{"id":"c8e4c41c-e79a-4020-bb3f-8e80168f26f7","arxiv_id":"2508.17535","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In the isotropic Heisenberg spin chain, higher-order cumulants of spin transfer scale with exponents identical across orders for spin-helix states (nu about 0.45) but differ for domain-wall states (nu3=nu4 about 2*nu1 about 1.3).","lead":"This paper numerically studies how higher-order fluctuations grow over time in a one-dimensional quantum magnet, and finds two distinct universal scaling behaviors depending on the initial state. The results offer a classification scheme for far-from-equilibrium transport that can be probed in cold-atom and quantum-simulator experiments.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Apparent ν≈0.45 for the 'anomalous diffusive' class is extracted from time windows shorter than one decade (Jt<λ/4), so it may be a preasymptotic transient; Fig. 1(e) even lists ν=1/2, not 0.45.","rationale":"The reader's weakest assumption already identifies the finite-time extraction as the key risk; I agree. I sharpen it: the 0.45 exponent is not merely 'maybe not asymptotic'; it is the sole basis for distinguishing a new universality class, and the accessible window is extremely short (Jt<24 for the largest λ used in the XZ-type analysis). A standard diffusive system with corrections would show exactly this kind of below-1/2 effective exponent in a short window. The Fig. 1(e) table inconsistency gives a concrete internal reason to suspect the 0.45 value is not the paper's own definitive summary. The numerical methods and convergence tests are solid as far as they go, and the FCS cross-check supports reproducibility, but those checks address accuracy within the chosen time window, not the asymptotic extrapolation. The proposed test—larger λ and local-exponent analysis—would settle whether the exponent drifts to 1/2. Until then, conditional acceptance is appropriate; I do not see grounds to reject, since the superdiffusive 0.65/1.3 result has more independent support and is less affected by this concern.","tokens_in":35089,"tokens_out":10888,"duration_ms":125308,"concrete_test":"Compute the local (logarithmic) exponent ν_eff(t) = d ln|ΔP1(λ,t)| / d ln t for the XZ-type MPDW/SH initial state within the boundary-free window, for λ=96 (L=96) and for a larger system such as λ=192 (L=192), which extends the window to Jt<48. If ν_eff increases monotonically toward 0.5 with increasing λ/time (or if a fit |ΔP1| = A t^ν (1 + B t^{-ω}) yields ν→0.5 with ω>0), then 0.45 is a preasymptotic transient and the anomalous-diffusive universality class is not established. Also verify which value (0.45 or 1/2) is actually obtained by a least-squares fit in the shaded region of Fig. 3(b), since Fig. 1(e) currently reports 1/2.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central classification rests on two exponent sets: ν1-4≈0.45 for XZ-type/SH states and ν1,2≈0.65, ν3,4≈1.3 for Ising-type/DW states. The former set is the most load-bearing: it is what makes the class 'anomalous diffusive' rather than ordinary diffusion. But the data used to infer it are confined to Jt<λ/4 for XZ-type MPDW states (main-text Fig. 3 gray region; SM Figs. S6, S7), and the largest wavelength used is λ=96, giving Jt<24. This is less than 1.4 decades of time. Within such a window, a normal diffusive process with scaling corrections (or a KPZ process with a slow crossover) can produce an effective exponent of ~0.45; the paper provides no error bars, no fit with correction-to-scaling terms, and no λ→∞ extrapolation. The internal summary table in Fig. 1(e) lists the anomalous-diffusion exponents as 1/2, 1/2, 1/2, 1/2, not 0.45, which undercuts the text's claim of a distinct exponent. If the true asymptotic ν is 1/2, the claimed two universality classes become one diffusive class with non-Gaussian fluctuations, and the central novelty disappears.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies temporal scaling of the first four cumulants of spin polarization transfer and of the contrast in the isotropic Heisenberg spin chain, starting from domain-wall (DW), spin-helix (SH), and two families of multi-periodic domain-wall (MPDW) states. Using TDVP and full counting statistics, the authors claim two distinct dynamical classes: a superdiffusive class for DW/Ising-type MPDW initial states with ν1=ν2≈0.65 and ν3=ν4≈1.3, with standardized cumulants γ3≈0.224 and γ4≈−0.093 deviating from KPZ predictions; and an 'anomalous diffusive' class for SH/XZ-type MPDW states with ν1=ν2=ν3=ν4≈0.45 and strongly non-Gaussian