Token | wanted |
Token ID | 2227 |
Feature activation | +3.48e-01 |
Pos logit contributions | |
rily | +0.017 |
han | +0.017 |
overed | +0.017 |
ivating | +0.017 |
ls | +0.017 |
Neg logit contributions | |
bisc | -0.011 |
count | -0.011 |
wake | -0.010 |
patch | -0.010 |
pile | -0.010 |
Token | to |
Token ID | 284 |
Feature activation | +3.84e-02 |
Pos logit contributions | |
spe | +0.008 |
overed | +0.008 |
led | +0.007 |
ls | +0.007 |
otte | +0.007 |
Neg logit contributions | |
bisc | -0.009 |
wake | -0.008 |
count | -0.008 |
snack | -0.008 |
patch | -0.008 |
Token | be |
Token ID | 307 |
Feature activation | -7.41e-01 |
Pos logit contributions | |
spe | +0.001 |
overed | +0.001 |
ls | +0.001 |
led | +0.001 |
otte | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
wake | -0.000 |
snack | -0.000 |
count | -0.000 |
patch | -0.000 |
Token | a |
Token ID | 257 |
Feature activation | +8.89e-01 |
Pos logit contributions | |
bisc | +0.017 |
wake | +0.015 |
count | +0.015 |
snack | +0.015 |
patch | +0.014 |
Neg logit contributions | |
spe | -0.019 |
overed | -0.018 |
led | -0.018 |
ls | -0.018 |
otte | -0.017 |
Token | good |
Token ID | 922 |
Feature activation | +4.33e-01 |
Pos logit contributions | |
led | +0.026 |
overed | +0.025 |
spe | +0.025 |
rily | +0.024 |
ls | +0.024 |
Neg logit contributions | |
bisc | -0.019 |
wake | -0.016 |
snack | -0.016 |
count | -0.016 |
patch | -0.015 |
Token | gol |
Token ID | 43791 |
Feature activation | +6.87e+00 |
Pos logit contributions | |
overed | +0.012 |
spe | +0.012 |
led | +0.012 |
rily | +0.012 |
ls | +0.012 |
Neg logit contributions | |
bisc | -0.009 |
snack | -0.008 |
wake | -0.008 |
count | -0.008 |
patch | -0.008 |
Token | fer |
Token ID | 2232 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
spe | +0.129 |
overed | +0.124 |
rily | +0.124 |
ls | +0.123 |
led | +0.123 |
Neg logit contributions | |
bisc | -0.194 |
count | -0.177 |
wake | -0.173 |
come | -0.171 |
snack | -0.168 |
Token | like |
Token ID | 588 |
Feature activation | +7.09e-01 |
Pos logit contributions | |
spe | +0.043 |
overed | +0.042 |
led | +0.041 |
ls | +0.040 |
rily | +0.039 |
Neg logit contributions | |
bisc | -0.038 |
count | -0.034 |
wake | -0.034 |
snack | -0.033 |
patch | -0.032 |
Token | his |
Token ID | 465 |
Feature activation | +7.38e-01 |
Pos logit contributions | |
spe | +0.016 |
led | +0.016 |
overed | +0.016 |
ls | +0.015 |
otte | +0.015 |
Neg logit contributions | |
bisc | -0.018 |
wake | -0.016 |
count | -0.016 |
snack | -0.016 |
come | -0.015 |
Token | dad |
Token ID | 9955 |
Feature activation | +7.45e-02 |
Pos logit contributions | |
led | +0.022 |
overed | +0.022 |
rily | +0.021 |
spe | +0.021 |
ivating | +0.021 |
Neg logit contributions | |
bisc | -0.014 |
patch | -0.012 |
snack | -0.012 |
count | -0.012 |
strip | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | +3.69e-01 |
Pos logit contributions | |
spe | +0.002 |
led | +0.002 |
ls | +0.002 |
overed | +0.002 |
otte | +0.002 |
Neg logit contributions | |
bisc | -0.002 |
wake | -0.002 |
snack | -0.002 |
count | -0.001 |
patch | -0.001 |
Token | |
Token ID | 220 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
led | +0.045 |
spe | +0.045 |
overed | +0.041 |
otte | +0.040 |
ls | +0.039 |
Neg logit contributions | |
bisc | -0.050 |
wake | -0.048 |
count | -0.047 |
snack | -0.047 |
go | -0.047 |
Token |
|
Token ID | 198 |
Feature activation | +7.08e-01 |
Pos logit contributions | |
spe | +0.035 |
led | +0.034 |
otte | +0.034 |
ize | +0.034 |
han | +0.033 |
Neg logit contributions | |
bisc | -0.018 |
snack | -0.014 |
come | -0.014 |
're | -0.013 |
go | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | +3.50e-01 |
Pos logit contributions | |
led | +0.021 |
overed | +0.020 |
ls | +0.020 |
spe | +0.020 |
rily | +0.019 |
Neg logit contributions | |
bisc | -0.014 |
count | -0.011 |
snack | -0.011 |
wake | -0.011 |
patch | -0.011 |
Token | The |
Token ID | 464 |
Feature activation | -3.57e-01 |
Pos logit contributions | |
led | +0.010 |
ls | +0.010 |
overed | +0.009 |
rily | +0.009 |
spe | +0.009 |
Neg logit contributions | |
bisc | -0.007 |
count | -0.006 |
wake | -0.006 |
snack | -0.006 |
patch | -0.006 |
Token | town |
Token ID | 3240 |
Feature activation | +2.29e+00 |
Pos logit contributions | |
bisc | +0.007 |
wake | +0.006 |
snack | +0.006 |
count | +0.006 |
patch | +0.006 |
Neg logit contributions | |
overed | -0.010 |
led | -0.010 |
spe | -0.010 |
ls | -0.010 |
rily | -0.010 |
Token | spe |
Token ID | 4125 |
Feature activation | +6.71e+00 |
Pos logit contributions | |
Acc | +0.057 |
otte | +0.056 |
ivating | +0.056 |
rily | +0.055 |
ls | +0.054 |
Neg logit contributions | |
count | -0.046 |
go | -0.045 |
pile | -0.045 |
wake | -0.044 |
bisc | -0.044 |
Token | ople |
Token ID | 643 |
Feature activation | +4.88e-01 |
Pos logit contributions | |
ls | +0.143 |
rily | +0.137 |
otte | +0.135 |
overed | +0.125 |
Acc | +0.125 |
Neg logit contributions | |
est | -0.165 |
count | -0.152 |
're | -0.149 |
bisc | -0.145 |
go | -0.139 |
Token | had |
Token ID | 550 |
Feature activation | +8.77e-01 |
Pos logit contributions | |
ls | +0.014 |
overed | +0.013 |
rily | +0.013 |
otte | +0.013 |
spe | +0.013 |
Neg logit contributions | |
bisc | -0.009 |
count | -0.009 |
wake | -0.009 |
come | -0.008 |
go | -0.008 |
Token | to |
Token ID | 284 |
Feature activation | -4.79e-01 |
Pos logit contributions | |
overed | +0.021 |
spe | +0.021 |
led | +0.021 |
ls | +0.021 |
otte | +0.021 |
Neg logit contributions | |
bisc | -0.020 |
wake | -0.018 |
count | -0.018 |
snack | -0.018 |
patch | -0.017 |
Token | find |
Token ID | 1064 |
Feature activation | -9.97e-01 |
Pos logit contributions | |
bisc | +0.006 |
wake | +0.005 |
count | +0.005 |
snack | +0.004 |
patch | +0.004 |
Neg logit contributions | |
spe | -0.017 |
overed | -0.017 |
led | -0.017 |
ls | -0.017 |
otte | -0.016 |
Token | a |
Token ID | 257 |
Feature activation | +5.75e-02 |
Pos logit contributions | |
bisc | +0.024 |
wake | +0.022 |
count | +0.022 |
snack | +0.021 |
patch | +0.021 |
Neg logit contributions | |
spe | -0.023 |
overed | -0.023 |
led | -0.022 |
ls | -0.022 |
otte | -0.022 |
Token | ! |
Token ID | 0 |
Feature activation | -1.11e-01 |
Pos logit contributions | |
spe | +0.003 |
ize | +0.003 |
ized | +0.003 |
ls | +0.003 |
overed | +0.003 |
Neg logit contributions | |
bisc | -0.004 |
nose | -0.003 |
carrot | -0.003 |
bacon | -0.003 |
coat | -0.003 |
Token | I |
Token ID | 314 |
Feature activation | +4.64e-01 |
Pos logit contributions | |
bisc | +0.003 |
wake | +0.002 |
count | +0.002 |
snack | +0.002 |
nose | +0.002 |
Neg logit contributions | |
spe | -0.003 |
led | -0.003 |
overed | -0.003 |
ls | -0.003 |
otte | -0.002 |
Token | 'm |
Token ID | 1101 |
Feature activation | -2.16e-01 |
Pos logit contributions | |
spe | +0.013 |
overed | +0.013 |
led | +0.013 |
ls | +0.013 |
rily | +0.013 |
Neg logit contributions | |
bisc | -0.009 |
wake | -0.008 |
count | -0.008 |
snack | -0.008 |
patch | -0.007 |
Token | the |
Token ID | 262 |
Feature activation | +3.96e-01 |
Pos logit contributions | |
come | +0.004 |
wake | +0.004 |
bisc | +0.004 |
go | +0.004 |
count | +0.004 |
Neg logit contributions | |
spe | -0.006 |
led | -0.006 |
edly | -0.006 |
rily | -0.006 |
ls | -0.006 |
Token | best |
Token ID | 1266 |
Feature activation | -1.88e-01 |
Pos logit contributions | |
led | +0.010 |
overed | +0.010 |
spe | +0.010 |
ls | +0.010 |
rily | +0.010 |
Neg logit contributions | |
bisc | -0.009 |
wake | -0.008 |
count | -0.008 |
snack | -0.007 |
patch | -0.007 |
Token | gol |
Token ID | 43791 |
Feature activation | +6.70e+00 |
Pos logit contributions | |
bisc | +0.005 |
go | +0.004 |
snack | +0.004 |
wake | +0.004 |
count | +0.004 |
Neg logit contributions | |
led | -0.004 |
otte | -0.004 |
rily | -0.004 |
Acc | -0.004 |
edly | -0.004 |
Token | fer |
Token ID | 2232 |
Feature activation | +2.06e+00 |
Pos logit contributions | |
rily | +0.123 |
spe | +0.123 |
ls | +0.121 |
led | +0.118 |
closely | +0.118 |
Neg logit contributions | |
bisc | -0.189 |
count | -0.171 |
snack | -0.167 |
come | -0.166 |
wake | -0.166 |
Token | in |
Token ID | 287 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
overed | +0.057 |
spe | +0.055 |
ivating | +0.054 |
ls | +0.054 |
rily | +0.054 |
Neg logit contributions | |
bisc | -0.041 |
wake | -0.038 |
count | -0.037 |
come | -0.037 |
snack | -0.036 |
Token | the |
Token ID | 262 |
Feature activation | -5.56e-01 |
Pos logit contributions | |
spe | +0.036 |
overed | +0.036 |
led | +0.036 |
ls | +0.035 |
rily | +0.034 |
Neg logit contributions | |
bisc | -0.044 |
wake | -0.040 |
count | -0.039 |
snack | -0.039 |
patch | -0.038 |
Token | world |
Token ID | 995 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
bisc | +0.011 |
wake | +0.010 |
count | +0.010 |
snack | +0.009 |
nose | +0.009 |
Neg logit contributions | |
spe | -0.015 |
overed | -0.015 |
led | -0.015 |
ls | -0.015 |
han | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | -3.35e-01 |
Pos logit contributions | |
overed | +0.041 |
spe | +0.041 |
led | +0.040 |
otte | +0.040 |
ls | +0.040 |
Neg logit contributions | |
bisc | -0.017 |
wake | -0.014 |
count | -0.014 |
snack | -0.014 |
come | -0.013 |
Token | ! |
Token ID | 0 |
Feature activation | -2.19e-01 |
Pos logit contributions | |
sadly | +0.018 |
overed | +0.018 |
spe | +0.017 |
ls | +0.017 |
ized | +0.017 |
Neg logit contributions | |
bisc | -0.021 |
bacon | -0.017 |
winter | -0.016 |
patch | -0.016 |
honey | -0.016 |
Token | I |
Token ID | 314 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
bisc | +0.006 |
patch | +0.005 |
wake | +0.005 |
snack | +0.005 |
count | +0.005 |
Neg logit contributions | |
spe | -0.005 |
overed | -0.005 |
ivating | -0.004 |
han | -0.004 |
otte | -0.004 |
Token | 'm |
Token ID | 1101 |
Feature activation | -5.50e-01 |
Pos logit contributions | |
spe | +0.001 |
overed | +0.001 |
led | +0.001 |
ls | +0.001 |
otte | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
wake | -0.001 |
snack | -0.001 |
count | -0.001 |
patch | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | +4.33e-01 |
Pos logit contributions | |
bisc | +0.012 |
wake | +0.011 |
count | +0.011 |
come | +0.011 |
snack | +0.011 |
Neg logit contributions | |
spe | -0.015 |
led | -0.014 |
overed | -0.014 |
ls | -0.014 |
edly | -0.014 |
Token | best |
Token ID | 1266 |
Feature activation | -2.72e-01 |
Pos logit contributions | |
spe | +0.011 |
overed | +0.010 |
ls | +0.010 |
led | +0.010 |
otte | +0.010 |
Neg logit contributions | |
bisc | -0.010 |
wake | -0.009 |
count | -0.009 |
snack | -0.009 |
patch | -0.008 |
Token | gol |
Token ID | 43791 |
Feature activation | +6.67e+00 |
Pos logit contributions | |
bisc | +0.007 |
snack | +0.006 |
count | +0.006 |
wake | +0.006 |
go | +0.006 |
Neg logit contributions | |
led | -0.007 |
otte | -0.006 |
spe | -0.006 |
rily | -0.006 |
overed | -0.006 |
Token | fer |
Token ID | 2232 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
rily | +0.125 |
spe | +0.125 |
ls | +0.122 |
closely | +0.119 |
led | +0.119 |
Neg logit contributions | |
bisc | -0.188 |
count | -0.169 |
snack | -0.164 |
come | -0.163 |
wake | -0.162 |
Token | in |
Token ID | 287 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
overed | +0.051 |
spe | +0.050 |
ivating | +0.048 |
rily | +0.048 |
ls | +0.048 |
Neg logit contributions | |
bisc | -0.045 |
wake | -0.040 |
count | -0.040 |
snack | -0.039 |
come | -0.039 |
Token | the |
Token ID | 262 |
Feature activation | -5.34e-01 |
Pos logit contributions | |
spe | +0.030 |
overed | +0.029 |
led | +0.028 |
ls | +0.028 |
otte | +0.027 |
Neg logit contributions | |
bisc | -0.042 |
wake | -0.037 |
count | -0.036 |
snack | -0.036 |
patch | -0.035 |
Token | world |
Token ID | 995 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
bisc | +0.011 |
wake | +0.009 |
count | +0.009 |
snack | +0.009 |
nose | +0.009 |
Neg logit contributions | |
spe | -0.014 |
ls | -0.014 |
led | -0.014 |
overed | -0.013 |
han | -0.013 |
Token | !" |
Token ID | 2474 |
Feature activation | +6.99e-01 |
Pos logit contributions | |
overed | +0.037 |
spe | +0.036 |
led | +0.036 |
otte | +0.036 |
ls | +0.036 |
Neg logit contributions | |
bisc | -0.016 |
wake | -0.013 |
count | -0.013 |
snack | -0.013 |
come | -0.012 |
Token | her |
Token ID | 607 |
Feature activation | -6.12e-01 |
Pos logit contributions | |
bisc | +0.050 |
wake | +0.045 |
count | +0.044 |
snack | +0.044 |
patch | +0.043 |
Neg logit contributions | |
spe | -0.025 |
overed | -0.024 |
led | -0.024 |
ls | -0.023 |
otte | -0.022 |
Token | not |
Token ID | 407 |
Feature activation | -6.69e-01 |
Pos logit contributions | |
wake | +0.014 |
bisc | +0.014 |
count | +0.014 |
snack | +0.013 |
go | +0.013 |
Neg logit contributions | |
led | -0.015 |
overed | -0.014 |
spe | -0.014 |
ivating | -0.014 |
ls | -0.014 |
Token | to |
Token ID | 284 |
Feature activation | -5.53e-01 |
Pos logit contributions | |
bisc | +0.021 |
wake | +0.018 |
snack | +0.018 |
count | +0.018 |
patch | +0.018 |
Neg logit contributions | |
spe | -0.011 |
overed | -0.010 |
ls | -0.010 |
otte | -0.010 |
led | -0.010 |
Token | investigate |
Token ID | 9161 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
bisc | +0.015 |
carrot | +0.012 |
ch | +0.012 |
est | +0.011 |
meats | +0.011 |
Neg logit contributions | |
closely | -0.011 |
sadly | -0.011 |
badly | -0.011 |
in | -0.011 |
ize | -0.011 |
Token | books |
Token ID | 3835 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
spe | +0.026 |
overed | +0.026 |
led | +0.025 |
ls | +0.025 |
otte | +0.024 |
Neg logit contributions | |
bisc | -0.026 |
wake | -0.023 |
count | -0.022 |
snack | -0.022 |
patch | -0.021 |
Token | hel |
Token ID | 2978 |
Feature activation | +6.51e+00 |
Pos logit contributions | |
spe | +0.048 |
overed | +0.046 |
led | +0.046 |
ls | +0.045 |
otte | +0.044 |
Neg logit contributions | |
bisc | -0.042 |
wake | -0.036 |
count | -0.036 |
snack | -0.035 |
patch | -0.034 |
Token | ves |
Token ID | 1158 |
Feature activation | -5.48e-01 |
Pos logit contributions | |
gol | +0.166 |
spe | +0.162 |
Acc | +0.161 |
narrative | +0.156 |
han | +0.152 |
Neg logit contributions | |
count | -0.133 |
snack | -0.128 |
wake | -0.126 |
nose | -0.124 |
chop | -0.123 |
Token | alone |
Token ID | 3436 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
bisc | +0.013 |
wake | +0.011 |
count | +0.011 |
snack | +0.011 |
patch | +0.011 |
Neg logit contributions | |
spe | -0.014 |
overed | -0.013 |
led | -0.013 |
ls | -0.013 |
otte | -0.013 |
Token | , |
Token ID | 11 |
Feature activation | +3.77e-01 |
Pos logit contributions | |
spe | +0.030 |
overed | +0.029 |
led | +0.029 |
ls | +0.028 |
rily | +0.028 |
Neg logit contributions | |
bisc | -0.024 |
wake | -0.022 |
count | -0.021 |
snack | -0.021 |
patch | -0.020 |
Token | but |
Token ID | 475 |
Feature activation | +6.83e-02 |
Pos logit contributions | |
spe | +0.011 |
overed | +0.010 |
led | +0.010 |
ls | +0.010 |
otte | +0.010 |
Neg logit contributions | |
bisc | -0.007 |
wake | -0.007 |
count | -0.006 |
snack | -0.006 |
patch | -0.006 |
Token | she |
Token ID | 673 |
Feature activation | +4.57e-01 |
Pos logit contributions | |
spe | +0.002 |
rily | +0.002 |
led | +0.002 |
overed | +0.002 |
Acc | +0.002 |
Neg logit contributions | |
bisc | -0.001 |
patch | -0.001 |
winter | -0.001 |
snack | -0.001 |
count | -0.001 |
Token | she |
Token ID | 673 |
Feature activation | +4.27e-01 |
Pos logit contributions | |
overed | +0.049 |
spe | +0.048 |
led | +0.048 |
rily | +0.047 |
ls | +0.045 |
Neg logit contributions | |
bisc | -0.059 |
wake | -0.055 |
count | -0.054 |
snack | -0.053 |
come | -0.051 |
Token | bowed |
Token ID | 35737 |
Feature activation | +3.74e-02 |
Pos logit contributions | |
spe | +0.011 |
overed | +0.011 |
rily | +0.011 |
led | +0.011 |
ls | +0.011 |
Neg logit contributions | |
bisc | -0.009 |
wake | -0.008 |
count | -0.008 |
snack | -0.008 |
patch | -0.008 |
Token | , |
Token ID | 11 |
Feature activation | -1.99e-01 |
Pos logit contributions | |
spe | +0.001 |
overed | +0.001 |
led | +0.001 |
ls | +0.001 |
otte | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
wake | -0.001 |
count | -0.001 |
snack | -0.001 |
patch | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | +5.42e-01 |
Pos logit contributions | |
bisc | +0.005 |
wake | +0.005 |
count | +0.005 |
snack | +0.005 |
come | +0.005 |
Neg logit contributions | |
overed | -0.004 |
spe | -0.004 |
led | -0.004 |
rily | -0.004 |
ls | -0.004 |
Token | town |
Token ID | 3240 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
spe | +0.015 |
overed | +0.015 |
led | +0.015 |
ls | +0.014 |
otte | +0.014 |
Neg logit contributions | |
bisc | -0.011 |
wake | -0.009 |
count | -0.009 |
snack | -0.009 |
patch | -0.009 |
Token | spe |
Token ID | 4125 |
Feature activation | +6.48e+00 |
Pos logit contributions | |
overed | +0.056 |
rily | +0.056 |
otte | +0.055 |
ivating | +0.055 |
ls | +0.054 |
Neg logit contributions | |
bisc | -0.040 |
wake | -0.038 |
count | -0.038 |
go | -0.036 |
come | -0.035 |
Token | ople |
Token ID | 643 |
Feature activation | +3.98e-01 |
Pos logit contributions | |
ls | +0.130 |
otte | +0.123 |
overed | +0.121 |
rily | +0.120 |
Acc | +0.113 |
Neg logit contributions | |
est | -0.164 |
count | -0.161 |
bisc | -0.158 |
wake | -0.151 |
go | -0.149 |
Token | stopped |
Token ID | 5025 |
Feature activation | +9.33e-01 |
Pos logit contributions | |
overed | +0.011 |
spe | +0.011 |
ls | +0.010 |
rily | +0.010 |
led | +0.010 |
Neg logit contributions | |
bisc | -0.008 |
wake | -0.008 |
count | -0.007 |
come | -0.007 |
patch | -0.007 |
Token | to |
Token ID | 284 |
Feature activation | -9.20e-01 |
Pos logit contributions | |
led | +0.023 |
spe | +0.023 |
ls | +0.023 |
overed | +0.023 |
rily | +0.023 |
Neg logit contributions | |
wake | -0.020 |
bisc | -0.020 |
come | -0.019 |
count | -0.019 |
go | -0.018 |
Token | admire |
Token ID | 21099 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
bisc | +0.016 |
snack | +0.011 |
patch | +0.010 |
pile | +0.010 |
nose | +0.010 |
Neg logit contributions | |
spe | -0.029 |
overed | -0.027 |
sadly | -0.026 |
ize | -0.026 |
ivating | -0.026 |
Token | her |
Token ID | 607 |
Feature activation | +8.91e-03 |
Pos logit contributions | |
overed | +0.033 |
spe | +0.033 |
led | +0.033 |
han | +0.032 |
ls | +0.032 |
Neg logit contributions | |
come | -0.026 |
wake | -0.025 |
count | -0.025 |
bisc | -0.025 |
go | -0.025 |
Token | noticed |
Token ID | 6810 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
led | +0.001 |
ls | +0.001 |
spe | +0.001 |
overed | +0.001 |
otte | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
wake | -0.001 |
count | -0.001 |
snack | -0.001 |
pile | -0.001 |
Token | something |
Token ID | 1223 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
spe | +0.005 |
otte | +0.005 |
overed | +0.004 |
rily | +0.004 |
led | +0.004 |
Neg logit contributions | |
bisc | -0.004 |
snack | -0.003 |
patch | -0.003 |
wake | -0.003 |
snow | -0.003 |
Token | strange |
Token ID | 6283 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
spe | +0.024 |
led | +0.024 |
overed | +0.024 |
ls | +0.024 |
rily | +0.022 |
Neg logit contributions | |
bisc | -0.026 |
wake | -0.022 |
count | -0.022 |
snack | -0.022 |
patch | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
spe | +0.030 |
led | +0.030 |
overed | +0.030 |
ls | +0.029 |
rily | +0.028 |
Neg logit contributions | |
bisc | -0.038 |
wake | -0.034 |
count | -0.033 |
snack | -0.033 |
patch | -0.032 |
Token | A |
Token ID | 317 |
Feature activation | -2.71e-01 |
Pos logit contributions | |
spe | +0.003 |
overed | +0.003 |
ls | +0.002 |
led | +0.002 |
otte | +0.002 |
Neg logit contributions | |
bisc | -0.003 |
wake | -0.003 |
count | -0.003 |
snack | -0.003 |
patch | -0.003 |
Token | gol |
Token ID | 43791 |
Feature activation | +6.47e+00 |
Pos logit contributions | |
bisc | +0.006 |
count | +0.006 |
're | +0.006 |
wake | +0.006 |
iver | +0.005 |
Neg logit contributions | |
spe | -0.006 |
ls | -0.006 |
gol | -0.005 |
ize | -0.005 |
Percy | -0.005 |
Token | fer |
Token ID | 2232 |
Feature activation | +8.38e-01 |
Pos logit contributions | |
spe | +0.119 |
rily | +0.119 |
ls | +0.117 |
Acc | +0.116 |
ivating | +0.114 |
Neg logit contributions | |
bisc | -0.183 |
count | -0.165 |
come | -0.160 |
wake | -0.158 |
paw | -0.155 |
Token | was |
Token ID | 373 |
Feature activation | +6.75e-01 |
Pos logit contributions | |
spe | +0.021 |
overed | +0.021 |
led | +0.020 |
ls | +0.020 |
otte | +0.020 |
Neg logit contributions | |
bisc | -0.019 |
wake | -0.017 |
count | -0.016 |
snack | -0.016 |
patch | -0.016 |
Token | hitting |
Token ID | 9008 |
Feature activation | -4.77e-01 |
Pos logit contributions | |
spe | +0.019 |
ls | +0.018 |
otte | +0.018 |
edly | +0.018 |
rily | +0.018 |
Neg logit contributions | |
bisc | -0.013 |
wake | -0.012 |
come | -0.012 |
count | -0.012 |
nose | -0.012 |
Token | golf |
Token ID | 13126 |
Feature activation | +9.10e-01 |
Pos logit contributions | |
bisc | +0.012 |
wake | +0.010 |
count | +0.010 |
snack | +0.010 |
patch | +0.010 |
Neg logit contributions | |
spe | -0.011 |
overed | -0.011 |
led | -0.011 |
otte | -0.010 |
ls | -0.010 |
Token | balls |
Token ID | 11333 |
Feature activation | +2.28e-02 |
Pos logit contributions | |
spe | +0.024 |
overed | +0.023 |
led | +0.023 |
otte | +0.023 |
ls | +0.022 |
Neg logit contributions | |
bisc | -0.020 |
wake | -0.017 |
count | -0.017 |
snack | -0.017 |
patch | -0.017 |
Token | on |
Token ID | 319 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
bisc | +0.018 |
wake | +0.016 |
count | +0.015 |
snack | +0.015 |
patch | +0.015 |
Neg logit contributions | |
spe | -0.021 |
overed | -0.021 |
led | -0.020 |
ls | -0.020 |
otte | -0.020 |
Token | the |
Token ID | 262 |
Feature activation | +5.25e-01 |
Pos logit contributions | |
led | +0.054 |
spe | +0.053 |
overed | +0.052 |
han | +0.051 |
ls | +0.050 |
Neg logit contributions | |
bisc | -0.043 |
snack | -0.043 |
come | -0.042 |
wake | -0.041 |
count | -0.041 |
Token | row |
Token ID | 5752 |
Feature activation | -5.72e-01 |
Pos logit contributions | |
spe | +0.014 |
overed | +0.013 |
otte | +0.013 |
ls | +0.013 |
led | +0.013 |
Neg logit contributions | |
bisc | -0.011 |
wake | -0.009 |
patch | -0.009 |
count | -0.009 |
snack | -0.009 |
Token | of |
Token ID | 286 |
Feature activation | +4.84e-01 |
Pos logit contributions | |
bisc | +0.012 |
wake | +0.011 |
count | +0.011 |
snack | +0.010 |
patch | +0.010 |
Neg logit contributions | |
spe | -0.015 |
overed | -0.015 |
led | -0.015 |
ls | -0.014 |
otte | -0.014 |
Token | books |
Token ID | 3835 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
spe | +0.011 |
overed | +0.011 |
ls | +0.010 |
led | +0.010 |
otte | +0.010 |
Neg logit contributions | |
bisc | -0.012 |
wake | -0.010 |
count | -0.010 |
patch | -0.010 |
snack | -0.010 |
Token | hel |
Token ID | 2978 |
Feature activation | +6.32e+00 |
Pos logit contributions | |
spe | +0.032 |
overed | +0.031 |
led | +0.031 |
ls | +0.030 |
otte | +0.030 |
