INDEX
Explanations
The neuron fires on HTML page navigation elements, especially tokens tied to “navbar” and skip‐navigation links.
New Auto-Interp
Negative Logits
inned
-0.06
Sınıf
-0.06
QP
-0.06
Müz
-0.06
گران
-0.06
Estonia
-0.06
Sheikh
-0.06
физ
-0.06
Webseite
-0.06
Buccaneers
-0.06
POSITIVE LOGITS
_IRQ
0.07
�
0.06
ight
0.06
(Constants
0.06
.assertIsNot
0.06
clk
0.06
bachelor
0.06
-component
0.06
=<
0.06
.getSimpleName
0.06
Activations Density 0.000%