{"id":"f8d96ec0-dc3a-4fb1-9688-a43b6e956270","arxiv_id":"2411.19889","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Tropical subrepresentations of finite groups correspond exactly to weak isomorphism classes of weak group actions on valuated matroids.","lead":"This paper proves that, for finite groups, tropical subrepresentations of a group are the same thing as weak actions on valuated matroids, and it develops a module theory over tropical semirings to make the correspondence intrinsic. The result gives a dictionary between tropical geometry and matroid theory that could simplify how representations over idempotent semifields are classified.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem E(2)/Prop 6.13 is stated without the simplicity hypothesis and is false for non-simple matroids; the paper's own Example 6.14 (rank-1 uniform) gives a counterexample, so the abstract's unqualified quasi-free-module claim overreaches.","rationale":"I read the paper in good faith. The main technical development — weakly free and quasi-free modules, the structure of Aut(V_M) as H ⋉ V, and the correspondence in Corollary 6.10 — is carefully argued and, under the stated hypotheses (simple matroids, finite group G), appears correct. Theorem 6.7 and Corollary 6.10(1) do not obviously require simplicity; the proof for V_M goes through for loopless matroids, and loops reduce to a lower-dimensional loopless situation. However, the introduction's Theorem E(2), which asserts the image of Aut(Q_M) equals Aut_w(M) for an arbitrary valuated matroid, is directly contradicted by the paper's own Example 6.14. That example shows Q_M collapses to T for U_{1,3}, so the image is trivial while Aut_w is S_3. Thus the abstract's claim of an intrinsic description via quasi-free modules is false as stated. This is an internal inconsistency rather than a matter of disagreement with the field. The reader's verdict of CONDITIONAL remains appropriate, but the reasoning should be updated: the missing 'simple' hypothesis is load-bearing for the quasi-free-module statement, not for the V_M-based core correspondence. The finite-group hypothesis is correctly stated in Theorem E and is essential, as Example 6.11 shows. I find no error in the main proof of Corollary 6.10 under its assumptions, and I credit the paper for explicitly flagging both the infinite-group failure and the non-simple counterexample. The fix is straightforward: add 'simple' to the statements involving Q_M in the abstract and Theorem E, or replace it with the weaker 'loopless' where the V_M correspondence is concerned.","tokens_in":33335,"tokens_out":26793,"duration_ms":236279,"concrete_test":"Run the quotient construction of Section 6 for the rank-1 uniform valuated matroid U_{1,3} (w(B)=1 for every singleton). The bend relations from each 2-element circuit identify all standard basis vectors, so Q_M = T^3 / ⟨e_i ∼ e_j⟩ ≅ T. Then Aut(Q_M) = Aut(T) = T^×, whose image under π : GL_3(T) → S_3 is trivial, while Aut_w(U_{1,3}) = S_3. This reproduces Example 6.14 and directly falsifies the unqualified statement of Theorem E(2). As a second check, for the same matroid with a finite group (e.g., G = C_2), verify that the V_M-based correspondence of Corollary 6.10(1) still holds: Hom(G, Aut(V_M)) up to isomorphism reduces to conjugacy classes of Hom(G, S_3), matching weak isomorphism classes of weak G-actions. This confirms that simplicity is essential for the Q_M statement but not for the V_M correspondence.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central advertised claim in the introduction, Theorem E, has two parts. Part (2) says that for any valuated matroid M, the image of Aut(Q_M) under π : GL_n(T) → S_n is the weak automorphism group Aut_w(M). This is stated without the simplicity hypothesis. Section 6 begins with the standing assumption that all matroids are simple, and Proposition 6.13 is proved under that assumption. But Example 6.14 in the same section explicitly constructs the rank-1 uniform valuated matroid U_{1,3}, for which Q_M collapses to T, so Aut(Q_M)=T^× and its image in S_3 is trivial, while Aut_w(U_{1,3})=S_3. Thus the unqualified Theorem E(2) is false. Because the abstract advertises an 'intrinsic description of tropical subrepresentation via certain quasi-free modules,' a reader applying the theorem to non-simple valuated matroids (which occur naturally) will get an incorrect statement. This is an internal inconsistency between the introductory claim and the counterexample the paper itself provides, not merely an omitted reference. The V_M-based correspondence in Corollary 6.10(1) appears to survive without simplicity — its proof uses Theorem 6.7 and Theorem 5.9(c), neither of which needs the full simple hypothesis for loopless matroids — so the reader's claim that simplicity is load-bearing for Corollary 6.10(1) is not supported. The load-bearing issue is specifically the quasi-free-module variant, Theorem E(2)/Prop 6.13, as stated in the abstract and introduction.