INDEX

Explanations

introduces explanations

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

 Ugly

1.38

स्य

1.38

િક

1.35

 wildly

1.29

ቶችን

1.29

íng

1.27

전히

1.26

 lembaga

1.25

ńsk

1.24

üğünüz

1.24

POSITIVE LOGITS

there

1.72

 there

1.62

它是

1.50

There

1.39

 There

1.37

had

1.31

contains

1.30

 знаходиться

1.28

 contiene

1.28

 THERE

1.28

Activations Density 0.377%