INDEX
Explanations
references to figures and captions in a structured document
New Auto-Interp
Negative Logits
¾
-0.19
¿
-0.18
lero
-0.17
+
-0.16
Landing
-0.16
@
-0.15
([
-0.14
æħ§
-0.14
akit
-0.14
AIM
-0.14
POSITIVE LOGITS
setup
0.22
{0.17
*
0.17
-setup
0.16
®
0.16
_setup
0.15
ologie
0.14
idor
0.14
setup
0.14
{$0.14
Activations Density 0.010%