INDEX
Explanations
conversational fragments from an online exchange (possibly IRC), including usernames, colons, and common abbreviations and emoticons
New Auto-Interp
Negative Logits
myſelf
-1.06
Monfieur
-1.05
themſelves
-1.05
étoit
-1.00
ainfi
-0.97
Efq
-0.97
ſmall
-0.95
ſeveral
-0.95
leſs
-0.95
avoient
-0.94
POSITIVE LOGITS
his
0.63
In
0.61
It
0.59
But
0.59
0.59
of
0.56
Is
0.54
↵
0.52
0.51
허
0.51
Activations Density 0.307%