Papex

Open problem · Mathematics · since 1965 · Open

Birch and Swinnerton-Dyer conjecture

Does an elliptic curve’s L-function tell how many rational points it has?

Summary

Relates the number of rational points on an elliptic curve to the behaviour of its associated L-function at s = 1.

AI-assisted claims

Partial result claimed · 06 Oct 2026

Exact Birch–Swinnerton-Dyer Formula from Low Selmer Corank

“We prove the full Birch–Swinnerton-Dyer leading-term formula for every elliptic curve over ℚ whose full q-power Selmer group has corank zero or one at some prime q. The analytic and Mordell–Weil ranks equal that corank, and the Tate–Shafarevich group is finite.”

By
OpenAI
AI involvement
This repository contains mathematical manuscripts and supporting proof artifacts produced by an internal OpenAI model.
Model
an unreleased internal OpenAI model, as stated
Where
openai/math repository
Source last read
14 Oct 2026

Checks

1 response

  • 08 Oct 2026 · Response · Frank Calegari: “There are some results proving BSD under certain rank conditions. My first impression is that these are technical improvements on results already in the literature, and that they represent no genuine progress towards the higher-rank version of BSD.” Source

Claims as their sources state them. Listing a claim is not a check by Papex; the status stays the source’s.

All AI-assisted claims

Why it is open

An elliptic curve over the rationals can be written y² = x³ + ax + b with integers a and b [5]. Its rational points form a group, and Louis Mordell proved in 1922 that this group consists of a finite part and r independent points of infinite order; r is the rank, and the curve has only finitely many rational points exactly when r = 0 [5].

The L-function is an infinite product built from the number of solutions of the curve's equation modulo each prime p; since every elliptic curve over the rationals is now known to be modular, it extends to the whole complex plane [5, 6, 7]. The conjecture, as the Clay Mathematics Institute states it, says that its order of vanishing at s = 1, the analytic rank, equals the rank r, so that it vanishes at s = 1 exactly when there are infinitely many rational points [5, 7]. A refined form predicts the leading coefficient from the order of the Tate–Shafarevich group Ш, the regulator (a determinant built from the heights of independent rational points), a period, local factors and the number of points of finite order [5, 7].

Ш measures a failure of a local-to-global principle: Ernst Selmer's curve 3x³ + 4y³ + 5z³ = 0 has real and p-adic solutions for every p but no rational one, which gives a non-trivial element of Ш for the elliptic curve x³ + y³ + 60z³ = 0 [7]. Ш is conjectured to be finite, but this is not known in general [5, 7].

A resolution would be a proof for every elliptic curve over the rationals, or a curve on which the predicted equality fails. For the Clay Mathematics Institute's prize, a solution must appear in a qualifying outlet, at least two years must pass after publication, and it must gain general acceptance in the mathematics community [15]. Rational points are hard to produce: in Andrew Wiles's description, "there are still almost no methods for finding such points" on varieties in general [5].

What is known

Origins (1960s). Bryan Birch and Peter Swinnerton-Dyer found the conjecture experimentally in the early 1960s, using one of the early EDSAC computers at Cambridge [5]. Their computations, on curves from the congruent-number problem, suggested that the product, over primes p up to X, of the number of points modulo p divided by p grows like a constant times (log X)ʳ, where r is the rank; they published the conjecture in 1965 [1, 5, 7].

Ranks zero and one (1976–2001). John Coates and Wiles proved in 1976 that a curve with complex multiplication whose L-function does not vanish at s = 1 has only finitely many rational points [2, 5]. In 1983 Benedict Gross and Don Zagier showed that for a modular curve whose L-function vanishes at s = 1 but has non-zero derivative there, a construction of Kurt Heegner gives a rational point of infinite order [3, 5]. Victor Kolyvagin then showed that, for modular curves, non-vanishing at s = 1 gives rank 0 and a simple zero gives rank 1, with Ш finite in both cases [4, 5, 7]. Since 2001 every elliptic curve over the rationals is known to be modular [6], so the rank statement holds for all curves of analytic rank 0 or 1 [5, 7].

The formula, one prime at a time (to 2017). The refined formula is checked one prime p at a time, by comparing the powers of p on both sides (its "p-part"). Karl Rubin proved the formula for curves with complex multiplication of analytic rank 0, up to primes dividing the number of units of the associated imaginary quadratic field [7]. For other curves, theorems of Kazuya Kato and of Christopher Skinner and Eric Urban give the p-part in analytic rank 0 for primes p ≥ 3 of good ordinary reduction, under conditions on how the Galois group acts on points of order p [7, 8]; Wei Zhang proved an analogue in analytic rank 1 for p ≥ 5 [7]. In 2017 Dimitar Jetchev, Skinner and Xin Wan proved the p-part in analytic rank 1 for semistable curves at every prime p ≥ 5 of good reduction where the points of order p form an irreducible Galois representation [12]. Converse theorems start from the p-power Selmer group, which combines the rational points with the p-part of Ш: its corank is at least the rank, with equality exactly when the p-part of Ш is finite, and whether it is 0 or 1 is, in Zhang's words, "usually easy to check" [7]. Under conditions like those above, Selmer corank 0 or 1 forces analytic rank 0 or 1 [7, 8]. Zhang's survey motivates the Selmer group by an analogy with the Hodge conjecture [7].

