{"id":"e0df3fa3-7978-4f9f-9c02-4926d33245c9","arxiv_id":"2502.03389","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A proper CAT(0) space whose 2-sphere filling inequality has constant below 1/(6*sqrt(pi)) satisfies isoperimetric inequalities with exponent 1+delta for every delta>0, equivalent to asymptotic rank at most 2.","lead":"What happens when you try to fill a sphere-shaped surface with a 3D ball in a space with non-positive curvature? The paper proves a sharp threshold: below a precise Euclidean constant, the filling volume is almost linear in surface area, a new higher-dimensional gap theorem.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 4.1's proof appears to require, but not supply, extensions of all three sides of the geodesic triangle to complete geodesics; one half-line per vertex cannot do this.","rationale":"The reader's weakest_assumption is Lemma 4.1, and my stress test converges on the same lemma, with a specific defect in its proof. The central theorem's novel step is Theorem D, whose proof relies on Lemma 4.1 to obtain controlled density and Nagata dimension for minimal tetrahedra. If the density bound fails, the covering argument and the linear filling inequality fail, so the implication (3)⇒(2) in Theorem A lacks support. The specific issue is concrete: the proof claims one half-line per vertex extends all three sides of a geodesic triangle to complete geodesics, which is geometrically impossible for nondegenerate triangles. This is not a disagreement with the external result [Sta21, Prop. 80]; the cited literature may well contain a correct proof, and the paper may be repairable by quoting it more faithfully. But as written, the proof of the key lemma is incomplete. I therefore recommend a conditional acceptance: the central argument is plausible and no contradiction or circularity was found, but acceptance should require the authors to either correct the extension construction in Lemma 4.1 or replace the sketch by an explicit reference to the relevant [Sta21] statement. This does not reject the paper, and I agree with the reader's identification of Lemma 4.1 as the load-bearing premise.","tokens_in":29052,"tokens_out":43999,"duration_ms":435915,"concrete_test":"Inspect [Sta21, Lemmata 78–80] and verify whether the construction there attaches two half-lines per vertex (or otherwise extends each side to a complete geodesic) before deriving the density bound for geodesic triangles. If the cited construction differs from the paper's description, repair Lemma 4.1's proof accordingly. Independently test the bound in a concrete CAT(0) example, such as a flat cone of angle 3π with a Euclidean geodesic triangle: compute the minimal triangle's area density in balls centered at the cone point and check whether it is actually ≤ (3π/2)r^2 for all r.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The load-bearing premise is Lemma 4.1: for every minimal triangle u, area(u(B^2) ∩ B(x,r)) ≤ (3π/2)r^2. This bound drives Lemma 4.2 (Nagata dimension at most 2) and Theorem 4.3 (linear filling for minimal tetrahedra), hence Theorem A. The proof of Lemma 4.1 states: 'We first glue a half-line to every vertex of △ to obtain a CAT(0) space X′. Note that in X′ every side s of △ is contained in a complete geodesic c_s that intersects X precisely in s.' A single half-line attached at a vertex of a CAT(0) space cannot serve as the extension of two distinct sides incident to that vertex: the two extensions would have to be collinear with the same ray, which forces the two sides to be collinear. To extend each of the three sides to a complete geodesic, one must attach two half-lines at each vertex (one for each incident side), or otherwise adjoin antipodal directions in the space of directions. As written, the construction of the CAT(0) space X̂ and the minimal plane Ẑ, and hence the area-growth limit lim_{r→∞} area(f(Ẑ)∩B(x,r))/(πr^2) = 3/2, are not justified. The cited [Sta21, Lemmata 78 and 79] may supply the correct argument, but the paper's recollection is incomplete at exactly the point the reader identified as weakest. If Lemma 4.1 lacks a valid proof, the Nagata-dimension bound and the linear inequality for minimal tetrahedra are unsupported, and the proof of Theorem A collapses.