INDEX

Explanations

information or learning about topics

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

-scrollbar

-0.09

Ø

-0.09

urdu

-0.08

BorderStyle

-0.08

shit

-0.08

oppins

-0.08

forcements

-0.08

ckett

-0.08

Ã¢n

-0.08

çĻ

-0.07

POSITIVE LOGITS

 everything

0.15

 Everything

0.12

/about

0.11

 tudo

0.11

everything

0.11

Everything

0.10

iefs

0.10

 characteristics

0.10

 aspects

0.10

 properties

0.10

Activations Density 0.218%