INDEX

Explanations

offensive unit

relative, higher, oriented, why, darker, urban

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 Köln

0.38

 üç

0.38

 drei

0.36

 Patil

0.35

 THREE

0.35

五个

0.35

Ｒ

0.35

 Several

0.35

POSITIVE LOGITS

 ones

0.82

のもの

0.53

ones

0.50

 counterpart

0.49

 counterparts

0.46

 theirs

0.42

 тоже

0.40

ONES

0.39

wanath

0.39

 Ones

0.38

Activations Density 0.161%