{"id":"70c91c88-f4d3-49d4-bb35-0c20336590ae","arxiv_id":"1908.09177","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":2.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"This survey maps solution theory and dynamics of stochastic reaction-diffusion equations, reporting known results rather than proving new ones.","lead":"A survey chapter reviewing the main solution concepts and long-time dynamical tools for stochastic reaction-diffusion equations, from Ito calculus to rough paths and regularity structures. A useful entry map for applied mathematicians, with the caveat that the metastability formulas and self-referenced recent results need checking.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Section 3.2 rate function omits the drift and Laplacian; as printed it cannot yield the cited Kramers metastability asymptotics.","rationale":"I agree with the reader that the paper is best read as a survey whose central value is accuracy and attributability, and I agree with the reader's strongest claim that the Section 3.2 LDP display is a concrete mathematical flaw. The most load-bearing issue is not the bibliographic conflation of [123] and [124]—that is a serious mechanical defect but one that can be fixed without changing the mathematical narrative—whereas the wrong rate function is a substantive error in a topic (metastability) that the abstract explicitly advertises. The reader's weakest_assumption focuses on the unreachability of the authors' own preprints; I would instead pinpoint the LDP formula as the single most damaging accuracy failure because it is directly checkable and because the metastability discussion, including the appeal to [18], rests on it. My recommendation is unchanged from the reader's CONDITIONAL verdict: the classical expository core is standard and mostly correctly attributed, but the paper should not be accepted in its current form without correcting the rate function, fixing the duplicated 'lim inf' immediately after it, and resolving the reference-list conflation for [123]/[124]. No new verdict category is needed; the reader's conditional assessment already captures the appropriate level of concern.","tokens_in":41532,"tokens_out":5198,"duration_ms":54522,"concrete_test":"Compute the minimal value of the displayed F[0,T] for a path connecting u(0) ≡ −1 and u(T) ≡ +1 on S^1, using the linear interpolation γ(t,x) = −1 + 2t/T. This gives inf F = 2/T, which vanishes as T→∞. Then compute the minimal value of the corrected rate F_corr(γ) = (1/2)∫∫ |∂_t γ − (∂_x^2 γ + γ − γ^3)|^2 dx dt for the same transition; standard Freidlin–Wentzell theory (or the results cited as [18, 15]) gives a finite positive limit as T→∞ equal to the energy barrier between the two stable states. If the corrected infimum has a finite positive limit while the printed infimum is 2/T, then the displayed LDP cannot be the LDP for (3.12) and the metastability conclusion as written does not follow.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's Section 3.2 introduces the Allen-Cahn SPDE (3.12), du = [∂_x^2 u + u − u^3] dt + √(2ε) dW_Id, and then states that the solution satisfies an LDP with good rate function F[0,T](γ) = (1/2) ∫_0^T ∫_{S^1} |∂_t γ(t,x)|^2 dx dt. This is the rate function for a pure additive-noise process with no drift; for (3.12) the Freidlin–Wentzell rate function should contain the deterministic drift, i.e. (1/2)∫∫ |∂_t γ − (∂_x^2 γ + γ − γ^3)|^2 dx dt, up to the usual spatial regularity conventions. As printed, the minimal cost of a transition from u ≡ −1 to u ≡ +1 over time horizon T is 2/T (the linear interpolation has ∂_t γ = 2/T on a circle of length 1), which tends to 0 as T→∞. The correct action for the double-well drift has a finite, barrier-dependent limit as T→∞; this finite barrier is exactly what the survey invokes from [18] for Kramers-type metastable exit times. The displayed formula therefore does not support the subsequent metastability discussion: with the printed rate function, transition probabilities would not be exponentially small in 1/ε with a fixed exponent independent of T. This is not merely a typographical slip in one display; the paragraph explicitly uses this F to conclude that 'metastability occurs' and to connect to sharp exit-time asymptotics in [18, 15]. Since metastability is one of the central dynamical topics advertised in the abstract, the survey's accuracy in this section depends on correcting this rate function or substantially rewording the claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript is a survey of solution theory and dynamical systems for stochastic reaction-diffusion equations. It reviews