{"id":"0a28a079-d4ce-41a3-ad74-08e67c9f4d33","arxiv_id":"2608.00887","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper proves existence, uniqueness, W^{2,2} regularity and uniform space-likeness for the prescribed Lorentzian mean curvature problem with homogeneous capillary boundary condition on convex domains.","lead":"Mathematicians proved a long-sought existence and regularity theorem for a free-boundary version of the so-called maximal surface equation, a relativistic analogue of minimal surfaces. The result guarantees a unique solution with controlled slope on convex domains, and establishes that these solutions never contain 'light segments' where the graph becomes null.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.1's claimed data-only uniform space-likeness bound C=C(m,Ω,Λ) is not established by the proof as written; Step 4 yields θ depending on the solution, and a compactness argument is needed.","rationale":"The reader's weakest_assumption identified the ψ=0 restriction in the Bernstein estimate, but that restriction is intentional and not a flaw for the ψ=0 theorem; it is explicitly acknowledged and does not create a gap in the proof of Theorem 1.1. The more load-bearing concern is in the final step of Theorem 1.1 itself: the proof constructs θ and C from the L^{α+1} mass δ of the limiting tilt v, which depends on the particular solution, so the claimed uniformity in the data is not proven. This directly affects a headline conclusion of the theorem. The gap is repairable by a standard compactness argument, so the paper's core existence and regularity results remain credible; acceptance should be conditional on supplying that argument. A secondary minor issue is that in Theorem 2.11 the set Γ'=Ω∩∂B_r(x) is not compact as defined when ∂B_r(x) meets ∂Ω, but redefining Γ' as the closure repairs this without changing the argument.","tokens_in":29804,"tokens_out":47413,"duration_ms":401173,"concrete_test":"Carry out the compactness argument in detail: assume sup_{|ρ|≤Λ,∫ρ=0} ‖w_{u_ρ}‖_{L∞}=∞; choose ρ_n with essinf v_n→0; extract ρ_n⇀ρ in L∞(Ω) and u_n→u locally uniformly via Lemma 2.6; use Lemma 3.3 on the smooth approximating solutions to show ∫ v_n^{α+1}≤C v_n(o_n)^α→0 for chosen points o_n, force v=0 a.e., and contradict Theorem 1.3. Verify that the constant C in Lemma 3.3 is independent of the approximating sequence and that the passage through the exhaustion preserves the monotonicity formula; if this check fails, weaken the theorem statement to θ depending on the solution.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"In Step 4 of the proof of Theorem 1.1, the uniform space-likeness bound w_u ≤ ϑ^{-1} is obtained with ϑ=(δ/C)^{1/α}, where 2δ=∫_Ω v^{α+1} dx > 0 is the L^{α+1} mass of the tilt density v=(1-|Du|^2)^{1/2} of the limit maximizer. This δ is not shown to be bounded below by any function of m, Ω, and Λ; it depends on the particular ρ and u. Consequently the constant C in the theorem statement, claimed to depend only on (m,Ω,Λ), is not obtained by the written argument. The theorem as stated asserts a uniform estimate over all admissible ρ with |ρ|≤Λ; the proof only gives, for each fixed ρ, a θ>0 that may a priori tend to zero as ρ varies. This is a gap between statement and proof. It is fixable by a standard compactness argument: if no uniform θ existed, a sequence ρ_n with |ρ_n|≤Λ and essinf v_n→0 would, after weak-* extraction, converge to a maximizer with v=0 a.e., contradicting Theorem 1.3. However, this argument is not present in the paper. The gap does not affect the existence, uniqueness, or W^{2,2} regularity conclusions, but it does affect the advertised data-only uniformity of the space-like estimate.