{"id":"3626ee0c-2052-4401-842e-2c20639d469d","arxiv_id":"2608.27284","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper defines noncommutative cluster varieties for every reduced root system and marked surface, and uses new 'Jordan split' groups to coordinate flag configurations, double Bruhat cells, and moduli spaces of local systems.","lead":"The authors construct a common generalization of cluster varieties, called noncommutative cluster varieties, tied to every reduced root system and marked surface. They also introduce Jordan split groups, a broad family of algebraic groups built from Jordan algebras, and use them to give cluster coordinates and positivity tests on flag spaces, groups, and moduli spaces.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Definition 3.37 introduces T(R,{Ji}) by fiat; the cited recognition theorems classify existing R-graded Lie algebras and do not, as stated, prove existence for every compatible family, leaving Theorem 1.10's main input unproved.","rationale":"The reader's weakest_assumption correctly identifies the definition-by-fiat of T(R,{Ji}) as the most load-bearing gap. I read the surrounding pinning and cluster constructions as genuine independent progress: the Jordan pinning, root system calculus, Θ-Bruhat cells, and the noncommutative mutation formulas are developed in detail and would stand even if the existence question were resolved elsewhere. However, Theorem 1.10 is formally about groups whose Lie algebra is T(R,{Ji}); if that object is not proved to exist for every compatible collection, the central parameterization claim lacks its principal input. I considered the alternative concern that the F4 and exceptional cluster cases are explicitly left incomplete, but that only affects a narrower range of the theorem and is secondary to the foundational existence issue. The reader's CONDITIONAL verdict is therefore appropriate, and my stress-test does not move it.","tokens_in":63462,"tokens_out":11927,"duration_ms":120061,"concrete_test":"For the admissible family {J1=J2=R, J3=J(0,5)} of type B3 used in Example 4.2, write down the Lie bracket on T(B3,{R,R,J(0,5)}) explicitly using the generalized Tits construction from [BZ96] and verify the Jacobi identity and the prescribed Jordan factors. If the brackets can be written out and checked directly, the existence concern is resolved for the non-split B_p case; if the only route is to invoke the recognition theorems as a black box, Theorem 1.10 needs an additional explicit existence theorem for T(R,{Ji}).","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 3.8 defines T(R,{Ji}) (Definition 3.37) as 'the (adjoint form of) Lie algebra with R-grading and collection {Ji}' and says the Jordan algebras 'can be glued' to recover the Tits index, but no bracket construction, universal property, or existence theorem is supplied. The recognition theorems in Section 3.3 (Theorems 3.11–3.18) are one-directional: they assume an R-graded Lie algebra already exists and conclude it is a central extension of a model built from Jordan/associative/alternative data. They do not state, and the paper does not prove, that every compatible tuple in Theorem 3.35 arises from an actual Lie algebra. The identification theorem similarly describes restrictions on the factors of an existing algebra, not a converse construction. Since Theorem 1.10 parameterizes local systems for 'G of type R' whose Lie algebra is defined to be T(R,{Ji}), the parameterization has no object if existence fails for some compatible family. Uniqueness up to central extension is also asserted rather than established. The gap is aggravated by the mismatch between the paper's blanket 'field not of characteristic 2 or 3' assumption and the unspecified field hypotheses of the quoted [BM92, BZ96] classification results; if those theorems require characteristic 0, the claimed generality is not justified. This is a load-bearing gap in the main input of the paper, not a dispute with the substantial and largely self-contained pinning and cluster calculus that follows.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs noncommutative cluster varieties A_{R,S} for each reduced root system R and marked surface S, with Jordan-split algebraic groups as the main input. Part 1 develops the structure theory of Lie algebras graded by root systems, classifies which parabolic subgroups induce such gradings via the new Jordan-compatibility condition, and introduces Jordan split groups with Jordan pinnings, Jordan weights, a root system calculus, and R-Bruhat decompositions. Part 2 reviews and extends noncommutative cluster formalism in two forms: polygonal cluster algebras for lower noncommutative rank and grounded wiring networks for higher noncommutative rank. Part 3 constructs cluster-like coordinates on configuration spaces of decorated flags, and Part 4 applies these to Gauss decompositions, double R-Bruhat cells, positivity, and decorated local systems. The central claim, Theorem 1.10, is