{"id":"296c4887-de00-4ba4-be43-6ec9b8e4f1bd","arxiv_id":"2608.09700","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A general paracontrolled ansatz built from decorated trees and iterated paraproducts gives local well-posedness and BPHZ renormalisation for a broad class of subcritical parabolic singular SPDEs.","lead":"This paper builds a general recipe for solving a broad class of noisy nonlinear equations, extending the paracontrolled method to many singular SPDEs at once. It gives a way to write solutions as a main controlled term plus a smooth remainder, with proofs of local existence and renormalisation.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.1 as stated is not proved: the Section 6 contraction requires the stochastic data to lie in a strictly more regular space C^{|τ|'} than the statement assumes.","rationale":"The paper's central goal is a general paracontrolled solution theory for subcritical parabolic singular SPDEs. The main existence result, Theorem 1.1, is proved through the fixed point argument of Section 6. That argument explicitly supposes the stochastic objects are more regular than their natural degree, introducing a modified degree |·|′ and using (6.3) to obtain a decay factor 2^{-aε0} in the contraction estimates. Neither Theorem 1.1 nor Proposition 6.3 states this stronger regularity condition for the admissible stochastic data. This is not a disagreement with an external consensus but an internal mismatch: the proof's hypotheses are stronger than the theorem's statement. The reader's verdict identified exactly this point. The concern is load-bearing because without the extra margin, or without a replacement estimate using the gap between |I(τ)| and the integer k in the contraction, the map Θ^T in Proposition 6.3 is not shown to be a contraction. The paper may be repairable, and the BPHZ result in Section 8 can supply the margin in the renormalized setting, but Theorem 1.1 as written is not proved for arbitrary admissible data. The appropriate verdict remains CONDITIONAL: the framework is plausible and likely correct after either adding the missing hypothesis or patching the fixed point estimate.","tokens_in":78971,"tokens_out":16967,"duration_ms":154375,"concrete_test":"Re-derive the contraction estimate in Proposition 6.2 without using the modified degree: for each tree in the sum k<|I(τ)|<α, apply (6.2) with r1=k and r2=|I(τ)| to get ∥Z^{>a}_{I(τ)}∥_{C^k} ≤ 2^{-a(|I(τ)|-k)}∥Z_{I(τ)}∥_{C^{|I(τ)|}}. Check whether the resulting contraction constant can be made <1 by choosing a large, uniformly over the finite tree set and the relevant ball. If yes, Theorem 1.1 needs only a local rephrasing of Section 6; if no, the missing C^{|τ|′} hypothesis must be added to Theorem 1.1 or derived from the BPHZ margin for all admissible data.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 6's fixed point argument introduces a modified degree |·|′ and relies on the dyadic truncation estimate (6.3), ∥Z^{>a}_τ∥_{C^{|τ|}} ≲ 2^{-aε0}∥Z_τ∥_{C^{|τ|′}}, to make the map Λ^{k,α} in Proposition 6.2 a contraction and to close the estimates in Proposition 6.3. But Theorem 1.1 and Proposition 6.3 are stated for arbitrary admissible stochastic data with Zτ ∈ C^{|τ|}; no C^{|τ|′} hypothesis appears there. For data that are only known to lie in C^{|τ|}, the high-frequency truncation does not satisfy an estimate with a positive ε0 at the same exponent |τ|, so the displayed contraction bound is unjustified. Theorem 1.2 provides extra regularity only for the BPHZ-renormalized data, not for the arbitrary data quantified in Theorem 1.1. This is an internal statement-proof mismatch: the existence and continuity theorem is not established as stated.