INDEX
Explanations
news agency names
mentions of the news organization Reuters
New Auto-Interp
Negative Logits
sed
-0.68
icular
-0.67
naires
-0.66
haun
-0.66
sed
-0.66
adm
-0.62
tein
-0.61
comprom
-0.61
quer
-0.60
iencies
-0.60
POSITIVE LOGITS
Images
0.81
Reuters
0.76
Tonight
0.75
PLA
0.70
%)
0.70
)—
0.70
Pool
0.69
Coverage
0.69
Film
0.67
Wire
0.67
Activations Density 0.020%