that Jordan points of A_{R,S} parametrize an open dense set of decorated G-local systems, and that for positively structured groups the positive Jordan points correspond exactly to decorated positive representations of the surface group.","tokens_in":63797,"tokens_out":6172,"duration_ms":64962,"significance":"If the main construction is sound, this paper provides a substantial unifying framework: it simultaneously generalizes the cluster varieties of Fock-Goncharov, Goncharov-Shen, Berenstein-Retakh, and Goncharov-Kontsevich, and it gives a cluster parametrization for all groups admitting a Guichard-Wienhard positive structure. The paper contains genuine and largely self-contained contributions in Part 1: the Jordan-compatibility classification of parabolic-induced root gradings, the detailed Serre-relation checks in the root system calculus, the explicit pinnings of Spin(3,4) and Sp_4, and the concrete classification tables for real Lie algebras with A1-gradings are valuable and appear carefully worked out. The proposed framework also yields explicit positivity tests and a uniform account of the mapping class group action via noncommutative quasi-cluster automorphisms. However, the existence of the underlying Lie algebra T(R,{J_i}) for every compatible family is asserted rather than proved, and this gap directly feeds the main parametrization theorem. The paper's significance would be high if this gap is filled; in its current form the central existence input is not yet established.","major_comments":[{"comment":"Definition 3.37 defines T(R,{J_i}) as 'the (adjoint form of) Lie algebra with R-grading and collection {J_i}' and Section 3.8 asserts that the Jordan algebras 'can be glued' to obtain the Tits index, but no bracket construction, universal property, or converse existence theorem is supplied. The recognition theorems in Section 3.3 (Theorems 3.11–3.18) have the direction: if an R-graded Lie algebra exists, then it is a central extension of a model built from Jordan data; they do not state that every compatible tuple in Theorem 3.35 arises from an actual Lie algebra. Since Theorem 1.10 parameterizes local systems for 'G of type R' whose Lie algebra is defined to be T(R,{J_i}), a compatible family for which existence fails would leave the main object of the parameterization undefined. Please provide a construction of T(R,{J_i}) with the promised R-grading and prescribed Jordan factors, or cite and state precisely a converse existence theorem from Benkart-Moody or Benkart-Zelmanov, and also prove the asserted uniqueness up to central extension.","section":"§3.8, Definition 3.37"},{"comment":"The paper's blanket hypothesis is that K has characteristic not 2 or 3 (Theorem 1.5), but no field hypotheses are stated for the quoted recognition theorems. If those classifications require characteristic 0, algebraically closed fields, or additional finite-dimensionality assumptions, then the classification of compatible families in Theorem 3.35 and the apparent generality of Theorem 1.10 over arbitrary such fields are not justified. Please state the precise hypotheses under which Theorems 3.11–3.18 apply, and either prove directly that the char-not-2-or-3 assumption suffices or restrict the main statements to the fields for which the quoted classification is valid.","section":"§3.3, Theorems 3.11–3.18; Theorem 1.5"},{"comment":"The amalgamation step, which reduces configurations of n flags to configurations of triples, is described by analogy with the split case and is said to be extended in Theorem 12.2, but the stated dependence of that theorem on the new Jordan-split setting is not proved in the material provided before the statement. Since Theorem 1.10 builds the cluster variety A_{R,S} from these configuration spaces of decorated flags, this amalgamation statement is load-bearing for the surface version of the result. Please give a proof of Theorem 12.2 in the Jordan-split generality, or at minimum state explicitly which of its ingredients are proved in the present paper and which are imported from [GS19] and [GKNW26].","section":"§8.2, Step 2 and Theorem 12.2"}],"minor_comments":[{"comment":"In the displayed formula for the prefix R-Lusztig map, the last factor is written as x_{i_k}(x_k); it should presumably be x_{i_k}(v_k).","section":"Definition 4.42"},{"comment":"In the statement of Lemma 4.35 and its proof, the expression 'a1 j' is not defined; the authors should introduce this notation explicitly or replace it with the precise scalar depending on the connection between nodes i and j.","section":"Lemma 4.35"},{"comment":"The graphical root notation that lists coefficients over the nodes of a Dynkin diagram is used before it is explained; a one-sentence explanation near Example 3.24 would make this notation immediately readable.","section":"Example 3.24 and Figure 3"},{"comment":"Remark 6.7 asserts that the coloring rules determine a unique coloring and refers to [GKNW26]; since this uniqueness is used to construct