INDEX
Explanations
descriptions or mentions of paintings or portraits
references to portraits and their variations
New Auto-Interp
Negative Logits
duct
-0.75
onge
-0.72
inki
-0.71
uge
-0.70
anwhile
-0.68
ibel
-0.68
urious
-0.67
hov
-0.67
jab
-0.66
lower
-0.65
POSITIVE LOGITS
portrait
1.18
portraits
1.18
raits
1.08
painter
0.96
rait
0.96
photographer
0.89
Painting
0.86
paintings
0.86
photographers
0.85
Photographer
0.82
Activations Density 0.018%