{"id":"440f522f-b045-4eda-8726-a772e2dd4894","arxiv_id":"2608.03949","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For dual quermassintegrals, the Brunn-Minkowski inequality fails for q>n (and q>n+2 for origin-symmetric bodies), holds at q=n+2 for origin-symmetric bodies, and holds for 0<q≤n+1 for unconditional bodies.","lead":"This mathematics paper determines when a Brunn-Minkowski type inequality holds for dual quermassintegrals, a family of weighted volume functionals of convex bodies, once the exponent q exceeds the dimension n. It shows the inequality fails above sharp thresholds, proves it holds at the endpoint q=n+2 for centrally symmetric bodies, and confirms it for unconditional bodies on a broad range.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 6.3's singular-branch exclusion is not justified as written for 0<q<1: the claim 'exponent ≤ -(q-1) < -1' is false there, and this step underlies the singular Reilly formula behind Theorem 1.6.","rationale":"The paper's headline Theorem 1.6 rests on a long analytic chain, and the most fragile link is the singular weighted Reilly formula (Proposition 6.5). The reader's weakest-assumption identification of Lemma 6.4 and Lemma 6.3 is correct, but the specific defect I find is in the singular-branch exclusion inside Lemma 6.3: the displayed inequality is algebraically false for 0<q<1, which is part of the theorem's range. This is not an isolated typo, because the exclusion is needed for the derivative bounds (6.7) and hence for the vanishing of the inner boundary terms in Lemma 6.4. I do not believe the theorem is false: the indicial equation supplies a cheap lower bound β_m>2-q that repairs the proof, so the concern is a gap rather than a counterexample. The endpoint and negative results in Sections 3-5 do not depend on this step, and the independent Hadwiger argument for q=n+2 is sound. Since the reader's verdict is already CONDITIONAL for separate issues and this additional gap is local and likely fixable, I do not move the verdict; a conditional acceptance with a required verification of Lemma 6.3 is appropriate.","tokens_in":21930,"tokens_out":32749,"duration_ms":297331,"concrete_test":"Recompute the indicial exclusion in Lemma 6.3 for 0<q<1. Solve (6.9) for m=1, use the monotonicity of the positive root, and verify β_m > 2-q (or at least β_m > 1-q/2) for every m≥1. Then check that the energy exponent satisfies 2β_m^-+q-3 < -1, so the integral ∫_0^ε r^{2β_m^-+q-3} dr diverges. If the bound holds, add the missing line to Lemma 6.3 and re-run the B_ε(u)→0 estimate in Lemma 6.4 with the corrected exponent; if the bound fails for some q∈(0,1), the expansion and Proposition 6.5 need a different argument.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Proposition 6.5 is the engine of Theorem 1.6. Its proof imports Lemma 6.3 to get the expansion, the derivative bounds (6.7), and the integrability assertions used in Lemma 6.4 and Proposition 6.5. In Lemma 6.3, the singular branch r^{β_m^-}Y_m is excluded by asserting that the weighted H1-energy exponent 2β_m^-+q-3 = -(q-1)-2β_m satisfies '≤ -(q-1) < -1 since q>0'. For 0<q<1 this chain is false: -(q-1)=1-q>0, so the displayed inequality does not establish divergence of the energy integral. The range 0<q<1 is included in Theorem 1.6, and without the exclusion the asymptotic expansion (6.7) and the vanishing of the inner-sphere boundary terms in Lemma 6.4 would be unjustified. The gap appears repairable: from the indicial equation (6.9), for m≥1 and q<2 one has β_m>2-q, which gives 2β_m^-+q-3 < q-3 < -1; for q≥2, β_m>0 gives the same divergence. But that bound is not stated or proved in the manuscript, so the singular weighted Reilly formula is not fully established as written for part of the theorem's range.