INDEX

Explanations

http://links, URLs, emails

lists, code, and instructions

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

 putative

0.27

বার্গ

0.27

 rejoined

0.26

 đị

0.26

Plc

0.26

 junt

0.25

 incurring

0.25

 aswell

0.25

 Dion

0.25

 inclu

0.24

POSITIVE LOGITS

М

0.29

Фе

0.29

싫

0.28

 குடும்ப

0.28

 нами

0.28

னவே

0.27

 фонда

0.27

MATERIALS

0.27

 माला

0.27

 പ്രവർത്തന

0.26

Activations Density 0.007%