{"id":"3a20d486-1faf-41c2-84fe-e664b846b653","arxiv_id":"2412.11315","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":3.0,"correctness_risk":"high","formal_verification":"none","parameter_count":1,"one_line_summary":"A stabilizer-free weak Galerkin scheme for biharmonic equations on non-convex polytopal meshes is proposed, but the norm-equivalence proof contains a degree mismatch that undermines the stated estimates.","lead":"The paper proposes a stabilizer-free weak Galerkin finite element method for biharmonic equations on non-convex polygon meshes, claiming optimal error estimates. The proof of the key stability estimate uses a test function whose degree exceeds the space the argument allows, so the main convergence claims are not supported as written.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 4.4 applies (2.2) to a bubble test function of degree 2N+k−1 while r=2N+k−2; the norm equivalence (4.9) and the main estimates are unsupported as written, though redefining r=2N+k−1 may repair the proof.","rationale":"Reader's weakest assumption is exactly right. The degree count is unambiguous: r=2N+k−2 cannot contain an element of degree 2N+k−1. Since (2.2) is the bridge between the discrete weak derivative and the boundary jumps, using a test function outside P_r invalidates (4.10), the lower bound of Lemma 4.4, and hence the uniqueness and error theorems. I also checked whether the flaw is fatal or cosmetic. A consistent redefinition r=2N+k−1 appears to close the gap: Lemma 4.1 only needs r≥2N+k−2; Lemma 4.2/4.3 are degree-independent; Q_r projections in §5–§7 remain at least as accurate when r increases. So I would not reject the mathematical idea outright, but a revision must make the parameter choice coherent. Secondary issue: Lemma 4.2 and the extension arguments in Lemma 4.3 are imported from the author's unpublished [23], which claims the same non-convex stabilizer-free setting; this is a missing-support flag that should be resolved by including proofs or citing a published version. I recommend CONDITIONAL rather than REJECT because the identified gap has a concrete, likely one-line repair; if the repair check fails, rejection is warranted.","tokens_in":16180,"tokens_out":15704,"duration_ms":139739,"concrete_test":"Redo Lemma 4.4 with r redefined as 2N+k−1 throughout: (i) verify φ=(vb−v0)l_kφ_{e_k} now lies in P_r; (ii) check Lemma 4.1 still holds with the same r (its test function has degree 2N+k−2≤r); (iii) trace every later use of Q_r in (5.3), (6.1), and §7, confirming no estimate loses h-powers when the projection space is enlarged. If all steps close, the original gap is a typo-level parameter choice and the paper's conclusions can stand after replacing r; if any approximation estimate or boundary term changes, the claimed h^{k−1}/h^{k+1} rates fail.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Section 4 fixes r=2N+k−2 in Lemma 4.1 and uses this r throughout. Lemma 4.4 substitutes φ=(vb−v0)l_kφ_{e_k} into identity (2.2). The factor l_kφ_{e_k} has degree 1+2(N−1)=2N−1, and v0∈P_k with vb∈P_p, p≤k, so generically deg φ=2N+k−1, one degree above P_r. Identity (2.2) is only valid for test functions in P_r; the discrete weak derivative ∂²_{ij,w}v is the P_r polynomial defined by (2.1) against P_r test functions, so the equality (∂²_{ij,w}v,φ)=... has no justification for φ outside P_r. Therefore the displayed formula (4.10) is unproved, and with it (4.11), (4.13), and the lower bound in (4.9). Since Theorem 4.5 and Theorem 6.3 both invoke (4.9), the existence/uniqueness and the H² error estimate are unsupported. The defect appears repairable: setting r=2N+k−1 makes Lemma 4.1's test function Φ_B∂²_{ij}v0 (degree 2N+k−2) admissible and keeps all Q_r approximation estimates valid because r≥k−2; hence this is an off-by-one consistency error rather than a demonstrated counterexample.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a stabilizer-free weak Galerkin (WG) finite element method for the biharmonic equation on polytopal meshes without a convexity restriction. The discrete weak second-order derivative is computed with polynomial degree r = 2N + k - 2 on an N-edge element, and bubble-function arguments are used to prove a norm equivalence between the WG energy and a discrete H^2 norm. On this basis the paper claims existence, uniqueness, and optimal-order error estimates in the discrete H^2 norm and in L^2, thereby extending the stabilizer-free WG framework of Xiu and Zhang [42] to non-convex polytopal meshes.","tokens_in":16466,"tokens_out":19378,"duration_ms":130817,"significance":"If the results are correct, the paper would provide a useful extension of stabilizer-free WG methods: it removes the convexity restriction, keeps the scheme symmetric and positive definite, and