the polygonal quivers, a short proof or a precise quotation of the relevant statement in [GKNW26] would improve self-containedness.","section":"Section 6.2, Remark 6.7"}],"recommendation":"major_revision","confidential_remarks":"The paper is very ambitious and, if the construction of T(R,{J_i}) can be supplied or properly attributed, it would be a significant contribution to cluster theory and higher Teichmüller theory. The main reason for major revision rather than rejection is that the existence gap is local in principle: it concerns the entry point of the theory rather than the substantial and well-developed root system calculus that follows. The authors should also be asked to verify that the quoted Benkart-Moody and Benkart-Zelmanov classification theorems have exactly the field hypotheses needed for the blanket char-not-2-or-3 assumption. I would not recommend acceptance before these two points are resolved."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe headline is that this is a serious, ambitious paper, and the most novel part—the Jordan split framework and the cluster coordinate machine—is largely credible. But the stress-test note lands: Definition 3.37 introduces T(R,{J_i}) by fiat, and that is a genuine gap that makes the main moduli-space theorem conditional.\n\nWhat is new and good: A_{R,S} genuinely generalizes several cluster varieties (Fock–Goncharov, Goncharov–Shen, Goncharov–Kontsevich, Berenstein–Fomin–Zelevinsky), with the Cp and F4 cases described as new. The Jordan split groups and Jordan pinnings give a clean organizing framework for attaching Jordan algebras to root systems. The classification of Jordan compatible subsets (Theorem 3.28), the Weyl group injections in Section 3.9, and the root system calculus in Section 4 are worked out in detail, with concrete pinnings (SL6, Sp12, Spin(3,4)) that make the claims checkable. These parts look solid. There is no circularity or fitted empirical input, and the self-citations are used appropriately, as building blocks rather than as a substitute for argument.\n\nThe weak spot is exactly where the stress-test note points. The recognition theorems quoted from Benkart–Moody and Benkart–Zelmanov classify existing R-graded Lie algebras; they do not prove that every compatible family in Theorem 3.35 actually occurs. Section 3.8 says the Jordan algebras “can be glued” to recover the Tits index, but no bracket construction, universal property, or existence theorem is supplied. Uniqueness up to central extension is also asserted rather than proved. Since Theorem 1.10 parameterizes G-local systems for G whose Lie algebra is T(R,{J_i}), the parameterization has no object if existence fails for some compatible family. The field hypothesis is also sloppy: the paper assumes characteristic not 2 or 3 throughout, while the quoted classifications may require characteristic 0. These are fixable, but they are load-bearing.\n\nPart 3 is harder to verify from the text: it is a long construction that leans on [GKNW26] and prior work, and the F4/exceptional cases are explicitly not fully explored. That is not a fatal flaw for a paper of this scope, but it means Part 3 reads more like a claim-check than a proof-read from the available text.\n\nNet: this paper deserves a serious referee, not a desk reject. The right recommendation is major revision: add a proof or a precise reference for existence and uniqueness of T(R,{J_i}), state the field hypotheses of the classification theorems cleanly, and either prove the F4 cases or mark them conjectural. If the existence gap is closed, this is a substantial contribution. I would bring it to a reading group working on cluster algebras or higher Teichmüller theory, and I would cite it for the root system calculus and the Jordan pinning framework.\n\nBest,","headline":"An ambitious and largely credible construction of noncommutative cluster varieties whose main input—the existence of the Lie algebras T(R,{J_i})—is asserted, not proved, making Theorem 1.10 conditional.","tokens_in":64288,"tokens_out":3629,"would_cite":true,"duration_ms":37678,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":["13F60","17B70","17C30","20G15","22E46"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper introduces Jordan split groups and constructs a noncommutative cluster variety $\\mathcal{A}_{R,S}$ whose Jordan points parameterize decorated local systems on any marked surface; for positive structures these points are exactly…","keywords":["noncommutative cluster varieties","Jordan split groups","quadratic Jordan algebras","root system grading","decorated local systems","higher Teichmüller spaces","positive representations","double Bruhat cells"],"falsifier":"Take the smallest higher-noncommutative-rank case not previously covered, for example a $C_3$-graded group with $J_1=J_2=M_k(\\mathbb{C})$ and $J_3=H_k(\\mathbb{C})$, pick two triangulations of a punctured surface, compute the same decorated local system's coordinates in both seeds, and check