{"id":"bdf588c3-ee0f-483f-beee-c55a2bcaa4ca","arxiv_id":"2508.09613","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":6.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"Gap-SBM claims optimal L2 and H1 convergence for Dirichlet and Neumann problems by integrating the variational form over the gap between the surrogate and true boundaries, using distance-map geometry.","lead":"This paper introduces Gap-SBM, a reworking of the Shifted Boundary Method that fills the gap between the approximate and true boundaries with extended numerical fields, and claims provably optimal error rates for Dirichlet and Neumann problems. The appeal for generalists: boundary accuracy on meshes that do not fit the geometry, often the limiting factor in moving-boundary simulations, would no longer be the weak link.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified — the supplied full text is unreadable, so no mathematical claim can be checked; the key unverified step is the consistency of gap quadrature and shift operators.","rationale":"The reader's verdict was UNVERDICTED with low confidence, based on the unreadable full text. My stress-test pass reaches the same conclusion. The strongest claim is genuinely strong: optimal L2 and H1 convergence for both Dirichlet and Neumann problems on unfitted meshes. Such a claim depends on a delicate chain of geometric approximation and numerical integration. The abstract itself highlights the two most fragile links: the distance-map construction of the gap and the approximate quadrature formulas plus shift operators. If the distance map only gives a coarse geometric approximation, or the quadrature rules are not sufficiently accurate on the irregular gap elements, the variational consistency would degrade and the optimal rates would fail. However, without access to the proofs, I cannot confirm that this failure occurs. The manuscript, as supplied, has no readable equations, no theorem statements, no tables, and no figures. Therefore the scientifically honest position is not to accept or reject the mathematical claim, but to mark it unverified. I am not raising a new objection beyond the reader's identified weakest assumption; I am agreeing that it is the load-bearing premise and that it cannot be assessed from the available text. The proposed test—recomputing or re-reading the actual consistency lemma once a readable copy is available—would settle whether the concern lands. Until then, the verdict should remain unchanged: UNVERDICTED.","tokens_in":11829,"tokens_out":2632,"duration_ms":31885,"concrete_test":"Obtain the machine-readable LaTeX/PDF of arXiv:2508.09613 and locate the lemma/theorem bounding the consistency error of the gap integrals. Verify that (i) the distance-map approximation of the gap geometry is proved accurate enough to contribute only O(h^{k+1}) to the H1 error and O(h^{k+2}) to the L2 error, where k is the polynomial degree, and (ii) the approximate quadrature over cut gap elements is exact, or its error is explicitly bounded, on the mapped polynomial spaces used in the extension. If either bound is missing, or if it relies on regularity assumptions stronger than those stated in the convergence theorems, then the optimal-rate claim is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that Gap-SBM attains optimal L2 and H1 convergence for Dirichlet and Neumann problems. For that claim to hold, the variational formulation over the surrogate domain plus the approximated gap must be variationally consistent to the correct order: the geometric error from approximating the gap with the distance map, and the quadrature error from integrating over arbitrarily cut gap elements, must not exceed the interpolation error of the finite element space. The abstract names these as the second and third stages, but the available manuscript body consists only of unreadable placeholder glyphs. No theorem statement, proof, or table of results can be inspected. This is not an internal inconsistency I can point to; it is a missing-evidence situation. The only honest assessment is that the optimal-convergence claim is unverified from the material provided. No mathematical objection is being manufactured, but no independent support (machine-checked proof, reproducible code, or readable derivation) is visible either.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes Gap-SBM, a Shifted Boundary Method for unfitted mesh treatment of Dirichlet and Neumann boundary conditions. The abstract describes a three-stage construction: (i) using a distance map between surrogate and true boundaries to approximate the gap geometry, (ii) extending solution/test representations from the surrogate domain into the gap, and (iii) applying approximate quadrature and shift operators to integrate an extended variational formulation. The claimed contributions are provable optimal accuracy in the L2- and H1-norms of the error for both boundary condition types, supported by an extensive set of two-dimensional numerical tests. However, the supplied full text consists entirely of unreadable placeholder glyphs; no equations, theorem statements, proofs, tables, or figures can be inspected. Consequently, the technical content of the paper cannot be evaluated from the material provided.","tokens_in":11997,"tokens_out":2257,"duration_ms":28980,"significance":"If the claimed results hold, this would be a meaningful contribution to the Shifted Boundary Method literature: simultaneous optimal H1 and L2 error estimates for both Dirichlet and Neumann problems on unfitted meshes, with explicit treatment of the geometric and quadrature consistency needed for the gap region, would strengthen the theoretical foundation of SBM. The abstract's three-stage formulation is reasonable in outline, and there is no indication in the abstract of fitted parameters or calibration, which is a point in favor of the work's scientific framing. That said, because the manuscript body is unreadable, the mathematical derivations, regularity assumptions, and numerical evidence cannot be verified. The significance is