Token | tree |
Token ID | 5509 |
Feature activation | +2.26e+00 |
Pos logit contributions | |
ours | +0.434 |
coconut | +0.422 |
Enjoy | +0.412 |
oral | +0.382 |
rust | +0.370 |
Neg logit contributions | |
aces | -0.564 |
amas | -0.551 |
coat | -0.550 |
robe | -0.529 |
forehead | -0.528 |
Token | with |
Token ID | 351 |
Feature activation | -8.89e-01 |
Pos logit contributions | |
ours | +0.142 |
coconut | +0.139 |
Enjoy | +0.132 |
oral | +0.123 |
rust | +0.121 |
Neg logit contributions | |
aces | -0.195 |
amas | -0.192 |
coat | -0.191 |
robe | -0.185 |
forehead | -0.185 |
Token | leaves |
Token ID | 5667 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
aces | +0.071 |
amas | +0.066 |
forehead | +0.065 |
coat | +0.065 |
drops | +0.065 |
Neg logit contributions | |
ours | -0.059 |
oral | -0.057 |
Enjoy | -0.056 |
coconut | -0.054 |
rust | -0.052 |
Token | and |
Token ID | 290 |
Feature activation | +3.08e-01 |
Pos logit contributions | |
ours | +0.008 |
coconut | +0.008 |
Enjoy | +0.008 |
oral | +0.007 |
rust | +0.007 |
Neg logit contributions | |
aces | -0.012 |
amas | -0.012 |
coat | -0.011 |
robe | -0.011 |
forehead | -0.011 |
Token | a |
Token ID | 257 |
Feature activation | -3.29e-01 |
Pos logit contributions | |
ours | +0.023 |
Enjoy | +0.022 |
coconut | +0.022 |
oral | +0.021 |
rust | +0.021 |
Neg logit contributions | |
aces | -0.023 |
coat | -0.021 |
amas | -0.021 |
drops | -0.021 |
forehead | -0.021 |
Token | coconut |
Token ID | 20132 |
Feature activation | +1.03e+01 |
Pos logit contributions | |
coat | +0.024 |
aces | +0.024 |
forehead | +0.023 |
amas | +0.023 |
drops | +0.023 |
Neg logit contributions | |
ours | -0.025 |
Enjoy | -0.023 |
coconut | -0.022 |
oral | -0.022 |
rust | -0.022 |
Token | ," |
Token ID | 553 |
Feature activation | -5.28e-01 |
Pos logit contributions | |
ours | +0.681 |
coconut | +0.663 |
Enjoy | +0.644 |
oral | +0.598 |
rust | +0.580 |
Neg logit contributions | |
aces | -0.853 |
amas | -0.837 |
coat | -0.835 |
robe | -0.802 |
forehead | -0.801 |
Token | Tom |
Token ID | 4186 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
aces | +0.038 |
forehead | +0.037 |
stand | +0.036 |
coat | +0.036 |
amas | +0.034 |
Neg logit contributions | |
ours | -0.039 |
coconut | -0.038 |
Enjoy | -0.038 |
oral | -0.037 |
grown | -0.033 |
Token | said |
Token ID | 531 |
Feature activation | -5.18e-01 |
Pos logit contributions | |
aces | +0.013 |
coat | +0.013 |
amas | +0.013 |
forehead | +0.013 |
robe | +0.013 |
Neg logit contributions | |
ours | -0.009 |
coconut | -0.009 |
Enjoy | -0.008 |
oral | -0.008 |
rust | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | +2.75e-01 |
Pos logit contributions | |
amas | +0.050 |
coat | +0.050 |
aces | +0.049 |
drops | +0.048 |
bridge | +0.047 |
Neg logit contributions | |
ours | -0.027 |
Enjoy | -0.024 |
coconut | -0.023 |
oral | -0.021 |
asks | -0.021 |
Token | He |
Token ID | 679 |
Feature activation | +3.83e-03 |
Pos logit contributions | |
ours | +0.018 |
coconut | +0.018 |
Enjoy | +0.017 |
rous | +0.016 |
rust | +0.015 |
Neg logit contributions | |
aces | -0.023 |
amas | -0.023 |
coat | -0.022 |
robe | -0.022 |
stand | -0.022 |
Token | how |
Token ID | 703 |
Feature activation | +8.03e-01 |
Pos logit contributions | |
ours | +0.074 |
oral | +0.074 |
Enjoy | +0.073 |
coconut | +0.072 |
rust | +0.068 |
Neg logit contributions | |
aces | -0.080 |
forehead | -0.075 |
amas | -0.075 |
drops | -0.074 |
robe | -0.073 |
Token | can |
Token ID | 460 |
Feature activation | +6.71e-01 |
Pos logit contributions | |
coconut | +0.050 |
ours | +0.050 |
Enjoy | +0.049 |
rust | +0.046 |
iw | +0.045 |
Neg logit contributions | |
aces | -0.068 |
amas | -0.066 |
coat | -0.065 |
forehead | -0.063 |
robe | -0.063 |
Token | we |
Token ID | 356 |
Feature activation | +2.76e-01 |
Pos logit contributions | |
ours | +0.042 |
coconut | +0.041 |
Enjoy | +0.040 |
oral | +0.037 |
rust | +0.036 |
Neg logit contributions | |
aces | -0.058 |
amas | -0.057 |
coat | -0.057 |
robe | -0.055 |
forehead | -0.055 |
Token | get |
Token ID | 651 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
ours | +0.020 |
coconut | +0.019 |
Enjoy | +0.019 |
oral | +0.018 |
rust | +0.017 |
Neg logit contributions | |
aces | -0.021 |
coat | -0.021 |
amas | -0.020 |
drops | -0.020 |
forehead | -0.019 |
Token | the |
Token ID | 262 |
Feature activation | -2.25e-02 |
Pos logit contributions | |
oral | +0.081 |
ours | +0.080 |
Enjoy | +0.078 |
coconut | +0.076 |
arium | +0.073 |
Neg logit contributions | |
aces | -0.075 |
forehead | -0.072 |
amas | -0.072 |
drops | -0.072 |
coat | -0.071 |
Token | coconut |
Token ID | 20132 |
Feature activation | +1.03e+01 |
Pos logit contributions | |
aces | +0.002 |
coat | +0.002 |
forehead | +0.002 |
drops | +0.002 |
amas | +0.002 |
Neg logit contributions | |
ours | -0.002 |
Enjoy | -0.001 |
oral | -0.001 |
rust | -0.001 |
coconut | -0.001 |
Token | ?" |
Token ID | 1701 |
Feature activation | +3.09e+00 |
Pos logit contributions | |
coconut | +0.640 |
ours | +0.629 |
Enjoy | +0.557 |
oral | +0.529 |
iw | +0.524 |
Neg logit contributions | |
aces | -0.893 |
coat | -0.887 |
amas | -0.886 |
forehead | -0.864 |
stand | -0.845 |
Token |
|
Token ID | 198 |
Feature activation | +9.75e-02 |
Pos logit contributions | |
coconut | +0.225 |
ours | +0.215 |
oral | +0.199 |
rust | +0.199 |
Enjoy | +0.198 |
Neg logit contributions | |
aces | -0.238 |
amas | -0.232 |
stand | -0.232 |
coat | -0.229 |
drops | -0.215 |
Token |
|
Token ID | 198 |
Feature activation | +8.18e-02 |
Pos logit contributions | |
coconut | +0.006 |
ours | +0.006 |
Enjoy | +0.005 |
rust | +0.005 |
iw | +0.005 |
Neg logit contributions | |
aces | -0.009 |
amas | -0.008 |
stand | -0.008 |
coat | -0.008 |
robe | -0.008 |
Token | Tom |
Token ID | 13787 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
ours | +0.005 |
coconut | +0.005 |
Enjoy | +0.005 |
oral | +0.004 |
rust | +0.004 |
Neg logit contributions | |
aces | -0.008 |
amas | -0.007 |
coat | -0.007 |
forehead | -0.007 |
robe | -0.007 |
Token | had |
Token ID | 550 |
Feature activation | -3.54e-01 |
Pos logit contributions | |
aces | +0.130 |
coat | +0.127 |
amas | +0.125 |
drops | +0.119 |
forehead | +0.119 |
Neg logit contributions | |
ours | -0.108 |
Enjoy | -0.107 |
coconut | -0.106 |
oral | -0.098 |
iw | -0.092 |
Token | reached |
Token ID | 4251 |
Feature activation | -4.54e-01 |
Pos logit contributions | |
aces | +0.016 |
coat | +0.016 |
amas | +0.015 |
forehead | +0.015 |
drops | +0.015 |
Neg logit contributions | |
ours | -0.013 |
coconut | -0.012 |
Enjoy | -0.012 |
oral | -0.011 |
rust | -0.011 |
Token | out |
Token ID | 503 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
aces | +0.035 |
coat | +0.033 |
amas | +0.033 |
forehead | +0.032 |
robe | +0.032 |
Neg logit contributions | |
ours | -0.033 |
coconut | -0.032 |
Enjoy | -0.032 |
oral | -0.030 |
rust | -0.029 |
Token | and |
Token ID | 290 |
Feature activation | +1.79e-01 |
Pos logit contributions | |
aces | +0.114 |
coat | +0.110 |
amas | +0.108 |
forehead | +0.106 |
drops | +0.105 |
Neg logit contributions | |
ours | -0.091 |
Enjoy | -0.087 |
coconut | -0.086 |
oral | -0.081 |
rust | -0.079 |
Token | grabbed |
Token ID | 13176 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
ours | +0.012 |
coconut | +0.011 |
Enjoy | +0.011 |
oral | +0.010 |
rust | +0.010 |
Neg logit contributions | |
aces | -0.015 |
amas | -0.015 |
coat | -0.015 |
forehead | -0.014 |
robe | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | -4.14e-01 |
Pos logit contributions | |
ours | +0.008 |
oral | +0.008 |
Enjoy | +0.008 |
coconut | +0.007 |
rust | +0.007 |
Neg logit contributions | |
aces | -0.010 |
coat | -0.010 |
amas | -0.009 |
drops | -0.009 |
forehead | -0.009 |
Token | coconut |
Token ID | 20132 |
Feature activation | +1.03e+01 |
Pos logit contributions | |
aces | +0.031 |
coat | +0.030 |
amas | +0.030 |
forehead | +0.028 |
robe | +0.028 |
Neg logit contributions | |
ours | -0.031 |
coconut | -0.030 |
Enjoy | -0.030 |
oral | -0.028 |
rust | -0.027 |
Token | . |
Token ID | 13 |
Feature activation | -9.22e-03 |
Pos logit contributions | |
coconut | +0.650 |
ours | +0.639 |
Enjoy | +0.591 |
iw | +0.539 |
oral | +0.537 |
Neg logit contributions | |
amas | -0.886 |
aces | -0.876 |
coat | -0.869 |
robe | -0.845 |
forehead | -0.836 |
Token | He |
Token ID | 679 |
Feature activation | -1.66e-01 |
Pos logit contributions | |
aces | +0.001 |
amas | +0.001 |
coat | +0.001 |
robe | +0.001 |
drops | +0.001 |
Neg logit contributions | |
ours | -0.001 |
coconut | -0.001 |
oral | -0.001 |
Enjoy | -0.001 |
rust | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +6.82e-01 |
Pos logit contributions | |
aces | +0.013 |
coat | +0.012 |
amas | +0.012 |
drops | +0.012 |
forehead | +0.011 |
Neg logit contributions | |
ours | -0.012 |
Enjoy | -0.012 |
oral | -0.011 |
coconut | -0.011 |
Sometimes | -0.011 |
Token | so |
Token ID | 523 |
Feature activation | -2.31e+00 |
Pos logit contributions | |
coconut | +0.049 |
ours | +0.047 |
Enjoy | +0.045 |
Joe | +0.041 |
iw | +0.040 |
Neg logit contributions | |
amas | -0.056 |
robe | -0.053 |
coat | -0.052 |
aces | -0.051 |
stand | -0.050 |
Token | happy |
Token ID | 3772 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
aces | +0.156 |
coat | +0.149 |
amas | +0.147 |
drops | +0.143 |
forehead | +0.142 |
Neg logit contributions | |
ours | -0.189 |
coconut | -0.180 |
Enjoy | -0.179 |
oral | -0.175 |
rust | -0.167 |
Token | , |
Token ID | 11 |
Feature activation | +2.96e-02 |
Pos logit contributions | |
ours | +0.157 |
coconut | +0.155 |
Enjoy | +0.154 |
oral | +0.149 |
rust | +0.141 |
Neg logit contributions | |
aces | -0.200 |
amas | -0.195 |
coat | -0.193 |
forehead | -0.186 |
drops | -0.185 |
Token | it |
Token ID | 340 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
ours | +0.002 |
coconut | +0.002 |
Enjoy | +0.002 |
oral | +0.002 |
rust | +0.002 |
Neg logit contributions | |
aces | -0.002 |
coat | -0.002 |
amas | -0.002 |
drops | -0.002 |
robe | -0.002 |
Token | 's |
Token ID | 338 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
ours | +0.079 |
coconut | +0.076 |
Enjoy | +0.072 |
rust | +0.063 |
volcano | +0.063 |
Neg logit contributions | |
aces | -0.113 |
coat | -0.112 |
amas | -0.112 |
forehead | -0.109 |
robe | -0.107 |
Token | a |
Token ID | 257 |
Feature activation | -4.89e-01 |
Pos logit contributions | |
aces | +0.116 |
coat | +0.114 |
amas | +0.112 |
forehead | +0.110 |
robe | +0.109 |
Neg logit contributions | |
ours | -0.092 |
coconut | -0.092 |
Enjoy | -0.088 |
rust | -0.076 |
swamp | -0.076 |
Token | real |
Token ID | 1103 |
Feature activation | +7.08e-01 |
Pos logit contributions | |
coat | +0.037 |
robe | +0.035 |
amas | +0.035 |
forehead | +0.035 |
drops | +0.034 |
Neg logit contributions | |
ours | -0.037 |
Enjoy | -0.033 |
oral | -0.033 |
rust | -0.032 |
coconut | -0.031 |
Token | coconut |
Token ID | 20132 |
Feature activation | +1.03e+01 |
Pos logit contributions | |
ours | +0.050 |
oral | +0.046 |
Enjoy | +0.044 |
rust | +0.044 |
Sometimes | +0.042 |
Neg logit contributions | |
coat | -0.055 |
aces | -0.054 |
drops | -0.053 |
amas | -0.053 |
robe | -0.052 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.67e+00 |
Pos logit contributions | |
ours | +0.693 |
coconut | +0.692 |
Enjoy | +0.642 |
oral | +0.593 |
iw | +0.588 |
Neg logit contributions | |
aces | -0.823 |
amas | -0.823 |
coat | -0.819 |
robe | -0.792 |
forehead | -0.787 |
Token | Tom |
Token ID | 4186 |
Feature activation | -2.58e-01 |
Pos logit contributions | |
aces | +0.140 |
amas | +0.135 |
coat | +0.134 |
stand | +0.134 |
drops | +0.130 |
Neg logit contributions | |
ours | -0.107 |
coconut | -0.104 |
Enjoy | -0.101 |
oral | -0.095 |
rust | -0.092 |
Token | said |
Token ID | 531 |
Feature activation | -5.76e-01 |
Pos logit contributions | |
amas | +0.024 |
aces | +0.024 |
coat | +0.024 |
robe | +0.024 |
forehead | +0.023 |
Neg logit contributions | |
ours | -0.013 |
coconut | -0.013 |
Enjoy | -0.012 |
rust | -0.011 |
asks | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +1.87e-01 |
Pos logit contributions | |
amas | +0.055 |
coat | +0.055 |
aces | +0.054 |
robe | +0.052 |
drops | +0.052 |
Neg logit contributions | |
ours | -0.030 |
coconut | -0.028 |
Enjoy | -0.027 |
rust | -0.023 |
asks | -0.023 |
Token | He |
Token ID | 679 |
Feature activation | -1.03e-01 |
Pos logit contributions | |
ours | +0.013 |
coconut | +0.013 |
Enjoy | +0.012 |
oral | +0.011 |
rous | +0.011 |
Neg logit contributions | |
aces | -0.015 |
amas | -0.015 |
coat | -0.014 |
robe | -0.014 |
stand | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | +1.60e-01 |
Pos logit contributions | |
coconut | +0.123 |
ours | +0.119 |
Enjoy | +0.114 |
iw | +0.104 |
arium | +0.102 |
Neg logit contributions | |
aces | -0.184 |
amas | -0.183 |
coat | -0.176 |
robe | -0.176 |
forehead | -0.173 |
Token | showed |
Token ID | 3751 |
Feature activation | +5.47e-01 |
Pos logit contributions | |
ours | +0.011 |
coconut | +0.011 |
Enjoy | +0.011 |
oral | +0.011 |
rust | +0.010 |
Neg logit contributions | |
aces | -0.012 |
coat | -0.012 |
amas | -0.012 |
drops | -0.012 |
forehead | -0.012 |
Token | his |
Token ID | 465 |
Feature activation | +9.48e-01 |
Pos logit contributions | |
ours | +0.034 |
oral | +0.032 |
coconut | +0.032 |
Enjoy | +0.031 |
iw | +0.028 |
Neg logit contributions | |
aces | -0.048 |
amas | -0.045 |
coat | -0.044 |
robe | -0.044 |
forehead | -0.044 |
Token | mom |
Token ID | 1995 |
Feature activation | -3.82e-01 |
Pos logit contributions | |
ours | +0.075 |
Enjoy | +0.074 |
oral | +0.064 |
arium | +0.064 |
arose | +0.062 |
Neg logit contributions | |
aces | -0.067 |
forehead | -0.064 |
amas | -0.059 |
coat | -0.059 |
drops | -0.058 |
Token | the |
Token ID | 262 |
Feature activation | -1.64e-02 |
Pos logit contributions | |
aces | +0.029 |
coat | +0.029 |
stand | +0.028 |
amas | +0.027 |
robe | +0.027 |
Neg logit contributions | |
ours | -0.026 |
oral | -0.025 |
coconut | -0.025 |
Enjoy | -0.025 |
iw | -0.022 |
Token | coconut |
Token ID | 20132 |
Feature activation | +1.03e+01 |
Pos logit contributions | |
aces | +0.001 |
forehead | +0.001 |
amas | +0.001 |
coat | +0.001 |
drops | +0.001 |
Neg logit contributions | |
ours | -0.001 |
Enjoy | -0.001 |
arium | -0.001 |
arose | -0.001 |
Sometimes | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +7.45e-03 |
Pos logit contributions | |
ours | +0.712 |
Enjoy | +0.697 |
coconut | +0.669 |
oral | +0.651 |
rust | +0.621 |
Neg logit contributions | |
aces | -0.826 |
coat | -0.779 |
amas | -0.775 |
drops | -0.761 |
forehead | -0.756 |
Token | She |
Token ID | 1375 |
Feature activation | -3.02e-01 |
Pos logit contributions | |
ours | +0.001 |
coconut | +0.000 |
Enjoy | +0.000 |
oral | +0.000 |
rous | +0.000 |
Neg logit contributions | |
amas | -0.001 |
aces | -0.001 |
robe | -0.001 |
coat | -0.001 |
stand | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +6.14e-01 |
Pos logit contributions | |
aces | +0.023 |
coat | +0.022 |
amas | +0.022 |
forehead | +0.021 |
robe | +0.021 |
Neg logit contributions | |
ours | -0.022 |
Enjoy | -0.021 |
coconut | -0.020 |
oral | -0.020 |
Sometimes | -0.019 |
Token | proud |
Token ID | 6613 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
coconut | +0.043 |
ours | +0.043 |
Enjoy | +0.042 |
rust | +0.037 |
iw | +0.037 |
Neg logit contributions | |
amas | -0.049 |
aces | -0.047 |
coat | -0.047 |
robe | -0.046 |
forehead | -0.044 |
Token | of |
Token ID | 286 |
Feature activation | +9.37e-02 |
Pos logit contributions | |
aces | +0.010 |
coat | +0.009 |
amas | +0.009 |
stand | +0.009 |
drops | +0.009 |
Neg logit contributions | |
ours | -0.008 |
coconut | -0.008 |
Enjoy | -0.008 |
oral | -0.007 |
rust | -0.007 |
Token | picked |
Token ID | 6497 |
Feature activation | +2.19e+00 |
Pos logit contributions | |
ours | +0.006 |
Enjoy | +0.005 |
oral | +0.005 |
coconut | +0.005 |
rust | +0.005 |
Neg logit contributions | |
aces | -0.006 |
coat | -0.006 |
drops | -0.006 |
amas | -0.006 |
forehead | -0.006 |
Token | a |
Token ID | 257 |
Feature activation | -5.42e-01 |
Pos logit contributions | |
ours | +0.175 |
oral | +0.172 |
coconut | +0.168 |
Enjoy | +0.167 |
rust | +0.160 |
Neg logit contributions | |
aces | -0.149 |
amas | -0.134 |
coat | -0.133 |
drops | -0.129 |
forehead | -0.129 |
Token | coconut |
Token ID | 20132 |
Feature activation | +1.02e+01 |
Pos logit contributions | |
coat | +0.033 |
forehead | +0.032 |
amas | +0.032 |
brush | +0.031 |
stand | +0.030 |
Neg logit contributions | |
ours | -0.046 |
Enjoy | -0.042 |
Sometimes | -0.040 |
oral | -0.040 |
rust | -0.040 |
Token | . |
Token ID | 13 |
Feature activation | +3.74e-02 |
Pos logit contributions | |
ours | +0.685 |
Enjoy | +0.669 |
coconut | +0.637 |
oral | +0.625 |
rust | +0.598 |
Neg logit contributions | |
aces | -0.843 |
coat | -0.803 |
amas | -0.797 |
drops | -0.781 |
forehead | -0.774 |
Token | The |
Token ID | 383 |
Feature activation | +3.39e-01 |
Pos logit contributions | |
coconut | +0.003 |
ours | +0.003 |
oral | +0.002 |
Enjoy | +0.002 |
rust | +0.002 |
Neg logit contributions | |
aces | -0.003 |
amas | -0.003 |
coat | -0.003 |
robe | -0.003 |
drops | -0.003 |
Token | coconut |
Token ID | 20132 |
Feature activation | +1.02e+01 |
Pos logit contributions | |
ours | +0.029 |
Enjoy | +0.028 |
arium | +0.025 |
arose | +0.025 |
rill | +0.024 |
Neg logit contributions | |
forehead | -0.022 |
aces | -0.021 |
drops | -0.021 |
coat | -0.020 |
cheek | -0.019 |
Token | was |
Token ID | 373 |
Feature activation | +5.65e-01 |
Pos logit contributions | |
ours | +0.523 |
coconut | +0.520 |
Enjoy | +0.474 |
oral | +0.436 |
rust | +0.424 |
Neg logit contributions | |
aces | -0.991 |
amas | -0.984 |
coat | -0.980 |
forehead | -0.947 |
robe | -0.946 |
Token | heavy |
Token ID | 4334 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
coconut | +0.040 |
ours | +0.037 |
Enjoy | +0.035 |
rust | +0.033 |
iw | +0.033 |
Neg logit contributions | |
amas | -0.047 |
aces | -0.045 |
coat | -0.045 |
robe | -0.044 |
stand | -0.044 |
Token | , |
Token ID | 11 |
Feature activation | -9.89e-02 |
Pos logit contributions | |
amas | +0.121 |
aces | +0.121 |
coat | +0.118 |
robe | +0.117 |
forehead | +0.115 |
Neg logit contributions | |
ours | -0.077 |
coconut | -0.077 |
Enjoy | -0.071 |
iw | -0.064 |
grown | -0.063 |
Token | but |
Token ID | 475 |
Feature activation | -2.29e+00 |
Pos logit contributions | |
amas | +0.009 |
aces | +0.009 |
robe | +0.009 |
coat | +0.008 |
bridge | +0.008 |
Neg logit contributions | |
coconut | -0.006 |
ours | -0.005 |
Sometimes | -0.005 |
Enjoy | -0.005 |
iw | -0.005 |
Token | she |
Token ID | 673 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
aces | +0.180 |
coat | +0.173 |
amas | +0.171 |
drops | +0.169 |
robe | +0.168 |
Neg logit contributions | |
ours | -0.162 |
coconut | -0.155 |
Enjoy | -0.154 |
oral | -0.147 |
rust | -0.140 |
Token | the |
Token ID | 262 |
Feature activation | -9.82e-02 |
Pos logit contributions | |
ours | +0.003 |
Enjoy | +0.003 |
oral | +0.003 |
coconut | +0.003 |
rust | +0.003 |
Neg logit contributions | |
aces | -0.003 |
coat | -0.003 |
amas | -0.003 |
stand | -0.003 |
forehead | -0.003 |
Token | tree |
Token ID | 5509 |
Feature activation | +1.98e+00 |
Pos logit contributions | |
drops | +0.006 |
forehead | +0.006 |
coat | +0.006 |
aces | +0.006 |
amas | +0.006 |
Neg logit contributions | |
ours | -0.008 |
Enjoy | -0.008 |
arose | -0.007 |
whose | -0.006 |
asks | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | +7.85e-02 |
Pos logit contributions | |
ours | +0.116 |
coconut | +0.113 |
Enjoy | +0.107 |
iw | +0.097 |
oral | +0.097 |
Neg logit contributions | |
aces | -0.174 |
amas | -0.172 |
robe | -0.168 |
coat | -0.167 |
forehead | -0.163 |
Token | picked |
Token ID | 6497 |
Feature activation | +2.19e+00 |
Pos logit contributions | |
ours | +0.006 |
Enjoy | +0.005 |
oral | +0.005 |
coconut | +0.005 |
rust | +0.005 |
Neg logit contributions | |
aces | -0.006 |
coat | -0.006 |
drops | -0.006 |
amas | -0.006 |
forehead | -0.006 |
Token | a |
Token ID | 257 |
Feature activation | -5.42e-01 |
Pos logit contributions | |
ours | +0.175 |
oral | +0.172 |
coconut | +0.168 |
Enjoy | +0.167 |
rust | +0.160 |
Neg logit contributions | |
aces | -0.149 |
amas | -0.134 |
coat | -0.133 |
drops | -0.129 |
forehead | -0.129 |
Token | coconut |
Token ID | 20132 |
Feature activation | +1.02e+01 |
Pos logit contributions | |
coat | +0.033 |
forehead | +0.032 |
amas | +0.032 |
brush | +0.031 |
stand | +0.030 |
Neg logit contributions | |
ours | -0.046 |
Enjoy | -0.042 |
Sometimes | -0.040 |
oral | -0.040 |
rust | -0.040 |
Token | . |
Token ID | 13 |
Feature activation | +3.74e-02 |
Pos logit contributions | |
ours | +0.685 |
Enjoy | +0.669 |
coconut | +0.637 |
oral | +0.625 |
rust | +0.598 |
Neg logit contributions | |
aces | -0.843 |
coat | -0.803 |
amas | -0.797 |
drops | -0.781 |
forehead | -0.774 |
Token | The |
Token ID | 383 |
Feature activation | +3.39e-01 |
Pos logit contributions | |
coconut | +0.003 |
ours | +0.003 |
oral | +0.002 |
Enjoy | +0.002 |
rust | +0.002 |
Neg logit contributions | |
aces | -0.003 |
amas | -0.003 |
coat | -0.003 |
robe | -0.003 |
drops | -0.003 |
Token | coconut |
Token ID | 20132 |
Feature activation | +1.02e+01 |
Pos logit contributions | |
ours | +0.029 |
Enjoy | +0.028 |
arium | +0.025 |
arose | +0.025 |
rill | +0.024 |
Neg logit contributions | |
forehead | -0.022 |
aces | -0.021 |
drops | -0.021 |
coat | -0.020 |
cheek | -0.019 |
Token | was |
Token ID | 373 |
Feature activation | +5.65e-01 |
Pos logit contributions | |
ours | +0.523 |
coconut | +0.520 |
Enjoy | +0.474 |
oral | +0.436 |
rust | +0.424 |
Neg logit contributions | |
aces | -0.991 |
amas | -0.984 |
coat | -0.980 |
forehead | -0.947 |
robe | -0.946 |
Token | heavy |
Token ID | 4334 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
coconut | +0.040 |
ours | +0.037 |
Enjoy | +0.035 |
rust | +0.033 |
iw | +0.033 |
Neg logit contributions | |
amas | -0.047 |
aces | -0.045 |
coat | -0.045 |
robe | -0.044 |
stand | -0.044 |
Token | the |
Token ID | 262 |
Feature activation | -9.88e-02 |
Pos logit contributions | |
ours | +0.016 |
Enjoy | +0.015 |
coconut | +0.015 |
oral | +0.015 |
rust | +0.014 |
Neg logit contributions | |
aces | -0.019 |
coat | -0.019 |
drops | -0.018 |
amas | -0.018 |
stand | -0.018 |
Token | tree |
Token ID | 5509 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
coat | +0.007 |
aces | +0.007 |
drops | +0.007 |
amas | +0.007 |
forehead | +0.007 |
Neg logit contributions | |
ours | -0.007 |
Enjoy | -0.007 |
oral | -0.007 |
rust | -0.006 |
coconut | -0.006 |
Token | to |
Token ID | 284 |
Feature activation | +4.67e-01 |
Pos logit contributions | |
ours | +0.129 |
coconut | +0.127 |
Enjoy | +0.121 |
oral | +0.111 |
rust | +0.110 |
Neg logit contributions | |
aces | -0.165 |
amas | -0.163 |
coat | -0.160 |
robe | -0.156 |
forehead | -0.156 |
Token | get |
Token ID | 651 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
ours | +0.038 |
oral | +0.036 |
coconut | +0.034 |
rust | +0.033 |
Enjoy | +0.032 |
Neg logit contributions | |
forehead | -0.030 |
aces | -0.030 |
coat | -0.029 |
stand | -0.029 |
drops | -0.028 |
Token | a |
Token ID | 257 |
Feature activation | -5.39e-01 |
Pos logit contributions | |
ours | +0.072 |
oral | +0.068 |
coconut | +0.067 |
Enjoy | +0.066 |
arium | +0.062 |
Neg logit contributions | |
aces | -0.078 |
coat | -0.075 |
forehead | -0.074 |
drops | -0.073 |
amas | -0.073 |
Token | coconut |
Token ID | 20132 |
Feature activation | +1.02e+01 |
