INDEX
Explanations
references to the Boston Marathon and related events
New Auto-Interp
Negative Logits
metics
-0.84
oshop
-0.74
obyl
-0.74
choes
-0.71
ARS
-0.71
Debor
-0.70
onom
-0.67
ORK
-0.66
ongyang
-0.65
orum
-0.64
POSITIVE LOGITS
runner
1.05
athon
0.95
runners
0.91
marathon
0.90
Runner
0.88
NING
0.74
consecut
0.72
Peaks
0.72
runners
0.71
bow
0.70
Activations Density 0.003%