INDEX

Explanations

instructions received

describing evolution and change

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

تى

0.49

 गलत

0.49

 طے

0.45

 او

0.45

0.44

itten

0.44

0.43

Nha

0.43

Nous

0.43

POSITIVE LOGITS

 pró

0.57

 breeder

0.54

 consenting

0.54

 personas

0.53

 refug

0.52

 mand

0.52

 debería

0.51

 refueling

0.51

 bookshelf

0.51

 recreating

0.51

Activations Density 0.001%