classical Itô-type mild, weak, strong, martingale, and variational solution concepts; quasilinear pathwise mild solutions; Walsh-type random field solutions; rough-path mild solutions for fractional noise; and regularity-structure renormalized solutions. On the dynamics side, it discusses random dynamical systems, stability concepts, large deviations and metastability, random invariant manifolds, and random pullback attractors, including recent results of the authors. The survey is intended as an accessible reference for a broad audience and is organized largely by solution-theoretic framework and by dynamical quantity.","tokens_in":41790,"tokens_out":3295,"duration_ms":35838,"significance":"If the accuracy issues identified below are corrected, the paper would be a useful survey that collects many recent developments in one place. Its classical sections (2.1–2.3, 2.5, 2.7, 3.1, 3.4) report standard material with generally correct attributions, and it gives a readable account of rough-path and quasilinear methods that are not always covered in survey form. The authors are also explicit about which results are their own recent work and provide proof sketches for these items. The main value of the paper is therefore bibliographic and expository; its reliability depends on the correctness of the statements and references it reports.","major_comments":[{"comment":"The large-deviation statement for the Allen-Cahn SPDE (3.12) is incorrect as printed. The displayed rate function F[0,T](γ) = (1/2)∫_0^T∫_{S^1}|∂_t γ|^2 dx dt is the rate function for a pure additive-noise process with no drift, whereas for du = [∂_x^2 u + u − u^3] dt + √(2ε) dW_Id the Freidlin–Wentzell rate function should contain the full deterministic drift, i.e. (1/2)∫_0^T∫_{S^1}|∂_t γ − (∂_x^2 γ + γ − γ^3)|^2 dx dt (up to the usual spatial regularity conventions). This is not a cosmetic issue: with the printed rate function, the minimal cost of a transition from u ≡ −1 to u ≡ +1 over a time horizon T is 2/T (attained by the linear interpolation on S^1), which tends to zero as T → ∞, so the exit probabilities would not be exponentially small in 1/ε with a fixed, T-independent exponent. The paragraph explicitly uses this F to conclude that metastability occurs and to connect to the sharp exit-time asymptotics of [18, 15]. The displayed formula must be corrected and the subsequent discussion reworded accordingly.","section":"Section 3.2 and references [123], [124]"},{"comment":"The bibliography lists the same arXiv identifier, 1802.10016, for both [123] and [124], although they are cited as two distinct sources supporting different results: Theorem 2.15 (pathwise mild solutions for quasilinear SPDEs) is attributed to [123], and Theorem 3.15 (rough center manifolds) is attributed to [124]. As a survey whose central value is trustworthy attribution, the paper must give distinct, locatable references for these two results, or explicitly state that they are two papers sharing an identifier but distinguish them by title and version. Without this, the reader cannot verify either theorem from the reference list.","section":"Section 2.4, proof sketch after Theorem 2.15"}],"minor_comments":[{"comment":"The displayed solution formula in the proof sketch contains an apparent typo: it reads −∫_0^t A_v(s)U^v(t,s)A_v(s)∫_s^t g(u(r)) dW_U(r) ds, with A_v(s) appearing twice, whereas Definition 2.14 and the preceding derivation give a single A(u(s)) in that term. This should be corrected for consistency with (2.19).","section":"Section 3.3, paragraph after Eq. (3.15)–(3.16)"},{"comment":"The sentence 'Since H is finite-dimensional we obtain that S_c(t) := e^{tA_c} is a group' is inaccurate: H is an infinite-dimensional Hilbert space throughout the paper, and the relevant finite-dimensional object is the center subspace H_c. The wording should be changed to 'Since H_c is finite-dimensional'.","section":"Section 2.6, Remark 4"},{"comment":"There are two remarks labeled '4)' in Section 2.6: one about decomposing the integral into one-dimensional integrals and fractional calculus, and a later one about additive fractional noise. The second of these should be labeled '5)'.","section":"Section 3.2, metastability discussion"},{"comment":"The invariant measure written in (3.13) is described as 'technically ... to be understood more precisely using the Gaussian free field' but no precise definition is given. Since this is a survey, a citation to a rigorous treatment of the Gibbs measure for the Allen-Cahn SPDE would help the reader.","section":"Section 3.4, Remark 1"}],"recommendation":"major_revision","confidential_remarks":"The bibliographic duplication of [123] and [124] is a concrete, checkable problem and should be fixed before publication. The rate-function error in Section 3.2 is also a substantive mathematical mistake, though it is localized and repairable; the rest of the survey appears to be a fair representation of the literature. I recommend major revision rather than rejection because the central expository value of the paper can be restored with targeted corrections."