INDEX

Explanations

lassen, zen, arrange

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

-0.10

-0.09

Ã·

-0.09

Sta

-0.09

 Sherlock

-0.09

ujet

-0.09

 Caldwell

-0.09

ãģ¨ãģª

-0.09

deaux

-0.08

unde

-0.08

POSITIVE LOGITS

 having

0.36

 Having

0.29

having

0.29

 à¤ķà¤°à¤µ

0.27

Having

0.26

 have

0.22

 Ð¸Ð¼ÐµÑĤÑĮ

0.22

 memiliki

0.20

 Have

0.20

 avoir

0.20

Activations Density 0.109%