{"id":"3052b0f2-ba5f-48f9-bbb0-0bf5e31f0b62","arxiv_id":"2505.10374","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper proves that the derived categories of modules over the dual numbers, in the bounded, bounded below, and strictly bounded ranges, and all derived categories of hereditary abelian categories, have strongly unique dg enhancements.","lead":"The paper proves that several derived categories associated to the ring of dual numbers, and to any hereditary abelian category, have a unique and rigid dg-enhancement structure. The relevance is for mathematicians using derived categories as geometric or representation-theoretic invariants, since it shows that those invariants do not depend on the chosen higher-categorical model.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem A(b) for the strictly bounded case rests on an in-preparation citation [6]; until that uniqueness result is public or replaced, the proof is not self-contained.","rationale":"The paper's strongest closed claim is Theorem A. The hereditary case (Theorem A(a)) is argued through Proposition 3.9 using published inputs from [4] and the classical decomposition of objects in derived categories of hereditary categories; I did not find a specific flaw there. The bounded and bounded below cases for the dual numbers (Section 4.1) are written out in reasonable detail, including the coproduct and compactness arguments. The remaining cases in Section 4.2 are the problem: the strictly bounded case relies entirely on [6], an in-preparation paper by the same authors, and the bounded above finitely generated case is sketched rather than proved. The reader's verdict of CONDITIONAL is therefore appropriate: the paper should not be rejected, because the published parts appear sound and the missing cases are plausibly fillable, but it also should not be accepted unconditionally while a central uniqueness input is unavailable. My concern matches the reader's weakest_assumption, so I see no change to the verdict. A single concrete check would be to replace [6] by a published proof or to state explicitly that Theorem A(b) is conditional on [6].","tokens_in":25465,"tokens_out":14012,"duration_ms":140476,"concrete_test":"Attempt to prove uniqueness of enhancement for D^sb(Mod(k[ε])) directly from published results, e.g., the methods of [4] and [11], without invoking [6]. If such a proof can be written out, replace the citation [6] by that argument; if it cannot, then Theorem A(b) for ?=sb remains conditional on an unpublished paper and should be annotated accordingly in the final version.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The decisive input for Theorem A(b) in the strictly bounded case is the assertion in Section 4.2: 'As for bS^sb_p, uniqueness of enhancement holds thanks to [6].' Reference [6] is listed as an in-preparation paper by the same authors, with no public proof. This is load-bearing because the strong-uniqueness strategy requires the triangulated category to be known to have a unique enhancement before one can reduce to lifting exact autoequivalences; without a proved uniqueness theorem for D^sb(Mod(k[ε])), the whole strictly bounded case has no foundation in the current text. The subsequent sentences in Section 4.2—'one can easily show' and 'one can rather directly conclude'—are sketches, not proofs, and they do not compensate for the missing uniqueness input. In particular, the claim that every exact autoequivalence of bS^sb_p restricts to bS^c_p is not demonstrated, and the coproduct/product argument for bS^-_{p,fg} is only indicated. The parts of the paper that are written out appear coherent, so this is an incompleteness concern rather than an identified mathematical error.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves Theorem A: (a) the derived categories D^?(A) have strongly unique dg enhancements for ? = b, +, −, ∅ whenever A is a hereditary abelian category, and (b) the categories D^?