INDEX
Explanations
URLs of specific news websites
URLs and web addresses, particularly those associated with news outlets
New Auto-Interp
Negative Logits
tein
-0.77
rhy
-0.72
sucker
-0.71
ansky
-0.67
abase
-0.67
weed
-0.67
predicate
-0.66
notebooks
-0.65
sample
-0.63
siph
-0.63
POSITIVE LOGITS
biz
0.87
levision
0.82
Vaugh
0.71
arthed
0.71
Thrones
0.71
rencies
0.70
News
0.70
politics
0.69
millenn
0.68
Politics
0.66
Activations Density 0.072%