INDEX

Explanations

TYPE =

amounts and locations

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

ိုလ်

0.49

BOARD

0.49

垲

0.49

MUSIC

0.45

вшего

0.45

頎

0.45

 administración

0.44

یل

0.44

𝚢

0.44

Movies

0.43

POSITIVE LOGITS

im

0.54

 firepower

0.53

ag

0.53

re

0.50

 molle

0.50

projectlombok

0.49

 leaf

0.49

arn

0.48

 upsetting

0.48

Activations Density 0.003%