avoids stabilizers. The constructive bubble-function strategy is transparent, and the claimed optimal rates are concrete and falsifiable. The main caveat is that the key norm-equivalence lemma contains a genuine degree mismatch as written, so the central estimates are not yet supported; however, the defect appears to be a locally repairable off-by-one error rather than a demonstrated counterexample. The paper would also be stronger if its proof were self-contained with respect to the extension lemmas currently delegated to the author's arXiv preprint [23].","major_comments":[{"comment":"The admissible test-function space for identity (2.2) is P_r(T) with r = 2N + k - 2, but the function phi = (v_b - v_0) l_k phi_{e_k} has degree at most 2N - 1 + k = 2N + k - 1, because l_k is linear, phi_{e_k} has degree 2N - 2, and v_b - v_0 is generically of degree k. Thus (2.2) cannot be applied to this phi, and the displayed identity (4.10) is unjustified. Consequently the bounds (4.11) and the lower half of the norm equivalence (4.9) are not established; since (4.9) underpins the uniqueness proof in Theorem 4.5, the H^2 error estimate in Theorem 6.3, and the L^2 estimate in Theorem 7.1, the main results are unsupported as written. The defect appears repairable: setting r = 2N + k - 1 makes phi admissible, keeps Lemma 4.1 valid (the test function Phi_B partial^2_{ij} v_0 has degree 2N + k - 2), and only improves the Q_r projection estimates.","section":"Section 4, Lemma 4.4 (Eq. (4.10))"},{"comment":"The proof relies on extending v_b and the trace of v_0 from e_k to the element T and on Lemma 4.2, but Lemma 4.2 and the extension construction are attributed to the author's arXiv preprint [23], while Lemma 4.3 in this paper only describes the extension of v_g. Since the inequality (4.7) and the extension properties are load-bearing for (4.11), the paper is not self-contained. The revision should include these proofs or cite a peer-reviewed version containing them.","section":"Section 4, Lemma 4.4 and Lemma 4.2"},{"comment":"In bounding the terms ||(Q_0 u - Q_b u) n_i||_{partial T} and ||partial_i(Q_0 u) - Q_n(partial_i u)||_{partial T}, the proof passes directly to estimates involving u - Q_0 u on T. This omits the edge projection errors u - Q_b u and partial_i u - Q_n(partial_i u); a trace-plus-approximation argument for Q_b and Q_n is needed to obtain the factor h^{k-1} in (6.3). The missing step is standard and fixable, but as written the proof of (6.3), which is used in Theorems 6.3 and 7.1, is incomplete.","section":"Section 6, Lemma 6.2"}],"minor_comments":[{"comment":"The same symbol v_0 is used for the original interior polynomial and for its edge-trace extension, which makes the substitution into (2.2) ambiguous; please use a separate notation such as tilde{v}_0 for the extension.","section":"Section 4, Lemma 4.4"},{"comment":"The boundary contribution is displayed as +C h_T^{-1} int |v_b - v_0|^2 phi_{e_k} ds, whereas a direct calculation gives the signed term -h_T^{-1} n_i n_j int |v_b - v_0|^2 phi_{e_k} ds; the proof should take absolute values or specify the sum over i,j.","section":"Equation (4.10)"}],"recommendation":"major_revision","confidential_remarks":"The central idea is promising and the main flaw is an off-by-one degree mismatch that appears fixable, so I recommend major revision rather than rejection. Please ensure the revision addresses the Lemma 4.4 degree issue explicitly, fills the [23] extension gap, and corrects the Lemma 6.2 projection estimate. The paper relies heavily on self-citations; it would be helpful to verify that [23] is publicly available and that the novelty relative to [42] and [23] is clear."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a plausible extension of Xiu and Zhang's stabilizer-free WG method to non-convex polytopal meshes, and the bubble-function analysis is mostly standard. The paper's main theorems, however, rest on a norm-equivalence lemma whose proof contains a concrete degree mismatch. The test function φ=(v_b−v0)l_kφ_{e_k} has degree 2N+k−1, while the discrete weak derivative is defined only against P_r with r=2N+k−2, so (2.2) cannot be applied. That makes (4.10) and the lower bound of (4.9) unjustified. The reader's rejection on those grounds is fair.\n\nBut I don't think this is a fatal or even deep flaw. The stress-test note is right: setting r=2N+k−1 makes the argument go through. Lemma 4.1's test function Φ_B∂²_ij v0 then has degree 2N+k−2 ≤ r, and all the projection estimates in Section 6 still hold for the larger r. So the paper is an off-by-one parameter error away from being correct. That said, the error is load-bearing: uniqueness, H² error, and L² error all invoke (4.9). As written, the proofs do not establish the stated theorems.