INDEX
Explanations
descriptions of physical appearance, particularly related to faces
New Auto-Interp
Negative Logits
cedure
-0.18
åĵģ
-0.16
ediator
-0.16
readcr
-0.15
èħ
-0.15
edor
-0.15
esus
-0.15
овÑĸд
-0.15
iferay
-0.15
ittel
-0.15
POSITIVE LOGITS
(face
0.20
-face
0.20
/head
0.19
facial
0.18
/body
0.17
guard
0.17
faces
0.17
less
0.16
Rooney
0.15
Facial
0.15
Activations Density 0.038%