INDEX
Explanations
proper nouns related to executives or personalities
occurrences of the name "Claire."
New Auto-Interp
Negative Logits
okin
-0.92
otos
-0.79
odd
-0.77
raphics
-0.75
oresc
-0.75
oster
-0.74
astered
-0.73
enty
-0.73
enton
-0.72
airo
-0.72
POSITIVE LOGITS
child
0.80
rics
0.77
Vu
0.76
ments
0.75
coat
0.75
mentation
0.74
woman
0.72
mens
0.71
rences
0.71
keeper
0.71
Activations Density 0.028%