{"id":"64a500b1-9c4d-4487-9067-28a29a44bd27","arxiv_id":"1908.09795","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Finite perimeter sets with reduced boundary almost equal to their closure that are volume-constrained critical points of a C^3 elliptic anisotropic perimeter must be finite unions of disjoint equal-radius Wulff shapes.","lead":"The paper proves that, for a large class of crystalline surface-energy functionals, the only volume-preserving equilibrium shapes with sufficiently regular boundaries are finite unions of equal-sized Wulff crystals. This answers a substantial part of Maggi's conjecture for nonsmooth anisotropic perimeter problems.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Corollary 6.8 proof applies (27) to a rescaled variation f_t that is not locally supported; as written, volume-constrained stationarity does not imply the isoperimetric stationarity used to invoke Theorem 6.4.","rationale":"The reader's weakest assumption, the boundary-closure hypothesis H^n(Clos(∂*E)\\∂*E)=0, is a genuine limitation but it is explicit in the theorem and corollaries; it does not undermine the stated claim. The more load-bearing issue I find is in the proof of Corollary 6.8: the reduction from volume-constrained stationarity to isoperimetric stationarity uses a rescaled family f_t that is not a local variation because the rescaling factor is not identically 1 outside the compact support of h_t. Hence condition (27) cannot be applied to f_t. The intended conclusion is very likely correct and can be recovered by the standard Lagrange multiplier plus homothety argument, so I recommend CONDITIONAL rather than REJECT: Theorem 6.4 and Corollary 6.7 appear sound, but Corollary 6.8 needs a corrected proof. I also note a likely typographical issue in Claim 1, where '1/(nH(x))' should probably read 'H(x)/n' for the displayed inequality to imply the bound used in (20); if the published text is literal, the proof of Theorem 6.4 would need correction as well.","tokens_in":29297,"tokens_out":39259,"duration_ms":428791,"concrete_test":"Re-derive Corollary 6.8 without using f_t: from (27) and the Lagrange multiplier theorem, show there exists λ with δP_F(g) = λ ∫_{∂*E} g·n(E,·) dH^n for every compactly supported smooth g. Then evaluate both sides on the homothety vector field g(x)=x, either directly or through a careful truncation, and verify λ = n P_F(E)/((n+1)L(E)). If this succeeds, the corollary is correct but the proof must be revised; if the homothety evaluation fails for finite-perimeter sets with noncompact reduced boundary, the corollary needs a different argument.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"In the proof of Corollary 6.8, the authors define f_t(x) = (L(E)/L(h_t(E)))^(1/(n+1)) h_t(x) for an arbitrary local variation h and claim that (27) applies to f_t because f_t preserves volume. This is not justified: if h_t is a local variation in the sense of Definition 6.6, it is the identity outside a compact set K, so for x outside K one has f_t(x)=c_t x with c_t not identically 1 for a general h. The set {x : f_t(x) ≠ x} is unbounded, so f_t is not a local variation and (27) cannot be applied. The corollary is not established by the argument as written. The gap appears repairable via the standard Lagrange multiplier route: (27) gives δP_F = λ δL on compactly supported variations, and differentiating P_F((1+t)E) and L((1+t)E) under the homothety determines λ = n P_F(E)/((n+1)L(E)), which makes equality in Theorem 6.4 available. But the submitted proof needs this replacement.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper establishes an anisotropic Heintze-Karcher inequality for finite-perimeter sets E in R^{n+1} whose anisotropic mean curvature is bounded between 0 and c and whose reduced boundary satisfies H^n(Clos(∂*E)\\∂*E)=0: the inequality asserts L^{n+1}(E) ≤ n/(n+1) ∫_{∂E} 1/|hF| dH^n, with equality precisely when E is, up to a null set, a finite union of disjoint open Wulff shapes of radii at least n/c. The proof develops an anisotropic normal bundle, an anisotropic nearest-point projection, a Lusin (N) property for (n,h)-sets, and an anisotropic Steiner formula, and uses these tools to reduce the problem to the analysis of level sets of the anisotropic distance function. As a corollary, the paper claims that volume-constrained