INDEX
Explanations
The neuron activates specifically on occurrences of the surname “Impastato.”
New Auto-Interp
Negative Logits
matchups
-0.07
M
-0.06
matchup
-0.06
mingle
-0.06
вей
-0.06
获得
-0.06
B
-0.06
eighteen
-0.06
=?",
-0.06
.Future
-0.06
POSITIVE LOGITS
suffix
0.07
0.06
знач
0.06
doctr
0.06
secret
0.06
userName
0.06
lav
0.06
UST
0.06
(prod
0.06
Diagnostic
0.06
Activations Density 0.000%