INDEX
Explanations
comparative and contrastive phrases or structures between two or more subjects
New Auto-Interp
Negative Logits
736
-0.16
ersen
-0.15
ednou
-0.15
pter
-0.14
atern
-0.14
ober
-0.14
053
-0.13
iae
-0.13
ectar
-0.13
NEO
-0.13
POSITIVE LOGITS
itch
0.18
osp
0.16
kers
0.16
uraa
0.16
Banc
0.15
unga
0.15
precation
0.15
lij
0.15
ongan
0.14
Matters
0.14
Activations Density 0.238%