{"id":"c27ff74f-deae-4625-8c65-148b4f529994","arxiv_id":"2505.16560","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Auslander-Reiten-Serre duality and almost-split conflations are established for extended hearts of proper connective dg algebras, together with a first Brauer-Thrall finiteness theorem.","lead":"This paper proves that extended hearts of proper connective differential graded algebras have a full Auslander-Reiten theory: almost-split conflations exist, and a first Brauer-Thrall finiteness theorem holds. The result broadens a classical tool from finite-dimensional algebras to the dg setting, giving explicit descriptions of Auslander-Reiten quivers.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proposition 2.11, the dual truncation equivalence used to define τ, is only sketched; its verification is load-bearing for Theorem 1.2.","rationale":"The reader's identification of Proposition 2.11 as the weakest point is accurate. The proof of Theorem 1.2 hinges on τ being a composite of verified equivalences; Corollary 2.10 is proved in detail, but its dual, Proposition 2.11, is explicitly only sketched. The t-exactness of D is standard, but the passage from the dual of Theorem 2.5 to the two quotients in Proposition 2.11 requires checking the kernel of the truncation and the image of the quotient subcategory; the compressed diagram does not supply this. I am not claiming the result is false: the paper contains substantial supporting material, including explicit examples and a detailed Harada–Sai argument, and the gap is plausibly fillable. However, because the dualisation also interacts with interval conventions and shifts (the proof shifts H^d_A by [−d+1]), a sign error here would change the definition of τ and hence the duality isomorphism. The listed concrete test settles the issue without altering the reader's CONDITIONAL verdict.","tokens_in":20103,"tokens_out":23526,"duration_ms":190875,"concrete_test":"Write out a complete proof of Proposition 2.11 by dualising the proof of Theorem 2.5 and Corollary 2.10 line by line: (i) prove the injective-resolution analogue of Lemma 2.7 for objects of thick(DA)[0,d], (ii) prove the analogue of Lemma 2.9 for t<d and P∈thick(DA)[0,d], and (iii) show that t<d sends add(DA[−d]) to zero and induces the claimed equivalence after quotient by add(DA[−d]⊕DA). As a minimal check, instantiate d=1 and A=H^0(A) a finite-dimensional algebra: verify explicitly whether the functor in Proposition 2.11 is the classical cokernel equivalence K[0,1](inj)/add(DΛ[−1]⊕DΛ) ≅ modΛ or whether a shift or quotient correction is needed; if the literal functor [d−1]∘t<d differs from the classical equivalence, adjust the definition of τ accordingly.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The construction of τ in diagram (3) rests on two equivalences: Corollary 2.10 and Proposition 2.11. The latter is introduced with 'we only give a sketch of the dualising procedure,' and the sketch does not verify the key points needed for Theorem 1.2. It asserts t-exactness of D with respect to the opposite t-structure, but t-exactness alone does not show that [d−1]∘t<d has kernel exactly add(DA[−d]), nor that the second quotient by add(DA[−d]⊕DA) is compatible with the induced functor; these are the properties that make τ a well-defined endo-equivalence of H^d_A. The displayed diagram also involves a shift [−d+1] and an identification of H^d_A[−d+1] with H^d_{Aop} that is not checked step-by-step. Since the binatural isomorphism η in the proof of Theorem 1.2 is built from τ and from relative Serre duality, a gap in the dual truncation equivalence leaves the central claim without a fully verified construction.