INDEX
Explanations
wrestling matches
This neuron fires on tokens related to professional wrestling broadcast commentary—particularly wrestler and show names (e.g. SmackDown, Champion, Joe, Sheamus) and match/action vocabulary.
New Auto-Interp
Negative Logits
Index
-0.07
.ReadAllText
-0.07
PLEX
-0.06
orpion
-0.06
View
-0.06
Mahmoud
-0.06
مهندسی
-0.06
sql
-0.06
_mes
-0.06
cultures
-0.06
POSITIVE LOGITS
대구
0.07
ázi
0.07
_ACTIVITY
0.07
тра
0.06
(create
0.06
بو
0.06
غراف
0.06
CGPoint
0.06
clazz
0.06
يح
0.06
Activations Density 0.011%