{"id":"9e39fdf0-0b96-42bf-814c-d3fa5fad606e","arxiv_id":"2507.16394","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The fully nonlinear Loewner-Nirenberg problem is shown to admit solutions when mu_Gamma^+>1-delta, in particular for sigma_k with k<=n/2, and whenever any admissible conformal metric exists.","lead":"This paper proves new existence theorems for the fully nonlinear Loewner-Nirenberg problem, which asks for a complete conformal metric with prescribed negative curvature on a manifold with boundary. It settles the threshold case k=n/2 for the sigma_k version on compact Riemannian manifolds, extending earlier work by the same authors.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central threshold case rests on the unproven τ-uniform interior gradient estimate of [10, Thm 7.1]; if that estimate degenerates at τ=1, the limit construction in §3.5 is unjustified.","rationale":"The paper is structurally coherent: Theorem 1.6 is supported by a self-contained compactness argument using Lemma 2.1 and C^{1,σ}-precompactness, the algebra in Lemma 2.1 and the explicit boundary barrier in Proposition 3.8 both check out, and the interior C^0 estimates in Section 3.2 are solid. The main risk is concentrated where the reader placed it: the τ→1 limit in Section 3.5 needs a local interior gradient estimate uniform in τ, and the paper imports this from the unpublished preprint [10, Theorem 7.1] without proof. The manuscript itself flags that the μ_Γ^+=1 threshold is where one-sided estimates fail and the half-space Liouville theorem becomes non-unique, so uniformity at exactly this value is non-obvious. The secondary reliance on verbatim transfer of boundary asymptotics and viscosity-convergence arguments from [14] and [48] is also a verification burden, but those are less likely to break because the new boundary barrier gives explicit control. My read found no internal contradiction and no circular definition; the load-bearing concern is external verification of a cited estimate. Therefore the reader's CONDITIONAL verdict stands without change.","tokens_in":30144,"tokens_out":22009,"duration_ms":212265,"concrete_test":"Independently re-derive [10, Theorem 7.1] for the regularized family (f^τ,Γτ) with Γ=Γ^+_{n/2}, τ∈[1/2,1], tracking the dependence of the interior gradient constant on τ. If the bound remains finite as τ→1 with dependence only on n, the C^0 oscillation on the ball, and the background geometry, the limit in §3.5 is justified; if any constant contains a factor (1-τ)^{-1} or depends on a uniform ellipticity lower bound that vanishes at τ=1, the limit u may fail to be C^{0,1}_loc and Theorem 1.2 is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The genuinely new content is the threshold case μ_Γ^+=1, handled in §3.5 by taking τ→1. There the proof does not establish a priori estimates; it imports [10, Theorem 7.1], a local interior gradient estimate depending on two-sided C^0 bounds, and asserts it holds uniformly for all τ≤1 and in the limit τ→1. This uniformity is essential: the family {u_τ} of smooth solutions to (3.1) is bounded in C^0_loc by Proposition 3.6, but without a τ-uniform interior C^1 bound there is no subsequence converging locally uniformly to a locally Lipschitz limit u. The paper does not reproduce the proof of [10, Theorem 7.1], and the threshold is delicate: the authors themselves note that one-sided C^0 gradient estimates fail at μ=1 and that the R^n_+ Liouville problem is non-unique there. If [10, Theorem 7.1] degenerates as τ→1 for cones with (1,0,...,0)∈∂Γ (which includes Γ^+_{n/2}), the limit construction in Theorem 1.7, and hence Theorem 1.2, is unsupported. Secondary dependencies are the verbatim transfers of boundary asymptotics and maximality from [14, Section 4]; those are also cited without proof, but they are plausible given the explicit boundary barrier in Proposition 3.8.