γ3≈−0.359, γ4≈−0.503. The central claim is that higher-order cumulants reveal universal classes invisible to the first-order (magnetization) dynamics.","tokens_in":35514,"tokens_out":4142,"duration_ms":52575,"significance":"If the claimed exponents hold asymptotically, the paper would provide a valuable classification of higher-order fluctuation dynamics in an integrable quantum spin chain, going beyond the established KPZ/superdiffusive picture and giving explicit predictions for cold-atom and quantum-gas-microscope experiments. The work has clear strengths: it uses two independent numerical methods (TDVP and FCS), provides convergence tests for bond dimension, time step, and system size in the Supplemental Material, and obtains consistent results from polarization-transfer, contrast, and subsystem-fluctuation analyses. The main weakness is that the most novel exponent—ν≈0.45 for the 'anomalous diffusive' class—is extracted from short finite-time windows without statistical uncertainties, and it is internally inconsistent with the table in Fig. 1(e), which lists 1/2 for the same class. The distinction between 0.45 and 1/2 is load-bearing for the paper's central classification.","major_comments":[{"comment":"The XZ-type 'anomalous diffusive' scaling ν1,2,3,4≈0.45 is extracted only from the gray windows Jt≲λ/4 (SM S6/S7), and the largest wavelength used is λ=96, so the fitting range is Jt≲24, i.e. less than 1.4 decades. No error bars, no correction-to-scaling fits, and no λ→∞ extrapolation are provided. Since ordinary diffusion would give ν=1/2, the data as presented cannot distinguish an asymptotic exponent 0.45 from a preasymptotic effective exponent produced by a diffusive process with scaling corrections. This is load-bearing for the claimed 'anomalous diffusive' class. Please provide a local-exponent analysis (e.g. ν(t) versus 1/t), fits with subleading corrections, and an estimate of the statistical uncertainty, or explicitly state the range of exponents compatible with the data.","section":"Fig. 3 and SM S6/S7"},{"comment":"There is a direct internal inconsistency in the central summary: the table in Fig. 1(e) lists the anomalous-diffusion exponents as '1/2 1/2 1/2 1/2', while the text and Fig. 3(b) report νn≈0.45 for the same class. This is not a cosmetic issue, because the entire novelty of the 'anomalous diffusive' class rests on the deviation from the standard diffusive exponent. The authors must decide which value is claimed, correct the table or the text, and explain why the 0.45 estimate is not merely a finite-time artifact.","section":"Fig. 1(e)"},{"comment":"The standardized cumulants γ3≈−0.359 and γ4≈−0.503 are reported without error bars or a quantitative statement of their time dependence. SM S7 shows that within the shaded 'boundary-free' regions these quantities evolve visibly with time and wavelength, especially for smaller λ. To support the claim that these are asymptotic non-Gaussian universal constants, the authors should define the fitting window, report the variation across λ and time in that window, and compare with the analogous uncertainty for the KPZ values. Otherwise the deviation from Gaussian statistics may again be a preasymptotic effect.","section":"Fig. 3(c)(d) and SM S7"},{"comment":"The same finite-window concern applies to the superdiffusive exponents: ν1,2≈0.65 and ν3,4≈1.3 are extracted from Jt≲λ/2, giving at most about 1.7 decades for λ=96, and no statistical errors are quoted. These values are close to the KPZ predictions 2/3 and 4/3. Without a quantitative fit that includes subleading corrections or a demonstration of stability over a wider time window, the claim that the third- and fourth-order exponents 'differ significantly' from KPZ rests mainly on the standardized cumulants, not on the exponents themselves. Please provide the same error analysis for the superdiffusive branch.","section":"Figs. 2 and 3(a)"}],"minor_comments":[{"comment":"The word 'cumulantes' in the abstract (and 'latter' written as 'later' in the text) should be corrected.","section":"Abstract and main text"},{"comment":"The caption calls the SH state 'diffusive' while quoting ν1≈0.45; if the authors maintain that 0.45 is a distinct 'anomalous diffusive' exponent, the caption should use that term consistently, otherwise readers will read 0.45 as a numerical approximation to 1/2.","section":"Fig. 2(b)"},{"comment":"The threshold-time extraction (decay to 0.4 for K1 and first maximum for K2,3,4) appears arbitrary; a brief statement of the sensitivity of νn to these thresholds would strengthen the analysis.","section":"SM S8/S9"},{"comment":"All figures show power-law guides without confidence intervals. Adding shaded fit ranges or quoted uncertainties would materially help the reader assess the 0.45 vs 1/2 and 0.65 vs 2/3 distinctions.