{"id":"048df5a6-ffb5-478c-beb4-5e9e67d6403d","arxiv_id":"1908.04963","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The spectral densities of the Gaussian, Laguerre, and Jacobi β-ensembles satisfy explicit linear differential equations of order β+1, derived uniformly for β=2,4 (and Gaussian β=6,2/3).","lead":"This paper derives explicit linear differential equations whose solutions are the eigenvalue densities of the classical random matrix ensembles, for the Gaussian, Laguerre, and Jacobi weights and several values of the symmetry parameter β. It also turns these equations into a uniform tool for moment recurrences and 1/N expansions of resolvents.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The β=4 Jacobi and β=6 Gaussian differential operators are asserted from suppressed computer algebra; the omitted elimination details in Lemma 2.2 and Proposition 2.3 leave the paper's central explicit formulas unverified from the text alone.","rationale":"I read the paper in good faith. The overall strategy is coherent: the β=2 case is proved in detail, the Selberg correlation framework is standard, and the dualities used for β=1 and β=2/3 are plausible and cited. The most load-bearing concern is not the analytic continuation of (2.11), which the reader highlighted; that issue is largely mitigated because (2.11) is an identity with polynomial coefficients in the parameters and the paper explicitly defines the needed objects via analytic continuation. The real weak point is the suppressed computer algebra for the β=4 Jacobi and β=6 Gaussian operators. Lemma 2.2 openly omits the lower-order coefficients, and the proof of Theorem 2.2 does not display the elimination; Proposition 2.3 is likewise terse. Since the paper's contribution is the explicit differential equations, these formulas must be independently reproducible. The reader's verdict is CONDITIONAL, and my concern supports that verdict: the correctness of the central formulas is plausible but not verifiable from the manuscript. I agree with the reader's overall assessment but locate the load-bearing gap differently, hence 'partial' agreement. A single symbolic recomputation along the lines above would settle whether the concern lands; if the formulas check out, the conditional concern would be resolved.","tokens_in":36586,"tokens_out":7399,"duration_ms":75055,"concrete_test":"Use a computer algebra system to eliminate J^(N)_{4,1}(x),...,J^(N)_{4,4}(x) from the 5×5 matrix system (2.18) for generic parameters a, b, N, obtaining the scalar equation for J^(N)_{4,0}(x), then change x→1/x and substitute the proportionality to ρ^(J)_{(1),4,N+1}(x) to obtain the density ODE; compare every coefficient with D^(J)_{4,N}ρ^(J)_{(1),4,N}=0 using the explicit operator (2.23). If all coefficients match for generic symbolic a, b, N, the β=4 Jacobi claim is verified; any mismatch falsifies the claim as written. The same elimination should be applied to (2.41) to check (2.38) for β=6.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is the explicit linear differential equations, most notably the fifth-order Jacobi operator D^(J)_{β,N} in Theorem 2.2 (β=4) and the seventh-order Gaussian operator D^(G)_{β,N} in Proposition 2.3 (β=6). The derivations reduce matrix differential systems (2.18) and (2.41) to scalar equations, but the reductions are not shown. Lemma 2.2 states that the lower-derivative coefficients in the β=4 equation are 'suppressed' and that the full expression is best obtained by computer algebra, while the proof of Theorem 2.2 simply says to 'undertake the same steps' as the β=2 proof. Proposition 2.3 similarly asserts that substituting the expressions from (2.41) 'then yields' the seventh-order equation, with no displayed algebra. Because the theorems are exactly these long coefficient formulas, a single transcription error would falsify the stated result. The checks mentioned (N=1,2, large-N limits, consistency with topological expansions) are not reproduced, and no computer algebra code is supplied. This is an omitted proof of the main deliverable, not a matter of exposition. The reader's named weakest assumption—analytic continuation of (2.11) to b′<−1—is less load-bearing: (2.11) is a polynomial-coefficient differential identity in Selberg parameters, so the analytic continuation is standard and the identity extends by analyticity once the Selberg integrals are defined. The genuine gap is the unverifiable elimination computation for β=4 and β=6.