Pith. sign in

REVIEW 2 major objections 4 minor 15 references

On algebraic integers all conjugates of which belong to a given compact subset of the complex plane

T0 review · 2 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read For finite unions of real intervals, logarithmic capacity above 1 guarantees infinitely many algebraic integers whose conjugates all stay in the set, and capacity below 1 guarantees only finitely many.

desk verdict A clearly written undergraduate exposition of capacity theory and Robinson's theorem, with no new results and one load-bearing reduction whose proof sketch is invalid as written. read the letter →

arxiv 1908.07569 v2 pith:7CSUEHPV submitted 2019-08-20 math.NT math.AGmath.CV

Marc Fersztand , Benjamin Gourevitch , Arnaud Rippol , Rubing Shen , Ian-Christopher Tanoh , Yicheng Zhou This is my paper · ORCID
classification math.NTmath.AGmath.CV MSC 11R0431A1511G3011G20
keywords algebraicintegerslogarithmiccapacitytransfinitediameterChebyshevpolynomialsPell-AbelequationhyperellipticcurvesRobinsontheoremequidistribution
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

The paper establishes a sharp split at capacity 1 for real intervals: if a finite union $E$ of intervals has logarithmic capacity greater than 1, then infinitely many algebraic integers exist whose conjugates all lie in $E$, at arbitrarily large degree; if capacity is below 1, only finitely many exist. For one interval this says length greater than 4 gives infinitely many such numbers and length less than 4 gives finitely many; the length-4 boundary is open except when the endpoints are integers. The proof uses potential theory, identifying capacity with the transfinite diameter and the Chebyshev constant, then uses hyperelliptic curves $y^2=D(x)$ and the Pell–Abel equation $P^2-DQ^2=M^2$ to produce Chebyshev-like polynomials that are perturbed into monic integer polynomials with all roots in $E$. One density statement, passed over with a sketch, extends the Pell–Abel case to arbitrary unions of intervals. Written for undergraduates, the paper makes a modern Bourbaki-seminar argument accessible.

What carries the argument

The load-bearing objects are the logarithmic capacity $\operatorname{Cap}(E)$—equivalently the transfinite diameter and the Chebyshev constant—and the hyperelliptic curve $C:y^2=D(x)$ with $D(X)=\prod_{j=0}^g (X-a_j)(X-b_j)$. On $C$, the Pell–Abel equation $P^2-DQ^2=M^2$ is equivalent to the divisor $r((\infty^-)-(\infty^+))$ being principal. The argument uses the unique differential form of the third kind $\eta=R(x)\,dx/y$ whose periods over the gaps $[b_{j-1},a_j]$ vanish; the numbers $r_j=r|\eta_j|/\pi$ control how many roots of $P$ lie in each interval $E_j$, so $P$ equioscillates on $E$. Finally, the perturbation lemma $|q|<M$ on $E$ preserves the $r$ roots of $P-q$ and transfers the construction to monic integer polynomials.

What would settle it

One could test the density step directly in a one-parameter family of interval configurations by computing, on the associated Jacobian variety, whether the torsion points that make $r((\infty^-)-(\infty^+))$ principal occur in every open set of parameters; an open region entirely free of such points would falsify Theorem 4.7 and break the reduction from the Pell–Abel case to the general case.

Watch

Extended reading notes

Core claim

The central claim, Theorem 4.3, states that every finite union $E$ of real intervals with $\operatorname{Cap}(E)>1$ contains infinitely many algebraic integers totally in $E$. The proof first assumes a Pell–Abel solution $P^2-DQ^2=M^2$ attached to the polynomial $D(X)=\prod_{j=0}^g (X-a_j)(X-b_j)$; in that case $P$ is the Chebyshev polynomial of $E$, $\operatorname{Cap}(E)=(M/2)^{1/r}$, and the roots of $P$ and $Q$ interlace with $r_j$ roots of $P$ inside each interval $E_j$. Writing $f=P+yQ$, the paper shows $df/f=r\eta$ for the canonical differential form of the third kind, whose periods compute $r_j=r|\eta_j|/\pi$. Raising $f$ to high powers and subtracting a small rational perturbation yields monic integer polynomials with arbitrarily large degree and all roots in $E$, proving the theorem in the solvable case. A density argument then extends the conclusion to all finite unions of intervals with capacity above 1.

