INDEX

Explanations

references to Asia and especially North Korea

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

BeginContext

-0.66

انجليز

-0.59

 pořád

-0.57

 Rainforest

-0.56

 createSprite

-0.54

parsedMessage

-0.53

 ChromeDriver

-0.53

Coven

-0.52

 Kết

-0.52

 گردد

-0.51

POSITIVE LOGITS

 Korea

1.68

 Korean

1.66

 Koreans

1.55

 korean

1.52

Korea

1.52

Korean

1.50

 korea

1.48

Kim

1.48

Kim

1.36

 Seoul

1.36

Activations Density 0.032%