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper investigates the Brunn-Minkowski inequality for q-th dual quermassintegrals, focusing on the range q>n posed by Sadovsky and Zhang. It establishes negative results: the inequality fails for arbitrary convex bodies when q>n, and fails for origin-symmetric convex bodies when q>n+2, via a second-variation argument and a dimension reduction to planar rectangles. It also proves the endpoint case q=n+2 for origin-symmetric bodies using Hadwiger's inequality for the polar moment of inertia, and for unconditional convex bodies it establishes the full range 0<q≤n+1 using a singular weighted Reilly formula and coordinate-slice Hardy inequalities. Applications to uniqueness of dual curvature measures are derived as consequences.","tokens_in":22174,"tokens_out":13369,"duration_ms":112238,"significance":"If the main results are correct, the paper makes a substantial contribution to the dual Brunn-Minkowski theory. Theorem 1.6 extends the known range for unconditional bodies up to q≤n+1 and gives an equality case that answers a question of Sadovsky and Zhang in that symmetry class. The counterexamples sharpen previous negative results, and the endpoint q=n+2 via Hadwiger's inequality is an elegant and nontrivial application. The paper is also careful in deriving many auxiliary estimates from external benchmarks such as the Kolesnikov-Milman Reilly formula and A2 weighted Poincaré theory, and in the parts I checked the technical estimates are coherent. However, two technical issues, one in the central proof of the unconditional theorem and one in a secondary theorem, need to be addressed before the results are fully established as written.","major_comments":[{"comment":"The exclusion of the singular branch r^{β_m^-}Y_m is not justified as written for 0<q<1. The displayed chain '2β_m^-+q-3 = -(q-1)-2β_m ≤ -(q-1)<-1 since q>0' is false in that range, because -(q-1)=1-q>0. This step is load-bearing: it is used to obtain the expansion (6.7), the integrability (6.8), the vanishing of inner-sphere boundary terms in Lemma 6.4, and hence the singular Reilly formula (6.13) in Proposition 6.5, which underlies Theorem 1.6. The gap appears repairable: from the indicial equation (6.9), for m≥1 and q<2 one has β_m>2-q, so 2β_m^-+q-3 = 1-q-2β_m < q-3 < -1, while for q≥2, β_m>0 gives the same divergence. Since neither bound is stated or proved in the manuscript, the proof of Lemma 6.3 must be corrected before the full range 0<q≤n+1 in Theorem 1.6 is established.","section":"Section 6.1, Lemma 6.3"},{"comment":"The identity '1/2·R_t +_p 1/2·R_-t = r_{p,t}R_0' used at the start of the proof is not correct for p<1 (and for p=0). For n=2, along u=(1,0) the support function of the L_p combination equals r_{p,t}, while along u=(1,1)/√2 it equals √2 independently of t. Since h_{R_0}(u) equals 1 on the first direction and √2 on the diagonal, equality would force r_{p,t}=1 for every t, which holds only for p=1. Consequently the formula for F''_{p,Q}(0), the threshold q_p^{(n)}, and the claimed L_p Brunn-Minkowski failure do not follow from the given argument. This does not affect Theorems 1.3-1.6, but Proposition 4.3 is a stated theorem and needs either a corrected proof or a corrected statement.","section":"Section 4.3, Proposition 4.3"}],"minor_comments":[{"comment":"The monotonicity of P(Q) is asserted from the integrand being increasing in Q, but P is a quotient of two increasing integrals; a direct proof of monotonicity would improve the presentation if Proposition 4.3 is retained.","section":"Section 4.3"},{"comment":"In the convergence statement, the phrase 'together with all its derivatives' should be qualified: the local uniform convergence is on compact subsets of B_{R_0}\\{0\\}, and near the origin the derivatives are not uniformly controlled for negative α.","section":"Section 6.1, Lemma 6.3"},{"comment":"The assertion that the ratios T_E^+(K_τ)/T_E^-(K_τ) and T_E^+(L_τ)/T_E^-(L_τ) vary continuously from 0 to ∞ and from ∞ to 0 is plausible but not proved; a short justification would make the intermediate value step fully transparent.","section":"Section 5, Step 2 of Theorem 5.1"}],"recommendation":"major_revision","confidential_remarks":"The central Theorem 1.6 is attractive and the proof strategy is credible, but the gap in Lemma 6.3, although repairable, is in the main line of argument, and the identity behind Proposition 4.3 appears genuinely false. The authors should be asked to repair the Lemma 6.3 estimate and to either correct or remove the L_p Brunn-Minkowski part. I do not see grounds for rejection if these issues are fixed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The main theorems here answer Sadovsky-Zhang's question in a satisfying way: failure for q>n in general, failure for symmetric bodies at q>n+2, a positive endpoint at q=n+2 via Hadwiger's moment-of-inertia inequality, and the full range 0<q≤n+1 for unconditional bodies. Theorems 1.4-1.6 are new and, as far as I can tell, correct. The dimension-reduction trick that lowers n to a planar rectangle is elegant, and the singular weighted Reilly formula with |x|^alpha is a serious technical piece. The Hardy inequalities with the offset parameter, and the equality cases, deserve credit. The applications to dual curvature measure uniqueness are a natural payoff.