Token | says |
Token ID | 1139 |
Feature activation | +1.52e-02 |
Pos logit contributions | |
aling | +0.023 |
alling | +0.021 |
irl | +0.019 |
astics | +0.019 |
ophone | +0.018 |
Neg logit contributions | |
guard | -0.022 |
poisoning | -0.022 |
injuries | -0.021 |
payoff | -0.020 |
last | -0.020 |
Token | , |
Token ID | 11 |
Feature activation | +2.10e-02 |
Pos logit contributions | |
guard | +0.015 |
poisoning | +0.014 |
injuries | +0.014 |
payoff | +0.012 |
unfortunately | +0.012 |
Neg logit contributions | |
aling | -0.032 |
alling | -0.029 |
astics | -0.027 |
irl | -0.027 |
sings | -0.027 |
Token | sad |
Token ID | 6507 |
Feature activation | +4.76e-02 |
Pos logit contributions | |
guard | +0.031 |
poisoning | +0.030 |
injuries | +0.030 |
payoff | +0.027 |
unfortunately | +0.027 |
Neg logit contributions | |
aling | -0.034 |
alling | -0.031 |
astics | -0.028 |
irl | -0.028 |
sings | -0.027 |
Token | . |
Token ID | 13 |
Feature activation | -2.94e-02 |
Pos logit contributions | |
poisoning | +0.069 |
injuries | +0.067 |
guard | +0.063 |
unfortunately | +0.063 |
payoff | +0.060 |
Neg logit contributions | |
aling | -0.072 |
dances | -0.064 |
sings | -0.064 |
alling | -0.062 |
astics | -0.062 |
Token | " |
Token ID | 366 |
Feature activation | -1.47e-02 |
Pos logit contributions | |
aling | +0.034 |
alling | +0.028 |
sings | +0.027 |
astics | +0.026 |
Hide | +0.026 |
Neg logit contributions | |
injuries | -0.053 |
poisoning | -0.053 |
guard | -0.052 |
hesitation | -0.049 |
payoff | -0.049 |
Token | Maybe |
Token ID | 13300 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
aling | +0.029 |
alling | +0.026 |
astics | +0.025 |
irl | +0.025 |
sings | +0.024 |
Neg logit contributions | |
poisoning | -0.017 |
injuries | -0.017 |
guard | -0.016 |
payoff | -0.015 |
unfortunately | -0.014 |
Token | the |
Token ID | 262 |
Feature activation | +2.79e-02 |
Pos logit contributions | |
poisoning | +0.120 |
guard | +0.116 |
injuries | +0.115 |
unfortunately | +0.103 |
payoff | +0.103 |
Neg logit contributions | |
aling | -0.191 |
alling | -0.172 |
astics | -0.166 |
irl | -0.165 |
sings | -0.161 |
Token | snail |
Token ID | 47374 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
poisoning | +0.043 |
injuries | +0.042 |
guard | +0.041 |
department | +0.039 |
unfortunately | +0.039 |
Neg logit contributions | |
aling | -0.042 |
alling | -0.036 |
irl | -0.036 |
astics | -0.034 |
sings | -0.034 |
Token | did |
Token ID | 750 |
Feature activation | -1.26e-03 |
Pos logit contributions | |
aling | +0.016 |
alling | +0.014 |
sings | +0.013 |
irl | +0.013 |
astics | +0.013 |
Neg logit contributions | |
injuries | -0.016 |
poisoning | -0.015 |
guard | -0.015 |
unfortunately | -0.015 |
payoff | -0.014 |
Token | not |
Token ID | 407 |
Feature activation | +3.73e-02 |
Pos logit contributions | |
aling | +0.003 |
alling | +0.002 |
astics | +0.002 |
irl | +0.002 |
sings | +0.002 |
Neg logit contributions | |
guard | -0.001 |
poisoning | -0.001 |
injuries | -0.001 |
unfortunately | -0.001 |
ho | -0.001 |
Token | like |
Token ID | 588 |
Feature activation | +9.32e-03 |
Pos logit contributions | |
guard | +0.051 |
poisoning | +0.051 |
injuries | +0.051 |
unfortunately | +0.047 |
payoff | +0.046 |
Neg logit contributions | |
aling | -0.065 |
alling | -0.057 |
astics | -0.054 |
irl | -0.053 |
sings | -0.051 |
Token | Maybe |
Token ID | 13300 |
Feature activation | +9.62e-02 |
Pos logit contributions | |
aling | +0.009 |
alling | +0.008 |
astics | +0.008 |
irl | +0.008 |
sings | +0.008 |
Neg logit contributions | |
poisoning | -0.006 |
injuries | -0.006 |
guard | -0.006 |
payoff | -0.005 |
unfortunately | -0.005 |
Token | it |
Token ID | 340 |
Feature activation | +3.14e-02 |
Pos logit contributions | |
poisoning | +0.125 |
injuries | +0.118 |
guard | +0.115 |
t | +0.108 |
ho | +0.108 |
Neg logit contributions | |
aling | -0.159 |
astics | -0.144 |
sings | -0.143 |
irl | -0.141 |
alling | -0.139 |
Token | is |
Token ID | 318 |
Feature activation | -3.37e-02 |
Pos logit contributions | |
guard | +0.044 |
poisoning | +0.042 |
department | +0.040 |
last | +0.040 |
injuries | +0.039 |
Neg logit contributions | |
aling | -0.054 |
alling | -0.050 |
irl | -0.046 |
astics | -0.046 |
essay | -0.043 |
Token | sad |
Token ID | 6507 |
Feature activation | +5.15e-02 |
Pos logit contributions | |
aling | +0.059 |
alling | +0.052 |
astics | +0.051 |
irl | +0.050 |
sings | +0.050 |
Neg logit contributions | |
guard | -0.045 |
poisoning | -0.042 |
injuries | -0.042 |
t | -0.039 |
ho | -0.039 |
Token | . |
Token ID | 13 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
poisoning | +0.089 |
injuries | +0.088 |
� | +0.087 |
hesitation | +0.087 |
guard | +0.085 |
Neg logit contributions | |
plays | -0.054 |
sings | -0.054 |
aling | -0.053 |
dances | -0.053 |
shows | -0.048 |
Token | Maybe |
Token ID | 6674 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
aling | +0.027 |
alling | +0.025 |
irl | +0.023 |
astics | +0.023 |
sings | +0.021 |
Neg logit contributions | |
guard | -0.021 |
poisoning | -0.021 |
injuries | -0.020 |
unfortunately | -0.019 |
payoff | -0.019 |
Token | it |
Token ID | 340 |
Feature activation | +2.54e-02 |
Pos logit contributions | |
poisoning | +0.132 |
guard | +0.127 |
injuries | +0.125 |
t | +0.119 |
ho | +0.114 |
Neg logit contributions | |
aling | -0.167 |
sings | -0.153 |
irl | -0.148 |
astics | -0.147 |
alling | -0.146 |
Token | is |
Token ID | 318 |
Feature activation | -3.42e-02 |
Pos logit contributions | |
guard | +0.037 |
poisoning | +0.036 |
injuries | +0.034 |
last | +0.033 |
department | +0.032 |
Neg logit contributions | |
aling | -0.042 |
alling | -0.039 |
astics | -0.036 |
irl | -0.036 |
essay | -0.033 |
Token | lonely |
Token ID | 21757 |
Feature activation | +3.50e-02 |
Pos logit contributions | |
aling | +0.059 |
alling | +0.053 |
astics | +0.051 |
irl | +0.051 |
sings | +0.050 |
Neg logit contributions | |
guard | -0.046 |
poisoning | -0.044 |
injuries | -0.043 |
t | -0.040 |
unfortunately | -0.040 |
Token | ." |
Token ID | 526 |
Feature activation | -7.52e-03 |
Pos logit contributions | |
ho | +0.059 |
injuries | +0.059 |
poisoning | +0.058 |
guard | +0.056 |
unfortunately | +0.056 |
Neg logit contributions | |
aling | -0.043 |
sings | -0.038 |
plays | -0.036 |
dances | -0.035 |
astics | -0.035 |
Token |
|
Token ID | 198 |
Feature activation | +2.82e-02 |
Pos logit contributions | |
aling | +0.014 |
alling | +0.012 |
astics | +0.012 |
sings | +0.012 |
irl | +0.011 |
Neg logit contributions | |
guard | -0.010 |
poisoning | -0.010 |
injuries | -0.009 |
payoff | -0.008 |
unfortunately | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | -2.19e-02 |
Pos logit contributions | |
guard | +0.034 |
injuries | +0.033 |
poisoning | +0.032 |
unfortunately | +0.032 |
payoff | +0.031 |
Neg logit contributions | |
aling | -0.037 |
alling | -0.032 |
astics | -0.030 |
sings | -0.030 |
ophone | -0.030 |
Token | She |
Token ID | 1375 |
Feature activation | +5.71e-03 |
Pos logit contributions | |
aling | +0.029 |
alling | +0.025 |
astics | +0.024 |
irl | +0.023 |
sings | +0.023 |
Neg logit contributions | |
guard | -0.039 |
poisoning | -0.038 |
injuries | -0.038 |
payoff | -0.036 |
unfortunately | -0.034 |
Token | thought |
Token ID | 1807 |
Feature activation | +5.66e-02 |
Pos logit contributions | |
guard | +0.007 |
poisoning | +0.007 |
injuries | +0.007 |
unfortunately | +0.006 |
payoff | +0.006 |
Neg logit contributions | |
aling | -0.011 |
alling | -0.010 |
astics | -0.009 |
irl | -0.009 |
ophone | -0.009 |
Token | , |
Token ID | 11 |
Feature activation | +3.35e-02 |
Pos logit contributions | |
poisoning | +0.066 |
extingu | +0.066 |
injuries | +0.066 |
guard | +0.065 |
payoff | +0.065 |
Neg logit contributions | |
aling | -0.102 |
irl | -0.091 |
alling | -0.089 |
dances | -0.088 |
sings | -0.088 |
Token | " |
Token ID | 366 |
Feature activation | -6.06e-03 |
Pos logit contributions | |
guard | +0.050 |
poisoning | +0.048 |
injuries | +0.048 |
payoff | +0.045 |
unfortunately | +0.043 |
Neg logit contributions | |
aling | -0.053 |
alling | -0.047 |
sings | -0.044 |
astics | -0.044 |
irl | -0.044 |
Token | Maybe |
Token ID | 13300 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
aling | +0.011 |
astics | +0.009 |
alling | +0.009 |
irl | +0.009 |
sings | +0.009 |
Neg logit contributions | |
poisoning | -0.009 |
injuries | -0.008 |
hesitation | -0.008 |
guard | -0.008 |
department | -0.008 |
Token | he |
Token ID | 339 |
Feature activation | +3.11e-02 |
Pos logit contributions | |
poisoning | +0.136 |
injuries | +0.127 |
ho | +0.126 |
extingu | +0.121 |
guard | +0.119 |
Neg logit contributions | |
aling | -0.167 |
astics | -0.148 |
sings | -0.144 |
irl | -0.144 |
plays | -0.142 |
Token | is |
Token ID | 318 |
Feature activation | -3.20e-02 |
Pos logit contributions | |
poisoning | +0.042 |
guard | +0.041 |
injuries | +0.038 |
last | +0.038 |
t | +0.036 |
Neg logit contributions | |
aling | -0.054 |
alling | -0.050 |
astics | -0.046 |
irl | -0.045 |
arg | -0.043 |
Token | hungry |
Token ID | 14720 |
Feature activation | +3.09e-02 |
Pos logit contributions | |
aling | +0.065 |
alling | +0.059 |
astics | +0.055 |
irl | +0.055 |
sings | +0.052 |
Neg logit contributions | |
poisoning | -0.035 |
injuries | -0.035 |
guard | -0.034 |
unfortunately | -0.031 |
payoff | -0.030 |
Token | or |
Token ID | 393 |
Feature activation | -3.39e-02 |
Pos logit contributions | |
injuries | +0.048 |
unfortunately | +0.046 |
ho | +0.046 |
poisoning | +0.046 |
hesitation | +0.045 |
Neg logit contributions | |
aling | -0.044 |
alling | -0.035 |
astics | -0.033 |
v | -0.033 |
plays | -0.033 |
Token | lonely |
Token ID | 21757 |
Feature activation | +3.40e-02 |
Pos logit contributions | |
aling | +0.055 |
alling | +0.048 |
astics | +0.045 |
irl | +0.044 |
sings | +0.044 |
Neg logit contributions | |
poisoning | -0.050 |
guard | -0.049 |
injuries | -0.049 |
unfortunately | -0.045 |
payoff | -0.045 |
Token | was |
Token ID | 373 |
Feature activation | -5.26e-02 |
Pos logit contributions | |
guard | +0.003 |
poisoning | +0.003 |
injuries | +0.003 |
unfortunately | +0.002 |
last | +0.002 |
Neg logit contributions | |
aling | -0.006 |
alling | -0.005 |
astics | -0.005 |
irl | -0.005 |
opl | -0.005 |
Token | scared |
Token ID | 12008 |
Feature activation | +3.17e-02 |
Pos logit contributions | |
aling | +0.101 |
alling | +0.091 |
irl | +0.088 |
astics | +0.088 |
ophone | +0.085 |
Neg logit contributions | |
injuries | -0.063 |
poisoning | -0.062 |
guard | -0.057 |
last | -0.055 |
though | -0.053 |
Token | too |
Token ID | 1165 |
Feature activation | +3.16e-02 |
Pos logit contributions | |
poisoning | +0.055 |
comb | +0.053 |
payoff | +0.053 |
ho | +0.052 |
guard | +0.051 |
Neg logit contributions | |
aling | -0.041 |
alling | -0.034 |
v | -0.030 |
plays | -0.030 |
zoom | -0.029 |
Token | . |
Token ID | 13 |
Feature activation | -2.56e-02 |
Pos logit contributions | |
poisoning | +0.040 |
guard | +0.039 |
injuries | +0.038 |
payoff | +0.037 |
unfortunately | +0.036 |
Neg logit contributions | |
aling | -0.057 |
alling | -0.050 |
irl | -0.048 |
astics | -0.048 |
sings | -0.045 |
Token | " |
Token ID | 366 |
Feature activation | -1.24e-02 |
Pos logit contributions | |
aling | +0.026 |
alling | +0.021 |
astics | +0.020 |
atters | +0.019 |
learn | +0.018 |
Neg logit contributions | |
poisoning | -0.051 |
guard | -0.051 |
injuries | -0.051 |
payoff | -0.049 |
ho | -0.046 |
Token | Maybe |
Token ID | 13300 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
aling | +0.026 |
alling | +0.023 |
astics | +0.022 |
irl | +0.022 |
sings | +0.021 |
Neg logit contributions | |
poisoning | -0.013 |
injuries | -0.013 |
guard | -0.013 |
payoff | -0.011 |
unfortunately | -0.011 |
Token | we |
Token ID | 356 |
Feature activation | +2.59e-02 |
Pos logit contributions | |
poisoning | +0.115 |
injuries | +0.109 |
guard | +0.108 |
ho | +0.101 |
payoff | +0.098 |
Neg logit contributions | |
aling | -0.193 |
alling | -0.174 |
astics | -0.168 |
irl | -0.166 |
sings | -0.163 |
Token | are |
Token ID | 389 |
Feature activation | -1.64e-02 |
Pos logit contributions | |
guard | +0.030 |
poisoning | +0.030 |
injuries | +0.029 |
unfortunately | +0.027 |
payoff | +0.027 |
Neg logit contributions | |
aling | -0.050 |
alling | -0.045 |
irl | -0.042 |
astics | -0.042 |
sings | -0.041 |
Token | lost |
Token ID | 2626 |
Feature activation | +8.37e-03 |
Pos logit contributions | |
aling | +0.030 |
alling | +0.027 |
astics | +0.026 |
irl | +0.025 |
irled | +0.024 |
Neg logit contributions | |
injuries | -0.021 |
poisoning | -0.021 |
guard | -0.020 |
unfortunately | -0.018 |
last | -0.018 |
Token | ." |
Token ID | 526 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
poisoning | +0.013 |
payoff | +0.013 |
unfortunately | +0.013 |
injuries | +0.013 |
ho | +0.013 |
Neg logit contributions | |
aling | -0.011 |
alling | -0.009 |
astics | -0.009 |
v | -0.008 |
ab | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | +2.56e-02 |
Pos logit contributions | |
aling | +0.011 |
astics | +0.009 |
learn | +0.009 |
irl | +0.009 |
alling | +0.009 |
Neg logit contributions | |
poisoning | -0.020 |
injuries | -0.019 |
guard | -0.019 |
ho | -0.018 |
payoff | -0.018 |
Token | is |
Token ID | 318 |
Feature activation | -4.47e-02 |
Pos logit contributions | |
injuries | +0.003 |
last | +0.003 |
poisoning | +0.003 |
or | +0.003 |
department | +0.003 |
Neg logit contributions | |
aling | -0.004 |
alling | -0.004 |
astics | -0.003 |
ophone | -0.003 |
beh | -0.003 |
Token | scared |
Token ID | 12008 |
Feature activation | +3.31e-02 |
Pos logit contributions | |
aling | +0.080 |
alling | +0.071 |
astics | +0.068 |
irl | +0.067 |
sings | +0.064 |
Neg logit contributions | |
injuries | -0.060 |
poisoning | -0.059 |
guard | -0.057 |
unfortunately | -0.053 |
last | -0.052 |
Token | too |
Token ID | 1165 |
Feature activation | +3.24e-02 |
Pos logit contributions | |
poisoning | +0.056 |
payoff | +0.053 |
guard | +0.052 |
comb | +0.051 |
injuries | +0.050 |
Neg logit contributions | |
aling | -0.046 |
alling | -0.039 |
v | -0.034 |
plays | -0.033 |
dances | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | -2.91e-02 |
Pos logit contributions | |
poisoning | +0.042 |
guard | +0.041 |
injuries | +0.039 |
payoff | +0.039 |
unfortunately | +0.038 |
Neg logit contributions | |
aling | -0.056 |
alling | -0.050 |
irl | -0.048 |
astics | -0.047 |
sings | -0.045 |
Token | " |
Token ID | 366 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
aling | +0.032 |
alling | +0.026 |
astics | +0.026 |
atters | +0.024 |
sings | +0.024 |
Neg logit contributions | |
injuries | -0.056 |
guard | -0.056 |
poisoning | -0.056 |
payoff | -0.052 |
unfortunately | -0.050 |
Token | Maybe |
Token ID | 13300 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
aling | +0.026 |
alling | +0.024 |
astics | +0.022 |
irl | +0.022 |
sings | +0.021 |
Neg logit contributions | |
poisoning | -0.014 |
injuries | -0.014 |
guard | -0.013 |
payoff | -0.012 |
unfortunately | -0.012 |
Token | we |
Token ID | 356 |
Feature activation | +2.46e-02 |
Pos logit contributions | |
poisoning | +0.122 |
injuries | +0.117 |
guard | +0.116 |
ho | +0.107 |
payoff | +0.104 |
Neg logit contributions | |
aling | -0.182 |
alling | -0.162 |
astics | -0.159 |
irl | -0.158 |
sings | -0.155 |
Token | should |
Token ID | 815 |
Feature activation | +3.09e-02 |
Pos logit contributions | |
guard | +0.029 |
poisoning | +0.029 |
injuries | +0.029 |
unfortunately | +0.026 |
payoff | +0.026 |
Neg logit contributions | |
aling | -0.047 |
alling | -0.042 |
irl | -0.040 |
astics | -0.039 |
sings | -0.038 |
Token | go |
Token ID | 467 |
Feature activation | -2.45e-02 |
Pos logit contributions | |
guard | +0.042 |
poisoning | +0.041 |
injuries | +0.041 |
unfortunately | +0.038 |
payoff | +0.037 |
Neg logit contributions | |
aling | -0.054 |
alling | -0.049 |
astics | -0.046 |
irl | -0.045 |
sings | -0.043 |
Token | out |
Token ID | 503 |
Feature activation | +2.66e-02 |
Pos logit contributions | |
aling | +0.043 |
alling | +0.038 |
astics | +0.035 |
irl | +0.035 |
atters | +0.033 |
Neg logit contributions | |
poisoning | -0.033 |
injuries | -0.033 |
guard | -0.032 |
unfortunately | -0.032 |
payoff | -0.031 |
Token | ." |
Token ID | 526 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
poisoning | +0.036 |
ho | +0.035 |
guard | +0.034 |
injuries | +0.033 |
payoff | +0.032 |
Neg logit contributions | |
aling | -0.047 |
alling | -0.041 |
astics | -0.037 |
irl | -0.037 |
sings | -0.035 |
Token | cloud |
Token ID | 6279 |
Feature activation | -9.46e-03 |
Pos logit contributions | |
aling | +0.054 |
alling | +0.049 |
astics | +0.046 |
irl | +0.046 |
sings | +0.044 |
Neg logit contributions | |
guard | -0.032 |
injuries | -0.032 |
poisoning | -0.032 |
payoff | -0.029 |
unfortunately | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | -2.42e-02 |
Pos logit contributions | |
aling | +0.015 |
alling | +0.013 |
c | +0.013 |
irl | +0.013 |
ophone | +0.012 |
Neg logit contributions | |
injuries | -0.013 |
poisoning | -0.013 |
payoff | -0.013 |
guard | -0.012 |
hesitation | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | +2.78e-02 |
Pos logit contributions | |
aling | +0.023 |
astics | +0.021 |
alling | +0.019 |
sings | +0.019 |
irl | +0.019 |
Neg logit contributions | |
guard | -0.049 |
injuries | -0.049 |
poisoning | -0.048 |
hesitation | -0.045 |
payoff | -0.045 |
Token |
|
Token ID | 198 |
Feature activation | +2.29e-02 |
Pos logit contributions | |
poisoning | +0.031 |
guard | +0.030 |
injuries | +0.030 |
unfortunately | +0.027 |
payoff | +0.027 |
Neg logit contributions | |
aling | -0.055 |
alling | -0.050 |
irl | -0.048 |
astics | -0.048 |
sings | -0.045 |
Token | " |
Token ID | 1 |
Feature activation | +3.31e-02 |
Pos logit contributions | |
guard | +0.022 |
poisoning | +0.021 |
injuries | +0.021 |
unfortunately | +0.019 |
payoff | +0.018 |
Neg logit contributions | |
aling | -0.049 |
alling | -0.045 |
astics | -0.043 |
irl | -0.042 |
sings | -0.041 |
Token | Maybe |
Token ID | 13300 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
poisoning | +0.048 |
hesitation | +0.047 |
injuries | +0.046 |
guard | +0.045 |
unfortunately | +0.044 |
Neg logit contributions | |
aling | -0.045 |
sings | -0.043 |
irl | -0.043 |
astics | -0.042 |
presses | -0.040 |
Token | we |
Token ID | 356 |
Feature activation | +3.20e-02 |
Pos logit contributions | |
poisoning | +0.146 |
injuries | +0.140 |
guard | +0.133 |
t | +0.128 |
payoff | +0.125 |
Neg logit contributions | |
aling | -0.153 |
irl | -0.139 |
sings | -0.137 |
astics | -0.135 |
alling | -0.133 |
Token | can |
Token ID | 460 |
Feature activation | +4.64e-02 |
Pos logit contributions | |
guard | +0.038 |
poisoning | +0.037 |
injuries | +0.036 |
unfortunately | +0.033 |
payoff | +0.033 |
Neg logit contributions | |
aling | -0.062 |
alling | -0.056 |
astics | -0.052 |
irl | -0.052 |
sings | -0.051 |
Token | ask |
Token ID | 1265 |
Feature activation | +5.39e-02 |
Pos logit contributions | |
guard | +0.077 |
poisoning | +0.075 |
injuries | +0.074 |
unfortunately | +0.070 |
payoff | +0.070 |
Neg logit contributions | |
aling | -0.065 |
irl | -0.059 |
alling | -0.058 |
astics | -0.054 |
atters | -0.052 |
Token | it |
Token ID | 340 |
Feature activation | +4.43e-02 |
Pos logit contributions | |
guard | +0.071 |
poisoning | +0.067 |
t | +0.064 |
unfortunately | +0.063 |
injuries | +0.063 |
Neg logit contributions | |
aling | -0.092 |
alling | -0.083 |
astics | -0.081 |
irl | -0.080 |
sings | -0.077 |
Token | to |
Token ID | 284 |
Feature activation | +4.29e-02 |
Pos logit contributions | |
poisoning | +0.051 |
guard | +0.051 |
injuries | +0.049 |
unfortunately | +0.045 |
payoff | +0.044 |
Neg logit contributions | |
aling | -0.084 |
alling | -0.076 |
astics | -0.075 |
irl | -0.073 |
sings | -0.073 |
Token | sorry |
Token ID | 7926 |
Feature activation | +6.23e-02 |
Pos logit contributions | |
aling | +0.069 |
alling | +0.062 |
astics | +0.059 |
irl | +0.058 |
sings | +0.056 |
Neg logit contributions | |
poisoning | -0.045 |
injuries | -0.044 |
guard | -0.044 |
unfortunately | -0.040 |
payoff | -0.039 |
Token | . |
Token ID | 13 |
Feature activation | -2.89e-02 |
Pos logit contributions | |
poisoning | +0.096 |
guard | +0.096 |
injuries | +0.092 |
payoff | +0.090 |
hesitation | +0.088 |
Neg logit contributions | |
aling | -0.092 |
sings | -0.078 |
alling | -0.077 |
plays | -0.075 |
dances | -0.074 |
Token |
|
Token ID | 198 |
Feature activation | +2.50e-02 |
Pos logit contributions | |
aling | +0.043 |
alling | +0.037 |
astics | +0.036 |
irl | +0.034 |
sings | +0.033 |
Neg logit contributions | |
poisoning | -0.047 |
injuries | -0.047 |
guard | -0.046 |
payoff | -0.042 |
ho | -0.042 |
Token |
|
Token ID | 198 |
Feature activation | +2.22e-02 |
Pos logit contributions | |
poisoning | +0.030 |
guard | +0.029 |
injuries | +0.028 |
unfortunately | +0.026 |
payoff | +0.026 |
Neg logit contributions | |
aling | -0.049 |
alling | -0.044 |
astics | -0.042 |
irl | -0.042 |
sings | -0.039 |
Token | " |
Token ID | 1 |
Feature activation | +2.79e-02 |
Pos logit contributions | |
poisoning | +0.026 |
injuries | +0.025 |
guard | +0.025 |
unfortunately | +0.023 |
payoff | +0.022 |
Neg logit contributions | |
aling | -0.044 |
alling | -0.039 |
astics | -0.038 |
irl | -0.037 |
sings | -0.035 |
Token | Maybe |
Token ID | 13300 |
Feature activation | +1.00e-01 |
Pos logit contributions | |
injuries | +0.041 |
poisoning | +0.040 |
hesitation | +0.039 |
department | +0.036 |
guard | +0.035 |
Neg logit contributions | |
aling | -0.043 |
astics | -0.038 |
irl | -0.037 |
Check | -0.036 |
alling | -0.036 |
Token | we |
Token ID | 356 |
Feature activation | +3.08e-02 |
Pos logit contributions | |
poisoning | +0.140 |
injuries | +0.131 |
guard | +0.130 |
ho | +0.126 |
t | +0.121 |
Neg logit contributions | |
aling | -0.161 |
irl | -0.141 |
astics | -0.141 |
alling | -0.137 |
sings | -0.135 |
Token | are |
Token ID | 389 |
Feature activation | -9.98e-03 |
Pos logit contributions | |
guard | +0.034 |
poisoning | +0.034 |
injuries | +0.034 |
unfortunately | +0.030 |
payoff | +0.030 |
Neg logit contributions | |
aling | -0.061 |
alling | -0.055 |
astics | -0.052 |
irl | -0.052 |
sings | -0.050 |
Token | both |
Token ID | 1111 |
Feature activation | +1.76e-02 |
Pos logit contributions | |
aling | +0.018 |
alling | +0.017 |
astics | +0.016 |
irl | +0.015 |
sings | +0.015 |
Neg logit contributions | |
poisoning | -0.013 |
injuries | -0.012 |
guard | -0.012 |
unfortunately | -0.011 |
payoff | -0.011 |
Token | bad |
Token ID | 2089 |
Feature activation | +5.58e-02 |
Pos logit contributions | |
poisoning | +0.028 |
guard | +0.028 |
injuries | +0.027 |
ho | +0.027 |
t | +0.027 |
Neg logit contributions | |
aling | -0.024 |
alling | -0.021 |
astics | -0.020 |
irl | -0.019 |
sings | -0.019 |
Token | fisher |
Token ID | 17685 |
Feature activation | -3.17e-02 |
Pos logit contributions | |
poisoning | +0.078 |
guard | +0.075 |
unfortunately | +0.074 |
t | +0.073 |
ho | +0.072 |
Neg logit contributions | |
aling | -0.082 |
sings | -0.080 |
astics | -0.079 |
dances | -0.078 |
irl | -0.076 |
Token | too |
Token ID | 1165 |
Feature activation | +3.51e-02 |
Pos logit contributions | |
poisoning | +0.081 |
guard | +0.080 |
injuries | +0.079 |
payoff | +0.072 |
unfortunately | +0.072 |
Neg logit contributions | |
aling | -0.105 |
alling | -0.092 |
sings | -0.091 |
astics | -0.088 |
irl | -0.086 |
Token | . |
Token ID | 13 |
Feature activation | -2.05e-02 |
Pos logit contributions | |
guard | +0.032 |
injuries | +0.032 |
poisoning | +0.031 |
unfortunately | +0.027 |
payoff | +0.026 |
Neg logit contributions | |
aling | -0.078 |
alling | -0.072 |
astics | -0.069 |
irl | -0.068 |
sings | -0.065 |
Token |
|
Token ID | 198 |
Feature activation | +3.37e-02 |
Pos logit contributions | |
aling | +0.029 |
alling | +0.025 |
astics | +0.024 |
irl | +0.022 |
sings | +0.022 |
Neg logit contributions | |
poisoning | -0.035 |
injuries | -0.035 |
guard | -0.034 |
payoff | -0.032 |
unfortunately | -0.031 |
Token |
|
Token ID | 198 |
Feature activation | +3.23e-02 |
Pos logit contributions | |
poisoning | +0.040 |
guard | +0.039 |
injuries | +0.038 |
unfortunately | +0.035 |
payoff | +0.035 |
Neg logit contributions | |
aling | -0.065 |
alling | -0.059 |
irl | -0.056 |
astics | -0.056 |
sings | -0.053 |
Token | " |
Token ID | 1 |
Feature activation | +3.06e-02 |
Pos logit contributions | |
poisoning | +0.039 |
injuries | +0.038 |
guard | +0.038 |
payoff | +0.034 |
unfortunately | +0.034 |
Neg logit contributions | |
aling | -0.063 |
alling | -0.056 |
astics | -0.054 |
irl | -0.052 |
sings | -0.050 |
Token | Maybe |
Token ID | 13300 |
Feature activation | +1.00e-01 |
Pos logit contributions | |
injuries | +0.046 |
poisoning | +0.045 |
hesitation | +0.044 |
department | +0.040 |
