{"id":"5afad35d-49e7-417a-9a86-c47e5224caa7","arxiv_id":"2508.01950","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Dimension-two twisted graded Calabi-Yau algebras on two-vertex quivers are classified into four families, A2(q), B2(q), J, D(q), with complete isomorphism criteria.","lead":"This mathematics paper classifies, up to isomorphism, all two-dimensional twisted Calabi-Yau algebras that can be built from a quiver with two vertices. It lists four explicit families and gives exact rules for when two members are the same, including two exceptional cases.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 3.3's normal-form change of basis is algebraically wrong: it uses eigenvectors of αβ and u_i=βv_i, but the τ-equivariance requires eigenvectors of (αβ)^T and u_i=α^T v_i. This invalidates the proof of the classification for M=[[0,2],[2,0]]; the theorem may survive, but needs correction.","rationale":"The paper's central classification Theorem 3.27 depends on Lemma 3.3 to show that every A satisfying Hypothesis 3.1 with adjacency matrix [[0,2],[2,0]] and identity Nakayama permutation is isomorphic to A2(q) or D(q). The proof of Lemma 3.3 rests on a change of basis intended to diagonalize τ. I checked the claimed identities τ(tilde a)=λ1 tilde b and τ(tilde b)=tilde a. For the notation in the lemma, τ(tilde a) has coordinates β^T u1 in the (a,c) basis, so the identities require β^Tβ v1=λ1 v1. The hypothesis αβ v1=λ1 v1 does not imply this. A concrete example with α=[[2,1],[0,3]] and β=I shows the asserted identities fail. A corrected construction uses eigenvectors of (αβ)^T and u_i=α^T v_i, which suggests the theorem is likely true, but the proof as written has a real gap. Lemma 3.4 also contains minor transposition typos, but those are less serious. Since the invalid step is load-bearing for completeness, the verdict should move from ACCEPT to CONDITIONAL pending a corrected proof. I disagree with the reader's weakest_assumption, which focused on external theorem hypotheses rather than this internal algebraic error.","tokens_in":36,"tokens_out":48104,"duration_ms":1136264,"concrete_test":"Check the asserted τ-identities on the explicit algebra defined by τ(a)=2b+d, τ(c)=3d, τ(b)=a, τ(d)=c on quiver (1.7) with P=I. Using the paper's Lemma 3.3 construction (v1=e1, u1=βv1=e1, so tilde a=b and tilde b=a), compute τ(tilde b)=τ(a)=2b, which is not equal to tilde a. If instead the construction is amended to take v1 as an eigenvector of (αβ)^T and u1=α^T v1, the identities hold and the algebra is A2(3/2). This test isolates the invalid step without disputing the final classification.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"In Lemma 3.3, for quiver (1.7) with P=I, the paper sets τ(a)=α11b+α12d, τ(c)=α21b+α22d, τ(b)=β11a+β12c, τ(d)=β21a+β22c, with α,β∈GL2(k). It then takes v1,v2 to be eigenvectors of αβ, sets u_i=βv_i, and defines tilde a=(u1)1b+(u1)2d, tilde b=(v1)1a+(v1)2c, etc. The asserted identities τ(tilde a)=λ1 tilde b and τ(tilde b)=tilde a are false in general. Indeed, τ(tilde a) has coordinates β^T u1 in the (a,c) basis, so the first identity requires β^Tβ v1=λ1 v1, while the hypothesis αβ v1=λ1 v1 gives no such relation. For example, take τ(a)=2b+d, τ(c)=3d, τ(b)=a, τ(d)=c, so α=[[2,1],[0,3]], β=I. With v1=e1, u1=βv1=e1, the construction gives tilde a=b, tilde b=a; then τ(tilde b)=τ(a)=2b≠tilde a, contradicting the claimed identity. A correct normal form uses eigenvectors of (αβ)^T and u_i=α^T v_i. Since Lemma 3.3 is the sole justification that every A with M=[[0,2],[2,0]] and P=I is A2(q) or D(q), the completeness part of Theorem 3.27 is not proven as written. The D(q) Jordan-block paragraph makes the same unsupported assertion.