{"id":"21717d67-6eac-4462-9439-b6c8d1b7cf0e","arxiv_id":"2603.12068","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"An asymptotic expansion of the supersymmetry characteristic function proves that off-diagonal S-matrix elements become Gaussian in the Ericson regime, with explicit 1/Ξ corrections.","lead":"Physicists derived the distribution of resonance-scattering amplitudes as resonances overlap, showing it becomes Gaussian in the Ericson regime. The work turns a 60-year-old empirical observation into an asymptotic theorem for time-reversal-non-invariant systems, with microwave and Monte Carlo checks.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (15) is ill-defined at n=1: Γ(n−1) diverges, yet the direct expansion of Eq. (11) gives a vanishing O(Ξ^{-2}) second-moment correction. The printed moment formula is therefore invalid as written, although the leading Gaussian and Eq. (16) survive.","rationale":"The leading claim — that the S-matrix element distribution becomes Gaussian in the Ericson regime — is credible and independently supported by Monte Carlo and microwave experiments. The n=1 singularity in Eq. (15) is a genuine, concrete internal inconsistency: the formula as printed gives an infinite correction to the second moment, while a direct expansion of the exact starting point shows the correction vanishes at that order. This does not overturn the Gaussian proof, because the resummed characteristic function Eq. (16) is compatible with the n≥2 terms and a zero n=1 term, and the leading-order Gaussian depends only on F_U(0)=1 plus an exponentially small connected-part tail. The reader's primary weakest assumption about finite-M channel-factor corrections is real but less severe than stated: corrections to the approximation (10) enter as O(Ξ q^2) and their leading contribution to the r-integral cancels, so they affect moments only at error order relative to the claimed subleading terms. The paper should nonetheless be corrected at Eq. (15) before being treated as the definitive proof, so the CONDITIONAL verdict stands unchanged.","tokens_in":12835,"tokens_out":34803,"duration_ms":314486,"concrete_test":"Evaluate ∂/∂q' of the r-integral in Eq. (11) at q'=0 for n=1 symbolically. If it vanishes as claimed, re-state Eq. (15) as n≥2 and verify that Eq. (16) is unchanged. Optionally, run a GUE Monte Carlo (≈30000 samples, N=200) at Ξ≈1.4 and fit the second moment of x_s to a + b/Ξ^2; the fitted b should be consistent with zero at the subleading order.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The explicit subleading moment formula, Eq. (15), contains Γ(n−1). For n=1 this is Γ(0), a pole, so the claimed O(Ξ^{−(n+1)}) correction to ⟨x_s^2⟩ is infinite — impossible for a finite moment. Directly expanding Eq. (11) at n=1, the m=1 Watson coefficient is proportional to ∫_0^1 dr [r^2 C + (1−r)^2(−C)], with C = [(g_a^++1)(g_b^++1) − 2(g_a^++g_b^++2)] / [(g_a^++1)^2(g_b^++1)^2]. The integral is zero. Hence the true O(Ξ^{−2}) correction to the second moment vanishes; Eq. (15) is only valid for n≥2, and the n=1 case must be listed separately. The resummed characteristic function in Eq. (16) is consistent with a zero n=1 correction, so the leading Gaussian proof survives, but the paper's 'explicit formulae for the moments' and the transition-correction statements built on Eq. (15) are incorrect as printed. The reader's channel-factor concern, Eq. (10), is less damaging: channel-factor corrections enter as O(Ξ q^2), and their leading r-integral cancels, so they do not compete at the claimed subleading order.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper derives, within the Heidelberg supersymmetry approach for β=2 (GUE) stochastic scattering, an asymptotic expansion in powers of 1/Ξ of the characteristic function of the off-diagonal S-matrix element S_ab. From this expansion it obtains the universal Gaussian distribution for the Ericson regime, explicit subleading moment corrections, and the corresponding cross-section distribution. The results are compared with microwave-network data and Monte Carlo GUE simulations at Ξ=1.424 and Ξ=9.55.","tokens_in":13213,"tokens_out":11078,"duration_ms":109705,"significance":"If the derivation is correct, this is a significant advance: the sixty-year-old heuristic Ericson Gaussian becomes a consequence of the supersymmetry formalism, with explicit finite-Ξ corrections. The paper starts from the exact characteristic function Eq. (4) obtained in prior work, rather than assuming a Gaussian form, and it validates the subleading results against experiment and simulations. The explicit formulas for the moments and distributions are potentially useful for analyzing data in the onset of the Ericson regime.","major_comments":[{"comment":"The subleading moment formula is not valid as written. It contains Γ(n−1), which diverges at n=1, yet the formula is presented for all moments without restriction. Expanding Eq. (11) directly at n=1 gives a vanishing O(Ξ^{-2}) correction to ⟨x_s^2⟩: the relevant Watson coefficient is an integral whose leading r' contribution cancels. Thus Eq. (15) must be restricted to n≥2, and the n=1 case must be stated separately. The resummed characteristic function in Eq. (16) is consistent with a zero n=1 correction, so the leading Gaussian result Eq. (14) survives, but the claim of explicit moment formulas for all moments is inaccurate as printed.","section":"Transition and higher order corrections, Eq. (15)"},{"comment":"The proof neglects the connected contribution R_s^{(c)}(k) with the statement that it is 'globally decaying with Ξ'. Since all later asymptotic results are obtained from the disconnected integral only, a bound or decay estimate for R_s^{(c)} is needed to justify the claim. In particular, the text should specify the order in Ξ at which the connected part first contributes, so that Eqs. (15)–(20) are not applied beyond their proven range.","section":"Derivation of the Universal Gaussian, Eqs. (6)–(7)"},{"comment":"The replacement F_U ≃ exp(−πΞ q') is derived in Appendix A in the limit of infinitely many channels with T_c ≃ 1/M. The experimental validation, however, uses a finite system (M=52) with two strong channels T_1=T_2=0.967. The paper asserts that additional terms in the channel-factor expansion contribute only at higher orders, but this is not demonstrated. Please give the first nonvanishing correction to Eq. (10) and show explicitly, e.g., by expanding the correction factor as 1+O(q'^2) and using the q'^{n-1} prefactor in Eq. (11), that it does not affect the orders quoted in Eqs. (15)–(20) for the parameters used in the experiment.","section":"Appendix A and Eq. (10)"}],"minor_comments":[{"comment":"State explicitly that the moment formula applies for n≥1; the zeroth moment is treated separately.","section":"Eq. (11)"},{"comment":"The sentence 'The explicit calculation is given in Appendix A' appears to refer to Appendix B, where the Gaussian derivation is carried out.","section":"After Eq. (12)"},{"comment":"The phrase 'globally decaying with Ξ' should be quantified, e.g., as exponentially small in Ξ, rather than left qualitative.","section":"Connected part, Eqs. (6)–(7)"}],"recommendation":"major_revision","confidential_remarks":"The n=1 issue in Eq. (15) is localized and apparently fixable; it does not undermine the leading-order Gaussian. The connected-part and channel-factor points require additional justification but do not appear fatal. I expect the paper can be accepted after a revision that corrects Eq. (15) and supplies the missing estimates."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"To my mind this paper is the real thing: a first derivation of the Ericson-regime Gaussian for β=2 from the exact supersymmetry characteristic function, not another heuristic or correlator-level argument. The starting point, Eq. (4), comes from earlier supersymmetry work and does not assume the Gaussian outcome. The Gaussian appears as the leading term of a genuine 1/Ξ expansion via Watson's lemma, and the subleading corrections are tested against a microwave experiment and Monte Carlo. If the expansion is right—and the numerics and experiment support it—this resolves a sixty-year-old question for the unitary case.\n\nWhat is new here: the asymptotic expansion itself, the explicit all-orders moment formulas, and the correction terms that describe the transition into Ericson behavior. Prior treatments treated the Gaussian as phenomenological or only handled correlators. The leading-order Gaussian (Eq. (14)) and the resummed characteristic function (Eq. (16)) are consistent.