{"id":"8e5d681d-950f-4936-b6b6-7c8d9449700a","arxiv_id":"2601.13249","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":3.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A survey of volume polynomials and their realization problems, connecting convex bodies, projective varieties, and algebraic matroids.","lead":"This paper is a survey of volume polynomials, a class of log-concave polynomials at the intersection of convex geometry, algebraic geometry, and matroid theory. It reviews recent realization results, states several conjectures, and explains applications to algebraic matroids.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 4.7 is the pivotal unproven input; the survey's covolume/symbol-theorem core depends on a universal-action result for which no proof is supplied.","rationale":"The reader's conditional verdict is appropriate: the survey's main new-structure results are quoted from unproved self-cited preprints. Among those, Theorem 4.7 is the single most load-bearing because it licenses the passage from 'g is dual to a monomial' to 'g acts on all volume polynomials', which is the core of the covolume duality and the symbol theorem. The reader identified the same broad class of assumptions but did not isolate this theorem; my stress test localizes the risk more sharply. Other parts of the survey are independently supported: Theorem 2.11 is published in [BH20], the nontriviality of the realization problem uses the published reverse Khovanskii–Teissier inequality, and the convex/matroid examples are internally consistent. Thus the concern does not change the verdict: the paper should remain conditional on the correctness of [GHM+], especially Theorem 4.7, and the concrete test above would either retire or confirm that concern. No ad hominem or rhetorical inflation is intended; this is purely a question of proof support.","tokens_in":18595,"tokens_out":20639,"duration_ms":206257,"concrete_test":"Obtain [GHM+] and isolate the proof of the forward direction of Theorem 4.7. Trace exactly where the implication (1)⇒(2) passes from the monomial test g(∂)∘x[μ] to arbitrary realizable f. If the proof reduces to a monomial test, expand f = Σ c_α x[α] and compute g(∂)∘f; check whether the binomial factors introduced by translating each x[α] to x[μ] are absorbed by a nonnegative linear change of coordinates or an explicit cone construction. A computational proxy: for k=Q, n=3, d=3, test a nontrivial covolume polynomial such as the one in Example 4.6 against a battery of realizable cubics f by checking that g(∂)f has nonnegative coefficients in degree 1; but the decisive check is the proof trace, since only it settles the universal quantifier. If the required lemma is absent or fails, Theorem 4.7 and the symbol-theorem corollaries are unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The survey's central structural claim is Theorem 4.7: g∈Q[∂] is a realizable covolume polynomial over k iff g(∂)∘f is a realizable volume polynomial for every realizable volume polynomial f. This is quoted from [GHM+], an arXiv preprint, and is not proved or marked as conditional in the paper. The nontrivial content is the move from Definition 4.1, which only tests g on one monomial x[μ], to a universal statement over all realizable f. The cited translation invariance ([Alu24, Remark 2.2]) explains why the choice of μ is irrelevant, but it does not by itself deliver the universal action; that is precisely where a hidden support or Leibniz-term assumption could live. If Theorem 4.7 fails, Corollary 4.8, the volume-polynomial symbol theorem (Theorem 4.10), and the operator-preservation results in §4.3 (truncation, polarization, normalization, interlacing, symmetric exclusion) lose their foundation. In contrast, the Lorentzian containment (Theorem 2.11) is published [BH20], and the strictness example in §3.2 rests on published reverse Khovanskii–Teissier inequalities; so the nontriviality of the realization problem is less vulnerable than the covolume-duality picture. The manuscript contains no proof sketch or limitation note addressing this gap.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper is a survey of recent developments on volume polynomials: the homogeneous degree-d polynomials f_C(x) = (1/d!)vol(x_1C_1+...+x_nC_n) of convex bodies and their algebraic analogues f_D(x) = (1/d!)