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies the fully nonlinear Loewner-Nirenberg problem on compact Riemannian manifolds with non-empty boundary. The main result, Theorem 1.2, asserts that for cones Γ with μ_Γ^+ > 1−δ, where δ depends on geometric bounds, there exists a maximal locally Lipschitz viscosity solution with boundary asymptotics u/d→1; in particular, it covers the threshold case k=n/2 for the σ_k problem. The proof is split into Theorem 1.6, which constructs an admissible conformal metric when μ_Γ^+ is close to 1 via a barrier/compactness argument, and Theorem 1.7, which produces solutions from the existence of any admissible metric. A Dirichlet version with positive boundary data is also proved. The genuinely new analytical content is the treatment of μ_Γ^+=1 by elliptic regularization τ→1.","tokens_in":30351,"tokens_out":31409,"duration_ms":294699,"significance":"If the proof is completed, the results are significant: they settle the σ_k-Loewner-Nirenberg problem for all k≤n/2, including the borderline case k=n/2, and provide a general existence criterion based on the presence of an admissible metric, without any assumption on μ_Γ^+. The paper contains several well-executed new tools: Lemma 2.1 (construction of admissible metrics for μ_Γ^+≥1 by keeping a term dropped by Yuan), Proposition 3.8 (an explicit annulus barrier), and the C^{1,σ}-compactness argument with boundary. The main caveat is that the threshold case imports its key interior gradient estimate from the preprint [10] and several convergence/asymptotics arguments from the authors' earlier work [14,48], rather than proving them in the present setting.","major_comments":[{"comment":"The proof of the threshold case τ0=1 relies on [10, Theorem 7.1], a local interior gradient estimate depending on two-sided C^0 bounds, asserted to hold uniformly for all τ≤1 and under perturbations of (f,Γ). This uniformity is what allows the subsequential convergence to u∈C^{0,1}_{loc} in the limit τ→1; without it the limit construction collapses. The estimate is imported from an arXiv preprint and is not proved or stated in the paper. Please either include a proof or state the precise theorem and verify that the family (f^τ,Γ^τ) satisfies its hypotheses uniformly, including at τ=1. This is load-bearing for Theorems 1.7 and 1.2.","section":"§3.5, proof of Theorem 1.7′; see also §1"},{"comment":"From f(λ)≤σ1(λ)/n (proved in Proposition 3.3) and f=1/2, one obtains σ1(−g_u^{−1}A_{g_u})≥n/2. Since σ1(−g^{−1}A_g)=−R_g/(2(n−1)), this yields R_{g_u}≤−n(n−1), not the stated R_{g_u}≤−2n(n−1). The comparison solution v in Lemma 3.5 solves R_{g_v}=−2n(n−1), so the comparison in Proposition 3.3 is in the wrong direction as written; the lower bounds in Propositions 3.3 and 3.6, and the lower boundary gradient estimate in Proposition 3.7, depend on this comparison. The factor in (3.3) and the displayed conformal-transformation formula for R_{g_v} need to be corrected.","section":"§3.2, Proposition 3.3 and Lemma 3.5"},{"comment":"The boundary asymptotics and maximality are transferred verbatim from [14, Section 4], and the viscosity-convergence argument is transferred from [48, Theorems 1.3 and 1.4]. The present setting includes μ_Γ^+=1 and (1,0,…,0)∈∂Γ, where the original hypotheses of [14] (μ_Γ^+>1) are not satisfied. Please spell out why the arguments of [14, Section 4] apply to the regularized family (f^τ,Γ^τ) uniformly as τ→1, or provide the necessary adaptations.","section":"§3.5, proof of Theorem 1.7′; §3.4"}],"minor_comments":[{"comment":"The displayed conformal transformation formula is missing the positive factor e^{−2e^{Nv}} on the first three terms; since Γ is a cone this does not affect the cone-membership conclusion of Lemma 2.1, but the identity as written is not correct.","section":"§2.1, equation (2.6)"},{"comment":"There are typographical errors: 'satifying' in (2.1) and a double '∈∈' in the proof of Theorem 2.8.","section":"§2.3, proof