INDEX

Explanations

when, use, reducing, among

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Sext

0.45

 Penet

0.44

 plag

0.44

ሰ

0.44

 penetrate

0.43

 erect

0.41

 sust

0.41

 Arabs

0.41

 Robbie

0.41

 propagate

0.40

POSITIVE LOGITS

 acidosis

0.45

ます

0.45

दूसरी

0.44

 dieses

0.44

νό

0.44

ホテル

0.43

オーダー

0.43

 umumnya

0.43

льный

0.43

 vidhan

0.43

Activations Density 0.001%