INDEX
Explanations
proper nouns, specifically names like "Winston" and "Prescott"
proper nouns, specifically names associated with prominent historical or public figures
New Auto-Interp
Negative Logits
*/(
-0.87
anguage
-0.82
people
-0.73
trl
-0.72
orpor
-0.72
onial
-0.72
cule
-0.72
venture
-0.70
edom
-0.69
vity
-0.69
POSITIVE LOGITS
endor
0.91
Recomm
0.75
Churchill
0.74
Ko
0.73
Ze
0.71
Fore
0.71
Tec
0.70
Bridgewater
0.70
Bunny
0.70
fur
0.69
Activations Density 0.104%