{"id":"7a7850ff-4b8e-41b3-ac79-72aa7b9f10f5","arxiv_id":"2506.02423","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Any bounded domain admitting a weak solution to a degenerate overdetermined elliptic problem with constant normal derivative is a ball; for ring-shaped domains, both boundaries must be balls but need not be concentric.","lead":"An overdetermined elliptic problem with a degenerate operator forces its domain to be a ball, even when the boundary is not smooth and the boundary gradient vanishes. The proof uses a symmetrization technique and extends the classical Serrin theorem to ring-shaped regions, where both boundary components must be balls.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proof of Theorem 1.5 rests on truncations that do not vanish on the inner boundary: the difference φ_t = [T^β_γ[u]]^t − T^β_γ[u] is not shown to lie in H^1_0(Ω), so the weak-form test-function step is unjustified and Lemmas 4.1–4.5 are left unproved.","rationale":"The reader’s stated weakest assumption, the unproved Lipschitz property of the zero extension in Section 3, is not the most serious issue. That assertion can be repaired: under (1.7) or (1.8), |∇u| is bounded on Ω by the boundary condition near ∂Ω and by continuity on compact subsets away from ∂Ω; for any x ∈ Ω, the ball B_{dist(x,∂Ω)}(x) is contained in Ω, so the segment from x to a nearest boundary point lies in Ω and the fundamental theorem gives |u(x)| ≤ L dist(x,∂Ω), hence the zero extension is Lipschitz. A similar argument applies to the ring-shaped extension in Section 4. The genuinely load-bearing problem is the inner-boundary trace of the truncations used in Theorem 1.5. The paper explicitly leaves Lemmas 4.1–4.5 unproved and describes the proof as a sketch, and the admissibility of the test function [T^β_γ[u]]^t − T^β_γ[u] in H^1_0(Ω) is not established. Since this test-function step is the mechanism by which the PDE enters the continuous Steiner symmetrization argument, the ring-shaped theorem is not rigorously supported. The Section 3 results appear plausible and the Section 4 gap is likely fixable, so the appropriate verdict remains conditional rather than rejection.","tokens_in":19402,"tokens_out":37794,"duration_ms":381485,"concrete_test":"For the radial annulus Ω = {R₁ < |x| < R₀}, take a radial solution u and choose 0 < γ < β < η − γ so that T^β_γ[u] is constant β in a thin shell near ∂Ω₁. Compute φ_t = [T^β_γ[u]]^t − T^β_γ[u] for small t by applying the one-dimensional CStS to horizontal slices: for a slice at height y, the set where T^β_γ[u] = β is a pair of intervals near x = ±√(R₁² − y²), and applying T_t shifts the endpoints of these intervals. Then evaluate the trace norm ∫_{∂Ω₁} |φ_t|² dH^{N−1} as t → 0. If it does not tend to 0, then φ_t is not in H^1_0(Ω), and the use of φ_t in the weak formulation (1.5) in the proof of Theorem 1.5 is invalid, confirming the need for an additional boundary term or a modified truncation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing gap is in Section 4, not Section 3. For the ring-shaped problem, the truncation T^β_γ[u] = min{β, (u−γ)_+} is used both in the stated Lemmas 4.2–4.5 and in the final test-function step of the proof of Theorem 1.5. On the inner boundary ∂Ω₁, the condition (1.12) gives u → η, so for β < η − γ the function T^β_γ[u] equals the positive constant β in a full neighborhood of ∂Ω₁; after the stated extension it also equals β in Ω₁. Its trace on ∂Ω₁ is therefore β, not 0. The continuous Steiner symmetrization [T^β_γ[u]]^t does not preserve pointwise boundary values: it acts on horizontal cross-sections, and for t > 0 the position of the level sets changes, so in general the trace of [T^β_γ[u]]^t on ∂Ω₁ is not β. Hence φ_t = [T^β_γ[u]]^t − T^β_γ[u] need not have zero trace on ∂Ω₁ and need not belong to H^1_0(Ω). The weak formulation (1.5) is only valid for test functions in H^1_0(Ω); if φ_t has nonzero trace on ∂Ω₁, integration by parts produces an unresolved boundary term involving c₁ on ∂Ω₁, which the sketch never computes. The paper explicitly states that the proofs of Lemmas 4.1–4.5 are omitted and that the argument is only sketched, but this trace issue is precisely the point where a 'straightforward modification' is not automatic. Without a cutoff or a boundary-term argument that makes the test function admissible, Theorem 1.5 is not proved as written.