INDEX
Explanations
initial introspection or context setting
New Auto-Interp
Negative Logits
Consequently
0.43
অতএব
0.42
entemente
0.42
=?";
0.41
果た
0.41
opos
0.41
مسلح
0.41
的重要
0.39
opo
0.38
استشهاد
0.38
POSITIVE LOGITS
being
0.64
seeing
0.63
when
0.59
definitely
0.59
honestly
0.57
initially
0.55
när
0.54
quando
0.54
khi
0.52
最初
0.52
Activations Density 0.001%