INDEX
Explanations
references to helicopters or related terminology
New Auto-Interp
Negative Logits
ess
-0.18
izzle
-0.16
emp
-0.15
veau
-0.15
èĥŀ
-0.14
-0.14
ój
-0.14
ners
-0.14
elerik
-0.14
naments
-0.14
POSITIVE LOGITS
dür
0.17
lob
0.16
icopter
0.16
.bc
0.16
ucid
0.15
insula
0.15
worker
0.15
Ñİ
0.14
Structured
0.14
ix
0.14
Activations Density 0.037%