{"id":"ac565c99-4a7b-4e7b-8b3d-368e24565865","arxiv_id":"2507.20082","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For arbitrary convex bodies K,L in R^n, the mixed area measure S_{K,...,K,L} is supported exactly on the closure of the (K,...,K,L)-extreme directions, which resolves Schneider's conjecture in R^3 and gives the upper bound in all dimensions.","lead":"This paper resolves a 1985 conjecture of Schneider on the support of mixed area measures: one inclusion is proved in every dimension, and the full characterization is proved in three dimensions, answering Minkowski's monotonicity problem there. It also derives a mixed Monge-Ampere version of the classical fact that zero-curvature surfaces are ruled.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the only delicate point is the Alexandrov-Fenchel equality condition in the lower-dimensional case, and the proof uses it in the non-degenerate regime.","rationale":"The reader identified the Alexandrov-Fenchel equality condition as the weakest assumption. I agree that this is the only deferred ingredient, but I do not believe it constitutes a live gap. In Lemma 4.4, after the case V(K',K,C)=0 is handled separately, the remaining case has V(K',K,C)>0. Since h_{K'}=h_K on the support of S_{K,C}, the equality V(K',K,C)=V(K,K,C) holds. The inequality chain together with the Alexandrov-Fenchel inequality forces V(K',K',C)=V(K',K,C), so the proportionality constant in Theorem 4.3 is exactly 1 and all diagonal mixed volumes are positive. Thus the equality condition is invoked precisely in the regime where the proportionality statement is non-vacuous. A potential counterexample involving two line segments in the plane has V(K,K)=V(L,L)=0 and V(K,L)>0, but then the Alexandrov-Fenchel inequality is strict, not an equality, so it does not test the equality condition. The rest of the proof, including the support-transfer through Lemma 2.8 and the induction step in Theorem 1.10, checks out. The upper-bound proof via Lipschitz covers and Hausdorff-measure estimates is also consistent. I therefore recommend keeping the reader's ACCEPT verdict unchanged, with the single concrete verification of the cited equality condition's hypotheses as a worthwhile check.","tokens_in":29361,"tokens_out":47176,"duration_ms":443441,"concrete_test":"Independently verify the exact hypotheses of Schneider [21, Theorem 7.4.2]: does the equality condition S_{K,C}=lambda S_{L,C} hold for all convex bodies, or only when both bodies have nonempty interior? Then re-run the proof of Lemma 4.4 with K replaced by K+epsilon B and pass epsilon to 0; if the equality condition requires full-dimensionality, the limiting argument must be supplied explicitly, and without it the conclusion u not in supp S_{L,C} would not follow.","verdict_should_be":"UNCHANGED","load_bearing_attack":"No significant objection identified. The central claims are supported by a coherent structure: Theorem 1.8 supplies the upper bound through a Lipschitz cover and tube-counting argument, and Theorem 1.10 is proved by a dimension induction that modifies K to \\hat K (Lemma 4.7) without changing the relevant mixed area measure, then transfers extremality via Lemma 1.20, Lemma 2.8, and the dim T(K,u)=1 case of Theorem 1.21. The most delicate step is the use of the Alexandrov-Fenchel equality condition (Theorem 4.3) in Lemmas 4.4 and 4.7, where one body can have empty interior. This is exactly the reader's flagged assumption. I do not find it load-bearing: in every invocation with V(K',K,C)>0, the chain of equalities forces V(K,K,C) and V(K',K',C) to be positive, so the proportionality formula is applied with well-defined positive constants. The classical 2D segment example that might falsify a naive reading of the equality condition has V(K,K)=V(L,L)=0 and does not satisfy the equality hypothesis of the theorem. The remaining analytical estimates, including the Lipschitz cover in Proposition 3.2 and the tube-counting bound in Theorem 3.1, are internally consistent, and the paper is explicit about the higher-dimensional obstruction in Theorem 1.21.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper addresses Schneider's conjectural description of the support of mixed area measures, equivalently the equality cases in Minkowski's monotonicity problem for mixed volumes. The two central results are Theorem 1.8, which proves that the mixed area measure vanishes on the set of directions that are not (C_1,...,C_{n-1})-exposed, and Theorem 1.10, which gives the full characterization supp S_{K[n-2],L} = cl{u : u is (K[n-2],L)-extreme} for arbitrary convex bodies K,L in R^n, thereby resolving Schneider's conjecture in dimension three. The upper bound is proved by a Lipschitz cover of k-singular boundary points together with a tube-counting estimate. The lower bound is proved by induction on dimension, using a body modification lemma (Lemma 4.7) that leaves the mixed area measure unchanged and transfers extremality through a projection step; the cases of Theorem 1.21 needed for the induction are proved in Section 4 and Section 6. The paper also transcribes the results to mixed Hessian measures, derives a mixed Hartman--Nirenberg--Pogorelov theorem, and identifies a genuinely higher-dimensional