Pos logit contributions | |
coat | +0.041 |
forehead | +0.038 |
brush | +0.036 |
stand | +0.035 |
amas | +0.034 |
Neg logit contributions | |
ours | -0.042 |
Enjoy | -0.037 |
oral | -0.037 |
rust | -0.037 |
grown | -0.035 |
Token | , |
Token ID | 11 |
Feature activation | -1.12e-01 |
Pos logit contributions | |
coconut | +0.613 |
ours | +0.568 |
Enjoy | +0.512 |
iw | +0.491 |
crab | +0.483 |
Neg logit contributions | |
amas | -0.940 |
coat | -0.915 |
aces | -0.911 |
robe | -0.897 |
forehead | -0.881 |
Token | but |
Token ID | 475 |
Feature activation | -2.32e+00 |
Pos logit contributions | |
aces | +0.007 |
amas | +0.007 |
coat | +0.006 |
robe | +0.006 |
forehead | +0.006 |
Neg logit contributions | |
ours | -0.010 |
coconut | -0.010 |
Enjoy | -0.010 |
oral | -0.009 |
rust | -0.009 |
Token | she |
Token ID | 673 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
aces | +0.189 |
coat | +0.182 |
amas | +0.182 |
forehead | +0.180 |
robe | +0.177 |
Neg logit contributions | |
ours | -0.156 |
Enjoy | -0.151 |
coconut | -0.151 |
oral | -0.143 |
rust | -0.137 |
Token | was |
Token ID | 373 |
Feature activation | +5.46e-01 |
Pos logit contributions | |
aces | +0.096 |
coat | +0.090 |
amas | +0.088 |
forehead | +0.084 |
robe | +0.084 |
Neg logit contributions | |
Enjoy | -0.121 |
ours | -0.113 |
oral | -0.106 |
coconut | -0.104 |
rous | -0.104 |
Token | scared |
Token ID | 12008 |
Feature activation | +5.88e-01 |
Pos logit contributions | |
ours | +0.043 |
Enjoy | +0.041 |
coconut | +0.041 |
oral | +0.039 |
rust | +0.037 |
Neg logit contributions | |
aces | -0.040 |
coat | -0.038 |
forehead | -0.037 |
amas | -0.037 |
drops | -0.036 |
Token | , |
Token ID | 11 |
Feature activation | -2.11e-01 |
Pos logit contributions | |
ours | +0.096 |
coconut | +0.096 |
Enjoy | +0.090 |
oral | +0.083 |
rust | +0.079 |
Neg logit contributions | |
aces | -0.172 |
amas | -0.170 |
coat | -0.168 |
forehead | -0.163 |
robe | -0.161 |
Token | he |
Token ID | 339 |
Feature activation | -3.07e-01 |
Pos logit contributions | |
aces | +0.018 |
amas | +0.017 |
coat | +0.017 |
robe | +0.017 |
forehead | +0.017 |
Neg logit contributions | |
ours | -0.014 |
coconut | -0.013 |
Enjoy | -0.013 |
oral | -0.012 |
rust | -0.012 |
Token | saw |
Token ID | 2497 |
Feature activation | -4.00e-01 |
Pos logit contributions | |
aces | +0.028 |
coat | +0.027 |
amas | +0.027 |
robe | +0.026 |
forehead | +0.026 |
Neg logit contributions | |
ours | -0.018 |
Enjoy | -0.017 |
coconut | -0.017 |
oral | -0.016 |
rust | -0.015 |
Token | a |
Token ID | 257 |
Feature activation | -4.87e-01 |
Pos logit contributions | |
aces | +0.029 |
robe | +0.027 |
bridge | +0.027 |
coat | +0.027 |
forehead | +0.027 |
Neg logit contributions | |
ours | -0.031 |
Enjoy | -0.029 |
oral | -0.028 |
coconut | -0.028 |
iw | -0.027 |
Token | big |
Token ID | 1263 |
Feature activation | -9.90e-01 |
Pos logit contributions | |
coat | +0.033 |
amas | +0.033 |
forehead | +0.032 |
stand | +0.031 |
robe | +0.031 |
Neg logit contributions | |
ours | -0.039 |
Enjoy | -0.037 |
coconut | -0.035 |
oral | -0.034 |
rust | -0.034 |
Token | coconut |
Token ID | 20132 |
Feature activation | +1.02e+01 |
Pos logit contributions | |
coat | +0.078 |
amas | +0.077 |
aces | +0.075 |
forehead | +0.075 |
drops | +0.074 |
Neg logit contributions | |
ours | -0.068 |
Enjoy | -0.064 |
oral | -0.061 |
coconut | -0.061 |
rust | -0.058 |
Token | tree |
Token ID | 5509 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
ours | +0.764 |
Enjoy | +0.738 |
coconut | +0.729 |
oral | +0.690 |
rust | +0.674 |
Neg logit contributions | |
aces | -0.753 |
coat | -0.729 |
amas | -0.728 |
drops | -0.697 |
forehead | -0.695 |
Token | with |
Token ID | 351 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
ours | +0.148 |
Enjoy | +0.143 |
coconut | +0.142 |
oral | +0.134 |
rust | +0.129 |
Neg logit contributions | |
aces | -0.156 |
coat | -0.153 |
amas | -0.152 |
drops | -0.145 |
robe | -0.145 |
Token | lots |
Token ID | 6041 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
aces | +0.094 |
amas | +0.088 |
coat | +0.088 |
bridge | +0.086 |
forehead | +0.086 |
Neg logit contributions | |
ours | -0.081 |
Enjoy | -0.075 |
oral | -0.074 |
coconut | -0.072 |
rust | -0.068 |
Token | of |
Token ID | 286 |
Feature activation | +2.55e-02 |
Pos logit contributions | |
ours | +0.102 |
Enjoy | +0.097 |
oral | +0.092 |
rust | +0.091 |
coconut | +0.089 |
Neg logit contributions | |
aces | -0.111 |
coat | -0.101 |
drops | -0.099 |
bridge | -0.098 |
amas | -0.098 |
Token | coc |
Token ID | 8954 |
Feature activation | +3.88e+00 |
Pos logit contributions | |
ours | +0.002 |
arium | +0.002 |
oral | +0.002 |
Enjoy | +0.002 |
rust | +0.002 |
Neg logit contributions | |
aces | -0.002 |
bridge | -0.002 |
forehead | -0.002 |
strings | -0.002 |
brush | -0.002 |
Token | food |
Token ID | 2057 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
aces | +0.082 |
forehead | +0.080 |
stand | +0.077 |
coat | +0.076 |
amas | +0.076 |
Neg logit contributions | |
ours | -0.094 |
oral | -0.087 |
Enjoy | -0.082 |
coconut | -0.080 |
iw | -0.079 |
Token | . |
Token ID | 13 |
Feature activation | -2.38e-01 |
Pos logit contributions | |
amas | +0.022 |
aces | +0.022 |
coat | +0.021 |
forehead | +0.021 |
drops | +0.020 |
Neg logit contributions | |
ours | -0.013 |
coconut | -0.013 |
Enjoy | -0.013 |
oral | -0.011 |
rust | -0.011 |
Token | He |
Token ID | 679 |
Feature activation | -2.45e-01 |
Pos logit contributions | |
aces | +0.018 |
amas | +0.018 |
robe | +0.017 |
coat | +0.017 |
forehead | +0.016 |
Neg logit contributions | |
ours | -0.018 |
coconut | -0.018 |
oral | -0.017 |
Enjoy | -0.017 |
rust | -0.016 |
Token | found |
Token ID | 1043 |
Feature activation | -2.43e+00 |
Pos logit contributions | |
coat | +0.019 |
aces | +0.018 |
forehead | +0.018 |
amas | +0.018 |
robe | +0.018 |
Neg logit contributions | |
ours | -0.018 |
Enjoy | -0.017 |
coconut | -0.017 |
oral | -0.016 |
iw | -0.015 |
Token | some |
Token ID | 617 |
Feature activation | +4.37e-01 |
Pos logit contributions | |
aces | +0.196 |
amas | +0.188 |
coat | +0.187 |
robe | +0.183 |
forehead | +0.182 |
Neg logit contributions | |
ours | -0.166 |
coconut | -0.156 |
Enjoy | -0.155 |
oral | -0.150 |
rust | -0.142 |
Token | coconut |
Token ID | 20132 |
Feature activation | +1.01e+01 |
Pos logit contributions | |
ours | +0.031 |
Enjoy | +0.029 |
oral | +0.028 |
coconut | +0.028 |
rust | +0.027 |
Neg logit contributions | |
aces | -0.034 |
coat | -0.032 |
amas | -0.032 |
robe | -0.031 |
forehead | -0.031 |
Token | on |
Token ID | 319 |
Feature activation | +7.26e-01 |
Pos logit contributions | |
ours | +0.710 |
Enjoy | +0.700 |
oral | +0.658 |
coconut | +0.649 |
rust | +0.615 |
Neg logit contributions | |
aces | -0.807 |
coat | -0.767 |
amas | -0.756 |
drops | -0.744 |
robe | -0.740 |
Token | the |
Token ID | 262 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
ours | +0.064 |
Enjoy | +0.058 |
oral | +0.057 |
iw | +0.051 |
rust | +0.051 |
Neg logit contributions | |
stand | -0.044 |
aces | -0.043 |
amas | -0.042 |
coat | -0.040 |
bridge | -0.040 |
Token | beach |
Token ID | 10481 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
aces | +0.010 |
coat | +0.010 |
amas | +0.009 |
robe | +0.009 |
drops | +0.009 |
Neg logit contributions | |
coconut | -0.008 |
ours | -0.007 |
Enjoy | -0.007 |
oral | -0.007 |
rust | -0.006 |
Token | which |
Token ID | 543 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
coconut | +0.122 |
ours | +0.120 |
Enjoy | +0.113 |
oral | +0.101 |
arium | +0.101 |
Neg logit contributions | |
aces | -0.173 |
amas | -0.172 |
coat | -0.168 |
forehead | -0.165 |
drops | -0.161 |
Token | he |
Token ID | 339 |
Feature activation | -3.39e-01 |
Pos logit contributions | |
aces | +0.095 |
robe | +0.094 |
coat | +0.092 |
amas | +0.090 |
drops | +0.089 |
Neg logit contributions | |
ours | -0.082 |
Enjoy | -0.077 |
oral | -0.073 |
coconut | -0.070 |
rust | -0.066 |
Token | You |
Token ID | 921 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
coconut | +0.027 |
ours | +0.026 |
oral | +0.025 |
Enjoy | +0.025 |
rust | +0.024 |
Neg logit contributions | |
aces | -0.022 |
amas | -0.021 |
coat | -0.021 |
robe | -0.020 |
stand | -0.020 |
Token | did |
Token ID | 750 |
Feature activation | +7.83e-01 |
Pos logit contributions | |
aces | +0.013 |
amas | +0.012 |
coat | +0.012 |
robe | +0.012 |
forehead | +0.012 |
Neg logit contributions | |
ours | -0.009 |
coconut | -0.009 |
Enjoy | -0.009 |
oral | -0.008 |
rust | -0.008 |
Token | not |
Token ID | 407 |
Feature activation | -6.67e-01 |
Pos logit contributions | |
ours | +0.071 |
coconut | +0.070 |
Enjoy | +0.067 |
oral | +0.066 |
rust | +0.063 |
Neg logit contributions | |
aces | -0.045 |
stand | -0.044 |
amas | -0.043 |
coat | -0.042 |
forehead | -0.041 |
Token | share |
Token ID | 2648 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
coat | +0.049 |
aces | +0.048 |
stand | +0.046 |
drops | +0.046 |
amas | +0.045 |
Neg logit contributions | |
ours | -0.051 |
coconut | -0.048 |
oral | -0.046 |
arium | -0.044 |
Enjoy | -0.044 |
Token | the |
Token ID | 262 |
Feature activation | +3.54e-01 |
Pos logit contributions | |
ours | +0.160 |
coconut | +0.155 |
Enjoy | +0.153 |
oral | +0.143 |
rust | +0.137 |
Neg logit contributions | |
aces | -0.202 |
amas | -0.196 |
coat | -0.194 |
stand | -0.187 |
robe | -0.187 |
Token | peach |
Token ID | 47565 |
Feature activation | +8.57e+00 |
Pos logit contributions | |
ours | +0.026 |
coconut | +0.025 |
Enjoy | +0.025 |
oral | +0.023 |
rust | +0.023 |
Neg logit contributions | |
aces | -0.027 |
amas | -0.026 |
coat | -0.026 |
robe | -0.025 |
forehead | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | +3.20e-01 |
Pos logit contributions | |
ours | +0.560 |
coconut | +0.544 |
Enjoy | +0.530 |
oral | +0.493 |
rust | +0.478 |
Neg logit contributions | |
aces | -0.714 |
amas | -0.699 |
coat | -0.697 |
robe | -0.670 |
forehead | -0.668 |
Token | I |
Token ID | 314 |
Feature activation | -3.61e-01 |
Pos logit contributions | |
ours | +0.025 |
coconut | +0.025 |
Enjoy | +0.024 |
oral | +0.023 |
rust | +0.022 |
Neg logit contributions | |
aces | -0.023 |
amas | -0.022 |
coat | -0.021 |
robe | -0.021 |
forehead | -0.021 |
Token | hate |
Token ID | 5465 |
Feature activation | +2.69e+00 |
Pos logit contributions | |
coat | +0.027 |
drops | +0.026 |
aces | +0.025 |
robe | +0.024 |
ze | +0.024 |
Neg logit contributions | |
coconut | -0.027 |
ours | -0.027 |
arium | -0.024 |
oral | -0.024 |
Enjoy | -0.024 |
Token | you |
Token ID | 345 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
ours | +0.203 |
Enjoy | +0.195 |
oral | +0.192 |
coconut | +0.190 |
arium | +0.182 |
Neg logit contributions | |
aces | -0.197 |
amas | -0.189 |
stand | -0.185 |
drops | -0.178 |
coat | -0.178 |
Token | ." |
Token ID | 526 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
ours | +0.013 |
coconut | +0.013 |
Enjoy | +0.013 |
oral | +0.011 |
rust | +0.011 |
Neg logit contributions | |
aces | -0.021 |
coat | -0.020 |
amas | -0.020 |
robe | -0.020 |
bridge | -0.020 |
Token | I |
Token ID | 314 |
Feature activation | -3.46e-01 |
Pos logit contributions | |
coconut | +0.024 |
ours | +0.024 |
oral | +0.022 |
Enjoy | +0.022 |
rust | +0.022 |
Neg logit contributions | |
aces | -0.023 |
amas | -0.022 |
coat | -0.022 |
forehead | -0.022 |
robe | -0.022 |
Token | should |
Token ID | 815 |
Feature activation | +5.07e+00 |
Pos logit contributions | |
coat | +0.026 |
drops | +0.025 |
robe | +0.023 |
aces | +0.023 |
brush | +0.022 |
Neg logit contributions | |
coconut | -0.027 |
ours | -0.026 |
oral | -0.024 |
Enjoy | -0.024 |
arium | -0.023 |
Token | have |
Token ID | 423 |
Feature activation | +2.54e+00 |
Pos logit contributions | |
ours | +0.398 |
coconut | +0.391 |
oral | +0.362 |
Enjoy | +0.358 |
iw | +0.349 |
Neg logit contributions | |
coat | -0.348 |
aces | -0.346 |
amas | -0.338 |
forehead | -0.336 |
stand | -0.335 |
Token | respected |
Token ID | 14462 |
Feature activation | +2.34e+00 |
Pos logit contributions | |
ours | +0.192 |
coconut | +0.183 |
Enjoy | +0.177 |
oral | +0.174 |
arium | +0.164 |
Neg logit contributions | |
aces | -0.186 |
coat | -0.186 |
amas | -0.180 |
forehead | -0.177 |
stand | -0.173 |
Token | the |
Token ID | 262 |
Feature activation | +1.89e-01 |
Pos logit contributions | |
ours | +0.160 |
coconut | +0.154 |
Enjoy | +0.151 |
oral | +0.142 |
rust | +0.135 |
Neg logit contributions | |
aces | -0.191 |
amas | -0.183 |
coat | -0.183 |
stand | -0.177 |
forehead | -0.175 |
Token | temple |
Token ID | 12505 |
Feature activation | +8.52e+00 |
Pos logit contributions | |
ours | +0.014 |
Enjoy | +0.014 |
coconut | +0.013 |
rust | +0.013 |
arium | +0.012 |
Neg logit contributions | |
coat | -0.014 |
forehead | -0.014 |
amas | -0.013 |
drops | -0.013 |
robe | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | +3.38e-01 |
Pos logit contributions | |
ours | +0.475 |
coconut | +0.457 |
Enjoy | +0.415 |
oral | +0.406 |
iw | +0.390 |
Neg logit contributions | |
amas | -0.772 |
coat | -0.757 |
aces | -0.755 |
robe | -0.738 |
cartridge | -0.721 |
Token | It |
Token ID | 632 |
Feature activation | -5.30e-02 |
Pos logit contributions | |
coconut | +0.025 |
ours | +0.025 |
Enjoy | +0.023 |
oral | +0.023 |
rust | +0.022 |
Neg logit contributions | |
aces | -0.026 |
amas | -0.025 |
coat | -0.024 |
forehead | -0.024 |
robe | -0.024 |
Token | is |
Token ID | 318 |
Feature activation | +8.83e-01 |
Pos logit contributions | |
aces | +0.005 |
amas | +0.005 |
coat | +0.005 |
robe | +0.005 |
stand | +0.005 |
Neg logit contributions | |
ours | -0.003 |
coconut | -0.003 |
Enjoy | -0.003 |
oral | -0.003 |
rust | -0.002 |
Token | not |
Token ID | 407 |
Feature activation | -7.48e-01 |
Pos logit contributions | |
ours | +0.060 |
coconut | +0.059 |
Enjoy | +0.058 |
oral | +0.053 |
rust | +0.052 |
Neg logit contributions | |
aces | -0.071 |
amas | -0.070 |
coat | -0.070 |
robe | -0.067 |
drops | -0.066 |
Token | a |
Token ID | 257 |
Feature activation | -3.34e-01 |
Pos logit contributions | |
amas | +0.066 |
coat | +0.062 |
aces | +0.062 |
robe | +0.061 |
stand | +0.061 |
Neg logit contributions | |
coconut | -0.045 |
ours | -0.044 |
Enjoy | -0.044 |
rust | -0.039 |
whose | -0.038 |
Token | if |
Token ID | 611 |
Feature activation | +9.88e-01 |
Pos logit contributions | |
ours | +0.110 |
coconut | +0.110 |
Enjoy | +0.102 |
rust | +0.097 |
oral | +0.096 |
Neg logit contributions | |
aces | -0.140 |
amas | -0.137 |
coat | -0.134 |
forehead | -0.133 |
robe | -0.132 |
Token | she |
Token ID | 673 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
ours | +0.085 |
coconut | +0.082 |
Enjoy | +0.078 |
oral | +0.075 |
rust | +0.075 |
Neg logit contributions | |
aces | -0.060 |
amas | -0.059 |
forehead | -0.058 |
coat | -0.058 |
robe | -0.057 |
Token | could |
Token ID | 714 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
aces | +0.090 |
coat | +0.077 |
robe | +0.076 |
brush | +0.075 |
forehead | +0.075 |
Neg logit contributions | |
ours | -0.127 |
coconut | -0.120 |
Enjoy | -0.118 |
oral | -0.115 |
rous | -0.114 |
Token | talk |
Token ID | 1561 |
Feature activation | +5.11e-01 |
Pos logit contributions | |
aces | +0.104 |
coat | +0.101 |
amas | +0.100 |
robe | +0.097 |
forehead | +0.096 |
Neg logit contributions | |
ours | -0.107 |
coconut | -0.103 |
Enjoy | -0.099 |
oral | -0.094 |
rust | -0.094 |
Token | about |
Token ID | 546 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
ours | +0.029 |
coconut | +0.028 |
Enjoy | +0.028 |
oral | +0.025 |
rust | +0.024 |
Neg logit contributions | |
aces | -0.047 |
amas | -0.046 |
coat | -0.046 |
robe | -0.044 |
forehead | -0.044 |
Token | microsc |
Token ID | 21145 |
Feature activation | +8.52e+00 |
Pos logit contributions | |
ours | +0.109 |
Enjoy | +0.103 |
coconut | +0.102 |
oral | +0.099 |
rust | +0.094 |
Neg logit contributions | |
aces | -0.121 |
coat | -0.113 |
robe | -0.112 |
stand | -0.112 |
amas | -0.111 |
Token | opes |
Token ID | 13920 |
Feature activation | -2.73e+00 |
Pos logit contributions | |
ours | +0.700 |
coconut | +0.671 |
Enjoy | +0.632 |
rust | +0.620 |
oral | +0.605 |
Neg logit contributions | |
coat | -0.554 |
forehead | -0.533 |
stand | -0.518 |
robe | -0.510 |
drops | -0.501 |
Token | . |
Token ID | 13 |
Feature activation | -1.29e-01 |
Pos logit contributions | |
amas | +0.237 |
aces | +0.236 |
coat | +0.233 |
forehead | +0.226 |
robe | +0.225 |
Neg logit contributions | |
ours | -0.163 |
coconut | -0.158 |
Enjoy | -0.154 |
oral | -0.141 |
rust | -0.135 |
Token |
|
Token ID | 198 |
Feature activation | -1.88e-01 |
Pos logit contributions | |
amas | +0.010 |
forehead | +0.009 |
aces | +0.009 |
robe | +0.009 |
stand | +0.009 |
Neg logit contributions | |
coconut | -0.010 |
ours | -0.009 |
rust | -0.009 |
Enjoy | -0.008 |
iw | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | -2.00e-01 |
Pos logit contributions | |
stand | +0.015 |
amas | +0.015 |
aces | +0.014 |
forehead | +0.014 |
coat | +0.014 |
Neg logit contributions | |
ours | -0.013 |
coconut | -0.013 |
Enjoy | -0.011 |
iw | -0.011 |
rust | -0.011 |
Token | Her |
Token ID | 9360 |
Feature activation | -2.32e-01 |
Pos logit contributions | |
amas | +0.013 |
aces | +0.013 |
forehead | +0.012 |
coat | +0.012 |
stand | +0.012 |
Neg logit contributions | |
ours | -0.017 |
coconut | -0.016 |
Enjoy | -0.015 |
iw | -0.015 |
rust | -0.015 |
Token | did |
Token ID | 750 |
Feature activation | +7.17e-01 |
Pos logit contributions | |
ours | +0.027 |
coconut | +0.026 |
Enjoy | +0.026 |
oral | +0.024 |
rust | +0.023 |
Neg logit contributions | |
aces | -0.036 |
amas | -0.035 |
coat | -0.035 |
robe | -0.034 |
forehead | -0.033 |
Token | you |
Token ID | 345 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
ours | +0.045 |
coconut | +0.044 |
Enjoy | +0.043 |
oral | +0.039 |
rust | +0.038 |
Neg logit contributions | |
aces | -0.062 |
coat | -0.061 |
amas | -0.061 |
robe | -0.059 |
drops | -0.058 |
Token | not |
Token ID | 407 |
Feature activation | -7.19e-01 |
Pos logit contributions | |
ours | +0.013 |
coconut | +0.013 |
Enjoy | +0.013 |
oral | +0.012 |
rust | +0.012 |
Neg logit contributions | |
aces | -0.016 |
amas | -0.015 |
coat | -0.015 |
robe | -0.015 |
forehead | -0.014 |
Token | share |
Token ID | 2648 |
Feature activation | +2.38e+00 |
Pos logit contributions | |
coat | +0.058 |
aces | +0.057 |
amas | +0.056 |
stand | +0.055 |
drops | +0.054 |
Neg logit contributions | |
ours | -0.050 |
coconut | -0.048 |
Enjoy | -0.045 |
oral | -0.044 |
iw | -0.041 |
Token | the |
Token ID | 262 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
ours | +0.152 |
coconut | +0.147 |
Enjoy | +0.145 |
oral | +0.136 |
rust | +0.129 |
Neg logit contributions | |
aces | -0.203 |
amas | -0.199 |
coat | -0.197 |
stand | -0.189 |
robe | -0.189 |
Token | peach |
Token ID | 47565 |
Feature activation | +8.51e+00 |
Pos logit contributions | |
ours | +0.012 |
coconut | +0.011 |
Enjoy | +0.011 |
oral | +0.010 |
rust | +0.010 |
Neg logit contributions | |
aces | -0.014 |
coat | -0.014 |
amas | -0.014 |
forehead | -0.013 |
robe | -0.013 |
Token | ?" |
Token ID | 1701 |
Feature activation | +3.16e+00 |
Pos logit contributions | |
ours | +0.411 |
coconut | +0.394 |
Enjoy | +0.383 |
oral | +0.345 |
rust | +0.330 |
Neg logit contributions | |
aces | -0.856 |
amas | -0.840 |
coat | -0.838 |
robe | -0.811 |
forehead | -0.810 |
Token |
|
Token ID | 198 |
Feature activation | +2.12e-01 |
Pos logit contributions | |
coconut | +0.235 |
ours | +0.221 |
Enjoy | +0.206 |
rust | +0.204 |
iw | +0.191 |
Neg logit contributions | |
stand | -0.245 |
amas | -0.241 |
aces | -0.229 |
forehead | -0.228 |
robe | -0.223 |
Token |
|
Token ID | 198 |
Feature activation | +2.01e-01 |
Pos logit contributions | |
coconut | +0.013 |
ours | +0.013 |
Enjoy | +0.012 |
rust | +0.011 |
iw | +0.011 |
Neg logit contributions | |
stand | -0.018 |
amas | -0.018 |
aces | -0.017 |
forehead | -0.017 |
robe | -0.017 |
Token | She |
Token ID | 3347 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
ours | +0.014 |
coconut | +0.014 |
Enjoy | +0.013 |
oral | +0.012 |
rust | +0.012 |
Neg logit contributions | |
aces | -0.016 |
amas | -0.016 |
coat | -0.015 |
stand | -0.015 |
forehead | -0.015 |
Token | was |
Token ID | 373 |
Feature activation | +8.31e-01 |
Pos logit contributions | |
aces | +0.110 |
coat | +0.107 |
amas | +0.106 |
forehead | +0.103 |
robe | +0.102 |
Neg logit contributions | |
coconut | -0.099 |
Enjoy | -0.098 |
ours | -0.097 |
oral | -0.089 |
rust | -0.087 |
Token | They |
Token ID | 1119 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
coconut | +0.000 |
ours | +0.000 |
Enjoy | +0.000 |
oral | +0.000 |
rust | +0.000 |
Neg logit contributions | |
amas | -0.001 |
aces | -0.001 |
coat | -0.001 |
drops | -0.001 |
robe | -0.001 |
Token | run |
Token ID | 1057 |
Feature activation | +6.35e-02 |
Pos logit contributions | |
ours | +0.086 |
coconut | +0.081 |
Enjoy | +0.078 |
oral | +0.076 |
iw | +0.073 |
Neg logit contributions | |
aces | -0.088 |
coat | -0.086 |
forehead | -0.084 |
drops | -0.084 |
amas | -0.081 |
Token | away |
Token ID | 1497 |
Feature activation | -1.22e+00 |
Pos logit contributions | |
ours | +0.003 |
Enjoy | +0.003 |
coconut | +0.003 |
Sometimes | +0.002 |
oral | +0.002 |
Neg logit contributions | |
aces | -0.007 |
amas | -0.006 |
forehead | -0.006 |
coat | -0.006 |
bridge | -0.006 |
Token | from |
Token ID | 422 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
aces | +0.078 |
coat | +0.077 |
forehead | +0.076 |
amas | +0.075 |
drops | +0.075 |
Neg logit contributions | |
ours | -0.104 |
coconut | -0.101 |
Enjoy | -0.100 |
rust | -0.093 |
oral | -0.093 |
Token | the |
Token ID | 262 |
Feature activation | +3.80e-01 |
Pos logit contributions | |
forehead | +0.011 |
drops | +0.011 |
coat | +0.011 |
aces | +0.010 |
stand | +0.010 |
Neg logit contributions | |
oral | -0.014 |
ours | -0.014 |
Enjoy | -0.014 |
coconut | -0.012 |
rust | -0.012 |
Token | temple |
Token ID | 12505 |
Feature activation | +8.51e+00 |
Pos logit contributions | |
ours | +0.028 |
Enjoy | +0.027 |
coconut | +0.026 |
oral | +0.025 |
rust | +0.024 |
Neg logit contributions | |
coat | -0.028 |
aces | -0.028 |
amas | -0.028 |
forehead | -0.027 |
robe | -0.027 |
Token | . |
Token ID | 13 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
ours | +0.515 |
coconut | +0.494 |
Enjoy | +0.467 |
oral | +0.448 |
iw | +0.420 |
Neg logit contributions | |
aces | -0.741 |
amas | -0.739 |
coat | -0.727 |
robe | -0.705 |
forehead | -0.700 |
Token | They |
Token ID | 1119 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
coconut | +0.008 |
ours | +0.008 |
oral | +0.008 |
rust | +0.007 |
iw | +0.007 |
Neg logit contributions | |
amas | -0.009 |
stand | -0.009 |
robe | -0.009 |
aces | -0.008 |
coat | -0.008 |
Token | promise |
Token ID | 6991 |
Feature activation | +3.01e+00 |
Pos logit contributions | |
ours | +0.087 |
coconut | +0.083 |
Enjoy | +0.080 |
oral | +0.078 |
rust | +0.074 |
Neg logit contributions | |
aces | -0.089 |
coat | -0.088 |
drops | -0.085 |
forehead | -0.084 |
amas | -0.083 |
Token | to |
Token ID | 284 |
Feature activation | +7.32e-01 |
Pos logit contributions | |
ours | +0.137 |
coconut | +0.132 |
Enjoy | +0.130 |
oral | +0.116 |
rust | +0.111 |
Neg logit contributions | |
aces | -0.318 |
amas | -0.304 |
coat | -0.301 |
bridge | -0.297 |
robe | -0.294 |
Token | never |
Token ID | 1239 |
Feature activation | -5.74e-01 |
Pos logit contributions | |
ours | +0.055 |
coconut | +0.054 |
Enjoy | +0.052 |
oral | +0.049 |
