{"id":"e0e7484d-4b06-42eb-8c40-35830d0e2975","arxiv_id":"2603.05292","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Each tropical vector bundle carries a convex chain whose values equal its equivariant Euler characteristic, yielding a combinatorial Riemann–Roch formula; for the tautological bundle of a matroid, the Euler characteristic equals the rank of global sections.","lead":"Tropical vector bundles, combinatorial analogs of vector bundles on toric varieties, are shown to have their Euler characteristics computable from a convex chain of polytopes, and the tautological bundle of any matroid satisfies an Euler-characteristic/global-section equality. The paper connects Khovanskii–Pukhlikov Ehrhart theory with matroid and toric geometry, and settles a question of Kaveh and Manon.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"False refinement-invariance identity (Prop 4.10) invalidates Theorem 4.11, so the proof of Theorem 1.1 does not go through.","rationale":"I read the paper in good faith and attempted to fill in the missing refinement argument flagged by the reader. That linearization step is standard. However, while checking the surrounding proof, I found a more serious, concrete flaw: the invariance theorem (Theorem 4.11) on which the proof of Theorem 4.13 explicitly relies is proved via Proposition 4.10, and Proposition 4.10's key identity is false. The proof of Proposition 4.10 uses the equality (4), which is an identity of alternating sums over all faces; it does not permit Möbius inversion to isolate a single face. The counterexample with the square cone and its diagonal refinement shows the claimed per-cone identity fails numerically. Moreover, the same counterexample extends to a line bundle on P^2, where the combinatorial χ defined in Definition 2.20 changes under refinement, contradicting Theorem 4.11. This is not a matter of missing detail or convention: the printed argument is internally inconsistent at a load-bearing step. The central theorem may be true, but the proof as written is unsupported, and the fix is not a minor clarification. Hence the verdict should move from CONDITIONAL to REJECT unless the authors supply a correct proof of refinement invariance or an alternative route to Theorem 4.13 that does not use the false identity.","tokens_in":19657,"tokens_out":52955,"duration_ms":439044,"concrete_test":"Compute the equivariant Euler characteristic of the toric line bundle L with a(v1)=0, a(v2)=2, a(v3)=0 on P^2 for u=(0,1) using Definition 2.20: first on the standard fan of P^2, then on the fan obtained by adding the diagonal ray δ with value 2. If the two values are not equal (1 vs −3), Theorem 4.11 is false and Proposition 4.10 is refuted.","verdict_should_be":"REJECT","load_bearing_attack":"The proof of Theorem 1.1 via Theorem 4.13 depends on Theorem 4.11, the invariance of the combinatorial Euler characteristic under fan refinement. Theorem 4.11 is proved from Proposition 4.10, whose identity (3) is false: (−1)^codim(σ) 1_{σ^∨} = Σ_{τ∈φ^{-1}(σ)} (−1)^codim(τ) 1_{τ^∨}. The proof claims 'Möbius inversion' on the face poset, but the right side of (4) is not a scalar combination of the functions 1_{F^∨}: for τ⊂F one has τ^∨ ⊃ F^∨, so the summands are indicators of larger cones. Explicit counterexample: in N=R^2, take σ=cone(e1,e2) and refine it by δ=cone(e1+e2). Then φ^{-1}(σ)={σ1,σ2} with σ1=cone(e1,δ), σ2=cone(δ,e2). The identity would give 1_{σ^∨}=1_{σ1^∨}+1_{σ2^∨}, but at u=(1,1) the left side is 1 and the right side is 2. Consequently the invariance in Theorem 4.11 fails even for line bundles: for a_ρ=(0,2,0) on P^2 and u=(0,1), Definition 2.20 gives χ_u=1 on the standard fan, while the pullback to the fan refined by the diagonal ray gives χ_u=−3. Since Theorem 4.13 explicitly replaces Σ by a refinement using Theorem 4.11, the proof of the central equality χ=α_E collapses.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops an Ehrhart-type theory for tropical vector bundles on complete toric varieties. It associates to a tropical vector bundle E a convex chain α_E and claims that its evaluation at a character u equals the equivariant Euler characteristic χ(X_Σ,E)_u (Theorem 1.1). From this it derives a Khovanskii–Pukhlikov-style Hirzebruch–Riemann–Roch formula (Corollary 1.3). The proof strategy is to establish invariance of the combinatorial Euler characteristic under fan refinement (Theorem 4.11), then refine the fan so that the order-statistic functions h_i are linear and apply Brianchon–Gram (Theorem 4.13). The paper also gives a split-resolution analogue for tropical vector bundles (Section 4.2) and proves that for the tautological matroid bundle E_M one has χ(X_m,E_M)_u = h^0(X_m,E_M)_u