INDEX
Explanations
mentions of the word "ão"
structures related to Brazilian culture
New Auto-Interp
Negative Logits
itionally
-0.80
nut
-0.78
nuts
-0.77
arily
-0.74
lake
-0.72
crew
-0.72
PAR
-0.71
ritic
-0.68
guide
-0.68
terms
-0.68
POSITIVE LOGITS
Paulo
1.34
ão
1.25
vernment
1.02
Janeiro
0.93
Bras
0.86
ð
0.85
Jiu
0.83
Grande
0.82
Lumpur
0.81
afia
0.78
Activations Density 0.010%