extingu | +0.040 |
Neg logit contributions | |
aling | -0.046 |
astics | -0.040 |
irl | -0.039 |
alling | -0.039 |
sings | -0.038 |
Token | we |
Token ID | 356 |
Feature activation | +3.14e-02 |
Pos logit contributions | |
poisoning | +0.127 |
injuries | +0.117 |
guard | +0.115 |
ho | +0.110 |
t | +0.110 |
Neg logit contributions | |
aling | -0.170 |
astics | -0.153 |
irl | -0.152 |
sings | -0.152 |
alling | -0.148 |
Token | should |
Token ID | 815 |
Feature activation | +3.69e-02 |
Pos logit contributions | |
guard | +0.035 |
poisoning | +0.035 |
injuries | +0.034 |
unfortunately | +0.031 |
payoff | +0.030 |
Neg logit contributions | |
aling | -0.062 |
alling | -0.057 |
astics | -0.053 |
irl | -0.053 |
sings | -0.051 |
Token | say |
Token ID | 910 |
Feature activation | +1.32e-02 |
Pos logit contributions | |
guard | +0.050 |
poisoning | +0.050 |
injuries | +0.049 |
unfortunately | +0.045 |
payoff | +0.044 |
Neg logit contributions | |
aling | -0.063 |
alling | -0.057 |
irl | -0.054 |
astics | -0.054 |
sings | -0.052 |
Token | sorry |
Token ID | 7926 |
Feature activation | +5.02e-02 |
Pos logit contributions | |
guard | +0.021 |
poisoning | +0.021 |
injuries | +0.021 |
unfortunately | +0.019 |
payoff | +0.019 |
Neg logit contributions | |
aling | -0.019 |
alling | -0.017 |
astics | -0.016 |
irl | -0.016 |
sings | -0.016 |
Token | to |
Token ID | 284 |
Feature activation | +2.73e-02 |
Pos logit contributions | |
poisoning | +0.060 |
injuries | +0.059 |
� | +0.057 |
unfortunately | +0.057 |
guard | +0.057 |
Neg logit contributions | |
aling | -0.087 |
sings | -0.082 |
dances | -0.081 |
plays | -0.080 |
astics | -0.079 |
Token | . |
Token ID | 13 |
Feature activation | -2.99e-02 |
Pos logit contributions | |
injuries | +0.028 |
poisoning | +0.028 |
guard | +0.028 |
payoff | +0.026 |
unfortunately | +0.025 |
Neg logit contributions | |
aling | -0.039 |
alling | -0.035 |
astics | -0.033 |
irl | -0.033 |
sings | -0.032 |
Token | He |
Token ID | 679 |
Feature activation | -2.86e-03 |
Pos logit contributions | |
aling | +0.037 |
astics | +0.031 |
alling | +0.029 |
sings | +0.027 |
plays | +0.027 |
Neg logit contributions | |
injuries | -0.057 |
poisoning | -0.056 |
guard | -0.054 |
ho | -0.052 |
hesitation | -0.051 |
Token | thinks |
Token ID | 6834 |
Feature activation | +6.47e-02 |
Pos logit contributions | |
aling | +0.005 |
alling | +0.004 |
vel | +0.004 |
astics | +0.004 |
etition | +0.004 |
Neg logit contributions | |
guard | -0.004 |
last | -0.004 |
poisoning | -0.004 |
injuries | -0.003 |
that | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | +3.08e-02 |
Pos logit contributions | |
poisoning | +0.069 |
injuries | +0.066 |
guard | +0.064 |
payoff | +0.062 |
extingu | +0.062 |
Neg logit contributions | |
aling | -0.124 |
irl | -0.114 |
alling | -0.111 |
sings | -0.110 |
dances | -0.109 |
Token | " |
Token ID | 366 |
Feature activation | -9.59e-03 |
Pos logit contributions | |
poisoning | +0.033 |
guard | +0.033 |
injuries | +0.033 |
unfortunately | +0.031 |
payoff | +0.029 |
Neg logit contributions | |
aling | -0.062 |
alling | -0.056 |
astics | -0.054 |
irl | -0.053 |
irled | -0.050 |
Token | Maybe |
Token ID | 13300 |
Feature activation | +1.00e-01 |
Pos logit contributions | |
aling | +0.020 |
alling | +0.018 |
astics | +0.017 |
irl | +0.017 |
sings | +0.017 |
Neg logit contributions | |
injuries | -0.010 |
poisoning | -0.010 |
guard | -0.009 |
unfortunately | -0.009 |
hesitation | -0.009 |
Token | I |
Token ID | 314 |
Feature activation | +3.12e-02 |
Pos logit contributions | |
poisoning | +0.135 |
injuries | +0.131 |
ho | +0.118 |
guard | +0.118 |
extingu | +0.117 |
Neg logit contributions | |
aling | -0.168 |
astics | -0.150 |
alling | -0.145 |
sings | -0.144 |
irl | -0.144 |
Token | was |
Token ID | 373 |
Feature activation | -3.46e-02 |
Pos logit contributions | |
guard | +0.034 |
poisoning | +0.033 |
injuries | +0.032 |
unfortunately | +0.029 |
payoff | +0.029 |
Neg logit contributions | |
aling | -0.063 |
alling | -0.058 |
astics | -0.055 |
irl | -0.054 |
sings | -0.052 |
Token | wrong |
Token ID | 2642 |
Feature activation | +3.89e-02 |
Pos logit contributions | |
aling | +0.066 |
alling | +0.061 |
irl | +0.055 |
cycl | +0.055 |
astics | +0.055 |
Neg logit contributions | |
poisoning | -0.041 |
guard | -0.039 |
injuries | -0.038 |
though | -0.036 |
that | -0.036 |
Token | . |
Token ID | 13 |
Feature activation | -3.12e-02 |
Pos logit contributions | |
poisoning | +0.058 |
injuries | +0.056 |
guard | +0.055 |
unfortunately | +0.054 |
ho | +0.052 |
Neg logit contributions | |
aling | -0.060 |
astics | -0.054 |
alling | -0.051 |
irl | -0.051 |
sings | -0.048 |
Token | Maybe |
Token ID | 6674 |
Feature activation | +9.14e-02 |
Pos logit contributions | |
aling | +0.054 |
alling | +0.049 |
irl | +0.046 |
astics | +0.046 |
sings | +0.043 |
Neg logit contributions | |
guard | -0.042 |
poisoning | -0.042 |
injuries | -0.041 |
unfortunately | -0.038 |
payoff | -0.038 |
Token | happy |
Token ID | 3772 |
Feature activation | +4.19e-02 |
Pos logit contributions | |
aling | +0.085 |
alling | +0.078 |
irl | +0.073 |
astics | +0.073 |
vel | +0.072 |
Neg logit contributions | |
guard | -0.069 |
poisoning | -0.062 |
they | -0.062 |
last | -0.061 |
payoff | -0.060 |
Token | . |
Token ID | 13 |
Feature activation | -2.11e-02 |
Pos logit contributions | |
poisoning | +0.064 |
injuries | +0.061 |
guard | +0.061 |
unfortunately | +0.059 |
ho | +0.056 |
Neg logit contributions | |
aling | -0.058 |
sings | -0.053 |
dances | -0.052 |
astics | -0.051 |
plays | -0.051 |
Token | They |
Token ID | 1119 |
Feature activation | +6.27e-04 |
Pos logit contributions | |
aling | +0.037 |
alling | +0.033 |
irl | +0.031 |
astics | +0.031 |
sings | +0.029 |
Neg logit contributions | |
guard | -0.029 |
poisoning | -0.028 |
injuries | -0.028 |
unfortunately | -0.026 |
payoff | -0.026 |
Token | ask |
Token ID | 1265 |
Feature activation | +5.15e-02 |
Pos logit contributions | |
guard | +0.001 |
poisoning | +0.001 |
injuries | +0.001 |
last | +0.001 |
payoff | +0.001 |
Neg logit contributions | |
aling | -0.001 |
alling | -0.001 |
astics | -0.001 |
irl | -0.001 |
irled | -0.001 |
Token | mom |
Token ID | 1995 |
Feature activation | +7.10e-02 |
Pos logit contributions | |
guard | +0.073 |
t | +0.072 |
poisoning | +0.071 |
comb | +0.070 |
storm | +0.068 |
Neg logit contributions | |
aling | -0.071 |
alling | -0.062 |
irl | -0.062 |
plays | -0.061 |
zoom | -0.060 |
Token | if |
Token ID | 611 |
Feature activation | +9.99e-02 |
Pos logit contributions | |
injuries | +0.097 |
poisoning | +0.092 |
guard | +0.090 |
unfortunately | +0.089 |
storm | +0.088 |
Neg logit contributions | |
aling | -0.116 |
alling | -0.101 |
plays | -0.097 |
sings | -0.095 |
ophone | -0.093 |
Token | they |
Token ID | 484 |
Feature activation | +4.73e-02 |
Pos logit contributions | |
injuries | +0.090 |
poisoning | +0.089 |
extingu | +0.085 |
guard | +0.080 |
t | +0.077 |
Neg logit contributions | |
aling | -0.179 |
sings | -0.170 |
irl | -0.168 |
astics | -0.168 |
rode | -0.165 |
Token | can |
Token ID | 460 |
Feature activation | +4.95e-02 |
Pos logit contributions | |
poisoning | +0.059 |
injuries | +0.059 |
guard | +0.059 |
unfortunately | +0.055 |
payoff | +0.052 |
Neg logit contributions | |
aling | -0.087 |
alling | -0.077 |
irl | -0.072 |
astics | -0.072 |
sings | -0.071 |
Token | talk |
Token ID | 1561 |
Feature activation | -5.97e-03 |
Pos logit contributions | |
poisoning | +0.080 |
injuries | +0.078 |
guard | +0.077 |
unfortunately | +0.075 |
payoff | +0.074 |
Neg logit contributions | |
aling | -0.073 |
alling | -0.064 |
irl | -0.062 |
astics | -0.059 |
sings | -0.054 |
Token | to |
Token ID | 284 |
Feature activation | +2.36e-02 |
Pos logit contributions | |
aling | +0.011 |
sings | +0.010 |
alling | +0.010 |
ophone | +0.010 |
astics | +0.009 |
Neg logit contributions | |
injuries | -0.007 |
poisoning | -0.007 |
guard | -0.006 |
ho | -0.006 |
unfortunately | -0.006 |
Token | the |
Token ID | 262 |
Feature activation | +1.29e-02 |
Pos logit contributions | |
poisoning | +0.030 |
guard | +0.030 |
t | +0.028 |
injuries | +0.028 |
payoff | +0.027 |
Neg logit contributions | |
aling | -0.039 |
alling | -0.035 |
irl | -0.034 |
sings | -0.034 |
astics | -0.033 |
Token | want |
Token ID | 765 |
Feature activation | +3.61e-02 |
Pos logit contributions | |
guard | +0.055 |
poisoning | +0.055 |
injuries | +0.054 |
payoff | +0.049 |
unfortunately | +0.049 |
Neg logit contributions | |
aling | -0.070 |
alling | -0.063 |
astics | -0.059 |
irl | -0.058 |
sings | -0.056 |
Token | to |
Token ID | 284 |
Feature activation | +2.01e-02 |
Pos logit contributions | |
poisoning | +0.041 |
injuries | +0.040 |
extingu | +0.039 |
department | +0.038 |
unfortunately | +0.037 |
Neg logit contributions | |
aling | -0.065 |
astics | -0.060 |
sings | -0.059 |
alling | -0.058 |
plays | -0.057 |
Token | move |
Token ID | 1445 |
Feature activation | -2.99e-02 |
Pos logit contributions | |
poisoning | +0.033 |
guard | +0.032 |
injuries | +0.032 |
unfortunately | +0.030 |
ho | +0.029 |
Neg logit contributions | |
aling | -0.030 |
alling | -0.026 |
astics | -0.025 |
irl | -0.025 |
sings | -0.023 |
Token | . |
Token ID | 13 |
Feature activation | -1.53e-02 |
Pos logit contributions | |
aling | +0.052 |
alling | +0.046 |
irl | +0.044 |
astics | +0.044 |
sings | +0.041 |
Neg logit contributions | |
poisoning | -0.040 |
injuries | -0.039 |
guard | -0.038 |
payoff | -0.037 |
unfortunately | -0.036 |
Token | " |
Token ID | 366 |
Feature activation | +2.84e-03 |
Pos logit contributions | |
aling | +0.028 |
alling | +0.025 |
irl | +0.024 |
astics | +0.023 |
sings | +0.022 |
Neg logit contributions | |
guard | -0.019 |
poisoning | -0.018 |
injuries | -0.018 |
unfortunately | -0.018 |
payoff | -0.017 |
Token | Maybe |
Token ID | 13300 |
Feature activation | +9.99e-02 |
Pos logit contributions | |
guard | +0.002 |
poisoning | +0.002 |
injuries | +0.002 |
Not | +0.002 |
payoff | +0.002 |
Neg logit contributions | |
aling | -0.007 |
alling | -0.006 |
irl | -0.006 |
astics | -0.006 |
sings | -0.006 |
Token | I |
Token ID | 314 |
Feature activation | +2.99e-02 |
Pos logit contributions | |
poisoning | +0.127 |
injuries | +0.122 |
guard | +0.118 |
ho | +0.111 |
t | +0.110 |
Neg logit contributions | |
aling | -0.174 |
alling | -0.156 |
astics | -0.155 |
irl | -0.153 |
sings | -0.149 |
Token | can |
Token ID | 460 |
Feature activation | +4.41e-02 |
Pos logit contributions | |
poisoning | +0.036 |
guard | +0.036 |
injuries | +0.036 |
unfortunately | +0.033 |
payoff | +0.032 |
Neg logit contributions | |
aling | -0.056 |
alling | -0.050 |
astics | -0.047 |
irl | -0.047 |
sings | -0.045 |
Token | stay |
Token ID | 2652 |
Feature activation | -3.63e-02 |
Pos logit contributions | |
poisoning | +0.062 |
guard | +0.062 |
injuries | +0.061 |
unfortunately | +0.056 |
payoff | +0.056 |
Neg logit contributions | |
aling | -0.075 |
alling | -0.067 |
irl | -0.063 |
astics | -0.062 |
sings | -0.058 |
Token | a |
Token ID | 257 |
Feature activation | -4.62e-04 |
Pos logit contributions | |
aling | +0.070 |
alling | +0.064 |
astics | +0.060 |
irl | +0.059 |
sings | +0.058 |
Neg logit contributions | |
poisoning | -0.042 |
guard | -0.042 |
injuries | -0.041 |
unfortunately | -0.037 |
payoff | -0.036 |
Token | little |
Token ID | 1310 |
Feature activation | +5.77e-03 |
Pos logit contributions | |
aling | +0.001 |
alling | +0.001 |
irl | +0.001 |
astics | +0.001 |
sings | +0.001 |
Neg logit contributions | |
poisoning | -0.001 |
guard | -0.001 |
injuries | -0.000 |
unfortunately | -0.000 |
t | -0.000 |
Token | it |
Token ID | 340 |
Feature activation | +2.23e-02 |
Pos logit contributions | |
aling | +0.038 |
alling | +0.035 |
astics | +0.033 |
ophone | +0.032 |
irl | +0.032 |
Neg logit contributions | |
poisoning | -0.012 |
injuries | -0.012 |
last | -0.011 |
that | -0.011 |
help | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | -2.55e-02 |
Pos logit contributions | |
guard | +0.029 |
poisoning | +0.027 |
injuries | +0.027 |
payoff | +0.026 |
unfortunately | +0.025 |
Neg logit contributions | |
aling | -0.040 |
alling | -0.036 |
irl | -0.034 |
astics | -0.034 |
sings | -0.032 |
Token |
|
Token ID | 198 |
Feature activation | +3.15e-02 |
Pos logit contributions | |
aling | +0.039 |
alling | +0.034 |
astics | +0.033 |
irl | +0.032 |
sings | +0.031 |
Neg logit contributions | |
guard | -0.040 |
poisoning | -0.040 |
injuries | -0.039 |
payoff | -0.036 |
unfortunately | -0.036 |
Token |
|
Token ID | 198 |
Feature activation | +2.99e-02 |
Pos logit contributions | |
poisoning | +0.039 |
guard | +0.039 |
injuries | +0.037 |
payoff | +0.035 |
unfortunately | +0.035 |
Neg logit contributions | |
aling | -0.059 |
alling | -0.053 |
irl | -0.051 |
astics | -0.051 |
sings | -0.048 |
Token | " |
Token ID | 1 |
Feature activation | +3.30e-02 |
Pos logit contributions | |
guard | +0.032 |
poisoning | +0.032 |
injuries | +0.032 |
unfortunately | +0.028 |
payoff | +0.028 |
Neg logit contributions | |
aling | -0.061 |
alling | -0.055 |
astics | -0.053 |
irl | -0.052 |
sings | -0.050 |
Token | Maybe |
Token ID | 13300 |
Feature activation | +9.97e-02 |
Pos logit contributions | |
injuries | +0.047 |
poisoning | +0.046 |
hesitation | +0.045 |
department | +0.044 |
guard | +0.043 |
Neg logit contributions | |
aling | -0.048 |
astics | -0.046 |
irl | -0.045 |
sings | -0.043 |
learn | -0.041 |
Token | we |
Token ID | 356 |
Feature activation | +2.83e-02 |
Pos logit contributions | |
poisoning | +0.141 |
injuries | +0.137 |
guard | +0.137 |
t | +0.126 |
payoff | +0.123 |
Neg logit contributions | |
aling | -0.157 |
irl | -0.139 |
astics | -0.138 |
sings | -0.137 |
alling | -0.137 |
Token | can |
Token ID | 460 |
Feature activation | +4.33e-02 |
Pos logit contributions | |
guard | +0.030 |
poisoning | +0.030 |
injuries | +0.029 |
unfortunately | +0.026 |
payoff | +0.026 |
Neg logit contributions | |
aling | -0.058 |
alling | -0.053 |
astics | -0.050 |
irl | -0.050 |
sings | -0.048 |
Token | find |
Token ID | 1064 |
Feature activation | +7.96e-03 |
Pos logit contributions | |
guard | +0.068 |
poisoning | +0.068 |
injuries | +0.067 |
payoff | +0.064 |
unfortunately | +0.063 |
Neg logit contributions | |
aling | -0.064 |
alling | -0.057 |
irl | -0.056 |
astics | -0.053 |
atters | -0.050 |
Token | another |
Token ID | 1194 |
Feature activation | +5.15e-02 |
Pos logit contributions | |
guard | +0.011 |
poisoning | +0.010 |
injuries | +0.010 |
unfortunately | +0.010 |
t | +0.009 |
Neg logit contributions | |
aling | -0.014 |
alling | -0.012 |
irl | -0.012 |
astics | -0.012 |
sings | -0.012 |
Token | one |
Token ID | 530 |
Feature activation | +6.49e-02 |
Pos logit contributions | |
injuries | +0.075 |
poisoning | +0.073 |
guard | +0.073 |
unfortunately | +0.067 |
department | +0.064 |
Neg logit contributions | |
aling | -0.080 |
sings | -0.071 |
c | -0.071 |
irl | -0.070 |
alling | -0.070 |
Token | hungry |
Token ID | 14720 |
Feature activation | +3.22e-02 |
Pos logit contributions | |
aling | +0.010 |
alling | +0.008 |
irl | +0.008 |
astics | +0.008 |
sings | +0.007 |
Neg logit contributions | |
guard | -0.010 |
poisoning | -0.010 |
injuries | -0.010 |
unfortunately | -0.010 |
payoff | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | -2.95e-02 |
Pos logit contributions | |
unfortunately | +0.050 |
guard | +0.049 |
injuries | +0.047 |
poisoning | +0.047 |
etsy | +0.047 |
Neg logit contributions | |
aling | -0.042 |
goals | -0.034 |
alling | -0.034 |
rides | -0.033 |
dances | -0.033 |
Token | They |
Token ID | 1119 |
Feature activation | -4.43e-03 |
Pos logit contributions | |
aling | +0.029 |
astics | +0.026 |
sings | +0.024 |
alling | +0.024 |
irl | +0.023 |
Neg logit contributions | |
guard | -0.060 |
poisoning | -0.060 |
injuries | -0.059 |
hesitation | -0.054 |
mercy | -0.054 |
Token | ask |
Token ID | 1265 |
Feature activation | +5.18e-02 |
Pos logit contributions | |
aling | +0.008 |
alling | +0.007 |
astics | +0.006 |
irl | +0.006 |
sings | +0.006 |
Neg logit contributions | |
guard | -0.006 |
poisoning | -0.006 |
injuries | -0.006 |
unfortunately | -0.005 |
payoff | -0.005 |
Token | Mom |
Token ID | 11254 |
Feature activation | +5.97e-02 |
Pos logit contributions | |
guard | +0.074 |
t | +0.068 |
poisoning | +0.067 |
payoff | +0.065 |
unfortunately | +0.063 |
Neg logit contributions | |
aling | -0.075 |
irl | -0.068 |
alling | -0.067 |
astics | -0.066 |
dances | -0.064 |
Token | if |
Token ID | 611 |
Feature activation | +9.97e-02 |
Pos logit contributions | |
injuries | +0.081 |
guard | +0.079 |
poisoning | +0.077 |
unfortunately | +0.075 |
payoff | +0.073 |
Neg logit contributions | |
aling | -0.098 |
alling | -0.085 |
sings | -0.083 |
plays | -0.081 |
astics | -0.080 |
Token | they |
Token ID | 484 |
Feature activation | +4.76e-02 |
Pos logit contributions | |
injuries | +0.093 |
poisoning | +0.088 |
extingu | +0.088 |
guard | +0.086 |
hesitation | +0.076 |
Neg logit contributions | |
aling | -0.176 |
irl | -0.172 |
sings | -0.171 |
astics | -0.169 |
plays | -0.163 |
Token | can |
Token ID | 460 |
Feature activation | +4.96e-02 |
Pos logit contributions | |
guard | +0.077 |
injuries | +0.075 |
poisoning | +0.074 |
unfortunately | +0.074 |
extingu | +0.072 |
Neg logit contributions | |
aling | -0.069 |
alling | -0.056 |
irl | -0.053 |
sings | -0.053 |
astics | -0.053 |
Token | have |
Token ID | 423 |
Feature activation | +1.97e-03 |
Pos logit contributions | |
poisoning | +0.089 |
guard | +0.089 |
injuries | +0.088 |
unfortunately | +0.086 |
ately | +0.085 |
Neg logit contributions | |
aling | -0.057 |
irl | -0.052 |
alling | -0.046 |
astics | -0.043 |
plays | -0.042 |
Token | a |
Token ID | 257 |
Feature activation | +1.24e-02 |
Pos logit contributions | |
guard | +0.003 |
ho | +0.003 |
poisoning | +0.003 |
department | +0.003 |
ately | +0.003 |
Neg logit contributions | |
aling | -0.003 |
irl | -0.002 |
astics | -0.002 |
sings | -0.002 |
alling | -0.002 |
Token | snack |
Token ID | 26906 |
Feature activation | +2.95e-02 |
Pos logit contributions | |
guard | +0.016 |
ately | +0.016 |
unfortunately | +0.016 |
injuries | +0.015 |
poisoning | +0.015 |
Neg logit contributions | |
aling | -0.019 |
irl | -0.018 |
sings | -0.016 |
astics | -0.016 |
alling | -0.016 |
Token | don |
Token ID | 836 |
Feature activation | +2.36e-02 |
Pos logit contributions | |
guard | +0.029 |
poisoning | +0.028 |
injuries | +0.028 |
unfortunately | +0.025 |
payoff | +0.025 |
Neg logit contributions | |
aling | -0.046 |
alling | -0.041 |
astics | -0.039 |
irl | -0.039 |
sings | -0.037 |
Token | 't |
Token ID | 470 |
Feature activation | +3.62e-02 |
Pos logit contributions | |
guard | +0.012 |
poisoning | +0.011 |
injuries | +0.011 |
unfortunately | +0.009 |
t | +0.009 |
Neg logit contributions | |
aling | -0.061 |
alling | -0.057 |
astics | -0.054 |
irl | -0.053 |
sings | -0.053 |
Token | have |
Token ID | 423 |
Feature activation | +3.13e-04 |
Pos logit contributions | |
poisoning | +0.050 |
guard | +0.050 |
injuries | +0.049 |
payoff | +0.046 |
last | +0.045 |
Neg logit contributions | |
aling | -0.061 |
alling | -0.055 |
astics | -0.052 |
irl | -0.052 |
irled | -0.049 |
Token | permission |
Token ID | 7170 |
Feature activation | +5.38e-02 |
Pos logit contributions | |
guard | +0.000 |
poisoning | +0.000 |
injuries | +0.000 |
unfortunately | +0.000 |
ho | +0.000 |
Neg logit contributions | |
aling | -0.001 |
alling | -0.001 |
astics | -0.001 |
irl | -0.001 |
sings | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -2.51e-02 |
Pos logit contributions | |
poisoning | +0.068 |
injuries | +0.068 |
guard | +0.067 |
unfortunately | +0.062 |
payoff | +0.061 |
Neg logit contributions | |
aling | -0.098 |
alling | -0.088 |
astics | -0.083 |
ophone | -0.081 |
irl | -0.077 |
Token | Maybe |
Token ID | 6674 |
Feature activation | +9.96e-02 |
Pos logit contributions | |
aling | +0.046 |
alling | +0.041 |
irl | +0.040 |
astics | +0.037 |
irled | +0.036 |
Neg logit contributions | |
poisoning | -0.031 |
injuries | -0.030 |
guard | -0.030 |
unfortunately | -0.030 |
sir | -0.029 |
Token | we |
Token ID | 356 |
Feature activation | +2.42e-02 |
Pos logit contributions | |
poisoning | +0.118 |
guard | +0.115 |
injuries | +0.111 |
t | +0.103 |
ho | +0.101 |
Neg logit contributions | |
aling | -0.179 |
astics | -0.160 |
alling | -0.160 |
irl | -0.158 |
sings | -0.156 |
Token | should |
Token ID | 815 |
Feature activation | +3.32e-02 |
Pos logit contributions | |
guard | +0.032 |
injuries | +0.032 |
poisoning | +0.032 |
payoff | +0.030 |
unfortunately | +0.029 |
Neg logit contributions | |
aling | -0.043 |
alling | -0.037 |
irl | -0.035 |
astics | -0.034 |
ophone | -0.034 |
Token | go |
Token ID | 467 |
Feature activation | -2.24e-02 |
Pos logit contributions | |
guard | +0.045 |
poisoning | +0.045 |
injuries | +0.044 |
unfortunately | +0.041 |
payoff | +0.040 |
Neg logit contributions | |
aling | -0.058 |
alling | -0.052 |
astics | -0.049 |
irl | -0.049 |
sings | -0.046 |
Token | back |
Token ID | 736 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
aling | +0.043 |
alling | +0.039 |
astics | +0.036 |
irl | +0.036 |
sings | +0.034 |
Neg logit contributions | |
poisoning | -0.027 |
guard | -0.027 |
injuries | -0.027 |
unfortunately | -0.025 |
payoff | -0.024 |
Token | ." |
Token ID | 526 |
Feature activation | -9.27e-03 |
Pos logit contributions | |
poisoning | +0.010 |
injuries | +0.010 |
guard | +0.009 |
unfortunately | +0.009 |
payoff | +0.008 |
Neg logit contributions | |
aling | -0.023 |
alling | -0.021 |
astics | -0.020 |
irl | -0.020 |
sings | -0.020 |
Token | was |
Token ID | 373 |
Feature activation | -4.04e-02 |
Pos logit contributions | |
guard | +0.026 |
poisoning | +0.026 |
injuries | +0.026 |
unfortunately | +0.023 |
payoff | +0.023 |
Neg logit contributions | |
aling | -0.037 |
alling | -0.034 |
astics | -0.032 |
irl | -0.031 |
sings | -0.030 |
Token | sad |
Token ID | 6507 |
Feature activation | +5.59e-02 |
Pos logit contributions | |
aling | +0.068 |
alling | +0.061 |
astics | +0.057 |
irl | +0.057 |
sings | +0.055 |
Neg logit contributions | |
poisoning | -0.056 |
injuries | -0.055 |
guard | -0.055 |
unfortunately | -0.052 |
payoff | -0.050 |
Token | and |
Token ID | 290 |
Feature activation | -3.08e-02 |
Pos logit contributions | |
guard | +0.099 |
payoff | +0.096 |
tha | +0.096 |
poisoning | +0.095 |
injuries | +0.095 |
Neg logit contributions | |
plays | -0.055 |
aling | -0.053 |
dances | -0.051 |
sings | -0.048 |
zoom | -0.045 |
Token | didn |
Token ID | 1422 |
Feature activation | +1.56e-02 |
Pos logit contributions | |
aling | +0.062 |
alling | +0.056 |
astics | +0.053 |
irl | +0.053 |
sings | +0.050 |
Neg logit contributions | |
guard | -0.034 |
poisoning | -0.033 |
injuries | -0.033 |
unfortunately | -0.030 |
payoff | -0.029 |
Token | 't |
Token ID | 470 |
Feature activation | +5.42e-02 |
Pos logit contributions | |
poisoning | +0.013 |
guard | +0.012 |
injuries | +0.012 |
payoff | +0.011 |
unfortunately | +0.010 |
Neg logit contributions | |
aling | -0.036 |
alling | -0.033 |
irl | -0.032 |
astics | -0.031 |
sings | -0.030 |
Token | know |
Token ID | 760 |
Feature activation | +9.96e-02 |
Pos logit contributions | |
guard | +0.069 |
injuries | +0.066 |
poisoning | +0.066 |
help | +0.063 |
last | +0.063 |
Neg logit contributions | |
aling | -0.095 |
alling | -0.089 |
astics | -0.084 |
irl | -0.083 |
irled | -0.081 |
Token | what |
Token ID | 644 |
Feature activation | +7.19e-02 |
Pos logit contributions | |
guard | +0.136 |
poisoning | +0.132 |
payoff | +0.127 |
injuries | +0.126 |
ho | +0.122 |
Neg logit contributions | |
aling | -0.162 |
alling | -0.144 |
astics | -0.137 |
irl | -0.136 |
sings | -0.132 |
Token | to |
Token ID | 284 |
Feature activation | +4.75e-02 |
Pos logit contributions | |
injuries | +0.074 |
poisoning | +0.066 |
guard | +0.064 |
unfortunately | +0.062 |
extingu | +0.060 |
Neg logit contributions | |
aling | -0.140 |
astics | -0.131 |
sings | -0.130 |
plays | -0.128 |
ophone | -0.127 |
Token | do |
Token ID | 466 |
Feature activation | +3.43e-02 |
Pos logit contributions | |
department | +0.094 |
injuries | +0.091 |
guard | +0.091 |
poisoning | +0.089 |
extingu | +0.088 |
Neg logit contributions | |
aling | -0.042 |
irl | -0.041 |
astics | -0.037 |
zoom | -0.036 |
sings | -0.036 |
Token | . |
Token ID | 13 |
Feature activation | -2.73e-03 |
Pos logit contributions | |
injuries | +0.056 |
ho | +0.056 |
guard | +0.055 |
poisoning | +0.054 |
extingu | +0.053 |
Neg logit contributions | |
aling | -0.047 |
astics | -0.042 |
irl | -0.040 |
alling | -0.039 |
vel | -0.037 |
Token | |
Token ID | 220 |
Feature activation | +4.76e-03 |
Pos logit contributions | |
aling | +0.005 |
alling | +0.004 |
astics | +0.004 |
irl | +0.004 |
sings | +0.004 |
Neg logit contributions | |
guard | -0.004 |
