{"id":"ca43d10f-59ff-47a4-9762-6bdbd3bcaf6c","arxiv_id":"2411.18458","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For Schrödinger operators with elliptic symmetric vertically independent coefficients and positive B∞ potentials, Lp regularity and Neumann problems are uniquely solvable for 1<p<2+ε and W^{1,p} estimates hold for 3/2−ε<p<3+ε above Lipschitz graphs.","lead":"This paper proves sharp ranges of p for boundary regularity, Neumann, and W^{1,p} estimates for a family of generalized Schrödinger equations with variable coefficients in regions above Lipschitz graphs. It extends earlier results for the classical Laplacian-Schrödinger operator to a wider operator class while keeping the known sharp exponent ranges.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proof of Theorem 1.3 for 2<p<2+epsilon uses L2 solvability on the truncated domain Omega_R, which is never established; the Rellich estimates of Section 4 are proved only for the full graph domain.","rationale":"Both advertised theorems share the same hinge: Theorem 1.3 for 2<p<2+epsilon is proved exclusively through the Omega_R real-variable argument, and Theorem 7.1 (Neumann) uses the same scheme. The paper's Section 4 Rellich estimates are stated for graph domains with no lateral boundary, so the truncated-domain solvability is not a direct corollary of the displayed estimates. The reader's identified weakest assumptions, x_d-independence and partial Dini continuity, are not the most fragile point: even with those assumptions in force, the truncation step is unjustified as written. I could not locate any passage establishing L2 solvability for Omega_R, and the phrase 'we omit the details' in the p=2 proof covers only the layer-potential invertibility in Omega, not the truncated case. The concern is concrete and testable: either add the missing estimate on Omega_R, or run the same real-variable argument directly on partial Omega. Both fixes are plausible, so the appropriate verdict remains conditional rather than rejection. The paper has substantial independent ingredients, including the Rellich identities, the Fefferman-Phong machinery, and the atomic Hardy-space argument for 1<p<2; the gap is an omitted proof step rather than a counterexample, but it should be explicitly requested in revision.","tokens_in":24081,"tokens_out":17295,"duration_ms":155092,"concrete_test":"Write out the L2 regularity estimate used in Section 5.2 for Omega_R: for v solving Lv+Vv=0 in Omega_R with trace (gR-h)phi on partial Omega cap partial Omega_R and zero data on the remaining boundary, prove that the L2 norm of (nabla v)_* over partial Omega cap partial Omega_R is controlled by the L2 norms of the tangential gradient of (gR-h)phi and of (gR-h)phi times m(x,V). In doing so, explicitly compute the Rellich boundary term on the lateral face {|x'|=R}. If this term is bounded by the right-hand side with a constant independent of R, the gap is closed. If not, replace the Omega_R truncation by a direct application of Theorem 5.3 on partial Omega, using only the global L2 solvability of Theorem 1.3 with p=2 on Omega; the concern is settled if (5.12) then follows without any L2 estimate on Omega_R.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In the proof of Theorem 1.3 for 2<p<2+epsilon, the argument is transferred to the truncated domain Omega_R. Inequality (5.12) is stated on the bottom face partial Omega cap partial Omega_R, and the functions v and w used in the verification of Theorem 5.3 are solutions of boundary value problems in Omega_R. The proof then invokes 'L2 solvability of the regularity problem' twice: once for v, to bound the L2 norm of (nabla v)_* over partial Omega cap partial Omega_R, and once to obtain the first inequality in the chain leading to (5.13). However, the only L2 regularity solvability established in Section 4 is the p=2 case of Theorem 1.3 in the full graph domain Omega above a Lipschitz graph, together with a decay condition at infinity. No L2 result is proved for Omega_R, whose boundary includes lateral faces {|x'|=R} and a top face. Moreover, the Rellich comparison (4.1) is particular to graph domains: on a lateral face n_d=0, so the left side of (4.2) vanishes while the right side contains the integral of partial_d w times partial_nu w over that face, a term not controlled by the tangential and m-weighted boundary quantities in (4.1) and not covered by Lemmas 4.5. Thus the real-variable argument for p>2 rests on an unproved estimate. This is load-bearing because it is exactly the mechanism that extends the range beyond p=2, the advertised sharp range in Theorem 1.3 and Theorem 7.1.