{"id":"107f0ea0-44c1-422e-825a-ed0fbff8823b","arxiv_id":"2411.11968","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For a two-parameter random-matrix Hamiltonian, the averaged quantum metric becomes a lower hemisphere in the ergodic phase and a cone with a pi/2 aperture angle near the integrable point, with a distinct intermediate non-ergodic regime.","lead":"The paper studies how the averaged quantum metric, a measure of how eigenstates change when a Hamiltonian's parameters change, behaves for random-matrix models. It finds a smooth hemispherical geometry in the chaotic regime and a conical defect with a right-angle aperture at the integrable point, with an intermediate non-ergodic regime in between.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (C8), the exact N=2 metric used to establish the conical defect, is negative for large r; the analytic basis for the claimed cone-to-hemisphere geometry is therefore unverified.","rationale":"I read the paper's central mathematical claim as the existence of a universal conical singularity (aperture π/2) and three scaling regimes for the averaged QGT of this random-matrix family. The all-GUE hemisphere derivation (Appendix B) is clean and parameter-free; the large-N scaling collapse in Figs. 3-4 is suggestive but lacks error bars and code. The most vulnerable step is the N=2 exact calculation: Eq. (C8) is the only fully analytic expression from which the cone is extracted, and it has an impossible large-r limit. This is an internal inconsistency, not a disagreement with consensus; it can be settled by direct numerical averaging of the defining integral. If the numerical average matches Eq. (16) only near r=0 and instead follows Eq. (9) at large r, the correct interpolation likely differs from C8 but the small-r cone may survive; if the average disagrees even near r=0, the N=2 conical claim falls. Either way, the paper needs a corrected Appendix C before the universal cone can be regarded as established. The broader domain extrapolation from diagonal random matrices to generic integrable systems remains a separate interpretive assumption, consistent with the reader's conditional verdict.","tokens_in":12988,"tokens_out":51838,"duration_ms":509398,"concrete_test":"Evaluate the N=2 ensemble average in Appendix C numerically (Monte Carlo over h1,h2,a,b,c,φ with the stated Gaussian measures) and compare Grr(r) and Gφφ(r) with Eqs. (16) at r=0.01, 0.1, 1, 10, 100. If the numerical averages disagree with Eq. (16) for r≳1, the exact N=2 metric is wrong; if they agree only for small r, the large-r crossover to the chaotic hemisphere claimed after Eq. (16b) is unsupported. A symbolic re-derivation of the integral in (C7) should also identify the sign or missing term.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The N=2 analytical result is Eq. (16b)/(C8): Grr = 1/4[ arccot(r√2)/(√2 r) − 1/(2+r²) ]. For r→∞, arccot(r√2)∼1/(r√2), so the bracket behaves as 1/(2r²) − 1/r² = −1/(2r²)+..., giving Grr∼−1/(8r²)<0. This contradicts the positive metric and the claimed crossover to the GUE hemisphere (Eq. 9); it also disagrees with the asymptotic Grr→1/(4r^4) stated in Eq. (C11). The conical angle quoted for N=2 comes from the small-r expansion of this same formula, and the large-N claim is benchmarked against the N=2 case, so an incorrect exact metric undermines the analytic pillar of the central claim. This may be a typo or a missing term, but as written the N=2 derivation is not self-consistent.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the averaged quantum geometric tensor Gαβ (Eq. 5) for two-parameter random-matrix Hamiltonians of the form H = H0 + xHx + yHy. For a GUE base H0 with GUE perturbations, the authors derive closed-form metric components (Eq. 9), embed the resulting two-dimensional manifold in three-dimensional Euclidean space, and identify it as a lower hemisphere. For a diagonal Gaussian base H0, which they associate with an integrable/Rosenzweig-Porter-like system, they derive an exact N=2 metric (Eqs. 16a and 16b), propose a conical singularity with aperture π/2 at the integrable point, and use large-N numerics to identify three scaling regimes—localized, intermediate non-ergodic extended, and ergodic—separated by r* = 1/√N. The paper connects the 1/r divergence of the metric to the loss of level repulsion and