Browse › Number theory › 003 The quasi-Riemann hypothesis
The Quasi-Riemann Hypothesis: A Zero-Free Half-Plane $\Re s\gt 7/8$
Open PDFCite Report a problem Add a reference or comment Interested in working on this Start a discussion
Edit NOTES.md · Existing issues · Source on GitHub
Abstract
We prove that all finite-order Hecke L-functions over $\mathbb Q(\sqrt{-3})$ and all Dirichlet L-functions are zero-free in the half-plane $\Re s\gt 7/8$, with the principal pole at s = 1 allowed. In particular, the Riemann zeta function is zero-free in this half-plane, proving the quasi-Riemann hypothesis.
If the PDF does not display here, open it directly: Open PDF.
Notes
No notes yet. Add a reference, a comment or your name with the buttons above, or edit NOTES.md on GitHub.
Automatically collected notes (unreviewed)
work in progress
Collected automatically by AI tools (Claude Opus 5.5 and GPT-6 Astra, 2026-10-23) and not checked by a person. May be wrong or incomplete. Report an error.
Follow-up works
- Baiying Liu, Improved zero-free half-planes for the quasi-Riemann hypothesis (arXiv:2610.12234, v1 2026-10-08, v2 2026-10-12). Varies three parameters (b, ℓ, κ) of this paper's argument to lower the boundary from 7/8 to approximately $0.874957019420098946$, claims optimality within the paper's exponent comparison system, and states the result is formalized in Lean.
- Yuanda Xu, Hejian Sang, Amplified moments and a refinement of the quasi-Riemann hypothesis (arXiv:2610.15291, v1 2026-10-11, v2 2026-10-15). Assuming this paper's analytic inputs, raises the boundary to 0.8748867 via a relaxed direct estimate and an amplified first moment of the reflected numerator.
- Dean Cureton, Jesse Thorner, A quasi-Riemann hypothesis for Dedekind zeta functions (hexagon:2610.00200, v1 2026-10-12, v2 2026-10-14). Proves ζ_K(s) ≠ 0 for Re s > 31/32 for every number field K, following the strategy of the release's Artin-conjecture work and citing this paper; proofs and Lean formalization attributed to AxiomProver.
- Christopher Frei, Efthymios Sofos and Martin Widmer, Some applications of the quasi-Riemann hypothesis to class groups and small generators of number fields (arXiv:2610.18491v2, 2026-10-14). Uses the OpenAI and Cureton–Thorner results to derive bounds for class-group torsion, least splitting primes and small generators. [family-level]
- Boyang Hu, A seven-eighths zero-free half-plane for unitary Weil L-functions (hexagon:2610.00216, 2026-10-13). Extends this paper's cubic-theta method to Hecke characters over fields containing 12th roots of unity and, via Brauer induction, claims Re s > 7/8 for L-functions of unitary Weil-group representations; states the proof has not been independently verified by a human. [AI-generated per source]
- Michael Rubinstein, Mateo Restrepo Borrero, A zero-free strip for the L-functions of the elliptic curves y²=x³+d (hexagon:2610.00220, 2026-10-13). Part II adapts this paper's 7/8 argument to infinite-order characters over Q(√−3), giving L(E_d,s) ≠ 0 for Re s > 11/8 (arithmetic normalization); its provenance states the complete proof has not been independently checked by a human. [AI-generated per source]
- Shubhrajit Bhattacharya, Unconditional implications of the Quasi-Riemann Hypothesis (hexagon:2610.00178, 2026-10-11). Collects published RH/GRH-conditional conclusions whose proofs need only a fixed zero-free half-plane, taking this paper's Theorem 1.1 as input and distinguishing Dirichlet from fixed-field Hecke inputs; states no human scrutiny was performed. [AI-generated per source]
- Manjul Bhargava, Rational points on elliptic curves that are primitive modulo a positive proportion of primes (hexagon:2610.00294, 2026-10-17). Lang–Trotter-type result for the point (2,2) on y²=x³−2x that cites this paper; its zero-free input is the Cureton–Thorner Dedekind half-plane, kept as an explicit hypothesis in its Lean deduction.
- Sayed Barez, Ramanujan's prime counting inequality holds for all x ≥ 38,358,837,683 (hexagon:2610.00375, 2026-10-19). Uses this paper's 7/8 half-plane for ζ as the hypothesis of its main deduction, determining Ramanujan's threshold previously known only under RH. [AI-generated per source]
Comments and discussion
- Persiflage, A Brief Digest of Some New OpenAI Results, galoisrepresentations.org (2026-10-08). Item 5 calls quasi-RH "a very strong intermediate result" and notes that applications needing the Dedekind-zeta version (e.g. Stark–Odlyzko-type bounds) see no improvement "at least as yet".