\n\nI agree with the reader's conditional verdict, and I want to flag the soft spots in the same proportions.\n\nFirst, Proposition 4.3 is not established as written. The identity 1/2·R_t +_p 1/2·R_{-t} = r_{p,t} R_0 is simply false for 0<p<1 and for p=0. The L_p combination of two rectangles is not a rectangle except at p=1; the support function in diagonal directions does not match r_{p,t}|u_1+u_2|. So the L_p threshold theorem as stated does not have a valid proof. This is a side result, though: Theorems 1.4-1.6 only use the p=1 case, where the rectangle identity holds. The paper should either correct the L_p statement or remove it.\n\nSecond, the claimed uniformity in the dimension reduction (Section 4.2, \"together with the first and second derivatives in t\") is asserted without proof. It is probably true, but it is load-bearing for Theorem 1.4 and needs a clean justification.\n\nThird, and most relevant to the core, Lemma 6.3 has a wrong inequality. The exclusion of the singular branch claims exponent ≤ -(q-1) < -1 since q>0; this is false for 0<q<1, and also not < -1 for 1<q<2. The gap is repairable: from the indicial equation, β_m > (2-q)/2 for q<2 and β_m>0 for q≥2, which gives exponent < -1 in all cases. But the manuscript does not say this, so the singular Reilly formula is not fully proved as written. The stress-test note is on target.\n\nNone of these problems appears to damage Theorems 1.3-1.6 themselves. The core arguments are credible and the flaws are fixable rather than fatal. I would send this to a serious referee. It belongs in a strong geometry journal after the L^p section is repaired or cut, and Lemma 6.3's exponent bound is corrected. I would cite it if I worked in this area.","headline":"Sharp dual Brunn-Minkowski thresholds are likely right, but two proof gaps (a false L^p calculation and a wrong exponent estimate in the singular Reilly step) need fixing.","tokens_in":22811,"tokens_out":5809,"would_cite":true,"duration_ms":45644,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["52A40","52A20","26D15"],"pacs":[],"model":"deepseek-v4-flash","headline":"For unconditional convex bodies, the dual Brunn-Minkowski inequality for q-th dual quermassintegrals holds for every $0<q\\le n+1$, with equality for smooth strictly convex bodies only when the two bodies are dilates; outside the…","keywords":["dual quermassintegral","Brunn-Minkowski inequality","unconditional convex body","singular weighted Reilly formula","Hardy inequality","dual curvature measure","polar moment of inertia","origin-symmetric convex body"],"falsifier":"Take the planar unconditional rectangles $K=[-2,2]\\times[-1,1]$ and $L=[-1,1]\\times[-2,2]$ with $q=3=n+1$; the theorem asserts $eV_3(K+L)^{1/3} \\ge eV_3(K)^{1/3}+eV_3(L)^{1/3}$, so a direct numerical evaluation of these radial integrals giving the reverse inequality would refute Theorem 1.6.","tokens_in":21682,"feed_emoji":"📐","tokens_out":11781,"duration_ms":99325,"temperature":0.7,"pith_summary":"This paper asks whether the Brunn-Minkowski inequality for q-th dual quermassintegrals can extend beyond $q=n$, and it answers the question with a near-complete map of where the inequality lives. For arbitrary convex bodies the inequality fails for every $q>n$, and for origin-symmetric bodies it fails for every $q>n+2$; at the endpoint $q=n+2$ it holds for origin-symmetric bodies. The main positive result is that for unconditional convex bodies the inequality holds for every $0<q\\le n+1$, with equality for smooth strictly convex bodies exactly when the two bodies are dilates. The stakes are that these are the natural size functionals of the dual Brunn-Minkowski theory, so knowing their concavity range controls when radial size behaves like a norm under Minkowski addition.","feed_headline":"Dual Brunn-Minkowski confirmed up to q=n+1 for unconditional bodies","feed_subtitle":"A singular weighted integration-by-parts identity extends the inequality