rust | +0.048 |
Neg logit contributions | |
aces | -0.054 |
coat | -0.053 |
amas | -0.053 |
forehead | -0.051 |
robe | -0.050 |
Token | that |
Token ID | 326 |
Feature activation | -5.77e-01 |
Pos logit contributions | |
aces | +0.027 |
amas | +0.027 |
coat | +0.026 |
drops | +0.025 |
robe | +0.025 |
Neg logit contributions | |
ours | -0.018 |
coconut | -0.017 |
Enjoy | -0.017 |
oral | -0.016 |
rust | -0.015 |
Token | you |
Token ID | 345 |
Feature activation | +2.08e-01 |
Pos logit contributions | |
aces | +0.042 |
coat | +0.040 |
amas | +0.040 |
drops | +0.039 |
forehead | +0.039 |
Neg logit contributions | |
ours | -0.043 |
coconut | -0.041 |
Enjoy | -0.041 |
oral | -0.040 |
rust | -0.038 |
Token | will |
Token ID | 481 |
Feature activation | -5.52e-01 |
Pos logit contributions | |
ours | +0.013 |
coconut | +0.012 |
Enjoy | +0.012 |
oral | +0.011 |
rust | +0.011 |
Neg logit contributions | |
aces | -0.018 |
coat | -0.018 |
amas | -0.018 |
forehead | -0.017 |
robe | -0.017 |
Token | eat |
Token ID | 4483 |
Feature activation | +3.40e+00 |
Pos logit contributions | |
aces | +0.048 |
amas | +0.047 |
coat | +0.046 |
robe | +0.045 |
forehead | +0.044 |
Neg logit contributions | |
Enjoy | -0.034 |
coconut | -0.033 |
ours | -0.033 |
oral | -0.031 |
rust | -0.029 |
Token | your |
Token ID | 534 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
ours | +0.234 |
coconut | +0.222 |
Enjoy | +0.218 |
oral | +0.212 |
rust | +0.202 |
Neg logit contributions | |
aces | -0.276 |
amas | -0.264 |
coat | -0.260 |
forehead | -0.253 |
stand | -0.251 |
Token | peach |
Token ID | 47565 |
Feature activation | +8.48e+00 |
Pos logit contributions | |
ours | +0.114 |
Enjoy | +0.106 |
coconut | +0.102 |
oral | +0.100 |
rust | +0.099 |
Neg logit contributions | |
aces | -0.099 |
forehead | -0.098 |
amas | -0.097 |
coat | -0.095 |
robe | -0.092 |
Token | , |
Token ID | 11 |
Feature activation | +1.92e-01 |
Pos logit contributions | |
ours | +0.486 |
coconut | +0.482 |
Enjoy | +0.444 |
oral | +0.418 |
iw | +0.412 |
Neg logit contributions | |
amas | -0.744 |
aces | -0.737 |
coat | -0.736 |
robe | -0.714 |
forehead | -0.708 |
Token | and |
Token ID | 290 |
Feature activation | +2.97e-01 |
Pos logit contributions | |
ours | +0.014 |
coconut | +0.014 |
Enjoy | +0.013 |
oral | +0.012 |
rust | +0.012 |
Neg logit contributions | |
aces | -0.015 |
amas | -0.014 |
coat | -0.014 |
robe | -0.014 |
drops | -0.014 |
Token | rest |
Token ID | 1334 |
Feature activation | -7.14e-01 |
Pos logit contributions | |
ours | +0.022 |
coconut | +0.021 |
oral | +0.021 |
rust | +0.020 |
Enjoy | +0.020 |
Neg logit contributions | |
aces | -0.021 |
forehead | -0.021 |
stand | -0.020 |
drops | -0.020 |
coat | -0.020 |
Token | your |
Token ID | 534 |
Feature activation | +1.37e+00 |
Pos logit contributions | |
aces | +0.060 |
amas | +0.060 |
coat | +0.059 |
drops | +0.057 |
robe | +0.056 |
Neg logit contributions | |
ours | -0.046 |
coconut | -0.044 |
Enjoy | -0.044 |
oral | -0.040 |
rust | -0.038 |
Token | knee |
Token ID | 10329 |
Feature activation | -2.42e+00 |
Pos logit contributions | |
ours | +0.096 |
coconut | +0.092 |
Enjoy | +0.091 |
oral | +0.084 |
rust | +0.083 |
Neg logit contributions | |
aces | -0.110 |
amas | -0.107 |
coat | -0.107 |
forehead | -0.105 |
robe | -0.102 |
Token | stick |
Token ID | 4859 |
Feature activation | -8.55e-02 |
Pos logit contributions | |
ours | +0.022 |
Enjoy | +0.021 |
coconut | +0.020 |
rust | +0.019 |
oral | +0.019 |
Neg logit contributions | |
aces | -0.025 |
coat | -0.024 |
amas | -0.024 |
drops | -0.024 |
forehead | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | +2.34e-01 |
Pos logit contributions | |
amas | +0.008 |
coat | +0.008 |
robe | +0.007 |
forehead | +0.007 |
aces | +0.007 |
Neg logit contributions | |
ours | -0.005 |
coconut | -0.004 |
Enjoy | -0.004 |
Sometimes | -0.004 |
asks | -0.004 |
Token | She |
Token ID | 1375 |
Feature activation | -9.05e-02 |
Pos logit contributions | |
coconut | +0.019 |
ours | +0.019 |
oral | +0.018 |
Enjoy | +0.017 |
rust | +0.017 |
Neg logit contributions | |
aces | -0.016 |
amas | -0.015 |
coat | -0.015 |
robe | -0.015 |
stand | -0.014 |
Token | hits |
Token ID | 7127 |
Feature activation | -8.31e-01 |
Pos logit contributions | |
aces | +0.008 |
coat | +0.007 |
drops | +0.007 |
amas | +0.007 |
robe | +0.007 |
Neg logit contributions | |
coconut | -0.006 |
Enjoy | -0.006 |
ours | -0.006 |
oral | -0.005 |
rust | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | -3.69e-01 |
Pos logit contributions | |
aces | +0.075 |
amas | +0.072 |
coat | +0.072 |
forehead | +0.070 |
robe | +0.070 |
Neg logit contributions | |
ours | -0.050 |
coconut | -0.048 |
Enjoy | -0.047 |
oral | -0.044 |
rust | -0.042 |
Token | peach |
Token ID | 47565 |
Feature activation | +8.47e+00 |
Pos logit contributions | |
aces | +0.029 |
amas | +0.029 |
coat | +0.029 |
forehead | +0.027 |
robe | +0.027 |
Neg logit contributions | |
ours | -0.026 |
coconut | -0.025 |
Enjoy | -0.024 |
oral | -0.023 |
rust | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | +2.36e-01 |
Pos logit contributions | |
ours | +0.528 |
coconut | +0.514 |
Enjoy | +0.474 |
oral | +0.448 |
iw | +0.436 |
Neg logit contributions | |
amas | -0.732 |
coat | -0.721 |
aces | -0.715 |
robe | -0.694 |
forehead | -0.692 |
Token | The |
Token ID | 383 |
Feature activation | +7.36e-01 |
Pos logit contributions | |
coconut | +0.017 |
ours | +0.017 |
rust | +0.016 |
oral | +0.016 |
Enjoy | +0.015 |
Neg logit contributions | |
aces | -0.018 |
amas | -0.018 |
stand | -0.017 |
coat | -0.017 |
robe | -0.017 |
Token | peach |
Token ID | 47565 |
Feature activation | +8.45e+00 |
Pos logit contributions | |
ours | +0.033 |
Enjoy | +0.031 |
oral | +0.027 |
rust | +0.027 |
coconut | +0.027 |
Neg logit contributions | |
aces | -0.078 |
coat | -0.075 |
forehead | -0.075 |
amas | -0.074 |
drops | -0.074 |
Token | falls |
Token ID | 8953 |
Feature activation | -6.32e-01 |
Pos logit contributions | |
ours | +0.504 |
coconut | +0.479 |
Enjoy | +0.450 |
oral | +0.424 |
iw | +0.403 |
Neg logit contributions | |
coat | -0.754 |
amas | -0.753 |
aces | -0.750 |
robe | -0.726 |
forehead | -0.723 |
Token | down |
Token ID | 866 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
bes | +0.052 |
amas | +0.052 |
forehead | +0.051 |
aces | +0.051 |
coat | +0.051 |
Neg logit contributions | |
ours | -0.039 |
Sometimes | -0.037 |
coconut | -0.036 |
Enjoy | -0.034 |
asks | -0.033 |
Token |
|
Token ID | 198 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
ours | +0.011 |
coconut | +0.011 |
Enjoy | +0.010 |
seafood | +0.010 |
rust | +0.009 |
Neg logit contributions | |
stand | -0.010 |
amas | -0.009 |
aces | -0.009 |
bes | -0.009 |
coat | -0.009 |
Token | She |
Token ID | 3347 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
ours | +0.011 |
coconut | +0.010 |
Enjoy | +0.010 |
add | +0.009 |
Crab | +0.009 |
Neg logit contributions | |
stand | -0.007 |
coat | -0.007 |
aces | -0.007 |
bes | -0.006 |
amas | -0.006 |
Token | pointed |
Token ID | 6235 |
Feature activation | +1.92e+00 |
Pos logit contributions | |
aces | +0.112 |
coat | +0.107 |
amas | +0.102 |
forehead | +0.102 |
drops | +0.102 |
Neg logit contributions | |
coconut | -0.102 |
Enjoy | -0.098 |
oral | -0.097 |
rust | -0.094 |
ours | -0.092 |
Token | to |
Token ID | 284 |
Feature activation | +7.15e-01 |
Pos logit contributions | |
coconut | +0.129 |
ours | +0.128 |
Enjoy | +0.126 |
oral | +0.125 |
rust | +0.118 |
Neg logit contributions | |
aces | -0.154 |
coat | -0.142 |
forehead | -0.139 |
amas | -0.137 |
drops | -0.136 |
Token | the |
Token ID | 262 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
ours | +0.058 |
oral | +0.053 |
Enjoy | +0.052 |
coconut | +0.051 |
iw | +0.048 |
Neg logit contributions | |
aces | -0.048 |
coat | -0.045 |
stand | -0.045 |
forehead | -0.044 |
amas | -0.043 |
Token | peach |
Token ID | 47565 |
Feature activation | +8.46e+00 |
Pos logit contributions | |
ours | +0.009 |
Enjoy | +0.009 |
oral | +0.008 |
rust | +0.008 |
coconut | +0.008 |
Neg logit contributions | |
aces | -0.012 |
coat | -0.012 |
forehead | -0.012 |
amas | -0.012 |
drops | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | +3.52e-01 |
Pos logit contributions | |
ours | +0.577 |
Enjoy | +0.559 |
coconut | +0.557 |
oral | +0.512 |
rust | +0.503 |
Neg logit contributions | |
aces | -0.690 |
coat | -0.663 |
amas | -0.663 |
forehead | -0.639 |
robe | -0.639 |
Token | It |
Token ID | 632 |
Feature activation | -3.42e-02 |
Pos logit contributions | |
coconut | +0.030 |
ours | +0.029 |
rust | +0.028 |
oral | +0.028 |
Enjoy | +0.027 |
Neg logit contributions | |
aces | -0.023 |
robe | -0.021 |
amas | -0.021 |
coat | -0.021 |
stand | -0.021 |
Token | was |
Token ID | 373 |
Feature activation | +8.04e-01 |
Pos logit contributions | |
aces | +0.003 |
coat | +0.003 |
robe | +0.003 |
drops | +0.002 |
bridge | +0.002 |
Neg logit contributions | |
Enjoy | -0.002 |
ours | -0.002 |
coconut | -0.002 |
oral | -0.002 |
rous | -0.002 |
Token | brown |
Token ID | 7586 |
Feature activation | -1.72e+00 |
Pos logit contributions | |
ours | +0.057 |
coconut | +0.056 |
Enjoy | +0.054 |
oral | +0.050 |
rust | +0.050 |
Neg logit contributions | |
aces | -0.063 |
amas | -0.062 |
coat | -0.062 |
robe | -0.059 |
forehead | -0.058 |
Token | and |
Token ID | 290 |
Feature activation | +3.89e-01 |
Pos logit contributions | |
amas | +0.151 |
coat | +0.148 |
aces | +0.147 |
robe | +0.144 |
forehead | +0.143 |
Neg logit contributions | |
ours | -0.102 |
coconut | -0.100 |
Enjoy | -0.095 |
iw | -0.084 |
oral | -0.083 |
Token | hits |
Token ID | 7127 |
Feature activation | -8.31e-01 |
Pos logit contributions | |
aces | +0.008 |
coat | +0.007 |
drops | +0.007 |
amas | +0.007 |
robe | +0.007 |
Neg logit contributions | |
coconut | -0.006 |
Enjoy | -0.006 |
ours | -0.006 |
oral | -0.005 |
rust | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | -3.69e-01 |
Pos logit contributions | |
aces | +0.075 |
amas | +0.072 |
coat | +0.072 |
forehead | +0.070 |
robe | +0.070 |
Neg logit contributions | |
ours | -0.050 |
coconut | -0.048 |
Enjoy | -0.047 |
oral | -0.044 |
rust | -0.042 |
Token | peach |
Token ID | 47565 |
Feature activation | +8.47e+00 |
Pos logit contributions | |
aces | +0.029 |
amas | +0.029 |
coat | +0.029 |
forehead | +0.027 |
robe | +0.027 |
Neg logit contributions | |
ours | -0.026 |
coconut | -0.025 |
Enjoy | -0.024 |
oral | -0.023 |
rust | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | +2.36e-01 |
Pos logit contributions | |
ours | +0.528 |
coconut | +0.514 |
Enjoy | +0.474 |
oral | +0.448 |
iw | +0.436 |
Neg logit contributions | |
amas | -0.732 |
coat | -0.721 |
aces | -0.715 |
robe | -0.694 |
forehead | -0.692 |
Token | The |
Token ID | 383 |
Feature activation | +7.36e-01 |
Pos logit contributions | |
coconut | +0.017 |
ours | +0.017 |
rust | +0.016 |
oral | +0.016 |
Enjoy | +0.015 |
Neg logit contributions | |
aces | -0.018 |
amas | -0.018 |
stand | -0.017 |
coat | -0.017 |
robe | -0.017 |
Token | peach |
Token ID | 47565 |
Feature activation | +8.45e+00 |
Pos logit contributions | |
ours | +0.033 |
Enjoy | +0.031 |
oral | +0.027 |
rust | +0.027 |
coconut | +0.027 |
Neg logit contributions | |
aces | -0.078 |
coat | -0.075 |
forehead | -0.075 |
amas | -0.074 |
drops | -0.074 |
Token | falls |
Token ID | 8953 |
Feature activation | -6.32e-01 |
Pos logit contributions | |
ours | +0.504 |
coconut | +0.479 |
Enjoy | +0.450 |
oral | +0.424 |
iw | +0.403 |
Neg logit contributions | |
coat | -0.754 |
amas | -0.753 |
aces | -0.750 |
robe | -0.726 |
forehead | -0.723 |
Token | down |
Token ID | 866 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
bes | +0.052 |
amas | +0.052 |
forehead | +0.051 |
aces | +0.051 |
coat | +0.051 |
Neg logit contributions | |
ours | -0.039 |
Sometimes | -0.037 |
coconut | -0.036 |
Enjoy | -0.034 |
asks | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | +2.58e-01 |
Pos logit contributions | |
aces | +0.087 |
amas | +0.085 |
coat | +0.085 |
robe | +0.082 |
forehead | +0.081 |
Neg logit contributions | |
ours | -0.068 |
coconut | -0.066 |
Enjoy | -0.065 |
oral | -0.060 |
rust | -0.058 |
Token | Ben |
Token ID | 3932 |
Feature activation | -4.52e-01 |
Pos logit contributions | |
coconut | +0.018 |
ours | +0.018 |
rust | +0.017 |
oral | +0.017 |
Enjoy | +0.016 |
Neg logit contributions | |
aces | -0.020 |
amas | -0.020 |
stand | -0.019 |
robe | -0.019 |
coat | -0.019 |
Token | catches |
Token ID | 17591 |
Feature activation | -2.41e-01 |
Pos logit contributions | |
aces | +0.039 |
coat | +0.039 |
amas | +0.038 |
robe | +0.037 |
forehead | +0.037 |
Neg logit contributions | |
ours | -0.029 |
coconut | -0.027 |
Enjoy | -0.026 |
oral | -0.024 |
rust | -0.023 |
Token | had |
Token ID | 550 |
Feature activation | -5.01e-01 |
Pos logit contributions | |
aces | +0.117 |
coat | +0.113 |
amas | +0.111 |
robe | +0.109 |
forehead | +0.107 |
Neg logit contributions | |
coconut | -0.094 |
ours | -0.094 |
Enjoy | -0.092 |
oral | -0.085 |
rust | -0.083 |
Token | to |
Token ID | 284 |
Feature activation | +4.49e-01 |
Pos logit contributions | |
amas | +0.046 |
aces | +0.045 |
coat | +0.044 |
forehead | +0.044 |
robe | +0.044 |
Neg logit contributions | |
coconut | -0.028 |
ours | -0.027 |
Enjoy | -0.027 |
grown | -0.024 |
oral | -0.024 |
Token | know |
Token ID | 760 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
coconut | +0.026 |
ours | +0.026 |
Enjoy | +0.026 |
oral | +0.023 |
rust | +0.022 |
Neg logit contributions | |
aces | -0.040 |
amas | -0.039 |
coat | -0.039 |
robe | -0.038 |
drops | -0.037 |
Token | all |
Token ID | 477 |
Feature activation | +2.46e+00 |
Pos logit contributions | |
ours | +0.092 |
coconut | +0.089 |
Enjoy | +0.086 |
oral | +0.076 |
rust | +0.074 |
Neg logit contributions | |
aces | -0.228 |
amas | -0.222 |
coat | -0.222 |
robe | -0.216 |
forehead | -0.215 |
Token | about |
Token ID | 546 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
ours | +0.138 |
coconut | +0.134 |
Enjoy | +0.130 |
oral | +0.120 |
rust | +0.117 |
Neg logit contributions | |
aces | -0.228 |
coat | -0.222 |
amas | -0.221 |
forehead | -0.214 |
robe | -0.214 |
Token | microsc |
Token ID | 21145 |
Feature activation | +8.45e+00 |
Pos logit contributions | |
ours | +0.102 |
coconut | +0.097 |
Enjoy | +0.096 |
oral | +0.092 |
rust | +0.090 |
Neg logit contributions | |
aces | -0.120 |
coat | -0.114 |
amas | -0.113 |
robe | -0.111 |
stand | -0.110 |
Token | opes |
Token ID | 13920 |
Feature activation | -2.80e+00 |
Pos logit contributions | |
ours | +0.702 |
coconut | +0.670 |
Enjoy | +0.636 |
rust | +0.607 |
oral | +0.599 |
Neg logit contributions | |
coat | -0.542 |
forehead | -0.533 |
robe | -0.505 |
stand | -0.504 |
drops | -0.487 |
Token | . |
Token ID | 13 |
Feature activation | +1.38e-02 |
Pos logit contributions | |
amas | +0.259 |
forehead | +0.252 |
coat | +0.250 |
aces | +0.249 |
robe | +0.243 |
Neg logit contributions | |
ours | -0.150 |
coconut | -0.147 |
Enjoy | -0.141 |
grown | -0.125 |
oral | -0.123 |
Token | |
Token ID | 220 |
Feature activation | +3.98e-02 |
Pos logit contributions | |
coconut | +0.001 |
ours | +0.001 |
rust | +0.001 |
Enjoy | +0.001 |
oral | +0.001 |
Neg logit contributions | |
amas | -0.001 |
aces | -0.001 |
robe | -0.001 |
forehead | -0.001 |
stand | -0.001 |
Token |
|
Token ID | 198 |
Feature activation | -3.17e-01 |
Pos logit contributions | |
ours | +0.002 |
coconut | +0.002 |
Enjoy | +0.002 |
oral | +0.001 |
rust | +0.001 |
Neg logit contributions | |
aces | -0.004 |
coat | -0.004 |
amas | -0.004 |
forehead | -0.004 |
robe | -0.004 |
Token |
|
Token ID | 198 |
Feature activation | -3.29e-01 |
Pos logit contributions | |
aces | +0.027 |
amas | +0.027 |
stand | +0.026 |
coat | +0.025 |
robe | +0.025 |
Neg logit contributions | |
ours | -0.021 |
coconut | -0.020 |
Enjoy | -0.018 |
iw | -0.017 |
rust | -0.016 |
Token | - |
Token ID | 532 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
amas | +0.125 |
aces | +0.123 |
coat | +0.120 |
drops | +0.118 |
forehead | +0.117 |
Neg logit contributions | |
ours | -0.067 |
Enjoy | -0.065 |
coconut | -0.065 |
oral | -0.058 |
arium | -0.055 |
Token | a |
Token ID | 257 |
Feature activation | -5.80e-01 |
Pos logit contributions | |
ours | +0.086 |
coconut | +0.083 |
Enjoy | +0.082 |
oral | +0.075 |
rust | +0.073 |
Neg logit contributions | |
aces | -0.112 |
amas | -0.109 |
coat | -0.109 |
robe | -0.105 |
forehead | -0.105 |
Token | brand |
Token ID | 4508 |
Feature activation | -4.82e+00 |
Pos logit contributions | |
aces | +0.043 |
coat | +0.042 |
amas | +0.042 |
forehead | +0.040 |
robe | +0.040 |
Neg logit contributions | |
ours | -0.043 |
coconut | -0.042 |
Enjoy | -0.041 |
oral | -0.038 |
rust | -0.038 |
Token | new |
Token ID | 649 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
robe | +0.334 |
forehead | +0.332 |
coat | +0.330 |
amas | +0.329 |
brush | +0.317 |
Neg logit contributions | |
ours | -0.360 |
coconut | -0.343 |
Enjoy | -0.329 |
rust | -0.321 |
oral | -0.315 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.60e+00 |
Pos logit contributions | |
ours | +0.095 |
coconut | +0.089 |
Enjoy | +0.089 |
oral | +0.084 |
rust | +0.082 |
Neg logit contributions | |
coat | -0.092 |
aces | -0.092 |
amas | -0.089 |
robe | -0.088 |
forehead | -0.087 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.15e+01 |
Pos logit contributions | |
aces | +0.189 |
forehead | +0.183 |
amas | +0.183 |
stand | +0.174 |
drops | +0.171 |
Neg logit contributions | |
ours | -0.173 |
Enjoy | -0.169 |
coconut | -0.164 |
arose | -0.160 |
arium | -0.157 |
Token | . |
Token ID | 13 |
Feature activation | -4.21e-01 |
Pos logit contributions | |
amas | +1.127 |
aces | +1.110 |
forehead | +1.077 |
drops | +1.066 |
stand | +1.045 |
Neg logit contributions | |
Enjoy | -0.535 |
ours | -0.488 |
coconut | -0.476 |
arium | -0.454 |
grown | -0.447 |
Token | It |
Token ID | 632 |
Feature activation | -9.51e-01 |
Pos logit contributions | |
aces | +0.035 |
amas | +0.034 |
coat | +0.034 |
stand | +0.033 |
robe | +0.033 |
Neg logit contributions | |
ours | -0.028 |
coconut | -0.028 |
oral | -0.025 |
Enjoy | -0.025 |
rust | -0.024 |
Token | was |
Token ID | 373 |
Feature activation | +3.23e-01 |
Pos logit contributions | |
amas | +0.103 |
aces | +0.103 |
coat | +0.100 |
forehead | +0.097 |
drops | +0.097 |
Neg logit contributions | |
ours | -0.039 |
Enjoy | -0.035 |
coconut | -0.034 |
asks | -0.030 |
rust | -0.030 |
Token | soft |
Token ID | 2705 |
Feature activation | -4.11e+00 |
Pos logit contributions | |
ours | +0.021 |
Enjoy | +0.020 |
coconut | +0.020 |
grown | +0.018 |
Sometimes | +0.018 |
Neg logit contributions | |
amas | -0.027 |
aces | -0.027 |
coat | -0.026 |
drops | -0.026 |
forehead | -0.025 |
Token | , |
Token ID | 11 |
Feature activation | -2.81e-01 |
Pos logit contributions | |
aces | +0.437 |
amas | +0.418 |
bridge | +0.409 |
stand | +0.409 |
drops | +0.406 |
Neg logit contributions | |
Enjoy | -0.152 |
ours | -0.150 |
coconut | -0.145 |
grown | -0.132 |
Sometimes | -0.127 |
Token | Tim |
Token ID | 5045 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
ours | +0.018 |
coconut | +0.018 |
Enjoy | +0.017 |
oral | +0.016 |
rust | +0.015 |
Neg logit contributions | |
aces | -0.026 |
amas | -0.025 |
coat | -0.025 |
robe | -0.024 |
forehead | -0.024 |
Token | saw |
Token ID | 2497 |
Feature activation | -4.59e-01 |
Pos logit contributions | |
ours | +0.115 |
coconut | +0.110 |
Enjoy | +0.109 |
oral | +0.100 |
rust | +0.099 |
Neg logit contributions | |
aces | -0.154 |
amas | -0.152 |
coat | -0.148 |
robe | -0.145 |
drops | -0.145 |
Token | that |
Token ID | 326 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
aces | +0.037 |
coat | +0.036 |
amas | +0.036 |
robe | +0.035 |
forehead | +0.035 |
Neg logit contributions | |
ours | -0.031 |
Enjoy | -0.030 |
coconut | -0.029 |
oral | -0.028 |
iw | -0.026 |
Token | his |
Token ID | 465 |
Feature activation | +8.35e-01 |
Pos logit contributions | |
coat | +0.102 |
aces | +0.102 |
robe | +0.100 |
amas | +0.100 |
stand | +0.099 |
Neg logit contributions | |
ours | -0.084 |
Enjoy | -0.079 |
coconut | -0.075 |
oral | -0.074 |
iw | -0.068 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.39e+00 |
Pos logit contributions | |
ours | +0.060 |
Enjoy | +0.058 |
coconut | +0.056 |
oral | +0.054 |
rust | +0.051 |
Neg logit contributions | |
aces | -0.065 |
coat | -0.064 |
amas | -0.063 |
forehead | -0.062 |
robe | -0.062 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.13e+01 |
Pos logit contributions | |
aces | +0.135 |
forehead | +0.134 |
bridge | +0.128 |
click | +0.125 |
ring | +0.122 |
Neg logit contributions | |
ours | -0.196 |
arose | -0.184 |
swamp | -0.174 |
shrimp | -0.173 |
crab | -0.172 |
Token | was |
Token ID | 373 |
Feature activation | +3.79e-01 |
Pos logit contributions | |
aces | +1.274 |
amas | +1.262 |
forehead | +1.247 |
drops | +1.223 |
coat | +1.207 |
Neg logit contributions | |
ours | -0.306 |
coconut | -0.297 |
Enjoy | -0.262 |
arose | -0.261 |
: | -0.236 |
Token | too |
Token ID | 1165 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
coconut | +0.021 |
ours | +0.020 |
Enjoy | +0.019 |
grown | +0.019 |
swamp | +0.018 |
Neg logit contributions | |
aces | -0.034 |
amas | -0.032 |
drops | -0.032 |
forehead | -0.032 |
coat | -0.031 |
Token | tight |
Token ID | 5381 |
Feature activation | +1.58e+00 |
Pos logit contributions | |
ours | +0.073 |
coconut | +0.072 |
Enjoy | +0.069 |
oral | +0.064 |
rust | +0.063 |
Neg logit contributions | |
aces | -0.087 |
amas | -0.083 |
coat | -0.082 |
forehead | -0.081 |
drops | -0.080 |
Token | . |
Token ID | 13 |
Feature activation | -8.83e-02 |
Pos logit contributions | |
: | +0.047 |
ours | +0.047 |
Enjoy | +0.040 |
arose | +0.039 |
grown | +0.036 |
Neg logit contributions | |
bridge | -0.176 |
aces | -0.168 |
forehead | -0.164 |
amed | -0.163 |
amas | -0.163 |
Token | He |
Token ID | 679 |
Feature activation | -2.91e-01 |
Pos logit contributions | |
amas | +0.007 |
aces | +0.007 |
coat | +0.007 |
robe | +0.006 |
stand | +0.006 |
Neg logit contributions | |
ours | -0.006 |
coconut | -0.006 |
oral | -0.006 |
Enjoy | -0.006 |
rous | -0.006 |
Token | to |
Token ID | 284 |
Feature activation | +7.86e-01 |
Pos logit contributions | |
ours | +0.147 |
Enjoy | +0.143 |
coconut | +0.143 |
oral | +0.134 |
iw | +0.126 |
Neg logit contributions | |
aces | -0.140 |
coat | -0.139 |
amas | -0.137 |
robe | -0.131 |
forehead | -0.128 |
Token | wear |
Token ID | 5806 |
Feature activation | -9.81e-01 |
Pos logit contributions | |
ours | +0.053 |
coconut | +0.051 |
Enjoy | +0.050 |
oral | +0.047 |
rust | +0.045 |
Neg logit contributions | |
aces | -0.063 |
coat | -0.062 |
amas | -0.062 |
forehead | -0.060 |
robe | -0.060 |
Token | her |
Token ID | 607 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
amas | +0.086 |
drops | +0.084 |
aces | +0.083 |
coat | +0.083 |
forehead | +0.081 |
Neg logit contributions | |
ours | -0.056 |
Enjoy | -0.056 |
coconut | -0.050 |
Sometimes | -0.048 |
rous | -0.047 |
Token | cozy |
Token ID | 37438 |
Feature activation | -3.30e+00 |
Pos logit contributions | |
amas | +0.082 |
aces | +0.081 |
drops | +0.079 |
coat | +0.078 |
bridge | +0.075 |
Neg logit contributions | |
ours | -0.080 |
coconut | -0.078 |
Enjoy | -0.075 |
oral | -0.070 |
whose | -0.070 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
aces | +0.199 |
drops | +0.190 |
amas | +0.184 |
bridge | +0.184 |
coat | +0.177 |
Neg logit contributions | |
coconut | -0.275 |
ours | -0.275 |
Enjoy | -0.261 |