of Theorem 1.6′ and §2.4"},{"comment":"After defining v(r) in (3.6), the boundary conditions are written as v(x)=δ on {r=r1}; the notation v(r1)=δ and v(r2)=m would be clearer.","section":"§3.3, Proposition 3.8"},{"comment":"The statement that the viscosity-convergence argument in the τ0=1 case is identical to [48, Theorem 1.4] is terse; since the present cone is fully nonlinear and not restricted to σ_k on annuli, a brief outline of the stability argument would improve readability.","section":"§3.4, proof of Theorem 1.10′"}],"recommendation":"major_revision","confidential_remarks":"The paper is honest about its dependencies on [10], [14] and [48]. The main risks are whether the refereeing process can verify the uniformity of the cited gradient estimate and whether the factor-of-two issue in Section 3.2 is resolved; the latter appears fixable without changing the overall strategy. The manuscript is within the scope of a PDE/geometric analysis journal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a real result, not a repackaging. The threshold case μΓ^+=1 (which includes σ_k for k=n/2) was open, and the paper cracks it via a clean two-step argument: first obtain an admissible metric for μ>1−δ by compactness plus Lemma 2.1, then run the machinery of Theorem 1.7 to get a maximal locally Lipschitz viscosity solution. Theorem 1.7 is itself new when (1,0,…,0) lies on ∂Γ and removes the extra cone assumption earlier work needed. The paper is careful, and the internal lemmas (especially Proposition 3.8) are explicit and checkable.\n\nWhere it is soft: the threshold case depends on an interior gradient estimate of Chu–Li–Li [10, Thm 7.1] that is imported, not reproved, and the uniformity as τ→1 and under perturbations of (f,Γ) is essential. If that estimate degenerates at τ=1, the limit construction in §3.5 collapses. The authors say so themselves, and they also transfer asymptotics/maximality arguments verbatim from their earlier [14] and [48]. That is a normal division of labor in this field, but it means the paper's central claim rests on a black box. The stress-test note is right: a referee should verify that [10] has exactly the stated uniformity. I do not see a circularity problem; the threshold result is not hidden in the cited prior work.\n\nOverall: the architecture is sound, the new lemmas are real, and the unresolved item is a specific external estimate, not a vague gap. If [10] holds up, this is a significant advance in conformal geometry and fully nonlinear PDE.\n\nRecommendation: send it to a serious referee. The paper deserves referee time, and the referee's main job is to check the imported estimate.","headline":"A genuine threshold result for the fully nonlinear Loewner-Nirenberg problem, with the main risk being the paper's reliance on an external gradient estimate whose uniformity at the critical case is asserted but not reproved.","tokens_in":31009,"tokens_out":1582,"would_cite":true,"duration_ms":16356,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35J60","35D40","53C21","58J05"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves existence of complete conformal metrics solving the fully nonlinear Loewner-Nirenberg problem whenever the cone parameter exceeds $1-\\delta$, covering the threshold case $k=n/2$.","keywords":["fully nonlinear Loewner-Nirenberg problem","Schouten tensor","sigma_k-curvature","admissible cone","viscosity solution","conformal metric","elliptic regularisation","Riemannian manifold with boundary"],"falsifier":"Look for a sequence of approximating problems at $\\mu_\\Gamma^+=1$ with two-sided $C^0$ bounds but $|\\nabla u|/u$ unbounded uniformly in the regularisation parameter; the paper itself records that one-sided variants fail at $\\mu_\\Gamma^+=1$, so the two-sided uniformity is exactly the load-bearing input. A cheaper computational check is to solve the radial equation on a Euclidean annulus for $\\Gamma^+_{n/2}$ and test whether the $C^0$-normalised gradient bound is independent of $\\tau$; a blow-up would falsify the central existence claim.","tokens_in":29866,"feed_emoji":"📐","tokens_out":9974,"duration_ms":94356,"temperature":0.7,"pith_summary":"This paper proves that the fully nonlinear Loewner-Nirenberg problem—the conformal uniformisation problem prescribing a symmetric function of the eigenvalues of the Schouten tensor—admits a complete conformal metric on any compact Riemannian manifold with non-empty boundary, provided the cone parameter $\\mu_\\Gamma^+$ exceeds $1-\\delta$. The constant $\\delta$ depends only on bounds for the Ricci curvature, injectivity radii, diameter, boundary mean curvature, and a H\\\"older norm. In particular, this settles the $\\sigma_k$-Loewner-Nirenberg problem for all $k\\le n/2$, including the borderline case $k=n/2$ that resisted earlier methods. A second group of results shows that a solution exists whenever any conformal metric already keeps $-g^{-1}A_g$ inside the cone, with no condition on $\\mu_\\Gamma^+$; the solution is smooth and unique when $(1,0,\\dots,0)$ lies in the interior of $\\Gamma$.","feed_headline":"Loewner-Nirenberg problem solved at the k=n/2 threshold","feed_subtitle":"Complete conformal metrics exist on curved manifolds with boundary for every cone parameter above 1−δ.","key_machinery":"Two devices carry the argument. The first is the cone parameter $\\mu_\\Gamma^+$, defined by $(-\\mu_\\Gamma^+,1,\\dots,1)\\in\\partial\\Gamma$, which measures how wide the cone $\\Gamma$ is; previous methods worked only for $\\mu_\\Gamma^+>1$, while $\\Gamma^+_{n/2}$ has $\\mu_\\Gamma^+=1$. The second is an explicit conformal deformation $g_N=e^{2e^{Nv}}g$ generated by a function $v\\ge1$ with no critical points; a direct eigenvalue computation shows $\\lambda(-g_N^{-1}A_{g_N})$ lies in $\\Gamma$ once $\\mu_\\Gamma^+$ is close enough to $1$, and a term that earlier arguments dropped is retained to extend the range to $\\mu_\\Gamma^+\\ge1$. Compactness of Riemannian manifolds with boundary under the uniform bounds, via harmonic-radius estimates and $C^{1,\\sigma}$-precompactness, converts this qualitative statement into a uniform gap $\\delta>0$. For the solution step, the $\\tau$-regularisation $f^\\tau(\\lambda)=f(\\tau\\lambda+(1-\\tau)\\sigma_1(\\lambda)e)/(\\tau+n(1-\\tau))$ and $\\Gamma^\\tau=\\{\\lambda:\\tau\\lambda+(1-\\tau)\\sigma_1(\\lambda)e\\in\\Gamma\\}$ makes the equation elliptic for $\\tau<1$; the continuity method, two-sided $C^0$ estimates, and the imported local interior gradient estimate produce solutions, and an explicit annulus barrier controls the boundary gradient and asymptotic behaviour.","core_discovery":"The central claim is Theorem 1.2: for any manifold in the compactness class $\\mathcal{M}^n_\\sigma$ with uniform bounds on Ricci curvature, injectivity radii, boundary mean curvature, and diameter, there exists $\\delta>0$ such that whenever $\\mu_\\Gamma^+>1-\\delta$, the equation $f(\\lambda(-g_u^{-1}A_{g_u}))=1/2$ with $u=0$ on $\\partial M$ has a maximal locally Lipschitz viscosity solution $g_u=u^{-2}g_0$ satisfying $u/d_{g_0}(\\cdot,\\partial M)\\to 1$. The proof splits into two independent statements. Theorem 1.6 constructs a conformal metric $g$ with $\\lambda(-g^{-1}A_g)\\in\\Gamma$ when $\\mu_\\Gamma^+>1-\\delta$, using a critical-point-free auxiliary function and $C^{1,\\sigma}$-compactness of manifolds with boundary. Theorem 1.7 turns any admissible conformal metric into a solution of the boundary-value problem, using elliptic regularisation and stability of two-sided interior gradient estimates as the regularisation parameter tends to $1$. Consequently the $\\sigma_k$ problem is solved for all $k\\le n/2$, and the Dirichlet version with positive boundary data is solved under the same admissible-metric hypothesis.","pith_inferences":["Beyond the paper: the uniform gap $\\delta$ in Theorem 