INDEX

Explanations

regions and provinces

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

 worldwide

-0.15

åħ¨çĲĥ

-0.12

 world

-0.12

ä¸ĸçķĮ

-0.11

 ÐºÑĢÐ°ÑĹÐ½

-0.10

 global

-0.10

 vere

-0.10

 countries

-0.10

 ÑģÑĤÑĢÐ°Ð½

-0.10

 Countries

-0.10

POSITIVE LOGITS

 regions

0.27

 Regions

0.22

 provinces

0.21

regions

0.20

 interior

0.18

 Northern

0.18

 central

0.18

Regions

0.17

à¸łà¸²à¸Ħ

0.17

åħ¨åĽ½

0.17

Activations Density 0.176%