{"id":"ce7a27ec-3703-4eac-85a9-343b18907e61","arxiv_id":"2411.15227","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For the Finsler p-Laplacian Dirichlet problem with u^q nonlinearity, the paper proves uniqueness of the positive solution for 3/2 < p < 2 and q close to p-1, and for any q < Q_{0,p} when the operator is close to the Euclidean one.","lead":"This paper proves that a Finsler (direction-dependent) version of the p-Laplacian equation has a unique positive solution when the exponent p is between 3/2 and 2 and the nonlinearity sits just above the sublinear threshold. It extends a 2023 uniqueness result of Brasco and Lindgren, which covered the smoother p > 2 case, into the harder singular regime.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Final contradiction in Theorem 1.2, Step 3 does not follow: the measure lower bounds on the sign sets of w_k are compatible with strong L^2 convergence to a sign-definite limit, and no uniform L^2 lower bound is proved.","rationale":"The reader's weakest_assumption pointed to the imported uniform regularity (Theorem 2.1) and integrability of the inverse gradient (Theorem 2.3). Those are indeed external pillars, and the paper would be stronger if their verification for the Finsler setting were supplied. However, the most load-bearing problem I found is internal and more severe: the final contradiction in the proof of Theorem 1.2 is logically incomplete. The argument up to (4.16) is plausible and the convergence of w_k to a sign-definite limit w is well motivated, but the step from the measure lower bound to a contradiction relies on an implication that is false without a lower bound on the L^2 mass of the sign parts. This is not a matter of missing regularity constants; it is a gap in the reasoning that establishes uniqueness. I therefore agree with the reader's CONDITIONAL verdict rather than moving to ACCEPT or REJECT: the central claim may be true and the method may be repairable, but the proof as written does not close the argument. The reader's mention that the final contradiction is 'written more quickly than it needs to be' anticipates this issue, but it is stronger than a style complaint: it is a genuine gap. If a referee can show a uniform L^2 lower bound for w_k^± from the existing estimates, or replace the measure bound with a maximum-principle argument, the theorem would be restored. Until then, the uniqueness of all positive solutions for q near p-1 in the range 3/2 < p < 2 remains an unproved assertion of the manuscript.","tokens_in":19972,"tokens_out":15236,"duration_ms":154648,"concrete_test":"Ask for a uniform lower bound on the weighted L^2 norm of each sign part of w_k, or a Harnack/maximum-principle argument for the singular linearized operator, to replace the measure bound in the display after (4.16). Concretely: re-derive the chain (4.16)-(4.11)-(Theorem 2.2) keeping explicit track of the L^2 norm of w_k^±; if the only k-independent quantity that survives is |Ω_k^±|, the contradiction is unsupported. As a minimal logical test, exhibit a sequence w_k ∈ W_0^{1,2}(Ω) with uniformly bounded ∫ |∇w_k|^2, strong L^2 limit w ≥ 0, w ≢ 0, and |{w_k < 0}| ≥ c > 0; such a sequence shows the claimed contradiction step is invalid without additional estimates.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The proof of Theorem 1.2 ends (Section 4, Step 3) with a non sequitur. After (4.16), the authors obtain only |Ω_k^±| ≥ 1/C' for the supports of the positive and negative parts of the normalized difference w_k. They then assert that this contradicts the strong L^2 convergence of w_k to a sign-definite, nontrivial limit w. But a measure lower bound does not prevent the L^2 mass of a sign part from collapsing: e.g., in any bounded domain, w_k = 1 - k^{-1} χ_E with |E| ≥ c has negative part supported on a set of measure ≥ c while w_k → 1 in L^2. No uniform lower bound on ∥w_k^±∥_{L^2} or on the corresponding weighted L^2 norm is derived anywhere in Step 3; the inequalities leading to the measure bound lose exactly the amplitude information needed for a contradiction. Since this is the only step that eliminates the possibility of two distinct positive solutions, the uniqueness conclusion of Theorem 1.2 is not established as written. The same issue does not affect Theorem 1.1, which uses non-degeneracy directly, but it is load-bearing for the paper's central extension to the singular range 3/2 < p < 2.