{"id":"86fa095e-6aee-41be-b19d-4b57ea8012d5","arxiv_id":"2411.17523","paper_version":4,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":1.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A survey of methods proving partition regularity for pairs of variables in homogeneous quadratic equations, with the remaining cases reduced to an open conjecture about vanishing correlations.","lead":"This paper surveys recent progress on the question of whether every finite coloring of the integers contains a monochromatic solution to equations like x^2 + y^2 = z^2, for pairs of variables. It explains the main proof ideas from multiplicative functions, ergodic theory, and the Q-trick, and lists the open problems that block full treatment of triples.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Missing complex conjugate in the Bochner-Herglotz representation (Eq. 16) invalidates the paper's reduction of density regularity to multiplicative-function averages; all subsequent unbarred product formulas need re-derivation.","rationale":"The reader's weakest assumption, Conjecture 3, governs only conditional Theorem 2.2 and the hardest remaining pairs. A more fundamental problem sits earlier in the unconditional route: the Bochner-Herglotz representation in §3.3.2 is stated without the necessary complex conjugation. On (Q_+,×), characters satisfy χ(r/s)=χ(r)\\overline{χ(s)}, so equation (16) should read ∫ f(r)\\overline{f(s)} dσ. With the printed f(r)f(s), even the diagonal case r=s=2 would force A(1)=∫ f(2)^2 dσ to be real and nonnegative for all measures σ arising from densities; this fails for any σ charging non-real-valued multiplicative functions, which the method explicitly needs (Archimedean characters n^{it} appear throughout Section 6). The defect propagates: Theorem 3.2's integrand should be f(P1)\\overline{f(P2)}, and condition (17) should be Hermitian. The later Q-trick exposes the inconsistency—B(f) in §6.1.2 contains (2mn)^{-it} while L(f,Q) in (65) contains f(2(Qm+1)Qn) unbarred. Because this reduction is the bridge from Definition 2.3 to the analytic positivity statements, the survey's proof of the unconditional Theorem 2.1 is not valid as written. This is an expositional/correctness issue, not a claim about the published theorems; it can likely be repaired by inserting conjugates, which is why I recommend CONDITIONAL rather than REJECT. A direct check against [22,23] or the point-mass calculation with f(2)=i would settle it.","tokens_in":86,"tokens_out":15021,"duration_ms":199166,"concrete_test":"Re-derive Eq. (16) from the standard Bochner-Herglotz theorem on Q_+: for every f∈M, f(r/s)=f(r)\\overline{f(s)}. Then check [22, §2] and [23, §3] for whether the actual statements of Theorems 3.2/3.3 use f(P1)\\overline{f(P2)}. Decisive numerical check: set σ=δ_f with f(2)=i (extended completely multiplicatively); the printed (16) would give A(1)=i^2=-1, contradicting A(1)=dΦ'(Λ)≥0. If the published papers use the barred version, the survey can be fixed; if not, Theorem 2.1's claimed reduction is unsupported.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Section 3.3.2, equation (16) claims dΦ'(r^{-1}Λ∩s^{-1}Λ)=∫_M f(r)f(s)dσ(f). This is not the Bochner-Herglotz representation on the multiplicative group Q_+. For a character f∈M, one has f(r/s)=f(r)f(s)^{-1}=f(r)\\overline{f(s)}, so the integrand must be f(r)\\overline{f(s)}. With the printed unbarred product, taking r=s=2 gives A(1)=∫ f(2)^2 dσ; for non-real-valued f this need not be real nonnegative, while A(1)=dΦ'(Λ)≥0. Thus (16) cannot hold for the measures arising from densities. The same missing conjugate propagates into the hypotheses of Theorems 3.2/3.3: condition (17) should read ∫ f(r)\\overline{f(s)}dσ≥0, and the target averages should contain f(P1)\\overline{f(P2)} rather than f(P1)f(P2). This is not cosmetic: the later Q-trick positivity (e.g., §6.1.2) explicitly uses (2mn)^{-it} in B(f), i.e., a conjugate factor, while the displayed L(f,Q) in (65) uses f(2(Qm+1)Qn); the two are consistent only if the representation carries a conjugate. Since the reduction to multiplicative functions is the bridge from Definition 2.3 to Theorems 3.2/3.3, the survey's account of the proof of Theorem 2.1 is internally inconsistent as written.