poisoning | -0.004 |
injuries | -0.004 |
unfortunately | -0.003 |
payoff | -0.003 |
Token | say |
Token ID | 910 |
Feature activation | +2.23e-02 |
Pos logit contributions | |
poisoning | +0.079 |
guard | +0.078 |
injuries | +0.077 |
unfortunately | +0.073 |
payoff | +0.072 |
Neg logit contributions | |
aling | -0.070 |
alling | -0.061 |
irl | -0.059 |
astics | -0.057 |
sings | -0.052 |
Token | sorry |
Token ID | 7926 |
Feature activation | +5.46e-02 |
Pos logit contributions | |
guard | +0.029 |
poisoning | +0.029 |
injuries | +0.028 |
unfortunately | +0.026 |
payoff | +0.025 |
Neg logit contributions | |
aling | -0.040 |
alling | -0.036 |
astics | -0.034 |
irl | -0.034 |
sings | -0.033 |
Token | and |
Token ID | 290 |
Feature activation | -2.17e-02 |
Pos logit contributions | |
guard | +0.068 |
poisoning | +0.067 |
injuries | +0.067 |
unfortunately | +0.062 |
payoff | +0.061 |
Neg logit contributions | |
aling | -0.093 |
dances | -0.085 |
sings | -0.085 |
astics | -0.084 |
plays | -0.083 |
Token | hug |
Token ID | 16225 |
Feature activation | +3.19e-02 |
Pos logit contributions | |
aling | +0.032 |
sings | +0.027 |
alling | +0.027 |
astics | +0.026 |
plays | +0.026 |
Neg logit contributions | |
poisoning | -0.033 |
injuries | -0.032 |
guard | -0.032 |
t | -0.030 |
ho | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | -2.32e-02 |
Pos logit contributions | |
guard | +0.046 |
poisoning | +0.045 |
injuries | +0.042 |
payoff | +0.042 |
t | +0.042 |
Neg logit contributions | |
aling | -0.046 |
sings | -0.041 |
irl | -0.041 |
astics | -0.040 |
alling | -0.040 |
Token | Maybe |
Token ID | 6674 |
Feature activation | +9.95e-02 |
Pos logit contributions | |
aling | +0.031 |
alling | +0.027 |
astics | +0.027 |
sings | +0.025 |
irl | +0.024 |
Neg logit contributions | |
guard | -0.041 |
poisoning | -0.041 |
injuries | -0.040 |
payoff | -0.036 |
unfortunately | -0.036 |
Token | later |
Token ID | 1568 |
Feature activation | +4.29e-02 |
Pos logit contributions | |
poisoning | +0.100 |
injuries | +0.094 |
ho | +0.093 |
guard | +0.093 |
t | +0.088 |
Neg logit contributions | |
aling | -0.180 |
sings | -0.168 |
astics | -0.165 |
irl | -0.163 |
alling | -0.161 |
Token | , |
Token ID | 11 |
Feature activation | +2.39e-02 |
Pos logit contributions | |
extingu | +0.047 |
poisoning | +0.047 |
injuries | +0.046 |
ousy | +0.042 |
guard | +0.042 |
Neg logit contributions | |
aling | -0.075 |
astics | -0.068 |
sings | -0.067 |
alling | -0.066 |
rides | -0.065 |
Token | they |
Token ID | 484 |
Feature activation | +7.47e-03 |
Pos logit contributions | |
injuries | +0.025 |
guard | +0.025 |
poisoning | +0.024 |
ho | +0.023 |
t | +0.022 |
Neg logit contributions | |
aling | -0.046 |
alling | -0.042 |
sings | -0.041 |
astics | -0.040 |
irl | -0.039 |
Token | would |
Token ID | 561 |
Feature activation | +9.73e-03 |
Pos logit contributions | |
guard | +0.008 |
poisoning | +0.008 |
injuries | +0.008 |
unfortunately | +0.007 |
payoff | +0.007 |
Neg logit contributions | |
aling | -0.015 |
alling | -0.014 |
astics | -0.013 |
irl | -0.013 |
sings | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | -3.06e-02 |
Pos logit contributions | |
guard | +0.015 |
poisoning | +0.015 |
injuries | +0.014 |
unfortunately | +0.013 |
payoff | +0.013 |
Neg logit contributions | |
aling | -0.015 |
alling | -0.013 |
irl | -0.013 |
astics | -0.013 |
sings | -0.012 |
Token | them |
Token ID | 606 |
Feature activation | +3.59e-02 |
Pos logit contributions | |
guard | +0.001 |
poisoning | +0.001 |
Not | +0.001 |
department | +0.001 |
t | +0.001 |
Neg logit contributions | |
aling | -0.002 |
irl | -0.001 |
astics | -0.001 |
alling | -0.001 |
sings | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -2.11e-02 |
Pos logit contributions | |
guard | +0.045 |
injuries | +0.044 |
poisoning | +0.044 |
payoff | +0.043 |
unfortunately | +0.042 |
Neg logit contributions | |
aling | -0.062 |
alling | -0.055 |
astics | -0.053 |
irl | -0.052 |
ophone | -0.050 |
Token |
|
Token ID | 198 |
Feature activation | +2.96e-02 |
Pos logit contributions | |
aling | +0.037 |
alling | +0.033 |
astics | +0.031 |
irl | +0.030 |
sings | +0.029 |
Neg logit contributions | |
guard | -0.029 |
poisoning | -0.029 |
injuries | -0.029 |
payoff | -0.026 |
unfortunately | -0.026 |
Token |
|
Token ID | 198 |
Feature activation | +2.20e-02 |
Pos logit contributions | |
poisoning | +0.036 |
guard | +0.035 |
injuries | +0.034 |
unfortunately | +0.032 |
payoff | +0.031 |
Neg logit contributions | |
aling | -0.057 |
alling | -0.051 |
astics | -0.049 |
irl | -0.048 |
sings | -0.046 |
Token | " |
Token ID | 1 |
Feature activation | +3.22e-02 |
Pos logit contributions | |
guard | +0.022 |
poisoning | +0.022 |
injuries | +0.022 |
unfortunately | +0.019 |
payoff | +0.019 |
Neg logit contributions | |
aling | -0.046 |
alling | -0.042 |
astics | -0.040 |
irl | -0.040 |
sings | -0.038 |
Token | Maybe |
Token ID | 13300 |
Feature activation | +9.94e-02 |
Pos logit contributions | |
poisoning | +0.043 |
injuries | +0.042 |
hesitation | +0.040 |
guard | +0.038 |
ful | +0.038 |
Neg logit contributions | |
aling | -0.053 |
astics | -0.048 |
alling | -0.046 |
irl | -0.045 |
sings | -0.045 |
Token | we |
Token ID | 356 |
Feature activation | +3.21e-02 |
Pos logit contributions | |
poisoning | +0.127 |
injuries | +0.121 |
guard | +0.118 |
t | +0.112 |
ho | +0.107 |
Neg logit contributions | |
aling | -0.164 |
astics | -0.149 |
sings | -0.147 |
irl | -0.146 |
alling | -0.143 |
Token | can |
Token ID | 460 |
Feature activation | +4.98e-02 |
Pos logit contributions | |
guard | +0.040 |
poisoning | +0.039 |
injuries | +0.039 |
unfortunately | +0.036 |
payoff | +0.035 |
Neg logit contributions | |
aling | -0.060 |
alling | -0.053 |
astics | -0.050 |
irl | -0.050 |
sings | -0.048 |
Token | fly |
Token ID | 6129 |
Feature activation | -4.27e-02 |
Pos logit contributions | |
guard | +0.084 |
poisoning | +0.082 |
injuries | +0.081 |
payoff | +0.076 |
unfortunately | +0.075 |
Neg logit contributions | |
aling | -0.067 |
alling | -0.058 |
irl | -0.058 |
astics | -0.055 |
atters | -0.050 |
Token | with |
Token ID | 351 |
Feature activation | -8.54e-03 |
Pos logit contributions | |
aling | +0.067 |
irl | +0.057 |
alling | +0.056 |
c | +0.056 |
astics | +0.054 |
Neg logit contributions | |
injuries | -0.064 |
poisoning | -0.063 |
unfortunately | -0.059 |
hesitation | -0.059 |
guard | -0.058 |
Token | the |
Token ID | 262 |
Feature activation | +7.99e-03 |
Pos logit contributions | |
aling | +0.013 |
irl | +0.012 |
alling | +0.011 |
sings | +0.011 |
astics | +0.011 |
Neg logit contributions | |
guard | -0.013 |
poisoning | -0.012 |
t | -0.012 |
department | -0.012 |
unfortunately | -0.011 |
Token | afraid |
Token ID | 7787 |
Feature activation | +5.34e-02 |
Pos logit contributions | |
aling | +0.090 |
alling | +0.081 |
astics | +0.078 |
irl | +0.077 |
ophone | +0.074 |
Neg logit contributions | |
poisoning | -0.064 |
injuries | -0.063 |
guard | -0.062 |
unfortunately | -0.058 |
last | -0.057 |
Token | and |
Token ID | 290 |
Feature activation | -4.26e-02 |
Pos logit contributions | |
poisoning | +0.073 |
guard | +0.067 |
payoff | +0.066 |
injuries | +0.065 |
extingu | +0.064 |
Neg logit contributions | |
aling | -0.087 |
sings | -0.078 |
dances | -0.076 |
alling | -0.076 |
plays | -0.074 |
Token | careful |
Token ID | 8161 |
Feature activation | +3.47e-02 |
Pos logit contributions | |
aling | +0.077 |
alling | +0.070 |
astics | +0.066 |
irl | +0.066 |
ophone | +0.062 |
Neg logit contributions | |
guard | -0.054 |
poisoning | -0.054 |
injuries | -0.053 |
unfortunately | -0.050 |
payoff | -0.048 |
Token | . |
Token ID | 13 |
Feature activation | -2.37e-02 |
Pos logit contributions | |
poisoning | +0.048 |
injuries | +0.047 |
guard | +0.046 |
payoff | +0.044 |
extingu | +0.041 |
Neg logit contributions | |
aling | -0.059 |
alling | -0.053 |
astics | -0.049 |
irl | -0.048 |
ophone | -0.048 |
Token | " |
Token ID | 366 |
Feature activation | -6.89e-03 |
Pos logit contributions | |
aling | +0.037 |
alling | +0.033 |
astics | +0.032 |
irl | +0.031 |
sings | +0.030 |
Neg logit contributions | |
guard | -0.038 |
poisoning | -0.036 |
injuries | -0.035 |
payoff | -0.033 |
unfortunately | -0.031 |
Token | Maybe |
Token ID | 13300 |
Feature activation | +9.94e-02 |
Pos logit contributions | |
aling | +0.014 |
alling | +0.013 |
astics | +0.012 |
irl | +0.012 |
sings | +0.012 |
Neg logit contributions | |
guard | -0.007 |
poisoning | -0.007 |
injuries | -0.007 |
unfortunately | -0.006 |
payoff | -0.006 |
Token | there |
Token ID | 612 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
poisoning | +0.135 |
injuries | +0.129 |
guard | +0.128 |
t | +0.121 |
ho | +0.121 |
Neg logit contributions | |
aling | -0.158 |
irl | -0.141 |
astics | -0.140 |
sings | -0.140 |
alling | -0.140 |
Token | are |
Token ID | 389 |
Feature activation | -1.50e-02 |
Pos logit contributions | |
injuries | +0.079 |
poisoning | +0.079 |
guard | +0.077 |
unfortunately | +0.073 |
payoff | +0.072 |
Neg logit contributions | |
aling | -0.072 |
alling | -0.062 |
astics | -0.058 |
irl | -0.058 |
sings | -0.058 |
Token | monsters |
Token ID | 10963 |
Feature activation | +2.18e-02 |
Pos logit contributions | |
aling | +0.022 |
astics | +0.019 |
sings | +0.019 |
alling | +0.019 |
irl | +0.019 |
Neg logit contributions | |
guard | -0.022 |
poisoning | -0.022 |
injuries | -0.020 |
payoff | -0.019 |
unfortunately | -0.019 |
Token | or |
Token ID | 393 |
Feature activation | -1.59e-02 |
Pos logit contributions | |
injuries | +0.026 |
poisoning | +0.026 |
guard | +0.025 |
unfortunately | +0.024 |
payoff | +0.024 |
Neg logit contributions | |
aling | -0.041 |
alling | -0.036 |
sings | -0.034 |
astics | -0.033 |
ophone | -0.033 |
Token | ghosts |
Token ID | 25899 |
Feature activation | +6.42e-03 |
Pos logit contributions | |
aling | +0.022 |
sings | +0.020 |
alling | +0.019 |
dances | +0.019 |
astics | +0.019 |
Neg logit contributions | |
poisoning | -0.025 |
guard | -0.025 |
unfortunately | -0.023 |
t | -0.023 |
injuries | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | +1.83e-02 |
Pos logit contributions | |
guard | +0.018 |
injuries | +0.016 |
poisoning | +0.015 |
unfortunately | +0.013 |
t | +0.013 |
Neg logit contributions | |
aling | -0.057 |
alling | -0.053 |
astics | -0.051 |
irl | -0.050 |
sings | -0.050 |
Token | " |
Token ID | 1 |
Feature activation | +3.33e-02 |
Pos logit contributions | |
guard | +0.014 |
Not | +0.012 |
injuries | +0.011 |
poisoning | +0.011 |
unfortunately | +0.011 |
Neg logit contributions | |
aling | -0.041 |
alling | -0.039 |
sings | -0.038 |
irl | -0.037 |
ophone | -0.037 |
Token | Me |
Token ID | 5308 |
Feature activation | -1.29e-02 |
Pos logit contributions | |
injuries | +0.046 |
poisoning | +0.044 |
hesitation | +0.041 |
ately | +0.040 |
guard | +0.040 |
Neg logit contributions | |
aling | -0.056 |
alling | -0.048 |
astics | -0.046 |
sings | -0.046 |
irl | -0.045 |
Token | too |
Token ID | 1165 |
Feature activation | +4.90e-02 |
Pos logit contributions | |
aling | +0.023 |
alling | +0.021 |
irl | +0.020 |
astics | +0.020 |
sings | +0.019 |
Neg logit contributions | |
poisoning | -0.016 |
injuries | -0.016 |
guard | -0.016 |
unfortunately | -0.014 |
payoff | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -1.91e-02 |
Pos logit contributions | |
poisoning | +0.055 |
injuries | +0.053 |
guard | +0.052 |
payoff | +0.046 |
unfortunately | +0.046 |
Neg logit contributions | |
aling | -0.092 |
alling | -0.082 |
irl | -0.081 |
astics | -0.079 |
sings | -0.077 |
Token | Maybe |
Token ID | 6674 |
Feature activation | +9.93e-02 |
Pos logit contributions | |
aling | +0.032 |
alling | +0.028 |
irl | +0.027 |
astics | +0.027 |
sings | +0.025 |
Neg logit contributions | |
guard | -0.027 |
poisoning | -0.027 |
injuries | -0.027 |
unfortunately | -0.025 |
payoff | -0.025 |
Token | we |
Token ID | 356 |
Feature activation | +2.69e-02 |
Pos logit contributions | |
poisoning | +0.128 |
injuries | +0.123 |
guard | +0.120 |
ho | +0.116 |
t | +0.115 |
Neg logit contributions | |
aling | -0.162 |
alling | -0.142 |
sings | -0.140 |
astics | -0.139 |
irl | -0.139 |
Token | should |
Token ID | 815 |
Feature activation | +3.72e-02 |
Pos logit contributions | |
guard | +0.031 |
poisoning | +0.031 |
injuries | +0.031 |
unfortunately | +0.028 |
payoff | +0.027 |
Neg logit contributions | |
aling | -0.052 |
alling | -0.047 |
astics | -0.044 |
irl | -0.044 |
sings | -0.043 |
Token | sell |
Token ID | 3677 |
Feature activation | -1.01e-05 |
Pos logit contributions | |
guard | +0.052 |
poisoning | +0.051 |
injuries | +0.049 |
unfortunately | +0.046 |
payoff | +0.045 |
Neg logit contributions | |
aling | -0.062 |
alling | -0.056 |
irl | -0.054 |
astics | -0.054 |
sings | -0.051 |
Token | the |
Token ID | 262 |
Feature activation | +2.44e-02 |
Pos logit contributions | |
aling | +0.000 |
astics | +0.000 |
rides | +0.000 |
irl | +0.000 |
alling | +0.000 |
Neg logit contributions | |
injuries | -0.000 |
unfortunately | -0.000 |
poisoning | -0.000 |
department | -0.000 |
guard | -0.000 |
Token | book |
Token ID | 1492 |
Feature activation | +1.13e-02 |
Pos logit contributions | |
poisoning | +0.032 |
injuries | +0.032 |
guard | +0.032 |
unfortunately | +0.031 |
t | +0.029 |
Neg logit contributions | |
aling | -0.040 |
alling | -0.036 |
irl | -0.034 |
astics | -0.034 |
sings | -0.034 |
Token | . |
Token ID | 13 |
Feature activation | -1.89e-02 |
Pos logit contributions | |
poisoning | +0.083 |
guard | +0.082 |
injuries | +0.082 |
payoff | +0.074 |
unfortunately | +0.073 |
Neg logit contributions | |
aling | -0.157 |
alling | -0.142 |
astics | -0.136 |
irl | -0.135 |
sings | -0.133 |
Token | Maybe |
Token ID | 6674 |
Feature activation | +9.60e-02 |
Pos logit contributions | |
aling | +0.033 |
alling | +0.030 |
irl | +0.028 |
astics | +0.028 |
sings | +0.026 |
Neg logit contributions | |
guard | -0.025 |
poisoning | -0.025 |
injuries | -0.025 |
unfortunately | -0.023 |
payoff | -0.022 |
Token | they |
Token ID | 484 |
Feature activation | +2.05e-02 |
Pos logit contributions | |
poisoning | +0.120 |
injuries | +0.114 |
guard | +0.110 |
t | +0.108 |
ho | +0.106 |
Neg logit contributions | |
aling | -0.166 |
sings | -0.147 |
astics | -0.146 |
alling | -0.145 |
irl | -0.141 |
Token | forgot |
Token ID | 16453 |
Feature activation | +4.96e-02 |
Pos logit contributions | |
guard | +0.023 |
poisoning | +0.022 |
injuries | +0.022 |
unfortunately | +0.020 |
payoff | +0.020 |
Neg logit contributions | |
aling | -0.041 |
alling | -0.037 |
astics | -0.035 |
irl | -0.035 |
sings | -0.034 |
Token | . |
Token ID | 13 |
Feature activation | -2.04e-02 |
Pos logit contributions | |
injuries | +0.058 |
poisoning | +0.057 |
guard | +0.056 |
ho | +0.056 |
unfortunately | +0.053 |
Neg logit contributions | |
aling | -0.086 |
sings | -0.078 |
alling | -0.076 |
ophone | -0.076 |
astics | -0.075 |
Token | Maybe |
Token ID | 6674 |
Feature activation | +9.93e-02 |
Pos logit contributions | |
aling | +0.032 |
alling | +0.028 |
astics | +0.027 |
irl | +0.026 |
sings | +0.025 |
Neg logit contributions | |
guard | -0.031 |
poisoning | -0.031 |
injuries | -0.031 |
unfortunately | -0.028 |
payoff | -0.028 |
Token | they |
Token ID | 484 |
Feature activation | +1.85e-02 |
Pos logit contributions | |
poisoning | +0.104 |
injuries | +0.100 |
guard | +0.093 |
t | +0.090 |
ho | +0.089 |
Neg logit contributions | |
aling | -0.184 |
sings | -0.174 |
astics | -0.166 |
alling | -0.164 |
plays | -0.162 |
Token | are |
Token ID | 389 |
Feature activation | -1.84e-02 |
Pos logit contributions | |
guard | +0.021 |
poisoning | +0.021 |
injuries | +0.021 |
unfortunately | +0.019 |
payoff | +0.019 |
Neg logit contributions | |
aling | -0.036 |
alling | -0.032 |
astics | -0.030 |
irl | -0.030 |
sings | -0.030 |
Token | late |
Token ID | 2739 |
Feature activation | +5.21e-02 |
Pos logit contributions | |
aling | +0.032 |
alling | +0.028 |
astics | +0.027 |
irl | +0.027 |
sings | +0.025 |
Neg logit contributions | |
guard | -0.025 |
poisoning | -0.025 |
injuries | -0.024 |
unfortunately | -0.023 |
payoff | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | -2.91e-02 |
Pos logit contributions | |
poisoning | +0.069 |
guard | +0.069 |
injuries | +0.066 |
unfortunately | +0.063 |
ho | +0.059 |
Neg logit contributions | |
aling | -0.089 |
alling | -0.079 |
astics | -0.078 |
sings | -0.076 |
irl | -0.073 |
Token | Maybe |
Token ID | 6674 |
Feature activation | +9.52e-02 |
Pos logit contributions | |
aling | +0.042 |
alling | +0.036 |
astics | +0.034 |
sings | +0.033 |
irl | +0.031 |
Neg logit contributions | |
guard | -0.048 |
poisoning | -0.048 |
injuries | -0.048 |
payoff | -0.043 |
unfortunately | -0.043 |
Token | different |
Token ID | 1180 |
Feature activation | -1.44e-02 |
Pos logit contributions | |
aling | +0.031 |
alling | +0.027 |
irl | +0.026 |
astics | +0.026 |
sings | +0.025 |
Neg logit contributions | |
guard | -0.030 |
poisoning | -0.029 |
injuries | -0.027 |
t | -0.026 |
ho | -0.025 |
Token | styles |
Token ID | 12186 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
aling | +0.021 |
alling | +0.018 |
astics | +0.017 |
irl | +0.017 |
sings | +0.016 |
Neg logit contributions | |
guard | -0.024 |
poisoning | -0.024 |
injuries | -0.023 |
unfortunately | -0.022 |
payoff | -0.021 |
Token | like |
Token ID | 588 |
Feature activation | -1.39e-04 |
Pos logit contributions | |
aling | +0.028 |
alling | +0.026 |
astics | +0.024 |
irl | +0.024 |
sings | +0.023 |
Neg logit contributions | |
injuries | -0.018 |
poisoning | -0.018 |
guard | -0.017 |
unfortunately | -0.016 |
payoff | -0.016 |
Token | stars |
Token ID | 5788 |
Feature activation | -9.11e-03 |
Pos logit contributions | |
aling | +0.000 |
alling | +0.000 |
irl | +0.000 |
astics | +0.000 |
sings | +0.000 |
Neg logit contributions | |
guard | -0.000 |
poisoning | -0.000 |
injuries | -0.000 |
unfortunately | -0.000 |
t | -0.000 |
Token | or |
Token ID | 393 |
Feature activation | -2.73e-02 |
Pos logit contributions | |
aling | +0.017 |
alling | +0.016 |
irl | +0.015 |
astics | +0.014 |
sings | +0.014 |
Neg logit contributions | |
injuries | -0.011 |
poisoning | -0.011 |
guard | -0.011 |
unfortunately | -0.010 |
payoff | -0.010 |
Token | spir |
Token ID | 9158 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
aling | +0.038 |
alling | +0.033 |
sings | +0.032 |
astics | +0.031 |
irl | +0.031 |
Neg logit contributions | |
guard | -0.046 |
poisoning | -0.045 |
injuries | -0.042 |
t | -0.042 |
unfortunately | -0.041 |
Token | als |
Token ID | 874 |
Feature activation | -2.15e-02 |
Pos logit contributions | |
astics | +0.270 |
irl | +0.265 |
aling | +0.259 |
essay | +0.246 |
sings | +0.243 |
Neg logit contributions | |
poisoning | -0.141 |
ho | -0.140 |
fourteen | -0.139 |
last | -0.138 |
t | -0.134 |
Token | . |
Token ID | 13 |
Feature activation | -2.29e-02 |
Pos logit contributions | |
aling | +0.039 |
alling | +0.035 |
irl | +0.032 |
sings | +0.031 |
astics | +0.031 |
Neg logit contributions | |
guard | -0.027 |
poisoning | -0.027 |
injuries | -0.027 |
unfortunately | -0.025 |
payoff | -0.025 |
Token | Everyone |
Token ID | 11075 |
Feature activation | +1.28e-03 |
Pos logit contributions | |
aling | +0.042 |
alling | +0.038 |
irl | +0.037 |
astics | +0.035 |
ophone | +0.033 |
Neg logit contributions | |
unfortunately | -0.027 |
poisoning | -0.027 |
guard | -0.026 |
t | -0.025 |
injuries | -0.025 |
Token | was |
Token ID | 373 |
Feature activation | -4.51e-02 |
Pos logit contributions | |
guard | +0.002 |
poisoning | +0.001 |
injuries | +0.001 |
unfortunately | +0.001 |
payoff | +0.001 |
Neg logit contributions | |
aling | -0.002 |
alling | -0.002 |
astics | -0.002 |
irl | -0.002 |
sings | -0.002 |
Token | so |
Token ID | 523 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
aling | +0.071 |
alling | +0.063 |
astics | +0.059 |
irl | +0.058 |
sings | +0.055 |
Neg logit contributions | |
guard | -0.069 |
poisoning | -0.068 |
injuries | -0.068 |
unfortunately | -0.063 |
payoff | -0.062 |
Token | on |
Token ID | 319 |
Feature activation | -7.54e-03 |
Pos logit contributions | |
guard | +0.005 |
poisoning | +0.005 |
injuries | +0.005 |
unfortunately | +0.004 |
last | +0.004 |
Neg logit contributions | |
aling | -0.007 |
alling | -0.006 |
astics | -0.006 |
irl | -0.006 |
sings | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -1.35e-02 |
Pos logit contributions | |
aling | +0.012 |
alling | +0.011 |
astics | +0.011 |
irl | +0.011 |
sings | +0.010 |
Neg logit contributions | |
guard | -0.011 |
poisoning | -0.011 |
injuries | -0.010 |
unfortunately | -0.010 |
t | -0.010 |
Token | lake |
Token ID | 13546 |
Feature activation | -5.03e-02 |
Pos logit contributions | |
aling | +0.022 |
alling | +0.020 |
astics | +0.019 |
irl | +0.018 |
sings | +0.017 |
Neg logit contributions | |
guard | -0.020 |
poisoning | -0.020 |
injuries | -0.019 |
unfortunately | -0.018 |
payoff | -0.017 |
Token | in |
Token ID | 287 |
Feature activation | -2.19e-02 |
Pos logit contributions | |
aling | +0.110 |
alling | +0.102 |
astics | +0.096 |
irl | +0.095 |
sings | +0.092 |
Neg logit contributions | |
guard | -0.047 |
poisoning | -0.046 |
injuries | -0.044 |
unfortunately | -0.040 |
ho | -0.039 |
Token | a |
Token ID | 257 |
Feature activation | -3.52e-02 |
Pos logit contributions | |
aling | +0.045 |
alling | +0.041 |
astics | +0.039 |
irl | +0.039 |
sings | +0.037 |
Neg logit contributions | |
guard | -0.023 |
poisoning | -0.023 |
injuries | -0.023 |
unfortunately | -0.020 |
payoff | -0.020 |
Token | kay |
Token ID | 34681 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
aling | +0.070 |
alling | +0.065 |
aled | +0.060 |
astics | +0.060 |
ophone | +0.059 |
Neg logit contributions | |
guard | -0.039 |
poisoning | -0.036 |
injuries | -0.036 |
last | -0.034 |
payoff | -0.033 |
Token | ak |
Token ID | 461 |
Feature activation | -3.60e-02 |
Pos logit contributions | |
sings | +0.246 |
dances | +0.229 |
presses | +0.221 |
aling | +0.220 |
alling | +0.216 |
Neg logit contributions | |
guard | -0.200 |
t | -0.189 |
poisoning | -0.176 |
ll | -0.163 |
side | -0.158 |
Token | . |
Token ID | 13 |
Feature activation | -9.20e-03 |
Pos logit contributions | |
aling | +0.080 |
alling | +0.074 |
astics | +0.071 |
irl | +0.071 |
sings | +0.068 |
Neg logit contributions | |
guard | -0.032 |
poisoning | -0.030 |
injuries | -0.028 |
unfortunately | -0.027 |
last | -0.025 |
Token | As |
Token ID | 1081 |
Feature activation | +2.61e-02 |
Pos logit contributions | |
aling | +0.018 |
alling | +0.017 |
irl | +0.016 |
astics | +0.015 |
irled | +0.015 |
Neg logit contributions | |
poisoning | -0.011 |
guard | -0.010 |
injuries | -0.010 |
unfortunately | -0.009 |
Papa | -0.009 |
Token | they |
Token ID | 484 |
Feature activation | +5.93e-04 |
Pos logit contributions | |
poisoning | +0.018 |
injuries | +0.018 |
guard | +0.017 |
unfortunately | +0.014 |
payoff | +0.014 |
Neg logit contributions | |
aling | -0.062 |
alling | -0.057 |
irl | -0.056 |
astics | -0.055 |
sings | -0.054 |
Token | were |
Token ID | 547 |
Feature activation | -2.33e-02 |
Pos logit contributions | |
poisoning | +0.001 |
guard | +0.001 |
injuries | +0.001 |
unfortunately | +0.001 |
payoff | +0.001 |
Neg logit contributions | |
aling | -0.001 |
alling | -0.001 |
irl | -0.001 |
astics | -0.001 |
sings | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | -1.92e-02 |
Pos logit contributions | |
aling | +0.032 |
spir | +0.029 |
sorts | +0.028 |
alling | +0.027 |
irl | +0.027 |
Neg logit contributions | |
guard | -0.045 |
t | -0.043 |
poisoning | -0.042 |
ll | -0.039 |
injuries | -0.039 |
Token | air |
Token ID | 1633 |
Feature activation | -1.17e-02 |
Pos logit contributions | |
aling | +0.030 |
alling | +0.027 |
astics | +0.025 |
irl | +0.025 |
sings | +0.023 |
Neg logit contributions | |
guard | -0.029 |
poisoning | -0.029 |
injuries | -0.029 |
payoff | -0.027 |
unfortunately | -0.027 |
Token | , |
Token ID | 11 |
Feature activation | +8.76e-03 |
Pos logit contributions | |
aling | +0.021 |
alling | +0.019 |
irl | +0.018 |
astics | +0.018 |
sings | +0.018 |
Neg logit contributions | |
guard | -0.014 |
poisoning | -0.014 |
injuries | -0.014 |
unfortunately | -0.013 |
payoff | -0.012 |
Token | leaving |
Token ID | 4305 |
Feature activation | -6.80e-03 |
Pos logit contributions | |
guard | +0.009 |
poisoning | +0.009 |
injuries | +0.009 |
unfortunately | +0.008 |
payoff | +0.008 |
Neg logit contributions | |
aling | -0.018 |
alling | -0.016 |
astics | -0.015 |
irl | -0.015 |
sings | -0.015 |
Token | a |
Token ID | 257 |
Feature activation | -1.72e-02 |
Pos logit contributions | |
aling | +0.010 |
alling | +0.009 |
irl | +0.009 |
sings | +0.009 |
c | +0.008 |
Neg logit contributions | |
guard | -0.010 |
injuries | -0.009 |
poisoning | -0.009 |
unfortunately | -0.008 |
extingu | -0.008 |
Token | spir |
