INDEX
Explanations
producing or collecting information
New Auto-Interp
Negative Logits
<unused450>
0.88
поскольку
0.84
porque
0.80
savvy
0.80
<unused970>
0.80
<unused108>
0.80
фильма
0.79
tragically
0.78
<unused2125>
0.78
zón
0.78
POSITIVE LOGITS
some
1.49
the
1.47
一些
1.22
any
1.21
bazı
1.17
various
1.15
các
1.15
了一些
1.14
بعض
1.14
all
1.09
Activations Density 0.268%