{"id":"c5bbb59a-7448-447d-91eb-a41a1695c2d5","arxiv_id":"2501.06822","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"For Schur representations in Azumaya algebras, the coarse moduli space is a quotient algebraic space, and the stack of twisted representations is equivalent to the corresponding quotient stack.","lead":"This paper constructs moduli spaces and stacks for representations of associative algebras inside Azumaya algebras, so that rational points of coarse moduli spaces can be interpreted through twisted representations. It matters because it extends quiver moduli results to arbitrary ground rings without stability conditions and identifies the true moduli stack with a concrete representation-theoretic object.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 6.4's proof of full faithfulness is incomplete: automorphism-group bijections alone do not force a functor between groupoids to be fully faithful, and essential surjectivity inherits the gap.","rationale":"The reader identified the terseness of Step 3 of Theorem 6.4 as a minor issue, but the manuscript's weakest-assumption analysis centered on the Schur restriction. I agree that the Schur restriction is an explicit and legitimate scope limitation, not a defect. However, the Step 2/3 argument is more load-bearing than merely terse: the stated proof of full faithfulness is logically insufficient, and essential surjectivity is derived from it. The theorem is very likely correct—the natural inverse functor appears constructible—but the paper does not contain that construction or a standard full-faithfulness argument. All other ingredients, including the representability of X in Theorem 3.5, the gerbe constructions in Section 5, and the Brauer-class computations in Section 8, appear sound and in line with existing literature. I therefore recommend accepting the paper conditional on completing the verification of full faithfulness/essential surjectivity of Φ, rather than rejecting it or declaring an actual falsehood.","tokens_in":30882,"tokens_out":21146,"duration_ms":239463,"concrete_test":"Write out an explicit quasi-inverse Ψ : [X/G^op/Q] → M. For an object (U,g,P,f), let M be the locally free rank-one left Λ^azu ⊗ O_U-module corresponding to the G_U-torsor P, set E' = Hom_{Λ^azu}(M, F^taut_{U'}), and define ρ' by descending the universal Schur representation along the G-torsor P using f : P → X. Check that Ψ∘Φ and Φ∘Ψ are naturally isomorphic, paying attention to Gm-weights on the splitting gerbe and to compatibility with the Q-structure. Concretely, test this for a non-split Azumaya algebra, such as a quaternion algebra over a field, and for a base algebraic space Q that is not a scheme. If the inverse exists and the natural transformations are isomorphisms, the proof gap is closed; if not, Theorem 6.4 is false.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The paper's central claim is the equivalence in Theorem 6.4. In Step 2 of its proof, the authors show that for a fixed twisted Schur representation, the comparison functor Φ induces a bijection on automorphism groups, via Corollary 2.4 and the Schur property. They then conclude: 'Since the fibers are groupoids, Φ is fully faithful on the fibers.' This inference is not valid in general: a functor between groupoids can induce isomorphisms on all automorphism groups yet fail to be full or faithful on Hom sets between non-isomorphic objects (e.g., two objects with no morphisms between them can both map to a single object). The subsequent Step 3, which claims essential surjectivity follows 'immediately from step 2' together with effectiveness of descent, therefore rests on an unproven full-faithfulness statement. The gap is likely repairable—the natural inverse sends (U,g,P,f) to E' = Hom_{Λ^azu}(M,F^taut_{U'}), where M is the rank-one Λ^azu-module associated with the G_U-torsor P, with ρ' obtained by descending the universal Schur representation along P—but the paper does not supply this verification. Since Corollary 6.5, Theorem 7.1, and Theorem 8.3 all depend on Theorem 6.4, the central derivation is not completely secure as written.