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the prescribed Lorentzian mean curvature (Born-Infeld) equation with a homogeneous capillary boundary condition on a bounded convex domain. The main theorem asserts existence of a unique zero-mean weak solution in W^{2,2}(Ω)∩C(Ω), together with a uniform space-likeness bound |Du|≤1-θ with θ depending only on the data. The proof proceeds in three parts: first, a variational analysis of the associated concave functional, including a no-light-segment theorem for the maximizer for arbitrary bounded capillary data; second, a free-boundary version of the Bartnik-Simon monotonicity formula and a Bernstein-type gradient estimate for classical solutions on strictly convex domains; third, an approximation argument by mollified data and smooth convex domains to obtain the weak solution to the original problem. The paper also includes a detailed appendix with the linear oblique-derivative theory used in the existence argument.","tokens_in":30090,"tokens_out":19722,"duration_ms":176187,"significance":"If the result holds, it is a meaningful advance: it appears to be the first existence and W^{2,2} regularity result for a non-Dirichlet boundary value problem for the prescribed Lorentzian mean curvature operator, and it establishes uniform space-likeness from data alone. The no-light-segment theorem for arbitrary bounded capillary data is a useful standalone contribution, and the free-boundary monotonicity formula is a genuine extension of the Bartnik-Simon machinery. The proofs are detailed, the external anchors (Bartnik-Simon, Lieberman) are appropriate, and the paper is transparent about the cases it cannot handle (general ψ, non-convex or merely Lipschitz domains). These strengths make the central claim credible and worth publishing after revision.","major_comments":[{"comment":"The theorem statement and abstract assert a constant C=C(m,Ω,Λ) for the tilt bound w_u≤C, i.e. a uniform space-likeness estimate depending only on the prescribed data. The proof, however, obtains θ=(δ/C)^{1/α}, where 2δ=∫_Ω v^{α+1}dx is the L^{α+1} mass of the tilt function v of the specific limit maximizer u. No lower bound for δ in terms of m, Ω, and Λ is supplied, and δ could a priori tend to 0 along a sequence of admissible data ρ with |ρ|≤Λ. Thus the data-only uniformity asserted in Theorem 1.1 is not established by the written argument. This does not affect the existence, uniqueness, or W^{2,2} regularity parts of the theorem, and the gap is repairable by a standard compactness argument: a sequence of maximizers with essinf v_n→0 would, after weak-* extraction, converge to a maximizer for a weak-* limit datum, and the monotonicity inequality would force that limit to satisfy v=0 a.e. and hence to contain interior light segments, contradicting Theorem 1.3. The authors should add this argument and state the resulting uniform bound explicitly.","section":"§4.2, Step 4 (proof of Theorem 1.1, after Eq. (4.11))"}],"minor_comments":[{"comment":"The displayed condition {x∈Ω: dist(x,∂Ω)>1/j}⊆Ω_j⋐Ω does not by itself imply Ω_j↗Ω; the authors should either state that the strictly convex smooth Ω_j can be chosen increasing, or adjust the wording to avoid claiming monotonicity that is not guaranteed by the given construction.","section":"§4.2, Step 0"},{"comment":"The phrase 'Let Ω⊆R^m be a C^2, convex domain' contains a stray comma; it should read 'a C^2 convex domain'. More substantively, the proof assumes u∈C^3 and says the general case follows by approximation, but the approximation step is only sketched; a brief justification that the boundary condition and convexity are preserved under the approximation would be helpful.","section":"Lemma 3.3, statement"},{"comment":"The notation N′ is used for ∂Ω∩B_r(x) and also as the capillary part of ∂Ω′; although the intended meaning is clear from context, a short parenthetical clarification would reduce the risk of confusion.","section":"Theorem 2.11, Step 2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is well within the scope of math.AP and the central mathematical strategy is sound. The only load-bearing issue I see is the uniformity gap in the space-likeness estimate in the proof of Theorem 1.1; it is local and fixable, so I recommend major revision rather than rejection. I have no concerns about attribution or novelty: the paper clearly builds on Bartnik-Simon and Lieberman and states its limitations honestly. The authors should also make the increasing-exhaustion construction in §4.2 fully explicit."