{"id":"47ae7092-a1c3-44cf-8d59-81e69810a367","arxiv_id":"2603.24190","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Chaotic nonlinear dynamics on a stratified social-network Hamiltonian thermalizes to Rayleigh-Jeans distributions whose condensation and Lorenz curves match observed wealth inequality, and yields KZ-like turbulence under pumping.","lead":"A nonlinear oscillator model on a real scientific-collaboration network, with diagonal energy stratification and weak nonlinearity, dynamically thermalizes to a Rayleigh-Jeans distribution whose low-energy condensation produces Lorenz curves resembling world wealth inequality. With low-mode pumping and high-mode absorption the same system shows algebraic Kolmogorov-Zakharov-like spectra.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"Incomplete thermalization of high-energy modes undermines the quantitative Lorenz-curve match claimed for the dynamical model.","rationale":"The reader correctly flags the E_m ↔ wealth identification as an uncalibrated modeling premise and therefore assigns CONDITIONAL. That premise is indeed soft, yet it is an interpretive layer rather than an internal inconsistency of the dynamical claim. The more immediate load-bearing weakness is that the dynamical claim itself—full RJ thermalization of the stratified sparse-network Hamiltonian—is only partially realized in the numerics of Sec. III. High-energy modes remain under-populated, so the Lorenz curves that are said to “match” world inequality either come from pure RJ theory (with free ε) or from incomplete numerical distributions that require ad-hoc bandwidth cuts. This does not overturn the mathematical/numerical core (chaos \to RJ for accessible modes, KZ-like spectra under pumping), but it does reinforce the reader’s call for more prominent caveats and keeps the verdict at CONDITIONAL rather than ACCEPT. The concrete test isolates whether the dynamical trajectories alone, without free parameters, already produce the reported inequality measures.","tokens_in":22793,"tokens_out":642,"duration_ms":6678,"concrete_test":"Recompute the Lorenz curves of Fig. 10 bottom panel using only the numerical \rho_m of the fully thermalized low-energy runs (m0 = 20, eta = 4, t = 2^30) without any post-hoc ε rescaling or bandwidth truncation; if the resulting Gini and the 2 %-wealth / 10 %-rich fractions deviate by more than ~15 % from the world-data benchmarks quoted in the abstract and Sec. V, the claimed quantitative match fails.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The strongest claim rests on the assertion that Hamiltonian dynamics (Eqs. 1–2, 8) produce the RJ distribution whose low-energy condensation then yields Lorenz curves comparable to world wealth data (Sec. V, Fig. 10). Section III and Figs. 5–7 show that this is only partially true: for initial modes with E_m0 > E_th (E_th ≈ 1–2.2 depending on eta) the numerical \rho_m remain far from the theoretical RJ curve even at t = 2^30, and the four highest modes cannot thermalize at all because energy conservation forbids the IPR growth needed for delocalization. The bottom panel of Fig. 10 therefore either (i) uses a well-thermalized low-energy state whose Gini (G = 0.46) is milder than world data or (ii) matches Gini by hand-adjusting ε after discarding the still-unthermalized high-energy tail. Consequently the quantitative similarity to real wealth inequality is not a direct dynamical prediction of the stratified network Hamiltonian; it is obtained only after selective use of the RJ formula or after free choice of the reduced energy ε.","agreement_with_reader":"partial"},"referee_report":{"model":"grok-4.5","summary":"The paper introduces a Hamiltonian model of social stratification (SSS) in which N oscillators are coupled by a real sparse social-network adjacency matrix, a diagonal disorder term that produces an approximately flat density of states (wealth layers), and a cubic nonlinearity. With two conserved quantities (energy and norm), the authors argue that above a chaos border the dynamics thermalizes to the Rayleigh–Jeans distribution ρ_m = T/(E_m − μ). At low total energy this distribution exhibits RJ condensation at the lowest modes; the resulting Lorenz curves are compared with world wealth inequality. A driven-dissipative variant with pumping at low modes and absorption at high modes is shown to produce an algebraic spectrum reminiscent of Kolmogorov–Zakharov turbulence. Numerical evidence is given for entropy growth, approach of mode occupations to RJ for selected initial conditions, and Lorenz/Gini statistics.","tokens_in":23134,"tokens_out":1745,"duration_ms":25801,"significance":"If the dynamical claims hold, the work supplies a concrete, numerically accessible example