INDEX
Explanations
file names and numerical data related to images and dates
New Auto-Interp
Negative Logits
ixel
-0.18
umlu
-0.17
inne
-0.15
igon
-0.14
å¹³
-0.14
estroy
-0.14
,tp
-0.14
ounge
-0.14
.persistent
-0.14
asley
-0.13
POSITIVE LOGITS
erman
0.15
/REC
0.15
avers
0.14
ÑĥлÑĭ
0.14
å¹´
0.14
/umd
0.14
bulk
0.14
áng
0.14
tÃŃ
0.14
-fashion
0.13
Activations Density 0.016%