{"id":"2076f39b-e46f-4e9b-8c15-7003ebfa981e","arxiv_id":"2505.18957","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Near-edge states in dense chaotic systems and in JT gravity have a universal, analytically computed fidelity susceptibility distribution that is heavy-tailed yet parametrically more rigid than bulk states.","lead":"This paper derives the statistical distribution of how much the first few quantum states above the ground state of a chaotic random matrix ensemble change under perturbations, and shows these edge states are unusually rigid. It maps the same result into two-dimensional gravity and topological string theory, giving a precise probe of near-ground-state black hole quantum mechanics.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The RMT edge result appears well supported, but the claim that low-dimensional gravity realizes the UFL rests on a leading-order KS-string identification that is labeled 'non-perturbative' without proof; that step is the load-bearing weak point.","rationale":"Good-faith reading: the paper is doing two things: (i) deriving a universal edge fidelity-susceptibility distribution for dense chaotic systems via the Kontsevich model, and (ii) claiming that low-dimensional gravity is the only microscopic realization. For (i), the argument is built on a standard supersymmetric sigma-model/Kontsevich framework and is supported by the provided numerics and Zenodo code; I see no reason to doubt the central RMT result, and the finite-size shift in Fig. 2 is a legitimate leading-correction effect at D=1000. For (ii), the bridge to gravity runs through the Kodaira-Spencer reduction and the claim that integrating out hidden open string modes implements JT's own ensemble. The end-matter computation, however, explicitly uses a free-field replacement at leading order in λKS before concluding a match with Eq. (6) and calling it non-perturbative. That is a real gap between what is shown and what is claimed. The Discussion's own 'weak link' sentence is an in-scope limitation statement and supports this reading. I therefore agree with the reader's weakest-assumption diagnosis. The verdict should remain conditional: the RMT result stands, but the strongest holographic claim should not be taken at face value until the all-orders status of the KS identification is clarified or a concrete microscopic construction of the hidden ensemble is provided.","tokens_in":12250,"tokens_out":7746,"duration_ms":73841,"concrete_test":"Compute the next-order correction to the KS correlation function in Eq. (9): evaluate Ẑ with one insertion of the cubic chiral-boson interaction (order λKS) and insert the result into Eq. (11). If the correction is nonvanishing and cannot be absorbed into a redefinition of c = exp(S0) and the brane coordinates, then the asserted non-perturbative equivalence after Eq. (11) fails and the gravitational realization of the UFL is unproven; if the correction vanishes or only renormalizes c, the concern is resolved. This analytic check directly tests the hidden-ensemble mechanism on which the density claim for low-dimensional gravity depends.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim has two parts: the RMT computation of the UFL fidelity distribution, Eq. (7), and the assertion that low-dimensional gravity realizes this universality class. The first part is credible: Eqs. (15)-(16) are only sketched, but the numerical check in Fig. 2 with the stated finite-size shift is consistent. The load-bearing weak point is the second part. The end-matter derivation of the gravity equivalence evaluates the KS correlation function Eq. (9) by replacing the interacting expectation value with the free-theory value after Eq. (11), explicitly 'to leading order in λKS ~ exp(-S0)'. The resulting cubic expression is then said to reproduce Eq. (6) with c = exp(S0), and the text calls this a 'non-perturbative equivalence'. No non-renormalization argument, symmetry protection, or all-orders resummation is supplied to justify dropping the cubic interaction of the chiral boson, so the mapping from the hidden open-string ensemble to the matrix ensemble is not established beyond leading order. If corrections at O(exp(-S0)) modify Z(z), then the identification c = exp(S0) and the entire UFL realization in JT gravity can fail even though the RMT computation is correct. The paper itself flags a 'weak link' in the Discussion; that limitation is precisely this step and should be decisive for how strongly the gravity realization is claimed.