Neg logit contributions | |
bisc | -0.051 |
wake | -0.047 |
count | -0.046 |
snack | -0.046 |
patch | -0.046 |
Token | ves |
Token ID | 1158 |
Feature activation | -5.91e-01 |
Pos logit contributions | |
Acc | +0.181 |
gol | +0.166 |
narrative | +0.162 |
scept | +0.162 |
han | +0.161 |
Neg logit contributions | |
snack | -0.117 |
count | -0.116 |
wake | -0.114 |
go | -0.113 |
come | -0.111 |
Token | . |
Token ID | 13 |
Feature activation | -1.65e-01 |
Pos logit contributions | |
bisc | +0.015 |
wake | +0.013 |
count | +0.013 |
patch | +0.013 |
snack | +0.013 |
Neg logit contributions | |
spe | -0.014 |
overed | -0.013 |
led | -0.013 |
otte | -0.013 |
ls | -0.013 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.93e-01 |
Pos logit contributions | |
bisc | +0.004 |
wake | +0.003 |
snack | +0.003 |
count | +0.003 |
patch | +0.003 |
Neg logit contributions | |
spe | -0.004 |
led | -0.004 |
ls | -0.004 |
overed | -0.004 |
rily | -0.004 |
Token | Tom |
Token ID | 13787 |
Feature activation | -8.54e-01 |
Pos logit contributions | |
bisc | +0.007 |
wake | +0.006 |
count | +0.006 |
snack | +0.006 |
patch | +0.005 |
Neg logit contributions | |
spe | -0.007 |
overed | -0.007 |
led | -0.007 |
ls | -0.007 |
otte | -0.007 |
Token | was |
Token ID | 373 |
Feature activation | -9.78e-01 |
Pos logit contributions | |
bisc | +0.014 |
snack | +0.010 |
wake | +0.010 |
count | +0.010 |
patch | +0.009 |
Neg logit contributions | |
led | -0.027 |
spe | -0.026 |
otte | -0.025 |
ls | -0.025 |
riet | -0.025 |
Token | pond |
Token ID | 16723 |
Feature activation | -8.16e-01 |
Pos logit contributions | |
bisc | +0.006 |
wake | +0.005 |
count | +0.005 |
snack | +0.005 |
patch | +0.004 |
Neg logit contributions | |
spe | -0.009 |
overed | -0.009 |
led | -0.009 |
ls | -0.009 |
otte | -0.009 |
Token | ?" |
Token ID | 1701 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
bisc | +0.021 |
wake | +0.018 |
count | +0.017 |
snack | +0.017 |
chop | +0.016 |
Neg logit contributions | |
spe | -0.018 |
overed | -0.017 |
led | -0.017 |
ls | -0.016 |
sadly | -0.016 |
Token |
|
Token ID | 198 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
led | +0.030 |
overed | +0.029 |
spe | +0.029 |
ls | +0.029 |
rily | +0.029 |
Neg logit contributions | |
bisc | -0.027 |
wake | -0.024 |
count | -0.024 |
snack | -0.023 |
chop | -0.022 |
Token |
|
Token ID | 198 |
Feature activation | -1.11e-03 |
Pos logit contributions | |
bisc | +0.005 |
wake | +0.005 |
count | +0.005 |
snack | +0.004 |
patch | +0.004 |
Neg logit contributions | |
spe | -0.006 |
overed | -0.006 |
led | -0.006 |
ls | -0.006 |
rily | -0.006 |
Token | The |
Token ID | 464 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
bisc | +0.000 |
wake | +0.000 |
count | +0.000 |
snack | +0.000 |
patch | +0.000 |
Neg logit contributions | |
overed | -0.000 |
led | -0.000 |
spe | -0.000 |
ls | -0.000 |
rily | -0.000 |
Token | gol |
Token ID | 43791 |
Feature activation | +6.25e+00 |
Pos logit contributions | |
led | +0.002 |
rily | +0.002 |
overed | +0.002 |
spe | +0.002 |
ls | +0.002 |
Neg logit contributions | |
bisc | -0.003 |
wake | -0.003 |
patch | -0.003 |
go | -0.003 |
count | -0.003 |
Token | fer |
Token ID | 2232 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
rily | +0.126 |
ls | +0.121 |
closely | +0.121 |
ivating | +0.120 |
spe | +0.119 |
Neg logit contributions | |
bisc | -0.166 |
count | -0.149 |
come | -0.144 |
snack | -0.139 |
wake | -0.137 |
Token | replied |
Token ID | 8712 |
Feature activation | -5.68e-01 |
Pos logit contributions | |
spe | +0.027 |
overed | +0.027 |
ls | +0.026 |
led | +0.026 |
otte | +0.026 |
Neg logit contributions | |
bisc | -0.025 |
wake | -0.022 |
count | -0.022 |
snack | -0.021 |
patch | -0.021 |
Token | , |
Token ID | 11 |
Feature activation | +2.38e-01 |
Pos logit contributions | |
bisc | +0.005 |
wake | +0.004 |
go | +0.004 |
come | +0.004 |
nose | +0.004 |
Neg logit contributions | |
otte | -0.022 |
spe | -0.022 |
narrative | -0.021 |
riet | -0.021 |
ivating | -0.020 |
Token | " |
Token ID | 366 |
Feature activation | +7.50e-01 |
Pos logit contributions | |
spe | +0.008 |
ls | +0.008 |
led | +0.007 |
han | +0.007 |
otte | +0.007 |
Neg logit contributions | |
bisc | -0.004 |
wake | -0.003 |
nose | -0.003 |
count | -0.003 |
come | -0.003 |
Token | I |
Token ID | 40 |
Feature activation | +4.26e-01 |
Pos logit contributions | |
rily | +0.021 |
led | +0.021 |
ls | +0.020 |
overed | +0.020 |
otte | +0.020 |
Neg logit contributions | |
bisc | -0.015 |
wake | -0.013 |
strip | -0.013 |
patch | -0.013 |
Be | -0.012 |
Token | sur |
Token ID | 969 |
Feature activation | +5.47e+00 |
Pos logit contributions | |
bisc | +0.009 |
wake | +0.007 |
count | +0.007 |
snack | +0.007 |
patch | +0.007 |
Neg logit contributions | |
spe | -0.011 |
overed | -0.011 |
led | -0.011 |
ls | -0.011 |
otte | -0.010 |
Token | fer |
Token ID | 2232 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
spe | +0.087 |
ls | +0.084 |
rily | +0.084 |
led | +0.083 |
overed | +0.083 |
Neg logit contributions | |
bisc | -0.165 |
wake | -0.156 |
count | -0.153 |
come | -0.147 |
snack | -0.144 |
Token | . |
Token ID | 13 |
Feature activation | +5.60e-02 |
Pos logit contributions | |
spe | +0.032 |
overed | +0.030 |
led | +0.030 |
ls | +0.030 |
rily | +0.029 |
Neg logit contributions | |
bisc | -0.024 |
wake | -0.019 |
snack | -0.019 |
count | -0.019 |
patch | -0.018 |
Token | Whenever |
Token ID | 21326 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
spe | +0.001 |
overed | +0.001 |
led | +0.001 |
ls | +0.001 |
otte | +0.001 |
Neg logit contributions | |
bisc | -0.002 |
wake | -0.001 |
count | -0.001 |
snack | -0.001 |
patch | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | +7.92e-01 |
Pos logit contributions | |
spe | +0.031 |
rily | +0.029 |
overed | +0.029 |
led | +0.029 |
otte | +0.027 |
Neg logit contributions | |
bisc | -0.047 |
wake | -0.045 |
count | -0.044 |
snack | -0.043 |
patch | -0.043 |
Token | sur |
Token ID | 969 |
Feature activation | +6.20e+00 |
Pos logit contributions | |
spe | +0.020 |
overed | +0.019 |
led | +0.019 |
ls | +0.019 |
otte | +0.018 |
Neg logit contributions | |
bisc | -0.019 |
wake | -0.016 |
count | -0.016 |
snack | -0.016 |
patch | -0.015 |
Token | fer |
Token ID | 2232 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
rily | +0.118 |
ls | +0.115 |
narrative | +0.115 |
Acc | +0.114 |
ivating | +0.111 |
Neg logit contributions | |
bisc | -0.154 |
wake | -0.149 |
count | -0.147 |
come | -0.141 |
chop | -0.134 |
Token | balanced |
Token ID | 12974 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
spe | +0.038 |
led | +0.036 |
overed | +0.036 |
ls | +0.035 |
otte | +0.034 |
Neg logit contributions | |
bisc | -0.036 |
wake | -0.031 |
snack | -0.030 |
count | -0.030 |
patch | -0.029 |
Token | on |
Token ID | 319 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
spe | +0.037 |
overed | +0.036 |
led | +0.035 |
ls | +0.035 |
otte | +0.034 |
Neg logit contributions | |
bisc | -0.041 |
wake | -0.036 |
count | -0.035 |
snack | -0.035 |
patch | -0.034 |
Token | the |
Token ID | 262 |
Feature activation | -3.59e-01 |
Pos logit contributions | |
spe | +0.033 |
led | +0.033 |
overed | +0.032 |
otte | +0.031 |
Acc | +0.031 |
Neg logit contributions | |
bisc | -0.031 |
count | -0.029 |
wake | -0.028 |
nose | -0.028 |
snack | -0.028 |
Token | board |
Token ID | 3096 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
bisc | +0.007 |
snack | +0.006 |
count | +0.006 |
wake | +0.006 |
patch | +0.006 |
Neg logit contributions | |
spe | -0.010 |
led | -0.010 |
overed | -0.010 |
ls | -0.010 |
rily | -0.009 |
Token | and |
Token ID | 290 |
Feature activation | -5.59e-01 |
Pos logit contributions | |
bisc | +0.032 |
pile | +0.029 |
go | +0.029 |
coat | +0.029 |
count | +0.029 |
Neg logit contributions | |
Acc | -0.029 |
otte | -0.029 |
narrative | -0.028 |
spe | -0.028 |
led | -0.027 |
Token | bumped |
Token ID | 38034 |
Feature activation | +9.44e-01 |
Pos logit contributions | |
bisc | +0.013 |
count | +0.011 |
wake | +0.011 |
snack | +0.011 |
patch | +0.010 |
Neg logit contributions | |
spe | -0.014 |
overed | -0.014 |
ls | -0.014 |
led | -0.014 |
rily | -0.014 |
Token | against |
Token ID | 1028 |
Feature activation | +8.33e-01 |
Pos logit contributions | |
narrative | +0.025 |
han | +0.023 |
spe | +0.022 |
otte | +0.021 |
ls | +0.020 |
Neg logit contributions | |
bisc | -0.021 |
snow | -0.021 |
nose | -0.020 |
paw | -0.020 |
wake | -0.020 |
Token | the |
Token ID | 262 |
Feature activation | -8.52e-01 |
Pos logit contributions | |
spe | +0.018 |
overed | +0.017 |
led | +0.017 |
ls | +0.017 |
otte | +0.016 |
Neg logit contributions | |
bisc | -0.023 |
wake | -0.020 |
count | -0.020 |
snack | -0.020 |
patch | -0.019 |
Token | books |
Token ID | 3835 |
Feature activation | +5.76e-01 |
Pos logit contributions | |
bisc | +0.018 |
count | +0.016 |
're | +0.015 |
wake | +0.015 |
feed | +0.015 |
Neg logit contributions | |
spe | -0.020 |
ls | -0.019 |
led | -0.019 |
overed | -0.018 |
ize | -0.018 |
Token | hel |
Token ID | 2978 |
Feature activation | +6.19e+00 |
Pos logit contributions | |
spe | +0.013 |
overed | +0.012 |
led | +0.012 |
ls | +0.012 |
otte | +0.012 |
Neg logit contributions | |
bisc | -0.015 |
wake | -0.014 |
count | -0.013 |
snack | -0.013 |
patch | -0.013 |
Token | ves |
Token ID | 1158 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
Acc | +0.167 |
narrative | +0.159 |
han | +0.152 |
ivating | +0.150 |
gol | +0.148 |
Neg logit contributions | |
snack | -0.126 |
go | -0.122 |
tip | -0.122 |
wake | -0.120 |
est | -0.119 |
Token | . |
Token ID | 13 |
Feature activation | -1.86e-01 |
Pos logit contributions | |
bisc | +0.038 |
wake | +0.034 |
count | +0.033 |
snack | +0.033 |
patch | +0.032 |
Neg logit contributions | |
spe | -0.035 |
overed | -0.034 |
led | -0.034 |
ls | -0.033 |
otte | -0.033 |
Token | He |
Token ID | 679 |
Feature activation | +7.95e-01 |
Pos logit contributions | |
bisc | +0.005 |
wake | +0.005 |
count | +0.004 |
snack | +0.004 |
patch | +0.004 |
Neg logit contributions | |
spe | -0.004 |
overed | -0.004 |
led | -0.004 |
ls | -0.004 |
rily | -0.004 |
Token | smiled |
Token ID | 13541 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
overed | +0.021 |
spe | +0.021 |
han | +0.020 |
ls | +0.020 |
rily | +0.020 |
Neg logit contributions | |
bisc | -0.017 |
wake | -0.015 |
count | -0.015 |
snack | -0.014 |
patch | -0.014 |
Token | as |
Token ID | 355 |
Feature activation | -5.12e-02 |
Pos logit contributions | |
spe | +0.031 |
ls | +0.030 |
otte | +0.030 |
han | +0.030 |
overed | +0.029 |
Neg logit contributions | |
bisc | -0.028 |
wake | -0.026 |
count | -0.025 |
patch | -0.024 |
nose | -0.024 |
Token | spotted |
Token ID | 13489 |
Feature activation | -1.89e-01 |
Pos logit contributions | |
bisc | +0.009 |
wake | +0.009 |
count | +0.008 |
patch | +0.008 |
come | +0.008 |
Neg logit contributions | |
overed | -0.015 |
spe | -0.015 |
rily | -0.014 |
led | -0.014 |
ls | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | +2.16e-01 |
Pos logit contributions | |
bisc | +0.004 |
wake | +0.003 |
snack | +0.003 |
count | +0.003 |
patch | +0.003 |
Neg logit contributions | |
spe | -0.005 |
overed | -0.005 |
led | -0.005 |
ls | -0.005 |
otte | -0.005 |
Token | row |
Token ID | 5752 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
spe | +0.006 |
ls | +0.005 |
overed | +0.005 |
otte | +0.005 |
ize | +0.005 |
Neg logit contributions | |
bisc | -0.005 |
count | -0.004 |
wake | -0.004 |
're | -0.004 |
snack | -0.004 |
Token | of |
Token ID | 286 |
Feature activation | +2.76e-01 |
Pos logit contributions | |
bisc | +0.026 |
wake | +0.024 |
patch | +0.023 |
pile | +0.022 |
come | +0.022 |
Neg logit contributions | |
otte | -0.045 |
ls | -0.045 |
Acc | -0.045 |
led | -0.044 |
overed | -0.044 |
Token | books |
Token ID | 3835 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
spe | +0.008 |
overed | +0.008 |
led | +0.007 |
ls | +0.007 |
otte | +0.007 |
Neg logit contributions | |
bisc | -0.006 |
wake | -0.005 |
count | -0.005 |
snack | -0.005 |
patch | -0.004 |
Token | hel |
Token ID | 2978 |
Feature activation | +6.19e+00 |
Pos logit contributions | |
spe | +0.024 |
otte | +0.022 |
ls | +0.022 |
led | +0.021 |
overed | +0.021 |
Neg logit contributions | |
bisc | -0.037 |
wake | -0.035 |
count | -0.034 |
patch | -0.034 |
come | -0.034 |
Token | ves |
Token ID | 1158 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
Acc | +0.189 |
gol | +0.184 |
narrative | +0.178 |
scept | +0.177 |
Bruce | +0.169 |
Neg logit contributions | |
go | -0.094 |
count | -0.093 |
snack | -0.093 |
est | -0.089 |
come | -0.089 |
Token | in |
Token ID | 287 |
Feature activation | +8.34e-01 |
Pos logit contributions | |
bisc | +0.034 |
wake | +0.031 |
count | +0.030 |
patch | +0.030 |
snack | +0.030 |
Neg logit contributions | |
spe | -0.031 |
overed | -0.029 |
otte | -0.029 |
ls | -0.029 |
led | -0.029 |
Token | her |
Token ID | 607 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
led | +0.020 |
overed | +0.020 |
spe | +0.019 |
ls | +0.019 |
Acc | +0.018 |
Neg logit contributions | |
bisc | -0.021 |
wake | -0.019 |
count | -0.018 |
snack | -0.018 |
strip | -0.018 |
Token | grandparents |
Token ID | 28571 |
Feature activation | -7.21e-01 |
Pos logit contributions | |
bisc | +0.003 |
wake | +0.002 |
count | +0.002 |
snack | +0.002 |
patch | +0.002 |
Neg logit contributions | |
spe | -0.004 |
overed | -0.004 |
led | -0.004 |
ls | -0.004 |
otte | -0.004 |
Token | ' |
Token ID | 6 |
Feature activation | -4.13e-01 |
Pos logit contributions | |
bisc | +0.016 |
wake | +0.013 |
snack | +0.013 |
count | +0.013 |
patch | +0.013 |
Neg logit contributions | |
spe | -0.019 |
led | -0.018 |
overed | -0.018 |
otte | -0.017 |
ls | -0.017 |
Token | B |
Token ID | 33 |
Feature activation | +3.76e-01 |
Pos logit contributions | |
spe | +0.002 |
overed | +0.001 |
led | +0.001 |
ls | +0.001 |
otte | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
wake | -0.001 |
snack | -0.001 |
count | -0.001 |
patch | -0.001 |
Token | ella |
Token ID | 12627 |
Feature activation | +3.21e-01 |
Pos logit contributions | |
spe | +0.008 |
overed | +0.008 |
led | +0.008 |
ls | +0.008 |
rily | +0.008 |
Neg logit contributions | |
bisc | -0.010 |
wake | -0.009 |
count | -0.009 |
snack | -0.008 |
patch | -0.008 |
Token | watched |
Token ID | 7342 |
Feature activation | -6.30e-01 |
Pos logit contributions | |
narrative | +0.009 |
rily | +0.009 |
otte | +0.009 |
ivating | +0.009 |
ls | +0.009 |
Neg logit contributions | |
bisc | -0.006 |
count | -0.006 |
wake | -0.005 |
nose | -0.005 |
patch | -0.005 |
Token | as |
Token ID | 355 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
wake | +0.013 |
go | +0.013 |
come | +0.012 |
count | +0.012 |
snack | +0.012 |
Neg logit contributions | |
spe | -0.017 |
narrative | -0.017 |
led | -0.016 |
ls | -0.016 |
otte | -0.016 |
Token | the |
Token ID | 262 |
Feature activation | +5.70e-01 |
Pos logit contributions | |
Cat | +0.004 |
go | +0.004 |
m | +0.004 |
hop | +0.004 |
jump | +0.004 |
Neg logit contributions | |
han | -0.008 |
led | -0.008 |
edly | -0.008 |
spe | -0.008 |
nes | -0.008 |
Token | sur |
Token ID | 969 |
Feature activation | +6.07e+00 |
Pos logit contributions | |
spe | +0.015 |
overed | +0.015 |
led | +0.015 |
ls | +0.015 |
otte | +0.014 |
Neg logit contributions | |
bisc | -0.012 |
wake | -0.010 |
count | -0.010 |
snack | -0.010 |
patch | -0.010 |
Token | fer |
Token ID | 2232 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
narrative | +0.121 |
rily | +0.116 |
ls | +0.116 |
Acc | +0.111 |
ivating | +0.111 |
Neg logit contributions | |
bisc | -0.152 |
wake | -0.145 |
count | -0.145 |
come | -0.138 |
chop | -0.132 |
Token | rode |
Token ID | 22075 |
Feature activation | +6.27e-01 |
Pos logit contributions | |
spe | +0.030 |
led | +0.029 |
overed | +0.028 |
ls | +0.028 |
sadly | +0.026 |
Neg logit contributions | |
bisc | -0.028 |
snack | -0.023 |
wake | -0.023 |
patch | -0.022 |
count | -0.021 |
Token | on |
Token ID | 319 |
Feature activation | +6.03e-01 |
Pos logit contributions | |
Acc | +0.018 |
otte | +0.017 |
overed | +0.016 |
spe | +0.015 |
itt | +0.015 |
Neg logit contributions | |
bisc | -0.013 |
go | -0.013 |
come | -0.013 |
count | -0.012 |
coat | -0.012 |
Token | the |
Token ID | 262 |
Feature activation | -4.64e-01 |
Pos logit contributions | |
led | +0.018 |
spe | +0.016 |
overed | +0.016 |
otte | +0.016 |
Acc | +0.016 |
Neg logit contributions | |
bisc | -0.012 |
count | -0.012 |
snack | -0.011 |
snow | -0.011 |
come | -0.011 |
Token | waves |
Token ID | 9813 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
bisc | +0.010 |
snack | +0.008 |
wake | +0.008 |
count | +0.008 |
patch | +0.008 |
Neg logit contributions | |
spe | -0.012 |
led | -0.012 |
overed | -0.012 |
ls | -0.012 |
otte | -0.011 |
Token | happy |
Token ID | 3772 |
Feature activation | +6.65e-01 |
Pos logit contributions | |
ls | +0.002 |
overed | +0.002 |
led | +0.002 |
otte | +0.002 |
spe | +0.002 |
Neg logit contributions | |
bisc | -0.002 |
wake | -0.002 |
count | -0.002 |
come | -0.002 |
patch | -0.002 |
Token | and |
Token ID | 290 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
otte | +0.019 |
led | +0.019 |
ivating | +0.019 |
ls | +0.019 |
han | +0.019 |
Neg logit contributions | |
come | -0.012 |
wake | -0.011 |
bisc | -0.011 |
count | -0.011 |
snack | -0.010 |
Token | the |
Token ID | 262 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
overed | +0.022 |
ls | +0.022 |
rily | +0.022 |
ivating | +0.021 |
spe | +0.021 |
Neg logit contributions | |
bisc | -0.026 |
wake | -0.024 |
count | -0.024 |
come | -0.024 |
patch | -0.023 |
Token | nation |
Token ID | 3277 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
rily | +0.031 |
otte | +0.030 |
Acc | +0.030 |
overed | +0.030 |
led | +0.029 |
Neg logit contributions | |
count | -0.021 |
wake | -0.021 |
go | -0.020 |
welcome | -0.020 |
come | -0.020 |
Token | pro |
Token ID | 386 |
Feature activation | +3.19e+00 |
Pos logit contributions | |
otte | +0.042 |
ivating | +0.041 |
rily | +0.041 |
overed | +0.041 |
ls | +0.041 |
Neg logit contributions | |
come | -0.027 |
count | -0.027 |
wake | -0.026 |
bisc | -0.026 |
go | -0.026 |
Token | spe |
Token ID | 4125 |
Feature activation | +6.07e+00 |
Pos logit contributions | |
rily | +0.061 |
spe | +0.061 |
han | +0.061 |
otte | +0.060 |
narrative | +0.060 |
Neg logit contributions | |
bisc | -0.075 |
come | -0.074 |
're | -0.072 |
go | -0.070 |
wake | -0.069 |
Token | red |
Token ID | 445 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
ls | +0.123 |
otte | +0.121 |
rily | +0.119 |
Acc | +0.116 |
overed | +0.114 |
Neg logit contributions | |
est | -0.153 |
count | -0.144 |
bisc | -0.138 |
go | -0.135 |
're | -0.134 |
Token | . |
Token ID | 13 |
Feature activation | -5.57e-01 |
Pos logit contributions | |
otte | +0.003 |
overed | +0.003 |
han | +0.003 |
spe | +0.003 |
ls | +0.003 |
Neg logit contributions | |
bisc | -0.003 |
wake | -0.002 |
count | -0.002 |
snack | -0.002 |
patch | -0.002 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -4.66e-01 |
Pos logit contributions | |
bisc | +0.010 |
're | +0.009 |
count | +0.009 |
snack | +0.008 |
wake | +0.008 |
Neg logit contributions | |
led | -0.017 |
ls | -0.017 |
overed | -0.016 |
ized | -0.015 |
ize | -0.015 |
Token | Once |
Token ID | 7454 |
Feature activation | +2.35e+00 |
Pos logit contributions | |
bisc | +0.010 |
wake | +0.008 |
snack | +0.008 |
count | +0.008 |
patch | +0.008 |
Neg logit contributions | |
spe | -0.013 |
overed | -0.012 |
led | -0.012 |
ls | -0.012 |
otte | -0.012 |
Token | upon |
Token ID | 2402 |
Feature activation | -3.96e-01 |
Pos logit contributions | |
edly | +0.082 |
ivating | +0.075 |
ize | +0.074 |
overed | +0.073 |
rily | +0.073 |
Neg logit contributions | |
go | -0.028 |
come | -0.028 |
pile | -0.028 |
rest | -0.028 |
're | -0.028 |
Token | walked |
Token ID | 6807 |
Feature activation | -3.38e-01 |
Pos logit contributions | |
spe | +0.015 |
overed | +0.014 |
led | +0.014 |
ls | +0.014 |
otte | +0.014 |
Neg logit contributions | |
bisc | -0.014 |
wake | -0.012 |
snack | -0.011 |
count | -0.011 |
patch | -0.011 |
Token | over |
Token ID | 625 |
Feature activation | +2.54e-01 |
Pos logit contributions | |
bisc | +0.009 |
count | +0.008 |
wake | +0.008 |
patch | +0.008 |
snack | +0.008 |
Neg logit contributions | |
spe | -0.007 |
narrative | -0.007 |
Acc | -0.007 |
ls | -0.007 |
led | -0.007 |
Token | to |
Token ID | 284 |
Feature activation | -8.65e-01 |
Pos logit contributions | |
spe | +0.006 |
overed | +0.006 |
led | +0.006 |
ls | +0.005 |
rily | +0.005 |
Neg logit contributions | |
bisc | -0.006 |
wake | -0.006 |
count | -0.006 |
snack | -0.006 |
patch | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -5.24e-01 |
Pos logit contributions | |
bisc | +0.029 |
patch | +0.024 |
snack | +0.023 |
wake | +0.023 |
nose | +0.023 |
Neg logit contributions | |
otte | -0.013 |
overed | -0.013 |
spe | -0.012 |
ize | -0.012 |
ls | -0.012 |
Token | books |
Token ID | 3835 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
bisc | +0.013 |
chop | +0.012 |
iver | +0.012 |
carrot | +0.011 |
feed | +0.011 |
Neg logit contributions | |
May | -0.010 |
ls | -0.010 |
issa | -0.010 |
otte | -0.010 |
spe | -0.009 |
Token | hel |
Token ID | 2978 |
Feature activation | +6.06e+00 |
Pos logit contributions | |
spe | +0.020 |
overed | +0.020 |
led | +0.019 |
ls | +0.019 |
otte | +0.018 |
Neg logit contributions | |
bisc | -0.030 |
wake | -0.027 |
count | -0.026 |
snack | -0.026 |
patch | -0.025 |
Token | ves |
Token ID | 1158 |
Feature activation | -8.60e-01 |
Pos logit contributions | |
gol | +0.173 |
Acc | +0.172 |
narrative | +0.167 |
ivating | +0.161 |
Bruce | +0.160 |
Neg logit contributions | |
go | -0.104 |
snack | -0.104 |
wake | -0.104 |
count | -0.102 |
strip | -0.101 |
Token | and |
Token ID | 290 |
Feature activation | -8.27e-01 |
Pos logit contributions | |
bisc | +0.023 |
wake | +0.020 |
count | +0.020 |
snack | +0.020 |
patch | +0.019 |
Neg logit contributions | |
spe | -0.018 |
overed | -0.018 |
led | -0.018 |
ls | -0.018 |
rily | -0.017 |
Token | began |
Token ID | 2540 |
Feature activation | -3.08e-01 |
Pos logit contributions | |
bisc | +0.016 |
wake | +0.014 |
count | +0.014 |
snack | +0.013 |
patch | +0.013 |
Neg logit contributions | |
spe | -0.024 |
overed | -0.023 |
led | -0.023 |
ls | -0.023 |
rily | -0.022 |
Token | to |
Token ID | 284 |
Feature activation | -3.59e-01 |
Pos logit contributions | |
bisc | +0.008 |
wake | +0.007 |
count | +0.007 |
snack | +0.007 |
patch | +0.007 |
Neg logit contributions | |
spe | -0.007 |
overed | -0.006 |
led | -0.006 |
ls | -0.006 |
otte | -0.006 |
Token | read |
Token ID | 1100 |
Feature activation | +3.64e-01 |
Pos logit contributions | |
bisc | +0.007 |
itten | +0.005 |
ch | +0.005 |
est | +0.005 |
umbs | +0.005 |
Neg logit contributions | |
spe | -0.010 |
closely | -0.009 |
nam | -0.009 |
proudly | -0.009 |
sadly | -0.009 |
Token | he |
