{"id":"b4326b16-f1f0-43f3-aa76-137860070f26","arxiv_id":"1908.01689","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"This paper reports superdiffusive energy spreading in 2D momentum-conserving nonlinear lattices, described by Lévy-stable distributions, with momentum diffusion no longer superdiffusive.","lead":"Large computer simulations show that energy spreads through two-dimensional momentum-conserving lattices faster than ordinary diffusion, meaning heat conduction in such systems does not follow the simple Fourier law. This matters for understanding thermal transport in thin materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Reported γ=1.818 and MSD β=1.27 for the 2D quartic lattice violate the paper's own Lévy relation β=3−γ (predicted 1.18), so the two diagnostics do not independently confirm the Lévy-stable description.","rationale":"I considered the reader's finite-time t ln t objection, which is real and explicitly acknowledged in Sec. IIIB. But it only affects whether the conductivity divergence is logarithmic or power-law; it does not threaten the central 'anomalous' classification. The γ−β mismatch is more load-bearing for the abstract's stronger 'well described by Lévy-stable distribution' claim. It is an internal contradiction with the paper's own relation in Sec. IIIB, visible in the reported numbers, and it specifically affects the quartic case. The paper deserves credit for the 1D benchmark (γ≈3/2), the honest admission about t ln t, the φ4 normal-diffusion control, and the side-peak disappearance observation. Those supports are independent of my concern. The proposed test would settle whether the inconsistency is a transient or a model failure. Because the qualitative anomalous conclusion and the FPU analysis likely survive, I do not recommend rejecting the paper; I keep the reader's CONDITIONAL, with the additional condition of reporting this consistency check and error bars.","tokens_in":20924,"tokens_out":9235,"duration_ms":98252,"concrete_test":"Use the raw Ny=1024 quartic correlation data to perform one joint asymptotic test: extend simulations to t≈2000 and, on the same time window, (i) fit the full rescaled profile t^{1/γ}ρ_E(i/t^{1/γ},t) with γ free (not only the peak) and (ii) fit MSD β; then test whether β approaches 3−γ. Also bootstrap over independent equilibrium runs to obtain confidence intervals for both exponents. If β remains ≈1.27 while γ≈1.82, the Lévy-stable description fails for the quartic lattice; if β drifts toward ≈1.18 and γ stays ≈1.82, the original exponents were finite-time transients and the reader's longer-time condition is met.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The quantitative core of the abstract is that 2D quartic energy diffusion is 'well described by the Lévy-stable distribution.' The paper supports this with two diagnostics. Sec. IIIA (Fig. 4) fits the peak-height decay with 1/γ=0.55, i.e. γ=1.818 for Ny=1024. Sec. IIIB (Fig. 7) fits the same data's MSD as t^β with β=1.27. According to the Lévy-walk relation stated in Sec. IIIB, β=3−γ, which predicts β=1.182 for γ=1.818. The observed β=1.27 lies above this by ~0.09. For FPU-β the same check is consistent (γ=1.504, predicted β=1.496, measured 1.51), so the inconsistency is specific to the case where the paper also admits the t^1.27 vs. t ln t ambiguity. No error bars are given, so one cannot tell whether 0.09 is a finite-time transient or a real breakdown of the Lévy-stable model. If the latter, the central-peak Lévy fit and the MSD analysis are not independent confirmations; the second moment is then dominated by tails not captured by the fitted γ, and the claim 'well described by Lévy-stable distribution' is not established.