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies overdetermined problems for the quasilinear elliptic equation (1.4), -div(g(|∇u|)∇u/|∇u|)=f(u), with g continuous and strictly increasing, g(0)=0, and f of the form continuous plus bounded variation. The main results are Theorems 1.1 and 1.3: for an arbitrary bounded domain Ω and u∈C^1(Ω) a positive weak solution satisfying the weak boundary conditions (1.6) (with c>0) or (1.8) (with c=0), Ω must be a ball. Theorem 1.5 extends the conclusion to ring-shaped domains Ω=Ω0∖Ω1, asserting that both Ω0 and Ω1 must be balls (not necessarily concentric). The proofs are based on continuous Steiner symmetrization (CStS): after extending u and using truncations, the authors prove that the derivative of the symmetrized Dirichlet-type energy with respect to the CStS parameter is o(t), invoke Brock's local-symmetry criterion (Proposition 2.7), and then deduce that locally symmetric level sets force the domain to be a ball. The paper also provides explicit examples showing that in the p-Laplacian case nonsymmetric solutions can exist in a ball with the required boundary data, so the theorems do not assert radial symmetry of u.","tokens_in":19797,"tokens_out":11163,"duration_ms":116235,"significance":"If the proofs are completed, the results would be a substantial advance: they would unify and extend earlier Serrin-type symmetry results by Brock [10], Fragalà–Gazzola–Kawohl [27], and others to settings with nonconstant f of bounded variation, potentially degenerate ellipticity, and completely nonsmooth boundaries, and they would give the first symmetry result for ring-shaped domains in this degenerate setting. The CStS framework is well suited to the problem, and the monotonicity lemma (Lemma 3.1) is an elegant and potentially reusable device. The paper is clearly organized and includes explicit examples that correctly avoid the false conclusion that the solution itself must be radial. However, the current manuscript contains several load-bearing unproved assertions: the global Lipschitzness of the zero extension, the admissibility of the inner-boundary truncation as a test function, and the omission of all proofs of Lemmas 4.1–4.5 in the proof of Theorem 1.5. The significance is therefore conditional on those gaps being resolved.","major_comments":[{"comment":"The statement 'It follows from (1.7) or (1.8) that u is Lipschitz in R^N' is asserted without proof. A C^1 function on an arbitrary bounded open set need not have bounded gradient, and extending by zero across a non-Lipschitz boundary need not produce a Lipschitz function; uniform convergence to constant boundary values together with |∇u| tending to a constant does not by itself control the oscillation of u in a cusp or give a global Lipschitz constant. This Lipschitz constant L is used repeatedly: in Proposition 2.4(10), in the inclusions (3.6)–(3.7), and in the pointwise bound (3.11) that underlies Lemma 3.5 and the test-function argument. A proof, or a substitute argument that avoids global Lipschitzness, must be supplied.","section":"Section 3, opening paragraph"},{"comment":"Lemmas 4.1–4.5 are stated without proofs, with the text saying only that they follow by 'straightforward modifications' of the previous section. These lemmas are not cosmetic: Lemma 4.2 provides the essential convexity inequality (4.1), Lemma 4.4 controls the h-integral term, and Lemma 4.5 controls the f-term in the weak formulation. Since the proof of Theorem 1.5 depends on all of them, each must be proved in full, especially because the two-sided truncation T^β_γ introduces boundary interactions with both ∂Ω0 and ∂Ω1 that are absent in Section 3.","section":"Section 4, Lemmas 4.1–4.5"},{"comment":"The proof uses φ_t = [T^{β1}_{γ1}[u]]^t − T^{β1}_{γ1}[u] as a test function in (1.5), but it is not shown that φ_t ∈ H^1_0(Ω). Since u→η on ∂Ω1 and u≡η in Ω1 after extension, and β1<η−γ1, the function T^{β1}_{γ1}[u] equals β1 in a full neighborhood of ∂Ω1 inside Ω; its trace on ∂Ω1 is therefore β1, not 0. Continuous Steiner symmetrization does not preserve pointwise boundary values, so the trace of [T^{β1}_{γ1}[u]]^t on ∂Ω1 is in general not β1. Hence φ_t does not have zero trace on the inner boundary and need not lie in H^1_0(Ω). The weak formulation (1.5) is only justified for H^1_0 test functions; using φ_t introduces an uncomputed boundary term on ∂Ω1. Remark 4.6 addresses only the case where Ω1 is already known to be a ball, which