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the extended heart H^d_A = D^{fd}(A)^{≤0} ∩ D^{fd}(A)^{>−d} of a proper connective dg algebra A. The main results are: (1) Theorem 1.2 establishes an Auslander–Reiten–Serre duality (τ, η) on H^d_A, with τ built explicitly from a derived Nakayama functor and truncation equivalences in diagram (3); (2) Corollary 1.3 concludes that H^d_A has almost-split conflations; (3) Section 3 proves a Harada–Sai lemma and a version of the first Brauer–Thrall conjecture for bounded connected components of the Auslander–Reiten quiver of H^d_A, assuming a d-connectedness condition; and (4) Section 4 gives an explicit algorithm for computing τ, illustrated by several examples in Section 5.","tokens_in":20295,"tokens_out":5269,"duration_ms":41522,"significance":"If Theorem 1.2 is fully justified, it is a meaningful generalization of the Auslander–Reiten–Serre duality for extended module categories proved by Zhou [Zho25], and it extends the known d-self-injective dg algebra case of Jin [Jin20]. The paper also provides a concrete and transparent construction of the Auslander–Reiten translation, together with worked examples and a Harada–Sai argument adapted to the dg setting. These are substantial and useful contributions. The manuscript is honest about its reliance on prior work: the characterization of inflations and deflations is imported from [Moc25], projective covers from [Bon24], and relative Serre duality from [Kel94]. The main weakness is that one load-bearing equivalence, Proposition 2.11, is only sketched, and the proof of Theorem 1.2 depends on it directly. The remaining issues are local and addressable.","major_comments":[{"comment":"The proof of Proposition 2.11 is explicitly only a sketch, and the skipped verification is load-bearing: the equivalence [d−1]∘t_{<d}: thick(DA)^{[0,d]}/(add(DA[−d]⊕DA)) → H^d_A is the right-hand leg of diagram (3), so τ in Theorem 1.2 is not fully constructed unless Proposition 2.11 is established. The sketch asserts t-exactness of D with respect to the opposite t-structure, but it does not verify that t_{<d} has kernel exactly add(DA[−d]) on thick(DA)^{[0,d]}, nor that the second quotient by add(DA[−d]⊕DA) is compatible with the induced functor, nor the identification H^d_A[−d+1] ≅ H^d_{A^{op}} used in the displayed diagram. I request a complete proof of the two equivalences, or a precise reference where the dual statements are proved in full, since Theorem 1.2 currently rests on this gap.","section":"Section 2, Proposition 2.11"},{"comment":"The proof of Lemma 2.9 invokes Lemma 2.8 to conclude Hom_A(P, t_{≤−d}Y)=0 and Hom_A(P, t_{≤−d}Y[1])=0. Lemma 2.8 is stated for P∈per(A)^{[−n,0]} and Y∈D^{fd}(A)^{≤0}, whereas Lemma 2.9 assumes P∈per(A)^{(−d,0]} and Y∈D^{fd}(A) arbitrary. The required vanishing therefore needs its own argument, or Lemma 2.8 must be reformulated to cover this case. This point is used in the fullness and kernel parts of Theorem 2.5 and also later in Proposition 3.22, so it should be fixed rather than left implicit.","section":"Section 2, Lemma 2.9"},{"comment":"In the definition of d-connectedness the index range is written as 0≤i≤1−d, and the same range appears in Theorem 3.17. For d>1 this set is empty, which cannot be the intended condition; from the proof of Theorem 3.20 and from Proposition 3.22, the intended range is 0≤i≤d−1. This appears to be a typo, but it occurs in the definition of a central notion, so it should be corrected explicitly.","section":"Definition 3.18 and Theorem 3.17"}],"minor_comments":[{"comment":"The exact sequence displayed in Remark 2.4 uses the notation H^d_A(X,Y) without defining the stable Hom notation, and the proof of Theorem 1.2 then writes a sequence with Hom_A(X,Y) and Hom_A(X,P^0_Y) that is not visibly the same as the displayed sequence. Please clarify the notation and the passage from the stable exact sequence to the cokernel computation.","section":"Section 2, Remark 2.4 and proof of Theorem 1.2"},{"comment":"The indexing in the induction step, especially the expressions M_{2t(n−1)−1}, M_{t(1)t(n−1)−(t(1)−1)}, and the morphisms g_2h_1, is hard to follow. A cleaner indexing or a short example would improve readability.","section":"Section 3, Lemma 3.12"},{"comment":"There are a few typographical slips: 'caluculate' in Example 4.2, 'definining' in Remark 3.13, and 'the the natural number' in Definition 3.10. These should be corrected.","section":"Global"},{"comment":"The statement 'Then there is no morphism starting or ending in 1' is terse; since this example is used to show that Theorem 3.17 is more general than Theorem 3.20, a short verification of this claim would be helpful.","section":"Section 5, Example 5.3"}],"recommendation":"major_revision","confidential_remarks":"The central ideas are sound and the paper is likely correct after a complete proof of Proposition 2.11 and the small fixes above. The heavy reliance on the first author's unpublished preprint [Moc25] for the characterization of inflations and deflations (Remark 1.1) and for Lemma 3.9 is a burden on the reader; the editor may wish to ensure that the cited statements are publicly available in a stable form. This does not affect my recommendation, which is driven by the local but load-bearing gap in Proposition 2.11."