INDEX

Explanations

multi-word quantities or lengths

abuse, learn, established, why

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 correg

0.35

 dobre

0.34

에서

0.34

 компет

0.33

ક

0.32

 ERISA

0.32

優れた

0.32

 polytopes

0.31

 nogle

0.31

}$

0.30

POSITIVE LOGITS

每一次

0.36

 আমিও

0.34

 নিজেও

0.33

quantidade

0.33

 Length

0.32

Length

0.31

长度

0.31

 তাছাড়া

0.31

 डायरेक्टली

0.31

 AMOUNT

0.31

Activations Density 0.001%