{"id":"3219bf14-1eff-4123-a87c-9beeb53e309b","arxiv_id":"1908.06615","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A minimizer of the obstacle problem under generalized Orlicz growth is shown to be continuous up to the boundary at capacity-fat points, and its gradient is globally higher integrable under a measure density condition.","lead":"This paper proves boundary continuity and higher integrability of the gradient for minimizers of an obstacle problem with generalized Orlicz growth. It extends earlier results from the non-obstacle case and yields new outcomes for Orlicz and double phase energies.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.2's proof applies Sobolev–Poincaré to ∇u under a normalization that only bounds ∇(u−f); display (7.7) is unjustified as written.","rationale":"The reader's verdict already marks the paper CONDITIONAL and lists both the unproved φ^- Poincaré inequality in Lemma 6.4 and the normalization issue in Theorem 1.2. I agree with the Lemma 6.4 concern: the paragraph in Section 6 explicitly asserts a 'necessary modification' of [15, Proposition 6.2.10] without proving it, and if no equivalent convex η exists under (A0),(A1) the measure-density-to-fatness bridge fails. I nevertheless take the Theorem 1.2 normalization issue as the more load-bearing of the two, because it sits inside the proof of the second headline theorem: display (7.7) applies Proposition 3.5 to u under hypotheses on ∇(u−f), and (7.13) splits the same gradient. The paper's own logic would require either extra smallness assumptions on ∇u and ∇f in the covering or a reformulation of (7.7) in terms of v and f. Both repairs seem plausible, so the verdict should remain CONDITIONAL, not REJECT. I do not see an argument that the main theorems are false; the issues are missing justifications in the written proof. The manuscript has independent value in the carefully assembled generalized-Orlicz machinery and the explicit strategy via two Caccioppoli inequalities plus Gehring's lemma.","tokens_in":22741,"tokens_out":26801,"duration_ms":261057,"concrete_test":"Re-derive displays (7.7) and (7.13) from Proposition 3.5, tracking the hypothesis ‖∇w‖_{L^{φ^{1/s}}(B)} ≤ 1 for each function w to which it is applied. Concretely, attempt to replace (7.6) by the three conditions ‖∇u‖_{L^{φ^{1/s}}(3B)} < 1, ‖∇(u−f)‖_{L^{φ^{1/s}}(3B)} < 1 and ‖∇ψ‖_{L^φ(2B)} < 1, and verify that a Vitali cover of Ω by balls satisfying them exists using only ∫_Ω φ(x,|∇u|) dx < ∞ and the assumed L^{1+δ} integrability of φ(x,|∇ψ|) and φ(x,|∇f|). If such a cover cannot be constructed, the proof of Theorem 1.2 has a real gap; if it can, the missing normalization is a cosmetic omission.","verdict_should_be":"UNCHANGED","load_bearing_attack":"At display (7.7), the interior Caccioppoli term is estimated as ⨏_{2B} φ(x, |u−u_{2B}|/diam(2B)) dx ≤ C(⨏_{2B} φ(x, |∇u|)^{1/s} dx)^s + C by quoting Proposition 3.5 for u. That proposition is stated only for functions w with ‖∇w‖_{L^{φ^{1/s}}(B)} ≤ 1. The hypotheses fixed in (7.6), however, are smallness of ‖∇(u−f)‖_{L^φ(3B)}, ‖∇(u−f)‖_{L^{φ^{1/s}}(3B)} and ‖∇ψ‖_{L^φ(2B)}. Since ∇u = ∇(u−f)+∇f, those assumptions give no control of ‖∇u‖_{L^{φ^{1/s}}(2B)}. The later estimate of the boundary term (7.13) also splits ∇(u−f) into ∇u and ∇f, so the missing normalization for ∇u cannot be avoided. Without an additional small-ball condition on ∇u (and on ∇f where its gradient enters the Sobolev–Poincaré step), the reverse Hölder inequality (7.14) and the Gehring conclusion are not justified as written. The gap appears fixable by adding such normalizations to the covering choice, which is why it supports a conditional verdict rather than rejection.