INDEX

Explanations

dom, DOM, domus, dominated, domain

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

omnia

-0.10

bands

-0.10

 Sons

-0.09

 neph

-0.09

İ

-0.09

leton

-0.09

ãĤ£

-0.09

wick

-0.09

burn

-0.08

Ä±lÄ±ÄŁÄ±yla

-0.08

POSITIVE LOGITS

estic

0.25

dom

0.16

ains

0.14

Dom

0.13

(dom

0.13

inant

0.13

ino

0.12

inion

0.12

Ã©st

0.12

dom

0.12

Activations Density 0.018%