Token ID | 9158 |
Feature activation | -1.47e-01 |
Pos logit contributions | |
aling | +0.027 |
alling | +0.024 |
irl | +0.023 |
astics | +0.022 |
sings | +0.021 |
Neg logit contributions | |
guard | -0.025 |
poisoning | -0.025 |
injuries | -0.024 |
unfortunately | -0.023 |
payoff | -0.022 |
Token | aling |
Token ID | 4272 |
Feature activation | -4.56e-02 |
Pos logit contributions | |
irl | +0.237 |
astics | +0.228 |
irling | +0.202 |
essay | +0.196 |
sings | +0.196 |
Neg logit contributions | |
ho | -0.194 |
poisoning | -0.182 |
ful | -0.181 |
last | -0.181 |
fourteen | -0.180 |
Token | trail |
Token ID | 8025 |
Feature activation | -4.09e-02 |
Pos logit contributions | |
aling | +0.078 |
alling | +0.070 |
astics | +0.065 |
irl | +0.064 |
sings | +0.062 |
Neg logit contributions | |
poisoning | -0.063 |
guard | -0.063 |
injuries | -0.062 |
unfortunately | -0.057 |
payoff | -0.056 |
Token | of |
Token ID | 286 |
Feature activation | -2.01e-02 |
Pos logit contributions | |
aling | +0.075 |
alling | +0.068 |
astics | +0.066 |
irl | +0.065 |
sings | +0.063 |
Neg logit contributions | |
guard | -0.052 |
poisoning | -0.050 |
injuries | -0.049 |
unfortunately | -0.045 |
t | -0.045 |
Token | black |
Token ID | 2042 |
Feature activation | -3.51e-02 |
Pos logit contributions | |
aling | +0.028 |
alling | +0.024 |
sings | +0.024 |
irl | +0.023 |
astics | +0.022 |
Neg logit contributions | |
guard | -0.033 |
poisoning | -0.032 |
injuries | -0.031 |
t | -0.031 |
unfortunately | -0.030 |
Token | ink |
Token ID | 16882 |
Feature activation | -2.92e-02 |
Pos logit contributions | |
aling | +0.052 |
alling | +0.046 |
astics | +0.042 |
irl | +0.041 |
sings | +0.041 |
Neg logit contributions | |
guard | -0.056 |
poisoning | -0.056 |
injuries | -0.055 |
unfortunately | -0.053 |
payoff | -0.051 |
Token | liked |
Token ID | 8288 |
Feature activation | -8.00e-03 |
Pos logit contributions | |
aling | +0.001 |
alling | +0.001 |
astics | +0.001 |
irl | +0.001 |
sings | +0.001 |
Neg logit contributions | |
guard | -0.001 |
poisoning | -0.001 |
injuries | -0.001 |
unfortunately | -0.001 |
last | -0.001 |
Token | to |
Token ID | 284 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
aling | +0.018 |
alling | +0.017 |
astics | +0.016 |
irl | +0.016 |
sings | +0.016 |
Neg logit contributions | |
injuries | -0.006 |
poisoning | -0.006 |
guard | -0.006 |
unfortunately | -0.005 |
t | -0.005 |
Token | paddle |
Token ID | 39517 |
Feature activation | -6.32e-02 |
Pos logit contributions | |
aling | +0.028 |
alling | +0.025 |
astics | +0.023 |
ophone | +0.022 |
sings | +0.022 |
Neg logit contributions | |
guard | -0.020 |
poisoning | -0.020 |
injuries | -0.019 |
last | -0.019 |
help | -0.018 |
Token | in |
Token ID | 287 |
Feature activation | -2.24e-02 |
Pos logit contributions | |
aling | +0.114 |
alling | +0.102 |
irl | +0.098 |
astics | +0.097 |
sings | +0.094 |
Neg logit contributions | |
injuries | -0.080 |
poisoning | -0.078 |
guard | -0.076 |
unfortunately | -0.073 |
payoff | -0.072 |
Token | the |
Token ID | 262 |
Feature activation | -2.48e-02 |
Pos logit contributions | |
aling | +0.046 |
alling | +0.042 |
irl | +0.040 |
astics | +0.040 |
sings | +0.038 |
Neg logit contributions | |
guard | -0.023 |
poisoning | -0.023 |
injuries | -0.022 |
unfortunately | -0.020 |
payoff | -0.020 |
Token | kay |
Token ID | 34681 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
aling | +0.046 |
alling | +0.042 |
astics | +0.040 |
irl | +0.039 |
sings | +0.037 |
Neg logit contributions | |
guard | -0.031 |
poisoning | -0.031 |
injuries | -0.031 |
unfortunately | -0.028 |
payoff | -0.028 |
Token | ak |
Token ID | 461 |
Feature activation | -3.26e-02 |
Pos logit contributions | |
sings | +0.221 |
aling | +0.220 |
alling | +0.214 |
dances | +0.204 |
presses | +0.200 |
Neg logit contributions | |
guard | -0.205 |
t | -0.189 |
poisoning | -0.183 |
ll | -0.167 |
last | -0.166 |
Token | and |
Token ID | 290 |
Feature activation | -4.27e-02 |
Pos logit contributions | |
aling | +0.067 |
alling | +0.061 |
astics | +0.058 |
irl | +0.057 |
ophone | +0.055 |
Neg logit contributions | |
injuries | -0.034 |
poisoning | -0.033 |
guard | -0.033 |
payoff | -0.030 |
unfortunately | -0.029 |
Token | splash |
Token ID | 22870 |
Feature activation | -6.06e-02 |
Pos logit contributions | |
aling | +0.074 |
alling | +0.066 |
astics | +0.062 |
irl | +0.061 |
sings | +0.058 |
Neg logit contributions | |
guard | -0.059 |
poisoning | -0.058 |
injuries | -0.057 |
unfortunately | -0.053 |
payoff | -0.052 |
Token | in |
Token ID | 287 |
Feature activation | -2.89e-02 |
Pos logit contributions | |
aling | +0.097 |
alling | +0.086 |
irl | +0.083 |
astics | +0.081 |
sings | +0.077 |
Neg logit contributions | |
injuries | -0.090 |
poisoning | -0.089 |
guard | -0.087 |
unfortunately | -0.083 |
payoff | -0.083 |
Token | the |
Token ID | 262 |
Feature activation | -2.38e-02 |
Pos logit contributions | |
aling | +0.059 |
alling | +0.054 |
irl | +0.051 |
astics | +0.051 |
sings | +0.049 |
Neg logit contributions | |
guard | -0.030 |
poisoning | -0.030 |
injuries | -0.029 |
unfortunately | -0.026 |
payoff | -0.026 |
Token | butterfly |
Token ID | 35113 |
Feature activation | -4.23e-02 |
Pos logit contributions | |
aling | +0.060 |
alling | +0.055 |
irl | +0.052 |
astics | +0.051 |
sings | +0.050 |
Neg logit contributions | |
injuries | -0.040 |
guard | -0.040 |
poisoning | -0.040 |
unfortunately | -0.036 |
payoff | -0.035 |
Token | , |
Token ID | 11 |
Feature activation | +2.19e-03 |
Pos logit contributions | |
aling | +0.089 |
alling | +0.082 |
astics | +0.079 |
irl | +0.077 |
aled | +0.074 |
Neg logit contributions | |
guard | -0.042 |
poisoning | -0.041 |
injuries | -0.039 |
unfortunately | -0.037 |
ho | -0.035 |
Token | tw |
Token ID | 665 |
Feature activation | -9.11e-02 |
Pos logit contributions | |
poisoning | +0.003 |
guard | +0.003 |
injuries | +0.003 |
unfortunately | +0.003 |
payoff | +0.002 |
Neg logit contributions | |
aling | -0.004 |
alling | -0.004 |
astics | -0.003 |
irl | -0.003 |
ophone | -0.003 |
Token | irling |
Token ID | 24297 |
Feature activation | -6.52e-02 |
Pos logit contributions | |
aling | +0.117 |
alling | +0.099 |
ophone | +0.097 |
sings | +0.090 |
cycl | +0.089 |
Neg logit contributions | |
ho | -0.151 |
injuries | -0.145 |
ful | -0.143 |
oles | -0.140 |
that | -0.139 |
Token | and |
Token ID | 290 |
Feature activation | -4.60e-02 |
Pos logit contributions | |
aling | +0.105 |
alling | +0.093 |
astics | +0.087 |
irl | +0.086 |
sings | +0.081 |
Neg logit contributions | |
guard | -0.097 |
poisoning | -0.096 |
injuries | -0.094 |
unfortunately | -0.088 |
payoff | -0.087 |
Token | spir |
Token ID | 9158 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
aling | +0.078 |
alling | +0.070 |
astics | +0.066 |
irl | +0.063 |
ophone | +0.062 |
Neg logit contributions | |
ho | -0.061 |
poisoning | -0.060 |
guard | -0.060 |
payoff | -0.059 |
unfortunately | -0.059 |
Token | aling |
Token ID | 4272 |
Feature activation | -4.45e-02 |
Pos logit contributions | |
astics | +0.279 |
irl | +0.275 |
aling | +0.257 |
essay | +0.249 |
sings | +0.246 |
Neg logit contributions | |
ho | -0.128 |
fourteen | -0.118 |
ful | -0.118 |
t | -0.117 |
last | -0.117 |
Token | in |
Token ID | 287 |
Feature activation | -2.84e-02 |
Pos logit contributions | |
aling | +0.086 |
alling | +0.077 |
astics | +0.074 |
irl | +0.073 |
sings | +0.069 |
Neg logit contributions | |
guard | -0.053 |
poisoning | -0.051 |
injuries | -0.049 |
payoff | -0.046 |
ho | -0.046 |
Token | the |
Token ID | 262 |
Feature activation | -2.36e-02 |
Pos logit contributions | |
aling | +0.047 |
alling | +0.042 |
irl | +0.039 |
astics | +0.038 |
sings | +0.037 |
Neg logit contributions | |
guard | -0.042 |
poisoning | -0.041 |
injuries | -0.040 |
unfortunately | -0.037 |
payoff | -0.036 |
Token | air |
Token ID | 1633 |
Feature activation | -1.74e-02 |
Pos logit contributions | |
aling | +0.043 |
alling | +0.038 |
astics | +0.037 |
irl | +0.037 |
ophone | +0.034 |
Neg logit contributions | |
guard | -0.030 |
poisoning | -0.030 |
injuries | -0.029 |
payoff | -0.029 |
last | -0.029 |
Token | . |
Token ID | 13 |
Feature activation | -2.07e-02 |
Pos logit contributions | |
aling | +0.032 |
alling | +0.029 |
irl | +0.027 |
astics | +0.027 |
sings | +0.026 |
Neg logit contributions | |
guard | -0.021 |
poisoning | -0.021 |
injuries | -0.021 |
unfortunately | -0.019 |
payoff | -0.018 |
Token | saw |
Token ID | 2497 |
Feature activation | -8.27e-03 |
Pos logit contributions | |
aling | +0.011 |
alling | +0.010 |
astics | +0.009 |
irl | +0.009 |
sings | +0.008 |
Neg logit contributions | |
guard | -0.007 |
poisoning | -0.007 |
injuries | -0.007 |
unfortunately | -0.007 |
payoff | -0.007 |
Token | a |
Token ID | 257 |
Feature activation | -1.78e-02 |
Pos logit contributions | |
aling | +0.012 |
alling | +0.011 |
sings | +0.011 |
ophone | +0.010 |
irl | +0.010 |
Neg logit contributions | |
poisoning | -0.012 |
injuries | -0.012 |
guard | -0.012 |
t | -0.011 |
payoff | -0.011 |
Token | man |
Token ID | 582 |
Feature activation | -1.83e-02 |
Pos logit contributions | |
aling | +0.030 |
alling | +0.027 |
irl | +0.026 |
astics | +0.025 |
sings | +0.025 |
Neg logit contributions | |
poisoning | -0.024 |
guard | -0.024 |
injuries | -0.023 |
unfortunately | -0.022 |
payoff | -0.021 |
Token | in |
Token ID | 287 |
Feature activation | -2.46e-02 |
Pos logit contributions | |
aling | +0.033 |
irl | +0.030 |
alling | +0.030 |
astics | +0.030 |
irled | +0.028 |
Neg logit contributions | |
guard | -0.022 |
poisoning | -0.022 |
injuries | -0.021 |
unfortunately | -0.020 |
last | -0.019 |
Token | a |
Token ID | 257 |
Feature activation | -3.88e-02 |
Pos logit contributions | |
aling | +0.042 |
alling | +0.038 |
astics | +0.035 |
irl | +0.035 |
sings | +0.033 |
Neg logit contributions | |
guard | -0.033 |
poisoning | -0.033 |
injuries | -0.031 |
t | -0.031 |
ho | -0.030 |
Token | kay |
Token ID | 34681 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
aling | +0.068 |
alling | +0.061 |
astics | +0.057 |
irl | +0.056 |
sings | +0.054 |
Neg logit contributions | |
guard | -0.053 |
poisoning | -0.053 |
injuries | -0.052 |
unfortunately | -0.047 |
payoff | -0.047 |
Token | ak |
Token ID | 461 |
Feature activation | -2.68e-02 |
Pos logit contributions | |
sings | +0.218 |
dances | +0.207 |
aling | +0.207 |
alling | +0.198 |
astics | +0.193 |
Neg logit contributions | |
guard | -0.211 |
poisoning | -0.194 |
t | -0.192 |
injuries | -0.176 |
last | -0.173 |
Token | . |
Token ID | 13 |
Feature activation | -2.82e-02 |
Pos logit contributions | |
aling | +0.058 |
alling | +0.053 |
irl | +0.052 |
astics | +0.052 |
sings | +0.050 |
Neg logit contributions | |
guard | -0.025 |
poisoning | -0.024 |
injuries | -0.022 |
unfortunately | -0.021 |
last | -0.020 |
Token | The |
Token ID | 383 |
Feature activation | +3.25e-03 |
Pos logit contributions | |
aling | +0.048 |
alling | +0.043 |
irl | +0.041 |
astics | +0.041 |
sings | +0.038 |
Neg logit contributions | |
guard | -0.039 |
poisoning | -0.039 |
injuries | -0.038 |
unfortunately | -0.036 |
payoff | -0.035 |
Token | man |
Token ID | 582 |
Feature activation | -2.17e-02 |
Pos logit contributions | |
guard | +0.003 |
poisoning | +0.003 |
injuries | +0.003 |
unfortunately | +0.002 |
payoff | +0.002 |
Neg logit contributions | |
aling | -0.007 |
alling | -0.007 |
astics | -0.006 |
irl | -0.006 |
sings | -0.006 |
Token | said |
Token ID | 531 |
Feature activation | +3.06e-02 |
Pos logit contributions | |
aling | +0.046 |
alling | +0.043 |
astics | +0.042 |
irl | +0.041 |
etition | +0.038 |
Neg logit contributions | |
guard | -0.023 |
poisoning | -0.020 |
injuries | -0.018 |
last | -0.018 |
unfortunately | -0.018 |
Token | Fro |
Token ID | 9734 |
Feature activation | -4.53e-02 |
Pos logit contributions | |
injuries | +0.060 |
guard | +0.059 |
poisoning | +0.057 |
payoff | +0.057 |
unfortunately | +0.053 |
Neg logit contributions | |
aling | -0.061 |
alling | -0.055 |
astics | -0.051 |
irl | -0.050 |
ophone | -0.050 |
Token | ggy |
Token ID | 19970 |
Feature activation | -3.42e-02 |
Pos logit contributions | |
aling | +0.067 |
irl | +0.058 |
alling | +0.058 |
astics | +0.057 |
sings | +0.052 |
Neg logit contributions | |
guard | -0.070 |
poisoning | -0.070 |
injuries | -0.069 |
payoff | -0.066 |
unfortunately | -0.065 |
Token | , |
Token ID | 11 |
Feature activation | +3.70e-02 |
Pos logit contributions | |
aling | +0.082 |
alling | +0.076 |
astics | +0.073 |
irl | +0.073 |
sings | +0.070 |
Neg logit contributions | |
guard | -0.023 |
poisoning | -0.023 |
injuries | -0.023 |
unfortunately | -0.019 |
payoff | -0.019 |
Token | in |
Token ID | 287 |
Feature activation | -1.55e-02 |
Pos logit contributions | |
injuries | +0.055 |
guard | +0.055 |
poisoning | +0.052 |
payoff | +0.051 |
unfortunately | +0.049 |
Neg logit contributions | |
aling | -0.060 |
alling | -0.053 |
sings | -0.049 |
astics | -0.047 |
irl | -0.047 |
Token | a |
Token ID | 257 |
Feature activation | -3.67e-02 |
Pos logit contributions | |
aling | +0.031 |
alling | +0.028 |
astics | +0.026 |
irl | +0.026 |
sings | +0.025 |
Neg logit contributions | |
guard | -0.017 |
poisoning | -0.017 |
injuries | -0.017 |
unfortunately | -0.015 |
payoff | -0.015 |
Token | kay |
Token ID | 34681 |
Feature activation | -1.43e-01 |
Pos logit contributions | |
aling | +0.067 |
alling | +0.061 |
astics | +0.057 |
irl | +0.056 |
sings | +0.054 |
Neg logit contributions | |
guard | -0.047 |
poisoning | -0.047 |
injuries | -0.046 |
unfortunately | -0.042 |
payoff | -0.041 |
Token | ak |
Token ID | 461 |
Feature activation | -3.23e-02 |
Pos logit contributions | |
aling | +0.212 |
sings | +0.210 |
alling | +0.203 |
dances | +0.195 |
astics | +0.189 |
Neg logit contributions | |
guard | -0.205 |
poisoning | -0.201 |
t | -0.193 |
injuries | -0.185 |
last | -0.175 |
Token | . |
Token ID | 13 |
Feature activation | -1.31e-02 |
Pos logit contributions | |
aling | +0.071 |
alling | +0.065 |
astics | +0.063 |
irl | +0.062 |
sings | +0.061 |
Neg logit contributions | |
guard | -0.029 |
poisoning | -0.029 |
injuries | -0.027 |
unfortunately | -0.025 |
ho | -0.024 |
Token | Fro |
Token ID | 9734 |
Feature activation | -5.62e-02 |
Pos logit contributions | |
aling | +0.027 |
alling | +0.024 |
astics | +0.023 |
irl | +0.023 |
ophone | +0.022 |
Neg logit contributions | |
poisoning | -0.014 |
guard | -0.014 |
injuries | -0.013 |
unfortunately | -0.012 |
ho | -0.012 |
Token | ggy |
Token ID | 19970 |
Feature activation | -5.24e-02 |
Pos logit contributions | |
aling | +0.092 |
alling | +0.081 |
astics | +0.077 |
irl | +0.077 |
sings | +0.072 |
Neg logit contributions | |
guard | -0.082 |
poisoning | -0.081 |
injuries | -0.080 |
unfortunately | -0.074 |
payoff | -0.074 |
Token | said |
Token ID | 531 |
Feature activation | +3.26e-02 |
Pos logit contributions | |
aling | +0.108 |
alling | +0.100 |
astics | +0.093 |
irl | +0.092 |
sings | +0.087 |
Neg logit contributions | |
guard | -0.055 |
poisoning | -0.054 |
injuries | -0.052 |
unfortunately | -0.048 |
ho | -0.047 |
Token | for |
Token ID | 329 |
Feature activation | +2.08e-02 |
Pos logit contributions | |
aling | +0.048 |
alling | +0.043 |
astics | +0.042 |
sings | +0.042 |
irl | +0.041 |
Neg logit contributions | |
guard | -0.029 |
poisoning | -0.028 |
injuries | -0.026 |
last | -0.026 |
t | -0.024 |
Token | a |
Token ID | 257 |
Feature activation | +1.44e-02 |
Pos logit contributions | |
guard | +0.026 |
poisoning | +0.026 |
injuries | +0.025 |
unfortunately | +0.023 |
payoff | +0.023 |
Neg logit contributions | |
aling | -0.038 |
alling | -0.035 |
astics | -0.033 |
irl | -0.033 |
sings | -0.031 |
Token | ride |
Token ID | 6594 |
Feature activation | -3.19e-02 |
Pos logit contributions | |
guard | +0.021 |
poisoning | +0.021 |
injuries | +0.020 |
unfortunately | +0.019 |
payoff | +0.019 |
Neg logit contributions | |
aling | -0.024 |
alling | -0.021 |
astics | -0.020 |
irl | -0.020 |
sings | -0.019 |
Token | in |
Token ID | 287 |
Feature activation | -1.62e-02 |
Pos logit contributions | |
aling | +0.059 |
alling | +0.053 |
astics | +0.051 |
irl | +0.050 |
sings | +0.047 |
Neg logit contributions | |
poisoning | -0.039 |
injuries | -0.039 |
guard | -0.039 |
payoff | -0.035 |
unfortunately | -0.035 |
Token | the |
Token ID | 262 |
Feature activation | -2.17e-02 |
Pos logit contributions | |
aling | +0.027 |
alling | +0.025 |
irl | +0.025 |
astics | +0.023 |
c | +0.022 |
Neg logit contributions | |
poisoning | -0.022 |
department | -0.021 |
guard | -0.021 |
injuries | -0.021 |
ho | -0.020 |
Token | kay |
Token ID | 34681 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
aling | +0.043 |
alling | +0.039 |
astics | +0.039 |
irl | +0.037 |
ophone | +0.037 |
Neg logit contributions | |
guard | -0.026 |
poisoning | -0.023 |
last | -0.022 |
injuries | -0.022 |
payoff | -0.022 |
Token | ak |
Token ID | 461 |
Feature activation | -2.25e-02 |
Pos logit contributions | |
sings | +0.234 |
dances | +0.213 |
aling | +0.210 |
alling | +0.205 |
presses | +0.201 |
Neg logit contributions | |
guard | -0.196 |
t | -0.179 |
poisoning | -0.174 |
ll | -0.157 |
last | -0.154 |
Token | and |
Token ID | 290 |
Feature activation | -4.52e-02 |
Pos logit contributions | |
aling | +0.043 |
alling | +0.039 |
astics | +0.036 |
irl | +0.035 |
ophone | +0.035 |
Neg logit contributions | |
injuries | -0.027 |
poisoning | -0.026 |
guard | -0.026 |
payoff | -0.024 |
unfortunately | -0.023 |
Token | her |
Token ID | 607 |
Feature activation | -8.92e-03 |
Pos logit contributions | |
aling | +0.096 |
alling | +0.088 |
astics | +0.083 |
irl | +0.083 |
sings | +0.079 |
Neg logit contributions | |
guard | -0.044 |
poisoning | -0.043 |
injuries | -0.042 |
unfortunately | -0.039 |
payoff | -0.038 |
Token | mom |
Token ID | 1995 |
Feature activation | +2.99e-02 |
Pos logit contributions | |
aling | +0.017 |
irl | +0.017 |
alling | +0.016 |
nic | +0.016 |
astics | +0.016 |
Neg logit contributions | |
guard | -0.008 |
poisoning | -0.008 |
injuries | -0.007 |
unfortunately | -0.007 |
payoff | -0.007 |
Token | said |
Token ID | 531 |
Feature activation | +2.55e-02 |
Pos logit contributions | |
injuries | +0.033 |
poisoning | +0.032 |
guard | +0.032 |
unfortunately | +0.029 |
payoff | +0.029 |
Neg logit contributions | |
aling | -0.059 |
alling | -0.054 |
astics | -0.051 |
irl | -0.051 |
sings | -0.049 |
Token | forth |
Token ID | 6071 |
Feature activation | -1.46e-03 |
Pos logit contributions | |
aling | +0.082 |
alling | +0.077 |
astics | +0.073 |
essay | +0.071 |
vel | +0.070 |
Neg logit contributions | |
poisoning | -0.059 |
guard | -0.059 |
though | -0.057 |
last | -0.056 |
while | -0.055 |
Token | in |
Token ID | 287 |
Feature activation | -2.39e-02 |
Pos logit contributions | |
aling | +0.003 |
alling | +0.003 |
irl | +0.003 |
astics | +0.003 |
sings | +0.003 |
Neg logit contributions | |
guard | -0.001 |
injuries | -0.001 |
poisoning | -0.001 |
payoff | -0.001 |
unfortunately | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | -2.95e-02 |
Pos logit contributions | |
aling | +0.040 |
alling | +0.036 |
irl | +0.033 |
astics | +0.032 |
sings | +0.031 |
Neg logit contributions | |
guard | -0.035 |
poisoning | -0.034 |
injuries | -0.032 |
t | -0.031 |
ho | -0.031 |
Token | sky |
Token ID | 6766 |
Feature activation | -1.91e-02 |
Pos logit contributions | |
aling | +0.060 |
alling | +0.054 |
astics | +0.054 |
irl | +0.051 |
ophone | +0.050 |
Neg logit contributions | |
guard | -0.032 |
injuries | -0.032 |
last | -0.031 |
poisoning | -0.030 |
payoff | -0.029 |
Token | , |
Token ID | 11 |
Feature activation | +2.89e-03 |
Pos logit contributions | |
aling | +0.041 |
alling | +0.037 |
astics | +0.035 |
irl | +0.035 |
sings | +0.034 |
Neg logit contributions | |
guard | -0.019 |
poisoning | -0.018 |
injuries | -0.018 |
unfortunately | -0.016 |
payoff | -0.016 |
Token | spir |
Token ID | 9158 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
poisoning | +0.004 |
guard | +0.003 |
injuries | +0.003 |
unfortunately | +0.003 |
last | +0.003 |
Neg logit contributions | |
aling | -0.005 |
alling | -0.005 |
astics | -0.005 |
irl | -0.005 |
ophone | -0.004 |
Token | alling |
Token ID | 9221 |
Feature activation | -5.34e-02 |
Pos logit contributions | |
astics | +0.227 |
irl | +0.222 |
essay | +0.202 |
aling | +0.198 |
sings | +0.196 |
Neg logit contributions | |
guard | -0.169 |
last | -0.167 |
poisoning | -0.165 |
ho | -0.165 |
t | -0.164 |
Token | higher |
Token ID | 2440 |
Feature activation | -2.02e-02 |
Pos logit contributions | |
aling | +0.090 |
alling | +0.080 |
astics | +0.074 |
irl | +0.073 |
sings | +0.070 |
Neg logit contributions | |
injuries | -0.075 |
poisoning | -0.075 |
guard | -0.075 |
unfortunately | -0.068 |
payoff | -0.068 |
Token | and |
Token ID | 290 |
Feature activation | -5.60e-02 |
Pos logit contributions | |
aling | +0.031 |
alling | +0.027 |
irl | +0.026 |
spir | +0.026 |
sings | +0.026 |
Neg logit contributions | |
poisoning | -0.030 |
injuries | -0.030 |
guard | -0.030 |
mercy | -0.028 |
unfortunately | -0.027 |
Token | higher |
Token ID | 2440 |
Feature activation | -1.97e-02 |
Pos logit contributions | |
aling | +0.091 |
astics | +0.086 |
alling | +0.084 |
irl | +0.082 |
essay | +0.081 |
Neg logit contributions | |
last | -0.076 |
guard | -0.076 |
payoff | -0.072 |
though | -0.071 |
poisoning | -0.070 |
Token | . |
Token ID | 13 |
Feature activation | -2.21e-02 |
Pos logit contributions | |
aling | +0.032 |
alling | +0.028 |
irl | +0.027 |
sings | +0.026 |
astics | +0.026 |
Neg logit contributions | |
poisoning | -0.028 |
guard | -0.028 |
injuries | -0.027 |
unfortunately | -0.025 |
department | -0.025 |
Token | can |
Token ID | 460 |
Feature activation | +4.05e-02 |
Pos logit contributions | |
injuries | +0.065 |
guard | +0.065 |
poisoning | +0.064 |
unfortunately | +0.058 |
payoff | +0.058 |
Neg logit contributions | |
aling | -0.091 |
alling | -0.081 |
astics | -0.077 |
irl | -0.077 |
sings | -0.074 |
Token | we |
Token ID | 356 |
Feature activation | +3.11e-02 |
Pos logit contributions | |
guard | +0.037 |
poisoning | +0.037 |
injuries | +0.036 |
unfortunately | +0.032 |
payoff | +0.031 |
Neg logit contributions | |
aling | -0.088 |
alling | -0.081 |
astics | -0.077 |
irl | -0.077 |
sings | -0.074 |
Token | go |
Token ID | 467 |
Feature activation | -3.31e-02 |
Pos logit contributions | |
guard | +0.051 |
poisoning | +0.051 |
injuries | +0.050 |
unfortunately | +0.047 |
payoff | +0.047 |
Neg logit contributions | |
aling | -0.045 |
alling | -0.039 |
astics | -0.036 |
irl | -0.036 |
sings | -0.034 |
Token | in |
Token ID | 287 |
Feature activation | -1.10e-02 |
Pos logit contributions | |
aling | +0.064 |
alling | +0.058 |
astics | +0.056 |
irl | +0.055 |
sings | +0.054 |
Neg logit contributions | |
guard | -0.038 |
poisoning | -0.037 |
injuries | -0.036 |
last | -0.033 |
unfortunately | -0.032 |
Token | the |
Token ID | 262 |
Feature activation | -3.14e-02 |
Pos logit contributions | |
aling | +0.019 |
alling | +0.017 |
irl | +0.017 |
astics | +0.016 |
c | +0.015 |
Neg logit contributions | |
injuries | -0.015 |
poisoning | -0.015 |
ho | -0.014 |
guard | -0.014 |
t | -0.013 |
Token | kay |
Token ID | 34681 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
aling | +0.061 |
alling | +0.055 |
astics | +0.052 |
irl | +0.051 |
ophone | +0.050 |
Neg logit contributions | |
guard | -0.039 |
poisoning | -0.037 |
injuries | -0.036 |
last | -0.034 |
payoff | -0.034 |
Token | ak |
Token ID | 461 |
Feature activation | -2.73e-02 |
Pos logit contributions | |
sings | +0.223 |
aling | +0.211 |
dances | +0.207 |
alling | +0.204 |
spir | +0.192 |
Neg logit contributions | |
guard | -0.201 |
t | -0.181 |
poisoning | -0.180 |
last | -0.161 |
injuries | -0.159 |
Token | ?" |
Token ID | 1701 |
Feature activation | +2.30e-03 |
Pos logit contributions | |
aling | +0.052 |
alling | +0.047 |
astics | +0.044 |
irl | +0.044 |
ophone | +0.042 |
Neg logit contributions | |
injuries | -0.034 |
poisoning | -0.032 |
guard | -0.032 |
payoff | -0.030 |
unfortunately | -0.029 |
Token | Her |
Token ID | 2332 |
Feature activation | -9.20e-03 |
Pos logit contributions | |
guard | +0.003 |
poisoning | +0.003 |
injuries | +0.003 |
unfortunately | +0.002 |
payoff | +0.002 |
Neg logit contributions | |
aling | -0.004 |
alling | -0.004 |
astics | -0.004 |
irl | -0.004 |
sings | -0.004 |
Token | dad |
Token ID | 9955 |
Feature activation | +6.75e-03 |
Pos logit contributions | |
aling | +0.021 |
alling | +0.020 |
irl | +0.020 |
astics | +0.020 |
sings | +0.019 |
Neg logit contributions | |
poisoning | -0.006 |
guard | -0.006 |
injuries | -0.006 |
unfortunately | -0.005 |
payoff | -0.005 |
Token | said |
Token ID | 531 |
Feature activation | +1.93e-02 |
Pos logit contributions | |
guard | +0.007 |
poisoning | +0.006 |
unfortunately | +0.006 |