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the generalized Schrödinger operator -div(A∇)+V in the region Ω above a Lipschitz graph, under the assumptions that A is elliptic, symmetric, x_d-independent and partially Dini continuous, and that 0<V lies in the class B∞. It claims three main results: unique solvability of the Lp regularity problem for 1<p<2+ε (Theorem 1.3), a W^{1,p} Neumann estimate for 3/2−ε<p<3+ε (Theorem 1.1), and Lp Neumann solvability for 1<p<2+ε (Theorem 7.1). The proofs combine the Fefferman-Phong-Shen maximal function, boundary L∞ and Hölder estimates, Rellich-type identities, layer potentials, and real-variable arguments in the spirit of Shen's work on -Δ+V.","tokens_in":24418,"tokens_out":5697,"duration_ms":49785,"significance":"If the results are correct, they give the first sharp-range Lp boundary estimates for generalized Schrödinger operators with variable coefficients, extending Shen's classical results for -Δ+V in Lipschitz graph domains and complementing the Kenig-Pipher theory for elliptic operators. The structural assumptions (1.2)-(1.4) are natural, and Remark 1.4 correctly identifies the role of x_d-independence. The paper is not circular: it relies on previously published results rather than assuming the target p-ranges. However, the manuscript is not self-contained at several load-bearing points, and the proofs of key truncated-domain estimates are missing.","major_comments":[{"comment":"The proof of Theorem 1.3 for 2<p<2+ε is carried out on the truncated domain Ω_R and invokes 'L2 solvability of the regularity problem' on ∂Ω∩∂Ω_R. However, Section 4 establishes L2 regularity solvability only on the full graph domain Ω (Theorem 4.1 and the layer-potential argument for p=2). No L2 regularity result is proved for Ω_R, whose boundary contains lateral faces {|x'|=R} and a top face. On a lateral face n_d=0, the left side of the Rellich identity (4.2) vanishes, so the comparison (4.1) cannot control the normal-derivative integrals on those faces. This missing estimate is load-bearing because it is exactly the mechanism that extends the range beyond p=2.","section":"§5.2, around (5.12)-(5.13)"},{"comment":"Lemma 4.5 is stated without proof and is described as 'essentially the same' as Lemmas 2.2, 2.6 and 2.7 in [22] with 'slightly modification'. These four inequalities are used repeatedly in the proof of Theorem 4.1, e.g. in (4.11)-(4.15). Since [22] treats -Δ+V while the present operator has variable coefficients A, the modification is not self-evident; a proof or a precise statement of the corresponding results in [22] under assumptions (1.1)-(1.4) is required before Theorem 4.1 can be accepted.","section":"Lemma 4.5, inequalities (4.7)-(4.10)"},{"comment":"The proof of Lemma 6.3 refers to 'Lemma 3.1' and 'Lemma 7.1', neither of which exists in the manuscript: Lemma 3.1 is not stated (Theorem 3.1 is, but is not a Lipsehitz-gradient estimate), and Section 7 (the appendix) contains no lemmas at all. Moreover, the line (6.15) invokes an L2 estimate on ∂Z_{tr} for the cylinder Z_{tr}, which is never proved. Since Lemma 6.3 is the key boundary reverse Hölder inequality used in Theorem 6.1, the W^{1,p} estimate for 2<p<3+ε rests on missing statements.","section":"§6.1, Lemma 6.3"},{"comment":"The proof of Theorem 7.1 asserts without proof both the L2 invertibility of 1/2 I + K_A on L2(∂Ω) and the atomic H1_at Neumann estimate (with only references [5,10,27] for Hardy-space background). The advertised by-product, Lp Neumann solvability for 1<p<2+ε, is therefore not established by the manuscript as written. Please provide the kernel estimates and invertibility argument, and prove the H1_at estimate or give a precise reference that covers exactly this operator and boundary condition.","section":"Appendix, Theorem 7.1"},{"comment":"The statement 'All the ranges of p are sharp' is asserted in the abstract and again in the introduction, but no proof or reference is provided for sharpness under the specific assumptions (1.1)-(1.4) and (1.5). Remark 1.4 concerns failure of global estimates when x_d-independence is dropped, which is not the same as optimality of 2+ε or 3+ε in the admissible class. Either supply a proof or a precise citation, or rephrase the claim.","section":"Abstract and Introduction"}],"minor_comments":[{"comment":"The sentence 'Integrating (7.1) by parts' refers to equation (7.1) in the appendix, but the intended reference is equation (6.5) or (6.9); please correct the cross-reference.","section":"§6.1, proof of Theorem 6.2"},{"comment":"There are several grammatical and typographical issues, e.g. 