{"id":"ca66883d-379e-4839-bcdf-9430eb7811cb","arxiv_id":"2507.12829","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Monodromy of Bethe eigenlines over real loci of cactus flower moduli spaces equals the virtual, mirabolic, and affine cactus group actions on tensor products of Kashiwara crystals.","lead":"This paper computes monodromy of Bethe eigenvectors for inhomogeneous and trigonometric Gaudin models over two real moduli loci and matches it with cactus group actions on crystals. It also proves a categorical equivalence between certain covering spaces of cactus flower moduli spaces and concrete coboundary monoidal categories.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Compact trigonometric monodromy (Thm 1.3(3)) rests on semisimplicity theorem 6.10(2), quoted from [IKR, Thm 8.15] and not reproved; Section 6.4 only arranges its hypothesis.","rationale":"I agree with the reader's identification of the weakest assumption. The compact trigonometric monodromy theorem (Theorem 1.3(3)) is a central advertised result, and its proof in Section 7.3 depends on the family E^comp_χ being a covering, which requires simple spectrum. That simple spectrum is obtained by combining cyclicity (Theorem 6.9) with the semisimplicity stated in Theorem 6.10(2), quoted from the authors' earlier preprint [IKR] rather than proved here. The ε-dependent parameter χ+ε/2µ in Section 6.4 is a legitimate reduction—it makes the hypothesis of Theorem 6.10(2) hold for every ε∈iR—but it does not replace the proof of that theorem. Thus the concern is about a dependency, not an internal contradiction. The inhomogeneous monodromy theorem 1.3(1) appears well-supported: Theorem 7.2 uses the published [HKRW] equivalence and the agreement of the monoidal structure maps φ and ψ, and Corollary 7.3 then transfers the vC_n action. The categorical equivalence Theorem 1.2 is plausible, though Proposition 5.4's extension over strata is asserted briefly. A second unverified input is Theorem 4.6 from the unpublished [GHR], which is needed for the split-case statements; I regard this as related but secondary to the semisimplicity dependency because the compact case would fail to be defined if the quoted semisimplicity failed. Overall, conditional acceptance is the right verdict: the paper should be accepted only once the quoted semisimplicity theorem and the [GHR] input are verifiable.","tokens_in":32211,"tokens_out":23743,"duration_ms":248661,"concrete_test":"Independently re-derive the semisimplicity statement of [IKR, Thm 8.15] for the case needed here: χ(ε) - ε/2 µ ∈ h_split, C ∈ F^comp_n, acting on V(λ)_µ. If the re-derivation goes through using only the degeneration to F_n(R) and the published Feigin-Frenkel-Rybnikov semisimplicity theorem, the quoted input is confirmed; if it requires an additional hypothesis (e.g., ε in a finite interval, or C away from a divisor), then Theorem 7.7 must be revised or the compact-case monodromy action marked conditional.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing premise for Theorem 7.7 is Theorem 6.10(2): for χ - ε/2 µ ∈ h_split and C ∈ F^comp_n, the algebra A^ε_χ(C) acts semisimply, hence (with cyclicity from Theorem 6.9) with simple spectrum on V(λ)_µ. This is exactly what makes E^comp_χ(λ)_µ → F^comp_n(-c,c) an unbranched covering, so the gAC_n monodromy action in Theorem 1.3(3) is defined. The paper quotes this result from [IKR, Thm 8.15] and does not reprove it. Section 6.4 sets χ(ε) = χ + ε/2 µ so that the hypothesis reduces to χ ∈ h_split, but that only invokes the quoted theorem; it does not supply its proof. If [IKR, Thm 8.15] has an unstated condition or is false, the covering E^comp_χ and the action of r and s_ij in Theorem 7.7 are not established. I found no internal inconsistency in the proof of Theorem 1.3(1); its external inputs are the published [HKRW] results, so the central inhomogeneous claim is on firmer ground than the compact trigonometric claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies the interplay between operadic coverings of the real cactus flower moduli spaces F_n(R) and concrete coboundary monoidal categories. The main categorical result (Theorem 1.2) establishes an equivalence between the category of Ξ-coloured operadic coverings of these moduli spaces and the category of Ξ-coloured concrete coboundary categories. The authors then apply this to Gaudin models: for a complex semisimple Lie algebra g, they show (Theorem 7.2) that the coverings