Token ID | 339 |
Feature activation | +5.80e-01 |
Pos logit contributions | |
led | +0.018 |
overed | +0.018 |
spe | +0.018 |
ivating | +0.017 |
ls | +0.017 |
Neg logit contributions | |
bisc | -0.018 |
patch | -0.018 |
come | -0.018 |
coat | -0.018 |
snack | -0.017 |
Token | could |
Token ID | 714 |
Feature activation | +3.26e-01 |
Pos logit contributions | |
overed | +0.015 |
spe | +0.015 |
ls | +0.015 |
rily | +0.015 |
otte | +0.015 |
Neg logit contributions | |
bisc | -0.012 |
wake | -0.011 |
count | -0.011 |
come | -0.011 |
snack | -0.011 |
Token | do |
Token ID | 466 |
Feature activation | +3.16e+00 |
Pos logit contributions | |
spe | +0.011 |
overed | +0.010 |
led | +0.010 |
ls | +0.010 |
otte | +0.010 |
Neg logit contributions | |
bisc | -0.005 |
wake | -0.004 |
count | -0.004 |
snack | -0.004 |
patch | -0.004 |
Token | anything |
Token ID | 1997 |
Feature activation | +1.89e+00 |
Pos logit contributions | |
spe | +0.087 |
overed | +0.086 |
otte | +0.084 |
han | +0.084 |
ls | +0.083 |
Neg logit contributions | |
count | -0.060 |
bisc | -0.059 |
snack | -0.058 |
wake | -0.057 |
come | -0.057 |
Token | if |
Token ID | 611 |
Feature activation | -5.36e-01 |
Pos logit contributions | |
overed | +0.050 |
otte | +0.048 |
han | +0.047 |
spe | +0.047 |
rily | +0.046 |
Neg logit contributions | |
bisc | -0.039 |
wake | -0.038 |
count | -0.038 |
come | -0.037 |
snack | -0.036 |
Token | he |
Token ID | 258 |
Feature activation | +6.03e+00 |
Pos logit contributions | |
bisc | +0.014 |
snack | +0.013 |
wake | +0.013 |
count | +0.013 |
patch | +0.012 |
Neg logit contributions | |
overed | -0.012 |
spe | -0.011 |
led | -0.011 |
ivating | -0.011 |
rily | -0.011 |
Token | set |
Token ID | 900 |
Feature activation | +5.65e-01 |
Pos logit contributions | |
spe | +0.149 |
overed | +0.148 |
ivating | +0.147 |
ls | +0.146 |
Acc | +0.142 |
Neg logit contributions | |
bisc | -0.137 |
snack | -0.126 |
wake | -0.125 |
patch | -0.122 |
nose | -0.121 |
Token | his |
Token ID | 465 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
spe | +0.013 |
overed | +0.013 |
led | +0.012 |
otte | +0.012 |
Acc | +0.012 |
Neg logit contributions | |
bisc | -0.014 |
wake | -0.013 |
snack | -0.013 |
count | -0.013 |
come | -0.012 |
Token | mind |
Token ID | 2000 |
Feature activation | +9.58e-01 |
Pos logit contributions | |
spe | +0.034 |
overed | +0.031 |
ls | +0.030 |
sadly | +0.030 |
led | +0.030 |
Neg logit contributions | |
bisc | -0.026 |
wake | -0.021 |
're | -0.020 |
patch | -0.020 |
count | -0.019 |
Token | to |
Token ID | 284 |
Feature activation | -3.24e-01 |
Pos logit contributions | |
overed | +0.022 |
spe | +0.022 |
led | +0.022 |
ls | +0.021 |
otte | +0.021 |
Neg logit contributions | |
bisc | -0.023 |
wake | -0.021 |
count | -0.020 |
snack | -0.020 |
patch | -0.020 |
Token | it |
Token ID | 340 |
Feature activation | -3.86e-01 |
Pos logit contributions | |
bisc | +0.006 |
wake | +0.004 |
snack | +0.004 |
patch | +0.004 |
count | +0.004 |
Neg logit contributions | |
spe | -0.010 |
overed | -0.009 |
led | -0.009 |
otte | -0.009 |
ls | -0.009 |
Token | rope |
Token ID | 17182 |
Feature activation | +8.04e-01 |
Pos logit contributions | |
spe | +0.006 |
overed | +0.005 |
ls | +0.005 |
ized | +0.005 |
ivating | +0.005 |
Neg logit contributions | |
bisc | -0.004 |
count | -0.003 |
wake | -0.003 |
snack | -0.003 |
chop | -0.003 |
Token | bouncing |
Token ID | 30713 |
Feature activation | +4.82e-02 |
Pos logit contributions | |
spe | +0.020 |
overed | +0.020 |
led | +0.020 |
ls | +0.019 |
otte | +0.019 |
Neg logit contributions | |
bisc | -0.018 |
wake | -0.016 |
count | -0.016 |
snack | -0.015 |
patch | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | -7.00e-02 |
Pos logit contributions | |
otte | +0.002 |
riet | +0.001 |
issa | +0.001 |
han | +0.001 |
spe | +0.001 |
Neg logit contributions | |
go | -0.001 |
come | -0.001 |
bisc | -0.001 |
patch | -0.001 |
pile | -0.001 |
Token | The |
Token ID | 383 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
bisc | +0.002 |
wake | +0.001 |
snack | +0.001 |
count | +0.001 |
patch | +0.001 |
Neg logit contributions | |
spe | -0.002 |
led | -0.002 |
ls | -0.002 |
otte | -0.002 |
rily | -0.002 |
Token | town |
Token ID | 3240 |
Feature activation | +2.40e+00 |
Pos logit contributions | |
bisc | +0.006 |
count | +0.005 |
wake | +0.005 |
snack | +0.004 |
're | +0.004 |
Neg logit contributions | |
overed | -0.006 |
ls | -0.006 |
spe | -0.006 |
ized | -0.006 |
led | -0.006 |
Token | spe |
Token ID | 4125 |
Feature activation | +5.91e+00 |
Pos logit contributions | |
otte | +0.071 |
Acc | +0.070 |
rily | +0.069 |
riet | +0.068 |
ivating | +0.068 |
Neg logit contributions | |
bisc | -0.038 |
pile | -0.037 |
patch | -0.037 |
go | -0.037 |
wake | -0.036 |
Token | ople |
Token ID | 643 |
Feature activation | +8.91e-01 |
Pos logit contributions | |
spe | +0.096 |
overed | +0.096 |
ls | +0.095 |
rily | +0.095 |
otte | +0.094 |
Neg logit contributions | |
bisc | -0.175 |
count | -0.160 |
wake | -0.159 |
est | -0.156 |
strip | -0.155 |
Token | gathered |
Token ID | 9272 |
Feature activation | +4.02e-01 |
Pos logit contributions | |
overed | +0.025 |
spe | +0.025 |
han | +0.024 |
rily | +0.024 |
ls | +0.024 |
Neg logit contributions | |
bisc | -0.016 |
wake | -0.016 |
count | -0.015 |
come | -0.015 |
patch | -0.014 |
Token | around |
Token ID | 1088 |
Feature activation | +7.10e-01 |
Pos logit contributions | |
otte | +0.009 |
spe | +0.009 |
issa | +0.008 |
Acc | +0.008 |
overed | +0.008 |
Neg logit contributions | |
bisc | -0.010 |
go | -0.010 |
snack | -0.009 |
snow | -0.009 |
patch | -0.009 |
Token | to |
Token ID | 284 |
Feature activation | -8.92e-01 |
Pos logit contributions | |
spe | +0.018 |
led | +0.017 |
otte | +0.016 |
ls | +0.016 |
rily | +0.016 |
Neg logit contributions | |
bisc | -0.017 |
wake | -0.015 |
count | -0.015 |
snack | -0.015 |
patch | -0.015 |
Token | watch |
Token ID | 2342 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
bisc | +0.011 |
wake | +0.008 |
count | +0.008 |
snack | +0.008 |
patch | +0.007 |
Neg logit contributions | |
spe | -0.032 |
led | -0.032 |
overed | -0.032 |
ls | -0.031 |
otte | -0.031 |
Token | Tom |
Token ID | 13787 |
Feature activation | +2.83e-01 |
Pos logit contributions | |
spe | +0.004 |
overed | +0.004 |
led | +0.004 |
ls | +0.004 |
otte | +0.003 |
Neg logit contributions | |
bisc | -0.003 |
wake | -0.003 |
count | -0.003 |
snack | -0.003 |
patch | -0.003 |
Token | called |
Token ID | 1444 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
spe | +0.008 |
overed | +0.008 |
ls | +0.007 |
led | +0.007 |
rily | +0.007 |
Neg logit contributions | |
bisc | -0.006 |
wake | -0.005 |
count | -0.005 |
patch | -0.005 |
snack | -0.005 |
Token | out |
Token ID | 503 |
Feature activation | -7.29e-01 |
Pos logit contributions | |
bisc | +0.052 |
wake | +0.045 |
count | +0.044 |
snack | +0.044 |
patch | +0.043 |
Neg logit contributions | |
spe | -0.047 |
overed | -0.046 |
led | -0.045 |
ls | -0.044 |
otte | -0.043 |
Token | to |
Token ID | 284 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
bisc | +0.017 |
wake | +0.015 |
count | +0.015 |
snack | +0.015 |
patch | +0.014 |
Neg logit contributions | |
spe | -0.018 |
led | -0.017 |
overed | -0.017 |
ls | -0.017 |
otte | -0.017 |
Token | the |
Token ID | 262 |
Feature activation | -7.63e-01 |
Pos logit contributions | |
bisc | +0.032 |
patch | +0.027 |
wake | +0.027 |
snack | +0.027 |
count | +0.027 |
Neg logit contributions | |
spe | -0.018 |
overed | -0.017 |
led | -0.016 |
otte | -0.016 |
ls | -0.016 |
Token | gol |
Token ID | 43791 |
Feature activation | +5.82e+00 |
Pos logit contributions | |
bisc | +0.017 |
wake | +0.015 |
patch | +0.015 |
snack | +0.015 |
count | +0.015 |
Neg logit contributions | |
led | -0.020 |
spe | -0.020 |
rily | -0.020 |
overed | -0.020 |
ls | -0.019 |
Token | fer |
Token ID | 2232 |
Feature activation | -2.25e-01 |
Pos logit contributions | |
rily | +0.112 |
spe | +0.112 |
ivating | +0.109 |
Acc | +0.108 |
ls | +0.107 |
Neg logit contributions | |
bisc | -0.160 |
count | -0.143 |
come | -0.140 |
snack | -0.136 |
wake | -0.136 |
Token | " |
Token ID | 366 |
Feature activation | +8.29e-01 |
Pos logit contributions | |
bisc | +0.005 |
wake | +0.004 |
snack | +0.004 |
count | +0.004 |
patch | +0.004 |
Neg logit contributions | |
spe | -0.006 |
overed | -0.006 |
led | -0.006 |
ls | -0.006 |
otte | -0.006 |
Token | Hey |
Token ID | 10814 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
spe | +0.017 |
led | +0.016 |
otte | +0.015 |
Acc | +0.015 |
overed | +0.015 |
Neg logit contributions | |
bisc | -0.022 |
count | -0.020 |
nose | -0.020 |
wake | -0.020 |
pile | -0.019 |
Token | , |
Token ID | 11 |
Feature activation | -5.37e-01 |
Pos logit contributions | |
spe | +0.026 |
ls | +0.025 |
overed | +0.025 |
led | +0.024 |
ize | +0.024 |
Neg logit contributions | |
bisc | -0.024 |
wake | -0.021 |
count | -0.021 |
patch | -0.020 |
snack | -0.020 |
Token | why |
Token ID | 1521 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
bisc | +0.012 |
wake | +0.011 |
count | +0.010 |
snack | +0.010 |
come | +0.010 |
Neg logit contributions | |
led | -0.014 |
spe | -0.014 |
overed | -0.014 |
ls | -0.014 |
rily | -0.013 |
Token | ! |
Token ID | 0 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
spe | +0.058 |
overed | +0.058 |
otte | +0.057 |
han | +0.056 |
Acc | +0.056 |
Neg logit contributions | |
bisc | -0.036 |
patch | -0.033 |
snack | -0.032 |
wake | -0.031 |
pile | -0.031 |
Token |
|
Token ID | 198 |
Feature activation | -8.67e-01 |
Pos logit contributions | |
bisc | +0.002 |
snack | +0.002 |
wake | +0.002 |
paw | +0.002 |
count | +0.002 |
Neg logit contributions | |
led | -0.003 |
spe | -0.003 |
ls | -0.003 |
overed | -0.003 |
ize | -0.003 |
Token |
|
Token ID | 198 |
Feature activation | -6.56e-01 |
Pos logit contributions | |
bisc | +0.019 |
wake | +0.017 |
count | +0.016 |
snack | +0.016 |
patch | +0.016 |
Neg logit contributions | |
spe | -0.023 |
led | -0.022 |
overed | -0.022 |
ls | -0.021 |
rily | -0.021 |
Token | The |
Token ID | 464 |
Feature activation | -6.97e-01 |
Pos logit contributions | |
bisc | +0.016 |
wake | +0.014 |
count | +0.013 |
snack | +0.013 |
patch | +0.013 |
Neg logit contributions | |
spe | -0.016 |
led | -0.016 |
overed | -0.016 |
ls | -0.015 |
rily | -0.015 |
Token | town |
Token ID | 3240 |
Feature activation | +2.01e+00 |
Pos logit contributions | |
bisc | +0.013 |
wake | +0.011 |
snack | +0.011 |
count | +0.011 |
patch | +0.011 |
Neg logit contributions | |
spe | -0.020 |
overed | -0.020 |
led | -0.020 |
ls | -0.020 |
otte | -0.019 |
Token | spe |
Token ID | 4125 |
Feature activation | +5.65e+00 |
Pos logit contributions | |
otte | +0.056 |
Acc | +0.056 |
ivating | +0.054 |
riet | +0.054 |
rily | +0.053 |
Neg logit contributions | |
bisc | -0.036 |
patch | -0.034 |
pile | -0.034 |
wake | -0.033 |
go | -0.033 |
Token | ople |
Token ID | 643 |
Feature activation | +5.46e-01 |
Pos logit contributions | |
ls | +0.107 |
rily | +0.107 |
otte | +0.106 |
overed | +0.104 |
Acc | +0.098 |
Neg logit contributions | |
bisc | -0.146 |
est | -0.141 |
count | -0.136 |
wake | -0.130 |
pile | -0.129 |
Token | were |
Token ID | 547 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
spe | +0.015 |
overed | +0.014 |
ls | +0.014 |
led | +0.014 |
otte | +0.014 |
Neg logit contributions | |
bisc | -0.011 |
wake | -0.010 |
count | -0.010 |
patch | -0.009 |
snack | -0.009 |
Token | sad |
Token ID | 6507 |
Feature activation | -6.32e-01 |
Pos logit contributions | |
bisc | +0.005 |
wake | +0.005 |
come | +0.004 |
count | +0.004 |
snack | +0.004 |
Neg logit contributions | |
spe | -0.005 |
led | -0.005 |
ls | -0.005 |
otte | -0.005 |
overed | -0.005 |
Token | and |
Token ID | 290 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
bisc | +0.011 |
wake | +0.010 |
snack | +0.010 |
come | +0.009 |
count | +0.009 |
Neg logit contributions | |
spe | -0.019 |
otte | -0.019 |
led | -0.018 |
overed | -0.018 |
han | -0.018 |
Token | scared |
Token ID | 12008 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
overed | +0.051 |
rily | +0.050 |
ize | +0.049 |
otte | +0.049 |
han | +0.049 |
Neg logit contributions | |
wake | -0.036 |
come | -0.036 |
go | -0.035 |
bisc | -0.033 |
nose | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | +1.76e-01 |
Pos logit contributions | |
spe | +0.059 |
overed | +0.057 |
led | +0.057 |
ls | +0.056 |
otte | +0.054 |
Neg logit contributions | |
bisc | -0.063 |
wake | -0.054 |
count | -0.053 |
snack | -0.053 |
patch | -0.052 |
Token |
|
Token ID | 198 |
Feature activation | -2.51e-01 |
Pos logit contributions | |
spe | +0.005 |
led | +0.004 |
ls | +0.004 |
overed | +0.004 |
otte | +0.004 |
Neg logit contributions | |
bisc | -0.004 |
wake | -0.003 |
count | -0.003 |
snack | -0.003 |
patch | -0.003 |
Token |
|
Token ID | 198 |
Feature activation | +2.01e-02 |
Pos logit contributions | |
bisc | +0.006 |
wake | +0.005 |
count | +0.005 |
snack | +0.005 |
patch | +0.005 |
Neg logit contributions | |
spe | -0.006 |
overed | -0.006 |
led | -0.006 |
ls | -0.006 |
otte | -0.006 |
Token | The |
Token ID | 464 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
led | +0.001 |
spe | +0.001 |
overed | +0.001 |
rily | +0.001 |
ls | +0.001 |
Neg logit contributions | |
bisc | -0.000 |
wake | -0.000 |
count | -0.000 |
patch | -0.000 |
snack | -0.000 |
Token | town |
Token ID | 3240 |
Feature activation | +2.45e+00 |
Pos logit contributions | |
bisc | +0.000 |
're | +0.000 |
patch | +0.000 |
count | +0.000 |
snack | +0.000 |
Neg logit contributions | |
spe | -0.000 |
overed | -0.000 |
led | -0.000 |
ls | -0.000 |
otte | -0.000 |
Token | spe |
Token ID | 4125 |
Feature activation | +5.65e+00 |
Pos logit contributions | |
otte | +0.075 |
ivating | +0.074 |
Acc | +0.074 |
rily | +0.073 |
overed | +0.072 |
Neg logit contributions | |
go | -0.035 |
wake | -0.035 |
pile | -0.035 |
count | -0.034 |
patch | -0.033 |
Token | ople |
Token ID | 643 |
Feature activation | +7.60e-01 |
Pos logit contributions | |
overed | +0.096 |
rily | +0.096 |
ls | +0.096 |
otte | +0.093 |
spe | +0.092 |
Neg logit contributions | |
bisc | -0.161 |
count | -0.151 |
wake | -0.148 |
est | -0.147 |
pile | -0.144 |
Token | were |
Token ID | 547 |
Feature activation | +1.71e-01 |
Pos logit contributions | |
overed | +0.024 |
rily | +0.023 |
spe | +0.023 |
ls | +0.023 |
han | +0.023 |
Neg logit contributions | |
wake | -0.012 |
count | -0.011 |
go | -0.011 |
come | -0.011 |
bisc | -0.010 |
Token | sad |
Token ID | 6507 |
Feature activation | -4.59e-01 |
Pos logit contributions | |
spe | +0.004 |
ls | +0.004 |
rily | +0.004 |
led | +0.004 |
riet | +0.004 |
Neg logit contributions | |
wake | -0.004 |
count | -0.003 |
come | -0.003 |
go | -0.003 |
paw | -0.003 |
Token | because |
Token ID | 780 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
bisc | +0.008 |
wake | +0.008 |
snack | +0.008 |
count | +0.008 |
patch | +0.008 |
Neg logit contributions | |
spe | -0.013 |
overed | -0.013 |
led | -0.013 |
han | -0.013 |
otte | -0.013 |
Token | they |
Token ID | 484 |
Feature activation | +4.46e-01 |
Pos logit contributions | |
rily | +0.003 |
led | +0.003 |
edly | +0.003 |
Acc | +0.003 |
riet | +0.003 |
Neg logit contributions | |
go | -0.003 |
winter | -0.003 |
come | -0.003 |
wake | -0.003 |
heat | -0.003 |
Token | ivered |
Token ID | 6396 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
bisc | +0.020 |
wake | +0.018 |
count | +0.018 |
snack | +0.018 |
patch | +0.017 |
Neg logit contributions | |
led | -0.015 |
overed | -0.014 |
spe | -0.013 |
ls | -0.013 |
rily | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | -6.10e-01 |
Pos logit contributions | |
bisc | +0.031 |
wake | +0.028 |
snack | +0.027 |
patch | +0.026 |
nose | +0.026 |
Neg logit contributions | |
spe | -0.046 |
otte | -0.044 |
han | -0.043 |
overed | -0.043 |
riet | -0.042 |
Token |
|
Token ID | 198 |
Feature activation | -7.95e-01 |
Pos logit contributions | |
bisc | +0.015 |
wake | +0.013 |
count | +0.013 |
snack | +0.013 |
patch | +0.012 |
Neg logit contributions | |
spe | -0.015 |
led | -0.014 |
overed | -0.014 |
ls | -0.014 |
otte | -0.014 |
Token |
|
Token ID | 198 |
Feature activation | -4.58e-01 |
Pos logit contributions | |
bisc | +0.018 |
wake | +0.016 |
count | +0.016 |
snack | +0.016 |
patch | +0.015 |
Neg logit contributions | |
spe | -0.020 |
overed | -0.019 |
led | -0.019 |
ls | -0.019 |
otte | -0.018 |
Token | " |
Token ID | 1 |
Feature activation | -5.49e-02 |
Pos logit contributions | |
bisc | +0.008 |
wake | +0.007 |
count | +0.007 |
snack | +0.007 |
patch | +0.006 |
Neg logit contributions | |
spe | -0.014 |
overed | -0.014 |
led | -0.013 |
rily | -0.013 |
ls | -0.013 |
Token | Br |
Token ID | 9414 |
Feature activation | -5.07e+00 |
Pos logit contributions | |
bisc | +0.001 |
count | +0.001 |
wake | +0.001 |
snack | +0.001 |
patch | +0.001 |
Neg logit contributions | |
led | -0.001 |
spe | -0.001 |
overed | -0.001 |
ls | -0.001 |
Acc | -0.001 |
Token | rr |
Token ID | 21062 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
bisc | +0.140 |
snack | +0.119 |
wake | +0.118 |
gloves | +0.117 |
coat | +0.117 |
Neg logit contributions | |
han | -0.106 |
overed | -0.102 |
narrative | -0.100 |
led | -0.096 |
uzz | -0.095 |
Token | , |
Token ID | 11 |
Feature activation | -9.87e-01 |
Pos logit contributions | |
bisc | +0.065 |
gloves | +0.056 |
snack | +0.056 |
patch | +0.055 |
coat | +0.055 |
Neg logit contributions | |
ls | -0.045 |
led | -0.041 |
spe | -0.040 |
oe | -0.036 |
te | -0.035 |
Token | it |
Token ID | 340 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
bisc | +0.023 |
count | +0.020 |
snack | +0.019 |
patch | +0.019 |
strip | +0.019 |
Neg logit contributions | |
ls | -0.023 |
spe | -0.023 |
led | -0.022 |
overed | -0.022 |
otte | -0.021 |
Token | 's |
Token ID | 338 |
Feature activation | -2.90e+00 |
Pos logit contributions | |
bisc | +0.031 |
snack | +0.026 |
count | +0.026 |
strip | +0.025 |
wake | +0.024 |
Neg logit contributions | |
led | -0.035 |
ls | -0.035 |
spe | -0.033 |
overed | -0.032 |
rily | -0.031 |
Token | freezing |
Token ID | 20884 |
Feature activation | -3.18e+00 |
Pos logit contributions | |
bisc | +0.067 |
count | +0.057 |
chop | +0.053 |
patch | +0.052 |
pile | +0.052 |
Neg logit contributions | |
led | -0.067 |
ls | -0.066 |
sadly | -0.064 |
rily | -0.064 |
ivating | -0.063 |
Token | cold |
Token ID | 4692 |
Feature activation | -2.48e+00 |
Pos logit contributions | |
led | +0.013 |
spe | +0.013 |
overed | +0.013 |
ls | +0.013 |
rily | +0.012 |
Neg logit contributions | |
bisc | -0.015 |
count | -0.013 |
wake | -0.013 |
snack | -0.013 |
patch | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | -7.46e-01 |
Pos logit contributions | |
bisc | +0.068 |
count | +0.060 |
wake | +0.059 |
snack | +0.058 |
patch | +0.057 |
Neg logit contributions | |
led | -0.051 |
spe | -0.050 |
overed | -0.050 |
ls | -0.049 |
rily | -0.046 |
Token |
|
Token ID | 198 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
bisc | +0.023 |
wake | +0.020 |
count | +0.020 |
snack | +0.020 |
patch | +0.020 |
Neg logit contributions | |
spe | -0.013 |
overed | -0.013 |
led | -0.013 |
ls | -0.012 |
otte | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | -7.89e-01 |
Pos logit contributions | |
bisc | +0.028 |
wake | +0.025 |
count | +0.024 |
snack | +0.024 |
patch | +0.024 |
Neg logit contributions | |
spe | -0.023 |
overed | -0.023 |
ls | -0.023 |
led | -0.022 |
otte | -0.022 |
Token | " |
Token ID | 1 |
Feature activation | -2.89e-01 |
Pos logit contributions | |
bisc | +0.016 |
wake | +0.014 |
count | +0.013 |
snack | +0.013 |
patch | +0.013 |
Neg logit contributions | |
spe | -0.022 |
overed | -0.022 |
led | -0.021 |
rily | -0.021 |
ls | -0.021 |
Token | Br |
Token ID | 9414 |
Feature activation | -5.04e+00 |
Pos logit contributions | |
bisc | +0.007 |
count | +0.007 |
wake | +0.007 |
snack | +0.006 |
patch | +0.006 |
Neg logit contributions | |
spe | -0.006 |
led | -0.006 |
overed | -0.006 |
ls | -0.006 |
otte | -0.006 |
Token | rr |
Token ID | 21062 |
Feature activation | -2.29e+00 |
Pos logit contributions | |
bisc | +0.138 |
wake | +0.115 |
snack | +0.114 |
carrot | +0.114 |
gloves | +0.113 |
Neg logit contributions | |
han | -0.110 |
narrative | -0.106 |
overed | -0.103 |
led | -0.098 |
oe | -0.095 |
Token | , |
Token ID | 11 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
bisc | +0.070 |
gloves | +0.059 |
snack | +0.059 |
patch | +0.058 |
chop | +0.057 |
Neg logit contributions | |
ls | -0.042 |
spe | -0.037 |
led | -0.036 |
han | -0.034 |
oe | -0.034 |
Token | it |
Token ID | 340 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
bisc | +0.030 |
count | +0.026 |
patch | +0.025 |
snack | +0.025 |
wake | +0.024 |
Neg logit contributions | |
ls | -0.022 |
spe | -0.022 |
overed | -0.022 |
led | -0.022 |
otte | -0.021 |
Token | is |
Token ID | 318 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
bisc | +0.038 |
snack | +0.032 |
count | +0.031 |
patch | +0.030 |
wake | +0.029 |
Neg logit contributions | |
led | -0.037 |
ls | -0.037 |
spe | -0.034 |
overed | -0.034 |
sadly | -0.032 |
Token | cold |
Token ID | 4692 |
Feature activation | -3.43e+00 |
Pos logit contributions | |
bisc | +0.036 |
count | +0.031 |
wake | +0.029 |
snack | +0.029 |
patch | +0.028 |
Neg logit contributions | |
led | -0.037 |
spe | -0.037 |
overed | -0.037 |
ls | -0.037 |
rily | -0.036 |
Token | hands |
Token ID | 2832 |
Feature activation | -2.29e+00 |
Pos logit contributions | |
bisc | +0.030 |
wake | +0.026 |
count | +0.025 |
snack | +0.025 |
patch | +0.024 |
Neg logit contributions | |
spe | -0.029 |
led | -0.028 |
overed | -0.028 |
ls | -0.027 |
otte | -0.026 |
Token | . |
Token ID | 13 |
Feature activation | -6.81e-01 |
Pos logit contributions | |
bisc | +0.060 |
count | +0.052 |
wake | +0.052 |
snack | +0.051 |
patch | +0.049 |
Neg logit contributions | |
spe | -0.050 |
led | -0.049 |
overed | -0.049 |
ls | -0.047 |
rily | -0.046 |
Token |
|
Token ID | 198 |
Feature activation | -7.48e-01 |
Pos logit contributions | |
bisc | +0.021 |
wake | +0.019 |
count | +0.018 |
snack | +0.018 |
patch | +0.018 |
Neg logit contributions | |
spe | -0.012 |
overed | -0.012 |
led | -0.012 |
ls | -0.011 |
otte | -0.011 |
Token |
|
Token ID | 198 |
Feature activation | -4.47e-01 |
Pos logit contributions | |
bisc | +0.018 |
wake | +0.016 |
count | +0.015 |
snack | +0.015 |
patch | +0.015 |
Neg logit contributions | |
spe | -0.018 |
overed | -0.018 |
led | -0.017 |
ls | -0.017 |
otte | -0.017 |
Token | " |
Token ID | 1 |
Feature activation | -3.74e-01 |
Pos logit contributions | |
bisc | +0.009 |
wake | +0.007 |
count | +0.007 |
snack | +0.007 |
patch | +0.007 |
Neg logit contributions | |
spe | -0.013 |
overed | -0.013 |
led | -0.013 |
ls | -0.012 |
rily | -0.012 |
Token | Br |
Token ID | 9414 |
Feature activation | -5.02e+00 |
Pos logit contributions | |
bisc | +0.010 |
count | +0.009 |
wake | +0.009 |
snack | +0.008 |
patch | +0.008 |
Neg logit contributions | |
