INDEX

Explanations

foreign languages

deployment strategy, tests, surface material

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

لاست

0.41

 extingu

0.41

Drv

0.41

 denominated

0.40

SError

0.39

 extinguish

0.39

 naudoj

0.39

 الكهربائيه

0.38

FN

0.38

 hydroelectric

0.38

POSITIVE LOGITS

 তরুণ

0.52

 талант

0.52

 уні

0.48

年轻

0.47

 उभर

0.47

 बॉबी

0.46

 कार्यकारिणी

0.46

龅

0.46

 lefty

0.46

 आध्यात्म

0.45

Activations Density 0.001%