formed by eigenlines of inhomogeneous Gaudin algebras yield a concrete coboundary category equivalent to the category of Kashiwara g-crystals. From this they deduce that the monodromy action of the virtual cactus group vC_n on Bethe eigenlines matches the action on tensor products of crystals (Theorem 1.3(1)), and similarly for the mirabolic and extended affine cactus groups in the split and compact trigonometric cases (Theorems 1.3(2),(3)). A connection to a combinatorial version of the Bezrukavnikov–Okounkov wall-crossing conjecture is discussed for minuscule highest weights.","tokens_in":32393,"tokens_out":12765,"duration_ms":126893,"significance":"This is a substantial contribution to the monodromy of Bethe vectors and its relation to crystals. The categorical equivalence of Theorem 1.2 is a clean and general statement that unifies the operadic covering perspective with concrete coboundary categories, extending earlier work of Halacheva–Kamnitzer–Rybnikov–Weekes. The monodromy computations for the virtual and mirabolic cactus groups are new and provide concrete realizations of these groups. The compact trigonometric case, if fully justified, gives a combinatorial avatar of the wall-crossing conjecture for minuscule slices, which is of independent interest. The proofs are detailed and make effective use of prior work on cactus flower spaces and Gaudin algebras; the main theorems are supported by explicit homotopies and cube decompositions, e.g., the proof of Theorem 4.9 and Proposition 5.2.","major_comments":[{"comment":"The covering property of E^{comp}_χ(λ)_µ over F^{comp}_n(-c,c) requires simultaneous cyclicity and semisimplicity of the algebras A^ε_{χ+ε/2 µ}(C) for all C in this neighborhood. While semisimplicity is reduced to Theorem 6.10(2), cyclicity is asserted to follow from Theorem 6.9, which is stated only for a fixed χ. In the compact case, the parameter χ is replaced by χ(ε) = χ + ε/2 µ, so Theorem 6.9 does not apply verbatim. The authors should supply a proof that the exceptional set of ε for which cyclicity fails has no accumulation at 0 uniformly in C (or else restrict the family differently), because without cyclicity the union of eigenlines is not a covering space and the monodromy action is undefined.","section":"Section 7.3, proof of Theorem 7.7"},{"comment":"The semisimplicity assertion for the compact real form is quoted as [IKR, Thm 8.15] and is not reproved. This result is load-bearing: it is exactly what makes E^{comp}_χ(λ)_µ → F^{comp}_n(-c,c) a covering space, and hence it underlies the entire gAC_n monodromy action in Theorem 1.3(3). Since [IKR] is a preprint, the referee cannot verify the theorem from the manuscript. Please state the theorem with all hypotheses explicitly and either include a proof or give a precise indication of where it is proved in [IKR], so that the compact case is self-contained to the extent possible.","section":"Section 6.4, Corollary 6.12 and its use in Theorem 7.7"}],"minor_comments":[{"comment":"The text refers to \"α_1^2 : F_2 → P^1\" as an isomorphism; this is likely a typo for the function δ_{12}. Please check and correct.","section":"Section 5.4"},{"comment":"The word \"domianant\" should be \"dominant\".","section":"Theorem 6.9"},{"comment":"The sentence \"we combine Theorem 4.11 and Theorem 7.3\" should refer to Corollary 7.3, since the preceding result is Corollary 7.3 rather than Theorem 7.3.","section":"Section 7.3"},{"comment":"The notation for the compact-family fibers is inconsistent: E^{comp}_χ(C, λ)_µ versus E_χ(λ)_µ. Please unify the notation.","section":"Section 7.3"},{"comment":"A small figure of the five boundary components of F_3(R)^+ would improve readability, though the verbal description is sufficient.","section":"Proposition 5.2"}],"recommendation":"major_revision","confidential_remarks":"The paper relies heavily on the companion preprint [IKR] for several key results, including the semisimplicity theorem that underpins the compact trigonometric monodromy. I recommend that the handling editor ensure that [IKR] is available and explicitly referred to with theorem numbers; if it is not yet peer-reviewed, the authors should be asked to include the necessary proofs or to make the dependence transparent. The core categorical and inhomogeneous monodromy results appear sound and are proved in detail."