INDEX

Explanations

also contributed, made it, had dropped

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rill

0.53

rophic

0.52

hews

0.48

్యూ

0.46

ಾ

0.46

ುದ

0.45

otroph

0.45

ುದು

0.45

agogue

0.43

示范

0.43

POSITIVE LOGITS

 purg

0.57

 versión

0.54

 până

0.52

Toh

0.51

Cuv

0.50

 versione

0.49

ˬ

0.49

 pung

0.49

 Prodi

0.48

 encontró

0.47

Activations Density 0.001%