{"id":"d24a93aa-d387-403c-af2e-e833b1e7512a","arxiv_id":"2411.15026","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":1,"one_line_summary":"For a generalized KPZ equation with competing local and nonlocal nonlinearities, mode-coupling theory predicts a rough phase with nonuniversal exponents and a crumpled phase with short-range order.","lead":"A theoretical study of a generalized Kardar-Parisi-Zhang growth equation predicts that its strongly fluctuating regime splits into a rough phase and a crumpled phase, with scaling exponents that depend continuously on model parameters. The result suggests that nonlocal growth rules can create new universality classes beyond standard surface roughening.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (15) yields χ<0 on part of the claimed crumpled interval: for γ∈(-2/3,-1/2), A(γ)<0 while B(γ)<0, so the selected MCT root is negative; the 2D crumpled phase is not supported by the paper's own MCT result.","rationale":"The paper is an honest, speculative theory letter, and I read it as attempting to establish two strong-coupling phases using one-loop RG, bare perturbation theory, and one-loop MCT. The exact pseudo-Galilean identity χ+z=2 and the one-loop RG separatrices are reasonable, and the authors explicitly flag the unproven marginality of γ and the breakdown of the theory for χ>1. However, the single most load-bearing quantitative step is Eq. (15), because it is used both to define the nonuniversal exponents and to locate the rough/crumpled boundary. That step contains a concrete algebraic inconsistency: the branch selected as χ=-B/A is positive only where A>0, but A changes sign inside the interval where the crumpled phase is claimed, so large parts of (-1.043,-0.142) have χ<0. This is not a matter of disagreeing with a consensus or of lacking numerical confirmation; it is a checkable internal property of the paper's own equations. The reader's weakest assumption (exact marginality of γ) is also real and well acknowledged, but the A-sign issue is more immediate: it means the MCT calculation as written does not establish the very phase boundary it is used to draw. A conditional verdict remains appropriate because the qualitative idea of a crumpled phase could still survive via the separate νe<0 argument or higher-order calculations, but the current derivation needs correction before the quantitative phase diagram is accepted.","tokens_in":16560,"tokens_out":10679,"duration_ms":98131,"concrete_test":"Evaluate Eq. (15) at γ=-0.6: if χ=-B/A≈-44.7 rather than >1, the claimed crumpled interval is not supported. More systematically, plot the valid MCT domain B(γ)<0 and A(γ)>0 over -1.383<γ<0.161 and recompute the χ=1 boundaries using only those branches; if the valid crumpled bands are separated by an invalid band, revise the phase diagram and the stated purple region in Fig. 1 accordingly.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The quantitative evidence for the 2D crumpled phase comes from setting χ=1 in Eq. (15), which gives γ=-1.043 and γ=-0.142, and the paper then calls the whole interval (-1.043,-0.142) crumpled. But Eq. (15) is χ=-B/A with B=18γ²+22γ-4 and A=36γ²+42γ+12. The condition for the selected root to be positive is B<0 and A>0. While B<0 on the whole stated interval (-1.383,0.161), A changes sign at γ=-2/3 and γ=-1/2 and is negative on (-2/3,-1/2). For example, at γ=-0.6, B=-10.72 and A=-0.24, giving χ≈-44.7. Thus the middle of the claimed crumpled interval is not a positive-χ MCT solution; the valid crumpled bands are instead roughly (-1.043,-2/3) and (-1/2,-0.142), separated by an invalid band where the quadratic has no positive root. The paper does not acknowledge this and draws a single purple crumpled region in Fig. 1(a). The bare-perturbation argument (Eq. 7) gives a different interval (-1.25,0), so it cannot independently rescue the MCT boundary. This internal sign/branch error directly undermines the central quantitative claim of a crumpled phase in 2D, separate from the separately acknowledged uncertainty about exact marginality of γ.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a generalized Kardar-Parisi-Zhang equation that adds a nonlocal nonlinearity proportional to lambda_1 to the usual local KPZ nonlinearity. Using a one-loop dynamic RG and a one-loop self-consistent mode-coupling theory (MCT), it claims that in two dimensions the strong-coupling regime splits into an algebraically rough phase with 0<chi(gamma)<1 and a crumpled phase with chi(gamma)>1, where gamma=lambda_1/lambda and chi is given by Eq. (15). The rough phase is predicted to have nonuniversal, continuously varying exponents, orientational long-range order and positional short-range order, while the crumpled phase has short-range order in both. The paper also extends these predictions to d>2 and speculates on the global RG fixed-point