\n\nWhat is genuinely new? The method itself—using squared bubble functions to handle non-convexity in a stabilizer-free WG scheme—is a reasonable incremental step beyond [42]. The implementation is simplified compared to stabilizer-based methods, and the analysis framework follows the author's established toolbox. I do think the introduction should acknowledge the overlap with the author's own [23], which appears to target the same setting; that is a citation and presentation issue, not a mathematical one.\n\nThis paper is for readers working on weak Galerkin methods for fourth-order problems on polytopal meshes. They will find the technique useful once the r parameter is corrected. It deserves a serious referee, but only after revision. I would not desk-reject it; I would send it out with a request to fix the degree count and rerun the proofs, and to clarify the relationship to [23].","headline":"Off-by-one degree error in Lemma 4.4 invalidates the central norm equivalence as written, but the fix is straightforward and the paper deserves a refereed revision.","tokens_in":17037,"tokens_out":4341,"would_cite":false,"duration_ms":33671,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["65N30","65N15","65N12","65N20"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that a stabilizer-free weak Galerkin method for the biharmonic equation achieves optimal H^2 and L^2 convergence on both convex and non-convex polytopal meshes, using bubble functions instead of stabilization terms.","keywords":["weak Galerkin","biharmonic equation","stabilizer-free","non-convex polytopal meshes","bubble functions","optimal error estimates","weak second-order partial derivative","discrete H2 norm"],"falsifier":"Check the degree of \\((v_b - v_0) l_k \\varphi_{e_k}\\) on a pentagon with \\(k=2\\): if it exceeds \\(r = 2N + k - 2\\), identity (4.10) does not follow from (2.2). One could then verify (4.10) numerically on a single non-convex element, or run the scheme on an L-shaped domain and compare the observed discrete \\($H^{2}$\\) error rate with the claimed \\($h^{{k-1}}$\\).","tokens_in":15890,"feed_emoji":"📐","tokens_out":7152,"duration_ms":56531,"temperature":0.7,"pith_summary":"The paper aims to establish that a simplified weak Galerkin (WG) finite element method can solve the biharmonic equation on polytopal meshes without any stabilization terms, even when the mesh elements are non-convex. It claims optimal-order error estimates in both the discrete $H^{2}$ norm and the $L^{2}$ norm, extending an earlier stabilizer-free WG scheme that was limited to convex polytopal meshes. The argument replaces stabilizers with higher-degree polynomials in the discrete weak second-order derivative and uses bubble functions as the analytical tool. If the claims hold, the method is symmetric, positive definite, and simpler to program while preserving the sparsity of the stiffness matrix.","feed_headline":"Optimal error rates claimed for biharmonic scheme on non-convex meshes","feed_subtitle":"A stabilizer-free weak Galerkin scheme aims for H^2 errors of order h^{k-1} and L^2 errors of order h^{k+1}.","key_machinery":"The key machinery is the discrete weak second-order partial derivative \\(\\$partial^{2}$_{ij,w,r,T}\\), defined on each element by the integration-by-parts identity \\((\\$partial^{2}$_{ij,w} v, \\phi)_T = (v_0, \\$partial^{2}$_{ji}\\phi)_T - \\langle v_b n_i, \\partial_j \\phi\\rangle_{\\partial T} + \\langle v_{g i}, \\phi n_j\\rangle_{\\partial T}\\) for all \\(\\phi \\in P_r(T)\\). On a non-convex polytopal element with \\(N\\) edges, the paper sets \\(r = 2N + k - 2\\) and builds a full bubble function \\(\\Phi_B = \\prod_{i=1}^N $l_i^{2}$\\), which vanishes on \\(\\partial T\\), together with edge bubbles \\(\\varphi_{e_k} = \\prod_{i \\ne k} $l_i^{2}$\\). These bubbles convert boundary jump terms into interior norms and yield the two-sided bound \\(C_1 \\|v\\|_{2,h} \\le \\left|\\left|\\left|v\\right|\\right|\\right| \\le C_2 \\|v\\|_{2,h}\\). That norm equivalence is what carries both the uniqueness proof and the optimal \\($H^{2}$\\) and \\($L^{2}$\\) error estimates.","core_discovery":"The central claim is that, for the biharmonic equation with Dirichlet and Neumann boundary conditions, the WG scheme defined by \\((\\$partial^{2}$_w u_h, \\$partial^{2}$_w v) = (f, v_0)\\) with no stabilization term has a unique solution and satisfies \\(\\left|\\left|\\left|u - u_h\\right|\\right|\\right| \\le C $h^{{k-1}}$ \\|u\\|_{k+1}\\) and \\(\\|e_0\\| \\le C $h^{{k+1}}$ \\|u\\|_{k+1}\\), provided the discrete weak second-order partial derivative is computed in a polynomial space of degree \\(r = 2N + k - 2\\) on an \\(N\\)-edge element. The proof obtains these rates by establishing a norm equivalence between the WG