Averages (2010–2015). Ordering curves by height, Manjul Bhargava and Arul Shankar proved that the average rank is bounded (its limit superior is at most 1.5) [10], then that it is less than 1.17, that a positive proportion of curves have rank 0 and, combined with Skinner and Urban's work, that a positive proportion satisfy the conjecture [11]. A 2014 preprint by Bhargava, Skinner and Zhang states that more than 66% of curves satisfy the rank part [9].

Recent claims. On 6 October 2026 OpenAI opened the public repository github.com/openai/math of "mathematical manuscripts and supporting proof artifacts produced by an internal OpenAI model", each cited there as an "OpenAI Math Release preprint" [13]. Its manuscript "Exact Birch–Swinnerton-Dyer Formula from Low Selmer Corank" (dated 3 October 2026, revised 7 October 2026) claims the full formula, the finiteness of Ш and the equality of rank and analytic rank for every elliptic curve over the rationals whose q-power Selmer group has corank 0 or 1 for some prime q: a special case that, by the claim itself, covers only curves of rank 0 or 1 [13]. It takes its rank and finiteness statements and its 2-part from two companion manuscripts in the same release, and the repository's catalogue adds that, with a result on Goldfeld's conjecture, it gives the full conjecture for a density-one set of quadratic twists of every elliptic curve over the rationals [13]. The history file describes the 7 October revision as removing "an obsolete introductory citation to a removed supporting manuscript"; on 14 October 2026 the Lean formalisation catalogue listed no formal proof of these results [13]. On 8 October 2026 Frank Calegari wrote on his blog Persiflage: "There are some results proving BSD under certain rank conditions. My first impression is that these are technical improvements on results already in the literature, and that they represent no genuine progress towards the higher-rank version of BSD" [14]. On 14 October 2026 the Clay Mathematics Institute still listed the conjecture as unsolved [15].

Main approaches

Heegner points and Euler systems. Build a rational point from the curve's modular parametrisation, relate its height to the derivative of an L-function of the curve over an imaginary quadratic field (the Gross–Zagier formula), and use Kolyvagin's classes, derived from Heegner points over larger fields, to control the Selmer group and Ш [3, 4, 7]. Strongest evidence: the rank part and finiteness of Ш in analytic rank at most 1 [7], and the p-part of the formula in analytic rank 1 [7, 12]. Main difficulty: the method rests on one point coming from the first derivative [5], and in Zhang's 2013 survey the analytic-rank-one theorem was still "by far the most general result" on the rank part [7].

Iwasawa theory. Study the curve over a tower of number fields, the cyclotomic Z_p-extension of the rationals, and prove the Iwasawa–Greenberg main conjecture, which equates a p-adic L-function with the characteristic ideal of a Selmer group over that tower [8]. Strongest evidence: the p-part of the formula in analytic rank 0, and the converse from Selmer corank 0 [7, 8]. Main difficulty: the theorems need conditions on the prime, its reduction type and the Galois action, and Zhang wrote that "It seems difficult to obtain the same result for small primes p, especially p = 2" [7].

Counting curves. Count integral orbits of binary quartic and ternary cubic forms to find the average size of Selmer groups, then combine with the converse and main-conjecture results [9, 10, 11]. Strongest evidence: bounded average rank and a positive proportion of curves satisfying the conjecture [10, 11], over 66% for the rank part in the 2014 preprint [9]. Main difficulty: these are statements about proportions of curves ordered by height, and the proportions proved so far fall short of all curves [9, 11].

Open sub-questions

  1. Rank two. Results on the rank part stop at analytic rank 1 [5, 7]. For one explicit curve of analytic rank 2, can one prove that its rank is 2, or that its Ш is finite?
  2. Small primes. Ye Tian, with Xinyi Yuan and Shou-Wu Zhang, proved the 2-part of the formula for many quadratic twists of the congruent-number curve y² = x³ − n²x [7]. Can the 2-part or the 3-part be proved for another explicit family of twists? (The 2026 manuscripts under Recent claims state results of this kind [13].)
  3. The rank-one converse. Zhang's 2013 survey proves that rank 1 with Ш finite implies analytic rank 1 when at least two distinct primes divide the conductor exactly once [7]. Can that condition on the conductor be removed?
  4. Beyond two thirds. Can the proportion of curves, ordered by height, proved to satisfy the rank part be raised above the 66% of Bhargava, Skinner and Zhang [9]?