(Mod(k[ε])) have strongly unique enhancements for ? = sb, b, +, together with D^-(mod(k[ε])). The proof strategy is to replace the derived categories of the dual numbers by the category bS of sequences of vector spaces via an equivalence with the homotopy category of minimal complexes (Proposition 2.4), classify all indecomposable objects in the sequence category (Corollary 1.16), and then show that every exact autoequivalence is the identity after a suitable conjugation, using uniqueness of enhancements as a black box. The bounded and bounded below cases are proved in detail in Section 4.1; the strictly bounded and bounded-above finitely generated cases are treated only by sketches in Section 4.2, with one load-bearing citation to an in-preparation paper.","tokens_in":25651,"tokens_out":4652,"duration_ms":49247,"significance":"If the announced results hold, they constitute a genuine advance: strong uniqueness of enhancements is established for several 'big' derived categories, beyond the bounded coherent setting, and the paper gives a complete classification of indecomposable objects in the category of sequences of vector spaces. The classification in Section 1 and the equivalence with the homotopy category of minimal complexes in Section 2 are carefully written and appear correct. The proof for hereditary categories in Proposition 3.9 is a clean reduction to the uniqueness theorem of [5] and published lemmas from [4]. However, the strictly bounded case of Theorem A(b) currently rests on an in-preparation citation, and the remaining cases in Section 4.2 are sketched rather than proved; these are load-bearing gaps that need to be addressed before the theorem can be accepted as stated.","major_comments":[{"comment":"The proof that bS^sb_p has a strongly unique enhancement is not self-contained: the text says 'uniqueness of enhancement holds thanks to [6]', but [6] is listed as an in-preparation paper by the same authors and no proof or preprint is available. This is load-bearing, because the whole strategy reduces strong uniqueness to the existence of dg lifts of exact autoequivalences only after uniqueness of enhancements is known. The manuscript should either include a proof of the uniqueness statement for D^sb(Mod(k[ε])), cite a publicly available reference, or explicitly mark that part of Theorem A(b) as conditional.","section":"§4.2, first paragraph"},{"comment":"The assertion that 'one can easily show that every exact autoequivalence of bS^sb_p restricts to an exact autoequivalence of bS^c_p' is not demonstrated. This restriction is not automatic from the classification alone: it requires knowing that bS^c_p is the subcategory of compact objects in bS^sb_p, or an equivalent argument. The paper never proves compactness of the finite-interval objects in bS^sb_p, so this step needs a written proof.","section":"§4.2, first paragraph"},{"comment":"The treatment of bS^-_{p,fg} ends with 'one can rather directly conclude using the universal properties of coproduct and product'. This is a sketch, not a proof. In particular, the text does not show in detail how the isomorphisms used to conjugate an autoequivalence to the identity on bS^b_{p,fg} are compatible with the fact that a coproduct in bS^-_{p,fg} is also a product, nor does it verify that every morphism between arbitrary objects of bS^-_{p,fg} is determined by its restrictions to the finite-type summands. These details are necessary for the claimed strong uniqueness.","section":"§4.2, second paragraph"},{"comment":"The reduction uses the statement that, for a hereditary abelian category A, the inclusion of B^?(A) in D^?(A) is an equivalence for ? = b, +, −, ∅. This is plausible and standard in the bounded and bounded-above cases, but the paper cites [21, Section 1.6] without commenting on whether the cited statement covers the unbounded case for arbitrary hereditary abelian categories. Please clarify the scope of the cited equivalence or give a direct argument for ? = ∅.","section":"§3.3, Proposition 3.9"}],"minor_comments":[{"comment":"The sentence 'In this paper is we are unable to settle...' contains a typo: 'is' should be deleted.","section":"Introduction, page 4"},{"comment":"The abstract emphasizes the bounded and bounded below cases for Mod(k[ε]), but Theorem A(b) also includes the strictly bounded case and D^-(mod(k[ε])); the abstract should reflect the full statement or explicitly say 'including'.","section":"Abstract and Introduction"},{"comment":"The parenthetical remark that S^d_{Z,{∞}} is closed under direct summands is stated without proof and with the comment that it is not needed; this is acceptable, but it would be helpful to give a reference or a one-sentence indication.","section":"§1.4, Remark 1.13"},{"comment":"The phrase 'D^?, for ? = sb,b,+,−,∅, and D^-_{fg}' is slightly confusing because D^? has also been used for D^?