INDEX

Explanations

code and abstracts

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

styleable

-0.55

Personensuche

-0.55

 وصلة

-0.54

SourceChecksum

-0.53

SuccessListener

-0.52

 typelib

-0.52

 Signalez

-0.51

Turning

-0.50

 communism

-0.49

 Leyden

-0.48

POSITIVE LOGITS

 fuite

0.59

嬸

0.53

isEnd

0.52

 sagesse

0.49

 Dumnezeu

0.49

dor

0.48

AndEndTag

0.48

__*/

0.47

={<

0.47

 Full

0.47

Activations Density 0.001%