obstruction to extending the induction.","tokens_in":29566,"tokens_out":12472,"duration_ms":124261,"significance":"If the arguments are correct, this is a major advance on a forty-year-old conjecture in convex geometry: it proves one direction of Schneider's conjecture in full generality and settles the full conjecture in the original three-dimensional setting of Minkowski. The proof strategy is coherent and, in the parts I checked, internally consistent. The paper is notably honest about its limitations: Theorem 1.21(b) and the discussion in Section 6.2 explicitly exhibit the obstruction that prevents the projection-induction method from working in dimension at least four. A particular strength is that the main proofs are written in full, including the delicate measure-theoretic upper bound and the explicit construction in Appendix A.","major_comments":[],"minor_comments":[{"comment":"There is a typo: \"Thoughout this paper\" should be \"Throughout this paper\"; similarly, the caption of Figure 1.3 reads \"Illustration of of Corollary 1.18\" and should be corrected.","section":"Section 1.2, Example 1.2"},{"comment":"The display in the middle of the proof is compressed: the reader has to combine monotonicity with the Alexandrov--Fenchel inequality to see that V(K',K,C_1,...,C_{n-2})^2 equals the product V(K',K',C_1,...,C_{n-2})V(K,K,C_1,...,C_{n-2}). Adding one sentence making this equality explicit would improve readability.","section":"Section 4.2, Lemma 4.4"},{"comment":"The convergence \\hat K_t \\to \\hat K is asserted with a reference to [21, Lemma 7.5.2]; since \\hat K_t is defined by intersecting a continuum of half-spaces, it would be helpful to note explicitly that h_{K_t} decreases to h_K as t\\downarrow 0 and hence the compact convex sets \\hat K_t form a nested family whose intersection is \\hat K.","section":"Section 4.3, Lemma 4.7(a)"},{"comment":"The statement that \"any two touching cones of K are either equal or disjoint\" is used without proof; it follows directly from Lemma 2.2 and could be stated there or in the proof of Lemma 6.3.","section":"Section 6.1, Lemma 6.3"},{"comment":"The convexity of the functions f and g is asserted but not verified; a parenthetical remark that h is positive semidefinite with vanishing x_1-derivative at x_1=0 would remove any doubt.","section":"Section 6.3, Example 6.8"}],"recommendation":"accept","confidential_remarks":"The paper is a strong fit for the journal and the referee report is positive. The only point that might merit attention from the editor is that the Alexandrov--Fenchel equality condition is imported from Schneider's monograph without reproof; I checked the relevant invocations and they are legitimate, so this does not affect my recommendation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague, the headline: this paper settles the upper-bound half of Schneider's 1985 conjecture for all convex bodies in R^n and the whole conjecture in R^3, and it does so with arguments that are actually written down. That is not common in this area. I read the full text and the proofs are complete enough to follow; the main novelty — the Lipschitz cover of k-singular boundary points in Proposition 3.2 and the local body modification in Lemma 4.7 — is real and does the work. The paper is honest about the higher-dimensional obstruction (Theorem 1.21(b), Section 6.2), and the mixed Hessian / Hartman-Nirenberg corollaries are nice consequences rather than padding.\n\nWhere are the soft spots? The reader flagged Lemma 4.7's reliance on the Alexandrov–Fenchel equality condition (Theorem 4.3) for possibly degenerate bodies. The paper cites this from Schneider's book and does not reprove it. I checked the invocation: in every place the proportionality is used, positivity of the mixed volumes forces the factors positive, so the equality condition is applied in the non-degenerate regime. I do not think it is load-bearing. Worth a referee note, not a blocker. There are minor presentation gaps — Lemma 4.4's AF argument skips one monotonicity step — but they are easily filled. The Section 6 counterexample construction in the appendix is intricate and one has to trust the analytic geometry of Lemma A.1; but that is for the illustrative obstruction, not for the main theorems.\n\nThe citation pattern looks clean: previous verified cases are cited, Schneider's conjecture is never assumed, and self-citations are to the authors' prior Alexandrov–Fenchel work which is genuinely relevant. No invented entities, no free parameters.