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces moduli stacks and spaces for Schur representations of a finitely presented associative algebra Lambda in an Azumaya algebra Lambda^azu over arbitrary ground rings. It defines a stack M of twisted Schur representations living on the gerbe of splittings of Lambda^azu, and proves (Theorem 6.4) that this stack is equivalent to the quotient stack [X/G^op/Q], where X is the quasiaffine scheme of Schur representations, G is the unit group, and Q = X/H^op is the coarse algebraic space. The paper then uses this equivalence to show that every rational point of Q acquires geometric origin after a canonical modification of the moduli problem (Theorem 7.1), and to compute Brauer classes of tautological Azumaya algebras on Q (Theorem 8.3).","tokens_in":31086,"tokens_out":6163,"duration_ms":59416,"significance":"If Theorem 6.4 holds, the paper provides a clean stack-theoretic framework for understanding why rational points on coarse moduli spaces of representations may fail to be represented by actual objects, and it gives a canonical repair via twisted representations and Azumaya algebras. The treatment is broad: it works over arbitrary ground rings, avoids stability conditions, and covers arbitrary finitely presented algebras, including group and path algebras. The exposition is largely concrete and checkable, with explicit constructions of the splitting gerbe, the comparison functor, and the tautological sheaf; the equivariance check in Lemma 6.3 is carried out in detail. The restriction to Schur representations is clearly stated and is a real limitation, as ordinary simple or stable representations need not be Schur over non-perfect fields. The main results are important and plausible, but the proof of the central equivalence Theorem 6.4 contains a nontrivial gap that needs to be addressed.","major_comments":[{"comment":"The proof of full faithfulness is incomplete. The argument verifies that the comparison functor Phi induces a bijection on automorphism groups of each object in a fiber, and then concludes that Phi is fully faithful because the fibers are groupoids. This inference is not valid: a functor between groupoids can induce isomorphisms on all automorphism groups while failing to be full or faithful on Hom sets between non-isomorphic objects (for example, two objects with no morphisms between them can both map to one object). To prove full faithfulness, the paper must compare Hom-sets between arbitrary objects of the twisted Schur stack and the quotient stack, for instance by identifying both with isomorphisms of the associated G-torsors, or by constructing an explicit inverse functor along the lines suggested in the stress-test note: send (U,g,P,f) to E' = Hom_{Lambda^azu}(M, F^taut_{U'}) with M the rank-one Lambda^azu-module attached to the G_U-torsor P. This verification is absent.","section":"Theorem 6.4, Step 3"},{"comment":"The proof of essential surjectivity is not supplied. The text states that every descent datum of the quotient stack arises from a descent datum of M, and says this is 'immediate from step 2', but full faithfulness on fibers does not imply that every object of the quotient stack is locally in the image of Phi. One must construct, for a given quadruple (U,g,P,f), a twisted Schur representation (E',rho') whose image under Phi is that quadruple, at least after passing to a cover, and then use descent to glue. The paper does not provide this construction. Since Corollary 6.5, Theorem 7.2, and Theorem 8.3 all depend on Theorem 6.4, this gap is load-bearing and requires a repaired proof.","section":"Theorem 6.4, Proof, overall"}],"minor_comments":[{"comment":"The reduction to the case where R is finitely generated over Z is too terse; the proof should explicitly justify that quasi-compactness and local finite presentation descend along the fppf extension R -> R0.","section":"Theorem 3.5, proof"},{"comment":"The statement that all possible GIT quotients are simultaneously contained as schematic open subsets in