INDEX
Explanations
words related to religious or revered concepts
references to sacredness and its variations
New Auto-Interp
Negative Logits
ILA
-0.84
OPA
-0.81
onne
-0.78
Agg
-0.78
EFF
-0.77
Streamer
-0.75
Newsletter
-0.74
POL
-0.74
OU
-0.73
UNCH
-0.73
POSITIVE LOGITS
sacred
1.12
maiden
0.97
rites
0.94
ceremonial
0.91
mant
0.87
rite
0.86
arte
0.86
relics
0.85
scripture
0.84
ificial
0.83
Activations Density 0.010%