{"id":"c9a24d29-5cd5-45e1-88df-e7cc0ca532ff","arxiv_id":"2412.16479","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper proves existence and uniqueness of solutions to rough differential equations driven by planarly branched rough paths for roughness 1/4 < alpha <= 1/3 via Banach's fixed point theorem.","lead":"This paper proves a fixed point theorem for differential equations driven by planarly branched rough paths, extending a central result of rough path theory to a new roughness range. The result is relevant to stochastic analysis and to the algebraic theory of post-Lie Hopf algebras.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Fixed-point map (3.15) does not preserve the initial controlled lift: equation (3.12) gives <•_a, ∫_0^0 F(Y)dX> = F^a(Y_0), so M(Y)_0 = Y_0 forces F^a(Y_0)=0; for generic F no fixed point in D^3_{X;α} exists.","rationale":"The reader's verdict flagged the unspecified initial controlled lift and the asserted completeness of D^3_{X;α}. The most load-bearing problem, however, is sharper: the map M defined in (3.15) does not preserve the fiber of controlled paths with a given initial controlled lift, because the integral lift (3.12) has nonzero higher-order components at time 0. The assertion in Theorem 3.8 that 'MY0 = Y0 = Y0' is simply false unless F^a(Y_0)=0 for all a. This is an internal inconsistency in the central fixed-point argument, not a missing technical lemma. A concrete scalar example with F(x)=x and X_t=t makes the failure immediate at t=0. The paper contains substantial algebraic and analytic computations, and it is plausible that a corrected fixed-point formulation can be found, but the theorem and proof as written cannot be accepted. I therefore recommend REJECT rather than CONDITIONAL: the claimed statement is not established, and the required change is at the level of the basic setup, not a minor clarification.","tokens_in":48714,"tokens_out":21695,"duration_ms":188527,"concrete_test":"Take d=n=1, F(x)=x, choose any α∈(1/4,1/3], and let X be the canonical level-3 lift of the smooth path X_t=t. Set Y_0=1 and evaluate the •-component of equation (3.29) at t=0 using definition (3.12): the left side is Y^•_0, while the right side is Y^•_0 + F(1) = Y^•_0 + 1. The equality fails for every choice of the initial derivative component Y^•_0. Hence no Y∈D^3_{X;α} satisfies the fixed-point equation, contradicting Theorem 3.8. If the authors intended the integral lift to have zero initial higher-order components, the test reveals the missing subtraction in (3.12).","verdict_should_be":"REJECT","load_bearing_attack":"Equation (3.12) defines the controlled lift of the integral by <[τ]_a, ∫_0^t Z_r·dX_r> := <τ, Z^a_t>. In particular, for τ=1 with [1]_a = •_a, <•_a, ∫_0^t F(Y)_r·dX_r> = F^a(Y_t). At t=0 this equals F^a(Y_0), not 0. The fixed-point map (3.15) therefore satisfies <•_a, M(Y)_0> = Y^{•_a}_0 + F^a(Y_0). A fixed point Y = M(Y) would require <•_a, Y_0> = Y^{•_a}_0 + F^a(Y_0), i.e. F^a(Y_0) = 0 for every a. This is not a harmless regularity condition: for d=n=1, F(x)=x, and any initial value Y_0 ≠ 0, no element of D^3_{X;α} satisfies (3.29). The proof of Theorem 3.8 explicitly asserts 'MY0 = Y0 = Y0' immediately after (3.27), which is false for generic F and Y0. The obstruction is at t=0, before any contraction or small-δ argument; it is independent of the completeness of D^3_{X;α}. Thus the fixed-point problem as stated is internally inconsistent, and Theorem 3.9 cannot hold in its present form.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops controlled planarly branched rough paths at truncation level N=3 and proves a universal limit theorem for RDEs driven by planarly branched rough paths with Hölder exponent α∈(1/4,1/3]. The main technical contributions are norm estimates for composing a controlled path with a C^3_b function (Theorems 2.8 and 2.9), a sewing-lemma construction of the rough integral (Theorem 3.2), and a Banach fixed point argument for local existence and uniqueness (Theorem 3.8), patched to [0,T] in Theorem 3.9.","tokens_in":49054,"tokens_out":14357,"duration_ms":118791,"significance":"If the proof can be completed, the result extends the universal limit theorem from geometric, classical, and branched rough paths to planarly branched rough paths in the roughness range (1/4,1/3], which is a natural