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies energy and momentum diffusion in two-dimensional nonlinear lattices (purely quartic, FPU-β, and φ4) by computing spatiotemporal correlation functions of energy and momentum fluctuations in equilibrium. For the momentum-conserving lattices, the authors extract a scaling exponent γ from the decay of the central peak height of the energy correlation and report superdiffusive values γ=1.818 (2D quartic) and γ=1.504 (2D FPU-β), both in the range 1<γ<2, and they claim that the rescaled profiles are well described by Lévy-stable distributions. The mean-square displacement (MSD) of the energy distribution grows as t^β with β=1.27 (quartic) and β=1.51 (FPU-β), supporting anomalous diffusion, while for the φ4 model the MSD grows linearly, indicating normal diffusion. The authors also find that momentum diffusion, superdiffusive in 1D, becomes normal or subdiffusive when the width Ny is increased to 1024, which contradicts the 1D hypothesis that anomalous heat transport is accompanied by momentum superdiffusion. They conclude that energy diffusion in 2D momentum-conserving lattices is anomalous and infer possible logarithmic (quartic) or power-law (FPU-β) divergence of thermal conductivity.","tokens_in":21220,"tokens_out":6310,"duration_ms":57320,"significance":"If the conclusions hold, this is one of the first direct numerical characterizations of energy diffusion in 2D momentum-conserving lattices using fluctuation correlation functions, and it clarifies that the 1D momentum-superdiffusion criterion does not extend to 2D. The 1D benchmarking against the nonlinear fluctuating hydrodynamics prediction γ=3/2 is a useful validation, and the use of two diagnostics (peak-height scaling and MSD) is a strength. However, the quantitative support for the Lévy-stable description is incomplete: for the quartic lattice the measured β=1.27 is inconsistent with the stated Lévy relation β=3−γ (which would give 1.18), and the paper explicitly acknowledges that t^1.27 and t ln t cannot be distinguished over the simulated time window. These issues materially affect the quantitative conclusions, although the qualitative anomalous-diffusion verdict appears robust.","major_comments":[{"comment":"For the 2D quartic lattice, the Lévy-walk relation β=3−γ stated in Sec. IIIB predicts β≈1.18 for the measured γ=1.818 (Fig. 4), whereas the reported MSD exponent is β=1.27 (Fig. 7). The paper does not address this ~0.09 discrepancy, and no error bars are provided, so it is unclear whether this is a finite-time transient or a genuine breakdown of the Lévy-stable description for the quartic case. As written, the MSD result does not independently confirm the abstract claim that energy diffusion is \"well described by the Lévy-stable distribution.\"","section":"Sec. IIIB, Fig. 7"},{"comment":"The paper states that for the quartic lattice \"the discrepancy in the scaling behavior between the power-law ∼ t^1.27 and the function of ∼ t ln t is so small that it is difficult to numerically determine\" which is correct, and later \"we cannot exclude that this may be a numerical coincidence.\" Given this admitted degeneracy, the fitted exponent β=1.27 and the inferred power-law divergence of thermal conductivity for the quartic lattice are not established; the text should clearly frame these as tentative and explicitly discuss the log-corrected alternative as equally compatible with the data.","section":"Sec. IIIB, Fig. 7"},{"comment":"The collapse of the rescaled correlation functions in Figs. 5 and 6 is performed using the same exponent γ that was extracted from the peak-height decay in Fig. 4. Consequently, the visible collapse is a self-consistency check rather than an independent validation of the Lévy-stable distribution; an independent test would require fitting the full profiles without inputting γ, or at least reporting a goodness-of-fit measure for the Lévy curves shown as the solid lines.","section":"Sec. IIIA, Figs. 5-6"},{"comment":"No error bars, confidence intervals, or fitting-window information is provided for any of the reported exponents (γ, β, µ). Because the paper's quantitative claims rest on small differences (e.g., γ=1.818 versus 1.504, β=1.27 versus t ln t, and the 1D-to-2D crossover), uncertainty quantification is needed to assess whether these differences and the dimensional crossover are statistically significant.","section":"Sec. IIIA-III C, Figs. 4, 7-9, 12"}],"minor_comments":[{"comment":"The sentence \"the scaling exponent γ reaches γ = 1.818 for the 1D purely quartic lattice\" should read \"for the 2D purely quartic lattice\", since the preceding sentence discusses the 1D values and the following values are for Ny=1024.","section":"Sec. IIIA, text after Fig. 6"},{"comment":"There are several typographical errors, including \"we we\" and \"of of\", and the notation in Eq. (4) is confusing because the left-hand side is written with a time-dependent argument; please rewrite the scaling relation