{"id":"47d2adbd-9a7b-42de-9044-4e0eb971c1fb","arxiv_id":"2506.12852","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Area-minimizing hypersurfaces are generically smooth in ambient dimension 11 after a C-infinity-small perturbation of the boundary or metric, and in dimensions 12 and up the singular set has dimension at most n-10-epsilon_n.","lead":"This paper proves that in 11-dimensional ambient space, almost every boundary curve (or metric) can be perturbed slightly so that the area-minimizing surface it spans is completely smooth. This settles the last open dimension for generic regularity of minimal hypersurfaces and improves the singular-set dimension bound in every higher dimension.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The R^11 borderline case depends on strong integrability of quadratic cones (Def. 3.12) via Cor. 3.34 and Thm 4.13; the paper cites [SS86, Prop. 2.7] but does not itself verify the degree-0 and degree-1 Jacobi-field classification, leaving the most load-bearing external premise untested in the text.","rationale":"The reader's weakest-assumption analysis and mine agree on the location of the most sensitive premise: the slow-decay dimension reduction in §7.7 rests on strong integrability of the quadratic cone factor, used in Corollary 3.34 and Theorem 4.13. I agree that this is the single most load-bearing condition for the new R^11 result. My emphasis is slightly different: the paper's Proposition A.4 delegates the entire verification to [SS86, Proposition 2.7], so the open risk is whether that citation covers exactly the classification needed for all minimizing quadratic cones, not whether the surrounding argument uses the condition correctly. I did not find an internal gap in the use of strong integrability: Lemma 4.12 does construct the required u∈Jac*_1(C)∩(Jacrot(C))^⊥, and Theorem 4.13 properly invokes Corollary 3.34 to obtain the proper subspace V. Given that [SS86] is a well-established source and the claim is plausible, this concern does not by itself overturn the result; it supports the reader's CONDITIONAL verdict pending expert or computational verification. Two other external premises feed the same strict-inequality step: the positive spectral gap for non-quadratic cones (Lemma 5.4, cited to [Wan24]) and the isolation of quadratic cones (Corollary A.5). Both would also need to be checked for the epsilon improvements, but they were not the specific reason the R^11 borderline enters, so I did not make them the headline concern. Overall, the paper is carefully written, the internal logic is coherent, and the correct verdict remains conditional until the external Jacobi-field classification is confirmed.","tokens_in":75632,"tokens_out":11445,"duration_ms":142293,"concrete_test":"For every minimizing quadratic cone C_{p,q}⊂R^8 (and more generally p+q≥6), independently recompute the spectrum of −(Δ_{L}+|A_L|^2) on the link using the explicit eigenfunctions on S^p×S^q, and via Eq. (3.4) list all eigenfunctions with homogeneity degrees γ=0 and γ=1. Verify that the γ=0 eigenspace is exactly the span of infinitesimal translations normal to C_{p,q} and that the γ=1 eigenspace is exactly the span of infinitesimal rotations; if an extra eigenfunction occurs, test Corollary 3.34 numerically by constructing the corresponding V_u and checking whether it can be a proper subspace. A clean symbolic/computer-algebra check of [SS86, Prop. 2.7] for the four smallest minimizing cases (p,q)=(3,3),(2,4),(4,2),(3,4) and their permutations would settle whether the load-bearing premise lands.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 1.1 for ambient dimension 11 is obtained by making the dimension bounds in Theorem 1.4 strict. The borderline term is the k=3, C=C°×R^3 case with C°⊂R^8 a minimizing quadratic cone. In that case the old bound dim spine C/(1+α(C)) equals 1, so the proof must use the improved dimension reduction of §7.7. That reduction, via Theorem 4.13 and Corollary 3.34, requires every degree-0 homogeneous Jacobi field on C° to be a translation and every degree-1 field to be a rotation: strong integrability. The paper asserts this for all minimizing quadratic cones in Proposition A.4 with the one-line proof 'follows from [SS86, Proposition 2.7]'. This is the most load-bearing external input: if some minimizing quadratic cone in R^8 had an extra degree-0 or degree-1 Jacobi field, then Remark 3.22 would fail, the subspace V_u in Corollary 3.34 need not be proper, and the claimed k→k−1 improvement in §7.7 would collapse. In that case the R^11 borderline term would remain exactly at dimension 1 and generic smoothness would not follow from the present argument. I found no internal inconsistency in how strong integrability is used; the risk is concentrated in whether the cited [SS86] result really gives the stated classification for all minimizing quadratic cones, including non-Simons cones C_{p,q} with p+q≥6. This is a citation-verification risk rather than an evident mathematical error.