spe | -0.008 |
led | -0.008 |
overed | -0.008 |
otte | -0.007 |
Acc | -0.007 |
Token | rr |
Token ID | 21062 |
Feature activation | -2.30e+00 |
Pos logit contributions | |
bisc | +0.133 |
snack | +0.112 |
wake | +0.112 |
pile | +0.112 |
patch | +0.111 |
Neg logit contributions | |
han | -0.111 |
narrative | -0.105 |
overed | -0.102 |
uzz | -0.098 |
led | -0.097 |
Token | , |
Token ID | 11 |
Feature activation | -9.03e-01 |
Pos logit contributions | |
bisc | +0.067 |
patch | +0.059 |
snack | +0.058 |
shade | +0.056 |
strip | +0.056 |
Neg logit contributions | |
ls | -0.042 |
led | -0.038 |
spe | -0.038 |
han | -0.036 |
te | -0.035 |
Token | you |
Token ID | 345 |
Feature activation | -2.95e-01 |
Pos logit contributions | |
bisc | +0.023 |
count | +0.020 |
snack | +0.019 |
patch | +0.019 |
wake | +0.019 |
Neg logit contributions | |
spe | -0.019 |
ls | -0.018 |
overed | -0.018 |
led | -0.018 |
otte | -0.017 |
Token | are |
Token ID | 389 |
Feature activation | -6.04e-01 |
Pos logit contributions | |
bisc | +0.006 |
strip | +0.005 |
snack | +0.005 |
patch | +0.005 |
shady | +0.005 |
Neg logit contributions | |
ls | -0.008 |
led | -0.007 |
sadly | -0.007 |
spe | -0.007 |
ivating | -0.006 |
Token | cold |
Token ID | 4692 |
Feature activation | -2.93e+00 |
Pos logit contributions | |
bisc | +0.015 |
wake | +0.013 |
count | +0.013 |
snack | +0.013 |
patch | +0.013 |
Neg logit contributions | |
spe | -0.014 |
overed | -0.014 |
led | -0.013 |
ls | -0.013 |
otte | -0.013 |
Token | display |
Token ID | 3359 |
Feature activation | -8.49e-02 |
Pos logit contributions | |
bisc | +0.015 |
wake | +0.011 |
snack | +0.011 |
count | +0.011 |
patch | +0.010 |
Neg logit contributions | |
spe | -0.031 |
overed | -0.030 |
led | -0.030 |
ls | -0.029 |
otte | -0.029 |
Token | their |
Token ID | 511 |
Feature activation | +5.08e-01 |
Pos logit contributions | |
bisc | +0.002 |
wake | +0.002 |
count | +0.002 |
snack | +0.002 |
patch | +0.002 |
Neg logit contributions | |
spe | -0.002 |
overed | -0.002 |
otte | -0.002 |
led | -0.002 |
ls | -0.002 |
Token | smiles |
Token ID | 21845 |
Feature activation | +4.61e-01 |
Pos logit contributions | |
spe | +0.014 |
led | +0.013 |
ls | +0.013 |
overed | +0.013 |
rily | +0.012 |
Neg logit contributions | |
bisc | -0.011 |
wake | -0.009 |
count | -0.009 |
patch | -0.009 |
snack | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | -2.81e-01 |
Pos logit contributions | |
spe | +0.010 |
overed | +0.009 |
ls | +0.009 |
led | +0.009 |
otte | +0.009 |
Neg logit contributions | |
bisc | -0.012 |
wake | -0.011 |
count | -0.011 |
snack | -0.011 |
patch | -0.011 |
Token | On |
Token ID | 1550 |
Feature activation | +2.44e-01 |
Pos logit contributions | |
bisc | +0.006 |
wake | +0.005 |
count | +0.005 |
snack | +0.005 |
patch | +0.005 |
Neg logit contributions | |
led | -0.007 |
spe | -0.007 |
overed | -0.007 |
ls | -0.007 |
otte | -0.007 |
Token | cold |
Token ID | 4692 |
Feature activation | -4.89e+00 |
Pos logit contributions | |
spe | +0.005 |
overed | +0.005 |
led | +0.005 |
ls | +0.005 |
otte | +0.005 |
Neg logit contributions | |
bisc | -0.006 |
wake | -0.006 |
count | -0.006 |
snack | -0.006 |
patch | -0.006 |
Token | days |
Token ID | 1528 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
bisc | +0.117 |
count | +0.095 |
wake | +0.089 |
patch | +0.089 |
snack | +0.088 |
Neg logit contributions | |
led | -0.128 |
ls | -0.118 |
overed | -0.117 |
spe | -0.115 |
han | -0.113 |
Token | the |
Token ID | 262 |
Feature activation | -4.23e-01 |
Pos logit contributions | |
bisc | +0.022 |
wake | +0.019 |
count | +0.019 |
snack | +0.019 |
patch | +0.018 |
Neg logit contributions | |
spe | -0.041 |
overed | -0.040 |
led | -0.039 |
ls | -0.039 |
rily | -0.039 |
Token | fireplace |
Token ID | 44250 |
Feature activation | -1.84e-01 |
Pos logit contributions | |
bisc | +0.010 |
count | +0.008 |
snack | +0.007 |
wake | +0.007 |
patch | +0.007 |
Neg logit contributions | |
spe | -0.011 |
led | -0.010 |
ls | -0.010 |
otte | -0.010 |
rily | -0.010 |
Token | would |
Token ID | 561 |
Feature activation | +5.21e-01 |
Pos logit contributions | |
bisc | +0.004 |
wake | +0.004 |
count | +0.004 |
snack | +0.004 |
patch | +0.003 |
Neg logit contributions | |
spe | -0.005 |
overed | -0.004 |
led | -0.004 |
ls | -0.004 |
otte | -0.004 |
Token | crack |
Token ID | 8469 |
Feature activation | -9.05e-01 |
Pos logit contributions | |
spe | +0.018 |
overed | +0.017 |
led | +0.017 |
ls | +0.017 |
rily | +0.017 |
Neg logit contributions | |
bisc | -0.007 |
snack | -0.005 |
count | -0.005 |
wake | -0.005 |
patch | -0.005 |
Token | iver |
Token ID | 1428 |
Feature activation | -2.39e+00 |
Pos logit contributions | |
bisc | +0.007 |
snack | +0.006 |
wake | +0.005 |
count | +0.005 |
patch | +0.005 |
Neg logit contributions | |
led | -0.019 |
overed | -0.019 |
spe | -0.019 |
ivating | -0.018 |
ls | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | -5.07e-01 |
Pos logit contributions | |
bisc | +0.065 |
count | +0.055 |
wake | +0.054 |
snack | +0.054 |
patch | +0.053 |
Neg logit contributions | |
spe | -0.052 |
led | -0.051 |
overed | -0.049 |
ls | -0.049 |
rily | -0.048 |
Token |
|
Token ID | 198 |
Feature activation | -6.77e-01 |
Pos logit contributions | |
bisc | +0.012 |
wake | +0.010 |
snack | +0.010 |
count | +0.010 |
nose | +0.010 |
Neg logit contributions | |
spe | -0.013 |
led | -0.012 |
ls | -0.012 |
overed | -0.012 |
otte | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | -4.49e-01 |
Pos logit contributions | |
bisc | +0.018 |
wake | +0.015 |
count | +0.015 |
snack | +0.015 |
patch | +0.015 |
Neg logit contributions | |
spe | -0.015 |
overed | -0.015 |
ls | -0.015 |
led | -0.014 |
otte | -0.014 |
Token | " |
Token ID | 1 |
Feature activation | -8.83e-02 |
Pos logit contributions | |
bisc | +0.009 |
wake | +0.007 |
count | +0.007 |
patch | +0.007 |
snack | +0.007 |
Neg logit contributions | |
spe | -0.013 |
overed | -0.013 |
led | -0.013 |
rily | -0.013 |
ls | -0.012 |
Token | Br |
Token ID | 9414 |
Feature activation | -4.89e+00 |
Pos logit contributions | |
bisc | +0.002 |
count | +0.002 |
wake | +0.002 |
snack | +0.002 |
patch | +0.002 |
Neg logit contributions | |
spe | -0.002 |
led | -0.002 |
overed | -0.002 |
otte | -0.002 |
ls | -0.002 |
Token | rr |
Token ID | 21062 |
Feature activation | -2.03e+00 |
Pos logit contributions | |
bisc | +0.134 |
snack | +0.111 |
carrot | +0.108 |
wake | +0.108 |
coat | +0.108 |
Neg logit contributions | |
han | -0.107 |
narrative | -0.105 |
overed | -0.104 |
led | -0.096 |
spe | -0.093 |
Token | , |
Token ID | 11 |
Feature activation | -9.38e-01 |
Pos logit contributions | |
bisc | +0.060 |
snack | +0.050 |
patch | +0.050 |
gloves | +0.050 |
coat | +0.049 |
Neg logit contributions | |
ls | -0.038 |
spe | -0.036 |
led | -0.035 |
te | -0.032 |
han | -0.032 |
Token | I |
Token ID | 314 |
Feature activation | -3.58e-01 |
Pos logit contributions | |
bisc | +0.024 |
count | +0.020 |
snack | +0.019 |
patch | +0.019 |
strip | +0.019 |
Neg logit contributions | |
spe | -0.020 |
ls | -0.020 |
overed | -0.020 |
led | -0.020 |
otte | -0.019 |
Token | 'm |
Token ID | 1101 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
bisc | +0.007 |
snack | +0.006 |
pile | +0.006 |
count | +0.006 |
nose | +0.006 |
Neg logit contributions | |
led | -0.009 |
spe | -0.009 |
ls | -0.009 |
sadly | -0.008 |
overed | -0.008 |
Token | cold |
Token ID | 4692 |
Feature activation | -3.19e+00 |
Pos logit contributions | |
bisc | +0.034 |
wake | +0.030 |
count | +0.029 |
snack | +0.029 |
patch | +0.028 |
Neg logit contributions | |
spe | -0.038 |
overed | -0.036 |
led | -0.036 |
ls | -0.035 |
otte | -0.034 |
Token | said |
Token ID | 531 |
Feature activation | -1.96e+00 |
Pos logit contributions | |
spe | +0.010 |
overed | +0.010 |
ivating | +0.010 |
rily | +0.010 |
ls | +0.010 |
Neg logit contributions | |
bisc | -0.007 |
count | -0.006 |
patch | -0.006 |
wake | -0.006 |
snack | -0.006 |
Token | , |
Token ID | 11 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
bisc | +0.027 |
wake | +0.023 |
count | +0.023 |
snack | +0.022 |
come | +0.021 |
Neg logit contributions | |
spe | -0.065 |
otte | -0.063 |
led | -0.063 |
overed | -0.062 |
ivating | -0.062 |
Token | " |
Token ID | 366 |
Feature activation | +4.27e-01 |
Pos logit contributions | |
spe | +0.003 |
led | +0.003 |
overed | +0.003 |
ls | +0.003 |
otte | +0.003 |
Neg logit contributions | |
bisc | -0.002 |
wake | -0.002 |
count | -0.002 |
snack | -0.002 |
patch | -0.002 |
Token | This |
Token ID | 1212 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
spe | +0.009 |
overed | +0.009 |
led | +0.009 |
ls | +0.009 |
rily | +0.009 |
Neg logit contributions | |
bisc | -0.011 |
wake | -0.010 |
count | -0.010 |
snack | -0.010 |
patch | -0.009 |
Token | is |
Token ID | 318 |
Feature activation | -1.70e+00 |
Pos logit contributions | |
bisc | +0.035 |
wake | +0.030 |
snack | +0.029 |
count | +0.028 |
iver | +0.028 |
Neg logit contributions | |
spe | -0.026 |
ls | -0.026 |
overed | -0.025 |
narrative | -0.024 |
han | -0.024 |
Token | called |
Token ID | 1444 |
Feature activation | -4.88e+00 |
Pos logit contributions | |
bisc | +0.045 |
wake | +0.039 |
snack | +0.038 |
count | +0.037 |
patch | +0.037 |
Neg logit contributions | |
spe | -0.037 |
overed | -0.036 |
led | -0.035 |
ls | -0.034 |
otte | -0.034 |
Token | striking |
Token ID | 8871 |
Feature activation | -2.22e+00 |
Pos logit contributions | |
bisc | +0.118 |
wake | +0.102 |
count | +0.099 |
snack | +0.099 |
patch | +0.098 |
Neg logit contributions | |
spe | -0.115 |
overed | -0.111 |
ls | -0.108 |
led | -0.107 |
otte | -0.106 |
Token | . |
Token ID | 13 |
Feature activation | -7.86e-01 |
Pos logit contributions | |
bisc | +0.053 |
wake | +0.046 |
count | +0.045 |
snack | +0.045 |
patch | +0.043 |
Neg logit contributions | |
spe | -0.054 |
led | -0.052 |
overed | -0.052 |
ls | -0.051 |
rily | -0.049 |
Token | Watch |
Token ID | 6305 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
bisc | +0.020 |
wake | +0.018 |
snack | +0.017 |
count | +0.017 |
patch | +0.017 |
Neg logit contributions | |
spe | -0.018 |
overed | -0.017 |
led | -0.017 |
ls | -0.016 |
otte | -0.016 |
Token | what |
Token ID | 644 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
bisc | +0.033 |
wake | +0.028 |
snack | +0.027 |
count | +0.027 |
patch | +0.027 |
Neg logit contributions | |
overed | -0.031 |
spe | -0.031 |
ls | -0.030 |
led | -0.030 |
han | -0.029 |
Token | I |
Token ID | 314 |
Feature activation | +2.06e-02 |
Pos logit contributions | |
spe | +0.028 |
overed | +0.027 |
led | +0.027 |
ls | +0.026 |
otte | +0.026 |
Neg logit contributions | |
bisc | -0.028 |
wake | -0.024 |
count | -0.023 |
snack | -0.023 |
patch | -0.023 |
Token | I |
Token ID | 40 |
Feature activation | +2.49e-01 |
Pos logit contributions | |
led | +0.017 |
ls | +0.017 |
rily | +0.017 |
overed | +0.017 |
sadly | +0.016 |
Neg logit contributions | |
bisc | -0.013 |
wake | -0.011 |
come | -0.011 |
count | -0.010 |
patch | -0.010 |
Token | like |
Token ID | 588 |
Feature activation | -7.69e-03 |
Pos logit contributions | |
spe | +0.008 |
overed | +0.008 |
led | +0.008 |
ls | +0.008 |
otte | +0.007 |
Neg logit contributions | |
bisc | -0.004 |
wake | -0.003 |
snack | -0.003 |
count | -0.003 |
patch | -0.003 |
Token | swimming |
Token ID | 14899 |
Feature activation | -2.46e+00 |
Pos logit contributions | |
bisc | +0.000 |
wake | +0.000 |
count | +0.000 |
snack | +0.000 |
patch | +0.000 |
Neg logit contributions | |
spe | -0.000 |
overed | -0.000 |
led | -0.000 |
ls | -0.000 |
otte | -0.000 |
Token | in |
Token ID | 287 |
Feature activation | +4.13e-02 |
Pos logit contributions | |
bisc | +0.070 |
wake | +0.063 |
count | +0.062 |
snack | +0.062 |
patch | +0.061 |
Neg logit contributions | |
spe | -0.049 |
overed | -0.047 |
led | -0.046 |
ls | -0.045 |
otte | -0.045 |
Token | the |
Token ID | 262 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
spe | +0.001 |
overed | +0.001 |
led | +0.001 |
ls | +0.001 |
otte | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
wake | -0.001 |
count | -0.001 |
snack | -0.001 |
patch | -0.001 |
Token | shallow |
Token ID | 19337 |
Feature activation | -4.87e+00 |
Pos logit contributions | |
bisc | +0.024 |
count | +0.022 |
snack | +0.020 |
wake | +0.020 |
chop | +0.020 |
Neg logit contributions | |
spe | -0.025 |
ls | -0.025 |
led | -0.025 |
narrative | -0.024 |
ized | -0.024 |
Token | water |
Token ID | 1660 |
Feature activation | -3.26e+00 |
Pos logit contributions | |
bisc | +0.107 |
count | +0.089 |
snack | +0.085 |
chop | +0.083 |
gloves | +0.083 |
Neg logit contributions | |
spe | -0.116 |
led | -0.112 |
ized | -0.111 |
ls | -0.109 |
closely | -0.107 |
Token | near |
Token ID | 1474 |
Feature activation | -6.75e-01 |
Pos logit contributions | |
bisc | +0.078 |
count | +0.066 |
wake | +0.065 |
snack | +0.065 |
patch | +0.060 |
Neg logit contributions | |
spe | -0.079 |
led | -0.076 |
overed | -0.076 |
ls | -0.075 |
rily | -0.074 |
Token | the |
Token ID | 262 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
bisc | +0.021 |
wake | +0.018 |
count | +0.018 |
snack | +0.018 |
nose | +0.017 |
Neg logit contributions | |
spe | -0.011 |
ls | -0.011 |
rily | -0.011 |
overed | -0.011 |
otte | -0.010 |
Token | rocks |
Token ID | 12586 |
Feature activation | -4.10e-02 |
Pos logit contributions | |
bisc | +0.023 |
count | +0.020 |
snack | +0.019 |
wake | +0.019 |
chop | +0.018 |
Neg logit contributions | |
spe | -0.025 |
ls | -0.024 |
overed | -0.023 |
led | -0.023 |
ivating | -0.023 |
Token | . |
Token ID | 13 |
Feature activation | -5.88e-01 |
Pos logit contributions | |
bisc | +0.001 |
wake | +0.001 |
count | +0.001 |
snack | +0.001 |
patch | +0.001 |
Neg logit contributions | |
spe | -0.001 |
overed | -0.001 |
led | -0.001 |
ls | -0.001 |
otte | -0.001 |
Token | rainbow |
Token ID | 27223 |
Feature activation | -6.06e-01 |
Pos logit contributions | |
spe | +0.013 |
overed | +0.012 |
ls | +0.012 |
led | +0.012 |
otte | +0.012 |
Neg logit contributions | |
bisc | -0.010 |
wake | -0.009 |
count | -0.009 |
snack | -0.008 |
patch | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | -6.79e-01 |
Pos logit contributions | |
bisc | +0.014 |
wake | +0.012 |
count | +0.012 |
snack | +0.012 |
patch | +0.012 |
Neg logit contributions | |
spe | -0.015 |
overed | -0.015 |
led | -0.015 |
ls | -0.014 |
otte | -0.014 |
Token |
|
Token ID | 198 |
Feature activation | -6.40e-01 |
Pos logit contributions | |
bisc | +0.017 |
wake | +0.015 |
count | +0.014 |
snack | +0.014 |
patch | +0.014 |
Neg logit contributions | |
spe | -0.016 |
overed | -0.016 |
led | -0.016 |
ls | -0.015 |
otte | -0.015 |
Token |
|
Token ID | 198 |
Feature activation | -3.08e-01 |
Pos logit contributions | |
bisc | +0.018 |
wake | +0.016 |
count | +0.016 |
snack | +0.016 |
nose | +0.015 |
Neg logit contributions | |
ls | -0.012 |
overed | -0.012 |
spe | -0.012 |
led | -0.011 |
otte | -0.011 |
Token | " |
Token ID | 1 |
Feature activation | +6.14e-02 |
Pos logit contributions | |
bisc | +0.007 |
wake | +0.006 |
snack | +0.006 |
count | +0.006 |
nose | +0.006 |
Neg logit contributions | |
spe | -0.008 |
overed | -0.008 |
ls | -0.008 |
led | -0.008 |
otte | -0.007 |
Token | Br |
Token ID | 9414 |
Feature activation | -4.87e+00 |
Pos logit contributions | |
spe | +0.001 |
overed | +0.001 |
led | +0.001 |
ivating | +0.001 |
otte | +0.001 |
Neg logit contributions | |
bisc | -0.002 |
wake | -0.002 |
count | -0.002 |
snack | -0.002 |
patch | -0.002 |
Token | ill |
Token ID | 359 |
Feature activation | -6.72e-01 |
Pos logit contributions | |
bisc | +0.133 |
carrot | +0.113 |
gloves | +0.111 |
coat | +0.110 |
snack | +0.110 |
Neg logit contributions | |
han | -0.106 |
overed | -0.104 |
narrative | -0.104 |
led | -0.095 |
spe | -0.093 |
Token | iant |
Token ID | 3014 |
Feature activation | +1.85e-01 |
Pos logit contributions | |
bisc | +0.014 |
wake | +0.012 |
snack | +0.011 |
count | +0.011 |
patch | +0.011 |
Neg logit contributions | |
spe | -0.019 |
led | -0.018 |
overed | -0.018 |
ls | -0.018 |
otte | -0.017 |
Token | !" |
Token ID | 2474 |
Feature activation | +3.84e-01 |
Pos logit contributions | |
overed | +0.005 |
spe | +0.005 |
ls | +0.004 |
led | +0.004 |
han | +0.004 |
Neg logit contributions | |
bisc | -0.004 |
wake | -0.004 |
snack | -0.003 |
count | -0.003 |
patch | -0.003 |
Token | Sam |
Token ID | 3409 |
Feature activation | +6.39e-01 |
Pos logit contributions | |
spe | +0.014 |
rily | +0.013 |
led | +0.013 |
overed | +0.013 |
otte | +0.013 |
Neg logit contributions | |
bisc | -0.006 |
count | -0.005 |
snack | -0.005 |
go | -0.005 |
come | -0.005 |
Token | agreed |
Token ID | 4987 |
Feature activation | +6.97e-01 |
Pos logit contributions | |
spe | +0.015 |
overed | +0.015 |
led | +0.015 |
ls | +0.015 |
otte | +0.014 |
Neg logit contributions | |
bisc | -0.015 |
wake | -0.013 |
count | -0.013 |
snack | -0.013 |
patch | -0.013 |
Token | Pete |
Token ID | 17542 |
Feature activation | +5.74e-01 |
Pos logit contributions | |
bisc | +0.004 |
wake | +0.003 |
snack | +0.003 |
patch | +0.003 |
count | +0.003 |
Neg logit contributions | |
spe | -0.004 |
led | -0.004 |
ls | -0.004 |
overed | -0.003 |
otte | -0.003 |
Token | tried |
Token ID | 3088 |
Feature activation | -4.14e-01 |
Pos logit contributions | |
spe | +0.015 |
overed | +0.015 |
ls | +0.014 |
led | +0.014 |
rily | +0.014 |
Neg logit contributions | |
bisc | -0.013 |
wake | -0.011 |
count | -0.011 |
snack | -0.011 |
patch | -0.011 |
Token | again |
Token ID | 757 |
Feature activation | +6.83e-01 |
Pos logit contributions | |
bisc | +0.012 |
wake | +0.010 |
count | +0.010 |
snack | +0.010 |
patch | +0.010 |
Neg logit contributions | |
spe | -0.008 |
overed | -0.008 |
led | -0.008 |
ls | -0.008 |
otte | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | +7.33e-02 |
Pos logit contributions | |
spe | +0.013 |
sadly | +0.012 |
otte | +0.011 |
overed | +0.011 |
ls | +0.011 |
Neg logit contributions | |
bisc | -0.020 |
wake | -0.017 |
snack | -0.017 |
patch | -0.016 |
count | -0.016 |
Token | " |
Token ID | 366 |
Feature activation | +7.33e-01 |
Pos logit contributions | |
spe | +0.002 |
led | +0.002 |
ls | +0.002 |
overed | +0.002 |
otte | +0.002 |
Neg logit contributions | |
bisc | -0.002 |
wake | -0.002 |
snack | -0.002 |
count | -0.002 |
patch | -0.001 |
Token | Br |
Token ID | 9414 |
Feature activation | -4.85e+00 |
Pos logit contributions | |
spe | +0.017 |
overed | +0.016 |
otte | +0.016 |
led | +0.015 |
ivating | +0.015 |
Neg logit contributions | |
bisc | -0.019 |
wake | -0.016 |
count | -0.016 |
snack | -0.016 |
patch | -0.016 |
Token | ush |
Token ID | 1530 |
Feature activation | -1.88e+00 |
Pos logit contributions | |
bisc | +0.142 |
snack | +0.120 |
wake | +0.119 |
coat | +0.119 |
carrot | +0.118 |
Neg logit contributions | |
overed | -0.100 |
han | -0.095 |
narrative | -0.088 |
led | -0.086 |
Acc | -0.084 |
Token | , |
Token ID | 11 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
bisc | +0.045 |
wake | +0.037 |
snack | +0.037 |
count | +0.036 |
patch | +0.035 |
Neg logit contributions | |
spe | -0.046 |
led | -0.045 |
overed | -0.045 |
ls | -0.044 |
otte | -0.042 |
Token | please |
Token ID | 3387 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
bisc | +0.026 |
wake | +0.023 |
count | +0.022 |
snack | +0.022 |
patch | +0.022 |
Neg logit contributions | |
spe | -0.027 |
overed | -0.026 |
led | -0.026 |
ls | -0.025 |
otte | -0.025 |
Token | call |
Token ID | 869 |
Feature activation | -2.90e+00 |
Pos logit contributions | |
bisc | +0.032 |
patch | +0.023 |
snack | +0.023 |
strip | +0.023 |
carrot | +0.022 |
Neg logit contributions | |
spe | -0.037 |
sadly | -0.035 |
overed | -0.035 |
otte | -0.034 |
led | -0.034 |
Token | my |
Token ID | 616 |
Feature activation | -7.19e-01 |
Pos logit contributions | |
bisc | +0.077 |
count | +0.065 |
wake | +0.065 |
patch | +0.064 |
snack | +0.063 |
Neg logit contributions | |
spe | -0.059 |
led | -0.057 |
otte | -0.057 |
ls | -0.057 |
overed | -0.057 |
Token | . |
Token ID | 13 |
Feature activation | -2.26e-01 |
Pos logit contributions | |
bisc | +0.001 |
wake | +0.001 |
count | +0.001 |
snack | +0.001 |
patch | +0.001 |
Neg logit contributions | |
spe | -0.001 |
overed | -0.001 |
led | -0.001 |
ls | -0.001 |
rily | -0.001 |
Token | Inside |
Token ID | 14384 |
Feature activation | +3.63e-02 |
Pos logit contributions | |
bisc | +0.006 |
wake | +0.005 |
count | +0.005 |
snack | +0.005 |
patch | +0.005 |
Neg logit contributions | |
spe | -0.005 |
led | -0.005 |
ls | -0.005 |
overed | -0.005 |
otte | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
spe | +0.001 |
overed | +0.001 |
led | +0.001 |
ls | +0.001 |
otte | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
wake | -0.001 |
count | -0.001 |
snack | -0.001 |
patch | -0.001 |
Token | filled |
Token ID | 5901 |
Feature activation | +2.96e-01 |
Pos logit contributions | |
bisc | +0.031 |
wake | +0.026 |
count | +0.026 |
snack | +0.026 |
patch | +0.025 |
Neg logit contributions | |
spe | -0.037 |
overed | -0.036 |
led | -0.036 |
ls | -0.036 |
otte | -0.035 |
Token | with |
Token ID | 351 |
Feature activation | +5.59e-01 |
Pos logit contributions | |
spe | +0.006 |
overed | +0.006 |
ls | +0.006 |
led | +0.006 |
Acc | +0.006 |
Neg logit contributions | |
bisc | -0.008 |
snack | -0.007 |
wake | -0.007 |
patch | -0.007 |
count | -0.007 |
Token | warm |
Token ID | 5814 |
Feature activation | -4.75e+00 |
Pos logit contributions | |
spe | +0.015 |
led | +0.014 |
overed | +0.014 |
ls | +0.014 |
rily | +0.013 |
Neg logit contributions | |
bisc | -0.012 |
wake | -0.011 |
count | -0.010 |
snack | -0.010 |
patch | -0.010 |
Token | air |
Token ID | 1633 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
bisc | +0.091 |
count | +0.085 |
pound | +0.079 |
chop | +0.076 |
wake | +0.075 |
Neg logit contributions | |
led | -0.130 |
ls | -0.119 |
overed | -0.115 |
han | -0.112 |
ized | -0.110 |
Token | and |
Token ID | 290 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
bisc | +0.027 |
wake | +0.023 |
count | +0.023 |
snack | +0.023 |
patch | +0.022 |
Neg logit contributions | |
spe | -0.026 |
overed | -0.025 |
led | -0.025 |
ls | -0.025 |
otte | -0.024 |
Token | fun |
Token ID | 1257 |
Feature activation | -3.48e-01 |
Pos logit contributions | |
bisc | +0.022 |
wake | +0.019 |
count | +0.018 |
snack | +0.018 |
patch | +0.018 |
Neg logit contributions | |
spe | -0.027 |
led | -0.026 |
overed | -0.026 |
ls | -0.025 |
otte | -0.024 |
Token | things |
Token ID | 1243 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
bisc | +0.008 |
wake | +0.007 |
count | +0.007 |
snack | +0.007 |
patch | +0.007 |
Neg logit contributions | |
spe | -0.009 |
overed | -0.008 |
led | -0.008 |
ls | -0.008 |
otte | -0.008 |
Token | to |
Token ID | 284 |
Feature activation | +5.91e-01 |
Pos logit contributions | |
spe | +0.033 |
overed | +0.032 |
led | +0.031 |
otte | +0.031 |
ls | +0.031 |
Neg logit contributions | |
bisc | -0.035 |
wake | -0.031 |
