INDEX
Explanations
creation
This neuron fires on the initial, often uppercase, words or headings at the very start of a text block (i.e. paragraph or article openings).
New Auto-Interp
Negative Logits
илання
-0.07
_tE
-0.06
：
-0.06
destinationViewController
-0.06
memo
-0.06
ظ
-0.06
ему
-0.06
付け
-0.06
#ac
-0.06
Ａ
-0.06
POSITIVE LOGITS
دن
0.06
indice
0.06
ovice
0.06
ebra
0.06
िकल
0.06
aceut
0.06
haust
0.06
')==
0.06
handles
0.06
{
↵
↵0.06
Activations Density 0.007%