∫_Y (Σ x_i D_i)^d for semiample divisors. It reviews realization problems for projection areas (Theorems 1.1–1.2, Conjectures 1.6–1.7), the Lorentzian characterization of quadratic and bivariate forms (Theorem 2.11, Example 2.2), the strict inclusion of volume polynomials over a field in the Lorentzian cone (Section 3.2), and the covolume/symbol-theoretic duality (Theorem 4.7, Theorem 4.10). The final section applies these ideas to algebraic polymatroids, including Proposition 5.5 and Corollary 5.6, and states open problems such as Question 5.7. Most displayed theorems are quoted without proof from the author's previous works, several of which are arXiv preprints or 'in preparation'.","tokens_in":18948,"tokens_out":11464,"duration_ms":111554,"significance":"If the quoted results are correct, the survey gives a valuable and accessible point of entry to a fast-moving area, with well-chosen concrete examples (e.g., Example 1.4 and Example 1.8) that illustrate the distinction between convex and algebraic realization. The Lorentzian baseline (Theorem 2.11) is published [BH20], and the nontriviality of the realization problem is supported by published reverse Khovanskii–Teissier inequalities. However, the manuscript's covolume/symbol core (Theorem 4.7, Corollary 4.8, Theorem 4.10) is inherited from the preprint [GHM+], and no proof or proof sketch is supplied. Because these results are load-bearing for the survey's structural claims and for the operator-preservation applications in Section 4.3, the survey should either include a careful statement of their logical status or provide enough of a proof/argument to let a reader assess the dependence. The paper otherwise reads as a reliable overview and is likely to be useful to the intended audience.","major_comments":[{"comment":"The equivalence (1)⇔(2) is quoted from [GHM+], an arXiv preprint, and is not proved or marked as conditional. This is the pivotal structural input: it is used to derive Corollary 4.8, the symbol theorem (Theorem 4.10), and the operator-preservation results in §4.3. If the preprint contains a hidden hypothesis (e.g., a support or Leibniz-term assumption, or a characteristic restriction), the survey's statements are incomplete. Please add (i) an explicit statement that Theorem 4.7 is an unpublished theorem of [GHM+], (ii) a proof sketch or at least a precise formulation of the hypotheses under which it is known, and (iii) a note in §4.3 stating which of the listed consequences depend on this theorem.","section":"§4.2 (Theorem 4.7)"},{"comment":"Definition 4.1 tests a covolume polynomial on a single monomial x^[μ], while Theorem 4.7(2) asserts an action on every realizable volume polynomial f. Remark 4.2 uses translation invariance to show independence of μ, but the step from one monomial to arbitrary realizable f is exactly the nontrivial content of Theorem 4.7; it is not a formal consequence of translation invariance. The manuscript should state this explicitly and not give the impression that Remark 4.2 reduces Theorem 4.7 to a routine check.","section":"§4.1 (Definition 4.1 and Remark 4.2)"},{"comment":"The example after 'For example, consider the cubic polynomial' is used to establish the proper inclusion V^d_n(R,k) ⊊ L^d_n for d≥3,n≥3. The Lorentzian property is verified by listing three Hessians, but the non-realizability of f is only asserted to follow from the reverse Khovanskii–Teissier inequality; no computation shows how the inequality applies to this specific f. Since this is the paper's main demonstration that the realization problem has a nontrivial answer, please add the missing verification or give the exact location in [Huh23, Example 14] and reproduce the key inequalities.","section":"§3.2"},{"comment":"Proposition 5.5 is quoted from [GHM+, Proposition 5.4] and is then used to derive Corollary 5.6 and Theorem 5.11. This is another load-bearing transfer from the same unpublished source. Please state the provenance clearly and, if the result depends on the same hypotheses as Theorem 4.7, say so. Otherwise the survey's application section rests on a result whose proof is not available to the reader.","section":"§5 (Proposition 5.5 and its use)"}],"minor_comments":[{"comment":"The text refers to pictures that do not appear in the arXiv version ('The above pictures show...' and the discussions of four matroids). Please add the figures or remove the references.","section":"§2.3 and §2.4"},{"comment":"The Young diagram is blank; the displayed polynomial suggests λ=(2,1). Please insert the intended diagram.","section":"Example 3.6"},{"comment":"'Limit of realizable volume polynomials' should specify the topology (coefficient-wise convergence, presumably) and, for non-algebraically-closed fields, the convention for semiample divisors should be clarified.","section":"Definition 3.1"},{"comment":"The strict inclusion is stated for any field k, while the cited reverse Khovanskii–Teissier inequality of [JL23] is stated for algebraically closed fields; clarify whether a base-change reduction is used.","section":"§3.2"},{"comment":"The endpoint positivity convention in the log-concavity criterion should be stated, to avoid ambiguity when p_0=0 or p_d=0.","section":"Example 2.2"},{"comment":"The sentence 'If this map is induced by an irreducible correspondence Γ... then, by [GHM+, Lemma 2.1], it preserves the classes of irreducible cycles up to a rational multiple' is unclear. Please define φ_T more precisely and rephrase the geometric interpretation.","section":"§4.3"}],"recommendation":"major_revision","confidential_remarks":"To the editor: this is a survey by the leading researcher in the area and is likely to be widely read. The main concern is provenance: several central theorems are from preprints by the author and collaborators, and the paper does not distinguish published from unpublished results. If the venue accepts surveys, this is acceptable provided the status is made explicit; if the journal requires original research, the paper would need a substantial new contribution. I also note the heavy self-citation, which is natural for a survey but should be balanced by an editorial note."