{"id":"82935ecd-2de3-4219-8d20-aa66226c6352","arxiv_id":"2506.19940","paper_version":3,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"This paper proves weak and strong convergence in covariance law for general Gaussian matrix ensembles to operator-valued semicircular families and constructs strongly convergent matrix models for interpolated free group factors.","lead":"This paper proves weak and strong convergence, in a precise operator-valued sense, for Gaussian random matrix ensembles to semicircular limit systems described by covariance kernels, including band matrices. It supplies a general framework for strong convergence and builds new matrix models for interpolated free group factors, objects central to von Neumann algebra theory.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 5.5's nuclearity check embeds B = L∞ into the C*-ultraproduct via discretization, but E^(n) is not asymptotically multiplicative on L∞; the argument supports only C(R/Z), leaving a gap in the verification of Theorem B's free-copy hypothesis.","rationale":"The reader's weakest assumption correctly identified the strong convergence of amalgamated free copies as the load-bearing hypothesis in Theorem B. The paper honestly flags that this is not implied by convergence of the covariance data and must be checked case-by-case. My stress-test found a concrete place where that check is not fully justified as written: Theorem 5.5 attempts to verify the free-copy strong convergence via a nuclearity/ultraproduct argument, but the claimed embedding of the full base algebra B = L∞(R/Z) into the C*-ultraproduct is invalid because the discretization maps E^(n) are not asymptotically multiplicative on L∞. This is not just a stylistic issue; the proof of injectivity of π relies on the diagram with B ⊗_max C*(X) and the nuclearity of B, and L∞ is not nuclear. The argument does go through for A = C(R/Z), which is nuclear, and the generating tuple b_ω is taken from C(R/Z), so the theorem can likely be repaired by restricting the embedding to A and observing that strong convergence for polynomials in the generators is all that Theorem B requires. Thus the concern is substantive but not fatal to the overall framework: Theorem A, Theorem B, and Theorem C (continuously weighted Wigner matrices) appear sound. The conditional verdict reflects that Theorem 5.5, one of the paper's advertised applications, currently has a gap in its written proof that needs a concrete fix before the claim is fully supported. The proposed test directly checks the failure of multiplicativity and then verifies the repaired argument.","tokens_in":46522,"tokens_out":28239,"duration_ms":280325,"concrete_test":"Take f = g = 1_{[0,1/2]} in L∞(R/Z) and compute limsup_{n→∞} ||E^(n)(fg) − E^(n)(f)E^(n)(g)||∞. If it is nonzero (it is at least 1/4), the map b ↦ (E^(n)(b))_U fails to be multiplicative on L∞, confirming the gap. Then rerun the injection/nuclearity diagram with A = C(R/Z) in place of B: verify that \\tildeπ: A ⊗_min C*(X) → ∏_U W*(D_n, X^(n)_free) is injective and that the canonical map C*(X) ↪ ∏_U C*(D_n, X^(n)_free) is well-defined, and confirm that strong convergence for polynomials in the continuous generators b_ω plus S is enough to apply Theorem B. If this repaired argument goes through, the theorem is valid despite the written gap.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"In Theorem 5.5, to verify the strong-convergence hypothesis of Theorem B, the proof constructs a *-homomorphism π: B ⊗_max C*(X) → ∏_U C*(D_n, X^(n)_free) with B = L∞(R/Z), asserting that b ↦ (b^(n))_U = (E^(n)(b))_U is asymptotically multiplicative. This fails for general L∞: for f = 1_{[0,1/2]}, the entries of E^(n)(f^2) − E^(n)(f)^2 equal n∫_{I_k} f^2 − n^2(∫_{I_k} f)^2, which is −1/4 on the boundary interval for all n, so the ultraproduct difference has norm at least 1/4. Thus the embedding of B is not defined. What the argument actually establishes is an embedding of A = C(R/Z), the continuous generating tuple b_ω. Because the covariance polynomials in Theorem B only involve the generating tuple b_ω ∈ C(R/Z), and the free-copy hypothesis also only concerns polynomials in these generators, the proof can be repaired by running the nuclearity argument on A and then applying Theorem B. But as written, the verification of the free-copy strong convergence for the full base algebra B is not justified. This is the most load-bearing soft spot: Theorem 5.5's claim of operator-valued strong convergence for shrinking band matrices depends on this step.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a framework for weak and strong convergence of random matrix models to operator-valued semicircular systems. It introduces covariance polynomials, which are non-commutative polynomials augmented by iterated applications of an operator-valued covariance matrix, and the associated notion of covariance laws. The main abstract results are Theorem A, giving weak convergence of general Gaussian matrix ensembles when the associated Choi matrices tend to zero, and Theorem B, giving strong convergence under a logarithmic rate condition on the Choi matrices plus an additional strong-convergence hypothesis on amalgamated free copies of the corresponding semicircular families. The paper then applies these results to continuously weighted Gaussian Wigner matrices (Theorem C), to Gaussian band matrices with shrinking width (Theorem 5.5), and to random matrix models for interpolated free group factors. The proofs use Wick expansions, the concentration and alignment estimates of Bandeira–Boedihardjo–van Handel, and the Haagerup–Thorbjørnsen linearization technique.","tokens_in":46737,"tokens_out":20167,"duration_ms":195239,"significance":"If the results are correct, the paper provides a systematic and general operator-valued extension of strong convergence, with a genuinely new formalism of covariance laws that is likely to be useful in future work. The proofs are detailed and make their hypotheses explicit, and the paper honestly flags the main limitation of Theorem B: the strong-convergence hypothesis on free copies is not