INDEX
Explanations
phrases enclosed in quotation marks
punctuated phrases, specifically those involving closing quotation marks
New Auto-Interp
Negative Logits
%.
-0.54
however
-0.54
-0.54
meanwhile
-0.53
-0.48
though
-0.47
.-
-0.45
garner
-0.44
ADVERTISEMENT
-0.43
↵↵
-0.43
POSITIVE LOGITS
")
3.56
").
3.29
"),
3.27
.")
3.13
");
3.07
"))
3.02
"]
2.84
"],
2.22
')
2.17
').
2.07
Activations Density 0.008%