{"id":"d4d96fd5-be25-4ff1-bc0d-a4bf652c518e","arxiv_id":"1908.03161","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For uniformly elliptic divergence-form operators with DKP coefficients on uniform Ahlfors regular domains, A∞ absolute continuity of elliptic measure is equivalent to uniform rectifiability of the boundary and to being a chord-arc domain.","lead":"This mathematics paper proves that for a broad class of elliptic equations with rough coefficients, a quantitative notion of boundary smoothness is exactly equivalent to a quantitative form of absolute continuity of the associated boundary measure. It gives the final large constant version of this equivalence using a new extrapolation technique.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 5.1 Base case: the pole XP is asserted to lie in P ⊂ Ω*, but P can include fattened boxes outside the sawtooth, so XP may not be in Ω*.","rationale":"The reader's weakest_assumption correctly identified the A∞ transference to sawtooth domains as the most delicate part. My independent reading confirms that the comparison (5.16)–(5.17) is the hinge, and I found a specific unproven assertion in its proof: the choice of the Green-function pole XP requires XP ∈ P ∩ Ω*, but the construction of P as the union of fattened neighbors of a boundary box I_i includes the adjacent outside Whitney box, whose fattened version is not contained in Ω*. Thus the written proof of Lemma 5.9 has a gap. However, the gap is very likely patchable: the center of I_i is inside Ω*, lies in P, and is at distance comparable to diam(P) from ∂P, so it should serve as a valid corkscrew point for P once the constants are made explicit. Because this fix is straightforward but the text as it stands is unsupported, the appropriate verdict is CONDITIONAL rather than an unqualified ACCEPT. I found no comparable issue in Section 4: the discrete-to-continuous small Carleson step is carefully argued, the constants are tracked, and the cases cover all radii. The extrapolation framework itself is coherent. The only soft spot is the pole selection in Theorem 5.1, which is precisely where the reader predicted failure could occur.","tokens_in":36788,"tokens_out":33186,"duration_ms":305173,"concrete_test":"In Lemma 5.9, verify that the center of the Whitney box I_i satisfies the corkscrew condition for the fundamental chord-arc domain P at scale diam(P), with uniform constants, and that this point lies in Ω*. If it does, the comparison (5.14)-(5.16) can be justified by setting XP to be that center. If no such point exists in Ω*∩P, the A∞ transference to sawtooth domains lacks a valid pole for the Green function G*.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"In the proof of Lemma 5.9 (Base case of Theorem 5.1), Green functions G*(XP,·) and GP(XP,·) are compared. The proof states 'as observed above XP ∈ P ⊂ Ω*'. This observation is not justified. P is defined as int(∪_{i'=1}^{m1} (I_{i'})*) where the union runs over all Whitney boxes I_{i'} meeting I_i. Since I_i is a boundary box of the sawtooth Ω*, there is an adjacent outside box J_i with I_i ∩ J_i ≠ ∅, and J_i^* is included in P but is not contained in Ω*. Hence P ⊄ Ω* in general. A corkscrew point XP for P at scale diam(P) could lie in the portion of P outside Ω*, making G*(XP,·) undefined. The comparison chain (5.14)–(5.16), and hence the local A∞ estimate (5.10), depends on XP belonging to both domains. This is load-bearing because Theorem 5.1 and Corollary 5.3 are the transference step (Step 3) essential for the extrapolation proof of Theorem 1.6. The gap is likely fixable by choosing XP inside I_i (or I_i^*), which lies in Ω* and is still a valid corkscrew for P at scale diam(P) since dist(center(I_i), ∂P) ≈ ℓ(I_i) ≈ diam(P). But as written, the assertion 'P ⊂ Ω*' is false, so the proof is incomplete at a critical point.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves the large-constant case of the equivalence, for uniform domains with Ahlfors regular boundary in R^n, n ≥ 3, between A∞ absolute continuity of elliptic measure for a divergence-form operator with Dahlberg–Kenig–Pipher coefficients and uniform rectifiability of the boundary, equivalently chord-arc regularity of the domain. The proof is by extrapolation of Carleson measures: assuming the small-constant result of the companion paper [HMMTZ], the authors reduce the main theorem to two technical steps, namely that a small discrete Carleson hypothesis implies a continuous small Carleson estimate on sawtooth subdomains, and that the A∞ property of elliptic measure transfers from a domain to its sawtooth subdomains with uniform constants. The paper also contains an optimality discussion and a corollary extending the result to operators satisfying an oscillation-type Carleson condition.","tokens_in":37057,"tokens_out":11350,"duration_ms":120759,"significance":"If accepted, this paper