Token | . |
Token ID | 13 |
Feature activation | -2.96e-03 |
Pos logit contributions | |
decision | +0.017 |
choice | +0.016 |
away | +0.016 |
ook | +0.013 |
taxi | +0.013 |
Neg logit contributions | |
sauce | -0.019 |
gum | -0.018 |
greens | -0.018 |
glasses | -0.017 |
ink | -0.017 |
Token | She |
Token ID | 1375 |
Feature activation | -8.47e-03 |
Pos logit contributions | |
sauce | +0.005 |
gum | +0.005 |
ink | +0.004 |
greens | +0.004 |
glasses | +0.004 |
Neg logit contributions | |
decision | -0.005 |
choice | -0.005 |
away | -0.005 |
ook | -0.004 |
doorstep | -0.004 |
Token | grabbed |
Token ID | 13176 |
Feature activation | -4.31e-03 |
Pos logit contributions | |
sauce | +0.014 |
gum | +0.014 |
greens | +0.013 |
triangles | +0.013 |
subtract | +0.013 |
Neg logit contributions | |
decision | -0.014 |
choice | -0.014 |
away | -0.013 |
promise | -0.012 |
sadly | -0.012 |
Token | the |
Token ID | 262 |
Feature activation | +2.71e-03 |
Pos logit contributions | |
sauce | +0.007 |
gum | +0.007 |
greens | +0.007 |
glasses | +0.006 |
pads | +0.006 |
Neg logit contributions | |
decision | -0.007 |
choice | -0.007 |
away | -0.007 |
ook | -0.006 |
taxi | -0.006 |
Token | fire |
Token ID | 2046 |
Feature activation | +6.92e-03 |
Pos logit contributions | |
decision | +0.004 |
choice | +0.004 |
away | +0.004 |
taxi | +0.003 |
doorstep | +0.003 |
Neg logit contributions | |
sauce | -0.005 |
gum | -0.005 |
greens | -0.005 |
pads | -0.004 |
subtract | -0.004 |
Token | extingu |
Token ID | 24995 |
Feature activation | +7.94e-02 |
Pos logit contributions | |
decision | +0.009 |
choice | +0.008 |
away | +0.008 |
taxi | +0.007 |
tha | +0.007 |
Neg logit contributions | |
gum | -0.014 |
sauce | -0.014 |
greens | -0.013 |
subtract | -0.013 |
glasses | -0.013 |
Token | isher |
Token ID | 4828 |
Feature activation | +2.57e-02 |
Pos logit contributions | |
decision | +0.116 |
ook | +0.114 |
checkpoint | +0.109 |
doorstep | +0.106 |
away | +0.103 |
Neg logit contributions | |
gum | -0.128 |
sauce | -0.126 |
subtract | -0.117 |
glasses | -0.115 |
mascara | -0.112 |
Token | and |
Token ID | 290 |
Feature activation | -4.85e-03 |
Pos logit contributions | |
decision | +0.038 |
choice | +0.037 |
away | +0.036 |
ook | +0.032 |
taxi | +0.030 |
Neg logit contributions | |
sauce | -0.047 |
gum | -0.046 |
greens | -0.044 |
glasses | -0.043 |
ink | -0.042 |
Token | put |
Token ID | 1234 |
Feature activation | -3.66e-03 |
Pos logit contributions | |
sauce | +0.007 |
gum | +0.007 |
greens | +0.007 |
triangles | +0.007 |
fiber | +0.007 |
Neg logit contributions | |
decision | -0.008 |
away | -0.008 |
choice | -0.008 |
taxi | -0.007 |
tha | -0.007 |
Token | out |
Token ID | 503 |
Feature activation | -4.84e-03 |
Pos logit contributions | |
sauce | +0.008 |
gum | +0.007 |
greens | +0.007 |
subtract | +0.007 |
pads | +0.007 |
Neg logit contributions | |
decision | -0.004 |
choice | -0.004 |
away | -0.004 |
forward | -0.004 |
taxi | -0.004 |
Token | the |
Token ID | 262 |
Feature activation | +3.35e-03 |
Pos logit contributions | |
sauce | +0.008 |
gum | +0.008 |
greens | +0.008 |
glasses | +0.007 |
subtract | +0.007 |
Neg logit contributions | |
decision | -0.008 |
choice | -0.008 |
away | -0.008 |
ook | -0.007 |
taxi | -0.007 |
Token | quickly |
Token ID | 2952 |
Feature activation | +2.32e-03 |
Pos logit contributions | |
decision | +0.012 |
choice | +0.011 |
away | +0.011 |
sadly | +0.010 |
exactly | +0.010 |
Neg logit contributions | |
gum | -0.013 |
sauce | -0.012 |
greens | -0.012 |
triangles | -0.012 |
squares | -0.012 |
Token | checked |
Token ID | 10667 |
Feature activation | +2.25e-02 |
Pos logit contributions | |
away | +0.004 |
decision | +0.004 |
choice | +0.003 |
taxi | +0.003 |
ook | +0.003 |
Neg logit contributions | |
gum | -0.004 |
triangles | -0.004 |
greens | -0.004 |
squares | -0.004 |
subtract | -0.004 |
Token | for |
Token ID | 329 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
decision | +0.044 |
choice | +0.041 |
away | +0.039 |
doorstep | +0.038 |
checkpoint | +0.038 |
Neg logit contributions | |
sauce | -0.031 |
gum | -0.030 |
ink | -0.026 |
glasses | -0.026 |
triangles | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | +9.93e-03 |
Pos logit contributions | |
sauce | +0.024 |
gum | +0.024 |
greens | +0.023 |
pads | +0.023 |
gloss | +0.023 |
Neg logit contributions | |
decision | -0.023 |
choice | -0.022 |
away | -0.021 |
taxi | -0.018 |
ook | -0.018 |
Token | fire |
Token ID | 2046 |
Feature activation | +1.84e-03 |
Pos logit contributions | |
decision | +0.016 |
choice | +0.015 |
away | +0.015 |
taxi | +0.013 |
ook | +0.013 |
Neg logit contributions | |
sauce | -0.017 |
gum | -0.017 |
greens | -0.016 |
glasses | -0.016 |
subtract | -0.015 |
Token | extingu |
Token ID | 24995 |
Feature activation | +7.60e-02 |
Pos logit contributions | |
decision | +0.002 |
choice | +0.002 |
away | +0.002 |
taxi | +0.002 |
doorstep | +0.002 |
Neg logit contributions | |
greens | -0.004 |
gum | -0.004 |
sauce | -0.004 |
subtract | -0.004 |
triangles | -0.003 |
Token | isher |
Token ID | 4828 |
Feature activation | +1.65e-02 |
Pos logit contributions | |
decision | +0.116 |
ook | +0.114 |
azel | +0.112 |
checkpoint | +0.108 |
doorstep | +0.108 |
Neg logit contributions | |
sauce | -0.115 |
gum | -0.113 |
glasses | -0.096 |
ink | -0.095 |
cr | -0.092 |
Token | , |
Token ID | 11 |
Feature activation | -4.20e-03 |
Pos logit contributions | |
decision | +0.022 |
choice | +0.021 |
away | +0.021 |
taxi | +0.018 |
ook | +0.017 |
Neg logit contributions | |
greens | -0.032 |
sauce | -0.032 |
gum | -0.031 |
subtract | -0.030 |
triangles | -0.030 |
Token | but |
Token ID | 475 |
Feature activation | +4.51e-02 |
Pos logit contributions | |
sauce | +0.009 |
fiber | +0.008 |
gloss | +0.008 |
gum | +0.008 |
greens | +0.008 |
Neg logit contributions | |
decision | -0.005 |
choice | -0.004 |
away | -0.004 |
exactly | -0.004 |
sadly | -0.004 |
Token | there |
Token ID | 612 |
Feature activation | +5.81e-03 |
Pos logit contributions | |
decision | +0.068 |
choice | +0.064 |
away | +0.063 |
ook | +0.055 |
taxi | +0.055 |
Neg logit contributions | |
sauce | -0.081 |
greens | -0.079 |
gum | -0.079 |
glasses | -0.076 |
subtract | -0.074 |
Token | wasn |
Token ID | 2492 |
Feature activation | +4.37e-03 |
Pos logit contributions | |
decision | +0.008 |
choice | +0.008 |
away | +0.007 |
ook | +0.007 |
taxi | +0.006 |
Neg logit contributions | |
greens | -0.011 |
sauce | -0.011 |
gum | -0.011 |
subtract | -0.010 |
pads | -0.010 |
Token | there |
Token ID | 612 |
Feature activation | +7.06e-04 |
Pos logit contributions | |
decision | +0.007 |
choice | +0.006 |
away | +0.006 |
doorstep | +0.005 |
checkpoint | +0.005 |
Neg logit contributions | |
sauce | -0.005 |
gum | -0.005 |
greens | -0.005 |
triangles | -0.004 |
ink | -0.004 |
Token | was |
Token ID | 373 |
Feature activation | -1.29e-02 |
Pos logit contributions | |
decision | +0.001 |
choice | +0.001 |
away | +0.001 |
ook | +0.001 |
taxi | +0.001 |
Neg logit contributions | |
sauce | -0.001 |
gum | -0.001 |
greens | -0.001 |
glasses | -0.001 |
ink | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | -3.24e-03 |
Pos logit contributions | |
sauce | +0.018 |
greens | +0.017 |
subtract | +0.017 |
gum | +0.017 |
pads | +0.016 |
Neg logit contributions | |
decision | -0.023 |
choice | -0.022 |
away | -0.022 |
taxi | -0.020 |
ook | -0.020 |
Token | girl |
Token ID | 2576 |
Feature activation | +8.27e-04 |
Pos logit contributions | |
sauce | +0.005 |
gum | +0.005 |
greens | +0.005 |
ink | +0.005 |
glasses | +0.005 |
Neg logit contributions | |
decision | -0.005 |
choice | -0.005 |
away | -0.005 |
ook | -0.005 |
ively | -0.005 |
Token | named |
Token ID | 3706 |
Feature activation | -1.95e-02 |
Pos logit contributions | |
decision | +0.002 |
choice | +0.002 |
away | +0.001 |
destination | +0.001 |
ently | +0.001 |
Neg logit contributions | |
sauce | -0.001 |
gum | -0.001 |
ink | -0.001 |
glasses | -0.001 |
greens | -0.001 |
Token | Far |
Token ID | 6755 |
Feature activation | +7.53e-02 |
Pos logit contributions | |
gum | +0.029 |
sauce | +0.029 |
greens | +0.028 |
subtract | +0.027 |
pads | +0.026 |
Neg logit contributions | |
decision | -0.033 |
away | -0.033 |
choice | -0.032 |
ook | -0.030 |
taxi | -0.029 |
Token | ah |
Token ID | 993 |
Feature activation | +1.34e-02 |
Pos logit contributions | |
decision | +0.108 |
choice | +0.093 |
doorstep | +0.092 |
iring | +0.088 |
ook | +0.088 |
Neg logit contributions | |
pads | -0.124 |
gum | -0.123 |
glasses | -0.122 |
Ol | -0.118 |
ink | -0.116 |
Token | . |
Token ID | 13 |
Feature activation | -3.62e-03 |
Pos logit contributions | |
decision | +0.021 |
choice | +0.020 |
away | +0.020 |
ook | +0.017 |
taxi | +0.017 |
Neg logit contributions | |
sauce | -0.023 |
gum | -0.023 |
greens | -0.022 |
glasses | -0.021 |
ink | -0.021 |
Token | She |
Token ID | 1375 |
Feature activation | +1.53e-02 |
Pos logit contributions | |
gum | +0.004 |
sauce | +0.004 |
greens | +0.004 |
glasses | +0.004 |
pads | +0.004 |
Neg logit contributions | |
decision | -0.008 |
choice | -0.007 |
away | -0.007 |
ook | -0.006 |
checkpoint | -0.006 |
Token | liked |
Token ID | 8288 |
Feature activation | -1.01e-02 |
Pos logit contributions | |
decision | +0.024 |
choice | +0.023 |
away | +0.023 |
ook | +0.021 |
taxi | +0.020 |
Neg logit contributions | |
sauce | -0.025 |
greens | -0.025 |
gum | -0.025 |
subtract | -0.024 |
triangles | -0.024 |
Token | playing |
Token ID | 2712 |
Feature activation | -1.74e-02 |
Pos logit contributions | |
greens | +0.022 |
Ol | +0.022 |
fiber | +0.022 |
pads | +0.022 |
gloss | +0.022 |
Neg logit contributions | |
away | -0.007 |
taxi | -0.007 |
home | -0.007 |
choice | -0.006 |
decision | -0.006 |
Token | her |
Token ID | 607 |
Feature activation | +2.78e-02 |
Pos logit contributions | |
gum | +0.010 |
greens | +0.009 |
Ol | +0.009 |
sauce | +0.009 |
lenses | +0.009 |
Neg logit contributions | |
away | -0.008 |
choice | -0.007 |
decision | -0.007 |
forward | -0.006 |
ively | -0.006 |
Token | mom |
Token ID | 1995 |
Feature activation | +2.00e-03 |
Pos logit contributions | |
decision | +0.041 |
choice | +0.040 |
away | +0.035 |
taxi | +0.032 |
doorstep | +0.030 |
Neg logit contributions | |
sauce | -0.053 |
gum | -0.051 |
greens | -0.051 |
subtract | -0.050 |
Ol | -0.048 |
Token | my |
Token ID | 1820 |
Feature activation | +4.60e-03 |
Pos logit contributions | |
decision | +0.003 |
choice | +0.003 |
away | +0.002 |
promise | +0.002 |
taxi | +0.002 |
Neg logit contributions | |
gum | -0.004 |
sauce | -0.004 |
greens | -0.003 |
triangles | -0.003 |
fiber | -0.003 |
Token | watched |
Token ID | 7342 |
Feature activation | +1.99e-02 |
Pos logit contributions | |
decision | +0.006 |
choice | +0.006 |
away | +0.006 |
taxi | +0.005 |
promise | +0.005 |
Neg logit contributions | |
gum | -0.008 |
sauce | -0.008 |
triangles | -0.008 |
greens | -0.008 |
squares | -0.008 |
Token | in |
Token ID | 287 |
Feature activation | +2.32e-03 |
Pos logit contributions | |
decision | +0.025 |
away | +0.024 |
choice | +0.024 |
forward | +0.019 |
taxi | +0.019 |
Neg logit contributions | |
sauce | -0.039 |
gum | -0.039 |
pads | -0.039 |
greens | -0.038 |
Ol | -0.038 |
Token | amaz |
Token ID | 40642 |
Feature activation | +7.47e-02 |
Pos logit contributions | |
decision | +0.004 |
choice | +0.003 |
away | +0.003 |
taxi | +0.003 |
purchase | +0.003 |
Neg logit contributions | |
sauce | -0.004 |
gum | -0.004 |
pads | -0.004 |
greens | -0.004 |
gloss | -0.004 |
Token | ement |
Token ID | 972 |
Feature activation | +3.31e-02 |
Pos logit contributions | |
decision | +0.077 |
choice | +0.068 |
away | +0.066 |
ook | +0.057 |
checkpoint | +0.055 |
Neg logit contributions | |
gum | -0.169 |
sauce | -0.168 |
greens | -0.159 |
triangles | -0.159 |
subtract | -0.157 |
Token | as |
Token ID | 355 |
Feature activation | +1.58e-02 |
Pos logit contributions | |
decision | +0.042 |
choice | +0.039 |
away | +0.039 |
ook | +0.032 |
taxi | +0.031 |
Neg logit contributions | |
sauce | -0.068 |
gum | -0.067 |
greens | -0.064 |
Ol | -0.064 |
glasses | -0.063 |
Token | the |
Token ID | 262 |
Feature activation | +6.69e-03 |
Pos logit contributions | |
decision | +0.023 |
choice | +0.023 |
away | +0.021 |
exactly | +0.019 |
taxi | +0.019 |
Neg logit contributions | |
sauce | -0.028 |
gum | -0.028 |
greens | -0.027 |
pads | -0.027 |
Ol | -0.026 |
Token | machine |
Token ID | 4572 |
Feature activation | -1.59e-03 |
Pos logit contributions | |
decision | +0.011 |
choice | +0.011 |
taxi | +0.009 |
away | +0.009 |
hotel | +0.008 |
Neg logit contributions | |
pads | -0.012 |
sauce | -0.011 |
Ol | -0.011 |
gum | -0.011 |
greens | -0.011 |
Token | flew |
Token ID | 13112 |
Feature activation | +7.96e-03 |
Pos logit contributions | |
gum | +0.003 |
sauce | +0.003 |
fiber | +0.003 |
Ol | +0.003 |
greens | +0.003 |
Neg logit contributions | |
decision | -0.002 |
choice | -0.002 |
away | -0.002 |
taxi | -0.002 |
destination | -0.002 |
Token | ." |
Token ID | 526 |
Feature activation | -5.98e-03 |
Pos logit contributions | |
decision | +0.009 |
away | +0.009 |
choice | +0.008 |
ook | +0.007 |
taxi | +0.007 |
Neg logit contributions | |
sauce | -0.013 |
gum | -0.013 |
greens | -0.013 |
glasses | -0.012 |
subtract | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | +3.51e-03 |
Pos logit contributions | |
sauce | +0.009 |
gum | +0.009 |
triangles | +0.009 |
greens | +0.009 |
glasses | +0.009 |
Neg logit contributions | |
decision | -0.010 |
choice | -0.010 |
away | -0.009 |
ook | -0.008 |
taxi | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | +3.76e-03 |
Pos logit contributions | |
decision | +0.007 |
choice | +0.007 |
away | +0.007 |
ook | +0.006 |
taxi | +0.006 |
Neg logit contributions | |
sauce | -0.004 |
gum | -0.004 |
greens | -0.004 |
glasses | -0.004 |
ink | -0.004 |
Token | Billy |
Token ID | 49640 |
Feature activation | +1.47e-02 |
Pos logit contributions | |
decision | +0.009 |
choice | +0.008 |
away | +0.008 |
ook | +0.008 |
taxi | +0.007 |
Neg logit contributions | |
sauce | -0.004 |
gum | -0.004 |
greens | -0.004 |
glasses | -0.003 |
subtract | -0.003 |
Token | was |
Token ID | 373 |
Feature activation | +4.84e-03 |
Pos logit contributions | |
decision | +0.021 |
choice | +0.020 |
away | +0.019 |
ook | +0.018 |
sadly | +0.017 |
Neg logit contributions | |
sauce | -0.026 |
gum | -0.026 |
greens | -0.026 |
ink | -0.025 |
triangles | -0.025 |
Token | ast |
Token ID | 6468 |
Feature activation | +7.26e-02 |
Pos logit contributions | |
decision | +0.007 |
choice | +0.007 |
away | +0.006 |
ook | +0.006 |
forward | +0.005 |
Neg logit contributions | |
sauce | -0.009 |
gum | -0.009 |
greens | -0.009 |
ink | -0.009 |
glasses | -0.008 |
Token | ounded |
Token ID | 6302 |
Feature activation | +4.72e-02 |
Pos logit contributions | |
decision | +0.086 |
choice | +0.084 |
away | +0.071 |
ator | +0.069 |
hotel | +0.067 |
Neg logit contributions | |
gum | -0.147 |
sauce | -0.143 |
pads | -0.139 |
gloss | -0.138 |
subtract | -0.137 |
Token | ! |
Token ID | 0 |
Feature activation | +3.18e-03 |
Pos logit contributions | |
decision | +0.077 |
choice | +0.074 |
away | +0.072 |
ook | +0.064 |
taxi | +0.063 |
Neg logit contributions | |
sauce | -0.079 |
gum | -0.079 |
greens | -0.074 |
glasses | -0.073 |
ink | -0.072 |
Token | He |
Token ID | 679 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
decision | +0.005 |
choice | +0.005 |
away | +0.005 |
ook | +0.004 |
forward | +0.004 |
Neg logit contributions | |
sauce | -0.005 |
greens | -0.005 |
gum | -0.005 |
glasses | -0.005 |
triangles | -0.005 |
Token | used |
Token ID | 973 |
Feature activation | -1.52e-02 |
Pos logit contributions | |
decision | +0.017 |
choice | +0.016 |
away | +0.016 |
ook | +0.015 |
exactly | +0.015 |
Neg logit contributions | |
sauce | -0.020 |
gum | -0.020 |
greens | -0.019 |
triangles | -0.018 |
glasses | -0.018 |
Token | the |
Token ID | 262 |
Feature activation | +4.02e-03 |
Pos logit contributions | |
gloss | +0.031 |
pads | +0.030 |
sauce | +0.030 |
greens | +0.029 |
gum | +0.029 |
Neg logit contributions | |
decision | -0.019 |
choice | -0.019 |
away | -0.017 |
purchase | -0.015 |
exactly | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | -3.23e-04 |
Pos logit contributions | |
away | +0.022 |
decision | +0.022 |
choice | +0.022 |
tha | +0.020 |
ook | +0.019 |
Neg logit contributions | |
sauce | -0.024 |
gum | -0.024 |
ink | -0.022 |
greens | -0.022 |
glasses | -0.022 |
Token | He |
Token ID | 679 |
Feature activation | +1.61e-02 |
Pos logit contributions | |
sauce | +0.001 |
gum | +0.001 |
greens | +0.000 |
glasses | +0.000 |
pads | +0.000 |
Neg logit contributions | |
decision | -0.001 |
choice | -0.001 |
away | -0.001 |
ook | -0.000 |
checkpoint | -0.000 |
Token | was |
Token ID | 373 |
Feature activation | +2.39e-03 |
Pos logit contributions | |
decision | +0.024 |
choice | +0.024 |
away | +0.023 |
promise | +0.021 |
taxi | +0.021 |
Neg logit contributions | |
gum | -0.026 |
sauce | -0.025 |
greens | -0.025 |
pads | -0.024 |
triangles | -0.024 |
Token | a |
Token ID | 257 |
Feature activation | +1.69e-02 |
Pos logit contributions | |
decision | +0.003 |
choice | +0.003 |
away | +0.003 |
ook | +0.003 |
forward | +0.003 |
Neg logit contributions | |
sauce | -0.005 |
gum | -0.005 |
greens | -0.005 |
pads | -0.004 |
ink | -0.004 |
Token | successful |
Token ID | 4388 |
Feature activation | +1.11e-03 |
Pos logit contributions | |
decision | +0.029 |
choice | +0.027 |
away | +0.027 |
ook | +0.025 |
ively | +0.024 |
Neg logit contributions | |
sauce | -0.028 |
gum | -0.028 |
greens | -0.025 |
glasses | -0.025 |
ink | -0.025 |
Token | clim |
Token ID | 5424 |
Feature activation | +7.22e-02 |
Pos logit contributions | |
decision | +0.002 |
choice | +0.001 |
away | +0.001 |
ook | +0.001 |
taxi | +0.001 |
Neg logit contributions | |
sauce | -0.002 |
gum | -0.002 |
greens | -0.002 |
glasses | -0.002 |
ink | -0.002 |
Token | ber |
Token ID | 527 |
Feature activation | +8.62e-03 |
Pos logit contributions | |
decision | +0.112 |
choice | +0.101 |
ently | +0.092 |
doorstep | +0.091 |
destination | +0.089 |
Neg logit contributions | |
cr | -0.105 |
gloss | -0.103 |
gum | -0.102 |
sauce | -0.102 |
elt | -0.101 |
Token | ! |
Token ID | 0 |
Feature activation | +4.62e-03 |
Pos logit contributions | |
decision | +0.012 |
choice | +0.011 |
away | +0.011 |
taxi | +0.009 |
ook | +0.009 |
Neg logit contributions | |
sauce | -0.017 |
gum | -0.016 |
greens | -0.016 |
ink | -0.015 |
glasses | -0.015 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.89e-02 |
Pos logit contributions | |
decision | +0.008 |
choice | +0.007 |
away | +0.007 |
ook | +0.007 |
taxi | +0.007 |
Neg logit contributions | |
sauce | -0.007 |
gum | -0.007 |
greens | -0.007 |
ink | -0.007 |
glasses | -0.007 |
Token | Once |
Token ID | 7454 |
Feature activation | +2.76e-02 |
Pos logit contributions | |
greens | +0.054 |
fiber | +0.054 |
sauce | +0.054 |
subtract | +0.053 |
gloss | +0.052 |
Neg logit contributions | |
away | -0.035 |
choice | -0.033 |
decision | -0.033 |
ively | -0.031 |
ook | -0.031 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.48e-02 |
Pos logit contributions | |
decision | +0.053 |
away | +0.051 |
choice | +0.050 |
doorstep | +0.047 |
checkpoint | +0.045 |
Neg logit contributions | |
sauce | -0.035 |
gum | -0.035 |
triangles | -0.034 |
Ben | -0.033 |
ink | -0.031 |
Token | it |
Token ID | 340 |
Feature activation | -5.70e-03 |
Pos logit contributions | |
sauce | +0.059 |
gum | +0.058 |
greens | +0.056 |
glasses | +0.054 |
ink | +0.053 |
Neg logit contributions | |
decision | -0.052 |
choice | -0.050 |
away | -0.049 |
ook | -0.042 |
taxi | -0.042 |
Token | . |
Token ID | 13 |
Feature activation | -7.18e-03 |
Pos logit contributions | |
sauce | +0.011 |
gum | +0.011 |
greens | +0.010 |
glasses | +0.010 |
ink | +0.010 |
Neg logit contributions | |
decision | -0.008 |
choice | -0.008 |
away | -0.008 |
ook | -0.006 |
taxi | -0.006 |
Token | She |
Token ID | 1375 |
Feature activation | +3.88e-03 |
Pos logit contributions | |
sauce | +0.014 |
gum | +0.014 |
greens | +0.013 |
mustard | +0.013 |
noodles | +0.013 |
Neg logit contributions | |
ook | -0.008 |
decision | -0.007 |
choice | -0.007 |
forward | -0.007 |
ively | -0.007 |
Token | was |
Token ID | 373 |
Feature activation | +3.64e-03 |
Pos logit contributions | |
choice | +0.006 |
decision | +0.006 |
away | +0.006 |
ook | +0.005 |
taxi | +0.005 |
Neg logit contributions | |
greens | -0.006 |
sauce | -0.006 |
gum | -0.006 |
triangles | -0.006 |
subtract | -0.006 |
Token | a |
Token ID | 257 |
Feature activation | +1.30e-02 |
Pos logit contributions | |
decision | +0.005 |
choice | +0.005 |
away | +0.004 |
forward | +0.004 |
ook | +0.004 |
Neg logit contributions | |
sauce | -0.007 |
gum | -0.007 |
greens | -0.007 |
fiber | -0.007 |
Ol | -0.007 |
Token | clim |
Token ID | 5424 |
Feature activation | +7.06e-02 |
Pos logit contributions | |
decision | +0.023 |
choice | +0.022 |
away | +0.022 |
ently | +0.021 |
ook | +0.021 |
Neg logit contributions | |
sauce | -0.020 |
gum | -0.020 |
glasses | -0.018 |
ink | -0.018 |
greens | -0.018 |
Token | ber |
Token ID | 527 |
Feature activation | +3.92e-03 |
Pos logit contributions | |
decision | +0.122 |
choice | +0.111 |
ently | +0.106 |
doorstep | +0.102 |
destination | +0.096 |
Neg logit contributions | |
elt | -0.094 |
cr | -0.092 |
o | -0.087 |
gloss | -0.086 |
cr | -0.085 |
Token | . |
Token ID | 13 |
Feature activation | -5.57e-03 |
Pos logit contributions | |
decision | +0.006 |
choice | +0.005 |
away | +0.005 |
taxi | +0.005 |
ook | +0.004 |
Neg logit contributions | |
sauce | -0.007 |
gum | -0.007 |
greens | -0.007 |
ink | -0.007 |
glasses | -0.007 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -3.35e-02 |
Pos logit contributions | |
sauce | +0.011 |
gum | +0.010 |
greens | +0.010 |
pads | +0.010 |
triangles | +0.010 |
Neg logit contributions | |
away | -0.007 |
ook | -0.006 |
choice | -0.006 |
taxi | -0.006 |
decision | -0.006 |
Token | L |
Token ID | 43 |
Feature activation | -8.10e-04 |
Pos logit contributions | |
greens | +0.060 |
sauce | +0.059 |
subtract | +0.059 |
fiber | +0.058 |
gloss | +0.057 |
Neg logit contributions | |
away | -0.045 |
choice | -0.044 |
decision | -0.043 |
ook | -0.040 |
taxi | -0.040 |
Token | ily |
Token ID | 813 |
Feature activation | -3.74e-03 |
Pos logit contributions | |
sauce | +0.001 |
ink | +0.001 |
gum | +0.001 |
glasses | +0.001 |
spaghetti | +0.001 |
Neg logit contributions | |
decision | -0.002 |
choice | -0.002 |
away | -0.002 |
rier | -0.001 |
checkpoint | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +7.23e-03 |
Pos logit contributions | |
away | +0.020 |
decision | +0.020 |
choice | +0.019 |
forward | +0.018 |
exactly | +0.018 |
Neg logit contributions | |
greens | -0.029 |
sauce | -0.029 |
fiber | -0.027 |
gum | -0.027 |
subtract | -0.027 |
Token | Tim |
Token ID | 5045 |
Feature activation | +1.23e-02 |
Pos logit contributions | |
taxi | +0.009 |
decision | +0.008 |
ook | +0.008 |
choice | +0.008 |
away | +0.008 |
Neg logit contributions | |
sauce | -0.014 |
gum | -0.013 |
fiber | -0.012 |
ink | -0.012 |
dots | -0.012 |
Token | my |
Token ID | 1820 |
Feature activation | +3.27e-03 |
Pos logit contributions | |
decision | +0.020 |
choice | +0.019 |
away | +0.017 |
promise | +0.017 |
ook | +0.015 |
Neg logit contributions | |
gum | -0.021 |
greens | -0.021 |
sauce | -0.020 |
triangles | -0.019 |
subtract | -0.019 |
Token | watched |
Token ID | 7342 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
decision | +0.005 |
choice | +0.005 |
away | +0.004 |
sadly | +0.004 |
promise | +0.004 |
Neg logit contributions | |
gum | -0.006 |
sauce | -0.006 |
greens | -0.006 |
ink | -0.005 |
subtract | -0.005 |
Token | with |
Token ID | 351 |
Feature activation | +2.47e-02 |
Pos logit contributions | |
decision | +0.019 |
away | +0.019 |
choice | +0.018 |
ively | +0.015 |
taxi | +0.015 |
Neg logit contributions | |
pads | -0.033 |
sauce | -0.033 |
gum | -0.032 |
glasses | -0.032 |
greens | -0.032 |
Token | amaz |
Token ID | 40642 |
Feature activation | +7.04e-02 |
