{"id":"205f9aa4-5cb1-48dd-b6f3-0d304b92447e","arxiv_id":"2507.07431","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For products of rectangular complex Ginibre matrices, the depth-to-width ratio Delta is the sharp threshold: Delta tending to 0 gives Airy kernel edge statistics, Delta tending to infinity gives Gaussian edge fluctuations.","lead":"Multiplying many random matrices moves the edge statistics of the product from Tracy-Widom (Airy kernel) behavior to Gaussian fluctuations, with the depth-to-width ratio as the exact order parameter. This maps the universality boundary for a natural random matrix family used in wireless and deep network models.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 2.2's derivative expansion (2.58) drops the 1/x0 term, so the sign and global decay used for Theorem 1.2 are unproven; the Airy-kernel proof is incomplete.","rationale":"The reader's weakest assumption points exactly at the global steepest-descent estimates in Theorem 1.2, particularly the incomplete expansion (2.58) in Lemma 2.2. I agree that this is the most load-bearing gap: Theorem 1.2 is one of the two central results, and its proof's final step (2.57) relies on the unverified sign/rate from Lemma 2.2 and the unexplained exponent -27/20. The rest of the paper gives independent structural support for the conclusion (the Fuss-Catalan edge for square factors, the single-matrix soft edge for one rectangular factor, and the Gaussian scaling matching Lyapunov variance formulas), so the issue is a genuine proof gap rather than a demonstrated falsehood. The abstract's inversion of the regimes is a real communication error but does not affect the validity of the theorems as stated. The appropriate disposition remains CONDITIONAL: the claimed results are plausible and partially checked, but the proof of Theorem 1.2 must be completed before the paper can be considered fully proven.","tokens_in":110,"tokens_out":6408,"duration_ms":480396,"concrete_test":"Re-derive (2.58) retaining the 1/x0 term: d/dy Re f(x0+iy) = y( rho^{3/2} sum_j (N+v_j+rho^{3/2} x0)^{-1} - 1/x0 ) + O(y^2). Then (i) verify this is negative for y>0 and positive for y<0 for all x0>q0 under Delta -> 0, using the defining equation for z0; and (ii) check whether this sign plus the contour geometry produces the claimed decay exponent -27/20 in (2.56). If (i) fails for some rectangularity profile (e.g. v_j growing at different rates), Lemma 2.2 is false and Theorem 1.2's proof must be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim of Theorem 1.2 (Airy edge as Delta -> 0) depends on the global steepest-descent estimate (2.56)-(2.57), whose proof via Lemma 2.2 contains an incomplete expansion. For z = x0 + iy, Re f_{M,N} has derivative d/dy Re f = Im f' = sum_j arg(1 + rho^{3/2} z/(N+v_j)) - arg z. For small y this equals y( rho^{3/2} sum_j 1/(N+v_j+rho^{3/2} x0) - 1/x0 ) + O(y^2), not the printed -rho^{3/2} y sum_j ... + O(Delta_{M,N}). The missing -1/x0 term is exactly what cancels at q0 (where rho^{3/2} q0 = z0 satisfies the defining equation (1.10)), so the sign for x0 > q0 is not established by the displayed computation. Since Lemma 2.2 is the only justification that Re f_{M,N} decreases along C1_global and hence that the global integral is O(e^{-epsilon rho^{-3/20}}), the uniform Airy limit (1.13) lacks a verified key ingredient. This is a proof gap, not evidence the theorem is false: the local Airy calculation (2.52) is standard and the edge parameter lambda_M matches known Fuss-Catalan checks.