{"id":"765ee7fa-80a8-4966-89dd-95db7b7b4c31","arxiv_id":"2506.14193","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Products of truncated unitary matrices exhibit a universal three-phase transition in local singular value statistics, interpolating between Gaussian and GUE universality as the modified depth-to-width ratio varies from infinity to zero.","lead":"This paper proves that products of many truncated Haar random unitary matrices show the same three-phase transition in local singular value statistics that was previously known for complex Gaussian matrices: Gaussian fluctuations for deep products, interpolating critical kernels at a critical depth, and standard GUE statistics for shallow products.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorems 1.1–1.3 omit a standing dimensional constraint; in the low-DWR regime the theorem's own parameters can violate it, leaving the central phase-transition claim undefined.","rationale":"The reader's weakest_assumption is the same one that strikes me as load-bearing: the explicit kernel representation (1.2) and the associated determinantal structure are only available under the dimensional constraint (1.1). My stress test sharpens this from a general caveat to a concrete failure mode: there are parameter values satisfying the stated hypotheses of Theorem 1.3(i) for which the objects used in the statement do not exist. This does not reveal a mathematical error in the asymptotic arguments, which for the regimes where (1.1) holds appear coherent and technically substantial; it reveals a missing hypothesis. The paper's own text flags the omitted u = 0 proof of Theorem 1.2(ii), and while that is a separate gap in verification, it does not change the conclusion: the correct disposition remains CONDITIONAL, exactly the reader's verdict. I therefore recommend no change to the verdict. The concrete test is deliberately analytic and cheap: it settles the missing-hypothesis issue without requiring new numerical experiments, and the positive control with m_1 = 3n confirms the intended regime is distinguishable from the invalid one.","tokens_in":24148,"tokens_out":17986,"duration_ms":192670,"concrete_test":"Check the M = 1, v_1 = 0, m_1 = 3n/2 case. Verify that (1.1) fails while ∆_{M,n} = 4/(3n) → 0, and that the left-hand side of (1.17), 2/(n+z) − 1/z − 1/(3n/2+z), is strictly negative for all z > 0 (common-denominator computation gives numerator −(3n/2 + z/2) < 0). If this is confirmed, Theorem 1.3(i) is undefined as stated. A positive control: replace m_1 by 3n, so (1.1) holds; then (1.17) has a unique positive root and the Airy limit should be reproducible.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The entire proof of the three-phase transition works from the double-contour correlation kernel (1.2), which the paper itself states is valid only when the dimensional constraint (1.1), n ≤ Σ_{j=1}^M (m_j − n − v_j), holds. But the theorems do not state (1.1) as a hypothesis, and the parameter regimes they impose do not force it. Concretely, take M = 1, v_1 = 0, m_1 = 3n/2. Then ∆_{M,n} = 2/n − 1/(3n/2) = 4/(3n) → 0, so Theorem 1.3(i) applies as written. Yet (1.1) fails: m_1 − n − v_1 = n/2 < n. In this case the edge-scaling construction collapses: equation (1.17) has no positive root, since its left-hand side tends to −∞ as z → 0+ and behaves like (m_1 − 2n)/z^2 = −n/(2z^2) < 0 as z → ∞, so ρ_{M,n} and λ_{M,n} in (1.15)–(1.16) are undefined. Thus Theorem 1.3(i) is not merely unproved but not well-defined for allowed parameters. Theorems 1.1 and 1.2 inherit the same issue through the kernel (1.2). The fix is to impose (1.1) as a standing assumption, or to prove the phase transition on the non-DPP side by other means; without this, the central claim overreaches.