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"I read the survey, and the reader's take mostly matches mine, with one caveat: the large-deviation issue in Section 3.2 is not just a localized typo. The displayed rate function F[0,T](γ) = (1/2)∫∫|∂tγ|^2 dx dt is the rate function for pure additive noise with no drift. For the Allen-Cahn equation (3.12), the Freidlin–Wentzell rate function must contain the deterministic drift: roughly (1/2)∫∫|∂tγ − (∂x^2γ + γ − γ^3)|^2 dx dt. As printed, the minimal cost of a transition from u ≡ −1 to u ≡ +1 over time horizon T is 2/T, which tends to zero as T grows. The paragraph explicitly uses this F to conclude that exit probabilities are exponentially small and to invoke the Kramers-type metastability asymptotics of [18]. That conclusion does not follow from the formula shown. This is a load-bearing error in a section advertised as covering metastability, one of the paper's central dynamical topics. It is fixable, but it must be fixed before the survey is used as a reference.\n\nThe classical core of the survey is genuinely good. Sections 2.1–2.3, 2.5, 2.7, 3.1, and most of 3.2 and 3.4 report standard results with mostly correct attributions. The didactic organization is a real service: a graduate student coming from deterministic PDEs can use this to get a map of mild, weak, strong, martingale, and variational solutions, the random field approach, rough paths, and regularity structures without drowning in the primary literature. The worked examples—especially Example 3.14 on the center manifold approximation—are helpful and correctly executed. The paper is honest about being a survey; the abstract says so and the content matches.\n\nTwo smaller issues are worth noting. The reference list assigns the same arXiv number, 1802.10016, to both [123] (pathwise mild solutions for quasilinear SPDEs) and [124] (rough center manifolds). That makes one of the two preprints effectively unlocatable from the bibliography. And Theorem 3.15, presented as an established result, rests on an arXiv preprint; the survey does give proof sketches, so I would not call it a fatal flaw, but the identifier conflation makes verification harder. The assertion in Remark 1 of Section 3.4 that no general random-attractor results exist for nonlinear multiplicative noise is a knowledge claim the survey cannot fully verify; the authors do cite [22] and [79] as partial exceptions, so this is a reasonable best-knowledge statement rather than a fabricated gap.\n\nOn the citation pattern: the authors cite themselves heavily in the recent-development sections, which is expected when they are among the few people working on rough-path dynamics and quasilinear SPDEs in this direction. Self-citation alone is not the problem here; the bibliography error is.\n\nWho is this for? A wide community, as the abstract says. A reader who wants a readable map of solution concepts and dynamical tools for SPDEs will get real value from the classical sections and the bibliography, once the reference list is corrected. The recent-development sections are useful as pointers but should be treated as preprints-in-progress rather than settled theory.\n\nRecommendation: yes, send this to a serious referee. A referee who knows SPDE dynamics can check the classical core quickly and will come back with a focused list of corrections, centered on Section 3.2 and the bibliography. The survey deserves that referee time. I would not cite it in my own work until the rate function is corrected and the preprint identifiers are disambiguated.","headline":"A useful survey of SPDE solution theory and dynamics with a solid classical core, but Section 3.2 prints an incorrect large-deviation rate function that undermines the metastability discussion, and the bibliography conflates two of the authors' own preprints under one arXiv identifier.","tokens_in":42456,"tokens_out":1956,"would_cite":false,"duration_ms":19454,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["60H15","37H05","35R60","60G22"],"pacs":[],"model":"deepseek-v4-flash","headline":"This survey argues that the right solution concept for an SPDE is dictated by the roughness of its noise, and that pathwise and