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a general paracontrolled ansatz for a large class of subcritical parabolic singular SPDEs. The solution is expanded as a sum of paraproducts of coefficient functions (elementary differentials) with stochastic reference distributions indexed by decorated trees. The stochastic data are constructed recursively via a new word-based regularity structure and a generalized Hairer–Kelly map. The paper also provides a fixed-point argument for local well-posedness, a renormalization scheme via preparation maps, and a BPHZ convergence result for the renormalized stochastic data using a spectral gap inequality. The two main theorems are Theorem 1.1 (local existence, uniqueness, and continuity for the paracontrolled expansion) and Theorem 1.2 (L^q convergence of BPHZ-renormalized stochastic data, independent of the mollifier).","tokens_in":79180,"tokens_out":8862,"duration_ms":77394,"significance":"If the results are correct, this is a substantial contribution: it gives the first general paracontrolled solution theory with a scope comparable to the decorated-tree ansatz in regularity structures, and it includes a systematic algebraic framework for stochastic data, renormalization, and convergence. The paper contains detailed, parameter-free algebraic derivations, two worked examples (gPAM and Phi^4_3), and a useful symbolic index. No machine-checked proofs are provided, but the arguments are explicit and largely self-contained apart from the cited 'to appear' works. However, as detailed below, the local well-posedness theorem currently has an unstated extra-regularity hypothesis, and the proof of the convergence theorem has a gap in the treatment of expectations.","major_comments":[{"comment":"Theorem 1.1 and Proposition 6.3 are stated for arbitrary admissible stochastic data with Z_tau in C^{|tau|}, but the fixed-point construction in Proposition 6.2 explicitly assumes the stronger regularity Z_tau in C^{|tau|'} with |tau|' > |tau| and uses the estimate (6.3), ||Z_tau^{>a}||_{C^{|tau|}} <= C 2^{-a epsilon_0} ||Z_tau||_{C^{|tau|'}}. The text in Section 6 acknowledges this: 'we will have to suppose that the stochastic objects (Z_I(tau))_tau are slightly more regular than the regularity given by their degree |I(tau)|.' No argument is given that all admissible data satisfy (6.3); Theorem 1.2 provides an extra margin only for the BPHZ-renormalised data under the spectral gap assumption, not for the arbitrary data quantified in Theorem 1.1. Therefore the local well-posedness claim is not proved as stated. The statement of Theorem 1.1 (and Proposition 6.3) must be revised to include the extra-regularity hypothesis, or a proof that admissible data automatically carry such a margin must be supplied.","section":"Section 6 (Eq. (6.3), Propositions 6.2 and 6.3; Theorem 1.1)"},{"comment":"The proof of Proposition 8.13 establishes convergence of the differences bZ^m1_tau - bZ^m2_tau via Lemma 8.10 and Lemma 8.12, but the treatment of the expectation term is only sketched. The assertion that |E[bZ^m1_tau - bZ^m2_tau]| tends to zero is justified by saying that bZ^m_tau = bPi^m_tau + F(bZ^m_sigma, sigma in T cup T^+) and that F is continuous in the bZ^m_sigma for smaller sigma; the functional form of F and its Lipschitz continuity are not spelled out. Since Theorem 1.2 asserts convergence in L^q for all q, the convergence of the first moment is load-bearing and should be proved explicitly, including the independence of the mollifier.","section":"Section 8.5 (Proposition 8.13, Theorem 1.2)"}],"minor_comments":[{"comment":"Please correct the typos 'renormsalition' in the abstract and 'paraliearisation' in the Section 5 heading; the spelling of 'renormalised/renormalization' is also inconsistent throughout.","section":"Abstract and Section 5 heading"},{"comment":"The proofs of Theorem 3.12, Proposition 8.2, and several propositions in Section 5 rely on [18, Theorem 1] and its variants, but [18] is listed as 'to appear'. Please state the precise results used or include them in an appendix so that the paper is self-contained.","section":"Reference [18]"},{"comment":"In the contraction proof, the choices of a and R are described as 'big enough' without explicit quantitative conditions; making the dependence of the contraction constant on epsilon_0, R, and the stochastic data explicit would improve readability.","section":"Section 6 (Proposition 6.2)"},{"comment":"The definition of admissibility refers to equations (4.2) and (4.7), but (4.2) was introduced for tau in T^+; please clarify the domain of the condition for tau in T.","section":"Definition 4.2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript depends heavily on the authors' own 'to appear' papers [18] and [19]; the editor may wish to verify that these are accessible to the referees. The main issue is the Section 6 hypothesis gap; the algebraic framework appears coherent and the two examples are well chosen."