INDEX
Explanations
references to the city of Arequipa in Peru
geographic locations or proper nouns
New Auto-Interp
Negative Logits
warm
-0.64
fart
-0.63
hook
-0.62
friendship
-0.61
football
-0.61
prank
-0.60
wanting
-0.60
clicks
-0.59
gift
-0.59
combined
-0.58
POSITIVE LOGITS
eda
4.84
eta
1.21
eto
1.21
ida
1.18
ede
1.16
eva
1.15
era
1.14
edo
1.13
edi
1.13
esi
1.10
Activations Density 0.007%