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper classifies, up to isomorphism, twisted graded Calabi-Yau algebras of dimension two on strongly connected two-vertex quivers with finite GK dimension. Relying on the Reyes-Rogalski presentation of such algebras as quotients of translation quivers by mesh relations, the authors prove that every such algebra is isomorphic to one of four families: A2(q), B2(q), J, or D(q). They also solve the isomorphism problem within these families, establish pairwise non-isomorphism results, and prove analogous isomorphism criteria for the larger families A_n(q) and B_n(q). The main classification is stated as Theorem 3.27.","tokens_in":14127,"tokens_out":5903,"duration_ms":63436,"significance":"If the proof is correct, the result is a complete isomorphism classification for a natural class of twisted graded Calabi-Yau algebras on two-vertex quivers, complementing the known one-vertex case and the earlier structural theorems of Reyes and Rogalski. The paper also gives useful isomorphism criteria for the modified preprojective algebra families A_n and B_n. A notable strength is the systematic use of the graded isomorphism theorem of Bell-Zhang via Gaddis's path-algebra version, which reduces ungraded isomorphism questions to graded ones. However, two lemmas in the classification section contain algebraic errors or inconsistencies that are load-bearing for the completeness claim, so the main theorem is not proven as written.","major_comments":[{"comment":"The normal-form argument is algebraically incorrect. With τ(a)=α11 b+α12 d, τ(c)=α21 b+α22 d, τ(b)=β11 a+β12 c, τ(d)=β21 a+β22 c, the proof takes v1 to be an eigenvector of αβ and sets u1=β v1. The asserted identities τ(tilde a)=λ1 tilde b and τ(tilde b)=tilde a require β^T β v1=λ1 v1 and α^T v1=β v1, respectively, but the hypothesis αβ v1=λ1 v1 gives neither. For example, with α=[[2,1],[0,3]], β=I, and v1=e1, the construction gives tilde a=b and tilde b=a, while τ(tilde b)=τ(a)=2b+d, which is not tilde a. Consequently, the reduction of every algebra with M=[[0,2],[2,0]] and P=I to A2(q) or D(q) is not established, and the completeness part of Theorem 3.27 is not proven as written. A correct normal form should use eigenvectors of (αβ)^T and u_i=α^T v_i; the Jordan-block paragraph for D(q) requires the same correction.","section":"Section 3, Lemma 3.3"},{"comment":"The displayed expression for ω in the case P=I2 is inconsistent with the stated arrow structure. For a quiver with adjacency matrix [[1,1],[1,1]] and P=I2, the condition τ(e_i V e_j)⊂e_j V e_i forces τ(b) to be a linear combination of arrows from e2 to e1, not a loop, and similarly for the other arrows. The text instead writes τ(a)=α1 a, τ(b)=α2 d, τ(c)=α3 b, τ(d)=α4 c and then ω=α1 a^2+α2 bd+α3 c^2+α4 bd, which does not match either the quiver (1.8) or the definition of J. Since Lemma 3.4 is the sole basis for the B2(q) and J cases of the classification, this lemma must be rewritten with a consistent labeling of arrows and a correct computation of ω before Theorem 3.27 can be accepted.","section":"Section 3, Lemma 3.4"},{"comment":"The Hilbert series verification for J is only partially supplied: after describing the leading-term reductions and possible path forms, the proof states that 'the remainder are left to the reader.' Because the Calabi-Yau property of J is one of the four families in Theorem 3.27, the omitted path counts should be filled in or at least summarized explicitly; as written, the equality h_J(t)=(I-Mt+Pt^2)^{-1} is not fully verified.","section":"Section 2, Lemma 2.4"}],"minor_comments":[{"comment":"In the line following equation (3.7), 'ℓ_1b+ell2d' should read 'ℓ_1b+ℓ_2d'.","section":"Section 3, Lemma 3.6"},{"comment":"The sentence 'the direct some of two (connected) algebras' contains a typo: 'some' should be 'sum'.","section":"Section 3, introductory paragraph"},{"comment":"The word 'autormorphism' should be 'automorphism'.","section":"Section 3, Lemma 3.2 proof"},{"comment":"In the sentence 'This p = q^{-1}, then the proof is complete', the word 'this' should be 'if'.","section":"Section 3, Proposition 3.14"},{"comment":"The notation for arrows in the quiver (1.8) is ambiguous in the text-only rendering; a clearer diagram or explicit source/target lists for a,b,c,d would help the reader verify the computations in Lemma 3.4.","section":"Section 3, Lemma 3.4"}],"recommendation":"major_revision","confidential_remarks":"The central classification is plausible and the use of external structural theorems is appropriate, but the proof of Lemma 3.3 contains a genuine linear-algebra error that invalidates the completeness argument for the M=[[0,2],[2,0]] case. Lemma 3.4 has similar internal inconsistencies. These are fixable within the scope of the manuscript, so I recommend major revision rather than rejection. The authors should also clarify the quiver drawings, since the text-only conventions for the arrow directions are not always clear."