of RJ thermalization and condensation on a sparse, stratified social-network Hamiltonian rather than on RMT or multimode-fiber models. That is a useful addition to the dynamical-thermalization literature (FPUT, NLIRM, optical fibers). The KZ-like cascade on the same stratified network is a natural extension of earlier RMT turbulence work. The social-wealth interpretation is more speculative and rests on an uncalibrated identification of linear eigen-energies with household wealth layers; its value is mainly as a provocative analogy that links RJ condensation to Lorenz curves, not as a calibrated economic theory. The numerical documentation of energy-shift effects and incomplete high-energy thermalization is honest and useful for the field.","major_comments":[{"comment":"Sec. III and Figs. 5–7: For initial modes with E_m0 ≳ E_th (≈ 1–2.2 depending on β) the numerical ρ_m remain far from the theoretical RJ curve even at t = 2^30, and the four highest modes cannot thermalize at all because energy conservation forbids the IPR growth required for delocalization. The abstract and Sec. V nevertheless present the RJ distribution (and its Lorenz curves) as the outcome of the stratified-network Hamiltonian dynamics. The manuscript should state more sharply which claims are supported by fully thermalized trajectories and which rest only on the theoretical RJ formula evaluated at a chosen total energy.","section":"Section III, Figs. 5-7"},{"comment":"Sec. V, bottom panel of Fig. 10 and caption: The quantitative similarity to world wealth data (G ≈ 0.8–0.9, ~50% of population owning ~2% of wealth) is obtained either from the theoretical RJ distribution at a freely chosen reduced energy ε = w_s/B, or by hand-matching Gini after discarding the still-unthermalized high-energy tail of a dynamical run (m0 = 1, ε_num = 0.16 rescaled to 0.24). The well-thermalized dynamical state (m0 = 20) yields only G = 0.46 and milder poor/rich fractions. The claim that the stratified Hamiltonian dynamics produces Lorenz curves comparable to world data is therefore not a direct, parameter-free prediction of the simulated trajectories; the free choice of ε (or of the effective bandwidth) should be acknowledged as such and the dynamical versus theoretical contributions separated.","section":"Section V, Fig. 10"},{"comment":"Abstract, Introduction and Sec. II.B: The identification of the linear eigen-energies E_m of H = D + f(A + κ H_GOE) with wealth layers of real households, and of the cubic nonlinearity plus sparse collaboration links with economic interactions, is stated as a modeling premise without independent empirical calibration of the map E_m ↔ wealth. This is an axiom of the paper, not a derived result. The social-stratification language in the abstract and title should be framed explicitly as an analogy whose quantitative success depends on the free parameter ε, rather than as a demonstrated dynamical explanation of real wealth inequality.","section":"Abstract, Introduction, Sec. II.B"},{"comment":"Sec. IV and Fig. 9: The algebraic exponent extracted for the SSS cascade is s0 = 1.52 ± 0.01, well above the KZ value s0 = 1 obtained for the RMT case. The text attributes the discrepancy to smaller N and to the relatively local energy couplings induced by f = 0.1. Given that the section is titled “KZ like turbulence” and the abstract claims “features of the Kolmogorov–Zakharov turbulence,” the manuscript should either demonstrate closer approach to s0 = 1 under controlled variation of N and f, or qualify the claim more carefully as a cascade with an algebraic spectrum that is only qualitatively KZ-like.","section":"Section IV, Fig. 9"}],"minor_comments":[{"comment":"Sec. II.A: “Hamilonian” → “Hamiltonian”.","section":"Section II.A"},{"comment":"Sec. II.C: “m = N, N−1, M−1, N−3” appears to be a typo for N−1, N−2, N−3.","section":"Section II.C"},{"comment":"Sec. VI: “the dynamical the RJ thermal distribution” → remove the extra “the”.","section":"Section VI"},{"comment":"Fig. 1 caption and text: density of states is shown for one random realization; a brief statement that other realizations give quantitatively similar ν(E) would strengthen reproducibility (already mentioned in the text but easy to miss).","section":"Figure 1"},{"comment":"The chaos border β_ch is never estimated for the SSS model (only asserted that β = 2, 4 lie above it). A short Lyapunov-exponent or entropy-production scan versus β would make the “above chaos border” claim more concrete, even if a precise border is hard to pin down.","section":"Section II.B / III"},{"comment":"Notation: both the conserved Hamiltonian energy and the linear energy ∑_m E_m ρ_m are called E in places; the paper already distinguishes them in Sec. III, but a consistent