INDEX

Explanations

ate

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

tvguidetime

-0.47

Which

-0.38

 которая

-0.36

 proposito

-0.36

 bestem

-0.36

 tillegg

-0.36

 näh

-0.35

več

-0.35

 HasFactory

-0.35

urysty

-0.35

POSITIVE LOGITS

ness

0.90

+#+#

0.84

the

0.81

nya

0.81

tagext

0.78

AccessorTable

0.75

 архивлан

0.75

setVerticalGroup

0.74

ring

0.73

 незавершена

0.73

Activations Density 0.042%