INDEX

Explanations

the word "Lake" followed by a body of water

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 newList

-0.85

parsedMessage

-0.81

 Ried

-0.80

 newVal

-0.80

 Aryan

-0.78

 Benfica

-0.78

 Meri

-0.78

 Machu

-0.78

NewLabel

-0.77

 Phan

-0.76

POSITIVE LOGITS

 Lake

1.79

Lake

1.56

 lakes

1.54

 lake

1.48

 Lakes

1.48

 LAKE

1.41

lakes

1.39

Lakes

1.30

lake

1.28

LAKE

1.26

Activations Density 0.003%