{"id":"82781a50-64f4-49e3-9fc7-464603261edb","arxiv_id":"1908.08895","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Every ribbon graph order has a Nakayama automorphism making it twisted 1-Calabi-Yau, and it is symmetric, hence 1-Calabi-Yau, precisely when the ribbon graph is bipartite or the field has characteristic two.","lead":"Ribbon graph orders, certain infinite-dimensional rings built from graphs on surfaces, are shown to be twisted 1-Calabi-Yau. They become genuinely 1-Calabi-Yau exactly when the underlying ribbon graph is bipartite, except in characteristic two.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proof of Lemma 2.16's zero case is a non sequitur; Theorem 2.5 and the twisted Calabi-Yau conclusion depend on it.","rationale":"The reader's rationale identifies Lemma 2.16 as the main concern, and our analysis confirms that this is the most load-bearing weak point: Theorem 2.5 and hence Corollary 2.7 depend on the Frobenius-form identity φ(qp)=φ(ε^⋆(p)q). The printed proof of Lemma 2.16 contains a genuinely unjustified equality in the zero case. This is not a mere stylistic gap: the equality is needed to show that the Frobenius form is ε^⋆-symmetrizing, which in turn produces the bimodule isomorphism. The reader's formal 'weakest_assumption' field names Lemma 2.11 (the basis), but the argument actually pivots on Lemma 2.16; hence agreement is partial. We did not find a counterexample to Lemma 2.16, and the lemma appears repairable via the symmetry of Lemma 2.15's pair list, so the appropriate verdict remains conditional on the author correcting the proof. Other potential concerns, such as Proposition 1.22's rank formula for arbitrary multiplicity maps, did not yield a concrete failure after checking the nodal singularity with m=(2,2); the module is still free over R with the stated rank. Thus no change to the reader's conditional verdict is warranted.","tokens_in":20112,"tokens_out":54787,"duration_ms":507721,"concrete_test":"Verify Lemma 2.16 directly on small ribbon graph orders: implement the completed path algebra over R=k[[t]] with t=Σ c_a, build the basis B from Lemma 2.11, and compute φ(qp)-φ(ε^⋆(p)q) for all p,q∈B for, say, the two-loop order k⟨⟨a,b⟩⟩/(a^2,b^2), the nodal singularity, and the circular graph orders Λ_n for n=2,3,4,5. If any difference is nonzero, Lemma 2.16 is false. Alternatively, fill the missing step in the proof: show that the set of pairs in Lemma 2.15 is closed under swapping and use ε^⋆(p)=±p to conclude φ(pq)=0 whenever φ(qp)=0; if this reconstruction fails for some pair, the lemma fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (Corollary 2.7) rests on Theorem 2.5, which asserts Λ^∨ ≅ Λ_{ε^⋆} as bimodules. The proof of Theorem 2.5 relies on Lemma 2.16, claiming φ(qp)=φ(ε^⋆(p)q) for all p,q. For the case φ(qp)=0 the proof prints: '0 = φ((ε^⋆)^2(qp)) = φ((ε^⋆)^2(p) ε^⋆(q)) = φ(ε^⋆(p)q).' The second equality is unjustified: (ε^⋆)^2(qp) is not (ε^⋆)^2(p) ε^⋆(q), and no earlier statement gives φ(p ε^⋆(q))=φ(ε^⋆(p)q). The first case only proved the desired identity when φ(qp)≠0, plus its contrapositive φ(ε^⋆(p)q)=0 ⇒ φ(qp)=0, not the converse. So the zero case is not established by the written argument. Since the Frobenius-form identity is the mechanism by which the canonical bimodule is identified with the twisted bimodule, the proof of the main theorem is incomplete as printed. The statement itself is likely repairable: Lemma 2.15's list of nonzero pairs is symmetric under swapping p and q (because (∂_m c_a, a^m) swaps to (∂_{n-m} c_{σ^m(a)}, (σ^m(a))^{n-m})), and ε^⋆(p)=±p for p∈B, so φ(qp)=0 implies (p,q) is not in the list and hence φ(pq)=0, giving φ(ε^⋆(p)q)=0. But this argument is absent.