INDEX
Explanations
references to political conflicts and interests
New Auto-Interp
Negative Logits
201
-0.22
umblr
-0.16
ec
-0.16
usterity
-0.15
tweeted
-0.15
tweet
-0.14
Fukushima
-0.14
kir
-0.14
ultipart
-0.14
Breitbart
-0.14
POSITIVE LOGITS
Iraq
0.35
Bush
0.34
Saddam
0.34
Bush
0.31
Iraqi
0.31
Iraq
0.29
اÙĦعراÙĤ
0.27
عراÙĤ
0.25
Baghdad
0.25
bush
0.23
Activations Density 0.052%