INDEX
Explanations
references to file types or image formats
New Auto-Interp
Negative Logits
**
-0.23
·
-0.21
..
-0.21
*
-0.20
[
-0.20
...
-0.20
~
-0.19
=
-0.18
<
-0.17
####
-0.17
POSITIVE LOGITS
andi
0.15
анÑĤаж
0.15
_IV
0.15
رز
0.15
"";č↵
0.15
abbo
0.14
axon
0.14
...↵↵↵↵
0.14
ilan
0.14
ideographic
0.14
Activations Density 0.484%