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This survey explains recent progress on partition regularity of homogeneous quadratic equations ax^2 + by^2 = cz^2, with emphasis on partition regularity with respect to pairs of variables. The central results are Theorem 2.1, asserting partition regularity with respect to (x,y) when ac or bc is a square, and the conditional Theorem 2.2, covering the case where (a+b)c is a square under Conjecture 3. The paper develops a systematic strategy: reformulate partition regularity as density regularity for parametrized pairs, pass via a Bochner-Herglotz representation on the multiplicative group to averages of completely multiplicative functions, split multiplicative functions into aperiodic and pretentious cases, and finally combine the two regimes through the Q-trick. It also presents applications to level sets of multiplicative functions, comparisons with earlier work, and a list of open problems, including vanishing of correlations along two irreducible quadratic forms and ergodic-theoretic formulations.","tokens_in":34932,"tokens_out":16352,"duration_ms":150738,"significance":"If the exposition is made correct, the paper is a valuable roadmap to a body of recent research by the author with Host, Klurman, and Moreira. Its conceptual decomposition—Bochner-Herglotz representation, Gowers uniformity of aperiodic multiplicative functions, concentration estimates for pretentious functions, and the Q-trick—is genuinely useful and appears to capture the structure of the existing proofs. The survey is also honest about its limitations: the main results are cited rather than proved, the conditional Theorem 2.2 is explicitly made dependent on Conjecture 3, and the open problems are stated precisely. For a survey, the level of transparency about the dependence on external papers and the clear separation of unconditional and conditional statements are strengths. The principal weakness is a systematic conjugation error in the foundational representation step, together with a missing conjugation in the definition of the Gowers norms; these are fixable but affect the advertised reduction from density regularity to multiplicative-function averages.","major_comments":[{"comment":"The Bochner-Herglotz representation on (Q_+, ×) is stated with f(r)f(s) instead of f(r)\\overline{f(s)}. For a character f in M, f(r/s) = f(r) f(s)^{-1} = f(r)\\overline{f(s)}, so the printed identity cannot hold. Taking r = s = 2 gives dΦ'(Λ) = ∫ f(2)^2 dσ(f), an expression that need not be real or nonnegative, while the left side is a density. Consequently condition (17) should be ∫ f(r)\\overline{f(s)} dσ ≥ 0, and the integrands in (18), (20), (65), (75), and (79) should contain f(P_1)\\overline{f(P_2)} rather than the unbarred product. The use of (2mn)^{-it} in §6.1.2 is consistent only with the conjugated representation, so the survey's account of the reduction from Definition 2.3 to Theorems 3.2 and 3.3 is internally inconsistent as written. This is not a purely cosmetic typo; the entire chain of displayed formulas leading to the Q-trick needs to be re-derived with the conjugate factor carried through.","section":"§3.3.2, Eq. (16); Theorems 3.2–3.3"},{"comment":"The inductive definition of the U^{s+1}(Z_N) norm omits a complex conjugation: the average should involve a · \\overline{a_h}, not a · a_h. As printed, the displayed formula for \\|a\\|_{U^2}^4 is not the standard U^2 norm for complex-valued sequences, and the equivalence (25) is not correct without the conjugate. Since Theorem 4.6 and the correlation-vanishing statements are expressed in terms of these norms, the definition should be corrected and any later use checked against the corrected definition.","section":"§4.2, Definition 4.2"},{"comment":"The Q-trick for the pair (m^2 + 2n^2, m^2 - 2n^2) is stated too tersely. In Case 2, the claimed equality G_{1,N}(f,K) = G_{2,N}(f,K) does not follow from the displayed condition (89) alone, because both G's contain contributions from primes p ≡ 1 mod 8, on which (89) imposes no condition; the expression in (88) also has P_j in both factors, which should be P_1 and P_2. Please add the missing condition on p ≡ 1 mod 8 or indicate explicitly where in [23] this case is proved.","section":"§6.5.1, Eqs. (88)–(91)"}],"minor_comments":[{"comment":"The summand in the definition of G_{d,N}(f,K) contains n^{-it}, but n is not defined there; it should be p^{-it}.","section":"§5.6, Eq. (59)"},{"comment":"The claim that the displayed double integral has positive real part for every t is justified by an unspecified computer calculation; please provide a closed form or a reference so that the claim is verifiable without