INDEX
Explanations
descriptions of something being visually appealing
references to the concept of attractiveness
New Auto-Interp
Negative Logits
sic
-0.78
aer
-0.76
bel
-0.72
roman
-0.71
lang
-0.71
iche
-0.70
cedented
-0.70
FIN
-0.69
orig
-0.69
elf
-0.69
POSITIVE LOGITS
attractive
1.04
proposition
0.92
propositions
0.86
attractiveness
0.84
lure
0.84
enticing
0.83
irresistible
0.77
enough
0.75
attraction
0.73
undermin
0.73
Activations Density 0.014%