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the edge statistics of singular values of products of independent rectangular complex Ginibre matrices with dimension sequence (N+v_j) x (N+v_{j-1}) and v_0=0. Building on the exact correlation kernel (1.2) and on the depth-to-width ratio (DWR) Δ_{M,N}=Σ_{j=0}^M 1/(N+v_j) introduced in [15], it proves two theorems. Theorem 1.1 states that when Δ_{M,N}→∞, after centering by Σ_j ψ(N+v_j+1-k) and scaling by ρ_{M,N}(k), the correlation functions converge to a Gaussian density for n=1 and vanish for n>1. Theorem 1.2 states that when Δ_{M,N}→0, with scaling ρ_{M,N} defined by (1.8) and centering log λ_{M,N} defined by (1.9)-(1.10), the correlation functions converge uniformly on compacts to the Airy kernel determinant. The proofs use steepest-descent analysis on the kernel (1.2), with local Taylor expansions and global decay estimates. Together with the moderate-DWR critical kernel from [15], the parameter Δ_{M,N} is claimed to be a sharp threshold between Gaussian and Airy edge statistics.","tokens_in":11688,"tokens_out":6436,"duration_ms":64824,"significance":"If the proof gaps are closed, the paper would establish a sharp, parameter-free phase transition for edge statistics of rectangular Ginibre products, extending the square-case results of [15] to non-square multiplicative chains. The claims are anchored by several external consistency checks: the M=2 square edge parameter λ_M reduces to the known Fuss-Catalan edge, the one-rectangular-factor limit matches the Wishart edge (√(N+v_1)+√N)^2, and the Gaussian variance in Theorem 1.1 matches the known Lyapunov variance. The derivation starts from the exact kernel and contains no fitting parameters or target-dependent normalization, which is a genuine strength. However, the global steepest-descent estimates that carry the main theorems are currently under-verified, so the paper is not yet at the standard of a journal publication.","major_comments":[{"comment":"The derivative expansion (2.58) drops the term coming from Im log z in d/dy Re f_{M,N}(x0+iy). The correct first-order coefficient is 1/x0 - ρ^{3/2} Σ_j 1/(N+v_j+ρ^{3/2}x0), not -ρ^{3/2} Σ_j ... . At x0=q0, equation (1.10) makes this coefficient vanish identically, so the printed formula does not even give the correct sign at the critical point; for x0>q0 the sign depends on a nontrivial comparison between 1/x0 and ρ^{3/2}Σ_j(N+v_j+ρ^{3/2}x0)^{-1} that is not supplied. Since Lemma 2.2 is the only justification for the decay of Re f along C1_global and hence for the global bound (2.56)-(2.57), the uniform Airy limit in Theorem 1.2 is not rigorously established as written.","section":"§2.2, Lemma 2.2, Eq. (2.58)"},{"comment":"The Gaussian theorem rests on uniform linear-decay estimates that are asserted rather than proved. Inequality (2.22) postulates an ε>0 with no argument controlling Im ψ uniformly on the vertical contour; (2.23) states an endpoint value (1/2)√N ρ^{1/2}(k)(1+o(1)) without derivation; and Lemma 2.1's inequality (2.25) is justified by a sketch on finitely many contour segments, with the uniformity of the error in (2.29) not quantified. These estimates are plausible and likely fixable, but they are load-bearing for the conclusion (2.27) that I2 and I_1^{global} are negligible, so Theorem 1.1 also needs a more complete proof.","section":"§2.1, Step 3, Eqs. (2.22)-(2.25)"}],"minor_comments":[{"comment":"The abstract's first paragraph reverses the nomenclature used in Section 1.1 and in the theorems: Gaussian fluctuations occur for Δ→∞ (high DWR) and the Airy kernel for Δ→0 (low DWR), while the abstract states the opposite.","section":"Abstract"},{"comment":"The subscript in ρ_{M,n}(k) should be ρ_{M,N}(k); the same inconsistency appears in (2.11).","section":"§2.1, Eq. (2.6)"},{"comment":"There are several typos: 'digama' (page 3), 'makig' (before (2.50)), 'maximun' and 'leftend' (page 9).","section":"Throughout"},{"comment":"The notation C_{ρ^{1/20}_{M,N}} and Σ_{ρ^{1/20}_{M,N}} for the truncated contours is hard to parse; it would help to define these explicitly as scaled versions of the contours in (2.53)-(2.54).","section":"§2.2, Eq. (2.52)"},{"comment":"Reference [4] contains a stray page number '255202 (2014)' after the EPL volume and page range.","section":"References"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a plausible and genuinely useful paper, but it is written as an announcement and the proof of the Airy-side theorem is incomplete as printed. The result itself—that Delta_{M,N} = sum 1/(N+v_j) is the sharp edge-universality parameter, giving Gaussian when Delta -> infinity and Airy when Delta -> 0—is the right classification, and the author has found a clean statement covering rectangular factors. The exact variance rho^2(k) = sum psi'(N+v_j+1-k) and the z_0-based scaling in Theorem 1.2 are new, and they are independently supported by the M=2 square Fuss-Catalan edge, the one-factor reduction, and known Lyapunov variance formulas. There are no fitted parameters. That is real progress.