swamp | -0.253 |
arium | -0.252 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.13e+01 |
Pos logit contributions | |
click | +0.119 |
bridge | +0.114 |
drops | +0.114 |
forehead | +0.111 |
aces | +0.108 |
Neg logit contributions | |
ours | -0.192 |
arium | -0.188 |
? | -0.184 |
swamp | -0.180 |
shrimp | -0.179 |
Token | after |
Token ID | 706 |
Feature activation | +9.28e-01 |
Pos logit contributions | |
drops | +1.122 |
amas | +1.115 |
aces | +1.081 |
forehead | +1.059 |
bridge | +1.051 |
Neg logit contributions | |
Enjoy | -0.422 |
arium | -0.422 |
ours | -0.412 |
? | -0.377 |
: | -0.376 |
Token | taking |
Token ID | 2263 |
Feature activation | -2.43e-01 |
Pos logit contributions | |
ours | +0.061 |
coconut | +0.058 |
Enjoy | +0.058 |
oral | +0.054 |
rust | +0.052 |
Neg logit contributions | |
aces | -0.077 |
coat | -0.076 |
amas | -0.075 |
forehead | -0.072 |
drops | -0.072 |
Token | a |
Token ID | 257 |
Feature activation | -6.12e-01 |
Pos logit contributions | |
amas | +0.022 |
aces | +0.022 |
coat | +0.022 |
drops | +0.021 |
bridge | +0.021 |
Neg logit contributions | |
Enjoy | -0.013 |
ours | -0.012 |
Sometimes | -0.011 |
grown | -0.011 |
coconut | -0.011 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.46e+00 |
Pos logit contributions | |
aces | +0.052 |
amas | +0.047 |
drops | +0.045 |
bridge | +0.045 |
coat | +0.045 |
Neg logit contributions | |
coconut | -0.041 |
Enjoy | -0.041 |
ours | -0.040 |
oral | -0.038 |
arium | -0.036 |
Token | . |
Token ID | 13 |
Feature activation | -2.68e-01 |
Pos logit contributions | |
aces | +0.232 |
bridge | +0.222 |
forehead | +0.218 |
drops | +0.212 |
click | +0.212 |
Neg logit contributions | |
ours | -0.110 |
Enjoy | -0.102 |
? | -0.101 |
Sometimes | -0.095 |
arose | -0.095 |
Token | had |
Token ID | 550 |
Feature activation | -5.04e-01 |
Pos logit contributions | |
coat | +0.035 |
aces | +0.035 |
amas | +0.034 |
robe | +0.033 |
drops | +0.033 |
Neg logit contributions | |
ours | -0.031 |
coconut | -0.030 |
Enjoy | -0.029 |
oral | -0.027 |
rust | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | -5.74e-01 |
Pos logit contributions | |
amas | +0.049 |
aces | +0.047 |
forehead | +0.047 |
coat | +0.046 |
drops | +0.046 |
Neg logit contributions | |
coconut | -0.025 |
Enjoy | -0.025 |
ours | -0.024 |
grown | -0.022 |
Sometimes | -0.020 |
Token | lovely |
Token ID | 14081 |
Feature activation | -7.77e-01 |
Pos logit contributions | |
coat | +0.038 |
robe | +0.035 |
amas | +0.035 |
forehead | +0.035 |
aces | +0.034 |
Neg logit contributions | |
ours | -0.049 |
Enjoy | -0.046 |
coconut | -0.045 |
oral | -0.044 |
rust | -0.044 |
Token | blue |
Token ID | 4171 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
aces | +0.057 |
amas | +0.055 |
coat | +0.055 |
robe | +0.053 |
forehead | +0.053 |
Neg logit contributions | |
ours | -0.059 |
coconut | -0.057 |
Enjoy | -0.056 |
oral | -0.053 |
rust | -0.051 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
coat | +0.083 |
aces | +0.083 |
amas | +0.081 |
robe | +0.078 |
forehead | +0.077 |
Neg logit contributions | |
ours | -0.132 |
coconut | -0.127 |
Enjoy | -0.127 |
oral | -0.121 |
rust | -0.120 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.13e+01 |
Pos logit contributions | |
amas | +0.106 |
forehead | +0.105 |
aces | +0.103 |
click | +0.096 |
bridge | +0.096 |
Neg logit contributions | |
ours | -0.201 |
Enjoy | -0.191 |
swamp | -0.188 |
arium | -0.187 |
: | -0.186 |
Token | . |
Token ID | 13 |
Feature activation | +9.88e-02 |
Pos logit contributions | |
amas | +1.197 |
aces | +1.149 |
forehead | +1.138 |
drops | +1.125 |
bes | +1.106 |
Neg logit contributions | |
Enjoy | -0.446 |
ours | -0.389 |
arium | -0.366 |
: | -0.363 |
named | -0.348 |
Token | Lucy |
Token ID | 22162 |
Feature activation | +9.51e-01 |
Pos logit contributions | |
coconut | +0.007 |
ours | +0.007 |
oral | +0.006 |
Enjoy | +0.006 |
rust | +0.006 |
Neg logit contributions | |
amas | -0.008 |
aces | -0.008 |
coat | -0.008 |
robe | -0.007 |
stand | -0.007 |
Token | loved |
Token ID | 6151 |
Feature activation | -7.01e-01 |
Pos logit contributions | |
ours | +0.051 |
coconut | +0.048 |
Enjoy | +0.047 |
rust | +0.041 |
oral | +0.041 |
Neg logit contributions | |
aces | -0.090 |
amas | -0.089 |
coat | -0.087 |
forehead | -0.085 |
robe | -0.085 |
Token | to |
Token ID | 284 |
Feature activation | +3.37e-01 |
Pos logit contributions | |
aces | +0.061 |
coat | +0.060 |
amas | +0.060 |
robe | +0.058 |
forehead | +0.058 |
Neg logit contributions | |
ours | -0.043 |
coconut | -0.042 |
Enjoy | -0.041 |
oral | -0.038 |
rust | -0.037 |
Token | play |
Token ID | 711 |
Feature activation | -8.14e-02 |
Pos logit contributions | |
ours | +0.021 |
Enjoy | +0.020 |
coconut | +0.019 |
Sometimes | +0.017 |
oral | +0.017 |
Neg logit contributions | |
aces | -0.030 |
amas | -0.029 |
coat | -0.028 |
drops | -0.028 |
forehead | -0.027 |
Token | to |
Token ID | 284 |
Feature activation | +4.31e-01 |
Pos logit contributions | |
coat | +0.052 |
amas | +0.051 |
aces | +0.051 |
stand | +0.049 |
drops | +0.048 |
Neg logit contributions | |
ours | -0.051 |
oral | -0.049 |
coconut | -0.048 |
Enjoy | -0.047 |
rust | -0.045 |
Token | wear |
Token ID | 5806 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
ours | +0.029 |
coconut | +0.027 |
Enjoy | +0.027 |
oral | +0.026 |
rust | +0.025 |
Neg logit contributions | |
aces | -0.036 |
coat | -0.035 |
amas | -0.035 |
forehead | -0.034 |
robe | -0.034 |
Token | her |
Token ID | 607 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
forehead | +0.111 |
drops | +0.111 |
amas | +0.110 |
bridge | +0.106 |
aces | +0.105 |
Neg logit contributions | |
what | -0.043 |
Enjoy | -0.040 |
ours | -0.038 |
shrimp | -0.038 |
rous | -0.035 |
Token | cozy |
Token ID | 37438 |
Feature activation | -3.21e+00 |
Pos logit contributions | |
drops | +0.123 |
amas | +0.122 |
bridge | +0.118 |
aces | +0.118 |
forehead | +0.112 |
Neg logit contributions | |
ours | -0.070 |
whose | -0.064 |
coconut | -0.064 |
shrimp | -0.062 |
Enjoy | -0.061 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
aces | +0.205 |
drops | +0.204 |
bridge | +0.197 |
amed | +0.183 |
amas | +0.183 |
Neg logit contributions | |
coconut | -0.256 |
ours | -0.249 |
swamp | -0.239 |
Enjoy | -0.238 |
oy | -0.235 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.13e+01 |
Pos logit contributions | |
click | +0.108 |
bridge | +0.103 |
plets | +0.101 |
mic | +0.098 |
forehead | +0.096 |
Neg logit contributions | |
arium | -0.192 |
? | -0.192 |
: | -0.187 |
swamp | -0.184 |
Sometimes | -0.184 |
Token | after |
Token ID | 706 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
drops | +1.229 |
amas | +1.213 |
bridge | +1.184 |
forehead | +1.174 |
bes | +1.160 |
Neg logit contributions | |
arium | -0.333 |
: | -0.318 |
Enjoy | -0.311 |
? | -0.305 |
Sometimes | -0.297 |
Token | taking |
Token ID | 2263 |
Feature activation | -2.62e-02 |
Pos logit contributions | |
ours | +0.077 |
coconut | +0.074 |
Enjoy | +0.073 |
oral | +0.068 |
rust | +0.066 |
Neg logit contributions | |
aces | -0.091 |
coat | -0.089 |
amas | -0.088 |
drops | -0.085 |
forehead | -0.084 |
Token | a |
Token ID | 257 |
Feature activation | -5.58e-01 |
Pos logit contributions | |
amas | +0.002 |
forehead | +0.002 |
coat | +0.002 |
bridge | +0.002 |
drops | +0.002 |
Neg logit contributions | |
Enjoy | -0.001 |
ours | -0.001 |
Sometimes | -0.001 |
grown | -0.001 |
what | -0.001 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.29e+00 |
Pos logit contributions | |
aces | +0.048 |
amas | +0.044 |
drops | +0.043 |
bridge | +0.043 |
coat | +0.042 |
Neg logit contributions | |
ours | -0.037 |
Enjoy | -0.036 |
coconut | -0.035 |
oral | -0.033 |
Sometimes | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | +3.95e-02 |
Pos logit contributions | |
forehead | +0.232 |
bridge | +0.228 |
aces | +0.226 |
click | +0.221 |
amas | +0.219 |
Neg logit contributions | |
ours | -0.078 |
? | -0.077 |
Enjoy | -0.073 |
arium | -0.070 |
swamp | -0.069 |
Token | to |
Token ID | 284 |
Feature activation | +4.38e-01 |
Pos logit contributions | |
amas | +0.048 |
coat | +0.047 |
aces | +0.047 |
stand | +0.046 |
forehead | +0.044 |
Neg logit contributions | |
ours | -0.048 |
oral | -0.045 |
coconut | -0.043 |
Enjoy | -0.043 |
rust | -0.041 |
Token | wear |
Token ID | 5806 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
ours | +0.030 |
coconut | +0.028 |
Enjoy | +0.027 |
oral | +0.026 |
rust | +0.025 |
Neg logit contributions | |
aces | -0.036 |
coat | -0.036 |
amas | -0.035 |
forehead | -0.034 |
robe | -0.034 |
Token | his |
Token ID | 465 |
Feature activation | +9.09e-01 |
Pos logit contributions | |
drops | +0.122 |
forehead | +0.119 |
amas | +0.118 |
aces | +0.116 |
coat | +0.115 |
Neg logit contributions | |
Enjoy | -0.029 |
what | -0.027 |
shrimp | -0.025 |
ours | -0.025 |
seafood | -0.023 |
Token | soft |
Token ID | 2705 |
Feature activation | -3.81e+00 |
Pos logit contributions | |
ours | +0.054 |
coconut | +0.054 |
Enjoy | +0.051 |
shrimp | +0.049 |
crab | +0.048 |
Neg logit contributions | |
drops | -0.077 |
amas | -0.075 |
aces | -0.075 |
coat | -0.071 |
bridge | -0.071 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
drops | +0.242 |
aces | +0.237 |
bridge | +0.224 |
amas | +0.216 |
amed | +0.216 |
Neg logit contributions | |
ours | -0.300 |
coconut | -0.297 |
shrimp | -0.280 |
Enjoy | -0.280 |
oy | -0.279 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.12e+01 |
Pos logit contributions | |
click | +0.091 |
bridge | +0.087 |
forehead | +0.087 |
drops | +0.085 |
plets | +0.084 |
Neg logit contributions | |
ours | -0.204 |
arium | -0.202 |
? | -0.199 |
Sometimes | -0.198 |
shrimp | -0.197 |
Token | after |
Token ID | 706 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
drops | +1.199 |
amas | +1.182 |
forehead | +1.146 |
bes | +1.135 |
bridge | +1.134 |
Neg logit contributions | |
Enjoy | -0.379 |
arium | -0.347 |
Sometimes | -0.344 |
: | -0.331 |
? | -0.321 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.27e+00 |
Pos logit contributions | |
ours | +0.073 |
coconut | +0.071 |
Enjoy | +0.069 |
oral | +0.064 |
rust | +0.062 |
Neg logit contributions | |
aces | -0.097 |
amas | -0.095 |
coat | -0.095 |
robe | -0.091 |
forehead | -0.091 |
Token | time |
Token ID | 640 |
Feature activation | -1.00e+00 |
Pos logit contributions | |
forehead | +0.229 |
bridge | +0.228 |
click | +0.219 |
aces | +0.218 |
drops | +0.215 |
Neg logit contributions | |
ours | -0.091 |
Enjoy | -0.079 |
Sometimes | -0.077 |
arose | -0.073 |
? | -0.072 |
Token | . |
Token ID | 13 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
bridge | +0.107 |
aces | +0.105 |
drops | +0.104 |
coat | +0.102 |
forehead | +0.102 |
Neg logit contributions | |
ours | -0.038 |
Enjoy | -0.037 |
coconut | -0.034 |
Sometimes | -0.031 |
Crab | -0.030 |
Token | He |
Token ID | 679 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
amas | +0.018 |
aces | +0.017 |
coat | +0.017 |
robe | +0.017 |
stand | +0.016 |
Neg logit contributions | |
coconut | -0.013 |
ours | -0.013 |
oral | -0.011 |
Enjoy | -0.011 |
rous | -0.011 |
Token | ; |
Token ID | 26 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
aces | +0.015 |
amas | +0.015 |
coat | +0.015 |
forehead | +0.014 |
drops | +0.014 |
Neg logit contributions | |
ours | -0.010 |
coconut | -0.010 |
Enjoy | -0.009 |
oral | -0.008 |
Sometimes | -0.008 |
Token | he |
Token ID | 339 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
ours | +0.075 |
coconut | +0.073 |
oral | +0.070 |
Enjoy | +0.069 |
rust | +0.067 |
Neg logit contributions | |
aces | -0.094 |
coat | -0.091 |
amas | -0.090 |
robe | -0.089 |
stand | -0.088 |
Token | has |
Token ID | 468 |
Feature activation | +7.69e-01 |
Pos logit contributions | |
aces | +0.010 |
coat | +0.010 |
amas | +0.010 |
robe | +0.010 |
forehead | +0.010 |
Neg logit contributions | |
ours | -0.007 |
coconut | -0.007 |
Enjoy | -0.007 |
oral | -0.007 |
rust | -0.006 |
Token | a |
Token ID | 257 |
Feature activation | -3.65e-01 |
Pos logit contributions | |
ours | +0.047 |
coconut | +0.046 |
Enjoy | +0.045 |
oral | +0.041 |
rust | +0.040 |
Neg logit contributions | |
aces | -0.066 |
amas | -0.066 |
coat | -0.065 |
forehead | -0.063 |
robe | -0.063 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
coat | +0.025 |
aces | +0.024 |
amas | +0.024 |
robe | +0.024 |
forehead | +0.024 |
Neg logit contributions | |
ours | -0.029 |
Enjoy | -0.028 |
coconut | -0.027 |
oral | -0.027 |
rust | -0.026 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.12e+01 |
Pos logit contributions | |
aces | +0.144 |
amas | +0.144 |
forehead | +0.140 |
stand | +0.131 |
bridge | +0.129 |
Neg logit contributions | |
ours | -0.155 |
coconut | -0.144 |
Enjoy | -0.144 |
crab | -0.138 |
shrimp | -0.135 |
Token | too |
Token ID | 1165 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
amas | +1.086 |
aces | +1.065 |
forehead | +1.035 |
bes | +1.004 |
drops | +1.001 |
Neg logit contributions | |
Enjoy | -0.549 |
ours | -0.519 |
coconut | -0.503 |
grown | -0.449 |
crab | -0.437 |
Token | ." |
Token ID | 526 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
ours | +0.099 |
coconut | +0.097 |
Enjoy | +0.094 |
oral | +0.090 |
rust | +0.088 |
Neg logit contributions | |
aces | -0.112 |
coat | -0.110 |
amas | -0.109 |
robe | -0.105 |
stand | -0.103 |
Token |
|
Token ID | 198 |
Feature activation | -1.70e-01 |
Pos logit contributions | |
stand | +0.124 |
amas | +0.118 |
aces | +0.116 |
coat | +0.112 |
forehead | +0.111 |
Neg logit contributions | |
ours | -0.128 |
coconut | -0.126 |
oral | -0.110 |
Enjoy | -0.109 |
rous | -0.108 |
Token |
|
Token ID | 198 |
Feature activation | -2.04e-01 |
Pos logit contributions | |
stand | +0.013 |
amas | +0.012 |
coat | +0.011 |
aces | +0.011 |
bes | +0.011 |
Neg logit contributions | |
ours | -0.013 |
coconut | -0.012 |
Enjoy | -0.011 |
seafood | -0.010 |
oy | -0.010 |
Token | Her |
Token ID | 9360 |
Feature activation | -3.21e-01 |
Pos logit contributions | |
stand | +0.013 |
coat | +0.013 |
aces | +0.012 |
amas | +0.012 |
bridge | +0.012 |
Neg logit contributions | |
ours | -0.018 |
coconut | -0.016 |
Enjoy | -0.016 |
volcano | -0.014 |
arium | -0.014 |
Token | loved |
Token ID | 6151 |
Feature activation | -9.64e-01 |
Pos logit contributions | |
ours | +0.012 |
coconut | +0.012 |
Enjoy | +0.012 |
oral | +0.011 |
rust | +0.010 |
Neg logit contributions | |
aces | -0.019 |
coat | -0.019 |
amas | -0.019 |
robe | -0.018 |
forehead | -0.018 |
Token | to |
Token ID | 284 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
aces | +0.075 |
coat | +0.075 |
amas | +0.074 |
stand | +0.072 |
drops | +0.071 |
Neg logit contributions | |
ours | -0.067 |
coconut | -0.063 |
oral | -0.063 |
Enjoy | -0.061 |
rust | -0.058 |
Token | play |
Token ID | 711 |
Feature activation | -3.23e-01 |
Pos logit contributions | |
ours | +0.007 |
coconut | +0.007 |
Enjoy | +0.007 |
oral | +0.006 |
rust | +0.006 |
Neg logit contributions | |
aces | -0.009 |
coat | -0.009 |
amas | -0.009 |
forehead | -0.008 |
robe | -0.008 |
Token | make |
Token ID | 787 |
Feature activation | -3.83e+00 |
Pos logit contributions | |
forehead | +0.029 |
aces | +0.028 |
amas | +0.028 |
drops | +0.028 |
coat | +0.027 |
Neg logit contributions | |
Enjoy | -0.018 |
ours | -0.017 |
swamp | -0.017 |
coconut | -0.016 |
whose | -0.016 |
Token | - |
Token ID | 12 |
Feature activation | -4.72e-01 |
Pos logit contributions | |
coat | +0.306 |
aces | +0.304 |
forehead | +0.301 |
amas | +0.298 |
robe | +0.291 |
Neg logit contributions | |
ours | -0.259 |
Enjoy | -0.238 |
coconut | -0.237 |
rust | -0.221 |
whose | -0.212 |
Token | bel |
Token ID | 6667 |
Feature activation | -1.12e+01 |
Pos logit contributions | |
amas | +0.031 |
aces | +0.031 |
forehead | +0.030 |
coat | +0.029 |
robe | +0.029 |
Neg logit contributions | |
ours | -0.039 |
coconut | -0.038 |
Enjoy | -0.037 |
rust | -0.036 |
oral | -0.035 |
Token | ieve |
Token ID | 12311 |
Feature activation | -2.07e+00 |
Pos logit contributions | |
coat | +0.892 |
robe | +0.840 |
drops | +0.839 |
amas | +0.817 |
forehead | +0.805 |
Neg logit contributions | |
ours | -0.814 |
Enjoy | -0.793 |
oral | -0.759 |
coconut | -0.725 |
rous | -0.715 |
Token | . |
Token ID | 13 |
Feature activation | -2.26e-01 |
Pos logit contributions | |
aces | +0.161 |
coat | +0.161 |
amas | +0.160 |
forehead | +0.157 |
drops | +0.152 |
Neg logit contributions | |
ours | -0.142 |
coconut | -0.138 |
Enjoy | -0.133 |
oral | -0.126 |
rust | -0.123 |
Token | Today |
Token ID | 6288 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
aces | +0.020 |
amas | +0.019 |
coat | +0.019 |
robe | +0.019 |
stand | +0.019 |
Neg logit contributions | |
coconut | -0.014 |
ours | -0.014 |
Enjoy | -0.013 |
oral | -0.012 |
rust | -0.011 |
Token | she |
Token ID | 673 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
ours | +0.089 |
coconut | +0.086 |
Enjoy | +0.084 |
oral | +0.078 |
rust | +0.074 |
Neg logit contributions | |
aces | -0.134 |
amas | -0.131 |
coat | -0.130 |
forehead | -0.126 |
robe | -0.125 |
Token | was |
Token ID | 373 |
Feature activation | +3.28e-01 |
Pos logit contributions | |
aces | +0.141 |
coat | +0.138 |
amas | +0.136 |
robe | +0.133 |
drops | +0.131 |
Neg logit contributions | |
ours | -0.100 |
coconut | -0.098 |
Enjoy | -0.095 |
oral | -0.089 |
rust | -0.084 |
Token | they |
Token ID | 484 |
Feature activation | +5.25e-01 |
Pos logit contributions | |
coat | +0.001 |
aces | +0.001 |
amas | +0.001 |
robe | +0.001 |
stand | +0.001 |
Neg logit contributions | |
ours | -0.001 |
coconut | -0.001 |
Enjoy | -0.001 |
oral | -0.001 |
rust | -0.001 |
Token | decided |
Token ID | 3066 |
Feature activation | +3.19e+00 |
Pos logit contributions | |
ours | +0.041 |
coconut | +0.040 |
Enjoy | +0.039 |
oral | +0.038 |
iw | +0.036 |
Neg logit contributions | |
coat | -0.035 |
aces | -0.035 |
robe | -0.035 |
cartridge | -0.033 |
amas | -0.032 |
Token | to |
Token ID | 284 |
Feature activation | +6.12e-01 |
Pos logit contributions | |
ours | +0.193 |
coconut | +0.185 |
Enjoy | +0.183 |
oral | +0.171 |
rust | +0.164 |
Neg logit contributions | |
aces | -0.280 |
coat | -0.278 |
amas | -0.273 |
forehead | -0.265 |
drops | -0.265 |
Token | be |
Token ID | 307 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
ours | +0.045 |
coconut | +0.044 |
Enjoy | +0.042 |
oral | +0.041 |
iw | +0.040 |
Neg logit contributions | |
aces | -0.045 |
coat | -0.045 |
amas | -0.044 |
robe | -0.043 |
stand | -0.043 |
Token | mail |
Token ID | 6920 |
Feature activation | -3.23e-01 |
Pos logit contributions | |
amas | +0.115 |
aces | +0.110 |
forehead | +0.106 |
coat | +0.104 |
stand | +0.103 |
Neg logit contributions | |
Enjoy | -0.069 |
ours | -0.066 |
coconut | -0.061 |
grown | -0.060 |
rous | -0.058 |
Token | carriers |
Token ID | 16651 |
Feature activation | -1.12e+01 |
Pos logit contributions | |
aces | +0.029 |
amas | +0.029 |
forehead | +0.029 |
coat | +0.028 |
drops | +0.028 |
Neg logit contributions | |
ours | -0.018 |
Enjoy | -0.018 |
coconut | -0.017 |
whose | -0.016 |
Sometimes | -0.015 |
Token | and |
Token ID | 290 |
Feature activation | +5.25e-02 |
Pos logit contributions | |
aces | +0.982 |
amas | +0.976 |
coat | +0.962 |
forehead | +0.944 |
drops | +0.941 |
Neg logit contributions | |
ours | -0.669 |
Enjoy | -0.652 |
coconut | -0.638 |
oral | -0.585 |
rust | -0.568 |
Token | deliver |
Token ID | 5203 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
ours | +0.004 |
coconut | +0.004 |
Enjoy | +0.004 |
oral | +0.004 |
rust | +0.004 |
Neg logit contributions | |
aces | -0.004 |
amas | -0.004 |
forehead | -0.004 |
coat | -0.004 |
stand | -0.003 |
Token | letters |
Token ID | 7475 |
Feature activation | -1.80e+00 |
Pos logit contributions | |
aces | +0.075 |
amas | +0.072 |
coat | +0.070 |
forehead | +0.068 |
robe | +0.068 |
Neg logit contributions | |
ours | -0.085 |
coconut | -0.082 |
Enjoy | -0.080 |
oral | -0.078 |
rust | -0.076 |
Token | to |
Token ID | 284 |
Feature activation | +6.24e-01 |
Pos logit contributions | |
amas | +0.195 |
coat | +0.190 |
forehead | +0.189 |
aces | +0.183 |
drops | +0.183 |
Neg logit contributions | |
ours | -0.069 |
Enjoy | -0.065 |
coconut | -0.065 |
whose | -0.056 |
Sometimes | -0.051 |
Token | their |
Token ID | 511 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
ours | +0.039 |
coconut | +0.037 |
Enjoy | +0.036 |
oral | +0.034 |
rust | +0.033 |
Neg logit contributions | |
aces | -0.053 |
amas | -0.052 |
coat | -0.051 |
forehead | -0.050 |
stand | -0.050 |
Token | . |
Token ID | 13 |
Feature activation | +2.06e-01 |
Pos logit contributions | |
aces | +0.134 |
amas | +0.131 |
coat | +0.130 |
drops | +0.127 |
robe | +0.126 |
Neg logit contributions | |
ours | -0.078 |
Enjoy | -0.078 |
coconut | -0.074 |
oral | -0.068 |
asks | -0.064 |
Token | We |
Token ID | 775 |
Feature activation | +9.14e-03 |
Pos logit contributions | |
ours | +0.012 |
Enjoy | +0.012 |
coconut | +0.011 |
oral | +0.010 |
Sometimes | +0.010 |
Neg logit contributions | |
aces | -0.019 |
coat | -0.019 |
amas | -0.019 |
drops | -0.018 |
forehead | -0.018 |
Token | love |
Token ID | 1842 |
Feature activation | -2.21e-01 |
Pos logit contributions | |
Enjoy | +0.000 |
ours | +0.000 |
coconut | +0.000 |
oral | +0.000 |
rust | +0.000 |
Neg logit contributions | |
amas | -0.001 |
aces | -0.001 |
stand | -0.001 |
coat | -0.001 |
robe | -0.001 |
Token | being |
Token ID | 852 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
amas | +0.018 |
aces | +0.018 |
stand | +0.017 |
coat | +0.017 |
forehead | +0.017 |
Neg logit contributions | |
ours | -0.015 |
coconut | -0.014 |
Enjoy | -0.013 |
oral | -0.013 |
rust | -0.013 |
Token | mail |
Token ID | 6920 |
Feature activation | -3.12e-01 |
Pos logit contributions | |
Enjoy | +0.007 |
whose | +0.006 |
ours | +0.006 |
shrimp | +0.006 |
coconut | +0.006 |
Neg logit contributions | |
drops | -0.016 |
coat | -0.016 |
amas | -0.016 |
aces | -0.016 |
forehead | -0.015 |
Token | carriers |
Token ID | 16651 |
Feature activation | -1.12e+01 |
Pos logit contributions | |
coat | +0.030 |
amas | +0.030 |
aces | +0.029 |
forehead | +0.029 |
drops | +0.029 |
Neg logit contributions | |
Enjoy | -0.014 |
ours | -0.013 |
coconut | -0.013 |
oral | -0.012 |
Crab | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | +2.13e-01 |
Pos logit contributions | |
amas | +0.994 |
coat | +0.985 |
aces | +0.983 |
drops | +0.951 |
forehead | +0.946 |
Neg logit contributions | |
ours | -0.639 |
Enjoy | -0.628 |
coconut | -0.614 |
oral | -0.561 |
rust | -0.538 |
Token | Can |
Token ID | 1680 |
Feature activation | +2.03e+00 |
Pos logit contributions | |
ours | +0.013 |
coconut | +0.012 |
Enjoy | +0.012 |
oral | +0.011 |
rust | +0.010 |
Neg logit contributions | |
aces | -0.019 |
coat | -0.019 |
amas | -0.019 |
robe | -0.018 |
forehead | -0.018 |
Token | we |
Token ID | 356 |
Feature activation | +2.78e-01 |
Pos logit contributions | |
ours | +0.125 |
coconut | +0.122 |
Enjoy | +0.118 |
oral | +0.109 |
rust | +0.106 |
Neg logit contributions | |
aces | -0.177 |
amas | -0.174 |
coat | -0.173 |
forehead | -0.167 |
robe | -0.167 |
Token | do |
Token ID | 466 |
Feature activation | +9.15e-01 |
Pos logit contributions | |
ours | +0.018 |
Enjoy | +0.018 |
coconut | +0.018 |
oral | +0.017 |
rust | +0.016 |
Neg logit contributions | |
aces | -0.023 |
amas | -0.022 |
coat | -0.022 |
robe | -0.021 |
drops | -0.021 |
Token | it |
Token ID | 340 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
coconut | +0.059 |
ours | +0.059 |
Enjoy | +0.054 |
iw | +0.053 |