Pos logit contributions | |
decision | +0.039 |
choice | +0.039 |
away | +0.036 |
taxi | +0.031 |
forward | +0.030 |
Neg logit contributions | |
sauce | -0.043 |
gum | -0.042 |
greens | -0.041 |
pads | -0.041 |
gloss | -0.040 |
Token | ement |
Token ID | 972 |
Feature activation | +3.75e-02 |
Pos logit contributions | |
decision | +0.076 |
choice | +0.066 |
away | +0.064 |
checkpoint | +0.055 |
doorstep | +0.054 |
Neg logit contributions | |
gum | -0.159 |
sauce | -0.157 |
subtract | -0.149 |
triangles | -0.149 |
greens | -0.149 |
Token | as |
Token ID | 355 |
Feature activation | +1.49e-02 |
Pos logit contributions | |
decision | +0.048 |
away | +0.045 |
choice | +0.045 |
ook | +0.037 |
taxi | +0.036 |
Neg logit contributions | |
sauce | -0.077 |
gum | -0.076 |
greens | -0.073 |
glasses | -0.072 |
ink | -0.070 |
Token | he |
Token ID | 339 |
Feature activation | -6.29e-03 |
Pos logit contributions | |
decision | +0.022 |
choice | +0.022 |
away | +0.022 |
taxi | +0.019 |
ook | +0.018 |
Neg logit contributions | |
sauce | -0.027 |
gum | -0.026 |
greens | -0.026 |
glasses | -0.025 |
pads | -0.025 |
Token | saw |
Token ID | 2497 |
Feature activation | -3.08e-04 |
Pos logit contributions | |
gum | +0.010 |
sauce | +0.010 |
greens | +0.010 |
fiber | +0.009 |
ink | +0.009 |
Neg logit contributions | |
away | -0.010 |
decision | -0.010 |
choice | -0.010 |
exactly | -0.009 |
destination | -0.009 |
Token | stars |
Token ID | 5788 |
Feature activation | +4.73e-03 |
Pos logit contributions | |
greens | +0.001 |
sauce | +0.001 |
fiber | +0.001 |
pads | +0.001 |
Ol | +0.001 |
Neg logit contributions | |
choice | -0.000 |
away | -0.000 |
decision | -0.000 |
exactly | -0.000 |
taxi | -0.000 |
Token | He |
Token ID | 679 |
Feature activation | +1.27e-02 |
Pos logit contributions | |
decision | +0.014 |
choice | +0.013 |
away | +0.013 |
taxi | +0.012 |
ook | +0.012 |
Neg logit contributions | |
sauce | -0.014 |
gum | -0.014 |
Ol | -0.013 |
greens | -0.013 |
ink | -0.013 |
Token | was |
Token ID | 373 |
Feature activation | +2.18e-03 |
Pos logit contributions | |
choice | +0.016 |
decision | +0.015 |
away | +0.015 |
taxi | +0.013 |
ook | +0.013 |
Neg logit contributions | |
greens | -0.025 |
subtract | -0.024 |
gum | -0.024 |
sauce | -0.024 |
triangles | -0.023 |
Token | a |
Token ID | 257 |
Feature activation | -5.11e-04 |
Pos logit contributions | |
decision | +0.003 |
choice | +0.003 |
taxi | +0.003 |
away | +0.003 |
ook | +0.003 |
Neg logit contributions | |
sauce | -0.004 |
greens | -0.004 |
gum | -0.004 |
fiber | -0.004 |
pads | -0.004 |
Token | very |
Token ID | 845 |
Feature activation | -1.01e-02 |
Pos logit contributions | |
sauce | +0.001 |
gum | +0.001 |
greens | +0.001 |
glasses | +0.001 |
ink | +0.001 |
Neg logit contributions | |
decision | -0.001 |
choice | -0.001 |
away | -0.001 |
ook | -0.001 |
ively | -0.001 |
Token | successful |
Token ID | 4388 |
Feature activation | -4.67e-03 |
Pos logit contributions | |
sauce | +0.017 |
gum | +0.017 |
greens | +0.016 |
glasses | +0.016 |
ink | +0.015 |
Neg logit contributions | |
decision | -0.017 |
choice | -0.016 |
away | -0.016 |
ook | -0.014 |
taxi | -0.013 |
Token | clim |
Token ID | 5424 |
Feature activation | +7.04e-02 |
Pos logit contributions | |
sauce | +0.009 |
gum | +0.009 |
greens | +0.009 |
glasses | +0.008 |
subtract | +0.008 |
Neg logit contributions | |
decision | -0.007 |
choice | -0.006 |
away | -0.006 |
taxi | -0.005 |
ook | -0.005 |
Token | ber |
Token ID | 527 |
Feature activation | +3.77e-03 |
Pos logit contributions | |
decision | +0.115 |
choice | +0.103 |
ently | +0.095 |
doorstep | +0.093 |
destination | +0.091 |
Neg logit contributions | |
cr | -0.096 |
elt | -0.093 |
pads | -0.091 |
gloss | -0.091 |
please | -0.090 |
Token | . |
Token ID | 13 |
Feature activation | -5.79e-04 |
Pos logit contributions | |
decision | +0.005 |
choice | +0.004 |
away | +0.004 |
taxi | +0.004 |
ook | +0.004 |
Neg logit contributions | |
greens | -0.008 |
sauce | -0.008 |
gum | -0.008 |
subtract | -0.007 |
ink | -0.007 |
Token | Every |
Token ID | 3887 |
Feature activation | +2.57e-02 |
Pos logit contributions | |
sauce | +0.001 |
gum | +0.001 |
ink | +0.001 |
greens | +0.001 |
Ol | +0.001 |
Neg logit contributions | |
decision | -0.001 |
choice | -0.001 |
away | -0.001 |
taxi | -0.001 |
checkpoint | -0.001 |
Token | day |
Token ID | 1110 |
Feature activation | +9.98e-03 |
Pos logit contributions | |
decision | +0.036 |
choice | +0.035 |
away | +0.033 |
taxi | +0.029 |
hotel | +0.027 |
Neg logit contributions | |
sauce | -0.051 |
greens | -0.050 |
gum | -0.049 |
subtract | -0.048 |
glasses | -0.048 |
Token | he |
Token ID | 339 |
Feature activation | +2.96e-03 |
Pos logit contributions | |
decision | +0.013 |
choice | +0.012 |
away | +0.012 |
taxi | +0.010 |
forward | +0.010 |
Neg logit contributions | |
sauce | -0.020 |
greens | -0.020 |
gum | -0.020 |
pads | -0.019 |
triangles | -0.019 |
Token | could |
Token ID | 714 |
Feature activation | +6.84e-03 |
Pos logit contributions | |
gum | +0.047 |
sauce | +0.046 |
greens | +0.046 |
subtract | +0.044 |
fiber | +0.044 |
Neg logit contributions | |
decision | -0.037 |
away | -0.036 |
choice | -0.034 |
ook | -0.032 |
destination | -0.031 |
Token | see |
Token ID | 766 |
Feature activation | -1.64e-03 |
Pos logit contributions | |
decision | +0.009 |
choice | +0.009 |
away | +0.009 |
ook | +0.008 |
taxi | +0.007 |
Neg logit contributions | |
sauce | -0.014 |
gum | -0.014 |
greens | -0.013 |
glasses | -0.013 |
ink | -0.012 |
Token | the |
Token ID | 262 |
Feature activation | +9.46e-03 |
Pos logit contributions | |
gum | +0.003 |
sauce | +0.003 |
gloss | +0.003 |
greens | +0.003 |
Ol | +0.003 |
Neg logit contributions | |
decision | -0.002 |
choice | -0.002 |
away | -0.002 |
ook | -0.002 |
exactly | -0.002 |
Token | sun |
Token ID | 4252 |
Feature activation | +1.11e-02 |
Pos logit contributions | |
decision | +0.016 |
choice | +0.015 |
away | +0.014 |
destination | +0.012 |
hotel | +0.012 |
Neg logit contributions | |
pads | -0.016 |
gloss | -0.016 |
Ol | -0.015 |
greens | -0.015 |
gum | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | +2.79e-03 |
Pos logit contributions | |
decision | +0.015 |
away | +0.013 |
choice | +0.013 |
destination | +0.012 |
taxi | +0.011 |
Neg logit contributions | |
gum | -0.022 |
sauce | -0.022 |
fiber | -0.022 |
Ol | -0.022 |
glasses | -0.021 |
Token | Far |
Token ID | 6755 |
Feature activation | +7.04e-02 |
Pos logit contributions | |
decision | +0.002 |
away | +0.002 |
taxi | +0.002 |
curiously | +0.002 |
ively | +0.002 |
Neg logit contributions | |
Ol | -0.006 |
pads | -0.006 |
gum | -0.006 |
fiber | -0.006 |
gloss | -0.006 |
Token | below |
Token ID | 2174 |
Feature activation | +6.96e-03 |
Pos logit contributions | |
decision | +0.116 |
choice | +0.108 |
doorstep | +0.104 |
hotel | +0.098 |
ook | +0.098 |
Neg logit contributions | |
gum | -0.110 |
sauce | -0.103 |
pads | -0.099 |
glasses | -0.097 |
ink | -0.095 |
Token | , |
Token ID | 11 |
Feature activation | -1.11e-02 |
Pos logit contributions | |
decision | +0.008 |
away | +0.008 |
choice | +0.007 |
forward | +0.006 |
ively | +0.006 |
Neg logit contributions | |
gum | -0.015 |
sauce | -0.015 |
fiber | -0.015 |
gloss | -0.015 |
greens | -0.014 |
Token | it |
Token ID | 340 |
Feature activation | -2.17e-02 |
Pos logit contributions | |
sauce | +0.022 |
gum | +0.021 |
gloss | +0.021 |
glasses | +0.021 |
greens | +0.020 |
Neg logit contributions | |
decision | -0.015 |
away | -0.014 |
choice | -0.014 |
forward | -0.012 |
taxi | -0.012 |
Token | could |
Token ID | 714 |
Feature activation | +9.75e-04 |
Pos logit contributions | |
gum | +0.041 |
sauce | +0.039 |
fiber | +0.037 |
greens | +0.037 |
glasses | +0.037 |
Neg logit contributions | |
decision | -0.031 |
away | -0.031 |
choice | -0.029 |
destination | -0.027 |
ook | -0.026 |
Token | see |
Token ID | 766 |
Feature activation | -1.89e-03 |
Pos logit contributions | |
decision | +0.001 |
choice | +0.001 |
away | +0.001 |
taxi | +0.001 |
ook | +0.001 |
Neg logit contributions | |
sauce | -0.002 |
gum | -0.002 |
greens | -0.002 |
glasses | -0.002 |
ink | -0.002 |
Token | spotted |
Token ID | 13489 |
Feature activation | -7.47e-03 |
Pos logit contributions | |
decision | +0.007 |
choice | +0.006 |
away | +0.006 |
ook | +0.006 |
doorstep | +0.005 |
Neg logit contributions | |
gum | -0.006 |
greens | -0.006 |
sauce | -0.006 |
subtract | -0.005 |
triangles | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | +1.22e-02 |
Pos logit contributions | |
greens | +0.014 |
subtract | +0.014 |
fiber | +0.014 |
pads | +0.013 |
sauce | +0.013 |
Neg logit contributions | |
choice | -0.009 |
exactly | -0.009 |
decision | -0.009 |
away | -0.009 |
taxi | -0.009 |
Token | row |
Token ID | 5752 |
Feature activation | +8.55e-03 |
Pos logit contributions | |
decision | +0.020 |
choice | +0.019 |
away | +0.018 |
taxi | +0.017 |
hotel | +0.015 |
Neg logit contributions | |
sauce | -0.020 |
greens | -0.020 |
pads | -0.019 |
gum | -0.019 |
subtract | -0.019 |
Token | of |
Token ID | 286 |
Feature activation | -1.83e-02 |
Pos logit contributions | |
decision | +0.010 |
choice | +0.009 |
away | +0.009 |
ook | +0.007 |
checkpoint | +0.007 |
Neg logit contributions | |
sauce | -0.019 |
gum | -0.018 |
greens | -0.017 |
ink | -0.017 |
glasses | -0.017 |
Token | books |
Token ID | 3835 |
Feature activation | +2.01e-02 |
Pos logit contributions | |
gloss | +0.028 |
pads | +0.028 |
subtract | +0.028 |
Ol | +0.027 |
greens | +0.027 |
Neg logit contributions | |
choice | -0.030 |
decision | -0.029 |
away | -0.029 |
taxi | -0.029 |
hotel | -0.026 |
Token | hel |
Token ID | 2978 |
Feature activation | +7.03e-02 |
Pos logit contributions | |
decision | +0.018 |
choice | +0.016 |
away | +0.016 |
ook | +0.012 |
taxi | +0.011 |
Neg logit contributions | |
sauce | -0.049 |
gum | -0.049 |
greens | -0.047 |
glasses | -0.046 |
ink | -0.046 |
Token | ves |
Token ID | 1158 |
Feature activation | +3.26e-03 |
Pos logit contributions | |
decision | +0.119 |
choice | +0.112 |
away | +0.105 |
promise | +0.100 |
forward | +0.100 |
Neg logit contributions | |
sauce | -0.103 |
triangles | -0.103 |
gum | -0.103 |
greens | -0.099 |
pads | -0.099 |
Token | in |
Token ID | 287 |
Feature activation | +4.97e-03 |
Pos logit contributions | |
away | +0.003 |
decision | +0.003 |
forward | +0.003 |
choice | +0.003 |
taxi | +0.003 |
Neg logit contributions | |
gum | -0.007 |
sauce | -0.007 |
fiber | -0.007 |
Ol | -0.007 |
subtract | -0.007 |
Token | her |
Token ID | 607 |
Feature activation | -9.92e-04 |
Pos logit contributions | |
away | +0.007 |
decision | +0.007 |
choice | +0.007 |
taxi | +0.006 |
destination | +0.005 |
Neg logit contributions | |
sauce | -0.009 |
gloss | -0.009 |
gum | -0.009 |
pads | -0.009 |
greens | -0.009 |
Token | grandparents |
Token ID | 28571 |
Feature activation | +1.05e-03 |
Pos logit contributions | |
sauce | +0.002 |
gum | +0.002 |
greens | +0.002 |
subtract | +0.002 |
ink | +0.002 |
Neg logit contributions | |
decision | -0.001 |
choice | -0.001 |
away | -0.001 |
taxi | -0.001 |
ook | -0.001 |
Token | ' |
Token ID | 6 |
Feature activation | +5.90e-04 |
Pos logit contributions | |
decision | +0.001 |
away | +0.001 |
choice | +0.001 |
doorstep | +0.001 |
taxi | +0.001 |
Neg logit contributions | |
sauce | -0.002 |
gum | -0.002 |
Ol | -0.002 |
greens | -0.002 |
fiber | -0.002 |
Token | they |
Token ID | 484 |
Feature activation | +7.28e-05 |
Pos logit contributions | |
decision | +0.009 |
away | +0.009 |
choice | +0.008 |
checkpoint | +0.008 |
ook | +0.008 |
Neg logit contributions | |
sauce | -0.008 |
gum | -0.008 |
ink | -0.008 |
greens | -0.008 |
glasses | -0.008 |
Token | went |
Token ID | 1816 |
Feature activation | +4.82e-03 |
Pos logit contributions | |
decision | +0.000 |
away | +0.000 |
choice | +0.000 |
taxi | +0.000 |
ook | +0.000 |
Neg logit contributions | |
sauce | -0.000 |
gum | -0.000 |
greens | -0.000 |
glasses | -0.000 |
ink | -0.000 |
Token | on |
Token ID | 319 |
Feature activation | -1.05e-02 |
Pos logit contributions | |
decision | +0.011 |
choice | +0.010 |
ently | +0.010 |
ively | +0.009 |
rily | +0.009 |
Neg logit contributions | |
gum | -0.005 |
sauce | -0.005 |
pads | -0.005 |
glasses | -0.005 |
shorts | -0.004 |
Token | the |
Token ID | 262 |
Feature activation | +1.70e-03 |
Pos logit contributions | |
gum | +0.013 |
sauce | +0.012 |
glasses | +0.012 |
ink | +0.011 |
greens | +0.010 |
Neg logit contributions | |
decision | -0.022 |
choice | -0.021 |
away | -0.021 |
rier | -0.020 |
ook | -0.020 |
Token | fer |
Token ID | 11354 |
Feature activation | +3.66e-02 |
Pos logit contributions | |
decision | +0.003 |
away | +0.003 |
choice | +0.003 |
ook | +0.003 |
ively | +0.003 |
Neg logit contributions | |
gum | -0.003 |
sauce | -0.003 |
greens | -0.002 |
glasses | -0.002 |
subtract | -0.002 |
Token | ris |
Token ID | 2442 |
Feature activation | +6.96e-02 |
Pos logit contributions | |
decision | +0.047 |
choice | +0.043 |
rier | +0.041 |
tha | +0.041 |
cancel | +0.040 |
Neg logit contributions | |
pads | -0.066 |
subtract | -0.064 |
gum | -0.063 |
scored | -0.061 |
drew | -0.060 |
Token | wheel |
Token ID | 7825 |
Feature activation | +2.28e-02 |
Pos logit contributions | |
decision | +0.138 |
away | +0.133 |
choice | +0.132 |
ook | +0.128 |
iring | +0.123 |
Neg logit contributions | |
gum | -0.082 |
pads | -0.081 |
subtract | -0.078 |
sauce | -0.078 |
glasses | -0.076 |
Token | together |
Token ID | 1978 |
Feature activation | -1.40e-03 |
Pos logit contributions | |
decision | +0.035 |
choice | +0.034 |
away | +0.033 |
ook | +0.028 |
taxi | +0.027 |
Neg logit contributions | |
sauce | -0.040 |
gum | -0.040 |
greens | -0.038 |
glasses | -0.037 |
ink | -0.037 |
Token | , |
Token ID | 11 |
Feature activation | -5.59e-03 |
Pos logit contributions | |
sauce | +0.003 |
gum | +0.003 |
greens | +0.003 |
Ol | +0.003 |
fiber | +0.003 |
Neg logit contributions | |
decision | -0.002 |
away | -0.002 |
choice | -0.002 |
ook | -0.001 |
forward | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | -5.28e-04 |
Pos logit contributions | |
sauce | +0.009 |
gum | +0.009 |
greens | +0.008 |
glasses | +0.008 |
ink | +0.008 |
Neg logit contributions | |
decision | -0.010 |
choice | -0.009 |
away | -0.009 |
ook | -0.008 |
taxi | -0.008 |
Token | Lily |
Token ID | 20037 |
Feature activation | -3.98e-03 |
Pos logit contributions | |
sauce | +0.001 |
gum | +0.001 |
greens | +0.001 |
glasses | +0.001 |
subtract | +0.001 |
Neg logit contributions | |
decision | -0.001 |
away | -0.001 |
choice | -0.001 |
ook | -0.001 |
taxi | -0.001 |
Token | eddy |
Token ID | 21874 |
Feature activation | -1.39e-04 |
Pos logit contributions | |
decision | +0.029 |
choice | +0.026 |
away | +0.024 |
taxi | +0.022 |
forward | +0.021 |
Neg logit contributions | |
sauce | -0.040 |
gum | -0.040 |
ink | -0.037 |
pads | -0.037 |
glasses | -0.037 |
Token | bear |
Token ID | 6842 |
Feature activation | -2.86e-03 |
Pos logit contributions | |
subtract | +0.000 |
greens | +0.000 |
gum | +0.000 |
sauce | +0.000 |
blond | +0.000 |
Neg logit contributions | |
decision | -0.000 |
choice | -0.000 |
away | -0.000 |
taxi | -0.000 |
doorstep | -0.000 |
Token | and |
Token ID | 290 |
Feature activation | -7.01e-03 |
Pos logit contributions | |
triangles | +0.006 |
subtract | +0.006 |
greens | +0.006 |
Ol | +0.006 |
squares | +0.005 |
Neg logit contributions | |
decision | -0.003 |
doorstep | -0.003 |
away | -0.003 |
choice | -0.003 |
taxi | -0.003 |
Token | watched |
Token ID | 7342 |
Feature activation | +1.49e-02 |
Pos logit contributions | |
Ol | +0.011 |
greens | +0.011 |
gum | +0.011 |
subtract | +0.010 |
sauce | +0.010 |
Neg logit contributions | |
choice | -0.011 |
decision | -0.011 |
away | -0.009 |
exactly | -0.009 |
taxi | -0.009 |
Token | in |
Token ID | 287 |
Feature activation | +8.01e-04 |
Pos logit contributions | |
decision | +0.019 |
away | +0.018 |
choice | +0.018 |
ook | +0.015 |
ively | +0.015 |
Neg logit contributions | |
sauce | -0.030 |
gum | -0.029 |
pads | -0.029 |
greens | -0.028 |
glasses | -0.028 |
Token | amaz |
Token ID | 40642 |
Feature activation | +6.96e-02 |
Pos logit contributions | |
decision | +0.001 |
choice | +0.001 |
away | +0.001 |
taxi | +0.001 |
purchase | +0.001 |
Neg logit contributions | |
sauce | -0.001 |
gum | -0.001 |
greens | -0.001 |
pads | -0.001 |
gloss | -0.001 |
Token | ement |
Token ID | 972 |
Feature activation | +3.44e-02 |
Pos logit contributions | |
decision | +0.077 |
choice | +0.068 |
away | +0.064 |
ook | +0.057 |
checkpoint | +0.056 |
Neg logit contributions | |
gum | -0.153 |
sauce | -0.151 |
triangles | -0.142 |
greens | -0.142 |
subtract | -0.141 |
Token | as |
Token ID | 355 |
Feature activation | +1.30e-02 |
Pos logit contributions | |
decision | +0.048 |
choice | +0.045 |
away | +0.045 |
ook | +0.038 |
taxi | +0.037 |
Neg logit contributions | |
sauce | -0.067 |
gum | -0.066 |
greens | -0.063 |
glasses | -0.062 |
ink | -0.061 |
Token | it |
Token ID | 340 |
Feature activation | -2.71e-02 |
Pos logit contributions | |
decision | +0.019 |
choice | +0.019 |
away | +0.018 |
ook | +0.016 |
exactly | +0.015 |
Neg logit contributions | |
sauce | -0.024 |
gum | -0.023 |
greens | -0.023 |
pads | -0.022 |
glasses | -0.022 |
Token | shrunk |
Token ID | 38900 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
gum | +0.049 |
fiber | +0.047 |
sauce | +0.047 |
subtract | +0.045 |
greens | +0.044 |
Neg logit contributions | |
decision | -0.040 |
choice | -0.038 |
ered | -0.037 |
away | -0.037 |
exactly | -0.036 |
Token | an |
Token ID | 281 |
Feature activation | +1.75e-02 |
Pos logit contributions | |
sauce | +0.028 |
gum | +0.028 |
subtract | +0.027 |
greens | +0.027 |
Ol | +0.027 |
Neg logit contributions | |
away | -0.017 |
decision | -0.016 |
choice | -0.016 |
ook | -0.012 |
ively | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | -9.29e-03 |
Pos logit contributions | |
sauce | +0.018 |
gum | +0.018 |
subtract | +0.018 |
Ol | +0.018 |
greens | +0.017 |
Neg logit contributions | |
decision | -0.010 |
choice | -0.010 |
away | -0.010 |
forward | -0.008 |
promise | -0.008 |
Token | He |
Token ID | 679 |
Feature activation | -1.38e-04 |
Pos logit contributions | |
sauce | +0.017 |
gum | +0.016 |
ink | +0.015 |
dots | +0.015 |
greens | +0.015 |
Neg logit contributions | |
decision | -0.012 |
choice | -0.012 |
taxi | -0.012 |
ook | -0.012 |
away | -0.011 |
Token | watched |
Token ID | 7342 |
Feature activation | +1.04e-02 |
Pos logit contributions | |
gum | +0.000 |
sauce | +0.000 |
subtract | +0.000 |
ink | +0.000 |
gloss | +0.000 |
Neg logit contributions | |
decision | -0.000 |
choice | -0.000 |
away | -0.000 |
sadly | -0.000 |
exactly | -0.000 |
Token | them |
Token ID | 606 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
away | +0.011 |
decision | +0.011 |
choice | +0.011 |
forward | +0.010 |
taxi | +0.009 |
Neg logit contributions | |
gum | -0.022 |
sauce | -0.022 |
pads | -0.021 |
subtract | -0.021 |
glasses | -0.021 |
Token | in |
Token ID | 287 |
Feature activation | -8.05e-04 |
Pos logit contributions | |
sauce | +0.020 |
gum | +0.019 |
greens | +0.019 |
glasses | +0.018 |
ink | +0.018 |
Neg logit contributions | |
decision | -0.017 |
choice | -0.016 |
away | -0.016 |
ook | -0.013 |
taxi | -0.013 |
Token | amaz |
Token ID | 40642 |
Feature activation | +6.95e-02 |
Pos logit contributions | |
gloss | +0.002 |
gum | +0.002 |
sauce | +0.002 |
pads | +0.002 |
subtract | +0.002 |
Neg logit contributions | |
decision | -0.001 |
choice | -0.001 |
away | -0.001 |
taxi | -0.001 |
forward | -0.001 |
Token | ement |
Token ID | 972 |
Feature activation | +3.34e-02 |
Pos logit contributions | |
decision | +0.081 |
choice | +0.071 |
away | +0.065 |
ook | +0.059 |
promise | +0.059 |
Neg logit contributions | |
gum | -0.148 |
sauce | -0.146 |
triangles | -0.140 |
greens | -0.139 |
glasses | -0.137 |
Token | and |
Token ID | 290 |
Feature activation | -6.49e-03 |
Pos logit contributions | |
decision | +0.039 |
away | +0.038 |
choice | +0.036 |
ook | +0.030 |
exactly | +0.029 |
Neg logit contributions | |
sauce | -0.072 |
gum | -0.072 |
greens | -0.068 |
glasses | -0.068 |
ink | -0.067 |
Token | even |
Token ID | 772 |
Feature activation | -1.13e-02 |
Pos logit contributions | |
gum | +0.012 |
sauce | +0.012 |
Ol | +0.011 |
fiber | +0.011 |
ink | +0.011 |
Neg logit contributions | |
decision | -0.010 |
away | -0.009 |
choice | -0.009 |
sadly | -0.009 |
exactly | -0.008 |
Token | laughed |
Token ID | 13818 |
Feature activation | -1.90e-02 |
Pos logit contributions | |
gum | +0.017 |
sauce | +0.016 |
greens | +0.015 |
glasses | +0.015 |
fiber | +0.015 |
Neg logit contributions | |
away | -0.022 |
decision | -0.021 |
choice | -0.020 |
forward | -0.018 |
ively | -0.018 |
Token | a |
Token ID | 257 |
Feature activation | +2.06e-03 |
Pos logit contributions | |
sauce | +0.042 |
gum | +0.042 |
fiber | +0.042 |
subtract | +0.041 |
Ol | +0.040 |
Neg logit contributions | |
away | -0.020 |
decision | -0.019 |
forward | -0.018 |
choice | -0.017 |
exactly | -0.015 |
Token | ran |
Token ID | 4966 |
Feature activation | +4.43e-03 |
Pos logit contributions | |
sauce | +0.015 |
pill | +0.014 |
greens | +0.014 |
gum | +0.014 |
subtract | +0.014 |
Neg logit contributions | |
choice | -0.015 |
decision | -0.014 |
taxi | -0.014 |
away | -0.014 |
passport | -0.012 |
Token | towards |
Token ID | 3371 |
Feature activation | +1.85e-02 |
Pos logit contributions | |
decision | +0.008 |
choice | +0.008 |
away | +0.007 |
ook | +0.007 |
tha | +0.007 |
Neg logit contributions | |
sauce | -0.007 |
gum | -0.007 |
glasses | -0.006 |
greens | -0.006 |
pads | -0.006 |
Token | the |
Token ID | 262 |
Feature activation | -4.97e-04 |
Pos logit contributions | |
decision | +0.029 |
away | +0.028 |
choice | +0.028 |
taxi | +0.024 |
ook | +0.024 |
Neg logit contributions | |
sauce | -0.032 |
gum | -0.032 |
greens | -0.031 |
glasses | -0.029 |
pads | -0.029 |
Token | circus |
Token ID | 33125 |
Feature activation | +3.31e-02 |
Pos logit contributions | |
sauce | +0.001 |
gum | +0.001 |
greens | +0.001 |
glasses | +0.001 |
triangles | +0.001 |
Neg logit contributions | |
decision | -0.001 |
choice | -0.001 |
away | -0.001 |
ook | -0.001 |
tha | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +1.04e-02 |
Pos logit contributions | |
decision | +0.053 |
choice | +0.050 |
away | +0.050 |
ook | +0.043 |
taxi | +0.042 |
Neg logit contributions | |
sauce | -0.057 |
gum | -0.056 |
greens | -0.054 |
glasses | -0.052 |
ink | -0.052 |
Token | Once |
Token ID | 4874 |
Feature activation | +6.92e-02 |
Pos logit contributions | |
decision | +0.018 |
choice | +0.017 |
away | +0.017 |
ook | +0.015 |
taxi | +0.015 |
Neg logit contributions | |
sauce | -0.016 |
gum | -0.016 |
greens | -0.015 |
triangles | -0.015 |
ink | -0.015 |
Token | there |
Token ID | 612 |
Feature activation | +6.87e-03 |
Pos logit contributions | |
decision | +0.157 |
choice | +0.149 |
away | +0.144 |
ook | +0.140 |
checkpoint | +0.140 |
Neg logit contributions | |
gum | -0.076 |
sauce | -0.073 |
spaghetti | -0.061 |
Ben | -0.058 |
ink | -0.053 |
Token | , |
Token ID | 11 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
decision | +0.013 |
away | +0.013 |
choice | +0.013 |
ook | +0.012 |
checkpoint | +0.011 |
Neg logit contributions | |
gum | -0.009 |
sauce | -0.009 |
glasses | -0.008 |
greens | -0.008 |
ink | -0.008 |
Token | he |
Token ID | 339 |
Feature activation | +1.92e-03 |
Pos logit contributions | |
decision | +0.025 |
choice | +0.024 |
away | +0.023 |
ook | +0.022 |
checkpoint | +0.021 |
Neg logit contributions | |
gum | -0.015 |
sauce | -0.014 |
glasses | -0.012 |
spaghetti | -0.012 |
ink | -0.012 |
Token | saw |
Token ID | 2497 |
Feature activation | +1.62e-03 |
Pos logit contributions | |
decision | +0.003 |
choice | +0.003 |
away | +0.003 |
ook | +0.003 |
destination | +0.003 |
Neg logit contributions | |
sauce | -0.003 |
greens | -0.003 |
gum | -0.003 |
fiber | -0.003 |
ink | -0.003 |
Token | lots |
Token ID | 6041 |
Feature activation | -1.30e-02 |
Pos logit contributions | |
decision | +0.003 |
