INDEX
Explanations
proper nouns, specifically names of news organizations and journalists
references to news sources and reporting
New Auto-Interp
Negative Logits
bragging
-0.80
naires
-0.67
tons
-0.67
Github
-0.66
appa
-0.64
istry
-0.62
Registry
-0.62
naire
-0.60
FIG
-0.60
ware
-0.60
POSITIVE LOGITS
correspondent
0.83
inion
0.76
BBC
0.74
coverage
0.72
Jazeera
0.71
Reporter
0.71
Jonah
0.70
uart
0.70
GOODMAN
0.69
radio
0.68
Activations Density 0.197%