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies modules over semirings, focusing on weakly free and quasi-free modules, and applies them to tropical representations and valuated matroidal representations. The main advertised result (Theorem E / Corollary 6.10) states that for a finite group G and a valuated matroid M, isomorphism classes of tropical subrepresentations whose underlying tropical linear space is isomorphic to the tropical linear space V_M correspond bijectively to weak isomorphism classes of weak G-actions on M, and that the image of Aut(Q_M) under the natural map to S_n is the weak automorphism group Aut_w(M). The paper also proves structure theorems for automorphism groups of weakly free modules and polyhedral cones, and classifies linear subgroups of (T^×)^n and (R_{≥0}^×)^n.","tokens_in":33637,"tokens_out":10192,"duration_ms":94121,"significance":"If the main correspondence is taken with the appropriate hypotheses, the paper provides a valuable generalization of Giansiracusa–Manaker's matroidal representations to the valuated setting, and it introduces a genuinely useful tool: the quasi-free module Q_M gives an embedding-independent way to study the intrinsic module-theoretic structure of a tropical representation. The proof of the correspondence for V_M (Theorem 6.7) is detailed and appears coherent, and the paper contains several useful structural results on automorphism groups, including the semidirect-product decomposition in Theorem 5.9 and the cone version in Corollary 5.11. The paper is also careful to give explicit examples, including Example 6.16 showing that weak automorphism groups can be proper subgroups of ordinary automorphism groups. However, the abstract and introduction state Theorem E(2) without the simplicity hypothesis that is used in the proof and is in fact necessary, as the paper's own Example 6.14 shows; this is a load-bearing overclaim that must be fixed before publication.","major_comments":[{"comment":"Theorem E(2) is stated for arbitrary valuated matroids, but the proof of Proposition 6.13 uses the standing simplicity assumption stated at the beginning of Section 6, and the statement is false without it. In the converse direction of Proposition 6.13, the map f(e_i)=τ_i^{-1}e_{σ(i)} is well-defined and invertible only when τ_i≠0_T, which the proof justifies by simplicity. Example 6.14 then explicitly shows that for the rank-1 uniform valuated matroid U_{1,3}, Q_M collapses to T, so the image of Aut(Q_M) in S_3 is trivial while Aut_w(U_{1,3})=S_3. Thus Theorem E(2) as printed in the abstract and introduction is false. The statement should be restricted to simple valuated matroids, or the definition of weak automorphism and the proof must be adapted to handle loops and parallel elements. I note that this failure does not appear to affect Theorem E(1)/Corollary 6.10(1), whose proof relies on Theorem 6.7 rather than on Proposition 6.13; the paper should make this distinction explicit.","section":"Abstract and §1.1, Theorem E; §6, Proposition 6.13; Example 6.14"},{"comment":"Case (b) of Theorem 5.9 is stated for an arbitrary weakly free module of rank n over R_{≥0}, but the proof invokes Lemma 4.6, which requires the hypothesis that M can be embedded into a free module. That embeddability hypothesis is present in case (a) but not in case (b). Lemma 4.6 also assumes finite presentation when a finite defining system is needed, and this finiteness condition is likewise absent from case (b). Since Theorem 5.9 is a central structural result and is used in Corollary 5.11, the statement should either add the missing hypotheses or give a proof that every weakly free R_{≥0}-module satisfies them.","section":"§5, Theorem 5.9(b), and proof of Theorem 5.9(1) using Lemma 4.6"},{"comment":"The statement and proof of Corollary 6.10 use the phrases 'isomorphism classes of tropical subrepresentations whose underlying tropical linear space is isomorphic to V_M' and 'isomorphism classes of homomorphisms G→Aut(V_M)' without defining the relevant notion of isomorphism for embedded tropical linear spaces or for subrepresentations. Earlier in the paper, Definition 5.8 defines equivalence of G-actions only up to conjugation by diagonal elements of Aut(V). Please state explicitly whether an isomorphism between tropical linear spaces is required to be induced by an element of GL_n(T), and prove that the reduction in the first sentence of the proof of Corollary 6.10 is valid under that definition.","section":"§6, Corollary 6.10 and its proof"}],"minor_comments":[{"comment":"In the definition