count | -0.030 |
snack | -0.030 |
patch | -0.029 |
Token | the |
Token ID | 262 |
Feature activation | +2.64e-01 |
Pos logit contributions | |
spe | +0.003 |
overed | +0.003 |
led | +0.003 |
ls | +0.003 |
otte | +0.003 |
Neg logit contributions | |
bisc | -0.004 |
wake | -0.003 |
count | -0.003 |
snack | -0.003 |
patch | -0.003 |
Token | icy |
Token ID | 30284 |
Feature activation | -2.22e+00 |
Pos logit contributions | |
spe | +0.008 |
overed | +0.008 |
led | +0.008 |
ls | +0.008 |
otte | +0.007 |
Neg logit contributions | |
bisc | -0.005 |
count | -0.004 |
wake | -0.004 |
snack | -0.004 |
patch | -0.003 |
Token | weather |
Token ID | 6193 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
bisc | +0.039 |
count | +0.033 |
wake | +0.031 |
snack | +0.028 |
chop | +0.028 |
Neg logit contributions | |
spe | -0.066 |
led | -0.066 |
ls | -0.065 |
overed | -0.064 |
ized | -0.061 |
Token | . |
Token ID | 13 |
Feature activation | +1.59e-01 |
Pos logit contributions | |
bisc | +0.040 |
count | +0.035 |
wake | +0.034 |
snack | +0.034 |
patch | +0.032 |
Neg logit contributions | |
spe | -0.038 |
overed | -0.037 |
led | -0.037 |
ls | -0.035 |
rily | -0.034 |
Token | " |
Token ID | 366 |
Feature activation | +5.77e-01 |
Pos logit contributions | |
led | +0.004 |
spe | +0.004 |
ls | +0.004 |
overed | +0.004 |
ized | +0.004 |
Neg logit contributions | |
bisc | -0.004 |
snack | -0.003 |
wake | -0.003 |
patch | -0.003 |
count | -0.003 |
Token | Br |
Token ID | 9414 |
Feature activation | -4.70e+00 |
Pos logit contributions | |
spe | +0.014 |
overed | +0.014 |
ls | +0.014 |
led | +0.014 |
rily | +0.013 |
Neg logit contributions | |
bisc | -0.014 |
wake | -0.012 |
count | -0.012 |
snack | -0.012 |
patch | -0.011 |
Token | rr |
Token ID | 21062 |
Feature activation | -2.07e+00 |
Pos logit contributions | |
bisc | +0.127 |
wake | +0.108 |
snack | +0.105 |
coat | +0.103 |
carrot | +0.103 |
Neg logit contributions | |
overed | -0.101 |
han | -0.101 |
led | -0.097 |
spe | -0.094 |
narrative | -0.093 |
Token | , |
Token ID | 11 |
Feature activation | -3.07e-01 |
Pos logit contributions | |
bisc | +0.054 |
snack | +0.045 |
wake | +0.044 |
patch | +0.044 |
chop | +0.043 |
Neg logit contributions | |
spe | -0.046 |
ls | -0.044 |
led | -0.044 |
han | -0.042 |
overed | -0.041 |
Token | it |
Token ID | 340 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
bisc | +0.008 |
wake | +0.007 |
count | +0.007 |
snack | +0.007 |
patch | +0.007 |
Neg logit contributions | |
spe | -0.007 |
overed | -0.007 |
led | -0.007 |
ls | -0.007 |
otte | -0.006 |
Token | 's |
Token ID | 338 |
Feature activation | -2.75e+00 |
Pos logit contributions | |
bisc | +0.021 |
count | +0.017 |
snack | +0.017 |
wake | +0.017 |
patch | +0.016 |
Neg logit contributions | |
spe | -0.030 |
led | -0.030 |
ls | -0.030 |
overed | -0.029 |
rily | -0.028 |
Token | so |
Token ID | 523 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
bisc | +0.068 |
count | +0.057 |
chop | +0.054 |
wake | +0.054 |
patch | +0.053 |
Neg logit contributions | |
spe | -0.061 |
led | -0.060 |
overed | -0.058 |
sadly | -0.058 |
ivating | -0.057 |
Token | to |
Token ID | 284 |
Feature activation | +5.17e-01 |
Pos logit contributions | |
bisc | +0.021 |
wake | +0.018 |
count | +0.018 |
snack | +0.018 |
nose | +0.017 |
Neg logit contributions | |
spe | -0.025 |
overed | -0.025 |
led | -0.024 |
ls | -0.024 |
otte | -0.024 |
Token | the |
Token ID | 262 |
Feature activation | -5.39e-01 |
Pos logit contributions | |
spe | +0.014 |
led | +0.013 |
ls | +0.013 |
overed | +0.013 |
Acc | +0.012 |
Neg logit contributions | |
bisc | -0.012 |
wake | -0.011 |
count | -0.011 |
snack | -0.011 |
come | -0.011 |
Token | brush |
Token ID | 14093 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
bisc | +0.012 |
're | +0.010 |
urger | +0.010 |
iver | +0.010 |
wake | +0.010 |
Neg logit contributions | |
closely | -0.011 |
overed | -0.011 |
Acc | -0.011 |
spe | -0.011 |
otte | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | +3.38e-01 |
Pos logit contributions | |
bisc | +0.008 |
wake | +0.007 |
count | +0.007 |
snack | +0.007 |
patch | +0.007 |
Neg logit contributions | |
spe | -0.008 |
overed | -0.008 |
led | -0.007 |
ls | -0.007 |
otte | -0.007 |
Token | " |
Token ID | 366 |
Feature activation | +9.98e-01 |
Pos logit contributions | |
spe | +0.008 |
overed | +0.008 |
led | +0.008 |
ls | +0.008 |
otte | +0.008 |
Neg logit contributions | |
bisc | -0.008 |
wake | -0.007 |
count | -0.007 |
snack | -0.007 |
patch | -0.007 |
Token | Br |
Token ID | 9414 |
Feature activation | -4.70e+00 |
Pos logit contributions | |
spe | +0.024 |
led | +0.022 |
overed | +0.022 |
otte | +0.022 |
rily | +0.021 |
Neg logit contributions | |
bisc | -0.025 |
wake | -0.022 |
count | -0.022 |
snack | -0.021 |
pile | -0.021 |
Token | ush |
Token ID | 1530 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
bisc | +0.139 |
snack | +0.118 |
wake | +0.118 |
carrot | +0.116 |
coat | +0.115 |
Neg logit contributions | |
overed | -0.095 |
han | -0.089 |
led | -0.082 |
narrative | -0.082 |
otte | -0.081 |
Token | , |
Token ID | 11 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
bisc | +0.058 |
wake | +0.048 |
snack | +0.048 |
count | +0.047 |
patch | +0.045 |
Neg logit contributions | |
spe | -0.057 |
overed | -0.055 |
led | -0.055 |
ls | -0.053 |
otte | -0.052 |
Token | can |
Token ID | 460 |
Feature activation | +7.74e-01 |
Pos logit contributions | |
bisc | +0.025 |
wake | +0.021 |
count | +0.021 |
snack | +0.021 |
patch | +0.020 |
Neg logit contributions | |
spe | -0.026 |
overed | -0.025 |
led | -0.024 |
ls | -0.024 |
otte | -0.024 |
Token | you |
Token ID | 345 |
Feature activation | -8.08e-01 |
Pos logit contributions | |
spe | +0.017 |
ls | +0.016 |
overed | +0.016 |
led | +0.016 |
otte | +0.015 |
Neg logit contributions | |
bisc | -0.021 |
wake | -0.017 |
snack | -0.017 |
patch | -0.017 |
count | -0.017 |
Token | call |
Token ID | 869 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
bisc | +0.016 |
snack | +0.010 |
est | +0.010 |
snow | +0.010 |
pile | +0.010 |
Neg logit contributions | |
otte | -0.022 |
sadly | -0.022 |
spe | -0.022 |
ivating | -0.022 |
han | -0.021 |
Token | man |
Token ID | 582 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
bisc | +0.013 |
count | +0.011 |
wake | +0.011 |
patch | +0.010 |
snack | +0.010 |
Neg logit contributions | |
spe | -0.018 |
led | -0.017 |
overed | -0.017 |
otte | -0.017 |
ls | -0.017 |
Token | cooked |
Token ID | 15847 |
Feature activation | -2.32e+00 |
Pos logit contributions | |
overed | +0.032 |
spe | +0.032 |
ls | +0.031 |
led | +0.031 |
otte | +0.030 |
Neg logit contributions | |
bisc | -0.028 |
wake | -0.025 |
count | -0.024 |
snack | -0.024 |
patch | -0.023 |
Token | all |
Token ID | 477 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
bisc | +0.068 |
wake | +0.060 |
count | +0.059 |
're | +0.057 |
strip | +0.056 |
Neg logit contributions | |
led | -0.043 |
rily | -0.043 |
ls | -0.040 |
overed | -0.039 |
ize | -0.038 |
Token | sorts |
Token ID | 10524 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
bisc | +0.031 |
wake | +0.028 |
snack | +0.028 |
count | +0.027 |
patch | +0.027 |
Neg logit contributions | |
overed | -0.035 |
spe | -0.034 |
ivating | -0.033 |
ls | -0.033 |
otte | -0.033 |
Token | of |
Token ID | 286 |
Feature activation | -2.09e-01 |
Pos logit contributions | |
spe | +0.051 |
otte | +0.050 |
Acc | +0.048 |
overed | +0.048 |
ivating | +0.047 |
Neg logit contributions | |
bisc | -0.013 |
est | -0.012 |
come | -0.011 |
welcome | -0.011 |
snack | -0.010 |
Token | cold |
Token ID | 4692 |
Feature activation | -4.70e+00 |
Pos logit contributions | |
bisc | +0.004 |
count | +0.003 |
wake | +0.003 |
patch | +0.003 |
nose | +0.003 |
Neg logit contributions | |
led | -0.006 |
ls | -0.006 |
spe | -0.006 |
rily | -0.005 |
ize | -0.005 |
Token | food |
Token ID | 2057 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
bisc | +0.095 |
count | +0.086 |
wake | +0.081 |
strip | +0.076 |
patch | +0.075 |
Neg logit contributions | |
led | -0.137 |
ls | -0.122 |
edly | -0.114 |
ized | -0.112 |
overed | -0.111 |
Token | in |
Token ID | 287 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
bisc | +0.039 |
wake | +0.034 |
count | +0.033 |
snack | +0.032 |
patch | +0.031 |
Neg logit contributions | |
spe | -0.033 |
led | -0.033 |
overed | -0.032 |
ls | -0.032 |
rily | -0.031 |
Token | the |
Token ID | 262 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
spe | +0.002 |
rily | +0.002 |
led | +0.002 |
overed | +0.002 |
ls | +0.002 |
Neg logit contributions | |
bisc | -0.004 |
wake | -0.004 |
count | -0.004 |
snack | -0.003 |
patch | -0.003 |
Token | big |
Token ID | 1263 |
Feature activation | -1.83e+00 |
Pos logit contributions | |
bisc | +0.021 |
count | +0.019 |
're | +0.017 |
wake | +0.017 |
snack | +0.016 |
Neg logit contributions | |
spe | -0.028 |
led | -0.026 |
ls | -0.026 |
rily | -0.025 |
ized | -0.025 |
Token | silver |
Token ID | 8465 |
Feature activation | -3.52e-01 |
Pos logit contributions | |
bisc | +0.038 |
count | +0.035 |
're | +0.033 |
wake | +0.032 |
iver | +0.031 |
Neg logit contributions | |
ls | -0.041 |
spe | -0.040 |
ized | -0.040 |
led | -0.039 |
sadly | -0.039 |
Token | and |
Token ID | 290 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
bisc | +0.007 |
wake | +0.006 |
count | +0.006 |
patch | +0.006 |
snack | +0.006 |
Neg logit contributions | |
spe | -0.007 |
overed | -0.007 |
otte | -0.007 |
led | -0.007 |
ls | -0.007 |
Token | w |
Token ID | 266 |
Feature activation | -7.15e-01 |
Pos logit contributions | |
bisc | +0.032 |
count | +0.028 |
wake | +0.028 |
snack | +0.028 |
patch | +0.027 |
Neg logit contributions | |
spe | -0.031 |
overed | -0.030 |
led | -0.030 |
ls | -0.030 |
rily | -0.029 |
Token | aded |
Token ID | 5286 |
Feature activation | -3.20e+00 |
Pos logit contributions | |
bisc | +0.019 |
wake | +0.017 |
count | +0.017 |
snack | +0.017 |
patch | +0.017 |
Neg logit contributions | |
spe | -0.016 |
otte | -0.014 |
overed | -0.014 |
ls | -0.014 |
led | -0.014 |
Token | in |
Token ID | 287 |
Feature activation | -6.73e-01 |
Pos logit contributions | |
bisc | +0.088 |
patch | +0.078 |
wake | +0.077 |
snack | +0.077 |
go | +0.077 |
Neg logit contributions | |
otte | -0.068 |
spe | -0.067 |
Acc | -0.064 |
overed | -0.063 |
ize | -0.062 |
Token | the |
Token ID | 262 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
bisc | +0.018 |
wake | +0.016 |
count | +0.015 |
snack | +0.015 |
patch | +0.015 |
Neg logit contributions | |
spe | -0.015 |
overed | -0.014 |
led | -0.014 |
ls | -0.014 |
rily | -0.013 |
Token | shallow |
Token ID | 19337 |
Feature activation | -4.67e+00 |
Pos logit contributions | |
bisc | +0.017 |
wake | +0.014 |
count | +0.014 |
snack | +0.013 |
patch | +0.013 |
Neg logit contributions | |
spe | -0.036 |
overed | -0.036 |
led | -0.035 |
ls | -0.035 |
otte | -0.035 |
Token | water |
Token ID | 1660 |
Feature activation | -3.18e+00 |
Pos logit contributions | |
bisc | +0.095 |
count | +0.078 |
wake | +0.075 |
snack | +0.075 |
chop | +0.071 |
Neg logit contributions | |
spe | -0.120 |
led | -0.118 |
ized | -0.116 |
ls | -0.116 |
overed | -0.114 |
Token | . |
Token ID | 13 |
Feature activation | -3.48e-01 |
Pos logit contributions | |
bisc | +0.080 |
wake | +0.069 |
count | +0.069 |
snack | +0.068 |
patch | +0.064 |
Neg logit contributions | |
spe | -0.073 |
overed | -0.071 |
led | -0.071 |
ls | -0.069 |
rily | -0.068 |
Token | He |
Token ID | 679 |
Feature activation | +2.31e-02 |
Pos logit contributions | |
bisc | +0.009 |
wake | +0.008 |
count | +0.008 |
snack | +0.008 |
patch | +0.008 |
Neg logit contributions | |
spe | -0.008 |
overed | -0.007 |
led | -0.007 |
ls | -0.007 |
otte | -0.007 |
Token | pointed |
Token ID | 6235 |
Feature activation | -9.28e-01 |
Pos logit contributions | |
spe | +0.001 |
overed | +0.001 |
led | +0.001 |
ls | +0.001 |
otte | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
wake | -0.000 |
count | -0.000 |
snack | -0.000 |
patch | -0.000 |
Token | to |
Token ID | 284 |
Feature activation | -8.95e-01 |
Pos logit contributions | |
bisc | +0.023 |
wake | +0.021 |
count | +0.020 |
patch | +0.020 |
snack | +0.020 |
Neg logit contributions | |
spe | -0.022 |
otte | -0.021 |
overed | -0.020 |
led | -0.020 |
ivating | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | -4.18e-01 |
Pos logit contributions | |
bisc | +0.019 |
wake | +0.016 |
count | +0.016 |
snack | +0.016 |
patch | +0.016 |
Neg logit contributions | |
spe | -0.016 |
overed | -0.015 |
ls | -0.015 |
led | -0.015 |
otte | -0.015 |
Token | " |
Token ID | 1 |
Feature activation | -1.84e-01 |
Pos logit contributions | |
bisc | +0.009 |
wake | +0.008 |
count | +0.008 |
snack | +0.008 |
patch | +0.008 |
Neg logit contributions | |
spe | -0.011 |
overed | -0.011 |
led | -0.010 |
ls | -0.010 |
otte | -0.010 |
Token | Look |
Token ID | 8567 |
Feature activation | -4.06e-01 |
Pos logit contributions | |
bisc | +0.005 |
wake | +0.005 |
count | +0.005 |
patch | +0.005 |
snack | +0.004 |
Neg logit contributions | |
spe | -0.004 |
overed | -0.003 |
narrative | -0.003 |
led | -0.003 |
ivating | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | -4.46e-01 |
Pos logit contributions | |
bisc | +0.010 |
wake | +0.008 |
count | +0.008 |
patch | +0.008 |
snack | +0.008 |
Neg logit contributions | |
ls | -0.010 |
led | -0.010 |
overed | -0.009 |
spe | -0.009 |
otte | -0.009 |
Token | there |
Token ID | 612 |
Feature activation | -8.14e-01 |
Pos logit contributions | |
bisc | +0.013 |
wake | +0.011 |
count | +0.011 |
patch | +0.011 |
snack | +0.011 |
Neg logit contributions | |
spe | -0.009 |
ls | -0.008 |
overed | -0.008 |
led | -0.008 |
otte | -0.008 |
Token | 's |
Token ID | 338 |
Feature activation | -4.66e+00 |
Pos logit contributions | |
bisc | +0.023 |
snack | +0.018 |
wake | +0.018 |
count | +0.018 |
iver | +0.018 |
Neg logit contributions | |
spe | -0.018 |
ls | -0.017 |
led | -0.015 |
overed | -0.014 |
sadly | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | -8.48e-01 |
Pos logit contributions | |
bisc | +0.141 |
itten | +0.119 |
wake | +0.117 |
snack | +0.112 |
tissue | +0.111 |
Neg logit contributions | |
spe | -0.073 |
sadly | -0.072 |
overed | -0.067 |
ls | -0.066 |
ivating | -0.065 |
Token | big |
Token ID | 1263 |
Feature activation | -9.97e-01 |
Pos logit contributions | |
bisc | +0.019 |
count | +0.016 |
est | +0.016 |
wake | +0.015 |
're | +0.015 |
Neg logit contributions | |
spe | -0.021 |
ls | -0.019 |
gol | -0.018 |
ize | -0.018 |
narrative | -0.018 |
Token | waterfall |
Token ID | 42467 |
Feature activation | -7.02e-01 |
Pos logit contributions | |
bisc | +0.022 |
count | +0.018 |
wake | +0.018 |
're | +0.017 |
snack | +0.017 |
Neg logit contributions | |
spe | -0.024 |
ls | -0.024 |
overed | -0.022 |
sadly | -0.022 |
gol | -0.022 |
Token | ahead |
Token ID | 4058 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
bisc | +0.011 |
wake | +0.010 |
count | +0.009 |
snack | +0.009 |
patch | +0.009 |
Neg logit contributions | |
spe | -0.022 |
overed | -0.022 |
otte | -0.021 |
led | -0.021 |
ls | -0.021 |
Token | !" |
Token ID | 2474 |
Feature activation | +3.38e-01 |
Pos logit contributions | |
bisc | +0.027 |
wake | +0.022 |
count | +0.022 |
snack | +0.022 |
patch | +0.021 |
Neg logit contributions | |
spe | -0.036 |
overed | -0.035 |
led | -0.035 |
ls | -0.034 |
rily | -0.033 |
Token | for |
Token ID | 329 |
Feature activation | +7.09e-01 |
Pos logit contributions | |
bisc | +0.055 |
wake | +0.048 |
count | +0.047 |
snack | +0.046 |
patch | +0.045 |
Neg logit contributions | |
spe | -0.057 |
ls | -0.056 |
led | -0.056 |
overed | -0.055 |
rily | -0.053 |
Token | yourself |
Token ID | 3511 |
Feature activation | -6.61e-01 |
Pos logit contributions | |
otte | +0.012 |
ls | +0.011 |
ize | +0.011 |
overed | +0.011 |
rily | +0.010 |
Neg logit contributions | |
bisc | -0.022 |
wake | -0.020 |
est | -0.019 |
patch | -0.019 |
count | -0.019 |
Token | . |
Token ID | 13 |
Feature activation | -3.51e-01 |
Pos logit contributions | |
bisc | +0.013 |
wake | +0.011 |
count | +0.011 |
snack | +0.011 |
patch | +0.011 |
Neg logit contributions | |
spe | -0.018 |
overed | -0.018 |
led | -0.018 |
otte | -0.017 |
ls | -0.017 |
Token | There |
Token ID | 1318 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
bisc | +0.009 |
wake | +0.008 |
count | +0.008 |
snack | +0.008 |
patch | +0.008 |
Neg logit contributions | |
spe | -0.008 |
overed | -0.008 |
led | -0.007 |
ls | -0.007 |
otte | -0.007 |
Token | are |
Token ID | 389 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
bisc | +0.025 |
wake | +0.019 |
snack | +0.018 |
count | +0.018 |
patch | +0.018 |
Neg logit contributions | |
ls | -0.025 |
spe | -0.025 |
led | -0.023 |
overed | -0.022 |
ize | -0.021 |
Token | plenty |
Token ID | 6088 |
Feature activation | -4.65e+00 |
Pos logit contributions | |
bisc | +0.029 |
wake | +0.025 |
snack | +0.024 |
count | +0.024 |
patch | +0.023 |
Neg logit contributions | |
spe | -0.027 |
ls | -0.026 |
led | -0.025 |
overed | -0.025 |
rily | -0.024 |
Token | for |
Token ID | 329 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
bisc | +0.122 |
count | +0.108 |
wake | +0.107 |
nose | +0.106 |
snack | +0.102 |
Neg logit contributions | |
ls | -0.100 |
led | -0.095 |
spe | -0.089 |
overed | -0.087 |
han | -0.085 |
Token | both |
Token ID | 1111 |
Feature activation | -2.58e+00 |
Pos logit contributions | |
spe | +0.021 |
ls | +0.020 |
overed | +0.020 |
led | +0.020 |
otte | +0.020 |
Neg logit contributions | |
bisc | -0.032 |
wake | -0.028 |
count | -0.027 |
snack | -0.027 |
patch | -0.027 |
Token | of |
Token ID | 286 |
Feature activation | -1.04e-01 |
Pos logit contributions | |
bisc | +0.058 |
wake | +0.050 |
count | +0.049 |
snack | +0.047 |
patch | +0.047 |
Neg logit contributions | |
spe | -0.067 |
ls | -0.064 |
led | -0.063 |
overed | -0.062 |
rily | -0.061 |
Token | us |
Token ID | 514 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
bisc | +0.002 |
wake | +0.002 |
snack | +0.002 |
count | +0.002 |
patch | +0.002 |
Neg logit contributions | |
spe | -0.003 |
overed | -0.003 |
led | -0.002 |
ls | -0.002 |
otte | -0.002 |
Token | ," |
Token ID | 553 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
bisc | +0.003 |
wake | +0.002 |
snack | +0.002 |
count | +0.002 |
patch | +0.002 |
Neg logit contributions | |
spe | -0.002 |
ls | -0.002 |
led | -0.002 |
overed | -0.002 |
otte | -0.002 |
Token |
|
Token ID | 198 |
Feature activation | -7.42e-01 |
Pos logit contributions | |
bisc | +0.006 |
wake | +0.005 |
count | +0.005 |
snack | +0.005 |
patch | +0.005 |
Neg logit contributions | |
spe | -0.010 |
led | -0.009 |
overed | -0.009 |
ls | -0.009 |
otte | -0.009 |
Token |
|
Token ID | 198 |
Feature activation | -4.99e-01 |
Pos logit contributions | |
bisc | +0.016 |
wake | +0.013 |
count | +0.013 |
snack | +0.013 |
patch | +0.013 |
Neg logit contributions | |
spe | -0.020 |
led | -0.020 |
overed | -0.020 |
ls | -0.019 |
otte | -0.019 |
Token | " |
Token ID | 1 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
bisc | +0.009 |
wake | +0.007 |
count | +0.007 |
patch | +0.007 |
snack | +0.007 |
Neg logit contributions | |
spe | -0.015 |
led | -0.015 |
overed | -0.015 |
ls | -0.015 |
rily | -0.014 |
Token | That |
Token ID | 2504 |
Feature activation | -7.89e-02 |
Pos logit contributions | |
bisc | +0.004 |
wake | +0.003 |
strip | +0.003 |
count | +0.003 |
Be | +0.003 |
Neg logit contributions | |
ls | -0.005 |
led | -0.005 |
rily | -0.005 |
overed | -0.005 |
sadly | -0.005 |
Token | is |
Token ID | 318 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
bisc | +0.002 |
count | +0.001 |
snack | +0.001 |
wake | +0.001 |
patch | +0.001 |
Neg logit contributions | |
spe | -0.002 |
led | -0.002 |
ls | -0.002 |
overed | -0.002 |
rily | -0.002 |
Token | called |
Token ID | 1444 |
Feature activation | -4.63e+00 |
Pos logit contributions | |
bisc | +0.037 |
wake | +0.032 |
count | +0.031 |
snack | +0.031 |
patch | +0.031 |
Neg logit contributions | |
spe | -0.032 |
overed | -0.030 |
led | -0.030 |
ls | -0.030 |
otte | -0.029 |
Token | a |
Token ID | 257 |
Feature activation | -9.99e-01 |
Pos logit contributions | |
bisc | +0.111 |
wake | +0.096 |
count | +0.095 |
snack | +0.094 |
patch | +0.092 |
Neg logit contributions | |
spe | -0.110 |
overed | -0.107 |
led | -0.105 |
ls | -0.105 |
otte | -0.102 |
Token | cell |
Token ID | 2685 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
bisc | +0.021 |
count | +0.019 |
iver | +0.018 |
snack | +0.018 |
gloves | +0.018 |
Neg logit contributions | |
spe | -0.024 |
otte | -0.021 |
han | -0.021 |
iana | -0.021 |
ls | -0.021 |
Token | ," |
Token ID | 553 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
bisc | +0.033 |
wake | +0.028 |
count | +0.028 |
snack | +0.028 |
patch | +0.027 |
Neg logit contributions | |
spe | -0.036 |
overed | -0.035 |
led | -0.035 |
ls | -0.034 |
otte | -0.033 |
Token | his |
Token ID | 465 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
spe | +0.028 |
overed | +0.028 |
ivating | +0.026 |
led | +0.025 |
ls | +0.025 |
Neg logit contributions | |
bisc | -0.025 |
patch | -0.022 |
nose | -0.022 |
wake | -0.022 |
snack | -0.021 |
Token | mum |
Token ID | 25682 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
spe | +0.033 |
overed | +0.033 |
led | +0.032 |
ls | +0.032 |
otte | +0.031 |
Neg logit contributions | |
bisc | -0.030 |
wake | -0.026 |
count | -0.026 |
snack | -0.025 |
patch | -0.024 |
Token | around |
Token ID | 1088 |
Feature activation | +5.85e-01 |
Pos logit contributions | |
bisc | +0.039 |
wake | +0.034 |
count | +0.033 |
snack | +0.033 |
patch | +0.032 |
Neg logit contributions | |
spe | -0.044 |
overed | -0.043 |
led | -0.042 |
otte | -0.042 |
ls | -0.041 |
Token | the |
Token ID | 262 |
Feature activation | -7.33e-01 |
Pos logit contributions | |
spe | +0.013 |
overed | +0.013 |
led | +0.013 |
ls | +0.013 |
otte | +0.012 |
Neg logit contributions | |
bisc | -0.015 |
wake | -0.013 |
count | -0.013 |
snack | -0.013 |
patch | -0.013 |
Token | symbol |
Token ID | 6194 |
Feature activation | +7.00e-01 |
Pos logit contributions | |
bisc | +0.015 |
wake | +0.014 |
count | +0.013 |
snack | +0.013 |
patch | +0.013 |
Neg logit contributions | |
spe | -0.020 |
overed | -0.020 |
led | -0.019 |
ls | -0.019 |
otte | -0.019 |
Token | . |
Token ID | 13 |
Feature activation | -2.48e-01 |
Pos logit contributions | |
spe | +0.017 |
overed | +0.017 |
led | +0.017 |
ls | +0.016 |
otte | +0.016 |
Neg logit contributions | |
bisc | -0.017 |
wake | -0.014 |
count | -0.014 |
snack | -0.014 |
patch | -0.014 |
Token | " |
Token ID | 366 |
Feature activation | +5.82e-01 |
Pos logit contributions | |
bisc | +0.006 |
snack | +0.005 |
wake | +0.005 |
nose | +0.005 |
pile | +0.005 |
Neg logit contributions | |
spe | -0.006 |
led | -0.006 |
ls | -0.006 |
otte | -0.006 |
overed | -0.005 |
Token | Br |
Token ID | 9414 |
Feature activation | -4.63e+00 |
Pos logit contributions | |
spe | +0.014 |
overed | +0.014 |