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies ribbon graph orders, understood as arrow ideal completions of path algebras of complete gentle quivers, and investigates their Calabi-Yau properties. The main result (Corollary 2.7) asserts that the derived Nakayama functor of a ribbon graph order is given by tensoring with a bimodule twisted by a canonical involution ε⋆, so that the relative Serre functor satisfies S ≅ (—)ε⋆ ∘ [1] and S² ≅ [2]; hence the order is twisted 1-Calabi-Yau and fractionally 2/2-Calabi-Yau. The paper further shows that a ribbon graph order is symmetric if and only if its graph is bipartite or the base field has characteristic two (Theorem 3.19), and derives analogues for twisted Brauer graph algebras (Theorem 3.11). The proofs are concrete: an explicit basis is exhibited, a Frobenius form is defined, and the canonical bimodule is identified with a twisted bimodule via an explicit isomorphism.","tokens_in":20480,"tokens_out":35367,"duration_ms":319970,"significance":"If the main theorem is correct, the paper gives a clean and explicit description of the canonical bimodule of ribbon graph orders and settles the symmetry question in terms of bipartiteness. The approach is constructive: the Frobenius form φ is written down, the bimodule isomorphism is explicit, and the reduction to circular graphs in the symmetry characterization is elegant. The paper is largely self-contained, using standard results on relative Serre duality from Iyama-Reiten and citing Kauer's thesis for the rank of the Cartan matrix. The applications to twisted Brauer graph algebras and the connection to dimer models make the results potentially useful beyond order theory. However, the proof of the key Frobenius-form identity in Lemma 2.16 contains a gap that must be repaired before the main theorem can be accepted.","major_comments":[{"comment":"The zero case of the proof is not established. The displayed sequence \"0 = φ((ε⋆)^2(qp)) = φ((ε⋆)^2(p)ε⋆(q)) = φ(ε⋆(p)q)\" is a non sequitur: since (ε⋆)^2 is the identity, the first equality is just φ(qp)=0, while the middle equality asserts φ(qp)=φ(pε⋆(q)), which is equivalent to the very identity being proved and is not available at that point. As Theorem 2.5# uses this lemma to prove that ϑ is a bimodule homomorphism, the main theorem is not proved as written. The statement is plausibly repairable: the support set in Lemma 2.15 is symmetric under swapping the two factors (with (∂_m c_a, a^m) corresponding to (∂_{n−m} c_{σ^m(a)}, (σ^m(a))^{n−m})), and since ε⋆ preserves B up to sign, φ(qp)=0 implies φ(pq)=0 and hence φ(ε⋆(p)q)=0. Please add this argument.","section":"§2.2.2, Lemma 2.16"}],"minor_comments":[{"comment":"The proof that B generates Λ would benefit from a sentence justifying that every nonzero path in a complete gentle quiver is a subpath of a power of a repetition-free cycle; as written, the decomposition \"p = a^{r n(a)+m}\" is asserted without proof.","section":"§2.2.1, Lemma 2.11"},{"comment":"The range \"1<m<n(a)\" should read \"1≤m<n(a)\" to include the case m=1, which is needed for consistency with the inverse map in part (c).","section":"Theorem 2.5#, part (b)"},{"comment":"Because ε⋆ is an involution, the notation (ε⋆)^2 in the zero case is just the identity; this makes the first step of the displayed argument confusing.","section":"§2.2.2, Lemma 2.16"},{"comment":"The letter φ is used for the map on arrows a↦φ(a) and later for the Frobenius form in Definition 2.13; please rename one of them to avoid ambiguity.","section":"Proposition 1.22(d)"},{"comment":"The implications (2)⇒(3) and (3)⇒(1) are skipped; since they are not used later this is acceptable, but a one-sentence indication or reference would improve readability.","section":"Proposition 3.15"}],"recommendation":"major_revision","confidential_remarks":"The gap in Lemma 2.16 is localized and appears repairable by the symmetry argument described in the major comment; if that repair is supplied, the main results are likely correct. The paper fits the journal's scope and is generally well written, but the current manuscript should not be accepted without the fix."