structure.","tokens_in":16946,"tokens_out":6202,"duration_ms":61765,"significance":"If the central claims hold, the paper would meaningfully extend the KPZ paradigm by showing that competing local and nonlocal nonlinearities can produce strong-coupling phases with nonuniversal scaling and a crumpled phase. Several elements are done carefully and are strengths: the pseudo-Galilean transformation is explicit and yields the exact relation chi+z=2; the one-loop MCT algebra is presented in enough detail in the Supplemental Material to be checked by the reader; and the gamma=0 limit reproduces the known KPZ MCT value chi=1/3 in 2D, providing a useful internal consistency check. However, the quantitative evidence for the 2D crumpled phase is weakened by a sign/branch error in the selected MCT root, and the continuously varying nonuniversal exponents rest on the unproven exact marginality of gamma. These issues are load-bearing for the paper's headline predictions.","major_comments":[{"comment":"The selected root chi=chi+ is not positive on the whole claimed crumpled interval. With d=2, A(gamma)=36gamma^2+42gamma+12 vanishes at gamma=-2/3 and -1/2 and is negative between those roots, while B(gamma)=18gamma^2+22gamma-4 is negative throughout (-1.383,0.161). On the subinterval (-2/3,-1/2), the quadratic (13) with C=0 has roots 0 and -B/A<0, so there is no positive-chi MCT solution; Eq. (15) itself gives chi=-44.7 at gamma=-0.6. The paper's own validity condition, stated just before Eq. (15) as 'B<0 and A>0, or B>0 and A<0', is violated in this band. Consequently, setting chi=1 in Eq. (15) supports crumpling only in the two disjoint bands (-1.043,-2/3) and (-1/2,-0.142), separated by a band with no positive-chi MCT solution. The continuous purple crumpled region in Fig. 1(a) and the abstract's statement of a 2D crumpled phase therefore need correction.","section":"Eq. (15) and Fig. 1(a)"},{"comment":"The continuously varying nonuniversal exponents chi(gamma) in the rough phase depend on gamma=lambda_1/lambda being exactly marginal at all loop orders. The one-loop RG shows no vertex corrections, but the paper explicitly states that unequal higher-loop renormalizations of lambda and lambda_1 would make gamma flow and replace the fixed line by isolated fixed points. Because the abstract and opening sections present nonuniversal scaling as an established result, the manuscript should either provide an argument for exact marginality or clearly label this as a conjecture at every point where it is used, rather than only in the closing paragraph.","section":"Final paragraph and central claim of nonuniversality"},{"comment":"The independent support for the crumpled phase comes from the bare-perturbation condition (7), which gives -1.25<gamma<0. After correcting the sign issue in Eq. (15), the MCT and bare-perturbation intervals do not coincide, and the invalid band (-2/3,-1/2) lies inside the bare-perturbation region but has no positive-chi MCT solution. The text says the two methods 'come to a similar conclusion', but this agreement is not obvious once the MCT interval is redrawn. The authors should state the corrected MCT interval, compare it explicitly with Eq. (7), and discuss whether the 2D crumpled phase is robust or an artifact of the one-loop Lorentzian closure.","section":"Eq. (7) and MCT comparison"},{"comment":"The MCT derivation assumes Lorentzian correlation functions, dominance of one-loop diagrams, and z<2 with chi>0; these assumptions are uncontrolled in the strong-coupling regime. Since the known MCT predictions for pure KPZ in d>2 disagree with several numerical studies cited in the introduction, the paper should state more prominently that the quantitative predictions for the rough and crumpled phases inherit this uncontrolled approximation. This is a caveat rather than a fatal defect, but it should accompany the central claims.","section":"MCT closure assumptions"}],"minor_comments":[{"comment":"The expression chi = -B + |B|/(2A) is ambiguous; it should be written as (-B+|B|)/(2A) for both chi+ and chi-.","section":"Eq. (14) and following line"},{"comment":"Figure 1(c) should mark the interval (-2/3,-1/2) where Eq. (15) has no positive-chi solution instead of drawing a single central crumpled region.","section":"Fig. 1(c)"},{"comment":"The sentence 'as A, depending upon gamma, decreases, chi grows, eventually exceeding unity' is valid only where A>0; it fails where A changes sign. This wording should be revised together with the corrected interval.","section":"Text near Eq. (15)"},{"comment":"The statement that chi>1 implies crumpling in d>2 is subject to the same sign-selection issue as in 2D; the corresponding gamma intervals should be rederived using the positivity conditions on the coefficients of the quadratic for each d.","section":"Higher-dimensional discussion"},{"comment":"The 