\n\nThe soft spots are concentrated in the proofs. The abstract has the regimes inverted: it says Gaussian in low DWR and Airy in high DWR, which is exactly backwards from Theorems 1.1 and 1.2. More seriously, Lemma 2.2 is not proven. The derivative printed in (2.58) drops the -1/x0 term. The actual d/dy Re f at x0+iy is y( rho^{3/2} sum_j 1/(N+v_j+rho^{3/2} x0) - 1/x0 ) + O(y^2). At x0 = q0 that bracket vanishes by (1.10); for x0 > q0 the sign depends on both terms. The printed expression drops the cancelling second piece and therefore cannot establish the monotonicity claimed in (2.55). Since that monotonicity feeds directly into the global bound (2.56)-(2.57), Theorem 1.2 lacks a verified key ingredient.\n\nLemma 2.1 and the global estimates are also only sketched: the linear decay (2.22)-(2.23) and the exponent -27/20 in (2.56) are asserted without derivation, and the Sigma_global contour with x2 = -C, y2 is given by fiat. None of this makes me think the theorems are false—the local Airy calculation (2.52) and the edge parameterization are standard and check out—but the paper is not a complete proof in its current form.\n\nWho this is for: people working on product random matrices and local universality. It deserves a serious referee; the result is important enough to go through refereeing despite the proof gaps. My recommendation: ask the author to fix the abstract, redo Lemma 2.2 with the correct derivative, and expand the global estimates. After those are addressed, the paper should be published.","headline":"Plausible sharp-threshold edge classification for rectangular product Ginibre, but the proof of the Airy-side theorem is incomplete because Lemma 2.2's key derivative is wrong as printed.","tokens_in":12366,"tokens_out":4316,"would_cite":false,"duration_ms":45327,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["60B20","15B52"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that the depth-to-width ratio $\\Delta_{M,N}=\\sum_{j=0}^M 1/(N+v_j)$ is a sharp threshold: if it tends to infinity, edge singular values of products of rectangular Ginibre matrices show Gaussian fluctuations; if it tends…","keywords":["products of random matrices","singular values","edge statistics","depth-to-width ratio","Ginibre matrices","Airy kernel","Tracy-Widom distribution","determinantal point process"],"falsifier":"Evaluate the coefficient of the linear term in $y$ of $d\\,\\Re f_{M,N}(x_0+iy)/dy$ at $y=0$, namely $\\sum_{j=0}^M \\rho_{M,N}^3/(N+v_j+\\rho_{M,N}^{3/2}x_0)^2 - 1/x_0^2$; Lemma 2.2 requires this coefficient to be negative for $x_0>q_0$. A concrete check is $M=1$, $N=100$, $v_1=0$ with $x_0$ just above $q_0$. If the sign is wrong for any profile, the global estimate behind Theorem 1.2 fails, even if the local Airy computation survives.","tokens_in":11158,"feed_emoji":"🎲","tokens_out":17920,"duration_ms":161236,"temperature":0.7,"pith_summary":"This paper studies the edge statistics of the logarithmic singular values of a product $Y_M=X_M\\cdots X_1$ of independent rectangular complex Ginibre matrices, where $X_j$ has size $(N+v_j)\\times(N+v_{j-1})$. Its central claim is that the depth-to-width ratio $\\Delta_{M,N}=\\sum_{j=0}^M 1/(N+v_j)$ is a sharp threshold: when $\\Delta_{M,N}\\to\\infty$, the rescaled correlation functions converge to a Gaussian one-point density and all higher correlation functions vanish; when $\\Delta_{M,N}\\to 0$, the same correlation functions converge to the Airy kernel determinant, giving Tracy-Widom GUE edge statistics. Combined with the moderate regime where