of weakly free module, the second minimal generating set is written '{y_n,\\ldots,y_n}'; it should be '{y_1,\\ldots,y_n}'.","section":"§3, Definition 3.1"},{"comment":"The introduction labels Theorem E as '(Corollary 6.10)', but Corollary 6.10 contains only the V_M-based correspondence; part (2) of Theorem E is Proposition 6.13. The cross-reference should be corrected.","section":"§1.1, Theorem E cross-reference"},{"comment":"The sentence 'the image of Aut(Q_M) in any finite group is trivial' is awkward and imprecise; it should say that the image of Aut(Q_M) under π: GL_n(T)→S_n is trivial.","section":"§6, Example 6.14"},{"comment":"The proof of Lemma 6.12 says it suffices to show that the set of nonzero coefficients contains a circuit, but this is only enough because the standing simplicity assumption guarantees that every circuit has size at least three. The proof should state this explicitly, especially since Example 6.14 shows the failure without simplicity.","section":"§6, Lemma 6.12"},{"comment":"The blanket assumption 'We will assume that all matroids are simple unless otherwise stated' should be recalled in the abstract and in Theorem E, since the abstract and theorem are currently unqualified and therefore overstate the scope of the quasi-free-module result.","section":"§6, opening paragraph"}],"recommendation":"major_revision","confidential_remarks":"The main issue is not a subtle algebraic error but a mismatch between the advertised scope and the proven statements: Theorem E(2) is false without the simplicity hypothesis, as the paper's own Example 6.14 demonstrates. The V_M-based correspondence in Corollary 6.10(1) appears sound, and the fix is local (add the hypothesis, adjust the abstract and introduction, and clarify the cross-references). I therefore recommend major revision rather than rejection. I saw no evidence of circularity or of dependence on the paper's own prior results in the central argument."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core of this paper is the valuated-matroid version of Giansiracusa–Manaker: for finite G, tropical subrepresentations whose underlying tropical linear space is V_M correspond to weak G-actions on the valuated matroid M. Corollary 6.10 is new, and the proof through Theorem 6.7 is careful and, as far as I can tell, correct. The classification of linear subgroups of the torus (Proposition 4.10) is also a tidy new piece, and the weakly free / quasi-free module framework gives a genuinely useful language for automorphism groups of tropical linear spaces.\n\nThe soft spots are real but local. Theorem E(2) / Proposition 6.13, as stated in the abstract and introduction, claims that the image of Aut(Q_M) in S_n equals Aut_w(M) for every valuated matroid. It is proved only under the standing simplicity assumption in Section 6, and the paper's own Example 6.14 shows it fails without it: for U_{1,3}, Q_M collapses to T, the image is trivial, while Aut_w is S_3. That is an internal inconsistency between the advertised claim and the counterexample the authors themselves provide, not a missing reference. It should be fixed by stating the simplicity hypothesis in Theorem E and the abstract, or by proving a version that handles non-simple matroids.\n\nThe reader's report flagged simplicity as load-bearing for Corollary 6.10(1) as well. I disagree with that part. The proof of Corollary 6.10 uses Theorem 6.7 and Theorem 5.9(c), neither of which requires the full simplicity hypothesis for loopless matroids; the failure is specifically in the quasi-free module Q_M. So the stress-test note is right on that point.\n\nA second, smaller issue: Theorem 5.9(b), for weakly free modules over R_{\\ge0}, is stated without an embeddability hypothesis, but the proof invokes Lemma 4.6, which requires embeddability into a free module. Either the hypothesis should be added or a separate argument supplied. This looks fixable.\n\nOverall, the paper is a solid contribution to tropical representation theory. The main equivalence is substantial, the proofs are mostly detailed and honest, and the flaws are overstatements in the presentation rather than deep conceptual errors. The authors even include the counterexample that exposes the overclaim in Theorem E(2), so they are aware of the edge case. With added hypotheses, the paper would be in good shape.