(A) earlier; a short reminder of the convention for the dual numbers would improve readability.","section":"§4, opening paragraph"},{"comment":"The reference [5] is cited as an arXiv preprint; since Theorem 3.6 is used as a black box throughout, it would be useful to state explicitly whether [5] has appeared in a refereed venue or is also still under review.","section":"§3.2, Theorem 3.6"}],"recommendation":"major_revision","confidential_remarks":"The paper is well within the scope of the journal and the main classification work is of independent interest. The main concern for the editor is the use of reference [6], an in-preparation paper by the same authors, at a load-bearing point of Theorem A(b); this should be resolved before publication. The remaining sketched arguments in Section 4.2 are likely fillable, but they are not mere presentation issues."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new content: strong uniqueness for the bounded and bounded below derived categories of all modules over k[ε], and uniformly for all four boundedness flavors in the hereditary case. The paper is transparent that the bounded hereditary case and the small categories were already known; the new results are the big categories and the uniform hereditary argument.\n\nThe sequence-category machinery in Sections 1–2 is the real workhorse. The classification of indecomposables is careful, and the comparison between Km and bS is explicit enough to be checked. The proof for ?=b,+, Section 4.1, is written out and looks sound: once you know every object is a coproduct of compact objects in the small subcategory, a trivial-on-the-compacts autoequivalence is forced to be trivial everywhere. That is a clean argument.\n\nProposition 3.9 (hereditary case) is short but not a black box: it reduces to published lemmas in [4] and uses the separately established uniqueness theorem [5]. Fine.\n\nThe soft spot is Section 4.2, and it is exactly where the reader and stress-test point. The strictly bounded case bS^sb_p begins with 'uniqueness of enhancement holds thanks to [6]', where [6] is an in-preparation paper by the same authors. This is not a stylistic complaint: Theorem 3.6 does not cover D^sb(Mod(k[ε])), so the strong-uniqueness reduction has no foundation in the current text without [6]. The rest of Section 4.2 is an acknowledged sketch: 'one can easily show' that autoequivalences restrict to bS^c_p, and the coproduct/product argument for bS^-_{p,fg} is only indicated. I see no hint of a real mathematical error in the sketched steps, and the parts that are written out are coherent. But two subcases of Theorem A(b) are not proved in this version.\n\nRecommendation: this deserves a serious referee rather than a desk reject. The ?=b,+ results and the hereditary theorem are solid and publishable. Before acceptance, the authors should either include the proof of uniqueness for D^sb or cite a public version of [6], and expand Section 4.2 into full proofs. If I were refereeing, I would ask for major revision with a clear path to verifying those two cases. I would cite this for the ?=b,+ results and the sequence technique. Maybe bring it to reading group after the gaps are patched.","headline":"New strong-uniqueness results for big derived categories over the dual numbers and a uniform hereditary argument, but two subcases of Theorem A(b) rest on an in-preparation citation and sketched proofs.","tokens_in":26235,"tokens_out":4986,"would_cite":true,"duration_ms":44849,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["18G80","14F08","18N40","18N60"],"pacs":[],"model":"deepseek-v4-flash","headline":"Derived categories over the dual numbers have strongly unique dg enhancements in the bounded, strictly bounded, and bounded-below cases, as do all derived categories of hereditary abelian categories.","keywords":["triangulated categories","dg enhancements","strong uniqueness","dual numbers","hereditary categories","category of sequences","indecomposable objects","derived categories"],"falsifier":"Compute the autoequivalence group of the sequence category $bS^+_p$ that models $D^+(\\operatorname{Mod}(k[\\varepsilon]))$; the paper's Section 4.1 proves it is trivial up to natural isomorphism. A concrete falsifier would be an exact autoequivalence that fixes every finite interval object $S^\\bullet_{a,b}$ but acts nontrivially on an infinite coproduct of intervals; exhibiting one would refute the main theorem, and the paper predicts none exists.","tokens_in":25227,"feed_emoji":"🔺","tokens_out":16939,"duration_ms":153946,"temperature":0.7,"pith_summary":"This paper establishes that the triangulated categories $D^b(\\operatorname{Mod}(k[\\varepsilon]))$, $D^{+}(\\operatorname{Mod}(k[\\varepsilon]))$, and $D^{\\mathrm{sb}}(\\operatorname{Mod}(k[\\varepsilon]))$, together with $D^{-}(\\mathrm{mod}(k[\\varepsilon]))$, have strongly unique dg enhancements: any two pretriangulated dg models of one of these categories are connected by a dg isomorphism whose induced equivalence is compatible with the identity functor. The same framework proves that every derived category $D^{?