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: this is a real step forward, not a repackaging. Theorem 1.2, the equivalence between operadic coverings and concrete coboundary categories, is genuinely new, and the three monodromy statements in Theorem 1.3 settle the conjecture from [IKR]. If you work on Bethe ansatz, crystals, or real loci of Gaudin systems, this paper is worth your time.\n\nWhat the paper does well: the construction in Sections 5.3–5.5 is careful, and the proof of the pentagon via the cube decomposition of F_3(R) is explicit enough to be checked. The main inhomogeneous monodromy result (Corollary 7.3) is on firmer ground, because it relies on the published [HKRW] results plus the new equivalence, and the argument reads coherently. The paper is also honest about what it quotes and what it proves.\n\nNow the soft spots, in proportion. First, Theorem 7.7, the compact trigonometric monodromy, needs semisimplicity of A^ε_χ(C) on V(λ)_μ. That is exactly [IKR, Thm 8.15], quoted and not reproved. Section 6.4 only arranges the hypothesis, it does not supply the proof. If that theorem has an unstated condition or is false, the covering and the gAC_n action are not established. I see no internal flaw in the paper's own argument, but this is a load-bearing external dependency. Second, Theorem 4.6, which describes the generators of the mirabolic cactus group, comes from [GHR], a paper 'in preparation' that is not public. This is also load-bearing for Theorem 7.6. A referee needs to see that proof before the split-case monodromy can be fully certified. Finally, the pentagon verification in Proposition 5.2 is compressed, though it looks checkable and the strategy is clear.\n\nWho this is for: people in quantum integrable systems, crystal theory, and geometric representation theory. It deserves a serious referee. My recommendation: send it to review, but ask the authors to either make the [GHR] input public or include the needed statements as an appendix, and ask one referee to verify [IKR, Thm 8.15] carefully. If those two items hold up, the paper is solid.","headline":"Serious extension of the cactus flower/Gaudin program; the inhomogeneous monodromy theorem is well supported, but the compact trigonometric version leans on a quoted semisimplicity theorem and the split version depends on an unpublished input.","tokens_in":33055,"tokens_out":1732,"would_cite":true,"duration_ms":20151,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["17B10","05E10","17B37","14N35","81R12"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that the monodromy of Bethe eigenlines for Gaudin models is captured exactly by crystal commutors, naive permutations, and cyclic rotations.","keywords":["cactus flower space","Gaudin model","Bethe vectors","monodromy","Kashiwara crystals","coboundary categories","virtual cactus group","affine Grassmannian wall-crossing"],"falsifier":"Pick $\\mathfrak{g} = \\mathfrak{sl}_2$, $n=3$, all $\\lambda_i$ the standard representation, and trace the joint eigenlines of the inhomogeneous Gaudin algebra along the generating loops of $\\mathrm{v}C_3$ in $\\overline{F}_3(\\mathbb{R})$; if the resulting permutation of lines in $V(\\lambda)_\\mu$ differs from the action of $\\mathrm{v}C_3$ on $B(\\omega_1)^3$ by Schützenberger commutors and factor permutations, Theorem 1.3(1) is false. The compact-form statement would be falsified more directly by any point $C \\in \\overline{F}^{\\mathrm{comp}}_n$ where the trigonometric Gaudin algebra has a repeated eigenvalue on a weight space, since the eigenline covering would be branched there.","tokens_in":31918,"feed_emoji":"🌵","tokens_out":11985,"duration_ms":113414,"temperature":0.7,"pith_summary":"The paper aims to show that the monodromy of Bethe eigenlines in Gaudin models is a purely combinatorial phenomenon, controlled by Kashiwara crystals. It establishes a one-to-one correspondence between operadic coverings of the real cactus flower spaces $\\overline{F}_n(\\mathbb{R})$ and concrete coboundary monoidal categories, then proves that the covering built from inhomogeneous Gaudin eigenlines corresponds exactly to the category of normal $g$-crystals. From this, it derives that the virtual cactus group $\\mathrm{v}C_n$ acts on the eigenlines by crystal commutors together with naive permutations of tensor factors, and that the mirabolic and extended affine cactus groups act in the trigonometric models by the same crystal commutors with, respectively, permutation and cyclic-rotation effects. The upshot is a uniform combinatorial description of how Bethe vectors