Token | happy |
Token ID | 3772 |
Feature activation | -4.29e+00 |
Pos logit contributions | |
boards | +0.056 |
Did | +0.056 |
auc | +0.055 |
uit | +0.055 |
cher | +0.054 |
Neg logit contributions | |
breath | -0.043 |
common | -0.034 |
job | -0.031 |
sleep | -0.031 |
care | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | +5.20e-01 |
Pos logit contributions | |
breath | +0.090 |
common | +0.071 |
job | +0.066 |
sleep | +0.066 |
long | +0.066 |
Neg logit contributions | |
boards | -0.110 |
Did | -0.107 |
uit | -0.106 |
auc | -0.106 |
cher | -0.105 |
Token | He |
Token ID | 679 |
Feature activation | +3.78e-01 |
Pos logit contributions | |
boards | +0.011 |
Did | +0.010 |
auc | +0.010 |
uit | +0.010 |
washer | +0.010 |
Neg logit contributions | |
breath | -0.014 |
common | -0.011 |
job | -0.011 |
sleep | -0.011 |
care | -0.011 |
Token | thought |
Token ID | 1807 |
Feature activation | +2.22e+00 |
Pos logit contributions | |
boards | +0.009 |
Did | +0.008 |
uit | +0.008 |
auc | +0.008 |
cher | +0.008 |
Neg logit contributions | |
breath | -0.009 |
common | -0.007 |
job | -0.007 |
long | -0.007 |
sleep | -0.007 |
Token | he |
Token ID | 339 |
Feature activation | +2.27e+00 |
Pos logit contributions | |
Did | +0.061 |
ode | +0.059 |
cher | +0.058 |
uit | +0.058 |
register | +0.058 |
Neg logit contributions | |
breath | -0.046 |
long | -0.038 |
sleep | -0.036 |
time | -0.035 |
common | -0.034 |
Token | found |
Token ID | 1043 |
Feature activation | +7.71e+00 |
Pos logit contributions | |
auc | +0.060 |
register | +0.059 |
boards | +0.055 |
uit | +0.055 |
bling | +0.054 |
Neg logit contributions | |
breath | -0.044 |
happily | -0.040 |
care | -0.038 |
sleep | -0.038 |
own | -0.036 |
Token | the |
Token ID | 262 |
Feature activation | +3.64e+00 |
Pos logit contributions | |
boards | +0.238 |
Did | +0.234 |
ode | +0.226 |
ains | +0.223 |
auc | +0.223 |
Neg logit contributions | |
breath | -0.142 |
care | -0.111 |
home | -0.109 |
long | -0.109 |
sleep | -0.101 |
Token | owner |
Token ID | 4870 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
boards | +0.088 |
Did | +0.086 |
auc | +0.086 |
ains | +0.084 |
cher | +0.083 |
Neg logit contributions | |
breath | -0.089 |
long | -0.072 |
job | -0.071 |
common | -0.070 |
time | -0.068 |
Token | of |
Token ID | 286 |
Feature activation | +8.80e-01 |
Pos logit contributions | |
auc | +0.056 |
boards | +0.055 |
ains | +0.054 |
Did | +0.053 |
Which | +0.051 |
Neg logit contributions | |
breath | -0.031 |
happily | -0.028 |
sleep | -0.028 |
, | -0.027 |
long | -0.027 |
Token | the |
Token ID | 262 |
Feature activation | +2.23e+00 |
Pos logit contributions | |
boards | +0.027 |
auc | +0.026 |
washer | +0.026 |
cher | +0.026 |
Did | +0.026 |
Neg logit contributions | |
breath | -0.016 |
common | -0.013 |
long | -0.012 |
sleep | -0.011 |
time | -0.011 |
Token | glove |
Token ID | 29144 |
Feature activation | +1.60e+00 |
Pos logit contributions | |
boards | +0.049 |
Did | +0.048 |
auc | +0.047 |
cher | +0.047 |
uit | +0.047 |
Neg logit contributions | |
breath | -0.056 |
common | -0.045 |
job | -0.045 |
long | -0.044 |
sleep | -0.044 |
Token | drop |
Token ID | 4268 |
Feature activation | +3.35e+00 |
Pos logit contributions | |
boards | +0.022 |
auc | +0.021 |
ains | +0.020 |
uit | +0.020 |
Did | +0.019 |
Neg logit contributions | |
breath | -0.026 |
sleep | -0.021 |
long | -0.020 |
home | -0.020 |
work | -0.020 |
Token | his |
Token ID | 465 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
ains | +0.094 |
ode | +0.092 |
cher | +0.091 |
boards | +0.090 |
Did | +0.089 |
Neg logit contributions | |
breath | -0.068 |
long | -0.052 |
care | -0.048 |
home | -0.048 |
time | -0.046 |
Token | wallet |
Token ID | 13008 |
Feature activation | +5.16e+00 |
Pos logit contributions | |
boards | +0.021 |
Did | +0.021 |
auc | +0.020 |
uit | +0.020 |
cher | +0.020 |
Neg logit contributions | |
breath | -0.026 |
common | -0.022 |
job | -0.021 |
sleep | -0.021 |
long | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | +7.17e-01 |
Pos logit contributions | |
Did | +0.155 |
ains | +0.150 |
Which | +0.150 |
auc | +0.149 |
boards | +0.145 |
Neg logit contributions | |
breath | -0.087 |
long | -0.067 |
home | -0.060 |
time | -0.058 |
job | -0.058 |
Token | He |
Token ID | 679 |
Feature activation | +8.58e-01 |
Pos logit contributions | |
boards | +0.018 |
Did | +0.018 |
auc | +0.017 |
uit | +0.017 |
washer | +0.017 |
Neg logit contributions | |
breath | -0.016 |
common | -0.012 |
job | -0.011 |
sleep | -0.011 |
long | -0.011 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.67e+00 |
Pos logit contributions | |
boards | +0.019 |
Did | +0.018 |
uit | +0.018 |
auc | +0.018 |
washer | +0.018 |
Neg logit contributions | |
breath | -0.021 |
common | -0.017 |
job | -0.017 |
sleep | -0.016 |
long | -0.016 |
Token | it |
Token ID | 340 |
Feature activation | +3.23e+00 |
Pos logit contributions | |
Did | +0.222 |
boards | +0.220 |
auc | +0.217 |
cher | +0.214 |
ode | +0.213 |
Neg logit contributions | |
breath | -0.144 |
long | -0.108 |
home | -0.107 |
care | -0.107 |
job | -0.103 |
Token | up |
Token ID | 510 |
Feature activation | -2.08e+00 |
Pos logit contributions | |
Did | +0.078 |
boards | +0.077 |
auc | +0.075 |
cher | +0.074 |
uit | +0.073 |
Neg logit contributions | |
breath | -0.076 |
common | -0.058 |
job | -0.058 |
long | -0.058 |
sleep | -0.057 |
Token | and |
Token ID | 290 |
Feature activation | -4.50e-01 |
Pos logit contributions | |
breath | +0.045 |
common | +0.037 |
sleep | +0.035 |
job | +0.034 |
care | +0.034 |
Neg logit contributions | |
boards | -0.050 |
uit | -0.049 |
Did | -0.048 |
auc | -0.047 |
washer | -0.047 |
Token | gave |
Token ID | 2921 |
Feature activation | +4.35e+00 |
Pos logit contributions | |
breath | +0.009 |
happily | +0.007 |
work | +0.007 |
sleep | +0.007 |
care | +0.007 |
Neg logit contributions | |
auc | -0.012 |
boards | -0.012 |
ains | -0.012 |
uit | -0.012 |
ode | -0.012 |
Token | it |
Token ID | 340 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
boards | +0.120 |
Did | +0.117 |
auc | +0.116 |
cher | +0.114 |
uit | +0.114 |
Neg logit contributions | |
breath | -0.088 |
care | -0.064 |
common | -0.063 |
job | -0.062 |
long | -0.061 |
Token | on |
Token ID | 319 |
Feature activation | -3.15e+00 |
Pos logit contributions | |
Did | +0.079 |
boards | +0.076 |
ains | +0.074 |
uit | +0.074 |
cher | +0.073 |
Neg logit contributions | |
breath | -0.042 |
job | -0.036 |
long | -0.031 |
idea | -0.031 |
day | -0.031 |
Token | the |
Token ID | 262 |
Feature activation | +2.28e+00 |
Pos logit contributions | |
breath | +0.067 |
common | +0.053 |
long | +0.051 |
job | +0.050 |
sleep | +0.050 |
Neg logit contributions | |
boards | -0.080 |
Did | -0.079 |
auc | -0.079 |
cher | -0.079 |
uit | -0.077 |
Token | ground |
Token ID | 2323 |
Feature activation | +1.74e-01 |
Pos logit contributions | |
Did | +0.053 |
cher | +0.050 |
uit | +0.050 |
boards | +0.050 |
auc | +0.049 |
Neg logit contributions | |
breath | -0.054 |
job | -0.047 |
long | -0.046 |
sleep | -0.043 |
home | -0.043 |
Token | . |
Token ID | 13 |
Feature activation | +7.90e-02 |
Pos logit contributions | |
cher | +0.006 |
Did | +0.006 |
boards | +0.006 |
uit | +0.006 |
auc | +0.006 |
Neg logit contributions | |
breath | -0.002 |
sleep | -0.001 |
long | -0.001 |
time | -0.001 |
job | -0.001 |
Token | He |
Token ID | 679 |
Feature activation | +6.27e-01 |
Pos logit contributions | |
boards | +0.001 |
Did | +0.001 |
lying | +0.001 |
ting | +0.001 |
uit | +0.001 |
Neg logit contributions | |
breath | -0.002 |
common | -0.002 |
job | -0.002 |
care | -0.002 |
sleep | -0.002 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.65e+00 |
Pos logit contributions | |
boards | +0.013 |
Did | +0.013 |
washer | +0.012 |
auc | +0.012 |
uit | +0.012 |
Neg logit contributions | |
breath | -0.016 |
common | -0.013 |
job | -0.013 |
sleep | -0.012 |
care | -0.012 |
Token | it |
Token ID | 340 |
Feature activation | +3.25e+00 |
Pos logit contributions | |
Did | +0.223 |
cher | +0.211 |
auc | +0.209 |
ode | +0.209 |
uit | +0.208 |
Neg logit contributions | |
breath | -0.143 |
long | -0.121 |
job | -0.114 |
care | -0.113 |
home | -0.112 |
Token | up |
Token ID | 510 |
Feature activation | -2.20e+00 |
Pos logit contributions | |
Did | +0.096 |
cher | +0.089 |
auc | +0.088 |
boards | +0.087 |
register | +0.085 |
Neg logit contributions | |
breath | -0.066 |
home | -0.051 |
job | -0.051 |
long | -0.051 |
work | -0.049 |
Token | and |
Token ID | 290 |
Feature activation | -2.22e-01 |
Pos logit contributions | |
breath | +0.045 |
common | +0.034 |
long | +0.032 |
job | +0.032 |
sleep | +0.032 |
Neg logit contributions | |
boards | -0.058 |
Did | -0.057 |
auc | -0.057 |
uit | -0.056 |
cher | -0.056 |
Token | started |
Token ID | 2067 |
Feature activation | +5.41e-01 |
Pos logit contributions | |
breath | +0.005 |
work | +0.004 |
care | +0.003 |
sleep | +0.003 |
long | +0.003 |
Neg logit contributions | |
uit | -0.006 |
auc | -0.006 |
boards | -0.006 |
register | -0.006 |
ains | -0.006 |
Token | walking |
Token ID | 6155 |
Feature activation | -5.35e-01 |
Pos logit contributions | |
boards | +0.015 |
auc | +0.014 |
Did | +0.014 |
uit | +0.014 |
cher | +0.014 |
Neg logit contributions | |
breath | -0.011 |
long | -0.008 |
common | -0.008 |
job | -0.008 |
sleep | -0.008 |
Token | big |
Token ID | 1263 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
Did | +0.058 |
boards | +0.057 |
cher | +0.056 |
uit | +0.055 |
auc | +0.055 |
Neg logit contributions | |
breath | -0.053 |
job | -0.041 |
common | -0.041 |
long | -0.041 |
sleep | -0.039 |
Token | red |
Token ID | 2266 |
Feature activation | +3.67e-01 |
Pos logit contributions | |
breath | +0.023 |
common | +0.018 |
job | +0.017 |
long | +0.017 |
sleep | +0.017 |
Neg logit contributions | |
boards | -0.027 |
Did | -0.027 |
auc | -0.026 |
uit | -0.026 |
cher | -0.026 |
Token | tie |
Token ID | 9839 |
Feature activation | +4.80e+00 |
Pos logit contributions | |
Did | +0.009 |
cher | +0.009 |
boards | +0.009 |
auc | +0.008 |
washer | +0.008 |
Neg logit contributions | |
breath | -0.009 |
job | -0.007 |
long | -0.007 |
common | -0.007 |
sleep | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | +6.19e-01 |
Pos logit contributions | |
Did | +0.153 |
ains | +0.146 |
auc | +0.146 |
Is | +0.145 |
Which | +0.143 |
Neg logit contributions | |
breath | -0.072 |
long | -0.057 |
home | -0.052 |
sleep | -0.052 |
time | -0.051 |
Token | He |
Token ID | 679 |
Feature activation | +8.48e-01 |
Pos logit contributions | |
boards | +0.013 |
Did | +0.013 |
washer | +0.012 |
auc | +0.012 |
uit | +0.012 |
Neg logit contributions | |
breath | -0.016 |
common | -0.013 |
job | -0.012 |
sleep | -0.012 |
care | -0.012 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.64e+00 |
Pos logit contributions | |
boards | +0.019 |
auc | +0.018 |
uit | +0.018 |
Did | +0.018 |
cher | +0.018 |
Neg logit contributions | |
breath | -0.021 |
common | -0.017 |
long | -0.016 |
sleep | -0.016 |
job | -0.016 |
Token | it |
Token ID | 340 |
Feature activation | +3.17e+00 |
Pos logit contributions | |
Did | +0.238 |
ode | +0.233 |
cher | +0.227 |
boards | +0.226 |
auc | +0.226 |
Neg logit contributions | |
breath | -0.129 |
long | -0.103 |
home | -0.099 |
care | -0.097 |
time | -0.090 |
Token | up |
Token ID | 510 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
Did | +0.085 |
boards | +0.080 |
auc | +0.079 |
cher | +0.078 |
ains | +0.076 |
Neg logit contributions | |
breath | -0.070 |
long | -0.054 |
job | -0.054 |
home | -0.053 |
sleep | -0.052 |
Token | and |
Token ID | 290 |
Feature activation | -2.96e-01 |
Pos logit contributions | |
breath | +0.049 |
common | +0.039 |
sleep | +0.037 |
job | +0.037 |
care | +0.036 |
Neg logit contributions | |
boards | -0.055 |
uit | -0.053 |
Did | -0.053 |
auc | -0.052 |
washer | -0.052 |
Token | looked |
Token ID | 3114 |
Feature activation | +2.28e+00 |
Pos logit contributions | |
breath | +0.006 |
happily | +0.005 |
work | +0.005 |
care | +0.005 |
sleep | +0.005 |
Neg logit contributions | |
auc | -0.008 |
boards | -0.008 |
uit | -0.008 |
ains | -0.007 |
ode | -0.007 |
Token | at |
Token ID | 379 |
Feature activation | +2.16e+00 |
Pos logit contributions | |
boards | +0.053 |
Did | +0.051 |
auc | +0.051 |
washer | +0.050 |
uit | +0.050 |
Neg logit contributions | |
breath | -0.052 |
common | -0.043 |
job | -0.040 |
sleep | -0.040 |
long | -0.039 |
Token | was |
Token ID | 373 |
Feature activation | +2.34e+00 |
Pos logit contributions | |
register | +0.058 |
cher | +0.055 |
uit | +0.054 |
auc | +0.054 |
ains | +0.053 |
Neg logit contributions | |
breath | -0.046 |
long | -0.040 |
time | -0.039 |
common | -0.036 |
sleep | -0.035 |
Token | a |
Token ID | 257 |
Feature activation | -2.13e-02 |
Pos logit contributions | |
boards | +0.064 |
cher | +0.063 |
register | +0.061 |
auc | +0.061 |
washer | +0.060 |
Neg logit contributions | |
breath | -0.047 |
long | -0.044 |
time | -0.042 |
common | -0.039 |
sleep | -0.037 |
Token | prize |
Token ID | 11596 |
Feature activation | +5.15e-01 |
Pos logit contributions | |
breath | +0.000 |
long | +0.000 |
common | +0.000 |
job | +0.000 |
sleep | +0.000 |
Neg logit contributions | |
boards | -0.001 |
uit | -0.001 |
Did | -0.001 |
auc | -0.001 |
cher | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +3.63e-01 |
Pos logit contributions | |
Did | +0.018 |
ains | +0.017 |
boards | +0.017 |
Does | +0.016 |
Is | +0.016 |
Neg logit contributions | |
breath | -0.007 |
long | -0.005 |
common | -0.004 |
sleep | -0.004 |
home | -0.004 |
Token | He |
Token ID | 679 |
Feature activation | +6.74e-01 |
Pos logit contributions | |
cher | +0.009 |
uit | +0.009 |
Did | +0.009 |
auc | +0.009 |
register | +0.009 |
Neg logit contributions | |
breath | -0.007 |
common | -0.006 |
long | -0.006 |
sleep | -0.006 |
excitement | -0.006 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.61e+00 |
Pos logit contributions | |
boards | +0.016 |
auc | +0.015 |
uit | +0.015 |
Did | +0.015 |
cher | +0.015 |
Neg logit contributions | |
breath | -0.016 |
common | -0.012 |
long | -0.012 |
sleep | -0.012 |
job | -0.012 |
Token | it |
Token ID | 340 |
Feature activation | +3.18e+00 |
Pos logit contributions | |
Did | +0.241 |
auc | +0.234 |
boards | +0.233 |
ode | +0.230 |
washer | +0.229 |
Neg logit contributions | |
breath | -0.126 |
long | -0.098 |
home | -0.090 |
care | -0.089 |
common | -0.086 |
Token | up |
Token ID | 510 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
Did | +0.095 |
boards | +0.089 |
auc | +0.087 |
cher | +0.087 |
ode | +0.085 |
Neg logit contributions | |
breath | -0.064 |
home | -0.049 |
long | -0.048 |
job | -0.046 |
sleep | -0.046 |
Token | and |
Token ID | 290 |
Feature activation | -1.53e-01 |
Pos logit contributions | |
breath | +0.040 |
common | +0.030 |
long | +0.029 |
job | +0.028 |
sleep | +0.028 |
Neg logit contributions | |
boards | -0.054 |
Did | -0.053 |
auc | -0.053 |
uit | -0.052 |
cher | -0.052 |
Token | knew |
Token ID | 2993 |
Feature activation | +2.18e+00 |
Pos logit contributions | |
breath | +0.003 |
happily | +0.003 |
work | +0.002 |
care | +0.002 |
sleep | +0.002 |
Neg logit contributions | |
auc | -0.004 |
boards | -0.004 |
ains | -0.004 |
uit | -0.004 |
register | -0.004 |
Token | he |
Token ID | 339 |
Feature activation | +9.07e-01 |
Pos logit contributions | |
boards | +0.075 |
auc | +0.075 |
Did | +0.075 |
ode | +0.074 |
ains | +0.073 |
Neg logit contributions | |
breath | -0.022 |
long | -0.021 |
home | -0.017 |
happily | -0.017 |
," | -0.016 |
Token | in |
Token ID | 287 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
Did | +0.016 |
auc | +0.015 |
cher | +0.015 |
uit | +0.015 |
boards | +0.015 |
Neg logit contributions | |
breath | -0.010 |
long | -0.008 |
work | -0.007 |
sleep | -0.007 |
job | -0.007 |
Token | the |
Token ID | 262 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
breath | +0.046 |
job | +0.037 |
common | +0.037 |
anna | +0.036 |
sleep | +0.036 |
Neg logit contributions | |
Did | -0.044 |
uit | -0.044 |
boards | -0.043 |
auc | -0.042 |
ting | -0.042 |
Token | water |
Token ID | 1660 |
Feature activation | +1.82e+00 |
Pos logit contributions | |
cher | +0.040 |
uit | +0.038 |
Did | +0.036 |
asing | +0.036 |
we | +0.036 |
Neg logit contributions | |
breath | -0.023 |
time | -0.020 |
long | -0.020 |
night | -0.019 |
home | -0.019 |
Token | . |
Token ID | 13 |
Feature activation | +5.40e-01 |
Pos logit contributions | |
cher | +0.062 |
Did | +0.062 |
Which | +0.060 |
uit | +0.059 |
boards | +0.059 |
Neg logit contributions | |
breath | -0.021 |
long | -0.013 |
home | -0.013 |
time | -0.013 |
sleep | -0.013 |
Token | He |
Token ID | 679 |
Feature activation | +8.38e-01 |
Pos logit contributions | |
boards | +0.012 |
Did | +0.012 |
auc | +0.012 |
uit | +0.011 |
washer | +0.011 |
Neg logit contributions | |
breath | -0.014 |
common | -0.011 |
job | -0.010 |
care | -0.010 |
sleep | -0.010 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.60e+00 |
Pos logit contributions | |
boards | +0.018 |
Did | +0.018 |
auc | +0.017 |
uit | +0.017 |
washer | +0.017 |
Neg logit contributions | |
breath | -0.021 |
common | -0.017 |
job | -0.016 |
sleep | -0.016 |
long | -0.016 |
Token | it |
Token ID | 340 |
Feature activation | +3.06e+00 |
Pos logit contributions | |
Did | +0.220 |
washer | +0.216 |
boards | +0.216 |
auc | +0.215 |
cher | +0.215 |
Neg logit contributions | |
breath | -0.138 |
long | -0.108 |
home | -0.103 |
care | -0.100 |
common | -0.100 |
Token | up |
Token ID | 510 |
Feature activation | -2.35e+00 |
Pos logit contributions | |
Did | +0.083 |
boards | +0.079 |
cher | +0.078 |
auc | +0.077 |
register | +0.075 |
Neg logit contributions | |
breath | -0.068 |
long | -0.051 |
home | -0.051 |
job | -0.049 |
sleep | -0.049 |
Token | and |
Token ID | 290 |
Feature activation | -2.75e-01 |
Pos logit contributions | |
breath | +0.054 |
common | +0.046 |
sleep | +0.044 |
care | +0.043 |
job | +0.043 |
Neg logit contributions | |
uit | -0.051 |
boards | -0.051 |
Did | -0.049 |
auc | -0.049 |
washer | -0.048 |
Token | showed |
Token ID | 3751 |
Feature activation | +5.43e+00 |
Pos logit contributions | |
breath | +0.006 |
happily | +0.005 |
common | +0.004 |
sleep | +0.004 |
work | +0.004 |
Neg logit contributions | |
boards | -0.007 |
auc | -0.007 |
uit | -0.007 |
register | -0.007 |
ains | -0.007 |
Token | it |
Token ID | 340 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
boards | +0.127 |
auc | +0.123 |
uit | +0.122 |
Did | +0.121 |
register | +0.119 |
Neg logit contributions | |
breath | -0.124 |
common | -0.100 |
job | -0.095 |
long | -0.092 |
sleep | -0.092 |
Token | in |
Token ID | 287 |
Feature activation | -1.74e+00 |
Pos logit contributions | |
auc | +0.087 |
boards | +0.085 |
ains | +0.083 |
uit | +0.082 |
Did | +0.082 |
Neg logit contributions | |
breath | -0.063 |
long | -0.053 |
sleep | -0.049 |
time | -0.048 |
," | -0.047 |
Token | the |
Token ID | 262 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
breath | +0.037 |
common | +0.029 |
job | +0.027 |
sleep | +0.027 |
long | +0.027 |
Neg logit contributions | |
boards | -0.044 |
Did | -0.043 |
auc | -0.043 |
uit | -0.043 |
cher | -0.042 |
Token | grass |
Token ID | 8701 |
Feature activation | -5.06e-01 |
Pos logit contributions | |
cher | +0.055 |
we | +0.049 |
bled | +0.049 |
boards | +0.049 |
uit | +0.049 |
Neg logit contributions | |
breath | -0.026 |
long | -0.026 |
home | -0.026 |
time | -0.025 |
job | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | +8.01e-01 |
Pos logit contributions | |
breath | +0.005 |
time | +0.003 |
home | +0.003 |
long | +0.003 |
sleep | +0.003 |
Neg logit contributions | |
cher | -0.019 |
Which | -0.018 |
register | -0.018 |
ode | -0.018 |
Did | -0.017 |
Token | He |
Token ID | 679 |
Feature activation | +8.37e-01 |
Pos logit contributions | |
boards | +0.020 |
cher | +0.020 |
auc | +0.020 |
register | +0.020 |
Did | +0.020 |
Neg logit contributions | |
breath | -0.017 |
common | -0.013 |
long | -0.013 |
sleep | -0.013 |
job | -0.013 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.60e+00 |
Pos logit contributions | |
boards | +0.017 |
Did | +0.017 |
uit | +0.017 |
auc | +0.017 |
cher | +0.016 |
Neg logit contributions | |
breath | -0.022 |
common | -0.018 |
job | -0.017 |
sleep | -0.017 |
long | -0.017 |
Token | it |
Token ID | 340 |
Feature activation | +3.20e+00 |
Pos logit contributions | |
auc | +0.242 |
boards | +0.236 |
Did | +0.235 |
uit | +0.233 |
cher | +0.232 |
Neg logit contributions | |
breath | -0.126 |
long | -0.094 |
care | -0.087 |
home | -0.087 |
common | -0.084 |
Token | up |
Token ID | 510 |
Feature activation | -2.31e+00 |
Pos logit contributions | |
Did | +0.077 |
boards | +0.076 |
auc | +0.076 |
cher | +0.075 |
uit | +0.073 |
Neg logit contributions | |
breath | -0.075 |
long | -0.058 |
common | -0.057 |
job | -0.057 |
sleep | -0.056 |
Token | and |
Token ID | 290 |
Feature activation | -1.78e-01 |
Pos logit contributions | |
breath | +0.053 |
common | +0.044 |
sleep | +0.042 |
job | +0.041 |
care | +0.041 |
Neg logit contributions | |
boards | -0.052 |
uit | -0.051 |
Did | -0.050 |
washer | -0.049 |
auc | -0.048 |
Token | looked |
Token ID | 3114 |
Feature activation | +2.43e+00 |
Pos logit contributions | |
breath | +0.004 |
work | +0.003 |
sleep | +0.003 |
happily | +0.003 |
care | +0.003 |
Neg logit contributions | |
auc | -0.005 |
boards | -0.005 |
uit | -0.005 |
register | -0.005 |
ains | -0.004 |
Token | at |
Token ID | 379 |
Feature activation | +2.37e+00 |
Pos logit contributions | |
boards | +0.051 |
Did | +0.050 |
auc | +0.048 |
washer | +0.048 |
cher | +0.048 |
Neg logit contributions | |
breath | -0.059 |
common | -0.050 |
job | -0.047 |
sleep | -0.046 |
purpose | -0.046 |
Token | a |
Token ID | 257 |
Feature activation | +3.92e-01 |
Pos logit contributions | |
washer | +0.080 |
cher | +0.078 |
boards | +0.077 |
register | +0.076 |
box | +0.074 |
Neg logit contributions | |
breath | -0.102 |
common | -0.095 |
time | -0.091 |
care | -0.089 |
long | -0.089 |
Token | light |
Token ID | 1657 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
boards | +0.011 |
Did | +0.011 |
auc | +0.010 |
uit | +0.010 |
cher | +0.010 |
Neg logit contributions | |
breath | -0.008 |
long | -0.006 |
common | -0.006 |
job | -0.006 |
sleep | -0.006 |
Token | novel |
Token ID | 5337 |
Feature activation | +1.22e-01 |
Pos logit contributions | |
boards | +0.041 |
Did | +0.041 |
uit | +0.039 |
cher | +0.038 |
register | +0.038 |
Neg logit contributions | |
breath | -0.038 |
common | -0.031 |
job | -0.030 |
long | -0.029 |
sleep | -0.029 |
Token | ! |
Token ID | 0 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
boards | +0.004 |
Did | +0.004 |
auc | +0.004 |
uit | +0.004 |
ting | +0.003 |
Neg logit contributions | |
breath | -0.002 |
common | -0.001 |
long | -0.001 |
job | -0.001 |
sleep | -0.001 |
Token | She |
Token ID | 1375 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
boards | +0.041 |
washer | +0.039 |
register | +0.039 |
uit | +0.039 |
cher | +0.038 |
Neg logit contributions | |
breath | -0.042 |
sleep | -0.037 |
Today | -0.037 |
common | -0.037 |
long | -0.036 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.60e+00 |
Pos logit contributions | |
Did | +0.034 |
boards | +0.034 |
uit | +0.034 |
auc | +0.033 |
cher | +0.032 |
Neg logit contributions | |
breath | -0.042 |
sleep | -0.036 |