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThis is a genuinely new result and a careful piece of work. The main theorem is the first existence and regularity theorem for a capillary (free boundary) problem for the Born-Infeld / prescribed Lorentzian mean curvature operator on bounded convex domains. The no-light-segment theorem for arbitrary bounded capillary data and the free-boundary monotonicity formula are real contributions, and the variational framework is clean. I checked the main steps—the maximizer lemma, the comparison principle, the CMC comparison in Theorem 2.11, and the Bernstein estimate—and they hold up.\n\nThe biggest soft spot is in the proof of Theorem 1.1, Step 4. The theorem promises C=C(m,Omega,Lambda), i.e. uniform space-likeness over all rho with |rho|<=Lambda. The written proof produces theta depending on delta=(1/2)∫ v^{alpha+1}, where v is the tilt of the limit maximizer for the particular rho. Nothing in the argument shows delta is bounded below by a function of m,Omega,Lambda. So the proof establishes, for each fixed rho, a positive theta, but not the advertised uniformity. This is fixable by a compactness argument—if theta_n ->0 along a sequence rho_n, extract a limit maximizer with v=0 a.e., contradicting Theorem 1.3—but that argument is not in the paper. The gap does not affect existence, uniqueness, or W^{2,2} regularity; it only affects the uniformity claim.\n\nWorth noting: the gradient estimate Theorem 3.4 handles only psi=0 on strictly convex C^2 domains, and the author says so explicitly. That limits the scope but is not a flaw in what is claimed.\n\nWho gets value from this: anyone working on Born-Infeld problems, prescribed Lorentzian mean curvature, or capillary free-boundary problems in geometric analysis. It deserves a serious referee. The core is sound and the main existence result is proven; the uniformity gap should be either closed or the theorem restated with theta depending on rho. I'd send it out.","headline":"A genuinely new capillary/free-boundary result for the Born-Infeld operator, with a real but fixable gap in the advertised data-only space-like bound.","tokens_in":30636,"tokens_out":2963,"would_cite":false,"duration_ms":28295,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35J93","53C50","35J66","35B45","49Q05","53A10","78A25"],"pacs":[],"model":"deepseek-v4-flash","headline":"On a bounded convex domain with zero capillary boundary condition, the prescribed Lorentzian mean curvature problem has a unique $W^{2,2}$ weak solution that is uniformly space-like and is exactly the unique maximizer of the associated…","keywords":["prescribed Lorentzian mean curvature","Born–Infeld","capillary boundary condition","free boundary","space-like graphs","light segments","W^{2,2} regularity","a priori gradient estimates"],"falsifier":"Compute the fixed-point solutions $u_\\varepsilon$ of Section 4.1 for a smooth strictly convex domain and a smooth zero-mean $\\rho$ with $\\|\\rho\\|_{L^\\infty}\\le\\Lambda$; they solve $H_{u_\\varepsilon}=\\varepsilon u_\\varepsilon-\\rho$ with $Du_\\varepsilon\\cdot n=0$. If $\\sup_\\Omega |Du_\\varepsilon|$ tends to 1 as $\\varepsilon\\to0$ for some such $\\rho$, then the uniform bound of Theorem 3.4 fails and the conclusion $|Du|\\le1-\\theta$ of Theorem 1.1 is false. A direct refutation of the no-light-segment claim would be a bounded zero-mean $\\rho$ on a convex $\\Omega$ whose unique maximizer of $I_{\\rho,0}$ has a light segment in the interior of $\\Omega$.","tokens_in":29577,"feed_emoji":"⚡","tokens_out":12584,"duration_ms":98565,"temperature":0.7,"pith_summary":"The paper establishes the first existence and regularity result for the prescribed Lorentzian mean curvature equation with a free boundary (homogeneous capillary) condition. On any bounded convex domain $\\Omega$, a bounded zero-mean datum $\\rho$ admits a unique, up to additive constants, weak solution $u\\in W^{2,2}(\\Omega)\\cap C(\\Omega)$, and the gradient satisfies $|Du|\\le 1-\\theta$ for some $\\theta\\in(0,1)$ depending only on the data. The same function is the unique maximizer of the Lorentzian area-plus-charge functional over 1-Lipschitz zero-mean functions, so the variational solution and the PDE solution coincide. The key step is showing that maximizers cannot contain interior light segments—segments where $|Du|=1$—and that a free-boundary monotonicity formula then upgrades this to uniform space-likeness. If correct, this supplies the capillary, or free-boundary, counterpart to the known Dirichlet theory, with