INDEX

Explanations

Guest Blogger

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

PRWEB

-0.79

 violate

-0.78

 złot

-0.77

湖南

-0.77

 المستخدم

-0.77

lineno

-0.77

quatre

-0.76

física

-0.76

 scout

-0.73

菱

-0.73

POSITIVE LOGITS

 succession

0.85

emi

0.85

 knowing

0.83

ेंद्र

0.81

 images

0.81

NB

0.81

 tors

0.81

мә

0.79

industrial

0.79

 applications

0.78

Activations Density 0.043%