care | -0.035 |
common | -0.035 |
long | -0.035 |
Token | it |
Token ID | 340 |
Feature activation | +2.92e+00 |
Pos logit contributions | |
Did | +0.211 |
boards | +0.207 |
auc | +0.206 |
washer | +0.204 |
uit | +0.204 |
Neg logit contributions | |
breath | -0.141 |
long | -0.115 |
care | -0.109 |
common | -0.108 |
home | -0.107 |
Token | up |
Token ID | 510 |
Feature activation | -2.46e+00 |
Pos logit contributions | |
Did | +0.068 |
boards | +0.067 |
auc | +0.065 |
cher | +0.064 |
uit | +0.064 |
Neg logit contributions | |
breath | -0.068 |
common | -0.055 |
long | -0.053 |
job | -0.053 |
sleep | -0.053 |
Token | and |
Token ID | 290 |
Feature activation | +2.43e-01 |
Pos logit contributions | |
breath | +0.055 |
common | +0.043 |
sleep | +0.042 |
job | +0.041 |
care | +0.041 |
Neg logit contributions | |
boards | -0.057 |
Did | -0.056 |
uit | -0.056 |
auc | -0.055 |
washer | -0.055 |
Token | carried |
Token ID | 5281 |
Feature activation | +2.87e+00 |
Pos logit contributions | |
uit | +0.006 |
auc | +0.006 |
boards | +0.006 |
Did | +0.006 |
register | +0.006 |
Neg logit contributions | |
breath | -0.005 |
care | -0.004 |
sleep | -0.004 |
long | -0.004 |
work | -0.004 |
Token | it |
Token ID | 340 |
Feature activation | +5.57e-01 |
Pos logit contributions | |
boards | +0.065 |
uit | +0.063 |
Did | +0.062 |
cher | +0.062 |
auc | +0.061 |
Neg logit contributions | |
breath | -0.066 |
common | -0.054 |
job | -0.051 |
sleep | -0.050 |
long | -0.049 |
Token | on |
Token ID | 319 |
Feature activation | -2.30e+00 |
Pos logit contributions | |
boards | +0.116 |
auc | +0.114 |
ains | +0.112 |
Did | +0.109 |
cher | +0.109 |
Neg logit contributions | |
breath | -0.095 |
long | -0.077 |
sleep | -0.071 |
job | -0.070 |
home | -0.068 |
Token | the |
Token ID | 262 |
Feature activation | +2.64e+00 |
Pos logit contributions | |
breath | +0.053 |
common | +0.043 |
job | +0.040 |
care | +0.040 |
sleep | +0.040 |
Neg logit contributions | |
boards | -0.052 |
uit | -0.051 |
Did | -0.051 |
auc | -0.049 |
register | -0.049 |
Token | ground |
Token ID | 2323 |
Feature activation | +2.85e-01 |
Pos logit contributions | |
boards | +0.061 |
cher | +0.061 |
Did | +0.060 |
auc | +0.060 |
uit | +0.059 |
Neg logit contributions | |
breath | -0.061 |
long | -0.051 |
job | -0.050 |
common | -0.049 |
sleep | -0.049 |
Token | . |
Token ID | 13 |
Feature activation | +5.09e-01 |
Pos logit contributions | |
cher | +0.010 |
Did | +0.010 |
boards | +0.009 |
uit | +0.009 |
register | +0.009 |
Neg logit contributions | |
breath | -0.003 |
long | -0.002 |
sleep | -0.002 |
time | -0.002 |
common | -0.002 |
Token | She |
Token ID | 1375 |
Feature activation | +6.21e-01 |
Pos logit contributions | |
boards | +0.011 |
Did | +0.011 |
uit | +0.011 |
auc | +0.010 |
washer | +0.010 |
Neg logit contributions | |
breath | -0.013 |
common | -0.011 |
job | -0.010 |
sleep | -0.010 |
long | -0.010 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.59e+00 |
Pos logit contributions | |
Did | +0.012 |
boards | +0.012 |
washer | +0.011 |
uit | +0.011 |
cher | +0.011 |
Neg logit contributions | |
breath | -0.017 |
common | -0.014 |
job | -0.014 |
sleep | -0.013 |
care | -0.013 |
Token | it |
Token ID | 340 |
Feature activation | +3.12e+00 |
Pos logit contributions | |
boards | +0.216 |
auc | +0.215 |
Did | +0.214 |
uit | +0.210 |
cher | +0.209 |
Neg logit contributions | |
breath | -0.141 |
long | -0.105 |
care | -0.102 |
common | -0.102 |
job | -0.101 |
Token | up |
Token ID | 510 |
Feature activation | -2.19e+00 |
Pos logit contributions | |
boards | +0.071 |
Did | +0.070 |
auc | +0.069 |
cher | +0.067 |
uit | +0.067 |
Neg logit contributions | |
breath | -0.076 |
common | -0.060 |
long | -0.059 |
job | -0.059 |
sleep | -0.058 |
Token | and |
Token ID | 290 |
Feature activation | -2.77e-01 |
Pos logit contributions | |
breath | +0.051 |
common | +0.044 |
sleep | +0.041 |
job | +0.041 |
care | +0.040 |
Neg logit contributions | |
boards | -0.047 |
uit | -0.047 |
Did | -0.046 |
lying | -0.045 |
washer | -0.044 |
Token | showed |
Token ID | 3751 |
Feature activation | +5.45e+00 |
Pos logit contributions | |
breath | +0.006 |
sleep | +0.005 |
common | +0.005 |
work | +0.005 |
care | +0.005 |
Neg logit contributions | |
auc | -0.007 |
boards | -0.007 |
uit | -0.007 |
register | -0.006 |
cher | -0.006 |
Token | it |
Token ID | 340 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
boards | +0.104 |
uit | +0.102 |
ting | +0.101 |
register | +0.099 |
auc | +0.098 |
Neg logit contributions | |
breath | -0.137 |
common | -0.121 |
job | -0.109 |
appetite | -0.107 |
sleep | -0.107 |
Token | found |
Token ID | 1043 |
Feature activation | +6.45e+00 |
Pos logit contributions | |
boards | +0.005 |
Did | +0.005 |
auc | +0.005 |
uit | +0.005 |
washer | +0.004 |
Neg logit contributions | |
breath | -0.006 |
common | -0.005 |
job | -0.005 |
sleep | -0.005 |
long | -0.005 |
Token | something |
Token ID | 1223 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
ains | +0.208 |
boards | +0.202 |
ode | +0.196 |
auc | +0.192 |
cher | +0.191 |
Neg logit contributions | |
breath | -0.103 |
long | -0.084 |
home | -0.080 |
care | -0.076 |
sleep | -0.074 |
Token | disgusting |
Token ID | 23374 |
Feature activation | +2.23e-01 |
Pos logit contributions | |
ains | +0.065 |
uit | +0.063 |
Did | +0.063 |
ode | +0.062 |
register | +0.061 |
Neg logit contributions | |
long | -0.032 |
breath | -0.029 |
common | -0.025 |
time | -0.025 |
idea | -0.023 |
Token | . |
Token ID | 13 |
Feature activation | +3.06e-01 |
Pos logit contributions | |
Did | +0.007 |
cher | +0.007 |
ains | +0.007 |
uit | +0.007 |
register | +0.007 |
Neg logit contributions | |
breath | -0.003 |
time | -0.002 |
sleep | -0.002 |
long | -0.002 |
idea | -0.002 |
Token | He |
Token ID | 679 |
Feature activation | +7.08e-01 |
Pos logit contributions | |
boards | +0.006 |
Did | +0.006 |
uit | +0.006 |
washer | +0.006 |
auc | +0.006 |
Neg logit contributions | |
breath | -0.008 |
common | -0.007 |
job | -0.006 |
care | -0.006 |
sleep | -0.006 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.59e+00 |
Pos logit contributions | |
boards | +0.015 |
auc | +0.015 |
uit | +0.015 |
Did | +0.015 |
cher | +0.014 |
Neg logit contributions | |
breath | -0.018 |
common | -0.015 |
sleep | -0.014 |
long | -0.014 |
job | -0.014 |
Token | it |
Token ID | 340 |
Feature activation | +3.11e+00 |
Pos logit contributions | |
Did | +0.234 |
boards | +0.230 |
cher | +0.227 |
auc | +0.225 |
washer | +0.225 |
Neg logit contributions | |
breath | -0.132 |
long | -0.104 |
care | -0.099 |
home | -0.097 |
time | -0.090 |
Token | up |
Token ID | 510 |
Feature activation | -2.25e+00 |
Pos logit contributions | |
Did | +0.095 |
boards | +0.090 |
cher | +0.089 |
auc | +0.088 |
ains | +0.088 |
Neg logit contributions | |
breath | -0.060 |
home | -0.045 |
long | -0.044 |
job | -0.042 |
happily | -0.042 |
Token | and |
Token ID | 290 |
Feature activation | -1.60e-01 |
Pos logit contributions | |
breath | +0.046 |
common | +0.035 |
long | +0.033 |
job | +0.033 |
sleep | +0.033 |
Neg logit contributions | |
boards | -0.060 |
Did | -0.058 |
auc | -0.058 |
uit | -0.057 |
cher | -0.057 |
Token | he |
Token ID | 339 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
breath | +0.003 |
sleep | +0.003 |
care | +0.003 |
long | +0.003 |
happily | +0.003 |
Neg logit contributions | |
auc | -0.004 |
boards | -0.004 |
uit | -0.004 |
register | -0.004 |
ains | -0.004 |
Token | didn |
Token ID | 1422 |
Feature activation | +8.31e-01 |
Pos logit contributions | |
boards | +0.024 |
uit | +0.024 |
auc | +0.024 |
Did | +0.023 |
register | +0.023 |
Neg logit contributions | |
breath | -0.025 |
common | -0.020 |
long | -0.020 |
sleep | -0.019 |
care | -0.019 |
Token | on |
Token ID | 319 |
Feature activation | -2.73e+00 |
Pos logit contributions | |
boards | +0.102 |
uit | +0.097 |
register | +0.097 |
cher | +0.096 |
Did | +0.096 |
Neg logit contributions | |
breath | -0.083 |
long | -0.065 |
time | -0.060 |
," | -0.059 |
, | -0.057 |
Token | the |
Token ID | 262 |
Feature activation | +2.80e+00 |
Pos logit contributions | |
breath | +0.057 |
common | +0.044 |
long | +0.042 |
sleep | +0.041 |
job | +0.041 |
Neg logit contributions | |
boards | -0.071 |
Did | -0.069 |
auc | -0.069 |
uit | -0.068 |
cher | -0.068 |
Token | ground |
Token ID | 2323 |
Feature activation | +6.42e-01 |
Pos logit contributions | |
Did | +0.065 |
boards | +0.065 |
auc | +0.064 |
cher | +0.064 |
uit | +0.062 |
Neg logit contributions | |
breath | -0.067 |
long | -0.055 |
job | -0.053 |
sleep | -0.053 |
home | -0.051 |
Token | . |
Token ID | 13 |
Feature activation | +4.26e-01 |
Pos logit contributions | |
cher | +0.023 |
register | +0.022 |
auc | +0.022 |
Which | +0.021 |
Did | +0.021 |
Neg logit contributions | |
breath | -0.008 |
long | -0.005 |
time | -0.005 |
sleep | -0.005 |
home | -0.004 |
Token | He |
Token ID | 679 |
Feature activation | +8.66e-01 |
Pos logit contributions | |
boards | +0.008 |
Did | +0.008 |
uit | +0.008 |
auc | +0.007 |
ting | +0.007 |
Neg logit contributions | |
breath | -0.012 |
common | -0.010 |
job | -0.009 |
sleep | -0.009 |
care | -0.009 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.58e+00 |
Pos logit contributions | |
Did | +0.017 |
boards | +0.016 |
washer | +0.015 |
lying | +0.015 |
uit | +0.014 |
Neg logit contributions | |
breath | -0.024 |
common | -0.021 |
job | -0.020 |
sleep | -0.019 |
purpose | -0.019 |
Token | it |
Token ID | 340 |
Feature activation | +3.23e+00 |
Pos logit contributions | |
boards | +0.222 |
Did | +0.220 |
auc | +0.219 |
cher | +0.216 |
washer | +0.215 |
Neg logit contributions | |
breath | -0.139 |
long | -0.102 |
care | -0.101 |
job | -0.097 |
common | -0.097 |
Token | up |
Token ID | 510 |
Feature activation | -2.32e+00 |
Pos logit contributions | |
Did | +0.077 |
boards | +0.076 |
auc | +0.074 |
cher | +0.074 |
uit | +0.073 |
Neg logit contributions | |
breath | -0.077 |
long | -0.059 |
common | -0.058 |
job | -0.058 |
sleep | -0.057 |
Token | with |
Token ID | 351 |
Feature activation | +2.65e+00 |
Pos logit contributions | |
breath | +0.051 |
common | +0.043 |
sleep | +0.041 |
job | +0.040 |
care | +0.039 |
Neg logit contributions | |
boards | -0.054 |
uit | -0.053 |
Did | -0.052 |
auc | -0.051 |
washer | -0.051 |
Token | his |
Token ID | 465 |
Feature activation | +2.14e-01 |
Pos logit contributions | |
boards | +0.063 |
uit | +0.062 |
Did | +0.062 |
auc | +0.061 |
cher | +0.060 |
Neg logit contributions | |
breath | -0.058 |
common | -0.047 |
job | -0.044 |
sleep | -0.044 |
long | -0.043 |
Token | mouth |
Token ID | 5422 |
Feature activation | +1.91e+00 |
Pos logit contributions | |
uit | +0.005 |
boards | +0.005 |
Did | +0.005 |
cher | +0.005 |
sticking | +0.005 |
Neg logit contributions | |
breath | -0.004 |
common | -0.004 |
sleep | -0.003 |
job | -0.003 |
care | -0.003 |
Token | an |
Token ID | 281 |
Feature activation | +9.67e-01 |
Pos logit contributions | |
ains | +0.208 |
cher | +0.206 |
ode | +0.206 |
ppers | +0.201 |
can | +0.199 |
Neg logit contributions | |
long | -0.065 |
sight | -0.055 |
breath | -0.051 |
time | -0.049 |
home | -0.049 |
Token | old |
Token ID | 1468 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
boards | +0.025 |
Did | +0.024 |
uit | +0.024 |
auc | +0.024 |
cher | +0.024 |
Neg logit contributions | |
breath | -0.020 |
common | -0.016 |
job | -0.015 |
sleep | -0.015 |
long | -0.015 |
Token | bone |
Token ID | 9970 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
breath | +0.038 |
common | +0.031 |
job | +0.029 |
long | +0.029 |
sleep | +0.029 |
Neg logit contributions | |
boards | -0.041 |
Did | -0.040 |
auc | -0.040 |
uit | -0.040 |
cher | -0.039 |
Token | . |
Token ID | 13 |
Feature activation | +6.13e-01 |
Pos logit contributions | |
ains | +0.050 |
auc | +0.048 |
Which | +0.048 |
ode | +0.048 |
Did | +0.047 |
Neg logit contributions | |
breath | -0.024 |
long | -0.023 |
, | -0.021 |
time | -0.020 |
day | -0.020 |
Token | He |
Token ID | 679 |
Feature activation | +8.70e-01 |
Pos logit contributions | |
auc | +0.015 |
Did | +0.015 |
cher | +0.015 |
boards | +0.014 |
uit | +0.014 |
Neg logit contributions | |
breath | -0.013 |
long | -0.011 |
sleep | -0.011 |
common | -0.011 |
job | -0.010 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.58e+00 |
Pos logit contributions | |
boards | +0.017 |
auc | +0.017 |
Did | +0.017 |
uit | +0.017 |
cher | +0.016 |
Neg logit contributions | |
breath | -0.023 |
common | -0.019 |
long | -0.018 |
sleep | -0.018 |
job | -0.018 |
Token | it |
Token ID | 340 |
Feature activation | +3.18e+00 |
Pos logit contributions | |
auc | +0.235 |
Did | +0.229 |
boards | +0.228 |
washer | +0.224 |
uit | +0.223 |
Neg logit contributions | |
breath | -0.130 |
long | -0.099 |
care | -0.093 |
home | -0.091 |
job | -0.089 |
Token | up |
Token ID | 510 |
Feature activation | -2.25e+00 |
Pos logit contributions | |
Did | +0.099 |
auc | +0.094 |
cher | +0.093 |
boards | +0.091 |
ains | +0.090 |
Neg logit contributions | |
breath | -0.060 |
home | -0.046 |
long | -0.045 |
happily | -0.042 |
job | -0.042 |
Token | and |
Token ID | 290 |
Feature activation | -1.56e-01 |
Pos logit contributions | |
breath | +0.044 |
common | +0.033 |
long | +0.032 |
job | +0.031 |
sleep | +0.031 |
Neg logit contributions | |
boards | -0.062 |
auc | -0.060 |
Did | -0.060 |
cher | -0.059 |
uit | -0.059 |
Token | was |
Token ID | 373 |
Feature activation | +2.06e+00 |
Pos logit contributions | |
breath | +0.003 |
happily | +0.003 |
work | +0.002 |
sleep | +0.002 |
long | +0.002 |
Neg logit contributions | |
auc | -0.004 |
uit | -0.004 |
ains | -0.004 |
boards | -0.004 |
register | -0.004 |
Token | about |
Token ID | 546 |
Feature activation | +3.10e-01 |
Pos logit contributions | |
boards | +0.052 |
auc | +0.050 |
uit | +0.049 |
Did | +0.049 |
ains | +0.048 |
Neg logit contributions | |
breath | -0.046 |
long | -0.037 |
sleep | -0.035 |
common | -0.035 |
job | -0.034 |
Token | found |
Token ID | 1043 |
Feature activation | +6.06e+00 |
Pos logit contributions | |
breath | +0.012 |
common | +0.009 |
long | +0.009 |
sleep | +0.009 |
job | +0.009 |
Neg logit contributions | |
boards | -0.011 |
uit | -0.011 |
auc | -0.011 |
Did | -0.011 |
cher | -0.011 |
Token | an |
Token ID | 281 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
boards | +0.174 |
cher | +0.168 |
Did | +0.167 |
ains | +0.166 |
auc | +0.164 |
Neg logit contributions | |
breath | -0.116 |
long | -0.093 |
home | -0.089 |
sleep | -0.086 |
care | -0.085 |
Token | arrow |
Token ID | 15452 |
Feature activation | +2.84e+00 |
Pos logit contributions | |
boards | +0.032 |
Did | +0.031 |
uit | +0.031 |
auc | +0.031 |
cher | +0.031 |
Neg logit contributions | |
breath | -0.036 |
common | -0.030 |
job | -0.028 |
sleep | -0.028 |
long | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | +3.28e-01 |
Pos logit contributions | |
auc | +0.087 |
boards | +0.085 |
Did | +0.084 |
ains | +0.083 |
uit | +0.082 |
Neg logit contributions | |
breath | -0.044 |
long | -0.035 |
sleep | -0.031 |
time | -0.031 |
," | -0.031 |
Token | He |
Token ID | 679 |
Feature activation | +7.66e-01 |
Pos logit contributions | |
cher | +0.008 |
auc | +0.008 |
register | +0.008 |
Did | +0.008 |
boards | +0.008 |
Neg logit contributions | |
breath | -0.007 |
long | -0.005 |
common | -0.005 |
sleep | -0.005 |
job | -0.005 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.57e+00 |
Pos logit contributions | |
boards | +0.015 |
Did | +0.015 |
uit | +0.014 |
auc | +0.014 |
cher | +0.014 |
Neg logit contributions | |
breath | -0.021 |
common | -0.017 |
job | -0.016 |
sleep | -0.016 |
long | -0.016 |
Token | it |
Token ID | 340 |
Feature activation | +3.10e+00 |
Pos logit contributions | |
auc | +0.227 |
boards | +0.223 |
Did | +0.221 |
uit | +0.219 |
cher | +0.219 |
Neg logit contributions | |
breath | -0.135 |
long | -0.102 |
care | -0.095 |
home | -0.095 |
common | -0.094 |
Token | up |
Token ID | 510 |
Feature activation | -2.41e+00 |
Pos logit contributions | |
Did | +0.074 |
boards | +0.073 |
auc | +0.072 |
cher | +0.071 |
uit | +0.070 |
Neg logit contributions | |
breath | -0.073 |
long | -0.057 |
common | -0.056 |
job | -0.056 |
sleep | -0.055 |
Token | and |
Token ID | 290 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
breath | +0.054 |
common | +0.044 |
sleep | +0.042 |
job | +0.042 |
care | +0.041 |
Neg logit contributions | |
boards | -0.056 |
Did | -0.055 |
uit | -0.055 |
auc | -0.053 |
washer | -0.053 |
Token | showed |
Token ID | 3751 |
Feature activation | +5.42e+00 |
Pos logit contributions | |
breath | +0.006 |
work | +0.004 |
happily | +0.004 |
sleep | +0.004 |
care | +0.004 |
Neg logit contributions | |
auc | -0.007 |
boards | -0.007 |
uit | -0.007 |
register | -0.007 |
ains | -0.007 |
Token | it |
Token ID | 340 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
boards | +0.120 |
uit | +0.115 |
Did | +0.114 |
auc | +0.114 |
washer | +0.113 |
Neg logit contributions | |
breath | -0.129 |
common | -0.108 |
job | -0.099 |
sleep | -0.098 |
long | -0.098 |
Token | in |
Token ID | 287 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
boards | +0.069 |
Did | +0.068 |
ains | +0.066 |
cher | +0.066 |
uit | +0.065 |
Neg logit contributions | |
breath | -0.049 |
long | -0.039 |
job | -0.036 |
sleep | -0.036 |
," | -0.035 |
Token | the |
Token ID | 262 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
breath | +0.034 |
common | +0.027 |
job | +0.025 |
sleep | +0.025 |
long | +0.025 |
Neg logit contributions | |
boards | -0.044 |
Did | -0.043 |
uit | -0.043 |
auc | -0.042 |
cher | -0.041 |
Token | ground |
Token ID | 2323 |
Feature activation | +4.70e-01 |
Pos logit contributions | |
cher | +0.046 |
boards | +0.043 |
uit | +0.043 |
Did | +0.041 |
washer | +0.040 |
Neg logit contributions | |
breath | -0.032 |
job | -0.027 |
long | -0.027 |
home | -0.027 |
time | -0.027 |
Token | . |
Token ID | 13 |
Feature activation | +7.11e-01 |
Pos logit contributions | |
cher | +0.015 |
boards | +0.015 |
auc | +0.015 |
uit | +0.015 |
Did | +0.015 |
Neg logit contributions | |
breath | -0.007 |
long | -0.005 |
sleep | -0.004 |
common | -0.004 |
time | -0.004 |
Token | She |
Token ID | 1375 |
Feature activation | +8.18e-01 |
Pos logit contributions | |
boards | +0.014 |
Did | +0.014 |
uit | +0.014 |
auc | +0.013 |
washer | +0.013 |
Neg logit contributions | |
breath | -0.019 |
common | -0.016 |
job | -0.015 |
sleep | -0.015 |
care | -0.015 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.57e+00 |
Pos logit contributions | |
boards | +0.016 |
Did | +0.016 |
auc | +0.016 |
uit | +0.015 |
washer | +0.015 |
Neg logit contributions | |
breath | -0.022 |
common | -0.018 |
job | -0.017 |
sleep | -0.017 |
long | -0.017 |
Token | it |
Token ID | 340 |
Feature activation | +3.13e+00 |
Pos logit contributions | |
boards | +0.227 |
Did | +0.224 |
auc | +0.224 |
cher | +0.223 |
washer | +0.221 |
Neg logit contributions | |
breath | -0.134 |
long | -0.101 |
care | -0.097 |
home | -0.093 |
common | -0.092 |
Token | up |
Token ID | 510 |
Feature activation | -2.18e+00 |
Pos logit contributions | |
boards | +0.072 |
Did | +0.072 |
auc | +0.070 |
cher | +0.070 |
uit | +0.069 |
Neg logit contributions | |
breath | -0.076 |
common | -0.059 |
long | -0.058 |
sleep | -0.058 |
job | -0.058 |
Token | and |
Token ID | 290 |
Feature activation | -1.60e-01 |
Pos logit contributions | |
breath | +0.051 |
common | +0.043 |
job | +0.041 |
sleep | +0.041 |
care | +0.040 |
Neg logit contributions | |
boards | -0.047 |
uit | -0.047 |
Did | -0.046 |
lying | -0.044 |
ting | -0.044 |
Token | put |
Token ID | 1234 |
Feature activation | +4.05e+00 |
Pos logit contributions | |
breath | +0.004 |
sleep | +0.003 |
care | +0.003 |
work | +0.003 |
long | +0.003 |
Neg logit contributions | |
boards | -0.004 |
auc | -0.004 |
uit | -0.004 |
register | -0.004 |
cher | -0.004 |
Token | it |
Token ID | 340 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
boards | +0.119 |
auc | +0.117 |
washer | +0.115 |
Did | +0.114 |
cher | +0.112 |
Neg logit contributions | |
breath | -0.078 |
care | -0.056 |
long | -0.055 |
common | -0.054 |
sleep | -0.052 |
Token | a |
Token ID | 257 |
Feature activation | -4.84e-01 |
Pos logit contributions | |
breath | +0.073 |
common | +0.059 |
job | +0.056 |
sleep | +0.054 |
care | +0.054 |
Neg logit contributions | |
boards | -0.062 |
Did | -0.060 |
uit | -0.060 |
auc | -0.058 |
ting | -0.058 |
Token | flower |
Token ID | 15061 |
Feature activation | +7.22e-01 |
Pos logit contributions | |
breath | +0.011 |
common | +0.009 |
job | +0.008 |
sleep | +0.008 |
care | +0.008 |
Neg logit contributions | |
boards | -0.012 |
Did | -0.011 |
auc | -0.011 |
uit | -0.011 |
washer | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +4.28e-01 |
Pos logit contributions | |
boards | +0.021 |
auc | +0.021 |
Did | +0.021 |
uit | +0.021 |
cher | +0.021 |
Neg logit contributions | |
breath | -0.012 |
sleep | -0.009 |
common | -0.009 |
long | -0.009 |
job | -0.008 |
Token | She |
Token ID | 1375 |
Feature activation | +5.58e-01 |
Pos logit contributions | |
boards | +0.009 |
Did | +0.008 |
uit | +0.008 |
ting | +0.008 |
auc | +0.008 |
Neg logit contributions | |
breath | -0.012 |
common | -0.010 |
job | -0.009 |
accomplishment | -0.009 |
breaths | -0.009 |
Token | gently |
Token ID | 15165 |
Feature activation | -3.71e+00 |
Pos logit contributions | |
boards | +0.011 |
Did | +0.011 |
uit | +0.010 |
auc | +0.010 |
washer | +0.010 |
Neg logit contributions | |
breath | -0.015 |
common | -0.013 |
job | -0.012 |
sleep | -0.012 |
long | -0.012 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.57e+00 |
Pos logit contributions | |
breath | +0.099 |
common | +0.081 |
sleep | +0.078 |
long | +0.078 |
job | +0.077 |
Neg logit contributions | |
boards | -0.074 |
auc | -0.072 |
Did | -0.071 |
uit | -0.071 |
cher | -0.070 |
Token | it |
Token ID | 340 |
Feature activation | +3.02e+00 |
Pos logit contributions | |
boards | +0.194 |
auc | +0.188 |
Did | +0.188 |
uit | +0.186 |
washer | +0.185 |
Neg logit contributions | |
breath | -0.162 |
common | -0.126 |
long | -0.121 |
sleep | -0.119 |
job | -0.119 |
Token | up |
Token ID | 510 |
Feature activation | -2.34e+00 |
Pos logit contributions | |
boards | +0.071 |
Did | +0.070 |
auc | +0.069 |
cher | +0.068 |
uit | +0.067 |
Neg logit contributions | |
breath | -0.072 |
common | -0.056 |
long | -0.056 |
sleep | -0.055 |
job | -0.055 |
Token | and |
Token ID | 290 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
breath | +0.053 |
common | +0.045 |
job | +0.043 |
sleep | +0.042 |
care | +0.041 |
Neg logit contributions | |
boards | -0.052 |
uit | -0.051 |
Did | -0.051 |
ting | -0.049 |
lying | -0.049 |
Token | held |
Token ID | 2714 |
Feature activation | +3.16e+00 |
Pos logit contributions | |
breath | +0.008 |
sleep | +0.006 |
common | +0.006 |
care | +0.006 |
long | +0.006 |
Neg logit contributions | |
boards | -0.008 |
auc | -0.008 |
uit | -0.007 |
register | -0.007 |
washer | -0.007 |
Token | it |
Token ID | 340 |
Feature activation | +1.46e+00 |
Pos logit contributions | |
uit | +0.067 |
boards | +0.067 |
Did | +0.067 |
cher | +0.066 |
ting | +0.066 |
Neg logit contributions | |
breath | -0.073 |
common | -0.062 |
job | -0.059 |
sleep | -0.058 |
long | -0.056 |
Token | on |
Token ID | 319 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
boards | +0.091 |
auc | +0.090 |
ode | +0.090 |
Did | +0.090 |