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a sharp isoperimetric gap theorem in dimension two for proper CAT(0) spaces. Theorem A states that a Euclidean filling inequality for Lipschitz 2-spheres with constant c < c_3 = 1/(6√π) for large areas is equivalent to a δ-isoperimetric inequality for every δ > 0, and to asymptotic rank at most 2. The proof introduces minimal tetrahedra and shows they satisfy a linear filling inequality (Theorem D), using a density bound for minimal triangles (Lemma 4.1), Nagata-dimension bounds, and a controlled-density filling theorem. The paper also proves a non-sharp Euclidean isoperimetric inequality for 2-spheres (Theorem C) and derives a genus version for closed surfaces (Corollary B).","tokens_in":29325,"tokens_out":28527,"duration_ms":264969,"significance":"If correct, this is the first higher-dimensional analogue of the sharp rank-one gap theorem and establishes near-linear fillings in rank-two CAT(0) spaces. The sharp constant c_3, the new minimal-tetrahedron technology, and the explicit use of Nagata dimension are strong contributions. The proof is not circular: the main theorem is derived from external results and independent lemmas, and no parameter is fitted to the desired inequality. However, the proof of the key density bound, Lemma 4.1, contains a gap that must be repaired; this is load-bearing for the central claim.","major_comments":[{"comment":"The proof of the density bound is not valid as written. The construction of X′ glues \"a half-line to every vertex\" of the geodesic triangle. In a nondegenerate triangle, the two sides incident at a vertex have different tangent directions, so a single half-line at that vertex can extend at most one of the two sides to a complete geodesic. Hence the assertion that every side s of △ is contained in a complete geodesic c_s intersecting X precisely in s does not follow from the described gluing. Since the subsequent area-growth computation lim_{r→∞} area(f(Ẑ)∩B(x,r))/(πr^2) = 3/2, and hence the inequality area(u(B^2)∩B(x,r)) ≤ (3π/2)r^2, depend on this extension, Lemma 4.1 is not established by the text. This lemma is load-bearing: it is used in Lemma 4.2 for the Nagata-dimension bound and in Theorem 4.3 for the linear filling inequality for minimal tetrahedra, which in turn is used in the proof of Theorem A (Theorem 7.3). The authors should either correct the construction (for example, attach two half-lines per vertex with the appropriate directions, or glue a Euclidean sector) or replace this sketch by a precise statement and proof reference for the degenerate-triangle case from [Sta21, Lemmata 78 and 79].","section":"Section 4, Lemma 4.1 (proof)"}],"minor_comments":[{"comment":"There are several typographical slips, for example \"isoperimetric inequalites\" (Section 1.1), \"the aymptotic rank\" (Section 1.2), and \"area(f )δ2\" (Theorem 7.3) where the exponent δ^2 is intended.","section":"Throughout"},{"comment":"After choosing s_1 ∈ (s/2, s), the text says \"by the coarea inequality, we can choose s such that length_φ(Π̂) ≤ 2 area(φ)/s\"; the choice of s was fixed in the statement, so this should presumably refer to s_1.","section":"Section 6, Proposition 6.5 proof"},{"comment":"The symbol a is used both as the increment in the desired inequality and as λx in the proof, which makes the argument harder to follow; renaming one of the two variables would improve clarity.","section":"Section 6, Lemma 6.1 proof"},{"comment":"The notation Z and Z′ for the maximal set and its subset clashes with the CAT(0) disc retract Z appearing in Lemma 4.1; using different letters would avoid confusion.","section":"Section 4, Lemma 4.2"}],"recommendation":"major_revision","confidential_remarks":"The main theorem appears significant and the overall strategy is coherent. The gap in Lemma 4.1 is local but load-bearing; I believe it can be repaired by a more explicit argument or a precise citation to [Sta21]. I recommend major revision rather than rejection. The reliance on prior work of one of the authors is legitimate given the subject, and I saw no circularity."