rust | +0.053 |
Neg logit contributions | |
aces | -0.074 |
amas | -0.074 |
forehead | -0.072 |
stand | -0.072 |
coat | -0.070 |
Token | they |
Token ID | 484 |
Feature activation | +2.19e-01 |
Pos logit contributions | |
aces | +0.043 |
amas | +0.042 |
coat | +0.042 |
robe | +0.040 |
forehead | +0.040 |
Neg logit contributions | |
ours | -0.025 |
coconut | -0.025 |
Enjoy | -0.025 |
oral | -0.021 |
rust | -0.021 |
Token | always |
Token ID | 1464 |
Feature activation | -9.77e-01 |
Pos logit contributions | |
ours | +0.014 |
coconut | +0.014 |
Enjoy | +0.014 |
oral | +0.013 |
rust | +0.012 |
Neg logit contributions | |
aces | -0.018 |
amas | -0.018 |
coat | -0.018 |
forehead | -0.017 |
robe | -0.017 |
Token | threw |
Token ID | 9617 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
aces | +0.083 |
amas | +0.082 |
coat | +0.081 |
robe | +0.078 |
drops | +0.077 |
Neg logit contributions | |
ours | -0.062 |
coconut | -0.060 |
Enjoy | -0.059 |
oral | -0.054 |
rust | -0.053 |
Token | the |
Token ID | 262 |
Feature activation | +2.74e-02 |
Pos logit contributions | |
ours | +0.014 |
coconut | +0.013 |
Enjoy | +0.013 |
oral | +0.012 |
rust | +0.012 |
Neg logit contributions | |
aces | -0.020 |
amas | -0.020 |
coat | -0.020 |
robe | -0.019 |
forehead | -0.019 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.15e+00 |
Pos logit contributions | |
coconut | +0.002 |
ours | +0.002 |
Enjoy | +0.002 |
oral | +0.002 |
rust | +0.002 |
Neg logit contributions | |
aces | -0.002 |
amas | -0.002 |
coat | -0.002 |
robe | -0.002 |
drops | -0.002 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.12e+01 |
Pos logit contributions | |
amas | +0.158 |
forehead | +0.152 |
aces | +0.149 |
bridge | +0.145 |
stand | +0.139 |
Neg logit contributions | |
ours | -0.147 |
coconut | -0.135 |
crab | -0.132 |
shrimp | -0.129 |
Enjoy | -0.127 |
Token | in |
Token ID | 287 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
amas | +1.030 |
aces | +0.996 |
forehead | +0.963 |
bes | +0.951 |
bridge | +0.935 |
Neg logit contributions | |
ours | -0.570 |
Enjoy | -0.560 |
coconut | -0.551 |
arium | -0.510 |
fired | -0.479 |
Token | the |
Token ID | 262 |
Feature activation | -2.20e-01 |
Pos logit contributions | |
ours | +0.056 |
coconut | +0.054 |
Enjoy | +0.053 |
oral | +0.048 |
rust | +0.046 |
Neg logit contributions | |
aces | -0.098 |
amas | -0.096 |
coat | -0.096 |
robe | -0.092 |
forehead | -0.092 |
Token | air |
Token ID | 1633 |
Feature activation | -3.38e+00 |
Pos logit contributions | |
aces | +0.017 |
amas | +0.016 |
coat | +0.016 |
forehead | +0.015 |
robe | +0.015 |
Neg logit contributions | |
ours | -0.016 |
coconut | -0.016 |
Enjoy | -0.015 |
oral | -0.014 |
rust | -0.014 |
Token | when |
Token ID | 618 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
amas | +0.323 |
aces | +0.322 |
forehead | +0.307 |
coat | +0.302 |
robe | +0.300 |
Neg logit contributions | |
coconut | -0.164 |
ours | -0.164 |
Enjoy | -0.149 |
arium | -0.136 |
crab | -0.131 |
Token | they |
Token ID | 484 |
Feature activation | +8.33e-02 |
Pos logit contributions | |
amas | +0.179 |
aces | +0.178 |
coat | +0.176 |
robe | +0.173 |
stand | +0.173 |
Neg logit contributions | |
ours | -0.127 |
coconut | -0.122 |
Enjoy | -0.119 |
oral | -0.112 |
rust | -0.111 |
Token | put |
Token ID | 1234 |
Feature activation | -1.73e+00 |
Pos logit contributions | |
aces | +0.099 |
amas | +0.096 |
coat | +0.094 |
drops | +0.091 |
robe | +0.091 |
Neg logit contributions | |
ours | -0.075 |
Enjoy | -0.075 |
coconut | -0.073 |
oral | -0.067 |
rust | -0.064 |
Token | on |
Token ID | 319 |
Feature activation | +6.78e-01 |
Pos logit contributions | |
amas | +0.175 |
aces | +0.172 |
drops | +0.172 |
coat | +0.169 |
forehead | +0.166 |
Neg logit contributions | |
ours | -0.082 |
Enjoy | -0.079 |
coconut | -0.067 |
rous | -0.065 |
Sometimes | -0.064 |
Token | her |
Token ID | 607 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
ours | +0.039 |
Enjoy | +0.037 |
coconut | +0.036 |
rous | +0.033 |
rust | +0.032 |
Neg logit contributions | |
aces | -0.061 |
amas | -0.061 |
coat | -0.059 |
drops | -0.059 |
bridge | -0.058 |
Token | lovely |
Token ID | 14081 |
Feature activation | -7.17e-01 |
Pos logit contributions | |
aces | +0.146 |
drops | +0.141 |
bridge | +0.139 |
amas | +0.139 |
coat | +0.138 |
Neg logit contributions | |
ours | -0.071 |
Enjoy | -0.064 |
coconut | -0.061 |
oral | -0.061 |
Sometimes | -0.060 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.15e+00 |
Pos logit contributions | |
aces | +0.064 |
drops | +0.061 |
bridge | +0.059 |
amas | +0.059 |
coat | +0.056 |
Neg logit contributions | |
ours | -0.043 |
Enjoy | -0.040 |
coconut | -0.040 |
shrimp | -0.038 |
oral | -0.038 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.12e+01 |
Pos logit contributions | |
bridge | +0.071 |
aces | +0.067 |
forehead | +0.067 |
drops | +0.065 |
ring | +0.061 |
Neg logit contributions | |
ours | -0.223 |
Enjoy | -0.210 |
arium | -0.207 |
ry | -0.206 |
Sometimes | -0.206 |
Token | and |
Token ID | 290 |
Feature activation | +8.17e-02 |
Pos logit contributions | |
amas | +1.119 |
drops | +1.076 |
forehead | +1.074 |
aces | +1.071 |
bridge | +1.068 |
Neg logit contributions | |
Enjoy | -0.431 |
arium | -0.412 |
ours | -0.406 |
: | -0.394 |
coconut | -0.368 |
Token | feel |
Token ID | 1254 |
Feature activation | -1.85e+00 |
Pos logit contributions | |
ours | +0.005 |
Enjoy | +0.005 |
coconut | +0.005 |
asks | +0.004 |
Sometimes | +0.004 |
Neg logit contributions | |
amas | -0.007 |
aces | -0.007 |
coat | -0.007 |
bridge | -0.006 |
robe | -0.006 |
Token | so |
Token ID | 523 |
Feature activation | -2.52e+00 |
Pos logit contributions | |
amas | +0.179 |
aces | +0.175 |
coat | +0.172 |
drops | +0.164 |
robe | +0.156 |
Neg logit contributions | |
Enjoy | -0.112 |
ours | -0.101 |
coconut | -0.092 |
rous | -0.092 |
Sometimes | -0.090 |
Token | fit |
Token ID | 4197 |
Feature activation | -4.71e+00 |
Pos logit contributions | |
aces | +0.214 |
coat | +0.212 |
amas | +0.212 |
drops | +0.198 |
robe | +0.194 |
Neg logit contributions | |
ours | -0.166 |
Enjoy | -0.165 |
coconut | -0.154 |
oral | -0.140 |
rous | -0.139 |
Token | and |
Token ID | 290 |
Feature activation | +6.99e-02 |
Pos logit contributions | |
amas | +0.479 |
coat | +0.442 |
aces | +0.441 |
forehead | +0.439 |
drops | +0.432 |
Neg logit contributions | |
ours | -0.202 |
Enjoy | -0.196 |
grown | -0.194 |
what | -0.179 |
fired | -0.167 |
Token | gave |
Token ID | 2921 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
aces | +0.048 |
amas | +0.047 |
coat | +0.047 |
forehead | +0.045 |
robe | +0.045 |
Neg logit contributions | |
ours | -0.027 |
coconut | -0.025 |
Enjoy | -0.025 |
oral | -0.023 |
rust | -0.022 |
Token | him |
Token ID | 683 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
aces | +0.112 |
coat | +0.109 |
amas | +0.109 |
robe | +0.105 |
forehead | +0.104 |
Neg logit contributions | |
ours | -0.079 |
coconut | -0.076 |
Enjoy | -0.074 |
oral | -0.071 |
rust | -0.066 |
Token | an |
Token ID | 281 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
aces | +0.129 |
coat | +0.124 |
robe | +0.120 |
amas | +0.119 |
forehead | +0.117 |
Neg logit contributions | |
ours | -0.134 |
coconut | -0.129 |
Enjoy | -0.125 |
oral | -0.124 |
rust | -0.117 |
Token | ancient |
Token ID | 6156 |
Feature activation | +4.04e-01 |
Pos logit contributions | |
ours | +0.080 |
coconut | +0.078 |
Enjoy | +0.077 |
oral | +0.072 |
rust | +0.070 |
Neg logit contributions | |
aces | -0.076 |
amas | -0.075 |
coat | -0.074 |
drops | -0.072 |
robe | -0.070 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
ours | +0.034 |
coconut | +0.033 |
Enjoy | +0.033 |
oral | +0.031 |
rust | +0.030 |
Neg logit contributions | |
aces | -0.026 |
amas | -0.025 |
coat | -0.025 |
robe | -0.024 |
forehead | -0.024 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
aces | +0.068 |
amas | +0.066 |
forehead | +0.064 |
bridge | +0.059 |
drops | +0.057 |
Neg logit contributions | |
ours | -0.231 |
Enjoy | -0.221 |
arose | -0.217 |
oral | -0.213 |
coconut | -0.212 |
Token | that |
Token ID | 326 |
Feature activation | -9.76e-01 |
Pos logit contributions | |
amas | +1.164 |
aces | +1.125 |
forehead | +1.117 |
drops | +1.116 |
coat | +1.085 |
Neg logit contributions | |
Enjoy | -0.472 |
ours | -0.431 |
coconut | -0.396 |
: | -0.361 |
arium | -0.352 |
Token | belonged |
Token ID | 19611 |
Feature activation | -2.02e+00 |
Pos logit contributions | |
aces | +0.093 |
amas | +0.093 |
coat | +0.091 |
forehead | +0.090 |
drops | +0.087 |
Neg logit contributions | |
ours | -0.051 |
coconut | -0.050 |
Enjoy | -0.048 |
rust | -0.043 |
Crab | -0.041 |
Token | to |
Token ID | 284 |
Feature activation | +4.88e-01 |
Pos logit contributions | |
amas | +0.223 |
forehead | +0.221 |
drops | +0.218 |
aces | +0.217 |
bridge | +0.212 |
Neg logit contributions | |
ours | -0.069 |
Enjoy | -0.068 |
coconut | -0.065 |
grown | -0.060 |
rust | -0.054 |
Token | his |
Token ID | 465 |
Feature activation | +9.61e-01 |
Pos logit contributions | |
ours | +0.028 |
coconut | +0.027 |
Enjoy | +0.026 |
oral | +0.024 |
rust | +0.023 |
Neg logit contributions | |
aces | -0.045 |
amas | -0.044 |
coat | -0.044 |
robe | -0.042 |
forehead | -0.042 |
Token | grand |
Token ID | 4490 |
Feature activation | -2.58e+00 |
Pos logit contributions | |
ours | +0.071 |
coconut | +0.067 |
Enjoy | +0.067 |
oral | +0.061 |
rust | +0.060 |
Neg logit contributions | |
coat | -0.072 |
robe | -0.071 |
amas | -0.069 |
forehead | -0.068 |
aces | -0.067 |
Token | she |
Token ID | 673 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
amas | +0.074 |
aces | +0.072 |
forehead | +0.071 |
coat | +0.071 |
robe | +0.069 |
Neg logit contributions | |
ours | -0.049 |
Enjoy | -0.045 |
coconut | -0.044 |
grown | -0.042 |
oral | -0.041 |
Token | put |
Token ID | 1234 |
Feature activation | -1.65e+00 |
Pos logit contributions | |
aces | +0.125 |
coat | +0.123 |
amas | +0.121 |
robe | +0.117 |
drops | +0.117 |
Neg logit contributions | |
ours | -0.104 |
coconut | -0.102 |
Enjoy | -0.098 |
oral | -0.093 |
rust | -0.089 |
Token | her |
Token ID | 607 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
amas | +0.149 |
aces | +0.145 |
drops | +0.143 |
forehead | +0.143 |
coat | +0.140 |
Neg logit contributions | |
ours | -0.093 |
Enjoy | -0.092 |
coconut | -0.082 |
rous | -0.077 |
iw | -0.077 |
Token | new |
Token ID | 649 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
amas | +0.116 |
aces | +0.115 |
bridge | +0.112 |
drops | +0.111 |
coat | +0.109 |
Neg logit contributions | |
ours | -0.094 |
Enjoy | -0.088 |
coconut | -0.087 |
shrimp | -0.079 |
crab | -0.078 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
ours | +0.125 |
coconut | +0.119 |
Enjoy | +0.119 |
oral | +0.108 |
rust | +0.107 |
Neg logit contributions | |
aces | -0.140 |
amas | -0.136 |
drops | -0.131 |
bridge | -0.128 |
coat | -0.128 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
amas | +0.079 |
forehead | +0.077 |
bridge | +0.068 |
stand | +0.068 |
click | +0.067 |
Neg logit contributions | |
ours | -0.204 |
Enjoy | -0.198 |
arium | -0.191 |
swamp | -0.187 |
in | -0.187 |
Token | on |
Token ID | 319 |
Feature activation | +7.11e-01 |
Pos logit contributions | |
amas | +1.073 |
drops | +1.003 |
forehead | +0.996 |
aces | +0.990 |
bes | +0.965 |
Neg logit contributions | |
Enjoy | -0.517 |
arium | -0.462 |
ours | -0.452 |
: | -0.440 |
coconut | -0.437 |
Token | . |
Token ID | 13 |
Feature activation | -1.70e-01 |
Pos logit contributions | |
ours | +0.045 |
coconut | +0.044 |
Enjoy | +0.043 |
oral | +0.040 |
rust | +0.038 |
Neg logit contributions | |
aces | -0.060 |
amas | -0.059 |
coat | -0.058 |
forehead | -0.057 |
drops | -0.056 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +4.85e-01 |
Pos logit contributions | |
aces | +0.016 |
amas | +0.015 |
robe | +0.015 |
coat | +0.015 |
stand | +0.015 |
Neg logit contributions | |
ours | -0.010 |
coconut | -0.010 |
Enjoy | -0.008 |
oral | -0.008 |
rust | -0.008 |
Token | Once |
Token ID | 7454 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
ours | +0.027 |
coconut | +0.026 |
Enjoy | +0.023 |
rust | +0.022 |
oral | +0.022 |
Neg logit contributions | |
aces | -0.046 |
amas | -0.046 |
robe | -0.046 |
coat | -0.045 |
stand | -0.044 |
Token | upon |
Token ID | 2402 |
Feature activation | +1.70e+00 |
Pos logit contributions | |
aces | +0.185 |
coat | +0.184 |
robe | +0.184 |
amas | +0.179 |
stand | +0.178 |
Neg logit contributions | |
Enjoy | -0.122 |
coconut | -0.120 |
ours | -0.112 |
rust | -0.103 |
iw | -0.099 |
Token | You |
Token ID | 1639 |
Feature activation | -7.17e-01 |
Pos logit contributions | |
aces | +0.122 |
amas | +0.119 |
coat | +0.118 |
forehead | +0.116 |
robe | +0.113 |
Neg logit contributions | |
Enjoy | -0.079 |
ours | -0.075 |
oral | -0.075 |
coconut | -0.072 |
rust | -0.071 |
Token | are |
Token ID | 389 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
amas | +0.068 |
coat | +0.067 |
aces | +0.066 |
drops | +0.064 |
stand | +0.063 |
Neg logit contributions | |
Enjoy | -0.041 |
coconut | -0.038 |
ours | -0.037 |
asks | -0.033 |
oral | -0.033 |
Token | very |
Token ID | 845 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
coat | +0.153 |
aces | +0.151 |
amas | +0.150 |
drops | +0.146 |
forehead | +0.141 |
Neg logit contributions | |
Enjoy | -0.084 |
ours | -0.074 |
coconut | -0.074 |
Crab | -0.065 |
rust | -0.064 |
Token | good |
Token ID | 922 |
Feature activation | +1.47e+00 |
Pos logit contributions | |
ours | +0.148 |
Enjoy | +0.144 |
coconut | +0.143 |
oral | +0.130 |
rust | +0.129 |
Neg logit contributions | |
aces | -0.165 |
amas | -0.161 |
coat | -0.160 |
robe | -0.152 |
forehead | -0.152 |
Token | mail |
Token ID | 6920 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
ours | +0.088 |
Enjoy | +0.083 |
coconut | +0.082 |
oral | +0.076 |
rust | +0.072 |
Neg logit contributions | |
aces | -0.132 |
coat | -0.129 |
amas | -0.129 |
robe | -0.125 |
forehead | -0.124 |
Token | carriers |
Token ID | 16651 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
aces | +0.028 |
amas | +0.028 |
coat | +0.027 |
forehead | +0.027 |
drops | +0.026 |
Neg logit contributions | |
Enjoy | -0.011 |
ours | -0.010 |
coconut | -0.009 |
whose | -0.009 |
Joe | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
aces | +0.984 |
amas | +0.980 |
coat | +0.964 |
forehead | +0.938 |
drops | +0.936 |
Neg logit contributions | |
ours | -0.655 |
Enjoy | -0.641 |
coconut | -0.633 |
oral | -0.574 |
rust | -0.550 |
Token | You |
Token ID | 921 |
Feature activation | -4.00e-01 |
Pos logit contributions | |
ours | +0.013 |
coconut | +0.013 |
Enjoy | +0.012 |
oral | +0.011 |
rust | +0.011 |
Neg logit contributions | |
aces | -0.014 |
amas | -0.013 |
coat | -0.013 |
robe | -0.013 |
forehead | -0.012 |
Token | made |
Token ID | 925 |
Feature activation | -4.25e+00 |
Pos logit contributions | |
amas | +0.039 |
coat | +0.038 |
aces | +0.037 |
stand | +0.036 |
forehead | +0.036 |
Neg logit contributions | |
Enjoy | -0.022 |
coconut | -0.019 |
ours | -0.019 |
asks | -0.018 |
oral | -0.017 |
Token | your |
Token ID | 534 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
aces | +0.396 |
coat | +0.391 |
amas | +0.384 |
robe | +0.375 |
drops | +0.374 |
Neg logit contributions | |
ours | -0.241 |
coconut | -0.233 |
Enjoy | -0.230 |
oral | -0.207 |
rust | -0.202 |
Token | friends |
Token ID | 2460 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
ours | +0.083 |
coconut | +0.081 |
Enjoy | +0.079 |
oral | +0.073 |
rust | +0.070 |
Neg logit contributions | |
aces | -0.117 |
amas | -0.115 |
coat | -0.114 |
robe | -0.110 |
drops | -0.110 |
Token | ! |
Token ID | 0 |
Feature activation | -7.21e-02 |
Pos logit contributions | |
ours | +0.021 |
Enjoy | +0.020 |
coconut | +0.019 |
Joe | +0.017 |
grown | +0.016 |
Neg logit contributions | |
aces | -0.043 |
amas | -0.042 |
forehead | -0.042 |
drops | -0.041 |
coat | -0.041 |
Token | I |
Token ID | 314 |
Feature activation | -6.63e-01 |
Pos logit contributions | |
aces | +0.006 |
coat | +0.006 |
robe | +0.006 |
amas | +0.005 |
stand | +0.005 |
Neg logit contributions | |
coconut | -0.005 |
ours | -0.005 |
oral | -0.005 |
Enjoy | -0.005 |
rust | -0.005 |
Token | found |
Token ID | 1043 |
Feature activation | -2.39e+00 |
Pos logit contributions | |
coat | +0.044 |
robe | +0.043 |
aces | +0.041 |
cartridge | +0.040 |
drops | +0.039 |
Neg logit contributions | |
ours | -0.054 |
oral | -0.052 |
coconut | -0.052 |
Enjoy | -0.048 |
rust | -0.045 |
Token | a |
Token ID | 257 |
Feature activation | -5.36e-01 |
Pos logit contributions | |
aces | +0.193 |
amas | +0.187 |
coat | +0.184 |
robe | +0.178 |
forehead | +0.178 |
Neg logit contributions | |
ours | -0.165 |
coconut | -0.159 |
Enjoy | -0.155 |
oral | -0.150 |
rust | -0.144 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.24e+00 |
Pos logit contributions | |
coat | +0.037 |
robe | +0.035 |
amas | +0.034 |
forehead | +0.034 |
stand | +0.033 |
Neg logit contributions | |
ours | -0.045 |
Enjoy | -0.041 |
oral | -0.040 |
coconut | -0.040 |
rust | -0.039 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
aces | +0.162 |
forehead | +0.151 |
amas | +0.151 |
stand | +0.145 |
bridge | +0.144 |
Neg logit contributions | |
ours | -0.166 |
Enjoy | -0.158 |
coconut | -0.157 |
swamp | -0.145 |
crab | -0.145 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
aces | +1.056 |
amas | +1.043 |
forehead | +1.008 |
bes | +0.996 |
drops | +0.979 |
Neg logit contributions | |
Enjoy | -0.594 |
ours | -0.556 |
coconut | -0.528 |
iw | -0.467 |
grown | -0.459 |
Token | Lily |
Token ID | 20037 |
Feature activation | +4.50e-01 |
Pos logit contributions | |
aces | +0.171 |
coat | +0.166 |
amas | +0.165 |
drops | +0.162 |
forehead | +0.160 |
Neg logit contributions | |
ours | -0.070 |
Enjoy | -0.067 |
coconut | -0.065 |
oral | -0.056 |
Sometimes | -0.056 |
Token | looked |
Token ID | 3114 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
ours | +0.028 |
coconut | +0.027 |
Enjoy | +0.027 |
oral | +0.025 |
rust | +0.024 |
Neg logit contributions | |
aces | -0.039 |
amas | -0.038 |
coat | -0.038 |
forehead | -0.037 |
robe | -0.036 |
Token | at |
Token ID | 379 |
Feature activation | +2.76e+00 |
Pos logit contributions | |
coconut | +0.129 |
oral | +0.128 |
ours | +0.128 |
Enjoy | +0.128 |
rust | +0.114 |
Neg logit contributions | |
aces | -0.122 |
coat | -0.121 |
forehead | -0.118 |
amas | -0.117 |
drops | -0.114 |
Token | the |
Token ID | 262 |
Feature activation | -7.15e-02 |
Pos logit contributions | |
ours | +0.158 |
coconut | +0.153 |
Enjoy | +0.149 |
oral | +0.136 |
rust | +0.133 |
Neg logit contributions | |
aces | -0.254 |
amas | -0.249 |
coat | -0.248 |
robe | -0.239 |
forehead | -0.239 |
Token | put |
Token ID | 1234 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
ours | +0.029 |
coconut | +0.028 |
Enjoy | +0.027 |
oral | +0.025 |
rust | +0.025 |
Neg logit contributions | |
aces | -0.039 |
coat | -0.038 |
amas | -0.038 |
robe | -0.037 |
forehead | -0.037 |
Token | on |
Token ID | 319 |
Feature activation | +7.82e-01 |
Pos logit contributions | |
amas | +0.157 |
aces | +0.155 |
coat | +0.153 |
drops | +0.152 |
forehead | +0.151 |
Neg logit contributions | |
ours | -0.079 |
Enjoy | -0.075 |
coconut | -0.072 |
iw | -0.063 |
rous | -0.063 |
Token | her |
Token ID | 607 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
ours | +0.048 |
coconut | +0.046 |
Enjoy | +0.045 |
oral | +0.042 |
rust | +0.041 |
Neg logit contributions | |
aces | -0.068 |
amas | -0.067 |
coat | -0.067 |
robe | -0.064 |
forehead | -0.064 |
Token | cozy |
Token ID | 37438 |
Feature activation | -3.11e+00 |
Pos logit contributions | |
aces | +0.106 |
drops | +0.102 |
amas | +0.100 |
coat | +0.099 |
bridge | +0.098 |
Neg logit contributions | |
ours | -0.091 |
coconut | -0.088 |
Enjoy | -0.086 |
crab | -0.082 |
shrimp | -0.081 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.10e+00 |
Pos logit contributions | |
aces | +0.210 |
drops | +0.201 |
bridge | +0.191 |
amas | +0.183 |
coat | +0.178 |
Neg logit contributions | |
coconut | -0.247 |
ours | -0.239 |
swamp | -0.231 |
Enjoy | -0.230 |
crab | -0.225 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
drops | +0.067 |
click | +0.067 |
bridge | +0.066 |
hy | +0.066 |
forehead | +0.065 |
Neg logit contributions | |
ours | -0.212 |
arium | -0.203 |
ry | -0.197 |
Enjoy | -0.196 |
Sometimes | -0.195 |
Token | and |
Token ID | 290 |
Feature activation | -7.70e-03 |
Pos logit contributions | |
drops | +1.090 |
forehead | +1.081 |
aces | +1.080 |
amas | +1.079 |
bridge | +1.063 |
Neg logit contributions | |
arium | -0.422 |
ours | -0.417 |
Enjoy | -0.390 |
: | -0.366 |
Sometimes | -0.363 |
Token | went |
Token ID | 1816 |
Feature activation | -1.86e-01 |
Pos logit contributions | |
aces | +0.001 |
amas | +0.001 |
coat | +0.001 |
bridge | +0.001 |
drops | +0.001 |
Neg logit contributions | |
ours | -0.000 |
Enjoy | -0.000 |
coconut | -0.000 |
asks | -0.000 |
rust | -0.000 |
Token | to |
Token ID | 284 |
Feature activation | +5.05e-01 |
Pos logit contributions | |
aces | +0.019 |
amas | +0.018 |
coat | +0.018 |
drops | +0.017 |
forehead | +0.017 |
Neg logit contributions | |
ours | -0.009 |
Enjoy | -0.009 |
coconut | -0.008 |
oral | -0.007 |
Sometimes | -0.007 |
Token | bed |
Token ID | 3996 |
Feature activation | -2.92e+00 |
Pos logit contributions | |
coconut | +0.051 |
ours | +0.051 |
oral | +0.049 |
rust | +0.047 |
iw | +0.047 |
Neg logit contributions | |
stand | -0.024 |
forehead | -0.024 |
amas | -0.023 |
robe | -0.022 |
coat | -0.022 |
Token | with |
Token ID | 351 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
aces | +0.294 |
coat | +0.276 |
drops | +0.274 |
bridge | +0.272 |
pads | +0.262 |
Neg logit contributions | |
ours | -0.141 |
Enjoy | -0.124 |
arium | -0.117 |
oral | -0.115 |
whose | -0.110 |
Token | she |
Token ID | 673 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
aces | +0.128 |
amas | +0.124 |
coat | +0.124 |
robe | +0.120 |
forehead | +0.119 |
Neg logit contributions | |
ours | -0.067 |
coconut | -0.064 |
Enjoy | -0.064 |
oral | -0.056 |
grown | -0.053 |
Token | had |
Token ID | 550 |
Feature activation | -5.28e-01 |
Pos logit contributions | |
aces | +0.156 |
amas | +0.154 |
coat | +0.147 |
bridge | +0.141 |
robe | +0.141 |
Neg logit contributions | |
Enjoy | -0.062 |
asks | -0.055 |
ours | -0.053 |
crab | -0.050 |
lly | -0.050 |
Token | a |
Token ID | 257 |
Feature activation | -5.85e-01 |
Pos logit contributions | |
amas | +0.053 |
drops | +0.053 |
forehead | +0.051 |
coat | +0.050 |
aces | +0.050 |
Neg logit contributions | |
Enjoy | -0.026 |
coconut | -0.023 |
grown | -0.023 |
rous | -0.022 |
ours | -0.021 |
Token | charming |
Token ID | 23332 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
coat | +0.044 |
aces | +0.043 |
amas | +0.043 |
forehead | +0.042 |
robe | +0.042 |
Neg logit contributions | |
ours | -0.043 |
coconut | -0.040 |
Enjoy | -0.039 |
oral | -0.037 |
rust | -0.036 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
aces | +0.120 |
amas | +0.117 |
coat | +0.117 |
robe | +0.112 |
forehead | +0.112 |
Neg logit contributions | |
ours | -0.095 |
coconut | -0.092 |
Enjoy | -0.090 |
oral | -0.084 |
rust | -0.081 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
amas | +0.190 |
aces | +0.187 |
forehead | +0.180 |
stand | +0.174 |
bridge | +0.174 |
Neg logit contributions | |
Enjoy | -0.106 |
ours | -0.101 |
coconut | -0.096 |
arium | -0.096 |
crab | -0.093 |
Token | to |
Token ID | 284 |