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read this one. The headline result — AR-Serre duality for extended hearts H^d_A of proper connective dg algebras — is new in this generality, and the construction of τ via the derived Nakayama functor and truncation equivalences is a real improvement over the finite-dimensional case. The proof of Theorem 1.2 is detailed, the Harada-Sai lemma adaptation is careful, and the concrete AR quiver computations in Sections 4-5 give the paper a useful hands-on side.\n\nThe soft spot is exactly where the stress-test points: Proposition 2.11 is only sketched. The sketch asserts that the duality D is t-exact for the opposite t-structure, but that alone doesn't show the induced functor [d-1]∘t_{<d} has kernel precisely add(DA[-d]), nor that it descends through the quotient by add(DA[-d]⊕DA). The identification H^d_A[-d+1] ≅ H^d_{Aop} is also not checked. Since this proposition is one of the two equivalences feeding into τ, Theorem 1.2 is technically incomplete as written. I don't see an actual error — the dualization of Theorem 2.5 should go through — but the authors need to supply the full proof, not just a sketch.\n\nMinor issues: the index range in Definition 3.18 (\"0 ≤ i ≤ 1-d\") is a typo, should be d-1 (or d); Lemma 2.9's appeal to Lemma 2.8 is imprecise, the vanishing needs the filtration argument spelled out. The dependence on [Moc25] for Remark 1.1 and Lemma 3.9 is fine but a referee should verify those statements.\n\nOverall this is a solid paper for representation theorists working with dg algebras and extriangulated categories. Deserves a serious referee; I'd recommend engagement and request a full proof of Proposition 2.11 and typos fixed.","headline":"Solid generalization of Zhou's AR theorem to proper connective dg algebras, with a fixable sketch-gap in Proposition 2.11.","tokens_in":20879,"tokens_out":5112,"would_cite":true,"duration_ms":39181,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["18G80","16E45","16G70"],"pacs":[],"model":"deepseek-v4-flash","headline":"A proper connective dg algebra's extended heart admits an Auslander–Reiten–Serre duality, yielding almost-split conflations and a Brauer–Thrall analogue.","keywords":["Auslander–Reiten–Serre duality","extriangulated categories","proper connective dg algebras","extended heart","almost-split conflations","Brauer–Thrall conjecture","Harada–Sai lemma","derived Nakayama functor"],"falsifier":"A direct test would be to compute, for a concrete proper connective dg algebra (for example the graded path algebra of Example 5.1), both sides of the claimed duality theorem for a small pair of objects: the left side $H^d_A(X,Y)$ and the right side $D\\,\\mathrm{Hom}_{\\mathcal{D}^{fd}(A)}(Y,\\tau X[1])$, using the Section 4 algorithm for $\\tau$. If any such pair fails to be isomorphic as predicted, Theorem 1.2 falls. More narrowly, Proposition 2.11 would be falsified by exhibiting a morphism in $\\mathrm{thick}(DA)[0,d]$ that vanishes after applying $[d-1]\\circ t_{<d}$ but does not factor through $\\mathrm{add}(DA[-d]\\oplus DA)$.","tokens_in":19866,"feed_emoji":"📐","tokens_out":8145,"duration_ms":56767,"temperature":0.7,"pith_summary":"This paper claims that the extended heart $H^d_A$ of a proper connective dg algebra $A$ behaves like a higher-dimensional module category: it has an Auslander–Reiten–Serre duality, hence almost-split conflations, and its Auslander–Reiten quiver satisfies a version of the first Brauer–Thrall conjecture. The extended heart is the extriangulated category of finite-dimensional dg modules whose cohomology lies in degrees $(-d,0]$, a category