Pith. sign in

REVIEW 4 minor 1 cited by

The density of elliptic curves over $\mathbb{Q}_p$ with a rational 3-torsion point or a rational 3-isogeny

T0 review · 0 major / 4 minor · reviewed 2026-08-08 · deepseek-v4-flash

Pith's one-line read For primes $p \ge 3$, the paper proves exact rational-function formulas for the Haar-measure density of Weierstrass equations over $\mathbb{Z}_p$ that define elliptic curves over $\mathbb{Q}_p$ with a rational 3-torsion point, and…

desk verdict Exact p-adic densities for 3-torsion and 3-isogeny, with two independent proofs in the p>3 case and one externally supported p=3 entry that a referee should check. read the letter →

arxiv 2502.08583 v2 pith:7Y3R37MP submitted 2025-02-12 math.NT

classification math.NT MSC 11G0714H5211S80
keywords ellipticcurves3-torsionpoints3-isogeniesp-adicfieldsHaarmeasuremodularX(3)andX_1(3)integrationKodairatypes
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

For primes $p \ge 3$, this paper computes exact probabilities for two arithmetic properties of a random elliptic curve over the $p$-adic numbers. Sampling long Weierstrass equations with coefficients in $\mathbb{Z}_p$ by Haar measure, the authors prove that the chance the curve has a nonzero $\mathbb{Q}_p$-rational 3-torsion point is the explicit rational function of $p$ in Theorem 1.1, and the chance it carries a $\mathbb{Q}_p$-rational cyclic 3-isogeny is the explicit rational function in Theorem 1.2 for $p > 3$. Which of the two formulas applies depends only on whether $p$ is $1$ or $2$ modulo $3$, with a separate value at $p = 3$. These are exact densities, not asymptotics, and they record the full system of congruences that detect or exclude rational 3-torsion. The paper also derives a large-sieve upper bound for counting rational 3-torsion curves over $\mathbb{Q}$ in boxes and an asymptotic for $\ell$-torsion points as $p$ tends to infinity.

What carries the argument

The central machinery is the upgrade of the classical level-3 modular parametrizations to spaces of Weierstrass equations equipped with a regular differential. Tate normal form $y^2 + uxy + vy = x^3$ parametrizes curves with a marked nonzero 3-torsion point, the Hesse pencil $u(x^3+y^3+z^3) - 3vxyz = 0$ parametrizes curves with full level-3 structure, and a twisted version with the point defined over a quadratic extension parametrizes the nontrivial characters of a rational 3-isogeny. Forgetting the level structure yields covering maps $\pi_1 : A^2_{X_1(3)} \to A^2_X$ and $\pi_2 : A^2_{X(3)} \to A^2_X$ whose Jacobian determinants are $256v^2$ and $-559872\,u^2(u-v)^2(u^2+uv+v^2)^2$, respectively. The $p$-adic change-of-variables formula turns the counting of integral preimages into integrals of the $p$-adic absolute values of these Jacobians, which are then evaluated by a residue-class calculation. Bad-reduction contributions are supplied by pre-existing Kodaira-type densities, and the case $p = 3$ uses congruences for the 3-division polynomial $\psi_3$ modulo higher powers together with the canonical lift.

What would settle it

Fix $p=5$, for which Theorem 1.1 predicts density $25/62$ and Theorem 1.2 predicts $401/781$, and compute by exhaustive enumeration the fraction of tuples in $(\mathbb{Z}/5^m\mathbb{Z})^5$ for increasing $m$ whose Weierstrass equation has a rational 3-torsion point, respectively a root of $\psi_3$ in $\mathbb{Q}_5$; convergence to the predicted rational numbers would support the theorems, and a definite mismatch at any single $m$ would refute them. Independently, recompute Proposition 4.3's $\mathbb{F}_p$-point counts on $X_1(3)$ and $X(3)$ for $p=5$ or $p=7$ by direct computer algebra to check the good-reduction step.