for clarity.","section":"Sec. IIB and Sec. IIIB"},{"comment":"The figure captions do not state the integration time step, the number of ensemble realizations used for each width, or the fit ranges for the power-law fits in Fig. 4; adding these details would improve reproducibility.","section":"Fig. 1-3 captions"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is honest about the t ln t ambiguity, which is a plus. The main issue is that the Lévy-stable description is over-claimed for the quartic case given the β vs. 3−γ inconsistency; a revision that reconciles the two diagnostics or appropriately softens the claim would be acceptable. No concerns about novelty or authorship."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Mark,\n\nThis paper is worth a serious referee, but I don't think the Lévy-stable claim for the quartic lattice is as clean as the abstract implies.\n\nWhat is actually new: this is the first fluctuation-correlation study of energy and momentum diffusion in 2D momentum-conserving lattices, and it makes two observations that matter. The heat-mode side peaks disappear as the width grows, suggesting weaker heat-sound coupling in 2D. And while energy diffusion stays anomalous, momentum diffusion crosses from superdiffusive in 1D to normal/subdiffusive in 2D. That decoupling challenges the 1D momentum-superdiffusion criterion, and it should get cited. The 1D benchmarks look right—quartic gamma ≈ 3/2 and FPU ≈ 1.45 are consistent with nonlinear fluctuating hydrodynamics—and the phi^4 control shows normal diffusion as expected.\n\nThe soft spot is internal consistency of the quartic result. The paper uses beta = 3 − gamma from Lévy-walk theory, then reports gamma = 1.818 and beta = 1.27 for the 2D quartic lattice. Those imply beta = 1.18, about 0.09 below the measured value. No error bars are given, so one can't tell whether the gap is finite-time drift or something structural. The authors themselves say t^1.27 and t ln t cannot be distinguished and that they cannot exclude a numerical coincidence. Separately, the rescaled profiles in Figs. 5 and 6 use the same gamma extracted from the peak-height decay, so the collapse is a consistency check, not independent confirmation of the Lévy form. The FPU-beta case is internally consistent (gamma 1.504, beta 1.51) and stronger. The qualitative verdict for quartic—anomalous energy diffusion—survives because both diagnostics sit clearly above normal; the quantitative Lévy description is what isn't established.\n\nIf I were refereeing I'd ask for error bars and explicit fit windows, a convergence test for gamma and beta as a function of simulation time, and a decision on t^1.27 versus t ln t for quartic. I'd also soften the conclusion's \"can ensure anomalous heat conduction\" wording, since the authors rightly note there is no 2D connection theory. The citations look fair and complete. This one is aimed squarely at the low-dimensional heat transport and nonlinear fluctuating hydrodynamics crowd; they'll get value from the decoupling result even if the exact exponents remain unresolved.\n\nSend it to review, conditional on revision. I'd bring it to group; the tension between the two diagnostics is a good discussion.","headline":"Worth refereeing: the 2D energy-vs-momentum decoupling is a real observation, but the quartic lattice's Lévy-stable claim is not quantitatively established.","tokens_in":21745,"tokens_out":3701,"would_cite":true,"duration_ms":33675,"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":"Energy diffusion in 2D momentum-conserving nonlinear lattices is anomalous and follows Lévy-stable distributions.","keywords":["anomalous heat transport","energy diffusion","Lévy-stable distribution","two-dimensional nonlinear lattices","momentum conservation","Fermi-Pasta-Ulam lattice","thermal conductivity divergence","fluctuation correlation function"],"falsifier":"Extend the equilibrium simulation of the purely quartic 2D lattice (width 1024) to correlation times beyond $t=200$ and compare the MSD $\\langle \\Delta x^2(t)\\rangle_E$ against $t^{1.27}$ and $t\\ln t$: if the data bend toward $t\\ln t$, the power-law diffusion exponent claimed for the quartic lattice is a finite-window artifact rather than the asymptotic law.","tokens_in":20713,"feed_emoji":"🔥","tokens_out":9537,"duration_ms":79779,"temperature":0.7,"pith_summary":"The