INDEX

Explanations

by the

personal topics and contrasts

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

abit

0.50

uffs

0.48

するので

0.48

older

0.46

vět

0.46

:'],

0.45

areas

0.44

alit

0.43

Subsec

0.43

rings

0.43

POSITIVE LOGITS

0.46

 closely

0.46

 Armenia

0.44

 Valores

0.44

 outra

0.43

 Canaria

0.43

 KPMG

0.43

 Taiwan

0.43

 Pyrénées

0.43

 Dubai

0.43

Activations Density 0.002%