{"id":"ecd0eab6-42fc-4fde-804c-47d910b003e3","arxiv_id":"2611.17039","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Under nonnegative weighted sectional curvature with density controls, an Abresch-Gromoll-type upper excess estimate e(x) <= C_A (h^{A+1}/s)^{1/A} is proved, together with a weighted law of cosines and lower excess bounds.","lead":"On curved spaces carrying a density weight, this paper proves comparison inequalities including a weighted law of cosines and an upper bound on the excess of geodesic triangles under a weighted curvature condition. The results extend classical comparison geometry to manifolds with density, complementing an earlier weighted excess estimate due to Wei and Wylie, with applications to rigidity and topology questions.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek/deepseek-v4-flash","headline":"Lemma 5.4 quotes the key tangential Hessian bound Hess r_y - phi_{r_y}g <= e^{-2phi_{r_y}}/r_y g from the author's unpublished preprint [11]; Theorem 5.1 is unsupported if that bound is absent or needs extra hypotheses.","rationale":"The reader's weakest_assumption identifies exactly the point that my pass converged on, so I agree. The internal algebra of Propositions 2.2, 2.4, 2.6, Theorem 3.1, Proposition 3.3, the lower excess estimates, and the barrier argument in Theorem 5.1 checks out; conditional on (5.2), the ODE barrier, support-function smoothing, and t* optimization are coherent. I found no additional hidden flaw. The zero-density recovery requiring sec >= 0 rather than Ric >= 0 is explicitly stated by the authors [b52] and is a limitation, not a defect. Since the reader already returned CONDITIONAL and my concern is the same one, I do not adjust the verdict; the condition is that [11] must be available or the bound must be proved in this paper.","tokens_in":23135,"tokens_out":14414,"duration_ms":96123,"concrete_test":"Obtain arXiv:2605.24407 and inspect Theorem 3.1 and its proof: do they establish, for an arbitrary basepoint y under exactly sec_phi >= 0, a0 <= phi <= 0, |grad phi| <= c0/r_p, the tangential bound Hess r_y - phi_{r_y} g <= e^{-2 phi_{r_y}}/r_y g (or an equivalent bound from which it follows)? If yes, the dependency in Lemma 5.4 is verified and Theorem 5.1 is supported. If the statement or proof differs, a self-contained derivation, or a citation to a published version, must be added before Theorem 5.1 should be accepted.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 5.1 depends on Lemma 5.4, whose proof begins by invoking the tangential bound Hess r_y - phi_{r_y} g <= e^{-2 phi_{r_y}}/r_y g and referring to the proof of [11, Theorem 3.1]. Reference [11] is arXiv:2605.24407, a same-author preprint with no published version cited, so this is not a routine quotation of a published result. The bound is needed for distance functions from arbitrary basepoints y: the barrier later applies it to r_p, r_q, r_x and to the shifted support distances r_{p_epsilon}, r_{q_epsilon}, r_{x_epsilon}. The trace step from the bound to (5.2) is elementary and uses only a0 <= phi <= 0 and |grad phi| <= c0/r_p; therefore the entire force of sec_phi >= 0 is carried by the quoted bound. If [11] does not prove that bound under exactly these hypotheses, or proves it only for the basepoint p of the density control, Lemma 5.4, the weighted Laplacian comparison, the barrier argument, and Theorem 5.1 all lack support. I checked the surrounding algebra and found no internal inconsistency; the paper's Remark 3.5 [b165] also reroutes the reader to [11] for this same bound, so the dependency is acknowledged but not discharged.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek/deepseek-v4-flash","summary":"The paper studies the modified Hessian MHess u = Hess u - <∇u, ∇φ> g on Riemannian manifolds with density, with measure dμ = e^{-(n+1)φ} dvol. Sections 2-4 develop structural identities for radial functions: a radial-tangential decomposition, spectral and trace identities, a comparison identity relating the modified Hessians of two radial functions, and a characterization of when those Hessians are proportional. These are then used to prove a hinge comparison theorem under the bound MHess u ≤ K(r_p)g along a minimizing geodesic, yielding a weighted law of cosines and lower excess estimates. Section 5 states the main new application, Theorem 5.1: under sec_φ ≥ 0, a_0 ≤ φ ≤ 0, and |∇φ| ≤ c_0/r_p, the excess of a point x relative to a minimizing