therefore conditional on content that is currently inaccessible to review.","major_comments":[{"comment":"The full text of the manuscript is unreadable: it consists of placeholder glyphs and repeated nonsensical strings, with no coherent equations, theorem statements, proofs, lemmas, or numeric tables. The central claim of the paper—provable optimal L2 and H1 convergence for Dirichlet and Neumann problems—is precisely a mathematical theorem backed by derivations and experiments, none of which can be inspected. This is a load-bearing presentation failure: without readable technical content, no aspect of the claimed analysis can be confirmed or refuted. The manuscript must be resubmitted in a legible form before substantive review can occur.","section":"Entire manuscript body"},{"comment":"The central technical premise—that approximate quadrature formulas and shift operators over the gap preserve variational consistency to the order required for optimal convergence—is asserted in the abstract but cannot be checked in the supplied text. In particular, the interaction between the geometry approximated by the distance map and the quadrature over arbitrarily cut gap elements is the natural place where the claimed rates could break down. Since the derivation is not visible, this remains a legitimate correctness risk rather than a demonstrated error. The authors should ensure that the full version contains an explicit consistency lemma quantifying the geometric and quadrature errors, and that this lemma is used in the final a priori estimates.","section":"Abstract, Stages 2–3"}],"minor_comments":[{"comment":"The abstract states 'provable optimal accuracy' but does not specify the polynomial degree of the finite element space or the regularity assumptions on the solution. Stating, for example, that the method is 'optimal in the sense that the error decays as h^{k+1} in L2 and h^k in H1 for P_k elements under H^{k+1} regularity' would make the claim more precise and easier to verify from the introduction.","section":"Abstract"}],"recommendation":"major_revision","confidential_remarks":"To the editor: The submitted PDF appears severely corrupted; the body text is unreadable placeholder content. I recommend checking the source file and, if this is an upload error, asking the authors to resubmit a legible manuscript. The abstract alone is not sufficient for any technical assessment. My major comments reflect the absence of readable content rather than a scientific flaw in the method itself."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe abstract promises a genuinely useful result: a shifted boundary method that provably reaches optimal L2 and H1 rates for both Dirichlet and Neumann problems on unfitted meshes. The three-stage construction—distance-map gap geometry, extension of trial and test fields, then approximate quadrature with shift operators—is a sensible way to get there, and it targets a known weak spot of earlier SBM variants. If the analysis holds, this is a real step forward.\n\nWhat I can verify from the abstract: the authors claim provable rates, not just observed behavior, and they describe the 2D tests as demonstrations, not as a fitting exercise. No free parameters or post-hoc tweaking are visible. That's a good sign. The citation of prior SBM work is missing from the abstract, so I can't judge novelty against the full literature, but the construction sounds distinct.\n\nThe soft spot is exactly where you'd expect: the consistency of the approximate quadrature and shift operators over the gap. If that estimate fails to match the interpolation order, the optimal rates fall apart. That is the load-bearing component, and it cannot be checked from the abstract. Our copy of the full text is full of placeholder glyphs—no theorems, no proofs, no tables. That's a pipeline problem, not a paper problem, but it means I only have the abstract to go on. Also note that 'extensive' 2D tests are not a substitute for proof, and the abstract doesn't mention 3D or more complex geometries.\n\nI see no reason to be suspicious of the claims; I also see no evidence to confirm them. The right move is to send this to a referee who works on SBM or unfitted finite elements, with explicit instructions to check the quadrature consistency lemma and the regularity assumptions in the convergence theorems. If those hold, the paper should get serious attention.\n\nRecommendation: accept for peer review, but flag that the provided PDF needs to be readable before any substantive verdict.","headline":"Plausible and potentially important SBM extension for optimal Neumann/Dirichlet accuracy, but the full text is unreadable in our copy—so it's a referee's job, not a desk decision.","tokens_in":12518,"tokens_out":2401,"would_cite":false,"duration_ms":25107,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["65N30","65N15","65N12"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proposes and analyzes Gap-SBM, a shifted boundary method that attains optimal error rates in the $L^2$ and $H^1$ norms for both Dirichlet and Neumann boundary conditions on unfitted meshes, with two-dimensional tests confirming th","keywords":["shifted boundary method","unfitted mesh","Dirichlet boundary condition","Neumann boundary condition","finite element error analysis","optimal convergence","distance map","variational formulation"],"falsifier":"Take a smooth curved domain (e.g., a disk) with a known exact solution and run uniform refinements of an unfitted triangular mesh with first-order elements; plot the $L^2$ and $H^1$ errors against mesh size on a log-log scale. If the slopes are not approximately 2 and 1, respectively, the claimed optimal rates fail. A more targeted probe is to reduce the quadrature order used only inside the gap elements: if the observed convergence rate drops accordingly, the gap quadrature is the limiting ingredient.","tokens_in":11697,"feed_emoji":"📐","tokens_out":6575,"duration_ms":60418,"temperature":0.7,"pith_summary":"This paper introduces and analyzes Gap-SBM, a Shifted Boundary Method for solving