{"id":"01fb1e65-dedc-4825-897a-8a32c237b973","arxiv_id":"2601.08789","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Root-of-unity quantum graph algebras have irreducible representations of maximal dimension l^{g·dim(g)+n·N}, and their small-quantum-group invariants have maximal dimension l^{g·dim(g)+N(n−1)−m}, with centers described explicitly.","lead":"At odd-order roots of unity, quantum graph algebras for a surface and a complex semisimple group are shown to have central localizations that are central simple algebras, with maximal representation dimensions computed exactly. The result fixes the size of state spaces entering non-semisimple TQFTs and skein quantization, uniformly for every simple Lie type.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Concern: App. C's asserted decomposition O_ε=⊕_{μ,ν} _μ(O_ε)_ν is unproved; if it fails, injectivity of Φ̂^ε_{g,n} (Prop. 5.9) collapses, taking domain and PI-degree results with it.","rationale":"The reader identified the same weakest assumption: injectivity of Φ̂^ε_{g,n}, with App. C as the under-proved base case. I agree this is the single most load-bearing step. My independent read confirms that the rest of the chain (quantum Killing form, integral forms, free module structure over Z0, PI-degree upper/lower bounds) is either imported from established results or proved in detail in the text. The only serious soft spot is App. C: the specialized weight-space decomposition is asserted rather than proved. It is plausible, because O_A is a free A-module with a basis of weight vectors, so the direct sum should specialize, and the leading coefficient ε^{(λ,σ)} is a nonzero root of unity. But the paper does not show that the defining conditions (157) characterize these specialized subspaces, nor that the direct sum is independent of the choice of representative in P/lP. Since domain and centralizer theorems collapse if this fails, the appropriate verdict is CONDITIONAL: accept only after App. C is expanded to a complete proof (or the check in concrete_test is carried out). The reader's ACCEPT with moderate confidence is not unreasonable, but a stress-test that takes the asserted step as the linchpin should require it to be nailed down.","tokens_in":83022,"tokens_out":25216,"duration_ms":227955,"concrete_test":"Fix G=SL_2, l=3, and let ε be a primitive 3rd root of unity. In O_A, choose nonzero weight vectors φ∈ _μ(O_A)_ν and ψ∈ _μ'(O_A)_ν with μ'−μ = 3 (so μ≡μ' mod 3). Specialize to O_ε and check whether φ|ε and ψ|ε are linearly independent over C. More generally, compute the joint eigenspaces of the left and right actions of K_1 and (K_1;3)_ε on O_ε and verify that they are indexed by all of P^2 (not merely P/3P × P/3P) and form a direct sum decomposition. If the eigenspaces for μ and μ+3 coincide or fuse, App. C's filtration is invalid; if they remain independent, the decomposition is justified and the injectivity proof can be completed.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claims rest on L^ε_{g,n} being a domain (Prop. 5.21) and on the centralizer theorem (Th. 5.13). Both are downstream of injectivity of the modified Alekseev morphism Φ̂^ε_{g,n} (Prop. 5.9). That injectivity reduces to two base cases: Φ^ε_{0,1} (imported from [12, Cor. 2.25], solidly supported by App. A) and Φ^ε_{1,0} (App. C). App. C is a one-paragraph 'adaptation' of [14, Th. 3.13]. The critical new step is the assertion that (156)–(157) specialize to a weight-space decomposition O_ε = ⊕_{μ,ν∈P} _μ(O_ε)_ν, and that the leading-term formula Φ^ε_{1,0}(β⊗1) = ε^{(λ,σ)}β⊗K_{λ+σ} + (lower terms) holds. No proof is given that the operators in (157) separate the indices μ,ν after specialization, nor that the subspaces corresponding to weights differing by lP are linearly independent. If the decomposition is not a direct sum (or the leading coefficient vanishes for some nonempty weight space), the filtration argument collapses. Since every subsequent structure—domain property (App. D uses injectivity of Φ̂^ε_{g,n}), centralizer theorem (Th. 5.13), and hence the PI-degree computations (Th. 5.24, Th. 5.28)—depends on this, the central claim is only as secure as this unproved step. The rest of the architecture is coherent and cross-referenced, but this specific gap is the load-bearing point.