Load-bearing premise

The load-bearing assumption is that for purposes of approximation, every union of intervals can be nudged so that the associated Pell–Abel equation $P^2-DQ^2=M^2$ has real polynomial solutions, and the paper gives only a sketch of this density fact, referring to the Bourbaki seminar for details.

Editorial extensions

If this is right

  • For a single interval, lengths below 4 yield only finitely many algebraic integers totally in it, lengths above 4 yield infinitely many, and the equality case length 4 is settled only for integer endpoints.
  • In the Pell–Abel case the proof is constructive: it produces explicit monic integer polynomials of unbounded degree whose roots all lie in $E$.
  • The equidistribution upgrade, Theorem 4.6, says the roots of the constructed polynomials converge in the weak-* sense to the equilibrium measure of $E$, partially answering the distribution question raised by Frobenius eigenvalues.
  • Combined with Fekete's theorem, the result leaves capacity exactly 1 as the only unresolved threshold for unions of real intervals.

Reading between the lines

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

  • The period formula $r_j=r|\eta_j|/\pi$ could be turned into an effective algorithm: compute the periods of the canonical differential form numerically and read off how many roots of the eventual integer polynomial lie in each interval, avoiding exhaustive coefficient search.
  • The length-4 case is plausibly sensitive to arithmetic: for integer endpoints the paper exhibits infinitely many numbers of the form $n+2\cos(2k\pi/m)$, while for non-integer endpoints no analogous construction exists, suggesting a rational-versus-irrational divide.
  • The open question in the conclusion suggests a stronger equidistribution conjecture: counting measures of conjugates of any infinite sequence of distinct algebraic integers totally in $E$ may converge to the equilibrium measure of $E$, not just for the constructed family.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 4 minor

Summary. The paper is an expository study, aimed at undergraduates, of algebraic integers all of whose conjugates lie in a fixed compact set K⊂C. It develops logarithmic capacity via three equivalent definitions (transfinite diameter, potential-theoretic capacity, Chebyshev constant), proves Fekete's finiteness theorem and the Fekete-Szegö theorem, and then presents a proof of Robinson's theorem: a finite union E of real intervals with Cap(E)>1 contains infinitely many algebraic integers totally in E. The proof follows Serre's Bourbaki seminar: the Pell-Abel case P^2-DQ^2=M^2 is worked out in detail, and the general case is reduced to it by a density statement about configurations for which the Pell-Abel equation has a solution. The paper also includes algorithms, numerical experiments, and appendices on measure theory and potential theory.

Significance. The paper is not a research article in the usual sense: the main theorems are classical (Fekete, Fekete-Szegö, Robinson) and the deepest input, the density of Pell-Abel configurations, is taken from Serre's seminar. Its value is pedagogical: it gives a largely self-contained development of capacity theory, a detailed treatment of the Pell-Abel construction, and a clear statement of the reduction. The Pell-Abel part of the Robinson proof is essentially sound, and the expository appendices are useful. However, the final reduction in §4.2.5 contains an invalid proof sketch of the needed density theorem, so the paper as submitted does not give a complete proof of Robinson's theorem; the gap is fixable by importing Theorem 4.7 as a black box from Serre.

