INDEX
Explanations
descriptors that evoke strong visual or emotional imagery
New Auto-Interp
Negative Logits
317
-0.16
iasi
-0.15
312
-0.15
athers
-0.15
uali
-0.14
IID
-0.14
812
-0.14
ILED
-0.14
redi
-0.14
owie
-0.13
POSITIVE LOGITS
ly
0.18
onest
0.16
Sir
0.15
اÙĨÙĩ
0.15
ogl
0.14
Sir
0.14
Bik
0.14
excuses
0.14
FUN
0.14
less
0.13
Activations Density 0.272%