led | +0.014 |
ls | +0.014 |
rily | +0.013 |
Neg logit contributions | |
bisc | -0.014 |
wake | -0.012 |
count | -0.012 |
snack | -0.012 |
patch | -0.011 |
Token | rr |
Token ID | 21062 |
Feature activation | -1.75e+00 |
Pos logit contributions | |
bisc | +0.120 |
coat | +0.099 |
wake | +0.099 |
snack | +0.098 |
carrot | +0.098 |
Neg logit contributions | |
overed | -0.108 |
han | -0.106 |
spe | -0.102 |
led | -0.102 |
narrative | -0.101 |
Token | , |
Token ID | 11 |
Feature activation | -4.41e-01 |
Pos logit contributions | |
bisc | +0.047 |
snack | +0.039 |
gloves | +0.038 |
coat | +0.038 |
patch | +0.037 |
Neg logit contributions | |
spe | -0.040 |
ls | -0.039 |
led | -0.037 |
ize | -0.035 |
overed | -0.035 |
Token | it |
Token ID | 340 |
Feature activation | -7.00e-01 |
Pos logit contributions | |
bisc | +0.012 |
wake | +0.011 |
count | +0.011 |
snack | +0.011 |
patch | +0.010 |
Neg logit contributions | |
spe | -0.009 |
overed | -0.009 |
led | -0.009 |
ls | -0.009 |
otte | -0.008 |
Token | 's |
Token ID | 338 |
Feature activation | -2.33e+00 |
Pos logit contributions | |
bisc | +0.012 |
wake | +0.011 |
count | +0.011 |
patch | +0.010 |
snack | +0.010 |
Neg logit contributions | |
spe | -0.021 |
overed | -0.021 |
otte | -0.020 |
rily | -0.020 |
han | -0.020 |
Token | so |
Token ID | 523 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
bisc | +0.060 |
wake | +0.053 |
count | +0.052 |
snack | +0.051 |
patch | +0.050 |
Neg logit contributions | |
spe | -0.053 |
overed | -0.051 |
led | -0.051 |
ls | -0.050 |
otte | -0.048 |
Token | . |
Token ID | 13 |
Feature activation | -6.79e-01 |
Pos logit contributions | |
bisc | +0.028 |
wake | +0.024 |
count | +0.024 |
snack | +0.024 |
patch | +0.023 |
Neg logit contributions | |
spe | -0.030 |
overed | -0.029 |
led | -0.028 |
ls | -0.028 |
otte | -0.027 |
Token | It |
Token ID | 632 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
bisc | +0.020 |
snack | +0.018 |
patch | +0.018 |
wake | +0.017 |
count | +0.017 |
Neg logit contributions | |
overed | -0.011 |
rily | -0.011 |
spe | -0.011 |
ivating | -0.011 |
ls | -0.011 |
Token | lives |
Token ID | 3160 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
bisc | +0.026 |
wake | +0.023 |
count | +0.022 |
patch | +0.021 |
snack | +0.021 |
Neg logit contributions | |
spe | -0.038 |
overed | -0.037 |
led | -0.036 |
otte | -0.036 |
han | -0.035 |
Token | in |
Token ID | 287 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
bisc | +0.031 |
wake | +0.030 |
snack | +0.029 |
patch | +0.029 |
count | +0.029 |
Neg logit contributions | |
spe | -0.022 |
overed | -0.021 |
han | -0.020 |
ivating | -0.019 |
led | -0.019 |
Token | the |
Token ID | 262 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
bisc | +0.009 |
wake | +0.008 |
count | +0.008 |
snack | +0.008 |
nose | +0.007 |
Neg logit contributions | |
spe | -0.005 |
ls | -0.004 |
rily | -0.004 |
led | -0.004 |
otte | -0.004 |
Token | hot |
Token ID | 3024 |
Feature activation | -4.61e+00 |
Pos logit contributions | |
bisc | +0.023 |
count | +0.021 |
snack | +0.019 |
're | +0.018 |
wake | +0.018 |
Neg logit contributions | |
ls | -0.023 |
spe | -0.023 |
narrative | -0.023 |
led | -0.022 |
iana | -0.022 |
Token | and |
Token ID | 290 |
Feature activation | -1.87e+00 |
Pos logit contributions | |
bisc | +0.104 |
count | +0.098 |
snack | +0.084 |
're | +0.084 |
icious | +0.084 |
Neg logit contributions | |
ls | -0.105 |
led | -0.104 |
han | -0.096 |
ized | -0.095 |
ize | -0.092 |
Token | dry |
Token ID | 5894 |
Feature activation | -2.45e+00 |
Pos logit contributions | |
bisc | +0.043 |
wake | +0.037 |
count | +0.036 |
snack | +0.036 |
patch | +0.035 |
Neg logit contributions | |
spe | -0.048 |
overed | -0.046 |
led | -0.046 |
ls | -0.045 |
otte | -0.044 |
Token | places |
Token ID | 4113 |
Feature activation | +5.25e-01 |
Pos logit contributions | |
bisc | +0.056 |
count | +0.049 |
snack | +0.046 |
wake | +0.045 |
chop | +0.042 |
Neg logit contributions | |
led | -0.062 |
ls | -0.061 |
rily | -0.060 |
spe | -0.060 |
overed | -0.059 |
Token | . |
Token ID | 13 |
Feature activation | -7.55e-01 |
Pos logit contributions | |
spe | +0.014 |
overed | +0.014 |
led | +0.013 |
ls | +0.013 |
otte | +0.013 |
Neg logit contributions | |
bisc | -0.011 |
wake | -0.010 |
count | -0.010 |
snack | -0.009 |
patch | -0.009 |
Token | It |
Token ID | 632 |
Feature activation | -9.20e-01 |
Pos logit contributions | |
bisc | +0.022 |
snack | +0.020 |
patch | +0.019 |
pile | +0.019 |
wake | +0.018 |
Neg logit contributions | |
overed | -0.013 |
ivating | -0.012 |
spe | -0.012 |
rily | -0.012 |
edly | -0.012 |
Token | man |
Token ID | 805 |
Feature activation | +5.07e-01 |
Pos logit contributions | |
spe | +0.007 |
otte | +0.007 |
overed | +0.006 |
ivating | +0.006 |
ls | +0.006 |
Neg logit contributions | |
bisc | -0.006 |
snack | -0.006 |
wake | -0.005 |
nose | -0.005 |
patch | -0.005 |
Token | who |
Token ID | 508 |
Feature activation | +5.72e-01 |
Pos logit contributions | |
spe | +0.012 |
overed | +0.011 |
led | +0.011 |
ls | +0.011 |
otte | +0.011 |
Neg logit contributions | |
bisc | -0.013 |
wake | -0.011 |
count | -0.011 |
snack | -0.011 |
patch | -0.011 |
Token | said |
Token ID | 531 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
overed | +0.015 |
spe | +0.015 |
han | +0.014 |
rily | +0.014 |
ls | +0.014 |
Neg logit contributions | |
bisc | -0.013 |
wake | -0.011 |
count | -0.011 |
snack | -0.011 |
patch | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | -7.32e-03 |
Pos logit contributions | |
bisc | +0.035 |
wake | +0.031 |
count | +0.030 |
snack | +0.030 |
patch | +0.029 |
Neg logit contributions | |
spe | -0.031 |
overed | -0.030 |
led | -0.029 |
ls | -0.029 |
otte | -0.028 |
Token | " |
Token ID | 366 |
Feature activation | +9.49e-01 |
Pos logit contributions | |
bisc | +0.000 |
wake | +0.000 |
count | +0.000 |
snack | +0.000 |
patch | +0.000 |
Neg logit contributions | |
spe | -0.000 |
overed | -0.000 |
led | -0.000 |
ls | -0.000 |
otte | -0.000 |
Token | Br |
Token ID | 9414 |
Feature activation | -4.61e+00 |
Pos logit contributions | |
spe | +0.021 |
overed | +0.020 |
led | +0.020 |
ls | +0.020 |
otte | +0.019 |
Neg logit contributions | |
bisc | -0.025 |
wake | -0.022 |
count | -0.022 |
snack | -0.022 |
patch | -0.021 |
Token | rr |
Token ID | 21062 |
Feature activation | -2.35e+00 |
Pos logit contributions | |
bisc | +0.115 |
wake | +0.096 |
snack | +0.094 |
pile | +0.092 |
patch | +0.092 |
Neg logit contributions | |
overed | -0.107 |
han | -0.107 |
led | -0.105 |
spe | -0.104 |
narrative | -0.101 |
Token | , |
Token ID | 11 |
Feature activation | -7.82e-01 |
Pos logit contributions | |
bisc | +0.060 |
patch | +0.051 |
snack | +0.051 |
wake | +0.050 |
shade | +0.048 |
Neg logit contributions | |
spe | -0.053 |
led | -0.050 |
ls | -0.048 |
han | -0.045 |
overed | -0.045 |
Token | it |
Token ID | 340 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
bisc | +0.019 |
wake | +0.017 |
count | +0.017 |
snack | +0.017 |
patch | +0.016 |
Neg logit contributions | |
spe | -0.018 |
overed | -0.018 |
led | -0.018 |
ls | -0.017 |
otte | -0.017 |
Token | 's |
Token ID | 338 |
Feature activation | -2.78e+00 |
Pos logit contributions | |
bisc | +0.028 |
snack | +0.023 |
wake | +0.022 |
count | +0.022 |
patch | +0.022 |
Neg logit contributions | |
led | -0.033 |
spe | -0.033 |
ls | -0.032 |
overed | -0.031 |
rily | -0.030 |
Token | so |
Token ID | 523 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
bisc | +0.066 |
wake | +0.056 |
count | +0.055 |
patch | +0.054 |
snack | +0.053 |
Neg logit contributions | |
spe | -0.062 |
led | -0.060 |
rily | -0.059 |
overed | -0.059 |
ls | -0.059 |
Token | temper |
Token ID | 4124 |
Feature activation | -8.12e-01 |
Pos logit contributions | |
spe | +0.078 |
overed | +0.077 |
otte | +0.075 |
led | +0.075 |
ls | +0.075 |
Neg logit contributions | |
bisc | -0.049 |
wake | -0.043 |
snack | -0.041 |
count | -0.040 |
patch | -0.040 |
Token | and |
Token ID | 290 |
Feature activation | +2.80e-01 |
Pos logit contributions | |
bisc | +0.021 |
wake | +0.018 |
count | +0.018 |
snack | +0.018 |
patch | +0.017 |
Neg logit contributions | |
spe | -0.019 |
overed | -0.018 |
led | -0.018 |
ls | -0.017 |
otte | -0.017 |
Token | often |
Token ID | 1690 |
Feature activation | +8.60e-01 |
Pos logit contributions | |
overed | +0.007 |
ls | +0.007 |
spe | +0.007 |
rily | +0.007 |
ivating | +0.007 |
Neg logit contributions | |
bisc | -0.007 |
count | -0.006 |
wake | -0.006 |
snack | -0.006 |
nose | -0.006 |
Token | threw |
Token ID | 9617 |
Feature activation | -7.44e-01 |
Pos logit contributions | |
spe | +0.021 |
led | +0.020 |
overed | +0.020 |
ls | +0.020 |
otte | +0.019 |
Neg logit contributions | |
bisc | -0.021 |
wake | -0.017 |
snack | -0.017 |
count | -0.017 |
patch | -0.017 |
Token | temper |
Token ID | 4124 |
Feature activation | -8.79e-01 |
Pos logit contributions | |
bisc | +0.019 |
snack | +0.016 |
wake | +0.016 |
count | +0.016 |
patch | +0.015 |
Neg logit contributions | |
spe | -0.018 |
overed | -0.018 |
otte | -0.017 |
led | -0.017 |
ls | -0.016 |
Token | tant |
Token ID | 24246 |
Feature activation | +4.28e+00 |
Pos logit contributions | |
bisc | +0.023 |
wake | +0.020 |
snack | +0.020 |
count | +0.020 |
patch | +0.019 |
Neg logit contributions | |
overed | -0.019 |
spe | -0.019 |
otte | -0.019 |
Acc | -0.018 |
rily | -0.017 |
Token | rums |
Token ID | 45241 |
Feature activation | +1.99e-01 |
Pos logit contributions | |
spe | +0.109 |
overed | +0.105 |
ls | +0.102 |
otte | +0.100 |
led | +0.100 |
Neg logit contributions | |
bisc | -0.091 |
count | -0.082 |
wake | -0.079 |
snack | -0.077 |
chop | -0.075 |
Token | . |
Token ID | 13 |
Feature activation | -2.19e-03 |
Pos logit contributions | |
spe | +0.004 |
led | +0.004 |
overed | +0.004 |
rily | +0.004 |
otte | +0.004 |
Neg logit contributions | |
bisc | -0.006 |
wake | -0.005 |
snack | -0.005 |
count | -0.005 |
patch | -0.005 |
Token | One |
Token ID | 1881 |
Feature activation | -3.06e+00 |
Pos logit contributions | |
bisc | +0.000 |
wake | +0.000 |
count | +0.000 |
snack | +0.000 |
patch | +0.000 |
Neg logit contributions | |
spe | -0.000 |
led | -0.000 |
overed | -0.000 |
ls | -0.000 |
otte | -0.000 |
Token | day |
Token ID | 1110 |
Feature activation | +9.06e-02 |
Pos logit contributions | |
bisc | +0.068 |
count | +0.057 |
wake | +0.056 |
snack | +0.055 |
patch | +0.053 |
Neg logit contributions | |
spe | -0.082 |
led | -0.077 |
overed | -0.075 |
ls | -0.075 |
otte | -0.073 |
Token | , |
Token ID | 11 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
spe | +0.002 |
overed | +0.002 |
led | +0.002 |
ls | +0.002 |
otte | +0.002 |
Neg logit contributions | |
bisc | -0.002 |
wake | -0.002 |
count | -0.002 |
snack | -0.002 |
patch | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
bisc | +0.025 |
snack | +0.022 |
wake | +0.021 |
count | +0.020 |
snow | +0.020 |
Neg logit contributions | |
spe | -0.043 |
overed | -0.041 |
ls | -0.041 |
led | -0.040 |
otte | -0.039 |
Token | boy |
Token ID | 2933 |
Feature activation | -9.61e-01 |
Pos logit contributions | |
bisc | +0.004 |
count | +0.003 |
wake | +0.003 |
snack | +0.003 |
're | +0.003 |
Neg logit contributions | |
spe | -0.005 |
ls | -0.004 |
otte | -0.004 |
overed | -0.004 |
led | -0.004 |
Token | who |
Token ID | 508 |
Feature activation | -3.04e-01 |
Pos logit contributions | |
bisc | +0.027 |
wake | +0.025 |
count | +0.024 |
snack | +0.024 |
patch | +0.024 |
Neg logit contributions | |
spe | -0.019 |
overed | -0.019 |
led | -0.018 |
ls | -0.018 |
rily | -0.018 |
Token | loved |
Token ID | 6151 |
Feature activation | +1.51e+00 |
Pos logit contributions | |
bisc | +0.009 |
wake | +0.008 |
count | +0.008 |
snack | +0.008 |
patch | +0.008 |
Neg logit contributions | |
spe | -0.006 |
led | -0.005 |
overed | -0.005 |
ls | -0.005 |
otte | -0.005 |
Token | to |
Token ID | 284 |
Feature activation | -2.05e-01 |
Pos logit contributions | |
spe | +0.032 |
overed | +0.030 |
led | +0.030 |
ls | +0.029 |
otte | +0.029 |
Neg logit contributions | |
bisc | -0.041 |
wake | -0.036 |
count | -0.036 |
snack | -0.035 |
patch | -0.035 |
Token | solve |
Token ID | 8494 |
Feature activation | +4.40e+00 |
Pos logit contributions | |
bisc | +0.004 |
est | +0.003 |
icious | +0.003 |
ch | +0.003 |
itten | +0.003 |
Neg logit contributions | |
proudly | -0.004 |
trophy | -0.004 |
happily | -0.004 |
they | -0.004 |
yacht | -0.004 |
Token | puzzles |
Token ID | 24367 |
Feature activation | +3.13e+00 |
Pos logit contributions | |
overed | +0.125 |
spe | +0.118 |
Acc | +0.116 |
otte | +0.115 |
han | +0.114 |
Neg logit contributions | |
count | -0.083 |
bisc | -0.081 |
snack | -0.080 |
snow | -0.079 |
pile | -0.079 |
Token | . |
Token ID | 13 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
spe | +0.095 |
overed | +0.095 |
Acc | +0.094 |
otte | +0.094 |
led | +0.094 |
Neg logit contributions | |
bisc | -0.050 |
count | -0.048 |
snack | -0.045 |
pile | -0.045 |
come | -0.045 |
Token | Every |
Token ID | 3887 |
Feature activation | +4.26e-02 |
Pos logit contributions | |
spe | +0.001 |
From | +0.001 |
Since | +0.001 |
Their | +0.001 |
Robert | +0.001 |
Neg logit contributions | |
bisc | -0.004 |
wake | -0.004 |
tissue | -0.004 |
patch | -0.004 |
snack | -0.004 |
Token | day |
Token ID | 1110 |
Feature activation | +1.44e+00 |
Pos logit contributions | |
spe | +0.001 |
overed | +0.001 |
led | +0.001 |
sadly | +0.001 |
ls | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
count | -0.001 |
wake | -0.001 |
nose | -0.001 |
chop | -0.001 |
Token | , |
Token ID | 11 |
Feature activation | +9.46e-01 |
Pos logit contributions | |
ls | +0.031 |
spe | +0.030 |
led | +0.029 |
overed | +0.029 |
sadly | +0.027 |
Neg logit contributions | |
bisc | -0.039 |
wake | -0.033 |
count | -0.032 |
patch | -0.032 |
iver | -0.032 |
Token | that |
Token ID | 326 |
Feature activation | -9.07e-01 |
Pos logit contributions | |
come | +0.001 |
wake | +0.001 |
bisc | +0.001 |
snack | +0.001 |
go | +0.001 |
Neg logit contributions | |
overed | -0.001 |
spe | -0.001 |
led | -0.001 |
narrative | -0.001 |
ls | -0.001 |
Token | day |
Token ID | 1110 |
Feature activation | +1.47e+00 |
Pos logit contributions | |
bisc | +0.021 |
wake | +0.019 |
count | +0.018 |
snack | +0.018 |
patch | +0.018 |
Neg logit contributions | |
spe | -0.023 |
overed | -0.022 |
led | -0.022 |
ls | -0.022 |
otte | -0.021 |
Token | and |
Token ID | 290 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
spe | +0.036 |
ls | +0.034 |
overed | +0.034 |
led | +0.034 |
otte | +0.034 |
Neg logit contributions | |
bisc | -0.035 |
wake | -0.031 |
snack | -0.030 |
count | -0.030 |
patch | -0.029 |
Token | she |
Token ID | 673 |
Feature activation | +4.21e-01 |
Pos logit contributions | |
ivating | +0.032 |
overed | +0.032 |
rily | +0.031 |
ls | +0.030 |
han | +0.030 |
Neg logit contributions | |
wake | -0.023 |
come | -0.023 |
bisc | -0.022 |
go | -0.022 |
patch | -0.022 |
Token | always |
Token ID | 1464 |
Feature activation | +8.71e-01 |
Pos logit contributions | |
overed | +0.012 |
rily | +0.011 |
ivating | +0.011 |
ls | +0.011 |
han | +0.011 |
Neg logit contributions | |
bisc | -0.009 |
wake | -0.008 |
count | -0.008 |
come | -0.008 |
patch | -0.008 |
Token | tre |
Token ID | 2054 |
Feature activation | +4.78e+00 |
Pos logit contributions | |
led | +0.018 |
spe | +0.018 |
otte | +0.017 |
sadly | +0.016 |
overed | +0.016 |
Neg logit contributions | |
bisc | -0.024 |
snack | -0.021 |
wake | -0.020 |
count | -0.019 |
strip | -0.019 |
Token | asured |
Token ID | 34006 |
Feature activation | +1.98e+00 |
Pos logit contributions | |
ls | +0.101 |
spe | +0.099 |
overed | +0.095 |
led | +0.094 |
ivating | +0.089 |
Neg logit contributions | |
bisc | -0.126 |
count | -0.120 |
wake | -0.115 |
patch | -0.111 |
pile | -0.109 |
Token | her |
Token ID | 607 |
Feature activation | +3.47e-01 |
Pos logit contributions | |
narrative | +0.063 |
ivating | +0.063 |
spe | +0.062 |
overed | +0.060 |
ls | +0.059 |
Neg logit contributions | |
coat | -0.029 |
come | -0.028 |
help | -0.027 |
snack | -0.027 |
count | -0.027 |
Token | toy |
Token ID | 13373 |
Feature activation | +9.66e-01 |
Pos logit contributions | |
overed | +0.010 |
spe | +0.010 |
led | +0.010 |
ls | +0.009 |
ivating | +0.009 |
Neg logit contributions | |
bisc | -0.007 |
snack | -0.006 |
wake | -0.006 |
count | -0.006 |
nose | -0.006 |
Token | car |
Token ID | 1097 |
Feature activation | +8.93e-01 |
Pos logit contributions | |
overed | +0.031 |
spe | +0.029 |
led | +0.029 |
Acc | +0.029 |
ivating | +0.028 |
Neg logit contributions | |
bisc | -0.016 |
coat | -0.015 |
count | -0.014 |
pile | -0.014 |
snack | -0.014 |
Token | made |
Token ID | 925 |
Feature activation | -2.57e-01 |
Pos logit contributions | |
overed | +0.023 |
spe | +0.022 |
ls | +0.022 |
led | +0.022 |
otte | +0.021 |
Neg logit contributions | |
bisc | -0.020 |
count | -0.018 |
wake | -0.018 |
snack | -0.018 |
patch | -0.017 |
Token | there |
Token ID | 612 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
spe | +0.008 |
overed | +0.007 |
han | +0.007 |
otte | +0.007 |
edly | +0.007 |
Neg logit contributions | |
bisc | -0.007 |
wake | -0.006 |
snack | -0.006 |
count | -0.006 |
go | -0.006 |
Token | was |
Token ID | 373 |
Feature activation | -1.07e+00 |
Pos logit contributions | |
bisc | +0.004 |
wake | +0.003 |
count | +0.003 |
snack | +0.003 |
patch | +0.003 |
Neg logit contributions | |
spe | -0.002 |
overed | -0.002 |
led | -0.002 |
ls | -0.002 |
sadly | -0.002 |
Token | a |
Token ID | 257 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
bisc | +0.021 |
snack | +0.018 |
wake | +0.017 |
count | +0.016 |
strip | +0.016 |
Neg logit contributions | |
spe | -0.034 |
overed | -0.031 |
ls | -0.031 |
led | -0.031 |
otte | -0.030 |
Token | three |
Token ID | 1115 |
Feature activation | +6.06e-01 |
Pos logit contributions | |
spe | +0.005 |
ls | +0.005 |
otte | +0.005 |
overed | +0.004 |
led | +0.004 |
Neg logit contributions | |
bisc | -0.004 |
count | -0.003 |
wake | -0.003 |
snack | -0.003 |
're | -0.003 |
Token | - |
Token ID | 12 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
spe | +0.014 |
overed | +0.013 |
led | +0.013 |
ls | +0.013 |
otte | +0.013 |
Neg logit contributions | |
bisc | -0.016 |
wake | -0.014 |
count | -0.014 |
snack | -0.013 |
patch | -0.013 |
Token | year |
Token ID | 1941 |
Feature activation | +4.28e+00 |
Pos logit contributions | |
overed | +0.017 |
rily | +0.015 |
ls | +0.015 |
led | +0.015 |
spe | +0.015 |
Neg logit contributions | |
bisc | -0.035 |
wake | -0.033 |
count | -0.032 |
chop | -0.031 |
snack | -0.031 |
Token | - |
Token ID | 12 |
Feature activation | +1.39e+00 |
Pos logit contributions | |
overed | +0.113 |
spe | +0.111 |
Acc | +0.110 |
ivating | +0.110 |
otte | +0.109 |
Neg logit contributions | |
bisc | -0.091 |
wake | -0.083 |
patch | -0.078 |
snack | -0.078 |
nose | -0.078 |
Token | old |
Token ID | 727 |
Feature activation | +2.35e+00 |
Pos logit contributions | |
overed | +0.034 |
rily | +0.032 |
spe | +0.031 |
led | +0.031 |
ivating | +0.030 |
Neg logit contributions | |
bisc | -0.034 |
wake | -0.029 |
count | -0.029 |
snack | -0.029 |
rest | -0.028 |
Token | girl |
Token ID | 2576 |
Feature activation | -5.54e-01 |
Pos logit contributions | |
overed | +0.085 |
ivating | +0.081 |
spe | +0.078 |
ized | +0.076 |
led | +0.076 |
Neg logit contributions | |
bisc | -0.036 |
wake | -0.030 |
coat | -0.029 |
nose | -0.029 |
paw | -0.028 |
Token | called |
Token ID | 1444 |
Feature activation | -1.88e+00 |
Pos logit contributions | |
bisc | +0.015 |
wake | +0.014 |
count | +0.013 |
snack | +0.013 |
patch | +0.013 |
Neg logit contributions | |
overed | -0.012 |
spe | -0.012 |
led | -0.011 |
ls | -0.011 |
rily | -0.011 |
Token | Joy |
Token ID | 14087 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
bisc | +0.052 |
wake | +0.047 |
count | +0.047 |
snack | +0.046 |
patch | +0.044 |
Neg logit contributions | |
spe | -0.039 |
led | -0.038 |
overed | -0.038 |
ls | -0.037 |
rily | -0.036 |
Token | he |
Token ID | 339 |
Feature activation | +5.80e-01 |
Pos logit contributions | |
led | +0.018 |
overed | +0.018 |
spe | +0.018 |
ivating | +0.017 |
ls | +0.017 |
Neg logit contributions | |
bisc | -0.018 |
patch | -0.018 |
come | -0.018 |
coat | -0.018 |
snack | -0.017 |
Token | could |
Token ID | 714 |
Feature activation | +3.26e-01 |
Pos logit contributions | |
overed | +0.015 |
spe | +0.015 |
ls | +0.015 |
rily | +0.015 |
otte | +0.015 |
Neg logit contributions | |
bisc | -0.012 |
wake | -0.011 |
count | -0.011 |
come | -0.011 |
snack | -0.011 |
Token | do |
Token ID | 466 |
Feature activation | +3.16e+00 |
Pos logit contributions | |
spe | +0.011 |
overed | +0.010 |
led | +0.010 |
ls | +0.010 |
otte | +0.010 |
Neg logit contributions | |
bisc | -0.005 |
wake | -0.004 |
count | -0.004 |
snack | -0.004 |
patch | -0.004 |
Token | anything |
Token ID | 1997 |
Feature activation | +1.89e+00 |
Pos logit contributions | |
spe | +0.087 |
overed | +0.086 |
otte | +0.084 |
han | +0.084 |
ls | +0.083 |
Neg logit contributions | |
count | -0.060 |
bisc | -0.059 |
snack | -0.058 |
wake | -0.057 |
come | -0.057 |
Token | if |
Token ID | 611 |
Feature activation | -5.36e-01 |
Pos logit contributions | |
overed | +0.050 |
otte | +0.048 |
han | +0.047 |
spe | +0.047 |
rily | +0.046 |
Neg logit contributions | |
bisc | -0.039 |
wake | -0.038 |
count | -0.038 |
come | -0.037 |
snack | -0.036 |
Token | he |
Token ID | 258 |
Feature activation | +6.03e+00 |
Pos logit contributions | |
bisc | +0.014 |
snack | +0.013 |
wake | +0.013 |
count | +0.013 |
patch | +0.012 |
Neg logit contributions | |
overed | -0.012 |
spe | -0.011 |
led | -0.011 |
ivating | -0.011 |
rily | -0.011 |
Token | set |
Token ID | 900 |
Feature activation | +5.65e-01 |
Pos logit contributions | |
spe | +0.149 |
overed | +0.148 |
ivating | +0.147 |
ls | +0.146 |
Acc | +0.142 |
Neg logit contributions | |
bisc | -0.137 |
snack | -0.126 |
wake | -0.125 |
patch | -0.122 |
nose | -0.121 |
Token | his |
Token ID | 465 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
spe | +0.013 |
overed | +0.013 |
led | +0.012 |
otte | +0.012 |
Acc | +0.012 |
Neg logit contributions | |
bisc | -0.014 |
wake | -0.013 |
snack | -0.013 |
count | -0.013 |
come | -0.012 |
Token | mind |
Token ID | 2000 |
Feature activation | +9.58e-01 |
Pos logit contributions | |
spe | +0.034 |
overed | +0.031 |
ls | +0.030 |
sadly | +0.030 |
led | +0.030 |
Neg logit contributions | |
bisc | -0.026 |
wake | -0.021 |
're | -0.020 |
patch | -0.020 |
count | -0.019 |
Token | to |
Token ID | 284 |
Feature activation | -3.24e-01 |
Pos logit contributions | |
overed | +0.022 |
spe | +0.022 |
led | +0.022 |
ls | +0.021 |
otte | +0.021 |
Neg logit contributions | |
bisc | -0.023 |
wake | -0.021 |
count | -0.020 |
snack | -0.020 |
patch | -0.020 |
Token | it |
Token ID | 340 |
Feature activation | -3.86e-01 |
Pos logit contributions | |
bisc | +0.006 |
wake | +0.004 |
snack | +0.004 |
patch | +0.004 |