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The main classification of dimension-two twisted Calabi-Yau algebras on two-vertex quivers has a real gap in Lemma 3.3. As written, the completeness portion of Theorem B is not established. The isomorphism criteria for the A_n(q) and B_n(q) families are new and convincing, and the overall strategy is sound, but I wouldn't cite the classification until the gap is patched.\n\nWhat's good: Theorem A gives clean criteria for isomorphisms of A_n(q) and B_n(q) using the graded isomorphism lemma of Bell-Zhang and Gaddis's adaptation; those proofs are straightforward and check out. The exceptional algebras J and D(q) are handled with concrete Hilbert series computations, and the paper is careful about the Nakayama automorphism. The reliance on Reyes-Rogalski's structural theorems is appropriate, not a weakness.\n\nThe soft spot: in Lemma 3.3, after ruling out P swapping, the paper picks eigenvectors v_i of αβ, sets u_i = β v_i, and claims τ(ã) = λ_i b̃ and τ(b̃) = ã. That's not what the matrix action gives. For a vector x in the (b,d) basis, τ(x) is β x in the (a,c) basis; for a vector y in the (a,c) basis, τ(y) is α y in the (b,d) basis. So with ã in the (b,d) basis with coordinates u_1 = β v_1, we have τ(ã) = β^2 v_1, and with b̃ in the (a,c) basis with coordinates v_1, τ(b̃) = α v_1. The identities would require β^2 v_1 = λ_1 v_1 and α v_1 = β v_1, which do not follow from αβ v_1 = λ_1 v_1. The stress-test note includes a concrete counterexample showing the construction fails. The reduction of an arbitrary τ to the normal form for A2(q) or D(q) is therefore not proven. This is load-bearing: the rest of Theorem B relies on it. The classification may still be true, but a corrected basis choice is needed—likely using eigenvectors of β α or the transposed pair, with the matching u_i = α v_i. There are also minor typos, such as a repeated bd term in the ω expression in Lemma 3.4 that should be db, and some Hilbert series counts in Lemma 2.4 are left to the reader.\n\nBottom line: this is a serious paper worth a referee, but I would not accept it in current form. The gap in Lemma 3.3 must be fixed. Once fixed, the classification is a useful contribution to the isomorphism problem for twisted Calabi-Yau algebras. For a reading group, it's worth a look as an application of the Reyes-Rogalski framework, but I'd wait for a corrected version before citing.","headline":"The classification is plausible and the A_n/B_n isomorphism theorems are solid, but the proof of Lemma 3.3 has a load-bearing change-of-basis error that currently invalidates the completeness of Theorem B.","tokens_in":14668,"tokens_out":4892,"would_cite":false,"duration_ms":53341,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["16E65","16P90","16S38","16W50"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper classifies all dimension-two twisted graded Calabi–Yau algebras on strongly connected two-vertex quivers, showing they fall into exactly four explicit families.","keywords":["twisted graded Calabi-Yau algebras","two-vertex quivers","translation quivers","isomorphism problem","preprojective algebras","Hilbert series","Gelfand-Kirillov dimension","Nakayama automorphism"],"falsifier":"A concrete test: search for a finite-Gelfand–Kirillov-dimension twisted graded Calabi–Yau algebra on a strongly connected two-vertex quiver whose adjacency matrix is not $\\begin{pmatrix}0&2\\\\2&0\\end{pmatrix}$ or $\\begin{pmatrix}1&1\\\\1&1\\end{pmatrix}$; the classification predicts none exists, so finding one refutes Lemma 3.2.","tokens_in":13515,"feed_emoji":"🧩","tokens_out":8090,"duration_ms":93192,"temperature":0.7,"pith_summary":"The paper solves the isomorphism problem for dimension-two twisted graded