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nYou should know this paper is the real thing: it proves the first higher-dimensional sharp isoperimetric gap theorem for CAT(0) spaces, matching the rank-one case. The equivalence between sub-Euclidean constants below c3=1/(6√π), δ-isoperimetric inequalities for all δ>0, and asymptotic rank at most 2 is new, and the minimal tetrahedra used to get the linear filling inequality are a genuine new tool. The paper also records a Euclidean filling inequality for 2-spheres (Theorem C) that apparently was missing from the literature. The overall strategy is careful, and the main estimates are explicit.\n\nThe reader's take is about right: no circularity, the self-citations to [Sta21] and [GL23] are used as building blocks, not as placeholders for the theorem. The proof is not formalized, and several steps are compressed, but the architecture is sound.\n\nThe soft spot is exactly where the reader put it: Lemma 4.1, the density bound for minimal triangles, is the load-bearing premise. As written, the proof says gluing one half-line to each vertex of the triangle extends all three sides to complete geodesics. That looks wrong on its face: a single half-line at a vertex cannot be the extension of two non-collinear sides. The authors explicitly defer the details to [Sta21, Lemmata 78 and 79], so this may be an exposition problem rather than a mathematical error. But a referee needs to check that the cited lemmas actually supply the needed construction, because if Lemma 4.1 fails, the Nagata-dimension bound and the linear inequality for tetrahedra collapse, and with them Theorem A.\n\nIf the cited lemmas check out, this is a strong paper that deserves publication in a top journal. It is aimed at people working in metric geometry, isoperimetric inequalities, and quantitative topology. I would bring it to our reading group and cite it. I recommend sending it to serious peer review rather than desk rejection.\n\nBest,\n[Your name]","headline":"First higher-dimensional sharp isoperimetric gap theorem; the proof hinges on one density lemma whose sketch is suspect but likely repairable via cited lemmas.","tokens_in":29925,"tokens_out":3172,"would_cite":true,"duration_ms":31158,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["53C23","49Q05","53C45"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that in a proper CAT(0) space, a Euclidean filling inequality for 2-spheres with constant below 1/(6√π) forces near-linear fillings and asymptotic rank at most two.","keywords":["CAT(0) spaces","isoperimetric inequalities","filling volume","asymptotic rank","minimal tetrahedra","minimal surfaces","Lipschitz spheres","gap theorem"],"falsifier":"A concrete test is to check whether any proper CAT(0) space of asymptotic rank 3, such as Euclidean $R^{3}$, admits a sequence of Lipschitz 2-spheres with large area that can be filled with volume at most c·area^(3/2) for some c < 1/(6√π); if such fillings existed, the equivalence (1)⇔(3) in Theorem A would be false. A more direct falsifier would be a minimal triangle in a CAT(0) space whose intersection with some ball has area exceeding (3π/2)$r^{2}$, which would violate Lemma 4.1 and break the linear inequality for minimal tetrahedra.","tokens_in":28803,"feed_emoji":"🔺","tokens_out":5523,"duration_ms":51627,"temperature":0.7,"pith_summary":"The paper establishes a sharp isoperimetric gap theorem in non-positive curvature, the first higher-dimensional analogue of the classical rank-one result. For a proper CAT(0) space, it shows that a Euclidean filling inequality for Lipschitz 2-spheres with any constant strictly below the sharp Euclidean threshold c3 = 1/(6√π) is equivalent to the existence of δ-isoperimetric inequalities with exponent 1+δ for every δ>0, and both are equivalent to the asymptotic rank being at most 2. A sympathetic reader should care because this connects a quantitative filling