Q is imprecise, since GIT quotients typically involve semistable points that are not Schur; it should be restricted to loci where objects are geometrically stable, as the preceding sentence suggests.","section":"Section 3, closing paragraph"},{"comment":"The condition that rho' is Schur is formulated by requiring that all pullbacks rho'_U to objects of the splitting gerbe are Schur; it may be simpler and more natural to require that the kernel of the adjoint map is exactly the scalars after pullback to a covering of the gerbe, but the current formulation is acceptable.","section":"Definition 6.1"},{"comment":"The paper uses [24] (SGA 1) for several stack-descent facts; for the benefit of readers, more precise pointers (e.g., the relevant Expose and proposition numbers for effectiveness of descent data on sheaves and torsors) would be helpful.","section":"Throughout"},{"comment":"The phrase 'the twisted forms \\tilde H of the group scheme H = PGL_n correspond to Azumaya algebras' is slightly telegraphic; the correspondence is via the non-abelian cohomology set H^1(-, PGL_n) and is standard, but a few words of clarification would improve readability.","section":"Introduction, page 3"}],"recommendation":"major_revision","confidential_remarks":"The proof gap in Theorem 6.4 is real but appears repairable; the rest of the paper is carefully written and the announced results are significant. I recommend major revision rather than rejection, provided the authors supply a complete proof of full faithfulness and essential surjectivity for the comparison functor, or equivalently construct the missing inverse functor."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"One thing to know: the paper is worth reading and refereeing, but the proof of its central equivalence, Theorem 6.4, has a gap that needs closing.\n\nWhat is actually new: the representability of Schur representations in Azumaya algebras over arbitrary ground rings (Theorem 3.5), the stack of twisted Schur representations as a quotient stack (Theorem 6.4), and the Brauer class formulas in Theorem 8.3. The paper is honest that the abstract \"geometric origin\" mechanism of Theorem 1.2 is well-known. What is not well-known is the concrete stack-theoretic translation for representations, and the paper does that carefully. Section 4's Grassmannian description for quivers is a nice independent check. Lemma 6.3 is a genuine piece of work: writing out the equivariance of the comparison map via the tensor identity E^∨ ⊗ E = End(E).\n\nThe soft spot is exactly where the stress-test note lands. Step 2 of Theorem 6.4 claims full faithfulness of Φ on fibers after checking only that Φ induces bijections on automorphism groups. That inference is not valid for groupoids, as the note says. Step 3 then makes essential surjectivity depend on Step 2, so the central equivalence is not fully proved as written. I think the gap is likely repairable — the natural inverse via the Hom sheaf Hom(F^taut, E') should work — but the authors need to supply the proof. A referee should ask for it. Also, the essential surjectivity argument is terse even apart from the gap.\n\nOther reservations are minor. The Schur restriction is real and the paper acknowledges it; Lemma 3.6 shows simple representations are Schur only when geometrically simple, so the framework excludes non-simple objects over imperfect fields. That is a scope limit, not a flaw. Reference [42] is in the bibliography but never cited; the authors should either cite it where relevant or remove it.