injuries | +0.005 |
ho | +0.005 |
Neg logit contributions | |
aling | -0.015 |
alling | -0.014 |
astics | -0.013 |
irl | -0.013 |
atters | -0.012 |
Token | river |
Token ID | 7850 |
Feature activation | -3.94e-02 |
Pos logit contributions | |
aling | +0.021 |
alling | +0.019 |
astics | +0.018 |
irl | +0.017 |
sings | +0.017 |
Neg logit contributions | |
injuries | -0.018 |
poisoning | -0.018 |
guard | -0.018 |
unfortunately | -0.017 |
payoff | -0.016 |
Token | . |
Token ID | 13 |
Feature activation | -4.37e-03 |
Pos logit contributions | |
aling | +0.075 |
alling | +0.067 |
astics | +0.064 |
irl | +0.063 |
sings | +0.060 |
Neg logit contributions | |
injuries | -0.047 |
poisoning | -0.046 |
guard | -0.046 |
payoff | -0.042 |
unfortunately | -0.042 |
Token | They |
Token ID | 1119 |
Feature activation | -9.87e-04 |
Pos logit contributions | |
aling | +0.007 |
alling | +0.006 |
astics | +0.006 |
irl | +0.006 |
sings | +0.005 |
Neg logit contributions | |
poisoning | -0.006 |
guard | -0.006 |
injuries | -0.006 |
unfortunately | -0.006 |
payoff | -0.006 |
Token | saw |
Token ID | 2497 |
Feature activation | -8.15e-03 |
Pos logit contributions | |
aling | +0.002 |
alling | +0.002 |
astics | +0.001 |
irl | +0.001 |
sings | +0.001 |
Neg logit contributions | |
guard | -0.001 |
poisoning | -0.001 |
injuries | -0.001 |
unfortunately | -0.001 |
payoff | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | -1.37e-02 |
Pos logit contributions | |
aling | +0.010 |
alling | +0.009 |
sings | +0.009 |
irl | +0.009 |
astics | +0.008 |
Neg logit contributions | |
injuries | -0.014 |
poisoning | -0.014 |
guard | -0.014 |
payoff | -0.013 |
t | -0.013 |
Token | kay |
Token ID | 34681 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
aling | +0.024 |
alling | +0.021 |
irl | +0.020 |
astics | +0.020 |
sings | +0.019 |
Neg logit contributions | |
poisoning | -0.018 |
guard | -0.018 |
injuries | -0.018 |
unfortunately | -0.016 |
payoff | -0.016 |
Token | ak |
Token ID | 461 |
Feature activation | -2.84e-02 |
Pos logit contributions | |
aling | +0.202 |
sings | +0.199 |
alling | +0.191 |
astics | +0.181 |
dances | +0.179 |
Neg logit contributions | |
guard | -0.224 |
poisoning | -0.209 |
t | -0.209 |
last | -0.185 |
injuries | -0.184 |
Token | on |
Token ID | 319 |
Feature activation | -2.58e-02 |
Pos logit contributions | |
aling | +0.061 |
alling | +0.056 |
astics | +0.054 |
irl | +0.053 |
sings | +0.051 |
Neg logit contributions | |
guard | -0.027 |
poisoning | -0.026 |
injuries | -0.026 |
unfortunately | -0.023 |
payoff | -0.022 |
Token | the |
Token ID | 262 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
aling | +0.045 |
alling | +0.040 |
irl | +0.039 |
astics | +0.038 |
sings | +0.036 |
Neg logit contributions | |
guard | -0.034 |
injuries | -0.034 |
poisoning | -0.033 |
department | -0.030 |
t | -0.030 |
Token | shore |
Token ID | 15191 |
Feature activation | -3.59e-02 |
Pos logit contributions | |
aling | +0.018 |
alling | +0.016 |
spir | +0.015 |
zoom | +0.014 |
nic | +0.014 |
Neg logit contributions | |
injuries | -0.025 |
guard | -0.024 |
poisoning | -0.024 |
unfortunately | -0.024 |
payoff | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | -1.65e-02 |
Pos logit contributions | |
aling | +0.064 |
alling | +0.057 |
astics | +0.053 |
irl | +0.052 |
ophone | +0.051 |
Neg logit contributions | |
injuries | -0.046 |
poisoning | -0.044 |
guard | -0.043 |
payoff | -0.040 |
mercy | -0.040 |
Token | hopped |
Token ID | 45230 |
Feature activation | -7.79e-02 |
Pos logit contributions | |
aling | +0.001 |
alling | +0.001 |
astics | +0.001 |
etition | +0.001 |
atters | +0.001 |
Neg logit contributions | |
guard | -0.000 |
last | -0.000 |
poisoning | -0.000 |
now | -0.000 |
to | -0.000 |
Token | on |
Token ID | 319 |
Feature activation | -1.18e-02 |
Pos logit contributions | |
aling | +0.136 |
alling | +0.121 |
astics | +0.116 |
irl | +0.113 |
sings | +0.109 |
Neg logit contributions | |
guard | -0.107 |
poisoning | -0.105 |
injuries | -0.099 |
unfortunately | -0.095 |
payoff | -0.094 |
Token | , |
Token ID | 11 |
Feature activation | +3.43e-03 |
Pos logit contributions | |
aling | +0.026 |
alling | +0.024 |
astics | +0.023 |
ophone | +0.022 |
irl | +0.022 |
Neg logit contributions | |
poisoning | -0.011 |
injuries | -0.010 |
guard | -0.010 |
last | -0.009 |
unfortunately | -0.009 |
Token | and |
Token ID | 290 |
Feature activation | -4.48e-02 |
Pos logit contributions | |
but | +0.004 |
while | +0.004 |
to | +0.004 |
though | +0.004 |
poisoning | +0.004 |
Neg logit contributions | |
aling | -0.006 |
astics | -0.006 |
irl | -0.006 |
alling | -0.006 |
atters | -0.005 |
Token | started |
Token ID | 2067 |
Feature activation | -2.10e-02 |
Pos logit contributions | |
aling | +0.088 |
alling | +0.081 |
astics | +0.079 |
irl | +0.078 |
atters | +0.075 |
Neg logit contributions | |
guard | -0.049 |
poisoning | -0.046 |
unfortunately | -0.044 |
though | -0.043 |
they | -0.043 |
Token | ped |
Token ID | 7190 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
aling | +0.048 |
alling | +0.043 |
irl | +0.043 |
ophone | +0.043 |
astics | +0.043 |
Neg logit contributions | |
to | -0.020 |
that | -0.016 |
while | -0.015 |
they | -0.014 |
guard | -0.013 |
Token | alling |
Token ID | 9221 |
Feature activation | -5.63e-02 |
Pos logit contributions | |
sings | +0.091 |
nic | +0.062 |
sights | +0.061 |
dances | +0.060 |
revers | +0.058 |
Neg logit contributions | |
since | -0.293 |
t | -0.291 |
guard | -0.282 |
then | -0.278 |
last | -0.275 |
Token | . |
Token ID | 13 |
Feature activation | -1.11e-02 |
Pos logit contributions | |
aling | +0.115 |
alling | +0.105 |
astics | +0.100 |
irl | +0.099 |
sings | +0.096 |
Neg logit contributions | |
guard | -0.059 |
poisoning | -0.057 |
injuries | -0.057 |
unfortunately | -0.051 |
payoff | -0.050 |
Token | She |
Token ID | 1375 |
Feature activation | -4.14e-03 |
Pos logit contributions | |
aling | +0.021 |
alling | +0.019 |
astics | +0.018 |
irl | +0.017 |
sings | +0.017 |
Neg logit contributions | |
poisoning | -0.013 |
guard | -0.012 |
injuries | -0.012 |
t | -0.012 |
unfortunately | -0.012 |
Token | felt |
Token ID | 2936 |
Feature activation | -4.76e-02 |
Pos logit contributions | |
aling | +0.008 |
astics | +0.007 |
alling | +0.007 |
etition | +0.007 |
atters | +0.007 |
Neg logit contributions | |
guard | -0.005 |
last | -0.004 |
t | -0.004 |
to | -0.004 |
while | -0.004 |
Token | like |
Token ID | 588 |
Feature activation | -9.21e-03 |
Pos logit contributions | |
aling | +0.103 |
astics | +0.095 |
alling | +0.094 |
irl | +0.092 |
ophone | +0.091 |
Neg logit contributions | |
poisoning | -0.039 |
guard | -0.039 |
unfortunately | -0.036 |
though | -0.036 |
injuries | -0.034 |
Token | and |
Token ID | 290 |
Feature activation | -3.77e-03 |
Pos logit contributions | |
aling | +0.055 |
alling | +0.051 |
astics | +0.050 |
irl | +0.049 |
atters | +0.046 |
Neg logit contributions | |
guard | -0.020 |
poisoning | -0.017 |
last | -0.016 |
payoff | -0.015 |
ho | -0.015 |
Token | Anne |
Token ID | 15397 |
Feature activation | +3.77e-03 |
Pos logit contributions | |
aling | +0.005 |
alling | +0.005 |
sings | +0.004 |
astics | +0.004 |
irl | +0.004 |
Neg logit contributions | |
poisoning | -0.006 |
injuries | -0.006 |
guard | -0.006 |
unfortunately | -0.006 |
ho | -0.006 |
Token | jumped |
Token ID | 11687 |
Feature activation | -6.56e-02 |
Pos logit contributions | |
guard | +0.005 |
last | +0.005 |
poisoning | +0.005 |
help | +0.005 |
ho | +0.004 |
Neg logit contributions | |
aling | -0.006 |
astics | -0.006 |
alling | -0.006 |
irl | -0.005 |
sings | -0.005 |
Token | in |
Token ID | 287 |
Feature activation | -1.77e-02 |
Pos logit contributions | |
aling | +0.115 |
astics | +0.101 |
alling | +0.100 |
sings | +0.097 |
irl | +0.096 |
Neg logit contributions | |
guard | -0.090 |
poisoning | -0.085 |
injuries | -0.081 |
last | -0.080 |
unfortunately | -0.079 |
Token | the |
Token ID | 262 |
Feature activation | -2.67e-02 |
Pos logit contributions | |
aling | +0.028 |
alling | +0.026 |
irl | +0.024 |
ped | +0.023 |
astics | +0.023 |
Neg logit contributions | |
guard | -0.026 |
poisoning | -0.025 |
t | -0.023 |
injuries | -0.023 |
ho | -0.022 |
Token | kay |
Token ID | 34681 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
aling | +0.052 |
astics | +0.048 |
vel | +0.047 |
alling | +0.046 |
irl | +0.046 |
Neg logit contributions | |
guard | -0.032 |
last | -0.030 |
payoff | -0.028 |
injuries | -0.027 |
poisoning | -0.026 |
Token | ak |
Token ID | 461 |
Feature activation | -2.73e-02 |
Pos logit contributions | |
sings | +0.237 |
dances | +0.222 |
aling | +0.210 |
alling | +0.202 |
astics | +0.200 |
Neg logit contributions | |
guard | -0.186 |
t | -0.170 |
poisoning | -0.159 |
less | -0.149 |
ll | -0.144 |
Token | and |
Token ID | 290 |
Feature activation | -5.60e-02 |
Pos logit contributions | |
aling | +0.060 |
alling | +0.055 |
astics | +0.053 |
irl | +0.052 |
sings | +0.050 |
Neg logit contributions | |
guard | -0.024 |
poisoning | -0.024 |
injuries | -0.023 |
unfortunately | -0.021 |
payoff | -0.020 |
Token | padd |
Token ID | 14098 |
Feature activation | -9.06e-02 |
Pos logit contributions | |
aling | +0.105 |
astics | +0.098 |
vel | +0.095 |
alling | +0.095 |
irl | +0.092 |
Neg logit contributions | |
they | -0.069 |
guard | -0.065 |
while | -0.062 |
last | -0.060 |
to | -0.059 |
Token | led |
Token ID | 992 |
Feature activation | -4.10e-02 |
Pos logit contributions | |
aling | +0.157 |
alling | +0.142 |
astics | +0.140 |
irl | +0.133 |
sings | +0.126 |
Neg logit contributions | |
guard | -0.127 |
poisoning | -0.117 |
injuries | -0.113 |
last | -0.108 |
payoff | -0.106 |
Token | away |
Token ID | 1497 |
Feature activation | +1.94e-02 |
Pos logit contributions | |
aling | +0.074 |
alling | +0.066 |
astics | +0.065 |
irl | +0.062 |
ophone | +0.059 |
Neg logit contributions | |
guard | -0.052 |
poisoning | -0.051 |
injuries | -0.050 |
last | -0.048 |
payoff | -0.048 |
Token | day |
Token ID | 1110 |
Feature activation | +6.54e-02 |
Pos logit contributions | |
injuries | +0.039 |
poisoning | +0.038 |
guard | +0.037 |
unfortunately | +0.034 |
payoff | +0.034 |
Neg logit contributions | |
aling | -0.059 |
alling | -0.053 |
irl | -0.051 |
astics | -0.050 |
sings | -0.048 |
Token | , |
Token ID | 11 |
Feature activation | +2.40e-02 |
Pos logit contributions | |
injuries | +0.085 |
guard | +0.077 |
poisoning | +0.076 |
mercy | +0.074 |
payoff | +0.073 |
Neg logit contributions | |
irl | -0.108 |
aling | -0.107 |
astics | -0.101 |
irled | -0.101 |
dances | -0.101 |
Token | they |
Token ID | 484 |
Feature activation | -7.71e-03 |
Pos logit contributions | |
injuries | +0.036 |
mercy | +0.036 |
guard | +0.033 |
ache | +0.032 |
payoff | +0.031 |
Neg logit contributions | |
irl | -0.034 |
plays | -0.031 |
irled | -0.030 |
dances | -0.030 |
astics | -0.029 |
Token | saw |
Token ID | 2497 |
Feature activation | -8.82e-03 |
Pos logit contributions | |
aling | +0.012 |
alling | +0.011 |
irl | +0.010 |
astics | +0.010 |
sings | +0.009 |
Neg logit contributions | |
injuries | -0.012 |
poisoning | -0.012 |
guard | -0.012 |
unfortunately | -0.011 |
mercy | -0.011 |
Token | a |
Token ID | 257 |
Feature activation | -1.51e-02 |
Pos logit contributions | |
aling | +0.013 |
alling | +0.011 |
irl | +0.011 |
sings | +0.010 |
astics | +0.010 |
Neg logit contributions | |
injuries | -0.014 |
guard | -0.014 |
poisoning | -0.014 |
payoff | -0.013 |
t | -0.013 |
Token | kay |
Token ID | 34681 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
aling | +0.029 |
alling | +0.026 |
astics | +0.025 |
irl | +0.024 |
sings | +0.023 |
Neg logit contributions | |
guard | -0.019 |
poisoning | -0.018 |
injuries | -0.018 |
unfortunately | -0.017 |
payoff | -0.016 |
Token | ak |
Token ID | 461 |
Feature activation | -2.86e-02 |
Pos logit contributions | |
sings | +0.209 |
aling | +0.203 |
alling | +0.194 |
dances | +0.191 |
astics | +0.188 |
Neg logit contributions | |
guard | -0.212 |
t | -0.195 |
poisoning | -0.191 |
last | -0.171 |
injuries | -0.171 |
Token | on |
Token ID | 319 |
Feature activation | -2.30e-02 |
Pos logit contributions | |
aling | +0.060 |
alling | +0.055 |
astics | +0.054 |
irl | +0.053 |
sings | +0.052 |
Neg logit contributions | |
guard | -0.028 |
poisoning | -0.027 |
injuries | -0.025 |
unfortunately | -0.024 |
ho | -0.023 |
Token | the |
Token ID | 262 |
Feature activation | -1.44e-02 |
Pos logit contributions | |
aling | +0.041 |
alling | +0.037 |
irl | +0.035 |
astics | +0.035 |
sings | +0.033 |
Neg logit contributions | |
guard | -0.029 |
injuries | -0.029 |
poisoning | -0.029 |
t | -0.027 |
department | -0.026 |
Token | shore |
Token ID | 15191 |
Feature activation | -3.17e-02 |
Pos logit contributions | |
aling | +0.020 |
alling | +0.018 |
astics | +0.016 |
irl | +0.016 |
spir | +0.016 |
Neg logit contributions | |
injuries | -0.022 |
guard | -0.022 |
poisoning | -0.022 |
unfortunately | -0.021 |
payoff | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | -2.61e-02 |
Pos logit contributions | |
aling | +0.058 |
alling | +0.052 |
astics | +0.048 |
irl | +0.047 |
ophone | +0.046 |
Neg logit contributions | |
injuries | -0.040 |
poisoning | -0.039 |
guard | -0.038 |
payoff | -0.035 |
unfortunately | -0.035 |
Token | on |
Token ID | 319 |
Feature activation | -4.22e-03 |
Pos logit contributions | |
aling | +0.122 |
alling | +0.108 |
astics | +0.103 |
irl | +0.101 |
sings | +0.097 |
Neg logit contributions | |
guard | -0.111 |
poisoning | -0.108 |
injuries | -0.105 |
unfortunately | -0.099 |
payoff | -0.097 |
Token | their |
Token ID | 511 |
Feature activation | -6.21e-03 |
Pos logit contributions | |
aling | +0.008 |
alling | +0.007 |
astics | +0.007 |
irl | +0.007 |
sings | +0.007 |
Neg logit contributions | |
guard | -0.005 |
poisoning | -0.005 |
injuries | -0.005 |
unfortunately | -0.004 |
payoff | -0.004 |
Token | bikes |
Token ID | 16715 |
Feature activation | -3.38e-02 |
Pos logit contributions | |
aling | +0.010 |
alling | +0.009 |
astics | +0.008 |
irl | +0.008 |
sings | +0.008 |
Neg logit contributions | |
guard | -0.010 |
poisoning | -0.009 |
injuries | -0.009 |
unfortunately | -0.009 |
payoff | -0.009 |
Token | and |
Token ID | 290 |
Feature activation | -4.79e-02 |
Pos logit contributions | |
aling | +0.068 |
alling | +0.062 |
astics | +0.059 |
irl | +0.058 |
sings | +0.056 |
Neg logit contributions | |
guard | -0.036 |
poisoning | -0.036 |
injuries | -0.035 |
unfortunately | -0.032 |
payoff | -0.031 |
Token | started |
Token ID | 2067 |
Feature activation | -2.36e-02 |
Pos logit contributions | |
aling | +0.083 |
alling | +0.074 |
astics | +0.070 |
irl | +0.070 |
sings | +0.065 |
Neg logit contributions | |
guard | -0.065 |
poisoning | -0.064 |
injuries | -0.063 |
unfortunately | -0.059 |
payoff | -0.058 |
Token | ped |
Token ID | 7190 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
aling | +0.056 |
alling | +0.051 |
irl | +0.049 |
astics | +0.049 |
ophone | +0.049 |
Neg logit contributions | |
guard | -0.016 |
injuries | -0.016 |
poisoning | -0.015 |
payoff | -0.013 |
to | -0.013 |
Token | aling |
Token ID | 4272 |
Feature activation | -4.38e-02 |
Pos logit contributions | |
sings | +0.088 |
dances | +0.057 |
sights | +0.049 |
azes | +0.045 |
nic | +0.042 |
Neg logit contributions | |
guard | -0.324 |
t | -0.317 |
last | -0.294 |
Not | -0.289 |
since | -0.288 |
Token | . |
Token ID | 13 |
Feature activation | -2.71e-02 |
Pos logit contributions | |
aling | +0.083 |
alling | +0.076 |
astics | +0.071 |
irl | +0.071 |
sings | +0.067 |
Neg logit contributions | |
poisoning | -0.052 |
guard | -0.052 |
injuries | -0.051 |
unfortunately | -0.047 |
payoff | -0.046 |
Token |
|
Token ID | 198 |
Feature activation | +2.22e-02 |
Pos logit contributions | |
aling | +0.049 |
alling | +0.045 |
astics | +0.042 |
irl | +0.042 |
sings | +0.040 |
Neg logit contributions | |
guard | -0.033 |
poisoning | -0.033 |
injuries | -0.033 |
unfortunately | -0.031 |
payoff | -0.030 |
Token |
|
Token ID | 198 |
Feature activation | +1.71e-02 |
Pos logit contributions | |
poisoning | +0.027 |
guard | +0.026 |
injuries | +0.026 |
payoff | +0.023 |
ho | +0.023 |
Neg logit contributions | |
aling | -0.043 |
alling | -0.039 |
irl | -0.037 |
astics | -0.037 |
sings | -0.034 |
Token | When |
Token ID | 2215 |
Feature activation | +5.24e-02 |
Pos logit contributions | |
poisoning | +0.018 |
guard | +0.018 |
injuries | +0.018 |
payoff | +0.016 |
unfortunately | +0.016 |
Neg logit contributions | |
aling | -0.035 |
alling | -0.032 |
astics | -0.030 |
irl | -0.030 |
sings | -0.029 |
Token | saw |
Token ID | 2497 |
Feature activation | -6.82e-03 |
Pos logit contributions | |
aling | +0.084 |
alling | +0.077 |
astics | +0.073 |
irl | +0.072 |
sings | +0.068 |
Neg logit contributions | |
guard | -0.047 |
poisoning | -0.047 |
injuries | -0.044 |
unfortunately | -0.041 |
payoff | -0.040 |
Token | a |
Token ID | 257 |
Feature activation | -1.92e-02 |
Pos logit contributions | |
aling | +0.009 |
alling | +0.007 |
sings | +0.007 |
irl | +0.007 |
astics | +0.007 |
Neg logit contributions | |
injuries | -0.012 |
guard | -0.011 |
poisoning | -0.011 |
t | -0.011 |
payoff | -0.011 |
Token | big |
Token ID | 1263 |
Feature activation | -3.77e-02 |
Pos logit contributions | |
aling | +0.037 |
alling | +0.033 |
astics | +0.031 |
irl | +0.030 |
sings | +0.029 |
Neg logit contributions | |
guard | -0.024 |
poisoning | -0.023 |
injuries | -0.022 |
unfortunately | -0.021 |
payoff | -0.020 |
Token | , |
Token ID | 11 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
aling | +0.075 |
alling | +0.068 |
astics | +0.065 |
irl | +0.063 |
ophone | +0.060 |
Neg logit contributions | |
guard | -0.043 |
poisoning | -0.043 |
injuries | -0.041 |
unfortunately | -0.038 |
payoff | -0.037 |
Token | red |
Token ID | 2266 |
Feature activation | -4.52e-02 |
Pos logit contributions | |
aling | +0.023 |
alling | +0.021 |
astics | +0.020 |
aled | +0.019 |
irled | +0.019 |
Neg logit contributions | |
guard | -0.014 |
poisoning | -0.013 |
last | -0.012 |
injuries | -0.012 |
but | -0.012 |
Token | kay |
Token ID | 34681 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
aling | +0.081 |
alling | +0.073 |
irl | +0.070 |
astics | +0.069 |
sings | +0.066 |
Neg logit contributions | |
injuries | -0.056 |
poisoning | -0.056 |
guard | -0.056 |
unfortunately | -0.051 |
payoff | -0.050 |
Token | ak |
Token ID | 461 |
Feature activation | -3.69e-02 |
Pos logit contributions | |
sings | +0.213 |
aling | +0.207 |
alling | +0.201 |
dances | +0.196 |
astics | +0.189 |
Neg logit contributions | |
guard | -0.202 |
t | -0.194 |
poisoning | -0.190 |
ho | -0.170 |
last | -0.162 |
Token | . |
Token ID | 13 |
Feature activation | -2.08e-02 |
Pos logit contributions | |
aling | +0.086 |
alling | +0.080 |
astics | +0.078 |
irl | +0.077 |
sings | +0.075 |
Neg logit contributions | |
guard | -0.029 |
poisoning | -0.027 |
that | -0.026 |
t | -0.025 |
ho | -0.024 |
Token | Lily |
Token ID | 20037 |
Feature activation | -2.36e-02 |
Pos logit contributions | |
aling | +0.038 |
alling | +0.034 |
astics | +0.032 |
irl | +0.032 |
sings | +0.030 |
Neg logit contributions | |
guard | -0.027 |
poisoning | -0.027 |
injuries | -0.026 |
unfortunately | -0.024 |
payoff | -0.024 |
Token | said |
Token ID | 531 |
Feature activation | +3.06e-02 |
Pos logit contributions | |
aling | +0.046 |
alling | +0.043 |
astics | +0.039 |
irl | +0.038 |
sings | +0.037 |
Neg logit contributions | |
guard | -0.027 |
poisoning | -0.027 |
unfortunately | -0.025 |
injuries | -0.024 |
ho | -0.024 |
Token | , |
Token ID | 11 |
Feature activation | +3.86e-02 |
Pos logit contributions | |
injuries | +0.032 |
poisoning | +0.029 |
guard | +0.029 |
payoff | +0.025 |
extingu | +0.025 |
Neg logit contributions | |
aling | -0.061 |
alling | -0.056 |
irl | -0.056 |
astics | -0.056 |
sings | -0.055 |
Token | food |
Token ID | 2057 |
Feature activation | +2.34e-02 |
Pos logit contributions | |
guard | +0.011 |
injuries | +0.010 |
poisoning | +0.010 |
ho | +0.010 |
unfortunately | +0.010 |
Neg logit contributions | |
aling | -0.013 |
alling | -0.012 |
astics | -0.011 |
irl | -0.011 |
sings | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | +5.51e-03 |
Pos logit contributions | |
payoff | +0.040 |
injuries | +0.038 |
Girls | +0.037 |
fourteen | +0.037 |
extingu | +0.037 |
Neg logit contributions | |
aling | -0.025 |
astics | -0.023 |
irl | -0.021 |
rides | -0.021 |
nic | -0.020 |
Token | He |
Token ID | 679 |
Feature activation | +2.60e-04 |
Pos logit contributions | |
poisoning | +0.007 |
guard | +0.007 |
injuries | +0.007 |
unfortunately | +0.007 |
t | +0.006 |
Neg logit contributions | |
aling | -0.010 |
alling | -0.009 |
irl | -0.008 |
astics | -0.008 |
sings | -0.008 |
Token | saw |
Token ID | 2497 |
Feature activation | -5.06e-03 |
Pos logit contributions | |
poisoning | +0.000 |
guard | +0.000 |
injuries | +0.000 |
unfortunately | +0.000 |
payoff | +0.000 |
Neg logit contributions | |
aling | -0.000 |
alling | -0.000 |
astics | -0.000 |
irl | -0.000 |
sings | -0.000 |
Token | a |
Token ID | 257 |
Feature activation | -1.82e-02 |
Pos logit contributions | |
aling | +0.005 |
nic | +0.004 |
alling | +0.004 |
rides | +0.004 |
zoom | +0.004 |
Neg logit contributions | |
guard | -0.009 |
injuries | -0.009 |
t | -0.009 |
ho | -0.009 |
poisoning | -0.009 |
Token | kay |
Token ID | 34681 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
aling | +0.025 |
irl | +0.023 |
alling | +0.022 |
astics | +0.022 |
sings | +0.021 |
Neg logit contributions | |
injuries | -0.027 |
guard | -0.027 |
poisoning | -0.026 |
unfortunately | -0.025 |
payoff | -0.025 |
Token | ak |
Token ID | 461 |
Feature activation | -2.69e-02 |
Pos logit contributions | |
sings | +0.216 |
aling | +0.204 |
dances | +0.198 |
alling | +0.197 |
presses | +0.187 |
Neg logit contributions | |
guard | -0.208 |
t | -0.200 |
poisoning | -0.199 |
injuries | -0.171 |
last | -0.169 |
Token | by |
Token ID | 416 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
aling | +0.055 |
alling | +0.051 |
astics | +0.048 |
irl | +0.048 |
sings | +0.046 |
Neg logit contributions | |
guard | -0.028 |
poisoning | -0.028 |
injuries | -0.027 |
unfortunately | -0.025 |
payoff | -0.024 |
Token | the |
Token ID | 262 |
Feature activation | -1.44e-02 |
Pos logit contributions | |
aling | +0.027 |
alling | +0.024 |
astics | +0.023 |
irl | +0.023 |
sings | +0.022 |
Neg logit contributions | |
guard | -0.016 |
poisoning | -0.016 |
injuries | -0.016 |
unfortunately | -0.014 |
payoff | -0.014 |
Token | river |
Token ID | 7850 |
Feature activation | -3.50e-02 |
Pos logit contributions | |
aling | +0.023 |
alling | +0.020 |
astics | +0.019 |
irl | +0.018 |
sings | +0.018 |
Neg logit contributions | |
guard | -0.020 |
poisoning | -0.020 |
injuries | -0.020 |
unfortunately | -0.019 |
t | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | -8.73e-03 |
Pos logit contributions | |
aling | +0.072 |
alling | +0.066 |
astics | +0.063 |
irl | +0.062 |
sings | +0.060 |
Neg logit contributions | |
guard | -0.036 |
poisoning | -0.036 |
injuries | -0.035 |
unfortunately | -0.031 |
payoff | -0.031 |
Token | to |
Token ID | 284 |
Feature activation | +2.60e-02 |
Pos logit contributions | |
injuries | +0.013 |
poisoning | +0.012 |
guard | +0.012 |
unfortunately | +0.012 |
payoff | +0.010 |
Neg logit contributions | |
aling | -0.047 |
alling | -0.043 |
astics | -0.042 |
irl | -0.041 |
sings | -0.041 |
Token | join |
Token ID | 4654 |
Feature activation | -3.25e-02 |
Pos logit contributions | |
guard | +0.040 |
poisoning | +0.040 |
injuries | +0.039 |
unfortunately | +0.038 |
t | +0.037 |
Neg logit contributions | |
aling | -0.040 |
alling | -0.035 |
astics | -0.033 |
irl | -0.033 |
sings | -0.031 |
Token | them |
Token ID | 606 |
Feature activation | +2.38e-02 |
Pos logit contributions | |
aling | +0.071 |
alling | +0.066 |
astics | +0.062 |
irl | +0.062 |
sings | +0.060 |
Neg logit contributions | |
guard | -0.029 |
poisoning | -0.029 |
injuries | -0.028 |
unfortunately | -0.025 |
payoff | -0.024 |
Token | in |
Token ID | 287 |
Feature activation | -6.98e-03 |
Pos logit contributions | |
injuries | +0.028 |
poisoning | +0.028 |
guard | +0.028 |
unfortunately | +0.026 |
payoff | +0.026 |
Neg logit contributions | |
aling | -0.045 |
alling | -0.040 |
astics | -0.039 |
irl | -0.037 |
sings | -0.037 |
Token | the |
Token ID | 262 |
Feature activation | -2.41e-02 |
Pos logit contributions | |
aling | +0.012 |
alling | +0.011 |
irl | +0.011 |
astics | +0.011 |
sings | +0.010 |
Neg logit contributions | |
poisoning | -0.009 |
guard | -0.009 |
injuries | -0.008 |
unfortunately | -0.008 |
t | -0.008 |
Token | kay |
Token ID | 34681 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
aling | +0.046 |
alling | +0.043 |
astics | +0.039 |
irl | +0.038 |
ophone | +0.038 |
Neg logit contributions | |
