INDEX
Explanations
online conversations
The main thing this neuron does is pick up on vague, non-specific qualifiers and approximations (e.g. “some,” “kind of,” “something,” “rough estimate,” “timeline”).
New Auto-Interp
Negative Logits
IAN
-0.07
OG
-0.06
rated
-0.06
(serializer
-0.06
felt
-0.06
SEL
-0.06
ادگی
-0.06
ชาย
-0.06
pack
-0.06
放送
-0.06
POSITIVE LOGITS
Πλη
0.07
.decor
0.07
spd
0.07
.FindElement
0.07
'<
0.06
िए
0.06
_Db
0.06
сервер
0.06
.Alignment
0.06
Aspect
0.06
Activations Density 0.082%