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper derives linear differential equations of order β+1 satisfied by the spectral densities of the classical Gaussian, Laguerre, and Jacobi β-ensembles, together with companion inhomogeneous equations for the resolvents. The explicit equations are given for β=2 and β=4 in all three classical cases, and for β=6 in the Gaussian case, with the cases β=1 and β=2/3 obtained via known β↔4/β dualities. The method uses Selberg correlation integrals: the differential-difference system (2.11) is converted to a matrix differential system whose elimination yields the scalar equation. Applications include recurrences for spectral moments and their 1/N expansion coefficients, first-order differential equations for the resolvent expansion coefficients in the topological expansion, and differential equations characterizing the soft- and hard-edge scaled densities.","tokens_in":36908,"tokens_out":4577,"duration_ms":45875,"significance":"If the displayed formulas are correct, this is a substantial contribution: it unifies and extends known β=2 results (Götze–Tikhomirov, Ledoux, Haagerup–Thorbjørnsen) and gives new explicit characterizations for β=1, β=4, and, in the Gaussian case, β=6 and β=2/3. The framework is principled, resting on external prior results (the Forrester differential-difference system, Selberg integral dualities, and known moment formulas) rather than on the target differential equations themselves. The applications to moment recurrences and to 1/N expansion coefficients (Harer–Zagier type recursions) are of independent interest, and the paper honestly reports machine checks and consistency with topological expansions. However, a substantial part of the central derivation is suppressed computer algebra, and the paper does not make that algebra reproducible.","major_comments":[{"comment":"The derivation of the β=4 Jacobi differential equation is not actually presented. Lemma 2.2 explicitly states that the lower-order coefficient forms are 'suppressed' and are 'obtained most efficiently by computer algebra', and the proof of Theorem 2.2 says one 'undertakes the same steps' as in the β=2 proof without displaying the elimination from the 5×5 system (2.18). Since the content of Theorem 2.2 is exactly the full coefficient formula (2.23), this is an omitted verification of the paper's central claim. Please include the full elimination, or provide a computer algebra script/ancillary file that produces (2.23) and (2.26) from (2.18), along with the verification commands.","section":"Lemma 2.2 and Theorem 2.2 (§2.1)"},{"comment":"The step from the 7×7 matrix differential system (2.41) to the seventh-order scalar operator (2.38) is asserted with the phrase 'then yields', and no intermediate algebra is displayed. As with the β=4 Jacobi case, this is a load-bearing omitted proof: the displayed operator (2.38) is the main deliverable of that subsection. Please supply the elimination details or a machine-readable derivation that the reader can execute to reproduce (2.38) and (2.40).","section":"Proposition 2.3 (§2.3)"},{"comment":"The statements that equations have been 'checked for N=1,2 using computer algebra' and that Proposition 3.11 'has been checked against [72] up to l=6' are not reproducible from the manuscript because no code or detailed outputs are provided. Given the length of (2.23), (2.38), and the recurrence coefficients in §3, a transcription error is a real possibility and the verification is essential. Please provide the computer algebra code, or at minimum the explicit commands and a small sample of output, so the checks can be independently repeated.","section":"Verification statements throughout §2 and §3.3"}],"minor_comments":[{"comment":"The operator for β=2/3 contains powers such as (κ−1)^{7/2} with κ=1/3, i.e. negative base raised to half-integer powers; the branch of (κ−1)^{1/2} and the sense in which the resulting operator is real on the space of even densities should be specified.","section":"Equation (2.38) and (2.40)"},{"comment":"The recurrence in Proposition 3.4 is written as a bare sum; the right-hand side '= 0' is missing. Also, the indices in the display and the stated ranges 'k>12' and 'k>6' in Proposition 3.5 should be checked for consistency.","section":"Equation (3.9)"},{"comment":"The notation for the zero extension of the expansion coefficients is inconsistent, e.g. 