external software.","section":"§6.1.2, Endgame I"},{"comment":"There are small typographical errors: 'Heglotz' should be 'Herglotz', the van der Waerden reference has 'Bewis' for 'Beweis', and Section 6.5.1 contains 'mutliplicative' for 'multiplicative'.","section":"§3.1 and references"},{"comment":"The character denoted χ3 in the sentence following (89) should presumably be χ2.","section":"§6.5.1, after Eq. (89)"}],"recommendation":"major_revision","confidential_remarks":"The paper is a survey of the author's own recent results, and the level of self-citation is appropriate for the topic. The editor may wish to verify that the main cited sources [21], [22], and [23] are publicly available and that the statements attributed to them match the published or preprint versions. The primary concern is the conjugation error in Eqs. (16)–(17) and its propagation into the core reduction; this is fixable but requires a careful pass through the entire chain of displayed formulas."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is a survey, so it has no new theorems, but it does something almost as useful for the field: it gives a coherent, readable account of the recent proof that Pythagorean pairs are partition regular and of the broader method covering generalized homogeneous quadratics. The bird's-eye view in Section 3.2 and the detailed but accessible explanation of the Q-trick in Section 6 are genuinely good. I also credit the author for being explicit about what is conditional: Theorem 2.2 depends on Conjecture 3, and the survey states plainly that the needed two-irreducible-quadratics vanishing is unknown even for Liouville. That honesty is a real strength.\n\nNow the soft spots. The stress-test note is correct. Equation (16) is not the Bochner-Herglotz representation on the multiplicative group Q_+. For f in the dual, f(r/s) = f(r) f(s)^{-1} = f(r)\\overline{f(s)}, so the integrand must carry a conjugate. The printed unbarred product cannot represent a positive definite function; taking r=s gives a nonnegative real number as an integral of f(r)^2, which is impossible for non-real-valued f. This typo propagates into (17), (18), (20), (65), and the aperiodic-correlation statements. That said, the later appearance of (2mn)^{-it} in Section 6.1.2 shows the author knows the correct conjugate factor, so this is a systematic notational error in the exposition, not an error in the underlying published results. It still needs a global fix: a reader trying to follow the proof sketch cannot reconstruct the reduction as written. There is also an obvious typo in (88), where P_j appears in both factors.\n\nThe heavy self-citation is not a problem for a survey of the author's own recent work; the cited results are published and the sketches are traceable. The conditional results are clearly labeled. The main mathematical content, as far as I can tell, is sound.\n\nBottom line: this is a useful survey by a principal contributor, and the flaws are fixable exposition errors rather than load-bearing mathematical ones. With the conjugate corrected throughout, it will be a good entry point to the area. I would send it to peer review.","headline":"A solid, honest survey of a real recent advance, worth publishing after a global fix of a missing complex conjugate that as written breaks the key reduction.","tokens_in":35485,"tokens_out":4118,"would_cite":true,"duration_ms":53052,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["05D10","11N37","11B30","37A44"],"pacs":[],"model":"deepseek-v4-flash","headline":"Homogeneous quadratic equations $ax^2+by^2=cz^2$ are pairwise partition regular when $ac$ or $bc$ is a square, and the remaining Rado-triple cases reduce to one conjecture.","keywords":["partition regularity","Pythagorean triples","homogeneous quadratic equations","multiplicative functions","Gowers uniformity","concentration inequalities","Q-trick","arithmetic Ramsey theory"],"falsifier":"Compute, or bound away from zero, the Cesàro average over $m,n\\le N$ of $\\lambda(m^2+2n^2)\\lambda(m^2-2n^2)$, where $\\lambda$ is the completely multiplicative sign function with $\\lambda(p)=-1$ at every prime, and let $N\\to\\infty$. A non-zero limit would disprove Conjecture 3 and destroy conditional Theorem 2.2; a proof of vanishing for even one explicit pair of irreducible quadratics would be the first confirmatory evidence for the conjecture in the genuinely irreducible