'referring the reader to for a detailed presentation are list below' and 'We list some properties of A∞ weight, as below'; these should be cleaned up.","section":"§2 and §3"},{"comment":"Theorem 3.1 states assumptions (1.1)-(1.2) and (1.5), but its proof uses Proposition 3.4 and the fundamental solution estimates that depend on the partial Dini condition (1.4); the theorem's hypotheses should include (1.4).","section":"Theorem 3.1"},{"comment":"The normalization of the atom in the definition of H1_{1,at} in §5.1 is stated as ∥∇tan a∥_{L2(∂Ω)} ≤ |B(P,r)∩∂Ω|^{-1/2}; this is standard, but the text should also specify the support condition precisely, as it is used in the proof of Theorem 5.1.","section":"§5.1, definition of H1_{1,at}"}],"recommendation":"major_revision","confidential_remarks":"The central idea is promising and the paper is not circular, but the manuscript is not ready in its current form. The most serious issue is the unproved L2 solvability on truncated domains Ω_R and cylinders Z_{tr}; this is load-bearing for the 2<p<2+ε range. The broken cross-references to Lemma 3.1 and Lemma 7.1 suggest an incomplete revision. I would recommend major revision rather than rejection, since the gap seems addressable by adding the missing estimates or by restructuring the argument to avoid truncated domains."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the paper is a serious extension of Shen's Neumann/regularity theory for Schrödinger operators to variable coefficients, and the main theorems are probably true. But the proof of the p>2 range has a load-bearing hole: the argument moves to truncated domains Ω_R and invokes L2 solvability there, while the only L2 solvability proved in Section 4 is for the full graph domain, via Rellich estimates that are particular to graph boundaries. On the lateral faces of Ω_R the normal has no vertical component; the Rellich identity (4.2) does not control the resulting boundary terms. So (5.12) and the chain leading to (5.13) are not justified as written. The same issue appears in Lemma 6.3, where estimates on ∂Z_{tr} are used. This is not cosmetic: the p>2 extension is the advertised new range. It may be repairable—by proving L2 solvability for Ω_R directly, or by a limiting argument—but it is not in the manuscript.\n\nThe genuinely new content is worth stating. The class of operators is natural: A elliptic, symmetric, x_d-independent, with partial Dini continuity. This is the first paper to extend Shen's 1994 result for −Δ+V to variable coefficients, and the W^{1,p} range 3/2−ε<p<3+ε appears new even for −Δ+V. Section 3's De Giorgi-Nash type boundary estimates and fundamental solution bounds are substantial and seem correctly adapted from Shen's program. There is no circularity, no fitted parameters, and no obvious invented entities.\n\nOther soft spots, in decreasing order: the L2 solvability via layer potential is sketched rather than proved; Lemma 4.5 is quoted as 'essentially the same' as [22] with the variable-coefficient modification left to the reader; the H1_at Neumann step in the Appendix is asserted; sharpness of the ranges is claimed but no argument is given; and cross-references are broken (Lemma 3.1, Lemma 7.1).\n\nThe paper is for specialists in elliptic boundary value problems. I think it deserves a serious referee, but the referee should require a complete proof of the p>2 case. I would not cite the p>2 results as they stand.","headline":"Plausible and interesting extension of Shen's Schrödinger BVP results to variable coefficients, but the p>2 regularity proof has an unproved L2 solvability step on truncated domains.","tokens_in":24974,"tokens_out":5253,"would_cite":false,"duration_ms":46979,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35J10","35J25","35B45"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves sharp $L^p$ regularity, Neumann, and $W^{1,p}$ estimates for generalized Schrödinger operators $\\operatorname{div}(A\\nabla)+V$ in Lipschitz graph domains when $A$ is symmetric, independent of the vertical variable, and…","keywords":["Schrödinger operator","Lipschitz graph","Lp boundary value problem","regularity problem","Neumann problem","W^{1,p} estimate","Fefferman-Phong-Shen maximal function","B∞ potential"],"falsifier":"A concrete check is to solve the regularity problem for $A=I$, $V\\equiv 1$ on a wedge-like Lipschitz graph domain with compactly supported boundary data and compare $\\|(\\nabla u)^*\\|_{L^p(\\partial\\Omega)}$ with $\\|\\nabla_{\\tan}g\\|_{L^p}+\\|g\\|_{L^p}$ for $p$ ranging through $(1,2+\\varepsilon)$; a single failure inside the claimed range refutes Theorem 1.3, while uniform success at the endpoint $p=2+\\varepsilon$ would show the stated range is not sharp.","tokens_in":23871,"feed_emoji":"📐","tokens_out":11951,"duration_ms":93760,"temperature":0.7,"pith_summary":"The paper aims to prove that the generalized Schrödinger operator $-\\operatorname{div}(A\\nabla)+V$ in the region above a Lipschitz graph has the same sharp $L^p$ boundary regularity theory as the classical Laplacian, despite rough coefficients and a large positive potential. Specifically, it claims unique solvability of the $L^p$ regularity problem for $1<p<2+\\varepsilon$, unique solvability of the $L^p$ Neumann problem in the same range, and $W^{1,p}$ estimates for $3/2-\\varepsilon<p<3+\\varepsilon$, with all ranges sharp. If true, these are the first sharp-range estimates for variable-coefficient Schrödinger operators, extending the previously known result for $-\\Delta+V$, which covered only $1<p\\le 2$ for the Neumann problem. The boundary data enters through the tangential gradient together with a potential-weighted term $g\\,m(x,V)$, where $m(x,V)$ is the Fefferman-Phong-Shen maximal function.","feed_headline":"Sharp Lp bounds for Schrödinger equations on Lipschitz graphs","feed_subtitle":"New proof extends the known 1<p≤2 Neumann range to sharp 1<p<2+ε with W^{1,p} estimates too.","key_machinery":"The argument is carried by three interacting pieces. The Fefferman-Phong-Shen maximal function $m(x,V)=\\inf\\{1/r:\\psi(x,r)\\le 1\\}$ with $\\psi(x,r)=r^{2-d}\\int_{B(x,r)}V\\,dy$ encodes the scale at which the potential is critical and appears as a weight in every boundary estimate. The Rellich identity (4.2), valid because $A$ is symmetric and $x_d$-independent, yields the boundary comparability $\\int_{\\partial\\Omega}|\\partial u/\\partial\\nu|^2\\,d\\sigma \\sim \\int_{\\partial\\Omega}(|\\nabla_{\\tan}u|^2+|u|^2m(x,V)^2)\\,d\\sigma$, which is the $L^2$ core of the proof. The De Giorgi-Nash type Hölder estimate (Theorem 3.1), obtained by a perturbation argument using fundamental solution bounds with decay $(1+|x-y|m(x,V))^{-k}$, supplies the pointwise control that feeds the weak reverse Hölder inequality and the passage from $L^2$ to the full $p$ ranges.","core_discovery":"The central claim is that the $L^p$ regularity problem $Lu+Vu=0$ in $\\Omega$, $u=g$ on $\\partial\\Omega$ is uniquely solvable for $1<p<2+\\varepsilon$, with the estimate $\\|(\\nabla u)^*\\|_{L^p(\\partial\\Omega)}\\le C(\\|\\nabla_{\\tan}g\\|_{L^p(\\partial\\Omega)}+\\|g\\,m(x,V)\\|_{L^p(\\partial\\Omega)})$, under the assumptions that $A$ is elliptic, symmetric, $x_d$-independent and partially Dini continuous and that $0<V\\in B_\\infty$. The companion $W^{1,p}$ estimate gives $\\|\\nabla u\\|_{L^p(\\Omega)}+\\|V^{1/2}u\\|_{L^p(\\Omega)}\\le C(\\|f\\|_{L^p(\\Omega)}+\\|G\\|_{B^{-1/p,p}(\\partial\\Omega)})$ for $3/2-\\varepsilon<p<3+\\varepsilon$ for the Neumann problem, and the $L^p$ Neumann problem is solvable for $1<p<2+\\varepsilon$. All these ranges are sharp, and the same method yields the Dirichlet version of the $W^{1,p}$ estimate described in Remark 1.2.","pith_inferences":["An implication left implicit is that the same Rellich comparability should yield $L^p$ solvability of the Dirichlet problem for $Lu+Vu=0$ for $2-\\varepsilon<p<\\infty$ by duality with the regularity problem, mirroring the Laplace theory.","Because the proof uses local boundary estimates together with a real-variable argument, the same sharp ranges should transfer to bounded Lipschitz domains by standard localization, with the potential condition understood locally.","A testable extension is to replace partial Dini continuity of $A$ by a Hölder modulus and quantify the resulting $\\varepsilon$; the scheme suggests $\\varepsilon$ should depend only on the modulus, the ellipticity constant, and the Lipschitz constant.","As $V\\to 0$ the weight term $g\\,m(x,V)$ disappears and the estimates reduce to classical tangential-gradient control, indicating that the data space introduced here is a natural potential-dependent analogue of the usual regularity data."],"forward_implications":["The $L^p$ Neumann problem for $Lu+Vu=0$ is uniquely solvable for $1<p<2+\\varepsilon$, improving the previously known $1<p\\le 2$ range even for the classical operator $-\\Delta+V$.","For the classical equation $-\\Delta u+Vu=0$ on Lipschitz graph domains, the regularity problem is now solvable in the sharp range $1<p<2+\\varepsilon$, not just at $p=2$.","The $W^{1,p}$ estimate controls both $\\|\\nabla u\\|_{L^p(\\Omega)}$ and the potential-weighted term $\\|V^{1/2}u\\|_{L^p(\\Omega)}$ for $3/2-\\varepsilon<p<3+\\varepsilon$, with a Dirichlet analogue stated in Remark 1.2.","The stated ranges $1<p<2+\\varepsilon$ and $3/2-\\varepsilon<p<3+\\varepsilon$ are sharp, so no wider $p$-range can hold under the same structural assumptions.","The weighted boundary condition $\\|g\\,m(x,V)\\|_{L^p}$ shows that large potentials contribute a coercive boundary term, so the admissible data space is naturally tied to the potential through the maximal function."],"supporting_citations":[{"why":"Supplies the previously known $-\\Delta+V$ Neumann result for $1<p\\le 2$ that is extended, and the Fefferman-Phong-Shen maximal function lemmas used throughout.","marker":"[22]"},{"why":"Provides the perturbation scheme and fundamental solution estimates underlying the De Giorgi-Nash type Hölder estimate.","marker":"[24]"},{"why":"Establishes the $L^p$ Neumann and regularity framework for $x_d$-independent coefficients that motivates the sharp range $1<p<2+\\varepsilon$.","marker":"[14]"},{"why":"Supplies the atomic Hardy space argument used for the extension to $1<p<2$.","marker":"[3]"},{"why":"Provides the real-variable theorem used to push the estimates from $L^2$ to $2<p<2+\\varepsilon$.","marker":"[25]"},{"why":"Supplies the reduction of $W^{1,p}$ estimates to weak reverse Hölder inequalities for Neumann solutions.","marker":"[8]"}],"fun_headline_variants":["Lp regularity for Schrödinger: sharp 2+ε bound on Lipschitz graphs","Schrödinger Neumann problem: sharp range extended beyond p=2","New sharp bounds for Schrödinger Lp and W1,p on Lipschitz graphs","Sharp Lp and W1,p ranges for Schrödinger on Lipschitz graphs","Lp Neumann solvability sharpened to 2+ε for Schrödinger"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the coefficient matrix $A$ is symmetric and independent of the vertical variable $x_d$, because the Rellich identity that compares normal and tangential boundary integrals requires the term $\\partial_d A$ to vanish; the paper notes in Remark 1.4 that without this condition the global estimates fail for every $p$.","fun_headline_variants_meta":{"raw":{"variants":["Lp regularity for Schrödinger: sharp 2+ε bound on Lipschitz graphs","Schrödinger Neumann problem: sharp range extended beyond p=2","New sharp bounds for Schrödinger Lp and W1,p on Lipschitz graphs","Sharp Lp and W1,p ranges for Schrödinger on Lipschitz graphs","Lp Neumann solvability sharpened to 2+ε for Schrödinger"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000596,"raw_usage":{"total_tokens":2856,"prompt_tokens":1079,"completion_tokens":1777,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":695,"completion_tokens_details":{"reasoning_tokens":1671}},"tokens_in":695,"tokens_out":1777,"duration_ms":12408,"temperature":1.0,"reasoning_tokens":1671,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T11:10:49.661513+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete check is to solve the regularity problem for $A=I$, $V\\equiv 1$ on a wedge-like Lipschitz graph domain with compactly supported boundary data and compare $\\|(\\nabla u)^*\\|_{L^p(\\partial\\Omega)}$ with $\\|\\nabla_{\\tan}g\\|_{L^p}+\\|g\\|_{L^p}$ for $p$ ranging through $(1,2+\\varepsilon)$; a single failure inside the claimed range refutes Theorem 1.3, while uniform success at the endpoint $p=2+\\varepsilon$ would show the stated range is not sharp.","supporting_citations":[{"cited_title":"Kenig and J","cited_arxiv_id":null,"evidence_quote":"Establishes the $L^p$ Neumann and regularity framework for $x_d$-independent coefficients that motivates the sharp range $1<p<2+\\varepsilon$."},{"cited_title":"Dahlberg and C","cited_arxiv_id":null,"evidence_quote":"Supplies the atomic Hardy space argument used for the extension to $1<p<2$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the real-variable theorem used to push the estimates from $L^2$ to $2<p<2+\\varepsilon$."},{"cited_title":"Geng, W 1,p estimates for elliptic problems with Neumann boundary conditions in Lipschitz domains , Adv","cited_arxiv_id":null,"evidence_quote":"Supplies the reduction of $W^{1,p}$ estimates to weak reverse Hölder inequalities for Neumann solutions."}],"review_version":1}