INDEX

Explanations

both ... and

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ÐĲÐ»ÐµÐºÑģÐ°Ð½

-0.29

 ÐŁÐ¾ÑĩÐµÐ¼

-0.27

 Ð¿Ð¾ÑģÐºÐ¾Ð»ÑĮ

-0.25

 Ð·Ð°ÑĢÐµÐ³Ð¸ÑģÑĤ

-0.24

æĹłè®ºæĺ¯

-0.23

æĹ¢æĺ¯

-0.23

 Ð¿ÑĥÑĤÐµÑĪ

-0.22

 ÐĶÐ¼Ð¸ÑĤÑĢ

-0.22

æĹ¢åı¯ä»¥

-0.22

 Ð¿Ð¾Ð´ÑĢÐ¾Ð±Ð½

-0.20

POSITIVE LOGITS

and

0.32

åĴĮç¤¾ä¼ļ

0.28

åĴĮ

0.27

 Ð¸

0.26

 ÙĪØ¨ÙĬÙĨ

0.26

åĴĮæĬĢæľ¯

0.25

åĴĮåıĳå±ķ

0.24

 vÃł

0.23

ìĻĢ

0.23

åĴĮæĶ¯æĮģ

0.22

Activations Density 0.580%