INDEX

Explanations

same

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

 accett

-0.51

-0.47

onAttach

-0.47

↵

-0.46

↵↵

-0.45

 thermique

-0.45

it

-0.45

”

-0.43

or

-0.43

<eos>

-0.43

POSITIVE LOGITS

rungsseite

0.83

 Italijanski

0.80

 للمعارف

0.80

 дописавши

0.77

 resourceCulture

0.74

 autorytatywna

0.72

 betweenstory

0.72

SPATH

0.69

 ویکی‌پدی

0.67

 typelib

0.66

Activations Density 0.001%