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

 дописавши

-0.70

 виправивши

-0.69

RegressionTest

-0.63

 jScrollPane

-0.63

PathParam

-0.61

crisy

-0.60

IsContent

-0.58

Diwedd

-0.57

UpInside

-0.57

onded

-0.57

POSITIVE LOGITS

 myſelf

0.75

 auroit

0.70

 étoient

0.68

Efq

0.66

 ainfi

0.65

 avoient

0.64

 siapa

0.62

 myself

0.62

 aveug

0.62

 مشارکت‌کنندگان

0.59

Activations Density 0.009%