direct relevance to Born–Infeld electrostatics and space-like capillary hypersurfaces in Minkowski spacetime.","feed_headline":"Free-boundary Born–Infeld problem has unique solutions","feed_subtitle":"For a convex vessel, the zero-contact-angle electrostatic potential exists, is unique, and never reaches the light cone.","key_machinery":"The central objects are the tilt function $w_u=(1-|Du|^2)^{-1/2}$, which measures the hyperbolic cosine of the angle between the graph's normal and the vertical direction and blows up as the graph approaches light-like behavior, and its reciprocal $v=w^{-1}=\\sqrt{1-|Du|^2}$, the Lorentzian volume density. Three mechanisms carry the argument. First, the no-light-segment comparison: if a maximizer had an interior segment with $u(y)-u(x)=|y-x|$, comparison with the constant-mean-curvature cones of formula (2.7) produces a contradiction, so bounded forcing and bounded capillary data cannot generate light segments by themselves. Second, the free-boundary monotonicity formula of Lemma 3.3, an improvement of the earlier Dirichlet-side formula, asserts that for free-boundary solutions on convex $C^2$ domains, $C\\exp(4(\\Lambda^2R^2+1))v(o)^\\alpha \\ge R^{-m}\\int_{E_R}v^{\\alpha+1}\\,dx + R^{2-m}\\int_{E_R}|D^2u|^2\\,dx$ with $\\alpha<1/m$; this holds even when the Lorentzian balls touch the boundary. Third, a Bernstein-type boundary gradient estimate (Theorem 3.4) for smooth strictly convex domains with zero capillary datum gives uniform space-likeness of the approximating classical solutions, which are then produced by a fixed-point argument using boundary regularity theory.","core_discovery":"The paper's central claim is Theorem 1.1: let $\\Omega\\subset\\mathbb{R}^m$ be a bounded convex domain, let $\\rho$ be a bounded zero-mean function, and set the capillary boundary datum $\\psi=0$. Then the free-boundary prescribed mean curvature problem, formally $-\\operatorname{div}(Du/\\sqrt{1-|Du|^2})=\\rho$ in $\\Omega$ with $Du\\cdot n/\\sqrt{1-|Du|^2}=0$ on $\\partial\\Omega$, has a weak solution $u\\in W^{2,2}(\\Omega)\\cap C(\\Omega)$; the solution is unique up to additive constants, and there is $\\theta\\in(0,1)$ depending only on $m$, $\\Omega$, and $\\|\\rho\\|_{L^\\infty}$ such that $|Du|\\le 1-\\theta$ almost everywhere. The proof identifies $u$ as the unique maximizer of the functional $I_{\\rho,0}(v)=\\int_\\Omega(\\sqrt{1-|Dv|^2}+\\rho v)$ among 1-Lipschitz zero-mean functions, and then shows this maximizer is regular. The decisive steps are a no-light-segment theorem, valid for arbitrary bounded capillary data, and a monotonicity formula that controls the average of the Lorentzian volume density and of the second derivatives by the value of the density at any point; together they rule out light-like behavior and yield uniform space-likeness.","pith_inferences":["Because the no-light-segment theorem holds for arbitrary bounded capillary data, the missing ingredient for a full existence theorem with nonzero $\\psi$ appears to be only a Bernstein-type boundary gradient estimate of the kind proved for $\\psi=0$; a suitable such estimate would likely carry the same exhaustion-and-compactness route through unchanged.","The free-boundary monotonicity formula probably implies a rigidity statement beyond maximizers: any strictly space-like free-boundary solution on a convex domain with bounded mean curvature that is light-like at one point must be light-like everywhere, so combining Lemma 3.3 with a no-light-segment result could characterize uniform space-likeness for a wider class of solutions.","In Born–Infeld electrostatics, the homogeneous capillary condition is the natural zero-contact-angle or boundary-tangential-field condition; extending the theorem to measure-valued charges would give a variational construction of the potential for point or sheet charges inside a convex cavity."],"forward_implications":["For every bounded zero-mean $\\rho$ on a bounded convex domain, the homogeneous capillary Born–Infeld problem has a zero-mean weak solution with finite tilt, so the free-boundary electrostatic potential exists, is unique, and never reaches the light cone.","The weak solution is exactly the unique maximizer of the Lorentzian area-plus-charge functional over 