for all characters u (Theorem 5.7), interpreted as vanishing of higher cohomologies.","tokens_in":20032,"tokens_out":12197,"duration_ms":115974,"significance":"If its main theorem is correct, the paper gives a clean combinatorial formula for the equivariant Euler characteristic of tropical vector bundles and answers a question of Kaveh–Manon for tautological matroid bundles. The convex-chain viewpoint is natural, the split-resolution construction is potentially useful, and the Fano-plane and U_{2,3} examples are welcome. However, the central proof has a serious gap: Proposition 4.10, which underpins Theorem 4.11 and hence Theorem 4.13, is false as stated. Because the main theorem depends on this unproved/false step, the paper is not yet in publishable form, although the overall approach may be salvageable.","major_comments":[{"comment":"The claimed refinement identity is false. Take N=R^2, σ=cone(e1,e2), and refine by the ray δ=cone(e1+e2). Then φ^{-1}(σ)={σ1,σ2} with σ1=cone(e1,δ) and σ2=cone(δ,e2), all of codimension 0. Equation (3) becomes 1_{σ^∨}=1_{σ1^∨}+1_{σ2^∨}. At u=(1,1), both σ1^∨ and σ2^∨ contain u, so the right-hand side equals 2 while the left-hand side equals 1. The 'Möbius inversion' step is invalid because the functions 1_{τ^∨} are not ordered with the appropriate triangularity: for τ⊂σ one has τ^∨⊃σ^∨. Since Theorem 4.11 is proved directly from this identity, and Theorem 4.13 explicitly invokes Theorem 4.11 to replace Σ by a refinement, the proof of Theorem 1.1 collapses at this point.","section":"§4.3, Proposition 4.10, Eq. (3)"},{"comment":"The proof asserts that 'after replacing Σ with a refinement of Σ, we can assume that the h_i are linear on each cone of Σ'. This simultaneous linearization of the order-statistic functions is not proved; it is plausible via a common refinement of hyperplane arrangements, but as written it is an unproved assumption. More importantly, the replacement of Σ uses Theorem 4.11, whose proof is invalid because of the false Proposition 4.10. Thus the reduction to a sum of line-bundle contributions and the application of Brianchon–Gram are not justified. The authors need either a correct proof of refinement invariance or a direct argument for Theorem 4.13.","section":"§4.3, proof of Theorem 4.13"},{"comment":"The involution argument proving cancellation in Corollary 5.5 is not rigorous as written. The condition 'if |S_{k-1} \\ S_k|>1' is impossible because S_{k-1}⊂S_k; presumably |S_k \\ S_{k-1}| was intended. More seriously, the text does not verify that the proposed map preserves the defining property of Π_u(S), namely that the distinguished flat is the first one at which ⟨u,e_{S_t}⟩=1, nor that the map is a sign-reversing involution. The coefficient computation for rank(S) is therefore asserted rather than demonstrated. Since this is the proof of Theorem 1.4, the vanishing statement is not established as written.","section":"§5.2, proof of Theorem 5.7"}],"minor_comments":[{"comment":"There are numerous typos and formatting issues: the title is given as 'EHRHART THEOR Y', 'Gröbner' is misspelled, and 'fGF(M)' is used without a consistent definition in the notation list.","section":"Global"},{"comment":"The paper explicitly notes that no higher cohomology functors are defined for tropical vector bundles. The phrase 'vanishing of higher cohomologies' in the abstract and Theorem 1.4 should therefore be understood only as the equality χ=h^0; this caveat should be stated in the main theorem statement itself, not only in a remark.","section":"Remark 1.5"},{"comment":"The compatibility proof for the split bundles F^k is hard to follow: it does not explicitly define the direct-sum basis for the total space or verify the compatibility condition on overlaps of cones. This is not central to Theorem 1.1, but a clearer proof would improve the paper.","section":"§4.2, Proposition 4.2"},{"comment":"In the proof, the statement that the exponential sum is determined by its values on the rays of τ is justified by smoothness, but it only requires that τ is a full-dimensional cone generated by its rays; smoothness is not needed. The reasoning should be rephrased.","section":"§4.3, Theorem 4.3"},{"comment":"The reference [CHK] is incomplete ('arXiv:' with no number). Also, the manuscript lacks page numbers in the citation to [KM25], which should be supplied.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The false Proposition 4.10 is a genuine load-bearing error, not merely a gap in exposition. If the authors can supply a correct proof of refinement invariance, or an alternative proof of Theorem 4.13 avoiding the false identity, the central result may be recoverable. The tautological-bundle section also needs a complete involution argument. I therefore recommend major revision rather than rejection, but the next version must contain a correct proof of the key invariance step."