{"id":"9a729ed3-3efd-4e91-85bd-582e1eaefb6a","arxiv_id":"1908.05178","paper_version":4,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"For elliptic random matrices with 0<|ϱ|<1, the normalized solution norm decays as t^{-1/2} at critical coupling, rigorously confirming and correcting the Mehlig-Chalker formula.","lead":"This paper rigorously proves the formula for the long-time decay of a large system of randomly coupled linear differential equations with elliptically correlated coefficients, covering the previously open intermediate correlation regime. It also corrects an error in the physics literature and extends the result to general variance profiles.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Kernel (2.15) is off by a factor N: stated Theorem 2.7 contradicts its own elliptic specialization (2.17) and the Ginibre limit.","rationale":"The reader's identified weak point (applicability of the [15] local law to the block matrix H^Z_α) is a reasonable but not decisive concern; the paper explicitly flags the reliance on Example 2.11 of [15], and no concrete failure was identified. However, a more direct and more severe problem appears in the central statement itself: the kernel definition (2.15) is off by a factor N relative to (2.17) and to the proof of (6.8). The normalized trace on the left side of (2.14) is O(1), yet (2.15) as written yields an expression of order 1/N for the elliptic case. This is not a matter of interpretation; it is a direct algebraic inconsistency. Because Theorem 2.7 is the main technical result, the paper cannot be accepted in its current form. The intended formula is almost certainly N^{-1}∑_{ij}, and the proofs of Theorems 2.1 and 2.6 use the corrected version, so the overall approach likely survives a revision. The verdict should therefore be CONDITIONAL: require the authors to correct the normalization in (2.15), (2.16), and Remark 2.8, and to re-verify all subsequent displays that depend on this kernel. The concern is concrete, testable by a finite-N simulation, and does not rely on any external convention.","tokens_in":37472,"tokens_out":19199,"duration_ms":162089,"concrete_test":"Specialize to the Ginibre case (ϱ=0), take ζ1=ζ2=2, and compute both sides explicitly. Since S=(1/N)11^T and b(ζ)=−1/ζ, (2.15) gives K = N^{-2}·1^T(4I−(1/N)11^T)^{-1}1 = 1/(3N). The paper's (2.17) gives 1/(ζ1ζ2−1)=1/3. These differ by a factor N. Independently, simulate an N×N Ginibre matrix with variance 1/N, compute the normalized trace TrN(X−2I)^{-1}(X*−2I)^{-1} for N=1000, and compare the empirical value with 1/3 versus 1/(3N). The simulation will match 1/3, settling that (2.15) needs N^{-1} instead of N^{-2}.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central formula for the kernel in Theorem 2.7, K(ζ1,ζ2) = N^{-2}∑_{ij}[(D^{-1}_{b(ζ1)b(ζ2)} − S)^{-1}]_{ij} in (2.15), is internally inconsistent with the paper's own elliptic-case reduction (2.17), and appears to be missing a factor N. With the paper's definition s_ij = E|x_ij|^2 = 1/N for the elliptic ensemble (Section 2.1.1), S = (1/N)11^T. For constant b, the sum of entries of M = (D^{-1}_{b1b2} − S)^{-1} is N b1b2/(1−b1b2), so (2.15) yields K = b1b2/(N(1−b1b2)), while (2.17) states K = b1b2/(1−b1b2). The proof of (6.8) contains the same mismatch: the limit of TrN of the (3,1) block is (1/N)∑ r_i with r = (D^{-1}_{b1b2} − S)^{-1}1, whereas K as defined in (2.15) equals N^{-2}∑ r_i = N^{-1}·(1/N)∑ r_i, a factor N smaller than the trace being approximated. Thus, as written, Theorem 2.7 gives a vanishing limit for the normalized trace, contradicting both the standard Ginibre limit 1/(ζ1ζ2−1) and the paper's own proof of Theorem 2.1. The intended formula is evidently K = N^{-1}∑_{ij}[...], but the stated theorem and Remark 2.8 need correction.