Feature activation | +3.63e-01 |
Pos logit contributions | |
amas | +1.229 |
aces | +1.188 |
bes | +1.160 |
forehead | +1.131 |
drops | +1.094 |
Neg logit contributions | |
Enjoy | -0.419 |
: | -0.337 |
arium | -0.316 |
coconut | -0.315 |
grown | -0.314 |
Token | wear |
Token ID | 5806 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
Enjoy | +0.022 |
ours | +0.021 |
coconut | +0.021 |
oral | +0.019 |
rust | +0.018 |
Neg logit contributions | |
aces | -0.032 |
amas | -0.032 |
coat | -0.031 |
drops | -0.030 |
robe | -0.029 |
Token | , |
Token ID | 11 |
Feature activation | -1.72e-01 |
Pos logit contributions | |
amas | +0.104 |
drops | +0.101 |
aces | +0.101 |
forehead | +0.100 |
coat | +0.097 |
Neg logit contributions | |
Enjoy | -0.042 |
what | -0.038 |
grown | -0.037 |
: | -0.036 |
ours | -0.036 |
Token | and |
Token ID | 290 |
Feature activation | -5.25e-03 |
Pos logit contributions | |
aces | +0.016 |
amas | +0.015 |
coat | +0.014 |
robe | +0.014 |
drops | +0.014 |
Neg logit contributions | |
Enjoy | -0.010 |
ours | -0.009 |
coconut | -0.009 |
whose | -0.009 |
Sometimes | -0.008 |
Token | she |
Token ID | 673 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
aces | +0.000 |
amas | +0.000 |
coat | +0.000 |
robe | +0.000 |
forehead | +0.000 |
Neg logit contributions | |
ours | -0.000 |
Enjoy | -0.000 |
coconut | -0.000 |
oral | -0.000 |
rust | -0.000 |
Token | He |
Token ID | 679 |
Feature activation | -2.20e-01 |
Pos logit contributions | |
coat | +0.009 |
amas | +0.009 |
aces | +0.009 |
robe | +0.009 |
drops | +0.009 |
Neg logit contributions | |
ours | -0.008 |
coconut | -0.008 |
Enjoy | -0.008 |
oral | -0.008 |
rust | -0.007 |
Token | wore |
Token ID | 12408 |
Feature activation | -9.15e-01 |
Pos logit contributions | |
coat | +0.018 |
aces | +0.018 |
robe | +0.017 |
amas | +0.017 |
drops | +0.016 |
Neg logit contributions | |
ours | -0.015 |
Enjoy | -0.014 |
coconut | -0.014 |
oral | -0.014 |
iw | -0.012 |
Token | his |
Token ID | 465 |
Feature activation | +8.95e-01 |
Pos logit contributions | |
amas | +0.095 |
forehead | +0.091 |
coat | +0.090 |
drops | +0.089 |
aces | +0.089 |
Neg logit contributions | |
coconut | -0.044 |
ours | -0.042 |
Enjoy | -0.040 |
rust | -0.037 |
Sometimes | -0.034 |
Token | big |
Token ID | 1263 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
coconut | +0.060 |
ours | +0.056 |
crab | +0.052 |
rust | +0.050 |
Enjoy | +0.050 |
Neg logit contributions | |
amas | -0.075 |
drops | -0.070 |
stand | -0.070 |
coat | -0.069 |
forehead | -0.068 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.07e+00 |
Pos logit contributions | |
aces | +0.078 |
amas | +0.077 |
coat | +0.076 |
drops | +0.074 |
forehead | +0.073 |
Neg logit contributions | |
ours | -0.077 |
coconut | -0.076 |
Enjoy | -0.072 |
oral | -0.068 |
rust | -0.067 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
forehead | +0.083 |
amas | +0.074 |
stand | +0.068 |
drops | +0.066 |
click | +0.064 |
Neg logit contributions | |
ours | -0.209 |
coconut | -0.208 |
swamp | -0.207 |
arium | -0.202 |
crab | -0.199 |
Token | because |
Token ID | 780 |
Feature activation | +7.55e-01 |
Pos logit contributions | |
drops | +1.101 |
amas | +1.101 |
aces | +1.066 |
bridge | +1.060 |
forehead | +1.058 |
Neg logit contributions | |
Enjoy | -0.468 |
coconut | -0.452 |
arium | -0.416 |
ours | -0.411 |
: | -0.400 |
Token | it |
Token ID | 340 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
ours | +0.044 |
coconut | +0.042 |
Enjoy | +0.042 |
oral | +0.039 |
rust | +0.037 |
Neg logit contributions | |
aces | -0.068 |
coat | -0.065 |
robe | -0.065 |
amas | -0.064 |
bridge | -0.063 |
Token | was |
Token ID | 373 |
Feature activation | +4.72e-01 |
Pos logit contributions | |
coconut | +0.052 |
ours | +0.051 |
rust | +0.047 |
swamp | +0.043 |
Enjoy | +0.042 |
Neg logit contributions | |
amas | -0.132 |
forehead | -0.130 |
aces | -0.127 |
coat | -0.124 |
stand | -0.124 |
Token | cold |
Token ID | 4692 |
Feature activation | +5.53e-01 |
Pos logit contributions | |
coconut | +0.031 |
ours | +0.030 |
Enjoy | +0.029 |
rust | +0.027 |
grown | +0.026 |
Neg logit contributions | |
amas | -0.040 |
aces | -0.039 |
coat | -0.038 |
forehead | -0.037 |
drops | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
coconut | +0.022 |
Enjoy | +0.022 |
ours | +0.022 |
Sometimes | +0.022 |
grown | +0.020 |
Neg logit contributions | |
forehead | -0.055 |
aces | -0.054 |
bes | -0.053 |
amas | -0.052 |
bridge | -0.052 |
Token | put |
Token ID | 1234 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
aces | +0.016 |
coat | +0.016 |
amas | +0.016 |
forehead | +0.015 |
robe | +0.015 |
Neg logit contributions | |
ours | -0.011 |
coconut | -0.010 |
Enjoy | -0.010 |
oral | -0.009 |
rust | -0.009 |
Token | on |
Token ID | 319 |
Feature activation | +6.94e-01 |
Pos logit contributions | |
amas | +0.169 |
aces | +0.166 |
drops | +0.165 |
coat | +0.163 |
bridge | +0.159 |
Neg logit contributions | |
ours | -0.083 |
Enjoy | -0.076 |
coconut | -0.067 |
iw | -0.063 |
Sometimes | -0.063 |
Token | the |
Token ID | 262 |
Feature activation | -4.34e-02 |
Pos logit contributions | |
ours | +0.036 |
Enjoy | +0.034 |
coconut | +0.034 |
Sometimes | +0.029 |
iw | +0.028 |
Neg logit contributions | |
aces | -0.067 |
amas | -0.065 |
coat | -0.064 |
bridge | -0.064 |
drops | -0.064 |
Token | new |
Token ID | 649 |
Feature activation | +1.71e+00 |
Pos logit contributions | |
aces | +0.004 |
drops | +0.004 |
amas | +0.004 |
bridge | +0.004 |
coat | +0.004 |
Neg logit contributions | |
ours | -0.002 |
shrimp | -0.002 |
Sometimes | -0.002 |
Enjoy | -0.002 |
oral | -0.002 |
Token | bath |
Token ID | 7837 |
Feature activation | -2.14e+00 |
Pos logit contributions | |
ours | +0.115 |
Enjoy | +0.107 |
coconut | +0.106 |
shrimp | +0.102 |
Sometimes | +0.101 |
Neg logit contributions | |
aces | -0.141 |
drops | -0.132 |
amas | -0.130 |
bridge | -0.129 |
coat | -0.122 |
Token | robe |
Token ID | 25481 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
aces | +0.071 |
bridge | +0.070 |
forehead | +0.070 |
drops | +0.064 |
ring | +0.060 |
Neg logit contributions | |
ours | -0.230 |
Enjoy | -0.212 |
ry | -0.209 |
Sometimes | -0.208 |
arium | -0.208 |
Token | and |
Token ID | 290 |
Feature activation | +5.83e-02 |
Pos logit contributions | |
amas | +1.064 |
aces | +1.035 |
drops | +1.035 |
forehead | +1.028 |
coat | +1.014 |
Neg logit contributions | |
ours | -0.534 |
Enjoy | -0.512 |
coconut | -0.480 |
arium | -0.477 |
Sometimes | -0.434 |
Token | it |
Token ID | 340 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
ours | +0.003 |
coconut | +0.003 |
Enjoy | +0.003 |
asks | +0.003 |
rust | +0.003 |
Neg logit contributions | |
aces | -0.005 |
amas | -0.005 |
coat | -0.005 |
forehead | -0.005 |
bridge | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | +5.18e-01 |
Pos logit contributions | |
ours | +0.039 |
arose | +0.032 |
asks | +0.030 |
arium | +0.029 |
lly | +0.029 |
Neg logit contributions | |
amas | -0.127 |
aces | -0.126 |
forehead | -0.122 |
bridge | -0.116 |
coat | -0.115 |
Token | a |
Token ID | 257 |
Feature activation | -5.48e-01 |
Pos logit contributions | |
ours | +0.031 |
coconut | +0.029 |
Enjoy | +0.029 |
what | +0.026 |
Sometimes | +0.026 |
Neg logit contributions | |
aces | -0.045 |
amas | -0.045 |
drops | -0.043 |
forehead | -0.042 |
coat | -0.042 |
Token | perfect |
Token ID | 2818 |
Feature activation | -3.13e+00 |
Pos logit contributions | |
aces | +0.043 |
amas | +0.042 |
coat | +0.041 |
robe | +0.039 |
drops | +0.039 |
Neg logit contributions | |
ours | -0.039 |
coconut | -0.038 |
Enjoy | -0.037 |
oral | -0.035 |
rust | -0.034 |
Token | still |
Token ID | 991 |
Feature activation | +9.06e-01 |
Pos logit contributions | |
aces | +0.110 |
coat | +0.107 |
drops | +0.100 |
robe | +0.100 |
bridge | +0.099 |
Neg logit contributions | |
ours | -0.124 |
Enjoy | -0.119 |
coconut | -0.119 |
oral | -0.108 |
iw | -0.108 |
Token | did |
Token ID | 750 |
Feature activation | +4.61e-01 |
Pos logit contributions | |
ours | +0.062 |
coconut | +0.061 |
Enjoy | +0.059 |
oral | +0.058 |
rust | +0.054 |
Neg logit contributions | |
aces | -0.074 |
coat | -0.072 |
forehead | -0.069 |
drops | -0.069 |
robe | -0.068 |
Token | not |
Token ID | 407 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
ours | +0.035 |
coconut | +0.034 |
Enjoy | +0.032 |
rust | +0.031 |
oral | +0.031 |
Neg logit contributions | |
aces | -0.034 |
amas | -0.032 |
stand | -0.032 |
coat | -0.032 |
forehead | -0.031 |
Token | want |
Token ID | 765 |
Feature activation | -4.38e-01 |
Pos logit contributions | |
aces | +0.084 |
coat | +0.082 |
amas | +0.080 |
drops | +0.078 |
forehead | +0.078 |
Neg logit contributions | |
ours | -0.068 |
coconut | -0.065 |
Enjoy | -0.063 |
oral | -0.059 |
rust | -0.057 |
Token | to |
Token ID | 284 |
Feature activation | +5.41e-01 |
Pos logit contributions | |
aces | +0.033 |
stand | +0.031 |
coat | +0.029 |
drops | +0.029 |
robe | +0.029 |
Neg logit contributions | |
ours | -0.033 |
oral | -0.031 |
coconut | -0.030 |
Enjoy | -0.030 |
rust | -0.029 |
Token | touch |
Token ID | 3638 |
Feature activation | +5.00e+00 |
Pos logit contributions | |
ours | +0.043 |
coconut | +0.040 |
Enjoy | +0.039 |
oral | +0.038 |
iw | +0.037 |
Neg logit contributions | |
aces | -0.038 |
coat | -0.038 |
stand | -0.037 |
drops | -0.037 |
amas | -0.037 |
Token | the |
Token ID | 262 |
Feature activation | +4.61e-02 |
Pos logit contributions | |
ours | +0.329 |
Enjoy | +0.320 |
coconut | +0.314 |
oral | +0.310 |
rust | +0.285 |
Neg logit contributions | |
aces | -0.425 |
amas | -0.395 |
coat | -0.393 |
robe | -0.385 |
drops | -0.384 |
Token | snakes |
Token ID | 26042 |
Feature activation | +2.33e+00 |
Pos logit contributions | |
ours | +0.003 |
Enjoy | +0.003 |
coconut | +0.003 |
rust | +0.003 |
oral | +0.003 |
Neg logit contributions | |
aces | -0.004 |
drops | -0.004 |
amas | -0.004 |
coat | -0.004 |
robe | -0.004 |
Token | , |
Token ID | 11 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
ours | +0.135 |
coconut | +0.133 |
Enjoy | +0.127 |
oral | +0.116 |
rust | +0.113 |
Neg logit contributions | |
amas | -0.208 |
aces | -0.207 |
coat | -0.205 |
forehead | -0.198 |
robe | -0.197 |
Token | but |
Token ID | 475 |
Feature activation | -2.10e+00 |
Pos logit contributions | |
ours | +0.001 |
coconut | +0.001 |
Enjoy | +0.001 |
oral | +0.001 |
rust | +0.001 |
Neg logit contributions | |
aces | -0.001 |
coat | -0.001 |
amas | -0.001 |
drops | -0.001 |
forehead | -0.001 |
Token | he |
Token ID | 339 |
Feature activation | -3.18e-01 |
Pos logit contributions | |
aces | +0.151 |
robe | +0.140 |
forehead | +0.138 |
coat | +0.135 |
drops | +0.134 |
Neg logit contributions | |
ours | -0.161 |
oral | -0.155 |
Enjoy | -0.153 |
coconut | -0.148 |
iw | -0.141 |
Token | of |
Token ID | 286 |
Feature activation | +5.90e-02 |
Pos logit contributions | |
aces | +0.134 |
amas | +0.131 |
coat | +0.131 |
robe | +0.126 |
forehead | +0.125 |
Neg logit contributions | |
ours | -0.095 |
coconut | -0.092 |
Enjoy | -0.090 |
oral | -0.083 |
rust | -0.081 |
Token | the |
Token ID | 262 |
Feature activation | -4.89e-02 |
Pos logit contributions | |
ours | +0.004 |
Enjoy | +0.004 |
oral | +0.004 |
coconut | +0.004 |
rust | +0.004 |
Neg logit contributions | |
coat | -0.004 |
aces | -0.004 |
stand | -0.004 |
amas | -0.004 |
forehead | -0.004 |
Token | dark |
Token ID | 3223 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
amas | +0.003 |
forehead | +0.003 |
coat | +0.003 |
aces | +0.003 |
robe | +0.003 |
Neg logit contributions | |
ours | -0.004 |
Enjoy | -0.003 |
coconut | -0.003 |
iw | -0.003 |
oral | -0.003 |
Token | forest |
Token ID | 8222 |
Feature activation | +2.51e+00 |
Pos logit contributions | |
ours | +0.097 |
Enjoy | +0.092 |
coconut | +0.090 |
oral | +0.083 |
rust | +0.080 |
Neg logit contributions | |
aces | -0.136 |
coat | -0.132 |
amas | -0.132 |
robe | -0.128 |
drops | -0.128 |
Token | . |
Token ID | 13 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
coconut | +0.114 |
ours | +0.108 |
Enjoy | +0.101 |
grown | +0.097 |
arium | +0.097 |
Neg logit contributions | |
amas | -0.252 |
aces | -0.247 |
coat | -0.241 |
forehead | -0.238 |
robe | -0.236 |
Token | Jane |
Token ID | 12091 |
Feature activation | +4.97e+00 |
Pos logit contributions | |
ours | +0.009 |
coconut | +0.008 |
rust | +0.008 |
oral | +0.008 |
crab | +0.007 |
Neg logit contributions | |
aces | -0.011 |
amas | -0.011 |
coat | -0.010 |
forehead | -0.010 |
drops | -0.010 |
Token | was |
Token ID | 373 |
Feature activation | +5.11e-01 |
Pos logit contributions | |
coconut | +0.253 |
ours | +0.251 |
Enjoy | +0.233 |
oral | +0.218 |
rust | +0.212 |
Neg logit contributions | |
amas | -0.475 |
aces | -0.475 |
coat | -0.465 |
robe | -0.455 |
forehead | -0.445 |
Token | never |
Token ID | 1239 |
Feature activation | -9.55e-01 |
Pos logit contributions | |
coconut | +0.031 |
Enjoy | +0.031 |
ours | +0.030 |
grown | +0.029 |
what | +0.028 |
Neg logit contributions | |
amas | -0.044 |
aces | -0.043 |
robe | -0.041 |
coat | -0.041 |
drops | -0.039 |
Token | able |
Token ID | 1498 |
Feature activation | -4.12e+00 |
Pos logit contributions | |
amas | +0.091 |
aces | +0.087 |
robe | +0.085 |
coat | +0.084 |
drops | +0.082 |
Neg logit contributions | |
coconut | -0.051 |
grown | -0.049 |
ours | -0.048 |
Enjoy | -0.047 |
fired | -0.045 |
Token | to |
Token ID | 284 |
Feature activation | +4.41e-01 |
Pos logit contributions | |
aces | +0.355 |
coat | +0.346 |
amas | +0.337 |
forehead | +0.332 |
drops | +0.330 |
Neg logit contributions | |
ours | -0.263 |
coconut | -0.244 |
Enjoy | -0.243 |
oral | -0.234 |
rust | -0.221 |
Token | go |
Token ID | 467 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
ours | +0.033 |
coconut | +0.031 |
Enjoy | +0.031 |
oral | +0.029 |
rust | +0.028 |
Neg logit contributions | |
aces | -0.033 |
coat | -0.033 |
amas | -0.032 |
drops | -0.031 |
forehead | -0.031 |
Token | ly |
Token ID | 306 |
Feature activation | -3.35e+00 |
Pos logit contributions | |
aces | +0.277 |
amas | +0.273 |
bridge | +0.269 |
coat | +0.265 |
cartridge | +0.263 |
Neg logit contributions | |
coconut | -0.214 |
Joe | -0.209 |
ours | -0.204 |
Enjoy | -0.200 |
oy | -0.199 |
Token | at |
Token ID | 379 |
Feature activation | +2.75e+00 |
Pos logit contributions | |
amas | +0.298 |
coat | +0.294 |
aces | +0.294 |
robe | +0.290 |
bridge | +0.277 |
Neg logit contributions | |
ours | -0.193 |
coconut | -0.192 |
Enjoy | -0.180 |
grown | -0.164 |
oral | -0.159 |
Token | Bob |
Token ID | 5811 |
Feature activation | +3.56e+00 |
Pos logit contributions | |
ours | +0.193 |
oral | +0.175 |
Enjoy | +0.170 |
coconut | +0.170 |
iw | +0.162 |
Neg logit contributions | |
aces | -0.221 |
forehead | -0.214 |
amas | -0.204 |
coat | -0.201 |
stand | -0.197 |
Token | and |
Token ID | 290 |
Feature activation | +2.02e-01 |
Pos logit contributions | |
ours | +0.208 |
coconut | +0.206 |
Enjoy | +0.199 |
oral | +0.172 |
rust | +0.171 |
Neg logit contributions | |
aces | -0.310 |
coat | -0.307 |
amas | -0.307 |
robe | -0.300 |
bridge | -0.297 |
Token | said |
Token ID | 531 |
Feature activation | -6.61e-01 |
Pos logit contributions | |
ours | +0.014 |
Enjoy | +0.013 |
coconut | +0.013 |
oral | +0.013 |
rust | +0.012 |
Neg logit contributions | |
aces | -0.016 |
coat | -0.016 |
forehead | -0.016 |
amas | -0.015 |
drops | -0.015 |
Token | : |
Token ID | 25 |
Feature activation | +5.15e+00 |
Pos logit contributions | |
aces | +0.056 |
coat | +0.054 |
amas | +0.054 |
forehead | +0.052 |
robe | +0.052 |
Neg logit contributions | |
ours | -0.043 |
coconut | -0.041 |
Enjoy | -0.040 |
oral | -0.038 |
rust | -0.037 |
Token |
|
Token ID | 198 |
Feature activation | -2.94e-02 |
Pos logit contributions | |
ours | +0.434 |
coconut | +0.412 |
rous | +0.389 |
rust | +0.381 |
oral | +0.378 |
Neg logit contributions | |
stand | -0.328 |
amas | -0.327 |
aces | -0.313 |
robe | -0.306 |
coat | -0.304 |
Token |
|
Token ID | 198 |
Feature activation | -3.87e-02 |
Pos logit contributions | |
stand | +0.002 |
amas | +0.002 |
aces | +0.002 |
coat | +0.002 |
bridge | +0.002 |
Neg logit contributions | |
ours | -0.002 |
coconut | -0.002 |
rust | -0.002 |
rous | -0.002 |
Enjoy | -0.002 |
Token | " |
Token ID | 1 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
aces | +0.002 |
coat | +0.002 |
amas | +0.002 |
stand | +0.002 |
drops | +0.002 |
Neg logit contributions | |
ours | -0.003 |
coconut | -0.003 |
Enjoy | -0.003 |
rust | -0.003 |
arium | -0.003 |
Token | Don |
Token ID | 3987 |
Feature activation | -2.59e+00 |
Pos logit contributions | |
aces | +0.150 |
coat | +0.147 |
amas | +0.146 |
robe | +0.140 |
forehead | +0.140 |
Neg logit contributions | |
ours | -0.129 |
coconut | -0.123 |
Enjoy | -0.120 |
oral | -0.112 |
rust | -0.108 |
Token | 't |
Token ID | 470 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
aces | +0.179 |
coat | +0.174 |
amas | +0.168 |
robe | +0.164 |
forehead | +0.162 |
Neg logit contributions | |
ours | -0.207 |
Enjoy | -0.206 |
coconut | -0.204 |
oral | -0.197 |
Sometimes | -0.183 |
Token | who |
Token ID | 508 |
Feature activation | +3.73e+00 |
Pos logit contributions | |
aces | +0.008 |
coat | +0.008 |
stand | +0.008 |
amas | +0.008 |
forehead | +0.008 |
Neg logit contributions | |
ours | -0.009 |
coconut | -0.009 |
oral | -0.009 |
Enjoy | -0.009 |
rust | -0.008 |
Token | was |
Token ID | 373 |
Feature activation | +5.55e-01 |
Pos logit contributions | |
ours | +0.279 |
coconut | +0.261 |
Enjoy | +0.259 |
oral | +0.249 |
rust | +0.227 |
Neg logit contributions | |
coat | -0.276 |
aces | -0.276 |
robe | -0.269 |
amas | -0.265 |
stand | -0.263 |
Token | asking |
Token ID | 4737 |
Feature activation | -4.49e-01 |
Pos logit contributions | |
ours | +0.042 |
coconut | +0.040 |
Enjoy | +0.040 |
oral | +0.039 |
rust | +0.036 |
Neg logit contributions | |
aces | -0.041 |
coat | -0.040 |
amas | -0.039 |
forehead | -0.038 |
robe | -0.038 |
Token | him |
Token ID | 683 |
Feature activation | -1.72e+00 |
Pos logit contributions | |
aces | +0.038 |
amas | +0.037 |
coat | +0.037 |
forehead | +0.036 |
robe | +0.036 |
Neg logit contributions | |
ours | -0.029 |
coconut | -0.028 |
Enjoy | -0.027 |
oral | -0.025 |
rust | -0.025 |
Token | the |
Token ID | 262 |
Feature activation | -8.71e-02 |
Pos logit contributions | |
aces | +0.144 |
robe | +0.136 |
coat | +0.136 |
forehead | +0.133 |
amas | +0.132 |
Neg logit contributions | |
ours | -0.111 |
coconut | -0.107 |
Enjoy | -0.100 |
oral | -0.100 |
rust | -0.098 |
Token | question |
Token ID | 1808 |
Feature activation | +6.10e+00 |
Pos logit contributions | |
coat | +0.007 |
aces | +0.006 |
robe | +0.006 |
forehead | +0.006 |
drops | +0.006 |
Neg logit contributions | |
ours | -0.007 |
coconut | -0.006 |
Enjoy | -0.006 |
rous | -0.006 |
rust | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
ours | +0.379 |
coconut | +0.367 |
Enjoy | +0.359 |
oral | +0.332 |
rust | +0.322 |
Neg logit contributions | |
aces | -0.526 |
amas | -0.514 |
coat | -0.513 |
robe | -0.494 |
forehead | -0.493 |
Token |
|
Token ID | 198 |
Feature activation | -8.98e-02 |
Pos logit contributions | |
aces | +0.009 |
amas | +0.009 |
coat | +0.008 |
stand | +0.008 |
robe | +0.008 |
Neg logit contributions | |
ours | -0.009 |
coconut | -0.009 |
oral | -0.008 |
Enjoy | -0.008 |
rust | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | -9.54e-02 |
Pos logit contributions | |
aces | +0.008 |
coat | +0.008 |
amas | +0.008 |
stand | +0.007 |
robe | +0.007 |
Neg logit contributions | |
ours | -0.006 |
coconut | -0.005 |
Enjoy | -0.005 |
oral | -0.005 |
rust | -0.005 |
Token | The |
Token ID | 464 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
aces | +0.008 |
amas | +0.008 |
coat | +0.008 |
forehead | +0.008 |
robe | +0.008 |
Neg logit contributions | |
Enjoy | -0.006 |
coconut | -0.006 |
ours | -0.006 |
oral | -0.005 |
rust | -0.005 |
Token | le |
Token ID | 443 |
Feature activation | +5.21e+00 |
Pos logit contributions | |
coat | +0.080 |
drops | +0.077 |
forehead | +0.076 |
robe | +0.075 |
aces | +0.074 |
Neg logit contributions | |
ours | -0.086 |
Enjoy | -0.077 |
coconut | -0.075 |
arium | -0.073 |
rous | -0.073 |
Token | € |
Token ID | 26391 |
Feature activation | -1.96e+00 |
Pos logit contributions | |
aces | +0.060 |
amas | +0.059 |
coat | +0.057 |
robe | +0.055 |
forehead | +0.055 |
Neg logit contributions | |
Enjoy | -0.075 |
ours | -0.070 |
coconut | -0.069 |
oral | -0.068 |
rust | -0.066 |
Token | â„¢ |
Token ID | 8151 |
Feature activation | -5.52e+00 |
Pos logit contributions | |
forehead | +0.176 |
aces | +0.174 |
coat | +0.165 |
robe | +0.165 |
drops | +0.164 |
Neg logit contributions | |
ours | -0.116 |
Enjoy | -0.113 |
coconut | -0.105 |
iw | -0.095 |
oral | -0.095 |
Token | re |
Token ID | 260 |
Feature activation | -6.83e-02 |
Pos logit contributions | |
aces | +0.528 |
coat | +0.518 |
amas | +0.516 |
robe | +0.499 |
forehead | +0.495 |
Neg logit contributions | |
ours | -0.293 |
coconut | -0.282 |
Enjoy | -0.274 |
oral | -0.249 |
rust | -0.241 |
Token | the |
Token ID | 262 |
Feature activation | -4.05e-02 |
Pos logit contributions | |
coat | +0.007 |
amas | +0.007 |
aces | +0.006 |
drops | +0.006 |
stand | +0.006 |
Neg logit contributions | |
Enjoy | -0.004 |
coconut | -0.003 |
ours | -0.003 |
Joe | -0.003 |
what | -0.003 |
Token | best |
Token ID | 1266 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
amas | +0.003 |
aces | +0.003 |
coat | +0.003 |
forehead | +0.003 |
drops | +0.003 |
Neg logit contributions | |
coconut | -0.003 |
Enjoy | -0.003 |
ours | -0.003 |
oral | -0.003 |
rust | -0.003 |
Token | vendor |
Token ID | 18371 |
Feature activation | +5.13e+00 |
Pos logit contributions | |
aces | +0.106 |
coat | +0.105 |
amas | +0.104 |
robe | +0.102 |
forehead | +0.101 |
Neg logit contributions | |
ours | -0.094 |
coconut | -0.092 |
Enjoy | -0.088 |
oral | -0.084 |
rust | -0.084 |
Token | â |
Token ID | 22940 |
Feature activation | -9.13e-01 |
Pos logit contributions | |
ours | +0.301 |
Enjoy | +0.298 |
coconut | +0.297 |
iw | +0.260 |
oral | +0.255 |
Neg logit contributions | |
amas | -0.457 |
aces | -0.444 |
coat | -0.441 |
forehead | -0.426 |
drops | -0.422 |
Token | € |
Token ID | 26391 |
Feature activation | -1.94e+00 |
Pos logit contributions | |
aces | +0.062 |
amas | +0.061 |
coat | +0.058 |
forehead | +0.058 |
bridge | +0.057 |
Neg logit contributions | |
Enjoy | -0.073 |
ours | -0.069 |
coconut | -0.066 |
oral | -0.066 |
rust | -0.064 |
Token | , |
Token ID | 11 |
Feature activation | -1.47e-01 |
Pos logit contributions | |
forehead | +0.171 |
aces | +0.171 |
coat | +0.162 |
amas | +0.162 |
drops | +0.162 |
Neg logit contributions | |
ours | -0.114 |
Enjoy | -0.113 |
coconut | -0.106 |
oral | -0.096 |
grown | -0.095 |
Token | said |
Token ID | 531 |
Feature activation | -6.95e-01 |
Pos logit contributions | |
aces | +0.012 |
amas | +0.012 |
coat | +0.012 |
robe | +0.011 |
forehead | +0.011 |
Neg logit contributions | |
ours | -0.010 |
coconut | -0.009 |
Enjoy | -0.009 |
oral | -0.008 |
rust | -0.008 |
Token | Midnight |
Token ID | 27960 |
Feature activation | -5.25e+00 |
Pos logit contributions | |
aces | +0.061 |
amas | +0.060 |
coat | +0.060 |
drops | +0.057 |
robe | +0.056 |
Neg logit contributions | |
Enjoy | -0.043 |
ours | -0.042 |
coconut | -0.039 |
oral | -0.037 |
Sometimes | -0.035 |
Token | and |
Token ID | 290 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
ours | +0.020 |
coconut | +0.020 |
Enjoy | +0.019 |
Crab | +0.019 |
rust | +0.018 |
Neg logit contributions | |
