 |
Pos logit contributions | |
Brown | +0.051 |
| | +0.050 |
Later | +0.050 |
Buzz | +0.048 |
heads | +0.048 |
Neg logit contributions | |
happened | -0.043 |
happens | -0.043 |
ield | -0.040 |
Dan | -0.039 |
meant | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | -5.31e-03 |
Pos logit contributions | |
Brown | +0.092 |
Buzz | +0.087 |
| | +0.084 |
Later | +0.083 |
heads | +0.082 |
Neg logit contributions | |
happens | -0.065 |
ield | -0.063 |
happened | -0.062 |
Dan | -0.057 |
meant | -0.055 |
Token | bunny |
Token ID | 44915 |
Feature activation | -7.20e-03 |
Pos logit contributions | |
ge | +0.003 |
heads | +0.002 |
| | +0.002 |
Brown | +0.002 |
ide | +0.002 |
Neg logit contributions | |
happens | -0.003 |
It | -0.003 |
meant | -0.003 |
happened | -0.003 |
ield | -0.003 |
Token | liked |
Token ID | 8288 |
Feature activation | -5.09e-03 |
Pos logit contributions | |
Brown | +0.016 |
heads | +0.015 |
ge | +0.015 |
Later | +0.014 |
Buzz | +0.014 |
Neg logit contributions | |
happens | -0.010 |
happened | -0.010 |
ield | -0.008 |
meant | -0.008 |
Dan | -0.008 |
Token | to |
Token ID | 284 |
Feature activation | -5.65e-02 |
Pos logit contributions | |
Brown | +0.013 |
| | +0.012 |
Buzz | +0.012 |
Ruby | +0.012 |
heads | +0.012 |
Neg logit contributions | |
happens | -0.006 |
happened | -0.005 |
ield | -0.005 |
Dan | -0.004 |
meant | -0.004 |
Token | play |
Token ID | 711 |
Feature activation | -5.45e-02 |
Pos logit contributions | |
Brown | +0.093 |
ge | +0.089 |
Ell | +0.081 |
heads | +0.080 |
Hope | +0.080 |
Neg logit contributions | |
happens | -0.101 |
It | -0.100 |
happened | -0.098 |
ield | -0.097 |
Dan | -0.096 |
Token | hide |
Token ID | 7808 |
Feature activation | -4.70e-02 |
Pos logit contributions | |
ge | +0.115 |
Brown | +0.115 |
heads | +0.107 |
ppa | +0.105 |
| | +0.103 |
Neg logit contributions | |
It | -0.071 |
happens | -0.071 |
happened | -0.068 |
ield | -0.066 |
at | -0.058 |
Token | and |
Token ID | 290 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
Brown | +0.114 |
Hope | +0.103 |
| | +0.103 |
Ell | +0.103 |
Ruby | +0.100 |
Neg logit contributions | |
happens | -0.049 |
ed | -0.048 |
ield | -0.044 |
enough | -0.044 |
aw | -0.043 |
Token | seek |
Token ID | 5380 |
Feature activation | -5.37e-02 |
Pos logit contributions | |
ppa | +0.122 |
ide | +0.122 |
Hope | +0.122 |
Brown | +0.121 |
stret | +0.120 |
Neg logit contributions | |
happens | -0.185 |
It | -0.177 |
happened | -0.176 |
J | -0.170 |
Dan | -0.169 |
Token | with |
Token ID | 351 |
Feature activation | -4.81e-02 |
Pos logit contributions | |
Brown | +0.123 |
Mic | +0.114 |
| | +0.112 |
Hope | +0.112 |
Buzz | +0.109 |
Neg logit contributions | |
happens | -0.064 |
happened | -0.062 |
enough | -0.058 |
ield | -0.058 |
meant | -0.057 |
Token | his |
Token ID | 465 |
Feature activation | -2.10e-02 |
Pos logit contributions | |
heads | +0.095 |
ge | +0.095 |
Brown | +0.092 |
| | +0.090 |
ital | +0.086 |
Neg logit contributions | |
happens | -0.071 |
It | -0.071 |
Dan | -0.070 |
Bill | -0.066 |
happened | -0.065 |
Token | friends |
Token ID | 2460 |
Feature activation | -4.30e-02 |
Pos logit contributions | |
Brown | +0.040 |
| | +0.039 |
Later | +0.038 |
heads | +0.038 |
ge | +0.037 |
Neg logit contributions | |
happened | -0.033 |
happens | -0.032 |
ield | -0.031 |
Dan | -0.029 |
meant | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | +3.10e-04 |
Pos logit contributions | |
Brown | +0.091 |
Buzz | +0.085 |
| | +0.083 |
heads | +0.082 |
Later | +0.081 |
Neg logit contributions | |
happens | -0.059 |
ield | -0.058 |
happened | -0.056 |
Dan | -0.052 |
meant | -0.049 |
Token | Ben |
Token ID | 3932 |
Feature activation | +8.71e-03 |
Pos logit contributions | |
Brown | +0.028 |
Buzz | +0.026 |
walk | +0.025 |
Hope | +0.024 |
seek | +0.024 |
Neg logit contributions | |
happens | -0.050 |
ield | -0.048 |
happened | -0.046 |
oke | -0.044 |
Dan | -0.042 |
Token | like |
Token ID | 588 |
Feature activation | -5.47e-03 |
Pos logit contributions | |
happens | +0.014 |
ield | +0.013 |
happened | +0.013 |
Dan | +0.012 |
meant | +0.011 |
Neg logit contributions | |
Brown | -0.016 |
Buzz | -0.016 |
seek | -0.015 |
| | -0.015 |
Ruby | -0.015 |
Token | to |
Token ID | 284 |
Feature activation | -4.69e-02 |
Pos logit contributions | |
Brown | +0.012 |
Buzz | +0.012 |
Later | +0.011 |
Ruby | +0.011 |
| | +0.011 |
Neg logit contributions | |
happens | -0.007 |
happened | -0.007 |
ield | -0.007 |
Dan | -0.006 |
meant | -0.005 |
Token | play |
Token ID | 711 |
Feature activation | -5.67e-02 |
Pos logit contributions | |
Brown | +0.040 |
Buzz | +0.034 |
Ruby | +0.031 |
Later | +0.031 |
Fl | +0.031 |
Neg logit contributions | |
happens | -0.125 |
ield | -0.122 |
happened | -0.121 |
Dan | -0.116 |
meant | -0.111 |
Token | hide |
Token ID | 7808 |
Feature activation | -4.73e-02 |
Pos logit contributions | |
Brown | +0.113 |
heads | +0.104 |
ge | +0.102 |
| | +0.101 |
Hope | +0.099 |
Neg logit contributions | |
happens | -0.086 |
happened | -0.085 |
ield | -0.082 |
It | -0.075 |
Dan | -0.073 |
Token | and |
Token ID | 290 |
Feature activation | -1.00e-01 |
Pos logit contributions | |
Brown | +0.105 |
Ruby | +0.095 |
| | +0.095 |
Hope | +0.095 |
Buzz | +0.095 |
Neg logit contributions | |
happens | -0.059 |
ield | -0.056 |
happened | -0.055 |
ed | -0.054 |
Dan | -0.052 |
Token | seek |
Token ID | 5380 |
Feature activation | -5.69e-02 |
Pos logit contributions | |
Brown | +0.102 |
Hope | +0.094 |
ppa | +0.086 |
ind | +0.086 |
Elsa | +0.085 |
Neg logit contributions | |
Dan | -0.222 |
happened | -0.222 |
happens | -0.215 |
Joe | -0.214 |
Bill | -0.208 |
Token | in |
Token ID | 287 |
Feature activation | -4.34e-02 |
Pos logit contributions | |
Brown | +0.122 |
Buzz | +0.110 |
Mic | +0.110 |
| | +0.110 |
Hope | +0.107 |
Neg logit contributions | |
happens | -0.077 |
happened | -0.077 |
ield | -0.072 |
enough | -0.069 |
meant | -0.068 |
Token | the |
Token ID | 262 |
Feature activation | -3.29e-02 |
Pos logit contributions | |
Brown | +0.096 |
Buzz | +0.091 |
heads | +0.088 |
Tweet | +0.088 |
Flo | +0.086 |
Neg logit contributions | |
Dan | -0.057 |
David | -0.054 |
happens | -0.052 |
happened | -0.050 |
Bill | -0.050 |
Token | park |
Token ID | 3952 |
Feature activation | -5.38e-02 |
Pos logit contributions | |
Brown | +0.055 |
ge | +0.052 |
heads | +0.051 |
Buzz | +0.050 |
ide | +0.048 |
Neg logit contributions | |
happens | -0.061 |
happened | -0.059 |
ield | -0.056 |
Dan | -0.055 |
David | -0.053 |
Token | . |
Token ID | 13 |
Feature activation | +1.34e-03 |
Pos logit contributions | |
Brown | +0.111 |
Buzz | +0.101 |
Ruby | +0.098 |
| | +0.097 |
Mid | +0.097 |
Neg logit contributions | |
happens | -0.079 |
ield | -0.077 |
happened | -0.076 |
Dan | -0.072 |
meant | -0.066 |
Token | they |
Token ID | 484 |
Feature activation | -5.47e-03 |
Pos logit contributions | |
ield | +0.052 |
happens | +0.051 |
oke | +0.046 |
happened | +0.043 |
fault | +0.043 |
Neg logit contributions | |
While | -0.038 |
seek | -0.038 |
lf | -0.037 |
games | -0.037 |
Later | -0.037 |
Token | decide |
Token ID | 5409 |
Feature activation | +1.33e-02 |
Pos logit contributions | |
Brown | +0.009 |
Buzz | +0.008 |
Fl | +0.007 |
Ruby | +0.007 |
Later | +0.007 |
Neg logit contributions | |
happens | -0.011 |
ield | -0.010 |
happened | -0.010 |
Dan | -0.010 |
meant | -0.009 |
Token | to |
Token ID | 284 |
Feature activation | -6.61e-02 |
Pos logit contributions | |
happens | +0.019 |
happened | +0.018 |
ield | +0.018 |
Dan | +0.016 |
meant | +0.015 |
Neg logit contributions | |
Brown | -0.028 |
Buzz | -0.027 |
Ruby | -0.026 |
| | -0.026 |
Fl | -0.026 |
Token | play |
Token ID | 711 |
Feature activation | -3.74e-02 |
Pos logit contributions | |
Brown | +0.098 |
Buzz | +0.092 |
Fl | +0.090 |
Later | +0.090 |
Ruby | +0.089 |
Neg logit contributions | |
happens | -0.133 |
ield | -0.132 |
happened | -0.126 |
Dan | -0.119 |
meant | -0.117 |
Token | hide |
Token ID | 7808 |
Feature activation | -4.03e-02 |
Pos logit contributions | |
Brown | +0.079 |
Buzz | +0.073 |
Bee | +0.070 |
Mic | +0.070 |
Hope | +0.069 |
Neg logit contributions | |
happens | -0.055 |
happened | -0.053 |
ield | -0.051 |
It | -0.046 |
Dan | -0.045 |
Token | and |
Token ID | 290 |
Feature activation | -9.96e-02 |
Pos logit contributions | |
Brown | +0.084 |
Buzz | +0.078 |
Ell | +0.076 |
Mic | +0.075 |
Ruby | +0.075 |
Neg logit contributions | |
happens | -0.058 |
ield | -0.054 |
happened | -0.053 |
Dan | -0.051 |
ed | -0.050 |
Token | seek |
Token ID | 5380 |
Feature activation | -4.75e-02 |
Pos logit contributions | |
Brown | +0.101 |
ide | +0.101 |
Hope | +0.090 |
ind | +0.090 |
ppa | +0.089 |
Neg logit contributions | |
happens | -0.226 |
happened | -0.210 |
Dan | -0.203 |
meant | -0.194 |
ield | -0.193 |
Token | . |
Token ID | 13 |
Feature activation | +1.99e-02 |
Pos logit contributions | |
Brown | +0.098 |
Buzz | +0.091 |
Mic | +0.088 |
| | +0.087 |
Tweet | +0.086 |
Neg logit contributions | |
happens | -0.072 |
happened | -0.069 |
ield | -0.066 |
meant | -0.061 |
Dan | -0.060 |
Token | Tom |
Token ID | 4186 |
Feature activation | -8.40e-03 |
Pos logit contributions | |
happens | +0.036 |
happened | +0.035 |
ield | +0.033 |
meant | +0.029 |
oke | +0.029 |
Neg logit contributions | |
Later | -0.031 |
Brown | -0.031 |
shelf | -0.030 |
Buzz | -0.030 |
| | -0.030 |
Token | is |
Token ID | 318 |
Feature activation | +4.78e-03 |
Pos logit contributions | |
Brown | +0.018 |
Buzz | +0.017 |
seek | +0.016 |
Ruby | +0.016 |
Later | +0.016 |
Neg logit contributions | |
happens | -0.012 |
happened | -0.011 |
ield | -0.011 |
Dan | -0.010 |
meant | -0.009 |
Token | the |
Token ID | 262 |
Feature activation | -6.83e-04 |
Pos logit contributions | |
happens | +0.009 |
ield | +0.009 |
happened | +0.008 |
Dan | +0.007 |
oke | +0.007 |
Neg logit contributions | |
Brown | -0.008 |
seek | -0.008 |
Buzz | -0.008 |
Ruby | -0.007 |
| | -0.007 |
Token | who |
Token ID | 508 |
Feature activation | +1.03e-03 |
Pos logit contributions | |
Brown | +0.023 |
Buzz | +0.023 |
Ruby | +0.023 |
hind | +0.023 |
Mid | +0.022 |
Neg logit contributions | |
ield | -0.022 |
happened | -0.021 |
Dan | -0.020 |
happens | -0.020 |
meant | -0.019 |
Token | loved |
Token ID | 6151 |
Feature activation | -1.31e-02 |
Pos logit contributions | |
happens | +0.002 |
happened | +0.002 |
ield | +0.002 |
meant | +0.001 |
Dan | +0.001 |
Neg logit contributions | |
Brown | -0.002 |
Buzz | -0.002 |
Ruby | -0.002 |
Later | -0.002 |
heads | -0.002 |
Token | to |
Token ID | 284 |
Feature activation | -4.91e-02 |
Pos logit contributions | |
Brown | +0.029 |
Buzz | +0.028 |
Ruby | +0.028 |
| | +0.027 |
Fl | +0.027 |
Neg logit contributions | |
happens | -0.017 |
happened | -0.017 |
ield | -0.016 |
Dan | -0.015 |
David | -0.013 |
Token | play |
Token ID | 711 |
Feature activation | -5.01e-02 |
Pos logit contributions | |
Brown | +0.068 |
Ruby | +0.065 |
Fl | +0.063 |
Later | +0.062 |
Buzz | +0.061 |
Neg logit contributions | |
ield | -0.102 |
happened | -0.101 |
happens | -0.101 |
Dan | -0.099 |
David | -0.094 |
Token | hide |
Token ID | 7808 |
Feature activation | -5.00e-02 |
Pos logit contributions | |
Brown | +0.094 |
ge | +0.090 |
| | +0.088 |
Bee | +0.086 |
Hope | +0.086 |
Neg logit contributions | |
ield | -0.076 |
happens | -0.073 |
happened | -0.073 |
It | -0.070 |
Dan | -0.065 |
Token | and |
Token ID | 290 |
Feature activation | -9.94e-02 |
Pos logit contributions | |
Brown | +0.123 |
Hope | +0.116 |
Ruby | +0.116 |
Ell | +0.115 |
Elsa | +0.112 |
Neg logit contributions | |
ed | -0.046 |
happens | -0.044 |
ield | -0.043 |
aw | -0.042 |
enough | -0.041 |
Token | seek |
Token ID | 5380 |
Feature activation | -5.51e-02 |
Pos logit contributions | |
Hope | +0.115 |
ide | +0.112 |
Brown | +0.111 |
ppa | +0.107 |
Elsa | +0.104 |
Neg logit contributions | |
happened | -0.190 |
happens | -0.189 |
Dan | -0.184 |
J | -0.183 |
Joe | -0.179 |
Token | . |
Token ID | 13 |
Feature activation | +2.01e-03 |
Pos logit contributions | |
Brown | +0.129 |
Mic | +0.121 |
Buzz | +0.120 |
| | +0.119 |
Hope | +0.119 |
Neg logit contributions | |
happened | -0.059 |
happens | -0.059 |
ield | -0.057 |
enough | -0.056 |
meant | -0.052 |
Token | One |
Token ID | 1881 |
Feature activation | +1.54e-02 |
Pos logit contributions | |
happens | +0.003 |
ield | +0.003 |
happened | +0.003 |
Dan | +0.002 |
meant | +0.002 |
Neg logit contributions | |
Brown | -0.004 |
Buzz | -0.004 |
Ruby | -0.004 |
seek | -0.004 |
| | -0.004 |
Token | day |
Token ID | 1110 |
Feature activation | +3.34e-02 |
Pos logit contributions | |
happens | +0.028 |
happened | +0.027 |
ield | +0.027 |
Dan | +0.025 |
meant | +0.024 |
Neg logit contributions | |
Brown | -0.026 |
Buzz | -0.025 |
Ruby | -0.024 |
seek | -0.023 |
| | -0.023 |
Token | , |
Token ID | 11 |
Feature activation | +2.09e-02 |
Pos logit contributions | |
ield | +0.061 |
happens | +0.061 |
fault | +0.056 |
Dan | +0.054 |
meant | +0.054 |
Neg logit contributions | |
Brown | -0.050 |
Buzz | -0.049 |
seek | -0.048 |
walk | -0.047 |
