INDEX

Explanations

provided/given

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 practicing

-0.54

?}

-0.51

kaŭ

-0.50

kész

-0.50

-0.49

 practising

-0.48

nologies

-0.48

hyrchwyd

-0.47

roppo

-0.47

 Runde

-0.46

POSITIVE LOGITS

you

0.87

 EconPapers

0.60

󠁣

0.59

 everyone

0.59

us

0.58

//

0.57

GHIJKLM

0.57

annica

0.56

elemField

0.56

msgTypes

0.55

Activations Density 0.166%