with a formal notion of short exact sequences even though it is not abelian. The central theorem constructs a $k$-linear equivalence $\\tau:H^d_A\\to H^d_A$ and a binatural isomorphism $H^d_A(X,Y)\\cong D\\,\\mathrm{Hom}_{\\mathcal{D}^{fd}(A)}(Y,\\tau X[1])$, which is precisely the duality needed to guarantee almost-split conflations and irreducible morphisms. A corollary states that $H^d_A$ is an extriangulated category with almost-split conflations, so non-projective indecomposables are targets of almost-split conflations starting at $\\tau M$ and non-injective indecomposables are sources of almost-split conflations ending at $\\tau^{-1}N$. The paper then proves a Harada–Sai lemma for the extended heart and uses it to show that any bounded connected component of the Auslander–Reiten quiver is finite, and under a connectivity condition is the whole quiver.","feed_headline":"Almost-split conflations exist in extended hearts of dg algebras","feed_subtitle":"New duality gives Auslander–Reiten translations and a Brauer–Thrall result for extended hearts.","key_machinery":"The object that carries the argument is the extended heart $H^d_A=\\mathcal{D}^{fd}(A)^{\\leq 0}\\cap\\mathcal{D}^{fd}(A)^{>-d}$, viewed as an extension-closed subcategory of the derived category, hence extriangulated. The mechanism is the pair of quotient equivalences \n$$t_{>-d}:\\frac{\\mathrm{per}(A)[-d,0]}{\\mathrm{add}(A[d]\\oplus A)}\\xrightarrow{\\sim}H^d_A,\\qquad [d-1]\\circ t_{<d}:\\frac{\\mathrm{thick}(DA)[0,d]}{\\mathrm{add}(DA[-d]\\oplus DA)}\\xrightarrow{\\sim}H^d_A,$$\n together with the derived Nakayama functor $\\nu_A=-\\otimes^{\\mathrm{L}}_A D(A)$, which restricts to an equivalence $\\mathrm{per}(A)\\to\\mathrm{thick}(DA)$ and hence links the two quotients. The Auslander–Reiten translation $\\tau$ is the composite of these equivalences, and the duality isomorphism $\\eta$ is produced from the relative Serre duality formula. A secondary mechanism is the Harada–Sai lemma proved in Section 3, which bounds the length of any non-zero chain of non-isomorphisms between indecomposables of bounded total cohomology; this bound is what forces bounded AR components to be finite.","core_discovery":"The central claim is Theorem 1.2: for a proper connective dg algebra $A$ with $d>0$ such that $A\\in H^d_A$, there is an Auslander–Reiten–Serre duality $(\\tau,\\eta)$. The construction of $\\tau$ is explicit: one moves from the heart to $(d+1)$-term perfect complexes via the truncation functor $t_{>-d}$, applies the derived Nakayama functor $[-d]\\circ\\nu_A$, and returns through the dual truncation $[d-1]\\circ t_{<d}$, so that $\\tau=[d-1]\\circ t_{<d}\\circ([-d]\\circ\\nu_A)\\circ(t_{>-d})^{-1}$. The binatural isomorphism $\\eta$ is assembled from the relative Serre duality formula $\\mathrm{Hom}_A(P,M)\\cong D\\,\\mathrm{Hom}_A(M,\\nu_A P)$ and from exact sequences induced by derived projective covers; the proof reduces to checking that the two truncation functors induce equivalences on the relevant quotient categories. The paper also states, as Theorem 3.20, the first Brauer–Thrall analogue: if $A$ is $d$-connected, any bounded connected component of the Auslander–Reiten quiver is finite and equals the whole quiver.","pith_inferences":["If the duality sketch in Proposition 2.11 is filled in, the same quotient-category mechanism should construct Auslander–Reiten–Serre duality for any interval cut out by a co-t-structure with enough projectives and injectives in a proper dg setting.","The explicit $\\tau$-algorithm of Section 4 suggests a practical method: given a $d$-term projective resolution, the Auslander–Reiten translation is obtained by applying the derived Nakayama functor and taking successive kernels; this could be implemented for graded quiver algebras with relations.","The $d$-connectedness condition is one possible analogue of 'connected representation-infinite' for extended hearts, and the Brauer–Thrall theorem may be a first step toward classifying extended hearts whose AR quiver is finite and complete.","The paper notes the Harada–Sai bound is known to be sharp for $d=1$ but not for $d>1$; a sharper bound, if it exists, would likely strengthen the finiteness