Fl | -0.047 |
Token | my |
Token ID | 1820 |
Feature activation | -1.74e-03 |
Pos logit contributions | |
Brown | +0.005 |
Buzz | +0.005 |
Ruby | +0.005 |
seek | +0.005 |
heads | +0.005 |
Neg logit contributions | |
happened | -0.008 |
happens | -0.008 |
ield | -0.007 |
meant | -0.007 |
Dan | -0.007 |
Token | loved |
Token ID | 6151 |
Feature activation | -1.90e-02 |
Pos logit contributions | |
Brown | +0.003 |
Buzz | +0.003 |
Ruby | +0.003 |
heads | +0.003 |
seek | +0.003 |
Neg logit contributions | |
happens | -0.003 |
happened | -0.003 |
ield | -0.003 |
Dan | -0.002 |
meant | -0.002 |
Token | to |
Token ID | 284 |
Feature activation | -5.49e-02 |
Pos logit contributions | |
Brown | +0.043 |
Buzz | +0.043 |
Ruby | +0.042 |
| | +0.041 |
Fl | +0.041 |
Neg logit contributions | |
happens | -0.023 |
happened | -0.023 |
ield | -0.022 |
Dan | -0.020 |
David | -0.018 |
Token | play |
Token ID | 711 |
Feature activation | -4.96e-02 |
Pos logit contributions | |
Brown | +0.067 |
Ruby | +0.061 |
Buzz | +0.060 |
Fl | +0.060 |
Later | +0.058 |
Neg logit contributions | |
happened | -0.125 |
happens | -0.124 |
ield | -0.122 |
Dan | -0.120 |
David | -0.115 |
Token | hide |
Token ID | 7808 |
Feature activation | -4.97e-02 |
Pos logit contributions | |
Brown | +0.090 |
ge | +0.088 |
Hope | +0.086 |
Bee | +0.085 |
| | +0.084 |
Neg logit contributions | |
happened | -0.076 |
happens | -0.074 |
ield | -0.074 |
It | -0.071 |
Dan | -0.064 |
Token | and |
Token ID | 290 |
Feature activation | -9.93e-02 |
Pos logit contributions | |
Brown | +0.123 |
Hope | +0.118 |
Ruby | +0.116 |
Ell | +0.116 |
Buzz | +0.114 |
Neg logit contributions | |
ed | -0.046 |
happens | -0.043 |
ield | -0.041 |
aw | -0.040 |
Dan | -0.039 |
Token | seek |
Token ID | 5380 |
Feature activation | -5.57e-02 |
Pos logit contributions | |
Hope | +0.115 |
Brown | +0.108 |
ide | +0.104 |
Elsa | +0.101 |
ind | +0.101 |
Neg logit contributions | |
happened | -0.197 |
happens | -0.197 |
Dan | -0.187 |
meant | -0.185 |
J | -0.183 |
Token | with |
Token ID | 351 |
Feature activation | -4.95e-02 |
Pos logit contributions | |
Brown | +0.117 |
Buzz | +0.109 |
| | +0.108 |
Mic | +0.108 |
Hope | +0.108 |
Neg logit contributions | |
happened | -0.075 |
happens | -0.074 |
ield | -0.070 |
enough | -0.070 |
meant | -0.068 |
Token | his |
Token ID | 465 |
Feature activation | -2.06e-02 |
Pos logit contributions | |
Brown | +0.085 |
| | +0.083 |
heads | +0.082 |
Buzz | +0.081 |
seek | +0.078 |
Neg logit contributions | |
happens | -0.084 |
Dan | -0.083 |
happened | -0.080 |
ield | -0.079 |
David | -0.078 |
Token | friends |
Token ID | 2460 |
Feature activation | -4.18e-02 |
Pos logit contributions | |
Brown | +0.035 |
Buzz | +0.033 |
| | +0.032 |
Later | +0.032 |
heads | +0.031 |
Neg logit contributions | |
happened | -0.037 |
happens | -0.037 |
ield | -0.035 |
Dan | -0.034 |
meant | -0.031 |
Token | . |
Token ID | 13 |
Feature activation | +1.40e-03 |
Pos logit contributions | |
Brown | +0.080 |
Buzz | +0.076 |
| | +0.073 |
Ruby | +0.072 |
seek | +0.072 |
Neg logit contributions | |
happens | -0.067 |
ield | -0.065 |
happened | -0.064 |
Dan | -0.059 |
meant | -0.056 |
Token | Benny |
Token ID | 44275 |
Feature activation | -6.55e-03 |
Pos logit contributions | |
happens | +0.010 |
ield | +0.009 |
happened | +0.009 |
It | +0.009 |
Dan | +0.009 |
Neg logit contributions | |
Brown | -0.007 |
heads | -0.006 |
lf | -0.006 |
Buzz | -0.006 |
Hope | -0.006 |
Token | loved |
Token ID | 6151 |
Feature activation | -2.06e-02 |
Pos logit contributions | |
Brown | +0.013 |
heads | +0.013 |
Buzz | +0.013 |
hind | +0.013 |
throne | +0.012 |
Neg logit contributions | |
happened | -0.009 |
happens | -0.009 |
ield | -0.008 |
meant | -0.008 |
Dan | -0.008 |
Token | to |
Token ID | 284 |
Feature activation | -5.35e-02 |
Pos logit contributions | |
Brown | +0.049 |
| | +0.046 |
Buzz | +0.046 |
Ruby | +0.046 |
ide | +0.046 |
Neg logit contributions | |
happens | -0.024 |
happened | -0.024 |
ield | -0.021 |
Dan | -0.019 |
meant | -0.018 |
Token | play |
Token ID | 711 |
Feature activation | -5.40e-02 |
Pos logit contributions | |
Brown | +0.080 |
ge | +0.071 |
heads | +0.071 |
Ell | +0.069 |
| | +0.067 |
Neg logit contributions | |
happens | -0.108 |
happened | -0.107 |
ield | -0.103 |
Dan | -0.102 |
It | -0.099 |
Token | hide |
Token ID | 7808 |
Feature activation | -4.76e-02 |
Pos logit contributions | |
Brown | +0.111 |
ge | +0.110 |
heads | +0.104 |
| | +0.102 |
Hope | +0.100 |
Neg logit contributions | |
happened | -0.070 |
happens | -0.070 |
It | -0.068 |
ield | -0.066 |
ed | -0.058 |
Token | and |
Token ID | 290 |
Feature activation | -9.90e-02 |
Pos logit contributions | |
Brown | +0.113 |
Hope | +0.104 |
Ell | +0.102 |
| | +0.102 |
Ruby | +0.101 |
Neg logit contributions | |
ed | -0.050 |
happens | -0.047 |
ield | -0.045 |
enough | -0.045 |
Dan | -0.043 |
Token | seek |
Token ID | 5380 |
Feature activation | -5.39e-02 |
Pos logit contributions | |
Hope | +0.115 |
Brown | +0.115 |
ppa | +0.112 |
ide | +0.112 |
ind | +0.108 |
Neg logit contributions | |
happens | -0.184 |
happened | -0.178 |
Dan | -0.176 |
Joe | -0.175 |
It | -0.172 |
Token | with |
Token ID | 351 |
Feature activation | -5.28e-02 |
Pos logit contributions | |
Brown | +0.112 |
| | +0.102 |
Mic | +0.101 |
Hope | +0.101 |
Buzz | +0.100 |
Neg logit contributions | |
happened | -0.074 |
happens | -0.073 |
ield | -0.069 |
enough | -0.068 |
meant | -0.067 |
Token | his |
Token ID | 465 |
Feature activation | -2.26e-02 |
Pos logit contributions | |
heads | +0.101 |
Brown | +0.101 |
| | +0.097 |
ge | +0.095 |
Hope | +0.091 |
Neg logit contributions | |
Dan | -0.079 |
happens | -0.077 |
happened | -0.074 |
It | -0.073 |
Bill | -0.073 |
Token | friends |
Token ID | 2460 |
Feature activation | -4.06e-02 |
Pos logit contributions | |
Brown | +0.041 |
| | +0.039 |
Later | +0.038 |
heads | +0.038 |
Buzz | +0.038 |
Neg logit contributions | |
happened | -0.037 |
happens | -0.037 |
ield | -0.035 |
Dan | -0.033 |
meant | -0.031 |
Token | . |
Token ID | 13 |
Feature activation | -5.28e-03 |
Pos logit contributions | |
Brown | +0.079 |
Buzz | +0.075 |
| | +0.073 |
seek | +0.071 |
Later | +0.071 |
Neg logit contributions | |
happens | -0.062 |
ield | -0.060 |
happened | -0.059 |
Dan | -0.055 |
meant | -0.052 |
Token | ila |
Token ID | 10102 |
Feature activation | -8.15e-04 |
Pos logit contributions | |
Brown | +0.001 |
Buzz | +0.001 |
seek | +0.001 |
Ruby | +0.001 |
Fl | +0.001 |
Neg logit contributions | |
happens | -0.001 |
ield | -0.001 |
happened | -0.001 |
Dan | -0.001 |
meant | -0.001 |
Token | liked |
Token ID | 8288 |
Feature activation | -6.17e-03 |
Pos logit contributions | |
Brown | +0.001 |
Buzz | +0.001 |
seek | +0.001 |
| | +0.001 |
Ruby | +0.001 |
Neg logit contributions | |
ield | -0.001 |
happens | -0.001 |
happened | -0.001 |
Dan | -0.001 |
David | -0.001 |
Token | to |
Token ID | 284 |
Feature activation | -5.27e-02 |
Pos logit contributions | |
Brown | +0.014 |
Buzz | +0.013 |
Ruby | +0.013 |
| | +0.013 |
seek | +0.013 |
Neg logit contributions | |
happens | -0.008 |
happened | -0.008 |
ield | -0.008 |
Dan | -0.007 |
David | -0.006 |
Token | play |
Token ID | 711 |
Feature activation | -5.53e-02 |
Pos logit contributions | |
Brown | +0.061 |
Buzz | +0.053 |
Fl | +0.051 |
Later | +0.050 |
Ruby | +0.050 |
Neg logit contributions | |
happens | -0.126 |
happened | -0.123 |
ield | -0.122 |
Dan | -0.119 |
David | -0.114 |
Token | hide |
Token ID | 7808 |
Feature activation | -5.15e-02 |
Pos logit contributions | |
Brown | +0.112 |
| | +0.102 |
heads | +0.101 |
ge | +0.099 |
Hope | +0.097 |
Neg logit contributions | |
happens | -0.084 |
happened | -0.083 |
ield | -0.082 |
It | -0.076 |
Dan | -0.072 |
Token | and |
Token ID | 290 |
Feature activation | -9.88e-02 |
Pos logit contributions | |
Brown | +0.120 |
| | +0.112 |
Ell | +0.109 |
Buzz | +0.109 |
Hope | +0.107 |
Neg logit contributions | |
happens | -0.058 |
ield | -0.055 |
ed | -0.054 |
Dan | -0.053 |
aw | -0.052 |
Token | seek |
Token ID | 5380 |
Feature activation | -5.56e-02 |
Pos logit contributions | |
Brown | +0.104 |
Hope | +0.092 |
ind | +0.088 |
ppa | +0.085 |
charged | +0.083 |
Neg logit contributions | |
Dan | -0.221 |
happened | -0.219 |
happens | -0.211 |
Joe | -0.209 |
J | -0.208 |
Token | with |
Token ID | 351 |
Feature activation | -5.83e-02 |
Pos logit contributions | |
Brown | +0.125 |
Buzz | +0.113 |
| | +0.113 |
Mic | +0.112 |
Hope | +0.109 |
Neg logit contributions | |
happens | -0.073 |
happened | -0.072 |
ield | -0.067 |
meant | -0.064 |
enough | -0.063 |
Token | her |
Token ID | 607 |
Feature activation | -3.47e-02 |
Pos logit contributions | |
Brown | +0.099 |
heads | +0.090 |
| | +0.089 |
Buzz | +0.089 |
Later | +0.088 |
Neg logit contributions | |
happens | -0.104 |
Dan | -0.103 |
happened | -0.102 |
ield | -0.096 |
David | -0.095 |
Token | shadow |
Token ID | 9082 |
Feature activation | -3.92e-02 |
Pos logit contributions | |
Brown | +0.062 |
Buzz | +0.059 |
Later | +0.058 |
| | +0.058 |
heads | +0.057 |
Neg logit contributions | |
happened | -0.058 |
happens | -0.058 |
Dan | -0.055 |
ield | -0.054 |
meant | -0.050 |
Token | . |
Token ID | 13 |
Feature activation | -2.86e-04 |
Pos logit contributions | |
Brown | +0.077 |
Buzz | +0.072 |
| | +0.070 |
Later | +0.068 |
Fl | +0.068 |
Neg logit contributions | |
happens | -0.063 |
happened | -0.060 |
ield | -0.059 |
Dan | -0.053 |
meant | -0.052 |
Token | Lily |
Token ID | 20037 |
Feature activation | +7.76e-03 |
Pos logit contributions | |
Brown | +0.028 |
Buzz | +0.027 |
seek | +0.026 |
Ruby | +0.025 |
| | +0.025 |
Neg logit contributions | |
happens | -0.034 |
ield | -0.033 |
happened | -0.032 |
Dan | -0.029 |
oke | -0.029 |
Token | liked |
Token ID | 8288 |
Feature activation | -1.37e-02 |
Pos logit contributions | |
happens | +0.013 |
ield | +0.013 |
happened | +0.012 |
Dan | +0.011 |
oke | +0.011 |
Neg logit contributions | |
Brown | -0.014 |
Buzz | -0.013 |
seek | -0.013 |
| | -0.013 |
walk | -0.012 |
Token | to |
Token ID | 284 |
Feature activation | -5.44e-02 |
Pos logit contributions | |
Brown | +0.032 |
Buzz | +0.030 |
Ruby | +0.030 |
| | +0.030 |
Fl | +0.030 |
Neg logit contributions | |
happens | -0.017 |
happened | -0.016 |
ield | -0.015 |
Dan | -0.014 |
David | -0.012 |
Token | play |
Token ID | 711 |
Feature activation | -5.25e-02 |
Pos logit contributions | |
Brown | +0.061 |
Buzz | +0.054 |
Fl | +0.053 |
Ruby | +0.053 |
Later | +0.052 |
Neg logit contributions | |
happens | -0.129 |
happened | -0.127 |
ield | -0.126 |
Dan | -0.124 |
David | -0.118 |
Token | hide |
Token ID | 7808 |
Feature activation | -4.79e-02 |
Pos logit contributions | |
Brown | +0.108 |
| | +0.097 |
ge | +0.096 |
Hope | +0.095 |
heads | +0.095 |
Neg logit contributions | |
happens | -0.079 |
happened | -0.078 |
ield | -0.075 |
It | -0.070 |
Dan | -0.067 |
Token | and |
Token ID | 290 |
Feature activation | -9.87e-02 |
Pos logit contributions | |
Brown | +0.113 |
| | +0.103 |
Ell | +0.103 |
Buzz | +0.102 |
Hope | +0.102 |
Neg logit contributions | |
happens | -0.055 |
ed | -0.051 |
ield | -0.050 |
happened | -0.048 |
enough | -0.046 |
Token | seek |
Token ID | 5380 |
Feature activation | -5.56e-02 |
Pos logit contributions | |
Brown | +0.104 |
Hope | +0.095 |
ppa | +0.087 |
Elsa | +0.086 |
ind | +0.085 |
Neg logit contributions | |
happened | -0.219 |
Dan | -0.215 |
happens | -0.213 |
Joe | -0.206 |
meant | -0.206 |
Token | in |
Token ID | 287 |
Feature activation | -3.97e-02 |
Pos logit contributions | |
Brown | +0.122 |
Buzz | +0.110 |
Mic | +0.110 |
| | +0.109 |
Hope | +0.107 |
Neg logit contributions | |
happens | -0.075 |
happened | -0.074 |
ield | -0.069 |
enough | -0.067 |
meant | -0.067 |
Token | the |
Token ID | 262 |
Feature activation | -3.04e-02 |
Pos logit contributions | |
Brown | +0.089 |
Buzz | +0.085 |
heads | +0.081 |
Tweet | +0.081 |
Flo | +0.080 |
Neg logit contributions | |
Dan | -0.051 |
David | -0.048 |
happens | -0.046 |
happened | -0.045 |
Matt | -0.044 |
Token | house |
Token ID | 2156 |
Feature activation | -4.05e-02 |
Pos logit contributions | |
Brown | +0.049 |
ge | +0.046 |
Buzz | +0.044 |
heads | +0.044 |
ide | +0.043 |
Neg logit contributions | |
happens | -0.059 |
happened | -0.055 |
ield | -0.055 |
Dan | -0.053 |
It | -0.050 |
Token | . |
Token ID | 13 |
Feature activation | -1.86e-03 |
Pos logit contributions | |
Brown | +0.085 |
Buzz | +0.080 |
| | +0.076 |
Ruby | +0.076 |
Mid | +0.076 |
Neg logit contributions | |
happens | -0.058 |
ield | -0.057 |
happened | -0.055 |
Dan | -0.050 |
meant | -0.050 |
Token | She |
Token ID | 1375 |
Feature activation | -2.30e-02 |
Pos logit contributions | |
Brown | +0.026 |
Buzz | +0.024 |
heads | +0.023 |
lf | +0.023 |
ge | +0.023 |
Neg logit contributions | |
happens | -0.013 |
happened | -0.012 |
ield | -0.012 |
It | -0.012 |
Dan | -0.011 |
Token | loved |
Token ID | 6151 |
Feature activation | -2.27e-02 |
Pos logit contributions | |
Brown | +0.046 |
Buzz | +0.045 |
ge | +0.045 |
heads | +0.044 |