ains | +0.089 |
Neg logit contributions | |
breath | -0.052 |
long | -0.040 |
," | -0.037 |
time | -0.037 |
sleep | -0.036 |
Token | the |
Token ID | 262 |
Feature activation | +3.28e+00 |
Pos logit contributions | |
breath | +0.060 |
common | +0.047 |
long | +0.045 |
sleep | +0.045 |
job | +0.044 |
Neg logit contributions | |
boards | -0.072 |
Did | -0.070 |
auc | -0.070 |
cher | -0.069 |
uit | -0.069 |
Token | ground |
Token ID | 2323 |
Feature activation | +6.01e-01 |
Pos logit contributions | |
boards | +0.081 |
Did | +0.081 |
auc | +0.081 |
cher | +0.080 |
uit | +0.079 |
Neg logit contributions | |
breath | -0.074 |
long | -0.065 |
sleep | -0.061 |
job | -0.061 |
common | -0.058 |
Token | . |
Token ID | 13 |
Feature activation | +7.56e-01 |
Pos logit contributions | |
cher | +0.021 |
Did | +0.021 |
boards | +0.020 |
register | +0.020 |
uit | +0.020 |
Neg logit contributions | |
breath | -0.007 |
long | -0.004 |
sleep | -0.004 |
time | -0.004 |
common | -0.004 |
Token | She |
Token ID | 1375 |
Feature activation | +9.07e-01 |
Pos logit contributions | |
boards | +0.016 |
Did | +0.015 |
uit | +0.015 |
auc | +0.015 |
cher | +0.015 |
Neg logit contributions | |
breath | -0.020 |
common | -0.016 |
job | -0.015 |
sleep | -0.015 |
long | -0.015 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.57e+00 |
Pos logit contributions | |
boards | +0.020 |
Did | +0.019 |
auc | +0.019 |
uit | +0.019 |
cher | +0.019 |
Neg logit contributions | |
breath | -0.022 |
common | -0.018 |
job | -0.018 |
sleep | -0.017 |
long | -0.017 |
Token | it |
Token ID | 340 |
Feature activation | +3.19e+00 |
Pos logit contributions | |
auc | +0.226 |
Did | +0.225 |
boards | +0.225 |
washer | +0.223 |
uit | +0.219 |
Neg logit contributions | |
breath | -0.134 |
long | -0.105 |
care | -0.098 |
common | -0.097 |
home | -0.094 |
Token | up |
Token ID | 510 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
Did | +0.078 |
boards | +0.077 |
auc | +0.075 |
cher | +0.074 |
uit | +0.073 |
Neg logit contributions | |
breath | -0.075 |
long | -0.058 |
common | -0.058 |
sleep | -0.057 |
job | -0.057 |
Token | and |
Token ID | 290 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
breath | +0.049 |
common | +0.041 |
job | +0.039 |
sleep | +0.039 |
care | +0.038 |
Neg logit contributions | |
boards | -0.046 |
uit | -0.046 |
Did | -0.045 |
lying | -0.044 |
ting | -0.043 |
Token | waved |
Token ID | 26834 |
Feature activation | +6.35e-02 |
Pos logit contributions | |
breath | +0.002 |
care | +0.002 |
work | +0.002 |
sleep | +0.002 |
common | +0.002 |
Neg logit contributions | |
boards | -0.003 |
auc | -0.003 |
uit | -0.003 |
register | -0.003 |
Did | -0.002 |
Token | it |
Token ID | 340 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
ting | +0.001 |
Did | +0.001 |
cher | +0.001 |
can | +0.001 |
lying | +0.001 |
Neg logit contributions | |
breath | -0.002 |
common | -0.002 |
job | -0.001 |
appetite | -0.001 |
discussion | -0.001 |
Token | of |
Token ID | 286 |
Feature activation | +3.57e+00 |
Pos logit contributions | |
Did | +0.154 |
ode | +0.152 |
ains | +0.151 |
cher | +0.146 |
register | +0.144 |
Neg logit contributions | |
breath | -0.100 |
long | -0.091 |
time | -0.087 |
sleep | -0.084 |
," | -0.081 |
Token | a |
Token ID | 257 |
Feature activation | +3.94e-01 |
Pos logit contributions | |
auc | +0.098 |
boards | +0.097 |
washer | +0.095 |
cher | +0.095 |
can | +0.092 |
Neg logit contributions | |
breath | -0.071 |
long | -0.062 |
sleep | -0.059 |
common | -0.058 |
excitement | -0.056 |
Token | skull |
Token ID | 14511 |
Feature activation | +2.53e+00 |
Pos logit contributions | |
boards | +0.010 |
auc | +0.010 |
uit | +0.010 |
Did | +0.010 |
cher | +0.009 |
Neg logit contributions | |
breath | -0.008 |
common | -0.007 |
long | -0.007 |
job | -0.006 |
sleep | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | -2.10e-01 |
Pos logit contributions | |
Did | +0.072 |
boards | +0.071 |
auc | +0.071 |
ains | +0.069 |
uit | +0.068 |
Neg logit contributions | |
breath | -0.042 |
long | -0.036 |
sleep | -0.034 |
time | -0.033 |
," | -0.032 |
Token | cross |
Token ID | 3272 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
breath | +0.005 |
long | +0.004 |
common | +0.004 |
sleep | +0.004 |
care | +0.004 |
Neg logit contributions | |
auc | -0.005 |
boards | -0.005 |
uit | -0.005 |
washer | -0.005 |
register | -0.005 |
Token | bones |
Token ID | 35095 |
Feature activation | +7.56e+00 |
Pos logit contributions | |
boards | +0.050 |
auc | +0.050 |
Did | +0.049 |
ode | +0.049 |
washer | +0.049 |
Neg logit contributions | |
breath | -0.041 |
common | -0.032 |
long | -0.032 |
care | -0.030 |
home | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | +6.41e-01 |
Pos logit contributions | |
ode | +0.194 |
auc | +0.192 |
washer | +0.192 |
uit | +0.189 |
Did | +0.189 |
Neg logit contributions | |
breath | -0.141 |
long | -0.133 |
happily | -0.118 |
time | -0.114 |
," | -0.110 |
Token | She |
Token ID | 1375 |
Feature activation | +4.44e-01 |
Pos logit contributions | |
boards | +0.014 |
Did | +0.014 |
uit | +0.013 |
auc | +0.013 |
washer | +0.013 |
Neg logit contributions | |
breath | -0.017 |
common | -0.013 |
job | -0.013 |
care | -0.012 |
sleep | -0.012 |
Token | knew |
Token ID | 2993 |
Feature activation | +2.17e+00 |
Pos logit contributions | |
boards | +0.010 |
auc | +0.010 |
uit | +0.010 |
Did | +0.010 |
cher | +0.010 |
Neg logit contributions | |
breath | -0.010 |
common | -0.008 |
sleep | -0.008 |
long | -0.008 |
job | -0.008 |
Token | that |
Token ID | 326 |
Feature activation | +3.61e+00 |
Pos logit contributions | |
boards | +0.061 |
auc | +0.059 |
Did | +0.058 |
ains | +0.057 |
uit | +0.056 |
Neg logit contributions | |
breath | -0.041 |
long | -0.034 |
home | -0.032 |
sleep | -0.031 |
job | -0.030 |
Token | meant |
Token ID | 4001 |
Feature activation | +9.21e-01 |
Pos logit contributions | |
boards | +0.099 |
uit | +0.098 |
auc | +0.098 |
cher | +0.097 |
ode | +0.095 |
Neg logit contributions | |
breath | -0.065 |
long | -0.055 |
sleep | -0.054 |
time | -0.053 |
common | -0.053 |
Token | in |
Token ID | 287 |
Feature activation | -1.94e+00 |
Pos logit contributions | |
boards | +0.038 |
Did | +0.037 |
auc | +0.037 |
uit | +0.037 |
cher | +0.036 |
Neg logit contributions | |
breath | -0.032 |
common | -0.025 |
sleep | -0.024 |
job | -0.024 |
long | -0.024 |
Token | the |
Token ID | 262 |
Feature activation | +1.35e+00 |
Pos logit contributions | |
breath | +0.042 |
common | +0.034 |
job | +0.032 |
sleep | +0.032 |
long | +0.031 |
Neg logit contributions | |
boards | -0.047 |
Did | -0.046 |
uit | -0.046 |
auc | -0.045 |
washer | -0.044 |
Token | park |
Token ID | 3952 |
Feature activation | -9.12e-01 |
Pos logit contributions | |
cher | +0.037 |
boards | +0.035 |
uit | +0.034 |
register | +0.034 |
auc | +0.034 |
Neg logit contributions | |
breath | -0.026 |
sleep | -0.023 |
long | -0.022 |
job | -0.022 |
home | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | +2.19e-01 |
Pos logit contributions | |
breath | +0.018 |
common | +0.014 |
job | +0.013 |
sleep | +0.012 |
care | +0.012 |
Neg logit contributions | |
boards | -0.025 |
auc | -0.024 |
Did | -0.024 |
uit | -0.024 |
washer | -0.024 |
Token | He |
Token ID | 679 |
Feature activation | +7.30e-01 |
Pos logit contributions | |
boards | +0.005 |
Did | +0.005 |
auc | +0.005 |
cher | +0.005 |
uit | +0.005 |
Neg logit contributions | |
breath | -0.005 |
common | -0.004 |
sleep | -0.004 |
job | -0.004 |
long | -0.004 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.56e+00 |
Pos logit contributions | |
boards | +0.015 |
Did | +0.015 |
auc | +0.014 |
uit | +0.014 |
washer | +0.014 |
Neg logit contributions | |
breath | -0.019 |
common | -0.016 |
job | -0.015 |
sleep | -0.015 |
long | -0.015 |
Token | it |
Token ID | 340 |
Feature activation | +3.23e+00 |
Pos logit contributions | |
auc | +0.255 |
cher | +0.252 |
Did | +0.251 |
washer | +0.249 |
boards | +0.249 |
Neg logit contributions | |
breath | -0.105 |
long | -0.086 |
home | -0.079 |
care | -0.076 |
work | -0.071 |
Token | up |
Token ID | 510 |
Feature activation | -2.23e+00 |
Pos logit contributions | |
Did | +0.085 |
boards | +0.083 |
auc | +0.082 |
cher | +0.082 |
register | +0.080 |
Neg logit contributions | |
breath | -0.071 |
long | -0.054 |
home | -0.053 |
job | -0.052 |
sleep | -0.052 |
Token | and |
Token ID | 290 |
Feature activation | -1.55e-01 |
Pos logit contributions | |
breath | +0.049 |
common | +0.040 |
sleep | +0.037 |
job | +0.037 |
care | +0.036 |
Neg logit contributions | |
boards | -0.054 |
Did | -0.053 |
uit | -0.052 |
auc | -0.051 |
washer | -0.051 |
Token | ran |
Token ID | 4966 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
breath | +0.003 |
work | +0.003 |
sleep | +0.003 |
long | +0.003 |
care | +0.003 |
Neg logit contributions | |
auc | -0.004 |
uit | -0.004 |
boards | -0.004 |
register | -0.004 |
ains | -0.004 |
Token | to |
Token ID | 284 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
boards | +0.023 |
ting | +0.022 |
uit | +0.022 |
lying | +0.021 |
pped | +0.020 |
Neg logit contributions | |
breath | -0.033 |
uth | -0.031 |
common | -0.030 |
appetite | -0.030 |
idea | -0.029 |
Token | on |
Token ID | 319 |
Feature activation | -2.63e+00 |
Pos logit contributions | |
ains | +0.111 |
auc | +0.107 |
boards | +0.105 |
Did | +0.105 |
ode | +0.101 |
Neg logit contributions | |
breath | -0.060 |
long | -0.054 |
home | -0.048 |
job | -0.048 |
sleep | -0.047 |
Token | the |
Token ID | 262 |
Feature activation | +2.70e+00 |
Pos logit contributions | |
breath | +0.059 |
common | +0.047 |
job | +0.044 |
sleep | +0.044 |
care | +0.044 |
Neg logit contributions | |
boards | -0.062 |
uit | -0.060 |
Did | -0.060 |
auc | -0.059 |
register | -0.058 |
Token | floor |
Token ID | 4314 |
Feature activation | +1.47e+00 |
Pos logit contributions | |
auc | +0.067 |
cher | +0.067 |
Did | +0.067 |
boards | +0.067 |
uit | +0.066 |
Neg logit contributions | |
breath | -0.056 |
long | -0.048 |
job | -0.048 |
sleep | -0.046 |
common | -0.045 |
Token | . |
Token ID | 13 |
Feature activation | +4.82e-01 |
Pos logit contributions | |
cher | +0.053 |
Which | +0.051 |
Did | +0.050 |
uit | +0.050 |
ains | +0.049 |
Neg logit contributions | |
breath | -0.014 |
long | -0.010 |
time | -0.010 |
sleep | -0.008 |
week | -0.008 |
Token | He |
Token ID | 679 |
Feature activation | +7.32e-01 |
Pos logit contributions | |
boards | +0.012 |
Did | +0.011 |
auc | +0.011 |
uit | +0.011 |
cher | +0.011 |
Neg logit contributions | |
breath | -0.011 |
common | -0.009 |
job | -0.008 |
sleep | -0.008 |
long | -0.008 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.56e+00 |
Pos logit contributions | |
boards | +0.014 |
Did | +0.014 |
uit | +0.014 |
washer | +0.014 |
auc | +0.013 |
Neg logit contributions | |
breath | -0.020 |
common | -0.016 |
job | -0.016 |
sleep | -0.015 |
care | -0.015 |
Token | it |
Token ID | 340 |
Feature activation | +3.11e+00 |
Pos logit contributions | |
Did | +0.230 |
auc | +0.230 |
boards | +0.228 |
uit | +0.224 |
cher | +0.222 |
Neg logit contributions | |
breath | -0.127 |
long | -0.095 |
care | -0.091 |
home | -0.090 |
common | -0.087 |
Token | up |
Token ID | 510 |
Feature activation | -2.21e+00 |
Pos logit contributions | |
Did | +0.080 |
boards | +0.077 |
auc | +0.076 |
cher | +0.076 |
uit | +0.074 |
Neg logit contributions | |
breath | -0.070 |
long | -0.054 |
job | -0.053 |
sleep | -0.052 |
home | -0.052 |
Token | and |
Token ID | 290 |
Feature activation | -2.52e-01 |
Pos logit contributions | |
breath | +0.049 |
common | +0.040 |
sleep | +0.038 |
job | +0.037 |
care | +0.037 |
Neg logit contributions | |
boards | -0.052 |
uit | -0.051 |
Did | -0.050 |
washer | -0.049 |
auc | -0.049 |
Token | looked |
Token ID | 3114 |
Feature activation | +2.24e+00 |
Pos logit contributions | |
breath | +0.005 |
work | +0.004 |
happily | +0.004 |
care | +0.004 |
sleep | +0.004 |
Neg logit contributions | |
auc | -0.007 |
uit | -0.007 |
boards | -0.007 |
register | -0.006 |
ains | -0.006 |
Token | at |
Token ID | 379 |
Feature activation | +2.23e+00 |
Pos logit contributions | |
boards | +0.052 |
Did | +0.050 |
auc | +0.049 |
washer | +0.049 |
cher | +0.049 |
Neg logit contributions | |
breath | -0.051 |
common | -0.043 |
job | -0.040 |
sleep | -0.040 |
purpose | -0.038 |
Token | a |
Token ID | 257 |
Feature activation | +9.47e-01 |
Pos logit contributions | |
boards | +0.196 |
ains | +0.191 |
cher | +0.190 |
ode | +0.188 |
register | +0.187 |
Neg logit contributions | |
breath | -0.111 |
long | -0.089 |
home | -0.088 |
sleep | -0.082 |
care | -0.081 |
Token | big |
Token ID | 1263 |
Feature activation | -2.23e+00 |
Pos logit contributions | |
boards | +0.028 |
auc | +0.027 |
Did | +0.027 |
ting | +0.027 |
uit | +0.027 |
Neg logit contributions | |
breath | -0.017 |
long | -0.014 |
job | -0.014 |
common | -0.013 |
sleep | -0.012 |
Token | rock |
Token ID | 3881 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
breath | +0.051 |
common | +0.041 |
job | +0.039 |
sleep | +0.038 |
long | +0.038 |
Neg logit contributions | |
boards | -0.053 |
Did | -0.051 |
auc | -0.051 |
uit | -0.051 |
cher | -0.050 |
Token | . |
Token ID | 13 |
Feature activation | +6.46e-01 |
Pos logit contributions | |
ains | +0.050 |
boards | +0.050 |
uit | +0.049 |
Did | +0.049 |
cher | +0.048 |
Neg logit contributions | |
breath | -0.037 |
long | -0.028 |
sleep | -0.028 |
job | -0.027 |
day | -0.027 |
Token | He |
Token ID | 679 |
Feature activation | +8.02e-01 |
Pos logit contributions | |
boards | +0.015 |
cher | +0.015 |
auc | +0.015 |
Did | +0.015 |
register | +0.015 |
Neg logit contributions | |
breath | -0.014 |
common | -0.011 |
sleep | -0.011 |
long | -0.011 |
job | -0.011 |
Token | picked |
Token ID | 6497 |
Feature activation | +7.56e+00 |
Pos logit contributions | |
boards | +0.015 |
Did | +0.015 |
washer | +0.014 |
uit | +0.014 |
auc | +0.014 |
Neg logit contributions | |
breath | -0.022 |
common | -0.019 |
job | -0.018 |
sleep | -0.017 |
care | -0.017 |
Token | it |
Token ID | 340 |
Feature activation | +3.07e+00 |
Pos logit contributions | |
Did | +0.228 |
auc | +0.227 |
boards | +0.226 |
cher | +0.224 |
uit | +0.223 |
Neg logit contributions | |
breath | -0.133 |
long | -0.099 |
care | -0.092 |
home | -0.091 |
job | -0.091 |
Token | up |
Token ID | 510 |
Feature activation | -2.21e+00 |
Pos logit contributions | |
Did | +0.074 |
boards | +0.073 |
auc | +0.072 |
cher | +0.071 |
uit | +0.070 |
Neg logit contributions | |
breath | -0.072 |
long | -0.056 |
job | -0.055 |
sleep | -0.055 |
common | -0.055 |
Token | and |
Token ID | 290 |
Feature activation | -2.22e-01 |
Pos logit contributions | |
breath | +0.053 |
common | +0.045 |
sleep | +0.043 |
job | +0.043 |
care | +0.043 |
Neg logit contributions | |
boards | -0.045 |
uit | -0.045 |
Did | -0.044 |
lying | -0.043 |
sticking | -0.042 |
Token | felt |
Token ID | 2936 |
Feature activation | +8.01e-01 |
Pos logit contributions | |
breath | +0.005 |
sleep | +0.004 |
common | +0.004 |
long | +0.004 |
care | +0.004 |
Neg logit contributions | |
boards | -0.005 |
auc | -0.005 |
uit | -0.005 |
register | -0.005 |
Did | -0.005 |
Token | how |
Token ID | 703 |
Feature activation | +7.72e-01 |
Pos logit contributions | |
boards | +0.023 |
Did | +0.022 |
uit | +0.022 |
auc | +0.022 |
cher | +0.022 |
Neg logit contributions | |
breath | -0.015 |
common | -0.011 |
sleep | -0.010 |
job | -0.010 |
long | -0.010 |
Token | soon |
Token ID | 2582 |
Feature activation | -5.23e+00 |
Pos logit contributions | |
boards | +0.044 |
Did | +0.044 |
uit | +0.044 |
auc | +0.043 |
cher | +0.042 |
Neg logit contributions | |
breath | -0.051 |
common | -0.042 |
job | -0.039 |
sleep | -0.039 |
care | -0.039 |
Token | nothing |
Token ID | 2147 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
breath | +0.133 |
common | +0.107 |
long | +0.104 |
sleep | +0.103 |
job | +0.102 |
Neg logit contributions | |
boards | -0.114 |
auc | -0.110 |
Did | -0.109 |
uit | -0.108 |
cher | -0.107 |
Token | but |
Token ID | 475 |
Feature activation | -7.51e-01 |
Pos logit contributions | |
breath | +0.031 |
common | +0.026 |
long | +0.025 |
sleep | +0.025 |
care | +0.024 |
Neg logit contributions | |
boards | -0.027 |
uit | -0.027 |
Did | -0.026 |
auc | -0.026 |
register | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
breath | +0.016 |
common | +0.012 |
long | +0.012 |
sleep | +0.012 |
care | +0.012 |
Neg logit contributions | |
boards | -0.021 |
auc | -0.020 |
uit | -0.020 |
Did | -0.020 |
cher | -0.020 |
Token | small |
Token ID | 1402 |
Feature activation | -3.60e+00 |
Pos logit contributions | |
boards | +0.004 |
Did | +0.004 |
uit | +0.004 |
auc | +0.004 |
washer | +0.004 |
Neg logit contributions | |
breath | -0.003 |
common | -0.002 |
sleep | -0.002 |
job | -0.002 |
care | -0.002 |
Token | spe |
Token ID | 693 |
Feature activation | -1.10e+01 |
Pos logit contributions | |
breath | +0.075 |
common | +0.063 |
job | +0.056 |
care | +0.055 |
sleep | +0.055 |
Neg logit contributions | |
Did | -0.082 |
boards | -0.081 |
auc | -0.081 |
uit | -0.080 |
washer | -0.080 |
Token | ck |
Token ID | 694 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
breath | +0.250 |
excitement | +0.236 |
sleep | +0.229 |
Today | +0.226 |
long | +0.226 |
Neg logit contributions | |
pped | -0.204 |
Did | -0.188 |
uit | -0.184 |
sticking | -0.182 |
pled | -0.181 |
Token | in |
Token ID | 287 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
boards | +0.026 |
Did | +0.025 |
washer | +0.025 |
auc | +0.025 |
uit | +0.024 |
Neg logit contributions | |
breath | -0.028 |
time | -0.022 |
long | -0.021 |
sleep | -0.021 |
common | -0.021 |
Token | the |
Token ID | 262 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
breath | +0.040 |
job | +0.031 |
anna | +0.031 |
sleep | +0.030 |
common | +0.030 |
Neg logit contributions | |
Did | -0.047 |
uit | -0.046 |
boards | -0.045 |
auc | -0.044 |
ting | -0.044 |
Token | sky |
Token ID | 6766 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
boards | +0.038 |
Did | +0.037 |
uit | +0.037 |
auc | +0.037 |
cher | +0.036 |
Neg logit contributions | |
breath | -0.037 |
common | -0.030 |
job | -0.028 |
sleep | -0.028 |
long | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | +4.16e-01 |
Pos logit contributions | |
breath | +0.021 |
common | +0.015 |
sleep | +0.015 |
long | +0.015 |
care | +0.014 |
Neg logit contributions | |
boards | -0.039 |
auc | -0.038 |
Did | -0.038 |
cher | -0.038 |
uit | -0.038 |
Token | the |
Token ID | 262 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
breath | +0.049 |
common | +0.038 |
sleep | +0.036 |
long | +0.036 |
job | +0.036 |
Neg logit contributions | |
boards | -0.059 |
Did | -0.057 |
auc | -0.057 |
uit | -0.057 |
cher | -0.057 |
Token | rain |
Token ID | 6290 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
boards | +0.037 |
auc | +0.036 |
Did | +0.035 |
uit | +0.035 |
washer | +0.034 |
Neg logit contributions | |
breath | -0.030 |
common | -0.024 |
job | -0.022 |
sleep | -0.021 |
long | -0.021 |
Token | and |
Token ID | 290 |
Feature activation | -8.64e-01 |
Pos logit contributions | |
breath | +0.022 |
common | +0.017 |
job | +0.016 |
long | +0.016 |
sleep | +0.016 |
Neg logit contributions | |
boards | -0.031 |
auc | -0.030 |
Did | -0.030 |
uit | -0.030 |
washer | -0.030 |
Token | take |
Token ID | 1011 |
Feature activation | +2.38e+00 |
Pos logit contributions | |
breath | +0.018 |
sleep | +0.014 |
common | +0.014 |
long | +0.014 |
care | +0.014 |
Neg logit contributions | |
boards | -0.022 |
uit | -0.022 |
auc | -0.022 |
Did | -0.022 |
register | -0.021 |
Token | a |
Token ID | 257 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
boards | +0.051 |
uit | +0.049 |
Did | +0.049 |
auc | +0.048 |
cher | +0.048 |
Neg logit contributions | |
breath | -0.059 |
common | -0.049 |
job | -0.046 |
sleep | -0.046 |
long | -0.045 |
Token | com |
Token ID | 401 |
Feature activation | -1.08e+01 |
Pos logit contributions | |
breath | +0.017 |
common | +0.016 |
uth | +0.015 |
breaths | +0.014 |
purpose | +0.014 |
Neg logit contributions | |
washer | -0.026 |
uit | -0.025 |
boards | -0.025 |
cher | -0.025 |
lying | -0.024 |
Token | fy |
Token ID | 24928 |
Feature activation | -7.65e+00 |
Pos logit contributions | |
breath | +0.196 |
common | +0.153 |
excitement | +0.145 |
purpose | +0.142 |
Sunday | +0.138 |
Neg logit contributions | |
boards | -0.281 |
auc | -0.278 |
washer | -0.277 |
cher | -0.274 |
uit | -0.264 |
Token | nap |
Token ID | 25422 |
Feature activation | -3.23e+00 |
Pos logit contributions | |
common | +0.139 |
breath | +0.136 |
uth | +0.119 |
job | +0.113 |
idea | +0.113 |
Neg logit contributions | |
uit | -0.184 |
boards | -0.169 |
box | -0.167 |
lying | -0.166 |
sticking | -0.165 |
Token | . |
Token ID | 13 |
Feature activation | +4.44e-01 |
Pos logit contributions | |
breath | +0.059 |
long | +0.042 |
sleep | +0.042 |
common | +0.041 |
job | +0.040 |
Neg logit contributions | |
Did | -0.092 |
boards | -0.092 |
auc | -0.090 |
register | -0.088 |
uit | -0.088 |
Token | The |
Token ID | 383 |
Feature activation | +2.32e+00 |
Pos logit contributions | |
boards | +0.012 |
auc | +0.011 |
Did | +0.011 |
cher | +0.011 |
uit | +0.011 |
Neg logit contributions | |
breath | -0.009 |
sleep | -0.007 |
common | -0.007 |
long | -0.007 |
job | -0.006 |
Token | tiger |
Token ID | 26241 |
Feature activation | -2.99e-01 |
Pos logit contributions | |
boards | +0.055 |
Did | +0.053 |
uit | +0.052 |
auc | +0.052 |
washer | +0.051 |
Neg logit contributions | |
breath | -0.053 |
common | -0.044 |
sleep | -0.040 |
job | -0.040 |
long | -0.040 |
Token | c |
Token ID | 269 |
Feature activation | -1.22e+00 |
Pos logit contributions | |
breath | +0.020 |
common | +0.017 |
job | +0.016 |
long | +0.016 |
sleep | +0.016 |
Neg logit contributions | |
boards | -0.016 |
Did | -0.016 |
washer | -0.016 |
auc | -0.016 |
cher | -0.016 |
Token | uddled |
Token ID | 32745 |
Feature activation | +5.08e-01 |
Pos logit contributions | |
breath | +0.041 |
common | +0.036 |
sleep | +0.034 |
long | +0.034 |
job | +0.034 |
Neg logit contributions | |
boards | -0.016 |
uit | -0.015 |
auc | -0.015 |
Did | -0.015 |
washer | -0.014 |
Token | up |
Token ID | 510 |
Feature activation | -4.34e+00 |
Pos logit contributions | |
boards | +0.011 |
auc | +0.010 |
Did | +0.010 |
uit | +0.010 |
cher | +0.010 |
Neg logit contributions | |
breath | -0.013 |
common | -0.011 |
sleep | -0.010 |
long | -0.010 |
job | -0.010 |
Token | in |
Token ID | 287 |
Feature activation | -2.43e+00 |
Pos logit contributions | |
breath | +0.102 |
common | +0.085 |
job | +0.080 |
care | +0.079 |
sleep | +0.078 |
Neg logit contributions | |
boards | -0.097 |
uit | -0.094 |
Did | -0.094 |
auc | -0.092 |
washer | -0.091 |
Token | a |
Token ID | 257 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
breath | +0.051 |
common | +0.040 |
sleep | +0.038 |
job | +0.038 |
long | +0.038 |
Neg logit contributions | |
boards | -0.062 |
Did | -0.060 |
auc | -0.060 |
uit | -0.060 |
cher | -0.059 |
Token | com |
Token ID | 401 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
breath | +0.026 |
common | +0.022 |
job | +0.020 |
care | +0.020 |
uth | +0.019 |
Neg logit contributions | |
boards | -0.028 |
washer | -0.028 |
auc | -0.027 |
Did | -0.027 |
uit | -0.026 |
Token | fy |
Token ID | 24928 |
Feature activation | -7.82e+00 |
Pos logit contributions | |
breath | +0.220 |