symbol (e.g. E_lin) throughout would help.","section":"Section III"}],"recommendation":"major_revision","confidential_remarks":"The physics core (RJ thermalization on a sparse stratified network, energy-shift effects, partial high-energy freezing) is publishable and honestly reported. The social-wealth framing is the main risk for the journal: it is an uncalibrated analogy that leans on free choice of ε to match empirical Gini. I would not reject on that ground alone if the authors reframe the social claims as analogy and cleanly separate dynamical results from theoretical RJ Lorenz curves; major revision is the right bar. Fit is reasonable for a statistical-mechanics journal that already publishes dynamical thermalization and wave-turbulence work; the social language should not be allowed to outrun the numerics."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The new piece is the SSS Hamiltonian: real Newman collaboration adjacency plus a strong diagonal stratification term (W=8) and weak GOE smoothing, plus the cubic nonlinearity. They show that even with sparse links and modest IPR this still thermalizes to the Rayleigh-Jeans distribution fixed by the two conserved quantities, and that pumping/absorption produces a KZ-like cascade on the same network. That is a genuine, if incremental, extension of their earlier RMT, fiber, and pure-network papers. The numerics are careful: symplectic integrator, honest reporting of energy-shift effects, incomplete high-mode thermalization, and entropy approach (Figs. 4–7). Citations to the FPUT literature, optical-fiber RJ work, and their own prior results are appropriate and not decorative.\n\nThe soft spot is exactly where the stress-test points. High-energy modes (E_m0 above roughly 1–2) remain far from RJ even at t=2^30, and the top four modes cannot delocalize at all under energy conservation. The Lorenz curves that look like world wealth data therefore come either from well-thermalized low-energy states (Gini ~0.46, milder than the world) or from hand-matching ε after discarding the unthermalized tail. The map E_m ↔ wealth layers is a modeling premise, not calibrated. So the quantitative similarity to inequality is not a direct dynamical output of the stratified network; it is an analogy that still requires free choice of reduced energy. That does not kill the paper, but it should be stated more prominently than it is.\n\nThis is for people already working on dynamical thermalization, wave turbulence, or econophysics who want a concrete sparse-network example with stratification. It is not a first-principles theory of inequality. Math and numerics look solid enough that a serious editor should send it to referees rather than desk-reject; expect requests to foreground the incomplete-thermalization and free-ε caveats. I would bring it to reading group if we are discussing RJ condensation or social-network Hamiltonians; I would cite the thermalization-with-stratification result if I needed that specific construction, but not the wealth claim itself.","headline":"Solid incremental numerics on RJ thermalization with a stratified sparse network; the wealth-inequality claim is an analogy that needs free ε, not a first-principles prediction.","tokens_in":23701,"tokens_out":548,"would_cite":false,"duration_ms":7734,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Nonlinear chaos on a stratified social network thermalizes agents to a Rayleigh-Jeans wealth distribution whose low-energy condensate matches observed household inequality.","keywords":["social stratification","Rayleigh-Jeans thermalization","wealth inequality","Lorenz curve","Kolmogorov-Zakharov turbulence","nonlinear oscillators","dynamical chaos","condensate"],"falsifier":"Compute Lorenz curves from the model’s steady-state occupations for a range of total energies and check whether their Gini coefficients and the fractions of population that own 2 % and 75 % of total wealth remain quantitatively close to the World Inequality Report numbers; a systematic mismatch would falsify the claimed correspondence.","tokens_in":23680,"feed_emoji":"⚖️","tokens_out":818,"duration_ms":8615,"temperature":0.7,"pith_summary":"The paper builds a Hamiltonian of agents whose linear couplings come from a real scientific-collaboration network plus a diagonal stratification term that sets wealth-like energy levels, and whose interactions are cubic. Because the dynamics conserve total energy and total probability norm, once the nonlinearity exceeds a chaos threshold the system thermalizes to the Rayleigh-Jeans occupation law fixed by those two integrals. At low total energy the law produces a condensate that piles most of the population into the poorest modes while a thin high-energy tail holds