Watch

Extended reading notes

Core claim

The paper's central claim is that two local probabilities are exact rational functions of the prime $p$. For $p \ge 3$, the Haar measure of the set of tuples $[a_1,a_2,a_3,a_4,a_6]$ in $\mathbb{Z}_p^5$ defining an elliptic curve with a nonzero $\mathbb{Q}_p$-rational 3-torsion point is $\frac{p^2(3p^6+4p^2-4p+4)}{8(p^8+p^6+p^4+p^2+1)}$ when $p \equiv 1 \pmod 3$, $\frac{p^2}{2(p^2+p+1)}$ when $p \equiv 2 \pmod 3$, and $\frac{3}{26}$ when $p = 3$. For $p > 3$, the measure of the set defining a $\mathbb{Q}_p$-rational cyclic 3-isogeny is $\frac{3p^4+3p^3+4p^2+4}{4(p^4+p^3+p^2+p+1)}$ when $p \equiv 1 \pmod 3$ and $\frac{p^4+p^3+2p^2+2}{2(p^4+p^3+p^2+p+1)}$ when $p \equiv 2 \pmod 3$. The proof treats good and bad reduction separately: for good reduction, the density is obtained from $\mathbb{F}_p$-point counts on the modular curves $X_1(3)$ and $X(3)$; for bad reduction, it is obtained from the Kodaira type, using pre-existing densities for each type. A second proof via $p$-adic integration computes the same torsion density from the Jacobian determinants of the forgetful maps, and the case $p = 3$ requires a separate analysis of the 3-division polynomial and the canonical lift.

Load-bearing premise

The load-bearing premise is that the Kodaira-type densities quoted from the external local-density computation (for split multiplicative reduction of type $I_m$ and for additive types $IV$ and $IV^*$) are correct; the final formulas inherit these numbers, and the paper does not re-derive them.

Editorial extensions

If this is right

  • For $p \ge 3$, the complete system of congruences characterizing elliptic curves over $\mathbb{Q}_p$ with a rational 3-torsion point is known, and its total Haar volume is the explicit rational function in Theorem 1.1.
  • If the density theorems are correct, the large-sieve argument gives an upper bound $\#S \ll N_1N_2/\sqrt{\min\{N_1,N_2\}}$ for elliptic curves with rational 3-torsion in an arbitrary box $[M_1,M_1+N_1] \times [M_2,M_2+N_2]$, and a nontrivial saving for short intervals.
  • For a fixed ordinary elliptic curve $\bar{E}$ over $\mathbb{F}_p$ with an $\mathbb{F}_p$-rational $p$-torsion point, exactly one lift in $p$ of the coefficient pairs modulo $p^2$ gives a $\mathbb{Q}_p$-rational $p$-torsion point, so the lifting probability is $1/p$; these lifts are cut out by congruences modulo $p^2$ and coincide with the canonical lift.
  • For a prime $\ell > 3$ with $p \neq \ell$, the density of curves with a $\mathbb{Q}_p$-rational $\ell$-torsion point tends to $\frac{1}{\ell-1} - \frac{\delta}{\ell^2-1}$ as $p \to \infty$ through a fixed residue class modulo $\ell$, where $\delta = 1$ for $p \equiv 1 \pmod \ell$ and $\delta = 0$ otherwise; the paper also expresses the finite-$p$ density in terms of $\mathbb{F}_p$-point counts on

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: the same $p$-adic integration scheme should yield exact densities for $\ell = 2, 3, 4, 5$, where $X(\ell) \cong \mathbb{P}^1$, with formulas depending on $p$ through the residue class modulo $\ell$; the paper expects this but does not compute it.
  • Editorial inference: since the torsion density is assembled from local point counts on modular curves, one should expect the density of curves with a prescribed local Galois structure on $E[\ell]$ to be a rational number whenever the relevant moduli space is rational over $\mathbb{Q}_p$, and not otherwise; testing $\ell = 7$ would discriminate between these regimes.
  • Editorial inference: combining the exact $p$-adic densities with the large sieve suggests that the short-interval count of elliptic curves with a rational 3-torsion point satisfies the improved bound in Corollary 7.2(2), a regime where the trivial geometry-of-numbers estimate is worse and the local $p$-adic information carries the argument.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

