I spent 40+ hours, and now I understand the mechanics of the proof of the L1 embedding conjecture by Open AI. I do not think I’ve internalized the proof yet, since I can’t explain how someone could have come up with it. Need more time to consolidate my understanding. minorfree.github.io/OpenAIDump/

Show Post

One solution: normalize desk-rejections. After a fast first assessment, papers are either desk-rejected, or go to full review. The current conf peer-review system has desk rejects as the exception, as it relied on self-selection prior to submission and small numbers. Neither of these hold anymore.

Show Post

I'm honored and excited to announce that this paper was selected to receive a Best Paper Award at FOCS 2026! Thanks to my coauthors Matt, Daochen, and Rain for including me in this work, and to the FOCS PC for selecting this despite it having a simple proof.

Show Post

TOCA-SV 2026: Friday 11/6 at Stanford (CoDa E160) sites.google.com/cs.stanford.... We have an exciting schedule featuring awesome invited speakers, as well poster and graduating-bits sessions (and, of course, lunch)!

Show Post

I've heard a lot of sad voices in our community who are "depressed" and "completely crushed" by the new wave of AI results. I'm actually very optimistic. I don't have particularly good rationales, but emotionally speaking, let me try to infect you with a bit of my optimism if you need it. 1/n

Show Post