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a universality class of the first levels (UFL) above the spectral edge of dense chaotic systems, characterized by a heavy-tailed fidelity susceptibility distribution with a D^{-1/3} scaling variable, and argues that low-dimensional gravity (specifically JT gravity) provides a microscopic realization of this class. The central RMT result is Eq. (7), P(g) proportional to p(g/D^{1/3}) exp(-D/(12g^3)), with p(x) a power-law series. The derivation uses a Kontsevich-model representation of the supersymmetric generating function, an exact Airy-function evaluation in the edge regime, and is compared with numerical simulations. The gravity connection is made through Kodaira-Spencer (KS) string theory, where the fidelity correlation function is represented as a correlation function of brane/anti-brane vertex operators, claimed to reproduce the matrix integral at the edge.","tokens_in":12540,"tokens_out":4401,"duration_ms":41546,"significance":"If the central claims hold, the paper identifies a genuinely new universal regime—rigid, parametrically inert states at the spectral edge—with a concrete observable (fidelity susceptibility) and a predicted scaling form that differs sharply from bulk behavior. The RMT edge computation is a useful technical contribution, and the paper provides reproducible numerical code and data (Zenodo), which is a strength. The string-theory identification, however, is the load-bearing step for the paper's headline claim that low-dimensional gravity realizes the UFL, and that step is only established at leading order. The significance of the gravity realization therefore remains conditional on a non-renormalization argument or an explicit all-orders statement, neither of which is present.","major_comments":[{"comment":"The equivalence between the KS string-theory correlation function and the matrix integral Eq. (6) is established only to leading order in λ_KS: the text explicitly replaces the interacting expectation value with the free-theory value 'to leading order in λKS ∼ exp(−S0)', and then concludes a 'non-perturbative equivalence'. No non-renormalization argument, symmetry protection, or all-orders resummation is supplied to control the cubic interaction of the chiral boson. Since the identification c = exp(S0) and the entire claim that JT gravity realizes the UFL rest on this step, the paper's headline gravity conclusion is not established beyond leading order; the authors themselves flag a 'weak link' in the Discussion, and this is precisely that step. The authors should either provide an argument that the free-field evaluation is exact for this correlation function, or explicitly restate the gravity realization as a leading-order/conjectural equivalence.","section":"End matter, Eq. (11) and following"},{"comment":"The closed-form Airy evaluation of Z(z) is the technical core from which Eq. (7) is derived, but the passage from Eq. (15) to Eq. (16) is not shown; the polynomials q_i(x) are presented only numerically (with the symbol ≃), and the final Gaussian transform to Eq. (7) is described only as a saddle-point integration. Because the coefficients appearing in p(x) (1, 7.12, 11.61, ...) are claimed to be exact to leading order in 1/D^{1/3}, the derivation should either be provided in the supplementary material or the coefficients should be explicitly labeled as numerically evaluated rather than derived in closed form.","section":"Supplemental Material, Eqs. (15)–(16)"},{"comment":"The numerical verification of Eq. (7) in Fig. 2 relies on a post hoc constant shift g → g + 0.47, and the figure does not show error bars; moreover, the number of independent disorder realizations used for the histograms is not stated. The claim of 'excellent agreement' is therefore not quantitatively supportable as presented. The authors should provide error bars or a statistical measure of the discrepancy, and either justify the shift as a controlled finite-size effect or treat it as a fitted parameter with its uncertainty reported.","section":"Fig. 2 and footnote [34]"}],"minor_comments":[{"comment":"The reference title 'Sypersymmetry in Disorder and Chaos' contains a typo and should be 'Supersymmetry in Disorder and Chaos'.","section":"References, Ref. [21]"},{"comment":"The spelling of 'Itzykson-Zuber' is inconsistent: the main text and Eq. (6) use 'Itzykson-Zuber', while the supplemental material uses 'Izykson-Zuber' (twice). Please unify the spelling.","section":"Throughout"},{"comment":"The phrase 'exact to leading order in 1/D^{1/3}' is ambiguous; it should specify what quantity is expanded and in which variable (e.g., corrections of relative order 1/D^{1/3} to the exponent or to the prefactor).","section":"Supplemental Material, final sentence"},{"comment":"The caption states the bulk distribution is