INDEX
Explanations
references to clothing and accessories
New Auto-Interp
Negative Logits
shirts
-0.16
dge
-0.15
bufsize
-0.15
shirt
-0.15
dresses
-0.15
yte
-0.14
Dresses
-0.14
arda
-0.14
762
-0.14
jeans
-0.14
POSITIVE LOGITS
hat
0.56
hat
0.47
hats
0.47
Hat
0.47
Hat
0.43
帽
0.42
Hats
0.38
cap
0.36
_hat
0.36
caps
0.35
Activations Density 0.075%