A new algorithm that is potentially relevant to the security of Dilithium

369 views
Skip to first unread message

Noah Stephens-Davidowitz

unread,
Nov 6, 2025, 5:18:27 PM (10 days ago) Nov 6
to pqc-forum

Natalia

unread,
Nov 6, 2025, 10:00:41 PM (9 days ago) Nov 6
to Noah Stephens-Davidowitz, pqc-forum, Natalia
This sounds very interesting and very difficult from what we have been studying so far. Care to make a Youtube video about it, so we can all follow along?

Natalia D'Onofrio, FRP, MBA
Florida Supreme Court Certified Mediator: Circuit, County, Family, Dependency
Florida Supreme Court Qualified Arbitrator
​The Florida Bar Certified Arbitrator & Mediator
Guardian ad Litem

Affinity Ventures, LLC
P.O. Box 2059
Palm Beach, FL 33480


--
You received this message because you are subscribed to the Google Groups "pqc-forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pqc-forum+...@list.nist.gov.
To view this discussion visit https://groups.google.com/a/list.nist.gov/d/msgid/pqc-forum/a0eec9e3-a6bf-4c91-8580-e232f0f6d871n%40list.nist.gov.

Muhammed Esgin

unread,
Nov 6, 2025, 10:34:03 PM (9 days ago) Nov 6
to pqc-forum, Natalia, pqc-forum, Noah Stephens-Davidowitz

Thanks for sharing this Noah! Very interesting and I agree with your points.

I was very worried when I saw your post that the attack works for s = q/(2k) and m >~ n^k. For k=1, that could have been very worrying for some schemes. Thankfully, their Theorem 1.5 requires k>=2; so clarifying this for everyone. That means m needs to be at least in the order of n^2.

-Muhammed
Reply all
Reply to author
Forward
0 new messages