critical points of the anisotropic perimeter among finite-perimeter sets with the same boundary-closure hypothesis are finite unions of equal-radius Wulff shapes.","tokens_in":29536,"tokens_out":15337,"duration_ms":150755,"significance":"If correct, the main theorem substantially extends the isotropic result of Delgadino and Maggi to elliptic integrands of class C^3 and gives a positive answer to Maggi's conjecture in the restricted class of sets satisfying H^n(Clos(∂*E)\\∂*E)=0. The proof is original and carefully structured; it leans explicitly on prior structural results ([33], [34], [31], [2], [11]) and introduces new tools (the anisotropic normal bundle and the anisotropic Steiner formula) that are likely to be of independent use. The equality analysis is detailed, and the paper is commendably explicit about the constants and hypotheses, with no free parameters fitted to the conclusion. The main caveat is that the proof of the central corollary, Corollary 6.8, contains a genuine gap (see major comment); additionally, the full conjecture without the boundary-closure hypothesis remains open, and the paper should state this limitation more prominently than it currently does.","major_comments":[{"comment":"The proof defines f_t(x) = (L(E)/L(h_t(E)))^(1/(n+1)) h_t(x) and claims that (27) applies to f_t because L(f_t(E)) = L(E) for every t. This is not justified: for a local variation h_t in the sense of Definition 6.6, h_t is the identity outside a compact set K, but f_t(x) = c_t x outside K with c_t = (L(E)/L(h_t(E)))^(1/(n+1)) generally different from 1. Hence {x : f_t(x) ≠ x} is unbounded and f_t is not a local variation, so the volume-constrained stationarity condition (27) does not apply. The corollary can be repaired by the standard Lagrange multiplier argument: (27) implies δP_F = λ δL for all compactly supported variations, and testing a sequence of radial cut-offs of the homothety vector field gives λ = n P_F(E)/((n+1)L(E)); then hF(V,x) = -λ n(E,x), so 0 < -hF·n = λ and equality holds in (17), allowing Theorem 6.4 to be invoked. This replacement should be incorporated.","section":"Section 6, proof of Corollary 6.8"}],"minor_comments":[{"comment":"In the proof, 'a′(0)' should be 'p′(0)'.","section":"Proof of Corollary 6.7"},{"comment":"The text 'finite uni on' contains a typo and should read 'finite union'.","section":"Statement of Theorem 6.4"},{"comment":"The author byline contains a garbled string 'S/suppress lawomir Kolasi´ nski'; it should read 'Sławomir Kolasiński'.","section":"Title page"},{"comment":"The expression H^n(Clos(∂*E) ∼ ∂*E) uses '∼' for set difference; the notation should be clarified (e.g., using '\\setminus') to avoid confusion with asymptotic equivalence.","section":"Notation throughout"},{"comment":"The introduction should state explicitly that Maggi's conjecture is proved only under the additional hypothesis H^n(Clos(∂*E)\\∂*E)=0, which is not part of the original conjecture; the abstract does state this, but a remark in the introduction would make the scope clearer.","section":"Introduction"}],"recommendation":"major_revision","confidential_remarks":"The gap in Corollary 6.8 is genuine, but it is local and repairable via a standard Lagrange multiplier argument; the central Theorem 6.4 appears sound. I recommend major revision rather than rejection. The paper is a strong contribution, but the main advertised corollary needs a corrected proof and the scope relative to Maggi's conjecture should be clarified."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: the main theorem (6.4) is a genuine advance and appears solid; the advertised corollary for volume-constrained critical points (6.8) has a gap in the proof as written.\n\nThe paper does something new. Delgadino-Maggi handled the isotropic area functional, and smooth anisotropic hypersurfaces were already known. Extending to finite-perimeter sets with C^3 elliptic integrands requires new machinery, and the authors build it carefully: the anisotropic normal bundle, the Lusin (N) condition for anisotropic (n,h)-sets, the anisotropic Steiner formula, and the level-set argument. The proof of Theorem 6.4 is long but structured, and I did not find any parameter-fitting or circular use of the target result. The self-citations are to prior structural lemmas, which is appropriate.