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The core idea is interesting and the two-method numerics are a real strength, but the main novelty—a distinct ν≈0.45 anomalous-diffusive class—is currently supported only by short-window, error-bar-free fits, and the paper itself contains a table that contradicts the text by listing 1/2 for that class. These issues are fixable within the manuscript's scope by adding careful exponent extraction with uncertainties and reconciling Fig. 1(e). I would not recommend rejection, but the manuscript is not yet at the level of statistical and internal consistency expected for the claimed universality classification."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe headline: the paper reports higher-order cumulant scaling in the isotropic Heisenberg chain for a family of initial states they call MPDW states, claiming two universality classes: a superdiffusive class with exponents ν1,2≈0.65 and ν3,4≈1.3, and an anomalous diffusive class with all four exponents ≈0.45. The second claim is the one to scrutinize; the first is on firmer ground.\n\nWhat's new: nobody has systematically extracted the third and fourth cumulant exponents for spin-helix initial states before, and the MPDW interpolation between domain-wall and helix is a useful construction. They also cross-check TDVP with full counting statistics and subsystem fluctuations, and the SM shows careful convergence tests in bond dimension, time step, and system size. That is real work and the data are presented transparently.\n\nSoft spots: the 0.45 exponent is the load-bearing claim that makes the diffusive class 'anomalous' rather than ordinary diffusion. It is extracted from time windows Jt<λ/4, which for the largest λ=96 means Jt<24 — less than 1.4 decades. There are no error bars, no fits with correction-to-scaling, and no λ→∞ extrapolation. A transient effective exponent of 0.45 is entirely plausible for an ordinary diffusive process. More troubling, the summary table in Fig. 1(e) lists the anomalous diffusion exponents as 1/2, not 0.45. The text says 0.45; the table says 1/2. That inconsistency needs to be resolved. If the true exponent is 1/2, the claimed two classes become one diffusive class with non-Gaussian fluctuations, and the central novelty collapses.\n\nFor the Ising-type class, the 0.65 exponent is consistent with prior work, and the ν3,4≈1.3 relation is plausible. The standardized cumulant values are interesting but also extracted from the same short windows, so they inherit the same caveat.\n\nWho this is for: people working on spin transport in 1D quantum magnets, KPZ universality in integrable systems, and cold-atom experiments. They will want to know about this, but they should read the SM before trusting the 0.45.\n\nRecommendation: This deserves a serious referee. It is not a desk reject. But the referee should press for error bars, a proper finite-time scaling analysis with corrections, and a resolution of the table/text discrepancy. If the 0.45 survives, it's a solid contribution; if not, the paper still has value as a careful numerical study of higher-order cumulants and the MPDW construction.","headline":"Worth a serious referee, but the central 0.45 exponent for the 'anomalous diffusive' class is not yet established — the paper's own summary table says 1/2.","tokens_in":35975,"tokens_out":3880,"would_cite":true,"duration_ms":38429,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"For the isotropic Heisenberg spin chain, the first four cumulants of spin transfer obey two distinct temporal scaling laws selected by the initial state: all orders scale as t^0.45 for helix-type states, versus t^0.65 and t^1.3 for domain-w","keywords":["higher-order cumulants","full counting statistics","isotropic Heisenberg chain","KPZ universality","superdiffusion","anomalous diffusion","spin helix","domain wall"],"falsifier":"Simulate or measure the same cumulants at larger wavelength and system size so that the boundary-free window extends well beyond Jt ≈ 100, then check whether the fitted exponents stay at 0.45, 0.65, and 1.3 or drift. If ν1 for XZ-type/spin-helix states trends toward 1/2 while γ3 and γ4 trend toward zero, the anomalous-diffusion class is a transient; similarly, extending Ising-type data beyond Jt ≈ λ/2 should show whether γ3 moves toward the KPZ Tracy-Widom value instead of remaining