ide | +0.044 |
Neg logit contributions | |
happened | -0.035 |
happens | -0.034 |
ield | -0.029 |
meant | -0.028 |
It | -0.026 |
Token | to |
Token ID | 284 |
Feature activation | -5.65e-02 |
Pos logit contributions | |
Brown | +0.053 |
Buzz | +0.052 |
Ruby | +0.051 |
| | +0.050 |
seek | +0.050 |
Neg logit contributions | |
happened | -0.026 |
happens | -0.026 |
ield | -0.024 |
Dan | -0.022 |
meant | -0.020 |
Token | play |
Token ID | 711 |
Feature activation | -5.17e-02 |
Pos logit contributions | |
Brown | +0.069 |
Buzz | +0.062 |
Ruby | +0.060 |
Fl | +0.060 |
Later | +0.060 |
Neg logit contributions | |
happened | -0.128 |
happens | -0.128 |
ield | -0.124 |
Dan | -0.122 |
David | -0.118 |
Token | hide |
Token ID | 7808 |
Feature activation | -5.08e-02 |
Pos logit contributions | |
ge | +0.100 |
Brown | +0.096 |
| | +0.093 |
charged | +0.093 |
heads | +0.089 |
Neg logit contributions | |
happened | -0.074 |
happens | -0.071 |
ield | -0.071 |
It | -0.070 |
at | -0.061 |
Token | and |
Token ID | 290 |
Feature activation | -9.87e-02 |
Pos logit contributions | |
Brown | +0.128 |
Hope | +0.118 |
Buzz | +0.118 |
| | +0.118 |
Ruby | +0.117 |
Neg logit contributions | |
ed | -0.045 |
happens | -0.039 |
ield | -0.038 |
aw | -0.037 |
enough | -0.037 |
Token | seek |
Token ID | 5380 |
Feature activation | -5.47e-02 |
Pos logit contributions | |
Hope | +0.108 |
Brown | +0.108 |
ide | +0.104 |
ind | +0.100 |
ge | +0.099 |
Neg logit contributions | |
happened | -0.197 |
happens | -0.195 |
meant | -0.186 |
Dan | -0.183 |
It | -0.179 |
Token | with |
Token ID | 351 |
Feature activation | -5.87e-02 |
Pos logit contributions | |
Brown | +0.111 |
Buzz | +0.103 |
| | +0.102 |
Mic | +0.101 |
Hope | +0.099 |
Neg logit contributions | |
happened | -0.078 |
happens | -0.076 |
ield | -0.073 |
enough | -0.072 |
meant | -0.072 |
Token | her |
Token ID | 607 |
Feature activation | -2.81e-02 |
Pos logit contributions | |
Brown | +0.106 |
| | +0.101 |
heads | +0.100 |
Buzz | +0.098 |
ge | +0.096 |
Neg logit contributions | |
happens | -0.096 |
Dan | -0.095 |
happened | -0.093 |
David | -0.089 |
Bill | -0.087 |
Token | friends |
Token ID | 2460 |
Feature activation | -4.11e-02 |
Pos logit contributions | |
Brown | +0.050 |
Buzz | +0.048 |
| | +0.047 |
Later | +0.047 |
heads | +0.046 |
Neg logit contributions | |
happened | -0.048 |
happens | -0.046 |
ield | -0.043 |
Dan | -0.043 |
meant | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
Brown | +0.079 |
Buzz | +0.075 |
| | +0.072 |
seek | +0.071 |
Ruby | +0.070 |
Neg logit contributions | |
happens | -0.065 |
ield | -0.063 |
happened | -0.063 |
Dan | -0.057 |
meant | -0.055 |
Token | . |
Token ID | 13 |
Feature activation | +1.81e-02 |
Pos logit contributions | |
Brown | +0.097 |
Buzz | +0.090 |
Ruby | +0.087 |
seek | +0.086 |
| | +0.086 |
Neg logit contributions | |
happens | -0.090 |
ield | -0.087 |
happened | -0.086 |
Dan | -0.080 |
meant | -0.075 |
Token | They |
Token ID | 1119 |
Feature activation | -3.34e-02 |
Pos logit contributions | |
happens | +0.034 |
ield | +0.034 |
happened | +0.032 |
meant | +0.028 |
fault | +0.027 |
Neg logit contributions | |
| | -0.027 |
Later | -0.027 |
Buzz | -0.027 |
ide | -0.026 |
Brown | -0.026 |
Token | like |
Token ID | 588 |
Feature activation | -9.25e-03 |
Pos logit contributions | |
Brown | +0.054 |
Buzz | +0.051 |
seek | +0.049 |
Ruby | +0.048 |
| | +0.048 |
Neg logit contributions | |
happens | -0.063 |
ield | -0.061 |
happened | -0.060 |
Dan | -0.055 |
meant | -0.053 |
Token | to |
Token ID | 284 |
Feature activation | -6.85e-02 |
Pos logit contributions | |
Brown | +0.019 |
Buzz | +0.018 |
seek | +0.017 |
Ruby | +0.017 |
Fl | +0.017 |
Neg logit contributions | |
happens | -0.014 |
ield | -0.013 |
happened | -0.013 |
Dan | -0.012 |
meant | -0.011 |
Token | run |
Token ID | 1057 |
Feature activation | -3.59e-02 |
Pos logit contributions | |
Brown | +0.082 |
Buzz | +0.072 |
Ruby | +0.068 |
Later | +0.067 |
Fl | +0.066 |
Neg logit contributions | |
happens | -0.160 |
happened | -0.154 |
ield | -0.154 |
Dan | -0.149 |
David | -0.141 |
Token | and |
Token ID | 290 |
Feature activation | -9.86e-02 |
Pos logit contributions | |
Brown | +0.068 |
Buzz | +0.064 |
Ruby | +0.061 |
Hope | +0.060 |
lf | +0.060 |
Neg logit contributions | |
happens | -0.059 |
ield | -0.056 |
happened | -0.056 |
Dan | -0.051 |
meant | -0.049 |
Token | jump |
Token ID | 4391 |
Feature activation | -3.34e-02 |
Pos logit contributions | |
Brown | +0.151 |
Buzz | +0.134 |
Later | +0.132 |
Hope | +0.130 |
ide | +0.127 |
Neg logit contributions | |
happens | -0.189 |
Dan | -0.188 |
happened | -0.185 |
ield | -0.175 |
Bill | -0.172 |
Token | and |
Token ID | 290 |
Feature activation | -7.88e-02 |
Pos logit contributions | |
Brown | +0.067 |
Buzz | +0.061 |
Hope | +0.060 |
| | +0.059 |
Ruby | +0.059 |
Neg logit contributions | |
happens | -0.052 |
happened | -0.048 |
ield | -0.048 |
Dan | -0.044 |
meant | -0.042 |
Token | slide |
Token ID | 10649 |
Feature activation | -3.08e-02 |
Pos logit contributions | |
Brown | +0.111 |
Later | +0.098 |
Buzz | +0.095 |
lf | +0.095 |
Hope | +0.094 |
Neg logit contributions | |
happens | -0.162 |
Dan | -0.158 |
happened | -0.155 |
ield | -0.150 |
Bill | -0.145 |
Token | . |
Token ID | 13 |
Feature activation | +2.01e-02 |
Pos logit contributions | |
Brown | +0.064 |
Buzz | +0.060 |
Ruby | +0.058 |
seek | +0.058 |
Mid | +0.058 |
Neg logit contributions | |
happens | -0.045 |
happened | -0.042 |
ield | -0.042 |
Dan | -0.039 |
meant | -0.036 |
Token | They |
Token ID | 1119 |
Feature activation | -3.48e-02 |
Pos logit contributions | |
happens | +0.040 |
ield | +0.040 |
happened | +0.037 |
meant | +0.034 |
ier | +0.033 |
Neg logit contributions | |
Later | -0.029 |
While | -0.028 |
| | -0.028 |
Fl | -0.028 |
Buzz | -0.028 |
Token | Ben |
Token ID | 3932 |
Feature activation | +1.38e-02 |
Pos logit contributions | |
walk | +0.015 |
Buzz | +0.015 |
Brown | +0.015 |
sleep | +0.014 |
Hope | +0.014 |
Neg logit contributions | |
happens | -0.020 |
ield | -0.020 |
oke | -0.019 |
ement | -0.018 |
happened | -0.018 |
Token | liked |
Token ID | 8288 |
Feature activation | -1.10e-02 |
Pos logit contributions | |
happens | +0.022 |
ield | +0.022 |
happened | +0.021 |
Dan | +0.019 |
meant | +0.018 |
Neg logit contributions | |
Brown | -0.025 |
Buzz | -0.024 |
seek | -0.023 |
| | -0.023 |
Ruby | -0.023 |
Token | to |
Token ID | 284 |
Feature activation | -5.55e-02 |
Pos logit contributions | |
Brown | +0.025 |
Buzz | +0.024 |
Ruby | +0.023 |
| | +0.023 |
Fl | +0.023 |
Neg logit contributions | |
happens | -0.014 |
happened | -0.014 |
ield | -0.013 |
Dan | -0.012 |
David | -0.011 |
Token | play |
Token ID | 711 |
Feature activation | -5.94e-02 |
Pos logit contributions | |
Brown | +0.049 |
Buzz | +0.042 |
Later | +0.039 |
Ruby | +0.039 |
Fl | +0.039 |
Neg logit contributions | |
happens | -0.146 |
ield | -0.142 |
happened | -0.142 |
Dan | -0.138 |
David | -0.133 |
Token | hide |
Token ID | 7808 |
Feature activation | -5.25e-02 |
Pos logit contributions | |
Brown | +0.122 |
heads | +0.109 |
| | +0.109 |
Hope | +0.107 |
ge | +0.106 |
Neg logit contributions | |
happens | -0.089 |
happened | -0.088 |
ield | -0.085 |
It | -0.080 |
Dan | -0.075 |
Token | and |
Token ID | 290 |
Feature activation | -9.85e-02 |
Pos logit contributions | |
Brown | +0.120 |
| | +0.111 |
Ruby | +0.110 |
Hope | +0.109 |
Ell | +0.109 |
Neg logit contributions | |
happens | -0.062 |
ield | -0.059 |
ed | -0.057 |
Dan | -0.057 |
happened | -0.055 |
Token | seek |
Token ID | 5380 |
Feature activation | -5.81e-02 |
Pos logit contributions | |
Brown | +0.100 |
Hope | +0.088 |
ind | +0.084 |
ppa | +0.083 |
Elsa | +0.082 |
Neg logit contributions | |
Dan | -0.221 |
happened | -0.219 |
happens | -0.216 |
Joe | -0.211 |
Bill | -0.209 |
Token | in |
Token ID | 287 |
Feature activation | -3.99e-02 |
Pos logit contributions | |
Brown | +0.127 |
Buzz | +0.115 |
| | +0.115 |
Mic | +0.114 |
Hope | +0.112 |
Neg logit contributions | |
happens | -0.078 |
happened | -0.077 |
ield | -0.073 |
meant | -0.069 |
enough | -0.068 |
Token | their |
Token ID | 511 |
Feature activation | -4.74e-02 |
Pos logit contributions | |
Brown | +0.086 |
Buzz | +0.083 |
heads | +0.079 |
Tweet | +0.079 |
Flo | +0.077 |
Neg logit contributions | |
Dan | -0.053 |
David | -0.050 |
happens | -0.049 |
happened | -0.047 |
Bill | -0.047 |
Token | big |
Token ID | 1263 |
Feature activation | -5.53e-02 |
Pos logit contributions | |
Brown | +0.071 |
Buzz | +0.066 |
heads | +0.062 |
ge | +0.061 |
Later | +0.061 |
Neg logit contributions | |
happens | -0.095 |
happened | -0.093 |
ield | -0.090 |
Dan | -0.090 |
David | -0.083 |
Token | house |
Token ID | 2156 |
Feature activation | -4.85e-02 |
Pos logit contributions | |
Brown | +0.079 |
Buzz | +0.072 |
Later | +0.067 |
Fl | +0.067 |
Ruby | +0.065 |
Neg logit contributions | |
happens | -0.118 |
ield | -0.113 |
happened | -0.112 |
Dan | -0.106 |
David | -0.102 |
Token | . |
Token ID | 13 |
Feature activation | +9.49e-03 |
Pos logit contributions | |
Brown | +0.061 |
hind | +0.059 |
lf | +0.059 |
Buzz | +0.059 |
seek | +0.058 |
Neg logit contributions | |
happened | -0.033 |
happens | -0.029 |
Dan | -0.029 |
ield | -0.029 |
meant | -0.027 |
Token | Spot |
Token ID | 15899 |
Feature activation | -9.81e-03 |
Pos logit contributions | |
happens | +0.019 |
ield | +0.018 |
happened | +0.018 |
Dan | +0.017 |
meant | +0.016 |
Neg logit contributions | |
Brown | -0.014 |
Buzz | -0.013 |
seek | -0.013 |
Ruby | -0.013 |
| | -0.013 |
Token | loved |
Token ID | 6151 |
Feature activation | -2.01e-02 |
Pos logit contributions | |
Brown | +0.020 |
Buzz | +0.019 |
heads | +0.019 |
Later | +0.019 |
ge | +0.018 |
Neg logit contributions | |
happened | -0.014 |
happens | -0.014 |
ield | -0.012 |
meant | -0.012 |
Dan | -0.012 |
Token | to |
Token ID | 284 |
Feature activation | -5.81e-02 |
Pos logit contributions | |
Brown | +0.046 |
Buzz | +0.044 |
Ruby | +0.044 |
| | +0.043 |
ide | +0.043 |
Neg logit contributions | |
happens | -0.024 |
happened | -0.024 |
ield | -0.023 |
Dan | -0.021 |
meant | -0.019 |
Token | run |
Token ID | 1057 |
Feature activation | -4.43e-02 |
Pos logit contributions | |
Brown | +0.077 |
ge | +0.067 |
heads | +0.066 |
Ruby | +0.065 |
Ell | +0.065 |
Neg logit contributions | |
happened | -0.124 |
happens | -0.124 |
ield | -0.123 |
Dan | -0.123 |
It | -0.116 |
Token | and |
Token ID | 290 |
Feature activation | -9.84e-02 |
Pos logit contributions | |
Brown | +0.097 |
Buzz | +0.092 |
Bee | +0.091 |
Hope | +0.089 |
lf | +0.087 |
Neg logit contributions | |
happens | -0.059 |
happened | -0.054 |
ield | -0.053 |
It | -0.048 |
enough | -0.047 |
Token | play |
Token ID | 711 |
Feature activation | -6.24e-02 |
Pos logit contributions | |
Brown | +0.130 |
ind | +0.120 |
ide | +0.118 |
Hope | +0.117 |
Buzz | +0.116 |
Neg logit contributions | |
happened | -0.194 |
Dan | -0.190 |
happens | -0.188 |
Joe | -0.183 |
Bill | -0.181 |
Token | in |
Token ID | 287 |
Feature activation | -5.31e-02 |
Pos logit contributions | |
Brown | +0.126 |
heads | +0.115 |
ge | +0.115 |
Hope | +0.113 |
Bee | +0.112 |
Neg logit contributions | |
happens | -0.088 |
happened | -0.087 |
ield | -0.083 |
It | -0.080 |
ed | -0.074 |
Token | the |
Token ID | 262 |
Feature activation | -2.68e-02 |
Pos logit contributions | |
Brown | +0.128 |
heads | +0.127 |
ppa | +0.125 |
Buzz | +0.124 |
sie | +0.122 |
Neg logit contributions | |
Dan | -0.056 |
It | -0.055 |
Bill | -0.049 |
happens | -0.048 |
David | -0.047 |
Token | park |
Token ID | 3952 |
Feature activation | -6.33e-02 |
Pos logit contributions | |
Brown | +0.039 |
ge | +0.036 |
Buzz | +0.036 |
lf | +0.035 |
heads | +0.035 |
Neg logit contributions | |
happens | -0.054 |
ield | -0.053 |
happened | -0.052 |
Dan | -0.049 |
It | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | -4.21e-03 |
Pos logit contributions | |
Brown | +0.129 |
Buzz | +0.116 |
Ruby | +0.114 |
Hope | +0.112 |
Mic | +0.112 |
Neg logit contributions | |
happens | -0.095 |
ield | -0.094 |
happened | -0.092 |
Dan | -0.086 |
meant | -0.080 |
Token | girl |
Token ID | 2576 |
Feature activation | -5.65e-03 |
Pos logit contributions | |
Brown | +0.023 |
Buzz | +0.022 |
| | +0.021 |
Later | +0.021 |
Ruby | +0.021 |
Neg logit contributions | |
happens | -0.031 |
ield | -0.030 |
happened | -0.030 |
Dan | -0.028 |
meant | -0.027 |
Token | who |
Token ID | 508 |
Feature activation | +1.90e-03 |
Pos logit contributions | |
Brown | +0.007 |
Buzz | +0.007 |
Ruby | +0.006 |
seek | +0.006 |
Fl | +0.006 |
Neg logit contributions | |
happens | -0.013 |
ield | -0.012 |
happened | -0.012 |
Dan | -0.011 |
meant | -0.011 |
Token | loved |
Token ID | 6151 |
Feature activation | -2.45e-02 |
Pos logit contributions | |
happens | +0.004 |
ield | +0.004 |
happened | +0.004 |
Dan | +0.003 |
oke | +0.003 |
Neg logit contributions | |
Brown | -0.003 |
Buzz | -0.003 |
seek | -0.003 |
| | -0.003 |
sleep | -0.002 |
Token | to |
Token ID | 284 |
Feature activation | -4.69e-02 |
Pos logit contributions | |
