INDEX

Explanations

field names or specific locations

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 ÐĲÐ»ÐµÐºÑģÐ°Ð½

-0.28

èĵ¢

-0.26

buc

-0.25

äº²æĪļ

-0.25

å©»

-0.24

çİ¶

-0.24

æŁ¬åŁĶ

-0.23

å¢Ī

-0.23

æ´´

-0.22

 ÐĶÐ¼Ð¸ÑĤÑĢ

-0.22

POSITIVE LOGITS

-svg

0.27

 Mats

0.24

(XML

0.23

ê²¸

0.23

æ¡§

0.23

æ£Ģ

0.22

izin

0.22

olon

0.22

æł¸çĶµ

0.22

 Copp

0.22

Activations Density 0.003%

No Known Activations