INDEX

Explanations

phrase or clause completions

historical context and trends

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

coral

0.51

of

0.47

two

0.45

 hati

0.45

 مدت

0.45

логии

0.45

өлү

0.44

比特

0.43

 หน่อย

0.43

鵃

0.43

POSITIVE LOGITS

 Chowdh

0.50

KBS

0.45

CDS

0.45

 genere

0.43

 feast

0.42

 fallback

0.42

rán

0.41

Rg

0.41

 watchlist

0.41

 resourceful

0.41

Activations Density 0.000%