- DH, comment on Persiflage's post (2026-10-09). Points out that the release's Artin write-up contains a more general quasi-RH covering the Dedekind zeta functions Hooley's argument uses; Persiflage replies that Hooley's middle range still seems to need full GRH.
- Harald Helfgott, comment on Persiflage's post (2026-10-11). Looks forward to studying the proof and remarks that it seems to confirm a "convex-hull metaphor" and involves theta functions.
- Maksym Radziwiłł, Jared Duker Lichtman, Alexander Dunn, Peter Sarnak, Will Sawin, quoted in Quanta Magazine (2026-10-19). Radziwiłł says both QRH papers fall short of the area's expository standard and that people focus on the shorter 11/12 version; Dunn calls the proof "a surprising synthesis of well-known ideas"; Sawin doubts the methods reach 1/2.
- Hector Pasten, quoted in Scientific American (2026-10-08). Notes that "plenty of applications of the Riemann hypothesis did not need the full Riemann hypothesis" and that opinions differ on whether this is progress toward RH.
- Jared Duker Lichtman, X post as quoted on Emergent Mind (2026-10-06). Observes that the cubic family of L-functions appears essential to the proof, even for the corollary for ζ.
- Terence Tao, A sketch of the proof of non-existence of Siegel zeroes (2026-10-19). Remarks that the QRH papers rely on an "automorphic miracle" to get past the parity barrier and are still being digested. [family-level]
- Terence Tao, comment on his tenth-discriminant AI challenge post (2026-10-20). Combines the Selberg–Chowla formula with the 7/8 (resp. 11/12) half-plane to bound a hypothetical tenth class-number-one discriminant by d < 655 (resp. d < 1475); LemmaLab's comment reports d ≤ 653 from the 7/8 input.
- Abhishek Saha (@ObhishekSaha), post on X (2026-10-14). Describes follow-up developments to the quasi-RH proof, starting with the Cureton–Thorner Hexagon preprint of October 12 that builds on OpenAI's proof. [family-level]
- Abhishek Saha (@ObhishekSaha), post on X (2026-10-20). Reports a preprint by Baiying Liu showing that all finite-order Hecke L-functions are zero-free to the right of 0.8749, improving on 31/32.
- John Carlos Baez (@johncarlosbaez.bsky.social), post on Bluesky (2026-10-20). Opens a thread explaining the zero-free strip and the resulting bound on |π(n) − li(n)| for p > 7/8 and p > 11/12, noting "Nobody has checked the first proof", and pointing to follow-up papers. [family-level]
Formalization
- OpenAI, The quasi-Riemann hypothesis: Lean scope note (openai/math release, 2026-10). Formalizes the 7/8 half-plane for ζ, all Dirichlet L-functions and finite-order Hecke L-functions over $\mathbb Q(\sqrt{-3})$ (principal poles excluded), plus the logarithmic real-zero gap; the paper's later applications are not included. [family-level]
- Baiying Liu, Lean repository for the improved half-planes (GitHub, 2026-10). Lean 4 project accompanying arXiv:2610.12234, stated there to formalize the improved zero-free conclusions for the same Hecke family and its Dirichlet and ζ consequences.
- LW56e2, Quasi-Riemann boundary refinement in Lean (GitHub, accessed 2026-10-21). Supplies a formal development adapting the original proof to the rational boundary $4374785349/5000000000$, covering the original Hecke family and its Dirichlet and zeta consequences.
- Zhang-Liao, Quasi-Riemann hypothesis — standalone Lean project (GitHub, build reported 2026-10-07). Extracts OpenAI's zeta and Dirichlet proof dependencies and reports a successful build and standard-axiom check, while stating that Comparator was not separately run.
- davegoldblatt, Independent check of OpenAI's Lean proof that ζ(s) ≠ 0 for Re(s) > 7/8 (GitHub, checks dated 2026-10-06–07). Provides logs and reproduction instructions for agent-operated Comparator checks with Lean and nanoda against the zeta statement, without claiming review of the prose proof.
- Axiom Math, Dean Cureton and Jesse Thorner, Dedekind Zeta Zero-Free Region (GitHub, accessed 2026-10-21). AxiomProver-generated Lean development for the $31/32$ half-plane over arbitrary number fields and finite-order Hecke characters, using analytic-number-theory code adapted from the OpenAI release. [AI-generated per source] [family-level]
Related prior literature mentioned by others
- Persiflage and commenters on Hooley's conditional proof of Artin's conjecture (2026-10-09). Discussed as an application that needs quasi-RH for Dedekind zeta functions rather than Dirichlet L-functions, and possibly full GRH; the post also names Stark–Odlyzko discriminant bounds and Vinogradov–Korobov zero-free regions as context.
