Documentary state: EDITORIALLY_ADMITTED. The manuscript has a protected editorial admission, not a theorem certification.
OAI-26 family 002: the exact Birch–Swinnerton-Dyer formula in low Selmer corank
Documentary state: SOURCE-BASED DRAFT. The source announces a proof; this article does not validate that proof or certify the Birch–Swinnerton-Dyer conjecture.
The question in plain language
An elliptic curve is a cubic equation with a distinguished point whose rational solutions form an abelian group. Two different measurements of that curve often seem to agree. The arithmetic rank counts independent rational points of infinite order; the analytic rank counts how many times the curve's associated L-function vanishes at its central point.
The Birch–Swinnerton-Dyer (BSD) conjecture says more than just that these two ranks agree. It predicts the exact first nonzero Taylor coefficient of the L-function in terms of the height regulator of rational points, real periods, local correction factors, torsion, and the Tate–Shafarevich group. The last group measures failures of a local-to-global principle and is not known to be finite in general.
The October 7 OAI-26 manuscript claims a complete leading-coefficient formula for every elliptic curve over the rational numbers provided that the full power-Selmer group has corank zero or one at at least one prime. The condition concerns a Selmer group, not merely a small observed number of rational points.
A small example
Imagine measuring the same landscape by two methods. The number of independent directions you can travel is analogous to the rank of rational points. A measurement of volume along those directions is analogous to the BSD leading coefficient. Establishing the correct number of directions does not establish the exact volume: it leaves several multiplicative factors to determine.
At analytic rank zero, the BSD formula predicts a nonzero central L-value expressed through a real period, local factors, a finite Tate–Shafarevich group and squared rational torsion. At rank one, the central L-value vanishes and the first derivative is controlled by the same factors together with a height regulator. The regulator equals one by convention in rank zero. Higher ranks would involve higher derivatives and a higher-dimensional regulator, but they are not covered by the low-corank hypothesis here.
As a schematic numerical analogy, knowing that the product of several positive factors is finite cannot tell us their exact product. The manuscript claims an exact equality of such quantities, not just that both sides are finite or have the same order of vanishing. This analogy is not evidence for its proof.
What is known
For an elliptic curve over the rational numbers, let the analytic rank be the order of vanishing of its L-function at the central point, and let the arithmetic rank be the rank of its rational point group. For each prime, the full prime-power Selmer group has a corank. The main manuscript claims:
- If some prime has full power-Selmer corank zero or one, then the analytic rank and arithmetic rank both equal that corank.
- The entire Tate–Shafarevich group is finite.
- The leading Taylor coefficient obeys the full BSD formula, including every prime's contribution, not just its valuation at one selected auxiliary prime.
- No additional restriction is imposed on reduction type, rational torsion, rational isogenies, complex multiplication, or residual Galois representations.
With standard normalizations specified in its introduction, the predicted formula has the following shape:
Here the product is over finite primes, the period is taken with the manuscript's differential convention, and the regulator uses a basis of the full free Mordell–Weil group. Using a finite-index subgroup without correcting its squared index would give the wrong leading coefficient.
The manuscript presents the low-corank Selmer converse and a two-primary leading-term formula as companion inputs. It identifies the remaining new claim as the exact leading-term valuation at odd primes. The Family 002 index includes three manuscripts: this exact-formula paper; an all-prime Selmer converse; and a two-primary BSD formula. These are logically distinguishable: the converse establishes ranks and finiteness, the two-primary result addresses the prime two, and the main paper asserts the remaining complete equality.
The source overview also says that, together with Family 006 on quadratic twists, Family 002 would imply full BSD on a density-one subfamily. That is a combined conditional dependency statement about the claimed manuscripts, not a standalone theorem proved in this documentary article.
GCL verification state
The source-locked Family 002 inventory lists three papers with their October 2026 editions. This documentary pass compared the manuscript's public abstract and its available source introduction with the indexed headline, its rank/corank hypotheses, the normalizations of the formula, and the described division of work among the companion manuscripts.
A source-version warning matters here: the publisher's history describes an October revision that removed an obsolete supporting citation from the exact-BSD manuscript. Consequently, claims about dependencies must be checked against the exact pinned October 7 PDF and its bibliography; a prior or later repository edition is not automatically equivalent.
No GCL Lean declaration, Comparator challenge, formal axiom inventory, proof replay, or MATHCERT certificate was established as corresponding specifically to Family 002 in this pass. Corpus-level statistics do not justify claiming such a match. The discussion here is documentary, not source-semantic certification, a theorem check, or a new independent derivation.
Editorial gate: DRAFT_UNACCEPTED; NO_CERTIFICATION. Source-introduction review is not PDF fidelity review or mathematical adjudication.
What remains
The next source-critical reviewer should check the exact source-locked PDFs and their theorem numbers; verify the treatment of every prime, especially the distinction between two-primary and odd-prime arguments; and reconstruct the chain from the all-prime Selmer converse through the two-primary formula to the final leading coefficient.
In particular, review must test whether the specified Selmer corank is genuinely that of the full prime-power Selmer group, whether the Tate–Shafarevich finiteness conclusion is invoked without circularity, whether the Néron differential, periods, Tamagawa numbers and height regulator are normalized consistently, and whether any proof uses a withdrawn or superseded companion result.
A critic should deliver source-page and theorem-specific findings with a bounded ACCEPTED_EXPOSITION, REVISE or SOURCE_EXCEPTION disposition. Only after rendered Markdown/HTML review and protected admission may the article count as admitted documentary prose. None of those steps certifies the mathematical claim.
Sources and claim boundaries
- Exact Birch–Swinnerton-Dyer Formula from Low Selmer Corank — pinned PDF.
- The Selmer converse for elliptic curves at every prime — pinned PDF.
- The two-primary Birch–Swinnerton–Dyer formula in Selmer corank at most one — pinned PDF.
- OAI-26 overview, Family 002; source revision history is useful for investigating subsequent corrections but does not silently modify the pinned source.
- GCL editorial assignment OAI26-ED-01.
Evidence status: NO_CERTIFICATION. The complete claim is not certified by GCL or MATHCERT.
Claim boundary: This article summarizes a publisher-claimed full BSD theorem under a sharply stated low-Selmer-corank condition. It does not establish full BSD for unrestricted elliptic curves, independently validate the manuscripts, or upgrade any GCL certification status.
Indexed primary manuscript paths:
- Exact Birch–Swinnerton-Dyer Formula from Low Selmer Corank —
preprints/Exact-Birch-Swinnerton-Dyer-Formula-from-Low-Selmer-Corank-October-7-2026/exact-bsd-low-selmer-corank.pdf - The Selmer converse for elliptic curves at every prime —
preprints/The-Selmer-converse-for-elliptic-curves-at-every-prime-October-7-2026/main.pdf - The two-primary Birch–Swinnerton-Dyer formula in Selmer corank at most one —
preprints/The-two-primary-Birch-Swinnerton-Dyer-formula-in-Selmer-corank-at-most-one-October-6-2026/paper.pdf