INDEX
Explanations
references to furniture, specifically couch or sofa
New Auto-Interp
Negative Logits
EndProject
-0.58
мѣ
-0.55
concentration
-0.53
ruz
-0.52
scolaires
-0.52
PageIndex
-0.51
المناصب
-0.50
osoba
-0.49
sata
-0.49
CHL
-0.49
POSITIVE LOGITS
couch
3.88
Couch
2.84
couch
2.83
sofa
2.77
couches
2.55
Couch
2.52
sofas
2.19
Sofa
2.09
sofá
1.88
sofa
1.85
Activations Density 0.001%