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies uniqueness of positive solutions to the Finsler p-Laplacian Dirichlet problem -Δ_F^p u = u^q in Ω, u=0 on ∂Ω, for 3/2 < p < 2 and q > p-1. Theorem 1.1 claims uniqueness for q in (p-1, Q_{0,p}) when the Finsler norm F is sufficiently close to the Euclidean norm, assuming the Euclidean problem has a unique non-degenerate positive solution. Theorem 1.2 claims, for every bounded C^2 domain and every admissible F, the existence of q_0 > p-1 such that uniqueness holds for all q in (p-1, q_0). The proof linearizes around two hypothetical distinct solutions, normalizes their difference in a weighted L^2 norm, passes to a limit in a weighted Sobolev space using a Hardy-type embedding, and aims for a contradiction by showing the limit would be a sign-definite first eigenfunction of a weighted eigenvalue problem, while the approximating sign-changing differences have both sign sets of uniformly positive measure.","tokens_in":20081,"tokens_out":30313,"duration_ms":286502,"significance":"If the main results are correct, the paper would substantially extend the uniqueness theory of Brasco and Lindgren from the ground-state setting and the range p>2 to all positive solutions and the singular range 3/2 < p < 2, in the anisotropic Finsler framework. The paper contains a number of well-structured components: a blow-up argument for uniform L∞ bounds, a Hardy-type uniform weighted embedding (Theorem 2.2), a detailed treatment of the weighted eigenvalue problem (Section 3), and a clean use of non-degeneracy in Theorem 1.1. The proofs of Propositions 3.1 and 3.2 are essentially complete and the overall strategy is coherent. However, the final contradiction in Theorem 1.2 is not logically justified as written, and this is load-bearing for the paper's central extension.","major_comments":[{"comment":"The final contradiction is a non sequitur. The authors prove only the measure lower bound |Ω_k^±| ≥ 1/C' for the supports of the positive and negative parts of w_k, and then assert that this contradicts the strong L^2 convergence of w_k to a sign-definite nontrivial limit w. But a measure lower bound is compatible with collapse of the L^2 mass: for instance, in any bounded domain, w_k = 1 - k^{-1} χ_E with |E| ≥ c satisfies |{w_k < 0}| ≥ c while w_k → 1 strongly in L^2. No uniform lower bound on ||w_k^±||_{L^2} or on the corresponding weighted L^2 norm is derived in Step 3; equation (4.16) only relates the weighted norm of a sign part to its energy, and both can tend to zero simultaneously. Since this is the only step that excludes the existence of two distinct positive solutions, the proof of Theorem 1.2 is incomplete as written.","section":"§4, Step 3 (after Eq. (4.16))"},{"comment":"The uniform Hardy-type inequality is proved for N > 2, since the exponent σ0 = 2 + 2(2+s0)/(N-2) is undefined for N = 2 and the Sobolev exponent 2^* = 2N/(N-2) is infinite in that case. The main theorems allow N = 2, and Step 3 uses the measure bound derived from Theorem 2.2 for some 2 < σ < σ0. The gap is probably fixable by taking σ0 arbitrarily large and using the two-dimensional Sobolev embedding for finite exponents, but the N = 2 case is not covered by the proof as written.","section":"Theorem 2.2 and Corollary 2.1"},{"comment":"The uniform C^{1,γ} estimates and, in particular, the boundary-layer gradient lower bound |∇u_{F,q}| ≥ μ0 in Ω_τ are imported from [4, Theorem 2.5] with [18]. The paper states that these results apply for all p > 1, but the cited [4] concerns the case p > 2, while the present application is 3/2 < p < 2. Since Theorem 2.2, Corollary 2.1, and the compactness passage in Step 2 all depend on this uniformity, the authors should either provide a proof for the anisotropic p < 2 setting or give a precise reference that covers it. This is a missing justification for a load-bearing input, not merely a citation issue.","section":"Theorem 2.1"}],"minor_comments":[{"comment":"The exponent Q_{0,p} is not defined in the paper; it is only said to be defined in [2, Section 4]. The range q ∈ (p-1, Q_{0,p}) should be stated explicitly or the relevant condition should be reproduced.","section":"Introduction and Theorem 1.1"},{"comment":"There are several typos and formatting issues: 'finslerp-Laplacian' in section headings, 'largerly' for 'largely' in Section 1.3, and inconsistent uses of 'Hölder' vs. 