guard | -0.030 |
poisoning | -0.028 |
injuries | -0.027 |
last | -0.027 |
payoff | -0.027 |
Token | ak |
Token ID | 461 |
Feature activation | -2.35e-02 |
Pos logit contributions | |
sings | +0.204 |
aling | +0.203 |
alling | +0.196 |
dances | +0.187 |
etition | +0.181 |
Neg logit contributions | |
guard | -0.209 |
t | -0.179 |
poisoning | -0.179 |
last | -0.165 |
injuries | -0.163 |
Token | . |
Token ID | 13 |
Feature activation | -3.96e-02 |
Pos logit contributions | |
aling | +0.046 |
alling | +0.041 |
astics | +0.039 |
irl | +0.038 |
sings | +0.037 |
Neg logit contributions | |
injuries | -0.027 |
poisoning | -0.027 |
guard | -0.026 |
unfortunately | -0.024 |
payoff | -0.024 |
Token |
|
Token ID | 198 |
Feature activation | +1.69e-02 |
Pos logit contributions | |
aling | +0.057 |
alling | +0.049 |
astics | +0.047 |
irl | +0.045 |
sings | +0.045 |
Neg logit contributions | |
injuries | -0.066 |
poisoning | -0.066 |
guard | -0.065 |
unfortunately | -0.060 |
payoff | -0.060 |
Token |
|
Token ID | 198 |
Feature activation | +1.35e-02 |
Pos logit contributions | |
poisoning | +0.018 |
guard | +0.018 |
injuries | +0.018 |
unfortunately | +0.016 |
payoff | +0.016 |
Neg logit contributions | |
aling | -0.034 |
alling | -0.031 |
irl | -0.030 |
astics | -0.030 |
sings | -0.028 |
Token | " |
Token ID | 1 |
Feature activation | +2.51e-02 |
Pos logit contributions | |
poisoning | +0.015 |
injuries | +0.015 |
guard | +0.014 |
unfortunately | +0.013 |
payoff | +0.013 |
Neg logit contributions | |
aling | -0.027 |
alling | -0.025 |
astics | -0.024 |
irl | -0.023 |
sings | -0.022 |
Token | Anne |
Token ID | 15397 |
Feature activation | +1.05e-02 |
Pos logit contributions | |
poisoning | +0.016 |
injuries | +0.016 |
guard | +0.016 |
unfortunately | +0.014 |
payoff | +0.014 |
Neg logit contributions | |
aling | -0.015 |
alling | -0.013 |
astics | -0.012 |
sings | -0.012 |
irl | -0.012 |
Token | began |
Token ID | 2540 |
Feature activation | -1.37e-02 |
Pos logit contributions | |
guard | +0.015 |
poisoning | +0.015 |
injuries | +0.015 |
payoff | +0.014 |
unfortunately | +0.014 |
Neg logit contributions | |
aling | -0.017 |
alling | -0.015 |
astics | -0.014 |
irl | -0.014 |
sings | -0.013 |
Token | to |
Token ID | 284 |
Feature activation | -6.74e-03 |
Pos logit contributions | |
aling | +0.038 |
alling | +0.035 |
astics | +0.034 |
irl | +0.034 |
sings | +0.033 |
Neg logit contributions | |
guard | -0.004 |
poisoning | -0.004 |
injuries | -0.004 |
unfortunately | -0.003 |
payoff | -0.003 |
Token | build |
Token ID | 1382 |
Feature activation | -2.47e-02 |
Pos logit contributions | |
aling | +0.012 |
alling | +0.011 |
astics | +0.010 |
sings | +0.010 |
irl | +0.010 |
Neg logit contributions | |
poisoning | -0.008 |
guard | -0.008 |
injuries | -0.008 |
last | -0.008 |
payoff | -0.008 |
Token | a |
Token ID | 257 |
Feature activation | -1.84e-02 |
Pos logit contributions | |
aling | +0.057 |
alling | +0.052 |
astics | +0.049 |
ophone | +0.048 |
sings | +0.047 |
Neg logit contributions | |
injuries | -0.020 |
poisoning | -0.019 |
last | -0.019 |
guard | -0.019 |
unfortunately | -0.018 |
Token | kay |
Token ID | 34681 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
aling | +0.035 |
alling | +0.032 |
astics | +0.030 |
sings | +0.030 |
irl | +0.029 |
Neg logit contributions | |
guard | -0.023 |
injuries | -0.022 |
poisoning | -0.022 |
last | -0.020 |
unfortunately | -0.020 |
Token | ak |
Token ID | 461 |
Feature activation | -2.53e-02 |
Pos logit contributions | |
sings | +0.230 |
dances | +0.209 |
aling | +0.202 |
alling | +0.195 |
presses | +0.193 |
Neg logit contributions | |
guard | -0.198 |
t | -0.180 |
poisoning | -0.167 |
last | -0.162 |
injuries | -0.154 |
Token | . |
Token ID | 13 |
Feature activation | -2.25e-02 |
Pos logit contributions | |
aling | +0.053 |
alling | +0.049 |
astics | +0.047 |
irl | +0.046 |
sings | +0.045 |
Neg logit contributions | |
guard | -0.025 |
poisoning | -0.024 |
injuries | -0.024 |
unfortunately | -0.022 |
payoff | -0.021 |
Token | They |
Token ID | 1119 |
Feature activation | -9.58e-03 |
Pos logit contributions | |
aling | +0.038 |
alling | +0.034 |
irl | +0.033 |
astics | +0.032 |
sings | +0.030 |
Neg logit contributions | |
guard | -0.031 |
poisoning | -0.031 |
injuries | -0.029 |
unfortunately | -0.029 |
payoff | -0.028 |
Token | found |
Token ID | 1043 |
Feature activation | -2.04e-02 |
Pos logit contributions | |
aling | +0.016 |
alling | +0.014 |
astics | +0.014 |
irl | +0.013 |
sings | +0.013 |
Neg logit contributions | |
guard | -0.014 |
poisoning | -0.014 |
injuries | -0.013 |
unfortunately | -0.013 |
last | -0.012 |
Token | some |
Token ID | 617 |
Feature activation | +1.81e-02 |
Pos logit contributions | |
aling | +0.029 |
alling | +0.025 |
irl | +0.023 |
astics | +0.022 |
sings | +0.022 |
Neg logit contributions | |
injuries | -0.034 |
guard | -0.033 |
poisoning | -0.033 |
payoff | -0.030 |
unfortunately | -0.030 |
Token | hose |
Token ID | 31489 |
Feature activation | -3.68e-02 |
Pos logit contributions | |
aling | +0.010 |
nic | +0.009 |
alling | +0.009 |
irl | +0.009 |
astics | +0.008 |
Neg logit contributions | |
poisoning | -0.011 |
injuries | -0.011 |
guard | -0.011 |
unfortunately | -0.011 |
t | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
aling | +0.060 |
alling | +0.054 |
astics | +0.050 |
ophone | +0.049 |
c | +0.049 |
Neg logit contributions | |
injuries | -0.052 |
guard | -0.049 |
poisoning | -0.048 |
payoff | -0.046 |
unfortunately | -0.045 |
Token | They |
Token ID | 1119 |
Feature activation | -5.91e-03 |
Pos logit contributions | |
aling | +0.026 |
alling | +0.024 |
astics | +0.022 |
irl | +0.022 |
sings | +0.021 |
Neg logit contributions | |
guard | -0.020 |
poisoning | -0.020 |
injuries | -0.019 |
unfortunately | -0.018 |
payoff | -0.018 |
Token | have |
Token ID | 423 |
Feature activation | -1.61e-02 |
Pos logit contributions | |
aling | +0.010 |
alling | +0.009 |
astics | +0.009 |
irl | +0.008 |
sings | +0.008 |
Neg logit contributions | |
poisoning | -0.008 |
guard | -0.008 |
injuries | -0.008 |
last | -0.007 |
unfortunately | -0.007 |
Token | a |
Token ID | 257 |
Feature activation | -8.69e-03 |
Pos logit contributions | |
aling | +0.024 |
astics | +0.022 |
irl | +0.021 |
alling | +0.021 |
sings | +0.019 |
Neg logit contributions | |
guard | -0.026 |
injuries | -0.024 |
poisoning | -0.024 |
unfortunately | -0.023 |
ho | -0.023 |
Token | kay |
Token ID | 34681 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
aling | +0.015 |
irl | +0.013 |
alling | +0.013 |
astics | +0.013 |
sings | +0.012 |
Neg logit contributions | |
guard | -0.012 |
poisoning | -0.011 |
injuries | -0.011 |
unfortunately | -0.011 |
payoff | -0.010 |
Token | ak |
Token ID | 461 |
Feature activation | -2.98e-02 |
Pos logit contributions | |
aling | +0.207 |
alling | +0.199 |
sings | +0.196 |
astics | +0.181 |
presses | +0.175 |
Neg logit contributions | |
guard | -0.205 |
poisoning | -0.202 |
t | -0.198 |
injuries | -0.181 |
last | -0.176 |
Token | toy |
Token ID | 13373 |
Feature activation | -2.58e-03 |
Pos logit contributions | |
aling | +0.060 |
alling | +0.055 |
astics | +0.052 |
irl | +0.051 |
sings | +0.050 |
Neg logit contributions | |
injuries | -0.033 |
poisoning | -0.031 |
guard | -0.031 |
unfortunately | -0.028 |
payoff | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | -1.75e-02 |
Pos logit contributions | |
aling | +0.004 |
alling | +0.004 |
astics | +0.003 |
ophone | +0.003 |
am | +0.003 |
Neg logit contributions | |
injuries | -0.004 |
poisoning | -0.003 |
guard | -0.003 |
payoff | -0.003 |
unfortunately | -0.003 |
Token | It |
Token ID | 632 |
Feature activation | +4.39e-03 |
Pos logit contributions | |
aling | +0.031 |
alling | +0.028 |
astics | +0.027 |
irl | +0.026 |
sings | +0.025 |
Neg logit contributions | |
guard | -0.023 |
poisoning | -0.023 |
injuries | -0.023 |
unfortunately | -0.021 |
payoff | -0.020 |
Token | is |
Token ID | 318 |
Feature activation | -4.07e-02 |
Pos logit contributions | |
guard | +0.006 |
last | +0.005 |
poisoning | +0.005 |
they | +0.005 |
that | +0.005 |
Neg logit contributions | |
aling | -0.008 |
alling | -0.008 |
astics | -0.007 |
vel | -0.007 |
essay | -0.007 |
Token | cooking |
Token ID | 10801 |
Feature activation | -1.34e-02 |
Pos logit contributions | |
aling | +0.027 |
alling | +0.024 |
irl | +0.023 |
astics | +0.023 |
sings | +0.021 |
Neg logit contributions | |
guard | -0.021 |
poisoning | -0.020 |
injuries | -0.020 |
payoff | -0.018 |
unfortunately | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | -1.43e-02 |
Pos logit contributions | |
aling | +0.022 |
irl | +0.020 |
alling | +0.019 |
astics | +0.019 |
sings | +0.018 |
Neg logit contributions | |
guard | -0.019 |
injuries | -0.018 |
poisoning | -0.017 |
department | -0.017 |
ho | -0.017 |
Token |
|
Token ID | 198 |
Feature activation | +2.83e-02 |
Pos logit contributions | |
aling | +0.029 |
alling | +0.026 |
irl | +0.025 |
astics | +0.025 |
sings | +0.023 |
Neg logit contributions | |
poisoning | -0.015 |
guard | -0.015 |
t | -0.015 |
unfortunately | -0.014 |
Not | -0.014 |
Token |
|
Token ID | 198 |
Feature activation | +2.24e-02 |
Pos logit contributions | |
guard | +0.023 |
poisoning | +0.022 |
injuries | +0.020 |
t | +0.019 |
unfortunately | +0.019 |
Neg logit contributions | |
aling | -0.065 |
alling | -0.060 |
astics | -0.057 |
irl | -0.056 |
sings | -0.055 |
Token | " |
Token ID | 1 |
Feature activation | +2.85e-02 |
Pos logit contributions | |
Not | +0.014 |
t | +0.014 |
guard | +0.014 |
unfortunately | +0.012 |
poisoning | +0.012 |
Neg logit contributions | |
aling | -0.054 |
alling | -0.051 |
sings | -0.048 |
astics | -0.047 |
irl | -0.046 |
Token | Mom |
Token ID | 29252 |
Feature activation | +5.53e-02 |
Pos logit contributions | |
injuries | +0.031 |
guard | +0.031 |
poisoning | +0.030 |
mercy | +0.027 |
unfortunately | +0.026 |
Neg logit contributions | |
aling | -0.056 |
alling | -0.051 |
astics | -0.048 |
irl | -0.047 |
sings | -0.046 |
Token | my |
Token ID | 1820 |
Feature activation | +4.37e-02 |
Pos logit contributions | |
guard | +0.065 |
poisoning | +0.063 |
injuries | +0.063 |
unfortunately | +0.056 |
payoff | +0.055 |
Neg logit contributions | |
aling | -0.107 |
alling | -0.097 |
astics | -0.091 |
irl | -0.091 |
sings | -0.088 |
Token | , |
Token ID | 11 |
Feature activation | +5.25e-02 |
Pos logit contributions | |
guard | +0.031 |
injuries | +0.030 |
poisoning | +0.030 |
unfortunately | +0.025 |
payoff | +0.024 |
Neg logit contributions | |
aling | -0.104 |
alling | -0.096 |
astics | -0.092 |
irl | -0.091 |
sings | -0.089 |
Token | what |
Token ID | 644 |
Feature activation | +5.55e-02 |
Pos logit contributions | |
injuries | +0.068 |
guard | +0.068 |
poisoning | +0.064 |
payoff | +0.059 |
unfortunately | +0.058 |
Neg logit contributions | |
aling | -0.092 |
alling | -0.082 |
irl | -0.078 |
astics | -0.078 |
sings | -0.077 |
Token | are |
Token ID | 389 |
Feature activation | +1.39e-02 |
Pos logit contributions | |
poisoning | +0.078 |
guard | +0.077 |
injuries | +0.068 |
last | +0.068 |
t | +0.068 |
Neg logit contributions | |
aling | -0.099 |
alling | -0.091 |
astics | -0.079 |
irl | -0.078 |
aled | -0.077 |
Token | you |
Token ID | 345 |
Feature activation | +2.30e-02 |
Pos logit contributions | |
injuries | +0.017 |
guard | +0.017 |
poisoning | +0.016 |
payoff | +0.015 |
department | +0.015 |
Neg logit contributions | |
aling | -0.024 |
irl | -0.022 |
astics | -0.022 |
alling | -0.022 |
sings | -0.021 |
Token | ily |
Token ID | 813 |
Feature activation | -6.37e-03 |
Pos logit contributions | |
aling | +0.056 |
alling | +0.051 |
astics | +0.049 |
irl | +0.048 |
sings | +0.046 |
Neg logit contributions | |
guard | -0.030 |
poisoning | -0.030 |
injuries | -0.029 |
unfortunately | -0.026 |
payoff | -0.026 |
Token | says |
Token ID | 1139 |
Feature activation | +4.03e-02 |
Pos logit contributions | |
aling | +0.015 |
alling | +0.015 |
astics | +0.014 |
irl | +0.014 |
beh | +0.014 |
Neg logit contributions | |
guard | -0.004 |
last | -0.003 |
poisoning | -0.003 |
payoff | -0.003 |
that | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | +4.60e-02 |
Pos logit contributions | |
poisoning | +0.036 |
injuries | +0.034 |
guard | +0.033 |
extingu | +0.028 |
unfortunately | +0.028 |
Neg logit contributions | |
aling | -0.085 |
alling | -0.078 |
sings | -0.078 |
irl | -0.077 |
astics | -0.077 |
Token | " |
Token ID | 366 |
Feature activation | -8.10e-03 |
Pos logit contributions | |
guard | +0.064 |
poisoning | +0.064 |
injuries | +0.063 |
unfortunately | +0.057 |
payoff | +0.057 |
Neg logit contributions | |
aling | -0.078 |
alling | -0.069 |
astics | -0.065 |
irl | -0.064 |
sings | -0.063 |
Token | Ben |
Token ID | 11696 |
Feature activation | +9.16e-03 |
Pos logit contributions | |
aling | +0.016 |
alling | +0.014 |
astics | +0.013 |
irl | +0.013 |
sings | +0.013 |
Neg logit contributions | |
injuries | -0.010 |
poisoning | -0.010 |
guard | -0.009 |
unfortunately | -0.009 |
department | -0.008 |
Token | , |
Token ID | 11 |
Feature activation | +5.07e-02 |
Pos logit contributions | |
injuries | +0.008 |
guard | +0.008 |
poisoning | +0.007 |
unfortunately | +0.007 |
payoff | +0.006 |
Neg logit contributions | |
aling | -0.020 |
alling | -0.018 |
astics | -0.018 |
sings | -0.018 |
irl | -0.018 |
Token | do |
Token ID | 466 |
Feature activation | +1.90e-02 |
Pos logit contributions | |
injuries | +0.070 |
guard | +0.067 |
poisoning | +0.066 |
extingu | +0.060 |
department | +0.060 |
Neg logit contributions | |
aling | -0.082 |
alling | -0.072 |
sings | -0.071 |
irl | -0.070 |
astics | -0.069 |
Token | you |
Token ID | 345 |
Feature activation | +1.30e-02 |
Pos logit contributions | |
that | +0.009 |
they | +0.008 |
t | +0.007 |
guard | +0.007 |
injuries | +0.006 |
Neg logit contributions | |
aling | -0.052 |
alling | -0.050 |
aled | -0.046 |
acked | -0.045 |
wn | -0.045 |
Token | want |
Token ID | 765 |
Feature activation | +2.83e-02 |
Pos logit contributions | |
guard | +0.016 |
poisoning | +0.016 |
last | +0.015 |
t | +0.015 |
injuries | +0.015 |
Neg logit contributions | |
aling | -0.023 |
alling | -0.021 |
astics | -0.020 |
irl | -0.019 |
sings | -0.018 |
Token | to |
Token ID | 284 |
Feature activation | +1.40e-02 |
Pos logit contributions | |
unfortunately | +0.030 |
ho | +0.030 |
injuries | +0.029 |
poisoning | +0.029 |
comb | +0.029 |
Neg logit contributions | |
aling | -0.051 |
astics | -0.048 |
plays | -0.047 |
sings | -0.046 |
alling | -0.046 |
Token | marry |
Token ID | 12479 |
Feature activation | -1.82e-02 |
Pos logit contributions | |
guard | +0.025 |
poisoning | +0.025 |
injuries | +0.025 |
unfortunately | +0.024 |
payoff | +0.023 |
Neg logit contributions | |
aling | -0.018 |
alling | -0.016 |
astics | -0.014 |
irl | -0.014 |
sings | -0.013 |
Token | faster |
Token ID | 5443 |
Feature activation | -2.59e-03 |
Pos logit contributions | |
aling | +0.024 |
alling | +0.021 |
astics | +0.020 |
irl | +0.019 |
sings | +0.019 |
Neg logit contributions | |
poisoning | -0.020 |
guard | -0.020 |
injuries | -0.019 |
unfortunately | -0.018 |
payoff | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | -1.81e-02 |
Pos logit contributions | |
aling | +0.004 |
alling | +0.004 |
irl | +0.004 |
astics | +0.004 |
sings | +0.004 |
Neg logit contributions | |
poisoning | -0.004 |
guard | -0.003 |
injuries | -0.003 |
unfortunately | -0.003 |
department | -0.003 |
Token |
|
Token ID | 198 |
Feature activation | +2.53e-02 |
Pos logit contributions | |
aling | +0.034 |
alling | +0.031 |
irl | +0.030 |
astics | +0.030 |
sings | +0.028 |
Neg logit contributions | |
guard | -0.021 |
poisoning | -0.021 |
injuries | -0.020 |
unfortunately | -0.019 |
payoff | -0.019 |
Token |
|
Token ID | 198 |
Feature activation | +1.90e-02 |
Pos logit contributions | |
poisoning | +0.031 |
guard | +0.031 |
injuries | +0.030 |
unfortunately | +0.028 |
payoff | +0.027 |
Neg logit contributions | |
aling | -0.048 |
alling | -0.043 |
astics | -0.041 |
irl | -0.040 |
sings | -0.038 |
Token | The |
Token ID | 464 |
Feature activation | +1.08e-02 |
Pos logit contributions | |
guard | +0.020 |
poisoning | +0.020 |
injuries | +0.020 |
unfortunately | +0.018 |
payoff | +0.017 |
Neg logit contributions | |
aling | -0.039 |
alling | -0.035 |
astics | -0.033 |
irl | -0.033 |
sings | -0.032 |
Token | end |
Token ID | 886 |
Feature activation | +6.99e-02 |
Pos logit contributions | |
guard | +0.012 |
injuries | +0.011 |
payoff | +0.010 |
last | +0.010 |
poisoning | +0.010 |
Neg logit contributions | |
aling | -0.023 |
alling | -0.021 |
astics | -0.020 |
irl | -0.019 |
atters | -0.019 |
Token | . |
Token ID | 13 |
Feature activation | -2.11e-02 |
Pos logit contributions | |
poisoning | +0.076 |
guard | +0.076 |
injuries | +0.075 |
unfortunately | +0.068 |
payoff | +0.066 |
Neg logit contributions | |
aling | -0.140 |
alling | -0.127 |
astics | -0.121 |
irl | -0.120 |
sings | -0.116 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -3.07e-03 |
Pos logit contributions | |
aling | +0.041 |
alling | +0.037 |
astics | +0.035 |
irl | +0.034 |
sings | +0.033 |
Neg logit contributions | |
poisoning | -0.025 |
injuries | -0.025 |
guard | -0.025 |
unfortunately | -0.021 |
payoff | -0.021 |
Token | Once |
Token ID | 7454 |
Feature activation | +3.39e-02 |
Pos logit contributions | |
aling | +0.007 |
alling | +0.006 |
astics | +0.006 |
irl | +0.006 |
sings | +0.006 |
Neg logit contributions | |
guard | -0.003 |
poisoning | -0.002 |
injuries | -0.002 |
payoff | -0.002 |
t | -0.002 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.58e-02 |
Pos logit contributions | |
injuries | +0.051 |
poisoning | +0.051 |
extingu | +0.048 |
payoff | +0.047 |
guard | +0.047 |
Neg logit contributions | |
aling | -0.052 |
irl | -0.047 |
alling | -0.045 |
ophone | -0.045 |
sings | -0.043 |
Token | a |
Token ID | 257 |
Feature activation | -1.22e-03 |
Pos logit contributions | |
aling | +0.030 |
alling | +0.027 |
astics | +0.025 |
irl | +0.025 |
sings | +0.024 |
Neg logit contributions | |
guard | -0.019 |
poisoning | -0.019 |
injuries | -0.019 |
unfortunately | -0.017 |
payoff | -0.017 |
Token | he |
Token ID | 339 |
Feature activation | -3.47e-03 |
Pos logit contributions | |
guard | +0.040 |
injuries | +0.039 |
payoff | +0.038 |
department | +0.037 |
extingu | +0.036 |
Neg logit contributions | |
aling | -0.040 |
irl | -0.039 |
zoom | -0.035 |
plays | -0.035 |
astics | -0.034 |
Token | was |
Token ID | 373 |
Feature activation | -3.69e-02 |
Pos logit contributions | |
aling | +0.006 |
alling | +0.006 |
astics | +0.005 |
irl | +0.005 |
sings | +0.005 |
Neg logit contributions | |
guard | -0.005 |
poisoning | -0.005 |
injuries | -0.004 |
unfortunately | -0.004 |
payoff | -0.004 |
Token | playing |
Token ID | 2712 |
Feature activation | -2.74e-02 |
Pos logit contributions | |
aling | +0.066 |
alling | +0.060 |
essay | +0.058 |
ophone | +0.057 |
etition | +0.056 |
Neg logit contributions | |
poisoning | -0.048 |
to | -0.047 |
that | -0.044 |
injuries | -0.043 |
while | -0.043 |
Token | with |
Token ID | 351 |
Feature activation | -2.45e-02 |
Pos logit contributions | |
aling | +0.048 |
alling | +0.043 |
astics | +0.040 |
irl | +0.040 |
sings | +0.038 |
Neg logit contributions | |
guard | -0.037 |
poisoning | -0.037 |
injuries | -0.036 |
unfortunately | -0.034 |
payoff | -0.033 |
Token | his |
Token ID | 465 |
Feature activation | +1.63e-02 |
Pos logit contributions | |
aling | +0.033 |
irl | +0.029 |
alling | +0.028 |
astics | +0.027 |
sings | +0.025 |
Neg logit contributions | |
guard | -0.043 |
injuries | -0.041 |
poisoning | -0.040 |
ho | -0.040 |
payoff | -0.040 |
Token | friend |
Token ID | 1545 |
Feature activation | +5.10e-02 |
Pos logit contributions | |
payoff | +0.033 |
guard | +0.033 |
injuries | +0.033 |
poisoning | +0.033 |
unfortunately | +0.032 |
Neg logit contributions | |
irl | -0.014 |
nic | -0.012 |
zoom | -0.011 |
irled | -0.011 |
astics | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | +4.24e-02 |
Pos logit contributions | |
guard | +0.066 |
injuries | +0.064 |
payoff | +0.061 |
poisoning | +0.057 |
ho | +0.056 |
Neg logit contributions | |
aling | -0.089 |
alling | -0.080 |
c | -0.080 |
astics | -0.078 |
ophone | -0.076 |
Token | a |
Token ID | 257 |
Feature activation | -1.66e-02 |
Pos logit contributions | |
injuries | +0.064 |
guard | +0.062 |
poisoning | +0.059 |
payoff | +0.059 |
ho | +0.057 |
Neg logit contributions | |
aling | -0.064 |
alling | -0.058 |
astics | -0.054 |
ophone | -0.053 |
irl | -0.052 |
Token | little |
Token ID | 1310 |
Feature activation | -5.85e-03 |
Pos logit contributions | |
aling | +0.025 |
irl | +0.024 |
astics | +0.022 |
alling | +0.022 |
sings | +0.022 |
Neg logit contributions | |
injuries | -0.022 |
poisoning | -0.022 |
guard | -0.022 |
payoff | -0.020 |
ately | -0.020 |
Token | bird |
Token ID | 6512 |
Feature activation | -3.20e-02 |
Pos logit contributions | |
aling | +0.012 |
alling | +0.011 |
astics | +0.010 |
irl | +0.010 |
sings | +0.010 |
Neg logit contributions | |
injuries | -0.006 |
poisoning | -0.006 |
guard | -0.006 |
unfortunately | -0.005 |
payoff | -0.005 |
Token | named |
Token ID | 3706 |
Feature activation | -3.40e-02 |
Pos logit contributions | |
aling | +0.045 |
alling | +0.038 |
irl | +0.036 |
astics | +0.035 |
c | +0.034 |
Neg logit contributions | |
guard | -0.055 |
poisoning | -0.054 |
injuries | -0.054 |
payoff | -0.052 |
unfortunately | -0.050 |
Token | job |
Token ID | 1693 |
Feature activation | +5.80e-02 |
Pos logit contributions | |
guard | +0.012 |
poisoning | +0.012 |
injuries | +0.011 |
unfortunately | +0.011 |
t | +0.011 |
Neg logit contributions | |
aling | -0.010 |
alling | -0.009 |
astics | -0.009 |
irl | -0.009 |
sings | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | -3.83e-02 |
Pos logit contributions | |
poisoning | +0.074 |
guard | +0.074 |
injuries | +0.073 |
unfortunately | +0.068 |
ho | +0.068 |
Neg logit contributions | |
aling | -0.103 |
alling | -0.092 |
astics | -0.088 |
irl | -0.086 |
sings | -0.082 |
Token | I |
Token ID | 314 |
Feature activation | +2.37e-02 |
Pos logit contributions | |
aling | +0.066 |
alling | +0.059 |
irl | +0.055 |
astics | +0.055 |
sings | +0.051 |
Neg logit contributions | |
guard | -0.053 |
poisoning | -0.052 |
injuries | -0.051 |
unfortunately | -0.048 |
payoff | -0.047 |
Token | don |
Token ID | 836 |
Feature activation | +2.80e-02 |
Pos logit contributions | |
injuries | +0.030 |
poisoning | +0.030 |
guard | +0.030 |
unfortunately | +0.027 |
payoff | +0.027 |
Neg logit contributions | |
aling | -0.043 |
alling | -0.038 |
irl | -0.036 |
astics | -0.036 |
sings | -0.035 |
Token | 't |
Token ID | 470 |
Feature activation | +3.41e-02 |
Pos logit contributions | |
guard | +0.011 |
t | +0.009 |
poisoning | +0.008 |
injuries | +0.007 |
unfortunately | +0.007 |
Neg logit contributions | |
aling | -0.076 |
alling | -0.071 |
astics | -0.069 |
sings | -0.067 |
irl | -0.067 |
Token | know |
Token ID | 760 |
Feature activation | +7.04e-02 |
Pos logit contributions | |
poisoning | +0.044 |
injuries | +0.043 |
guard | +0.043 |
last | +0.040 |
payoff | +0.039 |
Neg logit contributions | |
aling | -0.060 |
alling | -0.056 |
astics | -0.053 |
irl | -0.053 |
irled | -0.051 |
Token | what |
Token ID | 644 |
Feature activation | +6.70e-02 |
Pos logit contributions | |
guard | +0.079 |
poisoning | +0.078 |
injuries | +0.076 |
payoff | +0.071 |
unfortunately | +0.068 |
Neg logit contributions | |
aling | -0.137 |
alling | -0.124 |
astics | -0.118 |
irl | -0.117 |
sings | -0.114 |
Token | to |
Token ID | 284 |
Feature activation | +3.78e-02 |
Pos logit contributions | |
injuries | +0.056 |
poisoning | +0.055 |
guard | +0.053 |
unfortunately | +0.049 |
ho | +0.046 |
Neg logit contributions | |
aling | -0.148 |
astics | -0.134 |
alling | -0.134 |
sings | -0.132 |
irl | -0.131 |
Token | do |
Token ID | 466 |
Feature activation | +2.11e-02 |
Pos logit contributions | |
guard | +0.057 |
poisoning | +0.057 |
injuries | +0.057 |
unfortunately | +0.055 |
payoff | +0.055 |
Neg logit contributions | |
aling | -0.054 |
irl | -0.049 |
astics | -0.047 |
alling | -0.047 |
sings | -0.044 |
Token | ." |
Token ID | 526 |
Feature activation | -1.90e-02 |
Pos logit contributions | |
injuries | +0.032 |
poisoning | +0.031 |
guard | +0.031 |
ho | +0.031 |
extingu | +0.029 |
Neg logit contributions | |
aling | -0.032 |
alling | -0.028 |
astics | -0.028 |
irl | -0.027 |
sings | -0.025 |
Token |
|
Token ID | 198 |
Feature activation | +1.93e-02 |
Pos logit contributions | |
aling | +0.024 |
sings | +0.021 |
irl | +0.021 |
alling | +0.020 |
astics | +0.020 |
Neg logit contributions | |
poisoning | -0.032 |
injuries | -0.031 |
department | -0.030 |
guard | -0.030 |
extingu | -0.029 |
Token | But |
Token ID | 1537 |
Feature activation | +5.33e-02 |
