INDEX
Explanations
names of Korean individuals
references to Korean names or titles related to Korean culture and politics
New Auto-Interp
Negative Logits
conv
-0.72
FedEx
-0.70
barn
-0.69
butterflies
-0.66
ranc
-0.65
cort
-0.65
20439
-0.65
Ples
-0.64
Conv
-0.62
Frenzy
-0.61
POSITIVE LOGITS
jin
1.27
wei
1.16
jong
1.05
Jong
1.03
qi
1.02
sung
1.01
Sung
1.00
Suk
0.98
Kong
0.98
jri
0.95
Activations Density 0.122%