a critical kernel appears, this gives a complete three-way classification of the soft edge by a single scalar parameter. The interest is that the individual rectangularities of the chain do not matter separately; only their aggregate $\\Delta_{M,N}$ decides the universality class.","feed_headline":"Depth-to-width ratio decides Gaussian vs Airy edge law","feed_subtitle":"One scalar, the summed reciprocal widths, sorts product-Ginibre edge statistics into Gaussian, critical, or Tracy-Widom.","key_machinery":"The load-bearing object is the double integral representation (1.2) of the correlation kernel, obtained from the exact joint densities of the singular values: an integral over a Hankel contour $C$ and a closed contour $\\Sigma$, with ratios of gamma functions encoding all $M+1$ matrix dimensions. The asymptotic analysis is a steepest-descent argument: after a change of variables and Stirling expansion, the kernel is written through a phase $f_{M,N}$ whose critical point is $q_0=z_0\\rho_{M,N}^{-3/2}$, with $z_0$ the root of (1.10). The contours are split into local pieces near $q_0$, where a cubic Taylor expansion produces the Airy kernel, and global pieces whose decay is controlled by the monotonicity of $\\Re f_{M,N}$ along the deformed contours. The depth-to-width ratio $\\Delta_{M,N}$, the sum of the reciprocal widths, is the parameter that selects the regime: it controls the variance scale $\\rho_{M,N}(k)$ in the Gaussian theorem and the small parameter in the Airy theorem.","core_discovery":"On its own terms, the paper proves two limit theorems for the determinantal point process of $\\log(Y_M^*Y_M)$. Theorem 1.1 states that if $\\Delta_{M,N}\\to\\infty$, then for fixed $k\\in\\mathbb{N}$ the $n$-point correlation functions, centred at $\\sum_{j=0}^M\\psi(N+v_j+1-k)$ and scaled by $\\rho_{M,N}(k)=\\left(\\sum_{j=0}^M\\psi'(N+v_j+1-k)\\right)^{1/2}$, converge uniformly on compacts to a single Gaussian density for $n=1$ and to zero for $n>1$. Theorem 1.2 states that if $\\Delta_{M,N}\\to 0$, then with scale $\\rho_{M,N}$ and centre $\\log\\lambda_M$ determined by the unique positive solution $z_0$ of $\\sum_{j=0}^M 1/(N+v_j+z)=1/z$, the $n$-point correlation functions converge uniformly on compacts to the determinant $\\det(K_{\\mathrm{Ai}}(\\xi_i,\\xi_j))_{i,j=1}^n$. The author presents this as completing the phase diagram: the critical kernel occupies the moderate regime, and $\\Delta_{M,N}$ interpolates between Gaussian and Airy statistics at the two ends.","pith_inferences":["A question the paper leaves open is the speed of the transition between the Airy and Gaussian regimes; the uniform formulation suggests one could test whether $\\Delta_{M,N}$ decaying like $1/\\log N$ still gives Airy statistics and where the crossover begins.","A practical reading is that for finite $N$ the single number $\\Delta_{M,N}$, not the individual widths $v_j$, tells a user whether edge fluctuations are Gaussian or Airy-like.","The mechanism points to the same DWR trichotomy for neighbouring product ensembles with similar kernel structures, such as products of truncated unitary matrices, where a modified depth-to-width ratio already appears in related work."],"forward_implications":["If Theorem 1.1 holds, the edge statistics in the high-DWR regime collapse to a single Gaussian density: the $n$-point correlation functions vanish for $n>1$.","If Theorem 1.2 holds, the edge statistics in the low-DWR regime match the universal soft edge of the GUE, encoded by the Airy kernel.","Together with the moderate-DWR critical kernel, the two theorems give a complete classification: Gaussian, critical, and Airy statistics correspond to $\\Delta_{M,N}\\to\\infty$, $\\Delta_{M,N}\\to\\gamma\\in(0,\\infty)$, and $\\Delta_{M,N}\\to 0$.","The low-DWR theorem handles chains whose matrices have different sizes, with the centre $\\log\\lambda_M$ and scale $\\rho_{M,N}$ determined by all the $v_j$'s through the critical equation.","The paper's stated motivation is that these edge laws matter for stability analysis