INDEX
Negative Logits
as
1.46
it
1.30
i
1.30
p
1.26
:
1.24
on
1.11
ž
1.11
he
1.10
os
1.10
oh
1.10
POSITIVE LOGITS
ла
1.10
یط
0.98
イベント
0.97
allegations
0.96
5
0.96
эг
0.95
oaths
0.93
<unused2172>
0.93
}',
0.93
ক
0.93
Activations Density 0.004%
as
it
i
p
:
on
ž
he
os
oh
ла
یط
イベント
allegations
5
эг
oaths
<unused2172>
}',
ক