discusses connections to spectral complexity and regularization time scales.","tokens_in":13112,"tokens_out":12156,"duration_ms":120697,"significance":"If the N=2 analytical calculation is corrected, the paper provides a clean, parameter-free example in which the averaged quantum metric distinguishes an ergodic phase (smooth spherical geometry) from an integrable limit (conical singularity) in a random-matrix family. The derivations are self-contained Gaussian integrals with no fitted constants, and the large-N scaling predictions are explicit and falsifiable. The main limitation is interpretive: the identification of a diagonal Gaussian ensemble with 'integrability' is an assumption based on Poisson level statistics, so the universality of the conical defect beyond this specific matrix ensemble is not established. Nevertheless, the concrete random-matrix result is a useful contribution if the analytic inconsistency described below is resolved.","major_comments":[{"comment":"The printed exact N=2 radial metric is not self-consistent. For r→∞, arccot(r√2) ~ 1/(r√2), so the bracket in Eq. (C8) behaves as 1/(2r²) − 1/(2+r²) ~ −1/(2r²), giving Grr → −1/(8r²) < 0. This violates positivity, contradicts the stated asymptotic Grr → 1/(4r^4) in Eq. (C11), and is incompatible with the claimed crossover to the chaotic metric Eq. (9), which for N=2 gives Grr ~ 1/(2r^4). Because this exact solution anchors the N=2 conical angle and the large-N interpretation, the derivation must be corrected or replaced. This may be a typographical error or a missing term, but as written the analytic pillar of the central claim is unverified.","section":"Appendix C, Eq. (C8); main text Eq. (16b)"},{"comment":"The paper treats a diagonal matrix with independent Gaussian entries as a representation of an integrable model solely because it has Poisson level statistics. This is a modeling assumption, not a consequence of integrability. Real integrable systems with local interactions or continuous symmetries may not share this singular geometry, and the final paragraph's broader claim about integrable points should either be restricted to the Rosenzweig-Porter-type ensemble or supported by a concrete check, e.g., computing the averaged QGT near an integrability-preserving perturbation of a local integrable spin chain. Without such a check, the conical defect is a feature of this ensemble rather than a proven universal signature of integrability.","section":"Geometry of Integrability Breaking, first paragraph; Conclusions"},{"comment":"The intermediate regime is asserted to be 'still spherical or very close to spherical,' but the evidence presented—saturation of Grr and Gϕϕ/r² at constants scaling as N—does not by itself determine the curvature or guarantee an isometric embedding as a round sphere. This statement is part of the claimed three-regime geometry. The authors should either compute the Gaussian curvature in this regime or explicitly label the spherical interpretation as a conjecture based on the constant components.","section":"Geometry of Integrability Breaking, paragraph after Fig. 4"}],"minor_comments":[{"comment":"The sentence 'Such a diagonal matrix exhibits Poisson level statistics and can therefore be regarded as a representation of an integrable model..' contains a duplicated period; this is a trivial typo.","section":"Geometry of Integrability Breaking"},{"comment":"The text says 'at β ≥ 1 the 1/r behaviour is observable,' but the parameter introduced in that paragraph is γ; β is undefined. This should be corrected to γ.","section":"Appendix E"},{"comment":"The appendix is titled 'Embedding in 3d pseudo-Euclidean space' while the main text uses 'Euclidean space' and the equations (Eqs. 11–13) use the Euclidean signature dZ² + dR² + R²dϕ². The signature used for the N=2 embedding should be stated consistently.","section":"Appendix C and main text"},{"comment":"The sentence 'Thus, we found the explicit expression for metric up to constant to be found numerically' is misleading, because Eq. (B17) subsequently fixes the constant analytically. This wording should be revised.","section":"Appendix B, subsection 2"}],"recommendation":"major_revision","confidential_remarks":"The first major comment is the deciding issue: the exact N=2 metric as printed is self-contradictory, and this is the analytic basis for the claimed conical geometry. This is very likely fixable by correcting a typo or integration error, in which case the paper could become publishable. The second major comment is a scope concern rather than a correctness error; I would recommend that the authors restrict their claims to the random-matrix family or add a concrete test for a physically local integrable model. The numerical scaling analysis is otherwise convincing and well presented."