count | +0.004 |
Neg logit contributions | |
spe | -0.010 |
overed | -0.009 |
led | -0.009 |
otte | -0.009 |
ls | -0.009 |
Token | happened |
Token ID | 3022 |
Feature activation | +1.92e+00 |
Pos logit contributions | |
spe | +0.031 |
overed | +0.031 |
rily | +0.030 |
led | +0.030 |
otte | +0.029 |
Neg logit contributions | |
bisc | -0.031 |
wake | -0.027 |
snack | -0.026 |
count | -0.026 |
patch | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | +8.24e-02 |
Pos logit contributions | |
spe | +0.055 |
otte | +0.053 |
led | +0.053 |
overed | +0.053 |
ls | +0.052 |
Neg logit contributions | |
bisc | -0.037 |
wake | -0.033 |
count | -0.031 |
come | -0.031 |
snack | -0.030 |
Token | He |
Token ID | 679 |
Feature activation | +4.19e-01 |
Pos logit contributions | |
spe | +0.002 |
otte | +0.002 |
led | +0.002 |
overed | +0.002 |
han | +0.002 |
Neg logit contributions | |
bisc | -0.002 |
wake | -0.002 |
nose | -0.001 |
paw | -0.001 |
patch | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +4.19e-01 |
Pos logit contributions | |
spe | +0.012 |
overed | +0.012 |
ivating | +0.012 |
rily | +0.012 |
otte | +0.012 |
Neg logit contributions | |
bisc | -0.008 |
wake | -0.007 |
count | -0.007 |
patch | -0.007 |
pile | -0.006 |
Token | not |
Token ID | 407 |
Feature activation | +9.07e-02 |
Pos logit contributions | |
spe | +0.010 |
overed | +0.009 |
ls | +0.009 |
led | +0.009 |
otte | +0.009 |
Neg logit contributions | |
bisc | -0.011 |
wake | -0.010 |
count | -0.010 |
come | -0.010 |
paw | -0.009 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
led | +0.002 |
spe | +0.002 |
otte | +0.002 |
ls | +0.002 |
overed | +0.002 |
Neg logit contributions | |
bisc | -0.003 |
count | -0.002 |
snack | -0.002 |
wake | -0.002 |
patch | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +4.14e-02 |
Pos logit contributions | |
spe | +0.034 |
overed | +0.033 |
led | +0.033 |
ls | +0.032 |
otte | +0.032 |
Neg logit contributions | |
bisc | -0.032 |
wake | -0.028 |
count | -0.027 |
snack | -0.027 |
patch | -0.026 |
Token | He |
Token ID | 679 |
Feature activation | +8.26e-01 |
Pos logit contributions | |
spe | +0.001 |
led | +0.001 |
overed | +0.001 |
otte | +0.001 |
ls | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
wake | -0.001 |
count | -0.001 |
paw | -0.001 |
snack | -0.001 |
Token | sc |
Token ID | 629 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
overed | +0.025 |
spe | +0.025 |
rily | +0.025 |
ivating | +0.024 |
otte | +0.024 |
Neg logit contributions | |
bisc | -0.016 |
wake | -0.014 |
come | -0.013 |
count | -0.013 |
patch | -0.012 |
Token | olded |
Token ID | 48959 |
Feature activation | -9.15e-01 |
Pos logit contributions | |
spe | +0.032 |
overed | +0.030 |
led | +0.030 |
ls | +0.029 |
otte | +0.029 |
Neg logit contributions | |
bisc | -0.069 |
wake | -0.062 |
count | -0.062 |
snack | -0.061 |
patch | -0.060 |
Token | the |
Token ID | 262 |
Feature activation | -2.80e-01 |
Pos logit contributions | |
bisc | +0.025 |
wake | +0.022 |
count | +0.022 |
snack | +0.022 |
patch | +0.021 |
Neg logit contributions | |
spe | -0.020 |
overed | -0.019 |
led | -0.019 |
ls | -0.018 |
otte | -0.018 |
Token | showed |
Token ID | 3751 |
Feature activation | +7.62e-01 |
Pos logit contributions | |
spe | +0.021 |
overed | +0.020 |
led | +0.020 |
ls | +0.020 |
otte | +0.020 |
Neg logit contributions | |
bisc | -0.022 |
wake | -0.019 |
count | -0.018 |
snack | -0.018 |
patch | -0.018 |
Token | daddy |
Token ID | 35822 |
Feature activation | -3.83e-01 |
Pos logit contributions | |
spe | +0.017 |
overed | +0.017 |
led | +0.016 |
ls | +0.015 |
ivating | +0.015 |
Neg logit contributions | |
bisc | -0.019 |
wake | -0.018 |
count | -0.018 |
snack | -0.018 |
come | -0.017 |
Token | that |
Token ID | 326 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
bisc | +0.009 |
wake | +0.008 |
snack | +0.008 |
count | +0.008 |
come | +0.008 |
Neg logit contributions | |
overed | -0.009 |
spe | -0.009 |
led | -0.009 |
ivating | -0.009 |
ls | -0.009 |
Token | she |
Token ID | 673 |
Feature activation | +1.03e-01 |
Pos logit contributions | |
bisc | +0.044 |
wake | +0.040 |
count | +0.039 |
snack | +0.039 |
patch | +0.038 |
Neg logit contributions | |
spe | -0.021 |
overed | -0.020 |
led | -0.019 |
ls | -0.019 |
otte | -0.018 |
Token | could |
Token ID | 714 |
Feature activation | +6.14e-02 |
Pos logit contributions | |
overed | +0.003 |
spe | +0.003 |
rily | +0.003 |
han | +0.003 |
ls | +0.003 |
Neg logit contributions | |
bisc | -0.002 |
wake | -0.002 |
count | -0.002 |
come | -0.002 |
patch | -0.002 |
Token | do |
Token ID | 466 |
Feature activation | +3.14e+00 |
Pos logit contributions | |
spe | +0.002 |
led | +0.002 |
ls | +0.002 |
overed | +0.002 |
rily | +0.002 |
Neg logit contributions | |
bisc | -0.001 |
count | -0.001 |
wake | -0.001 |
snack | -0.001 |
patch | -0.001 |
Token | it |
Token ID | 340 |
Feature activation | +6.45e-03 |
Pos logit contributions | |
spe | +0.089 |
overed | +0.087 |
otte | +0.086 |
rily | +0.086 |
led | +0.085 |
Neg logit contributions | |
count | -0.058 |
bisc | -0.057 |
snack | -0.056 |
come | -0.056 |
wake | -0.055 |
Token | all |
Token ID | 477 |
Feature activation | +4.16e-02 |
Pos logit contributions | |
spe | +0.000 |
overed | +0.000 |
led | +0.000 |
ls | +0.000 |
otte | +0.000 |
Neg logit contributions | |
bisc | -0.000 |
wake | -0.000 |
snack | -0.000 |
count | -0.000 |
patch | -0.000 |
Token | by |
Token ID | 416 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
overed | +0.001 |
spe | +0.001 |
ls | +0.001 |
rily | +0.001 |
otte | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
count | -0.001 |
wake | -0.001 |
snack | -0.001 |
come | -0.001 |
Token | herself |
Token ID | 5223 |
Feature activation | +2.44e-01 |
Pos logit contributions | |
spe | +0.024 |
ls | +0.023 |
led | +0.023 |
overed | +0.023 |
rily | +0.021 |
Neg logit contributions | |
bisc | -0.034 |
count | -0.032 |
wake | -0.031 |
snack | -0.031 |
patch | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | -2.82e-01 |
Pos logit contributions | |
spe | +0.006 |
overed | +0.006 |
led | +0.006 |
ls | +0.006 |
otte | +0.006 |
Neg logit contributions | |
bisc | -0.006 |
wake | -0.005 |
count | -0.005 |
snack | -0.005 |
patch | -0.005 |
Token | leaf |
Token ID | 12835 |
Feature activation | -3.21e-01 |
Pos logit contributions | |
rily | +0.013 |
spe | +0.013 |
ls | +0.013 |
closely | +0.013 |
sadly | +0.013 |
Neg logit contributions | |
bisc | -0.009 |
count | -0.008 |
're | -0.008 |
wake | -0.008 |
est | -0.007 |
Token | and |
Token ID | 290 |
Feature activation | -1.44e-02 |
Pos logit contributions | |
bisc | +0.008 |
patch | +0.007 |
wake | +0.007 |
snack | +0.007 |
count | +0.007 |
Neg logit contributions | |
spe | -0.008 |
overed | -0.008 |
led | -0.008 |
ls | -0.008 |
otte | -0.007 |
Token | watched |
Token ID | 7342 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
bisc | +0.000 |
wake | +0.000 |
count | +0.000 |
snack | +0.000 |
patch | +0.000 |
Neg logit contributions | |
spe | -0.000 |
overed | -0.000 |
led | -0.000 |
ls | -0.000 |
otte | -0.000 |
Token | it |
Token ID | 340 |
Feature activation | -7.42e-01 |
Pos logit contributions | |
bisc | +0.026 |
wake | +0.025 |
snack | +0.024 |
count | +0.023 |
come | +0.023 |
Neg logit contributions | |
spe | -0.024 |
led | -0.021 |
otte | -0.021 |
han | -0.021 |
narrative | -0.021 |
Token | crawl |
Token ID | 27318 |
Feature activation | -2.31e-01 |
Pos logit contributions | |
bisc | +0.017 |
're | +0.013 |
flavor | +0.013 |
gloves | +0.012 |
itten | +0.012 |
Neg logit contributions | |
overed | -0.018 |
closely | -0.018 |
happily | -0.018 |
sadly | -0.018 |
ls | -0.018 |
Token | away |
Token ID | 1497 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
bisc | +0.006 |
come | +0.005 |
go | +0.005 |
patch | +0.005 |
snack | +0.005 |
Neg logit contributions | |
spe | -0.006 |
otte | -0.006 |
narrative | -0.006 |
issa | -0.005 |
Acc | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | +4.18e-02 |
Pos logit contributions | |
han | +0.041 |
led | +0.040 |
spe | +0.040 |
otte | +0.040 |
narrative | +0.039 |
Neg logit contributions | |
bisc | -0.024 |
wake | -0.023 |
patch | -0.022 |
snack | -0.022 |
count | -0.022 |
Token | Lily |
Token ID | 20037 |
Feature activation | +4.08e-01 |
Pos logit contributions | |
spe | +0.001 |
ls | +0.001 |
led | +0.001 |
han | +0.001 |
otte | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
wake | -0.001 |
snack | -0.001 |
count | -0.001 |
patch | -0.001 |
Token | felt |
Token ID | 2936 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
spe | +0.009 |
led | +0.009 |
overed | +0.009 |
ls | +0.008 |
otte | +0.008 |
Neg logit contributions | |
bisc | -0.011 |
wake | -0.009 |
snack | -0.009 |
count | -0.009 |
patch | -0.009 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
spe | +0.039 |
overed | +0.038 |
led | +0.037 |
ls | +0.037 |
otte | +0.036 |
Neg logit contributions | |
bisc | -0.040 |
wake | -0.036 |
count | -0.035 |
snack | -0.034 |
patch | -0.034 |
Token | knowing |
Token ID | 6970 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
spe | +0.040 |
otte | +0.036 |
led | +0.036 |
han | +0.036 |
ls | +0.036 |
Neg logit contributions | |
bisc | -0.042 |
wake | -0.040 |
snack | -0.039 |
patch | -0.038 |
come | -0.038 |
Token | 'd |
Token ID | 1549 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
spe | +0.033 |
overed | +0.033 |
ls | +0.031 |
led | +0.031 |
otte | +0.031 |
Neg logit contributions | |
bisc | -0.036 |
wake | -0.032 |
count | -0.031 |
snack | -0.031 |
patch | -0.030 |
Token | been |
Token ID | 587 |
Feature activation | +8.34e-01 |
Pos logit contributions | |
spe | +0.042 |
overed | +0.042 |
led | +0.041 |
ls | +0.041 |
otte | +0.040 |
Neg logit contributions | |
bisc | -0.046 |
wake | -0.041 |
count | -0.040 |
snack | -0.040 |
patch | -0.039 |
Token | so |
Token ID | 523 |
Feature activation | -4.54e-01 |
Pos logit contributions | |
spe | +0.021 |
overed | +0.020 |
led | +0.020 |
ls | +0.019 |
rily | +0.019 |
Neg logit contributions | |
bisc | -0.020 |
wake | -0.018 |
count | -0.017 |
snack | -0.017 |
patch | -0.016 |
Token | scared |
Token ID | 12008 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
wake | +0.009 |
bisc | +0.008 |
come | +0.008 |
welcome | +0.008 |
count | +0.008 |
Neg logit contributions | |
overed | -0.012 |
spe | -0.012 |
han | -0.012 |
rily | -0.012 |
ize | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | +4.90e-01 |
Pos logit contributions | |
bisc | +0.024 |
wake | +0.023 |
come | +0.021 |
snack | +0.021 |
patch | +0.021 |
Neg logit contributions | |
overed | -0.033 |
spe | -0.033 |
led | -0.032 |
otte | -0.032 |
Acc | -0.031 |
Token | From |
Token ID | 3574 |
Feature activation | +2.45e+00 |
Pos logit contributions | |
spe | +0.012 |
overed | +0.012 |
led | +0.012 |
ls | +0.011 |
otte | +0.011 |
Neg logit contributions | |
bisc | -0.012 |
wake | -0.010 |
count | -0.010 |
snack | -0.010 |
patch | -0.010 |
Token | then |
Token ID | 788 |
Feature activation | -9.28e-02 |
Pos logit contributions | |
spe | +0.077 |
Acc | +0.068 |
rily | +0.065 |
narrative | +0.065 |
overed | +0.064 |
Neg logit contributions | |
patch | -0.046 |
snow | -0.045 |
coat | -0.044 |
bisc | -0.044 |
carrot | -0.043 |
Token | on |
Token ID | 319 |
Feature activation | +1.69e+00 |
Pos logit contributions | |
bisc | +0.003 |
wake | +0.003 |
count | +0.003 |
snack | +0.003 |
patch | +0.002 |
Neg logit contributions | |
spe | -0.002 |
overed | -0.002 |
led | -0.002 |
ls | -0.002 |
otte | -0.001 |
Token | , |
Token ID | 11 |
Feature activation | +3.61e-01 |
Pos logit contributions | |
Acc | +0.060 |
spe | +0.060 |
overed | +0.060 |
ivating | +0.059 |
rily | +0.057 |
Neg logit contributions | |
bisc | -0.021 |
patch | -0.019 |
coat | -0.018 |
nose | -0.018 |
count | -0.018 |
Token | S |
Token ID | 311 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
spe | +0.009 |
overed | +0.008 |
led | +0.008 |
ls | +0.008 |
otte | +0.008 |
Neg logit contributions | |
bisc | -0.009 |
wake | -0.008 |
count | -0.007 |
snack | -0.007 |
patch | -0.007 |
Token | illy |
Token ID | 6548 |
Feature activation | +2.42e-01 |
Pos logit contributions | |
bisc | +0.054 |
count | +0.049 |
wake | +0.049 |
snack | +0.048 |
pile | +0.047 |
Neg logit contributions | |
led | -0.035 |
overed | -0.034 |
ls | -0.034 |
spe | -0.034 |
otte | -0.033 |
Token | up |
Token ID | 510 |
Feature activation | +4.58e-01 |
Pos logit contributions | |
spe | +0.013 |
otte | +0.012 |
Acc | +0.012 |
narrative | +0.012 |
ivating | +0.011 |
Neg logit contributions | |
bisc | -0.019 |
wake | -0.017 |
snack | -0.017 |
count | -0.017 |
go | -0.017 |
Token | the |
Token ID | 262 |
Feature activation | -2.54e-01 |
Pos logit contributions | |
spe | +0.012 |
Acc | +0.012 |
overed | +0.012 |
otte | +0.012 |
led | +0.011 |
Neg logit contributions | |
bisc | -0.010 |
snack | -0.009 |
wake | -0.009 |
count | -0.009 |
patch | -0.009 |
Token | sword |
Token ID | 8429 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
bisc | +0.006 |
wake | +0.005 |
count | +0.005 |
snack | +0.005 |
patch | +0.005 |
Neg logit contributions | |
spe | -0.006 |
overed | -0.006 |
ls | -0.006 |
led | -0.006 |
rily | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | -5.16e-01 |
Pos logit contributions | |
otte | +0.036 |
Acc | +0.036 |
spe | +0.035 |
overed | +0.035 |
han | +0.034 |
Neg logit contributions | |
bisc | -0.023 |
count | -0.020 |
pile | -0.020 |
nose | -0.020 |
patch | -0.020 |
Token | He |
Token ID | 679 |
Feature activation | +2.25e-01 |
Pos logit contributions | |
bisc | +0.012 |
wake | +0.010 |
count | +0.010 |
snack | +0.010 |
patch | +0.010 |
Neg logit contributions | |
spe | -0.013 |
led | -0.013 |
overed | -0.013 |
ls | -0.013 |
otte | -0.013 |
Token | pret |
Token ID | 2181 |
Feature activation | +3.69e+00 |
Pos logit contributions | |
spe | +0.007 |
otte | +0.007 |
overed | +0.007 |
led | +0.007 |
han | +0.007 |
Neg logit contributions | |
bisc | -0.004 |
count | -0.004 |
wake | -0.004 |
pile | -0.004 |
patch | -0.003 |
Token | ends |
Token ID | 2412 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
spe | +0.071 |
led | +0.066 |
overed | +0.065 |
otte | +0.065 |
ls | +0.064 |
Neg logit contributions | |
bisc | -0.108 |
count | -0.098 |
wake | -0.096 |
snack | -0.094 |
patch | -0.091 |
Token | it |
Token ID | 340 |
Feature activation | +6.18e-02 |
Pos logit contributions | |
spe | +0.043 |
otte | +0.043 |
ivating | +0.043 |
led | +0.043 |
han | +0.042 |
Neg logit contributions | |
bisc | -0.040 |
wake | -0.037 |
come | -0.037 |
snack | -0.037 |
count | -0.036 |
Token | is |
Token ID | 318 |
Feature activation | -5.80e-01 |
Pos logit contributions | |
led | +0.001 |
ls | +0.001 |
spe | +0.001 |
overed | +0.001 |
rily | +0.001 |
Neg logit contributions | |
bisc | -0.002 |
wake | -0.001 |
snack | -0.001 |
count | -0.001 |
patch | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | +4.63e-01 |
Pos logit contributions | |
bisc | +0.016 |
wake | +0.014 |
count | +0.013 |
snack | +0.013 |
patch | +0.013 |
Neg logit contributions | |
overed | -0.011 |
led | -0.011 |
spe | -0.011 |
ls | -0.011 |
rily | -0.011 |
Token | shovel |
Token ID | 33677 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
spe | +0.012 |
overed | +0.012 |
ls | +0.012 |
led | +0.011 |
otte | +0.011 |
Neg logit contributions | |
bisc | -0.010 |
wake | -0.008 |
count | -0.008 |
snack | -0.008 |
patch | -0.008 |
Token | , |
Token ID | 11 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
spe | +0.008 |
overed | +0.008 |
led | +0.008 |
otte | +0.007 |
ls | +0.007 |
Neg logit contributions | |
bisc | -0.007 |
wake | -0.006 |
count | -0.006 |
snack | -0.006 |
patch | -0.006 |
Token | so |
Token ID | 523 |
Feature activation | +6.90e-01 |
Pos logit contributions | |
spe | +0.030 |
overed | +0.029 |
led | +0.029 |
ls | +0.029 |
otte | +0.028 |
Neg logit contributions | |
bisc | -0.029 |
wake | -0.025 |
count | -0.024 |
snack | -0.024 |
patch | -0.024 |
Token | he |
Token ID | 339 |
Feature activation | +2.01e+00 |
Pos logit contributions | |
otte | +0.015 |
han | +0.015 |
spe | +0.015 |
overed | +0.015 |
led | +0.015 |
Neg logit contributions | |
bisc | -0.018 |
snack | -0.016 |
count | -0.016 |
go | -0.015 |
wake | -0.015 |
Token | decided |
Token ID | 3066 |
Feature activation | +1.79e+00 |
Pos logit contributions | |
spe | +0.058 |
overed | +0.056 |
ivating | +0.054 |
otte | +0.054 |
narrative | +0.053 |
Neg logit contributions | |
bisc | -0.040 |
wake | -0.036 |
patch | -0.035 |
count | -0.035 |
come | -0.035 |
Token | to |
Token ID | 284 |
Feature activation | +9.04e-01 |
Pos logit contributions | |
ivating | +0.061 |
narrative | +0.060 |
spe | +0.059 |
led | +0.059 |
otte | +0.058 |
Neg logit contributions | |
come | -0.028 |
bisc | -0.027 |
go | -0.025 |
wake | -0.025 |
patch | -0.024 |
Token | help |
Token ID | 1037 |
Feature activation | +4.33e-02 |
Pos logit contributions | |
spe | +0.031 |
overed | +0.030 |
led | +0.030 |
ls | +0.030 |
otte | +0.030 |
Neg logit contributions | |
bisc | -0.012 |
wake | -0.008 |
snack | -0.008 |
count | -0.008 |
patch | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | +7.02e-01 |
Pos logit contributions | |
ize | +0.001 |
spe | +0.001 |
ls | +0.001 |
sadly | +0.001 |
overed | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
snack | -0.001 |
wake | -0.001 |
patch | -0.001 |
count | -0.001 |
Token | He |
Token ID | 679 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
led | +0.019 |
spe | +0.018 |
otte | +0.018 |
ls | +0.018 |
overed | +0.018 |
Neg logit contributions | |
bisc | -0.016 |
count | -0.013 |
snack | -0.013 |
wake | -0.013 |
pile | -0.012 |
Token | carefully |
Token ID | 7773 |
Feature activation | +1.63e+00 |
Pos logit contributions | |
spe | +0.028 |
overed | +0.028 |
led | +0.027 |
ls | +0.027 |
otte | +0.026 |
Neg logit contributions | |
bisc | -0.023 |
wake | -0.019 |
count | -0.019 |
snack | -0.019 |
patch | -0.019 |
Token | picked |
Token ID | 6497 |
Feature activation | +4.31e-01 |
Pos logit contributions | |
led | +0.038 |
overed | +0.037 |
spe | +0.036 |
ls | +0.036 |
ized | +0.034 |
Neg logit contributions | |
bisc | -0.039 |
wake | -0.035 |
snack | -0.034 |
count | -0.033 |
're | -0.032 |
Token | up |
Token ID | 510 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
spe | +0.010 |
otte | +0.009 |
Acc | +0.008 |
narrative | +0.008 |
overed | +0.008 |
Neg logit contributions | |
bisc | -0.012 |
snack | -0.011 |
go | -0.011 |
count | -0.011 |
carrot | -0.011 |
Token | new |
Token ID | 649 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
spe | +0.005 |
ls | +0.005 |
ize | +0.004 |
otte | +0.004 |
overed | +0.004 |
Neg logit contributions | |
bisc | -0.004 |
count | -0.004 |
wake | -0.003 |
're | -0.003 |
iver | -0.003 |
Token | animal |
Token ID | 5044 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
spe | +0.032 |
led | +0.032 |
overed | +0.031 |
ls | +0.030 |
otte | +0.030 |
Neg logit contributions | |
bisc | -0.021 |
wake | -0.017 |
snack | -0.017 |
count | -0.017 |
patch | -0.016 |
Token | at |
Token ID | 379 |
Feature activation | +2.47e-01 |
Pos logit contributions | |
bisc | +0.003 |
wake | +0.003 |
patch | +0.003 |
snack | +0.003 |
count | +0.003 |
Neg logit contributions | |
spe | -0.003 |
rily | -0.002 |
otte | -0.002 |
overed | -0.002 |
Acc | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
spe | +0.006 |
led | +0.006 |
overed | +0.006 |
ls | +0.005 |
rily | +0.005 |
Neg logit contributions | |
bisc | -0.006 |
wake | -0.006 |
snack | -0.006 |
count | -0.006 |
patch | -0.005 |
Token | zoo |
Token ID | 26626 |
Feature activation | +5.88e-02 |
Pos logit contributions | |
spe | +0.002 |
ls | +0.002 |
ize | +0.002 |
ivating | +0.002 |
overed | +0.002 |
Neg logit contributions | |
bisc | -0.003 |
're | -0.002 |
count | -0.002 |
wake | -0.002 |
gloves | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
spe | +0.001 |
overed | +0.001 |
led | +0.001 |
ls | +0.001 |
otte | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
wake | -0.001 |
count | -0.001 |
snack | -0.001 |
patch | -0.001 |
Token | It |
Token ID | 632 |
Feature activation | -9.44e-01 |
Pos logit contributions | |
overed | +0.003 |
spe | +0.003 |
rily | +0.003 |
ls | +0.003 |
led | +0.003 |
Neg logit contributions | |
bisc | -0.006 |
count | -0.005 |
snack | -0.005 |
wake | -0.005 |
patch | -0.005 |
Token | has |
Token ID | 468 |
Feature activation | +7.36e-01 |
Pos logit contributions | |
bisc | +0.014 |
wake | +0.013 |
count | +0.013 |
come | +0.012 |
pile | +0.012 |
Neg logit contributions | |
otte | -0.033 |
overed | -0.032 |
spe | -0.032 |
edly | -0.032 |
Acc | -0.032 |
Token | black |
Token ID | 2042 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
spe | +0.018 |
overed | +0.017 |
led | +0.017 |
ls | +0.017 |
rily | +0.016 |
Neg logit contributions | |
bisc | -0.017 |
wake | -0.015 |
count | -0.015 |
snack | -0.015 |
patch | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | -7.91e-01 |
Pos logit contributions | |
bisc | +0.027 |
wake | +0.025 |
count | +0.024 |
patch | +0.024 |
snack | +0.023 |
Neg logit contributions | |
led | -0.029 |
ls | -0.029 |
overed | -0.029 |
edly | -0.027 |
der | -0.026 |
Token | white |
Token ID | 2330 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
bisc | +0.014 |
wake | +0.011 |
count | +0.011 |
snack | +0.011 |
nose | +0.011 |
Neg logit contributions | |
spe | -0.025 |
overed | -0.024 |
otte | -0.024 |
led | -0.024 |
rily | -0.023 |
Token | . |
Token ID | 13 |
Feature activation | -1.74e-02 |
Pos logit contributions | |
bisc | +0.033 |
wake | +0.029 |
count | +0.028 |
snack | +0.028 |
patch | +0.028 |
Neg logit contributions | |
spe | -0.031 |
overed | -0.030 |
led | -0.030 |
ls | -0.029 |
otte | -0.028 |
Token | They |
Token ID | 1119 |
Feature activation | -5.96e-01 |
Pos logit contributions | |
bisc | +0.001 |
wake | +0.000 |
count | +0.000 |
snack | +0.000 |
patch | +0.000 |
Neg logit contributions | |
spe | -0.000 |
overed | -0.000 |
led | -0.000 |
ls | -0.000 |
otte | -0.000 |
Token | did |
Token ID | 750 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
bisc | +0.014 |
wake | +0.012 |
count | +0.012 |
snack | +0.012 |
patch | +0.012 |
Neg logit contributions | |
spe | -0.015 |
overed | -0.015 |
led | -0.014 |
ls | -0.014 |
otte | -0.014 |
Token | as |
Token ID | 355 |
Feature activation | +4.57e-01 |
Pos logit contributions | |
spe | +0.038 |
led | +0.035 |
overed | +0.034 |
otte | +0.034 |
han | +0.034 |
Neg logit contributions | |
bisc | -0.044 |
snack | -0.041 |
count | -0.041 |
wake | -0.040 |
patch | -0.038 |
Token | Mama |
Token ID | 40084 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
han | +0.015 |
overed | +0.015 |
mith | +0.013 |
led | +0.013 |
jected | +0.013 |
Neg logit contributions | |
snack | -0.008 |
snow | -0.007 |
honey | -0.007 |
bisc | -0.007 |
strip | -0.007 |
Token | said |
Token ID | 531 |
Feature activation | -8.06e-01 |
Pos logit contributions | |
led | +0.004 |
spe | +0.003 |
overed | +0.003 |