near 0.224.","tokens_in":35016,"feed_emoji":"🧲","tokens_out":8604,"duration_ms":94572,"temperature":0.7,"pith_summary":"This paper claims that in the one-dimensional isotropic Heisenberg model, the growth of the first four cumulants of spin polarization transfer and contrast is governed by only two universal scaling classes, selected by the initial state. For domain-wall-like (Ising-type) initial states the first two cumulants grow as t^0.65 while the third and fourth grow as t^1.3, with skewness and excess kurtosis values that deviate from the Kardar-Parisi-Zhang (KPZ) prediction. For spin-helix-like (XZ-type) states all four cumulants grow as t^0.45, with non-Gaussian standardized cumulants; the same exponent is recovered for genuine spin-helix states. The authors unify both behaviors by introducing multi-periodic domain-wall states that interpolate between the two experimental preparations, and confirm the scaling with two independent numerical approaches. If correct, higher-order cumulants become a sharp experimental probe for distinguishing transport universality classes that first-order measurements cannot separate.","feed_headline":"Spin-chain fluctuations split into two universal scaling classes","feed_subtitle":"Helix states: all cumulants grow as t^0.45; domain walls: third and fourth double the first two's exponent.","key_machinery":"The central objects are the nth-order cumulants of two nonlocal observables—spin polarization transfer Pn and wavevector-resolved contrast Kn—extracted from connected n-point correlation functions of local magnetization. The unifying device is the multi-periodic domain-wall (MPDW) family of initial states, a product state whose local spin orientation is controlled by an amplitude ratio and phase difference; the Ising-type member reproduces domain-wall dynamics and the XZ-type member reproduces spin-helix dynamics. Cumulant growth is characterized by power laws |ΔPn|∝t^νn and by threshold-time scaling τ∝λ^{1/νn}, and standardized cumulants γ3 and γ4 (skewness and excess kurtosis) identify non","core_discovery":"The paper's central claim is that higher-order cumulant dynamics in the isotropic Heisenberg spin chain is universal but initial-state dependent, and falls into exactly two classes captured by the first four cumulants. For domain-wall-like (Ising-type) initial states—including the pure domain wall—the spin polarization transfer cumulants obey |ΔP1|,|ΔP2| ∝ t^0.65 and |ΔP3|,|ΔP4| ∝ t^1.3; the standardized cumulants take skewness γ3≈0.224 and excess kurtosis γ4≈−0.093, values that agree with recent experiments and deviate from the KPZ/Tracy-Widom prediction. For spin-helix-like (XZ-type) initial states—including the genuine spin helix—all four cumulants obey |ΔPn| ∝ t^0.45 with γ3≈−0.359 and γ","pith_inferences":["The ν≈0.45 class is a candidate new universality class distinct from normal diffusion (ν=1/2) and KPZ (ν=2/3); a decisive test would be whether the exponent survives simulation windows longer than the boundary-free limit used here.","Varying the MPDW amplitude ratio between 0 and 1 should interpolate between the two classes; if exponents vary continuously with this ratio, the classification is controlled by the initial state's transverse-spin content, not by integrability.","The same full-counting-statistics analysis could be transferred to fermionic or doped one-dimensional systems to ask whether spin and charge sectors carry different cumulant scaling signatures, extending the paper's suggested spin-charge separation direction.","The standardized cumulant plateaus are experimentally checkable: measuring γ3≈−0.359 and γ4≈−0.503 in spin-helix ultracold-atom setups over longer hold times would confirm the anomalous-diffusion class; drift toward zero would indicate preasymptotic transients."],"forward_implications":["Spin-helix-type initial states (SH and XZ-type MPDW) give identical anomalous-diffusive exponents ν1=ν2=ν3=ν4≈0.45, with skewness γ3≈−0.359 and excess kurtosis γ4≈−0.503, so the higher-order cumulants are the only place the non-Gaussian nature shows up.","Domain-wall-type initial states (DW and Ising-type MPDW) give ν1=ν2≈0.65 and ν3=ν4≈1.3, with γ3≈0.224 and γ4≈−0.093, deviating from the KPZ/Tracy-Widom prediction for the fourth order.","The two scaling laws are confirmed through three independent extraction routes—spin polarization transfer, contrast threshold times, and subsystem fluctuations—so the classification is not an artifact of one observable.","Because the same exponents are obtained from spin-helix and XZ-type MPDW states, and from DW and Ising-type