\n\nSoft spots, in proportion. Eq. (15) as printed is wrong at n=1: Γ(n−1) diverges while the second moment is finite. The stress-test note gives the direct expansion of Eq. (11) at n=1, which yields a vanishing O(Ξ^{-2}) correction. So Eq. (15) is valid for n≥2 only and the n=1 case must be listed separately. That is a concrete error in a displayed formula, but it does not break the central argument, because Eq. (16) is consistent with a zero n=1 correction. The authors need to correct the 'explicit formulae' claim.\n\nTwo smaller caveats. The abstract overclaims scope: only β=2 is treated; β=1,4 are deferred to future work and that should be stated up front. And the channel-factor approximation Eq. (10), derived in the large-M, T_c ~ 1/M limit, is applied to a finite system with a few strong channels. The stress-test note argues that the leading channel-factor corrections cancel in the relevant integral, so this is probably less damaging than a first read suggests, but the regime of validity deserves one clear sentence.\n\nBottom line: this is a serious piece of work for statistical scattering theory and random-matrix people. It deserves a real referee. The n=1 issue is a fixable blemish; I would want it corrected before publication, but I would not desk-reject this.","headline":"A credible first derivation of the Ericson-regime Gaussian from the supersymmetry characteristic function; Eq. (15) needs a fix at n=1 but the main result holds.","tokens_in":13692,"tokens_out":3025,"would_cite":true,"duration_ms":29791,"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 paper proves that off-diagonal S-matrix elements in quantum chaotic scattering become Gaussian distributed in the Ericson regime, with explicit 1/Ξ corrections.","keywords":["Ericson regime","stochastic scattering","random matrix theory","supersymmetry","Gaussian universality","scattering matrix","quantum chaos","microwave networks"],"falsifier":"Compute the fourth moment of Re S_21 from the exact supersymmetry integral (i.e., with F_U unapproximated) for a two-channel system with T_1 = T_2 = 0.967 at Ξ = 1.424, and compare to the paper's Eq. (15) with n = 2. A statistical deviation beyond the stated O(1/Ξ^2) would show that the exp(−πΞq′) replacement fails at subleading order.","tokens_in":12718,"feed_emoji":"⚛️","tokens_out":7873,"duration_ms":70968,"temperature":0.7,"pith_summary":"At low energy, resonances in quantum scattering are isolated; as energy rises they overlap, and in the Ericson regime the cross section becomes a random function. For sixty years it was believed that in this regime the off-diagonal elements of the scattering matrix follow a universal Gaussian distribution, but no first-principles derivation existed. This paper derives that Gaussian analytically for time-reversal non-invariant (β=2) systems, starting from the exact supersymmetry integral representation of the characteristic function. An asymptotic expansion in powers of 1/Ξ — the ratio of average resonance width to mean level spacing — yields the Gaussian at leading order and explicit subleading corrections describing the transition. The results match microwave network experiments and Monte Carlo simulations, and the same method promises to extend to the time-reversal invariant cases.","feed_headline":"Ericson's universal Gaussian proven for quantum chaotic scattering","feed_subtitle":"Asymptotic expansion of the exact supersymmetry integral yields the 60-year-old Ericson result plus finite-overlap corrections.","key_machinery":"The load-bearing object is the exact characteristic function R_s(k) of the off-diagonal S-matrix element distribution, expressed as a two-dimensional integral over hermitian/antisymmetric supermatrix variables. The derivation proceeds by (i) a change of variables that removes the q′−2 singularity associated with the supermatrix Berezin integration, (ii) replacing the channel factor F_U by exp(−πΞq′) in the limit of many channels with T_c ≈ 1/M, and (iii) applying Watson's lemma to convert the integrals into an asymptotic series in 1/Ξ. The leading term gives a Gaussian characteristic function exp(−k^2/[2πΞ(g_a^++1)(g_b^++1)]), whose inverse Fourier transform is the universal Gaussian.","core_discovery":"The paper establishes that, in the Ericson regime, the rescaled real and imaginary parts ξ_s = √Ξ x_s of an off-diagonal scattering matrix element S_ab (a ≠ b) have the leading-order distribution P_s^(l)(ξ_s) = sqrt((g_a^+ + 1)(g_b^+ + 1)/2) exp(−π(g_a^+ + 1)(g_b^+ + 1) ξ_s^2/2), a zero-mean Gaussian