major comments (2)
  1. [§4.2.5, Theorem 4.7] The reduction from the general union-of-intervals case to the Pell-Abel case rests on the density of U_PA in U. The proof sketch given in the text is not valid as written: it asserts that one can lift ν: U→R^g/Z^g to a continuous function θ: U→Q^g, and that ν(u) is a torsion point iff θ(u)∈Q^g. But U is connected and Q^g is totally disconnected, so every continuous θ:U→Q^g is constant; the stated equivalence would force all ν(u) to coincide, which is not the case. No alternative argument is supplied to show that the preimage of Q^g is dense. Since this density statement is load-bearing for the proof of Robinson's theorem, the manuscript either needs a correct proof of Theorem 4.7 or should explicitly import it as a black box from Serre [2] with a precise reference, deleting the invalid 'main steps' paragraph.
  2. [§2.4, Theorem 2.10 proof] In the first half of the proof, after establishing ||F_n||_K^{1/n} ≤ δ_{n+1}(K) and t_n(K) ≤ ||F_n||_K^{1/n}, the displayed chain 'τ(K) > lim sup ||F_n||^{1/n} > lim inf ||F_n||^{1/n} > Cheb(K)' has the inequalities reversed: the correct conclusion is Cheb(K) ≤ lim inf ||F_n||^{1/n} ≤ lim sup ||F_n||^{1/n} ≤ τ(K). The later argument proving the reverse inequality makes the theorem true, but the displayed chain as written is false and should be corrected.
minor comments (4)
  1. [§3.1, Fekete's theorem proof] In the displayed lower bound, the exponent on δ_{d_n}(K) should be d_n(d_n-1)/2, not d_n(d_n-1), and the right-hand side should be written with absolute values (the product over ordered pairs is the square of the Vandermonde product). The intended argument goes through after this correction.
  2. [§4.2.4, Proposition 4.13] The notation 'M̃∈[0,M]\Q' conflicts with the later instruction to choose M̃ rational; it should be M̃∈[0,M]∩Q (with 0<M̃<M for Lemma 4.1 to apply).
  3. [§2.4, Theorem 2.10 statement] The displayed limit F_n(z)^{1/n} → exp(-U_{μK}(z)) is not well-posed for complex F_n(z); the proof actually establishes uniform convergence of (1/n)log|F_n(z)| (equivalently |F_n(z)|^{1/n}) on compact subsets of C\K. Please restate accordingly, since the later use in Theorem 3.2 is the logarithmic form.
  4. [Throughout] There are numerous small typographical slips (for example, the repeated use of √(-D(x)) versus i√(-D(x)) and some indexing inconsistencies in the interval endpoints); these do not affect the mathematics but should be cleaned up.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper is an expository derivation from capacity theory and independent results of Serre; no fitted parameter is renamed as a prediction and no self-citation is load-bearing.

full rationale

The derivation chain is: capacity theory (Fekete, Fekete-Szegö) -> Pell-Abel equation via hyperelliptic curves -> construction of monic integer polynomials whose roots lie in E -> reduction from the Pell-Abel case to the general case via density of U_PA. No term is defined in terms of the conclusion, no parameter is fitted to data and then called a prediction, and no self-citation is used as evidence. The Pell-Abel polynomials P,Q are constructed from the curve y^2 = D(x) and the proof explicitly produces an infinite family P_n - q_n with integer coefficients and all roots in E; this is a construction, not a fit. The reduction from the general case to the Pell-Abel case invokes Theorem 4.7, which is cited from Serre's Bourbaki article [2], an external source not authored by the paper's authors. The equidistribution statement in Theorem 4.6 is likewise referred to Serre [2]. These are independent supporting references, so by the hard rules they do not raise the circularity score. A possible topological objection to the sketch in §4.2.5 (a continuous map into Q^g would be constant) identifies a gap in the sketch, but it is a correctness concern rather than circularity: the density theorem is cited from external work, not derived from the paper's own target result. Overall the burden of circularity is effectively zero.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper relies on standard ZFC mathematics and on classical potential theory results. The two load-bearing black boxes are Theorem 4.7 (density of Pell-Abel configurations) and Theorem 4.6 (equidistribution), both credited to Serre [2] and not fully proved in this paper. No free parameters or invented entities appear.

assumptions (4)
  • standard math ZFC set theory, complex analysis, measure theory, and standard facts about Riemann surfaces (residue theorem, genus, period matrix invertibility).
    The paper builds on these foundations throughout, e.g., Theorem 4.2 is cited from [6].
  • domain assumption Existence and uniqueness of the equilibrium measure and its characterization via Frostman's theorem (Theorems 2.2 and 2.5).
    These are standard in logarithmic potential theory; the paper proves some parts and cites [4,7] for others.
  • domain assumption Theorem 4.7: U_PA is dense in U (the reduction from the general case to the Pell-Abel case).
    The paper gives only a sketch in Section 4.2.5 and refers to Serre [2]. This is load-bearing for Robinson's theorem.
  • domain assumption Theorem 4.6: the constructed polynomials' root counting measures converge to the equilibrium measure.
    Stated in Section 4.2.4 with the proof deferred to Serre [2]; not needed for Robinson's theorem but presented as a strengthening.

how reviews work

0 comments
Cite this review

Pith. "Pith review of On algebraic integers all conjugates of which belong to a given compact subset of the complex plane." pith.science (2026). https://pith.science/paper/7CSUEHPV

