INDEX
Explanations
different formats and conditions of book covers
New Auto-Interp
Negative Logits
owers
-0.16
reib
-0.14
oice
-0.14
obody
-0.14
433
-0.14
ensive
-0.14
pek
-0.14
ho
-0.14
ension
-0.13
/themes
-0.13
POSITIVE LOGITS
oute
0.15
format
0.15
èģĺ
0.15
/by
0.15
-only
0.15
ized
0.14
ÐľÐµÑģÑĤо
0.14
ita
0.14
iola
0.14
illes
0.13
Activations Density 0.020%