INDEX

Explanations

usefulness

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

 TestBed

-0.51

housie

-0.47

Viitteet

-0.47

Xna

-0.45

BeforeAll

-0.44

 "..\..\..\

-0.43

LookAnd

-0.42

بوابة

-0.42

帖最后由

-0.41

⏟

-0.40

POSITIVE LOGITS

 pozdě

0.75

 myö

0.71

 later

0.67

in

0.65

jahteraan

0.63

 someday

0.63

 senere

0.62

 prossima

0.61

 später

0.61

 down

0.61

Activations Density 0.004%