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies the local singular-value statistics of products Y_M = T_M ... T_1 of truncated unitary matrices. It introduces the modified depth-to-width ratio Delta_{M,n} in (1.3) and proves three asymptotic regimes for the correlation kernel of the determinantal point process of squared singular values, whenever the explicit double-contour formula (1.2) applies: for Delta_{M,n} to infinity the kernel becomes Gaussian (Theorem 1.1), for Delta_{M,n} to gamma in (0, infinity) it converges to the critical bulk and edge kernels (Theorem 1.2), and for Delta_{M,n} to 0 it converges to the Airy and sine kernels (Theorem 1.3). The proofs use steepest-descent analysis of the kernel (1.2), together with a new parameterization of the limiting density in the low-DWR regime, and the paper also states a simplified edge limit (Theorem 3.1).","tokens_in":24460,"tokens_out":8699,"duration_ms":88478,"significance":"The announced three-phase transition is a natural and valuable counterpart, for the truncated-unitary ensemble, of the Ginibre-product phase transition established in [4,25]. The modified DWR Delta_{M,n} is a useful unifying parameter, and the explicit low-DWR parameterizations leading to Theorem 1.3 and Theorem 3.1 are new. The paper gives detailed steepest-descent estimates for the main bulk and high-DWR cases and tracks error terms in several places. However, several load-bearing ingredients are currently missing or stated too broadly, so the results cannot be accepted in the present form.","major_comments":[{"comment":"Theorems 1.1-1.3 do not state the dimensional condition (1.1) as a hypothesis, although the proofs rely on the determinantal kernel (1.2), which the paper itself says is valid only when (1.1) holds. The parameter regimes in the theorems do not force (1.1): for M=1, v_1=0, m_1=3n/2, one has Delta_{M,n}=4/(3n) to 0, so Theorem 1.3(i) applies as written, but n <= m_1 - n - v_1 = n/2 fails, and equation (1.17) has no positive solution (the cross-multiplied equation gives z = -3n). Hence rho_{M,n}, lambda_{M,n}, and the statement of Theorem 1.3(i) are undefined in an allowed regime; Theorems 1.1 and 1.2 inherit the same problem through their use of kernel (1.2). The fix is to impose (1.1) as a standing assumption or to handle the non-DPP regime by other methods; without this, the theorems overreach.","section":"Section 1.2, Eqs. (1.1)-(1.3)"},{"comment":"The proof of the critical edge limit u=0 is omitted, with the text stating that it is similar to u in (0,1) and only the scaling functions need to be substituted. This is not a negligible variant: the contour analysis and the estimates leading to Lemma 2.2 are specific to the bulk scaling (1.11), while the edge statement (1.13)-(1.14) requires its own verification. Since Theorem 1.2(ii) is one of the three main limits announced in the abstract, the omitted proof must be supplied.","section":"Section 2.2, Theorem 1.2(ii)"},{"comment":"Lemma 3.4 is the global estimate that controls all contributions outside the local saddle-point domains in the proof of Theorem 1.3(ii), and its proof is explicitly omitted. The reference to Lemma 2.1 of [26] is not sufficient, because the present parameterization via the Fuss-Catalan-type density and the kernels eK_n is different; the monotonicity and steepest-descent details need to be checked in this setting.","section":"Section 3.1, Lemma 3.4"},{"comment":"Theorem 3.1 is presented as a theorem but its proof is only sketched. If it is intended as a corollary of Theorem 1.3(i) for the simplified kernel, the reduction should be made explicit; if it is an independent result, the omitted contour-deformation and saddle-point estimates should be provided. As written the reader cannot verify the displayed normalization in (3.10).","section":"Section 3.2, Theorem 3.1"}],"minor_comments":[{"comment":"Throughout the text there are numerous typographical errors (e.g., 'T runcated', 'intoduce', 'propoties', 'etablishing', 'Togeter', 'denstiy', 'Stielties'); these should be corrected.","section":"General"},{"comment":"Equation (1.12) has a misplaced parenthesis in the exponential factor; it should read e^{(g(xi)-g(eta))[nu]} K_n(g(xi),g(eta)).","section":"Eq. (1.12)"},{"comment":"In equation (2.44), the differential dw is written ambiguously after the t-integral; the order of integration should be displayed with parentheses.","section":"Section 2.2, Eq. (2.44)"},{"comment":"The two displays labeled (3.38) and (3.39) are identical, and the definitions of Sigma-tilde_+ and Sigma-tilde_- in (3.14) both use the same interval for theta rather than distinguishing the two conjugate branches; these appear to be typos.","section":"Section 3.1, Eqs. (3.38)-(3.39) and (3.14)"},{"comment":"The bibliography entry for [4] contains a stray '255202 (2014)' at the end.