INDEX

Explanations

questions starting with how

questions asking why or how long

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

correct

0.43

 Confirm

0.40

correcto

0.40

 correct

0.40

 confirm

0.39

confirm

0.39

Также

0.39

沔

0.39

innaker

0.39

ex

0.38

POSITIVE LOGITS

很难

0.44

 geometri

0.43

ायक

0.41

 sonunda

0.40

 bleached

0.40

سرائيل

0.39

ខ្ល

0.39

 কোনটা

0.39

 deregulation

0.39

 いや

0.39

Activations Density 0.000%