'W (G),k β := 0' appears with italic and roman subscripts in the same sentence; this should be cleaned up.","section":"Notation in §3.3"},{"comment":"The text notes that b' may be less than −1 and that the Selberg integral must be interpreted by analytic continuation; since (2.11) is a differential identity with polynomial coefficients in the parameters, a one-sentence justification of why the identity extends to this continued regime would make the derivation fully airtight.","section":"After (2.7), analytic continuation"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely correct in substance, and the authors are clearly experts; the β=2 Jacobi derivation is explicit and the use of external dualities is legitimate. The main concern is reproducibility of the long formulas in Theorem 2.2 and Proposition 2.3. I recommend asking the authors to provide a supplementary computer algebra file or an expanded appendix containing the full elimination. This is a standard request for formulas of this type and should be a straightforward fix. I would not reject on these grounds."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper delivers more than the title promises: a uniform Selberg-integral/duality route to linear ODEs for the spectral densities and resolvents in all three classical ensembles, for β=1,2,4, and new seventh-order Gaussian equations for β=6 and 2/3. The β=2 and β=1,4 Gaussian cases were known, but the Jacobi and Laguerre β=1,4 equations and the β=6/2/3 Gaussian equations are new, and the moment recursions and topological-expansion coefficient equations are useful side products. The derivation from the differential-difference system (2.11) is shown in full for β=2, and the duality arguments for β=1 are elegant and consistent with existing moment data. The paper is honest about what it checks and what it does not.\n\nThe soft spot is exactly where the stress-test puts it. The explicit fifth-order Jacobi operator in Theorem 2.2 and the seventh-order Gaussian operator in Proposition 2.3 are the central deliverables, but their derivation from the matrix systems (2.18) and (2.41) is asserted, not shown. Lemma 2.2 states that the lower-derivative coefficients are suppressed and that computer algebra is the efficient route; Proposition 2.3 says 'then yields' with no displayed algebra. Since a single transcription error would falsify the stated formula, this is an omitted proof of the main result, not a style quibble. That said, the paper does provide consistency checks — N=1,2, large-N limits, agreement with topological expansion coefficients up to l=6 in Proposition 3.11 — and the method is systematic, so my prior is that the formulas are correct. But a referee should not have to take that on faith. The analytic-continuation worry about (2.11) for b′<−1 is minor: the identity is polynomial-coefficient and extends by analyticity once the Selberg integrals are defined. The reliance on Forrester's earlier differential-difference system is legitimate; it is not circular.\n\nWho should read this: anyone working on β-ensemble asymptotics, moment recursions, or edge universality. The paper deserves a serious referee. My recommendation: send it out, and require the authors to supply the full coefficient algebra or the computer-algebra notebook for Lemma 2.2 and Proposition 2.3 before acceptance. With that, it's a solid contribution.","headline":"Genuinely useful paper with new explicit differential equations for classical β-ensemble densities and resolvents; the flagship β=4 and β=6 formulas rest on suppressed computer algebra that referees should ask to see.","tokens_in":37504,"tokens_out":1560,"would_cite":true,"duration_ms":18524,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["15B52","33C45","34A30","60B20"],"pacs":[],"model":"deepseek-v4-flash","headline":"The spectral density of every classical β-ensemble is an exact solution of a linear differential equation of order β+1, and the paper writes these equations down explicitly for β=1, 2, 4, 6, and 2/3.","keywords":["beta ensembles","spectral density","linear differential equations","resolvents","Selberg correlation integrals","duality relations","spectral moments","random matrix theory"],"falsifier":"Take a Jacobi parameter pair with b′<−1, compute $J^{{(N)}}$_{2,0}(x) by an independent method such as the closed form (2.8) in terms of the multivariate hypergeometric function, and check whether it satisfies the differential equation (2.14); equivalently, evaluate $ρ^{{(J)}}$_{(1),1,N}(x) by direct numerical quadrature for N=3 and test whether $D^{{(J)}}$_{1,N}ρ=0 