behave when parameters wind around real loci, with a direct bearing on the wall-crossing conjecture of Bezrukavnikov and Okounkov in the minuscule case.","feed_headline":"Gaudin eigenlines follow crystal commutors and permutations","feed_subtitle":"Bethe-line monodromy is shown to match Kashiwara crystals, giving a combinatorial window on wall-crossing.","key_machinery":"The mechanism carrying the argument is the equivalence between $\\Xi$-coloured operadic coverings of the real cactus flower spaces and $\\Xi$-coloured concrete coboundary monoidal categories. An operadic covering is a family of covering spaces over $\\overline{F}_n(\\mathbb{R}) \\times \\Xi^n$ and $\\overline{M}_{n+1}(\\mathbb{R}) \\times \\Xi^{n+1}$, together with gluing isomorphisms coming from the operadic maps $\\alpha_k$, $\\beta_k$, $\\gamma_k$ that describe how marked points bubble off at infinity or at other marked points. A concrete coboundary monoidal category is a monoidal category with an involutive commutor $\\sigma_{A,B} : A\\otimes B \\to B\\otimes A$ and a faithful monoidal functor to sets; the commutors on $n$-fold tensor products generate an action of the cactus group, and the faithful functor adds the naive permutation action of the symmetric group, together giving the virtual cactus group action. Bethe eigenlines of the inhomogeneous Gaudin algebras assemble into such an operadic covering, and the equivalence identifies this covering with the category of normal Kashiwara crystals, whose commutor is built from the Schützenberger involution. Once this identification is in place, the monodromy of the eigenlines is forced to match the crystal commutor and permutation action on products of crystals.","core_discovery":"The central discovery is that the whole family of joint eigenlines of the inhomogeneous Gaudin algebra $A_\\chi(C)$ acting on $V(\\lambda)_\\mu$, as $C$ ranges over the cactus flower space $\\overline{F}_n(\\mathbb{R})$, forms an operadic covering whose associated concrete coboundary category is the category of Kashiwara $g$-crystals. Therefore the monodromy of this covering under $\\mathrm{v}C_n = \\pi_1^{S_n}(\\overline{F}_n(\\mathbb{R}), \\infty)$ agrees with the action of $\\mathrm{v}C_n$ on $B(\\lambda_1) \\times \\cdots \\times B(\\lambda_n)$ by crystal commutors and naive permutation of tensor factors. For the trigonometric Gaudin model, the same covering, restricted to the split and compact real loci, yields monodromy actions of the mirabolic cactus group $\\mathrm{MC}_n$ and the extended affine cactus group $\\mathrm{gAC}_n$; these actions factor through $\\mathrm{v}C_n$, with the cyclic generator $r$ acting as cyclic rotation of the tensor factors. This is the theorem conjectured in the authors' previous work and proved here by combining the covering/category equivalence with degeneration of trigonometric Gaudin algebras to inhomogeneous ones.","pith_inferences":["Beyond the paper, the equivalence suggests that the monodromy of any operadic covering of $\\overline{F}_n(\\mathbb{R})$ can be read off from the underlying coboundary category, so crystal commutors could in principle be reconstructed from observed monodromy rather than from a chosen crystal model.","The paper's generalization remarks point toward analogous statements for braided monoidal categories, where the virtual cactus group would be replaced by the virtual braid group; such an extension would give a topological description of R-matrix monodromy for quantum group Bethe vectors, which the paper does not prove.","The minuscule wall-crossing interpretation implies a concrete test: the monodromy permutations computed here should coincide with the wall-crossing functors for these slices, a comparison the paper leaves as a conjecture.","The compact-form trigonometric result depends on a semisimplicity input taken from earlier work, so a failure of that input would affect mainly the compact trigonometric monodromy statement rather than the rational inhomogeneous result."],"forward_implications":["For any dominant weights $\\lambda_1,\\dots,\\lambda_n$, the covering of inhomogeneous Gaudin eigenlines over $\\overline{F}_n(\\mathbb{R})$ is isomorphic to the combinatorial cover with fibre $B(\\lambda_1)\\times\\cdots\\times B(\\lambda_n)$, so the virtual cactus group acts on eigenlines exactly as on crystal tensor products.","The monodromy of trigonometric Gaudin eigenlines over the split real locus factors through $\\mathrm{v}C_n$, with the