INDEX
Explanations
occurrences of the name "Mary"
New Auto-Interp
Negative Logits
')")
-0.61
omnia
-0.59
"),
-0.59
ícil
-0.58
Andorra
-0.57
Figaro
-0.55
आव
-0.55
rouges
-0.55
tencent
-0.55
']").
-0.55
POSITIVE LOGITS
shift
0.95
shift
0.89
shirt
0.82
Mary
0.79
shirt
0.78
Shift
0.78
Shirt
0.75
Shift
0.74
hift
0.71
shifts
0.71
Activations Density 0.055%