rough-path methods now cover quasilinear equations and multiplicative fractional noise, enabling…","keywords":["stochastic partial differential equations","reaction-diffusion equations","solution theory","rough paths","random dynamical systems","center manifolds","random attractors","metastability"],"falsifier":"Compute the small-noise large-deviation rate function for the Allen-Cahn SPDE (3.12) on a transition path between the two stable constant states: the displayed $F_{[0,T]}(\\gamma) = \\frac{1}{2}\\int_0^T \\int_{\\mathbb{S}^1} |\\partial_t \\gamma(t,x)|^2 \\, dx \\, dt$ has an infimum of order $1/T$, whereas the cited sharp exit-time asymptotics require a barrier-height action independent of $T$, so a direct check of whether the printed LDP reproduces those asymptotics would settle the matter.","tokens_in":41146,"feed_emoji":"🎲","tokens_out":8728,"duration_ms":88385,"temperature":0.7,"pith_summary":"The paper is a survey, and its pith is a map: each type of stochastic forcing—additive, multiplicative, fractional, space-time white—selects a solution concept, and the choice of concept determines which dynamical tools can be applied. For semilinear equations the classical Itô mild, weak, and variational theories are mutually consistent; for quasilinear equations the paper reports a pathwise mild solution theory; for rough multiplicative noise it reports rough-path solutions and a local center-manifold theorem. The survey then connects these solution theories to stability, metastability, random invariant manifolds, and random attractors. A sympathetic reader takes this as a working roadmap: if the noise is additive or linear multiplicative, reduce to a random PDE and use random dynamical systems; if the noise is rough, use pathwise or rough-path methods; if the equation is singular, renormalize with regularity structures.","feed_headline":"Rough paths unlock SPDE dynamics under fractional noise","feed_subtitle":"A survey maps Itô, variational, and rough-path methods; center manifolds survive irregular multiplicative noise.","key_machinery":"The machinery is a sequence of named objects: metric dynamical systems with the Wiener shift, the cocycle property that makes a solution map a random dynamical system, pathwise mild solutions defined through a generalized stochastic convolution (Definition 2.14), rough-path solutions encoded by a path component $u$ and an area component $z$, and the Lyapunov-Perron map whose fixed point is the graph of the center manifold. For quasilinear equations the load-bearing identity is the variation-of-constants formula (2.19); for rough equations it is the pair of equations (2.32)-(2.33), built from the sewing lemma. Regularity structures, built from abstract symbols with parabolic homogeneity, play the same role for singular SPDEs.","core_discovery":"On the paper's own terms, the discovery is organizational: the correct solution theory for an SPDE is forced by the regularity of its noise, and the resulting map extends far enough to cover quasilinear and rough equations. For semilinear SPDEs with Lipschitz coefficients, the Itô-based mild, weak, and variational notions agree under the stated conditions (Theorems 2.8 and 2.13). For quasilinear SPDEs whose linear part is a sectorial random operator depending on the solution, an adapted Itô formulation fails because the evolution operator is not future-measurable, and the paper reports a unique local pathwise mild solution (Theorem 2.15). For multiplicative fractional noise with Hurst index in $(1/3,1/2]$, the paper presents a unique local rough-path solution pair $(u,z)$ (Theorem 2.17). On the dynamical side, it reports that pathwise and rough-path methods turn the SPDE into a random dynamical system and yield a local random center manifold (Theorem 3.15), while additive and linear multiplicative noise admit random attractors through reduction to random PDEs.","pith_inferences":["If the reported rough-path center-manifold theorem extends to stable and unstable manifolds, the discrete-time Lyapunov-Perron scheme would also provide a multiplicative ergodic theorem for fractional-noise SPDEs, something the survey lists as open.","The large-deviation discussion suggests a check: the displayed rate function for the Allen-Cahn SPDE omits the drift, so the printed formula alone would not produce the barrier-dependent metastability asymptotics attributed to the cited literature; a corrected rate function would make that section self-contained.","The claimed absence of general random-attractor results for nonlinear multiplicative noise is likely a temporary state: the pathwise random dynamical system from