INDEX

Explanations

lists of bracketed numbers

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

 russo

-1.59

 vermel

-1.59

’)

-1.55

offering

-1.52

 pongo

-1.48

still

-1.48

sistent

-1.41

蟊

-1.41

should

-1.39

helping

-1.38

POSITIVE LOGITS

因為

1.38

 більш

1.37

توضیحات

1.37

تحميل

1.34

 apă

1.31

 Because

1.29

Día

1.28

作家

1.25

十一

1.23

يكل

1.23

Activations Density 0.103%