Brown | +0.056 |
Buzz | +0.054 |
Ruby | +0.052 |
| | +0.052 |
Fl | +0.052 |
Neg logit contributions | |
happened | -0.031 |
happens | -0.030 |
ield | -0.028 |
Dan | -0.025 |
meant | -0.024 |
Token | run |
Token ID | 1057 |
Feature activation | -4.31e-02 |
Pos logit contributions | |
Brown | +0.051 |
Buzz | +0.045 |
Fl | +0.042 |
Ruby | +0.042 |
Later | +0.042 |
Neg logit contributions | |
happens | -0.114 |
happened | -0.112 |
ield | -0.112 |
Dan | -0.107 |
David | -0.103 |
Token | and |
Token ID | 290 |
Feature activation | -9.82e-02 |
Pos logit contributions | |
Brown | +0.085 |
Buzz | +0.082 |
Bee | +0.079 |
Ruby | +0.077 |
Hope | +0.077 |
Neg logit contributions | |
happens | -0.066 |
ield | -0.062 |
happened | -0.062 |
Dan | -0.055 |
meant | -0.054 |
Token | play |
Token ID | 711 |
Feature activation | -6.41e-02 |
Pos logit contributions | |
Brown | +0.105 |
Buzz | +0.098 |
ide | +0.093 |
heads | +0.091 |
Hope | +0.088 |
Neg logit contributions | |
happened | -0.219 |
happens | -0.218 |
Dan | -0.215 |
David | -0.208 |
Bill | -0.206 |
Token | . |
Token ID | 13 |
Feature activation | -1.12e-02 |
Pos logit contributions | |
Brown | +0.136 |
heads | +0.127 |
Bee | +0.122 |
Buzz | +0.121 |
ge | +0.120 |
Neg logit contributions | |
happens | -0.083 |
happened | -0.081 |
ield | -0.078 |
It | -0.072 |
ed | -0.068 |
Token | One |
Token ID | 1881 |
Feature activation | +2.22e-02 |
Pos logit contributions | |
Brown | +0.022 |
Buzz | +0.020 |
seek | +0.020 |
Ruby | +0.020 |
| | +0.019 |
Neg logit contributions | |
happens | -0.018 |
ield | -0.017 |
happened | -0.017 |
Dan | -0.015 |
meant | -0.014 |
Token | day |
Token ID | 1110 |
Feature activation | +3.18e-02 |
Pos logit contributions | |
happens | +0.048 |
ield | +0.046 |
happened | +0.044 |
Dan | +0.041 |
oke | +0.040 |
Neg logit contributions | |
Brown | -0.030 |
seek | -0.028 |
Buzz | -0.028 |
walk | -0.027 |
Later | -0.027 |
Token | , |
Token ID | 11 |
Feature activation | +1.78e-02 |
Pos logit contributions | |
happens | +0.051 |
ield | +0.051 |
happened | +0.047 |
Dan | +0.045 |
meant | +0.043 |
Neg logit contributions | |
Brown | -0.057 |
Buzz | -0.056 |
Fl | -0.054 |
Ruby | -0.053 |
Later | -0.053 |
Token | explore |
Token ID | 7301 |
Feature activation | -4.58e-02 |
Pos logit contributions | |
Brown | +0.046 |
Ruby | +0.039 |
Later | +0.038 |
heads | +0.038 |
Buzz | +0.038 |
Neg logit contributions | |
happens | -0.093 |
happened | -0.092 |
ield | -0.092 |
Dan | -0.089 |
Bill | -0.084 |
Token | new |
Token ID | 649 |
Feature activation | -5.50e-02 |
Pos logit contributions | |
Brown | +0.092 |
heads | +0.086 |
Buzz | +0.085 |
Ruby | +0.082 |
| | +0.082 |
Neg logit contributions | |
happens | -0.069 |
happened | -0.066 |
ield | -0.064 |
Dan | -0.059 |
meant | -0.057 |
Token | places |
Token ID | 4113 |
Feature activation | -3.91e-02 |
Pos logit contributions | |
Brown | +0.081 |
Buzz | +0.073 |
Ruby | +0.070 |
Mid | +0.070 |
Fl | +0.069 |
Neg logit contributions | |
happens | -0.116 |
happened | -0.109 |
ield | -0.109 |
Dan | -0.102 |
meant | -0.099 |
Token | . |
Token ID | 13 |
Feature activation | +9.90e-03 |
Pos logit contributions | |
Brown | +0.069 |
Buzz | +0.065 |
Ruby | +0.063 |
seek | +0.062 |
| | +0.062 |
Neg logit contributions | |
happens | -0.067 |
ield | -0.065 |
happened | -0.064 |
Dan | -0.059 |
meant | -0.056 |
Token | One |
Token ID | 1881 |
Feature activation | +2.53e-02 |
Pos logit contributions | |
happens | +0.018 |
happened | +0.017 |
ield | +0.016 |
Dan | +0.014 |
meant | +0.014 |
Neg logit contributions | |
Brown | -0.018 |
Buzz | -0.017 |
| | -0.016 |
Later | -0.016 |
Fl | -0.016 |
Token | day |
Token ID | 1110 |
Feature activation | +4.31e-02 |
Pos logit contributions | |
happens | +0.052 |
ield | +0.050 |
happened | +0.048 |
Dan | +0.045 |
David | +0.043 |
Neg logit contributions | |
Brown | -0.037 |
Buzz | -0.035 |
seek | -0.035 |
| | -0.034 |
Later | -0.033 |
Token | , |
Token ID | 11 |
Feature activation | +2.66e-02 |
Pos logit contributions | |
ield | +0.076 |
happens | +0.075 |
fault | +0.068 |
oke | +0.066 |
happened | +0.065 |
Neg logit contributions | |
Brown | -0.071 |
Buzz | -0.068 |
Fl | -0.068 |
seek | -0.067 |
walk | -0.066 |
Token | he |
Token ID | 339 |
Feature activation | -2.26e-03 |
Pos logit contributions | |
happens | +0.046 |
happened | +0.044 |
ield | +0.044 |
Dan | +0.041 |
meant | +0.038 |
Neg logit contributions | |
Brown | -0.048 |
Buzz | -0.045 |
Ruby | -0.043 |
seek | -0.043 |
| | -0.043 |
Token | entered |
Token ID | 5982 |
Feature activation | +2.64e-04 |
Pos logit contributions | |
Brown | +0.004 |
Buzz | +0.004 |
seek | +0.004 |
Ruby | +0.004 |
| | +0.004 |
Neg logit contributions | |
happens | -0.004 |
ield | -0.004 |
happened | -0.004 |
Dan | -0.003 |
David | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | -3.54e-02 |
Pos logit contributions | |
happens | +0.000 |
happened | +0.000 |
ield | +0.000 |
Dan | +0.000 |
meant | +0.000 |
Neg logit contributions | |
Brown | -0.001 |
Buzz | -0.001 |
Ruby | -0.000 |
| | -0.000 |
seek | -0.000 |
Token | room |
Token ID | 2119 |
Feature activation | -1.92e-04 |
Pos logit contributions | |
Brown | +0.061 |
Buzz | +0.056 |
| | +0.055 |
Mid | +0.055 |
bes | +0.054 |
Neg logit contributions | |
happens | -0.060 |
ield | -0.059 |
happened | -0.058 |
Dan | -0.051 |
It | -0.049 |
Token | to |
Token ID | 284 |
Feature activation | -4.45e-02 |
Pos logit contributions | |
happens | +0.042 |
ield | +0.042 |
happened | +0.039 |
oke | +0.037 |
Dan | +0.035 |
Neg logit contributions | |
Brown | -0.041 |
seek | -0.040 |
Buzz | -0.039 |
sleep | -0.039 |
blankets | -0.038 |
Token | shout |
Token ID | 9191 |
Feature activation | +3.29e-03 |
Pos logit contributions | |
Brown | +0.073 |
Later | +0.067 |
Ruby | +0.066 |
Mid | +0.065 |
heads | +0.065 |
Neg logit contributions | |
happens | -0.081 |
happened | -0.080 |
Dan | -0.079 |
ield | -0.079 |
Bill | -0.073 |
Token | . |
Token ID | 13 |
Feature activation | +6.90e-03 |
Pos logit contributions | |
happens | +0.006 |
ield | +0.006 |
happened | +0.006 |
Dan | +0.005 |
meant | +0.005 |
Neg logit contributions | |
Brown | -0.005 |
seek | -0.005 |
Buzz | -0.005 |
Mid | -0.005 |
Ruby | -0.005 |
Token |
|
Token ID | 198 |
Feature activation | -3.11e-02 |
Pos logit contributions | |
happens | +0.012 |
happened | +0.011 |
ield | +0.011 |
meant | +0.009 |
Dan | +0.009 |
Neg logit contributions | |
Brown | -0.013 |
Buzz | -0.012 |
Ruby | -0.012 |
Later | -0.012 |
Fl | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | -4.13e-02 |
Pos logit contributions | |
Brown | +0.058 |
Buzz | +0.054 |
seek | +0.053 |
Ruby | +0.052 |
| | +0.052 |
Neg logit contributions | |
happens | -0.050 |
ield | -0.048 |
happened | -0.047 |
Dan | -0.045 |
David | -0.041 |
Token | Suddenly |
Token ID | 38582 |
Feature activation | +6.08e-02 |
Pos logit contributions | |
carpet | +0.053 |
Brown | +0.053 |
throne | +0.053 |
blankets | +0.052 |
cushion | +0.050 |
Neg logit contributions | |
ield | -0.082 |
Dan | -0.080 |
happens | -0.080 |
happened | -0.076 |
David | -0.074 |
Token | , |
Token ID | 11 |
Feature activation | +3.42e-02 |
Pos logit contributions | |
ield | +0.103 |
happens | +0.101 |
happened | +0.093 |
oke | +0.089 |
fault | +0.088 |
Neg logit contributions | |
Brown | -0.104 |
Buzz | -0.100 |
seek | -0.100 |
Ruby | -0.096 |
lf | -0.095 |
Token | a |
Token ID | 257 |
Feature activation | +1.32e-02 |
Pos logit contributions | |
happens | +0.058 |
ield | +0.056 |
happened | +0.055 |
Dan | +0.050 |
meant | +0.048 |
Neg logit contributions | |
Brown | -0.062 |
Buzz | -0.058 |
seek | -0.056 |
Ruby | -0.056 |
| | -0.055 |
Token | little |
Token ID | 1310 |
Feature activation | +2.69e-02 |
Pos logit contributions | |
happens | +0.027 |
ield | +0.026 |
happened | +0.026 |
Dan | +0.025 |
meant | +0.023 |
Neg logit contributions | |
Brown | -0.019 |
Buzz | -0.018 |
seek | -0.017 |
Ruby | -0.017 |
lf | -0.016 |
Token | girl |
Token ID | 2576 |
Feature activation | +1.81e-02 |
Pos logit contributions | |
happens | +0.062 |
ield | +0.059 |
happened | +0.059 |
Dan | +0.055 |
meant | +0.052 |
Neg logit contributions | |
Brown | -0.033 |
Buzz | -0.031 |
seek | -0.030 |
Ruby | -0.029 |
cushion | -0.028 |
Token | appeared |
Token ID | 4120 |
Feature activation | +8.79e-03 |
Pos logit contributions | |
happens | +0.029 |
ield | +0.028 |
fault | +0.026 |
happened | +0.026 |
Dan | +0.024 |
Neg logit contributions | |
Brown | -0.033 |
Buzz | -0.032 |
Ruby | -0.031 |
walk | -0.031 |
seek | -0.031 |
Token | sad |
Token ID | 6507 |
Feature activation | +2.27e-02 |
Pos logit contributions | |
happens | +0.050 |
ield | +0.048 |
happened | +0.045 |
Dan | +0.040 |
oke | +0.038 |
Neg logit contributions | |
Brown | -0.053 |
Buzz | -0.050 |
seek | -0.050 |
| | -0.048 |
Ruby | -0.048 |
Token | but |
Token ID | 475 |
Feature activation | +3.03e-02 |
Pos logit contributions | |
happens | +0.043 |
ield | +0.042 |
happened | +0.040 |
Dan | +0.038 |
David | +0.036 |
Neg logit contributions | |
Brown | -0.034 |
sleep | -0.032 |
ide | -0.031 |
Buzz | -0.031 |
seek | -0.031 |
Token | they |
Token ID | 484 |
Feature activation | +1.29e-02 |
Pos logit contributions | |
ield | +0.063 |
happens | +0.062 |
happened | +0.057 |
oke | +0.053 |
fault | +0.053 |
Neg logit contributions | |
seek | -0.041 |
Brown | -0.040 |
Buzz | -0.039 |
Ruby | -0.037 |
sleep | -0.037 |
Token | knew |
Token ID | 2993 |
Feature activation | +3.89e-02 |
Pos logit contributions | |
happens | +0.020 |
ield | +0.019 |
happened | +0.018 |
Dan | +0.017 |
David | +0.016 |
Neg logit contributions | |
Brown | -0.025 |
Buzz | -0.024 |
seek | -0.023 |
Ruby | -0.023 |
| | -0.023 |
Token | they |
Token ID | 484 |
Feature activation | +2.12e-02 |
Pos logit contributions | |
ield | +0.076 |
happens | +0.074 |
happened | +0.068 |
oke | +0.067 |
elt | +0.065 |
Neg logit contributions | |
Brown | -0.054 |
seek | -0.051 |
Buzz | -0.050 |
blankets | -0.050 |
sleep | -0.050 |
Token | had |
Token ID | 550 |
Feature activation | +4.69e-02 |
Pos logit contributions | |
happens | +0.034 |
ield | +0.034 |
Dan | +0.031 |
happened | +0.030 |
David | +0.028 |
Neg logit contributions | |
Brown | -0.037 |
Buzz | -0.036 |
walk | -0.035 |
seek | -0.035 |
| | -0.035 |
Token | experienced |
Token ID | 5924 |
Feature activation | +2.36e-02 |
Pos logit contributions | |
ield | +0.079 |
happens | +0.076 |
oke | +0.070 |
Dan | +0.068 |
happened | +0.066 |
Neg logit contributions | |
seek | -0.077 |
Brown | -0.076 |
Buzz | -0.075 |
sleep | -0.075 |
walk | -0.075 |
Token | something |
Token ID | 1223 |
Feature activation | +3.68e-02 |
Pos logit contributions | |
ield | +0.042 |
happens | +0.040 |
oke | +0.038 |
happened | +0.036 |
Dan | +0.036 |
Neg logit contributions | |
blankets | -0.036 |
sleep | -0.036 |
seek | -0.036 |
Brown | -0.035 |
Buzz | -0.035 |
Token | wonderful |
Token ID | 7932 |
Feature activation | +1.79e-02 |
Pos logit contributions | |
ield | +0.074 |
happens | +0.070 |
Dan | +0.066 |
happened | +0.065 |
David | +0.063 |
Neg logit contributions | |
Brown | -0.049 |
seek | -0.049 |
Buzz | -0.048 |
carpet | -0.048 |
ide | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | +2.56e-02 |
Pos logit contributions | |
ield | +0.038 |
happens | +0.036 |
oke | +0.035 |
Dan | +0.034 |
Everything | +0.033 |
Neg logit contributions | |
seek | -0.022 |
sleep | -0.021 |
walk | -0.020 |
lf | -0.020 |
ide | -0.020 |
Token | Now |
Token ID | 2735 |
Feature activation | +2.77e-02 |
Pos logit contributions | |
happens | +0.046 |
happened | +0.045 |
ield | +0.043 |
meant | +0.037 |
Dan | +0.036 |
Neg logit contributions | |
Brown | -0.044 |
| | -0.042 |
Buzz | -0.042 |
Ruby | -0.042 |
Later | -0.041 |
Token |
|
Token ID | 198 |
Feature activation | -5.65e-02 |
Pos logit contributions | |
throne | +0.115 |
Brown | +0.113 |
charged | +0.111 |
seek | +0.111 |
lf | +0.110 |
Neg logit contributions | |
It | -0.041 |
happens | -0.041 |
happened | -0.040 |
Dan | -0.036 |
ield | -0.032 |
Token | L |
Token ID | 43 |
Feature activation | +6.72e-03 |
Pos logit contributions | |
throne | +0.100 |
den | +0.091 |
hind | +0.087 |
blankets | +0.084 |
sleep | +0.084 |
Neg logit contributions | |
ield | -0.092 |
It | -0.091 |
happens | -0.091 |
happened | -0.081 |
aw | -0.080 |
Token | ily |
Token ID | 813 |
Feature activation | -1.31e-02 |
Pos logit contributions | |
happens | +0.007 |
happened | +0.007 |
ield | +0.006 |
meant | +0.005 |
Dan | +0.005 |
Neg logit contributions | |
Brown | -0.017 |
Buzz | -0.016 |
seek | -0.016 |
Fl | -0.015 |
throne | -0.015 |
Token | knew |
Token ID | 2993 |
Feature activation | +7.95e-03 |
Pos logit contributions | |
Brown | +0.028 |
Buzz | +0.026 |
| | +0.025 |
seek | +0.025 |
Ruby | +0.025 |
Neg logit contributions | |
happens | -0.018 |
ield | -0.017 |
happened | -0.017 |
Dan | -0.015 |
meant | -0.015 |
Token | she |
Token ID | 673 |
Feature activation | +1.95e-02 |
Pos logit contributions | |
happens | +0.014 |
ield | +0.014 |
happened | +0.013 |
Dan | +0.012 |
meant | +0.012 |
