INDEX

Explanations

a standard additional whole list three

bits needed

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

in

0.54

 accuses

0.52

ere

0.50

or

0.47

so

0.45

 accuse

0.44

The

0.43

 categor

0.43

 discredit

0.43

on

0.42

POSITIVE LOGITS

𝘭

0.62

 필요한

0.54

<unused25>

0.52

میم

0.52

 prochaine

0.52

<unused77>

0.51

 nødv

0.51

হৃত

0.50

 необходимых

0.50

 உண்மையில்

0.50

Activations Density 0.000%