INDEX

Explanations

race, unique, break, time

finding patterns in lists

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

 ENABLE

0.49

 знал

0.48

 непотпуним

0.48

 медве

0.48

應該

0.47

 authenticate

0.47

 PRODUCTS

0.47

 знали

0.47

 bear

0.46

ENO

0.46

POSITIVE LOGITS

0.46

ivät

0.46

ถนน

0.44

processes

0.43

type

0.43

كتوبر

0.42

 than

0.42

street

0.41

나

0.41

istema

0.41

Activations Density 0.002%