geodesic from p to q satisfies e(x) ≤ C_A (h^{A+1}/s)^{1/A} whenever h ≤ s/2, with A = (n-1)e^{-2a_0} + 2c_0 and C_A = (2A/(A-1))(2A/(A+1))^{1/A}. In the zero-density case the estimate reduces to an Abresch-Gromoll-type bound with exponent 1/(n-1), but under nonnegative sectional rather than nonnegative Ricci curvature.","tokens_in":23497,"tokens_out":9563,"duration_ms":60483,"significance":"If the key modified Hessian input is available under the stated hypotheses, the paper is a coherent and useful contribution. The algebraic core is sound: the radial-tangential decomposition, the trace identity, Proposition 2.5, the hinge comparison in Theorem 3.1, the weighted law of cosines in Proposition 3.3, the lower excess algebra in Section 4, and the barrier ODE optimization in Theorem 5.1 all check out. The final constant C_A is derived explicitly in terms of a_0, c_0, and n, with no free parameters or fitted constants. The main theorem gives a concrete, falsifiable estimate and recovers the classical form in the limit φ ≡ 0. The paper is also honest in Remark 3.5 and in the proof of Lemma 5.4 that the crucial tangential Hessian bound comes from the author's earlier preprint. The significance of the paper is therefore real but conditional on that unpublished input being valid under exactly the hypotheses used here.","major_comments":[{"comment":"Lemma 5.4 is the load-bearing input for Theorem 5.1, and its proof invokes the tangential bound Hess r_y - φ_{r_y} g ≤ e^{-2φ_{r_y}}/r_y g by reference to [11, Theorem 3.1]. Reference [11] is an unpublished same-author preprint (arXiv:2605.24407), so this is not a routine quotation of a published theorem. Moreover, the bound is needed for distance functions from arbitrary basepoints y, including the shifted basepoints p_ε, q_ε, x_ε in the Calabi support argument, while the density control |∇φ| ≤ c_0/r_p is tied to the fixed point p. If [11, Theorem 3.1] is proved only for the basepoint p of the density control, or under different hypotheses, then Lemma 5.4, the weighted Laplacian comparison, and Theorem 5.1 are unsupported. The dependency is acknowledged in Remark 3.5 and in the proof of Lemma 5.4, but it is not discharged. The author should either prove the tangential bound in this paper under the stated hypotheses or supply a published version containing it.","section":"Lemma 5.4 and proof of Theorem 5.1","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Even granting [11, Theorem 3.1], the trace step in the proof of Lemma 5.4 is not literally consistent as written. If the displayed inequality Hess r_y - φ_{r_y} g ≤ e^{-2φ_{r_y}}/r_y g is interpreted as a full tensor inequality and traced over all n directions, it would give a factor n, not n-1, in both the exponential term and the φ_{r_y} term. The conclusion (5.2) requires the bound on the tangential part with g_{r_y} in place of g, traced over the (n-1)-dimensional tangential subspace. The notation should be corrected so that the trace computation genuinely yields (5.2).","section":"Lemma 5.4, trace step","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The display for Δ_μ e has missing denominators: it should read (α+2β)/r_p + α/r_q, not α+2β r_p + α r_q.","section":"Proof of Theorem 5.1, after Lemma 5.4","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The sentence 'the tangential part of the modified Hessian satisfies Hess r_y - φ_{r_y} g ≤ ...' should specify the tangential metric g_{r_y} on the right-hand side and indicate that the trace is taken over the tangential directions; otherwise the displayed algebra cannot be followed.","section":"Lemma 5.4","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The 'Warped Law of Cosines' in Remark 3.5 is stated informally with an undefined constant C_0 and an unspecified mixed geodesic comparison; since it is not used in the paper, it should be clearly labeled as a heuristic or deferred to future work.","section":"Remark 3.5","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Reference [11] and reference [16] are both listed as preprints; if any of them has appeared in the meantime, the citation should be updated. For [11], which carries the main technical input, a published version would also resolve the major concern above.","section":"References","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"major_revision"},"author_rebuttal":null,"desk_editor":{"model":"deepseek/deepseek-v4-flash","letter":"Main thing to know: Theorem 5.1 