1.2 is obtained by contradiction and is non-constructive; the explicit eigenvalue computation in Lemma 2.1 suggests a quantitative lower bound in terms of the $C^1$ size of the metric and an upper bound on the Schouten tensor, which would indicate exactly how far past $\\mu=1$ the method reaches.","Beyond the paper: the same compactness argument shows that the existence statement is stable under $C^{1,\\sigma}$ perturbations of the background metric, so the admissible-metric hypothesis in Theorem 1.7 is effectively an open condition in that topology.","Beyond the paper: at the threshold $\\mu_\\Gamma^+=1$ with $(1,0,\\dots,0)\\in\\partial\\Gamma$, the solutions should be expected to be merely Lipschitz rather than smooth, consistent with the known non-differentiability examples on annuli; the paper does not claim otherwise.","Beyond the paper: a natural test of sharpness is whether $\\delta$ in Theorem 1.2 can be replaced by the full range $\\mu_\\Gamma^+>0$ in dimensions $n\\ge4$, or whether the known failure of one-sided boundary gradient estimates at $\\mu_\\Gamma^+\\le1$ marks a genuine obstruction."],"forward_implications":["For every $k\\le n/2$, the $\\sigma_k$-Loewner-Nirenberg problem on any compact Riemannian manifold with boundary in the class $\\mathcal{M}^n_\\sigma$ admits a complete locally Lipschitz viscosity solution with $u/d_{g_0}(\\cdot,\\partial M)\\to1$ at the boundary.","If $(1,0,\\dots,0)\\in\\Gamma$, the solution is smooth and is the unique continuous viscosity solution with $u=0$ on $\\partial M$.","The Dirichlet problem with positive boundary data is solvable whenever an admissible conformal metric exists; the solution is Lipschitz, and smooth and unique when $(1,0,\\dots,0)\\in\\Gamma$.","Existence follows from the mere existence of a conformal metric $g$ with $\\lambda(-g^{-1}A_g)\\in\\Gamma$, with no condition on $\\mu_\\Gamma^+$; this covers cases with $(1,0,\\dots,0)\\in\\partial\\Gamma$ that were previously open.","In dimension three, the fully nonlinear Loewner-Nirenberg problem is now solved for all admissible cones, combining Theorem 1.7 with the existing three-dimensional existence result [60]."],"supporting_citations":[{"why":"Supplies the local interior gradient estimate with two-sided $C^0$ bounds, asserted to hold uniformly for all $\\tau\\le1$ and under perturbations of $(f,\\Gamma)$, used to extract the limit solution as $\\tau\\to1$.","marker":"[10]"},{"why":"Provides the earlier existence result for $\\mu_\\Gamma^+>1$ and the boundary asymptotics and maximality arguments that are transferred verbatim in Section 3.5.","marker":"[14]"},{"why":"Gives the viscosity-limit argument and the non-differentiability examples; its Theorems 1.3 and 1.4 justify the limit passage at $\\tau=1$.","marker":"[48]"},{"why":"Source of the auxiliary critical-point-free construction for admissible conformal metrics, which the paper refines by retaining a dropped term to reach $\\mu_\\Gamma^+\\ge1$.","marker":"[60]"},{"why":"Supplies the harmonic-radius and $C^{1,\\sigma}$-compactness theory for manifolds with boundary used to convert the pointwise construction into the uniform gap $\\delta$.","marker":"[2]"},{"why":"Establishes complete conformal metrics with constant negative scalar curvature, used for the $C^0$ lower bound and for comparison functions in the boundary-value problem.","marker":"[6]"},{"why":"Provides global Hessian estimates and invertibility of the linearised operator used in the continuity method when the regularisation parameter is below $1$.","marker":"[23]"},{"why":"Gives global gradient and second-derivative estimates and linearised-operator invertibility for fully nonlinear conformal equations, used in the continuity method.","marker":"[31]"}],"fun_headline_variants":["Loewner-Nirenberg problem solved for k=n/2 and beyond","Sigma_k Loewner-Nirenberg solved for all k up to n/2","Conformal metric existence at k=n/2 