partial differential equations on meshes that do not conform to the domain boundary. The central claim is that, for both Dirichlet and Neumann boundary conditions, the method converges at the optimal rates in the $L^2$ and $H^1$ norms, meaning it is as accurate as a standard finite element method on a mesh that fits the boundary. The construction has three stages: approximate the geometry of the gap between the surrogate boundary and the true boundary using a distance map, extend the finite element fields across that gap, and integrate the resulting variational formulation with specially designed quadrature and shift rules. The paper proves a priori error estimates and reports extensive two-dimensional numerical tests that match the predicted rates. If correct, the method offers a practical route to curved-domain accuracy without requiring the mesh to align with the boundary.","feed_headline":"Gap-SBM reaches optimal convergence for Dirichlet and Neumann problems","feed_subtitle":"Filling the gap between surrogate and true boundaries lets nonconforming meshes match fitted-mesh accuracy.","key_machinery":"The key machinery is the three-stage gap construction. First, the distance map between the surrogate boundary (the boundary of the unfitted mesh domain) and the true boundary supplies an approximation of the gap geometry. Second, extension operators carry the finite element solution and test functions from the surrogate domain into that gap. Third, approximate quadrature formulas and specific shift operators are used to integrate the extended variational formulation. The distance map is the geometric engine: it converts boundary misalignment into a volumetric gap whose integration error can be controlled, while the shift operators keep the formulation consistent with the original boundary co","core_discovery":"The paper's central discovery is that the gap between the surrogate boundary and the true boundary can be treated as a geometric object that is integrated, rather than ignored or penalized. Given a distance map between the two boundaries, the method constructs an approximate gap region, extends the numerical solution and test functions from the surrogate domain into this region, and then evaluates the variational form using approximate quadrature formulas and shift operators. The outcome is a provable optimal-order error estimate: for polynomial degree $k$, the $H^1$ error is bounded by $C h^k$ and the $L^2$ error by $C h^{k+1}$, with constants independent of the mesh, under the regularity a","pith_inferences":["A natural next step, not tested in the paper, is extension to three dimensions, where the distance-map and cut-cell quadrature become more intricate but the same three-stage construction remains applicable.","The analysis implies a sharp diagnostic for implementation errors: holding the polynomial degree fixed and reducing only the gap quadrature order should degrade the $L^2$ convergence rate exactly to the order of the quadrature error; this is testable with the authors' own two-dimensional setup.","The same mechanism could be adapted to interface and level-set problems, where the 'true boundary' is an internal interface and the distance map is replaced by a signed distance function; the paper does not make this claim.","Because the construction needs only a distance map between surrogate and true boundaries, it should combine naturally with level-set descriptions of evolving domains, potentially avoiding remeshing between time steps; this remains an unstated consequence."],"forward_implications":["The method attains the same optimal $H^1$ and $L^2$ convergence rates for Dirichlet and Neumann problems on unfitted meshes as standard conforming finite elements.","The distance-map/gap construction replaces the need for a body-fitted mesh: curved boundaries are handled by the gap geometry rather than by curved elements or mesh alignment.","The identical three-stage construction serves both Dirichlet and Neumann boundary conditions, so the method applies uniformly to problems with mixed boundary conditions."],"supporting_citations":[],"fun_headline_variants":["Provably optimal SBM: integrate the boundary gap","Integrating the gap yields optimal SBM convergence","Gap-SBM: treat the gap as geometry for optimal accuracy","Optimal convergence for Dirichlet and Neumann by integrating the gap","Stop penalizing the gap: integrate it with Gap-SBM"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The optimal convergence claim rests on the assumption that the distance-map-based approximation of the gap, together with the chosen quadrature and shift operators, preserves variational consistency to an accuracy no lower than the finite element approximation order.","fun_headline_variants_meta":{"raw":{"variants":["Provably optimal SBM: integrate the boundary gap","Integrating the gap yields optimal SBM convergence","Gap-SBM: treat the gap as geometry for optimal accuracy","Optimal convergence for Dirichlet and Neumann by integrating the gap","Stop penalizing the gap: integrate it with Gap-SBM"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001568,"raw_usage":{"total_tokens":6058,"prompt_tokens":667,"completion_tokens":5391,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":411,"completion_tokens_details":{"reasoning_tokens":5309}},"tokens_in":411,"tokens_out":5391,"duration_ms":35377,"temperature":1.0,"reasoning_tokens":5309,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T20:57:42.378551+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a smooth curved domain (e.g., a disk) with a known exact solution and run uniform refinements of an unfitted triangular mesh with first-order elements; plot the $L^2$ and $H^1$ errors against mesh size on a log-log scale. If the slopes are not approximately 2 and 1, respectively, the claimed optimal rates fail. A more targeted probe is to reduce the quadrature order used only inside the gap elements: if the observed convergence rate drops accordingly, the gap quadrature is the limiting ingredient.","supporting_citations":[],"review_version":1}