Pos logit contributions | |
injuries | +0.042 |
hesitation | +0.038 |
poisoning | +0.038 |
extingu | +0.037 |
guard | +0.037 |
Neg logit contributions | |
aling | -0.051 |
astics | -0.045 |
irl | -0.045 |
alling | -0.044 |
learn | -0.042 |
Token | Tim |
Token ID | 5045 |
Feature activation | +6.77e-03 |
Pos logit contributions | |
guard | +0.064 |
injuries | +0.063 |
poisoning | +0.062 |
payoff | +0.057 |
t | +0.055 |
Neg logit contributions | |
aling | -0.097 |
alling | -0.088 |
irl | -0.086 |
astics | -0.084 |
sings | -0.081 |
Token | my |
Token ID | 1820 |
Feature activation | +3.81e-02 |
Pos logit contributions | |
guard | +0.013 |
injuries | +0.012 |
poisoning | +0.012 |
unfortunately | +0.011 |
payoff | +0.011 |
Neg logit contributions | |
aling | -0.009 |
alling | -0.007 |
astics | -0.007 |
irl | -0.007 |
c | -0.006 |
Token | , |
Token ID | 11 |
Feature activation | +5.10e-02 |
Pos logit contributions | |
poisoning | +0.016 |
guard | +0.014 |
injuries | +0.012 |
unfortunately | +0.009 |
last | +0.009 |
Neg logit contributions | |
aling | -0.105 |
alling | -0.100 |
astics | -0.092 |
aled | -0.091 |
irl | -0.090 |
Token | it |
Token ID | 340 |
Feature activation | +2.54e-02 |
Pos logit contributions | |
guard | +0.056 |
injuries | +0.055 |
poisoning | +0.055 |
payoff | +0.049 |
unfortunately | +0.048 |
Neg logit contributions | |
aling | -0.100 |
alling | -0.091 |
irl | -0.087 |
astics | -0.087 |
sings | -0.084 |
Token | 's |
Token ID | 338 |
Feature activation | -1.78e-02 |
Pos logit contributions | |
guard | +0.039 |
poisoning | +0.038 |
last | +0.037 |
t | +0.036 |
unfortunately | +0.036 |
Neg logit contributions | |
aling | -0.041 |
alling | -0.038 |
astics | -0.033 |
aled | -0.031 |
essay | -0.031 |
Token | cold |
Token ID | 4692 |
Feature activation | +1.95e-02 |
Pos logit contributions | |
aling | +0.031 |
alling | +0.028 |
irl | +0.026 |
astics | +0.026 |
sings | +0.025 |
Neg logit contributions | |
guard | -0.025 |
poisoning | -0.024 |
injuries | -0.024 |
payoff | -0.022 |
unfortunately | -0.022 |
Token | outside |
Token ID | 2354 |
Feature activation | -5.03e-03 |
Pos logit contributions | |
injuries | +0.032 |
guard | +0.031 |
poisoning | +0.031 |
payoff | +0.030 |
unfortunately | +0.030 |
Neg logit contributions | |
aling | -0.028 |
alling | -0.024 |
irl | -0.023 |
astics | -0.022 |
sings | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | -1.96e-02 |
Pos logit contributions | |
aling | +0.010 |
alling | +0.009 |
astics | +0.008 |
irl | +0.008 |
c | +0.008 |
Neg logit contributions | |
poisoning | -0.006 |
injuries | -0.006 |
guard | -0.005 |
payoff | -0.005 |
ho | -0.005 |
Token | You |
Token ID | 921 |
Feature activation | +3.27e-04 |
Pos logit contributions | |
aling | +0.036 |
alling | +0.033 |
irl | +0.033 |
astics | +0.031 |
irled | +0.030 |
Neg logit contributions | |
unfortunately | -0.023 |
poisoning | -0.023 |
guard | -0.022 |
injuries | -0.022 |
Papa | -0.022 |
Token | need |
Token ID | 761 |
Feature activation | +3.92e-02 |
Pos logit contributions | |
guard | +0.000 |
poisoning | +0.000 |
injuries | +0.000 |
unfortunately | +0.000 |
payoff | +0.000 |
Neg logit contributions | |
aling | -0.001 |
alling | -0.001 |
astics | -0.001 |
irl | -0.001 |
sings | -0.000 |
Token | to |
Token ID | 284 |
Feature activation | +1.35e-02 |
Pos logit contributions | |
aling | +0.000 |
alling | +0.000 |
astics | +0.000 |
irl | +0.000 |
sings | +0.000 |
Neg logit contributions | |
guard | -0.000 |
poisoning | -0.000 |
injuries | -0.000 |
unfortunately | -0.000 |
payoff | -0.000 |
Token | his |
Token ID | 465 |
Feature activation | +1.53e-02 |
Pos logit contributions | |
injuries | +0.013 |
poisoning | +0.012 |
last | +0.012 |
unfortunately | +0.011 |
guard | +0.011 |
Neg logit contributions | |
aling | -0.029 |
alling | -0.027 |
astics | -0.025 |
irl | -0.024 |
ab | -0.024 |
Token | mom |
Token ID | 1995 |
Feature activation | +5.47e-02 |
Pos logit contributions | |
guard | +0.011 |
poisoning | +0.011 |
injuries | +0.011 |
unfortunately | +0.009 |
payoff | +0.009 |
Neg logit contributions | |
aling | -0.037 |
alling | -0.034 |
astics | -0.032 |
irl | -0.032 |
sings | -0.031 |
Token | and |
Token ID | 290 |
Feature activation | -1.26e-03 |
Pos logit contributions | |
guard | +0.063 |
poisoning | +0.063 |
injuries | +0.062 |
unfortunately | +0.056 |
payoff | +0.055 |
Neg logit contributions | |
aling | -0.106 |
alling | -0.096 |
astics | -0.091 |
irl | -0.091 |
sings | -0.087 |
Token | ate |
Token ID | 15063 |
Feature activation | -2.27e-02 |
Pos logit contributions | |
aling | +0.002 |
alling | +0.002 |
astics | +0.002 |
irl | +0.002 |
sings | +0.002 |
Neg logit contributions | |
injuries | -0.002 |
guard | -0.002 |
poisoning | -0.002 |
t | -0.001 |
unfortunately | -0.001 |
Token | his |
Token ID | 465 |
Feature activation | +2.12e-02 |
Pos logit contributions | |
aling | +0.035 |
alling | +0.031 |
irl | +0.030 |
astics | +0.029 |
sings | +0.027 |
Neg logit contributions | |
guard | -0.034 |
poisoning | -0.032 |
injuries | -0.032 |
t | -0.031 |
payoff | -0.031 |
Token | vegetables |
Token ID | 13701 |
Feature activation | +1.52e-02 |
Pos logit contributions | |
poisoning | +0.028 |
guard | +0.028 |
injuries | +0.028 |
unfortunately | +0.026 |
payoff | +0.025 |
Neg logit contributions | |
aling | -0.037 |
alling | -0.033 |
astics | -0.031 |
irl | -0.031 |
sings | -0.029 |
Token | . |
Token ID | 13 |
Feature activation | -1.88e-02 |
Pos logit contributions | |
injuries | +0.021 |
guard | +0.021 |
payoff | +0.021 |
department | +0.020 |
poisoning | +0.020 |
Neg logit contributions | |
aling | -0.024 |
alling | -0.021 |
irl | -0.019 |
astics | -0.019 |
vel | -0.018 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +6.20e-03 |
Pos logit contributions | |
aling | +0.042 |
alling | +0.039 |
irl | +0.037 |
astics | +0.037 |
sings | +0.035 |
Neg logit contributions | |
poisoning | -0.016 |
guard | -0.015 |
injuries | -0.015 |
unfortunately | -0.014 |
t | -0.014 |
Token | Once |
Token ID | 7454 |
Feature activation | +4.68e-02 |
Pos logit contributions | |
injuries | +0.008 |
poisoning | +0.008 |
guard | +0.008 |
unfortunately | +0.007 |
payoff | +0.007 |
Neg logit contributions | |
aling | -0.011 |
alling | -0.010 |
irl | -0.010 |
astics | -0.010 |
sings | -0.009 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
poisoning | +0.074 |
injuries | +0.074 |
extingu | +0.072 |
payoff | +0.072 |
guard | +0.071 |
Neg logit contributions | |
aling | -0.062 |
irl | -0.059 |
ophone | -0.054 |
c | -0.052 |
alling | -0.052 |
Token | to |
Token ID | 284 |
Feature activation | +3.80e-02 |
Pos logit contributions | |
guard | +0.035 |
poisoning | +0.034 |
injuries | +0.032 |
unfortunately | +0.029 |
last | +0.029 |
Neg logit contributions | |
aling | -0.068 |
alling | -0.062 |
irl | -0.058 |
astics | -0.058 |
sings | -0.055 |
Token | work |
Token ID | 670 |
Feature activation | +2.15e-02 |
Pos logit contributions | |
guard | +0.049 |
poisoning | +0.049 |
injuries | +0.048 |
unfortunately | +0.044 |
t | +0.043 |
Neg logit contributions | |
aling | -0.068 |
alling | -0.061 |
irl | -0.059 |
astics | -0.058 |
sings | -0.054 |
Token | together |
Token ID | 1978 |
Feature activation | +6.56e-02 |
Pos logit contributions | |
guard | +0.027 |
poisoning | +0.027 |
injuries | +0.026 |
payoff | +0.024 |
unfortunately | +0.024 |
Neg logit contributions | |
aling | -0.040 |
alling | -0.036 |
astics | -0.034 |
irl | -0.033 |
sings | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | -2.43e-02 |
Pos logit contributions | |
guard | +0.083 |
injuries | +0.083 |
poisoning | +0.083 |
unfortunately | +0.075 |
payoff | +0.072 |
Neg logit contributions | |
aling | -0.117 |
alling | -0.105 |
astics | -0.101 |
irl | -0.098 |
sings | -0.094 |
Token | They |
Token ID | 1119 |
Feature activation | -2.60e-03 |
Pos logit contributions | |
aling | +0.036 |
alling | +0.031 |
irl | +0.029 |
astics | +0.029 |
sings | +0.027 |
Neg logit contributions | |
guard | -0.039 |
poisoning | -0.039 |
injuries | -0.038 |
unfortunately | -0.037 |
payoff | -0.036 |
Token | learned |
Token ID | 4499 |
Feature activation | +2.50e-02 |
Pos logit contributions | |
aling | +0.005 |
alling | +0.004 |
astics | +0.004 |
irl | +0.004 |
sings | +0.004 |
Neg logit contributions | |
guard | -0.004 |
poisoning | -0.004 |
injuries | -0.003 |
unfortunately | -0.003 |
payoff | -0.003 |
Token | that |
Token ID | 326 |
Feature activation | +4.85e-02 |
Pos logit contributions | |
injuries | +0.024 |
poisoning | +0.022 |
guard | +0.021 |
ho | +0.020 |
payoff | +0.018 |
Neg logit contributions | |
aling | -0.052 |
alling | -0.048 |
astics | -0.048 |
irl | -0.046 |
sings | -0.045 |
Token | even |
Token ID | 772 |
Feature activation | +1.18e-02 |
Pos logit contributions | |
guard | +0.057 |
poisoning | +0.057 |
injuries | +0.055 |
unfortunately | +0.052 |
payoff | +0.051 |
Neg logit contributions | |
aling | -0.094 |
alling | -0.086 |
astics | -0.080 |
irl | -0.079 |
ophone | -0.076 |
Token | though |
Token ID | 996 |
Feature activation | +6.97e-02 |
Pos logit contributions | |
injuries | +0.011 |
guard | +0.010 |
poisoning | +0.010 |
ho | +0.009 |
unfortunately | +0.008 |
Neg logit contributions | |
aling | -0.026 |
alling | -0.024 |
astics | -0.023 |
irl | -0.022 |
sings | -0.022 |
Token | they |
Token ID | 484 |
Feature activation | +1.14e-02 |
Pos logit contributions | |
poisoning | +0.068 |
guard | +0.066 |
injuries | +0.066 |
payoff | +0.057 |
extingu | +0.057 |
Neg logit contributions | |
aling | -0.140 |
irl | -0.127 |
alling | -0.127 |
astics | -0.125 |
sings | -0.123 |
Token | were |
Token ID | 547 |
Feature activation | -2.85e-02 |
Pos logit contributions | |
injuries | +0.015 |
guard | +0.015 |
poisoning | +0.015 |
unfortunately | +0.014 |
extingu | +0.013 |
Neg logit contributions | |
aling | -0.020 |
alling | -0.018 |
irl | -0.017 |
astics | -0.016 |
sings | -0.016 |
Token | on |
Token ID | 319 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
aling | +0.081 |
sings | +0.074 |
alling | +0.074 |
astics | +0.073 |
irl | +0.072 |
Neg logit contributions | |
guard | -0.051 |
poisoning | -0.051 |
injuries | -0.050 |
last | -0.046 |
unfortunately | -0.045 |
Token | a |
Token ID | 257 |
Feature activation | -1.55e-02 |
Pos logit contributions | |
aling | +0.018 |
alling | +0.016 |
irl | +0.015 |
astics | +0.015 |
aled | +0.014 |
Neg logit contributions | |
guard | -0.018 |
poisoning | -0.017 |
injuries | -0.016 |
ho | -0.016 |
hesitation | -0.015 |
Token | ch |
Token ID | 442 |
Feature activation | -8.16e-02 |
Pos logit contributions | |
aling | +0.027 |
alling | +0.024 |
astics | +0.023 |
irl | +0.023 |
sings | +0.022 |
Neg logit contributions | |
guard | -0.021 |
poisoning | -0.021 |
injuries | -0.020 |
unfortunately | -0.019 |
payoff | -0.019 |
Token | ute |
Token ID | 1133 |
Feature activation | -4.03e-02 |
Pos logit contributions | |
aling | +0.135 |
sings | +0.128 |
astics | +0.127 |
alling | +0.125 |
beh | +0.119 |
Neg logit contributions | |
guard | -0.104 |
injuries | -0.101 |
last | -0.096 |
poisoning | -0.095 |
t | -0.094 |
Token | . |
Token ID | 13 |
Feature activation | -2.45e-02 |
Pos logit contributions | |
aling | +0.086 |
alling | +0.079 |
astics | +0.076 |
irl | +0.075 |
sings | +0.073 |
Neg logit contributions | |
guard | -0.038 |
poisoning | -0.038 |
injuries | -0.037 |
unfortunately | -0.033 |
last | -0.032 |
Token | He |
Token ID | 679 |
Feature activation | -9.83e-03 |
Pos logit contributions | |
aling | +0.057 |
alling | +0.053 |
irl | +0.051 |
astics | +0.051 |
sings | +0.049 |
Neg logit contributions | |
poisoning | -0.017 |
t | -0.017 |
guard | -0.017 |
unfortunately | -0.016 |
Not | -0.016 |
Token | had |
Token ID | 550 |
Feature activation | -2.10e-02 |
Pos logit contributions | |
aling | +0.020 |
alling | +0.019 |
astics | +0.018 |
vel | +0.017 |
beh | +0.017 |
Neg logit contributions | |
guard | -0.010 |
last | -0.009 |
they | -0.009 |
that | -0.009 |
now | -0.009 |
Token | to |
Token ID | 284 |
Feature activation | -9.23e-03 |
Pos logit contributions | |
aling | +0.044 |
alling | +0.040 |
irl | +0.039 |
astics | +0.038 |
sings | +0.037 |
Neg logit contributions | |
poisoning | -0.020 |
injuries | -0.019 |
guard | -0.019 |
payoff | -0.018 |
last | -0.018 |
Token | go |
Token ID | 467 |
Feature activation | -2.94e-02 |
Pos logit contributions | |
aling | +0.016 |
alling | +0.014 |
astics | +0.013 |
ophone | +0.013 |
sings | +0.013 |
Neg logit contributions | |
injuries | -0.012 |
guard | -0.012 |
poisoning | -0.012 |
last | -0.012 |
payoff | -0.011 |
Token | down |
Token ID | 866 |
Feature activation | -1.44e-02 |
Pos logit contributions | |
aling | +0.059 |
alling | +0.054 |
astics | +0.053 |
irl | +0.052 |
sings | +0.051 |
Neg logit contributions | |
guard | -0.032 |
poisoning | -0.030 |
injuries | -0.029 |
last | -0.027 |
unfortunately | -0.026 |
Token | the |
Token ID | 262 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
aling | +0.027 |
alling | +0.024 |
astics | +0.022 |
irl | +0.022 |
ped | +0.021 |
Neg logit contributions | |
injuries | -0.018 |
poisoning | -0.017 |
guard | -0.017 |
ho | -0.016 |
unfortunately | -0.016 |
Token | special |
Token ID | 2041 |
Feature activation | +2.17e-02 |
Pos logit contributions | |
injuries | +0.036 |
guard | +0.036 |
poisoning | +0.036 |
payoff | +0.034 |
unfortunately | +0.033 |
Neg logit contributions | |
aling | -0.038 |
alling | -0.033 |
irl | -0.031 |
astics | -0.031 |
sings | -0.029 |
Token | - |
Token ID | 532 |
Feature activation | +1.13e-02 |
Pos logit contributions | |
injuries | +0.032 |
guard | +0.030 |
poisoning | +0.029 |
payoff | +0.029 |
unfortunately | +0.028 |
Neg logit contributions | |
aling | -0.034 |
alling | -0.030 |
astics | -0.029 |
irl | -0.028 |
sings | -0.028 |
Token | a |
Token ID | 257 |
Feature activation | -2.02e-02 |
Pos logit contributions | |
poisoning | +0.011 |
guard | +0.011 |
injuries | +0.010 |
unfortunately | +0.010 |
last | +0.010 |
Neg logit contributions | |
aling | -0.024 |
alling | -0.022 |
astics | -0.021 |
irl | -0.021 |
aled | -0.020 |
Token | spring |
Token ID | 6076 |
Feature activation | -2.88e-02 |
Pos logit contributions | |
aling | +0.036 |
alling | +0.032 |
astics | +0.030 |
irl | +0.030 |
sings | +0.028 |
Neg logit contributions | |
guard | -0.027 |
poisoning | -0.027 |
injuries | -0.026 |
unfortunately | -0.024 |
payoff | -0.024 |
Token | ! |
Token ID | 0 |
Feature activation | -2.15e-02 |
Pos logit contributions | |
aling | +0.051 |
alling | +0.046 |
irl | +0.044 |
astics | +0.043 |
sings | +0.042 |
Neg logit contributions | |
injuries | -0.037 |
poisoning | -0.036 |
guard | -0.036 |
unfortunately | -0.033 |
payoff | -0.032 |
Token | Inside |
Token ID | 14384 |
Feature activation | +8.51e-03 |
Pos logit contributions | |
aling | +0.035 |
alling | +0.031 |
irl | +0.030 |
astics | +0.029 |
sings | +0.027 |
Neg logit contributions | |
guard | -0.032 |
poisoning | -0.031 |
injuries | -0.031 |
unfortunately | -0.029 |
payoff | -0.029 |
Token | were |
Token ID | 547 |
Feature activation | -3.01e-02 |
Pos logit contributions | |
injuries | +0.013 |
poisoning | +0.012 |
mercy | +0.011 |
extingu | +0.011 |
payoff | +0.011 |
Neg logit contributions | |
aling | -0.014 |
irl | -0.013 |
alling | -0.012 |
c | -0.012 |
sings | -0.011 |
Token | lots |
Token ID | 6041 |
Feature activation | +2.57e-03 |
Pos logit contributions | |
aling | +0.045 |
alling | +0.039 |
astics | +0.037 |
irl | +0.036 |
sings | +0.035 |
Neg logit contributions | |
poisoning | -0.048 |
guard | -0.048 |
injuries | -0.048 |
unfortunately | -0.043 |
payoff | -0.043 |
Token | of |
Token ID | 286 |
Feature activation | -1.62e-03 |
Pos logit contributions | |
injuries | +0.002 |
poisoning | +0.002 |
guard | +0.002 |
unfortunately | +0.002 |
payoff | +0.002 |
Neg logit contributions | |
aling | -0.006 |
alling | -0.005 |
irl | -0.005 |
astics | -0.005 |
sings | -0.005 |
Token | colourful |
Token ID | 44799 |
Feature activation | -2.34e-02 |
Pos logit contributions | |
aling | +0.003 |
alling | +0.003 |
astics | +0.002 |
ophone | +0.002 |
irl | +0.002 |
Neg logit contributions | |
guard | -0.002 |
poisoning | -0.002 |
injuries | -0.002 |
last | -0.002 |
unfortunately | -0.002 |
Token | coils |
Token ID | 41331 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
aling | +0.038 |
alling | +0.033 |
astics | +0.031 |
irl | +0.031 |
sings | +0.029 |
Neg logit contributions | |
guard | -0.035 |
poisoning | -0.035 |
injuries | -0.034 |
unfortunately | -0.032 |
payoff | -0.031 |
Token | to |
Token ID | 284 |
Feature activation | +9.22e-03 |
Pos logit contributions | |
guard | +0.002 |
injuries | +0.002 |
poisoning | +0.002 |
unfortunately | +0.001 |
payoff | +0.001 |
Neg logit contributions | |
aling | -0.006 |
alling | -0.006 |
astics | -0.006 |
irl | -0.006 |
ophone | -0.006 |
Token | him |
Token ID | 683 |
Feature activation | +3.94e-03 |
Pos logit contributions | |
guard | +0.011 |
poisoning | +0.010 |
injuries | +0.010 |
payoff | +0.009 |
unfortunately | +0.009 |
Neg logit contributions | |
aling | -0.018 |
alling | -0.016 |
irl | -0.015 |
astics | -0.015 |
sings | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -2.99e-02 |
Pos logit contributions | |
poisoning | +0.003 |
guard | +0.003 |
injuries | +0.003 |
unfortunately | +0.003 |
last | +0.003 |
Neg logit contributions | |
aling | -0.009 |
alling | -0.008 |
irl | -0.008 |
astics | -0.008 |
sings | -0.008 |
Token | Tim |
Token ID | 5045 |
Feature activation | -2.24e-02 |
Pos logit contributions | |
aling | +0.057 |
alling | +0.052 |
irl | +0.050 |
astics | +0.049 |
sings | +0.047 |
Neg logit contributions | |
poisoning | -0.035 |
guard | -0.035 |
injuries | -0.034 |
unfortunately | -0.032 |
payoff | -0.031 |
Token | was |
Token ID | 373 |
Feature activation | -4.45e-02 |
Pos logit contributions | |
aling | +0.043 |
alling | +0.039 |
astics | +0.037 |
irl | +0.037 |
sings | +0.035 |
Neg logit contributions | |
guard | -0.026 |
poisoning | -0.026 |
injuries | -0.025 |
unfortunately | -0.024 |
payoff | -0.023 |
Token | so |
Token ID | 523 |
Feature activation | -2.02e-02 |
Pos logit contributions | |
aling | +0.082 |
alling | +0.074 |
astics | +0.070 |
irl | +0.070 |
sings | +0.067 |
Neg logit contributions | |
poisoning | -0.055 |
guard | -0.055 |
injuries | -0.054 |
unfortunately | -0.051 |
payoff | -0.049 |
Token | happy |
Token ID | 3772 |
Feature activation | +2.96e-02 |
Pos logit contributions | |
aling | +0.028 |
alling | +0.024 |
irl | +0.022 |
astics | +0.022 |
sings | +0.021 |
Neg logit contributions | |
injuries | -0.035 |
poisoning | -0.035 |
guard | -0.035 |
payoff | -0.032 |
unfortunately | -0.032 |
Token | and |
Token ID | 290 |
Feature activation | -3.22e-02 |
Pos logit contributions | |
injuries | +0.036 |
guard | +0.034 |
poisoning | +0.033 |
payoff | +0.031 |
ho | +0.031 |
Neg logit contributions | |
aling | -0.053 |
alling | -0.046 |
ped | -0.045 |
astics | -0.045 |
sings | -0.043 |
Token | thanked |
Token ID | 26280 |
Feature activation | +5.37e-02 |
Pos logit contributions | |
aling | +0.068 |
alling | +0.063 |
irl | +0.060 |
astics | +0.060 |
ophone | +0.057 |
Neg logit contributions | |
guard | -0.032 |
poisoning | -0.032 |
unfortunately | -0.029 |
injuries | -0.029 |
payoff | -0.027 |
Token | the |
Token ID | 262 |
Feature activation | +1.80e-02 |
Pos logit contributions | |
poisoning | +0.079 |
guard | +0.078 |
t | +0.075 |
injuries | +0.075 |
less | +0.073 |
Neg logit contributions | |
aling | -0.082 |
alling | -0.072 |
irl | -0.068 |
sings | -0.065 |
astics | -0.064 |
Token | lady |
Token ID | 10846 |
Feature activation | +1.10e-02 |
Pos logit contributions | |
guard | +0.009 |
poisoning | +0.009 |
injuries | +0.008 |
payoff | +0.007 |
last | +0.006 |
Neg logit contributions | |
aling | -0.048 |
alling | -0.045 |
astics | -0.043 |
irl | -0.043 |
ophone | -0.041 |
Token | was |
Token ID | 373 |
Feature activation | -3.30e-02 |
Pos logit contributions | |
guard | +0.015 |
injuries | +0.015 |
poisoning | +0.015 |
unfortunately | +0.013 |
payoff | +0.013 |
Neg logit contributions | |
aling | -0.020 |
alling | -0.018 |
astics | -0.017 |
irl | -0.017 |
sings | -0.016 |
Token | still |
Token ID | 991 |
Feature activation | -6.19e-03 |
Pos logit contributions | |
aling | +0.057 |
alling | +0.051 |
astics | +0.048 |
irl | +0.048 |
sings | +0.046 |
Neg logit contributions | |
guard | -0.045 |
poisoning | -0.045 |
injuries | -0.044 |
unfortunately | -0.041 |
payoff | -0.040 |
Token | by |
Token ID | 416 |
Feature activation | -3.70e-03 |
Pos logit contributions | |
aling | +0.010 |
alling | +0.009 |
astics | +0.008 |
irl | +0.008 |
sings | +0.008 |
Neg logit contributions | |
guard | -0.009 |
poisoning | -0.009 |
injuries | -0.009 |
unfortunately | -0.008 |
payoff | -0.008 |
Token | her |
Token ID | 607 |
Feature activation | +4.03e-03 |
Pos logit contributions | |
aling | +0.006 |
alling | +0.006 |
astics | +0.005 |
irl | +0.005 |
sings | +0.005 |
Neg logit contributions | |
guard | -0.005 |
poisoning | -0.005 |
injuries | -0.005 |
t | -0.004 |
department | -0.004 |
Token | side |
Token ID | 1735 |
Feature activation | +2.41e-02 |
Pos logit contributions | |
guard | +0.006 |
poisoning | +0.006 |
injuries | +0.005 |
unfortunately | +0.005 |
t | +0.005 |
Neg logit contributions | |
aling | -0.007 |
alling | -0.006 |
astics | -0.006 |
irl | -0.005 |
sings | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | -3.03e-02 |
Pos logit contributions | |
guard | +0.027 |
poisoning | +0.026 |
injuries | +0.025 |
payoff | +0.024 |
unfortunately | +0.023 |
Neg logit contributions | |
aling | -0.047 |
alling | -0.043 |
astics | -0.039 |
c | -0.039 |
irl | -0.039 |
Token | Mom |
Token ID | 11254 |
Feature activation | +1.83e-02 |
Pos logit contributions | |
aling | +0.056 |
alling | +0.051 |
irl | +0.049 |
astics | +0.048 |
sings | +0.045 |
Neg logit contributions | |
poisoning | -0.037 |
guard | -0.037 |
injuries | -0.036 |
unfortunately | -0.035 |
payoff | -0.033 |
Token | my |
Token ID | 1820 |
Feature activation | +7.30e-03 |
Pos logit contributions | |
guard | +0.025 |
poisoning | +0.024 |
injuries | +0.023 |
unfortunately | +0.022 |
last | +0.022 |
Neg logit contributions | |
aling | -0.032 |
alling | -0.028 |
astics | -0.027 |
irl | -0.027 |
sings | -0.025 |
Token | said |
Token ID | 531 |
Feature activation | +2.65e-02 |
Pos logit contributions | |
guard | +0.009 |
poisoning | +0.009 |
injuries | +0.009 |
unfortunately | +0.008 |
payoff | +0.008 |
Neg logit contributions | |
aling | -0.013 |
alling | -0.012 |
astics | -0.011 |
irl | -0.011 |
sings | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | +4.09e-02 |
Pos logit contributions | |
injuries | +0.030 |
guard | +0.028 |
poisoning | +0.028 |
payoff | +0.024 |
extingu | +0.024 |
Neg logit contributions | |
aling | -0.048 |
alling | -0.044 |
irl | -0.043 |
vel | -0.043 |
sings | -0.043 |
Token | � |
Token ID | 6184 |
Feature activation | -5.94e-02 |
Pos logit contributions | |
guard | +0.051 |
poisoning | +0.051 |
injuries | +0.050 |
unfortunately | +0.046 |
payoff | +0.045 |
Neg logit contributions | |
aling | -0.075 |
alling | -0.068 |
astics | -0.064 |
irl | -0.063 |
sings | -0.061 |
Token | a |
Token ID | 257 |
Feature activation | -4.50e-03 |
Pos logit contributions | |
aling | +0.018 |
alling | +0.016 |
astics | +0.015 |
irl | +0.015 |
sings | +0.014 |
Neg logit contributions | |
guard | -0.011 |
poisoning | -0.011 |
injuries | -0.011 |
unfortunately | -0.010 |
payoff | -0.010 |
Token | box |
Token ID | 3091 |
Feature activation | -2.18e-02 |
Pos logit contributions | |
aling | +0.008 |
alling | +0.007 |
astics | +0.007 |
irl | +0.007 |
sings | +0.007 |
Neg logit contributions | |
poisoning | -0.006 |
guard | -0.006 |
injuries | -0.006 |
unfortunately | -0.005 |
payoff | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | -1.12e-02 |
Pos logit contributions | |
aling | +0.045 |
alling | +0.041 |
astics | +0.039 |
irl | +0.038 |
sings | +0.037 |
Neg logit contributions | |
guard | -0.023 |
poisoning | -0.022 |
injuries | -0.022 |
unfortunately | -0.020 |
payoff | -0.019 |
Token | The |
Token ID | 383 |
Feature activation | +2.69e-03 |
Pos logit contributions | |
aling | +0.021 |
alling | +0.019 |
irl | +0.019 |
astics | +0.019 |
irled | +0.017 |
Neg logit contributions | |
guard | -0.013 |
poisoning | -0.013 |
unfortunately | -0.012 |
injuries | -0.012 |
payoff | -0.011 |
Token | man |
Token ID | 582 |
Feature activation | -2.57e-02 |
Pos logit contributions | |
guard | +0.003 |
poisoning | +0.003 |
injuries | +0.003 |
payoff | +0.002 |
unfortunately | +0.002 |
Neg logit contributions | |
aling | -0.006 |
alling | -0.005 |
astics | -0.005 |
irl | -0.005 |
sings | -0.005 |
Token | looked |
Token ID | 3114 |
Feature activation | -4.44e-02 |
Pos logit contributions | |
aling | +0.053 |
alling | +0.048 |
astics | +0.047 |
irl | +0.046 |
