INDEX
Explanations
food descriptions and evaluations
New Auto-Interp
Negative Logits
боÑĤ
-0.15
ysz
-0.15
Pie
-0.14
Bent
-0.13
indir
-0.13
Pie
-0.13
_superuser
-0.13
pent
-0.13
606
-0.13
spinner
-0.13
POSITIVE LOGITS
bites
0.20
presentation
0.19
TEXT
0.18
bite
0.18
texture
0.17
ratio
0.16
text
0.16
resentation
0.16
textures
0.16
presentation
0.16
Activations Density 0.052%