INDEX

Explanations

such as

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

ä»Ģä¹Īåĳ¢

-0.29

ä¸ĸçķĮç»ıæµİ

-0.22

å¤ªå¤ļäºĨ

-0.22

æĺ¯ä»Ģä¹Īåĳ¢

-0.21

rove

-0.19

éĻ·åħ¥äºĨ

-0.18

_ComCallableWrapper

-0.18

 uLocal

-0.18

uxtap

-0.17

Unified

-0.17

POSITIVE LOGITS

etc

0.30

çŃīçŃī

0.29

,and

0.26

çŃīæĸ¹éĿ¢

0.23

çŃīåİŁåĽł

0.23

***

0.22

çŃīå¤ļé¡¹

0.22

 ëĵ±ìĿĦ

0.21

çŃīæĸ¹å¼ı

0.20

ãģĿãģĹãģ¦

0.20

Activations Density 0.928%

No Known Activations