holds.","tokens_in":36375,"feed_emoji":"📐","tokens_out":7890,"duration_ms":75425,"temperature":0.7,"pith_summary":"The paper establishes that for the Gaussian, Laguerre, and Jacobi β-ensembles the spectral density is annihilated by a linear differential operator of order β+1, and that the resolvent satisfies the same operator acting on N⁻¹ times the resolvent, equal to an explicit inhomogeneous term. Explicit operators are given for β=2 and 4 in all three classical settings, for β=6 in the Gaussian case, and the β↔4/β duality supplies the companion β=1 and Gaussian β=2/3 results. These differential characterizations are then used to derive recurrences for spectral moments, first-order differential equations for the coefficients of the 1/N expansions of the resolvents, and edge-scaled ODEs. A sympathetic reader would care because this unifies previously scattered third- and fifth-order equations and extends them to new cases such as β=6 and β=2/3.","feed_headline":"Spectral densities obey exact linear ODEs of order β+1","feed_subtitle":"Gaussian, Laguerre, and Jacobi ensembles are covered, with β=6 and 2/3 new in the Gaussian case.","key_machinery":"The engine is the Selberg correlation integral $J^{{(N)}}$_{n,p}(x), an average of products of characteristic-polynomial factors, which satisfies the first-order differential-difference system (2.11). For even β, the spectral-density average $I^{{(J)}}$_{β,N}(x) is proportional to (-x)^{βN} $J^{{(N)}}$_{β,0}(1/x), so the components p=0,1,...,β form a closed matrix differential equation; eliminating the auxiliary components yields a scalar ODE of order β+1 for the density. The same structure, together with the duality (2.7) and analytic continuation to parameters with b′<−1, carries over to β=1, while the Gaussian and Laguerre cases are obtained by limiting procedures. The Stieltjes transform then turns the density ODE into the resolvent's inhomogeneous ODE.","core_discovery":"The central discovery is that for each classical weight, the averaged power of the characteristic polynomial, and hence the spectral density itself, solves a linear homogeneous ODE of order β+1. For example, in the Jacobi case with β=2, the operator $D^{{(J)}}$_{2,N} satisfies $D^{{(J)}}$_{2,N} $ρ^{{(J)}}$_{(1),2,N}=0, while applying the same operator to N⁻$¹W^{{(J)}}$_{2,N} gives an explicit polynomial right-hand side; analogous statements hold for β=4 and for the Laguerre and Gaussian ensembles. The paper obtains these by substituting the Selberg correlation integral representation into a first-order matrix differential system and eliminating auxiliary components, then uses dualities to reach β=1 and β=2/3. It also derives the corresponding moment recurrences, first-order differential equations for the 1/N expansion coefficients of the resolvents, and the edge-scaled equations.","pith_inferences":["Inference: The paper's elimination scheme is in principle available for any even β, so explicit seventh-order equations for the Laguerre and Jacobi ensembles at β=6 should be derivable by the same limiting procedure, not just for the Gaussian case.","Inference: The observed β-independence of the leading resolvent coefficient W_0^β(x) suggests that the higher coefficients may organize by powers of h=√κ−1/√κ; testing this pattern for the new β=6 and β=2/3 coefficients would sharpen the structural picture.","Inference: A direct numerical check of the Jacobi equations at parameters with b′<−1, beyond the N=1 and N=2 checks reported in the paper, would independently test whether the analytic-continuation step underlying the β=1 Jacobi result is sound."],"forward_implications":["Spectral moments of the Jacobi ensemble obey a third-order linear recurrence for β=2 and a fifth-order one for β=1,4, with analogous Laguerre and Gaussian recurrences; these recover known results and give new ones for β=6 and β=2/3.","The 1/N expansion coefficients W_l^β(x) of the scaled resolvents satisfy first-order differential-difference equations, providing a recursive generation scheme for these coefficients that avoids multi-point correlators.","At the soft edge, the scaled densities for β=2/3, 1, 2, 4, and 6 satisfy a single explicit differential equation D^{soft}_{β,∞}ρ^{soft}=0, and at the hard edge the β=1, 2, and 4 densities satisfy D^{hard}_{β,∞}ρ^{hard}=0.","The differential equations characterize the density even in cases where no closed-form expression is known, such as the β=2/3 soft-edge density."],"supporting_citations":[{"why":"Supplies the differential-difference system (2.11) for Selberg