setting.","tokens_in":34388,"feed_emoji":"🔢","tokens_out":12159,"duration_ms":105963,"temperature":0.7,"pith_summary":"Finitely color the positive integers. This article surveys a method that proves, for a wide family of homogeneous quadratic equations in three variables, that two of the three variables can always be found in the same color class. The flagship result is that the Pythagorean equation $x^2+y^2=z^2$ is partition regular with respect to every pair of variables, and more generally $ax^2+by^2=cz^2$ is partition regular with respect to $x,y$ whenever $ac$ or $bc$ is a square. The remaining pair cases, in which only $(a+b)c$ is a square, are proved conditional on a single conjecture about vanishing of correlations of aperiodic multiplicative functions along two irreducible binary quadratic forms. The article presents this as the first systematic route to a genuinely nonlinear homogeneity problem that was previously out of reach, with the heavy lifting split between Gowers-uniformity estimates and concentration estimates.","feed_headline":"Pythagorean pairs are monochromatic in every finite coloring","feed_subtitle":"Pairwise solutions of the Pythagorean equation always share a cell; the method extends to many quadratics.","key_machinery":"The machinery has four parts. The Bochner–Herglotz representation on the multiplicative group $(\\mathbb{Q}_+,\\times)$ expresses positive multiplicative densities as integrals over the compact space of completely multiplicative functions $f\\colon\\mathbb{N}\\to S^1$, turning density regularity into a positivity property for averages such as $\\liminf_{N\\to\\infty} E_{m,n\\in[N]} \\int f(P_1(m,n))f(P_2(m,n))\\,d\\sigma(f)$. For aperiodic multiplicative functions, meaning those with zero correlation with every Dirichlet character, Gowers-uniformity results of all orders give vanishing of the relevant correlations, so only the pretentious part of $\\sigma$ matters. For pretentious functions, which are close to $\\chi(n)n^{it}$ in the pretentious distance, concentration estimates of Turán–Kubilius type show that after restricting to a highly divisible lattice $Qn+1$ the function concentrates on a tractable oscillatory factor. The Q-trick then averages over $Q$ in a multiplicative Følner set; all non-trivial Archimedean characters wash out and only the atom $\\sigma(\\{1\\})>0$ survives, yielding the required positivity.","core_discovery":"On its own terms, the paper claims that the pair version of the long-standing Pythagorean-triples problem is now within reach of a unified method. Theorem 1.3 states that the Pythagorean equation is partition regular with respect to all pairs of variables; Theorem 2.1 extends this to all non-zero integers $a,b,c$ with $ac$ or $bc$ a square, for the pair $(x,y)$. Theorem 2.2 shows that if Conjecture 3 holds, then the same conclusion follows whenever $(a+b)c$ is a square, which includes equations like $x^2+2y^2=z^2$ in the pair $(x,z)$ and $x^2+y^2=2z^2$ in every pair. The claim is not just the individual theorems: it is that the route through parametric reformulation, the Bochner–Herglotz representation, an aperiodic/pretentious dichotomy, and the Q-trick is the right systematic way to attack dilation-invariant nonlinear partition-regularity problems.","pith_inferences":["An immediate editorial inference is that the Q-trick is likely to transfer to any dilation-invariant pattern whose parametrization involves one reducible and one irreducible binary quadratic form; the paper already indicates this for pairs of the form $(\\ell mn,\\ell'(m^2-n^2))$.","A second inference is that the hardest new input for full Rado-triple pairs is not special number-theoretic structure but the correlation vanishing in Conjecture 3; a proof for a single explicit pair of irreducible quadratics, even with logarithmic averages, would be a decisive test.","A third inference is that the aperiodic/pretentious dichotomy may serve as a template for higher-degree homogeneous polynomials, but the paper's own example involving $n^2+1$ shows such transfers must be checked case by case."],"forward_implications":["Every finite coloring of $\\mathbb{N}$ contains two same-colored integers $x,y$ and a third integer $z$ with $x^2+y^2=z^2$, and the same holds for each pairing of the three variables.","For $x^2+2y^2=z^2$, same-colored pairs exist for $(x,y)$ and $(y,z)$ unconditionally, while the remaining $(x,z)$ pair follows from Conjecture 3.","If Conjecture 3 holds, then for every Rado triple $(a,b,c)$ the equation $ax^2+by^2=cz^2$ is partition regular with respect to all three pairs of variables; this is unconditional when $a=c$ or $b=c$.","The underlying results are density regular, not merely partition regular: every set of positive multiplicative density contains the corresponding pairs in parametric form, so the conclusions survive colorings far from random."],"supporting_citations":[{"why":"Supplies the flagship Theorem 1.3 for Pythagorean pairs and introduces the Q-trick endgame.","marker":"[22]"},{"why":"Proves Theorem 2.1 and the conditional Theorem 2.2, the main results surveyed.","marker":"[23]"},{"why":"Establishes Gowers uniformity of aperiodic multiplicative functions and earlier pair partition-regularity results.","marker":"[21]"},{"why":"Provides the inverse theorem for Gowers norms used to verify uniformity of aperiodic functions.","marker":"[36]"},{"why":"Gives the linear concentration estimate for pretentious multiplicative functions used in the Q-trick.","marker":"[44]"},{"why":"Supplies the C-regular associates in quadratic integer rings needed to handle the infinite units of $\\mathbb{Z}[\\sqrt{2}]$.","marker":"[55]"},{"why":"Establishes existence of the limits of the relevant multilinear averages of multiplicative functions.","marker":"[20]"}],"fun_headline_variants":["Pythagorean pairs always land in the same cell","Every finite coloring has a monochromatic Pythagorean pair","A unified proof: quadratics have monochromatic variable pairs","Pairwise partition regularity for homogeneous quadratics","Solved: Pythagorean equation has monochromatic pairs in any coloring"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conditional part of the argument rests on Conjecture 3: any two unfactorable quadratic forms in two variables that are not multiples of each other will have vanishing correlations when fed an aperiodic multiplicative function, and this is not known even for one explicit pair of such forms with the sign-changing multiplicative function; without it, the treatment of pairs such as $m^2+2n^2$ and $m^2-2n^2$, and hence Theorem 2.2, collapses.","fun_headline_variants_meta":{"raw":{"variants":["Pythagorean pairs always land in the same cell","Every finite coloring has a monochromatic Pythagorean pair","A unified proof: quadratics have monochromatic variable pairs","Pairwise partition regularity for homogeneous quadratics","Solved: Pythagorean equation has monochromatic pairs in any coloring"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000414,"raw_usage":{"total_tokens":2094,"prompt_tokens":854,"completion_tokens":1240,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":470,"completion_tokens_details":{"reasoning_tokens":1163}},"tokens_in":470,"tokens_out":1240,"duration_ms":11896,"temperature":1.0,"reasoning_tokens":1163,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T12:02:08.080925+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute, or bound away from zero, the Cesàro average over $m,n\\le N$ of $\\lambda(m^2+2n^2)\\lambda(m^2-2n^2)$, where $\\lambda$ is the completely multiplicative sign function with $\\lambda(p)=-1$ at every prime, and let $N\\to\\infty$. A non-zero limit would disprove Conjecture 3 and destroy conditional Theorem 2.2; a proof of vanishing for even one explicit pair of irreducible quadratics would be the first confirmatory evidence for the conjecture in the genuinely irreducible setting.","supporting_citations":[{"cited_title":"Frantzikinakis, O","cited_arxiv_id":null,"evidence_quote":"Supplies the flagship Theorem 1.3 for Pythagorean pairs and introduces the Q-trick endgame."},{"cited_title":"Frantzikinakis, B","cited_arxiv_id":null,"evidence_quote":"Establishes Gowers uniformity of aperiodic multiplicative functions and earlier pair partition-regularity results."},{"cited_title":"Green, T","cited_arxiv_id":null,"evidence_quote":"Provides the inverse theorem for Gowers norms used to verify uniformity of aperiodic functions."},{"cited_title":"Klurman, A","cited_arxiv_id":null,"evidence_quote":"Gives the linear concentration estimate for pretentious multiplicative functions used in the Q-trick."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the C-regular associates in quadratic integer rings needed to handle the infinite units of $\\mathbb{Z}[\\sqrt{2}]$."},{"cited_title":"Frantzikinakis, B","cited_arxiv_id":null,"evidence_quote":"Establishes existence of the limits of the relevant multilinear averages of multiplicative functions."}],"review_version":1}