\n\nWho this is for: people working on moduli of quiver or algebra representations, and anyone interested in rational points on coarse moduli spaces. It deserves a serious referee, not a desk reject. I would want to see the gap in Theorem 6.4 fixed before publication, but the paper is sound in its main ideas and clearly written.","headline":"Solid and genuinely useful, but Theorem 6.4's proof has a repairable gap that needs to be closed.","tokens_in":31699,"tokens_out":2624,"would_cite":true,"duration_ms":23992,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14D22","14D23","14A20","16G10","16G20","16H05"],"pacs":[],"model":"deepseek-v4-flash","headline":"Schur representations of associative algebras in Azumaya algebras admit a genuine moduli stack, and twisting makes every rational point of the coarse space come from an actual representation.","keywords":["rational points","coarse moduli spaces","algebraic stacks","twisted representations","Azumaya algebras","Schur representations","non-abelian cohomology","Brauer group"],"falsifier":"A direct way to test the main equivalence is to compute the automorphism group of a twisted Schur representation $(S,E',\\rho')$ over a local ground ring $R$ for which $\\operatorname{Hom}_{\\mathcal{O}_{S'}}(E',F^{\\mathrm{taut}}_{S'})$ is a non-free rank-one module. Theorem 6.4 and Corollary 2.4 force this automorphism group to be $R^{\\times}$; any non-scalar automorphism found in such a computation would refute the full faithfulness of the comparison functor and hence the equivalence.","tokens_in":30580,"feed_emoji":"🌀","tokens_out":13053,"duration_ms":117284,"temperature":0.7,"pith_summary":"The paper proves that Schur representations of a finitely presented associative algebra $\\Lambda$ in an Azumaya algebra $\\Lambda^{\\mathrm{azu}}$ form a genuine moduli stack, over arbitrary ground rings and without stability conditions. The central theorem identifies this stack of twisted Schur representations, whose objects live on the gerbe of splittings of the Azumaya algebra, with the quotient stack $[X/G^{\\mathrm{op}}/Q]$, where $X=X^{\\Lambda^{\\mathrm{azu}}}_{\\Lambda/R}$ is the quasiaffine scheme of Schur representations. Consequently every rational point of the coarse moduli space $Q=X/H^{\\mathrm{op}}$ acquires geometric origin after modifying the moduli problem by twisting the Azumaya algebra. The proof also computes two Brauer classes on $Q$, relating the endomorphism algebra of the tautological sheaf to the non-abelian coboundary of the quotient torsor.","feed_headline":"Twisted representations realize every rational moduli point","feed_subtitle":"A stack equivalence turns coarse moduli points of algebra representations into actual twisted representations and computes the Brauer…","key_machinery":"The load-bearing machinery is the $\\mathbb{G}_m$-gerbe of splittings $V'$ of the Azumaya algebra $\\Lambda^{\\mathrm{azu}}$, carrying the tautological locally free sheaf $F^{\\mathrm{taut}}_{V'}$ of rank $n$ and weight one. A twisted Schur representation is a pair $(E',\\rho')$ where $E'$ is a locally free weight-one sheaf of rank $n$ on $V'$ and $\\rho':\\Lambda\\otimes\\mathcal{O}_{V'}\\to\\operatorname{End}(E')$ is Schur. The comparison functor $\\Phi$ converts $(V,E',\\rho')$ into the $G_V$-torsor $P_{V,E'}=\\operatorname{Hom}_{\\mathcal{O}_{V'}}(E',F^{\\mathrm{taut}}_{V'})$ and a $G_V$-equivariant morphism $P_{V,E'}\\to X$; the canonical isomorphism $E^{\\vee}\\otimes E\\cong\\operatorname{End}(E)$ carries the proof of equivariance. The target $[X/G^{\\mathrm{op}}/Q]$ consists of $G$-torsors $P\\to U$ with a $G$-equivariant map $P\\to X_U$ inducing a given $g:U\\to Q$.","core_discovery":"The authors' central claim is Theorem 6.4: the category of twisted Schur representations $\\mathcal{M}=\\mathcal{M}^{\\Lambda^{\\mathrm{azu}}}_{\\Lambda/R}$ is equivalent, as a category fibered over affine schemes, to the quotient stack $[X/G^{\\mathrm{op}}/Q]$. An object of $\\mathcal{M}$ over an affine scheme $V$ is a pair $(E',\\rho')$ with $E'$ a locally free sheaf of rank $n=\\deg(\\Lambda^{\\mathrm{azu}})$ and weight one on the $\\mathbb{G}_{m,V}$-gerbe $V'$ of splittings, and $\\rho':\\Lambda\\otimes\\mathcal{O}_{V'}\\to\\operatorname{End}(E')$ a Schur representation. The equivalence sends such a pair to the $G_V$-torsor $\\operatorname{Hom}_{\\mathcal{O}_{V'}}(E',F^{\\mathrm{taut}}_{V'})$ together with a $G_V$-equivariant map to $X$, and the key identity $E^{\\vee}\\otimes E\\cong\\operatorname{End}(E)$ makes the map equivariant. Corollary 6.5 then yields that $\\mathcal{M}$ is an Artin stack, a $\\mathbb{G}_m$-gerbe over $Q$, with $Q=X/H^{\\mathrm{op}}$ as its coarse moduli space. Applying the twisting formalism, any rational point of $Q$ becomes of geometric origin for the modified moduli problem, and the Brauer classes satisfy $[A_Q]=\\partial[X]$ and $[A_Q^0]=[\\Lambda^{\\mathrm{azu}}\\otimes\\mathcal{O}_Q]$.","pith_inferences":["A natural extension, not pursued in the paper, would relax the Schur condition to allow objects with non-scalar endomorphism rings; the automorphism groups would then no longer be $\\mathbb{G}_m$, and the gerbe structure and Brauer-class computations would have to be replaced by a different group.","The general twisting theorem of Section 1 applies to any free quotient $Q=X/H^{\\mathrm{op}}$; the same modification technique could be used for moduli of coherent sheaves through Kronecker modules, as the authors mention, or for other moduli problems with non-trivial automorphisms.","The identity $[A_Q]=\\partial[X]$ suggests a practical obstruction computation: in examples, the Brauer class of the endomorphism Azumaya algebra of the tautological bundle can be compared with the class of the splitting gerbe to decide whether a given rational point is geometric.","Over non-perfect fields, Lemma 3.6 shows that simple representations need not be Schur, so the stack misses many simple objects; checking whether the equivalence survives after base change to the perfection would clarify the arithmetic scope of the result."],"forward_implications":["The stack $\\mathcal{M}$ is an Artin stack and a $\\mathbb{G}_m$-gerbe over the algebraic space $Q$, with $Q$ as its coarse moduli space; the quotient-stack description gives a concrete representation-theoretic model for the gerbe.","For any rational point $g\\in Q(R)$, twisting by the torsor $g^*(X)$ produces a modified moduli problem with the same quotient $Q$, and the twisted form $\\widetilde{X}$ is the scheme of Schur representations of $\\Lambda$ in the twisted Azumaya algebra; in this modified problem $g$ has geometric origin.","If $H^1(S,\\mathrm{GL}_n)$ is a singleton, having geometric origin, being induced by a Schur representation, and being induced by a twisted Schur representation are equivalent for a rational point $g\\in Q(R)$.","The endomorphism Azumaya algebra $A_Q$ of the tautological sheaf has Brauer class $\\partial[X]$, while its commutant $A_Q^0$ has Brauer class $[\\Lambda^{\\mathrm{azu}}\\otimes\\mathcal{O}_Q]$; these are explicit obstructions attached to the coarse space.","Over fields, the geometrically stable representations are Schur, so the stable loci of all GIT stability conditions appear as open subschemes of $Q$, and the stack $\\mathcal{M}$ covers them without fixing a stability condition."],"supporting_citations":[{"why":"Establishes that twisted forms of PGL_n are Azumaya algebras, the correspondence that lets the moduli problem be formulated with an arbitrary Azumaya algebra $\\Lambda^{\\mathrm{azu}}$.","marker":"[23]"},{"why":"The quiver moduli problem whose rational points motivated this paper; supplies the type-map partition of $Q(S)$ generalized in Section 1.","marker":"[26]"},{"why":"Provides the theory of Artin stacks and quotient stacks used to identify $\\mathcal{M}$ with $[X/G^{\\mathrm{op}}/Q]$ and to conclude $\\mathcal{M}$ is an Artin stack.","marker":"[33]"},{"why":"Supplies the lemma that a free quotient by a flat group scheme is an algebraic space, used for $Q=X/H^{\\mathrm{op}}$ and Corollary 6.5.","marker":"[34]"},{"why":"Gives the Brauer group of a commutative ring and the theorem that commutants of Azumaya algebras are Azumaya, used in Lemma 8.2 and Theorem 8.3.","marker":"[7]"},{"why":"Introduces twisted sheaves and provides the observation that a weight-one locally free sheaf on a $\\mathbb{G}_m$-gerbe represents the gerbe's Brauer class, used to prove $[A_Q]=\\partial[X]$.","marker":"[13]"},{"why":"Provides the fppf descent and stack-axiom results used in Proposition 6.2 and in the full faithfulness and essential-surjectivity steps of Theorem 6.4.","marker":"[24]"},{"why":"Develops the moduli theory of twisted sheaves and splitting