etition | +0.043 |
Neg logit contributions | |
guard | -0.027 |
poisoning | -0.026 |
injuries | -0.024 |
unfortunately | -0.024 |
last | -0.023 |
Token | up |
Token ID | 510 |
Feature activation | -1.63e-02 |
Pos logit contributions | |
aling | +0.072 |
alling | +0.064 |
astics | +0.063 |
irl | +0.059 |
sings | +0.055 |
Neg logit contributions | |
poisoning | -0.066 |
guard | -0.064 |
injuries | -0.063 |
unfortunately | -0.061 |
though | -0.060 |
Token | at |
Token ID | 379 |
Feature activation | -1.47e-02 |
Pos logit contributions | |
aling | +0.020 |
alling | +0.017 |
irl | +0.015 |
ped | +0.014 |
irled | +0.014 |
Neg logit contributions | |
guard | -0.030 |
injuries | -0.029 |
poisoning | -0.029 |
payoff | -0.027 |
mercy | -0.027 |
Token | the |
Token ID | 262 |
Feature activation | +4.31e-03 |
Pos logit contributions | |
aling | +0.016 |
sings | +0.015 |
c | +0.014 |
alling | +0.014 |
aled | +0.014 |
Neg logit contributions | |
guard | -0.025 |
poisoning | -0.025 |
t | -0.025 |
injuries | -0.024 |
unfortunately | -0.024 |
Token | girl |
Token ID | 2576 |
Feature activation | -6.67e-03 |
Pos logit contributions | |
guard | +0.005 |
poisoning | +0.005 |
injuries | +0.005 |
unfortunately | +0.004 |
payoff | +0.004 |
Neg logit contributions | |
aling | -0.009 |
alling | -0.008 |
astics | -0.008 |
irl | -0.008 |
sings | -0.007 |
Token | , |
Token ID | 11 |
Feature activation | +1.79e-02 |
Pos logit contributions | |
aling | +0.016 |
irl | +0.015 |
alling | +0.015 |
astics | +0.015 |
irled | +0.014 |
Neg logit contributions | |
poisoning | -0.004 |
guard | -0.004 |
injuries | -0.004 |
unfortunately | -0.004 |
last | -0.003 |
Token | do |
Token ID | 466 |
Feature activation | +2.31e-02 |
Pos logit contributions | |
guard | +0.026 |
poisoning | +0.025 |
injuries | +0.024 |
unfortunately | +0.022 |
last | +0.021 |
Neg logit contributions | |
aling | -0.055 |
alling | -0.051 |
astics | -0.049 |
irl | -0.048 |
irled | -0.046 |
Token | ggy |
Token ID | 19970 |
Feature activation | -3.18e-02 |
Pos logit contributions | |
guard | +0.024 |
poisoning | +0.022 |
that | +0.022 |
injuries | +0.022 |
unfortunately | +0.022 |
Neg logit contributions | |
aling | -0.049 |
alling | -0.046 |
aled | -0.041 |
atters | -0.040 |
ophone | -0.039 |
Token | !" |
Token ID | 2474 |
Feature activation | -2.73e-03 |
Pos logit contributions | |
aling | +0.062 |
alling | +0.057 |
astics | +0.054 |
irl | +0.053 |
sings | +0.051 |
Neg logit contributions | |
guard | -0.036 |
poisoning | -0.036 |
injuries | -0.035 |
unfortunately | -0.032 |
payoff | -0.031 |
Token | and |
Token ID | 290 |
Feature activation | -3.14e-02 |
Pos logit contributions | |
aling | +0.005 |
alling | +0.004 |
astics | +0.004 |
irl | +0.004 |
sings | +0.004 |
Neg logit contributions | |
guard | -0.004 |
poisoning | -0.004 |
injuries | -0.004 |
unfortunately | -0.004 |
payoff | -0.004 |
Token | he |
Token ID | 339 |
Feature activation | -9.85e-03 |
Pos logit contributions | |
aling | +0.065 |
alling | +0.060 |
astics | +0.057 |
irl | +0.056 |
sings | +0.054 |
Neg logit contributions | |
guard | -0.032 |
poisoning | -0.032 |
injuries | -0.031 |
unfortunately | -0.028 |
payoff | -0.027 |
Token | ran |
Token ID | 4966 |
Feature activation | -4.63e-02 |
Pos logit contributions | |
aling | +0.017 |
alling | +0.016 |
astics | +0.015 |
irl | +0.014 |
atters | +0.014 |
Neg logit contributions | |
guard | -0.014 |
poisoning | -0.013 |
injuries | -0.012 |
unfortunately | -0.012 |
last | -0.012 |
Token | away |
Token ID | 1497 |
Feature activation | +2.08e-02 |
Pos logit contributions | |
aling | +0.091 |
alling | +0.082 |
astics | +0.078 |
irl | +0.078 |
sings | +0.075 |
Neg logit contributions | |
guard | -0.053 |
poisoning | -0.053 |
injuries | -0.051 |
unfortunately | -0.046 |
payoff | -0.046 |
Token | . |
Token ID | 13 |
Feature activation | -2.14e-02 |
Pos logit contributions | |
poisoning | +0.026 |
injuries | +0.026 |
guard | +0.025 |
unfortunately | +0.023 |
payoff | +0.023 |
Neg logit contributions | |
aling | -0.036 |
alling | -0.032 |
astics | -0.030 |
irl | -0.030 |
ped | -0.029 |
Token | |
Token ID | 220 |
Feature activation | -6.71e-03 |
Pos logit contributions | |
aling | +0.042 |
alling | +0.038 |
astics | +0.036 |
irl | +0.036 |
sings | +0.034 |
Neg logit contributions | |
guard | -0.024 |
poisoning | -0.024 |
injuries | -0.024 |
unfortunately | -0.022 |
payoff | -0.021 |
Token |
|
Token ID | 198 |
Feature activation | +2.63e-02 |
Pos logit contributions | |
aling | +0.015 |
alling | +0.014 |
astics | +0.013 |
irl | +0.013 |
sings | +0.013 |
Neg logit contributions | |
guard | -0.005 |
poisoning | -0.005 |
injuries | -0.005 |
unfortunately | -0.005 |
t | -0.004 |
Token |
|
Token ID | 198 |
Feature activation | +2.19e-02 |
Pos logit contributions | |
injuries | +0.034 |
poisoning | +0.034 |
guard | +0.033 |
payoff | +0.030 |
unfortunately | +0.029 |
Neg logit contributions | |
aling | -0.048 |
alling | -0.043 |
irl | -0.041 |
astics | -0.041 |
sings | -0.039 |
Token | had |
Token ID | 550 |
Feature activation | -3.74e-03 |
Pos logit contributions | |
injuries | +0.014 |
poisoning | +0.014 |
guard | +0.014 |
unfortunately | +0.014 |
payoff | +0.014 |
Neg logit contributions | |
aling | -0.018 |
alling | -0.016 |
astics | -0.015 |
sings | -0.015 |
irl | -0.015 |
Token | to |
Token ID | 284 |
Feature activation | +4.30e-03 |
Pos logit contributions | |
aling | +0.007 |
alling | +0.007 |
astics | +0.006 |
irl | +0.006 |
sings | +0.006 |
Neg logit contributions | |
guard | -0.004 |
poisoning | -0.004 |
injuries | -0.004 |
unfortunately | -0.004 |
payoff | -0.004 |
Token | come |
Token ID | 1282 |
Feature activation | -1.77e-02 |
Pos logit contributions | |
guard | +0.006 |
poisoning | +0.006 |
injuries | +0.006 |
unfortunately | +0.006 |
payoff | +0.005 |
Neg logit contributions | |
aling | -0.007 |
alling | -0.006 |
astics | -0.006 |
irl | -0.006 |
sings | -0.006 |
Token | with |
Token ID | 351 |
Feature activation | -8.16e-03 |
Pos logit contributions | |
aling | +0.031 |
alling | +0.029 |
astics | +0.026 |
irl | +0.026 |
ophone | +0.025 |
Neg logit contributions | |
guard | -0.023 |
poisoning | -0.023 |
injuries | -0.022 |
payoff | -0.022 |
unfortunately | -0.022 |
Token | her |
Token ID | 607 |
Feature activation | +3.95e-02 |
Pos logit contributions | |
aling | +0.014 |
alling | +0.013 |
irl | +0.012 |
astics | +0.012 |
sings | +0.011 |
Neg logit contributions | |
guard | -0.011 |
poisoning | -0.010 |
unfortunately | -0.010 |
department | -0.010 |
ho | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | -2.65e-02 |
Pos logit contributions | |
guard | +0.060 |
payoff | +0.058 |
unfortunately | +0.056 |
poisoning | +0.056 |
Not | +0.056 |
Neg logit contributions | |
aling | -0.053 |
ophone | -0.048 |
alling | -0.047 |
sings | -0.046 |
c | -0.046 |
Token | She |
Token ID | 1375 |
Feature activation | -6.68e-04 |
Pos logit contributions | |
aling | +0.042 |
alling | +0.038 |
astics | +0.036 |
irl | +0.035 |
sings | +0.034 |
Neg logit contributions | |
guard | -0.041 |
poisoning | -0.039 |
injuries | -0.039 |
payoff | -0.036 |
unfortunately | -0.035 |
Token | said |
Token ID | 531 |
Feature activation | +3.47e-02 |
Pos logit contributions | |
aling | +0.001 |
alling | +0.001 |
astics | +0.001 |
irl | +0.001 |
sings | +0.001 |
Neg logit contributions | |
guard | -0.001 |
poisoning | -0.001 |
injuries | -0.001 |
unfortunately | -0.001 |
payoff | -0.001 |
Token | he |
Token ID | 339 |
Feature activation | +3.33e-02 |
Pos logit contributions | |
poisoning | +0.037 |
guard | +0.037 |
injuries | +0.036 |
extingu | +0.034 |
payoff | +0.033 |
Neg logit contributions | |
aling | -0.064 |
sings | -0.059 |
irl | -0.058 |
alling | -0.057 |
astics | -0.057 |
Token | would |
Token ID | 561 |
Feature activation | +1.65e-02 |
Pos logit contributions | |
guard | +0.035 |
poisoning | +0.035 |
injuries | +0.034 |
unfortunately | +0.030 |
t | +0.030 |
Neg logit contributions | |
aling | -0.069 |
alling | -0.063 |
astics | -0.059 |
irl | -0.059 |
sings | -0.056 |
Token | be |
Token ID | 307 |
Feature activation | -4.68e-02 |
Pos logit contributions | |
guard | +0.023 |
poisoning | +0.022 |
injuries | +0.021 |
unfortunately | +0.020 |
department | +0.020 |
Neg logit contributions | |
aling | -0.028 |
alling | -0.025 |
irl | -0.024 |
astics | -0.023 |
sings | -0.022 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -4.17e-03 |
Pos logit contributions | |
aling | +0.056 |
alling | +0.052 |
astics | +0.050 |
irl | +0.049 |
sings | +0.047 |
Neg logit contributions | |
poisoning | -0.020 |
guard | -0.020 |
injuries | -0.020 |
unfortunately | -0.018 |
payoff | -0.017 |
Token | One |
Token ID | 3198 |
Feature activation | +1.79e-02 |
Pos logit contributions | |
aling | +0.008 |
alling | +0.008 |
irl | +0.007 |
astics | +0.007 |
sings | +0.007 |
Neg logit contributions | |
poisoning | -0.004 |
injuries | -0.004 |
guard | -0.004 |
unfortunately | -0.004 |
payoff | -0.004 |
Token | day |
Token ID | 1110 |
Feature activation | +4.63e-02 |
Pos logit contributions | |
guard | +0.020 |
poisoning | +0.019 |
injuries | +0.019 |
unfortunately | +0.017 |
payoff | +0.017 |
Neg logit contributions | |
aling | -0.036 |
alling | -0.033 |
astics | -0.031 |
irl | -0.031 |
sings | -0.030 |
Token | , |
Token ID | 11 |
Feature activation | +2.27e-02 |
Pos logit contributions | |
injuries | +0.054 |
guard | +0.051 |
poisoning | +0.050 |
payoff | +0.047 |
department | +0.045 |
Neg logit contributions | |
aling | -0.086 |
irl | -0.085 |
irled | -0.080 |
astics | -0.078 |
alling | -0.076 |
Token | a |
Token ID | 257 |
Feature activation | -2.15e-02 |
Pos logit contributions | |
injuries | +0.036 |
guard | +0.035 |
poisoning | +0.034 |
payoff | +0.032 |
department | +0.031 |
Neg logit contributions | |
aling | -0.034 |
alling | -0.030 |
irl | -0.030 |
astics | -0.029 |
irled | -0.028 |
Token | sw |
Token ID | 1509 |
Feature activation | -1.28e-01 |
Pos logit contributions | |
aling | +0.051 |
alling | +0.048 |
astics | +0.045 |
irl | +0.043 |
aled | +0.043 |
Neg logit contributions | |
guard | -0.017 |
poisoning | -0.017 |
injuries | -0.016 |
last | -0.014 |
unfortunately | -0.013 |
Token | an |
Token ID | 272 |
Feature activation | -6.47e-02 |
Pos logit contributions | |
aling | +0.229 |
sings | +0.210 |
alling | +0.204 |
astics | +0.200 |
vel | +0.188 |
Neg logit contributions | |
guard | -0.172 |
ho | -0.165 |
poisoning | -0.153 |
injuries | -0.145 |
t | -0.142 |
Token | named |
Token ID | 3706 |
Feature activation | -4.04e-02 |
Pos logit contributions | |
aling | +0.120 |
alling | +0.109 |
astics | +0.103 |
irl | +0.101 |
sings | +0.095 |
Neg logit contributions | |
guard | -0.081 |
poisoning | -0.079 |
injuries | -0.077 |
unfortunately | -0.071 |
ho | -0.070 |
Token | Sam |
Token ID | 3409 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
aling | +0.098 |
alling | +0.091 |
astics | +0.086 |
irl | +0.083 |
essay | +0.080 |
Neg logit contributions | |
poisoning | -0.028 |
that | -0.027 |
ho | -0.027 |
last | -0.026 |
guard | -0.025 |
Token | was |
Token ID | 373 |
Feature activation | -3.74e-02 |
Pos logit contributions | |
aling | +0.024 |
alling | +0.022 |
astics | +0.020 |
irl | +0.020 |
sings | +0.019 |
Neg logit contributions | |
guard | -0.017 |
poisoning | -0.017 |
injuries | -0.017 |
unfortunately | -0.015 |
payoff | -0.015 |
Token | swimming |
Token ID | 14899 |
Feature activation | -7.07e-02 |
Pos logit contributions | |
aling | +0.079 |
alling | +0.072 |
astics | +0.069 |
etition | +0.069 |
essay | +0.068 |
Neg logit contributions | |
that | -0.037 |
to | -0.037 |
poisoning | -0.034 |
help | -0.033 |
last | -0.033 |
Token | play |
Token ID | 711 |
Feature activation | -3.17e-02 |
Pos logit contributions | |
aling | +0.024 |
alling | +0.022 |
astics | +0.019 |
ophone | +0.019 |
irled | +0.018 |
Neg logit contributions | |
help | -0.016 |
poisoning | -0.016 |
guard | -0.016 |
injuries | -0.016 |
last | -0.015 |
Token | with |
Token ID | 351 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
aling | +0.057 |
alling | +0.051 |
astics | +0.046 |
ophone | +0.044 |
sings | +0.044 |
Neg logit contributions | |
poisoning | -0.044 |
guard | -0.042 |
injuries | -0.041 |
unfortunately | -0.039 |
while | -0.037 |
Token | her |
Token ID | 607 |
Feature activation | +1.79e-02 |
Pos logit contributions | |
aling | +0.022 |
alling | +0.020 |
astics | +0.019 |
irl | +0.019 |
sings | +0.018 |
Neg logit contributions | |
poisoning | -0.012 |
guard | -0.012 |
injuries | -0.011 |
unfortunately | -0.010 |
payoff | -0.010 |
Token | toys |
Token ID | 14958 |
Feature activation | +1.13e-02 |
Pos logit contributions | |
guard | +0.035 |
payoff | +0.034 |
injuries | +0.034 |
poisoning | +0.034 |
unfortunately | +0.033 |
Neg logit contributions | |
irl | -0.016 |
sings | -0.016 |
nic | -0.015 |
astics | -0.014 |
irled | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | -1.98e-02 |
Pos logit contributions | |
injuries | +0.019 |
payoff | +0.018 |
guard | +0.018 |
mercy | +0.018 |
hesitation | +0.017 |
Neg logit contributions | |
aling | -0.015 |
alling | -0.013 |
irl | -0.012 |
astics | -0.012 |
rides | -0.012 |
Token | c |
Token ID | 269 |
Feature activation | -8.42e-02 |
Pos logit contributions | |
aling | +0.028 |
alling | +0.024 |
irl | +0.023 |
astics | +0.023 |
sings | +0.023 |
Neg logit contributions | |
guard | -0.033 |
poisoning | -0.032 |
injuries | -0.032 |
unfortunately | -0.030 |
payoff | -0.030 |
Token | uddle |
Token ID | 24500 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
aling | +0.168 |
alling | +0.151 |
astics | +0.133 |
sings | +0.132 |
ophone | +0.129 |
Neg logit contributions | |
poisoning | -0.102 |
injuries | -0.099 |
guard | -0.094 |
ho | -0.087 |
t | -0.086 |
Token | with |
Token ID | 351 |
Feature activation | -1.54e-02 |
Pos logit contributions | |
aling | +0.026 |
alling | +0.023 |
irl | +0.022 |
astics | +0.022 |
sings | +0.020 |
Neg logit contributions | |
guard | -0.022 |
injuries | -0.022 |
poisoning | -0.022 |
payoff | -0.021 |
unfortunately | -0.020 |
Token | her |
Token ID | 607 |
Feature activation | +1.61e-02 |
Pos logit contributions | |
aling | +0.027 |
alling | +0.025 |
irl | +0.024 |
astics | +0.024 |
sings | +0.023 |
Neg logit contributions | |
guard | -0.020 |
poisoning | -0.020 |
injuries | -0.019 |
payoff | -0.018 |
unfortunately | -0.018 |
Token | t |
Token ID | 256 |
Feature activation | -2.22e-02 |
Pos logit contributions | |
guard | +0.032 |
payoff | +0.032 |
unfortunately | +0.030 |
poisoning | +0.030 |
injuries | +0.030 |
Neg logit contributions | |
irl | -0.014 |
nic | -0.014 |
sings | -0.013 |
c | -0.013 |
zoom | -0.012 |
Token | eddy |
Token ID | 21874 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
aling | +0.034 |
alling | +0.030 |
irl | +0.029 |
astics | +0.028 |
ophone | +0.027 |
Neg logit contributions | |
poisoning | -0.033 |
injuries | -0.033 |
payoff | -0.032 |
guard | -0.032 |
unfortunately | -0.031 |
Token | messy |
Token ID | 24097 |
Feature activation | +1.43e-02 |
Pos logit contributions | |
aling | +0.089 |
alling | +0.083 |
astics | +0.076 |
irl | +0.073 |
irled | +0.070 |
Neg logit contributions | |
guard | -0.062 |
poisoning | -0.062 |
injuries | -0.062 |
unfortunately | -0.057 |
last | -0.057 |
Token | . |
Token ID | 13 |
Feature activation | -3.61e-02 |
Pos logit contributions | |
injuries | +0.018 |
poisoning | +0.018 |
guard | +0.018 |
unfortunately | +0.017 |
payoff | +0.017 |
Neg logit contributions | |
aling | -0.025 |
alling | -0.022 |
astics | -0.022 |
irl | -0.022 |
sings | -0.021 |
Token | Like |
Token ID | 4525 |
Feature activation | +6.77e-03 |
Pos logit contributions | |
aling | +0.058 |
alling | +0.052 |
irl | +0.049 |
astics | +0.047 |
irled | +0.044 |
Neg logit contributions | |
poisoning | -0.051 |
guard | -0.050 |
injuries | -0.050 |
unfortunately | -0.050 |
ful | -0.048 |
Token | when |
Token ID | 618 |
Feature activation | +2.48e-02 |
Pos logit contributions | |
guard | +0.007 |
poisoning | +0.007 |
injuries | +0.007 |
unfortunately | +0.006 |
payoff | +0.006 |
Neg logit contributions | |
aling | -0.014 |
alling | -0.012 |
astics | -0.012 |
irl | -0.012 |
sings | -0.011 |
Token | I |
Token ID | 314 |
Feature activation | +1.68e-02 |
Pos logit contributions | |
poisoning | +0.029 |
guard | +0.029 |
injuries | +0.028 |
unfortunately | +0.026 |
payoff | +0.026 |
Neg logit contributions | |
aling | -0.047 |
alling | -0.042 |
irl | -0.041 |
astics | -0.041 |
sings | -0.039 |
Token | sne |
Token ID | 10505 |
Feature activation | -1.04e-01 |
Pos logit contributions | |
guard | +0.021 |
poisoning | +0.021 |
injuries | +0.021 |
unfortunately | +0.019 |
payoff | +0.019 |
Neg logit contributions | |
aling | -0.031 |
alling | -0.028 |
irl | -0.026 |
astics | -0.026 |
sings | -0.025 |
Token | e |
Token ID | 68 |
Feature activation | -5.90e-02 |
Pos logit contributions | |
aling | +0.122 |
alling | +0.112 |
etition | +0.097 |
astics | +0.094 |
nic | +0.092 |
Neg logit contributions | |
poisoning | -0.179 |
injuries | -0.179 |
ful | -0.176 |
guard | -0.169 |
help | -0.166 |
Token | ze |
Token ID | 2736 |
Feature activation | -1.43e-02 |
Pos logit contributions | |
aling | +0.090 |
alling | +0.080 |
astics | +0.073 |
irl | +0.070 |
sings | +0.065 |
Neg logit contributions | |
poisoning | -0.092 |
guard | -0.091 |
injuries | -0.090 |
unfortunately | -0.084 |
ho | -0.083 |
Token | . |
Token ID | 13 |
Feature activation | -3.96e-02 |
Pos logit contributions | |
aling | +0.027 |
alling | +0.024 |
astics | +0.023 |
irl | +0.023 |
sings | +0.022 |
Neg logit contributions | |
injuries | -0.017 |
guard | -0.017 |
poisoning | -0.017 |
payoff | -0.015 |
unfortunately | -0.015 |
Token | Ach |
Token ID | 26219 |
Feature activation | -1.64e-02 |
Pos logit contributions | |
aling | +0.064 |
alling | +0.057 |
irl | +0.054 |
astics | +0.053 |
irled | +0.050 |
Neg logit contributions | |
poisoning | -0.056 |
injuries | -0.055 |
guard | -0.054 |
unfortunately | -0.053 |
Not | -0.051 |
Token | oo |
Token ID | 2238 |
Feature activation | -3.65e-02 |
Pos logit contributions | |
aling | +0.027 |
alling | +0.025 |
astics | +0.023 |
irl | +0.022 |
sings | +0.021 |
Neg logit contributions | |
injuries | -0.023 |
poisoning | -0.023 |
guard | -0.022 |
unfortunately | -0.021 |
t | -0.020 |
Token | " |
Token ID | 1 |
Feature activation | +2.48e-02 |
Pos logit contributions | |
guard | +0.020 |
poisoning | +0.020 |
injuries | +0.020 |
unfortunately | +0.017 |
payoff | +0.017 |
Neg logit contributions | |
aling | -0.041 |
alling | -0.037 |
astics | -0.035 |
irl | -0.035 |
sings | -0.033 |
Token | R |
Token ID | 49 |
Feature activation | -2.36e-02 |
Pos logit contributions | |
guard | +0.031 |
poisoning | +0.030 |
injuries | +0.029 |
payoff | +0.027 |
unfortunately | +0.027 |
Neg logit contributions | |
aling | -0.047 |
alling | -0.042 |
astics | -0.040 |
irl | -0.040 |
sings | -0.038 |
Token | ib |
Token ID | 571 |
Feature activation | -3.53e-02 |
Pos logit contributions | |
aling | +0.040 |
alling | +0.036 |
irl | +0.031 |
astics | +0.031 |
ophone | +0.031 |
Neg logit contributions | |
injuries | -0.034 |
poisoning | -0.034 |
guard | -0.034 |
payoff | -0.030 |
ho | -0.030 |
Token | bit |
Token ID | 2545 |
Feature activation | +2.40e-02 |
Pos logit contributions | |
aling | +0.056 |
alling | +0.050 |
astics | +0.046 |
irl | +0.044 |
ophone | +0.041 |
Neg logit contributions | |
injuries | -0.053 |
poisoning | -0.052 |
guard | -0.051 |
ho | -0.049 |
last | -0.048 |
Token | , |
Token ID | 11 |
Feature activation | +2.11e-02 |
Pos logit contributions | |
poisoning | +0.024 |
injuries | +0.024 |
guard | +0.024 |
unfortunately | +0.021 |
payoff | +0.020 |
Neg logit contributions | |
aling | -0.050 |
alling | -0.045 |
irl | -0.044 |
astics | -0.043 |
sings | -0.043 |
Token | rib |
Token ID | 12183 |
Feature activation | -8.63e-02 |
Pos logit contributions | |
guard | +0.034 |
poisoning | +0.033 |
injuries | +0.032 |
unfortunately | +0.031 |
payoff | +0.030 |
Neg logit contributions | |
aling | -0.032 |
alling | -0.028 |
irl | -0.026 |
astics | -0.026 |
sings | -0.025 |
Token | bit |
Token ID | 2545 |
Feature activation | +2.35e-02 |
Pos logit contributions | |
aling | +0.139 |
alling | +0.127 |
ophone | +0.115 |
astics | +0.114 |
irl | +0.114 |
Neg logit contributions | |
injuries | -0.122 |
poisoning | -0.121 |
payoff | -0.113 |
guard | -0.112 |
last | -0.111 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.79e-02 |
Pos logit contributions | |
poisoning | +0.028 |
injuries | +0.027 |
guard | +0.027 |
unfortunately | +0.025 |
payoff | +0.024 |
Neg logit contributions | |
aling | -0.045 |
alling | -0.040 |
irl | -0.039 |
astics | -0.038 |
sings | -0.038 |
Token | Lily |
Token ID | 20037 |
Feature activation | -7.66e-03 |
Pos logit contributions | |
poisoning | +0.024 |
injuries | +0.024 |
guard | +0.023 |
department | +0.021 |
ho | +0.021 |
Neg logit contributions | |
aling | -0.032 |
alling | -0.028 |
sings | -0.027 |
irl | -0.027 |
astics | -0.026 |
Token | says |
Token ID | 1139 |
Feature activation | +1.27e-02 |
Pos logit contributions | |
aling | +0.013 |
alling | +0.012 |
irl | +0.011 |
astics | +0.011 |
vel | +0.011 |
Neg logit contributions | |
guard | -0.010 |
last | -0.009 |
poisoning | -0.009 |
payoff | -0.009 |
but | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | -3.42e-02 |
Pos logit contributions | |
guard | +0.011 |
injuries | +0.011 |
poisoning | +0.011 |
unfortunately | +0.010 |
payoff | +0.009 |
Neg logit contributions | |
aling | -0.028 |
alling | -0.026 |
astics | -0.025 |
irl | -0.025 |
sings | -0.023 |
Token | parent |
Token ID | 2560 |
Feature activation | +2.75e-02 |
Pos logit contributions | |
injuries | +0.000 |
poisoning | +0.000 |
guard | +0.000 |
unfortunately | +0.000 |
last | +0.000 |
Neg logit contributions | |
aling | -0.001 |
alling | -0.001 |
astics | -0.001 |
irl | -0.001 |
sings | -0.001 |
Token | smiled |
Token ID | 13541 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
guard | +0.030 |
poisoning | +0.029 |
injuries | +0.029 |
unfortunately | +0.026 |
payoff | +0.026 |
Neg logit contributions | |
aling | -0.056 |
alling | -0.051 |
astics | -0.048 |
irl | -0.048 |
sings | -0.046 |
Token | and |
Token ID | 290 |
Feature activation | -3.56e-02 |
Pos logit contributions | |
aling | +0.028 |
alling | +0.025 |
astics | +0.023 |
irl | +0.023 |
sings | +0.023 |
Neg logit contributions | |
guard | -0.016 |
injuries | -0.016 |
poisoning | -0.016 |
payoff | -0.014 |
unfortunately | -0.014 |
Token | said |
Token ID | 531 |
Feature activation | +3.59e-02 |
Pos logit contributions | |
aling | +0.087 |
alling | +0.080 |
astics | +0.076 |
irl | +0.074 |
atters | +0.073 |
Neg logit contributions | |
guard | -0.025 |
poisoning | -0.023 |
unfortunately | -0.022 |
payoff | -0.020 |
they | -0.020 |
Token | � |
Token ID | 6184 |
Feature activation | -4.62e-02 |
Pos logit contributions | |
injuries | +0.026 |
poisoning | +0.025 |
guard | +0.025 |
payoff | +0.020 |
unfortunately | +0.020 |
Neg logit contributions | |
aling | -0.085 |
alling | -0.078 |
astics | -0.075 |
irl | -0.075 |
sings | -0.073 |
Token | � |
Token ID | 95 |
Feature activation | -9.37e-02 |
Pos logit contributions | |
aling | +0.082 |
sings | +0.076 |
alling | +0.075 |
ophone | +0.072 |
c | +0.071 |
Neg logit contributions | |
guard | -0.060 |
poisoning | -0.055 |
t | -0.051 |
� | -0.048 |
ho | -0.048 |
Token | € |
Token ID | 26391 |
Feature activation | -5.27e-02 |
Pos logit contributions | |
aling | +0.134 |
sings | +0.134 |
goals | +0.130 |
irl | +0.130 |
am | +0.127 |
Neg logit contributions | |
poisoning | -0.124 |
t | -0.121 |
guard | -0.118 |
payoff | -0.114 |
� | -0.107 |
Token | � |
Token ID | 129 |
Feature activation | -9.27e-02 |
Pos logit contributions | |
aling | +0.115 |
alling | +0.108 |
aled | +0.100 |
alled | +0.094 |
irl | +0.090 |
Neg logit contributions | |
t | -0.049 |
poisoning | -0.048 |
injuries | -0.048 |
guard | -0.046 |
they | -0.046 |
Token | � |
Token ID | 241 |
Feature activation | -5.09e-03 |
Pos logit contributions | |
aling | +0.135 |
sings | +0.131 |
cycl | +0.131 |
ophone | +0.128 |
c | +0.125 |
Neg logit contributions | |
that | -0.109 |
poisoning | -0.105 |
t | -0.099 |
first | -0.094 |
yet | -0.094 |
Token | you |
Token ID | 5832 |
Feature activation | -4.24e-03 |
Pos logit contributions | |
aling | +0.009 |
alling | +0.008 |
astics | +0.008 |
irl | +0.008 |
sings | +0.007 |
Neg logit contributions | |
injuries | -0.007 |
poisoning | -0.007 |
guard | -0.006 |
unfortunately | -0.006 |
payoff | -0.006 |
Token | will |
Token ID | 481 |
Feature activation | +1.56e-02 |
Pos logit contributions | |
aling | +0.008 |
alling | +0.007 |
astics | +0.007 |
irl | +0.007 |
sings | +0.006 |
Neg logit contributions | |
guard | -0.005 |
poisoning | -0.005 |
injuries | -0.005 |
unfortunately | -0.005 |
t | -0.005 |