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the large-N linear system ∂_t u_t = -u_t + gX u_t with a random connectivity matrix X whose pairs (x_ij, x_ji) are correlated. It introduces an elliptic-type ensemble with general variance profile S and correlation matrix T, defines the self-consistent pseudospectrum and the extraspectral Dyson equation, and proves a high-probability limit for the normalized trace Tr_N f(X)g(X*) in Theorem 2.7. This trace asymptotics is applied to the elliptic ensemble to obtain an explicit Bessel-function expansion and the t^{-1/2} decay at critical coupling, correcting the formula of [28], and to general elliptic-type matrices to obtain the same power-law decay with a model-dependent coefficient A(S,T). The proof is based on Hermitization, the matrix Dyson equation, and the optimal local law for correlated Hermitian matrices from [15].","tokens_in":37823,"tokens_out":9731,"duration_ms":96950,"significance":"If the stated results are correct, they substantially generalize the earlier paper [14] from independent entries to general elliptic correlations, give the first rigorous treatment of the partially symmetric neural-network model, and pinpoint a concrete error in the final formula of [28]. Strengths of the manuscript include the high-probability (rather than expectation-only) statements, the treatment of non-identical variance profiles, and the explicit and internally consistent Bessel-function reduction in Section 7.1. The authors also honestly flag the delicate point that the 4N×4N Hermitized matrix H^Z_α satisfies only the weaker assumptions (C),(D) of [15] rather than (CD). However, the central kernel formula in Theorem 2.7 is stated with an incorrect normalization, and this must be corrected before the results as printed are valid.","major_comments":[{"comment":"The normalization of the kernel K is off by a factor N. For the elliptic ensemble, S = (1/N)11^T and b is a scalar. Writing c = (b1 b2)^{-1}, one has sum_{ij} [(c I - (1/N)11^T)^{-1}]_{ij} = N/(c - 1), so the right-hand side of (2.15) equals N^{-1} b1 b2/(1 - b1 b2), whereas (2.17) and the proof of Theorem 2.1 require K = b1 b2/(1 - b1 b2). The same mismatch occurs in the proof of (6.8): the (3,1) block of ∂_α M^Z_α|_{\\alpha=0} is D_r with r = (D^{-1}_{b1 b2} - S)^{-1} 1, and Tr_N of that block equals N^{-1} sum_i r_i = N^{-1} sum_{ij} [(D^{-1}_{b1 b2} - S)^{-1}]_{ij}, which is N times the quantity appearing in (2.15). Thus Theorem 2.7 as stated would give a vanishing limit for the normalized trace in the elliptic and Ginibre cases, contradicting the standard limit 1/(ζ1 ζ2 - 1) as well as the paper's own derivation. The intended definition is evidently K = N^{-1} sum_{ij} [(D^{-1}_{b1 b2} - S)^{-1}]_{ij}; this correction should be made in Theorem 2.7, in Remark 2.8, and in the independent-entry formula (2.16).","section":"Theorem 2.7, Eq. (2.15); Remark 2.8; proof of (6.8)"},{"comment":"The proof of Theorem 6.1 depends on applying [15, Theorem 2.1 and Corollary 2.3] to the 4N×4N Hermitian matrix H^Z_α, in which the strongly correlated entries x_ij and x_ji appear in widely separated blocks. The authors state that H^Z_α satisfies assumptions (C) and (D) of [15] via Example 2.11, but the verification is not carried out in the text. Since this is the load-bearing transfer of the local law to the block linearization, please add an explicit check of the correlation-decay and block-structure conditions, or quote precisely the part of [15, Example 2.11] that covers this case.","section":"Sections 5.1 and 6, application of [15] to H^Z_α"}],"minor_comments":[{"comment":"The displayed probability statement has the inequality \"(\\ge 1 - C_{\\epsilon,\\nu}/N^\\nu)\" placed after the closing parenthesis in a nonstandard way; it should read P( ... ) \\ge 1 - C_{\\epsilon,\\nu}/N^\\nu.","section":"Theorem 2.1, Eq. (2.2)"},{"comment":"In the sentence defining K in the proof of (6.8), the reference \"(2.16)\" should be \"(2.15)\" once the normalization correction is made.","section":"Proof of (6.8)"},{"comment":"The phrase \"using the relationship I_j(z) = I_j(z)\" appears to be a typo; the intended statement is presumably the conjugation identity for Bessel functions of complex argument.","section":"Section 7.1, after Eq. (7.6)"},{"comment":"The second sentence says \"This bound follows from Δ_ζ ≥ Δ_{|ζ|} by (7.4)\"; the cross-reference should be to Lemma 7.4 rather than Eq. (7.4).","section":"Corollary 7.6"}],"recommendation":"major_revision","confidential_remarks":"The factor-N error in the kernel definition is simple to repair but sits in the main theorem, so I recommend major revision rather than rejection. The reliance on