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is in good shape. The central result — χ(X_Σ,E)_u = α_E(u) for tropical vector bundles — is new, and the route through Khovanskii–Pukhlikov convex chains is well chosen. It resolves the Kaveh–Manon question about the tautological bundle of a matroid, which is a genuine advance, and the extension of Klyachko's split resolution to the tropical setting is a useful tool. The examples are consistent, and the combinatorial HRR corollary is a nice payoff.\n\nOn the stress-test: the claimed counterexample to Proposition 4.10 misfires. In the proposed refinement of the quadrant, φ^{-1}(σ) also contains the new ray δ = cone(e1+e2), not just the two full-dimensional cones. Including that ray, with its codimension sign, the identity becomes 1 = 1+1−1 at (1,1), so it holds. The invariance theorem is not falsified by that argument.\n\nThat said, there are genuine weaknesses, mostly in presentation. The proof of Theorem 4.13 asserts that after refinement the sorted Chern-root functions h_i are linear on every cone; this is a standard simultaneous-linearization fact, but it is not proved or even stated as a lemma. Given that Theorem 4.11 supplies invariance, the step is likely fine, but a referee will want it spelled out. The involution proof of Theorem 5.7 is cryptic and contains apparent typos — e.g. S_{k-1} \\ S_k when the sets are nested — and as written it does not convince. The authors should rewrite that paragraph carefully. Finally, Remark 1.5 honestly concedes that “vanishing of higher cohomologies” is an interpretation rather than a cohomological statement; that is acceptable for the paper, but the abstract and Theorem 1.4 should carry the same caveat.\n\nOverall: the mathematical core appears sound, the citation pattern is appropriate, and the new results are substantial within tropical geometry and matroid theory. The paper should go to peer review, with the requested revisions focused on the refinement lemma and the involution proof.","headline":"The main theorem is real — the stress-test counterexample to Proposition 4.10 drops a cone, so it does not land; the real soft spots are an asserted refinement step and a garbled involution proof in §5.","tokens_in":20566,"tokens_out":7486,"would_cite":true,"duration_ms":67507,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14T05","14M25","52B20","14C40"],"pacs":[],"model":"deepseek-v4-flash","headline":"For a tropical vector bundle on a toric variety, the equivariant Euler characteristic at a character u equals the value at u of an associated convex chain α_E, yielding a combinatorial Hirzebruch–Riemann–Roch formula.","keywords":["tropical vector bundles","toric varieties","Ehrhart theory","convex chains","Euler characteristic","Hirzebruch–Riemann–Roch","matroids","Bergman fan"],"falsifier":"Compute the equivariant Euler characteristic directly from its alternating-sum definition on any fan, and compute α_E(u) for the same tropical vector bundle; a disagreement on a single example would refute Theorem 1.1. For instance, take the Fano-plane bundle of Example 4.15, where h_1 is non-convex, refine the fan to linearize h_1, and compare the two computations.","tokens_in":19478,"feed_emoji":"📐","tokens_out":11017,"duration_ms":88578,"temperature":0.7,"pith_summary":"The paper establishes a precise equality for tropical vector bundles on toric varieties: the equivariant Euler characteristic at a character u equals the value at u of a convex chain α_E built from the bundle's equivariant Chern roots. Because convex chains are governed by Khovanskii–Pukhlikov's Ehrhart theory, this equality yields a combinatorial Hirzebruch–Riemann–Roch formula, in which a Todd-class differential operator applied to the integral of α_E recovers the Euler characteristic. The paper also proves that the tautological bundle of any matroid has equal Euler characteristic and global-section rank, interpreted as vanishing of higher cohomology. The key supporting steps are a split-bundle resolution for tropical vector bundles and invariance of the Euler characteristic under fan refinements.","feed_headline":"Euler characteristic of tropical bundles is a lattice-polytope count","feed_subtitle":"Chern-root convex chains yield an exact Riemann-Roch identity, and matroid tautological bundles satisfy Euler = h^0.","key_machinery":"Khovanskii–Pukhlikov theory of convex chains: a convex chain is a finite integer combination Σ n_i 1_{P_i} of indicator functions of polytopes; its lattice sum