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves two global regularity results for minimizers of obstacle problems with generalized Orlicz (Musielak-Orlicz) growth. Theorem 1.1 asserts that, under strict convexity and conditions (A0), (A1), (A1-n), (aInc), (aDec), a continuous minimizer attains the continuous boundary datum f(x0) at every boundary point x0 satisfying the capacity fatness condition (2.8). Theorem 1.2 asserts that, under the measure density condition (2.7) at every boundary point and higher integrability of the obstacle and boundary data, one has φ(x,|∇u|)∈L^{1+ε}(Ω) with the explicit estimate (1.3). The proofs combine Caccioppoli inequalities, a Sobolev-Poincaré inequality, Gehring's lemma, a comparison principle, and a boundary regularity theorem from Harjulehto-Hästö for the non-obstacle case. Lemma 6.4 bridges the measure density condition to capacity fatness when q<n.","tokens_in":23030,"tokens_out":8631,"duration_ms":84430,"significance":"If the identified gaps are repaired, the results are a valuable extension of known boundary continuity and higher integrability statements from the variable exponent and polynomial cases to generalized Orlicz growth, and they are new in the Orlicz and double phase settings. The explicit bound (1.3), with constants depending only on n, φ, and c*, is a concrete and useful feature. The overall strategy is sound and largely follows established techniques, and the manuscript is transparent about which ingredients come from the authors' monograph and companion papers. The central theorems are genuine new combinations rather than reformulations of cited results, and the reliance on earlier work by the same group is not circular in itself. Two load-bearing technical points, however, need to be fixed before the claims are fully justified.","major_comments":[{"comment":"The proof of Theorem 1.2 applies Proposition 3.5 to u in estimating the interior Caccioppoli term, but Proposition 3.5 requires the normalization ‖∇v‖_{L^{φ^{1/s}}(B)} ≤ 1. The normalization (7.6) gives smallness of ∇(u−f) and ∇ψ only, not of ∇u; since ∇u = ∇(u−f)+∇f and no smallness of ∇f is assumed in the relevant norm, the inequality (7.7) is not justified as written. The same missing normalization affects the boundary estimate (7.13), where ∇(u−f) is split into ∇u and ∇f. This is load-bearing for the conclusion (7.14) and hence for Theorem 1.2. The gap appears repairable by adding smallness conditions on ∇u and ∇f to the covering choice or by applying Proposition 3.5 to u−f and estimating the ∇f term separately, but the proof as written does not supply those conditions.","section":"Section 7, display (7.7)"},{"comment":"The proof of Lemma 6.4 uses a Poincaré inequality for the lower-envelope function φ⁻_{2B} under assumptions (A0) and (A1), saying it follows 'in the almost same way' as [15, Proposition 6.2.10] after a modification involving an equivalent convex Φ-function and [15, Lemma 4.3.2]. This inequality is not stated, proved, or precisely located in the cited source. Since (6.5) is the only bridge from the measure density condition (2.7) to the capacity fatness condition (2.8), the advertised application of Theorem 1.1 to measure-density domains depends on this missing argument. Please provide the full proof or an exact statement and reference; the main theorem itself remains valid if capacity fatness is assumed directly.","section":"Section 6, Lemma 6.4"}],"minor_comments":[{"comment":"In (7.6) the constant C appears in the smallness condition before its value is specified; state explicitly that C is the constant from Proposition 3.5.","section":"Section 7, setup of Theorem 1.2"},{"comment":"The sentence 'By monotone convergence [16, Theorem 4.1], u−f−1/j converges to u−f' should be rephrased, since the sequence u−f−1/j is not monotone; presumably the truncated positive parts max{u−f−1/j,0} are meant.","section":"Proof of Theorem 1.1"},{"comment":"Several symbols are corrupted in the rendering (for example 'u /greaterorequalslantψ' in the abstract and the expression for φ⁻_{2B} in Lemma 6.4); these should be fixed in the final typeset version.","section":"Abstract and Lemma 6.4"}],"recommendation":"major_revision","confidential_remarks":"The two major comments are both repairable in revision, but the missing Poincaré argument for Lemma 6.4 should be checked against the cited monograph during revision. The paper's contribution is a genuine new combination of known boundary regularity with obstacle-problem arguments, and the explicit estimate (1.3) is a strength."