{"id":"7c788e11-3e61-4d72-9b17-d26d6f43605f","arxiv_id":"2608.23250","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Parabolic Ahlfors-David regular sets are parabolic uniformly rectifiable exactly when they have big pieces of regular parabolic (bi-)Lipschitz images of n-dimensional space-time.","lead":"This paper proves that in parabolic space-time, having big pieces of regular parabolic Lipschitz images, having big pieces of regular parabolic bi-Lipschitz images, and being parabolic uniformly rectifiable are equivalent for Ahlfors-David regular sets. The result extends the classical David-Semmes characterization of uniformly rectifiable sets to the parabolic setting.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"P-UR ⇒ BPRPBI hinges on the unproved support/intersection strengthening of the cited corona decomposition in Remark 2.24; without it, Main Lemma 5.31 is not applicable.","rationale":"The reader's CONDITIONAL verdict is appropriate. The most load-bearing unverified point is Remark 2.24, not the omitted coding argument in Section 4.1 (standard and replaceable) and not any internal inconsistency I could locate. I read through the gluing lemmas in Section 3, the smooth rotation in Section 5.1, Main Lemma 5.31, and the induction in Section 5.4; they are internally coherent, and the use of Theorem 2.22 is the only place where the manuscript relies on a strengthening of a cited theorem without proof. If Remark 2.24 is valid, the induction step applies Main Lemma exactly as written; if it is not, direction P-UR ⇒ BPRPBI lacks a proof. Because this is an addressable verification issue rather than a demonstrated false step, the paper should remain CONDITIONAL: I do not recommend REJECT, and I cannot recommend ACCEPT until the support/intersection control is checked. This matches the reader's weakest assumption, so agreement is full and the verdict is unchanged.","tokens_in":40820,"tokens_out":40270,"duration_ms":390427,"concrete_test":"Independently re-derive Remark 2.24 from the construction in [BHH+23a, Theorem 3.2] (and its Theorem 3.1): for every coherent stopping-time regime S and every semi-coherent S' = S ∩ D(Q_0), verify that the graph ψ_{S'} produced there satisfies (2.23) with the same L and admits P_{S'} ∩ B((X_{Q_0},t_{Q_0}),κ diam(Q_0)) ≠ ∅ and supp ψ_{S'} ⊆ P_{S'} ∩ B((X_{Q_0},t_{Q_0}),κ diam(Q_0)), with κ depending only on n, the ADR constant and M_Σ. Concretely, track the Carleson-packing estimate used to select ψ_{S'} and check whether truncating at Q_0 preserves the cutoff functions that keep the graph supported in the ball. If the deduction fails, modify Lemma 5.31 to replace (5.35) by a weaker support estimate and rerun Case 2b; if it succeeds, Theorem 1.1 is not endangered by this gap.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central implication P-UR ⇒ BPRPBI (Theorem 5.1) rests on the induction in Section 5.4, and its Case 2b applies Main Lemma 5.31 to the semi-coherent truncation S'_0 = S_0 ∩ D(Q_0). For this application, the graph θ = ψ_{S'_0} must satisfy (5.34)–(5.35): the plane P_0 meets B((X_{Q_0},t_{Q_0}), κ~ diam(Q_0)) and supp θ is contained in the same ball. These are exactly the extra conclusions asserted in Theorem 2.22(4) and Remark 2.24. The paper concedes in Remark 2.24 that item (4) and the support/intersection statement do not appear in [BHH+23a, Theorem 3.2] and says they “can be easily deduced from the construction,” but no deduction is given. Since every semi-coherent subregime S_0 ∩ D(Q_0) has maximal cube Q_0, the cited construction must produce a graph whose support is controlled relative to Q_0, not relative to the original stopping-time maximal cube; this is precisely the quantitative dependence the induction needs. Without (5.35), Lemma 5.20 cannot produce the glue function g with support in B*, and the construction of F* in Main Lemma 5.31 stops. This is an unverified external input and an omitted proof, not an internal contradiction; if the deduction is valid, the argument is coherent.