INDEX

Explanations

[entity name] is/are

New Auto-Interp

Top Features by Cosine Similarity

Configuration

Prompts (Dashboard)

10,000 prompts, 128 tokens each

Dataset (Dashboard)

lmsys/lmsys-chat-1m

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ouri

-0.09

ahas

-0.08

ISR

-0.08

 KÃ¼ltÃ¼r

-0.08

à¹Ģà¸īà¸ŀà¸²à¸°

-0.08

 Fetish

-0.08

uggy

-0.08

ISR

-0.08

 incom

-0.08

POSITIVE LOGITS

æĺ¯ä¸Ģ

0.18

 adalah

0.14

 Ã¤r

0.12

æĺ¯ä¸Ģä¸ª

0.12

ãģ¨ãģ¯

0.11

 æĺ¯

0.11

ä¸Ģç§į

0.10

ä¹ĥ

0.10

(/

0.10

 ÎµÎ¯Î½Î±Î¹

0.09

Activations Density 0.228%