INDEX

Explanations

"should" and "must" followed by verbs

necessity and obligation

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

ர்

0.32

))

0.30

<0x0D>

0.27

were

0.27

অ

0.26

？

0.26

}$.

0.25

 robes

0.25

架

0.25

POSITIVE LOGITS

 avoid

0.30

 максимально

0.30

 помнить

0.30

 обязательно

0.30

 considerare

0.30

ников

0.29

 encourage

0.29

 beware

0.29

 unbedingt

0.29

 know

0.28

Activations Density 0.434%