INDEX
Explanations
hyphens, quotes, parenthesis and other punctuation when they are used around numbers
New Auto-Interp
Negative Logits
in
-1.31
IN
-0.88
يتيمه
-0.73
inl
-0.73
on
-0.71
ConstraintMaker
-0.69
en
-0.64
inb
-0.63
out
-0.59
inon
-0.59
POSITIVE LOGITS
faßt
0.60
this
0.56
acious
0.54
choly
0.51
asta
0.49
achen
0.48
verns
0.48
atech
0.47
daß
0.47
acy
0.46
Activations Density 14.481%