INDEX
Explanations
image captions
references to images or visual content
New Auto-Interp
Negative Logits
dies
-0.72
itiz
-0.66
uits
-0.62
rapp
-0.61
onite
-0.61
pow
-0.60
itent
-0.60
lease
-0.59
merce
-0.58
die
-0.58
POSITIVE LOGITS
caption
0.92
©
0.77
Image
0.76
çͰ
0.75
Gallery
0.72
copyright
0.71
Photograph
0.70
Fran
0.70
Images
0.68
Pic
0.68
Activations Density 0.021%