known to follow from strong convergence of the covariance data alone, and must be verified case by case. The applications to continuously weighted Wigner matrices and to interpolated free group factors are natural and nontrivial, and no parameters are fitted to force the stated limits. The paper also exhibits strong convergence for band matrices with shrinking width, which goes beyond the earlier weak-convergence results in the literature.","major_comments":[{"comment":"In the paragraph beginning 'On the other hand, we obtain a *-homomorphism from B⊗max C*(X)...', the proof asserts that the map b ↦ (E^(n)(b))_U is an embedding of B = L∞(R/Z) into the C*-ultraproduct because E^(n) is 'asymptotically multiplicative.' This is false for general L∞ functions: for f = 1_{[0,1/2]} and the boundary interval I_k containing 1/2, the diagonal entry of E^(n)(f^2) − E^(n)(f)^2 equals 1/4 for every n, so the ultraproduct difference has norm at least 1/4. Thus the asserted embedding of L∞(R/Z) is not defined. The argument does establish the corresponding statement for A = C(R/Z), the C*-algebra of the generating tuple b, because E^(n)(a) → a uniformly for continuous a. Since the covariance polynomials in Theorem B and the free-copy strong-convergence hypothesis only involve the generating tuple b_ω ∈ C(R/Z), the proof can be repaired by applying the nuclearity argument to A and then invoking Theorem B. As written, however, the verification of the free-copy strong-convergence hypothesis for the full base algebra B = L∞(R/Z) is unjustified, and this is a load-bearing step in the proof of Theorem 5.5.","section":"5.2, proof of Theorem 5.5"}],"minor_comments":[{"comment":"In the estimate for ∥g(b^(k), (X^(k,t))_{t=1}^{t0})∥, the final term is stated as max_{i,t} ∥X^(k,t)_i∥ without an exponent, but for a polynomial of degree > 1 this term should be raised to a power corresponding to the degree (with the constant M adjusted), so that the hypothesis of Lemma 4.10 is actually satisfied.","section":"4.3, proof of Theorem 4.11"},{"comment":"The sentence 'the isomorphism of the two tensor products holds since B = C(R/Z) is nuclear' is inaccurate: B is L∞(R/Z), and the nuclearity needed is that of C(R/Z), the subalgebra generated by the b_ω. The surrounding argument should be rephrased in terms of A = C(R/Z).","section":"5.2, proof of Theorem 5.5"},{"comment":"The explanation that ι_0(Y_{t(r1)}) and ι_0(Y_{t(r2)}) cannot be connected by a cumulant unless t(r1) = t(r2) is cryptic and appears to contain index inconsistencies (e.g., 't(m)' where 't(r2)' is meant). The claim is correct, but the paragraph should be rewritten for clarity.","section":"4.2, proof of Lemma 4.7"},{"comment":"Several displays contain garbled symbols, especially the limit expression in Lemma 3.9 and some formulas in Section 5.1; these should be repaired in the final version.","section":"3.3, Lemma 3.9 and elsewhere"}],"recommendation":"major_revision","confidential_remarks":"The only substantive issue I found is the gap in the proof of Theorem 5.5 described in the major comment. It is local and repairable: replacing the attempted embedding of L∞(R/Z) by the correct embedding of C(R/Z) and rerunning the same argument should fix the proof, since the strong-convergence hypothesis of Theorem B only involves the generating tuple b_ω ∈ C(R/Z). I do not believe the issue affects the central theorems (A, B, C). The authors' own caveats about Theorem B's hypothesis and about the L∞-weight case in Remark 5.2 are appropriately honest. I therefore recommend major revision rather than rejection; after the repair, the paper should be suitable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis is a serious paper, and the first thing you should know is that the covariance-polynomial framework is a real step forward, not a repackaging. The authors define covariance laws that encode iterated applications of the operator-valued covariance, and Theorem B reduces strong convergence of Gaussian matrix ensembles to strong convergence of amalgamated free copies. That reduction is the heart of the paper, and it is novel. Theorem C and Corollary D (strongly convergent matrix models for interpolated free group factors) are substantial applications. The proofs are thorough and use the right tools: BBvH23 estimates, HT linearization, and Shlyakhtenko's A-valued semicirculars. I found no circularity and no fitted parameters.\n\nThe main soft spot is exactly where the stress-test note points. In Theorem 5.5, to verify Theorem B's free-copy hypothesis for shrinking band matrices, the proof embeds B = L∞(R/Z) into the C*-ultraproduct via b ↦ (E^(n)(b))_U and asserts asymptotic multiplicativity. That is false for general L∞; for f = 1_{[0,1/2]}, the discrepancy E^(n)(f^2) − E^(n)(f)^2 has norm 1/4 on the boundary interval for all n. The argument only supports the continuous generators b_ω ∈ C(R/Z). Since the covariance polynomials and the free-copy hypothesis in Theorem B only involve those generators, the theorem can be repaired by running the nuclearity argument on the C*-algebra generated by b (commutative, hence nuclear) and then applying Theorem B. As written, the proof overclaims, but the gap is not load-bearing for the stated convergence result. I think the repair works, but the authors need to fix this step.\n\nA second, smaller caveat is inherent: Theorem B's sufficient condition (strong convergence of free copies) must be verified case-by-case, which the authors state openly. That is not a flaw, just a limit of the method.\n\nOverall: this deserves a serious referee. The framework and main results are important, and the band-matrix proof needs revision. I'd accept for peer review and expect a major revision at most.","headline":"A genuinely new operator-valued strong convergence framework, with a repairable gap in the band-matrix proof.","tokens_in":47346,"tokens_out":3640,"would_cite":true,"duration_ms":40285,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":null,"created_at":"2026-08-15T18:22:27.290197+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":null,"supporting_citations":[],"review_version":2}