Domain 08 · Riemann Hypothesis
Millennium status, 20 October 2026: Clay continues to classify the Riemann Hypothesis as Unsolved. An October release reports a zero-free region Re(s) > 7/8, with an accompanying Lean formalization. This is strictly weaker than proving that all nontrivial zeros lie on Re(s) = 1/2. The announcement does not discharge GCL’s retained review blockers. Full comparison.
The question in everyday language
Prime numbers seem irregular, yet their distribution is connected to a remarkable mathematical function called the Riemann zeta function. The Riemann Hypothesis predicts where certain zeros of this function must lie. It is one of the most important unsolved questions in mathematics.
GCL has assembled the definitions, historical results and common weaknesses in attempted proofs. Checking many zeros by computer cannot settle a claim about every zero.
What remains uncertain: the conjecture is open. Completed software or documentary checks do not mean that a proof has been accepted; the exact limits of the earlier work are stated in the technical record below.
Detailed research record
Research surface authority
LIVE: RH-001 canonical tracker #163, with route trackers in MATHSOLVE.
AUTHORITY: RH-WP00-source-normalization-equivalence-audit.md, registered protected campaign records, and governance/rh_001_authority_transition_r080.json for the latest protected Route C transition.
EXPOSITION: this page is a human-facing projection. The issue governs mutable research state; protected records govern promoted/certified claims.
Campaign ID: RH-001
Mathematical status: open conjecture
Programme state: WP00 promoted; WP01 and WP02 implemented, merged, and CI-passed but not formally promoted
Governance: ADR-0009; ADR-0010
Accessible research guide: RH-001 Research Guide
Canonical challenge
Every nontrivial zero of the meromorphically continued Riemann zeta function should have real part 1/2, with zeros counted using the campaign's fixed multiplicity and range conventions.
The target is the classical Riemann Hypothesis. Generalized variants, simplicity, Lindelöf, density-one results, pair correlation, random-matrix evidence, finite verification, and incomplete spectral proposals are separate statements.
Programme posture
RH-WP00 fixes the zeta, completed-function, xi, Xi, and Hardy Z normalizations; pole and zero taxonomy; symmetry and multiplicity conventions; exact core equivalences; source-bound arithmetic criteria; spectral obligations; false-proof seeds; theorem spine; proof debt; and certification boundary.
PR #90 implemented and merged the two authorized successor packages:
RH-WP01— an executable eliminative false-proof atlas;RH-WP02— a source-normalized theorem, criterion, computation, evidence, and barrier ledger.
Their deterministic replay and integrated programme-policy workflow passed. They are not formally promoted: the governing legacy review records retain promotion_recommended: false and blocking Referee findings pending independent source-locator review and a schema-bound or explicitly superseding promotion decision.
Mechanism generation, claimed-proof promotion, unrestricted numerical search, novelty claims, and any new zero-range certification remain closed.
Latest protected forward transition
On MATHSOLVE protected main at e3c15b95722048be5ff0f186d85d24b41fd57a81, RH-R080 protects the signed-projective normality interface for Route C. It proves that pointwise positivity is stronger than necessary for the abstract C4 normal-family implication: uniform kappa_delta control on every strict substrip is sufficient. Actual CCM kappa_delta control remains open, so this does not discharge actual-family C4, determinant convergence, or RH.
Canonical artifacts
- WP00 integrated audit
- WP01/WP02 integration dossier
- WP01 false-proof atlas
- WP02 theorem ledger
- Post-merge retained-blocker disposition
- Merge record PR #90
- Catalogue-integration decision ADR-0009
- Continuity and documentary authority decision ADR-0010
Claim boundary
The programme has not proved or disproved RH, established a new zero-free region, critical-line proportion, prime-counting error estimate, equivalent criterion, Hilbert–Pólya operator, newly certified zero computation, or novelty result. Implementation, repository merge, replay, CI success, public documentation, and documentary publication are not theorem support.