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves Theorem 1.1: for a parabolic Ahlfors-David regular set Σ⊆R^{n+1}, parabolic uniform rectifiability (P-UR) is equivalent both to having big pieces of regular parabolic Lipschitz images (BPRPLI) and to having big pieces of regular parabolic bi-Lipschitz images (BPRPBI). The direction BPRPLI⇒P-UR is proved in Theorem 4.1 through a quantitative bi-Lipschitz decomposition of regular parabolic maps (Theorem 4.2) and an embedding into a higher-dimensional space-time. The reverse direction P-UR⇒BPRPBI is proved in Theorem 5.1 by an inductive Carleson-packing argument that uses the corona decomposition of Theorem 2.22 and a large gluing/rotation lemma, Lemma 5.31. The final implication BPRPBI⇒BPRPLI is immediate from the definitions.","tokens_in":41117,"tokens_out":14595,"duration_ms":137591,"significance":"If the proof is correct, this is a natural and substantial extension of the David–Semmes characterization of uniformly rectifiable sets to the parabolic setting. The paper introduces a workable definition of regular parabolic (bi-)Lipschitz images, provides explicit constructions preserving the regular Lip(1,1/2) condition, and gives a simplified alternative proof of the Euclidean analogue. The BPRPLI⇒P-UR direction is largely self-contained and convincing. The main caveat is the reliance on a strengthening of the cited corona decomposition that is asserted but not proved in Remark 2.24; this is load-bearing for Theorem 5.1. If that step is supplied, the overall strategy is coherent and the claims are quantitatively precise.","major_comments":[{"comment":"The support and intersection assertions in Theorem 2.22(4) and Remark 2.24—namely that for every semi-coherent sub-regime S′ the associated plane P satisfies P∩B((X_{Q(S′)},t_{Q(S′)}),κ diam Q(S′))≠∅ and supp ψ⊆P∩B((X_{Q(S′)},t_{Q(S′)}),κ diam Q(S′))—are not proved. The remark explicitly states that these do not appear in [BHH+23a, Theorem 3.2] and says they \"can be easily deduced from the construction,\" but no deduction is given. This is load-bearing. In Section 5.4, Case 1, the proposed graph map F_{Q0}(x,t)=(ψ(x,t),x,t) is asserted to satisfy property (iii)(a) of H(a+b), i.e., to be the identity outside a ball; this requires supp ψ to be contained in that ball, which is exactly the unproved support assertion. In Case 2b, the application of Lemma 5.31 requires hypotheses (5.34) and (5.35), which are precisely the extra conclusions of Remark 2.24 for the regime S′_0=S_0∩D(Q_0). Without (5.35), Lemma 5.20 cannot produce the glue function g with support in B*, and the construction of F* breaks. The authors should either prove the strengthening, give a precise lemma number where it appears in the cited paper, or explain in detail why the construction in [BHH+23a] yields the quantitative support bound for every sub-regime.","section":"Section 2, Theorem 2.22(4) and Remark 2.24; Section 5.4, Cases 1 and 2b"}],"minor_comments":[{"comment":"The abstract cites the authors' earlier work as \"[BH12]\", but the bibliography contains [BH17] and no [BH12]; the reference should be corrected consistently.","section":"Abstract and Section 1"},{"comment":"The displayed proof is headed \"Proof of Theorem 4.2\", but it proves Theorem 4.1. In the same proof, the text says \"By Theorem 4.17\", which should be \"By Lemma 4.17\".","section":"Section 4.2, proof of Theorem 4.1"},{"comment":"In the first sentence of part (2), the map is called an \"(L,M)-regular parabolic Lipschitz map\", but the definition requires a bi-Lipschitz map; the equivalent formulation (2′) is correct.","section":"Definition 2.16(2)"},{"comment":"The statement says \"If Σ is parabolic uniformly rectifiable then E has big pieces...\", but the set should be Σ, not E.","section":"Theorem 5.1"},{"comment":"The proof refers to equations (5.21), (5.22), and (5.23), which belong to Section 5.2 and are not the hypotheses of Lemma 3.3. The equations should be renumbered or the relevant inequalities should be restated locally, since this makes verification of the gluing lemma unnecessarily hard.","section":"Section 3, proof of Lemma 3.3"},{"comment":"The sentence \"If ψ(X,t),ψ(Y,s), then\" appears to have a missing inequality and should read \"If ψ(X,t)≠ψ(Y,s), then\".","section":"Lemma 5.2, proof"},{"comment":"Equation (4.10) writes H^{n-1}(f(J_0)×{s}), which is notationally confusing because f(J_0) is already a subset of R^{n+1}; the intended fiber of f(J_0) at time s would be clearer as {X∈R^n : (X,s)∈f(J_0)}.","section":"Lemma 4.5, proof"}],"recommendation":"major_revision","confidential_remarks":"The paper is heavily self-referential, relying on the authors' own earlier theorems, but that is normal for this line of work and the central equivalence is new. The blocking issue is the unproved support/intersection strengthening in Remark 2.24, which is needed in both the base case and the induction step of Theorem 5.1. The internal numbering errors