energy norm and a discrete \\($H^{2}$\\) norm, with bubble functions providing the needed lower bound on the boundary jump terms.","pith_inferences":["If the degree-count issue in the bubble test function were repaired by raising \\(r\\) or choosing another bubble, the same error equation would still give the claimed rates, so the numerical convergence is likely robust even though the written proof has a gap.","The edge bubble construction could be reused to build stabilizer-free WG methods for other fourth-order operators such as the plate equation with variable coefficients, where boundary terms require the same kind of control.","A direct numerical test on a single non-convex element, measuring the discrete \\(H^2\\) error against \\(h^{k-1}\\) for \\(k=2,3\\), would settle whether the claimed rates hold in practice independent of the proof details."],"forward_implications":["Stabilizer-free WG schemes become available on general polytopal meshes, so non-convex cells need no special treatment at the discretization stage.","The method is symmetric and positive definite, preserving the size and global sparsity of the stiffness matrix while removing the usual stabilization parameters.","The claimed convergence rates \\(O(h^{k-1})\\) in the discrete \\(H^2\\) norm and \\(O(h^{k+1})\\) in \\(L^2\\) are optimal for polynomial degree \\(k\\), with the analysis stated to extend to \\(d \\ge 3\\) dimensions.","The bubble-function technique gives a template for proving norm equivalence in other stabilizer-free weak Galerkin schemes for fourth-order problems."],"supporting_citations":[{"why":"The stabilizer-free WG method for the biharmonic equation on convex polytopal meshes that this paper extends to non-convex elements.","marker":"[42]"},{"why":"Supplies the definition of the weak second-order partial derivative and the basic WG framework for biharmonic equations.","marker":"[31]"},{"why":"Provides the shape-regularity, trace, and inverse inequalities used throughout the error analysis.","marker":"[37]"},{"why":"Source of the edge extension and edge bubble-function lemmas used to control boundary jump terms without convexity assumptions.","marker":"[23]"}],"fun_headline_variants":["Stabilizer-free WG method hits optimal biharmonic error rates","Non-convex polytopal meshes tamed by stabilizer-free WG scheme","Bubble functions unlock optimal error bounds for biharmonic WG","Simplified WG method achieves optimal H^2 and L^2 accuracy","Stabilizer-free WG: optimal biharmonic rates on non-convex meshes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"In Lemma 4.4, the proof assumes that the edge bubble test function \\((v_b - v_0) l_k \\varphi_{e_k}\\) is a polynomial of degree at most \\(r = 2N + k - 2\\) in each \\(N\\)-edge element, so that the defining identity for the discrete weak derivative applies; a direct degree count gives one degree higher, so the key boundary identity (4.10) is not justified as written.","fun_headline_variants_meta":{"raw":{"variants":["Stabilizer-free WG method hits optimal biharmonic error rates","Non-convex polytopal meshes tamed by stabilizer-free WG scheme","Bubble functions unlock optimal error bounds for biharmonic WG","Simplified WG method achieves optimal H^2 and L^2 accuracy","Stabilizer-free WG: optimal biharmonic rates on non-convex meshes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000691,"raw_usage":{"total_tokens":3069,"prompt_tokens":826,"completion_tokens":2243,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":442,"completion_tokens_details":{"reasoning_tokens":2148}},"tokens_in":442,"tokens_out":2243,"duration_ms":14567,"temperature":1.0,"reasoning_tokens":2148,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T15:05:05.126000+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Check the degree of \\((v_b - v_0) l_k \\varphi_{e_k}\\) on a pentagon with \\(k=2\\): if it exceeds \\(r = 2N + k - 2\\), identity (4.10) does not follow from (2.2). One could then verify (4.10) numerically on a single non-convex element, or run the scheme on an L-shaped domain and compare the observed discrete \\($H^{2}$\\) error rate with the claimed \\($h^{{k-1}}$\\).","supporting_citations":[{"cited_title":"Xiu and S","cited_arxiv_id":null,"evidence_quote":"The stabilizer-free WG method for the biharmonic equation on convex polytopal meshes that this paper extends to non-convex elements."},{"cited_title":"W ang and J","cited_arxiv_id":null,"evidence_quote":"Supplies the definition of the weak second-order partial derivative and the basic WG framework for biharmonic equations."},{"cited_title":"W ang, and X","cited_arxiv_id":null,"evidence_quote":"Provides the shape-regularity, trace, and inverse inequalities used throughout the error analysis."}],"review_version":1}