{"id":"acf97673-5df5-47da-bdd8-98be2863efaf","arxiv_id":"2605.19818","paper_version":2,"verdict":"REJECT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"Compact non-flat constant-curvature manifolds and irreducible non-positive locally symmetric spaces are claimed to admit no conformal product structures, but the stated main theorem is not supported as written.","lead":"This paper claims that compact non-flat constant-curvature manifolds admit no conformal product structures, and that irreducible non-positively curved compact locally symmetric spaces also admit none. The main theorem as stated conflicts with the paper's own arXiv abstract, which says flat surfaces are exceptional.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.1 is false as stated: an explicit non-closed conformal product structure exists on a flat torus, contradicting the claim that flat g forces D=∇g.","rationale":"The single most load-bearing concern is not merely a proof gap but the falsehood of the central theorem. The paper's own abstract concedes 'flat surfaces are exceptional', and an explicit flat-torus example satisfies the rank-1 characterization (Lemma 2.3) with non-closed Lee form, directly contradicting Theorem 1.1. The reader's weakest_assumption (noncompactness of the leaf L_x) correctly identifies the place in the proof where the flat-surface exception escapes, but it understates the damage: the theorem is not just unproved, it is false without a dimension restriction. I agree with the rejection verdict, but I would strengthen the basis: the counterexample is explicit and checkable, so the overclaim is decisive. The reader's rationale did mention the abstract contradiction, so there is partial agreement, though the weakest_assumption field focused on a proof-technical issue rather than the concrete counterexample. The concrete test of verifying the flat-torus example settles the matter directly.","tokens_in":11998,"tokens_out":15263,"duration_ms":145753,"concrete_test":"Verify the flat torus construction: on T^2 with g=dx^2+dy^2, set phi=sin x, xi=cos(sin x) d_x + sin(sin x) d_y, theta=cos x dy. Check equation (2) for X=d_x and X=d_y, and compute dtheta=-sin x dx^dy != 0. If the verification holds, Theorem 1.1 is false as stated; if it fails, the counterexample is invalid and the concern would need revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's own arXiv abstract states 'flat surfaces are exceptional', immediately conflicting with Theorem 1.1, which permits no conformal product structure on flat manifolds except D=∇g. This is not merely a wording issue: an explicit counterexample exists on the flat torus. Let g=dx^2+dy^2 on T^2, choose phi=sin x, xi=cos(sin x) d_x + sin(sin x) d_y, and theta=cos x dy. Since the connection is flat, a direct computation gives grad_X xi = -theta(xi)X + theta^sharp <X,xi> for all X, so by Lemma 2.3 this is a rank-1 conformal product structure. However dtheta = -sin x dx^dy != 0, so theta is not closed and the Weyl connection D is not the Levi-Civita connection of g. This contradicts Theorem 1.1's 'unless g is flat and D=grad g'. The failure is localized in the rank-1 proof (Proposition 3.2 and Case ii, r=1): the leaf xi^perp through a maximum point of a=theta(xi) is not compact, so the asserted interior maximum of ||theta_0||^2 does not exist and equations (14)-(21) do not apply. The theorem requires at least a dimension restriction n>=3, and the rank-1 argument needs a genuine compactness replacement.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies conformal product structures—non-closed reducible Weyl connections—on compact Riemannian manifolds. Theorem 1.1 asserts that a compact manifold of constant sectional curvature admits no conformal product structure unless the metric is flat and the Weyl connection is the Levi-Civita connection. Theorem 3.5 claims a similar nonexistence for compact irreducible locally symmetric spaces of non-positive curvature. The proofs are split into rank-1 and rank≥2 cases: the rank-1 case uses a maximum principle on leaves of the distribution ξ^⊥, while the higher-rank case uses an involution operator S borrowed from [14] together with integral identities obtained by taking traces of curvature expressions.","tokens_in":12352,"tokens_out":15349,"duration_ms":146530,"significance":"If the theorems were correct as stated, they would give strong obstructions to conformal product structures on basic compact spaces, complementing recent results on Kähler and