0 major / 4 minor

Summary. This paper determines, for each prime p≥3, the Haar measure of the subset of Z_p^5 consisting of coefficients [a1,...,a6] for which the corresponding Weierstrass equation defines an elliptic curve over Q_p with a nonzero Q_p-rational 3-torsion point (Theorem 1.1), and for p>3 the analogous density for rational cyclic 3-isogenies (Theorem 1.2). The densities are explicit rational functions of p, depending only on p mod 3, with the exceptional value 3/26 at p=3. The proof strategy has three parts: (i) reduction to short Weierstrass form via measure-preserving changes of variables, (ii) counting F_p-points on modular curves X(3), X_1(3), and twisted versions, and (iii) a p-adic integration method based on Igusa's change-of-variables formula applied to the Jacobians of the covering maps. The paper also contains a self-contained treatment of p-torsion over Q_p (Proposition 6.13), an asymptotic for ℓ-torsion as p→∞, a large-sieve counting application for elliptic curves over Q with rational 3-torsion, and a family-of-twists analysis.

Significance. If correct, the two main theorems give exact unconditional local densities rather than asymptotics, and they are the first such results for torsion of order greater than 2 in this family. The paper is particularly strong in that Theorem 1.1 for p>3 is proved in two independent ways that agree: a reduction-type count (§4) and a p-adic integration computation (§5.2). The modular-curve parametrizations are explicit and the Jacobian determinants are stated in closed form, making the calculations reproducible. All numerical constants arise from a finite set of explicit inputs; no fitted parameters appear. The main external input is the set of Kodaira-type probabilities from Cremona–Sadek [11], which is used in the p=3 part of Theorem 1.1 and in the bad-reduction cases for p>3; the other parts of the argument are self-contained. The paper is clearly written and will be of interest to arithmetic statisticians and to researchers studying local-global questions for elliptic curves.

minor comments (4)
  1. [Section 6, first paragraph] The citation '[31, Theorem .6.4]' has a missing theorem number; it should refer to a valid statement in Silverman's book, for example the formal logarithm or the structure of E_1(Q_p).
  2. [Theorem 6.9] The typo 'Propostion' should read 'Proposition'. In the same theorem, the additive and split-multiplicative contributions are imported from [11, Proposition 2.2 and Proposition 2.5]; since the p=3 value 3/26 is the only part of Theorem 1.1 not corroborated by the independent p-adic integration proof, please add a sentence identifying the exact numerical densities taken from [11] (for Kodaira types IV, IV^*, and split I_m) so that this dependence can be checked without consulting the external paper.
  3. [Lemma 6.11] The phrase 'the smoothness of a map End(E_can) → End(E) is an isomorphism' is mangled; presumably the intended statement is that the reduction homomorphism on endomorphism rings is an isomorphism. Please rephrase.
  4. [Proposition 7.1 and Corollary 7.2] Proposition 7.1 assumes M_1 and M_2 are positive integers, but Corollary 7.2(1) applies it with M_1 = M_2 = 0; the hypotheses should be amended to allow nonnegative M_i.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the main densities are derived by independent modular-curve counting and p-adic integration, with the only self-citation being contextual and not load-bearing.

full rationale

The paper's central claims, Theorems 1.1 and 1.2, are derived through two independent routes: reduction-type counting using modular curves X1(3) and X(3), and p-adic integration via explicit bijections with Tate and Hesse normal forms. The p-adic integration proof does not import the conclusion of Theorem 1.1; instead it computes the same quantity from Jacobian determinants and Igusa's change-of-variables theorem, and Remark 5.11 notes agreement as a sanity check. No fitted parameters appear, and no quantity is defined in terms of the target density. The reliance on Cremona and Sadek [11] for Kodaira-type densities is an external, cited benchmark with no author overlap, and the paper explicitly states when it is used; this is ordinary mathematical citation, not circularity. The only self-citation is to [8] by Bhargava, Cremona, Fisher, and the first author, but it appears solely as context for related p-adic density results and is not load-bearing for the proofs here. The p=3 case uses [11] for IV and IV* densities, but those are independent external inputs, so any sensitivity there is a correctness/fragility concern, not circularity.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