scaled by D^{1/3}, but the ordinate label reads 'D1/3 Pbulk(g)'; please make the notation consistent and clarify the normalization of the histograms.","section":"Fig. 2 caption"}],"recommendation":"major_revision","confidential_remarks":"The RMT portion of the paper is a solid contribution, and the data/code availability is commendable. The gravity claim, however, rests on a leading-order KS identification that the authors themselves call a weak link; this is a load-bearing point that should be strengthened or the claims tempered. The paper's novelty relative to the authors' prior work [10,11] could also be clarified, since the KS framework and the free-field evaluation are largely taken from those references."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Hi,\n\nThe part of this paper worth caring about is the RMT calculation of the fidelity susceptibility distribution for the first levels above the spectral edge. Eq. (7) — the D^{1/3} scaling, the x^{-5/2} power law truncated by the exponential, the parametric rigidity of near-edge states — is new and looks right. The numerical check in Fig. 2 is consistent, and they ship code and data, which makes the central claim reproducible. The bulk correlated Levy distribution was known, and the Kontsevich edge sigma model existed; the new content is the edge UFL distribution and its interpretation via state geometry.\n\nThe soft spots are real, but they cluster in the second half. The Airy evaluation between Eqs. (15) and (16) is only sketched; the 0.47 finite-size shift in Fig. 2 is fit post hoc; and the figure has no error bars. None of that sinks the RMT result, though a referee should ask for the derivation and the error bars.\n\nThe bigger issue is the string-theory section. The KS calculation evaluates the correlation function by replacing the interacting expectation value with the free-field value, explicitly to leading order in λ_KS ~ exp(-S0), and then labels the result a 'non-perturbative equivalence'. That label is not supported by anything in the paper, and the paper itself flags the 'weak link' in the Discussion. If corrections at order exp(-S0) change Z(z), the identification c = exp(S0) and the gravity realization of the UFL could fail even though the RMT computation is correct. I don't think the paper proves the string side to all orders; it demonstrates a leading-order match. That is worth saying clearly, because the abstract's claim that the UFL is realized in low-dimensional gravity is stronger than what the end matter establishes.\n\nBottom line: this deserves a serious referee. The RMT part is a solid advance in a specialized corner, and the gravity realization is an interesting conjecture that needs either a non-renormalization argument or a softer claim. I'd send it to peer review with a request to tighten the Airy step, add error bars, and rewrite the string-theory equivalence as a leading-order match unless they can close the loop.\n\nWould I cite it? The RMT result, yes, if I worked on fidelity or edge statistics. It's the kind of paper you bring to a reading group and argue about the gravity section.","headline":"New edge fidelity statistics are solid RMT work with reproducible numerics; the gravity-realization claim is a leading-order match presented as non-perturbative and should be scoped down before it is taken at face value.","tokens_in":13105,"tokens_out":2525,"would_cite":true,"duration_ms":24792,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The first states above a chaotic spectral edge belong to their own rigid universality class, and two-dimensional gravity realizes it.","keywords":["universality class of the first levels","fidelity susceptibility","spectral edge","random matrix theory","Jackiw-Teitelboim gravity","Kodaira-Spencer theory","Kontsevich matrix model","quantum chaos"],"falsifier":"Take a random-matrix Hamiltonian of dimension $D$ with a fixed Gaussian perturbation and histogram the fidelity susceptibility of eigenstates within one near-edge spacing of the edge: if the histogram does not converge to Eq. (7) with the $x = g D^{-1/3}$ scaling and the $g^{-5/2}$ tail as $D \\to \\infty$, the UFL universality claim fails. A sparse Hamiltonian of the same dimension with edge-position fluctuations provides a direct control: if its edge histogram also matches Eq. (7), the density condition is not the controlling ingredient.","tokens_in":12063,"feed_emoji":"🛡️","tokens_out":10253,"duration_ms":74168,"temperature":0.7,"pith_summary":"In chaotic quantum systems whose spectral edge is pinned—systems that are “dense,” with roughly as many independent Hamiltonian parameters as Hilbert-space dimensions—the first few states