Einstein manifolds. The paper's use of the involution formalism and integral identities is natural and potentially powerful. However, Theorem 1.1 is false as stated: an explicit conformal product structure exists on a flat 2-torus with non-closed Lee form. The abstract actually concedes this by calling flat surfaces 'exceptional,' so the theorem needs a dimension restriction. In addition, the rank-1 proof relies on an unjustified leaf-compactness assertion, and the symmetric-space proof asserts a key identity without derivation. These are load-bearing gaps, though they appear fixable.","major_comments":[{"comment":"Theorem 1.1 is false as stated. On the flat torus T^2 with g=dx^2+dy^2, set f=sin x, ξ=cos f ∂_x + sin f ∂_y, θ=cos x dy. A direct computation gives ∇_X ξ = -θ(ξ)X + θ^♯⟨X,ξ⟩ for all X, so by Lemma 2.3 this is a rank-1 conformal product structure. But dθ = -sin x dx∧dy ≠ 0, so D ≠ ∇g. This contradicts the conclusion 'unless g is flat and D=∇g'. The abstract's statement that flat surfaces are exceptional is inconsistent with the theorem; the theorem must be restricted to dim M ≥ 3 or must state the n=2 exception explicitly.","section":"Theorem 1.1 / Abstract"},{"comment":"The proof asserts that the leaf L_x of ξ^⊥ through a maximum point x of a is compact, so that ||θ0||² attains a maximum in its interior. Compactness of M does not imply compactness of leaves; dense leaves on flat tori provide a direct counterexample to this line of reasoning. No argument for leaf compactness is given. This interior maximum is used to deduce equations (19)-(21) and the contradiction κ+a²=0; without it the rank-1 argument collapses. The rank-1 flat case in Theorem 1.1 inherits this gap. A replacement argument is needed (e.g., using the ODE ξ(a)=a² after θ0=0 when n≥3).","section":"Proposition 3.2"},{"comment":"The key identity H = (r-1)r||θ1||² + (n-r)(n-r-1)||θ2||² is asserted without proof. It is the link between the geometric bound H≤0 and the non-negative θ-terms in equation (37), and the subsequent contradiction depends on it. This identity should follow from equation (6) and the curvature tensor of the locally symmetric metric, but the computation is not given. Without a derivation, the conclusion θ=0 in the symmetric-space setting is unsupported.","section":"Theorem 3.5"}],"minor_comments":[{"comment":"The lemma is stated as an 'if and only if', but the proof only establishes the forward direction (existence of a conformal product structure implies equation (2)). The converse—that (2) yields a D-parallel splitting—is used implicitly and should be proved or explicitly cited.","section":"Lemma 2.3"},{"comment":"The full-text abstract omits the sentence 'flat surfaces are exceptional' that appears in the arXiv abstract. The two versions should be made consistent, especially since this exception is essential for the correct statement of Theorem 1.1.","section":"Abstract"},{"comment":"The projection π: TM → ξ^⊥ is used without definition. Please define it explicitly before first use.","section":"Equation (14)"},{"comment":"There is a typo in the running header: 'CUR V ATURE' should be 'CURVATURE'.","section":"Title header"},{"comment":"The sentence 'Since the Euclidean type is excluded' should be justified briefly: a flat compact locally symmetric space is reducible (a torus), so irreducibility rules it out.","section":"Theorem 3.5"}],"recommendation":"major_revision","confidential_remarks":"The manuscript has merit, but it is not publishable in its current form. The main theorem is demonstrably false without a dimension restriction, and the rank-1 proof has an unproven compactness assertion. However, the abstract already acknowledges the flat-surface exception, so the error in Theorem 1.1 appears to be a statement-level oversight that can be corrected. The symmetric-space proof also needs a substantial missing computation. I recommend major revision: the author should restate Theorem 1.1 with the correct dimension hypothesis, supply a valid compactness argument or a substitute for the rank-1 maximum principle, and provide the missing derivation of the H-identity in Theorem 3.5."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First the bottom line: Theorem 1.1 is overclaimed. There is a genuine non-trivial conformal product structure on the flat torus: take θ = cos x dy and ξ = cos(sin x)∂x + sin(sin x)∂y. It satisfies the rank-1 equation in Lemma 2.3, the line bundle spanned by ξ is D-parallel, yet dθ ≠ 0, so D is not the Levi-Civita connection. The paper's own abstract (in the metadata) admits 'flat surfaces are exceptional,' but the theorem and the κ ≤ 0 proof both claim flat implies D = ∇g. At minimum, the theorem needs a dimension restriction n ≥ 3.