Work on this in Papex

If you have a testable answer to one of the sub-questions above, you can draft it as a private hypothesis in Papex, or prepare the draft from Claude or ChatGPT with the Papex connector. You review the exact version, sign it and decide whether to publish it; a public record that names this problem is listed on this page. A Papex record shows who signed which version and when. It does not check a proof, and it is not a submission for the Clay Mathematics Institute prize.

Key references

  1. Birch, B. J. & Swinnerton-Dyer, H. P. F. (1965). Notes on elliptic curves. II. Journal für die reine und angewandte Mathematik 218, 79–108 (online 2009). 10.1515/crll.1965.218.79 · Checked 14 Oct 2026
  2. Coates, J. & Wiles, A. (1977). On the conjecture of Birch and Swinnerton-Dyer. Inventiones Mathematicae 39(3), 223–251. 10.1007/BF01402975 · Checked 14 Oct 2026
  3. Gross, B. H. & Zagier, D. B. (1986). Heegner points and derivatives of L-series. Inventiones Mathematicae 84(2), 225–320. 10.1007/BF01388809 · Checked 14 Oct 2026
  4. Kolyvagin, V. A. (1989). Finiteness of E(Q) and Ш(E, Q) for a subclass of Weil curves. Mathematics of the USSR-Izvestiya 32(3), 523–541. 10.1070/IM1989v032n03ABEH000779 · Checked 14 Oct 2026
  5. Wiles, A. The Birch and Swinnerton-Dyer conjecture. Official problem description, Clay Mathematics Institute (prepared for the Millennium Prize; its most recent reference is from 2001). Source · Checked 14 Oct 2026
  6. Breuil, C., Conrad, B., Diamond, F. & Taylor, R. (2001). On the modularity of elliptic curves over Q: wild 3-adic exercises. Journal of the American Mathematical Society 14(4), 843–939. 10.1090/S0894-0347-01-00370-8 · Checked 14 Oct 2026
  7. Zhang, W. (2013). The Birch–Swinnerton-Dyer conjecture and Heegner points: a survey. Current Developments in Mathematics 2013, 169–203. 10.4310/CDM.2013.v2013.n1.a3 · Checked 14 Oct 2026
  8. Skinner, C. & Urban, E. (2014). The Iwasawa main conjectures for GL2. Inventiones Mathematicae 195(1), 1–277 (published online 2013). 10.1007/s00222-013-0448-1 · Checked 14 Oct 2026
  9. Bhargava, M., Skinner, C. & Zhang, W. (2014). A majority of elliptic curves over Q satisfy the Birch and Swinnerton-Dyer conjecture. arXiv:1407.1826 (version 2, 17 July 2014; preprint, no journal reference listed on 2026-10-14). Source · Checked 14 Oct 2026
  10. Bhargava, M. & Shankar, A. (2015). Binary quartic forms having bounded invariants, and the boundedness of the average rank of elliptic curves. Annals of Mathematics 181(1), 191–242. 10.4007/annals.2015.181.1.3 · Checked 14 Oct 2026
  11. Bhargava, M. & Shankar, A. (2015). Ternary cubic forms having bounded invariants, and the existence of a positive proportion of elliptic curves having rank 0. Annals of Mathematics 181(2), 587–621. 10.4007/annals.2015.181.2.4 · Checked 14 Oct 2026
  12. Jetchev, D., Skinner, C. & Wan, X. (2017). The Birch and Swinnerton-Dyer formula for elliptic curves of analytic rank one. Cambridge Journal of Mathematics 5(3), 369–434. 10.4310/cjm.2017.v5.n3.a2 · Checked 14 Oct 2026
  13. OpenAI (2026). Exact Birch–Swinnerton-Dyer formula from low Selmer corank. OpenAI Math Release preprint, editions dated 3 and 7 October 2026, in the repository github.com/openai/math (opened 6 October 2026), with the repository's README.md, CONTENTS.md, history.md and Lean formalisation catalogue (lean/formalization.yaml). Source · Checked 14 Oct 2026
  14. Calegari, F. (2026). A brief digest of some new OpenAI results. Persiflage (blog), 8 October 2026. Source · Checked 14 Oct 2026
  15. Clay Mathematics Institute. Birch and Swinnerton-Dyer Conjecture (problem page, status "Unsolved") and Rules for the Millennium Prize Problems. Source · Checked 14 Oct 2026

Page reviewed 14 Oct 2026

Source

Open, as listed by Clay Mathematics Institute · Millennium Prize Problems. Checked 29 Sep 2026.

The status is the source’s. Papex does not decide whether a problem is solved.

Prize

Clay Mathematics Institute · $1,000,000

Prize rules

Only Clay Mathematics Institute awards this prize, under its own rules. Your Papex record is separate: a signed, dated record of your work, not a prize entry.

Records on this problem

No public record addresses this problem yet.

More in Mathematics