that the coordinate change equals the predicted noncommutative cluster mutation. A mismatch, or a collision of two distinct local systems in the Jordan-point map, would refute Theorem 1.10. A second check is positivity: if a formally real Jordan algebra yields a point that is positive in one seed but not in another, the claimed choice-independence of $\\mathcal{A}_{R,S}^{>0}$ fails.","tokens_in":63266,"feed_emoji":"📐","tokens_out":9212,"duration_ms":80886,"temperature":0.7,"pith_summary":"The paper constructs, for every reduced root system $R$ and every marked surface $S$, a noncommutative cluster variety $\\mathcal{A}_{R,S}$ whose Jordan-algebra points parameterize decorated $G$-local systems on $S$, where $G$ is a newly introduced class of algebraic groups called Jordan split groups of type $R$. This one construction simultaneously extends the cluster varieties used for split groups, the noncommutative cluster algebras built from Jordan algebras, and the polygonal cluster algebras of the authors' earlier work. In the disk case the same points parameterize configurations of decorated flags, and this yields noncommutative cluster structures on the double $R$-Bruhat cells of $G$. When the input Jordan algebras are formally real, the positive Jordan points are shown to be independent of all auxiliary choices and to coincide with the space of decorated positive representations of $\\pi_1(S)$. A sympathetic reader would take the paper's aim to be a unified explicit coordinate system for higher Teichmüller spaces and for positivity in a broad new family of real algebraic groups.","feed_headline":"One cluster variety coordinates local systems for every root type","feed_subtitle":"Jordan split groups unify flag configurations, double Bruhat cells, and positive representations in one cluster structure.","key_machinery":"The load-bearing object is the Jordan pinning of a Jordan split group of type $R$: for each simple root one keeps homomorphisms $SL_2(J_i) \\to G$ and coroot maps $\\check\\beta_i:\\Gamma(J_i)\\to G$, so the structure groups of the quadratic Jordan algebras $J_i$ replace the multiplicative torus. Each $J_i$ is turned into the $A_1$-graded Lie algebra $\\mathfrak{sl}_2(J_i)$ by the Tits-Kantor-Koecher construction, and the recognition theorems of Benkart-Moody and Benkart-Zelmanov (as cited in the paper) classify the compatible families $\\{J_i\\}$ that can be glued into $T(R,\\{J_i\\})$. On the cluster side, the paper defines seed groups—free groups generated by variables attached to quivers or networks, modulo 'angle' relations—and glues them by noncommutative mutation. The mutation of Jordan points is defined by an angle summation formula that uses the quadratic map $\\iota(v)$ to translate sums of vectors in $V$ into products in $\\Gamma(J)$; in higher noncommutative rank the same role is played by grounded wiring networks. These two mechanisms together carry the whole argument.","core_discovery":"The central claim is Theorem 1.10: for a Jordan split group $G$ of type $R$ with associated parabolic $P_\\Theta$, the Jordan points of the cluster variety $\\mathcal{A}_{R,S}$ parametrize an open and dense set of the space of decorated local systems on $S$, and when $G$ carries a positive structure the set of positive Jordan points $\\mathcal{A}_{R,S}^{>0}$ is independent of all choices and corresponds exactly to decorated positive representations of $\\pi_1(S)$. Behind the theorem lies a structural discovery about algebraic groups: a semisimple group can be assembled from copies of $SL_2(J)$ for quadratic Jordan algebras $J$, with the structure groups $\\Gamma(J)$ playing the role of the torus of a split group. This 'Jordan split' viewpoint lets the authors classify which parabolic subgroups induce a root-system grading (Jordan compatibility), build a root-system calculus of commutation and Weyl-group relations, and then repeat the Fock-Goncharov amalgamation strategy with the torus replaced by Jordan structure groups. The result is a noncommutative cluster variety whose mutations add vectors inside the Jordan algebra while multiplying structure-group elements, and whose positive points are defined by positive cones in the Jordan algebras.","pith_inferences":["If the main theorem is correct, the same Jordan-algebra input should produce a noncommutative $X$-cluster variety whose positive points parametrize framed rather than decorated local systems; the paper develops only the $A$-side, so this is a natural extension rather than a claim of the paper.","The dependence on the Jordan split presentation predicts explicit coordinate-change maps between the different cluster structures on the same group arising from different $R$-gradings, for example $SL_6$ as $A_5$, $A_3$, or $A_2$; computing these changes in small examples would test how canonical the new coordinates really are.","Because the positive-structure argument uses only formally real Jordan algebras and their positive cones, the same construction might define