No free parameters are fitted: all formulas are exact functions of p. The paper introduces no new entities. It does rely on several external pillars: Igusa's change-of-variables theorem, the Cremona-Sadek Kodaira statistics, the Tate/Hesse parametrizations of level-3 moduli spaces, and Chebotarev equidistribution for the asymptotic section.

assumptions (6)
  • standard math Igusa's p-adic change of variables formula (Theorem 5.1), including the Jacobian determinant rule for Haar measure.
    Used throughout Section 5 for the reduction from long Weierstrass to short and for the integrals defining the densities.
  • domain assumption Cremona-Sadek local density formulas for Kodaira types, especially split multiplicative type I_m density (p-1)^2/(2 p^{m+2}) and the densities of types IV and IV^* with Tamagawa number 3.
    Black-boxed in Propositions 4.4, 4.6, and Theorem 6.9; the final densities are linearly affected by these numbers.
  • domain assumption The bijections between A^2_{X1(3)}, A^2_{X(3)}, A^2_{Xpsi(3)}, A^2_{XF(3)} and moduli of (E,omega,P) from Tate and Hesse normal forms, valid over perfect fields with char not 2 or 3.
    Used in Section 3 and then Lemmas 5.5, 5.6, and 5.13 to transfer counting to Weierstrass coefficients.
  • standard math Hensel's lemma and Nagell-Lutz integrality for torsion points over Z_p.
    Used in Lemma 5.5 to show preimages under pi_1 are integral and in Section 6 for p-torsion lifting.
  • domain assumption Serre-Tate canonical lift and Serre's ramification criterion for ordinary elliptic curves (Theorem 6.12).
    Basis for Proposition 6.13 that random lifts have a Q_p-rational p-torsion point with probability 1/p; not needed for the main 3-torsion or 3-isogeny formulas.
  • standard math Geometric Chebotarev equidistribution of Frobenius conjugacy classes in GL2(F_l) as p varies.
    Used in Section 7.2 to derive the asymptotic probability for l-torsion; not used in Theorems 1.1 and 1.2.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The density of elliptic curves over $\mathbb{Q}_p$ with a rational 3-torsion point or a rational 3-isogeny." pith.science (2026). https://pith.science/paper/7Y3R37MP

@misc{pith2026250208583,
  author       = {Pith},
  title        = {Pith review of: The density of elliptic curves over $\mathbbQ_p$ with a rational 3-torsion point or a rational 3-isogeny},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7Y3R37MP}},
  note         = {Machine review of arXiv:2502.08583}
}
abstract

We determine the probability that a random Weierstrass equation with coefficients in the $p$-adic integers defines an elliptic curve with a non-trivial $3$-torsion point, or with a degree $3$ isogeny, defined over the field of $p$-adic numbers. We determine these densities by calculating the corresponding $p$-adic volume integrals and analyzing certain modular curves. Additionally, we explore the case of $\ell$-torsion for $\ell>3$ prime.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Counting odd genus $2$ curves with a marked rational $3$-torsion point

    math.NT 2026-07 accept novelty 7.0 of 10

    The number of minimal monic odd-degree genus-2 Weierstrass models of height ≤ X with a marked rational Jacobian 3-torsion point is c X^10 + o(X^10) for an effectively computable constant c.

Reference graph

Works this paper leans on

35 extracted references · 32 canonical work pages · cited by 1 Pith paper

  1. [11]

    Local and global densities for Weierstrass models of elliptic curves

    John Cremona and Mohammad Sadek. Local and global densities for Weierstrass models of elliptic curves. Math. Res. Lett. , 30(2):413–461, 2023

  2. [1]

    Jacobians of genus one curves

    Sang Yook An, Seog Young Kim, David C Marshall, Susan H Marshall, W illiam G McCallum, and Alexander R Perlis. Jacobians of genus one curves. Journal of Number Theory , 90(2):304–315, 2001

  3. [2]

    The Hesse pencil of plane cubic curves

    Michela Artebani and Igor V Dolgachev. The Hesse pencil of plane cubic curves. L’Enseignement Math´ ematique, 55(3):235–273, 2009

  4. [3]

    Shadow line distributions

    Jennifer S Balakrishnan, Mirela C ¸ iperiani, Barry Mazur, and Karl Rubin. Shadow line distributions. arXiv preprint arXiv:2409.00891 , 2024

  5. [4]

    How often does a cubic hypersurface have a rational point?