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper follows the structure and many of the estimates of [25] closely; the genuinely new elements are the modified Delta_{M,n} and the low-DWR parameterization. I recommend that the editor ask the authors to state clearly which parts are new relative to [25] and to supply the omitted proofs. There is no indication of misconduct, only a novelty-disclosure question that should be clarified."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper proves the three-phase transition for local singular value statistics of products of truncated unitary matrices, controlled by the single parameter Delta_{M,n}. For anyone who works on random matrix products, this is the real thing: the truncated unitary case was open, and the paper extends the known Ginibre picture with a unified parameter. But the theorems as stated omit the dimensional constraint (1.1), and that omission is not cosmetic. The kernel (1.2), which is the whole engine of the proofs, is only valid when (1.1) holds. The paper says this in Section 1.1 and then never puts it back into the theorem statements. Take M=1, v1=0, m1=3n/2: Delta_{M,n}=4/(3n) -> 0, so Theorem 1.3(i) applies as written, yet (1.1) fails and equation (1.17) has no positive root, so rho_{M,n} and lambda_M in (1.15)-(1.16) are undefined. The theorems are not merely unproved in that regime; they are not defined. The fix is simple: impose (1.1) as a standing hypothesis. If the authors want to claim the transition without it, they need an argument outside the DPP kernel.\n\nWhat is genuinely good: Theorems 1.1-1.3 are new for truncated unitary products. The Gaussian limit (Theorem 1.1) is worked out with contour deformation and explicit error bounds. The critical bulk limit (Theorem 1.2(i)) uses the Euler reflection formula and yields the theta-function kernel. The low-DWR parameterization - especially the explicit density in (3.4)-(3.6) and the edge scaling in Theorem 3.1 - looks useful and goes beyond what was available for Wishart-type models. The self-citation to [25] is fair: the framework is inherited, but the current theorems are proven from the kernel (1.2), not assumed.\n\nThe omissions in the proof matter: the u=0 edge case of Theorem 1.2 is dismissed in one line, Lemma 3.4 is stated without proof, and Theorem 3.1 is only sketched. These are load-bearing for the low-DWR bulk and edge results. Remark 1.1 also overreaches: claiming the transition for inverse Ginibre products (including the spherical ensemble) without proof should be labeled a conjecture or deleted.\n\nIf you work in random matrix theory, this is worth reading, but not citing yet. The main mathematical content is sound in the detailed sections, and the flaw is repairable. I would send it to peer review, with the clear instruction that (1.1) must be a standing assumption and the omitted proofs supplied. Recommendation: engage with it; ask for a revision.","headline":"The truncated-unitary three-phase transition is real and mostly proven, but the theorems drop the standing dimensional constraint and are undefined for some parameter choices.","tokens_in":25042,"tokens_out":5503,"would_cite":false,"duration_ms":50358,"reading_group":"maybe","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 proves that the local singular-value statistics of products of truncated unitary matrices undergo a three-phase transition governed by a single modified depth-to-width ratio.","keywords":["truncated unitary matrices","singular values","determinantal point process","phase transition","critical kernel","depth-to-width ratio","Airy kernel","sine kernel"],"falsifier":"Set $m_j=n+v_j$ for all but one factor so that the dimensional constraint $n\\le\\sum_j(m_j-n-v_j)$ fails while still letting $\\Delta_{M,n}\\to\\gamma\\in(0,\\infty)$, and compute local spacing statistics of the squared singular values numerically; if the limiting statistics are not $K_{\\mathrm{crit}}^{\\mathrm{(bulk)}}$ or $K_{\\mathrm{crit}}^{\\mathrm{(edge)}}$, the claimed universality of the phase transition fails at that boundary. Alternatively, verify the convergence in (1.6) directly for a product with $M\\sim n^2$ by Monte Carlo simulation of the kernel.","tokens_in":23901,"feed_emoji":"🎲","tokens_out":7224,"duration_ms":69236,"temperature":0.7,"pith_summary":"The paper's central claim is that the local singular-value statistics of products of independent truncated unitary matrices are controlled by one modified depth-to-width ratio, $\\Delta_{M,n}=\\sum_{j=0}^M \\frac{1}{n+v_j}-\\sum_{j=1}^M \\frac{1}{m_j}$. As $\\Delta_{M,n}\\to\\infty$ the