AI-GENERATED ASSESSMENT · 7 OCTOBER 2026
How much mathematical progress do 722 manuscripts represent? A catalogue-wide assessment of the claims, their significance, and their historical context.
THE CENTRAL ESTIMATES
These are editorial estimates calibrated to VibeMathed’s rubric and actual entries. They are not scores published by VibeMathed. Strict cutoffs apply: a score of 80 does not count above 80. Alternative proofs count separately as manuscripts.
| Cutoff | Preprints | Families | Preprints with ±5-point shifts |
|---|---|---|---|
| >60 | 95 | 64 | 42–173 |
| >70 | 27 | 19 | 13–42 |
| >80 | 7 | 5 | 3–11 |
| >87 | 3 | 2 | 0–5 |
The ranges shift every nonfixed score down or up by five points. They are sensitivity exercises, not statistical confidence intervals. Selected published topic anchors remain fixed.
Two quasi-Riemann proofs at 90; CM Hodge at 88; restricted geometric Langlands and Vlasov–Maxwell at 85; and two Anderson-model papers at 82. The three papers above 87 therefore represent two breakthroughs. CM Hodge’s placement just above 87 is especially debatable.
Inspect the seven manuscriptsVibeMathed rates matrix multiplication at 55 and explicitly excludes the size of the particular improvement from that rating. This review preserves the topic anchor, even for the proposed jump from 2.371177 to 2.25. Problem importance and amount of progress are different axes.
Read the calibration comparisonEvery enumerated main-list question was screened. Of 504 problem entries, 37 have a claimed full answer, 44 have claimed partial progress, 32 have related work that does not answer the listed assertion, and 391 have no matching claim identified. The dataset also preserves 12 grouping headings and 44 already-solved appendix entries.
These are row counts, not distinct conjectures or certified new discoveries. “Partial” can still mean a huge breakthrough: Hilbert’s tenth problem over the rationals is partial relative to the broader number-field question. The Millennium matches are partial for RH, BSD and Hodge; related only for Navier–Stokes and P versus NP; and no matching resolution for Yang–Mills.
Inspect every Wikipedia match and its scopeTHE WORK BEHIND THE FINDINGS
Choose a dataset, search, then open a row for its full reasoning and source links. Each record has a link you can share.
Loading the assessment…
A READING GUIDE
Descriptions below concern the announced claims. The associated records preserve hypotheses and qualifications.
QRH excludes zeta zeros from Re(s) > 7/8 at arbitrarily large heights. Classical zero-free regions thin with height; a fixed strip would cross a qualitative barrier and yield fixed-power savings in prime-counting errors. RH still requires the much stronger boundary 1/2.
Start with the alternate 11/12 proof, then compare the stronger 7/8 argument. Focus on the analytic continuation and control of the constructed objects.
Claims, scope and papersThe claim covers every CM abelian variety, in every dimension and codimension. Through existing theorems, it would also imply Tate for all abelian varieties over finite fields and the Hodge standard conjecture for abelian varieties in every characteristic.
This remains a special case of general Hodge, but “special case” understates its possible reach. Companion results concern K3 products and Kuga–Satake correspondences.
Claims, scope and papersThe proposed μ(π) = 2 means rational approximations cannot repeatedly beat q⁻² by any fixed power. It does not establish normal digits or bounded continued-fraction coefficients.
The Flint Hills consequence, convergence of Σ 1/(n³ sin² n), is particularly appealing. Both the quality and the sparsity of unusually close approximations matter.
Claims, scope and papersThe claim L(F₂) ≅ L(F₃) would show that different free groups yield the same von Neumann algebra, with consequences for all interpolated free factors.
When reading, scrutinize generation: exhibiting operators with the right joint distribution is insufficient without showing they generate the entire required factor.
Claims, scope and papersThe complex arithmetic exponent 2.25 would remove roughly one third of the remaining exponent gap to two from the VibeMathed anchor of 2.371177. This is an asymptotic arithmetic result, not a measured hardware speedup; the strongest bound is over the complex numbers.
Claims, scope and papersUnique Games concerns the NP-hardness of distinguishing almost satisfiable permutation-constraint systems from poorly satisfiable ones. Its reach comes from a network of sharp approximation consequences. It does not prove P ≠ NP, and its corollaries should not each become an independent breakthrough count.