condition to a coarse geometric invariant, and it provides the first evidence that the sharp rank-one gap phenomenon persists in higher dimensions. The proof introduces minimal tetrahedra, a new class of Lipschitz spheres built from four minimal triangles, and proves that they satisfy a linear filling inequality. Along the way the paper supplies a Euclidean isoperimetric inequality for null-homotopies of 2-spheres in CAT(0) spaces, which it notes was missing from the literature.","feed_headline":"Subthreshold fillings cap CAT(0) rank at two","feed_subtitle":"In proper CAT(0) spaces, a Euclidean filling constant below 1/(6√π) forces near-linear sphere fillings and asymptotic rank at most 2.","key_machinery":"The key machinery is the minimal tetrahedron: a Lipschitz 2-sphere composed of four minimal triangles, where a minimal triangle is an area-minimizing Lipschitz disc filling a geodesic triangle. The load-bearing property is a density bound for minimal triangles, area(u($B^{2}$) ∩ B(x,r)) ≤ (3π/2)$r^{2}$, which forces the image of a minimal tetrahedron to have Nagata dimension at most 2 with an absolute constant. This dimension control, combined with a filling theorem for bounded-density cycles with controlled filling radius, gives the linear isoperimetric inequality for minimal tetrahedra. The proof of Theorem A then decomposes an arbitrary large 2-sphere into a controlled number of discs with small boundary length and area, deforms it into a piecewise-minimal sphere made of minimal tetrahedra, and fills each tetrahedron linearly.","core_discovery":"The central discovery is Theorem A: for a proper CAT(0) space X, the following three conditions are equivalent: (1) there is a constant c < 1/(6√π) such that every Lipschitz 2-sphere of sufficiently large area can be filled by a Lipschitz 3-ball of volume at most c times area^(3/2); (2) for every δ > 0, every Lipschitz 2-sphere of any area can be filled with volume at most C(δ) times area^(1+δ); and (3) the asymptotic rank of X is at most 2. The proof that (3) implies (2) is the main challenge and rests on a new object, the minimal tetrahedron, a Lipschitz 2-sphere made of four minimal triangles spanning geodesic triangles. Theorem D states that in a proper CAT(0) space of asymptotic rank at most 2, every minimal tetrahedron satisfies a linear isoperimetric inequality, Fillvol(τ) ≤ μ·area(τ). This linear bound, combined with a quantitative triangulation of arbitrary spheres into controlled pieces, yields the near-linear filling inequality. The paper also proves the converse direction (1) ⇒ (3) in all dimensions with the sharp constant c_{n+1}, and derives Corollary B giving δ-isoperimetric inequalities for closed surfaces of any genus in rank-at-most-2 spaces.","pith_inferences":["One might expect the sharp threshold c3 = 1/(6√π) to govern homological fillings of general 2-cycles as well, not just Lipschitz sphere fillings, since the Euclidean plane and R^3 give the corresponding sharp constants.","The success of decomposing arbitrary spheres into minimal tetrahedra suggests that a similar strategy could be attempted for higher-dimensional spheres if canonical minimal building blocks with linear fillings can be constructed, although the paper notes the lack of nice subdivisions for n ≥ 2.","The paper's result could be tested computationally on explicit CAT(0) spaces such as products of trees or higher-rank symmetric spaces, where the asymptotic rank is known and filling bounds can be estimated numerically.","If the linear inequality for minimal tetrahedra could be strengthened to all fillings of 2-cycles, it would lead to genuine linear isoperimetric inequalities, resolving Gromov's conjecture in rank-2 CAT(0) spaces, but the authors state that a different strategy would be required."],"forward_implications":["If Theorem A is correct, every proper CAT(0) space with asymptotic rank at most 2 satisfies a δ-isoperimetric inequality for 2-spheres for every δ > 0, giving near-linear filling bounds in a wide class of non-positively curved spaces.","Corollary B extends the δ-isoperimetric inequality