INDEX
Explanations
the use of the preposition "at"
New Auto-Interp
Negative Logits
gether
-0.21
eldon
-0.16
head
-0.16
ruptcy
-0.16
iance
-0.16
aney
-0.15
enberg
-0.15
ercise
-0.14
oved
-0.14
lessly
-0.14
POSITIVE LOGITS
andard
0.21
onic
0.18
utor
0.17
roph
0.17
umn
0.17
oxic
0.17
mega
0.17
avic
0.17
omy
0.17
ention
0.16
Activations Density 0.054%