INDEX

Explanations

change, use, added, ability

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

ersch

-0.07

ä¿ı

-0.07

adu

-0.06

asih

-0.06

åĲįãģ®

-0.06

éĴ§

-0.06

bai

-0.06

uncate

-0.06

éĩĳé¢ĿçļĦ

-0.06

 Ð½Ð°Ð¿ÑĢ

-0.06

POSITIVE LOGITS

 certain

0.10

ä¸Ĭè¿°

0.10

 aforementioned

0.09

çļĦçĽ¸åħ³

0.09

 relevant

0.09

çĽ¸åħ³

0.09

æĸ¹æĸ¹éĿ¢éĿ¢

0.09

 accordingly

0.08

æŁĲäºĽ

0.08

 various

0.08

Activations Density 0.321%