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear [Colleague],\n\nThe paper delivers on its headline: it gives a general paracontrolled ansatz that, for the first time, has scope comparable to regularity structures in the subcritical parabolic regime. The construction is genuinely new and technically rich. The stochastic data are built recursively via a word-based encoding, using a generalized Hairer-Kelly map (Def 3.8) that sends decorated trees to words of iterated paraproducts. The paralinearisation (Thm 5.11) and the equivalence with a paracontrolled system (Prop 5.13) are the heart of the paper and they look correct. The renormalisation section (preparation maps, Prop 7.8) and the BPHZ convergence via the spectral gap (Thm 8.5, adapted from Bailleul-Hoshino) are carefully done; the key identity δ bZ_τ = Y^{bZ}_{j*Ψ(D_Ξτ)} is new and the proof outline is plausible. That the framework recovers gPAM and Φ4_3 is a reassuring sanity check.\n\nThe main problem is a genuine statement-proof mismatch in Section 6. The fixed-point argument that proves Prop 6.3 (and hence Theorem 1.1) requires the stochastic data to be slightly more regular than their natural Hölder degree. Section 6 says exactly this: it introduces a modified degree |·|' with ε0>0 and uses estimate (6.3) to make the map Λ^{k,α} a contraction. Proposition 6.2 states that hypothesis. But Theorem 1.1 and Prop 6.3 are stated for arbitrary admissible stochastic data in C^{|τ|}, with no mention of the extra margin. For data only known in C^{|τ|}, the truncated objects Z^{>a}_τ do not decay in the C^{|τ|} norm, so the contraction estimate does not follow. This is not a cosmetic gap in the write-up; as stated, the existence theorem is not established by the given proof.\n\nThe fix is straightforward: either add the extra regularity as an explicit hypothesis in Theorem 1.1/Prop 6.3 (e.g., data in C^{|τ|'} for some ε0>0), or state that the theorem applies to the BPHZ-renormalized data, whose convergence is proved in Section 8 with a small positive margin (Prop 8.13 gives convergence in B^{|τ|(κ)} for small κ). Since the modified degree is already introduced in Section 6, this is a matter of quantifying the theorem correctly.\n\nTwo further points. The paper leans heavily on the unpublished-to-appear work [18] by one of the authors for the local expansion theory of iterated paraproducts. That is not circular, but it means part of the proof cannot be fully checked from the present text. And the notation is dense, even by the standards of this field. But the algebraic structure is coherent, the derivations are parameter-free, and the examples match known results.\n\nThe paper is for specialists in singular SPDEs, particularly those working on paracontrolled calculus and global existence. I would bring it to the reading group, and I would cite it. It deserves a serious referee; the fixed-point issue is checkable and patchable, and the rest of the framework is worth careful scrutiny. Send it to review.","headline":"A strong, genuinely general paracontrolled framework, but the leading existence theorem is overclaimed: the proof needs a positive regularity margin on the data that is not stated.","tokens_in":79743,"tokens_out":6922,"would_cite":true,"duration_ms":57872,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["60H17","35R60","60H07","46E35","16T05"],"pacs":[],"model":"deepseek-v4-flash","headline":"A recursive ansatz encoded by decorated trees provides a general paracontrolled solution theory for a large class of singular SPDEs, with mollifier-independent convergence of the renormalised stochastic data.","keywords":["paracontrolled calculus","singular SPDEs","regularity structures","decorated trees","renormalisation","BPHZ","spectral gap inequality","Besov spaces"],"falsifier":"A concrete check is to compute the BPHZ limit of the reference distribution for a borderline tree such as Ξℓ I(Ξℓ) in a near-critical example; if the limit belongs to $C^{{|τ|}}$ but not to $C^{{|τ|+ε}}$ for any ε > 0, then the estimate (6.3) has ε0 = 0 and the contraction of Proposition 6.3 is no longer guaranteed.","tokens_in":78714,"feed_emoji":"📐","tokens_out":8290,"duration_ms":76453,"temperature":0.7,"pith_summary":"The paper claims that the paracontrolled