INDEX
Explanations
The neuron activates on terms indicating defeat or losing outcomes (e.g. “lost,” “loss,” “losing”) in match and fight reports.
New Auto-Interp
Negative Logits
preempt
-0.08
↵ ↵
-0.06
Nin
-0.06
проведения
-0.06
продовж
-0.06
Effects
-0.06
-0.06
_pair
-0.06
Elaine
-0.06
/************************************************************************************************
-0.06
POSITIVE LOGITS
undra
0.07
.ident
0.06
формы
0.06
funeral
0.06
adio
0.06
мови
0.06
��
0.06
*(
0.06
etal
0.06
stro
0.06
Activations Density 0.013%