INDEX

Explanations

simple terms

can confidently provide

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

KO

0.46

Kho

0.44

yg

0.44

kitten

0.43

okhttp

0.43

essä

0.42

positions

0.42

eed

0.42

ramient

0.41

 MeOH

0.41

POSITIVE LOGITS

 eloqu

0.49

 supremac

0.48

ீஸ்

0.48

૦

0.47

νοντας

0.47

 eloquence

0.46

א

0.46

レンズ

0.45

 idolat

0.45

数は

0.45

Activations Density 0.000%