INDEX
Explanations
Articles and publications
The neuron fires strongly on the first meaningful content word in each new text segment, effectively marking the beginning of an article or section.
New Auto-Interp
Negative Logits
(command
-0.06
[ii
-0.06
ращи
-0.06
depending
-0.06
/end
-0.06
interrupt
-0.06
بيانات
-0.06
.cpp
-0.06
depressive
-0.06
Meals
-0.06
POSITIVE LOGITS
yll
0.06
dig
0.06
lyn
0.06
Maher
0.06
aying
0.06
UTH
0.06
liberty
0.06
Climate
0.06
conceal
0.06
質
0.06
Activations Density 0.032%