should also be corrected. If the authors supply the missing deduction or a precise citation, the paper is publishable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bortz, Hyde, and Sharp prove a theorem the field has been waiting for: for parabolic Ahlfors-David regular sets, parabolic uniform rectifiability is equivalent to having big pieces of regular parabolic Lipschitz images, and also to having big pieces of regular parabolic bi-Lipschitz images. This is a genuine extension of the Euclidean David-Semmes/Azzam-Schul characterizations to space-time, and Theorem 1.1 is new. The overall architecture is coherent: one direction goes through a Jones-style bi-Lipschitz decomposition and a transference principle; the other uses corona decompositions, a smooth rotation lemma, and careful gluing lemmas. The rotation lemma (5.2) is intricate but seems correct, and the gluing lemmas are plausible, though I have not checked every estimate line-by-line.\n\nThe main soft spot is that the proof leans on an extension of Theorem 2.22(4) that is not actually proved. Remark 2.24 admits that the support and intersection control for graph functions on semi-coherent sub-regimes does not appear in the cited [BHH+23a, Theorem 3.2], saying it 'can be easily deduced from the construction.' No deduction is given. This control is exactly what makes Lemma 5.31 applicable in the induction of Section 5.4, Case 2b: without (5.35), the gluing function g fails and the construction of F* stops. This is not an internal contradiction—the argument is coherent if the missing deduction is valid—but it is an unverified external input and an omitted proof. A referee should ask the authors to prove this strengthening or state it with full details.\n\nOther issues are minor: several cross-reference typos (e.g., 'Theorem 4.2' at the end of Section 4.2, 'Theorem 4.17' for Lemma 4.17), and the standard coding argument in the proof of Theorem 4.2 is omitted. The coding argument is standard from Jones/David-Semmes, so that omission is acceptable, but should be commented on. The citation pattern is heavily self-referential, but that is natural since the authors built the parabolic corona machinery. The real issue is not self-citation; it is citing [BHH+23a] for a theorem that is stronger than what appears there.\n\nFor anyone working on parabolic uniform rectifiability or elliptic/parabolic boundary value problems, this paper matters. I believe the main result is likely correct and worth the referee time. Accept for peer review, and require the missing proof in Remark 2.24 plus correction of the small naming errors.","headline":"A significant parabolic analogue of the David-Semmes/Azzam-Schul characterization, but the proof depends on an unproved strengthening of the corona decomposition that a referee must check.","tokens_in":41642,"tokens_out":3089,"would_cite":true,"duration_ms":29965,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":["28A75"],"pacs":[],"model":"deepseek-v4-flash","headline":"For parabolic Ahlfors–David regular sets, parabolic uniform rectifiability is equivalent to having big pieces of regular parabolic Lipschitz images, and also to having big pieces of regular parabolic bi-Lipschitz images.","keywords":["parabolic uniform rectifiability","big pieces of Lipschitz images","regular parabolic bi-Lipschitz images","Lip(1,1/2) functions","Carleson measures","corona decomposition","space-time geometry"],"falsifier":"Construct a parabolic Ahlfors–David regular set that satisfies the Carleson β² condition and, for its corona decomposition, find one semi-coherent stopping-time regime whose approximating regular Lip(1,1/2) graph cannot be chosen with support inside any ball of radius comparable to the regime's top cube that also meets the graph's reference plane; such a configuration would refute Remark 2.24 and invalidate the P-UR to BPRPBI direction of Theorem 1.1.","tokens_in":40578,"feed_emoji":"📐","tokens_out":11540,"duration_ms":97836,"temperature":0.7,"pith_summary":"This paper aims to characterize parabolic uniformly rectifiable sets—closed subsets of space-time that are quantitatively well approximated by time-independent planes at every location and scale—by a purely structural property. The main theorem states that for any parabolic Ahlfors–David regular set, parabolic uniform rectifiability holds exactly when the set has big pieces of regular parabolic Lipschitz images of $n$-dimensional space-time, and exactly when it has big pieces of