Neg logit contributions | |
Brown | -0.014 |
Buzz | -0.013 |
seek | -0.012 |
Ruby | -0.012 |
Later | -0.012 |
Token | had |
Token ID | 550 |
Feature activation | +4.22e-02 |
Pos logit contributions | |
happens | +0.036 |
ield | +0.035 |
Dan | +0.034 |
oke | +0.031 |
David | +0.030 |
Neg logit contributions | |
Brown | -0.030 |
walk | -0.028 |
seek | -0.027 |
read | -0.027 |
Ell | -0.027 |
Token | to |
Token ID | 284 |
Feature activation | -3.38e-02 |
Pos logit contributions | |
ield | +0.079 |
happens | +0.077 |
oke | +0.076 |
ement | +0.073 |
Dan | +0.069 |
Neg logit contributions | |
seek | -0.063 |
walk | -0.062 |
sleep | -0.061 |
read | -0.060 |
Brown | -0.059 |
Token | choose |
Token ID | 3853 |
Feature activation | +6.78e-03 |
Pos logit contributions | |
Brown | +0.055 |
Buzz | +0.051 |
Ruby | +0.049 |
Fl | +0.048 |
| | +0.048 |
Neg logit contributions | |
happens | -0.063 |
ield | -0.063 |
happened | -0.061 |
Dan | -0.058 |
meant | -0.056 |
Token | what |
Token ID | 644 |
Feature activation | +3.80e-02 |
Pos logit contributions | |
happens | +0.013 |
ield | +0.013 |
happened | +0.013 |
Dan | +0.011 |
oke | +0.011 |
Neg logit contributions | |
Brown | -0.010 |
Buzz | -0.009 |
seek | -0.009 |
Fl | -0.009 |
Ruby | -0.009 |
Token | to |
Token ID | 284 |
Feature activation | -1.31e-02 |
Pos logit contributions | |
ield | +0.069 |
happens | +0.066 |
oke | +0.061 |
Dan | +0.060 |
ement | +0.059 |
Neg logit contributions | |
Brown | -0.061 |
seek | -0.060 |
Buzz | -0.059 |
Ruby | -0.058 |
Fl | -0.056 |
Token | do |
Token ID | 466 |
Feature activation | +1.44e-02 |
Pos logit contributions | |
Brown | +0.026 |
Buzz | +0.025 |
Ruby | +0.024 |
Fl | +0.024 |
Mid | +0.024 |
Neg logit contributions | |
happens | -0.020 |
happened | -0.019 |
ield | -0.019 |
Dan | -0.018 |
meant | -0.017 |
Token | mom |
Token ID | 1995 |
Feature activation | +9.80e-03 |
Pos logit contributions | |
happens | +0.015 |
happened | +0.014 |
ield | +0.014 |
Dan | +0.013 |
meant | +0.012 |
Neg logit contributions | |
Brown | -0.019 |
| | -0.018 |
Buzz | -0.018 |
Later | -0.017 |
Ell | -0.017 |
Token | smiled |
Token ID | 13541 |
Feature activation | +1.00e-02 |
Pos logit contributions | |
happens | +0.017 |
ield | +0.016 |
happened | +0.016 |
Dan | +0.015 |
meant | +0.014 |
Neg logit contributions | |
Brown | -0.018 |
Buzz | -0.017 |
Ruby | -0.016 |
seek | -0.016 |
Later | -0.016 |
Token | and |
Token ID | 290 |
Feature activation | -1.70e-02 |
Pos logit contributions | |
happens | +0.019 |
ield | +0.018 |
happened | +0.018 |
Dan | +0.017 |
David | +0.016 |
Neg logit contributions | |
Brown | -0.016 |
Buzz | -0.015 |
Ruby | -0.014 |
Later | -0.014 |
seek | -0.014 |
Token | told |
Token ID | 1297 |
Feature activation | +6.95e-03 |
Pos logit contributions | |
Brown | +0.034 |
Buzz | +0.032 |
seek | +0.031 |
Ruby | +0.031 |
Later | +0.031 |
Neg logit contributions | |
happens | -0.025 |
ield | -0.024 |
happened | -0.024 |
Dan | -0.022 |
meant | -0.020 |
Token | her |
Token ID | 607 |
Feature activation | +1.93e-02 |
Pos logit contributions | |
happens | +0.015 |
happened | +0.014 |
ield | +0.014 |
Dan | +0.013 |
meant | +0.013 |
Neg logit contributions | |
Brown | -0.010 |
Buzz | -0.009 |
seek | -0.009 |
| | -0.008 |
Ruby | -0.008 |
Token | that |
Token ID | 326 |
Feature activation | +4.59e-02 |
Pos logit contributions | |
happens | +0.031 |
ield | +0.031 |
happened | +0.029 |
Dan | +0.028 |
meant | +0.026 |
Neg logit contributions | |
Brown | -0.034 |
Buzz | -0.032 |
seek | -0.031 |
lf | -0.031 |
Later | -0.031 |
Token | she |
Token ID | 673 |
Feature activation | +3.89e-02 |
Pos logit contributions | |
ield | +0.079 |
happens | +0.078 |
happened | +0.074 |
Dan | +0.070 |
oke | +0.067 |
Neg logit contributions | |
Brown | -0.079 |
Buzz | -0.075 |
seek | -0.073 |
Fl | -0.072 |
Ruby | -0.072 |
Token | was |
Token ID | 373 |
Feature activation | +3.87e-02 |
Pos logit contributions | |
Dan | +0.063 |
ield | +0.062 |
happens | +0.059 |
David | +0.057 |
Bill | +0.055 |
Neg logit contributions | |
Brown | -0.066 |
Buzz | -0.063 |
sleep | -0.062 |
Ruby | -0.062 |
ie | -0.061 |
Token | very |
Token ID | 845 |
Feature activation | +1.56e-02 |
Pos logit contributions | |
happens | +0.066 |
ield | +0.064 |
happened | +0.062 |
Dan | +0.058 |
David | +0.053 |
Neg logit contributions | |
Brown | -0.070 |
Buzz | -0.067 |
seek | -0.064 |
Ruby | -0.064 |
Fl | -0.063 |
Token | proud |
Token ID | 6613 |
Feature activation | +2.35e-02 |
Pos logit contributions | |
happens | +0.028 |
happened | +0.026 |
ield | +0.026 |
Dan | +0.024 |
meant | +0.024 |
Neg logit contributions | |
Brown | -0.028 |
Buzz | -0.025 |
| | -0.025 |
Ruby | -0.025 |
lf | -0.025 |
Token | of |
Token ID | 286 |
Feature activation | +4.87e-04 |
Pos logit contributions | |
ield | +0.040 |
happens | +0.040 |
happened | +0.038 |
Dan | +0.037 |
David | +0.035 |
Neg logit contributions | |
Brown | -0.037 |
sleep | -0.036 |
seek | -0.036 |
blankets | -0.036 |
carpet | -0.036 |
Token | , |
Token ID | 11 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
happens | +0.042 |
happened | +0.039 |
ield | +0.039 |
Dan | +0.037 |
Well | +0.035 |
Neg logit contributions | |
Brown | -0.037 |
walk | -0.033 |
lf | -0.033 |
carpet | -0.033 |
Buzz | -0.033 |
Token | it |
Token ID | 340 |
Feature activation | +3.86e-02 |
Pos logit contributions | |
happens | +0.024 |
ield | +0.023 |
happened | +0.022 |
Dan | +0.020 |
meant | +0.019 |
Neg logit contributions | |
Brown | -0.026 |
Buzz | -0.025 |
seek | -0.024 |
Ruby | -0.024 |
Fl | -0.023 |
Token | is |
Token ID | 318 |
Feature activation | +4.63e-02 |
Pos logit contributions | |
ield | +0.062 |
Dan | +0.060 |
happens | +0.060 |
happened | +0.058 |
David | +0.056 |
Neg logit contributions | |
Brown | -0.067 |
Ruby | -0.064 |
Buzz | -0.064 |
| | -0.062 |
Fl | -0.062 |
Token | so |
Token ID | 523 |
Feature activation | +1.93e-02 |
Pos logit contributions | |
happens | +0.086 |
ield | +0.085 |
happened | +0.077 |
oke | +0.077 |
Dan | +0.073 |
Neg logit contributions | |
Brown | -0.073 |
carpet | -0.070 |
seek | -0.069 |
Later | -0.067 |
sleep | -0.067 |
Token | bright |
Token ID | 6016 |
Feature activation | +2.47e-02 |
Pos logit contributions | |
happens | +0.043 |
ield | +0.043 |
happened | +0.041 |
Dan | +0.039 |
oke | +0.038 |
Neg logit contributions | |
Brown | -0.024 |
Buzz | -0.022 |
seek | -0.021 |
Later | -0.021 |
Fl | -0.020 |
Token | today |
Token ID | 1909 |
Feature activation | +2.54e-02 |
Pos logit contributions | |
happens | +0.046 |
ield | +0.045 |
happened | +0.044 |
Dan | +0.042 |
meant | +0.039 |
Neg logit contributions | |
Brown | -0.038 |
Buzz | -0.036 |
seek | -0.034 |
ide | -0.034 |
Later | -0.034 |
Token | !' |
Token ID | 13679 |
Feature activation | +6.33e-03 |
Pos logit contributions | |
happens | +0.048 |
ield | +0.046 |
happened | +0.045 |
Dan | +0.043 |
meant | +0.041 |
Neg logit contributions | |
Brown | -0.037 |
Buzz | -0.035 |
ide | -0.035 |
seek | -0.034 |
lf | -0.034 |
Token |
|
Token ID | 198 |
Feature activation | -4.68e-02 |
Pos logit contributions | |
happens | +0.010 |
ield | +0.010 |
happened | +0.009 |
Dan | +0.008 |
meant | +0.008 |
Neg logit contributions | |
Brown | -0.012 |
Buzz | -0.011 |
Later | -0.011 |
| | -0.011 |
Ruby | -0.011 |
Token | The |
Token ID | 464 |
Feature activation | -4.57e-03 |
Pos logit contributions | |
Brown | +0.096 |
Buzz | +0.091 |
seek | +0.091 |
throne | +0.090 |
cushion | +0.089 |
Neg logit contributions | |
happens | -0.061 |
happened | -0.061 |
ield | -0.060 |
Dan | -0.057 |
David | -0.054 |
Token | second |
Token ID | 1218 |
Feature activation | +3.67e-02 |
Pos logit contributions | |
Brown | +0.009 |
| | +0.008 |
heads | +0.008 |
ge | +0.008 |
Later | +0.008 |
Neg logit contributions | |
happens | -0.008 |
happened | -0.007 |
ield | -0.007 |
Dan | -0.007 |
meant | -0.006 |
Token | friend |
Token ID | 1545 |
Feature activation | +1.30e-02 |
Pos logit contributions | |
happens | +0.093 |
ield | +0.090 |
happened | +0.087 |
David | +0.082 |
Dan | +0.081 |
Neg logit contributions | |
Buzz | -0.034 |
seek | -0.031 |
Brown | -0.029 |
ide | -0.029 |
ie | -0.028 |
Token | rotting |
Token ID | 44494 |
Feature activation | +2.51e-02 |
Pos logit contributions | |
ield | +0.069 |
Dan | +0.064 |
happens | +0.063 |
oke | +0.062 |
David | +0.062 |
Neg logit contributions | |
sleep | -0.055 |
carpet | -0.054 |
Brown | -0.054 |
blankets | -0.053 |
walk | -0.053 |
Token | in |
Token ID | 287 |
Feature activation | -1.35e-02 |
Pos logit contributions | |
ield | +0.050 |
happens | +0.049 |
Dan | +0.046 |
happened | +0.046 |
oke | +0.044 |
Neg logit contributions | |
Brown | -0.033 |
Buzz | -0.031 |
seek | -0.031 |
carpet | -0.031 |
ide | -0.031 |
Token | their |
Token ID | 511 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
Brown | +0.021 |
Buzz | +0.020 |
seek | +0.019 |
Ruby | +0.019 |
| | +0.019 |
Neg logit contributions | |
happens | -0.026 |
ield | -0.025 |
happened | -0.025 |
Dan | -0.023 |
meant | -0.022 |
Token | garden |
Token ID | 11376 |
Feature activation | -4.41e-03 |
Pos logit contributions | |
Brown | +0.017 |
Buzz | +0.016 |
seek | +0.015 |
Ruby | +0.015 |
Later | +0.014 |
Neg logit contributions | |
happens | -0.028 |
ield | -0.027 |
happened | -0.027 |
Dan | -0.025 |
David | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | +2.01e-02 |
Pos logit contributions | |
Brown | +0.007 |
Buzz | +0.006 |
seek | +0.006 |
lf | +0.006 |
ide | +0.006 |
Neg logit contributions | |
happens | -0.009 |
ield | -0.008 |
happened | -0.008 |
Dan | -0.007 |
David | -0.007 |
Token | It |
Token ID | 632 |
Feature activation | +2.61e-02 |
Pos logit contributions | |
happens | +0.032 |
happened | +0.029 |
ield | +0.028 |
oke | +0.025 |
Dan | +0.024 |
Neg logit contributions | |
Brown | -0.040 |
Buzz | -0.038 |
Later | -0.037 |
| | -0.037 |
Fl | -0.036 |
Token | smelled |
Token ID | 37609 |
Feature activation | +3.84e-02 |
Pos logit contributions | |
ield | +0.042 |
happens | +0.041 |
Dan | +0.039 |
happened | +0.038 |
David | +0.037 |
Neg logit contributions | |
Brown | -0.049 |
Buzz | -0.047 |
Ruby | -0.045 |
Fl | -0.045 |
seek | -0.045 |
Token | y |
Token ID | 331 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
ield | +0.070 |
happens | +0.070 |
Dan | +0.065 |
happened | +0.065 |
oke | +0.062 |
Neg logit contributions | |
Buzz | -0.058 |
Brown | -0.058 |
Fl | -0.055 |
blankets | -0.055 |
Later | -0.054 |
Token | ucky |
Token ID | 5309 |
Feature activation | +8.99e-03 |
Pos logit contributions | |
happens | +0.029 |
happened | +0.027 |
ield | +0.026 |
Dan | +0.025 |
meant | +0.023 |
Neg logit contributions | |
Buzz | -0.027 |
Brown | -0.027 |
ide | -0.026 |
Later | -0.025 |
Fl | -0.025 |
Token | and |
Token ID | 290 |
Feature activation | +1.20e-02 |
Pos logit contributions | |
happens | +0.016 |
ield | +0.015 |
happened | +0.015 |
Dan | +0.014 |
meant | +0.013 |
Neg logit contributions | |
Brown | -0.016 |
Buzz | -0.015 |
| | -0.014 |
Ruby | -0.014 |
seek | -0.014 |
Token | they |
Token ID | 484 |
Feature activation | -9.64e-03 |
Pos logit contributions | |
happens | +0.019 |
ield | +0.019 |
happened | +0.018 |
Dan | +0.016 |
David | +0.015 |
Neg logit contributions | |
Brown | -0.023 |
Buzz | -0.022 |
seek | -0.021 |
Ruby | -0.021 |
Fl | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | +1.50e-02 |
Pos logit contributions | |
Brown | +0.023 |
Buzz | +0.022 |
seek | +0.021 |
Ruby | +0.021 |
Fl | +0.021 |
Neg logit contributions | |
happens | -0.027 |
ield | -0.025 |
happened | -0.025 |
Dan | -0.023 |
meant | -0.022 |
Token | I |
Token ID | 314 |
Feature activation | +1.51e-02 |
Pos logit contributions | |
happens | +0.027 |
happened | +0.026 |
ield | +0.026 |
meant | +0.022 |
Dan | +0.022 |
Neg logit contributions | |
Buzz | -0.026 |
Brown | -0.025 |
Ruby | -0.025 |
Later | -0.024 |
Fl | -0.024 |
Token | was |
Token ID | 373 |
Feature activation | +2.41e-02 |
Pos logit contributions | |
happens | +0.025 |
ield | +0.024 |
happened | +0.023 |
Dan | +0.022 |
David | +0.020 |
Neg logit contributions | |
Brown | -0.028 |
Buzz | -0.027 |
seek | -0.026 |
| | -0.026 |
Ruby | -0.026 |
Token | so |
Token ID | 523 |
Feature activation | +1.47e-02 |
Pos logit contributions | |
happens | +0.046 |
ield | +0.045 |
happened | +0.042 |
Dan | +0.038 |
oke | +0.037 |
Neg logit contributions | |
Brown | -0.037 |
seek | -0.036 |
carpet | -0.036 |
sleep | -0.035 |
throne | -0.035 |
Token | excited |
Token ID | 6568 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
happens | +0.033 |
ield | +0.033 |
happened | +0.031 |
Dan | +0.029 |
David | +0.028 |
Neg logit contributions | |
Brown | -0.018 |
Buzz | -0.017 |
seek | -0.016 |
Later | -0.015 |
Ruby | -0.015 |
Token | , |
Token ID | 11 |
Feature activation | +1.95e-02 |
Pos logit contributions | |
happens | +0.018 |
ield | +0.018 |
happened | +0.017 |
Dan | +0.016 |
oke | +0.015 |
Neg logit contributions | |
Brown | -0.018 |
seek | -0.017 |
ide | -0.017 |
Buzz | -0.017 |
Later | -0.017 |
Token | too |
Token ID | 1165 |
Feature activation | +2.51e-02 |
