INDEX

Explanations

would

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

-0.49

evident

-0.48

ล์

-0.48

 Reference

-0.47

Of

-0.46

 devons

-0.45

ofag

-0.43

 Crusoe

-0.43

 reference

-0.42

POSITIVE LOGITS

IVEREF

0.68

(!__

0.63

SBATCH

0.63

queryInterface

0.63

webElementGuid

0.62

elemField

0.61

 wireType

0.60

хьтан

0.60

ImageContext

0.58

 like

0.58

Activations Density 0.030%