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read it. The two main theorems are genuinely new: boundary continuity for the obstacle problem under generalized Orlicz growth, and global higher integrability of the gradient, with new cases in Orlicz and double phase settings. The proof structure is transparent, the reduction from obstacle minimizers to superminimizers and the comparison principle are handled cleanly, and the Caccioppoli-plus-Gehring strategy is standard but fitted to the setting with care. The paper deserves a serious referee.\n\nThe soft spots are real but fixable, and the reader's report correctly identifies them.\n\nFirst, Lemma 6.4 uses a Poincaré inequality for the lower envelope function phi^- that is asserted rather than proved; the author says it follows from a modification of [15, Proposition 6.2.10], but the modification is not written out. This matters because the advertised implication from the measure density condition to capacity fatness goes through that lemma. If the modification fails, Theorem 1.1 still holds if capacity fatness is assumed directly, but the bridge to measure-density domains would be lost. The author should either supply the proof or state the result with a precise citation.\n\nSecond, the stress-test note about display (7.7) lands. In the interior case, Sobolev–Poincaré is applied to the function u, but the normalization chosen in (7.6) only controls gradients of u−f and ψ. There is no smallness assumption on ∇u itself. The same gap appears in the boundary estimate at (7.13), where ∇(u−f) is split into ∇u and ∇f. As written, the reverse Hölder inequality is not justified. This is not fatal; it looks like a fix by adding a smallness condition on the L^{phi^{1/s}} norm of ∇u, and on ∇f where the splitting requires it, to the covering choice. But it needs to be stated and proved.\n\nOn the citation pattern: the reliance on Harjulehto–Hästö is natural given the framework, and the central theorems are new combinations rather than restatements. No problems there.\n\nWho gets value from this paper: researchers in elliptic regularity under generalized Orlicz growth, especially people working on obstacle problems and boundary regularity. It is a solid, honest extension rather than a breakthrough. It deserves peer review, and with the two gaps patched it should be publishable. If I were the editor, I would send it out.","headline":"Genuinely new boundary continuity and global higher-integrability results for obstacle problems under generalized Orlicz growth, with two fixable technical gaps that a good referee should catch.","tokens_in":774,"tokens_out":1014,"would_cite":false,"duration_ms":29720,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["49N60","35J60","35B65","46E35"],"pacs":[],"model":"deepseek-v4-flash","headline":"Under generalized Orlicz growth, obstacle minimizers are continuous up to the boundary and their gradient energies self-improve.","keywords":["obstacle problem","generalized Orlicz growth","Musielak–Orlicz spaces","boundary continuity","higher integrability","capacity fatness","measure density condition","double phase growth"],"falsifier":"Attempt to prove the Poincaré inequality for the lower-envelope function under assumptions (A0) and (A1) following the modification described in Lemma 6.4. If the modification introduces an additive term, then the estimate used to deduce capacity fatness from measure density is not justified. Concretely, check whether the inequality holds for a specific double-phase integrand in dimension two where the measure-density condition holds but the asserted Poincaré estimate fails, which would invalidate the bridge.","tokens_in":22505,"feed_emoji":"📐","tokens_out":5361,"duration_ms":45944,"temperature":0.7,"pith_summary":"This paper proves two global regularity results for minimizers of the obstacle problem associated with the Dirichlet energy integral under generalized Orlicz growth. First, if a boundary point satisfies a capacity-fatness condition, the continuous representative of the minimizer attains the boundary datum continuously at that point. Second, if the boundary satisfies a measure-density condition uniformly, the energy density is integrable to a higher power over the whole domain, with a bound in terms of the energies of the