above the edge behave like nothing else in the spectrum. This paper establishes that these “first levels” form a universality class of their own (the UFL): their wavefunctions barely deform under external perturbations, and their fidelity susceptibility $g$ is drawn from the universal distribution $P(g) \\propto p(g D^{-1/3}) e^{-D/(12g^3)}$, with $p(x) \\sim x^{-5/2}$ for large $x$. The same distribution is reproduced by a Kodaira-Spencer string-theoretic description of two-dimensional Jackiw-Teitelboim gravity, making gravity the only known microscopically defined system that naturally harbors the UFL. If the argument is right, the first states of low-dimensional holography are far better defined than generic chaotic states, and a genuinely non-perturbative probe of the holographic principle becomes available.","feed_headline":"Chaotic edge states obey a universal rigidity law","feed_subtitle":"The same fidelity-susceptibility distribution appears in dense random matrices and in JT gravity, revealing how near-edge states resist…","key_machinery":"The load-bearing object is the Kontsevich matrix model, a supermatrix integral over “flavor” matrices with action $S(A) = c\\,\\mathrm{str}(X A + \\tfrac{1}{3} A^3)$, which is the universal $φ^4$ field theory of the spectral edge. The computation routes the fidelity-susceptibility generating function, a ratio of determinants, through a Hubbard-Stratonovich decoupling into this supermatrix integral, uses the Itzykson-Zuber identity to integrate out angular variables, and then solves the remaining radial integral exactly in terms of Airy functions. On the gravity side, the same objects appear as correlation functions of brane and anti-brane vertex operators $\\psi(x) = e^{\\Phi(x)}$ in a chiral-boson field theory on the spectral curve obtained from Kodaira-Spencer theory; normal ordering generates the super-Vandermonde determinant $s\\Delta(X)$, and free-field expectation values reproduce the cubic Kontsevich action, closing the triangle between random matrix theory, the Kontsevich model, and topological string theory.","core_discovery":"The paper claims that the first $O(1)$ levels above the spectral edge of a dense chaotic system form a distinct universality class: squeezed between a non-fluctuating edge and the repelling bulk, their wavefunctions are almost pinned. The quantitative signature is a universal distribution of the fidelity susceptibility, $P(g) \\propto p(g D^{-1/3}) e^{-D/(12 g^3)}$, where $p(x) \\approx x^{-5/2} + 7.12 x^{-7/2} + \\cdots$; the same power-law exponent as in the bulk is rescued by the scaling variable $x = g D^{-1/3}$, so even large $x$ values correspond to parametrically smaller susceptibilities than in the bulk. The identical distribution is then derived from a Kodaira-Spencer string theory description of JT gravity, where flavor-brane vertex operators represent the determinant insertions of the random-matrix computation and reproduce the Kontsevich matrix model exactly at the edge. The paper concludes that the UFL is a real prediction of two-dimensional gravity, not a random-matrix artifact, and that it is invisible to every semiclassical or perturbative expansion.","pith_inferences":["Editorial inference: the $D^{-1/3}$ scaling in Eq. (7) implies that near-edge state susceptibilities diverge only as $D^{1/3}$ with system size instead of as $D$, so control-error sensitivity of ground states in large dense quantum simulators should be suppressed by this factor.","Editorial inference: if JT gravity indeed performs its own ensemble through hidden open strings, then state-geometry observables such as fidelity susceptibility or entanglement near the edge may provide a sharper test of holography than spectral correlators.","Editorial inference: the UFL prediction for a heavy tail $g^{-5/2}$ in the scaled variable means that rare, anomalously large susceptibilities still occur with probability $\\propto g^{-5/2} D^{5/6}$; searching for these rare events in ion-trap or microwave experiments could confirm the universality class.","Editorial inference: the same edge-rigidity mechanism might explain numerically observed quasi-non-ergodic low-lying states in other dense random-matrix-like many-body systems, and could be checked in the SYK model if its parameter count is increased toward the dense regime."],"forward_implications":["UFL states are parametrically more rigid than bulk states: in the scaled variable $x = g D^{-1/3}$ the heavy tail $x^{-5/2}$ still decays with $g$, but typical susceptibilities are suppressed by $D^{-1/3}$ relative to the bulk.","The UFL is invisible to semiclassical $1/D$ expansions and to the topological expansion of the