1-Lipschitz zero-mean functions, so variational maximizers of this functional are automatically $W^{2,2}$ solutions of the Euler–Lagrange equation.","Any capillary maximizer, even with nonzero bounded boundary datum $\\psi$, has no interior light segments; the possible failure of existence for general $\\psi$ is therefore located entirely in boundary gradient control, not in interior light-like singularities.","The quantitative bound $|Du|\\le 1-\\theta$ with $\\theta$ depending only on $m$, $\\Omega$, and $\\|\\rho\\|_{L^\\infty}$ is part of the theorem, so the space-likeness margin is controlled by the data rather than by the approximating sequence."],"supporting_citations":[{"why":"Supplies the Dirichlet-side monotonicity formula that Lemma 3.3 improves, and the anti-peeling theorem (Theorem 2.9) used to show maximal light segments reach the boundary.","marker":"[BS82]"},{"why":"Supplies the Bernstein-type gradient-estimate technique on which Theorem 3.4 is modeled.","marker":"[Bar88]"},{"why":"Documents that light segments can genuinely obstruct weak solutions and provides the Born–Infeld background and Corollary 1.10 cited in the introduction.","marker":"[Bye+24]"},{"why":"Provides the boundary Hölder regularity theorem (Theorem A.4 here) that controls the approximating fixed-point solutions up to $C^{1,\\beta}$.","marker":"[Lie88]"},{"why":"Provides the bootstrap argument used to promote the fixed-point solutions to $C^\\infty$ in the proof of Theorem 4.1.","marker":"[Lie13]"},{"why":"Provides the global Schauder estimates, maximum principle, and boundary point lemma used for the linearized oblique derivative problems in Lemma A.2.","marker":"[GT01]"},{"why":"Provides the degree-theoretic fixed-point criterion (Theorem 4.4.3) used in Lemma 4.3 for the existence of fixed points.","marker":"[Llo78]"},{"why":"Provides the Lipschitz extension theorem used to pass the approximating maximizers to a uniform limit on the full domain $\\Omega$.","marker":"[EG15]"}],"fun_headline_variants":["Unique capillary Born-Infeld solution, no light segments","Capillary problem: unique solution with uniform space-likeness","Free-boundary Born-Infeld: unique weak solution, no light segments","Zero-contact-angle problem: unique regular solution exists","No light segments imply unique capillary solution"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that on smooth strictly convex domains with zero contact-angle boundary condition, every classical solution with prescribed bounded mean curvature is uniformly space-like with a bound controlled by the data; the paper proves this only for the zero capillary datum, and for nonzero capillary data the boundary step of the estimate is stated to break down.","fun_headline_variants_meta":{"raw":{"variants":["Unique capillary Born-Infeld solution, no light segments","Capillary problem: unique solution with uniform space-likeness","Free-boundary Born-Infeld: unique weak solution, no light segments","Zero-contact-angle problem: unique regular solution exists","No light segments imply unique capillary solution"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00054,"raw_usage":{"total_tokens":2607,"prompt_tokens":977,"completion_tokens":1630,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":593,"completion_tokens_details":{"reasoning_tokens":1550}},"tokens_in":593,"tokens_out":1630,"duration_ms":11869,"temperature":1.0,"reasoning_tokens":1550,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T15:17:07.872478+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the fixed-point solutions $u_\\varepsilon$ of Section 4.1 for a smooth strictly convex domain and a smooth zero-mean $\\rho$ with $\\|\\rho\\|_{L^\\infty}\\le\\Lambda$; they solve $H_{u_\\varepsilon}=\\varepsilon u_\\varepsilon-\\rho$ with $Du_\\varepsilon\\cdot n=0$. If $\\sup_\\Omega |Du_\\varepsilon|$ tends to 1 as $\\varepsilon\\to0$ for some such $\\rho$, then the uniform bound of Theorem 3.4 fails and the conclusion $|Du|\\le1-\\theta$ of Theorem 1.1 is false. A direct refutation of the no-light-segment claim would be a bounded zero-mean $\\rho$ on a convex $\\Omega$ whose unique maximizer of $I_{\\rho,0}$ has a light segment in the interior of $\\Omega$.","supporting_citations":[],"review_version":1}