is not available at this stage of the proof.","section":"Section 4, proof of Theorem 1.5, test function"},{"comment":"The claim that it suffices to prove that T^{β0}_{γ0}[u] is locally symmetric for every β0>γ0>0 is asserted without justification. Even if each truncation is locally symmetric in the sense of Definition 2.5, the passage to local symmetry of u itself requires an argument: one must show that the symmetries of the family of annular level sets {γ0 is adapted to arbitrary bounded domains without smoothness assumptions, and the c=0 case plus the BV-decomposed nonlinearity are genuinely new. The monotonicity lemma (Lemma 3.1) is a clean idea and does real work. The literature review is careful; the self-citation is peripheral and not part of the proof chain. I see no circular reasoning.\n\nThe soft spots, in proportion:\n- The unproved claim that the zero extension of u is Lipschitz (start of Section 3) looks worse than it is. Under (1.7) or (1.8), |∇u| is bounded near ∂Ω, and a standard closest-boundary-point argument gives |u(x)|≤C dist(x,∂Ω), which yields the Lipschitz extension. The paper should write this out, but it is not a fatal gap.\n- Section 4 is the real problem. Lemmas 4.1–4.5 are stated without proofs, and the proof of Theorem 1.5 is explicitly a sketch. The stress-test worry about the test function is not as devastating as stated: the truncation is constant in a full neighborhood of ∂Ω1, and for small t the CStS moves the relevant level sets only by O(t), so the difference [T]^t−T does vanish near ∂Ω1 if one uses the uniform positive distance from ∂Ω1 to the level set u=γ+β. That argument is not in the paper, but it is plausibly repairable. What remains is that the ring-shaped theorem is currently unproved as written because the lemmas are missing. That is a genuine gap in exposition and verification, not a trivial typo.\n- Minor: Lemma 3.5's assertion that the support of v^t is in Ω for small t should be justified, though it follows from the same positive-distance observation.\n\nWho gets value: anyone working on overdetermined problems, CStS, or degenerate quasilinear equations. The paper should go to a serious referee. I would recommend major revision: write out the Lipschitz argument, prove the Section 4 lemmas, and make the test-function admissibility explicit. The core ideas are sound; the presentation is not yet there.","headline":"A serious Serrin-type paper with a solid Section 3 and a Section 4 that is a promise rather than a proof; referee it, but require the ring-shaped lemmas and test-function admissibility to be written out.","tokens_in":20371,"tokens_out":18118,"would_cite":true,"duration_ms":206183,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35N25","53C24","35J70"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that for a broad class of possibly degenerate elliptic equations, any bounded domain admitting a positive weak solution with $u\\to 0$ and $|\\nabla u|\\to c$ uniformly near the boundary must be a ball; for ring-shaped…","keywords":["overdetermined elliptic problems","degenerate elliptic operators","nonsmooth domains","continuous Steiner symmetrization","local symmetry","ring-shaped domains","p-Laplacian","symmetry rigidity"],"falsifier":"Check the zero-extension Lipschitz claim directly: for a cusped bounded domain satisfying all hypotheses, compute $\\sup_{x,y}|\\bar u(x)-\\bar u(y)|/|x-y|$ for the zero extension $\\bar u$. If this quantity is infinite for some allowed solution, Theorem 1.1 has no proof as written; if one can prove from (1.6) alone that $u(x)\\le C\\,\\operatorname{dist}(x,\\partial\\Omega)$ near the cusp, the gap closes.","tokens_in":19176,"feed_emoji":"🔵","tokens_out":8954,"duration_ms":92253,"temperature":0.7,"pith_summary":"This paper proves a symmetry rigidity theorem for overdetermined elliptic problems: if a positive weak solution of a quasilinear equation with possibly degenerate ellipticity exists on an arbitrary bounded domain, with the value tending to zero and the gradient tending to a constant near the boundary, then the domain must be a ball. No smoothness of the boundary is assumed. The same conclusion holds when the boundary constant is zero, provided the gradient does not vanish near the boundary, and an analogous statement says that in a ring-shaped domain the inner and outer boundaries are both balls, but not necessarily concentric. The result matters because it