INDEX
Explanations
specific project names and acronyms
New Auto-Interp
Negative Logits
atorial
-0.82
okia
-0.76
Drei
-0.75
ṭ
-0.74
yong
-0.73
toires
-0.73
Ẹ
-0.73
ています
-0.72
непо
-0.72
clad
-0.72
POSITIVE LOGITS
Faster
1.16
Faster
1.13
faster
0.94
SSD
0.91
faster
0.91
hourglass
0.88
分割
0.87
FAN
0.81
Jesuits
0.80
niająca
0.80
Activations Density 0.045%