INDEX

Explanations

years'

legal or premature actions

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 secteurs

0.47

 Nord

0.46

 danske

0.46

ru

0.45

 Länder

0.45

ിക

0.44

 Danmark

0.44

usters

0.43

 Scandinavia

0.43

 kampus

0.43

POSITIVE LOGITS

confirm

0.51

筈

0.48

 createImage

0.47

injured

0.47

主动

0.46

伤

0.46

ждены

0.46

 மாய

0.45

colon

0.44

cede

0.44

Activations Density 0.001%