coat | -0.047 |
aces | -0.046 |
amas | -0.045 |
drops | -0.044 |
bes | -0.044 |
Token | said |
Token ID | 531 |
Feature activation | -8.52e-01 |
Pos logit contributions | |
aces | +0.010 |
amas | +0.010 |
coat | +0.010 |
robe | +0.009 |
forehead | +0.009 |
Neg logit contributions | |
ours | -0.007 |
coconut | -0.007 |
Enjoy | -0.007 |
oral | -0.006 |
rust | -0.006 |
Token | , |
Token ID | 11 |
Feature activation | -2.24e-01 |
Pos logit contributions | |
aces | +0.076 |
amas | +0.075 |
coat | +0.075 |
robe | +0.072 |
forehead | +0.072 |
Neg logit contributions | |
ours | -0.051 |
coconut | -0.049 |
Enjoy | -0.048 |
oral | -0.044 |
rust | -0.042 |
Token | " |
Token ID | 366 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
aces | +0.020 |
coat | +0.020 |
amas | +0.019 |
robe | +0.019 |
drops | +0.018 |
Neg logit contributions | |
ours | -0.013 |
Enjoy | -0.013 |
coconut | -0.013 |
oral | -0.012 |
rust | -0.011 |
Token | That |
Token ID | 2504 |
Feature activation | +1.59e+00 |
Pos logit contributions | |
aces | +0.145 |
amas | +0.142 |
coat | +0.141 |
forehead | +0.139 |
robe | +0.136 |
Neg logit contributions | |
Enjoy | -0.099 |
coconut | -0.097 |
ours | -0.095 |
oral | -0.089 |
rust | -0.088 |
Token | looks |
Token ID | 3073 |
Feature activation | +2.20e+00 |
Pos logit contributions | |
ours | +0.124 |
coconut | +0.120 |
Enjoy | +0.119 |
oral | +0.112 |
rust | +0.109 |
Neg logit contributions | |
aces | -0.113 |
amas | -0.110 |
coat | -0.109 |
robe | -0.105 |
forehead | -0.104 |
Token | very |
Token ID | 845 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
ours | +0.162 |
coconut | +0.160 |
Enjoy | +0.157 |
rust | +0.143 |
oral | +0.141 |
Neg logit contributions | |
aces | -0.165 |
amas | -0.164 |
coat | -0.162 |
robe | -0.155 |
forehead | -0.155 |
Token | nice |
Token ID | 3621 |
Feature activation | -7.55e-01 |
Pos logit contributions | |
ours | +0.180 |
oral | +0.171 |
coconut | +0.170 |
Enjoy | +0.167 |
arose | +0.160 |
Neg logit contributions | |
aces | -0.103 |
coat | -0.095 |
drops | -0.095 |
robe | -0.094 |
forehead | -0.094 |
Token | Jack |
Token ID | 3619 |
Feature activation | +1.35e+00 |
Pos logit contributions | |
aces | +0.064 |
amas | +0.064 |
coat | +0.063 |
forehead | +0.061 |
robe | +0.060 |
Neg logit contributions | |
ours | -0.048 |
coconut | -0.047 |
Enjoy | -0.046 |
oral | -0.041 |
rust | -0.040 |
Token | . |
Token ID | 13 |
Feature activation | -2.30e-01 |
Pos logit contributions | |
ours | +0.079 |
coconut | +0.077 |
Enjoy | +0.076 |
oral | +0.066 |
rust | +0.065 |
Neg logit contributions | |
amas | -0.117 |
coat | -0.117 |
drops | -0.115 |
aces | -0.115 |
forehead | -0.112 |
Token | You |
Token ID | 921 |
Feature activation | -6.35e-01 |
Pos logit contributions | |
aces | +0.015 |
amas | +0.015 |
robe | +0.014 |
coat | +0.014 |
stand | +0.014 |
Neg logit contributions | |
coconut | -0.019 |
ours | -0.018 |
oral | -0.017 |
rous | -0.017 |
rust | -0.017 |
Token | my |
Token ID | 1820 |
Feature activation | -6.35e-01 |
Pos logit contributions | |
aces | +0.145 |
coat | +0.143 |
amas | +0.143 |
forehead | +0.138 |
drops | +0.137 |
Neg logit contributions | |
ours | -0.107 |
coconut | -0.106 |
Enjoy | -0.103 |
oral | -0.095 |
rust | -0.094 |
Token | heard |
Token ID | 2982 |
Feature activation | +1.51e+00 |
Pos logit contributions | |
aces | +0.056 |
amas | +0.054 |
coat | +0.054 |
robe | +0.052 |
forehead | +0.052 |
Neg logit contributions | |
ours | -0.039 |
coconut | -0.038 |
Enjoy | -0.037 |
oral | -0.034 |
rust | -0.033 |
Token | the |
Token ID | 262 |
Feature activation | +2.34e-01 |
Pos logit contributions | |
ours | +0.088 |
coconut | +0.085 |
Enjoy | +0.083 |
oral | +0.077 |
rust | +0.074 |
Neg logit contributions | |
aces | -0.136 |
amas | -0.134 |
coat | -0.133 |
forehead | -0.129 |
stand | -0.129 |
Token | noise |
Token ID | 7838 |
Feature activation | -2.02e+00 |
Pos logit contributions | |
coconut | +0.019 |
ours | +0.019 |
Enjoy | +0.018 |
oral | +0.018 |
volcano | +0.017 |
Neg logit contributions | |
aces | -0.016 |
coat | -0.015 |
amas | -0.014 |
robe | -0.014 |
cartridge | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
aces | +0.203 |
amas | +0.196 |
coat | +0.196 |
robe | +0.195 |
forehead | +0.186 |
Neg logit contributions | |
ours | -0.096 |
coconut | -0.086 |
oral | -0.080 |
Enjoy | -0.079 |
arose | -0.079 |
Token | She |
Token ID | 1375 |
Feature activation | -1.97e-01 |
Pos logit contributions | |
amas | +0.007 |
aces | +0.006 |
coat | +0.006 |
robe | +0.006 |
drops | +0.006 |
Neg logit contributions | |
coconut | -0.010 |
ours | -0.010 |
Enjoy | -0.010 |
rust | -0.010 |
oral | -0.009 |
Token | came |
Token ID | 1625 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
aces | +0.014 |
coat | +0.014 |
amas | +0.014 |
drops | +0.014 |
forehead | +0.014 |
Neg logit contributions | |
ours | -0.015 |
Enjoy | -0.014 |
coconut | -0.014 |
oral | -0.013 |
rust | -0.013 |
Token | to |
Token ID | 284 |
Feature activation | +7.41e-01 |
Pos logit contributions | |
aces | +0.121 |
amas | +0.119 |
coat | +0.119 |
forehead | +0.113 |
robe | +0.113 |
Neg logit contributions | |
ours | -0.113 |
coconut | -0.110 |
Enjoy | -0.108 |
oral | -0.101 |
rust | -0.098 |
Token | the |
Token ID | 262 |
Feature activation | +2.71e-01 |
Pos logit contributions | |
ours | +0.049 |
coconut | +0.046 |
Enjoy | +0.044 |
oral | +0.042 |
rust | +0.040 |
Neg logit contributions | |
coat | -0.061 |
aces | -0.060 |
amas | -0.060 |
forehead | -0.059 |
stand | -0.059 |
Token | kitchen |
Token ID | 9592 |
Feature activation | +9.00e-01 |
Pos logit contributions | |
ours | +0.020 |
coconut | +0.019 |
Enjoy | +0.019 |
rust | +0.018 |
oral | +0.018 |
Neg logit contributions | |
amas | -0.020 |
aces | -0.020 |
coat | -0.020 |
forehead | -0.019 |
robe | -0.019 |
Token | and |
Token ID | 290 |
Feature activation | +2.00e-01 |
Pos logit contributions | |
ours | +0.046 |
coconut | +0.041 |
Enjoy | +0.039 |
temple | +0.036 |
Sometimes | +0.036 |
Neg logit contributions | |
aces | -0.091 |
coat | -0.085 |
robe | -0.083 |
bridge | -0.083 |
amas | -0.083 |
Token | , |
Token ID | 11 |
Feature activation | +6.64e-02 |
Pos logit contributions | |
aces | +0.073 |
coat | +0.072 |
amas | +0.070 |
drops | +0.067 |
forehead | +0.067 |
Neg logit contributions | |
ours | -0.069 |
Enjoy | -0.067 |
oral | -0.064 |
coconut | -0.063 |
rust | -0.063 |
Token | red |
Token ID | 2266 |
Feature activation | -1.82e+00 |
Pos logit contributions | |
coconut | +0.005 |
ours | +0.005 |
Enjoy | +0.004 |
iw | +0.004 |
oral | +0.004 |
Neg logit contributions | |
aces | -0.005 |
amas | -0.005 |
robe | -0.005 |
coat | -0.005 |
stand | -0.005 |
Token | apple |
Token ID | 17180 |
Feature activation | +2.93e+00 |
Pos logit contributions | |
aces | +0.149 |
coat | +0.146 |
amas | +0.145 |
drops | +0.137 |
forehead | +0.136 |
Neg logit contributions | |
ours | -0.122 |
Enjoy | -0.115 |
coconut | -0.113 |
rust | -0.111 |
oral | -0.110 |
Token | . |
Token ID | 13 |
Feature activation | +2.94e-01 |
Pos logit contributions | |
coconut | +0.176 |
ours | +0.171 |
Enjoy | +0.156 |
iw | +0.145 |
oral | +0.140 |
Neg logit contributions | |
amas | -0.264 |
aces | -0.256 |
coat | -0.255 |
forehead | -0.251 |
robe | -0.246 |
Token | She |
Token ID | 1375 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
coconut | +0.018 |
ours | +0.018 |
Enjoy | +0.017 |
oral | +0.017 |
rust | +0.017 |
Neg logit contributions | |
aces | -0.025 |
amas | -0.025 |
coat | -0.025 |
stand | -0.024 |
drops | -0.023 |
Token | was |
Token ID | 373 |
Feature activation | +6.80e-01 |
Pos logit contributions | |
aces | +0.018 |
coat | +0.017 |
amas | +0.016 |
drops | +0.016 |
robe | +0.016 |
Neg logit contributions | |
ours | -0.014 |
Enjoy | -0.014 |
coconut | -0.014 |
oral | -0.013 |
rous | -0.012 |
Token | very |
Token ID | 845 |
Feature activation | +2.12e+00 |
Pos logit contributions | |
coconut | +0.039 |
ours | +0.037 |
Enjoy | +0.037 |
iw | +0.032 |
what | +0.032 |
Neg logit contributions | |
amas | -0.064 |
robe | -0.060 |
cartridge | -0.059 |
stand | -0.059 |
coat | -0.058 |
Token | happy |
Token ID | 3772 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
ours | +0.170 |
coconut | +0.164 |
Enjoy | +0.162 |
oral | +0.156 |
rust | +0.149 |
Neg logit contributions | |
aces | -0.147 |
coat | -0.143 |
amas | -0.140 |
forehead | -0.136 |
drops | -0.136 |
Token | . |
Token ID | 13 |
Feature activation | +2.63e-01 |
Pos logit contributions | |
aces | +0.089 |
coat | +0.087 |
amas | +0.087 |
robe | +0.083 |
drops | +0.083 |
Neg logit contributions | |
ours | -0.079 |
coconut | -0.077 |
Enjoy | -0.075 |
oral | -0.071 |
rust | -0.069 |
Token |
|
Token ID | 198 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
coconut | +0.018 |
ours | +0.017 |
rust | +0.016 |
oral | +0.015 |
Enjoy | +0.015 |
Neg logit contributions | |
aces | -0.022 |
amas | -0.021 |
coat | -0.021 |
stand | -0.021 |
robe | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | +9.67e-02 |
Pos logit contributions | |
ours | +0.007 |
coconut | +0.007 |
Enjoy | +0.006 |
rust | +0.006 |
iw | +0.006 |
Neg logit contributions | |
stand | -0.009 |
aces | -0.009 |
amas | -0.009 |
coat | -0.009 |
robe | -0.008 |
Token | very |
Token ID | 845 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
ours | +0.031 |
coconut | +0.030 |
Enjoy | +0.029 |
rust | +0.026 |
oral | +0.026 |
Neg logit contributions | |
aces | -0.035 |
amas | -0.035 |
coat | -0.034 |
robe | -0.033 |
drops | -0.033 |
Token | sad |
Token ID | 6507 |
Feature activation | +6.44e-01 |
Pos logit contributions | |
ours | +0.150 |
coconut | +0.148 |
Enjoy | +0.147 |
oral | +0.141 |
rust | +0.133 |
Neg logit contributions | |
aces | -0.137 |
forehead | -0.133 |
amas | -0.132 |
coat | -0.131 |
robe | -0.126 |
Token | . |
Token ID | 13 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
ours | +0.043 |
coconut | +0.041 |
Enjoy | +0.040 |
oral | +0.038 |
rust | +0.036 |
Neg logit contributions | |
aces | -0.053 |
amas | -0.052 |
coat | -0.052 |
forehead | -0.050 |
robe | -0.050 |
Token | |
Token ID | 220 |
Feature activation | -1.24e-02 |
Pos logit contributions | |
aces | +0.012 |
amas | +0.012 |
coat | +0.011 |
robe | +0.011 |
stand | +0.011 |
Neg logit contributions | |
ours | -0.010 |
coconut | -0.010 |
Enjoy | -0.009 |
oral | -0.009 |
rust | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | -2.16e-01 |
Pos logit contributions | |
aces | +0.001 |
amas | +0.001 |
coat | +0.001 |
robe | +0.001 |
forehead | +0.001 |
Neg logit contributions | |
ours | -0.001 |
coconut | -0.001 |
Enjoy | -0.001 |
oral | -0.000 |
rust | -0.000 |
Token |
|
Token ID | 198 |
Feature activation | -2.20e-01 |
Pos logit contributions | |
stand | +0.017 |
amas | +0.017 |
aces | +0.017 |
coat | +0.017 |
robe | +0.016 |
Neg logit contributions | |
ours | -0.015 |
coconut | -0.014 |
Enjoy | -0.013 |
rust | -0.012 |
oral | -0.012 |
Token | But |
Token ID | 1537 |
Feature activation | +8.13e-01 |
Pos logit contributions | |
aces | +0.017 |
amas | +0.017 |
coat | +0.017 |
robe | +0.016 |
forehead | +0.016 |
Neg logit contributions | |
ours | -0.016 |
coconut | -0.015 |
Enjoy | -0.015 |
oral | -0.014 |
rust | -0.014 |
Token | the |
Token ID | 262 |
Feature activation | -1.33e-01 |
Pos logit contributions | |
ours | +0.073 |
coconut | +0.071 |
Enjoy | +0.069 |
oral | +0.068 |
rust | +0.065 |
Neg logit contributions | |
aces | -0.049 |
forehead | -0.047 |
coat | -0.045 |
amas | -0.044 |
drops | -0.043 |
Token | princess |
Token ID | 21752 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
aces | +0.010 |
amas | +0.010 |
coat | +0.010 |
drops | +0.009 |
robe | +0.009 |
Neg logit contributions | |
coconut | -0.010 |
ours | -0.010 |
Enjoy | -0.009 |
oral | -0.009 |
rust | -0.009 |
Token | didn |
Token ID | 1422 |
Feature activation | +3.97e-02 |
Pos logit contributions | |
aces | +0.116 |
coat | +0.113 |
amas | +0.111 |
robe | +0.108 |
forehead | +0.108 |
Neg logit contributions | |
ours | -0.087 |
coconut | -0.084 |
Enjoy | -0.083 |
oral | -0.075 |
rust | -0.072 |
Token | 't |
Token ID | 470 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
ours | +0.003 |
coconut | +0.003 |
Enjoy | +0.003 |
oral | +0.003 |
rust | +0.003 |
Neg logit contributions | |
aces | -0.003 |
coat | -0.002 |
amas | -0.002 |
forehead | -0.002 |
drops | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | +7.71e-02 |
Pos logit contributions | |
aces | +0.263 |
coat | +0.257 |
amas | +0.257 |
robe | +0.247 |
forehead | +0.247 |
Neg logit contributions | |
ours | -0.157 |
coconut | -0.152 |
Enjoy | -0.150 |
oral | -0.136 |
rust | -0.132 |
Token | fairy |
Token ID | 25607 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
coconut | +0.005 |
ours | +0.005 |
Enjoy | +0.005 |
oral | +0.004 |
rust | +0.004 |
Neg logit contributions | |
aces | -0.007 |
amas | -0.006 |
coat | -0.006 |
robe | -0.006 |
drops | -0.006 |
Token | find |
Token ID | 1064 |
Feature activation | -3.02e-01 |
Pos logit contributions | |
amas | +0.134 |
aces | +0.133 |
coat | +0.132 |
drops | +0.127 |
robe | +0.124 |
Neg logit contributions | |
ours | -0.102 |
coconut | -0.099 |
Enjoy | -0.096 |
oral | -0.087 |
arium | -0.086 |
Token | her |
Token ID | 607 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
coat | +0.028 |
aces | +0.028 |
amas | +0.028 |
drops | +0.027 |
robe | +0.026 |
Neg logit contributions | |
ours | -0.017 |
coconut | -0.017 |
Enjoy | -0.016 |
oral | -0.014 |
rust | -0.014 |
Token | way |
Token ID | 835 |
Feature activation | -1.80e+00 |
Pos logit contributions | |
aces | +0.107 |
amas | +0.106 |
coat | +0.105 |
drops | +0.103 |
bridge | +0.100 |
Neg logit contributions | |
coconut | -0.079 |
ours | -0.078 |
Enjoy | -0.074 |
oral | -0.069 |
Sometimes | -0.067 |
Token | home |
Token ID | 1363 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
aces | +0.188 |
amas | +0.185 |
bridge | +0.180 |
coat | +0.180 |
drops | +0.176 |
Neg logit contributions | |
coconut | -0.079 |
ours | -0.075 |
Enjoy | -0.072 |
Sometimes | -0.066 |
arium | -0.063 |
Token | . |
Token ID | 13 |
Feature activation | +4.46e-01 |
Pos logit contributions | |
coconut | +0.048 |
ours | +0.048 |
Enjoy | +0.044 |
arium | +0.041 |
Sometimes | +0.040 |
Neg logit contributions | |
amas | -0.118 |
aces | -0.117 |
coat | -0.115 |
bridge | -0.112 |
cartridge | -0.112 |
Token | The |
Token ID | 383 |
Feature activation | +4.78e-01 |
Pos logit contributions | |
ours | +0.027 |
coconut | +0.026 |
Enjoy | +0.025 |
oral | +0.023 |
rust | +0.022 |
Neg logit contributions | |
aces | -0.040 |
coat | -0.039 |
amas | -0.039 |
forehead | -0.037 |
robe | -0.037 |
Token | fairy |
Token ID | 25607 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
coconut | +0.029 |
ours | +0.028 |
Enjoy | +0.028 |
oral | +0.027 |
rust | +0.025 |
Neg logit contributions | |
aces | -0.044 |
amas | -0.042 |
coat | -0.041 |
robe | -0.040 |
drops | -0.040 |
Token | was |
Token ID | 373 |
Feature activation | +6.79e-01 |
Pos logit contributions | |
aces | +0.147 |
amas | +0.147 |
coat | +0.144 |
drops | +0.139 |
robe | +0.138 |
Neg logit contributions | |
ours | -0.090 |
coconut | -0.088 |
Enjoy | -0.083 |
oral | -0.079 |
rust | -0.075 |
Token | grateful |
Token ID | 14066 |
Feature activation | -2.00e+00 |
Pos logit contributions | |
ours | +0.036 |
coconut | +0.034 |
Enjoy | +0.033 |
asks | +0.033 |
grown | +0.032 |
Neg logit contributions | |
amas | -0.064 |
drops | -0.062 |
coat | -0.062 |
aces | -0.061 |
robe | -0.059 |
Token | big |
Token ID | 1263 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
ours | +0.145 |
coconut | +0.141 |
Enjoy | +0.140 |
oral | +0.131 |
rust | +0.127 |
Neg logit contributions | |
aces | -0.151 |
coat | -0.148 |
amas | -0.145 |
robe | -0.142 |
drops | -0.141 |
Token | and |
Token ID | 290 |
Feature activation | +2.12e-01 |
Pos logit contributions | |
amas | +0.092 |
coat | +0.088 |
aces | +0.088 |
forehead | +0.086 |
robe | +0.084 |
Neg logit contributions | |
ours | -0.062 |
coconut | -0.059 |
Enjoy | -0.058 |
grown | -0.051 |
oral | -0.051 |
Token | fluffy |
Token ID | 39145 |
Feature activation | -2.75e+00 |
Pos logit contributions | |
ours | +0.016 |
coconut | +0.016 |
Enjoy | +0.015 |
oral | +0.014 |
rust | +0.014 |
Neg logit contributions | |
amas | -0.016 |
aces | -0.016 |
coat | -0.015 |
robe | -0.014 |
forehead | -0.014 |
Token | that |
Token ID | 326 |
Feature activation | -9.05e-01 |
Pos logit contributions | |
amas | +0.246 |
aces | +0.244 |
forehead | +0.237 |
coat | +0.237 |
robe | +0.228 |
Neg logit contributions | |
ours | -0.157 |
coconut | -0.152 |
Enjoy | -0.143 |
grown | -0.132 |
crab | -0.126 |
Token | looked |
Token ID | 3114 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
amas | +0.082 |
aces | +0.081 |
coat | +0.080 |
forehead | +0.078 |
drops | +0.076 |
Neg logit contributions | |
ours | -0.053 |
coconut | -0.052 |
Enjoy | -0.049 |
rust | -0.045 |
oral | -0.045 |
Token | like |
Token ID | 588 |
Feature activation | -2.45e+00 |
Pos logit contributions | |
ours | +0.149 |
coconut | +0.144 |
Enjoy | +0.143 |
oral | +0.137 |
rust | +0.132 |
Neg logit contributions | |
aces | -0.110 |
coat | -0.106 |
amas | -0.102 |
robe | -0.101 |
forehead | -0.100 |
Token | a |
Token ID | 257 |
Feature activation | -4.37e-01 |
Pos logit contributions | |
amas | +0.220 |
aces | +0.213 |
coat | +0.213 |
forehead | +0.208 |
drops | +0.204 |
Neg logit contributions | |
ours | -0.146 |
coconut | -0.143 |
Enjoy | -0.137 |
oral | -0.121 |
grown | -0.119 |
Token | bear |
Token ID | 6842 |
Feature activation | -1.77e+00 |
Pos logit contributions | |
aces | +0.029 |
coat | +0.028 |
amas | +0.028 |
robe | +0.027 |
forehead | +0.027 |
Neg logit contributions | |
ours | -0.036 |
coconut | -0.035 |
Enjoy | -0.034 |
oral | -0.032 |
rust | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | +2.02e-01 |
Pos logit contributions | |
amas | +0.154 |
aces | +0.154 |
coat | +0.150 |
forehead | +0.148 |
robe | +0.144 |
Neg logit contributions | |
ours | -0.107 |
coconut | -0.103 |
Enjoy | -0.101 |
oral | -0.089 |
rust | -0.087 |
Token |
|
Token ID | 198 |
Feature activation | +2.98e-02 |
Pos logit contributions | |
coconut | +0.014 |
ours | +0.014 |
oral | +0.013 |
Enjoy | +0.013 |
rous | +0.012 |
Neg logit contributions | |
robe | -0.015 |
aces | -0.015 |
stand | -0.015 |
amas | -0.015 |
cartridge | -0.014 |
Token |
|
Token ID | 198 |
Feature activation | +2.93e-02 |
Pos logit contributions | |
coconut | +0.002 |
ours | +0.002 |
Enjoy | +0.002 |
iw | +0.002 |
rust | +0.002 |
Neg logit contributions | |
amas | -0.002 |
stand | -0.002 |
aces | -0.002 |
robe | -0.002 |
coat | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +8.98e-02 |
Pos logit contributions | |
amas | +0.443 |
robe | +0.412 |
cartridge | +0.411 |
bridge | +0.403 |
aces | +0.403 |
Neg logit contributions | |
coconut | -0.134 |
: | -0.120 |
whose | -0.109 |
ours | -0.106 |
what | -0.105 |
Token | It |
Token ID | 632 |
Feature activation | -3.36e-01 |
Pos logit contributions | |
ours | +0.006 |
coconut | +0.006 |
Enjoy | +0.006 |
oral | +0.005 |
rust | +0.005 |
Neg logit contributions | |
aces | -0.008 |
amas | -0.007 |
coat | -0.007 |
robe | -0.007 |
forehead | -0.007 |
Token | was |
Token ID | 373 |
Feature activation | +6.82e-01 |
Pos logit contributions | |
aces | +0.030 |
amas | +0.030 |
coat | +0.029 |
robe | +0.028 |
stand | +0.028 |
Neg logit contributions | |
ours | -0.020 |
coconut | -0.020 |
Enjoy | -0.018 |
oral | -0.018 |
rust | -0.017 |
Token | scared |
Token ID | 12008 |
Feature activation | +7.36e-01 |
Pos logit contributions | |
coconut | +0.041 |
ours | +0.040 |
Enjoy | +0.038 |
grown | +0.036 |
what | +0.036 |
Neg logit contributions | |
amas | -0.060 |
aces | -0.060 |
robe | -0.057 |
coat | -0.057 |
stand | -0.055 |
Token | and |
Token ID | 290 |
Feature activation | +2.25e-01 |
Pos logit contributions | |
ours | +0.036 |
coconut | +0.034 |
Sometimes | +0.033 |
Enjoy | +0.033 |
grown | +0.032 |
Neg logit contributions | |
aces | -0.073 |
amas | -0.073 |
coat | -0.071 |
robe | -0.071 |
forehead | -0.069 |
Token | so |
Token ID | 523 |
Feature activation | -2.31e+00 |
Pos logit contributions | |
ours | +0.014 |
coconut | +0.013 |
Enjoy | +0.013 |
oral | +0.012 |
rust | +0.012 |
Neg logit contributions | |
aces | -0.020 |
amas | -0.019 |
coat | -0.019 |
robe | -0.018 |
drops | -0.018 |
Token | it |
Token ID | 340 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
aces | +0.210 |
amas | +0.206 |
coat | +0.206 |
robe | +0.198 |
forehead | +0.198 |
Neg logit contributions | |
ours | -0.134 |
coconut | -0.130 |
Enjoy | -0.127 |
oral | -0.116 |
rust | -0.112 |
Token | started |
Token ID | 2067 |
Feature activation | +4.27e-01 |
Pos logit contributions | |
ours | +0.088 |
coconut | +0.087 |
Enjoy | +0.083 |
oral | +0.078 |
rust | +0.076 |
Neg logit contributions | |
aces | -0.121 |
amas | -0.121 |
coat | -0.118 |
robe | -0.114 |
stand | -0.113 |
Token | to |
Token ID | 284 |
Feature activation | +5.90e-01 |
Pos logit contributions | |
ours | +0.022 |
coconut | +0.021 |
Enjoy | +0.020 |
oral | +0.018 |
rust | +0.018 |
Neg logit contributions | |
aces | -0.042 |
amas | -0.041 |
coat | -0.041 |
robe | -0.040 |
forehead | -0.039 |
Token | hope |
Token ID | 2911 |
Feature activation | -1.96e+00 |
Pos logit contributions | |
coconut | +0.042 |
ours | +0.040 |
Enjoy | +0.039 |
swamp | +0.037 |
Sometimes | +0.036 |
Neg logit contributions | |
aces | -0.047 |
amas | -0.045 |
robe | -0.043 |
coat | -0.043 |
bridge | -0.041 |
Token | that |
Token ID | 326 |
Feature activation | -8.17e-01 |
Pos logit contributions | |
amas | +0.200 |
aces | +0.200 |
coat | +0.198 |
robe | +0.192 |
forehead | +0.187 |
Neg logit contributions | |
ours | -0.087 |
coconut | -0.087 |
Enjoy | -0.083 |
oral | -0.076 |
Sometimes | -0.075 |
Token | I |
Token ID | 314 |
Feature activation | -6.35e-01 |
Pos logit contributions | |
Enjoy | +0.012 |
ours | +0.011 |
coconut | +0.010 |
asks | +0.010 |
Joe | +0.009 |
Neg logit contributions | |
coat | -0.024 |
aces | -0.024 |
amas | -0.024 |
robe | -0.023 |
amed | -0.023 |
Token | know |
Token ID | 760 |
Feature activation | +2.32e+00 |
Pos logit contributions | |
aces | +0.055 |
coat | +0.054 |
robe | +0.052 |
amas | +0.052 |
drops | +0.052 |
Neg logit contributions | |
coconut | -0.039 |
ours | -0.039 |
Enjoy | -0.036 |
rust | -0.034 |
oral | -0.034 |
Token | that |
Token ID | 326 |
Feature activation | -8.69e-01 |
Pos logit contributions | |
ours | +0.122 |
coconut | +0.119 |
Enjoy | +0.111 |
oral | +0.108 |
rust | +0.105 |
Neg logit contributions | |
aces | -0.223 |
amas | -0.215 |
coat | -0.214 |
forehead | -0.209 |
robe | -0.206 |
Token | you |
Token ID | 345 |
Feature activation | +1.76e-02 |
Pos logit contributions | |
aces | +0.074 |
amas | +0.072 |
coat | +0.071 |
robe | +0.069 |
forehead | +0.069 |
Neg logit contributions | |
ours | -0.055 |
coconut | -0.053 |
Enjoy | -0.052 |
oral | -0.049 |
rust | -0.048 |
Token | can |
Token ID | 460 |
Feature activation | +6.59e-01 |
Pos logit contributions | |
ours | +0.001 |
coconut | +0.001 |
Enjoy | +0.001 |
oral | +0.001 |
rust | +0.001 |
Neg logit contributions | |
aces | -0.002 |
coat | -0.001 |
amas | -0.001 |
robe | -0.001 |
forehead | -0.001 |
Token | make |
Token ID | 787 |
Feature activation | -3.30e+00 |
Pos logit contributions | |
ours | +0.046 |
coconut | +0.045 |
Enjoy | +0.044 |
oral | +0.041 |
rust | +0.040 |
Neg logit contributions | |
aces | -0.052 |
amas | -0.050 |
coat | -0.050 |
robe | -0.048 |
forehead | -0.048 |
Token | something |
Token ID | 1223 |
Feature activation | +2.13e+00 |
