INDEX

Explanations

changes or variations

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

 ilan

-0.07

çĨĦ

-0.06

 finances

-0.06

uston

-0.06

ä¸įå¤ļ

-0.06

æºĲæºĲ

-0.06

RESS

-0.06

éĤªéŃĶ

-0.06

ORD

-0.06

alai

-0.06

POSITIVE LOGITS

 changes

0.23

åıĺåĮĸ

0.22

 change

0.22

çļĦåıĺåĮĸ

0.21

åıĺåĮĸåĴĮ

0.20

 Changes

0.19

åıĺåĮĸçļĦ

0.18

change

0.18

è®ĬåĮĸ

0.18

çļĦæĶ¹åıĺ

0.18

Activations Density 0.446%