Pos logit contributions | |
happens | +0.034 |
ield | +0.032 |
happened | +0.032 |
oke | +0.027 |
Dan | +0.027 |
Neg logit contributions | |
Brown | -0.033 |
seek | -0.032 |
Buzz | -0.032 |
Ruby | -0.031 |
Fl | -0.031 |
Token | ! |
Token ID | 0 |
Feature activation | +1.61e-02 |
Pos logit contributions | |
happens | +0.046 |
ield | +0.044 |
happened | +0.043 |
Dan | +0.039 |
meant | +0.038 |
Neg logit contributions | |
Brown | -0.041 |
Buzz | -0.039 |
seek | -0.038 |
lf | -0.037 |
Later | -0.037 |
Token |
|
Token ID | 198 |
Feature activation | -3.63e-02 |
Pos logit contributions | |
happens | +0.025 |
happened | +0.024 |
ield | +0.023 |
meant | +0.020 |
Dan | +0.020 |
Neg logit contributions | |
Brown | -0.032 |
Buzz | -0.031 |
Ruby | -0.030 |
Fl | -0.029 |
Later | -0.029 |
Token |
|
Token ID | 198 |
Feature activation | -4.88e-02 |
Pos logit contributions | |
Brown | +0.077 |
seek | +0.071 |
heads | +0.071 |
throne | +0.070 |
blankets | +0.069 |
Neg logit contributions | |
happens | -0.048 |
happened | -0.047 |
ield | -0.044 |
Dan | -0.044 |
It | -0.041 |
Token | M |
Token ID | 44 |
Feature activation | +2.84e-02 |
Pos logit contributions | |
Brown | +0.076 |
throne | +0.074 |
| | +0.072 |
blankets | +0.071 |
hind | +0.070 |
Neg logit contributions | |
happens | -0.084 |
ield | -0.083 |
happened | -0.080 |
Dan | -0.079 |
It | -0.075 |
Token | He |
Token ID | 679 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
happens | +0.005 |
happened | +0.004 |
ield | +0.004 |
Dan | +0.004 |
meant | +0.004 |
Neg logit contributions | |
Brown | -0.005 |
Buzz | -0.005 |
Later | -0.005 |
Ruby | -0.005 |
seek | -0.005 |
Token | opened |
Token ID | 4721 |
Feature activation | +1.04e-02 |
Pos logit contributions | |
Brown | +0.030 |
Buzz | +0.028 |
| | +0.027 |
Ruby | +0.027 |
Fl | +0.027 |
Neg logit contributions | |
happens | -0.023 |
happened | -0.022 |
ield | -0.022 |
Dan | -0.020 |
meant | -0.019 |
Token | it |
Token ID | 340 |
Feature activation | -1.69e-02 |
Pos logit contributions | |
happens | +0.020 |
ield | +0.019 |
happened | +0.019 |
Dan | +0.017 |
meant | +0.017 |
Neg logit contributions | |
Brown | -0.017 |
Buzz | -0.016 |
seek | -0.015 |
Ruby | -0.015 |
Later | -0.015 |
Token | and |
Token ID | 290 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
Brown | +0.030 |
Buzz | +0.028 |
| | +0.026 |
seek | +0.026 |
Ruby | +0.026 |
Neg logit contributions | |
happens | -0.030 |
ield | -0.029 |
happened | -0.029 |
Dan | -0.027 |
meant | -0.025 |
Token | found |
Token ID | 1043 |
Feature activation | +2.30e-02 |
Pos logit contributions | |
Brown | +0.029 |
Buzz | +0.027 |
| | +0.026 |
Ruby | +0.026 |
Later | +0.026 |
Neg logit contributions | |
happens | -0.024 |
happened | -0.023 |
ield | -0.023 |
Dan | -0.021 |
David | -0.020 |
Token | some |
Token ID | 617 |
Feature activation | +4.53e-03 |
Pos logit contributions | |
happens | +0.046 |
ield | +0.045 |
happened | +0.044 |
oke | +0.041 |
fault | +0.038 |
Neg logit contributions | |
Buzz | -0.032 |
Brown | -0.032 |
blankets | -0.031 |
Fl | -0.030 |
Ruby | -0.030 |
Token | y |
Token ID | 331 |
Feature activation | +4.75e-03 |
Pos logit contributions | |
happens | +0.010 |
ield | +0.009 |
happened | +0.009 |
Dan | +0.009 |
David | +0.008 |
Neg logit contributions | |
Brown | -0.006 |
Buzz | -0.006 |
seek | -0.005 |
blankets | -0.005 |
Fl | -0.005 |
Token | umm |
Token ID | 13929 |
Feature activation | +3.70e-02 |
Pos logit contributions | |
happens | +0.007 |
happened | +0.007 |
ield | +0.007 |
Dan | +0.006 |
meant | +0.006 |
Neg logit contributions | |
Brown | -0.009 |
Buzz | -0.009 |
ide | -0.009 |
Later | -0.009 |
seek | -0.009 |
Token | my |
Token ID | 1820 |
Feature activation | +7.87e-03 |
Pos logit contributions | |
ield | +0.045 |
happens | +0.045 |
Everything | +0.038 |
It | +0.038 |
Dan | +0.038 |
Neg logit contributions | |
ie | -0.081 |
ide | -0.078 |
ge | -0.075 |
lf | -0.075 |
Brown | -0.074 |
Token | food |
Token ID | 2057 |
Feature activation | -4.62e-03 |
Pos logit contributions | |
happens | +0.015 |
ield | +0.014 |
happened | +0.014 |
Dan | +0.013 |
meant | +0.012 |
Neg logit contributions | |
Brown | -0.013 |
Buzz | -0.012 |
seek | -0.012 |
Ruby | -0.012 |
Fl | -0.012 |
Token | inside |
Token ID | 2641 |
Feature activation | -1.67e-02 |
Pos logit contributions | |
Brown | +0.007 |
seek | +0.007 |
Buzz | +0.007 |
ide | +0.006 |
lf | +0.006 |
Neg logit contributions | |
happens | -0.009 |
ield | -0.009 |
happened | -0.008 |
Dan | -0.008 |
David | -0.007 |
Token | wanted |
Token ID | 2227 |
Feature activation | +2.61e-02 |
Pos logit contributions | |
Brown | +0.005 |
Buzz | +0.005 |
seek | +0.005 |
| | +0.005 |
Ruby | +0.004 |
Neg logit contributions | |
happens | -0.004 |
happened | -0.003 |
ield | -0.003 |
Dan | -0.003 |
meant | -0.003 |
Token | to |
Token ID | 284 |
Feature activation | -4.90e-02 |
Pos logit contributions | |
ield | +0.045 |
happens | +0.045 |
happened | +0.042 |
oke | +0.039 |
Dan | +0.038 |
Neg logit contributions | |
Brown | -0.044 |
Buzz | -0.043 |
seek | -0.042 |
blankets | -0.042 |
sleep | -0.042 |
Token | shut |
Token ID | 4423 |
Feature activation | +3.46e-03 |
Pos logit contributions | |
Brown | +0.072 |
Ell | +0.060 |
Ruby | +0.059 |
lf | +0.059 |
Mic | +0.058 |
Neg logit contributions | |
happens | -0.101 |
ield | -0.100 |
happened | -0.098 |
Dan | -0.097 |
It | -0.092 |
Token | the |
Token ID | 262 |
Feature activation | -4.11e-03 |
Pos logit contributions | |
happens | +0.006 |
happened | +0.006 |
ield | +0.006 |
Dan | +0.005 |
David | +0.005 |
Neg logit contributions | |
Brown | -0.006 |
Buzz | -0.006 |
seek | -0.006 |
Ruby | -0.006 |
| | -0.006 |
Token | cup |
Token ID | 6508 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
Brown | +0.006 |
carpet | +0.006 |
Fl | +0.006 |
Buzz | +0.006 |
shelf | +0.006 |
Neg logit contributions | |
happens | -0.008 |
happened | -0.008 |
ield | -0.008 |
Dan | -0.007 |
David | -0.007 |
Token | though |
Token ID | 996 |
Feature activation | +2.37e-02 |
Pos logit contributions | |
Brown | +0.023 |
Buzz | +0.022 |
seek | +0.021 |
Ruby | +0.021 |
ide | +0.021 |
Neg logit contributions | |
happens | -0.021 |
happened | -0.020 |
ield | -0.020 |
Dan | -0.018 |
David | -0.017 |
Token | and |
Token ID | 290 |
Feature activation | -1.78e-02 |
Pos logit contributions | |
happens | +0.043 |
ield | +0.041 |
happened | +0.039 |
Dan | +0.036 |
oke | +0.036 |
Neg logit contributions | |
Brown | -0.039 |
Buzz | -0.037 |
seek | -0.036 |
lf | -0.036 |
sleep | -0.036 |
Token | it |
Token ID | 340 |
Feature activation | +2.82e-02 |
Pos logit contributions | |
Brown | +0.033 |
Buzz | +0.031 |
Ruby | +0.030 |
| | +0.030 |
Later | +0.030 |
Neg logit contributions | |
happens | -0.030 |
happened | -0.029 |
ield | -0.028 |
Dan | -0.027 |
meant | -0.025 |
Token | was |
Token ID | 373 |
Feature activation | +3.84e-02 |
Pos logit contributions | |
happens | +0.043 |
ield | +0.042 |
happened | +0.040 |
Dan | +0.038 |
David | +0.037 |
Neg logit contributions | |
Brown | -0.056 |
Buzz | -0.053 |
seek | -0.052 |
Fl | -0.052 |
Ruby | -0.051 |
Token | very |
Token ID | 845 |
Feature activation | +1.28e-02 |
Pos logit contributions | |
happens | +0.067 |
happened | +0.065 |
ield | +0.064 |
Dan | +0.058 |
meant | +0.057 |
Neg logit contributions | |
Brown | -0.068 |
Buzz | -0.064 |
Ruby | -0.061 |
| | -0.061 |
seek | -0.060 |
Token | hard |
Token ID | 1327 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
happens | +0.025 |
happened | +0.024 |
ield | +0.024 |
Dan | +0.022 |
meant | +0.022 |
Neg logit contributions | |
Brown | -0.020 |
Buzz | -0.019 |
Ruby | -0.018 |
| | -0.018 |
Hope | -0.018 |
Token | heard |
Token ID | 2982 |
Feature activation | +1.36e-02 |
Pos logit contributions | |
Brown | +0.044 |
Buzz | +0.042 |
seek | +0.040 |
Ruby | +0.040 |
| | +0.040 |
Neg logit contributions | |
happens | -0.032 |
happened | -0.031 |
ield | -0.030 |
Dan | -0.027 |
meant | -0.027 |
Token | mom |
Token ID | 1995 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
happens | +0.025 |
ield | +0.024 |
happened | +0.024 |
Dan | +0.021 |
meant | +0.021 |
Neg logit contributions | |
Brown | -0.022 |
Buzz | -0.021 |
Later | -0.020 |
Ruby | -0.020 |
Fl | -0.020 |
Token | coming |
Token ID | 2406 |
Feature activation | -1.79e-02 |
Pos logit contributions | |
Brown | +0.026 |
Buzz | +0.025 |
seek | +0.025 |
Fl | +0.023 |
Ruby | +0.023 |
Neg logit contributions | |
happens | -0.019 |
happened | -0.018 |
ield | -0.018 |
meant | -0.017 |
Dan | -0.016 |
Token | back |
Token ID | 736 |
Feature activation | -2.76e-02 |
Pos logit contributions | |
Brown | +0.042 |
Buzz | +0.040 |
lf | +0.039 |
Mic | +0.039 |
Mid | +0.038 |
Neg logit contributions | |
happens | -0.020 |
Dan | -0.020 |
ield | -0.019 |
happened | -0.019 |
Joe | -0.017 |
Token | and |
Token ID | 290 |
Feature activation | -4.30e-02 |
Pos logit contributions | |
Brown | +0.056 |
Buzz | +0.054 |
Mid | +0.051 |
Ruby | +0.050 |
Later | +0.050 |
Neg logit contributions | |
happens | -0.042 |
happened | -0.041 |
ield | -0.040 |
Dan | -0.039 |
meant | -0.036 |
Token | ran |
Token ID | 4966 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
Brown | +0.082 |
Buzz | +0.079 |
Later | +0.074 |
Ruby | +0.074 |
Fl | +0.074 |
Neg logit contributions | |
happens | -0.069 |
happened | -0.068 |
ield | -0.065 |
Dan | -0.062 |
David | -0.059 |
Token | to |
Token ID | 284 |
Feature activation | -4.22e-02 |
Pos logit contributions | |
Brown | +0.025 |
Buzz | +0.024 |
lf | +0.023 |
Ruby | +0.023 |
Fl | +0.023 |
Neg logit contributions | |
happens | -0.013 |
happened | -0.012 |
ield | -0.012 |
Dan | -0.011 |
meant | -0.010 |
Token | hide |
Token ID | 7808 |
Feature activation | -2.03e-02 |
Pos logit contributions | |
Brown | +0.081 |
Buzz | +0.076 |
Later | +0.072 |
lf | +0.071 |
Fl | +0.071 |
Neg logit contributions | |
happens | -0.068 |
happened | -0.067 |
Dan | -0.066 |
ield | -0.065 |
David | -0.062 |
Token | under |
Token ID | 739 |
Feature activation | -1.88e-02 |
Pos logit contributions | |
Brown | +0.038 |
Buzz | +0.035 |
| | +0.034 |
Ruby | +0.034 |
seek | +0.034 |
Neg logit contributions | |
happens | -0.034 |
ield | -0.032 |
happened | -0.032 |
Dan | -0.031 |
meant | -0.028 |
Token | the |
Token ID | 262 |
Feature activation | -2.52e-02 |
Pos logit contributions | |
Brown | +0.037 |
Buzz | +0.035 |
Later | +0.034 |
heads | +0.033 |
ide | +0.033 |
Neg logit contributions | |
happens | -0.029 |
happened | -0.029 |
ield | -0.027 |
Dan | -0.027 |
David | -0.025 |
Token | table |
Token ID | 3084 |
Feature activation | -3.16e-02 |
Pos logit contributions | |
Buzz | +0.039 |
Brown | +0.039 |
| | +0.036 |
seek | +0.036 |
Tweet | +0.036 |
Neg logit contributions | |
happens | -0.051 |
happened | -0.049 |
ield | -0.047 |
Dan | -0.042 |
It | -0.042 |
Token | |
Token ID | 220 |
Feature activation | +1.52e-02 |
Pos logit contributions | |
happens | +0.011 |
ield | +0.011 |
happened | +0.011 |
Dan | +0.010 |
meant | +0.009 |
Neg logit contributions | |
Brown | -0.014 |
Buzz | -0.014 |
seek | -0.013 |
Ruby | -0.013 |
| | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -4.81e-02 |
Pos logit contributions | |
happens | +0.027 |
ield | +0.025 |
happened | +0.025 |
Dan | +0.023 |
meant | +0.022 |
Neg logit contributions | |
Brown | -0.027 |
Buzz | -0.025 |
Ruby | -0.024 |
seek | -0.024 |
Later | -0.024 |
Token |
|
Token ID | 198 |
Feature activation | -5.18e-02 |
Pos logit contributions | |
throne | +0.114 |
blankets | +0.112 |
carpet | +0.112 |
fort | +0.111 |
cushion | +0.110 |
Neg logit contributions | |
It | -0.043 |
happened | -0.042 |
Dan | -0.042 |
happens | -0.040 |
ield | -0.039 |
Token | Her |
Token ID | 9360 |
Feature activation | +3.90e-03 |
Pos logit contributions | |
throne | +0.089 |
blankets | +0.088 |
hind | +0.085 |
fort | +0.082 |
carpet | +0.081 |
Neg logit contributions | |
ield | -0.083 |
happened | -0.079 |
It | -0.078 |
happens | -0.076 |
No | -0.075 |
Token | mom |
Token ID | 1995 |
Feature activation | +3.74e-03 |
Pos logit contributions | |
happens | +0.007 |
ield | +0.007 |
happened | +0.007 |
Dan | +0.006 |
meant | +0.006 |
Neg logit contributions | |
Brown | -0.007 |
Buzz | -0.006 |
Ruby | -0.006 |
| | -0.006 |
seek | -0.006 |
Token | came |
Token ID | 1625 |
Feature activation | -1.25e-02 |
Pos logit contributions | |
happens | +0.006 |
happened | +0.005 |
ield | +0.005 |
Dan | +0.005 |
meant | +0.005 |
Neg logit contributions | |
Brown | -0.007 |
Buzz | -0.007 |
seek | -0.007 |
| | -0.007 |
Ruby | -0.007 |
Token | outside |
Token ID | 2354 |
Feature activation | -2.03e-02 |
Pos logit contributions | |
Brown | +0.027 |
Buzz | +0.026 |
Blue | +0.025 |
Fl | +0.025 |
lf | +0.025 |
Neg logit contributions | |
Dan | -0.015 |
happened | -0.015 |
ield | -0.014 |
David | -0.014 |
happens | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | -3.02e-02 |
Pos logit contributions | |
Brown | +0.041 |
Buzz | +0.038 |
Fl | +0.036 |
Mic | +0.036 |
Ruby | +0.036 |
Neg logit contributions | |
happened | -0.031 |
happens | -0.031 |
