INDEX
Explanations
phrases ending with punctuation, such as periods and quotation marks
sentences that end in a period
New Auto-Interp
Negative Logits
pit
-0.60
purse
-0.59
uly
-0.56
iber
-0.55
emaker
-0.54
mur
-0.54
tack
-0.53
savage
-0.53
ascus
-0.52
quir
-0.52
POSITIVE LOGITS
However
1.00
Additionally
0.94
Nevertheless
0.91
Therefore
0.91
Whereas
0.91
Regardless
0.90
Luckily
0.90
Though
0.90
Nonetheless
0.89
Consequently
0.89
Activations Density 1.206%