INDEX

Explanations

traits including yield, fire damage, offices, depression, blockage, or steps

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·Ð°ÑĢÐµÐ³Ð¸ÑģÑĤ

-0.26

 Ð¿ÑĢÐ¸ÑıÑĤÐ½

-0.23

è¾Ī

-0.23

à¸Ħà¸ĵà¸°à¸ģà¸£à¸£à¸¡

-0.22

 Ð¿Ð¾ÑģÐºÐ¾Ð»ÑĮ

-0.22

à¹Ģà¸Ĺà¸Ħà¹Ĥà¸Ļ

-0.21

 ÐĶÐ¼Ð¸ÑĤÑĢ

-0.21

à¸Ħà¸°à¹ģà¸Ļ

-0.20

POSITIVE LOGITS

ãĢģãĢģ

0.38

ãĢģ

0.35

ãĢģ↵↵

0.26

ï½¤

0.24

ï¼īãĢģ

0.24

ãģ¯ãĤĤãģ¡ãĤįãĤĵ

0.23

-CS

0.21

è¸¯

0.20

ãĢįãĢĮ

0.19

 including

0.19

Activations Density 0.465%