generators $t_i$ acting as elementary transpositions of the tensor factors.","The monodromy of trigonometric Gaudin eigenlines over the compact real locus factors through $\\mathrm{v}C_n$, and the extended affine cactus group acts by crystal commutors together with cyclic rotation of tensor factors.","The equivalence of categories means that every $\\Xi$-coloured operadic covering of $\\overline{F}_n(\\mathbb{R})$ determines a concrete coboundary category, and conversely every such category produces such a covering; in particular, the Gaudin covering recovers the whole category of normal $g$-crystals.","For sums of minuscule weights, the compact-form monodromy theorem is a combinatorial counterpart of the Bezrukavnikov–Okounkov wall-crossing conjecture for minuscule resolutions of slices in the affine Grassmannian."],"supporting_citations":[{"why":"Supplies the strategy: it proves the homogeneous Gaudin analogue for $\\overline{M}_{n+1}(\\mathbb{R})$ and constructs the $g$-crystal structure on eigenlines used here.","marker":"[HKR W]"},{"why":"Previous paper by the authors that defines the inhomogeneous and trigonometric Gaudin algebras on the cactus flower space and supplies the cyclicity and semisimplicity theorems on which the covering is built.","marker":"[IKR]"},{"why":"Introduced the cactus flower spaces $\\overline{F}_n$, the virtual cactus group, the deformation to $\\overline{M}_{n+2}$, and the fundamental group isomorphisms used to identify monodromy groups.","marker":"[IKLPR]"},{"why":"Established crystals as a coboundary category and the cactus group action generated by commutors, which is the algebraic object matched to the eigenline covering.","marker":"[HK]"},{"why":"Provides the cactus group as the relevant fundamental group of $\\overline{M}_{n+1}(\\mathbb{R})$ and the cube-complex description powering the topological computations.","marker":"[DJS]"},{"why":"Proves semisimplicity of shift-of-argument subalgebras, used to show the inhomogeneous Gaudin algebras act with simple spectrum over $\\overline{F}_n(\\mathbb{R})$.","marker":"[FFRy]"},{"why":"Extended homogeneous Gaudin algebras to $\\overline{M}_{n+1}$ and initiated the monodromy-of-Bethe-vectors program for the cactus group.","marker":"[R1]"},{"why":"Defines the mirabolic cactus group and the generators $t_i$ used in the split trigonometric monodromy theorem.","marker":"[GHR]"}],"fun_headline_variants":["Bethe line monodromy matches crystal commutors","Gaudin eigenlines realize crystal commutors","Monodromy of Bethe vectors equals crystal actions","Cactus flower monodromy follows crystal commutors","Bethe lines mirror Kashiwara crystal actions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that over the compact real form the trigonometric Gaudin algebra always splits the relevant weight space into distinct one-dimensional eigenspaces; this semisimplicity-with-simple-spectrum fact is taken from the authors' previous paper, not proved here.","fun_headline_variants_meta":{"raw":{"variants":["Bethe line monodromy matches crystal commutors","Gaudin eigenlines realize crystal commutors","Monodromy of Bethe vectors equals crystal actions","Cactus flower monodromy follows crystal commutors","Bethe lines mirror Kashiwara crystal actions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000554,"raw_usage":{"total_tokens":2684,"prompt_tokens":1037,"completion_tokens":1647,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":653,"completion_tokens_details":{"reasoning_tokens":1570}},"tokens_in":653,"tokens_out":1647,"duration_ms":14190,"temperature":1.0,"reasoning_tokens":1570,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T16:38:33.881909+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Pick $\\mathfrak{g} = \\mathfrak{sl}_2$, $n=3$, all $\\lambda_i$ the standard representation, and trace the joint eigenlines of the inhomogeneous Gaudin algebra along the generating loops of $\\mathrm{v}C_3$ in $\\overline{F}_3(\\mathbb{R})$; if the resulting permutation of lines in $V(\\lambda)_\\mu$ differs from the action of $\\mathrm{v}C_3$ on $B(\\omega_1)^3$ by Schützenberger commutors and factor permutations, Theorem 1.3(1) is false. The compact-form statement would be falsified more directly by any point $C \\in \\overline{F}^{\\mathrm{comp}}_n$ where the trigonometric Gaudin algebra has a repeated eigenvalue on a weight space, since the eigenline covering would be branched there.","supporting_citations":[],"review_version":1}