INDEX

Explanations

cover(s) or sleeve(s)

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

deal

-0.26

-minded

-0.23

iscal

-0.23

mony

-0.22

çº§

-0.22

 minded

-0.22

 Casc

-0.22

 breakpoint

-0.22

ìĪľ

-0.22

 dispersed

-0.22

POSITIVE LOGITS

 worn

0.33

 protect

0.28

æĬ¤

0.28

 cover

0.27

ç£¨æįŁ

0.27

 covers

0.26

 covering

0.25

 sleeves

0.25

æİ¥ç®¡

0.25

èħĲèļĢ

0.24

Activations Density 0.054%