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a clean, readable survey of the volume-polynomial program, and it does a real service by organizing the convex-geometric and algebraic-geometric realization problems side by side. The new material is mostly conjectures and examples — 1.6, 1.7, 1.11, 3.8, 3.12 — plus the covolume picture in §4. The examples are unusually explicit: I checked Example 1.8's eigenvalue calculation and the convex body claim; both work. The contrast between Conjectures 1.6 and 1.7 is genuinely illuminating.\n\nThe soft spot is exactly what the stress-test says. Theorem 4.7 is the load-bearing result for everything in §4 — Corollary 4.8, the symbol theorem 4.10, the operator-preservation results — and it is quoted from [GHM+], an arXiv preprint, with no proof sketch and no caveat that it is unpublished. From Definition 4.1, (1) only says g sends the single monomial x^[μ] to a realizable volume polynomial; (2) says it sends every realizable volume polynomial. The gap between those two is the whole theorem. It may be true, but the survey doesn't give the reader any way to see why. I'd want the referee to check whether [GHM+] actually proves it and whether there's a hidden hypothesis (support conditions, degree bounds) being swept under the rug.\n\nThe Lorentzian-containment part, by contrast, is on solid ground: Theorem 2.11 is [BH20], a published Annals paper, and the strictness example in §3.2 uses published reverse Khovanskii–Teissier inequalities. So the basic nontriviality of the realization problem is safe. The reader's instinct to mark this CONDITIONAL is right.\n\nThe self-citation is heavy, but it's honest — the theorems are attributed to specific references, not silently reused. The bigger issue is that the survey doesn't flag which references are published and which are 'in preparation.' A reader who doesn't already know the field could take the whole edifice as established. That's a fixable editorial problem, not a scientific one.\n\nFor a survey, this deserves a serious referee. The examples and conjectures are worth engaging with, and the open questions are well posed. But the referee should insist that either Theorem 4.7 be proven in an appendix or the paper clearly mark it as conjectural/unpublished — and the same for the other results quoted from preprints. As it stands, it's a reliable roadmap to a literature that is partly still in flux.","headline":"A lucid survey of the volume-polynomial program whose central duality theorem rests on an unproved, preprint-level input; worth refereeing, but the referee should demand a proof or a clear status flag.","tokens_in":19364,"tokens_out":3932,"would_cite":true,"duration_ms":39071,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14-06","52-06"],"pacs":[],"model":"deepseek-v4-flash","headline":"This survey argues that volume polynomials, defined in convex and algebraic geometry, are completely characterized among quadratic forms by the Lorentzian property, and that covolume polynomials are exactly the linear operators that preserv","keywords":["volume polynomials","Lorentzian polynomials","mixed volumes","covolume polynomials","algebraic matroids","realization problems","Alexandrov–Fenchel inequality","polymatroids"],"falsifier":"A concrete check: exhibit a single polynomial g for which g(∂) maps every realizable volume polynomial to a realizable volume polynomial, but g is not itself realizable as a covolume polynomial over k; Theorem 4.7 says no such g exists. Alternatively, produce a non-realizable Lorentzian quadratic form over a field, which would refute the degree-two characterization.","tokens_in":18512,"feed_emoji":"📐","tokens_out":3789,"duration_ms":38446,"temperature":0.7,"pith_summary":"The paper surveys the theory of volume polynomials: homogeneous polynomials that measure how the volume of Minkowski sums or intersections changes with scaling. Its central thesis is that these polynomials form a distinguished log-concave class whose realization problem — which polynomials actually arise — has a complete and surprisingly clean answer in low degree, matched by a sharp operator-theoretic characterization in general. A sympathetic reader comes away with a unified picture: quadratic volume polynomials over any field are precisely the Lorentzian forms, bivariate forms are exactly Lorentzian, and the dual notion of covolume polynomial is exactly the class of differential operators that preserve realizable volume polynomials. These results settle old questions, like Gurvits's conjecture, and tie the subject to algebraic matroids.","feed_headline":"Covolume polynomials are exactly the operators