INDEX
Explanations
sports-related terms and actions, especially in soccer matches
New Auto-Interp
Negative Logits
netflix
-0.92
onym
-0.83
agate
-0.76
azeera
-0.73
DragonMagazine
-0.71
ersive
-0.70
glomer
-0.70
chwitz
-0.69
hovah
-0.68
enegger
-0.68
POSITIVE LOGITS
halftime
0.96
timeout
0.90
stopp
0.79
scoreboard
0.78
midway
0.75
seconds
0.74
minute
0.74
ensuing
0.72
minute
0.72
margin
0.71
Activations Density 0.203%