step given the post-Lie algebraic structure. The paper's detailed estimates in Sections 2.2–2.3 and the explicit sewing argument in Section 3.1 are substantial, self-contained, and parameter-free. However, the fixed-point framework has unresolved issues concerning the initial controlled data; these issues are central to the claimed existence and uniqueness and must be fixed before the main result is established.","major_comments":[{"comment":"The assertion 'MY0 = Y0 = Y0' immediately after (3.27) is not justified. By (3.12) with τ=1, ⟨•_a, ∫_0^0 F(Y)·dX⟩ = F^a(Y_0), so the initial value of M(Y) in the full controlled sense differs from the constant path Y0 by the nonzero component F^a(Y_0) unless F^a(Y_0)=0 for all a. Thus the ball Bδ(Y,R), defined with the condition Y0=Y0, is not invariant under M, and the Banach fixed point theorem cannot be applied to this ball as written. This affects the existence part of Theorem 3.8 and hence Theorem 3.9.","section":"§3.2, Eq. (3.15) and (3.27), proof of Theorem 3.8"},{"comment":"Lemma 3.7 assumes Y0 = ˜Y0, i.e., equality of the full initial controlled values. In the uniqueness proof, the restrictions Y|[σ,σ+ε] and ˜Y|[σ,σ+ε] are known to satisfy equality only for the underlying path value at σ; the higher-order initial components need not agree. Therefore the application of Lemma 3.7 in the inequality preceding (3.30) is unsupported. The proof must either show that the full initial controlled lift is uniquely determined by the equation, or replace Lemma 3.7 by a statement that controls differences in the higher initial components.","section":"§3.2, Lemma 3.7 and Theorem 3.9 (uniqueness)"},{"comment":"The claim that (D^N_{X;α}, |||·|||_{X;α}) is a Banach space is asserted with reference to [15] but not proved. Since the Banach fixed point theorem is the central tool, this completeness statement is load-bearing and should be proved or made precise, including the role of initial data, which is not fixed in the definition of D^N_{X;α}.","section":"§2.1, after (2.8)"},{"comment":"The definition of the controlled lift of the integral is incomplete. The symbol [τ]_a is not defined, the indexing of forest components τ_1···τ_n is not specified, and no verification is given that the resulting path belongs to D^3_{X;α}; in particular, the remainder estimates for components such as rr_ab and r∨r r, which appear in (3.5), are not checked. A reference to [15] is not sufficient because the planar setting has additional forest components.","section":"§3.1, (3.11)–(3.12)"},{"comment":"The proof states that δ does not depend on the initial condition Y0, but the bounds in Lemma 3.6 and Lemma 3.7 depend on the initial higher-order components Y^{•_a}_0, Y^{•_a•_b}_0, and Y^{rr_ab}_0. Without a uniform bound on these components (or an argument showing they are determined by Y0=ξ), δ cannot be taken uniformly over the patching steps, and the global existence on [0,T] is not established.","section":"§3.2, Theorem 3.9 (patching)"}],"minor_comments":[{"comment":"The metric |||Y, ˜Y|||_{X, ˜X;α} is defined as |||Y−˜Y|||_{X;α}; when X≠˜X the difference of two controlled paths with respect to different rough paths is not an element of a single normed space. Please clarify the definition.","section":"§2.1, Eq. (2.8)"},{"comment":"In the final display of Step 4, the expression RZ^{•_a•_b}−R˜Z^{•_a•_b} is written as RZ^{•_a•_b}−R˜Z^{•_a•_a}; this appears to be a typo.","section":"§2.3, Theorem 2.9, final display"},{"comment":"The notations F^a(Y_s)1 and F^a(Y_s)^{•_b} are not explained; they should be written as the corresponding components of the controlled path F^a(Y).","section":"§3.2, Eq. (3.16)"},{"comment":"The definition σ := sup{t≥0 | Y_t = ˜Y_t on [0,T]} is ill-posed; it should likely be sup{t≥0 | Y_s=˜Y_s for all s∈[0,t]}.","section":"§3.2, Theorem 3.9"},{"comment":"Reference [13] is a lecture-note manuscript rather than a published source; please cite a published version if available.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The main risk is the handling of initial controlled data in the fixed-point argument. The algebraic estimates in Sections 2.2–2.3 appear substantial and likely reusable. If the authors reformulate the fixed-point problem with the full initial controlled jet as part of the data, as is standard for controlled rough paths, the result may be correct. In its present form, the proof of the main theorem is not complete."