excitement | +0.182 |
common | +0.168 |
purpose | +0.165 |
sight | +0.159 |
Neg logit contributions | |
washer | -0.267 |
boards | -0.254 |
auc | -0.249 |
cher | -0.238 |
ray | -0.232 |
Token | blanket |
Token ID | 18447 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
common | +0.162 |
breath | +0.152 |
uth | +0.139 |
argue | +0.139 |
accomplishment | +0.137 |
Neg logit contributions | |
box | -0.162 |
uit | -0.162 |
lying | -0.151 |
boards | -0.150 |
living | -0.148 |
Token | and |
Token ID | 290 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
breath | +0.024 |
sleep | +0.017 |
long | +0.017 |
common | +0.016 |
job | +0.016 |
Neg logit contributions | |
Did | -0.044 |
boards | -0.044 |
auc | -0.043 |
cher | -0.042 |
uit | -0.042 |
Token | soon |
Token ID | 2582 |
Feature activation | -5.48e+00 |
Pos logit contributions | |
breath | +0.031 |
common | +0.024 |
sleep | +0.024 |
long | +0.023 |
job | +0.023 |
Neg logit contributions | |
boards | -0.031 |
auc | -0.030 |
Did | -0.030 |
uit | -0.030 |
cher | -0.030 |
Token | he |
Token ID | 339 |
Feature activation | -4.22e-01 |
Pos logit contributions | |
breath | +0.135 |
common | +0.109 |
long | +0.105 |
sleep | +0.105 |
job | +0.104 |
Neg logit contributions | |
boards | -0.120 |
Did | -0.117 |
auc | -0.117 |
uit | -0.116 |
cher | -0.113 |
Token | She |
Token ID | 1375 |
Feature activation | +7.84e-01 |
Pos logit contributions | |
boards | +0.002 |
register | +0.002 |
Did | +0.002 |
Is | +0.002 |
can | +0.002 |
Neg logit contributions | |
breath | -0.002 |
common | -0.002 |
job | -0.002 |
care | -0.002 |
sleep | -0.002 |
Token | lay |
Token ID | 3830 |
Feature activation | -3.58e-01 |
Pos logit contributions | |
boards | +0.019 |
Did | +0.018 |
auc | +0.018 |
uit | +0.018 |
cher | +0.018 |
Neg logit contributions | |
breath | -0.018 |
common | -0.014 |
sleep | -0.013 |
job | -0.013 |
long | -0.013 |
Token | down |
Token ID | 866 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
breath | +0.008 |
home | +0.006 |
long | +0.006 |
common | +0.006 |
job | +0.006 |
Neg logit contributions | |
boards | -0.009 |
auc | -0.009 |
Did | -0.009 |
washer | -0.009 |
cher | -0.009 |
Token | in |
Token ID | 287 |
Feature activation | -2.48e+00 |
Pos logit contributions | |
breath | +0.042 |
common | +0.033 |
long | +0.033 |
sleep | +0.032 |
job | +0.032 |
Neg logit contributions | |
boards | -0.040 |
Did | -0.039 |
auc | -0.039 |
uit | -0.038 |
cher | -0.038 |
Token | a |
Token ID | 257 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
breath | +0.053 |
common | +0.043 |
job | +0.041 |
sleep | +0.039 |
long | +0.039 |
Neg logit contributions | |
boards | -0.060 |
uit | -0.059 |
Did | -0.059 |
auc | -0.058 |
register | -0.057 |
Token | com |
Token ID | 401 |
Feature activation | -1.04e+01 |
Pos logit contributions | |
breath | +0.027 |
common | +0.022 |
job | +0.020 |
care | +0.020 |
sleep | +0.020 |
Neg logit contributions | |
boards | -0.030 |
auc | -0.029 |
washer | -0.029 |
Did | -0.028 |
uit | -0.028 |
Token | fy |
Token ID | 24928 |
Feature activation | -7.73e+00 |
Pos logit contributions | |
breath | +0.221 |
common | +0.180 |
purpose | +0.171 |
excitement | +0.170 |
sight | +0.163 |
Neg logit contributions | |
washer | -0.252 |
boards | -0.250 |
auc | -0.243 |
cher | -0.220 |
ray | -0.218 |
Token | position |
Token ID | 2292 |
Feature activation | -4.08e-01 |
Pos logit contributions | |
breath | +0.157 |
common | +0.152 |
care | +0.134 |
uth | +0.134 |
idea | +0.130 |
Neg logit contributions | |
uit | -0.167 |
box | -0.166 |
boards | -0.164 |
washer | -0.155 |
lying | -0.153 |
Token | and |
Token ID | 290 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
breath | +0.007 |
common | +0.005 |
long | +0.005 |
sleep | +0.005 |
job | +0.005 |
Neg logit contributions | |
boards | -0.012 |
Did | -0.011 |
auc | -0.011 |
cher | -0.011 |
uit | -0.011 |
Token | soon |
Token ID | 2582 |
Feature activation | -5.53e+00 |
Pos logit contributions | |
breath | +0.033 |
common | +0.026 |
sleep | +0.025 |
job | +0.025 |
long | +0.025 |
Neg logit contributions | |
boards | -0.032 |
Did | -0.031 |
auc | -0.031 |
uit | -0.031 |
cher | -0.031 |
Token | fell |
Token ID | 3214 |
Feature activation | -8.23e-01 |
Pos logit contributions | |
breath | +0.131 |
common | +0.108 |
job | +0.101 |
sleep | +0.099 |
long | +0.096 |
Neg logit contributions | |
boards | -0.127 |
uit | -0.120 |
Did | -0.120 |
register | -0.119 |
auc | -0.118 |
Token | and |
Token ID | 290 |
Feature activation | -2.53e+00 |
Pos logit contributions | |
breath | +0.126 |
common | +0.094 |
sleep | +0.088 |
long | +0.088 |
job | +0.088 |
Neg logit contributions | |
boards | -0.191 |
Did | -0.186 |
auc | -0.185 |
uit | -0.184 |
cher | -0.182 |
Token | curled |
Token ID | 41108 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
breath | +0.059 |
common | +0.049 |
job | +0.045 |
long | +0.045 |
sleep | +0.045 |
Neg logit contributions | |
boards | -0.057 |
Did | -0.056 |
uit | -0.055 |
cher | -0.055 |
auc | -0.055 |
Token | up |
Token ID | 510 |
Feature activation | -3.36e+00 |
Pos logit contributions | |
breath | +0.032 |
common | +0.027 |
sleep | +0.026 |
job | +0.026 |
care | +0.026 |
Neg logit contributions | |
boards | -0.024 |
uit | -0.023 |
Did | -0.023 |
cher | -0.023 |
auc | -0.023 |
Token | in |
Token ID | 287 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
breath | +0.084 |
common | +0.071 |
job | +0.067 |
care | +0.066 |
sleep | +0.066 |
Neg logit contributions | |
boards | -0.068 |
Did | -0.066 |
uit | -0.066 |
auc | -0.065 |
cher | -0.064 |
Token | a |
Token ID | 257 |
Feature activation | -5.73e-01 |
Pos logit contributions | |
breath | +0.028 |
common | +0.023 |
job | +0.021 |
long | +0.021 |
sleep | +0.021 |
Neg logit contributions | |
Did | -0.028 |
boards | -0.028 |
uit | -0.028 |
auc | -0.027 |
cher | -0.027 |
Token | com |
Token ID | 401 |
Feature activation | -1.04e+01 |
Pos logit contributions | |
breath | +0.012 |
common | +0.010 |
care | +0.010 |
breaths | +0.009 |
job | +0.009 |
Neg logit contributions | |
washer | -0.013 |
Did | -0.013 |
boards | -0.013 |
auc | -0.012 |
lying | -0.012 |
Token | fy |
Token ID | 24928 |
Feature activation | -7.44e+00 |
Pos logit contributions | |
breath | +0.208 |
excitement | +0.161 |
common | +0.156 |
purpose | +0.150 |
Sunday | +0.147 |
Neg logit contributions | |
washer | -0.258 |
boards | -0.255 |
auc | -0.252 |
cher | -0.248 |
uit | -0.235 |
Token | spot |
Token ID | 4136 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
breath | +0.149 |
common | +0.147 |
idea | +0.128 |
care | +0.128 |
powers | +0.125 |
Neg logit contributions | |
uit | -0.156 |
box | -0.150 |
lying | -0.148 |
living | -0.142 |
sticking | -0.137 |
Token | to |
Token ID | 284 |
Feature activation | -1.85e+00 |
Pos logit contributions | |
breath | +0.003 |
common | +0.002 |
long | +0.002 |
sleep | +0.002 |
job | +0.002 |
Neg logit contributions | |
boards | -0.005 |
Did | -0.005 |
auc | -0.005 |
uit | -0.005 |
cher | -0.005 |
Token | take |
Token ID | 1011 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
breath | +0.036 |
common | +0.027 |
sleep | +0.026 |
long | +0.025 |
care | +0.025 |
Neg logit contributions | |
boards | -0.052 |
auc | -0.050 |
Did | -0.050 |
uit | -0.049 |
washer | -0.049 |
Token | a |
Token ID | 257 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
boards | +0.045 |
uit | +0.043 |
Did | +0.042 |
ting | +0.042 |
cher | +0.042 |
Neg logit contributions | |
breath | -0.049 |
common | -0.040 |
job | -0.037 |
sleep | -0.036 |
long | -0.036 |
Token | for |
Token ID | 329 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
Did | +0.116 |
auc | +0.116 |
register | +0.114 |
washer | +0.113 |
boards | +0.113 |
Neg logit contributions | |
breath | -0.080 |
long | -0.061 |
common | -0.059 |
home | -0.056 |
happily | -0.055 |
Token | food |
Token ID | 2057 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
boards | +0.052 |
auc | +0.052 |
washer | +0.051 |
Did | +0.050 |
uit | +0.050 |
Neg logit contributions | |
breath | -0.038 |
common | -0.029 |
long | -0.029 |
care | -0.028 |
sleep | -0.027 |
Token | and |
Token ID | 290 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
Did | +0.038 |
auc | +0.038 |
boards | +0.037 |
register | +0.036 |
ains | +0.036 |
Neg logit contributions | |
breath | -0.018 |
long | -0.013 |
common | -0.013 |
sleep | -0.013 |
home | -0.012 |
Token | made |
Token ID | 925 |
Feature activation | +2.83e+00 |
Pos logit contributions | |
breath | +0.023 |
long | +0.019 |
sleep | +0.018 |
care | +0.018 |
common | +0.018 |
Neg logit contributions | |
auc | -0.027 |
register | -0.026 |
boards | -0.026 |
uit | -0.026 |
washer | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | -1.57e-01 |
Pos logit contributions | |
boards | +0.063 |
Did | +0.062 |
auc | +0.061 |
uit | +0.060 |
cher | +0.060 |
Neg logit contributions | |
breath | -0.070 |
common | -0.056 |
long | -0.054 |
sleep | -0.054 |
job | -0.054 |
Token | com |
Token ID | 401 |
Feature activation | -1.01e+01 |
Pos logit contributions | |
breath | +0.003 |
common | +0.003 |
care | +0.002 |
sleep | +0.002 |
job | +0.002 |
Neg logit contributions | |
boards | -0.004 |
Did | -0.004 |
washer | -0.004 |
uit | -0.004 |
register | -0.004 |
Token | fy |
Token ID | 24928 |
Feature activation | -6.67e+00 |
Pos logit contributions | |
breath | +0.193 |
common | +0.144 |
Sunday | +0.143 |
Wednesday | +0.141 |
purpose | +0.136 |
Neg logit contributions | |
washer | -0.262 |
boards | -0.260 |
cher | -0.257 |
auc | -0.255 |
ode | -0.246 |
Token | home |
Token ID | 1363 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
breath | +0.128 |
common | +0.113 |
anna | +0.098 |
care | +0.097 |
� | +0.094 |
Neg logit contributions | |
uit | -0.155 |
boards | -0.152 |
lying | -0.147 |
box | -0.147 |
ode | -0.143 |
Token | in |
Token ID | 287 |
Feature activation | -2.45e+00 |
Pos logit contributions | |
breath | +0.027 |
long | +0.019 |
sleep | +0.018 |
common | +0.018 |
job | +0.018 |
Neg logit contributions | |
Did | -0.052 |
boards | -0.052 |
auc | -0.051 |
uit | -0.050 |
cher | -0.049 |
Token | the |
Token ID | 262 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
breath | +0.051 |
common | +0.040 |
sleep | +0.038 |
job | +0.038 |
long | +0.038 |
Neg logit contributions | |
boards | -0.063 |
Did | -0.061 |
auc | -0.061 |
uit | -0.061 |
cher | -0.060 |
Token | woods |
Token ID | 16479 |
Feature activation | -7.38e-01 |
Pos logit contributions | |
boards | +0.028 |
Did | +0.028 |
auc | +0.028 |
uit | +0.027 |
cher | +0.027 |
Neg logit contributions | |
breath | -0.025 |
common | -0.020 |
long | -0.019 |
sleep | -0.019 |
job | -0.019 |
Token | to |
Token ID | 284 |
Feature activation | -9.46e-01 |
Pos logit contributions | |
breath | +0.000 |
common | +0.000 |
long | +0.000 |
job | +0.000 |
sleep | +0.000 |
Neg logit contributions | |
boards | -0.000 |
Did | -0.000 |
auc | -0.000 |
uit | -0.000 |
cher | -0.000 |
Token | make |
Token ID | 787 |
Feature activation | +3.20e+00 |
Pos logit contributions | |
breath | +0.019 |
common | +0.014 |
sleep | +0.014 |
job | +0.014 |
long | +0.014 |
Neg logit contributions | |
boards | -0.027 |
Did | -0.026 |
auc | -0.026 |
uit | -0.025 |
cher | -0.025 |
Token | it |
Token ID | 340 |
Feature activation | -3.75e-01 |
Pos logit contributions | |
boards | +0.069 |
auc | +0.066 |
Did | +0.066 |
uit | +0.066 |
cher | +0.065 |
Neg logit contributions | |
breath | -0.079 |
common | -0.066 |
job | -0.061 |
sleep | -0.060 |
long | -0.060 |
Token | even |
Token ID | 772 |
Feature activation | -3.10e+00 |
Pos logit contributions | |
breath | +0.009 |
common | +0.007 |
job | +0.007 |
sleep | +0.007 |
long | +0.007 |
Neg logit contributions | |
boards | -0.008 |
auc | -0.008 |
Did | -0.008 |
uit | -0.008 |
washer | -0.008 |
Token | more |
Token ID | 517 |
Feature activation | -4.05e+00 |
Pos logit contributions | |
breath | +0.085 |
common | +0.071 |
sleep | +0.068 |
job | +0.068 |
long | +0.068 |
Neg logit contributions | |
boards | -0.060 |
Did | -0.057 |
uit | -0.057 |
auc | -0.057 |
cher | -0.055 |
Token | com |
Token ID | 401 |
Feature activation | -1.01e+01 |
Pos logit contributions | |
breath | +0.088 |
common | +0.065 |
uth | +0.061 |
care | +0.061 |
discussion | +0.060 |
Neg logit contributions | |
boards | -0.094 |
washer | -0.094 |
lying | -0.094 |
auc | -0.089 |
pled | -0.088 |
Token | fy |
Token ID | 24928 |
Feature activation | -6.03e+00 |
Pos logit contributions | |
breath | +0.245 |
common | +0.191 |
excitement | +0.186 |
Sunday | +0.180 |
sight | +0.178 |
Neg logit contributions | |
boards | -0.224 |
auc | -0.214 |
cher | -0.213 |
washer | -0.210 |
register | -0.204 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.99e+00 |
Pos logit contributions | |
breath | +0.106 |
common | +0.085 |
care | +0.072 |
job | +0.071 |
excitement | +0.071 |
Neg logit contributions | |
boards | -0.170 |
uit | -0.163 |
washer | -0.162 |
ting | -0.162 |
cher | -0.162 |
Token | They |
Token ID | 1119 |
Feature activation | +6.23e-01 |
Pos logit contributions | |
breath | +0.041 |
common | +0.032 |
sleep | +0.030 |
job | +0.030 |
long | +0.030 |
Neg logit contributions | |
boards | -0.051 |
Did | -0.050 |
auc | -0.050 |
uit | -0.050 |
cher | -0.049 |
Token | both |
Token ID | 1111 |
Feature activation | -2.22e+00 |
Pos logit contributions | |
boards | +0.014 |
Did | +0.014 |
auc | +0.014 |
uit | +0.013 |
cher | +0.013 |
Neg logit contributions | |
breath | -0.015 |
common | -0.012 |
sleep | -0.012 |
long | -0.012 |
job | -0.011 |
Token | gigg |
Token ID | 30442 |
Feature activation | -2.80e+00 |
Pos logit contributions | |
breath | +0.059 |
common | +0.052 |
job | +0.049 |
purpose | +0.047 |
long | +0.047 |
Neg logit contributions | |
boards | -0.042 |
Did | -0.040 |
auc | -0.040 |
washer | -0.040 |
uit | -0.039 |
Token |
|
Token ID | 198 |
Feature activation | -1.78e+00 |
Pos logit contributions | |
breath | +0.001 |
common | +0.001 |
job | +0.001 |
accomplishment | +0.001 |
long | +0.001 |
Neg logit contributions | |
boards | -0.001 |
Did | -0.001 |
uit | -0.001 |
auc | -0.001 |
ting | -0.001 |
Token |
|
Token ID | 198 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
breath | +0.045 |
job | +0.036 |
common | +0.035 |
breaths | +0.035 |
accomplishment | +0.033 |
Neg logit contributions | |
boards | -0.038 |
Did | -0.038 |
uit | -0.037 |
washer | -0.035 |
auc | -0.035 |
Token | She |
Token ID | 3347 |
Feature activation | +2.46e-01 |
Pos logit contributions | |
breath | +0.044 |
common | +0.035 |
job | +0.033 |
sleep | +0.032 |
care | +0.032 |
Neg logit contributions | |
boards | -0.041 |
Did | -0.041 |
uit | -0.038 |
auc | -0.037 |
washer | -0.037 |
Token | had |
Token ID | 550 |
Feature activation | +3.13e+00 |
Pos logit contributions | |
boards | +0.005 |
Did | +0.005 |
uit | +0.005 |
auc | +0.005 |
cher | +0.005 |
Neg logit contributions | |
breath | -0.006 |
common | -0.005 |
sleep | -0.005 |
long | -0.005 |
job | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | -4.49e-01 |
Pos logit contributions | |
auc | +0.068 |
boards | +0.067 |
uit | +0.066 |
cher | +0.065 |
Did | +0.065 |
Neg logit contributions | |
breath | -0.078 |
long | -0.064 |
common | -0.063 |
sleep | -0.062 |
care | -0.060 |
Token | com |
Token ID | 401 |
Feature activation | -1.00e+01 |
Pos logit contributions | |
breath | +0.009 |
common | +0.007 |
care | +0.006 |
sleep | +0.006 |
breaths | +0.006 |
Neg logit contributions | |
boards | -0.011 |
washer | -0.011 |
uit | -0.011 |
Did | -0.011 |
box | -0.011 |
Token | fy |
Token ID | 24928 |
Feature activation | -6.17e+00 |
Pos logit contributions | |
breath | +0.191 |
purpose | +0.143 |
Sunday | +0.140 |
Wednesday | +0.137 |
common | +0.134 |
Neg logit contributions | |
boards | -0.257 |
washer | -0.250 |
cher | -0.250 |
auc | -0.248 |
pled | -0.231 |
Token | bed |
Token ID | 3996 |
Feature activation | -8.05e-01 |
Pos logit contributions | |
breath | +0.127 |
common | +0.116 |
care | +0.108 |
uth | +0.104 |
bye | +0.103 |
Neg logit contributions | |
uit | -0.135 |
box | -0.135 |
boards | -0.130 |
lying | -0.125 |
sticking | -0.123 |
Token | and |
Token ID | 290 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
breath | +0.015 |
long | +0.012 |
job | +0.011 |
sleep | +0.011 |
time | +0.011 |
Neg logit contributions | |
Did | -0.023 |
auc | -0.022 |
ode | -0.022 |
ains | -0.022 |
cher | -0.022 |
Token | lots |
Token ID | 6041 |
Feature activation | -2.65e+00 |
Pos logit contributions | |
breath | +0.025 |
long | +0.021 |
sleep | +0.020 |
common | +0.020 |
care | +0.020 |
Neg logit contributions | |
auc | -0.023 |
uit | -0.023 |
boards | -0.023 |
Did | -0.023 |
cher | -0.023 |
Token | of |
Token ID | 286 |
Feature activation | +1.06e-02 |
Pos logit contributions | |
breath | +0.072 |
common | +0.060 |
job | +0.057 |
sleep | +0.056 |
long | +0.056 |
Neg logit contributions | |
boards | -0.052 |
Did | -0.050 |
auc | -0.049 |
uit | -0.049 |
washer | -0.048 |
Token | a |
Token ID | 257 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
breath | +0.063 |
common | +0.052 |
job | +0.050 |
sleep | +0.050 |
long | +0.049 |
Neg logit contributions | |
boards | -0.049 |
Did | -0.048 |
uit | -0.047 |
auc | -0.047 |
cher | -0.046 |
Token | modern |
Token ID | 3660 |
Feature activation | -3.42e+00 |
Pos logit contributions | |
breath | +0.035 |
common | +0.028 |
accomplishment | +0.027 |
breaths | +0.027 |
care | +0.026 |
Neg logit contributions | |
Did | -0.030 |
washer | -0.029 |
lying | -0.027 |
boards | -0.027 |
living | -0.027 |
Token | cave |
Token ID | 11527 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
breath | +0.089 |
common | +0.077 |
care | +0.072 |
accomplishment | +0.071 |
excitement | +0.068 |
Neg logit contributions | |
box | -0.065 |
uit | -0.063 |
washer | -0.063 |
boards | -0.062 |
auc | -0.061 |
Token | with |
Token ID | 351 |
Feature activation | +2.71e+00 |
Pos logit contributions | |
breath | +0.027 |
common | +0.022 |
long | +0.021 |
sleep | +0.021 |
job | +0.021 |
Neg logit contributions | |
boards | -0.032 |
uit | -0.031 |
Did | -0.031 |
auc | -0.030 |
cher | -0.030 |
Token | a |
Token ID | 257 |
Feature activation | -2.91e-01 |
Pos logit contributions | |
boards | +0.071 |
uit | +0.069 |
cher | +0.069 |
auc | +0.069 |
Did | +0.068 |
Neg logit contributions | |
breath | -0.057 |
long | -0.044 |
common | -0.043 |
sleep | -0.042 |
care | -0.042 |
Token | com |
Token ID | 401 |
Feature activation | -9.94e+00 |
Pos logit contributions | |
breath | +0.006 |
common | +0.005 |
care | +0.005 |
anna | +0.005 |
breaths | +0.005 |
Neg logit contributions | |
Did | -0.007 |
washer | -0.006 |
boards | -0.006 |
lying | -0.006 |
cher | -0.006 |
Token | fy |
Token ID | 24928 |
Feature activation | -6.34e+00 |
Pos logit contributions | |
breath | +0.191 |
Sunday | +0.141 |
excitement | +0.139 |
Wednesday | +0.135 |
common | +0.131 |
Neg logit contributions | |
washer | -0.257 |
boards | -0.256 |
auc | -0.251 |
cher | -0.242 |
ode | -0.233 |
Token | bed |
Token ID | 3996 |
Feature activation | -1.87e+00 |
Pos logit contributions | |
breath | +0.131 |
common | +0.127 |
care | +0.117 |
discussion | +0.107 |
excitement | +0.105 |
Neg logit contributions | |
uit | -0.138 |
box | -0.131 |
lying | -0.130 |
sticking | -0.126 |
boards | -0.125 |
Token | . |
Token ID | 13 |
Feature activation | -1.84e-01 |
Pos logit contributions | |
breath | +0.024 |
long | +0.020 |
job | +0.018 |
sleep | +0.018 |
time | +0.017 |
Neg logit contributions | |
Did | -0.061 |
auc | -0.059 |
Which | -0.059 |
ode | -0.058 |
cher | -0.058 |
Token | One |
Token ID | 1881 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
breath | +0.004 |
common | +0.003 |
job | +0.003 |
care | +0.002 |
long | +0.002 |
Neg logit contributions | |
boards | -0.005 |
Did | -0.005 |
uit | -0.005 |
auc | -0.005 |
washer | -0.005 |
Token | day |
Token ID | 1110 |
Feature activation | -4.61e+00 |
Pos logit contributions | |
breath | +0.018 |
common | +0.012 |
breaths | +0.011 |
accomplishment | +0.011 |
excitement | +0.010 |
Neg logit contributions | |
washer | -0.037 |
auc | -0.035 |
lying | -0.035 |
boards | -0.035 |
Did | -0.035 |
Token | the |
Token ID | 262 |
Feature activation | +2.99e+00 |
Pos logit contributions | |
breath | +0.075 |
common | +0.058 |
sleep | +0.055 |
long | +0.055 |
job | +0.054 |
Neg logit contributions | |
boards | -0.100 |
Did | -0.099 |
uit | -0.098 |
auc | -0.098 |
cher | -0.097 |
Token | spider |
Token ID | 19230 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
auc | +0.062 |
Did | +0.061 |
boards | +0.061 |
washer | +0.061 |
lying | +0.060 |
Neg logit contributions | |
breath | -0.073 |
common | -0.062 |
care | -0.056 |
� | -0.053 |
content | -0.053 |
Token | had |
Token ID | 550 |
Feature activation | +3.12e+00 |
Pos logit contributions | |
boards | +0.021 |
auc | +0.021 |
uit | +0.021 |
Did | +0.021 |
register | +0.021 |
Neg logit contributions | |
breath | -0.028 |
sleep | -0.024 |
time | -0.023 |
happily | -0.023 |
care | -0.023 |
Token | loved |
Token ID | 6151 |
Feature activation | -4.15e-01 |
Pos logit contributions | |
boards | +0.070 |
auc | +0.069 |
Did | +0.068 |
washer | +0.067 |
uit | +0.066 |
Neg logit contributions | |
breath | -0.077 |
common | -0.062 |
sleep | -0.061 |
long | -0.060 |
care | -0.060 |
Token | his |
Token ID | 465 |
Feature activation | -6.57e-02 |
Pos logit contributions | |
breath | +0.009 |
long | +0.007 |
common | +0.007 |
sleep | +0.007 |
care | +0.006 |
Neg logit contributions | |
boards | -0.011 |
Did | -0.011 |
cher | -0.011 |
auc | -0.011 |
register | -0.011 |
Token | com |
Token ID | 401 |
Feature activation | -9.91e+00 |
Pos logit contributions | |
breath | +0.001 |
common | +0.001 |
� | +0.001 |
bye | +0.001 |
care | +0.001 |
Neg logit contributions | |
lying | -0.001 |
boards | -0.001 |
washer | -0.001 |
sticking | -0.001 |
auc | -0.001 |
Token | fy |
Token ID | 24928 |
Feature activation | -5.49e+00 |
Pos logit contributions | |
breath | +0.180 |
excitement | +0.127 |
Sunday | +0.125 |
common | +0.123 |
Wednesday | +0.117 |
Neg logit contributions | |
boards | -0.272 |
auc | -0.262 |
pled | -0.258 |
washer | -0.257 |
uit | -0.254 |
Token | cob |
Token ID | 22843 |
Feature activation | -1.85e+00 |
Pos logit contributions | |
breath | +0.121 |
common | +0.111 |
� | +0.095 |
care | +0.094 |
content | +0.093 |
Neg logit contributions | |
boards | -0.116 |
uit | -0.115 |
box | -0.110 |
sticking | -0.109 |
lying | -0.108 |
Token | web |
Token ID | 12384 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
breath | +0.061 |
sleep | +0.052 |
common | +0.051 |
job | +0.051 |
long | +0.051 |
Neg logit contributions | |
boards | -0.025 |
Did | -0.024 |
uit | -0.023 |
register | -0.023 |
auc | -0.022 |
Token | home |
Token ID | 1363 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
breath | +0.004 |
common | +0.003 |
sleep | +0.003 |
job | +0.003 |
long | +0.003 |
Neg logit contributions | |
boards | -0.007 |
Did | -0.007 |
auc | -0.006 |
uit | -0.006 |
cher | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | +4.84e-01 |
Pos logit contributions | |
breath | +0.027 |
sleep | +0.019 |
long | +0.019 |
job | +0.017 |
home | +0.017 |
Neg logit contributions | |
Did | -0.060 |
boards | -0.059 |
auc | -0.059 |
uit | -0.057 |
cher | -0.057 |
Token | around |
Token ID | 1088 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
boards | +0.040 |
ting | +0.039 |
