INDEX

Explanations

See

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

See

-1.41

 percent

-0.93

See

-0.90

percent

-0.70

SEE

-0.69

 Otherwise

-0.62

 См

-0.60

Otherwise

-0.58

querade

-0.56

 Percent

-0.54

POSITIVE LOGITS

 متعلقه

0.81

ing

0.63

 volks

0.60

 sphinct

0.59

 carapace

0.57

 whereof

0.56

enumi

0.56

 Lleg

0.55

invokeLater

0.54

 Folsom

0.54

Activations Density 0.095%