}(A)$, for $?\\in\\{b,+,-,\\emptyset\\}$, of a hereditary abelian category $A$ has a strongly unique enhancement. The proof works by replacing the derived category over the dual numbers with an equivalent category of sequences of vector spaces, where all indecomposable objects can be classified. This gives a linear-algebra model in which the question of dg liftability of exact autoequivalences can be settled by direct coproduct arguments, without any ample-set or Fourier--Mukai machinery.","feed_headline":"Derived categories over dual numbers have strongly unique enhancements","feed_subtitle":"The proof classifies all indecomposable objects and shows every exact self-equivalence lifts to a dg functor.","key_machinery":"The load-bearing object is the category $S(A)$ of $\\mathbb{Z}$-indexed sequences $X^\\bullet=(X_i,d_i:X_i\\to X_{i+1})$ in an abelian category, together with its variant $bS(A)$ whose morphism spaces add a second, 'type $\\varepsilon$' component; for $A=\\operatorname{Mod}(k)$, the homotopy category of minimal complexes over $k[\\varepsilon]$ is equivalent to $bS$, and the derived category $D(\\operatorname{Mod}(k[\\varepsilon]))$ is equivalent to the right half $bS_p$ of a semiorthogonal decomposition $bS=\\langle bS_a,bS_p\\rangle$. The classification result that does the work (Corollary 1.16) states that the indecomposable objects of $S$ are exactly the interval objects $S^\\bullet_{a,b}$ with $a\\in\\mathbb{Z}\\cup\\{-\\infty\\}$, $b\\in\\mathbb{Z}\\cup\\{\\infty\\}$, $a\\le b$, and that in the cases covered every object is a coproduct of such intervals. This complete decomposability converts the abstract question of whether every exact autoequivalence has a dg lift into checking that a functor fixing the finite interval objects fixes all morphisms.","core_discovery":"On the paper's own terms, the central result is Theorem A. For every hereditary abelian category $A$ and every $?\\in\\{b,+,-,\\emptyset\\}$, the derived category $D^{?}(A)$ has a strongly unique enhancement, meaning that any two dg enhancements are isomorphic in a way that respects the given equivalence to $D^{?}(A)$. For the dual numbers $k[\\varepsilon]$ over a field, the categories $D^{\\mathrm{sb}}(\\operatorname{Mod}(k[\\varepsilon]))$, $D^b(\\operatorname{Mod}(k[\\varepsilon]))$, $D^+(\\operatorname{Mod}(k[\\varepsilon]))$, and $D^{-}(\\mathrm{mod}(k[\\varepsilon]))$ all have strongly unique enhancements, while the unbounded and bounded-above cases for all modules are left open. The proof's core is the assertion that, in the categories it handles, every object is completely decomposable into explicitly classified indecomposable sequence objects, which reduces every exact autoequivalence to one that is the identity on the finite objects and then, by a coproduct argument, on everything.","pith_inferences":["The sequence-category method could plausibly be adapted to other finite-dimensional algebras, such as $k[x]/(x^n)$ or other self-injective algebras; the paper does not attempt this.","For the two open cases $D^\\emptyset(\\operatorname{Mod}(k[\\varepsilon]))$ and $D^{-}(\\operatorname{Mod}(k[\\varepsilon]))$, the paper reduces the obstruction to a purely linear-algebraic question about derivations on the category of sequences, so the problem can be investigated independently of dg categories.","The classification of indecomposable sequence objects may be independently useful for studying phantom morphisms, since the paper shows that in the unbounded case every phantom morphism is of type $\\varepsilon$ and factors through the interval objects.","If the announced uniqueness theorem cited as $[6]$ fails, the strictly bounded case might still be recoverable from the rest of the paper's machinery, because the complete-decomposability classification and the coproduct argument would remain intact."],"forward_implications":["Every exact autoequivalence of $D^b(\\operatorname{Mod}(k[\\varepsilon]))$ and of $D^+(\\operatorname{Mod}(k[\\varepsilon]))$ is isomorphic to the identity and therefore admits a dg lift, so any two dg enhancements of these categories are isomorphic over the identity.","The derived categories of every hereditary abelian category, including quasi-coherent and coherent sheaves on smooth curves over a field, now have strongly unique enhancements uniformly for $?