INDEX

Explanations

question and answer pairs

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

じゃなくて

-1.45

each

-1.31

 also

-1.31

 resultaten

-1.31

Dichloroprop

-1.30

 APPELLATE

-1.30

 ваши

-1.29

ateľ

-1.24

 each

-1.24

 your

-1.24

POSITIVE LOGITS

1.70

1.27

).

1.26

 Especially

1.25

хи

1.25

öffentlichung

1.22

 Unless

1.20

 suena

1.16

 Weird

1.16

ycznych

1.16

Activations Density 0.023%