INDEX

Explanations

revision and rumors

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

-0.09

<![

-0.08

 cest

-0.08

ICY

-0.08

amma

-0.08

 &#8203;&#8203;

-0.07

eway

-0.07

Hin

-0.07

Cog

-0.07

POSITIVE LOGITS

LookAndFeel

0.08

mium

0.08

 Erotische

0.08

 -*-č\n

0.08

-------------</

0.08

Aur

0.08

ì¹´ëĿ¼

0.08

andaÅŁ

0.07

----------</

0.07

<|begin_of_text|>

0.07

Activations Density 0.651%