INDEX

Explanations

which helps

built from the ground up

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 Maximum

0.42

न्ग

0.42

 kinases

0.42

 Trusts

0.42

Aha

0.42

herr

0.41

 Pharmacology

0.41

DEN

0.41

ČR

0.41

aline

0.40

POSITIVE LOGITS

 breach

0.48

クト

0.48

reden

0.48

orthodox

0.46

 congr

0.45

膳

0.44

raphe

0.44

此同时

0.44

 spectacle

0.43

 pleine

0.43

Activations Density 0.000%