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper gives a concrete and plausible description of the canonical bimodule of a ribbon graph order: it is the regular bimodule twisted by the involution coming from a polarization, so the derived Nakayama functor is just twisting by that involution. That is a real advance over the existing analogies with gentle and Brauer graph algebras, and the six-way symmetry criterion in Theorem 3.19 (bipartite graph or char 2) is new. The author works through explicit bases and a Frobenius form, and the main structure is clear. The reduction to circular graph orders in Section 3 is tidy.\n\nThe main soft spot is real. Lemma 2.16 asserts that φ(qp)=φ(ε*(p)q) for all p,q. The nonzero case is handled. For the zero case the proof prints '0 = φ((ε*)^2(qp)) = φ((ε*)^2(p)ε*(q)) = φ(ε*(p)q)'. The middle equality does not follow from anything stated: ε* is an algebra endomorphism, not an anti-endomorphism, so ε*(qp) is not ε*(q)ε*(p) in general; and applying ε* twice gives the identity, so the first equality is just φ(qp)=0. As printed, the argument only establishes the contrapositive direction, not the converse. That is a genuine gap in the proof of Theorem 2.5 and hence Corollary 2.7. It is probably not a fatal flaw: Lemma 2.15's characterization of nonvanishing pairs is symmetric under swapping the two entries, so when φ(qp)=0 one should be able to conclude φ(pq)=0, and since ε*(p)=±p for basis elements, that gives φ(ε*(p)q)=0. But that argument is absent, and the author should supply it.\n\nSecondary: Proposition 3.15 skips the proofs of two implications, and one benchmark result is quoted from Kauer's PhD thesis, which is not easy to verify. Those are minor. The citation pattern is honest; the forthcoming paper is only referenced as future work.\n\nOverall: the central construction is credible, the gap is specific and repairable, and the paper has enough new content to merit a serious referee. I would send it to peer review and ask the author to fix Lemma 2.16 in the revision.","headline":"Explicit Nakayama automorphism for ribbon graph orders, with a genuine but repairable gap in the proof of Lemma 2.16; worth refereeing.","tokens_in":20978,"tokens_out":3440,"would_cite":true,"duration_ms":31026,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["16G30","16E35","16G10","16G20","16H05"],"pacs":[],"model":"deepseek-v4-flash","headline":"Every ribbon graph order is twisted 1-Calabi-Yau, and exactly 1-Calabi-Yau when its ribbon graph is bipartite or the base field has characteristic two.","keywords":["ribbon graph orders","Calabi-Yau categories","derived Nakayama functor","Brauer graph algebras","complete gentle quivers","symmetric orders","polarizations","Frobenius forms"],"falsifier":"Compute the $R$-basis of the circular graph order $\\Lambda_3$ of Lemma 3.12 over a field of odd characteristic: either exhibit an element outside the span of $B$ in Lemma 2.11, which would break Theorem 2.5, or verify that no bimodule isomorphism $\\Lambda^{\\vee} \\cong \\Lambda$ exists, consistent with Theorem 3.19; the sign-trace calculation in Lemma 3.13 is the concrete place to look.","tokens_in":19913,"feed_emoji":"🎀","tokens_out":14790,"duration_ms":128192,"temperature":0.7,"pith_summary":"Every ribbon graph order — an infinite-dimensional algebra built from a quiver with two arrows entering