INDEX

Explanations

references to measurements of light or illumination

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

GraphicsUnit

-0.59

fillment

-0.47

head

-0.46

vings

-0.43

 UIImage

-0.43

vimento

-0.43

чого

-0.43

GREES

-0.42

Head

-0.41

complexType

-0.41

POSITIVE LOGITS

 heads

2.58

 Heads

2.56

Heads

2.30

heads

1.47

 cabezas

1.04

 têtes

0.83

sizeCache

0.82

 resourceCulture

0.78

NameInMap

0.70

PhysRevLett

0.70

Activations Density 0.001%