the local law from [15] is appropriate, but the authors should make the verification of assumptions (C) and (D) for H^Z_α explicit before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a genuine contribution. It closes the 0<|ϱ|<1 gap for elliptic ensembles, rigorously proves and corrects the Chalker–Mehlig formula, and extends the result to variance profiles. The proof architecture—extraspectral Dyson equation, 4x4 Hermitization, stability analysis—is serious and mostly sound. But the main theorem as stated has a factor-N error in the kernel (2.15). The stress-test note is right. I checked the algebra: for the elliptic case S=(1/N)11^T, the inverse has entry sum N/(b1b2−1), so (2.15) gives b1b2/(N(1−b1b2)), not b1b2/(1−b1b2) as claimed in Remark 2.8. Inside the proof of (6.8) they derive that the limit of the resolvent product trace is (1/N)∑ r_i = N^{-1}∑_{ij} M_{ij}, which is N times the stated K. So the proof actually establishes the corrected kernel K = N^{-1}∑_{ij}[...]. As written, Theorem 2.7 would predict a vanishing trace for Ginibre, which is false. This is a serious statement-level flaw, but it is localized: the machinery and the proofs produce the right object, and the elliptic specialization (2.17) is consistent with the corrected normalization. The fix is mechanical—change N^{-2} to N^{-1} in (2.15) and adjust Remark 2.8 and the intro where the kernel is quoted—but it must be done before the paper is publishable. There is also a smaller reliance issue: the optimal local law from [15] is applied to 4N×4N block matrices with correlated entries in distant positions, which the authors flag and justify via Example 2.11. That is a reasonable transfer but not reproved; a referee should push for a precise verification. Self-citation is not a problem here; the cited results are prior independent theorems. Bottom line: this deserves a full referee report, but the verdict should be major revision, not accept as is. Do not let the factor N slip through.","headline":"Substantial paper with a real factor-N blunder in the main kernel formula; the proof actually proves the corrected version.","tokens_in":38372,"tokens_out":7485,"would_cite":true,"duration_ms":66161,"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":"For a large random network with elliptic correlations between reciprocal connections, the averaged squared solution at critical coupling decays as $t^{-1/2}$ with an explicit, corrected prefactor.","keywords":["non-Hermitian random matrices","elliptic ensemble","elliptic-type random matrices","extraspectral Dyson equation","local law","neural networks","power law decay","resolvent products"],"falsifier":"Simulate the elliptic ensemble at critical coupling with a fixed $0<|\\varrho|<1$ and large $N$, averaging $\\|u_t\\|^2$ over initial conditions and comparing the coefficient of $t^{-1/2}$ with $[1+|\\varrho|^2-2\\Re((\\varrho+|\\varrho|^2)/(\\varrho+1))]/(2\\sqrt{\\pi})$. A systematic mismatch in the prefactor or a different decay exponent as $t$ grows within the admissible window $t\\le N^c$ would disprove the central asymptotic formula.","tokens_in":37283,"feed_emoji":"🧠","tokens_out":9294,"duration_ms":83246,"temperature":0.7,"pith_summary":"This paper proves a precise asymptotic description of the long-time decay for a large linear system $\\partial_t u=-u+gXu$, where the random matrix $X$ has elliptic correlations: reciprocal pairs $(x_{ij},x_{ji})$ may be correlated with a complex parameter $\\varrho$, with $|\\varrho|<1$, and different pairs are independent. The central result is that at the critical coupling $g=(1+|\\varrho|^2+2\\Re\\varrho)^{-1/2}$, the expectation of the squared norm of the solution over random initial data decays like $t^{-1/2}$ with an explicit prefactor that depends only on $\\varrho$, for every elliptic ensemble with $0<|\\varrho|<1$. This corrects an earlier non-rigorous formula for the same quantity. For the more general elliptic-type ensemble, with entrywise non-negative reciprocal-correlation matrix $T$, the paper proves the same $t^{-1/2}$ law with a coefficient $A(S,T)$ built from the variance profile $S$ and $T$. The proof is powered by a deterministic approximation of $\\operatorname{Tr}_N f(X)g(X^*)$ by a double