choice | +0.003 |
away | +0.003 |
taxi | +0.002 |
ook | +0.002 |
Neg logit contributions | |
sauce | -0.003 |
gum | -0.003 |
greens | -0.003 |
subtract | -0.002 |
pads | -0.002 |
Token | The |
Token ID | 383 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
decision | +0.004 |
sadly | +0.004 |
doorstep | +0.004 |
choice | +0.004 |
destination | +0.004 |
Neg logit contributions | |
cr | -0.010 |
dots | -0.010 |
greens | -0.009 |
fiber | -0.009 |
sauce | -0.009 |
Token | rain |
Token ID | 6290 |
Feature activation | -3.14e-02 |
Pos logit contributions | |
gum | +0.018 |
sauce | +0.018 |
greens | +0.017 |
subtract | +0.017 |
glasses | +0.016 |
Neg logit contributions | |
decision | -0.030 |
choice | -0.029 |
away | -0.027 |
taxi | -0.026 |
doorstep | -0.025 |
Token | turned |
Token ID | 2900 |
Feature activation | +4.33e-03 |
Pos logit contributions | |
sauce | +0.055 |
gum | +0.054 |
greens | +0.052 |
subtract | +0.051 |
glasses | +0.050 |
Neg logit contributions | |
decision | -0.052 |
choice | -0.049 |
away | -0.047 |
taxi | -0.044 |
tha | -0.041 |
Token | into |
Token ID | 656 |
Feature activation | +6.44e-03 |
Pos logit contributions | |
away | +0.005 |
forward | +0.005 |
decision | +0.005 |
taxi | +0.004 |
choice | +0.004 |
Neg logit contributions | |
sauce | -0.009 |
fiber | -0.009 |
subtract | -0.009 |
gum | -0.009 |
greens | -0.009 |
Token | ha |
Token ID | 387 |
Feature activation | -1.73e-02 |
Pos logit contributions | |
choice | +0.008 |
decision | +0.008 |
away | +0.008 |
exactly | +0.007 |
security | +0.007 |
Neg logit contributions | |
pads | -0.012 |
fiber | -0.012 |
gloss | -0.012 |
greens | -0.012 |
pill | -0.011 |
Token | ilst |
Token ID | 11750 |
Feature activation | +6.91e-02 |
Pos logit contributions | |
sauce | +0.032 |
gum | +0.031 |
greens | +0.031 |
ink | +0.028 |
subtract | +0.028 |
Neg logit contributions | |
decision | -0.025 |
choice | -0.024 |
away | -0.023 |
ook | -0.023 |
or | -0.021 |
Token | ones |
Token ID | 1952 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
decision | +0.117 |
choice | +0.112 |
away | +0.109 |
ook | +0.096 |
checkpoint | +0.093 |
Neg logit contributions | |
sauce | -0.114 |
gum | -0.112 |
greens | -0.104 |
ink | -0.103 |
glasses | -0.103 |
Token | and |
Token ID | 290 |
Feature activation | +1.14e-02 |
Pos logit contributions | |
greens | +0.024 |
sauce | +0.023 |
gum | +0.023 |
subtract | +0.022 |
glasses | +0.022 |
Neg logit contributions | |
away | -0.014 |
decision | -0.014 |
choice | -0.013 |
taxi | -0.012 |
ively | -0.012 |
Token | the |
Token ID | 262 |
Feature activation | +2.64e-04 |
Pos logit contributions | |
away | +0.010 |
decision | +0.010 |
choice | +0.009 |
sadly | +0.009 |
destination | +0.009 |
Neg logit contributions | |
pill | -0.026 |
greens | -0.025 |
fiber | -0.024 |
mascara | -0.024 |
gloss | -0.024 |
Token | thunder |
Token ID | 18355 |
Feature activation | -1.58e-02 |
Pos logit contributions | |
decision | +0.001 |
choice | +0.000 |
taxi | +0.000 |
hotel | +0.000 |
away | +0.000 |
Neg logit contributions | |
pads | -0.000 |
Ol | -0.000 |
sauce | -0.000 |
greens | -0.000 |
gum | -0.000 |
Token | roared |
Token ID | 45615 |
Feature activation | -1.98e-03 |
Pos logit contributions | |
fiber | +0.029 |
sauce | +0.028 |
gum | +0.028 |
subtract | +0.027 |
vitamins | +0.027 |
Neg logit contributions | |
away | -0.023 |
decision | -0.021 |
taxi | -0.021 |
choice | -0.020 |
ively | -0.019 |
Token | go |
Token ID | 467 |
Feature activation | +3.62e-03 |
Pos logit contributions | |
choice | +0.003 |
decision | +0.003 |
away | +0.003 |
taxi | +0.003 |
purchase | +0.003 |
Neg logit contributions | |
sauce | -0.005 |
gum | -0.005 |
greens | -0.005 |
pads | -0.005 |
gloss | -0.005 |
Token | ! |
Token ID | 0 |
Feature activation | +4.05e-03 |
Pos logit contributions | |
decision | +0.006 |
choice | +0.006 |
away | +0.005 |
ook | +0.005 |
tha | +0.005 |
Neg logit contributions | |
sauce | -0.006 |
gum | -0.006 |
glasses | -0.006 |
greens | -0.006 |
pads | -0.006 |
Token | Everyone |
Token ID | 11075 |
Feature activation | -4.95e-03 |
Pos logit contributions | |
decision | +0.006 |
taxi | +0.005 |
away | +0.005 |
choice | +0.005 |
ook | +0.005 |
Neg logit contributions | |
sauce | -0.007 |
gum | -0.007 |
greens | -0.007 |
ink | -0.007 |
triangles | -0.006 |
Token | watched |
Token ID | 7342 |
Feature activation | +7.59e-03 |
Pos logit contributions | |
sauce | +0.008 |
gum | +0.008 |
greens | +0.008 |
glasses | +0.008 |
ink | +0.008 |
Neg logit contributions | |
decision | -0.008 |
away | -0.008 |
choice | -0.008 |
taxi | -0.007 |
ook | -0.007 |
Token | in |
Token ID | 287 |
Feature activation | -7.39e-03 |
Pos logit contributions | |
decision | +0.009 |
away | +0.009 |
choice | +0.009 |
forward | +0.007 |
ively | +0.007 |
Neg logit contributions | |
sauce | -0.016 |
gum | -0.016 |
pads | -0.015 |
glasses | -0.015 |
greens | -0.015 |
Token | amaz |
Token ID | 40642 |
Feature activation | +6.91e-02 |
Pos logit contributions | |
sauce | +0.014 |
gum | +0.014 |
gloss | +0.013 |
pads | +0.013 |
Ol | +0.013 |
Neg logit contributions | |
decision | -0.011 |
choice | -0.010 |
away | -0.010 |
taxi | -0.008 |
forward | -0.008 |
Token | ement |
Token ID | 972 |
Feature activation | +3.43e-02 |
Pos logit contributions | |
decision | +0.077 |
choice | +0.068 |
away | +0.065 |
ook | +0.058 |
checkpoint | +0.056 |
Neg logit contributions | |
gum | -0.147 |
sauce | -0.147 |
triangles | -0.142 |
greens | -0.140 |
subtract | -0.140 |
Token | as |
Token ID | 355 |
Feature activation | +8.46e-03 |
Pos logit contributions | |
decision | +0.043 |
away | +0.040 |
choice | +0.040 |
ook | +0.033 |
taxi | +0.032 |
Neg logit contributions | |
sauce | -0.072 |
gum | -0.072 |
greens | -0.067 |
glasses | -0.067 |
ink | -0.066 |
Token | it |
Token ID | 340 |
Feature activation | -2.70e-02 |
Pos logit contributions | |
decision | +0.013 |
choice | +0.012 |
away | +0.012 |
ook | +0.010 |
taxi | +0.010 |
Neg logit contributions | |
sauce | -0.016 |
gum | -0.015 |
greens | -0.015 |
glasses | -0.014 |
pads | -0.014 |
Token | took |
Token ID | 1718 |
Feature activation | -1.20e-02 |
Pos logit contributions | |
gum | +0.049 |
sauce | +0.048 |
fiber | +0.044 |
greens | +0.043 |
ink | +0.043 |
Neg logit contributions | |
decision | -0.044 |
choice | -0.042 |
away | -0.041 |
ered | -0.036 |
destination | -0.036 |
Token | off |
Token ID | 572 |
Feature activation | -3.78e-03 |
Pos logit contributions | |
sauce | +0.023 |
gum | +0.023 |
Ol | +0.022 |
gloss | +0.022 |
greens | +0.022 |
Neg logit contributions | |
away | -0.017 |
decision | -0.016 |
choice | -0.016 |
forward | -0.014 |
taxi | -0.013 |
Token | much |
Token ID | 881 |
Feature activation | -1.80e-02 |
Pos logit contributions | |
sauce | +0.028 |
gum | +0.028 |
greens | +0.026 |
triangles | +0.026 |
ink | +0.026 |
Neg logit contributions | |
decision | -0.033 |
choice | -0.032 |
away | -0.030 |
checkpoint | -0.027 |
tha | -0.026 |
Token | . |
Token ID | 13 |
Feature activation | -4.66e-03 |
Pos logit contributions | |
sauce | +0.033 |
gum | +0.032 |
greens | +0.032 |
subtract | +0.031 |
glasses | +0.031 |
Neg logit contributions | |
decision | -0.026 |
away | -0.026 |
choice | -0.025 |
taxi | -0.021 |
ook | -0.021 |
Token | One |
Token ID | 1881 |
Feature activation | -7.95e-03 |
Pos logit contributions | |
sauce | +0.006 |
gum | +0.006 |
greens | +0.006 |
glasses | +0.006 |
ink | +0.006 |
Neg logit contributions | |
decision | -0.009 |
choice | -0.009 |
away | -0.009 |
ook | -0.008 |
taxi | -0.008 |
Token | day |
Token ID | 1110 |
Feature activation | -2.55e-04 |
Pos logit contributions | |
sauce | +0.013 |
gum | +0.013 |
greens | +0.012 |
ink | +0.012 |
glasses | +0.012 |
Neg logit contributions | |
decision | -0.014 |
choice | -0.013 |
away | -0.013 |
ook | -0.011 |
checkpoint | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | -4.76e-04 |
Pos logit contributions | |
gum | +0.000 |
sauce | +0.000 |
greens | +0.000 |
glasses | +0.000 |
ink | +0.000 |
Neg logit contributions | |
decision | -0.000 |
choice | -0.000 |
away | -0.000 |
checkpoint | -0.000 |
doorstep | -0.000 |
Token | Far |
Token ID | 6755 |
Feature activation | +6.88e-02 |
Pos logit contributions | |
gum | +0.000 |
glasses | +0.000 |
greens | +0.000 |
Ol | +0.000 |
ink | +0.000 |
Neg logit contributions | |
decision | -0.001 |
choice | -0.001 |
away | -0.001 |
ook | -0.001 |
doorstep | -0.001 |
Token | ah |
Token ID | 993 |
Feature activation | +4.99e-03 |
Pos logit contributions | |
decision | +0.124 |
choice | +0.114 |
doorstep | +0.113 |
cancel | +0.104 |
ook | +0.103 |
Neg logit contributions | |
pads | -0.093 |
gum | -0.090 |
glasses | -0.085 |
Ol | -0.085 |
dots | -0.083 |
Token | went |
Token ID | 1816 |
Feature activation | +3.45e-03 |
Pos logit contributions | |
decision | +0.008 |
away | +0.008 |
choice | +0.007 |
ook | +0.007 |
taxi | +0.006 |
Neg logit contributions | |
sauce | -0.009 |
greens | -0.009 |
gum | -0.009 |
subtract | -0.008 |
glasses | -0.008 |
Token | to |
Token ID | 284 |
Feature activation | -3.73e-03 |
Pos logit contributions | |
decision | +0.005 |
away | +0.005 |
choice | +0.005 |
ook | +0.004 |
taxi | +0.004 |
Neg logit contributions | |
sauce | -0.006 |
gum | -0.006 |
greens | -0.006 |
glasses | -0.006 |
triangles | -0.006 |
Token | the |
Token ID | 262 |
Feature activation | -3.01e-03 |
Pos logit contributions | |
sauce | +0.006 |
gum | +0.006 |
greens | +0.006 |
glasses | +0.006 |
pads | +0.006 |
Neg logit contributions | |
decision | -0.006 |
choice | -0.006 |
away | -0.006 |
ook | -0.005 |
taxi | -0.005 |
Token | park |
Token ID | 3952 |
Feature activation | +3.32e-02 |
Pos logit contributions | |
greens | +0.006 |
sauce | +0.006 |
gum | +0.005 |
pads | +0.005 |
subtract | +0.005 |
Neg logit contributions | |
decision | -0.004 |
away | -0.004 |
choice | -0.004 |
taxi | -0.004 |
hotel | -0.003 |
Token | " |
Token ID | 366 |
Feature activation | +6.31e-03 |
Pos logit contributions | |
taxi | +0.009 |
decision | +0.008 |
ook | +0.008 |
choice | +0.008 |
away | +0.008 |
Neg logit contributions | |
sauce | -0.014 |
gum | -0.013 |
fiber | -0.012 |
ink | -0.012 |
dots | -0.012 |
Token | Welcome |
Token ID | 14618 |
Feature activation | -1.14e-03 |
Pos logit contributions | |
away | +0.009 |
forward | +0.008 |
taxi | +0.007 |
checkpoint | +0.007 |
Only | +0.007 |
Neg logit contributions | |
sauce | -0.012 |
ink | -0.011 |
gum | -0.011 |
glasses | -0.011 |
pens | -0.011 |
Token | home |
Token ID | 1363 |
Feature activation | +5.40e-03 |
Pos logit contributions | |
sauce | +0.002 |
gum | +0.002 |
greens | +0.002 |
ink | +0.002 |
glasses | +0.002 |
Neg logit contributions | |
decision | -0.002 |
choice | -0.002 |
away | -0.002 |
ook | -0.001 |
taxi | -0.001 |
Token | , |
Token ID | 11 |
Feature activation | -9.08e-03 |
Pos logit contributions | |
decision | +0.007 |
choice | +0.007 |
away | +0.007 |
ook | +0.006 |
taxi | +0.005 |
Neg logit contributions | |
sauce | -0.011 |
gum | -0.011 |
greens | -0.010 |
glasses | -0.010 |
ink | -0.010 |
Token | B |
Token ID | 347 |
Feature activation | +1.58e-02 |
Pos logit contributions | |
sauce | +0.015 |
gum | +0.015 |
greens | +0.014 |
subtract | +0.014 |
glasses | +0.014 |
Neg logit contributions | |
decision | -0.015 |
choice | -0.015 |
away | -0.014 |
taxi | -0.013 |
ook | -0.012 |
Token | oun |
Token ID | 977 |
Feature activation | +6.88e-02 |
Pos logit contributions | |
decision | +0.021 |
choice | +0.020 |
away | +0.019 |
ook | +0.016 |
taxi | +0.016 |
Neg logit contributions | |
sauce | -0.031 |
gum | -0.031 |
greens | -0.030 |
glasses | -0.029 |
ink | -0.029 |
Token | cy |
Token ID | 948 |
Feature activation | +1.73e-02 |
Pos logit contributions | |
decision | +0.113 |
choice | +0.106 |
away | +0.094 |
ook | +0.090 |
doorstep | +0.086 |
Neg logit contributions | |
sauce | -0.109 |
gum | -0.104 |
Ol | -0.101 |
cr | -0.100 |
drew | -0.095 |
Token | !", |
Token ID | 40754 |
Feature activation | -3.85e-03 |
Pos logit contributions | |
decision | +0.024 |
choice | +0.023 |
away | +0.023 |
ook | +0.019 |
taxi | +0.019 |
Neg logit contributions | |
sauce | -0.033 |
gum | -0.033 |
greens | -0.032 |
glasses | -0.031 |
subtract | -0.031 |
Token | they |
Token ID | 484 |
Feature activation | -6.43e-03 |
Pos logit contributions | |
sauce | +0.007 |
gum | +0.006 |
greens | +0.006 |
triangles | +0.006 |
fiber | +0.006 |
Neg logit contributions | |
decision | -0.005 |
away | -0.005 |
choice | -0.005 |
taxi | -0.005 |
tha | -0.005 |
Token | said |
Token ID | 531 |
Feature activation | -1.55e-03 |
Pos logit contributions | |
gum | +0.012 |
sauce | +0.012 |
greens | +0.011 |
triangles | +0.011 |
ink | +0.011 |
Neg logit contributions | |
decision | -0.009 |
choice | -0.009 |
away | -0.009 |
taxi | -0.008 |
promise | -0.008 |
Token | , |
Token ID | 11 |
Feature activation | -2.49e-02 |
Pos logit contributions | |
sauce | +0.003 |
gum | +0.003 |
greens | +0.003 |
pads | +0.003 |
subtract | +0.003 |
Neg logit contributions | |
decision | -0.002 |
choice | -0.002 |
away | -0.002 |
ook | -0.002 |
forward | -0.002 |
Token | Tim |
Token ID | 5045 |
Feature activation | +1.84e-02 |
Pos logit contributions | |
sauce | +0.020 |
glasses | +0.020 |
greens | +0.019 |
lenses | +0.019 |
Ol | +0.019 |
Neg logit contributions | |
decision | -0.013 |
choice | -0.012 |
away | -0.011 |
exactly | -0.010 |
taxi | -0.010 |
Token | was |
Token ID | 373 |
Feature activation | +7.82e-03 |
Pos logit contributions | |
decision | +0.029 |
choice | +0.028 |
away | +0.026 |
ook | +0.023 |
promise | +0.023 |
Neg logit contributions | |
greens | -0.033 |
sauce | -0.031 |
gum | -0.031 |
subtract | -0.030 |
triangles | -0.030 |
Token | a |
Token ID | 257 |
Feature activation | +1.88e-02 |
Pos logit contributions | |
decision | +0.014 |
choice | +0.013 |
away | +0.013 |
ook | +0.011 |
checkpoint | +0.011 |
Neg logit contributions | |
sauce | -0.013 |
gum | -0.012 |
greens | -0.012 |
glasses | -0.011 |
ink | -0.011 |
Token | great |
Token ID | 1049 |
Feature activation | +6.18e-03 |
Pos logit contributions | |
decision | +0.031 |
away | +0.029 |
choice | +0.029 |
ook | +0.027 |
ively | +0.026 |
Neg logit contributions | |
sauce | -0.031 |
gum | -0.031 |
greens | -0.028 |
ink | -0.028 |
glasses | -0.028 |
Token | tree |
Token ID | 5509 |
Feature activation | +2.45e-02 |
Pos logit contributions | |
decision | +0.008 |
away | +0.008 |
choice | +0.007 |
tha | +0.006 |
ook | +0.006 |
Neg logit contributions | |
sauce | -0.012 |
gum | -0.012 |
greens | -0.011 |
glasses | -0.011 |
ink | -0.011 |
Token | clim |
Token ID | 5424 |
Feature activation | +6.86e-02 |
Pos logit contributions | |
decision | +0.035 |
choice | +0.034 |
away | +0.033 |
ook | +0.028 |
taxi | +0.028 |
Neg logit contributions | |
sauce | -0.046 |
gum | -0.045 |
greens | -0.044 |
glasses | -0.043 |
subtract | -0.042 |
Token | ber |
Token ID | 527 |
Feature activation | +5.07e-04 |
Pos logit contributions | |
decision | +0.120 |
choice | +0.110 |
doorstep | +0.095 |
away | +0.095 |
hotel | +0.093 |
Neg logit contributions | |
cr | -0.101 |
elt | -0.098 |
gloss | -0.092 |
pads | -0.091 |
o | -0.088 |
Token | . |
Token ID | 13 |
Feature activation | -4.42e-03 |
Pos logit contributions | |
decision | +0.001 |
choice | +0.001 |
away | +0.001 |
forward | +0.001 |
taxi | +0.001 |
Neg logit contributions | |
sauce | -0.001 |
greens | -0.001 |
gum | -0.001 |
fiber | -0.001 |
ink | -0.001 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -3.03e-02 |
Pos logit contributions | |
sauce | +0.006 |
gum | +0.006 |
greens | +0.006 |
glasses | +0.006 |
ink | +0.006 |
Neg logit contributions | |
decision | -0.008 |
choice | -0.008 |
away | -0.008 |
ook | -0.007 |
taxi | -0.007 |
Token | Once |
Token ID | 7454 |
Feature activation | +2.35e-02 |
Pos logit contributions | |
sauce | +0.043 |
gum | +0.042 |
greens | +0.040 |
glasses | +0.039 |
subtract | +0.038 |
Neg logit contributions | |
decision | -0.057 |
choice | -0.055 |
away | -0.054 |
ook | -0.049 |
taxi | -0.048 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
away | +0.045 |
decision | +0.044 |
choice | +0.042 |
doorstep | +0.041 |
checkpoint | +0.040 |
Neg logit contributions | |
sauce | -0.029 |
gum | -0.028 |
triangles | -0.027 |
Ben | -0.027 |
ink | -0.027 |
Token | a |
Token ID | 257 |
Feature activation | +2.50e-03 |
Pos logit contributions | |
greens | +0.013 |
gloss | +0.013 |
triangles | +0.012 |
sauce | +0.012 |
mascara | +0.012 |
Neg logit contributions | |
exactly | -0.008 |
security | -0.008 |
home | -0.008 |
taxi | -0.007 |
decision | -0.007 |
Token | big |
Token ID | 1263 |
Feature activation | -5.11e-03 |
Pos logit contributions | |
decision | +0.004 |
choice | +0.004 |
away | +0.004 |
taxi | +0.003 |
ook | +0.003 |
Neg logit contributions | |
sauce | -0.004 |
gum | -0.004 |
greens | -0.004 |
subtract | -0.004 |
pads | -0.004 |
Token | bucket |
Token ID | 19236 |
Feature activation | -9.52e-03 |
Pos logit contributions | |
sauce | +0.009 |
gum | +0.008 |
greens | +0.008 |
glasses | +0.008 |
ink | +0.008 |
Neg logit contributions | |
decision | -0.008 |
away | -0.008 |
choice | -0.008 |
ook | -0.007 |
taxi | -0.007 |
Token | of |
Token ID | 286 |
Feature activation | -1.87e-02 |
Pos logit contributions | |
greens | +0.021 |
sauce | +0.021 |
gum | +0.021 |
subtract | +0.020 |
glasses | +0.020 |
Neg logit contributions | |
decision | -0.010 |
away | -0.009 |
choice | -0.009 |
ook | -0.008 |
taxi | -0.008 |
Token | so |
Token ID | 523 |
Feature activation | -1.46e-02 |
Pos logit contributions | |
gloss | +0.023 |
subtract | +0.023 |
pads | +0.022 |
greens | +0.022 |
gum | +0.022 |
Neg logit contributions | |
choice | -0.037 |
decision | -0.036 |
away | -0.035 |
taxi | -0.032 |
ook | -0.030 |
Token | apy |
Token ID | 12826 |
Feature activation | -7.18e-02 |
Pos logit contributions | |
sauce | +0.027 |
greens | +0.026 |
gum | +0.026 |
glasses | +0.025 |
subtract | +0.025 |
Neg logit contributions | |
decision | -0.021 |
choice | -0.021 |
away | -0.020 |
ook | -0.018 |
taxi | -0.017 |
Token | water |
Token ID | 1660 |
Feature activation | -7.00e-03 |
Pos logit contributions | |
subtract | +0.105 |
Ol | +0.102 |
gloss | +0.101 |
triangles | +0.095 |
gum | +0.095 |
Neg logit contributions | |
choice | -0.125 |
decision | -0.120 |
away | -0.119 |
taxi | -0.105 |
promise | -0.098 |
Token | . |
Token ID | 13 |
Feature activation | -7.34e-03 |
Pos logit contributions | |
subtract | +0.016 |
greens | +0.015 |
Ol | +0.015 |
fiber | +0.015 |
gum | +0.015 |
Neg logit contributions | |
away | -0.007 |
decision | -0.007 |
choice | -0.006 |
taxi | -0.006 |
forward | -0.005 |
Token | She |
Token ID | 1375 |
Feature activation | -2.01e-03 |
Pos logit contributions | |
gum | +0.013 |
sauce | +0.012 |
gloss | +0.012 |
Ol | +0.012 |
ink | +0.011 |
Neg logit contributions | |
decision | -0.012 |
choice | -0.011 |
away | -0.010 |
ook | -0.010 |
doorstep | -0.010 |
Token | said |
Token ID | 531 |
Feature activation | -8.43e-03 |
Pos logit contributions | |
gum | +0.003 |
greens | +0.003 |
sauce | +0.003 |
subtract | +0.003 |
glasses | +0.003 |
Neg logit contributions | |
decision | -0.004 |
choice | -0.004 |
away | -0.003 |
ook | -0.003 |
taxi | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | -2.99e-02 |
Pos logit contributions | |
sauce | +0.018 |
greens | +0.018 |
gum | +0.018 |
pads | +0.018 |
glasses | +0.017 |
Neg logit contributions | |
decision | -0.009 |
away | -0.009 |
choice | -0.009 |
taxi | -0.007 |
ook | -0.007 |
Token | into |
Token ID | 656 |
Feature activation | -1.62e-02 |
Pos logit contributions | |
decision | +0.003 |
choice | +0.003 |
away | +0.003 |
ook | +0.003 |
taxi | +0.002 |
Neg logit contributions | |
sauce | -0.005 |
gum | -0.005 |
greens | -0.005 |
glasses | -0.005 |
ink | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -9.42e-03 |
Pos logit contributions | |
pads | +0.033 |
gloss | +0.033 |
gum | +0.031 |
sauce | +0.031 |
fiber | +0.030 |
Neg logit contributions | |
decision | -0.021 |
choice | -0.020 |
away | -0.019 |
purchase | -0.016 |
taxi | -0.015 |
Token | bucket |
Token ID | 19236 |
Feature activation | -9.16e-03 |
Pos logit contributions | |
subtract | +0.015 |
Ol | +0.014 |
sauce | +0.014 |
fiber | +0.014 |
gloss | +0.014 |
Neg logit contributions | |
decision | -0.016 |
choice | -0.015 |
away | -0.014 |
taxi | -0.014 |
home | -0.013 |
Token | of |
Token ID | 286 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
gum | +0.020 |
sauce | +0.020 |
Ol | +0.019 |
fiber | +0.019 |
subtract | +0.019 |
Neg logit contributions | |
decision | -0.010 |
away | -0.009 |
choice | -0.009 |
ook | -0.008 |
forward | -0.007 |
Token | so |
Token ID | 523 |
Feature activation | -9.37e-03 |
Pos logit contributions | |
gloss | +0.020 |
subtract | +0.019 |
pads | +0.019 |
Ol | +0.019 |
storms | +0.018 |
Neg logit contributions | |
choice | -0.021 |
decision | -0.021 |
away | -0.020 |
taxi | -0.020 |
hotel | -0.018 |
Token | apy |
Token ID | 12826 |
Feature activation | -7.12e-02 |
Pos logit contributions | |
sauce | +0.019 |
gum | +0.018 |
glasses | +0.018 |
greens | +0.017 |
pads | +0.017 |
Neg logit contributions | |
decision | -0.012 |
away | -0.012 |
choice | -0.012 |
ook | -0.010 |
forward | -0.010 |
Token | water |
Token ID | 1660 |
Feature activation | -1.75e-02 |
Pos logit contributions | |
subtract | +0.119 |
Ol | +0.113 |
gloss | +0.108 |
blond | +0.105 |
gum | +0.101 |
Neg logit contributions | |
away | -0.110 |
choice | -0.109 |
decision | -0.104 |
taxi | -0.088 |
forward | -0.087 |
Token | and |
Token ID | 290 |
Feature activation | -1.62e-02 |
Pos logit contributions | |
subtract | +0.037 |
gum | +0.036 |
Ol | +0.036 |
sauce | +0.036 |
fiber | +0.035 |
Neg logit contributions | |
decision | -0.020 |
away | -0.019 |
choice | -0.018 |
forward | -0.016 |
taxi | -0.015 |
Token | started |
Token ID | 2067 |
Feature activation | -2.84e-02 |
Pos logit contributions | |
gum | +0.027 |
subtract | +0.027 |
Ol | +0.027 |
gloss | +0.026 |
mascara | +0.026 |
Neg logit contributions | |
decision | -0.023 |
choice | -0.022 |
away | -0.021 |
sadly | -0.019 |
exactly | -0.019 |
Token | turning |
Token ID | 6225 |
Feature activation | -2.46e-02 |
Pos logit contributions | |
fiber | +0.057 |
gum | +0.056 |
sauce | +0.055 |
gloss | +0.054 |
subtract | +0.054 |
Neg logit contributions | |
decision | -0.033 |
away | -0.029 |
choice | -0.029 |
home | -0.028 |
ook | -0.026 |
Token | around |
Token ID | 1088 |
Feature activation | +7.08e-03 |
Pos logit contributions | |
subtract | +0.053 |
gloss | +0.051 |
sauce | +0.051 |
fiber | +0.051 |
gum | +0.050 |
Neg logit contributions | |
away | -0.027 |
forward | -0.025 |
decision | -0.025 |
choice | -0.023 |
taxi | -0.020 |
Token | it |
Token ID | 340 |
Feature activation | -1.93e-02 |
Pos logit contributions | |
subtract | +0.010 |
sauce | +0.010 |
fiber | +0.010 |
pill | +0.010 |
smoked | +0.010 |
Neg logit contributions | |
decision | -0.004 |
away | -0.004 |
choice | -0.004 |
forward | -0.004 |
up | -0.004 |
Token | with |
Token ID | 351 |
Feature activation | +1.59e-02 |
Pos logit contributions | |
sauce | +0.043 |
gum | +0.042 |
greens | +0.042 |
fiber | +0.041 |
glasses | +0.041 |
Neg logit contributions | |
decision | -0.020 |
away | -0.019 |
choice | -0.019 |
forward | -0.015 |
exactly | -0.014 |
Token | lots |
Token ID | 6041 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
choice | +0.013 |
home | +0.012 |
forward | +0.011 |
exactly | +0.011 |
tightly | +0.010 |
Neg logit contributions | |
storms | -0.035 |
pill | -0.035 |
Indians | -0.033 |
Tigers | -0.033 |
forest | -0.033 |
Token | of |
Token ID | 286 |
Feature activation | -2.45e-02 |
Pos logit contributions | |
sauce | +0.035 |
gum | +0.034 |
greens | +0.034 |
glasses | +0.033 |
Ol | +0.033 |
Neg logit contributions | |
decision | -0.008 |
away | -0.008 |
choice | -0.007 |
ook | -0.004 |
taxi | -0.004 |
Token | so |
Token ID | 523 |
Feature activation | -1.11e-02 |
Pos logit contributions | |
storms | +0.035 |
Indians | +0.034 |
pads | +0.034 |
tern | +0.032 |
Ol | +0.031 |
Neg logit contributions | |
choice | -0.042 |
decision | -0.039 |
away | -0.038 |
