INDEX

Explanations

specific times or specific names

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 hassle

0.46

賑

0.43

 milling

0.43

 склон

0.43

 состоит

0.42

 botany

0.41

cens

0.41

молча

0.41

時に

0.41

ਣ

0.41

POSITIVE LOGITS

ט

0.55

ត្រូវបាន

0.54

klady

0.51

诏

0.50

Whitelist

0.50

Bie

0.48

 Nieder

0.47

贝尔

0.47

 Brett

0.47

يت

0.46

Activations Density 0.001%

No Known Activations