INDEX
Explanations
dates, particularly in the format of months and years
dates and numerical sequences
New Auto-Interp
Negative Logits
enhagen
-0.79
ogle
-0.71
elsen
-0.70
psey
-0.66
scrolls
-0.63
anamo
-0.60
selfies
-0.57
lees
-0.56
microphones
-0.56
ãĤ½
-0.55
POSITIVE LOGITS
Ö¼
0.76
guiActiveUnfocused
0.65
Caucasus
0.61
ĨĴ
0.61
CEPT
0.59
Controller
0.59
elsius
0.57
irreversible
0.57
omaly
0.57
SPONSORED
0.57
Activations Density 1.056%