@misc{pith2026190807569,
  author       = {Pith},
  title        = {Pith review of: On algebraic integers all conjugates of which belong to a given compact subset of the complex plane},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7CSUEHPV}},
  note         = {Machine review of arXiv:1908.07569}
}
abstract

The study of Frobenius endomorphism provides numerous information about its corresponding Abelian variety. To understand the action of the Frobenius endomorphism, one may be interested in its eigenvalues. According to Weil's third conjecture ("Riemann hypothesis over finite fields"), they all have absolute value less than or equal to $2g\sqrt{p}$. Thus, the eigenvalues of the Frobenius endomorphism all belong to the same compact subset of the complex plane, and are roots of the same monic polynomial with integer coefficients (the characteristic polynomial of the Frobenius endomorphism). Such complex numbers are called algebraic integers "totally" in a compact subset, which means algebraic integers all conjugates of which belong to a same given compact subset of the complex plane. The study of such algebraic integers helps to understand the eigenvalues of the Frobenius endomorphism, especially their distribution. In this paper, we will study the following question : under which conditions a compact subset of the complex plane has a finite or infinite number of algebraic integers "totally" in it ? The problem can be studied in light of the notion of capacity of a compact subset, which comes from potential theory. In this paper, we will present the theory of capacity and some theorems (Fekete, Szeg\"o, Robinson) derived from it that partially answer the question: in the case of a union of real segments, when the capacity is smaller (resp. larger) than 1, it contains a finite (resp. infinite) number of algebraic integers totally in it. For instance, for real line segments, the limit length is 4. This paper is written as part of a collective project conducted in \'Ecole Polytechnique (France). It is aimed towards undergraduate audience in mathematics, with basic knowledge in algebra, topology, analysis, and dwells into a modern topic of research.

Figures

Figures reproduced from arXiv: 1908.07569 by the authors.

Figure 2
Figure 2. Number of new algebraic integers in function of the [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Number of new algebraic integers found in function of the degree of the polynomials traversed [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Distribution of algebraic integers of lower degree for different segments compared to that of the roots [PITH_FULL_IMAGE:figures/full_fig_p010_4.png] view at source ↗
Figures from the paper (9 more)
Figure 5
Figure 5. Figure 5: examples of points totally in K = [−2, 2] • if sin(θ) = 0, then z ∈ R. In order to verify z + 1 z ∈ [−2, 2], studying the sign of this expression shows us that z = ±1. • if ρ = 1 ρ , ρ = 1, since the modulus is a non-negative number. In all cases, z ∈ U. It follows fro…
Figure 6
Figure 6. Figure 6: Fekete points for n ∈ {8, 16} and equilibrium measure with K = U 23 [PITH_FULL_IMAGE:figures/full_fig_p023_6.png]
Figure 7
Figure 7. Figure 7: Frostman’s theorem U µK (z0) ≥ VK. By using the lower semi-continuity of U µK , there exists a closed ball denoted by B with radius r > 0 and center z0, on which we have U µK (z) > VK − 1/2n. Then, we have B ∩ Kn = ∅, and µK(B) = a > 0 because z0 is in the support of µ…
Figure 8
Figure 8. Figure 8: Case g = 2: C as a Riemann surface (above) or a curve on the affine plane (bellow) [PITH_FULL_IMAGE:figures/full_fig_p050_8.png]
Figure 9
Figure 9. Figure 9: Case g = 2: a symplectic basis of H1(C; Z) . The genus g, which intuitively corresponds to the number of holes in the surface, also corresponds to the dimension of the space of holomorphic forms. Proposition 4.3 The complex dimension of the space of holomorphic 1-forms…
Figure 10
Figure 10. Figure 10: Shape of P, case g = 1, r1 = 4, r2 = 6 This proposition is the very key to the proof of Robinson’s theorem. Because of its importance, it is very enlightening to prove it in the case g = 0 (in other words, E = [a, b] is a single segment). It allows to show the major s…
Figure 11
Figure 11. Figure 11: Shape of P, in the case g = 0, r = 8 Proof. A representative of f = P + yQ in E is f(x) = P(x) + iy1(x)Q(x), with y1(x) = p −D(x) since D(x) 6 0 for all x ∈ E. The Pell-Abel equation implies that |f(x)| 2 = P(x) 2 − D(x)Q(x) 2 = M2 , therefore f has a constant magnitu…
Figure 12
Figure 12. Figure 12: Case where g = 2: Sg j=0 Γj is the boundary of the medial hemi-surface. Proof. Let us denote by Γj the cycles which cover the segments [aj , bj ] (cf [PITH_FULL_IMAGE:figures/full_fig_p056_12.png]
Figure 13
Figure 13. Figure 13: Shape of P in the case g = 1, r0 = 4, r1 = 6 Proof. The beginning of the proof is identical to the case where g = 0. Let us consider j ∈ {0, . . . , g} and the segment Ej = [aj , bj ]. A representative of f on Ej is f(x) = P(x) + i p −D(x)Q(x). We have |f| = M constan…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