\n\nI would send this to peer review. It deserves a serious referee who can push on the missing hypotheses and check whether the Q_M story can be repaired for non-simple matroids. For researchers in tropical geometry and matroid theory, the valuated matroidal representation correspondence will be worth citing regardless.","headline":"Valuated matroidal representations are a real advance, but Theorem E(2) on quasi-free modules overreaches and needs a simplicity hypothesis; fixable and worth refereeing.","tokens_in":34217,"tokens_out":3444,"would_cite":true,"duration_ms":30705,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["12K10","14T10","05B35","05E10"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that for a finite group $G$, isomorphism classes of tropical subrepresentations whose underlying tropical linear space is isomorphic to $V_M$ correspond one-to-one with weak isomorphism classes of weak $G$-actions on the…","keywords":["matroid","valuated matroid","representation","tropical geometry","tropical representation","tropical linear space","weakly free module","quasi-free module"],"falsifier":"Compute the rank-1 uniform valuated matroid on three elements: every permutation lies in $\\operatorname{Aut}_w(M)=S_3$, yet the bend relations identify all three generators, so $Q_M\\cong\\mathbb{T}$ and the image of $\\operatorname{Aut}(Q_M)$ in $S_3$ is trivial; this demonstrates exactly where the simple-matroid hypothesis is needed, and a simple matroid whose $Q_M$ fails to be quasi-free would refute Lemma 6.12.","tokens_in":33096,"feed_emoji":"🌴","tokens_out":11691,"duration_ms":91996,"temperature":0.7,"pith_summary":"This paper tries to establish that representation theory over the tropical semifield $\\mathbb{T}$ is a combinatorial subject. Its central result is a one-to-one correspondence: for a finite group $G$ and a valuated matroid $M$, isomorphism classes of tropical subrepresentations whose underlying tropical linear space is isomorphic to $V_M$ are exactly the weak isomorphism classes of weak $G$-actions on $M$. This matters because a problem about groups acting on infinite geometric spaces becomes a problem about symmetries of a finite matroid. Along the way the authors introduce weakly free and quasi-free modules over semirings, show that the coordinate module $Q_M$ of a tropical linear space is quasi-free and recovers $V_M$ by dualizing, and identify the automorphism group of $V_M$ as a semidirect product of diagonal scalars and the weak automorphism group of $M$.","feed_headline":"Finite tropical representations are weak actions on valuated matroids","feed_subtitle":"Finite tropical group actions are weak actions on valuated matroids, so the theory becomes combinatorial.","key_machinery":"The central object is the quotient module $Q_M=\\mathbb{T}^n/{\\sim}$, the coordinate module of the tropical linear space $V_M$, obtained by imposing bend relations for every subset of size $d+1$. The module $Q_M$ is quasi-free: it has a quasi-basis, a minimal generating set in which any relation $x_i=\\sum_j c_jx_j$ forces $c_j=\\delta_{ij}$, and for the semirings here quasi-free implies weakly free, so automorphisms act by permuting and rescaling the quasi-basis. Dualizing gives $Q_M^*\\cong V_M$, which lets automorphism questions for $V_M$ be translated into automorphism questions for $Q_M$. The second workhorse is Theorem 5.9: for a weakly free $\\mathbb{T}$-module or a finitely presented $\\mathbb{T}$-linear space, $\\operatorname{Aut}(M)$ is a semidirect product $H \\rtimes V$, where $H\\subseteq S_n$ is the image of the projection and $V$ is a partition subspace (vectors whose coordinates are constant on each block of an equivalence relation). Because $H^i(G,V)=0$ for finite $G$ acting on such rational subspaces, finite group actions are classified by homomorphisms $G\\to H$.","core_discovery":"For a valuated matroid $M$ on $[n]$ with rank $d$, let $V_M$ be the tropical linear space in $\\mathbb{T}^n$ cut out by the bend relations of the circuits. The paper proves that the image of $\\operatorname{Aut}(V_M)$ under the natural projection $\\operatorname{GL}_n(\\mathbb{T})\\to S_n$ is exactly the weak automorphism group $\\operatorname{Aut}_w(M)$: those permutations $\\sigma$ for which there is a map $\\tau:[n]\\to\\mathbb{T}$ with $w(\\sigma(B))=(\\prod_{i\\in B}\\tau(i))w(B)$ for every basis $B$ (Theorem 6.7). It then proves that for finite $G$, isomorphism classes of tropical subrepresentations with underlying space isomorphic to $V_M$ correspond one-to-one with weak isomorphism classes of weak $G$-actions on $M$ (Corollary 6.10). The proof works through the quotient module $Q_M=\\mathbb{T}^n/{\\sim}$ obtained from the bend relations: $Q_M$ is quasi-free, its dual is $V_M$, and $\\operatorname{Aut}(Q_M)$ projects onto $\\operatorname{Aut}_w(M)$. This generalizes the earlier matroidal-representation construction over the Boolean semifield and makes the correspondence independent of the chosen embedding of $V_M$.","pith_inferences":["The authors leave implicit that $Q_M$ could serve as an intrinsic starting point for tropical representation theory, letting one define $G$-actions on the coordinate module rather than on an embedded tropical linear