'Occam's razor' global flow diagram is clearly speculative, but the caption should state explicitly that the fixed lines and fixed points in the strong-coupling region are conjectured and not derived from the calculations.","section":"Fig. 2 caption"}],"recommendation":"major_revision","confidential_remarks":"The sign/branch error is concrete and correctable in principle, so I do not recommend rejection. However, the headline prediction of a 2D crumpled phase should not be stated as established until the MCT interval is corrected and the dependence on the unproven marginality of gamma is presented as an explicit conjecture throughout. The paper should also clarify which quantitative results are new relative to the closely related Ref. [34], since several structural ingredients appear there already."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things about this paper. First, it is the first attempt to characterize the strong-coupling regime of the G-KPZ equation, and it predicts a split into an algebraically rough phase with nonuniversal exponents and a crumpled phase. That is a substantive new claim for a paradigmatic nonequilibrium growth model. Second, the quantitative evidence for the 2D crumpled phase has a sign/branch problem in the MCT equation that the paper does not acknowledge; the claimed crumpled interval is partly invalid.\n\nThe authors solve a one-loop MCT self-consistency equation for chi(gamma), recovering chi=1/3 at gamma=0 (pure KPZ) and using the exact relation chi+z=2 from the pseudo-Galilean invariance. The rough-phase prediction, 0<chi<1 over intervals near the separatrices, is internally coherent. They are also admirably explicit about the unproven assumption that gamma is exactly marginal, and about the theory breaking down for chi>1.\n\nThe soft spots are real. The 2D MCT equation is chi = -B/A with B=18 gamma^2+22 gamma-4 and A=36 gamma^2+42 gamma+12. The paper selects the B<0 branch because B=0 gives the separatrix, but then plots chi=-B/A for all B<0 without enforcing A>0. On the interval (-2/3,-1/2), A is negative, so chi is negative there; for example at gamma=-0.6, chi is about -44.7. The valid crumpled bands (chi>1) are actually (-1.043,-2/3) and (-1/2,-0.142), separated by a gap with no positive root. Figure 1(b) draws one continuous purple region. Also, the A values quoted at the separatrices appear swapped: gamma=0.161 gives A about 19.69 and gamma=-1.383 gives A about 22.77, not the other way around. These are fixable, but they undercut the specific quantitative claim.\n\nBeyond that, the whole crumpled phase is inferred from chi>1 in a theory that breaks down there, so it is a speculation, not a derivation. The nonuniversal chi(gamma) depends entirely on gamma staying marginal at all loop orders, which the authors admit is unproven. One-loop MCT with Lorentzian correlations is uncontrolled. None of this is disqualifying for a research letter, but it means the paper is a proposal with a flaw in its central numerical support.\n\nWho is it for? People working on nonlocal growth equations and MCT applications. It deserves a serious referee—there is a real model and a real calculation—but I would not treat the crumpled phase as established. Send it to review with a request to fix the sign issue, report the correct intervals, and soften the claims.","headline":"First serious strong-coupling analysis of the G-KPZ equation, but the crumpled-phase claim rests on an MCT branch mistake that punches a hole in the quoted interval.","tokens_in":17441,"tokens_out":3478,"would_cite":false,"duration_ms":30549,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["82C28","82C31"],"pacs":[],"model":"deepseek-v4-flash","headline":"A generalized KPZ surface equation has two strong-coupling phases, one rough and one crumpled.","keywords":["generalized KPZ equation","nonlocal nonlinearity","roughness exponent","crumpled phase","strong-coupling phase","nonuniversal scaling","roughening transition","mode-coupling theory"],"falsifier":"Run pseudospectral simulations of Eq. (2) in two dimensions for several $\\gamma$ values in $(-1.383,0.161)$, measure the height-difference structure factor at long times to extract $\\chi(\\gamma)$, and look for the predicted divergence of $\\langle(\\nabla h)^2\\rangle$ for $\\gamma$ in $(-1.043,-0.142)$. The central claim fails if $\\chi$ does not vary continuously with $\\gamma$, or if the surface variance does not grow as $L^{2\\chi}$ with $\\chi>1$ in the alleged crumpled interval.","tokens_in":16357,"feed_emoji":"📐","tokens_out":9518,"duration_ms":85898,"temperature":0.7,"pith_summary":"The paper studies a generalized Kardar-Parisi-Zhang equation for growing surfaces that adds a nonlocal nonlinearity to the usual local one. In two dimensions the authors argue that the strong-coupling regime, which ordinary perturbative renormalization