INDEX
Explanations
references to portraits and imagery related to faces and representations in art
New Auto-Interp
Negative Logits
-quarters
-0.18
asing
-0.17
ew
-0.17
æ´²
-0.16
odia
-0.15
Ùij
-0.15
elda
-0.15
kr
-0.15
asin
-0.15
elder
-0.15
POSITIVE LOGITS
raits
0.25
folios
0.23
age
0.21
smouth
0.21
ance
0.18
ugal
0.18
ability
0.18
lier
0.17
ive
0.17
æ¹¾
0.17
Activations Density 0.050%