of | -0.034 |
taxi | -0.034 |
Token | apy |
Token ID | 12826 |
Feature activation | -7.09e-02 |
Pos logit contributions | |
sauce | +0.022 |
gum | +0.021 |
greens | +0.021 |
glasses | +0.020 |
pads | +0.020 |
Neg logit contributions | |
decision | -0.015 |
choice | -0.014 |
away | -0.014 |
ook | -0.012 |
taxi | -0.012 |
Token | bubbles |
Token ID | 25037 |
Feature activation | -2.21e-02 |
Pos logit contributions | |
Ol | +0.121 |
Indians | +0.117 |
Gary | +0.114 |
subtract | +0.113 |
ccoli | +0.110 |
Neg logit contributions | |
choice | -0.089 |
away | -0.082 |
decision | -0.079 |
of | -0.079 |
- | -0.072 |
Token | . |
Token ID | 13 |
Feature activation | -6.68e-03 |
Pos logit contributions | |
Ol | +0.052 |
fiber | +0.052 |
subtract | +0.051 |
blond | +0.050 |
gum | +0.049 |
Neg logit contributions | |
away | -0.016 |
decision | -0.015 |
forward | -0.014 |
choice | -0.013 |
under | -0.012 |
Token | Finally |
Token ID | 9461 |
Feature activation | -1.08e-02 |
Pos logit contributions | |
sauce | +0.013 |
gum | +0.012 |
ink | +0.012 |
Ol | +0.012 |
triangles | +0.012 |
Neg logit contributions | |
decision | -0.008 |
taxi | -0.008 |
choice | -0.007 |
doorstep | -0.007 |
azel | -0.007 |
Token | she |
Token ID | 673 |
Feature activation | -1.11e-02 |
Pos logit contributions | |
sauce | +0.022 |
gum | +0.021 |
greens | +0.021 |
glasses | +0.020 |
pads | +0.020 |
Neg logit contributions | |
decision | -0.014 |
choice | -0.014 |
away | -0.013 |
ook | -0.011 |
taxi | -0.011 |
Token | polished |
Token ID | 23895 |
Feature activation | -5.43e-03 |
Pos logit contributions | |
greens | +0.019 |
subtract | +0.018 |
gum | +0.018 |
lenses | +0.017 |
blond | +0.017 |
Neg logit contributions | |
choice | -0.015 |
away | -0.015 |
decision | -0.015 |
ook | -0.015 |
promise | -0.014 |
Token | don |
Token ID | 836 |
Feature activation | +5.25e-03 |
Pos logit contributions | |
decision | +0.022 |
choice | +0.021 |
away | +0.021 |
ook | +0.018 |
taxi | +0.018 |
Neg logit contributions | |
sauce | -0.020 |
gum | -0.019 |
greens | -0.018 |
glasses | -0.018 |
ink | -0.018 |
Token | 't |
Token ID | 470 |
Feature activation | +1.58e-02 |
Pos logit contributions | |
decision | +0.010 |
choice | +0.010 |
away | +0.010 |
ook | +0.008 |
checkpoint | +0.008 |
Neg logit contributions | |
gum | -0.008 |
sauce | -0.008 |
greens | -0.007 |
ink | -0.007 |
glasses | -0.007 |
Token | know |
Token ID | 760 |
Feature activation | +3.40e-04 |
Pos logit contributions | |
away | +0.019 |
decision | +0.019 |
choice | +0.018 |
ook | +0.015 |
forward | +0.014 |
Neg logit contributions | |
gum | -0.034 |
sauce | -0.034 |
greens | -0.033 |
triangles | -0.032 |
glasses | -0.032 |
Token | ," |
Token ID | 553 |
Feature activation | -4.94e-03 |
Pos logit contributions | |
decision | +0.000 |
choice | +0.000 |
away | +0.000 |
taxi | +0.000 |
ook | +0.000 |
Neg logit contributions | |
sauce | -0.001 |
gum | -0.001 |
greens | -0.001 |
glasses | -0.001 |
subtract | -0.001 |
Token | his |
Token ID | 465 |
Feature activation | -1.51e-02 |
Pos logit contributions | |
sauce | +0.009 |
pads | +0.009 |
gum | +0.009 |
greens | +0.009 |
triangles | +0.009 |
Neg logit contributions | |
decision | -0.007 |
choice | -0.007 |
away | -0.007 |
taxi | -0.006 |
tha | -0.006 |
Token | sister |
Token ID | 6621 |
Feature activation | -6.97e-02 |
Pos logit contributions | |
sauce | +0.029 |
gum | +0.029 |
greens | +0.028 |
pads | +0.027 |
subtract | +0.027 |
Neg logit contributions | |
decision | -0.021 |
choice | -0.021 |
away | -0.019 |
ook | -0.017 |
taxi | -0.016 |
Token | replied |
Token ID | 8712 |
Feature activation | -1.43e-03 |
Pos logit contributions | |
sauce | +0.107 |
gum | +0.106 |
triangles | +0.100 |
greens | +0.099 |
fiber | +0.096 |
Neg logit contributions | |
decision | -0.133 |
choice | -0.131 |
sadly | -0.114 |
exactly | -0.112 |
promise | -0.111 |
Token | . |
Token ID | 13 |
Feature activation | -2.69e-02 |
Pos logit contributions | |
sauce | +0.003 |
gum | +0.003 |
greens | +0.003 |
pads | +0.003 |
subtract | +0.003 |
Neg logit contributions | |
decision | -0.002 |
choice | -0.002 |
away | -0.002 |
ook | -0.001 |
forward | -0.001 |
Token |
|
Token ID | 198 |
Feature activation | -1.53e-03 |
Pos logit contributions | |
sauce | +0.053 |
gum | +0.051 |
ink | +0.051 |
pens | +0.049 |
gloss | +0.048 |
Neg logit contributions | |
tha | -0.031 |
decision | -0.030 |
curiously | -0.030 |
taxi | -0.030 |
ook | -0.029 |
Token |
|
Token ID | 198 |
Feature activation | -2.41e-03 |
Pos logit contributions | |
sauce | +0.002 |
gum | +0.002 |
greens | +0.002 |
glasses | +0.002 |
ink | +0.002 |
Neg logit contributions | |
decision | -0.003 |
choice | -0.003 |
away | -0.003 |
ook | -0.003 |
taxi | -0.003 |
Token | They |
Token ID | 2990 |
Feature activation | -5.79e-03 |
Pos logit contributions | |
sauce | +0.003 |
gum | +0.003 |
greens | +0.003 |
glasses | +0.003 |
subtract | +0.003 |
Neg logit contributions | |
decision | -0.005 |
choice | -0.005 |
away | -0.004 |
ook | -0.004 |
checkpoint | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +8.46e-03 |
Pos logit contributions | |
sauce | +0.013 |
pads | +0.013 |
greens | +0.013 |
gum | +0.013 |
subtract | +0.012 |
Neg logit contributions | |
decision | -0.011 |
choice | -0.011 |
away | -0.011 |
forward | -0.010 |
taxi | -0.010 |
Token | cloth |
Token ID | 16270 |
Feature activation | +6.06e-03 |
Pos logit contributions | |
decision | +0.017 |
choice | +0.016 |
away | +0.016 |
taxi | +0.015 |
ook | +0.014 |
Neg logit contributions | |
sauce | -0.011 |
greens | -0.011 |
pads | -0.010 |
gum | -0.010 |
subtract | -0.010 |
Token | and |
Token ID | 290 |
Feature activation | +2.86e-03 |
Pos logit contributions | |
decision | +0.009 |
choice | +0.009 |
away | +0.009 |
ook | +0.007 |
taxi | +0.007 |
Neg logit contributions | |
sauce | -0.011 |
gum | -0.011 |
greens | -0.010 |
glasses | -0.010 |
ink | -0.010 |
Token | some |
Token ID | 617 |
Feature activation | -1.80e-02 |
Pos logit contributions | |
choice | +0.004 |
decision | +0.004 |
away | +0.004 |
ook | +0.004 |
taxi | +0.004 |
Neg logit contributions | |
greens | -0.004 |
sauce | -0.004 |
gum | -0.004 |
triangles | -0.004 |
subtract | -0.004 |
Token | so |
Token ID | 523 |
Feature activation | -2.51e-03 |
Pos logit contributions | |
storms | +0.023 |
pads | +0.022 |
greens | +0.022 |
subtract | +0.022 |
joked | +0.022 |
Neg logit contributions | |
choice | -0.033 |
decision | -0.031 |
away | -0.030 |
taxi | -0.028 |
promise | -0.026 |
Token | apy |
Token ID | 12826 |
Feature activation | -6.79e-02 |
Pos logit contributions | |
sauce | +0.004 |
greens | +0.004 |
gum | +0.004 |
glasses | +0.004 |
triangles | +0.004 |
Neg logit contributions | |
decision | -0.004 |
choice | -0.004 |
away | -0.004 |
ook | -0.003 |
taxi | -0.003 |
Token | water |
Token ID | 1660 |
Feature activation | -1.01e-02 |
Pos logit contributions | |
blond | +0.099 |
subtract | +0.099 |
triangles | +0.097 |
pads | +0.097 |
pill | +0.096 |
Neg logit contributions | |
choice | -0.104 |
away | -0.097 |
decision | -0.092 |
taxi | -0.082 |
of | -0.082 |
Token | and |
Token ID | 290 |
Feature activation | -6.62e-03 |
Pos logit contributions | |
subtract | +0.022 |
Ol | +0.022 |
greens | +0.022 |
triangles | +0.021 |
gum | +0.021 |
Neg logit contributions | |
decision | -0.009 |
choice | -0.009 |
away | -0.009 |
taxi | -0.007 |
home | -0.007 |
Token | got |
Token ID | 1392 |
Feature activation | -9.09e-03 |
Pos logit contributions | |
Ol | +0.011 |
greens | +0.010 |
pill | +0.010 |
triangles | +0.010 |
storms | +0.010 |
Neg logit contributions | |
choice | -0.009 |
decision | -0.009 |
away | -0.008 |
exactly | -0.008 |
sadly | -0.007 |
Token | to |
Token ID | 284 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
greens | +0.016 |
pads | +0.016 |
sauce | +0.016 |
fiber | +0.015 |
subtract | +0.015 |
Neg logit contributions | |
choice | -0.012 |
decision | -0.012 |
away | -0.012 |
forward | -0.011 |
exactly | -0.011 |
Token | work |
Token ID | 670 |
Feature activation | -4.27e-03 |
Pos logit contributions | |
decision | +0.024 |
choice | +0.024 |
away | +0.023 |
purchase | +0.019 |
ook | +0.019 |
Neg logit contributions | |
sauce | -0.029 |
gum | -0.028 |
pads | -0.028 |
greens | -0.028 |
glasses | -0.027 |
Token | lost |
Token ID | 2626 |
Feature activation | -6.19e-03 |
Pos logit contributions | |
decision | +0.031 |
choice | +0.030 |
away | +0.030 |
ook | +0.027 |
taxi | +0.027 |
Neg logit contributions | |
sauce | -0.021 |
gum | -0.020 |
greens | -0.018 |
glasses | -0.018 |
ink | -0.018 |
Token | my |
Token ID | 616 |
Feature activation | +9.20e-03 |
Pos logit contributions | |
sauce | +0.013 |
greens | +0.012 |
gum | +0.012 |
subtract | +0.012 |
pads | +0.012 |
Neg logit contributions | |
decision | -0.008 |
away | -0.008 |
choice | -0.007 |
taxi | -0.006 |
ook | -0.006 |
Token | doll |
Token ID | 3654 |
Feature activation | +2.74e-03 |
Pos logit contributions | |
decision | +0.014 |
choice | +0.014 |
away | +0.014 |
ook | +0.012 |
taxi | +0.011 |
Neg logit contributions | |
sauce | -0.016 |
gum | -0.016 |
greens | -0.015 |
glasses | -0.015 |
ink | -0.015 |
Token | ," |
Token ID | 553 |
Feature activation | -1.65e-03 |
Pos logit contributions | |
decision | +0.004 |
choice | +0.004 |
away | +0.004 |
ook | +0.003 |
checkpoint | +0.003 |
Neg logit contributions | |
sauce | -0.005 |
gum | -0.005 |
glasses | -0.005 |
greens | -0.005 |
ink | -0.005 |
Token | his |
Token ID | 465 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
pads | +0.004 |
greens | +0.003 |
lenses | +0.003 |
aces | +0.003 |
pill | +0.003 |
Neg logit contributions | |
choice | -0.002 |
decision | -0.002 |
away | -0.002 |
taxi | -0.002 |
promise | -0.001 |
Token | sister |
Token ID | 6621 |
Feature activation | -6.79e-02 |
Pos logit contributions | |
sauce | +0.031 |
gum | +0.031 |
greens | +0.030 |
glasses | +0.029 |
ink | +0.029 |
Neg logit contributions | |
decision | -0.026 |
choice | -0.024 |
away | -0.024 |
ook | -0.021 |
taxi | -0.020 |
Token | replied |
Token ID | 8712 |
Feature activation | +5.66e-03 |
Pos logit contributions | |
gum | +0.105 |
sauce | +0.105 |
greens | +0.100 |
triangles | +0.100 |
pads | +0.099 |
Neg logit contributions | |
choice | -0.125 |
decision | -0.125 |
sadly | -0.109 |
promise | -0.106 |
exactly | -0.102 |
Token | . |
Token ID | 13 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
decision | +0.008 |
choice | +0.007 |
away | +0.007 |
ook | +0.006 |
taxi | +0.006 |
Neg logit contributions | |
sauce | -0.011 |
gum | -0.011 |
greens | -0.011 |
pads | -0.011 |
glasses | -0.010 |
Token |
|
Token ID | 198 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
pads | +0.024 |
fiber | +0.024 |
dots | +0.024 |
sauce | +0.024 |
gum | +0.023 |
Neg logit contributions | |
taxi | -0.010 |
where | -0.010 |
ook | -0.010 |
sadly | -0.010 |
tha | -0.009 |
Token |
|
Token ID | 198 |
Feature activation | +1.27e-02 |
Pos logit contributions | |
decision | +0.022 |
choice | +0.022 |
away | +0.021 |
ook | +0.019 |
taxi | +0.019 |
Neg logit contributions | |
sauce | -0.016 |
gum | -0.015 |
greens | -0.015 |
glasses | -0.014 |
ink | -0.014 |
Token | " |
Token ID | 1 |
Feature activation | -7.14e-03 |
Pos logit contributions | |
decision | +0.022 |
away | +0.021 |
choice | +0.021 |
ook | +0.018 |
taxi | +0.018 |
Neg logit contributions | |
sauce | -0.020 |
gum | -0.020 |
greens | -0.019 |
glasses | -0.019 |
ink | -0.018 |
Token | cheese |
Token ID | 9891 |
Feature activation | +9.75e-03 |
Pos logit contributions | |
sauce | +0.034 |
gum | +0.033 |
greens | +0.032 |
glasses | +0.031 |
ink | +0.031 |
Neg logit contributions | |
decision | -0.027 |
choice | -0.025 |
away | -0.025 |
ook | -0.022 |
taxi | -0.022 |
Token | !" |
Token ID | 2474 |
Feature activation | -6.55e-03 |
Pos logit contributions | |
choice | +0.011 |
decision | +0.010 |
away | +0.009 |
taxi | +0.007 |
promise | +0.007 |
Neg logit contributions | |
subtract | -0.022 |
blond | -0.021 |
Ol | -0.021 |
greens | -0.021 |
blonde | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | +3.70e-03 |
Pos logit contributions | |
sauce | +0.011 |
gum | +0.011 |
subtract | +0.011 |
pads | +0.011 |
fiber | +0.011 |
Neg logit contributions | |
taxi | -0.008 |
choice | -0.008 |
sadly | -0.007 |
decision | -0.007 |
away | -0.007 |
Token |
|
Token ID | 198 |
Feature activation | +1.87e-03 |
Pos logit contributions | |
away | +0.006 |
decision | +0.006 |
choice | +0.006 |
ook | +0.005 |
taxi | +0.005 |
Neg logit contributions | |
sauce | -0.006 |
gum | -0.006 |
greens | -0.006 |
glasses | -0.006 |
pads | -0.006 |
Token | D |
Token ID | 35 |
Feature activation | -5.37e-04 |
Pos logit contributions | |
away | +0.003 |
decision | +0.003 |
choice | +0.003 |
ook | +0.003 |
taxi | +0.003 |
Neg logit contributions | |
sauce | -0.003 |
gum | -0.003 |
greens | -0.003 |
glasses | -0.003 |
ink | -0.003 |
Token | ot |
Token ID | 313 |
Feature activation | -6.77e-02 |
Pos logit contributions | |
sauce | +0.001 |
gum | +0.001 |
ink | +0.001 |
greens | +0.001 |
glasses | +0.001 |
Neg logit contributions | |
decision | -0.001 |
choice | -0.001 |
away | -0.001 |
taxi | -0.001 |
checkpoint | -0.001 |
Token | sniff |
Token ID | 26300 |
Feature activation | -8.75e-03 |
Pos logit contributions | |
gum | +0.114 |
fiber | +0.111 |
subtract | +0.108 |
greens | +0.107 |
sauce | +0.105 |
Neg logit contributions | |
away | -0.088 |
decision | -0.087 |
choice | -0.087 |
sadly | -0.086 |
ie | -0.084 |
Token | ed |
Token ID | 276 |
Feature activation | +3.62e-03 |
Pos logit contributions | |
sauce | +0.016 |
subtract | +0.016 |
gum | +0.016 |
greens | +0.015 |
glasses | +0.014 |
Neg logit contributions | |
away | -0.013 |
decision | -0.013 |
choice | -0.012 |
ook | -0.011 |
ively | -0.011 |
Token | the |
Token ID | 262 |
Feature activation | +7.47e-03 |
Pos logit contributions | |
away | +0.004 |
decision | +0.004 |
choice | +0.004 |
taxi | +0.003 |
ook | +0.003 |
Neg logit contributions | |
sauce | -0.008 |
gum | -0.008 |
subtract | -0.008 |
pads | -0.007 |
greens | -0.007 |
Token | air |
Token ID | 1633 |
Feature activation | -2.65e-03 |
Pos logit contributions | |
decision | +0.014 |
choice | +0.014 |
away | +0.012 |
taxi | +0.012 |
promise | +0.011 |
Neg logit contributions | |
subtract | -0.011 |
joked | -0.010 |
greens | -0.010 |
Ol | -0.010 |
gloss | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | -1.61e-02 |
Pos logit contributions | |
Ol | +0.006 |
subtract | +0.006 |
fiber | +0.006 |
gum | +0.006 |
pill | +0.006 |
Neg logit contributions | |
decision | -0.003 |
away | -0.003 |
choice | -0.003 |
sadly | -0.002 |
rily | -0.002 |
Token | his |
Token ID | 465 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
decision | +0.000 |
choice | +0.000 |
away | +0.000 |
ook | +0.000 |
taxi | +0.000 |
Neg logit contributions | |
sauce | -0.001 |
greens | -0.001 |
gum | -0.001 |
pads | -0.001 |
subtract | -0.001 |
Token | bubble |
Token ID | 14310 |
Feature activation | -1.76e-02 |
Pos logit contributions | |
decision | +0.020 |
choice | +0.019 |
away | +0.017 |
taxi | +0.016 |
doorstep | +0.016 |
Neg logit contributions | |
subtract | -0.018 |
greens | -0.018 |
sauce | -0.018 |
gum | -0.018 |
Ol | -0.017 |
Token | maker |
Token ID | 16009 |
Feature activation | -7.40e-03 |
Pos logit contributions | |
subtract | +0.031 |
gum | +0.031 |
sauce | +0.030 |
greens | +0.030 |
fiber | +0.029 |
Neg logit contributions | |
decision | -0.026 |
away | -0.025 |
choice | -0.024 |
ook | -0.022 |
taxi | -0.021 |
Token | with |
Token ID | 351 |
Feature activation | +1.72e-02 |
Pos logit contributions | |
sauce | +0.015 |
gum | +0.015 |
fiber | +0.015 |
greens | +0.015 |
Ol | +0.015 |
Neg logit contributions | |
decision | -0.009 |
choice | -0.008 |
away | -0.008 |
ook | -0.007 |
doorstep | -0.007 |
Token | so |
Token ID | 523 |
Feature activation | -3.72e-03 |
Pos logit contributions | |
choice | +0.025 |
decision | +0.023 |
taxi | +0.020 |
away | +0.020 |
exactly | +0.020 |
Neg logit contributions | |
gloss | -0.029 |
subtract | -0.029 |
Ol | -0.028 |
pads | -0.028 |
fiber | -0.028 |
Token | apy |
Token ID | 12826 |
Feature activation | -6.75e-02 |
Pos logit contributions | |
sauce | +0.007 |
gum | +0.007 |
greens | +0.007 |
glasses | +0.007 |
ink | +0.007 |
Neg logit contributions | |
decision | -0.005 |
choice | -0.005 |
away | -0.005 |
ook | -0.004 |
taxi | -0.004 |
Token | water |
Token ID | 1660 |
Feature activation | -1.41e-02 |
Pos logit contributions | |
subtract | +0.109 |
Ol | +0.107 |
please | +0.106 |
blond | +0.105 |
Mic | +0.101 |
Neg logit contributions | |
choice | -0.093 |
of | -0.086 |
away | -0.084 |
decision | -0.080 |
- | -0.076 |
Token | and |
Token ID | 290 |
Feature activation | -1.19e-02 |
Pos logit contributions | |
subtract | +0.032 |
Ol | +0.032 |
fiber | +0.031 |
greens | +0.031 |
blond | +0.030 |
Neg logit contributions | |
decision | -0.013 |
away | -0.012 |
choice | -0.012 |
taxi | -0.010 |
forward | -0.010 |
Token | blew |
Token ID | 17948 |
Feature activation | -1.48e-02 |
Pos logit contributions | |
Ol | +0.018 |
mascara | +0.017 |
subtract | +0.017 |
gum | +0.016 |
lenses | +0.016 |
Neg logit contributions | |
choice | -0.017 |
decision | -0.017 |
exactly | -0.015 |
taxi | -0.014 |
away | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
fiber | +0.032 |
subtract | +0.032 |
Ol | +0.032 |
gloss | +0.031 |
smoked | +0.030 |
Neg logit contributions | |
away | -0.015 |
forward | -0.015 |
decision | -0.013 |
choice | -0.013 |
up | -0.012 |
Token | Bub |
Token ID | 19645 |
Feature activation | +1.40e-03 |
Pos logit contributions | |
ink | +0.027 |
sauce | +0.027 |
gum | +0.027 |
pens | +0.026 |
glasses | +0.026 |
Neg logit contributions | |
decision | -0.017 |
ook | -0.016 |
taxi | -0.016 |
choice | -0.016 |
doorstep | -0.014 |
Token | are |
Token ID | 389 |
Feature activation | -5.07e-03 |
Pos logit contributions | |
decision | +0.008 |
choice | +0.008 |
promise | +0.007 |
ook | +0.007 |
away | +0.007 |
Neg logit contributions | |
greens | -0.010 |
lenses | -0.009 |
subtract | -0.009 |
sauce | -0.008 |
pads | -0.008 |
Token | you |
Token ID | 345 |
Feature activation | -1.12e-02 |
Pos logit contributions | |
sauce | +0.009 |
greens | +0.009 |
gum | +0.009 |
glasses | +0.009 |
pads | +0.008 |
Neg logit contributions | |
decision | -0.008 |
choice | -0.007 |
away | -0.007 |
ook | -0.006 |
taxi | -0.006 |
Token | doing |
Token ID | 1804 |
Feature activation | -1.97e-02 |
Pos logit contributions | |
greens | +0.019 |
pads | +0.018 |
sauce | +0.017 |
fiber | +0.017 |
glasses | +0.017 |
Neg logit contributions | |
decision | -0.018 |
choice | -0.017 |
away | -0.015 |
exactly | -0.015 |
promise | -0.015 |
Token | ?" |
Token ID | 1701 |
Feature activation | -5.00e-04 |
Pos logit contributions | |
greens | +0.039 |
pads | +0.039 |
sauce | +0.038 |
gum | +0.037 |
glasses | +0.037 |
Neg logit contributions | |
decision | -0.026 |
choice | -0.025 |
away | -0.024 |
ook | -0.020 |
taxi | -0.019 |
Token | her |
Token ID | 607 |
Feature activation | -1.99e-02 |
Pos logit contributions | |
pads | +0.001 |
sauce | +0.001 |
greens | +0.001 |
gum | +0.001 |
cr | +0.001 |
Neg logit contributions | |
decision | -0.001 |
choice | -0.001 |
away | -0.001 |
sadly | -0.001 |
taxi | -0.001 |
Token | sister |
Token ID | 6621 |
Feature activation | -6.74e-02 |
Pos logit contributions | |
sauce | +0.039 |
gum | +0.038 |
greens | +0.036 |
glasses | +0.036 |
ink | +0.035 |
Neg logit contributions | |
decision | -0.028 |
choice | -0.026 |
away | -0.026 |
ook | -0.022 |
taxi | -0.021 |
Token | asked |
Token ID | 1965 |
Feature activation | +1.35e-02 |
Pos logit contributions | |
fiber | +0.129 |
pill | +0.123 |
vitamins | +0.121 |
pads | +0.121 |
gum | +0.120 |
Neg logit contributions | |
sadly | -0.093 |
choice | -0.088 |
decision | -0.088 |
curiously | -0.084 |
promise | -0.079 |
Token | , |
Token ID | 11 |
Feature activation | -6.72e-03 |
Pos logit contributions | |
decision | +0.020 |
choice | +0.019 |
away | +0.019 |
ook | +0.016 |
taxi | +0.016 |
Neg logit contributions | |
sauce | -0.025 |
gum | -0.024 |
greens | -0.024 |
pads | -0.023 |
ink | -0.023 |
Token | coming |
Token ID | 2406 |
Feature activation | +1.08e-03 |
Pos logit contributions | |
fiber | +0.013 |
pads | +0.013 |
sauce | +0.012 |
gum | +0.012 |
pill | +0.012 |
Neg logit contributions | |
decision | -0.009 |
sadly | -0.008 |
curiously | -0.008 |
away | -0.008 |
taxi | -0.007 |
Token | into |
Token ID | 656 |
Feature activation | -5.84e-03 |
Pos logit contributions | |
decision | +0.002 |
choice | +0.002 |
away | +0.002 |
destination | +0.002 |
exactly | +0.002 |
Neg logit contributions | |
sauce | -0.001 |
glasses | -0.001 |
gum | -0.001 |
ink | -0.001 |
triangles | -0.001 |
Token | her |
Token ID | 607 |
Feature activation | +6.09e-03 |
Pos logit contributions | |
pads | +0.010 |
sauce | +0.010 |
gum | +0.010 |
greens | +0.010 |
gloss | +0.010 |
Neg logit contributions | |
decision | -0.009 |
choice | -0.009 |
away | -0.008 |
taxi | -0.007 |
ook | -0.007 |
Token | in |
Token ID | 287 |
Feature activation | +7.44e-03 |
Pos logit contributions | |
decision | +0.015 |
choice | +0.014 |
away | +0.013 |
forward | +0.012 |
exactly | +0.012 |
Neg logit contributions | |
sauce | -0.031 |
gum | -0.031 |
fiber | -0.030 |
subtract | -0.029 |
greens | -0.029 |
Token | the |
Token ID | 262 |
Feature activation | +3.31e-03 |
Pos logit contributions | |
decision | +0.011 |
choice | +0.011 |
away | +0.011 |
taxi | +0.009 |
ook | +0.009 |
Neg logit contributions | |
sauce | -0.013 |
gum | -0.013 |
greens | -0.013 |
pads | -0.012 |
gloss | -0.012 |
Token | warm |
Token ID | 5814 |
Feature activation | -2.24e-02 |
Pos logit contributions | |
decision | +0.006 |
choice | +0.006 |
away | +0.006 |
taxi | +0.005 |
hotel | +0.005 |
Neg logit contributions | |
subtract | -0.005 |
greens | -0.005 |
gum | -0.004 |
sauce | -0.004 |
triangles | -0.004 |
Token | , |
Token ID | 11 |
Feature activation | -1.07e-02 |
Pos logit contributions | |
gum | +0.027 |
subtract | +0.027 |
sauce | +0.027 |
greens | +0.026 |
Ol | +0.025 |
Neg logit contributions | |
decision | -0.046 |
away | -0.045 |
choice | -0.045 |
taxi | -0.039 |
hotel | -0.039 |
Token | so |
Token ID | 523 |
Feature activation | -3.69e-03 |
Pos logit contributions | |
subtract | +0.016 |
gum | +0.016 |
sauce | +0.016 |
greens | +0.015 |
fiber | +0.015 |
Neg logit contributions | |
decision | -0.018 |
away | -0.017 |
choice | -0.017 |
taxi | -0.015 |
forward | -0.015 |
Token | apy |
Token ID | 12826 |
Feature activation | -6.72e-02 |
Pos logit contributions | |
sauce | +0.007 |
gum | +0.006 |
greens | +0.006 |
glasses | +0.006 |
subtract | +0.006 |
Neg logit contributions | |
decision | -0.005 |
choice | -0.005 |
away | -0.005 |
ook | -0.005 |
taxi | -0.005 |
Token | water |
Token ID | 1660 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
subtract | +0.122 |
blond | +0.112 |
Ol | +0.109 |
gloss | +0.104 |
blonde | +0.104 |
Neg logit contributions | |
away | -0.095 |
choice | -0.095 |
decision | -0.088 |
hotel | -0.082 |
taxi | -0.079 |
Token | . |
Token ID | 13 |
Feature activation | -1.21e-02 |
Pos logit contributions | |
subtract | +0.023 |
Ol | +0.023 |
fiber | +0.022 |
gum | +0.022 |
greens | +0.021 |
Neg logit contributions | |
decision | -0.010 |
away | -0.010 |
choice | -0.009 |
taxi | -0.008 |
forward | -0.008 |
Token | She |
Token ID | 1375 |
Feature activation | -5.56e-03 |
Pos logit contributions | |
sauce | +0.021 |
gum | +0.020 |
triangles | +0.019 |
glasses | +0.019 |
greens | +0.019 |
Neg logit contributions | |
decision | -0.018 |
choice | -0.017 |
ook | -0.016 |
away | -0.016 |
taxi | -0.015 |
Token | spl |
Token ID | 4328 |
Feature activation | -1.68e-02 |
Pos logit contributions | |
subtract | +0.009 |
gum | +0.009 |
greens | +0.009 |
sauce | +0.009 |
fiber | +0.008 |
Neg logit contributions | |
decision | -0.008 |
choice | -0.008 |
away | -0.008 |
hotel | -0.007 |
ook | -0.007 |
Token | ashed |
Token ID | 5263 |
Feature activation | -2.82e-02 |
Pos logit contributions | |
gum | +0.026 |
sauce | +0.025 |
greens | +0.024 |
glasses | +0.024 |
