INDEX
Explanations
observation, inspection
The neuron detects numbered list markers (e.g. “1.”, “2.”, etc.) and other structured step or bullet formatting in procedural text.
New Auto-Interp
Negative Logits
_frames
-0.06
.Substring
-0.06
ala
-0.06
�
-0.06
polynomial
-0.06
blows
-0.06
Barton
-0.06
ろ
-0.06
/gpio
-0.06
apore
-0.06
POSITIVE LOGITS
MUSIC
0.07
supern
0.06
ras
0.06
Commentary
0.06
（
0.06
souha
0.06
ingroup
0.06
shots
0.06
plein
0.06
Swagger
0.06
Activations Density 0.043%