contour integral whose kernel is expressed through the solution of the extraspectral Dyson equation.","feed_headline":"Elliptic correlations slow network decay to t^-1/2","feed_subtitle":"Rigorous proof with explicit prefactor replaces a non-rigorous physics formula and covers all partially symmetric connectivity.","key_machinery":"The load-bearing object is the self-consistent pseudo-resolvent $b(\\zeta)\\in\\mathbb{C}^N$, defined as the maximal holomorphic solution of the extraspectral Dyson equation $$1+(\\zeta+Tb(\\zeta))b(\\zeta)=0,\\qquad b(\\zeta)\\sim -1/\\zeta\\ \\text{as }\\zeta\\to\\infty,$$ subject to the spectral-radius condition $r(D_{|b(\\zeta)|^2}S)<1$. It plays the role of the diagonal of the resolvent of $X$ outside the pseudospectrum. The proof machinery is a Hermitization: the product of resolvents $(X-\\zeta_1)^{-1}(X^*-\\zeta_2)^{-1}$ is recovered as an $\\alpha$-derivative of the resolvent of a $4N\\times 4N$ Hermitian block matrix $H^Z_\\alpha$ at $\\alpha=0$. The resolvent of $H^Z_\\alpha$ is compared, with an optimal local law, to the solution of the corresponding matrix Dyson equation; the stability operator of that equation is shown invertible with control $\\Delta_Z^{-1}$, and analytic perturbation theory extracts the singularity of $K$ at $\\zeta_1=\\zeta_2=\\zeta_*$ that produces the power-law decay.","core_discovery":"The paper's main theorem is that for elliptic-type random matrices satisfying assumptions (A), (B), and (2.C-F), the normalized trace of $f(X)g(X^*)$ converges with overwhelming probability to $$\\frac{1}{(2\\pi i)^2}\\oint_\\gamma\\oint_\\gamma f(\\zeta_1)g(\\zeta_2)K(\\zeta_1,\\zeta_2)\\,d\\zeta_1\\,d\\zeta_2,$$ where $K(\\zeta_1,\\zeta_2)=N^{-2}\\sum_{i,j}[(D^{-1}_{b(\\zeta_1)b(\\zeta_2)}-S)^{-1}]_{ij}$ and $b(\\zeta)$ is the unique holomorphic solution of the extraspectral Dyson equation $1+(\\zeta+Tb(\\zeta))b(\\zeta)=0$ with $b(\\zeta)\\sim -1/\\zeta$ at infinity. The singularities of this kernel near the rightmost point of the deterministic pseudospectrum control the long-time behavior: at critical coupling the double integral is dominated by a residue that produces the $t^{-1/2}$ decay and the explicit prefactor in the elliptic case. In the elliptic case, the kernel simplifies to $K(\\zeta_1,\\zeta_2)=b(\\zeta_1)b(\\zeta_2)/(1-b(\\zeta_1)b(\\zeta_2))$, which reduces the problem to a Bessel-function series and yields formulas (2.3) and (2.5). The paper also states the elliptic-type variant with coefficient $A(S,T)$ given explicitly in (7.16).","pith_inferences":["If the same kernel controls eigenvector overlaps, the method may yield rigorous overlap statistics for elliptic-type matrices, where only Gaussian results are known.","The critical-decay exponent being independent of $\\varrho$ for all $|\\varrho|<1$ suggests a wider universality class for partially symmetric random connectivity; one testable extension is to sparse or heavy-tailed connectivity matrices satisfying the same moment and correlation assumptions.","The formulas (2.3) and (2.5) could be compared directly to simulations of nonlinear neural rate models near instability, since the linearized dynamics at critical coupling is the regime the paper analyzes."],"forward_implications":["For every $0<|\\varrho|<1$, the averaged squared norm at critical coupling decays as $t^{-1/2}$, the same exponent as the fully asymmetric case, while the fully symmetric case decays as $t^{-3/2}$.","The explicit formula (2.3) replaces the earlier non-rigorous expression and can be used as a benchmark for finite-$N$ numerical studies of random linear networks.","For elliptic-type matrices with non-negative $T$, the decay exponent $t^{-1/2}$ is universal and the coefficient $A(S,T)$ is computable from $S$, $T$, and the Perron eigenvectors of $D_{|b(\\zeta_*)|^2}S$.","The trace formula for $f(X)g(X^*)$ provides a deterministic equivalent for correlated non-Hermitian matrices that holds with high probability, not just in expectation."],"supporting_citations":[{"why":"Supplies the optimal local law for Hermitian random matrices with correlated entries, used to compare the resolvent of the $4N\\times4N$ block matrix to the MDE solution.","marker":"[15]"},{"why":"The authors' previous