INDEX
Explanations
This neuron activates on references to Amazon’s Kindle products.
New Auto-Interp
Negative Logits
.As
-0.08
_by
-0.07
اوية
-0.06
arrest
-0.06
Advertisement
-0.06
(z
-0.06
.prefix
-0.06
fighter
-0.06
з
-0.06
.\"
-0.06
POSITIVE LOGITS
Kindle
0.13
Harley
0.07
Audi
0.07
ucchini
0.07
Carnival
0.07
Sony
0.06
Suite
0.06
LEGO
0.06
iking
0.06
ample
0.06
Activations Density 0.001%