INDEX

Explanations

clear: both

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

𝘽

-0.81

Grandpa

-0.73

Makefile

-0.72

Associate

-0.71

 Leute

-0.71

 områ

-0.70

associate

-0.69

 Wyatt

-0.68

فات

-0.66

toppers

-0.66

POSITIVE LOGITS

fix

0.71

そうです

0.69

 fixed

0.66

 Grateful

0.64

ium

0.62

 Vise

0.61

 genital

0.61

 Bomber

0.61

 cache

0.60

CACHE

0.60

Activations Density 0.056%