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things about arXiv:2412.16479. First, it is the first serious attempt at a universal limit theorem for planarly branched rough paths in the roughness window (1/4,1/3], and the composition and stability estimates in Section 2 are detailed and largely credible. Second, the fixed-point map at the center of the proof is internally inconsistent; as written, it cannot have a fixed point for generic smooth F and nontrivial initial data.\n\nThe issue is at t=0. Equation (3.12) defines the controlled lift of an integral by <•_a, ∫_0^t Z·dX> = <1, Z^a_t>, so in particular <•_a, ∫_0^t F(Y)·dX> = F^a(Y_t). The map M(Y)=Y_0+∫_0^• F(Y)·dX then has <•_a, M(Y)_0> = Y_0^{•_a} + F^a(Y_0). A fixed point Y=M(Y) forces F^a(Y_0)=0 for every a. This is not a harmless condition: for d=n=1, F(x)=x, and any ξ≠0, no element of D^3_{X;α} satisfies the equation. The proof of Theorem 3.8 explicitly uses MY0=Y0, which is false. The obstruction is before any contraction or small-δ argument, so it does not depend on the completeness gap (which is real but secondary).\n\nWhat the paper does well: the algebra for the composition of a controlled planarly branched path with a regular function is worked out in detail, with explicit bounds. The integration theorem via the sewing lemma (Theorem 3.2) appears sound. The positioning with respect to the post-Lie Hopf algebra literature is honest and the references are relevant.\n\nWhat needs fixing: the fixed point should be set on the base path (or on controlled paths with the initial derivative determined by the equation), not on the full controlled-lift space with an arbitrary initial lift. The authors also need to prove completeness of D^3_{X;α} and specify the initial controlled data in Definition 3.4 and Theorem 3.9. The patching step needs a δ uniform in the initial data, which is not shown.\n\nWho it is for: researchers working on rough DEs on homogeneous spaces and post-Lie Hopf algebras. The computations may be worth refereeing even if the current main theorem fails.\n\nMy recommendation: send it to peer review rather than desk-reject. A serious referee can identify the fixed-point issue and the authors may be able to repair it. As it stands, though, the universal limit theorem is not proven in this paper.","headline":"First planarly branched universal limit theorem attempt, but the fixed-point map is inconsistent at t=0 and the main theorem as stated fails.","tokens_in":49570,"tokens_out":10928,"would_cite":false,"duration_ms":90495,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["60L20","60L50","60H99","34K50","37H10","05C05"],"pacs":[],"model":"deepseek-v4-flash","headline":"Unique solutions for RDEs driven by planarly branched rough paths","keywords":["planarly branched rough paths","universal limit theorem","controlled rough paths","Banach fixed point theorem","post-Lie algebra","rough differential equations","Munthe-Kaas-Wright Hopf algebra","Holder roughness"],"falsifier":"Construct a sequence of controlled planarly branched rough paths satisfying Definition 2.5 whose remainder norms tend to zero but whose pointwise limit fails the required remainder estimates, which would disprove completeness and remove the guaranteed fixed point; alternatively, exhibit an explicit pair of controlled paths for which the contraction constant in Lemma 3.7 cannot be made smaller than one on any time interval of positive length.","tokens_in":48488,"feed_emoji":"📐","tokens_out":3053,"duration_ms":30645,"temperature":0.7,"pith_summary":"This paper establishes the universal limit theorem for planarly branched rough paths with roughness in the range half-open interval from one quarter to one third, the first such result for this class of drivers. It proves that rough differential equations driven by such paths admit a unique solution on the whole time interval, extending earlier existence-and-uniqueness theorems from classical, geometric, and branched rough paths. The proof works by lifting the equation to a space of controlled planarly branched rough paths and applying the Banach fixed point theorem, after verifying that composition with smooth vector fields preserves the controlled structure and satisfies explicit