JT path integral; only non-perturbative open-string (flavor-brane) probes detect it.","JT gravity, through its Kodaira-Spencer completion, is the only microscopically defined system known to realize the UFL “naturally,” because it supplies its own ensemble average via hidden open-string degrees of freedom integrated out in the reduction.","Sparse systems such as few-body chaotic Hamiltonians or SYK-like models do not exhibit the UFL: their edges fluctuate from sample to sample, and averaging destroys the fine-grained signatures.","The UFL distribution is universal across dense realizations, including random matrix ensembles, quantum graphs, dense Haar-random quantum circuits, and two-dimensional gravity."],"supporting_citations":[{"why":"Introduces the notion of dense systems and the universality class of the first levels, defining the edge spectrum that the paper analyzes.","marker":"[4]"},{"why":"Establishes the quantitative equivalence between the JT gravitational path integral and a matrix model in the $1/D$ expansion, the baseline the paper extends beyond perturbation theory.","marker":"[9]"},{"why":"Provides the Kodaira-Spencer string-theory framework for JT gravity, which the paper uses for the gravitational realization of the UFL.","marker":"[10]"},{"why":"Supplies the non-perturbative edge description of topological strings from which the brane correlation functions and Kontsevich action are obtained.","marker":"[11]"},{"why":"Defines the Kontsevich model, the universal edge matrix integral that is the central computational object.","marker":"[20]"},{"why":"Gives the method for computing fidelity susceptibility distributions in bulk chaotic states that the paper adapts to the edge.","marker":"[27]"},{"why":"Establishes the Kontsevich model as the edge extension of the nonlinear sigma model, justifying the contour and edge analysis.","marker":"[22]"},{"why":"Provides the earlier correlated Levy distribution for bulk fidelity susceptibility used for comparison with the edge result.","marker":"[19]"}],"fun_headline_variants":["Edge states share a universal rigidity signature","First levels above chaos edge are universally rigid","Universal fidelity law for edge states in chaos","Gravity and random matrices agree on edge rigidity","Edge-state rigidity is universal across chaos"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The UFL exists only in “dense” systems, where the number of statistically independent Hamiltonian parameters is comparable to the Hilbert-space dimension $D$, so the spectral edge is pinned at a non-fluctuating position; the gravity realization additionally assumes JT gravity incorporates its own ensemble through hidden open-string degrees of freedom in the Kodaira-Spencer reduction.","fun_headline_variants_meta":{"raw":{"variants":["Edge states share a universal rigidity signature","First levels above chaos edge are universally rigid","Universal fidelity law for edge states in chaos","Gravity and random matrices agree on edge rigidity","Edge-state rigidity is universal across chaos"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000455,"raw_usage":{"total_tokens":2253,"prompt_tokens":881,"completion_tokens":1372,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":497,"completion_tokens_details":{"reasoning_tokens":1307}},"tokens_in":497,"tokens_out":1372,"duration_ms":8027,"temperature":1.0,"reasoning_tokens":1307,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:22:55.167114+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a random-matrix Hamiltonian of dimension $D$ with a fixed Gaussian perturbation and histogram the fidelity susceptibility of eigenstates within one near-edge spacing of the edge: if the histogram does not converge to Eq. (7) with the $x = g D^{-1/3}$ scaling and the $g^{-5/2}$ tail as $D \\to \\infty$, the UFL universality claim fails. A sparse Hamiltonian of the same dimension with edge-position fluctuations provides a direct control: if its edge histogram also matches Eq. (7), the density condition is not the controlling ingredient.","supporting_citations":[{"cited_title":"Kontsevich, Commun","cited_arxiv_id":null,"evidence_quote":"Defines the Kontsevich model, the universal edge matrix integral that is the central computational object."},{"cited_title":"Altland and J","cited_arxiv_id":null,"evidence_quote":"Establishes the Kontsevich model as the edge extension of the nonlinear sigma model, justifying the contour and edge analysis."},{"cited_title":"Sierant, A","cited_arxiv_id":null,"evidence_quote":"Provides the earlier correlated Levy distribution for bulk fidelity susceptibility used for comparison with the edge result."}],"review_version":1}