INDEX
Explanations
sentence-ending punctuation, indicating the end of thoughts or statements
New Auto-Interp
Negative Logits
oro
-0.18
ir
-0.17
un
-0.17
id
-0.15
↵
-0.15
z
-0.15
[
-0.15
iv
-0.15
*
-0.15
"
-0.15
POSITIVE LOGITS
ascar
0.15
rient
0.15
®,
0.15
á»ĵi
0.15
alone
0.14
tog
0.14
â̲
0.14
ogle
0.14
,č↵
0.14
иÑĩа
0.13
Activations Density 0.103%