\\in\\{b,+,-,\\emptyset\\}$.","Strong uniqueness is no longer confined to bounded derived categories or to settings with an ample set of objects; the new cases include the bounded-below category $D^+$ and the finitely generated bounded-above case $D^{-}(\\mathrm{mod}(k[\\varepsilon]))$.","For each covered category, the proof supplies a concrete route to showing that every exact autoequivalence fixes the finite objects and then extends by coproducts, so the same template can be applied to other categories with a complete-decomposability classification."],"supporting_citations":[{"why":"Supplies the general uniqueness theorem for derived and geometric categories (Theorem 3.6) that reduces strong uniqueness to proving every exact autoequivalence has a dg lift.","marker":"[5]"},{"why":"Provides the technical lemmas used in Proposition 3.9 to choose the isomorphism between enhancements so that it respects the canonical objects with zero differential.","marker":"[4]"},{"why":"Theorem 7.1 gives the dg lift for exact autoequivalences of the finite/perfect subcategory, the key input in Section 4.","marker":"[12]"},{"why":"Announced work by the same authors that supplies uniqueness of enhancement for the strictly bounded category, on which the proof of the $D^{\\mathrm{sb}}$ case depends.","marker":"[6]"},{"why":"Supplies the standard fact that for hereditary categories the subcategory of objects with zero differential is equivalent to the whole derived category, used to conclude strong uniqueness.","marker":"[21]"}],"fun_headline_variants":["Strong uniqueness proven for dual numbers derived categories","Hereditary derived categories all have strongly unique enhancements","Classifying indecomposables proves enhancement uniqueness","Dual numbers: bounded derived categories have unique enhancements","Theorem A: every hereditary derived category has strong uniqueness"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a uniqueness theorem, cited to the authors' own paper still 'in preparation', really holds for the strictly bounded derived category over the dual numbers; if it fails, the proof of that case collapses.","fun_headline_variants_meta":{"raw":{"variants":["Strong uniqueness proven for dual numbers derived categories","Hereditary derived categories all have strongly unique enhancements","Classifying indecomposables proves enhancement uniqueness","Dual numbers: bounded derived categories have unique enhancements","Theorem A: every hereditary derived category has strong uniqueness"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000271,"raw_usage":{"total_tokens":1565,"prompt_tokens":817,"completion_tokens":748,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":433,"completion_tokens_details":{"reasoning_tokens":675}},"tokens_in":433,"tokens_out":748,"duration_ms":7035,"temperature":1.0,"reasoning_tokens":675,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:10:52.038669+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the autoequivalence group of the sequence category $bS^+_p$ that models $D^+(\\operatorname{Mod}(k[\\varepsilon]))$; the paper's Section 4.1 proves it is trivial up to natural isomorphism. A concrete falsifier would be an exact autoequivalence that fixes every finite interval object $S^\\bullet_{a,b}$ but acts nontrivially on an infinite coproduct of intervals; exhibiting one would refute the main theorem, and the paper predicts none exists.","supporting_citations":[{"cited_title":"Canonaco, A","cited_arxiv_id":null,"evidence_quote":"Provides the technical lemmas used in Proposition 3.9 to choose the isomorphism between enhancements so that it respects the canonical objects with zero differential."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Theorem 7.1 gives the dg lift for exact autoequivalences of the finite/perfect subcategory, the key input in Section 4."},{"cited_title":"Canonaco, A","cited_arxiv_id":null,"evidence_quote":"Announced work by the same authors that supplies uniqueness of enhancement for the strictly bounded category, on which the proof of the $D^{\\mathrm{sb}}$ case depends."},{"cited_title":"Krause, Derived categories, resolutions, and Brown representability , in: Interactions between homotopy theory and algebra, 101–139, Contemp","cited_arxiv_id":null,"evidence_quote":"Supplies the standard fact that for hereditary categories the subcategory of objects with zero differential is equivalent to the whole derived category, used to conclude strong uniqueness."}],"review_version":1}