INDEX

Explanations

copyright holder or author

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

liest

-0.10

nek

-0.08

Ã¡nd

-0.08

 conform

-0.08

 Charm

-0.08

lier

-0.08

mq

-0.07

 Karn

-0.07

elin

-0.07

pending

-0.07

POSITIVE LOGITS

 author

0.29

 Author

0.23

 authors

0.22

Author

0.21

author

0.19

 Authors

0.18

 AUTHOR

0.17

-author

0.16

 writer

0.16

.author

0.16

Activations Density 0.181%

No Known Activations