whose variance is determined by the transmission coefficients through g_c^+ = 2/T_c − 1. This is not a central-limit argument; it follows from an asymptotic evaluation of the exact characteristic function integral obtained in the supersymmetry formulation of stochastic scattering. The next-order term P_s^(sl)(ξ_s) is a quartic polynomial times the same Gaussian divided by Ξ, with sign set by 1 −","pith_inferences":["The same asymptotic strategy applied to the β=1 (orthogonal) and β=4 (symplectic) cases would test whether the Gaussian Ericson distribution is common to all three Wigner-Dyson symmetry classes; the authors say the method carries over, but the technical work is left for future publications.","Because the channel-factor approximation exp(−πΞq′) is exact only for infinitely many weak channels, a careful numerical check of the subleading formulas against the exact F_U for few strong channels could sharpen the range of validity; this is a direct, testable extension of the paper's results.","The 'universality emerging in a universality' framing suggests that other statistics of the scattering matrix — such as Wigner-Smith time delays, conductance statistics, or shot-noise power — might be re-derived by the same asymptotic route, potentially revealing further universal layers.","The proof explicitly avoids a Central Limit Theorem mechanism, deriving the Gaussian from the analytic structure of the supersymmetry integral; that distinction may matter for physical intuition, since it implies the Gaussian is tied to the Ericson large-Ξ limit rather than to the addition of many independent contributions."],"forward_implications":["In the Ericson regime, both real and imaginary parts of any off-diagonal S-matrix element are Gaussian with variance fixed by the transmission coefficients; the 1960 heuristic is now a proven consequence of the random-matrix/supersymmetry framework.","The explicit 1/Ξ corrections give the full transition: for Ξ ≈ 1 the distributions are visibly non-Gaussian, with a shift at zero proportional to 3(g_a^+ g_b^+ − g_a^+ − g_b^+ − 3)/(8πΞ) √(2(g_a^++1)(g_b^++1)).","The cross-section distribution is exponential in the Ericson regime, with a correction that explains observed deviations from the pure exponential at σ=0; normalization yields ⟨σ_ab⟩ = 2/(πΞ(g_a^++1)(g_b^+1)) + O(1/Ξ^2).","The convergence to the Ericson limit is fast: the subleading term suffices already at Ξ ≈ 1.4, as demonstrated by small deviations in Monte Carlo simulations using the full problem."],"fun_headline_variants":["60-year Ericson regime finally derived","Universal Gaussian in chaotic scattering: proven and tested","Ericson transition: exact theory for quantum chaos","Overlapping resonances yield universal Gaussian — derived","From isolated to Ericson: quantum scattering solved"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The derivation approximates the exact channel factor by exp(−πΞq′) in the limit of infinitely many weak channels, and then applies the resulting asymptotic formulas to finite systems with a few strong channels; if the neglected finite-channel corrections were comparable to the claimed subleading terms, the explicit correction formulas would change.","fun_headline_variants_meta":{"raw":{"variants":["60-year Ericson regime finally derived","Universal Gaussian in chaotic scattering: proven and tested","Ericson transition: exact theory for quantum chaos","Overlapping resonances yield universal Gaussian — derived","From isolated to Ericson: quantum scattering solved"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000158,"raw_usage":{"total_tokens":1043,"prompt_tokens":707,"completion_tokens":336,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":451,"completion_tokens_details":{"reasoning_tokens":267}},"tokens_in":451,"tokens_out":336,"duration_ms":3680,"temperature":1.0,"reasoning_tokens":267,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T05:48:03.552198+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the fourth moment of Re S_21 from the exact supersymmetry integral (i.e., with F_U unapproximated) for a two-channel system with T_1 = T_2 = 0.967 at Ξ = 1.424, and compare to the paper's Eq. (15) with n = 2. A statistical deviation beyond the stated O(1/Ξ^2) would show that the exp(−πΞq′) replacement fails at subleading order.","supporting_citations":[],"review_version":1}