INDEX
Explanations
descriptive features related to clothing and fashion
New Auto-Interp
Negative Logits
XT
-0.15
tie
-0.14
lass
-0.14
Vict
-0.14
tent
-0.14
Pressure
-0.14
gram
-0.14
Nest
-0.14
deut
-0.14
Font
-0.13
POSITIVE LOGITS
Manson
0.15
.enumer
0.15
_STRUCTURE
0.15
Ref
0.15
Mul
0.14
lateral
0.14
STORE
0.13
engo
0.13
Hogan
0.13
ref
0.13
Activations Density 0.015%