INDEX
Explanations
the word "actually" used in various contexts to emphasize a statement
New Auto-Interp
Negative Logits
fós
-0.65
kana
-0.64
canaux
-0.64
normaux
-0.64
stessi
-0.63
fermés
-0.63
comigo
-0.62
medes
-0.59
culturelles
-0.59
TestBed
-0.59
POSITIVE LOGITS
actually
1.52
actually
1.35
Actually
1.18
ACTUALLY
1.16
Actually
1.11
literally
0.97
literally
0.94
really
0.92
basically
0.87
egentligen
0.85
Activations Density 0.072%