correlation kernel becomes Gaussian; as $\\Delta_{M,n}\\to\\gamma\\in(0,\\infty)$ it converges to explicit critical kernels; and as $\\Delta_{M,n}\\to0$ it returns to the standard Airy kernel at the edge and the sine kernel in the bulk. This gives products of truncated unitary matrices the same three-phase diagram previously found for products of complex Gaussian matrices. A sympathetic reader would care because singular values of such products appear in wireless communications and chaotic dynamics, and the result reduces many separate limiting statements to a single parameter.","feed_headline":"A single ratio steers singular-value statistics through three phases","feed_subtitle":"The same depth-to-width ratio decides whether singular-value statistics are Gaussian, critical, or GUE.","key_machinery":"The load-bearing object is the modified depth-to-width ratio $\\Delta_{M,n}$ of (1.3), which combines all layer widths $n+v_j$ and row counts $m_j$ into one parameter. It enters the kernel through sums of digamma functions, and the proofs analyze the explicit double contour integral representation (1.2) of the correlation kernel. In the high-DWR regime a steepest-descent argument with a scaled contour through $1-k$ yields the Gaussian kernel; in the moderate regime Euler's reflection formula transforms the kernel into an integral whose residue evaluation produces the $\\theta$-function critical kernels; in the low regime a cubic saddle at a point $z_0$ defined by the rational equation (1.17) gives the Airy kernel, and a separate parameterization of the limiting density yields the sine kernel. The two critical kernels (1.7)-(1.8) are named limit objects that interpolate between GUE statistics and classical Gaussian behavior.","core_discovery":"The discovery is a complete characterization of local singular-value correlations for $Y_M=T_M\\cdots T_1$, where each $T_j$ is an $(n+v_j)\\times(n+v_{j-1})$ truncation of a Haar-distributed unitary matrix. The squared singular values form a determinantal point process with kernel $K_n(x,y)$ written as a double contour integral, and the paper proves three uniform limits for this kernel. In the high-depth regime the rescaled kernel tends to $\\frac{1}{\\sqrt{2\\pi}}e^{-\\eta^2/2}$, so singular-value positions become independent Gaussian fluctuations; in the moderate-depth regime it tends to the bulk and edge critical kernels $K_{\\mathrm{crit}}^{\\mathrm{(bulk)}}$ and $K_{\\mathrm{crit}}^{\\mathrm{(edge)}}$ defined through a Jacobi $\\theta$ function and a gamma-function contour integral; and in the low-depth regime it tends to the Airy kernel at the soft edge and the sine kernel in the bulk. The paper also states the same three-phase transition for products of complex Ginibre matrices and for products of Ginibre with inverse Ginibre matrices.","pith_inferences":["A natural extension not pursued in the paper is testing whether the same modified ratio controls local statistics for products with non-Haar or weakly dependent truncations; the kernel argument would likely need a new determinantal structure.","If the critical edge kernel also appears in last-passage percolation models as the cited literature suggests, the phase transition here may be a random-matrix manifestation of a hard-to-soft transition visible in other probabilistic models.","A concrete testable consequence of Theorem 1.1 is that eigenvalue repulsion disappears in the deep-narrow limit; simulations of spacing distributions for $M\\gg n$ should look essentially independent, while $M\\sim\\gamma n$ should show the critical kernel's signature.","The explicit edge scaling in Theorem 1.3(i), with $z_0$ and $\\lambda_M$, suggests a finite-$M$ correction rate of order $\\rho_{M,n}^{-1/2}$; a numerical check of that rate could confirm that the low-depth regime is reached quickly as $\\Delta_{M,n}\\to0$."],"forward_implications":["If the theorems are correct, the single parameter $\\Delta_{M,n}$ predicts which of the three statistics regimes applies, so fixed-$M$ and growing-$M$ limits are unified rather than treated as separate cases.","In the deep-narrow regime, local singular values become effectively independent: the normalized correlation function converges to a Gaussian, so deterministic scaling plus independent normal fluctuations emerges.","In the balanced regime, the critical kernels provide the interpolation between GUE and Gaussian behavior