\n\nNow the soft spots. The extra hypothesis H^n(Clos(d*E) \\ d*E) = 0 is load-bearing and is not part of Maggi's conjecture, so the full conjecture remains open. The paper is upfront about this, but readers should not mistake Corollary 6.8 for a full resolution.\n\nThe more immediate issue is in the proof of Corollary 6.8. The authors define f_t(x) = (L(E)/L(h_t(E)))^(1/(n+1)) h_t(x) and claim that because f_t preserves volume, equation (27) applies to it. But if h_t is a local variation, it equals the identity outside some compact set; on that set f_t is a constant dilation, not the identity (unless the volume ratio is exactly 1). So f_t is not a local variation and (27) cannot be invoked. The derivative computation itself is algebraically fine, but the premise is not satisfied. This is a real gap, though it looks repairable: from (27) one should get the Lagrange multiplier relation dP_F = lambda dL on compactly supported variations, then determine lambda by differentiating homotheties. The submitted proof needs that additional step.\n\nSo the main theorem is worth engaging with, and the paper deserves a serious referee. The referee should ask the authors to fix the Corollary 6.8 argument; with that fix, the paper would be in good shape. I would cite Theorem 6.4 even now, and I would take the paper to a reading group for the GMT audience.","headline":"A strong anisotropic Heintze-Karcher theorem, but Corollary 6.8 has a real gap in its proof that looks repairable.","tokens_in":30122,"tokens_out":3073,"would_cite":true,"duration_ms":31709,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["49Q20","49Q10"],"pacs":[],"model":"deepseek-v4-flash","headline":"Under a mild boundary-closure hypothesis, every finite-perimeter set with anisotropic mean curvature in $(0,c]$ satisfies a sharp Heintze-Karcher inequality, and equality forces finite unions of disjoint open Wulff shapes of radius at…","keywords":["anisotropic isoperimetric problem","Wulff shape","finite perimeter sets","anisotropic mean curvature","Heintze-Karcher inequality","varifolds","critical points","normal bundle"],"falsifier":"Search for a compact smooth hypersurface $\\Sigma\\subset\\mathbb{R}^3$ of genus one and a $\\mathscr{C}^3$ elliptic integrand $F$ for which the anisotropic mean curvature of $\\Sigma$ is a positive constant. If one exists, the set enclosed by $\\Sigma$ satisfies the boundary regularity hypotheses but is not a Wulff shape, contradicting the equality case of the theorem.","tokens_in":29064,"feed_emoji":"🧊","tokens_out":11173,"duration_ms":105957,"temperature":0.7,"pith_summary":"The paper attacks a rigidity conjecture for the anisotropic isoperimetric problem at the low regularity of finite-perimeter sets. It shows that, for a $\\mathscr{C}^3$ elliptic integrand $F$, any finite-perimeter set $E$ whose reduced boundary is almost closed and whose anisotropic mean curvature lies between $0$ and $c$ satisfies the anisotropic Heintze-Karcher inequality $L^{n+1}(E) \\le \\frac{n}{n+1}\\int_{\\partial E} \\frac{1}{|h_F(V,x)|}\\,dH^n(x)$. If equality holds, $E$ must coincide, up to a null set, with a finite union of disjoint open Wulff shapes of radii at least $n/c$. Consequently, under the same boundary condition, volume-constrained critical points of the anisotropic perimeter are precisely finite unions of disjoint equal-radius Wulff shapes. The result extends the smooth-boundary rigidity theorem to finite-perimeter sets, up to the stated boundary condition.","feed_headline":"Equal-radius Wulff shapes are the only anisotropic critical points","feed_subtitle":"A sharp Heintze-Karcher inequality forces finite-perimeter critical sets to be Wulff unions.","key_machinery":"The load-bearing object is the generalized anisotropic unit normal bundle $N^F(A)=\\{(a,\\operatorname{grad}F(u)):(a,u)\\in N(A)\\}$ together with the anisotropic nearest-point projection $\\xi_A^F$ and the anisotropic reach $r^F_A$. The proof establishes an $n$-dimensional Lusin (N) property for $N^F$ on anisotropic $(n,h)$-sets, which lets mean-curvature bounds on the boundary be transferred to the $\\mathscr{C}^{1,1}$ level sets of the anisotropic distance function. A coarea and area integration of the Jacobian of the anisotropic normal map yields the volume estimate, and the equality case uses the Steiner-formula theorem to show the