INDEX
Explanations
mentions of the name "Erlich" with high activation values
names of individuals or entities associated with a context
New Auto-Interp
Negative Logits
Independence
-0.78
bard
-0.70
sneak
-0.70
extract
-0.68
Sneak
-0.67
tang
-0.63
Alleg
-0.62
abs
-0.62
GH
-0.61
extracted
-0.61
POSITIVE LOGITS
lich
2.64
passer
2.13
gard
1.18
pest
1.12
sted
1.01
lear
0.97
ndra
0.93
rek
0.91
llor
0.87
rish
0.87
Activations Density 0.037%