INDEX

Explanations

go

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

TintMode

-0.61

Демографія

-0.60

Опис

-0.60

نیم

-0.59

$_['

-0.59

batore

-0.58

 distanciation

-0.57

 Strain

-0.57

 باخ

-0.56

 sưu

-0.56

POSITIVE LOGITS

AndEndTag

0.63

UnknownFields

0.56

 GenerationType

0.46

放

0.46

Witam

0.42

 cousins

0.42

TRON

0.41

 boxes

0.40

 math

0.40

 siblings

0.40

Activations Density 0.001%