INDEX

Explanations

He

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

ंदीखरीदारी

-1.12

MessageOf

-0.88

 يتيمه

-0.82

AndEndTag

-0.72

DoubleQuotes

-0.66

afficheront

-0.66

 kaarangay

-0.65

 للاسماء

-0.65

 Савезне

-0.63

 Roskov

-0.63

POSITIVE LOGITS

He

0.74

He

0.68

 tapping

0.60

he

0.59

 Para

0.54

 Pará

0.52

 tapped

0.51

EqualsAnd

0.48

Para

0.46

he

0.45

Activations Density 0.001%