INDEX
Negative Logits
in
1.58
↵
1.46
ل
1.18
ే
1.02
es
0.99
ан
0.95
ir
0.92
at
0.91
is
0.90
isches
0.90
POSITIVE LOGITS
1.06
ﻦ
0.79
^{-}0.71
人
0.70
ן
0.70
is
0.67
т
0.67
み
0.66
t
0.65
ﺪ
0.65
Activations Density 1.161%
in
↵
ل
ే
es
ан
ir
at
is
isches
ﻦ
^{-}人
ן
is
т
み
t
ﺪ