INDEX
Explanations
geographic names and locations, particularly in Africa
New Auto-Interp
Negative Logits
_mt
-0.17
-m
-0.17
bourg
-0.16
enen
-0.16
moid
-0.15
ABCDEFG
-0.15
adar
-0.14
èĴĻ
-0.14
oggle
-0.14
onto
-0.14
POSITIVE LOGITS
Gut
0.20
690
0.17
683
0.17
-g
0.17
932
0.16
-G
0.16
931
0.16
_g
0.16
ÙĬز
0.15
Gus
0.15
Activations Density 0.074%