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, what you should know: the paper's central geometric claim — a conical singularity in the averaged quantum metric at the integrable point — is well motivated and has numerical support, but the printed N=2 analytic metric, Eq. (C8), is not self-consistent. For large r, Grr from (C8) behaves as −1/(8r^2), whereas Eq. (C11) states it should go as +1/(4r^4). This is not a cosmetic typo: the same asymptotic is used to argue the crossover to the GUE hemisphere. The small-r expansion that produces the cone angle is unaffected in its leading term, so the conical defect may survive a correction, but as written the analytic pillar is shaky.\n\nWhat is genuinely new: the full two-parameter averaged metric for the all-GUE model, Eqs. (9), is derived without free parameters and the lower-hemisphere embedding is exact. The N=2 integrable-breaking calculation, despite the error, is a serious attempt at a closed form; the 1/r divergence of Grr and the Gϕϕ ~ r behavior near r=0 are what give the cone with aperture π/2. The large-N numerics in Figs. 3 and 4 show a clean three-regime structure with a scaling collapse in r/√N; that part is suggestive and consistent with the Rosenzweig-Porter phases.\n\nThe soft spots are these. First, the algebraic error just mentioned; it needs a corrected derivation before the N=2 claim is taken as established. Second, the model for the integrable point is a diagonal matrix with independent Gaussian entries, justified only by Poisson level statistics. Real integrable systems have local interactions and continuous symmetries, so the conical defect may be a feature of this ensemble rather than of integrability per se. The paper is appropriately cautious in the framing, but the title and abstract push the broader interpretation. Third, the 'spherical' geometry of the intermediate non-ergodic phase is asserted without a quantitative curvature check; Fig. 4 shows saturation but not sphericity. Fourth, the numerics have no error bars or code, which makes the scaling collapse hard to evaluate.\n\nThe reasoning in the appendices is otherwise careful, the Gaussian integrals are parameter-free, and the interpretation leans on the authors' prior framework, which is fine. The paper deserves a serious referee: the core conjecture is testable and may well be right. I would ask for a corrected Eq. (C8), error bars, a direct curvature diagnostic in the intermediate regime, and softer wording about integrability. Conditional accept after major revision, not a reject.","headline":"The N=2 analytic metric as printed is wrong at large r, but the conical-defect claim is numerically supported and worth refereeing.","tokens_in":13725,"tokens_out":4121,"would_cite":false,"duration_ms":36671,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81Q50","15B52"],"pacs":["05.45.Mt","03.65.-w","05.30.-d"],"model":"deepseek-v4-flash","headline":"The averaged quantum metric of a random-matrix family develops a universal conical singularity at the integrable point, with aperture π/2, tracing the loss of level repulsion.","keywords":["quantum geometric tensor","quantum metric","quantum chaos","integrability","random matrix theory","conical singularity","level repulsion","non-ergodic extended phase"],"falsifier":"Compute the averaged quantum metric for a genuinely integrable local spin chain, perturbed by a small non-integrable term along two coupling directions, and check whether the metric diverges as $1/r$ and the embedded surface approaches a cone of aperture $\\pi/2$; if either fails, the conical defect is not a universal signature of integrability.","tokens_in":12736,"feed_emoji":"📐","tokens_out":9386,"duration_ms":80469,"temperature":0.7,"pith_summary":"This paper asks whether the geometry of Hilbert space, as encoded in the quantum geometric tensor, can tell an ergodic (chaotic) Hamiltonian