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the specialization L^ε_{g,n} at odd roots of unity ε of the quantum graph algebras associated to a simply-connected complex semisimple group G and an oriented surface Σ°_{g,n} of genus g with n punctures and one boundary component. The main structural claims are: (i) L^ε_{g,n} is a free module of rank l^{(2g+n) dim(g)} over the central subalgebra Z_0(L^ε_{g,n}) ≅ O(G)^{⊗(2g+n)}; (ii) the central localization Q(L^ε_{g,n}) is a division algebra and a central simple algebra of PI-degree l^{g dim(g)+nN}; (iii) the analogous statement for the invariant subalgebra L^{u^ε}_{g,n} under the small quantum group u^ε has PI-degree l^{g dim(g)+(n-1)N-m} for (g,n)≠(0,1); (iv) the centers Z(L^ε_{g,n}) and Z(L^{u^ε}_{g,n}) are integrally closed, Noetherian, and are described by explicit multiplication isomorphisms. The proofs use a modified Alekseev morphism \\widehatΦ^ε_{g,n}, the quantum moment map, factorizability of u^ε, and a reduction to classical invariant theory. The representation-theoretic consequences via PI-ring theory are also drawn: simple modules have dimension at most the PI-degree, every central character is realized, and generic central characters have a unique simple module of maximal dimension.","tokens_in":83305,"tokens_out":15308,"duration_ms":149887,"significance":"If the structural theorems are correct, this is a substantial and uniform advance: it extends the root-of-unity structure theory of quantum graph algebras to all simple Lie types and to punctured surfaces, giving explicit PI-degrees, freeness ranks, and center descriptions, with applications to stated skein algebras and non-semisimple TQFTs. The paper is careful and mostly transparent: it provides machine-checkable algebraic constructions (the modified Heisenberg double and modified Alekseev morphism are defined and proved within the paper), gives explicit formulas for ranks and PI-degrees, and isolates the dependence on prior results. However, one load-bearing point—the injectivity of the modified Alekseev morphism—rests on an appendix that is too terse and, as stated, contains a questionable weight-space decomposition. The gap is localized and appears repairable, but it must be fixed before the main theorems can be considered fully established.","major_comments":[{"comment":"The decomposition O_ε = ⊕_{μ,ν∈P} _μ(O_ε)_ν, with _μ(O_ε)_ν defined by specializing (157), is not a direct sum as stated. In Γ^Q_ε one has K_i^l = 1 (eq. (86)), so the operators in (157) distinguish only μ,ν modulo lP; the eigenspaces for μ and μ+lϖ coincide. This affects the injectivity of Φ^ε_{1,0} and hence of \\widehatΦ^ε_{g,n} (Prop. 5.9), which is used for the domain property (Prop. 5.21), the centralizer theorem (Th. 5.13), and the PI-degree computations (Th. 5.24, Th. 5.28). A repair is possible by defining _μ(O_ε)_ν as the image of the A-weight space (via the canonical basis of O_A), which gives a genuine direct-sum decomposition, or by replacing P with P/lP and adapting the filtration argument. As written, the proof is incomplete and the statement is misleading.","section":"Appendix C, leading-term formula"},{"comment":"The claimed formula Φ^ε_{1,0}(β⊗1) = ε^{(λ,σ)} β⊗K_{λ+σ} + (lower terms) is asserted without proof. In the generic case this follows from explicit commutation relations in the Heisenberg double; after specialization one must verify that the lower terms are strictly ordered with respect to a well-founded order and that the leading coefficient cannot vanish. The collapse of weight spaces modulo lP makes this a nontrivial check, not a verbatim adaptation of [14, Th. 3.13]. Please supply a complete argument or a precise reference covering the specialized setting.","section":"Appendix C, leading-term formula"}],"minor_comments":[{"comment":"The notation (K_i; l)_{q_i} is retained while the text says the equations are specialized; please specify the specialized operators and their eigenvalue formulas to avoid ambiguity.","section":"Eq. (157)"},{"comment":"Once the index set is understood modulo lP, the phrase 'maximal for the order ≤' needs justification: the order must be shown to be well-founded on the relevant finite set and that no terms above the chosen maximal index can appear in the expansion.","section":"Prop. C.1"},{"comment":"The proof of Cor. D.1 is very brief; if the filtration T is not fully described in the final version, a complete definition or a precise reference to [14, Prop. 5.7] with the modifications needed at the root of unity would help the reader.","section":"Appendix D"},{"comment":"The subalgebra \\hatZ_0(L^ε_{g,n}) is introduced in mid-proof; a displayed definition before the theorem would improve readability.","section":"Th. 5.24"}],"recommendation":"major_revision","confidential_remarks":"The App. C issue is real and load-bearing, but it appears fixable within the scope of the paper. If the authors provide a corrected weight-space argument and prove the leading-term formula, the paper should be acceptable. I do not see evidence of circularity beyond what the authors disclose; the cross-references between Th. 5.28 and Th. 6.1 are carefully separated. The paper is a strong contribution to the same research program, and the main architecture is coherent."