'Holder' in the text.","section":"Throughout"},{"comment":"The proof explicitly treats only the case s < 0; the case s ≥ 0 is dismissed with 'We only deal with the case s < 0'. A short sentence explaining why the other case is easier would improve readability.","section":"Theorem 2.2 proof"},{"comment":"The convergence of A_k to A∞ in Z_δ^c requires a dominated-convergence argument in the t-integration when μ' = 0, because the integrand D^2H(t∇ũ_k + (1-t)∇ṽ_k) is singular near t = 0. The paper does not spell this out; the statement 'A_k → A∞ uniformly in Z_δ^c' is not immediate.","section":"§4, Step 2"},{"comment":"The derivation of (4.12) for the double integral weight ∫_0^1 (tũ_k+(1-t)ṽ_k)^{q_k-1} dt is only sketched. Since this weight is not literally of the form |u_k|^{q_k-1}, a few lines explaining the reduction to Theorem 2.2 would be helpful.","section":"Corollary 2.1 and Eq. (4.12)"}],"recommendation":"major_revision","confidential_remarks":"The gap in Step 3 of Theorem 1.2 is serious: the final contradiction is not justified by the measure lower bound, and no alternative argument is supplied. The theorem may still be true, but the current proof does not establish it. Theorem 1.1 appears to be on much firmer ground and could potentially be published separately if the authors are able to fix or remove the defective Step 3. The imported regularity input (Theorem 2.1) also deserves scrutiny, since the reference [4] treats the p > 2 case. I would not recommend rejection outright, because the overall strategy is plausible and the needed fix may be local, but the authors must supply a genuine uniform lower bound on the sign-part masses or find another way to derive the contradiction."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. The genuinely new part is the singular range 3/2 < p < 2 for the Finsler p-Laplacian, and the Hardy-based compact embedding in Section 2.2 is a different mechanism from the p > 2 computation in Brasco-Lindgren. The linearized eigenvalue setup is adapted carefully, and the identification λ(Ω;U) = p(p−1)λ_F(Ω) is a nice piece of work. That part deserves credit.\n\nThe soft spot is load-bearing. Step 3 of Theorem 1.2 ends by claiming that the measure lower bound |Ω_k^±| ≥ 1/C′ contradicts strong L^2 convergence of w_k to a sign-definite limit. That inference is false. A measure bound on the supports says nothing about L^2 mass: in a bounded domain, w_k = 1 − k^{−1}χ_E with |E| > 0 has a negative part supported on a set of positive measure while w_k → 1 in L^2. Nothing in Step 3 derives a uniform lower bound on ‖w_k^±‖_{L^2} or on the corresponding weighted norm. Since that is the only step eliminating two distinct positive solutions, Theorem 1.2 is not established as written.\n\nOther issues are comparatively minor. The proof leans on Theorem 2.1 and Theorem 2.3 imported without proof; a referee would want verification that the uniformity in F and q is exactly as stated. The remark that the result extends to p > 2 via [19, Lemma 3] is asserted, not proved. There are also typos and some compressed limit passages, but the overall architecture is coherent.\n\nDon't read this as the paper being careless. The authors know the right toolkit, and the embedding section is a real contribution. But Theorem 1.2 needs a repair: either replace the measure argument with a weighted L^2 lower bound, or change the contradiction. Theorem 1.1, which uses non-degeneracy directly and does not need Step 3, may survive intact.\n\nVerdict: worth sending to a serious referee, but with the expectation of substantial revision; conditional acceptance at best.","headline":"The paper makes a real advance in the singular Finsler p-Laplacian regime, but Theorem 1.2's final contradiction is a non sequitur as written.","tokens_in":20824,"tokens_out":2048,"would_cite":false,"duration_ms":21859,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35J92","35A02","35B45","35J60"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that a Finsler p-Laplacian Dirichlet problem has at most one positive solution when 3/2 < p < 2 and the exponent q sits just above p−1.","keywords":["Finsler p-Laplacian","anisotropic elliptic equations","uniqueness of positive solutions","linearized method","weighted Sobolev embedding","Hardy inequality","p-Laplacian singular case","Lane-Emden problem"],"falsifier":"Compute or numerically test the boundary gradient of normalized solutions for a family of $C^{3,\\alpha}$-near-Euclidean Finsler norms with $p<2$ and $q$ approaching $p-1$: if for some admissible domain and exponent the infimum of $|\\nabla u|$ on the boundary layer tends to zero, the uniform lower bound behind the compactness step is false. Alternatively, exhibit one Finsler norm $F$ and one exponent $q$ in the stated interval with two distinct positive solutions, for example on a $C^2$ domain where the Euclidean problem already admits multiple positive solutions.","tokens_in":19582,"feed_emoji":"📐","tokens_out":7588,"duration_ms":76425,"temperature":0.7,"pith_summary":"This paper proves that the anisotropic