completes a long program and settles in full the free-boundary direction for the Dahlberg–Kenig–Pipher class: A∞ absolute continuity of elliptic measure forces uniform rectifiability, with no smallness assumption on the Carleson norm. The extrapolation mechanism and the transference of A∞ to sawtooth subdomains are substantive new tools that are likely to be useful beyond this specific theorem. The proof is carefully organized and makes explicit which parts are drawn from the companion small-constant paper; the reliance on companion and preprint results is transparent and appropriate for a two-part work. The optimality examples in Section 6 and the oscillation-variant Corollary 6.3 strengthen the paper's contribution.","major_comments":[],"minor_comments":[{"comment":"The notation for fattened boxes is inconsistent: I∗∗ is defined twice, once as (1+2τ)I and once as (1+4τ)I; the second occurrence should be I∗∗∗, matching the later use of I∗∗∗ in the definition of U∗∗_Q.","section":"§2.2, paragraph after (2.20)"},{"comment":"There is a typographical error in the definition of β_Q: 'β Q; =' should read 'β_Q :='.","section":"§3, display (3.12)"},{"comment":"In the display 'dist( J, ∂Ω) = dist(J_i, ~Q_i)' the first argument should be J_i rather than J; the current text is a typographical slip.","section":"§5, display (5.30)"},{"comment":"The construction of P after (5.12) is potentially confusing: the sentence 'let m1 denote the maximal number of Whitney boxes intersecting I_i' could be read as including boxes outside Ω*, but the actual union used for P is the relabeled union from (5.11), whose boxes belong to the index set N* and therefore have interiors contained in Ω*. An explicit sentence to this effect would eliminate the ambiguity and justify the assertion P ⊂ Ω*.","section":"§5, Lemma 5.9"}],"recommendation":"minor_revision","confidential_remarks":"For the editor: I specifically checked the stress-test concern about Lemma 5.9, namely that the pole XP might lie in P but outside Ω* because P could include fattened Whitney boxes adjacent to I_i that are not part of the sawtooth. On careful reading this concern does not land: the union defining P is the relabeled union from (5.11), and the boxes in that relabeling are drawn from N*, all of which satisfy int((I_i)*) ⊂ Ω*. Thus P ⊂ Ω* and XP, chosen as a corkscrew point for P, does lie in Ω*. The proof is sound at that point, though the wording should be sharpened. The main theorem is conditional on the companion small-constant paper [HMMTZ] and on the cited preprint [CHMT]; this is a normal division of labor but the editor may wish to confirm that those works are under review or accepted at a comparable venue."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe paper settles the large constant case of the DKP/free boundary problem: A-infinity elliptic measure implies uniform rectifiability of the boundary, without a smallness assumption on the Carleson norm of grad A. That is the genuinely new content, and it is a major result. The proof strategy is clear: extrapolation from the small constant case, using a discrete Carleson measure formulation, with two key technical steps—small Carleson hypothesis implies a continuous Carleson estimate on sawtooth domains (Section 4), and transference of A-infinity to sawtooth domains (Section 5). The paper is honest about what is known, states the previously established companion results cleanly, and includes an optimality discussion. The math is coherent and the main line holds up; the dependence on companion papers is normal for a second part.\n\nThe soft spot is real. In Lemma 5.9, the proof asserts P subset Omega* and places the pole XP in both domains. That assertion is false as written: P is a fundamental chord-arc subdomain obtained by fattening all Whitney boxes meeting I_i, including boxes on the outside of the sawtooth boundary, so P can protrude outside Omega*. If XP lies in that protrusion, G*(XP, .) is undefined and the comparison at the heart of the base case collapses. The fix is easy—choose XP near the center of I_i, which lies in Omega*, and is still a corkscrew for P at scale diam(P)—but the paper needs that correction. This is not a manufactured objection; the text literally says 'XP in P subset Omega*'.\n\nOne other thing: the proof is long and depends on several results in companion papers, so it is not something I could verify line-by-line. I saw no circularity; the small-constant theorem is a different theorem from the target conclusion. Citation pattern looks appropriate.