INDEX
Explanations
references to a specific person or location: "ala"
mentions of the name "Ayala."
New Auto-Interp
Negative Logits
frames
-0.66
smoked
-0.65
downed
-0.63
batt
-0.63
tick
-0.62
smoke
-0.61
guiActive
-0.60
impe
-0.59
Vers
-0.59
peer
-0.58
POSITIVE LOGITS
ala
4.33
alle
1.44
alan
1.41
alia
1.40
alam
1.34
ali
1.30
alo
1.29
ela
1.25
alin
1.19
ila
1.17
Activations Density 0.017%