norm bounds. If the argument is correct, it closes the planarly branched case at truncation level three and provides the analytic foundation for solving rough differential equations on homogeneous spaces whose tangent structure is a post-Lie algebra.","feed_headline":"Unique solutions for RDEs driven by planarly branched rough paths","feed_subtitle":"A fixed-point proof extends existence and uniqueness into the one-quarter-to-one-third Holder regime for post-Lie drivers.","key_machinery":"The central objects are controlled planarly branched rough paths: paths taking values in the truncated Munthe-Kaas-Wright Hopf algebra $H^{{<=2}}$_{MKW}, whose increments are required to match the driving rough path X through the coproduct, with remainder terms RY of orders 3alpha, 2alpha, and $\\alpha$. The argument is carried by the norm |||Y|||_{X;$\\alpha$} on the space $D^{3}$_{X;$\\alpha$} of such controlled paths, together with Theorems 2.8 and 2.9, which bound the norm and stability of the composition Z = F(Y) with a $C^{3}$_b function F. Integration against X is defined through the sewing lemma, and the fixed-point map M(Y) = Y_0 + integral_0^bullet F(Y_r) dX_r is shown to map a small metric ball into itself and to be a contraction, yielding local existence and uniqueness.","core_discovery":"The central claim is Theorem 3.9: for any Holder exponent $\\alpha$ in (1/4, 1/3] and any planarly branched rough path X in $PBRP^{3}$_alpha above a path in R^d, there is a unique controlled planarly branched rough path Y in $D^{3}$_{X;$\\alpha$} solving the integral equation Y_t = Y_0 + integral_0^t F(Y_r) dX_r for all t in [0,T]. This is the planarly branched analogue of the universal limit theorem, proved here by a fixed-point argument rather than by regularity-structure methods. The proof's load-bearing steps are a norm bound for the controlled path obtained by composing a controlled planarly branched rough path with a smooth function, a stability estimate comparing two such compositions, and a contraction estimate for the fixed-point map on a small time interval, after which the solution is extended step by step to the full interval.","pith_inferences":["The completeness of D^3_{X;alpha}, asserted by analogy with the branched case, is a genuine gap: if the space were incomplete, the contraction mapping would have no guaranteed fixed point, so a direct completeness proof would harden the theorem.","The same composition-and-contraction scheme should extend to any connected graded Hopf algebra whose coproduct satisfies estimates comparable to Theorems 2.8 and 2.9; the level-three planar case is the first nontrivial test.","A natural testable extension is to reformulate the result in the 1/alpha-variation topology, matching the setting of the branched universal limit theorem and allowing driving paths with jumps.","The post-Lie structure underlying planarly branched rough paths suggests that the solution map should interact well with Lie group integrators on homogeneous spaces, though the paper itself does not develop that application."],"forward_implications":["The universal limit theorem now covers planarly branched rough paths in the three-step regime, alongside the classical, geometric, and branched cases.","The fixed-point construction gives a genuine existence-and-uniqueness statement for rough differential equations on manifolds whose tangent vector fields form a post-Lie algebra.","The stability estimates imply that solutions depend continuously on the driving rough path and on the initial lift, because differences of solutions are controlled by the distance between drivers and initial data.","The local-to-global extension argument shows that the solution exists up to any fixed time T without imposing smallness on T, provided the vector fields are sufficiently regular and bounded.","The proof provides a template for extending the universal limit theorem to arbitrary truncation levels within connected graded Hopf algebras, as the authors indicate they will do in a forthcoming paper."],"supporting_citations":[{"why":"Introduces rough paths and the original universal limit theorem for weakly geometric drivers.","marker":"[22]"},{"why":"Introduces branched rough paths and controlled branched rough paths, the conceptual template for the paper's controlled objects.","marker":"[12]"},{"why":"Establishes