INDEX
Explanations
frequent conjunctions and punctuation that emphasize reasoning and contrasts in statements
New Auto-Interp
Negative Logits
illac
-0.15
oss
-0.15
toMatch
-0.15
anax
-0.15
nty
-0.14
asonic
-0.14
argas
-0.14
.onStart
-0.14
Zar
-0.14
wc
-0.14
POSITIVE LOGITS
plus
0.18
284
0.17
Plus
0.16
дан
0.16
onom
0.16
Plus
0.15
PLUS
0.15
cul
0.15
hence
0.15
äch
0.14
Activations Density 0.282%