rough-path solutions supplies the cocycle, and the usual compactness estimates should carry over once a suitable absorbing set is found."],"forward_implications":["For semilinear SPDEs with Lipschitz coefficients, mild and variational solutions coincide, so dynamical results proved in one setting transfer to the other.","Quasilinear SPDEs with sectorial, Lipschitz-in-$u$ operators admit local pathwise mild solutions, opening cross-diffusion population models to the random-dynamical-systems toolbox.","SPDEs driven by fractional Brownian motion with Hurst index in $(1/3,1/2]$ and smooth multiplicative diffusion have unique local rough-path solutions, and the solution map is a genuine random dynamical system.","Under a spectral gap between center and stable directions, rough SPDEs have a local random center manifold, so finite-dimensional stochastic bifurcation analysis becomes possible.","For additive and linear multiplicative noise, pullback random attractors exist whenever the reduced random PDE has an absorbing compact set."],"supporting_citations":[{"why":"Supplies the classical Itô framework for mild, weak, strong, and martingale solutions, including the key existence theorem 2.8.","marker":"[51]"},{"why":"Provides the variational solution framework and the existence-uniqueness theorem 2.13.","marker":"[130]"},{"why":"Supports the unique local pathwise mild solution for quasilinear SPDEs stated as Theorem 2.15.","marker":"[123]"},{"why":"Gives the rough-path construction of local mild solutions for equation (2.26) and the sewing-lemma estimates behind Theorem 2.17.","marker":"[106]"},{"why":"Supports the existence of the local random center manifold stated as Theorem 3.15.","marker":"[124]"},{"why":"Provides the regularity-structures theory used for renormalized solutions of singular SPDEs like the Nagumo equation.","marker":"[99]"},{"why":"Establishes the random dynamical system and cocycle formalism underlying Sections 3.3 and 3.4.","marker":"[4]"},{"why":"Supplies the pullback attractor criterion stated as Theorem 3.21.","marker":"[156]"}],"fun_headline_variants":["Noise regularity picks the SPDE solution theory","Rough paths extend SPDE dynamics to fractional noise","Survey: Itô, variational, and rough-path SPDE methods","Center manifolds survive multiplicative fractional noise","SPDE solution theory matched to noise smoothness"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the results reported as established theorems, especially the quasilinear pathwise mild solution theorem and the rough center-manifold theorem, are correct and traceable to the cited sources.","fun_headline_variants_meta":{"raw":{"variants":["Noise regularity picks the SPDE solution theory","Rough paths extend SPDE dynamics to fractional noise","Survey: Itô, variational, and rough-path SPDE methods","Center manifolds survive multiplicative fractional noise","SPDE solution theory matched to noise smoothness"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000283,"raw_usage":{"total_tokens":1646,"prompt_tokens":897,"completion_tokens":749,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":513,"completion_tokens_details":{"reasoning_tokens":675}},"tokens_in":513,"tokens_out":749,"duration_ms":7058,"temperature":1.0,"reasoning_tokens":675,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:21:23.834380+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the small-noise large-deviation rate function for the Allen-Cahn SPDE (3.12) on a transition path between the two stable constant states: the displayed $F_{[0,T]}(\\gamma) = \\frac{1}{2}\\int_0^T \\int_{\\mathbb{S}^1} |\\partial_t \\gamma(t,x)|^2 \\, dx \\, dt$ has an infimum of order $1/T$, whereas the cited sharp exit-time asymptotics require a barrier-height action independent of $T$, so a direct check of whether the printed LDP reproduces those asymptotics would settle the matter.","supporting_citations":[{"cited_title":"Liu and M","cited_arxiv_id":null,"evidence_quote":"Provides the variational solution framework and the existence-uniqueness theorem 2.13."},{"cited_title":"Hesse and A","cited_arxiv_id":null,"evidence_quote":"Gives the rough-path construction of local mild solutions for equation (2.26) and the sewing-lemma estimates behind Theorem 2.17."},{"cited_title":"Pathwise mild solutions for quasilinear stochastic partial differential equations","cited_arxiv_id":"1802.10016","evidence_quote":"Supports the existence of the local random center manifold stated as Theorem 3.15."},{"cited_title":"Schmalfuß","cited_arxiv_id":null,"evidence_quote":"Supplies the pullback attractor criterion stated as Theorem 3.21."}],"review_version":1}