    Lea Beneish and Christopher Keyes. How often does a cubic hype rsurface have a rational point? arXiv preprint arXiv:2405.06584, 2024

  6. [5]

    Higher composition laws IV: The parametrizat ion of quintic rings

    Manjul Bhargava. Higher composition laws IV: The parametrizat ion of quintic rings. Annals of Math- ematics, 167:53–94, 2008

  7. [6]

    The proport ion of plane cubic curves over Q that everywhere locally have a point

    Manjul Bhargava, John Cremona, and Tom Fisher. The proport ion of plane cubic curves over Q that everywhere locally have a point. International Journal of Number Theory , 12(04):1077–1092, 2016. 31

  8. [7]

    The proport ion of genus one curves over Q defined by a binary quartic that everywhere locally have a point

    Manjul Bhargava, John Cremona, and Tom Fisher. The proport ion of genus one curves over Q defined by a binary quartic that everywhere locally have a point. International Journal of Number Theory , 17(04):903–923, 2021

Show all 35 references
  1. [8]

    The density of polynomials of degree n over Zp having exactly r roots in Qp

    Manjul Bhargava, John Cremona, Tom Fisher, and Stevan Gajo vi´ c. The density of polynomials of degree n over Zp having exactly r roots in Qp. Proc. Lond. Math. Soc. (3) , 124(5):713–736, 2022

  2. [9]

    The Magma alg ebra system

    Wieb Bosma, John Cannon, and Catherine Playoust. The Magma alg ebra system. I. The user language. J. Symbolic Comput., 24(3-4):235–265, 1997. Computational algebra and number theory (London, 1993)

  3. [10]

    Counting elliptic curves with prescrib ed level structures over number fields

    Peter Bruin and Filip Najman. Counting elliptic curves with prescrib ed level structures over number fields. Journal of the London Mathematical Society , 105(4):2415–2435, 2022

  4. [12]

    On a probabilistic local-global principle for torsion on elliptic curves

    John Cullinan, Meagan Kenney, and John Voight. On a probabilistic local-global principle for torsion on elliptic curves. Journal de th´ eorie des nombres de Bordeaux, 34(1):41–90, 2022

  5. [13]

    Report on Igusa’s local zeta function

    Jan Denef. Report on Igusa’s local zeta function. Ast´ erisque, 201:203, 1990

  6. [14]

    Die Typen der Multiplikatorenringe elliptischer Funkt ionenk¨ orper

    Max Deuring. Die Typen der Multiplikatorenringe elliptischer Funkt ionenk¨ orper. In Abhandlungen aus dem mathematischen Seminar der Universit¨ at Hamburg , volume 14 (1), pages 197–272. Springer Berlin/Heidelberg, 1941

  7. [15]

    The invariants of a genus one curve

    Tom Fisher. The invariants of a genus one curve. Proceedings of the London Mathematical Society , 97(3):753–782, 2008

  8. [16]

    The hessian of a genus one curve

    Tom Fisher. The hessian of a genus one curve. Proceedings of the London Mathematical Society , 104(3):613–648, 2012

  9. [17]

    A formula for the Jacobian of a genus one curve of a rbitrary degree

    Tom Fisher. A formula for the Jacobian of a genus one curve of a rbitrary degree. Algebra & Number Theory, 12:2123–2150, 2018

  10. [18]

    Everywhere local solub ility for hypersurfaces in products of projective spaces

    Tom Fisher, Wei Ho, and Jennifer Park. Everywhere local solub ility for hypersurfaces in products of projective spaces. Research in Number Theory , 7:1–27, 2021

  11. [19]

    Counting elliptic curves wit h prescribed torsion

    Robert Harron and Andrew Snowden. Counting elliptic curves wit h prescribed torsion. Journal f¨ ur die reine und angewandte Mathematik (Crelles Journal) , 2017(729):151–170, 2017