cannot reach, splits into two distinct phases: an algebraically rough phase with $0<\\chi(\\gamma)<1$ and a crumpled phase with $\\chi(\\gamma)>1$. The roughness exponent $\\chi$ is a continuous function of the nonlinearity ratio $\\gamma=\\lambda_1/\\lambda$, so scaling is nonuniversal and governed by a fixed line rather than isolated fixed points. This matters because the standard two-dimensional KPZ equation has no roughening transition, while this generalized model predicts one, and because the crumpled phase signals a genuine breakdown of the height-gradient description.","feed_headline":"Generalized KPZ surface can turn rough or crumpled in 2D","feed_subtitle":"Nonuniversal scaling exponents set by the local-to-nonlocal nonlinearity ratio mark a new phase structure.","key_machinery":"The central object is the generalized KPZ equation (2), whose nonlocal vertex is built from the longitudinal projection operator $Q_{ij}(\\mathbf{k})=k_ik_j/k^2$, so that the second nonlinearity is nonlocal in $\\nabla h$. A pseudo-Galilean invariance fixes $\\chi+z=2$ exactly in the strong-coupling phases. The argument then runs through two linked calculations: a one-loop dynamic renormalization-group flow that locates the perturbatively unstable region between the separatrices $g_2=0.161\\sqrt{g}$ and $g_2=-1.383\\sqrt{g}$, and a one-loop mode-coupling theory (MCT) that computes the zero-frequency self-energy and correlation-function amplitudes and equates the universal amplitude ratio $\\Gamma^2/(\\lambda^2 D)$ obtained from the two sets of one-loop diagrams. Equating the two expressions yields a quadratic equation for $\\chi$; the branch $\\chi_+$ is selected because it reduces to the known KPZ result $\\chi=1/3$ at $\\gamma=0$ in two dimensions, and it produces Eq. (15). The continuous $\\gamma$-dependence of $\\chi$ rests on the absence of vertex renormalization at one loop, which keeps $\\gamma$ marginal and makes a fixed line, not a fixed point, characterize the rough phase.","core_discovery":"On its own terms, the paper establishes that the strong-coupling phase of the generalized KPZ equation in $d=2$ is not uniform. For $-1.383<\\gamma<0.161$ with $\\gamma=\\lambda_1/\\lambda$, the surface is algebraically rough with $0<\\chi<1$; within that interval, for $-1.043<\\gamma<-0.142$, the predicted $\\chi$ exceeds one, which makes $\\langle(\\nabla h)^2\\rangle$ grow with system size and identifies a crumpled phase with positional and orientational short-range order. In the rough phase the surface has orientational long-range order and positional short-range order, and the exponents $\\chi(\\gamma)$ and $z(\\gamma)=2-\\chi(\\gamma)$ vary continuously with $\\gamma$, with $\\chi$ given by $\\chi=(-18\\gamma^2-22\\gamma+4)/(36\\gamma^2+42\\gamma+12)$. The same one-loop mode-coupling construction applied in $d=3$ yields the same rough-versus-crumpled structure, and the authors argue by extension that it holds for all $d>2$.","pith_inferences":["If the fixed-line picture survives higher-order corrections, this model would be one of the rare nonequilibrium surface problems with continuously tunable scaling exponents, and any numerical test should report $\\chi$ as a function of $\\gamma$ rather than a single exponent.","The paper's own caveat is the sharpest test: unequal infinite renormalizations of $\\lambda$ and $\\lambda_1$ at higher loop order would replace the fixed line by isolated fixed points, so a two-loop or nonperturbative calculation of the flow of $\\gamma$ would decide between continuous and discrete exponents.","Before the crumpled phase itself is probed, the sign change of the effective diffusion coefficient $\\nu_e$ at $\\gamma\\simeq-1.25$ and $\\gamma\\simeq0$ offers a cleaner early-warning observable, since a negative $\\nu_e$ at large scales is the physical mechanism the paper associates with crumpling.","The orientational-order language invites a direct comparison with membrane crumpling: if the nonlocal term acts as an effective bending stiffness, adding chiral or conserving terms, as the paper suggests, should push the crumple boundary to larger $|\\gamma|$."],"forward_implications":["In two dimensions the generalized model admits a roughening transition between a logarithmically rough weak-coupling phase and an algebraically rough strong-coupling phase, behavior the pure two-dimensional KPZ equation does not exhibit.","In the algebraically rough phase the scaling exponents vary continuously with the nonlinearity ratio $\\gamma$, so the model family displays nonuniversal, parameter-dependent scaling rather than isolated universality classes.","For $\\gamma$ inside the predicted crumpled interval, the mean-square height gradient diverges with system size, and the paper concludes that higher-order nonlinear terms absent from Eq. (2) become relevant at the crumpling threshold.","The