threshold","k=n/2 critical case of Loewner-Nirenberg resolved","Loewner-Nirenberg with k=n/2 solved on curved manifolds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the local interior gradient estimate imported from the literature remains valid uniformly as the regularisation parameter approaches $1$ and as the cone approaches the critical case $\\mu_\\Gamma^+=1$; if this uniformity fails, the limiting argument that produces the solution collapses.","fun_headline_variants_meta":{"raw":{"variants":["Loewner-Nirenberg problem solved for k=n/2 and beyond","Sigma_k Loewner-Nirenberg solved for all k up to n/2","Conformal metric existence at k=n/2 threshold","k=n/2 critical case of Loewner-Nirenberg resolved","Loewner-Nirenberg with k=n/2 solved on curved manifolds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000539,"raw_usage":{"total_tokens":2722,"prompt_tokens":1219,"completion_tokens":1503,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":835,"completion_tokens_details":{"reasoning_tokens":1405}},"tokens_in":835,"tokens_out":1503,"duration_ms":14066,"temperature":1.0,"reasoning_tokens":1405,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T15:10:45.286185+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Look for a sequence of approximating problems at $\\mu_\\Gamma^+=1$ with two-sided $C^0$ bounds but $|\\nabla u|/u$ unbounded uniformly in the regularisation parameter; the paper itself records that one-sided variants fail at $\\mu_\\Gamma^+=1$, so the two-sided uniformity is exactly the load-bearing input. A cheaper computational check is to solve the radial equation on a Euclidean annulus for $\\Gamma^+_{n/2}$ and test whether the $C^0$-normalised gradient bound is independent of $\\tau$; a blow-up would falsify the central existence claim.","supporting_citations":[{"cited_title":"Liouville theorems for conformally invariant fully nonlinear equations. I","cited_arxiv_id":"2311.07542","evidence_quote":"Supplies the local interior gradient estimate with two-sided $C^0$ bounds, asserted to hold uniformly for all $\\tau\\le1$ and under perturbations of $(f,\\Gamma)$, used to extract the limit solution as $\\tau\\to1$."},{"cited_title":"The $\\sigma_k$-Loewner-Nirenberg problem on Riemannian manifolds for $k<\\frac{n}{2}$","cited_arxiv_id":"2310.01346","evidence_quote":"Provides the earlier existence result for $\\mu_\\Gamma^+>1$ and the boundary asymptotics and maximality arguments that are transferred verbatim in Section 3.5."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the viscosity-limit argument and the non-differentiability examples; its Theorems 1.3 and 1.4 justify the limit passage at $\\tau=1$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Source of the auxiliary critical-point-free construction for admissible conformal metrics, which the paper refines by retaining a dropped term to reach $\\mu_\\Gamma^+\\ge1$."},{"cited_title":"Anderson, A","cited_arxiv_id":null,"evidence_quote":"Supplies the harmonic-radius and $C^{1,\\sigma}$-compactness theory for manifolds with boundary used to convert the pointwise construction into the uniform gap $\\delta$."},{"cited_title":"A viles and R","cited_arxiv_id":null,"evidence_quote":"Establishes complete conformal metrics with constant negative scalar curvature, used for the $C^0$ lower bound and for comparison functions in the boundary-value problem."},{"cited_title":"Guan, Complete conformal metrics of negative Ricci curvature on compact manifolds with boundary, Int","cited_arxiv_id":null,"evidence_quote":"Provides global Hessian estimates and invertibility of the linearised operator used in the continuity method when the regularisation parameter is below $1$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives global gradient and second-derivative estimates and linearised-operator invertibility for fully nonlinear conformal equations, used in the continuity method."}],"review_version":1}