\n\nFor whom: experts in harmonic analysis, PDEs, and geometric measure theory. It deserves a serious referee. I would not desk reject; I would send it to a strong referee with the specific request to check the XP placement in Lemma 5.9 and the surrounding sawtooth comparison. After that correction, this should be accepted.","headline":"Settles the large-constant DKP/free-boundary equivalence, and the main line holds up; the proof has a fixable gap in Lemma 5.9 where P is claimed to lie inside the sawtooth domain.","tokens_in":37629,"tokens_out":4138,"would_cite":true,"duration_ms":43317,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35J25","42B37","31B35"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that for divergence-form elliptic operators with coefficients satisfying the DKP Carleson condition, quantitative absolute continuity of elliptic measure forces the boundary to be uniformly rectifiable and the domain to…","keywords":["elliptic measure","uniform rectifiability","chord-arc domain","Dahlberg-Kenig-Pipher condition","Carleson measure","A∞ weights","extrapolation","sawtooth domains"],"falsifier":"Search for a uniform domain Ω⊂R^n with Ahlfors regular boundary and a uniformly elliptic matrix A satisfying (H1)–(H2) whose elliptic measure is in A∞(σ) but whose boundary is not uniformly rectifiable, for instance containing a set of positive surface measure with no tangent plane; the theorem asserts no such pair exists, so finding one, or even a single scale where the discrete sawtooth estimate (3.5) fails for a bounded harmonic function, would refute the central claim.","tokens_in":36569,"feed_emoji":"📐","tokens_out":6729,"duration_ms":66909,"temperature":0.7,"pith_summary":"The paper closes a circle of equivalences: in a uniform domain with Ahlfors regular boundary, if a divergence-form elliptic operator with coefficients satisfying the DKP condition has elliptic measure in A∞ with respect to surface measure, then the boundary is uniformly rectifiable and the domain is chord-arc. The reverse implication was already known; the new contribution is the forward direction in the large-constant case, where the Carleson norm of the coefficient gradient is merely finite rather than sufficiently small. A reader should care because this is a quantitative free-boundary statement: analytic regularity of the Dirichlet problem, measured by A∞ weights, is exactly equivalent to concrete geometric regularity of the boundary. The proof works by extrapolating a previously established small-constant theorem and by transferring the A∞ property from the original domain to arbitrary sawtooth subdomains with uniform constants.","feed_headline":"No small constants: elliptic measure regularity forces rectifiability","feed_subtitle":"A∞ elliptic measure, uniform rectifiability, and chord-arc geometry are equivalent for DKP operators.","key_machinery":"Two objects carry the argument. The first is a self-improvement theorem for discrete Carleson measures (Theorem 3.1): if a measure m with small Carleson norm controls a second measure m̃ on every sawtooth region where m is small, then m̃ is globally Carleson. The second is the family of sawtooth subdomains Ω_{F,Q} built from dyadic cubes on the boundary and fattened Whitney boxes; the main technical step, Theorem 5.1, transfers ω_L∈A∞(σ) from Ω to every such sawtooth with constants independent of the stopping family F and cube Q, by comparing Green functions in fundamental chord-arc subdomains and using reverse Hölder estimates for the kernel. The small discrete Carleson hypothesis is converted into a small continuous Carleson bound in the sawtooth domain (Section 4), which allows the small-constant theorem to be applied; monotone convergence then passes from compactly contained sawtooths to general ones.","core_discovery":"The central result (Theorem 1.6) states that for a uniform domain Ω⊂R^n, n≥3, with Ahlfors regular boundary, and a uniformly elliptic matrix A satisfying (H1) and (H2) — Lipschitz coefficients with |∇A|δ(·) bounded and |∇A|²δ(·) a Carleson measure with finite norm — the following are equivalent: elliptic measure ω_L for L=-div(A∇) belongs to A∞(σ), the boundary ∂Ω is uniformly rectifiable, and Ω is a chord-arc domain. The new load-bearing implication is that A∞ forces uniform rectifiability. The proof reduces to symmetric matrices, then uses a discrete Carleson extrapolation theorem with two discrete measures built from |∇A|²δ and from |∇u|²δ for bounded harmonic functions u; under a small discrete Carleson hypothesis on sawtooth subdomains, the previous small-constant theorem makes those sawtooths chord-arc, and an extrapolation step upgrades the resulting estimates to the whole domain.","pith_inferences":["The extrapolation theorem is plausibly a general bootstrapping device: any scale-invariant small-constant geometric conclusion that is stable under sawtooth restriction could be promoted to a large-constant statement, so similar results may hold for other elliptic or p-harmonic measure settings where an analogous small-constant theorem exists.","The transference mechanism suggests a testable refinement: the A∞ constants on sawtooth subdomains should control, quantitatively, the 'big pieces of