conclusions."],"forward_implications":["The extended heart $H^d_A$ is an extriangulated category with almost-split conflations, so the Auslander–Reiten quiver of $H^d_A$ is defined and can be computed by knitting.","Every non-projective indecomposable $M$ is the end of an almost-split conflation $\\tau M\\to X\\to M$, and every non-injective indecomposable $N$ is the start of an almost-split conflation $N\\to Y\\to\\tau^{-1}N$.","For a $d$-connected algebra, every bounded connected component of the Auslander–Reiten quiver is finite and, in fact, is the entire quiver.","The category $\\mathrm{per}(A)[-d,0]$ is a $(d-1)$-Auslander extriangulated category, making the first equivalence a higher analogue of known results for $0$-Auslander categories.","The Harada–Sai bound $\\ell(b)=(b(d-1)+2)^b$ gives a uniform vanishing result for long chains of non-isomorphisms between indecomposables of bounded cohomology."],"supporting_citations":[{"why":"Supplies derived projective covers and the equality $\\mathrm{DProj}(A)=\\mathrm{add}(A)$, the starting point for the truncation equivalence.","marker":"[Bon24]"},{"why":"Defines Auslander–Reiten–Serre duality and almost-split conflations in extriangulated categories, and provides the theorem that converts the duality into almost-split conflations.","marker":"[INP24]"},{"why":"Gives the Ext-finiteness and extriangulated structure used to invoke the almost-split conflation criterion.","marker":"[AMY19]"},{"why":"Provides the relative Serre duality formula linking Hom to the derived Nakayama functor, and the idempotent completeness of the derived category.","marker":"[Kel94]"},{"why":"Proves the finite-dimensional algebra case that this paper generalizes, including the construction of the Auslander–Reiten translation there.","marker":"[Zho25]"},{"why":"Establishes that the derived Nakayama functor restricts to an equivalence between perfect complexes and $\\mathrm{thick}(DA)$.","marker":"[KY14]"},{"why":"Supplies the classical Harada–Sai lemma and Brauer–Thrall arguments that Section 3 adapts to the extended heart.","marker":"[ASS06]"},{"why":"Gives the Krull–Schmidt property used to conclude that almost-split conflations exist.","marker":"[Kra15]"}],"fun_headline_variants":["Almost-split conflations in extended hearts","AR duality for extended hearts of DG algebras","Brauer–Thrall analogue for extended hearts","Extended hearts admit AR duality","DG algebra hearts get almost-split conflations"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The construction of the translation rests on the claim that a certain standard duality behaves well with respect to the truncation filtrations; this claim is asserted rather than proved, and the result that relies on it (Proposition 2.11) is justified only by a sketch.","fun_headline_variants_meta":{"raw":{"variants":["Almost-split conflations in extended hearts","AR duality for extended hearts of DG algebras","Brauer–Thrall analogue for extended hearts","Extended hearts admit AR duality","DG algebra hearts get almost-split conflations"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000627,"raw_usage":{"total_tokens":2874,"prompt_tokens":891,"completion_tokens":1983,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":507,"completion_tokens_details":{"reasoning_tokens":1917}},"tokens_in":507,"tokens_out":1983,"duration_ms":13797,"temperature":1.0,"reasoning_tokens":1917,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:59:53.299068+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct test would be to compute, for a concrete proper connective dg algebra (for example the graded path algebra of Example 5.1), both sides of the claimed duality theorem for a small pair of objects: the left side $H^d_A(X,Y)$ and the right side $D\\,\\mathrm{Hom}_{\\mathcal{D}^{fd}(A)}(Y,\\tau X[1])$, using the Section 4 algorithm for $\\tau$. If any such pair fails to be isomorphic as predicted, Theorem 1.2 falls. More narrowly, Proposition 2.11 would be falsified by exhibiting a morphism in $\\mathrm{thick}(DA)[0,d]$ that vanishes after applying $[d-1]\\circ t_{<d}$ but does not factor through $\\mathrm{add}(DA[-d]\\oplus DA)$.","supporting_citations":[],"review_version":1}