Did | +0.038 |
auc | +0.038 |
washer | +0.038 |
Neg logit contributions | |
breath | -0.057 |
common | -0.051 |
purpose | -0.047 |
job | -0.047 |
sleep | -0.045 |
Token | and |
Token ID | 290 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
breath | +0.031 |
common | +0.026 |
job | +0.024 |
sleep | +0.023 |
care | +0.023 |
Neg logit contributions | |
boards | -0.034 |
Did | -0.033 |
uit | -0.032 |
ting | -0.032 |
auc | -0.032 |
Token | saw |
Token ID | 2497 |
Feature activation | +4.85e+00 |
Pos logit contributions | |
breath | +0.034 |
common | +0.027 |
sleep | +0.026 |
long | +0.026 |
job | +0.026 |
Neg logit contributions | |
boards | -0.032 |
auc | -0.031 |
uit | -0.031 |
Did | -0.031 |
cher | -0.030 |
Token | a |
Token ID | 257 |
Feature activation | -4.52e-01 |
Pos logit contributions | |
cher | +0.136 |
ode | +0.134 |
uit | +0.132 |
boards | +0.131 |
Did | +0.130 |
Neg logit contributions | |
breath | -0.093 |
long | -0.081 |
excitement | -0.076 |
care | -0.072 |
home | -0.071 |
Token | big |
Token ID | 1263 |
Feature activation | -3.61e+00 |
Pos logit contributions | |
breath | +0.009 |
common | +0.007 |
long | +0.006 |
job | +0.006 |
sleep | +0.006 |
Neg logit contributions | |
boards | -0.012 |
auc | -0.012 |
uit | -0.012 |
Did | -0.012 |
cher | -0.012 |
Token | com |
Token ID | 401 |
Feature activation | -9.89e+00 |
Pos logit contributions | |
breath | +0.072 |
common | +0.056 |
job | +0.052 |
sleep | +0.052 |
long | +0.052 |
Neg logit contributions | |
boards | -0.094 |
Did | -0.092 |
uit | -0.091 |
auc | -0.091 |
washer | -0.090 |
Token | fy |
Token ID | 24928 |
Feature activation | -6.20e+00 |
Pos logit contributions | |
breath | +0.186 |
excitement | +0.134 |
common | +0.133 |
Sunday | +0.132 |
Wednesday | +0.126 |
Neg logit contributions | |
washer | -0.264 |
boards | -0.264 |
auc | -0.263 |
cher | -0.252 |
register | -0.246 |
Token | couch |
Token ID | 18507 |
Feature activation | -3.21e-01 |
Pos logit contributions | |
breath | +0.134 |
common | +0.117 |
care | +0.109 |
idea | +0.100 |
breaths | +0.100 |
Neg logit contributions | |
uit | -0.139 |
boards | -0.134 |
box | -0.133 |
washer | -0.130 |
lying | -0.129 |
Token | . |
Token ID | 13 |
Feature activation | +2.27e-01 |
Pos logit contributions | |
breath | +0.005 |
long | +0.004 |
sleep | +0.004 |
job | +0.004 |
time | +0.003 |
Neg logit contributions | |
Did | -0.009 |
cher | -0.009 |
ains | -0.009 |
Which | -0.009 |
auc | -0.009 |
Token | The |
Token ID | 383 |
Feature activation | +2.24e+00 |
Pos logit contributions | |
boards | +0.005 |
Did | +0.005 |
auc | +0.005 |
uit | +0.005 |
washer | +0.005 |
Neg logit contributions | |
breath | -0.005 |
common | -0.004 |
job | -0.004 |
sleep | -0.004 |
long | -0.004 |
Token | little |
Token ID | 1310 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
boards | +0.063 |
Did | +0.061 |
uit | +0.060 |
auc | +0.060 |
washer | +0.060 |
Neg logit contributions | |
breath | -0.042 |
common | -0.033 |
job | -0.029 |
sleep | -0.029 |
long | -0.029 |
Token | Would |
Token ID | 10928 |
Feature activation | +1.79e-01 |
Pos logit contributions | |
boards | +0.030 |
uit | +0.028 |
register | +0.028 |
auc | +0.028 |
cher | +0.027 |
Neg logit contributions | |
breath | -0.028 |
Today | -0.026 |
sleep | -0.025 |
bye | -0.024 |
common | -0.024 |
Token | ye |
Token ID | 9838 |
Feature activation | -9.43e-01 |
Pos logit contributions | |
boards | +0.004 |
Did | +0.004 |
uit | +0.004 |
auc | +0.004 |
cher | +0.004 |
Neg logit contributions | |
breath | -0.004 |
common | -0.004 |
job | -0.003 |
sleep | -0.003 |
care | -0.003 |
Token | be |
Token ID | 307 |
Feature activation | +7.48e-02 |
Pos logit contributions | |
breath | +0.019 |
common | +0.015 |
long | +0.014 |
sleep | +0.014 |
job | +0.014 |
Neg logit contributions | |
boards | -0.025 |
auc | -0.024 |
Did | -0.024 |
uit | -0.024 |
register | -0.024 |
Token | interested |
Token ID | 4609 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
cher | +0.002 |
uit | +0.002 |
Did | +0.002 |
washer | +0.002 |
ting | +0.002 |
Neg logit contributions | |
breath | -0.002 |
common | -0.001 |
sleep | -0.001 |
purpose | -0.001 |
job | -0.001 |
Token | in |
Token ID | 287 |
Feature activation | -2.88e+00 |
Pos logit contributions | |
breath | +0.028 |
common | +0.022 |
job | +0.021 |
sleep | +0.021 |
care | +0.021 |
Neg logit contributions | |
boards | -0.026 |
Did | -0.026 |
uit | -0.026 |
cher | -0.025 |
auc | -0.025 |
Token | com |
Token ID | 401 |
Feature activation | -9.87e+00 |
Pos logit contributions | |
breath | +0.062 |
common | +0.049 |
job | +0.046 |
sleep | +0.046 |
long | +0.045 |
Neg logit contributions | |
boards | -0.071 |
Did | -0.070 |
uit | -0.069 |
auc | -0.068 |
cher | -0.068 |
Token | in |
Token ID | 259 |
Feature activation | -2.62e+00 |
Pos logit contributions | |
breath | +0.265 |
excitement | +0.224 |
sight | +0.210 |
purpose | +0.203 |
happily | +0.203 |
Neg logit contributions | |
boards | -0.178 |
washer | -0.173 |
auc | -0.170 |
ode | -0.166 |
cher | -0.166 |
Token | ' |
Token ID | 6 |
Feature activation | -3.71e-01 |
Pos logit contributions | |
breath | +0.058 |
common | +0.046 |
long | +0.043 |
sleep | +0.043 |
job | +0.043 |
Neg logit contributions | |
boards | -0.063 |
Did | -0.061 |
auc | -0.061 |
uit | -0.061 |
cher | -0.061 |
Token | aboard |
Token ID | 15500 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
breath | +0.007 |
common | +0.006 |
job | +0.006 |
sleep | +0.006 |
long | +0.006 |
Neg logit contributions | |
boards | -0.010 |
Did | -0.010 |
uit | -0.010 |
auc | -0.009 |
cher | -0.009 |
Token | me |
Token ID | 502 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
breath | +0.038 |
common | +0.031 |
care | +0.030 |
job | +0.030 |
excitement | +0.030 |
Neg logit contributions | |
uit | -0.034 |
cher | -0.034 |
boards | -0.034 |
washer | -0.033 |
Did | -0.033 |
Token | ship |
Token ID | 4074 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
breath | +0.041 |
common | +0.033 |
excitement | +0.031 |
job | +0.030 |
accomplishment | +0.030 |
Neg logit contributions | |
boards | -0.039 |
cher | -0.039 |
washer | -0.039 |
uit | -0.038 |
auc | -0.038 |
Token | away |
Token ID | 1497 |
Feature activation | -2.41e+00 |
Pos logit contributions | |
boards | +0.030 |
uit | +0.030 |
Did | +0.029 |
auc | +0.029 |
cher | +0.029 |
Neg logit contributions | |
breath | -0.025 |
common | -0.020 |
job | -0.019 |
sleep | -0.018 |
care | -0.018 |
Token | , |
Token ID | 11 |
Feature activation | +5.07e-01 |
Pos logit contributions | |
breath | +0.050 |
common | +0.040 |
job | +0.038 |
sleep | +0.037 |
care | +0.037 |
Neg logit contributions | |
boards | -0.060 |
uit | -0.058 |
Did | -0.058 |
cher | -0.057 |
auc | -0.057 |
Token | a |
Token ID | 257 |
Feature activation | -5.45e-01 |
Pos logit contributions | |
boards | +0.013 |
auc | +0.012 |
uit | +0.012 |
Did | +0.012 |
washer | +0.012 |
Neg logit contributions | |
breath | -0.011 |
common | -0.009 |
long | -0.009 |
sleep | -0.009 |
job | -0.008 |
Token | bright |
Token ID | 6016 |
Feature activation | -3.68e+00 |
Pos logit contributions | |
breath | +0.011 |
common | +0.008 |
care | +0.008 |
sleep | +0.008 |
job | +0.007 |
Neg logit contributions | |
Did | -0.014 |
boards | -0.014 |
washer | -0.014 |
uit | -0.014 |
cher | -0.014 |
Token | blue |
Token ID | 4171 |
Feature activation | -2.19e+00 |
Pos logit contributions | |
breath | +0.071 |
common | +0.060 |
care | +0.054 |
job | +0.053 |
powers | +0.052 |
Neg logit contributions | |
uit | -0.087 |
lying | -0.085 |
Did | -0.085 |
boards | -0.084 |
auc | -0.084 |
Token | spe |
Token ID | 693 |
Feature activation | -9.86e+00 |
Pos logit contributions | |
breath | +0.049 |
common | +0.038 |
sleep | +0.036 |
long | +0.036 |
job | +0.036 |
Neg logit contributions | |
boards | -0.054 |
auc | -0.053 |
Did | -0.053 |
cher | -0.052 |
uit | -0.052 |
Token | ck |
Token ID | 694 |
Feature activation | +2.24e+00 |
Pos logit contributions | |
breath | +0.230 |
excitement | +0.226 |
Today | +0.217 |
sleep | +0.213 |
job | +0.209 |
Neg logit contributions | |
pped | -0.154 |
ode | -0.154 |
pled | -0.150 |
nut | -0.149 |
we | -0.148 |
Token | in |
Token ID | 287 |
Feature activation | -6.09e-01 |
Pos logit contributions | |
boards | +0.048 |
Did | +0.048 |
washer | +0.047 |
auc | +0.047 |
Which | +0.044 |
Neg logit contributions | |
breath | -0.059 |
long | -0.045 |
time | -0.044 |
happily | -0.043 |
home | -0.043 |
Token | the |
Token ID | 262 |
Feature activation | +2.19e+00 |
Pos logit contributions | |
breath | +0.012 |
common | +0.010 |
job | +0.010 |
sleep | +0.009 |
anna | +0.009 |
Neg logit contributions | |
Did | -0.015 |
uit | -0.014 |
boards | -0.014 |
auc | -0.014 |
ting | -0.014 |
Token | sky |
Token ID | 6766 |
Feature activation | -5.52e-01 |
Pos logit contributions | |
boards | +0.051 |
Did | +0.050 |
auc | +0.049 |
uit | +0.049 |
register | +0.048 |
Neg logit contributions | |
breath | -0.050 |
common | -0.041 |
job | -0.038 |
sleep | -0.038 |
long | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | +8.49e-01 |
Pos logit contributions | |
breath | +0.009 |
common | +0.007 |
long | +0.006 |
sleep | +0.006 |
job | +0.006 |
Neg logit contributions | |
boards | -0.016 |
auc | -0.016 |
Did | -0.016 |
cher | -0.016 |
uit | -0.016 |
Token | Soon |
Token ID | 15894 |
Feature activation | -4.03e+00 |
Pos logit contributions | |
boards | +0.007 |
Did | +0.006 |
uit | +0.006 |
auc | +0.006 |
washer | +0.006 |
Neg logit contributions | |
breath | -0.006 |
common | -0.004 |
job | -0.004 |
sleep | -0.004 |
long | -0.004 |
Token | , |
Token ID | 11 |
Feature activation | -4.76e-02 |
Pos logit contributions | |
breath | +0.063 |
common | +0.050 |
sleep | +0.042 |
job | +0.039 |
own | +0.037 |
Neg logit contributions | |
boards | -0.122 |
uit | -0.118 |
Did | -0.117 |
register | -0.115 |
ting | -0.114 |
Token | she |
Token ID | 673 |
Feature activation | -2.93e-03 |
Pos logit contributions | |
breath | +0.001 |
long | +0.001 |
job | +0.001 |
excitement | +0.001 |
common | +0.001 |
Neg logit contributions | |
auc | -0.001 |
cher | -0.001 |
boards | -0.001 |
uit | -0.001 |
ode | -0.001 |
Token | had |
Token ID | 550 |
Feature activation | +2.67e+00 |
Pos logit contributions | |
breath | +0.000 |
sleep | +0.000 |
long | +0.000 |
happily | +0.000 |
common | +0.000 |
Neg logit contributions | |
uit | -0.000 |
register | -0.000 |
boards | -0.000 |
auc | -0.000 |
Did | -0.000 |
Token | a |
Token ID | 257 |
Feature activation | -1.67e-01 |
Pos logit contributions | |
boards | +0.054 |
Did | +0.052 |
uit | +0.051 |
auc | +0.050 |
cher | +0.049 |
Neg logit contributions | |
breath | -0.071 |
common | -0.059 |
job | -0.056 |
sleep | -0.056 |
long | -0.055 |
Token | com |
Token ID | 401 |
Feature activation | -9.80e+00 |
Pos logit contributions | |
breath | +0.003 |
common | +0.003 |
care | +0.002 |
sleep | +0.002 |
anna | +0.002 |
Neg logit contributions | |
Did | -0.004 |
uit | -0.004 |
boards | -0.004 |
washer | -0.004 |
box | -0.004 |
Token | fy |
Token ID | 24928 |
Feature activation | -6.20e+00 |
Pos logit contributions | |
breath | +0.184 |
Sunday | +0.139 |
common | +0.135 |
Wednesday | +0.133 |
purpose | +0.133 |
Neg logit contributions | |
boards | -0.255 |
auc | -0.248 |
cher | -0.246 |
washer | -0.244 |
pled | -0.236 |
Token | home |
Token ID | 1363 |
Feature activation | -9.16e-01 |
Pos logit contributions | |
breath | +0.127 |
common | +0.122 |
care | +0.110 |
anna | +0.104 |
bye | +0.104 |
Neg logit contributions | |
uit | -0.135 |
box | -0.130 |
lying | -0.123 |
boards | -0.123 |
sticking | -0.121 |
Token | of |
Token ID | 286 |
Feature activation | +7.66e-01 |
Pos logit contributions | |
breath | +0.015 |
long | +0.011 |
sleep | +0.011 |
job | +0.011 |
home | +0.010 |
Neg logit contributions | |
Did | -0.028 |
boards | -0.027 |
auc | -0.027 |
ains | -0.026 |
cher | -0.026 |
Token | her |
Token ID | 607 |
Feature activation | -5.10e-01 |
Pos logit contributions | |
boards | +0.021 |
auc | +0.021 |
cher | +0.020 |
washer | +0.020 |
register | +0.020 |
Neg logit contributions | |
breath | -0.016 |
sleep | -0.013 |
long | -0.012 |
time | -0.012 |
home | -0.012 |
Token | own |
Token ID | 898 |
Feature activation | +5.62e-01 |
Pos logit contributions | |
breath | +0.010 |
common | +0.009 |
care | +0.008 |
anna | +0.007 |
excitement | +0.007 |
Neg logit contributions | |
washer | -0.012 |
lying | -0.012 |
box | -0.012 |
boards | -0.012 |
uit | -0.012 |
Token | away |
Token ID | 1497 |
Feature activation | -3.09e+00 |
Pos logit contributions | |
boards | +0.012 |
ting | +0.012 |
register | +0.011 |
uit | +0.011 |
washer | +0.011 |
Neg logit contributions | |
breath | -0.015 |
common | -0.013 |
appetite | -0.012 |
uth | -0.012 |
anna | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | +7.07e-01 |
Pos logit contributions | |
anna | +0.084 |
breath | +0.082 |
appetite | +0.076 |
� | +0.076 |
night | +0.074 |
Neg logit contributions | |
where | -0.042 |
lying | -0.041 |
ting | -0.040 |
who | -0.039 |
coming | -0.039 |
Token | They |
Token ID | 1119 |
Feature activation | +3.63e-01 |
Pos logit contributions | |
Did | +0.015 |
boards | +0.014 |
Why | +0.014 |
auc | +0.014 |
uit | +0.013 |
Neg logit contributions | |
breath | -0.019 |
job | -0.014 |
accomplishment | -0.014 |
sleep | -0.014 |
breaths | -0.014 |
Token | had |
Token ID | 550 |
Feature activation | +3.30e+00 |
Pos logit contributions | |
boards | +0.008 |
auc | +0.008 |
uit | +0.008 |
Did | +0.008 |
cher | +0.008 |
Neg logit contributions | |
breath | -0.009 |
common | -0.007 |
sleep | -0.007 |
long | -0.007 |
care | -0.007 |
Token | no |
Token ID | 645 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
boards | +0.066 |
Did | +0.065 |
uit | +0.064 |
auc | +0.063 |
washer | +0.062 |
Neg logit contributions | |
breath | -0.086 |
common | -0.071 |
job | -0.068 |
sleep | -0.067 |
long | -0.066 |
Token | com |
Token ID | 401 |
Feature activation | -9.76e+00 |
Pos logit contributions | |
breath | +0.042 |
common | +0.037 |
anna | +0.036 |
aire | +0.034 |
azel | +0.034 |
Neg logit contributions | |
register | -0.035 |
box | -0.034 |
uit | -0.034 |
lying | -0.034 |
sticking | -0.033 |
Token | bs |
Token ID | 1443 |
Feature activation | +7.63e-03 |
Pos logit contributions | |
breath | +0.212 |
common | +0.164 |
job | +0.158 |
purpose | +0.156 |
Sunday | +0.154 |
Neg logit contributions | |
boards | -0.240 |
auc | -0.233 |
cher | -0.230 |
Did | -0.229 |
uit | -0.224 |
Token | and |
Token ID | 290 |
Feature activation | -2.32e+00 |
Pos logit contributions | |
boards | +0.000 |
Did | +0.000 |
auc | +0.000 |
ode | +0.000 |
washer | +0.000 |
Neg logit contributions | |
breath | -0.000 |
time | -0.000 |
long | -0.000 |
home | -0.000 |
sleep | -0.000 |
Token | no |
Token ID | 645 |
Feature activation | -1.83e+00 |
Pos logit contributions | |
breath | +0.061 |
long | +0.051 |
common | +0.050 |
care | +0.049 |
sleep | +0.048 |
Neg logit contributions | |
uit | -0.050 |
boards | -0.049 |
auc | -0.046 |
ains | -0.046 |
register | -0.045 |
Token | fun |
Token ID | 1257 |
Feature activation | -5.91e+00 |
Pos logit contributions | |
breath | +0.039 |
common | +0.033 |
anna | +0.032 |
Today | +0.032 |
azel | +0.032 |
Neg logit contributions | |
uit | -0.038 |
register | -0.037 |
sticking | -0.037 |
lying | -0.037 |
boards | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | +8.22e-01 |
Pos logit contributions | |
breath | +0.128 |
common | +0.104 |
job | +0.095 |
sleep | +0.092 |
long | +0.092 |
Neg logit contributions | |
boards | -0.142 |
Did | -0.137 |
auc | -0.136 |
washer | -0.136 |
uit | -0.135 |
Token | they |
Token ID | 484 |
Feature activation | +2.40e+00 |
Pos logit contributions | |
breath | +0.019 |
long | +0.015 |
common | +0.014 |
sleep | +0.014 |
time | +0.014 |
Neg logit contributions | |
cher | -0.030 |
boards | -0.030 |
ode | -0.030 |
uit | -0.029 |
auc | -0.029 |
Token | had |
Token ID | 550 |
Feature activation | +4.23e+00 |
Pos logit contributions | |
auc | +0.062 |
boards | +0.058 |
uit | +0.056 |
register | +0.055 |
ting | +0.054 |
Neg logit contributions | |
breath | -0.053 |
long | -0.044 |
care | -0.044 |
time | -0.043 |
sleep | -0.042 |
Token | not |
Token ID | 407 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
auc | +0.103 |
boards | +0.103 |
washer | +0.095 |
register | +0.093 |
ting | +0.093 |
Neg logit contributions | |
breath | -0.101 |
common | -0.084 |
long | -0.082 |
care | -0.082 |
time | -0.081 |
Token | taken |
Token ID | 2077 |
Feature activation | +7.70e-01 |
Pos logit contributions | |
breath | +0.038 |
common | +0.031 |
sleep | +0.030 |
long | +0.030 |
care | +0.030 |
Neg logit contributions | |
boards | -0.035 |
auc | -0.034 |
uit | -0.033 |
Did | -0.033 |
register | -0.032 |
Token | the |
Token ID | 262 |
Feature activation | +2.39e+00 |
Pos logit contributions | |
boards | +0.016 |
uit | +0.015 |
ting | +0.015 |
Did | +0.015 |
register | +0.015 |
Neg logit contributions | |
breath | -0.019 |
common | -0.015 |
sleep | -0.015 |
job | -0.015 |
idea | -0.014 |
Token | com |
Token ID | 401 |
Feature activation | -9.73e+00 |
Pos logit contributions | |
boards | +0.053 |
Did | +0.051 |
uit | +0.051 |
auc | +0.051 |
cher | +0.050 |
Neg logit contributions | |
breath | -0.058 |
common | -0.047 |
sleep | -0.044 |
job | -0.044 |
long | -0.044 |
Token | bs |
Token ID | 1443 |
Feature activation | -9.99e-01 |
Pos logit contributions | |
breath | +0.169 |
excitement | +0.117 |
common | +0.112 |
sleep | +0.105 |
long | +0.103 |
Neg logit contributions | |
boards | -0.287 |
auc | -0.283 |
washer | -0.280 |
cher | -0.275 |
register | -0.271 |
Token | . |
Token ID | 13 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
breath | +0.019 |
common | +0.014 |
sleep | +0.013 |
long | +0.013 |
job | +0.013 |
Neg logit contributions | |
boards | -0.028 |
Did | -0.027 |
auc | -0.027 |
uit | -0.027 |
cher | -0.026 |
Token | They |
Token ID | 1119 |
Feature activation | +2.46e-01 |
Pos logit contributions | |
Did | +0.025 |
boards | +0.024 |
auc | +0.023 |
uit | +0.022 |
washer | +0.022 |
Neg logit contributions | |
breath | -0.024 |
job | -0.017 |
sleep | -0.017 |
common | -0.017 |
accomplishment | -0.017 |
Token | learned |
Token ID | 4499 |
Feature activation | +3.30e+00 |
Pos logit contributions | |
boards | +0.005 |
Did | +0.005 |
uit | +0.005 |
auc | +0.005 |
cher | +0.005 |
Neg logit contributions | |
breath | -0.006 |
common | -0.005 |
job | -0.005 |
sleep | -0.005 |
long | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | +7.07e-01 |
Pos logit contributions | |
boards | +0.078 |
Did | +0.074 |
uit | +0.074 |
cher | +0.073 |
auc | +0.073 |
Neg logit contributions | |
breath | -0.078 |
common | -0.066 |
long | -0.062 |
job | -0.062 |
sleep | -0.061 |
Token | , |
Token ID | 11 |
Feature activation | +4.89e-01 |
Pos logit contributions | |
breath | +0.037 |
long | +0.027 |
common | +0.027 |
sleep | +0.027 |
job | +0.027 |
Neg logit contributions | |
boards | -0.054 |
Did | -0.054 |
auc | -0.054 |
uit | -0.052 |
register | -0.052 |
Token | the |
Token ID | 262 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
boards | +0.014 |
auc | +0.013 |
register | +0.013 |
cher | +0.013 |
uit | +0.013 |
Neg logit contributions | |
breath | -0.010 |
sleep | -0.008 |
long | -0.008 |
content | -0.007 |
common | -0.007 |
Token | bird |
Token ID | 6512 |
Feature activation | +1.51e+00 |
Pos logit contributions | |
boards | +0.057 |
Did | +0.056 |
uit | +0.055 |
auc | +0.055 |
cher | +0.055 |
Neg logit contributions | |
breath | -0.054 |
common | -0.044 |
sleep | -0.041 |
job | -0.040 |
care | -0.040 |
Token | had |
Token ID | 550 |
Feature activation | +3.36e+00 |
Pos logit contributions | |
boards | +0.030 |
Did | +0.029 |
uit | +0.029 |
auc | +0.029 |
cher | +0.028 |
Neg logit contributions | |
breath | -0.040 |
common | -0.033 |
job | -0.031 |
sleep | -0.031 |
long | -0.031 |
Token | a |
Token ID | 257 |
Feature activation | -1.17e-02 |
Pos logit contributions | |
boards | +0.071 |
auc | +0.069 |
Did | +0.068 |
uit | +0.067 |
register | +0.066 |
Neg logit contributions | |
breath | -0.087 |
common | -0.071 |
long | -0.069 |
sleep | -0.069 |
care | -0.068 |
Token | com |
Token ID | 401 |
Feature activation | -9.66e+00 |
Pos logit contributions | |
breath | +0.000 |
common | +0.000 |
breaths | +0.000 |
anna | +0.000 |
care | +0.000 |
Neg logit contributions | |
Did | -0.000 |
uit | -0.000 |
washer | -0.000 |
boards | -0.000 |
box | -0.000 |
Token | fy |
Token ID | 24928 |
Feature activation | -5.93e+00 |
Pos logit contributions | |
breath | +0.176 |
Sunday | +0.133 |
common | +0.130 |
Wednesday | +0.130 |
purpose | +0.129 |
Neg logit contributions | |
boards | -0.246 |
washer | -0.243 |
cher | -0.241 |
auc | -0.240 |
pled | -0.233 |
Token | home |
Token ID | 1363 |
Feature activation | -8.15e-01 |
Pos logit contributions | |
breath | +0.113 |
common | +0.106 |
care | +0.090 |
Today | +0.086 |
� | +0.086 |
Neg logit contributions | |
uit | -0.140 |
box | -0.134 |
lying | -0.129 |
boards | -0.128 |
ode | -0.126 |
Token | , |
Token ID | 11 |
Feature activation | +5.27e-01 |
Pos logit contributions | |
breath | +0.012 |
long | +0.009 |
sleep | +0.009 |
home | +0.009 |
job | +0.008 |
Neg logit contributions | |
Did | -0.026 |
boards | -0.026 |
auc | -0.025 |
Which | -0.025 |
cher | -0.024 |
Token | and |
Token ID | 290 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
boards | +0.015 |
auc | +0.015 |
register | +0.014 |
cher | +0.014 |
washer | +0.014 |
Neg logit contributions | |
breath | -0.011 |
sleep | -0.008 |
long | -0.008 |
content | -0.008 |
common | -0.008 |
Token | the |
Token ID | 262 |
Feature activation | +2.26e+00 |
Pos logit contributions | |
breath | +0.026 |
long | +0.020 |
sleep | +0.020 |
common | +0.020 |
care | +0.020 |
Neg logit contributions | |
boards | -0.031 |
auc | -0.030 |
uit | -0.030 |
Did | -0.029 |
register | -0.029 |
Token | to |
Token ID | 284 |
Feature activation | -5.82e-01 |
Pos logit contributions | |
Did | +0.046 |
ains | +0.043 |
cher | +0.043 |
auc | +0.042 |
boards | +0.042 |
Neg logit contributions | |
breath | -0.025 |
long | -0.020 |
job | -0.018 |
home | -0.017 |
work | -0.017 |
Token | carry |
Token ID | 3283 |
Feature activation | +2.82e+00 |
Pos logit contributions | |
breath | +0.012 |
sleep | +0.010 |
long | +0.009 |
care | +0.009 |
common | +0.009 |
Neg logit contributions | |
boards | -0.016 |
Did | -0.016 |
auc | -0.015 |
register | -0.015 |
cher | -0.015 |
Token | supplies |
Token ID | 9416 |
Feature activation | +1.80e+00 |
Pos logit contributions | |
Did | +0.077 |
boards | +0.073 |
auc | +0.072 |
cher | +0.071 |
washer | +0.071 |
Neg logit contributions | |
breath | -0.060 |
long | -0.047 |
care | -0.046 |
common | -0.045 |
home | -0.044 |
Token | and |
Token ID | 290 |
Feature activation | -1.13e+00 |
Pos logit contributions | |
Did | +0.052 |
boards | +0.050 |
auc | +0.050 |
ains | +0.048 |
Is | +0.048 |
Neg logit contributions | |
breath | -0.033 |
long | -0.025 |
common | -0.024 |
job | -0.024 |
care | -0.023 |
Token | a |
Token ID | 257 |
Feature activation | +7.09e-02 |
Pos logit contributions | |
breath | +0.026 |
sleep | +0.021 |
long | +0.021 |
care | +0.021 |
job | +0.020 |
Neg logit contributions | |
boards | -0.027 |
Did | -0.027 |
auc | -0.027 |
uit | -0.027 |
register | -0.026 |
Token | com |
Token ID | 401 |
Feature activation | -9.65e+00 |
Pos logit contributions | |