subtract | +0.024 |
Neg logit contributions | |
away | -0.029 |
decision | -0.029 |
choice | -0.027 |
ook | -0.027 |
or | -0.025 |
Token | â |
Token ID | 22940 |
Feature activation | +1.26e-02 |
Pos logit contributions | |
taxi | +0.002 |
decision | +0.002 |
ook | +0.002 |
away | +0.002 |
choice | +0.002 |
Neg logit contributions | |
sauce | -0.002 |
gum | -0.002 |
greens | -0.002 |
ink | -0.002 |
fiber | -0.002 |
Token | € |
Token ID | 26391 |
Feature activation | +1.19e-02 |
Pos logit contributions | |
decision | +0.022 |
away | +0.021 |
choice | +0.021 |
tha | +0.019 |
ook | +0.018 |
Neg logit contributions | |
sauce | -0.020 |
gum | -0.020 |
glasses | -0.019 |
greens | -0.019 |
triangles | -0.018 |
Token |
|
Token ID | 198 |
Feature activation | -9.30e-03 |
Pos logit contributions | |
decision | +0.020 |
choice | +0.019 |
away | +0.018 |
ook | +0.018 |
tha | +0.018 |
Neg logit contributions | |
sauce | -0.020 |
gum | -0.018 |
greens | -0.018 |
ink | -0.018 |
glasses | -0.017 |
Token |
|
Token ID | 198 |
Feature activation | -6.63e-03 |
Pos logit contributions | |
sauce | +0.017 |
gum | +0.016 |
greens | +0.015 |
glasses | +0.015 |
subtract | +0.015 |
Neg logit contributions | |
decision | -0.014 |
choice | -0.013 |
away | -0.013 |
ook | -0.013 |
taxi | -0.012 |
Token | S |
Token ID | 50 |
Feature activation | +4.96e-03 |
Pos logit contributions | |
sauce | +0.012 |
gum | +0.011 |
greens | +0.011 |
ink | +0.011 |
glasses | +0.011 |
Neg logit contributions | |
away | -0.010 |
decision | -0.009 |
ook | -0.009 |
taxi | -0.009 |
choice | -0.009 |
Token | amy |
Token ID | 14814 |
Feature activation | -6.71e-02 |
Pos logit contributions | |
decision | +0.006 |
choice | +0.006 |
away | +0.006 |
ook | +0.005 |
tha | +0.005 |
Neg logit contributions | |
sauce | -0.010 |
gum | -0.010 |
greens | -0.010 |
glasses | -0.010 |
triangles | -0.010 |
Token | was |
Token ID | 373 |
Feature activation | -8.80e-03 |
Pos logit contributions | |
subtract | +0.129 |
gloss | +0.115 |
greens | +0.115 |
fiber | +0.114 |
gum | +0.114 |
Neg logit contributions | |
decision | -0.089 |
away | -0.082 |
choice | -0.080 |
hotel | -0.075 |
tha | -0.074 |
Token | so |
Token ID | 523 |
Feature activation | -2.29e-03 |
Pos logit contributions | |
fiber | +0.018 |
sauce | +0.018 |
greens | +0.018 |
gum | +0.018 |
Ol | +0.017 |
Neg logit contributions | |
decision | -0.011 |
choice | -0.010 |
forward | -0.010 |
exactly | -0.010 |
taxi | -0.009 |
Token | happy |
Token ID | 3772 |
Feature activation | +5.66e-03 |
Pos logit contributions | |
sauce | +0.005 |
gum | +0.005 |
greens | +0.004 |
ink | +0.004 |
glasses | +0.004 |
Neg logit contributions | |
decision | -0.003 |
choice | -0.003 |
away | -0.003 |
ook | -0.003 |
forward | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | -8.94e-03 |
Pos logit contributions | |
decision | +0.007 |
choice | +0.006 |
away | +0.006 |
exactly | +0.005 |
ook | +0.005 |
Neg logit contributions | |
greens | -0.012 |
sauce | -0.012 |
gum | -0.012 |
Ol | -0.012 |
subtract | -0.011 |
Token | He |
Token ID | 679 |
Feature activation | +7.70e-03 |
Pos logit contributions | |
sauce | +0.016 |
gum | +0.016 |
ink | +0.015 |
glasses | +0.015 |
greens | +0.015 |
Neg logit contributions | |
decision | -0.012 |
away | -0.011 |
ook | -0.011 |
choice | -0.011 |
taxi | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | -2.29e-02 |
Pos logit contributions | |
sauce | +0.021 |
gum | +0.021 |
greens | +0.020 |
triangles | +0.020 |
subtract | +0.020 |
Neg logit contributions | |
decision | -0.011 |
choice | -0.010 |
away | -0.010 |
forward | -0.008 |
taxi | -0.008 |
Token | " |
Token ID | 366 |
Feature activation | -1.36e-02 |
Pos logit contributions | |
sauce | +0.039 |
gum | +0.037 |
pads | +0.036 |
ink | +0.035 |
glasses | +0.035 |
Neg logit contributions | |
decision | -0.036 |
away | -0.034 |
choice | -0.033 |
forward | -0.032 |
taxi | -0.032 |
Token | My |
Token ID | 3666 |
Feature activation | +1.04e-02 |
Pos logit contributions | |
sauce | +0.026 |
glasses | +0.024 |
ink | +0.023 |
greens | +0.023 |
gum | +0.023 |
Neg logit contributions | |
away | -0.020 |
decision | -0.018 |
ively | -0.016 |
iring | -0.016 |
or | -0.016 |
Token | mum |
Token ID | 25682 |
Feature activation | -2.67e-02 |
Pos logit contributions | |
decision | +0.016 |
choice | +0.015 |
away | +0.013 |
taxi | +0.012 |
promise | +0.012 |
Neg logit contributions | |
sauce | -0.020 |
gum | -0.019 |
greens | -0.019 |
triangles | -0.018 |
subtract | -0.018 |
Token | 's |
Token ID | 338 |
Feature activation | -7.35e-03 |
Pos logit contributions | |
sauce | +0.050 |
greens | +0.049 |
gum | +0.049 |
triangles | +0.049 |
squares | +0.048 |
Neg logit contributions | |
decision | -0.039 |
choice | -0.038 |
promise | -0.032 |
away | -0.032 |
tha | -0.030 |
Token | poorly |
Token ID | 13455 |
Feature activation | -6.69e-02 |
Pos logit contributions | |
sauce | +0.014 |
gum | +0.013 |
subtract | +0.013 |
greens | +0.013 |
triangles | +0.013 |
Neg logit contributions | |
decision | -0.012 |
choice | -0.011 |
away | -0.009 |
taxi | -0.009 |
promise | -0.009 |
Token | , |
Token ID | 11 |
Feature activation | -1.58e-02 |
Pos logit contributions | |
subtract | +0.138 |
greens | +0.138 |
Ol | +0.137 |
triangles | +0.134 |
squares | +0.132 |
Neg logit contributions | |
decision | -0.087 |
choice | -0.087 |
away | -0.071 |
promise | -0.064 |
or | -0.056 |
Token | and |
Token ID | 290 |
Feature activation | -1.91e-03 |
Pos logit contributions | |
greens | +0.030 |
sauce | +0.029 |
subtract | +0.029 |
gum | +0.029 |
Ol | +0.029 |
Neg logit contributions | |
decision | -0.022 |
choice | -0.020 |
away | -0.018 |
exactly | -0.017 |
promise | -0.016 |
Token | I |
Token ID | 314 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
greens | +0.004 |
sauce | +0.004 |
triangles | +0.004 |
Ol | +0.004 |
gum | +0.004 |
Neg logit contributions | |
decision | -0.003 |
choice | -0.002 |
away | -0.002 |
exactly | -0.002 |
promise | -0.002 |
Token | wanted |
Token ID | 2227 |
Feature activation | +1.10e-02 |
Pos logit contributions | |
decision | +0.017 |
choice | +0.017 |
away | +0.016 |
promise | +0.014 |
ook | +0.014 |
Neg logit contributions | |
sauce | -0.021 |
gum | -0.021 |
greens | -0.021 |
triangles | -0.020 |
glasses | -0.020 |
Token | to |
Token ID | 284 |
Feature activation | -1.01e-02 |
Pos logit contributions | |
decision | +0.012 |
away | +0.012 |
choice | +0.012 |
home | +0.010 |
exactly | +0.009 |
Neg logit contributions | |
greens | -0.023 |
gloss | -0.023 |
pads | -0.023 |
Ol | -0.022 |
subtract | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
decision | +0.009 |
choice | +0.008 |
away | +0.008 |
ook | +0.007 |
taxi | +0.007 |
Neg logit contributions | |
sauce | -0.008 |
gum | -0.007 |
greens | -0.007 |
glasses | -0.007 |
ink | -0.007 |
Token | He |
Token ID | 679 |
Feature activation | +7.11e-03 |
Pos logit contributions | |
sauce | +0.027 |
Ol | +0.027 |
glasses | +0.027 |
gum | +0.026 |
ink | +0.026 |
Neg logit contributions | |
decision | -0.017 |
choice | -0.016 |
ook | -0.015 |
away | -0.013 |
iring | -0.013 |
Token | called |
Token ID | 1444 |
Feature activation | -2.12e-02 |
Pos logit contributions | |
decision | +0.012 |
choice | +0.012 |
away | +0.012 |
ook | +0.011 |
taxi | +0.011 |
Neg logit contributions | |
gum | -0.011 |
sauce | -0.010 |
greens | -0.010 |
subtract | -0.010 |
triangles | -0.010 |
Token | his |
Token ID | 465 |
Feature activation | -1.50e-02 |
Pos logit contributions | |
sauce | +0.039 |
gum | +0.039 |
greens | +0.037 |
glasses | +0.036 |
ink | +0.036 |
Neg logit contributions | |
decision | -0.032 |
choice | -0.030 |
away | -0.030 |
ook | -0.026 |
taxi | -0.025 |
Token | big |
Token ID | 1263 |
Feature activation | -2.37e-02 |
Pos logit contributions | |
gum | +0.033 |
sauce | +0.032 |
greens | +0.032 |
subtract | +0.032 |
Ol | +0.031 |
Neg logit contributions | |
decision | -0.018 |
choice | -0.018 |
taxi | -0.013 |
away | -0.013 |
promise | -0.013 |
Token | sister |
Token ID | 6621 |
Feature activation | -6.68e-02 |
Pos logit contributions | |
gum | +0.048 |
greens | +0.048 |
pads | +0.047 |
sauce | +0.047 |
gloss | +0.047 |
Neg logit contributions | |
choice | -0.033 |
decision | -0.032 |
away | -0.026 |
taxi | -0.025 |
hotel | -0.022 |
Token | , |
Token ID | 11 |
Feature activation | -1.11e-02 |
Pos logit contributions | |
gum | +0.159 |
fiber | +0.158 |
sauce | +0.152 |
greens | +0.152 |
Crunch | +0.151 |
Neg logit contributions | |
decision | -0.047 |
forward | -0.046 |
choice | -0.043 |
away | -0.042 |
taxi | -0.042 |
Token | Lily |
Token ID | 20037 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
greens | +0.021 |
sauce | +0.021 |
gum | +0.020 |
fiber | +0.020 |
gloss | +0.020 |
Neg logit contributions | |
choice | -0.013 |
exactly | -0.013 |
taxi | -0.013 |
decision | -0.013 |
where | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
gum | +0.033 |
sauce | +0.033 |
greens | +0.032 |
fiber | +0.032 |
subtract | +0.032 |
Neg logit contributions | |
decision | -0.009 |
away | -0.009 |
choice | -0.008 |
ook | -0.008 |
taxi | -0.008 |
Token | " |
Token ID | 366 |
Feature activation | -2.99e-03 |
Pos logit contributions | |
gum | +0.023 |
sauce | +0.023 |
pads | +0.023 |
Ol | +0.022 |
dots | +0.022 |
Neg logit contributions | |
choice | -0.016 |
decision | -0.016 |
ook | -0.015 |
away | -0.014 |
taxi | -0.013 |
Token | L |
Token ID | 43 |
Feature activation | -2.84e-03 |
Pos logit contributions | |
sauce | +0.004 |
gum | +0.004 |
greens | +0.004 |
triangles | +0.004 |
glasses | +0.004 |
Neg logit contributions | |
away | -0.006 |
decision | -0.005 |
choice | -0.005 |
ook | -0.005 |
taxi | -0.005 |
Token | He |
Token ID | 679 |
Feature activation | -4.86e-03 |
Pos logit contributions | |
sauce | +0.028 |
gum | +0.027 |
triangles | +0.026 |
ink | +0.026 |
greens | +0.026 |
Neg logit contributions | |
decision | -0.020 |
choice | -0.019 |
forward | -0.019 |
away | -0.018 |
ook | -0.018 |
Token | washed |
Token ID | 18989 |
Feature activation | -2.43e-02 |
Pos logit contributions | |
greens | +0.008 |
gum | +0.008 |
fiber | +0.007 |
lenses | +0.007 |
pads | +0.007 |
Neg logit contributions | |
choice | -0.007 |
decision | -0.007 |
promise | -0.007 |
ook | -0.006 |
sadly | -0.006 |
Token | it |
Token ID | 340 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
sauce | +0.046 |
subtract | +0.046 |
greens | +0.046 |
gloss | +0.045 |
pads | +0.045 |
Neg logit contributions | |
decision | -0.031 |
away | -0.030 |
choice | -0.030 |
forward | -0.025 |
or | -0.024 |
Token | with |
Token ID | 351 |
Feature activation | +9.36e-03 |
Pos logit contributions | |
greens | +0.030 |
gum | +0.030 |
Ol | +0.030 |
pill | +0.030 |
subtract | +0.030 |
Neg logit contributions | |
decision | -0.014 |
away | -0.013 |
choice | -0.013 |
forward | -0.010 |
exactly | -0.010 |
Token | so |
Token ID | 523 |
Feature activation | -1.21e-02 |
Pos logit contributions | |
choice | +0.012 |
decision | +0.011 |
home | +0.010 |
exactly | +0.009 |
purchase | +0.009 |
Neg logit contributions | |
storms | -0.019 |
pill | -0.019 |
gloss | -0.018 |
pads | -0.018 |
bows | -0.018 |
Token | apy |
Token ID | 12826 |
Feature activation | -6.68e-02 |
Pos logit contributions | |
greens | +0.024 |
sauce | +0.024 |
pads | +0.024 |
glasses | +0.023 |
lenses | +0.023 |
Neg logit contributions | |
decision | -0.015 |
away | -0.014 |
choice | -0.014 |
ook | -0.012 |
forward | -0.012 |
Token | water |
Token ID | 1660 |
Feature activation | -1.30e-02 |
Pos logit contributions | |
Ol | +0.110 |
pads | +0.109 |
blond | +0.107 |
gloss | +0.107 |
Indians | +0.106 |
Neg logit contributions | |
choice | -0.099 |
decision | -0.088 |
away | -0.084 |
of | -0.072 |
promise | -0.071 |
Token | and |
Token ID | 290 |
Feature activation | -1.45e-02 |
Pos logit contributions | |
Ol | +0.030 |
subtract | +0.030 |
blond | +0.029 |
greens | +0.029 |
mascara | +0.029 |
Neg logit contributions | |
decision | -0.011 |
away | -0.011 |
choice | -0.011 |
forward | -0.010 |
taxi | -0.008 |
Token | dried |
Token ID | 16577 |
Feature activation | -7.70e-03 |
Pos logit contributions | |
Ol | +0.022 |
Gary | +0.021 |
pill | +0.021 |
greens | +0.020 |
storms | +0.020 |
Neg logit contributions | |
choice | -0.022 |
decision | -0.021 |
exactly | -0.018 |
sadly | -0.018 |
promise | -0.017 |
Token | it |
Token ID | 340 |
Feature activation | -1.19e-02 |
Pos logit contributions | |
sauce | +0.015 |
gum | +0.015 |
greens | +0.015 |
glasses | +0.014 |
subtract | +0.014 |
Neg logit contributions | |
decision | -0.010 |
away | -0.010 |
choice | -0.010 |
ook | -0.008 |
ively | -0.008 |
Token | with |
Token ID | 351 |
Feature activation | +1.05e-02 |
Pos logit contributions | |
sauce | +0.026 |
gum | +0.026 |
greens | +0.026 |
glasses | +0.026 |
Ol | +0.026 |
Neg logit contributions | |
decision | -0.013 |
away | -0.012 |
choice | -0.012 |
forward | -0.009 |
exactly | -0.009 |
Token | a |
Token ID | 257 |
Feature activation | +1.18e-02 |
Pos logit contributions | |
sauce | +0.024 |
subtract | +0.024 |
greens | +0.024 |
pads | +0.024 |
gum | +0.024 |
Neg logit contributions | |
decision | -0.016 |
choice | -0.016 |
away | -0.016 |
forward | -0.013 |
ively | -0.012 |
Token | toy |
Token ID | 13373 |
Feature activation | +9.69e-03 |
Pos logit contributions | |
decision | +0.019 |
choice | +0.018 |
away | +0.018 |
taxi | +0.016 |
ook | +0.014 |
Neg logit contributions | |
sauce | -0.020 |
greens | -0.020 |
subtract | -0.020 |
pads | -0.019 |
gum | -0.019 |
Token | symbol |
Token ID | 6194 |
Feature activation | -1.74e-02 |
Pos logit contributions | |
decision | +0.013 |
choice | +0.013 |
away | +0.013 |
ook | +0.010 |
taxi | +0.010 |
Neg logit contributions | |
sauce | -0.018 |
gum | -0.018 |
greens | -0.018 |
subtract | -0.017 |
glasses | -0.017 |
Token | in |
Token ID | 287 |
Feature activation | +2.54e-03 |
Pos logit contributions | |
Ol | +0.040 |
greens | +0.039 |
subtract | +0.038 |
fiber | +0.038 |
joked | +0.037 |
Neg logit contributions | |
decision | -0.017 |
choice | -0.016 |
away | -0.016 |
forward | -0.013 |
exactly | -0.013 |
Token | so |
Token ID | 523 |
Feature activation | +2.87e-05 |
Pos logit contributions | |
choice | +0.004 |
decision | +0.004 |
away | +0.004 |
purchase | +0.003 |
ook | +0.003 |
Neg logit contributions | |
pads | -0.005 |
sauce | -0.005 |
gloss | -0.005 |
greens | -0.005 |
gum | -0.004 |
Token | apy |
Token ID | 12826 |
Feature activation | -6.66e-02 |
Pos logit contributions | |
decision | +0.000 |
choice | +0.000 |
away | +0.000 |
ook | +0.000 |
taxi | +0.000 |
Neg logit contributions | |
sauce | -0.000 |
greens | -0.000 |
gum | -0.000 |
pads | -0.000 |
glasses | -0.000 |
Token | water |
Token ID | 1660 |
Feature activation | -8.64e-03 |
Pos logit contributions | |
Ol | +0.134 |
pads | +0.132 |
Indians | +0.130 |
ccoli | +0.124 |
Gary | +0.124 |
Neg logit contributions | |
choice | -0.079 |
away | -0.077 |
decision | -0.069 |
forward | -0.054 |
of | -0.054 |
Token | , |
Token ID | 11 |
Feature activation | -4.45e-04 |
Pos logit contributions | |
Ol | +0.021 |
subtract | +0.020 |
greens | +0.019 |
fiber | +0.019 |
triangles | +0.019 |
Neg logit contributions | |
decision | -0.008 |
choice | -0.007 |
away | -0.007 |
forward | -0.006 |
or | -0.005 |
Token | it |
Token ID | 340 |
Feature activation | -2.42e-02 |
Pos logit contributions | |
sauce | +0.001 |
greens | +0.001 |
pads | +0.001 |
Ol | +0.001 |
subtract | +0.001 |
Neg logit contributions | |
decision | -0.001 |
choice | -0.001 |
away | -0.000 |
exactly | -0.000 |
forward | -0.000 |
Token | would |
Token ID | 561 |
Feature activation | +9.79e-03 |
Pos logit contributions | |
gum | +0.044 |
greens | +0.043 |
subtract | +0.043 |
sauce | +0.042 |
fiber | +0.041 |
Neg logit contributions | |
decision | -0.035 |
choice | -0.034 |
away | -0.030 |
exactly | -0.027 |
sadly | -0.027 |
Token | become |
Token ID | 1716 |
Feature activation | -6.30e-03 |
Pos logit contributions | |
decision | +0.015 |
choice | +0.014 |
away | +0.014 |
promise | +0.012 |
taxi | +0.012 |
Neg logit contributions | |
gum | -0.017 |
pads | -0.017 |
sauce | -0.017 |
glasses | -0.017 |
greens | -0.017 |
Token | in |
Token ID | 287 |
Feature activation | +7.91e-04 |
Pos logit contributions | |
decision | +0.008 |
away | +0.007 |
choice | +0.007 |
ook | +0.006 |
taxi | +0.006 |
Neg logit contributions | |
sauce | -0.013 |
gum | -0.012 |
greens | -0.012 |
subtract | -0.012 |
glasses | -0.012 |
Token | the |
Token ID | 262 |
Feature activation | +5.41e-03 |
Pos logit contributions | |
decision | +0.001 |
choice | +0.001 |
away | +0.001 |
taxi | +0.001 |
ook | +0.001 |
Neg logit contributions | |
sauce | -0.001 |
greens | -0.001 |
pads | -0.001 |
gum | -0.001 |
gloss | -0.001 |
Token | warm |
Token ID | 5814 |
Feature activation | -1.86e-02 |
Pos logit contributions | |
decision | +0.011 |
choice | +0.010 |
away | +0.010 |
taxi | +0.009 |
hotel | +0.009 |
Neg logit contributions | |
sauce | -0.007 |
greens | -0.007 |
gum | -0.007 |
subtract | -0.007 |
pads | -0.007 |
Token | , |
Token ID | 11 |
Feature activation | -1.01e-02 |
Pos logit contributions | |
sauce | +0.021 |
gum | +0.021 |
greens | +0.020 |
subtract | +0.019 |
glasses | +0.019 |
Neg logit contributions | |
decision | -0.040 |
choice | -0.039 |
away | -0.039 |
taxi | -0.035 |
ook | -0.034 |
Token | so |
Token ID | 523 |
Feature activation | +4.43e-03 |
Pos logit contributions | |
subtract | +0.016 |
sauce | +0.015 |
greens | +0.015 |
gum | +0.015 |
gloss | +0.015 |
Neg logit contributions | |
decision | -0.016 |
choice | -0.015 |
away | -0.015 |
taxi | -0.014 |
forward | -0.013 |
Token | apy |
Token ID | 12826 |
Feature activation | -6.65e-02 |
Pos logit contributions | |
decision | +0.006 |
choice | +0.006 |
away | +0.006 |
taxi | +0.005 |
ook | +0.005 |
Neg logit contributions | |
sauce | -0.008 |
greens | -0.008 |
gum | -0.008 |
glasses | -0.008 |
subtract | -0.008 |
Token | water |
Token ID | 1660 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
subtract | +0.112 |
blond | +0.105 |
Ol | +0.102 |
gloss | +0.099 |
greens | +0.098 |
Neg logit contributions | |
choice | -0.103 |
away | -0.098 |
decision | -0.096 |
taxi | -0.086 |
hotel | -0.085 |
Token | . |
Token ID | 13 |
Feature activation | -1.83e-04 |
Pos logit contributions | |
subtract | +0.030 |
Ol | +0.029 |
greens | +0.028 |
gum | +0.028 |
blond | +0.028 |
Neg logit contributions | |
decision | -0.014 |
away | -0.013 |
choice | -0.013 |
forward | -0.010 |
taxi | -0.010 |
Token | His |
Token ID | 2399 |
Feature activation | +1.70e-02 |
Pos logit contributions | |
sauce | +0.000 |
gum | +0.000 |
gloss | +0.000 |
pads | +0.000 |
ink | +0.000 |
Neg logit contributions | |
decision | -0.000 |
choice | -0.000 |
taxi | -0.000 |
ook | -0.000 |
away | -0.000 |
Token | mom |
Token ID | 1995 |
Feature activation | -1.86e-02 |
Pos logit contributions | |
decision | +0.021 |
choice | +0.020 |
away | +0.017 |
taxi | +0.016 |
doorstep | +0.015 |
Neg logit contributions | |
sauce | -0.036 |
greens | -0.035 |
gum | -0.035 |
subtract | -0.035 |
triangles | -0.034 |
Token | and |
Token ID | 290 |
Feature activation | +1.03e-02 |
Pos logit contributions | |
gum | +0.035 |
greens | +0.035 |
sauce | +0.034 |
subtract | +0.033 |
fiber | +0.033 |
Neg logit contributions | |
decision | -0.024 |
choice | -0.023 |
ook | -0.022 |
tha | -0.022 |
away | -0.021 |
Token | som |
Token ID | 3870 |
Feature activation | -2.86e-02 |
Pos logit contributions | |
decision | +0.038 |
choice | +0.037 |
away | +0.036 |
taxi | +0.031 |
ook | +0.031 |
Neg logit contributions | |
sauce | -0.037 |
gum | -0.037 |
greens | -0.036 |
glasses | -0.035 |
subtract | -0.034 |
Token | ers |
Token ID | 364 |
Feature activation | +6.22e-03 |
Pos logit contributions | |
greens | +0.047 |
sauce | +0.046 |
gum | +0.044 |
subtract | +0.044 |
glasses | +0.043 |
Neg logit contributions | |
away | -0.046 |
decision | -0.045 |
choice | -0.045 |
ook | -0.040 |
or | -0.039 |
Token | ault |
Token ID | 1721 |
Feature activation | +7.74e-03 |
Pos logit contributions | |
decision | +0.009 |
away | +0.008 |
choice | +0.008 |
ook | +0.007 |
taxi | +0.007 |
Neg logit contributions | |
sauce | -0.012 |
gum | -0.011 |
greens | -0.011 |
glasses | -0.011 |
ink | -0.011 |
Token | ?" |
Token ID | 1701 |
Feature activation | -5.47e-03 |
Pos logit contributions | |
decision | +0.009 |
away | +0.009 |
choice | +0.009 |
forward | +0.007 |
taxi | +0.007 |
Neg logit contributions | |
greens | -0.016 |
sauce | -0.016 |
gum | -0.016 |
subtract | -0.015 |
fiber | -0.015 |
Token | Her |
Token ID | 2332 |
Feature activation | -5.28e-03 |
Pos logit contributions | |
sauce | +0.010 |
gum | +0.009 |
pads | +0.009 |
greens | +0.009 |
triangles | +0.009 |
Neg logit contributions | |
decision | -0.008 |
choice | -0.008 |
away | -0.007 |
ook | -0.007 |
taxi | -0.006 |
Token | sister |
Token ID | 6621 |
Feature activation | -6.64e-02 |
Pos logit contributions | |
sauce | +0.010 |
gum | +0.010 |
greens | +0.010 |
pads | +0.009 |
subtract | +0.009 |
Neg logit contributions | |
decision | -0.008 |
choice | -0.007 |
away | -0.007 |
ook | -0.006 |
taxi | -0.006 |
Token | was |
Token ID | 373 |
Feature activation | -4.69e-03 |
Pos logit contributions | |
sauce | +0.123 |
pads | +0.120 |
gum | +0.120 |
fiber | +0.120 |
ink | +0.114 |
Neg logit contributions | |
decision | -0.094 |
sadly | -0.093 |
choice | -0.090 |
exactly | -0.081 |
promise | -0.080 |
Token | very |
Token ID | 845 |
Feature activation | -5.80e-03 |
Pos logit contributions | |
fiber | +0.010 |
sauce | +0.010 |
greens | +0.010 |
pads | +0.010 |
gum | +0.009 |
Neg logit contributions | |
decision | -0.005 |
choice | -0.005 |
exactly | -0.005 |
forward | -0.005 |
ook | -0.005 |
Token | patient |
Token ID | 5827 |
Feature activation | -2.07e-02 |
Pos logit contributions | |
sauce | +0.011 |
gum | +0.011 |
greens | +0.010 |
pads | +0.010 |
ink | +0.010 |
Neg logit contributions | |
decision | -0.009 |
choice | -0.009 |
away | -0.008 |
ook | -0.007 |
taxi | -0.007 |
Token | and |
Token ID | 290 |
Feature activation | +4.25e-03 |
Pos logit contributions | |
sauce | +0.044 |
gum | +0.044 |
greens | +0.043 |
subtract | +0.042 |
Ol | +0.042 |
Neg logit contributions | |
decision | -0.025 |
choice | -0.023 |
away | -0.022 |
ook | -0.017 |
forward | -0.017 |
Token | said |
Token ID | 531 |
Feature activation | -8.27e-03 |
Pos logit contributions | |
decision | +0.008 |
choice | +0.007 |
exactly | +0.007 |
away | +0.007 |
sadly | +0.007 |
Neg logit contributions | |
greens | -0.006 |
sauce | -0.006 |
gum | -0.006 |
triangles | -0.006 |
Ol | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | -8.98e-03 |
Pos logit contributions | |
sauce | +0.006 |
gum | +0.006 |
greens | +0.005 |
glasses | +0.005 |
ink | +0.005 |
Neg logit contributions | |
decision | -0.003 |
choice | -0.003 |
away | -0.003 |
ook | -0.003 |
taxi | -0.003 |
Token | filled |
Token ID | 5901 |
Feature activation | +8.48e-03 |
Pos logit contributions | |
gum | +0.013 |
subtract | +0.013 |
greens | +0.013 |
gloss | +0.013 |
sauce | +0.012 |
Neg logit contributions | |
choice | -0.014 |
decision | -0.014 |
away | -0.014 |
ook | -0.013 |
taxi | -0.013 |
Token | it |
Token ID | 340 |
Feature activation | -1.76e-03 |
Pos logit contributions | |
decision | +0.013 |
choice | +0.013 |
away | +0.013 |
ook | +0.011 |
taxi | +0.011 |
Neg logit contributions | |
sauce | -0.015 |
gum | -0.015 |
greens | -0.014 |
glasses | -0.014 |
ink | -0.013 |
Token | with |
Token ID | 351 |
Feature activation | +1.86e-02 |
Pos logit contributions | |
sauce | +0.004 |
gum | +0.003 |
greens | +0.003 |
glasses | +0.003 |
ink | +0.003 |
Neg logit contributions | |
decision | -0.002 |
choice | -0.002 |
away | -0.002 |
ook | -0.002 |
taxi | -0.002 |
Token | so |
Token ID | 523 |
Feature activation | -3.70e-03 |
Pos logit contributions | |
choice | +0.030 |
decision | +0.030 |
away | +0.027 |
exactly | +0.025 |
taxi | +0.024 |
Neg logit contributions | |
gloss | -0.031 |
Ol | -0.030 |
subtract | -0.029 |
pads | -0.028 |
gum | -0.028 |
Token | apy |
Token ID | 12826 |
Feature activation | -6.63e-02 |
Pos logit contributions | |
