INDEX

Explanations

explain or describe

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 192 tokens each

Dataset (Dashboard)

chanind/lmsys-chat-1m-qwen-2.5

Embeds

IFrame

Link

Negative Logits

().'/

-0.23

 sign

-0.22

BaseContext

-0.21

éĩįè¦ģåİŁåĽł

-0.21

_builtin

-0.20

 readdir

-0.20

è¦ģä¹Ī

-0.20

".$_

-0.19

+"<

-0.19

å°±ç»Ļå¤§å®¶

-0.19

POSITIVE LOGITS

akest

0.23

edio

0.21

voke

0.21

 tremend

0.20

 carr

0.20

 compaÃ±ero

0.20

ä¸¹

0.20

lobs

0.19

idth

0.19

 Åŀa

0.19

Activations Density 0.001%

No Known Activations