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies generic regularity of area-minimizing hypersurfaces. For a smooth closed oriented (n-1)-dimensional boundary Gamma in R^{n+1}, Theorem 1.1 asserts that arbitrary C^infinity-small perturbations Gamma' make all minimizing integral currents with boundary [[Gamma']] equal to [[Sigma']] for a smooth oriented hypersurface Sigma', with sing Sigma' empty when n+1 <= 11 and dim sing Sigma' <= n-10-epsilon_n otherwise. Theorem 1.2 is the analogous statement for area-minimization in integral homology classes under generic metric perturbations. The proofs are via a foliation-type argument: a collection F of pairwise disjoint minimizing boundaries carrying a Lipschitz time function T is analyzed through a covering tree (Section 6), and Theorem 1.4 bounds dim T(sing F) and the level sets dim(sing F cap T^{-1}(t)). The new technical content is a second-order blow-up analysis near cylindrical hypercones C = C^circ x R^k (Sections 3-4), organized around a discrete Almgren-type decay order: fast decay gives improved Holder exponents, slow decay gives a k -> k-1 improvement of the spine dimension under a strong-integrability assumption. Sections 8-9 reduce the Plateau and homology theorems to Theorem 1.4 by constructing generic families with the required properties.","tokens_in":76035,"tokens_out":19260,"duration_ms":221035,"significance":"If correct, the paper settles generic smoothness of minimizing hypersurfaces in ambient dimension 11 and improves the generic singular-set bound in all dimensions >= 12; this is a substantial step beyond the previous R^9/R^10 results. The manuscript is unusually careful: Sections 7-9 give detailed proofs of the covering-tree estimates, the spectral quantities alpha(C), Delta^{qd}, and Delta^{non-qd} are defined independently of the target conclusion, and the principal external inputs (Zhu's spectral gap, the Edelen-Szekelyhidi Liouville theorem, Simon's non-concentration estimates, and the quantitative gap theorem from Wan's GAFA paper) are cited explicitly rather than re-derived. The fast/slow decay dichotomy and the use of beta-harmonic polynomials for degree-one Jacobi fields are new and convincing. The main caveat is the verification burden for strong integrability, which is load-bearing for the R^11 borderline case.","major_comments":[{"comment":"The proof that every quadratic hypercone is strongly integrable is a one-line citation to [SS86, Proposition 2.7]. This property is what makes the subspace V_u in Corollary 3.34 proper, and the k -> k-1 dimension reduction in Section 7.7 is exactly what removes the R^11 borderline term; if any minimizing quadratic cone in R^8 failed strong integrability, or if unexpected degree-one Jacobi fields on C = C^circ x R^k arose from components not controlled by Jac0(C^circ) and Jac1(C^circ), Theorem 1.1 would not follow from the present argument. Please quote the precise statement of [SS86, Proposition 2.7], indicate how it rules out all homogeneous Jacobi fields of degrees 0 and 1 beyond translations and rotations, and state explicitly that it covers the non-Simons cones C_{p,q} with p+q >= 6. Relatedly, Remark 3.22 asserts an equivalence between strong integrability of C^circ and equalities involving Jac*_{0,0}(C), Jac*_{1,0}(C), Jac*_{0,1}(C), and Jacrot(C); the text before it records only inclusions (Lemma 3.21), so the missing argument should be supplied. This is a verification request rather than an assertion of error, but as written the most specialized load-bearing premise is left entirely to a citation.","section":"Appendix A.4, Proposition A.4, used through Corollary 3.34 and Theorem 4.13 in Section 7.7"}],"minor_comments":[{"comment":"The statement says the limiting Jacobi field u is 'L^2(C cap B_1)-orthogonal to Jacrot(C)^perp'; this should read 'orthogonal to Jacrot(C)'.","section":"Lemma 4.12"},{"comment":"The constant kappa_C used in Lemma H.5 is not defined before the displayed identity; please introduce it when the graph domain D_Phi is fixed.","section":"Appendix H, after equation (H.10)"},{"comment":"The displayed definition of epsilon_n is broken across a line inside the outer min; please ensure the brace matching is unambiguous so the reader can compare the two terms.","section":"Corollary 1.6(ii)"},{"comment":"The phrase 'graph_C h_j over a connected exhaustion of C cap B_{1/4}' could be clarified by specifying the domain of h_j and the choice of normal, since the sign and positivity of the limiting Jacobi field are important.","section":"Proposition 7.6, Claim 7.8"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is very impressive and appears internally coherent. My concern is strictly about the verification burden for Proposition A.4: strong integrability is a one-line citation to [SS86, Proposition 2.7] but is load-bearing for the borderline R^11 argument through Corollary 3.34 and Theorem 4.13. If the authors quote the exact statement, confirm it covers all minimizing quadratic cones including the non-Simons ones, and justify the equality asserted in Remark 3.22, I would be happy to accept. I found no evidence of circularity in the spectral quantities or the quantitative decay orders."