regular parabolic bi-Lipschitz images. Regular parabolic maps here fix the time variable up to translation and have spatial components that are Lip(1,1/2) functions with a Carleson measure on their local flatness. If correct, the result gives a concrete geometric description of parabolic uniformly rectifiable sets and completes the parabolic analogue of the classical big-pieces characterization of uniform rectifiability.","feed_headline":"Big regular graph patches characterize parabolic uniform rectifiability","feed_subtitle":"Every uniformly rectifiable space-time set contains large smooth graph patches at all scales, and conversely.","key_machinery":"The central machinery has three pieces. First, the corona decomposition theorem (Theorem 2.22, from the cited work [BHH+23a]) which, for any parabolic uniformly rectifiable set, produces a disjoint decomposition into 'bad' cubes with a Carleson packing bound and coherent stopping-time regimes, each regime carrying an $(L,M)$-regular Lip(1,1/2) graph that approximates the set at every cube in the regime to within a controlled multiple of the cube's size. Second, a gluing lemma (Lemma 3.3) that assembles a regular Lip(1,1/2) function from a well-separated family of cubes, provided each piece is locally regular with controlled support or is a bump-weighted regular function; this is what lets the authors patch local images together without losing the Carleson $\\gamma$-number bound. Third, a smooth parabolic rotation map (Lemma 5.2) built from Givens rotations: $F(X,t)=(g_{\\psi(X,t)}X,t)$ with $\\psi(X,t)=\\theta\\,\\eta(\\rho(X,t)/R)$, which coincides with a fixed rotation on $B(0,R)$, is the identity outside $B(0,2R)$, preserves parabolic balls, and has spatial components that are regular Lip(1,1/2) functions with constants independent of the angle between the two reference planes. These three tools, together with the bi-Lipschitz decomposition of regular parabolic Lipschitz maps (Theorem 4.2), carry the whole argument.","core_discovery":"On the paper's own terms, the central discovery is Theorem 1.1: for a closed, parabolic Ahlfors–David regular set $\\Sigma\\subseteq\\mathbb{R}^{n+1}$, the three properties (1) $\\Sigma$ is parabolic uniformly rectifiable, (2) $\\Sigma$ has big pieces of regular parabolic Lipschitz images of $n$-dimensional space-time, and (3) $\\Sigma$ has big pieces of regular parabolic bi-Lipschitz images of $n$-dimensional space-time, are equivalent. The proof runs through the chain BPRPLI $\\Rightarrow$ P-UR $\\Rightarrow$ BPRPBI $\\Rightarrow$ BPRPLI, where the last implication is trivial. The first implication is proved by a quantitative bi-Lipschitz decomposition of regular parabolic Lipschitz maps, an extension into higher-dimensional space-time, and a transfer argument that reduces the problem to checking that regular parabolic bi-Lipschitz images satisfy the parabolic geometric lemma. The second, harder implication starts from a corona decomposition of the P-UR set into regular Lip(1,1/2) graphs and uses an induction on the Carleson packing constant; at each step it rotates the reference planes of the approximating images into a common plane using a Givens-rotation map that preserves regularity, glues the rotated images by a local graph, and applies a Main Lemma to produce one large regular parabolic bi-Lipschitz image.","pith_inferences":["A testable extension suggested by the proof: the rotation lemma's constants are uniform for angles $0\\le\\theta\\le\\pi/2$; it would be worth checking whether the gluing induction still works for larger angles or whether the constants must degrade, which would show where the parabolic geometry differs from the Euclidean one.","The bi-Lipschitz decomposition theorem for regular parabolic Lipschitz maps (Theorem 4.2) may be strong enough to yield a parabolic traveling-salesman-type criterion in space-time, with the $\\gamma$-numbers playing the role of local flatness measures; the paper does not pursue this.","The induction scheme proves more than the theorem states: it shows that the Carleson packing constant controls the quality (L, M, κ, η) of the big-piece images, so an explicit quantitative dependence of the BPRPBI constants on the P-UR constant could be extracted from Section 5.4.","The proof's direct gluing approach suggests that any ambient space admitting a corona decomposition and a regular slow-rotation map may enjoy the same equivalence; the Heisenberg group is a natural candidate, but this is our inference, not