sauce | +0.007 |
gum | +0.007 |
greens | +0.007 |
glasses | +0.007 |
ink | +0.007 |
Neg logit contributions | |
decision | -0.005 |
choice | -0.005 |
away | -0.005 |
ook | -0.004 |
taxi | -0.004 |
Token | water |
Token ID | 1660 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
subtract | +0.109 |
gloss | +0.109 |
Ol | +0.108 |
blond | +0.106 |
pads | +0.101 |
Neg logit contributions | |
choice | -0.097 |
away | -0.096 |
decision | -0.091 |
taxi | -0.079 |
of | -0.074 |
Token | . |
Token ID | 13 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
Ol | +0.025 |
subtract | +0.025 |
gum | +0.024 |
blond | +0.024 |
greens | +0.024 |
Neg logit contributions | |
decision | -0.012 |
away | -0.011 |
choice | -0.011 |
forward | -0.009 |
taxi | -0.009 |
Token | He |
Token ID | 679 |
Feature activation | -7.99e-04 |
Pos logit contributions | |
sauce | +0.028 |
gum | +0.027 |
glasses | +0.027 |
ink | +0.027 |
gloss | +0.026 |
Neg logit contributions | |
decision | -0.018 |
ook | -0.017 |
choice | -0.017 |
away | -0.016 |
forward | -0.016 |
Token | went |
Token ID | 1816 |
Feature activation | +9.13e-04 |
Pos logit contributions | |
gum | +0.001 |
greens | +0.001 |
subtract | +0.001 |
sauce | +0.001 |
gloss | +0.001 |
Neg logit contributions | |
decision | -0.001 |
choice | -0.001 |
away | -0.001 |
ook | -0.001 |
promise | -0.001 |
Token | back |
Token ID | 736 |
Feature activation | +6.87e-05 |
Pos logit contributions | |
decision | +0.001 |
choice | +0.001 |
away | +0.001 |
tha | +0.001 |
ook | +0.001 |
Neg logit contributions | |
sauce | -0.002 |
gum | -0.002 |
glasses | -0.001 |
greens | -0.001 |
ink | -0.001 |
Token | ?" |
Token ID | 1701 |
Feature activation | -7.47e-03 |
Pos logit contributions | |
Ol | +0.018 |
glasses | +0.018 |
gum | +0.017 |
sauce | +0.017 |
blond | +0.017 |
Neg logit contributions | |
decision | -0.010 |
choice | -0.010 |
away | -0.010 |
exactly | -0.008 |
promise | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | +3.02e-03 |
Pos logit contributions | |
sauce | +0.014 |
pads | +0.014 |
dots | +0.013 |
fiber | +0.013 |
glasses | +0.013 |
Neg logit contributions | |
sadly | -0.009 |
away | -0.009 |
decision | -0.009 |
choice | -0.009 |
curiously | -0.009 |
Token |
|
Token ID | 198 |
Feature activation | +2.24e-03 |
Pos logit contributions | |
away | +0.004 |
ook | +0.004 |
decision | +0.004 |
choice | +0.004 |
taxi | +0.004 |
Neg logit contributions | |
sauce | -0.006 |
greens | -0.005 |
pads | -0.005 |
gum | -0.005 |
glasses | -0.005 |
Token | His |
Token ID | 6653 |
Feature activation | +6.14e-03 |
Pos logit contributions | |
away | +0.004 |
decision | +0.003 |
ook | +0.003 |
choice | +0.003 |
taxi | +0.003 |
Neg logit contributions | |
sauce | -0.004 |
gum | -0.004 |
greens | -0.003 |
glasses | -0.003 |
pads | -0.003 |
Token | oldest |
Token ID | 13325 |
Feature activation | -2.90e-02 |
Pos logit contributions | |
decision | +0.007 |
choice | +0.007 |
away | +0.006 |
taxi | +0.005 |
ook | +0.005 |
Neg logit contributions | |
sauce | -0.013 |
gum | -0.013 |
greens | -0.013 |
glasses | -0.012 |
pads | -0.012 |
Token | sister |
Token ID | 6621 |
Feature activation | -6.62e-02 |
Pos logit contributions | |
sauce | +0.058 |
greens | +0.057 |
gum | +0.057 |
pads | +0.055 |
gloss | +0.054 |
Neg logit contributions | |
decision | -0.039 |
choice | -0.039 |
away | -0.032 |
tha | -0.029 |
promise | -0.029 |
Token | replied |
Token ID | 8712 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
gum | +0.131 |
sauce | +0.128 |
fiber | +0.126 |
pads | +0.124 |
greens | +0.123 |
Neg logit contributions | |
decision | -0.088 |
choice | -0.084 |
sadly | -0.079 |
exactly | -0.071 |
curiously | -0.067 |
Token | in |
Token ID | 287 |
Feature activation | -3.03e-02 |
Pos logit contributions | |
sauce | +0.025 |
gum | +0.024 |
pads | +0.024 |
greens | +0.024 |
subtract | +0.024 |
Neg logit contributions | |
decision | -0.012 |
away | -0.012 |
choice | -0.011 |
forward | -0.010 |
ook | -0.010 |
Token | a |
Token ID | 257 |
Feature activation | -2.08e-02 |
Pos logit contributions | |
pads | +0.050 |
sauce | +0.048 |
gloss | +0.047 |
gum | +0.047 |
greens | +0.046 |
Neg logit contributions | |
decision | -0.049 |
choice | -0.048 |
away | -0.047 |
purchase | -0.040 |
forward | -0.039 |
Token | loud |
Token ID | 7812 |
Feature activation | -3.68e-02 |
Pos logit contributions | |
sauce | +0.038 |
pads | +0.038 |
gum | +0.038 |
gloss | +0.036 |
fiber | +0.036 |
Neg logit contributions | |
decision | -0.033 |
choice | -0.032 |
away | -0.027 |
promise | -0.025 |
taxi | -0.024 |
Token | and |
Token ID | 290 |
Feature activation | -7.57e-03 |
Pos logit contributions | |
pads | +0.083 |
gum | +0.081 |
sauce | +0.079 |
fiber | +0.077 |
Ol | +0.076 |
Neg logit contributions | |
decision | -0.048 |
choice | -0.044 |
away | -0.043 |
taxi | -0.034 |
deed | -0.029 |
Token | me |
Token ID | 502 |
Feature activation | -1.99e-02 |
Pos logit contributions | |
decision | +0.028 |
away | +0.028 |
choice | +0.027 |
ook | +0.026 |
taxi | +0.024 |
Neg logit contributions | |
greens | -0.032 |
sauce | -0.032 |
gum | -0.031 |
fiber | -0.030 |
subtract | -0.029 |
Token | ?" |
Token ID | 1701 |
Feature activation | -1.23e-02 |
Pos logit contributions | |
sauce | +0.041 |
gum | +0.041 |
greens | +0.038 |
subtract | +0.038 |
ink | +0.038 |
Neg logit contributions | |
decision | -0.027 |
choice | -0.025 |
away | -0.024 |
taxi | -0.020 |
promise | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | -2.42e-03 |
Pos logit contributions | |
sauce | +0.023 |
pads | +0.022 |
fiber | +0.022 |
dots | +0.021 |
gum | +0.021 |
Neg logit contributions | |
sadly | -0.015 |
taxi | -0.014 |
away | -0.014 |
choice | -0.014 |
decision | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -2.66e-03 |
Pos logit contributions | |
sauce | +0.004 |
gum | +0.004 |
greens | +0.004 |
glasses | +0.004 |
ink | +0.004 |
Neg logit contributions | |
away | -0.004 |
decision | -0.004 |
choice | -0.004 |
ook | -0.003 |
taxi | -0.003 |
Token | An |
Token ID | 2025 |
Feature activation | +4.60e-03 |
Pos logit contributions | |
sauce | +0.005 |
gum | +0.004 |
greens | +0.004 |
ink | +0.004 |
glasses | +0.004 |
Neg logit contributions | |
away | -0.004 |
decision | -0.004 |
choice | -0.004 |
ook | -0.004 |
taxi | -0.004 |
Token | ch |
Token ID | 354 |
Feature activation | -6.60e-02 |
Pos logit contributions | |
decision | +0.007 |
choice | +0.006 |
away | +0.006 |
ook | +0.005 |
taxi | +0.005 |
Neg logit contributions | |
sauce | -0.009 |
gum | -0.009 |
greens | -0.008 |
glasses | -0.008 |
ink | -0.008 |
Token | or |
Token ID | 273 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
gum | +0.137 |
sauce | +0.136 |
greens | +0.134 |
subtract | +0.131 |
glasses | +0.131 |
Neg logit contributions | |
decision | -0.083 |
choice | -0.079 |
away | -0.075 |
ook | -0.070 |
ively | -0.069 |
Token | said |
Token ID | 531 |
Feature activation | -2.14e-02 |
Pos logit contributions | |
sauce | +0.027 |
greens | +0.026 |
gum | +0.026 |
fiber | +0.024 |
ink | +0.024 |
Neg logit contributions | |
decision | -0.022 |
choice | -0.022 |
away | -0.021 |
tha | -0.020 |
ently | -0.020 |
Token | , |
Token ID | 11 |
Feature activation | -4.73e-02 |
Pos logit contributions | |
sauce | +0.052 |
pads | +0.051 |
greens | +0.051 |
subtract | +0.050 |
fiber | +0.050 |
Neg logit contributions | |
away | -0.018 |
decision | -0.016 |
choice | -0.015 |
taxi | -0.015 |
sadly | -0.014 |
Token | " |
Token ID | 366 |
Feature activation | -2.55e-02 |
Pos logit contributions | |
fiber | +0.095 |
sauce | +0.094 |
pads | +0.091 |
gum | +0.091 |
vitamins | +0.089 |
Neg logit contributions | |
sadly | -0.055 |
taxi | -0.054 |
forward | -0.054 |
tha | -0.051 |
decision | -0.051 |
Token | Yes |
Token ID | 5297 |
Feature activation | +1.46e-02 |
Pos logit contributions | |
sauce | +0.058 |
ink | +0.054 |
gum | +0.052 |
glasses | +0.052 |
pens | +0.051 |
Neg logit contributions | |
away | -0.028 |
Only | -0.021 |
forward | -0.021 |
or | -0.020 |
taxi | -0.020 |
Token | to |
Token ID | 284 |
Feature activation | -4.93e-03 |
Pos logit contributions | |
greens | +0.019 |
gloss | +0.019 |
lenses | +0.019 |
subtract | +0.019 |
pads | +0.019 |
Neg logit contributions | |
away | -0.005 |
home | -0.005 |
taxi | -0.004 |
where | -0.004 |
decision | -0.004 |
Token | play |
Token ID | 711 |
Feature activation | -4.58e-03 |
Pos logit contributions | |
pads | +0.010 |
triangles | +0.010 |
pill | +0.010 |
gum | +0.010 |
subtract | +0.010 |
Neg logit contributions | |
choice | -0.006 |
decision | -0.005 |
away | -0.005 |
taxi | -0.004 |
promise | -0.004 |
Token | all |
Token ID | 477 |
Feature activation | +2.07e-02 |
Pos logit contributions | |
sauce | +0.009 |
gum | +0.009 |
greens | +0.009 |
glasses | +0.009 |
subtract | +0.009 |
Neg logit contributions | |
decision | -0.006 |
away | -0.006 |
choice | -0.006 |
ook | -0.005 |
taxi | -0.005 |
Token | day |
Token ID | 1110 |
Feature activation | -4.55e-03 |
Pos logit contributions | |
decision | +0.039 |
choice | +0.037 |
away | +0.035 |
doorstep | +0.033 |
destination | +0.032 |
Neg logit contributions | |
sauce | -0.031 |
gum | -0.029 |
greens | -0.028 |
ink | -0.028 |
triangles | -0.028 |
Token | , |
Token ID | 11 |
Feature activation | +3.61e-04 |
Pos logit contributions | |
subtract | +0.008 |
sauce | +0.007 |
greens | +0.007 |
gum | +0.007 |
gloss | +0.007 |
Neg logit contributions | |
away | -0.007 |
decision | -0.007 |
choice | -0.006 |
ook | -0.006 |
forward | -0.005 |
Token | but |
Token ID | 475 |
Feature activation | +5.32e-02 |
Pos logit contributions | |
decision | +0.000 |
choice | +0.000 |
away | +0.000 |
taxi | +0.000 |
ook | +0.000 |
Neg logit contributions | |
sauce | -0.001 |
gum | -0.001 |
subtract | -0.001 |
ink | -0.001 |
glasses | -0.001 |
Token | today |
Token ID | 1909 |
Feature activation | -2.29e-03 |
Pos logit contributions | |
decision | +0.069 |
choice | +0.067 |
away | +0.058 |
taxi | +0.054 |
exactly | +0.054 |
Neg logit contributions | |
subtract | -0.103 |
greens | -0.102 |
lenses | -0.098 |
sauce | -0.098 |
glasses | -0.096 |
Token | was |
Token ID | 373 |
Feature activation | +7.33e-03 |
Pos logit contributions | |
sauce | +0.004 |
gum | +0.004 |
greens | +0.004 |
glasses | +0.004 |
ink | +0.004 |
Neg logit contributions | |
decision | -0.004 |
choice | -0.003 |
away | -0.003 |
ook | -0.003 |
taxi | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | +7.62e-03 |
Pos logit contributions | |
decision | +0.008 |
choice | +0.008 |
exactly | +0.007 |
badly | +0.007 |
forward | +0.006 |
Neg logit contributions | |
greens | -0.016 |
subtract | -0.016 |
fiber | -0.015 |
lenses | -0.015 |
pads | -0.015 |
Token | special |
Token ID | 2041 |
Feature activation | -2.82e-03 |
Pos logit contributions | |
decision | +0.013 |
choice | +0.012 |
away | +0.012 |
taxi | +0.010 |
ook | +0.010 |
Neg logit contributions | |
sauce | -0.013 |
greens | -0.012 |
gum | -0.012 |
subtract | -0.012 |
pads | -0.012 |
Token | day |
Token ID | 1110 |
Feature activation | +8.13e-03 |
Pos logit contributions | |
greens | +0.005 |
subtract | +0.005 |
pads | +0.005 |
triangles | +0.005 |
sauce | +0.005 |
Neg logit contributions | |
decision | -0.005 |
choice | -0.005 |
away | -0.004 |
taxi | -0.004 |
hotel | -0.004 |
Token | to |
Token ID | 284 |
Feature activation | -8.75e-03 |
Pos logit contributions | |
decision | +0.009 |
choice | +0.009 |
away | +0.009 |
ook | +0.008 |
taxi | +0.007 |
Neg logit contributions | |
sauce | -0.013 |
gum | -0.012 |
greens | -0.012 |
glasses | -0.012 |
ink | -0.012 |
Token | go |
Token ID | 467 |
Feature activation | -1.17e-03 |
Pos logit contributions | |
sauce | +0.017 |
gum | +0.017 |
greens | +0.016 |
triangles | +0.016 |
pads | +0.016 |
Neg logit contributions | |
decision | -0.012 |
choice | -0.012 |
away | -0.011 |
doorstep | -0.010 |
promise | -0.009 |
Token | to |
Token ID | 284 |
Feature activation | +4.89e-03 |
Pos logit contributions | |
sauce | +0.002 |
gum | +0.002 |
glasses | +0.002 |
ink | +0.001 |
rubbed | +0.001 |
Neg logit contributions | |
decision | -0.002 |
choice | -0.002 |
away | -0.002 |
ively | -0.002 |
rily | -0.002 |
Token | bed |
Token ID | 3996 |
Feature activation | +2.20e-02 |
Pos logit contributions | |
decision | +0.008 |
choice | +0.008 |
away | +0.008 |
ook | +0.007 |
taxi | +0.007 |
Neg logit contributions | |
sauce | -0.008 |
gum | -0.008 |
greens | -0.008 |
glasses | -0.007 |
pads | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | +4.84e-03 |
Pos logit contributions | |
decision | +0.030 |
choice | +0.029 |
away | +0.028 |
ook | +0.024 |
taxi | +0.023 |
Neg logit contributions | |
sauce | -0.043 |
gum | -0.042 |
greens | -0.041 |
glasses | -0.040 |
ink | -0.039 |
Token | But |
Token ID | 887 |
Feature activation | +4.81e-02 |
Pos logit contributions | |
decision | +0.007 |
choice | +0.006 |
away | +0.006 |
taxi | +0.006 |
sadly | +0.006 |
Neg logit contributions | |
sauce | -0.009 |
gum | -0.008 |
triangles | -0.008 |
Ol | -0.008 |
greens | -0.008 |
Token | Lily |
Token ID | 20037 |
Feature activation | -2.33e-03 |
Pos logit contributions | |
decision | +0.096 |
choice | +0.093 |
away | +0.090 |
ook | +0.081 |
taxi | +0.080 |
Neg logit contributions | |
sauce | -0.064 |
gum | -0.062 |
greens | -0.061 |
glasses | -0.057 |
triangles | -0.057 |
Token | didn |
Token ID | 1422 |
Feature activation | +1.97e-03 |
Pos logit contributions | |
greens | +0.004 |
sauce | +0.004 |
triangles | +0.004 |
squares | +0.004 |
gum | +0.004 |
Neg logit contributions | |
decision | -0.004 |
choice | -0.004 |
away | -0.003 |
promise | -0.003 |
doorstep | -0.003 |
Token | 't |
Token ID | 470 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
decision | +0.004 |
away | +0.004 |
choice | +0.004 |
ook | +0.003 |
checkpoint | +0.003 |
Neg logit contributions | |
gum | -0.003 |
sauce | -0.003 |
ink | -0.002 |
glasses | -0.002 |
subtract | -0.002 |
Token | want |
Token ID | 765 |
Feature activation | -8.06e-03 |
Pos logit contributions | |
decision | +0.016 |
choice | +0.015 |
away | +0.015 |
ook | +0.013 |
taxi | +0.012 |
Neg logit contributions | |
sauce | -0.020 |
gum | -0.020 |
greens | -0.019 |
triangles | -0.019 |
glasses | -0.019 |
Token | to |
Token ID | 284 |
Feature activation | -1.06e-02 |
Pos logit contributions | |
greens | +0.018 |
Ol | +0.017 |
pads | +0.016 |
subtract | +0.016 |
triangles | +0.016 |
Neg logit contributions | |
choice | -0.008 |
decision | -0.008 |
away | -0.008 |
home | -0.007 |
forward | -0.006 |
Token | post |
Token ID | 1281 |
Feature activation | +2.22e-02 |
Pos logit contributions | |
sauce | +0.008 |
gum | +0.008 |
greens | +0.008 |
glasses | +0.008 |
pads | +0.007 |
Neg logit contributions | |
decision | -0.007 |
choice | -0.006 |
away | -0.006 |
taxi | -0.006 |
doorstep | -0.005 |
Token | office |
Token ID | 2607 |
Feature activation | +5.21e-03 |
Pos logit contributions | |
decision | +0.025 |
choice | +0.024 |
away | +0.023 |
ook | +0.019 |
taxi | +0.018 |
Neg logit contributions | |
sauce | -0.049 |
gum | -0.048 |
greens | -0.047 |
glasses | -0.046 |
ink | -0.045 |
Token | . |
Token ID | 13 |
Feature activation | +1.84e-03 |
Pos logit contributions | |
decision | +0.007 |
choice | +0.007 |
away | +0.007 |
ook | +0.006 |
taxi | +0.006 |
Neg logit contributions | |
sauce | -0.010 |
gum | -0.010 |
greens | -0.010 |
fiber | -0.009 |
ink | -0.009 |
Token | On |
Token ID | 1550 |
Feature activation | +3.15e-03 |
Pos logit contributions | |
decision | +0.003 |
choice | +0.003 |
away | +0.003 |
taxi | +0.003 |
ook | +0.003 |
Neg logit contributions | |
sauce | -0.003 |
gum | -0.003 |
greens | -0.003 |
triangles | -0.003 |
glasses | -0.003 |
Token | the |
Token ID | 262 |
Feature activation | +4.96e-03 |
Pos logit contributions | |
decision | +0.006 |
away | +0.005 |
choice | +0.005 |
ook | +0.005 |
checkpoint | +0.005 |
Neg logit contributions | |
sauce | -0.005 |
gum | -0.005 |
greens | -0.004 |
glasses | -0.004 |
ink | -0.004 |
Token | way |
Token ID | 835 |
Feature activation | +4.44e-02 |
Pos logit contributions | |
away | +0.008 |
decision | +0.008 |
choice | +0.008 |
ook | +0.007 |
ively | +0.006 |
Neg logit contributions | |
sauce | -0.008 |
gum | -0.008 |
greens | -0.008 |
glasses | -0.008 |
ink | -0.008 |
Token | , |
Token ID | 11 |
Feature activation | -7.99e-03 |
Pos logit contributions | |
decision | +0.093 |
choice | +0.090 |
away | +0.083 |
ind | +0.080 |
promise | +0.077 |
Neg logit contributions | |
gum | -0.048 |
Sam | -0.047 |
sauce | -0.045 |
Ben | -0.043 |
pads | -0.042 |
Token | he |
Token ID | 339 |
Feature activation | -8.76e-04 |
Pos logit contributions | |
sauce | +0.013 |
gum | +0.012 |
greens | +0.012 |
glasses | +0.011 |
ink | +0.011 |
Neg logit contributions | |
decision | -0.014 |
choice | -0.013 |
away | -0.013 |
ook | -0.012 |
taxi | -0.011 |
Token | met |
Token ID | 1138 |
Feature activation | -1.23e-02 |
Pos logit contributions | |
sauce | +0.001 |
gum | +0.001 |
greens | +0.001 |
subtract | +0.001 |
fiber | +0.001 |
Neg logit contributions | |
decision | -0.001 |
away | -0.001 |
choice | -0.001 |
ook | -0.001 |
destination | -0.001 |
Token | some |
Token ID | 617 |
Feature activation | -2.58e-02 |
Pos logit contributions | |
sauce | +0.020 |
greens | +0.019 |
gum | +0.019 |
fiber | +0.018 |
subtract | +0.018 |
Neg logit contributions | |
decision | -0.021 |
choice | -0.020 |
away | -0.020 |
ook | -0.018 |
taxi | -0.018 |
Token | people |
Token ID | 661 |
Feature activation | -8.14e-03 |
Pos logit contributions | |
sauce | +0.046 |
gum | +0.044 |
greens | +0.044 |
subtract | +0.043 |
Ol | +0.043 |
Neg logit contributions | |
decision | -0.039 |
choice | -0.038 |
away | -0.035 |
taxi | -0.033 |
ook | -0.030 |
Token | dad |
Token ID | 9955 |
Feature activation | +1.15e-03 |
Pos logit contributions | |
decision | +0.023 |
away | +0.023 |
choice | +0.022 |
ook | +0.021 |
ively | +0.020 |
Neg logit contributions | |
gum | -0.022 |
sauce | -0.022 |
glasses | -0.020 |
ink | -0.019 |
subtract | -0.019 |
Token | outside |
Token ID | 2354 |
Feature activation | +3.22e-02 |
Pos logit contributions | |
decision | +0.002 |
choice | +0.002 |
away | +0.002 |
taxi | +0.002 |
tha | +0.002 |
Neg logit contributions | |
sauce | -0.002 |
gum | -0.002 |
greens | -0.002 |
glasses | -0.002 |
ink | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +9.02e-03 |
Pos logit contributions | |
decision | +0.054 |
choice | +0.052 |
away | +0.051 |
ook | +0.045 |
taxi | +0.044 |
Neg logit contributions | |
sauce | -0.053 |
gum | -0.052 |
greens | -0.050 |
glasses | -0.048 |
ink | -0.048 |
Token | They |
Token ID | 1119 |
Feature activation | +4.51e-03 |
Pos logit contributions | |
decision | +0.015 |
choice | +0.014 |
away | +0.014 |
ook | +0.013 |
taxi | +0.012 |
Neg logit contributions | |
gum | -0.015 |
sauce | -0.014 |
Ol | -0.014 |
greens | -0.013 |
glasses | -0.013 |
Token | looked |
Token ID | 3114 |
Feature activation | +3.10e-02 |
Pos logit contributions | |
decision | +0.009 |
choice | +0.009 |
away | +0.009 |
ook | +0.008 |
taxi | +0.008 |
Neg logit contributions | |
sauce | -0.006 |
gum | -0.005 |
greens | -0.005 |
glasses | -0.005 |
ink | -0.005 |
Token | at |
Token ID | 379 |
Feature activation | +4.75e-02 |
Pos logit contributions | |
decision | +0.057 |
choice | +0.054 |
checkpoint | +0.052 |
tha | +0.050 |
doorstep | +0.050 |
Neg logit contributions | |
sauce | -0.048 |
gum | -0.045 |
triangles | -0.043 |
spaghetti | -0.043 |
ink | -0.043 |
Token | the |
Token ID | 262 |
Feature activation | -4.86e-03 |
Pos logit contributions | |
decision | +0.084 |
choice | +0.080 |
away | +0.079 |
ook | +0.071 |
checkpoint | +0.068 |
Neg logit contributions | |
sauce | -0.075 |
gum | -0.074 |
greens | -0.069 |
glasses | -0.068 |
ink | -0.068 |
Token | terrible |
Token ID | 7818 |
Feature activation | -1.63e-02 |
Pos logit contributions | |
greens | +0.008 |
pads | +0.008 |
sauce | +0.008 |
joked | +0.008 |
gloss | +0.008 |
Neg logit contributions | |
decision | -0.008 |
choice | -0.008 |
away | -0.007 |
taxi | -0.007 |
destination | -0.006 |
Token | step |
Token ID | 2239 |
Feature activation | +2.28e-02 |
Pos logit contributions | |
greens | +0.027 |
sauce | +0.027 |
pads | +0.027 |
gum | +0.026 |
glasses | +0.026 |
Neg logit contributions | |
decision | -0.027 |
choice | -0.026 |
away | -0.025 |
taxi | -0.021 |
ook | -0.020 |
Token | and |
Token ID | 290 |
Feature activation | +6.38e-03 |
Pos logit contributions | |
decision | +0.029 |
away | +0.028 |
choice | +0.027 |
ook | +0.023 |
taxi | +0.023 |
Neg logit contributions | |
sauce | -0.045 |
gum | -0.044 |
greens | -0.044 |
glasses | -0.043 |
fiber | -0.042 |
Token | Tim |
Token ID | 5045 |
Feature activation | +1.59e-02 |
Pos logit contributions | |
away | +0.009 |
decision | +0.009 |
choice | +0.009 |
exactly | +0.008 |
sadly | +0.008 |
Neg logit contributions | |
gum | -0.011 |
sauce | -0.011 |
pill | -0.011 |
greens | -0.011 |
fiber | -0.011 |
Token | had |
Token ID | 550 |
Feature activation | -1.41e-03 |
Pos logit contributions | |
decision | +0.006 |
choice | +0.006 |
away | +0.006 |
ook | +0.005 |
taxi | +0.005 |
Neg logit contributions | |
sauce | -0.004 |
gum | -0.004 |
greens | -0.004 |
subtract | -0.003 |
triangles | -0.003 |
Token | stayed |
Token ID | 9658 |
Feature activation | +3.44e-03 |
Pos logit contributions | |
greens | +0.003 |
pill | +0.003 |
gloss | +0.003 |
sauce | +0.003 |
dots | +0.003 |
Neg logit contributions | |
exactly | -0.002 |
choice | -0.002 |
decision | -0.001 |
away | -0.001 |
where | -0.001 |
Token | away |
Token ID | 1497 |
Feature activation | +1.06e-03 |
Pos logit contributions | |
decision | +0.005 |
choice | +0.004 |
away | +0.004 |
ook | +0.004 |
taxi | +0.004 |
Neg logit contributions | |
sauce | -0.007 |
gum | -0.007 |
greens | -0.006 |
glasses | -0.006 |
ink | -0.006 |
Token | from |
Token ID | 422 |
Feature activation | +1.46e-02 |
Pos logit contributions | |
decision | +0.001 |
choice | +0.001 |
away | +0.001 |
ook | +0.001 |
taxi | +0.001 |
Neg logit contributions | |
sauce | -0.002 |
gum | -0.002 |
greens | -0.002 |
glasses | -0.002 |
subtract | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | +6.59e-04 |
Pos logit contributions | |
decision | +0.024 |
choice | +0.023 |
away | +0.023 |
ook | +0.019 |
taxi | +0.019 |
Neg logit contributions | |
sauce | -0.025 |
gum | -0.024 |
greens | -0.024 |
glasses | -0.023 |
subtract | -0.022 |
Token | tent |
Token ID | 11105 |
Feature activation | +4.30e-02 |
Pos logit contributions | |
decision | +0.001 |
choice | +0.001 |
taxi | +0.001 |
away | +0.001 |
promise | +0.001 |
Neg logit contributions | |
greens | -0.001 |
cr | -0.001 |
pads | -0.001 |
fiber | -0.001 |
sauce | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +6.63e-03 |
Pos logit contributions | |
decision | +0.058 |
away | +0.055 |
choice | +0.055 |
taxi | +0.045 |
ook | +0.044 |
Neg logit contributions | |
gum | -0.085 |
sauce | -0.084 |
greens | -0.083 |
subtract | -0.080 |
Ol | -0.079 |
Token |
|
Token ID | 198 |
Feature activation | +3.45e-03 |
Pos logit contributions | |
decision | +0.009 |
choice | +0.009 |
away | +0.009 |
ook | +0.008 |
checkpoint | +0.008 |
Neg logit contributions | |
sauce | -0.012 |
greens | -0.012 |
gum | -0.011 |
triangles | -0.011 |
gloss | -0.011 |
Token |
|
Token ID | 198 |
Feature activation | +8.24e-03 |
Pos logit contributions | |
decision | +0.007 |
choice | +0.007 |
away | +0.007 |
ook | +0.006 |
taxi | +0.006 |
Neg logit contributions | |
sauce | -0.005 |
gum | -0.004 |
greens | -0.004 |
glasses | -0.004 |
ink | -0.004 |
Token | Then |
Token ID | 6423 |
Feature activation | -4.58e-03 |
Pos logit contributions | |
decision | +0.016 |
choice | +0.016 |
away | +0.015 |
ook | +0.014 |
tha | +0.013 |
Neg logit contributions | |
sauce | -0.011 |
gum | -0.011 |
greens | -0.010 |
glasses | -0.010 |
ink | -0.010 |
Token | , |
Token ID | 11 |
Feature activation | -7.97e-03 |
Pos logit contributions | |
sauce | +0.008 |
gum | +0.008 |
greens | +0.007 |
glasses | +0.007 |
ink | +0.007 |
Neg logit contributions | |
decision | -0.007 |
choice | -0.007 |
away | -0.007 |
ook | -0.006 |