ield | -0.030 |
Dan | -0.028 |
meant | -0.027 |
Token | said |
Token ID | 531 |
Feature activation | +4.02e-02 |
Pos logit contributions | |
Brown | +0.066 |
Buzz | +0.064 |
ide | +0.062 |
| | +0.062 |
hind | +0.062 |
Neg logit contributions | |
happened | -0.040 |
happens | -0.039 |
Dan | -0.033 |
ield | -0.033 |
meant | -0.033 |
Token | , |
Token ID | 11 |
Feature activation | -2.35e-03 |
Pos logit contributions | |
ield | +0.071 |
happens | +0.068 |
Dan | +0.067 |
meant | +0.063 |
happened | +0.063 |
Neg logit contributions | |
Brown | -0.061 |
Buzz | -0.060 |
Mid | -0.058 |
Fl | -0.058 |
Ruby | -0.058 |
Token | " |
Token ID | 366 |
Feature activation | +2.01e-02 |
Pos logit contributions | |
Brown | +0.005 |
Later | +0.005 |
Fl | +0.005 |
Buzz | +0.005 |
Mid | +0.005 |
Neg logit contributions | |
happens | -0.003 |
ield | -0.003 |
happened | -0.003 |
Dan | -0.003 |
meant | -0.002 |
Token | there |
Token ID | 612 |
Feature activation | +2.39e-02 |
Pos logit contributions | |
happens | +0.032 |
ield | +0.031 |
happened | +0.030 |
Dan | +0.028 |
meant | +0.026 |
Neg logit contributions | |
Brown | -0.034 |
Buzz | -0.032 |
seek | -0.031 |
Ruby | -0.031 |
| | -0.031 |
Token | was |
Token ID | 373 |
Feature activation | +2.47e-02 |
Pos logit contributions | |
ield | +0.047 |
happens | +0.047 |
oke | +0.040 |
fault | +0.040 |
Dan | +0.039 |
Neg logit contributions | |
ide | -0.034 |
| | -0.034 |
Fl | -0.034 |
Brown | -0.033 |
Buzz | -0.033 |
Token | a |
Token ID | 257 |
Feature activation | -2.59e-02 |
Pos logit contributions | |
happens | +0.033 |
happened | +0.033 |
ield | +0.031 |
Dan | +0.029 |
meant | +0.027 |
Neg logit contributions | |
Brown | -0.053 |
Buzz | -0.049 |
seek | -0.048 |
heads | -0.048 |
Ruby | -0.048 |
Token | very |
Token ID | 845 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
| | +0.043 |
Brown | +0.042 |
ge | +0.041 |
ide | +0.041 |
Later | +0.040 |
Neg logit contributions | |
happened | -0.043 |
happens | -0.042 |
ield | -0.041 |
Dan | -0.040 |
David | -0.038 |
Token | big |
Token ID | 1263 |
Feature activation | -4.10e-03 |
Pos logit contributions | |
Brown | +0.023 |
Buzz | +0.021 |
| | +0.021 |
Fl | +0.021 |
Ruby | +0.020 |
Neg logit contributions | |
happens | -0.022 |
happened | -0.021 |
ield | -0.021 |
Dan | -0.019 |
meant | -0.018 |
Token | family |
Token ID | 1641 |
Feature activation | -1.20e-02 |
Pos logit contributions | |
Brown | +0.007 |
Buzz | +0.006 |
Later | +0.006 |
Ruby | +0.006 |
| | +0.006 |
Neg logit contributions | |
happens | -0.008 |
ield | -0.008 |
happened | -0.008 |
Dan | -0.007 |
David | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | +1.94e-02 |
Pos logit contributions | |
Brown | +0.023 |
Buzz | +0.022 |
Ruby | +0.022 |
heads | +0.021 |
hind | +0.021 |
Neg logit contributions | |
ield | -0.019 |
happened | -0.018 |
happens | -0.018 |
Dan | -0.017 |
meant | -0.016 |
Token | It |
Token ID | 632 |
Feature activation | +1.50e-02 |
Pos logit contributions | |
happens | +0.037 |
happened | +0.034 |
ield | +0.034 |
oke | +0.030 |
fault | +0.030 |
Neg logit contributions | |
Brown | -0.031 |
| | -0.031 |
Buzz | -0.029 |
seek | -0.029 |
shelf | -0.029 |
Token | included |
Token ID | 3017 |
Feature activation | -1.47e-02 |
Pos logit contributions | |
happens | +0.022 |
ield | +0.022 |
happened | +0.021 |
Dan | +0.019 |
David | +0.018 |
Neg logit contributions | |
Brown | -0.030 |
Buzz | -0.029 |
| | -0.028 |
seek | -0.028 |
Ruby | -0.028 |
Token | two |
Token ID | 734 |
Feature activation | -1.84e-03 |
Pos logit contributions | |
Brown | +0.023 |
Buzz | +0.021 |
Ruby | +0.020 |
seek | +0.020 |
| | +0.020 |
Neg logit contributions | |
happens | -0.029 |
ield | -0.028 |
happened | -0.028 |
Dan | -0.026 |
meant | -0.024 |
Token | mom |
Token ID | 1995 |
Feature activation | +7.28e-03 |
Pos logit contributions | |
Brown | +0.003 |
Buzz | +0.003 |
seek | +0.002 |
Ruby | +0.002 |
| | +0.002 |
Neg logit contributions | |
happens | -0.004 |
ield | -0.004 |
happened | -0.004 |
Dan | -0.003 |
meant | -0.003 |
Token | ! |
Token ID | 0 |
Feature activation | +1.33e-02 |
Pos logit contributions | |
happens | +0.023 |
ield | +0.022 |
happened | +0.022 |
Dan | +0.020 |
meant | +0.019 |
Neg logit contributions | |
Brown | -0.020 |
Buzz | -0.019 |
seek | -0.018 |
Ruby | -0.018 |
Fl | -0.018 |
Token | |
Token ID | 220 |
Feature activation | +1.44e-02 |
Pos logit contributions | |
happens | +0.021 |
ield | +0.020 |
happened | +0.020 |
Dan | +0.019 |
David | +0.017 |
Neg logit contributions | |
Brown | -0.025 |
Buzz | -0.024 |
seek | -0.023 |
Ruby | -0.023 |
Later | -0.022 |
Token |
|
Token ID | 198 |
Feature activation | -3.94e-02 |
Pos logit contributions | |
happens | +0.025 |
ield | +0.024 |
happened | +0.024 |
Dan | +0.022 |
meant | +0.021 |
Neg logit contributions | |
Brown | -0.025 |
Buzz | -0.024 |
Ruby | -0.023 |
seek | -0.023 |
| | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | -4.54e-02 |
Pos logit contributions | |
throne | +0.078 |
Brown | +0.076 |
seek | +0.076 |
carpet | +0.076 |
fort | +0.075 |
Neg logit contributions | |
happens | -0.049 |
happened | -0.049 |
It | -0.047 |
Dan | -0.047 |
ield | -0.045 |
Token | Jimmy |
Token ID | 40335 |
Feature activation | +5.99e-05 |
Pos logit contributions | |
throne | +0.073 |
carpet | +0.069 |
hind | +0.069 |
fort | +0.068 |
cushion | +0.068 |
Neg logit contributions | |
happens | -0.074 |
ield | -0.074 |
happened | -0.072 |
It | -0.068 |
Dan | -0.067 |
Token | and |
Token ID | 290 |
Feature activation | -2.20e-02 |
Pos logit contributions | |
happens | +0.000 |
ield | +0.000 |
happened | +0.000 |
Dan | +0.000 |
meant | +0.000 |
Neg logit contributions | |
Brown | -0.000 |
Buzz | -0.000 |
seek | -0.000 |
Ruby | -0.000 |
| | -0.000 |
Token | his |
Token ID | 465 |
Feature activation | -6.38e-03 |
Pos logit contributions | |
Brown | +0.032 |
Buzz | +0.030 |
Later | +0.028 |
Ruby | +0.028 |
Fl | +0.028 |
Neg logit contributions | |
happens | -0.045 |
ield | -0.044 |
happened | -0.044 |
Dan | -0.042 |
David | -0.040 |
Token | family |
Token ID | 1641 |
Feature activation | -2.55e-02 |
Pos logit contributions | |
Brown | +0.011 |
Buzz | +0.010 |
Ruby | +0.009 |
seek | +0.009 |
| | +0.009 |
Neg logit contributions | |
happens | -0.012 |
ield | -0.011 |
happened | -0.011 |
Dan | -0.011 |
meant | -0.010 |
Token | thanked |
Token ID | 26280 |
Feature activation | -2.47e-02 |
Pos logit contributions | |
Brown | +0.051 |
Buzz | +0.049 |
Ruby | +0.047 |
seek | +0.047 |
Later | +0.047 |
Neg logit contributions | |
happens | -0.038 |
happened | -0.036 |
ield | -0.036 |
Dan | -0.033 |
meant | -0.031 |
Token | the |
Token ID | 262 |
Feature activation | +8.52e-03 |
Pos logit contributions | |
Brown | +0.043 |
Buzz | +0.041 |
seek | +0.040 |
Ruby | +0.039 |
Later | +0.039 |
Neg logit contributions | |
happens | -0.043 |
ield | -0.041 |
happened | -0.041 |
Dan | -0.039 |
David | -0.037 |
Token | police |
Token ID | 1644 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
happens | +0.013 |
ield | +0.012 |
happened | +0.012 |
Dan | +0.011 |
meant | +0.011 |
Neg logit contributions | |
Brown | -0.017 |
Buzz | -0.016 |
seek | -0.015 |
Ruby | -0.015 |
| | -0.015 |
Token | played |
Token ID | 2826 |
Feature activation | -3.10e-02 |
Pos logit contributions | |
Brown | +0.011 |
Buzz | +0.010 |
Ruby | +0.010 |
Later | +0.010 |
seek | +0.010 |
Neg logit contributions | |
happens | -0.010 |
happened | -0.009 |
ield | -0.009 |
Dan | -0.008 |
meant | -0.008 |
Token | with |
Token ID | 351 |
Feature activation | -3.90e-02 |
Pos logit contributions | |
Brown | +0.051 |
Buzz | +0.046 |
| | +0.044 |
Hope | +0.044 |
Ruby | +0.043 |
Neg logit contributions | |
happens | -0.059 |
happened | -0.057 |
ield | -0.056 |
Dan | -0.053 |
It | -0.049 |
Token | the |
Token ID | 262 |
Feature activation | -4.57e-03 |
Pos logit contributions | |
Brown | +0.067 |
Buzz | +0.062 |
Later | +0.060 |
seek | +0.060 |
| | +0.059 |
Neg logit contributions | |
happens | -0.071 |
happened | -0.068 |
ield | -0.067 |
Dan | -0.064 |
David | -0.059 |
Token | dough |
Token ID | 15756 |
Feature activation | -3.20e-02 |
Pos logit contributions | |
Brown | +0.007 |
Later | +0.007 |
Buzz | +0.006 |
| | +0.006 |
heads | +0.006 |
Neg logit contributions | |
happened | -0.009 |
happens | -0.009 |
ield | -0.008 |
Dan | -0.008 |
meant | -0.008 |
Token | ball |
Token ID | 2613 |
Feature activation | -3.65e-02 |
Pos logit contributions | |
Brown | +0.062 |
Buzz | +0.057 |
Later | +0.055 |
| | +0.055 |
Hope | +0.055 |
Neg logit contributions | |
happens | -0.051 |
happened | -0.049 |
ield | -0.048 |
Dan | -0.045 |
meant | -0.043 |
Token | together |
Token ID | 1978 |
Feature activation | -3.67e-02 |
Pos logit contributions | |
Brown | +0.068 |
Buzz | +0.063 |
Later | +0.060 |
Hope | +0.060 |
heads | +0.060 |
Neg logit contributions | |
happens | -0.060 |
happened | -0.057 |
ield | -0.057 |
Dan | -0.052 |
meant | -0.050 |
Token | . |
Token ID | 13 |
Feature activation | +3.55e-03 |
Pos logit contributions | |
Brown | +0.071 |
Buzz | +0.066 |
| | +0.064 |
Ruby | +0.063 |
Later | +0.063 |
Neg logit contributions | |
happens | -0.059 |
happened | -0.057 |
ield | -0.056 |
Dan | -0.052 |
meant | -0.049 |
Token | They |
Token ID | 1119 |
Feature activation | -2.85e-02 |
Pos logit contributions | |
happens | +0.005 |
ield | +0.005 |
happened | +0.005 |
Dan | +0.005 |
David | +0.004 |
Neg logit contributions | |
Brown | -0.007 |
Buzz | -0.007 |
seek | -0.006 |
Ruby | -0.006 |
Later | -0.006 |
Token | had |
Token ID | 550 |
Feature activation | +2.49e-02 |
Pos logit contributions | |
Brown | +0.061 |
heads | +0.058 |
Buzz | +0.057 |
Later | +0.057 |
Ruby | +0.055 |
Neg logit contributions | |
happens | -0.037 |
happened | -0.037 |
Dan | -0.034 |
ield | -0.033 |
meant | -0.032 |
Token | lots |
Token ID | 6041 |
Feature activation | +5.56e-03 |
Pos logit contributions | |
happens | +0.041 |
ield | +0.040 |
happened | +0.038 |
Dan | +0.035 |
meant | +0.033 |
Neg logit contributions | |
Brown | -0.046 |
Buzz | -0.044 |
seek | -0.042 |
Ruby | -0.042 |
Fl | -0.042 |
Token | of |
Token ID | 286 |
Feature activation | -1.43e-02 |
Pos logit contributions | |
happens | +0.008 |
ield | +0.008 |
happened | +0.007 |
Dan | +0.007 |
meant | +0.006 |
Neg logit contributions | |
Brown | -0.012 |
Buzz | -0.011 |
seek | -0.011 |
Ruby | -0.011 |
| | -0.010 |
Token | the |
Token ID | 262 |
Feature activation | -3.03e-02 |
Pos logit contributions | |
Brown | +0.028 |
Buzz | +0.026 |
seek | +0.025 |
Ruby | +0.025 |
heads | +0.025 |
Neg logit contributions | |
happens | -0.027 |
happened | -0.026 |
ield | -0.026 |
Dan | -0.025 |
meant | -0.023 |
Token | sea |
Token ID | 5417 |
Feature activation | -3.87e-02 |
Pos logit contributions | |
Brown | +0.045 |
heads | +0.039 |
Buzz | +0.039 |
Ruby | +0.039 |
lf | +0.039 |
Neg logit contributions | |
happens | -0.062 |
ield | -0.060 |
happened | -0.058 |
Dan | -0.057 |
David | -0.053 |
Token | . |
Token ID | 13 |
Feature activation | +2.96e-05 |
Pos logit contributions | |
Brown | +0.076 |
Buzz | +0.069 |
Ruby | +0.068 |
Mid | +0.067 |
Fl | +0.067 |
Neg logit contributions | |
happens | -0.062 |
ield | -0.060 |
happened | -0.059 |
Dan | -0.055 |
meant | -0.051 |
Token | They |
Token ID | 1119 |
Feature activation | -1.95e-02 |
Pos logit contributions | |
happens | +0.000 |
ield | +0.000 |
happened | +0.000 |
Dan | +0.000 |
meant | +0.000 |
Neg logit contributions | |
Brown | -0.000 |
Buzz | -0.000 |
Fl | -0.000 |
Later | -0.000 |
| | -0.000 |
Token | wanted |
Token ID | 2227 |
Feature activation | +1.56e-02 |
Pos logit contributions | |
Brown | +0.033 |
Buzz | +0.032 |
seek | +0.031 |
| | +0.030 |
Ruby | +0.030 |
Neg logit contributions | |
happens | -0.035 |
ield | -0.034 |
happened | -0.033 |
Dan | -0.030 |
meant | -0.028 |
Token | to |
Token ID | 284 |
Feature activation | -5.72e-02 |
Pos logit contributions | |
ield | +0.030 |
happens | +0.029 |
oke | +0.027 |
ement | +0.027 |
happened | +0.027 |
Neg logit contributions | |
Buzz | -0.023 |
walk | -0.023 |
seek | -0.023 |
Brown | -0.022 |
sleep | -0.022 |
Token | go |
Token ID | 467 |
Feature activation | -2.88e-02 |
Pos logit contributions | |
Brown | +0.091 |
Ruby | +0.082 |
Mid | +0.081 |
Fl | +0.081 |
Buzz | +0.079 |
Neg logit contributions | |
happens | -0.111 |
ield | -0.108 |
happened | -0.107 |
Dan | -0.106 |
Bill | -0.098 |
Token | on |
Token ID | 319 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
Brown | +0.051 |
Buzz | +0.048 |
seek | +0.047 |
| | +0.046 |
Ruby | +0.046 |
Neg logit contributions | |
happens | -0.050 |
ield | -0.048 |
happened | -0.047 |
Dan | -0.043 |
meant | -0.041 |
Token | an |
Token ID | 281 |
Feature activation | -3.94e-02 |
Pos logit contributions | |
Brown | +0.037 |
Buzz | +0.035 |
seek | +0.034 |
Ruby | +0.034 |
| | +0.034 |
Neg logit contributions | |
happens | -0.023 |
ield | -0.022 |
happened | -0.022 |
Dan | -0.019 |
meant | -0.018 |
Token | adventure |
Token ID | 8855 |
Feature activation | -2.56e-02 |
Pos logit contributions | |
