{"id":"a335d567-6fb3-41d7-ab15-a1ea31cc5096","arxiv_id":"1908.07569","paper_version":2,"verdict":"UNVERDICTED","confidence":"HIGH","novelty_score":2.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A pedagogical proof, following Serre, that a finite union of real intervals with logarithmic capacity greater than 1 contains infinitely many algebraic integers with all conjugates in it.","lead":"This paper explains when a compact region of the complex plane contains finitely or infinitely many algebraic integers, whose conjugate numbers all stay inside the region. It is an expository treatment of classical theorems in potential theory and algebraic curves, aimed at advanced undergraduates.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 4.7's sketched proof asserts a continuous map U→Q^g; since U is connected and Q^g is totally disconnected, such a map is constant, so the density argument as written is invalid.","rationale":"In good faith, this paper is a collective expository project, not a research announcement. Its capacity theory sections are largely standard and mostly self-contained, and the algorithmic section is honestly compared with the state of the art. The main thing that must be true for the paper's stated purpose is that it gives a usable proof of Robinson's theorem, Theorem 4.3. That proof splits into a Pell-Abel construction (§4.2.4) and a reduction from the Pell-Abel case to the general union-of-intervals case (§4.2.5). The reduction depends on Theorem 4.7, the density of U_PA in U. I agree with the reader that this is the weakest load-bearing point, and I found a sharper concrete problem: the sketched proof asserts a continuous map from the connected open set U into Q^g, which is impossible unless constant. Even a charitable reading cannot make the written claim true; one would need a local lift to R^g plus a separate density argument. The theorem is attributed to Serre [2], so the paper's scientific conclusion may still be correct, but the text's own proof of the reduction is not. This does not change the overall verdict: the paper remains an expository review whose most delicate step is sourced to a citation, so UNVERDICTED remains the appropriate classification rather than ACCEPT or REJECT. If the authors intend the paper to stand as a self-contained proof, they should replace the invalid sentence with a correct argument or explicitly state that Theorem 4.7 is assumed from Serre.","tokens_in":50492,"tokens_out":11814,"duration_ms":612048,"concrete_test":"Test the Abel-Jacobi lift for g=1. Parametrize U by real elliptic curves y^2=(x-a0)(x-b0)(x-a1)(x-b1), and compute ν(u)=∫_{∞-}^{∞+} dx/y in R/Z. Show there exist u with ν(u) of finite order and u with ν(u) not of finite order; since U is connected, no continuous θ:U→Q can satisfy ν(u)=θ(u) mod Z for all u. This falsifies the sketched proof. Then check whether replacing the target by R^g and proving density of the preimage of Q^g via dominance or torsion-density in the Jacobian recovers Theorem 4.7, or whether the paper must explicitly rely on Serre's proof.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that must hold is the proof of Robinson's theorem (Theorem 4.3). The last reduction in §4.2.5 requires Theorem 4.7, namely that U_PA is dense in U. The sketch says: \"we can lift ν: U→R^g/Z^g in order to obtain a continuous function θ: U→Q^g, and ν(u) is a point of finite multiplicity in R^g/Z^g if and only if θ(u)∈Q^g.\" This cannot be right as stated: U is connected and Q^g is totally disconnected, so every continuous θ:U→Q^g is constant. Hence the asserted lift does not exist unless the image is a single rational point, in which case the finiteness condition would be trivial and false for many u. At best one might obtain a locally defined continuous lift into R^g and then need a separate argument that the preimage of Q^g is dense. The paper gives no such argument. Because Theorem 4.7 is load-bearing — without it the Pell-Abel case does not imply the general union-of-intervals case — the proof of the reduction is not merely deferred but internally unsupported at this step. The theorem is cited from Serre [2], so the underlying mathematics may still be true and recoverable, but the text's own justification is not valid.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","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.","tokens_in":50726,"tokens_out":11726,"duration_ms":113514,"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":[{"comment":"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.","section":"§4.2.5, Theorem 4.7"},{"comment":"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.","section":"§2.4, Theorem 2.10 proof"}],"minor_comments":[{"comment":"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.","section":"§3.1, Fekete's theorem proof"},{"comment":"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).","section":"§4.2.4, Proposition 4.13"},{"comment":"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.","section":"§2.4, Theorem 2.10 statement"},{"comment":"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.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"This is an expository collective-project paper rather than an original research contribution. Its main value is pedagogical, and the novelty is limited. The most serious issue for a referee is the invalid proof sketch of Theorem 4.7; the theorem itself is true and is cited from Serre, so the paper can be made reliable by replacing the sketch with a proper black-box citation. The paper would then be suitable for an expository or education-oriented venue, though its novelty may be too limited for a standard research journal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read this as an undergraduate collective project, not a research paper. It is an exposition of capacity theory and Robinson's theorem following Serre, and the best parts are genuinely good: the proof of Fekete's theorem, the Hilbert lemniscate construction, the Chebyshev constant equivalence, and the Pell-Abel machinery in Section 4.2.4 are all laid out carefully. The only original contribution is a degree-7 enumeration algorithm, which is weaker than the degree-13 results in the cited reference [3]. So there is no new mathematics here, and the authors do not claim otherwise.