most of the wealth—exactly the pattern seen in global Lorenz curves. Adding weak pumping at the bottom and absorption at the top generates a steady Kolmogorov–Zakharov-like cascade, offering a dynamical caricature of wealth flowing from workers to oligarchs. The concrete numerical match between the model’s Gini coefficients and real-country data is what makes the construction more than a formal analogy.","feed_headline":"Chaos on social networks produces Rayleigh-Jeans wealth inequality","feed_subtitle":"Low-energy condensate piles most people into the poorest modes, matching world Gini data","key_machinery":"The Rayleigh-Jeans distribution ρ_m = T/(E_m − μ) determined solely by the two integrals of motion (total energy and unit norm). It is the micro-canonical equilibrium of the chaotic oscillator system and the object whose low-temperature condensation directly produces the observed wealth inequality.","core_discovery":"Above a chaos border the stratified social-network Hamiltonian (adjacency matrix of a real collaboration network plus diagonal wealth levels plus cubic nonlinearity) undergoes purely dynamical thermalization to the Rayleigh-Jeans distribution fixed by the two conserved quantities; the resulting low-energy condensate yields Lorenz curves whose Gini values and poor/rich fractions are comparable to world wealth statistics.","pith_inferences":["If the energy–wealth map is only approximate, the same thermalization mechanism could still generate inequality on any stratified network whose spectrum is roughly flat, suggesting the result is robust to the precise choice of social graph.","Empirical time series of household wealth mobility could be compared with the model’s relaxation rate ~β² to test whether real economies sit above or below the chaos border.","Replacing the scientific-collaboration network by a denser social graph (e.g., online platforms) should lower the chaos threshold and accelerate condensation, a prediction open to numerical check."],"forward_implications":["Low total societal energy (wealth) necessarily produces a large poor condensate and high Gini, independent of microscopic network details once chaos is present.","A steady wealth cascade from low to high layers can be sustained by continuous injection at the bottom and absorption at the top, reproducing Kolmogorov–Zakharov-like spectra.","Lorenz curves of the model are controlled by a single dimensionless ratio of total wealth to spectral bandwidth, offering a one-parameter description of national inequality.","Negative-temperature states (energy above the spectral center) are dynamically allowed and would invert the inequality pattern, concentrating wealth at the richest modes."],"fun_headline_variants":["Social-network chaos thermalizes to Rayleigh-Jeans wealth condensate","Stratified oscillators hit chaos border and form RJ low-energy wealth pileup","Network Hamiltonian chaos yields RJ distribution matching world wealth Gini","Dynamical thermalization on social links creates RJ wealth inequality","Energy-stratified network chaos drives RJ condensate like global wealth gaps"],"cache_read_input_tokens":16512,"weakest_assumption_plain":"The linear eigen-energies of the adjacency-plus-diagonal matrix can be identified with real household wealth layers, and the cubic nonlinearity plus sparse social links constitute a faithful dynamical model of economic interactions.","fun_headline_variants_meta":{"raw":{"variants":["Social-network chaos thermalizes to Rayleigh-Jeans wealth condensate","Stratified oscillators hit chaos border and form RJ low-energy wealth pileup","Network Hamiltonian chaos yields RJ distribution matching world wealth Gini","Dynamical thermalization on social links creates RJ wealth inequality","Energy-stratified network chaos drives RJ condensate like global wealth gaps"]},"model":"grok-4.5","effort":"low","cost_usd":0.0042,"raw_usage":{"total_tokens":1228,"prompt_tokens":696,"num_sources_used":0,"completion_tokens":71,"cost_in_usd_ticks":42000000,"prompt_tokens_details":{"text_tokens":696,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":461,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":696,"tokens_out":71,"duration_ms":5068,"temperature":1.0,"reasoning_tokens":461,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-13T19:02:00.666897+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Compute Lorenz curves from the model’s steady-state occupations for a range of total energies and check whether their Gini coefficients and the fractions of population that own 2 % and 75 % of total wealth remain quantitatively close to the World Inequality Report numbers; a systematic mismatch would falsify the claimed correspondence.","supporting_citations":[],"review_version":1}