INDEX

Explanations

introducing a topic or context

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

otts

-0.10

(I

-0.09

ÅĦ

-0.09

wang

-0.09

 Heath

-0.09

-0.08

Nab

-0.08

 pist

-0.08

à¸£à¸£

-0.08

POSITIVE LOGITS

ï¼Įå¥¹

0.11

ï¼Įä»ĸ

0.10

she

0.10

Â¸

0.10

essel

0.09

ãĢģ&#8203;

0.09

azole

0.09

,{"

0.09

 ê·¸ëħĢëĬĶ

0.09

Activations Density 0.119%