INDEX
Explanations
mentions of the country Iraq or related entities
references to Iraq and its people
New Auto-Interp
Negative Logits
Wilde
-0.77
vo
-0.73
DK
-0.71
woo
-0.68
pi
-0.65
glove
-0.65
Cu
-0.63
Pier
-0.62
ynthesis
-0.62
clips
-0.61
POSITIVE LOGITS
Iraqi
3.67
Iraqis
3.10
Iraq
2.78
Iraq
2.35
Baghdad
2.31
Iranian
2.11
Afghan
2.10
Saddam
2.07
Kurdish
2.05
Yemeni
2.02
Activations Density 0.018%