INDEX

Explanations

referring or talking about issues

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

 ÐĶÐ¼Ð¸ÑĤÑĢ

-0.26

 ÐĲÐ»ÐµÐºÑģÐ°Ð½

-0.23

æµģæ·Į

-0.23

 Ð¼Ð¾ÑīÐ½

-0.22

æĸ°ä¸Ģè½®

-0.22

 Ð¿ÑĢÐ¸ÑıÑĤÐ½

-0.22

æĶ¶ç¼©

-0.21

modern

-0.21

ç©¹

-0.21

kah

-0.21

POSITIVE LOGITS

 diffs

0.25

/disc

0.24

åħ·ä½ĵåĨħå®¹

0.22

REET

0.22

/lic

0.22

 Loud

0.22

Ã¤ren

0.21

 ÑĦÐ°ÐºÑĤ

0.21

/sub

0.21

azine

0.21

Activations Density 0.060%