\n\nThis is a serious paper that deserves a serious referee. I would send it out. For a reader working on mixed volumes, Alexandrov–Fenchel equality, or Hessian measures, this is a must-cite and worth reading in full. My verdict: accept with minor revisions.","headline":"Resolves the upper bound of Schneider's conjecture in all dimensions and the full conjecture in R^3, with honest tools and clean exposition; the main theorems are as advertised.","tokens_in":30151,"tokens_out":2343,"would_cite":true,"duration_ms":19290,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["52A39","52A40","35J96"],"pacs":[],"model":"deepseek-v4-flash","headline":"The support of mixed area measures is characterized exactly in three dimensions, and in one direction in every dimension, which resolves a 1985 conjecture there.","keywords":["mixed volumes","mixed area measures","convex bodies","touching cones","extreme directions","mixed Hessian measures","Alexandrov–Fenchel inequality","homogeneous Monge–Ampère equations"],"falsifier":"Construct two convex bodies $K,L\\subset\\mathbb{R}^3$ and a direction $u$ that lies in the support of $S_{K,K,L}$ but is not a limit of $(K,K,L)$-extreme directions; that would overturn the paper's main theorem.","tokens_in":29100,"feed_emoji":"📐","tokens_out":7759,"duration_ms":73873,"temperature":0.7,"pith_summary":"The paper takes up an old question in convex geometry: when does equality hold in the monotonicity of mixed volumes? It reformulates this as the problem of locating the support of a mixed area measure $S_{C_1,\\ldots,C_{n-1}}$, and proves a decisive part of a 1985 conjectural characterization. In every dimension, the support is contained in the closure of the set of extreme directions; in three dimensions, the support is exactly that closure, settling the original equality problem for arbitrary convex bodies. The same methods yield a support characterization for mixed Hessian measures of convex functions and a mixed analogue of the classical theorem that a surface with vanishing Gaussian curvature is ruled.","feed_headline":"Mixed-volume equality cases in 3-D are now fully solved","feed_subtitle":"Characterizes the support of mixed area measures and proves one direction of the 1985 conjecture in all dimensions.","key_machinery":"The argument is carried by mixed area measures and the facial geometry of convex bodies. The central object is the touching cone $T(K,u)$, the cone of normal directions that meet $u$ in its relative interior; a direction is $(C_1,\\ldots,C_{n-1})$-extreme when, for every subset $I$, the span of the perpendicular spaces $T(C_i,u)^\\perp$ has dimension at least $|I|$. The upper bound uses a Lipschitz covering of $k$-singular boundary points, refining a classical covering result, to show that tubes of radius $t$ around such points have boundary area $O(t^{n-k-1})$. The lower bound in three dimensions is driven by a measure-preserving modification $\\hat K$ of $K$ that eliminates one-dimensional touching cones inside a cap, constructed via the equality case of the Alexandrov–Fenchel inequality, combined with a projection lemma that lets the proof descend one dimension.","core_discovery":"The central claim is that for any convex bodies $K,L\\subset\\mathbb{R}^n$, the support of the mixed area measure $S_{K,\\ldots,K,L}$ equals the closure of the set of directions $u$ that are $(K,\\ldots,K,L)$-extreme. Together with the companion statement that the mixed area measure vanishes on all directions that are not exposed, this proves the full conjecture in dimension $n=3$ and one full direction of it in every dimension. The result gives a geometric characterization of equality in mixed-volume monotonicity: equality holds exactly when the two bodies share supporting hyperplanes in every direction lying in this support. A direct corollary is that, for the measures covered by the theorem, every extreme direction is a limit of exposed directions.","pith_inferences":["If the equality condition of the Alexandrov–Fenchel inequality is reproved in the full degenerate generality needed here, the measure-preserving modification construction would likely extend to arbitrary mixed area measures, upgrading the $\\mathbb{R}^3$ result to all dimensions.","The established equivalence between the convex-body and convex-function formulations suggests that the new support characterization should transfer to homogeneous mixed Monge–Ampère equations in all dimensions, not just the two-function case treated here.","A concrete test: for smooth strictly convex bodies in $\\mathbb{R}^4$, one could numerically approximate $S_{K,K,K,L}$ and compare its support with the closed set of extreme directions; a mismatch would disprove the full conjecture, and a match would support the remaining open direction.","The projection obstruction indicates that non-projection methods, likely inspired by affine rigidity or by the line-foliation structure of the flat-surface theorem, are required for the full conjecture in higher dimensions."],"forward_implications":["Minkowski's monotonicity equality problem is fully resolved in $\\mathbb{R}^3$: for any pair of convex bodies, the support of $S_{K,K,L}$ is exactly the closure of the extreme directions.","In every dimension, the mixed area measure assigns zero mass to directions that are not exposed, so the support is always contained in the closure of the extreme directions for arbitrary convex bodies.","For mixed Hessian measures of convex functions, the same support characterization holds in dimension two, and the inclusion holds in all dimensions.","If the mixed Hessian measure $H_{f,g}$ vanishes on an open connected domain, then outside planar regions the domain is foliated by affine lines on which both functions are affine, giving a mixed analogue of the classical flat-surface theorem.","A projection-based induction that would prove the full conjecture in higher dimensions is blocked: the paper exhibits bodies in which a direction is extreme after projection