INDEX
Explanations
references to religious texts and the phrase "word of the Lord."
New Auto-Interp
Negative Logits
chner
-0.16
ouns
-0.16
ibal
-0.15
Delta
-0.14
ewriter
-0.14
hair
-0.14
643
-0.14
ses
-0.14
hair
-0.14
sk
-0.14
POSITIVE LOGITS
ring
0.16
Walls
0.16
éĶĭ
0.15
ittel
0.15
[dir
0.15
ebra
0.14
esson
0.14
ajs
0.14
walls
0.14
arching
0.14
Activations Density 0.085%