15 extracted references · 15 canonical work pages

  1. [3]

    Moscow Journal of Combinatorics and Number Theory, 1, 01 2011

    Valérie Flammang, Georges Rhin et Qiang wu : The totally real algebraic integers with diameter less than 4. Moscow Journal of Combinatorics and Number Theory, 1, 01 2011

    work page 2011

  2. [2]

    Serre : Distribution asymptotique des valeurs propres des endomorphismes de Frobenius [d’après Abel, Chebyshev, Robinson, ...].Séminaire Bourbaki, n◦1146, 2018

    J.-P. Serre : Distribution asymptotique des valeurs propres des endomorphismes de Frobenius [d’après Abel, Chebyshev, Robinson, ...].Séminaire Bourbaki, n◦1146, 2018

    work page 2018

  3. [1]

    Pierre Deligne : La conjecture de weil : I.Publications Mathématiques de l’IHÉS, 43:273–307, 1974

    work page 1974

  4. [4]

    Springer Science & Business Media, 2013

    Edward B Saff et Vilmos Totik : Logarithmic potentials with external fields, volume 316. Springer Science & Business Media, 2013

    work page 2013

  5. [5]

    The Journal of Online Mathematics and Its Applications, 6, 2006

    Robert Mayans : The chebyshev equioscillation theorem. The Journal of Online Mathematics and Its Applications, 6, 2006

    work page 2006

  6. [6]

    Springer Berlin Heidelberg, Berlin, Heidelberg, 1992

    Jean-Benoît Bost : Introduction to Compact Riemann Surfaces, Jacobians, and Abelian Varieties, pages 64–211. Springer Berlin Heidelberg, Berlin, Heidelberg, 1992

    work page 1992

  7. [7]

    Ransford : Potential theory in the complex plane, volume 28

    T. Ransford : Potential theory in the complex plane, volume 28. Cambridge University Press, 1995

    work page 1995

  8. [8]

    E. B. Saff : Logarithmic potential theory with applications to approximation theory.Surveys in Approx- imation Theory, 5 (2010), 165–200

    work page 2010

Show all 15 references
  1. [9]

    I.Burgos Gil et R.Menares : Equidistribución, teoría del potencial y aplicaciones aritméticas

    J. I.Burgos Gil et R.Menares : Equidistribución, teoría del potencial y aplicaciones aritméticas. Notes de cours

  2. [10]

    Tsuji : Potential theory in modern function theory, volume 4

    M. Tsuji : Potential theory in modern function theory, volume 4. Chelsea Publ. Co., New York, 1975

    1975

  3. [11]

    Fekete et G

    M. Fekete et G. Szegö : On algebraic equations with integral coefficients whose roots belong to a given point set. Mathematische Zeitschrift, 63(1):158–172, 1955

    1955

  4. [12]

    Ransford : Computation of logarithmic capacity

    T. Ransford : Computation of logarithmic capacity. Computational Methods and Function Theory, 10(2):555–578, 2011

    2011

  5. [13]

    Rémond : Conjectures uniformes sur les variétés abéliennes.Quart.,J

    G. Rémond : Conjectures uniformes sur les variétés abéliennes.Quart.,J. Math.(to appear), 2017

    2017

  6. [14]

    In Mémoires de la S

    Bertrand et W aldschmidt: Quelques travaux récents en théorie des nombres transcendants. In Mémoires de la S. M. F. 2e série, volume 2, pages 107–119. 1980

    1980

  7. [15]

    In Cahiers du séminaire d’histoire des mathématiques 2e série, volume 3, pages 79–84

    Marc Hindry : Géométrie arithmétique. In Cahiers du séminaire d’histoire des mathématiques 2e série, volume 3, pages 79–84. 1993. 76

    1993

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.