INDEX

Explanations

introduces specific terms

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

显

-1.75

 fidé

-1.66

 传统

-1.60

 个性

-1.58

 方案

-1.56

 célè

-1.52

 advirtió

-1.52

鹪

-1.52

减

-1.51

 zwölf

-1.48

POSITIVE LOGITS

and

1.73

ског

1.50

balanceOf

1.41

 seulement

1.33

 који

1.33

in

1.30

 دیگری

1.29

อีก

1.29

 other

1.28

 voks

1.27

Activations Density 0.001%