correlation integrals that drives all the scalar eliminations.","marker":"[28]"},{"why":"Provides the Jacobi duality (2.7) and the Selberg integral background needed to relate β to 4/β and to identify parameters requiring analytic continuation.","marker":"[31]"},{"why":"Supplies the Gaussian duality (1.5) connecting the β and 4/β ensembles.","marker":"[3]"},{"why":"Supplies the moment and resolvent dualities for global spectrum fluctuations used to reach β=1 and β=2/3.","marker":"[23]"},{"why":"Provides the large-N topological expansion framework and the computed resolvent coefficients used to check the differential equations.","marker":"[38]"},{"why":"Supplies the Gaussian β-ensemble moments and prior density ODEs that the present method reproduces and extends.","marker":"[72]"},{"why":"Supplies the Jacobi moment recurrence that Proposition 3.1 recovers, and the moment reciprocity laws used in the discussion.","marker":"[14]"},{"why":"Supplies the initial Jacobi and Laguerre spectral moments needed to fix the inhomogeneous terms in the resolvent equations.","marker":"[61]"}],"fun_headline_variants":["Resolvents of classical ensembles solve linear ODEs","Order β+1 ODEs for spectral densities of β ensembles","Gaussian, Laguerre, Jacobi spectra obey exact ODEs","Dualities give ODEs for β=1 and β=2/3 spectra"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the differential-difference system (2.11) for Selberg correlation integrals remains valid under the analytic continuation that sends the Jacobi parameter b′ below −1; if this fails, the derived β=1 and β=4 Jacobi equations, and the Laguerre and Gaussian equations built from them, do not follow.","fun_headline_variants_meta":{"raw":{"variants":["Resolvents of classical ensembles solve linear ODEs","Order β+1 ODEs for spectral densities of β ensembles","Gaussian, Laguerre, Jacobi spectra obey exact ODEs","Dualities give ODEs for β=1 and β=2/3 spectra"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000283,"raw_usage":{"total_tokens":1686,"prompt_tokens":975,"completion_tokens":711,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":591,"completion_tokens_details":{"reasoning_tokens":631}},"tokens_in":591,"tokens_out":711,"duration_ms":6884,"temperature":1.0,"reasoning_tokens":631,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:27:14.676423+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a Jacobi parameter pair with b′<−1, compute $J^{{(N)}}$_{2,0}(x) by an independent method such as the closed form (2.8) in terms of the multivariate hypergeometric function, and check whether it satisfies the differential equation (2.14); equivalently, evaluate $ρ^{{(J)}}$_{(1),1,N}(x) by direct numerical quadrature for N=3 and test whether $D^{{(J)}}$_{1,N}ρ=0 holds.","supporting_citations":[{"cited_title":"Forrester,Recurrence equations for the computation of correlations in the 1/r2 quantum many body system, J","cited_arxiv_id":null,"evidence_quote":"Supplies the differential-difference system (2.11) for Selberg correlation integrals that drives all the scalar eliminations."},{"cited_title":"Forrester, Log-gases and random matrices, Princeton University Press, Princeton, NJ, ( 2010)","cited_arxiv_id":null,"evidence_quote":"Provides the Jacobi duality (2.7) and the Selberg integral background needed to relate β to 4/β and to identify parameters requiring analytic continuation."},{"cited_title":"Baker and P .J","cited_arxiv_id":null,"evidence_quote":"Supplies the Gaussian duality (1.5) connecting the β and 4/β ensembles."},{"cited_title":"Dumitriu and A","cited_arxiv_id":null,"evidence_quote":"Supplies the moment and resolvent dualities for global spectrum fluctuations used to reach β=1 and β=2/3."},{"cited_title":"Forrester, A.A","cited_arxiv_id":null,"evidence_quote":"Provides the large-N topological expansion framework and the computed resolvent coefficients used to check the differential equations."},{"cited_title":"Witte and P .J","cited_arxiv_id":null,"evidence_quote":"Supplies the Gaussian β-ensemble moments and prior density ODEs that the present method reproduces and extends."},{"cited_title":"Cunden, F","cited_arxiv_id":null,"evidence_quote":"Supplies the Jacobi moment recurrence that Proposition 3.1 recovers, and the moment reciprocity laws used in the discussion."},{"cited_title":"Mezzadri, A.K","cited_arxiv_id":null,"evidence_quote":"Supplies the initial Jacobi and Laguerre spectral moments needed to fix the inhomogeneous terms in the resolvent equations."}],"review_version":1}