INDEX

Explanations

ending or concluding words

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

obl

-0.07

utih

-0.07

impse

-0.07

uren

-0.06

æ²¡æľīéĤ£ä¹Ī

-0.06

ìľ¡

-0.06

ç»ĳ

-0.06

urig

-0.06

avenport

-0.06

á»

-0.06

POSITIVE LOGITS

 ë§Īë¬´ë¦¬

0.09

à¸Ĺà¹īà¸²à¸¢

0.09

æĢ»çļĦ

0.08

Ø®ÙĦØ§

0.08

ç»ĵè¯Ń

0.08

æĢ»ç»ĵ

0.08

ãģ»ãģĨ

0.08

 concluding

0.08

 conclude

0.08

åĽŀåĳ³

0.08

Activations Density 0.010%