beyond q=n in a natural symmetry class.","key_machinery":"The carrying object is the singular weighted Reilly formula (Proposition 6.5), an integration-by-parts identity for the operator $L_\\alpha u=|x|^{-\\alpha}\\,\\mathrm{div}(|x|^\\alpha \\nabla u)$ with $\\alpha=q-n$, obtained by excising a small ball around the origin and letting its radius tend to zero. The inner-sphere boundary terms vanish because the Neumann solution has a controlled expansion near the singularity, and the Hessian remains square-integrable against the weight $|x|^\\alpha$. Two coordinate-slice Hardy inequalities with an offset parameter $c=(|x|^2-x_i^2)^{1/2}$, one for $0<\\alpha<3$ with constant $3-\\alpha$ and one for negative $\\alpha$ with constant $-\\alpha$, dominate the tangential-gradient terms, giving nonnegativity of the Hessian functional and strict log-concavity of the weighted volume along unconditional perturbations.","core_discovery":"The central claim is a sharp threshold picture for the dual Brunn-Minkowski inequality $eV_q(K+L)^{1/q} \\ge eV_q(K)^{1/q} + eV_q(L)^{1/q}$. When $q>n$ the inequality is shown to fail for arbitrary convex bodies by testing the local second variation at the Euclidean ball. In the origin-symmetric class it fails for all $q>n+2$ through a dimension-reduction construction that collapses an $n$-dimensional body to a product of a planar rectangle and a thin cube; at the endpoint $q=n+2$ the inequality holds for all origin-symmetric bodies via a classical inequality for the polar moment of inertia, with equality only for dilates. The main theorem establishes the inequality in the full range $0<q\\le n+1$ for unconditional convex bodies and characterizes equality for smooth strictly convex unconditional bodies as equality exactly when the bodies are dilates. This extends the known $0<q\\le n$ range and settles the equality question for the unconditional class.","pith_inferences":["The dimension-reduction counterexample for symmetric bodies makes it plausible that any positive range for general origin-symmetric bodies cannot exceed $q=n+2$; the open interval $n<q<n+2$ is exactly where a strengthening of the Hardy estimates would have to be found.","The excision-and-limit treatment of the singular weight is likely to work for other radially singular densities of the form $|x|^{q-n}$ with different symmetry classes, so the same machinery could test Brunn-Minkowski behavior for partial symmetries beyond unconditional bodies.","At $q=1$ the derived uniqueness statement must allow dilates because the dual curvature measure in question is dilation-invariant; this suggests that the dilation ambiguity at $q=1$ is a structural feature of $(1,q)$-dual curvature measures rather than a defect of the proof."],"forward_implications":["The dual Brunn-Minkowski inequality now holds for all $0<q\\le n+1$ in the unconditional class, extending the previously known $0<q\\le n$ range for symmetric bodies.","At the endpoint $q=n+2$ the inequality holds for all origin-symmetric bodies, so the only remaining open range for the symmetric problem is $n<q<n+2$.","The equality characterization for smooth strictly convex unconditional bodies settles the equality case of the earlier symmetric-range inequality within this class.","The Brunn-Minkowski inequalities imply corresponding Minkowski inequalities for dual curvature measures, yielding uniqueness of the $(1,q)$-th dual curvature measure for $0<q\\le n+1$ in the unconditional class and for $q=n+2$ in the origin-symmetric class.","The counterexample thresholds show that the range $q\\le n$ is sharp without symmetry and $q\\le n+2$ is the natural boundary with origin symmetry, since the inequality fails beyond those ranges."],"supporting_citations":[{"why":"Poses the $q>n$ problem and supplies the $0<q\\le n$ symmetric-case inequality whose equality case the paper answers in the unconditional class.","marker":"[23]"},{"why":"Provides the weighted second-variation formula and the smooth weighted Reilly identity that the singular Reilly argument adapts.","marker":"[17]"},{"why":"Supplies the polar-moment-of-inertia Brunn-Minkowski inequality used to prove the endpoint $q=n+2$.","marker":"[10]"},{"why":"Proves the unconditional endpoint $q=n+2$ and a partial range, the direct predecessor extended here to the full range $0<q\\le n+1$.","marker":"[12]"},{"why":"Establishes