boards | +0.002 |
Did | +0.002 |
uit | +0.002 |
auc | +0.002 |
cher | +0.002 |
Neg logit contributions | |
breath | -0.002 |
common | -0.001 |
sleep | -0.001 |
job | -0.001 |
long | -0.001 |
Token | fy |
Token ID | 24928 |
Feature activation | -6.20e+00 |
Pos logit contributions | |
breath | +0.173 |
common | +0.126 |
Sunday | +0.121 |
Wednesday | +0.117 |
purpose | +0.117 |
Neg logit contributions | |
boards | -0.269 |
washer | -0.265 |
auc | -0.263 |
cher | -0.262 |
uit | -0.249 |
Token | seat |
Token ID | 5852 |
Feature activation | +9.65e-02 |
Pos logit contributions | |
breath | +0.134 |
common | +0.117 |
care | +0.102 |
job | +0.100 |
sleep | +0.100 |
Neg logit contributions | |
uit | -0.141 |
boards | -0.141 |
box | -0.134 |
auc | -0.133 |
washer | -0.132 |
Token | for |
Token ID | 329 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
Did | +0.003 |
boards | +0.003 |
Which | +0.003 |
ains | +0.003 |
auc | +0.003 |
Neg logit contributions | |
breath | -0.002 |
long | -0.001 |
job | -0.001 |
sleep | -0.001 |
home | -0.001 |
Token | him |
Token ID | 683 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
boards | +0.032 |
auc | +0.031 |
Did | +0.031 |
washer | +0.030 |
cher | +0.030 |
Neg logit contributions | |
breath | -0.027 |
long | -0.021 |
common | -0.021 |
sleep | -0.021 |
care | -0.020 |
Token | to |
Token ID | 284 |
Feature activation | -2.41e+00 |
Pos logit contributions | |
breath | +0.021 |
common | +0.016 |
sleep | +0.015 |
long | +0.015 |
job | +0.015 |
Neg logit contributions | |
boards | -0.032 |
Did | -0.032 |
auc | -0.031 |
uit | -0.031 |
register | -0.030 |
Token | place |
Token ID | 1295 |
Feature activation | +1.63e+00 |
Pos logit contributions | |
boards | +0.057 |
washer | +0.057 |
register | +0.056 |
Did | +0.055 |
uit | +0.055 |
Neg logit contributions | |
breath | -0.038 |
common | -0.031 |
care | -0.027 |
sleep | -0.027 |
anna | -0.026 |
Token | had |
Token ID | 550 |
Feature activation | +3.49e+00 |
Pos logit contributions | |
Did | +0.033 |
uit | +0.033 |
boards | +0.033 |
cher | +0.032 |
ode | +0.031 |
Neg logit contributions | |
breath | -0.043 |
common | -0.035 |
job | -0.035 |
long | -0.034 |
home | -0.034 |
Token | a |
Token ID | 257 |
Feature activation | -2.81e-01 |
Pos logit contributions | |
cher | +0.083 |
auc | +0.082 |
uit | +0.081 |
register | +0.079 |
washer | +0.079 |
Neg logit contributions | |
breath | -0.080 |
long | -0.069 |
common | -0.064 |
time | -0.063 |
sleep | -0.063 |
Token | lot |
Token ID | 1256 |
Feature activation | -2.33e+00 |
Pos logit contributions | |
breath | +0.005 |
common | +0.004 |
sleep | +0.004 |
care | +0.004 |
job | +0.004 |
Neg logit contributions | |
boards | -0.007 |
Did | -0.007 |
uit | -0.007 |
washer | -0.007 |
auc | -0.007 |
Token | of |
Token ID | 286 |
Feature activation | -4.12e-01 |
Pos logit contributions | |
breath | +0.045 |
long | +0.038 |
common | +0.037 |
time | +0.035 |
home | +0.035 |
Neg logit contributions | |
Did | -0.064 |
cher | -0.062 |
ode | -0.059 |
uit | -0.058 |
ains | -0.058 |
Token | com |
Token ID | 401 |
Feature activation | -9.61e+00 |
Pos logit contributions | |
breath | +0.008 |
common | +0.006 |
job | +0.005 |
sleep | +0.005 |
long | +0.005 |
Neg logit contributions | |
boards | -0.012 |
Did | -0.011 |
auc | -0.011 |
uit | -0.011 |
cher | -0.011 |
Token | fy |
Token ID | 24928 |
Feature activation | -6.07e+00 |
Pos logit contributions | |
breath | +0.222 |
common | +0.178 |
purpose | +0.170 |
excitement | +0.170 |
Sunday | +0.170 |
Neg logit contributions | |
boards | -0.229 |
auc | -0.203 |
ting | -0.197 |
uit | -0.196 |
pled | -0.196 |
Token | chairs |
Token ID | 18791 |
Feature activation | -1.73e+00 |
Pos logit contributions | |
breath | +0.116 |
common | +0.091 |
job | +0.083 |
sleep | +0.082 |
care | +0.082 |
Neg logit contributions | |
boards | -0.166 |
uit | -0.159 |
Did | -0.157 |
auc | -0.156 |
ting | -0.154 |
Token | and |
Token ID | 290 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
breath | +0.023 |
long | +0.017 |
sleep | +0.016 |
time | +0.014 |
home | +0.014 |
Neg logit contributions | |
Did | -0.057 |
Which | -0.056 |
cher | -0.056 |
auc | -0.056 |
ode | -0.054 |
Token | cush |
Token ID | 24736 |
Feature activation | -2.39e+00 |
Pos logit contributions | |
breath | +0.024 |
long | +0.022 |
excitement | +0.019 |
sleep | +0.019 |
common | +0.018 |
Neg logit contributions | |
uit | -0.028 |
auc | -0.028 |
cher | -0.028 |
ode | -0.027 |
register | -0.027 |
Token | ions |
Token ID | 507 |
Feature activation | +1.24e-02 |
Pos logit contributions | |
breath | +0.084 |
common | +0.067 |
long | +0.067 |
breaths | +0.067 |
happily | +0.066 |
Neg logit contributions | |
Did | -0.030 |
Which | -0.030 |
Which | -0.027 |
Is | -0.026 |
Does | -0.026 |
Token | but |
Token ID | 475 |
Feature activation | +1.98e-01 |
Pos logit contributions | |
boards | +0.014 |
auc | +0.014 |
register | +0.013 |
uit | +0.013 |
washer | +0.013 |
Neg logit contributions | |
breath | -0.013 |
sleep | -0.010 |
long | -0.010 |
common | -0.010 |
care | -0.010 |
Token | only |
Token ID | 691 |
Feature activation | -2.55e+00 |
Pos logit contributions | |
boards | +0.005 |
auc | +0.004 |
uit | +0.004 |
Did | +0.004 |
cher | +0.004 |
Neg logit contributions | |
breath | -0.005 |
common | -0.004 |
long | -0.004 |
job | -0.004 |
sleep | -0.004 |
Token | in |
Token ID | 287 |
Feature activation | -3.06e+00 |
Pos logit contributions | |
breath | +0.054 |
long | +0.048 |
sleep | +0.044 |
time | +0.043 |
job | +0.043 |
Neg logit contributions | |
boards | -0.069 |
auc | -0.069 |
uit | -0.066 |
ains | -0.064 |
register | -0.063 |
Token | his |
Token ID | 465 |
Feature activation | -7.78e-01 |
Pos logit contributions | |
breath | +0.062 |
common | +0.047 |
long | +0.045 |
sleep | +0.045 |
job | +0.044 |
Neg logit contributions | |
boards | -0.084 |
auc | -0.082 |
Did | -0.080 |
cher | -0.080 |
washer | -0.079 |
Token | big |
Token ID | 1263 |
Feature activation | -4.09e+00 |
Pos logit contributions | |
breath | +0.013 |
common | +0.011 |
anna | +0.009 |
� | +0.009 |
uth | +0.009 |
Neg logit contributions | |
boards | -0.020 |
washer | -0.020 |
box | -0.020 |
uit | -0.020 |
register | -0.019 |
Token | com |
Token ID | 401 |
Feature activation | -9.59e+00 |
Pos logit contributions | |
breath | +0.070 |
common | +0.067 |
uth | +0.059 |
bye | +0.058 |
anna | +0.057 |
Neg logit contributions | |
box | -0.100 |
boards | -0.099 |
washer | -0.096 |
uit | -0.093 |
ode | -0.093 |
Token | fy |
Token ID | 24928 |
Feature activation | -5.91e+00 |
Pos logit contributions | |
breath | +0.187 |
common | +0.148 |
Sunday | +0.140 |
purpose | +0.139 |
excitement | +0.137 |
Neg logit contributions | |
boards | -0.246 |
auc | -0.239 |
washer | -0.237 |
cher | -0.231 |
ode | -0.226 |
Token | chair |
Token ID | 5118 |
Feature activation | -7.64e-01 |
Pos logit contributions | |
breath | +0.115 |
common | +0.107 |
� | +0.092 |
care | +0.089 |
job | +0.088 |
Neg logit contributions | |
boards | -0.144 |
box | -0.140 |
uit | -0.140 |
washer | -0.133 |
auc | -0.130 |
Token | upstairs |
Token ID | 26148 |
Feature activation | -2.23e+00 |
Pos logit contributions | |
breath | +0.013 |
sleep | +0.010 |
long | +0.010 |
job | +0.009 |
common | +0.009 |
Neg logit contributions | |
Did | -0.022 |
boards | -0.022 |
auc | -0.022 |
register | -0.021 |
ains | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | +4.78e-01 |
Pos logit contributions | |
breath | +0.040 |
common | +0.029 |
long | +0.029 |
sleep | +0.029 |
job | +0.028 |
Neg logit contributions | |
boards | -0.062 |
Did | -0.062 |
auc | -0.062 |
uit | -0.061 |
cher | -0.061 |
Token | He |
Token ID | 679 |
Feature activation | +9.39e-01 |
Pos logit contributions | |
boards | +0.013 |
auc | +0.013 |
Did | +0.013 |
cher | +0.013 |
uit | +0.013 |
Neg logit contributions | |
breath | -0.009 |
common | -0.007 |
sleep | -0.007 |
long | -0.006 |
job | -0.006 |
Token | little |
Token ID | 1310 |
Feature activation | -4.92e+00 |
Pos logit contributions | |
breath | +0.035 |
common | +0.026 |
care | +0.024 |
breaths | +0.023 |
sleep | +0.023 |
Neg logit contributions | |
Did | -0.050 |
washer | -0.049 |
boards | -0.049 |
auc | -0.048 |
register | -0.047 |
Token | bird |
Token ID | 6512 |
Feature activation | -2.23e-01 |
Pos logit contributions | |
breath | +0.109 |
common | +0.096 |
accomplishment | +0.083 |
care | +0.083 |
excitement | +0.082 |
Neg logit contributions | |
washer | -0.113 |
auc | -0.112 |
boards | -0.107 |
Did | -0.106 |
register | -0.105 |
Token | . |
Token ID | 13 |
Feature activation | -1.17e-02 |
Pos logit contributions | |
breath | +0.006 |
common | +0.005 |
job | +0.005 |
excitement | +0.005 |
care | +0.005 |
Neg logit contributions | |
Did | -0.004 |
auc | -0.004 |
boards | -0.004 |
washer | -0.004 |
uit | -0.004 |
Token | The |
Token ID | 383 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
breath | +0.000 |
common | +0.000 |
accomplishment | +0.000 |
anna | +0.000 |
job | +0.000 |
Neg logit contributions | |
Did | -0.000 |
boards | -0.000 |
lying | -0.000 |
sticking | -0.000 |
uit | -0.000 |
Token | bird |
Token ID | 6512 |
Feature activation | +1.97e+00 |
Pos logit contributions | |
Did | +0.037 |
lying | +0.037 |
ter | +0.035 |
auc | +0.035 |
washer | +0.035 |
Neg logit contributions | |
breath | -0.056 |
common | -0.051 |
care | -0.046 |
azel | -0.045 |
breaths | -0.045 |
Token | wanted |
Token ID | 2227 |
Feature activation | +3.27e+00 |
Pos logit contributions | |
Did | +0.031 |
boards | +0.030 |
washer | +0.029 |
auc | +0.029 |
cher | +0.027 |
Neg logit contributions | |
breath | -0.061 |
common | -0.052 |
job | -0.050 |
excitement | -0.048 |
sleep | -0.048 |
Token | to |
Token ID | 284 |
Feature activation | -5.38e-01 |
Pos logit contributions | |
boards | +0.084 |
Did | +0.082 |
auc | +0.081 |
uit | +0.081 |
cher | +0.080 |
Neg logit contributions | |
breath | -0.068 |
common | -0.054 |
job | -0.050 |
sleep | -0.050 |
long | -0.049 |
Token | fly |
Token ID | 6129 |
Feature activation | +1.68e-01 |
Pos logit contributions | |
breath | +0.013 |
common | +0.012 |
anna | +0.012 |
breaths | +0.011 |
uth | +0.011 |
Neg logit contributions | |
lying | -0.009 |
uit | -0.009 |
sticking | -0.009 |
ode | -0.009 |
cher | -0.009 |
Token | high |
Token ID | 1029 |
Feature activation | -5.32e+00 |
Pos logit contributions | |
Did | +0.003 |
lying | +0.003 |
uit | +0.003 |
boards | +0.003 |
cher | +0.003 |
Neg logit contributions | |
breath | -0.004 |
common | -0.003 |
anna | -0.003 |
uth | -0.003 |
accomplishment | -0.003 |
Token | in |
Token ID | 287 |
Feature activation | -1.99e+00 |
Pos logit contributions | |
breath | +0.145 |
common | +0.127 |
care | +0.122 |
excitement | +0.120 |
accomplishment | +0.118 |
Neg logit contributions | |
boards | -0.095 |
ting | -0.090 |
auc | -0.085 |
lying | -0.085 |
washer | -0.084 |
Token | the |
Token ID | 262 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
breath | +0.047 |
common | +0.039 |
job | +0.036 |
sleep | +0.035 |
anna | +0.035 |
Neg logit contributions | |
Did | -0.044 |
boards | -0.042 |
auc | -0.041 |
uit | -0.041 |
ting | -0.041 |
Token | The |
Token ID | 383 |
Feature activation | +2.61e+00 |
Pos logit contributions | |
boards | +0.016 |
Did | +0.015 |
uit | +0.015 |
auc | +0.015 |
washer | +0.015 |
Neg logit contributions | |
breath | -0.014 |
common | -0.011 |
job | -0.010 |
sleep | -0.010 |
long | -0.010 |
Token | two |
Token ID | 734 |
Feature activation | -9.88e-02 |
Pos logit contributions | |
uit | +0.088 |
auc | +0.085 |
cher | +0.085 |
we | +0.084 |
boards | +0.084 |
Neg logit contributions | |
breath | -0.037 |
long | -0.028 |
job | -0.026 |
idea | -0.026 |
excitement | -0.025 |
Token | of |
Token ID | 286 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
breath | +0.002 |
long | +0.002 |
care | +0.002 |
excitement | +0.002 |
job | +0.002 |
Neg logit contributions | |
auc | -0.002 |
uit | -0.002 |
Did | -0.002 |
boards | -0.002 |
cher | -0.002 |
Token | them |
Token ID | 606 |
Feature activation | +9.81e-01 |
Pos logit contributions | |
breath | +0.028 |
common | +0.023 |
sleep | +0.022 |
long | +0.022 |
care | +0.022 |
Neg logit contributions | |
boards | -0.020 |
Did | -0.019 |
auc | -0.019 |
cher | -0.019 |
uit | -0.019 |
Token | polished |
Token ID | 23895 |
Feature activation | +7.58e-01 |
Pos logit contributions | |
boards | +0.023 |
auc | +0.022 |
Did | +0.022 |
uit | +0.022 |
register | +0.021 |
Neg logit contributions | |
breath | -0.023 |
sleep | -0.018 |
long | -0.018 |
common | -0.017 |
care | -0.017 |
Token | the |
Token ID | 262 |
Feature activation | +3.41e+00 |
Pos logit contributions | |
Did | +0.021 |
boards | +0.021 |
auc | +0.020 |
cher | +0.020 |
register | +0.020 |
Neg logit contributions | |
breath | -0.016 |
common | -0.012 |
job | -0.011 |
long | -0.011 |
sleep | -0.011 |
Token | belt |
Token ID | 10999 |
Feature activation | +2.63e+00 |
Pos logit contributions | |
auc | +0.072 |
boards | +0.069 |
Did | +0.068 |
cher | +0.068 |
rooms | +0.066 |
Neg logit contributions | |
breath | -0.089 |
long | -0.077 |
job | -0.072 |
sleep | -0.071 |
time | -0.070 |
Token | just |
Token ID | 655 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
Did | +0.071 |
boards | +0.070 |
auc | +0.069 |
uit | +0.068 |
cher | +0.068 |
Neg logit contributions | |
breath | -0.051 |
common | -0.038 |
long | -0.038 |
job | -0.037 |
sleep | -0.037 |
Token | the |
Token ID | 262 |
Feature activation | +2.39e+00 |
Pos logit contributions | |
breath | +0.020 |
care | +0.015 |
long | +0.015 |
sleep | +0.015 |
job | +0.015 |
Neg logit contributions | |
auc | -0.030 |
boards | -0.030 |
Did | -0.029 |
washer | -0.028 |
cher | -0.028 |
Token | way |
Token ID | 835 |
Feature activation | -1.00e+00 |
Pos logit contributions | |
auc | +0.064 |
boards | +0.062 |
cher | +0.060 |
uit | +0.060 |
rooms | +0.060 |
Neg logit contributions | |
breath | -0.049 |
long | -0.041 |
job | -0.040 |
time | -0.039 |
sleep | -0.038 |
Token | his |
Token ID | 465 |
Feature activation | -2.03e-02 |
Pos logit contributions | |
breath | +0.021 |
common | +0.016 |
long | +0.016 |
job | +0.015 |
sleep | +0.015 |
Neg logit contributions | |
boards | -0.026 |
Did | -0.026 |
auc | -0.025 |
uit | -0.025 |
cher | -0.025 |
Token | Everyone |
Token ID | 11075 |
Feature activation | +2.61e+00 |
Pos logit contributions | |
boards | +0.014 |
Did | +0.013 |
uit | +0.013 |
auc | +0.013 |
cher | +0.013 |
Neg logit contributions | |
breath | -0.011 |
common | -0.009 |
long | -0.008 |
sleep | -0.008 |
job | -0.008 |
Token | was |
Token ID | 373 |
Feature activation | +8.83e-01 |
Pos logit contributions | |
boards | +0.051 |
uit | +0.051 |
auc | +0.049 |
Did | +0.049 |
cher | +0.048 |
Neg logit contributions | |
breath | -0.070 |
common | -0.057 |
sleep | -0.055 |
care | -0.055 |
long | -0.055 |
Token | so |
Token ID | 523 |
Feature activation | -5.58e+00 |
Pos logit contributions | |
boards | +0.018 |
Did | +0.018 |
uit | +0.018 |
auc | +0.018 |
washer | +0.018 |
Neg logit contributions | |
breath | -0.022 |
common | -0.018 |
job | -0.017 |
sleep | -0.017 |
care | -0.017 |
Token | excited |
Token ID | 6568 |
Feature activation | -5.76e+00 |
Pos logit contributions | |
breath | +0.111 |
common | +0.080 |
sleep | +0.080 |
job | +0.079 |
breaths | +0.078 |
Neg logit contributions | |
washer | -0.149 |
Did | -0.147 |
boards | -0.146 |
auc | -0.144 |
cher | -0.139 |
Token | to |
Token ID | 284 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
breath | +0.131 |
common | +0.105 |
job | +0.098 |
sleep | +0.098 |
long | +0.097 |
Neg logit contributions | |
boards | -0.139 |
Did | -0.132 |
uit | -0.132 |
auc | -0.132 |
washer | -0.130 |
Token | put |
Token ID | 1234 |
Feature activation | +3.22e+00 |
Pos logit contributions | |
breath | +0.022 |
common | +0.017 |
job | +0.017 |
long | +0.016 |
sleep | +0.016 |
Neg logit contributions | |
uit | -0.024 |
boards | -0.024 |
cher | -0.023 |
Did | -0.023 |
auc | -0.023 |
Token | them |
Token ID | 606 |
Feature activation | +1.97e+00 |
Pos logit contributions | |
Did | +0.088 |
boards | +0.086 |
ode | +0.082 |
auc | +0.082 |
washer | +0.081 |
Neg logit contributions | |
breath | -0.072 |
care | -0.054 |
common | -0.053 |
long | -0.052 |
home | -0.051 |
Token | on |
Token ID | 319 |
Feature activation | -3.71e+00 |
Pos logit contributions | |
boards | +0.041 |
Did | +0.041 |
auc | +0.039 |
register | +0.038 |
uit | +0.038 |
Neg logit contributions | |
breath | -0.052 |
common | -0.041 |
long | -0.040 |
sleep | -0.040 |
job | -0.040 |
Token | ! |
Token ID | 0 |
Feature activation | +2.71e-01 |
Pos logit contributions | |
breath | +0.076 |
common | +0.059 |
long | +0.056 |
job | +0.056 |
sleep | +0.055 |
Neg logit contributions | |
boards | -0.097 |
Did | -0.095 |
auc | -0.094 |
uit | -0.093 |
cher | -0.093 |
Token |
|
Token ID | 198 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
boards | +0.007 |
Did | +0.007 |
auc | +0.007 |
uit | +0.007 |
washer | +0.006 |
Neg logit contributions | |
breath | -0.006 |
common | -0.005 |
job | -0.004 |
sleep | -0.004 |
long | -0.004 |
Token |
|
Token ID | 198 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
breath | +0.041 |
job | +0.032 |
common | +0.031 |
breaths | +0.030 |
care | +0.029 |
Neg logit contributions | |
boards | -0.040 |
Did | -0.040 |
uit | -0.038 |
washer | -0.038 |
auc | -0.037 |
Token | amazed |
Token ID | 24562 |
Feature activation | -3.59e+00 |
Pos logit contributions | |
breath | +0.099 |
breaths | +0.075 |
common | +0.072 |
anna | +0.072 |
job | +0.071 |
Neg logit contributions | |
Did | -0.131 |
washer | -0.127 |
lying | -0.124 |
auc | -0.123 |
can | -0.122 |
Token | that |
Token ID | 326 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
breath | +0.078 |
common | +0.062 |
sleep | +0.058 |
job | +0.058 |
long | +0.058 |
Neg logit contributions | |
boards | -0.089 |
Did | -0.086 |
auc | -0.086 |
uit | -0.086 |
cher | -0.084 |
Token | he |
Token ID | 339 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
boards | +0.048 |
uit | +0.048 |
cher | +0.047 |
ode | +0.047 |
Did | +0.046 |
Neg logit contributions | |
breath | -0.035 |
long | -0.028 |
common | -0.028 |
excitement | -0.027 |
care | -0.026 |
Token | asked |
Token ID | 1965 |
Feature activation | +2.31e+00 |
Pos logit contributions | |
boards | +0.003 |
Did | +0.003 |
cher | +0.003 |
washer | +0.003 |
uit | +0.003 |
Neg logit contributions | |
breath | -0.003 |
common | -0.003 |
job | -0.003 |
sleep | -0.003 |
care | -0.002 |
Token | " |
Token ID | 366 |
Feature activation | +4.46e+00 |
Pos logit contributions | |
boards | +0.062 |
Did | +0.062 |
ode | +0.061 |
uit | +0.060 |
cher | +0.060 |
Neg logit contributions | |
breath | -0.047 |
common | -0.036 |
long | -0.035 |
care | -0.035 |
home | -0.035 |
Token | What |
Token ID | 2061 |
Feature activation | +6.12e+00 |
Pos logit contributions | |
auc | +0.075 |
boards | +0.073 |
washer | +0.072 |
box | +0.067 |
register | +0.065 |
Neg logit contributions | |
breath | -0.134 |
time | -0.108 |
� | -0.107 |
sight | -0.107 |
week | -0.106 |
Token | made |
Token ID | 925 |
Feature activation | +3.81e+00 |
Pos logit contributions | |
uit | +0.150 |
ting | +0.141 |
register | +0.139 |
boards | +0.138 |
auc | +0.136 |
Neg logit contributions | |
breath | -0.132 |
common | -0.128 |
long | -0.128 |
time | -0.125 |
," | -0.121 |
Token | the |
Token ID | 262 |
Feature activation | +2.45e+00 |
Pos logit contributions | |
boards | +0.093 |
Did | +0.090 |
auc | +0.089 |
uit | +0.089 |
cher | +0.088 |
Neg logit contributions | |
breath | -0.085 |
common | -0.068 |
job | -0.064 |
sleep | -0.064 |
long | -0.064 |
Token | rainbow |
Token ID | 27223 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
boards | +0.055 |
auc | +0.054 |
washer | +0.053 |
Did | +0.053 |
ains | +0.052 |
Neg logit contributions | |
breath | -0.055 |
common | -0.046 |
care | -0.042 |
sleep | -0.041 |
anna | -0.041 |
Token | ?". |
Token ID | 43634 |
Feature activation | -2.88e+00 |
Pos logit contributions | |
boards | +0.032 |
Did | +0.031 |
uit | +0.031 |
auc | +0.031 |
cher | +0.030 |
Neg logit contributions | |
breath | -0.024 |
common | -0.019 |
job | -0.018 |
long | -0.018 |
sleep | -0.018 |
Token | Jill |
Token ID | 22772 |
Feature activation | +3.47e-01 |
Pos logit contributions | |
breath | +0.073 |
common | +0.060 |
job | +0.057 |
sleep | +0.057 |
long | +0.056 |
Neg logit contributions | |
boards | -0.061 |
Did | -0.059 |
auc | -0.059 |
uit | -0.059 |
cher | -0.058 |
Token |
|
Token ID | 198 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
boards | +0.015 |
Did | +0.015 |
auc | +0.015 |
uit | +0.015 |
washer | +0.014 |
Neg logit contributions | |
breath | -0.016 |
common | -0.012 |
job | -0.012 |
care | -0.011 |
sleep | -0.011 |
Token |
|
Token ID | 198 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
breath | +0.037 |
job | +0.029 |
breaths | +0.027 |
common | +0.027 |
care | +0.026 |
Neg logit contributions | |
Did | -0.034 |
uit | -0.032 |
boards | -0.032 |
auc | -0.031 |
washer | -0.030 |
Token | Tom |
Token ID | 13787 |
Feature activation | +2.50e-01 |
Pos logit contributions | |
breath | +0.045 |
job | +0.036 |
breaths | +0.035 |
idea | +0.034 |
common | +0.034 |
Neg logit contributions | |
Did | -0.033 |
uit | -0.029 |
boards | -0.029 |
cher | -0.028 |
Is | -0.028 |
Token | and |
Token ID | 290 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
boards | +0.006 |
Did | +0.005 |
uit | +0.005 |
auc | +0.005 |
cher | +0.005 |
Neg logit contributions | |
breath | -0.006 |
common | -0.005 |
sleep | -0.005 |
care | -0.005 |
long | -0.005 |
Token | Sam |
Token ID | 3409 |
Feature activation | -2.61e-01 |
Pos logit contributions | |
breath | +0.036 |
common | +0.029 |
job | +0.028 |
sleep | +0.026 |
long | +0.026 |
Neg logit contributions | |
Did | -0.036 |
boards | -0.036 |
uit | -0.036 |
auc | -0.035 |
cher | -0.035 |
Token | got |
Token ID | 1392 |
Feature activation | +3.74e+00 |
Pos logit contributions | |
breath | +0.007 |
common | +0.006 |
sleep | +0.006 |
care | +0.006 |
long | +0.005 |
Neg logit contributions | |
boards | -0.005 |
auc | -0.005 |
Did | -0.005 |
uit | -0.005 |
cher | -0.005 |
Token | scared |
Token ID | 12008 |
Feature activation | -3.95e+00 |
Pos logit contributions | |
boards | +0.072 |
auc | +0.070 |
Did | +0.070 |
washer | +0.069 |
uit | +0.068 |
Neg logit contributions | |
breath | -0.100 |
common | -0.085 |
job | -0.080 |
sleep | -0.080 |
long | -0.079 |
Token | . |
Token ID | 13 |
Feature activation | +6.56e-01 |
Pos logit contributions | |
breath | +0.079 |
common | +0.062 |
job | +0.057 |
sleep | +0.057 |
long | +0.056 |
Neg logit contributions | |
boards | -0.105 |
Did | -0.103 |
auc | -0.102 |
uit | -0.102 |
cher | -0.100 |
Token | They |
Token ID | 1119 |
Feature activation | -1.54e-01 |
Pos logit contributions | |
boards | +0.015 |
Did | +0.015 |
auc | +0.015 |
uit | +0.014 |
washer | +0.014 |
Neg logit contributions | |
breath | -0.016 |
common | -0.012 |
job | -0.012 |
care | -0.011 |
sleep | -0.011 |
Token | tried |
Token ID | 3088 |
Feature activation | +1.79e+00 |
Pos logit contributions | |
breath | +0.004 |
common | +0.003 |
job | +0.003 |
sleep | +0.003 |
long | +0.003 |
Neg logit contributions | |
boards | -0.003 |
Did | -0.003 |
auc | -0.003 |
uit | -0.003 |
cher | -0.003 |
Token | to |
Token ID | 284 |
Feature activation | -4.12e-01 |
Pos logit contributions | |
