INDEX

Explanations

systems and components

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

é£ĺ

-0.23

ideo

-0.23

dish

-0.22

hone

-0.21

UCE

-0.20

æĿ¥åĽŀ

-0.20

zza

-0.20

åıįæŃ£

-0.19

er

-0.19

uish

-0.19

POSITIVE LOGITS

ç³»åĪĹäº§åĵģ

0.29

äº§åĵģ

0.26

è§£åĨ³æĸ¹æ¡Ī

0.25

éĢĤçĶ¨äºİ

0.24

å¹¿æ³Ľ

0.24

 products

0.23

 ÑģÐµÑĢÐ¸Ð¸

0.23

é«ĺæķĪ

0.21

äº§åĵģçļĦ

0.21

ãĤ·ãĥªãĥ¼ãĤº

0.21

Activations Density 0.060%