the low range $0<q\\le 1$ and the $L_p$ Brunn-Minkowski framework for dual quermassintegrals.","marker":"[26]"},{"why":"Gives earlier counterexamples for $q>2n+1$ and the local stability threshold that the sharper dimension-reduction counterexample improves.","marker":"[28]"},{"why":"Defines the dual curvature measures and their variational formulas, used in the uniqueness applications of Section 8.","marker":"[13]"},{"why":"Introduces the $(p,q)$-th dual curvature measure whose $(1,q)$ uniqueness is derived from the new Brunn-Minkowski inequalities.","marker":"[22]"}],"fun_headline_variants":["Unconditional bodies extend dual Brunn-Minkowski to q=n+1","Dual Brunn-Minkowski: unconditional bodies hold up to q=n+1","Sharp threshold: dual Brunn-Minkowski works for unconditional until n+1","Dual Brunn-Minkowski: fails generally, unconditional extended to n+1","Unconditional bodies save dual Brunn-Minkowski beyond q=n"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof depends on the singular weighted Reilly formula: after cutting a small ball around the origin out of the body and shrinking it, the boundary terms on that inner sphere must contribute nothing and the second derivatives of the auxiliary Neumann solution must stay integrable against $|x|^{q-n}$; if that failed, an uncontrolled contribution from the origin would break the strict log-concavity that produces the inequality.","fun_headline_variants_meta":{"raw":{"variants":["Unconditional bodies extend dual Brunn-Minkowski to q=n+1","Dual Brunn-Minkowski: unconditional bodies hold up to q=n+1","Sharp threshold: dual Brunn-Minkowski works for unconditional until n+1","Dual Brunn-Minkowski: fails generally, unconditional extended to n+1","Unconditional bodies save dual Brunn-Minkowski beyond q=n"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001605,"raw_usage":{"total_tokens":6379,"prompt_tokens":915,"completion_tokens":5464,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":531,"completion_tokens_details":{"reasoning_tokens":5360}},"tokens_in":531,"tokens_out":5464,"duration_ms":32132,"temperature":1.0,"reasoning_tokens":5360,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T14:48:42.796648+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the planar unconditional rectangles $K=[-2,2]\\times[-1,1]$ and $L=[-1,1]\\times[-2,2]$ with $q=3=n+1$; the theorem asserts $eV_3(K+L)^{1/3} \\ge eV_3(K)^{1/3}+eV_3(L)^{1/3}$, so a direct numerical evaluation of these radial integrals giving the reverse inequality would refute Theorem 1.6.","supporting_citations":[{"cited_title":"Math.480(2025), Paper No","cited_arxiv_id":null,"evidence_quote":"Poses the $q>n$ problem and supplies the $0<q\\le n$ symmetric-case inequality whose equality case the paper answers in the unconditional class."},{"cited_title":"Kolesnikov and Emanuel Milman,Poincar´ e and Brunn-Minkowski inequalities on the boundary of weighted Riemannian manifolds, Amer","cited_arxiv_id":null,"evidence_quote":"Provides the weighted second-variation formula and the smooth weighted Reilly identity that the singular Reilly argument adapts."},{"cited_title":"Hadwiger,Konkave Eik¨ orperfunktionale und h¨ ohere Tr¨ agheitsmomente, Comment","cited_arxiv_id":null,"evidence_quote":"Supplies the polar-moment-of-inertia Brunn-Minkowski inequality used to prove the endpoint $q=n+2$."},{"cited_title":"Weighted centro-affine Poincar\\'e inequalities","cited_arxiv_id":"2606.04774","evidence_quote":"Proves the unconditional endpoint $q=n+2$ and a partial range, the direct predecessor extended here to the full range $0<q\\le n+1$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the low range $0<q\\le 1$ and the $L_p$ Brunn-Minkowski framework for dual quermassintegrals."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives earlier counterexamples for $q>2n+1$ and the local stability threshold that the sharper dimension-reduction counterexample improves."},{"cited_title":"2, 325-388","cited_arxiv_id":null,"evidence_quote":"Defines the dual curvature measures and their variational formulas, used in the uniqueness applications of Section 8."},{"cited_title":"Math.329(2018), 85-132","cited_arxiv_id":null,"evidence_quote":"Introduces the $(p,q)$-th dual curvature measure whose $(1,q)$ uniqueness is derived from the new Brunn-Minkowski inequalities."}],"review_version":2}