INDEX

Explanations

of/the

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

Which

-0.54

quels

-0.54

 hvilket

-0.53

 calendriers

-0.51

 laquelle

-0.50

 opportunità

-0.49

wanych

-0.49

Lähteet

-0.49

 lesquels

-0.49

 lequel

-0.49

POSITIVE LOGITS

 oprot

0.73

 that

0.65

even

0.64

 sice

0.60

 even

0.59

XmlType

0.55

tze

0.54

nloa

0.53

STATUS

0.52

 accordingly

0.52

Activations Density 0.006%