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper completes the CMS program by handling ambient dimension 11 and improving the singular-set bound in higher dimensions. The genuinely new step is going one blow-up level deeper: modeling minimizers as graphs of Jacobi fields over cylindrical hypercones and introducing a discrete Almgren frequency with a fast/slow decay dichotomy. In the slow-decay case the \"effective spine\" dimension reduction is a real new idea, and the algebra in Corollary 1.6 correctly identifies R^11 as the borderline where the old cone-splitting argument fails exactly at quadratic cylindrical cones. The paper is honest about its limits (Remark 1.8) and careful with external inputs: Zhu's spectral gap, Edelen–Székelyhidi's Liouville theorem, and Simon's non-concentration estimates are all cited precisely. Theorems 1.1, 1.2 and the quantitative Theorem 1.4 are proved in detail, and I found no internal contradiction.\n\nThe main soft spot is the one flagged by the stress test: the borderline R^11 case depends on strong integrability of quadratic cones (Definition 3.12), i.e., every degree-0 homogeneous Jacobi field is a translation and every degree-1 field is a rotation. The paper cites [SS86, Proposition 2.7] in Proposition A.4 with a one-line proof. The risk is real because this is load-bearing: if some minimizing quadratic cone in R^8 had an extra degree-0 or degree-1 Jacobi field, the proper subspace V_u in Corollary 3.34 would not be proper and the k-to-(k-1) improvement in Section 7.7 would collapse. That said, this is a citation-verification risk rather than an evident mathematical error. The cited result is plausible and standard, and the paper explicitly lists strong integrability as a hypothesis in Remark 1.7, so the reliance is transparent. I would not call this a fatal flaw.\n\nThe paper deserves a serious referee. It is long and technical, but the proof structure is clear and the new ideas are substantial. I would bring it to a reading group and would cite it if I worked in generic regularity.","headline":"Strong extension of the CMS generic-regularity program that closes the R^11 borderline via a second-order Jacobi-field dichotomy; the main proof seems solid, with the residual risk concentrated in an external Jacobi-field classification cited to Simon–Solomon.","tokens_in":76595,"tokens_out":2114,"would_cite":true,"duration_ms":22021,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["49Q20","49Q15","53A10","58E12"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that tiny perturbations of boundaries or metrics make area-minimizing hypersurfaces smooth in ambient dimension 11, and shrink singular sets in all higher dimensions.","keywords":["generic regularity","Plateau problem","area-minimizing hypersurfaces","singular set","Jacobi fields","minimizing hypercones","Almgren frequency","dimension 11"],"falsifier":"A direct falsifier would be a $C^\\infty$-small perturbation problem in $\\mathbb{R}^{11}$ for which every perturbed boundary still admits a minimizing current with a singular point; that would contradict Theorem 1.1. A more structural test is to compute the Jacobi-field spectrum of a candidate minimizing quadratic hypercone in $\\mathbb{R}^8$: if $\\mathrm{Jac}_0$ or $\\mathrm{Jac}_1$ contains anything beyond translations or rotations, the strong-integrability premise behind the slow-decay dimension reduction is violated, and the borderline argument would not apply to that cone.","tokens_in":75443,"feed_emoji":"📐","tokens_out":10423,"duration_ms":111852,"temperature":0.7,"pith_summary":"For Plateau's problem, the paper proves that a smooth closed oriented submanifold boundary $\\Gamma \\subset \\mathbb{R}^{n+1}$ can be perturbed $C^\\infty$-slightly to $\\Gamma'$ so that every area-minimizing integral current with boundary $\\Gamma'$ is a smooth embedded hypersurface when $n+1 \\leq 11$, and has singular set of dimension at most $n-10-\\epsilon_n$ when $n+1 \\geq 12$. The same conclusion holds for minimizers in a fixed integral homology class on a closed manifold, after a $C^\\infty$-small perturbation of the Riemannian metric. This extends generic regularity from $\\mathbb{R}^{10}$ to $\\mathbb{R}^{11}$, and it improves the existing singular-set bound in all higher dimensions. The