Claims, scope and papersSidorenko’s proposed counterexample suppresses a bipartite pattern below its random-graph benchmark. Hadwiger’s would overturn a central colouring/minor relationship. Borsuk’s failure in dimension nine uses a striking geometry of rank-one projectors from vectors in four dimensions.
A negative answer changes the search for the right structural principle.
Explore the counterexamplesRelativistic particles have bounded speed, but their momenta can still diverge. The proposed global theorem addresses arbitrary smooth data in the stated one-species class, without smallness or symmetry.
The proof summary organizes control around signed force cancellations and the time required for momentum doublings. The hard part is making these estimates close.
Claims, scope and papersAnderson spectral localization and delocalization, the pure spin-one Heisenberg gap, and quantum ferromagnetic magnetization/Bloch laws form a remarkable cluster. Spectral type does not automatically settle all transport questions. In magnetization, the order of infinite-volume and zero-field limits is essential.
Explore the Heisenberg resultsErdős’s reciprocal-sum conjecture says that a set A of positive integers with divergent Σ 1/a contains arbitrarily long arithmetic progressions. This reaches much sparser sets than a positive-density hypothesis, including the primes, and is accompanied here by quantitative Szemerédi claims.
Claims, scope and papersHISTORICAL CALIBRATION
A conditional order-of-magnitude judgment: several dozen, with an illustrative range of 20–50. This is not a measured median or a prediction of awards.
The reference set covers every medalist from 1998 through 2026 using official citations and contemporary work profiles. There was no award in 2000. Later achievements are excluded from earlier award portfolios; the review did not audit each laureate’s complete publication record.
The historical examples vary. Lafforgue and Ngô completed central programmes through deep new machinery. Maynard and Avila supplied multiple connected breakthroughs. Voevodsky, Hairer and Scholze developed mathematical languages with broad reach. A theorem tally cannot capture these different combinations of centrality, depth, methods, synthesis and influence.
The proposed 28 portfolios use 234 distinct repo families without reusing a family. This prevents one kind of double counting, but shared inputs and dependencies still matter. Some clusters could split; others require much more evidence about novelty and reusable methods before the comparison is persuasive.
Birational geometry with Birkar; Kakeya, restriction and smoothing with Wang and Tao; critical phenomena with Duminil-Copin, Smirnov and Werner; kinetic PDE with Villani and Deng; Langlands with Lafforgue and Ngô.
Inspect all 28 proposed portfoliosSeveral solved equations do not automatically equal regularity structures. Several arithmetic conjectures do not automatically equal perfectoid spaces. The strongest assessment must inspect what the proofs make possible beyond their immediate targets.
Inspect all 30 historical profilesMETHODS, LIMITS & SOURCES
All 372 family summaries were screened. Within the 64 families rated above 60, all 195 manuscript scopes were considered separately. Remaining manuscripts inherit their below-threshold family estimate. The 95-paper count therefore differs from the 195 papers merely associated with a highly rated family.
Wikipedia matching preserves bundled questions and known duplicates. “Claimed full answer” means that the announced statement appears to answer the listed assertion, not that its proof is certified. “Related only” deliberately excludes tempting keyword matches such as mutually unbiased bases versus SIC-POVMs.
A Lean link beside a family is not a blanket certification of every manuscript. In family 157, the linked scope concerns a positive list-colouring bound rather than the Hadwiger counterexample. In family 266, the formal statement has a weaker MUB bound than the headline. In family 260, selected auxiliary statements do not cover the full spacetime Penrose claims. Conversely, the QRH, π-exponent and Vlasov–Maxwell scope documents target substantial headline content.
Correctness, fidelity of the formal statement, novelty, and significance are distinct questions. This assessment concentrates on catalogue scope and conditional significance. The published workbook records the original screening judgments; the website presents those judgments without claiming an additional proof review.
Repo snapshot: adc7f1241b42e322a6451854ab7e4b4c146bf78a. Wikipedia revision: 1378950817. VibeMathed dataset retrieved 7 October 2026. Individual records link to source papers and official medal citations.
Attribution: VibeMathed scores and dataset material, CC BY 4.0. Wikipedia wording, Wikipedia contributors, CC BY-SA. Repository claims are attributed to OpenAI. The match classifications, significance estimates, portfolio comparisons and commentary are AI-generated analysis.
Download the complete Excel workbook Download the data as JSON