minimizer, the obstacle, and the boundary datum. These results recover the known polynomial and variable-exponent cases and are new for Orlicz and double-phase growth.","feed_headline":"Boundary continuity for obstacle minimizers with Orlicz growth","feed_subtitle":"New global regularity results that cover Orlicz and double-phase energy densities.","key_machinery":"The load-bearing machinery is a chain of inequalities built from two Caccioppoli inequalities: one interior estimate involving the obstacle function and one boundary-localized estimate involving the boundary datum. These combine with a Sobolev–Poincaré inequality for generalized Orlicz functions and Gehring's lemma to upgrade the integrability of the energy density. For boundary continuity, the key object is the relative Sobolev capacity, and the capacity-fatness condition (2.8), along with a transferred regularity theorem for boundary points from the unconstrained case. The bridging lemma relies on a Poincaré inequality for the lower-envelope function constructed from the integrand.","core_discovery":"The paper's central claim is Theorem 1.1: for a strictly convex generalized Orlicz integrand satisfying assumptions (A0), (A1), (A1-n), (aInc), and (aDec), the limit of the minimizer at a boundary point equals the boundary datum whenever the complement of the domain is capacity-fat at that point. Theorem 1.2 states that under a uniform measure-density condition and higher integrability of the obstacle and boundary data, the minimizer satisfies an explicit global higher integrability estimate for its energy density. The paper also proves a bridging lemma, Lemma 6.4, showing that the measure-density condition implies capacity fatness when the lower growth exponent q is less than the dimension n.","pith_inferences":["The capacity-fatness condition in Theorem 1.1 is likely not the weakest possible; the argument suggests that any boundary condition implying regularity of boundary points in the unconstrained problem would transfer to the obstacle problem.","Tracking constants in Gehring's lemma could yield an explicit, though not sharp, higher integrability exponent in terms of the growth bounds, which may be of interest for quantitative regularity estimates.","The fragile step is the Poincaré inequality inside Lemma 6.4; if that inequality fails, the advertised application to measure-density domains needs an alternative proof, while Theorem 1.1 itself remains valid whenever capacity fatness is assumed directly.","One could test the boundary-continuity theorem for double-phase energies on non-Lipschitz boundaries to compare the geometric content of capacity fatness with measure density."],"forward_implications":["Boundary continuity holds for any domain whose complement is capacity-fat at the point, which covers Lipschitz domains and many rougher domains without requiring measure density.","Under the measure-density hypothesis, the global higher integrability estimate gives a self-improvement of the energy class of the minimizer, which is a standard input for further Hölder regularity of the gradient.","The results unify previously separate treatments for polynomial growth, variable exponent growth, Orlicz growth, and double-phase growth.","The explicit estimate in Theorem 1.2 shows quantitatively that higher integrability of the obstacle and boundary data transfers to the minimizer, up to an additive constant and a power of the original energy.","The local-superminimizer structure and comparison principle imply that the minimizer is locally Hölder continuous in the set where it stays strictly above the obstacle."],"supporting_citations":[{"why":"Provides the capacity-fatness boundary regularity criterion used as the main tool in Theorem 1.1, along with the capacity comparability estimate (6.3).","marker":"[13]"},{"why":"Supplies the Sobolev–Poincaré inequality, density results, and the Poincaré inequality framework that Lemma 6.4 modifies.","marker":"[15]"},{"why":"Establishes local higher integrability for quasiminimizers, which Theorem 1.2 globalizes to the obstacle problem.","marker":"[17]"},{"why":"Provides the interior continuity strategy and the relation between obstacle