boards | +0.049 |
auc | +0.048 |
Did | +0.048 |
cher | +0.047 |
uit | +0.047 |
Neg logit contributions | |
breath | -0.035 |
common | -0.027 |
long | -0.025 |
sleep | -0.025 |
job | -0.025 |
Token | are |
Token ID | 389 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
boards | +0.001 |
Did | +0.001 |
auc | +0.001 |
uit | +0.001 |
cher | +0.001 |
Neg logit contributions | |
breath | -0.001 |
common | -0.001 |
sleep | -0.001 |
job | -0.001 |
long | -0.001 |
Token | on |
Token ID | 319 |
Feature activation | -4.89e+00 |
Pos logit contributions | |
boards | +0.050 |
auc | +0.048 |
uit | +0.048 |
Did | +0.048 |
cher | +0.047 |
Neg logit contributions | |
breath | -0.032 |
common | -0.024 |
long | -0.023 |
job | -0.022 |
sleep | -0.022 |
Token | the |
Token ID | 262 |
Feature activation | +1.51e+00 |
Pos logit contributions | |
breath | +0.120 |
anna | +0.101 |
bye | +0.099 |
night | +0.097 |
awk | +0.095 |
Neg logit contributions | |
lying | -0.090 |
Did | -0.087 |
uit | -0.087 |
sticking | -0.086 |
boards | -0.085 |
Token | track |
Token ID | 2610 |
Feature activation | +3.02e+00 |
Pos logit contributions | |
boards | +0.035 |
Did | +0.035 |
uit | +0.034 |
auc | +0.034 |
cher | +0.034 |
Neg logit contributions | |
breath | -0.035 |
common | -0.028 |
long | -0.028 |
job | -0.028 |
sleep | -0.027 |
Token | . |
Token ID | 13 |
Feature activation | +8.40e-01 |
Pos logit contributions | |
Did | +0.091 |
boards | +0.089 |
uit | +0.088 |
cher | +0.087 |
ode | +0.087 |
Neg logit contributions | |
breath | -0.047 |
long | -0.034 |
sleep | -0.033 |
job | -0.033 |
common | -0.032 |
Token | The |
Token ID | 383 |
Feature activation | +2.52e+00 |
Pos logit contributions | |
boards | +0.019 |
Did | +0.019 |
auc | +0.019 |
uit | +0.019 |
washer | +0.018 |
Neg logit contributions | |
breath | -0.020 |
common | -0.016 |
job | -0.015 |
sleep | -0.015 |
care | -0.015 |
Token | big |
Token ID | 1263 |
Feature activation | -2.34e+00 |
Pos logit contributions | |
boards | +0.061 |
Did | +0.059 |
uit | +0.059 |
auc | +0.058 |
cher | +0.058 |
Neg logit contributions | |
breath | -0.058 |
long | -0.047 |
common | -0.047 |
job | -0.046 |
sleep | -0.044 |
Token | train |
Token ID | 4512 |
Feature activation | +3.47e+00 |
Pos logit contributions | |
breath | +0.057 |
common | +0.047 |
sleep | +0.044 |
job | +0.044 |
care | +0.044 |
Neg logit contributions | |
boards | -0.051 |
auc | -0.049 |
Did | -0.049 |
uit | -0.048 |
washer | -0.048 |
Token | cannot |
Token ID | 2314 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
uit | +0.074 |
boards | +0.073 |
Did | +0.073 |
auc | +0.072 |
cher | +0.071 |
Neg logit contributions | |
breath | -0.086 |
common | -0.069 |
long | -0.068 |
job | -0.068 |
sleep | -0.068 |
Token | stop |
Token ID | 2245 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
breath | +0.048 |
content | +0.042 |
uth | +0.041 |
breaths | +0.041 |
anna | +0.041 |
Neg logit contributions | |
uit | -0.030 |
auc | -0.030 |
washer | -0.029 |
cher | -0.028 |
boards | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
breath | +0.028 |
common | +0.023 |
job | +0.021 |
sleep | +0.021 |
care | +0.020 |
Neg logit contributions | |
ting | -0.031 |
boards | -0.030 |
Did | -0.030 |
auc | -0.030 |
sticking | -0.030 |
Token | eat |
Token ID | 4483 |
Feature activation | +1.79e+00 |
Pos logit contributions | |
breath | +0.029 |
common | +0.023 |
job | +0.022 |
long | +0.021 |
sleep | +0.021 |
Neg logit contributions | |
Did | -0.026 |
uit | -0.026 |
boards | -0.026 |
auc | -0.025 |
register | -0.025 |
Token | it |
Token ID | 340 |
Feature activation | +7.60e-01 |
Pos logit contributions | |
boards | +0.049 |
Did | +0.048 |
auc | +0.048 |
uit | +0.048 |
cher | +0.048 |
Neg logit contributions | |
breath | -0.034 |
common | -0.026 |
long | -0.025 |
job | -0.025 |
sleep | -0.024 |
Token | , |
Token ID | 11 |
Feature activation | +8.59e-01 |
Pos logit contributions | |
boards | +0.021 |
auc | +0.020 |
washer | +0.019 |
uit | +0.019 |
cher | +0.019 |
Neg logit contributions | |
breath | -0.015 |
common | -0.011 |
job | -0.010 |
sleep | -0.010 |
excitement | -0.010 |
Token | or |
Token ID | 393 |
Feature activation | -3.31e-01 |
Pos logit contributions | |
boards | +0.022 |
auc | +0.022 |
cher | +0.022 |
uit | +0.021 |
washer | +0.021 |
Neg logit contributions | |
breath | -0.018 |
long | -0.014 |
sleep | -0.014 |
common | -0.014 |
job | -0.013 |
Token | trade |
Token ID | 3292 |
Feature activation | +4.31e+00 |
Pos logit contributions | |
breath | +0.008 |
long | +0.006 |
sleep | +0.006 |
care | +0.006 |
common | +0.006 |
Neg logit contributions | |
boards | -0.008 |
uit | -0.008 |
auc | -0.008 |
cher | -0.007 |
ains | -0.007 |
Token | it |
Token ID | 340 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
boards | +0.127 |
Did | +0.124 |
auc | +0.124 |
uit | +0.123 |
cher | +0.121 |
Neg logit contributions | |
breath | -0.074 |
common | -0.055 |
long | -0.051 |
job | -0.050 |
care | -0.050 |
Token | for |
Token ID | 329 |
Feature activation | +8.11e-01 |
Pos logit contributions | |
boards | +0.029 |
Did | +0.028 |
auc | +0.028 |
uit | +0.028 |
cher | +0.027 |
Neg logit contributions | |
breath | -0.020 |
common | -0.016 |
job | -0.014 |
sleep | -0.014 |
long | -0.014 |
Token | something |
Token ID | 1223 |
Feature activation | +6.29e-01 |
Pos logit contributions | |
boards | +0.019 |
Did | +0.018 |
uit | +0.018 |
cher | +0.018 |
auc | +0.018 |
Neg logit contributions | |
breath | -0.019 |
common | -0.015 |
job | -0.014 |
sleep | -0.014 |
long | -0.014 |
Token | else |
Token ID | 2073 |
Feature activation | -2.95e+00 |
Pos logit contributions | |
boards | +0.015 |
Did | +0.015 |
uit | +0.015 |
auc | +0.014 |
cher | +0.014 |
Neg logit contributions | |
breath | -0.014 |
common | -0.011 |
long | -0.011 |
job | -0.011 |
sleep | -0.011 |
Token | ." |
Token ID | 526 |
Feature activation | -2.05e+00 |
Pos logit contributions | |
breath | +0.052 |
common | +0.037 |
sleep | +0.034 |
job | +0.034 |
long | +0.033 |
Neg logit contributions | |
boards | -0.085 |
Did | -0.083 |
washer | -0.082 |
auc | -0.082 |
cher | -0.082 |
Token |
|
Token ID | 198 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
breath | +0.047 |
common | +0.036 |
job | +0.036 |
sleep | +0.034 |
care | +0.033 |
Neg logit contributions | |
Did | -0.050 |
boards | -0.050 |
uit | -0.048 |
auc | -0.047 |
washer | -0.046 |
Token | wherever |
Token ID | 14530 |
Feature activation | -1.36e-01 |
Pos logit contributions | |
boards | +0.001 |
auc | +0.001 |
Did | +0.001 |
uit | +0.001 |
register | +0.001 |
Neg logit contributions | |
breath | -0.000 |
common | -0.000 |
sleep | -0.000 |
long | -0.000 |
job | -0.000 |
Token | he |
Token ID | 339 |
Feature activation | -6.43e-01 |
Pos logit contributions | |
breath | +0.004 |
common | +0.003 |
job | +0.003 |
sleep | +0.003 |
long | +0.003 |
Neg logit contributions | |
boards | -0.003 |
Did | -0.003 |
auc | -0.003 |
uit | -0.003 |
cher | -0.003 |
Token | went |
Token ID | 1816 |
Feature activation | +4.81e-01 |
Pos logit contributions | |
breath | +0.015 |
common | +0.012 |
sleep | +0.012 |
job | +0.012 |
long | +0.012 |
Neg logit contributions | |
boards | -0.015 |
Did | -0.014 |
uit | -0.014 |
auc | -0.014 |
cher | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | +5.83e-01 |
Pos logit contributions | |
boards | +0.014 |
Did | +0.014 |
auc | +0.013 |
uit | +0.013 |
cher | +0.013 |
Neg logit contributions | |
breath | -0.008 |
common | -0.006 |
sleep | -0.006 |
long | -0.006 |
job | -0.006 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -4.90e-01 |
Pos logit contributions | |
boards | +0.014 |
Did | +0.014 |
uit | +0.014 |
auc | +0.014 |
washer | +0.014 |
Neg logit contributions | |
breath | -0.013 |
common | -0.010 |
job | -0.010 |
long | -0.009 |
care | -0.009 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
breath | +0.009 |
common | +0.007 |
sleep | +0.007 |
long | +0.006 |
job | +0.006 |
Neg logit contributions | |
boards | -0.014 |
auc | -0.013 |
Did | -0.013 |
uit | -0.013 |
cher | -0.013 |
Token | upon |
Token ID | 2402 |
Feature activation | +3.00e+00 |
Pos logit contributions | |
breath | +0.033 |
common | +0.028 |
care | +0.028 |
job | +0.028 |
breaths | +0.027 |
Neg logit contributions | |
washer | -0.022 |
boards | -0.020 |
ains | -0.020 |
uit | -0.020 |
auc | -0.020 |
Token | a |
Token ID | 257 |
Feature activation | -2.58e+00 |
Pos logit contributions | |
boards | +0.062 |
auc | +0.060 |
Did | +0.059 |
uit | +0.059 |
cher | +0.058 |
Neg logit contributions | |
breath | -0.078 |
common | -0.064 |
sleep | -0.061 |
long | -0.061 |
job | -0.061 |
Token | time |
Token ID | 640 |
Feature activation | -3.86e+00 |
Pos logit contributions | |
breath | +0.045 |
breaths | +0.035 |
common | +0.031 |
care | +0.028 |
excitement | +0.028 |
Neg logit contributions | |
Did | -0.069 |
boards | -0.066 |
uit | -0.066 |
pled | -0.065 |
lying | -0.065 |
Token | , |
Token ID | 11 |
Feature activation | +9.24e-01 |
Pos logit contributions | |
breath | +0.073 |
common | +0.056 |
care | +0.052 |
sleep | +0.051 |
job | +0.051 |
Neg logit contributions | |
boards | -0.103 |
uit | -0.102 |
Did | -0.101 |
lying | -0.101 |
box | -0.099 |
Token | there |
Token ID | 612 |
Feature activation | +1.19e+00 |
Pos logit contributions | |
boards | +0.018 |
Did | +0.017 |
auc | +0.017 |
uit | +0.017 |
cher | +0.016 |
Neg logit contributions | |
breath | -0.026 |
common | -0.021 |
long | -0.020 |
job | -0.020 |
sleep | -0.020 |
Token | cry |
Token ID | 3960 |
Feature activation | -2.85e+00 |
Pos logit contributions | |
breath | +0.031 |
common | +0.026 |
sleep | +0.024 |
anna | +0.023 |
breaths | +0.022 |
Neg logit contributions | |
Did | -0.046 |
cher | -0.046 |
washer | -0.046 |
boards | -0.045 |
uit | -0.045 |
Token | of |
Token ID | 286 |
Feature activation | -3.36e-01 |
Pos logit contributions | |
breath | +0.030 |
long | +0.016 |
sleep | +0.015 |
common | +0.014 |
job | +0.014 |
Neg logit contributions | |
boards | -0.104 |
Did | -0.102 |
auc | -0.100 |
uit | -0.099 |
washer | -0.099 |
Token | joy |
Token ID | 8716 |
Feature activation | -3.98e+00 |
Pos logit contributions | |
breath | +0.007 |
common | +0.005 |
sleep | +0.005 |
long | +0.005 |
job | +0.005 |
Neg logit contributions | |
boards | -0.009 |
Did | -0.009 |
auc | -0.009 |
uit | -0.009 |
washer | -0.009 |
Token | , |
Token ID | 11 |
Feature activation | -8.82e-01 |
Pos logit contributions | |
breath | +0.066 |
sleep | +0.044 |
common | +0.044 |
long | +0.043 |
care | +0.041 |
Neg logit contributions | |
boards | -0.121 |
Did | -0.119 |
auc | -0.118 |
uit | -0.115 |
register | -0.115 |
Token | the |
Token ID | 262 |
Feature activation | +1.68e+00 |
Pos logit contributions | |
breath | +0.019 |
common | +0.015 |
long | +0.014 |
sleep | +0.014 |
job | +0.014 |
Neg logit contributions | |
boards | -0.022 |
Did | -0.021 |
auc | -0.021 |
uit | -0.021 |
cher | -0.021 |
Token | boy |
Token ID | 2933 |
Feature activation | +1.42e-01 |
Pos logit contributions | |
boards | +0.042 |
auc | +0.040 |
Did | +0.040 |
uit | +0.040 |
cher | +0.039 |
Neg logit contributions | |
breath | -0.037 |
common | -0.029 |
long | -0.028 |
job | -0.028 |
sleep | -0.028 |
Token | rushed |
Token ID | 13999 |
Feature activation | +2.19e+00 |
Pos logit contributions | |
boards | +0.003 |
Did | +0.003 |
auc | +0.003 |
uit | +0.003 |
cher | +0.003 |
Neg logit contributions | |
breath | -0.003 |
common | -0.003 |
sleep | -0.003 |
long | -0.003 |
job | -0.003 |
Token | forward |
Token ID | 2651 |
Feature activation | -4.35e+00 |
Pos logit contributions | |
boards | +0.047 |
Did | +0.045 |
auc | +0.045 |
uit | +0.045 |
cher | +0.044 |
Neg logit contributions | |
breath | -0.055 |
common | -0.045 |
sleep | -0.043 |
job | -0.043 |
long | -0.043 |
Token | and |
Token ID | 290 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
breath | +0.087 |
common | +0.068 |
job | +0.064 |
sleep | +0.063 |
long | +0.062 |
Neg logit contributions | |
boards | -0.114 |
Did | -0.110 |
uit | -0.110 |
auc | -0.110 |
cher | -0.109 |
Token | the |
Token ID | 262 |
Feature activation | +2.78e+00 |
Pos logit contributions | |
breath | +0.034 |
common | +0.026 |
sleep | +0.026 |
care | +0.025 |
job | +0.025 |
Neg logit contributions | |
boards | -0.038 |
auc | -0.037 |
uit | -0.037 |
Did | -0.036 |
register | -0.035 |
Token | par |
Token ID | 1582 |
Feature activation | +8.26e-01 |
Pos logit contributions | |
boards | +0.070 |
auc | +0.068 |
Did | +0.068 |
uit | +0.068 |
cher | +0.066 |
Neg logit contributions | |
breath | -0.060 |
common | -0.047 |
long | -0.045 |
job | -0.045 |
sleep | -0.044 |
Token | could |
Token ID | 714 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
uit | +0.014 |
auc | +0.014 |
boards | +0.014 |
ting | +0.014 |
bling | +0.013 |
Neg logit contributions | |
breath | -0.023 |
care | -0.020 |
long | -0.020 |
happily | -0.020 |
," | -0.020 |
Token | eat |
Token ID | 4483 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
breath | +0.031 |
sleep | +0.024 |
common | +0.023 |
job | +0.023 |
care | +0.023 |
Neg logit contributions | |
boards | -0.042 |
Did | -0.041 |
auc | -0.041 |
cher | -0.041 |
uit | -0.040 |
Token | the |
Token ID | 262 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
boards | +0.048 |
cher | +0.047 |
auc | +0.047 |
washer | +0.047 |
Did | +0.046 |
Neg logit contributions | |
breath | -0.034 |
long | -0.026 |
common | -0.025 |
job | -0.025 |
sleep | -0.024 |
Token | pie |
Token ID | 2508 |
Feature activation | +1.60e+00 |
Pos logit contributions | |
boards | +0.045 |
Did | +0.043 |
auc | +0.043 |
uit | +0.043 |
cher | +0.042 |
Neg logit contributions | |
breath | -0.074 |
common | -0.062 |
job | -0.060 |
sleep | -0.060 |
long | -0.060 |
Token | . |
Token ID | 13 |
Feature activation | +2.11e-01 |
Pos logit contributions | |
Did | +0.050 |
auc | +0.049 |
boards | +0.049 |
cher | +0.048 |
uit | +0.047 |
Neg logit contributions | |
breath | -0.024 |
job | -0.017 |
long | -0.017 |
sleep | -0.017 |
common | -0.016 |
Token | But |
Token ID | 887 |
Feature activation | -5.72e-02 |
Pos logit contributions | |
boards | +0.005 |
Did | +0.005 |
washer | +0.004 |
uit | +0.004 |
auc | +0.004 |
Neg logit contributions | |
breath | -0.005 |
common | -0.004 |
sleep | -0.004 |
breaths | -0.004 |
long | -0.004 |
Token | her |
Token ID | 607 |
Feature activation | +1.85e-01 |
Pos logit contributions | |
breath | +0.001 |
common | +0.001 |
sleep | +0.001 |
long | +0.001 |
job | +0.001 |
Neg logit contributions | |
boards | -0.002 |
uit | -0.002 |
Did | -0.002 |
auc | -0.001 |
register | -0.001 |
Token | mom |
Token ID | 1995 |
Feature activation | +7.67e-01 |
Pos logit contributions | |
boards | +0.004 |
washer | +0.004 |
box | +0.004 |
lying | +0.004 |
uit | +0.004 |
Neg logit contributions | |
breath | -0.004 |
common | -0.003 |
care | -0.003 |
sleep | -0.003 |
anna | -0.003 |
Token | answered |
Token ID | 9373 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
auc | +0.023 |
boards | +0.023 |
Did | +0.023 |
ains | +0.023 |
uit | +0.022 |
Neg logit contributions | |
breath | -0.012 |
job | -0.008 |
long | -0.008 |
idea | -0.008 |
care | -0.008 |
Token | " |
Token ID | 366 |
Feature activation | +4.83e+00 |
Pos logit contributions | |
auc | +0.047 |
ode | +0.047 |
cher | +0.047 |
boards | +0.046 |
Did | +0.046 |
Neg logit contributions | |
breath | -0.013 |
long | -0.008 |
happily | -0.007 |
time | -0.007 |
sleep | -0.007 |
Token | No |
Token ID | 2949 |
Feature activation | -1.95e+00 |
Pos logit contributions | |
boards | +0.116 |
auc | +0.113 |
uit | +0.112 |
washer | +0.111 |
register | +0.110 |
Neg logit contributions | |
breath | -0.108 |
common | -0.086 |
care | -0.083 |
long | -0.083 |
job | -0.082 |
Token | be |
Token ID | 307 |
Feature activation | +4.48e-01 |
Pos logit contributions | |
boards | +0.014 |
auc | +0.013 |
Did | +0.013 |
uit | +0.013 |
cher | +0.013 |
Neg logit contributions | |
breath | -0.018 |
common | -0.015 |
sleep | -0.014 |
job | -0.014 |
long | -0.014 |
Token | careful |
Token ID | 8161 |
Feature activation | -3.04e+00 |
Pos logit contributions | |
Did | +0.011 |
cher | +0.011 |
uit | +0.011 |
boards | +0.011 |
auc | +0.010 |
Neg logit contributions | |
breath | -0.010 |
common | -0.008 |
sleep | -0.007 |
job | -0.007 |
long | -0.007 |
Token | with |
Token ID | 351 |
Feature activation | +1.69e+00 |
Pos logit contributions | |
breath | +0.060 |
common | +0.047 |
sleep | +0.043 |
job | +0.043 |
long | +0.042 |
Neg logit contributions | |
boards | -0.080 |
auc | -0.078 |
uit | -0.078 |
Did | -0.078 |
washer | -0.077 |
Token | the |
Token ID | 262 |
Feature activation | +2.99e+00 |
Pos logit contributions | |
boards | +0.044 |
auc | +0.042 |
Did | +0.041 |
washer | +0.041 |
register | +0.041 |
Neg logit contributions | |
breath | -0.038 |
common | -0.029 |
long | -0.029 |
sleep | -0.028 |
care | -0.028 |
Token | toy |
Token ID | 13373 |
Feature activation | +2.89e+00 |
Pos logit contributions | |
boards | +0.070 |
Did | +0.067 |
uit | +0.067 |
auc | +0.066 |
cher | +0.065 |
Neg logit contributions | |
breath | -0.072 |
common | -0.057 |
long | -0.056 |
job | -0.056 |
sleep | -0.055 |
Token | ." |
Token ID | 526 |
Feature activation | -2.55e+00 |
Pos logit contributions | |
boards | +0.077 |
Did | +0.077 |
auc | +0.074 |
uit | +0.074 |
cher | +0.073 |
Neg logit contributions | |
breath | -0.057 |
job | -0.043 |
common | -0.043 |
sleep | -0.042 |
long | -0.042 |
Token | She |
Token ID | 1375 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
breath | +0.067 |
common | +0.055 |
job | +0.053 |
breaths | +0.050 |
accomplishment | +0.049 |
Neg logit contributions | |
Did | -0.054 |
uit | -0.051 |
boards | -0.050 |
washer | -0.048 |
auc | -0.048 |
Token | smiled |
Token ID | 13541 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
breath | +0.004 |
common | +0.003 |
sleep | +0.003 |
care | +0.003 |
long | +0.003 |
Neg logit contributions | |
boards | -0.003 |
Did | -0.003 |
uit | -0.003 |
auc | -0.003 |
cher | -0.003 |
Token | and |
Token ID | 290 |
Feature activation | -1.95e+00 |
Pos logit contributions | |
breath | +0.027 |
common | +0.025 |
job | +0.022 |
care | +0.021 |
sleep | +0.021 |
Neg logit contributions | |
uit | -0.025 |
Did | -0.025 |
boards | -0.025 |
lying | -0.024 |
ting | -0.024 |
Token | hugged |
Token ID | 41130 |
Feature activation | +1.62e+00 |
Pos logit contributions | |
breath | +0.046 |
common | +0.037 |
long | +0.035 |
sleep | +0.035 |
job | +0.035 |
Neg logit contributions | |
boards | -0.045 |
auc | -0.044 |
Did | -0.044 |
uit | -0.044 |
register | -0.042 |
Token | them |
Token ID | 606 |
Feature activation | +2.44e+00 |
Pos logit contributions | |
boards | +0.036 |
uit | +0.036 |
ting | +0.035 |
Did | +0.035 |
auc | +0.034 |
Neg logit contributions | |
breath | -0.038 |
common | -0.032 |
job | -0.030 |
long | -0.029 |
sleep | -0.029 |
Token | , |
Token ID | 11 |
Feature activation | +3.76e-02 |
Pos logit contributions | |
breath | +0.079 |
common | +0.058 |
sleep | +0.053 |
anna | +0.050 |
accomplishment | +0.050 |
Neg logit contributions | |
boards | -0.137 |
Did | -0.134 |
uit | -0.132 |
sticking | -0.130 |
auc | -0.129 |
Token | he |
Token ID | 339 |
Feature activation | -1.97e-01 |
Pos logit contributions | |
boards | +0.001 |
Did | +0.001 |
uit | +0.001 |
auc | +0.001 |
washer | +0.001 |
Neg logit contributions | |
breath | -0.001 |
common | -0.001 |
sleep | -0.001 |
job | -0.001 |
care | -0.001 |
Token | found |
Token ID | 1043 |
Feature activation | +6.29e+00 |
Pos logit contributions | |
breath | +0.005 |
common | +0.004 |
sleep | +0.004 |
job | +0.004 |
long | +0.004 |
Neg logit contributions | |
boards | -0.004 |
Did | -0.004 |
uit | -0.004 |
auc | -0.004 |
cher | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +7.01e-01 |
Pos logit contributions | |
boards | +0.199 |
ains | +0.198 |
ode | +0.196 |
cher | +0.195 |
auc | +0.190 |
Neg logit contributions | |
breath | -0.097 |
long | -0.078 |
home | -0.078 |
care | -0.072 |
sleep | -0.069 |
Token | tap |
Token ID | 9814 |
Feature activation | +3.54e+00 |
Pos logit contributions | |
boards | +0.021 |
ting | +0.020 |
auc | +0.020 |
Did | +0.020 |
uit | +0.020 |
Neg logit contributions | |
breath | -0.012 |
long | -0.010 |
common | -0.010 |
job | -0.010 |
time | -0.009 |
Token | in |
Token ID | 287 |
Feature activation | -1.92e+00 |
Pos logit contributions | |
ains | +0.099 |
Did | +0.097 |
cher | +0.095 |
boards | +0.091 |
uit | +0.091 |
Neg logit contributions | |
breath | -0.067 |
long | -0.058 |
job | -0.052 |
, | -0.051 |
day | -0.051 |
Token | his |
Token ID | 465 |
Feature activation | +5.67e-03 |
Pos logit contributions | |
breath | +0.039 |
common | +0.030 |
long | +0.029 |
sleep | +0.028 |
job | +0.028 |
Neg logit contributions | |
boards | -0.052 |
auc | -0.050 |
cher | -0.050 |
Did | -0.050 |
uit | -0.049 |
Token | yard |
Token ID | 12699 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
boards | +0.000 |
Did | +0.000 |
cher | +0.000 |
auc | +0.000 |
uit | +0.000 |
Neg logit contributions | |
breath | -0.000 |
common | -0.000 |
job | -0.000 |
sleep | -0.000 |
long | -0.000 |
Token | . |
Token ID | 13 |
Feature activation | +7.07e-01 |
Pos logit contributions | |
cher | +0.039 |
auc | +0.037 |
Did | +0.037 |
boards | +0.037 |
register | +0.037 |
Neg logit contributions | |
breath | -0.015 |
sleep | -0.010 |
long | -0.010 |
time | -0.010 |
job | -0.009 |
Token | He |
Token ID | 679 |
Feature activation | +8.53e-01 |
Pos logit contributions | |
boards | +0.017 |
Did | +0.016 |
auc | +0.016 |
uit | +0.016 |
cher | +0.016 |
Neg logit contributions | |
breath | -0.016 |
common | -0.013 |
sleep | -0.012 |
long | -0.012 |
job | -0.012 |
Token | thought |
Token ID | 1807 |
Feature activation | +2.58e+00 |
Pos logit contributions | |
boards | +0.018 |
Did | +0.018 |
auc | +0.017 |
uit | +0.017 |
washer | +0.017 |
Neg logit contributions | |
breath | -0.022 |
common | -0.018 |
job | -0.017 |
sleep | -0.017 |
long | -0.017 |
Token | want |
Token ID | 765 |
Feature activation | +2.66e+00 |
Pos logit contributions | |
boards | +0.004 |
auc | +0.004 |
Did | +0.004 |
uit | +0.004 |
washer | +0.004 |
Neg logit contributions | |
breath | -0.005 |
common | -0.004 |
long | -0.004 |
job | -0.004 |
sleep | -0.004 |
Token | to |
Token ID | 284 |
Feature activation | -1.73e+00 |
Pos logit contributions | |
boards | +0.077 |
Did | +0.076 |
uit | +0.076 |
cher | +0.075 |
auc | +0.075 |
Neg logit contributions | |
breath | -0.048 |
common | -0.036 |
long | -0.035 |
sleep | -0.034 |
care | -0.034 |
Token | pick |
Token ID | 2298 |
Feature activation | +5.81e+00 |
Pos logit contributions | |
breath | +0.042 |
common | +0.035 |
breaths | +0.035 |
content | +0.035 |
anna | +0.034 |
Neg logit contributions | |
uit | -0.032 |
boards | -0.031 |
cher | -0.031 |
lying | -0.030 |
washer | -0.030 |
Token | flowers |
Token ID | 12734 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
Did | +0.147 |
boards | +0.143 |
uit | +0.140 |
washer | +0.139 |
cher | +0.139 |
Neg logit contributions | |
breath | -0.128 |
long | -0.101 |
common | -0.100 |
home | -0.099 |
care | -0.099 |
Token | ," |
Token ID | 553 |
Feature activation | -2.32e+00 |
Pos logit contributions | |
boards | +0.035 |
Did | +0.034 |
auc | +0.034 |
uit | +0.034 |
cher | +0.033 |
Neg logit contributions | |
breath | -0.024 |
common | -0.018 |
job | -0.017 |
sleep | -0.017 |
long | -0.017 |
Token | Mia |
