INDEX
Explanations
references to the Boko Haram group and related violent events
New Auto-Interp
Negative Logits
oop
-0.17
liner
-0.15
iller
-0.14
umen
-0.14
urg
-0.14
ÏĥÏĦÏģο
-0.14
panion
-0.14
ech
-0.14
anner
-0.14
ele
-0.14
POSITIVE LOGITS
_HARD
0.16
foot
0.15
idis
0.14
irim
0.14
èµŀ
0.14
rider
0.13
cư
0.13
nid
0.13
/car
0.13
iton
0.13
Activations Density 0.002%