a claim of the paper."],"forward_implications":["Every parabolic uniformly rectifiable set contains, at every point and scale, a large piece (a fixed positive fraction of the parabolic surface measure) of a regular parabolic bi-Lipschitz image of $n$-dimensional space-time, with constants depending only on dimension, the ADR constant, and the P-UR constant.","Conversely, the large-scale presence of regular parabolic Lipschitz images (without any lower Lipschitz bound) is enough to force parabolic uniform rectifiability, so in the parabolic category the lower bound is not a hidden extra assumption.","The equivalence closes the parabolic analogue of the classical big-pieces characterization of uniform rectifiability; the paper notes its proof also yields a new, simpler proof of the corresponding Euclidean fact.","The gluing and rotation lemmas supply a reusable method for combining regular parabolic bi-Lipschitz images whose reference planes are not parallel, which is precisely the difficulty that fails for mere Lipschitz graphs in the known Venetian-blinds obstruction.","Since regular parabolic bi-Lipschitz images are parabolic uniformly rectifiable, any finite or controlled union of such images glued by the Main Lemma remains parabolic uniformly rectifiable, so the class of P-UR sets is closed under the paper's gluing operation."],"supporting_citations":[{"why":"Provides Theorem 2.22, the corona decomposition by regular Lip(1,1/2) graphs that the P-UR to BPRPBI direction starts from and extends in Remark 2.24.","marker":"[BHH+23a]"},{"why":"Supplies the induction-on-packing constant scheme, the big-pieces-squared framework, and the stability of the geometric lemma under big pieces used in the proof of Theorem 4.1.","marker":"[BHH+22]"},{"why":"Provides the earlier approximation and induction ideas that Section 5.3 adapts to the parabolic setting.","marker":"[BH17]"},{"why":"Supplies the method of rotating reference planes and gluing Lipschitz images that the Main Lemma and Lemma 5.7 use.","marker":"[AS12]"},{"why":"The source of the bi-Lipschitz decomposition of Lipschitz maps that Theorem 4.2 adapts to regular parabolic maps.","marker":"[Jon88]"},{"why":"Provides the generalized decomposition and coding argument used to complete Theorem 4.2.","marker":"[DS93b]"},{"why":"Supplies Lemma 5.59, the discrete Carleson region decomposition that drives the inductive step in Section 5.4.","marker":"[HM14]"},{"why":"Establishes the equivalence between the Carleson γ-measure condition and the half-order time derivative in parabolic BMO, justifying the regularity notion for the spatial components.","marker":"[Hof97]"}],"fun_headline_variants":["Parabolic uniform rectifiability = big regular space-time patches","Big pieces of regular parabolic images iff parabolic uniform rectifiability","Big regular patches iff parabolic uniform rectifiability"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the decomposition of a uniformly rectifiable set into approximating graph patches can be refined so that every sub-patch's support sits inside a single bounded ball that intersects the patch's reference plane; the authors state this follows easily from the cited work but do not give the full proof, and the later gluing argument depends on it.","fun_headline_variants_meta":{"raw":{"variants":["Parabolic uniform rectifiability = big regular space-time patches","Big pieces of regular parabolic images iff parabolic uniform rectifiability","Big regular patches iff parabolic uniform rectifiability"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001273,"raw_usage":{"total_tokens":5236,"prompt_tokens":1004,"completion_tokens":4232,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":620,"completion_tokens_details":{"reasoning_tokens":4179}},"tokens_in":620,"tokens_out":4232,"duration_ms":32289,"temperature":1.0,"reasoning_tokens":4179,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-28T00:18:07.705697+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a parabolic Ahlfors–David regular set that satisfies the Carleson β² condition and, for its corona decomposition, find one semi-coherent stopping-time regime whose approximating regular Lip(1,1/2) graph cannot be chosen with support inside any ball of radius comparable to the regime's top cube that also meets the graph's reference plane; such a configuration would refute Remark 2.24 and invalidate the P-UR to BPRPBI direction of Theorem 1.1.","supporting_citations":[],"review_version":1}