INDEX

Explanations

importantly, crucially, paramount

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

芒果

1.43

jupyter

1.41

 nvidia

1.37

RDF

1.33

 AIAA

1.33

Waterfall

1.32

mimeTypes

1.29

ropshire

1.28

osterol

1.28

zechoslovak

1.27

POSITIVE LOGITS

Ke

1.93

Ve

1.90

De

1.89

Der

1.88

Le

1.86

Ar

1.86

Ne

1.86

Per

1.81

Es

1.80

Ge

1.79

Activations Density 1.699%