the universal limit theorem for branched rough paths in the variation setting, the result being extended here.","marker":"[10]"},{"why":"Defines planarly branched rough paths and studies rough differential equations on homogeneous spaces, supplying the motivating framework.","marker":"[5]"},{"why":"Constructs the Munthe-Kaas-Wright Hopf algebra of planar rooted forests on which planarly branched rough paths are based.","marker":"[26]"},{"why":"Supplies the controlled rough path framework and the sewing lemma used to define integration against X.","marker":"[11]"},{"why":"Gives the sewing lemma in the form used to construct the rough integral of controlled paths.","marker":"[7]"},{"why":"Provides the Banach space theory for controlled branched rough paths, cited for the analogous completeness assertion.","marker":"[15]"}],"fun_headline_variants":["Universal limit theorem for planarly branched rough paths","Fixed-point proof for RDEs with planarly branched drivers","Unique solutions for RDEs on planarly branched rough paths","Extending universal limit theorem to planarly branched case"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The Banach fixed point argument assumes that the space of controlled planarly branched rough paths, with its chosen norm, is complete, but the paper only asserts this by analogy with the branched case and gives no proof.","fun_headline_variants_meta":{"raw":{"variants":["Universal limit theorem for planarly branched rough paths","Fixed-point proof for RDEs with planarly branched drivers","Unique solutions for RDEs on planarly branched rough paths","Extending universal limit theorem to planarly branched case"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000281,"raw_usage":{"total_tokens":1655,"prompt_tokens":924,"completion_tokens":731,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":540,"completion_tokens_details":{"reasoning_tokens":665}},"tokens_in":540,"tokens_out":731,"duration_ms":6066,"temperature":1.0,"reasoning_tokens":665,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T10:33:11.846782+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a sequence of controlled planarly branched rough paths satisfying Definition 2.5 whose remainder norms tend to zero but whose pointwise limit fails the required remainder estimates, which would disprove completeness and remove the guaranteed fixed point; alternatively, exhibit an explicit pair of controlled paths for which the contraction constant in Lemma 3.7 cannot be made smaller than one on any time interval of positive length.","supporting_citations":[{"cited_title":"Lyons, Differential equations driven by rough signals, Rev","cited_arxiv_id":null,"evidence_quote":"Introduces rough paths and the original universal limit theorem for weakly geometric drivers."},{"cited_title":"Gubinelli, Ramification of rough paths, Journal of Di fferential Equations, 248(4) (2010), 693-721","cited_arxiv_id":null,"evidence_quote":"Introduces branched rough paths and controlled branched rough paths, the conceptual template for the paper's controlled objects."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the universal limit theorem for branched rough paths in the variation setting, the result being extended here."},{"cited_title":"Curry, K","cited_arxiv_id":null,"evidence_quote":"Defines planarly branched rough paths and studies rough differential equations on homogeneous spaces, supplying the motivating framework."},{"cited_title":"Munthe-Kaas and W.M","cited_arxiv_id":null,"evidence_quote":"Constructs the Munthe-Kaas-Wright Hopf algebra of planar rooted forests on which planarly branched rough paths are based."},{"cited_title":"Gubinelli, Controlling rough paths, Journal of Functional Analysis, 216(1) (2004), 86-140","cited_arxiv_id":null,"evidence_quote":"Supplies the controlled rough path framework and the sewing lemma used to define integration against X."},{"cited_title":"Feyel and A","cited_arxiv_id":null,"evidence_quote":"Gives the sewing lemma in the form used to construct the rough integral of controlled paths."},{"cited_title":"Kelly, It ˆo corrections in stochastic equations, Ph.D","cited_arxiv_id":null,"evidence_quote":"Provides the Banach space theory for controlled branched rough paths, cited for the analogous completeness assertion."}],"review_version":1}