Lipschitz graphs' constants of the boundary, so one could search for an explicit bound on those geometric constants in terms of C0 and θ.","For DKP operators whose elliptic measure fails A∞, the theorem predicts that the boundary cannot be uniformly rectifiable; the paper's optimality examples indicate that such failures can arise as limits of operators with finite but growing Carleson norms, which may serve as a template for probing the sharpness of the quantitative constants.","The oscillation-based Corollary 6.3 suggests the natural endpoint of the theory is a condition phrased entirely in terms of local oscillation of A, and one could test whether the equivalence persists under a weak-L∞ or BMO-type Carleson version of that condition."],"forward_implications":["For DKP operators, the three notions — A∞ elliptic measure, uniform rectifiability of the boundary, and chord-arc geometry — stand or fall together in uniform domains with Ahlfors regular boundary.","Since ω_L∈A∞ is equivalent to L^p solvability of the Dirichlet problem, the result gives a geometric characterization of L^p solvability, a quantitative analogue of the Wiener criterion adapted to singular L^p data.","Corollary 6.3 replaces the pointwise gradient hypotheses (H1)–(H2) by a Carleson condition on the local oscillation of A, so the equivalence survives under a weaker, more natural condition on the coefficients.","The optimality examples show the Carleson condition cannot simply be dropped: without it, elliptic measure can be singular with respect to surface measure, so the A∞ hypothesis genuinely carries the geometric conclusion.","All constants in the equivalence depend only on dimension, ellipticity, uniformity, Ahlfors regularity, and the A∞ constants, so the result is quantitatively stable under perturbation of the background parameters."],"supporting_citations":[{"why":"Proves the small-constant version (Theorem 3.10) that the extrapolation argument bootstraps.","marker":"[HMMTZ]"},{"why":"Establishes that DKP coefficients give A∞ elliptic measure in Lipschitz and chord-arc domains, providing the known (3)⇒(1) direction.","marker":"[KP]"},{"why":"Supplies the extrapolation theorem for discrete Carleson measures and the Carleson measure estimate for harmonic functions.","marker":"[HMM1]"},{"why":"Proves the converse direction of Theorem 3.7, that Carleson measure estimates for harmonic functions imply uniform rectifiability.","marker":"[GMT]"},{"why":"Provides the reduction to symmetric matrices and the A∞-iff-Carleson perturbation theorem used to transfer and extend the results.","marker":"[CHMT]"},{"why":"Contains the prior characterization of chord-arc domains used for the (2)⇒(3) direction.","marker":"[AHMNT]"},{"why":"Gives big-piece Lipschitz approximation of chord-arc domains, used both for (3)⇒(2) and for passing A∞ from Lipschitz subdomains to chord-arc domains.","marker":"[DJ]"},{"why":"Introduces the extrapolation method for Carleson measures on which the proof's bootstrap is based.","marker":"[LM]"}],"fun_headline_variants":["A∞ elliptic measure forces uniform rectifiability","Large constant case solved: A∞ implies rectifiability","Elliptic measure A∞ and rectifiability: now equivalent","No small constants: A∞ elliptic measure yields rectifiability"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the A∞ property of elliptic measure transfers from Ω to every sawtooth subdomain Ω*_{F,Q} with uniform constants; if the Green-function comparison behind that transfer fails below scale τℓ(I_i)/8, the small-constant hypothesis cannot be applied on the sawtooth and the extrapolation collapses.","fun_headline_variants_meta":{"raw":{"variants":["A∞ elliptic measure forces uniform rectifiability","Large constant case solved: A∞ implies rectifiability","Elliptic measure A∞ and rectifiability: now equivalent","No small constants: A∞ elliptic measure yields rectifiability"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000732,"raw_usage":{"total_tokens":3307,"prompt_tokens":1012,"completion_tokens":2295,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":628,"completion_tokens_details":{"reasoning_tokens":2228}},"tokens_in":628,"tokens_out":2295,"duration_ms":17712,"temperature":1.0,"reasoning_tokens":2228,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:21:31.019459+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Search for a uniform domain Ω⊂R^n with Ahlfors regular boundary and a uniformly elliptic matrix A satisfying (H1)–(H2) whose elliptic measure is in A∞(σ) but whose boundary is not uniformly rectifiable, for instance containing a set of positive surface measure with no tangent plane; the theorem asserts no such pair exists, so finding one, or even a single scale where the discrete sawtooth estimate (3.5) fails for a bounded harmonic function, would refute the central claim.","supporting_citations":[],"review_version":1}