INDEX

Explanations

policy reforms and changes

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

å¾Ģè¿Ķ

-0.11

hon

-0.09

ä¸įå¦¨

-0.09

æĪĺçģ«

-0.09

æĬķå½±

-0.08

äº«åıĹ

-0.08

å¾ĹçŁ¥

-0.08

æľīå¤ļå°ĳ

-0.08

 dispro

-0.08

CurrentValue

-0.08

POSITIVE LOGITS

æİªæĸ½

0.15

äºĨä¸Ģç³»åĪĹ

0.14

æĶ¹éĿ©

0.13

çĿĢæīĭ

0.13

ä¸Ģç³»åĪĹ

0.12

æķ´æĶ¹æİªæĸ½

0.12

å½»åºķ

0.11

åĪ¶åº¦æĶ¹éĿ©

0.11

 reforms

0.11

åĩºåı°

0.11

Activations Density 0.070%