Brown | +0.064 |
Ruby | +0.059 |
Fl | +0.058 |
Buzz | +0.057 |
| | +0.056 |
Neg logit contributions | |
happens | -0.075 |
ield | -0.072 |
happened | -0.072 |
Dan | -0.070 |
David | -0.066 |
Token | , |
Token ID | 11 |
Feature activation | -1.65e-02 |
Pos logit contributions | |
Brown | +0.043 |
seek | +0.041 |
Buzz | +0.041 |
ide | +0.040 |
| | +0.039 |
Neg logit contributions | |
happens | -0.047 |
ield | -0.044 |
happened | -0.043 |
Dan | -0.039 |
oke | -0.038 |
Token | friend |
Token ID | 1545 |
Feature activation | +1.49e-02 |
Pos logit contributions | |
Brown | +0.001 |
Buzz | +0.001 |
seek | +0.001 |
Ruby | +0.001 |
| | +0.001 |
Neg logit contributions | |
happens | -0.002 |
ield | -0.002 |
happened | -0.002 |
Dan | -0.002 |
meant | -0.002 |
Token | , |
Token ID | 11 |
Feature activation | -1.18e-02 |
Pos logit contributions | |
happens | +0.026 |
ield | +0.026 |
happened | +0.023 |
ement | +0.022 |
oke | +0.022 |
Neg logit contributions | |
Brown | -0.023 |
Buzz | -0.023 |
walk | -0.023 |
ie | -0.022 |
ide | -0.022 |
Token | Jim |
Token ID | 5395 |
Feature activation | +9.77e-04 |
Pos logit contributions | |
Brown | +0.023 |
Buzz | +0.022 |
seek | +0.022 |
Ruby | +0.021 |
| | +0.021 |
Neg logit contributions | |
happens | -0.019 |
ield | -0.018 |
happened | -0.017 |
Dan | -0.015 |
meant | -0.014 |
Token | , |
Token ID | 11 |
Feature activation | -6.93e-03 |
Pos logit contributions | |
happens | +0.002 |
ield | +0.002 |
happened | +0.001 |
Dan | +0.001 |
meant | +0.001 |
Neg logit contributions | |
Brown | -0.002 |
Buzz | -0.002 |
seek | -0.002 |
Ruby | -0.002 |
Fl | -0.002 |
Token | decided |
Token ID | 3066 |
Feature activation | +2.29e-02 |
Pos logit contributions | |
Brown | +0.012 |
Buzz | +0.012 |
seek | +0.011 |
Ruby | +0.011 |
| | +0.011 |
Neg logit contributions | |
happens | -0.012 |
ield | -0.012 |
happened | -0.010 |
Dan | -0.009 |
oke | -0.009 |
Token | to |
Token ID | 284 |
Feature activation | -6.24e-02 |
Pos logit contributions | |
ield | +0.042 |
happens | +0.042 |
oke | +0.040 |
happened | +0.038 |
ement | +0.038 |
Neg logit contributions | |
seek | -0.034 |
walk | -0.033 |
Brown | -0.032 |
Later | -0.032 |
Buzz | -0.030 |
Token | go |
Token ID | 467 |
Feature activation | -4.35e-02 |
Pos logit contributions | |
Brown | +0.081 |
Buzz | +0.075 |
Ruby | +0.070 |
seek | +0.070 |
Later | +0.069 |
Neg logit contributions | |
happens | -0.137 |
ield | -0.133 |
happened | -0.132 |
Dan | -0.124 |
meant | -0.119 |
Token | for |
Token ID | 329 |
Feature activation | -4.17e-02 |
Pos logit contributions | |
Brown | +0.083 |
Buzz | +0.079 |
seek | +0.076 |
Ruby | +0.076 |
| | +0.075 |
Neg logit contributions | |
happens | -0.069 |
ield | -0.066 |
happened | -0.066 |
Dan | -0.060 |
meant | -0.056 |
Token | a |
Token ID | 257 |
Feature activation | -4.23e-02 |
Pos logit contributions | |
Brown | +0.077 |
Buzz | +0.073 |
seek | +0.072 |
| | +0.071 |
Ruby | +0.071 |
Neg logit contributions | |
happens | -0.068 |
ield | -0.066 |
happened | -0.064 |
Dan | -0.058 |
meant | -0.056 |
Token | walk |
Token ID | 2513 |
Feature activation | -2.53e-02 |
Pos logit contributions | |
Brown | +0.040 |
Buzz | +0.033 |
Later | +0.031 |
Fl | +0.030 |
charged | +0.030 |
Neg logit contributions | |
happens | -0.108 |
happened | -0.108 |
ield | -0.103 |
Dan | -0.099 |
David | -0.097 |
Token | . |
Token ID | 13 |
Feature activation | +3.57e-03 |
Pos logit contributions | |
Brown | +0.042 |
Buzz | +0.039 |
seek | +0.038 |
Ruby | +0.037 |
| | +0.037 |
Neg logit contributions | |
happens | -0.047 |
ield | -0.045 |
happened | -0.045 |
Dan | -0.041 |
meant | -0.039 |
Token | on |
Token ID | 319 |
Feature activation | -2.26e-02 |
Pos logit contributions | |
happens | +0.019 |
ield | +0.018 |
happened | +0.017 |
Dan | +0.017 |
David | +0.016 |
Neg logit contributions | |
Brown | -0.016 |
lf | -0.015 |
seek | -0.015 |
carpet | -0.014 |
cushion | -0.014 |
Token | it |
Token ID | 340 |
Feature activation | -4.84e-03 |
Pos logit contributions | |
Brown | +0.036 |
Buzz | +0.033 |
seek | +0.032 |
Ruby | +0.032 |
| | +0.031 |
Neg logit contributions | |
happens | -0.044 |
ield | -0.042 |
happened | -0.042 |
Dan | -0.039 |
meant | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | +1.86e-02 |
Pos logit contributions | |
Brown | +0.007 |
seek | +0.007 |
Buzz | +0.007 |
lf | +0.007 |
Later | +0.007 |
Neg logit contributions | |
happens | -0.009 |
ield | -0.009 |
happened | -0.009 |
Dan | -0.008 |
David | -0.008 |
Token | They |
Token ID | 1119 |
Feature activation | -2.24e-02 |
Pos logit contributions | |
happens | +0.033 |
ield | +0.033 |
happened | +0.032 |
meant | +0.028 |
Dan | +0.028 |
Neg logit contributions | |
Buzz | -0.031 |
Brown | -0.030 |
Later | -0.029 |
Ruby | -0.029 |
Fl | -0.029 |
Token | wanted |
Token ID | 2227 |
Feature activation | +1.30e-02 |
Pos logit contributions | |
Brown | +0.049 |
Buzz | +0.047 |
| | +0.046 |
Ruby | +0.046 |
seek | +0.046 |
Neg logit contributions | |
happens | -0.029 |
happened | -0.028 |
ield | -0.027 |
Dan | -0.025 |
meant | -0.023 |
Token | to |
Token ID | 284 |
Feature activation | -6.17e-02 |
Pos logit contributions | |
happens | +0.023 |
ield | +0.022 |
happened | +0.021 |
oke | +0.020 |
Dan | +0.019 |
Neg logit contributions | |
Brown | -0.022 |
Buzz | -0.020 |
seek | -0.020 |
blankets | -0.020 |
sleep | -0.019 |
Token | go |
Token ID | 467 |
Feature activation | -2.82e-02 |
Pos logit contributions | |
Brown | +0.107 |
Mid | +0.105 |
Ruby | +0.101 |
Buzz | +0.100 |
Fl | +0.100 |
Neg logit contributions | |
happens | -0.108 |
ield | -0.106 |
Dan | -0.104 |
happened | -0.103 |
Bill | -0.097 |
Token | near |
Token ID | 1474 |
Feature activation | -2.78e-03 |
Pos logit contributions | |
Brown | +0.054 |
Buzz | +0.051 |
Ruby | +0.049 |
Mid | +0.049 |
Fl | +0.049 |
Neg logit contributions | |
happens | -0.045 |
ield | -0.043 |
happened | -0.043 |
Dan | -0.040 |
meant | -0.037 |
Token | it |
Token ID | 340 |
Feature activation | -3.65e-02 |
Pos logit contributions | |
Brown | +0.004 |
walk | +0.004 |
Buzz | +0.004 |
blankets | +0.004 |
lf | +0.004 |
Neg logit contributions | |
happens | -0.005 |
ield | -0.005 |
happened | -0.005 |
oke | -0.004 |
Well | -0.004 |
Token | and |
Token ID | 290 |
Feature activation | -4.96e-02 |
Pos logit contributions | |
Brown | +0.076 |
Buzz | +0.073 |
Mid | +0.070 |
Ruby | +0.069 |
| | +0.069 |
Neg logit contributions | |
happens | -0.054 |
ield | -0.051 |
happened | -0.050 |
Dan | -0.047 |
meant | -0.046 |
Token | see |
Token ID | 766 |
Feature activation | +5.26e-03 |
Pos logit contributions | |
Brown | +0.086 |
Buzz | +0.081 |
Mid | +0.078 |
Ruby | +0.078 |
Fl | +0.077 |
Neg logit contributions | |
happens | -0.090 |
happened | -0.085 |
ield | -0.084 |
Dan | -0.082 |
meant | -0.076 |
Token | said |
Token ID | 531 |
Feature activation | +2.78e-02 |
Pos logit contributions | |
Brown | +0.014 |
Buzz | +0.013 |
seek | +0.013 |
Ruby | +0.013 |
Later | +0.013 |
Neg logit contributions | |
happens | -0.010 |
ield | -0.010 |
happened | -0.010 |
Dan | -0.009 |
meant | -0.008 |
Token | yes |
Token ID | 3763 |
Feature activation | +1.00e-03 |
Pos logit contributions | |
ield | +0.055 |
happens | +0.055 |
Dan | +0.052 |
ays | +0.052 |
fault | +0.052 |
Neg logit contributions | |
Brown | -0.036 |
Mid | -0.034 |
Bun | -0.034 |
Fl | -0.034 |
Buzz | -0.032 |
Token | and |
Token ID | 290 |
Feature activation | -2.16e-02 |
Pos logit contributions | |
happens | +0.002 |
happened | +0.002 |
ield | +0.002 |
Dan | +0.002 |
Well | +0.002 |
Neg logit contributions | |
Brown | -0.002 |
Later | -0.001 |
ide | -0.001 |
walk | -0.001 |
sleep | -0.001 |
Token | they |
Token ID | 484 |
Feature activation | -3.44e-02 |
Pos logit contributions | |
Brown | +0.036 |
Buzz | +0.033 |
seek | +0.033 |
Ruby | +0.033 |
Fl | +0.032 |
Neg logit contributions | |
happens | -0.039 |
ield | -0.039 |
happened | -0.037 |
Dan | -0.034 |
meant | -0.032 |
Token | started |
Token ID | 2067 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
Brown | +0.058 |
Buzz | +0.056 |
seek | +0.053 |
heads | +0.053 |
Ruby | +0.053 |
Neg logit contributions | |
happens | -0.062 |
happened | -0.060 |
ield | -0.059 |
Dan | -0.055 |
meant | -0.053 |
Token | to |
Token ID | 284 |
Feature activation | -6.55e-02 |
Pos logit contributions | |
Brown | +0.026 |
Buzz | +0.024 |
seek | +0.024 |
Later | +0.023 |
walk | +0.023 |
Neg logit contributions | |
happens | -0.028 |
ield | -0.027 |
happened | -0.027 |
Dan | -0.024 |
meant | -0.024 |
Token | speed |
Token ID | 2866 |
Feature activation | -3.17e-02 |
Pos logit contributions | |
Buzz | +0.080 |
Brown | +0.078 |
Ruby | +0.074 |
heads | +0.072 |
Mid | +0.071 |
Neg logit contributions | |
happens | -0.145 |
ield | -0.140 |
happened | -0.140 |
Dan | -0.137 |
Bill | -0.133 |
Token | down |
Token ID | 866 |
Feature activation | -3.36e-02 |
Pos logit contributions | |
Brown | +0.047 |
Buzz | +0.042 |
seek | +0.042 |
| | +0.042 |
Later | +0.042 |
Neg logit contributions | |
happens | -0.064 |
ield | -0.062 |
happened | -0.061 |
Dan | -0.056 |
meant | -0.054 |
Token | the |
Token ID | 262 |
Feature activation | -2.60e-02 |
Pos logit contributions | |
Brown | +0.042 |
seek | +0.039 |
| | +0.038 |
Buzz | +0.038 |
walk | +0.038 |
Neg logit contributions | |
happens | -0.074 |
ield | -0.073 |
happened | -0.070 |
oke | -0.064 |
Dan | -0.064 |
Token | street |
Token ID | 4675 |
Feature activation | -3.47e-02 |
Pos logit contributions | |
Buzz | +0.032 |
Brown | +0.031 |
Ruby | +0.028 |
Bee | +0.028 |
Mid | +0.028 |
Neg logit contributions | |
happens | -0.060 |
happened | -0.056 |
ield | -0.056 |
Dan | -0.052 |
It | -0.051 |
Token | together |
Token ID | 1978 |
Feature activation | -4.61e-02 |
Pos logit contributions | |
Brown | +0.052 |
seek | +0.048 |
ide | +0.047 |
| | +0.046 |
walk | +0.046 |
Neg logit contributions | |
happens | -0.070 |
ield | -0.067 |
happened | -0.066 |
Dan | -0.060 |
David | -0.058 |
Token | on |
Token ID | 319 |
Feature activation | -1.87e-02 |
Pos logit contributions | |
Brown | +0.050 |
lf | +0.048 |
Mid | +0.045 |
Buzz | +0.045 |
Mic | +0.044 |
Neg logit contributions | |
happens | -0.047 |
happened | -0.041 |
ield | -0.040 |
enough | -0.038 |
It | -0.037 |
Token | your |
Token ID | 534 |
Feature activation | -3.18e-02 |
Pos logit contributions | |
Brown | +0.034 |
Buzz | +0.032 |
| | +0.032 |
Ruby | +0.031 |
seek | +0.031 |
Neg logit contributions | |
happens | -0.032 |
happened | -0.030 |
ield | -0.029 |
Dan | -0.027 |
enough | -0.026 |
Token | own |
Token ID | 898 |
Feature activation | -3.38e-02 |
Pos logit contributions | |
Brown | +0.054 |
| | +0.052 |
Buzz | +0.051 |
Ell | +0.051 |
ide | +0.050 |
Neg logit contributions | |
happens | -0.057 |
ield | -0.053 |
happened | -0.052 |
fault | -0.049 |
enough | -0.048 |
Token | ." |
Token ID | 526 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
Brown | +0.043 |
Buzz | +0.040 |
| | +0.039 |
Ruby | +0.038 |
Ell | +0.037 |
Neg logit contributions | |
happens | -0.076 |
ield | -0.074 |
happened | -0.071 |
Dan | -0.067 |
enough | -0.067 |
Token |
|
Token ID | 198 |
Feature activation | -4.21e-02 |
Pos logit contributions | |
ge | +0.040 |
seek | +0.039 |
Brown | +0.039 |
lf | +0.038 |
heads | +0.038 |
Neg logit contributions | |
It | -0.016 |
happens | -0.014 |
happened | -0.014 |
ield | -0.013 |
Dan | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -7.04e-02 |
Pos logit contributions | |
seek | +0.099 |
Brown | +0.099 |
throne | +0.096 |
heads | +0.095 |
charged | +0.093 |
Neg logit contributions | |
happened | -0.046 |
happens | -0.044 |
It | -0.040 |
Dan | -0.038 |
ield | -0.036 |
Token | L |
Token ID | 43 |
Feature activation | +8.15e-03 |
Pos logit contributions | |
throne | +0.147 |
den | +0.137 |
ge | +0.135 |
charged | +0.129 |
| | +0.128 |
Neg logit contributions | |
aw | -0.081 |
ield | -0.077 |
It | -0.076 |
happens | -0.074 |
� | -0.069 |
Token | ily |
Token ID | 813 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
happens | +0.009 |
happened | +0.008 |
ield | +0.008 |
meant | +0.006 |
oke | +0.006 |
Neg logit contributions | |
Brown | -0.021 |
seek | -0.019 |
Buzz | -0.019 |
throne | -0.019 |
Later | -0.019 |
Token | held |
Token ID | 2714 |
Feature activation | -7.79e-05 |
Pos logit contributions | |
Brown | +0.028 |
Buzz | +0.027 |
| | +0.026 |
seek | +0.026 |
Ruby | +0.026 |
Neg logit contributions | |
happens | -0.021 |
happened | -0.020 |
ield | -0.020 |
Dan | -0.017 |
meant | -0.017 |
Token | the |
Token ID | 262 |
Feature activation | +7.76e-03 |
Pos logit contributions | |
Brown | +0.000 |
Buzz | +0.000 |
seek | +0.000 |
Later | +0.000 |
| | +0.000 |
Neg logit contributions | |
happens | -0.000 |
ield | -0.000 |
happened | -0.000 |
Dan | -0.000 |
meant | -0.000 |
Token | bird |
Token ID | 6512 |
Feature activation | -1.86e-02 |
Pos logit contributions | |
happens | +0.015 |
ield | +0.015 |
happened | +0.015 |
Dan | +0.013 |
meant | +0.013 |
Neg logit contributions | |
Brown | -0.012 |
Buzz | -0.011 |
| | -0.011 |
seek | -0.011 |
Ruby | -0.011 |