INDEX

Explanations

presidential elections

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

 Everett

-0.10

âĶĤ

-0.10

 Voter

-0.10

ä¿

-0.10

uned

-0.09

 Electoral

-0.09

 nominee

-0.09

clinic

-0.08

 finalist

-0.08

elter

-0.08

POSITIVE LOGITS

 elections

0.34

 election

0.33

 Ø§ÙĨØªØ®Ø§Ø¨Ø§Øª

0.26

election

0.22

 presidential

0.22

lection

0.21

-election

0.20

Ø®Ø§Ø¨

0.20

éĢī

0.19

 Election

0.19

Activations Density 0.065%