minimizers and local superminimizers.","marker":"[3]"},{"why":"Gives the Caccioppoli-iteration and local-minimizer regularity tools used in the proof of Theorem 5.8.","marker":"[18]"},{"why":"Used for modular–norm equivalence, uniqueness of the minimizer, and Sobolev embedding properties of generalized Orlicz spaces.","marker":"[16]"},{"why":"Supplies the Gehring lemma that upgrades integrability in the proof of Theorem 1.2.","marker":"[10]"},{"why":"Provides the outline for the boundary-continuity proof and the notion of regular boundary points.","marker":"[22]"}],"fun_headline_variants":["Global boundary continuity for Orlicz obstacle minimizers","Higher integrability for minimizers under generalized Orlicz growth","Orlicz and double phase: global regularity for obstacle minimizers","Beyond polynomial growth: global regularity for obstacle minimizers"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof that measure density implies capacity fatness (Lemma 6.4) relies on a Poincaré inequality for the lower-envelope function that is asserted without a full derivation; if that inequality is wrong, the advertised boundary-continuity application to measure-density domains collapses, though Theorem 1.1 still holds whenever capacity fatness is assumed directly.","fun_headline_variants_meta":{"raw":{"variants":["Global boundary continuity for Orlicz obstacle minimizers","Higher integrability for minimizers under generalized Orlicz growth","Orlicz and double phase: global regularity for obstacle minimizers","Beyond polynomial growth: global regularity for obstacle minimizers"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001355,"raw_usage":{"total_tokens":5390,"prompt_tokens":722,"completion_tokens":4668,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":338,"completion_tokens_details":{"reasoning_tokens":4600}},"tokens_in":338,"tokens_out":4668,"duration_ms":30729,"temperature":1.0,"reasoning_tokens":4600,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:41:41.083309+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Attempt to prove the Poincaré inequality for the lower-envelope function under assumptions (A0) and (A1) following the modification described in Lemma 6.4. If the modification introduces an additive term, then the estimate used to deduce capacity fatness from measure density is not justified. Concretely, check whether the inequality holds for a specific double-phase integrand in dimension two where the measure-density condition holds but the asserted Poincaré estimate fails, which would invalidate the bridge.","supporting_citations":[{"cited_title":"Harjulehto and P","cited_arxiv_id":null,"evidence_quote":"Provides the capacity-fatness boundary regularity criterion used as the main tool in Theorem 1.1, along with the capacity comparability estimate (6.3)."},{"cited_title":"Harjulehto and P","cited_arxiv_id":null,"evidence_quote":"Supplies the Sobolev–Poincaré inequality, density results, and the Poincaré inequality framework that Lemma 6.4 modifies."},{"cited_title":"Harjulehto, P","cited_arxiv_id":null,"evidence_quote":"Establishes local higher integrability for quasiminimizers, which Theorem 1.2 globalizes to the obstacle problem."},{"cited_title":"Björn and J","cited_arxiv_id":null,"evidence_quote":"Provides the interior continuity strategy and the relation between obstacle minimizers and local superminimizers."},{"cited_title":"Harjulehto, P","cited_arxiv_id":null,"evidence_quote":"Gives the Caccioppoli-iteration and local-minimizer regularity tools used in the proof of Theorem 5.8."},{"cited_title":"Harjulehto, P","cited_arxiv_id":null,"evidence_quote":"Used for modular–norm equivalence, uniqueness of the minimizer, and Sobolev embedding properties of generalized Orlicz spaces."},{"cited_title":"Giusti: Direct Methods in the Calculus of V ariations, World Scientific, Singapore, 2003","cited_arxiv_id":null,"evidence_quote":"Supplies the Gehring lemma that upgrades integrability in the proof of Theorem 1.2."},{"cited_title":"Heinonen, T","cited_arxiv_id":null,"evidence_quote":"Provides the outline for the boundary-continuity proof and the notion of regular boundary points."}],"review_version":1}