Token ID | 32189 |
Feature activation | -1.67e+00 |
Pos logit contributions | |
breath | +0.056 |
job | +0.047 |
anna | +0.044 |
common | +0.044 |
sleep | +0.043 |
Neg logit contributions | |
Did | -0.050 |
auc | -0.049 |
boards | -0.048 |
washer | -0.046 |
register | -0.045 |
Token | said |
Token ID | 531 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
breath | +0.043 |
common | +0.035 |
job | +0.035 |
purpose | +0.032 |
sleep | +0.032 |
Neg logit contributions | |
boards | -0.036 |
washer | -0.034 |
Did | -0.034 |
uit | -0.034 |
auc | -0.034 |
Token | . |
Token ID | 13 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
boards | +0.053 |
Did | +0.052 |
auc | +0.051 |
uit | +0.051 |
cher | +0.051 |
Neg logit contributions | |
breath | -0.034 |
common | -0.026 |
job | -0.024 |
sleep | -0.023 |
long | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | -1.77e+00 |
Pos logit contributions | |
boards | +0.032 |
Did | +0.032 |
auc | +0.031 |
uit | +0.031 |
cher | +0.031 |
Neg logit contributions | |
breath | -0.024 |
common | -0.018 |
job | -0.017 |
sleep | -0.017 |
long | -0.017 |
Token |
|
Token ID | 198 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
breath | +0.043 |
job | +0.034 |
common | +0.033 |
breaths | +0.031 |
excitement | +0.031 |
Neg logit contributions | |
Did | -0.040 |
boards | -0.039 |
uit | -0.039 |
auc | -0.038 |
washer | -0.037 |
Token | " |
Token ID | 1 |
Feature activation | +5.30e+00 |
Pos logit contributions | |
breath | +0.039 |
job | +0.030 |
sleep | +0.029 |
idea | +0.029 |
common | +0.029 |
Neg logit contributions | |
Did | -0.040 |
boards | -0.037 |
auc | -0.037 |
uit | -0.037 |
washer | -0.036 |
Token | up |
Token ID | 510 |
Feature activation | -3.06e+00 |
Pos logit contributions | |
breath | +0.009 |
common | +0.008 |
job | +0.007 |
long | +0.007 |
sleep | +0.007 |
Neg logit contributions | |
boards | -0.008 |
uit | -0.008 |
auc | -0.008 |
Did | -0.008 |
washer | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
breath | +0.061 |
common | +0.047 |
job | +0.044 |
sleep | +0.044 |
long | +0.043 |
Neg logit contributions | |
boards | -0.081 |
auc | -0.078 |
uit | -0.078 |
Did | -0.078 |
washer | -0.076 |
Token | Lily |
Token ID | 20037 |
Feature activation | +8.12e-01 |
Pos logit contributions | |
boards | +0.028 |
Did | +0.027 |
uit | +0.026 |
auc | +0.026 |
register | +0.026 |
Neg logit contributions | |
breath | -0.030 |
common | -0.024 |
job | -0.022 |
care | -0.021 |
breaths | -0.021 |
Token | learned |
Token ID | 4499 |
Feature activation | +2.97e+00 |
Pos logit contributions | |
boards | +0.014 |
Did | +0.013 |
uit | +0.013 |
washer | +0.013 |
auc | +0.013 |
Neg logit contributions | |
breath | -0.024 |
common | -0.021 |
job | -0.020 |
long | -0.019 |
sleep | -0.019 |
Token | that |
Token ID | 326 |
Feature activation | +2.80e+00 |
Pos logit contributions | |
boards | +0.065 |
cher | +0.063 |
auc | +0.063 |
Did | +0.062 |
ains | +0.061 |
Neg logit contributions | |
breath | -0.074 |
common | -0.063 |
long | -0.062 |
job | -0.061 |
care | -0.060 |
Token | sometimes |
Token ID | 3360 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
boards | +0.066 |
uit | +0.064 |
cher | +0.064 |
auc | +0.064 |
register | +0.062 |
Neg logit contributions | |
breath | -0.066 |
common | -0.053 |
long | -0.053 |
sleep | -0.053 |
job | -0.053 |
Token | it |
Token ID | 340 |
Feature activation | +1.44e+00 |
Pos logit contributions | |
breath | +0.041 |
common | +0.033 |
sleep | +0.030 |
job | +0.029 |
long | +0.029 |
Neg logit contributions | |
boards | -0.044 |
Did | -0.043 |
washer | -0.042 |
ting | -0.041 |
register | -0.041 |
Token | 's |
Token ID | 338 |
Feature activation | +2.86e+00 |
Pos logit contributions | |
uit | +0.030 |
boards | +0.030 |
Did | +0.029 |
cher | +0.028 |
auc | +0.028 |
Neg logit contributions | |
breath | -0.036 |
care | -0.031 |
common | -0.030 |
job | -0.030 |
sleep | -0.030 |
Token | important |
Token ID | 1593 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
boards | +0.072 |
auc | +0.069 |
uit | +0.067 |
Did | +0.067 |
cher | +0.067 |
Neg logit contributions | |
breath | -0.064 |
common | -0.052 |
job | -0.050 |
long | -0.050 |
sleep | -0.049 |
Token | to |
Token ID | 284 |
Feature activation | -4.85e-01 |
Pos logit contributions | |
breath | +0.024 |
common | +0.018 |
sleep | +0.017 |
long | +0.017 |
job | +0.017 |
Neg logit contributions | |
boards | -0.028 |
Did | -0.027 |
uit | -0.026 |
auc | -0.026 |
washer | -0.026 |
Token | listen |
Token ID | 6004 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
breath | +0.010 |
common | +0.008 |
sleep | +0.007 |
job | +0.007 |
care | +0.007 |
Neg logit contributions | |
boards | -0.013 |
Did | -0.013 |
auc | -0.013 |
cher | -0.013 |
washer | -0.013 |
Token | and |
Token ID | 290 |
Feature activation | -5.35e-01 |
Pos logit contributions | |
cher | +0.005 |
Which | +0.005 |
Did | +0.005 |
uit | +0.005 |
ains | +0.005 |
Neg logit contributions | |
breath | -0.002 |
work | -0.002 |
long | -0.002 |
sleep | -0.002 |
care | -0.001 |
Token | smiled |
Token ID | 13541 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
breath | +0.012 |
common | +0.010 |
job | +0.010 |
sleep | +0.009 |
purpose | +0.009 |
Neg logit contributions | |
boards | -0.012 |
Did | -0.012 |
washer | -0.012 |
auc | -0.012 |
cher | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +4.77e-01 |
Pos logit contributions | |
breath | +0.032 |
common | +0.028 |
care | +0.024 |
job | +0.024 |
uth | +0.024 |
Neg logit contributions | |
boards | -0.032 |
washer | -0.031 |
Did | -0.031 |
ting | -0.031 |
lying | -0.031 |
Token | She |
Token ID | 1375 |
Feature activation | +7.02e-01 |
Pos logit contributions | |
boards | +0.010 |
auc | +0.010 |
Did | +0.010 |
washer | +0.010 |
uit | +0.009 |
Neg logit contributions | |
breath | -0.012 |
common | -0.010 |
job | -0.009 |
sleep | -0.009 |
care | -0.009 |
Token | picked |
Token ID | 6497 |
Feature activation | +6.82e+00 |
Pos logit contributions | |
boards | +0.018 |
uit | +0.017 |
auc | +0.017 |
Did | +0.017 |
cher | +0.017 |
Neg logit contributions | |
breath | -0.015 |
common | -0.012 |
sleep | -0.011 |
long | -0.011 |
job | -0.011 |
Token | up |
Token ID | 510 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
Did | +0.170 |
cher | +0.168 |
ode | +0.163 |
boards | +0.163 |
uit | +0.162 |
Neg logit contributions | |
breath | -0.153 |
long | -0.121 |
care | -0.121 |
home | -0.116 |
job | -0.115 |
Token | the |
Token ID | 262 |
Feature activation | +2.89e+00 |
Pos logit contributions | |
breath | +0.045 |
common | +0.034 |
long | +0.034 |
care | +0.032 |
job | +0.032 |
Neg logit contributions | |
boards | -0.061 |
cher | -0.061 |
Did | -0.060 |
auc | -0.060 |
uit | -0.059 |
Token | big |
Token ID | 1263 |
Feature activation | -2.44e+00 |
Pos logit contributions | |
boards | +0.070 |
Did | +0.068 |
uit | +0.068 |
auc | +0.067 |
cher | +0.067 |
Neg logit contributions | |
breath | -0.065 |
common | -0.052 |
long | -0.049 |
job | -0.049 |
sleep | -0.049 |
Token | pan |
Token ID | 3425 |
Feature activation | +1.89e+00 |
Pos logit contributions | |
breath | +0.057 |
common | +0.050 |
sleep | +0.048 |
job | +0.045 |
uth | +0.044 |
Neg logit contributions | |
boards | -0.053 |
auc | -0.051 |
washer | -0.050 |
register | -0.050 |
Did | -0.049 |
Token | and |
Token ID | 290 |
Feature activation | -3.38e-01 |
Pos logit contributions | |
Did | +0.063 |
cher | +0.060 |
Which | +0.060 |
boards | +0.059 |
ains | +0.059 |
Neg logit contributions | |
breath | -0.028 |
home | -0.018 |
sleep | -0.018 |
happily | -0.017 |
long | -0.017 |
Token | the |
Token ID | 262 |
Feature activation | +2.01e+00 |
Pos logit contributions | |
breath | +0.008 |
common | +0.006 |
long | +0.006 |
sleep | +0.006 |
care | +0.006 |
Neg logit contributions | |
boards | -0.008 |
uit | -0.008 |
auc | -0.007 |
Did | -0.007 |
cher | -0.007 |
Token | to |
Token ID | 284 |
Feature activation | -1.22e+00 |
Pos logit contributions | |
boards | +0.058 |
Did | +0.057 |
uit | +0.057 |
auc | +0.056 |
cher | +0.056 |
Neg logit contributions | |
breath | -0.045 |
common | -0.035 |
long | -0.034 |
sleep | -0.033 |
care | -0.033 |
Token | play |
Token ID | 711 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
breath | +0.029 |
common | +0.023 |
long | +0.022 |
job | +0.022 |
sleep | +0.021 |
Neg logit contributions | |
uit | -0.026 |
boards | -0.026 |
Did | -0.026 |
auc | -0.025 |
cher | -0.025 |
Token | with |
Token ID | 351 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
boards | +0.034 |
register | +0.033 |
auc | +0.033 |
uit | +0.033 |
Did | +0.033 |
Neg logit contributions | |
breath | -0.022 |
sleep | -0.017 |
common | -0.017 |
job | -0.017 |
long | -0.017 |
Token | her |
Token ID | 607 |
Feature activation | +8.30e-01 |
Pos logit contributions | |
boards | +0.050 |
Did | +0.049 |
cher | +0.049 |
uit | +0.048 |
auc | +0.048 |
Neg logit contributions | |
breath | -0.043 |
common | -0.035 |
long | -0.033 |
sleep | -0.033 |
care | -0.032 |
Token | blocks |
Token ID | 7021 |
Feature activation | +5.99e-01 |
Pos logit contributions | |
boards | +0.019 |
Did | +0.019 |
uit | +0.019 |
washer | +0.019 |
cher | +0.019 |
Neg logit contributions | |
breath | -0.019 |
common | -0.015 |
care | -0.014 |
sleep | -0.014 |
long | -0.014 |
Token | again |
Token ID | 757 |
Feature activation | -6.43e+00 |
Pos logit contributions | |
boards | +0.017 |
Did | +0.017 |
auc | +0.017 |
uit | +0.017 |
cher | +0.017 |
Neg logit contributions | |
breath | -0.010 |
common | -0.007 |
sleep | -0.007 |
long | -0.007 |
job | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | +1.37e+00 |
Pos logit contributions | |
breath | +0.146 |
common | +0.109 |
job | +0.104 |
excitement | +0.104 |
breaths | +0.103 |
Neg logit contributions | |
Did | -0.150 |
boards | -0.145 |
uit | -0.144 |
ting | -0.141 |
cher | -0.140 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -6.91e-01 |
Pos logit contributions | |
boards | +0.033 |
Did | +0.033 |
auc | +0.032 |
uit | +0.032 |
cher | +0.031 |
Neg logit contributions | |
breath | -0.032 |
common | -0.024 |
job | -0.023 |
sleep | -0.022 |
care | -0.022 |
Token | Tom |
Token ID | 13787 |
Feature activation | +6.70e-01 |
Pos logit contributions | |
breath | +0.015 |
common | +0.012 |
sleep | +0.012 |
long | +0.011 |
time | +0.011 |
Neg logit contributions | |
boards | -0.016 |
uit | -0.016 |
cher | -0.016 |
auc | -0.016 |
Did | -0.016 |
Token | and |
Token ID | 290 |
Feature activation | -2.56e+00 |
Pos logit contributions | |
boards | +0.011 |
cher | +0.011 |
Did | +0.011 |
auc | +0.011 |
uit | +0.010 |
Neg logit contributions | |
breath | -0.020 |
common | -0.017 |
sleep | -0.017 |
care | -0.016 |
job | -0.016 |
Token | Mia |
Token ID | 32189 |
Feature activation | -1.85e+00 |
Pos logit contributions | |
breath | +0.051 |
common | +0.040 |
sleep | +0.038 |
long | +0.038 |
care | +0.038 |
Neg logit contributions | |
boards | -0.070 |
uit | -0.067 |
auc | -0.067 |
Did | -0.067 |
cher | -0.066 |
Token | . |
Token ID | 13 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
breath | +0.067 |
common | +0.054 |
job | +0.050 |
sleep | +0.048 |
excitement | +0.048 |
Neg logit contributions | |
boards | -0.074 |
cher | -0.073 |
Did | -0.072 |
washer | -0.072 |
uit | -0.071 |
Token | We |
Token ID | 775 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
boards | +0.040 |
uit | +0.039 |
auc | +0.039 |
Did | +0.039 |
cher | +0.038 |
Neg logit contributions | |
breath | -0.040 |
common | -0.033 |
sleep | -0.031 |
job | -0.031 |
long | -0.031 |
Token | will |
Token ID | 481 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
boards | +0.016 |
uit | +0.016 |
Did | +0.016 |
cher | +0.016 |
auc | +0.015 |
Neg logit contributions | |
breath | -0.035 |
common | -0.030 |
long | -0.029 |
job | -0.028 |
sleep | -0.028 |
Token | not |
Token ID | 407 |
Feature activation | -2.49e+00 |
Pos logit contributions | |
breath | +0.039 |
common | +0.031 |
long | +0.029 |
job | +0.029 |
breaths | +0.029 |
Neg logit contributions | |
uit | -0.034 |
Did | -0.032 |
boards | -0.032 |
cher | -0.032 |
register | -0.031 |
Token | fight |
Token ID | 1907 |
Feature activation | +1.58e+00 |
Pos logit contributions | |
breath | +0.057 |
common | +0.046 |
accomplishment | +0.043 |
content | +0.042 |
breaths | +0.042 |
Neg logit contributions | |
Did | -0.053 |
washer | -0.053 |
uit | -0.052 |
ode | -0.052 |
boards | -0.052 |
Token | again |
Token ID | 757 |
Feature activation | -6.59e+00 |
Pos logit contributions | |
boards | +0.034 |
Did | +0.033 |
uit | +0.032 |
auc | +0.032 |
cher | +0.032 |
Neg logit contributions | |
breath | -0.040 |
common | -0.032 |
sleep | -0.030 |
job | -0.030 |
long | -0.030 |
Token | ." |
Token ID | 526 |
Feature activation | -1.94e+00 |
Pos logit contributions | |
breath | +0.134 |
common | +0.095 |
job | +0.089 |
excitement | +0.089 |
sleep | +0.088 |
Neg logit contributions | |
Did | -0.172 |
boards | -0.168 |
ting | -0.166 |
ains | -0.164 |
cher | -0.163 |
Token |
|
Token ID | 198 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
breath | +0.055 |
job | +0.045 |
accomplishment | +0.042 |
common | +0.042 |
idea | +0.041 |
Neg logit contributions | |
Did | -0.038 |
boards | -0.037 |
auc | -0.034 |
register | -0.033 |
uit | -0.033 |
Token |
|
Token ID | 198 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
breath | +0.054 |
job | +0.044 |
breaths | +0.041 |
common | +0.041 |
care | +0.040 |
Neg logit contributions | |
Did | -0.046 |
uit | -0.043 |
boards | -0.043 |
washer | -0.042 |
auc | -0.042 |
Token | Mom |
Token ID | 29252 |
Feature activation | -6.80e-01 |
Pos logit contributions | |
breath | +0.059 |
job | +0.048 |
common | +0.048 |
sleep | +0.047 |
idea | +0.046 |
Neg logit contributions | |
Did | -0.037 |
boards | -0.035 |
uit | -0.034 |
auc | -0.034 |
washer | -0.033 |
Token | and |
Token ID | 290 |
Feature activation | -2.60e+00 |
Pos logit contributions | |
breath | +0.011 |
sleep | +0.007 |
work | +0.007 |
care | +0.006 |
happily | +0.006 |
Neg logit contributions | |
Did | -0.021 |
cher | -0.021 |
boards | -0.020 |
auc | -0.020 |
Which | -0.020 |
Token | not |
Token ID | 407 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
breath | +0.071 |
long | +0.055 |
job | +0.053 |
common | +0.053 |
sleep | +0.052 |
Neg logit contributions | |
boards | -0.096 |
Did | -0.094 |
auc | -0.094 |
uit | -0.092 |
cher | -0.091 |
Token | to |
Token ID | 284 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
breath | +0.024 |
common | +0.018 |
long | +0.018 |
care | +0.018 |
job | +0.018 |
Neg logit contributions | |
boards | -0.035 |
auc | -0.035 |
Did | -0.034 |
uit | -0.034 |
cher | -0.034 |
Token | do |
Token ID | 466 |
Feature activation | +8.58e-01 |
Pos logit contributions | |
breath | +0.037 |
common | +0.030 |
job | +0.028 |
long | +0.027 |
sleep | +0.027 |
Neg logit contributions | |
boards | -0.036 |
uit | -0.036 |
Did | -0.035 |
auc | -0.034 |
cher | -0.034 |
Token | naughty |
Token ID | 45128 |
Feature activation | -3.57e+00 |
Pos logit contributions | |
boards | +0.018 |
Did | +0.017 |
auc | +0.017 |
uit | +0.017 |
cher | +0.017 |
Neg logit contributions | |
breath | -0.022 |
common | -0.018 |
long | -0.018 |
job | -0.018 |
sleep | -0.018 |
Token | things |
Token ID | 1243 |
Feature activation | +9.52e-01 |
Pos logit contributions | |
breath | +0.069 |
common | +0.053 |
sleep | +0.049 |
job | +0.049 |
long | +0.049 |
Neg logit contributions | |
boards | -0.097 |
Did | -0.094 |
auc | -0.093 |
uit | -0.093 |
cher | -0.092 |
Token | again |
Token ID | 757 |
Feature activation | -7.39e+00 |
Pos logit contributions | |
Did | +0.028 |
auc | +0.028 |
boards | +0.027 |
ains | +0.027 |
uit | +0.026 |
Neg logit contributions | |
breath | -0.016 |
care | -0.012 |
job | -0.012 |
long | -0.012 |
work | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | +7.81e-01 |
Pos logit contributions | |
breath | +0.128 |
common | +0.096 |
long | +0.089 |
job | +0.089 |
sleep | +0.089 |
Neg logit contributions | |
boards | -0.218 |
auc | -0.212 |
Did | -0.211 |
uit | -0.210 |
cher | -0.209 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.64e-01 |
Pos logit contributions | |
boards | +0.020 |
Did | +0.020 |
uit | +0.020 |
auc | +0.020 |
washer | +0.019 |
Neg logit contributions | |
breath | -0.016 |
common | -0.012 |
job | -0.012 |
long | -0.011 |
sleep | -0.011 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
breath | +0.005 |
common | +0.004 |
sleep | +0.003 |
long | +0.003 |
time | +0.003 |
Neg logit contributions | |
boards | -0.007 |
auc | -0.007 |
cher | -0.007 |
Did | -0.007 |
register | -0.007 |
Token | there |
Token ID | 612 |
Feature activation | +7.63e-01 |
Pos logit contributions | |
breath | +0.031 |
common | +0.026 |
breaths | +0.026 |
job | +0.026 |
care | +0.025 |
Neg logit contributions | |
washer | -0.021 |
boards | -0.019 |
auc | -0.019 |
uit | -0.019 |
lying | -0.019 |
Token | was |
Token ID | 373 |
Feature activation | -1.53e-01 |
Pos logit contributions | |
boards | +0.012 |
Did | +0.012 |
auc | +0.012 |
uit | +0.012 |
cher | +0.011 |
Neg logit contributions | |
breath | -0.023 |
common | -0.020 |
sleep | -0.019 |
job | -0.019 |
long | -0.019 |
Token | red |
Token ID | 2266 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
breath | +0.004 |
common | +0.003 |
care | +0.003 |
sleep | +0.003 |
breaths | +0.003 |
Neg logit contributions | |
washer | -0.004 |
boards | -0.004 |
Did | -0.004 |
auc | -0.004 |
uit | -0.004 |
Token | sweater |
Token ID | 35685 |
Feature activation | -7.26e-02 |
Pos logit contributions | |
breath | +0.039 |
common | +0.032 |
sleep | +0.030 |
job | +0.030 |
long | +0.030 |
Neg logit contributions | |
boards | -0.034 |
Did | -0.033 |
uit | -0.033 |
auc | -0.033 |
cher | -0.032 |
Token | that |
Token ID | 326 |
Feature activation | +4.45e+00 |
Pos logit contributions | |
breath | +0.001 |
long | +0.001 |
job | +0.001 |
sleep | +0.001 |
day | +0.001 |
Neg logit contributions | |
ains | -0.002 |
Did | -0.002 |
auc | -0.002 |
ode | -0.002 |
cher | -0.002 |
Token | he |
Token ID | 339 |
Feature activation | +7.45e-01 |
Pos logit contributions | |
register | +0.117 |
ains | +0.114 |
boards | +0.113 |
Did | +0.113 |
auc | +0.112 |
Neg logit contributions | |
breath | -0.088 |
long | -0.079 |
sleep | -0.075 |
common | -0.073 |
job | -0.072 |
Token | loved |
Token ID | 6151 |
Feature activation | -3.53e-01 |
Pos logit contributions | |
boards | +0.016 |
auc | +0.016 |
uit | +0.016 |
Did | +0.016 |
register | +0.016 |
Neg logit contributions | |
breath | -0.018 |
common | -0.015 |
sleep | -0.014 |
long | -0.014 |
job | -0.014 |
Token | very |
Token ID | 845 |
Feature activation | -6.40e+00 |
Pos logit contributions | |
breath | +0.007 |
common | +0.005 |
long | +0.005 |
sleep | +0.005 |
job | +0.005 |
Neg logit contributions | |
boards | -0.009 |
auc | -0.009 |
Did | -0.009 |
uit | -0.009 |
cher | -0.009 |
Token | much |
Token ID | 881 |
Feature activation | -5.20e+00 |
Pos logit contributions | |
breath | +0.162 |
breaths | +0.123 |
common | +0.121 |
excitement | +0.117 |
sleep | +0.116 |
Neg logit contributions | |
boards | -0.134 |
washer | -0.132 |
auc | -0.126 |
Did | -0.116 |
lying | -0.116 |
Token | . |
Token ID | 13 |
Feature activation | -7.68e-02 |
Pos logit contributions | |
breath | +0.098 |
common | +0.074 |
job | +0.067 |
sleep | +0.067 |
long | +0.066 |
Neg logit contributions | |
boards | -0.145 |
Did | -0.140 |
uit | -0.140 |
auc | -0.139 |
cher | -0.138 |
Token | It |
Token ID | 632 |
Feature activation | +3.21e+00 |
Pos logit contributions | |
breath | +0.002 |
common | +0.001 |
sleep | +0.001 |
long | +0.001 |
job | +0.001 |
Neg logit contributions | |
boards | -0.002 |
Did | -0.002 |
auc | -0.002 |
cher | -0.002 |
uit | -0.002 |
Token | was |
Token ID | 373 |
Feature activation | +2.62e+00 |
Pos logit contributions | |
register | +0.075 |
boards | +0.073 |
cher | +0.069 |
ains | +0.069 |
uit | +0.069 |
Neg logit contributions | |
breath | -0.075 |
sleep | -0.067 |
long | -0.066 |
care | -0.063 |
common | -0.063 |
Token | an |
Token ID | 281 |
Feature activation | +5.16e-01 |
Pos logit contributions | |
boards | +0.073 |
auc | +0.069 |
cher | +0.068 |
register | +0.068 |
Did | +0.068 |
Neg logit contributions | |
breath | -0.053 |
long | -0.043 |
common | -0.043 |
sleep | -0.041 |
job | -0.040 |
Token | but |
Token ID | 475 |
Feature activation | +5.71e-01 |
Pos logit contributions | |
boards | +0.032 |
auc | +0.031 |
cher | +0.031 |
washer | +0.030 |
uit | +0.030 |
Neg logit contributions | |
breath | -0.023 |
long | -0.017 |
sleep | -0.017 |
common | -0.016 |
excitement | -0.016 |
Token | they |
Token ID | 484 |
Feature activation | +6.90e-01 |
Pos logit contributions | |
boards | +0.016 |
uit | +0.015 |
auc | +0.015 |
cher | +0.015 |
Did | +0.015 |
Neg logit contributions | |
breath | -0.012 |
common | -0.009 |
long | -0.009 |
job | -0.008 |
care | -0.008 |
Token | are |
Token ID | 389 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
boards | +0.016 |
Did | +0.015 |
auc | +0.015 |
uit | +0.015 |
cher | +0.015 |
Neg logit contributions | |
breath | -0.017 |
sleep | -0.014 |
common | -0.014 |
care | -0.014 |
work | -0.013 |
Token | not |
Token ID | 407 |
Feature activation | -1.94e+00 |
Pos logit contributions | |
boards | +0.053 |
Did | +0.052 |
auc | +0.051 |
uit | +0.051 |
cher | +0.051 |
Neg logit contributions | |
breath | -0.059 |
common | -0.048 |
job | -0.046 |
sleep | -0.045 |
long | -0.045 |
Token | as |
Token ID | 355 |
Feature activation | -4.31e+00 |
Pos logit contributions | |
breath | +0.043 |
common | +0.035 |
job | +0.033 |
sleep | +0.033 |
� | +0.032 |
Neg logit contributions | |
cher | -0.044 |
lying | -0.043 |
Did | -0.043 |
washer | -0.043 |
boards | -0.042 |
Token | fun |
Token ID | 1257 |
Feature activation | -6.91e+00 |
Pos logit contributions | |
breath | +0.082 |
common | +0.062 |
job | +0.058 |
long | +0.058 |
sleep | +0.058 |
Neg logit contributions | |
boards | -0.120 |
Did | -0.117 |
uit | -0.116 |
auc | -0.116 |
cher | -0.114 |
Token | as |
Token ID | 355 |
Feature activation | -4.37e+00 |
Pos logit contributions | |
breath | +0.143 |
common | +0.118 |
care | +0.105 |
job | +0.104 |
sleep | +0.101 |
Neg logit contributions | |
boards | -0.176 |
ting | -0.171 |
uit | -0.168 |
auc | -0.167 |
register | -0.165 |
Token | the |
Token ID | 262 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
breath | +0.099 |
common | +0.079 |
job | +0.075 |
sleep | +0.074 |
long | +0.073 |
Neg logit contributions | |
boards | -0.104 |
Did | -0.101 |
auc | -0.100 |
uit | -0.100 |
cher | -0.099 |
Token | hose |
Token ID | 31489 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
boards | +0.039 |
Did | +0.038 |
uit | +0.038 |
auc | +0.037 |
cher | +0.037 |
Neg logit contributions | |
breath | -0.038 |
common | -0.030 |
job | -0.029 |
long | -0.029 |
sleep | -0.029 |
Token | . |
Token ID | 13 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
Did | +0.060 |
boards | +0.057 |
uit | +0.057 |
cher | +0.056 |
ode | +0.056 |
Neg logit contributions | |
breath | -0.028 |
common | -0.019 |
sleep | -0.019 |
work | -0.019 |
care | -0.018 |
Token |
|
Token ID | 198 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
boards | +0.029 |
Did | +0.029 |
auc | +0.028 |
uit | +0.028 |
washer | +0.027 |
Neg logit contributions | |
breath | -0.025 |
common | -0.018 |
job | -0.018 |
sleep | -0.017 |
care | -0.017 |