anisotropic reach is at least $n/c$, then a totally-umbilical classification: a connected $\\mathscr{C}^{1,1}$ hypersurface all of whose $F$-principal curvatures are equal is the boundary of a Wulff shape.","core_discovery":"The central claim is Theorem 6.4. Let $F$ be an elliptic integrand of class $\\mathscr{C}^3$, let $E\\subseteq\\mathbb{R}^{n+1}$ be a finite-perimeter set with $H^n(\\mathrm{Clos}(\\partial^*E)\\setminus\\partial^*E)=0$, and let $V=v_n(\\partial^*E)$ be its boundary varifold. Suppose $V$ has no singular first variation, its anisotropic mean curvature satisfies $0<-h_F(V,x)\\cdot n(E,x)\\le c$, and $h_F(V,\\cdot)$ is $\\mathscr{C}^{0,\\alpha}$ on compact subsets of the $\\mathscr{C}^{1,\\alpha}$ regular part of $\\operatorname{spt}\\|V\\|$. Then the anisotropic Heintze-Karcher inequality holds. Equality holds exactly when $E$ differs by a set of measure zero from a finite union of disjoint open Wulff shapes of radii not smaller than $n/c$. Corollary 6.8 draws the critical-point conclusion: a finite-volume finite-perimeter set satisfying the same boundary-closure condition that is a volume-constrained critical point of $P_F$ must be a finite union of disjoint equal-radius Wulff shapes.","pith_inferences":["The condition $H^n(\\mathrm{Clos}(\\partial^*E)\\setminus\\partial^*E)=0$ is used to replace $E$ by an open set with the same essential boundary; it is plausible that a finer argument could remove it, which would settle the unrestricted finite-perimeter conjecture.","A quantitative stability statement should follow from the same proof: sets with small Heintze-Karcher deficit should be close in $L^1$ to a finite union of Wulff shapes, with a distance bound controlled by the deficit and the curvature bounds.","The level-set and normal-bundle machinery may extend to inequalities involving prescribed nonconstant anisotropic mean curvature or weighted perimeters, not only the constant-bound case treated here."],"forward_implications":["Volume-constrained critical points of any $\\mathscr{C}^3$ elliptic anisotropic perimeter, among finite-perimeter sets with the boundary-closure condition, are finite unions of disjoint equal-radius Wulff shapes.","The Heintze-Karcher inequality is rigid: its equality cases are exactly Wulff-shape unions with radii at least $n/c$, so any set achieving equality is explicitly classified.","Smooth-boundary rigidity is recovered: a smooth closed hypersurface with constant positive anisotropic mean curvature must be a Wulff sphere.","The Lusin (N) property for anisotropic normal bundles is established for a broad class of sets, providing a tool for further measure-theoretic inequalities.","Any set satisfying the hypotheses whose volume exceeds the Heintze-Karcher bound cannot be a volume-constrained critical point, giving an explicit volume obstruction."],"supporting_citations":[{"why":"Supplies the Heintze-Karcher inequality for finite-perimeter sets and the replacement argument that turns $E$ into an open set with the same essential boundary.","marker":"[33]"},{"why":"Provides the distance-function, nearest-point projection, reach, and normal-bundle formalism for arbitrary closed sets.","marker":"[31]"},{"why":"Gives the isotropic Lusin (N) argument for bounded-mean-curvature sets that Section 4 adapts to the anisotropic normal bundle.","marker":"[34]"},{"why":"Defines anisotropic $(n,h)$-sets and supplies the weak maximum principle that turns a bounded anisotropic mean curvature into an $(n,h)$-set condition.","marker":"[11]"},{"why":"Regularity theorem for varifolds whose first variation with respect to an elliptic integrand is controlled; it yields the $\\mathscr{C}^{1,\\alpha}$ regular part of the boundary.","marker":"[2]"},{"why":"The standard geometric measure theory reference for the area, coarea, and constancy theorems used in the integral estimates.","marker":"[17]"},{"why":"The smooth-boundary rigidity result for constant higher-order anisotropic mean curvatures that the finite-perimeter theorem extends.","marker":"[21]"},{"why":"The isotropic finite-perimeter critical-point result whose integration strategy is reworked here with anisotropic tools.","marker":"[9]"},{"why":"Defines positive reach and curvature measures, used when the Steiner-formula step forces the anisotropic reach to be at least $n/c$.","marker":"[16]"},{"why":"The