result for independent entries that this paper extends to elliptic correlations; its stability-operator estimates are reused with modifications.","marker":"[14]"},{"why":"Prior non-rigorous formula for the time decay as a function of asymmetry that the paper proves and corrects.","marker":"[28]"},{"why":"The neural-network model with asymmetric connectivity that motivates the partially symmetric case and whose analysis is made rigorous.","marker":"[27]"},{"why":"Theory of quadratic vector equations on the upper half-plane, used for the non-negative $T$ case and the Stieltjes representation of $b$.","marker":"[3]"},{"why":"Hermitization of non-Hermitian matrices, the starting point that reduces resolvent products to resolvents of larger Hermitian matrices.","marker":"[17]"},{"why":"Dyson equation with linear self-energy; supplies the spectral-support and stability lemmas used to define the pseudospectrum set $E$ and to extend $b$.","marker":"[5]"}],"fun_headline_variants":["Elliptic correlations: exact decay for random ODE systems","Rigorous t^-1/2 decay law from elliptic random matrices","Correcting asymmetric-network decay with elliptic correlations","Elliptic random ODEs: precise long-time decay proved","Asymmetric connectivity slows to t^-1/2 via elliptic trace formula"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the known Wigner-type approximation of a resolvent by a deterministic matrix still works for the $4N\\times 4N$ block matrix built from $X$ and $X^*$, even though the strongly correlated pairs $x_{ij}$ and $x_{ji}$ lie in widely separated positions within that block matrix; the proof relies on the general correlation conditions from [15] for this transfer.","fun_headline_variants_meta":{"raw":{"variants":["Elliptic correlations: exact decay for random ODE systems","Rigorous t^-1/2 decay law from elliptic random matrices","Correcting asymmetric-network decay with elliptic correlations","Elliptic random ODEs: precise long-time decay proved","Asymmetric connectivity slows to t^-1/2 via elliptic trace formula"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000286,"raw_usage":{"total_tokens":1708,"prompt_tokens":997,"completion_tokens":711,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":613,"completion_tokens_details":{"reasoning_tokens":626}},"tokens_in":613,"tokens_out":711,"duration_ms":6776,"temperature":1.0,"reasoning_tokens":626,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:20:59.280077+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate the elliptic ensemble at critical coupling with a fixed $0<|\\varrho|<1$ and large $N$, averaging $\\|u_t\\|^2$ over initial conditions and comparing the coefficient of $t^{-1/2}$ with $[1+|\\varrho|^2-2\\Re((\\varrho+|\\varrho|^2)/(\\varrho+1))]/(2\\sqrt{\\pi})$. A systematic mismatch in the prefactor or a different decay exponent as $t$ grows within the admissible window $t\\le N^c$ would disprove the central asymptotic formula.","supporting_citations":[{"cited_title":"Erdős, T","cited_arxiv_id":null,"evidence_quote":"Supplies the optimal local law for Hermitian random matrices with correlated entries, used to compare the resolvent of the $4N\\times4N$ block matrix to the MDE solution."},{"cited_title":"Erdős, T","cited_arxiv_id":null,"evidence_quote":"The authors' previous result for independent entries that this paper extends to elliptic correlations; its stability-operator estimates are reused with modifications."},{"cited_title":"Mehlig and J","cited_arxiv_id":null,"evidence_quote":"Prior non-rigorous formula for the time decay as a function of asymmetry that the paper proves and corrects."},{"cited_title":"Martí, N","cited_arxiv_id":null,"evidence_quote":"The neural-network model with asymmetric connectivity that motivates the partially symmetric case and whose analysis is made rigorous."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Theory of quadratic vector equations on the upper half-plane, used for the non-negative $T$ case and the Stieltjes representation of $b$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Hermitization of non-Hermitian matrices, the starting point that reduces resolvent products to resolvents of larger Hermitian matrices."}],"review_version":1}