INDEX
Negative Logits
,
1.26
2
1.12
3
1.11
،
1.09
leri
1.05
dir
1.04
die
1.00
I
1.00
daniel
0.99
dır
0.95
POSITIVE LOGITS
ak
1.36
ad
1.33
as
1.28
ap
1.20
ing
1.17
ら
1.17
at
1.15
ia
1.15
ня
1.15
ag
1.10
Activations Density 0.000%
,
2
3
،
leri
dir
die
I
daniel
dır
ak
ad
as
ap
ing
ら
at
ia
ня
ag