INDEX

Explanations

then [pronoun/article/preposition]

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

-0.27

çļĦéĩįçĤ¹

-0.23

mod

-0.23

å¦«

-0.23

éĩįçĤ¹åħ³æ³¨

-0.23

thus

-0.23

tered

-0.23

 hence

-0.22

åıĤä¼ļ

-0.22

illus

-0.21

POSITIVE LOGITS

äººå®¶

0.24

ä¿Ĺè¯Ŀè¯´

0.24

è¯´æĺ¯

0.24

 rumor

0.23

 never

0.23

éĤ£æĺ¯

0.23

 contrad

0.22

åĪ¶éĢł

0.22

 saying

0.22

 ideal

0.21

Activations Density 0.007%