taxi | -0.006 |
Token | with |
Token ID | 351 |
Feature activation | +1.20e-02 |
Pos logit contributions | |
sauce | +0.011 |
greens | +0.010 |
gum | +0.010 |
subtract | +0.010 |
glasses | +0.010 |
Neg logit contributions | |
away | -0.007 |
decision | -0.006 |
choice | -0.006 |
taxi | -0.005 |
ook | -0.005 |
Token | Spot |
Token ID | 15899 |
Feature activation | +8.76e-03 |
Pos logit contributions | |
away | +0.015 |
choice | +0.013 |
decision | +0.013 |
taxi | +0.011 |
exactly | +0.011 |
Neg logit contributions | |
greens | -0.025 |
gloss | -0.024 |
subtract | -0.024 |
sauce | -0.024 |
pads | -0.023 |
Token | anytime |
Token ID | 17949 |
Feature activation | +7.10e-03 |
Pos logit contributions | |
decision | +0.011 |
choice | +0.011 |
away | +0.011 |
forward | +0.009 |
taxi | +0.008 |
Neg logit contributions | |
sauce | -0.018 |
greens | -0.017 |
subtract | -0.017 |
gum | -0.017 |
ink | -0.017 |
Token | you |
Token ID | 345 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
away | +0.009 |
decision | +0.009 |
choice | +0.009 |
taxi | +0.008 |
ook | +0.007 |
Neg logit contributions | |
sauce | -0.014 |
gum | -0.013 |
greens | -0.013 |
subtract | -0.013 |
gloss | -0.013 |
Token | want |
Token ID | 765 |
Feature activation | -9.29e-03 |
Pos logit contributions | |
gum | +0.029 |
sauce | +0.028 |
triangles | +0.027 |
gloss | +0.027 |
greens | +0.027 |
Neg logit contributions | |
away | -0.014 |
choice | -0.014 |
decision | -0.014 |
promise | -0.013 |
forward | -0.011 |
Token | ." |
Token ID | 526 |
Feature activation | +4.06e-03 |
Pos logit contributions | |
Ol | +0.021 |
gloss | +0.021 |
subtract | +0.021 |
sauce | +0.020 |
greens | +0.020 |
Neg logit contributions | |
away | -0.009 |
choice | -0.009 |
decision | -0.008 |
forward | -0.007 |
home | -0.007 |
Token |
|
Token ID | 198 |
Feature activation | +1.28e-02 |
Pos logit contributions | |
decision | +0.007 |
choice | +0.007 |
away | +0.007 |
ook | +0.006 |
taxi | +0.006 |
Neg logit contributions | |
sauce | -0.006 |
gum | -0.006 |
greens | -0.006 |
glasses | -0.006 |
ink | -0.006 |
Token |
|
Token ID | 198 |
Feature activation | +1.36e-02 |
Pos logit contributions | |
decision | +0.026 |
choice | +0.025 |
away | +0.025 |
ook | +0.022 |
taxi | +0.022 |
Neg logit contributions | |
sauce | -0.017 |
gum | -0.017 |
greens | -0.016 |
glasses | -0.015 |
ink | -0.015 |
Token | S |
Token ID | 50 |
Feature activation | +1.33e-02 |
Pos logit contributions | |
decision | +0.025 |
away | +0.024 |
choice | +0.024 |
ook | +0.021 |
taxi | +0.021 |
Neg logit contributions | |
sauce | -0.021 |
gum | -0.020 |
greens | -0.019 |
glasses | -0.019 |
ink | -0.018 |
Token | ara |
Token ID | 3301 |
Feature activation | -2.04e-03 |
Pos logit contributions | |
decision | +0.020 |
choice | +0.019 |
away | +0.019 |
ook | +0.016 |
checkpoint | +0.016 |
Neg logit contributions | |
sauce | -0.024 |
gum | -0.024 |
greens | -0.022 |
glasses | -0.022 |
ink | -0.022 |
Token | and |
Token ID | 290 |
Feature activation | -1.50e-02 |
Pos logit contributions | |
gum | +0.004 |
sauce | +0.004 |
greens | +0.004 |
ink | +0.004 |
gloss | +0.004 |
Neg logit contributions | |
decision | -0.002 |
choice | -0.002 |
away | -0.002 |
promise | -0.002 |
ook | -0.002 |
Token | around |
Token ID | 1088 |
Feature activation | +8.46e-03 |
Pos logit contributions | |
decision | +0.002 |
choice | +0.002 |
away | +0.002 |
ook | +0.002 |
taxi | +0.002 |
Neg logit contributions | |
sauce | -0.003 |
gum | -0.003 |
greens | -0.003 |
glasses | -0.003 |
ink | -0.003 |
Token | her |
Token ID | 607 |
Feature activation | -8.71e-03 |
Pos logit contributions | |
decision | +0.016 |
choice | +0.015 |
away | +0.014 |
ook | +0.013 |
taxi | +0.013 |
Neg logit contributions | |
sauce | -0.013 |
gum | -0.013 |
greens | -0.012 |
glasses | -0.012 |
ink | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | +2.67e-03 |
Pos logit contributions | |
sauce | +0.015 |
gum | +0.015 |
greens | +0.014 |
glasses | +0.014 |
ink | +0.014 |
Neg logit contributions | |
decision | -0.014 |
choice | -0.013 |
away | -0.013 |
ook | -0.011 |
taxi | -0.011 |
Token | |
Token ID | 220 |
Feature activation | +1.91e-02 |
Pos logit contributions | |
decision | +0.005 |
choice | +0.004 |
away | +0.004 |
ook | +0.004 |
taxi | +0.004 |
Neg logit contributions | |
sauce | -0.004 |
gum | -0.004 |
greens | -0.004 |
glasses | -0.004 |
subtract | -0.004 |
Token |
|
Token ID | 198 |
Feature activation | +1.37e-03 |
Pos logit contributions | |
decision | +0.041 |
choice | +0.039 |
away | +0.039 |
ook | +0.034 |
taxi | +0.034 |
Neg logit contributions | |
sauce | -0.023 |
gum | -0.022 |
greens | -0.021 |
glasses | -0.020 |
ink | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | +5.32e-03 |
Pos logit contributions | |
decision | +0.003 |
choice | +0.003 |
away | +0.003 |
tha | +0.003 |
taxi | +0.003 |
Neg logit contributions | |
gum | -0.001 |
sauce | -0.001 |
greens | -0.001 |
glasses | -0.001 |
pads | -0.001 |
Token | One |
Token ID | 3198 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
decision | +0.011 |
choice | +0.011 |
away | +0.011 |
ook | +0.010 |
tha | +0.010 |
Neg logit contributions | |
sauce | -0.007 |
gum | -0.006 |
greens | -0.006 |
glasses | -0.006 |
ink | -0.006 |
Token | day |
Token ID | 1110 |
Feature activation | -1.63e-05 |
Pos logit contributions | |
sauce | +0.020 |
gum | +0.018 |
ink | +0.018 |
greens | +0.017 |
glasses | +0.016 |
Neg logit contributions | |
decision | -0.030 |
choice | -0.029 |
away | -0.028 |
destination | -0.027 |
ently | -0.026 |
Token | , |
Token ID | 11 |
Feature activation | -1.27e-03 |
Pos logit contributions | |
sauce | +0.000 |
gum | +0.000 |
ink | +0.000 |
glasses | +0.000 |
greens | +0.000 |
Neg logit contributions | |
decision | -0.000 |
choice | -0.000 |
away | -0.000 |
checkpoint | -0.000 |
doorstep | -0.000 |
Token | L |
Token ID | 406 |
Feature activation | +3.14e-03 |
Pos logit contributions | |
sauce | +0.002 |
gum | +0.002 |
glasses | +0.001 |
ink | +0.001 |
greens | +0.001 |
Neg logit contributions | |
decision | -0.003 |
away | -0.003 |
choice | -0.003 |
checkpoint | -0.002 |
doorstep | -0.002 |
Token | ulu |
Token ID | 15712 |
Feature activation | +7.00e-03 |
Pos logit contributions | |
decision | +0.007 |
choice | +0.006 |
away | +0.006 |
checkpoint | +0.006 |
destination | +0.006 |
Neg logit contributions | |
sauce | -0.004 |
ink | -0.004 |
gum | -0.003 |
triangles | -0.003 |
gloss | -0.003 |
Token | was |
Token ID | 373 |
Feature activation | +7.72e-03 |
Pos logit contributions | |
decision | +0.003 |
choice | +0.003 |
away | +0.002 |
exactly | +0.002 |
sadly | +0.002 |
Neg logit contributions | |
greens | -0.003 |
sauce | -0.003 |
gum | -0.003 |
triangles | -0.003 |
subtract | -0.003 |
Token | excited |
Token ID | 6568 |
Feature activation | +2.71e-02 |
Pos logit contributions | |
decision | +0.010 |
choice | +0.010 |
away | +0.009 |
forward | +0.009 |
exactly | +0.009 |
Neg logit contributions | |
sauce | -0.016 |
greens | -0.015 |
gum | -0.015 |
ink | -0.014 |
subtract | -0.014 |
Token | to |
Token ID | 284 |
Feature activation | -6.21e-03 |
Pos logit contributions | |
decision | +0.042 |
choice | +0.039 |
away | +0.039 |
ook | +0.032 |
taxi | +0.031 |
Neg logit contributions | |
sauce | -0.050 |
gum | -0.049 |
greens | -0.048 |
glasses | -0.046 |
triangles | -0.046 |
Token | see |
Token ID | 766 |
Feature activation | -6.50e-03 |
Pos logit contributions | |
sauce | +0.011 |
gum | +0.011 |
greens | +0.010 |
glasses | +0.010 |
subtract | +0.010 |
Neg logit contributions | |
decision | -0.009 |
away | -0.009 |
choice | -0.009 |
ook | -0.008 |
taxi | -0.007 |
Token | what |
Token ID | 644 |
Feature activation | -2.57e-03 |
Pos logit contributions | |
sauce | +0.013 |
gum | +0.013 |
greens | +0.012 |
glasses | +0.012 |
ink | +0.012 |
Neg logit contributions | |
decision | -0.009 |
choice | -0.008 |
away | -0.008 |
ook | -0.007 |
taxi | -0.007 |
Token | he |
Token ID | 339 |
Feature activation | +1.40e-02 |
Pos logit contributions | |
greens | +0.005 |
lenses | +0.005 |
sauce | +0.005 |
fiber | +0.005 |
pads | +0.005 |
Neg logit contributions | |
choice | -0.003 |
decision | -0.003 |
away | -0.003 |
taxi | -0.002 |
security | -0.002 |
Token | could |
Token ID | 714 |
Feature activation | +2.04e-02 |
Pos logit contributions | |
decision | +0.024 |
choice | +0.024 |
away | +0.023 |
taxi | +0.020 |
checkpoint | +0.020 |
Neg logit contributions | |
sauce | -0.022 |
gum | -0.022 |
greens | -0.021 |
triangles | -0.021 |
ink | -0.020 |
Token | do |
Token ID | 466 |
Feature activation | +2.01e-02 |
Pos logit contributions | |
decision | +0.033 |
away | +0.032 |
choice | +0.032 |
ook | +0.028 |
taxi | +0.027 |
Neg logit contributions | |
sauce | -0.034 |
gum | -0.033 |
greens | -0.032 |
glasses | -0.031 |
subtract | -0.031 |
Token | with |
Token ID | 351 |
Feature activation | +1.10e-02 |
Pos logit contributions | |
decision | +0.032 |
choice | +0.030 |
away | +0.030 |
ook | +0.026 |
taxi | +0.026 |
Neg logit contributions | |
sauce | -0.035 |
gum | -0.034 |
greens | -0.033 |
glasses | -0.032 |
ink | -0.031 |
Token | his |
Token ID | 465 |
Feature activation | +6.71e-03 |
Pos logit contributions | |
decision | +0.013 |
choice | +0.013 |
away | +0.012 |
exactly | +0.011 |
taxi | +0.010 |
Neg logit contributions | |
sauce | -0.022 |
greens | -0.022 |
Ol | -0.021 |
pads | -0.021 |
gloss | -0.021 |
Token | lemon |
Token ID | 18873 |
Feature activation | -2.81e-02 |
Pos logit contributions | |
decision | +0.009 |
choice | +0.009 |
taxi | +0.007 |
promise | +0.007 |
purchase | +0.007 |
Neg logit contributions | |
sauce | -0.014 |
greens | -0.013 |
Ol | -0.013 |
smoked | -0.013 |
gum | -0.013 |
Token | the |
Token ID | 262 |
Feature activation | +4.92e-04 |
Pos logit contributions | |
decision | +0.023 |
away | +0.022 |
choice | +0.022 |
forward | +0.019 |
ively | +0.018 |
Neg logit contributions | |
gum | -0.039 |
fiber | -0.038 |
sauce | -0.038 |
pads | -0.038 |
Ol | -0.036 |
Token | room |
Token ID | 2119 |
Feature activation | +6.24e-03 |
Pos logit contributions | |
decision | +0.001 |
choice | +0.001 |
away | +0.001 |
ook | +0.001 |
taxi | +0.001 |
Neg logit contributions | |
sauce | -0.001 |
gum | -0.001 |
greens | -0.001 |
glasses | -0.001 |
subtract | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -4.68e-03 |
Pos logit contributions | |
decision | +0.008 |
choice | +0.007 |
away | +0.007 |
ook | +0.006 |
doorstep | +0.006 |
Neg logit contributions | |
gum | -0.013 |
sauce | -0.013 |
fiber | -0.012 |
Ol | -0.012 |
greens | -0.012 |
Token | Kate |
Token ID | 16693 |
Feature activation | +1.35e-02 |
Pos logit contributions | |
gum | +0.008 |
sauce | +0.008 |
mustard | +0.008 |
glasses | +0.008 |
pads | +0.008 |
Neg logit contributions | |
decision | -0.007 |
choice | -0.007 |
away | -0.006 |
ook | -0.006 |
doorstep | -0.006 |
Token | was |
Token ID | 373 |
Feature activation | +7.13e-03 |
Pos logit contributions | |
decision | +0.020 |
away | +0.020 |
choice | +0.019 |
ook | +0.017 |
sadly | +0.017 |
Neg logit contributions | |
sauce | -0.024 |
gum | -0.023 |
greens | -0.023 |
glasses | -0.022 |
fiber | -0.021 |
Token | so |
Token ID | 523 |
Feature activation | +1.32e-02 |
Pos logit contributions | |
decision | +0.011 |
choice | +0.010 |
away | +0.010 |
ook | +0.009 |
taxi | +0.009 |
Neg logit contributions | |
sauce | -0.013 |
gum | -0.013 |
greens | -0.012 |
glasses | -0.012 |
ink | -0.012 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.44e-02 |
Pos logit contributions | |
decision | +0.019 |
choice | +0.018 |
away | +0.018 |
ook | +0.015 |
taxi | +0.015 |
Neg logit contributions | |
sauce | -0.025 |
gum | -0.025 |
greens | -0.024 |
glasses | -0.023 |
ink | -0.023 |
Token | with |
Token ID | 351 |
Feature activation | +8.12e-03 |
Pos logit contributions | |
decision | +0.019 |
choice | +0.018 |
away | +0.018 |
ook | +0.015 |
taxi | +0.015 |
Neg logit contributions | |
sauce | -0.028 |
gum | -0.028 |
greens | -0.027 |
glasses | -0.027 |
subtract | -0.026 |
Token | her |
Token ID | 607 |
Feature activation | +5.99e-03 |
Pos logit contributions | |
decision | +0.015 |
away | +0.015 |
choice | +0.015 |
ook | +0.014 |
taxi | +0.013 |
Neg logit contributions | |
sauce | -0.012 |
gum | -0.012 |
greens | -0.011 |
ink | -0.011 |
glasses | -0.011 |
Token | new |
Token ID | 649 |
Feature activation | -8.70e-04 |
Pos logit contributions | |
away | +0.010 |
decision | +0.009 |
choice | +0.009 |
ook | +0.008 |
taxi | +0.008 |
Neg logit contributions | |
sauce | -0.010 |
gum | -0.010 |
ink | -0.010 |
triangles | -0.009 |
glasses | -0.009 |
Token | circle |
Token ID | 9197 |
Feature activation | -1.12e-02 |
Pos logit contributions | |
sauce | +0.001 |
gum | +0.001 |
triangles | +0.001 |
ink | +0.001 |
glasses | +0.001 |
Neg logit contributions | |
decision | -0.002 |
away | -0.002 |
choice | -0.002 |
tha | -0.002 |
ively | -0.002 |
Token | but |
Token ID | 475 |
Feature activation | +3.80e-02 |
Pos logit contributions | |
sauce | +0.005 |
gum | +0.005 |
Ol | +0.005 |
greens | +0.005 |
glasses | +0.005 |
Neg logit contributions | |
decision | -0.003 |
away | -0.003 |
choice | -0.003 |
exactly | -0.002 |
forward | -0.002 |
Token | they |
Token ID | 484 |
Feature activation | +8.29e-03 |
Pos logit contributions | |
decision | +0.050 |
choice | +0.046 |
away | +0.044 |
exactly | +0.037 |
promise | +0.037 |
Neg logit contributions | |
sauce | -0.075 |
glasses | -0.074 |
gum | -0.073 |
greens | -0.073 |
pads | -0.073 |
Token | are |
Token ID | 389 |
Feature activation | -6.58e-03 |
Pos logit contributions | |
decision | +0.012 |
choice | +0.011 |
away | +0.011 |
promise | +0.010 |
ook | +0.009 |
Neg logit contributions | |
gum | -0.016 |
sauce | -0.016 |
greens | -0.015 |
glasses | -0.015 |
gloss | -0.015 |
Token | not |
Token ID | 407 |
Feature activation | +4.51e-03 |
Pos logit contributions | |
gloss | +0.014 |
sauce | +0.013 |
gum | +0.013 |
Ol | +0.013 |
greens | +0.013 |
Neg logit contributions | |
decision | -0.008 |
choice | -0.007 |
away | -0.007 |
exactly | -0.007 |
forward | -0.006 |
Token | always |
Token ID | 1464 |
Feature activation | -1.15e-02 |
Pos logit contributions | |
decision | +0.006 |
choice | +0.006 |
away | +0.006 |
exactly | +0.005 |
prisoner | +0.005 |
Neg logit contributions | |
gloss | -0.009 |
Ol | -0.009 |
sauce | -0.009 |
mascara | -0.009 |
gum | -0.009 |
Token | bad |
Token ID | 2089 |
Feature activation | +6.96e-03 |
Pos logit contributions | |
gloss | +0.022 |
Ol | +0.022 |
sauce | +0.022 |
gum | +0.022 |
greens | +0.021 |
Neg logit contributions | |
decision | -0.017 |
choice | -0.016 |
away | -0.016 |
promise | -0.013 |
forward | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | +4.13e-03 |
Pos logit contributions | |
decision | +0.009 |
choice | +0.009 |
away | +0.008 |
forward | +0.006 |
deed | +0.006 |
Neg logit contributions | |
Ol | -0.015 |
sauce | -0.014 |
gum | -0.014 |
greens | -0.014 |
gloss | -0.014 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.64e-02 |
Pos logit contributions | |
decision | +0.006 |
choice | +0.006 |
away | +0.006 |
ook | +0.005 |
checkpoint | +0.005 |
Neg logit contributions | |
sauce | -0.007 |
gum | -0.007 |
greens | -0.007 |
glasses | -0.007 |
pads | -0.007 |
Token | Once |
Token ID | 7454 |
Feature activation | +2.36e-02 |
Pos logit contributions | |
greens | +0.047 |
subtract | +0.047 |
sauce | +0.046 |
fiber | +0.046 |
gloss | +0.045 |
Neg logit contributions | |
away | -0.035 |
choice | -0.035 |
decision | -0.034 |
ook | -0.032 |
forward | -0.031 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.48e-02 |
Pos logit contributions | |
decision | +0.042 |
away | +0.041 |
choice | +0.040 |
doorstep | +0.038 |
checkpoint | +0.037 |
Neg logit contributions | |
sauce | -0.033 |
gum | -0.032 |
triangles | -0.032 |
ink | -0.030 |
Ben | -0.030 |
Token | a |
Token ID | 257 |
Feature activation | -8.67e-03 |
Pos logit contributions | |
sauce | +0.020 |
gum | +0.019 |
greens | +0.019 |
subtract | +0.018 |
glasses | +0.017 |
Neg logit contributions | |
decision | -0.029 |
choice | -0.028 |
away | -0.028 |
ook | -0.024 |
taxi | -0.024 |
Token | my |
Token ID | 616 |
Feature activation | -2.46e-04 |
Pos logit contributions | |
sauce | +0.028 |
greens | +0.027 |
pads | +0.027 |
gum | +0.026 |
triangles | +0.026 |
Neg logit contributions | |
decision | -0.019 |
choice | -0.018 |
away | -0.017 |
taxi | -0.015 |
tha | -0.014 |
Token | skirt |
Token ID | 23967 |
Feature activation | -1.73e-02 |
Pos logit contributions | |
sauce | +0.000 |
smoked | +0.000 |
greens | +0.000 |
Ol | +0.000 |
subtract | +0.000 |
Neg logit contributions | |
decision | -0.000 |
choice | -0.000 |
promise | -0.000 |
away | -0.000 |
doorstep | -0.000 |
Token | go |
Token ID | 467 |
Feature activation | -8.71e-03 |
Pos logit contributions | |
sauce | +0.035 |
greens | +0.035 |
gum | +0.034 |
triangles | +0.033 |
glasses | +0.033 |
Neg logit contributions | |
decision | -0.023 |
choice | -0.022 |
away | -0.021 |
taxi | -0.018 |
destination | -0.017 |
Token | ?" |
Token ID | 1701 |
Feature activation | -1.38e-02 |
Pos logit contributions | |
sauce | +0.017 |
gum | +0.016 |
greens | +0.016 |
glasses | +0.015 |
triangles | +0.015 |
Neg logit contributions | |
decision | -0.013 |
away | -0.012 |
choice | -0.012 |
taxi | -0.010 |
ook | -0.010 |
Token | she |
Token ID | 673 |
Feature activation | -3.01e-02 |
Pos logit contributions | |
sauce | +0.024 |
gum | +0.023 |
pads | +0.022 |
greens | +0.022 |
fiber | +0.022 |
Neg logit contributions | |
decision | -0.021 |
choice | -0.021 |
away | -0.021 |
ook | -0.019 |
ively | -0.019 |
Token | cried |
Token ID | 16896 |
Feature activation | -2.70e-03 |
Pos logit contributions | |
sauce | +0.057 |
gum | +0.056 |
pill | +0.054 |
fiber | +0.054 |
greens | +0.053 |
Neg logit contributions | |
sadly | -0.040 |
curiously | -0.040 |
decision | -0.037 |
choice | -0.036 |
away | -0.034 |
Token | . |
Token ID | 13 |
Feature activation | -2.06e-02 |
Pos logit contributions | |
sauce | +0.004 |
gum | +0.004 |
greens | +0.004 |
glasses | +0.004 |
ink | +0.004 |
Neg logit contributions | |
decision | -0.004 |
choice | -0.004 |
away | -0.004 |
ook | -0.004 |
taxi | -0.004 |
Token |
|
Token ID | 198 |
Feature activation | +6.33e-04 |
Pos logit contributions | |
sauce | +0.041 |
gum | +0.039 |
fiber | +0.037 |
pads | +0.037 |
noodles | +0.037 |
Neg logit contributions | |
decision | -0.025 |
ook | -0.024 |
ively | -0.024 |
away | -0.024 |
choice | -0.024 |
Token |
|
Token ID | 198 |
Feature activation | +2.03e-03 |
Pos logit contributions | |
decision | +0.001 |
away | +0.001 |
choice | +0.001 |
ook | +0.001 |
taxi | +0.001 |
Neg logit contributions | |
sauce | -0.001 |
gum | -0.001 |
greens | -0.001 |
glasses | -0.001 |
pads | -0.001 |
Token | M |
Token ID | 44 |
Feature activation | +2.07e-03 |
Pos logit contributions | |
decision | +0.003 |
away | +0.003 |
choice | +0.003 |
ook | +0.003 |
taxi | +0.003 |
Neg logit contributions | |
sauce | -0.003 |
gum | -0.003 |
greens | -0.003 |
glasses | -0.003 |
ink | -0.003 |
Token | ummy |
Token ID | 13513 |
Feature activation | -3.35e-02 |
Pos logit contributions | |
decision | +0.003 |
away | +0.003 |
choice | +0.003 |
ook | +0.003 |
taxi | +0.003 |
Neg logit contributions | |
sauce | -0.004 |
gum | -0.004 |
ink | -0.003 |
subtract | -0.003 |
glasses | -0.003 |
Token | had |
Token ID | 550 |
Feature activation | +3.92e-03 |
Pos logit contributions | |
gum | +0.029 |
greens | +0.029 |
sauce | +0.028 |
subtract | +0.028 |
triangles | +0.026 |
Neg logit contributions | |
decision | -0.028 |
away | -0.028 |
choice | -0.027 |
taxi | -0.023 |
forward | -0.023 |
Token | never |
Token ID | 1239 |
Feature activation | +1.17e-02 |
Pos logit contributions | |
decision | +0.005 |
choice | +0.005 |
away | +0.005 |
taxi | +0.004 |
exactly | +0.004 |
Neg logit contributions | |
greens | -0.008 |
sauce | -0.008 |
gum | -0.007 |
gloss | -0.007 |
Ol | -0.007 |
Token | been |
Token ID | 587 |
Feature activation | +9.39e-03 |
Pos logit contributions | |
away | +0.018 |
decision | +0.018 |
choice | +0.017 |
checkpoint | +0.015 |
home | +0.014 |
Neg logit contributions | |
greens | -0.021 |
sauce | -0.020 |
gum | -0.020 |
pill | -0.018 |
fiber | -0.018 |
Token | raised |
Token ID | 4376 |
Feature activation | +1.50e-02 |
Pos logit contributions | |
away | +0.012 |
decision | +0.012 |
choice | +0.011 |
forward | +0.010 |
ook | +0.010 |
Neg logit contributions | |
greens | -0.019 |
sauce | -0.019 |
gum | -0.018 |
subtract | -0.018 |
fiber | -0.017 |
Token | up |
Token ID | 510 |
Feature activation | +1.68e-02 |
Pos logit contributions | |
away | +0.021 |
decision | +0.020 |
choice | +0.019 |
forward | +0.017 |
exactly | +0.016 |
Neg logit contributions | |
sauce | -0.030 |
greens | -0.030 |
gum | -0.029 |
subtract | -0.029 |
gloss | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | -6.72e-03 |
Pos logit contributions | |
away | +0.023 |
decision | +0.023 |
choice | +0.022 |
forward | +0.018 |
ook | +0.018 |
Neg logit contributions | |
sauce | -0.033 |
gum | -0.033 |
greens | -0.032 |
subtract | -0.032 |
gloss | -0.031 |
Token | But |
Token ID | 887 |
Feature activation | +3.82e-02 |
Pos logit contributions | |
sauce | +0.014 |
gloss | +0.013 |
gum | +0.013 |
mustard | +0.013 |
pads | +0.013 |
Neg logit contributions | |
decision | -0.007 |
taxi | -0.006 |
ook | -0.006 |
checkpoint | -0.006 |
choice | -0.006 |
Token | one |
Token ID | 530 |
Feature activation | +1.57e-05 |
Pos logit contributions | |
decision | +0.055 |
choice | +0.052 |
away | +0.052 |
ook | +0.044 |
taxi | +0.044 |
Neg logit contributions | |
sauce | -0.073 |
gum | -0.071 |
greens | -0.069 |
glasses | -0.067 |
subtract | -0.066 |
Token | day |
Token ID | 1110 |
Feature activation | -4.77e-03 |
Pos logit contributions | |
decision | +0.000 |
choice | +0.000 |
away | +0.000 |
ook | +0.000 |
tha | +0.000 |
Neg logit contributions | |
sauce | -0.000 |
gum | -0.000 |
greens | -0.000 |
ink | -0.000 |
glasses | -0.000 |
Token | , |
Token ID | 11 |
Feature activation | -1.52e-03 |
Pos logit contributions | |
sauce | +0.010 |
gum | +0.009 |
greens | +0.009 |
glasses | +0.009 |
subtract | +0.009 |
Neg logit contributions | |
decision | -0.006 |
choice | -0.006 |
away | -0.006 |
ook | -0.005 |
taxi | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | -4.69e-03 |
Pos logit contributions | |
sauce | +0.002 |
pads | +0.002 |
greens | +0.002 |
gum | +0.002 |
gloss | +0.002 |
Neg logit contributions | |
decision | -0.002 |
choice | -0.002 |
away | -0.002 |
taxi | -0.002 |
exactly | -0.002 |
Token | rules |
Token ID | 3173 |
Feature activation | +7.89e-03 |
Pos logit contributions | |
pads | +0.006 |
sauce | +0.006 |
greens | +0.006 |
gloss | +0.006 |
gum | +0.006 |
Neg logit contributions | |
decision | -0.005 |
choice | -0.005 |
away | -0.004 |
taxi | -0.004 |
hotel | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | +7.72e-03 |
Pos logit contributions | |
decision | +0.010 |
away | +0.010 |
choice | +0.010 |
ook | +0.008 |
forward | +0.008 |
Neg logit contributions | |
sauce | -0.016 |
gum | -0.016 |
greens | -0.015 |
glasses | -0.015 |
ink | -0.015 |
Token | Too |
Token ID | 14190 |
Feature activation | -4.58e-02 |
Pos logit contributions | |
decision | +0.011 |
away | +0.011 |
choice | +0.011 |
ook | +0.010 |
checkpoint | +0.010 |
Neg logit contributions | |
sauce | -0.014 |
gum | -0.013 |
greens | -0.013 |
triangles | -0.012 |
glasses | -0.012 |
Token | much |
Token ID | 881 |
Feature activation | -4.56e-02 |
Pos logit contributions | |
greens | +0.099 |
sauce | +0.096 |
subtract | +0.093 |
gloss | +0.092 |
triangles | +0.091 |
Neg logit contributions | |
decision | -0.055 |
away | -0.054 |
choice | -0.054 |
taxi | -0.046 |
badly | -0.040 |
Token | food |
Token ID | 2057 |
Feature activation | -2.26e-02 |
Pos logit contributions | |
Ol | +0.089 |
gloss | +0.081 |
greens | +0.079 |
Ted | +0.078 |
Carl | +0.077 |
Neg logit contributions | |
choice | -0.069 |
decision | -0.066 |
away | -0.055 |
promise | -0.052 |
taxi | -0.050 |
Token | can |
Token ID | 460 |
Feature activation | -3.60e-03 |
Pos logit contributions | |
