INDEX
Explanations
mentions of candidates in a political context
New Auto-Interp
Negative Logits
hir
-0.15
Backend
-0.15
yper
-0.14
ighet
-0.14
.newInstance
-0.14
extras
-0.14
оба
-0.14
обÑĭ
-0.14
خب
-0.13
оÑĤв
-0.13
POSITIVE LOGITS
lear
0.15
ç¹Ķ
0.15
efore
0.14
_approval
0.14
nah
0.14
ennon
0.14
duk
0.14
Cle
0.14
anden
0.14
461
0.14
Activations Density 0.048%