approach to singular stochastic PDEs can be organised once and for all rather than equation by equation: every solution is written in a universal recursive form, a sum of paraproducts of explicit coefficient functions with stochastic reference distributions indexed by decorated trees, plus a smooth remainder. It proves that for a large class of subcritical parabolic equations this ansatz is closed under the solution map, giving local existence, uniqueness, and continuous dependence on the initial condition and the stochastic data. It also proves that the BPHZ-renormalised stochastic data converge in L^q independently of the mollifier. A sympathetic reader would care because this turns a powerful but bespoke technique into a systematic theory whose algebraic structure is fixed by decorated trees.","feed_headline":"A recursive tree ansatz now covers a broad class of singular SPDEs","feed_subtitle":"Decorated trees describe the expansion; BPHZ renormalised data converge independently of the mollifier.","key_machinery":"The load-bearing object is the generalised Hairer-Kelly map Ψ, a recursive map defined from the coaction on decorated trees that sends each tree τ to a linear combination of decorated words ⟨τ1,…,τn⟩^k_ℓ X^r; on these words one builds the regularity structure of iterated paraproducts, with model Π^Z⟨τ1,…,τn⟩ = P_ℓ(Z_{τ1},…,Z_{τn}). This map carries the argument because it converts paralinearisation into a reconstruction problem: the modified model (eΠ, eg) of a tree equals the word model applied to Ψ, so the paracontrolled representation of any modelled distribution follows from the reconstruction theorem. The convergence proof runs on a second structure, a regularity-integrability structure over words, together with the identity δbZτ = $Y^{{bZ}}$_{j*Ψ(DΞτ)} that links the Malliavin derivative of the renormalised stochastic data to the word model.","core_discovery":"The central discovery is that the entire paracontrolled expansion can be generated from two ingredients: a generalised Hairer-Kelly map Ψ that sends decorated trees into decorated words, and recursive definitions that build the reference distributions Zτ from convolutions with the kernel, pointwise products with the noises, and iterated paraproducts of strictly smaller trees. The paper establishes that these Zτ form an admissible stochastic data set, that the paracontrolled ansatz is equivalent to a full paracontrolled system, and that the BPHZ-renormalised data converge in L^q independently of the mollifier (Theorem 1.2). In particular, Theorem 1.1 provides a unique local solution of the regularised renormalised equation whose paracontrolled expansion has remainder in C^γ and depends continuously on the initial condition and the stochastic data.","pith_inferences":["Beyond the paper, the same word structure could serve as a dictionary between paracontrolled models and regularity-structure models, since Ψ identifies the algebraic skeleton shared by both expansions.","The recursive combinatorics appears independent of the smoothing gain β, so adapting the ansatz to dispersive equations would mainly require replacing the convolution and paraproduct estimates, a step the paper suggests but does not carry out.","A testable consequence of Theorem 1.2 is that any two mollifiers produce the same renormalised solution limit; comparing the BPHZ data numerically for a concrete model would check this directly.","If the extra-regularity margin needed for the fixed point is supplied by the BPHZ regularity gain, the ansatz could be applied to global-in-time constructions across the full subcritical regime, extending the global results known for particular models."],"forward_implications":["For every subcritical parabolic equation in the class, the local solution has the same structural description: a finite paracontrolled expansion with coefficients given by elementary differentials and a C^γ remainder.","The BPHZ-renormalised stochastic data converge in L^q for every q ≥ 1 and the limit is independent of the mollifier, so the renormalised objects can be defined canonically.","The fixed-point construction gives continuity of the solution and its paracontrolled remainder with respect to both the initial condition and the stochastic data.","The general ansatz specialises to the known paracontrolled