INDEX

Explanations

flexibility, limitations, improves, avoids, power

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

ager

-0.07

 Notes

-0.06

 sometimes

-0.06

LR

-0.06

 otherwise

-0.06

iffs

-0.06

-notes

-0.06

åĲ¦åĪĻ

-0.06

æĺ¾å¾ĹæľīäºĽ

-0.06

 Ð·Ð°Ð²

-0.06

POSITIVE LOGITS

åķıé¡ĮãĤĴ

0.09

è¿ĻäºĽéĹ®é¢ĺ

0.09

éĹ®é¢ĺè§£åĨ³

0.08

erala

0.08

 Ð·Ð°ÑĤÐ¾

0.07

åģ¥åº·äº§ä¸ļ

0.07

ENAME

0.07

 hopefully

0.07

Ø±ÙĪÙĨ

0.07

æŀľçĦ¶

0.07

Activations Density 0.249%