INDEX

Explanations

Research/organizations

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

原始内容存档于

-0.71

oredCriteria

-0.71

hyrchwyd

-0.69

 lodge

-0.68

 Lodge

-0.66

inga

-0.65

 Drugs

-0.65

WireFormatLite

-0.63

 gili

-0.62

GEBURTSDATUM

-0.61

POSITIVE LOGITS

 TextAppearance

0.56

%)$

0.53

mål

0.51

totta

0.49

 Sheppard

0.48

må

0.47

 tenuto

0.47

quini

0.47

 weft

0.46

"</

0.46

Activations Density 0.128%