and leaving each vertex and relations of length two — has a derived Nakayama functor given by twisting by an explicit involution; this makes its derived category twisted 1-Calabi-Yau and fractionally 2/2-Calabi-Yau. The same construction passes to anti-commutative finite-dimensional quotients, the twisted Brauer graph algebras, whose perfect derived categories are twisted 0-Calabi-Yau. The paper also decides exactly when the twist is trivial: a ribbon graph order is symmetric, hence 1-Calabi-Yau, precisely when its underlying graph is bipartite or the base field has characteristic two. The payoff is that these infinite-dimensional orders share the homological self-duality of finite-dimensional Frobenius algebras, with a graph-theoretic switch controlling symmetry.","feed_headline":"Every ribbon graph order is twisted 1-Calabi-Yau","feed_subtitle":"An involution describes its Serre functor; bipartite ribbon graphs make it genuinely symmetric.","key_machinery":"The load-bearing object is the Frobenius form $\\varphi = \\omega^{\\vee}$ on $\\Lambda$, assembled from a polarization $\\varepsilon : Q_1 \\to \\{+,-\\}$ that gives opposite signs to the two arrows leaving each vertex. The form satisfies the twisted symmetry identity $\\varphi(qp)=\\varphi(\\varepsilon^{\\star}(p)q)$, and the map $p \\mapsto p\\cdot \\varphi$ is shown to be a bimodule isomorphism $\\Lambda_{\\varepsilon^{\\star}} \\to \\Lambda^{\\vee}$. A basis lemma, Lemma 2.11, underpins the calculation: the idempotents, the two cyclic paths $x_i,y_i$ at each vertex, and the proper powers $a^m$ inside repetition-free cycles form an $R$-basis of the completed path algebra, which makes the pairing explicit.","core_discovery":"Corollary 2.7 states that for any ribbon graph order $\\Lambda$ the canonical bimodule $\\Lambda^{\\vee} = \\mathrm{Hom}_R(\\Lambda,R)$ is isomorphic to $\\Lambda_{\\varepsilon^{\\star}}$, the bimodule whose right action is twisted by an involution $\\varepsilon^{\\star}$ determined by a polarization of the quiver. Consequently the derived Nakayama functor $\\nu_\\Lambda$ is isomorphic to $(-)^{\\varepsilon^{\\star}}$, and the relative Serre functor satisfies $S_\\Lambda \\cong (-)^{\\varepsilon^{\\star}} \\circ [1]$ and $S_\\Lambda^2 \\cong [2]$. The paper reads this as $\\Lambda$ being twisted 1-Calabi-Yau and fractionally 2/2-Calabi-Yau. Theorem 3.19 then characterizes symmetry: $\\Lambda$ is symmetric, equivalently $S_\\Lambda \\cong [1]$, exactly when the ribbon graph is bipartite or the base field has characteristic two.","pith_inferences":["The paper leaves implicit that the same Frobenius form is literally the shared object behind both ribbon graph orders and twisted Brauer graph algebras, so homological invariants computed on the finite-dimensional quotient may lift to the infinite-dimensional order.","Because every finite-dimensional gentle algebra is a quotient of a ribbon graph order, the twisted 1-Calabi-Yau statement offers a possible uniform explanation of Calabi-Yau phenomena in derived categories of gentle algebras and their surface models; checking that transfer is a natural next step.","A testable extension: Remark 3.18 shows the full equivalence in Theorem 3.19 can fail without roots of $-1$ for all multiplicities, so one can ask exactly which multiplicity maps still force symmetry in the absence of that root condition."],"forward_implications":["The category $\\mathrm{Perf}_{\\mathrm{fd}}(\\Lambda)$ of perfect complexes with finite-dimensional cohomology is twisted 1-Calabi-Yau with Serre functor square $[2]$, so its Auslander-Reiten quiver consists of homogeneous tubes of rank one or two.","Every ribbon graph order is weakly symmetric: $\\Lambda^{\\vee} \\otimes_\\Lambda P_i \\cong