INDEX

Explanations

design, conversion, targeted, need, required

modern, parameter, fast, key

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

0.47

ｐ

0.45

Researchers

0.44

are

0.43

指導

0.43

RI

0.42

wiring

0.42

is

0.42

erg

0.42

襄

0.41

POSITIVE LOGITS

 fêtes

0.53

 poveć

0.52

 classico

0.51

 pluie

0.51

 judul

0.51

 clásico

0.50

 negeri

0.49

 पॉप

0.48

 paga

0.48

入住

0.48

Activations Density 0.003%