  12. [20]

    Algebraic geometry, volume 52

    Robin Hartshorne. Algebraic geometry, volume 52. Springer Science & Business Media, 2013

  13. [21]

    An introduction to the theory of local zeta functions , volume 14

    Jun-ichi Igusa. An introduction to the theory of local zeta functions , volume 14. American Mathematical Soc., 2000

  14. [22]

    Algebraic geometry and arithmetic curves , volume 6

    Qing Liu. Algebraic geometry and arithmetic curves , volume 6. Oxford Graduate Texts in Mathe, 2002

  15. [23]

    Elliptic cur ves and formal groups

    Jonathan Lubin, Jean-Pierre Serre, and John Tate. Elliptic cur ves and formal groups. In Lecture notes prepared in connection with the seminars held at the Summer I nstitute on Algebraic Geometry, Whitney Estate, Woods Hole, Massachusetts , volume 6, 1964

  16. [24]

    Multiplicative functions inshort intervals

    Kaisa Matom¨ aki and Maksym Radziwi/suppress l/suppress l. Multiplicative functions inshort intervals. Annals of Math- ematics, pages 1015–1056, 2016

  17. [25]

    A note on elliptic curves with ordinary reduct ion

    Tetsuo Nakamura. A note on elliptic curves with ordinary reduct ion. Archiv der Mathematik , 60:440– 445, 1993

  18. [26]

    On p-torsion of p-adic elliptic curves with additive reduction

    Ren´ e Pannekoek. On p-torsion of p-adic elliptic curves with additive reduction. arXiv preprint arXiv:1211.5833, 2012. 32

  19. [27]

    Counting elliptic curves with an isogeny of degree three

    Maggie Pizzo, Carl Pomerance, and John Voight. Counting elliptic curves with an isogeny of degree three. Proc. Amer. Math. Soc. Ser. B , 7:28–42, 2020

  20. [28]

    Families of elliptic curves with constant modp representations

    Karl Rubin and Alice Silverberg. Families of elliptic curves with constant modp representations. Elliptic curves, modular forms, & Fermat’s last theorem (Hong Kong, 1 993), 1:148–161, 1995

  21. [29]

    Une formule de masse pour les extensions t otalement ramifiees de degres donne d’un corps local

    Jean-Pierre Serre. Une formule de masse pour les extensions t otalement ramifiees de degres donne d’un corps local. CR Acad. Sc. Paris A , 286:1031–1036, 1978

  22. [30]

    Families of elliptic curves ordered by conductor

    Ananth N Shankar, Arul Shankar, and Xiaoheng Wang. Families of elliptic curves ordered by conductor. arXiv preprint arXiv:1904.13063 , 2019

  23. [31]

    The arithmetic of elliptic curves , volume 106

    Joseph H Silverman. The arithmetic of elliptic curves , volume 106. Springer Science & Business Media, 2009

  24. [32]

    The discrete logarithm problem on elliptic curves of trace one

    Nigel P Smart. The discrete logarithm problem on elliptic curves of trace one. J. Cryptology, 12(3):193– 196, 1999

  25. [33]

    Algorithm for determining the type of a singular fiber in an elliptic pencil

    John Tate. Algorithm for determining the type of a singular fiber in an elliptic pencil. In Modular Functions of One Variable IV: Proceedings of the Internatio nal Summer School, University of Antwerp, RUCA, July 17–August 3, 1972 , pages 33–52. Springer, 2006

  26. [34]

    The greatest common valuatio n of φn and ψ2 n at points on elliptic curves

    Paul Voutier and Minoru Yabuta. The greatest common valuatio n of φn and ψ2 n at points on elliptic curves. Journal of Number Theory , 229:16–38, 2021

  27. [35]

    Local solubility of a family of ternary conics over a biprojective base I

    Cameron Wilson. Local solubility of a family of ternary conics over a biprojective base I. arXiv preprint arXiv:2409.10688, 2024. Stevan Gajovi´ c, MPIM Bonn; Charles University Prague, Facultyof Mathematics and Physics, Department of Algebra; Mathematical Institute MISANU E-m...

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.