INDEX
Explanations
stop words
The neuron strongly activates on common English function words—especially short articles and prepositions like “the,” “for,” and “of.”
New Auto-Interp
Negative Logits
کل
-0.06
Iran
-0.06
Nobody
-0.06
clearfix
-0.06
Kurds
-0.06
navegador
-0.06
_detail
-0.06
Defence
-0.06
Creatures
-0.06
ие
-0.06
POSITIVE LOGITS
OMP
0.06
PERF
0.06
PMC
0.06
(DBG
0.06
imu
0.06
_tgt
0.06
_HTTP
0.06
_cor
0.06
�
0.06
Μ
0.06
Activations Density 0.102%