INDEX

Explanations

foreign language content

safety, estimation, code blocks

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

Beginning

0.52

將

0.47

Wed

0.46

itism

0.45

Mountains

0.44

ändig

0.43

ojan

0.42

ibrated

0.41

يد

0.41

itative

0.41

POSITIVE LOGITS

 रिश्तों

0.50

 बल

0.50

>?

0.49

 फसल

0.49

pel

0.48

 mezzanine

0.47

 dangereux

0.47

 receta

0.46

 sosten

0.46

 llevó

0.46

Activations Density 0.000%