P_i$ for each indecomposable projective module $P_i$.","For a twisted Brauer graph algebra $A$, there is an algebra involution $\\varepsilon$ with $A^{*} \\cong A_{\\varepsilon}$, so $\\mathrm{Perf}(A)$ is twisted 0-Calabi-Yau and fractionally 0/2-Calabi-Yau.","When the ribbon graph is bipartite or the base field has characteristic two, $\\Lambda$ is symmetric ($S_\\Lambda \\cong [1]$), and if the field contains the relevant roots of $-1$, the twisted algebra $A$ is isomorphic to an ordinary Brauer graph algebra.","Any dimer model or dessin d'enfants, together with a choice of node multiplicities, gives rise to a 1-Calabi-Yau category $\\mathrm{Perf}_{\\mathrm{fd}}(\\Lambda)$ and a 0-Calabi-Yau category $\\mathrm{Perf}(A)$."],"supporting_citations":[{"why":"introduces the order-theoretic ribbon graph construction and the Gorenstein property used to set up the relative Serre functor.","marker":"[KR01]"},{"why":"provides the definition of complete gentle quivers on which ribbon graph orders are built.","marker":"[Rin11]"},{"why":"supplies the relative Serre functor theorem and the symmetric-order criterion used in Theorems 1.25 and 1.27.","marker":"[IR08]"},{"why":"supplies the Frobenius algebra method, including symmetrizing forms and the Brauer-Nesbitt-Nakayama theorem, that the bimodule computation follows.","marker":"[SY11]"},{"why":"gives the rank formula for the Cartan matrix in terms of bipartite connected components, needed for Proposition 3.15.","marker":"[Kau98]"},{"why":"provides the proof that Brauer graph algebras are symmetric, used in Theorem 3.11(b).","marker":"[Sch18]"},{"why":"identifies Brauer graph algebras with symmetric special biserial algebras over algebraically closed fields, used in the final comparison.","marker":"[Sch15]"},{"why":"supplies the notion of twisted fractionally Calabi-Yau categories in which the paper states its main consequences.","marker":"[HI11]"}],"fun_headline_variants":["Every ribbon graph order twisted 1-CY","Bipartite ribbon graphs are symmetric and 1-CY","Ribbon graph orders: twisted 1-CY, symmetric if bipartite","Twist makes ribbon graph orders 1-CY","Ribbon graphs: twisted 1-CY, plain 1-CY if bipartite"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument stands and falls with Lemma 2.11's claim that the listed paths form an $R$-basis of the completed path algebra; if the completion contained any element not uniquely expressible as a power series in the central element $z$ times those monomials, the Frobenius form and the bimodule isomorphism would be undefined.","fun_headline_variants_meta":{"raw":{"variants":["Every ribbon graph order twisted 1-CY","Bipartite ribbon graphs are symmetric and 1-CY","Ribbon graph orders: twisted 1-CY, symmetric if bipartite","Twist makes ribbon graph orders 1-CY","Ribbon graphs: twisted 1-CY, plain 1-CY if bipartite"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.002042,"raw_usage":{"total_tokens":7873,"prompt_tokens":781,"completion_tokens":7092,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":397,"completion_tokens_details":{"reasoning_tokens":7000}},"tokens_in":397,"tokens_out":7092,"duration_ms":49793,"temperature":1.0,"reasoning_tokens":7000,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:27:10.849830+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the $R$-basis of the circular graph order $\\Lambda_3$ of Lemma 3.12 over a field of odd characteristic: either exhibit an element outside the span of $B$ in Lemma 2.11, which would break Theorem 2.5, or verify that no bimodule isomorphism $\\Lambda^{\\vee} \\cong \\Lambda$ exists, consistent with Theorem 3.19; the sign-trace calculation in Lemma 3.13 is the concrete place to look.","supporting_citations":[],"review_version":1}