reason the last open dimension yields is a refined analysis of the borderline tangent cones $C^\\circ \\times \\mathbb{R}^k$, with $C^\\circ$ a quadratic hypercone in $\\mathbb{R}^8$: one blow-up level deeper, the minimizer is modeled by Jacobi fields whose decay is either strictly faster than linear or exactly linear, and each case improves one of the two competing estimates. A sympathetic reader would take the paper's central claim to be that area-minimizing hypersurfaces are generically smooth up through ambient dimension $11$ in both the boundary and homology formulations.","feed_headline":"Area minimizers in R^11 become generically smooth","feed_subtitle":"A tiny nudge of the boundary or metric makes singular sets vanish through dimension 11.","key_machinery":"The machinery is the analysis of Jacobi fields on cylindrical hypercones $C = C^\\circ \\times \\mathbb{R}^k$ whose factor $C^\\circ$ is regular, strictly stable, strictly minimizing, and strongly integrable; all minimizing quadratic hypercones satisfy these properties. The load-bearing object is the decay order $G_C(u;\\varrho)$, a discrete Almgren-type frequency with a monotonicity property, applied to the Jacobi field $u$ obtained by Simon non-concentration estimates when a minimizer is well-approximated by $C$. The key gap in the homogeneity spectrum, either degree $1$ or degree $\\geq 1+\\Delta^{>1}_C$, drives the fast/slow decay dichotomy: fast decay improves the Hölder separation of leaves, slow decay uses Corollary 3.34 to find a proper effective spine subspace $V_u \\subsetneq \\mathrm{spine}\\, C$ on which nearby singular points concentrate. A covering tree structure (Proposition 6.1) packages the coarse dimension and coarse Hölder data into the Hausdorff dimension bounds for $\\mathcal{T}(\\mathrm{sing}\\,\\mathcal{F})$ and for the level sets $\\mathcal{T}^{-1}(t)$. The whole construction works uniformly for a foliation of pairwise-disjoint minimizers with a Lipschitz time function $\\mathcal{T}$, which is what lets Theorem 1.4 feed the Plateau and homology theorems.","core_discovery":"The central discovery is a dichotomy for the singular behavior that blocks generic regularity in ambient dimension $11$. Near a point modeled by a cylindrical hypercone $C = C^\\circ \\times \\mathbb{R}^k$ with $C^\\circ$ a quadratic hypercone, the paper classifies the decay of the Jacobi field $u$ that describes the minimizer as a graph over $C$: either the decay order is $> 1$ (fast decay), in which case the tangent cone is unique and the separation between disjoint leaves improves by $\\Delta^{>1}_C$; or the decay order equals $1$ (slow decay), in which case strong integrability of $C^\\circ$ produces a proper linear subspace $V_u$ of the spine along which $u$ is translation-invariant. This proper 'effective spine' drops the coarse dimension of the singular set by one, so the previously borderline ratio $\\dim \\mathrm{spine}\\, C/(1+\\alpha(C)) = 3/3 = 1$ for $C^\\circ \\times \\mathbb{R}^3$ becomes an effective ratio below $1$. Feeding both improvements into a covering-tree argument yields Theorem 1.4, whose dimension formulas imply $d^{\\mathrm{img}}_n < 1$ exactly when $n+1 \\leq 11$; Theorems 1.1 and 1.2 then follow by arranging Plateau boundaries or Riemannian metrics into a foliation of minimizers with disjoint supports. The upshot is that generic perturbations of the boundary or metric make minimizing hypersurfaces smooth in ambient dimension $11$, and cut the singular-set dimension to $n-10-\\epsilon_n$ in dimensions $n+1 \\geq 12$.","pith_inferences":["A testable extension recommended by the paper's own Remark 1.8: before attempting generic regularity in $\\mathbb{R}^{12}$, one should determine whether $\\mathbb{R}^8$ contains minimizing hypercones that are not strictly stable, strictly minimizing, or strongly integrable; any such cone would enlarge the borderline family beyond the quadratic cylinders this argument controls.","The fast/slow decay dichotomy suggests a general recipe for other borderline regularity problems: when the standard spine-dimension over Hölder-exponent ratio equals $1$, one extra blow-up level at the Jacobi-field scale can split the obstruction into a separation-improving case and a spine-dimension-reducing case, a transfer one might attempt in obstacle-type or mean-convex flow settings.","In the slow-decay case the effective spine is determined by the zero set of a homogeneous polynomial $p_3(r,y)$ (e.g., $y_1^3 - r^2 y_1$ for $k=3$), so one expects the singular set to concentrate on a codimension-one subspace of the spine; this refined stratification is implicit in Section 1.5 and could be made explicit in concrete examples such as foliations over Simons cones."],"forward_implications":["In $\\mathbb{R}^{11}$, the set of boundaries whose every minimizing current is smooth