- Alexander Dunn and Maksym Radziwiłł, Bias in cubic Gauss sums: Patterson's conjecture (Annals of Mathematics 200, 2024; arXiv:2109.07463v3). Cited in Rubinstein and Restrepo Borrero's follow-up, this work develops cubic-Gauss-sum transformations that also enter the quasi-Riemann arguments. [family-level]
Possibly relevant literature (found by the agent, not checked)
- Gautami Bhowmik, Imre Z. Ruzsa, Average Goldbach and the Quasi-Riemann Hypothesis (arXiv:1711.06442, Anal. Math. 2018). Relates the average order of Goldbach representations to zero-free half-planes of quasi-RH type, a direct application area for Theorem 1's statement about ζ.
Digests and expositions
- Ying Xie, What the Machine Invented: Mining New Mathematical Instruments from OpenAI's October 2026 Release of AI-Generated Mathematics (arXiv:2610.17090, 2026-10-12). Survey whose family-003 entry extracts the device of twisting the Möbius sum by all sextic residue symbols and classes it as "combination as architecture, leaning new".
- Claude Opus 5.5 and OpenAI Astra, at Michael O. Rubinstein's request, Zeros of a whole family: an accessible guide to the 7/8 argument for ζ(s) (working notes, accessed 2026-10-21). Explains the family-wide continuation argument, two evaluations of the probe and moment estimates, describing itself as a guide rather than an independent proof. [AI-generated per source]
- Singh-AP, ai-math-papers-explained: The Quasi-Riemann Hypothesis (GitHub, 2026-10). Popular-level explainer of this paper with a proof overview at three levels; text written with AI assistance and slides/audio generated with NotebookLM. [AI-generated per source]
News and blogs
- Konstantin Kakaes, AI Proved the Quasi-Riemann Hypothesis. What Happens Now?, Quanta Magazine (2026-10-19). Feature on this paper and its 11/12 companion, noting that the 199-page 7/8 paper is Lean-verified while experts study the shorter 11/12 one, and discussing their use of earlier cubic-Gauss-sum techniques.
- Zvi, New Math from OpenAI, LessWrong (2026-10-09). Roundup explaining applications of quasi-RH (deterministic modular square roots, prime counts) and collecting reactions.
- Shiva Kintali (@ShivaKintali), post on X (2026-10-08). Announces a short proof aiming to "substantially simplify OpenAI's proof" of the quasi-RH for some 1−c without optimizing c, obtaining c=1/48. [family-level]
Elsewhere
- Hexagon: A quasi-Riemann hypothesis for Dedekind zeta functions (human and AI; human understanding: some)
- Hexagon: A seven-eighths zero-free half-plane for unitary Weil L-functions (AI-generated; human understanding: some)
- Hexagon: A zero-free strip for the $L$-functions of the elliptic curves $y^2=x^3+d$ (AI-generated; human understanding: some)
- Hexagon: Rational points on elliptic curves that are primitive modulo a positive proportion of primes (human and AI; human understanding: some)
- VibeMathed: Problem record
- VibeMathed: Problem record
- whataifound: Registry entry
- Emergent Mind: family 003
Independent sites; their content is not reviewed here.
Cite
@misc{OAI:The-Quasi-Riemann-Hypothesis-September-30-2026,
author = {{OpenAI}},
title = {{The Quasi-Riemann Hypothesis:
A Zero-Free Half-Plane $\mathrm{Re}(s)>7/8$}},
howpublished = {OpenAI Math Release preprint
\href{https://github.com/openai/math/blob/main/preprints/The-Quasi-Riemann-Hypothesis-September-30-2026/paper.pdf}{OAI:The-Quasi-Riemann-Hypothesis-September-30-2026}},
year = {2026}
}Lean formalization
OpenAI provides Lean files for this family: lean/docs/003.md. The Lean statements have not been checked by people against the paper's claims.
Family 003: The quasi-Riemann hypothesis
Proves that every Dirichlet L-function, including $\zeta(s)$, is zero-free in $\Re s\gt 7/8$, resolving the quasi-Riemann hypothesis. The same half-plane is zero-free for every finite-order Hecke L-function over $\mathbb Q(\sqrt{-3})$. A companion gives a different proof of the zero-free half-plane $\Re s\gt 11/12$.
Other papers in this family:
- The Quasi-Riemann Hypothesis (alternate 11/12 proof) · October 5, 2026
- Uniform exclusion of Landau–Siegel zeros · October 19, 2026 Community-revisedNotes