\n\nWhat is actually new and good: the rank-1 equation in the Lee-form coefficient a, and the S-involution machinery adapted from [14], are the right tools for this problem. The non-flat constant curvature obstruction and the extension to irreducible non-positive symmetric spaces are natural targets, and the integral identity (34) is a neat trick. The paper follows the existing literature closely and is honestly written.\n\nSoft spots, in proportion: the rank-1 proof (Prop 3.2) assumes the leaf L_x through a maximum of a is compact, so that ||θ0||² attains an interior maximum. That is unproven and load-bearing; dense leaves occur even on flat tori, so this is not a minor technicality. Second, the symmetric-space proof asserts the identity H = (r−1)r||θ1||² + (n−r)(n−r−1)||θ2||² without derivation. It is presumably the same computation as in the κ ≤ 0 case, but it needs to be written out. There are also minor typos (e.g., the garbled 'S￿').\n\nFor whom: researchers working on Weyl connections and conformal product structures. The methods could be reused, but the current version is not citable as is. I would send it to peer review with a clear request for major revision rather than desk reject it; the defects are identifiable and likely fixable, and the non-flat results may well survive intact.","headline":"The main theorem is false as stated—flat tori give a counterexample—but the underlying methods are promising and the paper deserves a chance at major revision.","tokens_in":12767,"tokens_out":9730,"would_cite":false,"duration_ms":94254,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["53C18","53C35","53C20"],"pacs":[],"model":"deepseek-v4-flash","headline":"A compact constant-curvature manifold admits a conformal product structure only in the trivial flat case: flat metric and Weyl connection equal to Levi-Civita.","keywords":["conformal product structure","Weyl connection","Lee form","constant sectional curvature","locally symmetric space","holonomy reduction","compact manifold","space form"],"falsifier":"Construct a compact flat Riemannian torus of dimension at least three that carries a conformal product structure whose Lee form is not closed. The theorem forbids it, and the rank-1 proof's compact-leaf assumption fails exactly in this setting, so such an example (or a rigorous obstruction to its existence) would settle the claim.","tokens_in":11918,"feed_emoji":"📐","tokens_out":4636,"duration_ms":48763,"temperature":0.7,"pith_summary":"The paper proves that a compact Riemannian manifold with constant sectional curvature carries a conformal product structure — a reducible Weyl connection — only when the metric is flat, the dimension is at least three, and the Weyl connection coincides with the Levi-Civita connection. In particular, every compact non-flat space form is free of such structures, and flat surfaces are the sole exception. The argument splits into rank-1 and higher-rank cases. For rank 1, the function a=θ(ξ) is constant on the leaves of the foliation ξ⊥; extremality of a and a curvature comparison force a contradiction on positive curvature. For non-positive curvature, an integral identity forces θ=0 or the flat case. The same machinery, with a one-parameter family of curvature computations on the involution S, extends to compact irreducible locally symmetric spaces of non-positive curvature.","feed_headline":"Non-flat compact space forms forbid conformal products","feed_subtitle":"Flat metrics in dimensions three and up only allow the trivial Weyl connection; flat surfaces are the exception.","key_machinery":"The machinery is the orthogonal involution S that encodes the splitting TM=T1⊕T2 (eigenvalues +1 and −1), together with the Lee form θ of the Weyl connection. From the conformal Koszul formula the paper derives two independent expressions for the curvature action R_{X,Y}S: one algebraic expression (Equation 8) and one geometric expression from constant sectional curvature or from the symmetric-space curvature formula. Equating these and splitting into symmetric/skew and S-commuting/anti-commuting parts yields algebraic constraints that force dθ=0 or an integral identity. For rank 1, the key object is the vector field ξ spanning T1 and the function a=θ(ξ), whose gradient and