Calabi–Yau algebras that come from a strongly connected quiver with two vertices. It proves that every such algebra is isomorphic to one of four explicit algebras: two one-parameter families $A_2(q)$ and $B_2(q)$, and two exceptional algebras $J$ and $D(q)$. Within the parameterized families, the only coincidences are $A_2(q) \\cong A_2(q^{-1})$ and $D(q) \\cong D(q^{-1})$. The result gives a complete list of isomorphism classes for this class of noncommutative algebras, which are twisted relatives of preprojective algebras. A complete classification matters because these algebras are a testbed for the isomorphism problem in noncommutative algebra, and the list is short enough to be useful for checking further properties.","feed_headline":"Four families classify all two-vertex twisted Calabi-Yau algebras","feed_subtitle":"Every such algebra is one of four explicit families, with q and 1/q identified in two of them.","key_machinery":"The load-bearing object is the pair $(M,P)$: the adjacency matrix $M$ of the quiver and the permutation matrix $P$ coming from the Nakayama automorphism. A known structure theorem for dimension-two twisted graded Calabi–Yau algebras says the matrix-valued Hilbert series is $(I - Mt + Pt^2)^{-1}$, and that $M$, $M^T$, and $P$ pairwise commute; combined with finite Gelfand–Kirillov dimension this forces the spectral radius of $M$ to be 2. That fact plus commutativity leaves only the two adjacency matrices above. A second mechanism is the graded isomorphism theorem for quotients of path algebras by homogeneous ideals, which lets the authors replace arbitrary algebra isomorphisms by graded ones, so the classification can be carried out by linear changes of variables on the arrows.","core_discovery":"The central claim, Theorem 3.27, is that the class is exactly $\\{A_2(q), B_2(q), J, D(q)\\}$. More precisely, if $A = A(Q,\\tau)$ is a twisted graded Calabi–Yau algebra of dimension two with finite Gelfand–Kirillov dimension, $Q$ strongly connected, and $|Q_0|=2$, then $A$ is isomorphic to one of these algebras. The four families are pairwise non-isomorphic except $A_2(q)\\cong A_2(q^{-1})$ and $D(q)\\cong D(q^{-1})$. The proof first restricts the adjacency matrix of $Q$ to $\\begin{pmatrix}0&2\\\\2&0\\end{pmatrix}$ or $\\begin{pmatrix}1&1\\\\1&1\\end{pmatrix}$, then normalizes the twisting map $\\tau$ by changes of basis, reducing the relations to the four displayed forms.","pith_inferences":["The method suggests a general recipe for classifying twisted graded Calabi–Yau algebras on larger strongly connected quivers: first solve the spectral-radius and commutativity constraints on $(M,P)$, then normalize $\\tau$ by eigenvector changes of basis; the hard part is likely the isomorphism step, where the exceptional families $J$ and $D(q)$ show that non-schurian quivers produce unexpected cas","The $q \\leftrightarrow q^{-1}$ identification in $A_2(q)$ and $D(q)$ hints at an underlying orientation-reversal duality; testing whether these isomorphisms lift to derived equivalences or Morita equivalences would be a natural next step the paper does not address.","Because the theorem assumes finite Gelfand–Kirillov dimension, the paper leaves open the infinite-growth case on two-vertex quivers; exploring it might reveal additional families that degenerate to these four in the finite-growth limit.","The Hilbert-series criterion is algorithmic: given a quiver and $\\tau$, one can compute $M$ and $P$ and check whether $\\rho(M)=2$ and $M, M^T, P$ commute, giving a fast necessary condition for membership in the classified families."],"forward_implications":["Every isomorphism class in the two-vertex setting is represented by one of four explicit presentations, so any invariant of these algebras can be checked on the finite list.","The isomorphism problem for the two-vertex algebras is completely settled: $A_2(q)$ and $D(q)$ have only the $q \\leftrightarrow q^{-1}$ identification, $B_2(q)$ has no parameter identifications, and $J$ is a singleton.","For $n \\geq 3$, Theorem A classifies the $A_n(q)$ and $B_n(q)$ families: $A_n(q)$ has dihedral symmetry, while $B_n(q)$ has only rotational