apart from an integrable one. For a two-parameter family of random matrices, the authors compute the averaged quantum metric and find that the ergodic regime has smooth spherical geometry, while the integrable limit develops a conical defect with aperture $\\pi/2$. The $1/r$ divergence of the metric that produces the cone is traced to the loss of level repulsion, that is, to the Poisson spectral statistics of the integrable limit. The paper also identifies three scaling regimes in the metric, localized, intermediate non-ergodic, and ergodic, separated by the scale $r^* = 1/\\sqrt{N}$.","feed_headline":"A conical singularity marks the integrable point of the quantum metric","feed_subtitle":"In a two-parameter random-matrix model, the averaged quantum metric turns from sphere to cone with aperture π/2 at the integrable limit.","key_machinery":"The central object is the averaged quantum geometric tensor $G_{\\alpha\\beta} = (1/N)\\sum_n \\sum_{m\\neq n}\\langle n|\\partial_\\alpha H|m\\rangle\\langle m|\\partial_\\beta H|n\\rangle/(E_n-E_m)^2$, evaluated at infinite temperature. Because the antisymmetric part of this averaged tensor vanishes for Hermitian Hamiltonians, the real part defines a Riemannian metric on the two-dimensional parameter space $(r,\\phi)$. The argument proceeds by embedding this metric into Euclidean space through $R^2(r)=G_{\\phi\\phi}$ and $(dZ/dr)^2+(dR/dr)^2=G_{rr}$; solving these equations maps the metric to a surface whose shape is the main diagnostic. For the fully random family the surface is a hemisphere; for the diagonal-plus-random family the small-$r$ asymptotics give $R \\approx Z \\approx \\sqrt{\\pi/(4\\sqrt{2})}\\,\\sqrt{r}$, which is a cone with aperture $\\pi/2$. The mechanism behind the $1/r$ divergence is the disappearance of level repulsion: with Poisson statistics the energy denominators in the geometric tensor can become arbitrarily small, and a cutoff $\\mu$ of order the level spacing $1/N$ controls the divergence.","core_discovery":"The central claim is that for Hamiltonians $H = \\Lambda_0 + xH_x + yH_y$, with $\\Lambda_0$ a diagonal matrix of independent Gaussian entries and $H_x, H_y$ drawn from the Gaussian Unitary Ensemble, the averaged quantum metric is singular at the integrable point $r=0$: near the origin $ds^2 \\sim \\sqrt{N}\\,(dr^2/r + 2r\\,d\\phi^2)$, whose isometric embedding is a cone with aperture $\\pi/2$. For $N=2$ the metric is computed exactly and shows the same conical small-$r$ behaviour, and for large $N$ numerics confirm it. For large $r$ the metric crosses over to the pure random-matrix result $ds^2 \\sim (dr^2 + r^2 d\\phi^2)/2$, the lower hemisphere. The paper interprets the cone as a geometric signature of integrability, the crossover scale $r^* = 1/\\sqrt{N}$ as the localization transition, and the intermediate window as the non-ergodic extended phase. It also notes that the cone angle matches the conical singularity found earlier at a quantum critical point, suggesting that integrable and critical points share a singular-geometry signature.","pith_inferences":["If the conical defect is generic for integrable points rather than an artefact of the diagonal random matrix, the quantum metric becomes a diagnostic of integrability that requires no level-spacing statistics: a two-parameter scan near a suspected integrable point would reveal the cone directly.","The paper expects stronger divergence for integrability-preserving perturbations that commute with $\\Lambda_0$, but does not compute that case; an analytic or numerical check of the commuting case would show whether the cone is set by spectral statistics alone or also by the structure of the perturbation matrix elements.","Because the metric is linked to a late-time autocorrelation function, the $1/r$ divergence at small perturbation implies a slow $1/t$ tail in the corresponding correlation function; measuring that tail would be a direct experimental test of the conical geometry.","The fact that the same aperture $\\pi/2$ appeared earlier at a quantum critical point raises the question, not addressed here, of whether the cone angle is universal across integrable and critical points or whether different universality classes produce different angles."],"forward_implications":["The averaged quantum metric, not just the fidelity susceptibility at a single