treatments of gPAM and Φ4_3 and recovers their standard enhanced noise sets."],"supporting_citations":[{"why":"Supplies the original paracontrolled framework that this paper generalises to an arbitrary subcritical equation.","marker":"[33]"},{"why":"Provides the regularity structure machinery, including models, reconstruction, and multilevel Schauder estimates used throughout.","marker":"[37]"},{"why":"Gives the decorated-tree ansatz and renormalised equation in regularity structures that the present paracontrolled ansatz adapts to the word setting.","marker":"[6]"},{"why":"Provides the word regularity structure and local expansion results for iterated paraproducts on which the construction of the Zτ rests.","marker":"[18]"},{"why":"Supplies regularity-integrability structures and the spectral gap framework used in Section 8 for the convergence proof.","marker":"[14]"},{"why":"Introduces the diagram-free spectral gap approach to stochastic estimates that is adapted here to decorated trees and words.","marker":"[48]"},{"why":"Extends the spectral gap and BPHZ analysis to decorated trees, serving as the baseline for Theorem 1.2.","marker":"[45]"},{"why":"Gives the paracontrolled representation of modelled distributions and the extension procedure used to build admissible stochastic data.","marker":"[12]"},{"why":"Supplies the high-order paracontrolled calculus, especially star derivatives and the generalised derivatives used for the coefficients.","marker":"[3]"},{"why":"Establishes BPHZ renormalisation and preparation maps in regularity structures, the formalism used for the renormalised equation.","marker":"[16]"}],"fun_headline_variants":["Recursive tree ansatz tames singular SPDEs broadly","BPHZ renormalised data converge for paracontrolled SPDEs","Generalised Hairer–Kelly map lifts paraproducts in SPDEs","Recursive definition yields full paracontrolled solution theory"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The contraction argument in Section 6 needs the stochastic data to be a little more regular than their natural Hölder degree; if that extra margin vanishes, the fixed-point map is not shown to contract.","fun_headline_variants_meta":{"raw":{"variants":["Recursive tree ansatz tames singular SPDEs broadly","BPHZ renormalised data converge for paracontrolled SPDEs","Generalised Hairer–Kelly map lifts paraproducts in SPDEs","Recursive definition yields full paracontrolled solution theory"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000296,"raw_usage":{"total_tokens":1692,"prompt_tokens":896,"completion_tokens":796,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":512,"completion_tokens_details":{"reasoning_tokens":718}},"tokens_in":512,"tokens_out":796,"duration_ms":6625,"temperature":1.0,"reasoning_tokens":718,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T12:24:38.782477+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete check is to compute the BPHZ limit of the reference distribution for a borderline tree such as Ξℓ I(Ξℓ) in a near-critical example; if the limit belongs to $C^{{|τ|}}$ but not to $C^{{|τ|+ε}}$ for any ε > 0, then the estimate (6.3) has ε0 = 0 and the contraction of Proposition 6.3 is no longer guaranteed.","supporting_citations":[{"cited_title":"Bruned, A","cited_arxiv_id":null,"evidence_quote":"Gives the decorated-tree ansatz and renormalised equation in regularity structures that the present paracontrolled ansatz adapts to the word setting."},{"cited_title":"Local expansion properties of paracontrolled systems","cited_arxiv_id":"2412.12670","evidence_quote":"Provides the word regularity structure and local expansion results for iterated paraproducts on which the construction of the Zτ rests."},{"cited_title":"Hairer, R","cited_arxiv_id":null,"evidence_quote":"Extends the spectral gap and BPHZ analysis to decorated trees, serving as the baseline for Theorem 1.2."},{"cited_title":"Bailleul, M","cited_arxiv_id":null,"evidence_quote":"Gives the paracontrolled representation of modelled distributions and the extension procedure used to build admissible stochastic data."},{"cited_title":"Bruned, M","cited_arxiv_id":null,"evidence_quote":"Establishes BPHZ renormalisation and preparation maps in regularity structures, the formalism used for the renormalised equation."}],"review_version":1}