INDEX
Explanations
references to martial arts rankings and styles
New Auto-Interp
Negative Logits
icina
-0.17
Wrest
-0.16
iggers
-0.16
Agent
-0.14
ernet
-0.14
̧
-0.14
iesen
-0.14
gov
-0.14
ritch
-0.14
_agent
-0.13
POSITIVE LOGITS
belt
0.24
Sense
0.20
Belt
0.20
sense
0.20
Sense
0.20
belts
0.20
belt
0.18
sense
0.18
kicks
0.17
kicking
0.17
Activations Density 0.035%