INDEX
Explanations
references to political candidates and election-related context
New Auto-Interp
Negative Logits
²
-0.15
flank
-0.14
feud
-0.14
arshal
-0.14
rough
-0.14
Fresh
-0.14
ringe
-0.13
Confeder
-0.13
Or
-0.13
_gs
-0.13
POSITIVE LOGITS
abler
0.23
ledo
0.19
عاÙĦ
0.19
edb
0.18
ingroup
0.18
reland
0.18
IIIK
0.18
atest
0.18
acomment
0.18
оÐ
0.17
Activations Density 0.609%