INDEX
Explanations
proper nouns related to news agencies or media outlets
names of news agencies and publications
New Auto-Interp
Negative Logits
gone
-0.71
rous
-0.67
turn
-0.67
stood
-0.66
proof
-0.64
hall
-0.64
rams
-0.63
yz
-0.63
href
-0.62
gm
-0.62
POSITIVE LOGITS
IMAGES
0.75
Photo
0.74
Miko
0.73
PLIED
0.73
FILE
0.71
Images
0.70
Photographer
0.70
toggle
0.68
ENN
0.64
verett
0.63
Activations Density 0.070%