symmetry.","The results confirm that these algebras are twisted versions of preprojective algebras of type $A$; setting all $q_i = 1$ recovers the ordinary preprojective algebras."],"supporting_citations":[{"why":"Supplies the characterization that $A(Q,\\tau)$ is twisted graded Calabi–Yau of dimension two iff the matrix Hilbert series is $(I - Mt + Pt^2)^{-1}$, and that $M$, $M^T$, and $P$ commute; every enumeration step leans on it.","marker":"[8]"},{"why":"Gives the theorem that isomorphisms between quotients of path algebras by homogeneous ideals are graded, letting the classification use graded maps.","marker":"[4]"},{"why":"The Diamond Lemma is used to count normal forms and compute the Hilbert series of $B_n(q)$, $J$, and $D(q)$.","marker":"[2]"},{"why":"Provides the Hilbert series calculation for the $A_n(q)$ preprojective-type algebras.","marker":"[3]"},{"why":"The isomorphism lemma for connected graded algebras underlies the graded-isomorphism theorem applied throughout.","marker":"[1]"},{"why":"Connects twisted Calabi–Yau algebras to homological regularity, supporting the dimension-two framework.","marker":"[7]"}],"fun_headline_variants":["Four families classify all two-vertex twisted CY algebras","Two-vertex quivers: exactly four families of twisted CY algebras","All 2-vertex twisted CY algebras are one of four families","Four explicit families cover 2-vertex twisted CY algebras","Classification complete: 2-vertex twisted CY algebras are four families"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The classification rests on the known theorem that every such algebra has Hilbert series $(I - Mt + Pt^2)^{-1}$ with $M$, $M^T$, and $P$ commuting; if that theorem carries hidden hypotheses about the Nakayama automorphism or the grading, the list of possible quivers could be incomplete.","fun_headline_variants_meta":{"raw":{"variants":["Four families classify all two-vertex twisted CY algebras","Two-vertex quivers: exactly four families of twisted CY algebras","All 2-vertex twisted CY algebras are one of four families","Four explicit families cover 2-vertex twisted CY algebras","Classification complete: 2-vertex twisted CY algebras are four families"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00131,"raw_usage":{"total_tokens":5260,"prompt_tokens":783,"completion_tokens":4477,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":399,"completion_tokens_details":{"reasoning_tokens":4391}},"tokens_in":399,"tokens_out":4477,"duration_ms":39819,"temperature":1.0,"reasoning_tokens":4391,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T05:15:47.074433+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete test: search for a finite-Gelfand–Kirillov-dimension twisted graded Calabi–Yau algebra on a strongly connected two-vertex quiver whose adjacency matrix is not $\\begin{pmatrix}0&2\\\\2&0\\end{pmatrix}$ or $\\begin{pmatrix}1&1\\\\1&1\\end{pmatrix}$; the classification predicts none exists, so finding one refutes Lemma 3.2.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the characterization that $A(Q,\\tau)$ is twisted graded Calabi–Yau of dimension two iff the matrix Hilbert series is $(I - Mt + Pt^2)^{-1}$, and that $M$, $M^T$, and $P$ commute; every enumeration step leans on it."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the theorem that isomorphisms between quotients of path algebras by homogeneous ideals are graded, letting the classification use graded maps."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The Diamond Lemma is used to count normal forms and compute the Hilbert series of $B_n(q)$, $J$, and $D(q)$."},{"cited_title":"Etingof and C.-H","cited_arxiv_id":null,"evidence_quote":"Provides the Hilbert series calculation for the $A_n(q)$ preprojective-type algebras."},{"cited_title":"Bell and J","cited_arxiv_id":null,"evidence_quote":"The isomorphism lemma for connected graded algebras underlies the graded-isomorphism theorem applied throughout."},{"cited_title":"Reyes, D","cited_arxiv_id":null,"evidence_quote":"Connects twisted Calabi–Yau algebras to homological regularity, supporting the dimension-two framework."}],"review_version":1}