gerbes that underlies the definition of twisted Schur representations in Section 5.","marker":"[35]"}],"fun_headline_variants":["Every rational moduli point is a twisted representation","Twisted Schur reps give geometric origin to all points","Moduli stack equivalence: twisted reps = quotient stack","Rational points on moduli spaces become twisted reps","No stability needed: twisted reps cover rational points"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole construction is limited to Schur representations, meaning that at every prime point the only endomorphisms of the representation in the Azumaya algebra are scalar multiplications; if a representation has any non-scalar endomorphism it is not an object of the stack, and the main equivalence does not apply.","fun_headline_variants_meta":{"raw":{"variants":["Every rational moduli point is a twisted representation","Twisted Schur reps give geometric origin to all points","Moduli stack equivalence: twisted reps = quotient stack","Rational points on moduli spaces become twisted reps","No stability needed: twisted reps cover rational points"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000509,"raw_usage":{"total_tokens":2506,"prompt_tokens":998,"completion_tokens":1508,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":614,"completion_tokens_details":{"reasoning_tokens":1434}},"tokens_in":614,"tokens_out":1508,"duration_ms":11110,"temperature":1.0,"reasoning_tokens":1434,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T20:50:57.394360+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct way to test the main equivalence is to compute the automorphism group of a twisted Schur representation $(S,E',\\rho')$ over a local ground ring $R$ for which $\\operatorname{Hom}_{\\mathcal{O}_{S'}}(E',F^{\\mathrm{taut}}_{S'})$ is a non-free rank-one module. Theorem 6.4 and Corollary 2.4 force this automorphism group to be $R^{\\times}$; any non-scalar automorphism found in such a computation would refute the full faithfulness of the comparison functor and hence the equivalence.","supporting_citations":[{"cited_title":"Grothendieck: Le groupe de Brauer","cited_arxiv_id":null,"evidence_quote":"Establishes that twisted forms of PGL_n are Azumaya algebras, the correspondence that lets the moduli problem be formulated with an arbitrary Azumaya algebra $\\Lambda^{\\mathrm{azu}}$."},{"cited_title":"Hoskins, F","cited_arxiv_id":null,"evidence_quote":"The quiver moduli problem whose rational points motivated this paper; supplies the type-map partition of $Q(S)$ generalized in Section 1."},{"cited_title":"Laumon, L","cited_arxiv_id":null,"evidence_quote":"Provides the theory of Artin stacks and quotient stacks used to identify $\\mathcal{M}$ with $[X/G^{\\mathrm{op}}/Q]$ and to conclude $\\mathcal{M}$ is an Artin stack."},{"cited_title":"Laurent, S","cited_arxiv_id":null,"evidence_quote":"Supplies the lemma that a free quotient by a flat group scheme is an algebraic space, used for $Q=X/H^{\\mathrm{op}}$ and Corollary 6.5."},{"cited_title":"Auslander, O","cited_arxiv_id":null,"evidence_quote":"Gives the Brauer group of a commutative ring and the theorem that commutants of Azumaya algebras are Azumaya, used in Lemma 8.2 and Theorem 8.3."},{"cited_title":"de Jong: A result of Gabber","cited_arxiv_id":null,"evidence_quote":"Introduces twisted sheaves and provides the observation that a weight-one locally free sheaf on a $\\mathbb{G}_m$-gerbe represents the gerbe's Brauer class, used to prove $[A_Q]=\\partial[X]$."},{"cited_title":"Grothendieck: Revˆ etements ´ etales et groupe fondamental (SGA 1)","cited_arxiv_id":null,"evidence_quote":"Provides the fppf descent and stack-axiom results used in Proposition 6.2 and in the full faithfulness and essential-surjectivity steps of Theorem 6.4."},{"cited_title":"Lieblich: Moduli of twisted sheaves and generalized Azumaya algebras","cited_arxiv_id":null,"evidence_quote":"Develops the moduli theory of twisted sheaves and splitting gerbes that underlies the definition of twisted Schur representations in Section 5."}],"review_version":1}