point, carries a sharp geometric signature of the ergodic-to-integrable crossover: a universal cone of aperture $\\pi/2$ at the integrable point.","The same calculation yields three distinct scaling regimes for the metric components, separated by the crossover $r^* = 1/\\sqrt{N}$, matching the localized, non-ergodic extended, and ergodic phases.","The conical singularity traces directly to the loss of level repulsion, so observing the cone in the metric can be read as geometric evidence of Poisson-like spectral statistics in the unperturbed system.","The cone angle matches the conical singularity found previously for a quantum critical point, suggesting a common geometric signature of critical slowing down and integrability.","With a finite energy cutoff regularizing the metric, the integrable $1/r$ regime appears when the probe time $t\\sim 1/\\mu$ exceeds $N$, consistent with the known scaling of the localization transition time with system size."],"supporting_citations":[{"why":"Supplies the two-parameter random-matrix family $H=H_0+xH_x+yH_y$ and the quantum-metric statistical setup that this Letter generalizes to the full geometry.","marker":"[32]"},{"why":"Introduces the averaged geometric tensor as a probe of quantum chaos and provides the scaling expectations near integrable points that motivate the calculation.","marker":"[8]"},{"why":"Defines the random-matrix model with localized, non-ergodic extended, and ergodic phases whose phase structure underpins the three scaling regimes identified here.","marker":"[15]"},{"why":"Provides the perturbation-sensitivity analysis of (multi)fractal eigenstates used to interpret the intermediate non-ergodic regime.","marker":"[35]"},{"why":"Introduces the random-energy model, the diagonal Gaussian ensemble that the paper uses as its integrable limit.","marker":"[33]"},{"why":"Documents the conical singularity with the same aperture $\\pi/2$ in the ground-state metric at a quantum critical point, the precedent the paper compares its integrable-point cone to.","marker":"[4]"}],"fun_headline_variants":["Quantum metric forms cone at integrable point","Cone with half-angle π/2 signals integrability","Random-matrix metric: singular cone for integrable phase","Ergodic to integrable: geometry cone transition"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The broad interpretation leans on treating a diagonal matrix with independent Gaussian entries, justified only by its Poisson level statistics, as a faithful stand-in for a generic integrable model; if real integrable Hamiltonians with local interactions do not share this singular geometry, the cone is a property of that ensemble rather than of integrability.","fun_headline_variants_meta":{"raw":{"variants":["Quantum metric forms cone at integrable point","Cone with half-angle π/2 signals integrability","Random-matrix metric: singular cone for integrable phase","Ergodic to integrable: geometry cone transition"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000299,"raw_usage":{"total_tokens":1727,"prompt_tokens":944,"completion_tokens":783,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":560,"completion_tokens_details":{"reasoning_tokens":729}},"tokens_in":560,"tokens_out":783,"duration_ms":8516,"temperature":1.0,"reasoning_tokens":729,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T18:04:27.473289+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the averaged quantum metric for a genuinely integrable local spin chain, perturbed by a small non-integrable term along two coupling directions, and check whether the metric diverges as $1/r$ and the embedded surface approaches a cone of aperture $\\pi/2$; if either fails, the conical defect is not a universal signature of integrability.","supporting_citations":[{"cited_title":"Berry and P","cited_arxiv_id":null,"evidence_quote":"Supplies the two-parameter random-matrix family $H=H_0+xH_x+yH_y$ and the quantum-metric statistical setup that this Letter generalizes to the full geometry."},{"cited_title":"Classifying and measuring the geometry of the quantum ground state manifold","cited_arxiv_id":"1305.0568","evidence_quote":"Documents the conical singularity with the same aperture $\\pi/2$ in the ground-state metric at a quantum critical point, the precedent the paper compares its integrable-point cone to."}],"review_version":1}