INDEX

Explanations

willingness and purpose

giving insight or to intercept

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 делу

0.62

 иссле

0.55

 новости

0.55

 offre

0.52

rico

0.52

 серь

0.51

 consente

0.51

viene

0.50

 трёх

0.50

POSITIVE LOGITS

ia

0.49

 Tavern

0.42

শিয়া

0.41

 Overland

0.40

ersen

0.39

تمام

0.39

াবেন

0.39

 fasten

0.38

 McKin

0.38

Ive

0.38

Activations Density 0.000%