to all closed Lipschitz surfaces of any genus in rank-at-most-2 CAT(0) spaces, via Gromov's systolic inequality.","The implication (1) ⇒ (3) shows that a Euclidean filling inequality with a constant below the sharp Euclidean threshold rules out asymptotic rank larger than 2 in any dimension, generalizing the rank-one gap theorem.","The linear isoperimetric inequality for minimal tetrahedra provides a new finite unit that may be useful for studying filling problems in CAT(0) spaces of low rank.","The paper's Euclidean isoperimetric inequality for 2-spheres (Theorem C) fills a gap in the literature and holds for all CAT(0) spaces, independent of rank."],"supporting_citations":[{"why":"Supplies the density bound for minimal triangles that underlies Lemma 4.1 and the Nagata dimension bound.","marker":"[Sta21]"},{"why":"Provides fillings with controlled density and bounded filling radius for bounded-density cycles, used in Theorem 4.3.","marker":"[GL23]"},{"why":"Supplies the iterative filling-decomposition method adapted to prove the Euclidean isoperimetric inequality for 2-spheres in Theorem C.","marker":"[Wen08b]"},{"why":"Solves Plateau's problem in metric spaces and gives the regularity of minimal discs used in Theorem 2.2.","marker":"[L W17]"},{"why":"Extends the Plateau solution to singular curves, yielding a Lipschitz minimal disc with connected fibers.","marker":"[Cre22]"},{"why":"Provides the sharp Euclidean isoperimetric inequality for curves and the CAT(0) disc retract structure used in Lemma 2.3 and elsewhere.","marker":"[LW18b]"},{"why":"Shows that Assouad dimension bounds Nagata dimension, used in Lemma 4.2 to control covering multiplicity.","marker":"[LDR15]"},{"why":"Gives the isomorphism between integral current and Lipschitz chain homologies, used in Theorem 4.3 to fill by coning in a simplicial complex.","marker":"[RS09]"},{"why":"Contributes the idea of using minimizing maps with controlled density to obtain linear isoperimetric inequalities, acknowledged in the proof of Theorem D.","marker":"[Whi84]"},{"why":"Supplies Gromov's systolic inequality for closed surfaces, used in Corollary B to reduce genus.","marker":"[Gro83]"}],"fun_headline_variants":["Subthreshold volume gap forces rank ≤2 in CAT(0)","Below 1/(6√π): near-linear sphere fillings","Minimal tetrahedra unlock isoperimetric gap","Sharp filling threshold caps CAT(0) rank at two","Euclidean filling constant implies near-linear fills"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument depends on the density bound for minimal triangles, area(u($B^{2}$) ∩ B(x,r)) ≤ (3π/2)$r^{2}$, which comes from the structure theory of minimal surfaces; if that bound failed, the Nagata cover and the linear filling for minimal tetrahedra would not follow.","fun_headline_variants_meta":{"raw":{"variants":["Subthreshold volume gap forces rank ≤2 in CAT(0)","Below 1/(6√π): near-linear sphere fillings","Minimal tetrahedra unlock isoperimetric gap","Sharp filling threshold caps CAT(0) rank at two","Euclidean filling constant implies near-linear fills"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000266,"raw_usage":{"total_tokens":1661,"prompt_tokens":1045,"completion_tokens":616,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":661,"completion_tokens_details":{"reasoning_tokens":535}},"tokens_in":661,"tokens_out":616,"duration_ms":6704,"temperature":1.0,"reasoning_tokens":535,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T04:55:08.425047+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete test is to check whether any proper CAT(0) space of asymptotic rank 3, such as Euclidean $R^{3}$, admits a sequence of Lipschitz 2-spheres with large area that can be filled with volume at most c·area^(3/2) for some c < 1/(6√π); if such fillings existed, the equivalence (1)⇔(3) in Theorem A would be false. A more direct falsifier would be a minimal triangle in a CAT(0) space whose intersection with some ball has area exceeding (3π/2)$r^{2}$, which would violate Lemma 4.1 and break the linear inequality for minimal tetrahedra.","supporting_citations":[],"review_version":1}