\n\nThe real soft spot is the final reduction in Section 4.2.5. Theorem 4.7, the density of U_PA in U, is load-bearing: without it the Pell-Abel case does not imply Robinson for arbitrary unions of intervals. The stress-test note is right about what is wrong. The text says one can lift nu: U -> R^g/Z^g to a continuous function theta: U -> Q^g and then use the density of Q^g. As written that cannot work: U is connected and Q^g is totally disconnected, so such a continuous map would be constant. The intended argument probably involves a lift into R^g and a separate density statement for torsion points, and the theorem itself is cited from Serre [2], so the underlying fact is almost certainly true and recoverable. But the paper's own justification is not valid at this step. For a paper that promises a proof, this is not a minor typo; it should be fixed or explicitly deferred to Serre.\n\nThere are also minor typographical issues: in the Fekete theorem proof the product should carry absolute values and a square factor, and in Theorem 2.10 two displayed inequalities have the wrong direction, though the intended equalities are clear.\n\nWho gets value from this paper? Undergraduates with a solid background in analysis and algebra who want a guided tour through capacity theory and the Pell-Abel route to Robinson's theorem. Sections 1 through 3 and most of Section 4 are useful for that audience. The defect in Section 4.2.5 means the final step is essentially imported from Serre rather than proved here.\n\nIf this were submitted to a research venue I would desk reject it. If submitted to an expository journal, it could be refereed after the authors fix the Theorem 4.7 paragraph or explicitly mark it as a citation to Serre. As it stands, I would not cite it in my own work.","headline":"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.","tokens_in":51296,"tokens_out":2262,"would_cite":false,"duration_ms":163623,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11R04","31A15","11G30","11G20"],"pacs":[],"model":"deepseek-v4-flash","headline":"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.","keywords":["algebraic integers","logarithmic capacity","transfinite diameter","Chebyshev polynomials","Pell-Abel equation","hyperelliptic curves","Robinson theorem","equidistribution"],"falsifier":"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.","tokens_in":50295,"feed_emoji":"📏","tokens_out":14098,"duration_ms":136383,"temperature":0.7,"pith_summary":"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.","feed_headline":"Capacity above 1: infinite algebraic integers inside real intervals","feed_subtitle":"For a single interval, length 4 is the dividing line; shorter is finite, longer infinite, the boundary open.","key_machinery":"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.","core_discovery":"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.","pith_inferences":["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."],"forward_implications":["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."],"supporting_citations":[{"why":"supplies the Bourbaki-seminar proof of Robinson's theorem that this paper expands, including the density step and equidistribution theorem.","marker":"[2]"},{"why":"source of the Fekete–Szegö theorem used to construct Hilbert lemniscates and handle compact sets of capacity at least 1.","marker":"[11]"},{"why":"reference for strict convexity of the logarithmic energy and uniqueness of the equilibrium measure.","marker":"[4]"},{"why":"reference for compact Riemann surfaces, period matrices, and Jacobians used in the hyperelliptic-curve argument.","marker":"[6]"},{"why":"source for the generalized minimum principle and backward-pulling of harmonic functions used in capacity estimates and regularity.","marker":"[7]"},{"why":"source for the equioscillation theorem used to identify P as the Chebyshev polynomial of E.","marker":"[5]"}],"fun_headline_variants":["Capacity >1: infinitely many algebraic integers","Interval length 4: threshold for infinite algebraic integers","Set capacity decides finiteness of algebraic integers","For finite interval unions, capacity 1 is the breakpoint","Algebraic integers infinite if set capacity >1"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"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.","fun_headline_variants_meta":{"raw":{"variants":["Capacity >1: infinitely many algebraic integers","Interval length 4: threshold for infinite algebraic integers","Set capacity decides finiteness of algebraic integers","For finite interval unions, capacity 1 is the breakpoint","Algebraic integers infinite if set capacity >1"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00048,"raw_usage":{"total_tokens":2472,"prompt_tokens":1140,"completion_tokens":1332,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":756,"completion_tokens_details":{"reasoning_tokens":1259}},"tokens_in":756,"tokens_out":1332,"duration_ms":13308,"temperature":1.0,"reasoning_tokens":1259,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:03:59.004591+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"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.","supporting_citations":[{"cited_title":"Serre : Distribution asymptotique des valeurs propres des endomorphismes de Frobenius [d’après Abel, Chebyshev, Robinson, ...].Séminaire Bourbaki, n◦1146, 2018","cited_arxiv_id":null,"evidence_quote":"supplies the Bourbaki-seminar proof of Robinson's theorem that this paper expands, including the density step and equidistribution theorem."},{"cited_title":"Fekete et G","cited_arxiv_id":null,"evidence_quote":"source of the Fekete–Szegö theorem used to construct Hilbert lemniscates and handle compact sets of capacity at least 1."},{"cited_title":"Springer Science & Business Media, 2013","cited_arxiv_id":null,"evidence_quote":"reference for strict convexity of the logarithmic energy and uniqueness of the equilibrium measure."},{"cited_title":"Springer Berlin Heidelberg, Berlin, Heidelberg, 1992","cited_arxiv_id":null,"evidence_quote":"reference for compact Riemann surfaces, period matrices, and Jacobians used in the hyperelliptic-curve argument."},{"cited_title":"Ransford : Potential theory in the complex plane, volume 28","cited_arxiv_id":null,"evidence_quote":"source for the generalized minimum principle and backward-pulling of harmonic functions used in capacity estimates and regularity."},{"cited_title":"The Journal of Online Mathematics and Its Applications, 6, 2006","cited_arxiv_id":null,"evidence_quote":"source for the equioscillation theorem used to identify P as the Chebyshev polynomial of E."}],"review_version":1}