INDEX

Explanations

numbers and non-english words

key improvements and explanations

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

𝓰

0.48

𝚊

0.46

Gł

0.42

NavLink

0.42

putBoolean

0.41

 forgo

0.41

gister

0.41

 grotesque

0.39

 Nguyên

0.39

nasium

0.38

POSITIVE LOGITS

 dieron

0.41

 প্রাণ

0.41

 หรือ

0.40

 suelen

0.39

 tương

0.39

 atque

0.38

 famed

0.38

or

0.38

 โปร

0.38

XII

0.38

Activations Density 0.103%