polynomial parallel-volume theorem used to conclude positive reach from the anisotropic Steiner formula.","marker":"[22]"}],"fun_headline_variants":["Wulff unions: only anisotropic critical points","Equal-radius Wulff shapes forced by sharp inequality","Anisotropic critical sets must be Wulff unions","Finite-perimeter minimizers: Wulff shapes only"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument collapses if a finite-perimeter set has singular boundary points outside its reduced boundary: the assumption $H^n(\\mathrm{Clos}(\\partial^*E)\\setminus\\partial^*E)=0$ is what lets the proof replace $E$ by an open set with the same essential boundary and control the singular part.","fun_headline_variants_meta":{"raw":{"variants":["Wulff unions: only anisotropic critical points","Equal-radius Wulff shapes forced by sharp inequality","Anisotropic critical sets must be Wulff unions","Finite-perimeter minimizers: Wulff shapes only"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000138,"raw_usage":{"total_tokens":1097,"prompt_tokens":831,"completion_tokens":266,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":447,"completion_tokens_details":{"reasoning_tokens":201}},"tokens_in":447,"tokens_out":266,"duration_ms":3327,"temperature":1.0,"reasoning_tokens":201,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:01:32.897434+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Search for a compact smooth hypersurface $\\Sigma\\subset\\mathbb{R}^3$ of genus one and a $\\mathscr{C}^3$ elliptic integrand $F$ for which the anisotropic mean curvature of $\\Sigma$ is a positive constant. If one exists, the set enclosed by $\\Sigma$ satisfies the boundary regularity hypotheses but is not a Wulff shape, contradicting the equality case of the theorem.","supporting_citations":[{"cited_title":"Uniqueness of singular convex hypersurfaces with lower bounded k-th mean curvature","cited_arxiv_id":"1908.05952","evidence_quote":"Supplies the Heintze-Karcher inequality for finite-perimeter sets and the replacement argument that turns $E$ into an open set with the same essential boundary."},{"cited_title":"Fine properties of the curvature of arbitrary closed sets","cited_arxiv_id":"1708.01549","evidence_quote":"Provides the distance-function, nearest-point projection, reach, and normal-bundle formalism for arbitrary closed sets."},{"cited_title":"The spherical image of singular varieties of bounded mean curvature","cited_arxiv_id":"1903.10379","evidence_quote":"Gives the isotropic Lusin (N) argument for bounded-mean-curvature sets that Section 4 adapts to the anisotropic normal bundle."},{"cited_title":"The Area Blow Up set for bounded mean curvature submanifolds with respect to elliptic surface energy functionals","cited_arxiv_id":"1901.03514","evidence_quote":"Defines anisotropic $(n,h)$-sets and supplies the weak maximum principle that turns a bounded anisotropic mean curvature into an $(n,h)$-set condition."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Regularity theorem for varifolds whose first variation with respect to an elliptic integrand is controlled; it yields the $\\mathscr{C}^{1,\\alpha}$ regular part of the boundary."},{"cited_title":"Geometric measure theory","cited_arxiv_id":null,"evidence_quote":"The standard geometric measure theory reference for the area, coarea, and constancy theorems used in the integral estimates."},{"cited_title":"Compact embedd ed hypersurfaces with constant higher order anisotropic mean curvatures","cited_arxiv_id":null,"evidence_quote":"The smooth-boundary rigidity result for constant higher-order anisotropic mean curvatures that the finite-perimeter theorem extends."},{"cited_title":"Alexandrov’s theo rem revisited","cited_arxiv_id":null,"evidence_quote":"The isotropic finite-perimeter critical-point result whose integration strategy is reworked here with anisotropic tools."},{"cited_title":"Curvature measures","cited_arxiv_id":null,"evidence_quote":"Defines positive reach and curvature measures, used when the Steiner-formula step forces the anisotropic reach to be at least $n/c$."},{"cited_title":"Does polynom ial parallel volume imply con- vexity? Math","cited_arxiv_id":null,"evidence_quote":"The polynomial parallel-volume theorem used to conclude positive reach from the anisotropic Steiner formula."}],"review_version":1}