INDEX

Explanations

to

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

 owing

-0.47

dit

-0.47

Oneof

-0.47

-0.45

 blit

-0.43

<bos>

-0.43

–

-0.43

 shir

-0.43

-0.42

 comp

-0.42

POSITIVE LOGITS

tagHelperRunner

0.78

UnsafeEnabled

0.76

addContainerGap

0.71

Hochspringen

0.69

 betweenstory

0.66

 JpaRepository

0.66

 дописавши

0.66

Referencies

0.65

 متعلقه

0.65

 avoient

0.63

Activations Density 0.001%