paper aims to settle whether heat transport in two-dimensional momentum-conserving nonlinear lattices is anomalous by simulating energy diffusion directly with fluctuation correlation functions, bypassing finite-size conductivity simulations. It reports that energy spreads superdiffusively in the 2D purely quartic and FPU-$\\beta$ lattices, with profiles well described by Lévy-stable distributions, and that the mean-square energy displacement grows faster than linearly. If correct, thermal conductivity in these 2D lattices diverges with system size, and the 1D diagnostic tying anomalous heat conduction to momentum superdiffusion fails in 2D. It also finds that adding an on-site potential restores normal Gaussian diffusion.","feed_headline":"Simulations find energy spreads faster than normal in 2D lattices","feed_subtitle":"Heat fluctuations follow Lévy-stable laws, so thermal conductivity should grow with system size.","key_machinery":"The central object is the spatiotemporal correlation function of energy fluctuations, $\\rho_E(i,t)$, whose central heat-mode peak rescales as $t^{1/\\gamma}\\rho_E(i/t^{1/\\gamma},t)$ and is fitted by the Lévy-stable distribution $f_{\\rm LW}^{\\gamma}$; the companion momentum correlation $\\rho_P(i,t)$ tracks sound modes. The scaling exponent $\\gamma$ extracted from the peak-height decay $H_c^E \\sim t^{-1/\\gamma}$, together with the MSD exponent $\\beta$ from $\\langle \\Delta x^2(t)\\rangle_E \\sim t^{\\beta}$, carries the classification: $1<\\gamma<2$ and $\\beta>1$ mark superdiffusion.","core_discovery":"Using equilibrium energy and momentum fluctuation correlation functions in two-dimensional square lattices with vector displacements, the paper finds that energy diffusion in the momentum-conserving purely quartic and FPU-$\\beta$ lattices is superdiffusive: the heat-mode profile rescales with exponents $\\gamma = 1.818$ and $\\gamma = 1.504$ for width 1024 and is well fitted by Lévy-stable distributions, while the mean-square displacement grows as $t^{1.27}$ and $t^{1.51}$, respectively. In the same 2D systems the momentum correlation peaks decay with exponent $\\mu = 0.402$ (quartic case), i.e. momentum diffusion is no longer superdiffusive, so the 1D rule that anomalous heat transport is corroborated by momentum superdiffusion does not extend to 2D. The $\\phi^4$ lattice with an on-site potential shows Gaussian normal diffusion regardless of width.","pith_inferences":["A testable next step is to apply the same fluctuation-correlation analysis to a 2D coupled-rotator lattice with momentum conservation but normal 1D conduction; if its energy diffusion is normal, momentum conservation would be necessary but not sufficient for anomalous 2D heat transport, mirroring the 1D rotator exception.","The reported $\\mu = 0.402$ subdiffusive momentum decay in 2D suggests that sound-mode damping, not the heat mode itself, carries the dimensional crossover; comparing $\\mu$ across widths could give a sharper diagnostic than $\\gamma$.","If logarithmic corrections are genuine for the quartic lattice, the apparent power-law spread $t^{1.27}$ could be an effective exponent drifting toward $t\\ln t$; a clean way to test this is to plot $\\langle\\Delta x^2\\rangle_E/(t\\ln t)$ versus time and look for a plateau."],"forward_implications":["In the 2D FPU-$\\beta$ lattice, thermal conductivity should diverge as a power law, roughly $\\kappa \\sim N^{0.51}$, matching direct nonequilibrium simulations with vector displacements.","In the 2D purely quartic lattice, the conductivity divergence is at least logarithmic; whether it is exactly $\\ln N$ or a weak power law is left unresolved.","The 1D criterion that anomalous heat transport is corroborated by momentum superdiffusion does not hold in 2D, where momentum spreads normally or subdiffusively while energy remains superdiffusive.","Harmonic interactions accelerate energy diffusion, so the nature of the divergence depends on the balance between harmonic and purely anharmonic forces.","Adding a $\\phi^4$ on-site potential gives normal diffusion in both 1D and 2D, confirming that momentum conservation is the key ingredient for anomalous 2D heat transport."],"supporting_citations":[{"why":"Supplies the nonlinear fluctuating hydrodynamics prediction of Lévy-3/2 scaling that anchors the scaling analysis in 1D and the