INDEX
Explanations
references to civilian casualties and the impact of violence on families
New Auto-Interp
Negative Logits
Fav
-0.17
ovna
-0.16
urette
-0.15
shalt
-0.15
_vlog
-0.15
adiens
-0.14
Äħż
-0.14
inux
-0.14
leston
-0.14
ľ´
-0.14
POSITIVE LOGITS
account
0.16
country
0.16
j
0.15
collateral
0.14
Dickens
0.14
ãĥ³ãĥĨ
0.14
occo
0.14
j
0.14
bull
0.14
ood
0.14
Activations Density 0.339%