INDEX
Explanations
phrases or names containing "ahan"
repeated references to specific names, particularly "Kahan."
New Auto-Interp
Negative Logits
raltar
-0.78
erence
-0.75
iculty
-0.75
icult
-0.74
eries
-0.73
erences
-0.72
enser
-0.71
sheet
-0.71
erential
-0.70
pse
-0.70
POSITIVE LOGITS
ufact
1.08
igans
0.85
lain
0.77
Kirin
0.73
igan
0.72
Marino
0.72
lich
0.71
Assass
0.71
wine
0.67
ONT
0.67
Activations Density 0.030%