but not extreme for the original body, so any higher-dimensional proof needs a new ingredient.","The support result implies that, for the measures covered, every extreme direction is a limit of exposed directions, a fact previously known only in special cases."],"supporting_citations":[{"why":"Supplies the standard theory of mixed volumes, mixed area measures, projection formulas, and the equality case of the Alexandrov–Fenchel inequality used throughout.","marker":"[21]"},{"why":"Formulates the conjectural support characterization in terms of touching cones and verifies it for polytopes and other special classes.","marker":"[19]"},{"why":"Provides the classical theorem that a surface with vanishing Gaussian curvature is ruled, which serves as the model for the mixed analogue proved here.","marker":"[6]"},{"why":"Gives the analytic characterization, via mixed discriminants, of when positivity holds in the smooth case, used in the discussion of the smooth support and in the Hessian-measure setting.","marker":"[15]"},{"why":"Establishes the correspondence between mixed area measures and mixed Hessian measures, which is used to transfer the main results to convex functions.","marker":"[7]"},{"why":"Provides prior equality-case results and techniques for special classes, and is invoked in extending the main theorem to combinations with previously known cases.","marker":"[23]"},{"why":"Gives earlier support results for zonoids and polyoids, and supplies tools used in the analysis of support under projection.","marker":"[8]"},{"why":"Defines mixed Hessian measures and proves the continuity property needed to extend the definitions to non-smooth convex functions.","marker":"[24]"},{"why":"Provides the classical covering of k-singular boundary points that is refined and used in the upper-bound proof.","marker":"[1]"}],"fun_headline_variants":["Mixed-volume equality cases fully solved in 3D","Schneider's conjecture proven for convex bodies in R^3","Extreme directions reveal mixed area measure support","Minkowski monotonicity equality characterized in all dimensions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The lower-bound proof assumes that the equality case of the Alexandrov–Fenchel inequality, cited from the literature, holds without hidden degeneracy exceptions for the possibly non-smooth, possibly lower-dimensional bodies that appear in the construction of the modified body.","fun_headline_variants_meta":{"raw":{"variants":["Mixed-volume equality cases fully solved in 3D","Schneider's conjecture proven for convex bodies in R^3","Extreme directions reveal mixed area measure support","Minkowski monotonicity equality characterized in all dimensions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000723,"raw_usage":{"total_tokens":3181,"prompt_tokens":823,"completion_tokens":2358,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":439,"completion_tokens_details":{"reasoning_tokens":2293}},"tokens_in":439,"tokens_out":2358,"duration_ms":17147,"temperature":1.0,"reasoning_tokens":2293,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T17:52:09.824326+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct two convex bodies $K,L\\subset\\mathbb{R}^3$ and a direction $u$ that lies in the support of $S_{K,K,L}$ but is not a limit of $(K,K,L)$-extreme directions; that would overturn the paper's main theorem.","supporting_citations":[{"cited_title":"Schneider","cited_arxiv_id":null,"evidence_quote":"Supplies the standard theory of mixed volumes, mixed area measures, projection formulas, and the equality case of the Alexandrov–Fenchel inequality used throughout."},{"cited_title":"Schneider","cited_arxiv_id":null,"evidence_quote":"Formulates the conjectural support characterization in terms of touching cones and verifies it for polytopes and other special classes."},{"cited_title":"Hartman and L","cited_arxiv_id":null,"evidence_quote":"Provides the classical theorem that a surface with vanishing Gaussian curvature is ruled, which serves as the model for the mixed analogue proved here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the analytic characterization, via mixed discriminants, of when positivity holds in the smooth case, used in the discussion of the smooth support and in the Hessian-measure setting."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the correspondence between mixed area measures and mixed Hessian measures, which is used to transfer the main results to convex functions."},{"cited_title":"Shenfeld and R","cited_arxiv_id":null,"evidence_quote":"Provides prior equality-case results and techniques for special classes, and is invoked in extending the main theorem to combinations with previously known cases."},{"cited_title":"Hug and P","cited_arxiv_id":null,"evidence_quote":"Gives earlier support results for zonoids and polyoids, and supplies tools used in the analysis of support under projection."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines mixed Hessian measures and proves the continuity property needed to extend the definitions to non-smooth convex functions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the classical covering of k-singular boundary points that is refined and used in the upper-bound proof."}],"review_version":1}