INDEX
Explanations
The neuron responds to occurrences of the word “Discord” (in any casing) in the text.
New Auto-Interp
Negative Logits
corrupt
-0.07
años
-0.07
Smoke
-0.07
Beds
-0.07
YRO
-0.07
Animals
-0.06
GeneratedValue
-0.06
exposure
-0.06
descend
-0.06
interviews
-0.06
POSITIVE LOGITS
妙
0.06
NHL
0.06
msgid
0.06
τησε
0.06
güvenilir
0.06
нул
0.06
klad
0.06
Serious
0.06
_$
0.06
就会
0.06
Activations Density 0.003%