// the find
teorth/optimizationproblems
A collection of optimization problems in mathematics
A crowdsourced table of best-known upper and lower bounds for ~90 open optimization constants in math (cap sets, moving sofa, matrix multiplication exponent, Ramsey numbers, etc.), maintained by Terence Tao, Damek Davis, and Paata Ivanisvili. It's for mathematicians and researchers who want a single place to check the current state of the art on a specific open problem instead of hunting through arXiv.
The changelog is genuinely rigorous — every bound update is dated, attributed, linked to a paper or repo, and corrected publicly when attribution turns out wrong (see the 4b entry fixing credit to Beigel/Gasarch). Several entries link out to Lean 4 formalizations with a registry ID, so some of the 'crowdsourced' claims are machine-checked, not just asserted. The asterisk convention for 'minimal verification' bounds is a real signal, not decoration — it tells you which numbers to double-check before citing. Broad coverage in one place is itself useful: cap sets, Grothendieck constants, the moving sofa problem, and prime gap work all sit in one table instead of forty scattered blog posts.
This isn't code — it's a Jekyll site over hand-edited markdown files, so normal repo signals (tests, structure, issue tracker discipline) don't tell you anything useful here. There's no machine-readable data source (JSON/CSV/API) behind the table, so you can't pull a single constant's history programmatically without scraping the markdown. The README is explicit that submissions only meet 'minimal standards of plausibility' and aren't certified, which means unverified (asterisked) bounds can sit for months before anyone checks them properly. Verification is inconsistent per entry — some have Lean proofs, most just have a linked PDF or GitHub repo you're expected to trust.