INDEX
Explanations
proper nouns related to individuals
mentions of key individuals, particularly those with the surname "Kahn" or "Og"
New Auto-Interp
Negative Logits
ylum
-0.78
ory
-0.74
ity
-0.73
arity
-0.71
ophobia
-0.68
terday
-0.67
Mahjong
-0.64
fully
-0.62
ophy
-0.62
ophobic
-0.62
POSITIVE LOGITS
rils
0.81
onga
0.80
thin
0.78
selage
0.75
leneck
0.74
Machine
0.72
elig
0.70
Kahn
0.70
gered
0.69
aez
0.66
Activations Density 0.018%