Borcherds' finiteness problem for singular-weight automorphic products and the Gritsenko-Nikulin conjecture on strongly reflective modular forms
Let be an even lattice of signature with and a finite-index subgroup of . A non-constant holomorphic modular form for is reflective if its zeros lie on mirrors of reflections in , and strongly reflective if all these zeros are simple. The weight of a non-constant form is at least , the singular weight. Borcherds products (automorphic or modular products) are the forms with infinite product expansions produced by the singular theta lift. Borcherds asked in 1995, as Problem 3 of his list of open problems: are there a finite or infinite number of automorphic forms of singular weight that can be written as modular products? Gritsenko and Nikulin conjectured (Conjecture 2.5.5, stated for their Lie reflective automorphic forms, which have simple zeros on mirrors) that such forms, and the lattices that carry them, are finite in number up to rescaling the lattice. In the form studied here: are there only finitely many strongly reflective modular forms on , , up to conjugation and scalar multiples?
- Result
- Proved(see note)
- Status
- Candidate (review pending)
- AI contribution
- AI-assisted
- Method
- Argument
- Field
- Automorphic forms on orthogonal groups; Borcherds products
- Posed by
- Richard Borcherds (Invent. Math. 120, 1995, Section 17, Problem 3); Gritsenko and Nikulin (Conjecture 2.5.5, preprint 1996, published 1998)
- Year posed
- 1995
- Years open
- 31y
- Solved
- 2026-11-12
- Model
- Generative AI (model not named in the source)
- Vendor
- —
- Collaborators
- Jialin Li, Haowu Wang
- Verification
- Unreviewed
- Publication
- Preprint
- Significance
- 30 / 100
- Disclosed cost
- —
- Wikipedia
- No dedicated article
What was actually shown
Theorem 1.3 of arXiv:2611.16819 (Jialin Li and Haowu Wang, Wuhan University, 16 pages): for each fixed there are only finitely many reflective modular forms on , , whose zero mirrors all have multiplicity at most , up to the equivalence with and . All weights, multiplier systems, finite-index subgroups and even lattices of signature are allowed. With this is finiteness of strongly reflective modular forms and, through Wang-Williams, of holomorphic automorphic products of singular weight. Corollary 1.4: up to conjugation, only finitely many arithmetic groups have a freely generated algebra of modular forms. The theorem gives finiteness, not a classification, and it is stated with the multiplicity bound : finiteness of all reflective forms at once is not claimed.
What the AI did
From the paper's Disclosure of AI use: the authors developed the initial mathematical framework and the proof ideas themselves, and then used generative AI interactively in mathematical exploration, including literature searches and the development, testing and refinement of parts of the proof. No model or vendor is named. The authors state that they verified the final arguments independently and take full responsibility for the paper.
Verification
No independent mathematician has checked this yet. Read by this site on 19 November 2026: the arXiv record and PDF of 2611.16819v1 (introduction, Theorem 1.3, the proof outline and the AI disclosure). Checked against the originals: Borcherds's 1995 paper, Section 17, Problem 3, reads as the paper quotes it; Gritsenko-Nikulin Part I (alg-geom/9610022) contains Conjecture 2.5.5, stated for Lie reflective forms, a subclass of the strongly reflective forms treated here, so the paper's version is a reformulation that covers it. Wang-Williams (arXiv 2301.13367) was opened to confirm that it settles the other half of Borcherds's Problem 3 and not the finiteness question. The proof was not refereed here; it rests on the higher-rank convergence theorem of Abert et al., Ma's volume estimates and H. C. Wang's finiteness theorem, none of which was re-checked. There is no formal proof. Candidate because a thirty-year-old problem is claimed on the authors' word.
Sources
- PaperarXivWang and Williams, On the non-existence of singular Borcherds products (the other half of Problem 3)
- Problem recordBorcherds, Automorphic forms on O_{s+2,2}(R) and infinite products (1995): Section 17, Problem 3Gritsenko and Nikulin, Automorphic forms and Lorentzian Kac-Moody algebras, Part I: Conjecture 2.5.5
Submitted by MellowGander562 on