that preserve volumes","feed_subtitle":"The survey also pins quadratic volume polynomials to Lorentzian forms and ties realizable volumes to algebraic matroids.","key_machinery":"Volume polynomials themselves — Minkowski-sum volume polynomials in convex geometry and intersection-product volume polynomials in algebraic geometry — with the Alexandrov–Fenchel inequality as the source of log-concavity. Lorentzian polynomials, defined by the Hessian condition or the M-convex support condition, form the ambient class. Covolume polynomials, defined via the action of R[∂] on R[x], are the dual objects; Theorem 4.7 is the key bridge showing that they are precisely the linear operators preserving realizable volume polynomials.","core_discovery":"The paper's core claim is that the set of volume polynomials over a field k is a proper, well-behaved subset of Lorentzian polynomials for degree d≥3 and at least 3 variables, but coincides with the Lorentzian class in the quadratic case and in the bivariate case. The main structural result, Theorem 4.7, characterizes realizable covolume polynomials over k as those polynomials g for which g(∂) sends every realizable volume polynomial to another realizable volume polynomial; taking limits gives the analogous statement for volume polynomials. Theorem 2.11 identifies Lorentzian polynomials as exactly those homogeneous polynomials with nonnegative coefficients whose support is M-convex and whose","pith_inferences":["The paper's conjectures suggest that the distinction between 'realizable' and 'limit of realizable' volume polynomials is the right seam: the algebraic realization problem has integrality obstructions that the convex one lacks, as the (1,1,1,1,1,3) example shows.","If Conjecture 3.12 is true, the algebraic geometric volume polynomials are exactly the closure of minors of convex-geometric ones, which would mean Hodge-theoretic positivity is fully generated by convex bodies.","A resolution of Question 5.7 would determine whether covolume polynomials carry genuinely new algebraic information; if duals of algebraic matroids are always algebraic, the support theory of covolume polynomials collapses into that of volume polynomials.","The triangular hyperfield appears throughout, suggesting the realization conditions may be expressible as Grassmannian Plücker relations over a tropical-style semiring."],"forward_implications":["Quadratic volume polynomials over any field are exactly the Lorentzian quadratic forms, so the realization problem in degree two is completely solved.","The existence of a Lorentzian cubic that is not a volume polynomial over any field disproves Gurvits's strong-log-concavity conjecture in three variables.","Since covolume polynomials preserve volume polynomials under differential action, every known inequality for volume polynomials can be applied to Schubert transforms to yield new mixed-volume inequalities.","The support of every realizable volume polynomial is an algebraic polymatroid, giving a new proof of Lindström's theorem for polymatroids and the closure of algebraic matroids under intersection.","The characterization of covolume polynomials gives a volume-polynomial analogue of the symbol theorem for Lorentzian operators."],"fun_headline_variants":["Volume polynomials are precisely volume-preserving operators","Volume polynomials pin down the volume-preserving class","Quadratic volume polynomials match Lorentzian forms","Volume polynomials and algebraic matroids: the link","Realizable covolume polynomials preserve realizability"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The survey's load-bearing premise is that the theorems quoted from the author's recent preprints — in particular Theorem 4.7 and Theorem 2.11 — are correct as stated; since they are not proved in these notes, any error in those sources would propagate.","fun_headline_variants_meta":{"raw":{"variants":["Volume polynomials are precisely volume-preserving operators","Volume polynomials pin down the volume-preserving class","Quadratic volume polynomials match Lorentzian forms","Volume polynomials and algebraic matroids: the link","Realizable covolume polynomials preserve realizability"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000112,"raw_usage":{"total_tokens":802,"prompt_tokens":553,"completion_tokens":249,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":297,"completion_tokens_details":{"reasoning_tokens":180}},"tokens_in":297,"tokens_out":249,"duration_ms":3472,"temperature":1.0,"reasoning_tokens":180,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T09:34:33.143447+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete check: exhibit a single polynomial g for which g(∂) maps every realizable volume polynomial to a realizable volume polynomial, but g is not itself realizable as a covolume polynomial over k; Theorem 4.7 says no such g exists. Alternatively, produce a non-realizable Lorentzian quadratic form over a field, which would refute the degree-two characterization.","supporting_citations":[],"review_version":1}