INDEX
Explanations
instances of color descriptors and their combinations
New Auto-Interp
Negative Logits
auen
-0.21
amil
-0.17
tas
-0.15
stown
-0.15
polym
-0.15
ibil
-0.15
lessly
-0.14
.poly
-0.14
qi
-0.14
DOI
-0.13
POSITIVE LOGITS
ì°
0.16
agal
0.15
باب
0.15
åª
0.14
strom
0.14
.componentInstance
0.14
pNet
0.14
endar
0.14
orial
0.14
äft
0.14
Activations Density 0.038%