S(α) and integral I(α) generalize the Ehrhart polynomial and volume. The paper encodes a tropical vector bundle E, defined by a piecewise linear map Φ_E : |Σ| → the lifted Bergman fan of a matroid, by a multi-valued support function h_E(x) = Σ_i [⟨u_{σ,i}, x⟩] on each maximal cone σ. This support function determines the convex chain α_E. The Khovanskii–Pukhlikov theorem, an exact Euler–Maclaurin formula sometimes called a multidimensional Riemann–Roch for polytopes, connects S and I, giving the combinatorial HRR formula. The proof als","core_discovery":"The central claim is Theorem 1.1: for a tropical vector bundle E on X_Σ, the convex chain α_E — constructed from the equivariant Chern roots of E — satisfies χ(X_Σ,E)_u = α_E(u) for every character u. The chain is derived from a multi-valued support function h_E(x) = Σ_i [⟨u_{σ,i}, x⟩] on each maximal cone σ. The proof first shows the Euler characteristic is invariant under refinements of the fan (Theorem 4.11), then passes to a refinement where the sorted Chern-root functions h_i are linear on each cone, so each h_i is the support function of a virtual polytope; Brianchon–Gram then sums the contributions to α_E. Combined with the Khovanskii–Pukhlikov theorem, this yields the combinatorial H","pith_inferences":["The unproved linearization step — that the sorted Chern-root functions can be made simultaneously linear on a refinement — is a concrete geometric property; testing it could either complete the proof or reveal counterexamples for more general piecewise linear maps.","The convex-chain encoding suggests that tropical vector bundles may admit an Ehrhart theory with multiplicities beyond matroids, possibly connecting to valuated matroids and matroid polytope invariants.","If the equality χ = h^0 holds for tautological bundles without a constructed cohomology theory, it raises the question of whether a derived category or homological algebra for tropical bundles can be defined so that the equality becomes a genuine vanishing theorem."],"forward_implications":["The Euler characteristic of a tropical vector bundle can be computed as S(α_E), a lattice-point count, making it accessible to Ehrhart-type algorithms.","The combinatorial HRR formula gives a Todd-correction to the volume of α_E that reproduces the Euler characteristic, a concrete analogue of the classical Hirzebruch–Riemann–Roch theorem.","Theorem 1.2 shows the equivariant Euler characteristic is invariant under passing to a refinement of the fan, so it is a well-defined invariant of the tropical bundle independent of the chosen fan presentation.","Theorem 5.7 shows that for matroid tautological bundles, higher cohomology vanishes in the only sense currently available, matching the representable case.","The split-resolution theorem extends Klyachko's K-theoretic decomposition from toric vector bundles to the tropical setting, providing a new tool for studying tropical bundles."],"fun_headline_variants":["Tropical bundles: Euler characteristic is a polytope count","Matroid bundles: Euler equals global sections","Chern roots become convex chains for tropical HRR","Riemann-Roch exact for tropical bundles and matroids","Tropical HRR: convex chains give exact Euler count"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The proof that the Euler characteristic matches the convex chain rests on the claim that after refining the fan, the sorted Chern-root functions become linear on every cone; the paper asserts this refinement exists but does not prove it.","fun_headline_variants_meta":{"raw":{"variants":["Tropical bundles: Euler characteristic is a polytope count","Matroid bundles: Euler equals global sections","Chern roots become convex chains for tropical HRR","Riemann-Roch exact for tropical bundles and matroids","Tropical HRR: convex chains give exact Euler count"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000778,"raw_usage":{"total_tokens":3279,"prompt_tokens":751,"completion_tokens":2528,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":495,"completion_tokens_details":{"reasoning_tokens":2464}},"tokens_in":495,"tokens_out":2528,"duration_ms":16401,"temperature":1.0,"reasoning_tokens":2464,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T18:44:45.321005+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the equivariant Euler characteristic directly from its alternating-sum definition on any fan, and compute α_E(u) for the same tropical vector bundle; a disagreement on a single example would refute Theorem 1.1. For instance, take the Fano-plane bundle of Example 4.15, where h_1 is non-convex, refine the fan to linearize h_1, and compare the two computations.","supporting_citations":[],"review_version":1}