INDEX

Explanations

Varied, somewhat negative text

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

ChildScrollView

-0.61

unted

-0.57

ht

-0.57

unting

-0.56

 remembered

-0.54

nting

-0.52

createComponent

-0.51

 referred

-0.50

blers

-0.50

 stunned

-0.50

POSITIVE LOGITS

tlement

0.63

ième

0.62

TagMode

0.62

󠁣

0.61

KommentareTeilen

0.58

shaft

0.56

Personensuche

0.55

💼

0.54

SourceChecksum

0.54

sing

0.54

Activations Density 0.076%