greens | +0.047 |
Ol | +0.046 |
gloss | +0.046 |
subtract | +0.046 |
sauce | +0.046 |
Neg logit contributions | |
decision | -0.027 |
choice | -0.027 |
away | -0.025 |
promise | -0.021 |
taxi | -0.020 |
Token | hurt |
Token ID | 5938 |
Feature activation | -1.72e-03 |
Pos logit contributions | |
gum | +0.007 |
sauce | +0.007 |
triangles | +0.006 |
greens | +0.006 |
gloss | +0.006 |
Neg logit contributions | |
decision | -0.005 |
choice | -0.005 |
away | -0.005 |
promise | -0.004 |
taxi | -0.004 |
Token | the |
Token ID | 262 |
Feature activation | +1.04e-02 |
Pos logit contributions | |
gloss | +0.004 |
greens | +0.004 |
subtract | +0.004 |
sauce | +0.004 |
gum | +0.004 |
Neg logit contributions | |
away | -0.002 |
decision | -0.002 |
choice | -0.002 |
ively | -0.001 |
ook | -0.001 |
Token | fish |
Token ID | 5916 |
Feature activation | -1.11e-03 |
Pos logit contributions | |
decision | +0.018 |
choice | +0.017 |
away | +0.015 |
taxi | +0.014 |
hotel | +0.014 |
Neg logit contributions | |
subtract | -0.018 |
greens | -0.017 |
gloss | -0.017 |
Ol | -0.017 |
joked | -0.016 |
Token | . |
Token ID | 13 |
Feature activation | +4.30e-03 |
Pos logit contributions | |
greens | +0.002 |
gum | +0.002 |
subtract | +0.002 |
sauce | +0.002 |
glasses | +0.002 |
Neg logit contributions | |
decision | -0.001 |
away | -0.001 |
choice | -0.001 |
taxi | -0.001 |
promise | -0.001 |
Token | Changing |
Token ID | 33680 |
Feature activation | -1.97e-02 |
Pos logit contributions | |
decision | +0.005 |
choice | +0.005 |
ook | +0.005 |
away | +0.005 |
checkpoint | +0.005 |
Neg logit contributions | |
sauce | -0.008 |
gum | -0.007 |
greens | -0.007 |
triangles | -0.007 |
glasses | -0.007 |
Token | huge |
Token ID | 3236 |
Feature activation | +1.28e-03 |
Pos logit contributions | |
sauce | +0.003 |
gum | +0.003 |
greens | +0.003 |
ink | +0.003 |
glasses | +0.003 |
Neg logit contributions | |
decision | -0.004 |
choice | -0.003 |
away | -0.003 |
ively | -0.003 |
ently | -0.003 |
Token | tree |
Token ID | 5509 |
Feature activation | +1.49e-02 |
Pos logit contributions | |
decision | +0.002 |
choice | +0.002 |
away | +0.002 |
ook | +0.002 |
taxi | +0.002 |
Neg logit contributions | |
sauce | -0.002 |
gum | -0.002 |
greens | -0.002 |
glasses | -0.002 |
ink | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | -3.62e-03 |
Pos logit contributions | |
decision | +0.024 |
choice | +0.023 |
away | +0.023 |
ook | +0.019 |
tha | +0.019 |
Neg logit contributions | |
sauce | -0.026 |
gum | -0.025 |
greens | -0.023 |
ink | -0.023 |
glasses | -0.023 |
Token | A |
Token ID | 317 |
Feature activation | -2.07e-02 |
Pos logit contributions | |
pads | +0.007 |
gloss | +0.007 |
gum | +0.007 |
sauce | +0.007 |
Ol | +0.007 |
Neg logit contributions | |
taxi | -0.004 |
decision | -0.004 |
choice | -0.004 |
away | -0.004 |
ook | -0.003 |
Token | little |
Token ID | 1310 |
Feature activation | +6.88e-03 |
Pos logit contributions | |
sauce | +0.039 |
greens | +0.038 |
gum | +0.038 |
subtract | +0.037 |
pads | +0.037 |
Neg logit contributions | |
decision | -0.029 |
choice | -0.028 |
away | -0.027 |
taxi | -0.025 |
ook | -0.023 |
Token | boy |
Token ID | 2933 |
Feature activation | +2.99e-04 |
Pos logit contributions | |
decision | +0.008 |
choice | +0.008 |
away | +0.008 |
taxi | +0.007 |
hotel | +0.006 |
Neg logit contributions | |
subtract | -0.014 |
sauce | -0.014 |
greens | -0.014 |
gum | -0.013 |
glasses | -0.013 |
Token | named |
Token ID | 3706 |
Feature activation | -2.06e-02 |
Pos logit contributions | |
decision | +0.000 |
choice | +0.000 |
away | +0.000 |
ook | +0.000 |
taxi | +0.000 |
Neg logit contributions | |
sauce | -0.001 |
gum | -0.001 |
greens | -0.001 |
glasses | -0.001 |
subtract | -0.001 |
Token | Tim |
Token ID | 5045 |
Feature activation | +2.05e-02 |
Pos logit contributions | |
sauce | +0.024 |
gum | +0.023 |
greens | +0.021 |
glasses | +0.020 |
ink | +0.020 |
Neg logit contributions | |
decision | -0.047 |
choice | -0.045 |
away | -0.044 |
ook | -0.039 |
taxi | -0.039 |
Token | went |
Token ID | 1816 |
Feature activation | +7.85e-03 |
Pos logit contributions | |
decision | +0.037 |
choice | +0.036 |
away | +0.036 |
ook | +0.032 |
taxi | +0.031 |
Neg logit contributions | |
sauce | -0.030 |
gum | -0.030 |
greens | -0.030 |
subtract | -0.028 |
glasses | -0.028 |
Token | to |
Token ID | 284 |
Feature activation | +2.57e-05 |
Pos logit contributions | |
decision | +0.013 |
choice | +0.013 |
away | +0.012 |
ook | +0.011 |
tha | +0.011 |
Neg logit contributions | |
sauce | -0.013 |
gum | -0.013 |
greens | -0.012 |
glasses | -0.012 |
ink | -0.012 |
Token | play |
Token ID | 711 |
Feature activation | -2.91e-03 |
Pos logit contributions | |
decision | +0.000 |
choice | +0.000 |
away | +0.000 |
taxi | +0.000 |
ook | +0.000 |
Neg logit contributions | |
sauce | -0.000 |
gum | -0.000 |
greens | -0.000 |
glasses | -0.000 |
pads | -0.000 |
Token | didn |
Token ID | 1422 |
Feature activation | +1.20e-02 |
Pos logit contributions | |
sauce | +0.017 |
gum | +0.017 |
greens | +0.016 |
glasses | +0.015 |
pads | +0.015 |
Neg logit contributions | |
decision | -0.023 |
choice | -0.022 |
away | -0.022 |
ook | -0.020 |
taxi | -0.020 |
Token | 't |
Token ID | 470 |
Feature activation | +2.47e-02 |
Pos logit contributions | |
decision | +0.027 |
away | +0.026 |
choice | +0.026 |
tha | +0.023 |
ook | +0.023 |
Neg logit contributions | |
gum | -0.014 |
sauce | -0.013 |
subtract | -0.012 |
ink | -0.012 |
Ol | -0.011 |
Token | know |
Token ID | 760 |
Feature activation | +2.21e-03 |
Pos logit contributions | |
decision | +0.041 |
choice | +0.039 |
away | +0.039 |
ook | +0.034 |
taxi | +0.034 |
Neg logit contributions | |
sauce | -0.041 |
gum | -0.040 |
greens | -0.038 |
glasses | -0.038 |
ink | -0.037 |
Token | how |
Token ID | 703 |
Feature activation | +1.08e-04 |
Pos logit contributions | |
decision | +0.003 |
away | +0.003 |
choice | +0.003 |
ook | +0.003 |
doorstep | +0.003 |
Neg logit contributions | |
sauce | -0.004 |
gum | -0.004 |
ink | -0.004 |
triangles | -0.004 |
greens | -0.004 |
Token | to |
Token ID | 284 |
Feature activation | +3.10e-03 |
Pos logit contributions | |
decision | +0.000 |
away | +0.000 |
choice | +0.000 |
ook | +0.000 |
taxi | +0.000 |
Neg logit contributions | |
sauce | -0.000 |
gum | -0.000 |
greens | -0.000 |
glasses | -0.000 |
pads | -0.000 |
Token | fix |
Token ID | 4259 |
Feature activation | -9.08e-03 |
Pos logit contributions | |
decision | +0.006 |
choice | +0.005 |
away | +0.005 |
ook | +0.005 |
forward | +0.005 |
Neg logit contributions | |
sauce | -0.005 |
gum | -0.005 |
subtract | -0.004 |
ink | -0.004 |
greens | -0.004 |
Token | it |
Token ID | 340 |
Feature activation | -1.78e-02 |
Pos logit contributions | |
sauce | +0.018 |
gum | +0.017 |
greens | +0.017 |
gloss | +0.017 |
pads | +0.017 |
Neg logit contributions | |
decision | -0.012 |
choice | -0.012 |
away | -0.012 |
ook | -0.009 |
taxi | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | -2.90e-03 |
Pos logit contributions | |
sauce | +0.035 |
gum | +0.035 |
greens | +0.034 |
glasses | +0.033 |
Ol | +0.033 |
Neg logit contributions | |
decision | -0.024 |
choice | -0.023 |
away | -0.023 |
ook | -0.018 |
taxi | -0.018 |
Token | Then |
Token ID | 3244 |
Feature activation | -1.07e-02 |
Pos logit contributions | |
gum | +0.006 |
sauce | +0.006 |
pads | +0.006 |
gloss | +0.006 |
Ol | +0.006 |
Neg logit contributions | |
decision | -0.003 |
choice | -0.003 |
away | -0.003 |
taxi | -0.003 |
doorstep | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | -9.72e-03 |
Pos logit contributions | |
sauce | +0.019 |
gum | +0.018 |
greens | +0.017 |
glasses | +0.017 |
ink | +0.017 |
Neg logit contributions | |
decision | -0.017 |
away | -0.016 |
choice | -0.016 |
ook | -0.014 |
checkpoint | -0.014 |
Token | wise |
Token ID | 10787 |
Feature activation | -1.74e-02 |
Pos logit contributions | |
pads | +0.018 |
sauce | +0.018 |
greens | +0.017 |
gum | +0.017 |
subtract | +0.017 |
Neg logit contributions | |
decision | -0.015 |
choice | -0.014 |
away | -0.013 |
taxi | -0.012 |
promise | -0.011 |
Token |
|
Token ID | 198 |
Feature activation | -3.99e-03 |
Pos logit contributions | |
sauce | +0.015 |
gum | +0.014 |
glasses | +0.014 |
triangles | +0.014 |
greens | +0.014 |
Neg logit contributions | |
decision | -0.015 |
choice | -0.014 |
away | -0.014 |
ook | -0.013 |
forward | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -1.53e-03 |
Pos logit contributions | |
sauce | +0.006 |
gum | +0.006 |
greens | +0.005 |
glasses | +0.005 |
triangles | +0.005 |
Neg logit contributions | |
decision | -0.007 |
choice | -0.007 |
away | -0.007 |
ook | -0.006 |
taxi | -0.006 |
Token | Billy |
Token ID | 49640 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
sauce | +0.002 |
gum | +0.002 |
greens | +0.002 |
glasses | +0.002 |
ink | +0.002 |
Neg logit contributions | |
decision | -0.003 |
choice | -0.003 |
away | -0.003 |
ook | -0.003 |
taxi | -0.003 |
Token | made |
Token ID | 925 |
Feature activation | -5.58e-03 |
Pos logit contributions | |
decision | +0.019 |
away | +0.018 |
choice | +0.017 |
sadly | +0.016 |
tha | +0.015 |
Neg logit contributions | |
gum | -0.028 |
sauce | -0.028 |
greens | -0.026 |
subtract | -0.026 |
mascara | -0.025 |
Token | a |
Token ID | 257 |
Feature activation | +1.24e-02 |
Pos logit contributions | |
sauce | +0.009 |
gum | +0.009 |
greens | +0.009 |
glasses | +0.009 |
subtract | +0.008 |
Neg logit contributions | |
decision | -0.009 |
choice | -0.009 |
away | -0.009 |
ook | -0.008 |
taxi | -0.007 |
Token | lot |
Token ID | 1256 |
Feature activation | -3.80e-02 |
Pos logit contributions | |
decision | +0.017 |
away | +0.017 |
choice | +0.016 |
ook | +0.014 |
ively | +0.014 |
Neg logit contributions | |
sauce | -0.024 |
gum | -0.023 |
greens | -0.022 |
glasses | -0.022 |
ink | -0.021 |
Token | of |
Token ID | 286 |
Feature activation | -1.62e-02 |
Pos logit contributions | |
sauce | +0.100 |
greens | +0.099 |
gum | +0.099 |
glasses | +0.097 |
pads | +0.096 |
Neg logit contributions | |
decision | -0.025 |
away | -0.024 |
choice | -0.023 |
ook | -0.017 |
taxi | -0.014 |
Token | new |
Token ID | 649 |
Feature activation | -2.30e-03 |
Pos logit contributions | |
sauce | +0.028 |
gloss | +0.028 |
pads | +0.028 |
Ol | +0.028 |
greens | +0.028 |
Neg logit contributions | |
decision | -0.027 |
choice | -0.026 |
away | -0.022 |
promise | -0.019 |
taxi | -0.019 |
Token | friends |
Token ID | 2460 |
Feature activation | +2.33e-03 |
Pos logit contributions | |
sauce | +0.004 |
gum | +0.004 |
greens | +0.004 |
subtract | +0.004 |
Ol | +0.004 |
Neg logit contributions | |
decision | -0.003 |
choice | -0.003 |
away | -0.003 |
ook | -0.003 |
taxi | -0.002 |
Token | that |
Token ID | 326 |
Feature activation | +5.31e-03 |
Pos logit contributions | |
decision | +0.003 |
away | +0.003 |
choice | +0.003 |
ook | +0.002 |
forward | +0.002 |
Neg logit contributions | |
sauce | -0.005 |
gum | -0.005 |
greens | -0.005 |
glasses | -0.004 |
subtract | -0.004 |
Token | day |
Token ID | 1110 |
Feature activation | -1.73e-03 |
Pos logit contributions | |
decision | +0.006 |
choice | +0.006 |
away | +0.006 |
forward | +0.005 |
taxi | +0.005 |
Neg logit contributions | |
subtract | -0.011 |
mascara | -0.011 |
brist | -0.011 |
lenses | -0.011 |
greens | -0.011 |
Token | studied |
Token ID | 9713 |
Feature activation | +1.61e-02 |
Pos logit contributions | |
choice | +0.038 |
decision | +0.037 |
away | +0.034 |
promise | +0.032 |
ook | +0.030 |
Neg logit contributions | |
gum | -0.039 |
sauce | -0.038 |
greens | -0.038 |
triangles | -0.037 |
squares | -0.035 |
Token | and |
Token ID | 290 |
Feature activation | -3.40e-03 |
Pos logit contributions | |
away | +0.022 |
decision | +0.022 |
choice | +0.022 |
forward | +0.018 |
exactly | +0.018 |
Neg logit contributions | |
sauce | -0.031 |
gum | -0.029 |
greens | -0.029 |
Ol | -0.028 |
gloss | -0.028 |
Token | I |
Token ID | 314 |
Feature activation | +1.82e-02 |
Pos logit contributions | |
sauce | +0.006 |
gum | +0.006 |
greens | +0.006 |
Ol | +0.006 |
pads | +0.006 |
Neg logit contributions | |
decision | -0.005 |
choice | -0.005 |
away | -0.004 |
checkpoint | -0.004 |
promise | -0.004 |
Token | learned |
Token ID | 4499 |
Feature activation | -3.97e-03 |
Pos logit contributions | |
choice | +0.029 |
decision | +0.028 |
promise | +0.025 |
away | +0.025 |
purchase | +0.023 |
Neg logit contributions | |
gum | -0.033 |
sauce | -0.032 |
greens | -0.032 |
triangles | -0.032 |
squares | -0.030 |
Token | a |
Token ID | 257 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
sauce | +0.007 |
greens | +0.007 |
gum | +0.007 |
pads | +0.007 |
Ol | +0.007 |
Neg logit contributions | |
decision | -0.006 |
choice | -0.006 |
away | -0.006 |
exactly | -0.005 |
taxi | -0.005 |
Token | lot |
Token ID | 1256 |
Feature activation | -3.45e-02 |
Pos logit contributions | |
decision | +0.020 |
away | +0.020 |
choice | +0.019 |
ook | +0.017 |
tha | +0.017 |
Neg logit contributions | |
sauce | -0.026 |
gum | -0.026 |
greens | -0.024 |
glasses | -0.024 |
ink | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | +9.31e-03 |
Pos logit contributions | |
sauce | +0.073 |
gum | +0.072 |
greens | +0.069 |
glasses | +0.068 |
ink | +0.067 |
Neg logit contributions | |
decision | -0.042 |
choice | -0.040 |
away | -0.040 |
ook | -0.032 |
taxi | -0.031 |
Token | Can |
Token ID | 1680 |
Feature activation | +5.65e-03 |
Pos logit contributions | |
decision | +0.013 |
choice | +0.012 |
away | +0.012 |
ook | +0.012 |
checkpoint | +0.011 |
Neg logit contributions | |
sauce | -0.017 |
gum | -0.016 |
ink | -0.015 |
greens | -0.015 |
glasses | -0.015 |
Token | I |
Token ID | 314 |
Feature activation | +2.07e-03 |
Pos logit contributions | |
decision | +0.010 |
choice | +0.010 |
away | +0.009 |
ook | +0.008 |
taxi | +0.008 |
Neg logit contributions | |
sauce | -0.009 |
gum | -0.009 |
greens | -0.008 |
glasses | -0.008 |
ink | -0.008 |
Token | have |
Token ID | 423 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
decision | +0.003 |
choice | +0.003 |
away | +0.003 |
ook | +0.002 |
taxi | +0.002 |
Neg logit contributions | |
sauce | -0.004 |
gum | -0.004 |
greens | -0.004 |
triangles | -0.004 |
pads | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.74e-02 |
Pos logit contributions | |
greens | +0.024 |
pads | +0.023 |
Ol | +0.023 |
sauce | +0.023 |
gloss | +0.023 |
Neg logit contributions | |
choice | -0.016 |
away | -0.015 |
decision | -0.014 |
exactly | -0.014 |
passport | -0.013 |
Token | today |
Token ID | 1909 |
Feature activation | -8.98e-03 |
Pos logit contributions | |
greens | +0.060 |
Ol | +0.058 |
subtract | +0.057 |
sauce | +0.057 |
gum | +0.057 |
Neg logit contributions | |
decision | -0.034 |
choice | -0.034 |
away | -0.031 |
forward | -0.025 |
taxi | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | +1.39e-02 |
Pos logit contributions | |
greens | +0.018 |
sauce | +0.018 |
gum | +0.018 |
Ol | +0.017 |
subtract | +0.017 |
Neg logit contributions | |
decision | -0.012 |
choice | -0.012 |
away | -0.010 |
taxi | -0.009 |
doorstep | -0.009 |
Token | She |
Token ID | 1375 |
Feature activation | +1.05e-02 |
Pos logit contributions | |
decision | +0.018 |
choice | +0.016 |
ook | +0.016 |
away | +0.016 |
checkpoint | +0.016 |
Neg logit contributions | |
gum | -0.025 |
sauce | -0.025 |
greens | -0.023 |
triangles | -0.023 |
glasses | -0.023 |
Token | has |
Token ID | 468 |
Feature activation | +3.83e-03 |
Pos logit contributions | |
decision | +0.016 |
choice | +0.016 |
away | +0.015 |
ook | +0.013 |
promise | +0.013 |
Neg logit contributions | |
sauce | -0.018 |
gum | -0.018 |
greens | -0.018 |
triangles | -0.017 |
subtract | -0.017 |
Token | a |
Token ID | 257 |
Feature activation | +1.59e-02 |
Pos logit contributions | |
choice | +0.005 |
decision | +0.004 |
home | +0.004 |
taxi | +0.004 |
away | +0.004 |
Neg logit contributions | |
greens | -0.008 |
subtract | -0.007 |
Carl | -0.007 |
triangles | -0.007 |
squares | -0.007 |
Token | fever |
Token ID | 17372 |
Feature activation | -3.55e-02 |
Pos logit contributions | |
decision | +0.027 |
choice | +0.026 |
away | +0.024 |
taxi | +0.022 |
promise | +0.021 |
Neg logit contributions | |
greens | -0.026 |
sauce | -0.026 |
subtract | -0.025 |
gum | -0.025 |
pads | -0.025 |
Token | and |
Token ID | 290 |
Feature activation | -1.10e-02 |
Pos logit contributions | |
greens | +0.084 |
Ol | +0.082 |
triangles | +0.082 |
subtract | +0.082 |
squares | +0.080 |
Neg logit contributions | |
choice | -0.033 |
decision | -0.033 |
away | -0.032 |
taxi | -0.026 |
forward | -0.025 |
Token | a |
Token ID | 257 |
Feature activation | +9.66e-03 |
Pos logit contributions | |
greens | +0.019 |
subtract | +0.018 |
triangles | +0.018 |
squares | +0.017 |
sauce | +0.016 |
Neg logit contributions | |
choice | -0.016 |
decision | -0.016 |
away | -0.015 |
taxi | -0.014 |
exactly | -0.013 |
Token | sore |
Token ID | 19597 |
Feature activation | -2.43e-02 |
Pos logit contributions | |
decision | +0.017 |
choice | +0.017 |
away | +0.015 |
taxi | +0.015 |
promise | +0.014 |
Neg logit contributions | |
greens | -0.015 |
subtract | -0.015 |
triangles | -0.015 |
pads | -0.015 |
squares | -0.015 |
Token | throat |
Token ID | 13589 |
Feature activation | -2.94e-02 |
Pos logit contributions | |
sauce | +0.035 |
greens | +0.034 |
gum | +0.033 |
triangles | +0.033 |
subtract | +0.033 |
Neg logit contributions | |
decision | -0.046 |
choice | -0.046 |
away | -0.043 |
taxi | -0.038 |
ook | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | -1.86e-03 |
Pos logit contributions | |
greens | +0.063 |
triangles | +0.062 |
subtract | +0.062 |
gum | +0.061 |
sauce | +0.061 |
Neg logit contributions | |
decision | -0.035 |
choice | -0.035 |
away | -0.033 |
taxi | -0.027 |
promise | -0.026 |
Token | . |
Token ID | 13 |
Feature activation | -1.08e-02 |
Pos logit contributions | |
decision | +0.008 |
choice | +0.008 |
away | +0.008 |
ook | +0.006 |
taxi | +0.006 |
Neg logit contributions | |
sauce | -0.011 |
gum | -0.011 |
greens | -0.011 |
glasses | -0.010 |
ink | -0.010 |
Token | |
Token ID | 220 |
Feature activation | +9.47e-03 |
Pos logit contributions | |
sauce | +0.017 |
gum | +0.016 |
glasses | +0.015 |
greens | +0.015 |
ink | +0.014 |
Neg logit contributions | |
decision | -0.019 |
choice | -0.018 |
away | -0.018 |
ook | -0.017 |
taxi | -0.016 |
Token |
|
Token ID | 198 |
Feature activation | -7.54e-03 |
Pos logit contributions | |
ook | +0.013 |
decision | +0.013 |
choice | +0.013 |
away | +0.012 |
ively | +0.012 |
Neg logit contributions | |
sauce | -0.017 |
gum | -0.016 |
greens | -0.016 |
glasses | -0.016 |
ink | -0.016 |
Token |
|
Token ID | 198 |
Feature activation | -4.52e-03 |
Pos logit contributions | |
sauce | +0.012 |
gum | +0.012 |
greens | +0.011 |
glasses | +0.011 |
ink | +0.011 |
Neg logit contributions | |
decision | -0.013 |
choice | -0.012 |
away | -0.012 |
ook | -0.011 |
taxi | -0.011 |
Token | The |
Token ID | 464 |
Feature activation | -1.17e-02 |
Pos logit contributions | |
sauce | +0.007 |
gum | +0.007 |
greens | +0.007 |
glasses | +0.007 |
ink | +0.006 |
Neg logit contributions | |
decision | -0.008 |
choice | -0.008 |
away | -0.007 |
ook | -0.007 |
taxi | -0.006 |
Token | end |
Token ID | 886 |
Feature activation | -3.92e-02 |
Pos logit contributions | |
sauce | +0.027 |
gum | +0.026 |
greens | +0.026 |
pads | +0.026 |
glasses | +0.025 |
Neg logit contributions | |
decision | -0.013 |
choice | -0.012 |
away | -0.011 |
taxi | -0.009 |
ook | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | -6.40e-03 |
Pos logit contributions | |
sauce | +0.079 |
gum | +0.078 |
greens | +0.077 |
subtract | +0.075 |
glasses | +0.074 |
Neg logit contributions | |
decision | -0.052 |
choice | -0.050 |
away | -0.048 |
ook | -0.040 |
taxi | -0.039 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.57e-02 |
Pos logit contributions | |
sauce | +0.008 |
gum | +0.008 |
greens | +0.007 |
glasses | +0.007 |
ink | +0.007 |
Neg logit contributions | |
decision | -0.013 |
choice | -0.013 |
away | -0.013 |
ook | -0.012 |
taxi | -0.011 |
Token | Once |
Token ID | 7454 |
Feature activation | +2.52e-02 |
Pos logit contributions | |
fiber | +0.048 |
subtract | +0.048 |
sauce | +0.048 |
greens | +0.047 |
gum | +0.046 |
Neg logit contributions | |
away | -0.032 |
decision | -0.030 |
choice | -0.030 |
taxi | -0.028 |
ively | -0.028 |
Token | upon |
Token ID | 2402 |
Feature activation | -1.75e-02 |
Pos logit contributions | |
decision | +0.046 |
away | +0.045 |
choice | +0.044 |
doorstep | +0.042 |
checkpoint | +0.041 |
Neg logit contributions | |
gum | -0.031 |
sauce | -0.031 |
triangles | -0.030 |
ink | -0.029 |
Ben | -0.029 |
Token | a |
Token ID | 257 |
Feature activation | -1.45e-02 |
Pos logit contributions | |
sauce | +0.021 |
gum | +0.020 |
greens | +0.019 |
glasses | +0.018 |
ink | +0.018 |
Neg logit contributions | |
decision | -0.038 |
choice | -0.037 |
away | -0.036 |
ook | -0.033 |
taxi | -0.032 |
Token | around |
Token ID | 1088 |
Feature activation | +2.67e-02 |
Pos logit contributions | |
decision | +0.020 |
choice | +0.019 |
away | +0.019 |
ook | +0.016 |
taxi | +0.015 |
Neg logit contributions | |
sauce | -0.028 |
gum | -0.027 |
greens | -0.026 |
glasses | -0.026 |
ink | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
away | +0.034 |
decision | +0.033 |
choice | +0.032 |
forward | +0.028 |
taxi | +0.027 |
Neg logit contributions | |
sauce | -0.054 |
gum | -0.054 |
greens | -0.053 |
subtract | -0.052 |
gloss | -0.052 |
Token |
|
Token ID | 198 |
Feature activation | -4.17e-03 |
Pos logit contributions | |
sauce | +0.020 |
gum | +0.020 |
ink | +0.019 |
glasses | +0.019 |
gloss | +0.019 |
Neg logit contributions | |
decision | -0.017 |
choice | -0.016 |
away | -0.016 |
ook | -0.015 |
taxi | -0.015 |
Token |
|
Token ID | 198 |
Feature activation | -2.35e-03 |
Pos logit contributions | |
sauce | +0.005 |
gum | +0.005 |
greens | +0.005 |
glasses | +0.005 |
triangles | +0.005 |
Neg logit contributions | |
decision | -0.008 |
choice | -0.008 |
away | -0.008 |
tha | -0.007 |
checkpoint | -0.007 |
Token | The |
Token ID | 464 |
Feature activation | -7.20e-03 |
Pos logit contributions | |
sauce | +0.003 |
gum | +0.003 |
greens | +0.003 |
Ol | +0.002 |
triangles | +0.002 |
Neg logit contributions | |
decision | -0.005 |
choice | -0.005 |
away | -0.005 |
tha | -0.004 |
checkpoint | -0.004 |
Token | end |
Token ID | 886 |
Feature activation | -3.73e-02 |
Pos logit contributions | |
sauce | +0.015 |
greens | +0.015 |
gum | +0.015 |
pads | +0.014 |
subtract | +0.014 |
Neg logit contributions | |
decision | -0.009 |
choice | -0.009 |
away | -0.008 |
taxi | -0.007 |
hotel | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | -6.43e-03 |
Pos logit contributions | |
sauce | +0.070 |
gum | +0.070 |
greens | +0.068 |
glasses | +0.066 |
subtract | +0.065 |
Neg logit contributions | |
decision | -0.054 |
choice | -0.052 |
away | -0.051 |
ook | -0.044 |
taxi | -0.042 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.37e-02 |
Pos logit contributions | |
sauce | +0.008 |
gum | +0.007 |
greens | +0.007 |
glasses | +0.007 |
ink | +0.006 |
Neg logit contributions | |
decision | -0.014 |
choice | -0.013 |
away | -0.013 |
ook | -0.012 |
taxi | -0.012 |
Token | Today |
Token ID | 8888 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
sauce | +0.042 |
subtract | +0.041 |
greens | +0.041 |
gum | +0.040 |
fiber | +0.040 |
Neg logit contributions | |
away | -0.033 |
choice | -0.032 |
decision | -0.032 |
ook | -0.029 |
taxi | -0.029 |
Token | , |
Token ID | 11 |
Feature activation | -6.80e-03 |
Pos logit contributions | |
sauce | +0.023 |
gum | +0.022 |
ink | +0.020 |
greens | +0.020 |
glasses | +0.020 |
Neg logit contributions | |
decision | -0.020 |
away | -0.019 |
choice | -0.019 |
ook | -0.017 |
checkpoint | -0.016 |
Token | Emma |
Token ID | 18966 |
Feature activation | +2.13e-02 |
Pos logit contributions | |
sauce | +0.011 |
greens | +0.011 |
subtract | +0.011 |
gloss | +0.011 |
gum | +0.011 |
Neg logit contributions | |
decision | -0.010 |
choice | -0.010 |
away | -0.010 |
taxi | -0.009 |
ook | -0.009 |