Pos logit contributions | |
aces | +0.298 |
coat | +0.294 |
amas | +0.291 |
robe | +0.282 |
drops | +0.282 |
Neg logit contributions | |
ours | -0.194 |
coconut | -0.186 |
Enjoy | -0.182 |
oral | -0.166 |
rust | -0.162 |
Token | beautiful |
Token ID | 4950 |
Feature activation | -9.00e-01 |
Pos logit contributions | |
ours | +0.147 |
coconut | +0.143 |
Enjoy | +0.140 |
oral | +0.130 |
rust | +0.126 |
Neg logit contributions | |
aces | -0.170 |
amas | -0.166 |
coat | -0.165 |
robe | -0.158 |
forehead | -0.158 |
Token | out |
Token ID | 503 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
aces | +0.079 |
coat | +0.079 |
amas | +0.079 |
forehead | +0.077 |
drops | +0.075 |
Neg logit contributions | |
ours | -0.052 |
Enjoy | -0.051 |
coconut | -0.050 |
oral | -0.045 |
Sometimes | -0.043 |
Token | of |
Token ID | 286 |
Feature activation | +1.67e-01 |
Pos logit contributions | |
aces | +0.124 |
amas | +0.121 |
coat | +0.121 |
robe | +0.117 |
forehead | +0.116 |
Neg logit contributions | |
ours | -0.087 |
coconut | -0.084 |
Enjoy | -0.082 |
oral | -0.076 |
rust | -0.074 |
Token | something |
Token ID | 1223 |
Feature activation | +2.13e+00 |
Pos logit contributions | |
ours | +0.010 |
coconut | +0.009 |
Enjoy | +0.009 |
oral | +0.008 |
rust | +0.008 |
Neg logit contributions | |
aces | -0.015 |
amas | -0.015 |
coat | -0.015 |
forehead | -0.014 |
robe | -0.014 |
Token | � |
Token ID | 6184 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
aces | +0.008 |
coat | +0.008 |
amas | +0.008 |
robe | +0.008 |
forehead | +0.008 |
Neg logit contributions | |
ours | -0.006 |
coconut | -0.006 |
Enjoy | -0.006 |
oral | -0.005 |
rust | -0.005 |
Token | � |
Token ID | 95 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
ours | +0.146 |
coconut | +0.142 |
Enjoy | +0.139 |
oral | +0.130 |
rust | +0.127 |
Neg logit contributions | |
aces | -0.150 |
amas | -0.147 |
coat | -0.146 |
robe | -0.140 |
forehead | -0.140 |
Token | € |
Token ID | 26391 |
Feature activation | -2.00e+00 |
Pos logit contributions | |
aces | +0.096 |
coat | +0.096 |
forehead | +0.095 |
amas | +0.095 |
robe | +0.092 |
Neg logit contributions | |
Enjoy | -0.077 |
ours | -0.075 |
oral | -0.068 |
coconut | -0.066 |
Sometimes | -0.063 |
Token | � |
Token ID | 129 |
Feature activation | -2.42e+00 |
Pos logit contributions | |
forehead | +0.169 |
aces | +0.163 |
robe | +0.156 |
coat | +0.156 |
cartridge | +0.153 |
Neg logit contributions | |
ours | -0.133 |
Enjoy | -0.128 |
coconut | -0.116 |
Sometimes | -0.112 |
oral | -0.112 |
Token | � |
Token ID | 241 |
Feature activation | -4.22e+00 |
Pos logit contributions | |
aces | +0.177 |
robe | +0.177 |
coat | +0.176 |
amas | +0.174 |
forehead | +0.168 |
Neg logit contributions | |
ours | -0.176 |
Enjoy | -0.175 |
coconut | -0.164 |
oral | -0.156 |
rust | -0.153 |
Token | The |
Token ID | 464 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
coat | +0.335 |
aces | +0.330 |
amas | +0.317 |
drops | +0.315 |
stand | +0.312 |
Neg logit contributions | |
ours | -0.297 |
coconut | -0.282 |
Enjoy | -0.261 |
oral | -0.255 |
rust | -0.254 |
Token | mountains |
Token ID | 12269 |
Feature activation | -3.54e-01 |
Pos logit contributions | |
coat | +0.103 |
amas | +0.103 |
aces | +0.102 |
forehead | +0.098 |
drops | +0.097 |
Neg logit contributions | |
Enjoy | -0.060 |
coconut | -0.058 |
ours | -0.058 |
oral | -0.056 |
temple | -0.051 |
Token | look |
Token ID | 804 |
Feature activation | +3.89e-01 |
Pos logit contributions | |
amas | +0.034 |
aces | +0.033 |
coat | +0.033 |
forehead | +0.032 |
stand | +0.032 |
Neg logit contributions | |
ours | -0.019 |
coconut | -0.018 |
Enjoy | -0.017 |
oral | -0.016 |
Sometimes | -0.014 |
Token | very |
Token ID | 845 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
ours | +0.025 |
Enjoy | +0.024 |
coconut | +0.023 |
Sometimes | +0.021 |
grown | +0.021 |
Neg logit contributions | |
amas | -0.034 |
coat | -0.033 |
aces | -0.033 |
forehead | -0.032 |
robe | -0.032 |
Token | clear |
Token ID | 1598 |
Feature activation | -1.00e+00 |
Pos logit contributions | |
ours | +0.161 |
coconut | +0.158 |
Enjoy | +0.155 |
oral | +0.150 |
rust | +0.144 |
Neg logit contributions | |
aces | -0.130 |
amas | -0.125 |
coat | -0.124 |
drops | -0.122 |
forehead | -0.120 |
Token | today |
Token ID | 1909 |
Feature activation | +1.58e+00 |
Pos logit contributions | |
coat | +0.092 |
amas | +0.092 |
aces | +0.091 |
forehead | +0.089 |
robe | +0.089 |
Neg logit contributions | |
ours | -0.056 |
coconut | -0.053 |
Enjoy | -0.052 |
grown | -0.046 |
Sometimes | -0.045 |
Token | used |
Token ID | 973 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
aces | +0.019 |
coat | +0.018 |
amas | +0.018 |
forehead | +0.018 |
drops | +0.017 |
Neg logit contributions | |
ours | -0.016 |
coconut | -0.015 |
Enjoy | -0.015 |
oral | -0.014 |
rust | -0.013 |
Token | the |
Token ID | 262 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
coconut | +0.047 |
ours | +0.045 |
Enjoy | +0.044 |
Sometimes | +0.034 |
rust | +0.034 |
Neg logit contributions | |
amas | -0.127 |
aces | -0.125 |
coat | -0.124 |
drops | -0.124 |
robe | -0.122 |
Token | flashlight |
Token ID | 38371 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
coconut | +0.008 |
Enjoy | +0.007 |
ours | +0.007 |
oral | +0.007 |
question | +0.007 |
Neg logit contributions | |
amas | -0.009 |
aces | -0.009 |
coat | -0.009 |
robe | -0.009 |
drops | -0.009 |
Token | to |
Token ID | 284 |
Feature activation | +4.80e-01 |
Pos logit contributions | |
amas | +0.109 |
aces | +0.104 |
coat | +0.101 |
bridge | +0.100 |
bes | +0.100 |
Neg logit contributions | |
coconut | -0.041 |
ours | -0.038 |
fired | -0.034 |
Enjoy | -0.033 |
arium | -0.032 |
Token | show |
Token ID | 905 |
Feature activation | +1.60e+00 |
Pos logit contributions | |
coconut | +0.030 |
Enjoy | +0.029 |
ours | +0.027 |
oral | +0.025 |
temple | +0.025 |
Neg logit contributions | |
aces | -0.042 |
amas | -0.041 |
coat | -0.039 |
robe | -0.039 |
drops | -0.038 |
Token | him |
Token ID | 683 |
Feature activation | -1.70e+00 |
Pos logit contributions | |
ours | +0.090 |
Enjoy | +0.090 |
coconut | +0.089 |
oral | +0.078 |
rust | +0.075 |
Neg logit contributions | |
coat | -0.148 |
amas | -0.145 |
aces | -0.145 |
drops | -0.139 |
robe | -0.138 |
Token | the |
Token ID | 262 |
Feature activation | +4.63e-02 |
Pos logit contributions | |
aces | +0.159 |
coat | +0.149 |
amas | +0.149 |
forehead | +0.149 |
robe | +0.147 |
Neg logit contributions | |
ours | -0.096 |
coconut | -0.090 |
Enjoy | -0.089 |
oral | -0.084 |
rust | -0.082 |
Token | way |
Token ID | 835 |
Feature activation | -1.80e+00 |
Pos logit contributions | |
ours | +0.003 |
coconut | +0.003 |
Enjoy | +0.003 |
oral | +0.003 |
rust | +0.003 |
Neg logit contributions | |
aces | -0.004 |
amas | -0.004 |
coat | -0.004 |
robe | -0.003 |
forehead | -0.003 |
Token | home |
Token ID | 1363 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
aces | +0.162 |
amas | +0.159 |
coat | +0.155 |
robe | +0.150 |
drops | +0.148 |
Neg logit contributions | |
coconut | -0.107 |
ours | -0.104 |
Enjoy | -0.100 |
Joe | -0.091 |
Sometimes | -0.090 |
Token | . |
Token ID | 13 |
Feature activation | +1.58e-02 |
Pos logit contributions | |
coconut | +0.055 |
ours | +0.054 |
Enjoy | +0.050 |
Joe | +0.046 |
temple | +0.045 |
Neg logit contributions | |
amas | -0.112 |
aces | -0.110 |
coat | -0.110 |
cartridge | -0.106 |
forehead | -0.105 |
Token | |
Token ID | 220 |
Feature activation | +5.08e-02 |
Pos logit contributions | |
ours | +0.001 |
coconut | +0.001 |
rust | +0.001 |
oral | +0.001 |
rous | +0.001 |
Neg logit contributions | |
amas | -0.001 |
aces | -0.001 |
stand | -0.001 |
robe | -0.001 |
bridge | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +2.66e-01 |
Pos logit contributions | |
aces | +0.162 |
amas | +0.162 |
coat | +0.154 |
robe | +0.152 |
forehead | +0.151 |
Neg logit contributions | |
ours | -0.102 |
coconut | -0.096 |
Enjoy | -0.092 |
Sometimes | -0.083 |
iw | -0.082 |
Token | I |
Token ID | 314 |
Feature activation | -5.83e-01 |
Pos logit contributions | |
ours | +0.017 |
coconut | +0.016 |
Enjoy | +0.015 |
oral | +0.015 |
rust | +0.014 |
Neg logit contributions | |
aces | -0.023 |
amas | -0.022 |
coat | -0.022 |
robe | -0.021 |
forehead | -0.021 |
Token | have |
Token ID | 423 |
Feature activation | +2.27e+00 |
Pos logit contributions | |
coat | +0.046 |
aces | +0.042 |
robe | +0.042 |
drops | +0.042 |
brush | +0.040 |
Neg logit contributions | |
ours | -0.042 |
coconut | -0.040 |
oral | -0.037 |
Enjoy | -0.036 |
rust | -0.036 |
Token | some |
Token ID | 617 |
Feature activation | +5.04e-01 |
Pos logit contributions | |
ours | +0.144 |
coconut | +0.139 |
Enjoy | +0.136 |
oral | +0.128 |
rust | +0.123 |
Neg logit contributions | |
aces | -0.195 |
coat | -0.189 |
amas | -0.189 |
forehead | -0.182 |
robe | -0.181 |
Token | band |
Token ID | 4097 |
Feature activation | -1.75e+00 |
Pos logit contributions | |
ours | +0.037 |
coconut | +0.035 |
Enjoy | +0.035 |
oral | +0.033 |
rust | +0.032 |
Neg logit contributions | |
aces | -0.039 |
coat | -0.038 |
amas | -0.037 |
forehead | -0.036 |
robe | -0.036 |
Token | ages |
Token ID | 1095 |
Feature activation | -7.63e+00 |
Pos logit contributions | |
amas | +0.128 |
cartridge | +0.114 |
forehead | +0.113 |
aces | +0.109 |
coat | +0.107 |
Neg logit contributions | |
Enjoy | -0.133 |
ours | -0.128 |
coconut | -0.117 |
iw | -0.115 |
rill | -0.114 |
Token | and |
Token ID | 290 |
Feature activation | +2.55e-01 |
Pos logit contributions | |
amas | +0.700 |
aces | +0.684 |
coat | +0.675 |
forehead | +0.652 |
drops | +0.651 |
Neg logit contributions | |
ours | -0.430 |
Enjoy | -0.416 |
coconut | -0.402 |
oral | -0.347 |
iw | -0.344 |
Token | some |
Token ID | 617 |
Feature activation | +4.98e-01 |
Pos logit contributions | |
ours | +0.017 |
coconut | +0.016 |
Enjoy | +0.016 |
oral | +0.015 |
rust | +0.014 |
Neg logit contributions | |
aces | -0.021 |
amas | -0.021 |
coat | -0.021 |
robe | -0.020 |
drops | -0.020 |
Token | medicine |
Token ID | 9007 |
Feature activation | -9.78e-01 |
Pos logit contributions | |
ours | +0.035 |
coconut | +0.034 |
Enjoy | +0.034 |
oral | +0.031 |
rust | +0.030 |
Neg logit contributions | |
aces | -0.039 |
amas | -0.038 |
coat | -0.038 |
robe | -0.036 |
forehead | -0.036 |
Token | . |
Token ID | 13 |
Feature activation | +2.69e-01 |
Pos logit contributions | |
amas | +0.087 |
aces | +0.085 |
coat | +0.085 |
forehead | +0.082 |
drops | +0.081 |
Neg logit contributions | |
ours | -0.058 |
Enjoy | -0.055 |
coconut | -0.054 |
oral | -0.049 |
arium | -0.046 |
Token | I |
Token ID | 314 |
Feature activation | -5.90e-01 |
Pos logit contributions | |
ours | +0.018 |
coconut | +0.017 |
Enjoy | +0.016 |
oral | +0.016 |
rust | +0.016 |
Neg logit contributions | |
aces | -0.022 |
amas | -0.022 |
coat | -0.021 |
forehead | -0.021 |
robe | -0.021 |
Token | made |
Token ID | 925 |
Feature activation | -4.35e+00 |
Pos logit contributions | |
aces | +0.119 |
amas | +0.119 |
forehead | +0.115 |
coat | +0.113 |
drops | +0.110 |
Neg logit contributions | |
coconut | -0.056 |
ours | -0.056 |
arose | -0.048 |
Enjoy | -0.047 |
asks | -0.047 |
Token | a |
Token ID | 257 |
Feature activation | -4.59e-01 |
Pos logit contributions | |
coat | +0.473 |
aces | +0.469 |
amas | +0.454 |
drops | +0.443 |
robe | +0.441 |
Neg logit contributions | |
ours | -0.181 |
coconut | -0.173 |
Enjoy | -0.168 |
Joe | -0.151 |
Sometimes | -0.148 |
Token | loud |
Token ID | 7812 |
Feature activation | -9.47e-01 |
Pos logit contributions | |
aces | +0.034 |
coat | +0.031 |
amas | +0.030 |
robe | +0.030 |
drops | +0.029 |
Neg logit contributions | |
coconut | -0.036 |
ours | -0.036 |
Enjoy | -0.034 |
oral | -0.033 |
rust | -0.032 |
Token | wh |
Token ID | 348 |
Feature activation | -2.58e+00 |
Pos logit contributions | |
aces | +0.073 |
coat | +0.068 |
robe | +0.067 |
amas | +0.064 |
cartridge | +0.063 |
Neg logit contributions | |
ours | -0.068 |
coconut | -0.067 |
Enjoy | -0.065 |
volcano | -0.063 |
question | -0.062 |
Token | ir |
Token ID | 343 |
Feature activation | +3.84e+00 |
Pos logit contributions | |
aces | +0.209 |
amas | +0.205 |
amed | +0.194 |
coat | +0.193 |
bridge | +0.193 |
Neg logit contributions | |
Enjoy | -0.183 |
coconut | -0.176 |
ours | -0.162 |
Crab | -0.154 |
Sometimes | -0.153 |
Token | ring |
Token ID | 1806 |
Feature activation | -6.11e+00 |
Pos logit contributions | |
iles | +0.260 |
: | +0.259 |
oral | +0.257 |
Joe | +0.255 |
coconut | +0.253 |
Neg logit contributions | |
coat | -0.262 |
cartridge | -0.260 |
robe | -0.256 |
amas | -0.252 |
bes | -0.228 |
Token | sound |
Token ID | 2128 |
Feature activation | -2.93e+00 |
Pos logit contributions | |
aces | +0.510 |
cartridge | +0.503 |
coat | +0.480 |
amas | +0.476 |
robe | +0.469 |
Neg logit contributions | |
ours | -0.338 |
arose | -0.320 |
volcano | -0.306 |
coconut | -0.305 |
Sometimes | -0.300 |
Token | . |
Token ID | 13 |
Feature activation | +2.68e-02 |
Pos logit contributions | |
amas | +0.334 |
aces | +0.323 |
coat | +0.322 |
robe | +0.321 |
cartridge | +0.312 |
Neg logit contributions | |
what | -0.085 |
: | -0.084 |
Joe | -0.080 |
why | -0.077 |
ours | -0.075 |
Token | He |
Token ID | 679 |
Feature activation | -2.40e-01 |
Pos logit contributions | |
ours | +0.002 |
coconut | +0.002 |
oral | +0.002 |
rous | +0.002 |
rust | +0.002 |
Neg logit contributions | |
amas | -0.002 |
aces | -0.002 |
stand | -0.002 |
robe | -0.002 |
coat | -0.002 |
Token | wanted |
Token ID | 2227 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
aces | +0.018 |
coat | +0.018 |
drops | +0.018 |
amas | +0.018 |
forehead | +0.017 |
Neg logit contributions | |
Enjoy | -0.017 |
ours | -0.017 |
coconut | -0.016 |
oral | -0.016 |
iw | -0.015 |
Token | to |
Token ID | 284 |
Feature activation | +6.14e-01 |
Pos logit contributions | |
ours | +0.097 |
coconut | +0.090 |
Enjoy | +0.089 |
oral | +0.089 |
rust | +0.084 |
Neg logit contributions | |
aces | -0.124 |
coat | -0.122 |
stand | -0.119 |
amas | -0.119 |
forehead | -0.117 |
Token | . |
Token ID | 13 |
Feature activation | +1.61e-01 |
Pos logit contributions | |
ours | +0.190 |
coconut | +0.184 |
Enjoy | +0.179 |
oral | +0.167 |
rust | +0.159 |
Neg logit contributions | |
aces | -0.295 |
amas | -0.289 |
coat | -0.286 |
robe | -0.278 |
forehead | -0.274 |
Token | He |
Token ID | 679 |
Feature activation | -4.54e-02 |
Pos logit contributions | |
coconut | +0.011 |
ours | +0.011 |
Enjoy | +0.011 |
oral | +0.010 |
rust | +0.010 |
Neg logit contributions | |
aces | -0.013 |
amas | -0.012 |
forehead | -0.012 |
robe | -0.012 |
coat | -0.012 |
Token | snaps |
Token ID | 23429 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
aces | +0.003 |
coat | +0.003 |
drops | +0.003 |
forehead | +0.003 |
amas | +0.003 |
Neg logit contributions | |
coconut | -0.003 |
oral | -0.003 |
Enjoy | -0.003 |
ours | -0.003 |
rust | -0.003 |
Token | the |
Token ID | 262 |
Feature activation | +5.35e-02 |
Pos logit contributions | |
ours | +0.110 |
coconut | +0.107 |
Enjoy | +0.106 |
oral | +0.099 |
rust | +0.094 |
Neg logit contributions | |
aces | -0.182 |
coat | -0.175 |
amas | -0.175 |
robe | -0.170 |
forehead | -0.170 |
Token | snow |
Token ID | 6729 |
Feature activation | -6.74e-01 |
Pos logit contributions | |
ours | +0.003 |
coconut | +0.003 |
Enjoy | +0.003 |
oral | +0.003 |
rust | +0.003 |
Neg logit contributions | |
aces | -0.005 |
amas | -0.004 |
coat | -0.004 |
drops | -0.004 |
robe | -0.004 |
Token | man |
Token ID | 805 |
Feature activation | -7.45e+00 |
Pos logit contributions | |
aces | +0.045 |
coat | +0.044 |
amas | +0.043 |
stand | +0.041 |
cartridge | +0.040 |
Neg logit contributions | |
coconut | -0.054 |
ours | -0.051 |
volcano | -0.050 |
crab | -0.046 |
crocod | -0.046 |
Token | 's |
Token ID | 338 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
aces | +0.597 |
coat | +0.592 |
amas | +0.591 |
robe | +0.574 |
drops | +0.553 |
Neg logit contributions | |
ours | -0.508 |
coconut | -0.500 |
Enjoy | -0.479 |
oral | -0.437 |
rust | -0.425 |
Token | arm |
Token ID | 3211 |
Feature activation | -6.21e-01 |
Pos logit contributions | |
aces | +0.083 |
coat | +0.082 |
amas | +0.082 |
drops | +0.078 |
robe | +0.078 |
Neg logit contributions | |
coconut | -0.103 |
ours | -0.101 |
Enjoy | -0.095 |
oral | -0.091 |
iw | -0.086 |
Token | . |
Token ID | 13 |
Feature activation | +1.61e-01 |
Pos logit contributions | |
amas | +0.063 |
coat | +0.061 |
robe | +0.060 |
forehead | +0.059 |
cartridge | +0.059 |
Neg logit contributions | |
ours | -0.025 |
coconut | -0.024 |
Sometimes | -0.021 |
whose | -0.019 |
: | -0.019 |
Token | He |
Token ID | 679 |
Feature activation | -3.44e-02 |
Pos logit contributions | |
coconut | +0.011 |
ours | +0.011 |
Enjoy | +0.010 |
oral | +0.010 |
rust | +0.010 |
Neg logit contributions | |
aces | -0.013 |
amas | -0.013 |
robe | -0.012 |
forehead | -0.012 |
coat | -0.012 |
Token | breaks |
Token ID | 9457 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
aces | +0.003 |
drops | +0.002 |
coat | +0.002 |
forehead | +0.002 |
amas | +0.002 |
Neg logit contributions | |
oral | -0.002 |
coconut | -0.002 |
Enjoy | -0.002 |
ours | -0.002 |
rust | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +3.18e-02 |
Pos logit contributions | |
amas | +0.157 |
aces | +0.155 |
forehead | +0.151 |
bridge | +0.149 |
bes | +0.148 |
Neg logit contributions | |
ours | -0.060 |
arose | -0.054 |
coconut | -0.053 |
Joe | -0.052 |
Enjoy | -0.050 |
Token | She |
Token ID | 1375 |
Feature activation | -4.12e-01 |
Pos logit contributions | |
ours | +0.002 |
coconut | +0.002 |
Enjoy | +0.002 |
rust | +0.002 |
oral | +0.002 |
Neg logit contributions | |
amas | -0.002 |
aces | -0.002 |
coat | -0.002 |
forehead | -0.002 |
drops | -0.002 |
Token | saw |
Token ID | 2497 |
Feature activation | -4.69e-01 |
Pos logit contributions | |
aces | +0.033 |
coat | +0.033 |
amas | +0.032 |
forehead | +0.031 |
robe | +0.031 |
Neg logit contributions | |
ours | -0.028 |
Enjoy | -0.027 |
coconut | -0.027 |
oral | -0.025 |
rust | -0.023 |
Token | the |
Token ID | 262 |
Feature activation | -1.81e-01 |
Pos logit contributions | |
aces | +0.042 |
amas | +0.042 |
coat | +0.042 |
forehead | +0.040 |
robe | +0.040 |
Neg logit contributions | |
ours | -0.028 |
coconut | -0.027 |
Enjoy | -0.026 |
oral | -0.024 |
rust | -0.023 |
Token | fire |
Token ID | 2046 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
aces | +0.017 |
coat | +0.016 |
amas | +0.016 |
robe | +0.016 |
stand | +0.016 |
Neg logit contributions | |
ours | -0.010 |
coconut | -0.010 |
Enjoy | -0.010 |
oral | -0.009 |
Sometimes | -0.008 |
Token | man |
Token ID | 805 |
Feature activation | -7.81e+00 |
Pos logit contributions | |
ours | +0.113 |
coconut | +0.109 |
Enjoy | +0.103 |
oral | +0.099 |
volcano | +0.098 |
Neg logit contributions | |
aces | -0.120 |
coat | -0.115 |
amas | -0.112 |
forehead | -0.107 |
stand | -0.107 |
Token | working |
Token ID | 1762 |
Feature activation | -3.84e-01 |
Pos logit contributions | |
aces | +0.636 |
coat | +0.620 |
amas | +0.620 |
robe | +0.595 |
forehead | +0.590 |
Neg logit contributions | |
ours | -0.528 |
coconut | -0.512 |
Enjoy | -0.500 |
oral | -0.466 |
rust | -0.450 |
Token | hard |
Token ID | 1327 |
Feature activation | +2.86e+00 |
Pos logit contributions | |
aces | +0.033 |
amas | +0.032 |
forehead | +0.032 |
coat | +0.032 |
robe | +0.031 |
Neg logit contributions | |
ours | -0.023 |
coconut | -0.022 |
Enjoy | -0.021 |
grown | -0.020 |
oral | -0.019 |
Token | to |
Token ID | 284 |
Feature activation | +3.75e-01 |
Pos logit contributions | |
ours | +0.154 |
Enjoy | +0.139 |
coconut | +0.139 |
Sometimes | +0.131 |
grown | +0.131 |
Neg logit contributions | |
aces | -0.260 |
amas | -0.256 |
coat | -0.256 |
forehead | -0.255 |
bridge | -0.249 |
Token | put |
Token ID | 1234 |
Feature activation | -1.72e+00 |
Pos logit contributions | |
coconut | +0.027 |
ours | +0.027 |
Enjoy | +0.027 |
oral | +0.024 |
rust | +0.024 |
Neg logit contributions | |
aces | -0.029 |
amas | -0.028 |
coat | -0.027 |
robe | -0.026 |
forehead | -0.026 |
Token | out |
Token ID | 503 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
aces | +0.151 |
amas | +0.150 |
coat | +0.148 |
forehead | +0.146 |
drops | +0.142 |
Neg logit contributions | |
ours | -0.101 |
coconut | -0.096 |
Enjoy | -0.096 |
Sometimes | -0.083 |
grown | -0.083 |
Token | water |
Token ID | 1660 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
ours | +0.272 |
coconut | +0.265 |
Enjoy | +0.260 |
oral | +0.244 |
rust | +0.237 |
Neg logit contributions | |
aces | -0.281 |
coat | -0.273 |
amas | -0.273 |
robe | -0.261 |
forehead | -0.261 |
Token | and |
Token ID | 290 |
Feature activation | +2.62e-01 |
Pos logit contributions | |
ours | +0.065 |
coconut | +0.060 |
Enjoy | +0.056 |
oral | +0.051 |
iw | +0.050 |
Neg logit contributions | |
amas | -0.094 |
aces | -0.091 |
forehead | -0.089 |
coat | -0.087 |
bridge | -0.086 |
Token | make |
Token ID | 787 |
Feature activation | -3.12e+00 |
Pos logit contributions | |
ours | +0.018 |
coconut | +0.017 |
Enjoy | +0.017 |
oral | +0.015 |
rust | +0.015 |
Neg logit contributions | |
amas | -0.021 |
aces | -0.021 |
coat | -0.021 |
robe | -0.020 |
forehead | -0.020 |
Token | more |
Token ID | 517 |
Feature activation | +3.71e+00 |
Pos logit contributions | |
aces | +0.268 |
coat | +0.264 |
amas | +0.263 |
robe | +0.251 |
forehead | +0.251 |
Neg logit contributions | |
ours | -0.200 |
coconut | -0.189 |
Enjoy | -0.185 |
oral | -0.170 |
rust | -0.169 |
Token | rain |
Token ID | 6290 |
Feature activation | -1.22e+00 |
Pos logit contributions | |
ours | +0.310 |
coconut | +0.303 |
Enjoy | +0.298 |
oral | +0.281 |
rust | +0.275 |
Neg logit contributions | |
aces | -0.243 |
amas | -0.236 |
coat | -0.235 |
robe | -0.224 |
forehead | -0.223 |
Token | bows |
Token ID | 25435 |
Feature activation | -7.62e+00 |
Pos logit contributions | |
amas | +0.065 |
aces | +0.063 |
cartridge | +0.059 |
forehead | +0.058 |
bridge | +0.055 |
Neg logit contributions | |
ours | -0.114 |
Sometimes | -0.101 |
coconut | -0.100 |
swamp | -0.098 |
temple | -0.097 |
Token | . |
Token ID | 13 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
amas | +0.853 |
coat | +0.842 |
forehead | +0.819 |
aces | +0.814 |
robe | +0.790 |
Neg logit contributions | |
ours | -0.260 |
coconut | -0.255 |
volcano | -0.222 |
temple | -0.214 |
Enjoy | -0.200 |
Token | They |
Token ID | 1119 |
Feature activation | +1.27e+00 |
Pos logit contributions | |
coconut | +0.013 |
ours | +0.013 |
oral | +0.011 |
Enjoy | +0.011 |
rous | +0.011 |
Neg logit contributions | |
aces | -0.016 |
robe | -0.016 |
amas | -0.016 |
stand | -0.015 |
coat | -0.015 |
Token | laugh |
Token ID | 6487 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
ours | +0.089 |
coconut | +0.087 |
Enjoy | +0.085 |
oral | +0.079 |
rust | +0.077 |
Neg logit contributions | |
aces | -0.100 |
coat | -0.098 |
amas | -0.097 |
forehead | -0.094 |
robe | -0.094 |
Token | and |
Token ID | 290 |
Feature activation | +2.59e-01 |
Pos logit contributions | |
ours | +0.004 |
Enjoy | +0.004 |
Sometimes | +0.004 |
what | +0.004 |
: | +0.004 |
Neg logit contributions | |
coat | -0.013 |
amas | -0.013 |
aces | -0.013 |
drops | -0.012 |
bes | -0.012 |
Token | play |
Token ID | 711 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
Enjoy | +0.015 |
ours | +0.015 |
coconut | +0.015 |
asks | +0.013 |
rust | +0.013 |
Neg logit contributions | |
amas | -0.023 |
aces | -0.023 |
coat | -0.022 |
bridge | -0.022 |
robe | -0.021 |