in deep neural networks and communication systems, where products of rectangular random matrices arise."],"supporting_citations":[{"why":"Introduces the depth-to-width ratio and the moderate-DWR critical kernel, and supplies the proof strategy the paper adapts.","marker":"[15]"},{"why":"Derives the exact joint density of the singular values and establishes the determinantal point process structure underlying the correlation functions.","marker":"[5]"},{"why":"Provides the double integral representation (1.2) of the correlation kernel that the entire asymptotic analysis starts from.","marker":"[14]"},{"why":"Establishes the fixed-depth, large-width sine and Airy limits that this paper extends to arbitrary rectangularity profiles.","marker":"[16]"},{"why":"Derives universal Gaussian distributions for finite-size Lyapunov exponents in the deep regime, supporting the high-DWR side of the trichotomy.","marker":"[3]"},{"why":"Identifies the crossover from integrable to chaotic local statistics in the physics literature, which the DWR threshold makes rigorous.","marker":"[4]"}],"fun_headline_variants":["DWR flips Gaussian to Airy at matrix product edges","One ratio sets edge law: Gaussian or Airy","Sharp depth-to-width ratio governs edge fluctuations","Edge statistics of matrix products: Gaussian to Airy transition","DWR threshold: Gaussian to Airy at soft edge"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the global pieces of the deformed integration contours decay at the stated exponential rates; in particular, the sign of the decay rate of the phase near the critical point must be correct for every rectangularity profile.","fun_headline_variants_meta":{"raw":{"variants":["DWR flips Gaussian to Airy at matrix product edges","One ratio sets edge law: Gaussian or Airy","Sharp depth-to-width ratio governs edge fluctuations","Edge statistics of matrix products: Gaussian to Airy transition","DWR threshold: Gaussian to Airy at soft edge"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000853,"raw_usage":{"total_tokens":3673,"prompt_tokens":877,"completion_tokens":2796,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":493,"completion_tokens_details":{"reasoning_tokens":2718}},"tokens_in":493,"tokens_out":2796,"duration_ms":31938,"temperature":1.0,"reasoning_tokens":2718,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:47:38.080893+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate the coefficient of the linear term in $y$ of $d\\,\\Re f_{M,N}(x_0+iy)/dy$ at $y=0$, namely $\\sum_{j=0}^M \\rho_{M,N}^3/(N+v_j+\\rho_{M,N}^{3/2}x_0)^2 - 1/x_0^2$; Lemma 2.2 requires this coefficient to be negative for $x_0>q_0$. A concrete check is $M=1$, $N=100$, $v_1=0$ with $x_0$ just above $q_0$. If the sign is wrong for any profile, the global estimate behind Theorem 1.2 fails, even if the local Airy computation survives.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the depth-to-width ratio and the moderate-DWR critical kernel, and supplies the proof strategy the paper adapts."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Derives the exact joint density of the singular values and establishes the determinantal point process structure underlying the correlation functions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the double integral representation (1.2) of the correlation kernel that the entire asymptotic analysis starts from."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the fixed-depth, large-width sine and Airy limits that this paper extends to arbitrary rectangularity profiles."},{"cited_title":"Journal of Physics A 47(39), 395202, 35 (2014)","cited_arxiv_id":null,"evidence_quote":"Derives universal Gaussian distributions for finite-size Lyapunov exponents in the deep regime, supporting the high-DWR side of the trichotomy."},{"cited_title":"EPL (Europhys","cited_arxiv_id":null,"evidence_quote":"Identifies the crossover from integrable to chaotic local statistics in the physics literature, which the DWR threshold makes rigorous."}],"review_version":1}