expected exponent.","marker":"[20]"},{"why":"States the 1D hypothesis that anomalous heat conduction is corroborated by momentum superdiffusion, which the paper shows fails in 2D.","marker":"[27]"},{"why":"Reports power-law divergence of thermal conductivity in 2D FPU-β lattices with vector displacements, the benchmark for the FPU result here.","marker":"[31]"},{"why":"Reports logarithmically divergent conductivity for quartic scalar lattices, the comparison used for the t ln t fit of the MSD.","marker":"[32]"},{"why":"Provides the relation linking mean-square energy displacement to the heat-current autocorrelation, used to translate β into a conductivity prediction.","marker":"[34]"},{"why":"Introduced the energy fluctuation correlation function and peak-height scaling approach that the paper extends to 2D.","marker":"[36]"},{"why":"Defines the Lévy-walk/stable distribution f^γ_LW used to fit the rescaled heat-mode profiles.","marker":"[45]"}],"fun_headline_variants":["2D lattices show superdiffusive heat flow, conductivity grows with size","Heat spreads superdiffusively in 2D lattices, conductivity not constant","Superdiffusive heat in 2D lattices, no momentum link","2D lattices: heat superdiffusion, momentum diffusion normal","Heat in 2D lattices follows Lévy-stable diffusion, conductivity diverges"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The fitted exponents $\\gamma$ and $\\beta$ come from power-law fits over times up to $t=200$ and are assumed to be asymptotic; for the quartic lattice, $t^{1.27}$ and $t\\ln t$ cannot be told apart, so the quantitative exponents could be logarithmic corrections in disguise.","fun_headline_variants_meta":{"raw":{"variants":["2D lattices show superdiffusive heat flow, conductivity grows with size","Heat spreads superdiffusively in 2D lattices, conductivity not constant","Superdiffusive heat in 2D lattices, no momentum link","2D lattices: heat superdiffusion, momentum diffusion normal","Heat in 2D lattices follows Lévy-stable diffusion, conductivity diverges"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001952,"raw_usage":{"total_tokens":7644,"prompt_tokens":972,"completion_tokens":6672,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":588,"completion_tokens_details":{"reasoning_tokens":6571}},"tokens_in":588,"tokens_out":6672,"duration_ms":42530,"temperature":1.0,"reasoning_tokens":6571,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:05:48.959320+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Extend the equilibrium simulation of the purely quartic 2D lattice (width 1024) to correlation times beyond $t=200$ and compare the MSD $\\langle \\Delta x^2(t)\\rangle_E$ against $t^{1.27}$ and $t\\ln t$: if the data bend toward $t\\ln t$, the power-law diffusion exponent claimed for the quartic lattice is a finite-window artifact rather than the asymptotic law.","supporting_citations":[{"cited_title":"van Beijeren , journal Phys","cited_arxiv_id":null,"evidence_quote":"Supplies the nonlinear fluctuating hydrodynamics prediction of Lévy-3/2 scaling that anchors the scaling analysis in 1D and the expected exponent."},{"cited_title":"Li , author S","cited_arxiv_id":null,"evidence_quote":"States the 1D hypothesis that anomalous heat conduction is corroborated by momentum superdiffusion, which the paper shows fails in 2D."},{"cited_title":"Shiba and author N","cited_arxiv_id":null,"evidence_quote":"Reports power-law divergence of thermal conductivity in 2D FPU-β lattices with vector displacements, the benchmark for the FPU result here."},{"cited_title":"Wang , author B","cited_arxiv_id":null,"evidence_quote":"Reports logarithmically divergent conductivity for quartic scalar lattices, the comparison used for the t ln t fit of the MSD."},{"cited_title":"Liu , author P","cited_arxiv_id":null,"evidence_quote":"Provides the relation linking mean-square energy displacement to the heat-current autocorrelation, used to translate β into a conductivity prediction."},{"cited_title":"Zhao , journal Phys","cited_arxiv_id":null,"evidence_quote":"Introduced the energy fluctuation correlation function and peak-height scaling approach that the paper extends to 2D."},{"cited_title":"Zaburdaev , author S","cited_arxiv_id":null,"evidence_quote":"Defines the Lévy-walk/stable distribution f^γ_LW used to fit the rescaled heat-mode profiles."}],"review_version":1}