Finsler p-Laplacian Dirichlet problem $-\\Delta^F_p u = u^q$ with zero boundary data has at most one positive solution on a bounded $C^2$ domain when $3/2 < p < 2$ and $q$ is close to $p-1$ from above. The near-critical regime needs no restriction on the anisotropy: uniqueness is forced by the structure of the equation. A companion perturbation statement says that for each fixed admissible $q$, if the Euclidean p-Laplace problem has a unique non-degenerate positive solution, then every Finsler norm sufficiently close to the Euclidean norm in $C^{3,\\alpha}$ on the unit sphere inherits uniqueness. This moves a previously known ground-state uniqueness result to all positive solutions and into the singular exponent range $p<2$.","feed_headline":"Finsler p-Laplacian gains unique positive solutions","feed_subtitle":"Any two positive solutions coincide when q is close to p−1 and p lies in the singular range.","key_machinery":"The engine is the linearized method: normalize two hypothetical positive solutions by their $L^\\infty$ maxima, form their sign-changing difference $w_k$, and show that $w_k$ converges to a first eigenfunction of the weighted linear eigenvalue problem $\\lambda(\\Omega; U) = \\inf \\int_\\Omega \\langle D^2H(\\nabla U)\\nabla\\varphi, \\nabla\\varphi\\rangle \\, dx \\big/ \\int_\\Omega U^{p-2}\\varphi^2\\, dx$, whose only minimizers are $\\pm U$ and whose value equals $p(p-1)\\lambda_F(\\Omega)$. Two imported regularity inputs carry the argument: a uniform $C^{1,\\gamma}$ bound together with a boundary-layer gradient lower bound $|\\nabla u| \\ge \\mu_0$ for normalized solutions, and the integrability of $|\\nabla u|^{-r}$ for every $r < p-1$. These convert the classical Hardy inequality into a uniform weighted compact embedding, which is what lets the sequence of normalized differences pass to the limit.","core_discovery":"The central claim is that, under the stated hypotheses, the normalized difference of any two supposed positive solutions leads to a contradiction: it must change sign by a Picone-type comparison, but the linearized equation forces its limit to be a first eigenfunction of a weighted linear eigenvalue problem, and first eigenfunctions have one sign. The proof blows up solutions about their maximum points, uses a boundary gradient lower bound together with Hardy's inequality to obtain a uniform weighted compact embedding relative to $|\\nabla \\tilde u|$, and then passes to the limit as $q_k \\to p-1$. The limit function $w$ solves $-\\operatorname{div}(D^2H(\\nabla \\tilde u)\\nabla w) = p(p-1)\\lambda_F(\\Omega)\\, \\tilde u^{p-2} w$, and the only attainable minimizers of this weighted problem are $\\pm \\tilde u$, so the sign contradiction is unavoidable. The restriction $p > 3/2$ enters because the proof needs the inverse-gradient integrability estimate $|\\nabla u|^{-r} \\in L^1$ for $r = 2-p$, which requires $r < p-1$.","pith_inferences":["The open interval $1 < p \\le 3/2$ is a natural target for a counterexample search: the proof uses $p > 3/2$ only to make $r = 2-p$ admissible in the inverse-gradient integrability estimate, so any failure of uniqueness on that interval would most plausibly stem from exactly that input.","Theorem 1.1 suggests that uniqueness is structurally stable under small anisotropic perturbations of the Euclidean norm; a quantitative question follows naturally, namely how large the admissible $C^{3,\\alpha}$ distance $\\delta_1$ can be for a fixed $q$, and whether uniqueness persists away from the near-critical regime.","The Hardy-type weighted compact embedding built in the proof depends only on the uniform boundary gradient lower bound, not on the specific polynomial form $u^q$; it could therefore be reused for nonlinearities with boundary singularities or for systems driven by the same anisotropic operator."],"forward_implications":["For every bounded $C^2$ domain $\\Omega$ and $3/2 < p < 2$, there is $q_0 = q_0(N,\\Omega,p) > p-1$ such that the anisotropic Lane-Emden problem has exactly one positive solution whenever $p-1 < q < q_0$.","The earlier ground-state-only uniqueness result for the isotropic p-Laplacian is upgraded to uniqueness of all positive solutions, and the exponent range now includes the singular p-Laplacian case $p<2$.","Near the Euclidean norm in $C^{3,\\alpha}(S^{N-1})$, uniqueness persists for each fixed $q \\in (p-1, Q_{0,p})$ whenever the Euclidean problem has a unique non-degenerate positive solution.","Along any sequence $q_k \\to p-1$, normalized solutions converge in $C^{1,\\beta}$ to the first eigenfunction of $-\\Delta^F_p$, and the normalization constants $M_k^{q_k+1-p}$ converge to the first eigenvalue $\\lambda_F(\\Omega)$."],"supporting_citations":[{"why":"Supplies the linearized-method template, the weighted eigenvalue problem, and the uniform $C^{1,\\gamma}$ regularity theorem that this proof imports as Theorem 2.1.","marker":"[4]"},{"why":"Provides the blow-up argument that bounds $M^{q+1-p}$ as $q$ approaches $p-1$.","marker":"[19]"},{"why":"Gives the global $C^{1,\\alpha}$ boundary regularity behind the uniform estimates in Theorem 2.1.","marker":"[18]"},{"why":"Proves the anisotropic inverse-gradient integrability $|\\nabla u|^{-r}$ used with $r=2-p$, which forces the condition $p > 3/2$.","marker":"[8]"},{"why":"Proves the corresponding p-Laplacian inverse-gradient integrability invoked alongside the anisotropic version.","marker":"[9]"},{"why":"Supplies the classical Hardy inequality that the paper converts into the uniform weighted embedding near the boundary.","marker":"[6]"},{"why":"Defines $Q_{0,p}$ and supplies the Liouville theorems used in the near-Euclidean perturbation theorem.","marker":"[2]"},{"why":"Provides the Picone-type inequality in the Finsler setting used to show that the difference of two distinct positive solutions must change sign.","marker":"[3]"}],"fun_headline_variants":["Finsler p-Laplacian positive solutions are unique","Only one positive solution for Finsler p-Laplacian","Uniqueness proven for Finsler p-Laplacian equations","Finsler p-Laplacian has a single positive solution","Positive solutions unique in Finsler p-Laplacian case"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof leans on an imported regularity estimate: any solution normalized by its maximum has a uniformly controlled first derivative and, near the boundary, a gradient size bounded away from zero, with constants independent of the exponent; if that boundary lower bound ever fails, the Hardy-type compactness step that drives the contradiction collapses.","fun_headline_variants_meta":{"raw":{"variants":["Finsler p-Laplacian positive solutions are unique","Only one positive solution for Finsler p-Laplacian","Uniqueness proven for Finsler p-Laplacian equations","Finsler p-Laplacian has a single positive solution","Positive solutions unique in Finsler p-Laplacian case"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000313,"raw_usage":{"total_tokens":1746,"prompt_tokens":877,"completion_tokens":869,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":493,"completion_tokens_details":{"reasoning_tokens":783}},"tokens_in":493,"tokens_out":869,"duration_ms":8128,"temperature":1.0,"reasoning_tokens":783,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:35:16.578068+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute or numerically test the boundary gradient of normalized solutions for a family of $C^{3,\\alpha}$-near-Euclidean Finsler norms with $p<2$ and $q$ approaching $p-1$: if for some admissible domain and exponent the infimum of $|\\nabla u|$ on the boundary layer tends to zero, the uniform lower bound behind the compactness step is false. Alternatively, exhibit one Finsler norm $F$ and one exponent $q$ in the stated interval with two distinct positive solutions, for example on a $C^2$ domain where the Euclidean problem already admits multiple positive solutions.","supporting_citations":[{"cited_title":"Brasco and E","cited_arxiv_id":null,"evidence_quote":"Supplies the linearized-method template, the weighted eigenvalue problem, and the uniform $C^{1,\\gamma}$ regularity theorem that this proof imports as Theorem 2.1."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the blow-up argument that bounds $M^{q+1-p}$ as $q$ approaches $p-1$."},{"cited_title":"Castorina, G","cited_arxiv_id":null,"evidence_quote":"Proves the anisotropic inverse-gradient integrability $|\\nabla u|^{-r}$ used with $r=2-p$, which forces the condition $p > 3/2$."},{"cited_title":"Damascelli and B","cited_arxiv_id":null,"evidence_quote":"Proves the corresponding p-Laplacian inverse-gradient integrability invoked alongside the anisotropic version."},{"cited_title":"Brezis and M","cited_arxiv_id":null,"evidence_quote":"Supplies the classical Hardy inequality that the paper converts into the uniform weighted embedding near the boundary."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines $Q_{0,p}$ and supplies the Liouville theorems used in the near-Euclidean perturbation theorem."},{"cited_title":"Brasco and G","cited_arxiv_id":null,"evidence_quote":"Provides the Picone-type inequality in the Finsler setting used to show that the difference of two distinct positive solutions must change sign."}],"review_version":1}