is both dense and open, hence Baire generic; singular minimizers become avoidable by arbitrary small boundary perturbations up through ambient dimension $11$.","In every ambient dimension $n+1 \\geq 12$, a generically chosen minimizer has singular set of Hausdorff dimension at most $n-10-\\epsilon_n$, improving the earlier $n-9-\\epsilon'_n$ bound by one full dimension.","In the closed-manifold homology setting, generic metrics force minimizers in any nonzero integral homology class to be smooth embedded hypersurfaces for $n+1 \\leq 11$, with the same improved singular bound above; this extends Schoen–Yau's positive scalar curvature obstruction up to dimension $11$ and, via Lohkamp's reduction, carries the positive mass theorem to those dimensions.","Theorem 1.4's formulas make the threshold exact: $d^{\\mathrm{img}}_n < 1$ and $d^{\\mathrm{dom}}_n < 0$ are equivalent to $n+1 \\leq 11$, so the improvement comes precisely from the second-order Jacobi-field analysis rather than from a slack in the dimension count."],"supporting_citations":[{"why":"Previous foliation-plus-separation strategy that established generic regularity in dimensions 9 and 10 and supplies the framework this paper refines.","marker":"[CMS23a]"},{"why":"Earlier improved singular-set bound n-9-epsilon'_n that Theorem 1.1 strengthens to n-10-epsilon_n.","marker":"[CMS24]"},{"why":"Cylindrical tangent cone and non-concentration estimates used to model minimizers one blow-up level deeper as graphs over cylinders.","marker":"[Sim93]"},{"why":"Provides the uniqueness and non-concentration estimates and the strong-integrability setting for quadratic cones.","marker":"[Sim94]"},{"why":"Harnack-type decay for positive Jacobi fields on cones, used to control separation of disjoint minimizers.","marker":"[Sim08]"},{"why":"Liouville theorem for positive Jacobi fields on cylindrical cones that identifies the positive field as a multiple of r^{-2} and prevents cancellation of slow decay.","marker":"[ES24]"},{"why":"Spectral analysis showing quadratic hypercones are strongly integrable, the premise behind the effective spine.","marker":"[SS86]"},{"why":"Hardt-Simon generic boundary regularity in R^8 and the foliation by minimizers with isolated singularities that the current argument generalizes.","marker":"[HS85]"},{"why":"Establishes that alpha(C) is at least alpha_n with equality only for quadratic hypercones, the spectral input fixing the borderline family.","marker":"[Zhu18]"},{"why":"Supplies the positive gap Delta^{non-qd}_n for non-quadratic cones and supporting Jacobi-field decay estimates used in Section 5 and Appendix C.","marker":"[Wan24]"}],"fun_headline_variants":["Generic regularity for area minimizers proven in dimension 11","In dimension 11, almost all minimizing hypersurfaces are smooth","Small perturbations of boundary or metric smooth area minimizers in dim 11","Generic regularity for minimizing hypersurfaces at dimension 11"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The slow-decay half of the proof assumes that every minimizing quadratic hypercone $C^\\circ$ is strongly integrable: its only homogeneous Jacobi fields of degree $0$ and $1$ are translations and rotations; if some minimizing quadratic cone in $\\mathbb{R}^8$ had an exotic degree-zero or degree-one Jacobi field, the effective-spine reduction from $k$ to $k-1$ in Section 7.7 and the $\\mathbb{R}^{11}$ conclusion would fail.","fun_headline_variants_meta":{"raw":{"variants":["Generic regularity for area minimizers proven in dimension 11","In dimension 11, almost all minimizing hypersurfaces are smooth","Small perturbations of boundary or metric smooth area minimizers in dim 11","Generic regularity for minimizing hypersurfaces at dimension 11"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000948,"raw_usage":{"total_tokens":4056,"prompt_tokens":966,"completion_tokens":3090,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":582,"completion_tokens_details":{"reasoning_tokens":3018}},"tokens_in":582,"tokens_out":3090,"duration_ms":25668,"temperature":1.0,"reasoning_tokens":3018,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T00:37:06.961452+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct falsifier would be a $C^\\infty$-small perturbation problem in $\\mathbb{R}^{11}$ for which every perturbed boundary still admits a minimizing current with a singular point; that would contradict Theorem 1.1. A more structural test is to compute the Jacobi-field spectrum of a candidate minimizing quadratic hypercone in $\\mathbb{R}^8$: if $\\mathrm{Jac}_0$ or $\\mathrm{Jac}_1$ contains anything beyond translations or rotations, the strong-integrability premise behind the slow-decay dimension reduction is violated, and the borderline argument would not apply to that cone.","supporting_citations":[],"review_version":1}