higher Teichmüller spaces over other ordered fields, or in $p$-adic settings, by replacing the Euclidean Jordan algebras with suitable positive cones.","The cluster mutation formulas give an algorithm for composing Lusztig parameters under group multiplication; making this algorithm explicit for rank-one $B_p$ and $G_2$ groups would turn the paper's structural result into a computational tool for positive representations."],"forward_implications":["For any marked surface and any Jordan split group of type $R$, the cluster variety $\\mathcal{A}_{R,S}$ gives explicit coordinates on decorated local systems, so holonomies along curves can be written directly in cluster coordinates.","For groups with positive structure, the positive Jordan points $\\mathcal{A}_{R,S}^{>0}$ are choice-independent and coincide with decorated positive representations; each cluster seed yields concrete positivity tests expressed as generalized minors.","On a disk, the construction gives noncommutative cluster structures on double $R$-Bruhat cells $P_\\Theta u P_\\Theta \\cap P_\\Theta^{\\mathrm{opp}} v P_\\Theta^{\\mathrm{opp}}$, generalizing the Berenstein-Fomin-Zelevinsky cluster algebras on double Bruhat cells.","The cluster modular group $\\Gamma_{R,S}$—containing the mapping class group, outer automorphisms, Weyl groups at punctures, and braid groups at boundary components—acts on $\\mathcal{A}_{R,S}$ by noncommutative quasi-cluster automorphisms, giving an explicit realization of mapping class group action.","For type $A_p$, $\\mathcal{A}_{R,S}$ is birationally equivalent to the moduli space of untwisted rank 1 local systems on the spectral surface, and for $C_p$ there is a symmetric spectral description; this links the new coordinates to spectral-network methods."],"supporting_citations":[{"why":"Classifies Lie algebras graded by simply laced root systems, supplying the recognition theorem for $A_p$, $D_p$, and $E$-type gradings.","marker":"[BM92]"},{"why":"Extends recognition to all reduced root systems and gives the generalized Tits construction $T(R,J)$ used to define $T(R,\\{J_i\\})$.","marker":"[BZ96]"},{"why":"Provides the classification of real and complex Jordan algebras and the Shirshov-Cohn theorem that two-generated subalgebras are special.","marker":"[McC04]"},{"why":"Introduces the parametrization of totally positive unipotent semigroups that the $R$-Lusztig map generalizes.","marker":"[Lus94]"},{"why":"Defines positive structures relative to a subset $\\Theta$ of simple roots, which the paper identifies with its Jordan-algebra positive structures.","marker":"[GW25]"},{"why":"Establishes the A/X cluster variety picture for split groups and the amalgamation strategy for configuration spaces that $\\mathcal{A}_{R,S}$ extends.","marker":"[FG06]"},{"why":"Provides cluster coordinates for all split real Lie groups and the cluster modular group $\\Gamma_{R,S}$ acting on them.","marker":"[GS19]"},{"why":"Introduces noncommutative cluster varieties via plabic networks and the spectral description that the type-$A_p$ case generalizes.","marker":"[GK24]"},{"why":"Gives cluster structures on double Bruhat cells of split groups, which the double $R$-Bruhat cell structure generalizes.","marker":"[BFZ05]"},{"why":"Supplies the polygonal cluster algebras and their seed-group and angle machinery used directly for noncommutative rank 1.","marker":"[GKNW26]"}],"fun_headline_variants":[],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The construction assumes that for every reduced root system $R$ and every compatible family of Jordan algebras $\\{J_i\\}$, there really is a Lie algebra $T(R,\\{J_i\\})$ with the prescribed root-system grading and prescribed Jordan factors—the paper invokes the Benkart-Moody and Benkart-Zelmanov recognition theorems to glue the Jordan algebras together, and if those theorems fail for some field or some compatible family, the parametrization of $G$-local systems would lose its main input.","fun_headline_variants_meta":{"error":"Client error '402 Payment Required' for url 'https://api.deepseek.com/chat/completions'\nFor more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/402"},"cache_creation_input_tokens":0},"created_at":"2026-08-28T20:10:46.029704+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the smallest higher-noncommutative-rank case not previously covered, for example a $C_3$-graded group with $J_1=J_2=M_k(\\mathbb{C})$ and $J_3=H_k(\\mathbb{C})$, pick two triangulations of a punctured surface, compute the same decorated local system's coordinates in both seeds, and check that the coordinate change equals the predicted noncommutative cluster mutation. A mismatch, or a collision of two distinct local systems in the Jordan-point map, would refute Theorem 1.10. A second check is positivity: if a formally real Jordan algebra yields a point that is positive in one seed but not in another, the claimed choice-independence of $\\mathcal{A}_{R,S}^{>0}$ fails.","supporting_citations":[],"review_version":1}