is the genuinely new piece. It gives an Abresch-Gromoll-type upper excess estimate under nonnegative weighted sectional curvature with density controls, complementing Wei-Wylie's Bakry-Emery Ricci version. The comparison with Kennard-Wylie-Yeroshkin is fair: they treat positive weighted sectional curvature rigidity, not this estimate. The structural work in Sections 2-3 is mostly elementary but clean: radial-tangential decomposition, the comparison identity, the hinge comparison, and the weighted law of cosines all check out. The lower excess estimates in Section 4 follow from the law of cosines as claimed. The constants are derived from the stated hypotheses, not fitted, and the paper is honest that the zero-density recovery needs sec >= 0 rather than Ric >= 0.\n\nThe soft spot is exactly where the stress-test points. Lemma 5.4, and therefore the barrier argument for Theorem 5.1, begins with the tangential bound Hess r_y - phi_{r_y} g <= e^{-2 phi_{r_y}}/r_y g, quoted from [11, Theorem 3.1], the author's own unpublished arXiv preprint. This is not a routine quotation of a published theorem. The bound is applied to distance functions from arbitrary basepoints y, including the shifted support points, so the question is whether [11] proves it under precisely these hypotheses and for arbitrary centers. The rest of Lemma 5.4 from that bound is elementary, and I found no internal inconsistency in the excess algebra, the ODE barrier, or the t* minimization. The dependency is acknowledged, but it is not discharged. The paper would be fully checkable if the author either supplies a self-contained proof of that tangential bound or points to a published version of [11] with exactly these hypotheses.\n\nMinor notes: the Calabi-support handling of cut points is standard and adequate, and the singularity at r_p = 0 is consciously avoided in the barrier region. I do not see a separate circularity problem; the reliance on [11] is sequential, not circular, but it is load-bearing. I would not cite Theorem 5.1 in my own work until the [11] dependency is resolved. If the bound in [11] holds as stated, the paper is a solid contribution to comparison geometry for manifolds with density; if not, Theorem 5.1 collapses.\n\nThis deserves a serious referee, not a desk reject. Send it out with the explicit instruction to verify [11, Theorem 3.1] for arbitrary basepoints under sec_phi >= 0, a0 <= phi <= 0, |grad phi| <= c0/r_p. Conditional acceptance is the right posture until that check is done.","headline":"New weighted-sectional upper excess bound, but its key Hessian input is quoted from the author's unpublished preprint [11].","tokens_in":24075,"tokens_out":1560,"duration_ms":17750},"pith_extraction":{"msc":["53C20","53C21","53C22"],"pacs":[],"model":"deepseek/deepseek-v4-flash","headline":"Hinge comparison and excess estimates via modified Hessian bounds on manifolds with density","keywords":["manifolds with density","weighted sectional curvature","modified Hessian","hinge comparison","excess estimates","Abresch-Gromoll inequality","weighted Laplacian comparison"],"falsifier":"Find a complete n-dimensional manifold with density (n≥3) satisfying sec_φ ≥ 0, a_0 ≤ φ ≤ 0, |∇φ| ≤ c_0/r_p, but for which the upper excess estimate (5.1) fails for some triple of points p,q,x with h ≤ s/2. Alternatively, demonstrate that the tangential bound from [11, Theorem 3.1] does not hold under these hypotheses.","tokens_in":24528,"feed_emoji":"📐","tokens_out":985,"duration_ms":182504,"temperature":0.7,"pith_summary":"This paper tries to establish comparison geometry results for Riemannian manifolds equipped with a density function, using a modified Hessian tensor. It derives structural identities for the modified Hessian of radial functions, proves a hinge comparison theorem yielding a weighted law of cosines and lower excess bounds, and under nonnegative weighted sectional curvature obtains an Abresch-Gromoll-type upper excess estimate. These results extend classical comparison theorems to the weighted setting, with potential implications for understanding the geometry and topology of manifolds with density.","feed_headline":"Hinge comparison and excess estimates via modified Hessian bounds on manifolds with…","feed_subtitle":"Extends classical comparison geometry to weighted manifolds using a modified Hessian tensor, yielding weighted law of cosines and…","key_machinery":"The key machinery is the modified Hessian tensor MHess u = Hess u - ⟨∇u,∇φ⟩g, its radial-tangential decomposition, and the comparison identity relating modified Hessians of two radial functions. The hinge comparison theorem uses integration of the second derivative of u along a geodesic, while the upper excess estimate employs a weighted Laplacian comparison (Lemma 5.4) and a barrier argument with a radial function satisfying an ODE.","core_discovery":"The central discovery is a hinge comparison theorem (Theorem 3.1) that provides an integral inequality along a geodesic hinge under an upper bound on the modified Hessian of a radial function. This yields a weighted law of cosines and lower bounds on the excess function. Additionally, under nonnegative weighted sectional curvature and density controls, an upper excess estimate (Theorem 5.1) of Abresch-Gromoll type is obtained, with exponent depending on dimension and density bounds.","pith_inferences":["The modified Hessian framework may be applied to prove Toponogov-type triangle comparison theorems in the weighted setting, as suggested by the hinge comparison and mixed geodesic triangle considerations.","The upper excess estimate could be used to study the topology of manifolds with nonnegative weighted sectional curvature, analogous to applications of the classical Abresch-Gromoll inequality in Ricci curvature.","The radial-tangential decomposition and comparison identity may facilitate Hessian and Laplacian comparison for other radial potentials, beyond the squared distance function."],"forward_implications":["Provides a weighted law of cosines and lower excess estimates under modified Hessian bounds, generalizing classical Euclidean and Riemannian results.","Yields an Abresch-Gromoll-type upper excess estimate under nonnegative weighted sectional curvature, with explicit dependence on dimension and density bounds.","The structural identities for the modified Hessian (radial-tangential decomposition, spectral representation, comparison identity) offer tools for further comparison geometry in manifolds with density.","In the zero-density case, the upper excess estimate recovers the classical Abresch-Gromoll inequality (with improved constant for n=3) but under the stronger assumption of nonnegative sectional curvature rather than Ricci curvature."],"supporting_citations":[],"fun_headline_variants":[],"cache_read_input_tokens":0,"weakest_assumption_plain":"The upper excess estimate (Theorem 5.1) depends on the tangential bound Hess r_y - φ_{r_y} g ≤ $e^{{-2φ_{r_y}}$}/r_y g, quoted from the author's unpublished preprint [11, Theorem 3.1]; if that bound does not hold under the stated hypotheses (sec_φ ≥ 0, a_0 ≤ φ ≤ 0, |∇φ| ≤ c_0/r_p), Theorem 5.1 is unsupported.","fun_headline_variants_meta":{"raw":{"variants":[]},"model":"deepseek/deepseek-v4-flash","effort":"low","cost_usd":5.7e-05,"raw_usage":{"cost":5.65292e-05,"is_byok":false,"cost_details":{"upstream_inference_cost":5.65292e-05,"upstream_inference_prompt_cost":5.61652e-05,"upstream_inference_completions_cost":3.64e-07},"total_tokens":824,"prompt_tokens":822,"completion_tokens":2,"prompt_tokens_details":{"audio_tokens":0,"video_tokens":0,"cached_tokens":256,"cache_write_tokens":0},"completion_tokens_details":{"audio_tokens":0,"image_tokens":0,"reasoning_tokens":0}},"tokens_in":822,"tokens_out":2,"cost_source":"openrouter_usage","duration_ms":1770,"image_pages":[],"temperature":1.0,"billed_attempts":1,"reasoning_tokens":0,"transport_provider":"openrouter","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-11-13T07:25:46.888933+00:00","model_set":{"reader":"deepseek/deepseek-v4-flash"},"pipeline_version":"pith-pipeline@v1.4.1-daily-deepseek-or-medium","input_kind":null,"input_chars":null,"input_total_chars":null,"passes_missing":null,"published_standing":null,"published_confidence":null,"objection_audit":null,"run_record":null,"basis":{"kind":"not_recorded","source":null,"chars_given":null,"chars_total":null},"receipt":null,"falsifier":"Find a complete n-dimensional manifold with density (n≥3) satisfying sec_φ ≥ 0, a_0 ≤ φ ≤ 0, |∇φ| ≤ c_0/r_p, but for which the upper excess estimate (5.1) fails for some triple of points p,q,x with h ≤ s/2. Alternatively, demonstrate that the tangential bound from [11, Theorem 3.1] does not hold under these hypotheses.","supporting_citations":[],"lane_accuracy":{"status":"not yet measured","lane":{"model":"deepseek/deepseek-v4-flash","input_kind":null},"url":"https://pith.science/accuracy"},"review_version":1}