MPDW states, the MPDW initial states form a unified experimental platform for studying both universality classes.","The results settle, for these initial states, the open question of whether higher-order cumulants follow universal power laws in diffusive and superdiffusive spin dynamics."],"supporting_citations":[{"why":"Supplies the infinite-temperature scaling relation ν_{2n}=nν_2 for cumulants that the paper compares and contrasts with its finite-temperature results.","marker":"[30]"},{"why":"Experimental observation of KPZ superdiffusion in the domain-wall state; supplies the ν1≈0.65 baseline and the experimental setup the paper extends to higher orders.","marker":"[32]"},{"why":"Experimental observation of diffusive spin transport in a spin-helix state; supplies the ν1≈0.45 baseline reproduced by XZ-type MPDW states.","marker":"[33]"},{"why":"Experimental measurement of higher-order magnetization cumulants in a Heisenberg chain; provides the γ4≈−0.093 value that the Ising-type class reproduces.","marker":"[36]"},{"why":"Supplies the efficient cumulant-evolution/full-counting-statistics algorithm used as an independent numerical cross-check.","marker":"[37]"},{"why":"Supplies the tensor-network implementation used for the matrix-product-state time-evolution simulations.","marker":"[58]"},{"why":"Establishes that diffusive transport can contain distinct universality classes visible in full counting statistics, supporting the anomalous-diffusive interpretation.","marker":"[73]"}],"fun_headline_variants":["Spin-chain cumulants reveal two universal scaling classes","Two universal fluctuation classes from spin-chain cumulants","Spin-chain cumulants: KPZ breakdown plus new universal scaling","Anomalous diffusive and superdiffusive: spin-chain cumulant universality"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The extracted exponents are assumed to be true asymptotic power laws, but they are measured only in finite time windows bounded by boundary effects (Jt < λ/2 for Ising-type and Jt < λ/4 for XZ-type states); if longer-time simulations show a crossover, for instance 0.45 drifting toward 1/2, the claimed universality classes would not hold.","fun_headline_variants_meta":{"raw":{"variants":["Spin-chain cumulants reveal two universal scaling classes","Two universal fluctuation classes from spin-chain cumulants","Spin-chain cumulants: KPZ breakdown plus new universal scaling","Anomalous diffusive and superdiffusive: spin-chain cumulant universality"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000286,"raw_usage":{"total_tokens":1540,"prompt_tokens":783,"completion_tokens":757,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":527,"completion_tokens_details":{"reasoning_tokens":700}},"tokens_in":527,"tokens_out":757,"duration_ms":9041,"temperature":1.0,"reasoning_tokens":700,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T16:51:32.041463+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate or measure the same cumulants at larger wavelength and system size so that the boundary-free window extends well beyond Jt ≈ 100, then check whether the fitted exponents stay at 0.45, 0.65, and 1.3 or drift. If ν1 for XZ-type/spin-helix states trends toward 1/2 while γ3 and γ4 trend toward zero, the anomalous-diffusion class is a transient; similarly, extending Ising-type data beyond Jt ≈ λ/2 should show whether γ3 moves toward the KPZ Tracy-Widom value instead of remaining near 0.224.","supporting_citations":[{"cited_title":"Krajnik, J","cited_arxiv_id":null,"evidence_quote":"Supplies the infinite-temperature scaling relation ν_{2n}=nν_2 for cumulants that the paper compares and contrasts with its finite-temperature results."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Experimental observation of KPZ superdiffusion in the domain-wall state; supplies the ν1≈0.65 baseline and the experimental setup the paper extends to higher orders."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Experimental observation of diffusive spin transport in a spin-helix state; supplies the ν1≈0.45 baseline reproduced by XZ-type MPDW states."},{"cited_title":"Rosenberg, T","cited_arxiv_id":null,"evidence_quote":"Experimental measurement of higher-order magnetization cumulants in a Heisenberg chain; provides the γ4≈−0.093 value that the Ising-type class reproduces."},{"cited_title":"Gopalakrishnan, A","cited_arxiv_id":null,"evidence_quote":"Establishes that diffusive transport can contain distinct universality classes visible in full counting statistics, supporting the anomalous-diffusive interpretation."}],"review_version":1}