space.","The finite-group hypothesis is likely not merely technical: for infinite groups such as $\\mathbb{T}^{\\times}$, scalar actions already produce tropical subrepresentations with no matroid counterpart, so a full infinite theory would need to track the cohomology of the diagonal part.","A testable extension would repair the simple-matroid assumption by modifying the quotient $Q_M$ to keep loops and parallel elements as distinguished generators, potentially restoring the correspondence for all valuated matroids.","The partition-subspace decomposition suggests that equivariant tropical geometry reduces to representation theory of the weak automorphism group, so questions about $G$-invariant tropical linear spaces could be attacked by first classifying subgroups of $\\operatorname{Aut}_w(M)$."],"forward_implications":["Finite group actions on tropical linear spaces can be studied through finite combinatorial data: weak automorphisms of the underlying valuated matroid.","The automorphism group of a tropical linear space splits as a semidirect product of diagonal rescalings and the weak automorphism group, so symmetries decompose into permutations and coordinate-wise scaling.","Because $Q_M$ is defined intrinsically from the matroid and its dual recovers $V_M$, the correspondence does not depend on how the tropical linear space is embedded in $\\mathbb{T}^n$.","For finite $G$, every linear action on $V_M$ is equivalent to one that permutes a quasi-basis of $Q_M$ up to scalars, and non-isomorphic actions are detected by homomorphisms $G\\to\\operatorname{Aut}_w(M)$."],"supporting_citations":[{"why":"Supplies the description of tropical linear spaces generated by vectors $v_I$ for corank-one independent sets, used to show $V_M$ is weakly free.","marker":"[Fre13]"},{"why":"Proves that the dual of $Q_M$ is $V_M$, linking the quotient module to the tropical linear space.","marker":"[GG18]"},{"why":"Introduces tropical subrepresentations and matroidal representations over the Boolean semifield; this paper extends that notion to valuated matroids.","marker":"[GM20]"},{"why":"Defines valuated matroids and projective equivalence of valuations, which underlies the definition of weak automorphisms.","marker":"[DW92]"},{"why":"Proves that finitely generated submodules of free tropical modules are weakly free; used for $V_M$ as a module.","marker":"[Wag91]"},{"why":"Proves the split exact sequence describing $\\operatorname{GL}_n(R)$ as permutations plus diagonal units, used for the projection to $S_n$.","marker":"[JMT23]"},{"why":"Supplies the lifting of actions on weakly free modules to free modules and the dualization tools used in Section 6.","marker":"[JMT24]"},{"why":"Provides the vanishing of group cohomology $H^k(G,V)$ for finite $G$, used to split automorphism groups and classify actions.","marker":"[Bro12]"}],"fun_headline_variants":["Tropical subreps are weak actions on valuated matroids","Valuated matroids classify tropical subrepresentations","Tropical representations reduce to matroidal weak actions","Quasi-free modules unlock tropical subrepresentations","Weak actions on valuated matroids are tropical subreps"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The correspondence in Corollary 6.10 is proved under the standing assumption that all matroids in Section 6 are simple and that $G$ is finite; dropping simplicity makes $Q_M$ collapse (the rank-1 uniform matroid gives $Q_M\\cong\\mathbb{T}$), and dropping finiteness creates scalar actions with no matroid counterpart.","fun_headline_variants_meta":{"raw":{"variants":["Tropical subreps are weak actions on valuated matroids","Valuated matroids classify tropical subrepresentations","Tropical representations reduce to matroidal weak actions","Quasi-free modules unlock tropical subrepresentations","Weak actions on valuated matroids are tropical subreps"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00107,"raw_usage":{"total_tokens":4477,"prompt_tokens":935,"completion_tokens":3542,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":551,"completion_tokens_details":{"reasoning_tokens":3462}},"tokens_in":551,"tokens_out":3542,"duration_ms":21483,"temperature":1.0,"reasoning_tokens":3462,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T05:46:27.370548+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the rank-1 uniform valuated matroid on three elements: every permutation lies in $\\operatorname{Aut}_w(M)=S_3$, yet the bend relations identify all three generators, so $Q_M\\cong\\mathbb{T}$ and the image of $\\operatorname{Aut}(Q_M)$ in $S_3$ is trivial; this demonstrates exactly where the simple-matroid hypothesis is needed, and a simple matroid whose $Q_M$ fails to be quasi-free would refute Lemma 6.12.","supporting_citations":[],"review_version":1}