removes smoothness and non-degeneracy assumptions that earlier proofs needed, and because the conclusion is best possible: explicit non-radial solutions exist in balls, and nonconcentric ring domains occur.","feed_headline":"Overdetermined PDEs force every bounded domain to be a ball","feed_subtitle":"Even with rough boundaries and degenerate operators, constant normal derivative forces roundness.","key_machinery":"The proof is carried by continuous Steiner symmetrization (CStS), a one-parameter flow $T_t(u)$ that rearranges each superlevel set of $u$ into intervals of the same length in a fixed direction, preserving measure and decreasing convex gradient energies. Building on the local-symmetry criterion, if the $G$-energy difference $\\int G(|\\nabla u_t|)-\\int G(|\\nabla u|)$ is $o(t)$ as $t\\to 0$ for strictly convex $G$, then $u$ is locally symmetric in that direction; a locally symmetric function's superlevel sets are countable unions of disjoint balls, and connectedness of $\\Omega$ makes $\\Omega$ a single ball. The approximation part truncates $u$ by levels $(u-\\gamma)_+$ near the boundary, with a monotonicity lemma showing the lowest truncation gives the worst energy drop, plus co-area and bounded-variation estimates controlling the boundary terms; for ring-shaped domains the truncation $T^\\beta_\\gamma[u]$ handles both inner and outer boundaries.","core_discovery":"On the paper's own terms, the central discovery is that this symmetry is a robust geometric constraint: for any bounded domain $\\Omega$ with no regularity assumed on $\\partial\\Omega$, any $f$ admitting a decomposition into a continuous part plus a function of bounded variation, and any strictly increasing $g$ with $g(0)=0$, a positive weak solution of $-\\operatorname{div}(g(|\\nabla u|)\\nabla u/|\\nabla u|)=f(u)$ with $u\\to 0$ and $|\\nabla u|\\to c>0$ uniformly near $\\partial\\Omega$ forces $\\Omega$ to be a ball. If $c=0$, the same conclusion holds under the stronger uniform condition $0<|\\nabla u|<\\varepsilon$ near the boundary. For ring-shaped domains, with the value and normal derivative prescribed on both boundary components and $02$ and the capillary mean-curvature operator fall inside the theorem, and the solution need not itself be radially symmetric inside the ball.","The $c=0$ case is included under the nonvanishing-gradient condition (1.8), so the result does not require the normal derivative constant to be positive.","For ring-shaped domains the conclusion stops short of concentricity: nonconcentric examples exist, so the theorem is optimal.","If one boundary component of a ring domain is already known to be a ball, the normal-derivative data on that component can be dropped, as noted in Remark 4.6."],"supporting_citations":[{"why":"Defines the continuous Steiner symmetrization flow and proves the equimeasurability, monotonicity, semigroup, and homotopy properties used in Section 2.","marker":"[7]"},{"why":"Supplies the decomposition of locally symmetric functions into annuli and the energy-derivative criterion that turns an $o(t)$ energy drop into local symmetry.","marker":"[8]"},{"why":"Gives the earlier symmetrization proof for the same class of equations with $c>0$ and the boundary-truncation approximation scheme that this paper adapts and replaces.","marker":"[10]"},{"why":"Provides the weak formulation of the boundary conditions and the comparability of $u$ with distance to the boundary used to set up the problem on arbitrary domains.","marker":"[59]"},{"why":"Removes growth conditions on $g$ for the geometric proof of symmetry under smoothness assumptions, marking the baseline of degenerate ellipticity this paper targets.","marker":"[27]"},{"why":"Sets up the overdetermined problem on ring-shaped domains with $00$ and the boundary-truncation approximation scheme that this paper adapts and replaces."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the weak formulation of the boundary conditions and the comparability of $u$ with distance to the boundary used to set up the problem on arbitrary domains."},{"cited_title":"Fragal\\`a, F","cited_arxiv_id":null,"evidence_quote":"Removes growth conditions on $g$ for the geometric proof of symmetry under smoothness assumptions, marking the baseline of degenerate ellipticity this paper targets."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Sets up the overdetermined problem on ring-shaped domains with $0