INDEX

Explanations

just before a consequential word

new users and development

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

}=\

0.41

句

0.40

odef

0.38

骰

0.38

 culp

0.37

인은

0.37

 mago

0.37

手中的

0.37

0.36

arch

0.36

POSITIVE LOGITS

更好的

0.56

 bessere

0.55

 جدید

0.53

 bättre

0.52

ใหม่

0.52

 інших

0.51

 newUser

0.51

Shipping

0.50

 новые

0.50

 nových

0.50

Activations Density 0.009%