INDEX
Explanations
This neuron fires on tokens related to numeric quantification of cost or value—prices, percentages, and other amount markers.
New Auto-Interp
Negative Logits
��
-0.07
cạnh
-0.06
match
-0.06
Зап
-0.06
cake
-0.06
cous
-0.06
tpl
-0.06
ोप
-0.06
Inserted
-0.06
thing
-0.06
POSITIVE LOGITS
.assertNull
0.06
Amnesty
0.06
>";↵
0.06
abs
0.06
primaryKey
0.06
Essays
0.06
сни
0.06
ایان
0.06
]-$
0.06
RECEIVE
0.06
Activations Density 0.346%