leaf-constancy pr","core_discovery":"The central claim is Theorem 1.1: on a compact constant-sectional-curvature manifold (M,g), the only possible conformal product structure is the trivial one — g flat and D the Levi-Civita connection — provided dim M ≥ 3. Since a non-closed reducible Weyl connection defines a conformal product structure, the result eliminates all non-trivial reducible Weyl connections on such manifolds. The proof brings the Lee form θ and the involution S (with the two D-parallel distributions as eigenspaces) into a curvature identity, then contrasts two evaluations of the same curvature expression. In the non-positive curvature case the identity integrates to a non-negative expression that must vanish, forci","pith_inferences":["The leaf-compactness assumption in the rank-1 proof suggests that genuinely flat tori with irrational foliations could host counterexamples in dimension two, where compact leaves fail; the paper's flat-surface exception is consistent with this failure.","A direct testable extension: construct a flat product torus with a non-closed 1-form and check whether the corresponding Weyl connection has reducible holonomy; the theorem predicts no such structure in dimensions at least three, so this is a concrete place to probe the boundary of the result.","The integral identity (34) has a form that may generalize to any manifold with parallel curvature tensor or even to non-locally-symmetric non-positive curvature, suggesting a broader obstruction than constant sectional curvature.","The irreducibility argument for symmetric spaces suggests that any compact quotient of a non-compact irreducible symmetric space is likewise devoid of conformal product structures, and the result could extend to reducible symmetric spaces via de Rham factors."],"forward_implications":["On compact spherical space forms (constant positive curvature), every Weyl connection is either non-reducible or the Levi-Civita connection, so no local product splitting with a non-closed Lee form can exist.","On flat tori of dimension at least three, any conformal product structure must be trivial, meaning the only reducible Weyl connection is the Levi-Civita connection of the flat metric.","The same obstruction applies to compact irreducible locally symmetric spaces of non-positive curvature, for example compact quotients of hyperbolic space.","The results constrain the possible holonomy reductions of Weyl connections on these manifolds, complementing the classification of holonomies of torsion-free affine connections.","Passing to finite covers or universal covers does not create conformal product structures where none exist on the compact base."],"fun_headline_variants":["Conformal products vanish on compact curved space forms","Curved compact manifolds reject conformal splitting","Only trivial conformal products on flat manifolds in dim ≥3","Flat surfaces alone allow nontrivial conformal products","Compact constant-curvature manifolds: conformal products die"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The rank-1 proof assumes that the leaf L_x through a point where a=θ(ξ) is extremal is compact, so that the squared norm of θ0 attains a maximum inside the leaf; leaf compactness is not guaranteed by compactness of M, and dense leaves on flat tori already violate it.","fun_headline_variants_meta":{"raw":{"variants":["Conformal products vanish on compact curved space forms","Curved compact manifolds reject conformal splitting","Only trivial conformal products on flat manifolds in dim ≥3","Flat surfaces alone allow nontrivial conformal products","Compact constant-curvature manifolds: conformal products die"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000175,"raw_usage":{"total_tokens":1046,"prompt_tokens":588,"completion_tokens":458,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":332,"completion_tokens_details":{"reasoning_tokens":379}},"tokens_in":332,"tokens_out":458,"duration_ms":5313,"temperature":1.0,"reasoning_tokens":379,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T05:04:36.547612+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a compact flat Riemannian torus of dimension at least three that carries a conformal product structure whose Lee form is not closed. The theorem forbids it, and the rank-1 proof's compact-leaf assumption fails exactly in this setting, so such an example (or a rigorous obstruction to its existence) would settle the claim.","supporting_citations":[],"review_version":2}