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this paper proves the expected PI-degree formulas for root-of-unity specializations of quantum graph algebras L^ε_{g,n} and their small-quantum-group invariants L^{u^ε}_{g,n} for all simple types and all (g,n), describes the centers, and shows freeness over the Frobenius center. If the proofs hold, these are the definitive structural results in the area. The main new engine is a modified Alekseev morphism Φ̂^ε_{g,n} into an extended Heisenberg double Ĥ, plus a centralizer theorem (Th 5.13) identifying L^{u^ε}_{g,n} as the centralizer of the quantum moment map. The architecture is coherent and the paper is unusually upfront about its dependencies and deferred proofs.\n\nThe soft spot is exactly where you'd expect: everything downstream of injectivity of Φ̂^ε_{g,n} (Prop 5.9). That reduces to two base cases: Φ^ε_{0,1}, imported from [12], and Φ^ε_{1,0}, handled in App. C. The latter is one paragraph: it says the weight-space decomposition O_ε = ⊕_{μ,ν} _μ(O_ε)_ν specializes from the A-form, and then uses it to get the leading-term filtration. No proof is given that this decomposition is a direct sum after specialization, nor that the leading coefficients are nonzero on each weight space. If that decomposition fails, the filtration collapses, and with it the domain property and the PI-degree computations. The reader's stress test flags this, and I think the flag is fair. The rest of the paper's dependencies—on [12] for the (0,1) base case and on [14] for generic-q structure—are standard and clearly cited.\n\nThat said, I don't read this as a fatal flaw. The step is probably fixable with a more careful argument (Lusztig's canonical basis should control the specialized weights), but it is a genuine gap in the current version, not a cosmetic one. A referee should ask for a complete proof of App. C before signing off on Th 5.24 and 5.28.\n\nWho is this for? People working on skein algebras, non-semisimple TQFT state spaces, and quantum group invariants. It deserves a serious referee. I would send it out, with the explicit request that the referee verify App. C and the reduction to [12, Cor. 2.25]. If the gap is closed, it's an acceptance; if not, it's a major revision.","headline":"Strong results on PI degrees of root-of-unity graph algebras, but injectivity of the modified Alekseev morphism rests on a sketched filtration in App. C.","tokens_in":84075,"tokens_out":3274,"would_cite":true,"duration_ms":29235,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["17B37","20G42","16T20"],"pacs":[],"model":"deepseek-v4-flash","headline":"Quantum graph algebras specialized at odd roots of unity are shown to have central localizations that are central simple algebras, with exact PI degrees controlling all irreducible representation dimensions.","keywords":["quantum graph algebras","roots of unity","central simple algebras","PI degree","small quantum group","quantum moment map","Alekseev morphism","character varieties"],"falsifier":"For the smallest case g=sl_2 and l=3, compute Phi^epsilon_{1,0} on a non-zero element of a highest weight space and check whether the leading-term coefficient in the Appendix C filtration is indeed non-zero; a zero leading term would produce an element in the kernel, collapsing the domain and PI-degree results.","tokens_in":82691,"feed_emoji":"🌀","tokens_out":4656,"duration_ms":46073,"temperature":0.7,"pith_summary":"The paper studies the specialization of quantum graph algebras at odd roots of unity. It proves that both the full algebra and its subalgebra of elements invariant under a small quantum group are domains whose central localizations are central simple algebras, and it computes their PI degrees exactly. These numbers bound the largest possible dimension of any irreducible representation, so the result pins down the representation theory of these algebras. A sympathetic reader would care because these algebras quantize character varieties and appear in non-semisimple topological quantum field theories.","feed_headline":"PI degrees computed for root-of-unity graph algebras","feed_subtitle":"The bounds on