INDEX

Explanations

the related calls

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

fac

-0.08

Fac

-0.07

fac

-0.07

ìĤ´

-0.07

oret

-0.07

Fac

-0.07

ellate

-0.07

erate

-0.07

æĺł

-0.07

eson

-0.07

POSITIVE LOGITS

åĵī

0.08

 inequality

0.08

 relationship

0.07

éĹ®é¢ĺåĴĮ

0.07

ä»Ĭå¤©è¦ģ

0.07

æĿ¡ä»¶åĴĮ

0.07

 equality

0.06

 phenomenon

0.06

rega

0.06

sel

0.06

Activations Density 0.009%

No Known Activations