INDEX
Explanations
copyright-related information
references to copyright and media content
New Auto-Interp
Negative Logits
poon
-0.77
bite
-0.77
¬
-0.67
yt
-0.67
rational
-0.66
opian
-0.65
pill
-0.64
mind
-0.64
©¶æ¥µ
-0.64
sanity
-0.64
POSITIVE LOGITS
window
1.20
FILE
0.97
Protesters
0.89
FILE
0.88
PHOTO
0.86
IMAGES
0.82
Melania
0.75
photo
0.72
depiction
0.71
Ivanka
0.71
Activations Density 0.101%