ls | +0.003 |
otte | +0.003 |
Neg logit contributions | |
bisc | -0.004 |
wake | -0.003 |
count | -0.003 |
snack | -0.003 |
coat | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | +5.90e-01 |
Pos logit contributions | |
bisc | +0.018 |
wake | +0.015 |
count | +0.015 |
snack | +0.015 |
patch | +0.014 |
Neg logit contributions | |
spe | -0.021 |
overed | -0.021 |
led | -0.021 |
ls | -0.020 |
otte | -0.020 |
Token | and |
Token ID | 290 |
Feature activation | +1.26e-01 |
Pos logit contributions | |
overed | +0.013 |
spe | +0.013 |
led | +0.012 |
rily | +0.012 |
ls | +0.012 |
Neg logit contributions | |
bisc | -0.016 |
wake | -0.013 |
count | -0.013 |
snack | -0.013 |
patch | -0.013 |
Token | the |
Token ID | 262 |
Feature activation | +5.32e-01 |
Pos logit contributions | |
spe | +0.002 |
overed | +0.002 |
ls | +0.002 |
led | +0.002 |
ivating | +0.002 |
Neg logit contributions | |
bisc | -0.004 |
snack | -0.003 |
wake | -0.003 |
patch | -0.003 |
count | -0.003 |
Token | blue |
Token ID | 4171 |
Feature activation | -2.19e-01 |
Pos logit contributions | |
narrative | +0.013 |
spe | +0.013 |
ls | +0.013 |
led | +0.012 |
edly | +0.012 |
Neg logit contributions | |
bisc | -0.010 |
iver | -0.009 |
count | -0.009 |
wake | -0.009 |
chop | -0.008 |
Token | berries |
Token ID | 20853 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
bisc | +0.005 |
wake | +0.004 |
count | +0.004 |
strip | +0.004 |
chop | +0.004 |
Neg logit contributions | |
ls | -0.006 |
overed | -0.006 |
led | -0.006 |
spe | -0.006 |
rily | -0.006 |
Token | Lad |
Token ID | 12862 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
bisc | +0.021 |
're | +0.017 |
snack | +0.016 |
count | +0.016 |
iver | +0.016 |
Neg logit contributions | |
spe | -0.022 |
narrative | -0.019 |
overed | -0.019 |
sadly | -0.018 |
Ap | -0.018 |
Token | der |
Token ID | 1082 |
Feature activation | +6.98e-01 |
Pos logit contributions | |
bisc | +0.027 |
wake | +0.025 |
count | +0.024 |
snack | +0.023 |
patch | +0.023 |
Neg logit contributions | |
spe | -0.022 |
overed | -0.021 |
otte | -0.021 |
ls | -0.021 |
led | -0.020 |
Token | for |
Token ID | 329 |
Feature activation | +6.23e-01 |
Pos logit contributions | |
overed | +0.018 |
otte | +0.017 |
spe | +0.017 |
ivating | +0.017 |
rily | +0.017 |
Neg logit contributions | |
bisc | -0.014 |
count | -0.014 |
patch | -0.013 |
wake | -0.013 |
snack | -0.013 |
Token | its |
Token ID | 663 |
Feature activation | -2.18e-01 |
Pos logit contributions | |
spe | +0.014 |
overed | +0.014 |
led | +0.013 |
ls | +0.013 |
otte | +0.013 |
Neg logit contributions | |
bisc | -0.016 |
wake | -0.014 |
count | -0.014 |
snack | -0.013 |
patch | -0.013 |
Token | help |
Token ID | 1037 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
bisc | +0.004 |
wake | +0.003 |
pile | +0.003 |
patch | +0.003 |
're | +0.003 |
Neg logit contributions | |
spe | -0.006 |
overed | -0.006 |
sadly | -0.006 |
edly | -0.006 |
ls | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
bisc | +0.029 |
snack | +0.025 |
count | +0.025 |
wake | +0.025 |
patch | +0.024 |
Neg logit contributions | |
spe | -0.021 |
led | -0.019 |
sadly | -0.019 |
overed | -0.019 |
rily | -0.019 |
Token | And |
Token ID | 843 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
bisc | +0.003 |
wake | +0.003 |
count | +0.003 |
snack | +0.003 |
patch | +0.002 |
Neg logit contributions | |
spe | -0.003 |
led | -0.003 |
overed | -0.003 |
ls | -0.003 |
otte | -0.003 |
Token | they |
Token ID | 484 |
Feature activation | -5.71e-01 |
Pos logit contributions | |
overed | +0.004 |
led | +0.004 |
ivating | +0.004 |
ls | +0.004 |
Acc | +0.004 |
Neg logit contributions | |
bisc | -0.004 |
count | -0.004 |
wake | -0.004 |
come | -0.004 |
go | -0.004 |
Token | always |
Token ID | 1464 |
Feature activation | +5.82e-01 |
Pos logit contributions | |
bisc | +0.013 |
patch | +0.012 |
count | +0.012 |
wake | +0.012 |
come | +0.012 |
Neg logit contributions | |
overed | -0.014 |
ls | -0.013 |
rily | -0.013 |
spe | -0.013 |
ivating | -0.013 |
Token | went |
Token ID | 1816 |
Feature activation | -6.15e-01 |
Pos logit contributions | |
spe | +0.013 |
led | +0.013 |
overed | +0.012 |
otte | +0.012 |
sadly | +0.012 |
Neg logit contributions | |
bisc | -0.015 |
snack | -0.013 |
wake | -0.013 |
count | -0.012 |
patch | -0.012 |
Token | on |
Token ID | 319 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
bisc | +0.014 |
snack | +0.013 |
count | +0.013 |
wake | +0.013 |
come | +0.013 |
Neg logit contributions | |
spe | -0.016 |
overed | -0.015 |
led | -0.015 |
otte | -0.015 |
rily | -0.015 |
Token | , |
Token ID | 11 |
Feature activation | +6.25e-01 |
Pos logit contributions | |
ls | +0.017 |
led | +0.016 |
overed | +0.016 |
spe | +0.015 |
otte | +0.015 |
Neg logit contributions | |
bisc | -0.016 |
wake | -0.014 |
count | -0.014 |
patch | -0.014 |
pile | -0.013 |
Token | Lily |
Token ID | 20037 |
Feature activation | -4.73e-01 |
Pos logit contributions | |
spe | +0.015 |
overed | +0.013 |
led | +0.013 |
rily | +0.013 |
narrative | +0.013 |
Neg logit contributions | |
bisc | -0.017 |
snack | -0.016 |
wake | -0.015 |
nose | -0.015 |
count | -0.015 |
Token | 's |
Token ID | 338 |
Feature activation | -7.15e-01 |
Pos logit contributions | |
bisc | +0.009 |
wake | +0.007 |
snack | +0.007 |
count | +0.007 |
patch | +0.007 |
Neg logit contributions | |
spe | -0.014 |
overed | -0.014 |
led | -0.014 |
ls | -0.014 |
rily | -0.013 |
Token | mom |
Token ID | 1995 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
bisc | +0.019 |
count | +0.017 |
wake | +0.015 |
're | +0.015 |
chop | +0.015 |
Neg logit contributions | |
spe | -0.014 |
otte | -0.014 |
ls | -0.013 |
overed | -0.013 |
led | -0.013 |
Token | asked |
Token ID | 1965 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
bisc | +0.033 |
count | +0.028 |
snack | +0.028 |
wake | +0.027 |
nose | +0.026 |
Neg logit contributions | |
led | -0.028 |
spe | -0.028 |
otte | -0.027 |
ls | -0.026 |
han | -0.025 |
Token | her |
Token ID | 607 |
Feature activation | -7.12e-01 |
Pos logit contributions | |
bisc | +0.060 |
patch | +0.052 |
count | +0.052 |
pile | +0.051 |
strip | +0.051 |
Neg logit contributions | |
otte | -0.012 |
led | -0.011 |
ize | -0.010 |
ls | -0.010 |
sadly | -0.010 |
Token | to |
Token ID | 284 |
Feature activation | -1.07e+00 |
Pos logit contributions | |
bisc | +0.016 |
wake | +0.015 |
count | +0.015 |
snack | +0.015 |
come | +0.014 |
Neg logit contributions | |
overed | -0.018 |
led | -0.018 |
spe | -0.018 |
ivating | -0.017 |
rily | -0.017 |
Token | perform |
Token ID | 1620 |
Feature activation | -8.26e-02 |
Pos logit contributions | |
bisc | +0.019 |
est | +0.016 |
itten | +0.014 |
ch | +0.014 |
snow | +0.013 |
Neg logit contributions | |
han | -0.028 |
otte | -0.028 |
proudly | -0.027 |
spe | -0.026 |
ize | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | +8.63e-01 |
Pos logit contributions | |
bisc | +0.002 |
wake | +0.002 |
count | +0.002 |
snack | +0.002 |
patch | +0.002 |
Neg logit contributions | |
overed | -0.002 |
led | -0.002 |
spe | -0.002 |
ls | -0.002 |
otte | -0.002 |
Token | little |
Token ID | 1310 |
Feature activation | +7.13e-02 |
Pos logit contributions | |
ls | +0.020 |
spe | +0.019 |
otte | +0.019 |
ize | +0.018 |
led | +0.018 |
Neg logit contributions | |
bisc | -0.020 |
count | -0.017 |
're | -0.017 |
paw | -0.016 |
wake | -0.016 |
Token | dance |
Token ID | 9280 |
Feature activation | -3.65e-01 |
Pos logit contributions | |
ls | +0.002 |
spe | +0.002 |
led | +0.002 |
otte | +0.002 |
closely | +0.002 |
Neg logit contributions | |
bisc | -0.001 |
're | -0.001 |
count | -0.001 |
iver | -0.001 |
chop | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +1.48e-01 |
Pos logit contributions | |
bisc | +0.001 |
snack | +0.001 |
wake | +0.001 |
count | +0.001 |
patch | +0.001 |
Neg logit contributions | |
overed | -0.001 |
spe | -0.001 |
otte | -0.001 |
led | -0.001 |
ls | -0.001 |
Token |
|
Token ID | 198 |
Feature activation | -3.51e-01 |
Pos logit contributions | |
led | +0.005 |
overed | +0.004 |
ls | +0.004 |
ized | +0.004 |
spe | +0.004 |
Neg logit contributions | |
bisc | -0.003 |
paw | -0.002 |
count | -0.002 |
Cat | -0.002 |
wake | -0.002 |
Token |
|
Token ID | 198 |
Feature activation | -1.33e-01 |
Pos logit contributions | |
bisc | +0.007 |
wake | +0.006 |
count | +0.006 |
snack | +0.006 |
patch | +0.006 |
Neg logit contributions | |
led | -0.010 |
spe | -0.010 |
overed | -0.010 |
rily | -0.009 |
ls | -0.009 |
Token | They |
Token ID | 2990 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
bisc | +0.002 |
count | +0.002 |
wake | +0.002 |
snack | +0.002 |
patch | +0.002 |
Neg logit contributions | |
led | -0.004 |
rily | -0.004 |
overed | -0.004 |
ls | -0.004 |
spe | -0.004 |
Token | broke |
Token ID | 6265 |
Feature activation | -7.51e-02 |
Pos logit contributions | |
bisc | +0.024 |
wake | +0.022 |
count | +0.022 |
come | +0.021 |
patch | +0.021 |
Neg logit contributions | |
overed | -0.038 |
ls | -0.037 |
rily | -0.037 |
spe | -0.037 |
han | -0.036 |
Token | some |
Token ID | 617 |
Feature activation | -2.47e+00 |
Pos logit contributions | |
snow | +0.002 |
m | +0.002 |
strip | +0.002 |
nose | +0.002 |
paw | +0.002 |
Neg logit contributions | |
Acc | -0.002 |
overed | -0.002 |
narrative | -0.002 |
riet | -0.002 |
han | -0.002 |
Token | of |
Token ID | 286 |
Feature activation | +6.47e-02 |
Pos logit contributions | |
bisc | +0.052 |
wake | +0.045 |
count | +0.044 |
snack | +0.043 |
patch | +0.042 |
Neg logit contributions | |
spe | -0.067 |
overed | -0.065 |
led | -0.064 |
ls | -0.063 |
otte | -0.062 |
Token | the |
Token ID | 262 |
Feature activation | +2.43e-01 |
Pos logit contributions | |
overed | +0.001 |
led | +0.001 |
spe | +0.001 |
ls | +0.001 |
otte | +0.001 |
Neg logit contributions | |
bisc | -0.002 |
snack | -0.002 |
count | -0.002 |
wake | -0.001 |
pile | -0.001 |
Token | things |
Token ID | 1243 |
Feature activation | +5.90e-01 |
Pos logit contributions | |
overed | +0.007 |
spe | +0.007 |
led | +0.007 |
ls | +0.007 |
otte | +0.007 |
Neg logit contributions | |
bisc | -0.005 |
wake | -0.004 |
snack | -0.004 |
count | -0.004 |
patch | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -4.54e-01 |
Pos logit contributions | |
spe | +0.015 |
overed | +0.015 |
Acc | +0.015 |
otte | +0.015 |
led | +0.014 |
Neg logit contributions | |
bisc | -0.013 |
wake | -0.012 |
come | -0.012 |
go | -0.011 |
count | -0.011 |
Token | They |
Token ID | 1119 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
bisc | +0.011 |
count | +0.009 |
wake | +0.009 |
snack | +0.009 |
paw | +0.009 |
Neg logit contributions | |
led | -0.012 |
ls | -0.011 |
overed | -0.011 |
spe | -0.011 |
otte | -0.011 |
Token | of |
Token ID | 286 |
Feature activation | -1.96e-01 |
Pos logit contributions | |
bisc | +0.030 |
count | +0.025 |
wake | +0.025 |
patch | +0.024 |
snack | +0.024 |
Neg logit contributions | |
spe | -0.031 |
ls | -0.030 |
led | -0.030 |
overed | -0.029 |
rily | -0.028 |
Token | noise |
Token ID | 7838 |
Feature activation | -9.99e-01 |
Pos logit contributions | |
bisc | +0.004 |
patch | +0.004 |
count | +0.004 |
wake | +0.003 |
strip | +0.003 |
Neg logit contributions | |
led | -0.005 |
ls | -0.005 |
spe | -0.005 |
ize | -0.004 |
closely | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -1.67e-02 |
Pos logit contributions | |
bisc | +0.026 |
wake | +0.022 |
snack | +0.022 |
count | +0.022 |
patch | +0.022 |
Neg logit contributions | |
spe | -0.022 |
led | -0.022 |
ls | -0.021 |
overed | -0.021 |
rily | -0.020 |
Token | |
Token ID | 220 |
Feature activation | +3.03e-01 |
Pos logit contributions | |
bisc | +0.000 |
wake | +0.000 |
patch | +0.000 |
snack | +0.000 |
count | +0.000 |
Neg logit contributions | |
spe | -0.000 |
overed | -0.000 |
ivating | -0.000 |
rily | -0.000 |
ls | -0.000 |
Token |
|
Token ID | 198 |
Feature activation | -3.05e-01 |
Pos logit contributions | |
spe | +0.007 |
overed | +0.007 |
led | +0.007 |
ls | +0.007 |
otte | +0.007 |
Neg logit contributions | |
bisc | -0.007 |
wake | -0.006 |
count | -0.006 |
snack | -0.006 |
patch | -0.006 |
Token | One |
Token ID | 3198 |
Feature activation | -2.57e+00 |
Pos logit contributions | |
bisc | +0.008 |
wake | +0.007 |
count | +0.007 |
snack | +0.007 |
patch | +0.007 |
Neg logit contributions | |
spe | -0.007 |
overed | -0.007 |
ls | -0.006 |
led | -0.006 |
otte | -0.006 |
Token | day |
Token ID | 1110 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
bisc | +0.064 |
count | +0.053 |
wake | +0.050 |
patch | +0.050 |
pile | +0.050 |
Neg logit contributions | |
spe | -0.060 |
led | -0.058 |
ls | -0.056 |
overed | -0.054 |
otte | -0.053 |
Token | , |
Token ID | 11 |
Feature activation | +1.29e-01 |
Pos logit contributions | |
spe | +0.004 |
overed | +0.004 |
led | +0.004 |
ls | +0.004 |
otte | +0.004 |
Neg logit contributions | |
bisc | -0.003 |
wake | -0.003 |
count | -0.003 |
patch | -0.003 |
snack | -0.003 |
Token | he |
Token ID | 339 |
Feature activation | -4.02e-01 |
Pos logit contributions | |
spe | +0.003 |
overed | +0.003 |
led | +0.003 |
han | +0.003 |
ls | +0.003 |
Neg logit contributions | |
bisc | -0.003 |
snack | -0.003 |
count | -0.003 |
wake | -0.003 |
nose | -0.003 |
Token | went |
Token ID | 1816 |
Feature activation | -3.94e-01 |
Pos logit contributions | |
bisc | +0.009 |
wake | +0.007 |
count | +0.007 |
snack | +0.007 |
patch | +0.007 |
Neg logit contributions | |
spe | -0.011 |
overed | -0.011 |
led | -0.011 |
ls | -0.010 |
otte | -0.010 |
Token | shopping |
Token ID | 9735 |
Feature activation | +4.68e-01 |
Pos logit contributions | |
bisc | +0.009 |
wake | +0.008 |
count | +0.008 |
snack | +0.008 |
patch | +0.008 |
Neg logit contributions | |
spe | -0.011 |
led | -0.010 |
overed | -0.010 |
rily | -0.010 |
otte | -0.010 |
Token | having |
Token ID | 1719 |
Feature activation | +9.52e-01 |
Pos logit contributions | |
bisc | +0.027 |
wake | +0.024 |
count | +0.024 |
snack | +0.023 |
patch | +0.023 |
Neg logit contributions | |
spe | -0.028 |
overed | -0.027 |
led | -0.027 |
ls | -0.027 |
otte | -0.026 |
Token | fun |
Token ID | 1257 |
Feature activation | -8.10e-01 |
Pos logit contributions | |
spe | +0.027 |
overed | +0.026 |
led | +0.026 |
ls | +0.025 |
otte | +0.025 |
Neg logit contributions | |
bisc | -0.020 |
wake | -0.017 |
count | -0.016 |
snack | -0.016 |
patch | -0.016 |
Token | together |
Token ID | 1978 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
bisc | +0.019 |
snack | +0.017 |
count | +0.017 |
pile | +0.017 |
patch | +0.017 |
Neg logit contributions | |
spe | -0.020 |
otte | -0.020 |
riet | -0.019 |
overed | -0.019 |
rily | -0.019 |
Token | . |
Token ID | 13 |
Feature activation | -3.99e-01 |
Pos logit contributions | |
spe | +0.005 |
overed | +0.005 |
led | +0.005 |
ls | +0.005 |
otte | +0.005 |
Neg logit contributions | |
bisc | -0.004 |
wake | -0.004 |
count | -0.004 |
snack | -0.004 |
patch | -0.004 |
Token | In |
Token ID | 554 |
Feature activation | -7.54e-01 |
Pos logit contributions | |
bisc | +0.009 |
count | +0.009 |
snack | +0.008 |
wake | +0.008 |
go | +0.008 |
Neg logit contributions | |
led | -0.010 |
rily | -0.009 |
ls | -0.009 |
otte | -0.009 |
spe | -0.009 |
Token | the |
Token ID | 262 |
Feature activation | -2.17e+00 |
Pos logit contributions | |
bisc | +0.018 |
count | +0.017 |
snack | +0.016 |
patch | +0.016 |
wake | +0.016 |
Neg logit contributions | |
led | -0.018 |
spe | -0.018 |
overed | -0.018 |
ls | -0.018 |
rily | -0.017 |
Token | end |
Token ID | 886 |
Feature activation | -3.53e+00 |
Pos logit contributions | |
bisc | +0.033 |
wake | +0.025 |
count | +0.024 |
snack | +0.023 |
nose | +0.023 |
Neg logit contributions | |
spe | -0.067 |
narrative | -0.065 |
Acc | -0.062 |
han | -0.062 |
sadly | -0.062 |
Token | , |
Token ID | 11 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
bisc | +0.062 |
count | +0.052 |
wake | +0.052 |
snack | +0.051 |
patch | +0.049 |
Neg logit contributions | |
spe | -0.108 |
led | -0.105 |
overed | -0.105 |
otte | -0.102 |
rily | -0.102 |
Token | they |
Token ID | 484 |
Feature activation | -1.22e+00 |
Pos logit contributions | |
count | +0.039 |
go | +0.037 |
come | +0.037 |
snack | +0.036 |
bisc | +0.036 |
Neg logit contributions | |
rily | -0.035 |
edly | -0.034 |
Acc | -0.034 |
led | -0.034 |
overed | -0.034 |
Token | became |
Token ID | 2627 |
Feature activation | +3.04e-02 |
Pos logit contributions | |
bisc | +0.026 |
patch | +0.024 |
wake | +0.024 |
count | +0.024 |
come | +0.023 |
Neg logit contributions | |
overed | -0.033 |
spe | -0.031 |
ls | -0.031 |
led | -0.030 |
han | -0.030 |
Token | good |
Token ID | 922 |
Feature activation | +6.99e-01 |
Pos logit contributions | |
overed | +0.001 |
led | +0.001 |
spe | +0.001 |
ls | +0.001 |
rily | +0.001 |
Neg logit contributions | |
bisc | -0.001 |
count | -0.001 |
snack | -0.001 |
patch | -0.001 |
go | -0.001 |
Token | sleep |
Token ID | 3993 |
Feature activation | +3.23e-01 |
Pos logit contributions | |
bisc | +0.001 |
itten | +0.001 |
ch | +0.001 |
est | +0.001 |
strip | +0.001 |
Neg logit contributions | |
spe | -0.002 |
sadly | -0.002 |
otte | -0.002 |
rily | -0.002 |
han | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +5.44e-01 |
Pos logit contributions | |
spe | +0.007 |
overed | +0.007 |
led | +0.007 |
ls | +0.007 |
otte | +0.006 |
Neg logit contributions | |
bisc | -0.009 |
wake | -0.008 |
count | -0.007 |
snack | -0.007 |
patch | -0.007 |
Token | He |
Token ID | 679 |
Feature activation | +7.93e-01 |
Pos logit contributions | |
spe | +0.011 |
overed | +0.011 |
led | +0.010 |
rily | +0.010 |
otte | +0.010 |
Neg logit contributions | |
bisc | -0.015 |
wake | -0.013 |
snack | -0.013 |
count | -0.013 |
patch | -0.013 |
Token | asked |
Token ID | 1965 |
Feature activation | -6.38e-01 |
Pos logit contributions | |
led | +0.018 |
spe | +0.017 |
sadly | +0.017 |
ls | +0.016 |
otte | +0.016 |
Neg logit contributions | |
bisc | -0.020 |
snack | -0.017 |
count | -0.017 |
wake | -0.016 |
welcome | -0.016 |
Token | his |
Token ID | 465 |
Feature activation | +1.84e-01 |
Pos logit contributions | |
bisc | +0.026 |
itten | +0.022 |
patch | +0.022 |
iver | +0.022 |
icious | +0.022 |
Neg logit contributions | |
sadly | -0.006 |
spe | -0.005 |
happily | -0.005 |
led | -0.004 |
proudly | -0.004 |
Token | mom |
Token ID | 1995 |
Feature activation | -2.56e+00 |
Pos logit contributions | |
spe | +0.004 |
ls | +0.004 |
otte | +0.004 |
sadly | +0.004 |
overed | +0.003 |
Neg logit contributions | |
bisc | -0.005 |
wake | -0.004 |
're | -0.004 |
count | -0.004 |
patch | -0.004 |
Token | if |
Token ID | 611 |
Feature activation | -6.50e-01 |
Pos logit contributions | |
bisc | +0.071 |
wake | +0.057 |
snack | +0.057 |
paw | +0.056 |
strip | +0.056 |
Neg logit contributions | |
spe | -0.053 |
ls | -0.049 |
otte | -0.048 |
sadly | -0.048 |
led | -0.047 |
Token | it |
Token ID | 340 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
bisc | +0.016 |
wake | +0.015 |
come | +0.015 |
snack | +0.015 |
count | +0.015 |
Neg logit contributions | |
overed | -0.015 |
ivating | -0.015 |
rily | -0.015 |
spe | -0.015 |
led | -0.015 |
Token | was |
Token ID | 373 |
Feature activation | -7.27e-01 |
Pos logit contributions | |
bisc | +0.037 |
iver | +0.033 |
icious | +0.032 |
count | +0.031 |
feed | +0.031 |
Neg logit contributions | |
led | -0.018 |
spe | -0.018 |
sadly | -0.017 |
ls | -0.016 |
overed | -0.015 |
Token | time |
Token ID | 640 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
bisc | +0.019 |
count | +0.014 |
strip | +0.014 |
Grass | +0.014 |
chop | +0.014 |
Neg logit contributions | |
sadly | -0.016 |
rily | -0.015 |
led | -0.015 |
happily | -0.015 |
otte | -0.015 |
Token | for |
Token ID | 329 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
bisc | +0.040 |
wake | +0.034 |
count | +0.034 |
snack | +0.034 |
patch | +0.033 |
Neg logit contributions | |
spe | -0.036 |
overed | -0.035 |
led | -0.035 |
ls | -0.034 |
otte | -0.033 |
Token | what |
Token ID | 644 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
spe | +0.006 |
ls | +0.006 |
led | +0.006 |
overed | +0.005 |
ize | +0.005 |
Neg logit contributions | |
bisc | -0.008 |
wake | -0.007 |
count | -0.007 |
snack | -0.007 |
nose | -0.007 |
Token | he |
Token ID | 339 |
Feature activation | +2.94e+00 |
Pos logit contributions | |
spe | +0.035 |
ls | +0.032 |
led | +0.031 |
overed | +0.030 |
ize | +0.029 |
Neg logit contributions | |
bisc | -0.058 |
wake | -0.049 |
snack | -0.049 |
count | -0.048 |
nose | -0.048 |
Token | could |
Token ID | 714 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
led | +0.057 |
ls | +0.051 |
spe | +0.049 |
overed | +0.048 |
otte | +0.046 |
Neg logit contributions | |
bisc | -0.087 |
snack | -0.076 |
nose | -0.075 |
wake | -0.074 |
count | -0.072 |
Token | see |
Token ID | 766 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
spe | +0.027 |
overed | +0.027 |
ize | +0.026 |
ized | +0.026 |
closely | +0.026 |
Neg logit contributions | |
bisc | -0.026 |
ch | -0.019 |
snack | -0.018 |
snow | -0.018 |
nose | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | -2.72e-02 |
Pos logit contributions | |
spe | +0.023 |
led | +0.023 |
ls | +0.022 |
overed | +0.022 |
rily | +0.021 |
Neg logit contributions | |
bisc | -0.029 |
wake | -0.025 |
snack | -0.025 |
count | -0.024 |
patch | -0.024 |
Token | One |
Token ID | 1881 |
Feature activation | -2.77e+00 |
Pos logit contributions | |
bisc | +0.001 |
wake | +0.001 |
patch | +0.001 |
snack | +0.001 |
count | +0.001 |
Neg logit contributions | |
spe | -0.001 |
overed | -0.000 |
Percy | -0.000 |
sadly | -0.000 |
han | -0.000 |
Token | day |
Token ID | 1110 |
Feature activation | +5.06e-01 |
Pos logit contributions | |
bisc | +0.080 |
count | +0.067 |
're | +0.064 |
pile | +0.063 |
wake | +0.063 |
Neg logit contributions | |
spe | -0.053 |
led | -0.048 |
sadly | -0.044 |
ls | -0.044 |
closely | -0.043 |
Token | he |
Token ID | 339 |
Feature activation | -8.96e-02 |
Pos logit contributions | |
led | +0.013 |
spe | +0.013 |
overed | +0.013 |
ls | +0.013 |
otte | +0.012 |
Neg logit contributions | |
bisc | -0.011 |
wake | -0.010 |
count | -0.009 |
patch | -0.009 |
snack | -0.009 |
Token | visited |
Token ID | 8672 |
Feature activation | +5.22e-01 |
Pos logit contributions | |
bisc | +0.002 |
wake | +0.002 |
count | +0.002 |
snack | +0.002 |
patch | +0.002 |
Neg logit contributions | |
spe | -0.002 |
led | -0.002 |
overed | -0.002 |
ls | -0.002 |
otte | -0.002 |
Token | his |
Token ID | 465 |
Feature activation | +7.95e-01 |
Pos logit contributions | |
overed | +0.010 |
led | +0.010 |
spe | +0.010 |
ls | +0.010 |
ize | +0.009 |
Neg logit contributions | |
bisc | -0.015 |
wake | -0.013 |
count | -0.013 |
patch | -0.012 |
snack | -0.012 |
Token | grandma |
Token ID | 49890 |
Feature activation | -3.62e-01 |
Pos logit contributions | |
spe | +0.021 |
ls | +0.020 |
overed | +0.020 |
otte | +0.020 |
ize | +0.019 |
Neg logit contributions | |
bisc | -0.018 |
wake | -0.014 |
count | -0.014 |
snack | -0.014 |
're | -0.013 |