INDEX

Explanations

change, increase, and revision

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

å°±çŁ¥éģĵ

-0.25

èĢĮåľ¨

-0.19

Pac

-0.19

è¿Ĳè½¬

-0.19

ä¸īçĤ¹

-0.19

ç²Ĵ

-0.18

_filtered

-0.18

à¸ģà¸²

-0.17

æĿ¯

-0.17

çĽĪ

-0.17

POSITIVE LOGITS

åıĺåĮĸ

0.25

åĩºäºİ

0.21

 changes

0.21

ffen

0.21

 resc

0.20

avors

0.20

éī´äºİ

0.20

changes

0.19

å¤īãĤıãĤĭ

0.19

å±ĬæĹ¶

0.19

Activations Density 0.016%