for truncated unitary products, and they coincide with the critical kernels found for Ginibre products.","In the shallow-wide regime, the product ensemble exhibits GUE universality despite $M\\to\\infty$: the Airy kernel at the soft edge and the sine kernel in the bulk.","The paper's remark extends the same three-phase statement to products of complex Ginibre matrices and to Ginibre-inverse-Ginibre products, including the spherical ensemble as a special case."],"supporting_citations":[{"why":"Establishes that squared singular values of products of truncated unitary matrices form a determinantal point process and provides the explicit correlation kernel.","marker":"[22]"},{"why":"Characterizes when the eigenvalues form a determinantal point process, giving the dimensional constraint (1.1) and the double contour integral kernel (1.2) used throughout the proofs.","marker":"[12]"},{"why":"Established the same three-phase transition for products of complex Ginibre matrices and supplies the critical kernels used as the limit objects here.","marker":"[25]"},{"why":"Independently discovered the transition in the physics literature and introduced the critical bulk and edge kernels that appear in Theorem 1.2.","marker":"[4]"},{"why":"Proved sine and Airy kernel universality for fixed numbers of Ginibre products, providing the low-DWR GUE-statistics framework that is extended here.","marker":"[26]"},{"why":"Supplies the explicit limiting density of squared singular values under the assumption $m_j/n\\to2$, used in the bulk parameterization of the low-depth regime.","marker":"[30]"}],"fun_headline_variants":["Three phases in singular-value statistics of truncated unitary products","Depth-to-width ratio sets three-phase singular-value law","From Gaussian to GUE: universal transition in product matrices","Truncated unitary products: three universal statistics regimes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Everything rests on the explicit double contour integral kernel for the determinantal point process, which is only valid when the dimensional constraint $n\\le\\sum_{j=1}^M(m_j-n-v_j)$ holds; if that constraint fails, the kernel formula and with it all three theorems are not justified.","fun_headline_variants_meta":{"raw":{"variants":["Three phases in singular-value statistics of truncated unitary products","Depth-to-width ratio sets three-phase singular-value law","From Gaussian to GUE: universal transition in product matrices","Truncated unitary products: three universal statistics regimes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000725,"raw_usage":{"total_tokens":3199,"prompt_tokens":842,"completion_tokens":2357,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":458,"completion_tokens_details":{"reasoning_tokens":2293}},"tokens_in":458,"tokens_out":2357,"duration_ms":18407,"temperature":1.0,"reasoning_tokens":2293,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T00:17:55.081708+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Set $m_j=n+v_j$ for all but one factor so that the dimensional constraint $n\\le\\sum_j(m_j-n-v_j)$ fails while still letting $\\Delta_{M,n}\\to\\gamma\\in(0,\\infty)$, and compute local spacing statistics of the squared singular values numerically; if the limiting statistics are not $K_{\\mathrm{crit}}^{\\mathrm{(bulk)}}$ or $K_{\\mathrm{crit}}^{\\mathrm{(edge)}}$, the claimed universality of the phase transition fails at that boundary. Alternatively, verify the convergence in (1.6) directly for a product with $M\\sim n^2$ by Monte Carlo simulation of the kernel.","supporting_citations":[{"cited_title":"and Stivigny, D.: Singular value statistics of matrix products with truncated unitary matrices","cited_arxiv_id":null,"evidence_quote":"Establishes that squared singular values of products of truncated unitary matrices form a determinantal point process and provides the explicit correlation kernel."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Characterizes when the eigenvalues form a determinantal point process, giving the dimensional constraint (1.1) and the double contour integral kernel (1.2) used throughout the proofs."},{"cited_title":"Random Matrices Theory Appl","cited_arxiv_id":null,"evidence_quote":"Supplies the explicit limiting density of squared singular values under the assumption $m_j/n\\to2$, used in the bulk parameterization of the low-depth regime."}],"review_version":1}