exact relation $\\chi+z=2$ ties the dynamic exponent to the roughness exponent, so both exponents are nonuniversal together in the rough phase.","In $d>2$ the same two-phase structure appears, including ranges where the roughening transition is directly from a smooth phase to a crumpled phase."],"supporting_citations":[{"why":"This reference introduces the generalized KPZ equation and derives the weak-coupling RG flows and logarithmic roughening that the present strong-coupling analysis extends.","marker":"[34]"},{"why":"This supplemental material supplies the diagrammatic self-energy and correlation-function integrals that produce Eqs. (11), (12), and the quadratic equation for $\\chi$.","marker":"[35]"},{"why":"This reference provides the one-loop MCT scheme for the pure KPZ equation and the $\\chi=1/3$ value in two dimensions that the $\\gamma=0$ limit of Eq. (15) must reproduce.","marker":"[9]"},{"why":"This reference gives the mode-coupling framework for coupled growth equations used to set up the amplitude-ratio calculation.","marker":"[16]"},{"why":"This reference defines the roughness and dynamic exponents and the height-correlation function used to state all scaling predictions.","marker":"[3]"},{"why":"This reference supplies the earlier rough-versus-crumpled phase analysis for kinetic growth with surface relaxation that motivates the conjectured global RG flow topology.","marker":"[42]"}],"fun_headline_variants":["KPZ surface shows rough and crumpled phases","2D KPZ surface turns rough or crumpled","Rough and crumpled phases in KPZ surface","KPZ surface: two strong-coupling phases"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole prediction stands on two linked assumptions: the bare nonlinearity ratio $\\gamma=\\lambda_1/\\lambda$ never renormalizes, even at higher loop orders, and the one-loop mode-coupling calculation with Lorentzian correlation shapes captures the true strong-coupling physics.","fun_headline_variants_meta":{"raw":{"variants":["KPZ surface shows rough and crumpled phases","2D KPZ surface turns rough or crumpled","Rough and crumpled phases in KPZ surface","KPZ surface: two strong-coupling phases"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000514,"raw_usage":{"total_tokens":2475,"prompt_tokens":899,"completion_tokens":1576,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":515,"completion_tokens_details":{"reasoning_tokens":1515}},"tokens_in":515,"tokens_out":1576,"duration_ms":12194,"temperature":1.0,"reasoning_tokens":1515,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T14:36:30.137719+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run pseudospectral simulations of Eq. (2) in two dimensions for several $\\gamma$ values in $(-1.383,0.161)$, measure the height-difference structure factor at long times to extract $\\chi(\\gamma)$, and look for the predicted divergence of $\\langle(\\nabla h)^2\\rangle$ for $\\gamma$ in $(-1.043,-0.142)$. The central claim fails if $\\chi$ does not vary continuously with $\\gamma$, or if the surface variance does not grow as $L^{2\\chi}$ with $\\chi>1$ in the alleged crumpled interval.","supporting_citations":[{"cited_title":"Logarithmic or al- gebraic: Roughening of an active Kardar-Parisi-Zhang surface,","cited_arxiv_id":null,"evidence_quote":"This reference introduces the generalized KPZ equation and derives the weak-coupling RG flows and logarithmic roughening that the present strong-coupling analysis extends."},{"cited_title":"See Supplemental Material for the intermediate calcu- lational details, which includes Refs. [5, 34, 46, 47]","cited_arxiv_id":null,"evidence_quote":"This supplemental material supplies the diagrammatic self-energy and correlation-function integrals that produce Eqs. (11), (12), and the quadratic equation for $\\chi$."},{"cited_title":"Upper critical dimension of the Kardar-Parisi-Zhang equation,","cited_arxiv_id":null,"evidence_quote":"This reference provides the one-loop MCT scheme for the pure KPZ equation and the $\\chi=1/3$ value in two dimensions that the $\\gamma=0$ limit of Eq. (15) must reproduce."},{"cited_title":"Scaling and universality in coupled driven diffusive models,","cited_arxiv_id":null,"evidence_quote":"This reference gives the mode-coupling framework for coupled growth equations used to set up the amplitude-ratio calculation."},{"cited_title":"Barab´ asi and H","cited_arxiv_id":null,"evidence_quote":"This reference defines the roughness and dynamic exponents and the height-correlation function used to state all scaling predictions."},{"cited_title":"Rough or crumpled: Phases in kinetic growth with surface relaxation,","cited_arxiv_id":null,"evidence_quote":"This reference supplies the earlier rough-versus-crumpled phase analysis for kinetic growth with surface relaxation that motivates the conjectured global RG flow topology."}],"review_version":1}