INDEX
Explanations
shall followed by verb or not
New Auto-Interp
Negative Logits
NA
0.37
نتی
0.35
naden
0.35
Shadows
0.33
watery
0.33
sembles
0.32
Lightning
0.32
Figures
0.32
books
0.32
োদ্ধ
0.31
POSITIVE LOGITS
<0xA2>
0.34
inoltre
0.34
এইরূপে
0.31
yrıca
0.30
எம்
0.30
eric
0.29
&=
0.29
шенный
0.29
تحتوي
0.28
officielle
0.28
Activations Density 0.001%