INDEX
Explanations
Programming objects
This neuron detects type‐parameter or template syntax—that is, the angle-bracketed constructs (“<…>”) used for generics or templates.
New Auto-Interp
Negative Logits
isol
-0.07
linkage
-0.07
िश
-0.07
pose
-0.06
_One
-0.06
alias
-0.06
Rahman
-0.06
mente
-0.06
ging
-0.06
shin
-0.06
POSITIVE LOGITS
活
0.07
adır
0.07
ゝ
0.07
●●●●●●●●
0.06
어떤
0.06
علم
0.06
manuscripts
0.06
англ
0.06
문
0.06
div
0.06
Activations Density 0.024%