irreducible representations follow exactly, settling the representation theory of these algebras.","key_machinery":"The argument is carried by a modified Alekseev morphism bPhi^epsilon_{g,n} that intertwines the quantum moment map with the representation theory of the small quantum group, together with the double centralizer theorem. Injectivity of bPhi^epsilon_{g,n} makes L^epsilon_{g,n} a domain, and the quantum moment map identifies the invariant subalgebra L^{u^epsilon}_{g,n} as the centralizer of a subalgebra in L^epsilon_{g,n}, which allows the PI degree to be computed from the center of L^epsilon_{g,n}.","core_discovery":"For a root of unity of odd order, the specialized graph algebra L^epsilon_{g,n} is a free module of rank l^{(2g+n)dim(g)} over its central subalgebra Z_0, and its central localization Q(L^epsilon_{g,n}) is a division algebra, central simple of PI degree l^{g·dim(g)+nN}, where N is the number of positive roots. For the invariant subalgebra L^{u^epsilon}_{g,n}, the central localization is central simple of PI degree l^{g·dim(g)+N(n-1)-m} when (g,n) is not (0,1). From the theory of PI rings, every simple module has dimension at most this degree, every central character arises from some simple module, and generic central characters admit a unique simple module of exactly this dimension.","pith_inferences":["If the same framework is extended to roots of unity beyond the odd, coprime-to-D case (which the paper flags as future work), the central-simplicity and PI-degree structure would plausibly survive, though the small quantum group would no longer be factorizable and new arguments would be needed.","Because the center equals the trace ring, these algebras are likely maximal orders over their centers; verifying this would complete the description of their noncommutative algebraic geometry.","The compatibility of the central extension with the mapping class group action suggests the PI-degree results could be made equivariant, leading to a refined picture of the action on irreducible representations."],"forward_implications":["Irreducible representations of L^epsilon_{g,n} and L^{u^epsilon}_{g,n} have dimension bounded by the respective PI degrees, and the bound is attained exactly on the complement of the discriminant variety.","The centers of both algebras are integrally closed Noetherian rings, described by the subalgebras Z_0 and Z_1, and they coincide with the trace rings of the algebras.","The projection to the graph algebra of the small quantum group makes L^epsilon_{g,n} a central extension of the endomorphism algebra of a TQFT state space, so the representation-theoretic results transfer to that setting.","The PI-degree formula covers all simply-connected complex semi-simple groups uniformly, without needing a presentation of the algebra by generators and relations."],"fun_headline_variants":["Root-of-unity graph algebras become central simple division algebras","PI degrees give tight bounds on simple modules for graph algebras","PI degree bounds simple module dimensions at roots of unity","Odd root of unity yields explicit PI degree for graph algebras"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The entire structure theorem rests on the injectivity of the modified Alekseev morphism, and in particular on the base case Phi^epsilon_{1,0}, whose proof in Appendix C relies on a weight-space filtration that is asserted rather than fully proved.","fun_headline_variants_meta":{"raw":{"variants":["Root-of-unity graph algebras become central simple division algebras","PI degrees give tight bounds on simple modules for graph algebras","PI degree bounds simple module dimensions at roots of unity","Odd root of unity yields explicit PI degree for graph algebras"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001253,"raw_usage":{"total_tokens":4937,"prompt_tokens":672,"completion_tokens":4265,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":416,"completion_tokens_details":{"reasoning_tokens":4199}},"tokens_in":416,"tokens_out":4265,"duration_ms":29952,"temperature":1.0,"reasoning_tokens":4199,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T10:46:55.037935+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For the smallest case g=sl_2 and l=3, compute Phi^epsilon_{1,0} on a non-zero element of a highest weight space and check whether the leading-term coefficient in the Appendix C filtration is indeed non-zero; a zero leading term would produce an element in the kernel, collapsing the domain and PI-degree results.","supporting_citations":[],"review_version":1}