INDEX
Explanations
parts of things
The neuron fires on words that label or introduce subdivisions of a document—terms like “portion,” “component,” “section,” or “part” that mark outline segments.
New Auto-Interp
Negative Logits
قوان
-0.06
referer
-0.06
مراجع
-0.06
successors
-0.06
$t
-0.05
ripple
-0.05
�
-0.05
apg
-0.05
Cbd
-0.05
embali
-0.05
POSITIVE LOGITS
component
0.09
.Color
0.08
part
0.08
components
0.08
région
0.07
portion
0.07
portions
0.07
Parts
0.07
Shoot
0.07
feature
0.07
Activations Density 0.136%