INDEX
Explanations
instances of mathematical notation with superscripts and negative exponents
New Auto-Interp
Negative Logits
-0.67
A
-0.63
-0.62
from
-0.60
C
-0.60
D
-0.58
R
-0.57
from
-0.56
G
-0.55
d
-0.53
POSITIVE LOGITS
^(@)
1.05
iſt
0.98
}}"></
0.96
homonymie
0.93
Beſ
0.93
་་
0.91
ſelf
0.90
―――――
0.88
raiſ
0.87
$_"
0.87
Activations Density 0.012%