INDEX
Explanations
articles
This neuron identifies magazine-style section headings or content labels (e.g. “Article,” “Feature,” “Reviews,” “Interview,” “Cover Story,” table-of-contents entries).
New Auto-Interp
Negative Logits
_alias
-0.06
_registro
-0.06
iki
-0.06
Πρω
-0.06
roll
-0.06
------------------------------------------------
-0.06
.xlsx
-0.06
าณ
-0.06
.diag
-0.06
_tab
-0.06
POSITIVE LOGITS
lobe
0.07
리
0.06
","\
0.06
몽
0.06
_inds
0.06
Barcl
0.06
بج
0.06
ær
0.06
safeg
0.06
oček
0.06
Activations Density 0.016%